research letters\(\def\hfill{\hskip 5em}\def\hfil{\hskip 3em}\def\eqno#1{\hfil {#1}}\)

IUCrJ
Volume 4| Part 5| September 2017| Pages 569-574
ISSN: 2052-2525

ζ-Glycine: insight into the mechanism of a polymorphic phase transition

crossmark logo

aISIS Facility, Rutherford–Appleton Laboratory, Harwell Science and Innovation Campus, Didcot, Oxfordshire OX11 0QX, UK, bSchool of Chemistry and Centre for Science at Extreme Conditions, The University of Edinburgh, King's Buildings, W. Mains Road, Edinburgh EH9 3FJ, UK, cScuola Internazionale Superiore di Studi Avanzati, Via Bonomea 265, Trieste 34136, Italy, dDepartment of Chemistry and Biochemistry, University of California San Diego, La Jolla, CA 92093, USA, and eOak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831, USA
*Correspondence e-mail: degironc@sissa.it, kucukben@gmail.com, s.parsons@ed.ac.uk

Edited by L. R. MacGillivray, University of Iowa, USA (Received 8 May 2017; accepted 25 July 2017; online 1 September 2017)

Glycine is the simplest and most polymorphic amino acid, with five phases having been structurally characterized at atmospheric or high pressure. A sixth form, the elusive ζ phase, was discovered over a decade ago as a short-lived intermediate which formed as the high-pressure ɛ phase transformed to the γ form on decompression. However, its structure has remained unsolved. We now report the structure of the ζ phase, which was trapped at 100 K enabling neutron powder diffraction data to be obtained. The structure was solved using the results of a crystal structure prediction procedure based on fully ab initio energy calculations combined with a genetic algorithm for searching phase space. We show that the fate of ζ-glycine depends on its thermal history: although at room temperature it transforms back to the γ phase, warming the sample from 100 K to room temperature yielded β-glycine, the least stable of the known ambient-pressure polymorphs.

1. Introduction

Glycine, the simplest amino acid, exhibits greater phase diversity than any of the other naturally occurring amino acids. All phases consist of the zwitterionic form +H3N—CH2—CO2 and three polymorphs are known under ambient conditions. The α and β forms are both monoclinic (space groups P21/n and P21, respectively), with crystal structures composed of hydrogen-bonded layers. The γ form is trigonal (P31) and displays a three-dimensional hydrogen-bonded network featuring chains disposed about 31 screw axes. Of the three polymorphs, γ-glycine is the most thermodynamically stable at room temperature (Perlovich et al., 2001[Perlovich, G., Hansen, L. & Bauer-Brandl, A. (2001). J. Therm. Anal. Calorim. 66, 699-715.]).

The effect of pressure on the polymorphs of glycine has been extensively investigated. Although α-glycine persists to 23 GPa (Murli et al., 2003[Murli, C., Sharma, S. M., Karmakar, S. & Sikka, S. K. (2003). Physica B, 339, 23-30.]), β-glycine undergoes a phase transition at only 0.8 GPa to the δ phase (P21/n; Dawson et al., 2005[Dawson, A., Allan, D. R., Belmonte, S. A., Clark, S. J., David, W. I., McGregor, P. A., Parsons, S., Pulham, C. R. & Sawyer, L. (2005). Cryst. Growth Des. 5, 1415-1427.]; Goryainov et al., 2005[Goryainov, S. V., Kolesnik, E. N. & Boldyreva, E. V. (2005). Phys. B Condens. Matter, 357, 340-347.]). This transition is displacive and fully reversible as a result of the topological similarity of the β and δ phases, consisting of a concerted inversion of the O—C—C—N torsion angles in half the molecules which otherwise retain their positions and orientations. Application of pressure to γ-glycine leads to ɛ-glycine (Pn; Boldyreva et al., 2005[Boldyreva, E. V., Ivashevskaya, S. N., Sowa, H., Ahsbahs, H. & Weber, H.-P. (2005). Z. Kristallogr. Cryst. Mater. 220, 50-57.]; Dawson et al., 2005[Dawson, A., Allan, D. R., Belmonte, S. A., Clark, S. J., David, W. I., McGregor, P. A., Parsons, S., Pulham, C. R. & Sawyer, L. (2005). Cryst. Growth Des. 5, 1415-1427.]; Moggach et al., 2015[Moggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315-5328.]). The structure changes substantially, the threefold helices being replaced by a layered structure. The transition is sluggish and reconstructive, but is complete at between 4 and 5 GPa. Note that some authors refer to the δ and ɛ phases as β′ and δ, respectively.

The γ- to ɛ-glycine transition shows considerable hysteresis and the ɛ phase persists on decompression to 0.28 GPa (Moggach et al., 2015[Moggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315-5328.]). Complete release of applied load yields yet another phase, ζ-glycine, which exists for as little as 30 min at room temperature before transforming back to the γ phase (Bordallo et al., 2008[Bordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748-8759.]; Goryainov et al., 2006[Goryainov, S. V., Boldyreva, E. V. & Kolesnik, E. N. (2006). Chem. Phys. Lett. 419, 496-500.]; Moggach et al., 2015[Moggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315-5328.]). The ζ phase, which has been characterized by X-ray powder diffraction and Raman spectroscopy, is only observed upon decompression of the ɛ form; application of pressure to the γ phase yields the ɛ phase directly.

Although the ζ phase was first identified over a decade ago, its structure is unknown. On the basis of Raman data it has been proposed to have a layered structure similar to all the other phases except for γ-glycine (Bordallo et al., 2008[Bordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748-8759.]). In this paper we describe the solution of the crystal structure of ζ-glycine using high-quality neutron powder diffraction measurements in combination with crystal structure prediction based on fully first-principles total energy calculations and an improved genetic search algorithm for searching the phase space.

2. Experimental

2.1. Neutron powder diffraction

A sample of γ-glycine-d5 (CDN Isotopes, Canada) was obtained by non-photochemical laser-induced nucleation of a supersaturated solution in D2O (Zaccaro et al., 2001[Zaccaro, J., Matic, J., Myerson, A. S. & Garetz, B. A. (2001). Cryst. Growth Des. 1, 5-8.]). Colourless needle-like crystals were allowed to grow for 2 h at room temperature and then lightly crushed to form a polycrystalline sample. Pressure was applied to the sample using a Paris–Edinburgh press modified for use at low temperature, utilizing zirconia-toughened alumina anvils. The sample was enclosed in a null-scattering TiZr encapsulated gasket. A mixture of deuterated methanol and ethanol (4:1 v/v) was used as a pressure transmitting medium with a lead pellet as pressure marker. Diffraction data were obtained on the PEARL instrument at ISIS (Bull et al., 2016[Bull, C. L., Funnell, N. P., Tucker, M. G., Hull, S., Francis, D. J. & Marshall, W. G. (2016). High. Press. Res. 36, 493-511.]) and the resulting diffaction corrected for anvil attenuation. The sample was compressed to 5.56 GPa to yield a phase-pure sample of ɛ-glycine. The temperature was then rapidly lowered to 200 K and the load reduced to 0 tonnes. Then the temperature was further reduced to 100 K. Neutron powder diffraction data were collected over the following 5 h. Following data collection, the sample was warmed to room temperature while monitoring the diffraction pattern. Further experimental details are available in the supporting information.

2.2. Crystal structure prediction

The evolutionary algorithm as implemented in the USPEX package (Lyakhov et al., 2013[Lyakhov, A. O., Oganov, A. R., Stokes, H. T. & Zhu, Q. (2013). Comput. Phys. Commun. 184, 1172-1182.]) was used to search for the low-energy structures of glycine with Z = 2, 3 or 4 molecules in the unit cell. In the first generation 30 structures were created randomly. The energetically least favourable 20% of the population was discarded. A fingerprint analysis of the remaining structures was performed (Oganov & Valle, 2009[Oganov, A. R. & Valle, M. (2009). J. Chem. Phys. 130, 104504-104512.]). The structures whose fingerprint was within an adimensional cosine distance of 0.01 from any lower energy structure were also discarded. The remaining structures were then eligible as parents and allowed to `procreate'. Thirty new structures of the next generation were created from the parents through the following operations: heredity (crossover of two structures) (40%), soft mutation (translation and rotation based on an estimate of soft vibrational modes) (20%), rotation of the molecule (20%) and random structure generation (20%). In addition, the three best parents were cloned directly to the next generation. In all simulations, the maximum number of generations was 20.

2.3. Ab initio calculations

For every structure generated by USPEX, the geometry and cell relaxation were calculated using a van der Waals density functional (vdW-DF) as implemented in the QUANTUM ESPRESSO package (Giannozzi et al., 2009[Giannozzi, P. et al. (2009). J. Phys. Condens. Matter, 21, 395502.]). A kinetic energy cutoff of 80 Ryd and a charge density cutoff of 560 Ryd were used. The Brillouin zone sampling resolution was gradually increased over three steps during relaxation, from 2π × 0.12 Å−1 to 2π × 0.10 Å−1 and finally 2π × 0.08 Å−1. Energies and geometries from the last step with the densest k-point grid were used throughout the study. Projector augmented wave (PAW) pseudopotentials were taken from the PSLibrary v0.1 (Küçükbenli et al., 2014[Küçükbenli, E., Monni, M., Adetunji, B. I., Ge, X., Adebayo, G. A., Marzari, N., de Gironcoli, S. & Dal Corso, A. (2014). arXiv, 1404.3015.]). By using this setup all structures were fully relaxed within a convergence of less than 0.1 mRyd for the absolute total energy per molecule, 0.5 mRyd a.u.−1 for the forces on the atoms and less than 0.005 GPa for the pressure obtained from the stress tensor.

3. Results

3.1. Crystal structure prediction

Conventional methods for indexing and solving the structure of ζ-glycine failed to yield models which reproduced the experimental data. However, independently of the diffraction experiments, theoretical work had been carried out to assess the predictive power of a newly developed crystal structure prediction (CSP) workflow using glycine as the test case. Although a number of CSP surveys of glycine have been described (Zhu et al., 2012[Zhu, Q., Oganov, A. R., Glass, C. W. & Stokes, H. T. (2012). Acta Cryst. B68, 215-226.]; Lund et al., 2015[Lund, A. M., Pagola, G. I., Orendt, A. M., Ferraro, M. B. & Facelli, J. C. (2015). Chem. Phys. Lett. 626, 20-24.]; Chisholm et al., 2005[Chisholm, J. A., Motherwell, S., Tulip, P. R., Parsons, S. & Clark, S. J. (2005). Cryst. Growth Des. 5, 1437-1442.]), none has so far identified candidates which fit the experimental data for ζ-glycine, though a combined CSP–powder diffraction study has very recently yielded the crystal structure of glycine dihydrate (Xu et al., 2017[Xu, W., Zhu, Q. & Hu, C. T. (2017). Angew. Chem. 129, 2062-2066.]).

A critical component of CSP is the method of evaluating the lattice energies of candidate phases. This determines both energy ranking and the direction of the phase-space search. The highest quality energies are obtained using first-principles calculations based on periodic density functional theory (DFT) (Hohenberg & Kohn, 1964[Hohenberg, P. & Kohn, W. (1964). Phys. Rev. 136, B864-B871.]; Reilly et al., 2016[Reilly, A. M. et al. (2016). Acta Cryst. B72, 439-459.]), but despite some spectacular successes (Woodley & Catlow, 2008[Woodley, S. M. & Catlow, R. (2008). Nat. Mater. 7, 937-946.]) molecular crystals are challenging because of the difficulties in accounting for dispersion interactions (Reilly et al., 2016[Reilly, A. M. et al. (2016). Acta Cryst. B72, 439-459.]).

In this study energies were calculated (QUANTUM ESPRESSO; Giannozzi et al., 2009[Giannozzi, P. et al. (2009). J. Phys. Condens. Matter, 21, 395502.]) with a recently developed functional (vdW-DF; Berland et al., 2015[Berland, K., Cooper, V. R., Lee, K., Schröder, E., Thonhauser, T., Hyldgaard, P. & Lundqvist, B. I. (2015). Rep. Prog. Phys. 78, 066501.]) which takes dispersion into account via a non-local functional of the overall charge density. The method yields reliable energy differences between phases of molecular materials (Sabatini et al., 2012[Sabatini, R., Küçükbenli, E., Kolb, B., Thonhauser, T. & de Gironcoli, S. (2012). J. Phys. Condens. Matter, 24, 424209.]), including the ambient-pressure forms of glycine, without suffering from the transferability problems of empirical methods. The combination of this method for energy evaluation with the genetic algorithm for searching phase space (USPEX; Lyakhov et al., 2013[Lyakhov, A. O., Oganov, A. R., Stokes, H. T. & Zhu, Q. (2013). Comput. Phys. Commun. 184, 1172-1182.]) assuming two, three or four molecules per unit cell yielded a very rich series of hypothetical structures for glycine (Fig. 1[link]). Fuller details are available in the supporting information.

[Figure 1]
Figure 1
Enthalpy versus volume distribution for all the structures obtained via ab initio crystal structure search within the lowest 2 kJ mol−1 range of the most stable γ phase. Crowding around each polymorph indicates multiple encounters with the same phase during the phase-space exploration.

In addition to all the experimentally observed phases of glycine, the CSP analysis predicted several new phases within 2 kJ mol−1 of the most stable (γ) phase. In particular, the survey had identified a potential new polymorph in P1 with one molecule per unit cell (Z = 1; a = 5.0168, b = 4.7491 and c = 4.0593 Å; α = 95.8383, β = 105.6522 and γ = 64.8726°) that was very close in energy to the known β polymorph. This was proposed as ζ-glycine based on the agreement between previous mixed-phase X-ray diffraction experiments (Boldyreva et al., 2005[Boldyreva, E. V., Ivashevskaya, S. N., Sowa, H., Ahsbahs, H. & Weber, H.-P. (2005). Z. Kristallogr. Cryst. Mater. 220, 50-57.]) and theoretical assessment (see supporting information). The structural parameters of the new phase, which had been predicted completely ab initio, were found to form a suitable starting model for refinement of the structure against the neutron diffraction data.

3.2. Neutron powder diffraction

Neutron powder diffraction data suitable for structure solution were collected for ζ-glycine by trapping the phase at 100 K, extending its lifetime for long enough to collect a high-quality diffraction pattern.

The powder diffraction data were indexed using the unit-cell dimensions obtained in the structure prediction multiplied by a factor of 0.98. This accounts for the tendency of the method used for CSP to overestimate slightly the unit-cell dimensions of other glycine polymorphs (Table S1). Pawley fitting yielded the following cell dimensions for ζ-glycine: a = 4.9307 (3), b = 4.54798 (4) and c = 3.9191 (3) Å, α = 95.550 (5), β = 105.250 (5) and γ = 64.938 (6)°, and V = 76.797 (11) Å3. These values can be transformed to a body-centred triclinic setting containing two molecules per unit cell, which clarifies the structural relationship of ζ-glycine with the other phases. These transformed cell dimensions are compared with those of the other Z = 2 phases in Table 1[link]. The predicted crystal structure of ζ-glycine was refined using Rietveld methods (TOPAS-Academic; Coelho, 2015[Coelho, A. (2015). TOPAS-Academic. Version 5. Coelho Software, Brisbane, Australia.]). The final fit is shown in Fig. 2[link](a).

Table 1
Cell dimensions of β-, ɛ- and ζ-glycine

In all cases Z = 2. All values are from this work.

  ɛ (100 K) ζ (100 K) β (298 K) ζ (298 K)
Symmetry Pn I1 P21 I1
a (Å) 5.0230 (4) 5.1000 (4) 5.0907 (2) 5.1029 (16)
b (Å) 5.9846 (4) 6.2850 (3) 6.25954 (16) 6.3450 (12)
c (Å) 5.4946 (5) 5.4295 (3) 5.38710 (19) 5.4331 (18)
α (°) 90 85.815 (5) 90 85.91 (3)
β (°) 114.654 (8) 114.456 (5) 113.261 (4) 114.26 (3)
γ (°) 90 104.136 (5) 90 103.55 (3)
Volume (Å3) 150.12 (2) 153.545 (17) 157.710 (10) 155.85 (9)
[Figure 2]
Figure 2
(a) Rietveld fit of the neutron powder diffraction pattern of ζ-glycine at 100 K (blue = observed, red = calculated). In addition to the peaks from ζ-glycine, the pattern also shows the presence of residual ɛ- and a trace of γ-glycine. Other peaks arise from the sample environment, namely the lead pressure marker and the Al2O3 and ZrO2 components of the anvils of the pressure cell. (b) Rietveld fit of the neutron powder diffraction pattern of β-glycine (contaminated with ζ- and a trace of γ-glycine) at 290 K. A 1 Å d spacing approximates to 4837 µs in time-of-flight.

When the sample of ζ-glycine prepared in this study was warmed from 100 K, the peaks due to the residual ɛ phase were first seen to disappear at 250 K. At 290 K the peaks due to ζ-glycine began to diminish, to yield not the expected γ phase but rather the β phase (Fig. 2[link]b). Persistence of the ζ polymorph after the disappearance of the ɛ polymorph shows that the transition to β-glycine is from ζ-glycine rather than ɛ-glycine; the initial disappearance of ɛ-glycine is the continuation of the original ɛ to ζ transition halted by the temperature decrease.

4. Discussion

4.1. Structural relationships and phase stability

The data in Table 1[link] show that the unit-cell dimensions of the β, ɛ and ζ polymorphs are very closely related to each other. The structures, along with that of γ-glycine, are shown in Figs. S13–S16. The β, ɛ and ζ phases are layered structures in which the molecules in the unit cell reside near positions equivalent to [[{1 \over 4}, {1 \over 4}, {1 \over 4}]] and [[{3 \over 4}, {3 \over 4}, {3 \over 4}]] (Table S4). A crystallographic fractional coordinate of [1 \over 4] can be transformed to [3 \over 4] either by adding [1 \over 2] or by inverting it and adding 1. The β, ɛ and ζ polymorphs can simply be viewed as the result of various combinations of these operations along the x, y and z directions of the unit cell. In the ζ phase the relationship is x + [1 \over 2], y + [1 \over 2], z + [1 \over 2] (I-centring); in the ɛ phase it is x + [1 \over 2], −y + 1, z + [1 \over 2] (an n-glide plane), while in the β phase it is −x + 1, y + [1 \over 2], −z + 1 (a 21 screw axis).

In the case of the β and ζ phases the symmetry operations preserve conformation, and this is why the cell dimensions of these two phases are so similar. In ɛ-glycine the two zwitterionic molecules are related by an n-glide and have opposite conformations, the N1—C2—C1—O2 torsion angles being ±16.0 (5)°. In ζ-glycine the torsion angle increases to 35.1 (4)° and all the molecules have the same conformation. The ɛ- to ζ-glycine transition can thus be viewed as a switch of conformation of half the molecules in the structure, and in this regard it is similar to the β- to δ-glycine transition at 0.8 GPa. Although γ-glycine shares the polarity of the ɛ and ζ phases along c, and since its space group is chiral all three molecules in the unit cell share the same conformation [τ(N1—C2—C1—O2) = 15.6 (1)°; Kvick et al., 1980[Kvick, Å., Canning, W. M., Koetzle, T. F. & Williams, G. J. B. (1980). Acta Cryst. B36, 115-120.]], its structure consists of a three-dimensional hydrogen-bonded network rather than layers.

Symmetry-adapted perturbation theory (PSI4 software; Turney et al., 2012[Turney, J. M. et al. (2012). WIREs Comput. Mol. Sci. 2, 556-565.]) and PIXEL calculations (Gavezzotti, 2005[Gavezzotti, A. (2005). Z. Kristallogr. 220, 499-510.]) (see supporting information) both indicate that the strongest intermolecular interaction in ζ-glycine, with an energy in excess of −100 kJ mol−1, is a `head-to-tail' hydrogen bond formed between ammonium and carboxyl­ate groups. Repetition of this interaction builds a chain which runs along the c direction of the crystal structure. This same chain is present in all six known polymorphs, and the phases differ in the way these chains interact.

In ζ-glycine the chains are arranged in the ac plane, related by lattice repeats along a, forming a layer (Fig. 3[link]a). This layer motif confirms Boldyreva's conclusion regarding the layered nature of the ζ phase made on the basis of vibrational spectroscopy (Bordallo et al., 2008[Bordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748-8759.]). The layers are very similar, both in terms of geometry and in the molecule–molecule energies, to those seen for the ɛ phase. The intermolecular contact energy across the N1—H5⋯O1 hydrogen bonds (−26 kJ mol−1) is only about a quarter of that of the head-to-tail linkages, a result of the repulsive influence of neighbouring carboxyl­ate groups.

[Figure 3]
Figure 3
Intermolecular interactions in ζ-glycine. (a) Layers formed in the ac plane, viewed along b. (b) Stacking of the layers, viewed along c.

In the ζ phase all layers are connected through equivalent N1—H4⋯O contacts (Fig. 3[link]b). The layer separation is b/2 = 3.14 Å. Although the intermolecular contact energy for the pairs of molecules linked by N1—H4⋯O1 contacts (−59.4 kJ mol−1) is equivalent to a strong hydrogen bond, the geometry of the contact deviates markedly from linearity [∠N1—H4⋯O1 = 121.1 (10)°] with a relatively long H⋯O distance [2.274 (5) Å], and the interaction is more plausibly classified as a simple electrostatic interaction involving ammonium and carboxylate groups.

The relative energies (in kJ mol−1) of the six phases of glycine, calculated using DFT and including the approximation for van der Waals interactions described above, are: γ 0, α 0.084, β 1.055, ζ 1.070, ɛ 1.832 and δ 2.019. The energies refer to geometry-optimized structures at 0 K and 0 Pa. The ordering of the α, β and γ phases agrees with previous work (Marom et al., 2013[Marom, N., DiStasio, R. A., Atalla, V., Levchenko, S., Reilly, A. M., Chelikowsky, J. R., Leiserowitz, L. & Tkatchenko, A. (2013). Angew. Chem. Int. Ed. 52, 6629-6632.]; Perlovich et al., 2001[Perlovich, G., Hansen, L. & Bauer-Brandl, A. (2001). J. Therm. Anal. Calorim. 66, 699-715.]; Sabatini et al., 2012[Sabatini, R., Küçükbenli, E., Kolb, B., Thonhauser, T. & de Gironcoli, S. (2012). J. Phys. Condens. Matter, 24, 424209.]), while the total range, which spans 2 kJ mol−1, agrees with Hunter's recent results on typical polymorph energy differences (Hunter & Prohens, 2017[Hunter, C. A. & Prohens, R. (2017). CrystEngComm, 19, 23-26.]). The ζ phase is marginally less stable than the previously known ambient-pressure phases, but more stable than either of the high-pressure phases.

4.2. The fate of ζ-glycine on warming

Previous work (Boldyreva et al., 2005[Boldyreva, E. V., Ivashevskaya, S. N., Sowa, H., Ahsbahs, H. & Weber, H.-P. (2005). Z. Kristallogr. Cryst. Mater. 220, 50-57.]; Bordallo et al., 2008[Bordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748-8759.]; Goryainov et al., 2006[Goryainov, S. V., Boldyreva, E. V. & Kolesnik, E. N. (2006). Chem. Phys. Lett. 419, 496-500.]; Moggach et al., 2015[Moggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315-5328.]) using X-ray and neutron powder diffraction and vibrational spectroscopy has shown that ζ-glycine transforms spontaneously to the γ phase at room temperature. By contrast, in the present investigation, warming the sample from 100 K to room temperature yielded β-glycine rather than the expected γ phase. The thermal history of the samples used in this and previous work were different: in previous studies all manipulations were carried out at room temperature, whereas here the sample had been cooled. Nevertheless the difference in behaviour is surprising, and this is the first time, to our knowledge, that a transition to the metastable β phase of glycine has been observed. Similar sensitivity of phase formation to thermal history has been observed, for example in paracetamol (Qi et al., 2008[Qi, S., Avalle, P., Saklatvala, R. & Craig, D. Q. M. (2008). Eur. J. Pharm. Biopharm. 69, 364-371.]; Rossi et al., 2003[Rossi, A., Savioli, A., Bini, M., Capsoni, D., Massarotti, V., Bettini, R., Gazzaniga, A., Sangalli, M. E. & Giordano, F. (2003). Thermochim. Acta, 406, 55-67.]), while in glycine itself the temperature of the thermal γ-to-α phase transition can be increased by ca 10 K by annealing the sample (Perlovich et al., 2001[Perlovich, G., Hansen, L. & Bauer-Brandl, A. (2001). J. Therm. Anal. Calorim. 66, 699-715.]).

Over the course of the ɛζγ transition (see the movie in the supporting information), the layers of ɛ-glycine first slide over one another with small molecular rotations to give ζ-glycine. Larger reorientations that disrupt the layer structure are needed as the system transforms to the γ phase.

As discussed above, the β and ζ phases are closely related. Both consist of chains composed of head-to-tail N1—H3⋯O2 hydrogen bonds which are connected into a layer via N1—H5⋯O1 hydrogen bonds. The difference between the phases is that the c axis is a polar direction in the ζ phase but not in the β phase. Therefore, in the ζβ transition (see the movie in the supporting information) the layer structure is retained but a rotation of layers with respect to one another is needed. The lack of a strong directional preference in the electrostatic contacts between the layers (see above) may explain why such phase transitions involving rearrangements of layer stacking can occur.

The structure of ζ-glycine is significant because the phase is akin to a supramolecular reactive intermediate, providing insight into the mechanism of solid-state phase transitions. The results presented here, along with those of previous studies (Boldyreva et al., 2005[Boldyreva, E. V., Ivashevskaya, S. N., Sowa, H., Ahsbahs, H. & Weber, H.-P. (2005). Z. Kristallogr. Cryst. Mater. 220, 50-57.]; Bordallo et al., 2008[Bordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748-8759.]; Goryainov et al., 2006[Goryainov, S. V., Boldyreva, E. V. & Kolesnik, E. N. (2006). Chem. Phys. Lett. 419, 496-500.]; Moggach et al., 2015[Moggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315-5328.]), indicate that two transformation pathways are available to ζ-glycine, one leading to the β phase and the other to the γ phase. One possibility is that the activation barrier to the β phase is lower, and it is possible that seeds of this phase begin to form at low temperature. In this interpretation, at ambient temperature there is enough thermal energy to form the thermodynamically more stable γ phase. Alternatively, the topological similarity of the ɛ and γ phases, in which all the head-to-tail chains have the same polarity, may enable a residual amount of the high-energy ɛ phase to seed the formation of γ-glycine. Further theoretical modelling and experimental investigations would provide valuable insight into the physical source of the difference between the low- and ambient-temperature behaviour.

5. Related literature

For additional literature relating to the supporting information, see Brandenburg (2004[Brandenburg, K. (2004). DIAMOND. Crystal Impact GbR, Bonn, Germany.]), Destro et al. (2000[Destro, R., Roversi, P., Barzaghi, M. & Marsh, R. E. (2000). J. Phys. Chem. A, 104, 1047-1054.]), Fortes et al. (2007[Fortes, A. D., Wood, I. G., Alfredsson, M., Vocadlo, L., Knight, K. S., Marshall, W. G., Tucker, M. G. & Fernandez-Alonso, F. (2007). High Pressure Res. 27, 201-212.], 2012[Fortes, A. D., Wood, I. G., Alfredsson, M., Vocadlo, L., Knight, K. S., Marshall, W. G., Tucker, M. G. & Fernandez-Alonso, F. (2012). High Pressure Res. 32, 337-337.]), Gavezzotti (2011[Gavezzotti, A. (2011). New J. Chem. 35, 1360-1368.]), Hohenstein & Sherrill (2010a[Hohenstein, E. G. & Sherrill, C. D. (2010a). J. Chem. Phys. 133, 014101.],b[Hohenstein, E. G. & Sherrill, C. D. (2010b). J. Chem. Phys. 133, 104107.], 2012[Hohenstein, E. G. & Sherrill, C. D. (2012). WIREs Comput. Mol. Sci. 2, 304-326.]), Jeziorski et al. (1994[Jeziorski, B., Moszynski, R. & Szalewicz, K. (1994). Chem. Rev. 94, 1887-1930.]), Frisch et al. (2009[Frisch, M. J. et al. (2009). GAUSSIAN09. Gaussian Inc., Wallingford, Connecticut, USA.]), Macrae et al. (2008[Macrae, C. F., Bruno, I. J., Chisholm, J. A., Edgington, P. R., McCabe, P., Pidcock, E., Rodriguez-Monge, L., Taylor, R., van de Streek, J. & Wood, P. A. (2008). J. Appl. Cryst. 41, 466-470.]), Sheldrick (2001[Sheldrick, G. M. (2001). SHELXTL-XP. Version 6.01. University of Göttingen, Germany.]), Spek (2003[Spek, A. L. (2003). J. Appl. Cryst. 36, 7-13.]) and Stone (2013[Stone, A. J. (2013). The Theory of Intermolecular Forces. Oxford University Press.]).

Supporting information


Computing details top

zeta-glycine-d5 (zeta_glycine_100K) top
Crystal data top
C2D5NO2γ = 104.136 (5)°
Mr = 80.10V = 153.54 (2) Å3
Triclinic, I1Z = 2
Hall symbol: I1Cell parameters included in refinement
a = 5.1000 (4) ÅDx = 1.732 Mg m3
b = 6.2850 (3) ÅNeutron radiation
c = 5.4295 (3) ÅT = 100 K
α = 85.815 (5)°Particle morphology: Block
β = 114.456 (5)°Colourless
Data collection top
PEARL, ISIS
diffractometer
Scan method: Time-of-flight
Radiation source: Pulsed neutron sourceAbsorption correction: empirical (using intensity measurements)
Calculated using custom ISIS software.
Specimen mounting: TiZr Capsule
Data collection mode: transmission
Refinement top
Least-squares matrix: selected elements only0 restraints
Rp = 2.86622 constraints
Rwp = 3.088Weighting scheme based on measured s.u.'s
Excluded region(s): excluded short TOF region above(Δ/σ)max = <0.001
Profile function: Voigt function convoluted with a back-to-back double exponential.Background function: Chebychev polynomial
65 parametersPreferred orientation correction: Modelling of preferred orientation was found to be unnecessary.
Special details top

Geometry. Distances and angles were all constrained in the refinement so the sus are 0 [this won't be the case if the values are relcalculated from the using the coordinate data block above - getting the right values depends on having the fill covariance matrix available].

Refinement. The data in this cif come from a Rietveld refinement against neutron powder data. Details of refinement are given in the paper, and the Topas instructuion file is under _iucr_refine_instructions_details.

Checkcif output:

091_ALERT_1_A No Wavelength found in CIF - 0.71073 Ang Assumed Please Check

The data were collected by the time of flight method

155_ALERT_4_C The Triclinic Unitcell is NOT Reduced ·········. Please Do ! 128_ALERT_4_G Alternate Setting for Input Space Group I1 P1 Note

The setting is chosen to show the relationship to the epsilon phase. The matrix relating the prediction and refinement cell settings is [-1 1 0] [ 0 -1 1] [ 1 0 1]

161_ALERT_4_C Missing or Zero su (esd) on x-coordinate for ··· N1 162_ALERT_4_C Missing or Zero su (esd) on y-coordinate for ··· N1 163_ALERT_4_C Missing or Zero su (esd) on z-coordinate for ··· N1

This atom was used to fix the origin.

340_ALERT_3_C Low Bond Precision on C-C Bonds ··············· 0.0080 Ang. 201_ALERT_2_G Isotropic non-H Atoms in Main Residue(s) ······. 5 Why ? 210_ALERT_3_G No Anisotropic ADP's Found in CIF ············.. Please Check

The structure was refined against medium resolution neutron powder data.

741_ALERT_1_C Bond Calc 1.25 (2), Rep 1.25000 ······ Missing su O1 -C1 1.555 1.555 # 1 741_ALERT_1_C Bond Calc 1.250 (11), Rep 1.25000 ······ Missing su O2 -C1 1.555 1.555 # 2 ···plus numerous similar alerts

The s.u.s on the diatnces and angles are all exactly 0 as they were constrained during refinement.

017_ALERT_1_G Check Consistency of Scattering Type D for H1 017_ALERT_1_G Check Consistency of Scattering Type D for H2 017_ALERT_1_G Check Consistency of Scattering Type D for H3 017_ALERT_1_G Check Consistency of Scattering Type D for H4 017_ALERT_1_G Check Consistency of Scattering Type D for H5

The sample was deuterated.

154_ALERT_1_G The su's on the Cell Angles are Equal ·········. 0.00500 Degree 720_ALERT_4_G Number of Unusual/Non-Standard Labels ·········. 5 Note 802_ALERT_4_G CIF Input Record(s) with more than 80 Characters ! Info

No action taken

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) top
xyzBiso*/Beq
N10.711440.681320.779260.58 (3)
C10.540 (2)0.777 (2)1.1166 (9)0.58 (3)
C20.518 (2)0.7995 (16)0.8265 (8)0.58 (3)
O10.308 (3)0.766 (4)1.1491 (13)0.58 (3)
O20.788 (2)0.7714 (17)1.2992 (6)0.58 (3)
H10.2893 (15)0.732 (3)0.6865 (9)0.69 (4)
H20.584 (4)0.9728 (15)0.7886 (14)0.69 (4)
H30.7178 (14)0.7244 (11)0.5909 (7)0.87 (5)
H40.625 (2)0.5116 (2)0.7800 (17)0.87 (5)
H50.9255 (5)0.7260 (15)0.9341 (9)0.87 (5)
Geometric parameters (Å, º) top
O1—C11.2500N1—H41.0500
O2—C11.2500N1—H51.0500
N1—C21.4800C2—H11.0900
C1—C21.5300C2—H21.0900
N1—H31.0500
O1—C1—O2126.00H3—N1—H5110.00
O1—C1—C2117.00N1—C2—C1112.00
O2—C1—C2117.00N1—C2—H1108.00
C2—N1—H5109.00N1—C2—H2108.00
C2—N1—H3109.00C1—C2—H1109.00
C2—N1—H4109.00C1—C2—H2109.00
H4—N1—H5110.00H1—C2—H2110.00
H3—N1—H4110.00
O1—C1—C2—N1144.9 (4)O2—C1—C2—N135.1 (4)
epsilon-glycine-d5 (epsilon_glycine_100K) top
Crystal data top
C2D5NO2V = 150.12 (2) Å3
Mr = 80.10Z = 2
Monoclinic, PnCell parameters included in refinement
Hall symbol: P -2yacDx = 1.772 Mg m3
a = 5.0230 (4) ÅNeutron radiation
b = 5.9846 (4) ÅT = 100 K
c = 5.4946 (5) ÅParticle morphology: Block
β = 114.654 (8)°Colourless
Data collection top
PEARL, ISIS
diffractometer
Scan method: Time-of-flight
Radiation source: Pulsed neutron sourceAbsorption correction: empirical (using intensity measurements)
Calculated using custom ISIS software.
Specimen mounting: TiZr Capsule
Data collection mode: transmission
Refinement top
Least-squares matrix: selected elements only0 restraints
Rp = 2.86621 constraints
Rwp = 3.088Weighting scheme based on measured s.u.'s
Excluded region(s): excluded short TOF region above(Δ/σ)max = <0.001
Profile function: Voigt function convoluted with a back-to-back double exponential.Background function: Chebychev polynomial
65 parametersPreferred orientation correction: Modelling of preferred orientation was found to be unnecessary.
Special details top

Geometry. Distances and angles were all constrained in the refinement so the sus are 0 [this won't be the case if the values are relcalculated from the using the coordinate data block above - getting the right values depends on having the fill covariance matrix available].

Refinement. The data in this cif come from a Rietveld refinement against neutron powder data. Details of refinement are given in the paper.

Checkcif output:

091_ALERT_1_A No Wavelength found in CIF - 0.71073 Ang Assumed Please Check

Time of flight data.

340_ALERT_3_B Low Bond Precision on C-C Bonds ··············· 0.0110 Ang. 201_ALERT_2_G Isotropic non-H Atoms in Main Residue(s) ······. 5 Why ? 210_ALERT_3_G No Anisotropic ADP's Found in CIF ············.. Please Check

The structure was refined against medium resolution neutron powder data.

161_ALERT_4_C Missing or Zero su (esd) on x-coordinate for ··· N1 163_ALERT_4_C Missing or Zero su (esd) on z-coordinate for ··· N1

These coordinates were used to fix the origin.

741_ALERT_1_C Bond Calc 1.25 (2), Rep 1.25000 ······ Missing su O1 -C1 1.555 1.555 # 1 741_ALERT_1_C Bond Calc 1.250 (14), Rep 1.25000 ······ Missing su O2 -C1 1.555 1.555 # 2 ···Plus similar alerts

The s.u.s on the diatnces and angles are all exactly 0 as they were constrained during refinement.

017_ALERT_1_G Check Consistency of Scattering Type D for H1 017_ALERT_1_G Check Consistency of Scattering Type D for H2 017_ALERT_1_G Check Consistency of Scattering Type D for H3 017_ALERT_1_G Check Consistency of Scattering Type D for H4 017_ALERT_1_G Check Consistency of Scattering Type D for H5

The sample was deuterated.

720_ALERT_4_G Number of Unusual/Non-Standard Labels ·········. 5 Note 802_ALERT_4_G CIF Input Record(s) with more than 80 Characters ! Info

No action taken

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) top
xyzBiso*/Beq
O10.292 (3)0.795 (5)1.147 (3)0.42 (4)
O20.747 (2)0.667 (2)1.2731 (10)0.42 (4)
N10.7340.6927 (12)0.780.42 (4)
C10.514 (2)0.751 (3)1.1073 (15)0.42 (4)
C20.500 (2)0.806 (2)0.8301 (13)0.42 (4)
H10.2887 (10)0.752 (4)0.6782 (16)0.50 (5)
H20.525 (4)0.986 (2)0.816 (2)0.50 (5)
H30.7163 (17)0.738 (2)0.5891 (10)0.63 (6)
H40.709 (3)0.5191 (12)0.790 (2)0.63 (6)
H50.9385 (9)0.743 (2)0.9265 (15)0.63 (6)
Geometric parameters (Å, º) top
O1—C11.2500N1—H41.0500
O2—C11.2500N1—H51.0500
N1—C21.4800C2—H11.0900
C1—C21.5300C2—H21.0900
N1—H31.0500
O1—C1—O2126.00H3—N1—H5110.00
O1—C1—C2117.00N1—C2—C1112.00
O2—C1—C2117.00N1—C2—H1108.00
C2—N1—H5109.00N1—C2—H2108.00
C2—N1—H3109.00C1—C2—H1109.00
C2—N1—H4109.00C1—C2—H2109.00
H4—N1—H5110.00H1—C2—H2110.00
H3—N1—H4110.00
O1—C1—C2—N1164.0 (5)O2—C1—C2—N116.0 (5)
beta-glycine-d5 (beta_glycine_290K) top
Crystal data top
C2D5NO2V = 157.71 (1) Å3
Mr = 80.10Z = 2
Monoclinic, P21Cell parameters included in refinement
Hall symbol: P 2ybDx = 1.686 Mg m3
a = 5.0907 (2) ÅNeutron radiation
b = 6.25954 (16) ÅT = 290 K
c = 5.38710 (19) ÅParticle morphology: Block
β = 113.261 (4)°Colourless
Data collection top
PEARL, ISIS
diffractometer
Scan method: Time-of-flight
Radiation source: Pulsed neutron sourceAbsorption correction: empirical (using intensity measurements)
Calculated using custom ISIS software.
Specimen mounting: TiZr Capsule
Data collection mode: transmission
Refinement top
Least-squares matrix: selected elements only0 restraints
Rp = 3.22021 constraints
Rwp = 3.550Weighting scheme based on measured s.u.'s
Excluded region(s): excluded short TOF region above(Δ/σ)max = <0.001
Profile function: Voigt function convoluted with a back-to-back double exponential.Background function: Chebychev polynomial
62 parametersPreferred orientation correction: Modelling of preferred orientation was found to be unnecessary.
Special details top

Geometry. Distances and angles were all constrained in the refinement so the sus are 0 [this won't be the case if the values are relcalculated from the using the coordinate data block above - getting the right values depends on having the fill covariance matrix available].

Refinement. The data in this cif come from a Rietveld refinement against neutron powder data. Details of refinement are given in the paper, and the Topas instructuion file is under _iucr_refine_instructions_details.

Checkcif output:

091_ALERT_1_A No Wavelength found in CIF - 0.71073 Ang Assumed Please Check

The data were collected by the time of flight method

162_ALERT_4_C Missing or Zero su (esd) on y-coordinate for ··· N1

This atom was used to fix the origin.

340_ALERT_3_C Low Bond Precision on C-C Bonds ··············· 0.0070 Ang. 201_ALERT_2_G Isotropic non-H Atoms in Main Residue(s) ······. 5 Why ? 210_ALERT_3_G No Anisotropic ADP's Found in CIF ············.. Please Check

The structure was refined against medium resolution neutron powder data.

741_ALERT_1_C Bond Calc 1.252 (14), Rep 1.25000 ······ Missing su O1 -C1 1.555 1.555 # 1 741_ALERT_1_C Bond Calc 1.250 (9), Rep 1.25000 ······ Missing su O2 -C1 1.555 1.555 # 2 741_ALERT_1_C Bond Calc 1.480 (8), Rep 1.48000 ······ Missing su N1 -C2 1.555 1.555 # 3 ···plus other alerts like this

The s.u.s on the diatnces and angles are all exactly 0 as they were constrained during refinement.

017_ALERT_1_G Check Consistency of Scattering Type D for H1 017_ALERT_1_G Check Consistency of Scattering Type D for H2 017_ALERT_1_G Check Consistency of Scattering Type D for H3 017_ALERT_1_G Check Consistency of Scattering Type D for H4 017_ALERT_1_G Check Consistency of Scattering Type D for H5

The sample was deuterated

720_ALERT_4_G Number of Unusual/Non-Standard Labels ·········. 5 802_ALERT_4_G CIF Input Record(s) with more than 80 Characters ! Info

No action taken.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) top
xyzBiso*/Beq
O10.410 (2)0.378 (3)0.5987 (13)1.33 (4)
O20.8795 (16)0.3217 (13)0.7432 (8)1.33 (4)
N10.8522 (7)0.250.2370 (7)1.33 (4)
C10.6367 (16)0.3537 (16)0.5629 (10)1.33 (4)
C20.6146 (14)0.3646 (14)0.2716 (9)1.33 (4)
H10.4136 (8)0.292 (2)0.1385 (10)1.60 (5)
H20.621 (3)0.5315 (14)0.2161 (12)1.60 (5)
H30.8504 (12)0.2867 (10)0.0461 (9)2.00 (6)
H40.824 (2)0.08499 (18)0.2520 (17)2.00 (6)
H51.0465 (9)0.2989 (13)0.3890 (11)2.00 (6)
Geometric parameters (Å, º) top
O1—C11.2500N1—H41.0500
O2—C11.2500N1—H51.0500
N1—C21.4800C2—H11.0900
C1—C21.5300C2—H21.0900
N1—H31.0500
O1—C1—O2126.00H3—N1—H5110.00
O1—C1—C2117.00N1—C2—C1112.00
O2—C1—C2117.00N1—C2—H1108.00
C2—N1—H5109.00N1—C2—H2108.00
C2—N1—H3109.00C1—C2—H1109.00
C2—N1—H4109.00C1—C2—H2109.00
H4—N1—H5110.00H1—C2—H2110.00
H3—N1—H4110.00
O1—C1—C2—N1156.3 (3)O2—C1—C2—N123.7 (3)
zeta-glycine-d5 (zeta_glycine_290K) top
Crystal data top
C2D5NO2γ = 103.55 (3)°
Mr = 80.10V = 155.85 (9) Å3
Triclinic, I1Z = 2
Hall symbol: I1Cell parameters included in refinement
a = 5.1029 (16) ÅDx = 1.707 Mg m3
b = 6.3450 (12) ÅNeutron radiation
c = 5.4331 (18) ÅT = 290 K
α = 85.91 (3)°Particle morphology: Block
β = 114.26 (3)°Colourless
Data collection top
PEARL, ISIS
diffractometer
Scan method: Time-of-flight
Radiation source: Pulsed neutron sourceAbsorption correction: empirical (using intensity measurements)
Calculated using custom ISIS software.
Specimen mounting: TiZr Capsule
Data collection mode: transmission
Refinement top
Least-squares matrix: selected elements only0 restraints
Rp = 3.22022 constraints
Rwp = 3.550Weighting scheme based on measured s.u.'s
Excluded region(s): excluded short TOF region above(Δ/σ)max = <0.001
Profile function: Voigt function convoluted with a back-to-back double exponential.Background function: Chebychev polynomial
62 parametersPreferred orientation correction: Modelling of preferred orientation was found to be unnecessary.
Special details top

Geometry. Distances and angles were all constrained in the refinement so the sus are 0 [this won't be the case if the values are relcalculated from the using the coordinate data block above - getting the right values depends on having the fill covariance matrix available].

Refinement. The data in this cif come from a Rietveld refinement against neutron powder data. Details of refinement are given in the paper.

Checkcif output:

091_ALERT_1_A No Wavelength found in CIF - 0.71073 Ang Assumed Please Check

The data were collected by the time of flight method

340_ALERT_3_B Low Bond Precision on C-C Bonds ··············· 0.0400 Ang. 201_ALERT_2_G Isotropic non-H Atoms in Main Residue(s) ······. 5 Why ? 210_ALERT_3_G No Anisotropic ADP's Found in CIF ············.. Please Check

The structure was refined against medium resolution neutron powder data.

155_ALERT_4_C The Triclinic Unitcell is NOT Reduced ·········. Please Do ! 128_ALERT_4_G Suggested Setting for Input Space-Group I1 P1 Note

The setting is chosen to show the relationship to the epsilon phase. The matrix relating the prediction and refinement cell settings is [-1 1 0] [ 0 -1 1] [ 1 0 1]

161_ALERT_4_C Missing or Zero su (esd) on x-coordinate for ··· N1 162_ALERT_4_C Missing or Zero su (esd) on y-coordinate for ··· N1 163_ALERT_4_C Missing or Zero su (esd) on z-coordinate for ··· N1

This atom was used to fix the origin.

723_ALERT_1_C Torsion Calc -147.21, Rep -149 (2) Dev··· 1.79 Sigma O1 -C1 -C2 -N1 1.555 1.555 1.555 1.555 1 741_ALERT_1_C Bond Calc 1.25 (11), Rep 1.25000 ······ Missing su O1 -C1 1.555 1.555 # 1 741_ALERT_1_C Bond Calc 1.25 (7), Rep 1.25000 ······ Missing su O2 -C1 1.555 1.555 # 2

···plus numerous similar alerts

The s.u.s on the diatnces and angles are all exactly 0 as they were constrained during refinement.

017_ALERT_1_G Check Consistency of Scattering Type D for H1 017_ALERT_1_G Check Consistency of Scattering Type D for H2 017_ALERT_1_G Check Consistency of Scattering Type D for H3 017_ALERT_1_G Check Consistency of Scattering Type D for H4 017_ALERT_1_G Check Consistency of Scattering Type D for H5

The sample was deuterated.

154_ALERT_1_G The su's on the Cell Angles are Equal ·········. 0.03000 Deg. 720_ALERT_4_G Number of Unusual/Non-Standard Labels ·········. 5 802_ALERT_4_G CIF Input Record(s) with more than 80 Characters ! Info

No action taken

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) top
xyzBiso*/Beq
N10.711440.681320.779260.7 (2)
C10.527 (13)0.767 (13)1.112 (5)0.7 (2)
C20.500 (11)0.784 (11)0.821 (5)0.7 (2)
O10.297 (16)0.75 (2)1.148 (7)0.7 (2)
O20.778 (11)0.767 (11)1.293 (3)0.7 (2)
H10.277 (7)0.703 (18)0.686 (5)0.8 (3)
H20.54 (2)0.955 (11)0.772 (8)0.8 (3)
H30.731 (8)0.737 (7)0.599 (5)1.0 (3)
H40.633 (12)0.5123 (11)0.766 (10)1.0 (3)
H50.917 (3)0.724 (9)0.943 (5)1.0 (3)
Geometric parameters (Å, º) top
O1—C11.2500N1—H41.0500
O2—C11.2500N1—H51.0500
N1—C21.4800C2—H11.0900
C1—C21.5300C2—H21.0900
N1—H31.0500
O1—C1—O2126.00H3—N1—H5110.00
O1—C1—C2117.00N1—C2—C1112.00
O2—C1—C2117.00N1—C2—H1108.00
C2—N1—H5109.00N1—C2—H2108.00
C2—N1—H3109.00C1—C2—H1109.00
C2—N1—H4109.00C1—C2—H2109.00
H4—N1—H5110.00H1—C2—H2110.00
H3—N1—H4110.00
O1—C1—C2—N1149 (2)O2—C1—C2—N131 (2)
 

Acknowledgements

We thank Professor Qiang Zhu of the University of Nevada Las Vegas who put our two groups in contact when he learned of our shared interest in ζ-glycine, and Drs L. Martin-Samos, L. Paulatto and F. Pellegrini for suggestions on improving the manuscript. We also thank Dr Andrew Alexander for assistance with crystal growth of γ-glycine-d5.

Funding information

The following funding is acknowledged: Science and Technology Facilities Council/ISIS (award No. RB1520301); Engineering and Physical Sciences Research Council (award No. DTA); SISSA, CINECA and PRACE (award No. 2011050736); H2020 European Research Council (award No. 676531).

References

First citationBerland, K., Cooper, V. R., Lee, K., Schröder, E., Thonhauser, T., Hyldgaard, P. & Lundqvist, B. I. (2015). Rep. Prog. Phys. 78, 066501.  Web of Science CrossRef PubMed Google Scholar
First citationBoldyreva, E. V., Ivashevskaya, S. N., Sowa, H., Ahsbahs, H. & Weber, H.-P. (2005). Z. Kristallogr. Cryst. Mater. 220, 50–57.  CAS Google Scholar
First citationBordallo, H. N., Boldyreva, E. V., Buchsteiner, A., Koza, M. M. & Landsgesell, S. (2008). J. Phys. Chem. B, 112, 8748–8759.  Web of Science CrossRef PubMed CAS Google Scholar
First citationBrandenburg, K. (2004). DIAMOND. Crystal Impact GbR, Bonn, Germany.  Google Scholar
First citationBull, C. L., Funnell, N. P., Tucker, M. G., Hull, S., Francis, D. J. & Marshall, W. G. (2016). High. Press. Res. 36, 493–511.  Web of Science CrossRef Google Scholar
First citationChisholm, J. A., Motherwell, S., Tulip, P. R., Parsons, S. & Clark, S. J. (2005). Cryst. Growth Des. 5, 1437–1442.  Web of Science CrossRef CAS Google Scholar
First citationCoelho, A. (2015). TOPAS-Academic. Version 5. Coelho Software, Brisbane, Australia.  Google Scholar
First citationDawson, A., Allan, D. R., Belmonte, S. A., Clark, S. J., David, W. I., McGregor, P. A., Parsons, S., Pulham, C. R. & Sawyer, L. (2005). Cryst. Growth Des. 5, 1415–1427.  Web of Science CSD CrossRef CAS Google Scholar
First citationDestro, R., Roversi, P., Barzaghi, M. & Marsh, R. E. (2000). J. Phys. Chem. A, 104, 1047–1054.  Web of Science CSD CrossRef CAS Google Scholar
First citationFortes, A. D., Wood, I. G., Alfredsson, M., Vocadlo, L., Knight, K. S., Marshall, W. G., Tucker, M. G. & Fernandez-Alonso, F. (2007). High Pressure Res. 27, 201–212.  Web of Science CrossRef CAS Google Scholar
First citationFortes, A. D., Wood, I. G., Alfredsson, M., Vocadlo, L., Knight, K. S., Marshall, W. G., Tucker, M. G. & Fernandez-Alonso, F. (2012). High Pressure Res. 32, 337–337.  Google Scholar
First citationFrisch, M. J. et al. (2009). GAUSSIAN09. Gaussian Inc., Wallingford, Connecticut, USA.  Google Scholar
First citationGavezzotti, A. (2005). Z. Kristallogr. 220, 499–510.  Web of Science CrossRef CAS Google Scholar
First citationGavezzotti, A. (2011). New J. Chem. 35, 1360–1368.  Web of Science CrossRef CAS Google Scholar
First citationGiannozzi, P. et al. (2009). J. Phys. Condens. Matter, 21, 395502.  Web of Science CrossRef PubMed Google Scholar
First citationGoryainov, S. V., Boldyreva, E. V. & Kolesnik, E. N. (2006). Chem. Phys. Lett. 419, 496–500.  Web of Science CrossRef CAS Google Scholar
First citationGoryainov, S. V., Kolesnik, E. N. & Boldyreva, E. V. (2005). Phys. B Condens. Matter, 357, 340–347.  Web of Science CrossRef CAS Google Scholar
First citationHohenberg, P. & Kohn, W. (1964). Phys. Rev. 136, B864–B871.  CrossRef Web of Science Google Scholar
First citationHohenstein, E. G. & Sherrill, C. D. (2010a). J. Chem. Phys. 133, 014101.  Web of Science CrossRef PubMed Google Scholar
First citationHohenstein, E. G. & Sherrill, C. D. (2010b). J. Chem. Phys. 133, 104107.  Web of Science CrossRef PubMed Google Scholar
First citationHohenstein, E. G. & Sherrill, C. D. (2012). WIREs Comput. Mol. Sci. 2, 304–326.  Web of Science CrossRef CAS Google Scholar
First citationHunter, C. A. & Prohens, R. (2017). CrystEngComm, 19, 23–26.  Web of Science CrossRef CAS Google Scholar
First citationJeziorski, B., Moszynski, R. & Szalewicz, K. (1994). Chem. Rev. 94, 1887–1930.  CrossRef CAS Web of Science Google Scholar
First citationKüçükbenli, E., Monni, M., Adetunji, B. I., Ge, X., Adebayo, G. A., Marzari, N., de Gironcoli, S. & Dal Corso, A. (2014). arXiv, 1404.3015.  Google Scholar
First citationKvick, Å., Canning, W. M., Koetzle, T. F. & Williams, G. J. B. (1980). Acta Cryst. B36, 115–120.  CSD CrossRef CAS IUCr Journals Web of Science Google Scholar
First citationLund, A. M., Pagola, G. I., Orendt, A. M., Ferraro, M. B. & Facelli, J. C. (2015). Chem. Phys. Lett. 626, 20–24.  Web of Science CrossRef CAS PubMed Google Scholar
First citationLyakhov, A. O., Oganov, A. R., Stokes, H. T. & Zhu, Q. (2013). Comput. Phys. Commun. 184, 1172–1182.  Web of Science CrossRef CAS Google Scholar
First citationMacrae, C. F., Bruno, I. J., Chisholm, J. A., Edgington, P. R., McCabe, P., Pidcock, E., Rodriguez-Monge, L., Taylor, R., van de Streek, J. & Wood, P. A. (2008). J. Appl. Cryst. 41, 466–470.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationMarom, N., DiStasio, R. A., Atalla, V., Levchenko, S., Reilly, A. M., Chelikowsky, J. R., Leiserowitz, L. & Tkatchenko, A. (2013). Angew. Chem. Int. Ed. 52, 6629–6632.  Web of Science CrossRef CAS Google Scholar
First citationMoggach, S. A., Marshall, W. G., Rogers, D. M. & Parsons, S. (2015). CrystEngComm, 17, 5315–5328.  Web of Science CSD CrossRef CAS Google Scholar
First citationMurli, C., Sharma, S. M., Karmakar, S. & Sikka, S. K. (2003). Physica B, 339, 23–30.  Web of Science CrossRef CAS Google Scholar
First citationOganov, A. R. & Valle, M. (2009). J. Chem. Phys. 130, 104504–104512.  Web of Science CrossRef PubMed Google Scholar
First citationPerlovich, G., Hansen, L. & Bauer-Brandl, A. (2001). J. Therm. Anal. Calorim. 66, 699–715.  Web of Science CrossRef CAS Google Scholar
First citationQi, S., Avalle, P., Saklatvala, R. & Craig, D. Q. M. (2008). Eur. J. Pharm. Biopharm. 69, 364–371.  Web of Science CrossRef PubMed CAS Google Scholar
First citationReilly, A. M. et al. (2016). Acta Cryst. B72, 439–459.  Web of Science CrossRef IUCr Journals Google Scholar
First citationRossi, A., Savioli, A., Bini, M., Capsoni, D., Massarotti, V., Bettini, R., Gazzaniga, A., Sangalli, M. E. & Giordano, F. (2003). Thermochim. Acta, 406, 55–67.  Web of Science CrossRef CAS Google Scholar
First citationSabatini, R., Küçükbenli, E., Kolb, B., Thonhauser, T. & de Gironcoli, S. (2012). J. Phys. Condens. Matter, 24, 424209.  Web of Science CrossRef PubMed Google Scholar
First citationSheldrick, G. M. (2001). SHELXTL-XP. Version 6.01. University of Göttingen, Germany.  Google Scholar
First citationSpek, A. L. (2003). J. Appl. Cryst. 36, 7–13.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationStone, A. J. (2013). The Theory of Intermolecular Forces. Oxford University Press.  Google Scholar
First citationTurney, J. M. et al. (2012). WIREs Comput. Mol. Sci. 2, 556–565.  Web of Science CrossRef CAS Google Scholar
First citationWoodley, S. M. & Catlow, R. (2008). Nat. Mater. 7, 937–946.  Web of Science CrossRef PubMed CAS Google Scholar
First citationXu, W., Zhu, Q. & Hu, C. T. (2017). Angew. Chem. 129, 2062–2066.  CrossRef Google Scholar
First citationZaccaro, J., Matic, J., Myerson, A. S. & Garetz, B. A. (2001). Cryst. Growth Des. 1, 5–8.  Web of Science CrossRef CAS Google Scholar
First citationZhu, Q., Oganov, A. R., Glass, C. W. & Stokes, H. T. (2012). Acta Cryst. B68, 215–226.  Web of Science CrossRef IUCr Journals Google Scholar

This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.

IUCrJ
Volume 4| Part 5| September 2017| Pages 569-574
ISSN: 2052-2525