Introduction

Several strategies have been explored to extend the optical absorption of TiO2 into the visible light region including doping of metal or nonmetal ions, dye sensitization, surface modification, and coupling with narrow bandgap semiconductors [1,2,3,4,5,6,7,8]. Among these, a remarkable enhancement in the photoelectrochemical (PEC) performance of TiO2 under light has been obtained by coupling low bandgap semiconductors such as CdS [4], CdSe [5], Ag2S [6], Sb2S3 [7], and Bi2S3 [8]. However, to date, all such heterojunction-based solar cells do not meet the theoretical value for light-harvesting efficiency reported in the literature [9]. This is because of the one of the limits that a large amount of the incident solar radiation is not absorbed by the solar cells. This deterrent has resulted in search for the different light-harvesting sensitizers to enhance the light absorption capacity in the visible and near infra-red (IR) region of the solar spectrum.

Mercury sulfide (HgS), a group II–VI compound with cubic phase [9] having bulk bandgap about 0.5 eV [10], which is tunable in the range of about 1.9–2.6 eV [9], suggests it an alternative candidate to above discussed sensitizers. The literature also reveals that the conduction band position of HgS and TiO2 w.r.t. vacuum is favorable [11], which proposes TiO2/HgS heterojunction configuration, a potential candidate for photovoltaic applications. However, the literature survey indicates that the photovoltaic studies of HgS sensitized TiO2 photoelectrode are not reported earlier. And so, it is important to explore HgS as a sensitizer for TiO2 photoelectrode toward photovoltaic application. Thus, in the present work, we demonstrate the temporal chemical solution deposition of HgS nanocrystals over spin-coated titanium dioxide for photovoltaic application.

Experimental details

Materials

TiO2 (P-25, Degussa, Germany), acetyl acetone, p-hydroxybenzoic acid, ethanol, mercury chloride, and sodium thiosulfate were analytical reagent (AR) grade and used without further purification. Fluorine-doped tin oxide (FTO) is used as substrate for deposition of TiO2.

Synthesis of TiO2 film by spin-coating technique

In the present work, TiO2 photoelectrodes were prepared onto FTO substrates using spin-coating technique at room temperature. Prior to deposition, FTO substrates were washed with detergent and then cleaned with double-distilled water (DDW). Again, the substrates were rinsed ultrasonically with DDW before deposition of thin films. Initially, TiO2 nanoparticles (P-25, Degussa, Germany), precursor for TiO2, ethanol, and acetyl acetone, solvents, and catalyst were used for the preparation of gel of TiO2 using the sonication method. First, 2 g of TiO2 powder (Degussa P-25), 8.5 ml of ethanol, and 1.5 ml acetyl acetone are added together in a beaker and stirred together for 30 min. Second, 1 g p-hydroxybenzoic acid and catalyst were added to the above mixture and allowed for continuous stirring for next 8 h at room temperature until a uniform mixture is formed. Third, the obtained homogeneous mixture was sonicated for 30 min after transferring them into a clean 10-ml beaker with the frequency of 20 kHz using ultrasonication bath.

For the preparation of TiO2 samples, 1 ml of TiO2 gel was introduced over the cleaned substrate and then rotated for 2 min. Furthermore, the coated samples were annealed at 400 °C for 45 min.

Deposition of HgS over spin-coated TiO2

The deposition process of HgS involves mercury chloride (HgCl2) and sodium thiosulphate (Na2S2O3) as a precursor of mercury ions (Hg2+) and sulfide ions (S2−), respectively. First, for the deposition of HgS films, aqueous solution of HgCl2 (0.04 M) and Na2S2O3 (0.08 M) in 10 ml of DDW was prepared in two different beakers. Initially, the beaker containing HgCl2, source of Hg2+, is kept on constant stirring at room temperature, and Na2S2O3, source of S2−, is added slowly in the first bath. The formed mixture is stirred well to dissolve the formed precipitate, which leads to the homogeneous solution. Now, the deposition process was considered to be based on slow release of Hg2+ and S2− ions in the solution which then condense as ion by ion or cluster by cluster mechanism on the surface of substrate, as a result of heterogeneous nucleation reported earlier [8].

Now, spin-coated FTO/TiO2 samples were introduced vertically into the bath of HgS, for the purpose of the chemical solution deposition of HgS at 70 °C for three different deposition times, namely 30, 45, and 60 min. The samples were nomenclated as A, B, and C, respectively. The porous film of TiO2 was observed to be coated with the enhanced brownish shade of the photoanodes with the increase in the deposition time. This might be due to the enhanced level of deposition of nanocrystalline HgS onto the bare photoanodes of TiO2 with increased sensitization time.

Thereafter, to enhance the crystallinity of HgS crystals, samples A–C were annealed for 30 min at 200 °C.

Cell assembly

For the measurement of photoelectrochemical cell (PEC) performance, we have used the aqueous solution of polysulfide electrolyte which was prepared using the 1 M of sodium sulfide (Na2S), 1 M of sulfur powder, and 1 M of sodium hydroxide (NaOH). For cell assembly, the carbon suit of FTO was used as a counter electrode. All three samples A–C and counter electrode were clamped together into a sandwich-type configuration with droplet of polysulfide electrolyte injected in between them, for the fabrication of solar cells.

Characterization of HgS, TiO2, and TiO2/HgS thin films

The structural analysis was carried out using X-ray diffractometry by means of an automated Bruker D8 advance X-ray diffractometer. Optical absorption study was carried out using a JASCO UV–Vis spectrophotometer (V-630) at room temperature. A scanning electron microscope (SEM) with an energy-dispersive spectrometer (EDS) attachment [S-4800 Type-II (Hitachi High Technology Corporation Tokyo, Japan)] was used to study morphological and elemental analysis of the samples. Contact angle measurement (CAM) has been measure with the help of lab made facility using DSC-HX7V cyber shot camera with 16.2 Mega pixels sensor and 10x optical zoom. The cell performance was measured by a semiconductor characterization unit (Keithley 2420 source meter) under illumination of 85 mW/cm2. Electrochemical impedance spectroscopy (EIS) was studied by potentiostat/galvanostat (IVIUM Vertex model).

Results and discussion

Structural analysis

The XRD patterns of the HgS and TiO2 are shown in Fig. 1a, b, which reveals the hexagonal phase for the deposited HgS crystals (JCPDS data file: 421408) and anatase phase with tetragonal structure for the spin-coated TiO2 films (JCPDS data file: 84-1285).

Fig. 1
figure 1

2θ versus intensity plot of a HgS, b TiO2, and c HgS sensitized TiOphotoanodes

However, the XRD pattern shown in Fig. 1c comprises combinations of two sets of patterns: one of them is assigned to TiO2 (JCPDS data file: 84-1285) and other originates from the HgS (JCPDS data file: 421408). This confirms the formation of heterojunction between TiO2 and HgS crystals deposited using chemical solution technique.

Morphological analysis and contact angle measurement of HgS, TiO2, and TiO2/HgS films

Figure 2a–c shows the surface morphology analysis of HgS, TiO2, and TiO2/HgS films, respectively. Figure 2a shows that the substrate is uniformly covered with HgS, having spherical morphology. The flake-like network of TiO2 crystals is observed from Fig. 2b. However, from Fig. 2c, it is observed that in case of TiO2/HgS sample, surface is covered with enhanced granular structure. This depicts the coating of HgS nanoparticles over the entire surface of porous TiO2 films.

Fig. 2
figure 2

Scanning electron micrographs and contact angle measurement of a HgS, b TiO2, and c TiO2/HgS photoanodes

Besides as shown in Fig. 2a–c, the wettability of the samples studied using CAM for TiO2/HgS film is 31°, which is smaller than the bare TiO2 film (52°). The improved wettability of TiO2/HgS after sensitization is helpful for the diffusion of electrolyte and charge transport at the interfaces between photoanode and electrolyte [12].

Elemental analysis

Figure 3a–c shows a comparative elemental analysis of HgS, TiO2, and TiO2/HgS photoelectrodes, respectively. The presence of emission lines of ‘Ti’, ‘Hg’, ‘S’, and ‘O’ in the energy-dispersive X-ray spectroscopy (EDXS) spectra for TiO2/HgS film gives qualitative confirmation of the chemical bath deposition (CBD) of the HgS over the TiO2 film, which matches with TiO2/HgS heterojunction formation analysis discussed above in XRD analysis.

Fig. 3
figure 3

EDXS spectrum of HgS, TiO2, and HgS sensitized TiO2 photoelectrodes

Optical properties

Figure 4 displays the UV–visible spectra curves of the TiO2 films loaded with HgS nanocrystals for different deposition times. It is clear that both absorbance and absorption range of the TiO2 films are proportional to the number density of HgS nanocrystals. The observed shift of the absorbance edge toward the red edge suggests that the TiO2 film after HgS deposition may cover the entire visible and near IR region of the solar spectrum. Such enhanced absorption in the visible region, which causes the blue shift in the energy bandgap for TiO2/HgS films (Fig. 4b), indicates the lowering of the conduction band of the TiO2 [13]. This may provide a driving force in the charge separation process at the interface of HgS and TiO2.

Fig. 4
figure 4

a Wavelength-dependent absorption spectra of bare TiO2 and TiO2/HgS films. b Optical energy bandgap of bare TiO2, HgS, and TiO2/HgS film

Electrochemical impedance spectroscopy

Figure 5a shows electrochemical impendence spectra (EIS) curves for the cells A–C under the dark condition. It comprises two semicircles: first gives the charge transfer resistance at the counter electrode/polysulfide electrolyte interface, and the second circle represents charge transfer and transport at the TiO2/HgS/polysulfide interface.

Fig. 5
figure 5

a Nyquist plot for cells A, B, and C. b Bode plot for cells A, B, and C

In the present work, EIS result shows the enlarged radius of the second semicircle in the Nyquist plot for cell A. This is probably due to the very less number and size of HgS nanocrystals, in case of sample A. However, for increased deposition time up to 45 min, there is a decrease in the radius of the semicircle in the Nyquist plot for cell B. This may be because of the enhancement in the size and number of HgS nanocrystals over TiO2, which is in agreement with optical absorption study discussed above. Further, as shown in Fig. 5a, in case of cell C, there is a sudden decrease in the radius of the semicircle in the Nyquist plot. Such a result may be attributed to the aggregation of HgS, leading to clogging of the TiO2 pores. The apparent red shift in the optical absorption spectra for sample C supports such an aggregation.

Figure 5b shows the Bode plots of the solar cells A–C. The curve peak of the spectrum can be used to determine the electron lifetime according to the equation τ = 1/(2пf) [14], where τ = electron lifetime and f = maximum frequencies from Bode plot, and is presented in Table 1.

Table 1 Electron lifetime for the different electrode systems

Thus, according to the results of the optical absorption, electron transfer, and life time, the optimal coating of HgS may have sped up the charge transport and results in the enhanced photocurrent for cell B. This suggests 45 min as appropriate sensitization time for HgS nanocrystals over TiO2 in the present study.

PEC performance

Figure 6 shows current density–voltage (J–V) characteristics of HgS sensitized TiO2 solar cells. The device B shows a better performance than the devices A and C, respectively. The obtained photovoltaic parameters are tabulated in Table 2.

Fig. 6
figure 6

Photocurrent density (Jsc) versus photovoltage (Voc) characteristic curves for cells A, B and C

Table 2 Solar cell parameters for samples A, B, and C

As discussed earlier, with increase of sensitization time, the optical absorption spectra as shown in Fig. 4a clearly show an enhanced loading of the sensitizer into the photoelectrodes, which is evident from the improved optical absorption in the visible region. However, the enlarged arc of Nyquist plot for cell A suggests the deposition of less number of HgS nanocrystals for the deposition time of 30 min. Thus, instead of maximum life time, cell A gives rise to less current density Jsc and open-circuit voltage Voc under illumination, as tabulated in Table 2.

Nevertheless, in case of sample B, the stretched optical absorbance in the far visible region compared to sample A reveals increased size and number of HgS nanocrystals for the deposition time of 45 min. Such an enhanced absorption of light in the visible region leads to the generation of increased number density of excitons at the electrode–electrolyte interface. This gives rise to the enhanced photocurrent J and hence Jsc in case of cell B, even if the lifetime is observed to be decreased from 12 to 6.2 ms.

Conversely, the observed declined performance for cell C may be attributed to the aggregation of HgS nanocrystals, resulting in the decrease of the access of electrolyte into the pores of photoanode. This may have increased the recombination probability in the film, which may be confirmed from the minimized life time in case of cell C.

Conclusion

We have demonstrated the successful chemical bath deposition of hexagonal HgS crystals over the anatase TiO2 photoanode for solar cell application, which is first of its kind. The process of the HgS deposition on the surface of the TiO2 film was heterogeneous nucleation, which was visibly affected by the deposition time. The appropriate deposition time was determined as 45 min for the TiO2/HgS heterojunction using electron transfer and lifetime measurements through EIS. For the deposition time beyond 45 min, i.e., 60 min, the photovoltaic properties of TiO2/HgS-based cell started dropping due to the increased surface-charge recombination.