No abstract

The metal−organic framework (MOF) H 3 [(Cu 4 Cl) 3 − (BTTri) 8 , H 3 BTTri = 1,3,5-tris( 1 H-1,2,3-triazol-5-yl)benzene] (CuBT-Tri) is a precatalyst for biomedically relevant nitric oxide (NO) release from S-nitrosoglutathione (GSNO). The questions of the number and nature of the catalytically most active, kinetically dominant sites are addressed. Also addressed is whether or not the well-defined structural geometry of MOFs (as solid-state analogues of molecular compounds) can be used to generate specific, testable hypotheses about, for example, if intrapore vs exterior surface metal sites are more catalytically active. Studies of the initial catalytic rate vs CuBTTri particle external surface area to interior volume ratio show that intrapore copper sites are inactive within the experimental error (≤1.7 × 10 −5% of the observed catalytic activity)restated, the traditional MOF intrapore metal site catalysis hypothesis is disproven for the current system. All observed catalysis occurs at exterior surface Cu sites, within the experimental error. Fourier transform infrared (FT-IR) analysis of CN − -poisoned CuBTTri reveals just two detectable Cu sites at a ca. ≥0.5% detection limit, those that bind three or one CN − ("Cu(CN) 3 " and "CuCN"), corresponding to the CN − binding expected for exterior surface, 3-coordinate (Cu surface ) and intrapore, 5-coordinate (Cu pore ) sites predicted by the idealized, metal-terminated crystal structure. Two-coordinate Cu defect sites are ruled out at the ≥0.5% FT-IR detection limit as such defect sites would have been detectable by the FT-IR studies of the CN − -poisoned catalyst. Size-selective poisoning studies of CuBTTri exterior surface sites reveal that 1.3 (±0.4)% of total copper in 0.6 ± 0.4 μm particles is active. That counting of active sites yields a normalized turnover frequency (TOF), TOF norm = (4.9 ± 1.2) × 10 − 2 mol NO (mol Cu surface ) −1 s −1 (in water, at 20 min, 25 °C, 1 mM GSNO, 30% loss of GSNO, and 1.3 ± 0.4 mol % Cu surface ) a value ∼100× higher than the TOF calculated without active site counting. Overall, Ockham's razor interpretation of the data is that exterior surface, Cu surface sites are the catalytically most active sites present at a 1.3 (±0.4)% level of total Cu.

Evidence is provided for a previously unknown proton-coupled electron transfer (PCET) mechanism for nitric oxide (NO) release from endogenous S-Nitrosoglutathione (GSNO) catalyzed by the metal−organic framework (MOF) H 3 [(Cu 4 Cl) 3 (BTTri) 8 ] (CuBTTri) in the presence of glutathione (GSH). The balanced reaction stoichiometry, active site characterization, and the experimental rate law are used to systematically disprove competing mechanistic hypotheses, leading unexpectedly to PCET as the proposed basic mechanism. The PCET mechanism contrasts traditionally proposed, either formally Cu II to Cu I redox or other formally Cu II Lewis acid mechanisms for NO generation by other Cu-based catalysts. The proposed PCET mechanism sets the stage for mechanistically guided syntheses of improved Cu-MOF catalysts for GSNO to NO conversion and for computational investigations on the CuBTTri/GSNO/GSH/NO system as further tests of the well-defined structure of CuBTTri, structures of hypothesized reaction intermediates, and specific questions and hypotheses generated by the proposed PCET mechanism.

Metal–organic frameworks (MOFs) have high porosity and surface area, making them ideal candidates for adsorption-mediated applications. One high-value application is the removal of uremic toxins from solution for dialysis. Previous studies have reported adsorptive removal of the uremic toxin p-cresyl sulfate from solution via zirconium-based MOFs, but a specific analysis of parameters contributing to adsorptive uptake is needed to clarify differences in uptake performance between MOFs. We synthesized zirconium 1,3,5-benzenetricarboxylate (MOF-808) and an iron-based analog, MIL-100­(Fe), and compared their adsorptive uptake with previously reported values of other zirconium-based MOFs. MIL-100­(Fe) adsorbed three times more p-cresyl sulfate from solution on a per mass basis than MOF-808 and had a greater adsorption efficiency than 75% of previously reported Zr-based MOFs. We compared p-cresyl sulfate uptake by MOFs as a function of BET surface area, number of aromatic carbons in the organic linker, internal cage diameter, and pore window diameter. There is poor correlation between p-cresyl sulfate uptake and each of the variables considered, but the number of aromatic carbons of the MOF linker was a better predictor of uptake than BET surface area (R 2 = 0.7034 and 0.1430, respectively), and pore window aperture was a better predictor of uptake than the pore cage diameter (R 2 = 0.4780 and 0.0383, respectively). We hypothesize that the greater adsorptive capacity of MIL-100­(Fe) compared to MOF-808 results from direct coordination of p-cresyl sulfate to vacant metal sites in the MOF, and the total adsorption may be accounted for by some combination of adsorptive interactions occurring at both metal and organic linker sites near to the exterior particle surface. The adsorptive uptake of p-cresyl sulfate by MIL-100­(Fe) was observed to increase with p-cresyl sulfate content, mass of MIL-100­(Fe), and volume of p-cresyl sulfate solution; the mass of MIL-100­(Fe) had the greatest effect on total adsorption.

Thermogravimetric analysis (TGA) is a technique which can probe chemisorption of substrates onto metal organic frameworks. A TGA method was developed to examine the catalytic oxidation of S-nitrosoglutathione (GSNO) by the MOF H3[(Cu4Cl)3(BTTri)8] (abbr. Cu-BTTri; H3BTTri = 1,3,5-tris­(1H-1,2,3-triazol-5-yl)­benzene), yielding glutathione disulfide (GSSG) and nitric oxide (NO). Thermal analysis of reduced glutathione (GSH), GSSG, GSNO, and Cu-BTTri revealed thermal resolution of all four analytes through different thermal onset temperatures and weight percent changes. Two reaction systems were probed: an aerobic column flow reaction and an anaerobic solution batch reaction with gas agitation. In both systems, Cu-BTTri was reacted with a 1 mM GSH, GSSG, or GSNO solution, copiously rinsed with distilled–deionized water (dd-H2O), dried (25 °C, < 1 Torr), and assessed by TGA. Additionally, stock, effluent or supernatant, and rinse solutions for each glutathione derivative within each reaction system were assessed by mass spectrometry (MS) to inform on chemical transformations promoted by Cu-BTTri as well as relative analyte concentrations. Both reaction systems exhibited chemisorption of glutathione derivatives to the MOF by TGA. Mass spectrometry analyses revealed that in both systems, GSH was oxidized to GSSG, which chemisorbed to the MOF whereas GSSG remained unchanged during chemisorption. For GSNO, chemisorption to the MOF without reaction was observed in the aerobic column setup, whereas conversion to GSSG and subsequent chemisorption was observed in the anaerobic batch setup. These findings suggest that within this reaction system, GSSG is the primary adsorbent of concern with regards to strong binding to Cu-BTTri. Development of similar thermal methods could allow for the probing of MOF reactivity for a wide range of systems, informing on important considerations such as reduced catalytic efficiency from poisoning, recyclability, and loading capacities of contaminants or toxins with MOFs.