Brought to you by:
Paper

Ag-decorated GaN for high-efficiency photoreduction of carbon dioxide into tunable syngas under visible light

, , , , , , and

Published 14 October 2021 © 2021 IOP Publishing Ltd
, , Citation Wei Huang et al 2021 Nanotechnology 32 505722 DOI 10.1088/1361-6528/ac28d7

0957-4484/32/50/505722

Abstract

Visible light-driven photoreduction of CO2 and H2O to tunable syngas is an appealing strategy for both artificial carbon neutral and Fischer–Tropsch processes. However, the development of photocatalysts with high activity and selectivity remains challenging. For this case, we here design a hybrid catalyst, synthesized by in situ deposition of Ag crystals on GaN nanobelts, that delivers a tunable H2/CO ratio between 0.5 and 3 under visible light irradiation (λ > 400 nm). The obtained photocatalyst delivers a maximal turnover frequency value of 3.85 h–1 and a corresponding yield rate of 2.12 mmol h–1 g–1 for CO production, while the photocatalytic activity keeps stable during five cycling tests. Additionally, syngas can be detected even at λ > 600 nm. Experiments and mechanistic studies reveal that the existence of Ag crystals not only extends the light absorption region but also promotes the charge transfer efficiency, and thereby leading to a photocatalytic improvement. Accordingly, the present work affords an opportunity for developing an efficient photo-driven system by using solar energy to alleviate CO2 emissions.

Export citation and abstract BibTeX RIS

1. Introduction

The photoconversion of gaseous CO2 (that is, photocatalytic CO2 reduction reaction (pCO2RR)) offers an underlying option to mitigate the excessive emission of anthropogenic CO2 [15]. Thus far, great advancements have been achieved towards converting CO2 to C1, C2, or C2+ products through pCO2RR by using solar energy [612]. Among them, syngas (H2/CO mixture), is a key chemical feedstock for the industrially adopted Fischer–Tropsch (FT) route [1315], which has been traditionally obtained by coal gasification and natural gas reforming [16]. Also, it should be noted that different CO/H2 ratio is required for the different downstream product [9, 15]. Therefore, given the significance of syngas, the synthesis of tunable syngas via pCO2RR is highly desirable.

Although various semiconductor pCO2RR catalysts have been developed [1, 6, 7, 15], most studies are only focus on metal oxides (including TiO2, Fe2O3 and ZnO). It has been documented that CO2 molecules are typically adsorbed on oxygen vacancies of metal oxide surfaces [9, 17, 18]; however, at the same time, the adsorption of O2 in solution is also strengthened, and then fill the vacancy sites, leading to an inert surface [17]. Recent studies demonstrated that GaN is a highly promising candidate for solar fuel applications due to its wide direct transition bandgap (3.42 eV) along with high carrier mobility, thermal conductivity and nontoxicity [17, 1926]. In comparison with most reported metal oxides, the conduction band minimum of GaN is more negative, and hence it is kinetically more probable for the reduction of stable CO2 [17, 27]. Meantime, GaN has also shown relatively high stability during harsh photocatalysis reactions with more considerable corrosion resistance [27]. In addition, the built-in electric field induces an enhanced charge separation efficiency [28]. Nevertheless, the wide bandgap of GaN limits its use to effectively harvest a wide range of solar spectrum. Currently, the construction of heterojunction by introducing plasmonic metals (e.g. Au and Ag) is an effective strategy to tune electronic and photocatalytic properties of the GaN-based photocatalysts [22, 2730]. The resulting Mott–Schottky effect will significantly promote the transfer of photogenerated electrons from the semiconductors to the metal nanoparticles [31]. Meanwhile, the loaded Au and Ag metals can capture the irradiation energy, resulting in high energy electrons on their surface via the localized surface plasmon resonance effects [22, 32]. Of the plasmonic metals studied, Ag has been widely investigated for photocatalysis due to its relatively low cost and high electrical conductivity [33]. To date, however, the exploration of Ag/GaN hybrid photocatalyst that can precisely control the H2/CO ratio is still lacking, which largely hinders their further industrial applications.

Inspired by the pioneering progress and our previous work [8, 1012], we design an Ag crystals-decorated GaN nanobelts photocatalyst that is capable of efficiently producing syngas with a wide range of H2/CO ratio (0.5–3) under visible light irradiation (λ > 400 nm). Notably, the generated syngas can further detect under light irradiation up to 600 nm. The obtained maximal turnover frequency (TOF) value and the corresponding yield rate is 3.85 h−1 and 2.12 mmol h−1 g−1 for CO production, respectively. This performance is comparable to the recently reported catalysts (see details in table S1 (available online at stacks.iop.org/NANO/32/505722/mmedia)), such as porous metallic and magnetic Co–C composite [34], SnS2/SnO2 [35], Cu/Pt/TiO2 [36], Pt/C–In2O3 [37], and In2S3–CdIn2S4 [38]. Interestingly, the catalyst shows pCO2RR activity toward syngas production in low­-concentration CO2 atmosphere. Moreover, it also shows impressive pCO2RR stability during five cycling tests. Detailed experimental characterizations confirm that the introduced Ag not only extends the light absorption region but also promotes the charge transfer efficiency, both of which enable more electrons to participate in the conversion of CO2 to CO, resulting in improved pCO2RR activity and selectivity.

2. Experimental section

2.1. Chemicals

AgNO3, sodium citrate solution, NaBH4, CH3CN, [Ru(bpy)3]Cl2·6H2O, Triethanolamine (TEOA) and acetonitrile were bought from Sigma-Aldrich. The GaN nanobelts were purchased from Hefei Kejing Material Technology Co., Led. (http://kjmti.com/). These chemicals were used without further purification.

2.2. Synthesis of Ag/GaN

Ag/GaN was prepared by a deposition–reduction method, which has been reported in our previously reported work [39]. Firstly, 50 mg GaN was added into a beaker containing deionized water (50 ml) and stirred at room-temperature for 30 min. Subsequently, 0, 3, 5, 7 or 10 ml AgNO3 solution (0.01 M) and 10 ml sodium citrate solution (0.01 M) were added drop-wise to the above solution under stirring for 1 h, respectively. Finally, samples with different amounts of Ag loadings were obtained, and denoted as Ag/GaN-x, in which x denotes the mass fraction of Ag.

2.3. Characterizations

X-ray diffraction (XRD) patterns were performed with a Bruker D8 diffractometer with Cu–Kα radiation at a scan rate of 5° min−1 in the range of 10°–80°. Room-temperature UV–visible (UV–vis) diffuse reflectance spectroscopy was recorded on a PerkinElmer Lambda 950 UV–vis–NIR spectrophotometer. The field-emission-gun scanning electron microscopy (SEM) instruments (Verios 460L of FEI) were applied in the SEM characterization with an accelerating voltage of 15 kV. Transmission electron microscopy (TEM) and energy dispersive spectroscopy (EDX) images were achieved via TEM instruments (namely JEOL JEM-2100) at an accelerating voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) results were recorded by a Kratos AXIS Ultra DLD system with the Al Kα radiation as the x-ray source. Meanwhile, the C1s peak has been fixed at the binding energy of 284.8 eV. Fluorescence emission decay spectra were measured with a full-featured steady-state transient spectrum analyzer spectrometer (FLS980, edinburgh instrument).

2.4. Electrochemical measurements

The photocatalytic CO2 reactions were performed in a 50 ml reactor with 300 W Xenon lamp irradiation through the quartz cover (>400 nm, Beijing Perfect Light, PLS-SXE300D, China) with temperature maintained at 30 °C. The whole system consists of [Ru(bpy)3]Cl2·6H2O (0.005 mmol), Ag/GaN-x (0.5 mg) and solvent [10 ml in total, CH3CN/TEOA/H2O = 3:1:1 (volume ratio)]. Before the experiment, the reactor was evacuated with a vacuum pump, and then CO2 gas was aerated until the reaction pressure reached 1.8 bar.

During electrochemical tests, the concentration of gaseous samples was measured using an on-line gas chromatography (GC, 7890B Agilent) equipped with a flame ionization detector and a thermal conductivity detector. The liquid products were analyzed through 1H spectrum using a nuclear magnetic resonance (NMR, AVANCE AV III 400 Bruker) equipment, and dimethylsulfoxide was used as an internal standard for quantification. In addition, the possible liquid product was also analyzed by high-performance liquid chromatography (HPLC, Agilent 1260 Infinity II) using 1 mM sulfuric acid as eluent. In the work, no liquid product can be detected.

The isotopic labeling experiment was conducted under the same condition except using 13CO2 and gas chromatography-mass spectrometry (GC-MS, QP2020 Micropacked column). The TOF is defined as the number of moles of product produced per mole of catalytic sites and per hour in the reaction.

3. Results and discussion

The details for the synthesis of hybrid photocatalysts are described in the experimental section. Here, four samples with different Ag mass loadings on GaN were synthesized and denoted as Ag/GaN-x (x = 0.8, 1.7, 2.2 or 3.5). As depicted in SEM and TEM images (figures 1(a) and (b)), GaN displayed a nanobelt-like morphology with a width of 100–500 nm and length of 1–10 μm. After in situ depositing Ag crystals on surface GaN nanobelts, a hybrid photocatalyst (namely Ag/GaN-3.5) were obtained, as confirmed by the TEM image (figure 1(c)). The particle size of Ag crystals was measured to be about 10–30 nm.

Figure 1.

Figure 1. Morphological characterizations of Ag/GaN-3.5. (a) SEM and (b) TEM images of pristine GaN. (c) TEM image and (d), (e) HRTEM images of Ag/GaN-3.5. (f) EDX mapping of Ga, N, and Ag in Ag/GaN-3.5.

Standard image High-resolution image

As revealed by the high-resolution TEM (HRTEM) image (figures 1(d) and (e)), the lattice fringes with a distance of 0.272 and 0.233 nm were observed, which correspond to the (100) and (111) crystallographic planes of GaN and Ag, respectively. EDX mapping analysis also confirmed the deposition of Ag nanoparticles on the GaN surface (figure 1(f)).

XRD pattern of Ag/GaN-3.5 was shown in figure 2(a). Interestingly, all the diffraction peaks are well-matched with the hexagonal GaN (JCPDS No. 50-0792), confirming that the Ag deposition on GaN nanobelts is well-dispersed with a low loading amount. The Ag/GaN-3.5 sample was examined by Raman spectroscopy (figure S1), which yields two peaks at 561.7 and 712.2 cm–1, corresponding to the E2 and A1 modes of GaN [40], respectively. The elemental compositions and valence states of Ag, Ga, and N of Ag/GaN-3.5 were then examined by XPS. The survey spectrum (figure S2) verified the presence of Ag, Ga and N in Ag/GaN, which agrees with the EDX results (figure 1(f)). The high-resolution Ag 3d XPS spectrum of Ag/GaN demonstrated two peaks centered at binding energies of 367.1 and 373.2 eV that index to the metallic state of Ag, ruling out the existence of Ag2O or other silver oxides (figure 2(b)) [41]. In figure 2(c), the spin–orbit constituent of Ga 2p could be deconvoluted into two peaks with binding energies of 1117.2 and 1144.1 eV assigned to the 2p3/2 and 2p1/2 valence state of Ga3+, respectively. This is consistent with previous work [40]. It should also be noted that there is a negative shift of 0.5 eV for Ga 2p in Ag/GaN-3.5 in comparison to pristine GaN, suggesting the strong electronic interaction between the Ag and GaN [41]. Figure S3 presents the N 1s peak located at 397.3 eV, which could be ascribed to the N–Ga bond. In addition, the peak at 395.5 eV was originated from the Ga Auger LMM transitions. As shown in UV–vis spectra (figure 2(d)), increased absorption in the visible range was observed owing to the surface plasmon resonance effect of Ag, which is beneficial to the utilization of low photonic energy long-wavelength light [42].

Figure 2.

Figure 2. Structural analyses of GaN and Ag/GaN-3.5. (a) XRD patterns. (b), (c) High-resolution XPS spectra for Ag 3d and Ga 2p. (d) UV–vis spectra.

Standard image High-resolution image

We next investigated the pCO2RR activity of the as-obtained Ag/GaN-x in an airtight gas circulation system (figure S4) under visible-light irradiation (λ > 400 nm) with [Ru(bpy)3]Cl2·6H2O as the photosensitizer and TEOA as the electron donor [42, 43]. According to GC, NMR, and HPLC measurements, CO and H2 were confirmed to be the only gaseous products from CO2 photoreduction (figures S5–S7). When replacing CO2 with Ar, a much fewer amount of H2 was detected, implying that CO was generated from pCO2RR. In addition, when no [Ru(bpy)3]Cl2 was added, Ag/GaN-3.5 achieved low CO and H2 yield rates of 0.01 and 0.005 μmol g–1 h–1, respectively (figure S8(a)). Furthermore, in the case of only [Ru(bpy)3]Cl2 being added, no CO or H2 can be detected (figure S8(b)). Noticeably, as shown in figures 3(a) and (b), the yield rates, ratios and selectivities for CO and H2 products depend strongly on the amount of Ag. As the Ag loading increased as 0, 0.8, 1.7, 2.2 to 3.5 in Ag/GaN-x, the CO yield rate reached 0.13, 1.16, 1.65, 1.89 to 2.12 mmol g–1 h–1, respectively. At the same time, the corresponding H2 yield rate was determined to be 0.38, 2.27, 2.48, 1.85, and 1.04 mmol g–1 h–1, respectively. In this case, the H2/CO ratios for the above five photocatalysts were measured to be 2.92, 1.96, 1.50, 0.98, and 0.49, respectively. This concomitant production of tunable H2/CO mixture (from 0.5 to 3) would provide the possibility to synthesize a wide range of value-added chemicals through the FT process [15]. Of note, the achieved maximal yield rates for H2 and CO are comparable to the best results reported in the literatures (table S1). The above results also demonstrated that the deposited Ag is the main active site for CO2 photoreduction, which matches well with previous studies [12, 44]. To further improve the pCO2RR performance of GaN-based materials, strategies including the introduction of defects, the construction of hierarchical morphology, and the combination of suitable p-type semiconductors have been developed. In this regard, the TOF values of Ag/GaN-x were calculated to evaluate the catalytic efficiency of Ag depositions (figure 3(c)). The results show that GaN, Ag/GaN-0.8, Ag/GaN-1.7, Ag/GaN-2.2, and Ag/GaN-3.5 gives a TOF value of 0.69, 2.61, 3.05, 3.32 and 3.85 h–1, respectively.

Figure 3.

Figure 3. pCO2RR performance of Ag/GaN-x under visible-light irradiation. (a) The yield rates and H2/CO ratios and (b) the corresponding selectivity for H2 and CO production. (c) Calculated TOF values. (d) Isotope labeling experiment by using 13CO2 as feedstock. (e) The yield rates and H2/CO ratios on Ag/GaN-3.5 under irradiation with different wavelengths. (f) Stability test of Ag/GaN-3.5.

Standard image High-resolution image

To verify the carbon source of produced CO, an isotopic experiment with 13CO2 was then performed, as depicted in figure 3(d). As seen, a signal at m/z = 29 was observed, which could be ascribed to the generated 13CO. This strongly proved that CO was originated from photocatalysis of CO2 via pCO2RR. More interestingly, Ag/GaN-3.5 still displayed high pCO2RR activity for syngas production with a CO yield rate of 0.09 mmol g–1 h–1 and a corresponding H2/CO ratio of ∼1:2 when the long-wavelength light (λ > 600 nm) was applied (figure 3(e)). It should be emphasized that the usage of long-wavelength light plays an important role in promoting the utilization efficiency of solar energy [13, 42, 44].

The stability of Ag/GaN-3.5 was checked by five sequential CO2 photoreduction experiments under visible-light irradiation (λ > 400 nm). As indicated in figure 3(f), the yield rates for CO and H2 remained stable during the test, suggesting the good reusability of Ag/GaN-3.5. Furthermore, the post-reaction catalyst was characterized by XRD result (figure S9), both of which confirmed the compositional and structural stability of Ag/GaN-3.5.

Previous studies indicated that the conversion of low­-concentration CO2 is of great importance for the practical application of pCO2RR technology [43, 45]. However, to date, very few photocatalysts can effectively adsorb and activate low­-concentration CO2 molecules [43]. Interestingly, our designed Ag/GaN-3.5 exhibited a catalytic performance toward syngas production in a 10% CO2 atmosphere (the typical CO2 concentration in flue gas [46]) with corresponding yield rates of 4.05 and 1.99 μmol g–1 h–1 for CO and H2 production, respectively (figures S10 and S11).

To uncover the underlying reasons for the improved pCO2RR activity of Ag/GaN-3.5, the electrochemical impedance spectra were first explored, as shown in figure 4(a). Ag/GaN-3.5 showed a smaller semicircle radius compared to GaN under visible-light irradiation, suggesting the faster charge transfer kinetics and thereby enabling more photoexcited electrons to participate in photocatalysis CO2 [42]. This result was also confirmed by the increased photocurrent density observed on Ag/GaN-3.5 (figure 4(b)). Meanwhile, the photoluminescence (PL) spectra in figure 4(c) manifested that a suppressed recombination of the electron–hole pairs occurred on Ag/GaN-3.5, indicating an improved efficiency for charge transfer and separation resulted from the Ag deposition. Further, the Mott–Schottky plots (figure S12) of GaN and Ag/GaN-3.5 showed two positive slopes, implying Ag/GaN-3.5 still maintain an n-type characteristic [33]. Moreover, it is noted that the formation of heterojunctions in Ag/GaN-3.5 increased the conduction band position, which promoted the separation and migration of photogenerated carriers, and then induced a higher pCO2RR performance [31].

Figure 4.

Figure 4. Mechanistic insights. (a) Nyquist plots. (b) Photocurrent-time profiles. (c) PL spectra. (d) Schematic illustration of photocatalytic CO2 reduction for Ag/GaN-3.5.

Standard image High-resolution image

It has been well recognized that Ag possessed a stronger adsorption affinity toward CO2 in comparison with H2O, enabling it exhibits higher activity for CO2 photoreduction [12, 47, 48]. Meantime, previous studies have shown that GaN is extremely reactive for H2 production directly from H2O under UV and visible light illumination [20, 27]. Taking into these considerations and the above observations and analyses, a possible mechanism for photoreduction of CO2 and H2O to tunable syngas on Ag/GaN has been proposed [13, 2226, 49], as depicted in figure 4(d). Under visible-light irradiation, the photosensitizer [Ru(bpy)3]2+ was photoexcited and oxidized to [Ru(bpy)3]3+, producing a larger amount of electron. Then, these photoelectrons are transferred to Ag/GaN surface, where CO2 and H2O molecules are adsorbed. Some fraction of the generated photoelectrons were used to reduce CO2 to CO on Ag sites, while another fraction of ones on GaN catalyzed the coupling of H+ to form H2. After that, [Ru(bpy)3]3+ was reduced to the initial state [Ru(bpy)3]2+ by TEOA. By flexibly adjusting the Ag loading amount in Ag/GaN photocatalyst, it is easy to control the H2/CO ratios in syngas, which is highly desirable for potential applications.

4. Conclusions

In conclusion, a hybrid photocatalyst of well-defined Ag crystals anchoring on GaN nanobelts has been prepared and was evaluated for CO2 photoreduction. As a result, the obtained composites can produce syngas with a widely tunable H2/CO ratio (from 0.5 to 3) under visible-light irradiation (λ > 400 nm). The maximal TOF and yield rate for CO formation was measured to be 3.85 h–1 and 2.12 mmol g–1 h–1, respectively, achieved by Ag/GaN-3.5. Interestingly, the catalyst also showed activity toward syngas synthesis (with an H2/CO ratio of 1:2) under light irradiation up to 600 nm. Moreover, Ag/GaN-3.5 demonstrated feasibility for utilization of low-concentration CO2 (i.e. flue gas). Besides, it could remain a high activity for five sequential CO2 reductions. Detailed experimental measurements clarified that the Ag deposition not only increases the absorption at longer wavelengths but also promotes the charge transfer and separation efficiency. Meantime, the construction of dual active sites toward CO2 and H2O adsorption synergistically induced remarkable improvement in pCO2RR activity. In short, this work provides a valuable guideline for the development of efficient and robust GaN-based photocatalysts for syngas synthesis by using clean solar energy. Future investigation includes the development of GaN nanobelts for the conversion of CO2 to valuable fuels under visible light irradiation.

Acknowledgments

This work was financially supported by National Natural Science Foundation of China (6202780115) and the Science and Technology Development Fund of Tianjin Education Commission for Higher Education (2018KJ126).

Data availability statement

All data that support the findings of this study are included within the article (and any supplementary files).

Authorship contribution statement

Wei Huang and Dejin Zhou: Investigation, Data curation, Visualization, Validation, Writing—original draft. John Lee, Jiaqiang Sun, Shusheng Zhang and Hong Xu: Methodology, Resources, Funding acquisition. Jun Luo and Xijun Liu: Conceptualization, Methodology, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Please wait… references are loading.
10.1088/1361-6528/ac28d7