Enhancement of Proton Conductivity in Nonporous Metal–Organic

by modulation of their structures is a key target but remains a challenging task. .... Each Ln(III) center is coordinated to eight oxygens from si...
0 downloads 0 Views 1MB Size
Subscriber access provided by University of Sunderland

Article

Enhancement of proton conductivity in non-porous metal-organic frameworks: the role of framework proton density and humidity Simona Pili, Peter Rought, Daniil I. Kolokolov, Longfei Lin, Ivan da Silva, Yongqiang Cheng, Christopher Marsh, Ian P. Silverwood, Victoria García-Sakai, Ming Li, Jeremy J. Titman, Lyndsey Knight, Luke L. Daemen, Anibal J. Ramirez-Cuesta, Chiu C. Tang, Alexander G. Stepanov, Sihai Yang, and Martin Schröder Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.8b02765 • Publication Date (Web): 24 Sep 2018 Downloaded from http://pubs.acs.org on September 25, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Enhancement of proton conductivity in non-porous metal-organic frameworks: the role of framework proton density and humidity

Simona Pili,1† Peter Rought,1†, Daniil I. Kolokolov,2,3† Longfei Lin,1 Ivan da Silva,4 Yongqiang Cheng,5 Christopher Marsh,1 Ian P. Silverwood,4 Victoria García-Sakai,4 Ming Li,6 Jeremy J. Titman,7 Lyndsey Knight,7 Luke L. Daemen,5 Anibal J. Ramirez-Cuesta,5 Chiu C. Tang,8 Alexander G. Stepanov,2,3 Sihai Yang1* and Martin Schröder1*

1. School of Chemistry, University of Manchester, Manchester M13 9PL, U.K. [email protected]; [email protected] 2. Novosibirsk State University, Novosibirsk 630090, Russia. 3. Boreskov Institute of Catalysis, Siberian Branch of Russian Academy of Sciences, Prospekt Akademika Lavrentieva 5, Novosibirsk 630090, Russia 4. ISIS Pulsed Neutron and Muon Source, Rutherford Appleton Laboratory, Oxfordshire OX11 0QX, U.K. 5. Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, United States. 6. School of Engineering, University of Nottingham, Nottingham NG7 2RD, U.K. 7. School of Chemistry, University of Nottingham, Nottingham NG7 2RD, U.K. 8. Diamond Light Source, Harwell Science and Innovation Campus, Oxfordshire OX11 0DE, U.K. †

These authors contribute equality to this work.

1 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

ABSTRACT Owing to their inherent pore structure, porous metal-organic frameworks (MOFs) can undergo post-synthetic modification, such as loading extra-framework proton carriers. However, strategies for improving the proton conductivity for non-porous MOFs are largely lacking, although increasing numbers of non-porous MOFs exhibit promising proton conductivities. Often, high humidity is required for non-porous MOFs to achieve high conductivities, but to date no clear mechanisms have been experimentally identified. Here we describe the new materials MFM-550(M), [M(HL1)], (H4L1 = biphenyl-4,4'-diphosphonic acid; M = La, Ce, Nd, Sm, Gd, Ho), MFM-550(Ba), [Ba(H2 L1)], and MFM-555(M), [M(HL2)], (H4L2 = benzene-1,4-diphosphonic acid; M = La, Ce, Nd, Sm, Gd, Ho), and report enhanced proton conductivities in these non-porous materials by (i) replacing the metal ion to one with a lower oxidation state, (ii) reducing the length of the organic ligand, and (iii) introducing additional acidic protons on MOF surface. Increased framework proton density in these materials can lead to an enhancement in proton conductivity of up to four orders of magnitude. Additionally, we report a comprehensive investigation using in situ 2H NMR and neutron spectroscopy, coupled with molecular dynamic modelling, to elucidate the role of humidity in assembling interconnected networks for proton hopping. This study constructs a relationship between framework proton density and the corresponding proton conductivity in non-porous MOFs, and directly explains the role of both surface protons and external water in assembling the proton conducting pathways.

2 ACS Paragon Plus Environment

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

INTRODUCTION The development of proton conductors based upon MOFs is of increasing interest because of their very rich structural diversity owing to the availability of very many combinations of metal ions/cluster and organic linkers.1-14 Enhancement of the proton conductivity of MOFs by modulation of their structures is a key target but remains a challenging task. To date, the majority of research efforts have been focused on porous MOFs primarily because their extended pore structure enables design and post-synthetic modification to achieve targeted properties, and three methods have been reported to improve proton conductivity in porous MOFs.1-14 These involve the introduction of proton carriers as counter ions,1-3 the functionalisation of the ligands with proton-rich functional groups (which remain accessible in the pore of resultant MOFs)4-7 and the doping of MOFs with neutral proton carrier guests.1,8-14 Recently, the impact of structural defects on the proton conductivity of modified MOFs has also been reported.15,16 More recently, attempts to control proton conductivity have been achieved by using metal cations of different sizes in isostructural MOFs,17 and it was found that increasing the size of the metal cation enhanced the proton conductivity of the resultant MOF.18 However, in contrast, design and modification of non-porous MOFs for improved proton conductivity have been very rarely explored, not least because of the limited strategy for structural modification. The mechanism of proton dynamics within MOFs is often hypothesised by using a combination of Xray crystallographic information and the activation energy obtained from variable temperature impedance measurements. This provides a general characterisation of the proton dynamics being driven by either Grötthus or Vehicular mechanisms but does not provide a concrete explanation which can be used to further develop materials with enhanced proton conductivity. A range of experimental and computational techniques have been used to gain an understanding of the rotational and translational motion of guest molecules confined within MOFs;17-19 however, similar investigations into proton dynamics are much rarer.22-23 Quasielastic neutron scattering (QENS) has been applied recently to study the proton dynamics in a new phosphonate-based material allowing for the first time, confirmation of the model of “free diffusion inside a sphere” as a mechanism for proton conduction within MOFs.22 Another recent study also used QENS analysis in combination with molecular dynamics simulations in order to gain a greater understanding of the role that hydrogen bonded water networks have on proton transfer in UiO-66(Zr)-(CO2H)2.23 Solid state 2H NMR spectroscopy has been used to confirm the rapid exchange of 2H in the layered structure of PCMOF-3, although it was challenging to confirm in this case whether 2H solely or the water molecules were mobile.24 On the other hand, 2H NMR spectroscopy has been successfully applied recently to probe the molecular mobility of both water and charge-carrier protonic species in the solid heteropolyacid (HPA) hydrates, which are widely known for their proton conducting properties.25 Herein, we discuss the correlation between framework proton density and the corresponding proton conductivity in non-porous MOF systems and the application of three approaches to increase the concentration of protons in a given MOF without significantly modifying its structure. A new family of lanthanide phosphonate complexes, MFM-550(M) of formula [M(HL1)] (H4L1 = biphenyl-4,4'-diphosphonic acid; M = La, Ce, Nd, Sm, Gd, Ho) were synthesised, characterised and studied by impedance measurements. Our first methodology to increase the proton conductivity in MFM-550(M) involved the 3 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

synthesis of MFM-550(Ba), [Ba(H2L1)], which is isostructural to MFM-550(M) but has an excess of protons derived from the replacement of a +3 metal with a +2 metal of similar size. We are thus tuning the proton conductivity of a MOF by using two metal ions which are of a similar size but exhibit different oxidation states. The second approach was to use a shorter analogue of the organic ligand in order to decrease the distance between the metal chains in which the mobile protons are localised, and consequently increase the number of protons in a given volume, leading to higher framework proton density. Thus, a secondary family of lanthanide complexes, MFM-555(M) with formula [M(HL2)] (H4L2 = benzene-1,4-diphosphonic acid; M = La, Ce, Nd, Sm, Gd, Ho) were synthesised and characterised. The third method was based upon loading H2SO4 on the external surface of these non-porous MOFs to improve the surface proton concentration. Impedance spectroscopic measurements of all the synthesised materials were recorded at different temperatures and relative humidity (% RH), and up to four orders of magnitude enhancement on proton conductivity was observed. In order to gain a deeper mechanistic understanding of the proton dynamics in these materials, we conducted inelastic neutron scattering (INS), QENS, and ab initio molecular dynamics (MD) experiments on MFM-555(Ho). We also report a comprehensive in situ solid state 2H NMR spectroscopic study of MFM-555(Ho) at various levels of hydration with D2O. Overall, this study provides fundamental insights at the molecular level of the complex proton dynamics within this MOF and highlights the importance of surface hydration of MOF crystallites, reflected by the dependence of humidity on the observed proton conductivity.

EXPERIMENTAL SECTION Materials and characterisations. Starting materials were purchased from Acros Organics and from Sigma Aldrich and used without further purification. Elemental analyses were performed on a CE-440 Elemental Analyser. Magic angle spinning (MAS) 1H and {1H}

31

P NMR spectra were obtained on a Bruker Avance III spectrometer in the Centre of

Biomolecular Science at University of Nottingham. Impedance analyses were performed on a Solartron SI1260 Impedance analyser over a frequency range of 0.1 Hz to 1 MHz at an amplitude of 100 mV and 0 mV DC rest voltage. High resolution in situ variable temperature (VT) synchrotron powder X-ray diffraction (PXRD) data were collected on Beamline I11 at Diamond Light Source using a multi-analysing crystal detector and monochromated radiation (λ = 0.8269 Å). Neutron vibrational spectra of MFM-555(Ho) were measured using the VISION (BL-16B) instrument at the Spallation Neutron Source, Oak Ridge National Laboratory. The dynamics of the protons were probed using the neutron spectrometer IRIS at the ISIS Neutron Facility. 2H NMR experiments were performed at 61.424 MHz on a Bruker Avance-400 spectrometer using a high power probe with a 5 mm horizontal solenoid coil at Boreskov Institute of Catalysis, Novosibirsk. Full experimental details of all techniques used can be found in the Supplementary Information (SI).

Synthesis of [La(C12H8P2O6H)], MFM-550(La). H4L1 (0.0450 g, 0.14 mmol) and LaCl3·7H2O (0.0531 g, 0.14 mmol) were suspended in water (6 mL). A 2.6 M solution of HNO3 in H2O (1.5 mL), and a 0.3 M solution of piperazine in H2O (1.5 mL) were added to the mixture in a PARR pressure vessel. The vessel was sealed, heated in an oven at 210 °C for 72 h and cooled to room temperature. The crystalline white powder obtained was washed with DMF and water to remove non-reacted ligand, and then with

4 ACS Paragon Plus Environment

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

acetone. Yield: 0.05 g (0.11 mmol), 79 %. Elemental analysis (%): anal. calc. for C12H9O6P2La: C 32.03, H 2.02, N 0.00; found: C 32.12, H 1.99, N 0.00. FTIR: ν (cm-1) = 3070 (w), 2358 (w), 1917 (w), 1604 (w), 1541 (w), 1482 (w), 1312 (m), 1138 (s), 1113 (s), 1055 (m), 968 (s), 807 (s), 720 (s), 703 (s), 645 (w), 629 (w). 1H MAS NMR (600 MHz) δ ppm: 7.40 (s, 8H, Ar-H), 18.56 (w, 1H, O-H). 31P{1H} MAS NMR (600 MHz) δ ppm: 10.67.

Synthesis of [M(C12H8P2O6H)], MFM-550(M) and [M(C6H4P2O6H)], MFM-555(M). The same synthetic conditions described for MFM-550(La) were used to give MFM-550(M) as crystalline powder. For MFM-555(M) (M = La, Ce, Nd, Sm, Gd, Ho), H4L2 (0.14 mmol) was substituted for H4L1. LaCl3·7H2O, CeCl3·7H2O, Nd(NO3)3·6H2O, Sm(NO3)3·6H2O, GdCl3·6H2O and Ho(CH3CO2)3·6H2O were used as inorganic metal salts for the synthesis of the corresponding MOF (see SI for details).

RESULTS AND DISCUSSION Synthesis, structures, solid state NMR spectroscopy and thermal stabilities. The family of isostructural MOFs, MFM-550(M) (M = La, Ce, Nd, Sm, Gd, Ho), were synthesised by mixing the ligand H4L1 and the relevant metal salt (1:1 ratio) in H2O in the presence of HNO3 and piperazine. The reaction mixture was sealed in a hydrothermal reactor and heated at 210 oC. MFM-550(M) was formed after 3 days as a highly crystalline powder. Except for alterations in the ligand and metal salt used, the same synthetic conditions were employed to synthesise MFM-550(Ba) and MFM-555(M) (M = La, Ce, Nd, Sm, Gd, Ho). All complexes were isolated as microcrystalline powders in pure phase, except for MFM-550(Ho) which contains a very small amount of unknown impurities [see Figure S1 for additional details on the structural analysis of MFM-550(Ho)]. Their crystal structures were solved from high-resolution synchrotron PXRD data using ab initio methods and Rietveld refinements (Figure S1-S3). These compounds all crystallise in the monoclinic system and are isostructural with small changes tuned by the presence of different Ln(III) cations in the framework (Figure 1). The ionic radius of the lanthanide metals progressively decreases from La(III) to Gd(III). Consequently, small contractions in the unit cell parameters and shorter Ln-Ln distances are observed on moving from [La(HLn)] to [Gd(HLn)] (Table S1). Each phosphonate group consists of three crystallographically independent oxygen atoms O(1), O(2) and O(3) (Figure 1c), and the coordination geometry at each Ln(III) centre is distorted bicapped trigonal-prismatic with a {MO8} node (Figure 1d). Each Ln(III) centre is coordinated to eight oxygens from six phosphonate groups, two pairs of O(2) centres bridging two adjacent Ln, two O(1) and two O(3). The bond lengths discussed below refer to MFM-550(Ce) (Table S1). O(1) and O(2) are coordinated by chelating modes to Ce(III) with bond distances of 2.5218(53) Å for Ce'-O(1) and 2.8511(60) Å for Ce'-O(2). O(2) also acts as a bridge to the adjacent Ce(III) cation [Ce''O(2), 2.2602(46) Å], and O(3) is bonded to a third metal cation, [Ce'''-O(3), 2.4249(62) Å]. The distance between adjacent Ce(III) cations is between 4.166 Å and 4.254 Å. Extension of this trimetallic cluster arrangement along the c- and a-axes results in the formation of Ln-phosphonate 2D-sheets which are pillared by the organic ligands along the b-axis to afford a three dimensional framework (Figure 1e). The main difference between MFM-550(M) and MFM-555(M) is the distance between the adjacent metal chains (where the free protons are located, see below) and hence, a variation in the number of protons in a given volume. In each case of MFM-550(M) and MFM-555(M), one free proton for each ligand (PO-H) 5 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

is delocalised in the framework, while in MFM-550(Ba) there are two free protons per ligand, as predicted by the charge balance in the formula. Detection of protons from X-ray experiments can often be inconclusive, and in order to quantify the protons and investigate their environment, solid-state 1H and {1H} 31

P MAS NMR spectra were recorded for H4L1, MFM-550(La) and MFM-550(Ba) at room temperature

(Figure 2). Comparing the 1H spectra, the intensity ratio of aromatic protons to free –OH groups varies as expected confirming that substituting La(III) with Ba(II) doubles the number of free protons in the framework in order to achieve charge balance. The presence of O-H groups close to the P centre in MFM550(La) and MFM-550(Ba) is supported by {1H}-31P dipolar correlation spectra which show two cross peaks corresponding to contacts between both proton sites and the phosphonate P; the more intense cross peak is assigned to the short-range P-O-H contact, while the less intense cross peak corresponds to H-C-C-P contact. These MOFs show impressive thermal stability to above 500 °C as confirmed by TGA and VT-PXRD measurements (Figure S8-S12). The absence of weight loss at low temperatures, characteristic of loss of solvent molecules trapped inside the pores, confirms MFM-550(M) and MFM-555(M) to be very compact frameworks, in agreement with their crystal structures. The slightly higher thermal stability (by about 50 oC) of MFM-555(M) compared with MFM-550(M) is likely due to the more compact structure of MFM-555(M) due to the use of a shorter ligand.

Proton conductivity study. Impedance measurements on all complexes were recorded at 99%, 70% and 45% RH at 20 oC to investigate any correlation between the proton conductivity and the relative humidity (Table S5). At 99% RH, the Nyquist plots show a semi-circle in the high frequency region and a tail at low frequencies, representing the bulk resistance and blocking of protons at the electrodes, respectively, consistent with ion migration. At reduced humidity, the resistivity of all samples increases and in the frequency range investigated the tail in the Nyquist plot is not present. Reduction of resistivity with increasing relative humidity is a well-known response observed in PEMs and MOFs when their proton conductivity is water-mediated,26-31 the proton conductivity of water-mediated materials relying on both water molecules and hydrogen-bonding interactions between the water molecules and the protons within the framework.31 It is noteworthy that the proton conductivities measured at 99% RH and 20 oC for the series MFM-550(M) and MFM-555(M) are very similar, as expected since the materials are isostructural and only small changes in the unit cell parameters are observed with varying metal ions. The proton conductivity of MFM-550(M) at 20 oC and 99% RH is ~10-6 S cm-1, which is lower than many reported MOFs under similar conditions.32-35 We sought to improve the proton conductivities of these materials with a focus on the availability of free protons and their location in the structure. Firstly, we argued that by substituting Ln(III) with Ba(II) ions would require the loss of positive charge to be made up by additional protons in the framework. The case discussed here goes beyond simply doping the Ln(III) framework with Ba(II) defects, rather the successful synthesis of the Ba-form of the framework has been achieved. Interestingly, an increase of proton conductivity of one order of magnitude was measured for MFM-550(Ba) (~10-5 S cm-1) at 99% RH and 20

6 ACS Paragon Plus Environment

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

o

C (Figure 3). This demonstrates the success of strategy and confirms that the proton conductivity of a MOF

can be increased by the use of a metal with lower oxidation state without modifying the overall structure. The second methodology targeted the increase of the number of protons in a fixed cell volume. This was accomplished with the use of benzene-1,4-diphosphonic acid, H4L2, a shortened version of the H4L1 linker. This decreases the distance between the metal phosphonate chains and thus increases the concentration of protons in a fixed volume, 1.4 times ongoing from MFM-550(M) to MFM-555(M). On going from MFM-550(M) to MFM-555(M) at 99% RH and 20 oC, the proton conductivity shows a remarkable increase of two orders of magnitude from 10-6 S cm-1 to 10-4 S cm-1 (Figure 3). Despite increasing the proton density to a lesser extent than seen for MFM-550(Ba), this method has a greater impact on the increase in proton conductivity, highlighting the importance of the accessibility of the mobile protons, particularly in non-porous MOFs. This result shows for the first time that increasing the density of active protons in non-porous MOF structures positively impacts on their proton conducting property. The third methodology involved doping MFM-555(Ho) with 2M H2SO4 to increase the concentration of surface H+. The introduction of proton carriers as guest molecules into the cavities of porous MOFs is widely reported.8-14 However, the surface effect of such doping in non-porous MOFs is yet to be explored. The stability and full retention of the crystal structure of MFM-555(Ho) on treatment with H2SO4 were verified by PXRD (Figure S16). Significantly, treatment of MFM-555(Ho) with H2SO4 results in four orders of magnitude increase in proton conductivity (from 10-6 S cm-1 to 10-2 S cm-1) compared with MFM-550(M), and two orders of magnitude increase of conductivity (from 10-4 to 10-2 S cm-1) compared to the MFM555(Ho) at 99% RH and 20 oC (Figure 3a). Interestingly, the proton conductivity of H2SO4 treated MFM555(Ho) is comparable to the best-behaving proton-conductive MOFs8,33,34 to date and, to the best of our knowledge, is the highest performing non-porous MOF under the same conditions (Table S7). These results are accompanied by an additional of 0.31 mmol g-1 of acidic sites residing on the MOF surface compared to the pristine MFM-555(Ho) (see SI for details of pH and NH3-temperature programmed desorption experiments), indicating that the proton dynamics also plays a key role in the resultant conductivity.22,36,37 Variable temperature impedance measurements at 99% RH revealed that the activation energy was reduced from 0.32 eV to 0.12 eV on addition of H2SO4 to MFM-555(Ho), suggesting an increase in the efficiency of proton transfer across the surface of acid-treated MFM-555(Ho) (Figures S17 and S18). A general linear relationship between the framework proton density and proton conductivity is observed for all non-porous MOFs and 2D coordination polymers reported to date (Figure 3b), and, more interestingly, compared to literature examples, a steeper slope (Figure 3b) is observed for MOFs reported in this study, demonstrating the high efficiency of the design strategies developed here.

Inelastic neutron scattering (INS), quasi elastic neutron scattering (QENS), and ab initio molecular dynamics (MD) studies of proton dynamics in MFM-555(Ho). The MFM-555(M) materials show similar conductivities, and the diamagnetic Ho-analogue was selected for further detailed investigations. The DFTcalculated INS spectrum derived from the crystal structure of MFM-555(Ho) shows an excellent agreement with the experimental spectrum, allowing the assignment of all vibrational modes, particularly for the 7 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

phosphonate moiety (Figure S25, Table S9). The O-P-O-H wagging and P-O-H stretching modes were assigned to peaks at 203 and 208 cm-1, respectively, while P-O stretching modes were assigned at 878 and 979 cm-1. The only discrepancy between experimental and simulated spectra is the observation of a notable INS peak at ~21.6 cm-1 (or 2.7 meV) in the experimental spectrum, which was attributed to a tunnelling mode and this peak disappears with increasing temperature (Figures S22). The nature of this tunnelling is beyond the scope of this report and in-depth discussion will be reported separately. MD simulations were conducted to observe the proton dynamics within MFM-555(Ho) at 300 and 800 K, and similar results were obtained at both temperatures (Figure S23). Figure 4a shows the MD trajectories of the H-bonded -PO2(OH) groups. Proton hopping between -PO2(OH) groups via -PO3 rotation is clearly observed, suggesting a potential proton transfer pathway along the Ho-PO2(OH) moieties in MFM-555(Ho) via a Grotthus-type mechanism. Figure 4b shows the mean square displacement of all the atoms and highlights the notable diffusion of the H atoms within -PO2(OH) groups through the MOF structure. To further understand the influence of humidity on proton dynamics in MFM-555(Ho), QENS experiments were conducted between 248 and 353 K for both dehydrated and hydrated MFM-555(Ho). For the dehydrated sample, despite a loss of elastic intensity, no QENS broadening was observed with increasing temperature (Figure S26), suggesting that any motions, if occurring, were too slow to be captured within the resolution of the spectrometer. As with the INS experiment, evidence was found for the presence of a tunnelling mode at low temperature, manifested by extra intensity on the shoulder of the elastic peak at ~0.4 meV (Figure S27). For the hydrated sample, broadening of the elastic peak was observed between 248 and 353 K (Figure S28). Analysis of the half-width of the half maximum (HWHM, Γ) of the QENS spectra recorded as a function of Q2 was successfully fitted with the Hall-Ross diffusion model [equation 1]38 (Figure 5a). ℏ

Γ =  1 − exp 

  〈  〉  

(1)

where  is the average residence time and 〈  〉 is the mean square jump length. The only exception to this model was the fit at 248 K, which showed almost no broadening. Extraction of  and 〈  〉 (Table S10 and

S11) allowed calculation of the diffusion coefficient, D at all temperatures (Table S11) [equation 2].  =

〈  〉

(2)



The mean jump-length,  increases from 1 Å at low temperature to 3 Å at elevated temperature,

suggesting that QENS is not subject to the dynamics of a single hydrogen bonded atom. Instead, at low temperature the gyration of a single O-H is observed, arising from water molecules adsorbed on the MOF surface. At elevated temperature, proton-transfer occurs between the surface P-OH and adsorbed water to form H3O+. The increased population of H3O+ with temperature leads to the increase in the observed gyration radii. The residence time, , decreases as a function of temperature giving an activation energy of ~3 kJ mol-1

(Figure S29) consistent with the observation of fast local gyrations with a small energy barrier. The diffusion coefficient, D, is not constant over the temperature range measured due to its dependence on jump-length, showing a transition temperature of ~302 K (Figure 5b). Thus, the results of MD are consistent with the proton migration pathway derived from the structure of MFM-555(Ho), and the neutron scattering study 8 ACS Paragon Plus Environment

Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

builds on this by considering the influence that hydration of accessible phosphonate groups on the surface of the MOF has on the overall proton dynamics. Solid state 2H NMR study of proton dynamics in d-MFM-555(Ho). In order to investigate further the proton dynamics and its correlation with the conductivity, solid state 2H NMR spectroscopy was employed to study the deuterated material, d-MFM-555(Ho), in which the active O-H groups were exchanged with O-D. Three samples were sealed at varying hydration levels with D2O [dehydrated (I), partially hydrated (II) and saturated (III)] (Figure S31). It is worth noting that the average pore aperture for the framework channels is ~1 Å, and is too small to host any guest water molecules. This is consistent with the lack of apparent water adsorption by MFM-555(Ho) (Figure S32). Thus, the primary hydrated proton conduction pathway resides on the external surface of the MOF crystallites. The 2H NMR spectra of dehydrated d-MFM-555(Ho) was measured between 123 and 403 K (Figure 6a), in this region the line shape is given by a very broad, Gaussian-type pattern. Even at 123 K, a Pakepowder pattern, which is typical for static deuterons, was not observed. This observation is distinct to reported studies on zeolites and anhydrous tungsten HPAs, where static patterns were commonly seen,34 and suggests that even at 123 K, the O-D groups in d-MFM-555(Ho) are partially mobile. A very broad Gaussian-type pattern indicates that these motions are heavily restricted and the O-D groups are localized in a disordered manner with various orientations and mobility.37 Moreover, the line-width of Gaussian patterns (∆ν1/2 > 150 kHz) suggests that the motions responsible for the line shape are slow (10-4 –10-5 s) and that a tunneling mechanism is responsible for the motion,36,37 entirely consistent with neutron scattering results. With increasing temperature, the effective width of the pattern decreases, indicating that the motion becomes faster and deuterons at different positions are readily exchanging between accessible sites. At 123 K, the Gaussian pattern has ∆ν1/2 ~210 kHz, which corresponds to a quadrupolar coupling constant Q0 ~250 kHz. This is comparable with values reported for O-D groups in zeolites and HPAs (234 and 210 kHz, respectively).35,39 The temperature dependence of the Gaussian line-width ∆ν1/2 allows the estimation of the apparent transverse relaxation time T2* ~1/(π∆ν1/2), which enables the calculation of the activation barrier (~0.25 kJ mol-1) (Figure S35). Interestingly, such a low energy barrier suggests migration of the deuterons between neighboring sites via a tunneling mechanism. For comparison, in dehydrated tungsten HPA, the surface migration of protons is characterized by an activation barrier of ~70 kJ mol-1, corresponding to the proton hopping between neighboring oxygens sites.36 The 2H NMR spectra of D2O saturated d-MFM-555(Ho) show that at 123-480 K only one type of signal was observed indicating that all accessible deuterons exhibit similar mobility (Figures 6e). Moreover, the observed temperature-dependence for line shape is fully reversible between 123 and 480 K. Below 273 K, the line shape shows a Gaussian-type pattern similar to the dehydrated sample (denoted as State 1). At 273 K and above, the line shape changes to a Lorentzian-type isotropic pattern (denoted as State 2) with an effective width ~2.6 times narrower than that of State 1. The drastic decrease of ∆ν1/2 indicates that the deuterons in State 2 are considerably more mobile than that in State 1 (Figure 6f). Moreover, the Lorentziantype pattern indicates that above 273 K, the proton dynamics is sufficiently efficient to homogenously 9 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

average the O-D orientations in space and time to provide an isotropic dynamic pathway.25,39 Below 273K in State 1, similar to the dehydrated sample, the Gaussian patterns are characterized by a very large ∆ν1/2 with no trace of typical patterns for immobile water molecules present. This indicates that the water is strongly confined in State 1 but not completely immobilised.37 Such a situation is distinct to the HPAs, where surface O-D groups, hydrogen bonded D2O and hydronium ion (D3O+) were all observable at low temperatures, even under conditions when all surface hydroxyl groups are bound to a water molecule.24 The analysis for ∆ν1/2 temperature dependence in State 1 shows a very small activation barrier, EtH1 ~0.5 kJ mol-1 (Figure S35), consistent with the tunneling process. In contrast, in State 2, the barrier is increased to ~7 kJ mol-1, suggesting that this isotropic tumbling is an Arrhenius activation process, and the rate constant is characterized by a relatively small pre-exponential factor of ~106 s-1, compared with that (~1013 s-1) for the anhydrous HPAs.36 The latter is characterized by a motion based upon individual hops of a single deuteron from site to site,24 whereas in d-MFM-555(Ho), a proton migration via formationdecomposition of [MOF·(D2O)n] complex is responsible to the motion in State 2. The dynamic behavior of the partially hydrated d-MFM-555(Ho) is similar to the saturated case (Figure S33a), with an important difference: above 273 K, the pattern is composed of signals seen in State 1 and State 2 simultaneously. This indicates that the mobile state for proton species on the surface is strictly dependent on the water coverage as the exchange between State 1 and State 2, if presents, must be very slow; additional discussion of the partially hydrated sample is given in supporting information. Thus, this 2H NMR study confirms that the mobile proton carrier observed in State 2 originates from the interaction between the O-D groups on MOF surface and adsorbed water species (e.g., D2O, D3O+). Only a fully interconnected coverage of the MOF surface affords a continuous pathway for the proton hopping. Below 273 K, the water is strongly bound to the surface and largely immobile. While above 273 K, mobile species (e.g., H3O+) can be observed; in contrast, this transition in mobility occurs at 100 K for HPA hydrates.24 This result directly explains the observed high humidity dependence on conductivity of the nonporous MFM-555(Ho): below the saturation conditions, water is only partially adsorbed, decreasing the effective connectivity of the hydrogen bonding network assembled by the mobile species. Therefore, the mobile species migrate over the surface very slowly (kaH2 ~105 s-1 at 303 K) and the exchange with nonhydrated surface hydroxyl groups is even slower (kex < 103 s-1 across the whole range). In contrast, under high humidity conditions, all surface protons are able to form mobile species (e.g., H3O+), increasing the charge carrier density significantly and thus inducing the proton conductivity. Analysis of 2H NMR spin relaxation for fully hydrated d-MFM-555(Ho). While being very broad, the lineshapes of all 2H NMR spectra of the hydrated d-MFM-555(Ho) are featureless (Figures 6a). Thus, the only way to uncover details of their mobility is to measure the relaxation time, particularly for the spin-lattice T1 and the apparent T2* relaxations. The spin relaxation of the fully hydrated sample is studied because its behaviour is fully reversible and the activation energies obtained are directly relevant to the proton conduction.

10 ACS Paragon Plus Environment

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

The temperature dependence of experimental T1 and T2* relaxation times in D2O-saturated d-MFM555(Ho) shows that that even in State 1, the deuterons are involved in very fast internal motions and the minimum T1 is found at 170 K (Figure 6f). These are fast but very restricted librations in a cone of the O-D bond with a semi-cone angle θlib ~15○, an activation barrier Elib of 15 kJ mol-1 and a pre-exponential factor klib0 of 0.4x1012 s-1. Such a barrier is typical for the cleavage of a single hydrogen bond. The presence of such a fast motion is not contradictory to the broad line-shape due to the small angle of the libration. At 250 K, another sharp minimum was observed with an activation barrier Eex of 45 kJ mol-1. This value is consistent with that obtained from variable temperature impedance measurements on MFM-555 (Figure S13), suggesting that this motion is correlated to the proton dynamics responsible for conductivity within these MOFs. Above 300 K, the T1 curve shows exactly the same slope as T2* with Eiso of 7 kJ mol-1, indicating that the same motion is being observed in both cases. This result suggests that once formed, the interconnected mobile species exhibits a slow isotropic re-orientation over the MOF surface. The transition between two populations shows excellent agreements with the results observed by QENS. These observations can be further rationalised: the surface O-D groups interact with the immobilised D2O molecules to form a new species, e.g., D3O+ ions. Such a process can be described by a general scheme: A + B ↔ C, where A (O-D) and B (D2O) are only involved in local librations and represent State 1. C (D3O+) represents the more mobile State 2, which, in addition to local motions, also shows an isotropic reorientation. This general scheme allows a satisfactory fit of the observed relaxation data and allows the determination of all kinetic parameters of the motions (Figure 6f). Notably, it allows derivation of the D3O+ complex formation and decomposition rate constants: forward process, k12 (E12 = 45 kJ mol-1; k120 = 0.7x1012 s-1) and backward process, k21 (E21 = 40 kJ mol-1; k210 = 0.4x107 s-1), showing that balance is clearly shifted towards the formation of D3O+. The isotropic reorientation of the latter is governed by Eiso of 7 kJ mol-1 and kiso0 of 0.5x106 s-1. This confirms that although the D3O+ isotropic diffusion is slow, the rate-determining step in the proton hopping process is unambiguously the formation of the complex, where the humidity plays a critical role.

CONCLUSIONS This work focuses on non-porous MOFs as proton conductors. We have systematically investigated the correlation between framework proton density and proton conductivity, and in non-porous MOFs increasing framework proton density is a highly efficient strategy to enhance proton conductivity. This has been successfully demonstrated in a family of lanthanide materials MFM-550(M) (M = La, Ce, Nd, Sm, Gd, Ho) by (i) replacing the metal ion to one with a lower oxidation state, (ii) reducing the length of the organic ligand, and (iii) introducing additional acidic protons on MOF surface. Impedance measurements suggest that all three strategies are successful with an enhancement in proton conductivity for the modified materials of up to four orders of magnitude. This paves the way to the design of new non-porous MOFs with improved conductivity. Subsequently, MFM-555(Ho) was studied by neutron scattering, MD modelling and solid state 2

H NMR spectroscopy. MD simulations identified a potential route for proton transfer within the MOF based

on shortening and lengthening of the P-O-H bonds with proton trajectories showing diffusion along the 11 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

phosphonate backbone of the MOF structure. QENS studies confirm that diffusion in the hydrated material occurs via a Hall-Ross mechanism, with two separate regimes obtained from the temperature dependence of the mean square jump length and diffusion coefficient. The low temperature regime is associated with water adsorbed to the MOF surface, while the elevated temperature regime is derived from the gyration of an increasing population of H3O+. The temperature dependence of the retention time is consistent with fast gyrations that have a low energy barrier (~3 kJ mol-1). 2H NMR spectroscopic studies suggest that D2O interacts with P-OH moieties on the MOF surface, and this was supported by water vapour sorption. Line shape and spin relaxation analysis of samples measured at varying hydration levels highlighted the critical importance for the MOF to be saturated by water in order to form an extensive hydrogen bonding network, in excellent agreement with the variable humidity impedance results. In the water-saturated sample, proton dynamics appear to be facilitated by a general scheme in which the water is adsorbed on the MOF surface, leading to formation and slow isotropic reorientation of the resultant hydronium ions, which is the fundamental step for the entire proton hopping process, in excellent agreement with the QENS study.

REFERENCES (1) Sadakiyo, M.; Yamada, T.; Kitagawa, H. High proton conductivity of one-dimensional ferrous oxalate dihydrate. J. Am. Chem. Soc. 2009, 131, 9906–9907. (2) Bao, S.; Wu, Y.; Li, N.; Zheng, L. Proton conductivities manipulated by the counter-anions in 2D Co-Ca coordination frameworks. Eur. J. Inorg. Chem. 2016, 27, 4476–4482. (3) You, Y.; Xue, C.; Tian, Z.; Liu, S.; Ren, X. Three orders of magnitude enhancement of proton conductivity of porous coordination polymers by incorporating ion-pairs into a framework. Dalton Trans. 2016, 45, 7893–7899. (4) Maity, D. K.; Otake, K.; Ghosh, S.; Kitagawa, H.; Ghoshal, D. Sulfonic group functionalized mixed ligand coordination polymers: synthesis, characterization, water sorption, and proton conduction studies. Inorg. Chem. 2017, 56, 1581−1590. (5) Yang, F.; Huang, H.; Wang, X.; Li, F.; Gong, Y.; Zhong, C.; Li, J. Proton conductivities in functionalized UiO-66: tuned properties, thermogravimetry mass, and molecular simulation analyses. Cryst. Growth Des. 2015, 15, 5827−5833. (6) Shigematsu, A.; Yamada, T.; Kitagawa, H. Wide control of proton conductivity in porous coordination polymers. J. Am. Chem. Soc. 2011, 133, 2034-2036. (7) Zhou, L.; Deng, W.; Wang, L.; Xu, G.; Yin, S.; Liu, Q. Lanthanide-potassium biphenyl-3,3’-disulfonyl4,4’-dicarboxylate frameworks: gas sorption, proton conductivity, and luminescent sensing of metal ions. Inorg. Chem. 2016, 55, 6271−6277. (8) Sadakiyo, M.; Yamada, T.; Kitagawa, H. Proton conductivity control by ion substitution in a highly proton-conductive metal–organic framework. J. Am. Chem. Soc. 2014, 136, 13166–13169. (9) Liu, J.; Zou, X.; Liu, C.; Cai, K. ; Zhao, N.; Zheng, W.; Zhu, G. Ionothermal synthesis and protonconductive properties of NH2-MIL-53 MOF nanomaterials. CrystEngComm. 2016, 18, 525-528. 12 ACS Paragon Plus Environment

Page 13 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

(10) Bureekaew, S.; Horike, S.; Higuchi, M.; Mizuno, M.; Kawamura, T.; Tanaka, D.; Yanai, N.; Kitagawa, S. One-dimensional imidazole aggregate in aluminium porous coordination polymers with high proton conductivity. Nat. Mater. 2009, 8, 831–836. (11) Umeyama, D.; Horike, S.; Inukai, M.; Hijikata, Y.; Kitagawa, S. Confinement of mobile histamine in coordination nanochannels for fast proton transfer. Angew. Chem. Int. Ed. 2011, 50, 11706–11709. (12) Hurd, J.; Vaidhyanathan, R.; Thangadurai, V.; Ratcliffe, C. I.; Moudrakovski, I. L.; Shimizu, G. K. H. Anhydrous proton conduction at 150 °C in a crystalline metal-organic framework. Nat. Chem. 2009, 1, 705– 710. (13) Ye, Y.; Wu, X.; Yao, Z.; Wu, L.; Cai, Z.; Wang, L.; Ma, X.; Chen, Q. H.; Zhang, Z.; Xiang, S. Metal– organic frameworks with a large breathing effect to host hydroxyl compounds for high anhydrous proton conductivity over a wide temperature range from subzero to 125 °C. J. Mater. Chem. A 2016, 4, 4062-4070. (14) Phang, W. J.; Lee, W. R.; Yoo, K.; Ryu, D. W.; Kim, B.; Hong, C. S. pH-dependent proton conducting behavior in a metal-organic framework material. Angew. Chem. Int. Ed. 2014, 53, 8383–8387. (15) Taylor, J. M.; Komatsu, T.; Dekura, S.; Otsubo, K.; Takata, M.; Kitagawa, H.. The role of a three dimensionally ordered defect sublattice on the acidity of a sulfonated metal–organic framework. J. Am. Chem. Soc. 2015, 137, 11498-11506. (16) Inukai, M.; Horike, S.; Itakura, T.; Shinozaki, R.; Ogiwara, N.; Umeyama, D.; Nagarkar, S.; Nishiyama, Y.; Malon, M.; Hayashi, A.; Ohhara, T.; Kiyanagi, R.; Kitagawa, S. Encapsulating mobile proton carriers into structural defects in coordination polymer crystals: high anhydrous proton conduction and fuel cell application. J. Am. Chem. Soc. 2016, 138, 8505-8511. (17) Su, X.; Yao, Z.; Ye, Y.; Zeng, H.; Xu, G.; Wu, L.; Ma, X.; Chen, Q.-H.; Wang, L.; Zhang, Z.; Xiang, S. 40-Fold enhanced intrinsic proton conductivity in coordination polymers with the same proton-conducting pathway by tuning metal cation nodes. Inorg. Chem. 2016, 55, 983–986. (18) Wong, N. E.; Ramaswamy, P.; Lee, A. S.; Gelfand, B. S.; Bladek, K. J.; Taylor, J. M.; Spasyuk, D. M.; Shimizu, G. K. H. Tuning intrinsic and extrinsic proton conduction in metal–organic frameworks by the lanthanide contraction. J. Am. Chem. Soc. 2017, 139, 14676−14683. (19) Kolokolov, D. I.; Jobic, H.; Rives, S.; Yot, P. G.; Ollivier, J.; Trens, P.; Stepanov, A. G.; Maurin, G. Diffusion of benzene in the breathing metal–organic framework MIL-53(Cr): A joint experimental– computational investigation. J. Phys. Chem. C. 2015, 119, 8217−8225. (20) Kolokolov, D. I.; Maryasov, A. G.; Ollivier, J.; Freude, D.; Haase, J.; Stepanov, A. G.; Jobic, H. Uncovering the rotation and translational mobility of benzene confined in UiO-66 (Zr) metal–organic framework by the 2H NMR–QENS experimental toolbox. J. Phys. Chem. C. 2017, 121, 2844−2857. (21) Jiang, X.; Duan, H.; Khan, S. I.; Garcia-Garibay, M. A. Diffusion-controlled rotation of triptycene in a metal-organic framework (MOF) sheds light on the viscosity of MOF-confined solvent. ACS Cent. Sci. 2016, 2, 608−613. (22) Pili, S.; Argent, S. P.; Morris, C. G.; Rought, P.; García-Sakai, V.; Silverwood, I. P.; Easun, T. L.; Li, M.; Warren, M. R.; Murray, C. A.; Tang, C. C.; Yang, S. Schröder, M. Proton conduction in a phosphonate-

13 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

based metal-organic framework mediated by intrinsic "free diffusion inside a sphere". J. Am. Chem. Soc. 2016, 138, 6352−6355. (23) Borges, D. D.; Devautour-Vinot, S.; Jobic, H. ; Ollivier, J. ; Nouar, F. ; Semino, R.; Devic, T.; Serre, C.; Paesani, F.; Maurin, G. Proton transport in a highly conductive porous zirconium-based metal-organic framework: molecular insight. Angew. Chem. Int. Ed. 2015, 55, 3919-3924. (24) Taylor, J. M.; Mah, R. K.; Moudrakovski, I. L.; Ratcliffe, C. I.; Vaidhyanathan, R. Shimizu, G. K. H. Facile proton conduction via ordered water molecules in a phosphonate metal−organic framework. J. Am. Chem. Soc. 2010, 132, 14055–14057. (25) Kolokolov, D. I.; Kazantsev, M. S.; Luzgin, M. V.; Jobic, H.; Stepanov, A. G. Characterization and dynamics of the different protonic species in hydrated 12-tungstophosphoric acid studied by 2H NMR. J. Phys. Chem. C. 2014, 118, 30023−30033. (26) Ponomareva, V. G.; Kovalenko, K. A.; Chupakhin, A. P.; Dybtsev, D. N.; Shutova, E. S.; Fedin, V. P. Imparting high proton conductivity to a metal-organic framework material by controlled acid impregnation. J. Am. Chem. Soc. 2012, 134, 15640–15643. (27) Pardo, E.; Train, C.; Gontard, G.; Boubekeur, K.; Fabelo, O.; Liu, H.; Dkhil, B.; Lloret, F.; Nakagawa, K.; Tokoro, H.; Ohkoshi S. and Verdaguer, M. High proton conduction in a chiral ferromagnetic metal–organic quartz-like framework. J. Am. Chem. Soc., 2011, 133, 15328–15331. (28) Dey, C.; Kundu, T.; Banerjee, R. Reversible phase transformation in proton conducting Strandberg-type POM based metal organic material. Chem. Commun. 2012, 48, 266-268. (29) Okawa, H.; Sadakiyo, M.; Yamada, T.; Maesato, M.; Ohba, M.; Kitagawa, H. Proton-conductive magnetic metal–organic frameworks, {NR3(CH2COOH)}[MaIIMbIII(ox)3]: effect of carboxyl residue upon proton conduction. J. Am. Chem. Soc. 2013, 135, 2256–2262. (30) Colodrero, R. M. P.; Olivera-Pastor, P.; Losilla, R. R.; Hernandez-Alonso, D.; Aranda, M. A. G.; LeonReina, L.; Rius, J.; Demadis, K. D.; Moreau, B.; Villemin, D.; Palomino, M.; Rey, F.; Cabeza, A. High proton conductivity in a flexible, cross-linked, ultramicroporous magnesium tetraphosphonate hybrid framework. Inorg. Chem. 2012, 51, 7689–7698. (31) Yoon, M.; Suh, K.; Natarajan, S.; Kim, K. Proton conduction in metal-organic frameworks and related modularly built porous solids. Angew. Chem. Int. Ed. 2013, 52, 2688–2700. (32) Ramaswamy, P.; Matsuda, R.; Kosaka, W.; Akiyama, G.; Jeon, H. J.; Kitagawa, S. Highly proton conductive nanoporous coordination polymers with sulfonic acid groups on the pore surface. Chem. Commun. 2014, 50, 1144-1146. (33) Nagarkar, S. S.; Unni, S. M.; Sharma, A.; Kurungot, S.; Ghosh, S. K. Two-in-one: inherent anhydrous and water-assisted high proton conduction in a 3D metal-organic framework. Angew. Chem. Int. Ed. 2014, 53, 2638-2642. (34) Bazaga-Garcia, M.; Colodrero, R. M. P.; Papadaki, M.; Garczrek, P.; Zon, J.; Olivera-Pastor, P.; Losilla, E. R.; León-Reina, L.; Aranda, M. A. G.; Choquesillo-Lazarte, D.; Demadis, D. K.; Cabeza, A. Guest molecule-responsive functional calcium phosphonate frameworks for tuned proton conductivity. J. Am.Chem. Soc. 2014, 136, 5731-5739. 14 ACS Paragon Plus Environment

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

(35) Sen, S.; Yamada, T.; Kitagawa, H.; Bharadwaj, P. K. 3D coordination polymer of Cd(II) with an imidazolium-based linker showing parallel polycatenation forming channels with aligned imidazolium groups. Cryst. Growth Des. 2014, 14, 1240-1244. (36) Kolokolov, D. I.; Kazantsev, M. S.; Luzgin, M. V.; Jobic, H.; Stepanov, A. G. Direct 2H NMR observation of the proton mobility of the acidic sites of anhydrous 12-tungstophosphoric acid. ChemPhysChem. 2013, 14, 1783 –1786. (37) Hassan, J. Analysis of 2H NMR spectra of water molecules on the surface of nano-silica material MCM-41: Deconvolution of the signal into a Lorentzian and a powder pattern line shapes. Physica B. 2012, 407, 179-183. (38) Hall, P. L. and Ross, D. K. Incoherent neutron scattering function for molecular diffusion in lamellar systems. Mol. Phys. 1978, 36, 1549-1554. (39) Ernst, H.; Freude, D.; Pfeifer, H.; Wolf I.; Multinuclear NMR studies of acid sites in zeolites. Stud. Surf. Sci. Catal. 1994, 84 381-385.

Acknowledgement We thank EPSRC (EP/I011870), ERC (AdG 742041), Royal Society and University of Manchester for funding. We are especially grateful to STFC and ISIS Neutron Facility for access to the Beamline IRIS, to Diamond Light Source for access to Beamline I11 and to ORNL for access to Beamline VISION. The computing resources were made available through the VirtuES and ICEMAN projects, funded by Laboratory Directed Research and Development program at ORNL. DIK and AGS acknowledge financial support within the framework of the budget project #АААА-А17-117041710084-2 of the Boreskov Institute of Catalysis.

Additional information Supplementary information is available in the online version of the paper. Correspondence and requests for materials should be addressed to S.Y. and M.S. CCDC-1827844-1827855 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

Competing financial interests The authors declare no competing financial interests.

15 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

Figures and Tables:

Figure 1. Views of structures of MFM-550(M) and MFM-555(M) obtained from high resolution synchrotron PXRD data. Views of the structure of MFM-550 (a) and MFM-555 (b) along the c-axis. (c) View of the coordination node of the phosphonate group of the ligand. (d) Perspective view of the coordination environment at Ln(III) centres. (e) View of the extended Ln-phosphonate 2D-sheet along the b-axis.

16 ACS Paragon Plus Environment

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Figure 2. Comparison of solid-state 1H NMR spectra of (a) H4L1, (b) MFM-550(Ba) and (c) MFM-550(La). Correlation spectrum of 31P to H in (d) MFM-550(Ba) and (e) MFM-550(La).

17 ACS Paragon Plus Environment

Chemistry of Materials

b)

0.1

0.01

Proton conductivity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 22

1E-3

1E-4

1E-5

MFM-550(Ln) MFM-550(Ba) MFM-555(Ln) Non-porous MOFs/CPs in literature

1E-6

1E-7 1

3

5

7

9

11

13

15

17

19

21

23

25

Framework proton density (H+/nm-3)

Figure 3. (a) Proton conductivity for MFM-550(M), MFM-555(M), MFM-550(Ba) and H2SO4@MFM555(Ho) measured at 99% RH and 20 oC. (b) Correlation between framework proton density and proton conductivity in non-porous MOFs or CPs (coordination polymers) in this work and literature (dotted lines are guides to the eye). Detailed data are summarised in Table S7.

18 ACS Paragon Plus Environment

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

a

b

Figure 4. (a) Trajectories of the H bonded PO2(OH) groups in MFM-555(Ho) at 800 K. For clarity the trajectories of other atoms are not shown, and there is no diffusion associated with those. (b) Mean square displacement of the atoms in MFM-555(Ho). Red: H bonded PO2(OH) groups; blue: other H atoms on the linker; other colours correspond to Ho, P and O. Only H bonded PO2(OH) groups are diffusing.

19 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

a

Page 20 of 22

b

Figure 5. (a) HWHM dependence with Q2 of hydrated MFM-555(Ho) and fitting with the Hall-Ross diffusion model;36 (b) temperature dependence of diffusion coefficients obtained from Hall-Ross fitting, showing two temperature regimes with a transition temperature of ~302 K.

20 ACS Paragon Plus Environment

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Figure 6. (a) Experimental 2H NMR spectra showing temperature dependence of the line-shape for dehydrated d-MFM-555(Ho). Simulated 2H NMR spectral line-shapes for typical cases: (b) a Gaussian pattern with ∆ν1/2 ~212 kHz; (c) a static unaveraged Pake-powder pattern with Q0 of 250 kHz, η ~ 0; and (d) its evolution for the typical case of 2-site flips of the O-D bond within the water molecule geometry. (e) Experimental 2H NMR spectra showing line-shape temperature dependence for fully hydrated d-MFM555(Ho). (f) Experimental temperature dependence of the T1 (brown ■) and T2* (blue ●) relaxation time of the fully hydrated d-MFM-555(Ho); the simulation is shown by solid lines.

21 ACS Paragon Plus Environment

Chemistry of Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 22

TOC

22 ACS Paragon Plus Environment