Topotactic elimination of water across a C-C ligand bond in a dense 3-D metal-organic framework

Article abstract of DOI:10.1039/c4cc06136e

Upon heating, lithium l-malate undergoes topotactic dehydration to form a phase containing the unsaturated fumarate ligand, in which the original 3-D framework remains intact. Insight into this unusual transformation has been obtained by single crystal X-ray diffraction, MAS-NMR, in situ powder X-ray diffraction and DFT calculations. This journal is

Full text of DOI:10.1039/c4cc06136e

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Topotactic elimination of water across a C–C
ligand bond in a dense 3-D metal–organic
framework†
Cite this: Chem. Commun., 2014,
50, 13292
Received 6th August 2014,
Accepted 8th September 2014
Hamish H.-M. Yeung,ab Monica Kosa,c John M. Griffin,d Clare P. Grey,d
Dan T. Majorc and Anthony K. Cheetham*a
DOI: 10.1039/c4cc06136e
www.rsc.org/chemcomm
Upon heating, lithium L-malate undergoes topotactic dehydration to
We describe herein the first example of a topochemical elimina-
form a phase containing the unsaturated fumarate ligand, in which the tion reaction within a non-porous MOF, which occurs spontaneously
original 3-D framework remains intact. Insight into this unusual in a single crystal-to-single crystal manner despite a lack of accessible
transformation has been obtained by single crystal X-ray diffraction, porosity in the conventional sense. The crystal structure and thermo-
MAS-NMR, in situ powder X-ray diffraction and DFT calculations.
gravimetry of lithium L-malate [Li2(L-C4H4O5)], 1, was described in a
previous article, in which we reported that the relative flexibility of
In recent years, metal–organic frameworks (MOFs) have opened Li–O bonds, compared to stronger transition metal–oxygen bonds,
up new avenues in the search for novel functional materials, allows MOF frameworks to distort enough to accommodate ligand
owing to their wide compositional scope, modular structure solution solutions.16 1 undergoes a two-step mass loss on heating,
and facile synthesis.1–3 Coordination bonding between metal whereby decomposition of the main framework at 400 1C is
ions and bridging organic ligands gives rise to extended 1-, 2- preceded by a loss of approx. 12 wt%, which corresponds to the
and 3-dimensional structures, with diverse properties, including combined mass of one proton and one hydroxyl group per malate
gas sorption,4 magnetism,5 chirality6 and multiferroicity,7 arising ligand. Upon closer investigation, it was observed that large single
from the synergy between the constituent organic and inorganic crystals remained intact after various heat treatments between 280–
units. MOFs have also been used as templates for post-synthetic 350 1C, allowing the crystal structure of the resulting mixed-ligand
functionalization8 and topochemical reactions,9–11 in which the MOF, [Li2(L-C4H4O5)1Àx(C4H2O4)x], 2(x), to be determined by single
orientations of the organic molecules in the precursor crystal crystal X-ray diffraction for various values of x(see Table S1 in the ESI†).
direct the structure of the product.12,13 The crystal engineering
The crystal structure of 2(x) consists of a 3-D framework of LiO4
approach to topochemical reactions has traditionally been domi- tetrahedra bridged by dicarboxylate L-malate and fumarate ligands,
nated by non-covalent forces, which can accommodate the crystal which reside in the same crystallographic location with fractional
strain accompanying bond breaking and formation.14 Despite occupancies of (1 À x) and x, respectively (Fig. 1). The trans-
their relative rigidity, however, coordination bonds impart extra configuration of the fumarate ligand suggests that water has been
stability to organic molecules, allowing access to more demanding eliminated from across the L-malate C–C bond in a stereospecific
conditions, such as higher temperatures and pressures, as well as manner directed by its fixed conformation within the 3-D MOF
giving excellent control over molecular conformations.15
(Scheme 1). All compositions of 2(x) analysed were found to be
isostructural with 1 in space group R3, indicating that framework
chirality is preserved even when 80% of ligands present are achiral.
Whilst the parent compound 1 exhibits disorder between two
crystallographic positions of the L-malate hydroxyl group,16 the struc-
ture determinations of 2(x) only found significant electron density in
one position, O5. This suggests a preference for dehydration of the
minor site over the major site, which may arise from local differences
in stability owing to hydrogen bonding. Further evidence was
obtained from MAS-NMR spectra of 2(x), which revealed small
but significant changes in chemical shifts upon partial dehydration
(see S2, Fig. S1 and Table S2, ESI†). 1H and 13C NMR spectra of 2(0.77)
in D2O, and FTIR spectra of 2(x) at different stages of dehydration are
consistent with the above observations (see Fig. S2–S4, ESI†).
a Department of Materials Science and Metallurgy, Cambridge University,
Charles Babbage Road, Cambridge, UK CB3 0FS. E-mail: akc30@cam.ac.uk
b International Center for Young Scientists, WPI Center for Materials
Nanoarchitectonics, National Institute of Materials Science, 1-1 Namiki, Tsukuba,
Ibaraki, 305-0044, Japan
c Department of Chemistry, Faculty of Exact Sciences and the Lise Meitner-Minerva
Center of Computational Quantum Chemistry, Bar Ilan University, Ramat Gan,
Israel IL-52900
d Department of Chemistry, Cambridge University, Lensfield Road, Cambridge,
UK CB2 1EW
† Electronic supplementary information (ESI) available: Experimental proce-
dures, crystallography, NMR, DFT, and XRD. CCDC 1018021–1018024. For ESI
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c4cc06136e
13292 | Chem. Commun., 2014, 50, 13292--13295
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
Fig. 1 Ortep extended asymmetric unit of 2(0.77) determined by single
crystal X-ray diffraction, showing the original L-malate atoms joined by
black lines and the unsaturated fumarate ligand (dashed blue lines). ADP
ellipsoids are shown at 50% (20% for H).
Fig. 2 VT-powder X-ray diffraction data for 2(x) in R3, showing the
expansion, topotactic dehydration and decomposition processes in blue,
purple and red, respectively.
(Fig. 2 and Fig. S6–S10, ESI†). Whilst the variations in unit cell
parameters are highly anisotropic, the average volumetric thermal
expansion of each process is relatively constant (aV = 97(8) Â
10À6 KÀ1, 115(18) Â 10À6 KÀ1 and 103(9) Â 10À6 KÀ1, for (i)–(iii),
Scheme 1 Stereospecificity of the topotactic dehydration of Li2(L-malate)
directed by its conformation in the MOF.
The relative formation enthalpies of 2(x) at various composi- respectively). It should be noted that the transition temperatures
tions were calculated at 0 K using Density Functional Theory were lower than observed in single crystal studies, partly due to
(DFT) within the PBE functional approximation,17 alone and kinetic effects of powder particle size reduction.
with dispersion corrections D218 and D319 (Table S3, ESI†).
From 30 1C to 150 1C (process i), thermal expansion of 1
Enthalpies were also calculated including zero-point energy occurs predominantly in the crystallographically equivalent
and vibrational contributions at 300 K, showing similar trends a- and b-directions, with negligible change in c. The mechanical
and so will not be discussed here. The dehydration reaction was properties of 1, as reported previously,16 depend strongly on the
shown to be endothermic, as expected for elimination of water hydrogen bonded trimer of hydroxyl groups, which causes
to form a CQC bond. At the PBE + D3 level of theory, the contraction in the ab-plane. At raised temperatures, however,
complete dehydration reaction requires 81.8 kJ molÀ1 and they may be more easily broken by thermal vibrations than the
extrapolation of the data to x = 0 reveals that this includes stronger covalent bonds which connect the MOF along the
13.7(3) kJ molÀ1 needed to break the strong hydrogen bonding c-axis. Hence the coefficients of thermal expansion in the two
between hydroxyl groups before the first water is lost (Fig. S5, orthogonal directions, aa and ac, are 49.0(1.5) Â 10À6 KÀ1 and
ESI†). The unfavourable enthalpy is primarily offset, of course, À1(11) Â 10À6
K
À1, respectively.
In the region 150–240 1C (process ii), the c-axis undergoes
by the entropy of the water vapour at the decomposition
temperature. On the basis of a value for the entropy of water rapid expansion (ac = 311(19) Â 10À6
K
À1), whilst the a-axis
À1). Although these
of 188.8 J molÀ1 KÀ1 20
160 1C, in good agreement with our observations.
,
decomposition would be expected at contracts significantly (aa = À98(7) Â 10À6
K
values are one order of magnitude larger than common materials
The structural evolution of 2(x) with increasing temperature was and similar to the colossal positive and negative thermal expan-
studied using in situ variable-temperature (VT) powder X-ray diffrac- sion observed in silver(I) hexacyanocobaltate(III)-type materials
tion. Non-linear shifts in the positions of the Bragg peaks were recently reported,21 the behaviour in this system is irreversible
observed and Le Bail refinement of the unit cell parameters revealed owing to the accompanying dehydration process. The effect of
three distinct processes: (i) expansion of the unit cell of 1, hydrogen bonding has already been weakened below 150 1C, and
(ii) dehydration to form 2(x), and (iii) expansion of the unit cell so expansion along c must be due to the lengthening of the ligand
of the dehydrated MOF alongside decomposition to form Li2CO3 and the increase in its rigidity (owing to p-conjugation).
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 13292--13295 | 13293

View Article Online
ChemComm
Communication
Above 240 1C, the distinct change in aa and ac (24(2) Â 10À6 KÀ1 theoretical calculations involving hybrid systems such as MOFs,
and 55(8) Â 10À6
K
À1, respectively), indicates that the main as observed previously.22–24
dehydration process is essentially complete. The resulting MOF
In summary, we have described the topotactic transforma-
undergoes anisotropic thermal expansion up until complete tion of a dense 3-D MOF, 1, in which the stereospecific
decomposition to form Li2CO3, which precludes the formation of elimination of water occurs across the ligand’s central C–C
the fully dehydrated material, Li2(C4H2O4), 2(1). From single crystal bond to form a CQC double bond with a trans-configuration.
studies, we estimate that the maximum extent of dehydration in Remarkably, despite the lack of conventional porosity, the
this system is in the region 0.8 r xmax r 1.
water molecules produced can escape the framework leaving
Attempts to synthesize hypothetical 2(1) directly from fumaric the ligand–lithium binding intact, and so the resulting frame-
acid via both solution and solvothermal crystallization resulted work retains the connectivity and crystallographic symmetry
%
in the discovery of a polymorphic phase in P1, 3, whose frame- found in the starting material. Furthermore, the dehydration
work connectivity is very different to 2(1) (I1O2 and I3O0, respec- process results in colossal and irreversible positive and negative
tively, according to the notation of Cheetham et al.).2 Notably, thermal expansion of the MOF framework. This is another
2(1) was not observed under any conditions employed, and so a striking example of the remarkable flexibility that is exhibited
detailed comparison of the two structures was undertaken (see by many MOFs,16,21,25 resulting in this case from the combination
S3 and Fig. S11–S13, ESI†). Several similarities in the structures of chemical substitution and framework dynamics induced by
of 2(1) and 3 were found, but the average Li–O bond valence sum thermal treatment. Our discovery reveals a new route to Post
of 3 is 6% lower than that of 2(1) (1.02 and 1.08, respectively), Synthetic Modification in dense MOFs, via the thermal dehydra-
suggesting that the ionic bonding in 2(1) is stronger. In contrast, tion of the ligand moiety. Phase 2 could potentially undergo
the X-ray density of 3 is 1.809 g cmÀ3, 25% greater than that of further functionalization by addition to the CQC group, but
the hypothetical fully dehydrated phase, 2(1), suggesting that unlike phase 3 is unobtainable via conventional means. The
dispersion forces are stronger in 3.
broad temperature and time-dependent nature of the thermal
The relative enthalpies of 3 and 2(1) were calculated using DFT dehydration allow for good stoichiometry control in the resulting
methods at 0 K, revealing that, without dispersion corrections, mixed-ligand MOF. On the other hand, our observations demon-
2(1) is 9.6 kJ molÀ1 more stable than 3. However, when dispersion strate that the additional stability imparted by MOF formation to
corrections are applied, the relative energies are reversed, such organic molecules (L-malic acid decomposes completely to form
that 3 is more stable than 2(1) by 12.44–13.20 kJ molÀ1, depending gaseous products above 150 1C) allows simple and efficient
on the level of theory (see Table S4, ESI† for values) This confirms thermal reactions in the solid state to be utilised that might
that dispersion forces related to the packing density dominate the otherwise require more complex reaction routes in solution. By
synthetic landscape for the formation of Li2(C4H2O4) in solution, stabilizing the molecule in a 3-D framework, the functionalization
rather than the ionic contributions from Li–O bonding, and that also occurs in a stereospecific manner, which demonstrates that
topotactic dehydration offers a unique route to the otherwise such solid state reactions of MOFs could be usefully integrated
unobtainable phase 2 (Scheme 2). Furthermore, it emphasizes the into more mainstream organic synthesis methodology.
importance of full consideration of non-covalent interactions in
We thank Giulio Lampronti and Sebastian Henke for assis-
tance with VT-XRD measurements and acknowledge the financial
support of the EPSRC and ERC (Advanced Investigator Award,
AKC). HHMY acknowledges support from the World Premier
International Research Center Initiative on Materials Nanoarchi-
tectonics (WPI-MANA) from MEXT, Japan.
Notes and references
1 S. Kitagawa, R. Kitaura and S. Noro, Angew. Chem., Int. Ed., 2004, 43,
2334–2375.
2 A. K. Cheetham, C. N. R. Rao and R. K. Feller, Chem. Commun., 2006,
4780–4795.
´
3 G. Ferey, Chem. Soc. Rev., 2008, 37, 191–214.
4 X. Zhao, B. Xiao, A. J. Fletcher, K. M. Thomas, D. Bradshaw and
M. J. Rosseinsky, Science, 2004, 306, 1012–1015.
5 M. Kurmoo, Chem. Soc. Rev., 2009, 38, 1353–1379.
6 D. Bradshaw, J. B. Claridge, E. J. Cussen, T. J. Prior and
M. J. Rosseinsky, Acc. Chem. Res., 2005, 38, 273–282.
7 P. Jain, V. Ramachandran, R. J. Clark, H. D. Zhou, B. H. Toby,
N. S. Dalal, H. W. Kroto and A. K. Cheetham, J. Am. Chem. Soc.,
2009, 131, 13625–13627.
8 S. M. Cohen, Chem. Rev., 2012, 112, 970–1000.
9 G. M. J. Schmidt, Pure Appl. Chem., 1971, 27, 647–678.
10 W. L. Leong and J. J. Vittal, Chem. Rev., 2011, 111, 688–764.
11 G. K. Kole and J. J. Vittal, Chem. Soc. Rev., 2013, 42, 1755–1775.
12 R. Medishetty, L. L. Koh, G. K. Kole and J. J. Vittal, Angew. Chem., Int.
Ed., 2011, 50, 10949–10952.
Scheme 2 Reaction routes from precursor ligands L-malic acid and fumaric
acid (H2(L-mal) and H2(fum), respectively) to 1–3 via MOF assembly and
topotactic transformation, and the calculated relative formation enthalpies of
each phase at the PBE + D3 level of theory.
13294 | Chem. Commun., 2014, 50, 13292--13295
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
13 M.-H. Xie, X.-L. Yang and C.-D. Wu, Chem. – Eur. J., 2011, 17, 20 J. D. Cox, D. D. Wagman and V. A. Medvedev, CODATA Key Values
11424–11427.
for Thermodynamics, Hemisphere Publishing Co., New York, 1984.
21 A. L. Goodwin, M. Calleja, M. J. Conterio, M. T. Dove, J. S. O. Evans,
D. A. Keen, L. Peters and M. G. Tucker, Science, 2008, 319, 794–797.
22 H. H.-M. Yeung, M. Kosa, M. Parrinello, P. M. Forster and
A. K. Cheetham, Cryst. Growth Des., 2011, 11, 221–230.
14 K. Biradha and R. Santra, Chem. Soc. Rev., 2013, 42, 950–967.
15 H. H.-M. Yeung and A. K. Cheetham, Dalton Trans., 2014, 43,
95–102.
¨
16 H. H.-M. Yeung, W. Li, P. J. Saines, T. K. J. Koster, C. P. Grey and
A. K. Cheetham, Angew. Chem., Int. Ed., 2013, 52, 5544–5547.
17 J. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77,
3865–3868.
23 B. K. Chang, N. C. Bristowe, P. D. Bristowe and A. K. Cheetham,
Phys. Chem. Chem. Phys., 2012, 14, 7059–7064.
24 H. H.-M. Yeung, M. Kosa, M. Parrinello and A. K. Cheetham, Cryst.
Growth Des., 2013, 13, 3705–3715.
18 S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
19 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 25 L. Sarkisov, R. L. Martin, M. Haranczyk and B. Smit, J. Am. Chem.
132, 154104.
Soc., 2014, 136, 2228–2231.
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 13292--13295 | 13295