Photocatalytic Overall Water Splitting over MIL-125(Ti) upon CoPi and Pt Co-catalyst Deposition.

Photocatalytic overall water splitting is realized by taking advantage of the semiconductive and porous properties of MIL-125(Ti). CoPi and Pt are deposited into MIL-125(Ti) in two steps. The co-catalysts CoPi and Pt not only act as reactive sites for oxygen and hydrogen evolution, respectively, but also improve the photogenerated charge separation efficiency. The above conclusions are supported by the photoelectrical and photophysical results.

Nowadays, shortage of fossil fuels and their unrestricted use have led to austere climate change.1 Photocatalytic water splitting has been extensively studied, as it provides an eco‐friendly method of producing hydrogen and oxygen, which could be used as renewable energy. However, few photocatalysts have been reported to be able to split water.2, 3, 4 To realize water splitting, the photocatalyst should have a suitable thermodynamic potential, that is, the bottom level of the conduction band has to be more negative than the reduction potential of H+/H2 (0 V vs. NHE), whereas the top level of the valence band has to be more positive than the oxidation potential of O2/H2O (1.23 V vs. NHE). Nevertheless, at least a band gap of 2.3 eV should be needed for a water splitting photocatalyst, owing to the overpotential of O2 and H2 evolution from water.5 Co‐catalysts are often used to lower the activation energy and decrease the O2 and H2 evolution overpotentials. To date, various co‐catalysts have been proposed to effectively improve the activity of photocatalytic H2 or O2 evolution from water,6, 7, 8 and Pt is considered the most suitable H2 evolution co‐catalyst due to its large work function,9, 10, 11 whereas IrO2, RuO2, and CoPi are regarded the best co‐catalysts for O2 evolution reaction.12, 13, 14, 15, 16, 17 Metal–organic frameworks (MOFs) are an attractive class of porous solids with large specific surface areas. MOFs are widely used in gas storage, separation, catalysis, thin films, magnetism, or drug delivery.18, 19 Some MOFs display semiconductive behavior and, therefore, represent a new type of promising photocatalyst.20, 21, 22, 23, 24 MOFs have also been used for different photocatalytic applications, for example, hydrogen evolution,25 CO2 reduction,26 pollutant degradation,27 and organic syntheses.28 However, to the best of our knowledge, most studies have focused on the half reaction of water splitting over MOFs.29, 30, 31, 32 In fact, most MOFs are able to split water thermodynamically (Table 1). The major obstacle to water splitting over MOFs maybe the absence of reactive sites, resulting in large overpotentials, as discussed above.

Table 1

HOMO and LUMO positions of some typical MOFs.

Name	HOMO [eV]	LUMO [eV]	Band gap [eV]	Ref.
MIL‐125(Ti)	2.40	−1.40	3.80	33, 34, 35
NH2‐MIL‐125(Ti)	2.40	−0.05	2.45	33, 34, 35
UiO‐66	2.07	−0.6	2.67	36
NH2‐UiO‐66	1.95	−0.8	2.75	37
PCN‐224(Zn)	1.25	−0.58	1.83	38
PCN‐222	1.35	−0.40	1.75	39
Al‐ATA	1.55	−1.15	2.70	30
Al‐TA	3.17	−0.58	3.75	30
MIL‐53(Fe)	2.42	−0.46	2.88	40

MOFs offer a tantalizing platform for depositing co‐catalysts to improve the photocatalytic performance, because of their porous structure and rich chemical functionality. In our previous work, we proved that photocatalytic overall water splitting over MOFs is possible by constructing appropriate reaction sites. H2‐evolution‐active Ni2+ was introduced into O2‐evolution‐active Al−ATA through the coordination between Ni2+ and −NH2 in ATA, and the obtained Al−ATA−Ni is able to split water into H2 and O2 under UV/Vis light irradiation.41 MIL‐125(Ti), consisting of Ti and terephthalic acid, is a typical MOF. MIL‐125(Ti) displays photocatalytic activities towards hydrogen evolution from water in the presence of sacrificial agents.42 In this work, Pt and CoPi are co‐deposited onto MIL‐125(Ti) in two steps, which would act as H2 and O2 reaction sites, respectively. As expected, the as‐prepared MIL‐125(Ti)‐CoPi‐Pt produces stoichiometric H2 and O2 evolution from pure water with no external bias. Photoelectrical results suggest that CoPi and Pt play essential roles in decreasing the overpotential of O2 and H2 evolution from water, respectively. This work may shed new light on developing new efficient overall water splitting MOFs based photocatalysts.

To determine the crystal structure and purity of the as‐prepared samples, the XRD patterns were measured. The XRD pattern of MIL‐125(Ti) shows that all the diffraction peaks are in good agreement with that previously reported,43 as presented by Figure 1 a. Thermogravimetric (TG) analysis further excludes the presence of amorphous TiO2 in MIL‐125(Ti) (Figure S1).The peak positions of MIL‐125(Ti)‐CoPi, MIL‐125(Ti)‐Pt, and MIL‐125(Ti)‐CoPi‐Pt are identical to that of MIL‐125(Ti), suggesting that Pt and CoPi introduction does not destroy the framework of MIL‐125(Ti). Only a higher diffraction intensity of MIL‐125(Ti)‐CoPi‐Pt is observed, which suggests an aging process of the whole framework structure during the deposition of CoPi and Pt. Figure 1 b shows the N2 adsorption–desorption isotherms at 77 K of MIL‐125(Ti) and MIL‐125(Ti)‐CoPi‐Pt. The isotherm displays the typical type‐IV isotherm with a typical H3 hysteresis loop. At relatively high (P/P 0) range, it shows high adsorption, indicating the presence of accumulation pores. The BET surface areas of MIL‐125(Ti) and MIL‐125(Ti)‐CoPi‐Pt are calculated to be 734.485 and 511.014 m2 g−1, respectively. The smaller BET surface area of MIL‐125(Ti)‐CoPi‐Pt compared to that of MIL‐125(Ti) suggests the successful deposition of Pt and CoPi into MIL‐125(Ti). Figure 1 c shows the DRS spectra of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt. The onset of MIL‐125(Ti) is located at 351 nm. This value is in good consistent with that previously reported.34 No obviously difference is observed after CoPi loading, whereas Pt loading slightly improves the absorption in the visible light region, owing to the surface plasmonic effect.44, 45

Figure 1

a) XRD patterns of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt. b) N2 adsorption–desorption isotherms of MIL‐125(Ti) and MIL‐125(Ti)‐CoPi‐Pt. c) DRS spectra of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt.

Photocatalytic overall water splitting was investigated over the as‐prepared MIL‐125(Ti)‐CoPi‐Pt under UV/Vis light irradiation. It can be seen clearly from Figure 3 a that hydrogen and oxygen is produced stoichiometrically from pure water without external bias. The evolution rates of H2 and O2 are 42.33 and 21.33 μL h−1 respectively. The results suggest that MIL‐125(Ti)‐CoPi‐Pt is an efficient photocatalyst towards overall water splitting. Besides, it is found that the deposition sequence of CoPi and Pt has an effect on the photocatalytic activity. When Pt is deposited first, the obtained MIL‐125(Ti)‐Pt‐CoPi displays a much lower activity than MIL‐125(Ti)‐CoPi‐Pt under the same experimental conditions (Figure S2). In addition, the gas evolution rate decreased after 3 h over MIL‐125(Ti)‐CoPi‐Pt under UV/Vis irradiation, suggesting a relative stability (Figure S3). Poor stability is very common for most MOFs, and the framework structure is easily destroyed during the photocatalytic process.21, 46, 47, 48

Figure 3

a) Linear sweep voltammograms and b) impedance analysis upon chopped UV/Vis light illumination of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt. The scan rate is 20 mV s−1.

The half reactions of water splitting were further investigated to check the roles of Pt and CoPi. As can be seen from Figure 2 b, the oxygen evolution rate is improved more than twofold after CoPi deposition, from 143 to 307 μL h−1. Meanwhile, Figure 2 c shows the H2 generation rate is almost 25 times higher over MIL‐125(Ti)‐Pt than that over MIL‐125(Ti) under the same experiment conditions. These results are in good agreement with the reported results that Pt or CoPi can greatly improve the photocatalytic half reaction of water splitting.

Figure 2

a) Photocatalytic overall water splitting over MIL‐125(Ti)‐CoPi‐Pt. Reaction conditions: 30 mg of catalyst, 30 mL water. b) Photocatalytic O2 evolution reaction over MIL‐125(Ti) and MIL‐125(Ti)‐CoPi. Reaction conditions: 30 mg of catalyst, 30 mL water, 30 mg AgNO3. c) Photocatalytic H2 evolution reaction over MIL‐125(Ti) and MIL‐125(Ti)‐Pt. Reaction conditions: 30 mg of catalyst, 24 mL water, 6 mL CH3OH. Light source, Xe lamp (300 W).

The linear sweep voltammograms are investigated in order to reveal the role of co‐catalysts, and the results are shown in Figure 3 a. The onset potential for oxygen generation shifts negatively from 1.38 to 1.25 V after CoPi deposition, whereas the onset potentials for hydrogen evolution shift positively from −0.45 to −0.15 V after loading with Pt. MIL‐125(Ti)‐CoPi‐Pt displays both the cathodic and anodic potential shift of MIL‐125(Ti)‐CoPi and MIL‐125(Ti)‐Pt. All of the results confirm our assumption that CoPi and Pt could effectively reduce the overpotential of the oxygen and hydrogen evolution reactions, respectively.

Figure 3 b shows the impedance analyses (Nyquist plots) of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt under UV/Vis light illumination. The Nyquist plots of the samples consist of one dominant semicircle, whose diameter is related to the charge‐transfer resistance at the sample/electrolyte interface. The larger the diameter of the semicircle is, the larger the charge‐transfer resistance will be.49 As can be seen, the charge‐transfer resistance of MIL‐125(Ti) is the largest, and it can be reduced by either Pt or CoPi, suggesting faster photogenerated charge‐transfer kinetics after the MIL‐125(Ti) was modified with Pt or CoPi.

The properties of the excited state of the samples are further investigated by photoluminescence (PL) spectra. It is well accepted that the peak intensity corresponds to the separation capacity of photogenerated charge carriers.50 The higher the intensity is, the lower the possibility that photogenerated charge carriers participate in the photocatalytic process. Figure 4 suggests that the as‐prepared samples display a maximum emission at 468 nm, which is ascribed to the π–π* transition of the benzene ring.51 The maximum intensity is as follows the following order: MIL‐125(Ti) > MIL‐125(Ti)‐CoPi > MIL‐125(Ti)‐Pt > MIL‐125(Ti)‐CoPi‐Pt. The results imply that MIL‐125(Ti)‐CoPi‐Pt should display the highest photocatalytic activity, which is in agreement with both the photocatalytic and photoelectrical results.

Figure 4

PL spectra of MIL‐125(Ti), MIL‐125(Ti)‐Pt, MIL‐125(Ti)‐CoPi, and MIL‐125(Ti)‐CoPi‐Pt excited at 300 nm.

Based on the above results and discussions, a mechanism was proposed to elucidate the enhanced performance of photocatalytic water splitting. As illustrated in Scheme 1, upon excitation of photons with energy larger than 3.5 eV [energy gap between HOMO and LUMO orbital of MIL‐125(Ti)], electrons and holes are generated. The photogenerated electrons migrate towards Pt, while the photogenerated holes migrate towards CoPi. Through this way, the photogenerated charge carriers are spatially separated. After that, the electrons and holes participate in reduction and oxidation reactions, respectively. Pt and CoPi not only act as reactive sites, but also suppress the electron–hole recombination, as suggested by the PL results.

Scheme 1

Proposed photocatalytic mechanism of MIL‐125(Ti) for photocatalytic overall water splitting.

In summary, we have proven the assumption that photocatalytic overall water splitting over MOFs can be realized upon depositing of appropriate co‐catalysts. MIL‐125(Ti)‐CoPi‐Pt produces stoichiometric H2 and O2 from pure water under UV/Vis light irradiation. The co‐catalysts (CoPi and Pt) not only can reduce the overpotential of the half reactions of water splitting, but also improve the efficiency of photogenerated electron–hole separation, as suggested by the photoelectrical results. This work may provide a common way to realize water splitting over MOFs by taking advantage of their porous structure.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.