Thickness dependent properties of NiO thin films on ITO/PET flexible substrates for solar cell applications Authors: Ibrahim Aldawood, Ahmad Althumali, Syed Mansoor Ali, Saad Aldawood Published: 2026 ABSTRACT AbstractNiO thin films (10–40 nm) were grown on polyethylene terephthalate (PET) substrates coated with indium tin oxide (ITO) using molecular beam epitaxy (MBE) to evaluate their potential for flexible optoelectronic applications. The X-ray diffraction (XRD) analysis reveals a single NiO (002) textured growth with strain-related peak shifts arising from the flexible substrate. The X-ray photoelectron spectroscopy (XPS) analysis confirms that Ni²⁺ is the dominant oxidation state. Surface morphology and roughness were investigated using scanning electron microscopy (SEM) and atomic force microscopy (AFM), showing a clear evolution from discontinuous dome-like islands at ~ 10 nm thickness to a compact and smooth film at 40 nm, with the surface roughness decreasing from 2.96 to 0.26 nm, respectively. Optical characterisation using photoluminescence (PL) and UV-visible absorption spectra shows that film thickness strongly influences emission and absorption: PL intensity reaches a maximum for the 30 nm NiO film due to reduced non-radiative recombination losses. Thicker films exhibit a redshift and slight bandgap narrowing (3.98 → 3.92 eV), attributed to defect-induced band tails. Electrical measurements show a decrease in resistivity from 0.3806 to 0.1477 Ω·cm as thickness increases, consistent with enhanced hole concentration from Ni vacancies and oxygen interstitials. Overall, increasing the NiO thickness improves film uniformity, optical performance, and p-type conductivity, highlighting 30–40 nm NiO layers as promising hole-transport materials for flexible solar and optoelectronic applications.Keywords: NiO thin films, molecular beam epitaxy deposition, flexible optoelectronics, surface morphology, photoluminescence, bandgap narrowing, p-type conductivity, hole-transport layer.Subject terms: Materials science, Nanoscience and technology, Optics and photonics, Physics Introduction Transition metal oxides have gained considerable attention due to their high performance in different applications1,2. Among these, nickel oxide (NiO) stands out due to its low cost, robustness, superior electrochemical stability, intrinsic p-type semiconducting behaviour, and high optical transparency3,4. Additionally, its deep valence band level (~ 5.4 eV) and diverse fabrication methods make it a particularly versatile material3,4. NiO is one of the few p-type semiconducting oxides that exhibits a large direct optical bandgap ranging from 3.6 to 4 eV5. It is a highly promising material that is used as a hole transport layer (HTL) for various applications, including photovoltaics (PVs), dye-sensitised solar cell (DSSC) technologies6, light-emitting diodes7,8, gas sensors9, catalysis10, and Li-ion batteries as anode materials11. The integration of functional oxide materials such as NiO onto flexible substrates has become an area of interest for next-generation optoelectronic applications12–15. Despite the fabrication of NiO films using several growth techniques for solar cell applications, such as the sol-gel process16, the pulsed laser deposition (PLD) method17,18, and the atomic layer deposition (ALD) method19, no studies have reported the use of the molecular beam epitaxy (MBE) growth method to deposit undoped or doped NiO film for solar cell applications. Most of the research has predominantly focused on rigid substrates such as glass; the growth of NiO thin films onto flexible substrates, polyethylene terephthalate (PET) coated with indium tin oxide (ITO) using MBE, exhibits novel opportunities. The performance of NiO in such applications strongly depends on the growth techniques employed and the film thickness, both of which directly influence its structural, optical and electrical properties18–23. In particular, the use of thin NiO films (e.g., 7 nm PLD-grown18 or 20 nm sputter-grown20 may lead to incomplete coverage of the substrate, resulting in insufficient blocking of photogenerated electrons18,20. Moreover, thinner NiO layers tend to exhibit smaller grain sizes, thereby increasing the density of grain boundary defects in the overlying solar cell absorber layer21. In contrast, excessively thick NiO films (e.g., 90 nm) can introduce intrinsic photon absorption losses18. Taken together, these thickness-dependent effects negatively impact the overall device performance18,20,21. Despite these observations, a systematic examination of thickness-dependent behaviour on flexible ITO-coated PET substrates using high-precision deposition techniques remains limited. Therefore, this study investigates the structural, optical, and electrical evolution of NiO thin films deposited on ITO-coated PET substrates with varying thicknesses using the MBE growth method. Determining the optimal NiO thickness that provides a suitable balance between optical transparency, surface coverage, and electrical conductivity are the main objectives of this work, thereby providing fundamental insight and practical guidance for the integration of NiO thin films into flexible optoelectronic devices, including solar cells. Experimental details Sample growth A deposition series of NiO films with varying thickness was performed on ITO-coated PET substrates using MBE. Prior to deposition, all substrates underwent a cleaning process in an ultrasonic bath, starting with isopropanol followed by deionised water for 10 min each. After loading the substrates into the growth chamber, all the deposited NiO thin films were grown while maintaining the Ni effusion cell temperature at 1350 ℃, with a partial pressure of 5 × 10− 6 mbar. High-purity oxygen gas (99.999%) was introduced into the growth chamber from an external oxygen cylinder through a precision needle valve, which was used to regulate and maintain the oxygen partial pressure at mbar during the deposition process. The substrate temperature was initially maintained at room temperature; however, during the deposition process, the substrate temperature gradually increased due to thermal radiation from the high-temperature Ni effusion cell, reaching a maximum temperature of approximately 55 °C by the end of the deposition process. No pre-or post-deposition annealing treatments were performed in order to avoid potential thermal deformation or degradation of the flexible ITO-coated PET substrates. These controlled growth conditions ensured good reproducibility and uniformity of the deposited NiO thin films. The film thickness and growth rate were calibrated prior to the deposition on ITO-coated PET substrates by depositing NiO thin films on a single-crystal SrTiO3 (STO) (001) substrate under identical growth conditions. The use of the STO(001) substrate was necessary to obtain reliable X-ray reflectivity (XRR) measurements due to its atomically smooth surface and high structural uniformity, which can enable clear interference fringes and accurate thickness determination. In contrast, the relatively high surface roughness and structural complexity of the flexible ITO-coated PET substrates limited the quality and reliability of the XRR analysis. The thickness of the reference NiO films deposited on STO(001) was determined using XRR analysis combined with GenX 3 simulation software and further confirmed by cross-sectional TEM measurements. Based on these calibration experiments, the NiO growth rate was estimated to be approximately 2.66 Å/min. The desired film thicknesses were subsequently achieved by controlling the deposition time while maintaining identical growth conditions for all samples. The detailed XRR and TEM calibration results obtained from the STO(001) reference samples are part of a separate study focused on single-crystal substrates and will be reported in a future publication. Characterisations The crystal structure of the thin films was characterised by X-ray diffraction (XRD). This measurement was performed using a Rigaku SmartLab diffractometer, with radiation source of λ = 1.54050 Å, using the mode. The obtained XRD patterns were smoothed to enhance clarity and analysis. The analysis of the XRD data was conducted utilising crystallographic information files (CIF) references from the Inorganic Crystal Structure Database (ICSD). These are; ICSD 19,755 for the MgO substrate, ICSD 112,324 for the NiO film, and ICSD 128,532 for the CoO film. The measurements of the optical properties were conducted using ultraviolet–visible spectroscopy (UV–Vis) with an Agilent Cary 60 UV-Vis spectrophotometer. Elemental chemical analysis was performed by energy-dispersive spectroscopy (EDX), using a JSM-7800 F field-emission scanning electron microscope (SEM) at 20 kV, over an area of 240 180 . The system is equipped with a dual EDX detectors from Oxford Instruments (Ultim Max 170, with a sensor area of approximately 170 mm2). The elemental composition quantifications were conducted using Oxford Aztec software (version 6.0). Atomic force microscopy (AFM) was used to assess the surface roughness. This was carried out using Bruker NanoScope V in tapping mode under ambient air conditions. RTESPA-300 cantilever tip was used during the measurements over an area of 2 2 μm², with a scan rate of 1 Hz and a high image resolution of 512 512 pixels, to provide detailed surface features. In order to determine the oxidation states of elements near the film surfaces, the NiO films were examined by X-ray photoelectron spectroscopy (XPS), utilising Thermo fisher Multilab 2000, using a monochromatic Al Kα X-ray source (1486.7 eV). The XPS measurements were calibrated with respect to the C1s peak (binding energy = 284.2 eV). XPS peak fitting was performed using CasaXPS software, employing a standard Shirley background and a mixed Gaussian–Lorentzian peak shape (70% Gaussian, 30% Lorentzian). Photoluminescence (PL) spectra measurements of NiO thin films were performed at room temperature using a spectrofluorometer (JASCO FP-8200) with an excitation wavelength of 325 nm. Hall-effect measurements were conducted using the Van der Pauw configuration to investigate the electrical properties of the NiO thin film. Results and discussion Structural and surface morphological investigations The XRD analysis was performed to investigate the crystallinity as well as the structural properties of NiO thin films with varying thicknesses deposited by MBE on ITO-coated PET substrates. The obtained results reveal characteristic features corresponding to both the NiO films and the substrate, as shown in Fig. 1. A small hump at approximately 15°, along with three distinct reflection peaks at 26°, 46.8°, and 53.8°, was observed, which correspond to the planes of PET present in all samples24. The dominant peak at 26° is attributed to the (001) plane, consistent with previously reported studies on semi-crystalline PET structures24–27. In addition, a diffraction pattern located at 21.64° corresponds to the (211) crystallographic plane of the polycrystalline ITO, as identified by ICSD code no. 50,848 and previously reported in the literature28,29. The peak reflected at 41.16° is assigned to the (002) plane of NiO, which indicates nanocrystalline NiO grown under ultra-high vacuum MBE conditions. X-ray diffraction patterns of NiO thin films with varying thicknesses deposited on ITO-coated PET substrates. The observed NiO peaks exhibit a shift towards a lower angle in comparison with NiO powder (43.4°) without a substrate, according to ICSD code no. 112,324, corresponding to a shift of approximately 2.24°. This shift is consistently present across all film thicknesses, while the substrate peak remains unchanged. Therefore, the observed shift in the NiO diffraction peak may be associated with strain/lattice distortion effects induced by the underlying ITO-coated PET substrate. Similar substrate-induced peak shifts have previously been reported in other oxide thin film systems deposited on flexible substrates24,26. Such as ZnO films deposited on ITO-coated PET substrates, where the XRD presented peak shifts due to internal stress and lattice distortion as a result of substrate effects26. In addition, Can et al.24 reported substrate-dependent diffraction peak shifts for PANI films grown on ITO-coated PET substrates, which can be assigned to strain and crystallographic orientation changes induced by the flexible substrate. These findings further support that the observed NiO peak shifts are primarily driven by substrate-induced strain. The XRD patterns reveal a single diffraction peak corresponding to the (002) plane in all thin films, with no secondary phases detected, indicating the phase purity of the NiO films. Secondary electron SEM images were acquired to investigate the morphological evolution of NiO thin films with increasing thickness grown on ITO-coated PET substrates. Figure 2a shows a 10 nm thick NiO thin film, exhibiting localised island formation, resulting in incomplete coverage, likely driven by the roughness of the underlying ITO-coated PET substrate. This implies that at the initial state of the growth, the NiO film follows the morphology of the substrate by filling uneven surface features, hence the deposition of a flat continuous layer is unlikely. As the thickness increases to 20 and 30 nm, the NiO film becomes progressively more homogeneous, as evidenced by surface morphology in the SEM images shown in Fig. 2b and c. A further increase in film thickness to 40 nm (Fig. 2d) results in smoother and more compact film morphology, indicating that the film has grown well and the substrate surface is fully covered, which improves the adhesive nature of the NiO films to the ITO-coated PET substrate. This morphological improvement plays an important role in flexible solar cell applications, in which homogeneity and continuity are essential for optimal charge transport and device stability30. Therefore, these compact and continuous NiO thin films are suitable for optoelectronic device fabrication30. SEM images of NiO thin films with varying thicknesses grown on ITO-coated PET substrates using MBE for flexible solar cell applications: (a) 10 nm, (b) 20 nm, (c) 30 nm, and (d) 40 nm. Energy-dispersive X-ray spectroscopy (EDX) was conducted to examine the elemental composition and depth profile behaviour of NiO thin films grown on ITO-coated PET substrates. Table 1 presents a systematic change in the atomic percentages of Ni, O, C, and In + Sn as the NiO thickness increases. The relative atomic percentages of Ni and O increase as increasing the thickness, starting from 0.88 at% and 33.15 at% at 10 nm to 3.89 at% and 46.93 at% at 40 nm, respectively. This increase arises from the increased electron interaction volume within the NiO layer as the film becomes thicker. The relatively high oxygen signal in the EDX spectra reflects a cumulative O contribution from the entire multilayer structure, encompassing the PET substrate, ITO layer, and NiO thin film. In contrast, the carbon peak, which is attributed to the PET substrate, shows a decrease in the atomic percentage from 62.94 at% at 10 nm to 45.15 at% at 40 nm thickness. This decrease is an indication of signal attenuation from the substrate, as the NiO film thickness increases. The In and Sn signals originating from the ITO layer, exhibiting a slight increase in the atomic percentage, despite the thickness of the ITO layer remaining constant across all samples. This change is attributed to variations in the electron beam penetration depth and interaction volume as the NiO thin film becomes thicker and more uniform, which alter the proportion of X-rays generated from the underlying ITO region. Therefore, as the NiO layer suppresses deeper electron penetration, it affects how deeply the electron beam excites the ITO layer beneath. Particularly, this behaviour can also be influenced by the ITO-coated substrate with a thickness of 130 nm in which it exists as a thin film, rather than a bulk material. However, the observed slight variations in the In and Sn atomic percentages are not expected to indicate an actual compositional change in the ITO layer but rather result from these interaction and excitation effects. It is worth noting that there is an overlap between In and Sn peaks, since they have a similar excitation energies of Lα at 3.29 keV and Lα at 3.44 keV, respectively. Table 1Atomic percentages of Ni, O, In + Sn, and C in NiO thin films deposited on ITO-coated PET substrate derived from EDX. In and Sn are grouped due to peak overlap in EDX spectra.Thickness (nm)Ni (at%)O (at%)In + Sn (at%)C (at%)100.8833.153.0262.94202.0239.813.5054.67302.7942.833.8150.57403.8946.934.0045.17 Atomic percentages of Ni, O, In + Sn, and C in NiO thin films deposited on ITO-coated PET substrate derived from EDX. In and Sn are grouped due to peak overlap in EDX spectra. Atomic force microscopy (AFM) was conducted to investigate the surface characteristics of NiO thin films deposited on ITO-coated PET substrates as a function of film thickness. Figures 3a-d presents high-resolution 2D AFM images of all films obtained using a scan area of 2 × 2 μm² for all thin films. To ensure reliable and representative roughness measurements, multiple scans were conducted at randomly selected regions of each sample for all NiO thicknesses. The reported Rq values represent the average of these measurements together with their corresponding standard deviations, as illustrated in Fig. 3e. All AFM images were acquired under identical scanning conditions (2 × 2 μm² scan area, tapping mode) using a consistent colour scale to enable direct comparison of the surface morphology among the films. The results confirm that the observed reduction in surface roughness with increasing NiO thickness is not associated with local surface variations, but rather reflects improved film uniformity and densification. 2D AFM topography images of NiO thin films with different thicknesses: a) 10 nm, b) 20 nm, c) 30 nm and d) 40 nm. e) Root mean square surface roughness (Rq) as a function of NiO film thickness. The findings demonstrate an inverse relationship between the root mean square surface roughness (Rq) of NiO thin films and the film thickness, decreasing from 2.960.18 nm at 10 nm to 0.26 0.01 nm at 40 nm NiO thickness, as illustrated in Fig. 3e. The deposited film begins to fill native surface irregularities of the ITO-coated PET substrate, resulting in higher surface roughness compared with the other samples. A similar decrease in substrate-induced roughness after the deposition of the NiO layer was observed by Xiao et al.31 and Yang et al.32, who stated that depositing a thin NiO layer smoothened the ITO surfaces and enhanced uniformity. When the NiO thin film thickness increased to 20 and 30 nm, higher continuity and improvement of the surface roughness were obtained, resulting in a reduction in the surface roughness to 1.670.50 nm and 0.390.11 nm, respectively. This is consistent with Gostishchev et al.33, who found the surface roughness of the NiO layer deposited by ion beam sputtering to be approximately 0.84 nm at 20 nm. In addition, Abdullah et al.34 obtained a low surface roughness of approximately 0.4 nm NiO thin films, indicating highly crystalline and nanostructured NiO films. The surface roughness further improved at 40 nm thickness, reaching an Rq value of 0.26 0.01 nm, indicating a nanocrystalline morphology with minimal grain boundaries. A very high uniformity and smoothness of the surface facilitate the full coverage of the absorber layer within the device since it assists the transfer of charge carrier35. These findings align with previously reported studies, which observed a similar trend of smooth NiO surfaces36. It is worth mentioning that a surface roughness of 0.3 –0.4 nm was reported for a 10 nm NiO thin film grown by ALD, confirming the feasibility of obtaining such low roughness at nanoscale thicknesses37. The surface roughness analysis reveals an improvement in the morphology, resulting in a decrease in surface roughness (Rq) as the NiO film thickness increases from 10 nm to 40 nm, indicating a transition from a discontinuous to a compact and nanocrystalline morphology. This observation is further supported by SEM imaging (Fig. 2), where thinner films exhibit localised domes or accumulations, while thicker films display dense and uniform surfaces. Therefore, these morphological enhancements have a strong influence on the optical and electrical characteristics of the device performance, which will be further discussed in the following sections. In order to obtain insight into the surface chemical oxidation states of the NiO thin films, XPS was performed. The XPS analysis was carried out on the thickest NiO thin film (40 nm) to confirm the chemical composition, oxidation state, and phase purity of the deposited NiO on the flexible ITO-coated PET substrate. The XPS investigation was intended to verify the successful formation of NiO rather than to establish thickness-dependent chemical variations among the grown films. The Ni 2p₃/₂ core-level spectrum of a representative 40 nm NiO film grown on a flexible ITO-coated PET substrate is shown in Fig. 4. The spectrum exhibits a characteristic main Ni peak at approximately 854 eV, accompanied by a distinct shoulder near 855.5 eV and a shake-up satellite peak at a higher binding energy around 861 eV38–40. These features confirm the formation of NiO with Ni²⁺ as the dominant oxidation state, in agreement with the expected chemical structure observed in the XRD results. An XPS spectrum of the Ni 2p₃/₂ core-level of 40 nm NiO film grown on a flexible ITO–coated PET substrate. The shoulder peak feature in the Ni 2p₃/₂ spectrum is attributed to multiplet splitting and nonlocal screening processes involving transient charge transfer between Ni and O atoms, which give rise to complex final-state configurations that may appear as Ni3⁺-like spectral character, rather than a true change in the bulk oxidation state40–43. The nonlocal screening stabilises the system by neutralising the core hole charge through interactions with surrounding oxygen states41–43. Furthermore, the observed satellite features arise from multiplet splitting and final-state effects associated with interactions between the Ni 2p core hole and unpaired electrons in the partially filled Ni 3d orbitals41,44. This effect is particularly pronounced in transition metal compounds such as NiO due to the strong electron correlation and the influence of nonlocal screening interactions. Consequently, a core hole is generated following the emission of a core electron during the XPS process. The core hole can be partially screened by charge transfer from neighbouring atoms, notably O 2p valence states, resulting in the appearance of satellite features at higher binding energies41,44. Optical analysis Room temperature photoluminescence (PL) spectroscopy was conducted to identify the optical emission properties of NiO thin films deposited on ITO-coated PET substrates. As shown in Fig. 5a, the acquired emission was analysed in the range of wavelengths from 350 to 650 nm, including the bare ITO-coated PET substrate, to distinguish the contributions from the NiO layers. The PL spectrum of the ITO-coated PET substrate presented three distinct peaks positioned at approximately 370 nm, 390 nm, and 408 nm. The emitted peak at 370 nm could be attributed to near-UV emission of the PET polymer. Whereas the two peaks emitted at 390 nm and 408 nm are assigned to the near band edge (NBE) transition of the ITO, since the photogeneration of the holes near the valence band recombine with electrons in the donor levels via a radiative process45. These findings are consistent with a previously reported study which observed similar emission behaviour in ITO-coated PET-based structures46. Additionally, these peaks remained present in all samples after the deposition of NiO films with different thicknesses; however, their intensity and sharpness were changed, indicating the presence and the contribution of the NiO layer. a) Room-temperature photoluminescence emission spectra of variant thicknesses of NiO thin films on flexible substrate ITO-coated PET. b) Highlights the region between 590 nm and 645 nm wavelengths to show the emission originating from the NiO thin film. A reduction in the intensity of the substrate-related peaks was observed at a NiO thickness of 10 nm, which may be attributed to several factors, such as thickness, morphology47,48, crystallinity, and surface modification48. In addition, the intensity reduction in the UV to the visible regions (390 nm and 408 nm) occurs due to less electron density in the valence band associated with the thin NiO layer, resulting in a suppression of the excitonic transitions between the bands49,50. Furthermore, this decrease can result from the quenching centres that are induced by defect states, which suppress the excitation of the electrons and enhancing the non-radiative transitions49. When the NiO thickness increased to 20 nm, a moderate increase in PL intensity was observed, suggesting improved structural quality of the film. The emission observed for the 30 nm thickness of NiO shows a remarkable enhancement in the peak intensity and sharpness at 370 nm, 390 nm, and especially 408 nm. This improvement suggests that the 30 nm thickness of the NiO exhibits better surface morphology and reduced surface defect density, assisting in minimising non-radiative recombination, thereby increasing the PL emission47,51. Among them, the peak at 408 nm exhibited the highest intensity, indicating a better balance between the film thickness and defect density. Despite the XRD patterns of all NiO films exhibit similar peak positions and widths, indicating comparable crystallinity. A noticeable decrease in PL intensity was observed for all characteristic peaks when the thickness of the NiO was increased to 40 nm. The reduction in the intensity became lower than that recorded for 10 nm thickness. This behaviour may be associated with an increase in defect-related non-radiative recombination centres, such as nickel vacancies, oxygen-related defects, and interstitial states, which become more influential as the film thickness increases49,52. These defect states can trap photogenerated charge carriers and suppress radiative recombination, resulting in reduced PL intensity. Similar thickness-dependent PL behaviour associated with defect-related non-radiative recombination has been reported in oxide semiconductor thin films53. In addition, Auger recombination may become dominant at high carrier concentrations since the extra energy resulting from an electron-hole pair is transfered to a free carrier rather than be emitted as a photon54. Thus, a reduction in PL intensity will be observed since the reabsorption of the emitted photons by free carriers would further weaken the emission. It is worth noting that two broad emitted peaks located at 601 nm and 632 nm were observed in all NiO thin films with different thicknesses (Fig. 5b), which are assigned to orange-red and red emission, respectively. These peaks are not detected in the bare ITO-coated PET spectrum, confirming that they originate from the NiO layers. These emissions are attributed to deep-level defects, originating from the transitions between the conduction band and oxygen interstitial defect states (Oi)49,55. This observation supports the possible contribution of interstitial oxygen as a defect in facilitating radiative transitions in the visible region. These findings play an important role in understanding and optimising the NiO optical behaviour in such applications as flexible solar cells by controlling the defect levels and emission efficiency. Therefore, although increased thickness generally yields stronger emission, the 30 nm thickness of NiO film presented the highest emission intensity at longer wavelengths. This further supports that a 30 nm NiO thickness can offer a favourable configuration to achieve optimal optical performance as an efficient HTL in solar cell applications. The optical properties of the NiO thin film layer were further investigated using the UV-Vis spectrophotometer to study the influence of the thickness variation on the ITO-coated PET substrate. Figure 6 represents the absorption spectra of the bare ITO-coated PET substrate alongside the NiO thickness variation ranging from 10 nm to 40 nm. The optical absorbance spectra as a function of wavelength for different NiO thin film thicknesses deposited on ITO-coated PET substrates. It can be seen that all NiO samples with different thicknesses exhibited higher absorption than the bare substrate, starting from the UV region (320 nm) towards the visible region, in agreement with previously reported study21,23. A consistent redshift in the absorption edge was observed as NiO thickness increased, indicating a reduction in the optical bandgap. This narrowing is likely attributed to the relaxation of strain at the NiO/ITO interface and the evolution of the film structure towards bulk properties21,56. In addition, the absorption spectra indicate that NiO films exhibit and maintain low absorption in the visible range, allowing a high amount of photon flux to pass through and reach the absorber layer. Consequently, more photons are available for photoexcitation, resulting in an enhancement of charge carrier generation, which may contribute to the efficiency of device performance. To quantify the bandgap variation, the Tauc relation, Eq. (1) was employed by plotting as a function of photon energy assuming a direct allowed transition. The optical bandgap was extracted by linear fitting of the absorption edge region followed by extrapolation of the fitted line to 57,] 58], as illustrated in Fig. 7. The fitting was performed within the most linear region of the spectra to ensure accurate bandgap determination. Furthermore, the uncertainty associated with the linear fitting procedure was evaluated and incorporated into the extracted bandgap values, as presented in inset in Fig. 7. Tauc plot for bandgap measurements of NiO with different thickness deposited on ITO-coated PET substrates. The inset shows the estimated bandgap as a function of NiO film thickness. Where α corresponds to the absorption coefficient, h refers to Planck’s constant, is the photon’s frequency, is the bandgap, B represents a proportionality constant, and n is equal to 1/2 or 2 for direct and indirect transitions, respectively. Thus, the absorption coefficient can be calculated utilising the following Eqs.59,60. These values illustrate a systematic decrease in the bandgap as NiO thickness increases, starting from 3.98 eV for the 10 nm NiO thickness, followed by a gradual decrease to 3.92 eV for 40 nm NiO thickness. The decrease observed in the bandgap can be attributed to the increase in the defect formation of the NiO thin films, mainly Ni vacancies and oxygen interstitials, as the film thickness increases, which will be further discussed in the next section. Thus, these defects can lead to the introduction of additional energy levels and trap states within the NiO bandgap, thereby influencing the optical properties21,61. The broadening of the absorption edge observed for thicker films qualitatively indicates increased structural disorder and defect states. These findings are closely aligned with previous studies which reported a similar redshift alongside a reduction trend in the bandgap energy as the thickness of NiO thin films increased21,23,61. Electrical properties Hall-effect measurements were employed to investigate the thickness dependence and the resistivity profile of the NiO films deposited on ITO-coated PET substrates, which is illustrated in Fig. 8. The electrical resistivity of the NiO thin films is strongly affected by microstructural defects within the crystallites, mainly nickel vacancies and oxygen interstitials62,63. In addition, film thickness, microstructure of the film, composition, growth conditions, and methods have a significant impact on the electrical performance62,64,65. It is worth mentioning that the nature of the substrate can affect the electrical performance of the NiO film, which would influence the deposition of the film and interfacial quality and carrier transport across the heterojunction63. Variation of resistivity as a function of the NiO thin film thickness deposited on ITO-coated PET substrates, where 0 nm corresponds to the bare ITO-coated PET substrate without NiO deposition. The used bare substrate exhibits a low resistivity of approximately 0.1719 Ω·cm, indicating high electrical conductivity needed for solar cell applications. The resistivity significantly increased after the deposition of a thin 10 nm NiO layer to 0.3806 Ω·cm, suggesting charge traps impeding carrier transport across the NiO/ITO interface due to the suppression of the defects at low thickness. As the NiO thickness increases, the resistivity decreases, reaching 0.1477 Ω·cm, suggesting enhanced film continuity and increased p-type conduction. This behaviour may be associated with the increased formation of acceptor-type defects, primarily Ni vacancies and oxygen interstitials63,64. Simultaneously, defect-related states may contribute to bandgap reduction, as observed in the corresponding optical measurements, which can be correlated with the decrease in resistivity as NiO thickness increases. The defect density in NiO thin films can be estimated by analysing the carrier concentration, based on the assumption that each ionised acceptor-type defect, such as a nickel vacancy (VNi) or an oxygen interstitial (Oi) generates one hole in the lattice66. Under this assumption, the experimentally measured carrier concentration can provide an indirect indication of the concentration of acceptor-related defects. As the film thickness increases, the carrier concentration rises noticeably, from approximately for the 10 nm film to for the 40 nm film. This upward trend suggests that thicker NiO films may contain a higher concentration of ionised acceptor-related defects. The observed increase in carrier concentration may contribute to the reduction in electrical resistivity and the slight narrowing of the optical bandgap with increasing film thickness. A higher concentration of acceptor-related defects may enhance the p-type conductivity through an increase in hole concentration. Additionally, the increased defect population may introduce localised states within the band structure, which may modify the optical transitions and contribute to the bandgap narrowing. Together, these observations suggest that thickness-dependent defect-related effects may influence both the electrical and optical properties of the NiO films. These findings demonstrate the potential suitability of NiO as HTL in optoelectronic applications and agree with previous studies reported in the literature64,67. The combined characterisation results indicate that the thickness of the NiO films strongly influences their structural, morphological, optical, and electrical properties. AFM analysis revealed a reduction in surface roughness with increasing film thickness, suggesting improved film uniformity and densification. This behaviour is consistent with the structural evolution observed in the XRD analysis. The improved film continuity at higher thicknesses also influences the optical and electrical behaviour of the films, contributing to the observed variation in optical bandgap and electrical conductivity. In addition, the PL and XPS analyses suggest the presence of defect-related states that may contribute to the thickness-dependent optical and electrical response of the NiO thin films. Conclusion Nanocrystalline cubic NiO thin films were synthesised on ITO-coated PET substrates with varying thicknesses using MBE for flexible solar cell applications. As the thickness increased to 20 and 30 nm, the NiO thin films became progressively more homogeneous, illustrating signs of nanocrystallinity and improved uniformity of the surface, as evidenced by SEM and AFM findings. Furthermore, the surface roughness was further reduced at 40 nm thickness, reaching an Rq value of 0.26 0.01 nm. The smooth and well-defined surface indicates the adhesive nature of the NiO films to the ITO-coated PET substrate, which facilitates the full coverage of the absorber layer within the device since it assists the transfer of the charge carriers. The XRD and XPS analyses together confirm the formation of NiO thin films with Ni²⁺ as the dominant oxidation state, corresponding to the cubic phase with no additional impurities detected. A remarkable enhancement in the PL peak intensity and sharpness for the 30 nm thickness of NiO was observed, suggesting that a better surface morphology assists in reducing non-radiative recombination losses, thereby increasing the PL emission, which can offer a favourable configuration for achieving optimal optical performance as an efficient HTL in solar cell applications. The decrease in the bandgap from 3.96 eV for 10 nm NiO thickness to 3.91 eV for 40 nm NiO thickness was observed. As the NiO thickness increases, the resistivity decreases, reaching 0.1477 Ω·cm, suggesting enhanced film continuity and increased p-type conduction, indicating the potential of NiO for flexible optoelectronic applications. Therefore, the optimised NiO thin films can be further employed as hole transport layers (HTLs) in flexible solar cell devices to investigate their photovoltaic performance and bending stability for practical flexible optoelectronic applications. Acknowledgements This work was funded by the Researcher Supporting Program at King Saud University, Riyadh, Saudi Arabia, under project number (ORF-2026-328). Author contributions I.A. grew the NiO thin films, conducted the experiments, analysed the data, and wrote the manuscript. A.A. analysed the data and reviewed the manuscript. S.M.A. acquired the data and reviewed the manuscript. S.A. acquired the data and reviewed the manuscript. All authors reviewed and approved the final manuscript. Data availability The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.