Spray-Deposited Aluminum-Doped Zinc Oxide as an Efficient Electron Transport Layer for Inverted Organic Solar Cells

: Spray-deposited thin films of zinc oxide (ZnO) and aluminum-doped zinc oxide (Al-ZnO) are characterized in detail to get insight into the role of a dopant in the matrix. ZnO and Al-ZnO are implemented as electron transport layers (ETLs) in inverted organic solar cells (IOSCs) with PTB7-Th as a donor and IEICO-4F as a nonfullerene acceptor, forming the bulk heterojunction (BHJ) photoactive layer. Organic solar cells (OSCs) based on the ZnO ETL exhibit a short-circuit current density ( J SC ) of 24.46 mA/cm 2 and an open-circuit voltage ( V OC ) of 0.68 V, yielding a power conversion efficiency (PCE) of 9.3%. A solar cell based on the Al-ZnO ETL yields a higher J SC of 25.16 mA/cm 2 and a V OC of 0.71 V, resulting in a PCE of 10.5%, which indicates that Al doping improves the device performance. Time-delayed collection field (TDCF) measurements yielded field-independent charge generation for both devices. Furthermore, steady-state photoluminescence (PL), time-resolved PL, and transient absorption measurements confirm reduction in the number of defect states in Al-ZnO thin films compared to ZnO thin films and efficient charge transfer, yielding an overall improved IOSC device performance.


INTRODUCTION
Organic solar cells (OSCs) exhibit several appealing properties, such as the possibility for roll-to-roll fabrication, light-weight, and device flexibility, favorable for solar cell development and deployment. 1 More than 19% efficient OSCs using donor polymers and small-molecule nonfullerene acceptors (NFAs) have been obtained by extensive research and development.-4F) are promising materials for OSCs due to their high performance. 5Power conversion efficiencies (PCEs) >12% of single-junction OSCs with PTB7-Th as an electron donor and IEICO-4F as a nonfullerene acceptor have been achieved using optimized device structures and advanced fabrication processes. 5verted organic solar cells (IOSCs) have several advantages compared to normal structures, such as device stability, extraction of electrons by inorganic metal oxide electron transport layers (ETLs), and the requirement of low-workfunction metal electrodes. 6,7−11 The extraction of electrons via the ETL in IOSCs depends on two factors: (1) the work function and (2) the conductivity of the ETL.Metal oxides with low work functions are preferred, resulting in lower series resistance and larger built-in potential.Zinc oxide (ZnO) is one of the most common transparent metal oxides that has been employed as the electron transport layer material in IOSCs due to its higher electron mobility, non-toxicity, solution processability, optical transparency, and enhanced electrical properties.The electron affinity and ionization energy of ZnO are ∼4.3 and 7.8 eV, respectively, which facilitates both hole blocking and electron extraction. 12The electron affinity of ZnO is also matched with the lowest unoccupied molecular orbital (LUMO) levels of many common polymers used in OSCs. 13,14wever, ZnO is associated with defect states, resulting in nonradiative charge carrier recombination losses, and hence detrimental to the device performance. 15,16Several approaches have been implemented to enhance the morphological, electrical, and chemical properties of ZnO, for instance, cross-linked layers, self-assembled monolayers, surface modification, and chemical doping.Doping of ZnO with boron (B), aluminum (Al), gallium (Ga), indium (In), etc., gives rise to improved layer properties, with Al-doped ZnO (Al-ZnO) being particularly important for solar cell applications. 17,18Details of the state-of-the-art IOSC devices using ZnO as the ETL with various dopants and different polymers including the deposition technique used are shown in Table 1.
Mostly, the spin coating technique has been used to deposit the ETL in inverted organic solar cells.During spin coating, a significant amount of material is lost in the initial stage of substrate spinning.Other deposition techniques, such as blade coating, spray, slot-die coating, screen printing, and inkjet printing, offer the advantage of easy scalability to large-area deposition on flexible as well as rigid substrates. 20,33,34Spray coating is a low-cost technique, and the deposition on a large area is eminently possible.However, many parameters influence the film properties during the deposition, namely, the deposition temperature, the solution concentration, the distance between the nozzle and heating surface, the atomization pressure, etc.
We deposited ZnO and Al-ZnO by the low-cost spray technique and used these as the ETLs in IOSCs.Spraydeposited ZnO and Al-ZnO thin films were characterized by their electrical, optical, and morphological properties.Lastly, the performance of solar cells based on different ETLs and PTB7-Th:IEICO-4F as a photoactive layer is presented here.

EXPERIMENTAL SECTION
Detailed about experiment and characterization techniques are available in the Supporting Information.

RESULTS AND DISCUSSION
Spray-deposited ZnO and Al-ZnO thin films are used as the ETLs in IOSCs.The J−V curves of IOSCs of the structures ITO/ZnO/PTB7-Th:IEICO-4F/MoO 3 /Ag and ITO/Al-ZnO/PTB7-Th:IEICO-4F/MoO 3 /Ag are shown in Figure 1a.The fabrication process of the IOSC along with the schematic diagram is provided in the supporting information (Figure S1).The detailed J−V parameters of IOSCs are summarized in Table 2.The IOSC based on the spraydeposited Al-ZnO ETL showed a PCE of 10.48%, while the IOSC based on the spray-deposited ZnO ETL showed a PCE of 9.29%.The short-circuit current density (J SC ) of the Al-ZnO-based ETL was 25.16 mA/cm 2 , which is higher than that of the IOSC based on the undoped ZnO ETL.The improvement in J SC may be due to the change in the conductivity and morphology of the Al-ZnO layer after Al doping.Moreover, the J−V parameters as a function of light intensity were quantified to explore the charge recombination mechanism.Figure 1b shows the dependence of J SC on the incident light intensity (I), which could be expressed via the following equation Under short-circuit conditions, if the bimolecular recombination is insignificant, α approaches ∼1.The α values were 0.95 for the IOSC with ZnO and 1.00 for the IOSC with Al-ZnO ETLs.This indicates that the bimolecular recombination is lower in the case of the IOSC with Al-ZnO as the ETL.
Second, the impact of illumination intensity on the opencircuit voltage (V OC ) is analyzed for IOSCs using ZnO or Al-ZnO as the ETL, as shown in Figure 1c.The V OC of the device depends on the light intensity, 35 and it can be expressed by i k j j j j j y where n is the ideality factor, k is the Boltzmann constant, T is the temperature in Kelvin, and q is the charge of the electron.The slope (S) can be obtained from a semilogarithmic plot by linear fitting.If S is 1 kT/q, then the IOSC characteristics can be described by the drift-diffusion model, where quasi-Fermi levels are aligned across the IOSC without any recombination caused by trap states.In fact, Al-ZnO-based IOSCs exhibit less trap-assisted recombination than ZnO-based IOSCs, as indicated by the S values of 1.27 and 1.03 for ZnO and Al-ZnO-based IOSCs, respectively.
The external quantum efficiency (EQE) of IOSCs based on ZnO and Al-ZnO is presented in Figure 1d.The EQE spectra show consistent responses across a broad spectral range.The maximum spectral response is observed with above 80% in the range of 600−700 nm.J SC values obtained from EQE spectra match the J SC values acquired from the J−V characteristics.Thus, the enhancement of J SC of Al-ZnO-based IOSCs is further confirmed by the EQE measurements.
We performed time-delayed collection field (TDCF) measurements to investigate the field dependence of charge generation for these two devices. 36A nanosecond laser pulse with a fluence of 0.1 μJ cm −2 was used to photoexcite the active layer of the device and to generate charges.To extract the photogenerated charges, a collection field was applied after 10 ns.The total extracted charge, namely, Q tot , was measured as a function of the applied bias to investigate a possible field dependence on charge generation.However, both devices showed field-independent charge generation and reduced geminate recombination, as shown in Figure S2 in the Supporting Information.
The work functions of ZnO and Al-ZnO ETLs were measured to acquire more understanding of the reasons for the voltage enhancement of IOSCs.The work function values of ITO/ZnO and ITO/Al-ZnO measured from PESA are 4.31 and 4.12 eV, respectively (Figure S3a).The enhancement in V OC of the Al-ZnO-based IOSC might be due to the changes in their work functions.The energy barrier height is the difference between the work function of the ETL and the LUMO level of IEICO-4F (Figure S3b).This energy barrier is also called a Schottky barrier.The Schottky barrier for the Al-ZnO/IEICO-4F interface (∼0.22 eV) is smaller than that of the ZnO/IEICO-4F interface (∼0.41 eV).This reduces charge recombination and improves electron transport from the PTB7-Th:IEICO-4F photoactive layer to the Al-ZnO ETL layer because of better alignment of energy levels. 22,37Thus, the difference in work function between the ZnO and Al-ZnO ETL appears to improve the V OC and fill factor (FF) of the IOSC devices.
The sheet resistance values of ITO/ZnO and ITO/Al-ZnO measured by the four-probe method are 13.1 and 11.3 Ω cm 2 , respectively.When 1 atomic % Al is added to the ZnO thin film, the Al is ionized to Al 3+ and substitutes Zn 2+ .One free valence electron is produced when one zinc atom is replaced by the aluminum atom, and thus, 1 atomic % aluminum atoms create a large number of free electrons in the Al-doped ZnO films, and thus, the resistivity decreases.The improvement of the conductivity reduces the device's ohmic losses.The improvement of J SC in the case of Al-ZnO as the ETL is due to a change in the conductivity of the ZnO thin film upon the addition of Al.The FFs of the IOSCs are slightly lower compared to the reported values of the FF for this donor/ acceptor system.The roughness of the ETL plays an important role in interface between the ETL and active layer and affects the FF of the device.Atomic force microscopy (AFM) was used to study the film roughness, and the surface topography of spray-deposited ZnO and Al-ZnO thin films at the scale of 5 μm × 5 μm is shown in Figure 2a,b, while Figure 2c displays the corresponding surface height histograms extracted from those images.The spray-deposited Al-ZnO film exhibited a smoother surface compared to the ZnO thin films.Also, the Al-ZnO film has a lower root-mean-square roughness (R rms ) value of 9.54 nm (ZnO = 14.0 nm) and thus shifts the entire height distribution toward lower values.The change in the topography of the Al-ZnO thin films compared to the ZnO thin films is due to a change in the crystallite size and hence a reduction in the average roughness of the Al-ZnO thin films.The change in the roughness is also caused by scattering of Al atoms.The maximum FF obtained for Al-ZnO was 58.63%, which is higher compared to the IOSC fabricated with ZnO as the ETL.The lower FF of both devices compared to the reported data is a consequence of the high roughness of spraydeposited ZnO and Al-ZnO layers as compared to spin-coated films. 24,25he XRD patterns of ZnO and Al-ZnO are shown in Figure S4.Spray-deposited ZnO and Al-ZnO films showed peaks of the (100), ( 002), ( 101), ( 102), ( 110), (103), and (112) planes, while no diffraction peaks were detected, which belong to metallic Al, Zn, and Al 2 O 3 . 18The intensity of Al-ZnO, especially at the (002) plane, is less compared to ZnO films, and the peak is slightly shifted toward higher values of 2θ.This indicates the successful doping of Al atoms into the ZnO lattice.The decrement in the intensity of the Al-ZnO film is due to the incorporation of the Al atom into the ZnO lattice, which is responsible for the disorder in the lattice due to the size of Zn 2+ since Al 3+ is not equal in size and hence blander XRD intensity of Al-ZnO. 38Lattice parameters details, such as average crystallite size, micro strain, dislocation density, and the number of crystallite size, are shown in Table 3.
The crystallite size calculated from the Scherrer formula 39 (available in the Supporting Information) decreased in Al-ZnO (18.48 nm) compared to undoped zinc oxide (22.96 nm).The slight decrease in the crystallite size is due to the doping of Al atoms into ZnO.The micro strain is increased in Al-ZnO thin films compared to undoped ZnO.The change in the micro strain is due to the development of stress in the lattice due to the doping with Al atoms.The dislocation density (δ) and no  crystallite size (N) are also shown in Table 3.The values of δ and N depend on the crystallite size of the thin films.Contact angle measurements were carried out with water to further explicate the improvement in the performance of Al-ZnO-based IOSCs.The lower contact angle of (90.2°) for the ZnO film reveals its hydrophilic nature, and the higher contact angle for Al-ZnO (100.6°)shows the hydrophobic nature of its surface (Figure 3).The hydrophobic surface is beneficial for the intimate contact between the ETL and the photoactive layer, which leads to improvements in the device performance.
To further investigate the optical properties of ZnO and Al-ZnO, steady-state photoluminescence (PL) spectra were recorded (Figure 4a).Generally, ZnO films have two emission bands, which relate to the ultraviolet emission band and a broad visible emission band.The intensity of the ultraviolet emission band at ∼390 nm is determined by the band-to-band transitions, and it is due to the recombination of photogenerated electrons and holes in the conduction and valence bands, respectively.The visible emission band at ∼500 nm is due to the transitions between band edge states and localized states present in the semiconductor band gap. 40The high intensity of the band-to-band peak detected in Al-ZnO thin films, and the slight shift toward lower wavelength compared to the undoped ZnO thin film is due to a relapse of the crystal structure after doping.Moreover, the peak intensity at 450− 550 nm in Al-ZnO is lower than that of undoped ZnO.It can be understood that ZnO films are passivated by Al atoms, leading to the reduction of oxygen vacancies of ZnO, which reduces recombination and thus improves the power conversion efficiency of the IOSCs.The optical band gaps (E g ) of ZnO and Al-ZnO calculated using the Tauc plot are 3.23 and 3.30 eV, respectively (Figure 4b). 41e performed time-resolved photoluminescence (TRPL) experiments to investigate the photoexcited-state dynamics.ZnO and Al-ZnO thin films were measured upon photoexcitation at 370 nm.Selected time-integrated spectra are shown in Figure 5a,b, while the transients were tracked at the photoluminescence (PL) peak position and are presented in Figure 5c.The Al-ZnO sample exhibited a shorter PL lifetime compared to the ZnO sample, which appears due to a reduction in the number of defects states in ZnO as also indicated by the steady-state PL spectra (Figure 5c).The shortened lifetime is directive of efficient charge transfer/  extraction to Al-ZnO and hence the improvement in V OC and J SC of IOSCs. 42or an in-depth comparison of the recombination losses causing the differences in J SC , we carried out transient absorption (TA) spectroscopy measurements in the ps−ns time range and subsequently evaluated the excited-state dynamics.The ZnO and Al-ZnO thin films were photoexcited at 350 nm, where both components of the ZnO and Al-ZnO absorb, and the acquired TA data is presented in Figure 5d,e.We clearly observed the ground-state bleach signal at 370 nm without any further spectral shift during time evolution.For comparison, we track the kinetics (Figure 5f) of the groundstate bleaching, which yielded a prolonged decay time of the Al-ZnO sample.This is a consequence of less non-radiative recombination, hence facilitating efficient charge extraction, and in line with the observed higher photocurrent.In conclusion, our TRPL measurements reveal more efficient charge extraction, while the difference in TA kinetics is assigned to losses by non-radiative recombination.

CONCLUSIONS
In summary, we successfully fabricated nonfullerene-based IOSCs by spray-deposited ZnO and Al-ZnO ETLs.Al-ZnO ETL-based devices exhibited an ∼12.8% higher efficiency compared to the IOSC based on ZnO as the ETL.The enhancement in the V OC of Al-ZnO-based cells is due to efficient charge transfer supported by the electron affinity of the ETL.The overall improvement in Al-ZnO-based devices is due to enhanced optical properties and optimal surface morphology leading to efficient charge transfer.Transient absorption spectroscopy revealed less non-radiative recombination in devices with Al-ZnO.Time-resolved PL measurements exhibited shorter lifetimes for Al-ZnO compared to its counterpart, hence more efficient charge extraction, which is in line with the increased photovoltaic performance of the IOSC.The Al-ZnO thin films deposited by a low-cost spray coating technique can be useful for fabricating the large-area IOSC devices.

Figure 1 .
Figure 1.(a) J−V curve of IOSCs using ZnO and Al-ZnO as ETLs, (b) current density versus light intensity of IOSCs, (c) open-circuit voltage versus light intensity of IOSCs, and (d) EQE of IOSCs based on spray-deposited ZnO and Al-ZnO.

Figure 2 .
Figure 2. AFM images of the spray-deposited thin films: (a) ZnO, (b) Al-ZnO, and (c) surface height histograms of ZnO and Al-ZnO films.

Figure 3 .
Figure 3. Contact angle measurement for (a) ZnO and (b) Al-ZnO films deposited by spray technique.

Figure 5 .
Figure 5. Time-resolved PL measurements following photoexcitation at 370 nm.(a) Selected time-integrated spectra of ZnO and (b) Al-ZnO, (c) associated PL transients tracked at the respective peak positions of the ZnO and Al-ZnO films, (d) selected time-integrated TA spectra of ZnO upon photoexcitation at 350 nm, (e) selected time-integrated TA spectra of Al-ZnO, and (f) picosecond−nanosecond TA kinetics of ZnO and Al-ZnO thin films.

Table 1 .
Solar Cell Performance with Doped ZnO Used as the ETL in Inverted Organic Solar Cells with Different Polymers

Table 2 .
Figures of Merit of IOSC-Based Devices

Table 3 .
Lattice Parameters of Spray-Deposited ZnO and Al-ZnO Thin Films