Abstract

Despite several attempts to enhance the electrical and charge carrier transport characteristics of porous silicon (PSi), the requisite conditions for optimally synthesizing n-PSi with appealing optoelectronic properties are yet to be achieved. Therefore, this research explores the effect of the chemical ratio of precursor materials (HF:C2H6O:H2O2) on the surface morphology, crystalline structure, and optical and electric properties of PSi. The PSi was produced by photoelectrochemical etching followed by anodization of the n-type Si under light illumination. The properties of the as-prepared PSi were studied by means of microscopic and spectroscopic techniques. The HF:C2H6O:H2O2 chemical ratio was optimized at 2 : 1 : 1. A metal–semiconductor–metal (MSM) ultraviolet photodetector (Pt/n-PSi/Pt) was fabricated, which exhibited high performances under UV light (365 nm) illumination. The photodetector was shown to be highly stable and reliable with a rapid rise time of 0.56 s at a bias voltage of +5 V. The MSM photodetector displayed responsivity (Rp) of 9.17 A/m at 365 nm, which significantly exceeds the values reported for TiC/porous Si/Si in some contemporary research. The photodetector fabricated from n-PSi, synthesized at an optimum chemical ratio (2 : 1 : 1) exhibited the best photodetection performance, possibly due to the high porosity and defect-free state of the n-PSi thin films.

1. Introduction

Porous silicon (PSi) has shown great potential, particularly in the manufacture of light-emitting diodes (LEDs), on account of its visible light emission at room temperature. The very large and reactive internal surface of the PSi layer makes it characteristically different from its bulk counterpart. Thus, the internal surface of the PSi layer plays a vital role in determining its intrinsic properties [1]. The simple and cost-effective fabrication associated with its good optical properties further increases its potential application in integrated optics and photonics [2]. As a result, several attempts have been made to enhance the electrical and charge carrier mobility attributes of PSi [3]. Nonetheless, the optimal settings for synthesizing n-PSi with the sought-after optoelectronic properties are yet to be attained. For instance, n-PSi is presently being investigated as a precursor material for the manufacture of ultraviolet photodetectors (UVPDs) because of its high visibility and UV photoluminescence under ambient conditions [4]. Given such demand, this study is carried out to optimize the electrical and optical attributes of n-PSi as input material for the manufacture of metal–semiconductor–metal (MSM) UVPDs.

Specifically, this paper examines the effect of the chemical ratio of precursor materials (HF:C2H6O:H2O2) on the structural and optoelectronic properties of PSi. The duration of photoelectrochemical etching, applied voltage or current, and illumination intensity on the properties of the synthesized n-PSi were also explored. In addition, the photoresponse of MSM UVPDs fabricated from different PSi layers (synthesized at different chemical ratios) in the dark and under illumination was also studied.

2. Experimental Details

Before the photoelectrochemical etching for the growth of n-PSi, single crystalline wafers of n-type Si (111) with a thickness range of 205–306 µm were cut into squares of 1.5 cm2, which were then cleaned using the Radio Corporation of America (RCA) technique. An anodization cell was fabricated from Teflon. The Si wafer was affixed and sealed through an O-ring, while its front side was subjected to electrolyte mix solution of hydrogen fluoride (HF) solution (48%), ethanol (C2H6O, 99.9%), and hydrogen peroxide (H2O2).

2.1. Preparation of Porous Si (PSi)

The first step involves the immersion of the Si wafer into the electrolyte mix solution of HF, ethanol, and H2O2. Pure H2O2 was included to eliminate the H2 gas bubbles at the sample surface that were produced in the course of anodization [5]. The Si wafer and a platinum wire were utilized as the anode and cathode, respectively. The PSi layer was synthesized inside the HF solution (HF:C2H6O:H2O2 = 2 : 1 : 1) using the DC process at a fixed current density of 45 mA/cm2 and an etching duration of 30 min under external incandescent light. The HF, ethanol (C2H6O), and H2O2 were mixed at different ratios (1 : 1 : 0, 1 : 3 : 0, and 2 : 1 : 1). The as-prepared PSi layers were then utilized to fabricate MSM-UVPD. To ensure the formation of a Schottky contact (resulting from its higher work function), a ≈ 150 nm thick platinum (Pt) layer was placed over the PSi surface. The MSM-UVPD structure comprises two interdigitated Schottky contacts (electrodes with five fingers) with a length of ≈3.3 mm, a width of 30 µm, and a spacing of 400 µm, as shown in Figure 1. The electron beam evaporation technique was employed to deposit the metal finger mask by means of the ATO 306 vacuum coater.

2.2. Morphological and Optical Characterizations

The structure of the as-prepared PSi was determined via X-ray diffraction using the PANalytical X’PertPro Diffractometer equipped with Cu–K radiation (wavelength 1.54056 Å) and operated at 40 kV and 30 mA. The surface morphology of the as-obtained PSi layer was analyzed using a high-resolution field emission scanning electron microscope (FESEM NOVA NANO SEM 450). Photoluminescence (PL) spectra of the samples were recorded using Jobin Yvon HR 800 UV spectrophotometer equipped with a He–Cd laser with a wavelength, , of 325 nm at 20 mW. Current–voltage (IV) and current–time (It) plots of the fabricated MSM-UVPD were recorded using the Keithley electrometer (model 6517 A).

3. Results and Discussion

3.1. Morphological Observations

The porosity (P) of the synthesized PSi layer was calculated using Equation (1).where M1 and M2 represent the weight of the sample pre- and postanodization, respectively, while M3 denotes the weight of the sample following removal of the PSi layer using 3% KOH for 60 s [5]. The porosity is ∼70.33%. The thickness of the porous layer (d) was calculated to be 4.07 nm using Equation (2) [6]:

The FESEM micrographs of the n-PSi samples for different chemical ratios (1 : 1 : 0, 1 : 3 : 0, and 2 : 1 : 1) are shown in Figure 2. The dark and bright regions denote the nucleation of Si nanocrystallites and intergranular pores, respectively. Here, the crystallinity of Si structure declines, thereby increasing the density of the pores. Thus, it can be interpreted that the pore size and density of n-PSi nanostructures can be modified by regulating the chemical ratio of HF:C2H6O:H2O2 in solution. Figure 3 depicts the FESEM cross-sectional features of n-PSi prepared under different chemical ratios (HF:C2H6O:H2O2). The thickness of the layer varies significantly with the chemical ratio, with values of 22.11, 32.60, and 40.75 μm for 1 : 1 : 0, 1 : 3 : 0, and 2 : 1 : 1, respectively.

3.2. Surface Topography Characterization using AFM

It is imperative to emphasize that the AFM method provides images of the surface of pores, which makes it disadvantageous compared with the SEM technique [7]. Nonetheless, this technique can be performed using minute magnitude of forces [8]. The AFM image of PSi grown onto n-type Si (Figure 4) shows the usual rough surface morphology of the nanostructure. The 3D AFM image of the as-anodized PSi layer grown onto n-type for etching duration of 30 min at 45 mA cm−2 shows that the n-PSi film is uniform with many columnar-shaped grains. The scanned area film of 5 μm × 5 µm appears smooth with a root mean square (RMS) of ∼47.5 nm for roughness.

3.3. Structural Analysis

The XRD spectra of the PSi layers were obtained over 2θ range of 25°–30° (Figure 5). The relatively high intensity and broad peak centered at ∼28.42° can be indexed as (111) reflection of the Si cubic structure, which is consistent with JCPDS card no. 00-041-1426. No effect of the chemical ratio is noticeable. The increase in lattice constant in PSi silicon is observed, which can be attributed to an increase in induced stress by the formation of a native oxide on the pore surface.

The crystallite size (D) was derived using Scherrer’s equation as presented in Equation (3) [9]:where λ (= 1.5406 Å), B, and symbolize the incident X-ray wavelength, full-width half maximum (FWHM), and Bragg’s angle of diffraction, respectively. The crystallite size increases with higher ethanol concentration from 10.87 nm for 1 : 1 : 0 to 14.38 nm for 1 : 3 : 0, but slightly increases to 14.51 nm with the addition of hydrogen peroxide at the expense of ethanol (2 : 1 : 1).

The strain in the direction of the (111) diffraction plane of n-PSi was calculated using the basic equation expressed below (Equation 4):where C and C0 represent the experimental lattice constants of PSi and bulk Si from the JCPDS card (≈0.542 nm), respectively. It can be observed from Table 1 that the chemical ratio affects the strain, where it decreases with increasing ethanol concentration and decreased further with the addition of hydrogen peroxide. Furthermore, the positive value of indicates tensile character.

3.4. Photoluminescence Spectroscopy

The PL emission spectra of PSi synthesized at different chemical ratios (HF:C2H6O:H2O2) under room temperature are shown in Figure 6. A well-defined and distinct peak is discernible around 624.96 nm. As the concentration of ethanol and hydrogen peroxide were increased, this prominent emission peak made a blue shift, the width reduced and the intensity increased, suggesting the influence of the chemical ratio. The observed blue shift can be related to the quantum confinement effect [9, 10]. The reduction in particle (grains) size due to the variation in HF:C2H6O:H2O2 chemical ratio (from 1 : 1 : 0 to 1 : 3 : 0) may account for the strong quantum size confinement that explains the increases in energy band gap [11, 12] (see Table 2). Afterward, the simple photoelectrochemical technique for the preparation of PSi nanostructures presents different properties, indicating enhanced thermal stability, thus confirming their applicability as thermal insulators [13].

3.5. Fabrication of MSM Photodetector

Figure 7 presents a schematic illustration of the MSM photodetector (Pt/n-PSi/Pt) fabricated in this study. The Pt grid comprises two interdigital contacts (electrodes), which consist of five fingers each. The dimensions of each finger, in terms of width, spacing, and length are ≈230 µm, ≈400 µm, and 3.3 mm, respectively, with an active area of about 0.25 cm2. The Schottky contacts (Pt/n-PSi) are determined based on PSi electron affinity and Pt work function. Given the electron affinity of the semiconductor and the nature of metal, the junction exhibits a Schottky behavior [14], where the charge carriers are mobilized unrestrained, into or outside the semiconductor [15]. In this case, minimum resistance is observed as the contact is traversed. Since the n-PSi is an n-type semiconductor, the work function value of the used metal is proximate or less compared to its electron affinity [16]. The theoretical work function value of n-PSi (111) is 2.06 eV, which is considerably dependent on concentration of O2 vacancies produced in the synthesis process [17], whereas that of Pt is 4.08 eV [18]. The rectifying properties of the MSM photodetector are confirmed by the development of a Schottky barrier at the Pt–Si junction, which indicates its efficiency as a diode.

3.6. Electrical Characterization of the MSM Photodetector

IV curves of the Pt/n-PSi/Pt UV photodetector were obtained for different experimental conditions that include dark UV light illumination (365 nm/1.5 mW cm−2), and bias voltage range of +5 V to −5 V. Figure 8 displays the Schottky IV property under both forward and reverse bias conditions at +5 V. The Schottky contact was detectable at the interface linking the Pt electrodes to n-PSi nanostructure. The measurement of current both in the dark and under illumination involves placing the photodetector in a black box to prevent incident light in the measurements. At a bias voltage of +5 V, the dark current (1.81 × 10−6 A) is lower than values in related contemporary research [19]. In this context, it is imperative to emphasize that the signal-to-noise (S/N) ratio of the detector is strongly dependent on the low dark current [20]. The measured high photocurrent of 109 × 10−6 A under UV illumination can be ascribed to the large surface area-to-volume ratio and improved crystal structure of the synthesized n-PSi nanostructures. In addition, the use of higher etching/anodization current density remarkably improved the IV response of the MSM UV photodetector. Moreover, the conversion of the optical signal to an electrical signal via the inflow of charges inside the circuit indicates the successful completion of the photodetector. This implies that the IV properties of the MSM photodetector are dependent on the anodization current density.

The novel structural arrangement of the MSM photodetector, i.e., two Schottky contacts linked end-to-end, oriented positively toward the variation of reverse/forward bias voltage from −5 to +5 V, as it is recognized that sensitivity is improved as the bias voltage is increased. In addition, at higher bias voltage, n-PSi can absorb light intensely to produce additional e–h+pairs inside the space charge area. Therefore, the production of an electric field inside the space charge region, which results in the partition of the charge carriers, is dependent on the applied high bias voltage, thus generating an increasing rise in thermionic emission to enhance the mobility of the carriers. Accordingly, there is an exponential rise in current with the applied voltage, which surpasses the potential barrier. The bias voltage provides adequate energy to the electrons to disrupt the barrier height and generate a strong diffusion current [21], resulting in the shift of holes/electrons in the direction of positive/negative terminals, which accounts for the generation of photocurrent in the exterior contact.

The results showed that the sample prepared at HF:C2H6O:H2O2 chemical ratio, current density, and etching duration of 2 : 1 : 1, 45 mA cm−2, and 30 min, respectively, exhibited a very high photocurrent, possibly due to the considerable resistivity and Schottky barrier height. The sample prepared at 1 : 1 : 0 displayed the highest dark current, which can be attributed to other carrier transport mechanisms. For instance, by disregarding the thermionic emission over the Schottky barrier, the transport of charges can be ensured through the mechanism of direct tunneling of carriers [22]. At chemical ratio of 2 : 1 : 1, the dark/leakage current is lowest, indicating improved electrical properties at this ratio. It also suggests the influence of hydrogen peroxide in enhancing photocurrent detection. Thus, photodetection using MSM can be achieved with optical transparency in the UV region, which is consistent with an earlier study by Hadjersi and Gabouze [23]. The ideality factor and Schottky barrier height certainly confirm the efficiency of the fabricated photodetector.

This study further evaluated the effect of chemical ratio on UV photodetection (under UV light illumination at wavelength of 365 nm) at +3, +5, and +7 V bias voltages that are applied at an interval of 20 s (Figure 9). It is observed that the fabricated photodetector exhibits exceptional repeatability and stability at the selected bias voltages (+3, +5, and +7 V) with a rectangular-shaped It curve for each on/off cycle. The recovery and rise times of the photodetector were calculated from the I–t curves. It is discernible that the recovery and rise times are very swift for the different bias voltages. Regardless of the anodization current density, the spectral sensitivity was observed to be optimum at an applied bias voltage of +5 V. Afterwards, evaluating the photoresponse behavior of the detectors was maintained at a bias voltage of 5 V, while the HF:C2H6O:H2O2 ratio was varied. Thus, the spectral responsivity values of the three photodetectors (derived from different HF:C2H6O:H2O2 ratios) were evaluated at +5 V. After the application of thermionic emission–diffusion [21], the Schottky barrier height () and ideality factor (n) values were determined using the underlying equations (Equations 5 and 6):where I0, T, k, , A, and q denote saturation anodization current, temperature, Boltzmann constant, effective Richardson coefficient (≈112 A cm−2 K2), area of the Schottky contact (≈0.25 cm2), and electronic charge. The intercept of the ln IV plot was utilized to determine the value. A slight variation in was observed with varying chemical ratios. This is attributed to differential modification of the surface roughness of PSi, which significantly affects the n-PSi/Pt interface [24]. The average value of 0.96 eV was derived for n-PSi photodetector fabricated at the optimum HF:C2H6O:H2O2 chemical ratio of 2 : 1 : 1 at a bias voltage of +5 V. The high porosity achieved at large anodization chemical ratio accounts for the rise in , resulting in a shift in Fermi level toward the valence band, thereby reducing the dark current. The obtained high ideality factor values are consistent with the higher density of interfacial states [25]. The observed enhancement can also be attributed to the presence of two heterojunctions, substrate–PSi interface and n-PSi–Pt boundary [26]. Moreover, the presence of a heterogeneous barrier height can lead to a higher increase in n compared to the values reported in contemporary studies [27].

Furthermore, the It curve was utilized for the calculation of sensitivity (S) of the photodetector at a bias voltage of +5 V using the underlying relation (Equation 7) [10]:where denotes photocurrent, and represents the dark current density. The stability and repeatability of the photocurrent values of MSM photodetectors are ascertained during each on/off phase. The response and recovery times for the chemical ratio of 2 : 1 : 1 are 0.56 and 0.25 s, respectively, which are comparatively lesser than the values reported in relevant studies [28]. The obtained fast response at the optimum anodization chemical ratio (2 : 1 : 1) is possibly due to the defect-free nature, high quality, and considerable photoactive surface area of the synthesized n-PSi. Therefore, it is crucial to note that the rapid response of the photodetector is typically associated with the short transit time of photogenerated carriers that have long lifetimes [14], while the gradual response duration is associated with the incidence of grain margins and increase in surface adsorption of O2 [26].

The photodetection performance can be assessed based on its responsivity (R), which is can be calculated as the ratio of device photocurrent to incident optical power [29], as expressed below (Equation 8):where E denotes the photon beam energy obtained by means of a conventional UV power meter (Newport power meter, Model 2936-C, USA), A is the surface area of the photodetector. The responsivity (in AW−1) of the photodetector was obtained under ambient conditions in the range of 300–900 nm at a bias voltage of +5 V.

The photoresponses of the MSM photodetectors at different chemical (HF:C2H6O:H2O2) ratios are depicted in Figure 10, indicating the peak wavelength (λp) and peak responsivity (Rp) values. Regardless of the anodization chemical ratio, the responsivity spectra showed a distinctly observable peak at 365 nm. The Rp is highest (9.17 AW−1) for the photodetector prepared at optimum anodization chemical ratio (2 : 1 : 1). The apparent rise in responsivity as anodization chemical ratio is increased can be due to the specific surface area and varying thickness of the active area of n-PSi layers [14]. The maximum response (9.17 AW−1) of the photodetector is achieved for the sample with the highest layer thickness with a porosity that is seven times higher than values noted in related studies [30]. The high-intensity peak discernible at 365 nm originates from the upper Pt contacts, as previously observed for Au–Si-based devices [31]. Meanwhile, the slight shift in peak response from 365.29 to 365.45 is due to a structural change in the n-PSi layer.

The enhanced responsivity at lower wavelengths may be attributable to an increase in photocurrent. This is pertinent as increase in porosity at higher anodization current density generates a considerable number of eh+ pairs in the depletion region. This is anticipated since the depth of penetration reduces with wavelength, thus decreasing absorption in the Si substrate and concomitantly enhancing the depletion region. Therefore, the amount of eh+ pairs produced in the depletion region increases with decrease in wavelength. Additional observations reveal that the light responsivity declines abruptly at a lesser wavelength (blue and near UV), largely attributable to the absorption of photons in the n-PSi film, thus opposing its influx into the depletion region [23]. The responsivity value obtained in this study is higher compared to the value reported by Hadjersi and Gabouze [23].

The percentage quantum efficiency () of the photodetector is described on the basis of the amount of accumulated charge carriers required to generate the photocurrent () for a specific incident optical power () using Equation (9):

The quantum efficiency of the photodetector in terms of responsivity (R) gives the underlying Equation (10) [32]:where G denotes the current gain of the photodetector, while the other symbols have been explained in previous equations. The G can be derived from the ratio of accumulated electrons per unit of time in the dark to the number of absorbed photons to generate a photoelectron per unit of time [33]. G is derived from the IV properties using Equation (11):

At 365 nm (intensity of 1.5 mW cm−2) and a bias voltage of +5 V, the calculated value of G is ∼60.22 (Table 3). The reversibility of the photodetector was determined through the measurement of optical response (dynamic response time) when subjected to UV illumination. Table 3 outlines the performance parameters of the photodetectors fabricated from n-PSi prepared using different HF:C2H6O:H2O2 ratios. It can be easily deduced that the photodetector fabricated from n-PSi synthesized at the optimum chemical ratio (2 : 1 : 1) exhibited the best photodetection performance.

4. Conclusion

This study showed that the intrinsic properties of n-PSi can be controlled by varying the HF:C2H6O:H2O2 chemical ratio. The HF:C2H6O:H2O2 chemical ratio was, thus, optimized at 2 : 1 : 1. The photodetector fabricated from n-PSi synthesized at optimum chemical ratio (2 : 1 : 1), using planer interdigitated electrodes, exhibited excellent photodetection performance. The detector displayed a rapid photoresponse with a rise time of 0.56 s. The obtained responsivity (Rp) value of 9.17 at 365 nm is significantly higher than values reported for TiC/porous Si/Si. The good stability and repeatability, rapid response, and recovery of the MSM UV photodetector were credited to the high porosity and defect-free feature of the n-PSi thin films.

Data Availability

All th data used have been included in this manuscript.

Additional Points

Highlights. Investigates the influence of the chemical ratio of precursor materials (HF:C2H6O:H2O2) on the structural, morphological, optical, and electric properties of PSi. The PSi was produced by photoelectrochemical etching followed by subsequent anodization of the n-type Si under light illumination. The HF:C2H6O:H2O2 chemical ratio was optimized at 2 : 1 : 1. A MSM ultraviolet photodetector (Pt/n-PSi/Pt) was fabricated. The responsivity of the laser-annealed n-Psi with different chemical ratio can reach as high as 9.17 A/W. The current gain under 365 nm UV illumination is 60.22.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Authors’ Contributions

Asad A Thahe: conceptualization, methodology, experiment, and writing-original draft preparation. M.A. Qaeed and A. Mindil: data curation and original draft co-writing. Suhail Najm Abdullah: writing-reviewing, editing, and methodology. Hasan Alqaraghuli: visualization and investigation. Ibrahim M. Badawy: validation. Yasser Saleh Mustafa Alajerami and Ammar AL-Farga: testing, conceptualization, and methodology.

Acknowledgments

The authors thankfully recognize the financial support from the University of Fallujah and the University of Jeddah.