Controlled Construction of Copper Phthalocyanine/α‐Fe₂O₃ Ultrathin S‐Scheme Heterojunctions for Efficient Photocatalytic CO₂ Reduction under Wide Visible‐Light Irradiation

Abstract

The wide visible‐light‐driven CO₂ reduction to acquire solar fuels is a highly desired green route. Herein, novel ultrathin copper phthalocyanine (CuPc)/α‐Fe₂O₃ heterojunctions as efficient wide visible‐light‐driven photocatalysts for CO₂ reduction are controllably synthesized by the hydroxyl‐induced self‐assembly of CuPc onto ultrathin α‐Fe₂O₃ as‐pre‐prepared through an Al³⁺‐regulated hydrothermal method. The optimized CuPc/Fe₂O₃ heterojunction exhibits about 15‐fold high photoactivity for reducing CO₂ to CO and CH₄ compared with reported Fe₂O₃ nanoparticles. The exceptional photoactivity is mainly attributed to the enhanced S‐scheme charge transfer and separation in the resulting closely contacted heterojunction, the extended visible‐light range from molecularly disperse CuPc, and its provided central metal cation (Cu²⁺) with favorable catalytic function for CO₂ activation, mainly by means of the dual‐wavelength photocurrent action spectra, the electrochemical reduction tests, and the in situ diffuse reflectance infrared Fourier transform spectra (DRIFTS). This investigation provides new insight about designing and constructing novel metal phthalocyanine (MPc)‐involved S‐scheme heterojunction photocatalysts.

Novel copper phthalocyanine (CuPc)/α‐Fe₂O₃ ultrathin S‐scheme heterojunctions as efficient wide visible‐light‐driven photocatalysts for CO₂ reduction are controllably synthesized by a hydroxyl inducing assembly process. The optimized CuPc/Fe₂O₃ heterojunction exhibits about 15‐fold high photoactivity for reducing CO₂ to CO and CH₄ compared with the reported Fe₂O₃ nanoparticles.

1. Introduction

As the principal product of fossil fuel consumption, CO₂ emissions beyond the recycling capacity of nature has caused the greenhouse effect, which seriously impacts on the climate environment and the development of human society.^{[ 1 ]} Learning from nature, photocatalytic conversion of CO₂ into value‐added products by the semiconductors has attracted extensive attention, as a promising eco‐friendly and low‐cost technology.^{[ 2} , ^{3 ]} Hematite (α‐Fe₂O₃) as a classic metal–oxide–semiconductor offers significant potential for photocatalysis, owing to its narrow bandgap (2.0–2.2 eV), chemical stability, low‐cost and environmental friendliness, etc.^{[ 4} , ⁵ , ^{6 ]} Especially, the excellent oxidation ability of photogenerated holes (h⁺) for water oxidation is an important semi‐reaction during the photocatalytic CO₂ reduction process as usual.^{[ 7 ]} However, its low reduction potential of photogenerated electrons (e⁻) deriving from the low conduction band leads to insufficiently competent in CO₂ reduction reaction by itself.^{[ 8 ]} Moreover, the short migration distance of holes in α‐Fe₂O₃ (less than 4 nm) seriously inhibits the charge separation; meanwhile, unsatisfactory surface catalytic sites should also be overcome.^{[ 4} , ^{9 ]} Therefore, the development of highly efficient α‐Fe₂O₃‐based photocatalysts for CO₂ reduction is full of opportunities and challenges.

In general, to improve the photocatalytic performance of α‐Fe₂O₃, morphology control is considered to be one of the most effective strategies.^{[ 10} , ¹¹ , ^{12 ]} Compared with the common 0D and 1D morphology, the 2D especially for the ultrathin 2D materials (lateral size larger than 100 nm and the thickness typically less than 5 nm) exhibits unprecedented physical, electronic, chemical, and optical properties due to the electron confinement in 2D.^{[ 13 ]} In prospect of ultrathin 2D α‐Fe₂O₃, the photogenerated holes are much easier to migrate to the surfaces, and a larger specific surface area with more catalytic sites is exposed, which are favorable to improve the charge separation and then enhance the photocatalytic activities. Hence, it is meaningful to design a feasible method to synthesize ultrathin α‐Fe₂O₃. Gao and co‐workers reported a facile Al³⁺‐intervened hydrothermal method for large‐scaled preparation of cell‐unit‐thick single‐crystalline hematite (α‐Fe₂O₃) nanosheets with a thickness of 1.3 nm, and then, they were converted to magnetic maghemite (γ‐Fe₂O₃) for utilization in Li‐ion battery.^{[ 14 ]} Undoubtedly, it is not only necessary but also feasible to study the relationship between thickness and photocatalytic performance. However, few works related to ultrathin α‐Fe₂O₃‐based photocatalysts have been reported to date.

In addition, how to overcome the drawback of over‐positive conduction band on α‐Fe₂O₃ is still a challenge. In recent years, according to the band structure characteristics of α‐Fe₂O₃, construction of Z‐scheme nanocomposite based on α‐Fe₂O₃ as the oxidization photocatalyst (OP) with an efficient reduction photocatalyst (RP) has been considered as one of the most effective tactics to overcome the bottleneck of insufficient reduction ability of photogenerated electrons on α‐Fe₂O₃ for improving its photocatalytic performance.^{[ 15} , ¹⁶ , ^{17 ]} Wong and co‐workers constructed a hierarchical Z‐scheme α‐Fe₂O₃/g‐C₃N₄ hybrid with a significantly improved performance for photocatalytic CO₂ reduction.^{[ 18 ]} Based on the generally accepted Z‐scheme heterojunction system, Yu and co‐workers proposed a Step‐scheme (S‐scheme) one, and the charge separation at the interface between RP and OP could be further boosted by an internal electric field (IEF) directing from RP to the OP due to the matched Fermi level.^{[ 19} , ^{20 ]} Thus, choosing a proper RP with a matched band structure and Fermi level to construct the ultrathin 2D α‐Fe₂O₃‐based S‐scheme heterojunctions as efficient photocatalysts is feasible by forming a strong interaction interface for promoting the charge transfer and separation. In particular, it is highly desired to endow lots of catalytic sites such as nanoclusters or single atoms for CO₂ reduction.^{[ 21 ]}

Compared with the widely used g‐C₃N₄ as the RP, metal phthalocyanine (MPc) as a class of visible‐light response organic photocatalyst is a metal‐central and macrocyclic molecule with a planar conjugated array of 18‐p electrons.^{[ 22} , ^{23 ]} Especially, the high lowest unoccupied molecular orbital (LUMO) and suitable highest occupied molecular orbital (HOMO) energy levels make it possibly used as the RP in the S‐scheme heterojunction photocatalysts.^{[ 24 ]} Moreover, various central metal ion with M‐N₄ structures in different MPc could play as a catalytic site, similar to the single atom catalysis.^{[ 25 ]} It has been demonstrated in our previous work that the zinc phthalocyanine (ZnPc) is used as RP to successfully construct the Z‐scheme nanocomposite with BiVO₄.^{[ 26 ]} Remarkably, compared with BiVO₄ (E _g ≈ 2.4 eV), α‐Fe₂O₃ has a wider spectral response range (λ ≤ 600 nm) and a higher matching degree with common MPc materials (550 ≤ λ ≤ 750 nm) in terms of solar spectral utilization, which is expected to achieve effective utilization of full visible spectrum. In addition, it is desired to confirm that the positive positions of conduction band and related Fermi level are favored to build a profitable interface driving force by the possible IEF between the α‐Fe₂O₃ and MPc. Moreover, it is important to select an MPc with appropriate center metal.^{[ 27 ]} As various studies have shown that Cu species has a special performance in the catalytic reaction of CO₂ reduction, it is anticipated that copper phthalocyanine (CuPc) modification would improve the photocatalytic performance of CO₂ reduction of α‐Fe₂O₃ under wide visible‐light irradiation.^{[ 28} , ^{29 ]}

In this work, we fabricate a wide visible‐light‐driven S‐scheme CuPc/Fe₂O₃ ultrathin heterojunction by a simple hydroxyl inducing assembly process for efficient CO₂ reduction for the first time. In the assembly process, the pure‐phase ultrathin α‐Fe₂O₃ with abundant surface hydroxyl groups induces the highly dispersed CuPc to uniformly assemble on it. Moreover, mechanisms of S‐scheme charge separation and catalytic process are clarified by systematic characterizations. This work provides a vital strategy for the further development of a new type of α‐Fe₂O₃‐based S‐scheme heterojunction photocatalysts with excellent performance.

2. Results and Discussion

2.1. Optimized Synthesis of Ultrathin α‐Fe₂O₃

It is meaningful to investigate the effects of additional aluminum ions (Al³⁺) on the properties of as‐prepared α‐Fe₂O₃, because Al³⁺ is the crystal face regulator in the synthetic process.^{[ 14 ]} First, crystal phase compositions of different xUTFO UTFO = ultrathin α‐Fe₂O₃, x = the molar ratio percentage of Al³⁺ to Fe³⁺ in the reactants, and x = 0.25, 0.5, and 0.75, respectively), FO (without addition of Al³⁺), and Fe₂O₃ nanoparticle (NPFO) samples were characterized by X‐ray diffraction (XRD) patterns. As shown in Figure S1a, Supporting Information, it is noticed that some impurities are formed in FO and 0.75UTFO samples, but 0.25UTFO and 0.5UTFO with pure phase of α‐Fe₂O₃ (JCPDS No. 33‐0664) can be obtained in this case. Especially for the 0.5UTFO sample, the relative intensity ratio of the (110) to the (104) facet peak is three times higher than that of the as‐prepared NPFO, because that Al³⁺ ions can be adsorbed on the (001) surface, leading to the preferred growth along [100] direction.^{[ 14} , ^{31 ]} Moreover, a blue shift appears on the absorption edges of 0.5UTFO comparing with other samples via the UV–vis diffuse reflectance spectra (DRS), as shown in Figure S1b, Supporting Information, which is attributed to the quantum size effect.^{[ 32 ]}

Scanning electron microscopy (SEM) images were further used to analyze the morphologies of as‐prepared α‐Fe₂O₃ (Figure S2, Supporting Information), and one can see that the resulting FO without additive Al³⁺ is irregular particle shape (Figure S2a, Supporting Information), whereas the 2D nanosheet morphology could be obtained by introducing Al³⁺. Noticeably, the 0.5UTFO (Figure S2c, Supporting Information) shows a thinner and smoother nanosheet structure compared with 0.25UTFO (Figure S2b, Supporting Information) and 0.75UTFO (Figure S2d, Supporting Information), respectively. In addition, the ultrathin nanosheet‐like structure of the 0.5UTFO is further confirmed by transmission electron microscopy (TEM) image (Figure S3a, Supporting Information). The high‐resolution TEM (HRTEM) image shows clear lattice fringes with an interfringe spacing of 0.25 and 0.16 nm (Figure S3b, Supporting Information), matching well with the lattice distance of (110) and (211) planes of hexagonal α‐Fe₂O₃, respectively.^{[ 14 ]} The selected area electron diffraction (SAED) pattern as the inset further reveals that the nanosheets are hexagonal single crystals with (001) orientation, which is consistent with the XRD results.^{[ 33 ]} Moreover, the atomic force microscopy (AFM) image (Figure S3c, Supporting Information) demonstrates that the thickness of 0.5UTFO is about 1.4 nm, which agrees well with the reported result,^{[ 14 ]} and as shown in Figure S4a,b, Supporting Information, the 0.25UTFO (≈3.1 nm) is slightly more than twice as thick as 0.5UTFO.

Subsequently, the NPFO as a referential sample was prepared according to our previous reported method, and then, the photocatalytic activities of the as‐prepared xUTFO and NPFO samples for CO₂ reduction were evaluated under visible‐light irradiation (λ > 420 nm). As shown in Figure S3d, Supporting Information, 0.5UTFO shows the best activity, which is nearly four times higher than that for as‐prepared NPFO. In the inset of Figure S3d, Supporting Information, the surface photovoltage spectroscopy (SPS) responses confirm that the ultrathin structure of 0.5UTFO is indeed more favorable for charge separation. Moreover, the charge separation related to hydroxyl radical fluorescence spectra (FS) is tested, when the photocatalyst with enough water oxidation capacity is excited in the aqueous environment, and it will produce hydroxyl radicals as captured by the coumarin to produce 7‐hydroxycoumarin, whose fluorescence signal, hence, can be quantified to reflect the charge separation situation of as‐tested sample. In general, a larger signal intensity means a higher photogenerated charge separation. As shown in Figure S5a, Supporting Information, the results of FS are consistent with SPS response. As expected, the 0.5UTFO sample also exhibits the highest photocatalytic activity for phenol degradation under visible‐light irradiation (Figure S5b, Supporting Information). The above‐mentioned results clearly indicate that the controllably synthesized ultrathin α‐Fe₂O₃ could exhibit the good charge separation and then the improved photocatalytic performance.

2.2. Effects of Modifying CuPc on Ultrathin α‐Fe₂O₃

Based on the XRD patterns of the 0.5UTFO and yCuPc/UTFO, as shown in Figure S6, Supporting Information, no detectable characteristic peaks of CuPc are observed, but only pure phase of α‐Fe₂O₃ for all samples modified with different contents of CuPc, which maybe attributed to the small amount of CuPc with uniform dispersion on the surface of 0.5UTFO. The DRS of the 0.5UTFO and yCuPc/UTFO is shown in Figure  1a, and the results indicate that the introduction of CuPc does not change the absorption band edge of 0.5UTFO, while endows a strong characteristic absorption in the region of 550–750 nm.^{[ 26} , ^{34 ]} As expected, light absorption intensity is increasing with the amount of CuPc. Noticeably, the thickness of 1.5CuPc/UTFO is ≈1.8 nm, which is 0.4 nm thicker than 0.5UTFO according to the AFM images, as shown in Figure 1b, so it suggests that monodispersed CuPc molecules are assembled on the 0.5UTFO. Moreover, the high angle annular dark field scanning transmission microscopy (HAADF‐STEM) (Figure 1c) and the related energy dispersive X‐ray analysis (EDX) mapping (Figure 1d) further confirm that the CuPc is indeed evenly dispersed on the 0.5UTFO ultrathin nanosheets. From the viewpoint of geometric matching (2D/2D), it is accessible that CuPc modifying on UTFO can form an ultrathin heterojunction, and possibly brings with a strong interface interaction.

Figure 1.

a) UV–vis DRS of 0.5UTFO and yCuPc/UTFO, b) AFM image and the corresponding height profile as the inset of 1.5CuPc/UTFO, c) HAADF‐STEM, and d) the corresponding EDX mapping images of elemental Fe, O, Al, and N of 1.5CuPc/UTFO (y stands for the mass ratio percentage of CuPc to 0.5UTFO, y = 1, 1.5, and 2, respectively).

To confirm the elemental chemical states and the possible strong interface interaction between the CuPc and the 0.5UTFO, X‐ray photoelectron spectroscopy (XPS) was measured. As shown in Figure  2a, the O 1s spectra of 0.5UTFO and 1.5CuPc/UTFO samples are fitted into two peaks located at 529.5 and 531.6 eV, which are attributed to crystal lattice oxygen and hydroxyl oxygen, respectively.^{[ 17 ]} Noticeably, the intensity of the hydroxyl oxygen peak at 531.6 eV obviously decreases after be modified by such a small amount of CuPc. As shown in Figure 2b, the binding energy (BE) of Fe 2p element with characteristic two main peaks located at 712.0 and 725.8 eV is attributed to Fe³⁺ in α‐Fe₂O₃,^{[ 35 ]} and the BE has a negative shift after CuPc modification. Meanwhile, the BE of N 2p in the 1.5CuPc/UTFO has a positive shift comparing with pure CuPc (Figure 2c), whereas a slightly positive shift of Cu 2p BE is observed, as shown in Figure 2d. Based on the above‐mentioned XPS analysis, it suggests that a hydroxy group related close interface interaction between the CuPc with 0.5UTFO is formed in the ultrathin heterojunction composite and associated with a process of electron transfer from CuPc to 0.5UTFO in the interface.

Figure 2.

XPS for a) O 1s, b) Fe 2p of 0.5UTFO and 1.5CuPc/UTFO, and for c) N 1s and d) Cu 2p of pure CuPc and 1.5CuPc/UTFO.

The effect of CuPc modification on the charge separation of 0.5UTFO is explored by the hydroxyl radical‐related fluorescence spectra (FS). As shown in Figure  3a, the FS signal intensity of 0.5UTFO is significantly increased by introducing an appropriate amount of CuPc, and especially, the one for the 1.5CuPc/UTFO is approximately four times as higher as 0.5UTFO. Moreover, the photoelectrochemical (PEC) measurement further supports the FS results (Figure 3b). The photocatalytic activity of 0.5UTFO is significantly improved after introducing CuPc, and the optimized 1.5CuPc/UTFO heterojunction composite could achieve nearly threefold enhancement (Figure 3c), which is ≈15 times higher than the NPFO; meanwhile, a certain amount of oxidation product O₂ is also produced, as shown in Figure S7, Supporting Information. Based on the previous results within our group, it is deduced that the generated CO, CH₄, and O₂ are all derived from H₂O and CO₂.^{[ 26 ]} In addition, the 1.5CuPc/UTFO sample exhibits good stability for CO₂ reduction reaction after five cycles (Figure 3d).

Figure 3.

a) Fluorescence spectra related to the formed hydroxyl radicals, b) PEC I–V curves, c) photocatalytic activities for CO₂ conversion under visible light irradiation for 4 h of 0.5UTFO and yCuPc/UTFO, and d) photocatalytic cycling test of 1.5CuPc/UTFO.

2.3. Discussion on Mechanism

To reveal the mechanism of enhanced charge separation and improved photocatalytic activities on the as‐prepared ultrathin heterojunction composite, the following experiments were designed. According to the energy band alignment of CuPc and UTFO, the S‐scheme charge transfer mechanism is possible. For this, the SPS responses of 0.5UTFO and 1.5CuPc/UTFO are test in a N₂ atmosphere (Figure  4a,b). Based on the working principle of SPS that when the photocatalyst is excited by the irradiation of light, the photogenerated electrons and holes would reach the surface of material. When the electrons reach the surface and then are captured by the trapping agent (such as oxygen), there is a potential difference on the surface of the photocatalyst, which generates an SPS signal. When nitrogen is used to remove oxygen in the air, SPS signals will not be generated, because surface photogenerated electrons will not be captured. Under this condition, SPS signals can be generated by constructing a channel, which can quickly transfer electrons on the surface of the material, and then, constructing a heterojunction has this function. As shown in Figure 4a, there are no detectable SPS responses for both 0.5UTFO and 1.5CuPc/UTFO, indicating no charge transfer and separation under monochromatic beam probing. However, with an additional 660 nm monochromatic excitation beam, the 1.5CuPc/UTFO exhibits a detectable SPS response in the range of 300–600 nm. Similarly, when the 1.5CuPc/UTFO is exposed to an additional 520 nm monochromatic excitation beam, an SPS response can be detected when probing from 600 to 750 nm. According to the UV–vis DRS spectra (Figure 1a), the light absorption of introduced CuPc is between the range of 550 and 750 nm, whereas 0.5UTFO absorbs light at wavelength smaller than 600 nm. Thus, it is clear that a charge separation‐related SPS response can be detected only when both of CuPc and 0.5UTFO are excited by the excitation and probing beams at the same time, naturally conforming to an S‐scheme charge transfer mechanism.

Figure 4.

SPS responses of UTFO and 1.5CuPc/UTFO in N₂ atmosphere assisted with a) 660 nm and b) 520 nm monochromatic light beam. c) Normalized photocurrent action spectra of UTFO and 1.5CuPc/UTFO under different monochromatic light irradiation. d) Photocatalytic activities for CO₂ conversion under different monochromatic beam irradiation for NPFO, UTFO, and 1.5CuPc/UTFO.

The proposed S‐scheme mechanism is further confirmed by the monochromatic photocurrent action spectra for the samples of 0.5UTFO and 1.5CuPc/UTFO. As shown in Figure 4c, nearly no photocurrent is detected over the 0.5UTFO under the excitation of monochromatic light with a wavelength longer than 580 nm, but an obvious photocurrent is recorded at 580 nm and increased with the decrease in the excitation wavelength. In construct, obvious photocurrent of 1.5CuPc/UTFO is detected under 700 and 660 nm irradiation, which is mainly attributed to the charge separation from the excitation of CuPc, whereas the photocurrent intensity is nearly equal to 0.5UTFO, as the wavelength continues to decrease, because that only the 0.5UTFO can be excited. Noticeably, the photocurrent density of 1.5CuPc/UTFO is dramatically increased when the wavelength is shorter than 580 nm and with continuous irradiation under additional 660 nm monochromatic excitation beam. These results are in good agreement with the above‐mentioned results of the single‐wavelength‐assisted SPS response. Then, the photocatalytic CO₂ reduction was performed under single wavelength irradiation and assisted by a single wavelength. As shown in Figure 4d, for the pure 0.5UTFO sample, the activity under 520 nm single wavelength light irradiation is not significantly changed under the condition of simultaneous introduction of 660 nm assisted light, whereas the activity of the 1.5CuPc/UTFO sample under dual‐wavelength light irradiation is significantly improved, which is about four times higher than that of the 0.5UTFO sample. Therefore, its activity result also proves the S‐scheme mechanism of heterojunction composite.

To verify the catalytic function of the coordination metal in CuPc, the electrochemical reduction tests were carried over 0.5UTFO, 1.5H₂Pc/UTFO, and 1.5CuPc/UTFO samples under the N₂‐ or CO₂‐bubbling systems, as shown in Figure  5a,b. The results show that 0.5UTFO and 1.5H₂Pc/UTFO exhibit similar onset potentials in the N₂‐ and CO₂‐bubbling systems, whereas 1.5CuPc/UTFO exhibits a most positive reduction potential, especially in the presence of CO₂, indicating that the central metal in CuPc has a crucial catalytic effect on the CO₂ reduction reaction. Moreover, a series of controllable experiments are designed to further confirm the key role of Cu²⁺ in catalyzing the reduction of CO₂, as shown in Figure S8, Supporting Information. The SPS responses and FS tests are carried on 0.5UTFO, 1.5H₂Pc/UTFO, and 1.5CuPc/UTFO. The SPS responses and FS spectra both follow the orders: 0.5UTFO<1.5H₂Pc/UTFO<1.5CuPc/UTFO (Figure S8a,b, Supporting Information), which is in agreement with the photocatalytic CO₂ reduction activities (Figure S8c, Supporting Information) and phenol degradation activities (Figure S8d, Supporting Information). This indicates that the central metal (Cu²⁺) of CuPc could catalyze the reduction reaction of CO₂. Furthermore, the CO₂ and H₂O adsorption capacities of CuPc as an important factor to photocatalytic CO₂ conversion were investigated by in situ diffuse reflectance infrared Fourier transform spectra (DRIFTS) at room temperature. As shown in Figure 5c,d, the sample was purged first with CO₂ gas in advance of passing through H₂O for 30 min, and then with high purity N₂ for 1 h. The experimental data of the in situ DRIFTS spectra were obtained by subtracting the background spectrum. As shown in Figure 5c, several vibration bands can be observed in the range of 2280–2390 cm⁻¹, which are ascribed to the asymmetric stretching vibration of CO₂.^{[ 36 ]} The characteristic peaks of CO₂ over 1.5CuPc/UTFO are higher than that of over 0.5UTFO, whereas more efficient elimination of the adsorbed CO₂ over 1.5CuPc/UTFO than 0.5UTFO is observed under visible‐light irradiation for 5 or 15 min, indicating that the introduction of CuPc is favorable for CO₂ adsorption and followed photocatalytic reduction reaction. Similarly, the peaks in the range of 3550–3750 cm⁻¹ are allocated to the surface O—H stretching of H₂O,^{[ 37 ]} which are also eliminated more completely over 1.5CuPc/UTFO with more adsorbed water than 0.5UTFO. These results suggest that the excellent CO₂ reduction performance of 1.5CuPc/UTFO is not only attributed to the good CO₂ adsorption performance of CuPc, but also owing to the efficient water oxidation process.

Figure 5.

Electrochemical reduction curves of 0.5UTFO, 1.5H₂Pc/UTFO, and 1.5CuPc/UTFO in a) N₂‐ and b) CO₂‐bubbled systems. In situ DRIFTS of adsorbed c) CO₂ and d) H₂O on UTFO (black line) and 1.5CuPc/UTFO (red line) (solid line: in dark for 30 min, and dot line: under visible‐light irradiation for 5 and 15 min).

In addition, an effective S‐scheme charge separation between the CuPc and the 0.5UTFO should depend on the strong driving force from IEF. For this, the scanning kelvin probe (SKP) (Figure  6a) and Mott–Schottky plots (Figure S9, Supporting Information) were performed on 0.5UTFO, CuPc, and 1.5CuPc/UTFO, respectively. It is demonstrated that IEF direction from CuPc to 0.5UTFO is formed after close contact due to the different work functions, as shown in Figure 6b,c, which also explains the results of XPS (Figure 2). As a result, IEF is conducive to driving S‐scheme charge transfer in the process of photocatalytic reaction. Finally, the schematic diagram of photocatalytic reaction process is shown in Figure 6d. In this heterojunction system, the CuPc and 0.5UTFO can absorb two completely complementary spectra in the range of whole visible light, and then, the S‐scheme charge separation is driven by IEF, so as to boost the reaction of CO₂ reduction on CuPc with Cu²⁺ as a catalytic center, whereas H₂O oxidation occurs on UTFO. In addition, the important role CuPc is also feasible to be replaced with other transitional MPc, such as FePc and CoPc, and the as‐prepared 1.5CuPc/UTFO heterojunction exhibits the best photocatalytic activity for CO₂ conversion (Figure S10a, Supporting Information), due to the highest charge separation (Figure S10b, Supporting Information) and the best catalytic ability for CO₂ reduction (Figure S10c,d, Supporting Information).

Figure 6.

a) SKP maps of 0.5UTFO, CuPc, and CuPc/UTFO. Band diagram and Fermi levels of CuPc and 0.5UTFO b) before and c) after contacting. d) S‐scheme mechanism of photogenerated charge transfer and subsequently induced main reactions in the fabricated CuPc/UTFO heterojunction.

3. Conclusion

In summary, novel ultrathin CuPc/UTFO heterojunction with a Janus structure has been fabricated by a facile hydroxyl‐induced self‐assembly process, which exhibits excellent visible‐light photocatalytic activity for CO₂ reduction. This is mainly originated from the S‐scheme charge transfer driven by the strong IEF and sufficient utilization of wide visible‐light spectrum. The optimized 1.5CuPc/UTFO exhibits the best activity with 15 times higher than the as‐prepared NPFO, which is attributed to the synergistic effects of the better charge separation and preferable catalytic role of central Cu²⁺ centers of CuPc. In addition, this strategy is applicable to other MPcs for the heterojunction construction. Therefore, this work not only develops a facile synthetic strategy for wide‐spectrum Fe₂O₃‐based photocatalysts, but also provides a highly competitive RP for construction of novel S‐scheme heterojunction photocatalysts with wide visible‐light responsive for efficient solar fuel production.

4. Experimental Section

4.1.

4.1.1.

Synthesis of Ultrathin Fe₂O₃ Nanosheets

The ultrathin α‐Fe₂O₃ nanosheets (UTFO) were fabricated based on the previous report.^{[ 14 ]} Typically, 1 mmol of Fe(NO₃)₃·9H₂O and 0.5 (0, 0.25, or 0.75) mmol of Al₂(SO₄)₃·18H₂O were dissolved in 10 mL of deionized water under magnetic stirring to form a transparent solution. Subsequently, 3 mL of triethylamine was added into the transparent solution and keep stirring for 40 min. Then, the mixed solution was transferred and sealed in a 50 mL Teflon‐lined autoclave, which was maintained at 160 °C for 24 h. The red precipitates at the bottom of autoclave were collected by centrifugation, washed by water and ethanol for several times, and then dried at 60 °C. After that, the powders were heated at 200 °C for 1 h in muffle furnace. The obtained samples were denoted as xUTFO (x = the molar ratio of added Al³⁺ to Fe³⁺, x = 0.25, 0.5, and 0.75, respectively), and the one without added Al³⁺ was denoted as FO.

Synthesis of yCuPc/UTFO Heterojunctions

A series of yCuPc/UTFO samples were synthesized using a simple assembly method. Typically, 7.5 mg (5 or 10 mg) of CuPc powder was dissolved into 100 mL dimethyl formamide (DMF) by sonication for 0.5 h and then stirring for 24 h to obtain the CuPc solution. Subsequently, 0.5 g of 0.5UTFO was added into the above‐mentioned CuPc solution; then, the mixture was dispersed in ultrasound for 30 min and keep stirring for 6 h. Finally, the dark red sample was collected by centrifugation, washed by water and ethanol for three times, and then dried at 80 °C for 12 h. The obtained samples were denoted as yCuPc/UTFO (y = the mass ratio percentage of the added CuPc to 0.5UTFO, y = 1, 1.5, and 2, respectively).

Synthesis of NPFO

The NPFO was fabricated based on the previous report.^{[ 30 ]} Typically, 10 mL of 5 wt% ammonia solution was added into a 50 mL Teflon‐lined autoclave, and a 10 mL weighing bottle containing a solution of 0.8 g Fe(NO₃)₃ in 8 mL n‐butylalcohol was placed in the autoclave with a support between the autoclave bottom and the weighing bottle. The Teflon‐lined autoclave was kept at 140 °C for 6 h. After being cooled naturally to the room temperature, the sample was washed for several times with ethanol and deionized water, in turn, and then dried in an oven at 80 °C to obtain α‐Fe₂O₃. Finally, the samples were calcined in air at 400 °C for 2 h.

Characterization Techniques

Powder XRD was performed on a Bruker D8 advance diffractometer with Cu Kα1 radiation (λ = 1.5406 Å). UV–vis DRS were recorded on a Shimadzu UV2700 spectrophotometer. Morphologies and element mappings of samples were performed in SEM (ZEISS Sigma 500), TEM (JEOL‐F200) with an energy‐dispersive X‐ray spectrometric microanalysis (EDS) unit, and an AFM (Bruker MultiMode 8). The XPS analysis was recorded on a Kratos‐AXIS ULTRA DLD system with the monochromatic Al Kα X‐rays, and the BE was calibrated internally by the adventitious C 1s peak at 284.4 eV. The steady‐state SPS measurement of the samples was recorded by a home‐built apparatus; the signals were recorded on a phase‐locked amplifier (SRS, SR830) synchronized with a light chopper (SR540), and the monochromatic light was obtained with passing light from a 500 W xenon lamp across a double prism monochromator (SBP300).

In situ DRIFTS were collected from 650 to 4000 cm⁻¹ at a spectral resolution of 4 cm⁻¹ with 64 scans on a FTIR spectrometer (Nicolet is 50) equipped with a high‐sensitive mercury cadmium telluride (MCT) detector cooled by liquid N₂ and a DRIFT cell (Harrick) with three windows including an specialized irradiation window. Each sample was first put in a drying oven at 110 °C for 2 h, and then was purged with high purified N₂ (30 mL min⁻¹) at room temperature for 1 h in the in situ tank, to eliminate the physical adsorbed water and other impurities; then, the background spectra were recorded. First, 30 mL min⁻¹ pure CO₂ was introduced into water, and then, the mixture was through the sample surface for 1 h; the spectra, were recorded with subtracting from the background spectrum in dark conditions. Second, the CO₂ and water were turned off, and 30 mL min⁻¹ pure N₂ was introduced, and the spectra were recorded with subtracting from the background spectrum in simulated sunlight. Thus, the obtained in situ DRIFTS could actually reflect the vibration peaks of the group on the catalyst surfaces in different atmosphere and at different times.

Conflict of Interest

The authors declare no conflict of interest.

Data Availability Statement

Research data are not shared.

Supporting information

Supplementary Material