betaFeedback

WorksPhysics and Astronomy

Additive engineering for Sb2_2S3_3 indoor photovoltaics with efficiency exceeding 17%

Abstract

Indoor photovoltaics (IPVs) have attracted increasing attention for sustainably powering Internet of Things (IoT) electronics. Sb2_2S3_3 is a promising IPV candidate material with a bandgap of ~1.75 eV, which is near the optimal value for indoor energy harvesting. However, the performance of Sb2_2S3_3 solar cells is limited by nonradiative recombination, closely associated with the poor-quality absorber films. Additive engineering is an effective strategy to improved the properties of solution-processed films. This work shows that the addition of monoethanolamine (MEA) into the precursor solution allows the nucleation and growth of Sb2_2S3_3 films to be controlled, enabling the deposition of high-quality Sb2_2S3_3 absorbers with reduced grain boundary density, optimized band positions and increased carrier concentration. Complemented with computations, it is revealed that the incorporation of MEA leads to a more efficient and energetically favorable deposition for enhanced heterogeneous nucleation on the substrate, which increases the grain size and accelerates the deposition rate of Sb2_2S3_3 films. Due to suppressed carrier recombination and improved charge-carrier transport in Sb2_2S3_3 absorber films, the MEA-modulated Sb2_2S3_3 solar cell yields a power conversion efficiency (PCE) of 7.22% under AM1.5G illumination, and an IPV PCE of 17.55% under 1000 lux white light emitting diode (WLED) illumination, which is the highest yet reported for Sb2_2S3_3 IPVs. Furthermore, we construct high performance large-area Sb2_2S3_3 IPV modules to power IoT wireless sensors, and realize the long-term continuous recording of environmental parameters under WLED illumination in an office. This work highlights the great prospect of Sb2_2S3_3 photovoltaics for indoor energy harvesting.

  • Corresponding authors:

[email protected] (Jiejie Chen)

[email protected] (Robert L. Z. Hoye)

[email protected] (Ru Zhou)

Introduction

The rapid development of Internet of Things (IoT) technologies is leading to an ongoing exponentially growing market of smart devices[1, 2]. Such end nodes are usually designed for cyclic operation and low power consumption (microwatts to milliwatts), and having a reliable and sustainable long-term power supply is critical for the success of IoT technologies. Currently, autonomous IoT nodes are mostly powered using batteries. However, the short lifespan of batteries not only limits the power consumption and size of IoT devices, but also restricts the applications to the cases which are compatible with battery replacement and maintenance. Solely relying on batteries to power IoT devices might not sustain the rapidly growing size of the IoT ecosystem as it proceeds to a trillion nodes, where it is predicted that >100>100 billion batteries will have to be disposed of each year[3]. In this regard, integrated energy systems for harvesting ambient energy (e.g., indoor lighting, mechanical energy, or thermal energy) are now expected to serve as alternatives, or work complementarily to batteries[1, 4, 5]. Indoor photovoltaics (IPVs), which capture energy from ambient lighting (either from artificial light sources, or daylight), have significant potential to provide sustainable power for driving wireless IoT nodes that communicate using a range of protocols, such as Bluetooth low energy, RFID tags, LoRa, passive Wi-Fi, Zigbee, ANT, etc[6, 7]. Indeed, IPVs are deployable in view of their reliance on radiative energy transfer and indoor lighting is ubiquitously available and predictable. Furthermore, IPVs afford relatively high energy density compared to other ambient energy harvesting technologies[3]. Therefore, the development of high performance IPVs is important for sustainable IoT applications[8]. Indoor illuminances typically range from 200-500 lux for private areas, and 500-1500 lux for office, commercial and industrial areas. For commonly used cold white light emitting diodes (WLEDs), an illuminance of 500-1000 lux (0.140.14-0.28 mW cm−20.28\,\mathrm{mW\,cm^{-2}}) is 300-700 times lower compared to “1-sun” (AM1.5G, 100 mW cm−2100\,\mathrm{mW\,cm^{-2}})[9]. The narrow emission spectra of indoor light sources (e.g., LEDs and fluorescent lamps (FLs)) range from 400 to 700 nm, which determines the optimal bandgap for indoor light-absorbing materials to be around 1.80-2.00 eV[10]. Therefore, crystalline silicon (c-Si), which dominates the outdoor PVs market, would not be well suited for indoor energy harvesting, due to its small bandgap (~1.12 eV1.12\ \mathrm{eV}), as well as high dark current densities, which together limit its performance under the lower irradiances from indoor lighting[11].

The state-of-the-art commercial solution up to now for IPVs is hydrogenated amorphous silicon (a-Si:H), with a bandgap of 1.71.7-2.0 eV2.0\ \mathrm{eV}. The PCEs of commercially available standard a-Si:H module devices typically range from 4.4%4.4\%-9.2%9.2\%[3]. To date, emerging next-generation solar cells, including organic photovoltaics (OPVs), dye-sensitized solar cells (DSSCs) and lead-halide perovskite solar cells (LHPSCs) have also been investigated for IPV applications[8, 12, 13, 14, 15, 16]. However, DSSCs and OPVs are limited by the use of toxic solvents and expensive small molecules; LHPSCs involve the use of high contents of lead, which exceeds the limits of the EU restriction of hazardous substances directive (RoHS), and this may act as a barrier to commercialization. A stable material that can be simply processed and is comprised of RoHS-compliant elements is missing.

Sb2S3\mathrm{Sb_2S_3} is an earth-abundant, low-toxicity and stable material with a bandgap of ~1.75 eV1.75\ \mathrm{eV}, close to the optimal bandgap value for IPV application[10]. Moreover, Sb2S3\mathrm{Sb_2S_3} possesses excellent materials and optoelectronic properties, including simple binary composition, quasi-one-dimensional crystal structure, high absorption coefficients (10410^4-105 cm−110^5\ \mathrm{cm}^{-1}), low melting points (~500 ∘C500\ ^\circ\mathrm{C}) and high vapor pressures. These merits enable the low-temperature fabrication of high performance, flexible and lightweight Sb2S3\mathrm{Sb_2S_3} devices for powering IoT electronics. According to calculations by Hoye et al., the spectroscopic limited maximum efficiency (SLME) reaches 47% under 1000 lux white LED lighting[3]. However, up to now, the development of Sb2S3\mathrm{Sb_2S_3} solar cells has mainly focused on its performance under 1-sun illumination (i.e., outdoor photovoltaics). Since the illumination intensity and emission spectra of indoor light sources in most building and office environments, such as FLs and LEDs, are strikingly different from 1-sun conditions, it is necessary to carefully study the IPV performance of Sb2S3\mathrm{Sb_2S_3} photovoltaics[4, 17]. As far as we know, only Zheng et al. reported the photovoltaic performance under low-intensity illuminations for Sb2S3\mathrm{Sb_2S_3} devices, which delivered 16.37% indoor efficiency illuminated by a 1000 lux WLED[18]. The application of Sb2S3Sb_2S_3 devices for powering IoT electronics has, to our knowledge, never been demonstrated.

Thanks to endeavors from a growing community of researchers focusing on this chalcogenide photovoltaic materials system, great achievements have been made for Sb2S3Sb_2S_3 solar cells. To date, the mesoporous and planar Sb2S3Sb_2S_3 photovoltaics have delivered record efficiencies of 7.5% and 8.0% under 1-sun illumination, respectively[19, 20]. However, such efficiency values still lag far behind the theoretical maximum efficiency of 28.64% according to the Shockley–Queisser limit[21]. The severe charge-carrier recombination at grain boundaries (GBs) and interfaces is one of the critical problems that limits the performance of Sb2S3Sb_2S_3 solar cells[21, 22]. Over the past decade, researchers have explored a variety of approaches to enhance the efficiency of Sb2S3Sb_2S_3 solar cells, such as ion doping[21, 23], additive engineering[18, 24], surface passivation[25], configuration optimization[26], etc. Additive engineering has attracted tremendous attention because it is a simple strategy to regulate the deposition processes as well as the film properties for solution-processed Sb2S3Sb_2S_3 films, such as deposition rate, growth orientation, film morphology, etc[20]. To date, a number of additives, including sodium dodecyl sulfate[18], thioacetamide[20], tartaric acid[27], tetramethylammonium hexafluorophosphonate[24], and ammonium sulfide[28] have been investigated to adjust the chemical bath deposition or hydrothermal deposition processes of Sb2S3Sb_2S_3 films. For instance, Wang et al. used thioacetamide together with sodium thiosulfate to act as dual sulfur sources and achieved noticeable improvements in the device efficiency due to the advantageous hydrolysis cooperation between two sulfur sources[20]. Han et al. employed tetramethylammonium hexafluorophosphonate which could coordinate with Sb3+Sb^{3+} due to the multidentate anchoring and further improve the morphology of Sb2S3Sb_2S_3 films as well as reduce the trap states[24]. Therefore, additive engineering is promising to optimize the film properties and further enhance the photovoltaic performance of Sb2S3Sb_2S_3 solar cells.

As a complexing agent that has hydrophilic hydroxyl group (−OH-\mathrm{OH}) and amino group (−NH2-\mathrm{NH}_2), monoethanolamine (MEA) is expected to be capable of effectively regulating the nucleation and growth environment for film deposition[29, 30]. In this work, we obtained high-quality Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films through an additive-modulated hydrothermal deposition method by incorporating a small amount of MEA into the precursor solution. This additive engineering strategy increases the grain size and accelerates the deposition rate of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films, which further improve the optoelectronic properties of absorber films. The MEA-modulated planar Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cells delivered an impressive PCE of 7.22% under 1-sun illumination and a remarkable indoor PCE of 17.55% under 1000 lux LED illumination. We further demonstrated the use of large-area Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} modules to power low-power portable electronics.

Results and discussion

Deposition of high-quality Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films

In this work, we employed a hydrothermal deposition method to deposit Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films onto FTO/SnO2/CdS\mathrm{FTO/SnO}_{2}\mathrm{/CdS} substrate. Antimony potassium tartrate (APT) and sodium thiosulfate were used as antimony source and sulfur source, respectively. A small amount of MEA was used as an additive in the precursor solution to regulate the deposition of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films. Figure 1a shows top-view scanning electron microscopy (SEM) images of as-deposited Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} thin films prepared without the additive (the control sample), and with different concentrations of MEA additives. For the convenience of description, the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films obtained with MEA concentrations of 3 μg/mL\mu\mathrm{g}/\mathrm{mL}, 4 μg/mL\mu\mathrm{g}/\mathrm{mL}, and 5 μg/mL\mu\mathrm{g}/\mathrm{mL} in the hydrothermal deposition are labeled as MEA-3, MEA-4, and MEA-5, respectively. As shown, the control sample displays grain sizes smaller than 5 μm\mu\mathrm{m}, consistent with that reported for hydrothermally-deposited Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films[21]. It is worth noting that we have previously established that SEM images are suitable for evaluating the grain size of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films[21]. With the addition of MEA, the grain size of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films involves a significant increase, and some Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} grains exceed 10 μm\mu\mathrm{m} with an MEA concentration of 4 μg/mL\mu\mathrm{g}/\mathrm{mL}. The increase in grain size is accompanied by a decrease in GB density, which can be defined as the GB length per unit area. As shown in Figure 1c, the GB density of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films decreases from 0.434 ±\pm 0.027 μm cm−2\mu\mathrm{m}\,\mathrm{cm}^{-2} for the control sample to 0.319 ±\pm 0.006, 0.303 ±\pm 0.010, and 0.229 ±\pm 0.006 μm cm−2\mu\mathrm{m}\,\mathrm{cm}^{-2} for the MEA-3, MEA-4, and MEA-5 samples, respectively. The reduced GB density of the absorber films is expected to favor the suppression of non-radiative recombination[31]. The cross-sectional SEM images (Figure 1b) reveal that the Sb2S3\mathrm{Sb_2S_3} films shows a gradual increase of thickness with the addition of MEA. The film thickness increases from 328 ± 11 nm for the control sample to 399 ± 8 nm, 428 ± 6 nm, and 480 ± 5 nm for the MEA-3, MEA-4, and MEA-5 samples, respectively. This implies that the addition of MEA to the precursor solution accelerates the deposition rate of Sb2S3\mathrm{Sb_2S_3} films. The 2D and 3D morphology spatial maps of atomic force microscopy (AFM) characterization (Supplementary Fig. 1) further indicate that the MEA-4 film samples show larger grain sizes with fewer GBs compared to the control samples. Moreover, the addition of MEA also leads to a reduction in the root-mean-square (RMS) roughness of absorber films from 18.4 nm (the control sample) to 16.2 nm (the MEA-4 sample). The flatter Sb2S3\mathrm{Sb_2S_3} films should be beneficial for the formation of high-quality heterojunction between the absorber layer and the solution-processed hole transport layer (HTL) on top.

Image not extracted

Figure 1. Morphological and structural properties of Sb2S3\mathrm{Sb_2S_3} films. a, b Top-view and cross-sectional SEM images and c GB densities of hydrothermally deposited Sb2S3\mathrm{Sb_2S_3} thin films prepared without (control) and with MEA additives. Three samples were measured to determine the mean GB density values shown, and the error bars represent the standard deviation. d, e XRD patterns and texture coefficients of the dominant (020), (240), (211) and (221) peaks of the control film and MEA-Sb2S3\mathrm{Sb_2S_3}

films prepared with different concentrations of MEA additives.

X-ray diffraction (XRD) patterns of Sb2S3\mathrm{Sb_2S_3} films (Figure 1d) show that the diffraction peaks can be indexed to pure orthorhombic Sb2S3\mathrm{Sb_2S_3} (JCPDS #42-1393), with the background peaks from FTO substrates as marked by the asterisks[22]. The diffraction peaks of Sb2S3\mathrm{Sb_2S_3} films reveal negligible shifts with the addition of MEA, indicating that the MEA additive could not be incorporated into the Sb2S3\mathrm{Sb_2S_3} lattice. Compared to the control film, the MEA-Sb2S3\mathrm{Sb_2S_3} films exhibit slightly reduced intensities for (020) and (240) peaks as well as enhanced intensities for (211) and (221) peaks. The texture coefficients (TC) of (020), (240), (211) and (221) planes of Sb2S3\mathrm{Sb_2S_3} films are further calculated to evaluate the changes in the film orientation, as shown in Figure 1e. It can be seen that the addition of MEA results in the (211) and (221) preferred orientations for Sb2S3\mathrm{Sb_2S_3} films. As we know, Sb2S3\mathrm{Sb_2S_3} has a quasi-1D crystal structure consisting of numerous parallel (Sb4S6)n(\mathrm{Sb_4S_6})_n ribbons with van der Waals interactions between them[22]. Hence the carrier transport along (Sb4S6)n(\mathrm{Sb_4S_6})_n ribbons is much more efficient than the charge hopping between ribbons, and the enhanced (hk1)(hk1) preferred orientation should favor for the charge-carrier transport in the vertically-structured devices.

Conductive AFM (c-AFM) characterization of the control and MEA-4 films (Supplementary Fig. 2a and b) reveal that the MEA-4 samples display higher current on the film surface than the control sample. The surface current intensity distributions (Supplementary Fig. 2c and d) show that both samples have obvious current fluctuations at GBs, which might be caused by the reduction in the carrier density at GBs[21]. Compared to the control sample, the MEA-4 films show reduced current fluctuations and increased average current intensity. This should be associated with the decrease of GBs for the MEA-4 sample. The alleviated current fluctuation on the absorber films would facilitate the collection of charge-carriers. To further evaluate the conductivity change of Sb2S3\mathrm{Sb_2S_3} films, we prepared devices with the simple device structure of FTO/Sb2S3\mathrm{Sb_2S_3}/Au and measured II-VV curves in the dark (Supplementary Fig. 3). The conductivity of films can be calculated by the equation[26]: I=σ0AL−1VI=\sigma_0 A L^{-1}V, where σ0\sigma_0 is the conductivity, AA represents the area of the apparatus (0.06 cm2^2), LL is the thickness of Sb2_2S3_3 layer (here the thicknesses of the control and MEA-4 films are 339 nm and 427 nm, respectively), and II and VV are the current and the voltage, respectively. The estimated conductivities of the control and MEA-4 films are 3.55×10−53.55 \times10^{-5} and 5.96×10−55.96 \times10^{-5} S cm−1^{-1}, respectively. The increase in the conductivity of MEA-4 film might be associated to the increase in grain size, consistent with the c-AFM results[32].

Mechanisms of additive engineering

We now investigate the underlying mechanism for the impact of the MEA additives on the film properties of Sb2_2S3_3. First, we hypothesize that the MEA additive is absent in as-deposited Sb2_2S3_3 films, since the decomposition temperature of MEA molecules (~200 °C) is much lower than the post-annealing temperature for Sb2_2S3_3 films (~370 °C)[33]. The X-ray photoelectron spectroscopy (XPS) and Raman characterization of the control and MEA-4 films were performed to test this hypothesis. XPS spectra (Supplementary Fig. 4) and Raman spectra (Supplementary Fig. 5) reveal that both of the control and MEA-4 film samples share nearly identical characteristic peaks, indicating that the incorporation of MEA additives exerts no impact on the elemental composition and chemical states of Sb2_2S3_3 films. Hence we speculate that MEA only plays roles in the hydrothermal deposition process of Sb2_2S3_3 absorber films.

of [Sb(MEA)3]3+[\mathrm{Sb}(\mathrm{MEA})_{3}]^{3+} into [Sb(MEA)2]3+[\mathrm{Sb}(\mathrm{MEA})_{2}]^{3+} and further into [SbMEA]3+[\mathrm{SbMEA}]^{3+} on the CdS substrate with the addition of MEA. c The binding energy between the Sb3+\mathrm{Sb}^{3+} complexes ([SbMEA]3+[\mathrm{SbMEA}]^{3+} and APT) and the CdS substrate. d The reaction energy profile diagram of [SbMEA]3+[\mathrm{SbMEA}]^{3+} binding to the CdS substrate and releasing the MEA ligand. e Schematic illustration of the evolution of the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} deposition process on the substrate with the addition of MEA.

In this study, MEA as the complexing agent was incorporated into the precursor solution to regulate the deposition process. MEA can be uniformly dissolved into the precursor solution. To investigate the possible mechanisms of additive engineering, density functional theory (DFT) calculations were employed to examine the role of ligands and their interactions with the substrate. The models of [Sb(MEA)n]3+[\mathrm{Sb}(\mathrm{MEA})_{n}]^{3+} complexes were constructed with different coordination numbers, where n=1n = 1, 2, and 3. Coordination numbers exceeding 3 were excluded from consideration due to significant steric hindrance among MEA molecules. As shown in Figure 2a, [Sb(MEA)3]3+[\mathrm{Sb}(\mathrm{MEA})_{3}]^{3+} manifests the lowest formation energy, thereby signifying it as the most stable coordination state. Moreover, the dissociation process of [Sb(MEA)3]3+[\mathrm{Sb}(\mathrm{MEA})_{3}]^{3+} into [Sb(MEA)2]3+[\mathrm{Sb}(\mathrm{MEA})_{2}]^{3+} and subsequently into [Sb(MEA)]3+[\mathrm{Sb}(\mathrm{MEA})]^{3+} necessitates an external energy input of merely 1.003 eV/complex and 0.943 eV/complex, respectively (Figure 2b and Supplementary Fig. 6). This indicates that continuous thermal energy input into the system allows the coexistence of the three forms of MEA-complexed Sb3+\mathrm{Sb}^{3+} during the hydrothermal deposition stage. Ultimately, Sb3+\mathrm{Sb}^{3+} might bind to the CdS substrate in the form of [SbMEA]3+[\mathrm{SbMEA}]^{3+}, owing to the complete exposure of Sb3+\mathrm{Sb}^{3+} in this configuration. This specific configuration of the ligand is advantageous for the deposition and growth of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} on the substrate. In the hydrothermal deposition of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films, Sb3+\mathrm{Sb}^{3+} ions was introduced in the form of C8H4K2O12Sb2\mathrm{C}_{8}\mathrm{H}_{4}\mathrm{K}_{2}\mathrm{O}_{12}\mathrm{Sb}_{2}, necessitating an examination of the binding capacities of this specific coordination complex at the interface. The strong chelating effect of the tartrate ions introduces complexity to the binding process of Sb3+\mathrm{Sb}^{3+} to the substrate, thereby impeding their effective stabilization on the substrate surface[32]. Moreover, the considerable steric hindrance presented by the molecular structure of C8H4K2O12Sb2\mathrm{C}_{8}\mathrm{H}_{4}\mathrm{K}_{2}\mathrm{O}_{12}\mathrm{Sb}_{2} further restricts the direct interaction between Sb3+\mathrm{Sb}^{3+} and the substrate. Nevertheless, the incorporation of MEA can markedly mitigate this limitation. The unique bridging capacity of MEA allows its molecular termini to adhere to the substrate, acting as a molecular bridge that facilitates the connection between Sb3+\mathrm{Sb}^{3+} with the CdS substrate. As shown in Figure 2c, compared to APT, the [Sb(MEA)]3+[\mathrm{Sb}(\mathrm{MEA})]^{3+} complex exhibits a more negative binding energy with the CdS substrate. Hence the bridging effect of MEA significantly enhances the affinity and subsequent binding of Sb3+\mathrm{Sb}^{3+} to the CdS substrate, effectively overcoming the problems posed by the chelating effect of tartrate ions and the steric hindrance of APT. During this process, MEA possesses the capability to sequester free Sb3+\mathrm{Sb}^{3+} within the solution, thereby elevating the likelihood of their interaction with the substrate. This dual functionality not only acts as a physical bridge but also enhances the local concentration of Sb3+\mathrm{Sb}^{3+} in proximity to the substrate surface, increasing the deposition rate of Sb3+\mathrm{Sb}^{3+} on the substrate. Furthermore, computational analyses have substantiated that the release of ligands facilitated by the addition of MEA requires minimal energy input, underscoring the high efficiency of MEA in promoting the attachment of Sb3+\mathrm{Sb}^{3+} to the substrate and further the reaction with S2−\mathrm{S}^{2-} to form Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} (Figure 2d). Therefore, the MEA additive plays a crucial role in overcoming the challenges faced by APT, culminating in a more efficient and energetically favorable deposition process.

Image not extracted

Figure 2. Mechanisms of additive engineering for the deposition of Sb2_2S3_3 films. a The formation energy of [Sb(MEA)n_n]3+^{3+}, n=1n = 1, 2, and 3. b Schematic diagram depicting the sequential dissociation

Based on these discussions, the incorporation of MEA into the precursor contributes to the precise control of the deposition conditions, which is essential for the preparation of high-quality Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films[34]. We schematically depict the evolution of the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} deposition process on the substrate with the incorporation of MEA, as shown in Figure 2e. It is known that, during the initial hydrothermal deposition stage, two typical nucleation processes occur, i.e., homogeneous nucleation in the precursor solution and heterogeneous nucleation on the substrate[21]. In the typical hydrothermal deposition process, the strong chelating effect of ATP would result in the formation of many aggregates in the precursor solutions[32]. These APT aggregates would act as nucleation centers for homogeneous nucleation and growth of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} in the precursor solution. As a result, a large number of suspended Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} particles would form in the precursor solution during the hydrothermal deposition without the use of additives[29]. In contrast, the incorporation of MEA results in the formation of MEA-complexed Sb3+\mathrm{Sb}^{3+}, which would decrease the amount of APT aggregates and lead to a more uniform distribution of Sb3+^{3+} in the precursor solution. The formation of fewer aggregates leads to increases in the concentration of Sb3+^{3+} in the precursor solution, as well as increases in the density of reactive ions on the substrate. The deposition process of Sb2_2S3_3 with the presence of MEA complexing agent can be described by Equations (S1-S8) (Supplementary Note S1)[35]. According to computational analyses, the unique bridging capacity of MEA leads to a decrease in the energy barrier of heterogeneous nucleation and thus promotes the nucleation of Sb2_2S3_3 on the substrate. This further decreases the sites for homogeneous nucleation, in good agreement with the reduction of suspended particles in the reactor at the end of hydrothermal deposition (Supplementary Fig. 7). As a result, the enhanced heterogeneous nucleation and suppressed homogeneous nucleation lead to the formation of dense, uniform and large-grain Sb2_2S3_3 films[21].

Here, to exclude the influence of pH on the deposition rate of Sb2_2S3_3, we measured the pH of the precursor solution without and with the addition of MEA (Supplementary Fig. 8), which reveals that the addition of a small amount of MEA exerts no significant change in the pH of the precursor solution. Therefore, the changes in the morphology and structural properties of Sb2_2S3_3 films should be associated with the addition of MEA. The multifaceted role of MEA plays a pivotal part in regulating the nucleation and growth processes of Sb2_2S3_3 films, resulting in the deposition of ideal absorber films with minimized GB density for constructing solar cells.

Device performance of Sb2_2S3_3 solar cells

We prepared planar Sb2_2S3_3 solar cells with a typical p-i-n configuration of FTO/SnO2_2/CdS/Sb2_2S3_3/Spiro-OMeTAD/Au, as illustrated in Figure 3a. The band structures of each functional layers in Sb2_2S3_3 solar cells are shown in Figure 3b. The energy levels of Sb2_2S3_3 were obtained from ultraviolet photoelectron spectroscopy (UPS) measurements (Supplementary Fig. 9), while the energy levels of other functional layers are compiled from the literature. The detailed analysis of UPS spectra of the control and MEA-4 Sb2_2S3_3 film samples is given in Supplementary Note S2. The corresponding Fermi level (EFE_\mathrm{F}), valence band maximum (VBM) and conduction band minimum (CBM) of the control and MEA-4 Sb2S3\mathrm{Sb_2S_3} films are summarized in Supplementary Table 1. Here, the bandgap values of the control and MEA-Sb2S3\mathrm{Sb_2S_3} films are determined to be 1.74 and 1.73 eV, respectively, based on the Tauc plots of UV-vis-NIR absorption spectra (Supplementary Fig. 10). As shown, the energy levels of Sb2S3\mathrm{Sb_2S_3} match well with the electron transport layer (CdS) and HTL (Spiro-OMeTAD), favoring for the efficient charge transport from absorber layer towards two electrodes. Moreover, compared to the control Sb2S3\mathrm{Sb_2S_3} film, the Fermi level of MEA-Sb2S3\mathrm{Sb_2S_3} is shifted up towards CBM. This indicates an increased electron carrier concentration for n-type Sb2S3\mathrm{Sb_2S_3}. Compared to the control film, the upward shift of CBM for the MEA-4 film facilitates the photogenerated carrier transport from the absorber layer to the electron transport layer. Figure 3c gives current density-voltage (JJ-VV) curves of best-performing control and MEA-Sb2S3\mathrm{Sb_2S_3} solar cells. As shown, the control device delivers a VOCV_{\mathrm{OC}} of 766 mV, a JSCJ_{\mathrm{SC}} of 14.89 mA cm−2^{-2}, an FF of 54.49%, and a PCE of 6.22%, while the MEA-4 device yields an enhanced PCE of 7.22%, with a VOCV_{\mathrm{OC}} of 787 mV, a JSCJ_{\mathrm{SC}} of 16.12 mA cm−2^{-2}, and an FF of 56.92%. That is, the incorporation of MEA affords a 16% relative efficiency enhancement for Sb2S3\mathrm{Sb_2S_3} solar cells compared to the control device (or 1% absolute efficiency increase). The performance enhancement of Sb2S3\mathrm{Sb_2S_3} solar cells should be associated with the improvements in materials and optoelectronic properties of the Sb2S3\mathrm{Sb_2S_3} absorber films. The corresponding external quantum efficiency (EQE) spectra of both devices are shown in Figure 3d. It can be seen that both devices exhibit a wide light response ranging from 350 to 750 nm. Compared to the control device, the MEA-4 device exhibits an enhanced EQE value close to 90% at around 520 nm. The obtained integrated JSCJ_{\mathrm{SC}} values are 14.83 and 16.02 mA cm−2^{-2} for the control and MEA-4 devices, respectively. The values are consistent with the JSCJ_{\mathrm{SC}} obtained from JJ-VV curves (within 5% deviation). Figure 3e-Figure 3h displays the distribution statistics of the PCE, VOCV_{\mathrm{OC}}, JSCJ_{\mathrm{SC}}, and FF for the Sb2S3\mathrm{Sb_2S_3} devices, and the mean and champion performance parameters are summarized in Table 1. As shown, the photovoltaic performance of Sb2S3\mathrm{Sb_2S_3} solar cells first increases and then decreases with the increasing amount of MEA. MEA-Sb2S3\mathrm{Sb_2S_3} devices show enhanced photovoltaic performance compared to the control sample, and an optimal device efficiency can be obtained based on the MEA-4 films. The MEA-4 device delivers an enhanced average efficiency of 6.88% in contrast to that of 5.99% for the control device.

Image not extracted

Figure 3. Photovoltaic performances of Sb2_2S3_3 solar cells under AM1.5G illumination. a Schematic illustration of the device configuration of planar Sb2_2S3_3 solar cells. b Band alignment of the components of Sb2_2S3_3 solar cells. c J-V curves of best-performing control and MEA-Sb2_2S3_3 solar cells, measured under 1-sun (AM1.5G, 100 mW cm−2^{-2}) illumination. d EQE spectra of the control and MEA-4 Sb2_2S3_3 solar cells. e-h Statistics of the performance parameters of the control and MEA-Sb2_2S3_3 devices obtained with the addition of different concentration of MEA.

DevicesVOCV_{\mathrm{OC}} (mV)JSCJ_{\mathrm{SC}} (mA cm−2^{-2})FF (%)PCE (%)
Control764±5764 \pm5 (774774)14.01±0.5514.01 \pm0.55 (15.1915.19)53.07±1.2753.07 \pm1.27 (55.1055.10)5.99±0.185.99 \pm0.18 (6.226.22)
MEA-3773±4773 \pm4 (782782)14.81±0.6614.81 \pm0.66 (15.8215.82)54.96±1.0454.96 \pm1.04 (56.8556.85)6.52±0.216.52 \pm0.21 (6.816.81)
MEA-4788±4788 \pm4 (795795)15.36±0.4315.36 \pm0.43 (16.1216.12)55.57±0.9855.57 \pm0.98 (56.9256.92)6.88±0.186.88 \pm0.18 (7.227.22)
MEA-5772±5772 \pm5 (784784)14.87±0.4914.87 \pm0.49 (15.8115.81)54.48±0.9654.48 \pm0.96 (56.5856.58)6.61±0.236.61 \pm0.23 (6.916.91)

Table 1. Photovoltaic performance parameters of the control and MEA-Sb2_2S3_3 solar cells with the addition of different concentration of MEA, measured under 1-sun (AM1.5G, 100 mW cm−2^{-2}) illumination. Format: mean ± standard deviation (best value)

It is worth noting that, since the addition of MEA affects the thickness of the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films, the control and MEA-Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} devices were constructed based on absorber films with different thicknesses. In order to reveal the impact of the film thickness on the device performance, we further compared the performance of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cells based on the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films with different thickness prepared by using conventional hydrothermal precursor solutions without the addition of MEA. It is found that, the device based on the 430 nm-thick Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} film that shares a similar thickness as that of the MEA-4 film involves a slight efficiency drop compared to the 339 nm-thick film (Supplementary Fig. 11a and b). This indicates that the performance enhancement of the MEA-Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} devices is not caused by an increase in light absorption. The SEM image (Supplementary Fig. 11c) further reveals that, compared to the 339 nm-thick film, the grain size of the 430 nm-thick Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films shows no negligible changes. However, the XRD patterns reflect that the (hklhkl) orientation of this 430 nm-thick Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} film is also slightly enhanced compared to the 339 nm-thick film (Supplementary Fig. 11d), similar to the scenario for MEA-Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films. Therefore, it is reasonable to speculate that the enhanced (hklhkl) orientation of the MEA-4 Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} film might be related to the increase of film thickness. The underlying mechanism for the impact of the film thickness on the film orientation needs further study. In addition, a slight drop in the device efficiency was observed when the concentration of MEA increased up to 5 μg/mL\mu\mathrm{g}/\mathrm{mL}, and this might be caused by a continuous increase in the film thickness. To verify this hypothesis, we optimized the thickness of absorption layer based on the device performance of Sb2S3\mathrm{Sb_2S_3} solar cells (Supplementary Fig. 12a and b and Supplementary Table 2). As reflected by the corresponding thickness of the MEA-4 films prepared for different times of hydrothermal growth (Supplementary Fig. 12c-e), the optimal thickness for Sb2S3\mathrm{Sb_2S_3} layer is around 428 nm, which can be obtained after 180 min of hydrothermal deposition. Further increases in the film thickness would result in the performance degradation of Sb2S3\mathrm{Sb_2S_3} solar cells. Based on such discussions, we conclude that the positive impact of the addition of MEA into the precursor solutions should be responsible for the enhanced performance of MEA-Sb2S3\mathrm{Sb_2S_3} solar cells. Moreover, in § 2.1, we propose that -OH and -NH2_2 groups in MEA contribute to form coordination complexes with Sb3+\mathrm{Sb^{3+}} in the precursor solution. In order to confirm this, we also investigated the use of diethanolamine (DEA) and triethanolamine (TEA), which share similar coordination groups, as the additives in the precursor solution to deposit Sb2S3\mathrm{Sb_2S_3} films (Supplementary Fig. 13). The corresponding device performance reveals that the incorporation of DEA and TEA additives also contribute to efficiency improvements of Sb2S3\mathrm{Sb_2S_3} solar cells (Supplementary Fig. 14).

We further performed various characterization of Sb2S3\mathrm{Sb_2S_3} solar cells to study the device physics and explore the underlying mechanisms responsible for the performance enhancements. Space charge limited current (SCLC) density measurements of electron-only device (FTO/CdS/Sb2S3/PCBM/Au\mathrm{FTO}/\mathrm{CdS}/\mathrm{Sb_2S_3}/\mathrm{PCBM}/\mathrm{Au}) are given in Figure 4a. Generally, the curves can be divided into three parts, the Ohmic region (low voltage), the trap-filled limiting (TFL) region (intermediate voltage), and the trap-free region (high voltage). At the low voltage region, the curve is generally linear. When the voltage exceeds the trap-filled limiting voltage (VTFLV_{\mathrm{TFL}}), the current increases dramatically, implying that the injected carriers have completely occupied the trap states[26]. The VTFLV_{\mathrm{TFL}} values for the control and MEA-4 device are estimated to be 0.486 V and 0.414 V, respectively. The trap state density (nτn_{\tau}) of Sb2S3\mathrm{Sb_2S_3} films can be evaluated according to the equation VTFL=qnτL2/2εrε0V_{\mathrm{TFL}} = q n_{\tau} L^2 / 2\varepsilon_{\mathrm{r}}\varepsilon_0, where qq is the elementary charge, nτn_{\tau} is the trap state density, LL is the thickness of the absorber film (here 339 nm and 427 nm for the control and MEA-4 Sb2S3\mathrm{Sb_2S_3} films, respectively), ε0\varepsilon_0 is the permittivity of free space (8.85×10−128.85 \times10^{-12} F m−1^{-1}), and εi\varepsilon_i represents the relative permittivity of Sb2_2S3_3 (6.676.67). Then the estimated nτn_\tau for the control device and the MEA-4 Sb2_2S3_3 device are 3.12×10153.12 \times10^{15} and 1.67×10151.67 \times10^{15} cm−3^{-3}, respectively. The reduced trap-state density for the MEA-4 sample indicates that the addition of MEA can effectively reduce the density of trap states in Sb2_2S3_3 films[36]. We also performed the dependence of JSCJ_\mathrm{SC} and VOCV_\mathrm{OC} on the light intensity for both devices to reveal the trap-assisted charge recombination loss mechanisms, as shown in Figure 4b and Figure 4c. The relationship between JSCJ_\mathrm{SC} (or VOCV_\mathrm{OC}) and the light intensity II can be described by the equations JSC∝IαJ_\mathrm{SC} \propto I^\alpha and VOC∝(nkBT/q)ln⁡(I)V_\mathrm{OC} \propto(nk_\mathrm{B}T/q)\ln(I), where II is the light intensity, kBk_\mathrm{B} is the Boltzmann constant, TT is the absolute temperature, qq is the elementary charge, α\alpha and nn reflect the level of charge-carrier recombination[37, 38]. The calculated α\alpha for the control and MEA-4 Sb2_2S3_3 device are 0.9040.904 and 0.9410.941, respectively. Ideally α\alpha equals 11, and α<1\alpha< 1 means the loss of photogenerated carriers caused by the incomplete charge collection. The increment in α\alpha for the MEA-4 device confirms the improvement of the carrier collection rate[38, 39]. For trap-free solar cells, the slope of VOCV_\mathrm{OC} versus ln⁡(I)\ln(I) is kBT/qk_\mathrm{B}T/q. By performing linear fitting of these curves, the slopes for the control and MEA-4 devices are obtained to be 1.53kBT/q1.53k_\mathrm{B}T/q and 1.41kBT/q1.41k_\mathrm{B}T/q, respectively. The reduced slope for the MEA-4 device indicates that the trap-assisted charge-carrier recombination is suppressed. Since the Urbach energy (EUE_\mathrm{U}) is capable of evaluating the band tailing effect, we derived EUE_\mathrm{U} value from the EQE spectrum by performing linear fitting of the ln⁡(EQE)\ln(\mathrm{EQE}) vs EUE_\mathrm{U} plots (Supplementary Fig. 15)[20]. The EUE_\mathrm{U} values of the control and MEA-4 device were fitted to be 32.3632.36 and 27.9827.98 meV, respectively. The reduction in EUE_\mathrm{U} suggests the suppression of nonradiative recombination probably due to the decrease of defect density, thus explaining the improvement of VOCV_\mathrm{OC} for MEA-Sb2_2S3_3 devices.

Image not extracted

Figure 4. Device physics and charge-carrier kinetics. a SCLC measurements of the control and MEA-4 devices based on the electron-only device configuration of FTO/CdS/Sb2_2S3_3/PCBM/Au. b, c The dependence of JSCJ_{\mathrm{SC}} and VOCV_{\mathrm{OC}} on the light intensity for the control and MEA-4 Sb2_2S3_3 solar cells. d, g Time-resolved absorption spectra obtained at 1, 10, 100 and 1000 ps, excited by a 400 nm laser pulse, e, h 2D TA spectroscopy pseudo-color images for the time-resolved absorption spectra, and f, i Transient kinetic decay (scatter) and corresponding bi-exponential curve fittings (solid line) monitored at 560 nm of the control and MEA-4 Sb2_2S3_3 films. ΔA\Delta A is defined as the variation of absorption.

The JJ-VV curves of the control and MEA-4 solar cells measured under dark conditions are collected to evaluate the carrier recombination and collection in the devices (Supplementary Fig. 16a). According to the abrupt junction JJ-VV equation as expressed in Equation (S15), and its formula manipulation as given in Equations (S16-S18), the parameters of diode ideality factor (AA), the reverse saturation current density (J0J_0), the series resistance (RSR_{\mathrm{S}}), and the shunt conductance (GG, i.e., 1/RSH1/R_{\mathrm{SH}}) can be extracted (Supplementary Fig. 16b-d and Supplementary Table 3)[40]. The detailed analysis is given in Supplementary Note S3. It can be seen that, compared to the control device, the MEA-4 device has lower RSR_{\mathrm{S}} and GG, which implies an improvement in the carrier extraction capability. The AA values extracted from the control and MEA-4 devices are 2.33 and 1.87, respectively; the reduced AA value indicates the suppression of defect states in the absorber films[25]. The calculated J0J_{0} is reduced from 9.43×10−69.43 \times10^{-6} mA cm−2^{-2} for the control device to 1.49×10−61.49 \times10^{-6} mA cm−2^{-2} for the MEA-4 device. The main reason for the generation of J0J_{0} is the charge-carrier recombination caused by deep-level defects in the devices[41], and the reduced J0J_{0} implies the suppression of the defect-induced recombination, contributing to the increase of VOCV_{\mathrm{OC}} for the MEA-4 device.

The electrochemical impedance spectroscopy was also employed to study the carrier recombination and transport behaviors of devices. The Nyquist plots of the impedance spectra for the control and MEA-4 devices were measured under dark conditions at the bias voltage of 0.7 V (Supplementary Fig. 17), and the resistance and capacitance parameters were obtained by performing the curve fitting based on the equivalent circuit diagram given in the inset (Supplementary Table 4). Here RserR_{\mathrm{ser}}, and RrecR_{\mathrm{rec}} in the equivalent circuit represent the internal series resistance and recombination resistance, respectively. CPE consists of the interface capacitance (CPE-T) and the ideal capacitance (CPE-P)[42]. RrecR_{\mathrm{rec}} is related to charge-carrier recombination and its value is equal to the diameter of the semicircle in the Nyquist plots. As shown, the MEA-4 device delivers an increased recombination resistance at the CdS/Sb2_2S3_3 interface (Rrec=16.02R_{\mathrm{rec}} = 16.02 kΩ cm2^{2}) compared to the control device (Rrec=9.62R_{\mathrm{rec}} = 9.62 kΩ cm2^{2}). The significant increase in the recombination resistance is suggested to suppress the charge recombination and improve the charge collection in solar cells.

Ultrafast transient absorption (TA) spectroscopy measurements were performed to understand the charge-carrier kinetics of FTO/SnO2_2/CdS/Sb2_2S3_3 films. As shown in Figure 4d and g, the TA spectra of both the control sample and the MEA-4 sample display distinct negative ground state bleach (GSB) peaks and positive photoinduced absorption (PIA) peaks. The GSB peaks observed at the wavelengths of 460–510 nm and 610–680 nm can be attributed to the state filling of CdS and the ground state absorption of Sb2_2S3_3, respectively. The PIA peak at the wavelength of 520 to 620 nm can be ascribed to the formation of sulfide radicals (S−\mathrm{S}^{-}) as a result of the localization of photogenerated holes on the S atom within the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} lattice[43, 44]. The transient dynamics are extracted from the pseudocolor images for the TA spectra of both samples (Figure 4e and h). The transient kinetic decays monitored at 560 nm for both the control and MEA-4 films are presented in Figure 4f and i, respectively, which can be well fitted by a phenomenological biexponential equation y=A1exp⁡(−t/τ1)+A2exp⁡(−t/τ2)y=A_{1}\exp(-t/\tau_{1})+A_{2}\exp(-t/\tau_{2}), where τ1\tau_{1} and τ2\tau_{2} denote short-lived and long-lived carrier lifetimes. The fitting results are listed in Supplementary Table 5. The gradual decrease of the PIA peak can be attributed to the decay of trapped holes, i.e., the S⁻ species, which we here attribute to the nonradiative carrier recombination in Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films. The MEA-4 sample delivers higher τav\tau_{\mathrm{av}} values (6.61 ns) compared to the control sample (4.93 ns). The longer lifetime in the MEA-4 sample suggests the suppression of the bulk charge-carrier recombination. Therefore, the TA analysis indicates a slower charge recombination rate in MEA-Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films. The extended lifetime of minority hole carriers would contribute to the enhancement in VOCV_{\mathrm{OC}}[21].

IPV performance of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} devices and demonstration for powering IoT electronics

The bandgaps of the Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films obtained by hydrothermal deposition in this study are 1.73-1.74 eV, which are very close to the optimum value of bandgap for indoor photovoltaics under white light illumination (1.9-2.0 eV)[7], as shown in Figure 5a. Figure 5b shows the emission spectra of AM 1.5G, 3000 K WLED and 2700 K FL, which reveals that the intensity of indoor light sources is 100-1000 times lower than that of 1-sun[8].

Image not extracted

Figure 5. IPV performance of Sb2_2S3_3 solar cells. a Calculated efficiency limit of Sb2_2S3_3 devices under 1-sun (AM1.5G, 100 mW cm−2^{-2}), 3000 K WLED at 1000 lux, and 2700 K FL at 1000 lux illuminations, respectively.[46] b Comparison of the emission spectra of AM1.5G solar, 3000 K WLED at 1000 lux, and 2700 K FL at 1000 lux. The spectral intensities of WLED and FL are amplified by 50 and 5 times, respectively, in order for these spectra to be legible in this plot. c Spectral irradiance and integrated power density of a 3000 K WLED at 1000, 500, and 200 lux. d, e JJ–VV curves of the MEA-4 Sb2_2S3_3 solar cells measured under the illumination of LED and FL at 1000, 500, and 200 lux. f Device stability of unencapsulated MEA-4 Sb2_2S3_3 solar cells stored in a cabinet with 15% relative humidity at ambient temperature.

Firstly, we measured the emission power spectra of 3000 K WLED and 2700 K FL light sources at 200, 500, and 1000 lux illuminance, as shown in Figure 5c and Supplementary Fig. 18. The corresponding optical power densities of 3000 K WLED were calculated to be 291.9 μ\muW cm−2^{-2} (1000 lux), 145.2 μ\muW cm−2^{-2} (500 lux), 62.0 μ\muW cm−2^{-2} (200 lux), respectively, while the 2700 K FL delivers 296.7 μ\muW cm−2^{-2} (1000 lux), 153.1 μ\muW cm−2^{-2} (500 lux), 64.6 μ\muW cm−2^{-2} (200 lux), respectively (Supplementary Table 6). All IPV measurements were performed in a black box to prevent the influence of stray light from the environment (Supplementary Fig. 19)[17]. The JJ-VV curves for the bestperforming MEA-4 Sb2S3\mathrm{Sb}_2\mathrm{S}_3 solar cell measured under WLED and FL illumination are shown in Figure 5d and Figure 5e, respectively. The corresponding IPV performance parameters are summarized in Table 2. As shown, under the WLED illuminations of 1000, 500, and 200 lux, the PCEs of this device are 17.55%, 17.17%, and 16.41%, respectively, with the output power densities being 51.23, 24.93, 10.18 μW cm−2\mu\mathrm{W}\,\mathrm{cm}^{-2}, respectively; under the FL illuminations of 1000, 500, and 200 lux, the device delivers the PCEs of 16.54%, 15.92%, and 15.51%, respectively, with the output power densities being 49.09, 24.37, 10.02 μW cm−2\mu\mathrm{W}\,\mathrm{cm}^{-2}, respectively. As expected, Sb2S3\mathrm{Sb}_2\mathrm{S}_3 solar cells exhibit excellent IPV performance under low-intensity indoor light illuminations. To the best of our knowledge, the IPV efficiency of 17.55% is the highest reported thus far for Sb2S3\mathrm{Sb}_2\mathrm{S}_3 photovoltaics under indoor lighting. The dependences of photovoltaic parameters (PCE, VOCV_{\mathrm{OC}}, JSCJ_{\mathrm{SC}}, FF, RSR_{\mathrm{S}}, and RSHR_{\mathrm{SH}}) of the MEA-4 device on the indoor light intensity are depicted in Supplementary Fig. 20. As shown, JSCJ_{\mathrm{SC}} is linearly proportional to the illumination intensity, VOCV_{\mathrm{OC}} decreases approximately logarithmically with the illumination intensity, and FF involves very small changes under such low illumination intensities. This enables the Sb2S3\mathrm{Sb}_2\mathrm{S}_3 device to yield the best IPV efficiency under 1000 lux. Moreover, compared to the device performance measured under the standard AM1.5G illumination, the low-intensity illumination leads to a significant reduction in JSCJ_{\mathrm{SC}} and VOCV_{\mathrm{OC}} and a remarkable enhancement in FF under the indoor light illumination[45]. Such photovoltaic parameters involve similar variation trend with those previously reported in the literature[18]. The detailed discussions are given in Supplementary Note S4. We further recorded the device stability of the unencapsulated Sb2S3\mathrm{Sb}_2\mathrm{S}_3 solar cell stored in a cabinet with 15% relative humidity at ambient temperature (Figure 5f). The device retains 96.75% of its initial efficiency after one-month storage. The excellent device stability under indoor environments demonstrates the great prospect of Sb2S3\mathrm{Sb}_2\mathrm{S}_3 solar cells for IoT applications.

MEA-4 deviceVOCV_{\mathrm{OC}} (mV)JSCJ_{\mathrm{SC}} (mA cm−2^{-2})FF (%)PCE (%)RSR_{\mathrm{S}} (kΩ cm2^2)RSHR_{\mathrm{SH}} (kΩ cm2^2)
1-sun78716.1256.927.220.0070.23
1000 lux LED6060.13463.0917.550.4448.31
500 lux LED5790.068562.8617.170.7686.57
200 lux LED5570.028963.2216.412.17206.66

Table 2. Indoor photovoltaic performance of MEA-4 Sb2_2S3_3 solar cells measured under WLED (3000 K) and white light FL (2700 K) at 1000, 500, and 200 lux.

1000 lux FL6040.12963.0016.540.4449.47
500 lux FL5710.069661.3215.920.9075.55
200 lux FL5310.029763.5415.511.97226.47

Table 3.

Having developed efficient Sb2S3\mathrm{Sb_2S_3} IPVs, we are now in the position to prototype these for the first time for powering IoT wireless electronics. It is known that a large number of IoT devices typically require very little energy with power consumption ranging from several μW\mu\mathrm{W} to a few mW\mathrm{mW}, which depends on the application and communication protocol[9]. In this work, we used the RSL 10 solar cell multi-sensor platform (RSL10-SOLARSENS-GEVK) manufactured by ON Semiconductor, which is a comprehensive development platform for battery-less IoT applications in smart buildings, smart homes, and industrial sectors (Supplementary Fig. 21). This board is based on the industry’s lowest-power Bluetooth® low-power radio (RSL10) and features sensors for environmental sensing: the BME280 smart environmental sensor. The platform also features a lightweight, low-profile 47 F storage capacitor, a programming and debugging interface, and an interface to a solar cell. To drive this sensor platform via an external power supply, a solar module with an VOCV_{\mathrm{OC}} higher than 2.65 V and an average input power exceeding 50 μW\mu\mathrm{W} is needed. The JJ-VV curve of as-prepared 1 cm2\mathrm{cm^2} area MEA-4 Sb2S3\mathrm{Sb_2S_3} solar cell (Supplementary Fig. 22) demonstrates a PCE of 15.77% under 1000 lux WLED illumination, coupled with a VOCV_{\mathrm{OC}} of 0.596 V, a JSCJ_{\mathrm{SC}} of 124 μA cm−2\mu\mathrm{A}\,\mathrm{cm^{-2}} and an FF of 62.39%. Compared with the 0.06 cm2\mathrm{cm^2} small-area cell, the device performance of this 1 cm2\mathrm{cm^2} area device only shows about a 10% decrease in PCE. We then obtained an Sb2S3\mathrm{Sb_2S_3} module (5×\times1 cm2\mathrm{cm^2}) by connecting five Sb2S3\mathrm{Sb_2S_3} devices together in series. As shown in Figure 6a, the JJ-VV curve of this Sb2S3\mathrm{Sb_2S_3} module has a VOCV_{\mathrm{OC}} of 2.92 V and an output power of 187.17 μW\mu\mathrm{W} at maximum power point under 1000 lux WLED illumination, which is sufficient to drive the RSL10-SOLARSENS-GEVK platform. Compared with commercial a-Si:H IPV products in the market, the Sb2S3\mathrm{Sb_2S_3} module here is of competitive performance in terms of PCE, VOCV_{\mathrm{OC}} and output power[3, 13].

Image not extracted

Figure 6. Demonstration for powering IoT electronics. a JJ-VV curve of Sb2_2S3_3 solar module (5×\times1 cm2^2). b The picture showing the connection between the sensor platform and the Sb2_2S3_3 module. c Schematic circuit diagram of an IPV-driven sensor device. d-f Temperature, humidity, and atmospheric pressure variations recorded by the IPV-driven sensor in the office for nearly 14 hours.

The connection between the sensor and the Sb2S3\mathrm{Sb_2S_3} module is shown in Figure 6b. Figure 6c illustrates an IPV-driven sensor device, in which the Sb2_2S3_3 module converts light energy into electrical energy and drives the sensor to operate stably. The sensor communicates with the mobile phone via Bluetooth Low Energy, which enables the real-time monitoring and collection of environmental data. Supplementary Fig. 23 presents plots of the temperature, humidity, and atmospheric pressure measured in the laboratory by the Sb2_2S3_3 module-driven sensor platform running continuously under 1000 lux WLED illumination for 140 min in ambient air without encapsulation. Furthermore, we placed it on a desk in a common office area (with a light intensity of ∼\sim726 lux) to successfully monitor the environmental parameters for nearly 14 hours, as shown in Figure 6d-Figure 6f, confirming that the Sb2_2S3_3 module can drive IoT devices under indoor light continuously and stably.

Conclusion

In this work, we demonstrate an effective additive engineering strategy to improve the optoelectronic properties of Sb2_2S3_3 films and enhance the photovoltaic performance of corresponding planar Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cells. The multifaceted role of MEA plays a pivotal part in regulating the nucleation and growth of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films for enhanced heterogeneous nucleation on the substrate. Compared to the control sample, the MEA-modulated Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} films show reduced GB density, optimized band positions, and increased carrier concentration, affording suppressed charge-carrier recombination and efficient charge collection in solar devices. The corresponding MEA-Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cell achieves a state-of-the-art efficiency of 7.22% under AM1.5G illumination, much higher than that of the control device (6.22%). This device delivers an IPV efficiency of 17.55% under 1000 lux WLED illumination, which, to our knowledge, is the highest reported so far for Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cells. Furthermore, we constructed an Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} module (5×1 cm25 \times1\ \mathrm{cm}^{2}) and successfully reported the use of this module to power IoT wireless devices for the first time. This study demonstrates the broad prospect of Sb2S3\mathrm{Sb}_{2}\mathrm{S}_{3} solar cells for IPV applications.

Acknowledgements

X.C. and X.S. contributed equally to this work. The financial support by the National Natural Science Foundation of China (nos. 52371219 and U19A2089), and the National Natural Science Foundation of Hefei City (no. 2022024) is greatly acknowledged. Y. F., J. Y. and R. L. Z. H. acknowledge funding from a UKRI Frontier Grant (no. EP/X022900//1). Y.-T. H. and R. L. Z. H. acknowledge funding from the Engineering and Physical Sciences Research Council (grant no. EP/V014498/2). R. L. Z. H. thanks the Royal Academy of Engineering for support through the research fellowships scheme (no. RF/201718/17101), and also the Henry Royce Institute for support through the Industrial Collaborative Programme, funded by EPSRC (no. EP/X527257/1).

Author contributions

R.Z. and X.C. designed the experiments and analyzed the data, and also conceived of the idea for this manuscript, with input from R.L.Z.H on the idea of IPV development and prototyping. X.C. carried out the experiments and device optimizations. X.S. carried out the DFT calculations. J.Z., P.X., L.W., Y.F., J.Y., Y.-T. H., B.Y., D.-J. X, T.C., J.C., R.L.Z.H. and R.Z. assisted in experiments and data analysis. R.Z., X.C., and R.L.Z.H. wrote the manuscript. All authors commented on the manuscript.

Conflict of interest

The authors declare no conflict of interest.

References

  1. [1]Zhu, Z. et al. Indoor Photovoltaic Fiber with an Efficiency of 25.53% under 1500 Lux Illumination. Adv. Mater. 36, 2304876 (2024).DOI
  2. [2]Wojciechowski, K. & Forgács, D. Commercial Applications of Indoor Photovoltaics Based on Flexible Perovskite Solar Cells. ACS Energy Lett. 7, 3729-3733 (2022).DOI
  3. [3]Pecunia, V., Occhipinti, L. G. & Hoye, R. L. Z. Emerging Indoor Photovoltaic Technologies for Sustainable Internet of Things. Adv. Energy Mater. 11, 2100698 (2021).DOI
  4. [4]Polyzoidis, C., Rogdakis, K. & Kymakis, E. Indoor Perovskite Photovoltaics for the Internet of Things—Challenges and Opportunities toward Market Uptake. Adv. Energy Mater. 11, 2101854 (2021).
  5. [5]Ye, C. et al. An Integrated Solar Panel with a Triboelectric Nanogenerator Array for Synergistic Harvesting of Raindrop and Solar Energy. Adv. Mater. 35, 2209713 (2023).DOI
  6. [6]Guo, Z. L., Jena, A. K. & Miyasaka, T. Halide Perovskites for Indoor Photovoltaics: The Next Possibility. ACS Energy Lett. 8, 90-95 (2023).
  7. [7]Yan, B. et al. Indoor photovoltaics awaken the world’s first solar cells. Sci. Adv. 8, eadc9923 (2022).DOI
  8. [8]Cui, Y. et al. Wide-gap non-fullerene acceptor enabling high-performance organic photovoltaic cells for indoor applications. Nat. Energy 4, 768-775 (2019).DOI
  9. [9]Müller, D. et al. Indoor Photovoltaics for the Internet-of-Things - A Comparison of State-of-the-Art Devices from Different Photovoltaic Technologies. ACS Appl. Energy Mater. 6, 10404-10414 (2023).DOI
  10. [10]Zhao, C. et al. Evaporated tellurium thin films for p-type field-effect transistors and circuits. Nat Nanotechnol 15, 53-58 (2020).DOI
  11. [11]He, X. et al. 40.1% Record Low-Light Solar-Cell Efficiency by Holistic Trap-Passivation using Micrometer-Thick Perovskite Film. Adv Mater 33, e2100770 (2021).DOI
  12. [12]Kim, G. et al. Transparent Thin-Film Silicon Solar Cells for Indoor Light Harvesting with Conversion Efficiencies of 36% without Photodegradation. ACS Appl Mater Interfaces 12, 27122-27130 (2020).DOI
  13. [13]Li, M. et al. Indoor Thin-Film Photovoltaics: Progress and Challenges. Adv. Energy Mater. 10, 2000641 (2020).DOI
  14. [14]Freitag, M. et al. Dye-sensitized solar cells for efficient power generation under ambient lighting. Nat. Photonics 11, 372-378 (2017).DOI
  15. [15]Opoku, H. et al. A tailored graft-type polymer as a dopant-free hole transport material in indoor perovskite photovoltaics. J. Mater. Chem. A 9, 15294-15300 (2021).
  16. [16]Tavakkolnia, I. et al. Organic photovoltaics for simultaneous energy harvesting and high-speed MIMO optical wireless communications. Light Sci. Appl. 10, 41 (2021).DOI
  17. [17]Cui, Y. et al. Accurate photovoltaic measurement of organic cells for indoor applications. Joule 5, 1016-1023 (2021).DOI
  18. [18]Zheng, J. et al. Enhanced hydrothermal heterogeneous deposition with surfactant additives for efficient Sb₂S₃ solar cells. Chem. Eng. J. 446, 136474 (2022).DOI
  19. [19]Choi, Y. C. et al. Highly Improved Sb₂S₃ Sensitized-Inorganic-Organic Heterojunction Solar Cells and Quantification of Traps by Deep-Level Transient Spectroscopy. Adv. Funct. Mater. 24, 3587-3592 (2014).DOI
  20. [20]Wang, S. et al. A Novel Multi-Sulfur Source Collaborative Chemical Bath Deposition Technology Enables 8%-Efficiency Sb₂S₃ Planar Solar Cells. Adv. Mater. 34, 2206242 (2022).DOI
  21. [21]Liu, X. et al. Grain Engineering of Sb₂S₃ Thin Films to Enable Efficient Planar Solar Cells with High Open-Circuit Voltage. Adv. Mater. 36, 2305841 (2024).DOI
  22. [22]Kondrōtas, R., Chen, C. & Tang, J. Sb₂S₃ Solar Cells. Joule 2, 857-878 (2018).DOI
  23. [23]Tang, R. et al. n-Type Doping of Sb₂S₃ Light-Harvesting Films Enabling High-Efficiency Planar Heterojunction Solar Cells. ACS Appl Mater Interfaces 10, 30314-30321 (2018).DOI
  24. [24]Han, J. et al. Multidentate anchoring through additive engineering for highly efficient Sb₂S₃ planar thin film solar cells. J. Mater. Sci. Technol. 89, 36-44 (2021).DOI
  25. [25]Qi, Y., Li, Y. & Lin, Q. Engineering the charge extraction and trap states of Sb₂S₃ solar cells. Appl. Phys. Lett. 120, 221102 (2022).DOI
  26. [26]Zhou, R. et al. Bulk Heterojunction Antimony Selenosulfide Thin-Film Solar Cells with Efficient Charge Extraction and Suppressed Recombination. Adv. Funct. Mater. 34, 2308021 (2024).DOI
  27. [27]Huang, Y. et al. Efficient In Situ Sulfuration Process in Hydrothermally Deposited Sb₂S₃ Absorber Layers. ACS Appl. Mater. Interfaces 14, 54822-54829 (2022).DOI
  28. [28]Huang, Y. et al. A Robust Hydrothermal Sulfuration Strategy toward Effective Defect Passivation Enabling 6.92% Efficiency Sb₂S₃ Solar Cells. Sol. RRL 7, 2201115 (2023).DOI
  29. [29]Sun, Y. et al. Novel non-hydrazine solution processing of earth-abundant Cu₂ZnSn(S,Se)₄ absorbers for thin-film solar cells. J. Mater. Chem. A 1, 6880-6887 (2013).DOI
  30. [30]Manders, J. R. et al. Solution-Processed Nickel Oxide Hole Transport Layers in High Efficiency Polymer Photovoltaic Cells. Adv. Funct. Mater. 23, 2993-3001 (2013).DOI
  31. [31]Jung, M. et al. Perovskite precursor solution chemistry: from fundamentals to photovoltaic applications. Chem Soc Rev 48, 2011-2038 (2019).DOI
  32. [32]Wang, X. et al. Manipulating the Electrical Properties of Sb₂(S,Se)₃ Film for High-Efficiency Solar Cell. Adv. Energy Mater. 10, 2002341 (2020).DOI
  33. [33]Rashid, H. U. et al. Solvent Degradation in CO₂ Capture Process from Power Plant Flue Gas. Theor. Exp. Chem. 49, 371-375 (2014).DOI
  34. [34]Shi, W., Song, S. & Zhang, H. Hydrothermal synthetic strategies of inorganic semiconducting nanostructures. Chem Soc Rev 42, 5714-5743 (2013).DOI
  35. [35]Hankare, P. P. et al. Synthesis and characterization of tin sulphide thin films grown by chemical bath deposition technique. J. Alloys Compd. 463, 581-584 (2008).DOI
  36. [36]Wang, C. et al. Interfacial defect healing of In₂S₃/Sb₂(S,Se)₃ heterojunction solar cells with a novel wide-bandgap InOCl passivator. J. Mater. Chem. A 11, 19914-19924 (2023).DOI
  37. [37]Zhou, J. et al. Colloidal SnO₂-Assisted CdS Electron Transport Layer Enables Efficient Electron Extraction for Planar Perovskite Solar Cells. Sol. RRL 5, 2100494 (2021).DOI
  38. [38]Han, J. et al. Alcohol Vapor Post-Annealing for Highly Efficient Sb₂S₃ Planar Heterojunction Solar Cells. Sol. RRL 3, 1900133 (2019).DOI
  39. [39]Shi, X. et al. Nanorod-textured Sb₂(S,Se)₃ bilayer with enhanced light harvesting and accelerated charge extraction for high-efficiency Sb₂(S,Se)₃ solar cells. Chem. Eng. J. 437, 135341 (2022).DOI
  40. [40]Hegedus, S. S. & Shafarman, W. N. Thin-film solar cells: device measurements and analysis. Prog. Photovoltaics Res. Appl. 12, 155-176 (2004).DOI
  41. [41]Li, J. et al. Manipulating the morphology of CdS/Sb₂S₃ heterojunction using a Mg-doped tin oxide buffer layer for highly efficient solar cells. J. Energy Chem. 66, 374-381 (2022).DOI
  42. [42]Zhou, J. et al. Unraveling the roles of mesoporous TiO₂ framework in CH₃NH₃PbI₃ perovskite solar cells. Sci. China Mater. 63, 1151-1162 (2020).DOI
  43. [43]Tang, R. et al. Hydrothermal deposition of antimony selেনosulfide thin films enables solar cells with 10% efficiency. Nat. Energy 5, 587-595 (2020).DOI
  44. [44]Lian, W. et al. Revealing composition and structure dependent deep-level defect in antimony trisulfide photovoltaics. Nat Commun 12, 3260 (2021).DOI
  45. [45]Shen, K. et al. CdTe solar cell performance under low-intensity light irradiance. Sol. Energy Mater. Sol. Cells 144, 472-480 (2016).DOI
  46. [46]Guillemoles, J. F. et al. Guide for the perplexed to the Shockley-Queisser model for solar cells. Nat. Photonics 13, 501-505 (2019).DOI

Paper details

vlnc:W_dgjg5mch
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 6

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents