An Improved Analysis of the Sevoflurane⋯Benzene Structure
by CP-FTMW Spectroscopy
Nathan A. Seifert, Cristobal Perez, Daniel P. Zaleski, Justin L. Neill, Brooks H. Pate
University of Virginia
Alberto Lesarri, Montserrat Vallejo-López
Universidad de Valladolid
Emilio J. Cocincero, Fernando Castaño
Universidad del País Vasco
Isabelle Kleiner
Laboratore Interuniversitaire des Systèmes Atmosphériques (LISA)
A recap from last year
2012, MH13:
What we could achieve in previous study:
• Spectrum complicated by
high barrier, six-fold internal rotation
• BELGI-C1 fit of internal rotation with
135 a-type transitions,
determined V6 = 32.88(3) cm-1
• Experimental rs coordinates for benzene
hydrogens using sevoflurane⋯benzene-d1
• Sevoflurane monomer coordinates
derived from monomer FTMW study1
• Sensitivity insufficient for any 13C isotopologues
in natural abundance!
Open problems from last year:
• Sensitivity of instrument too low to achieve full experimental structure
• μa = 2.3 D, μc = 1.7 (DFT values); no c-types detected in original spectrum, however.
• The existence of other conformers unclear (Suhm and coworkers report two in IR study2)
1. Lesarri, A., et al. Phys. Chem. Chem. Phys. 2010, 12, 9624.
2. Dom, J. J., et al. Phys. Chem. Chem. Phys. 2011, 13, 14142.
Experimental
New improvements to 2-8 GHz
(see Cristobal’s talk, TH10, for more information)
• New horns with improved directionality
enable 5 pulsed nozzles (1-2 nozzles previously)
• New TWT (Appl. Sys. Engin. 167X/KU) with 600W peak output
(200W peak power previously)
• New oscilloscope (Tek DPO73304D) enables detection of
8 4μs pulse frames at an effective rate of 3.3 Hz with 40μs
detection gate (previous 2-8GHz rate was ~0.5Hz!)
• High phase stability at 2-8 GHz allows for
long term averaging, even with stops to repressurize
the chamber for routine maintenance
Data sets obtained:
1) 0.2% sevoflurane / 0.2% benz in Ne, 10.8 million averages
2) 0.2% sevo / 0.2% benz-d1 in Ne, 6.5 million averages
Spectroscopic Results
Some facts about…
Sevoflurane + benzene-d1
sevoflurane monomer
Dynamic range: 20000:1
6.2 million averages
Average FWHM: 60 kHz
10,360 lines over 4:1 SNR
(1.7 MHz-1!)
sevoflurane-benzene-d1 (6x)
13C -sevoflurane-benzene-d
1
1
(33x)
The real success of this spectrum:
We now have enough sensitivity for
13C isotopologues in natural abundance!
BELGI Fit: The old…
(Ab-initio dipoles used for intensities)
J = 10 - 9, Ka ≥ 1
J=9-8
J = 5 - 4, Ka ≥ 2
BELGI Fit: ..and the new
(Ab-initio dipoles used for intensities)
J = 10 - 9, Ka ≥ 1
J=9-8
J = 5 - 4, Ka ≥ 2
One more comparison for good measure…
Old 2-8 GHz spectrum
(Scaled for noise)
New 2-8 GHz spectrum
(scaled for noise)
Improved BELGI-C1 Fit Results
Parameter
Lit. Label
Operator
Value
OA (MHz)
PAM value
B (MHz)
PAM value
A
"
B
"
Pa2
J a2
Pb2
J b2
509.684(15)
515.776(15)
365.017(12)
361.052(12)
C (MHz)
C
Pc2
338.2424(84)
PAM value
V6 (cm-1)
RHORHO
(unitless)
F (MHz)
DELTA (MHz)
DAB (MHz)
DACI (MHz)
DJ (kHz)
DJK (kHz)
DK (kHz)
C11 (MHz)
BK2 (MHz)
AK2 (MHz)
"
V6
Jc2
(1-cos(6α))/2
336.1133(84)
32.8688(27)
Ρ
PαPa
0.170557(54)
F
Fab
Dab
Daci
ΔJ
ΔJK
ΔK
V6bc
V6K
FK
Pa2
2
Pα {Pa,Pb}
{Pa,Pb}
{Pa,Pc}
-P4
-P2Pa2
-Pa4
(1-cos(6α))(Pb2-Pc2)
(1-cos(6α))(Pa2)
Pα2Pa2
[3452.484]
-0.0306(44)
25.40(24)
17.945(81)
0.0117(23)
0.648(85)
-7.85(81)
-13.45(40)
-13.2(48)
-0.234(28)
-----------------------
-------------
Fit Statistics
N (A/B states)
σAB (kHz)
N (E states)
σE (kHz)
N (total)
σ (kHz)
Direction Cosines
cos(λa)
cos(λb)
cos(λc)
------------
Experiment
0.961
-0.214
0.176
178
6.3
123
9.7
301
7.7
M06-2X
0.951
-0.230
0.208
• High-barrier six-fold internal rotation
• Consistent with C6 symmetry:
• Spin weights of 10:44:22:18 consistent
with observed relative intensities
• Fit includes new c-types and a-types with Ka ≤ 6
•
•
(A,E1) doublets fit as vt = 0 -> 0 transitions
(B,E2) doublets fit as vt = 1 -> 1 transitions C3 basis
J=0, K=0
Conformational Search
• MM force field conformer scan detected 95 conformers
within 43 kJ mol-1 (M06-2X/6-311++g(d,p) energies, ZPE corrected)
Conformers reported
by Suhm & coworkers1
(~15:1 in
Xe cryosol’n)
All structures & values derived using
MP2/6-311++g(d,p) + counterpoise corrections
1. Dom, J. J., et al. Phys. Chem. Chem. Phys. 2011, 13, 14142.
Isotopologue Assignments
• 6 2D positions X 10 13C positions = 60 total possible 13C/2D isotopologues
• However, 1-2D and 2-2D signals much weaker than other 4, so only 40 max are detectable
•
A single overnight Autofit run detected 22 13C/2D isotopologues
Final assignments (too many to list in detail here!):
• Refined six 2D benzene isotopologue fits (avg. ~250 lines/isotopologue)
• 33 2D/13C isotopologues (other 7 are buried)
• All six 13C benzene isotopologues
(Sevoflurane 13C isotopologues are complicated
by internal rotation, so they’ve been left alone. But
they’re there and quite strong!)
What can we do with all of this data?
• Full r0 structure fit, including fluorines!
(Thanks to Z. Kisiel for updating STRFIT to allow this many isotopic species in the input file!)
Structural Results
Parameters fixed:
• Sevoflurane hydrogen positions
• <(F-C-F) in perfluoro groups
• (r[C-F] was floated)
r0 σfit = 0.0494 amu Å2
MP2/6-311++g(d,p) + BSSE corrections
r0 experimental structure
Take home points:
• CF⋯H interactions are weakly contributing
• CF⋯H bonds in CSD database, on average, ~2.6 Å1
• CH⋯π interaction very dominant!
• HCCH⋯Ph: 3.5 Å (CCSD)2
• CHX3⋯Ph: 3.69 Å (X = H) ≤ r ≤ 3.39 Å (X = F)3
1) Schneider, H-J. Phys Chem Chem Phys. 2012, 3, 1381.
2) Fuiji, A., et al. J. Phys. Chem. A. 2004, 108, 2652.
3) Dey, R. C.; Seal, P.; Chakrabarti, S. J. Phys. Chem. A. 2009, 113, 10113.
Conclusions
Take home points
•
CP-FTMW sensitivity enables full r0 structural determination in one experiment
•
•
Can’t do much better, outside radioisotopic substitution of F (obviously not realistic in the slightest)
BELGI fit improved: includes new a-types and c-types with Ka ≤ 6, fit RMS = 7.7 kHz
•
Some issues with fitting the E1/E2 states in particular
or
M06-2X/6-311++g(d,p) optimized geometries
What else could be hiding in the spectrum?
• Higher order clusters
• Sevo-(benz)2?
• What wins -- T-shaped benzene dimer + sevoflurane, or benzene at two different
binding sites?
?
Acknowledgements
Pate Group
Brooks Pate
Cristobal Perez
Simon Lobsiger
Luca Evangelisti
Thanks for your time!
Brent Harris
Amanda Steber
Nathan Seifert
Newcastle University
Daniel Zaleski
Brightspec
Justin Neill
Universidad de Valladolid
Alberto Lesarri
Montserrat Vallejo-López
UPV-EHV
Emilio J. Cocincero
Fernando Castaño
LISA
Isabelle Kleiner
Thanks to the NSF for funding:
MRI-R2, Award CHE-0960074
Thanks to Zbiginew Kisiel for
his improvements to STRFIT!
Isotopologue Assignments - Autofit
Top 18 Autofit Results
A (MHz)
B (MHz)
C (MHz)
OMC (kHz)
Assignment
504.2
506.7
508.9
503.1
505.9
502.2
500.4
510.5
501.1
506.3
509.4
503.2
499.8
503.5
512.2
504.8
507.0
503.4
356.78
355.84
356.40
356.64
353.52
355.16
356.77
353.76
355.70
356.44
355.34
355.07
355.45
353.68
359.93
354.57
357.42
355.43
336.55
337.14
331.70
336.42
335.12
334.46
336.55
332.30
333.91
335.44
335.36
335.14
335.80
335.67
332.23
334.83
335.32
334.84
9.7
10.3
10.5
12.1
12.2
12.6
12.7
13.2
13.5
14.0
14.1
14.2
14.3
14.6
14.8
15.4
15.6
15.7
3-2D
4-2D
No fit
3-2D/13C(1)
4-2D/2-13C
3-2D/3-13C
3-2D
No fit
2
5- D/13C(4)
5-2D
6-2D/13C(2)
3-2D/4-13C
1-2D
2
4- D/5-13C
No fit
2
6- D/13C(1)
2-2D
5-2D/13C(1)
…but first 1000 fits all have OMC <24 kHz!
(See my talk on Thursday morning (RC12), for more info on Autofit)
• A total of 22 13C/2D isotopologues were assigned
from a single autofit run (there are very likely more)
• 9.89 million triples; frequency window ±20 MHz
• Fit using 10010-909, 927-826, 1046-945
• Fit checked with 10 transitions, 7 ≤ J ≤ 10, 0 ≤ Ka ≤ 4
(thus, each fit to the left is a 13-line fit)
Structural Results
Slight rotation of r0/rs coordinates
noticable w.r.t. calculated geometry:
Low-lying torsional mode contribution?