Unravelling the transformation from a type-I to type-II MA₃Bi₂I₉-based heterostructure photocatalyst via energy band engineering

Abstract

The photocatalytic dissociation of hydroiodic acid (HI) utilizing halide perovskites offers an environmentally benign and economically viable approach for hydrogen production under ambient temperature conditions. With lead-halide perovskites showing encouraging efficacy in the domain of photocatalytic hydrogen generation, this work focused on developing a lead-free Bi-based hybrid perovskite, specifically MA₃Bi₂I₉ (MABI), which was successfully synthesized in a heterostructure configuration, wherein the MABI perovskite was in situ grown around amorphous MoS₂. This research underscores that for heterostructures made of amorphous MoS₂ and MABI, the doping of phosphorus not only modified the energy levels but it also altered the crucial bandgap values of amorphous MoS₂. The shifted energy levels of MoS₂ relative to MABI resulted in unique energy band arrangements for the three composites. A transition of the heterojunction from type I to type II was observed with the phosphorus-doped MoS₂-containing composites. Among all three composites, P50_MoS₂/MABI possessed advantageous band alignment, facilitating the most efficient photogenerated charge separation and transport. Under optimal reaction parameters, a hydrogen evolution rate of 1176 µmol g⁻¹ h⁻¹ can be attained for P50_MoS₂/MABI composites.

---

Introduction

To meet the need for clean and renewable energy, H₂ stands as a potential candidate against fossil fuels as well as a solution for mitigating the environment deterioration. Among the various strategies adopted for H₂ generation, photocatalysis stands as the most promising approach to mitigate the energy crisis.^1–3 To date, the photocatalytic solar energy conversion process has been realized with various oxide-based materials like TiO₂, CaTiO₃, etc.^4–6 However, the last few years have witnessed the tremendous utilization of lead halide-based perovskites for solar H₂ to electricity generation.^7–9 The halide-based perovskites were widely explored for photovoltaic and optoelectronic applications due to their efficient light absorption property, large extinction coefficient, high carrier mobility, larger carrier lifetime, tunable bandgap and low exciton binding energy.^10–12 These properties also make them a suitable candidate for photocatalysis for solar H₂ generation.^13–15 Despite the good photocatalytic activity shown by the lead-based halide perovskite,^16–18 the high toxicity of lead and the material instability limit its utility for large-scale applications.¹⁹ Therefore, there is a need to explore stable and lead-free perovskite materials.²⁰ One alternative is the substitution of Pb with Sn since it belongs to the same group. However, the intrinsic instability of Sn²⁺ against air and moisture readily oxidizes Sn²⁺ to the Sn⁴⁺ state, degrading the material. A recent study highlighted the bismuth-based perovskite as a substitution to the lead-based perovskite, as Bi³⁺ is isoelectronic with Pb²⁺ with comparable electronegativity and similar ionic radii.^21–23 These are extremely stable in ambient conditions, unlike Pb- and Sn-based perovskite materials. A class of bismuth-based perovskites of general formula A₃Bi₂X₉, where A is a monovalent cation (Cs⁺ or MA⁺) and X is a halide anion (Cl⁻, Br⁻, and I⁻), was recently used as a photoactive material in both photovoltaic application and photocatalytic HER.^24,25 Of these materials, MA₃Bi₂I₉ (MABI) has been less widely investigated. Its structure consists of metal halide octahedra layers with a void between the two layers filled by MA⁺ cations, and the octahedral (Bi₂I₉)³⁻ clusters in MA₃Bi₂I₉ are surrounded by MA⁺ cations. A heterojunction of MABI/Pt for HI splitting has also been reported.²⁶ Such heterojunction structures prevent charge recombination and favour efficient charge transfer by forming a Schottky barrier.²⁷ However, the use of the expensive Pt metal is a significant hindrance for practical application. Thus, the exploration of alternative inorganic materials with high capability of efficient separation and transfer for enhanced H₂ generation is extremely crucial.²⁸

MoS₂, classified as a transition metal dichalcogenide, has been extensively investigated as an electrocatalyst due to its lower overpotential, which facilitates an enhanced hydrogen evolution rate.²⁹ In theoretical aspects, a material is regarded as a potentially effective HER catalyst when the free energy of the adsorbed atomic hydrogen is close to that of the reactant or product (i.e., ).³⁰ The edge sites of MoS₂ have been recognized for promising H₂ evolution activities.³¹ Subsequent investigations utilizing MoS₂ revealed that the HER activity is indeed directly correlated with the edge length, rather than the surface area.³² The preferential exposure of edge sites was more prominently observed in amorphous MoS₂, with relatively minimal surface area rather than crystalline 2D MoS₂, which possesses a higher surface area but a diminished number of active edge sites. The abundance of exposed edge sites in amorphous MoS₂ has facilitated the advancement of more efficient HER catalysts.^33,34 Alongside its high activity, good stability against strong acids and affordability against noble Pt metal, MoS₂ is a promising material that can assist in the photocatalytic HER process.

The pivotal function of element doping lies in its capacity to modify the inherent architecture of the original molecule, thereby exerting a significant influence on the formation of chemical bonds, the processes of adsorption and desorption, and the Gibbs free energy of reactions.^35,36 The stability of the doped configurations was evaluated through the formation energy associated with P-doping (E_formation). Negative values of formation energies, significantly below zero, indicate that phosphorus can be readily incorporated into the MoS₂ structure, resulting in a stable configuration.³⁷

The difference in electronegativity between phosphorus and sulfur atoms prompts a rearrangement of electrons around the molybdenum and sulfur atoms. The electronic interactions resulting from phosphorus doping can attenuate the S–H_ads bonds that develop on the catalyst's surface.³⁸

The weakened S–H_ads bonds facilitate a balance in the adsorption and desorption dynamics between the catalyst and hydrogen atoms, thereby augmenting the catalytic efficiency. Phosphorus doping increases the number of new states near the Fermi level and subsequently drives the Fermi level downward towards the valence band. The Fermi level denotes the highest occupied electronic energy state at absolute zero temperature. This increase in the density of states (DOS) proximate to E_F serves as a direct indicator that the MoS₂ material, now containing phosphorus dopants, exhibits enhanced conductivity, which is essential as it promotes electron transfer during light exposure, thus accelerating the rates of photocatalytic reactions.³⁷

In this work, we have delineated a series of composites composed of MA₃Bi₂I₉ and variants of amorphous MoS₂, which together establish heterostructures aimed at enhancing photocatalytic H₂ evolution in HI solution upon exposure to visible light irradiation (Fig. S1). Through UPS analysis, it was discerned that phosphorus doping in MoS₂ instigates a downward trend in the VBM. As the phosphorus doping concentration increases, the VBM was observed to shift into a considerably deeper negative energy domain. This doping not only modifies the band alignment of MoS₂, but also alters the bandgaps of the three MoS₂ variants. The relative positions of the CBM and VBM of both MABI and MoS₂ constituents within the composites engender three distinct scenarios. The doping has induced a transition from a type I heterostructure observed in MoS₂/MABI to a type II heterojunction noted in the other two doped variants of composites. This heterojunction transition has significantly influenced the photocatalytic HER activities, resulting in varied HER performance across the three composites. The electronic band engineering, which is attributed to the change in elemental composition, could impact the charge transfer characteristics in photocatalytic HER activities. Even though all heterostructures exhibited improved HER activity as compared to pristine MABI, the highest photocatalytic H₂ generation rates were delivered by P50_MoS₂/MABI composites. Pristine MABI showed negligible HER activity under visible white light LED. In contrast, for the P50_MoS₂/MABI composites, the registered maximum HER activity was 1176 µmol g⁻¹ h⁻¹, which was 100 times higher than that for pristine MABI.

Results and discussion

The powder X-ray diffraction analysis of MA₃Bi₂I₉ (MABI) revealed diffraction peaks at angles of 12.46°, 12.79°, 14.40°, 17.02°, 24.56°, 25.31°, 25.76°, 26.95°, 29.06°, 31.64°, 32.31°, and 39.13°, which can be attributed to the (100), (101), (103), (105), (006), (202), (203), (216), and (208) crystallographic planes, respectively, as illustrated in Fig. 1a. Due to the high crystallinity of MABI, no discernible broad peaks characteristic of the amorphous MoS₂ were detected in the diffractogram of the MABI composites, as shown in Fig. 1a. The XRD pattern obtained for MA₃Bi₂I₉ was found to be in remarkable concurrence with previously reported data.²⁶ The comparative PXRD patterns for the P-doped amorphous MoS₂ and the pristine amorphous MoS₂ are displayed in Fig. 1b. Five prominent peaks at 16.72°, 28.62°, 37.8°, 47.34° and 55.42°, corresponding to the (002), (100), (103), (105) and (110) planes of MoS₂, respectively, were identified.³⁹ The lack of crystallinity in amorphous MoS₂ makes the peaks broader and shifted from the standard reference. The stability of the perovskite-like materials was a significant concern. Hence, the ambient stability of the synthesized MABI was evaluated. MABI was subjected to aqueous HI medium for a duration of 53 hours, after which the PXRD pattern was re-recorded. The PXRD pattern obtained from the light-exposed MABI is presented in Fig. S2. The results indicated negligible changes in the PXRD pattern, suggesting that MABI possesses exceptional stability even after 53 hours of photocatalytic hydrogen generation.

Fig. 1 XRD patterns of the (a) three MABI composites and of (b) the 3 variants of MoS₂.

The light absorption characteristics of the synthesized MABI and its various composites were analyzed utilizing UV-vis absorption spectroscopy. The resulting pristine MABI powder demonstrated a significant absorbance within the visible light spectrum and a broad absorption range spanning up to 500–600 nm, as depicted in Fig. 2a. MABI possesses a bandgap of 1.98 eV (shown in Fig. S3), which aligns remarkably well with findings from prior investigations.⁴⁰ This favourable bandgap of MABI is appealing for further investigation into its potential as a photocatalyst for H₂ generation applications. In this regard, amorphous MoS₂ and phosphorus-doped MoS₂ were combined with MABI to achieve enhanced H₂ evolution performance. The solid-state UV-visible spectra of all samples (MABI, MoS₂/MABI, P50_MoS₂/MABI, P100_MoS₂/MABI) demonstrated that the incorporation of MoS₂ onto the MABI substrate has significantly broadened the absorption spectrum of solar radiation. Moreover, P-doped MoS₂ further extended the absorption tail up to 750 nm. Consequently, the P-doped MABI heterostructures were anticipated to absorb a wide spectrum of sunlight.¹⁵ The bandgaps of MoS₂, P50_MoS₂, and P100_MoS₂ were calculated to be 1.45 eV, 1.43 eV, and 1.2 eV, respectively, as illustrated in Fig. S4.

Fig. 2 Solid-state UV-visible spectra of (a) MABI and its three composites and of the (b) amorphous MoS₂ variants (inset: full spectra).

The fundamental reason for the bandgap narrowing after phosphorus doping is the higher orbital energy of the P 2p orbital compared with the S 2p orbital. Phosphorus doping leads to an increase in the charge states close to the Fermi level (E_F). The PDOS analysis shows that the Mo 3d orbital and the P 2p orbital are the key orbitals involved in the electronic restructuring upon P doping.

Due to the introduction of phosphorus atoms, there is an enhancement in coupling between the Mo 3d orbital and the P 2p orbital. This interaction introduces new electronic states closer to the Fermi level, effectively filling in parts of the bandgap and reducing the overall energy difference between the valence band maximum and the conduction band minimum. The degree of this coupling is improved by increasing the concentration of phosphorus. As a result of these changes in the electronic structure, the bandgap becomes more narrow.³⁷Fig. 2b depicts portions of the normalised solid-state UV-visible spectra of three variants of MoS₂.

To elucidate the morphology and structural characteristics of MABI and its composites, as well as the interactions between MABI and MoS₂ within the heterostructures, FESEM and FETEM analyses were conducted. Fig. 3a and b exhibit the FESEM images of pristine MABI and MABI composites, respectively. The pristine MABI microcrystals exhibited a high degree of crystallinity, characterized by hexagonal shape and smooth surface morphology.⁴¹ Furthermore, following the integration of amorphous MoS₂ with MABI, a microstructure emerged where the MoS₂ were observed to be securely attached to the MABI surface, indicating a close integration of MABI and MoS₂ clusters. This robust heterojunction might be attributed to the growth of the MABI microcrystal structures around the MoS₂ clusters. The high-resolution FESEM image of amorphous MoS₂ is displayed in Fig. S5. The size distributions of MoS₂ and MABI are depicted in Fig. S6 and S7, respectively. The mean sizes of amorphous MoS₂ and MABI were 3.55 µm and 9.38 µm, respectively. Energy dispersive X-ray spectroscopy (EDX) was employed to analyze the elemental distribution within the composites. The elemental mapping of the MABI composite (Fig. 3d–g) reveals the presence of I, Bi atoms (contributed by MABI) and Mo, S (contributed by MoS₂), which were uniformly dispersed throughout the crystal structure.

Fig. 3 FESEM top view images of the (a) pristine MABI and (b) MABI composite. EDX elemental mapping of (c) all the elements, (d) I, (e) Bi, (f) Mo, and (g) S for the MABI composites.

TEM and HRTEM analysis of MABI heterostructures provides valuable insights regarding their crystal structural characteristics, which are essential for understanding and optimizing their performances in HER activities. TEM analysis of MABI heterostructures typically shows distinct regions of both MABI and amorphous MoS₂. The MABI microcrystal exhibited a hexagonal crystal structure, while amorphous MoS₂ generally has layered structures, as depicted in Fig. 4a.

Fig. 4 (a) TEM and (b) HRTEM images of the MABI composites. Inset of (b) shows the magnified images of MABI and MoS₂.

The high-resolution TEM images of the MABI composites (Fig. 4b) revealed the presence of crystal lattices of both MABI and MoS₂ components. It also showed the mixing of two interfaces between these two materials, which is of particular interest, as it plays a crucial role in the heterostructure's properties.⁴² The d-spacings of MABI and MoS₂ were calculated as 0.23 nm and 0.29 nm, respectively.

X-ray photoelectron spectroscopy (XPS) analysis was utilized for a more comprehensive examination. The XPS survey spectrum of the synthesized P50_MoS₂/MABI, illustrated in Fig. 5a, reveals the existence of C, N, Bi, I, Mo, S, P elements. The high-resolution XPS of phosphorus present in the P50_MoS₂/MABI composites, displayed in Fig. 5b, depicts the presence of P 2p_1/2 and P 2p_3/2 doublets. The high-resolution XPS spectrum of I 3d is depicted in Fig. 5c, which reveals two peaks at 630.36 eV and 618.732 eV for pristine MABI, corresponding to the I 3d_3/2 and I 3d_5/2 states of the I⁻ ions, respectively. These peaks associated with the 3d_3/2 and 3d_5/2 states exhibited a shift of 0.23 eV towards higher binding energy, following the integration of P50_MoS₂ in the P50_MoS₂/MABI composites. The high-resolution XPS spectrum of Bi4f (Fig. 5d) exhibits two peaks at binding energies of 156.5 eV and 161.88 eV for pristine MABI, which are attributed to the doublet state of +3 oxidation states, assigned as Bi 4f_7/2 and Bi 4f_5/2, respectively. A shift towards a higher binding energy by 0.31 eV was observed in the P50_MoS₂/MABI composites compared to the pristine MABI, indicating an efficient transfer of photogenerated electrons to the anchored P50_MoS₂.⁴⁰

Fig. 5 XPS survey scan of (a) P50_MoS₂/MABI. High-resolution deconvoluted XPS spectra of (b) P 2p, (c) I 3d, (d) Bi 4f and (e) Mo 3d for the composites and pristine components.

For the P50_MoS₂/MABI composites, the Mo 3d peaks appeared at 229.32 and 232.28 eV, as depicted in Fig. 5e, and are attributed to Mo 3d_5/2 and Mo 3d_3/2, respectively, for the doublet of Mo⁴⁺.⁴³ Also, the sulfur atom-related 2s peak was observed at 226.38 eV for S²⁻. In addition, the peak at 235.48 eV corresponding to the Mo⁶⁺ of MoO₃ was observed. For P50_MoS₂/MABI composites, the Mo 3d state demonstrated a negative shift of approximately 0.65 eV when compared to pristine P50_MoS₂, as illustrated in Fig. 5e. These findings consistently indicate that the P50_MoS₂ component presented in the composites was accepting photoelectrons, which leads to a decrease in the binding energy of the Mo 3d state. The observed shifts in binding energy towards higher values for Bi and I, and lower values for Mo, indicate an efficient interfacial charge transfer from MABI to P50_MoS₂ through a robust heterojunction, ultimately enhancing the photocatalytic performance of P50_MoS₂/MABI.⁴⁴

To investigate the role of three variants of amorphous MoS₂ as a photoelectron carrier within the charge transfer mechanism of MABI composites, a time-resolved photoluminescence (TRPL) investigation was undertaken, as illustrated in Fig. 6a. The MABI compound and three distinct variants of MABI composites display decay phenomena at approximately 630 nm, characterized by diverse patterns that reflect their unique rates of charge recombination. The lifetimes corresponding to these decay processes exhibited a significant correlation with the recombination dynamics of electron–hole pairs. The pristine MABI demonstrated an average lifetime of 8.2 ns, which is the highest among all, as depicted in Fig. 6a. The PL decay of MABI was markedly accelerated in the presence of the three MoS₂ variants, necessitating the application of a two-exponential function to fit the PL kinetics. Hence, the mean lifetimes for all three MABI composites were notably diminished, indicating an expedited charge transfer mechanism within the heterostructures. Consequently, the molybdenum dichalcogenides component can efficiently facilitate the extraction of photogenerated electrons from MABI due to the favourable alignment of energy levels.⁴⁵ Among the three heterostructures analysed, the P50_MoS₂/MABI and P100_MoS₂/MABI composites exhibited higher lifetimes in comparison to the MoS₂/MABI configuration. Specifically, P50_MoS₂/MABI and P100_MoS₂/MABI recorded lifetimes of 4.2 ns and 3.7 ns, respectively. In contrast, MoS₂/MABI demonstrated the lowest recorded lifetime of 2.91 ns. The TRPL decay parameters were included in tabular form (Table S1).

Fig. 6 (a) TRPL decay at 630 nm, (b) transient photocurrent responses under visible light irradiation, and (c) EIS Nyquist plots of the pristine MABI and MABI composites.

The P50_MoS₂/MABI and P100_MoS₂/MABI composites were identified as type II heterostructures, in contrast to the MoS₂/MABI composite, which forms a type I heterostructure. In type I systems, the significant spatial overlap between photo-generated charge carriers enhances radiative recombination, typically resulting in shorter carrier lifetimes. Conversely, type II heterostructures facilitate a more effective spatial separation of electrons and holes, as evidenced by the prolonged TRPL lifetimes observed in the P50_MoS₂/MABI and P100_MoS₂/MABI composites. This reduction in charge carrier recombination corresponds to more efficient charge separation and migration.⁴⁶ The photoelectrochemical (PEC) evaluation was employed to furnish substantial evidence regarding charge separation and transfer within the composites. The photo-current responses for the unmodified MABI and its various composites were recorded over multiple 10-second illumination intervals and are depicted in Fig. 6b. Clearly, under the illumination condition, the photocurrent intensity was noted to remain relatively stable, while it rapidly diminished to zero upon cessation of light, indicating a swift photocurrent response to the on–off light conditions. It was noted that the P100_MoS₂/MABI composites exhibited the highest photocurrent, whereas the photocurrent progressively declined for the P50_MoS₂/MABI and MoS₂/MABI composites, respectively. As expected, the pristine MABI displayed the lowest photocurrent when subjected to the same electrical bias. In the realm of transient photocurrent, a higher value signifies effective and efficient charge mobility within the composites, which frequently correlates with the superior hydrogen evolution rate among all tested samples.⁴⁷ These findings were corroborated by electrochemical impedance spectroscopy (EIS) measurements, as illustrated in Fig. 6c. The P100_MoS₂/MABI composite, exhibiting the smallest semicircle arc, signifies the lowest interfacial charge transfer resistance, thereby facilitating expedited charge transfer at the interface. In contrast, the interfacial charge transfer resistance for the other MABI composites progressively increased, peaking for the pristine MABI.⁴⁸ This indicates that charge migration was most arduous for the pristine MABI, consequently resulting in the lowest photocurrent. This observation elucidates that under certain scenarios, P-doped amorphous MoS₂-based heterostructures displayed superior photo-electrochemical properties.⁴⁹ Notably, along with the extended carrier lifetimes for P50_MoS₂/MABI and P100_MoS₂/MABI composites, they also demonstrate both higher photocurrent responses and lower charge transfer resistance compared to the MoS₂/MABI material with type I heterojunctions. These results reinforce the conclusion that type II heterostructures promote more efficient charge separation and transfer processes. Under certain conditions, it might accelerate the photocatalytic reductive H₂ evolution reaction rate.⁴⁶

In the context of photocatalytic H₂ generation reactions, the MABI composites were synthesized through the incorporation of different variants of MoS₂ and precursors, notably BiI₃ and MAI, within a mixed solution of MABI-saturated aqueous HI/H₃PO₂, utilizing the method of in situ crystallization. The photocatalytic H₂ evolution efficiency of the synthesized samples was evaluated under visible white light LED irradiation. The three fundamental elements in this photocatalytic process are the photocatalyst, the illumination source, and the reaction medium. To validate this assertion, a sequence of controlled photocatalytic experiments were performed, during which each component was systematically omitted. Under these controlled conditions, no H₂ gas was observed in the absence of visible light irradiation, and the MABI microcrystals underwent disintegration when HI was absent from the reaction medium. Nevertheless, the photocatalytic activities were verified via gas chromatography, with the visible light acting upon a photocatalyst immersed in the MABI-saturated HI/H₃PO₂ solution. The H₂ evolution rates for all samples are compiled and illustrated as a function of light exposure duration in Fig. 7a. The H₂ evolution activity of pristine MABI was too low to be determined. This subpar performance may be attributed to a limited number of surface reaction sites present on MABI surfaces, resulting in a diminished generation of photogenerated electrons and a delayed transfer to the reaction medium.⁴⁹ Unexpectedly, the HER performance exhibited a significant enhancement upon the incorporation of amorphous MoS₂ with MABI microcrystals. P50_MoS₂/MABI composites showcased the highest HER activity, achieving 1176 µmol g⁻¹ h⁻¹, the most notable performance across the various composite formulations.

Fig. 7 (a) Comparative study of H₂ evolution of all the MABI composites (10 wt%) and (b) H₂ evolution activity of the P50_MoS₂/MABI composites during a 53 h white light LED experiment.

The MoS₂/MABI and P100_MoS₂/MABI composites demonstrated H₂ evolution rates in the range of 300–450 µmol g⁻¹ h⁻¹, much lower HER activity compared to P50_MoS₂/MABI. The formation of a robust heterojunction contributes to the aforementioned enhancement of the HER performance. A comparative performance summary table (Table S2) has been included in the SI. In terms of H₂ evolution activity, all the MABI composites not only surpassed pristine MABI, but also displayed superior HER performance relative to the MABI samples decorated with Pt.²⁶ Additionally, it is pertinent to note that the best performing P50_MoS₂/MABI composites exhibited commendable stability in HER activities over an extended duration (Fig. 7b). The photocatalytic H₂ evolution was conducted utilizing P50_MoS₂/MABI composites under white LED illumination for a period of 53 hours. There was no significant decline in HER activity over 8 consecutive cycles, each spanning 5 hours with a one-hour interval for the evacuation and purging of inert nitrogen gas. However, following prolonged irradiation of 53 hours, a substantial decline in HER activity was observed during the final cycle. This reduction in measured activity may be attributed to an excessive accumulation of I₃⁻. However, XPS and TEM analysis of the P50_MoS₂/MABI samples were performed after 53 hours of HER activity, and no significant degradation was observed following this duration (Fig. S8 and S9). This phenomenon can be elucidated through the in situ crystallization process, wherein amorphous MoS₂ clusters were effectively integrated on the MABI surface (Fig. 3b and 4b) resulting in the probability of segregation between MABI and MoS₂ diminishing during magnetic stirring.

A crucial technique, UPS analysis, was employed to assess the VBM of three variants of MoS₂ as well MABI. This investigation revealed that MABI has its VBM position at −5.74 eV against vacuum, whereas the VBM of pristine MoS₂ was positioned at −5.43 eV against vacuum (Fig. S10 and S11). The obtained bandgap of MABI and pristine MoS₂ were 1.98 eV and 1.45 eV, respectively. Hence, the CBM values of MABI and MoS₂ were calculated as −3.76 eV and −3.98 eV, respectively. Notably, upon P-doping, the VBM of the doped MoS₂ exhibited a downward shift for both P50_MoS₂ and P100_MoS₂. The VBM of P50_MoS₂, obtained from UPS analysis was −5.76 eV (Fig. S12). For P100_MoS₂, the VBM was calculated as −6.27 eV (Fig. S13). Hence, it was clearly evident that the doping of phosphorus showed a trend towards deeper negative energy level for VBM. The bandgap of P100_MoS₂ followed a declining trend compared to pristine MoS₂. The calculated bandgaps of both P50_MoS₂ and P100_MoS₂ were 1.43 eV and 1.24 eV, respectively. Summing up both the data of VBM and bandgap, the obtained CBM values of P50_MoS₂ and P100_MoS₂ were −4.33 eV and −5.03 eV, respectively. For scale conversion from vacuum to RHE, we usually consider that −4.44 eV against vacuum is equivalent to 0 V against RHE. This conversion works well under ideal conditions when pH is 0 (i.e., C_H⁺ = 1 M). However, if the pH changes, the RHE scale also shifts either in an upward or downward directions. Here, in this project, we use 57 wt% HI and 50 wt% H₃PO₂ in a 5 : 1 v/v ratio. The pK_a values of HI and H₃PO₂ are 2 × 10⁹ and 1.1, respectively. Assuming HI as a strong acid of high concentration (7.5 M) and H₃PO₂ as a very weak acid, we assume that the protons present in water are solely coming from HI. The concentration of H⁺ in the solution would be near about 7.5(5/6) = 6.25 M. Converting it into pH scale, the calculated pH of the solution would be −log₁₀(6.25) = −0.796.⁵⁰ Using the Nernst equation, the redox potential E(H⁺/H₂) would be (E⁰ + 0.04) V against RHE. After considering the pH effect on the RHE scale, if we compare both the vacuum scale and RHE scale side by side, it will move towards positive direction against vacuum scale, i.e., upwards. After considering the pH effect, during the scale conversion from vacuum to RHE, 0 V against RHE would be equivalent to −4.4 V against vacuum scale. Hence, after conversion from the vacuum scale to the RHE scale, the VBM values of MoS₂, P50_MoS₂, and P100_MoS₂ are 1.03 eV, 1.36 eV, and 1.87 eV, respectively. Meanwhile, the CBM values of MoS₂, P50_MoS₂, and P100_MoS₂ are −0.42 eV, −0.07 eV and 0.63 eV, respectively.

As delineated in Scheme 1 for MoS₂/MABI, the VBM of pristine MoS₂ was situated considerably higher than that of MABI, while the CBM resides below that of MABI. This alignment of band positions exemplifies a classical type I heterostructure, wherein both the CBM and VBM of the pristine MoS₂ component were encompassed within the respective CBM and VBM of the MABI component. Consequently, both photogenerated electrons and holes were confined within the individual MoS₂ component of the composites. Such an arrangement enhances the probability of radiative recombination in MABI composites, potentially resulting in lower average TRPL lifetime and lower photocurrent, leading towards reduced HER activities. In contrast, for the P-doped variants of MoS₂, a downward shift towards a deeper negative energy region was observed, as depicted in Scheme 1. This downward shift created a favourable scenario for a type II heterojunction, where the CBM and VBM of the P-doped MoS₂ component were positioned below or in close proximity to those of MABI, yielding a spatial separation of electrons and holes in P50_MoS₂/MABI and P100_MoS₂/MABI composites.

Scheme 1 Schematic of the band alignment of three variants of MABI composites and the mechanism of H₂ evolution in aqueous HI.

Photogenerated electrons were predominantly localized in the conduction band of the MoS₂ component, while holes were situated in the valence band of MABI. This spatial separation facilitates extended TRPL lifetimes and was beneficial for the transfer of photogenerated charges in both P50_MoS₂/MABI and P100_MoS₂/MABI composites. The photocurrent response and charge transfer resistance data further corroborate this observation. Between the P50_MoS₂/MABI and P100_MoS₂/MABI composites, the latter one exhibited better photoelectrochemical performances despite both having the type II heterostructures. The driving force for electron transfer (ΔG) matters because it strongly affects the transfer rate (as per Marcus theory), and it is linked to how the CBM shifts.^51,52

For two of the composites (MoS₂/MABI, P50_MoS₂/MABI) the CBM of MoS₂ components were situated above and in close proximity to the redox potential E(H⁺/H₂), which is −4.55 eV relative to vacuum or 0 V relative to NHE.¹⁸ This specific configuration of the energy bands was extremely crucial for efficient proton reduction, where MoS₂ and P50_MoS₂ could efficiently transfer the photoelectrons towards the protons present in the reaction medium. Conversely, the CBM of P100_MoS₂ was found to be positioned much deeper relative to E(H⁺/H₂). Although the relative energy band alignment between MABI and P100_MoS₂ created a type II heterojunction which might be favourable for effective charge extraction from MABI towards P100_MoS₂, this particular aforementioned band position of P100_MoS₂ was very detrimental for HER activities. The role of molybdenum dichalcogenides was to extract the photoelectron from MABI and transfer it towards the protons present in the reaction medium. Hence, P100_MoS₂ fails to serve as a photoelectron carrier. Despite the most efficient charge separation and transfer, P100_MoS₂ could not facilitate proton reduction, as its CBM was situated significantly below E(H⁺/H₂).

An understanding of the distinct photocatalytic and photoelectrochemical characteristics can be elucidated through the analysis and comparison of the energy profile diagrams corresponding to the three composites, as depicted in Scheme 1. As previously noted, among the three composites MoS₂/MABI, P50_MoS₂/MABI, and P100_MoS₂/MABI; the P-doped composites exhibited type II heterostructures. For the initial two composites, the MoS₂ component was instrumental in facilitating proton reduction. This observation clearly indicates that P50_MoS₂/MABI fulfils two of the most critical criteria for being an effective photocatalyst. In contrast, the other two composites fall short in one of these essential characteristics, adversely impacting their photocatalytic performance. Consequently, both MoS₂/MABI and P100_MoS₂/MABI exhibited HER activities within a comparable range, while P50_MoS₂/MABI demonstrated the best HER performance, as depicted in Fig. 7a. Among the remaining two composites, although the P100_MoS₂/MABI composites with type II heterostructures showcased superior TRPL lifetime and photocurrent response, MoS₂/MABI with type I heterostructure exhibited marginally superior HER performance.

Conclusion

The impact of phosphorus doping on the bandgap and energy levels has been thoroughly examined. This investigation reveals that the bandgaps of amorphous pristine MoS₂ and P50_MoS₂ are 1.45 eV and 1.43 eV, respectively. It decreases to 1.2 eV for P100_MoS₂ following phosphorus doping. As a result of interactions between P 2p with Mo 3d, new electronic states closer to the Fermi level arise. These changes in the electronic structure were particularly responsible for the narrow bandgaps. Simultaneously, the VBM energy level of two variants of phosphorus-doped MoS₂ have undergone a downward transition toward a more negative energy region. As a result, the P50_MoS₂/MABI and P100_MoS₂/MABI composites have formed a type II heterojunction, while the MoS₂/MABI composites manifest as a type I heterojunction. The transition from type I to type II heterostructures in MABI composites leads to improved charge separation and transport, as reflected by the increased photocurrent, reduced charge transfer resistance, and extended carrier lifetimes in the P50_MoS₂/MABI and P100_MoS₂/MABI composites. Conversely, for MoS₂/MABI and P50_MoS₂/MABI, the MoS₂ component played a pivotal role in facilitating proton reduction. However, for P100_MoS₂/MABI, the MoS₂ component does not contribute to proton reduction as it obstructs its capability to function as a photoelectron carrier for efficient proton reduction. This finding indicates that P50_MoS₂/MABI meets two of the most essential requirements for being a proficient photocatalyst. In contrast, the other two composites do not satisfy one of these vital characteristics, thereby diminishing their photocatalytic efficacy. Hence, for P100_MoS₂/MABI, despite showing superior charge separation and segregation, it performed worst in terms of HER activity among all three composites. Meanwhile, P50_MoS₂/MABI exhibits the most advantageous HER performance of 1176 µmol g⁻¹ h⁻¹.

Author contributions

Tamal Pal: conceptualization, methodology, validation, formal analysis, investigation, data curation, writing – original draft, visualization; Soumalya Bhowmik: validation, investigation, writing – review & editing, visualization; Sushant Sharma: methodology, resources; Ameer Suhail: resources, data curation; Nageswara Rao Peela: resources; Chivukula V. Sastri: supervision, project administration; and Parameswar Krishnan Iyer: supervision, project administration, funding acquisition.

Conflicts of interest

The authors declare no conflicts of interests.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: synthesis of materials, photochemical and photoelectrochemical measurements. See DOI: https://doi.org/10.1039/d5se01135c.

Acknowledgements

PKI acknowledges financial support from DST, India, through the projects DST/TSG/PT/2009/23, DST/TMD/IC-MAP/2K20/03 and DST/CRG/2019/002164; Deity, India, no. 5(9)/2012-NANO (Vol. II); Max-Planck-Gesellschaft IGSTC/MPG/PG(PKI)/2011A/48; and the SPARC project SPARC/2018-2019/P1097/SL. NRP acknowledges financial support from DST SERB, New Delhi, India under a core research grant (CRG, file no. CRG/2022/005144). CVS acknowledges financial support from DST SERB, New Delhi, India through the grant code CRG/2023/000456. TP, ASN and ASM are thankful to the Ministry of Education, Govt. of India, for the financial support. SB acknowledges financial support and a fellowship from PMRF, Ministry of Education, India (grant number 1900823). The authors are greatly indebted to Dr Koushik Rout and AMRC at IIT Mandi (India), for access to XPS and UPS facilities. The Department of Chemistry, Centre for Nanotechnology and CIF, IIT Guwahati, are acknowledged for instrument facilities.

References

1. M. R. Shaner, H. A. Atwater, N. S. Lewis and E. W. McFarland, Energy Environ. Sci., 2016, 9, 2354–2371.
2. Q. Hassan, A. M. Abdulateef, S. A. Hafedh, A. Al-samari, J. Abdulateef, A. Z. Sameen, H. M. Salman, A. K. Al-Jiboory, S. Wieteska and M. Jaszczur, Int. J. Hydrogen Energy, 2023, 48, 17383–17408.
3. F. Ueckerdt, C. Bauer, A. Dirnaichner, J. Everall, R. Sacchi and G. Luderer, Nat. Clim. Change, 2021, 11, 384–393.
4. F. Li, G. Liu, F. Liu and S. Yang, Chemosphere, 2023, 324, 138277.
5. J. Wang, S. Sun, L. Pan, Z. Xu, H. Ding and W. Li, Materials, 2019, 12, 3369.
6. P. A. Demina, D. V. Voronin, E. V. Lengert, A. M. Abramova, V. S. Atkin, B. V. Nabatov, A. P. Semenov, D. G. Shchukin and T. V. Bukreeva, ACS Omega, 2020, 5, 4115–4124.
7. R. D. Yukta Chavan, A. Mahapatra, D. Prochowicz, P. Yadav, P. K. Iyer and S. Satapathi, ACS Appl. Mater. Interfaces, 2023, 15, 53351–53361.
8. H. Baishya, R. D. Adhikari, M. J. Patel, D. Yadav, T. Sarmah, M. Alam, M. Kalita and P. K. Iyer, J. Energy Chem., 2024, 94, 217–253.
9. R. Das Adhikari, H. Baishya, M. J. Patel, D. Yadav and P. K. Iyer, Small, 2024, 20, e2404588.
10. M. E. Kamminga, G. A. de Wijs, R. W. A. Havenith, G. R. Blake and T. T. M. Palstra, Inorg. Chem., 2017, 56, 8408–8414.
11. N. K. Tailor, S. Mishra, T. Sharma, A. K. De and S. Satapathi, J. Phys. Chem. C, 2021, 125, 9891–9898.
12. M. Roknuzzaman, C. Zhang, K. Ostrikov, A. Du, H. Wang, L. Wang and T. Tesfamichael, Sci. Rep., 2019, 9, 718.
13. S. Chen, H. Yin, P. Liu, Y. Wang and H. Zhao, Adv. Mater., 2023, 35, e2203836.
14. T. Pal, S. Bhowmik, A. Suhail, S. Sharma, N. R. Peela, C. V. Sastri and P. K. Iyer, ACS Appl. Nano Mater., 2024, 7, 12163–12170.
15. T. Pal, S. Bhowmik, A. S. Tanwar, A. Suhail, N. R. Peela, C. V. Sastri and P. K. Iyer, Energy Adv., 2024, 2965–2971.
16. S. Park, W. J. Chang, C. W. Lee, S. Park, H.-Y. Ahn and K. T. Nam, Nat. Energy, 2016, 2, 16185.
17. X. Zhao, S. Chen, H. Yin, S. Jiang, K. Zhao, J. Kang, P. F. Liu, L. Jiang, Z. Zhu, D. Cui, P. Liu, X. Han, H. G. Yang and H. Zhao, Matter, 2020, 3, 935–949.
18. W. Guan, Y. Li, Q. Zhong, H. Liu, J. Chen, H. Hu, K. Lv, J. Gong, Y. Xu, Z. Kang, M. Cao and Q. Zhang, Nano Lett., 2021, 21, 597–604.
19. A. L. Wani, A. Ara and J. A. Usmani, Interdiscip. Toxicol., 2015, 8, 55–64.
20. Y. Tang, C. H. Mak, J. Zhang, G. Jia, K.-C. Cheng, H. Song, M. Yuan, S. Zhao, J.-J. Kai, J. C. Colmenares and H.-Y. Hsu, Adv. Mater., 2023, 35, e2207835.
21. A. Mondal and S. Gupta, Mater. Lett., 2024, 366, 136572.
22. J.-G. Park, S. W. Park and K.-H. Hong, Nanotechnology, 2022, 33, 485706.
23. C. Soykan and H. Gocmez, Results Phys., 2019, 13, 102278.
24. J. Feng, C. H. Mak, G. Jia, B. Han, H.-H. Shen, S. P. Santoso, J.-J. Kai, M. Yuan, H. Song, J. C. Colmenares and H.-Y. Hsu, Adv. Energy Mater., 2024, 14, 2402785.
25. M. Miodyńska, T. Klimczuk, W. Lisowski and A. Zaleska-Medynska, Catal. Commun., 2023, 177, 106656.
26. Y. Guo, G. Liu, Z. Li, Y. Lou, J. Chen and Y. Zhao, ACS Sustain. Chem. Eng., 2019, 7, 15080–15085.
27. S. Purohit, K. L. Yadav and S. Satapathi, Adv. Mater. Interfaces, 2022, 9, 2200058.
28. J. Ran, J. Zhang, J. Yu, M. Jaroniec and S. Z. Qiao, Chem. Soc. Rev., 2014, 43, 7787–7812.
29. I. Alfa, H. Y. Hafeez, J. Mohammed, S. Abdu, A. B. Suleiman and C. E. Ndikilar, Int. J. Hydrogen Energy, 2024, 71, 1006–1025.
30. J. K. Nørskov, T. Bligaard, A. Logadottir, J. R. Kitchin, J. G. Chen, S. Pandelov and U. Stimming, J. Electrochem. Soc., 2005, 152, J23.
31. B. Hinnemann, P. G. Moses, J. Bonde, K. P. Jørgensen, J. H. Nielsen, S. Horch, I. Chorkendorff and J. K. Nørskov, J. Am. Chem. Soc., 2005, 127, 5308–5309.
32. T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch and I. Chorkendorff, Science, 2007, 317, 100–102.
33. L. Wu, A. Longo, N. Y. Dzade, A. Sharma, M. M. R. M. Hendrix, A. A. Bol, N. H. de Leeuw, E. J. M. Hensen and J. P. Hofmann, ChemSusChem, 2019, 12, 4383–4389.
34. H. Belhadj, Y. Messaoudi, M. R. Khelladi and A. Azizi, ChemistrySelect, 2024, 9, e202302564.
35. P. Liu, J. Zhu, J. Zhang, P. Xi, K. Tao, D. Gao and D. Xue, ACS Energy Lett., 2017, 2, 745–752.
36. H. Li, C. Tsai, A. L. Koh, L. Cai, A. W. Contryman, A. H. Fragapane, J. Zhao, H. S. Han, H. C. Manoharan, F. Abild-Pedersen, J. K. Nørskov and X. Zheng, Nat. Mater., 2016, 15, 48–53.
37. C. Peng, L. Song, L. Wang, F. Yang, J. Ding, F. Huang and Y. Wang, ACS Appl. Energy Mater., 2021, 4, 4887–4896.
38. W. Xiao, P. Liu, J. Zhang, W. Song, Y. P. Feng, D. Gao and J. Ding, Adv. Energy Mater., 2017, 7, 1602086.
39. Y. Wen, D. Li, D. Jin and H. Jin, Mater. Sci. Semicond. Process., 2024, 172, 108077.
40. K. Ahmad, W. Raza, P. Kumar, M. Q. Khan, A. Alsalme and H. Kim, Chem. Eur J., 2023, 29, e202300250.
41. M. Zhang, L. Lei and X. Zheng, Chem. Eur J., 2025, 31, e202404004.
42. Y. Wu, P. Wang, X. Zhu, Q. Zhang, Z. Wang, Y. Liu, G. Zou, Y. Dai, M.-H. Whangbo and B. Huang, Adv. Mater., 2018, 30, 1704342.
43. Y.-T. Ho, C.-H. Ma, T.-T. Luong, L.-L. Wei, T.-C. Yen, W.-T. Hsu, W.-H. Chang, Y.-C. Chu, Y.-Y. Tu, K. P. Pande and E. Y. Chang, Phys. Status Solidi Rapid Res. Lett., 2015, 9, 187–191.
44. J. Yu and X. Xu, Energy Mater. Adv., 2022, 2022(1) DOI:10.34133/2022/9836095.
45. T. Wang, D. Yue, X. Li and Y. Zhao, Appl. Catal., B, 2020, 268, 118399.
46. X. Meng, S. Wang, C. Zhang, C. Dong, R. Li, B. Li, Q. Wang and Y. Ding, ACS Catal., 2022, 12, 10115–10126.
47. R. Xie, K. Fang, Y. Liu, W. Chen, J. Fan, X. Wang, Y. Ren and Y. Song, J. Mater. Sci., 2020, 55, 11919–11937.
48. S. Chakraborty, D. Kalita, S. Agarwal, S. Vashishth, N. Mathew, S. Maity, D. Goud, A. Rao, S. C. Peter, A. K. Singh and M. Eswaramoorthy, Chem. Mater., 2024, 36, 6547–6557.
49. W. Li, F. Wang, Z. Zhang and S. Min, Chem. Commun., 2021, 57, 7774–7777.
50. K. F. Lim, J. Chem. Educ., 2006, 83, 1465.
51. M. Xiao, M. Hao, M. Lyu, E. G. Moore, C. Zhang, B. Luo, J. Hou, J. Lipton-Duffin and L. Wang, Adv. Funct. Mater., 2019, 29, 1905683.
52. R. Das Adhikari, M. J. Patel, H. Baishya, D. Yadav, M. Kalita, M. Alam and P. K. Iyer, Chem. Soc. Rev., 2025, 54, 3962–4034.

---