Electrons and Hydroxyl Radicals Synergistically Boost the Catalytic Hydrogen Evolution from Ammonia Borane over Single Nickel Phosphides under Visible Light Irradiation.

From the perspective of tailoring the reaction pathways of photogenerated charge carriers and intermediates to remarkably enhance the solar-to-hydrogen energy conversion efficiency, we synthesized the three low-cost semiconducting nickel phosphides Ni2P, Ni12P5 and Ni3P, which singly catalyzed the hydrogen evolution from ammonia borane (NH3BH3) in the alkaline aqueous solution under visible light irradiation at 298 K. The systematic investigations showed that all the catalysts had higher activities under visible light irradiation than in the dark and Ni2P had the highest photocatalytic activity with the initial turnover frequency (TOF) value of 82.7 min-1, which exceeded the values of reported metal phosphides at 298 K. The enhanced activities of nickel phosphides were attributed to the visible-light-driven synergistic effect of photogenerated electrons (e-) and hydroxyl radicals (.OH), which came from the oxidation of hydroxide anions by photogenerated holes. This was verified by the fluorescent spectra and the capture experiments of photogenerated electrons and holes as well as hydroxyl radicals in the catalytic hydrogen evolution process.

Introduction

In consideration of the global environment problems, hydrogen energy has become a promising alternative to fossil fuels.1, 2 The efficient storage and release of hydrogen are main barriers during developing the hydrogen economy based on fuel cells.3, 4 Ammonia borane (NH3BH3) with low toxicity and high hydrogen content (19.6 wt%) is an excellent chemical hydrogen storage material.5, 6, 7 Up to now, a variety of catalysts based on non‐noble metals such as Ni and Co have been explored for the hydrogen evolution from NH3BH3 (NH3BH3+2H2O→NH4BO2+3H2), which is a thermodynamics‐controlled process.8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 However, their activities still remain low compared with the noble metal catalysts.19, 20, 21, 22, 23, 24, 25, 26, 27 Thus, it is highly desirable to explore new catalytic systems using low‐cost and efficient catalysts for hydrogen evolution.

Compared with the thermocatalysis, the visible‐light‐driven photocatalysis is considered as a friendly route for solar‐to‐chemical energy conversion.28, 29, 30 Importantly, the photogenerated charge carriers in semiconductors can promote many redox reactions including hydrogen evolution from small molecules such as water.31, 32 For photocatalytic splitting of NH3BH3, the key is to explore low‐cost catalytically active semiconductors. It has been shown that non‐precious metal phosphides composed of P and earth abundant metals such as Ni have been used in water splitting as cocatalysts in photocatalysis and catalysts in electrocatalysis.33, 34 However, no attention has been paid to using these metal phosphides as photocatalysts without photosensitizers or photogenerated electron acceptors such as metal nanoparticles probably due to that they have narrow band gaps and thus have no enough redox potential to directly split water.35 Different from the water splitting with non‐spontaneous characteristics, NH3BH3, which has weak B−N (∼117 kJ ⋅ mol−1) and B−H (∼430 kJ ⋅ mol−1) bonds,25, 36 is relatively easy to split though the catalytic reaction. In addition, photogenerated charge carriers such as electrons and highly active groups such as hydroxyl radicals, which are often used to promote chemical reactions, can benefit the cleavage of B−H and B−N bonds in the photocatalytic splitting of NH3BH3. Bearing these aspects in mind, we considered that if semiconducting metal phosphides such as NixPy are directly used as catalysts under visible light irradiation, the efficiency of hydrogen evolution from NH3BH3 will be enhanced through the synergistic effect of photogenerated electrons and hydroxyl radicals.

Herein, we reported a series of low‐cost metal phosphides Ni2P, Ni12P5 and Ni3P with different structures, which were for the first time used as single catalysts for hydrogen evolution from NH3BH3 in the alkaline aqueous solution under visible light irradiation at 298 K. Compared with the activities of all the catalysts in the dark, their visible‐light‐driven activities were enhanced and Ni2P had the highest activity. Moreover, the photocatalytic hydrogen evolution mechanism was also discussed on the basis of the related experiments.

Results and Discussion

Synthesis Strategy

Visible‐light‐responsive materials with low cost were selected to catalyze hydrogen evolution from NH3BH3 since they can harvest visible light and then produce separated charge carriers including holes and electrons, which can promote the catalytic reaction.37 Among the catalytically active species, nickel phosphides were selected as photocatalysts on the basis of three considerations. Firstly, nickel phosphides are active components for catalytic hydrogen evolution from NH3BH3, but their activities are still low. Secondly, the photogenerated charge carriers induced by the visible light might benefit the enhanced catalytic reaction rate of nickel phosphides. Thirdly, three nickel phosphides Ni2P, Ni12P5 and Ni3P with different structures, where there are different electron transfer characteristics from Ni to P,38, 39 provided us a chance to systematically study the photocatalytic chemistry of metal phosphides. Since OH− ions in the aqueous solution are easily oxidized by photogenerated holes from semiconductors to form highly reactive hydroxyl radicals.40 The inorganic base NaOH was selected as the resource of hydroxyl radicals, which were beneficial for the cleavage of B−N bonds in NH3BH3. Once the holes were consumed by OH− ions, the remaining photogenerated electrons could efficiently participate in the splitting of NH3BH3. In the light of the above contents, the visible‐light‐driven synergistic effect of photogenerated electrons and hydroxyl radicals could lead to the efficient catalytic hydrogen evolution from NH3BH3 in the alkaline aqueous solution over the nickel phosphides with different structures (Figure 1).

Figure 1

Schematic illustration of the catalytic hydrogen evolution procedure over single semiconducting nickel phosphides with different structures under visible light irradiation.

Structural Characterization

The crystalline phases of as‐synthesized nickel phosphides were characterized using PXRD. The results showed that the pure phases of Ni2P, Ni12P5 and Ni3P were synthesized (Figure S1–S3). The surface areas of Ni2P, Ni12P5 and Ni3P were 33.2, 11.8 and 20.9 m2/g (Figure S4–S6). The TEM images showed that the particles of Ni2P, Ni12P5 and Ni3P were in the nanoscale (Figure 2 and S7). The lattice fringes with spacing of 0.221, 0.193 and 0.180 nm corresponded to the (111) plane of Ni2P, (240) plane of Ni12P5 and (222) plane of Ni3P. The XPS investigations showed that the Ni 2p3/2 edges in the three nickel phosphides were deconvolved into three peaks (Figure 3). The peaks at 853.0, 852.7 and 852.4 eV were assigned to the positive charge Niδ+ for Ni2P, Ni12P5 and Ni3P, respectively.22, 41, 42 The value of δ was δ(Ni2P)>δ(Ni12P5)>δ(Ni3P). The peaks at 855.9, 856.0 and 855.9 eV were consistent to the Ni2+ for Ni2P, Ni12P5 and Ni3P,43 respectively, due to the surface oxidation. The peaks at 861.3, 860.9 and 860.5 eV were assigned to the satellites of Ni 2p3/2. The spectra of Ni 2p1/2 were similar to those of Ni 2p3/2. In the P 2p3/2 spectra (Figure S8), the peaks at 128.9, 129.1 and 129.3 eV confirmed the presence of Pδ− for Ni2P, Ni12P5 and Ni3P,44 respectively, indicating that there were different characteristics of charge transfer from Ni to P in the three nickel phosphides with different structures.

Figure 2

TEM and HRTEM mages of (a, c) Ni2P and (b,d) Ni12P5.

Figure 3

XPS patterns of Ni 2p for (a) Ni2P, (b) Ni12P5 and (c) Ni3P.

In the UV‐vis spectra, Ni2P, Ni12P5 and Ni3P had the visible light absorption (Figure S9). The steep absorption edge suggested the electron transfer from valence band to conduction band. In order to verify the electron transfer of the nickel phosphides, we studied their transient photocurrent density under visible light irradiation (λ≥420 nm). The results showed that the transient photocurrent appeared in the nickel phosphides and Ni2P had the highest photocurrent density (Figure 4a). These results also confirmed that the three nickel phosphides had semiconducting characteristics and Ni2P had higher catalytic reduction ability. It should be noted that the photocurrent response of Ni2P and Ni3P were positive while the photocurrent response of Ni12P5 was negative, indicating that Ni2P and Ni3P were n‐type semiconductors and Ni12P5 was p‐type semiconductor.45 These different semiconducting characteristics might lead to generating different charge carriers in the conductive process, which might further influence the photocatalytic properties of corresponding catalysts. In addition, the EIS experiments were carried out to get further insights into the hydrogen evolution ability of nickel phosphides. It was found that Ni2P and Ni3P had the smallest and biggest charge transfer resistance, respectively (Figure 4b), which was consistent with the corresponding transient photocurrent density.

Figure 4

(a) Time versus transient photocurrent density and (b) Nyquist plots of the three nickel phosphides.

Catalytic Performance and Mechanism

In order to investigate the effect of visible light irradiation on the catalytic performance, the hydrogen evolution from NH3BH3 in the alkaline aqueous solution over Ni2P, Ni12P5 and Ni3P was measured. As shown in Figure 5, the activities of three catalysts were enhanced under visible light irradiation compared with their activities in the dark. Specially, Ni2P exhibited the highest activity with the initial turnover frequency (TOF) value of 82.7 min−1, which was the highest in the values of reported metal phosphide catalysts (Table 1). The 87.5, 78.7 and 88.2 % of activity enhancement of Ni2P, Ni12P5 and Ni3P was caused by the contribution of visible light irradiation comparable to the dark reaction, respectively. The different activity enhancement could be attributed to the different light absorption ability and the different charge separation efficiency with increasing the Ni/P ratio from 2 to 3 in these nickel phosphides, which might lead to the different usage efficiency of photogenerated charge carriers and hydroxyl radicals. In addition, the characteristic of electron transfer from Ni to P in the three nickel phosphides with different Ni/P ratios was different (Figure 3), which led to their different photocatalytic hydrogen evolution activities. It should be noted that the crystalline phases of three nickel phosphides were maintained after the catalytic hydrogen evolution reaction (Figure S1–S3), indicating that these photocatalysts were stable in the present alkaline environment.

Figure 5

(a) Time versus volume of hydrogen evolution from the alkaline aqueous NH3BH3 over the three nickel phosphides under different conditions and (b) the initial TOF values.

Table 1

Activities of catalysts for hydrogen evolution from NH3BH3.

Catalyst	TOF (min−1)	Reference
Ni2P	82.7	This work
Ni0.8W0.2	25.0	6
Ni	19.6	14
Co/(CeOx)0.91/NGH	79.5	15
Cu0.72Co0.18Mo0.1	46	16
CoP	72.2	17
Ni0.7Co1.3P	58.4	18
Pd74Ni26/MCN	246.8	20
CuxCo1−xO−GO	70.0	24
Co/CTF	42.3	25
Co/C3N4‐580	93.8	37

The formation of radical intermediates under the light irradiation of photoactive materials in aqueous environment plays an important role in the photocatalytic reactions.46, 47 It is known that hydroxyl radicals can react with TA and generate TAOH, which emits fluorescence at around 426 nm. To investigate the effect of hydroxyl radicals, which generated in a process of oxidation of OH− or H2O by photogenerated holes, on the catalytic activity, we selected TA as fluorescence probe to trace the hydroxyl radicals. The results showed that the concentration of OH− affected the amount of produced hydroxyl radicals and there was different intensity of fluorescence peaks, which appeared at about 426 nm (Figure 6). More importantly, the photocatalytic hydrogen evolution rate increased with increasing the base concentration (Figure 7). In detail, Ni2P had the highest activity in 0.5 M of NaOH. However, its activity decreased when the alkaline concentration increased to 0.7 M. The high‐concentration base induced the formation of more hydroxyl radicals, which participated in the cleavage of chemical bonds in NH3BH3, while the excessive hydroxyl radicals impeded the adsorption of catalytic substrates NH3BH3 and H2O on the surface of catalyst. Besides, the base NaOH also acted as a catalyst promoter for the hydrolytic dehydrogenation of NH3BH3.16, 17, 18 This suggested that the formation of hydroxyl radicals benefited the catalytic activity enhancement under visible light irradiation. To further make sure the existence of hydroxyl radicals, we tested the photocatalytic activity of Ni2P, where 2‐propanol was chosen as the scavenger of hydroxyl radicals. The results showed the activity decreased due to the decrease of hydroxyl radicals (Figure 8), indicating that the formation of hydroxyl radicals was a positive factor for photocatalytic hydrogen evolution from NH3BH3.

Figure 6

Fluorescent spectra of Ni2P in (a) 10.0, (b) 30.0 and (c) 50.0 mM of alkaline aqueous solution and TA (5 mM) under visible light irradiation.

Figure 7

Time versus volume of hydrogen evolution from NH3BH3 in the aqueous solution with different alkaline concentrations over Ni2P under visible light irradiation.

Figure 8

Time versus volume of hydrogen evolution from the alkaline aqueous NH3BH3 over Ni2P without scavenger or in the presence of 2‐propanol, K2Cr2O7 and KI under visible light irradiation.

In the present catalytic system, the hydroxyl radicals generated through the oxidation of OH− ions by photogenerated holes.48, 49 In order to confirm this judgment, we measured the photocatalytic activity of Ni2P, where KI was used as capture agent of holes. The results showed that the catalytic activity decreased due to the decrease of holes (Figure 8), suggesting that the formation of holes had a positive effect on the catalytic reaction rate in the alkaline aqueous solution. Besides the photogenerated holes, the photogenerated electrons also played an important role in the catalytic reaction.50 To check this, K2Cr2O7 was selected as scavenger of electrons in the catalytic process. The results showed that the photocatalytic activity decreased after the introduction of K2Cr2O7 (Figure 8), indicating that the photocatalytic hydrogen evolution was remarkably affected by the photogenerated electrons.

In view of the above results regarding hydroxyl radicals and photogenerated charge carriers, a possible mechanism for photocatalytic hydrogen evolution from NH3BH3 in the alkaline aqueous solution was proposed (Figure 9). After the absorption of visible light by nickel phosphides, the photogenerated electrons and holes formed, separated and migrated to the catalyst surface. In the meantime, the hydroxyl radicals generated through the reaction between photogenerated holes and OH− ions. For the decomposition of NH3BH3 for hydrogen evolution, the cleavage of B−N bonds, which form through enjoying lone pair electrons between NH3 and BH3 groups, should first happen. On the basis of that the hydroxyl radicals benefited the enhancement of photocatalytic hydrogen evolution from NH3BH3 in the experiment, it could be rationally deduced that the hydroxyl radicals attacked NH3BH3 to break the B−N bond and the photogenerated electrons attacked the B−H bond to generate H2 with the assistance of water.

Figure 9

Possible mechanism for the hydrogen evolution from NH3BH3 in the alkaline aqueous solution over nickel phosphides under visible light irradiation.

To further understand the role of visible light irradiation in the activity enhancement of catalysts, the catalytic performance of Ni2P under different light intensity was investigated. It was found that increasing the light intensity resulted in an almost linear enhancement of catalytic activity (Figure 10). These phenomena could be attributed to that more electrons and hydroxyl radicals generated from Ni2P in the alkaline environment under high‐intensity visible light irradiation, leading to that more photogenerated electrons and hydroxyl radicals were used to attack and break B−H and B−N bonds in NH3BH3

Figure 10

(a) Time versus volume of hydrogen evolution from the alkaline aqueous NH3BH3 over Ni2P under visible light irradiation with different intensities and (b) the dependence of initial TOF values on the light intensity.

In the present reaction, water molecules participate in the hydrogen evolution.5, 51 In order to confirm this judgment, we used D2O instead of H2O to investigate the kinetic isotope effect (KIE) in the catalytic reaction over Ni2P, Ni12P5 and Ni3P. The results showed that the hydrolysis of NH3BH3 in D2O had a low reaction rate compared to that in H2O under light irradiation (Figure 11 and S10). The KIE constants were 2.9, 2.8 and 2.4 calculated according to the hydrogen evolution rate. This indicated that water molecules were involved in the hydrogen evolution from the aqueous NH3BH3, which was similar to the hydrogen evolution through the hydrolysis of NaBH4.52

Figure 11

Time versus volume of hydrogen evolution from NH3BH3 in H2O or D2O over Ni2P and Ni3P under visible light irradiation.

It is known that the hydrogen evolution from aqueous NH3BH3 is a thermodynamics‐controlled process.53 In order to evaluate the effect of reaction temperature on the photocatalysis, we measured the photocatalytic performance of Ni2P under the photothermal condition. The results showed that compared with the photocatalytic activity of Ni2P at 298 K, its activity was significantly enhanced under photothermal condition and its initial TOF value was 138.3 min−1 (Figure S11), which was much higher than the value (82.7 min−1) at room temperature.

The practical application of catalysts requires their good recycle durability. So the long‐time catalytic durability test of Ni2P under visible light irradiation was carried out at 298 K. The results showed that the catalyst had good durability and 100 % of H2 selectivity even after 20 cycles of catalysis (Figure 12a). After the catalytic cycles, the crystalline phase of Ni2P was unchanged (Figure 12b). With increasing the number of catalytic cycles, the rate of hydrogen evolution decreased, which might be due to the agglomeration of catalyst particles after the long‐time photocatalytic reaction (Figure 12c). In addition, the increase of boron species in the catalytic system might also lead to the gradual decrease of hydrogen evolution rate.5

Figure 12

Durability test for the hydrogen evolution from the alkaline aqueous NH3BH3 over Ni2P, (b) the PXRD pattern and (c) the TEM image of catalyst after 20 cycles of catalysis.

Conclusions

In summary, three semiconducting nickel phosphides Ni2P, Ni12P5 and Ni3P with different structures were synthesized. The compounds were used as single catalysts for hydrogen evolution from NH3BH3 under visible light irradiation. Among all the catalysts, Ni2P had the highest room‐temperature activity. In the catalytic process, the synergistic effect of photogenerated electrons and hydroxyl radicals in the alkaline aqueous condition enhanced the activities of nickel phosphides under visible light irradiation. This effect has been experimentally testified by the fluorescent spectra and the capture of photogenerated charge carriers and hydroxyl radicals as well as the related hydrogen evolution performance of the catalysts. Based on the obtained results, a new mechanism for the photocatalytic hydrogen evolution from NH3BH3 in the alkaline aqueous solution has been presented. This work sheds light on exploring low‐cost and visible‐light‐responsive semiconductors with well‐defined structures, which is beneficial for rationally designing new high‐performance photocatalytic systems toward a greener world using clean energy.

Experimental Section

Chemicals

Nickel chloride hexahydrate (NiCl2 ⋅ 6H2O, Sinopharm Chemical Reagent Co. Ltd, >99 %), sodium hypophosphite (NaH2PO2, Aladdin, 99 %), trisodium citrate dihydrate (Na3C6H5O7 ⋅ 2H2O, Sinopharm Chemical Reagent Co. Ltd, >99 %), phosphorus red (P, Aladdin, 99 %), sodium acetate (CH3COONa, Sinopharm Chemical Reagent Co. Ltd, >99 %), nickel acetate tetrahydrate (Ni(CH3COO)2 ⋅ 4H2O, Sinopharm Chemical Reagent Co. Ltd, >99 %), potassium dichromate (K2Cr2O7, Sinopharm Chemical Reagent Co. Ltd, 99.8 %), potassium hydroxide (KOH, Sinopharm Chemical Reagent Co. Ltd, >85 %), potassium iodide (KI, J&K Chemical, >99 %), sodium hydroxide (NaOH, Sinopharm Chemical Reagent Co. Ltd, >96 %), ammonia borane (NH3BH3, Aldrich, 97 %), terephtalic acid (C8H6O4, TA, J&K Chemical, 99 %), 2‐propanol ((CH3)2CHOH, J&K Chemical, 99.9 %) and deuterium oxide (D2O, J&K Chemical, 99.8 % atom D) were obtained without purification.

Synthesis and Catalytic Study

For the synthesis of Ni2P,22 the mixture of NiCl2 ⋅ 6H2O (1.00 g), Na3C6H5O7 ⋅ 2H2O (0.25 g) and NaOH (2.80 g) in water (45 mL) was stirred for 1 h to give a green solid. Then, the green solid (0.25 g) and NaH2PO2 (1.25 g) were heated at 573 K for 2 h in an Ar atmosphere to give Ni2P. For the synthesis of Ni12P5,39 the mixture of Ni(CH3COO)2 ⋅ 4H2O (0.95 g) and phosphorus red (0.70 g) was dissolved in water (30 mL) under stirring for 20 min. Then, the mixture was hydrothermally treated at 473 K for 12 h to give Ni12P5. For the synthesis of Ni3P,54 the mixture of NiCl2 ⋅ 6H2O (5.0 g), NaH2PO2 (24.4 g) and CH3COONa (2.9 g) was dissolved in water (100 mL). The KOH was used to adjust the solution pH to 8.0. Then the solution was heated at 363 K for 1 h to give a black solid. Later, the black solid was calcined at 673 K for 1 h in an Ar atmosphere. Finally, the product was washed in HCl solution, water and ethanol, resulting in Ni3P.

The catalytic reaction was performed under visible light irradiation at 298 K or in the dark at 298 K in the aqueous solution (3.0 mL) of NH3BH3 (1.71 mmol) and NaOH with different concentrations. The molar ratio of NixPy/NH3BH3 was 0.03 in the catalytic reaction. The reaction temperature was kept at 298 K through the cooling water circulating pump.

Catalyst Characterization

The transmission electron microscopy (TEM, JEM‐2010) was applied to confirm the morphologies of samples. The UV‐vis, the fluorescent and the surface area measurements were conducted on a Shimadzu UV‐3600 spectrometer, a FLS920 spectrometer and an Autosorb‐iQ2‐MP adsorption equipment, respectively. Powder X‐ray diffraction (PXRD) and X‐ray photoelectron spectroscopy (XPS) were performed using an X‐ray diffractmeter (Panalytical X‐Pert) and an instrument (ESCALAB250, Thermo VG Corp.). An electrochemical station (PGSTAT302N, Switzerland) was used to measure the transient photocurrent and electrochemical impedance spectroscopy (EIS) in the 1.0 M of Na2SO4 aqueous solution. An Ag/AgCl electrode, an indium tin oxide (ITO) glass covered with metal phosphides and a Pt plate were selected as the reference electrode, the working electrode and the counter electrode, respectively.

Calculation Method

The initial TOF value was calculated at the NH3BH3 conversion of 72 %. The calculation was shown as follow:

In the equation, nNH3BH3 is the molar amount of NH3BH3 at its conversion of 72 %, ncatalyst is the total molar amount of nickel phosphide and t is the time at the NH3BH3 conversion of 72 %.

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.