I. PERSISTENT RADICALS FOR DYNAMIC NUCLEAR POLARIZATION
II. SYNTHESIS OF SUBSTITUTED INDOLES
by
Olesya Haze
B.S. cum laude, with High Distinction in Chemistry;
B.A. cum laude, with Distinction in Mathematics
University of Rochester, NY, 2006
SUBMITTED TO THE DEPARTMENT OF CHEMISTRY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN CHEMISTRY
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
SEPTEMBER 2015
© Massachusetts Institute of Technology, 2015. All rights reserved.
Signature of Author ............................................................................................................
Department of Chemistry
August 17, 2015
Certified by .........................................................................................................................
Timothy M. Swager
John D. MacArthur Professor of Chemistry
Thesis Advisor
Accepted by .......................................................................................................................
Robert W. Field
Haslam and Dewey Professor of Chemistry
Chairman, Departmental Committee of Graduate Studies
-1 -
This doctoral thesis has been examined by the following committee of the
Department of Chemistry:
Professor Stephen L. Buchwald: .......................................................................................
Chairman
Professor Timothy M. Swager: ...........................................................................................
Thesis Advisor
Professor Timothy F. Jamison: ...........................................................................................
Department of Chemistry
-2 -
In loving memory of D-dog
-3 -
Acknowledgements
I am greatly indebted to Tim Swager for being the ever patient, supportive, and
understanding advisor during my time at MIT. He was generous with his time, and
allowed for complete scientific freedom while still maintaining the reasonable safety net
of subgroup meetings. I could never fully express my gratitude to Tim.
I thank my thesis committee members, Steve Buchwald and Tim Jamison for their
support and helpful suggestions with regard to this thesis. In particular, Steve always
reminded me to keep my focus on what’s really important in life. In class, in our annual
meetings, and via sideline observations I’ve learned so much from you! You gave me
the courage to do things “my way”, and for that I thank you. Cross-couplings being a rite
of passage for an organic chemist, one of Steve’s ligands (of course!) solved a synthetic
problem I was having, improving my lab life.
Rick Danheiser’s lab was my home for the first couple of years at MIT. Rick taught me
so much about how to write and how to organize my thoughts, what retrosynthetic
analysis was all about, and the importance of running a neat lab. Thank you for your
guidance and support.
The many wonderful yearmates, labmates, officemates, and collaborators - the amazing
minds of MIT you made my time here the time of largest intellectual and personal
growth. I chose to come to MIT mostly because of the people, and you did not
disappoint! My friend and roommate of two years, Ricky (Weerawat) Runguphan, thank
you for the wonderful stories of chemistry/plant biology interface and for the introduction
to the secrets and tastes of homemade Thai cuisine. Danheiser group labmates Takeo
Sakai, Chung-Yang (Dennis) Huang, Jun Chul (JC) Choi, thank you for the insightful
chemistry (and glassware!!!) discussions, sock-court badminton, and an introduction to
authentic sushi. I am grateful to Tom Willumstad for his helpful collaboration on the
indole paper, and to Tammy Lam for helping me get started in lab.
Great many thanks to my best friend Jan Schnorr. Jan one of the most conscientious
people I know. You are always ready to help, to listen, and you have the most
reasonable, almost german, way of looking at problems. It’s equally awesome to share
happy and frustrating moments with you. I am watching you, Dr CEO.
-4 -
Jose Miguel Lobez Gomeras, you amaze me. You accomplish in one week what four
very skilled and hardworking people would accomplish in.. one week. Wait a minute,
now I get the name thing! Sure, getting a PhD in chemistry is not enough of challenge,
learn Japanese while you are at it! With all your fantastic work ethic, studiously honed
natural talent, fashion sense, and ability to maintain stable long term relationships you
still never made me feel inadequate. That is the greatest talent of all. I really hope Apple
continues to do well, so that we can have our little love/hate exchanges every time a
new iDevice is announced (and when it ends up in your pocketses on release day). It
was a wonderful learning experience, an inspiration, and just a fun ride to work side by
side. Thank you for the advice, support, and for believing in me. No! You’re a towel!
Rebecca Parkhurst, my trusty Zumba buddy, WIC teammate, and year/labmate friend.
Thank you for welcoming me into the US citizen ranks. I still have the card you put
together then. Just to put it in perspective, I did not keep the “letter from the president”
that came in the naturalization packet! (Granted, it was from G.W. Bush.) Thank you for
helping make my transition to the Swager lab seamless, you have a fine sense of
people’s comfort zones and an exquisite ability to make people feel welcome.
Tim may be hands off, but he has a knack for selecting just the right mix of people to
join the group, and this made the TMS lab a fabulous place to work and grow. I’d like to
thank the rest of the Swager group for being fantastic co-workers. I’m grateful to Eric
Dane for introducing me to the wonder-world of DNP. The DNPeople: Yi Wei, Matt
Kiesewetter, Derik Frantz, and Joe Walish thank you for being such a great team to
bounce about radical ideas with. Joe Walish, thank you also for introducing me to the
entrepreneurial world and experience of grant writing with DYNUPOL. My office mates
Kelvin Frazier and Fumi Ishiwari, were always ready to share the latest exciting
chemistry developments or frustrations. Grace Han, the little ray of sunshine in our
basement office, thank you for being such a wonderful friend. Stef Sydlik, all the way
from ISN your confidence and perseverance have inspired me, thank you! Shuang Liu
thank you for the insightful discussions and for being a wonderful lab citizen. Jon Weis,
it was a privilege to work in the same lab with you. Balta Bonillo you were there during
the late shifts which were usually either disappointing or extremely exciting. You did not
laugh at me when I ran to you ecstatic that I finally got a biradical to stay red. It turned
-5 -
out to be a dud when it came to DNP, but I promised it would be yours, BaltaRad it is.
Sorry. And thank you for your share of the excitement.
I thank Jennifer Weisman, for helping me navigate the logistics of being a non-resident
thesis status student, and Kathy Sweeney, Caitlin McDowel for making sure the lab ran
smoothly and everything was in its place in its time.
A great big thank you also goes out to my collaborators in Griffin group. I
thank Bob Griffin for the opportunity to see my compounds in action at the cutting edge
of solid state NMR. Björn Corzilius, Andy Smith, Jennifer Mathies, and Vlad Michaelis,
thank you for your hard work and innovative ideas and the EPR and DNP studies. Björn
deserves a special thank you for expediently, coherently, and patiently explaining the
basics, and not so basics, of spectroscopy and DNP to me and proofreading relevant
chapters of this thesis. I wish you utmost success in your academic career.
I thank John Kurhanewicz at UCSF for welcoming me to his lab for a couple of days to
try out experiments, and Mark Van Criekinge, for helping with the dissolution DNP
polarizer and “letting me drive”. It was quite an experience to see the polarization build
up late into the night with Mark almost as pumped as I was. Good times!
I’d like to thank my sister Diana Makhaldiani and her partner in crime Rich Lalor for
helping me settle in Cambridge, apartment hunting for me, and taking the starving
graduate student out for a “random dinner” just at the right time. Dinka, thank you for
having my back, always.
Finally, I don’t think I could finish writing this thesis without the support and motivation of
my three boys. Egi - my redheaded step child - the happiest, cuddliest Vizsla of all. You
are such a loving and forgiving soul, thank you for keeping your soft eyes on me at all
times, and for warming my feet even as I type this. I look forward to more birding
adventures with you. Tim - the most amazing child a mother could ask for. Your focus
and curiosity are fascinating, and your gentle heart makes mine burst with pride. My
wonderful husband - Joe - thank you for believing in me when even my own faith had
failed. I would not be the person I am today without your unwavering support and
encouragement. -6 -
I. PERSISTENT RADICALS FOR DYNAMIC NUCLEAR POLARIZATION
II. SYNTHESIS OF SUBSTITUTED INDOLES
by
Olesya Haze
Submitted to the Department of Chemistry on August 17th, 2015 in Partial Fulfillment of
the Requirements for the Degree of Doctor of Philosophy in Chemistry
Abstract:
Part I: An brief introduction to dynamic nuclear polarization (DNP) is provided including
a discussion of polarization mechanisms and development of new polarization methods
in a historical context. Efficient synthesis of highly water-soluble BDPA derivatives that
preserve the desirable DNP properties of BDPA and expand its application to aqueous
systems is described. The narrow line radical applications in magnetic resonance
spectroscopic imaging (MRSI) are investigated focusing on thermal mixing (TM) DNP of
pyruvic acid. Design considerations and efforts toward the synthesis of bi- and
multiradicals for Cross Effect (CE) DNP, and experiments performed with mixtures of
radicals and covalently bound hetero-biradicals are reported.
Part II: Indoles are important heterocycles because of their presence as common
structural motifs in natural products and pharmaceutical candidates. Second generation
Danheiser benzannulation–tandem cyclization approach was developed and applied
toward the synthesis of highly substituted indoles. Substrate synthesis, benzannulation
results and product elaboration are described.
Thesis Supervisor: Timothy M. Swager
Title: John D. MacArthur Professor of Chemistry
-7 -
Table of Contents
Part I: Persistent Radicals for
Dynamic Nuclear Polarization .................................................................................10
Chapter 1 Dynamic Nuclear Polarization ..................................................................10
1.1 Introduction to DNP .................................................................................................11
1.2 Polarization transfer mechanisms ...........................................................................15
1.3 Major advancements in DNP technology ................................................................20
1.4 Contemporary DNP methods ..................................................................................22
Chapter 2 Water-Soluble Narrow-Line Radicals for DNP-Enhanced Solid State
NMR ...............................................................................................................................27
2.1 Introduction: DNP and Solid State NMR .................................................................28
2.2 Polarization transfer mechanisms: Solid Effect ......................................................29
2.3 DNP with narrow-line radicals ................................................................................31
2.4 Synthesis and properties of water-soluble narrow-line BDPA-based radicals ........35
2.5 DNP MAS SSNMR results with SA–BDPA .............................................................44
2.6 Synthesis and properties sulfonated BDPA derivatives ..........................................51
2.7 Summary ................................................................................................................55
2.8 Experimental ...........................................................................................................56
Chapter 3 Narrow-Line Radicals for Hyperpolarized MRS Imaging .......................72
3.1 Introduction: MRI and MRSI ...................................................................................73
3.2 Improving the SNR in MRSI with Dissolution DNP .................................................75
3.3 Polarization transfer mechanisms: Thermal Mixing ................................................77
3.4 Dissolution DNP with narrow line radicals and its applications to MRSI .................80
3.5 Hyperpolarization of Pyruvate with SA–BDPA and SAH-BDPA results...................84
3.6 Summary ................................................................................................................91
3.7 Experimental ...........................................................................................................91
-8 -
Chapter 4 Radical Mixtures, Biradicals, and Multiradicals for Cross Effect DNP ..........................................................................................................92
4.1 Polarization Transfer Mechanisms: Cross Effect ....................................................93
4.2 Broad radicals as CE polarization agents ...............................................................93
4.3 Broad biradicals as CE polarization agents ............................................................95
4.4 New nitroxide triradicals for CE and TM DNP .........................................................95
4.5 Hetero-biradicals ..................................................................................................101
4.6 The ideal case for CE: two narrow line radicals ....................................................109
4.7 Summary ...............................................................................................................111
4.8 Experimental .........................................................................................................112
Part II: Synthesis of Highly Substituted Indoles ..................................................118
II.1 Introduction .............................................................................................................119
II.2 Danheiser benzannulation ......................................................................................120
II.3 Second generation DBAN - tandem cyclization approach ......................................123
II.4 Substrate synthesis ................................................................................................125
II.5 D-BAN results .........................................................................................................129
II.6 Product elaboration .................................................................................................135
II.7 Summary ...............................................................................................................138
II.8 Experimental ...........................................................................................................139
-9 -
Part I: Persistent Radicals for
Dynamic Nuclear Polarization
Chapter 1
Dynamic Nuclear Polarization
e↑N↑
e↑N↓
NMR
ZQ
EPR
EPR
DQ
e ↓ N↑
NMR
e↓N↓
+ small state mixing at
high fields
-10-
1.1 Introduction to DNP
Dynamic Nuclear Polarization (DNP)1 is a method of increasing signal to noise ratio2 in
NMR and in imaging techniques that rely on NMR, namely MRI and magnetic
resonance spectroscopic imaging (MRSI). This chapter provides a brief introduction to
the method and some historical milestones. Literature reviews listed here are an
expedient way for an interested reader to become familiar with the field and current
developments.
NMR signal arises when nuclear spins are placed in an external magnetic
field (B0) and the degenerate (2I+1, where I is the nuclear spin) nuclear energy levels
split due to Zeeman effect. The absorption of RF radiation matching the energy of the
splitting produces the NMR signal. The same thing happens when an electron (spin 1/2)
is placed in a magnetic field, producing EPR signal (Figure 1)
Figure 1. Energy level splitting of an electron or a spin 1/2 nucleus in an external
magnetic field B0 due to Zeeman effect. (Because electron gyromagnetic ratio is
negative, ms =–1/2 is the lower energy state, for a nucleus with positive gyromagnetic
number , e.g. 1H, the energy of parallel alignment I =1/2 level is lower.)
1
For a quick introduction to DNP written in a language accessible to an average organic chemist see
Atsarkin, V. A. J. Phys.: Conf. Ser. 2011, 324, 012003. Three DNP themed issues in AMR and PCCP are
a good starting point for anyone interested in learning more about DNP: Appl. Magn. Reson. 2008, 34,
213–544. Phys. Chem. Chem. Phys., 2010, 12, 5737–5920. Appl. Magn. Reson. 2012, 43, 1613–7507.
2 Taking into account receiver coil response and thermal noise, NMR/MRI SNR scales as ~(ɣ )3(B )2
n
0
Hoult, D. I.; Richards, R. E. J. Magn. Reson. 1976, 24, 71.
-11-
Thermal equilibrium polarization (P), which is the difference between the populations of
the two Zeeman energy levels (n+ – n-) divided by the total number of spins (n+ + n-)
determines the magnitude of the NMR signal (or how many photons of energy ΔE =
ɣħB0 can be absorbed). Polarization in a system of I = 1/2 nuclear spins with
gyromagnetic ratio ɣ, in magnetic field B0, at temperature T is given by equation 1.
(1)
The definition of the gyromagnetic ration is in SI units of angular frequency (Rad•T-1•s-1).
Confusingly, ɣ is often given in units of frequency (MHz/T), which would correctly
describe ɣ/2π. E.g., “ɣC = 10.705 MHz/T” actually is “ɣC/(2π) = 10.705 MHz/T.” When
plotting the polarization equation (Figure 2), the Plank’s constant should then be used
instead of its reduced form.
99.999 %
100
11.43 %
10
Spin Polarization [%]
3.14 %
1
0.26 %
0.1
0.017 %
electron
1H
13C
15N
B0 = 14T (600 MHz 1H NMR)
0.01
0.001
1
0.0048 %
0.0012 %
10
100
RT
Temperature [K]
Figure 2. Nuclear and electron polarization as a function of temperature at 14 T.
-12-
From Figure 2 we see that at room temperature proton and carbon polarization is
exceedingly small (0.0048% and 0.0012% respectively) even at relatively high field (14
T which corresponds to 600 MHz NMR). From equation 1 and Figure 2 several ways of
increasing polarization are apparent. Brute force3 approach is to conduct NMR
experiments at high field and low temperature. An increase in field strength results in an
increase in polarization. However, building magnets stronger than 20–30 T for routine
NMR is currently cost-prohibitive, and also technically challenging, so this is not a
feasible way to improve NMR. A significant polarization increase occurs at low
temperature and this has been harnessed in “temperature jump” experiments.4 Last but
not least, compared to nuclei, electrons have a much higher polarization at a given field
and temperature. This is due to their relatively high gyromagnetic ratio (mostly because
of their higher charge to mass ratio) compared to, e.g., 1H or 13C. Electron5 and nuclear
gyromagnetic ratios are given by equations 2 and 3.
(2)
(3)
DNP is the technique that transfers this high electron spin polarization to the nuclei. The
maximum enhancement of nuclear polarization provided by DNP6 is then given by
equation 4.
(4)
3
Roberts, L. D.; Dabbs,J. W. T. Annu. Rev. Nucl. Sci. 1961, 11, 175.
Abragam, A.; Proctor, W. G. Phys. Rev. 1958. Joo, C.-G.; Hu, K.-N.; Bryant, J. A.; Griffin, R. G. J. Am.
Chem. Soc. 2006, 128, 9428.
5 NIST value for the electron is given rather than the /2π
6 Equation 4 holds for T≥2 K. At lower temperature electron polarization is almost unity while nuclear
polarization is still increasing, leading to overall lower theoretical maximum DNP enhancements. See
Figure 2.
4
-13-
Examples of maximum polarization that can be obtained by transferring electron
polarization to the common nuclei used for DNP are in table below:
Nucleus
ɣn/2π
εDNP
1H
42.576 MHz/T
658
13C
10.705 MHz/T
2618
15N
-4.316 MHz/T
6493
89Y
2.0864 MHz/T
13432
Electron and nuclear spins communicate through hyperfine interactions (scalar
coupling), and electron-nuclear dipolar interactions. Scalar relaxation is the return of a
spin system to equilibrium via time-dependent fluctuations in indirect dipole-dipole
interactions or “through chemical bonds”. Scalar coupling or Fermi contact interaction
(also known as J-coupling in NMR and isotropic hyperfine coupling in EPR) is
modulated by rapid chemical exchange or by rapid relaxation of one of the spins. Scalar
coupling leads only to zero quantum (flip-flop) transitions. Dipolar coupling (direct
magnetic dipole-dipole interaction) is affected (modulated) by rotational and
translational diffusion (tumbling of molecules in solution, methyl group rotations, etc.).
Modulation (or fluctuation in intensity) of a particular dipole-dipole interaction gives rise
to the fluctuating local magnetic field “seen” by a spin. These fluctuations drive the
relaxation of the spins toward equilibrium state. At equilibrium the energy level
populations are determined by Boltzmann distribution and all coherences such as
transverse magnetization have decayed. The process of changing the energy level
populations requires a spin transition or a flip, e.g. spin up to spin down. The energy
released (or absorbed) during this transition has to go (or come from) somewhere. For
example, nuclear or electron spin-lattice relaxation (characterized by T1) occurs when
the energy of the spin transition is matched by a lattice motion (translations rotations
and internal motions of molecules). Microwaves can drive a spin transition, and mutual
“flip-flop” transitions between electrons and nuclei are also possible. Different DNP
mechanisms harness the electron-nuclear dipole couplings to transfer electron
polarization to nuclei. We’ll discuss DNP mechanisms in the order they were
discovered.
-14-
1.2 Polarization transfer mechanisms
Overhauser7 first predicted the possibility of transferring polarization between spin
systems by saturating the EPR line with resonant microwaves in 1953. The theory was
then put in practice by Carver and Slichter.8 Their paper announcing the observation of
DNP in 7Li metal, although it submitted two months after, was published ahead of the
theoretical paper by Overhauser, indicative of the skepticism toward DNP present
among the magnetic resonance community at the time. Carver and Slichter next
demonstrated that DNP was not limited to metals by achieving OE enhancement of 1H
in liquid ammonia9 via scalar coupling (granted, the experiment was not completely
“metal-free” as the source of electrons was sodium dissolved in ammonia).
Overhauser effect (OE)10 is a two spin (one electron and one nuclear spin) process.
OE stems from both the scalar and dipolar cross relaxation between
electrons and nuclei. It was originally thought that OE is not effective in
systems lacking mobile electrons, which limited its application to metals
and liquid solutions. Also at fields higher than 1 T the nuclear longitudinal
(spin-lattice) relaxation rate is much higher than the electron-nuclear
dipolar relaxation and the OE enhancement due to dipolar cross relaxation
approaches zero. Thus OE was thought to be ineffective at high fields.
However, Griffin demonstrated11 that at high fields (5T) the scalar
component of the relaxation can produce large signal enhancements in
liquid samples. In insulating solids OE was not expected to produce any
enhancement as a result of the lack of mobile electrons and rapid
molecular tumbling. However, OE has been observed in carbazole – an
insulating solid – at 1.4 T.12 Recently, OE in polystyrene has been studied
7
Overhauser, A. W. Phys. Rev. 1953, 92, 411.
Carver, T. P.; Slichter, C. P. Phys. Rev. 1953, 92, 212. For for a wonderful first hand historical account of
discovery of DNP see Slichter, C. P. Rep. Prog. Phys. 2014, 77, 072501.
9 Carver, T. P.; Slichter, C.P. Phys. Rev. 1956, 102, 975.
10 Solomon, I. Phys. Rev. 1955, 99, 559
11 Loening, N. M.; Rosay, M.; Weis, V.; Griffin, R. G. J. Am. Chem. Soc. 2002, 124, 8808.
12 Hu, J. Z.; Solum, M. S.; Wind, R. A.; Nilsson, B. L.; Peterson, M. A.; Pugmire, R. J.; Grant, D. M. J.
Phys. Chem. A 2000, 104, 4413.
8
-15-
at high fields (9.4–18.8 T), and was found to increase linearly with field.7
The effect is thought to originate from fluctuating hyperfine interactions - a
property completely determined by the polarization agent supplying the
unpaired electrons. Overhauser Effect relies on irradiation of the allowed
EPR transitions and can be easily saturated at low microwave power. The
μw radiation is applied to the allowed single quantum (SQ) electron
transitions and the enhancement is generated by the difference in the rate
of zero quantum (ZQ) and double quantum (DQ) cross relaxation. OE is
the only DNP mechanism that can be applied effectively in liquids.13
Jeffries14 and his students Abraham and Kedzie proposed saturating the nominally
forbidden zero and double quantum EPR transitions to induce electron nuclear
transitions and demonstrated the effect in 60Co. Later, Uebersfeld15, and Abragam and
Proctor16 independently applied the same approach to polarize LiF, and named it the
solid effect.
solid effect (SE)17 is also a two spin process where the nuclear spin is dipolar coupled
to one electron spin. In the SE the μw irradiation is applied to the
nominally forbidden transitions (either a flip-flop ZQ or flip-flip DQ
transition) which become partially allowed as a result of the protonelectron dipolar interactions. Because the probability of the ZQ and DQ
transitions is low, high microwave power is required for efficient SE. SE is
operative when the paramagnetic centers are localized (immobile), and
are at low enough concentration that electrons are not coupled to each
13
Some methods (e.g., TJ and Dissolution DNP described below) enable liquid state NMR and MRI
measurements using other DNP mechanisms, however the polarization is carried out in solids state. For
an early review of applications of OE to liquids see Hausser, K. H.; Stehlik, D. Adv. Magn. Reson. 1968, 3,
79.
14 Jeffries, C. D. Phys. Rev. 1957, 106, 164; Abraham, M.; Kedzie, R. W.; Jeffries, C. D. Phys. Rev. 1957,
106, 165.
15 Erb E.: Motchane J.L.; Uebersfeld J. CR Hebd. Acad. Sci. 1958, 246, 2121.
16 Abragam A.; Proctor, W. G. C. R. Hebd. Seances Acad. Sci. 1958, 246, 2253. Abragam, A “The
Principles of Nuclear Magnetism” Oxford: Clarendon, 1961. Abragam, A.; Goldman, M. Rep. Prog. Phys.
1978, 41, 395.
17 Jeffries, C. D. “Dynamic Nuclear Orientation” New York: Wiley 1965. For a recent review Karabanov, A.;
Kwiatkowski, G.; Köckenberger, W. Molecular Physics 2014, 112, 1838.
-16-
other. The SE mechanism depends on nuclear spin diffusion18 to distribute
polarization to the bulk nuclei. To take advantage of SE a polarization
agent with both the homogeneous EPR linewidth (δ) and the
inhomogeneous spectral breadth (Δ) that are smaller than the nuclear
Larmor frequency (δ, Δ < ω0I) is required. The efficiency of nuclear
polarization transfer via solid effect depends on mixing of states in
electron spin subspaces which is decreases with increasing B0,19 so SE
DNP enhancements scales20 as ω0I-2. The deterioration in SE efficiency at
fields over 3T can be compensated for by using higher microwave power.
Shortly after the development of SE, high energy physicists studying spin-dependent
scattering needed to increase proton concentration in their polarized accelerator targets
The solution was to work in frozen alcohols doped with radicals rather than single
crystals of lanthanide salts. The use of frozen solutions and higher fields led to
inhomogeneous broadening of the EPR lines causing overlap of the forbidden
transitions (leading to differential SE or complete cancelation of the solid effect) and
enhancements plummeted. Necessity then led to the discovery of cross effect DNP by
Kessenikh and Manenkov.21
cross effect (CE)22 is dominant when in a three spin system, two sufficiently strongly
dipolar-coupled radicals interact with a nucleus with non-zero spin. The
inhomogeneous breadth (Δ) of the complete EPR spectrum must be larger
than or equal the nuclear Larmor frequency, while homogeneous width
must remain small (δ < ω0I ≤ Δ). The energy matching condition is then
ωS1–ωS2 = ω0I, where ωS1 and ωS2 are electron Larmor frequencies of two
18
Khutsishvili, G. R. Sov. Phys. Uspehki 1969, 11, 802.
Ni, Q. Z.; Daviso, E.; Can, T. V.; Markhasin, E.; Jawla, S. K.; Swager, T. M.; Temkin, R. J.; Herzfeld, J.;
Griffin, R. G. Acc. Chem. Res. 2013, 46, 1933.
20 Abragam, A.; Goldman, M. “Nuclear Magnetism: Order and Disorder”, Oxford University Press, 1982.
21 Kessenikh, A. V.; Manenkov, A. A. Sov. Phys. Solid State, 1963, 5, 835.
22 Pine, A.; Gibby, M. G.; Waugh, J. S. J. Chem. Phys. 1972, 56, 1776. Bennett, A. E.; Rienstra, C. M.;
Auger, M.; Lakshmi, K. V.; Griffin, R. G. J. Chem. Phys. 1995, 103, 6951. Barnes, A. B.; Paepe, G. D.;
Wel, P. C. A. v. d.; Hu, K.-N.; Joo, C.-G.; Bajaj, V. S.; Mak-Jurkauskas, M. L.; Sirigiri, J. R.; Herzfeld, J.;
Temkin, R. J.; Griffin, R. G. Appl. Magn. Reson. 2008, 34, 237.
19
-17-
different dipolar-coupled radicals (or of two different individual spin
packets within inhomogeneously broad EPR lines of two identical
radicals). Unlike in SE, microwave irradiation in CE drives an allowed
transition, saturating one of the two EPR lines ωS1 or ωS2, or a spin packet
within an inhomogeneously broad EPR line. CE DNP enhancements scale
with ω0I-1. CE can be accomplished using (arranged in increasing
theoretical efficiency) high concentrations of a single broad line radical, a
mixture of broad and narrow radicals, a homo-biradical consisting of two
broad radicals covalently bound in such a way as to optimize spin-spin
coupling, a hetero-biradical consisting of a narrow line radical and a broad
line radical, or a hetero-biradical consisting of two narrow line radicals with
different g-values. The last case is theoretically ideal, but is not practical,
because in order to meet the matching condition two persistent narrow line
radicals with perfectly spaced g-values would have to be covalently bound
at the perfect distance to enable just the right amount of spin-spin
coupling. Lacking a reliable method of predictably tuning radical EPR
frequency by altering the chemical structure, and considering the dearth of
known stable narrow-line organic radicals23 and the synthetic difficulty
associated with their preparation and derivatization ideal CE biradicals are
not yet within our reach. Also, because ω0I is different for different nuclei,
one biradical could not be used to polarize both 1H and
13C,
for example.
However, the “second best” option -covalently bound broad and narrow
radical pair - is a challenging, but reasonable target which could
theoretically be used for a variety of nuclei. (see Chapter 4 for further
discussion of design considerations for CE polarization agents).
At the same time as cross effect theoretical description was being rounded out,
Provotorov24 proposed a many-particle theory based on quantum statistics and spin
23
Only two currently small, although growing families of narrow line radicals are available: trityl-type and
BDPA-type. See Chapter 2 for discussion of narrow line radicals.
24 Provotorov, B. N. Sov. Phys. JETP, 1961, 14, 1126.
-18-
thermodynamics which took into all spin-spin processes including spin-spin cross
relaxation not accounted for by CE. The theory was further developed by Buishvili, and
expanded to low temperature (where Provotorov theory did not apply) by Borghini and
Kozhushner.25 Atsarkin26 demonstrated TM polarization in Cr3+-doped alumina crystal by
showed that the 27Al nuclei polarized at 1.8 K reached the same spin-temperature as
the paramagnetic ions. Wind27 continued the work and described indirect thermal mixing
when both allowed and forbidden EPR transitions are activated.
thermal mixing (TM)28 involves a homogeneously broad EPR line (δ ≥ ω0I), from
multiple dipolar coupled electrons. Inhomogeneously broadened radicals
can also be used so long as the cross-relaxation between spin packets is
faster than spin-lattice relaxation. TM is a generalization of the 3-spin CE
to the real many-particle spin-coupled system, and in many ways the two
mechanisms are similar. TM is dominant at low temperatures (which slow
spin-lattice relaxation) and/or low nuclear larmor frequencies. The low
NMR frequency can be achieved by performing DNP at low field <5T, or by
polarizing low ɣ nuclei. TM requires high concentrations29 of paramagnetic
polarizing agent to achieve strong spin-spin interactions at high fields.
However, most TM experiments are conducted at low fields because TM is
more efficient then, and only 10–40 mM radical is required. Different nuclei
present in the sample achieve the same spin temperature30 under TM.
The allowed EPR transition is irradiated off resonance and the frequency
of microwave irradiation which produce maximum enhancements is
determined by the radicals. The efficiency of nuclear polarization transfer
25
Buishvili, L. L. Sov. Phys. JETP 1965, 22, 1277. Borghini, M. Phys. Rev. Lett. 1968, 20, 419.
Kozhushner, M. A. Sov. Phys. JETP 1969, 29, 136
26 Atsarkin, V. A.; Mefed, A. E.; Rodak, M. I. Phys. Lett. A 1968, 27, 57.
27 Wind, R.; Duijvestijn, M.; Van Der Lugt, C.; Manenschijn, A.; Vriend, J. Prog. Nucl. Magn. Reson.
Spectrosc. 1985, 17, 33.
28 Goldman, M. “Spin Temperature and Nuclear Magnetic Resonance in Solids”, Oxford University Press,
London, 1970.
29 In practice, the optimal concentration appears to depend on the radical (possibly dictated by the spin
distribution within the molecule) and the same concentration for a given radical tends to be used by
different groups to achieve the highest enhancements independent of the DNP mechanism.
30 This was shown experimentally in LiF Cox, S.; Bouffard, V.; Goldman, M. J. Phys. C: Solid State Phys.
1973, 6, L100.
-19-
via thermal mixing scales as B0-2 so at higher temperatures (~90 – 100 K)
CE with a well designed polarization agent is more efficient (and is in fact
the most efficient way of polarization transfer so far).
1.3 Major advancements in DNP technology
In 60’s, 70’s and 80’s DNP was the domain of physicists and physical chemists. They
have developed theoretical models describing the different polarization transfer
mechanisms. Experiments were conducted at low magnetic fields (~1.4 T) using
klystrons producing 40 GHz microwaves. Paramagnetic impurities present in samples
(e.g. coal or inorganic salts) were the spin sources. In order to become relevant in
chemistry, biology, and medicine a major improvement in instrumentation was needed
along with access to new radicals which could be doped into the various samples
chemically and biologically interesting samples.
The instrumentation improvement was provided by Griffin and Temkin31
groups in the early ‘90s. They developed a powerful microwave source -a gyrotronwhich generated high frequency microwaves required for DNP at high fields. Coupled to
a 5 T room temperature MAS SSNMR setup, this 20 W 140 GHz beast produced SE
DNP enhancements ~10 for 1H and ~40 for
13C
of polystyrene doped with a narrow line
radical. Next came experiments at low temperature in aqueous glassing matrix. There
was no water-soluble narrow line radical available at the time, so TEMPO was used.
Enhancements from SE in this case turned out to be very small, as expected with an
inhomogeneously broad radical. However, irradiating closer to the EPR line produced
large power-dependent TM enhancements (ε= 185 for 1H using 1 W microwave power).
Since then, Griffin group has been developing DNP instruments incorporating stronger
32
Becerra, L.; Gerfen, G.; Temkin, R.; Singel, D.; Griffin, R. G. Phys. Rev. Lett. 1993, 71, 3561.
-20-
magnets and higher frequency microwave sources. In collaboration with Swager and
Tordo groups they have also developed several biradicals32 for use in CE DNP.
Next great leap for DNP was the introduction of trityl-type radicals together
with dissolution DNP by Golman, Thaning, and Ardenkjær-Larsen.33 This method was
aimed at medical imaging (hyperpolarized MRSI) from the beginning. Dissolution DNP
combines the brute force temperature jump approach with thermal mixing to obtain very
high enhancements. The experiments are conducted at very low temperature (~1.4 K)
and low field (~3 T) using a low-power Gunn diode microwave source. The polarized
sample is then rapidly dissolved and the NMR acquisition is performed in liquid state.
Commercial equipment is now available to screen polarization agents, targets and
glassing matrixes in vitro, and for polarization in sterile conditions for injection into live
imaging subjects. Dissolution DNP is discussed further in Chapter 3.
Together these two advancements have rejuvenated DNP and number of
publications have since grown accordingly (Figure 3).
32
Hu, K.-N. K.; Yu, H.-H. H.; Swager, T. M. T.; Griffin, R. G. J. Am. Chem. Soc. 2004, 126, 10844. Song,
C.; Hu, K.-N.; Joo, C.-G.; Swager, T. M.; Griffin, R. G. J. Am. Chem. Soc. 2006, 128, 11385. Matsuki, Y.;
Maly, T.; Ouari, O.; Karoui, H.; Le Moigne, F.; Rizzato, E.; Lyubenova, S.; Herzfeld, J.; Prisner, T.; Tordo,
P.; Griffin, R. G. Angew. Chem. Int. Ed. 2009, 48, 4996. Dane, E. L.; Corzilius, B.; Rizzato, E.; Stocker, P.;
Maly, T.; Smith, A. A.; Griffin, R. G.; Ouari, O.; Tordo, P.; Swager, T. M. J. Org. Chem. 2012, 77, 1789.
Kiesewetter, M. K.; Corzilius, B.; Smith, A. A.; Griffin, R. G.; Swager, T. M. J. Am. Chem. Soc. 2012, 134,
4537. Kiesewetter, M. K.; Michaelis, V. K.; Walish, J. J.; Griffin, R. G.; Swager, T. M. J. Phys. Chem. B
2014, 118, 1825.
33 Ardenkjær-Larsen, J. H.; Fridlund, B.; Gram, A.; Hansson, G.; Lerche, M. H.; Servin, R.; Thaning, M.;
Golman, K. Proc. Natl. Acad. Sci. USA 2003, 100, 10158. Golman, K.; In’t Zandt, R.; Thaning, M. Proc.
Natl. Acad. Sci. USA 2003, 100, 11270. Thaning, M. (2000) U.S. Patent 6,013,810.
For a recent review of dissolution DNP see Köckenberger, W. eMagRes, 2014, 3, 161.
-21-
500
'03 First Dissolution DNP by Golman and Ardenkjær-Larsen
Beginning of Hyperpolarized Imaging Revolution
450
400
'93 Griffin Group develops Gyrotron
The Start of DNP MAS SSNMR at High Fields
Number of Publications
350
'63 Kessenikh and Manenkov
Propose Cross Effect
300
250
'58 Abragam and Proctor Describe Solid Effect
DNP Moves onto Insulating Solids
200
150
'53 Overhauser Predicts DNP
Carver and Slichter
Show OE Polarization of Li
100
0
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
50
Year
Figure 3. Number of DNP-related publications per year (red)34. (Additional projected
papers for 2015 in blue.)
1.4 Contemporary DNP methods
Different types of DNP methods have been developed over time and found their
applications in liquid35 and solid state NMR36, and MRSI.37 The field is rapidly evolving,
and significant improvements are being made as a result of systematic studies.
34
Data obtained from SciFinder Scholar using topic search terms: Dynamic Nuclear Polarization,
Hyperpolarized MRI, Hyperpolarized MRSI. Duplicates were removed. Current as of August 1, 2015.
35 Davis, A. L.; Day, I. J. Dynamic Nuclear Polarization: Applications to Liquid-State NMR Spectroscopy;
John Wiley & Sons, Ltd: Chichester, UK, 2007; pp. 1–9.
36 For a review of spin alignment methods used to improve biomolecular NMR see Ardenkjær-Larsen, J.
H.; Boebinger, G. S.; Comment, A.; Duckett, S.; Edison, A. S.; Engelke, F.; Griesinger, C.; Griffin, R. G.;
Hilty, C.; Maeda, H.; Parigi, G.; Prisner, T.; Ravera, E.; van Bentum, J.; Vega, S.; Webb, A.; Luchinat, C.;
Schwalbe, H.; Frydman, L. Angew. Chem. Int. Ed. 2015, 54, 9162. For a review of DNP applied to protein
structure elucidation see Su, Y.; Andreas, L.; Griffin, R. G. Annu. Rev. Biochem. 2015, 84, 465.
37 Kurhanewicz, J.; Bok, R.; Nelson, S. J.; Vigneron, D. B. J. Nucl. Med. 2008, 49, 341.
-22-
Contemporary DNP-enhanced solid state NMR is done at high field, using high
frequency microwaves, and relies on CE with custom nitroxide biradicals.38 The details
of the involved polarization transfer mechanisms are still evolving.39 New MAS
equipment40 and polarization agents are being actively developed.
With DNP-SSNMR experiments, DNP and NMR are performed at the
same field and temperature in a MAS rotor.41 With dissolution DNP that is not the case,
so several improvements to the technique are possible. Dissolution DNP is improved by
scavenging free radicals after dissolution (filter or reduction42) and by using
immobilized43 radicals. The polarization time required by TM at low temperature is
reduced by lanthanides.44 Advanced pulse sequences45 are modulating the “single shot”
limitation of the dissolution DNP to allow 2D correlations, and cross polarization (CP)46
reduces instrument time.
While dissolution DNP yields liquid NMR and MRSI samples, the
polarization is done in solid state. Solution-state DNP47 which relies on OE at room
temperature has to balance the benefit obtained from DNP enhancement versus the
NMR line broadening due caused by the presence of the radicals in the sample. The
38
Maly, T.; Debelouchina, G. T.; Bajaj, V. S.; Hu, K.-N.; Joo, C.-G.; Mak-Jurkauskas, M. L.; Sirigiri, J. R.;
van der Wel, P. C. A.; Herzfeld, J.; Temkin, R. J.; Griffin, R. G. J. Chem. Phys. 2008, 128, 052211. Ni, Q.
Z.; Daviso, E.; Can, T. V.; Markhasin, E.; Jawla, S. K.; Swager, T. M.; Temkin, R. J.; Herzfeld, J.; Griffin, R.
G. Acc. Chem. Res. 2013, 46, 1933. Michaelis, V. K.; Ong, T. C.; Kiesewetter, M. K.; Frantz, D. K.; Walish,
J. J.; Ravera, E.; Luchinat, C.; Swager, T. M.; Griffin, R. G. Isr. J. Chem. 2014, 54, 207.
39 Solid State DNP mechanisms review Can, T. V.; Ni, Q. Z.; Griffin, R. G. J. Magn. Reson. 2015, 253, 23.
40 Thrurber, K. R.; Tycko, R. J. Chem. Phys. 2012, 137, 084508. Thurber, K. R.; Potapov, A.; Yau, W.-M.,
Tycko, R. J. Magn. Reson. 2013, 226, 100.
41 Currently there is no cavity used with MAS DNP. Microwaves are just blasted out of a horn onto the
rotor in "free space".
42 Miéville, P.; Ahuja, P.; Sarkar, R.; Jannin, S.; Vasos, P. R.; Gerber-Lemaire, S.; Mishkovsky, M.;
Comment, A.; Gruetter, R.; Ouari, O.; Tordo, P.; Bodenhausen, G. Angew. Chem. Int. Ed. 2010, 49, 6182.
43 Gajan, D.; Bornet, A.; Vuichoud, B.; Milani, J.; Melzi, R.; van Kalkeren, H. A.; Veyre, L.; Thieuleux, C.;
Conley, M. P.; Grüning, W. R.; Schwarzwälder, M.; Lesage, A.; Copéret, C.; Bodenhausen, G.; Emsley, L.;
Jannin, S. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 14693.
44Gordon, J. W.; Fain, S. B.; Rowland, I. J. Magn. Reson. Med. 2012, 68, 1949.
45 Frydman, L.; Scherf, T.; Lupulescu, A. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 15858. Donovan, K. J.;
Frydman, L. J. Magn. Reson. 2012, 225, 115.
46Weis, V.; Griffin, R. G. Solid State Nucl. Magn. Reson. 2006, 29, 66. Gadian, D. G.; Panesar, K. S.;
Perez Linde, A. J.; Horsewill, A. J.; Köckenberger, W.; Owers-Bradley, J. R. Phys. Chem. Chem. Phys.
2012, 14, 5397. Batel, M.; Däpp, A.; Hunkeler, A.; Meier, B. H.; Kozerke, S.; Ernst, M. Phys. Chem.
Chem. Phys. 2014, 16, 21407.
47Loening, N. M.; Rosay, M.; Weis, V.; Griffin, R. G. J. Am. Chem. Soc. 2002, 124, 8808.
-23-
paramagnetic relaxation results in nuclear T2 shortening, so OE in liquids is more useful
in to nuclei with intrinsically long T2 (enhancements for 1H and
19F
<
31P, 13C
and
15N).
The chemical shift range should also be considered, because the line broadening
results in loss of resolution. Liquid state polarization methods relying on OE are being
improved by predicting and optimizing the Fermi contact interactions.48
Studies of the glassing matrix49 effects on polarization transfer and spin
diffusion inform optimal sample composition.
A variety of DNP methods has been developed using a different
experimental setups to take advantage of the DNP mechanisms described above,
however a stable polarizing agent is a universal requirement.50
Examples of the polarizing agents that have been recently used as the
source of the unpaired electrons in the DNP experiments are given in Figure 4. The
structure of the polarizing agent has a direct effect (although not completely predictable)
on DNP mechanism and the enhancement of the NMR signal via DNP. Because
biological systems require aqueous environment, high water solubility is essential in the
development of new polarizing agents. Narrow EPR linewidth (≤ 40 MHz at 211 MHz 1H
Larmor frequency) is required for SE experiments as mentioned above, and biradicals
with EPR resonances combining narrow homogeneous linewidth with inhomogeneous
breadth matching the nuclear Larmor frequency will give the largest enhancements via
CE. From the point of view of a synthetic chemist, the relaxation pathways and the
resulting DNP mechanisms are important to consider in the design and synthesis of
polarization agents - persistent organic radicals (and occasionally chelated
paramagnetic ions).
48
Wang, X.; Isley, W. C., III; Salido, S.; Sun, Z.; Song, L.; Cramer, C. J.; Dorn, H. Chem. Sci. 2015. DOI:
10.1039/C5SC02499D. (in print)
49 Kurdzesau, F.; van den Brandt, B.; Comment, A.; Hautle, P.; Jannin, S.; van der Klink, J. J.; Konter, J.
A. J Phys. D Appl. Phys. 2008, 41, 155506. Ludwig, C.; Marin-Montesinos, I,; Saunders, M. G.; Gunther,
U. L. J. Am. Chem. Soc. 2010, 132, 2508.
Lumata, L.; Kovacs, Z.; Malloy, C.; Sherry, A. D.; Merritt, M. Phys. Med. Biol. 2011, 56, N85.
50 Maly, T.; Debelouchina, G. T.; Bajaj, V. S.; Hu, K.; Joo, C.; Mak-Jurkauskas, M. L.; Sirigiri, J. R.; van der
Wel, P. C. A.; Herzfeld, J.; Temkin, R. J.; Griffin, R. G. J. Chem. Phys. 2008, 128, 052211
-24-
Narrow line radicals used for SE and TM DNP
R
Cl
RO 2C
CH2OH
O
Cl
Cl Cl
Cl
Cl
Cl
Cl
O
RR
S
S
R
HN
R
R
S
S
S
S
R
R
O
R
S
S
R
R
CO2Na
R= Me Finland Trityl (cTAM aka CT-03)
R= CH2CH2OH OX063 (aka OXO-63&OX63)
R= CH2CH2OMe OX063Me
S
R
R
S
R
O 2N
t-Bu
O•
N
t-Bu
NO 2
R
t-Bu
Galvinoxyl
DPPH
COOH
R NH
t-Bu
O
N•
S
S
S
S
S
R
R
NO 2
COOH
S
R
R
CO2Na
O
CO2Me
S
S
S
N R
O
S
S
S
R
O
R
R
R=
R N
S
NaO 2C
O
RR
Broad radicals used for TM and CE DNP
R
R
CO2Me
O
CO2R
R= H
R= Na
Narrow line biradicals used for TM DNP
R NH HN
O
O
Cl
BA-BDPA
S
O
Cl
Cl
Cl
BDPA
O
MeO 2C
CO2R
O
O•
O•
O•
N
N
N
TEMPO
OH
O
TEMPOL
4-oxo-TEMPO
Broad biradicals designed for CE DNP
•O
N
O
N
O•
O
N
N
H
H
BTurea
•O
N
S
O
N
•O
N
N
N
S
O
O
n
N
O
N
•O
N
N
H
H
BTthio-3
O
H
N
N
O
X
X
•O
N
O
bTbtk-py
O•
•O
O
N O•
O
O
O
N
H
X
X
X
X = S, SO, SO2
bCTbk
N
N O•
O
bTbtk
X
N O•
O
O
TOTAPOL
O•
X
N
X
X = S, SO, SO2
O
N
O•
N
H
OH
•O
O
O
•O
O•
N O•
O
bTbk
N
N
H
BTthio-2
N
O
N
n= 2,3,4
O
•O
•O
BTnE
•O
O
S
N
O•
BTthio-1
•O
O
O•
N
O
N
N
H
O•
N
R
O
R = H PyPol
R = (CH 2CH2O)4Me AMUPol
BTOXA
Figure 4. Some persistent radicals employed by DNP.
BDPA Loening, N. M.; Rosay, M.; Weis, V.; Griffin, R. G. J. Am. Chem. Soc. 2002, 124, 8808. Lumata, L.;
Ratnakar, S. J.; Jindal, A.; Merritt, M.; Comment, A.; Malloy, C.; Sherry, A. D.; Kovacs, Z. Chem. Eur. J.
2011, 17, 10825.
BA-BDPA Muñoz-Gómez, J. L.; Monteagudo, E.; Lloveras, V.; Parella, T.; Veciana, J.; Vidal-Gancedo, J.
Org. Biomol. Chem. 2015, 13, 2689.
Polychlorinated trityls Paniagua, J. C.; Mugnaini, V.; Gabellieri, C.; Feliz, M.; Roques, N.; Veciana, J.;
Pons, M. Phys. Chem. Chem. Phys. 2010, 12, 5824.
-25-
OX063 This is probably the most used radical in dissolution DNP and hyperpolarized MRSI. It is
commercially available. J. H. Ardenkjær-Larsen, B. Fridlund, A. Gram, G. Hansson, L. Hansson, M. H.
Lerche, R. Servin, M. Thaning and K. Golman, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 10158.
BTnE K.-N. Hu, H.-h. Yu, T. M. Swager, and R. G. Griffin, J. Am. Chem. Soc. 2004, 126, 10844 ︎.
TOTAPOL Song, C.; Hu, K.-N.; Joo, C.-G.; Swager, T. M.; Griffin, R. G. J. Am. Chem. Soc. 2006, 128,
11385.
bTbk Matsuki, Y.; Maly, T.; Ouari, O.; Karoui, H.; Le Moigne, F.; Rizzato, E.; Lyubenova, S.; Herzfeld, J.;
Prisner, T.; Tordo, P.; Griffin, R. G. Angew. Chem. Int. Ed. 2009, 48, 4996.
bTbtk Dane, E. L.; Corzilius, B.; Rizzato, E.; Stocker, P.; Maly, T.; Smith, A. A.; Griffin, R. G.; Ouari, O.;
Tordo, P.; Swager, T. M. J. Org. Chem. 2012, 77, 1789.
bTbtk-PY Kiesewetter, M. K.; Corzilius, B.; Smith, A. A.; Griffin, R. G.; Swager, T. M. J. Am. Chem. Soc.
2012, 134, 4537.
BTurea BTOXA K.-N. Hu, C. Song, H. Yu, T. M. Swager, R. G. Griffin, J. Chem. Phys. 2008, 128, 052302
BTthio Michaelis, V. K.; Ong, T. C.; Kiesewetter, M. K.; Frantz, D. K.; Walish, J. J.; Ravera, E.; Luchinat,
C.; Swager, T. M.; Griffin, R. G. Isr. J. Chem. 2014, 54, 207.
PyPol, AMUPol Sauvée, C.; Rosay, M.; Casano, G.; Aussenac, F.; Weber, R. T.; Ouari, O.; Tordo, P.
Angew. Chem. Int. Ed. 2013, 52, 10858.
DPPH Lumata, L.; Merritt, M.; Khemtong, C.; Ratnakar, S. J.; van Tol, J.; Yu, L.; Song, L.; Kovacs, Z. RSC
Adv 2012, 2, 12812.
Galvinoxyl Lumata, L. L.; Merritt, M. E.; Malloy, C. R.; Sherry, A. D.; van Tol, J.; Song, L.; Kovacs, Z. J.
Magn. Reson. 2013, 227, 14.
Trityl biradicals Macholl, S.; Jóhannesson, H.; Ardenkjaer-Larsen, J. H. Phys. Chem. Chem. Phys.
2010, 12, 5804.
-26-
Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-Enhanced
Solid State NMR
Δ = 28 MHz
FWHM
DNP enhanced
MAS NMR
(SO3Na)n
mw on
mw off
1H
Enhancement
MAS DNP
30
40
50
60
70
n(NaO3S)
ε = 94
80
n(NaO3S)
180
(SO3Na)n
190
80
180
Magnetic Field (mT)
SA-BDPA
13C
60
Chemical Shift (ppm)
(SO3Na)n
13C
EPR
40
Chemical Shift (ppm)
Adapted and reprinted in part with permission from:
Haze, O.; Corzilius, B.; Smith, A. A.; Griffin, R. G.; Swager, T. M.
“Water-Soluble Narrow-Line Radicals for Dynamic Nuclear Polarization”
J. Am. Chem. Soc. 2012, 134, 14287.
and
Smith, A. A.; Corzilius, B.; Haze, O.; Swager, T. M.; Griffin, R. G. J. Chem. Phys. 2013,
139, 214201.
Copyright 2012 American Chemical Society
Contributed: Bjorn Corzilius, Andy Smith, R.G. Griffin
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.1 Introduction: DNP and Solid State NMR
Although the development of magic angle spinning (MAS) experiments in the 1980s
made solid NMR a practical research tool,1 the technique, however, suffers from
relatively poor sensitivity especially for low abundance nuclei and those with low
gyromagnetic ratios (γn). Among these nuclei, 13C and 15N are particularly important
given their prevalence in biological samples. Dynamic nuclear polarization (DNP) is
used to significantly enhance the limited sensitivity of NMR studies of small molecules
and complex biological systems.2 DNP in general and DNP-enabled solid state NMR3 in
particular are currently experiencing a renaissance. This is driven by the interest in
experiments at high magnetic field, and enabled by gyrotrons capable of delivering
microwaves at the high frequency and power required for such experiments.4 In addition
to increased sensitivity and decreased measurement times, DNP MAS SSNMR allows,
for example, the study of proteins that can not be crystallized (e.g. transmembrane
proteins) and, therefore, can not probed by X-ray crystallography.5 Recently, the
1
For historical reviews, see: Hennel, J. W.; Klinowski, J. Top. Curr. Chem. 2005, 246, 1. Andrew, E. R. In
Solid State NMR Studies of Biopolymers, McDermott, A. E.; Polenova, T., Eds.; John Wiley & Sons:
Chichester, 2010; pp. 83–97.
2 Wind, R. A.; Duijvestijn, M. J.; van der Lugt, C.; Manenschijn, A.; Vriend, J. Prog. Nucl. Magn. Reson.
Spectrosc. 1985, 17, 33. Gerfen, G. J.; Becerra, L. R.; Hall, D. A.; Griffin, R. G.; Temkin, R. J.; Singel, D.
J. J. Chem. Phys. 1995, 102, 9494. Hall, D. A.; Maus, D. C.; Gerfen, G. J.; Inati, S. J.; Becerra, L. R.;
Dahlquist, F. W.; Griffin, R. G. Science 1997, 276, 930. Rosay, M. M. Ph.D. Thesis, Massachusetts
Institute of Technology, Cambridge, MA, 2001. Rosay, M.; Weis, V.; Kreischer, K. E.; Temkin, R. J.; Griffin,
R. G. J. Am. Chem. Soc. 2002, 124, 3214. Rosay, M.; Tometich, L.; Pawsey, S.; Bader, R.; Schauwecker,
R.; Blank, M.; Borchard, P. M.; Cauffman, S. R.; Felch, K. L.; Weber, R. T.; Temkin, R. J.; Griffin, R. G.;
Maas, W. E. Phys. Chem. Chem. Phys. 2010, 12, 5850.
3 For a recent review of the effect of DNP on solid state NMR see Lee, D.; Hediger, S.; De Paëpe, G.
Solid State Nucl. Magn. Reson. 2015, 66-67, 6.
4 Maly, T.; Debelouchina, G. T.; Bajaj, V. S.; Hu, K.-N.; Joo, C.-G.; Mak-Jurkauskas, M. L.; Sirigiri, J. R.;
van der Wel, P. C. A.; Herzfeld, J.; Temkin, R. J.; Griffin R. G. J. Chem. Phys. 2008, 128, 052211.
5 Griffiths, J. M.; Lakshmi, K. V.; Bennett, A. E.; Raap, J.; Vanderwielen, C. M.; Lugtenburg, J.; Herzfeld,
J.; Griffin, R. G. J. Am. Chem. Soc. 1994, 116, 10178. Rienstra, C. M.; Hohwy, M.; Hong, M.; Griffin, R. G.
J. Am. Chem. Soc. 2000, 122, 10979. Reif, B.; Jaroniec, C. P.; Rienstra, C. M.; Hohwy, M.; Griffin, R. G. J.
Magn. Reson. 2001, 151, 320. Rosay, M.; Lansing, J. C.; Haddad, K. C.; Bachovchin, W. W.; Herzfeld, J.;
Temkin, R. J.; Griffin, R. G. J. Am. Chem. Soc. 2003, 125, 13626. Mak-Jurkauskas, M. L.; Bajaj, V. S.;
Hornstein, M. K.; Belenky, M.; Griffin, R. G.; Herzfeld, J. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 883.
Bajaj, V. S.; Mak-Jurkauskas, M. L.; Belenky, M.; Herzfeld, J.; Griffin, R. G. Proc. Natl. Acad. Sci. U.S.A.
2009, 106, 9244.
-28-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
structure of insoluble amyloid fibrils (associated with Alzheimer's disease) have been
elucidated with the help of DNP.6
DNP relies on the transfer of polarization from relatively highly polarized
unpaired electron spins to nuclear spins (e.g., 1H or 13C) via microwave excitation of the
electron transitions. Unpaired electrons are supplied by the polarizing agents, which in
most cases are persistent radical species. In practice, DNP MAS SSNMR is performed
with frozen glass-forming solutions or a solid mixture of the sample and the polarization
agent. The sample is continuously irradiated from a high power microwave source
(gyrotron) with a frequency at or near the Larmor frequency of the electron spins.
Several mechanisms for polarization transfer have been identified, in this chapter will
focus on Solid Effect (SE) DNP.
2.2 Polarization transfer mechanisms: Solid Effect
The solid effect7 (SE) DNP is a two-spin process that relies on the polarization transfer
between a single electron spin and a nuclear spin. As described in Chapter 1, at thermal
equilibrium (Scheme 1, a) there’s a large energy –and therefore population– difference
between the different orientations of the electron spin with respect to the magnetic field,
while the difference due to nuclear spin is small. Qualitatively8, irradiating the SSNMR
sample at the microwave frequency (ω) equal to the electron Larmor frequency (ωe or
ω0S) saturates the EPR transition (Scheme 1, grey arrows, c). This does not change the
population difference relevant to NMR, so no SE enhancement (ε) is observed. As the
frequency is swept, two enhancement maxima appear at the matching conditions ω =
6
Debelouchina, G. T.; Bayro, M. J.; van der Wel, P. C. A.; Caporini, M. A.; Barnes, A. B.; Rosay, M.; Maas,
W. E.; Griffin, R. G. Phys. Chem. Chem. Phys. 2010, 12, 5911. For a review of MAS SSNMR of
biomacromolecules see Su, Y.; Andreas, L.; Griffin, R. G. Ann. Rev. Biochem. 2015, 84, 465.
7 Jefferies, C. D. Phys. Rev. 1957, 106, 164. Jeffries, C. D. Phys. Rev. 1960, 117, 1056. Abragam, A.;
Proctor, W. G. C. R. Hebd. Seances Acad. Sci. 1958, 246, 2253. Wenckebach, W. T. Appl. Magn. Reson.
2008, 34, 227. Smith, A. A.; Corzilius, B.; Barnes, A. B.; Maly, T.; Griffin, R. G. J. Chem. Phys. 2012, 136,
015101. Karabanov, A.; Kwiatkowski, G.; Köckenberger, W. Molecular Physics 2014, 112, 1838. For a
recent review of DNP mechanisms in insulating solids including SE, OE and CE see T. V. Can, Q. Z. Ni,
R. G. Griffin, J. Magn. Reson. 2015, 253, 23.
8 For a quantitative description of the SE DNP experiment see Abragam, A.; Goldman, M. Rep. Prog.
Phys. 1978, 41, 395, and other SE references above.
-29-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
(ωe + ωN) and ω = (ωe – ωN) corresponding to the zero-quantum (both electron and
nuclear spins flip, but because they are opposite to begin with, overall spin quantum
number change is zero) and double-quantum transitions (both electron and nuclear
spins flip, and because they were the same to begin with, overall spin quantum number
change is two), (Scheme 1, b, d). In actual DNP experiments, the microwave frequency
is fixed and the field is swept. So essentially it is the energy level splitting that is swept
and as the splitting matches the microwave frequency enhancement maxima appear.
Because smaller energy splittings require stronger fields to match the microwave
frequency, the peaks in field swept profiles appear in opposite order. I.e. the higher
energy “higher frequency” zero quantum transition (ωe + ωN) which gives negative NMR
enhancement occurs at a lower field, while the “lower frequency” double quantum
transition (ωe – ωN) which gives rise to the positive NMR enhancement occurs at higher
field.
Scheme 1. Energy Levels and Transitions Involved in Polarization Transfer from
!"#
Electrons to Nuclei
$%&'%&$()*+(,-*.
PI
ω
ω0Se
ω
ω0Se −–ωω0I N
e↑N↑
e↑N↓
NMR
ZQ
EPR
EPR
DQ
e ↓ N↑
NMR
e↓N↓
+ small state mixing at
high fields
ω
ω0Se ++ωω0I N
ω
!"#$ %$ R8*6(-31 526-3?@-0?2)1 -/1 -1 >8)*0?2)1 2>1 0.(1 4?*32C-D(1 >3(:8()*71 ?)1 0.(1 *-/(1 2>1 0.(1 C(66M
3(/26D(B1 /26?B1 (>>(*0%
/-0(1 (-*.1 20.(31 *2456(0(671 ,(*-8/(1 0.(1 3-0(1 9:%1 ;<=1 2>1 0.(1 526-3?@-0?2)1 03-)/>(31 ?/
4-A?4-61 -01 B?>>(3()01 4?*32C-D(1 >3(:8()*?(/1 !!&"&!E!&#&!E" *2/"!1 -)B1 !#&"&!E!&!
!E" *2/"# %
ωe
ωe – ω N
F)1 >-*0G1 C.()1 0.(1 9HI1 6?)(1 ?/1 )-332C1 *245-3(B1 C?0.1 !E" *2/"JG1 52/?0?D(1 -)B
ωe
ωe +>3(:8()*71
ωN
)(K-0?D(1 526-3?@-0?2)1 03-)/>(3/1 2**831 ?)1 0C21
*2456(0(671 /(5-3-0(B1
3(K?2)/%
'.?/1 *-/(1 ?/1 +)2C)1 -/1 0.(1 LC(66M3(/26D(B1 /26?B1 (>>(*0N1 -)B1 O?K%1 #1 /.2C/1 0.(1 3(/860M
?)K1 )8*6(-31 526-3?@-0?2)1 -/1 -1 >8)*0?2)1 2>1 0.(1 4?*32C-D(1 >3(:8()*7%1 P2C(D(3G1 ?>1 0.(
9HI1 6?)(1 C?B0.1 ?/1 )201 )-332C1 *245-3(B1 C?0.1 !E"*2/"!G1 52/?0?D(1 -)B1 )(K-0?D(1 52M
6-3?@-0?2)1 03-)/>(3/1 *245()/-0(1 (-*.1 20.(31 5-3067%1 '.?/1 *-/(1 ?/1 +)2C)1 -/1 0.(1 LB?>M
>(3()0?-61 /26?B1 (>>(*0N1 -)B1 O?K%1 Q1 /.2C/1 .2C1 0.?/1 *245()/-0?2)1 ->>(*0/1 0.(1 3(/860M
Equilibrium
Positive-/1
ε -1 >8)*0?2)1No
Negative
ε
?)K1 )8*6(-31 526-3?@-0?2)1
2>1 ε
0.(1 4?*32C-D(1
>3(:8()*7%
a
b
c
d
PI
-30-
ω0S
ω0S + ω0I
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
In order to avoid exciting both the zero and the double-quantum transitions
simultaneously which would lead to a cancellation of negative and positive
enhancements (i.e., differential SE9), a radical with a narrow EPR line is used. The
overall EPR linewidth is dependent on the homogeneous EPR linewidth (δ), and
inhomogeneous spectral breadth (∆). The homogeneous linewidth is affected by the
electron spin-lattice and spin-spin relaxation times T1e and T2e, with shorter lifetimes
resulting in broader EPR line. The inhomogeneous spectral breadth is a result of ganisotropy and hyperfine coupling, with more spherically symmetric radicals with fewer
coupled nuclei having narrower EPR lines. SE DNP method requires a polarization
agent with both the homogeneous EPR linewidth (δ) and the inhomogeneous spectral
breadth (Δ) smaller than the nuclear Larmor frequency (ω0I). Trityl-type and
bis(diphenylene)-2-phenyl-allyl (BDPA)-based radicals satisfy these requirements at
high magnetic fields for SE DNP of 1H.
2.3 DNP with narrow-line radicals
2.3.1 Trityl-type radicals
The trityl-type radicals10 such as CT-03 and OX06311 (Figure 1) exhibit narrow EPR
linewidth (e.g., Δ = 50 MHz at 140 GHz for OX063) and have been used successfully in
DNP NMR experiments and in vivo EPR imaging of tissue oxygenation and pH.12 A
large variety of trityl-type radicals13 has been prepared which allows applications in
9
Henstra, A.; Dirksen, P.; Wenckebach, W. T. Phys. Lett. A 1988, 134, 134.
Triarylmethyl (TAM) radicals: Ardenkjaer-Larsen, J. H.; Laursen, I.; Leunbach, I.; Ehnholm, G.;
Wistrand, L.-G.; Petersson, J. S.; Golman, K. J. Magn. Reson. 1998, 133, 1. Reddy, T. J.; Iwama, T.;
Halpern, H. J.; Rawal, V. H. J. Org. Chem. 2002, 67, 4635. Lurie, D.; Li, H.; Petryakov, S.; Zweier, J. L.
Magn. Reson. Med. 2002, 47, 181. Paniagua, J. C.; Mugnaini, V.; Gabellieri, C.; Feliz, M.; Roques, N.;
Veciana, J.; Pons, M. Phys. Chem. Chem. Phys. 2010, 12, 5824.
11 Anderson, S.; Golman, K.; Rise, F.; Wikstrom, H.; Wistrand, L.- G., U.S. Patent 5,530,140, 1996.
12 Bobko, A. A.; Dhimitruka, I.; Zweier, J. L.; Khramtsov, V. V. J. Am. Chem. Soc. 2007, 129, 7240. Liu, Y.;
Villamena, F. A.; Sun, J.; Xu, Y.; Dhimitruka, I.; Zweier, J. L. J. Org. Chem. 2008, 73, 1490. Corzilius, B;
Smith, A. A.; Griffin, R. G. J. Chem. Phys. 2012, 137, 054201. Michaelis, V. K.; Corzilius, B.; Smith, A. A.;
Griffin, R. G. J. Phys. Chem. B 2013, 117, 14894.
13 NMR literature tends to be rather sloppy when referring to trityl-type radicals, often calling OX063,
“Trityl” and not specifying whether the radical is in the carboxylic acid or sodium carboxylate form. This is
important because the chemical structure of the radical has significant impact on its solubility and
performance as a polarizing agent.
10
-31-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
aqueous and non-aqueous environments. However, lengthy syntheses14 contribute to
the high cost of these radicals and limits their use as routine polarizing agents.
Cl
RO 2C
R
Cl
Cl Cl
CO2R
Cl
O
O
O
MeO 2C
CO2Me
S
NaO 2C
RR
R
S
S
CO2Na
Cl
Cl
Cl
Cl
Cl
Cl
Cl
O
O
O
O
O
Gomberg Trityl
O
S
CO2R
R= H
R= Na
O
S
R
S
S
S
O
R
R
S
S
R
R
O
R
S
S
R
CO2Me
R
CO2Na
R= Me Finland Trityl (cTAM aka CT-03)
R= CH2CH2OH OX063 (aka OXO-63 and OX63)
Figure 1. Structures of several triarylmethyl radicals.
2.3.2 BDPA-based radicals
BDPA (Figure 2 a) is an air-stable persistent radical that shows no dimerization in solid
state or in solution. First synthesized by Koelsch,15 the properties of this radical were so
unexpected that the JACS refused to publish the results. Thirty years and several
studies16 later and the original paper was resubmitted and published. While trityl-type
radicals rely on lack of hyperfine coupling to achieve the narrow line width, BDPA has
21 protons coupled to the unpaired electron and it still has a narrow EPR linewidth17 (D
≈ 25 MHz at 140 GHz). This is a result of the vanishingly small g-anisotropy of the
BDPA: spin density is distributed over both fluorenyl blades, and partially on the phenyl
group, which are twisted by 30º forming a very symmetrical propeller.18 (Figure 2, c and
b)
14
Dhimitruka, I.; Velayutham, M.; Bobko, A. A.; Khramtsov, V. V.; Villamena, F. A.; Hadad, C. M.; Zweier,
J. L. Bioorg. Med. Chem. Lett. 2007, 17, 6801.
15 Koelsch, C. F. J. Am. Chem. Soc. 1957, 79, 4439
16 Keevil, N. B. J. Am. Chem. Soc. 1937, 59, 2104. Wertz, J. E.; Koelsch, C. F.; Vivo, J. L. J. Chem. Phys.
1955, 23, 2194. Kreevoy, M. M. Tetrahedron, 1958, 2, 354.
17 De Boer, W. J. Low Temp. Phys. 1976, 22, 185.
18 For crystal structures of BDPA and McLachlan π-spin densities see: Azuma, N.; Ozawa, T.; Yamauchi,
J. Bull. Chem. Soc. Jpn. 1994, 67, 31.
-32-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
bis(diphenylene)-2-phenyl-allyl
BDPA
a
b
c
Figure 2. a) Structure of BDPA b) X-ray structure of BDPA from ref. 18 c) positive (blue)
and negative (red) spin densities of BDPA calculated using B3LYP/6-31G(d) plotted at
0.001 and -0.001 isovalue respectively.
Although Azuma has reported McLachlan spin densities alongside the x-ray structures
of BDPA complexes with benzene and acetone, for the purposes of visualization the
spin densities were re-calculated using B3LYP/6-31G(d) as implemented in Spartan 14.
They are plotted at 0.001 and -0.001 isovalues (lower than usual 0.002) in order to
visualize the small negative spin densities19 on the phenyl ring resulting from spin
polarization.
The ability of BDPA to transfer polarization was investigated via SE DNP in
a polystyrene matrix20 and recently via dissolution DNP in sulfolane21 and to study the
stereochemistry, kinetics, and mechanism of olefin polymerization22.
While BDPA was proven to be an excellent polarization agent, its utility in
biologically relevant applications is limited by its lack of solubility in aqueous media.
Only one water-soluble BDPA derivative has been described.23 Eric Dane synthesized
WS-BDPA radical (Scheme 2) as part of his PhD work in Swager lab. He used
19
Negative spin densities are not part of the SOMO, but do give rise to hyperfine coupling.
Wind, R. A.; Duijvestijn, M. J.; van der Luat, C.; Manenschijn, A.; Vriend, J. Prog. Nucl. Magn. Reson.
Spectrosc. 1985, 17, 33. Duijvestijn, M. J.; Wind, R. A.; Smidt, J. Physica B+C 1986, 138, 147. Afeworki,
M.; McKay, R. A.; Schaefer, J. Macromolecules 1992, 25, 4084. Becerra, L. R.; Gerfen, G. J.; Temkin, R.
J.; Singel, D. J.; Griffin, R. G. Phys. Rev. Lett. 1993, 71, 3561.
21 Lumata, L.; Ratnakar, S. J.; Jindal, A.; Merritt, M.; Comment, A.; Malloy, C.; Sherry, A. D.; Kovacs, Z.
Chem.–Eur. J. 2011, 17, 10825.
22 Chen, C.-H.; Shih, W.-C.; Hilty, C. J. Am. Chem. Soc. 2015, 137, 6965.
23 Dane, E. L.; Swager, T. M. J. Org. Chem. 2010, 75, 3533.
20
-33-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
functionalized fluorene carrying ester groups which later were converted into carboxylic
acids imparting water solubility.
Scheme 2. Water-soluble BDPA Synthesis by E. Dane (Scheme from ref. 23 Copyright
American Chemical Society, 2010.)
The solution EPR of the WS-BDPA24 (Figure 3) resembles that of BDPA, however, its
water solubility is limited, and DNP experiments with this compound were not pursued.
Figure 3. The room temperature 9 GHz EPR of 0.1 mM BDPA and 10 mM WS-BDPA
(Figure from ref. 23, Copyright American Chemical Society, 2010.)
24
Small hyperfine couplings are not resolved in the pictured spectra.
-34-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.4 Synthesis and properties of water-soluble narrow-line BDPA-based radicals
2.4.1 Synthetic approaches
Because BDPA performed so well in the DNP experiments to date, we set out to make a
water soluble derivative that would not alter the electronic properties of the BDPA core.
Because very little of the spin density is located on the phenyl ring18, we aimed to attach
the solubilizing groups there. Taking into account the starting materials available and
ease of synthesis, derivative 1 in Scheme 3 was proposed.
Scheme 3. Proposed Water-Soluble BDPA Derivative and its Synthesis
NMe 2
COOH
O
COOH
Br
Br
NaOH, EtOH
Br
COOH
reflux 1 h
N
1) CO2Cl2, cat DMF,
DCM, 0 ºC to rt
Fluorene, t-BuONa
H
DMA
0 ºC to rt
Me 2N
N
H
89%
91%
H
2) cat DMAP
DCM, 0 ºC to rt
NMe 2
NMe 2
yield and stability problems
2
HOOC-BDPAH
1) oxidize
2) 1,3 propanesultone
N+
Me Me
O
SO3–
N
•
N+
Me Me
SO3–
1
However, as the synthesis progressed it became apparent that the stability of the amide
in the para position of the phenyl ring was going to be a problem. While HOOC-BDPAH
was prepared in high yield using conditions reported previously,25 the amide 2
decomposed26 on standing in solution and attempted chromatography yielding several
colorful fractions. Decomposition route was not investigated, however it’s possible that
amide 2 is air oxidized to the corresponding radical (possibly with the tertiary amine
25
Mi, Q., Ph.D. Thesis, Northwestern University, Evanston, IL, 2009. Dane, E. L.; Maly, T.; Debelouchina,
G. T.; Griffin, R. G.; Swager, T. M. Org. Lett. 2009, 11, 1871.
26 Similar amide had oxidized and decomposed in solution and upon heating Dane, E. L. Design,
Synthesis, and Characterization of Conjugated Polymers and Functional Paramagnetic Materials for
Dynamic Nuclear Polarization. Massachusetts Institute of Technology, Cambridge, MA, September 2010.
-35-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
groups acting as an intramolecular base in our case or just the excess amine from the
crude reaction mixture acting as a base) following by photodegradation pathways
similar to the ones described by Fox27 in BDPA (Scheme 4). Because of the amide
stability and purification problems, and because the radical stability in propanesultone
was not certain, taking in consideration that BDPA itself decomposed in strong acids,
this route was abandoned in favor of a simpler approach.
Scheme 4. Possible Degradation Products of Amide 2
NMe 2
NMe 2
O
O
N
N
NMe 2
O
OH
O
N
H
O2, DCM
?
NMe 2
SiO2
2
NMe 2
NMe 2
O
NMe 2
O
N
•
NMe 2
Electrophilic aromatic sulfonation was chosen as a simple alternate route
to a water-soluble BDPA. Cerfontain has studied aromatic sulfonations in detail, and in
particular sulfonation of fluorene28 in sulfuric acid. Regiochemistry of the sulfonation with
sulfuric acid (Scheme 5) is similar to nitration29 with fuming nitric acid — both give 2,5and 2,7-fluorene as main difunctionalization products.
Scheme 5. Sulfonation of Fluorene
27
Breslin, D.; Fox, M. J. Phys. Chem. 1993, 97, 13341.
Schaasberg-Nienhuis, Z.; Cerfontain, H.; Kortekaas, T. J. Chem. Soc. Perkin II 1979, 844.
29 Morgan, G. T.; Thomason, R. W. J. Chem. Soc. 1926, 129, 2691. Dewar, M. J. S.; Urch, D. S. J. Chem.
Soc. 1958, 3079.
28
-36-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
HO 3S
8
1
7
2
6
SO3H
5
4
56 ± 4%
98.3% H 2SO4 rt
3
HO 3S
104% H 2SO4 rt
SO3H
SO3H
SO3H
30 ± 4%
81 ± 3%
SO3H
+other regioisomers
+other regioisomers
Sulfonation proceeds with typical biphenyl reactivity, and the first sulfonyl group is
incorporated at C2 position (Scheme 5, green dot). Once the fluorene is disubstituted,
the meta-directing sulpho-group reinforces the phenyl group directing ability directing
sulfonation to C5 (or C7 for minor isomer). The 4,5-sulfonation is prevented by the
strong steric hindrance of the sulpho-group in the bay region of the fluorene (Scheme
6).
Scheme 6. Steric Hindrance in the Bay Region of Fluorene
SO3H
SO3H
SO
SO3H
3H
SO
3H
The BDPA precursor BDPAH was prepared in four steps from fluorene and
benzaldehyde using published procedures30 with minor changes (Scheme 7).
Scheme 7. BDPAH Synthesis
Ph
H
PhCHO
KOH
BnOH
95 ºC 1.5 h
Ph
Br 2
HOAc
rt 0.5 h
90%
Br
Br
>95%
30
Ph
Br
Fluorene
t-BuONa
DMA
rt 1 h
NaOH
EtOH
rt 3 h
>95%
H
BDPAH
85%
Kuhn, R.; Neugebauer, A. Monatsh. Chem. 1964, 95, 3. Plater, M.; Kemp, S.; Lattmann, E. J. Chem.
Soc., Perkin Trans.1 2000, 971.
-37-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Koelsch had reported that BDPA was stable in mild acid and could be
recrystallized from acetic acid. However, he found that treatment with aqueous chromic
acid in acetic acid resulted in decomposition of the radical to benzoic acid and
fluorenone. Later, Screttas proposed31 that BDPA is oxidized in concentrated H2SO4 and
in TFA to give BDPA cation. In our hands, BDPA radical decomposed in fuming sulfuric
acid. Instead, BDPA precursor, BDPAH, was used in the reaction with fuming sulfuric
acid.32 Commonly, aromatic sulfonic acids are isolated by simple dilution of the reaction
mixture with water followed by filtration. In our case, however, the product was highly
soluble in water at any pH. The reaction mixture was made alkaline with sodium
hydroxide and concentrated. The SA–BDPA– then had to be separated from the large
amount of sodium sulfate produced. After screening several solvents (water, ethanol,
isopropanol, acetonitrile, acetone, THF, DMA, etc., and their mixtures) separation of the
SA–BDPA– was accomplished with DMSO (Na2SO4 solubility in DMSO is very low33).
SA–BDPA– was then precipitated by addition of benzene and washed with DCM and
dried to remove solvent traces. Oxidation with silver nitrate yielded a mixture of the
extremely water-soluble sulfonated radicals SA–BDPA (Scheme 8). Here, again,
working with aqueous solutions separation of the byproduct salts was challenging.
Reverse phase chromatography was used to separate the inorganic salts, but the
oligosulfonated BDPA radicals were not resolved.
31
Screttas, C. G. J. Chem. Soc. Perkin II 1975, 165.
Concurrently this approach was used to impart water-solubility on polychlorinated trityl radical by Juliá
and coworkers. Mesa, J. A.; Velázquez-Palenzuela, A.; Brillas, E.; Torres, J. L.; Juliá, L. Tetrahedron 2011,
67, 3119.
33 Na SO solubility in DMSO is <0.1 g/L Vorob'ev, A. V.; Mustafin, D. I.; Tsivileva, O. M. Zh. Obshch.
2
4
Khim., 1989, 59, 2196. Gaylord Chemical Company, LLC. Slidell, LA, USA 2007, Bulletin #102B.
32
-38-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Scheme 8. Synthesis of Water-Soluble SA–BDPA Radicals
1) t-BuOK, DMF;
2) AgNO3
H
BDPA
BDPAH
1) H 2SO 4 (fuming);
2) NaOH
X
(SO3Na)ni
(SO3Na)ni
(SO3Na)ni
Na +
ni(NaO3S)
ni(NaO3S)
(SO3Na)ni
AgNO3
H 2O
ni(NaO3S)
ni(NaO3S)
(SO3Na)ni
(SO3Na)ni
SA-BDPA –
SA-BDPA
ni = i , i = 1, 2
ni = i , i = 1, 2
2.4.2 Properties of the water-soluble BDPA derivative
The HRMS analysis showed a mixture of compounds34 containing four to
seven sulfonyl groups. SA–BDPA is air-stable both in solution and as a solid. Unlike
BDPA, however, SA–BDPA does not partition into organic solvents and is soluble in
water in all proportions.
The shape of the UV−vis spectrum, the redox behavior, and the EPR
spectrum of SA–BDPA closely resemble those of BDPA. The sulfonate groups cause a
red shift of the absorption maxima of the carbanion and SA–BDPA radical ~20 nm
relative to those of BDPA anion and BDPA. As shown in Figure 4, SA–BDPA in water
has a strong absorption in the visible region (λmax = 508 nm, D0 → D2) and a weak
absorption in the near-IR region (λmax = 881 nm, D0 → D1). BDPA in dichloromethane
has similar absorption maxima (485 and 859 nm, respectively).23,27
34
Major molecular ion corresponded to 5 sulfonyl groups
-39-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.5
628 nm
508 nm
1.5
4
-1
-1
ε ( x10 M cm )
2.0
1.0
SA-BDPA SA-BDPA–
881 nm
0.5
0
400
500
600
700
800
900
1000
1100
Wavelength (nm)
Figure 4. Absorption spectra and (inset) photographs of aqueous solutions of SA–BDPA
and the corresponding carbanion.
Similar to BDPA15, SA–BDPA is reduced35 in aqueous NaOH to SA–
BDPA– in the presence of catalytic amounts of acetone, tetrahydrofuran (THF), dimethyl
sulfoxide, or ascorbic acid. As shown in Figure 5, aqueous reduction of SA–BDPA with
NaOH exhibited a clean isosbestic point, indicating a one to one conversion of the
radical to the anion. The mechanism of the reduction is not obvious. Screttas has
reported that Lewis-bases such as triethylamine and THF increased the rate of
reduction of BDPA by butyl lithium.36 Screttas posited that role of the THF is “to provide
substrate orbitals of correct symmetry to receive a delocalized electron from the
organometallic species”. In our case the organometallic species is replaced by NaOH
and the role Lewis bases play in the reduction of SA–BDPA is not clear. The SA–BDPA–
carbanion can be reoxidized to SA–BDPA chemically (e.g., by AgNO3) or
electrochemically.
35
SA–BDPA is more easily reduced than BDPA and required that all glassware and silica gel used for
purification was acid-washed prior to use.
36 Screttas, C. G. J. Chem. Soc., Chem. Commun. 1971, 406.
-40-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.0
1 min
3 min
5 min
9 min
λmax = 508 nm (radical)
1.8
543.5544.0
544.5545.0
0.43
1.6
0.42
1.4
0.41
OD
1.2
λmax =628 (carbanion)
1.0
0.8
0.6
0.4
881 nm (radical)
0.2
0
400
500
600
700
800
900
1000
Wavelength [nm]
Figure 5. Reduction of SA–BDPA radical with NaOH in water.
The EPR spectrum of SA–BDPA (Figure 6) shows no evidence of
aggregate formation or radical dimerization in liquid or frozen solution. Frozen pulsed
EPR spectra were collected by Andy Smith in Griffin lab. The solution EPR spectrum
features a hyperfine pattern typical of BDPA-type radicals, with nine lines resulting from
coupling to eight protons with similar coupling constants of ∼5.1 MHz (Figure 6A).
A
B
351
352
353
C
346
347
348
4993
4994
4995
4996
Magnetic Field (mT)
Figure 6. EPR spectra of SA–BDPA: (A) 1 mM solution in water (9.856 GHz, rt, cw
EPR); (B) 1 mM frozen solution in 60/40 (v/v) glycerol-d8/D2O (9.745 GHz, 80 K, echodetected); (C) 1 mM frozen solution in 60/40 (v/v) glycerol-d8/D2O (140.0 GHz, 80 K,
echo-detected).
-41-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
The hyperfine coupling constant at H is directly proportional to the spin density at the
carbon to which it is attached, according to McConnell equation37 (eq 1)
a(H) = QρC
(1)
A comparison of the hyperfine splitting in the liquid EPR spectra of SA–BDPA (in water)
and BDPA (in toluene) shown in Figure 7 gives some information about the extent and
regiochemistry of the sulfonation. Based on spin density calculations and simulations,
18,38
the large hyperfine couplings were previously assigned to the hydrogens at C1 and
C3, and the corresponding symmetry-related hydrogens in BDPA. Based on the
calculated spin densities, the smaller hyperfine splitting observed for BDPA (Figure 7
right), can be attributed to the hydrogen attached to C2 and the three symmetry-related
hydrogens (Figure 7, blue dots). Importantly, the absence of this smaller hyperfine
splitting in the EPR spectrum of SA–BDPA is consistent with C2 and the three
symmetry-related carbon atoms being sulfonated. This regiochemistry for sulfonation of
BDPA is analogous to that observed for the sulfonation of fluorene. Preferential
sulfonation at C2 and the three symmetry-related carbons on the fluorenyl blades is
additionally supported by results on the chlorosulfonation of BDPA (vide infra). As
described above, mass spectral analysis showed the major SA–BDPA component
contained five sulfonyl groups. It is plausible that the fifth sulfonyl group is located at the
para position39 of the phenyl ring.
In frozen solution, the linewidth [D = 26 MHz at 9 GHz, full width at halfmaximum (fwhm)] is dominated by unresolved hyperfine couplings to protons and
perfectly matches the envelope of the solution spectrum (Figure 6B). At 140 GHz, the
line broadens insignificantly to D = 28 MHz as a result of a very small g anisotropy
(Figure 6C). This small g anisotropy qualifies SA–BDPA as an interesting new polarizing
agent for SE DNP at even higher fields (e.g., 9.4 T). Trityl’s linewidth, on the other hand,
37
Wertz, J. E. and Bolton, J. R. Electron Spin Resonance, Elementary Theory and Practical Applications;
Chapman and Hall: New York, 1986.
38 Watanabe, K.; Yamauchi, J.; Ohya-Nishiguchi, H.; Deguchi, Y.; Ishizu, K. Bull. Inst. Chem. Res., Kyoto
Univ., 1975, 53,161. Can, T. V.; Caporini, M. A.; Mentink-Vigier, F.; Corzilius, B.; Walish, J. J.; Rosay, M.;
Maas, W. E.; Baldus, M.; Vega, S.; Swager, T. M.; Griffin, R. G. J. Chem. Phys. 2014, 141, 064202.
39 Cerfontain, H.; Schaasberg-Nienhuis, Z. R. H. J. Chem. Soc. Perkin II 1974, 536.
-42-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
increases linearly with the external field because g anisotropy is the dominant
broadening mechanism in that case.40
C1
C2
C4
C3
BDPA
Figure 7. X-Band, rt, cw EPR spectra of 50 μM aqueous SA–BDPA (left) and 50μM
BDPA in toluene41 (right) showing ~1 MHz hyperfine coupling present in BDPA is
removed by sulfonation at C2 and symmetry related (blue) carbons.
Encouraged by SA–BDPA’s desirable properties, we proceeded to test it
as a polarization agent for SE DNP.
40
Hu, K.-N.; Bajaj, V. S.; Rosay, M.; Griffin, R. G. J. Chem. Phys. 2007, 126, No. 044512.
Figure taken from Can, T. V.; Caporini, M. A.; Mentink-Vigier, F.; Corzilius, B.; Walish, J. J.; Rosay, M.;
Maas, W. E.; Baldus, M.; Vega, S.; Swager, T. M.; Griffin, R. G. J. Chem. Phys. 2014, 141, 064202.
41
-43-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.5 DNP MAS SSNMR results with SA–BDPA
DNP MAS SSNMR experiments were carried out by Björn Corzilius in Griffin lab. Figure
8 shows the field-dependent DNP enhancement at 5 T (140 GHz microwave frequency)
obtained with a 40 mM frozen solution of SA–BDPA in 60/30/10 (v/v) glycerol-d8/D2O/
H2O compared with that of the commonly used trityl polarizing agent OX063. The glassforming glycerol/water mixture is crucial because it prevents phase separation of the
solutes and allows for efficient nuclear spin diffusion. Furthermore, it acts as a
cryoprotectant to prevent potential cold denaturation of proteins. 1H signal enhancement
by DNP was directly observed via a Bloch decay as function of the external magnetic
field. The two radicals gave frequency profiles typical of the well-resolved SE, with a
positive peak and a negative peak separated by twice the Larmor frequency of the
polarized nucleus and centered around the EPR resonance. As expected, the smaller
linewidth of SA–BDPA was retained in the DNP field profile. The enhancement factors
were determined with a Hartmann−Hahn cross-polarization (CP) step to 13C at the
respective fields of maximum enhancement (1 M [13C]urea was added to provide
sufficient 13C for detection of thermal equilibrium polarization). SA–BDPA yielded an
NMR signal enhancement (ε) of 61 by comparison of the on and off signals after a time
of 1.3 TB (where TB is the time constant of polarization buildup) at a microwave power of
6 W; this is ~30% higher than the enhancement obtained with OX063 under the same
conditions.42
42
Corzilius, B.; Smith, A. A.; Griffin, R. G. J. Chem. Phys. 2012, 137, No. 054201.
-44-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
60
SA-BDPA
trityl (OX063)
20
H Enhancement
0
-20
1
H Enhancement
40
1
-40
-60
4970
100
80
60
40
20
0
0
5
10
15
Microwave Power (W)
4975
4980
4985
4990
4995
Magnetic Field (mT)
Figure 8. Field-dependent 1H DNP enhancement by 40 mM SA–BDPA (red) in 60/30/10
(v/v) glycerol-d8/D2O/H2O compared with that of trityl OX063 (green) recorded at a
microwave power of ~6 W. Inset: power dependence of the 1H enhancements measured
at the respective field maxima by 1H−13C CP. The trityl OX063 comparison data
(recorded under similar conditions) were taken from ref. 45.
Previous studies of the solid effect have shown that the polarization
enhancement is accompanied by a decrease in the time constant at which nuclear
longitudinal polarization builds up.45 Therefore, careful analysis of the buildup dynamics
and extrapolation of the signal intensity at infinite polarization time are crucial to prevent
misinterpretation of data (Figure 9). We report enhancement values extrapolated to
infinite polarization time.
-45-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
13.4 W
1.0
Relative 1H Polarization
ε = 94; TB = 31 s
0.8
8.9 W
ε = 72; TB = 34 s
0.6
3.6 W
0.4
ε = 42; TB = 36 s
0.2
T1 = 43 s
0.0
0
50
100
150
200
mw off
×10
250
Polarization Time (s)
Figure 9. Polarization buildup curves obtained with 40 mM SA–BDPA in 60/30/10 (v/v)
glycerol-d8/D2O/H2O at different microwave power levels and without microwave
irradiation (multiplied by a factor of 10 to enhance visibility).
The buildup time constants TB were found to be between 43 and 31 s depending on the
incident microwave power and obtained enhancements (Figure 9). The reduction of TB
by SE DNP compared with T1 leads to a further increase in sensitivity as a result of
more rapid recycling of NMR experiments.43 At the highest microwave power, the
enhancement factor was ε = 94, while the gain in sensitivity was ε(T1/TB)1/2 = 110.
However, the overall TB values (including T1) were ~50% larger in comparison with
those of trityl OX06345 under similar conditions, which is attributed to less efficient
longitudinal relaxation enhancement of protons by the paramagnetic species.
In order to assess potential electron spin saturation during SE DNP experiments,
electron spin-lattice relaxation time constants of frozen 1 mM SA–BDPA and trityl
OX063 solutions in 60/40 (v/v) glycerol-d8/D2O have been measured using saturation
recovery at 140 GHz. T1e measurements were performed by Andy Smith in Griffin lab.
After a sufficiently long presaturation pulse the electron polarization is read out via a
Hahn echo sequence. Experiments are performed at the magnetic field corresponding
43
Smith, A. A.; Corzilius, B.; Barnes, A. B.; Maly, T.; Griffin, R. G. J. Chem. Phys. 2012, 136, No. 015101.
-46-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
to the respective maximum of the EPR spectrum. Relaxation curves are shown in
Figure 10.
Figure 10. Relaxation curves of SA–BDPA (left) and trityl OX063 (right). Open circles
represent experimental data while the solid line is a monophasic exponential fit.
Monophasic exponential fitting using equation 1 yielded relaxation time constants T1 of
55.9 and 1.28 ms, respectively.
(1)
The significantly slower spin−lattice relaxation of the SA–BDPA electron
spin is consistent with the less efficient longitudinal relaxation enhancement of protons
by the paramagnetic species. This longer spin−lattice relaxation time constant might
also lead to a “saturation effect” observed in microwave-power-dependent
measurements of the 1H DNP enhancement: while trityl showed a near-linear power
dependence, the enhancements obtained with SA–BDPA were more than 50% larger at
the lowest applied power but approached those obtained with trityl at the highest
gyrotron output power available (Figure 8 inset).
Having identified the necessary DNP parameters, SA–BDPA was then
used in 1D DNP-enhanced CPMAS and 2D 13C−13C correlation spectra of a 0.1 M
solution of uniformly 13C-labeled proline.44 13C−13C mixing was achieved using spin
44
These measurements were performed by Björn Corzilius in Griffin lab
-47-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
diffusion assisted by a dipolar-assisted rotational resonance (DARR) field applied to
1H.45
Mixing was either limited to a one-bond distance (Figure 11, green) or allowed to
occur among all of the nuclei in the small molecule (Figure 11, red) by allowing spin
diffusion for a mixing period of 2 or 20 ms, respectively. In each case all of the expected
cross-peaks were present and clearly resolved. The 1H signal enhancement was
determined to be ε = 50 by comparison of the 1D signal amplitudes with and without
microwave irradiation (Figure 11, right). The slight decrease in the DNP enhancement
compared with the experiments using urea could be caused by less efficient irradiation
of the entire sample volume (a fully packed NMR rotor was used for proline, while the
urea experiments were conducted in a center-packed rotor). Additionally, different 1H
relaxation properties induced by alicyclic side chain dynamics could lead to lower
13
C Chemical Shift (ppm)
equilibrium polarization.
CH2OH/α
CO
30
δ βγ
CHOH
40
mw on, 8 scans
50
60
mw off
768 scans ×50
70
2 ms
20 ms
80
mw off, 8 scans ×10
180
mw off, 8 scans
190
180
13
80
60
40
180
C Chemical Shift (ppm)
140
13
100
60
20
C Chemical Shift (ppm)
Figure 11. (left) DNP-enhanced 2D 13C−13C correlation spectrum (spin diffusion) using
mixing times of 2 ms (green) and 20 ms (red) and (right) 13C CPMAS spectra of 0.1 M
[U-13C5]proline polarized by 40 mM SA–BDPA in 60/30/10 (v/v) glycerol-d8/D2O/H2O
under 8.9 W microwave irradiation.
45
Takegoshi, K.; Nakamura, S.; Terao, T. Chem. Phys. Lett. 2001, 344, 631.
-48-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
In addition to the signals attributed to the SE, another feature centered at
the EPR resonance field of 4983 mT was observed for SA–BDPA. Experiments have
shown a 1H enhancement of ~8 independent of the applied microwave power over the
range of 2.4−10 W. This flat power dependence together with the symmetric shape and
increased width of this feature led us to conclude that the underlying mechanism is
solely based on direct saturation of the EPR resonance and that nuclear polarization is
induced via cross-relaxation, similar to the Overhauser effect. It is unlikely that the cross
effect (CE) or thermal mixing (TM) caused this peak because both mechanisms require
a much larger EPR linewidth and typically yield a DNP field profile with regions of
positive and negative enhancements and an overall width comparable to the EPR
linewidth. For direct comparison, an overlay of the DNP field profile and the EPR
spectrum of SA–BDPA is shown in Figure 12. In fact, we attributed the feature in the
center of the trityl profile to CE/TM.46
Figure 12. Comparison between DNP field profile (red) and EPR spectrum (blue) of
SA–BDPA.46
46
The difference in applied microwave frequency between the MAS DNP experiment (139.65 GHz) and
the EPR experiment (140.0 GHz) has been corrected for by shifting the EPR spectrum to the resonance
field under MAS DNP conditions.
-49-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
However, the SA–BDPA profile does not show any sign of CE/TM, in accordance with
the significantly reduced linewidth of SA–BDPA compared with trityl (D = 28 MHz vs 50
MHz). Further investigation41 using the same sample of SA–BDPA concluded that the
feature is in fact a result of the Overhauser effect DNP. Overhauser effect was not
predicted in insulating solids, however, it has been observed and identified previously47
in BDPA-doped crystalline carbazole. The origin of the effect was traced to the
fluctuating hyperfine interactions, and the magnitude of the hyperfine couplings is
proportional to the polarization enhancements.
In addition to assessing the practical applications of SA–BDPA (to
increase signal to noise) in SSNMR experiments, it was also used to further the study of
the SE DNP mechanism. The results48 suggest that in addition to the standard solid
effect condition,49 higher order transitions involving up to 4 nuclei may play a major role
in polarization transfer via DNP. The higher order transitions appear to accelerate the
polarization of the nearby nuclei but this polarization is not transferred to the bulk nuclei.
This finding led to the conclusion that SE DNP performance can be improved by
reducing the number of 1H near the electron, while maintaining 1H concentration in the
bulk sample. Design of future polarization agents is guided by these findings.
47
Hu, J. Z.; Solum, M. S.; Wind, R. A.; Nilsson, B. L.; Peterson, M. A.; Pugmire, R. J.; Grant, D. M. J.
Phys. Chem. A 2000, 104, 4413.
48 Smith, A. A.; Corzilius, B.; Haze, O.; Swager, T. M.; Griffin, R. G. J. Chem. Phys. 2013, 139, 214201.
49 As described in Chapter 2.2, SE is usually described as a two-spin mechanism involving one electron
and one nucleus
-50-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.6 Synthesis and properties sulfonated BDPA derivatives
Although SA–BDPA containing mainly the quintuply sulfonylated BDPA performed well
in the DNP experiments, a single component material rather than a mixture was
desired. Alternative work up and purification conditions were investigated. Longer
sulfonation reaction time was used to obtain a higher fraction of fully sulfonylated
product (7-SO3H). The reaction mixture was then diluted with water and treated with
aqueous barium hydroxide. Surprisingly, SA–BDPA– barium salt stayed in solution and
the fine barium sulfate precipitate was filtered out or removed by centrifugation. Ion
exchange resin was used to convert the barium salt of the SA–BDPA to the sodium50
salt and to remove excess Ba(OH)2. The oxidation can also be achieved with cheaper
and easier to remove PbO2. Purification conditions were also optimized to allow the
separation of the BDPA(SO3Na)7, although the yield was reduced.
SA–BDPA is an effective narrow-line DNP agent with outstanding water
solubility. However, access to a family of peripherally substituted BDPA derivatives
would be advantageous because experience has indicated that a single polarization
agent may not suffice for all analytes.51 For example, DNP NMR experiments may be
compromised if the polarizing agent binds to the analyte, which would cause significant
paramagnetic broadening of the NMR signals of interest. Therefore, we set out to
develop a modular route to access differentially functionalized water-soluble BDPA
derivatives. Because of our previous success with sulfonation, we chose chlorosulfonic
acid52 for our modular approach.
Aromatic hydrocarbons react with excess chlorosulfonic acid to produce
sulfonyl chlorides which are easier to handle and are much less prone to hydrolysis than
carboxylic acid chlorides, but still react with nucleophiles. Reactions of fluorene with
50
This was mainly done to reduce molecular weight of the product and to avoid further exposure to
soluble barium2+.
51 Blazina, D.; Reynolds, S.; Slade, R. Hypersense Application Note: Influence of Trityl Radical on the
DNP Process; Oxford Instruments Molecular Biotools, Ltd: Oxon, U.K., 2006.
52 Prabhakaran, P. C.; Sharma, A.; Gorman, C. Chlorosulfonic Acid; John Wiley & Sons, Ltd: Chichester,
UK, 2001.
-51-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
chlorosulfonic acid were reported and the regiochemistry of the sulfonation is similar to
that with sulfuric acid.53
The reaction of HOOC–BDPAH54 with chlorosulfonic acid (Scheme 8) produced
chlorosulfonate 3, which could be hydrolyzed or reacted with nucleophiles prior to
oxidation to give both ionic and neutral polarization agents.
Scheme 8. Chlorosulfonation of HOOC-BDPA
SO2Cl
COOH
COOH
ClSO 3H
H
ClO2S
H
ClO2S
SO2Cl
3
HOOC-BDPAH
BDPAH can also be functionalized under these conditions, however the
carboxylic acid derivative was chosen because it eliminated any ambiguity over the
regiochemistry of sulfonation on the phenyl ring. Carboxylic acid deactivates the ring,
and the sulfonation proceeds only on the fluorenyl rings. At room temperature
chlorosulfonation proceeds stepwise and stops at 4-SO2Cl as observed by ESI– MS
53
Bassin, J. P.; Cremlyn, R. J.; Swinbourne, F. J. Phosphorus Sulfur 1992, 72, 157.
Mi, Q., Ph.D. Thesis, Northwestern University, Evanston, IL, 2009. Dane, E. L.; Maly, T.; Debelouchina,
G. T.; Griffin, R. G.; Swager, T. M. Org. Lett. 2009, 11, 1871.
54
-52-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
R
SO2Cl
ClO2S
H
ClO2S
a) H 2O, reflux;
NaOH,
PbO 2, H 2O;
COOH
IRP64–Na +
or
b) NH 4OH, H 2O, rt;
PbO 2
SO2Cl
or
c) diethanolamine, THF;
AgNO3, H 2O;
R'
RR
R
6a R=SO 3Na, R'=COONa
6b R=SO 2NH 2, R'=COOH
6c R=SO 2N(CH 2CH2OH)2, R'=COOH
x
6b
x
6a
x
6c
5G
Figure 13. Modular approach to water-soluble BDPA-based radicals. X-Band solution
EPR spectra of the corresponding radicals maintain the hyperfine structure of BDPA
core (6a and 6c in water, 6b in pyruvic acid).
Treatment of 3 with sodium hydroxide in water followed by PbO2 oxidation
and cation exchange gave 6a, a highly water-soluble ionic radical similar to SA–BDPA
(Figure 13). Replacing NaOH with ammonium hydroxide and diethanolamine gave
neutral radicals CN-BDPA 6b and SN-BDPA 6c, respectively. Liquid state EPR spectra
of these radicals are almost identical, as expected. However, 6c formed aggregates in
glycerol/water mixtures, as indicated by a prominent shoulder in it’s EPR spectrum at 80
K (Figure 14A).
-53-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Figure 14. A: 140 GHz field-swept electron-spin echo detected EPR spectrum of a
frozen 40 mM solution of SN-BDPA (6c) in glycerol-d8/D2O/H2O (60/30/10 v/v/v). B:
Nutation curve acquired by varying the length of the flip-pulse in the Hahn echo
sequence. Nutation curves are shown for two field positions, indicated by arrows in A.
C: Distribution of nutation frequencies obtained by Fourier transformation of curves
shown in B.
Electron spin transient nutation experiments can be used to determine spin multiplicity
in solid state EPR.55 In our case, the SN-BDPA radical is spin 1/2 and aggregates will
have larger spin quantum numbers. Nutation curves (obtained by pulse length variation)
were recorded by Andy Smith at Griffin lab at magnetic fields corresponding to the
center of the spectrum and to the distinct shoulder (satellite) of the central line. Nutation
frequencies of 3.5 and 4.8 MHz are obtained for the center and the satellite,
respectively. The ratio of these frequencies is ~0.7, corresponding to the expected ratio
between an S = 1/2 and an S = 1 spin system. The absence of the 4.8 MHz peak in the
frequency distribution of the central line is explained by a collapse of the high-spin
properties under the strong microwave field for vanishing electron-electron interactions.
55
Astashkin, A.V.; Schweiger, A. Chem. Phys. Lett., 1990, 174, 595.
-54-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Preliminary SE DNP experiments with SN-BDPA 6c showed very low
enhancements likely caused by aggregation. The spin 1 dimer reduces the
concentration of the spin 1/2 species and causes faster relaxation of protons. The cause
of aggregation is likely the carboxylic acid dimerization, and carboxylate analogues may
perform better. Also, CN-BDPA and SN-BDPA may be used to polarize pyruvate via
dissolution DNP in metabolic imaging.56
2.7 Summary
Air-stable, highly water-soluble organic radicals containing a 1,3-bis(diphenylene)-2phenylallyl (BDPA) core were synthesized. A sulfonated derivative, SA–BDPA, retains
the narrow electron paramagnetic resonance linewidth (<30 MHz at 5 T) of the parent
BDPA in highly concentrated glycerol/water solutions (40 mM), which enables its use as
polarizing agent for solid effect dynamic nuclear polarization (SE DNP). A sensitivity
enhancement of 110 was obtained in high-field magic-angle-spinning (MAS) NMR
experiments. The ease of synthesis and high maximum enhancements obtained with
the BDPA-based radicals constitute a major advance over the trityl-type narrow-line
polarization agents.
56
Kurhanewicz, J.; Bok, R.; Nelson, S. J.; Vigneron, D. B. J. Nucl. Med. 2008, 49, 341.
-55-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
2.8 Experimental
Materials. All chemicals, reagents, and solvents were used as received from
commercial sources (Aldrich, Acros, Fluka) without further purification. Ion exchange
resins were washed with MeOH and deionized water prior to use. C18-reverse phase
column chromatography gel was washed with 1% acetic acid in ethanol prior to use.
Any glassware that was washed in a base bath was rinsed with 10 % HCl followed by
deionized water and acetone prior to oven drying. (This is necessary as some of the
radicals described here are reduced under basic environments.)
Instrumentation. Proton nuclear magnetic resonance (1H NMR) spectra and carbon
nuclear magnetic resonance (13C NMR) spectra were recorded on a Bruker Avance-400
(400 MHz) NMR spectrometer. Chemical shifts for protons are reported in parts per
million downfield from tetramethylsilane and are referenced to residual protium in the
NMR solvent. Chemical shifts for carbon are reported in parts per million downfield from
tetramethylsilane and are referenced to the carbon resonances of the solvent. The
mass spectrometry data were obtained using Waters LCT Premier TOF equipped with
an ESI probe. IR spectra were recorded using Nicolet 6700 FT-IR (Thermo Scientific)
equipped with ATR probe (Ge window). Uv/Vis spectra were collected on Agilent 8453
spectrophotometer.
EPR. 9 GHz EPR experiments (rt) were performed on Bruker EMX spectrometer,
equipped with an ER 4199HS cavity and a Gunn diode microwave source operating at
9.8565 GHz. Aqueous samples were measured using Wilmad WG-808-Q Suprasil Low
Temperature Aqueous Cell. A microwave power of 5 mW was used in conjunction with
100 kHz field modulation with 0.1 mT amplitude. 1024 field points were recorded over a
width of 5 mT; the conversion time was 20.48 ms with an integration time constant of
5.12 ms. The spectrum was recorded in a single sweep.
The 9 GHz EPR (frozen) field profile was performed by Andy Smith at Griffin lab on a
Bruker Elexsys E580 pulsed-EPR spectrometer, equipped with a MD-5-W1 probe with
TE011 mode cavity operating at 9.745 GHz. The field profile was acquired at 80 K using
-56-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
a Hahn echo pulse sequence using 16 ns π/2- and π-pulses, with a 100 ns delay
between pulses. 80 shots were acquired at 1024 field points, without phase cycling.
140 GHz EPR experiments were performed on a custom-built 140 GHz EPR
spectrometer at 80 K.57 ~300 nL of sample was contained in a 0.40 mm inner diameter/
0.55 outer diameter quart capillary. The spectrum in Figure 6, C is obtained with a Hahn
echo using π/2 and π-pulse lengths of 30 ns and 60 ns, and a pulse separation of 200
ns. 200 transients were recorded at 131 field points, using a 4-step phase cycle. The
spectrum in Figure 14, A was acquired with the same pulse sequence, using 400
transients at 201 field points, and a 2-step phase cycle. Spin-lattice relaxation times of
SA–BDPA and trityl (OX063) in Figure 10 were obtained by applying a 3 ms saturating
pulse at the center of the EPR spectrum. After a recovery delay, a Hahn echo was used
to measure the polarization recovery. The length of the recovery delay is varied to
determine the spin-lattice relaxation time. π/2 and π pulses were 30 ns and 60 ns for
SA–BDPA, and 35 ns and 70 ns for trityl. The pulse separation was 200 ns in both
cases. 100 transients were recorded at each point for SA–BDPA, and 200 transients at
each point for trityl, using a 4-step phase cycle in both cases. Nutation curves in Figure
14, B were obtained using a pulse with variable length, followed by a 200 ns delay and
π-pulse of 60 ns. 400 transients were recorded at each point, with a 2-step phase cycle.
DNP experimental conditions. DNP experiments were performed by Björn Corzilius in
Griffin lab using a custom-built instrument, consisting of a 139.65 GHz gyrotron58 and a
212 MHz (1H frequency) MAS NMR spectrometer (courtesy of D. Ruben). The
spectrometer utilizes a custom-built cryogenic MAS DNP probe with triple-channel (1H,
13C, 15N)
rf resonant circuit and 4 mm rotor diameter Revolution NMR stator.
Microwaves are guided to the sample via circular overmoded waveguides whose inner
volume has been corrugated to reduce mode conversion from HE11 mode into higher
order modes and to reduce ohmic losses.
57
Smith, A. A.; Corzilius, B.; Bryant, J. A.; DeRocher, R.; Woskov, P. P.; Temkin, R. J.; Griffin, R. G. J.
Magn. Reson. 2012, submitted.
58 Becerra, L. R.; Gerfen, G. J. ; Temkin, R. J.; Singel, D. J.; Griffin, R. G. Phys. Rev. Lett. 1993, 71, 3561.
-57-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
The sample is contained within sapphire rotors (Inasco, Inc) with 0.7 mm wall thickness
within a volume of ~4 mm height. Samples were prepared by dissolving radicals in a
60/30/10 (v/v) glycerol-d8/D2O/H2O mixture to give a concentration of 10 mM. For
standard experiments 1 M 13C-urea was added to provide 13C for NMR detection.
Further experiments were performed using 0.1 mM 13C5-proline; in this case the sample
was fully packed in a rotor, so that the entire active NMR volume was used.
Experiments were performed under MAS with a spinning frequency of ωr/2π = 5 kHz at
a temperature of ~82 K, measured directly outside the MAS stator. The field dependent
DNP enhancement profile was recorded by detecting the 1H FID intensity during a Bloch
decay while sweeping the external magnetic field value. At each field point the
spectrometer frequency was re-set, and the probe circuit was retuned. The
enhancement values were subsequently scaled to the maximum ε obtained via crosspolarization to 13C because a significant non-enhanceable 1H background signal caused
by the stator material, spacers, etc., impedes any direct measurement of ε by
comparison of 1H signal intensities recorded with and without microwave irradiation. All
one-dimensional spectra were recorded using a 16-pulse presaturation train (phase
alternating in +x and +y) of 108° flip angle and γB1/2π = 50 kHz separated by 5 ms each
on both 1H and 13C channels. 1H polarization was allowed to build up during a variable
recovery time and polarization was then read out by a Bloch decay or cross-polarization
to 13C. For all pulses (incl. CP and TPPM decoupling) γB1/2π = 100 kHz has been used.
Build-up curves have been obtained at the field of maximum ε by varying the recovery
time between presaturation and detection, and enhancement values were obtained after
mono-exponentially fitting by dividing the pre-exponential factor of the on-signal by that
of the off-signal.
-58-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Synthesis
BDPAH (1) was prepared according to the published procedures.59
SA–BDPA (3): 100 mg (0.239 mmol, 1 equiv) of BDPAH (1) and 1 mL fuming sulfuric
acid were heated at 85 ºC while stirring for 5 min. The resulting brown solution was
carefully diluted with 20 mL of deionized water. The resulting salmon–colored solution
was stirred at room temperature for 16 h, and then filtered to remove any insoluble
material. The filtrate was treated with 16 mL of 10% NaOH and concentrated. The
resulting deep-blue powder was extracted with 125 mL of DMSO. The resulting sodium
salt SA–BDPA– (2) (λmax= 628 nm) was precipitated from the DMSO solution with 75 mL
of benzene, and washed with DCM. The dark blue powder was dissolved in 3 mL of
water and 102 mg (0.598 mmol, 2.5 equiv) of AgNO3 in 2 mL of water was added while
stirring. The resulting red mixture stirred at room temperature for 3 h and then was
filtered and concentrated to give 200 mg red powder. This was chromatographed on
C18 reverse phase column eluting with 5% H2O-CH3CN to give 189 mg (85%) of deepred glassy solid. EPR (9 GHz, rt, water) g=2.0027. UV–Vis (water): λmax = 508 (13000),
881(900) nm (ε). This material (a mixture of sulfonation products BDPA(SO3Na)4-7) was
used in DNP experiments.
Alternatively, 0.53 g (1.27 mmol, 1 equiv) of BDPAH (1) and 1 mL fuming sulfuric
acid were heated at 85 °C while stirring for 6 h. The resulting black solution was
carefully diluted with 10 mL of deionized water, and the pH of resulting salmon-colored
59
Plater, M.; Kemp, S.; Lattmann, E. J. Chem. Soc. Perkin Trans. 1 2000, 971–979.
-59-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
solution was brought to 9 by addition of aqueous Ba(OH)2. BaSO4 was filtered out and
the blue filtrate was treated with DOWEX HRC-W2 Na+ form. The resulting violet-blue
solution was shaken with 5 g of PbO2 until the color changed to deep red (ca 30 min),
and then filtered onto Amberlite IRC-50 (–COOH form). The ion exchange resin was
separated, and the deep red solution was adsorbed onto RP C18 silica gel in vacuo at
50 °C. This was dry-loaded onto C18 reverse phase column and eluted with 200 proof
ethanol followed by 20% H2O-EtOH to give 400 mg (28%) of a deep red glassy solid
(BDPA(SO3Na)7). IR (ATR, Ge) 3449, 1640, 1390, 1184, 1112, and 1041 cm-1. ESI– was
obtained after ion exchange60 on DOWEX HRC-W2 (H+ form) ESI HRMS (m/z) [M–
2H]2– calcd for C33H21O21S7• 487.4237, found 487.4225. Further elution with 40% H2OEtOH yields additional material, which consists of a mixture of sulfonation products
BDPA(SO3Na)4-7.
Reaction time and temperature may be adjusted to increase or reduce the amount of
sulfonation. The sulfonation can be carried out at rt over 3 days to produce completely
water-soluble product mixture.
60
MS of polysulfonates is close to impossible without ion exchange. Sometimes HPLC can help, but not
in our case. Holcapek, M.; Volna, K.; Jandera, P.; Kolarova, L.; Lemr, K.; Exner, M.; Cirkva, A. J. Mass.
Spectrom. 2004, 39, 43. Holcapek, M.; Jandera, P.; Prikryl, J. Dyes Pigm. 1999, 43, 127.
-60-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
HOOC-BDPAH (4) was prepared according to the published procedures.54
SO2Cl
COOH
ClSO3H
H
COOH
ClO2S
H
ClO2S
SO2Cl
HOOC-BDPAH(SO2Cl)4 (5): 0.50 g (1.08 mmol, 1 equiv) of finely powdered HOOC–
BDPAH (4) was stirred with 2.0 mL of chlorosulfonic acid in a sealed pressure-release
vial at room temperature for 24 h. The resulting deep-green solution was slowly poured
over 50 g of ice. Ice was allowed to melt, and the mixture was filtered, washing with
deionized water. The resulting pale pink powder was dried at rt, 100 mTorr and
precipitated from benzene-THF to give 0.68 g (74%) of 5 as a rosy powder. IR (ATR,
Ge) 3420, 3077, 1719, 1602, 1405, 1373, 1344, 1166, 1071 and 1004 cm-1; 1H NMR
(400 MHz, THF-d8) δ: 9.51 (s, br, 1H), 9.22 (d, J = 1.5 Hz, 1H), 8.55 (d, 8.3 Hz, 1H),
8.53 (s, 2H), 8.42 (dd, J = 8.3, 1.5 Hz, 1H), 8.39 (d, J = 8.3 Hz, 1H), 8.29 (d, J = 8.2 Hz,
2H), 8.25 (dd, J = 8.2, 1.5 Hz, 2H), 8.10 (dd, J = 8.3, 1.7 Hz, 1H), 7.81 (d, J = 8.3 Hz,
2H), 6.95 (s, 1H), 6.88 (d, J = 8.3 Hz, 2H), 6.53 (d, J = 1.7 Hz, 1H); 13C NMR (100 MHz,
THF-d8) δ: 166.2 (sp2, CO), 147.7, 147.1, 147.0, 145.7, 145.6, 145.4, 144.9, 144.1,
140.9, 140.5, 135.8, 132.3, 130.8, 129.0, 128.8, 128.5, 127.7, 125.7, 125.0, 124.3,
124.1, 123.9, 123.1 and 53.9 (sp3, CH); HRMS (ESI) [M–H]– calcd for C34H17Cl4O10S4,
852.846; found 852.837. -61-
8.55 (d, 8.3 Hz, 1H)
8.42 (dd, J = 8.3, 1.5 Hz, 1H)
ClO2S
ClO2S
6.95 (s, 1H)
9.22 (d, J = 1.5 Hz, 1H)
8.25 (dd, J = 8.2, 1.5 Hz, 2H)
8.29 (d, J = 8.2 Hz, 2H)
H
SO2Cl
SO2Cl
COOH
8.39 (d, J = 8.3 Hz, 1H)
8.10 (dd, J = 8.3, 1.7 Hz, 1H)
7.81 (d, J = 8.3 Hz, 2H)
6.88 (d, J = 8.3 Hz, 2H)
6.53 (d, J = 1.7 Hz, 1H)
9.51 (s, br, 1H)
8.53 (s, 2H)
in THF-d8
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
-62-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
SO3Na
SO2Cl
COOH
ClO2S
COONa
H
ClO2S
NaO3S
NaO3S
SO2Cl
5
SO3Na
6a
CA-BDPA (6a): 0.10 g (0.12 mmol, 1 equiv) of 5 was boiled in 20 mL of deionized water
for 15 minutes until a light red homogeneous solution was obtained. NaOH was then
added to pH 10, and the resulting deep blue solution (anion λmax= 611 (20500) nm (ε) )
was shaken with 0.3 g of PbO2 for 15 min (consumption of the anion was monitored by
UV/Vis) and filtered onto Amberlite IRP-64 (COOH form) ion exchange resin. The deep
red mixture (pH 6) was filtered and concentrated. Chromatography on RP C18 silica gel
eluting with 20 % H2O-CH3CN gave 0.09 g (84%) of deep red powder. UV-Vis (water):
λmax= 496 (13200), 865 (960) nm (ε); EPR (9 GHz, rt, water) g=2.0027.
-63-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
SO2NH 2
SO2Cl
COOH
COOH
ClO2S
NH 4OH, H 2O, rt;
PbO 2
H
ClO2S
H 2NO 2S
H 2NO 2S
SO2Cl
SO2NH 2
CN-BDPA (6b): 0.20 g (0.24 mmol, 1 equiv) of 5 was added to 0.5 mL of NH4OH
(28-30% NH3 basis) in 20 mL of deionized water and the resulting deep blue mixture
stirred for 1 h. Reaction progress was monitored by HRMS (ESI) [M]– calcd for
C34H25N4O10S4–, 777.0459; found 777.0444; The resulting deep blue solution was
shaken with 0.5 g of PbO2 for 15 min (consumption of the anion was monitored by UV/
Vis) and filtered onto Amberlite IRP-64 (COOH form) ion exchange resin. The deep red
mixture (pH 6) was filtered and concentrated. Chromatography on RP C18 silica gel
eluting with 20 % H2O-CH3CN gave 0.15 g (82%) of deep red powder. EPR (9 GHz, rt,
pyruvic acid) g=2.0027.
-64-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
SO2N(CH2CH2OH)2
SO2Cl
COOH
ClO2S
COOH
H
ClO2S
SO2Cl
(HOH2CH2C)2NO2S
(HOH2CH2C)2NO2S
5
SO2N(CH2CH2OH)2
6b
SN-BDPA (6c): Diethanolamine (0.5 mL, 5 mmol, 40 equiv) was added to the solution of
5 (0.108 g, 0.126 mmol, 1 equiv) in 12 mL of THF. The resulting deep blue solution
stirred at rt until deep blue oil separated at the bottom of the flask. The remaining clear
supernatant was decanted, the oil was washed with 20 mL THF, and then, diluted with
20 mL of nanopure water. Anion UV-Vis (water): λmax= 639(5000), 834(4000) nm (ε).
Silver nitrate (64.2 mg, 0.384 mmol, 3 equiv) in 3 mL of water was added to the above
solution, and the resulting red mixture stirred at rt for 0.5 h. The mixture was acidified to
pH 2.5 with 1 M HCl and filtered. The filtrate was washed with 0.1 M HCl, and dried to
give 0.136 g (96 %) of 6b as a deep red powder. UV-Vis (water): λmax= 514 (3300) nm
(ε); EPR (9 GHz, rt, water) g=2.0027; HRMS (ESI) [M–H]– calcd for C50H58N4O18S4•,
1129.2556; found 1129.2461.
-65-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Calculations
Spin density calculations were performed using MacSPARTAN '14 build 14.119p1.0.0
(Oct 31 2014) Input geometry was obtained from ZECZOW.cml from Cambridge
structural database. Truncated verbose output file is included below.--(Site specific preferences)
... THRESH
9
... SCF_CONVERGENCE 7
... SMALL_PROD_XCMAT 9
... BASIS_LIN_DEP_THRESH 5
... SCF_ALGORITHM
DIIS_GDM
... MAXSCF
100
... MAXDIIS
50
... THRESHDIIS
-1 (i.e. don't switch on delta-E)
... ONEEXE_SPAR TRUE
... GUI
GUI_SPARTAN
... TERSE_OUTPUT TRUE
Processing $rem in input file
... EXCHANGE
B3LYP
... CORRELATION
none (built-in)
... BASIS
6-31G(d)
... UNRESTRICTED
TRUE (setting default UHF)
... VARTHRESH 2 (default DFT)
... INCDFT
TRUE (default DFT)
... GUI
GUI_SPARTAN
... TERSE_OUTPUT
TRUE
... SCF_GUESS
READ
Writing REM_CC_EA
0
---------------------------------------------------Standard Nuclear Orientation (Angstroms)
I Atom
X
Y
Z
---------------------------------------------------1
C
1.219733 -0.102799 0.120673
2
C
-0.012028 0.567142 0.056798
3
C
-1.270121 -0.050084 -0.038560
4
C
3.523454 -0.524196 -0.006156
5
C
2.526951 0.417458 -0.328897
6
C
2.885442 1.560395 -1.051034
7
C
4.208145 1.739901 -1.410462
8
C
5.178930 0.826206 -1.056073
9
C
4.846351 -0.321216 -0.361856
10
C
2.879965 -1.650252 0.674031
11
C
1.498001 -1.405764 0.732917
12
C
0.687235 -2.325255 1.398457
-66-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
13
C
1.258038 -3.460189 1.950545
14
C
2.608435 -3.698375 1.840670
15
C
3.441477 -2.790626 1.209577
16
C
-3.587799 -0.388197 0.025575
17
C
-2.566500 0.500743 0.391686
18
C
-2.886217 1.646491 1.128215
19
C
-4.219406 1.890799 1.435824
20
C
-5.215314 1.030299 1.039008
21
C
-4.913483 -0.124203 0.331379
22
C
-2.987040 -1.520419 -0.685411
23
C
-1.597368 -1.319839 -0.703761
24
C
-0.793491 -2.226121 -1.393157
25
C
-1.376160 -3.324635 -1.991081
26
C
-2.744421 -3.523965 -1.935410
27
C
-3.566302 -2.612504 -1.299425
28
C
0.037633 2.060294 0.050685
29
C
0.769354 2.741018 0.995560
30
C
0.903099 4.123519 0.941336
31
C
0.281530 4.807521 -0.064948
32
C
-0.478528 4.173086 -1.000914
33
C
-0.598429 2.787513 -0.937673
34
H
2.229684 2.202580 -1.292552
35
H
4.453803 2.508246 -1.912116
36
H
6.084301 0.986600 -1.291183
37
H
5.513007 -0.957424 -0.132439
38
H
-0.247389 -2.171885 1.473316
39
H
0.708793 -4.083460 2.411097
40
H
2.975423 -4.496196 2.202821
41
H
4.376821 -2.947044 1.148295
42
H
-2.205053 2.244863 1.412166
43
H
-4.446447 2.668907 1.930566
44
H
-6.119371 1.228023 1.250968
45
H
-5.601274 -0.722759 0.062702
46
H
0.144190 -2.089082 -1.449903
47
H
-0.829575 -3.953606 -2.448180
48
H
-3.122725 -4.297170 -2.338731
49
H
-4.507424 -2.734933 -1.284603
50
H
1.189459 2.258024 1.697965
51
H
1.418729 4.585438 1.593608
52
H
0.381713 5.749418 -0.113167
53
H
-0.917372 4.664912 -1.684561
54
H
-1.126297 2.334058 -1.584774
---------------------------------------------------Molecular Point Group
C1 NOp = 1
Largest Abelian Subgroup
C1 NOp = 1
Nuclear Repulsion Energy = 3090.4104410058 hartrees
-67-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
There are
110 alpha and
109 beta electrons
theFileMan(): MAXOPENFILES=206 MAX_SUB_FILE_NUM=16
Maximum size of a physical file is 2.0 GB, maximum size of a tmp-file is 32.0 GB
Requested basis set is 6-31G(d)
There are 174 shells and 537 basis functions
Total Memory Limit 2048 MB
Mega-Array Size
142 MB
MEM_STATIC part
140 MB
MacSPARTAN '14 Quantum Mechanics Program: (x86/Darwin) build 14.119
(3.1.Pw)
A cutoff of 1.0D-09 yielded 7153 shell pairs
There are 75352 function pairs
Evaluating contribution to one-electron hamiltonian from nuclear
Smallest overlap matrix eigenvalue = 2.37E-04
Multipole matrices computed through 2nd order
Scale SEOQF with 1.000000e-02/1.000000e-02/1.000000e-01
Standard Electronic Orientation quadrupole field applied
Nucleus-field energy = -0.0000000140 hartrees
An unrestricted hybrid HF-DFT SCF calculation will be performed using Pulay DIIS +
Geometric Direct Minimization Exchange: 0.2000 Hartree-Fock + 0.0800 Slater +
0.7200 Becke Correlation: 0.8100 LYP + 0.1900 VWN1RPA Using SG-1 standard
quadrature grid SCF converges when RMS gradient is below 1.0E-07 Exchange:
0.2000 Hartree-Fock + 0.0800 Slater + 0.7200 Becke Correlation: 0.8100 LYP + 0.1900
VWN1RPA Using SG-1 standard quadrature grid using static scheduler for incremental
DFT
--------------------------------------Cycle
Energy
DIIS Error
--------------------------------------1 -1269.5408503518
2.88E-03
2 -1269.7642699330
4.47E-04
3 -1269.7957152946
3.55E-04
4 -1269.8074641942
1.28E-04
5 -1269.8111568006
3.19E-05
6 -1269.8149571893
1.47E-05
7 -1269.8174143062
1.43E-05
8 -1269.8189851621
4.60E-06
9 -1269.8196109074
3.57E-06
10 -1269.8199246611
1.17E-06
11 -1269.8199473341
5.85E-07
12 -1269.8199540359
2.48E-07
13 -1269.8199549052
1.08E-07
14 -1269.8199550990
3.29E-08 Convergence criterion met
--------------------------------------<S^2> = 0.7868
SCF time: CPU 687.07 s wall 749.43 s
-68-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
Analysis of SCF Wavefunction
Ground-State Mulliken Net Atomic Charges
Atom
Charge (a.u.) Spin (a.u.)
-------------------------------------------------------1C
-0.025932
0.414700
2C
-0.009907
-0.215903
3C
-0.021642
0.422046
4C
0.050924
0.073509
5C
0.050321
-0.090678
6C
-0.150349
0.097923
7C
-0.110725
-0.042402
8C
-0.106124
0.093751
9C
-0.156472
-0.033758
10 C
0.040991
0.067823
11 C
0.066946
-0.086285
12 C
-0.157965
0.097672
13 C
-0.103165
-0.042155
14 C
-0.110713
0.091200
15 C
-0.153007
-0.034014
16 C
0.049090
0.072736
17 C
0.050464
-0.092112
18 C
-0.153325
0.103306
19 C
-0.109185
-0.043399
20 C
-0.105391
0.095117
21 C
-0.158933
-0.034178
22 C
0.052780
0.070796
23 C
0.049421
-0.087729
24 C
-0.151808
0.095531
25 C
-0.106578
-0.042343
26 C
-0.110591
0.092794
27 C
-0.154255
-0.034100
28 C
0.050215
0.038598
29 C
-0.141526
-0.015094
30 C
-0.098193
0.007723
31 C
-0.107353
-0.012494
32 C
-0.097447
0.007687
33 C
-0.139796
-0.015228
34 H
0.118498
-0.003975
35 H
0.104833
0.001358
36 H
0.105770
-0.004309
37 H
0.099422
0.001240
38 H
0.117092
-0.004112
39 H
0.106877
0.001340
40 H
0.105879
-0.004172
41 H
0.099512
0.001241
42 H
0.115536
-0.004176
-69-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
43 H
0.105560
0.001382
44 H
0.106271
-0.004374
45 H
0.099932
0.001256
46 H
0.118188
-0.004065
47 H
0.107035
0.001360
48 H
0.106492
-0.004243
49 H
0.099232
0.001232
50 H
0.113778
0.001023
51 H
0.113034
-0.000341
52 H
0.108537
0.000639
53 H
0.112884
-0.000347
54 H
0.114867
0.001001
-------------------------------------------------------Sum of atomic charges = 0.000000
Sum of spin charges = 1.000000
----------------------------------------------------------------Cartesian Multipole Moments
----------------------------------------------------------------Charge (ESU x 10^10)
0.0000
Dipole Moment (Debye)
X
0.0414
Y
0.7601
Z
-0.0668
Tot
0.7641
Quadrupole Moments (Debye-Ang)
XX -167.8044 XY
0.3537 YY -165.2505
XZ
-0.0209 YZ
-0.0893 ZZ -183.5166
Traceless Quadrupole Moments (Debye-Ang)
QXX
13.1583 QYY
20.8199 QZZ -33.9782
QXY
1.0610 QXZ
-0.0627 QYZ
-0.2678
Octapole Moments (Debye-Ang^2)
XXX
-3.6032 XXY -17.6058 XYY
5.4629
YYY
25.2380 XXZ
-2.5061 XYZ -19.9828
YYZ
-1.6144 XZZ
-0.3492 YZZ
3.1811
ZZZ
-0.5272
Traceless Octapole Moments (Debye-Ang^2)
XXX -67.6423 YYY 281.2497 ZZZ
33.9216
XXY -296.5265 XXZ -23.6480 XYY
77.4117
XYZ -299.7421 XZZ
-9.7694 YYZ -10.2735
YZZ
15.2768
Hexadecapole Moments (Debye-Ang^3)
XXXX -8301.0105 XXXY -15.0344 XXYY -2439.6549
XYYY
18.3245 YYYY -5827.5545 XXXZ -96.5374
XXYZ
12.6297 XYYZ
66.6282 YYYZ -19.7673
XXZZ -1806.2742 XYZZ
5.7825 YYZZ -1190.7366
XZZZ
-1.6810 YZZZ
2.7693 ZZZZ -1566.1997
Traceless Hexadecapole Moments (Debye-Ang^3)
-70-
Part I: Chapter 2
Water-Soluble Narrow-Line Radicals for DNP-enhanced SSNMR
XXXX 18505.6015 XXXY -1986.8820 XXXZ -8714.8677
XXYY -5794.7730 XXYZ 1403.1837 XXZZ -12710.8285
XYYY 1515.8081 XYYZ 7469.8140 XYZZ 471.0739
XZZZ 1245.0536 YYYY 209.0536 YYYZ -1844.3751
YYZZ 5585.7194 YZZZ 441.1914 ZZZZ 7125.1092
-----------------------------------------------------------------
-71-
DNP
Chapter 3
Narrow-Line Radicals for Hyperpolarized MRS Imaging
Kurhanewicz
Lab
ance Specification and Services
mance
:
C, 15N, 29Si and other spin 1/2
nuclei
13
Typically 15 minutes to 6 hours
All current spectrometer
platforms
Operation with 5 and 10mm
probes
‘Trityl radical’ typical usage
0.2-5mg per sample
ecification
Integrated 94GHz with 0.5GHz
sweep, (user selectable) up to
100mW
3.35T actively shielded
65L from minimum level
Typical <2L per dissolution9
<3.0m
Shorte
r data
collect
of muc
ion
h
materia smaller amou times and/or
l.
nts of
the us
e
sample
■ Be
nefits
improv for quaterna
e
ry carb
elucidat structure co
ons th
at
nf
io
NMR sp n. This coul irmation an will
d
d
ectral
assignm provide mor
e relia
ents in
■ Th
bl
a shor
e abili
ter tim e
e.
identif ty using dire
y
ct
metab complex biom detection of 13
onomics
ixtures
C
to
.
in
■ Us
ing th
chemica e enhanced
signa
l kinetic
experim l to carry ou
t
ents.
HyperS
ense, an
easily
with yo instrument
th
ur labo
spectro
ratory at configures
meter.
and ex
isting
2860mm (~113”)
■
6 bar min
6 bar min, grade 99.999% pure
from cylinder
Vacuum exhaust line or suitable
safety compliant filter
Europe: 230V, 50Hz, 2.5A single
phase and 415V, 50Hz 20A three
phase
USA: 110V, 60Hz, 3A single phase
and 208V, 60Hz, 20A three phase
Japan: 100V, 50-60Hz, 3A single
phase and 200V, 60Hz, 20A
three phase
W 1668mm (~66”)
x D 1038mm (~41”)
Microwa
ve
Energy
CUSA
8403 Cross e
Park
Drive, Suite 3F
Austin Texas 78754 e Tel: +1 512 339
0640
eFax: +1 512 339 0620
E-mail: molecularbiotools@ma.oxinst.com
Trityl ra
dical
Enhan
ced 13C
Spectra
■ Af
l Conte
ter po
larisatio
nt
makes
n, carb
de
on
tection
nuclei
quicker
are en
hanced
compa
. This
red to
conven
tional
E
C-DNP
natura of acetaminop
l abun
hen
dance,
2µmol
(300µg
)
13
103
8m
m(
~41
”)
tion.
C
12
13
NMR.
A
DC
B
solvent
SpinLab Hyperpolarizer
”)
(~66
8mm
166
F
4 hour
Limit of
s polar
isation,
detecti
on S/N
sin
of 10 on gle scan
the sm
allest
peak
2860mm (~113”)
xford-instruments.com
morewas
information
Thisfor
work
carried
tudy of Chemical Kinetics by DNP NMR – Presented at SMASH 2006, Damir Blazina, Steven Reynolds, Oxford Instruments
.*
ological Samples - Presented at ENC 2006, Christian Ludwig1, Abdul Hamid-Emwas1, Damir Blazina2, Steven Reynolds2, Andrew
nstitute for Cancer Studies, University of Birmingham, Birmingham, UK, 2Oxford Instruments Molecular Biotools Ltd., Oxon, UK.*
A. Proc. Nat. Acad. Sci. 2002, 99, 15858.
gy requires a license from the Weizmann Institute of Science.
ar Magnetic Resonance Spectroscopy of Hyperpolarised Solutions, L. Frydman and D. Blazina, Nature Physics, in press.
ob Slade, Graham Hutton, Damir Blazina.*
B.; Gram, A.; Hansson, G.; Hansson, L.; Lerche, M. H.; Servin, R.; Thaning, M.; Golman, K. Proc. Nat. Acad. Sci. 2003, 100, 10158.
ol system with 65% helium insert level at microwave power 100mW over 30 minutes.
s with a temperature of 1.4K over a 6 hour polarisation and dissolution, 100mW microwave power with helium level at 65%.
com.
F
out in collaboration with Kurhanewicz laboratory at UCSF and
Joe Walish at DyNuPol, Inc
HyperSense and SpinLab are trademarks of Oxford Instruments and GE Healthcare
Ref: Hyp/0307
Oxford Instruments Molecular Biotools Ltd and provides outline information only, which (unless agreed by the company in
reproduced for any purpose or form part of any order or contract or regarded as the representation relating to the products
ments’ policy is one of continued improvement. The company reserves the right to alter, without notice the specification,
y product or service. HyperSense is a registered trademark of Oxford Instruments Molecular Biotools Ltd. and the Oxford
Oxford Instruments plc or its subsidiaries.
o use.
What
is D
NP?
Japan
Dynami
c Nucle
ar Polar
strongly Tomioka,
2-11-6,
isatioKoto-ku
polarise
n (DNP
d nucle
) is the
sensitiv
ar spins
method
enhanc
in solut
Tokyo ity135-0047
ement
ion, there for obtaining
for your
by delive
applica
In a gla
ring
tio
ssy
sta(0)
Tel:
te the3fre5245 3261 n.
tempe+81
e electr
rature,
on of th
in the
polarisa
e radica
sence of 4472
tion to (0) pre
Fax:
+81
l, at low
a
nearby3 5245
nuclei us magnetic field
ing micro
can tra
nsfer th
wave irra
E-mail: molecularbiotools@oxinst.co.jp
e
dia
HyperS
en
to levels se allows the
user to
never se
dr
improv
ements en before. Th ive NMR sens
attainab
e sensiti
itivity
the po
tent
le
vity
Nuclear ial to increas with 13C, 15N
and 29Si
e
Magne
tic Reso the breadth
and yo
of your have
ur appl
ications nance (NMR)
experim
.
ents
Exampl
es are:
60L from minimum level
5V signal
Fax: +44 (0) 1865 393 333
E-mail: molecularbiotools@oxinst.co.uk
HyperS
DNP P ense - the in
olarise
-vitro
r
Selectable <3.9K, ≥1.4K8
600kg approx
GE Healthcare
Oxford Instruments
Hyper
Sense ® Molecular Biotools
Delive
ring Se UK
nsiTubney
Abingdon
tivitWoods,
y t5QX
Oxon OX13
o NMR
Tel: +44 (0) 1865 393 200
Water and methanol
www.oxford-instruments.com
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.1 Introduction: MRI and MRSI
Magnetic resonance imaging (MRI) is one of the most powerful non-invasive tools
available to modern medicine. The contrast in MRI1 comes from four sources: the proton
density (protons of bulk tissue water, fat, proteins, etc.), the spin lattice (T1) and spinspin (T2) relaxation times of the tissue protons, and flow (e.g., loss of signal from rapidly
flowing blood). The T1 describes the recovery of longitudinal magnetization, while T2
measures the loss of phase coherence in the transverse plane. The emphasis on or
weighting of each of these sources of contrast by changing the pulse sequence
parameters determines the image composition.
Currently, high MRI contrast necessary for diagnostic imaging is achieved
using gadolinium-based contrast agents. These Gd3+ complexes decrease T1 and T2 of
the tissue in their local environment, and therefore allow for greater contrast. However,
there has been concern about toxicity of gadolinium complexes in patients with kidney
dysfunction.2 FDA is also currently evaluating the risk associated with the gadolinium
deposits found in the brain after repeated use of gadolinium-based contrast agents for
magnetic resonance imaging (MRI).3 Alternative methods for achieving high MRI
contrast are needed. Also, although extremely useful in characterizing anatomical
details, MRI provides little or no information about the biochemical makeup of the tissue,
which contains critical information related to a wide variety of disease pathways,
progression, and treatment efficacy.
Magnetic Resonance Spectroscopic Imaging (MRSI) combines
spectroscopic (NMR) and imaging methods to observe metabolic processes in tissues.
MRSI adds the ability to detect “chemical shift” differences in endogenous, smallmolecule metabolites thereby enabling the direct imaging of metabolic processes. MRSI
has been applied in the study and diagnosis of a large number of disorders including
Weishaupt, D.; Kockli, V. D.; Marincek, B. “How Does MRI Work?” 2nd Ed. Springer, Heidelberg, 2006.
2 FDA Drug Safety Communications Safety Announcement 9-09-2010 http://www.fda.gov/Drugs/
DrugSafety/ucm223966.htm Accessed 7-27-2015
3 FDA Drug Safety Communications Safety Announcement 7-27-2015 http://www.fda.gov/downloads/
Drugs/DrugSafety/UCM455390.pdf Accessed 7-27-2015
1
-73-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
diabetes mellitus,4 cystic fibrosis,5 rheumatism6, and cancer7 (in particular the brain8
breast,9 and prostate10). Understanding in vivo metabolism is necessary in order to
improve diagnosis and therapy of these diseases.
Both MRI and MRSI are limited by the inherently low sensitivity of the
underlying nuclear magnetic resonance. Thermal-equilibrium Boltzmann polarization of
1H
detected by MRI is low, and it is even lower for 13C and 15N – the nuclei important for
MRSI – because of their low gyromagnetic ratios (ɣ). MRSI is further limited by the low
natural abundance of the MRS-active 13C, and 15N nuclei and low concentration of the
metabolites in question.11 A significant improvement in sensitivity is necessary for MRSI
to effectively monitor metabolism. Because the sensitivity problem is a result of the low
nuclear spin polarization, the logical solution is dynamic nuclear polarization (DNP)
described in Chapter 1.
4
Shulman, G. I.; Rothman, D. L.; Jue, T.; Stein, P.; DeFronzo, R. A.; Shulman, R. G. N. Engl. J. Med.
1990, 322, 223. Magnusson, I.; Rothman D. L.; Katz, L.D.; Shulman, R.G.; Shulman, G. I. J. Clin. Invest.
1992, 90, 1323.
5 Dimand, R. J.; Moonen, C. T.; Chu, S. C.; Bradbury, E. M.; Kurland, G.; Cox, K. L. Pediatr. Res. 1988,
24, 243.
6 Ackerman, J. J.; Grove, T. H.; Wong, G. G.; Gadian, D. G.; Radda, G. K. Nature 1980, 283, 167.
7 Pretlow, T. G. II.; Harris, B. E.; Bradley, E. L. J.; Bueschen, A. J.; Lloyd, K. L.; Pretlow, T. P. Cancer. Res.
1985, 45, 442.
8 Chuang, C.F.; Chan, A. A.; Larson, D.; Verhey, L.J.; McDermott, M.; Nelson, S.J.; Pirzkall, A. Technol.
Cancer Res. Treat. 2007, 6, 375.
9 Katz-Brull, R.; Lavin, P. T.; Lenkinski, R. E. J. Natl. Cancer. Inst. 2002, 94, 1197. Bolan, P. J.; Nelson, M.
T.; Yee, D.; Garwood, M. Breast Cancer Res. 2005, 7, 149.
10 Thomas M. A.; Narayan P.; Kurhanewicz J.; Jajodia P.; Weiner M. W. J. Magn. Reson. 1990, 87; 610.
Kurhanewicz, J.; Vigneron, D. B.; Nelson, S. J.; Hricak, H.; MacDonald, J. M.; Konety, B.; Narayan, P.
Urology 1995, 45, 459. Kurhanewicz, J.; Vigneron, D. B.; Hricak, H.; Narayan, P.; Carroll, P.; Nelson, S. J.
Radiology 1996, 198, 795. Males, R. G.; Vigneron, D. B.; Star-Lack, J.; Falbo, S. C.; Nelson, S. J.; Hricak,
H.; Kurhanewicz, J. Magn. Reson. Med. 2000, 43, 17.
11 13C and 15N are chosen because of their longer T and larger chemical shift range for these molecules
1
allows for higher spectral resolution
-74-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.2 Improving the SNR in MRSI with Dissolution DNP
Dissolution dynamic nuclear polarization (DDNP)12 was developed by ArdenkjærLarsen and Golman with the aim of improving liquid state NMR signal-to-noise ratio and
for its application to in vitro and in vivo MRSI with hyperpolarized endogenous small
molecule tracers. The original equipment13 reported by Ardenkjær-Larsen was
redesigned to enable sterile use14 and was commercialized first by Nycomed, later
changing hands from Amersham to the current owner GE Healthcare.15 Sensitivity
improvements by a factor of >10,000 have been achieved in MRSI16 using DDNP.
In a typical DDNP experiment, an endogenous small molecule tracer
(usually 13C or 15N labeled to maximize SNR) is co-frozen with a paramagnetic
polarization agent (a persistent free radical) and placed in a magnetic field in liquid He.
A glass-forming matrix is used if the analyte is not selfglassing, however this dilutes the
tracer and therefore is not desirable. The frozen sample is then irradiated with
microwaves to transfer the relatively high electron polarization to the desired nuclei.17
The buildup of polarization is monitored with a rudimentary SSNMR coil, and when
polarization is complete the sample is rapidly dissolved with hot solvent (water or
buffer). The polarization agent is filtered out (in part to minimize paramagnetic
relaxation) and the hyperpolarized tracer solution is ready for injection into the subject
for MRSI study or for transfer to the NMR magnet for in vitro studies. The dissolution
process and the transfer of the hyperpolarized solution has to occur on a timescale of
12
Ardenkjær-Larsen, J. H.; Fridlund, B.; Gram, A.; Hansson, G.; Lerche, M. H.; Servin, R.; Thaning, M.;
Golman, K. Proc. Natl. Acad. Sci. USA 2003, 100, 10158. Golman, K.; In’t Zandt, R.; Thaning, M. Proc.
Natl. Acad. Sci. USA 2003, 100, 11270. For a recent review of DDNP see Köckenberger, W. eMagRes,
2014, 3, 161.
11 Wolber, J.; Ellner, F.; Fridlund, B.; Gram, A.; Jóhannesson, H.; Hansson, G.; Hansson, L. H.; Lerche, M.
H.; Månsson, S.; Servin, R.; Thaning, M.; Golman, K.; Ardenkjaer-Larsen, J. H. Nucl. Instrum. Methods
Phys. Res., Sect. A 2004, 526, 173.
14 Ardenkjær-Larsen, J. H.; Leach, A. M.; Clarke, N.; Urbahn, J.; Anderson, D.; Skloss, T. W. NMR
Biomed. 2011, 24, 927.
15 GE Diamond SPINlab Polarizer
16 Golman, K.; Ardenkjær-Larsen, J. H.; Petersson, J. S.; Mansson, S.; Leunbach, I.;Proc. Natl. Acad. Sci.
USA 2003, 100, 10435.
17 Typically low-ɣ are polarized directly, however Cross Polarization (CP) schemes have been developed
to effectively polarize 13C via hyperpolarized 1H. Batel, M.; Däpp, A.; Hunkeler, A.; Meier, B. H.; Kozerke,
S.; Ernst, M. Phys. Chem. Chem. Phys. 2014, 16, 21407.
-75-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
seconds to effectively compete with the depolarization of the nuclei (determined by their
relaxivity). Hyperpolarized molecules infused in vivo undergo enzyme catalyzed
reactions18 or other metabolic processes the products of which can also be imaged
given long enough T1. Endogenous tracers such as pyruvic, and fumaric acids, choline,
and acetyl-L-carnitine (Scheme 1) have been chosen for MRSI because of the long T1
of the carbonyl carbons and nitrogen and biochemical significance of these analytes
and their metabolic products. For example, elevated levels of lactate are present in
cancerous growths and the ratio of [1-13C]lactate/[1-13C]pyruvate can be used to image
tumors and evaluate success of treatment.
Scheme 1. Commonly Used Hyperpolarized MRSI Tracers
O
H 3C
O
13C
OH
O
pyruvic-1-13 C acid
18
HO
13C
13C
+NMe
OH
O
fumaric acid-1,4-13 C2
+NMe
HO
13CO
O
3
3H
-
choline bicarbonate -13C
Bowen, S.; Hilty, C. Angew. Chem. Int. Ed. 2008, 47, 5235.
-76-
13C
H 3C
O
O
(R)
3
Cl -
OH
acetyl-1-13 C-L-carnitine hydrochloride
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.3 Polarization transfer mechanisms: Thermal Mixing
Dissolution DNP utilizes relatively low power microwave source at low temperature (1 –
1.4 K) and moderately high (3.35 – 5 T) magnetic field. At this temperature and field
strength, thermal mixing becomes the dominant polarization transfer mechanism.
Thermal mixing19 (TM) requires a homogeneously broad EPR line (linewidth equal or
larger than the nuclear Larmor frequency, D ≥ ωN) from multiple dipolar coupled
electrons. This condition requires high concentrations of paramagnetic polarizing agent.
Polarization transfer process due to TM is usually described using spin-temperature
formalism20 because of the large number of strongly dipolar coupled electrons and
nuclei. While the details are left to the literature,21 one important outcome of the
Borghini model of TM mechanism is that for electronic linewidth greater than the nuclear
Larmor frequency (D ≥ ωN), the microwave frequencies at which polarization maxima
(as determined by SSNMR probe) occur are independent of the nuclei being polarized
(1H, 13C, 15N all obtain the same spin-temperature) and are determined by the
polarization agent. In addition, because dynamic cooling of the electron dipole–dipole
reservoir by the microwave irradiation is more efficient when the heat capacity of this
reservoir is smaller, narrow linewidth radicals (ideally, D = ωN) result in more efficient TM
and higher nuclear polarization.22
It’s also important to note why signal enhancements obtained with
dissolution DNP are much higher than those of SE MAS SSNMR discussed in Chapter
2. Dissolution DNP is routinely reported to give SNR enhancements of 10’000 to
19
For current theoretical description of thermal mixing see Hovav, Y.; Feintuch, A.; Vega, S. Phys. Chem.
Chem. Phys. 2013 ,15, 188.
20 Redfield, A. G. Phys. Rev. 1955, 98, 1787. Provotorov, B. N. Sov. Phys. JETP 1962, 14, 1126.
Borghini,M. Phys. Rev. Lett. 1968, 20, 419. Goldman, M. Spin temperature and nuclear magnetic
resonance in solids, Oxford University Press, Oxford, 1970. Wenckebach, W. T.; Swanenburg, T. J. B.;
Poulis, N. J. Phys. Rep. 1974, 14, 181. Abragam, A.; Goldman, M. Rep. Prog. Phys. 1978, 41, 395.
Duijvestijn, M. J.; Wind, R. A.; Schmidt, J. Physica B & C 1986, 138, 147.
21 Ardenkjaer-Larsen, J. H.; Macholl, S.; Johanesson, H. Appl. Magn. Res. 2008, 34, 509. Jannin, S.;
Comment, A.; Kurdzesau, F.; Konter, J. A.; Hautle, P.; van der Brandt, B.; van der Klink, J. J. J. Chem..
Phys. 2008, 128, 241102.
22 Goertz, S. T.; Harmsen, J .; Heckmann, J.; Heb, Ch.; Meyer, W. ; Radtke, E.; Reicherz, G. Nucl.
Instrum. Methods 2004, 526, 43. Heckmann, J.; Meyer, W.; Radtke, E.; Reicherz, G.; Goertz, S. T. Phys.
Rev. B 2006, 74, 134418. Lumata, L.; Jindal, A. K.; Merritt, M. E.; Malloy, C. R.; Sherry, A. D.; Kovacs, Z.
J. Am. Chem. Soc. 2011, 133, 8673.
-77-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
60’00023 depending on the nucleus being polarized, while much more moderate
numbers of 20–400 (mostly for 1H) are seen for DNP-enhanced solid state NMR
experiments. Because SE relies on the microwave excitation of forbidden transitions,
the probability of these transitions is low. Thermal mixing, on the other hand, is a two
step process where all transitions are allowed,24 so it’s more efficient. Lower microwave
power can be used because at the low temperature utilized for TM , the spin-lattice
relaxation is slow and the EPR transition can be easily saturated. However, the
increased efficiency is not enough so to account for the large enhancement difference.
The difference is a result of the fact that SE DNP enhanced SSNMR samples are
polarized and analyzed at the same temperature (usually 80K or 100K), while with
dissolution DNP polarization is done at low temperature (ca 1K) and data acquisition is
performed after dissolution (at 300 K). A rough estimate of signal enhancement
attributable to dissolution DNP (εDDNP) which does not include the polarization lost via
relaxation during dissolution and sample transfer is given by equation 1. The
enhancement is the ratio of the intensity of NMR (or MRSI) signal (IDDNP)observed after
dissolution DNP to the the signal intensity from a sample at thermal equilibrium at room
temperature (Ithermal). Both the field at which DNP is carried out (BDNP) and at which the
NMR signal is acquired (BNMR) and the corresponding temperatures (TDNP and TNMR)
factor into the equation in addition to the polarization enhancement achieved by thermal
mixing (εTM).25
(1)
The increase in polarization as a result of the temperature jump26 (Figure 1) is
responsible for a factor of ~102 enhancement (TNMR/TDNP = 1.4K/300K). The DNPproper due to the thermal mixing produces enhancements on the order of 100-600
23
Compared to thermal signal at room temperature
D. G.; Meyer, W. Annu. Rev. Nucl. Part. Sci. 1997, 47, 67. Goertz,S.T. Nucl.Instrum.Methods
2004, 526, 28.
25 Joo, C.-G.;Hu, K.-N.; Bryant, J. A.; Griffin, R. G. J. Am. Chem. Soc., 2006, 128, 9428.
26 Purcell, E.M.; Pound, R.V. Phys. Rev. 1951, 81, 279. Abragam, A.; Proctor, W.G. Phys. Rev. 1958,109,
1441.
24Crabb,
-78-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
depending on the ɣN –comparable to SE DNP. Maximum theoretical DNP27
enhancements are given by the ratio of the gyromagnetic ratios εDNP = ɣe/ɣN at T≥ 2K. At
1.4 K the maximum theoretical DNP enhancements decrease. The maximum
enhancement for 13C is ~2600 and 1H ~660 in the high temperature regime, but drops to
~390 and ~100 respectively at 1.4 K . (ɣe = 28 000 MHz/T, ɣ13C = 10.705 MHz/T, ɣ1H =
42.576 MHz/T)
1.4K
𝜀DNP
𝜀TJ
300K
Figure 1. Electron and nuclear spin polarization as a function of temperature at 3.4 T.
Approximately 102 of the polarization increase measured at 300 K is a result of
temperature jump (green) created by rapid dissolution, the other factor of 102 is
ascribed to polarization transfer from electrons to nuclei (theoretical maximum 13C
polarization at 1.4 K is used as an example)
27
see Chapter 1
-79-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.4 Dissolution DNP with narrow line radicals and its applications to MRSI
Dissolution DNP has been successfully applied for real time imaging28 of metabolic
processes in heart and liver using hyperpolarized labeled tracers. Hyperpolarized MRSI
use in cancer diagnosis and treatment monitoring has also been successful.29 The first
human Phase 1 clinical trial30 of prostate cancer metabolic imaging with hyperpolarized
[1-13C]pyruvate was recently completed. The study evaluated the safety and feasibility
of using pyruvic acid as a tracer for characterizing changes in tumor metabolism. The
safety of the agent was confirmed and elevated [1-13C]lactate/[1-13C]pyruvate was found
in regions of biopsy-proven cancer (Figure 2).
Figure 2. 1D 13C dynamic MRSI data from a patient injected with hyperpolarized
[1-13C]pyruvate. (A) Axial T2-weighted image (red arrows show region of cancer). (B)
Dynamic 13C spectra. (C) Plot of 1D localized hyperpolarized pyruvate and lactate data
from the slice that overlapped the region of prostate cancer. (D) Plot of 1D localized
28
Merritt, M. E.; Harrison, C.; Storey, C.; Jeffrey, F. M.; Sherry, F. M.; Malloy, C. R. Proc. Natl. Acad. Sci.
USA 2007, 104, 19773. Gallagher, F. A.; Kettunen, M. I.; Brindle, K. M. Prog. Nucl. Magn. Reson.
Spectrosc. 2009, 55, 285. Malloy,C.R.; Merritt, M.E.; Sherry, A.D. NMR Biomed. 2011, 24, 979. Flori, A.;
Liserani, M.; Frijia, F.; Giovannetti, G.; Lionetti, V.; Casieri, V.; Positano, V.; Aquaro, G. D.; Recchia, F. A.;
Santarelli, M. F.; Landini, L.; Ardenkjaer-Larsen, J. H.; Menichetti, L. Contrast Media Mol. Imaging 2015,
10, 194.
29 For a recent review of MRSI with hyperpolarized nuclei see Kurhanewicz, J.; Vigneron, D. B.; Brindle,
K.; Chekmenev, E. Y.; Comment, A.; Cunningham, C. H.; DeBerardinis, R. J.; Green, G. G.; Leach, M. O.;
Rajan, S. S.; Rizi, R. R.; Ross, B. D.; Warren, W. S.; Malloy, C. R. Neoplasia 2011, 13, 81.
30 Nelson, S. J.; Kurhanewicz, J.; Vigneron, D. B.; Larson, P. E. Z.; Harzstark, A. L.; Ferrone, M.; Van
Criekinge, M.; Chang, J. W.; Bok, R.; Park, I.; Reed, G.; Carvajal, L.; Small, E. J.; Munster, P.; Weinberg,
V. K.; Ardenkjaer-Larsen, J. H.; Chen, A. P.; Hurd, R. E.; Odegardstuen, L.-I.; Robb, F. J.; Tropp, J.;
Murray, J. A. Sci. Transl. Med. 2013, 5, 198ra108.
-80-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
hyperpolarized pyruvate and lactate data from the slice that overlapped a contralateral
region of the prostate. (Figure taken from ref. 30)
Because narrow line radicals are more efficient at thermal mixing
polarization transfer, almost exclusively water-soluble trityl-type radicals31 (such as
Finland and OX063, see Section 2.3.1) have been used for MRSI studies.
BDPA – a narrow line radical32 with very small inhomogeneous broadening – has not
been used broadly because of its limited solubility in the biologically relevant
selfglassing analytes. However Lumata and Kovacs33 were able to successfully polarize
[1-13C]pyruvic acid using sulfolane as a co-solvent reaching P=11% and εDDNP=14000 in
liquid state compared to εDDNP=12000 with OX063 under the same conditions.34 Ideally,
however, the liquid tracer is used neat during polarization which avoids dilution of the
spins and maximizes polarization. This is not always possible with every polarizing
agent due to different solubility profiles of the radicals, and because not all tracers are
self-glassing liquids.
DNP (and hyperpolarized MRSI in particular) is currently very active as a
field with new targets and polarizing agents being actively developed by several groups
around the world. Just a few months ago a benzyl-BDPA derivative was reported by a
group in Spain.35 BA-BDPA (Figure 3 A) effectively hyperpolarized [1-13C]pyruvic acid
31
Nitroxides have been explored for DDNP, but lower polarization was obtained. Lumata, L.; Jindal, A. K.;
Merritt, M. E.; Malloy, C. R.; Sherry, A. D.; Kovacs, Z. J. Am. Chem. Soc. 2011, 133, 8673.
Also stability of the nitroxides in the analyte limits their application, e.g., TEMPO and TOTAPOL
decompose in neat pyruvic acid.
32 See Section 2.3.2
33 Lumata, L.; Ratnakar, S. J.; Jindal, A.; Merritt, M.; Comment, A.; Malloy, C.; Sherry, A. D.; Kovacs, Z.
Chem. Eur. J. 2011, 17, 10825.
34While the enhancements obtained with BDPA-sulfolane system are high, injecting sulfolane into human
subjects is not desirable and dilution of the pyruvate tracer leads to lower overall signal.
35Muñoz-Gómez, J. L.; Monteagudo, E.; Lloveras, V.; Parella, T.; Veciana, J.; Vidal-Gancedo, J. Org.
Biomol. Chem. 2015, 13, 2689.
-81-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
using dissolution DNP and achieving ε = 2886 liquid state NMR enhancement36 (Figure
3) compared to ε = 2886 for BDPA/sulfolane and ε = 2658 with OX063 under same
conditions. Here the liquid state enhancement was calculated by taking the ratio of the
integrated NMR area of the sample in the hyperpolarized state measured 10 s after
dissolution at 14.1 T over the thermal NMR signal measured after the sample returned
to thermal polarization.
The capacity of DNP-MRS for metabolic imaging and the monitoring of
treatment efficacy can be further enhanced by providing researchers with radicals that
provide increased enhancement and/or are applicable for use with a greater variety of
samples.
36
The difference in enhancement value between the two reports Ref 33 vs Ref 35 is due to the different
method of calculating enhancements, different magnetic fields the NMR signal was obtained at (9.4 T vs
14 T), and different length of time after sample transfer. In general, DNP is a rapidly evolving field with a
multitude of experimental setups, lack of universal theoretical treatment and agreed upon definition of
enhancement. As a result, DNP enhancements can only be meaningfully compared when the
experiments are carried out under the same conditions, on the same instrument, and data worked up in
the same way. Experimental conditions such as glassing matrix (neat, sulfolane, water/DMSO, glycerol/
water) etc. and whether the radical and the target tracer is in its COOH or COONa form matter greatly.
E.g. The same Veciana has reported ε = 39 214 of [1-13C]pyruvate liquid state NMR using OX063.
Gabellieri, C.; Mugnaini, V.; Paniagua, J. C.; Roques, N.; Oliveros, M.; Feliz, M.; Veciana, J.; Pons, M.
Angew. Chem. Int. Ed. 2010, 49, 3360. Cross-group comparisons are not useful at this time.
-82-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
Figure 3. Structure of the BA-BDPA polarization agent (A); Frequency sweep
comparing P(+) and P(-) for BA-BDPA and BDPA (B); Buildup curves showing solid
state signal enhancement with 40 mM BA-BDPA is slightly higher and polarization build
up is faster than with OX063 (C); Liquid state 13C DNP-NMR spectra of [1-13C1]pyruvic
acid and the thermal spectrum (D). (B, C, D taken from ref. 35)
-83-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.5 Hyperpolarization of Pyruvate with SA–BDPA and SAH-BDPA results
The modular BDPA synthesis, described in Chapter 2, allows for the creation of tailored
derivatives with tunable solubilities that are compatible with a spectrum of MRS probes
and will allow the rapid optimization of hyperpolarization. This chemistry has the ability
to form the basis for a new set of polarization agents. Initially, however, we wanted to
test the performance of sulfonated BDPA in dissolution DNP. SAH7-BDPA and it’s
sodium salt were prepared as described in Chapter 2. (Scheme 2)
Scheme 2. Synthesis of Sulfonated BDPA for Dissolution DNP
SO3Na
H
BDPAH
1) H 2SO 4 (fuming), 85ºC, 6 h;
NaO 3S
2) Ba(OH) 2, H 2O
3) Dowex HRC-W2(Na +), H 2O
4) PbO 2, Amberlite IRC-50(H+)
NaO 3S
NaO 3S
SO3H
SO3Na
SO3H
HO 3S
Dowex
HRC-W2(H +)
H 2O
HO 3S
SO3Na
HO 3S
SO3H
NaO 3S
HO 3S
SA7–BDPA
SAH7–BDPA
Preliminary DNP studies were conducted using a commercial polarizer –
HyperSenseTM – designed for in vitro work and produced by Oxford Instruments, PLC.
The DNP experiments were carried out at the at the UCSF Hyperpolarized MRI
Technology Resource Center under the guidance of Professor John Kurhanewicz and
instrumental support from Mark Van Criekinge. Stability, solubility, and glass formation in
addition to the DNP enhancements were evaluated with the new radicals. The
[1-13C]pyruvic acid was chosen as the target tracer because of its important applications
in cancer imaging. CN-BDPA and SAH7–BDPA are soluble in neat pyruvic acid (>100
mM), however the solutions lost color (and EPR signal intensity) over two weeks on
exposure to light and air. The sodium salt (SA7–BDPA) is soluble in 8% water-pyruvic
acid and appeared to be less photo- and air-sensitive. Because DDNP samples are not
normally reused (unlike SSNMR samples) as long as the radical is stable in solution for
several hours, it is a viable option for MRSI. From practical medical application point of
view, however, premixed shelf-stable polarizing agent/tracer solutions would be better
-84-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
suited for clinical use. Both neat and 8% water solutions formed glassy solids when
frozen in liquid N2, and no radical partitioning was observed. After cooling to 1.4 K in the
polarizer, a microwave frequency sweep was performed to determine the optimum
excitation frequency for the subsequent polarization studies (Figure 4).
P(+) at 94.087 GHz
1.0
Normalized 13C NMR Signal Intensity
0.8
0.6
0.4
0.2
0
−0.2
−0.4
−0.6
−0.8
−1.0
P(–) at 94.127 GHz
94060
94080
94100
Frequency [MHz]
94120
94140
Figure 4. 13C microwave DNP spectrum of 40 mM SA7–BDPA in 8 wt% water/[1-13C]
pyruvic acid(black) and 40 mM SAH7–BDPA in neat [1-13C]pyruvic acid (red) at 1.4 K
and 3.35 T.
The 13C NMR signal as a function of microwave frequency was normalized to maximum
value. A full microwave sweep was obtained for the SA7–BDPA, to locate the maximum
polarization frequency. Only the positive peak was scanned for SAH7–BDPA to minimize
instrument time. Because of their very similar chemical structure positive polarization
maximum for both radicals occurs at the same frequency (94.087 GHz). This is exactly
the same as reported for BA-BDPA35 in neat pyruvic acid, and BDPA and OX063 in
-85-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
sulfolane/pyruvic acid33(94.06 GHz).The difference between the two polarization
maxima for the SA7–BDPA is 40 MHz which is consistent with thermal mixing
mechanism. The SE mechanism would result in ~70 MHz separation for 13C (ωN = 35.96
MHz). Signal intensity using 40 mM CN-BDPA was very low, either due to radical
decomposition in pyruvic acid or due to poor polarization. Because of this, only P(+)
peak was scanned (Figure 5) and further experiments with CN-BDPA were not pursued.
P(+) at 94.096 GHz
800
13C
NMR Signal Intensity
600
400
200
SO2NH 2
COOH
0
−200
H 2NO 2S
H 2NO 2S
−400
SO2NH 2
CN–BDPA
94040
94060
94080
94100
Frequency [MHz]
94120
94140
Figure 5. 13C microwave DNP spectrum of 40 mM CN–BDPA in neat [1-13C] pyruvic
acid at 1.4 K and 3.35 T.
Next, the polarization buildup time constant (TB) and amplitude as a
function of radical concentration were briefly evaluated. Too little radical leads to
inefficient polarization resulting in longer instrument time. Too much radical results in
accelerated paramagnetic relaxation, shorter nuclear T1, and lower final polarization.
High radical concentration can also affect polarization decay after dissolution by
shortening nuclear T1. For SA7–BDPA (Figure 6) the polarization buildup time constant
TB decreased from 3400 s to 1030 s when concentration was increased from 20 mM to
-86-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
40 mM radical. The maximum solid state polarization amplitude (observed as solid state
13C
NMR signal intensity) was only slightly improved. Liquid state polarization was not
assessed, but if the radical is filtered out or reduced during dissolution, paramagnetic
relaxation is not a concern and faster polarization buildup is preferred. The buildup rate
and signal intensity are comparable to those obtained with BA-BDPA and OX06335
under similar conditions.
11000
40 mM SA7-BDPA
20 mM SA7-BDPA
10000
9000
13C
NMR Signal Intensity [a.u.]
40 mM Amplitude: 9926; Time constant: 1030 s
8000
7000
6000
5000
4000
20 mM Amplitude: 8700; Time constant: 3400 s
3000
2000
1000
0
0
1000
2000
3000
4000
5000
6000
Time [s]
Figure 6. Polarization buildup curves (3.35 T, 1.4 K) of 8% H2O/[1-13C]pyruvic acid
doped with 20 mM (blue) and with 40 mM (red) SA7–BDPA radical irradiated at P(+) =
94.087 GHz. Build up time constant TB is reduced from 3400 s to 1030 s with increased
radical concentration.
For SAH7–BDPA, 40 mM and 60 mM radical concentrations were
evaluated and the effect was similar: TB decreased from 3127 s to 1000 s, but the final
signal intensity was unaffected. Polarization buildup is slower in SAH7–BDPA sample.
This may be due to a number of factors. The glassing matrix in the two sets of
experiments is different: SAH7–BDPA is in neat pyruvic acid, while SA7–BDPA glassing
-87-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
matrix contains 8% water which may affect spin diffusion37 dynamics, or it just creates a
better glass with more homogeneously dispersed radicals. The SA7–BDPA is a salt and
this may affect the interaction of the radical with the tracer. Neutral (-COOH) and anionic
(-COONa) polychlorinated trityl radicals have been reported38 to give a factor of 150
enhancement difference (εDDNP = 198 vs 29925) although no origin of this difference
was proposed. BDPA and BA–BDPA have also been shown35,39 to provide the highest
polarization at 40 mM concentration.
Having obtained baseline performance indicators for the sulfonated BDPA
radicals we investigated ways to increase the maximum polarization. The beneficial
effect of Gd3+ – as a T1e relaxation agent – on final polarization has been reported before
our experiments40 and systematically studied since41. Because sulfonated BDPA
radicals have a very long T1e (56 ms cf.1.3 ms for OX063) it was reasonable to expect
that addition of Gd3+ would give some improvement. Effective concentrations for
electronic (T1e) relaxation vary depending on the radical and glassing matrix. For
example, in glycerol/water maximum polarization obtained with OX06342 can be
quadrupled by addition of 1 mM Gd-HP-DO3A. In sulfolane/pyruvic acid the effect is
less pronounced, with maximum polarization doubled with 2.5 mM Gd3+. At higher
concentrations Gd-DOTA is used as an MRI contrasting agent to shorten nuclear T1 and
T2. In our case, increasing nuclear relaxation is detrimental, so minimal Gd3+
concentration expected to give improvements was used. Adding 1 mM Gd-DOTA
(Figure 7) to 40 mM SA7–BDPA radical doped 8 wt% water/[1-13C] pyruvic acid resulted
in doubling of the solid state polarization.
37
Hovav, Y.; Feintuch, A.; Vega, S. J. Chem. Phys. 2011, 134. Lumata, L.; Jindal, A. K.; Merritt, M. E.;
Malloy, C. R.; Sherry, A. D.; Kovacs, Z. J. Am. Chem. Soc. 2011, 133, 8673.
38 Paniagua, J. C.; Mugnaini, V.; Gabellieri, C.; Feliz, M.; Roques, N.; Veciana, J.; Pons, M. Phys. Chem.
Chem. Phys. 2010, 12, 5824.
39 Merritt, M. E.; Malloy, C. R.; Sherry, A. D.; Kovacs, Z. J. Phys. Chem. A 2012, 116, 5129.
40 Ardenkjær-Larsen, J. H.; Macholl, S.; Johannesson, H. Appl. Magn. Reson. 2008, 34, 509.
41Gordon, J. W.; Fain, S. B.; Rowland, I. J. Magn. Reson. Med. 2012, 68, 1949. Lumata, L.; Merritt, M. E.;
Malloy, C. R.; Sherry, A. D.; Kovacs, Z. J. Phys. Chem. A 2012, 116, 5129. Friesen-Waldner, L.; Chen, A.;
Mander, W.; Scholl, T. J.; McKenzie, C. A. J. Magn. Reson. 2012, 223, 85. Flori, A.; Liserani, M.; Bowen,
S.; Ardenkjaer-Larsen, J. H.; Menichetti, L. J. Phys. Chem. A 2015, 119, 1885.
42 Lumata, L.; Merritt, M. E.; Malloy, C. R.; Sherry, A. D.; Kovacs, Z. J. Phys. Chem. A 2012, 116, 5129.
-88-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
20000
40 mM SA7-BDPA
40 mM SA7-BDPA + 1mM Gd3+
18000
13C
NMR Signal Intensity [a.u.]
16000
With Gd3+: Amplitude: 18253; Time constant: 1180 s
14000
12000
10000
Amplitude: 9926; Time constant: 1030 s
8000
6000
4000
2000
Gd-DOTA
0
0
1000
2000
3000
4000
5000
6000
Time [s]
Figure 7. Effect of Gd3+ doping: polarization buildup curves (3.35 T, 1.4 K) of 8% H2O/
[1-13C]pyruvic acid doped with 40 mM SA7–BDPA radical (red) and with 40 mM SA7–
BDPA radical and 1 mM Gd-DOTA (green, structure inset) irradiated at P(+) = 94.087
GHz.
Interestingly, addition of Gd–DOTA to the 60 mM SAH7–BDPA sample in neat pyruvic
acid (Figure 8) did not result in a significant improvement of the signal amplitude even
though the buildup time constant decreased. This may be due to several factors: Gd3+ is
generally more effective at lower radical concentrations, glassing matrix (lack of water)
may have an effect, and sodium ions may, again, play a role.
-89-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
14000
With Gd3+: Amplitude: 13479; Time constant: 604.9 s
13C
NMR Signal Intensity [a.u.]
12000
Amplitude: 11573; Time constant: 1009 s
10000
8000
6000
4000
2000
60 mM SAH7-BDPA
60 mM SAH7-BDPA + 1mM Gd3+
0
0
1000
2000
3000
4000
5000
6000
Time [s]
Figure 8. Effect of Gd3+ doping: polarization buildup curves (3.35 T, 1.4 K) of neat
[1-13C]pyruvic acid doped with 60 mM SAH7–BDPA radical (red) and with 60 mM SAH7–
BDPA radical and 1 mM Gd-DOTA irradiated at P(+) = 94.087 GHz.
-90-
Part I: Chapter 3
Narrow-Line Radicals for MRS Imaging with DNP
3.6 Summary
The extent of the work described in this chapter was limited by the time and resources
available to us at UCSF at the time. Further experiments including looking into the effect
of water (including deuteration to reduce T1N relaxation pathways) in the glassing matrix;
converting solid state 13C NMR intensity into % polarization using a thermal equilibrium
signal obtained from the same sample; measurements of liquid state enhancements
and nuclear T1; assessing the effect of gadolinium at 40 mM SAH7–BDPA, and its effect
on frequency of polarization maxima. Also methods for filtration and/or reduction of the
radicals during the dissolution need to be investigated. This should be simple given that
sulfonated BDPA radicals are readily reduced in basic buffers used for DDNP and
filtration setups are already available for OX063. Our preliminary results showed that
sulfonated BDPA radicals perform on par or better than state of the art trityl radical
OX063, and the data obtained was sufficient to give rise to a successful STTR grant
application for DYNUPOL, Inc.
3.7 Experimental
For the DNP experiments, the samples were polarized in the HyperSense polarizer
(Oxford Instruments, Tubney Woods, UK). Prior to microwave irradiation, the samples
(25 μL) were quickly frozen in liquid nitrogen to ensure glass formation. The frozen
sample was immediately inserted into the polarizer (1.4 K) then irradiated with
microwaves (100 mW) at a frequency near the ESR frequency (approximately 94 GHz
at 3.35 T) of the radical. The 13C microwave DNP and polarization buildup data were
plotted with DataGraph. Polarization buildup was obtained by fitting 13C signal intensity
vs time using exponential decay equation 2. Where Imax is the maximum 13C NMR signal
(arbitrary units), TB is the polarization buildup time constant (seconds), and k is offset.
(2)
-91-
Chapter 4
Radical Mixtures, Biradicals, and Multiradicals for Cross Effect DNP
Adapted and reprinted in part with permission from:
Michaelis, V. K.; Smith, A. A.; Corzilius, B.; Haze, O.; Swager, T. M.; Griffin, R. G.
“High-Field 13C Dynamic Nuclear Polarization with a Radical Mixture”
J. Am. Chem. Soc. 2013, 135, 2935.
Copyright 2013 American Chemical Society
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
4.1 Polarization Transfer Mechanisms: Cross Effect
The cross effect (CE) is a three-spin mechanism, involving two dipolar coupled
electrons and a nucleus. Microwaves are used to saturate the EPR transition of one
electron which then undergoes a flip-flop-flip process with the other electron and
polarizes the nucleus. CE becomes the dominant mechanism when the homogeneous
linewidth of the radicals is less than the nuclear Larmor frequency, (δ < ωn) while
inhomogeneous spectral breadth is large (Δ > ωn), and the electron exchange coupling
is weak. This allows for the matching condition ωn = ωe1–ωe2, where ωe1 and ωe2 are the
EPR frequencies of the two different radicals or two individual spin packets within an
inhomogeneously broadened EPR line. At high fields where contemporary DNP is
performed, cross effect, which is based on allowed transitions, is the most efficient
polarization transfer mechanism because of its ω−1 field dependence.
A number of custom polarizing agents has been designed for CE DNP in
the past few years, with solid state NMR of biomacromolecules driving the demand. The
emphasis now is on designing water-soluble polarization agents, although some of the
hydrophobic radicals have been adapted for use in aqueous samples by using
surfactants.1 The CE matching condition can be satisfied by a variety of polarizing agent
configurations. We’ll discuss them in order of polarization efficiency.
4.2 Broad radicals as CE polarization agents
Nitroxides such as TEMPO2, TEMPOL, or TEMPAMINE3, used in high enough
concentration (20–40 mM),4 achieve very good 1H enhancements (ε = 50–85 at 5 T and
1
Kiesewetter, M. K.; Michaelis, V. K.; Walish, J. J.; Griffin, R. G.; Swager, T. M. J. Phys. Chem. B 2014,
118, 1825. Lelli, M.; Rossini, A. J.; Casano, G.; Ouari, O.; Tordo, P.; Lesage, A.; Emsley, L. Chem.
Commun. 2014, 50, 10198.
2 Gerfen, G. J.; Becerra, L. R.; Hall, D. A.; Griffin, R. G.; Temkin, R. J.; Singel, D. J. J. Chem. Phys. 1995,
102, 9494.
3 Farrar, C. T.; Hall, D. A.; Gerfen, G. J.; Inati, S. J.; Griffin, R. G. J. Chem. Phys. 2001, 114, 4922.
4 EPR spectrum of TEMPO at higher concentrations is also homogeneously broadened and TM
mechanism becomes an important contributor to DNP. Farrar, C. T.; Hall, D. A.; Gerfen, G. J.; Inati, S. J.;
Griffin, R. G. J. Chem. Phys. 2001, 114, 4922.
-93-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
70–100K). This is because the nitroxide EPR spectrum is inhomogeneously broadened
(D ~ 600 MHz at 5 T) and when two nitroxide radicals are found near and orthogonal to
each other the frequency difference between the gyy and gxx or gyy and gzz components
of the g tensor matches the 1H Larmor frequency. In other words, ω1H “fits inside” the
EPR spectrum of the nitroxide radical. A low temperature EPR spectrum of a generic
nitroxide plotted as a function of magnetic field strength in mT is shown in Figure 1; the
distance (in mT) between the spectral lines that satisfies the CE matching condition is
ωn/ɣe.
ωn
γe
e1
e2
gyy
gxx
gzz
Field [mT]
Figure 1. Frozen EPR spectrum of a generic nitroxide radical showing the CE condition
being met by spin packets at e1 and e2.
-94-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
4.3 Broad biradicals as CE polarization agents
Biradicals incorporating two nitroxides have been developed5 to minimize the
paramagnetic broadening and signal quenching6 caused by high concentration of
unpaired electrons needed for CE (Scheme 1). Covalent linking makes sure that there is
always a second unpaired electron in the vicinity to complete the three-spin flip-flop-flip
process, and controls the electron-electron dipolar coupling. Further, orthogonal
arrangement of the nitroxides in rigid biradicals such as bTbK7 or SPIROPOL ensures
that the relative g-tensor orientations satisfy the CE matching condition with maximal
spectral density.
Scheme 1. A Sampling of Nitroxide Biradical Polarization Agents
N
O
•O
O
N
O•
•O
n
O
N
O
O
n= 2,3,4
•O
N
OH
TOTAPOL
•O
O
N
O
O
bTbk
BTnE
N
H
O
O•
O
X
•O
N
O
X
X
N O•
X
X = S, SO, SO2
bTbtk-py AKA SPIROPOL
O
O
N O•
N
O
N
N
H
O•
N
R
O
R = H PyPol
R = (CH 2CH2O)4Me AMUPol
O
4.4 New nitroxide triradicals for CE and TM DNP
The report of a nitroxide triradical DOTOPA-TEMPO8 by Tycko group served as a
motivation for us to investigate some other trinitroxides. DOTOPA-TEMPO was twice as
effective as TOTAPOL at polarizing 1H at 80 K using a low power microwave source (30
5
For a review of recent developments in polarization agents see Michaelis, V. K.; Ong, T. C.; Kiesewetter,
M. K.; Frantz, D. K.; Walish, J. J.; Ravera, E.; Luchinat, C.; Swager, T. M.; Griffin, R. G. Isr. J. Chem.
2014, 54, 207.
6 Corzilius, B.; Andreas, L. B.; Smith, A. A.; Ni, Q. Z.; Griffin, R. G. J. Magn. Reson. 2014, 240, 113.
7 A1WO2010108995 A1, 2010
8 Thurber, K. R.; Yau, W.-M.; Tycko, R. J. Magn. Reson. 2010, 204, 303.
-95-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
mW at 264.0 GHz) at 9.4 T in glycerol-water matrix. The EPR spectrum (Figure 2) of
DOTOPA-TEMPO shows splitting as a result of dipolar spin-spin coupling.
Figure 2. Room temperature X-band EPR spectra of 0.5 mM solutions of TOTAPOL and
DOTOPA-TEMPO (Figure from ref. 8)
We started our investigation with a known triradical TriT9 (Figure 3, A). It
has not been used for DNP before, and its solubility in biologically relevant solvents was
yet to be determined. The EPR reported in the original paper (Figure 3, B) showed
similar splitting pattern to DOTOPA-TEMPO, indicative of conformers with both weak
and strong exchange coupling (J) and some dipolar broadening, so we were hopeful.
Authors noted that the EPR spectrum varied considerably with solvent, which in our
case means TriT performance may vary significantly depending on what conformations
are favored in a particular DNP matrix.
9
Golubev, V.; Rashba, Y. É. Bull. Acad. Sci. USSR Chem. Sci. 1982, 31, 2445.
-96-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Figure 3. Structure of TriT and its room temperature X-band EPR spectrum (from ref. 9)
in heptane.
The TriT hydrate crystallizes out of the aqueous reaction mixture of 4amino-TEMPO10 with formaldehyde in near quantitative yield, and the compound is a
beautiful rosy color (Figure 4).
•O
O•
N
O
H
NH 2
H 2O
H
rt
N
N
N
O•
N
N
•H 2O
95%
N
O•
TriT
Figure 4. Left: Condensation of 4-amino-TEMPO and formaldehyde gives the
trinitroxide hexahydrotriazine. Right: Crystal structure of TriT (retrieved from CCDC, ID=
CINTUO, submitted by ref. 11, water not shown).
The crystal structure11 of this hexahydrotriazine (Figure 4, right) is an extremely
aesthetically pleasing example of non-bonding electron pair repulsion driving one of the
three substituents on the 6-membered ring into the pseudo-axial orientation.
10
11
TEMPAMINE is water soluble.
Shevyrev, A. A.; Lobkovskaya, R. M.; Shibaeva, R. P. Kristallografiya, 1984, 29, 279.
-97-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Although, as reported previously, this compound is well-behaved (it is
thermally and air stable), it did not dissolve in 60/40 glycerol-water mixture commonly
used in DNP SSNMR experiments. Surprisingly, it was also completely insoluble in
DMSO (DMSO/water being of interest because it’s considered to be the next best
gIassing matrix).
Initial attempts at endowing TriT with water solubility via quaternization
with MeI did not succeed (Scheme 2). Reaction with up to 4 equivalents of MeI in
chloroform did not produce a precipitate, and the water-insoluble material isolated after
concentration closely resembled TriT. Dissolving TriT in neat MeI and allowing for the
reaction to proceed for a day at room temperature gave the expected dark orange
precipitate. However, the isolated solid was still not soluble in water or aqueous
glycerol.
Scheme 2. Attempt at Quaternization of TriT
1 to 4 equiv MeI
•O
N
N
N
O•
CHCl 3
No water soluble material obtained
possibly no reaction
rt and/or Δ
N
N
•O
•H 2O
neat MeI
N
rt, 24 h
N
N
Me
N+
O•
?
N
I–
N
O•
N
O•
dark orange precipitate
not water soluble
Next, 1,3-propanesultone (1,3-PS, which we didn’t get a chance to use in Chapter 2)
was employed to make a water-soluble derivative of the TriT (Scheme 3). The initial
suspension of TriT in warm neat 1,3-PS produces a red solution which quickly gells as
the reaction progresses.
-98-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Scheme 3. Reaction of TriT with 1,3-Propanesultone
•O
N
N
N
N
N
•H 2O
O•
O O
S
O
•O
N
N
N+
O•
N
N
neat 80 ºC, 10 min
-O
3S
N
N
O•
O•
The gel collapses into a sticky residue when diethyl ether is added to wash out excess
1,3-PS, the product can be resuspended in DCM, although it does not re-gel. Two
cycles of washing with ether and DCM were used to remove any remaining 1,3-PS,
after which a very hygroscopic red solid remained. Because we were interested in a
water-soluble triradical, and this compound was clearly yearning for some water, it was
dissolved in water and concentrated (adding ethanol to azeotrope off the water). This
resulted in a powdery, what is presumably, hydrate12 ws-TriT which was much easier to
handle. Ease of handling becomes important as successful polarizing agents start to be
used by non-chemists. Unfortunately X-ray quality crystals, or anything better than a
fine powder, did not form in the variety of solvent systems tried. Mass spectrum of the
ws-TriT showed the TriT fragment in ESI+ mode, however the molecular ion was not
observed. This is consistent with our previous trouble of observing sulfate ions via MS.
It’s reasonable to assume the reaction with 1,3-propanesultone only occurs once per
molecule of TriT because nucleophilicity of the other two nitrogens would be significantly
dampened. The water solubility of ws-TriT was great and solutions with much higher
concentration than necessary for DNP13 could easily be prepared in glycerol/water.
The 140 GHz (5 T) frozen EPR spectrum of 2 mM ws-TriT (6 mM radical)
was collected by Jennifer Mathies and Joe Walish (of Griffin and Swager groups,
respectively) and is shown in Figure 5. The first derivative shows a characteristic
nitroxide powder pattern with some broadening as a result of electron-electron coupling
between the three radicals.
12
13
Precipitation from dry ethanol with diethyl ether gave back the very hygroscopic material.
Usually CE DNP samples use ~20 mM nitroxide biradicals.
-99-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
gxx
2 mM ws-TriT in 60/40 glycerol-water
20 K 140 GHz (5 T)
1st derivative (simulated modulation 5 G)
gyy
gzz
4975
4980
4985
4990
4995
5000
Field [mT]
Figure 5. Frozen 140 GHz EPR of 2 mM ws-TriT (black) and its 1st derivative (green).
Collected by Jennifer Mathies and Joe Walish.
Although TriT was not soluble in glycerol-water, it formed inclusion
complexes with ethanol and did in fact dissolve in ethanol-water mixtures. A literature
search turned up some successful conditions for DNP experiments using ethanolwater14 as a glassing matrix. DNP 1H enhancements with both TriT (in 66/23/11 ethanold6/D2O/H2O v/v/v) and ws-TriT (in 60/30/10 glycerol-d8/D2O/H2O v/v/v) were measured
by Vlad Michaelis and Joe Walish (of Griffin and Swager groups, respectively). In these
preliminary15 experiments, both triradicals performed worse than TOTAPOL in the same
MAS NMR rotor and under the similar conditions (TriT16 ε = 20 ; ws-TriT17 ε = 45;
TOTAPOL ε = 120). The lower enhancement of TriT compared to ws-TriT can be a
14
Kurdzesau, F.; van den Brandt, B.; Comment, A.; Hautle, P.; Jannin, S.; van der Klink, J. J.; Konter, J.
A. J Phys. D Appl. Phys. 2008, 41, 155506. Mango S. Nucl. Instrum. Methods A 2004, 526, 1.
15 A full field sweep was not conducted, thus the enhancement was not necessarily measured at optimal
field.
16 Single point was taken at 4978 G, at 8 W microwave power 139.66 GHz.
17 Four points were collected, ε of 45 was obtained at 4978 G using 4 W power at 69 K. Higher power
yields higher enhancement.
-100-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
result of the inferior glassing matrix, and also be a result of a different spacial
arrangement of the nitroxides in the triradical. The overall poor performance of the
triradicals could be due to too strong scalar coupling (the nitroxides are at a similar
distance to each other as in BTOXA radical, which does not perform as well as
TOTAPOL), and/or unfavorable mutual orientation for dipolar coupling which doesn’t
allow for the CE matching condition to be satisfied.
These problems could potentially be solved18 by using a slightly longer
linker which would provide more distance between the radicals lowering J-coupling, and
more flexibility increasing the chance of correct orientation. The TriTXL19 reported
alongside with TriT could fit the bill (Scheme 4). TriTXL and its water-soluble derivative
would make interesting test subjects for CE and TM DNP experiments.
Scheme 4. Proposed CE/TM DNP Polarization Agents TriTXL (known) and ws-TriTXL
N
O•
N
N
•O
O O
S
O
N
N
O•
N
N
N+
N
-O
N
3S
N
TriTXL
O•
N
O•
ws-TriTXL
O•
4.5 Hetero-biradicals
The next most effective CE polarization agent configuration is a mixture of a narrow line
radical and a broad radical. Such narrow-broad mixture consisting of trityl-type radical
(perdeuterated Finland trityl) and TEMPOL was investigated by Hu and Griffin20 (Figure
18
For a paper detailing effects of structure on DNP efficiency of biradicals see: Ysacco, C.; Rizzato, E.;
Virolleaud, M.-A.; Karoui, H.; Rockenbauer, A.; Le Moigne, F.; Siri, D.; Ouari, O.; Griffin, R. G.; Tordo, P.
Phys. Chem. Chem. Phys. 2010, 12, 5841.
19 Golubev, V.; Rashba, Y. É. Bull. Acad. Sci. USSR Chem. Sci. 1982, 31, 2445.
20 Hu, K.-N.; Bajaj, V. S.; Rosay, M.; Griffin, R. G. J. Chem. Phys. 2007, 126, 044512.
-101-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
6). Next to ideal (EPR) frequency separation is present between the nitroxides’ gyy
component and several trityl-type radicals employed in DNP. The relatively long T1e of
trityl-type radicals allows for microwave saturation of the EPR transition, and short T1e of
the nitroxides results in rapid recycling of polarization. Much higher enhancement (ε ︎=
162 ± 20) was obtained with the mixture compared to the individual components at
equivalent spin concentration (ε ︎= ~15 and ε ︎= ~55 for CT-03(d36) and TEMPOL
respectively).
B
A
C
Figure 6. Low temperature 140 GHz EPR spectra ︎(top)︎ and 1H DNP enhancement
profiles (︎lower panels)︎ with 40 mM perdeuterated CT-03 (A), 40 mM TEMPOL (B), and ︎
1:1 mixture of TEMPOL and CT-OH(d36) (C) at 139.66 MHz in d6-DMSO/D2O/H2O
60:34:6 w/w/w solutions at 90 K. T (Figure from ref. 21)
This successful pairing made it interesting to try the SA–BDPA described in Chapter 2 in
combination with a broad radical. TEMPOL was chosen because of its ready availability.
A field profile of the mixture of 20 mM SA–BDPA and TEMPOL was collected, and
enhancement measured by Vlad Michaelis and Joe Walish (of Griffin and Swager
groups, respectively).
-102-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
100
(SO3Na)ni
OH
(SO3Na)ni
80
N
O•
TEMPOL
ni(NaO3S)
ni(NaO3S)
(SO3Na)ni
SA-BDPA
40
ni = i , i = 1, 2
20
1H
enhancement
60
0
−20
−40
4965
4970
4975
4980
4985
4990
Field [mT]
Figure 7. Field profile of 20 mM SA–BDPA and 20 mM TEMPOL in 60/30/10 glycerol-d8/
D2O/H2O v/v/v at 69 K, 139.66 GHz, 8 W microwave power. 1H enhancements were
calculated after a CP step to 13C.
The enhancement, ~85, was not as high as expected for this very narrow-broad radical
pairing. The CT-03 (not deuterated)/TEMPOL mixture gave ε~115 using the same
experimental setup. Low temperature EPR of the mixture was not collected, but as
apparent from the field profile (Figure 7), the maximum enhancement is no longer
aligned with SA–BDPA EPR line. If the EPR line is shifted because of the matrix being
changed by the presence of TEMPOL the spacing (e1–e2) may no longer be optimal for
1H
polarization. If the EPR line is shifted because of a specific interaction with TEMPOL
that may mean that the two radicals are now too strongly coupled to provide efficient
CE. It is possible that there is some other fundamental difference between trityl-type
and BDPA-type narrow line radicals which is affecting the performance of their mixtures.
BDPA-type radicals are more symmetric and their line width is almost completely
determined by hyperfine coupling whereas trityl-type radicals exhibit slight anisotropic
broadening. The longer T1e of SA–presenceBDPA may also play a role (although there
is some evidence that it is shortened in the presence of other radicals, see section 4.6).
-103-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Other broad radicals, e.g, TEMPO, TEMPONE, or DPPH could be tested and T1e of SA–
BDPA measured in these mixtures to determine the source of underwhelming
performance.
Prior to the (disappointing) DNP data on the mixture of the SA–BDPA/
TEMPOL becoming available, efforts were directed toward the synthesis of a watersoluble BDPA-TEMPO biradical. Previously, the synthesis of a BDPA-TEMPO biradical
(Scheme 5) was reported by Eric Dane.21 Because of its lack of solubility in aqueous
media, however, it was not tested in DNP experiments.
Scheme 5. BDPA–TEMPO Biradical Synthesis by Eric Dane (Scheme from ref. 22)
In retrospect, this particular tethering of BDPA and TEMPO radicals was
not a good system to emulate. The low temperature EPR shows a significant shift in the
BDPA absorption line (Figure 8), similar to the effect we suspect above in the mixture of
SA–BDPA/TEMPOL. There’s still a 196 MHz separation between the nitroxide gyy and
the BDPA component, which should potentially allow CE polarization of 1H (ωn = 211
MHz at this field, 5 T). However, if the perturbation is due to strong J-coupling enabled
by the short tether,22 CE is not going to be the dominant relaxation pathway for the
21
Dane, E. L.; Maly, T.; Debelouchina, G. T.; Griffin, R. G.; Swager, T. M. Org. Lett. 2009, 11, 1871.
A very relevant study of the effect of tether structure in Trityl-TEMPO biradicals has come out since: Liu,
Y.; Villamena, F. A.; Rockenbauer, A.; Song, Y.; Zweier, J. L. J. Am. Chem. Soc. 2013, 135, 2350.
22
-104-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
radical centered on BDPA (it will relax via J-coupling instead of dipolar couplingdependent CE).
Figure 8. Echo-detected 140 GHz 20 K EPR spectra of BDPA–TEMPO (black),
TEMPO (red), and BDPA (blue) in toluene. (Figure from ref. 21)
In order to make a water-soluble version of this biradical, we turned back
to the modular approach developed in Chapter 2. At first in a gunshot approach,
sulfonylchloride 1 (Scheme 6) was reacted with 4-amino-TEMPO, oxidized, and the
mixture chromatographed to separate the different products BDPA(SO3Na)x(SO2NHTEMPO)y.
Scheme 6. Synthesis of SA–BDPA–TEMPO Biradicals
SO2Cl
SO2Cl
ClO2S
NH 2
BDPA(SO 3Na) x(SO 2NH-Tempo) y
H
N
ClO2S
SO2Cl
O•
1
-105-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Two fractions with the most promising EPR spectra (with double integral indicating 1.1
to 1 and 2 to 1 ratio of BDPA core to nitroxide, see Figure 9) were tested in SSNMR
DNP experiments by Griffin group. Enhancements of 7 and 23 were obtained for two of
the fractions.
ε = 23
BDPA(SO3Na)6.5(SO2NH-Tempo)0.5
ε=7
3460
BDPA(SO3Na)6(SO2NH-Tempo)1
3480
3500
3520
3540
3560
Field [G]
Figure 9. X-Band EPR (room temperature) of SA–BDPA–TEMPO biradicals and
corresponding 1H enhancements.
These enhancements were not very high, but we hoped this was a result
of the sample quality. There was some concern over the radical concentration in the
compounds, because the SA–BDPA–TEMPO proved to be difficult to maintain in the
desired oxidation state. The SA–BDPA is prone to reduction in the presence of base,
amines23 in particular, so any stray 4-amino-TEMPO facilitated reduction of the SA–
BDPA radical. Nitroxides are acid sensitive and required very careful neutralization.
Hopeful that the low enhancements were not a result of the intrinsic nature
of the SA–BDPA–TEMPO biradicals, a more controlled approach to the biradical was
23
See Chapter 2 for some discussion of reduction of BDPA radicals in the presence of bases.
-106-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
attempted. The acid chloride24 from the reaction of HOOC–BDPAH with neat
chlorosulfonic acid was isolated and reacted with 4-amino-TEMPO in the presence of
NaOH (Figure 10). The more reactive acid chloride (vs sulfonyl chlorides) was expected
to react with the more nucleophilic amine (vs NaOH) forming the amide, followed by
hydrolysis of the sulfonyl chlorides. On addition of the amine and sodium hydroxide, the
reaction solution turned blue indicating BDPA anion formation. However as reaction
progressed (exposed to air) red color indicative of the sulfonated BDPA radical
developed. It appeared that oxygen (or excess TEMPAMINE25) was oxidizing the the
anion without any extra oxidant present. The progress of the hydrolysis of the sulfonyl
chlorides was monitored by MS. When hydrolysis was complete, careful workup was
required to maintain the biradical. Weak acid cation exchange resin (COOH form) was
used to neutralize the reaction mixture, excess amine was extracted with ether, and
then a weak acid COONa-form exchange resin was used to ensure all counterions were
Na+. After reverse phase chromatography only a 6% yield of the biradical was isolated.
The low yield is partially a result of side reactions with methanol (used to maintain
homogeneous reaction mixture), and decomposition during reaction and
chromatography.
24
In Chapter 2 an aqueous workup was used which hydrolyzed the carboxylic acid chloride but not the
sulfonyl chlorides.
25 This is the opposite of the previously observed reduction of the SA–BDPA radical in the presence of
base.
-107-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
NH 2
SO3Na O
SO2Cl
COOH
ClSO 3H
H
COCl
N
H
N
O•
ClO2S
H
ClO2S
NaOH, THF, MeOH
NaO 3S
Na +
NaO 3S
SO2Cl
O•
N
SO3Na
HOOC-BDPAH
air;
Ion exchange resins
in work up
SO3Na
ε=2
O
BaltaRad
N
O•
N
H
NaO 3S
NaO 3S
3460
3480
3500
3520
3540
3560
SO3Na
BaltaRad
Field [G]
Figure 10. Synthesis and X-Band room temperature EPR of BaltaRad.26
However, the 40 mg that was made was enough for DNP experiments, and Vlad
Michaelis and Joe Walish (of Griffin and Swager groups, respectively) tested it. The 1H
enhancement, however, was only 2.27 This DNP experiment was done in the same time
frame as the SA–BDPA/TEMPOL mixture and is consistent with the strong J-coupling
interfering with CE mechanism. Using a longer linker between the nitroxide and BDPAcore perhaps would benefit here.
26
I promised Balta to name this biradical after him before I knew how lousy of a polarizing agent it was
going to turn out.
27 Full field sweep was not obtained so the 1H enhancement was not optimized and it’s likely the
maximum enhancement frequency was missed. However, enhancement of 2 is not a promising result.
-108-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
4.6 The ideal case for CE: two narrow line radicals
When near-spherically symmetric narrow-line radicals are used as mixtures for CE28
DNP, there is no concern about the orientation of the radicals in space to maximize EPR
spectral density. On the other hand, the separation between EPR lines of the two
radicals must match the Larmor frequency of the nucleus that is being polarized. This
matching condition is satisfied for 13C (and nuclei with similarly low ɣn) by SA–BDPA29
described in Chapter 2 and trityl-type radical OX063 (Figure 11). Direct polarization of
13C
at 82 K was performed by Vlad Michaels, Andy Smith, and Björn Corzilius of Griffin
group.30 The mixture of the two radicals outperformed the individual radicals (at equal
spin concentration) with enhancements εSA–BDPA = 300 due to SE, εOX063 = 480 from a
combination of SE and CE, and εmixture = 620.
The outstanding performance of the SA–BDPA/OX063 mixture is a result
of several contributing factors. The near-perfect spacing of the EPR lines ensures that
the matching condition for CE is satisfied by a majority of unpaired electrons. The long
T1e of SA–BDPA (28.9 ms31)allowed for the SA–BDPA EPR transition to be easily
saturated by microwave irradiation, and ensured that the polarization recovery occurred
mostly via dipolar coupling (in our case CE DNP). On the other hand, the relatively short
T1e (1.4 ms) of OX063 allowed for the CE mechanism to be recycled rapidly leading to
and overall faster polarization of the nuclei. Another beneficial effect that was observed
in the mixture of the two radicals is that SA–BDPA T1e was actually shortened (but not
too much, down to 3.6 ms) by the presence of OX063. This is a result of paramagnetic
relaxation, analogous to adding Gd3+ in the dissolution DNP experiments described in
Chapter 3. The result was that all DNP mechanisms originating from SA–BDPA became
more efficient.
28
This should also hold true for TM
Haze, O.; Corzilius, B.; Smith, A. A.; Griffin, R. G.; Swager, T. M. J. Am. Chem. Soc. 2012, 134, 14287.
30 Michaelis, V. K.; Smith, A. A.; Corzilius, B.; Haze, O.; Swager, T. M.; Griffin, R. G. J. Am. Chem. Soc.
2013, 135, 2935.
31 This T1e value was measured using a 40 mM sample of SA–BDPA and is lower than previously reported
T1e = 56 ms measured using 1 mM sample in the same solvent system. OX063 does not show this
dependence of T1e on concentration. (T1e(1mM) 1.28 ms; T1e(40 mM) = 1.4 ms)
29
-109-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
Figure 11. 140 GHz frozen EPR spectra of SA–BDPA and OX063 (A). Field-dependent
13C
DNP enhancement profiles of SA–BDPA (B), OX063 (C), and a 1:1 mixture (D) at 82
K, irradiated with 139.66 GHz, 8 W microwaves, MAS 4.8 kHz. (Figure from ref. 30)
Theoretically, the performance of these two radicals in CE DNP could be improved by
covalently linking them together. This would allow for lower total concentration of
unpaired electrons and would improve NMR resolution. CE in biradicals may also be
increased leading to higher DNP gain.
-110-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
4.7 Summary
Several combinations of radicals can be used in CE/TM DNP. Biradicals tend to perform
better than monoradicals because of the increased local concentration of spins
(necessary for the three spin CE and multi-spin TM mechanisms) and improved dipolar
coupling. Two nitroxide triradicals were prepared and tested in CE experiments,
however the enhancements obtained were not larger than those obtained with
TOTAPOL biradical likely as a result of strong J-coupling. Hetero-biradicals combining a
narrow line and a broad line radical increase the efficiency of the DNP by providing
higher spectral density (more unpaired electrons) at the microwave irradiation
frequency. These biradicals can be synthesized by covalently linking two different
radicals, or they can be approximated by using a mixture of two different radicals. Unlike
OX063/TEMPOL mixture reported previously, preliminary results show that neither SA–
BDPA/TEMPOL mixture nor the covalently bound biradical enable efficient CE. If the
preliminary experiments are confirmed by further studies, the fundamental difference
between the trityl-type and BDPA-type narrow line radicals should be investigated to
determine the cause of this ineffectiveness. A mixture of two narrow line radicals - the
newly prepared SA–BDPA and the commercially available OX063 - representing an
almost ideal32 polarizing system for CE was investigated for direct 13C CE DNP and
showed outstanding performance with ε > 600 in MAS SSNMR experiment.
32
The ideal case would be a properly spaced heterobiradical made up of these two radicals.
-111-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
4.8 Experimental
Materials. DNP samples were prepared using
13C3-glycerol-d8,
D2O and H2O in a
60:30:10 mixture. OX063 and SA–BDPA samples were prepared using 40 mM radical
concentration (40 mM electrons), while the mixture was a 20 mM OX063 : 20 mM SA–
BDPA. Samples were mixed to give a homogeneous solution at 298 K and packed
within 4 mm sapphire rotors equipped with a Vespel drive cap and a Kel-F top cap.
Typical fill volumes were between 40 and 50 μL.
13C3-glycerol-d8
and D2O were
purchased from Cambridge Isotope Laboratories (Andover, MA) and used without
further purification. OX063 was a gift from Jan-Hendrik Ardenkjaer-Larsen and Klaus
Golman (Nycomed Inc., Malmo, Sweden).
EPR. High field Electron Paramagnetic Resonance ws-TriT, SA–BDPA, and OX063
EPR spectra were obtained by Jennifer Mathies, Andy Smith, Vlad Michaelis and Björn
Corzilius at 140 GHz and 80 K, using 1-2 mM solutions dissolved in 60:40 (v/v)
glycerol:D2O. Spectra were recorded by integration of a Hahn echo at various field
points. For SA–BDPA, 131 field points were acquired between 4989 mT and 5002 mT,
using a π/2–τ–π timing of 30 ns–200 ns–60 ns, and taking 200 shots at each point with
a 4-step phase cycle. For OX063, 241 field points were acquired between 4988 mT and
5000 mT, using a timing of 35 ns–200 ns–70 ns, and taking 400 shots at each point with
a 4-step phase cycle. Electron spin-lattice relaxation times (T1S) for SA–BDPA, OX063
and mixture samples were acquired by saturating the center of the EPR spectrum for
1-3 ms, followed by a variable delay, and then a Hahn echo for detection. The delay is
varied, and the results are fit to a monoexponential curve to obtain the relaxation time.
T1S for trityl was recorded at a field position of 4993 mT and T1S for SA–BDPA were
recorded at 4994.85 mT. For the 40 mM trityl sample, 100 time points are taken with a
delay up to 8 ms, with 400 shots taken per point, and the Hahn echo timing is 41 ns–
700 ns–82 ns. For the 40 mM SA–BDPA sample, 50 time points are taken with a delay
up to 200 ms, with 200 shots taken per point, and a timing of 49 ns–350 ns–96 ns. For
the trityl resonance in the mixture, 50 time points are taken with a delay up to 30 ms,
with 400 shots taken per point, and a timing of 38 ns–800 ns –76 ns. Finally, for the SA–
-112-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
BDPA resonance in the mixture, 50 time points are taken with a delay up to 30 ms, with
400 shots taken per point, and a timing of 38 ns–600 ns–76 ns.
DNP. Dynamic nuclear polarization magic-angle spinning NMR experiments were
performed by Vlad Michaelis with Joe Walish on a custom home-built instrument,
consisting of a 212 MHz (1H, 5 T) NMR spectrometer (courtesy of Dr. David Ruben,
FBML, MIT) and a 140.6 GHz cyclotron maser (gyrotron) generating high power
microwaves up to 15 W. Spectra were recorded on a home-built cryogenic 4 mm
quadrupole resonance (1H,
13C, 15N
& e-) DNP NMR probe equipped with a Kel-F stator
(Revolution NMR, Fort Collins, CO). Microwaves are guided to the sample via circular
overmoded waveguide whose inner surface has been corrugated to reduce mode
conversion and ohmic losses. Sample temperatures were maintained below 85 K, with a
spinning frequency, ωr/2π = 4.8 kHz. Directly polarized
13C
experiments were acquired
under continuous microwave irradiation, using a Bloch sequence and high-power
TPPM33 proton decoupling (1H and
13C
- γB1/2π = 100 kHz). Recycle delays were
determined using a saturation recovery experiment yielding TB between 167 and 287
seconds. The recycle delay was chosen as TB x 1.26, yielding optimum sensitivity per
unit of time. 13C detected DNP field-profiles were performed by sweeping the main NMR
field using a sweep coil between 4974 and 4988 mT (211.8 and 212.3 MHz, 1H Larmor
frequency).
33
Bennett, A.E.; Rienstra, C.M.; Auger, M.; Lakshmi, K.V.; Griffin, R.G. J. Chem. Phys. 1995, 103, 6951.
-113-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
•O
O•
N
O
H
NH 2
N
N
H 2O
H
N
rt
O•
N
N
•H 2O
95%
N
O•
%
TriT
TriT. Was prepared using the procedure reported by Golubev (see ref 8).
TEMPAMINE (0.467 g, 2.73 mmol) was dissolved in 3 mL H2O and aqueous solution of
formaldehyde (0.3 mL, 13 M, 3.81 mmol, 1.4 equiv34) was added dropwise at rt. Within 1
min an orange-rosy precipitate formed and was filtered out and washed with iced water.
The product was dried in a desiccator to constant weight yielding 0.45 g (87%35) of rosy
powder. ESI MS of this compound shows the main 548 amu peak in both ESI+ and ESI–
modes. Appearance and EPR (9 GHz, room temperature) are consistent with those
reported in literature. TriT is not soluble in water or DMSO to an appreciable extent. It is
however, soluble in acidic water and ethanol. 34
35
Larger excess of formaldehyde (e.g. 3) gives higher yields.
On smaller scale yields upward of 95% were observed as reported.
-114-
Part I: Chapter 4
•O
Biradicals and Multiradicals for Cross Effect DNP
N
N
N
N
N
•H 2O
O•
O O
S
O
•O
N
N
N+
N
N
neat 80 ºC, 10 min
-O
O•
3S
N
N
O•
O•
TriT
ws-TriT
•H 2O
ws-TriT. A 50 mL pressure seal vial was charged with 150 mg (0.26 mmol) of TriT and 1
g (8.2 mmol) of 1,3-propanesultone (pre-warmed to 32 ºC to melt). A stir bar was added
and the capped vial was heated in an 80 ºC bath. The resulting suspension became a
dark red solution over 5 min and then gelled. After additional 5 min of heating, the gel
was diluted with 20 mL Et2O leaving a red sticky precipitate on flask walls. This was
suspended in DCM and re-precipitated with E2O twice. The remaining hygroscopic red
precipitate was dissolved in 5 mL H2O, 20 mL of ethanol was added and the solution
was concentrated to a powdery beige ws-TriT.36 EPR (139.997 GHz, 60/40 glycerol/
water, 20 K) gxx=2.0091, gyy=2.0062, and gzz = 2.0022, 14
︎ N hyperfine︎ Axx = 6.8 G, Ayy = 7.1
G, and Azz= 37 G. A small sample was reduced with phenylhydrazine for 1H NMR. The
spectrum is complicated by the variety of conformations.37 The molecular ion of this
compound was not observed by ESI MS. In positive mode (injected in water/THF/
ACN) )parent TriT fragment was observed at 548 amu and in negative mode the [TriT
+Cl-] was present at 585 amu38. The physical properties of the compound however
clearly indicate that it is not the starting material.
36
ws-TriT can be precipitated with Et2O out of EtOH solution, and the resulting fluffy powder is very
hygroscopic.
37 Also, the sample contained excess water so the peaks around 4 ppm are obscured.
38 HRMS was not available at the time these spectra were collected.
-115-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
-116-
Part I: Chapter 4
Biradicals and Multiradicals for Cross Effect DNP
NH 2
SO3Na
SO2Cl
COOH
ClSO 3H
H
O
COCl
N
ClO2S
H
O•
N
H
O•
ClO2S
N
NaOH, THF, MeOH
SO2Cl
NaO 3S
air;
Ion exchange resins NaO 3S
in work up
SO3Na
BaltaRad
HOOC-BDPAH
HOOC-BDPAH (300 mg, 0.65 mmol) in 20 mL DCM was added to 0.5 mL of
chlorosulfonic acid. The reaction mixture stirred at room temperature until 4 sulfonyl
groups were present as indicated by MS ESI- and then poured into 50 mL hexanes. The
precipitate was filtered, washed with hexanes and dissolved in 20% MeOH/THF
(methanol is necessary to keep the intermediate reaction products in solution) and 200
mg of 4-amino-TEMPO and 200 mg of NaOH was added. The resulting blue solution
stirred at rt open to air. Red color developed over 24 h. The reaction mixture was filtered
through Amberlite IRC-50 (COOH form), concentrated and washed with THF and Et2O.
The product was then redissolved in water and stirred with IRC-50 (COONa form) for 5
min and concentrated. Chromatography on C18 reverse phase gave 40 mg (6%) of
BaltaRad as a red-brown glassy solid. ESI MS (after H+ exchange) [M]– calcd for
C43H38N2O14S4: 934.12, found 934 (very weak) and fragments. EPR (9 GHz, rt, water) g
= 2.0028. This material is very pH labile- easily reduced in presence of base. The
carbanion is not stable in strongly oxidizing environments – PbO2 can not be used to
bring up the oxidation state of the BDPA carbanion core, for example.
-117-
Part II: Synthesis of Highly Substituted Indoles
R4
R3
N
R2
R1
Z
This work was carried out in the Danheiser laboratory at MIT in part in collaboration with
Dr Tammy Lam. Only synthetic work performed directly by O. Haze is included below.
For a complete story, see reference:
Adapted and reprinted in part with permission from:
Willumstad, T. P.; Haze, O.; Mak, X. Y.; Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
“Batch and Flow Photochemical Benzannulations Based on the Reaction of Ynamides
and Diazo Ketones. Application to the Synthesis of Polycyclic Aromatic and
Heteroaromatic Compounds.”
J. Org. Chem. 2013, 78, 11450.
Copyright 2013 American Chemical Society
Part II Synthesis of Substituted Indoles
II.1 Introduction
Benzo-fused nitrogen heterocycles1 are of interest as they are common substructures in
natural products and potential drug candidates. The focus of this thesis work is the
development of approaches to highly substituted indoles. While methodology for the
synthesis of C2 and C3 functionalized indoles is well developed,2 indoles that are
substituted on the benzenoid portion still pose a synthetic challenge.
Direct substitution of indole provides access to a limited number of
derivatives, with regioselectivity being the main obstacle. A more efficient approach is to
cyclize a highly substituted aniline derivative. The N-2-oxoethyl aniline derivatives
(carrying a carbonyl group beta to the nitrogen) readily cyclize to form indoles.3 A
properly positioned acetal group can be similarly employed4 (Scheme 1). However, the
problem of preparing these aniline derivatives regioselectively still remains.
Scheme 1. Aniline Cyclizations as Approaches to Indoles
H
R
CHO
H
R
R
N
N
N
R'
R'
R'
CH(OR)2
CH(OR)2
R
NH
R'
Convergent benzannulation strategies5 provide the most attractive approach for the
regiocontrolled construction of multiply substituted benzenoid aromatic systems.
1
For a review of the synthesis of benzo-fused heterocycles see Majhi, T. P.; Achari, B.; Chattopadhyay, P.
Heterocycles 2007, 71, 1011.
2Joule, J. A. Indoles in Science of Synthesis 2000, 10, 361. Beller, M. et at. Adv. Synth. Catal. 2008, 350,
2153. G. R. Humphrey and J. T. Kuethe Chem. Rev. 2006, 106, 2875. Cacchi S.; Fabrizi G. Chem. Rev.
2005, 105, 2873. Zeni, G.; Larock, R. C. Chem. Rev. 2004, 104, 2285.
3 Julia, M.; Tchernoff, G. Bull. Soc. Chim. Fr. 1960, 741. Joule, J. A. In Science of Synthesis; Thomas, E.
J., Ed.; Thieme: Stuttgart, 2000; Vol. 10, pp 390-391 and 461-462.
4 Nordlander, J. E.; Catalane, D. B.; Kotian, K. D.; Stevens, R. M.; Haky, J. E. J. Org. Chem. 1981, 46,
778. Sundberg, R. J.; Laurino, J. P. J. Org. Chem. 1984, 49, 249.
5 Reviews: Kotha, S.; Misra, S.; Halder, S. Tetrahedron 2008, 64, 10775. Saito, S.; Yamamoto, Y. Chem.
Rev. 2000, 100, 2901.
-119-
Part II Synthesis of Substituted Indoles
II.2 Danheiser benzannulation
R1
•
O
OH
R4
R3
R4
R2
X
+
R2
R3
X
R1
Figure 1. Vinylketenes react with alkynes to give highly substituted phenols.
Danheiser benzannulation6 (DBAN) provides an efficient route to highly substituted
phenols. (Figure 1) At the core of this pericyclic cascade (Scheme 2) is a [2+2]
cycloaddition of a vinylketene7 with an alkyne. The vinylketene intermediate can be
generated via the thermal or photochemical ring opening of a cyclobutenone. The 4
electron electrocyclic cleavage is reversible and the equilibrium favors the
cyclobutenone. As a result, only a low concentration of vinylketene is present which is
beneficial in this reaction because vinylketenes have a tendency to dimerize and
polymerize. Arylacetylenes, alkynyl ethers, alkynyl thioethers, ynamines, and ynamides
have been used as the alkyne partner.
Alternatively, the vinylketene intermediate can be generated via
photochemical Wolff rearrangement of an α-diazoketone (Scheme 2, second
generation). The mechanism of Wolff rearrangement is subject to some debate, with
studies pointing to a concerted process for s-Z constrained α-diazoketones and a
combination of stepwise and concerted for s-E substrates. In the cases described here,
6
Danheiser, R. L., Gee, S. K. J. Org. Chem. 1984, 49, 1672. R. L. Danheiser, R. G. Brisbois, J. J.
Kowalczyk, and R. F. Miller J. Am. Chem. Soc. 1990, 112, 3093. For a recent review of Danheiser
benzannulation and its use in tandem synthetic strategies see Willumstad, T. P. Synthesis of Highly
Substituted Benzo-Fused Nitrogen Heterocycles via Tandem Benzannulation/Cyclization Strategies.
Ph.D. Thesis, Massachusetts Institute of Technology, Cambridge, MA, September, 2013.
A related, complementary strategy has been reported by Liebeskind and Moore. Liebeskind, L. S.; Iyer,
S.; Jewell, C. F. J. Org. Chem. 1986, 51, 3065. Perri, S. T.; Foland, L. D.; Decker, O. H. H.; Moore, H. W.
J. Org. Chem. 1986, 51, 3067.
7 For reviews of the chemistry of vinylketenes, see Danheiser, R. L.; Dudley, G. B.; Austin, W. F.
“Alkenylketenes”, In Science of Synthesis; Danheiser, R. L., Ed.; Thieme: Stuttgart, 2006; Vol. 23; pp
493−568. Tidwell, T. T. Ketenes, 2nd ed.; John Wiley & Sons: Hoboken, NJ, 2006; pp 206−214.
-120-
Part II Synthesis of Substituted Indoles
the α-diazoketones either favor s-Z conformations (R1 =H) or are constrained in a ring
and the nitrogen extrusion and 1,2 shift are presumed to occur in concert (Scheme 2,
inset). Advantages of the second generation method include ability to use thermally
sensitive substrates and access to polycyclic aromatic and heteroaromatic systems not
easily available via cyclobutenone-based approaches (first generation DBAN and
Liebeskind and Moore strategies). The vinylketene intermediate is intercepted in a [2+2]
cycloaddition by the alkyne reaction partner.
Scheme 2. Danheiser Benzannulation Pericyclic Cascade
R1
OH
R4
O
Δ or hν
+
R2
R3
R4
Z
N
Z
R5
R5
Z
R2
R3
R4
R3
N
R2
R5
Z
O
R3
N
R5
R1
R2
O
R4
Z
O
•
N
R5
R5
•
R1
R3
second generation
6 electron
electrocyclic
closure
4 electron
electrocyclic
cleavage
N
R2
R1
[2+2]
cycloaddition
Z
photochemical
Wolff rearrangement
O
first generation
R4
R 2 R1
s-Z
tautomerization
O
•
N2 +
R3
R1
4 electron
electrocyclic
cleavage
R1
hν
R2
N
R4
O
R3
R3
R2
R1
-121-
O
O–
R3
•
N
R2
N
hν
R2
R1
R3
R1
Part II Synthesis of Substituted Indoles
The regioselectivity of the [2+2] cycloaddition, which is driven by
maximizing the orbital overlap between the HOMO of the alkyne and LUMO of the
vinylketene (Figure 2), dictates overall regiochemistry of the benzannulation. A bond is
formed between the alkyne carbon carrying the highest HOMO coefficient and the
carbonyl carbon of the vinylketene.
O
O
O
•
O
OR'
R4
OR'
R4
N
N
R5
R5
R1
R1
R2
R3
R2
R3
Figure 2. [2+2] Cycloaddition regiochemistry is determined by orbital coefficients.
The resulting vinyl cyclobutenone undergoes a reversible 4 electron
electrocyclic cleavage to generate a dienylketene which rapidly cyclizes and
subsequently tautomerizes to provide a highly substituted phenol. Full control of the
regiochemistry is maintained throughout the reaction, making DBAN attractive as a
component of a tandem cyclization strategy.
-122-
Part II Synthesis of Substituted Indoles
II.3 Second generation DBAN - tandem cyclization approach
Given an appropriate choice of substrates, benzannulation products can be further
elaborated to furnish highly substituted indoles. In particular, benzannulations utilizing
ynamides8 can give meta-carbamophenols (substituted meta-hydroxy carbamates),
which are already setup for cyclization to 4-hydroxy or 6-hydroxy indoles (Scheme 3).
Two tandem strategies using masked aldehydes for acetal and phenoxide cyclizations
(Scheme 3, Type A and Type B respectively) were examined as part of this work.
Scheme 3. Cyclization Pathways to Highly Substituted Indoles9
OH
OH
CHO
R3
Type A
R2
R1
Z
OH
R3
CHO
Type B
N
R1
Z
R2
R4
R2
N
R2
NH
R1
R3
R3
N
HO
Z
R4
Z
Overall, the indoles can be obtained from the benzannulation products,
which in turn are obtained from the reaction of a diazoketone and an ynamide (Scheme
4). Starting with a cyclic diazo ketone, tricyclic indole heterocycles can be produced,
increasing the utility of this transformation.
8
For recent reviews on the chemistry of ynamides, see: DeKorver, K. A.; Li, H.; Lohse, A. G.; Hayashi, R.;
Lu, Z.; Zhang, Y.; Hsung, R. P. Chem. Rev. 2010, 110, 5064 and Evano, G.; Coste, A.; Jouvin, K. Angew.
Chem. Int. Ed. 2010, 49, 2840.
9 Z is a protecting group, aldehydes are also protected
-123-
Part II Synthesis of Substituted Indoles
Scheme 4. Disassembling Highly Substituted Indoles
OH
OH
R3
R2
R3
N
R2
R1
R2
R
N
HO
R4
Z
Z
R4
R2
N
R2
CHO
R3
N
R5
R4
O
R3
H
R1
R5
OH
R3
N2
R3
NH
R1
R2
CHO
O
CHO
N2
H
Z
Z
N
CHO
When R2---R3 is a benzene ring further variations, including a construction of a
tetracyclic fused heterocycle, are possible (Scheme 5). A similar approach was used in
the synthesis of the tricyclic core of salvilenone.10
Scheme 5. Retrosynthetic Analysis
OH
OH
O
CHO
N
N
R
R
Z
Z
H 3C
CH 3
O
CH 3
O
H 3C
Salvilenone
R. L.; Helgason, A. L. J. Am. Chem. Soc. 1994, 116, 9471.
-124-
N
R
CH 3
10 Danheiser,
CHO
N2
Part II Synthesis of Substituted Indoles
II.4 Substrate synthesis
II.4.1 Synthesis of Ynamides
Compared to ynamines which could also produce aniline derivatives as benzannulation
products, ynamides exhibit increased air and temperature stability, and attenuated
nucleophilicity leading to fewer side reactions. Additionally, ease of synthesis via Nalkynylation makes ynamides attractive in benzannulation-cyclization strategy leading to
indoles.
Ynamide 7 used to to fine-tune the photochemical benzannulation
conditions and the two ynamides (8, 9) required for Type A and Type B cyclization
pathways (See Scheme 3) were prepared by the copper mediated N-alkynylation11 of
carbamates using the procedure reported by our laboratory (eq 1, 2, 4). The requisite
carbamates 1 and 2 are known and were obtained by N-acylation of primary amines
according to literature procedures.12 Carbamate 3 was prepared by Curtius
rearrangement of phenylacetic acid, trapping the intermediate isocyanate with
trimethylsilylethanol.13 Terminal alkynes were halogenated by the method of
Hofmeister14 to give the alkynyl bromides 4–6.15 Ynamides 7 and 8 were obtained using
the standard N-Alkynylation conditions (Method A). However, the reaction of carbamate
3 with alkynyl bromide 6 (eq 4) gave a mixture of the desired ynamide 9, unreacted
carbamate, homodimer 10, and iodoalkyne 11.
11
Dunetz, J. R.; Danheiser, R. L. Org. Lett. 2003, 5, 4011. Kohnen, A. L.; Dunetz, J. R.; Danheiser, R. L.
Org. Synth. 2007, 84, 88.
12 Carbamate 2 was prepared by Masaki Hayashi as reported by: Kozmin, S. A.; Iwama, T.; Huang, Y.;
Rawal, V. H. J. Am. Chem. Soc. 2002, 124, 4628.
13 Kita, Y.; Haruta, J.; Yasuda, H.; Fukunaga, K.; Shirouchi, Y.; Tamura, Y. J. Org. Chem. 1982, 47, 2697.
14 Hofmeister, H.; Annen, K.; Laurent, H. Angew. Chem. Int. Ed. 1984, 23, 727.
15 Octynyl bromide 4 was prepared as reported in Kohnen, A. L.; Mak, X. Y.; Lam, T. Y.; Dunetz, J.
R.;Danheiser R. L. Tetrahedron, 2006, 62, 3815. Alkynyl bromide 5 was prepared by Dennis Huang as
reported in Villeneuve, K.; Riddell, N.; Jordan, R. W.; Tsui, G. C.; Tam, W. Org. Lett. 2004, 6, 4543.
Alkynyl bromide 6 was prepared by the procedure reported in Lam, T. L. Synthesis of Indoles via a
Tandem Benzannulation-Cyclization Strategy. Ph.D. Thesis, Massachusetts Institute of Technology,
Cambridge, MA, September, 2008 from 4,4-Dimethoxybut-1-yne: Deng, J.; Wang, Y. P.; Danheiser, R. L.
Org. Synth. 2015, 92, 13.
-125-
Part II Synthesis of Substituted Indoles
Hex
H
NBS, cat. AgNO3
Hex
acetone, rt, 20 h
Br
4
64%
O
Cl
Method A
MeO
DCM, 0 ºC to rt, 1.5 h
N
H
36%
MeO
Me
(1)
Me
Hex
7
1
O
O
MeO
N
62%
O
CH 3NH 2, Et 3N,
OMe
O
OSit-BuMe2
+
N
H
Method A
MeO
75%
Br
2
5
N
(2)
8
OSit-BuMe2
1.5 equiv Al
cat. HgCl 2
Et 2O, reflux;
H
H
OMe
Br 1 equiv CH(OMe) ;
3
H 2O – 80 ºC to rt
61%
OMe
COOH
SiMe3
Ph
HO
85 ºC, 20 h
78%
N
H
OMe
Br
acetone, rt, 1 h
6
6
Teoc
(3)
OMe
99%
DPPA, Et 3N
toluene, 85 ºC, 2 h;
Ph
NBS, cat. AgNO3
Ph
Method A: 42%
N
Teoc
Method B: 73%
(4)
3
9
MeO
OMe
I
OMe
OMe
CH(OMe)2
OMe
OMe
10
11
Method A: 1 equiv KHMDS, THF, 0 ºC, 20 min; 25 equiv pyridine, bring to rt; 1 equiv
CuI, 2 h; add 1.5-2.1 equiv of alkynyl bromide over 1h, rt, 18 h
Method B: 1 equiv KHMDS, THF, 0 ºC, 20 min; 25 equiv pyridine, bring to rt; 1 equiv
CuBr, 1.5 h; add 1.3 equiv of alkynyl bromide over 30 min, rt, 1 h
-126-
Part II Synthesis of Substituted Indoles
The presence of the iodoalkyne 11 and the unreacted carbamate complicated the
purification and required multiple rounds of column chromatography. Pure ynamide was
obtained in 42% yield16 at best. Presumably the iodoalkyne is generated via coppermediated halogen-halogen exchange.17 Because CuI is the sole source of iodide in this
reaction, a modification18 of the procedure to use CuBr as the Cu(I) species was then
examined (eq 4, Method B). The modified protocol avoids the formation of the
iodoalkyne and requires only 1.3 equivalents of the bromoalkyne (instead of 2.1
equivalents). The reaction time was reduced from 20 h to 1 h, while the yield increased
from 42% to 73%, and pure product was obtained after a single column chromatography
using neutral alumina. The reaction aesthetics were also drastically improved: instead of
a chunky brown slurry when CuI is used, the CuBr mediated reaction mixture is a fine
brilliant-green suspension.19 Copper bromide does not present any handling difficulties
over CuI, and is readily available and comparably priced.
The formation of iodoalkynes under our standard reaction conditions is not
unique to the reaction of 3 and 6, e.g. in the reaction of 1 with octynylbromide (eq 1),
iodooctyne was isolated in 25% yield. The improved N-alkynylation conditions (Method
B) should be applicable in the reactions where the less reactive iodoalkyne is formed
through the halogen-halogen exchange. 16
Although, Lam had reported 57-69% yield of 9 requiring 2–4 purifications via column chromatography.
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L. J. Org. Chem. 2013, 78, 9396.
17 Abe, H.; Suzuki, H. Bull. Chem. Soc. Jpn. 1999, 72, 787.
18 Other variables which were explored but did not provide improvement include toluene and acetonitrile
as solvents, different rates and temperatures for the addition of alkynyl bromide, and addition of AgNO3.
19 Presumably CuIBr or CuI2 can form via ion exchange with KI present in the reaction mixture, this would
be followed by decomposition to give CuI or CuBr2 and I2 responsible for the brown color. Concentrated
aqueous solutions of CuBr2 are green, here the pyridine may play the role of the coordinating ligand
giving the green suspension.
-127-
Part II Synthesis of Substituted Indoles
II.4.2 Synthesis of α-Diazo Ketones
Three types of unsaturated α-diazo ketones20 (acyclic, cyclic and aryl) were prepared to
demonstrate the scope of the second generation benzannulation in the tandem
approach. α-Diazo ketones were prepared from the corresponding ketones (eq 5–8) via
the detrifluoroacetylative diazo transfer21 developed in our laboratory.
O
Me
2 equiv MeLi
Me
OH Et 2O –78 ºC to rt;
0.3 M HCl
Me
O
1.25 equiv HMDS
1.25 equiv nBuLi
Me THF 0 ºC to –78 ºC;
Me
56-61%
Br
O
KOH
OEt
1.2 equiv
CF 3CO2CH2CF 3
O
MeLi, Et 2O
OH
PhMe reflux
O
CF3
O
N2
Me
(5)
61-66%
O
Me
49%
O
O
1 equiv H 2O
1.5 equiv Et 3N
CH 3CN
1.4 equiv MsN 3 Me
rt, 2.5 h
Me
–78 ºC to rt
44%
O
O
Me
HMDS, n-BuLi
THF 0 ºC to –78 ºC;
TFETFA;
Et 3N, H 2O, MsN 3
CH 3CN, rt
O
N2
51%
O
N2
Me
(6)
(7)
N2
(8)
82%
72%
20
For reviews on the synthesis and chemistry of diazo ketones, see: Maas, G. Angew. Chem., Int. Ed.
2009, 48, 8186. Zhang, Y.; Wang, J. Chem. Commun. 2009, 5350. Zhang, Y.; Wang, J. Tetrahedron 2008,
64, 6577. Doyle, M. P.; McKervey, M. A.; Ye, T. Synthesis of α-Diazocarbonyl Compounds. Modern
Catalytic Methods for Organic Synthesis of Diazo Compounds: from Cyclopropanes to Ylides; Wiley &
Sons: New York, 1998; pp 1−60. Regitz, M.; Maas, G. Diazo Compounds: Properties and Synthesis;
Academic Press: Orlando, Fl, 1986.
21These diazo ketones were previously reported. The yields shown here are for the reactions I carried out.
Danheiser, R. L.; Miller, R. F.; Brisbois, R. G.; Park, S. Z. J. Org. Chem., 1990, 55, 1959. Danheiser, R. L.;
Miller, R.F.; Brisbois, R.G. Org. Synth. 1996, 73, 134. Danheiser et al. J. Am. Chem. Soc. 1990, 112,
3093. Cyclobut-1-enecarboxylic acid: Campbell, A.; Rydon, H. N. J. Chem. Soc. 1953, 3002. Lee, J. C.;
Parker, K. A.; Sampson, N. S. J. Am. Chem. Soc. 2006, 128, 4578.
-128-
Part II Synthesis of Substituted Indoles
II.5 D-BAN results
Recently the combination of the cyclobutenone-based benzannulation with several
heterocyclization and annulation processes was used as an attractive route to indoles
bearing multiple substituents on the benzenoid ring.22 Lam has applied three cyclization
methods in the first generation benzannulation-cyclization sequence for the synthesis of
indoles. A limitation of this tandem strategy, however, is that it cannot provide access to
indoles with certain substitution patterns and to indoles fused to additional ring systems.
As demonstrated below, the “second generation” variant of the benzannulation strategy
addresses these limitations and extends the scope of this chemistry to include the
construction of a number of highly substituted polycyclic indoles. The amino phenols
produced by the second generation benzannulation were cyclized via two of the
methods identified by Lam.
Photochemical benzannulation conditions were optimized23 using the
reaction of acetylcyclohexene diazo ketone with ynamide 7 (eq 9). A two step procedure
involving photochemical Wolff rearrangement of the diazo ketone and thermal 4 electron
electrocyclic cleavage of the vinyl cyclobutenone (see Scheme 2) was necessary for the
the reaction completion.
O
O
N2
+
MeO
OH
N
Me
hv Hanovia 450 W
CH2Cl2 10 h;
n-Hex
O
PhMe reflux 1.5 - 2 h
7
N
85%
Hex
(9)
OMe
Me
In the benzannulations where a highly reactive vinyl aldoketene is formed
(such as eq 9, and Table 3 entries 1 and 2) slow addition of an excess of the diazo
ketone was found to be beneficial. Benzannulations with ynamide 9 were aimed at
acetal cyclization previously used by Lam to provide highly substituted indoles.
22
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L. J. Org. Chem. 2013, 78, 9396.
This reaction was originally studied by Xiao Yin Mak: Mak, X. Y. Tandem Benzannulation-Ring Closing
Metathesis Strategy for the Synthesis of Benzo-Fused Nitrogen Heterocycles. Massachusetts Institute of
Technology, Cambridge, MA, September 2008.
23
-129-
Part II Synthesis of Substituted Indoles
In the first case (Table 3, entry 1) the Teoc group in the benzannulation
product 15 was smoothly removed with TBAF, but treatment with HCl at room
temperature (as described by Lam) gave a mixture of the desired indole and the
cyclized 2-methoxyindoline intermediate. However, heating the reaction mixture at reflux
after addition of HCl completed the elimination and provided the desired indole 16 in
good yield. This indole readily oxidizes in air to the corresponding p-indoloquinone,
requiring the workup and chromatography solvents to be purged with Ar prior to use.
Table 3. Tandem Benzannulation Acetal Cyclization Strategy for the Synthesis of
Indoles
O
O
R2
N2
Ph
N
OR
O
SiMe3
DBAN
R1
R3
CH(OMe)2
R3
R2
N
R1
OMe
Teoc
OR
R3
deprotection,
cyclization
N
R2
Bn
R1
Bn
9
OMe
entry
diazo ketone
conditions
DBAN product
O
OH
N2
Me
1
Me
slow addition of
2.4 equiv 12
Hanovia 450W
DCM, rt, 5 h;
toluene reflux 1h
12
OH
N2
2
slow addition of
2.4 equiv 13
Hanovia 450W,
DCM, rt,10 h;
toluene reflux 2.5 h
13
O
3
N
17
N2
1) 1.5 equiv 14
Hanovia 450W
DCM, rt, 2.5 h;
toluene reflux 2 h
OTf
14
Me
82
N
Me
Bn
16
CH(OMe)2
Teoc
59
1) TBAF, THF, rt, 24 h;
then add 6 M HCl
65 ºC 1 h
OTf
63
N
2) NaH, PhNTf 2
THF, 0 ºC to rt, 1 h
Bn
Bn
19
CH(OMe)2
N
2) NaH, PhNTf 2
THF, 0 ºC to rt, 1 h
Teoc
TBAF, THF, rt, 18 h;
then add 6 M HCl
65 ºC 40 min
Bn
N
yield (%)
indole
OH
59
Me
conditions
CH(OMe)2
Me
15
O
yield (%)
Teoc
OTf
60
TfOH, DCM, rt;
H 2O, NaHCO 3
76
N
Bn
Bn
21
22
-130-
Part II Synthesis of Substituted Indoles
Product 17 resulting from the tandem benzannulation-cyclization of the αdiazoacetylcyclobutene (13) with ynamide 9 also required heating at reflux to cyclize.
The resulting indole appeared to be even more air sensitive, and was converted to its
triflate derivative (19) prior to purification (Table 3 entry 2). The necessity of this
transformation in order to isolate the stable indole 19 does not detract from the utility of
the method, as the triflyl group provides a handle for further elaboration of the indole
product.
Benzannulation of the benzosuberone-derived diazo ketone (Scheme 6)
provided phenol 20 in good yield. This benzannulation does not require a slow addition
of an excess of the diazo ketone because the more sterically congested ketoketene that
is formed via the Wolff rearrangement is less prone to dimerization and polymerization.
Scheme 6. Benzosuberone Cyclization
O
N2
OH
9
Hanovia 450W
DCM, rt, 2.5 h;
CH(OMe)2
TBAF, THF, rt, 24 h;
toluene reflux 2 h
N
Teoc
decomposition
then add 6 M HCl,
65 ºC, 1 h
Bn
14
20
79%
PMB-Br;
TBAF;
HCl
OPMB
N
Bn
24
7%
Removal of the Teoc protecting group with TBAF also appeared (by TLC) to proceed
well. Unfortunately, the acidic conditions required for the cyclization step led to complete
decomposition. Possibly, because of the high electron density on the naphthalene, the
unprotected product is easily oxidized under acidic conditions. PMB protection of the
phenol followed by deprotection of the Teoc and cyclization did not afford the desired
-131-
Part II Synthesis of Substituted Indoles
indole 24 in a synthetically useful yield. Attempts at replacement of the phenol with an
electron withdrawing group, including cyanation and carbonylation failed. Next the
phenol was converted to the triflate derivative 21 because the triflyl group was thought
to increase air stability of the final indole, and further attempts at cyclization of 21 were
made. TBAF is known to cleave aromatic triflates, so several acidic conditions, including
HCl, TFA, and TsOH, were tested instead in an attempt to simultaneously deprotect the
Teoc and the acetal groups and effect cyclization. Unfortunately, Teoc proved to be
difficult to cleave under these conditions and acetal deprotection was followed only by
decomposition of the substrate. An alternative route was considered: because triflyl
group was complicating the TBAF deprotection step, we tried to replace it. Attempts at
functionalization of 21 via cross-coupling failed mostly leading to the hydrolysis of the
triflyl function followed by oxidation to the corresponding quinomethide.
Finally, a superacid TfOH (pKa ~ –12, water and DCE)24 was used to
effect the cyclization of 21 to the desired naphthoindole 22 (Table 3, entry 3).
Presumably upon treatment with TfOH, Teoc is removed and a dicationic intermediate
23 is formed but is unable to cyclize under strongly acidic conditions (Scheme 7). Lack
of cyclization is beneficial here, because it allows for deprotection to complete without a
competing dimerization of the indole product. When water is added, hydronium
becomes the strongest acid in the reaction mixture no longer strong enough to stall
cyclization. Dehydration occurs within seconds (as monitored by 1H NMR) and prompt
neutralization of the reaction mixture is needed to prevent dimerization of the indole.
Scheme 7. Proposed Intermediate in TfOH-Mediated Cyclization
OTf
OTf
CH(OMe) 2
N
Bn
O
H
H
TfOH
DCM
N
Teoc
H
Bn
H
(-OTf)2
21
23
24
Raamat, E.; Kaupmees, K.; Ovsjannikov, G.; Trummal, A.; Kütt, A.; Saame, J.; Koppel, I.; Kaljurand, I.;
Lipping, L.; Rodima, T.; Pihl, V.; Koppel, I. A.; Leito, I. J. Phys. Org. Chem. 2013, 26, 162.
-132-
Part II Synthesis of Substituted Indoles
N-Allyl ynamides serve as another useful building block in our tandem
benzannulation/heterocyclization strategy22 for the synthesis of highly substituted
indoles. Oxidative cleavage of the allyl double bond following benzannulation affords an
aldehyde that undergoes phenoxide cyclization and dehydration to generate the desired
6-hydroxyindole upon treatment with K2CO3 followed by brief exposure to HCl.
Application of the phenoxide cyclization protocol in conjunction with the “second
generation” benzannulation provides access to very highly substituted indoles whose
synthesis would be difficult to achieve by alternative methods. Benzannulation of
ynamide 8 with excess of diazo ketone 12 proceeded well under conditions described
above. The dihydroxylation of the benzannulation product 26 followed by oxidative
cleavage yielded aldehyde 28 which cyclized upon treatment with K2CO3 to the
intermediate 3-hydroxy indoline. The dehydration of the intermediate and a
simultaneous deprotection of the primary alcohol was completed by addition of excess
hydrochloric acid. (Table 4, entry 1) The cyclization and dehydration can be effected by
K2CO3 alone, (Table 4, entry 2) provided sufficient heating and time is allowed. Thus it is
possible to maintain the TBS protecting group if it is necessary in a sythesis.
-133-
Part II Synthesis of Substituted Indoles
Table 4. Tandem Benzannulation Phenoxide Cyclization Strategy for the Synthesis of
Indoles
O
MeO
O
R2
R2
R3
N
oxidative cleavage;
cyclization
HO
H
N
N
HO
CO2Me
CO2Me
8
entry diazo ketone
OSit-BuMe 2
OSit-BuMe2
DBAN product
conditions
OR
yield (%)
conditions
N2
1
Me
slow addition of
2.4 equiv 12
Hanovia 450W
DCM, rt, 7 h;
toluene reflux 3 h
Me
Me
Me
HO
N
76-78
CO2Me
OSit-BuMe 2
12
yield (%)
indole
Me
O
Me
H
DBAN
N2
R3
R2
R3
1) OsO4, NMO
THF/H2O 24-40 h;
NaIO 4-SiO2, DCM HO
rt, 10 min;
2) K 2CO 3 iPrOH 75 ºC;
HCl
26
N
52-59
CO2Me
OH
29
O
N2
2
slow addition of
2.4 equiv 13
Hanovia 450W,
DCM, rt, 6-10 h;
toluene reflux 1 h
HO
N
CO2Me
OSit-BuMe 2
13
30
-134-
62-75
1) OsO4, NMO
THF/H2O 48 h;
HO
NaIO 4-SiO2, DCM
rt, 20 min;
2) K 2CO 3 iPrOH 75 ºC
N
62
CO2Me
OSit-BuMe 2
33
Part II Synthesis of Substituted Indoles
II.6 Product elaboration
Thermolysis of cyclobutarenes provides access to o-quinodimethanes, whose utility as
highly reactive dienes in Diels-Alder cycloadditions is well documented.25 Indoles 19
and 33 incorporate a benzocyclobutene moiety, and were expected to undergo ring
opening to the corresponding o-quinodimethanes. Intermediate 34 from the thermal ring
opening of indole 33 successfully participated in [4+2] cycloadditions with activated
dienophiles (Scheme 8). Reaction with N-phenylmaleimide gave the expected product
(35) in acceptable yield. Regioselectivity of the cycloaddition was probed in the reaction
with ethyl propiolate. No selectivity was observed with ca. 60:40 inseparable mixture of
regioisomers 36 and 37 forming in 62% yield.
Scheme 8. Tricyclic Indoles via Diels-Alder Reactions of o-Quinodimethanes
Ph
N
Ph
N
O
N
HO
p-xylene
180 ºC, 40 h
CO 2Me
HO
CO 2Me
33
N
CO 2Me
OSit-BuMe 2
OSit-BuMe 2
H
H
O
(4 equiv)
N
HO
O
O
OSit-BuMe 2
35
44%
34
O
OEt
(4 equiv)
CO2Et
EtO 2C
HO
N
HO
CO 2Me
36
OSit-BuMe 2
N
CO2Me
62% as 40:60 mixture
25
OSit-BuMe 2
37
Charlton, J. L.; Alauddin, M. M. Tetrahedron 1987, 43, 2873. Segura, J. L.; Martin, N. Chem. Rev. 1999,
99, 3199. Sadana, A. K., Saini, R. K., Billups, W. E. Chem. Rev., 2003, 103, 1539. Toyota, S.; Iwanaga,
T., Science of Synthesis, 2010, 45, 752.
-135-
Part II Synthesis of Substituted Indoles
While indole 33 reacted as expected, the linear fused indole 19 did not undergo ring
opening at 180 ºC and decomposed at higher temperatures. This recalcitrance can be
explained by the necessity of breaking aromaticity in both the benzenoid and the pyrrole
portion of the indole during ring cleavage.(Scheme 9)
Scheme 9. Linear Fused Cyclobutaindole Fails to Undergo Ring Opening
OTf
OTf
p-xylene
N
19
180 ºC
N
Bn
Bn
In a classic example of this, a larger dearomatization energy is required for the ring
opening of linear 1,2-dihydrocyclobuta[b]naphthalene (39) than that of 1,2dihydrocyclobuta[a]naphthalene (38) to afford the o-quinodimethanes 4
‌ 1 and ‌40,
respectively. (Scheme 10) This difference is reflected in the higher reaction temperature
necessary to achieve reaction of 39 with maleic anhydride.26
Scheme 10. Generation of 1,2-Dimethylene-1,2-Dihydronaphthalene and 2,3Dimethylene-2,3-Dihydronaphthalene by Thermally Induced Ring-Opening Reactions
O
O
O
DEP
200 ºC
O
O
O
53%
38
40
O
DEP
250 ºC
O
O
61%
39
26
O
O
41
Cava, M. P.; Shirley, R. L.; Erickson, B. W., J. Org. Chem., 1962, 27, 755.
-136-
O
Part II Synthesis of Substituted Indoles
The tandem sequence provides substituted indoles which can be further
functionalized using the triflate group as a handle. For example, triflate 22 was
converted to the 4-ethylindole 25 via Suzuki-Miyaura coupling (Scheme 11). The
combination of Buchwald’s SPhos and Cs2CO3 with THF was necessary to effect the
coupling because this substrate decomposed in DMF and acetonitrile on standing, and
the triflate was readily cleaved with K2CO3, KF, and even Hünig’s base.
Scheme 11. Product Elaboration via Suzuki Cross-Coupling
5 mol% Pd(OAc) 2
7.5 mol% SPhos
3 equiv Cs2CO 3
2.5 equiv Et 3B
OTf
N
Bn
Et
THF, rt, 1 h
N
79%
22
Bn
25
Although it was not investigated in detail, the propensity of the electron
rich N-benzyl 4-hydroxyindoles (16, 18, 22) to oxidize to corresponding pindoloquinones (Scheme 12) may be of interest. The autoxidation is clean and
represents a potentially viable route to 4,7-indoloquinones which are an important class
of bioreductive alkylating agents and are also present as structural motifs in antifungal
and cytotoxic natural products.27
27
Exiguamines: Brastianos, H.C. et al. J. Am. Chem. Soc. 2006, 128, 16046. Discorhabdins: Y.
Harayama, Y. Kita. Curr. Org. Chem. 2005, 9, 1567
-137-
Part II Synthesis of Substituted Indoles
Scheme 12. Air Oxidation of 4-Hydroxyindoles to 4,7-Indoloquinones and Related
Compounds
O
OH
O
Me
MeN
Me
N
Me
16
Bn
N
Me
Bn
O
O
O
NH 2
O
NMe
HO
N
H
O
OH
O
+
Me N
Me
N
18
Bn
OH
N
O
Bn
O
Exiguamine A
Br
Br
N
O
N
N
Bn
20c
Bn
H
N
H
NH
O
Discohabdin C
II.7 Summary
Readily available diazo ketones and ynamides were used in a tandem second
generation benzannulation/cyclization strategy to obtain a variety of substituted indoles.
This approach provides a convergent and efficient synthesis of polycyclic indoles not
accessible via previously reported strategies. Products incorporating the
benzocyclobutane moiety serve as o-quinodimethane precursors for Diels-Alder and
dipolar cycloadditions. Indole products can be further functionalized via cross-couplings,
C-H insertions, or oxidation.
-138-
Part II Synthesis of Substituted Indoles
II.8 Experimental
General Procedures. All reactions were performed in flame-dried or oven-dried
glassware under a positive pressure of argon and stirred magnetically unless otherwise
indicated. Air- and moisture-sensitive liquids and solutions were transferred via syringe
or cannula and were introduced into reaction vessels through rubber septa. Reaction
product solutions and chromatography fractions were concentrated by using a Büchi
rotary evaporator at 15-20 mmHg and then at 0.05 mmHg (vacuum pump) unless
otherwise indicated. Thin layer chromatography was performed on EMD (Merck)
precoated glass-backed silica gel 60 F-254 plates. Column chromatography was
performed on Sorbent Technologies silica gel 60 (32-63 μm or 40-63 μm) or Aldrich
aluminum oxide (deactivated, neutral, Brockmann III, standard grade, ~150 mesh, 58Å).
Slow addition from syringes were performed with an Orion M365 multi-range variable
rate infusion pump manufactured by Thermo Electron Corporation.
Materials. Commercial grade reagents and solvents were used without further
purification except as indicated below. (a) Distilled under argon from calcium hydride:
pyridine, and triethylamine. (b) Purified by pressure filtration through activated alumina:
dichloromethane, diethyl ether, and tetrahydrofuran. (c) Purified by pressure filtration
through activated alumina and Cu(II) oxide: toluene.
Instrumentation. Melting points were determined with a Fisher-Johns melting point
apparatus and are uncorrected. Infrared spectra were obtained using a Perkin Elmer
2000 FT-IR spectrophotometer. 1H NMR spectra were recorded on Bruker Avance 400
(400 MHz), and Bruker Avance 600 (600 MHz) spectrometers. 1H NMR chemical shifts
are expressed in parts per million (ppm) downfield relative to tetramethylsilane (with the
chloroform resonance at 7.26 ppm and others as reported28 as a standard). 13C NMR
spectra were recorded on Bruker Avance 400 (100 MHz) and Bruker Avance 600 (125
MHz) spectrometers. 13C NMR chemical shifts are expressed in parts per million (ppm)
28
Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.; Nudelman, A.; Stoltz, B. M.; Bercaw, J.
E.; Goldberg, K. I. Organometallics 2010, 29, 2176.
-139-
Part II Synthesis of Substituted Indoles
downfield relative to tetramethylsilane (with the chloroform resonance at 77.16 ppm as a
standard). High resolution mass spectra (HRMS) were measured on Bruker Daltonics
APEXII 3 Tesla Fourier Transform and Bruker Daltonics APEXIV 4.7 Tesla mass
spectrometers. Elemental analyses were performed by Atlantic Microlab, Inc. Norcross,
GA 30091
-140-
Part II Synthesis of Substituted Indoles
O
O
MeO
N
H
2
OSit-BuMe2
+
MeO
N
Br
5
8
OSit-BuMe2
Methyl allyl(4-(tert-butyldimethylsilyloxy)but-1-ynyl)carbamate (8). A 100 mL, threeneck, round-bottom flask equipped with a rubber septum, glass stopper, and a pressure
equalizing addition funnel, fitted with an argon inlet adapter, was charged with
carbamate 2 (0.660 g, 5.74 mmol, 1.0 equiv) and THF (25 mL). The solution was cooled
at 0 °C while KHMDS (0.91 M in THF, 6.3 mL, 5.7 mmol, 1.0 equiv) was added via
syringe over 4 min. The resulting white slurry was stirred 20 min at 0 °C, and then
pyridine (11.5 mL, d = 0.978 g/mL, 142.4 mmol, 25 equiv) was added via syringe, and
the reaction mixture was allowed to warm to rt. CuI (1.08 g, 5.67 mmol, 0.99 equiv) was
added in one portion against a positive flow of argon giving a green-yellow mixture. The
flask was wrapped in aluminum foil and the reaction mixture was stirred 2 h at rt,
developing a brown-green color. A solution of bromoalkyne 5 (3.00 g, 11.4 mmol, 2.0
equiv) in THF (15 mL) was added via the addition funnel over 1 h. The resulting mixture
was stirred 18 h at rt and then diluted with Et2O (100 mL) and washed with a mixture of
saturated aq NaCl and concentrated aq NH4OH (2:1, 3x30 mL). The combined aqueous
layers were extracted with Et2O (2x20 mL), and the combined organic extracts were
washed with 5% aq HCl (20 mL), DI water (50 mL), saturated aq NaCl (80 mL), dried
over Na2SO4, filtered, and concentrated to give 2.90 g of a brown oil. Column
chromatography on 200 g of silica gel (eluting with 5% EtOAc–hexanes) afforded 1.26 g
(75%) of ynamide 8 as a yellow oil: IR (film) 3085, 2955, 2930, 2857, 2262, 1732, 1646,
1445, 1391, 1330, 1298, 1233, and 1105 cm-1; 1H NMR (400 MHz, CDCl3) δ: 5.85 (ddt,
J = 17.0, 10.3, 6.0 Hz, 1H), 5.26 (dd, J = 18.8, 1.2 Hz, 1H), 5.23 (dd, J = 10.0, 1.2 Hz,
1H), 4.04 (d, J = 6.0 Hz, 2H), 3.80 (s, 3H), 3.71 (t, J = 6.4 Hz, 2H), 2.51 (t, J = 7.0 Hz,
2H), 0.90 (t, J = 2.8 Hz, 9H), and 0.07 (t, J = 3.2 Hz, 6H); 13C NMR (100 MHz, CDCl3) δ:
156.0, 131.8, 118.3, 74.4, 67.0, 62.3, 53.9, 52.7, 25.9, 22.9, 18.4, and -5.1; HRMS (ESI)
[M+Na]+ calcd for C15H27NO3Si: 320.1652, found 320.1650.
-141-
Part II Synthesis of Substituted Indoles
8
-142-
Part II Synthesis of Substituted Indoles
Bn
Ph
N
H
3
Teoc
+
Br
N
Teoc
OMe
OMe
6
9
CH(OMe)2
N-Benzyl-N-2-(trimethylsilyl)ethoxycarbonyl-4,4-dimethoxybut-1-ynylamine (9). A
250 mL, two-neck round-bottom flask equipped with a rubber septum and argon inlet
adapter was charged with carbamate 3 (4.02 g, 16.0 mmol, 1.0 equiv) and 50 mL of
THF. The solution was cooled at 0 °C while KHMDS solution (0.91 M in THF, 18.0 mL,
16.4 mmol, 1.0 equiv) was added via syringe over 5 min. The mixture was stirred for 20
min at 0 °C, and then pyridine (32.0 mL, 31.3 g, 396 mmol, 25 equiv) was added via
syringe. The reaction mixture was warmed to rt, and CuBr (2.30 g, 16.0 mmol, 1.0
equiv) was added in one portion giving a cloudy mixture. The flask was wrapped in
aluminum foil and the reaction mixture was stirred at rt for 1.5 h. A solution of bromo
alkyne 6 (4.10 g, 21.2 mmol, 1.3 equiv) in 35 mL of THF was added via syringe over 1
h. The resulting emerald green mixture was stirred at rt for 1 h, and then diluted with
200 mL of diethyl ether. The reaction mixture was washed with three 100 mL portions of
a 2:1 mixture of brine and concd aq NH4OH solution, 100 mL of water, and 100 mL of
brine, and dried over Na2SO4, filtered, and concentrated to give 7.0 g of a red-brown oil.
Column chromatography on 300 g of neutral alumina (elution with 5% EtOAc–hexanes)
afforded 4.23 g (73%) of ynamide 9 as a pale yellow oil: 1H NMR (400 MHz, CDCl3) δ
7.37–7.26 (m, 5H), 4.59 (s, 2H), 4.45 (t, J = 5.7 Hz, 1H), 4.27 (m, 2H), 3.32 (s, 6H), 2.58
(d, J = 5.7 Hz, 2H), 1.05 (m, 2H), 0.04 (s, 9H);
13C
NMR (100 MHz, CDCl3) δ 155.8,
136.5, 128.5, 128.0, 102.8, 65.7, 53.7, 53.4, 23.9, 17.7, -1.4 (alkyne peaks were not
observed); HRMS (ESI) [M+Na]+ calcd for C19H29NO4Si: 386.1758, found 386.1765.
-143-
Part II Synthesis of Substituted Indoles
9
-144-
Part II Synthesis of Substituted Indoles
9
-145-
Part II Synthesis of Substituted Indoles
O
O
N2
Me
13
1-(Cyclobut-1-en-1-yl)-2-diazoethanone (13). A 250 mL, two-neck, round-bottom flask
equipped with a rubber septum and argon inlet adapter was charged with 1,1,1,3,3,3hexamethyldisilazane (4.8 mL, 3.7 g, 23 mmol, 1.1 equiv) and 40 mL of THF. nButyllithium solution (2.31 M in hexane, 9.90 mL, 22.9 mmol, 1.1 equiv) was then added
at 0 ºC over 5 min. The resulting yellow solution was stirred at 0 ºC for 1 h, and then
cooled to –78 ºC. A solution of 1-acetylcyclobutene (2.0 g, 20.8 mmol, 1 equiv) in 40 mL
of THF was added via cannula over 30 min. The reaction mixture was stirred at –78 ºC
for 30 min, and then 2,2,2-trifluoroethyl trifluoroacetate (3.80 mL, 5.60 g, 28.4 mmol, 1.4
equiv) was added rapidly (1 s) by syringe in one portion. After 10 min, the reaction
mixture was diluted with 150 mL of Et2O and transferred into a separatory funnel
containing 100 mL of 1 M aqueous HCl solution. The aqueous layer was separated and
extracted with three 50 mL portions of Et2O, and the combined organic phases were
washed with brine, dried over Na2SO4, filtered, and concentrated in a 300 mL roundbottom flask to give 9.2 g of a green oil. The flask was equipped with a stir bar and a
septum, and flushed with argon. Acetonitrile (40 mL), water (0.375 mL, 0.375 g, 20.8
mmol, 1 equiv), and triethylamine (4.35 mL, 3.16 g, 31.2 mmol, 1.5 equiv) were then
added giving a bright orange solution. A solution of methanesulfonyl azide (4.69 g, 38.7
mmol, 1.9 equiv) in 30 mL of CH3CN was added over 10 min, and the resulting solution
was stirred at rt for 4 h, and then concentrated to a volume of ca. 15 mL. The residue
was diluted with 150 mL of Et2O and washed with three 50 mL portions of cold 5%
aqueous NaOH solution. The combined aqueous layers were extracted with three 50
mL portions of Et2O, and the combined organic phases were washed with 40 mL of
brine, dried over Na2SO4, filtered, and concentrated to afford 3.48 g of a dark orange oil.
Column chromatography on 35 g of silica gel (elution with 25% Et2O-pentane) provided
-146-
Part II Synthesis of Substituted Indoles
1.302 g (51%) of the diazo ketone 1329 as orange solid: 1H NMR (400 MHz, CDCl3) δ
6.59 (s, 1H), 5.41 (s, 1H), 2.72 (m, 2H), 2.50 (m, 2H).
13
29
Brisbois, R. G. Application of ɑ-Diazo Ketones to the Synthesis of Highly Substituted Aromatic
Compounds. PhD Thesis, Massachusetts Institute of Technology, May 1990.
-147-
Part II Synthesis of Substituted Indoles
O
O
N2
14
6-Diazo-6,7,8,9-tetrahydro-5H-benzocyclohepten-5-one (14). A 250 mL, two-neck,
round-bottom flask equipped with a rubber septum and argon inlet adapter was charged
with 1,1,1,3,3,3-hexamethyldisilazane (5.00 mL, 3.87 g, 24.0 mmol, 1.3 equiv) and 40
mL of THF. n-Butyllithium solution (2.60 M in hexane, 8.62 mL, 22.4 mmol, 1.2 equiv)
was then added at 0 ºC over 6 min. The resulting yellow solution was stirred at 0 ºC for
40 min, and then cooled to –78 ºC. A solution of benzosuberone (3.00 g, 18.7 mmol, 1
equiv) in 40 mL of THF was added via cannula over 25 min. The reaction mixture was
stirred at –78 ºC for 30 min and then 2,2,2-trifluoroethyl trifluoroacetate (3.25 mL, 4.76
g, 24.3 mmol, 1.3 equiv) was added rapidly (1 s) by syringe in one portion. After 10 min,
the reaction mixture was diluted with 150 mL of Et2O and transferred into a separatory
funnel containing 100 mL of 5% aqueous HCl solution. The aqueous layer was
separated and extracted with three 20 mL portions of Et2O, and the combined organic
phases were washed with brine, dried over Na2SO4, filtered, and concentrated in a 300
mL round-bottom flask to give 7.4 g of a tan oil. The flask was equipped with a stir bar
and rubber septum and flushed with argon. Acetonitrile (40 mL), water (0.33 mL, 0.33 g,
18.3 mmol, 1 equiv), and triethylamine (3.91 mL, 2.84 g, 28.1 mmol, 1.5 equiv) were
then added giving a foggy bright orange solution. A solution of methanesulfonyl azide
(3.80 g, 31.4 mmol, 1.7 equiv) in 40 mL of CH3CN was added via cannula over 15 min,
and the resulting solution was stirred at rt for 14 h and then concentrated to a volume of
ca. 15 mL. The residue was diluted with 100 mL of Et2O, washed with three 30 mL
portions of 3 M aqueous NaOH solution and 40 mL of brine, dried over MgSO4, filtered,
and concentrated to afford 3.03 g of a yellow solid. Column chromatography on 120 g of
silica gel (gradient elution with 10-15% EtOAc–hexanes) provided 2.50 g (72%) of the
diazo ketone 14 as yellow crystalline solid: mp 95–96 ºC; 1H NMR (400 MHz, CDCl3) δ
7.63 (dd, J = 7.6, 1.4 Hz, 1H), 7.41 (td, J = 7.5, 1.5 Hz, 1H), 7.33 (td, J = 7.5, 1.3 Hz,
1H), 7.20 – 7.13 (m, 1H), 2.86 (t, J = 6.9 Hz, 2H), 2.46 (t, J = 6.9 Hz, 2H), 2.09 (p, J =
6.9 Hz, 2H);
13C
NMR (100 MHz, CDCl3) δ 192.57, 138.94, 138.88, 131.89, 129.30,
-148-
Part II Synthesis of Substituted Indoles
127.52, 127.24, 31.42, 28.49 and 22.35.Other spectral characteristics were similar to
those reported previously.30
14
30
Kowalczyk, J. J. Annulation Approaches to Highly Substituted Aromatic Compounds. PhD Thesis,
Massachusetts Institute of Technology, September 1988.
-149-
Part II Synthesis of Substituted Indoles
14
-150-
Part II Synthesis of Substituted Indoles
O
O
N2
Me
+
Ph
N
OH
O
SiMe3
Me
Me
Me
12
9
N
15
OMe
CH(OMe)2
Teoc
Bn
OMe
2-(Trimethylsilyl)ethylbenzyl(2-(2,2-dimethoxyethyl)-3-hydroxy-4,5-dimethyl
phenyl) carbamate (15). A solution of ynamide 9 (0.340 g, 0.936 mmol, 1.0 equiv) in
CH2Cl2 (3.5 mL) was distributed evenly between two 25 cm quartz tubes (I.D. 15 mm)
fitted with rubber septa and argon inlet needles, and purged with argon (10 min). A 15
mL Pyrex tube was charged with diazo ketone 12 (0.310 g, 2.07 mmol, 2.4 equiv), fitted
with a rubber septum and argon inlet needle, and flushed with argon. CH2Cl2 (7 mL)
was added via syringe, the solution was purged with argon (10 min) and taken up in two
5 mL glass syringes fitted with 20 gauge, 20 cm steel needles and wrapped in aluminum
foil. The upper portions (ca. 10 cm) of the quartz tubes were wrapped in aluminum foil
and the tubes were positioned ca. 4 cm from a Hanovia 450 W lamp (quartz immersion
well, cooled by tap water). The quartz reaction tubes and the immersion well were
partially submerged in a room-temperature water bath contained in a 2 L beaker
wrapped in aluminum foil. The diazo ketone solution was added using a syringe pump
over 5 h while irradiating the reaction mixture. The Pyrex tube was rinsed with CH2Cl2
(0.5 mL), the rinse was taken up in the two syringes and added to the reaction tubes.
Irradiation was continued 1 h. The reaction mixtures were combined, concentrated in a
25 mL round-bottom flask, and diluted with toluene (5 mL). The flask was fitted with a
cold-finger condenser, and the solution was heated at reflux for 1 h, allowed to cool to
room temperature, and concentrated to give 0.625 g of an orange oil. Column
chromatography on 20 g neutral alumina (Brockman activity II, gradient elution 0–5%
EtOAc–hexanes) gave 0.292 g of a mixture of unreacted ynamide 9 and phenol 15.
Mixed fractions (0.057 g) were chromatographed using 5 g neutral alumina (Brockman
activity II, gradient elution 5–10% EtOAc–hexanes) to give additional 0.034 g of 9 and
15. The combined mixtures were chromatographed using 25 g silica gel (elution with
10:10:80 Et2O:Benzene:DCM) to give 0.255 g (59%) of phenol 15 as a yellow oil: IR
-151-
Part II Synthesis of Substituted Indoles
(film) 3314, 2952, 1698, 1619, 1573, 1454, 1407, 1311, 1251, 1116, 1093, 1074, 1041
cm-1; 1H NMR (400 MHz, CDCl3), (ca. 70:30 mixture of rotamers) for major rotamer: δ
8.07 (br s, 1H), 7.22–7.29 (m, 5H), 6.39 (br s, 1H), 4.72 (s, 2H), 4.05–4.50 (br m, 3H),
3.36 (s, 3H), 3.27 (s, 3H), 2.56 (d, J = 5.6 Hz, 2H), 2.17 (s, 3H), 2.16 (s, 3H), 0.91 (br m,
2H), -0.06 (s, 9H); additional resonances appeared for the minor rotamer at δ 6.47 (br s,
1H), 1.13 (br m, 2H), 0.08 (s, 9H);
13C
NMR (100 MHz, CDCl3), (mixture of two
rotamers) for major rotamer: δ 156.6, 154.5, 138.4, 137.8, 137.1, 129.6, 128.6, 127.8
124.7, 121.8, 118.8 105.8, 64.3, 55.2, 54.9, 53.7, 30.9, 20.1, 18.1, 12.3, -1.44;
additional resonances appeared for the minor rotamer at δ 129.1, 55.5; HRMS (ESI) [M
+Na]+ calcd for C25H37NO5Si: 482.2333, found 482.2336.
-152-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
15
OH
CH(OMe)2
Me
Me
N
15
Bn
-153S31
Teoc
Part II Synthesis of Substituted Indoles
OH
CH(OMe)2
OH
Me
Me
Me
N
15
Teoc
N
Me
Bn
16
Bn
1-Benzyl-5,6-dimethyl-4-hydroxyindole (16). A 25-mL recovery flask fitted with a
rubber septum and argon inlet needle was charged with a solution of phenol 15 (0.325
g, 0.712 mmol, 1.0 equiv) in 3.5 mL of THF. The solution was cooled at 0 °C while TBAF
solution (1 M in THF, 3.5 mL, 3.5 mmol, 5 equiv) was added via syringe over 3 min. The
resulting brown solution was stirred at rt for 18 h, and then cooled at 0 °C while 6 M
aqueous HCl solution (3.5 mL, 21.0 mmol, 30 equiv) was added dropwise over ca. 2
min. The resulting light yellow solution was stirred at 0 °C for 5 min, and then heated at
reflux for 40 min. The resulting blue-brown mixture was allowed to cool to rt, diluted with
80 mL of Et2O, and washed with two 30 mL portions of satd aq NaHCO3 solution. The
combined aqueous layers were extracted with three 15 mL portions of Et2O, and the
combined organic phases were washed with 50 mL of brine, dried over Na2SO4, filtered,
and concentrated to give 0.184 g of a brown oil. Column chromatography on 20 g of
silica gel (elution with 10% EtOAc− hexanes) gave 0.146 g (82%) of indole 16 as a tan
solid: IR (KBr pellet) 3592, 3054, 2986, 2924, 1570, 1491, 1454, 1442, 1297, 1265,
1236, 909, 705 cm−1; 1H NMR (400 MHz, CDCl3) δ 7.35 – 7.24 (m, 3H), 7.13 – 7.06 (m,
2H), 6.96 (d, J = 3.2 Hz, 1H), 6.76 (s, 1H), 6.49 (d, J = 3.0 Hz, 1H), 5.23 (s, 2H), 5.06 (s,
1H, OH), 2.36 (s, 3H), 2.26 (s, 3H);
13
C NMR (100 MHz, CDCl3) δ 146.1, 137.8, 136.2,
132.5, 128.8, 127.6, 126.7, 126.5, 116.5, 111.5, 103.4, 97.1, 50.1, 21.3, 11.2; HRMS
(ESI) [M + H]+ calcd for C17H18NO: 252.1383, found 252.1371.
-154-
Part II Synthesis of Substituted Indoles
16
-155-
Part II Synthesis of Substituted Indoles
16
-156-
Part II Synthesis of Substituted Indoles
O
Ph
O
N2
N
O
SiMe3
OH
CH(OMe)2
+
N
OMe
13
9
Teoc
Bn
17
OMe
2-(Trimethylsilyl)ethylbenzyl-(4-(2,2-dimethoxyethyl)-3-hydroxybenzocyclobutene)
-5-carbamate (17). A 25 mL quartz reaction tube (13 mm ID, 15 mm OD) fitted with a
rubber septum and an argon inlet needle was charged with ynamide 9 (0.351 g, 0.966
mmol, 1.0 equiv) and flushed with argon. Dichloromethane (3 mL) was then added, and
the reaction tube was positioned ca. 15 cm from a Hanovia 450 W lamp (quartz
immersion well, cooled by tap water). A solution of the diazo ketone 13 (0.287 g, 2.35
mmol, 2.4 equiv) in 7 mL of CH2Cl2 was added to the vigorously stirred reaction mixture
via syringe31 over 3.5 h. Irradiation was continued for 6.5 h and then the reaction
mixture was concentrated to give 0.7 g of a dark red-orange oil. A solution of this oil in
15 mL of toluene was transferred to a flask fitted with a cold-finger condenser with argon
inlet side arm and heated at reflux for 2 h. Concentration gave 0.68 g of a red-brown oil.
Column chromatography on 15 g of silica gel (elution with 15% EtOAc–hexanes)
afforded 0.262 g (59%) of carbamate 17 as a yellow oil: IR (film) 3583, 3297, 2952,
2832, 1698, 1671, 1594, 1431, 1251, 1116, 1073, 859, 838 cm-1; 1H NMR (400 MHz,
CDCl3, ca. 70:30 mixture of rotamers): 8.08 (s, 1H), 7.37-7.15 (b, 5H), 6.27 (s, 1H),
4.14–4.84 (m, 5H), 3.29-3.37 (m, 6H), 3.03–3.11 (m, 4H), 2.60 (d, 2H), 0.89-1.60 (m,
2H), –0.07 (s, 9H); additional resonances appeared for the minor rotamer at δ 6.37 (s,
1H) and 0.07 (s, 9H);
13C
NMR (100 MHz, CDCl3) δ 156.5, 150.5, 145.7, 140.8, 137.5,
130.70, 129.3, 128.8, 128.4, 127.6, 120.3, 116.0, 105.6, 64.1, 55.1, 54.7, 53.4, 30.8,
28.9, 27.1, 17.9, –1.62.
31
The needle and the syringe were wrapped in aluminum foil
-157-
Part II Synthesis of Substituted Indoles
17
-158-
Part II Synthesis of Substituted Indoles
Bn
N
Teoc
17
OH
CH(OMe)2
-159-
Part II Synthesis of Substituted Indoles
OH
CH(OMe)2
N
17
OH
Teoc
Bn
OTf
N
N
Bn
18
Bn
19
N-Benzyl-5,6-dihydro-1H-cyclobuta[f]indol-4-yl trifluoromethanesulfonate (19). A
25 mL recovery flask fitted with a rubber septum and an argon inlet needle was charged
with a solution of phenol 17 (0.170 g, 0.372 mmol, 1.0 equiv) in 2 mL of THF. The
solution was cooled at 0 ºC while TBAF solution (1M in THF, 2 mL, 2 mmol, 5 equiv)
was added via syringe over 1 min. The resulting brown solution was stirred at rt for 24 h,
and then cooled to 0 ºC while 6M HCl solution (2 mL, 12 mmol, 32 equiv)32 was added
dropwise over ca. 2 min. The resulting solution was stirred at 0 ºC for 5 min, and then
heated at reflux for 1 h. The resulting blue-brown mixture was allowed to cool to rt,
diluted with 20 mL of Et2O, and washed with 20 mL of satd aq NaHCO3 solution. The
combined aqueous layers were extracted with Et2O, and the combined organic phases
were washed with brine, dried over MgSO4 containing a small amount of decolorizing
carbon, filtered, and concentrated to give 0.191 g of hydroxyindole 18 as a brown oil
used in the next step without purification: 1H NMR (400 MHz, CDCl3) δ 7.31–7.24 (m,
3H), 7.09–7.11 (m, 2H), 6.98 (d, J = 3.2 Hz, 1H), 6.64 (m, 1H), 6.61 (s, 1H), 5.26 (s,
2H), 3.54 (s, 1H) 3.11– 3.15 (m, 4H).
32
Because of the facile air oxidation of the product indole to the corresponding p-quinone, HCl and all of
the workup and chromatography solutions were purged with argon prior to use.
-160-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
18
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
-161-
Part II Synthesis of Substituted Indoles
The hydroxyindole prepared in the previous step was transferred into a 25 mL recovery
flask equipped with a rubber septum and an argon inlet needle, dissolved in 3 mL of
THF, and cooled to 0 ºC. Sodium hydride (60% wt oil disp., 0.025 g, 0.63 mmol, 1.7
equiv) was then added and the reaction mixture was stirred for 2 min until no more gas
evolved. A solution of N-phenyltriflimide (0.210 g, 0.588 mmol, 1.6 equiv) in 2 mL of THF
was added, and the resulting mixture was stirred at rt for 1 h. Satd aq NaHCO3 (3 mL),
was added, and the resulting mixture was extracted with three 20 mL portions of Et2O.
The combined organic phases were washed with two 20 mL portions of 10% NaOH
solution and 20 mL of brine, dried over MgSO4, filtered and concentrated to give 0.225 g
of a brown oil. Column chromatography on 5 g of silica gel (elution with 10% EtOAc–
hexanes) gave 0.130 g of an oily white solid. Another column chromatography on 5 g of
silica gel (elution with 5% EtOAc–hexanes) gave 0.093 g of a colorless oil. The oil was
taken up in 4 mL of rt pentane, and the solution was cooled to –20 ºC. The resulting
white crystals were washed with cold pentane and dried under vacuum. A second crop
was collected by concentrating the mother liquor and the washes to ca. 0.5 mL and
cooling to –20 ºC. Crystallization provided 0.089 g (63% from 17) of indole 19 as a white
solid: mp 77–78 ºC; IR (KBr pellet) 2926, 1497, 1421, 1250, 1200, 1187, 998, 916, 727,
617 and 603 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.31–7.34 (m, 3H), 7.10–7.14 (m, 3H),
7.00 (s, 1H), 6.62 (d, J = 3.2 Hz, 1H), 5.31 (s, 2H), 3.18–3.36 (m, 4H);
13C
NMR (100
MHz, CDCl3) δ 140.5, 129.1, 128.3, 128.0, 126.9, 126.0, 120.6, 119.0 (q, J =320 Hz,
CF3), 105.3, 98.4, 50.9, 28.5, 27.1; Anal. Calcd for C21H33NO4Si: C, 64.41; H, 8.49; N,
3.58. Found: C, 64.50; H, 8.36; N, 3.53.
-162-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
19
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
-163-
Part II Synthesis of Substituted Indoles
Bn
S42
19
OTf
N
-164-
Part II Synthesis of Substituted Indoles
O
O
N2
Ph
N
OH
O
SiMe3
+
N
9
Teoc
Bn
OMe
14
CH(OMe)2
20
OMe
2-(Trimethylsilyl)ethylbenzyl(5-(2,2-dimethoxyethyl)-6-hydroxy-2,3-dihydro-1Hphenalen-4-yl)carbamate (20). Two 25 cm quartz tubes (13 mm ID, 15 mm OD) were
each charged with ynamide 9 (0.300 g, 0.825 mmol, 1.0 equiv) and diazo ketone 14
(0.231 g, 1.24 mmol, 1.5 equiv). A stir bar was added to each reaction tube, the tubes
were fitted with rubber septa and argon inlet needles, and purged with argon.
Dichloromethane (8.5 mL) was then added to each tube and the reaction tubes were
positioned ca. 3 cm from a Hanovia 450 W lamp (quartz immersion well, cooled by tap
water) in a room-temperature water bath contained in a 2 L beaker wrapped in
aluminum foil. The reaction solutions were irradiated for 2.5 h and then combined and
concentrated to give 1.129 g of an orange oil. A solution of this material in 20 mL of
toluene was transferred to a flask fitted with a cold-finger condenser. The solution was
heated at reflux for 2 h, allowed to cool to rt, and concentrated to give 1.15 g of an
orange oil. Column chromatography on 50 g of neutral alumina (Brockman activity II,
elution with 10% EtOAc–hexanes) gave 0.680 g (79%) of phenol 20 as a yellow oil: IR
(film) 3291, 2949, 2835, 1698, 1620, 1583, 1496, 1398, 1357, 1318, 1250, 1208, 1184,
1141, 1115 cm-1; 1H NMR (400 MHz, CDCl3, ca. 80:20 mixture of rotamers) for major
rotamer: δ 8.50 (s, 1H), 8.14 (d, J = 8.4 Hz, 1H), 7.41–7.36 (m, 1H), 7.31-7.10 (m, 6H),
4.84 (d, J = 13.9 Hz, 1H), 4.58 (d, J = 13.9 Hz, 1H), 4.44–4.04 (m, 3H), 3.37 (s, 3H),
3.27 (s, 3H), 3.01 (t, J = 5.9 Hz, 2H), 2.69 (d, J = 6.9 Hz, 2H), 2.84–2.42 (m, 2H), 1.81
(pq, J = 6.0 Hz, 2H), 0.95–0.75 (m, 2H), and 0.10 (s, 9H); additional resonances
appeared for the minor rotamer at δ: 8.48 (s, 1H), 8.13 (d, J = 8.3 Hz, 1H), 4.48 (d, J =
14.2 Hz, 1H) 3.39 (s, 3H), 3.28 (s, 3H), 2.73 (d, J = 6.9 Hz, 2H), 1.89 (pq, J = 6.1 Hz,
2H) 1.26–1.18 (m, 2H), 0.12 (s, 9H);
13C
NMR (100 MHz, CDCl3, mixture of two
rotamers) for major rotamer: δ 156.7, 150.8, 137.0, 136.3, 135.0, 130.1, 129.9, 128.5,
128.0, 125.7, 125.5, 125.2, 124.8, 120.6, 114.9, 106.0, 64.2, 55.1, 54.4, 53.0, 31.36,
-165-
Part II Synthesis of Substituted Indoles
31.34, 26.8, 22.8, 18.1, –1.5; additional resonances appeared for the minor rotamer at
δ: 155.7, 136.9, 135.5, 130.4, 130.2, 128.0, 125.1, 124.6, 115.4, 106.2, 64.3, 55.6, 54.3,
52.9, 31.5, 22.9, 18.2, –1.2; HRMS (ESI) [M+Na]+ calcd for C30H39NO5Si: 544.2490,
found 544.2502.
-166-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
20
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
-167-
Part II Synthesis of Substituted Indoles
Bn
Teoc
N
20
OH
CH(OMe)2
S34
-168-
Part II Synthesis of Substituted Indoles
OH
CH(OMe)2
N
OTf
Teoc
CH(OMe)2
N
Bn
Teoc
Bn
20
21
4-(Benzyl((2-(trimethylsilyl)ethoxy)carbonyl)amino)-5-(2,2-dimethoxyethyl)-2,3dihydro-1H-phenalen-6-yl trifluoromethanesulfonate (21). A 25 mL recovery flask
was charged with phenol 20 (0.233 g, 0.447 mmol, 1 equiv) and 4 mL of THF and
cooled to 0 ºC. Sodium hydride (60% wt oil disp., 0.025 g, 0.63 mmol, 1.4 equiv) was
then added and the reaction mixture was stirred for ca. 5 min until gas evolution ceased.
A solution of N-phenyltriflimide (0.224 g, 0.627 mmol, 1.4 equiv) in 2 mL of THF was
added, and the resulting brown mixture was allowed to warm to rt over 15 min. The
reaction mixture was cooled to 0 ºC, and 3 mL of satd aq NaHCO3 solution was added.
The resulting mixture was diluted with 15 mL of Et2O, separated and the aqueous layer
was extracted with two 15 mL portions of Et2O. The combined organic phases were
washed with two 20 mL portions of 2.5 M NaOH solution and 20 mL of brine, dried over
MgSO4, filtered and concentrated to give 0.329 g of a yellow oil. Column
chromatography on 15 g of silica gel (elution with 10% EtOAc–hexanes) gave 0.219 g
(75%) of trifluoromethanesulfonate 21 as a viscous colorless oil: IR (film) 3583, 3384,
2953, 2835, 1703, 1597, 1403, 1318, 1248, 1215, 1138, 1121, 1074, 943 and 729 cm-1;
1H
NMR (600 MHz, CDCl3) δ 8.00 (d, J = 8.5 Hz, 1H), 7.57 (m, 1H), 7.38–7.14 (m, 6H),
5.35 (d, J = 14.3 Hz ,1H), 4.76 (m, 1H), 4.16–4.56 (m, 3H), 3.35–3.46 (m, 3H), 3.29–
3.35 (m, 3H), 2.96–3.13 (m, 3H), 2.54 (m, 1H), 2.14–2.27(m, 1H), 1.27–1.96 (m, 4H),
0.87–1.00 (m, 2H), 0.00 (s, 9H); additional resonances for the minor rotamer appeared
at δ 5.19 (d, J=14 Hz, 1H), 4.70 (m, 1H) 0.20 (s, 9H);
13C
NMR (100 MHz, CDCl3,
mixture of two rotamers) for major rotamer: δ 156.7, 150.8, 137.0, 136.3, 135.0, 130.1,
129.9, 128.5, 128.0, 125.7, 125.5, 125.2, 124.8, 120.6, 114.9, 106.0, 64.2, 55.1, 54.4,
53.0, 31.36, 31.34, 26.8, 22.8, 18.1, –1.5; additional resonances appeared for the minor
rotamer at δ: 155.7, 136.9, 135.5, 130.4, 130.2, 128.0, 125.1, 124.6, 115.4, 106.2, 64.3,
-169-
Part II Synthesis of Substituted Indoles
55.6, 54.3, 52.9, 31.5, 22.9, 18.2, –1.2; HRMS (ESI) [M+Na]+ calcd for C31H85F3NO7SSi:
676.1983, found 676.1986.
-170-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
21
-171S35
Part II Synthesis of Substituted Indoles
OTf
CH(OMe)2
N
OTf
Teoc
N
Bn
Bn
21
22
N-Benzyl-1,2,3,10-tetrahydronaphtho[1,8-fg]indol-7-yl trifluoromethanesulfonate
(22) A 25 mL recovery flask was charged with carbamate 21 (0.189 g, 0.289 mmol, 1
equiv) and 5 mL of CH2Cl2. Trifluoromethanesulfonic acid (0.225 mL, 2.88 mmol, 10
equiv) was added via syringe in one portion producing a dark red-brown solution. The
reaction mixture was stirred vigorously at rt for 5 min and then 2 mL of water was
added. This biphasic reaction mixture turned light yellow within 30 sec. Satd aq
NaHCO3 (5 mL) was then added33, and the reaction mixture was diluted with 20 mL
CH2Cl2 and separated. The organic phase was washed with 10 mL of water and brine,
dried over MgSO4, filtered, and concentrated to give 0.117 g of a tan solid. Column
chromatography on 10 g of silica gel (elution with 20% benzene–hexanes34) followed by
a column chromatography on 6 g of silica gel (elution with 20% benzene–hexanes)
furnished 0.098 g (76%) of indole 22 as a white powder: mp 112–113 ºC (dec); 1H NMR
(600 MHz, CDCl3) δ 7.97 (d, J = 8.7 Hz, 1H), 7.36–7.40 (m, 1H), 7.25–7.35 (m, 4H),
7.15 (dd, J = 10.0, 3.9 Hz, 1H), 6.96 (d, J = 7.3 Hz, 2H), 6.82 (d, J = 3.4 Hz, 1H), 5.69
(s, 2H), 3.41 (t, J = 6.2 Hz, 2H), 3.06 (t, J = 6.2 Hz, 2H), 2.02–1.99 (m, 2H); 13C NMR
(100 MHz, CDCl3) δ 139.0, 136.6, 136.1, 135.0, 134.4, 129.2, 127.8, 127.2, 125.6,
124.5, 123.9, 122.6, 121.7, 119.1 (q, J = 310 Hz), 118.6, 118.3, 98.1, 53.2, 31.4, 26.9,
22.7; HRMS (ESI) [M+H]+ calcd for C23H18F3NO3S: 446.1032, found 446.1020.
33Quick
neutralization is essential once the product indole forms to avoid acid-promoted dimerization/
decomposition.
34 In a different experiment indole 22 was purified by eluting with Et O–pentane. This avoided
2
decomposition that was observed when benzene–hexanes and EtOAc–hexanes eluents were used.
-172-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
22
-173S43
Part II Synthesis of Substituted Indoles
Et
OTf
N
N
Bn
Bn
22
25
N-Benzyl-7-ethyl-1,2,3,10-tetrahydronaphtho[1,8-fg]indole (25) A 25 mL pear flask
was charged with Pd(OAc)2 (2.5 mg, 0.011 mmol, 5 mol %), SPhos (6.6 mg, 0.016
mmol, 7.5 mol %), and Cs2CO3 (0.208 g, 0.638 mmol, 3 equiv). The flask was equipped
with a rubber septum fitted with an argon inlet needle and flushed with argon. A solution
of triflate 22 (0.095 g, 0213 mmol, 1 equiv) in 4.5 mL of THF was added, followed by a
solution of Et3B (1 M in THF, 0.533 mL, 0.533 mmol, 2.5 equiv). The orange-yellow
reaction mixture was stirred at rt for 1 h, diluted with 15 mL of Et2O, and filtered through
5 g of silica gel with the aid of two 20 mL portions of Et2O. The resulting yellow solution
was concentrated to give 0.061 g of an off-white solid. This was dissolved in 50 mL of
hot pentane and crystallized out at –78 ºC to give 0.055 g (79%) of indole 25 as an offwhite powder: mp 114–115 ºC; IR (film) 3054, 2936, 1453,1386, 1355, 1265, 739 cm−1;
1H
NMR (600 MHz, CDCl3) δ 8.07 (d, J = 9 Hz, 1H), 7.27–7.34 (m, 5H), 7.13 (d, J = 3.6
Hz, 1H), 7.01 (d, J = 7.2 Hz, 2H), 6.81 (d, J = 3.0 Hz, 1H), 5.70 (s, 2H), 3.45–3.48 (m,
4H), 3.11 (t, J = 6 Hz, 2H), 2.02–2.06 (m, 2H), 1.49 (t, J = 7.2 Hz, 3H);
13C
NMR (150
MHz, CDCl3) δ 140.0, 136.3, 134.5, 134.1, 129.9, 129.3, 129.0, 127.5, 127.4, 126.2,
125.8, 122.2, 114.8, 100.1, 53.2, 32.1, 27.2, 23.1, 22.9, 15.5; HRMS (ESI) [M+H]+ calcd
for C24H23N: 326.1903, found 326.1900.
-174-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
25
-175S44
Part II Synthesis of Substituted Indoles
N2
Me
+
Me
12
Me
O
O
MeO
Me
N
HO
8
N
CO2Me
OSit-BuMe 2
OSit-BuMe2
26
Methylallyl(2-(2-(tert-butyldimethylsilyloxy)ethyl)-3-hydroxy-4,5-dimethylphenyl)
carbamate (26). Two 25 cm quartz tubes (13 mm ID, 15 mm OD) fitted with rubber
septa and argon inlet needles were charged with ynamide 8 (tube A: 0.168 g, 0.565
mmol, 1.0 equiv; tube B: 0.151 g, 0.508 mmol, 1.0 equiv) and 2 mL of CH2Cl2 each. A 25
mL recovery flask was charged with a solution of diazo ketone 12 (0.340 g, 2.74 mmol,
2.6 equiv) in 7 mL of CH2Cl2. All solutions were purged with argon for 15 min. The upper
portions (ca. 10 cm) of the quartz reaction tubes were wrapped in aluminum foil and the
tubes were positioned ca. 5 cm from a Hanovia 450 W lamp (quartz immersion well,
cooled by tap water). The reaction tubes and the immersion well were partially
submerged in a room-temperature water bath contained in a 2 L beaker wrapped in
aluminum foil.35 The diazo ketone solution was taken up into two 5 mL gastight syringes
fitted with 20 gauge, 20 cm steel needles and wrapped in aluminum foil, and added to
the reaction tubes using a syringe pump over 7 h (1 mL CH2Cl2 rinse).36 Irradiation was
continued for 20 min, and then the reaction mixtures were combined and concentrated
in a 25 mL round-bottomed flask. The residue was diluted with 7 mL of toluene, the flask
was fitted with a cold-finger condenser with an argon inlet sidearm, and the solution was
heated at reflux for 3 h, allowed to cool to rt, and then concentrated to give 0.7 g of a
dark-orange oil. Column chromatography on 50 g of silica gel (elution with 10% EtOAc–
hexanes) gave 0.328 g (78%) of phenol 26 as a yellow oil which crystallized upon
refrigeration: mp 48–49 ºC; IR (neat) 3273, 2859, 2930, 1709, 1620, 1574, 1449, 1388,
1313, 1259, 1214, 1192, 1147, 1092, 1039 cm-1; 1H NMR (400 MHz, CDCl3) δ 8.59 (s,
35
The reaction tubes were positioned such as to allow efficient stirring with a magnetic stir motor. This
required that the bottoms of the reaction tubes were lower than the lamp’s anode, as opposed to having
the solution aligned with the center of the arc.
36 The bath temperature rose to 29 ºC during the irradiation.
-176-
Part II Synthesis of Substituted Indoles
1H), 6.51 (s, 1H), 5.85–5.95 (m, 1H), 5.12 (m, 2 H), 3.70–4.30 (m, 4H), 3.62 (s, 3H),
2.78 (m, 2H), 2.23 (s, 3H), 2.17 (s, 3H), 0.92 (s, 9H), 0.08 (d, J = 5.6 Hz, 6H);
13C
NMR
(100 MHz, CDCl3) δ 156.6, 154.7, 137.6, 136.7, 133.3, 124.7, 122.4, 121.2, 118.4, 65.3,
54.3, 53.0, 29.5, 25.9, 20.2, 18.4, 12.3, –5.5 ; HRMS (ESI) [M+Na]+ calcd for
C21H35NO4Si: 416.2228, found 416.2218.
-177-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
26
-178S36
Part II Synthesis of Substituted Indoles
Me
Me
Me
HO
N
OH
Me
HO
Me
H
HO
N
HO
N
CO2Me
OSit-BuMe 2
26
Me
CO2Me
OSit-BuMe 2
27
O
CO2Me
OSit-BuMe 2
28
Methyl 2-(2-(tert-butyldimethylsilyl)oxy)ethyl)-3-hydroxy-4,5-dimethylphenyl(2oxoethyl)carbamate (28). A 50 mL recovery flask fitted with a rubber septum and argon
inlet needle was charged with alkene 26 (0.325 g, 0.826 mmol, 1.0 equiv), 6 mL of THF,
2 mL of H2O, and OsO4 (4 wt% in H2O, 0.11 mL, 0.11 g, 0.018 mmol, 0.02 equiv). The
solution was stirred at rt for 10 min and then NMO (0.130 g, 1.11 mmol, 1.3 equiv) was
added. The resulting cloudy tan mixture was stirred at rt for 24 h. Aqueous NaHSO3
solution (1.0 M, 9.0 mL, 9.0 mmol, 11 equiv) was added, and the reaction mixture was
stirred at rt for 10 min, diluted with 15 mL of brine and extracted with three 20 mL
portions of EtOAc. The combined organic phases were dried over MgSO4, filtered, and
concentrated to give 0.401 g of diol 27 as a pale yellow oil used in the next step without
purification: IR (neat) 3281, 2955, 2931, 2884, 2860, 1685, 1619, 1573, 1455, 1391,
1321, 1260, 1197, 1164, 1092, 1039, 900, 839, 781, 732 cm-1; 1H NMR (400 MHz,
CDCl3), (ca. 50:50 mixture of rotamers) δ 8.69 and 8.60 (s, 1H), 6.60 and 6.42 (s, 1H),
3.81– 3.73 (m, 2H), 3.73–3.59 (m, 6H), 2.94–2.70 (m, 3H), 2.22 (two d, J = 8.9 Hz, 3H),
2.17 (s, 3H), 0.92 (two s, 9H), 0.10–0.08 (m, 6H).
-179-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
27
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
-180-
Part II Synthesis of Substituted Indoles
A 25 mL recovery flask was charged with NaIO4 supported on silica gel (0.68
mmol/g, 2.5 g, 1.7 mmol, 2.0 equiv) and 5 mL of CH2Cl2. A solution of the diol 27
prepared above in 5 mL of CH2Cl2 was added to the stirring suspension via pipet (0.5
mL CH2Cl2 rinse). The reaction mixture was stirred at rt for 0.5 h and then filtered
through a sintered glass funnel with the aid of three 3 mL portions of CH2Cl2.
Concentration of the filtrate gave 0.316 g (97%) of aldehyde 28 as a pale yellow oil: IR
(neat) 3259, 2955, 2931, 2885, 2859, 1706, 1619, 1574, 1451, 1387, 1331, 1260, 1136,
1093, 1053, 1038, 899, 839, 782 cm-1; 1H NMR (600 MHz, CDCl3), (ca. 80:20 mixture of
rotamers) for major rotamer: δ 9.71 (s, 1H), 8.59 (s, 1H), 6.61 (s, 1H), 4.43 (d, J = 18
Hz, 1H), 4.03 (d, J = 18 Hz, 1H), 3.87 (m, 2H), 3.68 (s, 3H), 2.88 (m, 2H), 2.23 (s, 3H),
2.17 (s, 3H), 0.92 (s, 9H), 0.09 (d, J = 6 Hz, 6H); additional resonances appeared for the
minor rotamer at δ 8.65 (s, 1H), 4.34 (d, J = 18, 2H), 3.75 (s, 3H);
13C
NMR (100 MHz,
CDCl3) δ 197.3, 156.9, 154.9, 137.8, 137.2, 125.3, 122.2, 120.6, 65.2, 61.1, 53.5, 29.3,
25.9, 20.1, 18.4, 12.3, –5.5; additional resonance appeared for the minor rotamer at δ
120.4; HRMS (ESI) [M+H]+ calcd for C20H33NO5Si: 396.2201, found 396.2216.
-181-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
28
-182S46
Part II Synthesis of Substituted Indoles
Me
Me
Me
H
HO
N
O
Me
N
HO
CO2Me
CO2Me
OSit-BuMe 2
OH
28
29
4,5-Dimethyl-6-hydroxy-7-(2-hydroxyethyl)-1-methoxycarbonylindole (29). A 50 mL
recovery flask equipped with a cold finger condenser with an argon inlet sidearm was
charged with aldehyde 28 (0.316 g, 0.800 mmol, 1.0 equiv), 40 mL of isopropanol, and
K2CO3 (0.11 g, 0.80 mmol, 1.0 equiv). The pale-yellow reaction mixture was heated at
75 ºC for 2.5 h. The oil bath was removed, the warm mixture was diluted with 15 mL of
H2O, and then treated with 5 mL of aq 1M HCl to adjust the pH to 1. The resulting
mixture was allowed to stir at rt for 0.5 h and then concentrated to a volume of ca. 10
mL. The resulting heterogeneous mixture was diluted with 200 mL of Et2O and washed
with two 40 mL portions of H2O and 40 mL of brine. The aqueous layers were extracted
with two 20 mL portions of Et2O, and the combined organic phases were dried over
MgSO4, filtered, and concentrated to afford 0.209 g of a brown oil. Column
chromatography on 7 g of silica gel (elution with 50% EtOAc–hexanes) afforded 0.130 g
(62%) of indole 29 as a white solid: mp 116–117 °C; IR (KBr pellet) 3472, 3311, 3153,
3113, 2994, 2952, 2891, 1719, 1560, 1450, 1351, 1294, 1252, 1156, 1041, 930 cm-1; 1H
NMR (400 MHz, CDCl3) δ 8.79 (s, 1H), 7.42 (d, J = 4.0 Hz, 1H), 6.56 (d, J = 4.0 Hz,
1H), 4.36 (m, 2H), 3.93 (s, 3H), 3.13 (t, J = 4.8 Hz, 2H), 2.43 (br s, 1H), 2.41 (s, 3H),
2.30 (s, 3H);
13C
NMR (100 MHz, CDCl3) δ 153.0, 152.4, 133.7, 127.8, 125.6, 125.4,
121.5, 111.9, 107.4, 66.3, 54.0, 31.1, 15.8, 12.6; HRMS (ESI) [M+H]+ calcd for
C14H17NO4: 264.1230, found 264.1231.
-183-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
29
-184S47
Part II Synthesis of Substituted Indoles
O
MeO
O
N2
13
N
+
HO
8
30
N
CO2Me
OSit-BuMe 2
OSit-BuMe2
Methyl allyl(4-(2-((tert-butyldimethylsilyl)oxy)ethyl)-3-hydroxybenzocyclobutene)
-5-carbamate (30) An oven-dried Pyrex photochemical reactor vessel containing a
water-cooled quartz immersion well and a Hanovia 450 W lamp was charged with
ynamide 8 (2.11 g, 7.10 mmol, 1 equiv), diazo ketone 13 (1.08 g, 8.85 mmol, 1.25
equiv), and 80 mL of CH2Cl2. The reaction solution was purged with CH2Cl2-saturated
argon37 and irradiated for 2 h. Argon flow was continued as a means of agitating the
reaction mixture. A second portion of the diazo ketone (1.08 g, 8.85 mmol, 1.25 equiv) in
20 mL of CH2Cl2 was then added, and irradiation was continued for 2 h. The reaction
solution was transferred to a 500 mL round-bottom flask and concentrated to give 4.55 g
of a red-orange oil. Toluene (200 mL) was then added and the flask was equipped with
a reflux condenser carrying an argon inlet adapter. The solution was heated at reflux for
2 h, and then concentrated to give 5 g of a red-brown oil. Column chromatography on
120 g of silica gel (elution with 10% EtOAc–hexanes) gave 1.493 g (54%)38 of
carbamate 30 as a white powder: mp 129–130 °C; IR (KBr pellet) 3317, 2928, 1679,
1593, 1457, 1432, 1310, 1258, 1148, 1089, 982, 921, 859, 834 and 773 cm-1; 1H NMR
(400 MHz, CDCl3) δ 8.64 (s, 1H), 6.45 (s, 1H), 5.87–5.93 (m, 1H), 5.11–5.14 (m, 2H),
4.31 (dd, J = 19.1, 10.1 Hz, 1H), 3.59–3.91(m, 6H), 3.09–3.15 (m, 4H), 2.74–2.85 (m,
2H), 0.92 (s, 9H), 0.10 (d, J = 4.1 Hz, 6H);
13C
NMR (100 MHz, CDCl3) δ 156.7, 151.0,
145.6, 133.3, 130.8, 124.2, 118.5, 118.4, 115.5, 65.3, 54.4, 53.0, 29.7, 29.2, 27.4, 26.0,
18.5, –5.4; Anal. Calcd for C21H33NO4Si: C, 64.41; H, 8.49; N, 3.58. Found: C, 64.50; H,
8.36; N, 3.53.
37
CH2Cl2-saturated argon was obtained by adding a tall Drechel bottle filled with CH2Cl2 into the argon
line.
38On scales up to 0.250 g using the setup and procedure as for 26 61-75% yield was obtained.
-185-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
30
-186S37
Part II Synthesis of Substituted Indoles
OH
HO
HO
N
HO
N
CO2Me
OSit-BuMe 2
30
H
HO
N
CO2Me
OSit-BuMe 2
31
O
CO2Me
OSit-BuMe 2
32
Methyl 2-oxoethyl(4-(2-((tert-butyldimethylsilyl)oxy)ethyl)-3-hydroxybenzocyclo
butene)-5-carbamate (32). A 100 mL recovery flask fitted with a rubber septum and
argon inlet needle was charged with alkene 30 (1.22 g, 3.12 mmol, 1.0 equiv), 22.5 mL
of THF, 7.5 mL of H2O, OsO4 (4 wt% in H2O, 0.5 mL, 0.08 mmol, 0.03 equiv), and NMO
(0.552 g, 4.71 mmol, 1.5 equiv). The resulting yellow mixture was stirred at rt for 26 h.
Aqueous NaHSO3 solution (2.0 M, 9.0 mL, 9.0 mmol, 9 equiv) was added, and the
reaction mixture was stirred at rt for 10 min. The resulting mixture was transferred to a
separatory funnel containing 60 mL of brine and extracted with three 40 mL portions of
Et2O. The combined organic phases were dried over MgSO4, filtered, and concentrated
to give 1.4 g of diol 31 as an off-white foam used in the next step without purification: IR
(neat) 3345, 2939, 2858, 1680, 1595, 1455, 1391, 1254, 1084, 914, 836 and 777 cm-1;
1H
NMR (400 MHz, CDCl3), (ca. 50:50 mixture of rotamers): δ 8.75 (s, 1H), 6.55 (s, 1H),
3.55–3.91 (m, 9H), 3.35–3.50 (m, 3H), 3.05–3.15 (m, 4H), 2.61–2.85 (m, 2H), 0.90 (s,
9H), 0.09 (m, 6H); additional resonances appeared for the second rotamer at δ 8.63 (s,
1H), 6.38 (s, 1H).
-187-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
31
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
-188-
Part II Synthesis of Substituted Indoles
A 100 mL recovery flask was charged with NaIO4 supported on silica gel (0.68
mmol/g, 10.6 g, 7.2 mmol, 2.3 equiv) and 20 mL of CH2Cl2. A solution of the diol 31
prepared above in 15 mL of CH2Cl2 was added to the stirring suspension via pipet (two
5 mL CH2Cl2 rinses). The reaction mixture was stirred at rt for 20 min and then filtered
through 1 g of silica gel with the aid of CH2Cl2. Concentration of the filtrate gave 1.08 g
(88%) of aldehyde 32 as a white powder: mp 134–135 ºC; 1H NMR (400 MHz, CDCl3),
(ca. 80:20 mixture of rotamers) for major rotamer: δ 9.69 (s, 1H), 8.65 (s, 1H), 6.58 (s,
1H), 3.75–4.51 (m, 4H), 3.68 (s, 3H), 3.05–3.14 (m, 4H), 2.87–2.90 (m, 2H), 0.92 (s,
9H), 0.10 (d, J = 2.8 Hz, 6H); additional resonances appeared for the minor rotamer at δ
8.71 (s, 1H), 3.35 (s, 3H);
13C
NMR (100 MHz, CDCl3) δ 197.2, 157.0, 151.2, 146.2,
140.5, 131.5, 124.1, 115.1, 65.3, 61.3, 53.6, 29.6, 29.2, 27.5, 26.0, 18.5, –5.42; Anal.
Calcd for C20H31NO5Si: C, 61.04; H, 7.94; N, 3.56. Found: C, 60.77; H, 7.96; N, 3.52.
-189-
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
32
-190S49
Part II Synthesis of Substituted Indoles
H
HO
O
N
N
HO
CO2Me
CO2Me
OSit-BuMe 2
OSit-BuMe 2
32
33
N-Carbomethoxy 4-(2-((tert-butyldimethylsilyl)oxy)ethyl)-5-hydroxy-6,7-dihydro3H-cyclobuta[e]indole (33). A 300 mL round-bottom flask was charged with aldehyde
32 (0.735 g, 1.87 mmol, 1.0 equiv), 100 mL of isopropanol, and K2CO3 (0.260 g, 1.89
mmol, 1.0 equiv). The reaction flask was equipped with a reflux condenser fitted with an
argon inlet adapter, and the mixture was heated at 80 ºC for 10 h. The resulting black
mixture was allowed to cool to rt, neutralized with 3.3 mL of aq 0.6 M HCl, and
concentrated to a volume of ca. 20 mL. This mixture was extracted with three 50 mL
portions of Et2O, and the combined organic phases were washed with brine, dried over
MgSO4, and concentrated to afford 0.82 g of a brown oil. This oil was taken up in 200
mL of hexanes and filtered, and the filtrate was concentrated to give 0.577 g of a lightbrown oil. Column chromatography on 25 g of silica gel (elution with 5% EtOAc–
hexanes) afforded 0.337 g (48%) of indole 33 as a white powder: mp 90–91 °C; IR (KBr
pellet) 3248, 2928, 2857, 1751, 1608, 1548, 1465, 1440, 1371, 1336, 1273, 1256, 1174,
1145, 1041, 901, 786 and 719 cm-1; 1H NMR (400 MHz, CDCl3) δ 9.14 (s, 1H), 7.45 (d,
J = 4.0 Hz, 1H), 6.40 (d, J = 4.0 Hz, 1H), 4.30 (m, 2H), 3.94 (s, 3H), 3.24–3.30 (m, 4H),
3.08 (t, J = 4.4 Hz, 2H), 0.98 (s, 9H), 0.18 (s, 6H);
13C
NMR (100 MHz, CDCl3) δ 152.4,
149.2, 136.9, 136.7, 128.0, 127.2, 121.4, 114.2, 105.6, 67.2, 54.0, 32.3, 28.6, 28.0,
26.1, 18.5, –5.31; Anal. Calcd for C20H29NO4Si: C, 63.97; H, 7.78; N, 3.73. Found: C,
64.20; H, 7.50; N, 3.46.
-191-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
33
-192S50
Part II Synthesis of Substituted Indoles
Ph
N
O
Ph
+
O
N
H
H
O
N
HO
O
HO
CO 2Me
N
CO 2Me
OSit-BuMe 2
OSit-BuMe 2
35
33
Methyl 4-(2-((tert-butyldimethylsilyl)oxy)ethyl)-5-hydroxy-7,9-dioxo-8-phenyl-6a,
7,8,9,9a,10-hexahydroisoindolo[5,6-e]indole-3(6H)-carboxylate (35). A 35 mL
pressure tube39 fitted with a rubber septum was charged with N-phenylmaleimide (0.183
g, 1.06 mmol, 3.6 equiv) and flushed with argon. A solution of indole 33 (0.110 g, 0.293
mmol, 1 equiv) in 12 mL of p-xylene was added, the pressure tube was sealed with a
Teflon plug, and heated at 180 ºC for 40 h. The reaction mixture was allowed to cool to
rt and concentrated to give 0.376 g of a dark orange oil. Column chromatography on 15
g of silica gel (elution with 20% EtOAc–hexanes) gave 0.070 g (44%) of indole 35 as a
white powder: mp 172−174 °C; IR (KBr pellet) 3449, 2926, 1625, 1569, 1497, 1455,
1438, 1380, 1320, 1287, 1250, 1220, 1140, 998, 916, 840 and 727 cm-1; 1H NMR (400
MHz, CDCl3) δ 9.24 (s, 1H), 7.19–7.50 (m, 4H), 6.97 (d, J = 7.2 Hz, 2H), 6.59 (d, J = 3.6
Hz, 1H), 4.23–4.28 (m, 2H), 3.94 (s, 3H), 2.85–3.80 (m, 8H), 0.92 (s, 9H), 0.11 (s, 3H),
0.09 (s, 3H);
13C
NMR (100 MHz, CDCl3) δ 179.0, 178.8, 152.22, 152.17, 134.6, 132.1,
129.4, 129.1, 128.5, 126.7, 126.5, 123.8, 119.8, 113.6, 106.2, 67.2, 54.1, 40.29, 40.26,
31.8, 26.2, 26.0, 22.7, 18.4, –5.42, –5.46; HRMS (ESI) [M + H]+ calcd for C30H36N2O6Si:
549.2415, found 549.2419.
39
Ace 8648-078648-07 TUBE,PRESSURE,35ML,150psi,25.4MMOD,17.8CM LONG,COMPLETE WITH
#15 FRONT-SEAL PLUG
-193-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
35
-194S51
Part II Synthesis of Substituted Indoles
EtO 2C
O
+
N
HO
OEt
N
HO
CO 2Me
CO 2Me
OSit-BuMe 2
OSit-BuMe 2
33
36&37
8-Ethyl 3-methyl 4-(2-((tert-butyldimethylsilyl)oxy)ethyl)-5-hydroxy-3H-benzo[e]
indole-3,8(6H,9H)-dicarboxylate (36) and 7-Ethyl 3-methyl 4-(2-((tert-butyldimethyl
silyl)oxy)ethyl)-5-hydroxy-3H-benzo[e]indole-3,7(6H,9H)-dicarboxylate (37) A 35
mL pressure tube40 was charged with a solution of indolocyclobutene 33 (0.109 g, 0.290
mmol, 1 equiv) in 10 mL of p-xylene followed by ethyl propiolate (108 µL, 1.07 mmol,
3.7 equiv). The tube was sealed with a Teflon plug and heated at 180 ºC for 40 h. The
reaction mixture was allowed to cool to rt and concentrated to give 0.159 g of an orange
solid. Column chromatography on 15 g of silica gel (elution with 10% EtOAc–hexanes)
gave 0.085 g (62%) of an inseparable ca. 40:60 mixture of isomers 36 and 37 as a
white powder: IR (KBr pellet) 3250, 2956, 2931, 2884, 2858, 1750, 1714, 1667, 1440,
1337, 1292, 1257, 1230, 1144, 1038, 918, 838, 781 and 719 cm-1; 1H NMR (600 MHz,
CDCl3), (ca. 60:40 mixture of regioisomers) for major isomer: δ 9.24 (s, 1H), 7.53 (d, J =
2.8 Hz, 1H), 7.34–7.35 (m, 1H), 6.67 (d, J = 2.4 Hz, 1H), 4.35–4.39 (m, 4H), 4.02 (s,
3H), 3.72–3.82 (m, 2H), 3.67–3.71(m, 2H), 3.22–3.24 (m, 2H), 1.43–1.45 (m, 3H), 1.03
(s, 9H), 0.21 (s, 6H); for the minor isomer: δ 9,23 (s, 1H), 7.52 (d, J = 2.4 Hz, 1H), 7.24–
7.35 (m, 1H), 6.56 (d, J = 2.4 Hz, 1H), 4.35–4.39 (m, 4H), 4.02 (s, 3H), 3.72–3.82 (m,
2H), 3.67–3.71(m, 2H), 3.22–3.24 (m, 2H), 1.43–1.45 (m, 3H), 1.0 (s, 9H), 0.21 (s, 6H);
13C
NMR (100 MHz, CDCl3) δ 153.0, 152.6, 152.2, 137.2, 134.9, 134.1, 129.0, 127.0,
125.8, 124.7, 123.8, 123.3, 123.1, 119.1, 117.6, 112.5, 112.4, 106.5, 106.3, 67.3, 60.7,
60.6, 54.1, 31.7, 31.7, 28.4, 26.7, 26.3, 26.1, 26.0, 24.4, 18.5, 18.4, 14.6, –5.3, –5.4;
HRMS [M+H]+ calcd for C25H35NO6Si: 474.2306, found 474.2312.
40Ace
8648-078648-07 TUBE,PRESSURE,35ML,150psi,25.4MMOD,17.8CM LONG,COMPLETE WITH
#15 FRONT-SEAL PLUG
-195-
Willumstad, T. P.; Haze, O.; Mak, X. Y.;
Lam, T. Y.; Wang, Y.-P.; Danheiser, R. L.
Batch and Flow Photochemical Benzannulations Based on the
Reaction of Ynamides and Diazo Ketones. Application to the
Synthesis of Polycyclic Aromatic and Heteroaromatic Compounds
Part II Synthesis of Substituted Indoles
36&37
-196S52