C O M M U N I C A T I O N S
of the crystal from the rotation axis, and the direction in which
this force is applied within the crystal depends on the orientation
of the crystal relative to the rotation axis. It is reasonable to propose
that a different dehydration mechanism, leading to the formation
of the SA-I phase, might be induced if the magnitude of the
centrifugal force is sufficiently high and if this force is applied in
2
an appropriate direction within the SA‚3H O crystal. The fact that
this effect depends on both the position of the crystal relative to
the rotation axis and on the orientation of the crystal relative to
this axis can explain why some crystals can undergo the “new”
dehydration mechanism to yield SA-I, whereas other crystals still
yield SA-beta. In addition, rapid sample rotation can produce
acoustic vibration, and we cannot, at this stage, exclude the
possibility that acoustic vibration might also influence the mech-
Figure 2. Solid-state 13C NMR spectra recorded as a function of time during
dehydration of SA‚3H2O (150 scans (5 h) per spectrum). The tuning and
match of the NMR probe were re-established at regular intervals during
the experiment. Note that the anomalously low signal intensity in the first
spectrum is observed reproducibly and is attributed to instrumental factors.
2
anism for dehydration of SA‚3H O. Further experiments are in
2
Is SA-I produced directly by dehydration of SA‚3H O or by a
sample rotation induced transformation from another phase (e.g.,
SA-beta) that is produced initially upon dehydration?
progress to yield a more detailed fundamental understanding of the
change in polymorphic product distribution induced by rapid sample
rotation.
First, it is well-known that sample rotation in solid-state MAS
While the SA-I phase obtained by dehydration of SA‚3H
conditions of rapid sample rotation is a known polymorph (although
not previously known to be accessible by dehydration of SA‚3H O),
2
O under
9
NMR experiments can increase the sample temperature. Our
9
temperature calibration experiments using lead nitrate indicate that,
2
for a nominal sample temperature of 293 K (recorded using the
thermocouple in the solid-state NMR probe), the actual sample
temperature for rotation at 3 kHz was 294 K and at 7 kHz was 317
K. To assess independently whether such increases of temperature
could influence the polymorphic product distribution, experiments
it is conceivable that, for other systems, carrying out solid-state
desolvation processes under rapid sample rotation may provide a
viable route for the formation of new, hitherto unknown, polymor-
phic forms.
of type (i) (dehydration of SA‚3H
2
O with no sample rotation) were
Acknowledgment. We are grateful to Research Councils UK
for financial support through a Basic Technology grant, and to Dr.
Rob Jenkins for technical assistance.
carried out at elevated temperature (313 and 323 K). These
experiments produced only SA-beta, with no detectable amounts
of SA-I (Figure 1e).
Second, it is conceivable that the SA-I phase obtained under
conditions of sample rotation is not formed directly by dehydration
References
(1) (a) Dunitz, J. D. Pure Appl. Chem. 1991, 63, 177. (b) Caira, M. R. Top.
Curr. Chem. 1998, 198, 163. (c) Bernstein, J.; Davey, R. J.; Henck, J.-O.
Angew. Chem., Int. Ed. 1999, 38, 3441. (d) Bernstein, J. Polymorphism
in Molecular Crystals; Oxford University Press: Oxford, 2002.
(2) (a) Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1964, 1996. (b)
Schmidt, G. M. J. Pure Appl. Chem. 1971, 27, 647. (c) Thomas, J. M.
Philos. Trans. R. Soc. 1974, 277, 251.
of SA‚3H
2
O, but is instead produced from an initially formed phase
(SA-beta) by a transformation that is induced by sample rotation.
13
To assess this issue, high-resolution solid-state C NMR has been
used as an in situ probe of the dehydration process. Solid-state 13
NMR spectra recorded as a function of time during dehydration
are shown in Figure 2, and isotropic 13C chemical shifts (measured
independently of the in situ study) for SA‚3H
polymorphs of SA are listed in ref 10. At the start of the in situ
experiment, the signal at 181.8 ppm is characteristic of the SA‚3H
C
(3) Wei, K. T.; Ward, D. L. Acta Crystallogr. 1977, B33, 522.
(
4) (a) Hsu, L. Y.; Nordman, C. E. Acta Crystallogr. 1983, C39, 690. (b)
Helmholdt, R. B.; Sonneveld, E. J.; Schenk, H. Z. Kristallogr. 1998, 213,
596.
2
O and the anhydrous
(
5) Sharma, S. K.; Jotshi, C. K.; Kumar, S. Thermochim. Acta 1991, 184, 9.
6) The PENCIL rotor (Varian) comprises a zirconia sleeve, a removable drive
tip, a sample spacer, and a Teflon end-cap. The end-cap designed for
high-temperature operation has a vent hole to relieve pressure in the rotor,
which provides an escape route for loss of water from the rotor during
the in situ dehydration experiments.
(
2
O
starting material. The intensity of this signal decreases as a function
of time, while a new signal at 181.2 ppm arises from the formation
SA-beta, and a weaker signal at 182.1 ppm indicates the concomi-
(7) Powder XRD patterns were recorded on a Bruker D8 instrument operating
in transmission mode (Cu KR1 radiation).
tant formation of SA-I.11 The solid-state C NMR results indicate
13
12
(8) All solid-state C NMR spectra were recorded under conditions of 13C
13
1
1
r H cross-polarization, high-power H decoupling, and MAS, at 75.48
MHz on a Chemagnetics CMX-Infinity 300 spectrometer (CP contact time,
1 ms; recycle delay, 120 s).
that both SA-beta and SA-I are produced from the start of the
dehydration process, and there is no evidence (e.g., from integrated
peak areas for SA-I and SA-beta as a function of time) that SA-I
is produced by a transformation from SA-beta. In further support
of this conclusion, pure SA-beta (obtained by dehydration of SA‚
(
9) (a) Van Gorkom, L. C. M.; Hook, J. M.; Logan, M. B.; Hanna, J. V.;
Wasylishen, R. E. Magn. Reson. Chem. 1995, 33, 791. (b) Bielecki, A.;
Burum D. P. J. Magn. Reson. 1995, A116, 215. (c) Neue, G.; Dybowski,
C. Solid State Nucl. Magn. Reson. 1997, 7, 333.
1
3
-);
); SA-
(
10) Isotropic C chemical shifts (ppm): SA‚3H
2
O, 26.0 (CH
3
), 181.8 (CO
2
3H
2
O in experiments of type (i)) was subjected to sample rotation
-
-).
-
2
SA-I, 26.6 (CH
beta, 27.2 (CH
3
), 182.1 (CO ); SA-II, 26.8 (CH
2
3
), 183.6 (CO
for ca. 12 h (nominal temperature 293 K) in separate experiments
at 3 and 7 kHz; in each case, there was no change in the solid-
state 13C NMR spectrum.
Sample rotation produces a centrifugal force in the radial
direction with respect to the rotation axis, and the magnitude of
this force is higher as distance from the rotation axis increases.
For the polycrystalline sample within the solid-state NMR rotor,
the crystals are oriented randomly. Clearly, the magnitude of the
centrifugal force exerted on a given crystal depends on the distance
3
), 181.2 (CO
2
(
11) Similar changes in the signals for the methyl carbon environment (not
shown in Figure 2) are in agreement with these conclusions.
13
23
(12) High-resolution solid-state Na NMR spectra, again recorded as a function
of time during the dehydration process, also lead to the same conclusions
13
as the in situ solid-state C NMR study.
(13) In another experiment, the product (SA-I plus SA-beta) following complete
2
dehydration of SA‚3H O was subjected to sample rotation at 3 kHz for
ca. 12 h (in separate experiments at nominal temperatures of 293 and
1
3
393 K). There was no evidence from the solid-state C NMR spectra of
any change in the relative amounts of SA-I and SA-beta.
JA052668P
J. AM. CHEM. SOC.
9
VOL. 127, NO. 31, 2005 10833