Elucidating Improvements to MIL‐101(Cr)’s Porosity and Particle Size Distributions based on Innovations and Fine‐Tuning in Synthesis Procedures

Among the existing metal–organic frameworks (MOFs), MIL‐101(Cr) is renowned for its stability in air and water. As a result, MIL‐101(Cr) has numerous potential applications ranging from adsorptive cooling to catalysis. The industrial‐scale production of MIL‐101(Cr) is necessary before realizing these applications. Yet, there remain two main bottlenecks in preparing MIL‐101(Cr) in bulk: the toxicity of hydrofluoric acid (HF) used in conventional MIL‐101(Cr) synthesis and the challenge of ensuring that the as‐prepared MIL‐101(Cr) is highly porous with specific Brunauer–Emmett–Teller surface area (SBET) above 4000 m2 g−1. On the laboratory scale, MIL‐101(Cr) often presents SBET 2300–3500 m2 g−1. The synthesis and purification procedures often influence the yield, particle size, porosity, and other properties of MIL‐101(Cr). This critical review examines trends in MIL‐101(Cr) preparation procedures and the MOF's resulting properties to elucidate areas for improvement toward its real‐world applications. The purification processes for conventional HF‐based MIL‐101(Cr) whereby porosities vary despite the same synthesis approach are first investigated. Next, the reported additives for substituting HF and their influence on the resulting MIL‐101(Cr)’s porosity and particle size are discussed. The selection of additives may be application‐specific: exemplified in the examination of MIL‐101(Cr)’s preparation, its corresponding water sorption capacity, and desiccant‐related applications.


Literature Search
This review utilizes information from 446 articles (n = 446). We conducted the literature search in March 2023. First, we used the search parameters i) most relevant, ii) with the exact phrase "MIL-101," and iii) setting the year 2005 to 2022 to identify several works through Google Scholar (n = 913). Subsequently, we excluded the non-English articles, thesis works, reviews, and duplicates (n = 891). Since we are focusing on MIL-101(Cr) prepared through hydrothermal means, we further excluded 311 primary articles that solely discussed functionalized MIL-101(Cr), composites containing MIL-101(Cr) or MIL-101(Cr) from microwaveassisted syntheses, and other methods for n = 579 (see Supporting Information Table S1). Next, to compare the porosities of the MIL-101(Cr) samples, the remaining articles should report both S BET and V pore . Therefore, we excluded another 133 articles for n = 446.  or MIL-101(Cr)-containing composites that limited their discussion on pristine MIL-101(Cr) itself.
Several factors contribute to HF-MIL-101(Cr) samples having S BET below 2300 m 2 g −1 (Group 1, 49 entries). For instance, the partial pressure (P/P 0 ) range for S BET calculation may differ from the conventional P/P 0 range of 0.03-0.3, [34] such as 0.01-0.1 [35] or 0.05-0.2, [12] hindering comparisons between S BET values. Another variable is the HF:H 2 O molar ratios in the reaction mixture. Since HF is available as an aqueous solution, the weight percentage of HF in the solution may change over time as water evaporates, inadvertently increasing the molar ratio of HF in the reaction mixture. Existing studies illustrate that the nonoptimal molar ratios of water in MIL-101(Cr) syntheses affect the yield and porosity of the resulting HF-MIL-101(Cr) samples. [11,36,37] More importantly, the purification processes conducted may be inadequate in removing residual H 2 BDC, solvents, and side products within and around the pores of HF-MIL-101(Cr). [11,19,20] Purification or activation processes for HF-MIL-101(Cr) samples include washing with additional solvents (heated or at room temperature) such as N,N-dimethylformamide (DMF), ethanol or methanol, and ammonium fluoride (NH 4 F) solutions. Combinations of three or more solvents (including deionized water) may be used. Subsequently, the HF-MIL-101(Cr) samples may also be dried under ambient pressure or vacuum-dried. Table 2 presents the number of entries and percentages of HF-MIL-101(Cr) samples activated through several common approaches.
Following hydrothermal synthesis, one of the first steps in purifying HF-MIL-101(Cr) samples includes filtering them over large pore glass filters to remove recrystallized H 2 BDC [10] due to their limited solubility in water at room temperatures. As shown in Table 2, this approach was applied to a larger proportion of samples in Group 3 than in Group 2. Moreover, six out of eight entries with S BET above 4000 m 2 g −1 were purified by filtering out recrystallized H 2 BDC as residue. [10,14,29,[38][39][40] Residual H 2 BDC can also be removed by dissolution in alkaline media and alcohols. As organic amines and amides are more suitable for dissolving H 2 BDC than inorganic hydroxides, DMF washings are common when activating MIL-101(Cr) samples. [11] Based on the data in Table 2, this ubiquitous approach does not seem to increase the S BET to above 3500 m 2 g −1 , even with hot DMF. Only one of the eight entries with S BET above 4000 m 2 g −1 uses DMF (at ambient temperature) during purification, [41] suggesting that washing MIL-101(Cr) with DMF has limited efficacy because DMF, as a high boiling point solvent (153°C), may persist in MIL-101(Cr) after activation. Therefore, organic amines and amides with low boiling points may substitute DMF for preparing highly porous MIL-101(Cr). Furthermore, phasing out activation in DMF could permit greener and more cost-effective MIL-101(Cr) preparations on an industrial scale. [8] On the other hand, the results in Table 2 indicate that rinsing with hot ethanol or methanol increases S BET . Six of the eight MIL-101(Cr) samples with S BET above 4000 m 2 g −1 were also activated in hot ethanol. [14,29,[38][39][40][41] Notably, H 2 BDC is more soluble in methanol than in water at 25°C. Although this step also helps to remove free H 2 BDC, [42][43][44] the more polar nature of ethanol and methanol over DMF may imply that alcohols also remove inorganic impurities from MIL-101(Cr)'s pores. Chromium oxide is one such impurity. [30] Ethanol also has a lower boiling point (78°C) than DMF, thus facilitating its subsequent removal from MIL-101(Cr). Washing samples with NH 4 F solution is also an effective activation procedure. Typically, this step is introduced after washing the HF-MIL-101(Cr) samples with water and ethanol [11] and is followed by further washing with hot deionized water to remove residual salts. [14,15,43] The process of alkaline halide exchange could remove both inorganic and organic species from the pores of HF-MIL-101(Cr) samples. NH 4 F supplies F − ions for ion exchange with residual terephthalate anions (BDC 2− ) that are coordinated to the open metal sites of MIL-101(Cr), removing BDC 2− as soluble ammonium terephthalate salts. [29] Samples washed with NH 4 F solution constitute most of the samples in Group 3 (Table 2) and heating the NH 4 F solution seems to increase the washing efficacy. [31,[45][46][47][48][49][50][51] Seven out of the eight MIL-101(Cr) samples with S BET above 4000 m 2 g −1 were also activated in hot NH 4 F solutions. [14,29,[38][39][40][41]43] Unfortunately, NH 4 F causes acute toxicity, [21] thus limiting its use in scaled-up MIL-101(Cr) syntheses.
Overall, activating HF-MIL-101(Cr) through multiple solvents appears effective (Table 2). One reason may be the removal of both inorganic and organic species via solvents of different polarities. Lastly, drying the MIl-101(Cr) samples in vacuo may not remove all residual solvents to achieve higher S BET . The common practice appears to be vacuum-drying samples at 150°C [41,[47][48][49][50] to remove residual DMF but higher temperatures (e.g., 160°C) [52][53][54] are feasible. One reason is that the high affinity between F − ions and water molecules often hinders the evacuation of water molecules from HF-MIL-101(Cr) at 150°C. [55] However, the corresponding pressure range (e.g. 1 × 10 −5 Torr) is rarely reported. [56] Notably, this brief analysis excludes several factors such as the number of washing cycles per solvent, the volume of solvent used per cycle, and the sequence of solvent exchange. Similar analyses with larger data sets and statistical tools could pinpoint more accurate means of improving HF-MIL-101(Cr) purification processes.

Influence of Additives on MIL-101(Cr)'s Porosity and Particle Size
HF is a contact poison that quickly penetrates and damages tissues. [16,31] Therefore, substituting HF with less toxic alternatives or additive-free MIL-101(Cr) syntheses are attractive options. Table 3 presents the properties of the selected additives discussed herein.
As shown in Figure 7b, MIL-101(Cr) prepared using organic monocarboxylic acids does not have S BET above 4000 m 2 g −1 .
Like HNO 3 -MIL-101(Cr) and HCl-MIL-101(Cr), the samples synthesized using these organic acids mostly have S BET between 2300 and 3500 m 2 g −1 . Through an unconventional activation procedure, Wang et al. reported CH 3 COOH-MIL-101(Cr) with S BET = 3788 m 2 g −1 . Notably, this procedure involved a hydrothermal reaction at 180°C for 12 h followed by heating and stirring the reaction mixture in sodium acetate and ethanol before filtering out H 2 BDC and carrying our subsequent washing in hot ethanol and NH 4 F solution. [105] Likewise, improved activation procedures could raise S BET for other MIL-101(Cr) samples prepared using organic acids. While the reaction temperature and time for CH 3 COOH-MIL-101(Cr) can be reduced to 180°C and 4 h, [106] nine of the ten entries with S BET above 3200 m 2 g −1 were still synthesized at 200-220°C for 8 h or more. [16,21,97,[107][108][109][110][111][112] This observation may imply that lowering reaction temperatures and durations may also compromise MIL-101(Cr)'s porosity when using monocarboxylic acids as additives. However, the extent of mineralization achieved via monocarboxylic acids versus HF is unclear. . [104] On the contrary, Rallapalli et al. reported increasing S BET from C 6 F 5 COOH-MIL-101(Cr), HF-MIL-101(Cr) to CH 3 COOH-MIL-101(Cr). [16] Three other reports also recorded higher S BET in CH 3 COOH-MIL-101(Cr) than HF-MIL-101(Cr). [55,112,113] On the other hand, Figure 7b demonstrates how V pore above 2 cm 3 g −1 often occurs when larger molecular weight monocarboxylic acids, such as C 6 H 5 COOH and its derivatives and longchain CH 3 (CH 2 ) 16 COOH, modulate MIL-101(Cr) synthesis. [15,98] This result suggests that these monocarboxylic acids compete with H 2 BDC for coordination in the MIL-101(Cr) framework, introducing defects that increase V pore above the expected 2.15 cm 3 g −1 . [10,11,14] Hence, these acids could yield MIL-101(Cr) samples of higher porosities and are safer to handle than HF (Table 3).

Optimizing MIL-101(Cr) for Improved Water Sorption Capacity
Theoretically, materials with type V water sorption isotherms are ideal desiccants as they can absorb water at lower relative pressures and are more readily regenerated than other materials. [168] Therefore, given its high water uptake capacity and type V isotherm, [22,169,170] MIL-101(Cr) has promising applications as a solid polymeric desiccant. This section aims to correlate MIL-101(Cr)'s preparation and water sorption capacity for identifying MIL-101(Cr) samples with high water sorption capacities and highlight gaps in realizing its desiccant-related applications. Figure 10 presents simulated water molecules in mesoporous MIL-101(Cr) under increasing relative humidity. [30] Overall, MIL-101(Cr)'s water sorption isotherm can be divided into three sections: relative humidity (RH) < 35%, RH 35-50%, and RH 50-90%. [33,171] Initially, at RH < 35%, water coordinates to the open metal sites of MIL-101(Cr). [33,50,84,169,171] Specifically, the coordination stems from Coulombic interactions between the unsaturated Cr 3+ sites of MIL-101(Cr) and the oxygen atom of water molecules. [50,172] Molecular simulation results support these adsorbent-adsorbate interactions, including the weak coordination of water to saturated Cr 3+ sites. [30,50] 1D water chains form within MIL-101(Cr)'s pores as more water molecules interact with the unsaturated Cr-bonded water. [50,172] Within this low RH range, MIL-101(Cr)'s water sorption capacity varies linearly with RH. [33,171] Interestingly, water molecules do not interact with the carboxylate oxygen in MIL-101(Cr) with a low atom charge. [50,172] As RH increases to 35-50%, MIL-101(Cr) has a steep rise in water vapor uptake. [33,84,171] This enhanced water uptake occurs because water undergoes capillary condensation in the pores of MIL-101(Cr). [33,171] Hysteresis occurs during desorption due to the cavitation-induced evaporation of water, further illustrating that capillary condensation occurs within this RH range. [50,171] In molecular simulations, the 1D water chains lengthen and connect to form a monolayer on the inner mesoporous cages of MIL-101(Cr). [50] This monolayer changes the cages from hydrophobic to hydrophilic to induce capillary condensation in MIL-101(Cr). [50,173] For measurements collected over smaller RH intervals, two adsorption steps may be visible as water vapor fills the 29 Å and then the 34 Å mesoporous cages of MIL-101(Cr). [33,50,62,172,[174][175][176] This difference is because the 29 Å cages have a smaller inner surface with greater curvature than the 34 Å cages. [50] Furthermore, MIL-101(Cr)'s heat of adsorption decreases rapidly from about 80 kJ mol −1 at low RH to slightly above the water's heat of evaporation (40.69 kJ mol −1 ) with a minute increase in RH, demonstrating that water-water interactions govern adsorption after the Cr 3+ sites are saturated. [30,169,173] Water molecules also enter the super tetrahedra of MIL-101(Cr). [50] Although the smaller 8.6 Å super tetrahedra can repel adjacent water molecules, they also contain up to five water molecules at higher RH. [50] Third, as RH rises between 50% and 90%, water mainly adsorbs in the interparticulate voids of MIL-101(Cr) powder, raising the water uptake slightly. [33,171,174,177] Hence, MIL-101(Cr) is a suitable desiccant when ambient RH is 50% and above to realize its high water uptake capacities between 0.85 and 1.66 g g −1 (measured at T = 298 K, P/P 0 = 0.90-0.95) (Figure 11 and Supporting Information Table S2). [22,113,170,175] Moreover, the physisorption of water in MIL-101(Cr) favors the MOF's low-temperature regeneration. [178]
From the available data, HF-MIL-101(Cr) has a maximum water sorption capacity of 1.3 g g −1 . [33,50,180] However, HNO 3 -, [82,84] CH 3 COOH-, [113] TMAOH-, [34,170] and (additivefree)-MIL-101(Cr) [165] samples have water uptake capacities 1.3 g g −1 and above. Interestingly, none of these six samples was washed with NH 4 F solution. Hence, F − anions were absent and these MIL-101(Cr) samples have the formula sample were instrumental in understanding these differences. [30] At low RH, the fully fluorinated MIL-101(Cr) structure has different partial charges on its coordinatively unsaturated Cr 3+ sites versus the fully hydroxylated MIL-101(Cr) structure. [30] As a result, the fluorinated structure features a larger ratio of the number of water molecules close to the unsaturated Cr 3+ sites to the number of water molecules surrounding the coordinated Cr 3+ sites. [30] The hydroxylated structure favors hydrogen bonding between the hydroxyl group's oxygen atom and water's hydrogen atom, resulting in lower heat of adsorption toward higher water sorption at low RH. [30] This effect influences the isotherm's shape at low RH but becomes less important at higher RH whereby interactions between water molecules govern sorption behavior. [30] Nonetheless, the differing partial charges on coordinatively unsaturated Cr 3+ sites can affect capillary condensation and subsequent adsorption as RH increases. Therefore, fully hydroxylated MIL-101(Cr), like the six samples highlighted above, would favor higher water uptake capacities.
Conversely, MIL-101(Cr) with hydrophobic groups has also been investigated. These short-chain groups include -F and -CH 3 . [109] As shown in Table 4, MIL-101(Cr)-F and MIL-101(Cr)-CH 3 have lower water uptake than pristine MIL-101(Cr) even at T = 303 K, P/P 0 = 0.32. [109] However, unlike fully functionalized MIL-101(Cr)-F, partially fluorinated samples such as Diaz-Ramirez et al.'s MIL-101(Cr)-4F(1%) can still have a relatively higher water uptake capacity at T = 298 K, P/P 0 = 0.90 (Table 4). [109,195] This sample was prepared using tetrafluoroterephthalic acid instead of 2-fluoroterephthalic acid. [109,195] At T = 303 K, P/P 0 = 0.20, MIL-101(Cr)-4F(1%) had a higher water uptake of 7.81 versus 5.37 wt% for (additive-free)-MIL-101(Cr). [195] One reason is that the few molecules of water entering MIL-101(Cr)-4F(1%) at low RH are held in place through dipolar interactions with the fluorine atoms instead of coordinating with the Cr 3+ sites. [30,195] Hence, this MOF could also be suitable for low RH applications. Nonetheless, MIL-101(Cr) has also been functionalized with long-chain hydrocarbons. Specifically, Li et al. introduced n-propyl, n-hexyl and n-dodecyl groups to MIL-101(Cr)'s open Cr 3+ sites to attain increasingly hydrophobic samples from MIL-101(Cr)-pro, -hex to -dod. [82] Not only do the increasing carbon chain lengths result in decreasing water uptake capacity, but the samples' water contact angles also increased from 17.9°, 34.3°, 56.4°to 146.5°. [82] This result indicates that besides collecting water sorption isotherms, water contact angle measurements provide a quick estimate of a sample's hydrophilicity. [82,174] While MIL-101(Cr) and its functionalized counterparts are promising, there remain obstacles to their industrial applications. Figure 12 charts MIL-101(Cr)'s progress toward real-world applications using De Lange et al.'s framework for developing MOF-based adsorption-driven heat pumps. [191] Existing literature has demonstrated MIL-101(Cr)'s stability in air and water (Stage 0), its type V water sorption isotherms and high water uptake capacities (Stage 1) [22,169,170] and the feasibility of calculating  [191] and have been adapted to chart MIL-101(Cr)'s advances toward industrial applications.
its thermodynamic efficiency and working capacity (Stage 2). [191] Research into shaping MIL-101(Cr) into granules and coatings (Stage 3) and the resultant heat and mass transfer properties of such samples (Stage 4) are ongoing. For instance, there are several reports on compacting MIL-101(Cr) into granules [183,192] and using binders to prepare MIL-101(Cr) coatings. [84,95,174,191,196,197] The adhesion between MIL-101(Cr), the binder, and the coated surface is important for effective contact heat transfer because adsorption is an exothermic process. [6,191,198] The binder should also be chemically inert toward the MIL-101(Cr)-water working pair without hindering water sorption. [6,198] Moreover, the MIL-101(Cr) coatings or granules should retain high cyclic stability. [6,141,191,198] The scaled-up synthesis in Stage 5 directly influences Stage 6 because kilograms of MIL-101(Cr) could be required for system-level performance tests. [191,199] Finally, Stage 6 optimizes various application-specific parameters to ensure the proposed MIL-101(Cr)-based technology is sufficiently robust for commercial applications. [6,191] These system-level performance tests are paramount as laboratory prototypes [35,65,141,183,186,192,197] and numerical modeling [199] may not accurately represent realworld conditions.

Recommendations for Future Work and Conclusion
Despite being an MOF with at least 17 years of history, MIL-101(Cr) pends further research before its industrial applications. As discussed herein, there are two bottlenecks in MIL-101(Cr) preparations: the toxicity of HF and the challenge of raising MIL-101(Cr)'s S BET above 4000 m 2 g −1 . Figure 13 compares the current state and envisioned future of preparing and utilizing MIL-101(Cr) based on improvements discussed in this review. Reports on the effects of various additives on MIL-101(Cr) syntheses are instrumental toward realizing HF-free procedures. Specifically, the choice of additive or even additive-free syntheses influences MIL-101(Cr)'s porosity and particle size. Several additive-based syntheses (e.g., NaOH-and CH 3 COONa-MIL-101(Cr)) require further optimization and novel additives such as glycine may also be worth exploring. [200] Combinations of additives present another alternative for synthesizing MIL-101(Cr), including using sodium acetate and the surfactant cetyltrimethylammonium bromide in tandem. [28] Moreover, as exemplified by HF-MIL-101(Cr), hydrothermal synthesis should be accompanied by thorough activation or purification for highly porous MIL-101(Cr) for sorption applications. Removing recrystallized H 2 BDC by filtration and washing the crude MIL-101(Cr) with solvents of varying polarities often improve the sample's S BET . Finding greener low boiling point alternatives to DMF and NH 4 F is also another area for further study. In addition, the articles surveyed herein suggest that fully hydroxylated MIL-101(Cr) may have greater water sorption capacities than its fluorinated counterpart and that functionalized MIl-101(Cr) may be more relevant to low RH applications such as adsorption chillers and adsorption-driven heat pumps.
One existing limitation of this review is the limited sample size. More data would become available with time, permitting multivariate or similar analyses to correlate data more accurately. Standardized reporting of metrics, such as S BET and V pore and the pressure range for calculation could also facilitate comparison across the literature. Although it influences sorption kinetics and capacity, particle size is reported less frequently than S BET and V pore . For instance, larger grains facilitate mass transfer whereas smaller grains favor heat transfer. [201] There are also differences in the small-and large-scale syntheses of MIL-101(Cr) that require further study. [12,21] Lastly, the autoclaves used for synthesis present another variable as the local variations of heat transfer within the autoclave influence particle sizes whereby higher heating rates favor larger crystals. [97] www.advancedsciencenews.com www.advmatinterfaces.de Figure 13. Comparison of the current and envisioned future state of MIL-101(Cr)'s synthesis, activation, and applications.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.