1 Introduction

Among all types of solar cells, the quantum dot sensitized solar cells (QDSSCs) as a derivative from dye-sensitized solar cells with low cost and simple fabrication processes have been widely studied [1–3]. Although, the quantum dots sensitizers exhibit fantastic advantages in comparison with conventional dyes such as quantum confinement effect, higher extinction coefficient and multiple exciton excitation, the conversion efficiency of QDSSCs is still much lower than that of 13 % of DSSCs [4–6]. The back electron transfer or strong recombination processes of photogenerated carriers happened at the interface are known to be one of the main factors for such lower efficiency of QDSSCs [7]. For ZnO photoanodes, except for the above disadvantages, its surface is quite easy to be etched by the electrolyte. Therefore, it is believed that surface modification on the ZnO photoanodes is an effective way to overcome the challenge and finally enhance the photovoltaic performance of QDSSCs. So far, people usually use ZnS/ZnO heterostructure as photoanodes to enhance the photovoltaic performance of QDSSCs [8]. In this way, the ZnS compact layer in the coaxial structure not only retards the back transfer of electrons to the QDs and electrolyte, but also the formation of a ZnS layer on the ZnO nanorods reduces the photogenerated electron–hole recombination at the photoanode/electrolyte interface due to the reduced surface defects [8]. However, for CdS QDs sensitized ZnS/ZnO heterostructure solar cells, the higher energy level of ZnS conducting band than will form a potential barrier between ZnO nanorods and CdS QDs, which strongly suppress the charge transfer from the CdS QDs to ZnO photoanode [8]. If one can change the band alignment between CdS/ZnS/ZnO, the photovoltaic performance of ZnO QDSSCs will be further improved.

In the field of organic photovoltaic cells, the self-assembled monolayers (SAMs) have been widely used to modify the transparent electrodes or photoanodes for enhancing the photovoltaic performance [9–11]. SAMs is a two-dimensional molecular array that is spontaneously organized by adsorption of amphiphilic organic molecules on a solid inorganic surface [11]. It can be deposited on the surface of polymer [12], metal [13] and metal oxide [14, 15] in a simple and effective way with stable structure, high order and less defects. In this way, we can successfully adjust the properties of materials such as wettability, work function, and charge transfer. Therefore, we can prospect that introducing SAMs into the interface of ZnO/ZnS heterostructures can make the band alignment better, which will induce the enhancement of photovoltaic performance of QDSSCs. However, so far, only a few works have been reported about the utilization of SAMs in QDSSCs, and most of them were focused on the TiO2 photoanode [8, 16–18]. For instance, Bent et al. introduced different SAMs between CdS QDs and TiO2 photoanode and found that the nature of the SAMs tailgroups do not significantly affect the uptake of CdS quantum dots on TiO2 nor their optical properties, and the presence of the SAMs does enhance the power conversion efficiencies almost three times [18]. But they did not reveal the influence mechanism in detail.

In this work, we proposed to introduce 3-PPA SAMs into the interface of ZnO/ZnS heterostructures. By varying the deposition time of 3-PPA, we investigated their effects on the photovoltaic performances of ZnO/ZnS heterostructure quantum dot sensitized solar cells and revealed the influence mechanism in detail.

2 Experiments

2.1 Synthesis of ZnO nanorods

All chemicals (analytical grade reagents) were directly used without further purification. The indium tin oxide (ITO) substrates were sequentially cleaned with acetone, alcohol and deionized water for 15 min under ultrasonic treatment. The ZnO nanorods with good alignment were grown on ITO substrates by two-step chemical bath deposition (CBD) method, including a substrate treatment prior to the CBD growth. The detailed process can be found in our previous report [7, 8, 19–21]. First of all, ITO substrates were pretreated by spin coating the substrate with a 10 mM zinc acetate dehydrate (Zn(C2H3O2)2·2H2O) ethanol solution. For the CBD growth of ZnO nanorods, the aqueous solutions of 0.1 M zincnitrate hexahydrate (Zn(NO3)2·6H2O) and 0.1 M methenamine (C6H12N4) were first prepared and mixed together. In this case, the pretreated ITO substrates were put into the mixture solution for 5 h under 95 °C to grow ZnO nanorods.

2.2 Depositing SAMs of 3-PPA on ZnO nanorods

To realize the effective deposition of SAMs, the samples of ZnO nanorods arrays were dipped into 10 mM 3-PPA ethanol solutions for 1, 2, 3 min, respectively. The samples were named as 3-PPA(1 min)/ZnO, 3-PPA(2 min)/ZnO and 3-PPA(3 min)/ZnO. After ending the immersion step, the samples were rinsed with ethanol for 30 s to remove excess SAMs molecule [7].

2.3 Assemble of ZnS/ZnO and ZnS/3-PPA/ZnO heterostructures

The ZnS layer was in situ deposited on ZnO and 3-PPA/ZnO by successive ionic layer deposition and reaction (SILAR) [16]. The prepared pure ZnO, 3-PPA(1 min)/ZnO, 3-PPA(2 min)/ZnO and 3-PPA(3 min)/ZnO samples were immersed in a 0.1 M Zn(NO3)2 solution for 5 min. They were then rinsed with distilled water for 30 s to remove excess ions weakly bound to the surface of samples and immersed in a 0.1 M Na2S solution for another 5 min followed by another rinsing with distilled water. The above process can be defined as one cycle. In our case, we performed eight cycles to realize the formation of ZnS/ZnO and ZnS/3-PPA/ZnO heterostructures. Subsequently, the samples were thoroughly washed with ethanol and deionized water and then dried at room temperature. The obtained heterostructures were named as ZnS/ZnO, ZnS/3-PPA(1 min)/ZnO, ZnS/3-PPA(2 min)/ZnO and ZnS/3-PPA(3 min)/ZnO.

2.4 Deposition of CdS QDs

The CdS QDs were also in situ deposited on the surfaces of ZnS/ZnO and ZnS/3-PPA/ZnO heterostructures by a SILAR technique, which is quite similar to the deposition of ZnS layer described above. We use Cd(NO3)2 (0.1 M) solution as the Cd source. We also use Na2S (0.1 M) solution as S source. We performed the SILAR cycles for 12 times.

2.5 Cell fabrication

The photoelectrode of CdS/ZnS/3-PPA/ZnO was incorporated into thin layer sandwich-type cells. A 20 nm platinum-sputtered ITO substrate as the counter electrode and the working electrode were positioned face-to-face. The iodide-based electrolyte, consisting of 1 M LiI and 0.05 M I2 in alcohol, was injected into the interelectrode space by capillary action.

2.6 Characterization and measurements

The X-ray diffraction (XRD) patterns were recorded by a MAC Science MXP-18 X-ray diffractometer using a Cu target radiation source. Scanning electron microscopy (SEM) pictures were collected on a Hitachi, S-570 SEM. Transmission electron micrographs (TEM) and high-resolution transmission electron microscopy (HRTEM) images were taken on JEM-2100 transmission electron microscope. X-ray photoelectron spectra (XPS) with a resolution of 1.0 eV were recorded on a VG ESCALAB Mark II XPS using Mg Kα radiation (hν = 1253.6 eV). The photoluminescence (PL) measurements were performed on the Renishaw invia spectroscopy excited by a continuous He-Cd laser with a wavelength of 325 nm at a power of 2 mW. The Ultraviolet–Visible (UV–Vis) absorption spectra of each photoelectrode were recorded on a UV–Vis spectrophotometer (UV-5800PC, Shanghai Metash Instruments Co., Ltd) at room temperature. The performance of ultraviolet photoemission spectroscopy (UPS) was carried out with a helium discharge lamp (hν = 21.22 eV) in normal emission with a sample bias of −8 V. The photocurrent dependence on the voltage (I − V) were measured under AM 1.5 G simulated sunlight illumination (100 mW/cm2, Model 91160, Oriel).

3 Results and discussion

Figure 1 displays the XRD patterns of ZnO nanorods, ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO samples, respectively. From the XRD patterns, we can only see the (002), (103), and (004) characteristic peaks of wurtzite ZnO according to the JCPDS card No. 05-0664. The absence of diffraction peaks from ZnS maybe due to their rather thin thickness. However, the intensities of (002) diffraction peak decrease step by step after ZnS deposition and further 3-PPA modification, which provides an evidence that ZnS and 3-PPA have been successfully deposited on the ZnO nanorods.

Fig. 1
figure 1

XRD pattern of a sample ZnO, b 3-PPA(1 min)/ZnO and c ZnS/3-PPA(1 min)/ZnO

Full size image

Figure 2a shows the SEM (tilted view) image of as grown ZnO nanorods. We can see that the large-scale, high-density and vertically aligned ZnO nanorods are uniformly grown on the entire surface of ITO substrate. The average diameter is ~130 nm. Figure 2b displays the SEM image of 3-PPA(1 min)/ZnO sample, from which we can see that no obvious change happens on the surface of ZnO nanorods due to the 3-PPA deposition. Figure 2c shows the SEM image of ZnS/3-PPA(1 min)/ZnO sample. The surfaces of ZnO nanorods are no longer smooth and lots of small particles attach compactly on their surfaces, indicating the formation of ZnS/ZnO heterostructure. Figure 2d displays the SEM image of CdS/ZnS/3-PPA(1 min)/ZnO sample. The surfaces of each nanorod turn much rougher and some islands made of small particles appear on the top surface of nanorod as shown in the inset.

Fig. 2
figure 2

SEM images: a ZnO nanorods; b 3-PPA(1 min)/ZnO; c ZnS/3-PPA(1 min)/ZnO; d CdS/ZnS/3-PPA(1 min)/ZnO

Full size image

To further reveal the microscopic structure and morphology variation of all samples, we performed TEM measurement on ZnO, ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO samples. Figure 3a, b present the TEM and HRTEM images of single ZnO nanorod, which clearly shows that the average diameter of ZnO nanorod is ~130 nm. The well-resolved lattice fringe spacing can be distinguished to be 0.26 nm, corresponding to the typical wurtzite structure of ZnO. Figure 3c, d show the TEM and HRTEM images of single 3-PPA(1 min)/ZnO nanorod. From Fig. 3c, we can see that the surface of nanorod turns rough after depositing 3-PPA for 1 min. From Fig. 3d, we can observe a thin amorphous layer with 4.0 nm thickness on the surface of ZnO nanorods. Here we would like to mention that, to prepare the sample for TEM measurement, we utilized the blade to scrape the ZnO and 3-PPA(1 min)/ZnO nanorods off the ITO surface. They were collected on the copper screen, then we dropped a few drops of ethanol on the copper screen to remove the contaminations and dried them at room temperature. Since the amorphous layer did not appear on the surface of pure ZnO nanorods as shown in Fig. 3b, we can deduce that this thin amorphous layer on the surface of 3-PPA(1 min)/ZnO sample is originated from SAMs. Figure 3e, f show the TEM and HRTEM images of single ZnS/3-PPA(1 min)/ZnO nanorod. Obviously, a layer of nanoparticles cover the entire nanorod as shown in Fig. 3e. In Fig. 3f, we can distinguish lots of nanoparticles with a clear lattice spacing of 0.33 nm, which matches with the distance of the (111) plane of the cubic ZnS. Some of ZnS nanoparticles have been marked by red circles with average diameter of 3.6 nm. These results verified the successful deposition of 3-PPA on the surface of ZnO nanorods and the formation of ZnS/ZnO heterostructure.

Fig. 3
figure 3

TEM images of ZnO (a), 3-PPA/ZnO (c), ZnS/3-PPA(1 min)/ZnO (e), and HRTEM images of ZnO (b), 3-PPA/ZnO (d), ZnS/3-PPA(1 min)/ZnO (f)

Full size image

To further reveal the variation of surface chemical composition of 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO, we used XPS technique to characterize the samples. All XPS spectra have been presented in Fig. 4, where the binding energies have been calibrated by taking the carbon C1 s peak (285.0 eV) as reference. Figure 4a shows the XPS survey spectra obtained from 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO, in which all of the peaks can only be ascribed to Zn, O, C and P elements. The high resolution scans of O1 s XPS spectra from 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO have been displayed in Fig. 4b, c, respectively. We used the XPS peak curve-fitting procedure to fit the experimental spectra and all of them can be deconvoluted into several components. As show in Fig. 4a, the deconvolutions show the presence of four different O1 s peaks in the 3-PPA(1 min)/ZnO. The peak located at 529.9 eV can be assigned to the oxygen from zinc oxide(Zn–O) [22]. The peaks located at 531.2 and 532.2 eV can be ascribed to the oxygen in the phosphate group of P=O and P–O–H, respectively [23]. The peaks located at 533.1 eV is originated from the oxygen in the carboxyl group (–COOH) [23]. In comparison with Fig. 4b, three O1 s peaks are enough to well fit the experimental spectra of ZnS/3-PPA(1 min)/ZnO. Except for the oxygen in ZnO and –COOH, we can observe another peak located at 531.6 eV, which is ascribed to –P(O)(OH)2 [24]. Such kind of difference should be related to the further deposition of ZnS on the sample. Figure 4d shows the C1 s XPS spectra of 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO. The deconvolutions for C1 s XPS spectrum of 3-PPA(1 min)/ZnO indicate the presence of three different C1 s species in the sample, i.e. the carbon of the aliphatic chain (C–C/C–H) located at 284.51 eV [25, 26], the covalently bound carbon that binds to P of the phosphate group (–C–P(O)(OH)2) located at 285.77 eV [23], and the carbon of the carboxyl group (O–C=O) located at 288.43 eV [11]. After further depositing ZnS layer, the carbon peak from –C–P(O)(OH)2 disappears. Figure 4e shows the XPS spectrum for P2p core level lines of 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO. The peaks at 132.86 and 139.18 eV are attributed to the P2p1/2 and P2p3/2 peak, respectively [25, 27]. After depositing ZnS, the peak of P2p1/2 disappears and the binding energy shifts to the higher energy since the lone pair electrons of phosphorous ligand become different. Figure 4f shows the Zn2p XPS spectra of 3-PPA(1 min)/ZnO and ZnS/3-PPA(1 min)/ZnO. The peaks located at 1021.46 and 1044.57 eV are corresponding to the binding energy of Zn2p3/2 and Zn2p1/2, respectively. No obvious change happens after depositing ZnS layer. Figure 4g shows the XPS S2p spectrum of ZnS/3-PPA(1 min)/ZnO. Based on the above discussion and according to the molecule formula of 3-PPA, we can deduce that the above element species originate from 3-PPA/ZnO and, indicating the surfaces of ZnO nanorods have been successfully modified by 3-PPA for 3-PPA/ZnO nanorods. Meanwhile, both the presence of S signal and the obvious change of O1 s, C1 s and P2p signals prove that ZnS layer has been successfully deposited on the surface of 3-PPA(1 min)/ZnO.

Fig. 4
figure 4

The XPS spectra of 3-PPA(30 s)/ZnO nanorods and ZnS/3-PPA(1 min)/ZnO nanorods. a survey spectrum; b O1 s XPS spectrum of 3-PPA(30 s)/ZnO; c O1 s XPS spectrum of ZnS/3-PPA(1 min)/ZnO; d C1 s XPS spectrum; e P2p XPS spectra; f Zn2p XPS spectra; g S2p XPS spectrum

Full size image

To further prove the origination of thin amorphous layer in 3-PPA/ZnO sample, we performed the FT-IR measurements on ZnO nanorods and 3-PPA(1 min)/ZnO sample. Since these two samples are grown on ITO substrates, they were directly tested by FT-IR under reflectance configuration mode. Meanwhile, the 3-PPA powders were characterized by transmission FT-IR technique to make a reference. Figure 5 demonstrates the FT-IR spectra of ZnO NRs, 3-PPA powders and 3-PPA(1 min)/ZnO. In the FT-IR spectrum of ZnO NRs shown in Fig. 5a, the peak at 3060 ~ 3826 cm−1 can be ascribed to the hydroxyl stretching vibration ν(O–H) [27]. The peak at 2850 ~ 2966 cm−1 is known as the C–H stretching vibration ν(C–H) [27]. In the FT-IR spectrum of pure 3-PPA powder shown in Fig. 5b, the hydroxyl stretching vibration ν(C–H) located at 2969 cm−1 is originated from 3-PPA according to its molecule formula. In the carbonyl stretching region, the strong peak located at 1716 cm−1 is related to the C=O stretch of carboxylic group [27] and the peak occurring at 1267 cm−1 is ascribed to the C–O stretching in carboxylic groups [25]. The two characteristic peaks in the region of 900 ~ 1050 cm−1 are originated from P–O–H group [27]. The band at 1207 cm−1 is related to P=O stretching [25]. The position of ν(O–H), C=O, C–O and P–O–H characteristic peaks in the FT-IR spectrum of 3-PPA(1 min)/ZnO sample can be well matched with that of pure 3-PPA powder as shown in Fig. 5c. But the intensities of P–O–H and P=O stretching characteristic peaks turn much lower. Meanwhile, a new strong peak around 1060 cm−1 appears. It is typically assigned to the stretches of PO3 2− group anchored to the surface through a multidentate bonding, which involves both P–O and P=O terminations [28, 29]. These data can confirm that 3-PPA molecule is bonded to the surface through the phosphonic group.

Fig. 5
figure 5

FT-IR spectra of ZnO NRs, 3-PPA powders and 3-PPA(1 min)/ZnO

Full size image

Figure 6 shows the J–V curves of the CdS/ZnO, CdS/ZnS/ZnO and CdS/ZnS/3-PPA/ZnO solar cells under simulated AM 1.5 G solar illumination of 100 mW/cm2. The photovoltaic parameters including the short-circuit current density (Jsc), the open-circuit voltage (Voc), the fill factor (FF) and the photovoltaic conversion efficiency (η) are summarized in Table 1. From the result, we can see that the Voc of CdS/ZnS/ZnO solar cell was increased by 0.1 V due to the deposition of ZnS layer in comparison with that of CdS/ZnO solar cell. Generally, the maximum Voc of photovoltaic cells is considered to be related to the offset between the valence band (VB) of the electron donor and the conduction band (CB) of the electron acceptor [30]. Thus, the enhancement of Voc can be attributed to the larger energy offset between VB of CdS and CB of ZnS than that between the VB of CdS and the CB of ZnO. Meanwhile, the JSC increased from 1.24 to 2.12 mA/cm2. Normally, the Jsc is mainly dependent on the photocurrent intensity when the solar cell works. The photocurrent density is not only related to the concentration of photogenerated carriers, but also the process of electron transfer and recombination. To obtain the related information, we further performed the UV–Vis absorption spectra of ZnO, ZnS/ZnO and ZnS/3-PPA/ZnO samples. Obviously, as shown in Fig. 7, we can see that the photoabsorption of ZnS/ZnO exhibits a increase in comparison with ZnO nanorods, indicating the concentration of photogenerated carriers will increase in solar cell. On the other hand, we also performed PL measurement on the ZnO, ZnS/ZnO and ZnS/3-PPA/ZnO samples. The PL spectra have been presented in Fig. 8, which consists of a UV emission located at 378 nm and a green emission band centered at 546 nm. The UV emission is attributed to the near band edge (NBE) emission of ZnO [31–33]. The green emission band is originated from ionized oxygen vacancies [34–36]. As shown in Fig. 8, the UV emission for ZnS/ZnO heterostructure does not exhibit any shift in comparison with that of ZnO nanorods, which is consistent with the results in the previous report [37]. To evaluate the green emission band, all PL spectra have been normalized to the UV emission of ZnO nanorods as shown in Fig. 8. Obviously, the intensity of green emission for ZnS/ZnO heterostructure enhances a little bit, which means the ZnS covering does not suppress the surface defects of ZnO nanorods and can not contribute to the increase of Jsc. Therefore, we can deduce that the increase of Jsc is mainly related to the enhanced photoabsorption of ZnS/ZnO heterostructure.

Fig. 6
figure 6

J-V curves of the CdS/ZnO, CdS/ZnS/ZnO and CdS/ZnS/3-PPA/ZnO solar cells under simulated AM 1.5 G solar illumination of 100 mW/cm2

Full size image
Table 1 Photovoltaic parameters obtained from the J–V curves of QDSSCs using various photoelectrodes
Full size table
Fig. 7
figure 7

UV-visible absorption spectra of ZnO, ZnS/ZnO and ZnS/3-PPA/ZnO

Full size image
Fig. 8
figure 8

Photoluminescence spectra of ZnO, ZnS/ZnO and ZnS/3-PPA/ZnO

Full size image

After introducing 3-PPA on the interface of ZnS/ZnO heterostructure, we can see from Table 1 that the best photovoltaic performance comes from the CdS/ZnS//3-PPA(1 min)/ZnO solar cell, which owns the highest η of 0.65 %, Isc of 2.71 mA/cm2, Voc of 0.67 V and FF of 0.35. Compared to the CdS/ZnS/ZnO solar cell, the η of CdS/ZnS/3-PPA(1 min)/ZnO solar cell significant increases from 0.51 to 0.65 %. Such enhancement of η should be mainly attributed to the increase of Jsc from 2.12 to 2.71 mA/cm2. To explain the increase of Jsc, we can see from Fig. 7 that no obvious change appears for the photoabsorption with introducing 3-PPA on the interface of ZnS/ZnO heterostructure, indicating the deposition of 3-PPA dose not significantly influence the uptake of ZnS on the ZnO surfaces. That is to also to say, the photoabsorption can not contribute to the variation of Jsc in CdS/ZnS/3-PPA/ZnO solar cell. Let us see Fig. 8, as compared to the green emission of ZnS/ZnO heterostructure, the intensity of green emission for ZnS/3-PPA(1 min)/ZnO is slightly suppressed. It is well known that the surfaces of ZnO nanorods grown with CBD method are prone to absorb various kinds of functional groups. According to the chemical reaction in the solution, these functional groups should be related to the elements such as carbon, nitrogen, and hydrogen [22]. After depositing 3-PPA for 1 min, the rupture Zn–O bond on the ZnO nanorods surfaces are combined with POOH of 3-PPA, which passivates a part of surface defects in ZnO nanorods. It will inhibit the recombination of electrons at the anode/QDs/electrolyte interfaces [37] and finally prompt the separation of photocharges and induce the increase of Jsc in CdS/ZnS/3-PPA(1 min)/ZnO solar cell. With increasing the immersion time of 3-PPA from 1 to 3 min, we can see from Table 1 that the conversion efficiency of CdS/ZnS/3-PPA/ZnO solar cells decreases again, which is accompanied by the decrease of Jsc. For all the CdS/ZnS/3-PPA/ZnO solar cells, the variation of Voc can be ignored due to the very small difference between them, so that Jsc is the main reason for the variation of their photovoltaic performance. As shown in Fig. 8, we can observe that the intensity of green emission exhibits a huge enhancement step by step with increasing the immersion time of 3-PPA. Once the depositing time is further prolonged, the passivating process was apt to saturated and the surface defects will be increased again due to the attachment of excessive 3-PPA. Therefore, the Jsc decreases and photovoltaic performance of solar cells turn worse again.

To further reveal the effects of 3-PPA on the photovoltaic performance of solar cells, we utilized the UPS technique to determine the analyze the energy level of ZnO nanorods, 3-PPA/ZnO, CdS/ZnS and ZnS/3-PPA/ZnO. Measurements with the UPS were conducted in ultra-high vacuum (UHV) at ~10−10 mbar, by irradiating with 21.22 eV photons(He I line). For each sample, the work function is calculated from the UPS spectrum by subtracting the energy of the incident beam from the difference between the Fermi edge and the low-energy cut-off of secondary electrons (spectrum ‘‘width’’) [38, 39]. This technique yields the minimum absolute value of the work function of the surface under examination, and is independent of any experimental parameters, except for the photon energy of the UV source [38, 39].

The He I UPS spectra show the samples in kinetic energy scale of ZnO, 3-PPA(1 min)/ZnO, ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO, respectively. The difference between Fermi level and vacuum level can be calculated by the Einstein photoelectric law as follow [40]

$${\text{E}}_{\text{vac}} - {\text{E}}_{\text{F}} = {\text{h}}\upnu - {\text{E}}_{\text{Cutoff}}$$
(1)
$${\text{IP}} = {\text{E}}_{\text{VBM}} - {\text{E}}_{\text{vac}}$$
(2)
$${\text{EA}} = {\text{E}}_{\text{CBM}} - {\text{E}}_{\text{vac}}$$
(3)

where Ecutoff is the location of the inelastic cutoff and hν is the incident photon energy.

IP is the ionization potential. EVBM is the energy location of valence band maximum. ECBM is the energy location of conduction band maximum. EVBM can be determined by choosing the point of maximum inflection near EF. Since UPS only probes occupied states, ECBM cannot be determined from the UPS spectrum. We utilized the energy band gap (Eg) determined from UV–Vis absorption measurements [41]. All the EF − EVBM and Ecutoff for four samples have been illustrated in the UPS spectra. For ZnO nanorods as shown in Fig. 9a, we can deduce that IP of ZnO is 6.78 eV. Since the Eg can be deduced from absorption spectra to be 3.17 eV, we can determine that EA of ZnO is 3.61 eV. In a similar way, for 3-PPA(1 min)/ZnO as shown in Fig. 9b, we can deduce that the IP is 5.74 eV and EA is 2.57 eV. Obviously, we can see that the EA of 3-PPA(1 min)/ZnO is increased by 1.04 eV than that of ZnO. That is to say, the lowest unoccupied molecular orbital (LUMO) level and highest occupied molecular orbital (HOMO) of 3-PPA increase 1.04 eV than the ECBM and EVBM of ZnO. Likewise, ECBM and EVBM of ZnS/3-PPA(1 min)/ZnO decrease 0.69 eV than that of ZnS/ZnO heterostructure.

Fig. 9
figure 9

UPS spectrum of ZnO, 3-PPA(1 min)/ZnO, ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO

Full size image

Figure 10a shows the schematic diagram of the electronic structure as derived from the dynamic UPS spectra of ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO in the above part. Figure 10b illustrates the energy level diagram for the CdS/ZnS/ZnO and CdS/ZnS/3-PPA(1 min)/ZnO solar cells. Obviously, the 3-PPA actually serves as a buffer layer to make the band alignment of soalr cells better, since its HOMO and LUMO levels are just located at the energy region between ZnO and ZnS. Meanwhile, we can also find that the 3-PPA layer contacts with both ZnO and ZnS surface and it can work on them at the same time, so the final energy level of 3-PPA/ZnS is adjusted to the even lower location. That is to say, a relatively smaller energy barrier has been formed between CdS and ZnO, which not only works efficiently on retarding the back transfer of electrons, but also weaken the blocking effect on the charge injection from CdS to ZnO. We believe this is the main reason for the enhancing photovoltaic performance of solar cell with introducing 3-PPA layer into the interface of ZnS/ZnO heterostructure.

Fig. 10
figure 10

a Schematic diagram of the electronic structure as derived from the dynamic UPS spectra of ZnS/ZnO and ZnS/3-PPA(1 min)/ZnO; b the energy level diagram for the CdS/ZnS/ZnO and CdS/ZnS/3-PPA(1 min)/ZnO solar cells

Full size image

4 Conclusions

In this work, we tried to introduce 3-PPA SAMs into the interface of ZnO/ZnS heterostructure by a simple chemical immersion process and investigated its effects on the photovoltaic performances of ZnO/ZnS heterostructure QDSSCs. The results revealed that the 3-PPA can act as buffer layer in ZnO/ZnS heterostructure to align the energy level of solar cells, which finally results in the enhancement of photoelectric conversion efficiency. Our results may enlighten more researchers to explore and design the proper SAMs to modify the interface of QDSSCs and stimulate more theoretical and experimental investigations on their effects on the photovoltaic performance of QDSSCs.