Introduction

Metal halide perovskite (MHP) materials have occupied the frontlines of optoelectronic research for over a decade, since 2009. Although the initial pioneering research on perovskites was entirely for photovoltaic (PV) applications, there has recently been a shift to exploration of other applications such as thermoelectric,1,2 solar capacitors,1 supercapacitors,3,4 photoelectrochemical water splitting, photocatalysis,5,6 photodetectors,7,8,9,10 x-ray detectors,11,12 and many more.13,14,15 Considering PV application, MHPs have realized groundbreaking progress in terms of efficiency, with an over sixfold leap from efficiency of 3.8% (CH3NH3PbI3 perovskite) in 200916 to 25.6% in 2021 (CH(NH2)2PbI3 perovskite).17 However, MHPs face a notable disadvantage in terms of stable performance under ambient environmental conditions. One of the major causes for the degradation of perovskite solar cells (PSCs) is the presence of hygroscopic organic cations such as methyl ammonium (CH3NH3/MA) and formamidinium (CH(NH2)2/FA). Further, the presence of lead (Pb) in the composition makes them unsuitable for consideration as a green energy solution.18,19,20,21,22

Over the years, various plausible solutions have been suggested to realize stable, environmentally friendly PSCs with good efficiency.23,24 Research has been taken in multiple directions, with attempts to tailor the basic ABX3 perovskite structure at the A, B, and X sites to achieve the desired properties.25,26,27 In this regard, inorganic lead-free MHPs are being extensively pursued by researchers, as the materials possess stellar optoelectronic properties that can be tuned to achieve benchmark performance. The most popular alternatives for replacing Pb at the B site are germanium (Ge),10,28,29,30,31,32,33,34,35,36,37,38,39 tin (Sn),23,40,41,42,43,44,45 antimony (Sb),46,47,48,49 bismuth (Bi),50,51,52 indium (In),53,54,55 titanium (Ti),56,57,58,59 and tellurium (Te).60,61,62 Among these, Ge-based perovskites have proven to provide better stability and promising optoelectronic properties.10,28,29,35,37,63,64 However, one factor that limits further experimental exploration of germanium-based perovskites is the high cost of the necessary precursors, especially germanium iodide (GeI2). Also, the currently reported methods for the synthesis of CsGeI3 powders are solution-based,29 which is difficult to process for bulk and facile production.

In this work, we present a novel route for the synthesis of CsGeI3 (CGI) polycrystals using low-cost precursors from the available alternatives. Instead of using germanium iodide, we propose the use of pure germanium precursor separately along with pure iodine and cesium iodide precursors to obtain the necessary composition. This method of using germanium powders and iodine crystals separately is found to significantly reduce the precursor cost (with a rough estimate of the current market prices, 10 times lower cost) for the synthesis of germanium iodide-based perovskites. Also, the devised route is a solution-free technique, which is more facile, provides better crystallinity, and contributes to the enhanced stability of the synthesized polycrystals. Fortunately, compared to 3 days, there was very little degradation of the CGI polycrystals observed over 45 days. Further, the experimental results correlated closely with the simulation outcomes undertaken via density functional theory (DFT) studies. Additionally, theoretical studies were performed on the CGI device configuration using various electrode materials.

Experimental Procedure

Synthesis of CsGeI3 (CGI) Polycrystals

Inorganic lead-free MHP crystals have been successfully synthesized using the melt-quenching technique. In the present work, CGI polycrystals were synthesized using the solid-state reaction technique. Ge metal pieces (Alfa Aesar, 99.999%) and CsI powder (Alfa Aesar, 99.999%) were ground together for 2 h using an agate mortar and pestle. The finely ground sample was transferred to a clean quartz ampoule and loaded into the crystal puller. Resublimed iodine (I2, Merck, 99.8%) crystals were added to the mixture just before loading into the crystal puller because of the volatile nature of iodine. Initially, the temperature was increased to 150°C and maintained for 10 h, then increased to 925°C and maintained for 24 h to ensure mixing, further maintained at 700°C for 24 h for homogeneity, and eventually cooled to room temperature. The process flowchart is presented in Fig. 1. The melt-quenching technique described by Xiao et al.61 for the preparation of Cs2TeI6 crystals was modified according to the source materials and melting point of the compounds.

Fig. 1
figure 1

Process flowchart.

Full size image

Characterization of Synthesized CsGeI3 (CGI) Polycrystals

The structural analysis of the synthesized CGI polycrystals was carried out using x-ray diffraction (XRD) with Cu Kα [λ = 1.5405 Å] radiation at 40 kV and 15 mA in the 2θ range of 20°–60° (Rigaku MiniFlex 600). Raman spectroscopy was performed at an excitation wavelength of 532 nm using a LabRAM HR (UV). The prepared polycrystals were ground into powder, and the photoluminescence (PL) emission was studied using a Jasco FP-8300 fluorescence spectrometer using a 450 W xenon lamp. Also, the reflectance spectra were obtained for the powdered samples using a Lambda 950 UV–visible–near-infrared (NIR) spectrophotometer equipped with an integrating sphere (PerkinElmer, Waltham, MA, USA). Thermogravimetric analysis (TGA) was performed using a PerkinElmer TGA 4000 to observe the behaviour of the powders under temperature exposure.

Theoretical Studies

Quantum Espresso (QE) code was used for all the density functional theory (DFT)-based calculations.65 All calculations were performed with 3 × 3 × 2 supercells, for pure CGI structures, and a k-point grid of 7 × 7 × 3 was used for integrations over the Brillouin zone. We used various approximations including the local density approximation (LDA), generalized gradient approximation (GGA), and Heyd–Scuseria–Ernzerhof (HSE) for the calculations.66 Seo et al. showed that when using HSE to calculate the bandgaps of metal compounds which contain valence electrons in different orbitals, the mixing parameters must be optimized from typical values to see the variation in orbital overlapping.67,68 Thus, to obtain a more accurate bandgap for CsGeI3, the HSE calculation was used, with mixing parameters of 0.25 for all the calculations. Initially, we optimized the structures until the convergence of energy decreased to 5 × 10−5 eV and the value of force per atom reached 0.04 eV.69 SCAPS-1D software was used to determine quantum efficiency measurements.70

Results and Discussion

Figure 2a shows the XRD diffractogram of the as-synthesized CGI polycrystals. The XRD pattern of the CGI crystals synthesized using the melt-quenching technique showed an excellent match with existing CsGeI3 reports (JCPDS #01-085-1274).64 The XRD pattern of the powdered ingot was acquired immediately after breaking the quartz ampoule and after an interval of 3 days to evaluate the sample's stability. The XRD pattern shows peaks corresponding to Ge and CsI, which could indicate possible decomposition of CsGeI3 into its constituents CsI and Ge within just hours of exposure to an open environment. The XRD patterns obtained on day 3 further suggest the degradation of the sample, as shown in Fig. 2b. This agrees with the observations of Yue et al.64 However, it is notable that even after 45 days of storage under ambient conditions, there is no further degradation observable for the CGI polycrystals. Also, such behaviour suggests the suitability of the synthesis route for employing solution-free solid-state methods to synthesize inorganic MHPs.

Fig. 2
figure 2

(a) X-ray diffractogram of the as-synthesized CGI polycrystals. (b) Comparison of XRD patterns of CGI polycrystals: 3rd day and 45th day.

Full size image

The Raman spectra shown in Fig. 3 agree with the reported data.29 The strongest Raman peaks of the CGI crystal at 165 cm−1 can be attributed to the A1 (longitudinal mode).71 A minor peak detected at 337 cm−1 aligns with the overtone of CsI,72 thereby confirming the existence of CsI, as indicated by XRD analysis. The low phonon frequencies observed in CGI suggest that its transparency range could be extended into the far-IR region.73

Fig. 3
figure 3

Raman spectra of CGI polycrystals.

Full size image

Further, diffuse reflectance spectroscopy (DRS) was performed, and the bandgap was found to be 1.6 eV, very close to previously reported values. Also, there was no change observed in the bandgap measured after 45 days. Kubelka–Munk theory was applied to estimate the bandgap from the reflectance spectra, with n = 1/2. The bandgap estimation from the plot is presented in Fig. 4. The estimated bandgap suggests the suitability of the synthesized CGI polycrystals for use in optoelectronic applications after being made into pellet or thin-film format, as has been previously reported.10

Fig. 4
figure 4

Diffuse reflectance spectrum of the as-synthesized CGI polycrystals.

Full size image

To match the experimentally obtained energy bandgap of CsGeI3, theoretical studies were carried out. Accordingly, DFT calculations using QE code with different approximations were done to correlate with the DRS results. From Table I it is clear that out of all approximations, HSE calculation gives more accurate values of the electronic bandgap of CGI in comparison with the experimental outcome. The density of states is a crucial physical parameter for understanding the electronic properties of materials. Figure 5 shows the total density of states (TDOS) and partial density of states (PDOS) for CsGeI3 simulated using the HSE approximation. It is clear that the highest energy level of the valence band in CsGeI3 is primarily influenced by the halogen element iodine, while for the lowest occupied levels of the conduction band, Ge contributed the most.

Table I Estimated electronic bandgaps of CGI using different approximations
Full size table
Fig. 5
figure 5

Simulated total and partial density of states of CsGeI3.

Full size image

The experimentally obtained CsGeI3 polycrystals with the bandgap ranging between 1.5 and 1.6 eV (both theoretically and experimentally) have great potential for use in photovoltaics and photodetectors. In this direction, the device suitability of the CsGeI3 polycrystals was checked theoretically with different type of electrodes. Systematic simulations were undertaken to investigate how different electrodes can have a notable impact on the performance of CsGeI3 in photovoltaics and photodetection processes. Figure 6 depicts the simulation results of quantum efficiency studies. Figure 6a–c shows the circuit diagrams of the CGI device with different electrodes (Ag, Au, Pt). The estimated current density and incident photon-to-electron conversion efficiency (IPCE%) of the proposed devices clearly indicate the significant influence of different electrode materials on CsGeI3 performance. It can be seen from Fig. 6d and e that the current density of CGI starts increasing with lower bias voltage in the case of the Au electrode, while the maximum IPCE% was observed for the Ag electrode. Moreover, the highest IPCE% was found to be independent of the electrodes and was seen at a wavelength of 760 nm.

Fig. 6
figure 6

(a–c) Schematic of the device structure of CsGeI3 with (a) Ag, (b) Au, and (c) Pt electrodes. (d) Plot of current density vs bias voltage and (e) IPCE% vs wavelength with different electrodes.

Full size image

The photoluminescence studies conducted with an excitation wavelength of 660 nm reveal an intense broad peak centred around 798 nm and a peak of low intensity around 824 nm, as seen in Fig. 7. It can be observed that the emission at 798 nm corresponds to Eg = 1.55 eV, which is very close to the theoretical prediction (1.58 eV) as well as the experimentally observed (1.60 eV) bandgap. This high intense peak is attributed to electron transitions of the GeI3 matrix and matches with reported data in the literature.10 The peak around 824 nm is due to the perovskite matrix and observed for other inorganic MHPs as well.74 Inter-band transitions in the perovskite matrix give rise to these emissions.75

Fig. 7
figure 7

Photoluminescence (PL) spectrum of the as-synthesized CGI polycrystals.

Full size image

The results of the thermogravimetric analysis of the as-prepared CGI powders are presented in Fig. 8. Transition temperatures of 415°C, 450°C, 500°C, 720°C, and 800°C were found for the respective compounds. TGA was employed to investigate the decomposition of these perovskites under controlled, high-temperature conditions in air atmosphere from room temperature to 1000°C. The decomposition pathway under air atmosphere was accounted for by the apparent sharp weight loss observed. It was noted that the compound requires multiple steps for decomposition. The first stage is associated with the production of iodide molecules, escaping around 200–600°C. A subsequent steep weight loss in the second step is ascribable to the sublimation of GeI2. At around 600°C, weight loss of about 50% of the initial weight is observed. The compound has a melting point of 790°C, which can also be seen from the TGA curves, which indicate complete disintegration around the same temperature.

Fig. 8
figure 8

Thermogravimetric analysis (TGA) curve of the as-synthesized CGI polycrystals.

Full size image

Conclusions

In summary, CsGeI3 (CGI) polycrystals were synthesized using the solution-free melt-growth technique, and their stability up to 45 days was investigated. X-ray diffraction and Raman studies confirmed the CGI structure with the appearance of some impurity phases because of the decomposition. After 3 days, there was slight degradation of the polycrystals. However, up until the 45th day, they remained largely steady in comparison to the third day. Further, optical studies suggested that the bandgap of the synthesized powder was within the region of 1.6 eV. Also, a better correlation to the experimental results was achieved through DFT calculations with suitable approximation. The highest incident photon-to-electron conversion efficiency was observed for the CGI device with Ag electrodes. These CsGeI3 powders can be further used for producing pellets or thin films. Moreover, the devised method can be applied to produce inorganic MHPs with enhanced stability. The experimental and theoretical outcomes suggest the suitability of the CGI for various optoelectronic applications. However, the observed decomposition of the polycrystals hinders its practical applicability. Strategies such as doping or halide cation composition optimization can be used to further increase the stability of these polycrystals by addressing structural defects, controlling ion migration, and optimizing the electronic properties of the material.