Journal Information
Vol. 8. Issue 3.
Pages 2767-2776 (May - June 2019)
Share
Share
Download PDF
More article options
Visits
391
Vol. 8. Issue 3.
Pages 2767-2776 (May - June 2019)
Original Article
DOI: 10.1016/j.jmrt.2019.03.011
Open Access
Structural, morphological, optical, and gas sensing characteristics of ultraviolet-assisted photoelectrochemical etching derived AlInGaN nano-spikes
Visits
391
Way Foong Lim, Zainuriah Hassan, Hock Jin Quah
Corresponding author
Institute of Nano Optoelectronics Research and Technology (INOR), Universiti Sains Malaysia, 11800 Penang, Malaysia
This item has received
391
Visits

Under a Creative Commons license
Article information
Abstract
Full Text
Bibliography
Download PDF
Statistics
Figures (12)
Show moreShow less
Abstract

The formation of nano-dendritic like structure and nano-spikes in AlInGaN films via ultraviolet-assisted photoelectrochemical (PEC) etching at different current densities (5, 20, and 40mA/cm2) could be potentially deployed as the hydrogen sensor. The ability of nano-dendritic like structure and nano-spikes to provide large surface area to volume ratio could improve hydrogen (H) adsorption in the AlInGaN films, and thereby offering a greater sensitivity as compared to the as-grown film. The film subjected to PEC etching at 40mA/cm2 has demonstrated the highest sensitivity (79.6%), followed by that subjected to PEC etching at 20 and 5mA/cm2. The acquisition of the highest sensitivity in the aforementioned film suggested that nano-spikes (40mA/cm2) surpassed nano-dendritic like structures (5 and 20mA/cm2) in term of providing larger surface area to volume ratio for H adsorption. Moreover, the largest total dislocation density present in the nano-spikes film could be the reason contributing to the increased gas sensitivity because the dislocation could serve as the trapping sites to mediate the diffusion of the adsorbed H, and thus facilitating the H detection. As a result, a fast response time (105s) and recovery time (46s) was obtained.

Keywords:
AlInGaN
Photoelectrochemical
Nano-spikes
Dislocation
Sensitivity
Response
Full Text
1Introduction

With the advent of technology revolution nowadays toward Industry 4.0, especially in the field of electronics and optoelectronics, the III-nitrides semiconductors, namely the binary gallium nitride (GaN) and its ternary and quaternary alloys (indium gallium nitride (InGaN), aluminum gallium nitride (AlGaN), and aluminum indium gallium nitride (AlInGaN)) have left remarkable footprints for deployment as the core materials in energy efficient solid-state lighting to substitute the conventional lighting sources [1]. Numerous efforts have been also plunged to lessen the dependence on fossil fuels for energy supplies in the aerospace, automotive, chemical plants, and semiconductor manufacturing industries to ensure sustainable development of a nation [2–4]. The initiation of exploiting clean and renewable hydrogen (H2) energy as a source of energy supply would restrain the unpleasant threats posed by the fossil fuel, which has brought significant breakthrough in term of improving energy efficiency while reducing environmental pollution risks [2–4]. Nevertheless, a proper handling of H2 gas is a subject of matter that should not be taken lightly due to confrontation of severe hazardous phenomenon that might take place as a consequence of leaking of odorless, highly flammable, and highly volatile H2 gas beyond certain concentration (>4%) at room temperature (Troom) [3]. In order to deter the happening of unforeseen circumstances, a highly sensitive and reliable H2 gas sensor with the capability of detecting gas leak and operate under hostile environment is required [3].

To-date, GaN and its alloys have been also deemed for use as H2 gas sensor due to the possession of wide band gap, large critical electric field, high electron mobility, as well as good chemical stability, apart from emerging as the material-of-choice for solid-state lighting [3,5,6]. It was disclosed in recent times that H2 gas sensing performance of planar GaN films could be modulated via the formation of porous or nanostructured GaN, wherein a reduction in operating temperature to Troom as well as an enhancement in response time could be achieved [3,4,7]. With these, the deployment of Troom H2 gas sensing using the porous or nanostructured GaN would thus minimize the hazardous risk of ignition confronted by the planar GaN films, in which continuous heating in inflammable environment is essential for detection of H2 gas [8,9]. Further investigation has been spread over to the exploration of ternary and quaternary GaN-based alloys for use as H2 gas sensor [9]. Self-assembled InGaN/GaN nanowires grown using molecular beam epitaxy (MBE) have revealed the success of detecting H2 gas as low as 20ppm at 30°C [9]. In addition, plasma assisted MBE grown GaN/InGaN nanowire-based opto-chemical transducers operated at 80°C were able to detect H2 gas at 100ppb [10]. The growth of GaN nanowires using chemical vapor deposition technique coupled with platinum nanoparticles has demonstrated a tremendous improvement in response time and sensitivity as ultraviolet photodetector [11]. Besides, porous quaternary InAlGaN alloy formed via xenon-assisted photoelectrochemical (PEC) etching in acidic electrolyte has also demonstrated a better sensitivity toward H2 gas as compared to the non-porous InAlGaN at Troom[12].

Although significant improvement in term of H2 gas sensitivity in comparison to that of planar films has been demonstrated by porous or nanostructured GaN, InGaN, and/or InAlGaN, the underlying science behind the technological improvement of sensing performance has not been divulged. Thus far, it has been reported elsewhere that the presence of porous or nanostructures would provide a higher surface area to volume ratio, which has served as the key to the enhancement of gas sensing performance. Nevertheless, supporting evidences that could assist in substantiating the improved sensing characteristics are not well established. According to previous studies reporting about the advantages in term of a reduction in dislocation density and an improvement in stress relaxation of the films, accompanied with defect generation through pore formation, the dissimilar etching solution, etching current density, or etching duration could be the possible factors contributing to the alteration of structural, morphological, and optical properties of the films. However, significant influence of the porous or nanostructures in influencing the sensing characteristics of the films as well as the relationship has not been well delivered.

Therefore, it is of interest in this work to execute a systematic investigation in different aspects, encompassing structural, morphological, optical, and sensing characteristics of quaternary AlInGaN nanostructures produced via ultraviolet (UV)-assisted PEC etching for 5min in a diluted potassium hydroxide (KOH) solution at different current densities (J=5, 20, and 40mA/cm2).

2Experimental procedures

A commercially purchased unintentionally doped n-type 100nm thick Al0.1In0.1Ga0.8N epilayer grown on silicon substrate using plasma-assisted molecular beam epitaxy was diced into smaller pieces before subjected to UV-assisted PEC etching at different J of 5, 20, and 40mA/cm2 for 5min in a diluted 1% KOH electrolyte. Detailed description for the PEC etching has been discussed in Ref. [13]. Crystalline phase and orientation of the investigated samples were characterized using high-resolution X-ray diffraction (HR-XRD; Panalytical X’Pert PRO MRW PW3040) operated at 40kV and 30mA. Rocking curves (RCs) of symmetric (0002) and asymmetric (1 0 1¯ 2) scans were carried out at 40kV and 35mA. Field-emission electron microscopy (FESEM; FEI Nova NanoSEM 450) was utilized to examine surface morphology of the samples while atomic force microscopy (AFM; Dimension Edge, Bruker) was employed to examine surface topography of the samples. Optical characteristics of the samples were accessed using Raman spectroscopy (Horiba Jobin Yvon HR800UV) and photoluminescence (PL; Horiba Jobin Yvon HR800UV). In order to fabricate AlInGaN-based gas sensor, RF-magnetron sputtering (HHV Auto 500) was used to deposit platinum (Pt) contacts (area=5×10−3cm2) on the samples using shadow mask. Electrical measurements of the sensors (current–time (It)) were performed in a home-made chamber using Keithley 2400 source meter connected to personal computer and the flow of 0.1% H2 with 99.9% N2 gas was controlled using a digital mass flow controller.

3Results and discussion

Field-emission scanning electron microscopy (FESEM) has been used to examine surface morphology of the investigated samples (Fig. 1). Discrete nano-islands with dissimilar sizes and shapes were seen to distribute over the entire surface of the as-grown AlInGaN film (Fig. 1a). The occurrence of PEC etching induced on the AlInGaN film at different J (5 (Fig. 1b), 20 (Fig. 1c), and 40mA/cm2 (Fig. 1d)) has contributed to morphological change of the film surface. An initial formation of nano-dendritic like structures was observed in the film subjected to PEC etching at 5mA/cm2. Subsequent increase of J to 20mA/cm2 (Fig. 1c) has enhanced the formation of nano-dendritic like structures to an extent that a transformation of the structures to nano-spikes has taken place as the J approached 40mA/cm2 (Fig. 1d). It could be deduced from the figure that the dendrites have elongated sideways till joining the neighboring dendrites to form the nano-spikes structures.

Fig. 1.

Field emission scanning electron microscopy (FESEM) images of (a) as-grown and AlInGaN films obtained after PEC etching at (b) 5mA/cm2, (c) 20mA/cm2, and (d) 40mA/cm2.

(0.58MB).

Surface topography of the investigated film has been examined using atomic force microscopy (AFM) on a 5μm×5μm scanning area. It could be conveniently observed from Fig. 2 that protrusions with relatively similar height and width were distributed over surface of the as-grown film (Fig. 2a). Changes in term of height and width took place when the film was subjected to PEC etching at different J (Figs. 2b–d). The introduction of J at 5mA/cm2 has broadened the protrusions, owing to the occurrence of coalescence effect in the film. The discrete nano-islands observed in Fig. 1a could have shrunk vertically while expanding horizontally, forming the dendritic structure (Fig. 1b). This phenomenon has decreased much of the initial protrusion height. Nevertheless, the subsequent increase of J to 20mA/cm2 has encouraged the etching process to an extent that vertical shrinkage and/or horizontal expansion have taken place uncontrolledly, causing a huge variation in term of the protrusion height and width (Fig. 2c). As the J was further increased to 40mA/cm2, an exaggeration of etching process has triggered the formation of protrusions with dissimilar height and width extending horizontally and vertically throughout the entire surface of the film (Fig. 2d). The corresponding root-mean-square (RMS) roughnesses of the films were presented in Fig. 3 with respect to the etching J. It could be noticed from the figure that an increase of the J from 5mA/cm2 to 40mA/cm2 has contributed to an increase in the RMS roughness value. This finding was very much aligned with the topographies of the investigated films (Fig. 2), wherein an enhancement in dissimilarity of protrusion height and width was perceived as a function of J.

Fig. 2.

Atomic force microscopy images of (a) as-grown and AlInGaN films obtained after PEC etching at (b) 5mA/cm2, (c) 20mA/cm2, and (d) 40mA/cm2.

(0.38MB).
Fig. 3.

Root-mean-square (RMS) roughness of the investigated films with respect to PEC etching current density, J.

(0.08MB).

Crystalline phases present in the investigated films were detected using high resolution X-ray diffraction (HRXRD) as shown in Fig. 4. It could be deduced that diffraction peaks of aluminum nitride (AlN) (International Centre of Diffraction Data (ICDD) file no. of 03-065-0831) and Si (111) (ICDD file no. 01-073-1512), accompanied with the emergence of diffraction peak, ascribed to hexagonal phase of GaN (ICDD file no. of 00-050-0792) oriented in (0002) plane were revealed in all of the investigated films (Fig. 4). In the present scenario, the detected GaN peak was translated as the AlInGaN phase, having known that AlInGaN could be the solid solution of GaN (0.047nm) with the incorporation of trivalent indium (0.079nm) and aluminum (0.039nm) [13,14].

Fig. 4.

High resolution X-ray diffraction (HR-XRD) patterns of as-grown and AlInGaN films subjected to PEC etching at 5, 20, and 40mA/cm2. Insets (a) and (b) show typical rocking curves of symmetric ω-scan of (0002) plane and asymmetric ω-scan of (1 0 1¯ 2) plane, respectively, for AlInGaN films subjected to PEC etching at 5mA/cm2.

(0.14MB).

In addition, minute changes in terms of the peak position for (0002)-oriented AlInGaN diffraction peak to either lower or higher angles could be also observed along with the changes of J. It could be seen from Fig. 4 that the peak for AlInGaN film subjected to PEC etching at 5mA/cm2 was shifted to a lower angle while an opposite trend was attained by other films etched at larger J (20mA/cm2 and 40mA/cm2) with respect to the as-grown film. The changes in (0002)-oriented peak position could be supported through the calculation of lattice parameter c using Eqs. (1) and (2) by taking into consideration of θ0002 of the films and subsequently the interplanar spacing (dhkl) values:

where λ is the X-ray radiation wavelength, n is the diffraction order, dhkl is the interplanar spacing, (hkl) is Miller's index, and θ is the diffraction angle [15]. Fig. 5 presents the relationship of lattice parameter c with respect to J. A decrease in the lattice parameter c was observed along with the increase of J, except for the film subjected to PEC etching at 5mA/cm2. This finding was possible if lattice contraction has taken place in the films subjected to etching at 20 and 40mA/cm2 while on the other hand, an early stage of PEC etching at 5mA/cm2 has in fact caused the lattice to expand, rather than to contract. The occurrence of either lattice expansion or contraction could be proven through the determination of strain energy present in the investigated films.

Fig. 5.

Lattice parameters c and a of all the investigated InAlGaN films in comparison to as-grown film.

(0.07MB).

It has been noteworthy that any alteration in the distance between atoms sitting in a lattice of the film would result in strain formation. Depending on either the size of foreign atoms added to the lattice in comparison to that of the host atoms or the defect formation in the lattice with respect to charge compensation, strain generation would happen [16–18]. In order to substantiate the dependence of lattice parameter c on strain, Eq. (3) has been used to calculate the strain along the c-direction (ɛzz) [19]:

where ɛzz is the out-of-plane strain, c is the lattice parameter c of the porous and as-grown AlInGaN, and co is the lattice parameter of strain free AlInGaN, which could be determined using Vegard's law [20]. The corresponding ɛzz relationship is conveniently disclosed in Fig. 6 to be in an agreement with the changes in lattice parameter c with respect to J. In comparison, positive value (tensile) was perceived by the film subjected to etching at 5mA/cm2 while negative values (compressive) were detected for the films subjected to higher etching J. The changes in lattice parameter c and the ɛzz have in-turn contributed to an opposite trend for the lattice parameter a and the strain along the a-direction (ɛxx/yy), in which the increase of J from 0 to 20mA/cm2 and to 40mA/cm2 has resulted in an inverse proportional relationship.

Fig. 6.

Strains along the c-direction and a-direction as well as hydrostatic strain calculated for as-grown film and AlInGaN etched at different J.

(0.09MB).

The values of lattice parameter a (Fig. 5) were obtained by taking into consideration of the diffraction angle (θ1 0 1¯ 2) obtained from RC for asymmetric ω-scan of the samples as well as dhkl at (1 0 1¯ 2) plane substituted into Eqs. (1) and (2) while ɛxx/yy values were calculated using Eq. (4)[19]:

where ɛxx/yy is the in-plane strain, a is the lattice parameter a of the porous and as-grown AlInGaN, and ao is the lattice parameter a of strain free AlInGaN, which could be determined using Vegard's law [20]. Typical asymmetric (1 0 1¯ 2)ω-scan of the AlInGaN film subjected to PEC etching at 5mA/cm2 is presented in inset of Fig. 4.

In addition, hydrostatic strains (ɛh) present in all of the investigated samples were also calculated using Eq. (5)[21] by considering both ɛzz and ɛxx/yy values of the films:

where ɛh and v are the hydrostatic strain and Poisson's ratio, respectively. The calculated value of v for AlInGaN films was 0.228 by taking into consideration of reported v values for GaN (0.210) [22,23], AlN (0.203) [23,24], and InN (0.399) [23,25]. It could be observed from Fig. 6 that the trend demonstrated by ɛh with respect to J was in an agreement with the trend obtained for ɛxx/yy. This finding suggested the plausible generation of defects along a direction in the lattice of the films. On the other hand, the acquisition of larger lattice parameters a in all the films subjected to etching as compared to the as-grown film as well as the recognition of tensile ɛxx/yy proposed that a larger size defect might have been generated in the films after the etching process. The defect was likely to be In3+ interstitial, owing to its larger size (0.064nm) than Ga3+ (0.047nm) and Al3+ (0.039nm). Among the films subjected to etching at different J, the largest values of lattice parameter a and ɛh were perceived by the film etched at 5mA/cm2. As the J was increased from 5 to 40mA/cm2, a decrease in the lattice parameter a (Fig. 5) and ɛh (Fig. 6) was observed. This finding suggested the occurrence of more dissolution of In(OH)3 as a function of J, which resulted in the formation of more In3+ vacancies, surpassing that of In3+ interstitials. Due to the difficulty of In(OH)3 dissolution as compared to Ga3+ and Al3+, more In3+ would preferably reside at the interstitial sites, inducing higher electrostatic repulsion effect toward the neighboring Ga3+ and Al3+ cations and/or vacancies. The N3− anions would be also pulled toward the In3+ interstitials, expanding the lattice [13]. Nevertheless, the increase of J from 5 to 40mA/cm2 would allow the In3+ cations to be completely dissolved, forming In3+ vacancies. This would reduce much of the repulsion effect brought by the In3+ interstitials. As a result, the lattice would become less relaxed, decreasing the lattice parameter a[13].

Density of screw dislocation (Nscrew) and edge dislocation defects (Nedge) present in the investigated films have been calculated using Eqs. (6) and (7)[26,27] by taking into consideration of full-width half maximum (FWHM) of the RCs for symmetric (0002) and asymmetric (1 0 1¯ 2) planes, respectively:

where Nscrew, Nedge, β(0002), β(1 0 1¯ 2), bscrew, and bedge are the screw dislocation density, edge dislocation density, ω-scan full-width-half-maximum (FWHM) of (0002) plane, ω-scan FWHM of (1 0 1¯ 2) plane, c-type Burger's vector, and a-type Burger's vector, respectively. It could be worth noting from Fig. 7 that the investigated films have demonstrated larger Nedge values (∼(1.46–1.71)×1011cm−2) than that of Nscrew (∼(1.69–3.28)×109cm−2), revealing the dominance of dislocation defects along a direction of the lattice. Inevitably, the resultant of dislocation density, which is referred to as the total dislocation density (Ntotal=Nscrew+Nedge) was showing the similar trend with that of Nedge, whereby the increase of J from 5 to 40mA/cm2 has contributed to an increased Ntotal. The acquisition of a dissimilar trend for ɛh (Fig. 6) as well as lattice parameters a and c (Fig. 5) from that of Ntotal suggested that the increase in In(OH)3 dissolution with respect to J might not only cause the occurrence of lattice contraction, but also increased the total dislocation density in the films. In other words, the acquisition of an increased Ntotal with respect to J was related to the increase in In(OH)3 dissolution, which resulted in the formation of more In3+ vacancies.

Fig. 7.

Screw dislocation density, edge dislocation density, and total dislocation density for as-grown film and AlInGaN etched at different J.

(0.1MB).

Room temperature Raman spectroscopy has been used to determine optical phonon modes present in the investigated films as compared to that of the as-grown film (Fig. 8). It could be conveniently acquired from the Raman spectra that the strongest peak associated with Si (521cm−1) [28] was detected in all of the samples. Apart from the Si peak, additional four Raman peaks fitted to E2(high) InGaN [29], E2(high) AlGaN [30], E1(TO) AlN [31], and A1(LO) InGaN [32] were also detected, respectively at 567.9cm−1, 608.3cm−1, 693.6cm−1, and 729.8cm−1 in the as-grown film. The presence of these peaks that were shifted to lower wavenumber were also observed in the film subjected to etching at 5mA/cm2. Nonetheless, disappearance of the E2(high) InGaN and E2(high) AlGaN was noticed as the J was increased to 20mA/cm2, leaving only the Raman peaks allied with E1(TO) AlN and A1(LO) InGaN phonon modes with decreased peak intensities. Further increase of the J to 40mA/cm2 did not seem to suppress the peak intensities, but instead increasing the peak intensity along with the emergence of E2(high) AlN (651.8 and 666.2cm−1) [29] and re-appearance of E2(high) InGaN. When comparing among the Raman spectra acquired for the films subjected to 5mA/cm2 and 20mA/cm2, the absence of E2(high) AlGaN and E2(high) InGaN modes in the latter suggested that much of the In3+, Al3+, Ga3+, and N3− have been removed as a result of the etching process. However, the increase of J to 40mA/cm2 has caused the re-emergence of E2(high) InGaN phonon mode, which could be in conjunction with the nano-spikes structure formed in the film (Fig. 1d). It was predicted that the nano-spikes could have served as the platform for surface-enhanced Raman scattering to happen, and therefore revealing more phonon modes, which were in fact not detected in the film etched at 20mA/cm2.

Fig. 8.

Deconvoluted Raman spectra for as-grown and AlInGaN films subjected to PEC etching at different J.

(0.22MB).

Room temperature photoluminescence (PL) spectra of the investigated films are depicted in Fig. 9. Multi-emission peak in the blue-to-green-to-red region (400–700nm) was obtained from the investigated films with the strongest peak attributed to the blue emission region (479.6–485.3nm). In comparison, as the J was increased from 5 to 40mA/cm2, intensities of the emission peaks were significantly trimmed down. The decrease of PL emission peaks with respect to the J could be associated with the generation of density of states in the films after the PEC etching process, whereby the density of states might have served as the non-radiative recombination centers to degrade the localization of excited carriers for radiative recombination to take place [15]. By referring to the trend obtained for Ntotal (Fig. 7), the increased Ntotal was plausibly serve as the non-radiative recombination centers that influenced the changes in the PL intensity. On the other hand, it could be noted that the intensities of emission peak in the film subjected to PEC etching at 5mA/cm2 were higher than that of as-grown film. This phenomenon could be related to the above mentioned occurrence of In(OH)3 dissolution at slower pace at 5mA/cm2 via the detection of E2(high) InGaN and A1(LO) InGaN (Fig. 8). It has been disclosed that the presence of In3+ ions could inevitably serve as the radiative recombination centers in the film. Likewise, the absence of E2(high) InGaN and the decrease in intensity for A1(LO) InGaN for the film etched at 20mA/cm2 have suggested the disappearance of much of the In3+ contents in the film. Thus, a degradation in the PL intensity was observed as the J approached 20mA/cm2. Although both the optical phonon modes appeared again in the film etched at 40mA/cm2, the growth of elongated branches on the dendritic structures till the formation of nano-spikes would trigger a destructive interference effect, wherein the scatter light would be unable to propagate through the AlInGaN surface.

Fig. 9.

Room temperature photoluminescence (PL) spectra of as-grown and AlInGaN films subjected to PEC etching at different J.

(0.16MB).

UV–visible reflectance spectra acquired by the as-grown film and AlInGaN films subjected to PEC etching at 5, 20, and 40mA/cm2 have been employed to determine empirically the optical band gap (Eg) using Kubelka–Munk (KM) function by approximating the optical absorption from the diffuse reflectance spectrum, according to the following equation [33,34]:

where R is the diffuse reflectance and the F(R) function can be multiplied by hv using corresponding coefficient (n), which can be 1/2 and 2 for direct-band and indirect-band transition, respectively. A plot of (F(R)×hv)2 vs hv (not shown) would give a straight line, upon which the extrapolation to (F(R)×hv)2=0 would provide the Eg values. In comparison, the AlInGaN films subjected to PEC etching at different J have demonstrated smaller Eg values than that of the as-grown film. As the etching J was increased from 5 to 40mA/cm2, a decrease in the Eg values from 2.2447 to 2.1266eV (Fig. 10) was observed. The decrease in the Eg value with respect to J could be associated with an increase in In3+ vacancies in the film, which might have introduced extra levels in the band gap of AlInGaN films [34].

Fig. 10.

A relationship between optical band gap values of the investigated films.

(0.08MB).

Time dependence of sensor response in term of forward current (I) change at forward bias of 2.0V for the as-grown and etched AlInGaN films with (H2 on) and without H2 gas exposure (N2 on) for 4 cycles is shown in Fig. 11. A good repeatability in the I change and the ability to cycle the I in response to repeated exposure of H2 was achieved in all the films. An instantaneous rise in I after switching from N2 to H2 ambient for the samples might imply that diffusion of H2 through the Pt electrode with catalytic properties might not limit the time response of the sensors. Nevertheless, the dissimilar response time (trep) obtained for the sensors, wherein a longer trep was attained by the as-grown film (315s) as compared to that of 5 (266s), 20 (175s), and 40mA/cm2 (105s) might suggest that the trep was influenced by a chemical reaction between the H2 gas and the AlInGaN films, which might lead to the I change after switching of the gas ambient [35,36]. The largest I change for AlInGaN nano-spikes formed at 40mA/cm2 has contributed to the highest sensitivity for this sample (79.6%) as compared to nano-dendritic like structures formed at 20 (40.9%), and 5mA/cm2 (22.6%), as well as nano-islands for the as-grown film (7.1%). As the gas ambient was switched back to N2 after the H2 exposure, a faster recovery time (trec) was obtained by the AlInGaN nano-dendritic like and nano-spikes through the attainment of 276s, 175s, and 46s at 5, 20, and 40mA/cm2, respectively as compared to the as-grown film (311s).

Fig. 11.

I–t responses of (a) as-grown and AlInGaN films subjected to PEC etching at (b) 5mA/cm2, (c) 20mA/cm2, and (d) 40mA/cm2.

(0.12MB).

The improvement of trep and sensitivity for AlInGaN nano-spikes and nano-dendritic like structures might be attributed to the presence of larger surface area to volume ratio as a consequence of the PEC etching process. The increased surface area to volume ratio was expected to enhance the adsorption of hydrogen (H) atoms, which would be there after polarized at the interface between the Pt and AlInGaN layers because of the built-in electric field present in the depletion region [2,37]. It has been reported elsewhere that large surface area to volume ratio would enhance the polarization effect, leading to an increased electrical conduction [38] due to a narrowing of the depletion width [39] and thus enhancing the accumulation of electrons on the AlInGaN surface [40]. Furthermore, electrical conduction could be also enhanced via the accelerated diffusion process of the adsorbed H atoms [41] through increased number of trapping sites (defects) present either in the bulk or surface of the AlInGaN film [37,42,43]. The same reasons might facilitate desorption of the H atoms and thereby leading to an improvement in the trec.

An increasing trend in the Ntotal from 1.69×109cm−2 at 5mA/cm2 to 2.25×109and 3.28×109cm−2 at 20 and 40mA/cm2, respectively has evidenced that the increase of Ntotal could serve as the trapping sites [44] for H to diffuse through the nano-dendritic like and nano-spikes. As a result, an enhancement in trep, sensitivity, and trec of the sensors was obtained. Nevertheless, the attainment of a lower Ntotal for the nano-dendritic like structure obtained by etching at 5mA/cm2 than that of the as-grown sample (1.76×109cm−2) indicated that PEC etching at 5mA/cm2 might have removed much of the defects initially present in the as-grown sample. The generation of more dislocations at J>5mA/cm2 might therefore deteriorate peak intensity for (0002)-oriented AlInGaN peaks detected for the films (Fig. 4).

Fig. 12 presents time dependence I change at different forward biases (0.5V, 1.0V, and 2.0V) in response to H2 on and H2 off (N2) ambient for AlInGaN nano-spikes obtained at 40mA/cm2. A decrease in sensitivity of the AlInGaN nano-spikes from 79.6% to 11.9% was attained as the bias was decreased from 2.0V to 0.5V. The decrease of forward bias might not provide sufficient energy to promote the diffusion of adsorbed H atoms to the AlInGaN surface for interaction to take place, and therefore decreasing the polarization effect. With this, the number of electrons accumulated on the AlInGaN surface would be trimmed down and the transport of electrons across the interface would be affected [38], leading to a degraded electrical conduction. Though the sensitivity of AlInGaN nano-spikes was poorer at 0.5V, it was still viable to be potentially used as a low power consumption H2 sensor, owing to the ability to repeatedly adsorb and desorb H.

Fig. 12.

It responses at different biases (0.5, 1.0, and 2.0V) and different flow rates (inset) for AlInGaN nano-spikes formed at 40mA/cm2.

(0.13MB).

Consecutive I responses of the AlInGaN nano-spikes obtained by etching at 40mA/cm2 with respect to time as the H2 flow rate was increased from 400 to 4000sccm at a fixed bias of 2.0V are presented in inset of Fig. 12. A decrease in the trep and trec was observed as the flow rate was varied from 400 to 1000sccm. For instance, the trep taken was 105s and 62.1s, while trec was 46s and 8.2s at 400sccm and 1000sccm, respectively. The decrease of trep and trec indicated that mass transport of the H2 gas in and out from the sensors might improve the time response of the sensor [4]. As the flow rate was increased beyond 1000sccm, the trep and trec were increased from 75.8s to 84.5s and from 79.5s to 89.4s, respectively, suggesting that surface of the film might have been saturated with the adsorbed H atoms, and thus delaying the response of the film. However, the attained trep and trec were faster than that attained at 400sccm. Therefore, it was denoted that besides mass transport of the H2 gas, surface condition of AlInGaN might also affect the sensor performance. In addition, repeatability of the sensor to respond to different H2 flow rates and recover as the gas ambient was switched from H2 to N2 ambient indicated good adsorption and desorption behaviors of the adsorbed H on the AlInGaN nano-spikes.

4Conclusions

Ultraviolet (UV)-assisted photoelectrochemical (PEC) etching of AlInGaN films at different etching current densities (J) has successfully produced nano-dendritic like structure and nano-spikes, which could be used as the hydrogen (H2) sensors. As the J was increased from 5mA/cm2 to 40mA/cm2, a transformation of AlInGaN nano-islands present in the as-grown film to nano-dendritic like structure and nano-spikes was obtained. The presence of nano-dendritic like structure and nano-spikes in the AlInGaN films has provided larger surface area to volume ratio for hydrogen (H) adsorption to take place. In comparison, a greater sensitivity (79.6%) was demonstrated by the AlInGaN nano-spikes for H2 gas sensing as compared to that of nano-dendritic like (40.9% and 22.6% for AlINGaN film etched at 5mA/cm2 and 20mA/cm2, respectively). Moreover, a significant improvement in term of sensitivity could be obviously seen through the acquisition of approximately 10 times greater sensitivity for the film subjected to J at 40mA/cm2 (79.6%) with respect to that of the as-grown film (7.1%). The improvement in gas sensitivity could be also associated with the increase in total dislocation density (Ntotal) generated in the sample with respect to the J. The presence of the largest Ntotal in the film subjected to J of 40mA/cm2 has contributed to the fasted response (trep) and recovery (trec) time, which was 105s and 46s, respectively. The increase of forward bias to 2.0V has also enhanced sensitivity of the AlInGaN nano-spikes yet the increase of H2 flow rate beyond 1000sccm has created a H-saturated AlInGaN film surface, and thus the trep and trec became longer.

Conflicts of interest

The authors declare no conflicts of interest.

Acknowledgement

The authors would like to acknowledge Universiti Sains Malaysia and Long Term Research Grant (LRGS; 203/CINOR/6720013) for their financial support.

References
[1]
C.L. Zhou, A. Ghods, V.G. Saravade, P.V. Patel, K.L. Yunghans, C. Ferguson, et al.
Review – The current and emerging applications of III-nitrides.
ECS J Solid State Sci Technol, 6 (2017), pp. Q149-Q156
[2]
J.R. Huang, W.C. Hsu, I.H. Chen, W.C. Liu.
Comparative study of hydrogen sensing characteristics of a Pd/GaN schottky diode in air and N2 atmospheres.
Sens Actuators B Chem, 123 (2007), pp. 1040-1048
[3]
A. Zhong, T. Sasaki, K. Hane.
Comparative study of schottky diode type hydrogen sensors based on a honeycomb GaN nanonetwork and on a planar GaN film.
Int J Hydrogen Energy, 39 (2014), pp. 8564-8575
[4]
K. Al-Heuseen, M.R. Hashim.
Enhancing hydrogen sensitivity of porous GaN by using simple and low cost photoelectrochemical etching techniques.
Sens Actuators B Chem, 163 (2012), pp. 159-164
[5]
H.J. Quah, K.Y. Cheong.
Effects of post-deposition annealing ambient on chemical, structural, and electrical properties of RF magnetron sputtered Y2O3 gate on gallium nitride.
J Alloys Compd, 575 (2013), pp. 382-392
[6]
H.J. Quah, K.Y. Cheong.
Surface passivation of gallium nitride by ultrathin RF-magnetron sputtered Al2O3 gate.
ACS Appl Mater Interfaces, 5 (2013), pp. 6860-6863
[7]
B. Liu, W. Yang, J. Li, X. Zhang, P. Niu, X. Jiang.
Template approach to crystalline GaN nanosheets.
Nano Lett, 17 (2017), pp. 3195-3201
[8]
J. 7Sun, J. Xu, Y. Yu, P. Sun, F. Liu, G. Lu.
UV-activated room temperature metal oxide based gas sensor attached with reflector.
Sens Actuators B Chem, 169 (2012), pp. 291-296
[9]
S. 8Paul, A. Helwig, G. Muller, F. Furtmayr, J. Teubert, M. Eickhoff.
Opto-chemical sensor system for the detection of H2 and hydrocarbons based on InGaN/GaN nanowires.
Sens Actuators B Chem, 173 (2012), pp. 120-126
[10]
S. Paul, A. Helwig, G. Muller, F. Furtmayr, J. Teubert, M. Eickhoff.
Opto-chemical sensor system for the detection of H2 and hydrocarbons based on InGaN/GaN nanowires.
Sens Actuators B Chem, 173 (2012), pp. 120-126
[11]
X. Zhang, Q. Liu, B. Liu, W. Yang, J. Li, P. Niu, et al.
Giant UV photoresponse of a GaN nanowire photodetector through effective Pt nanoparticle coupling.
J Mater Chem C, 5 (2017), pp. 4319-4326
[12]
R. Radzali, Z. Hassan, N. Zainal, F.K. Yam.
Preparation of porous InAlGaN/Si (111) by photoelectrochemical etching for high performance hydrogen gas sensors at room temperature.
Sens Actuators B Chem, 213 (2015), pp. 276-284
[13]
W.F. Lim, H.J. Quah, Z. Hassan, R. Radzali, N. Zainal, F.K. Yam.
Alteration of structural and optical properties in quaternary Al0.1In0.1Ga0.8N films using ultraviolet assisted photo-electrochemical etching route.
J Alloys Compd, 649 (2015), pp. 337-347
[14]
S.D. Li, M.S. Sellers, C. Basaran, A.J. Schultz, D.A. Kofke.
Lattice strain due to an atomic vacancy.
Int J Mol Sci, 10 (2009), pp. 2798-2808
[15]
H.J. Quah, W.F. Lim, Z. Hassan, R. Radzali, N. Zainal, F.K. Yam.
Effects of ultraviolet-assisted electrochemical etching current densities on porous quaternary AlInGaN alloys.
[16]
B. Santara, P.K. Giri, K. Imakita, M. Fujii.
Microscopic origin of lattice contraction and expansion in undoped rutile TiO2 nanostructures.
J Phys D Appl Phys, 47 (2014),
[17]
V. Perebeinos, S.W. Chan, F. Zhang.
‘Madelung Model’ prediction for dependence of lattice parameter on nanocrystal size.
Solid State Commun, 123 (2002), pp. 295-297
[18]
W.F. Lim, H.J. Quah, Z. Hassan, R. Radzali, N. Zainal, F.K. Yam.
Porous quaternary Al0.1In0.1Ga0.8N film formation via photoelectrochemical etching in HF:C2H5OH electrolyte.
J Am Ceram Soc, 99 (2016), pp. 2395-2401
[19]
V.S. Harutyunyan, A.P. Aivazyan, E.R. Weber, Y. Kim, Y. Park, S.G. Subramanya.
High-resolution X-ray diffraction strain-stress analysis of GaN/sapphire heterostructures.
J Phys D Appl Phys, 34 (2001), pp. A35-A39
[20]
S. Zhou, M.F. Wu, S.D. Yao, J.P. Liu, H. Yang.
Chemical composition and elastic strain in AlInGaN quaternary films.
Thin Solid Films, 515 (2006), pp. 1429-1432
[21]
C. Kisielowski, J. Kruger, S. Ruvimov, T. Suski, J.W. Ager III, E. Jones, et al.
Strain-related phenomena in GaN thin films.
Phys Rev B, 54 (1996), pp. 17745-17753
[22]
A. Polian, M. Grimsditch, I. Grzegory.
Elastic constants of gallium nitride.
J Appl Phys, 79 (1996), pp. 3343-3344
[23]
M. Shur, R.F. Davis.
GaN-based materials and devices: growth, fabrication, characterization and performance.
World Scientific, (2004),
[24]
L.E. McNeil, M. Grimsditch, R.H. French.
Vibrational spectroscopy of aluminium nitride.
J Am Ceram Soc, 76 (1993), pp. 1132-1136
[25]
A.U. Sheleg, V.A. Savastenko.
Determination of elastic constants of hexagonal crystals from measured values of dynamic atomic displacements.
Izv Akad Nauk SSSR Neorg Mater, 15 (1979), pp. 1598-1602
[26]
S. Qu, S. Li, Y. Peng, X. Zhu, X. Hu, C. Wang, et al.
Influence of the growth temperature of AlN buffer on the quality and stress of GaN films grown on 6H-SiC substrate by MOVPE.
J Alloys Compd, 502 (2010), pp. 417-422
[27]
T. Metzger, R. Hopler, E. Born, O. Ambacher, M. Stutzmann, R. Stommer, et al.
Defect structure of epitaxial GaN films determined by transmission electron microscopy and triple-axis X-ray diffractometry.
Philos Magn A, 77 (1998), pp. 1013-1025
[28]
V. Grigorov, O. Angelov, D.D. Malinovska.
Structural properties of poly-Si thin films obtained by aluminum-induced crystallization in different atmospheres.
Nanostructures and advanced materials for applications in sensor, optoelectronic and photovoltaic technology, pp. 339-344
[29]
D. Liu, W.W. Lei, S. Qin, L.T. Hou, Z.W. Liu, W.L. Cui, et al.
Large scale synthesis of hexagonal corundum-type In2O3 by ball milling with enhanced lithium storage capabilities.
J Mater Chem A, 1 (2013), pp. 5274-5278
[30]
C.H. Yin, C. Liu, D.Y. Tao, Y.P. Zeng.
Annealing effect on structural and magnetic properties of Tb and Cr co-implanted AlGaN.
Front Mater Sci, 6 (2012), pp. 366-370
[31]
Y.D. Tian, J.C. Yan, Y. Zhang, Y.H. Zhang, X. Chen, Y.N. Guo, et al.
Formation and characteristics of AlGaN-based three-dimensional hexagonal nanopyramid semi-polar multiple quantum well.
Nanoscale, 821 (2016), pp. 11012-11018
[32]
S. Hernandez, R. Cusco, D. Pastor, L. Artus, K.P. O’Donnell, R.W. Martin, et al.
Raman-scattering study of the InGaN alloy over the whole composition range.
J Appl Phys, 98 (2005),
[33]
R. Lopez, R. Gomez.
Band-gap energy estimation from diffuse reflectance measurements on sol–gel and commercial TiO2: a comparative study.
J Sol–Gel Sci Technol, 61 (2012), pp. 1-7
[34]
S.M. Chang, R.A. Doong.
Interband transitions in sol–gel derived ZrO2 films under different calcination conditions.
Chem Mater, 19 (2007), pp. 4804-4810
[35]
J.J. Mares, J. Kristofik, E. Hulicius.
Influence of humidity on transport in porous silicon.
Thin Solid Films, 255 (1995), pp. 272-275
[36]
R.C. Erson, R.S. Muller, C.W. Tobias.
Investigations of porous silicon for vapor sensing.
Sens Actuators A, 23 (1990), pp. 835-839
[37]
S.J. Pearton, F. Ren.
Wide bandgap semiconductor one-dimensional nanostructures for applications in nanoelectronics and nanosensors.
Nanomater Nanotechnol, 3 (2013), pp. 1-15
[38]
Y.Y. Tsai, H.I. Chen, C.W. Hung, T.P. Chen, T.H. Tsai, K.Y. Chu, et al.
A hydrogen gas sensitive Pt-In0.5Al0.5O metal-semiconductor schottky diode.
J Electrochem Soc, 154 (2007), pp. J357-J361
[39]
F. Ren, S.J. Pearton.
Semiconductor device-based sensors for gas, chemical, and biomedical applications.
CRC Press, (2011),
[40]
S.H. Park, S.H. Kim, C.M. Lee.
Fabrication, structure, and hydrogen-gas-sensing properties of multiple networked GaN nanostructure gas sensors.
J Korean Phys Soc, 66 (2015), pp. 1062-1066
[41]
S. Karato.
The role of hydrogen in the electrical conductivity of the upper mantle.
Nature, 347 (1990), pp. 272-273
[42]
C.G.V. De Walle.
Hydrogen interactions with semiconductors and oxides.
Hydrogen in materials and vacuum systems, pp. 33-39
[43]
S. Pizzini.
Physical chemistry of semiconductor materials and processes.
John Wiley & Sons, (2015),
[44]
T. Hino, S. Tomiya, T. Miyajima, K. Yanashima, S. Hashimoto, M. Ikeda.
Characterization of threading dislocations in GaN epitaxial layers.
Appl Phys Lett, 76 (2000), pp. 3421-3423
Copyright © 2019. Brazilian Metallurgical, Materials and Mining Association
Journal of Materials Research and Technology

Subscribe to our newsletter

Article options
Tools
Cookies policy
To improve our services and products, we use cookies (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here.