A hybrid pyro-phototronic nanogenerator (HPyNG) for ultra-low light detection

Abstract

We report the fabrication of a hybrid pyro-phototronic nanogenerator (HPyNG) based on a heterojunction formed between two materials having different pyroelectric origins, viz. inorganic zinc oxide (ZnO) and an organic polyaniline–crystalline rubrene thin film. ZnO exhibits pyroelectric behaviour due to its non-centrosymmetric crystal structure, while the crystalline rubrene-containing organic matrix, a centrosymmetric material, demonstrates pyroelectric properties due to the surface layer polarization effect. The interface between these materials forms an efficient junction, which enhances photovoltage generation through synergistic pyroelectric and optoelectronic effects. However, the pyroelectric polarization effects generated in the ZnO film and the crystalline rubrene are found to oppose each other, introducing another unique dimension to this device's functionality. The figures of merit of the nanogenerator are highly competitive with existing technologies, and its ultra-low light detection sensitivity, down to 50 nW, further underscores its unique capabilities. This study transcends the mundane synthetic processes and involves a novel approach to nano-electronic device fabrication using multiple pyroelectric materials with promising applications in low-intensity light sensing.

---

1. Introduction

Nanogenerators, being the new sensation in optoelectronics, emerge as one of the most relevant energy-harnessing photovoltaic devices in this field. Their external power-free operating ability stands as the foremost quality driving their growing demand in optoelectronic fabrication.^1,2 However, with rapid technological advances and tireless efforts from researchers across the globe, staying relevant in energy-harnessing fabrication has been quite a challenging task. The quest for higher-performance rendering devices with better functionalities continues, and pyroelectric as well as piezoelectric nanomaterials are the new focus for the researchers in this regard.^2–4 Inorganic pyroelectric materials have always been the first choice for pyroelectric nanogenerators owing to their unquestionable efficiency and ease of synthesis.^5,6 However, organic pyroelectric materials have recently emerged as potential competitors to their counterparts.^7,8 Finding pyroelectricity through oxidized surface layer polarization without any non-centrosymmetric crystal structure is another recent breakthrough development for organic nanomaterials.^7,9 However, the crucial drawback of an organic pyroelectric device is its limited efficiency. So, deviating from the mundane approach of increasing efficiency by improving material properties through advanced synthetic processes, combining these pyroelectric materials with other compatible and efficient materials can be an alternate yet potentially efficient process to achieve enhanced performance. Hence, an inorganic–organic hybrid pyroelectric photo-active material could be the answer to the quest for a highly efficient pyroelectric nanogenerator with advanced functionalities.^10,11

Among the inorganic pyroelectric materials, the zinc oxide (ZnO) semiconductor is widely used because of its unique optoelectronic properties, such as higher excitonic binding energy (60 meV), excellent electronic properties, and good environment stability.^12–14 Being a wide-band gap semiconductor (band gap ∼3.3 eV), it helps in harvesting UV light, which has very high technological demand in the field of civil, space, and military applications such as pollution monitoring, flame detection, highly secure space communications, sterilization, early missile plume detection and so on.^15,16 Despite having such excellent optoelectronic properties, the major drawback of using ZnO in optoelectronic devices is its inherent defect states, which hamper the speed of the fabricated devices.¹⁷ However, this obstacle is overcome by the special crystal structure of ZnO, that is the c-axis oriented wurtzite phase of ZnO. This special structure produces a pyroelectric effect owing to its non-centrosymmetric crystal structure. When these crystal structures experience light-induced temperature fluctuations, the atoms within the crystal rearrange themselves to produce pyroelectric polarization. This results in a rapid current flow, known as the pyroelectric current. This pyroelectric current not only speeds up the functioning of the device but also enhances its overall photodetection capabilities. So, all these excellent optoelectronic properties place this special kind of ZnO well ahead of other members of the wide band gap transparent oxide family, such as titanium dioxide (TiO₂), nickel oxide (NiO), tin oxide (SnO₂), etc.^18,19

For the fabrication of a pyroelectric nanogenerator using ZnO, the formation of a proper junction with a ZnO-compatible material is an essential requisite. A good barrier junction helps to produce the built-in electric field within the device, which is the core functionality of a nanogenerator. Usually, other inorganic materials compatible with ZnO are the first choice for forming the required junction barrier.^20,21 However, some recent studies reveal the possibilities of using hybrid inorganic–organic materials for this purpose. These hybrid device geometries show excellent photodetection performance and add unique features to the fabricated devices.^22,23 So, in our approach of designing an inorganic–organic hybrid pyroelectric nanogenerator, an organic small molecule, rubrene, is chosen as the compatible organic material. Rubrene is by far the best-known organic small molecule for its excellent electronic properties in photo-detecting devices due to its high charge carrier mobility. However, the mobility and hence the optoelectronic applications of rubrene highly depend on its structures. Amorphous rubrene shows very poor mobility, whereas single crystal rubrene shows very impressive optoelectronic properties with very high mobility of the order of 13–40 cm² V⁻¹ s⁻¹.^7,24 Moreover, among the three polymorphs of rubrene, i.e., orthorhombic, triclinic and monoclinic, the most interesting carrier transport properties are shown by the orthorhombic one, because of the π-stacking of adjacent molecules in the direction of conduction. Apart from this excellent mobility, the most interesting fact about crystalline rubrene is the generation of pyroelectricity through surface layer polarization. A recent report has also demonstrated the successful fabrication of an organic pyro-phototronic nanogenerator using a crystalline rubrene containing thin film as the photo-active material.⁸ Although the synthesis of crystalline rubrene following the conventional routes is a difficult and tedious task, the plasma processing method adopted in this study renders this task quite easy and faster.²⁴

In this work, an effort has been made to integrate two pyroelectric materials of different origins in a single device geometry to unveil the unique carrier generation mechanism. As a result, we ended up fabricating a hybrid pyro-phototronic nanogenerator (HPyNG) comprising inorganic ZnO and organic crystalline rubrene. ZnO shows the pyroelectric effect due to its non-centrosymmetric crystal structure, whereas oxidized surface layer polarization is the origin of pyroelectricity in rubrene. It has been observed that the nanogenerator shows excellent optoelectronic performance by incorporating the best traits of both the materials. The nanogenerator is found to generate a high open circuit voltage of about 0.2 V. Detailed material and photoelectrical characterization of the materials and the devices is carried out and presented in this study. The light-induced electric field profile of the devices are determined using FDTD simulation, demonstrating the superiority of the combined ZnO/rubrene photo-active material for device fabrication. This study thus opens up a new pathway for utilizing multiple pyroelectric materials in single device geometry to fabricate high-performance nanogenerators.

2. Results and discussion

2.1 Material characterization

For any optoelectronic application, the UV-visible spectroscopic study of the active materials is crucial as it gives an idea of the possible optically active regions of the materials when incorporated into any device geometry. The UV-vis absorption study of the materials is carried out with a Shimadzu UV-2600 instrument, and the results are shown in Fig. 1(a). The UV-vis absorption spectrum of ZnO indicated by the red coloured line shows an absorption peak in the UV region owing to its wide band gap. The black coloured line indicates the UV-vis absorption spectrum of rubrene. In this case, the optical properties of rubrene are highly dominated by the surrounding PPA matrix. This combination of materials, i.e., rubrene in PPA, also shows a similar absorption behaviour as that of ZnO. So, this optical synergy makes the combination of rubrene and ZnO very effective for optoelectronic applications as they both excite and generate charge carriers with UV illumination. The rubrene sample shows a very feeble peak in the visible region because of the E_0→0 transition of the M-polarized band.²⁴ This peak is considered to be the signature peak of crystalline rubrene. However, the prominence of this peak is affected by the dominance of the surrounding PPA matrix. The absorbance peak in the UV region is mainly dominated by the plasma polymerized aniline and is attributed to the π → π* transition of the benzene ring.²⁴ This peak generally appears at the lower end of the UV region. However, in this particular case, the peak is red-shifted, creating synergy with the ZnO absorption band due to the interaction of PPA with rubrene. The blue coloured line represents the UV-visible absorbance spectrum of the combined ZnO and rubrene, showing enhanced light absorbance in the UV region due to the concomitant effect of both the materials. The combined effect is also reflected in the device behaviour, enhancing its performance in the UV region.

Fig. 1 (a) UV-vis absorption spectra of the photo-active materials, showing the synergistic optical absorption of both the photo-active materials. (b) Photoluminescence spectra of the active materials; the inset shows the enlarged PL spectra of rubrene and the ZnO/rubrene combination. The quenched PL intensity in the combination of materials indicates efficient charge transfer among them, (c) Time-resolved photoluminescence spectra of ZnO and the ZnO/rubrene combination, showing much faster decay in the combined system. (d) XPS depth profile of the rubrene sample showing the variation of carbon and oxygen components with etching time. This study indicates the presence of a surface oxidized layer, which is responsible for pyroelectricity in rubrene.

To investigate the emission behaviour of both the synthesized materials and the combined photo-active material, photoluminescence (PL) spectra were recorded using a JASCO FP-8300 spectrometer. All samples were excited with a 320 nm light source, and the results are shown in Fig. 1(b). The PL spectrum of the ZnO thin film, represented by the red line, displays a strong emission peak at around 378 nm, attributed to band-edge recombination of photogenerated carriers. Additional emission peaks that appear in the visible range are associated with various defect states in ZnO, such as oxygen vacancies and interstitial defects.²⁵ It is worth noting that the exact positions of these defect states can vary significantly depending on the substrate type, deposition conditions, and the morphology of ZnO nanoparticles.

The green line in the spectrum represents the PL characteristics of rubrene, which are largely dominated by the base PPA matrix. A PL emission peak is observed around 374 nm, attributed to the π–π* transition within the benzenoid units of PPA. Additional peaks in the visible range (400–500 nm) arise from transitions between polaronic and π-band (HOMO) structures of PPA.²⁶ Rubrene's characteristic PL peak, around 525 nm, corresponds to the M-axis polarized band within the tetracene backbone of the rubrene molecules.^24,27 Although rubrene typically exhibits strong photoluminescence, its intensity here is subdued due to its immersion within the PPA matrix, suggesting significant interaction between PPA and rubrene.

In the ZnO/rubrene binary system, represented by the pink line, the PL emission is significantly quenched compared to the ZnO spectrum alone. This quenching indicates enhanced interfacial charge transfer with reduced recombination rates. Additionally, a slight redshift in the UV region implies strong ZnO–rubrene interactions, making this composite material promising for optoelectronic applications. The interaction between ZnO and rubrene not only enhances charge transfer but may also increase device stability. ZnO is known for its stability and robustness, which could mitigate the degradation of rubrene when exposed to light or heat. This synergistic stability can extend the operational life of devices based on this composite material, making it suitable for durable, high-performance optoelectronic devices. These observations highlight the potential of the ZnO/rubrene combination for UV sensing applications.

Following the indications of significant charge transfer in the PL study, Time-Resolved Photoluminescence (TRPL) analysis was conducted to explore carrier dynamics. The TRPL study was carried out with the help of a Horiba Scientific Deltaflex Modular Fluorescence Lifetime System. TRPL enables the estimation of carrier lifetimes, charge transfer times, and charge transfer efficiencies.²⁸ Decay profiles for ZnO (red line) and the ZnO/rubrene binary system (green line) were recorded and are shown in Fig. 1(c). For these measurements, the samples were excited with a 375 nm LED and the PL peak at approximately 397 nm was used to obtain the decay profiles. The ZnO/rubrene system exhibits a notably faster decay rate than ZnO alone, resulting in a shorter carrier lifetime that promotes efficient photoinduced charge transport.

To gain further insight into charge carrier dynamics, the decay profiles were fitted using a tri-exponential decay function, yielding three distinct lifetimes: τ₁, τ₂ and τ₃. The first two lifetimes, τ₁ and τ₂, correspond to free and bound excitons, while τ₃ is associated with defect-state transitions. These lifetimes facilitate the calculation of the average carrier lifetime 〈τ〉, charge transfer time 〈τ_CT〉, and charge transfer efficiency (η), providing valuable insights into the photoelectric performance of the fabricated device.^29,30

The parameters are calculated using the following equations:

Here B₁, B₂ and B₃ are the normalized constants used in the tri-exponential function. The calculated average lifetime for ZnO is 3.32 ns, while in the ZnO/rubrene composite, it reduces to 1.87 ns. This shorter lifetime in the binary system enhances charge transport efficiency between ZnO and rubrene. A charge transfer time of 4.28 ns and a high charge transfer efficiency of approximately 43.69% further support the suitability of this composite material for optoelectronic applications. This high η value suggests a highly effective interface, where nearly half of the generated carriers successfully transfer between ZnO and rubrene. This efficiency value is promising for applications where consistent charge flow is required. For instance, in UV photodetectors, this high efficiency could translate into a stronger, more stable photocurrent, especially under low-light conditions. Table S1 in the ESI section† provides detailed lifetime values and parameter calculations. The structural confirmation of ZnO and rubrene was performed by X-ray diffraction and Raman analysis, respectively, as shown in the ESI part (Fig. S1 and S2).†

While pyroelectricity of ZnO is widely known and attributed to the non-centrosymmetricity of its crystal structure, the pyroelectric nature of the rubrene thin film remains a relatively unexplored phenomenon. Due to its centro-symmetric crystalline structure, Rubrene is not supposed to show any pyroelectric current. Nevertheless, a few recent studies have demonstrated very significant and novel oxidized surface layer polarization-dependent pyroelectric current generation over organic crystalline thin films.⁹ To quote in particular, Gogoi et al. reported the generation of pyroelectricity in crystalline rubrene thin films based on the same concept.⁷ In that report, they observed the deposition of an ultrathin amorphous PPA–rubrene layer over the crystalline rubrene thin film as a result of after-glow plasma inside the chamber. Oxidation of this particular surface layer as soon as it comes in contact with the atmosphere leads to polarization change in that layer which in turn contributes to the generation of pyroelectric current. Furthermore, the presence of this oxide layer is confirmed using the X-ray photoelectron depth profiling analysis which is shown in Fig. 1(d). The figure shows the variation in the atomic percentage of carbon and oxygen with etching of the rubrene surface. It has been found that the atomic percentage of oxygen is high in the top layer, and as we go towards to the bulk of the material, oxygen concentration decreases and the carbon atomic percentage increases. This study confirms the presence of a surface oxidized layer, which is responsible for pyroelectric phenomena in rubrene. The detailed survey spectra are shown in the ESI (Fig. S3).†

In order to carry out the morphological and structural analysis of both the samples, TEM analysis is performed with the help of a JEOL JEM 2100 plus instrument. To obtain the exact morphology, both the materials are directly deposited on a Cu based TEM grid during the deposition process itself. This technique compromises slightly with the quality of the obtained image but can provide the exact morphology of the synthesized materials. Fig. 2(a), showing the TEM image of ZnO, confirms the synthesis of a pin hole free continuous film composed of hexagonal nanoparticles. The FESEM image shown in the ESI section (Fig. S4†) also validates the formation of the smooth thin film. This compact film helps in minimizing the leakage current in the fabricated devices. The HRTEM image of ZnO, as shown in Fig. 2(b), gives a clear picture of lattice planes. This HRTEM analysis provides insight into the synthesis of a highly crystalline material. As this study aims to utilize the pyroelectric nature of ZnO, a strong crystalline behaviour is expected from the material. So, the synthesized highly crystalline ZnO enhances the overall performance of the device as shown in the subsequent sections. The lattice spacing of ZnO is also calculated from the inverse FFT image and found to be 0.26 nm, as shown in Fig. 2(c). This lattice spacing corresponds to the (002) crystal plane of ZnO. The appearance of this crystal plane confirms the synthesis of the c axis-oriented wurtzite phase of ZnO, which has a non-centrosymmetric structure and thus shows pyroelectric behaviour.^12,31

Fig. 2 (a) TEM image of ZnO showing the formation of a pin hole-free film comprising hexagonal ZnO nanoparticles, (b) HRTEM image of ZnO showing the presence of lattice planes, (c) inverse FFT HRTEM of ZnO for the calculation of lattice spacing, (d) TEM image of rubrene showing the formation of rubrene crystallites within the PPA matrix, (e) HRTEM image of rubrene showing the crystal planes, and (f) inverse FFT HRTEM image for the calculation of d-spacing.

The TEM analysis of the as grown rubrene–PPA matrix is also carried out, and Fig. 2(d) shows the morphology of the material. It has been observed that tiny rubrene crystals are grown and embedded in the PPA matrix. The FESEM image shown in the ESI section† further validates the findings. The crystalline rubrene is highly desirable in optoelectronic applications because of its high charge carrier mobility. Moreover, only the crystalline rubrene shows the pyroelectric effect.⁷ So, the formation of rubrene crystals within the PPA matrix fulfils our objective of combining two pyroelectric materials of different origins in a single device geometry. The HRTEM image shown in Fig. 2(e) also confirms the crystalline behaviour of rubrene by showing the crystal planes. The interplanar spacing, calculated from the inverse FFT image (shown in Fig. 2(f)), is found to be 0.33 nm. This lattice spacing corresponds to the (022) crystalline plane of triclinic rubrene which is centrosymmetric in nature.^7,32

2.2 Photoelectrical study

To explore the combined effect of two different pyroelectric materials in a single device geometry, an organic–inorganic hybrid device is fabricated with ZnO as the inorganic pyroelectric material and Rubrene as the organic counterpart. ZnO is widely known as a traditional inorganic pyroelectric material, with non-centrosymmetric crystallinity of ZnO being responsible for its pyroelectric nature. On the other hand, unlike traditional pyroelectric materials, rubrene relies on the surface layer polarization effect for showing pyroelectricity in this particular case. These two different origins of pyroelectricity make every minute detail of the fabricated device interesting as well as important.

As discussed in the experimental section of this manuscript, the fabricated device has the configuration ITO/ZnO/PPA-rubrene/Au (Device-1). The effective area of the device is measured to be 1 cm², where ZnO and PPA–rubrene act as photoactive materials. Both the materials generate pyroelectric current under illumination, albeit from different origins. Moreover, these two photo-active materials exhibit concomitant light absorption predominantly in the UV region, which is why the fabricated device is expected to show substantial efficiency compared to similar devices.

To conduct a detailed comparative study, two other devices are fabricated with ZnO and rubrene separately as photo-active materials, termed Device-2 (ITO/ZnO/Au) and Device-3 (ITO/PPA-rubrene/Au), respectively. Photoelectric analyses of all the three devices will thus eventually highlight the effect of combining ZnO and rubrene in the photo-physics of Device-1.

Generation of photovoltage upon illumination in all the three devices is evident from Fig. 3, showing the I–V characteristics of the devices. Fig. 3(a–c) shows the comparative I–V characteristics of the three devices in dark and illumination mode with a 365 nm light source. While all the three devices show photocurrent generation under illumination, Device-2 and Device-3 show contrasting behaviour in terms of the generated photocurrent and photovoltage. There is a distinct trade-off observed in both the devices: (i) a high photocurrent but lower photovoltage for ZnO-based Device-2 and (ii) a low photocurrent but higher photovoltage for the PPA–rubrene based Device-3. The ZnO based device shows almost negligible photovoltaic characteristics with a very minimal open circuit voltage (V_oc) of 4.5 mV. On the other hand, in the case of the rubrene based device, the photovoltaic effect is found to be quite significant with a V_oc of 48 mV. However, Device-1 with the combination of ZnO and rubrene as the photo-active material exhibits interesting characteristics as shown in Fig. 3(c). Device-1 shows generation of photocurrent in the μA range which is same as that in the ZnO based device while it also shows a significant photovoltage of 0.2 V, which surpasses that in the rubrene based device. Formation of a proper junction barrier between ZnO and rubrene is responsible for this photovoltaic nature, leading to the generation of a high open circuit voltage of 0.2 V. In other words, this hybrid Device-1 serves its purpose by harnessing the best qualities from both photo-active materials. The complete I–V characteristics of the devices are shown in the ESI section (Fig. S5†). Furthermore, to probe deeper into the device photo-physics, photoelectrical characterization of the high performing Device-1 are carried out and analysed. Fig. 1(d) shows the intensity dependent I–V characteristics of Device-1. The graph indicates an almost linear increase in photocurrent with increasing intensity of illumination with 365 nm light.

Fig. 3 I–V characteristics of all the devices, i.e., (a) Device-2, (b) Device-3 and (c) Device-1, showing the generation of photocurrent and photovoltage under illumination. The study shows that the combination of ZnO and rubrene as the photoactive layer, generates much higher photocurrent and photovoltage as compared to their individual counterparts. (d) Intensity dependent I–V characteristics of Device-1, showing the enhancement of photovoltage with increasing light intensity.

The pyroelectric effect is usually observed in a device when the pyroelectric material it contains experiences a rapid temperature fluctuation. A change in the dipole moment occurs momentarily inside the material due to temperature fluctuations which can be visualized as sharp spikes in the output I–t or V–t characteristics of the device.² In this particular study, although the device geometry contains pyroelectric materials, no temperature fluctuation induced by any heat treatment is applied. Rather, the V–t characteristics of the devices are recorded by switching light and the outcome reveals some key details about the photoelectrical characteristics. In order to carry out a comparative study, V–t characteristics of all the three fabricated devices are recorded under fluctuating 365 nm light under zero bias conditions and are presented along with their zoomed in view in Fig. 4(a–f). The first and foremost key point observed from the V–t characteristics is that the devices show pyroelectric behaviour under fluctuating light illumination. This implies that the devices show pyro-phototronic behaviour where light induced rapid temperature fluctuations play a pivotal role instead of any induced heat.³³

Fig. 4 The V–t response and its zoomed-in versions for Device-2 (a and b), Device-3 (c and d), and Device-1 (e and f) reveal the distinct characteristics of pyroelectric currents in the three devices. Additionally, the analysis highlights the trade-off between pyroelectric and photocurrents across all three devices.

Fig. 4(a) shows the V–t characteristic curve of the ZnO based Device-2. ZnO shows a very strong pyroelectric nature because of its highly crystalline non-centrosymmetric nature. As can be seen, in this device the pyroelectric voltage contributes almost 82.50% whereas photovoltage contributes only 17.50%. A single cycle on–off switching in the enlarged version of the graph is shown in Fig. 4(b).

Fig. 4(c) shows the V–t characteristic curve of rubrene based Device-3 whereas Fig. 4(d) shows its enlarged version. The nature of pyroelectric voltage is found to be quite different from that of the ZnO based device. Here, the pyroelectric voltage is found to be dominated in one particular direction. This distinctive nature of the generated pyroelectric voltage originates due to the entirely different mechanism of pyroelectricity happening in rubrene. Despite having a centrosymmetric crystalline structure, rubrene shows pyroelectric nature due to the polarization of the oxidized amorphous surface layer upon light illumination.⁷ The formation of the oxidized surface layer above the PPA-rubrene thin film has been discussed in detail in the earlier section. Although the origin of pyroelectricity and the direction of pyroelectric voltages are different in Device-2 and Device-3, the latter produces an almost similar amount of pyroelectric voltage and photovoltage i.e. 83% pyroelectric voltage and 17% photovoltage. This means both the devices are pyro dominated devices.

However, combining these two pyroelectric materials leads to a completely new pyro-phototronic nature as can be observed from the V–t characteristics of Device-1 presented in Fig. 4(e) and (f). A substantial rise in the generated photovoltage accompanied by a diminished pyro-voltage can be witnessed from the output V–t curve. The enhanced photovoltage is mainly dominated by a strong photovoltaic effect. The device shows 53% photo voltage and 47% pyro voltage. The enhancement of the photo voltage is attributed to the generation of a proper junction barrier between ZnO and rubrene while the decrease in the pyro voltage may be due to the opposite directional polarization in these two materials. The plausible reason for the occurrence of this opposite directional polarization is explained schematically in the following section. A comparison of the pyro-phototronic effect occurring in all the three devices is provided in the following table (Table 1).

Table 1 Relative amount of pyro and photo voltage, generated in the fabricated devices

Device	Photo-voltage	Pyro-voltage
Device-1 (ZnO/PPA–rubrene based)	53%	47%
Device-2 (ZnO based)	17.5%	82.5%
Device-3 (PPA–rubrene based)	17%	83%

---

A schematic explaining the probable pyro-voltage generation in all these three devices is shown in Fig. 5.

Fig. 5 Schematic of the device structures along with the schematic illustration of the generation of pyroelectric polarization within them: (a) Device-2, (b) Device-3 and (c) Device-1. Moreover, the KPFM study of the photo-active layers of all three devices in dark and illumination mode is also presented in order to find the surface potential.

It is evident from the V–t characteristics, shown in Fig. 4, that the light induced polarization effect is responsible for the pyroelectric behaviour of different photo-active layers in this particular case. In the ZnO-based Device-2, the pyroelectric mechanism is straightforward, with polarization generated by the atomic rearrangement of ZnO crystals due to light-induced transient heating.³⁴ In this case, the probable direction of polarization is found to be in the upward direction as shown in Fig. 5(a). On the other hand, in the rubrene-based Device-3, a thin oxidized layer forms on top of the PPA–rubrene layer due to after-glow plasma. Immediately after the RF discharge power is cut, the residual PPA–rubrene precursor in the experimental chamber creates an ultrathin amorphous layer above the rubrene film. This ultrathin layer further gets oxidized on exposure to the atmosphere. This oxidized surface layer is typically negatively charged, inducing positive charges beneath it toward the interface of the crystalline rubrene film. This process results in surface polarization directed downward. This phenomenon is schematically depicted in Fig. 5(b). It is important to mention that, unlike the ZnO based device, here the polarization is confined to the top surface layer only and as a result the pyroelectric behaviour is mainly dominated in one direction as observed in the V–t characteristic curve shown in Fig. 4(b).

When these two materials are combined in a single device, i.e., in Device-1, the opposite directional polarization, as already discussed, diminishes the overall pyro effect and as a result we observe feeble pyro response in the V–t characteristic curve of Device-1 (Fig. 5(c)). However, with the concomitant photo and pyro effect, the device came out as an excellent pyroelectric nanogenerator.

To understand the opposite directional polarization and investigate the photovoltage generation capabilities of the photo-active materials, Kelvin Probe Force Microscopy (KPFM) analysis was carried out under both dark and illuminated conditions. The KPFM measurements were performed using an MFP-3D-bio instrument (Asylum Research of Oxford Instruments), and the results are presented in Fig. 5. This analysis provides a clear depiction of the surface potential of the photo-active materials. It has been observed that the surface potential of the ZnO layer is negative, whereas the rubrene layer exhibits a positive surface potential. This opposite directional surface potential directly correlates with the opposite directional polarization effect, as described in the schematic diagram. So, this study serves as strong evidence for the proposed polarization mechanism of the fabricated devices. On the other hand, KPFM analysis of the photo-active materials was also conducted under UV illumination during the scanning process. This study gives a very clear picture about the photovoltage generation capability of the layers. It has been found that the surface potential of all the samples is enhanced with illumination. However, the enhancement of surface potential in the combined ZnO/rubrene layer is much higher as compared to the individual ZnO and rubrene layers. Additionally, the surface potential of the combined system is more uniformly distributed across the sample. So, this study clearly demonstrates the effectiveness of the ZnO/rubrene combination for the fabrication of nanogenerators.

The I–V and V–t characteristics show that Device-1 shows a much higher photo response as compared to the other devices due to the combined pyro & photo effects of both rubrene and ZnO. To further continue the photoelectrical characterization of the devices, the photoresponsivity of all three devices is measured with the help of a PVE 300 Quantum Efficiency measurement system and compared. The photoresponsivity curve of all three devices is shown in Fig. 6(a). It has been found that all the devices show maximum photoresponsivity in the UV region, which is the prime optical absorption region of both the materials, as observed in the UV-vis absorption study. This study also reveals much higher photo response in Device-1 as compared to other devices because of the synergistic optical excitation in ZnO and rubrene. Fig. 6(b) shows the performance enhancement of Device-1 w. r. t. Device-2 and Device-3. It has been found that the performance of Device-1 is enhanced by a factor of 688 times and 1125 times with respect to Device-2 and Device-3 respectively. So, it can be claimed that the combination of inorganic ZnO and organic rubrene can be an excellent candidate for the fabrication of nanogenerators. The unique but complementary optoelectronic properties, inorganic–organic hybrid geometry and fascinating pyro effect also make this device very interesting to study. The external quantum efficiency is also calculated and found to be maximum in the UV region. The EQE graph is shown in Fig. 6(c). Furthermore, the detectivity (D*) of the fabricated device is calculated as a function of light wavelength and is presented in Fig. 6(c). The detectivity is calculated using the following relation³⁵

where S is the effective surface area, R_λ is the photoresponsivity, q is the electronic charge and I_dark is the dark current of the device. The detectivity is calculated in self-powered mode and the maximum detectivity of the device is found to be around 4 × 10⁹ Jones, which is a very impressive value for self-powered photodetectors. This detectivity value indicates the ability of the device to detect very low intense UV light signals.

Fig. 6 (a) Comparative photoresponsivity study of all the devices, showing the superiority of the device made with combined ZnO and rubrene, (b) performance enhancement factor of Device-1 w. r. t. Device-2 and Device-3, (c) EQE and detectivity spectrum of Device-1, (d) power density of Device-1 as a function of light intensity, showing the excellent performance of the device as a nanogenerator, and (e) energy level diagram of Device-1, showing the carrier generation and transportation mechanism.

Another figure of merit parameter for nanogenerators is the power density. So, the power density of the device is calculated as a function of light intensity for 365 nm light illumination and shown in Fig. 6(d). Considering the volume of the device, the fabricated nanogenerator shows an excellent power density of about 65 mW cm⁻³ with an illumination light intensity of 2.5 mW cm⁻².³⁶ The power density increases linearly with increasing light intensity, reflecting the direct relationship between light intensity and the energy output of the nanogenerator.

To understand the device's position among existing self-powered devices in the literature, a comparative study is also conducted. The comparison table is provided in the ESI (Table S2).† The comparative study highlights that the fabricated device demonstrates performance comparable to existing state-of-the-art devices, making it a competitive candidate for practical applications.

Once the device performance determining parameters are calculated, it is worthwhile to investigate the working mechanism of the device. So, to study the carrier generation and transportation process, an energy level diagram is constructed and shown in Fig. 6(e). From the figure, it is evident that a type-II heterostructure is formed in between ZnO and rubrene, which is responsible for the photovoltaic effect shown by the device. Being wide band gap semiconductors, both ZnO and rubrene are excited by UV light illumination and the electrons are promoted to their respective conduction bands. In this way, charge separations take place. The energy band alignment of these two materials and the electrodes are such that, the excited electrons then transport to the external circuit via the Au electrode. Whereas the holes flow in the opposite direction through the ITO electrode. In this way, current flows in the circuit and the device works in self-powered mode. However, the device does not produce only photo induced charge carriers; rather, the carrier generation from the pyroelectric effect also contributes to the generated current. So, in order to understand the current generation through the combined pyro and photo effects, a schematic diagram is constructed to visualize the four-stage current generation process in the device. The schematic is shown in Fig. S6 of the ESI† part.

In order to support the obtained experimental results, the optical cross-section of the synthesized materials and the light induced electric field distribution of the fabricated devices are studied using commercially available software, Ansys Lumerical, which employed the FDTD (finite-difference time domain) method. The FDTD is a very powerful technique for understanding the optical characteristics of any optoelectronic material by solving the Maxwell's electromagnetic equations. The simulation details are included in the ESI section.† The simulation is executed by maneuvering the refractive index values (n and k) of ZnO and rubrene, extracted from Spectroscopic Ellipsometry (SE) analysis, while the standard dielectric constant values for ITO and Au films are sourced from the Lumerical Database.^37,38 Using the simulation, the optical absorptions of the photoactive materials are calculated and the obtained spectra are shown in Fig. S6† in the ESI section.† The obtained FDTD results show sheer resemblance with the experimental UV-vis absorption spectra as shown in Fig. 1(a) implying enhanced synergistic optoelectronic properties of the ZnO–rubrene combination.

To compare the light-induced electric field (E) distribution of all three fabricated devices, a power field monitor xz is set to observe the cross-sectional E-field distribution. Fig. 7(a)–(c) illustrate the simulated E-field distribution across the xz monitor of the fabricated devices: ITO/ZnO/rubrene/Au film (Device-1), ITO/ZnO/Au film (Device-2), and ITO/rubrene/Au film (Device-3), respectively, under illumination at three different wavelengths: 365 nm, 500 nm, and 600 nm. Under 365 nm light illumination, Device-1 exhibits a stronger induced electric field compared to Device-2 and Device-3. This stronger field enhances the likelihood of generating more electron–hole pairs in Device-1 than in the other two devices.³⁹ Additionally, for Device-1, the induced electric field is distributed across the photo-active area of the device, gradually decreasing towards the ITO and Au films. The maximum E-field intensity, however, is concentrated at the Rubrene–ZnO interface. Noteworthy is that both ITO and Au-films act as an electrical conductor in the device configuration that helps to collect the photogenerated carriers. So, the large area distribution of the electric field across the device active area can be directly correlated with the better performance of Device-1 as compared to Device-2 and Device-3.^40,41 The simulated results further confirm the experimentally obtained KPFM data.

Fig. 7 FDTD simulation: light induced electric field distributions at different illuminating wavelengths of 365 nm, 500 nm, and 600 nm for (a) Device-1 (b) Device-2 and (c) Device-3. The light induced electric field in Device-1 is stronger as compared to other two devices and it is distributed throughout the photoactive area of the device for 365 nm light illumination, leading to higher performance of the device as compared to other devices.

Moreover, when comparing the performance of the Device-1 at different wavelengths, it has been observed that the electric field is homogeneously distributed across the three interfaces (Au film–rubrene, rubrene–ZnO and ZnO–ITO) when illuminated with a 365 nm light source. But, in the case of illumination with sources of higher wavelengths, light gets transmitted through the photo-active material due to its high band-gap nature and gets absorbed by the Au-film as a result of plasmonic light absorption in Au. So, at higher wavelength this simulated E-field is mainly confined at the Au film–rubrene interface only. Moreover, the E-field diminishes significantly towards the ITO side, resulting in a considerable hindrance to the collection of photogenerated carriers at the ITO end. This particular finding thus corroborates the superior performance of Device-1 in the UV region. The data clearly support the spectral response profile obtained in the experiments (Fig. 6(a and c)). A similar electric field distribution is observed in Device-2 and Device-3 also, enabling them to function efficiently exclusively in the UV region. So, this simulation provides a very strong support for the obtained experimental results and helps in understanding the device functioning in detail.

The signature of a good photo detecting device is its ability to detect different intensities of light efficiently, specifically detecting ultra-low intense signals. Moreover, the photocurrent enhancement should follow a linear relationship with increasing intensity. So, an intensity dependent I–t characteristic of Device-1 is carried out and shown in Fig. 8(a). This study shows a linear relationship of photocurrent enhancement as a function light intensity. The device is found to detect a diverse range of intensities, especially its ability to detect intensity as low as 50 nW makes it very special. In order to understand the variation of photocurrent with light intensity, the change in photocurrent as a function of light intensity is plotted on a logarithmic scale and presented in the ESI part in Fig. S8.†

Fig. 8 (a) Intensity dependent I–t characteristics of the fabricated device (Device-1), demonstrating its capability to detect a wide range of light intensities, from ultralow to high. (b) Power law fitting for calculation of the β value, (c) plot of current density vs. light intensity, to calculate the Linear Dynamic Range (LDR) of the fabricated device, (d) the trade of between the pyro current and photocurrent with intensity, and (e) long range I–t response of the device confirming its stability under fluctuating light and the enlarged version of the long range I–t cycle (e1 & e2).

The power law fitting is done (shown in Fig. 8(b)) and the β value is calculated to be 0.98. This high β value implies efficient carrier generation and extraction over the recombination process.⁴² Apart from the β value, the linear dynamic range (LDR) value of the device is also calculated and shown in Fig. 8(c). The LDR value represents the range of illuminating light intensity over which the photoresponsivity of the device remains almost constant. From the figure, it is evident that the current density follows a linear relationship with a wider range of incident light intensity. The LDR value, calculated from the graph, is found to be approximately 107 dB. This value ranks among the best compared to its counterparts, highlighting the superiority of the fabricated device.⁴³

The photoelectrical analysis of the device reveals a trade-off between the photocurrent and pyro current. To gain a deeper understanding of this behaviour, we investigated the variation in these currents with increasing light intensity and the results are shown in Fig. 8(d). At low light intensity, the device exhibits a significantly higher pyro current compared to the photocurrent. However, as the light intensity increases, the photocurrent begins to dominate over the pyro current. During measurements, the device is illuminated from the back side, specifically the ITO side. At low illumination levels, primarily the ZnO layer efficiently absorb the light, and as a result, light-induced pyroelectric current of ZnO becomes dominant. In contrast, as the light intensity rises, it penetrates more deeply, reaching the ZnO/rubrene junction, thereby enhancing the photocurrent. This shift highlights the increased contribution of photogenerated carriers as light intensity grows, transitioning the device response from pyro current dominated to photocurrent dominated behaviour. This analysis provides insights into the interplay between pyro current and photocurrent in the device, which may be beneficial for applications that require tuning of the device's sensitivity under varying illumination conditions. The response time, i.e. the rise time and fall time of the device, is also calculated as a function of light intensity and included in the ESI part (Fig. S9).† This study gives a clear indication about the speed of the fabricated device.

Fig. 8(e) shows a long range I–t response of the device. Sometimes, organic component containing devices face the issue of stability. But the fabricated device demonstrates excellent stability over a long period of time (recorded up to 35 min) without performance degradation, underscoring the superiority of the photo-active layers. The plasma-based synthesis routes implemented here are found to be responsible for this enhanced stability. The photoelectrical analysis clearly implies the unique coexistence of photovoltaic and pyroelectric current in the fabricated hybrid PyNG and the essence of the whole process lies in the non-conventional plasma based synthetic route with optimized deposition parameters.

In order to gain more insights into the charge transport behaviour of the fabricated devices, impedance spectroscopic analysis was carried out for all three devices using a Solartron Impedance Analyzer. The measurements were performed in the frequency range of 1 MHz to 100 mHz with an AC perturbation voltage of 50 mV. The obtained Nyquist spectra of all three devices under dark conditions are shown in Fig. 9(a). To analyze the data, an equivalent circuit diagram was extracted using ZView software, which is shown in the inset of Fig. 9(a). In this circuit diagram, R_s, C_PE, and R_ct represent the contact resistance, interfacial capacitance, and charge transfer resistance, respectively.⁴⁴ The diameter of the semicircle in the Nyquist plot corresponds to the charge transfer resistance. From the figure, it is evident that the charge transfer resistance of the Rubrene/ZnO device (Device-1) is significantly lower compared to the pristine devices (Device-2 and Device-3). A lower charge transfer resistance indicates a higher electron–hole separation efficiency and easier charge transport across the device interfaces. This suggests that the performance of Device-1 should be superior to the pristine devices, as observed in the photoelectrical analysis.

Fig. 9 Impedance spectroscopic analysis of the fabricated devices. (a) Nyquist plot of all the devices in dark mode, showing smallest charge transfer resistance for Device-1. The inset shows the equivalent circuit diagram (b) Nyquist plot of Device-1 in dark and light mode, showing the reduction of charge transfer resistance with illumination.

To further investigate the light-dependent photoelectrical performance of Device-1, impedance spectroscopic analysis was conducted under UV illumination of intensity 1 mW cm⁻². The comparative Nyquist spectra of Device-1 in both dark and illuminated conditions are shown in Fig. 9(b). The results reveal that, upon illumination, the charge transfer resistance of the device decreases even further. This finding strongly supports the enhanced device performance under UV exposure, aligning well with the photoelectrical characteristics discussed in earlier sections. The observed reduction in charge transfer resistance under illumination confirms the improved carrier dynamics, reinforcing the potential of Device-1 for optoelectronic applications.

2.3 Stability and durability test

The performance and reliability of an optoelectronic device strongly depend on its stability and durability under diverse environmental conditions. To assess the robustness of the fabricated device (Device-1), we evaluated its performance under varying conditions of temperature, humidity, illumination frequency, and long-term ageing, as illustrated in Fig. 10.

Fig. 10 Stability and durability test of the fabricated device under variation of (a) temperature and (b) humidity. (c) Performance of the device with enhancing illuminating light frequency, (d) ageing effect of the device. All these tests under various diverse environmental conditions clearly establishes the superiority of the device for practical applications.

Fig. 10(a) presents the device's response to temperature variations. For this analysis, the device was placed on a hot plate, and its temperature was gradually increased while recording the corresponding I–t curves. The results indicate that the device's performance improves with increasing temperature up to approximately 200 °C. This enhancement is attributed to the decrease in resistivity of both the materials, which contributes to superior device efficiency. However, beyond this threshold the device performance reduces due to the degradation of the materials at high temperature.

To further examine the stability of the device under different environmental factors, we investigated its behaviour at varying humidity levels, as shown in Fig. 10(b). For this study, the device was enclosed in a sealed glass chamber, where water vapor was introduced in a controlled manner. A humidity sensor inside the chamber continuously monitored the local humidity levels. The results indicate a minimal decline in device performance with increasing humidity. Even under fully humid conditions, the device exhibits only a 31% reduction in performance, demonstrating its significant stability in humid environments.

Beyond environmental factors, we also analyzed the device's response to varying illumination frequencies to evaluate its ability to function in high-speed optoelectronic applications. As illustrated in Fig. 10(c), the transient response of the device was recorded while systematically increasing the frequency of the incident light. The device maintains its on–off switching behaviour even at higher frequencies, confirming its ability to process high-frequency optical signals. Additionally, the normalized frequency response was analyzed to determine the −3 dB cutoff frequency, with the corresponding graph included in the ESI (Fig. S10).†

Finally, long-term operational stability is a crucial factor, especially for devices incorporating organic materials. To assess its ageing effects, the device was tested over a span of 120 days, with I–t curves recorded at regular intervals, as shown in Fig. 10(d). Remarkably, the device retained excellent performance even after four months, exhibiting only a minor degradation of approximately 28%.

These comprehensive studies affirm the exceptional stability and durability of the fabricated device under diverse environmental and operational conditions. The device demonstrates outstanding resilience against temperature and humidity variations, maintains reliable operation at high frequencies, and exhibits prolonged stability over time. These attributes establish its potential for practical optoelectronic applications requiring long-term reliability.

3. Conclusions

In summary, a hybrid pyro-phototronic nanogenerator (HPyNG) has been fabricated using the heterojunction of inorganic zinc oxide (ZnO) and organic polyaniline–rubrene (PPA–rubrene). The synergistic optical properties of both the materials help in achieving very high photo response in the UV region. The photoluminescence properties of both the materials also complement the higher performance and the TRPL analysis shows very good charge transfer possibilities among these materials. The TEM analysis shows the formation of a compact ZnO thin film comprising hexagonal nanostructures and the TEM analysis of PPA–rubrene shows the formation of rubrene crystallites within the PPA matrix. Both the photo-active materials are pyroelectric in nature, although of different origin. The pyroelectric effect in ZnO occurs due to its non-centrosymmetric crystal structure, whereas surface polarization driven pyroelectric effects are observed in centrosymmetric crystalline rubrene. Not just the origin, the direction of polarization in these materials are also found to be opposite. So, this different origins and opposite directional polarization create an intriguing interplay in the device's photophysics, which makes the study much more interesting. The fabricated device shows very good performance in terms of photovoltage generation. For a comparative study, only ZnO- and only rubrene-based devices are also fabricated and it has been found that, the combination of materials is much more effective than the pristine ones. The junction created at the interface of ZnO–rubrene is found to be responsible for this higher performance. The light-induced electric field profile of the devices were determined using FDTD simulation, revealing that the generated electric field at the ZnO/rubrene interface is significantly stronger and well-distributed throughout the device's active area. This further confirms the superiority of the combined photo-active material for optoelectronic applications. So, the synergistic pyroelectric and photovoltaic effect makes this device an ideal pyro-phototronic nanogenerator. Apart from higher performance, the device shows excellent stability and the capability of detecting ultra-low intense light, down to as low as 50 nW, adds additional features to the device. The origin of the pyroelectric polarization is explained with a schematic diagram. So, this study sets a new paradigm for the fabrication of high-performance pyro-phototronic nanogenerators with this cutting-edge approach of using multiple pyroelectric materials in a hybrid device geometry.

4. Experimental details

The fabricated device, designed particularly for efficient photodetection, has an organic–inorganic hybrid pyroelectric photo-active layer in a vertical geometry where the photoactive materials viz. ZnO and PPA–rubrene are sandwiched in between two electrodes of ITO and Au.

The device fabrication process follows the three consecutive steps: (i) deposition of a ZnO layer on a patterned ITO coated glass substrate by pulsed DC magnetron sputtering process. (ii) Deposition of a rubrene layer over the ITO/ZnO sample by the plasma deposition method. (iii) Deposition of a Au electrode over the ITO/ZnO/rubrene sample.

For ZnO deposition, the base pressure inside a cylindrical plasma chamber is achieved up to 5 × 10⁻⁵ mbar. A mixture of Ar and O₂ gas in the ratio 2 : 1 is used as the sputtering gas for the sputter deposition from the Zn target of 99.99% purity. The deposition is optimized at a pressure of 5 × 10⁻² mbar with an applied DC power of 125 W and the deposition is carried out for 5 minutes. The optimized deposition parameters help in depositing a very compact ZnO thin film which helps in proper charge transportation in the device.

For the next step, the ITO/ZnO substrate is transferred to another plasma chamber for the deposition of the crystalline rubrene thin film by the plasma-based deposition process. The precursor used in this process is a solution of rubrene (C₄₂H₂₈) (Sigma-Aldrich 99.99%) in aniline (C₆H₅NH₂) solvent (Sigma-Aldrich 99.5%) (5 mg of Rubrene per ml of aniline). During deposition, the base pressure is maintained at 2 × 10⁻⁵ mbar inside the chamber and a working pressure of 2 × 10⁻² mbar is attained by injecting the prepared precursor solution in vapour form along with Ar gas into the chamber. In order to transform the liquid precursor into vapour form, the solution is passed through a heated channel and the flow is controlled using a vapour source mass flow controller. The plasma-based deposition process is initiated by applying a RF power of 75 W to the electrode and the deposition is carried out for 45 s. The relatively high RF power inhibits the formation of a smooth plasma polymerized aniline (PPA) background by distorting the polymer chain, while the energy efficient plasma conditions at high RF power activate the rubrene molecules to form crystalline films over it simultaneously. The chemical structure of rubrene is resistant to the applied high RF power due to its high molecular weight. Nevertheless, it is worth mentioning that low applied RF power reportedly produces amorphous rubrene which does not show pyroelectric behaviour.

Once the photoactive layers are deposited step by step, a 70 nm smooth Au film is deposited on top of it by magnetron sputtering to act as a counter electrode. That process completes the device fabrication protocol with the final device configuration of ITO/ZnO/Rubrene/Au film (Device-1).

To conduct a detailed comparative study to decode the impact of pyroelectric origins in device performances, two different devices are also fabricated with ZnO and crystalline rubrene separately as the respective photoactive materials. The devices have the configurations ITO/ZnO/Au film (Device-2) and ITO/rubrene/Au film (Device-3). The device fabrication protocol remains the same as that of Device-1. The photoelectrical characterization of all three devices is performed in a comparative manner.

Data availability

Additional details such as XRD data of ZnO, Raman spectra of rubrene, FESEM images of ZnO and rubrene, table of the kinetic parameters obtained from TRPL analysis, detailed I–V characteristics, schematic of the 4 stage pyro–photo current generation, FDTD simulation details, simulated optical absorption spectra, photocurrent generation as a function of light intensity, plot of rise and fall times as a function of light intensity, normalized frequency response of the device and a performance comparison table are provided in the ESI part.†

Author contributions

Santanu Podder and Jyotisman Bora: conceptualization, methodology, investigation, formal analysis, writing – original draft, writing – reviewing and editing (contributed equally); Khomdram Bijoykumar Singh: FDTD analysis, formal discussion, writing – reviewing and editing; Deepshikha Gogoi: formal discussion, writing – reviewing and editing; Bablu Basumatary: formal discussion; Arup Ratan Pal: conceptualization, methodology, resources, supervisions, validation, writing – review & editing.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

The work reported in this publication was financially supported by the Department of Science and Technology, Government of India, under the Autonomous S&T Institutions grant (Grant No. IASST/R&D/PSD/IHP-12/2023-24/1191-1199). The authors acknowledge the Sophisticated Analytical Instrumentation Centre (SAIC), IASST, Guwahati, for the instrument facilities FESEM, Raman, and HRTEM. The authors acknowledge CSIR-NEIST, Jorhat, for providing the X-Ray photoelectron spectroscopy facility. The authors also acknowledge CSS, IACS Kolkata for the TRPL facility. The authors thank Prof. Sarathi Kundu and Mr Saiyad Akhirul Ali of IASST, for providing the instrumentation facility for impedance spectroscopic analysis. The authors acknowledge the North East Centre for Biological Sciences and Healthcare Engineering (NECBH), IIT Guwahati and the Department of Biotechnology (DBT), Govt. of India with project no. BT/NER/143/SP44675/2023 for the Atomic Force Microscopy (AFM) facility.

References

1. S. Hajra, A. Ali, S. Panda, H. Song, P. M. Rajaitha, D. Dubal, A. Borras, P. In-Na, N. Vittayakorn, V. Vivekananthan, H. J. Kim, S. Divya and T. H. Oh, Adv. Energy Mater., 2024, 14, 2400025.
2. S. Korkmaz and A. Kariper, Nano Energy, 2021, 84, 105888.
3. H. Ryu and S. W. Kim, Small, 2021, 17, 1903469.
4. X. Li, Y. Li, Y. Li, J. Tan, J. Zhang, H. Zhang, J. Liang, T. Li, Y. Liu, H. Jiang and P. Li, ACS Appl. Mater. Interfaces, 2022, 14, 46789–46800.
5. D. Zhang, H. Wu, C. R. Bowen and Y. Yang, Small, 2021, 17, 2103960.
6. R. Mondal, M. A. M. Hasan, J. M. Baik and Y. Yang, Mater. Today, 2023, 66, 273–301.
7. D. Gogoi, A. A. Hussain, S. Biswasi and A. R. Pal, J. Mater. Chem. C, 2020, 8, 6450–6460.
8. B. Basumatary, D. Gogoi, S. Podder, J. Bora, K. B. Singh, S. D. Bora, A. R. Pal and D. S. Patil, Nano Energy, 2023, 114, 108655.
9. E. Meirzadeh, D. V. Christensen, E. Makagon, H. Cohen, I. Rosenhek-Goldian, E. H. Morales, A. Bhowmik, J. M. G. Lastra, A. M. Rappe, D. Ehre, M. Lahav, N. Pryds and I. Lubomirsky, Adv. Mater., 2019, 31, 1904733.
10. Y. M. Di, J. Y. Liu, M. H. Li, S. Q. Zhang, M. H. You and M. J. Lin, Small, 2022, 18, 2201159.
11. Q. Zhang, M. Li, L. Li, D. Geng, W. Chen and W. Hu, Chem. Soc. Rev., 2024, 53, 3096–3133.
12. S. Podder, B. Basumatary, D. Gogoi, J. Bora and A. R. Pal, Appl. Surf. Sci., 2021, 537, 147893.
13. Y. Zhang, H. Su, H. Li, Z. Xie, Y. Zhang and Y. Zhou, Nano Energy, 2021, 85, 105968.
14. P. S. Chauhan, M. Parekh, S. Sahoo, S. Kumar, A. Das Mahapatra, P. Sharma, V. Panwar, A. M. Rao and A. Misra, J. Mater. Chem. A, 2024, 12, 22725–22736.
15. S. M. Rekha, H. V. Neelamana and S. V. Bhat, ACS Appl. Electron. Mater., 2023, 5, 4051–4066.
16. W. Ouyang, J. Chen, Z. Shi and X. Fang, Appl. Phys. Rev., 2021, 8, 031315.
17. D. babu Padmanaban, P. Maguire and D. Mariotti, J. Mater. Chem. A, 2024, 12, 9212–9231.
18. B. Deka Boruah, Nanoscale Adv., 2019, 1, 2059–2085.
19. Z. Wang, R. Yu, X. Wang, W. Wu and Z. L. Wang, Adv. Mater., 2016, 28, 6880–6886.
20. V. Panwar, S. Nandi, M. Majumder and A. Misra, J. Mater. Chem. C, 2022, 10, 12487–12510.
21. Y. Zhang, M. Hu and Z. Wang, Nano Energy, 2020, 71, 104630.
22. J. Han, F. Wang, S. Han, W. Deng, X. Du, H. Yu, J. Gou, Q. J. Wang and J. Wang, Adv. Funct. Mater., 2022, 32, 2205150.
23. J. Wu and M. Gong, J. Appl. Phys., 2021, 130, 070905.
24. A. A. Hussain and A. R. Pal, J. Mater. Chem. C, 2017, 5, 1136–1148.
25. Z. Chen, J. Wang, H. Wu, J. Yang, Y. Wang, J. Zhang, Q. Bao, M. Wang, Z. Ma, W. Tress and Z. Tang, Nat. Commun., 2022, 13, 4387.
26. T. Barman, A. R. Pal and J. Chutia, Appl. Surf. Sci., 2014, 313, 286–292.
27. M. S. Kim, E. H. Cho, D. H. Park, H. Jung, J. Bang and J. Joo, Nanoscale Res. Lett., 2011, 6, 405.
28. T. Kirchartz, J. A. Márquez, M. Stolterfoht and T. Unold, Adv. Energy Mater., 2020, 10, 1904134.
29. S. Podder, J. Bora, S. Thakur, D. Gogoi, B. Basumatary, S. M. Borah, N. C. Adhikary and A. R. Pal, Mater. Chem. Phys., 2022, 275, 125290.
30. N. T. Khoa, S. W. Kim, D. H. Yoo, S. Cho, E. J. Kim and S. H. Hahn, ACS Appl. Mater. Interfaces, 2015, 7, 3524–3531.
31. S. L. Ou, F. P. Yu and D. S. Wuu, Sci. Rep., 2017, 7, 14251.
32. D. H. Park, S. G. Jo, Y. K. Hong, C. Cui, H. Lee, D. J. Ahn, J. Kim and J. Joo, J. Mater. Chem., 2011, 21, 8002–8007.
33. S. Sahare, P. Ghoderao, M. K. Sharma, M. Solovan, R. Aepuru, M. Kumar, Y. Chan, M. Ziółek, S. L. Lee and Z. H. Lin, Nano Energy, 2023, 107, 108172.
34. Z. Wang, R. Yu, C. Pan, Z. Li, J. Yang, F. Yi and Z. L. Wang, Nat. Commun., 2015, 6, 8401.
35. S. Podder and A. R. Pal, J. Appl. Phys., 2019, 126, 083108.
36. Y. Yang, S. Wang, Y. Zhang and Z. L. Wang, Nano Lett., 2012, 12, 6408–6413.
37. T. A. F. König, P. A. Ledin, J. Kerszulis, M. A. Mahmoud, M. A. El-Sayed, J. R. Reynolds and V. V. Tsukruk, ACS Nano, 2014, 8, 6182–6192.
38. P. B. Johnson and R. W. Christy, Phys. Rev. B, 1972, 6, 4370–4379.
39. S. Linic, P. Christopher and D. B. Ingram, Nat. Mater., 2011, 10, 911–921.
40. J. Kim, J. H. Yun, H. Kim, Y. Cho, H. H. Park, M. M. D. Kumar, J. Yi, W. A. Anderson and D. W. Kim, Sci. Rep., 2015, 5, 9256.
41. A. M. Al-Amri, P. H. Fu, K. Y. Lai, H. P. Wang, L. J. Li and J. H. He, Sci. Rep., 2016, 6, 28671.
42. X. Meng, Y. Du, W. Wu, N. B. Joseph, X. Deng, J. Wang, J. Ma, Z. Shi, B. Liu, Y. Ma, F. Yue, N. Zhong, P. H. Xiang, C. Zhang, C. G. Duan, A. Narayan, Z. Sun, J. Chu and X. Yuan, Adv. Sci., 2023, 10, 2300413.
43. W. L. Tsai, C. Y. Chen, Y. T. Wen, L. Yang, Y. L. Cheng and H. W. Lin, Adv. Mater., 2019, 31, 1900231.
44. W. Dong, C. Lu, M. Luo, Y. Liu, T. Han, Y. Ge, X. Xue, Y. Zhou and X. Xu, J. Colloid Interface Sci., 2022, 621, 374–384.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta00063g

‡ These authors contributed equally.

---