A Robust and Biocompatible Bismuth Ellagate MOF Synthesized Under Green Ambient Conditions

The first bioinspired microporous metal–organic framework (MOF) synthesized using ellagic acid, a common natural antioxidant and polyphenol building unit, is presented. Bi2O(H2O)2(C14H2O8)·nH2O (SU-101) was inspired by bismuth phenolate metallodrugs, and could be synthesized entirely from nonhazardous or edible reagents under ambient aqueous conditions, enabling simple scale-up. Reagent-grade and affordable dietary supplement-grade ellagic acid was sourced from tree bark and pomegranate hulls, respectively. Biocompatibility and colloidal stability were confirmed by in vitro assays. The material exhibits remarkable chemical stability for a bioinspired MOF (pH = 2–14, hydrothermal conditions, heated organic solvents, biological media, SO2 and H2S), attributed to the strongly chelating phenolates. A total H2S uptake of 15.95 mmol g–1 was recorded, representing one of the highest H2S capacities for a MOF, where polysulfides are formed inside the pores of the material. Phenolic phytochemicals remain largely unexplored as linkers for MOF synthesis, opening new avenues to design stable, eco-friendly, scalable, and low-cost MOFs for diverse applications, including drug delivery.


Contents 1. Chemicals and Synthesis
Chemical Reagents S1 Synthesis of SU-101 S1

Chemicals and Synthesis Chemical Reagents
The reagents were obtained as follows: reagent-grade ellagic acid (97%, isolated from chestnut tree bark) was purchased from Acros Organics, bismuth acetate (99%) was purchased from Alfa Aesar, glacial acetic acid (EMSURE glacial acetic acid, 100%) was purchased from Merck, supplemental ellagic acid (90%, isolated from pomegranate hulls) was purchased from PureBulk (PureBulk Inc, Roseburg, Oregon, USA), white vinegar was bought in a Swedish convenience store ('Perstorp Ättika', 24% acetic acid in water). All chemicals were used as-received without further purification. Acetic acid solutions were prepared with deionized water along with the acetic acid source.

Synthesis of SU-101
In a typical synthesis, 0.15 g reagent-grade ellagic acid (Acros Organics, 97% ellagic acid, isolated from chestnut tree bark, 0.5 mmol) and 0.38 g bismuth acetate (Alfa Aesar, 99%, 1 mmol) were added to 30 mL of a water and acetic acid mixture (6 vol.% acetic acid, made from glacial acetic acid, with the possibility of replacing with store-bought white vinegar) in common borosilicate glass beakers with PTFE stir bars. The resulting suspension (pH ≈ 2.3) was stirred at room temperature for 48 h, after which it was centrifuged at 8000 rpm for 10 min and placed in an oven (60 °C) to dry overnight. It was found that successively higher surface areas could be obtained after washing the as-synthesized material with water or with both water and ethanol. Yield (after washing with water and ethanol, then drying overnight at 60 °C): 0.28 g (76% of theoretical yield). Larger batches of SU-101 were synthesized using 3.3 g of ellagic acid (90 wt. %, isolated from pomegranate hulls, 10 mmol), sold as a dietary supplement (PureBulk Inc.), and 7.6 g bismuth acetate (20 mmol) which was added to a beaker containing 600 mL of a water and acetic acid mixture (6 vol.% acetic acid, made from glacial acetic acid, with the possibility of replacing with store-bought white vinegar). The solution was allowed to stir for 48 h before being centrifuged at 8000 rpm for 10 min, after which it was placed in an oven at 60 °C overnight. Yield after washing with water and ethanol, then drying overnight at 60 °C: 5.6 g (74% of theoretical yield).
The phase purity of both materials was confirmed by PXRD and elemental analysis. Calculated (%) for Bi2O(H2O)2(C14H2O8)·2H2O using reagent-grade ellagic acid: C 20 Larger crystals of SU-101 (> 2 µm) could be synthesized as a phase-mixture under hydrothermal conditions by adding 10 mg of reagent-grade ellagic acid (Acros Organics, 97% ellagic acid, isolated from chestnut tree bark, 0.03 mmol) and 24 mg bismuth acetate (Alfa Aesar, 99%, 0.06 mmol) to a 5 mL borosilicate 3.3 glass tube (Duran 12 x 100 mm, DWK Life Sciences) containing 3 mL of deionized water. The tube was then sealed with a polybutylene terephthalate (PBT) cap containing a PTFE seal, whereafter it was heated to 120 °C for 16 h. The contents were then filtered off and left to dry at ambient conditons.

Scanning electron microscopy
Scanning electron microscopy (SEM) images were collected on a JEOL JSM7401F SEM.

Transmission electron microscopy
Transmission electron microscopy (TEM) images were collected on a JEOL JEM 2100-LaB6 operating at 200 kV using a Gatan ORIUS 200 D detector.

3D electron diffraction (3DED)
Three-dimensional electron diffraction data were collected using a JEOL JEM2100-LaB6 TEM, equipped with a Timepix detector from Amsterdam Scientific Instruments, while continuously rotating the crystal at 0.45° s −1 . The experiment was carried out using Instamatic, 1 with data reduction being performed in XDS. 2 The acquired intensities were then used to solve the structure of SU-101 with SHELXT, 3 and refined using SHELXL, 4 with electron scattering factors extracted from SIR2014. 5 From the 3DED data, all non-hydrogen atoms could be located in the initial structure solution from SHELXT. A data completeness of 99% could be obtained due to the high symmetry of the crystals, belonging to the tetragonal space group P42/n (No. 86). Upon refinement against the acquired 3DED data, dynamical scattering, which is common to 3DED, led to an R1 value considered high in comparison to acceptable refinement statistics for SCXRD data. Therefore, the structure was subsequently refined against high-resolution synchrotron X-ray powder diffraction data ( Figure S4), despite confidence in the 3DED model.

High-resolution PXRD
High-resolution PXRD data for the structure refinement of SU-101 were collected at 11BM at the Advanced Photon Source (APS), Argonne National Laboratory, USA, using the dedicated mail-in system, for which the sample was loaded in a Kapton capillary and measured with an X-ray wavelength of 0.457863 Å (stepsize of 0.000998 Å) using a multi-analyzer detector assembly. The refinement of the electron-diffraction model against high-resolution PXRD data was carried out in TOPAS-Academic V6. 6 Topological analysis of the SU-101 framework was carried out using the software package ToposPro, 7 as well as Systre and 3dt (both part of the GAVROG package). 8,9 Fig. S4. Plot for the structure refinement of SU-101. High-resolution PXRD data were collected at 11BM at the APS, Argonne National Laboratory, USA. λ = 0.457863 Å.

Thermogravimetric analysis (TGA)
Thermogravimetric analysis data were gathered on a sample of SU-101 using a TA Instruments Discovery TGA. The sample was put into a platinum crucible and heated in air from 28 °C to 600 °C with a heating rate of 10 °C min -1 . The sum formula best matching the observed data was determined as Bi2O(H2O)2(C14H2O8)nH2O (n = 1, i.e. one rather than two non-coordinated water molecules, as was determined from the refinement against HR-PXRD data).

Sorption properties
Gas adsorption/desorption isotherms were recorded on a Micromeritics ASAP2020 surface area and porosity analyzer. Prior to the experiments, the samples were pretreated at 150 °C under vacuum for 10h. Nitrogen adsorption/desorption isotherms were recorded at liquid nitrogen temperature (-196 °C). A liquid nitrogen bath was used as temperature control. The Brunauer−Emmett−Teller (BET) specific surface area of SU-101 was calculated using the N2 adsorption points recorded at a relative pressure range p/p0 = 0.02 -0.10. Nitrogen (N2), carbon dioxide (CO2), and methane (CH4) adsorption-desorption isotherms were recorded at 0°C. An ice slurry bath was used as the temperature control for these experiments.

Stability in solvents and solutions
For the stability tests, 10 mg of SU-101 was added to a 5 mL glass vial fitted with a screw-cap. For every trial, 1 mL of each respective solvent or solution was added and the resulting dispersion was stirred at room temperature or 80 °C for 24 h unless otherwise specified.

Stability in the presence of L-cysteine and L-cystine
The integrity of SU-101 in the presence of L-cysteine and L-cystine at 37 °C was evaluated by preparing solutions with 1 mg mL -1 of the material and each respective compound in water. The SU-101 dispersions were then allowed to stir at 37 °C for 24 h before the MOF was retrieved by centrifugation (8000 rpm, 10 min). PXRD patterns were acquired after the powders were allowed to dry under ambient conditions. Residual crystalline L-cystine can be observed in the sample previously immersed in a solution of L-cystine.

pH-dependent stability
For the pH-dependent stability tests, 20 mg of as-synthesized SU-101 was immersed in 3 mL of stock solution, prepared from either NaOH or concentrated HCl, as to obtain the desired pH. Fig. S18. Powder X-ray diffraction patterns acquired of SU-101 after being stirred in aqueous solutions at various pH levels, at room temperature. The phase acquired at pH < 2 is tetragonal BiOCl. When exposed to 20 M NaOH the material is dissolved, forming a clear yellow solution. . Powder X-ray diffraction was collected on the pellet, while the supernatant was stored in an Eppendorf tube for dynamic light scattering (DLS) analysis. The evolution of the SU-101 particle size and ζ -potential as well as its structure were evaluated over time in presence of diverse physiological media (aqueous solution (Milli-Q water) and cell culture media «RPMI»). Duplicates of SU-101 particles were dispersed at 1 mg mL −1 by using an ultrasound water bath in desired media at 37 °C under continuous stirring. After different incubation times (1, 2, 4, 6 and 24 h), the colloidal stability was evaluated by dynamic light scattering (DLS; Zetasizer Nano, Malvern Instruments). Subsequently, in each incubation time and in each type of media, SU-101 samples were centrifugated (14.500 rpm, 10 min). The remaining solid were kept in order to collect the XRPD patterns. Profiles were generally collected in the 3° < 2θ < 30° range with a typical step size of 0.02° in continuous mode using a D8 Advance Bruker diffractometer (Cu Kα1 radiation, λ = 1.5406 Å).

Cytotoxicity studies
The cytotoxic activity of SU-101 as well as its precursors (57% of Bi(AOC)3 and 43% of ellagic acid) was analyzed by the colorimetric MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. 10,11 The promyelocytic cell line HL-60 (cells in suspension) was seeded 24h prior to the assay in 96-well plates at a density of 1 × 10 5 cells per well in RPMI (Roswell Park Memorial Institute medium) supplemented with 10% FBS. The particle suspensions were prepared as a dilution series with cell culture media (30 µL of the sample in aqueous solution were added to a final volume of 300 µL per well), yielding different concentrations (from 1000 to 8 µg . mL -1 ). Subsequently, all these treatments were added into the cells for 24 h, while being kept at 37 °C with a 5% CO2 atmosphere. S17 °C for 2 h) followed by a PBS washing with 100 µL, ending with 100 µL of dimethylsulfoxide (DMSO) added to each well. Absorbance was determined at λ= 539 nm under stirring. The percentage of cell viability was calculated by the absorbance measurements of control growth and test growth in the presence of the formulations at various concentration levels.

SO2 and H2S capture SO2 capture
The adsorption-desorption SO2 isotherms were carried out at 298, 303 and 308 K up to 1 bar in a Dynamic Gravimetric Gas/Vapour Sorption Analyser, DVS vacuum (Surface Measurement Systems Ltd). Before each experiment, as-synthesized SU-101 was activated at 120 °C and 1.7 × 10 -6 Torr for 6 h.

SO2 isosteric heat of adsorption
The acquired SO2 adsorption isotherms were used to estimate the isosteric heat of adsorption (Fig.  S18). The isosteric heat of adsorption was determinated by fitting a virial-type equation (Eq. S1) to 303 and 308 K SO2 adsorption isotherms. Where p is the pressure, n is the amount adsorbed and A0, A1, ... are the virial coefficients (A2 and higher terms can be negelcted at lower coverage values). A plot of ln(n/p) vs. n should give a straight line at low surface coverage (Fig. S19). The heat of adsorption was estimated to be -29.60 kJ mol -1 .

PXRD after humid SO2 adsorption experiments
Humid SO2 adsorption studies were carried out in a previously reported lab-made system. 12 The system contains two principal parts (Fig. S26): SO2 gas generator (left), a dropping funnel with conc. H2SO4 [1] connected to a schlenk flask with Na2SO3 (s) under stirring [2]; and a saturation chamber (right), constructed from a round-bottom flask with distilled water [3], connected to a sintered glass filter adapter [4] and a vacuum line [5]. The activated sample is placed on the glass filter adapter. The PXRD results indicated structural retention after exposure to humid SO2 (Fig. S27).

H2S breakthrough experiments
H2S experiments were carried out using a HP 5890 GC, where the exhausted gas was continuously injected to the gas chromatograph, acquiring a chromatogram for each injection. From the corresponding chromatogram the H2S signal was integrated to obtain its quantity. Knowing the H2S concentration from the feed, the H2S concentration can be calculated for each injection, as the saturation concentration is the original feed concentration. Dynamic breakthrough experiments were carried out in a lab-made system (Scheme S1).
Scheme S1. Representation of breakthrough dynamic system for H2S uptake experiments.
The H2S adsorption capacity for each cycle was calculated using Eq. S2, where VH2S represents the H2S volumetric capacity (cm 3 g -1 ), m the adsorbent mass (g), F the input flow rate (cm 3 min -1 ), Cf and Ct the influent and downstream H2S concentrations respectively (% vol), and t the time (min). 13

Eq. S2
As mentioned before, the adsorption column has a porous glass bed and a blank run was performed before each experiment to eliminate the adsorption contribution of the column. In Fig. S21 the black circles represent the adsorption of the column, and the others circles represent the MOF adsorption for each cycle. Then the corrected volumetric capacity 'VH2S,corr' for SU-101 was estimated using Eq. S3 for each cycle.

Eq. S3
The H2S adsorption capacity is often reported as qH2S (mol g -1 ), this value was roughly estimated with the volumetric adsorption capacity VH2S,corr (cm 3 g -1 ) and the ideal gas law Eq. S4, where p is the system pressure (77.3 kPa), T the measurement temperature (298 K), and R the ideal gas constant (8314.4598 cm 3 kPa K -1 mol -1 ).

Eq. S4
The H2S uptake for SU-101 was measured in this home-made system using the following conditions: average temperature of 298 K, a gas concentration of 4.3% vol H2S/N2 and a gas flow of 25 mL min -1 , p = 0.78 atm and 50 mg of SU-101 material were used for each adsorption cycle. Before carrying out the experiments the material was activated at 120 °C for 6 h under a flow of N2 (25 mL min -1 ). For each experiment, three independent measurements were performed.

PXRD after H2S experiments
Patterns were collected in Bragg-Brentano geometry with Cu Kα1 radiation (λ = 1.5406 Å) on a Rigaku ULTIMA IV. The powder patterns were recorded from 5 to 40° (2θ) in 0.02° steps and a scan rate of 0.2° min -1 .

Raman spectroscopy
The Raman experiments were measured on a DXR2 Thermo Scientific instrument with a lamp of 780 nm and 10X microscope objective for samples of SU-101 before and after H2S experiments.

FTIR Spectroscopy
FTIR spectra were measured (in-situ and at 25 °C) using an FTIR Nicolet 6700 spectrophotometer (DTGS detector) with a 4 cm -1 resolution equipped with a diffuse reflectance vacuum chamber with CaF2 windows.

Proposed mechanism for polysulfide formation
The formation of polysulfides from adsorbed H2S can proceed through a sequence of steps: in the first part, H2S is adsorbed within the SU-101 material (inside the pores) modifying the redox properties of H2S. The mechanism can be explained as below, which is adapted from previous studies. 14,15 I. H2S adsorption at the surface of the MOF: 2 ( ) → 2 ( ) II.