Effect of Low-Energy Implantation of In⁺ Ions on the Composition and Electronic Structure of Single-Crystal GaP(111)

Abstract

Using a complex of secondary and photoelectron spectroscopy methods, the effects of the implantation of In⁺ ions with an energy of E₀ = 1 keV at different doses and subsequent annealing on the composition, electronic, and crystal structure of the GaP(111) surface were studied. It is shown that in the dose range D ≈ 5 × 10¹⁴–5 × 10¹⁵ cm⁻² after annealing, nanocrystalline phases Ga_0.6In_0.4P are formed with surface dimensions d ≈ 10–30 nm, and at D ≥ 6 × 10¹⁶ cm⁻² nanofilm–Ga_0.6In_0.4P with a thickness of 30–35 nm. It has been found that the band gap of nanophases (E_g ≈ 2–2.3 eV) is much larger than Eg of the film (~1.85 eV). For the first time, information was obtained on the density of state of electrons in the valence band of nanophases and nanofilm GaInP.

1. Introduction

Interest in nanosized structures grown on the basis of GaP is due to their wide use and their promise in the creation of a monolithic optoelectronic integrated circuit (MOEIC), photoelectric generators (solar cells), optical and electronic pumping lasers, and microwave technology devices [1,2,3,4,5,6]. Therefore, electronic, optical, emission, and electrophysical properties and the influence of various effects on these properties are currently being intensively studied. In particular, to create an MOEIC on a Si substrate, it will be necessary to grow an epitaxial film with a direct gap transition on its surface [1,7]. GaP lattice constants and Si practically coincide with each other, so it is possible to obtain a heteroepitaxial GaP/Si system.

Methods of molecular beam epitaxy [1,2,3,4,5,6,7,8,9], liquid phase and vapor phase epitaxy [10,11,12,13,14,15,16], MOCVD [5,17,18,19], and high-energy ion implantation [20,21,22] are widely used to obtain three- and four-component layers on the GaP surface. In recent years, low-energy ion implantation is often used to obtain nanosized structures on the surface and in the near-surface region of materials [23,24,25,26,27,28,29,30,31,32].

In [20], GaAlA nanophases and nanofilms were obtained by implanting Al ions into GaP(111) and GaAs(111). It has been established that in the case of GaAlP nanophases with surface sizes less than 35–40 nm and a thickness of 3.5–4 nm, quantum size effects appear. However, to date, nanoscale structures of the GaInP type have not been obtained by low-energy ion implantation.

The results of such studies are very important for elucidating the basic mechanisms of the formation of nanofilm heteroepitaxial systems based on A³B⁵ binary semiconductors, as well as for the development of new opto- and nanoelectronic devices.

In this work, the composition, electronic, and crystal structure of GaInP nanostructures obtained by implanting In⁺ ions into GaP(111) with E₀ = 1 keV with different doses of D were studied for the first time. For comparison, the surface morphology of the GaInP/GaP(111) MBE film was also studied.

2. Experimental Method

Standard single-crystal samples of n-type GaP(111) (n = 3∙1017 cm⁻³) grown by the Czochralski method with a diameter of ~10 mm and a thickness of 1 mm were chosen as objects of study (samples were obtained from the Institute of Crystallography, Russia). Before ion implantation, GaP(111) was degassed under ultra-high vacuum conditions (P = 10⁻⁷ Pa) at T = 900 K for ~4 h. An ultra-high vacuum in the device was created by an oil-free pumping system: two zeolite sorption pumps and a cooled titanium getter pump. The experimental device consists of two chambers. In the first chamber, all technological operations (thermal heating, ion implantation) were carried out, and in the second chamber, the composition and electronic structure of the materials under research were studied using the methods of Auger electron spectroscopy (AES) (A.F.Ioffe Physical-Technical Institute, Russia) and ultraviolet photoelectron spectroscopy (UVES) (UV-1280, Shimadzu, Japan). The spectra of Auger electrons and photoelectrons were recorded using a four-grid electron analyzer. To determine the depth distribution profile of the atoms, a layer-by-layer Auger analysis was carried out by sputtering the sample surface with Ar⁺ ions with an energy of 1 keV at an incidence angle of ∼80–85° relative to the normal; the etching rate was ∼(5 ± 1) Å/min. Ultraviolet photoelectron spectra were recorded at photon energy hν ≈ 10.8 eV. The photon source was a standard gas-discharge hydrogen lamp. Studies of the morphology of the crystalline structure of the surface of the samples were carried out in standard installations of scanning electron microscopy (SEM) (JSM-7200F), atomic force microscopy AFM (XE-100), and high-energy electron diffraction RHEED (EMR-2). The width of the forbidden zone was determined by a UV-1280 spectrophotometer.

3. Experimental Results

Figure 1 shows the initial part of the Auger spectra of pure GaP(111) and GaP implanted with In⁺ ions with an energy of 1 keV at D = D_sat ≈ 6∙10¹⁶ cm⁻² before and after heating at a temperature T = 950 K (where Dsat—In ion saturation dose). Warm-up time ~30 min. It can be seen that after implantation of In⁺ ions, the intensity of the Ga and P Auger peaks in the spectrum decreases significantly and intense peaks characteristic of In appear. Ion-implanted layers are highly disordered up to amorphization. Calculations based on the full spectrum of Auger electrons of an ion-implanted sample using the formula [25] ${\mathrm{C}}_{\mathrm{x}}={\displaystyle \frac{{\mathrm{I}}_{\mathrm{x}}/{\mathrm{S}}_{\mathrm{x}}}{\sum {\mathrm{I}}_{\mathrm{i}}/{\mathrm{S}}_{\mathrm{i}}}}$ showed that the surface concentrations of In, Ga, and P are, respectively, 40–45 at.%, 30–35 at.%, and 25–30 at.%. After heating this system at T = 950 K, the surface concentration of these atoms changes sharply: C_Ga ≈ 28–33 at.%, C_In ≈ 20–22 at.%, C_P ≈ 48–52 at.%, i.e., a single-crystal film with an approximate composition of Ga_0.6In_0.4P with a thickness of ~30–35 Å was formed.

Figure 2 shows the concentration profiles of the distribution of In atoms over the depth h′ for GaP(111) implanted with In⁺ ions with E₀ = 1 keV at D = 6 × 10¹⁶ cm⁻², obtained before and after heating. It can be seen that after ion implantation, the surface concentration of In atoms is ~40–45 at.%. With an increase in h′ to ~8–10 Å, the C_In value sharply decreases by ~8–10 at.%. In the range h ≈ 10–25 nm, the In concentration does not change noticeably, and in the range h′ ≈ 25–80 Å, it decreases to a bullet. After heating at T = 950 K, the value of C_In on the surface decreases to 20–22 at.%, and it practically does not change until a depth of 35–40 Å. In the range h′ ≈ 40–100 Å, C_In monotonically decreases to 0, i.e., after heating, the concentration of atoms in the transition layer increases slightly.

Figure 1. Auger spectra: 1—pure GaP(111); 2—GaP(111), implanted with In⁺ ions with E₀ = 1 keV at D = 6 × 10¹⁶ cm⁻²; 3—after annealing at T = 950 K. Annealing time 40 min.

Figure 1. Auger spectra: 1—pure GaP(111); 2—GaP(111), implanted with In⁺ ions with E₀ = 1 keV at D = 6 × 10¹⁶ cm⁻²; 3—after annealing at T = 950 K. Annealing time 40 min.

A comparison of the areas under the C_In(h′) curves shows that after heating it noticeably decreases. Apparently, during the heating process, along with some diffusion of In atoms into the depths of GaP, their evaporation from the surface layers occurs.

Figure 2. Profiles of the distribution of In atoms over the depth h′ for GaP(111) implanted with In⁺ ions with E₀ = 1 keV at D = 6 × 10¹⁶ cm⁻² before (curve 1) and after heating at T = 950 K (curve 2).

Figure 2. Profiles of the distribution of In atoms over the depth h′ for GaP(111) implanted with In⁺ ions with E₀ = 1 keV at D = 6 × 10¹⁶ cm⁻² before (curve 1) and after heating at T = 950 K (curve 2).

Figure 3 shows SEM and AFM images of the surface of pure GaP(111) and GaP(111) with Ga_0.6In_0.4P surface films, obtained by MBE (h = 50 nm) and ion implantation (h = 0.35–0.4 nm) methods. From Figure 3, it is clear that the surface of single-crystal GaP(111) is atomically smooth. The estimated value of surface roughness, which was determined by AFM, is ~1 nm. Analysis of picture 3b shows that the MBE of the Ga_0.6In_0.4P film consists of individual blocks. In this case, the surface roughness increases to 5–6 nm (Figure 3d). The surface of Ga_0.6In_0.4P obtained by ion implantation in combination with heating is highly smooth, and the roughness is ~2 nm (Figure 3e).

Figure 4 shows HSED images of the surface of a pure GaP and GaP sample with a Ga_0.6In_0.4P film with a thickness of 3–3.5 nm. The images clearly reveal reflections characteristic of cubic lattices. However, in the case of a three-component film, the reflections are bolder and blurrier. This is apparently due to some discrepancy between the lattice parameters of GaP and GaInP (see Table 1).

Figure 5 shows the photoelectron spectra obtained after heating at T = 950 K (heating time 40 min) of GaP implanted with In⁺ ions with E₀ = 1 keV at doses of 10¹⁵ (curve 2) and 6 × 10¹⁶ cm⁻² (curve 3) [29,32]. In the first case, nanocrystalline phases with surface diameters d = 15–20 nm were formed, and in the second case, a Ga_0.6In_0.4P film was formed. These spectra well reflect the distribution of the density of the state of the valence band electrons. It can be seen that the spectrum of pure GaP exhibits maxima at binding energies of E_cb ≈–0.8 eV; –2.2 eV; and 4 eV, probably due to the excitation of electrons from the 4p and 4p + 4s states of Ga, as well as the hybridization of the 4s states of Ga with the 3s states of P (Figure 5, curve 1) [29,32]. In the case of the Ga_0.6In_0.4P film, the spectrum (Figure 5, curve 3) contains intense peaks with E_cb = −1.2; −3.3; and −5.6 eV, apparently associated with the excitation of electrons from the hybridized states of electrons 4s(Ga) + 5p(In), 4s(Ga) + 5p(In) + 3d(P), and 4s (Ga) + 5s(In) + 3d(P). The spectrum of GaP with Ga_0.6In_0.4P nanocrystals reveals some features characteristic of both GaP and Ga_0.6In_0.4P (Figure 5, curve 2).

To estimate the band gap of these samples, light absorption spectra were recorded (Figure 6), where K is the light absorption coefficient. It is known that at medium doses (5 × 10¹⁴–5∙10¹⁵ cm⁻²), the ion-implanted surface consists of two phases: the ion-doped phase and the undoped portions of the semiconductor [29]. In our case, at D = 10¹⁵ cm⁻², the average surface diameters of the nanophases were d ~ 15–20 nm, and at D = 6 × 10¹⁶ cm⁻² they were 25-30 nm. The distance between the centers of the phases is ~50–60 nm. In the case of D ≈ 6 × 10¹⁶ cm⁻², after heating, a continuous homogeneous layer (film) of Ga_0.6In_0.4P with h = 35–40 Å is formed. Figure 6 shows that in all dependences K(hν), first with increasing hν, the value of K practically does not change, and then K sharply decreases, where

K = I₀/I_n

I₀ is the intensity of light passing through GaP, and I_n is the intensity of light passing through GaP with a GaInP nanostructure.

Figure 6. Spectra of light transmission through GaP(111) (curve 1) and GaP(111) with nanophases with surface diameters d ≈ 15–20 nm (2), d = 25–30 nm (3), and Ga_0.6In_0.4P nanofilm (4).

Figure 6. Spectra of light transmission through GaP(111) (curve 1) and GaP(111) with nanophases with surface diameters d ≈ 15–20 nm (2), d = 25–30 nm (3), and Ga_0.6In_0.4P nanofilm (4).

In the case of GaP and GaP with a Ga_0.6In_0.4P film, a decrease in K is observed starting from hν ≈ 2.2 eV (curve 1) and hν ≈ 1.7 eV (curve 4), respectively, and in the case of GaP with Ga_0.6In_0.4P nanocrystals, a sharp decrease in K is observed in two values of hν [33]. The first decrease is associated with light absorption in areas coated with GaInP nanocrystals, and the second-in areas of GaP are not coated with nanocrystals. The extrapolation of a sharply decreasing part of the curves to the hν axis gives a value of E_g for a given material. It can be seen that E_g for GaP(111) is approximately equal to 2.36 eV, for a Ga_0.6In_0.4P film ~1.85 eV, and for nanocrystalline Ga_0.6In_0.4P phases with d = 25 nm ~2.3 eV and d = 25–30 nm ~2.3 eV. With increasing surface diameters d nanotimes, the K(hν) curves shift towards lower energies, i.e., E_g decreases (Table 1). Based on curve 2 (Figure 6), one can estimate the degree of coverage of the GaP(111) surface with GaInP Ѳ nanophases using the following formula [33]:

Ѳ = 1 − K.

In the case of GaP with GaInP nanocrystals, the value of Ѳ ≈ 0.40. Table 1 shows the values of E_g and Ѳ for the samples under study.

Table 1. Dimensions of nanostructures (thickness h and surface diameter d) and values of E_g and Ѳ of nanophases and GaInP films.

Studied Samples	h, nm	d, nm	Eg, eV	Ѳ, %
GaP(111)	Monocrystal sample		2.36	0
Ga0.6In0.4P/GaP(111)	3–3.5	15–20	2.3	40
	3–3.5	25–30	2.0	70
	3–3.5	continuous film	1.85	100

Table 1. Dimensions of nanostructures (thickness h and surface diameter d) and values of E_g and Ѳ of nanophases and GaInP films.

Studied Samples	h, nm	d, nm	Eg, eV	Ѳ, %
GaP(111)	Monocrystal sample		2.36	0
Ga0.6In0.4P/GaP(111)	3–3.5	15–20	2.3	40
	3–3.5	25–30	2.0	70
	3–3.5	continuous film	1.85	100

From Figure 5 and the table, it can be seen that, at the same thicknesses, the values of E_g for nanocrystalline phases are greater than for the film. These results show that in nanocrystalline phases, starting from a diameter of ≈ 25–30 nm, the value of E_g increases with decreasing d.

4. Conclusions

For the first time, nanocrystalline phases and Ga_0.6In_0.4P films with a thickness of 30–35 Å were obtained by the implantation of In⁺ ions in combination with annealing on the GaP(111) surface. Their parameters of energy bands and density of valence band electrons, as well as the degree of coverage of the GaP surface with Ga_0.6In_0.4P nanophases, were determined. It has been shown that in GaInP nanocrystals, the E_g value is greater than the E_g of the nanofilm.