Controlled Under
Pressure
DOI: 10.1002/anie.200((will be filled in by the editorial staff))
Pressure-Driven Orbital Reorientations and Coordination Sphere
Reconstructions in [CuF2(H2O)2(pyz)]
Alessandro Prescimone, Chelsey Morien, David Allan, John A. Schlueter,
Stan W. Tozer, Jamie L. Manson, Simon Parsons,* Euan K. Brechin,* and
Stephen Hill*
((Dedication----optional))
The application of pressure in the study of molecule-based materials
has gained considerable recent interest, in part due to their high
compressibilities, but also because the relevant electronic/magnetic
(low energy) degrees of freedom in such materials are often very
sensitive to pressure.[1-11] For example, small changes in the
coordination environment around a magnetic transition metal ion
can produce quite dramatic variations in both the on-site spin-orbit
coupling as well as the exchange interactions between such ions
when assembled into 3D networks.[2-5] However, perhaps the most
compelling reason to use pressure as a tool for understanding
magneto-structural correlations is the possibility of focusing
investigations on a single molecule or material, as opposed to using
chemical means to influence the coordination environment around a
metal center, e.g., by studying families of seemingly related
complexes that vary only in the identity of the coordinating ligands.
[]
Dr A. Prescimone, Prof. S. Parsons, Prof. E. K. Brechin,
EaStCHEM School of Chemistry, The University of
Edinburgh, West Mains Road, Edinburgh, EH9 3JJ, UK
E-mail: ebrechin@staffmail.ed.ac.uk; s.parsons@ed.ac.uk
Ms C. Morien, Dr. S. Tozer, Prof. S. Hill, National High
Magnetic Field Laboratory and Department of Physics
Florida State University, Tallahassee, FL 32310, USA
E-mail: shill@magnet.fsu.edu
Dr D. Allan, Diamond Light Source, Harwell, Science
Innovation Campus, Chilton, Oxfordshire, OX11 ODE, UK
Dr J. A. Schlueter, Materials Science Division, Argonne
National Laboratory, Argonne, IL 60439, USA
Dr J. L. Manson, Department of Chemistry and
Biochemistry, Eastern Washington University, Cheney, WA
99004, USA
[]
We acknowledge the support of the EPSRC (UK) and NSF
(CHE-0924374). We thank Diamond Light Source for
access to beamline I19 (proposal number MT6146) that
contributed to the results presented herein. Work at
Argonne, a U.S. Department of Energy Office of Science
laboratory, is operated under Contract No. DE-AC0206CH11357. Work performed at the NHMFL is supported
by the NSF (DMR-0654118) and the State of Florida. JLM
gratefully acknowledges support of the NSF through grant
No. DMR-1005825. ST acknowledges support from DOE
NNSA DE-FG52-10NA29659.
Supporting information for this article is available on the
WWW under http://www.angewandte.org or from the
author.
The latter approach obviously suffers from the ‘non-innocence’ of
the ligand, particularly in the solid state. The desire to study
increasingly complex materials under pressure has spurred the
development of sophisticated spectroscopic tools that can be
integrated with high-pressure instrumentation.[12-15] The study of
magneto-structural correlations requires not only precise
crystallographic data, but also detailed spectroscopic information
concerning the unpaired electron(s) that give rise to the magnetic
properties. Here, we separately employ X-ray and high-frequency
EPR spectroscopies to obtain high-resolution structural and
magnetic data from oriented single-crystal samples subjected to
pressures of up to 3.5 GPa.
We focus on the magnetic coordination polymer [CuF2(H2O)2(pyz)]
(1, pyz = pyrazine), which previous powder diffraction studies have
shown to undergo successive pressure-induced structural transitions,
both of which are believed to involve dramatic reorientations of the
Jahn-Teller (JT) axes associated with the CuII ions.[9] Importantly,
magnetic susceptibility studies reveal a pronounced change in the
effective dimensionality of the extended Cu···Cu exchange
interactions (from 2D to 1D) at the first of these transitions.[9] In the
present investigation, EPR measurements provide direct information
on the disposition of the magnetic dx2-y2 orbital upon entering the
first of the two high-pressure phases. Furthermore, while the
previous powder X-ray studies established the onset of the phase
transitions and the reorientations of the JT axes, we describe the
results of single-crystal diffraction experiments which not only yield
structural data of substantially higher precision, but which also
reveal the way in which relief of strain built-up in compressed Hbonds drives the system through successive phase transitions as
pressure is increased. We also identify for the first time an
additional high pressure phase characterized by a dimerization of 1D
chains.
Under ambient conditions complex 1 crystallizes in the monoclinic
space group P21/c, with one copper center in the asymmetric unit
(Table S1). The metal ion resides in a distorted octahedral
environment formed by two O-atoms from the two water molecules,
two pyrazine N-atoms and two fluorides, as illustrated in Figure 1.
The elongated JT axis lies along the N-Cu-N vector with the two
symmetry equivalent bonds measuring 2.454(6) Å; the shorter Cu-O
and Cu-F distances are 1.984(4) Å and 1.908(4) Å, respectively
(Table 1). The CuII ions are linked into 1D chains along the
crystallographic a-axis via the pyrazine. They are also linked into a
2D network within the bc-plane by a series of short Cu-OH···F-Cu
hydrogen bonds [2.623(4) Å and 2.607(4) Å (Figure 1)]. The
magnetic dx2-y2 orbital associated with the axially elongated CuII ion
is expected to lie within the CuF2O2 (bc-) plane, thereby adding
significance to the hydrogen bonding interactions. Indeed, extensive
1
experimental studies of 1 at ambient pressures reveal a low
temperature transition to an antiferromagnetic state (TN = 2.54 K)
with strong 2D character, indicating appreciable Cu···Cu exchange
within the bc-plane, with considerably weaker interactions along the
chains.[16] Under the application of external hydrostatic pressure the
first, and perhaps most obvious, effect is the compression of the unit
cell (Table S1). At 1.2 GPa the Cu-X (X = N, F, O) coordination
bond distances remain essentially unchanged, but a contraction in
both hydrogen bonds to 2.515(13) Å is observed; these distances are
amongst the shortest M-OH2···F-M H-bonds to have been observed,
pinpointing the build-up of strain in the compressed intermolecular
contacts.
transmitting media used for the various experiments, and possibly
also due to the different nature of the samples (fine powder versus
the more rigid/fragile single-crystals). We note also that the
pressures were calibrated at very different temperatures using
different manometers.
Table 1. Bond distances (Å) in 1 as function of pressure.
P/GPa
0
0.5
0.9
1.2
1.8
2.2
2.5
2.85
3.3
3.3
3.3
Cu1-F
1.908(4)
1.904(3)
1.907(2)
2.030(14)
1.898(2)
1.904(2)
1.906(2)
1.907(2)
1.886(15)
Cu2-F'
2.283(12)
Cu2-O
2.22(2)
Cu1-O
1.984(4)
1.975(4)
1.968(3)
1.987(6)
2.316(3)
2.331(3)
2.324(3)
2.320(3)
2.36(2)
Cu2-F
1.91(2)
Cu2-N
2.069(13)
Cu1-N
2.454(6)
2.430(4)
2.417(3)
2.441(7)
2.039(3)
2.036(3)
2.031(2)
2.027(2)
2.046(13)
Cu2-F˝
1.91(2)
Cu2-N'
2.016(13)
Use of single-crystal diffraction data enables location of the H-atom
positions from Fourier difference maps.[17] The mean Cu-O-H angle
in phase-II decreases from 108.3(9)° at 1.8 GPa to only 99.7(9)°
upon further increasing the pressure to 2.85 GPa (Table S3). A
search of the Cambridge Database shows that this almost
perpendicular orientation is quite unusual.[18] Nevertheless, it
preserves the near linearity in the OH···F contacts as the structure is
compressed.
Figure 1. Visualization of the 2D H-bonded network with the [OH···F]2 unit in the crystallographic bc plane in phase I (top) and in
phase IV (bottom). Color scheme: Cu = orange, O = red, F = yellow,
N = blue, C = gray and H = salmon. The dashed red lines indicate the
OH···F H-bonds.
Upon further increasing the pressure to 1.8 GPa, the expected phase
modification occurs: the monoclinic symmetry is retained, but the
Cu-N bonds are rather dramatically compressed (by 0.4 Å) to a
value of 2.039(3) Å, while the Cu-O bonds elongate (by 0.3 Å) to a
value of 2.316(3) Å. This is a clear indication that the JT axis has
reoriented from the N-Cu-N vector to the O-Cu-O vector. This
structural reorganization relieves the tension in the OH···F hydrogen
bonds, which increase to 2.702(3) Å and 2.626(3) Å. Previous highpressure studies performed on powder samples have reported this
phase change to be reversible.[9,10] However, the single-crystal
sample becomes polycrystalline upon release of the pressure.
Moreover, the first phase modification is not observed until
substantially higher pressures are achieved in the present
investigation (~1.8 GPa as opposed to 1.0 GPa[9] – see also the EPR
data below). These differences may be due to the different pressure
Figure 2. Disposition of the Jahn-Teller axes (higlighted by the dark
blue rods) in 1 in the three phases. Color scheme: Cu = orange, O =
red, F = yellow, N = blue, C = gray.
Further pressure increments reveal a second structural transition
between 2.85 and 3.3 GPa, and this transformation is even more
disruptive than the first. The two-fold symmetry is lost and the
structure becomes twinned in the triclinic space group P-1 (Table
S1). The cause of the symmetry loss lies in the ejection of one water
molecule per copper unit from two thirds of the chains, forcing them
to dimerize through the F-atoms in order to fill the vacant
coordination sites (Figures 1 and 2). Interestingly, the remaining one
third of the chains are unchanged, i.e., they have bond lengths and
angles equivalent to those seen prior to the 2nd phase transition.
Consequently, the structure now contains two different chain types
2
within the asymmetric unit. The structure is also twinned via a
pseudo two-fold rotation about the b-axis. The expelled water
molecules remain in the crystal lattice, located between the
monomeric and dimeric chains, held in place by the OH···F
hydrogen bonding network (Figure S1). The mean O···F distances
mediated by H-bonds is now 2.67(2) Å, spanning the range from
2.52(2) – 2.74(2) Å, while the average Cu-O-H angle is now
117.5(5)˚.
Figure 3. Experimental EPR spectra recorded at 10 K as a function of
the field orientation within the ab-plane of a single-crystal. The three
panels correspond to three different pressures and microwave
frequencies (indicated in the figure). Spectra were recorded every 10 o,
with 0o corresponding to the field along a; the field orientations
corresponding to the top and bottom traces are indicated in each
panel.
The high-pressure phase obtained here for a single-crystal is
different from phase-III described by Halder.[9] The latter phase is
monoclinic; the Cu-atoms remain 6-coordinate, but the JT axis shifts
to the F-Cu-F direction. The powder pattern calculated from the
triclinic structure described here, which we shall call phase-IV, is
quite clearly different from the experimental powder data reported
by Halder (see Fig. S2 in reference 9). It is possible that small
deviations in hydrostaticity that arise from use of polycrystalline or
single-crystal samples, or from the different hydrostatic media used
in the two studies, are responsible for the differences in ours and
Halder’s results. The transition to phase-IV reported here causes
major changes in the coordination environment of the CuII ions in
the dimerized chains (Table 1): in particular, the JT axes are now
oriented along the distorted O-Cu-F bonds (Figure 2), which also
undergo a significant reorientation relative to the O-Cu-O JT axes
associated with the monomeric chains. Most importantly, adjacent
CuII ions within the dimerized chains are now bridged directly by Fatoms, likely engendering appreciable exchange coupling between
these spins. The hydrogen bonding network is also affected: the CuOH···F-Cu pathways between the monomeric and dimerized chains
remain relatively unaffected; however, the linkages between dimeric
chains become considerably more complex (Fig. S1).
The reorientation of the CuII JT axes at the first pressure-induced
transition has a profound influence on the magnetic properties of 1.
The consequent reorientation of the ground state dx2-y2 orbital into the
CuF2N2 plane cuts off the 2D magnetic interactions within the bcplane, while promoting appreciable 1D correlations along the Cupyz-Cu chains. Previous magnetic susceptibility measurements have
provided indirect evidence for this JT reorientation on the basis of a
change in the dimensionality of the extended magnetic (Cu···Cu)
interactions.[9] On the other hand, single-crystal EPR studies can
provide direct information on the symmetries of the wave functions
associated with the unpaired magnetic electrons through their spinorbit mixing with excited ligand field states. This mixing manifests
itself in a characteristic anisotropy of the Lande g-tensor. EPR
measurements were performed initially at a low pressure of
0.67 GPa. Figure 3(a) displays experimental spectra recorded as a
function of field orientation at a frequency of 69.3 GHz and a
temperature of 10 K. The crystal was oriented with its c-axis
approximately perpendicular to the plane of rotation, i.e., the field
was rotated in the ab-plane. A single, sharp peak is observed that
displays strong angle dependence with Lande value extrema of
g// = 2.42 and g = 2.08 (see Figure 4); the larger value corresponds
to the direction parallel to the JT (a-) axis. These values compare
extremely well with published results performed under ambient
conditions using a commercial X-band (9 GHz) spectrometer.[16]
The next set of measurements were performed at 1.82 GPa, with the
65.7 GHz data displayed in Figure 3(b). This time, two signals are
observed: one displays virtually identical angle-dependence to the
low-pressure signal (see Figure 4); however, the second signal
displays a comparatively weaker angle-dependence, with low gvalues in the 2.05 to 2.10 range, suggesting that it corresponds to
field rotation within the plane of the dx2-y2 orbital ( orientation).
Similar measurements were performed at a frequency of 96 GHz
(see Figure 4). These data provide the first hint that some fraction of
the sample has transitioned into the second phase in which the JT
axis switches to the O-Cu-O bonds. Nevertheless, the transition
appears to be incomplete, which might signify a slight nonuniformity of the pressure within the cell. Also evident is a
broadening of the EPR signal, particularly in the regions of
maximum variation with angle. This indicates strains in the sample,
i.e., a degradation of the sample brought about either by thermal
cycling or non-hydrostatic pressures within the cell.
Figure 4. EPR peak positions seen in Figure 3 (together with data
obtained at other frequencies – not shown), plotted as their
corresponding Lande g-factors versus the orientation of the applied
field within the ab-plane of the crystal.
Upon further increasing the pressure to 2.1 GPa (Figure 3(c)), the
strongly angle-dependent signal (red/black data in Figure 4)
vanishes completely, and the remaining EPR intensity is observed in
a single, broad low-g signal (blue/green data in Figure 4). This highpressure EPR signal displays a weak angle-dependence (Figure 4)
that agrees remarkably well with bc-plane rotations performed under
3
ambient conditions.[16] However, in the present case, the field was
rotated within the ab-plane, thereby providing the most direct
evidence that the magnetic dx2-y2 orbital has switched from the
CuF2O2 plane to the CuF2N2 plane. Unfortunately, efforts to access
the 2nd phase transition were hindered by multiple failures of the
plastic pressure cells used for the EPR studies.[15] Future efforts that
use diamonds with smaller culets will allow us to reach pressures in
excess of 3 GPa more reliably.
We return to the 2nd phase transition that occurs at ~3 GPa. As noted
above, the dimerized chains involve direct Cu-F-N bridges, thus
leading to the formation of Cu2F2 dimers that are coupled along the
a-axis via the pyrazine. Magnetically, F-bridges have been shown to
behave very much like OH-bridges;[19] the latter have been more
widely characterized. In other words, the Cu2F2 dimers should have
similar properties to their Cu2(OH)2 counterparts, for which
extensive information can be found in the literature. On this basis,
the 104o Cu-F-Cu bridging angle would seem to suggest
antiferromagnetic coupling; ferromagnetism is typically limited to a
narrow angle range between 96o and 98o. Consequently, the
dimerized chains can likely be viewed as antiferromagnetic spin
ladders with different exchange coupling strength on the rungs and
along the rails, the latter determined by the pyrazine bridges. Such
chains would very likely possess a spin gap separating a singlet
(diamagnetic) ground state from a band of dispersive triplet excited
states;[20] the magnitude of this gap would be dependent on the ratio
of the two exchange constants (Jrung:Jladder), as well as their
magnitudes. Consequently, the dimerized chains would most likely
not contribute to the low-temperature magnetic susceptibility or the
EPR response. Because the monomer chains are essentially
indistinguishable from those in the intermediate pressure phase, it is
conceivable that the only magnetic signature of the 2nd phase
transformation would be a reduction of the susceptibility (or loss of
EPR intensity) by a factor of three due to the dimerization of two
thirds of the –Cu–pyz– chains.
It should be emphasized that the above discussion of the magnetism
within the high-pressure phase is somewhat speculative. Efforts are
currently under way aimed at measuring the pressure dependence of
the magnetic susceptibility of a crystal in phase-IV in order to
bolster our understanding. We also hope that the present
investigation may motivate ab-initio calculations aimed at
estimating the magnetic exchange interactions within the dimerized
spin ladders, thereby providing a means of confirming whether the
interaction within the Cu2F2 dimers is indeed antiferromagnetic.
Such information could then be used to determine the magnitude of
the spin gap associated with the dimerized chains. Finally, we note
that high magnetic fields could eventually prove very useful for
closing the singlet-triplet gap,[20] thereby ‘switching-on’ the
magnetism associated with the dimerized chains.
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
D. S. Clarke, S. P. Strong, P. M. Chaikin, E. I. Chashechkina, Science
1998, 279, 2071.
W. Fujita, K. Awaga, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A.
2000, 341, 389.
M. Mito, H. Deguchi, T. Tajiri, M. Yamashita, H. Miyasaka, Phys.
Rev. B. 2005, 72, 144421.
A. Sieber, R. Bircher, O. Waldmann, G. Carver, G. Chauboussant, H.
Muttka, H.-U. Güdel, Angew. Chem. Int. Ed. 2005, 44, 4239.
(a) A. Prescimone, C. J. Milios, J. Sanchez-Benitez, K. V. Kamenev,
C. Loose, J. Kortus, S. A. Moggach, M. Murrie, J. E. Warren, A. R.
Lennie, S. Parsons, E. K. Brechin, Dalton Trans. 2009, 4858.
K. W. Galloway, S. A. Moggach, P. Parois, A. R. Lennie, J. E.
Warren, E. K. Brechin, R. D. Peacock, R. Valiente, J. González, F.
Rodriguez, S. Parsons, M. Murrie, CrystEngComm., 2010, 12, 2516.
S. A. Moggach, S. Parsons, Spectrosc. Prop. Inorg. Organomet.
Compd. 2009, 40, 324.
P. Parois, S. A. Moggach, J. Sanchez-Benitez, K. V. Kamenev, A. R.
Lennie, J. E. Warren, E. K. Brechin, S. Parsons, M. Murrie, Chem.
Commun. 2010, 46, 1881.
G. J. Halder, K. W. Chapman, J. A. Schlueter, J. L. Manson, Angew.
Chem. Int. Ed. 2010, 49, 419, and references therein.
J. L. Mustfeldt, Z. Liu, S. Li, J. Kang, C. Lee, P. Jena, J. L. Manson, J.
A. Schlueter, G. L. Carr, M.-H. Whangbo, Inorg. Chem. 2011, 50,
6347.
A. L. Leitch, K. Lekin, S. M. Winter, L. A. Downie, H. Tsuruda, J. S.
Tse, M. Mito, S. Desgreniers, P. A. Dube, S. Zhang, Q. Liu, C. Jin, Y.
Ohishi, R. T. Oakley, J. Am. Chem. Soc. 2011, 133, 6051.
L. Merrill, W. A. Bassett, Rev. Sci. Instrum. 1974, 45, 290.
A. Dawson, D. R. Allan, S. Parsons, M. Ruf, J. Appl. Cryst., 2004, 37,
410.
J. Diederichs, A. K. Gangopadhyay, J. S. Schilling, Phys. Rev. B 1996,
54, R9662.
D. E. Graf, R. L. Stillwell, K. M. Purcell, S. W. Tozer, High Pressure
Research 2011, 31, 533.
J. L. Manson, M. M. Conner, J. A. Schlueter, A. C. McConnel, H. I.
Southerland, I. Malfant, T. Lancaster, S. J. Blundell, M. L. Brooks, F.
L. Pratt, J. Singleton, R. D. McDonald, C. Lee, M.-H. Whangbo,
Chem. Mater. 2008, 20, 7408.
P. W. Betteridge, J. R. Carruthers, R. I. Cooper, K. Prout, D. J. Watkin,
J. Appl. Cryst. 2003, 36, 1487.
F. H. Allen, Acta. Cryst. 2002, B58, 380.
S. C. Lee, R. H. Holm, Inorg. Chem. 1993, 32, 4745.
T. Giamarchi, A. M. Tsvelik, Phys. Rev. B 1999, 59, 11398.
Received: ((will be filled in by the editorial staff))
Published online on ((will be filled in by the editorial staff))
Keywords: High Pressure · EPR · X-ray crystallography · Jahn-Teller
reorientation · Cu(II)
4
Entry for the Table of Contents (Please choose one layout)
Layout 1:
Controlled Under Pressure
Alessandro Prescimone, Chelsey
Morien, David Allan, John Schlueter,
Stan Tozer, Jamie L. Manson, Simon
Parsons,* Euan K. Brechin,* and
Stephen Hill*
__________ Page – Page
Pressure-Driven Orbital Reorientations
and Coordination Sphere
Reconstructions in [CuF2(H2O)2(pyz)]
X-ray crystallography and highfrequency EPR measurements on
[CuF2(H2O)2(pyz)] reveal a series of
pronounced structural transitions that
involve successive ~90o reorientations
of the Jahn-Teller axes associated with
the CuII ions. The second transition
forces a dimerization involving two thirds
of the CuII sites due to ejection of one of
the water molecules from the
coordination sphere.
5
Supporting Information
Pressure-Driven
Orbital
Reorientations
and
Coordination
Sphere
Reconstructions
in
[CuF2(H2O)2(pyz)]
Alessandro Prescimone, Chelsey Morien, David R. Allan, John Schlueter, Stan Tozer, Jamie L. Manson,
Simon Parsons, Euan K. Brechin, and Stephen Hill
Table S1. Crystallographic details for the high pressure refinements of 1.
ambient
0.50 GPa
0.90 GPa
1.20 GPa
1.80 GPa
Chemical formula
C4H8CuF2 C4H8CuF2 C4H8CuF2 C4H8CuF2 C4H8CuF2N
N2O2
N2O2
N2O2
N2O2
2O2
Mr
217.66
217.66
217.66
217.66
217.66
Crystal
system, Monoclini Monoclini Monoclini Monoclini Monoclinic,
space group
c, P21/c
c, P21/c
c, P21/c
c, P21/c
P21/c
a, b, c (Å)
7.7004 (2), 7.6300 (6), 7.5857
7.6471
6.8372 (3),
7.5648 (3), 7.5027
(14), 7.456 (10), 7.551 7.5981 (9),
(2), 6.751 (5), 6.8191 7.0280 (8)
6.9033 (2) (14),
6.8168
(2)
(10)
(12)
, ,  (°)
90,
111.062
(2), 90
V (Å3)
90,
112.657
(8), 90
90,
113.769
(15), 90
90,
112.833
(8), 90
90, 115.095
(6), 90
375.27 (2) 360.12
(10)
349.44
(17)
362.9 (3)
330.64 (6)
Z
2
2
2
2
2
F(000)
218
218
218
218
218
Radiation type
Mo K
Mo K
Mo K
Synchrotro Mo K
n,  =
0.48590 Å
 (mm-1)
2.90
3.02
3.11
3.00
Crystal size (mm)
0.09 × 0.07 0.07 × 0.05 0.07 × 0.05 0.06 × 0.06 0.07 × 0.05
× 0.04
× 0.04
× 0.04
× 0.05
× 0.04
Diffractometer
Bruker
Kappa
Apex2
Bruker
Kappa
Apex2
Bruker
Kappa
Apex2
3.29
Area
Bruker
diffractom Kappa
eter
Apex2
6
Scan method
 &
scans
  scans
Tmin, Tmax
0.49, 0.89
0.68, 0.89
 scans
 scans
 scans
0.73, 0.88
0.54, 0.89
0.81, 0.88
measured,
2543, 990, 4080, 338, 1984, 335, 1119, 231, 1888, 318,
independent
and 715
315
294
187
273
observed
[I
>
2.0(I)] reflections
Rint
0.051
0.022
0.024
0.046
0.024
max (°)
30.2
26.0
26.0
15.7
26.3
R, wR(F2), S
0.056,
0.174,
1.00
0.026,
0.072,
1.03
0.022,
0.062,
0.98
0.046,
0.121,
1.00
0.019,
0.050, 1.01
reflections
964
336
328
221
305
parameters
68
68
68
68
68
No. of restraints
22
22
22
16
22
H-atom treatment
refined
Refined
Refined
refined
Refined
Weighting scheme Chebychev Chebychev Chebychev Chebychev Chebychev
polynomial polynomial polynomial polynomial polynomial
max, min (e Å-3) 0.95, -1.27 0.27, -0.22 0.20, -0.21 0.32, -0.46 0.19, -0.21
2.20 GPa
2.50 GPa
2.85 GPa
3.30 GPa
Chemical formula
C4H8CuF2N2O2
C4H8CuF2N2O2
C4H8CuF2N2O2
C12H24Cu3F6N6O6
Mr
217.66
217.66
217.66
652.99
Crystal system, space Monoclinic,
group
P21/c
Monoclinic,
P21/c
Monoclinic,
P21/c
Triclinic, P¯1
a, b, c (Å)
6.8385(7),
7.5892(12),
7.0257(11)
6.8228(6),
7.5581(10),
7.0043 (9)
6.8098(6),
7.5078(10),
6.9873 (9)
6.7987(8), 10.452
(6), 7.272 (2)
, ,  (°)
90,
90
V (Å3)
329.91 (8)
325.83 (7)
321.12 (7)
463.3 (3)
Z
2
2
2
1
F(000)
218
218
218
327
Radiation type
Synchrotron,  = Synchrotron,  = Synchrotron,  = Synchrotron,  =
0.48590 Å
0.48590 Å
0.48590 Å
0.48590 Å
 (mm-1)
3.30
115.204(7), 90,
90
3.34
115.566(6), 90,
90
3.39
115.987(6), 86.24(3), 115.844
(16), 88.71(2)
3.52
7
Crystal size (mm)
0.06 × 0.06 × 0.08 × 0.07 × 0.08 × 0.07 × 0.60 × 0.06 × 0.05
0.05
0.03
0.03
Diffractometer
Area
diffractometer
Area
diffractometer
Area
diffractometer
Area
diffractometer
Scan method
 scans
 scans
 scans
 scans
Tmin, Tmax
0.68, 0.88
0.81, 0.88
0.80, 0.87
0.70, 0.87
measured, independent 1543, 327, 293
and observed [I >
2.0(I)] reflections
1994, 458, 397
1958, 462, 400
2269, 608, 532
Rint
0.034
0.026
0.025
0.041
max (°)
15.6
17.7
17.9
16.1
R, wR(F2), S
0.022,
0.98
reflections
316
433
454
532
parameters
68
68
68
77
No. of restraints
22
22
22
13
H-atom treatment
geometrical
Geometrical
Geometrical
geometrical
Weighting scheme
Chebychev
polynomial
Chebychev
polynomial
Chebychev
polynomial
Chebychev
polynomial
max, min (e Å-3)
0.19, -0.20
0.25, -0.21
0.22, -0.22
1.10, -1.11
0.061, 0.023,
0.94
0.063, 0.023,
1.14
0.063, 0.109,
1.15
0.072,
8
Table S2. H-bond distances and Cu-O-H average angle in 1 as functions of pressure.
P/GPa
O-H…F
O-H…F
Cu-O-Haverage
0
2.623(4)
2.607(4)
113.0(9)˚
0.5
2.620(4)
2.599(4)
118.3(9)˚
09
2.610(3)
2.587(3)
108.7(9)˚
1.2
2.515(13)
2.515(13)
-
1.8
2.702(3)
2.626(3)
108.3(9)˚
2.2
2.694(2)
2.618(3)
103.2(9)˚
2.5
2.693(2)
2.608(2)
100.2(9)˚
2.85
2.691(2)
2.600(2)
99.7(9)˚
3.3
2.72(2)
2.74(2)
117.5(5)˚
2.69(3)
2.52(2)
Figure S1. Representation of the structure of 1 in phase IV, containing the F-bridged dimerized chains,
the monomeric chains and interstitial water molecules. Color scheme: Cu = orange, O = red, F = yellow,
N = blue, C = grey and H = white.
9
Experimental section
Table S3. Details of resolution, X-ray source and different crystals used as function of pressure.
P/kbar
0
5
9
12
18
22
25
28.5
33
lamda/Ang
0.71
0.71
0.71
0.48
0.71
0.48
0.48
0.48
0.48
theta_max/deg
30.223
26.043
25.947
15.663
26.306
15.644
17.654
17.854
16.062
d/Ang
0.71
0.81
0.81
0.89
0.80
0.89
0.79
0.78
0.87
crystal#
1
2
2
3
2
3
4
4
3
X-ray crystallography
High-pressure single-crystal X-ray diffraction experiments were carried out using a Merrill-Bassett
diamond anvil cell (half-opening angle 40o),[1] equipped with Boehler-Almax diamonds with 600 μm
culets and a tungsten gasket.[2] Petroleum ether was used as the hydrostatic medium and a small ruby chip
was loaded into the cell as the pressure calibrant, with the ruby fluorescence used to measure the
pressure.[3] High-pressure data collections were taken at 0.5, 0.90, 1.20, 1.8, 2.2, 2.5, 2.85 and 3.3 GPa all
at room temperature. Structures at 1.20, 2.2, 2.5, 2.85 and 3.3 GPa were collected using synchrotron
radiation of wavelength  = 0.4859 Å on a Rigaku diffractometer at Station I19 at the Diamond Light
Source, Harwell Science and Innovation Campus, whilst data at ambient, 0.5, 0.8 and 1.8 GPa were
collected on a Mo-Ka Bruker Smart APEX II diffractometer.[4] Integrations were carried out using the
program SAINT[5] and absorption corrections with the programs SADABS[6] and SHADE.[7] Refinements
of the compressed forms of 1 were carried out starting from the coordinates obtained from the model of
the previous pressure. The program CRYSTALS[8] was used to refine the structures of 1 against F2 using
all the reflections except for the 3.3 GPa dataset that had to be refined on F using reflections with I >2.
Due to the loss of the two-fold symmetry the 3.3 GPa dataset was twinned with the following twin law: (1 0 0, -0.194 1 -0.267, 0 0 -1). Unit cells and refinement parameters are reported in Table S1. Structures
and packing diagrams were drawn using DIAMOND.[9]
10
High-frequency EPR
The high-pressure EPR experiments were performed using a miniature plastic turnbuckle diamond anvil
cell (DAC) and nonmetallic gasket described in ref [10]. The dimensions of the DAC (Ø11.5  12.5 mm
long) permitted its insertion into a custom cylindrical microwave cavity made of OFHC copper without
completely suppressing its resonant characteristics, thus allowing cavity perturbation measurements
involving a small single-crystal within the DAC; the cavity perturbation technique is described
elsewhere.[11] Measurements were performed within a split-pair 7 T superconducting magnet, enabling in
situ orientation of the crystal, thereby enabling discrimination of the anisotropic EPR signal due to the
sample from background signals associated with the DAC.[12] A fiber optic for pressure calibration via
ruby fluorescence[3] was inserted into the cavity along the axis of the DAC without adversely affecting the
resonator Q-factor. Several pressure media were employed for this portion of the study in an attempt to
find the most hydrostatic fluid. Single crystals were freshly cleaved in such a manner to allow for visual
alignment with respect to the applied field to within 5°. A manuscript describing the high pressure, highfrequency EPR experimental technique is in preparation.[13]
References
[1] L. Merrill, W. A. Bassett, Rev. Sci. Instrum. 1974, 45, 290.
[2] S. A. Moggach, D. R. Allan, S. Parsons, J. E. Warren, J. Appl. Cryst. 2008, 41, 249.
[3] G. J. Piermarini, S. Block, J. D. Barnett, R. A. Forman, J. Appl. Phys. 1975, 46, 2774.
[4] Bruker-Nonius APEX-ll, Version V1, Madison, Wisconsin, USA, 2004.
[5] Bruker-Nonius SAINT version 7, Bruker-AXS: Madison, Wisconsin, USA, 2006.
[6] G. M. Sheldrick, SADABS Version 2004-1, Bruker-AXS: Madison, WIsconsin, USA, 2004.
[7] S. Parsons, S.SHADE, The University of Edinburgh: Edinburgh, United Kingdom, 2004.
[8] D. J. Watkin, K. Prout, J. R. Carruthers, P. W. Betteridge, R. I. Cooper, CRYSTALS Issue 12, Chemical Crystallography
Laboratory, University of Oxford: Oxford, UK, 2003.
[9] K. Brandenburg, H. Putz, DIAMOND, Crystal Impact: Bonn, Germany, 2005.
[10] D. E. Graf, R. L. Stillwell, K. M. Purcell, S. W. Tozer, High Pressure Research 2011, 31, 533.
[11] M. Mola, S. Hill, P. Goy, M. Gross, Rev. Sci. Inst. 2000, 71, 186.
[12] S. Takahashi, S. Hill, Rev. Sci. Inst. 2005, 76, 023114.
[13] C. C. Beedle, C. Morien, S. Tozer, S. Hill, in preparation.
11