Transient Charging Dynamics in Stacked Ga2O3/ZnO Photovoltaic Units Governing Wavelength-Selective Polarity Switching

Transient Charging Dynamics in Stacked Ga2O3/ZnO Photovoltaic Units Governing Wavelength-Selective Polarity Switching | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL Energy & Environmental Materials This is a preprint and has not been peer reviewed. Data may be preliminary. 15 January 2026 V1 Latest version Share on Transient Charging Dynamics in Stacked Ga2O3/ZnO Photovoltaic Units Governing Wavelength-Selective Polarity Switching Authors : Songqi Zhao , Man Zhao 0000-0002-9193-4073 [email protected] , Dayong Jiang , Yuhan Duan , Haoming Wei , Qingcheng Liang , and Rui Deng Authors Info & Affiliations https://doi.org/10.22541/au.176850563.39331966/v1 Published ENERGY & ENVIRONMENTAL MATERIALS Version of record Peer review timeline 220 views 83 downloads Contents Abstract Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Developing self-powered photodetectors capable of processing multi-dimensional optical signals is pivotal for next-generation photonic computing and secure communications. However, conventional device architectures are typically limited to single-mode intensity detection, and the complex photophysics within coupled photovoltaic units remain poorly understood. Here, we reveal a unified “Photovoltaic-Capacitance” coupling mechanism within a vertically stacked, self-powered Ga2O3/PEDOT:PSS and ZnO/Graphene architecture. We demonstrate that the system’s wavelength-selective, transient bipolar response is governed by the multifaceted charging and discharging dynamics between the two coupled units. The key to this mechanism is the ZnO/Graphene (3D/2D) interface, which we define as a novel “Photovoltaic Dynamic-Capacitor” (PDC) component, exhibiting a defined four-stage transient (instantaneous polarization, steady-state saturation, reverse discharge, and relaxation). This architecture enables the Ga2O3 unit (photovoltaic source) to dynamically charge the PDC under 270 nm illumination (+0.27 A/W), while 380 nm illumination directly activates the PDC itself, generating a reverse current (–0.009 A/W). This universal (proven with MgZnO) and dynamically-coupled architecture unlocks a new paradigm for self-powered, multi-dimensional optical processing. We leverage this unique behavior to implement a physical-layer secure communication protocol based on a innovative ternary optical logic (“1”, “0”, “-1”), offering enhanced anti-jamming capabilities rooted in a new photonic degree of freedom. Transient Charging Dynamics in Stacked Ga 2 O 3 /ZnO Photovoltaic Units Governing Wavelength-Selective Polarity Switching Songqi Zhao a , Man Zhao a,* , Dayong Jiang a,* , Yuhan Duan a , Haoming Wei a ，Qingcheng Liang a ，Rui Deng a a School of Materials Science and Engineering, Changchun University of Science and Technology, Changchun 130022, China * Corresponding author. E-mail address : [email protected] (M. Zhao) * Corresponding author. E-mail address : [email protected] (D. Y. Jiang) Abstract Developing self-powered photodetectors capable of processing multi-dimensional optical signals is pivotal for next-generation photonic computing and secure communications. However, conventional device architectures are typically limited to single-mode intensity detection, and the complex photophysics within coupled photovoltaic units remain poorly understood. Here, we reveal a unified “Photovoltaic-Capacitance” coupling mechanism within a vertically stacked, self-powered Ga 2 O 3 /PEDOT:PSS and ZnO/Graphene architecture. We demonstrate that the system’s wavelength-selective, transient bipolar response is governed by the multifaceted charging and discharging dynamics between the two coupled units. The key to this mechanism is the ZnO/Graphene (3D/2D) interface, which we define as a novel “Photovoltaic Dynamic-Capacitor” (PDC) component, exhibiting a defined four-stage transient (instantaneous polarization, steady-state saturation, reverse discharge, and relaxation). This architecture enables the Ga 2 O 3 unit (photovoltaic source) to dynamically charge the PDC under 270 nm illumination (+0.27 A/W), while 380 nm illumination directly activates the PDC itself, generating a reverse current (–0.009 A/W). This universal (proven with MgZnO) and dynamically-coupled architecture unlocks a new paradigm for self-powered, multi-dimensional optical processing. We leverage this unique behavior to implement a physical-layer secure communication protocol based on a innovative ternary optical logic (“1”, “0”, “-1”), offering enhanced anti-jamming capabilities rooted in a new photonic degree of freedom. Keywords: Photovoltaic-Capacitance Coupling, Bipolar Photoresponse, Self-Powered Photodetector, Secure UV Communication, ZnO/Ga 2 O 3 Introduction Next-generation photonic technologies, encompassing intelligent sensing, optical computing, and secure communications, are driving a paradigm shift in optoelectronic devices—moving from conventional signal perception to in-situ information processing, often termed in-sensor computing [1-12] . Central to this shift is the requirement for the device itself to be capable of directly resolving and processing multidimensional optical information—spanning the spectral, temporal, and polarization domains [13, 14] . However, most contemporary photodetector architectures remain “single-modal,” limited to responding to only one dimension: optical intensity. This architectural constraint forces complex processing tasks to be offloaded onto subsequent electronic circuitry, creating inherent latency and power-consumption bottlenecks. Consequently, the development of novel photonic architectures capable of processing multidimensional optical signals directly at the physical device level represents a critical and frontline challenge in the field. Among these dimensions, resolving the spectral (wavelength) domain is essential to applications in multiplexing and target recognition [15, 16] . To this end, bipolar photodetectors, typically realized by vertically stacking dissimilar materials, have been developed. These devices integrate distinct light-absorbing materials within a single pixel, often by stacking different “standard photovoltaic units.” This architecture enables a wavelength-selective capability to output either positive or negative photocurrents [17-19] . In effect, “wavelength” information is directly encoded into more readily processed “electrical symbols” significantly simplifying the complexity of back-end spectral processing. However, the vast majority of research has focused on optimizing the steady-state performance metrics of these devices—such as responsivity, detectivity, and spectral selectivity [20-23] —while the physical origins and functional potential of their transient current dynamics remain largely unexplored. This has created a fundamental disconnect in the field: On one hand, in the pursuit of high-performance bipolar detectors, the photocurrent polarity is often simplified to a static, wavelength-dependent output symbol, while its rich temporal dynamics are, by and large, overlooked [2, 24, 25] . On the other hand, intricate transient optoelectronic phenomena—such as photocurrent spikes and long-term relaxation—are, in fact, widely observed and reported in heterojunction studies, particularly those involving “dynamic capacitive units” [26-29] . Nevertheless, “transient dynamics” and “spectral bipolarity” have consistently been treated as two mutually exclusive research subjects. These transient effects are either dismissed as parasitic “bugs” in conventional photodetectors that must be eliminated, or they are harnessed as “features” for an entirely different application domain: neuromorphic computing [30-32] . Consequently, a core scientific question remains unaddressed: When a “standard photovoltaic unit” is vertically coupled with a “dynamic capacitive unit” possessing sophisticated transient dynamics, what is the constitutive physical mechanism governing the wavelength-selective current polarity reversal? Furthermore, how does this dynamic coupling between “steady-state” photovoltaic action and “transient” capacitive effects dictate the system’s complete spectral-temporal response? Herein, we directly address this central question by constructing a self-powered, vertically stacked architecture comprising a Ga 2 O 3 /PEDOT:PSS “photovoltaic source” and a ZnO/Graphene “dynamic capacitor”. We reveal and experimentally validate, for the first time, a “Photovoltaic-Capacitance” coupling mechanism, demonstrating that the system’s bipolar response originates not from steady-state competition but from the dynamic charging and discharging between the units. Central to this mechanism is our definition of the ZnO/Graphene interface as a novel PDC, for which we describe a four-stage dynamic model: (I) instantaneous polarization, (II) steady-state saturation, (III) reverse discharge, and (IV) relaxation. This architecture enables the Ga 2 O 3 unit (generating +0.27 A/W at 270 nm) to dynamically “charge” the PDC load, whereas 380 nm illumination (–0.009 A/W) directly activates the PDC itself. This coupled architecture, whose generalizability is validated using MgZnO, unlocks a new paradigm for multidimensional optical information processing and is ultimately leveraged to construct a innovative physical-layer secure communication system based on ternary (“1”, “0”, “-1”) optical logic. Results and Discussion To ultimately elucidate the intrinsic physical mechanism governing photocurrent polarity reversal in the stacked device, we must first deconstruct the unique optoelectronic behaviors of the two key functional units. We began by fabricating the bottom heterojunction unit, ITO-PET/ZnO/Graphene. The structural, morphological, and material properties of this unit are presented in Figure 1. A three-dimensional schematic (Figure 1a.) illustrates the layered heterojunction architecture, composed of graphene, zinc oxide (ZnO), and the ITO-PET substrate. Scanning electron microscopy (SEM) images provide direct morphological evidence; the cross-sectional view (Figure. 1b) confirms the respective layer thicknesses, including the thin graphene film, while the plan-view (Figure. 1c) reveals the continuous and uniform surface coverage. UV-visible absorption spectra (Figure. 1d) comparing the bare substrate to the ITO-PET/ZnO composite confirm that the ZnO layer significantly enhances absorption in the ultraviolet region. Furthermore, X-ray diffraction (XRD) data (Figure. 1e) provides crystallographic corroboration, confirming the successful fabrication of a well-crystallized ZnO thin film. Subsequently, the device performance was characterized under zero-bias conditions. The current-time (I-t) characteristic curve (Fig. 1f) immediately reveals a critical anomalous phenomenon: its photoresponse is not a conventional steady-state plateau typical of photovoltaic devices. Instead, it exhibits a distinct, spike-like transient response. Upon illumination (light-on), the photocurrent spikes instantaneously before slowly decaying. When the light is removed (light-off), the current sharply reverses polarity before gradually recovering to the baseline. This unusual transient behavior is indicative of a characteristic optoelectronic conversion mechanism. To investigate the physical origin of this unique transient response, we designed a set of optical path modulation experiments. As illustrated in Figure 2a, three different illumination configurations were employed: PD-1 (front-side illumination, where light directly strikes the Graphene surface), PD-2 (back-side illumination, where UV light must pass through the ITO-PET and ZnO layers before reaching the ZnO/Graphene interface), and PD-3 (filtered front-side illumination, where an additional ITO-PET/ZnO layer was placed above the device as an optical filter). The I-t characteristic curves for these configurations were measured from 280 nm to 380 nm (Figure 2b), with their corresponding responsivities plotted in Figures 2c and 2d. The results clearly indicate that the device’s photoresponse is strongly modulated by the optical transmission path. The front-side configuration (PD-1) yields the highest responsivity, as the incident light directly excites the active layer. Conversely, the responsivity of the back-side configuration (PD-2) is sharply suppressed at wavelengths below 320 nm; this suppression is directly correlated with the UV absorption edge of the ITO-PET substrate, which effectively filters the incident light. As the wavelength increases past 320 nm, the ITO-PET substrate becomes more transparent, allowing the responsivity for PD-2 to rise. Notably, both the PD-2 and PD-3 configurations exhibit a peak response at 380 nm. This phenomenon is attributed to the wavelength-selective absorption of the ZnO layer itself (which acts as a filter in both PD-2 and PD-3): for λ > 370 nm, the absorption coefficient of ZnO decreases, permitting more photons to penetrate the filtering layers and finally reach the active ZnO/Graphene interface to generate photocarriers. Critically, despite their opposing illumination directions (back-side vs. front-side), the responsivity curves and, importantly, the current polarities of PD-2 and PD-3 remain highly consistent. This set of control experiments provides evidence that the optoelectronic behavior is governed not by the direction of illumination, but rather by the effective intensity of the photons that successfully reach the ZnO/Graphene interface. This strongly indicates that the unique transient dynamics originate from a photo-induced charge transfer process occurring at the ZnO/Graphene interface itself. To physically corroborate this interfacial charge transfer process and elucidate its dynamic model, a series of spectroscopic and high-resolution electrical characterizations were performed. Compared to the pristine ZnO film, the steady-state photoluminescence (PL) spectrum of the ZnO/Graphene heterostructure reveals significant emission quenching (Figure 3a). Concurrently, its time-resolved photoluminescence (TRPL) decay curve (Figure 3b) is shortened by several orders of magnitude. This complementary spectroscopic evidence directly confirms that an ultrafast carrier transfer process occurs at the ZnO/Graphene interface upon illumination, effectively suppressing the radiative recombination pathways within the ZnO. Building upon this spectroscopic foundation, high-temporal-resolution I-t measurements (Figure 3c) enable the deconstruction of a single transient response cycle into four distinct phases, as detailed in the smoothed curve (Figure 3d): (I) light-on ascent, (II) light-on decay, (III) light-off reverse ascent, and (IV) light-off relaxation. Based on this collective evidence, we propose a four-stage “photocapacitive” dynamic model, illustrated schematically in Figure 3e. The complete dynamic evolution of this process can be described in four sequential stages. Stage I (Instantaneous Polarization): Upon illumination, a capacitor-like charge separation is established across the ZnO, generating the initial photocurrent spike. A negative capacitive region forms at the ZnO/Graphene interface due to the enrichment of electrons from the external circuit, driven by the hole accumulation (polarization) at the ITO/ZnO interface. Stage II (Steady-State Saturation): As illumination continues, this interfacial polarization reaches a steady state, analogous to a fully charged capacitor. This saturated internal field opposes further charge flow, causing the net current to decay to a lower, stable value. Stage III (Reverse Discharge): When the light is terminated, photogeneration ceases. The stored potential difference from the polarization now acts as the driving force for a compensation current. The electrons enriched at the ZnO/Graphene interface are rapidly driven back through the external circuit to the ITO/ZnO interface, mimicking a capacitor discharge and causing the sharp reverse current spike. Stage IV (Relaxation): Finally, the device exhibits a typical capacitive discharge characteristic. The discharge current gradually diminishes as the polarization potential decays, allowing the system to relax back to its initial dark state. Having defined the characteristics of the PDC, we proceeded to fabricate the second key component of the vertical “photovoltaic unit photocapacitor” coupling architecture: the ITO-PET/Ga 2 O 3 /PEDOT:PSS heterojunction device. The device structure is schematically illustrated in Figure 4a. Morphological evidence is provided by cross-sectional and plan-view SEM images (Figures 4b and 4c, respectively). Transmission electron microscopy (TEM) analysis further confirmed the amorphous nature of the deposited gallium oxide film (Figure 4d). The device exhibits a sharp absorption edge (Figure 4e) and a peak responsivity of 0.48 A/W at 270 nm (Figure 4f), indicating it functions as an efficient, solar-blind photovoltaic unit. In stark contrast to the PDC unit, this device’s I-t characteristic curve (Figure 4g) displays a general, stable photovoltaic response, entirely devoid of the spike-like transient phenomena. Notably, the I-V curve (Figure 4h) exhibits a drift, which is attributed to the built-in electric field originating from the Type-II heterojunction formed between the p-type PEDOT:PSS and the n-type Ga 2 O 3 . An energy band diagram (Figure 4i) illustrates how this built-in field facilitates the spatial separation of photogenerated electron-hole pairs. Electrons are transported through the Ga 2 O 3 conduction band to the ITO electrode, while holes migrate via the HOMO of the PEDOT:PSS to the silver electrode. This asymmetric carrier transport mechanism enables efficient photovoltaic conversion under zero-bias conditions. Consequently, this unit is defined as the standard “photovoltaic source” within our coupled architecture. We then integrated these two functionally dissimilar units—the “photovoltaic source” (Ga 2 O 3 ) and the “photovoltaic dynamic-capacitor” (ZnO)—into a vertical, back-to-back stack via a shared ITO electrode (Figure 5a), thereby realizing the “photovoltaic unit–photocapacitor” coupled architecture. This coupled structure immediately exhibited a singular optoelectronic behavior, governed by the interplay of its two constituent units. Spectral responsivity measurements (Figure 5b) confirmed a dual-peak response, with maxima at 270 nm and 380 nm. More importantly, the I-t characteristic curves (Figures 5c, d) reveal that the system undergoes a dynamic photocurrent polarity reversal, a process that is selectively triggered by the excitation wavelength. Furthermore, the device achieved an extremely low dark current (1.7 × 10 -7 mA, Figure 5e), which is attributed to effective carrier confinement at the heterojunction interfaces. To account for this wavelength-selective polarity reversal, we propose a unified “Photovoltaic-Capacitance” coupling mechanism based on our prior, independent definitions of the two subunits. This model perfectly explains all observed phenomena. The bipolar response originates from the combination of the photovoltaic effect from the Ga 2 O 3 /PEDOT:PSS p-n junction and the distinct photocapacitive nature of the ZnO/Graphene interface. Under 270 nm deep-UV illumination (Figure 6a), the top Ga 2 O 3 layer undergoes band-to-band absorption, generating abundant photocarriers. The built-in electric field of the Ga 2 O 3 /PEDOT:PSS junction separates these carriers, driving electrons toward the shared ITO electrode and holes to the PEDOT:PSS electrode. This process generates a net positive current, which spikes instantaneously. However, as illumination persists, the bottom ZnO/Graphene unit—which is in the dark—acts as the PDC load we previously defined. Electrons flowing from the top unit begin to accumulate at the ZnO/Graphene interface, effectively “charging” this capacitor. This charge accumulation creates an opposing field (an open-circuit-like effect), which suppresses further carrier transport and causes the steady-state current to gradually decay. Upon cessation of the light, this stored charge (the “capacitor”) discharges, producing the observed transient reverse current spike. In the intermediate 330 nm range, the Ga 2 O 3 absorption is weak, while the underlying ITO-PET substrate still blocks most light from reaching the ZnO. This combination results in a negligible net photoresponse (Figures 5c-d). Conversely, under 380 nm near-UV illumination (Figure 6b), the top Ga 2 O 3 layer is transparent. Photons pass through and are absorbed primarily by the bottom ZnO layer. Here, the ZnO/Graphene PDC unit itself is activated. Its local built-in field drives electrons to the ITO electrode—a current flow direction diametrically opposed to that of the 270 nm case. This generates the net negative photocurrent. Crucially, because this response originates from the PDC unit, the I-t curve inherently exhibits the same characteristic transient spike, decay, and light-off reverse spike, consistent with its intrinsic photocapacitive nature. Finally, to demonstrate the generalizability of this “Photovoltaic-Capacitance” coupling architecture, we fabricated an alternative “photovoltaic source” by replacing Ga 2 O 3 with magnesium-doped zinc oxide (MgZnO) (Mg:Zn = 0.2:0.8). This ITO-PET/MgZnO/PEDOT:PSS unit was then integrated with the same ITO-PET/ZnO/Graphene PDC. As shown in Figure 6c, this new composite device exhibits the exact same wavelength-selective polarity reversal. This result confirms that the observed bipolar transient behavior is a robust phenomenon of the architecture and the underlying dynamic coupling mechanism, not an artifact of a specific material choice. This spectral-polarity response, governed by transient dynamics, unlocks new possibilities for optical information processing that transcend conventional binary systems. We leveraged the device’s characteristic wavelength-polarity mapping (254 nm → “1”, 365 nm → “-1”, dark state → “0”) to construct a ternary optical channel. Building on this, we designed a physical-layer encryption protocol, as conceptualized in Figure 7a. As an example, a plaintext character (e.g., ‘0’ in “CUST 2025”) is first converted into a 6-bit raw ternary code [0,1,-1,-1,1,0] via a public mapping rule. Encryption is then performed by applying a pseudo-random private key stream (K = [0,0,-1,1,-1,-1]) via modulo-3 addition, which yields the encrypted signal [0,0,1,0,0,-1]. A comparison immediately shows the encrypted signal is now completely obscured. For an eavesdropper, lacking the private key K, the decryption task is confronted with a vast solution space of 3 6 = 729 combinations, effectively concealing the valid information. Figure 7b illustrates the complete system architecture and data transmission workflow. The process begins at the transmitter, where plaintext characters are converted to ASCII and subsequently mapped to the 6-bit ternary code. The private key stream (K) is then applied via modulo-3 addition to encrypt the raw code. The resulting optical command sequence drives the UV illumination to emit the corresponding optical signals. At the receiver, the bipolar photodetector, leveraging its unique polarity-resolving capability, accurately demodulates the incident optical signal sequence. The receiver system then applies the same private key K (via modulo-3 subtraction) to reverse the encryption, enabling lossless recovery of the original plaintext. This complete signal processing workflow ensures end-to-end communication security at the physical layer. The practical performance and decryption efficacy of this scheme were experimentally validated. As shown in the actual transmitted waveform in Figure 7c, the device’s response to the encrypted optical sequence is both clear and robust, with the positive and negative current pulses corresponding precisely to their logical definitions. Notably, a phenomenon once considered an ”anomaly” is converted into a key functional advantage here: the transient reverse current spike generated at the conclusion of each light pulse (originating from the PDC’s discharge characteristic) provides the receiver with a far more distinct and recognizable timing marker than the simple intensity decay found in traditional systems. This feature significantly enhances the detection fidelity of the “0” (dark) state, thereby reinforcing the system’s timing synchronization and data recovery reliability. By decoding this waveform and applying the correct private key, the receiver successfully recovered the “CUST 2025” message, visually demonstrating the effectiveness and reliability of the entire communication link. This application powerfully demonstrates the immense potential of our revealed physical mechanism for constructing innovative, high-security optical communication paradigms, offering a new design pathway for next-generation secure photonic technologies. In conclusion, the scientific narrative of this work originates from an anomalous transient response observed in the ZnO/Graphene heterojunction (Figure 1f). Through detailed mechanistic investigations (Figures 2–3), we first identified and defined this unit as a novel PDC. This PDC was then vertically integrated with a standard “photovoltaic source” (Ga 2 O 3 /PEDOT:PSS) (Figures 4–5), a coupling that resulted in the key phenomenon of wavelength-selective current polarity reversal. Ultimately, we revealed (Figure 6) that this behavior stems from a previously unrecognized “Photovoltaic-Capacitance” coupling mechanism, which is governed by the transient charging and discharging dynamics between the photovoltaic source and the dynamic capacitive load. This finding provides a successful answer to the core scientific question posed in the introduction. We have demonstrated that current polarity reversal in stacked photovoltaic devices is not governed by steady-state competition, but is instead dominated by transient dynamics. This generalizable architecture, together with its application in ternary secure communication (Figure 7), showcases an innovative paradigm: constructing an information-theoretic security framework directly from the device’s physical properties. This work provides a new design pathway for developing next-generation multidimensional photonic processors. Conclusion In summary, this work has elucidated the intrinsic physical mechanism governing the wavelength-selective, transient bipolar response in stacked, self-powered photovoltaic devices. We have demonstrated that this complex behavior is not a simple steady-state competition but is governed by a unified “Photovoltaic-Capacitance” coupling mechanism. The discovery of this mechanism was enabled by our fundamental definition of the ZnO/Graphene 3D/2D interface as a novel PDC, whose distinct four-stage charging and discharging dynamics are the key to the entire system’s operation. We have experimentally validated this model, showing how the Ga 2 O 3 photovoltaic “source” dynamically charges the PDC “load” (positive current), and how the PDC generates a reverse current upon its own activation (negative current). This “Photovoltaic-Capacitance” coupling model, proven to be a universal principle through MgZnO substitution, fundamentally shifts the design paradigm for multi-junction devices—moving beyond static band alignment to embrace transient, dynamic-controlled photonic logic. The successful implementation of a ternary (“1”, “0”, “-1”) optical logic for physical-layer secure communication is a direct demonstration of this new capability. This work opens a new avenue for designing self-powered, multi-dimensional photonic processors where spectral and temporal information can be processed in parallel at the sensor level. Experimental Section 4.1 Materials Zinc oxide (ZnO), gallium oxide (Ga 2 O 3 ), and magnesium-doped zinc oxide (MgZnO, Mg:Zn = 0.2:0.8) sputtering targets (>99.99% purity) were purchased from Zhongnuo New Materials Technology Co., Indium tin oxide-coated polyethylene terephthalate (ITO-PET) substrates (sheet resistance <40 Ω) were supplied by Huizhou Lanyu Electronic Technology Co., Ltd. Graphene (supported on PMMA/Cu) was sourced from Nanjing Muke Nano Technology Co., Ltd. An aqueous solution of PEDOT:PSS was acquired from Shenzhen Yilai Technology Co., Ltd. 4.2 Device Fabrication First, commercial ITO-PET substrates were sequentially cleaned by sonication in acetone, ethanol, and deionized (DI) water for 15 minutes each, followed by drying. A ZnO thin film was then deposited onto the cleaned substrate via radio frequency (RF) magnetron sputtering (Ar: O 2 = 40:10 sccm, 150 W, 4 Pa, 5400 s). For the graphene component, a commercial PMMA/Graphene/Cu stack was cut into 0.5 cm × 0.5 cm pieces. The underlying copper foil was etched away using a 3% ammonium persulfate solution, allowing the PMMA/Graphene film to float. After several rinses with DI water, the film was transferred onto the surface of the ZnO/ITO-PET substrate. The structure was dried at room temperature, and the PMMA layer was subsequently dissolved in an acetone bath to yield the Graphene/ZnO/ITO-PET structure. In parallel, for the top cell fabrication, a Ga2O3 thin film was deposited onto a separate, newly cleaned ITO-PET substrate via RF magnetron sputtering (Ar = 30 sccm, 120 W, 4 Pa, 3600 s). A MgZnO thin film was also prepared on another clean ITO-PET substrate under different conditions (Ar:O 2 = 40:10 sccm, 130 W, 4 Pa, 7200 s). Following deposition, the PEDOT:PSS solution was spin-coated onto the surfaces of both the Ga2O3 and MgZnO films using a two-step program (500 rpm for 10 s; 3000 rpm for 60 s). Both films were then baked at 90°C for 10 minutes. Finally, the devices were assembled and electrodes were prepared. The Graphene/ZnO/ITO-PET substrate was assembled back-to-back (PET side facing out) with either the PEDOT:PSS/Ga2O3/ITO-PET substrate or the PEDOT:PSS/MgZnO /ITO-PET substrate (also PET side facing out). For electrical contacts, silver paste was applied to the exposed ITO regions on the sides of the assembled stack and connected with silver wires. Concurrently, silver paste was applied to the central functional areas of the two film stacks (i.e., the top PEDOT:PSS and bottom Graphene layers), and silver wires were extended as the primary electrodes. 4.3 Device Characterization The surface and cross-sectional morphologies of the films were characterized using a cold field-emission scanning electron microscope (SEM, JEOL JSM-6701F). Phase analysis was conducted using a Rigaku Ultima VI X-ray diffractometer (XRD, 40 kV/20 mA). Optical absorption characteristics were determined using a Perkin Elmer Lambda 950 UV-Vis spectrophotometer. 4.4 Optoelectronic Measurements Optoelectronic properties were evaluated using a xenon lamp as illumination. Transient photocurrents were measured under ambient conditions. Spectral responsivity was acquired using a customized spectral response system operating at room temperature, also utilizing the xenon lamp. Current-voltage (I-V) and current-time (I-t) characteristics were characterized using a Keithley 2400 SourceMeter. Select I-t curves were additionally verified using a Keysight B1500A Semiconductor Analyzer. 4.5 Optical Communication Test The communication test was conducted in a well-illuminated indoor environment. A handheld UV lamp served as the signal source, and its on/off timing sequence was controlled by an Arduino program. The resulting output current signals were captured by the Keithley 2400 SourceMeter. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work was supported by the Jilin Provincial Scientific and Technological Development Program (Grant No. YDZJ202401562ZYTS). Figure 1. (a) Schematic illustration of the ITO-PET/ZnO/Graphene device. (b) Cross-sectional and (c) plan-view SEM images of the device. (d) Absorption spectra and (e) XRD patterns for the bare ITO-PET substrate and the ITO-PET/ZnO sample. (f) I-t characteristic of the ITO-PET/ZnO/Graphene device. Figure 2. (a) Schematic of the three different illumination configurations (PD-1 to PD-3) for the ITO-PET/ZnO/Graphene device. (b) I-t curves and (c) corresponding responsivity of the device under these three configurations, measured from 280 nm to 380 nm. (d) A magnified view of the responsivity in the 350–425 nm spectral range. Figure 3. (a) Steady-state PL emission spectra and (b) TRPL decay curves for the pristine ZnO film and the ZnO/Graphene heterostructure. (c) I-t response of the device to 365 nm illumination pulsed at 0.5 Hz. (d) A smoothed I-t curve detailing a single response cycle. (e) Schematic of the operating mechanism for the ZnO/Graphene PDC unit, detailing Stage I–II (under illumination) and Stage III–IV (light-off condition). Figure 4. (a) Schematic illustration of the ITO-PET/Ga 2 O 3 /PEDOT:PSS device. (b) Cross-sectional and (c) plan-view SEM images. (d) TEM diffraction pattern of the Ga 2 O 3 film. (e) Absorption spectrum. (f) Spectral responsivity. (g) I-t characteristic curve. (h) I-V characteristic curve. (i) Schematic of the device’s operating principle. Figure 5. (a) Schematic illustration of the vertically stacked (back-to-back) composite device. (b) Spectral responsivity curve. (c) I-t characteristic curves measured from 270 nm to 380 nm. (d) Schematic illustrating the evolution of the I-t response shape across the spectrum. (e) I-V characteristic curve measured in the dark. Figure 6. Mechanism for the generation of positive and negative photocurrents in the composite structure under different UV excitation wavelengths: (a) 270 nm illumination; (b) 380 nm illumination. (c) Spectral responsivity of a composite device fabricated with ITO-PET/MgZnO/PEDOT:PSS and ITO-PET/ZnO/Graphene, demonstrating the generalizability of the mechanism. Figure 7. (a) An example of the ternary cryptographic encoding scheme. (b) Schematic diagram of the complete optical communication system. (c) Input signals, corresponding output signals from the bipolar photodetector, and a schematic of the information decoding process. Reference [1] A. Pospischil, M. Humer, M. M. Furchi, D. Bachmann, R. Guider, T. Fromherz, T. Mueller, Nature Photonics . 2013 , 7 , 892-896.[2] T. Ouyang, X. Zhao, X. Xun, F. Gao, B. Zhao, S. Bi, Q. Li, Q. Liao, Y. Zhang, Advanced Science . 2023 , 10 , 2301585.[3] C. Bao, J. Yang, S. Bai, W. Xu, Z. Yan, Q. Xu, J. Liu, W. Zhang, F. Gao, Advanced Materials . 2018 , 30 , 1803422.[4] J. Zheng, D. Yang, D. Guo, L. Yang, J. Li, D. Ma, ACS Photonics . 2023 , 10 , 1382-1388.[5] F. Teng, K. Hu, W. Ouyang, X. Fang, Advanced Materials . 2018 , 30 , 1706262.[6] L. Yang, D. Yang, D. Guo, J. Zheng, S. Tao, J. Li, G. He, D. Ma, Advanced Functional Materials . 2022 , 32 , 2206993.[7] C. H. Kang, I. Dursun, G. Liu, L. Sinatra, X. Sun, M. Kong, J. Pan, P. Maity, E.-N. Ooi, T. K. Ng, Light: Science & Applications . 2019 , 8 , 94.[8] D. Han, Y. Yang, L. Meng, S. Hu, K. Liu, H. Lin, N. Liu, W. Zhang, J. Ye, Applied Physics Letters . 2023 , 123 .[9] S. Li, Z. Liu, M. Zhang, L. Yang, Y. Guo, W. Tang, Science China Technological Sciences . 2024 , 67 , 608-615.[10] B. Chen, J. Xu, S. Shi, L. Kong, X. Zhang, L. Li, ACS Applied Materials & Interfaces . 2024 , 16 , 28917-28927.[11] Z. Shi, X. Gao, M. Xie, T. Xie, Z. Li, H. Zhao, Y. Wang, X. Gao, ACS Applied Materials & Interfaces . 2025 .[12] T. Sun, X. Ran, M. Cao, J. Liu, Y. Su, L. Duan, D. Liang, Z. Xia, W. Gao, H. Zhao, Environmental Research . 2025 , 121927.[13] L. Bian, F. Cao, H. Zhao, F. Xiang, H. Sun, M. Wang, L. Li, Laser & Photonics Reviews . 2025 , 19 , 2401331.[14] W. Kim, H. Kim, T. J. Yoo, J. Y. Lee, J. Y. Jo, B. H. Lee, A. A. Sasikala, G. Y. Jung, Y. Pak, Nature Communications . 2022 , 13 , 720.[15] S. E. Ng, N. Yantara, N. A. Tu, E. Erdenebileg, P. W. F. Li, D. Sharma, Y. M. Lam, S. Mhaisalkar, A. Basu, A. Chattopadhyay, Advanced Materials . 2024 , 36 , 2406568.[16] D. Wu, Z. Mo, Y. Han, P. Lin, Z. Shi, X. Chen, Y. Tian, X. J. Li, H. Yuan, Y. H. Tsang, ACS Applied Materials & Interfaces . 2021 , 13 , 41791-41801.[17] W. Chen, D. Wang, W. Wang, Y. Kang, X. Liu, S. Fang, L. Li, Y. Luo, K. Liang, Y. Liu, Advanced Materials . 2024 , 36 , 2307779.[18] S. Xu, H. Ni, X. Zhang, C. Han, J. Qian, Inorganic Chemistry . 2023 , 62 , 14757-14763.[19] D. Zhao, Y. Wang, X. Sun, X. Wu, B. Li, S. Zhang, D. Gao, B. Liu, S. Gong, Z. Li, Small . 2024 , 20 , 2309827.[20] X. Liang, S. Wang, S. Zhang, C. Lin, F. Xie, R. Li, F. Zhang, C. Wen, L. Feng, C. Wan, Inorganic Chemistry Frontiers . 2023 , 10 , 1238-1254.[21] J. Ma, F. Zhang, Y. Xing, H. Jiang, Y. Tian, H. Ji, X. Chen, D. Wu, L. Zeng, X. Li, Advanced Science . 2503498.[22] Z. Lan, Y. S. Lau, L. Cai, J. Han, C. W. Suen, F. Zhu, Laser & Photonics Reviews . 2022 , 16 , 2100602.[23] B. Ouyang, H. Zhao, Z. L. Wang, Y. Yang, Nano Energy . 2020 , 68 , 104312.[24] W. Peng, Z. Pan, F. Li, Y. Cai, Y. He, Nano Energy . 2020 , 78 , 105268.[25] Y. Lu, L. Liu, R. Gao, Y. Xiong, P. Sun, Z. Wu, K. Wu, T. Yu, K. Zhang, C. Zhang, Light: Science & Applications . 2024 , 13 , 241.[26] S. Singh, R. Suthar, A. Tomimatsu, P. Rani, M. M. Matsushita, K. Awaga, S. Karak, Advanced Functional Materials . 2025 , 2425426.[27] M. Zhu, Y. Huang, Y. Huang, Z. Pei, Q. Xue, H. Li, H. Geng, C. Zhi, Advanced Functional Materials . 2016 , 26 , 4481-4490.[28] W. Han, K. Liu, J. Yang, X. Chen, Q. Ai, Y. Zhu, Z. Cheng, B. Li, L. Liu, D. Shen, Applied Surface Science . 2023 , 615 , 156371.[29] Y. Bo, M. Yang, Z. Qian, H. Bi, S. Lin, Nano Energy . 2025 , 138 , 110889.[30] D. Wang, X. Liu, Y. Kang, X. Wang, Y. Wu, S. Fang, H. Yu, M. H. Memon, H. Zhang, W. Hu, Nature Electronics . 2021 , 4 , 645-652.[31] H. Xu, Y. Weng, K. Chen, C. Wu, H. Hu, D. Guo, Advanced Optical Materials . 2025 , 13 , 2402238.[32] D. Wang, W. Wu, S. Fang, Y. Kang, X. Wang, W. Hu, H. Yu, H. Zhang, X. Liu, Y. Luo, Light: Science & Applications . 2022 , 11 , 227. Information & Authors Information Version history V1 Version 1 15 January 2026 Peer review timeline Published ENERGY & ENVIRONMENTAL MATERIALS Version of Record 15 Apr 2026 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection Energy & Environmental Materials Keywords sensors thin films two-dimensional materials Authors Affiliations Songqi Zhao Changchun University of Science and Technology View all articles by this author Man Zhao 0000-0002-9193-4073 [email protected] Changchun University of Science and Technology View all articles by this author Dayong Jiang Changchun University of Science and Technology View all articles by this author Yuhan Duan Changchun University of Science and Technology View all articles by this author Haoming Wei Changchun University of Science and Technology View all articles by this author Qingcheng Liang Changchun University of Science and Technology View all articles by this author Rui Deng Changchun University of Science and Technology View all articles by this author Metrics & Citations Metrics Article Usage 220 views 83 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Songqi Zhao, Man Zhao, Dayong Jiang, et al. Transient Charging Dynamics in Stacked Ga2O3/ZnO Photovoltaic Units Governing Wavelength-Selective Polarity Switching. Authorea . 15 January 2026. DOI: https://doi.org/10.22541/au.176850563.39331966/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.176850563.39331966/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe3f781f97b0db4',t:'MTc3OTIwMjc4MA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();