Photocatalytic Hydrogen Production from Glycerol Aqueous Solutions as Sustainable Feedstocks Using Zr-Based UiO-66 Materials under Simulated Sunlight Irradiation

There is an increasing interest in developing cost-effective technologies to produce hydrogen from sustainable resources. Herein we show a comprehensive study on the use of metal–organic frameworks (MOFs) as heterogeneous photocatalysts for H2 generation from photoreforming of glycerol aqueous solutions under simulated sunlight irradiation. The list of materials employed in this study include some of the benchmark Zr-MOFs such as UiO-66(Zr)-X (X: H, NO2, NH2) as well as MIL-125(Ti)-NH2 as the reference Ti-MOF. Among these solids, UiO-66(Zr)-NH2 exhibits the highest photocatalytic H2 production, and this observation is attributed to its adequate energy level. The photocatalytic activity of UiO-66(Zr)-NH2 can be increased by deposition of small Pt NPs as the reference noble metal co-catalyst within the MOF network. This photocatalyst is effectively used for H2 generation at least for 70 h without loss of activity. The crystallinity of MOF and Pt particle size were maintained as revealed by powder X-ray diffraction and transmission electron microscopy measurements, respectively. Evidence in support of the occurrence of photoinduced charge separation with Pt@UiO-66(Zr)-NH2 is provided from transient absorption and photoluminescence spectroscopies together with photocurrent measurements. This study exemplifies the possibility of using MOFs as photocatalysts for the solar-driven H2 generation using sustainable feedstocks.


Introduction
Fossil fuels are still employed to supply more than 80% of the world energy demand [1]. The combustion of fossil fuels is directly connected to air pollution, emissions of greenhouse gases and climate change, resulting in negative impacts for the environment and human health [1][2][3]. For these reasons, there is an urgent need to develop alternative sustainable and renewable energy vectors that favor the decarbonization of the current energy system [4,5]. In this context, hydrogen is recognized as an ideal carbon-free energy vector candidate [6][7][8]. Today, however, hydrogen is mostly produced via steam reforming of hydrocarbons at high temperature (>650-1000 • C), with most energy requirements from nonrenewable fossil fuels [6,9]. Therefore, important scientific and technological efforts are being made to produce H 2 in a sustainable and renewable manner [7,8]. Electrolysis processes are relatively efficient and mature technologies for decomposing H 2 O into H 2 and O 2 that can operate using, for example, renewable photovoltaic energy [10]. The economic viability of this technology is, however, still hampered due to the investment and operation costs [7,8,11,12]. Alternatively, solar-driven photocatalytic water splitting into H 2 and O 2 using heterogeneous catalysts is a simpler and cheaper technology with potential application at a medium-large scale [13][14][15]. For this purpose, it would be necessary further increase the relatively low efficiencies achieved using inorganic semiconductors or investigate other type of photocatalysts [13,15,16]. this purpose, it would be necessary further increase the relatively low efficiencies achieved using inorganic semiconductors or investigate other type of photocatalysts [13,15,16].
Today, metal-organic frameworks (MOFs) [17][18][19] are considered among the most versatile and tunable materials to act as active photocatalysts [20][21][22][23]. Photocatalytic overall water splitting using MOFs is still a challenging process that is in its infancy [24][25][26]. A more efficient though less ambitious process closer to real small-to-medium industrial applications is the photocatalytic hydrogen evolution reaction (HER) in the presence of sacrificial electron donors [27]. The research in the field of photocatalytic HER using MOFs is exceptional, and excellent achievements have been reached even under visible or sunlight irradiation [27−40]. Most of the reports in this area have employed sacrificial electron donors such as methanol, triethanolamine (TEOA) or triethylamine to enhance the H2 production [27−40]. The use of these sacrificial agents limits, however, the cost-effective real application. Alternatively, in a similar way to that reported for inorganic semiconductors [27,41], the use of biomass or industrial byproduct feedstocks for H2 production is envisioned as a promising solution [42−43]. In this context, it is worth mentioning that biodiesel production generates about a 10 w/w % of crude glycerol as byproduct [44]. In this scenario, the development of technologies such as photoreforming of glycerol aqueous solution ( Figure 1) will contribute to both the viability of the biodiesel market and the production of sustainable H2 [42,44]. Theoretically, complete glycerol photoreforming results in a H2-to-CO2 molar ratio of 2.33 ( Figure 1a). The proposed reaction mechanism of glycerol photoreforming mainly using inorganic semiconductors is complicated and still under investigation [45]. Some studies have proposed that the oxidation process could proceed through different parallel reaction pathways involving initial oxidative C-C scission, oxidation of primary or secondary carbons (carbonyl group formation) or light-driven dehydration (Figure 1b) [46]. Importantly, the reaction pathway determines the H2-to-CO2 molar ratio [46]. The higher extent of carbonyl group formation with respect to oxidative C-C rupture leads to the higher H2:CO2 molar ratio [47]. Some studies have also reported the formation of byproducts such as formaldehyde, methane or CO among other possible products during glycerol photoreforming [45]. The occurrence of other processes such as photocatalytic water splitting might also take place simultaneously [24]. In the area of MOF-based materials, as far as we know, there is only one example reporting the use of an inorganic semiconductor/MOF composite, termed as TiO2@HKUST-1, used for hydrogen production from photoreforming of glycerol aqueous solutions (Figure 1a) [48].  Proposed initial reaction sequences of glycerol photoreforming using Rh/TiO 2 . Note: Light-driven dehydration to hydroxyacetone constitutes a side reaction (i) [46].
With these precedents in mind, the present study investigated in a comprehensive manner the H 2 generation from photoreforming of glycerol aqueous solution using Zrbased UiO-66 materials under simulated sunlight irradiation. Particularly, three UiO-Nanomaterials 2022, 12, 3808 3 of 17 66(Zr)-X (X: H, NH 2 , NO 2 ) solids having different energy band levels were selected for the study. For comparison, the photoactivity of a Ti-based MOF reference material such as MIL-125(Ti)-NH 2 was tested. The most active UiO-66(Zr)-NH 2 photocatalyst of the series was further modified with Pt NPs as the noble metal reference co-catalyst. The photocatalyst activity and stability of the Pt@UiO-66(Zr)-NH 2 was also studied. The reaction mechanism of the most active sample was further studied by means of transition absorption and photoluminescence spectroscopies as well as by photocurrent measurements.

Materials
All the chemicals employed in this work were of analytical or HPLC grade and supplied by Merck.

Photocatalytic Hydrogen Generation from Glycerol Aqueous Solution
Briefly, 5 mg of MOF photocatalyst was placed in a quartz reactor (51 mL) containing a glycerol aqueous mixture (20 mL), and then the system was sonicated (450 W) for 20 min to obtain a good MOF dispersion. To remove the air from the reactor, the system was purged with argon for 1 h. The MOF suspension while stirring was irradiated under simulated sunlight irradiation (Hamamatsu Hg-Xe lamp-150W-L8253; Hamamatsu spotlight source-L9566-04; Hamamatsu light guide A10014-50-0110; Lasing AM 1.5 G type filter-81094). During the irradiation, the temperature and the pressure of the system were monitored. At the required time, the evolved gases were analyzed from the head space of the quartz reactor by direct connection to a Micro GC system (Agilent 490 Micro GC system equipped with a Molsieve 5 A • column) that employed argon as the carrier gas. These photocatalytic measurements were carried out at least in triplicate trials, and the presented data correspond to the average of these experiments.

Photocatalyst Preparation and Characterization
The selected MOFs under study include the series of Zr-based MOF with UiO-66 topology, namely, UiO-66(Zr), UiO-66(Zr)-NH 2 and UiO-66(Zr)-NO 2 . It should be remembered that the UiO-66 solids are considered as benchmark Zr-MOF photocatalysts [51]. One of the important features of these MOFs is that the presence of the electron donor or acceptor functional groups in the organic ligand can tune the energy level of the solids [51]. The photocatalytic activity of the most active UiO-66 solid under study was compared with that of MIL-125(Ti)-NH 2 as reference Ti-MOF [52]. These four MOFs were prepared by solvothermal method as previously reported and described in the SI [49,50]. Figure 2 shows the characteristic diffraction patterns of the four MOFs under study that are in good agreement with their respective simulated patterns and the thermogravimetric analyses of the MOFs that confirmed the thermal stability up to about 350 • C.
donor or acceptor functional groups in the organic ligand can tune the energy level of the solids [51]. The photocatalytic activity of the most active UiO-66 solid under study was compared with that of MIL-125(Ti)-NH2 as reference Ti-MOF [52]. These four MOFs were prepared by solvothermal method as previously reported and described in the SI [49,50]. Figure 2 shows the characteristic diffraction patterns of the four MOFs under study that are in good agreement with their respective simulated patterns and the thermogravimetric analyses of the MOFs that confirmed the thermal stability up to about 350 °C. XPS spectroscopy was used to characterize the oxidation state of the elements present in the MOFs under study. Figure 3 compiles the survey and XPS spectra of MOFs, while the corresponding deconvolutions of the elements can be found in the Supplementary Materials ( Figures S1-S4). The C 1s spectra show the presence of a band centered at 284.4 eV assigned to C-C sp 2 of the terephthalate organic ligand. The C 1s spectra of UiO-66(Zr)-NH2 and MIL-125(Ti)-NH2 exhibit an additional band at 285.5 eV due to the presence of the C-N bond in the amino functional groups. In the case of the UiO-66(Zr)-NO2, a band appearing at 286 eV due to the presence of the nitro group can be observed. The O 1s spectra of the four MOFs show a broad band that corresponds to the oxygen atoms present in the carboxylate (about 532 eV) and zirconium or titanium oxo clusters (529.8 eV). In the case of the UiO-66(Zr)-NO2 sample, a band due to the oxygen atoms present in the nitro group (~532.5 eV) can be also assigned. The N 1s spectra of the amino-or nitro-based MOFs show their characteristic band centered at about 399 or 405 eV, respectively. The Zr 3d spectra of the UiO-66 samples exhibit two main bands characteristic of the Zr(IV) ions in the oxo clusters at 182.1 and 184.4 eV that correspond to the Zr 3d5/2 and Zr 3d3/2, respectively. In the case of the MIL-125(Ti)-NH2 solid, the XPS Ti 2p due to the presence of Ti(IV) in the octameric oxo cluster resulted in the observation of Ti XPS spectroscopy was used to characterize the oxidation state of the elements present in the MOFs under study. Figure 3 compiles the survey and XPS spectra of MOFs, while the corresponding deconvolutions of the elements can be found in the Supplementary Materials ( Figures S1-S4). The C 1s spectra show the presence of a band centered at 284.4 eV assigned to C-C sp 2 of the terephthalate organic ligand. The C 1s spectra of UiO-66(Zr)-NH 2 and MIL-125(Ti)-NH 2 exhibit an additional band at 285.5 eV due to the presence of the C-N bond in the amino functional groups. In the case of the UiO-66(Zr)-NO 2 , a band appearing at 286 eV due to the presence of the nitro group can be observed. The O 1s spectra of the four MOFs show a broad band that corresponds to the oxygen atoms present in the carboxylate (about 532 eV) and zirconium or titanium oxo clusters (529.8 eV). In the case of the UiO-66(Zr)-NO 2 sample, a band due to the oxygen atoms present in the nitro group (~532.5 eV) can be also assigned. The N 1s spectra of the amino-or nitro-based MOFs show their characteristic band centered at about 399 or 405 eV, respectively. The Zr 3d spectra of the UiO-66 samples exhibit two main bands characteristic of the Zr(IV) ions in the oxo clusters at 182.1 and 184.4 eV that correspond to the Zr 3d 5/2 and Zr 3d 3/2 , respectively. In the case of the MIL-125(Ti)-NH 2 solid, the XPS Ti 2p due to the presence of Ti(IV) in the octameric oxo cluster resulted in the observation of Ti 2p 1/2 (458 eV) and Ti 2p 3/2 (464 eV) signals. In addition, FT-IR spectroscopy confirms the presence of the expected vibration bands of the carboxylate (around 1500 and 1300 cm −1 ), amino (3518-3382 cm −1 ) or nitro (1542 and 1498 cm −1 ) groups present in the terephthalate organic ligand of the MOFs ( Figure S5).
The porosity of the MOFs under study was measured by isothermal N 2 adsorption experiments (Table 1 and Figures S6-S9). The estimated BET surface area and pore volumes are summarized in Table 1. Remarkably is the higher porosity of MIL-125(Ti)-NH 2 with respect to the UiO-66 solids. SEM measurements showed that all the four MOFs under study exhibit average particle size distribution with values lower than 400 nm (Table 1 and Figures S10 and S11). SEM coupled to EDX reveals a good distribution of the elements in the samples (Figures S12-S15). 2p1/2 (458 eV) and Ti 2p3/2 (464 eV) signals. In addition, FT-IR spectroscopy confirm presence of the expected vibration bands of the carboxylate (around 1500 and 1300 c amino (3518-3382 cm −1 ) or nitro (1542 and 1498 cm −1 ) groups present in the terephth organic ligand of the MOFs ( Figure S5).     One of the important aspects in heterogeneous photocatalysis is to determine the energy level of the solid employed for this purpose. Three main factors that determine this diagram in the case of MOFs include their band gap, the highest occupied crystal orbital (HOCO) and the lowest unoccupied crystal orbital (LUCO). Furthermore, the optical properties of the solids were studied by UV-Vis diffuse reflectance spectroscopy. Figure S16 shows that the UiO-66(Zr) exhibits a main broad absorption in the UV range due to the presence of the zirconium oxo clusters and the terephthalate organic ligands. The presence of the nitro and specially the amino group in the terephthalate ligand of the UiO-66(Zr) results in bathochromic shift in the absorption toward visible light region. From these optical absorption data and the corresponding Tauc plots, the optical band gap of UiO-66(Zr), UiO-66(Zr)-NO 2 and UiO-66(Zr)-NH 2 was estimated to be 3.91, 3.27 and 2.81 eV, respectively ( Figure S16). Similarly, the optical band gap of the MIL-125(Ti)-NH 2 was estimated as 2.65 eV ( Figure S16). These results are in agreement with previous analogous reports and confirm the relatively simplicity to tune the band gap of MOFs by introducing one functional in the organic ligand [53]. Furthermore, XPS spectroscopy was also employed to characterize the valence band maximum of the UiO-66(Zr)-X (X:H, NH 2 , NO 2 ) and MIL-125(Ti)-NH 2 samples ( Figure S17). From these values and the optical band gap the conduction band energy minimum was calculated (see details in Supplementary Materials, Section S2. Characterization of the MOF-based materials). Figure 4 illustrates the energy band diagram of the MOF samples under study. The obtained results indicate that the presence of functional groups in the organic ligand reduces the band gap of the solids with respect to the parent MOF and determines the HOCO and LUCO values of the samples. As it is shown later, these energy level diagrams influence to a large extent the observed photocatalytic activity.

Photocatalytic Results
Initially, the activity of UiO-66(Zr), UiO-66(Zr)-NH2, UiO-66(Zr)-NO2 and MIL-125(Ti)-NH2 was evaluated for the generation of H2 from glycerol photoreforming in aqueous solution under simulated sunlight irradiation. The main role of glycerol is to act as an electron donor during the photocataytic reaction by scavenging the photogenerated holes [45]. During this process, glycerol becomes oxidized through several possible steps, resulting in the formation of CO2 and H2 ( Figure 1). Furthermore, the photogenerated

Photocatalytic Results
Initially, the activity of UiO-66(Zr), UiO-66(Zr)-NH 2 , UiO-66(Zr)-NO 2 and MIL-125(Ti)-NH 2 was evaluated for the generation of H 2 from glycerol photoreforming in aqueous solution under simulated sunlight irradiation. The main role of glycerol is to act as an elec-tron donor during the photocataytic reaction by scavenging the photogenerated holes [45]. During this process, glycerol becomes oxidized through several possible steps, resulting in the formation of CO 2 and H 2 ( Figure 1). Furthermore, the photogenerated electrons have the ability to reduce protons to H 2 [45]. Therefore, the evolved hydrogen during the photocatalytic process is determined by the ability of the MOF-based photocatalyst for both glycerol oxidation and proton reduction toward hydrogen production. As previously commented, glycerol photoreforming is a process that involves the generation of H 2 :CO 2 with a molar ratio of 2.33 ( Figure 1a). The results shown in Figure 5 indicate that the UiO-66(Zr)-NH 2 solid is the most active photocatalyst of the series for the HER with a H 2 :CO 2 molar ratio of 7.3 after 3 h of reaction. It should be commented that, in all cases, the conversion of glycerol was below 0.5%, and during the photocatalytic reaction, traces of methane were also detected. Furthermore, the used MOF-based photocatalysts retain their crystallinity and morphology, as revealed by XRD and SEM measurements, respectively (Figures S18 and S19). The higher activity of UiO-66(Zr)-NH 2 with respect to UiO-66(Zr)-NO 2 or UiO-66(Zr) is attributed to the higher ability of the former to absorb visible light. The higher photoactivity of UiO-66(Zr)-NH 2 compared to MIL-125(Ti)-NH 2 is remarkable, even given the slightly higher band gap and lower BET surface area and pore volume of the former solid (Table 1). This observation can be mainly attributed to the more negative LUCO value of UiO-66(Zr)-NH 2 with respect to MIL-125(Ti)-NH 2 to favor the HER. In order to put in context of the achieved photocatalytic results, the most active UiO-66(Zr)-NH 2 was compared with the existing analogous study using MOFs. It should be mentioned that the photocatalytic activity can depend to a large extent on the reaction conditions such as glycerol aqueous solution concentration, catalyst amount, temperature, reactor design or irradiation source and, thus, comparison with literature results should be cautiously taken. With these comments in mind, the achieved activity of UiO-66(Zr)-NH 2 (0.62 mmol g −1 h −1 after 3 h; H 2 :CO 2 7.3) is higher than that achieved using the Cu-MOF termed as HKUST-1 (0.1 mmol g −1 h −1 after 8 h, H 2 :CO 2~0 .9), both under simulated sunlight irradiation [48].  Of particular interest is also the observation of different H2-to-CO2 ratios as a function of the MOF employed as the photocatalyst. These observations may be associated with the occurrence of different photochemical reactions taking place simultaneously during the photoreforming of glycerol aqueous solutions with each specific MOF ( Figure  1). As commented in the introduction, the occurrence of primary or secondary carbon oxidation (carbonyl formation) to a larger extent than the oxidative C-C rupture ( Figure  1) would favor a higher H2-to-CO2 molar ratio. Therefore, it can be assumed that the use of UiO-66(Zr)-NO2 as the photocatalyst might favor the carbonyl formation pathway to a Of particular interest is also the observation of different H 2 -to-CO 2 ratios as a function of the MOF employed as the photocatalyst. These observations may be associated with the occurrence of different photochemical reactions taking place simultaneously during the photoreforming of glycerol aqueous solutions with each specific MOF ( Figure 1). As commented in the introduction, the occurrence of primary or secondary carbon oxidation (carbonyl formation) to a larger extent than the oxidative C-C rupture (Figure 1) would favor a higher H 2 -to-CO 2 molar ratio. Therefore, it can be assumed that the use of UiO-66(Zr)-NO 2 as the photocatalyst might favor the carbonyl formation pathway to a larger extent than UiO-66(Zr)-NH 2 or UiO-66(Zr) and, therefore the higher H 2 -to-CO 2 molar ratio observed ( Figure 5). This fact could be explained by considering the higher oxidation ability of UiO-66(Zr)-NO 2 with a more positive HOCO value (+2.22 V) than the other two UiO-66 solids (Figure 4). According to previous study, the higher acidity of metals nodes of UiO-66(Zr)-NO 2 with respect to UiO-66(Zr)-NH 2 or UiO-66(Zr) may also influence some steps of the reaction mechanism of aqueous glycerol photoreforming [54]. Furthermore, the use of UiO-66 photocatalysts results in an increase in the H 2 -to-CO 2 molar ratio as the reaction proceeds from 1 to 3 h, and these observations might be associated with the occurrence of the carbonyl pathways to a larger extent at the longer reaction time. For comparison, the similar H 2 -to-CO 2 ratio at 3 h observed for MIL-125(Ti)-NH 2 with respect to UiO-66(Zr)-NO 2 may be also due to the fact that both photocatalysts exhibit similar HOCO values that would favor hydrogen evolution through the carbonyl formation pathway. In the case of MIL-125(Ti)-NH 2 , the formation of H 2 at 1 h is accompanied with only traces of CO 2 , while as the reaction proceeds, more CO 2 is formed toward the theoretical 2.33 molar ratio. However, further studies are required to gain more insights about the complex reaction mechanism for glycerol photoreforming using MOF-based materials. Using the most active UiO-66(Zr)-NH 2 photocatalyst under study, the influence of the glycerol aqueous solution concentration on the H 2 production was studied. Figure 6 shows that the H 2 production increases along with the glycerol aqueous solution concentration up to 14 vol%. Further increase in the glycerol concentration resulted in slight decrease in activity. The optimum glycerol concentration observed of 14 vol% to maximize H 2 production can be interpreted considering that the number of photogenerated electronhole pairs can efficiently react up to this value. The use of higher glycerol concentration might be responsible for deactivation of the MOF by either poisoning the redox active sites or blocking the porosity. For comparison, the use of methanol instead glycerol as the sacrificial agent, with the same number of moles, results in a relatively higher activity for H 2 production (4.46 mmol g −1 at 3 h) with respect to glycerol (1.85 mmol g −1 at 3 h). Considering the relatively higher oxidation potential of methanol (0.016 V vs. NHE) with respect to glycerol (0.004 V vs. NHE) [55], it can be assumed that better diffusion of the former through the porous structure of the UiO-66(Zr)-NH 2 plays a significant role in the observed photocatalytic activity. Regardless of these comments, it should be remembered that, currently, methanol production is mainly through a catalytic reaction that involves the hydrogenation of CO with H 2 obtained from steam reforming of hydrocarbons, and, therefore, its use as the sacrificial electron donor for H 2 production is far from practical applications [6,9]. Therefore, the use of highly available glycerol as a byproduct of biodiesel can contribute simultaneously to the development of this market and to the development of sustainable technologies for H 2 generation.
To further study the H 2 generation from the photoreforming of glycerol aqueous solutions, the active UiO-66(Zr)-NH 2 material was modified with Pt NPs as the co-catalyst. Pt NPs are among the preferred reference noble metal NPs to promote the photocatalytic hydrogen generation [47]. For example, previous studies have reported the use of Pt NPs supported on MOFs such as UiO-66(Zr)-NH 2 as photocatalysts for the HER using methanol or TEOA as sacrificial agents [56][57][58]. Frequently, the use of Pt NPs loadings below 3 wt% is considered an adequate strategy to boost the photocatalytic activity for HER reaction [56,58]. In the present study, the innovation resides in the use for the first time of Pt NPs-supported UiO-66(Zr)-NH 2 at low metal loading (1 wt%) as photocatalyst for hydrogen production in a sustainable manner from glycerol photoreforming. PXRD of the as-prepared Pt NPs-supported UiO-66(Zr)-NH 2 by the photodeposition method shows that the resulting MOF-based material retains its initial crystallinity (Figure 7a). DF-STEM of the Pt@UiO-66(Zr)-NH 2 reveals an average particle size and standard deviation of Pt NPs of 1.27 and 0.43 nm, respectively (Figure 7b). Point EDX analyses confirmed the presence of Pt NPs within the MOF framework ( Figure S20). potential of methanol (0.016 V vs. NHE) with respect to glycerol (0.004 V vs. NHE) [5 can be assumed that better diffusion of the former through the porous structure o UiO-66(Zr)-NH2 plays a significant role in the observed photocatalytic activity. Reg less of these comments, it should be remembered that, currently, methanol producti mainly through a catalytic reaction that involves the hydrogenation of CO with H tained from steam reforming of hydrocarbons, and, therefore, its use as the sacri electron donor for H2 production is far from practical applications [6,9]. Therefore use of highly available glycerol as a byproduct of biodiesel can contribute simultaneo to the development of this market and to the development of sustainable technologie H2 generation. To further study the H2 generation from the photoreforming of glycerol aqu solutions, the active UiO-66(Zr)-NH2 material was modified with Pt NPs as co-catalyst. Pt NPs are among the preferred reference noble metal NPs to promot photocatalytic hydrogen generation [47].  HER using methanol or TEOA as sacrificial agents [56−58]. Frequently, the use of Pt NPs loadings below 3 wt% is considered an adequate strategy to boost the photocatalytic activity for HER reaction [56,58]. In the present study, the innovation resides in the use for the first time of Pt NPs-supported UiO-66(Zr)-NH2 at low metal loading (1 wt%) as photocatalyst for hydrogen production in a sustainable manner from glycerol photoreforming. PXRD of the as-prepared Pt NPs-supported UiO-66(Zr)-NH2 by the photodeposition method shows that the resulting MOF-based material retains its initial crystallinity (Figure 7a). DF-STEM of the Pt@UiO-66(Zr)-NH2 reveals an average particle size and standard deviation of Pt NPs of 1.27 and 0.43 nm, respectively (Figure 7b). Point EDX analyses confirmed the presence of Pt NPs within the MOF framework ( Figure S20).     After the characterization of the Pt@UiO-66(Zr)-NH 2 sample, this solid was employed as the photocatalyst for H 2 generation from glycerol aqueous solution (14 vol%). The results show higher H 2 photocatalytic production when using Pt@UiO-66(Zr)-NH 2 (2.3 mmol g −1 in 3 h) with respect to the parent UiO-66(Zr)-NH 2 (1.7 mmol g −1 in 3 h). Remarkably, the Pt@UiO-66(Zr)-NH 2 material can be used under the same reaction conditions for 70 h (Figure 9a) while retaining its crystallinity and morphology, as revealed by XRD and SEM measurements, respectively (Figure 9b,c). DF-STEM measurements (Figure 9d) coupled with EDX analyses ( Figure S22) also allowed us to confirm the absence of significant Pt NP aggregation (1.34 ± 0.45 nm) with respect to the fresh sample (1.27 ± 0.43 nm).

Reaction Mechanism
To gain some understanding about the photocatalytic performance and rea mechanism of Pt@UiO-66(Zr)-NH2 and UiO-66(Zr)-NH2 solids, several spectroscopi electrochemical measurements were carried out. Initially, the PL response of Ar-pu acetonitrile MOF suspensions having the same absorbance value (ca 0.35) at 340 corresponding with the excitation wavelength of the organic ligand, was measured ure 10 shows that the presence of Pt NPs (1 wt%) within the UiO-66(Zr)-NH2 solid r in a decrease in PL of 66% with respect to the parent UiO-66(Zr)-NH2. These resul interpreted in a way that considers the fact that the presence of Pt NPs avoids, at partially, the photoinduced charge recombination that is the process responsible fo observed photoluminescence emission. In fact, photocurrent measurements us Pt@UiO-66(Zr)-NH2/FTO (fluorine doped tin oxide) working electrode upon polariz and dark/illumination cycles allow us to confirm the occurrence of photoinduced c separation, a process that is favored in the presence of glycerol. This enhanceme photocurrent associated with the role of glycerol as the sacrificial electron donor c explained considering that it quenches the photogenerated holes and increases the tocurrent intensity. This effect agrees with the previous photocatalytic data in whic presence of glycerol in water increases the photocatalytic H2 production ( Figure 6).

Reaction Mechanism
To gain some understanding about the photocatalytic performance and reaction mechanism of Pt@UiO-66(Zr)-NH 2 and UiO-66(Zr)-NH 2 solids, several spectroscopic and electrochemical measurements were carried out. Initially, the PL response of Ar-purged acetonitrile MOF suspensions having the same absorbance value (ca 0.35) at 340 nm, corresponding with the excitation wavelength of the organic ligand, was measured. Figure 10 shows that the presence of Pt NPs (1 wt%) within the UiO-66(Zr)-NH 2 solid results in a decrease in PL of 66% with respect to the parent UiO-66(Zr)-NH 2 . These results are interpreted in a way that considers the fact that the presence of Pt NPs avoids, at least partially, the photoinduced charge recombination that is the process responsible for the observed photoluminescence emission. In fact, photocurrent measurements using a Pt@UiO-66(Zr)-NH 2 /FTO (fluorine doped tin oxide) working electrode upon polarization and dark/illumination cycles allow us to confirm the occurrence of photoinduced charge separation, a process that is favored in the presence of glycerol. This enhancement of photocurrent associated with the role of glycerol as the sacrificial electron donor can be explained considering that it quenches the photogenerated holes and increases the photocurrent intensity. This effect agrees with the previous photocatalytic data in which the presence of glycerol in water increases the photocatalytic H 2 production ( Figure 6).
TAS was also used to further study the reaction mechanism when using UiO-66(Zr)-NH 2 or Pt@UiO-66(Zr)-NH 2 as photocatalysts. Specifically, TAS spectrum of both MOFs recorded upon irradiation at ligand-centered absorption at 355 nm under Ar atmosphere clearly shows the presence of two bands (Figure 11a). The first band appearing from 350 to 450 nm and second one from 550 to 750 nm, both exhibiting different decays of kinetics and profiles. This fact suggests that these two absorbance bands correspond to different species. In the case of UiO-66(Zr)-NH 2 , the first region is characterized by transient absorbance decay kinetics with two components, one quick and intense, attributed to the occurrence of charge separation, and other residual, with a longer lifetime due to charge delocalization along the MOF crystal (Figure 11b). The second band observed in the region from 550-750 nm is mainly characterized by one component decay (Figure 11c, blue line). In the case of the Pt@UiO-66(Zr)-NH 2 sample, the presence of Pt mainly causes the quenching of the second component, inhibiting photoinduced charge delocalization in the MOF (Figure 11b). More specifically, the transient absorbance lifetime decay of UiO-66(Zr)-NH 2 (65 ns) is quenched in the presence of Pt NPs in UiO-66(Zr)-NH 2 (40 ns) (Figure 11c). In addition, the presence of Pt NPs results in a quenching of the transient absorbance spectrum of UiO-66(Zr)-NH 2 , particularly intense in the region from 550 to 750 nm. These observations can be interpreted considering that metallic Pt NPs have the ability of trapping photoinduced generated charges. Aiming to bring light to the nature of the observed photoinduced charge separation species for the most active Pt@UiO-66(Zr)-NH2 sample, additional TAS measurements in the presence of quenchers were carried out. Methanol was selected as the hole quencher due to its good electron-donor behavior, while molecular oxygen was chosen as the electron acceptor. Figures S23 and S24 show that the use of these two quenchers results in Aiming to bring light to the nature of the observed photoinduced charge separation species for the most active Pt@UiO-66(Zr)-NH 2 sample, additional TAS measurements in the presence of quenchers were carried out. Methanol was selected as the hole quencher due to its good electron-donor behavior, while molecular oxygen was chosen as the electron acceptor. Figures S23 and S24 show that the use of these two quenchers results in significant changes to the TAS measurements. In particular, the presence of methanol quenches the intensity signal and the spectra decay along the whole spectrum, specifically the region between 350 and 450 nm ( Figure S23). These observations suggest that the transient absorption observed mainly from 350 to 450 nm corresponds to photogenerated holes. The use of O 2 as the quencher also modifies the transient absorption profiles and decreases the decay lifetimes ( Figure S24). For instance, the presence of O 2 decreases the decay lifetimes from 65 to 38 ns recorded at 380 nm or from 45 to 30 ns at 680 nm. The observed quenching of transient absorbance intensity due to the presence of O 2 in the region from 550 to 750 nm is associated with the presence of photogenerated electrons. These observations agree with the role of supported Pt NPs as an electron reservoir responsible for the observed decrease in both signal intensity and lifetime of the transient species in the region from 550 to 750 nm with respect to the parent MOF. Figure 11d illustrates a simplified reaction mechanism showing the photoinduced electron transfer from the organic ligand to the metal node of the MOF. Then, the electrons present in the LUCO are transferred to Pt NPs where the reduction reactions occur, while oxidations take place in the HOCO localized in the organic ligand.
In addition to the above spectroscopic and electrochemical characterization, the influence of the light irradiation source with Pt@UiO-66(Zr)-NH 2 was also investigated. Figure 12 shows the highest photocatalytic activity achieved using UV-Vis light, followed by simulated sunlight irradiation and then visible light irradiation. In any case, these results confirm unambiguously that Pt@UiO-66(Zr)-NH 2 behaves as an efficient photocatalyst under both visible or simulated sunlight irradiation.

Conclusions
The present study has shown the possibility of using UiO-66 topology for H2 generation from photoreforming of glycerol aqueous solutions under simulated sunlight irradiation. The most active photocatalyst of the series was UiO-66(Zr)-NH2, followed by UiO-66(Zr)-NO2 and the less active UiO-66(Zr). The activity of MIL-125(Ti)-NH2 was lower than that of UiO-66(Zr)-NH2. The observed order of activity was mainly attributed to the energy level difference of the solids, where UiO-66(Zr)-NH2 has an appropriate band gap to absorb visible light together with adequate LUCO value to promote reduction reactions. Deposition of Pt NPs as the co-catalyst for hydrogen generation within the UiO-66(Zr)-NH2 network has significantly improved the photocatalytic activity of the MOF. In addition, the Pt@UiO-66(Zr)-NH2 photocatalyst was employed for about 70 h with good H2 generation and without loss of crystallinity or change in its morphology based on PXRD and SEM measurements, respectively. Furthermore, Pt particle size dis-

Conclusions
The present study has shown the possibility of using UiO-66 topology for H 2 generation from photoreforming of glycerol aqueous solutions under simulated sunlight irradiation. The most active photocatalyst of the series was UiO-66(Zr)-NH 2 , followed by UiO-66(Zr)-NO 2 and the less active UiO-66(Zr). The activity of MIL-125(Ti)-NH 2 was lower than that of UiO-66(Zr)-NH 2 . The observed order of activity was mainly attributed to the energy level difference of the solids, where UiO-66(Zr)-NH 2 has an appropriate band gap to absorb visible light together with adequate LUCO value to promote reduction reactions. Deposition of Pt NPs as the co-catalyst for hydrogen generation within the UiO-66(Zr)-NH 2 network has significantly improved the photocatalytic activity of the MOF. In addition, the Pt@UiO-66(Zr)-NH 2 photocatalyst was employed for about 70 h with good H 2 generation and without loss of crystallinity or change in its morphology based on PXRD and SEM measurements, respectively. Furthermore, Pt particle size distribution was also retained, as revealed by TEM measurements. Evidence in support of the occurrence of photoinduced charge separation with Pt@UiO-66(Zr)-NH 2 was obtained by TAS and photocurrent measurements. In addition, PL measurements using Pt@UiO-66(Zr)-NH 2 and UiO-66(Zr)-NH 2 indicated that the use of the Pt NPs avoids to some extent the occurrence of undesirable photoinduced charge recombination. The study of the influence of light irradiation on H 2 generation from glycerol aqueous solution using Pt@UiO-66(Zr)-NH 2 revealed that this photocatalyst exhibits good activity under both visible and simulated sunlight irradiation. The authors consider that this study will contribute to the field of solar-driven photocatalytic H 2 generation using MOFs and sustainable feedstocks as sacrificial agents.

Data Availability Statement:
The data presented in this study are available on request from the corresponding authors.