Lithium inclusion in indium metal-organic frameworks showing increased surface area and hydrogen adsorption

Investigation of counterion exchange in two anionic In-Metal-Organic Frameworks (In-MOFs) showed that partial replacement of disordered ammonium cations was achieved through the pre-synthetic addition of LiOH to the reaction mixture. This resulted in a surface area increase of over 1600% in {Li [In(1,3 − BDC)2]}n and enhancement of the H2 uptake of approximately 275% at 80 000 Pa at 77 K. This method resulted in frameworks with permanent lithium content after repeated solvent exchange as confirmed by inductively coupled plasma mass spectrometry. Lithium counterion replacement appears to increase porosity after activation through replacement of bulkier, softer counterions and demonstrates tuning of pore size and properties in MOFs.

Metal-Organic Frameworks (MOFs) are an emerging category of porous materials that consist of metal-containing units and organic linkers. 1,2They are promising materials for the storage of fuel gases like methane and hydrogen, but so far they fall short of targets for gravimetric and volumetric capacity set to make them superior to current fuel technologies. 3,4Recently, theoretical studies have shown that lithium doping is a viable way to increase hydrogen uptake near ambient conditions, through increased electrostatic interactions. 5Additionally, crystalline ionic MOFs are promising materials for ion exchange and separation. 6MOFs that are easily impregnated by lithium ions should possess several attributes. 7First, this MOF should possess considerable stability, to retain porosity after lithium inclusion and the subsequent activation process.Second, this MOF should be negatively charged, so that Li + ions can be held in the framework by electrostatic forces.2][13] In principle, it is suggested that the exchange of bulkier counterions for small lithium ions should be a facile method to increase the porosity of any MOF with an anionic framework.
In order to test the hypothesis outlined above, several synthetic methods were tried to introduce lithium into several MOFs.In(1, 3 − BDC) − 2 (BDC = 1,3-benzene dicarboxylate) and In(NDC) − 2 (NDC = 2,6-naphthalene dicarboxylate) were synthesized according to the literature procedures. 9,10fter the crystal synthesis, exchange reactions were performed with various lithium metal sources, including lithium chloride, lithium acetate, and lithium nitrate.Unexpectedly, these efforts led to either framework collapse (possibly due to exposure to atmospheric moisture) or no increase in porosity.It is suggested that the cations in the frameworks as reported were too bulky to be removed by a simple exchange process, and thus this approach was unsuccessful.
In our experimental studies, a direct synthesis method was used, which introduced LiOH directly to the solvothermal reaction mixture before sonication and heating.This was intended to bring about the immediate formation in solution of some proportion of ligand molecules that were deprotonated to form lithium carboxylates, which would then form a framework with the lithium ions close at hand for charge balancing during synthesis.This resulted in an In-BDC framework with significantly increased porosity compared to the version filled with disordered tetramethyl ammonium cations in the framework, with the greatest porosity found when a number of moles of lithium hydroxide equal to the moles of ligand present were added before solvothermal synthesis.Based on stoichiometric control, we found that addition of LiOH in ratios higher than 1:1 LiOH:ligand produced amorphous material as shown by powder X-Ray Diffraction (PXRD).Attempts to produce MOF with added NaOH or KOH produced a non-crystalline, non-porous product, presumably due to the drastically increased basicity of those solutions.
Both investigated compounds, } n , possess a very similar structure to that described in the literature 9 as verified by PXRD (Figure 1), which was that of one dimensional anionic nanotubes, with each indium connected to four isophthalates, shown in Figure 2. Extraneous peaks were found in the PXRD measurement of , and 34.5 • , likely indicating some impurity accompanied the starting materials. 9s shown in Figure 1, slight downshifting of the peaks in the PXRD for {Li[In(1,3-BDC) 2 ]} n could indicate a slightly larger unit cell for the experimental pattern, while slight changes of the intensity ratios could be attributed to different occupancies, with lithium replacing tetramethylammonium. 14 The dimensional variation of the unit cell could be influenced by a different test temperature-the experimental PXRDs were taken at room temperature, which should cause a slight expansion of the unit cell compared to the reported single crystal XRD taken at 100 K. 15 It could also arise from a greater degree of solvent remaining in the crystals for the PXRD pattern, as the PXRD was taken from sample directly removed from the solvent without activation.The difference in peak intensities may be caused by solvent molecules and counterions in the channels of the structure, which could not be crystallographically localized but contribute to the experimental X-ray diffraction. 16or example, the highest observed intensity in the experimental pattern occurs on the (2 1 0) plane of the unit cell, which includes the line describing the center of the nanotube, shown in Figure 2. Therefore, a great deal of electron density is shown on this plane in the experimental pattern which is absent in the pattern simulated from the Crystallographic Information File(CIF).This may be because the single-crystal solution reported did not include the contents of this solvent-accessible space inside the nanotube.The large amount of electron density in the center of the nanotube shown by the experimental pattern likely consists of molecules trapped within the nanotubes, in addition to the oxygen and carbon atoms on the framework also intersected by the plane.It cannot be ruled out that some slight shift in the framework structure may have occurred as a consequence of lithium inclusion.However, the similarity of the PXRD patterns indicates that any such shift is minor.
Following solvent exchange and activation, gas sorption measurements were performed on both materials.The 4 showed H 2 uptake of 40 cm 3 /g at 800 millibars, and hysteresis in its H 2 isotherm was consistent with small pore size.{Li[In(1,3-BDC) 2 ]} n shows increased pore size as well as increased N 2 and H 2 uptake.This is consistent with a hypothesized smaller Li + counterion in an otherwise very similar framework (Fig. 3).
After repeated washes of the material with dimethylformamide(DMF) and ethanol, and multiple cycles of overnight solvent exchange, ICP-MS measurements showed that {Li[In(1,3-BDC) 2 ]} n possessed an indium-lithium ratio of approximately one lithium per two indium.This provides evidence that Li + is permanently ionically attached in the pores as a counterion, indicating significant but incomplete counterion replacement.While the presence of lithium in the In-BDC framework can be preliminarily confirmed by the ICP results, the increased surface area, and hydrogen adsorption data, further crystallographic evidence was sought.However, single-crystal XRD characterization of {Li[In(1,3-BDC) 2 ]} n was unsuccessful.Instead, synthesis of the In(2,6-NDC) series was undertaken to further explore the incorporation of Li + ions in anionic MOFs from the crystallographic perspective.
Single crystal XRD was taken for multiple crystals of {Li[In(NDC) 2 ]} n , all of which were solved in the P4/n space group.The reaction mixture including LiOH contained crystals with two separate habits.Clear rods were measured which also had a framework identical to {(Et 2 NH 2 )[In(2,6-NDC) 2 * 2H 2 O * DEF]} n , while purple blocks were also isolated which had identical connectivity, but a unit cell with a short axis elongated from 8.2 Å to 8.7 Å.This corresponds to a change in the indiumindium distance, 10 and the pore length, of the same amount on that axis.However, the topology of these networks remained consistent, and significant electron density next to the indium in the clear rods found in {(Et ) + counterion could be crystallographically located (Fig. 4). 10 We were careful to verify this before the removal of disordered solvent electron density using the SQUEEZE routine of PLATON, which could interfere with such analysis.ICP results also confirmed the presence of significant lithium in the pores even after washes with DMF and ethanol, though the Li-In ratio found for this sample may not be accurate due to difficulties in acquiring a sample including only one type of crystal.
Pure samples of {Li[In(NDC) 2 ]} n crystals were not obtained in quantities necessary for gas sorption measurement, as all samples prepared with LiOH contained both types of crystals.However, this provides evidence that Li + can be successfully incorporated into multiple, topologically dissimilar negatively charged indium MOFs through direct solvothermal synthesis in the presence of LiOH.See supplementary material for details of synthesis, XRD, and ICP-MS measurements. 17n conclusion, we showed that partial replacement of disordered ammonium cations with lithium ions in two anionic In-MOFs was possible through the pre-synthetic addition of LiOH to the reaction mixture.The effects of the replacement of ammonium cations in anionic In-MOFs with Li + were investigated, resulting in a surface area increase of over 1600% in {Li[In (  10,18 This article is copyrighted as indicated in the article.Reuse of AIP content is subject to the terms at: http://aplmaterials.aip.org/about/rights_and_permissionsDownloaded to IP: 165.91.178.73On: Fri, 30 Jan 2015 21:00:57 lithium inclusion through ICP-MS measurements and the presence of significantly reduced counterion electron density in the lithiated version through X-ray analysis.This synthetic approach may be applicable to more MOFs possessing anionic frameworks, and replacement of bulky with small and hard counterions should be accompanied by an increase in pore volume.Furthermore, as ICP-MS indicated a molar ratio of one lithium per two indium over repeated measurements, synthetic tuning may produce a greater degree of counterion replacement and allow higher porosity than that reported here.This synthetic strategy also provides a facile method to tune the pore sizes of various anionic MOFs, for example, size-selective separation of various gas molecules through counterion exchange in topologically identical frameworks.19 Finally, this material may be promising for investigation of its Li + or other ion conduction properties.20