Influence of Mg Concentration on Overall Performance of APTES–ZnO/PANI Hybrids Flexible UV Photodetectors

Abstract

Zinc oxide (ZnO) nanoparticles combined with conducting polymers such as polyaniline (PANI) demonstrate promising potential in flexible ultraviolet (UV) photodetection applications. However, the overall performance of undoped ZnO in photodetectors is often limited by high dark current, low responsivity, and detectivity, attributable to the high density of intrinsic defects and recombination rates. This study was aimed at evaluating the influence of magnesium (Mg) concentration ($0.5\le \mathrm{x}\le 3.0\% \mathrm{m}\mathrm{o}\mathrm{l})$ on the structural and optical properties of 3-aminopropyltriethoxysilane (APTES)-modified ZnO/PANI hybrid matrix for ultraviolet (UV) photodetector applications. The novelty of this work lies in the dual strategy of Mg doping and surface modification intended to tailor the optoelectronic properties of ZnO nanoparticles (NPs). X-ray diffraction analysis confirmed the formation of a single-phase wurtzite ZnO. Photoluminescence measurements revealed a significant increase in photoemission intensity with increasing Mg concentration up to a maximum 2.0% mol. Incorporation of Mg remarkably modified the surface morphology and topography of the ZnO/PANI thin film, demonstrating an increase in both surface area and roughness. The Mg-ZnO/PANI photodetector with 1.0% mol of Mg doping concentration demonstrated excellent performance, with responsivity of 2.34 × 10⁻² A/W and detectivity of 1.56 × 10¹⁰ Jones. The effect of Mg doping concentration on the photoemission and photodetection is discussed in detail.

1. Introduction

Ultraviolet (UV) photodetectors (PDs) have been widely studied in recent years due to their wide-ranging applications in environmental monitoring, optical communication, biological research, and defence systems [1]. Among the various materials explored for UV detection, zinc oxide (ZnO) nanoparticles are promising candidate due to excellent optoelectronic properties such as wide bandgap (~3.37 eV) for absorption and emission of light in the UV region, high exciton binding energy (60 meV) at room temperature necessary for efficient photoemission applicable in light emitting diodes, excellent piezoelectric and pyroelectric properties for sensing energy harvesting and optoelectronic device applications, and high electron mobility [2]. ZnO crystallizes in a stable hexagonal wurtzite structure and can be tailored into diverse morphologies, including nanospheres, nanoflowers, nanowires, nanosheets, nanorods, and tetrapods [3,4]. The fundamental properties of ZnO can be tailored and fine-tuned by doping and controlling synthesis parameters such as reactant (precursor/surfactant/dopant) concentrations, temperature, and reaction time, among other things [5].

Additionally, ZnO with diverse particle morphologies can be cost-effectively prepared using a wide variety of synthesis methods, such as solid-state reaction [6], hydrothermal synthesis [7], sol–gel methods [8], co-precipitation [9], and the combustion or microwave combustion methods [10]. The choice of the synthesis method is guided by the anticipated particle morphology, desired properties, and intended applications. The co-precipitation method [7] is frequently used because of its key advantages, such as low processing temperatures, high product purity, and enhanced morphological control. The use of ZnO nanostructures has gained momentum in various key emerging optoelectronic applications, such as photovoltaics, displays, light-emitting diodes, different sensors (e.g., gas, bio, and UV), and photodetectors. In this work, we evaluated the effect of Mg doping on ZnO-based photodetectors, focusing on doping concentration and surface modification. The overall performance of undoped ZnO in photodetectors is often limited by high dark current and slow response times [11,12,13] attributable to the high density of intrinsic defects.

To address these challenges, researchers have turned to doping ZnO with various elements, such as magnesium (Mg) [2] and aluminum (Al) [14,15] because of their ability to tune the bandgap and enhance the material’s electrical and optical properties. Previous studies have shown that Mg²⁺ (with Ne electronic configuration) doping on ZnO does not alter the structure or contribute additional electronic states due to its ionic radius (0.066 nm) being slightly smaller than that of zinc ion (Zn²⁺) (0.074 nm), and when incorporated into the ZnO lattice, it functions as an isovalent substitute of Zn²⁺ [16,17,18]. Moreover, doping with Mg²⁺ has the propensity to influence the electrical properties of ZnO, leading to dramatic improvement in conductivity, often due to its impact on the electron-hole pair interactions [16]. Nurfani et al. [19] fabricated a Mg-doped ZnO photodetector using the spray deposition technique and observed a considerable increase in photocurrent, with UV peak sensitivity of 240.5, responsivity of 73.3 A/W, and detectivity of 1.6 × 10¹¹ Jones. The best rise and decay times of the photodetectors were 0.8 s and 0.7 s, respectively, at 5 V bias and 5% Mg doping concentration. The results of Kılınç et al. [14] and Nurfani et al. [17] demonstrated that Mg doping improved the electrical properties and the overall performance of the ZnO-based photodetectors.

ZnO nanoparticles are usually combined with conducting polymers such as polyaniline (PANI) to fabricate photodetectors. PANI has been found to offer unique electrical conductivity, environmental stability, and ease of processing [20]. These properties often improve the overall performance of photodetectors in comparison with those fabricated solely using ZnO nanoparticles. Combining ZnO with PANI has been found to produce a promising heterojunction that enhances the performance of the photodetector devices [21]. Tababouchet et al. [22] prepared nanocomposites of polyaniline–zinc oxide (PANI-ZnO) via in situ chemical oxidative polymerization, resulting in the formation of a hydrogen bond with the N-H groups of PANI, a phenomenon consistently observed across all composites. This result suggests that ZnO incorporation induces morphological transformations and fosters crystallite growth, with dimensions expanding twofold to 21.8 nm. The measured surface electrical conductivity using the four-point probe method recorded enhancements in conductivity up to 1.65 × 10⁻² S.cm⁻¹.

Zang et al. [21] fabricated a ZnO/PANI/ZnO photodetector device whose UV photocurrents were found to be two orders of magnitude higher than those of the devices prepared using only ZnO. When measured using a UV wavelength of 395 nm, the device displayed excellent responsivity of 0.00968 A/W and photocurrent of ~5.5 × 10⁻⁷ A. These relatively high values were attributed to the large specific surface area of the n-type ZnO nanosheets, and the p-n junction formed between the p-type PANI nano-porous film and the nanosheets. Hadizadeh et al. [23] fabricated a ZnO/PANI heterojunction UV photodetector via the spin-coating technique, producing a 30% increase in responsivity (R_Ph) and rise time (t_R) when illuminated from the PANI side rather than the opposite side. However, the best fall time was obtained when the device was illuminated from the ZnO side. This differentiated performance may be due to PANI acting as a superior electron–hole generator, facilitating faster and stronger photoresponse, while ZnO functioned as a superior electron/hole transporter, suggesting a more rapid return of electron/hole, leading to shorter fall times. Generally, ZnO exhibits strong UV absorption, while PANI possesses a broader absorption range, including the UV region. When both materials are combined, they form a p-n junction that improves UV light absorption, trapping, and charge transport, which makes them preferable for various UV applications [24]. Furthermore, the ZnO-PANI combination produces devices with long-term stability, enhanced photoresponse properties, and simultaneous improvement of t_R, R_Ph, and photocurrent. However, the overall performance of the devices can be further optimized through Mg doping of ZnO. Mg doping offers advantages such as a reduction in the density of intrinsic defects, engineering of the bandgap resulting in better alignment of the energy levels, and enhanced charge transport.

While Mg doping is essential for improved photodetection, optimization of the doping concentration plays a key role in the application of various devices. Lower doping concentration can lead to high charge carrier mobility, while high concentration can lead to higher recombination rates, thereby producing an undesirable outcome. Consequently, this work is set to investigate the effect of Mg concentration on the structural, optical, and electrical properties, as well as the UV photodetection capabilities, of ZnO/PANI hybrid devices. The objective was to determine the best Mg doping concentration to optimize the overall performance of the devices. We varied the doping concentrations ($0.5\le \mathrm{x}\le 3\% \mathrm{m}\mathrm{o}\mathrm{l})$ of the Mg and examined the effect on the crystal structure and optical characteristics. The results demonstrated that Mg doping enhanced the photoluminescence (PL) of ZnO, with the maximum PL intensity produced at the doping concentration of 2.0% mol. The PL emission profiles guided the choice of ZnO:Mg nanoparticle incorporation in the photodetectors, with those with the lowest doping concentration (0.5 and 1.0% mol Mg-doped ZnO) expected to demonstrate minimal radiative recombination and a subsequent improvement in photodetector device performance. Consequently, photodetectors incorporating hybrids of undoped ZnO and 1.0% mol Mg-doped ZnO NPs were fabricated separately, and their electrical properties were systematically evaluated. APTES was incorporated to improve particle surface passivation, dispersion, and film uniformity before PANI electrodeposition. This approach has not been reported in previous studies on Mg-ZnO/PANI photodetectors. Furthermore, nanoparticles with the lowest photoemissions were carefully selected to investigate baseline charge transfer and to better understand the physics of hybrid photodetector devices. This study highlights the benefits of Mg doping for the development of high-performance inorganic–organic hybrid PDs.

Notably, the emission colour of the ZnO doped with different concentrations of Mg was found to shift from the UV region towards the green–yellow–red range, demonstrating tunable luminescence properties that are highly promising for optoelectronic applications. In addition, the surface morphology showed significant modification: the surface area (SA) and root mean square roughness (Sq) increased from 439.39 nm and 600.39 nm (PANI) to 754.29 nm and 1015.42 nm (1.0% Mg–ZnO/PANI), respectively, further confirming the influence of Mg incorporation on surface topography. The UV photodetection results revealed a notable enhancement in the overall performance of the devices. The results of the photodetection experiments are presented and critically discussed.

2. Materials and Methods

2.1. Materials

Zinc acetate dihydrate (Zn(CH₃COO)₂·2H₂O, 99%, Sigma-Aldrich, St. Louis, MO, USA), potassium hydroxide (KOH, ≥85%, Sigma-Aldrich, St. Louis, MO, USA), and absolute ethanol (C₂H₅OH, 99.8%, Sigma-Aldrich, St. Louis, MO, USA) were used as received without further purification. Magnesium acetate (Mg(CH₃COO)₂·4H₂O, 99%, Sigma-Aldrich, St. Louis, MO, USA) was employed as the dopant precursor; 3-Aminopropyltriethoxysilane (APTES, ≥98%, Sigma-Aldrich, St. Louis, MO, USA) was used for surface functionalization of ZnO nanoparticles prior to polymer deposition. Aniline (CH₆NH₅, 99%, Sigma-Aldrich, St. Louis, MO, USA), deionized water (DI), nitric acid (HNO₃, 64%, Sigma-Aldrich, St. Louis, MO, USA), and hydrochloric acid (HCl, 37%, Sigma-Aldrich, St. Louis, MO, USA) were used to prepare the electrolyte for the electrodeposition process.

2.1.1. Synthesis of Undoped and Mg-Doped ZnO

ZnO nanoparticles were synthesized via a controlled precipitation method [25] as shown schematically in Figure 1a. To synthesize undoped ZnO nanoparticles, 3 g of zinc acetate was dissolved in 50 mL of ethanol under vigorous stirring at room temperature until a clear and homogeneous solution was obtained. Separately, 1.5 g of KOH was dissolved in 50 mL of ethanol. Both precursor solutions were then cooled in an ice bath to 0 °C to suppress premature particle growth. The cold KOH solution was then added dropwise into the zinc acetate precursor solution under continuous stirring for 2 h, resulting in the formation of a white colloidal suspension. Subsequently, 1 mL of APTES was added. For the Mg-doped ZnO nanoparticles, appropriate amounts (0.5–3% mol) of magnesium acetate were added to the zinc acetate reaction mixture to incorporate Mg dopants. The resulting suspension was stirred for 2 h to ensure that the reaction was complete, and the solution/mixture was homogeneous. The undoped and Mg-doped ZnO nanoparticles were isolated by centrifugation operated at 4000 rpm for 5 min, followed by three washing cycles with ethanol to remove residual byproducts. The precipitate was then dried overnight at 90 °C in an electric oven before annealing in air at a temperature of 200 °C for 2 h. The dried precipitate was ground into fine powder using a pestle and mortar.

2.1.2. PANI and ZnO/PANI Thin-Film Fabrication

Figure 1b illustrates the schematic representation of the fabrication process and structure of the photodetector device fabricated in this study. The fabrication process was as follows: the active layer solution was prepared by dissolving 0.01 M of aniline in 50 mL of deionized water at 70 °C; 0.5 M nitric acid (HNO₃) was dissolved in 50 mL of hydrochloric acid (HCl) as a supporting electrolyte for ionic conductivity and to provide additional protons. The prepared ZnO nanoparticles were uniformly dispersed in the aniline solution with vigorous, constant magnetic stirring. The resulting solution was then electrodeposited on ITO-coated PET substrate using a standard three-electrode configuration set-up, where the working electrode (WE) was the electrodeposited ITO-coated PET substrate, the counter electrode was platinum wire, and the reference electrode was Ag/AgCl (3 M KCl electrolyte). The deposition was facilitated using cyclic voltammetry, conducted as follows: the potential window was set from −0.350 V to +1.700 V (vs. Ag/AgCl) at a scan rate of 25 mV/s for 5 cycles using a Bio-Logic workstation. This resulted in a uniform electrodeposited PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI nanocomposites on the ITO surface. In the device, two copper (Cu) stripes with dimensions of 5 mm × 30 mm were placed on top of the device, serving as charge collectors, and a 10 V bias was applied during measurements.

2.2. Powder Characterization

The crystal structure of the powders was characterized by X-ray diffraction (Bruker BV 2D diffractometer, Bruker BV, Karlsruhe, Germany) using Cu-Kα (λ = 0.154 nm) radiation, with the 2$\theta$ range set from 5 to 80°. The GSAS II (Python 3.10.8, Version 5478) software was used for Rietveld refinement analysis based on the XRD data to determine the unit cell dimensions, quantify phases, and establish atomic position coordinates. Vesta (ver. 3.5.8) software was used to virtualise the ZnO unit cell. The electronic states and surface chemical composition of the ZnO powders were determined using a PHI Quantes scanning dual X-ray photoelectron microprobe (XPS) system with a 100 µm, 25 W, and 15 kV Al X-ray beam. Surface morphology and elemental composition of the powders were analyzed using a field emission scanning electron microscope (FE-SEM; Zeiss Ultra Plus 55, Tokyo, Japan) with the electron beam accelerated at 2.0 kV and 20.0 kV, respectively. The absorption properties were studied using an Cary 500 UV–Vis–NIR Spectrophotometer, and the spectra were recorded in the range of 200–800 nm. The room-temperature photoluminescence (PL) spectroscopy was conducted using a He–Cd laser (325 nm line, ~50–100 mW) as the excitation source.

2.3. Thin-Film Characterization

The vibrational modes and surface topography of the films were analyzed using the WITec Raman 300RAS+ system equipped with Zeiss objectives (10×, 20×, 50×, and 100×) and atomic force microscopy (AFM), respectively. Raman spectra were collected with a 532 nm excitation laser operated at <0.5 mW to avoid sample heating or damage. A grating of 600 g/mm (BLZ = 500 nm) was used. Raman mapping was performed over a 50 × 50 μm² area with a resolution of 300 × 300 pixels (lines × points) and an integration time of 1 s per point. Optical images were acquired using the 50× Zeiss objective. A high-precision piezoelectric scanning stage was used for both Raman mapping and AFM imaging, enabling accurate positioning and nanoscale resolution across the scanned areas. The AFM tapping mode was used to investigate the nanoscopic surface topography of the films. Measurements were taken from a 50 × 50 μm² scan area, and the surface roughness parameters were estimated from the AFM data. Data processing and Raman image reconstruction were performed using the WITec Project Five software (Build 5.0.3043, Plus version).

2.4. Electrochemical Characterization

The flexible thin-film electrode fabricated via electrodeposition was used as the working electrode (WE) in three electrode configurations. The Ag/AgCl was used as the reference electrode, and glassy carbon as the counter-electrode, immersed in a 3 M KOH solution used as an electrolyte. Electrochemical tests were accomplished with a Bio-Logic VMP-240 1-channel potentiostat (Knoxville, TN, USA) operated at room temperature. The CV curves were analyzed across scan rates ranging from 20 to 1000 mV s⁻¹ within 0–1.1 V. Electrochemical impedance spectroscopy (EIS) measurements of the deposited electrode films were taken by applying an alternating voltage of 10 mV over the frequency range of 100 kHz to 10 mHz.

2.5. Photodetection Characterization

The optoelectronic properties of the photodetectors were investigated using a Keithley 2602B Source Meter connected to a microprobe station. Current–voltage (I–V) measurements were conducted under two distinct conditions: complete darkness and controlled ultraviolet (UV) illumination. The darkness characteristics were recorded when the samples were placed inside a light-shielded chamber, ensuring that there was no incidental stray optical input. For illuminated measurements, a 395 nm UV LED with an output power of 2.10 × 10⁻⁴ W was used as a source of light. The effective illuminated area on the device was approximately 3 cm². The measured currents under illumination (I_light) and in the dark (I_dark) were compared at fixed bias voltages. In addition, time-dependent current (I–t) measurements were performed by periodically modulating the light source, which enabled assessment of the device’s switching capabilities.

3. Results and Discussion

3.1. Structural and Optical Properties

Figure 2a depicts the XRD patterns of the ZnO powder samples. All observed diffraction peaks correspond to the hexagonal wurtzite structure of ZnO (JCPDS card no. 36-1451) without secondary phases (e.g., MgO or other impurities) within the detection limit. This confirms the crystalline nature of both undoped ZnO and Mg-doped ZnO NPs. The patterns also suggest successful incorporation of Mg²⁺ into the ZnO lattice sites without phase segregation [17,26,27]. Figure 2b shows the hexagonal close-packed arrangement of Zn²⁺ and O²⁻ ions, with Zn atoms occupying tetrahedral sites and O²⁻ occupying the hexagonal close-packed lattice framework. Mg²⁺ ion dopants most likely substituted Zn²⁺ in lattice sites due to similar ionic radii and the same ⁺² oxidation state, which allows Mg²⁺ to occupy Zn²⁺ positions without significantly distorting the crystal structure, but slightly altering the lattice parameters [16,17]. The XRD patterns were evaluated using the Rietveld refinement method, and the results are displayed in

Table 1. Sample ID and GSAS refinement structural parameter values for ZnO NPs.

Sample ID	Lattice Parameters (Å)		Crystallite Size	Weighted Residual Factor	Goodness of Fit
Mg(%mol)	a (Å)	c (Å)	D (nm)	Rwp	χ2
0.0	3.25048	5.20949	9.8	8.34	5.02
0.5	3.25074	5.20589	6.3	5.88	3.54
1.0	3.25082	5.20510	8.0	6.47	3.73
1.5	3.25076	5.20884	5.4	9.88	3.76
2.0	3.25027	5.20816	8.7	7.30	3.85
3.0	3.25040	5.20467	14.7	7.52	5.40
Average	3.25058	5.20736	8.48	7.90	4.55
Std. Dev.	0.00023	0.00206	3.21	-	-

Std. Dev. means standard deviation.

. The lattice parameter a shows negligible variation with Mg concentration, while the c parameter exhibits slight fluctuations, pointing to the substitution of Zn²⁺ (ionic radius = 0.74 Å [27,28]) by Mg²⁺ with a relatively smaller ionic radius (ionic radius = 0.72 Å [26]). These subtle changes are consistent with the successful incorporation of Mg into the ZnO wurtzite lattice without causing any significant lattice distortion. The calculated crystallite sizes range from 5.4 to 14.7 nm, indicating that Mg doping influenced the nucleation and growth process of the NPs. Notably, the weighted residual factor (R_wp) values between 5.88 and 9.88% and goodness-of-fit (χ²) values between 3.54 and 5.40 confirm the reliability of the refinement. The Rietveld refined XRD patterns of all the samples were generated using the GSAS (Python 3.10.8, Version 5478) software. However, since the structures were similar, only the undoped and 2.0% mol Mg-doped ZnO NPs are illustrated in Figure 2c,d. The plots display the observed data (y_obs), calculated profile (y_calc), background (y_bkg), difference curve (y_diff), and Bragg reflection positions (tick marks (|)). The small residuals in the y_diff curves confirm the reliability of the refinement, while the well–matched Bragg peak positions demonstrate that both undoped and 2.0% Mg-doped ZnO NPs crystallize in the hexagonal wurtzite structure. A slight peak shift was observed for the 2.0% Mg sample, which is consistent with the substitution of Zn²⁺ by slightly smaller Mg²⁺ ions in the lattice.

The FTIR was used to determine the vibrational/bending frequency modes and the surface chemistry of the ZnO nanoparticles. The FTIR spectra of the undoped and Mg-doped ZnO NPs in Figure 3 show absorption peaks in the range of 400–800 cm^−1, which are consistent with the Zn–O stretching vibrations reported in the literature [29]. The spectra do not exhibit any distinct peaks that could be ascribed to Mg–O vibrations, suggesting that Mg is fully incorporated into the ZnO lattice sites. This observation is consistent with the XRD results shown in Figure 1a. The absorption features in the 1000–1600 cm⁻¹ region can be attributed to O–H bending and stretching vibrations, indicative of surface-adsorbed water or residual precursors [26,27]. The peaks located at 1000–1200 cm⁻¹, 3200–3500 cm⁻¹, and 2800–3000 cm⁻¹ correspond to the vibrational bands of Si–O, N–H, and C–H, respectively, and confirm the successful functionalization of the ZnO nanoparticles by APTES [30,31].

XPS was used to analyze the electronic state and elemental composition of the samples. Figure 4a shows the XPS survey spectra of ZnO, confirming the presence of Zn, O, and Si from the silica compound, which agrees with the FTIR observations. Figure 4b illustrates the C 1s spectrum with peaks at 284.73, 290.75, and 294.04 eV, corresponding to C–C/C–H (sp²/sp³ carbon) from the propyl chain and carbonate groups (O–C=O). These are due to the oxidized carbon species and π–π* satellites of carbonate species [32]. The O 1s peak was deconvoluted into three components with maxima at 528.61, 530.15, and 531.71 eV, as shown in Figure 4c. These are attributed to lattice oxygen in ZnO, confirming the presence of Zn–O bonding [33]; oxygen associated with vacancies or defect-related Zn–OH states [33]; and surface hydroxyl groups (–OH) or chemisorbed oxygen species, which are commonly present due to interactions with moisture or atmospheric oxygen [34]. The Zn 2p core-level spectrum shown in Figure 4d exhibits two distinct peaks at 1022.18 eV (Zn 2p₃/₂) and 1045.27 eV (Zn 2p₁/₂), consistent with Zn²⁺ oxidation states in ZnO [35,36]. Although the high-resolution XPS spectrum of Mg was not measured in this study, it has been reported in the literature [37] that Mg consists of 2p core levels with Mg 2p₃/₂ and Mg 2p₁/₂ doublets located at the binding energies of 49.5 and 50.4 eV, respectively. This close correspondence in electronic structure and ionic characteristics illustrates strong compatibility between Mg and Zn, supporting the effective incorporation of Mg into the Zn lattice.

Figure 5a–c depict the SEM micrographs of undoped, 1.0% Mg²⁺ and (c) 2.0% Mg²⁺—doped ZnO NPs. All the micrographs show distributions of hexagonally shaped nanostructures over a smooth surface. Similar morphologies have been reported previously for ZnO NPs [38,39,40,41]. The EDS spectrum of the 2.0% Mg-doped ZnO NPs is illustrated in Figure 5d, confirming the presence of Zn (39.84 wt%) and O (19.44 wt%) elements. Other elements such as C (35.70 wt%), Mg (1.32%), and Si (4.33 wt%) are associated with the APTES compound, which was used to modify the ZnO NPs. These results agree with the FTIR results in Figure 3, which confirmed the vibrational frequency modes associated with APTES.

Figure 6a illustrates the diffuse reflectance spectra (DRS) of undoped and Mg-doped ZnO NPs. The spectra reveal a distinct UV absorption band edge near 378 nm, attributed to the absorption band of ZnO [35,37]. Upon Mg doping, the 378 nm absorption band alternates between red and blue shifts, in agreement with the band gap values shown in Figure 6b and in

Table 2. The estimated values of the energy band gap and CIE coordinates.

Sample ID	Estimated Bandgap	CIE Coordinates	Correlated Colour Temperature
Mg (%mol)	(eV)	(x;y)	(K)
0.0	3.16	(0.3929:0.3978)	3838
0.5	3.21	(0.4229:0.4213)	2223
1.0	3.10	(0.4229:0.4213)	3383
1.5	3.02	(0.4256:0.4237)	3352
2.0	3.22	(0.4702:0.4246)	2667
3.0	3.11	(0.4671:0.4278)	2731

below. In addition, two absorption peaks were observed at 603 nm and 640 nm, which are attributed to defect-related electronic transitions induced by Mg incorporation into the ZnO lattice. The introduction of Mg caused insignificant lattice distortion and modified the defect structure of ZnO, resulting in the formation of deep-level states within the band gap. These states are mainly associated with oxygen vacancies (V_o), zinc vacancies (V_Zn), as well as Mg-related defect complexes such as Mg_Zn-V_o, which gave rise to absorption in the visible region [42,43,44]. The band gap was estimated from the Tauc plots (Figure 6b), derived from the Kubelka–Munk transformation [45,46,47]. At lower Mg concentration (0.5% mol), a slight blue shift is observed (3.21 eV), which can be attributed to the combined effects of reduced crystallite size (6.3 nm), quantum confinement [48], and the Burstein–Moss effect [49,50] arising from increased carrier concentration. However, at 1.0–1.5 Mg % mol, the band gap decreased from 3.10 to 3.02 eV, suggesting an increase in oxygen vacancies (Vₒ) [51,52,53,54,55], Mg-related defect states [56,57], and lattice strain, thereby introducing localized states within the band gap that enable lower energy optical transitions [58,59,60]. At 2.0% mol Mg doping concentration, the band gap widened to 3.22 eV as crystallite size increased to 8.7 nm, reaffirming the effects mentioned above. In addition, at higher Mg content (3.0% mol), the crystallite size increases significantly (14.7 nm), yet the band gap narrows again (3.11 eV), which can be associated with dopant clustering and enhanced defect density [61,62].

Figure 7a–c show the PL emission spectra of the undoped and Mg-doped ZnO NPs, recorded at an excitation wavelength of 325 nm using an air-cooled He-Cd laser source. The spectra consist of three emission peaks with maxima in UV (~390 nm), green (~547 nm), and red (~760 nm) regions. The PL emission spectra of 0.5% and 2.0% Mg-doped ZnO NPs were fitted using the Lorentzian functions, and distinct emission peaks located at around 390, 547, and 760 nm are shown in Figure 7b,c. The prominent ultraviolet (UV) emission around 390 nm is due to recombination of free excitons [63,64,65] in the ZnO bandgap. The 543 nm visible emission is associated with oxygen vacancies or zinc interstitials, acting as radiative centres within the bandgap [66,67], while the 764 nm deep-level emission is ascribed to oxygen interstitials leading to radiative transitions in the red-to-near-infrared region [68,69]. The lowest PL emission intensity recorded from 1.0% Mg-doped ZnO NPs is indicative of a reduced radiative recombination rate in this specific sample. This characteristic is favourable for optoelectronic applications, as it implies improved charge separation and longer carrier lifetimes. For this reason, the 1.0% Mg-doped ZnO sample was chosen for the fabrication of the thin-film photodetector device.

The Commission Internationale de l’Éclairage (CIE) chromaticity diagram in Figure 7d depicts the emission colour associated with radiative transitions from ZnO NPs doped with different concentrations of Mg. The estimated CIE colour coordinates and corresponding CCT (colour related temperature) values are presented in Table 2. These values were calculated using colour calculator software (Version 7.59). The results show that as the Mg concentration increases, the emission colour shifts from the UV region towards the green–yellow–red region, indicating tunable luminescence suitable for optoelectronic applications. This shift suggests that higher Mg doping concentrations enhanced defect-related emissions, resulting in a broader, redshifted emission [70,71].

3.2. Thin-Film Morphological Characterization

The 2D and 3D AFM images, together with colour height profiles of the fabricated photodetector thin films, are presented in Figure 8a–c. The images exhibit spherical particles and rough surface topography, which is consistent with the micrographs reported by Chilukusha et al. [72]. It can be observed that the height profiles of ZnO–PANI and 1.0% Mg–ZnO–PANI-based thin films exceed that of the PANI-only thin film. This is aligned with rougher surface parameters presented in

Table 3. AFM roughness parameters for the fabricated PDs.

Device	AFM Imaging Roughness Parameters
	SDR%	SA (nm)	SQ (nm)	SSK	SKU	Peak to Peak (nm)
PANI	34.51	439.39	600.39	7.12 × 10−5	4.39	5463.62
ZnO/PANI	58.98	619.14	781.05	−0.22	2.83	4767.55
1.0% Mg-ZnO/PANI	80.47	754.29	1015.42	−0.41	4.48	7669.16

. The AFM parameter reveals that incorporating undoped ZnO and 1.0% Mg–ZnO nanoparticles into PANI increases the surface roughness. These morphological and topological variations have direct implications for both electrochemical and photodetection applications. The surface area and root-mean-square values (RMS), respectively, increased from 439.39 nm and 600.39 nm (PANI) to 754.29 nm and 1015.42 nm (1.0% Mg–ZnO/PANI), indicating significant modification of the surface topography. The surface bearing ratio (SDR%) increased from 34.51% (PANI) to 80.47% (1.0% Mg–ZnO/PANI), showing a denser distribution of asperities on the surface. Skewness (SSK) values shifted from slightly positive in PANI (7.12 × 10⁻⁵) thin film to negative for ZnO/PANI (–0.22) and 1.0% Mg–ZnO/PANI (–0.41) thin films, suggesting a transition from peak-dominated to valley-dominated surfaces. Kurtosis (SKU) remained above 3 in all samples, confirming a leptokurtic distribution of surface features. The peak-to-peak height was highest for 1.0% Mg–ZnO/PANI (7669.16 nm), demonstrating the formation of rougher, irregular structures compared to undoped PANI (5463.62 nm). Improvement in roughness suggests more active sites, which indicates enhancement in responsivity, dark current, and stability for the 1.0% Mg-ZnO/PANI PD [73,74]. Table 3 also shows the skewness and kurtosis (>3) values, and the result indicates that PANI has a uniform and peak-dominated surface (Figure 8a) while ZnO/PANI and 1.0%Mg-ZnO/PANI (Figure 8b,c) surfaces are valley-dominated, which suggests improved charge carrier collection [75].

Raman topography and spectroscopy results are presented in Figure 9 below. PANI topography (Figure 9a) shows homogeneous distribution, consistent with the granular morphology observed in the AFM data (Figure 8a). The ZnO/PANI topography shown in Figure 9b reveals localized ZnO clusters (bright spots at 437 cm⁻¹) embedded in a PANI-dominated matrix. Figure 9c shows the topography of 1.0% Mg-ZnO/PANI, and the results reveal a more uniform distribution of Mg-ZnO features compared to ZnO/PANI, in agreement with reduced agglomeration observed in the AFM data (Figure 8c).

Figure 9d–k shows the Raman mapping and optical images of PANI, (b) ZnO/PANI, (c) 1.0% Mg-ZnO/PANI PDs. The mapping of PANI is shown in Figure 9d–j, and they display relatively uniform colour distribution, reflecting homogeneity in the polymer matrix. Minor intensity variations in the image suggest localized differences in crystallinity or chain alignment for PANI, in agreement with AFM observations in Figure 8a. Raman mapping for ZnO/PANI PD is presented in Figure 9b, and bright spots corresponding to ZnO aggregates were observed, suggesting an inhomogeneous distribution. This observation agrees with the AFM results in Figure 8b, where ZnO introduction increased surface roughness. The Raman image of the 1.0% Mg-ZnO/PANI hybrid device is shown in Figure 9d, and it exhibits enhanced uniformity compared to ZnO/PANI, with fewer bright clusters. The reduced clustering of ZnO (bright spots) indicates that Mg doping promotes better dispersion of ZnO within the PANI matrix.

Figure 9l,m depict the spectrum of PANI and ZnO. The PANI shown in Figure 9l exhibits PANI characteristic vibrational modes. The C=C stretching in the benzenoid (~1480 cm⁻¹) and quinoid (~1580 cm⁻¹) rings confirms the polymer’s emeraldine salt form [76]. The C–N stretching (~1220 cm⁻¹) and C–H bending (~1160 cm⁻¹) are typical polymer backbone [76]. The sharpness of these peaks suggests moderate crystallinity, while the absence of significant broadening indicates minimal structural disorder.

3.3. Electrochemistry

Figure 10a–d show the electrochemical performance of the fabricated PDs in a 3-electrode system immersed in the 3 M KOH electrolyte. The CV results for all the devices at various positive potentials in the 0.0 to 1.1 V range were measured at a single scan rate of 25 mV⁻¹ (Figure 10a), while the CV curves at different scan rates (Figure 10b–d) were scanned at a maximum potential window of 1.1 V. The voltage window was limited to 1.1 V to avoid electrolysis of water adsorbed on the electrode and/or electrolyte under ambient conditions. The PANI and ZnO/PANI CV curves at 25 mV⁻¹ are shown in Figure 10b,c, and they exhibit a constant quasi-rectangular shape, suggesting the electric double-layer capacitive (EDLC) behaviour, which signifies excellent electrical conductivity and fast ionic transport [77]. The CV curves of 1.0% Mg-ZnO/PANI PDs depicted in Figure 10d exhibit a constant quasi-rectangular EDLC shape with redox peaks at a potential window of ~0.11 and ~0.59 V. The presence of minor redox peaks indicates that a reversible redox reaction occurs in the electrode, suggesting minor redox pseudocapacitive contributions. These results suggest a potential for sensing and energy storage applications [78,79].

The comparison of the PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI CV curves is shown in Figure 11a below. The 1.0%MgZnO/PANI device exhibits the highest current density, suggesting enhanced charge transfer and redox activity compared to PANI and ZnO/PANI devices. An increase in current density for 1.0% Mg-ZnO/PANI device is likely due to improved electron mobility and synergistic effects, such as effective charge separation, reduced charge recombination, and faster interfacial electron transfer between 1.0% Mg-ZnO and PANI [80,81]. Electrochemical impedance spectroscopy (EIS) was performed to investigate the effects of Mg concentration on the electrochemical performance of the PDs, and the plots are shown in Figure 11b. The depicted EIS curves of PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI PDs demonstrated the Nyquist plots (imaginary impedance as a function of real impedance). The inset presents the estimated values of the resistance (R) of 0.58, 0.07, and 0.003 Ohms, respectively. These results indicate that the 1.0% Mg-ZnO/PANI PD has the least resistance, in contrast with PANI and ZnO/PANI, which is indicative of improved electron transport [82,83].

3.4. Photodetection

Figure 12a,b depicts the energy diagram and I-V characteristics in the dark and under 395 nm UV illumination. The energy diagram (Figure 12a) shows all components of the PD device, with their respective energy levels in vacuum. When incident photons strike the active interface (ZnO/PANI), they generate electron–hole pairs within the active layer. ZnO, acting as an efficient electron transport layer, facilitates the movement of photo-generated electrons, while PANI enhances hole mobility [81,84]. This process results in photo-generated charge separation, which gives rise to a photocurrent that can be measured with the help of Cu contacts. Light and dark current–voltage (I-V) characteristics of PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI are presented in Figure 12b. The PANI device shows the lowest current values in both dark and light conditions, underscoring the limited charge transport and photogeneration capability of the pure polymer. This aligns with previous work indicating that conducting polymers alone have limited responsivity, limiting their applications in UV photodetectors [85,86,87]. The ZnO/PANI (Figure 12b) device indicates that the incorporation of ZnO significantly improved current, particularly under UV light illumination. Heterojunction formation between PANI and ZnO is known to facilitate effective charge separation and transport [84,88]. Although the ZnO/PANI and 1.0% Mg-ZnO/PANI interfaces can be described as heterojunctions, the absence of rectifying behaviour in Figure 12b can be attributed to the nature of the device architecture and charge transport mechanism. The photodetectors operate in a photoconductive mode with two symmetric Cu electrodes acting as charge collectors, rather than as a conventional diode structure with asymmetric contacts. As a result, the current–voltage characteristics are dominated by bulk and interfacial photoconductive transport rather than junction-controlled rectification transport [84,88]. Upon exposure to UV light, all PD devices demonstrate a remarkable increase in current, confirming that they are sensitive to UV illumination. Notably, the 1.0% Mg-ZnO/PANI PD exhibits the highest photocurrent under both forward and reverse bias, which suggests that Mg doping significantly enhances charge transport and photogeneration efficiency. As reported in similar studies, the incorporation of Mg introduces shallow donor levels that modify the ZnO band structure, resulting in reduced recombination rates and subsequent enhancement of the intrinsic conductivity and interfacial charge transport [81,89].

To further investigate the performance of the PDs, some figures of merit (FOMs) were calculated. These include the photocurrent (I_ph), responsivity (R), detectivity (D), and external quantum efficiencies (EQE), which reflect the response and detection capabilities. The FOMs were determined using the following equations [90,91,92,93,94,95]:

$${\mathit{I}}_{\mathit{p}\mathit{h}}={\mathit{I}}_{\mathit{l}\mathit{i}\mathit{g}\mathit{h}\mathit{t}}-{\mathit{I}}_{\mathit{d}\mathit{a}\mathit{r}\mathit{k}}$$

(1)

$$\mathit{R}={\displaystyle \frac{{\mathit{I}}_{\mathit{p}\mathit{h}}}{{\mathit{P}}_{\mathit{l}\mathit{i}\mathit{g}\mathit{h}\mathit{t}}}}$$

(2)

$${\mathit{D}}^{\mathbf{*}}={\displaystyle \frac{\mathit{R}\sqrt{\mathit{A}}}{\sqrt{2\mathit{e}{\mathit{I}}_{\mathit{d}\mathit{a}\mathit{r}\mathit{k}}}}}$$

(3)

$$\mathit{E}\mathit{Q}\mathit{E}(\mathit{\%})=\mathit{R}{\displaystyle \frac{\mathit{h}\mathit{c}}{\mathit{e}\mathit{\lambda }}}\times 100,$$

(4)

where I_ligh, I_dark, and I_ph are current under the UV light illumination, dark current, and photocurrent, respectively. R is responsivity, and P_light is the incident power of the light source. A is the illuminated area of the device, e is the electrical charge, h is Planck’s constant, and c is the speed of light in a vacuum. The calculated values are presented in

Table 4. Electrical performance parameters for the fabricated PDs.

	Electrical Performance
Device ID	Photocurrent (µA)	Responsivity (A/W)	Detectivity (Jones)	EQE (%)
PANI	1.53	7.26 × 10−3	2.42 × 109	2.26
ZnO/PANI	3.98	1.90 × 10−2	1.46 × 1010	5.65
1.0% Mg-ZnO/PANI	4.92	2.34 × 10−2	1.56 × 1010	6.95

below. The 1.0% Mg-ZnO/PANI PD exhibits high values of R (2.34 × 10⁻² A/W), $D^{\mathbf{*}}$ (1.56 × 10¹⁰ Jones), and EQE (6.95%) compared to PANI (R = 7.26 × 10⁻² A/W, $D^{\mathbf{*}}=$ 1.56 × 10¹⁰ Jones), EQE = 2.26%) and ZnO/PANI (R = 1.90 × 10⁻² A/W, 1.46 × 10¹⁰ Jones, EQE = 5.65%), indicating improved PD performance.

A comparison of the parameters of the PDs is presented in

Table 5. The comparison of results between previously reported ZnO- or PANI-based hybrid PDs and our devices in this work.

Device	Photocurrent (µA)	Responsivity (A/W)	Detectivity (Jones)	Ref.
MgZnO/PANI	---	0.16 at 0 V	1.5 × 1011	[94]
PANI/ZnO nanowires	---	0.56 at 0 V	---	[74]
ZnO/EP/TiO2	22	4.46 at 0 V	63.2 × 1012	[96]
ZnO/PANI wires	---	0.06 at—1 V	2.0 × 1011	[97]
ZnO/PEDOS nanorods	---	247.7 at 2V	3.50 × 1012	[98]
1.0% Mg-ZnO/PANI	4.92	2.34 × 10−2 at 0 V	1.56 × 1010	This work

EP: polyaniline emeraldine salt; PEDOS: poly(3,4-ethylenedioxyselenophene).

. The 1.0% Mg-ZnO/PANI PDs exhibit a responsivity of 2.34 × 10⁻² A/W and detectivity of 1.56 × 10¹⁰ Jones, which are comparable to those reported in a similar study by Chen et al. [94], confirming that the incorporation of Mg ions improved the quality of p-n heterojunctions, effectively separating and transmitting photoelectron–hole pairs, thereby enhancing the performance of the PDs. This result provides a practical basis for optimizing the performance of the flexible hybrid PDs.

The photoresponses of PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI PDs under UV light irradiation under 395 nm at 30 s intervals are presented in Figure 13a,b. Figure 13a shows that the 1.0% Mg-ZnO/PANI PD exhibits the highest current response (4.3 mA) compared to ZnO/PANI (2.2 mA) and PANI (1.2 mA) PDs, indicating a significant enhancement in charge carrier generation and efficient separation of photo-generated electron-hole pairs. This enhancement is ascribed to the Mg doping in ZnO, which introduces shallow donor states and reduces recombination rates [81]. Figure 13a also shows that incorporating ZnO into PANI enhanced the current response. This improvement is attributable to the relatively high electron mobility of ZnO, which facilitates charge transfer. All three devices exhibit a rapid rise and decay in current, demonstrating a fast response when switched ON/OFF. The current remains consistent over multiple cycles, confirming good stability and reproducibility of the hybrid materials. The rise (t_rise) and fall (t_fall) times are critical parameters for evaluating the response time of a PD. The t_rise and t_fall are defined as the time taken by the device to reach 90% of its maximum value of current with illumination, and 10% of its value in the absence of illumination, respectively [99]. Figure 13b illustrates the enhanced time-dependent current response from 30 to 60 s for all devices. The measured t_rise and t_fall response of PANI, ZnO/PANI, and 1.0% Mg-ZnO/PANI PDs are (t_rise = 161 ms and t_fall = 80 ms), (t_rise = 168 ms and t_fall = 88 ms), and (t_rise = 329 ms and t_fall = 190 ms), respectively.

4. Conclusions

In conclusion, modified x% Mg:ZnO (x = 0.0–3.0% mol) NPs, ZnO/PANI, and 1.0% Mg-ZnO/PANI hybrid photodetectors were successfully synthesized and fabricated using simple co-precipitation and electrodeposition techniques, respectively. The XRD analysis confirmed crystallization of single-phase hexagonal wurtzite structures of ZnO NPs. This was further supported by the XPS and FTIR results. The SEM results revealed a smooth surface with some degree of agglomeration, while EDS confirmed the presence of all the elements. Photoluminescence and the CIE diagram, respectively, showed that the 1.0% Mg-ZnO sample exhibited the lowest photoemission intensity, and a strong green–yellow–red colour. Furthermore, the 1.0% Mg-ZnO/PANI PD device demonstrated superior performance under UV illumination. The cycling stability, photocurrent, and photoresponse indicate that 1.0% Mg-ZnO/PANI PD device has potential and should be considered for application in a flexible hybrid photodetector.