JOURNAL OF SOLID STATE CHEMISTRY 126, 184–188 (1996) ARTICLE NO. 0327 The Crystal Structure and Methyl Group Dynamics in the Room-Temperature and Low-Temperature Phases of Lithium Acetate Dihydrate G. J. Kearley,* B. Nicolai,* P. G. Radaelli,* and F. Fillaux² *Institut Laue-Langevin, B.P. 156X, 38042 Grenoble Cedex 09, France; and ²LASIR, CNRS, Rue Henri Dunant, 94320 Thiasis, France Received April 10, 1996; in revised form June 24, 1996; accepted July 2, 1996 the room-temperature crystal structure of LiAc ? 2H 2 O Neutron powder diffraction is used to determine the crystal persists down to liquid-helium temperatures. Even the structure of LiCD 3 COO ? 2D 2 O at 293 and 1.5 K. Lithium crystal structure of this phase has been questioned, and acetate dihydrate crystallizes in the Cmmm space group in the the most recent study was made with X-rays more than 25 room-temperature phase with a 5 6.82082(9) A ˚ , b 5 years ago (9). In the present paper we present the crystal 10.88842(12) A ˚ , c 5 6.59911(7) A ˚ , and Z 5 4. The CH 3 groups structures of the room-temperature and low-temperature are dynamically disordered, and there are short H bonds be- modifications of LiAc ? 2H 2 O as determined by neutron tween the water molecules and the acetate oxygen atoms. There powder diffraction, and relate this to the low-temperature is a phase transition to a low-temperature modification which motions of the hydrogen atoms. The use of neutron diffrac- has the space group Pman with a 5 6.70246(7) A ˚ , b 5 tion allows us to locate the deuterium atoms. 10.87932(11) A ˚ , c 5 6.56999(7) A ˚ , and Z 5 4. The CH 3 groups are ordered in this phase, but have large thermal ellipsoids even at 1.5 K. The H atoms of the water molecules also have EXPERIMENTAL large thermal ellipsoids even at 1.5 K, but the CH 3 group tunnel- ing spectrum shows no evidence for coupling with the motions Samples of LiCD 3 COO ? 2D 2 O were prepared via the of the water molecules.  1996 Academic Press, Inc. reaction of lithium carbonate with deuterated acetic acid in D 2 O. The sample was purified by recrystallization from D 2 O. The polycrystalline sample was placed in a standard INTRODUCTION thin-walled vanadium container, and temperature control was achieved using a standard liquid-helium cryostat. Rotational tunneling of small molecules or molecular Diffraction patterns were measured at a fixed neutron groups such as CH 3 at low temperatures has been inten- wavelength of 1.5941 A ˚ , using the D2B diffractometer (10) sively studied for the past 20 years. Until recently these at the Institut Laue Langevin, France. Rietveld refinement quantum motions were regarded as being entirely single of the patterns was achieved using the GSAS (11) suite of particle in nature (1), but more recently unequivocal evi- programs. Observed and calculated diffraction patterns are dence of extensively correlated rotor–rotor coupling has illustrated in Figs. 1a and 1b, and data collection and crystal accumulated (2–4). The first evidence for coupling of quan- structure parameters are given in Table 1. The tunneling tum rotors was obtained for lithium acetate dihydrate spectra of LiCH 3 COO ? 2D 2 O and LiCH 3 COO ? 2H 2 O (LiAc ? 2H 2 O, where Ac 5 CH 3 COO 2 or CD 3 COO 2 ), were recorded with a resolution of 12 eeV using the IN5 which showed extra splitting of the tunnel-split librational spectrometer at the Institut Laue Langevin, France. ground state (5–7). This was modeled in terms of isolated pairs of coupled tops. However, the coupled-pair model RESULTS AND DISCUSSION fails to account for the details of the librational spectrum of LiAc ? 2H 2 O (5), and recent experimental work with The first X-ray diffraction study of LiAc ? 2H 2 O pro- isotopically mixed LiCH 3 COO ? 2H 2 O/LiCH 2 DCOO ? posed the Cmm2 space group (12) but this was subject to 2H 2 O (8) revealed that the energy-level scheme of LiAc is some question regarding the interatomic distances (13). more complex than was originally thought. An alternative Infrared and Raman studies also militated against this interpretation based on long-range rotor–rotor coupling structure (14–16). In 1970 Galigne et al. (9) proposed the of pairs has been suggested. Cmmm space group with somewhat different cell parame- ters, but this is only plausible if the methyl groups are This work has been carried out on the assumption that 184 0022-4596/96 $18.00 Copyright  1996 by Academic Press, Inc. All rights of reproduction in any form reserved.