1. Design Principles
S-scheme heterojunction photocatalysts have similar charge transfer characteristics to direct Z-scheme heterojunctions, but there are some differences in design principles
[60,61,62][1][2][3]. Different from the Z-scheme photocatalysts, which are constructed with n-type semiconductors, the S-scheme photocatalyst consists of an oxidation semiconductor photocatalyst (OP) and a reduction semiconductor photocatalyst (RP). Therefore, the S-scheme heterojunction photocatalysts can be an n–n junction, p–n junction, n–p junction, and p–p junction as long as the fermi level and band structure are appropriate
[63,64,65][4][5][6]. In an S-scheme heterojunction photocatalyst, the OP needs to have a lower CB, VB, and fermi level than the OP, and the energy bands of the OP and RP are staggered . The difference in fermi levels causes the spontaneous transfer of some electrons from the RP to the OP at the contact surface, and the uneven charge distribution induces the built-in electric field from the RP to the OP. The energy bands near the interface are also bent by electrostatic repulsion. The carriers can transfer directionally under the built-in electric field, band bending, and coulomb force
[66,67][7][8]. This transfer way of carrier will promote the recombination of holes and electrons with weak redox ability and prolong the lifetime of the carriers with strong redox capabilities. Therefore, S-scheme heterojunction photocatalysts have enhanced redox ability.
Pure carbon nitride is usually an n-type semiconductor, and there are three types of carbon nitride-based S-scheme photocatalysts: n–n, p–n, and n–p type
[68,69,70][9][10][11]. The carbon nitride has quite a negative CB position, and the photogenerated electrons on it have strong reductive ability. Therefore, carbon nitride is typically used as an RP in S-scheme photocatalysts. Semiconductors whose CB, VB, and fermi levels are lower than carbon nitride can be used as the OP. When the carbon nitride-based S-scheme photocatalysts are excited by light, the electrons in the CB of the OP will recombine with the holes in the VB of carbon nitride under the coulomb force and built-in electric field. The photogenerated electrons in carbon nitride and the photogenerated holes in the OP can fully exert their strong redox capabilities
[71][12]. Zhang et al. prepared TpPa-1-COF/g-C
3N
4 S-scheme photocatalysts using a covalent organic framework and carbon nitride nanowires
[72][13]. The TpPa-1-COF has a lower CB, VB, and fermi level than g-C
3N
4 nanowires. The π-π conjugated heterointerface between the TpPa-1-COF and g-C
3N
4 has an enhanced interface electric field. The robust internal electric field makes the charge move quickly, which improves the charge separation and utilization of TpPa-1-COF/g-C
3N
4. Therefore, the TpPa-1-COF/g-C
3N
4 S-scheme heterojunction shows an excellent performance of photocatalytic hydrogen production (1153 μmol g
−1 h
−1). Chen et al. reported the IB/CNx S-scheme photocatalyst composed of iodine-doped BiOBr and nitrogen-deficient g-C
3N
4. The CB potential of IB is not negative enough, and the electrons on it have a low reduction ability. These electrons with a low reduction ability are driven by the internal electric field to recombine with the holes with a low oxidation ability on CNx. This charge transfer pathway leads to a prolonged lifetime of strong oxidizing holes on IB and strong reducing electrons on CNx. The IB/CNx S-scheme heterojunction has high-speed carrier migration and effective charge separation and exhibits enhanced photocatalytic activity in CO
2 reduction and tetracycline degradation
[73][14]. In addition, there are also some semiconductors whose CB, VB, and fermi level are higher than carbon nitride. At this time, carbon nitride can be used as the OP to construct S-scheme photocatalysts. However, the VB position of carbon nitride is not positive enough, and the oxidation capacity of this S-scheme heterojunction may be insufficient
[74][15].
2. Preparation of Carbon Nitride-Based S-scheme Photocatalysts
Carbon nitride is an organic polymer semiconductor that is easy to prepare and has stable physicochemical properties. There are many methods to construct S-scheme heterojunctions based on carbon nitride. The typical synthesis method of carbon nitride-based S-scheme photocatalysts includes coprecipitation, the high-temperature solid-state method, the hydrothermal method, the vapor deposition method, electrostatic self-assembly, the thermal polymerization method, etc.
[75,76,77][16][17][18]. Macyk et al. prepared TiO
2/C
3N
4/Ti
3C
2 composites with thermal polymerization and electrostatic self-assembly. The C
3N
4 layer is in situ grown on TiO
2 nanosheets using urea as a precursor, and then, Ti
3C
2 quantum dots are loaded on a core-shell TiO
2/C
3N
4 via electrostatic self-assembly. TiO
2/C
3N
4/Ti
3C
2 composite heterojunction photocatalysts follow the S-scheme charge transfer characteristics in CO
2 reduction and exhibit enhanced CO and CH
4 generation activity. The S-scheme charge transfer pathway between TiO
2 and C
3N
4 is beneficial to inhibit the strong redox electron and hole recombination. Ti
3C
2 can promote electron transfer and increase the CO
2 reduction rate
[78][19]. Chen et al. prepared bulk carbon nitride via thermal polymerization using melamine as a precursor and then added bulk carbon nitride to the solution containing V source and In source. The InVO
4 quantum dots/g-C
3N
4 S-scheme heterojunctions (InVO
4/CN) were finally synthesized via ultrasonic and hydrothermal cutting methods. The CO
2 reduction rate of 0D/2D hybrid InVO4/CN S-scheme photocatalyst reached 69.8 μmol g
−1 h
−1, and the content of CO in the product was as high as 93.3%. The InVO
4/CN obtained via this method had abundant defect sites, which were beneficial to the adsorption and activation of CO
2 [79][20].
Cui et al. prepared anatase TiO
2/H-doped rutile TiO
2/g-C
3N
4 (TiCN) double S-scheme heterojunction photocatalysts with the electrostatic self-assembly method. First, g-C
3N
4 nanosheets, anatase TiO
2 nanoparticles, and H-doped rutile TiO
2 nanorods are prepared separately, and then, the three materials are spontaneously assembled with electrostatic force in an ethanol–water mixture. Anatase TiO
2 nanoparticles and H-doped rutile TiO
2 nanorods are loaded on g-C
3N
4 ultra-thin nanosheets. The photocatalytic hydrogen evolution rate of this double S-scheme heterojunction catalyst is 62.37 mmol g
−1 h
−1, and the quantum efficiency at 365 nm is 45.9%. The ternary TiCN photocatalyst has suitable lattice matching and fast charge transport channels, which is beneficial to the separation and transfer of photogenerated carriers
[80][21]. Ye et al. prepared S-doped g-C
3N
4/N-doped MoS
2 S-scheme photocatalysts (NMS/SCN) using thiourea and copper chelator as raw materials via the thermal polymerization method. The S released by thiourea during calcination is doped in g-C
3N
4, and the N released by copper chelating agent is doped in MoS
2. The charge transfer between NMS and SCN follows the S-scheme. The electrons and holes are distributed on SCN and NMS, respectively, and the carrier recombination is effectively suppressed. The hydrogen evolution rate of NMS/SCN photocatalysis is 658.5 μmol g
−1 h
−1, which is 23 and 38 times that of NMS and SCN, respectively
[81][22]. In summary, the construction of an S-scheme photocatalyst is a simple and efficient way to enhance the photocatalytic performance of carbon nitride. In addition, defect engineering, interfacial engineering, and co-catalyst modification can be combined with the S-scheme heterojunction to promote the performance of carbon nitride-based heterojunction photocatalysts.
3. Techniques for the Identification of Carbon Nitride-Based S-scheme Photocatalysts
With the progress of characterization technology and the continuous exploration of researchers, the carrier transfer characteristics of an S-scheme photocatalyst has been gradually verified. An in situ irradiation X-ray photoelectron spectroscopy (ISIXPS) and kelvin probe force microscope (KPFM) are two characterization techniques that can directly prove the electronic transfer in S-scheme photocatalysts. XPS is a high-precision instrument for detecting the elemental and valence of solid materials. The binding energy in the spectrum is related to the change in electron density. After the element loses some electrons, its electron density decreases. Therefore, the binding energy of the elements that lose electrons will increase; on the contrary, the binding energies of the elements that get electrons will decrease. The changing trend of the binding energy of different elements can prove the formation of the interface electric field of an S-scheme photocatalyst and the direction of electron transfer under irradiation. For example, in the S-scheme photocatalyst constructed by α-Fe
2O
3 and g-C
3N
4, g-C
3N
4 has a smaller work function than α-Fe
2O
3. After contact, the electrons of carbon nitride transfer spontaneously to α-Fe
2O
3. The loss of electrons in g-C
3N
4 results in a positive shift in the binding energy of N and a decrease in the binding energy of Fe in α-Fe
2O
3 after obtaining electron. The inhomogeneous charge distribution forms an interfacial electric field from g-C
3N
4 to α-Fe
2O
3. When the composite photocatalyst is excited by light, the photogenerated electrons of α-Fe
2O
3 move to carbon nitride under the action of the interfacial electric field. At this time, g-C
3N
4 loses electrons and α-Fe
2O
3 gets electrons, so the binding energy of N and Fe shift positively and negatively, respectively
[83][23]. Therefore, ISIXPS can characterize the direction of electron transfer in S-scheme heterojunctions. KPFM can draw a topographic map and potential distribution map of the sample surface. Yu et al. used KPFM to study charge transfer characteristics in S-scheme photocatalysts composed of CdS and pyrene-based conjugated polymers (PT). The PT has a smaller work function than CdS according to the contact potential difference between pure CdS and pure PT. The electrons on PT are transferred to CdS and its surface is positively charged, so the surface potential of PT is higher compared to the CdS. The potential difference between CdS and PT indicates the existence of an interfacial electric field. The surface potential of PT decreases and the potential of CdS increases under light, indicating that photogenerated electrons are moved from CdS to PT
[82][24]. Therefore, KPFM can also visually characterize the electron transfer in S-scheme photocatalysts.
Other techniques can also indirectly prove or assist in proving the formation of S-scheme heterojunctions, such as the scanning kelvin probe (SKP), surface photovoltage (SPV), femtosecond transient absorption (fs-TA), and electron paramagnetic resonance (EPR). An SKP can test the work function of a semiconductor to get information about the fermi level and provide a basis for the identification of the interface electric fields in heterojunctions. Similar characterization technologies also include ultraviolet photoelectron spectroscopy (UPS)
[84][25]. The SPV signal can provide feedback information about charge transfer, which is similar to a transient photocurrent, electrochemical impedance spectroscopy (EIS), photoluminescence (PL), etc. The built-in electric field can drive the rapid transfer of charge carriers to form a fast and strong SPV response. The peaks of OH and O
2− in the EPR spectrum are related to the redox ability of photogenerated carriers in the photocatalyst. A different electronic transfer mechanism can lead to changes in the redox property of the heterojunction photocatalyst. The charge transfer characteristics in the photocatalyst can be judged by the trend of the radical signal. Hu et al. verified the charge transfer characteristics of g-C
3N
4/InVO
4 S-scheme photocatalysts by EPR
[85][26]. The single carbon nitride is not sufficient to oxidize water to produce ·OH, and the single InVO
4 is not sufficient to reduce oxygen to produce ·O
2−. The ERP spectrum of the g-C
3N
4/InVO
4 composite photocatalyst shows significantly enhanced signal peaks for ·O
2− and OH, indicating that the charge transfer characteristics in g-C
3N
4/InVO
4 is an S-scheme rather than type-I. The interface charge transfer characteristics of S-scheme photocatalysts can also be verified with DFT calculations, including the average planar electron density difference Δρ(z) and differential charge density map. Li et al. verified the charge transfer pathways of Fe
2O
3/C
3N
4 S-scheme heterojunction via DFT calculations
[86][27]. The average planar electron density curve and differential charge density map show that the electrons move from C
3N
4 to Fe
2O
3, resulting in the interfacial electric field being created from C
3N
4 to Fe
2O
3. Therefore, DFT calculations can provide a basis for judging the formation of the interface electric field. Fs-TA can simulate the quenching path and corresponding lifetime of carriers in semiconductors by extracting spectral attenuation features. Yu et al. used the fs-TA technique to study the photophysical processes of S-scheme heterojunctions such as cadmium sulfide/pyrene-ALT-difluorobenzothiadiazole (CdS/PBD) and detected the charge transfer signal of the S-scheme and the lifetime of this transfer process
[87][28].
A built-in electric field-driven S-scheme charge transfer can efficiently separate electrons and holes. Therefore, photoelectrochemical tests and photoluminescence spectra can also demonstrate the successful construction of S-scheme heterojunctions to some extent. The transfer of photogenerated electrons along the external circuit will generate a photocurrent. The higher the separation efficiency of photogenerated electrons and holes, the greater the photocurrent generated
[84,85][25][26]. The transient photocurrent response of p-C
3N
4/InVO
4 is significantly higher than that of the original p-C
3N
4 and InVO
4, indicating that p-C
3N
4/InVO4 has higher charge separation efficiency, which means that more electrons and holes can participate in the photocatalytic reaction
[85][26]. The EIS can also prove that the S-scheme heterojunction has excellent charge transfer characteristics. In the LaVO
4/g-C
3N
4 S-scheme heterojunction system, the composite has the smallest arc radius and the best photocatalytic performance. The original g-C
3N
4 has the largest arc radius and the worst photocatalytic activity, indicating that the S-scheme heterojunction is beneficial to reduce the charge transfer resistance and promote the carrier separation
[84][25]. The PL spectrum can represent the intensity of electron and hole recombination luminescence. Stronger PL peaks mean higher electron-hole complexation rates. The separation efficiency of electrons and holes can also be analyzed via a fluorescence lifetime. The fluorescence lifetime is equivalent to the existence time of photogenerated carriers. The longer the fluorescence lifetime of the photocatalyst means the better the separation of electrons and holes
[80,84][21][25].