Structural, electric, and optical properties of MgGa2Se4 epilayers grown by hot wall epitaxy method

Structural, electric, and optical properties of MgGa2Se4 epilayers grown by hot wall epitaxy method

Journal of Crystal Growth 361 (2012) 142–146 Contents lists available at SciVerse ScienceDirect Journal of Crystal Growth journal homepage: www.else...

651KB Sizes 0 Downloads 48 Views

Journal of Crystal Growth 361 (2012) 142–146

Contents lists available at SciVerse ScienceDirect

Journal of Crystal Growth journal homepage: www.elsevier.com/locate/jcrysgro

Structural, electric, and optical properties of MgGa2Se4 epilayers grown by hot wall epitaxy method S.H. You a, K.J. Hong a,n, T.S. Jeong b, C.J. Youn b a b

Department of Physics, Chosun University, Gwangju 501-759, South Korea School of Semiconductor and Chemical Engineering, Semiconductor Physics Research Center (SPRC), Chonbuk National University, Jeonju 561-756, South Korea

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 January 2012 Received in revised form 27 July 2012 Accepted 18 September 2012 Communicated by: J.B. Mullin Available online 29 September 2012

The epilayer growth of the MgGa2Se4 compounds was successfully achieved through the hot wall epitaxy method. The grown layer was accumulated along the o 116 4 direction onto the GaAs(100) substrate. From the Hall effect measurement, the mobility was determined to be 264 cm2/Vs at 293 K. At a high temperature range (T 4150 K), it tended to decrease as a function of T  3/2 by increasing the temperature, and increase as a function of T3/2 at the low-temperature range (To 100 K). In the photocurrent (PC) measurement, we observed the A, B, and C peaks corresponding to 529.9 (2.3398 eV), 495.2 (2.5037 eV), and 477.6 nm (2.5960 eV) at 10 K, respectively. Three peaks of A, B, and C were caused by the band-to-band transitions from the valence band state of G4(z), G5(x), and G5(y) to the conduction band state of G1(s), respectively. By comparing the results of PC and absorption, the temperature dependence of the optical bandgap energy was well interpreted using Varshni’s relation Eg(T) ¼2.3412  8.87  10  4 T2/(T þ251). & 2012 Elsevier B.V. All rights reserved.

Keywords: A1. Characterization A3. Hot wall epitaxy B1. MgGa2Se4 B2. Semiconducting ternary compounds

1. Introduction VI The AII–BIII 2 –C4 materials are attractive materials because of their applicability for solar cells, light emitting diodes, and photoconductive devices [1–3]. Magnesium gallium selenide (MgGa2Se4) is a new material satisfying such a requirement. Thus, in order to realize these applications, it is of primary importance to grow high quality epilayers and characterize the fundamental properties. But, it has been reported that the bulk of MgGa2Se4 crystals grew through the Bridgman method, and its characterization investigated through absorption and photoluminescence [4,5]. Therefore, the information on MgGa2Se4 is still very limited. MgGa2Se4 is a photoconductive material having the bandgap energy of 2.2 eV at room temperature. Therefore, it is important to investigate the conductivity change of photoconductive MgGa2Se4 caused by incident radiation. Thus, the photocurrent (PC) spectroscopic measurement had been studied for applications in photodetection and radiation measurements [6]. But, the PC-measurement analysis of MgGa2Se4 has not yet been reported. In PC measurements, absorbed photons with higher energy than the bandgap energy create electron and hole carriers in the conduction and valence bands, respectively. These carriers instantly flow out through the electrodes. Thereby, the obtained PC peak corresponds to the direct bandgap energy.

n

Corresponding author. Tel.: þ82 62 230 6637; fax: þ82 62 234 4326. E-mail address: [email protected] (K.J. Hong).

0022-0248/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jcrysgro.2012.09.030

In this study, photoconductive MgGa2Se4 epilayers were first grown by the hot wall epitaxy (HWE) method. The grown MgGa2Se4 layers were investigated by high-resolution X-ray diffraction (HRXRD), atomic force microscopy (AFM), Hall effect, PC, and absorption spectroscopy. From these results, we discussed the structural, electric, and optical properties of the MgGa2Se4 epilayers. In addition, the bandgap energy obtained from the PC peak position was discussed as a function of temperature together with that of the absorption experiment.

2. Experimental procedures Prior to the MgGa2Se4 layer, MgGa2Se4 polycrystalline was formed. The starting materials were 6 N purity shot-types of Mg, Ga, and Se. After the materials were weighed in stoichiometric proportions, they were sealed in a quartz tube for maintaining the vacuum atmosphere. At this time, the horizontal synthesis furnace was used for the formation of the MgGa2Se4 polycrystalline. The sealed ampoule was placed in the synthesis furnace and was continually rotated at a rate of 1 rpm. In order to avoid an explosion of the ampoule due to the selenium vapor pressure, the temperature of the ampoule was increased gradually to 1000 1C, which was then maintained for 48 h. Fig. 1 presents a schematic diagram of the HWE apparatus used for the MgGa2Se4 layer. The MgGa2Se4 layers were grown on semi-insulating GaAs (100). The GaAs substrate was cleaned ultrasonically for 1 min in successive baths of trichloroethylene, acetone, methanol, and

S.H. You et al. / Journal of Crystal Growth 361 (2012) 142–146

143

HRXRD intensity (arb. units)

XRD intensity (arb. units)

100

In order to grow the MgGa2Se4 layer, an ingot of synthesized MgGa2Se4 polycrystalline was used as a source material of HWE. At this time, the source temperature was fixed at 610 1C, which was obtained through experimental repetition. Thus, the MgGa2Se4 layers were grown by changing the substrate temperature from 360 to 420 1C. Fig. 2 presents the HRXRD rocking curves on the MgGa2Se4 layers grown as a function of the substrate temperature. The observation of this HRXRD rocking curves indicate that the grown layers have a high crystalline quality. As shown in Fig. 2, the HRXRD intensity of the layer on the 400 1C substrate temperature had the highest than that of the measured other samples. Fig. 3 displays the intensity and FWHM value of

(116)

MgGa2 Se4 45

55

65

75

400 °C 420 °C 380 °C 360 °C

-500 0 500 Rocking angle (arcsec)

1000

FWHM (arcsec)

Fig. 2. HRXRD rocking curves on the MgGa2Se4 layers grown as a function of the substrate temperature. Here, the subfigure shows the XRD curves on the specimen grown at the substrate of 400 1C.

500

10

400

8

300

6

200

4

100

2

0 340

360 380 400 420 Substrate temperature (°C)

440

Intensity ( x103, arb. units)

3.1. Growth and structural property

GaAs

Diffraction angle (2θ)

Fig. 1. Schematic diagram of the HWE apparatus used for the MgGa2Se4 layer.

3. Results and discussion

(400)

50

0 35

-1000

2-propanol and etched for 1 min in a solution of H2SO4:H2O2:H2O (5:1:1). The substrate was degreased in organic solvents and rinsed with deionized water (18.2 MO). After the substrate was dried off, the substrate was immediately loaded onto the substrate holder in Fig. 1 and was annealed at 580 1C for 20 min to remove the residual oxide on the surface of the substrate. To find the optimum growth conditions, the grown MgGa2Se4 layers were analyzed by HRXRD measurements. The thickness and stoichiometric composition of the MgGa2Se4 layers were measured by using an a-step profilometer and an energy dispersive X-ray spectrometer (EDS), respectively. Also, the electric properties were achieved by Hall effect measurement of the van der Pauw method with various temperatures. In order to take PC measurements, two Au electrodes with a coplanar geometry on MgGa2Se4 were fabricated by an e-beam evaporator and an Ohmic contact of electrodes was confirmed by the current–voltage measurement. After the electrodes were connected to a wire, the sample was mounted on the holder of a low-temperature cryostat. The PC spectrum measurement was done while monochromatic light emitted from a halogen lamp passed through a chopper to illuminate the sample. Thus, the optical absorption measurement was performed with a UV–vis–NIR spectrophotometer (Hitachi, U-3501) for a range of 400 to 800 nm. At this time, the PC and absorption experiments were measured while varying the temperature from 10 to 293 K.

400 °C

0

Fig. 3. Intensity and FWHM value of the HRXRD rocking curves on the MgGa2Se4 layers grown as a function of the substrate temperature.

the HRXRD rocking curves on the MgGa2Se4 layers grown as a function of the substrate temperature. As shown in Fig. 3, the intensity of the HRXRD curves increased with an increase in the substrate temperature. However, its intensity tended to decrease after an increase in the substrate temperature of 400 1C. On the contrary, the FWHM on the HRXRD curves decreased with an increasing substrate temperature. Then, it again increased after having a minimum value of 117 arcsec at 400 1C. These results indicate that the optimum temperature of the substrate is 400 1C. With a source temperature of 610 1C, the most suitable substrate temperature for growth turned out to be 400 1C. Thus, the MgGa2Se4 layers grown under these optimum conditions were obtained with a thickness of 2.8 mm and a growth rate of ˚ 1.39 A/sec. On the other hand, to confirm the orientation of the MgGa2Se4 layer, XRD analysis was used. The subfigure in Fig. 2 shows the XRD curves on the specimen grown on the substrate of 400 1C. As shown in subfigure of Fig. 2, two dominant peaks were observed. These correspond to the diffraction peaks of the MgGa2Se4 (116) and the GaAs (400) plane. The intensity of the

144

S.H. You et al. / Journal of Crystal Growth 361 (2012) 142–146

3.2. Temperature dependence of the Hall mobility

Table 1 Composition ratios of each element on the synthesized polycrystalline and the layer analyzed through the EDS measurement. Element

Mg Ga Se

Synthesized polycrystalline Before

After

14.29 28.57 57.14

14.65 28.64 56.71

Layer growth

14.83 28.55 56.62

Units: (%)

MgGa2Se4 (116) peak, located at a two theta of 47.7491, is very sharp and dominant. This peak position indicates that the grown MgGa2Se4 layer has a rhombohedral structure. Then, the orientation of the MgGa2Se4 layer grown on the GaAs (100) substrate was converted to the (116) plane. This phenomenon had been observed in the CdTe epilayer grown on the GaAs (100) too. Faurie et al. [7] reported that the growth of the CdTe (100) or (111) planes on GaAs (100) was possible by controlling the annealing temperatures and times of the substrate. So, we suggest that the orientation of the grown MgGa2Se4 (116) layers relate to the preannealing process used to remove the residual oxide on the surface of the substrate. Thus, no other peak beyond the MgGa2Se4 (116) plane was observed in the range of the XRD measurement. Therefore, the observation of the only (116) peak indicates that the MgGa2Se4 was grown epitaxially along the o1164 direction onto the GaAs (100) substrate. Fig. 4 shows the AFM surface-morphology of the epilayer grown below optimum growth conditions. As shown in Fig. 4, the surface morphology of the grown epilayer was smooth. The root-mean-square value roughness, Rms, of the surface was 0.8 nm. It suggests that the smooth surface and the low Rms value are attributed to low growth rate. The growth rate of MgGa2Se4 epilayer was actually a ˚ low value of 1.39 A/sec. Such low growth rate is one of the many benefits of having the HWE method. Thus, the stoichiometric composition on the MgGa2Se4 epilayers was measured using the EDS analysis. Table 1 presents the composition ratios of each element on the synthesized polycrystalline and the layer analyzed through the EDS measurement. As shown in Table 1, the rate of the Mg components increased a little bit, while that of the Se components decreased. But, it suggests that the component ratios of initial mole fraction were continuously maintained during the layer growth. The result of EDS indicates that the grown MgGa2Se4 epilayers were formed into the stoichiometric composition.

3.3. Bandgap investigation through PC and optical absorption spectra For photosensitive material, the volume lifetime is much larger than the surface lifetime; therefore, a maximum in the spectral response occurs when a transition is made from surface excitation (exciting photon energy greater than band gap) to volume excitation (exciting photon energy smaller than band gap) with increasing wavelength [10]. In the PC experiment, the absorbed photons with higher energy than the bandgap energy, create electron and hole carriers. If an external electric field is applied, the electrons and holes move in opposite directions. These carriers instantly flow out through the electrodes, and produce PC signals. These PC signals correspond to the direct bandgap energy. Therefore, it indicates that the maximum in the spectral

104

Mobility (cm2/ V sec)

Fig. 4. AFM surface-morphology of the epilayer grown below optimum growth conditions.

Hall effect measurements on the MgGa2Se4 epilayers were carried out by varying the temperature from 30 to 293 K. The measured mobility and carrier density at 293 K were 264 cm2/Vs and 6.65  1018 cm  3, respectively. Fig. 5 shows the temperature dependence of the Hall mobility. The mobility tended to increase during the lowering of temperature to 150 K, after which it decreased. As shown in Fig. 5, the mobility at the high temperature range (T4150 K) tended to decrease as a function of T  3/2 by increasing the temperature, and increase as a function of T3/2 at the low-temperature range (To100 K). This indicates that scattering at the high-temperature range is mainly due to the acoustic mode of lattice vibrations through a deformation potential and that scattering at the low-temperature range is most pronounced due to the impurity effect [8]. From the results of the Hall coefficient value, the grown MgGa2Se4 epilayer was confirmed to be of n-type. The grown sample was always n-type presumably due to slight stoichiometric deviations originating from an excess of selenium vacancies (VSe). At this time, the activation energy caused by these native defects was extracted out at 61.8 meV, from the relation between the reciprocal temperature and the carrier concentration. Kim et al. [9] reported a shallow donor level located at 0.07 eV below the bottom of the conduction band using photoluminescence spectroscopy. Therefore, our value corresponds to the shallow donor level of 61.8 meV, due to the VSe below the edge of the conduction band.

103 T3/2

102 Impurity scattering

101 1 10

T-3/2

Lattice scattering

102 Temperature (K)

103

Fig. 5. Temperature dependence of the Hall mobility.

S.H. You et al. / Journal of Crystal Growth 361 (2012) 142–146

PC response are the intrinsic transition caused by the band-toband transition [11,12]. Fig. 6 shows the PC response of the photoconductive MgGa2Se4 epilayer obtained at temperatures ranging from 10 to 293 K. As shown in Fig. 6, two peaks having strong intensity appeared at the temperature ranges between 100 and 293 K. However, in the PC-response spectrum, three peaks observed at 10 K (Infact, the 30–77 K spectra are not in Fig. 6. But the 50 and 70 K spectra were observed to be two peaks, too, and the appearance of three peaks started from 30 K.). Two peaks appeared at 524.2 (2.3652 eV) and 563.3 nm (2.2011 eV) in the PC spectrum of 293 K. The peak at 2.2011 eV corresponds to the band-to-band electron transition from G4(z) of the valence band to G1(s) of the conduction band. This peak is labeled as peak A. The peak at 2.3653 eV corresponds to the transition from G5(x) of the valence band to G1(s) of the conduction band, and is labeled peak B. Three peaks were observed in the PC spectrum in 10 K. The peaks at 529.9 nm (2.3398 eV) and 495.2 nm (2.5037 eV) among the 10 K PC peaks are associated with peaks A and B, respectively. And the peak at 477.6 nm (2.5960 eV) is attributed to the transition of electrons from G5(y) of the valence band to G1(s) of the conduction band, which is peak C. In order to identify the bandgap energy for MgGa2Se4, the optical absorption measurement, which is one of the traditional methods of the bandgap measurement, was conducted. Fig. 7 shows the optical absorption spectra obtained in the temperature range of 10 K to 293 K. From the optical absorption spectra, the relation between the optical absorption coefficient (a) and the incident photon energy (hn) was examined. The relation for a

145

direct bandgap between hn and a is given as ðahnÞ2  ðhnEg Þ,

ð2Þ

where h is the Planck constant. According to Eq. (2), (ahn)2 linearly depends upon the photon energy. Fig. 8 shows the plots of (ahn)2 versus photon energy for different temperatures. As shown in Fig. 8, the bandgap energy determined by extrapolating the linear portions of the respective curves to (ahn)2 ¼0. Fig. 9 displays the variation of the bandgap energy as a function of temperature on the MgGa2Se4 epilayer obtained from PC and absorption measurements. We have found that the PC peak positions are consistent with the absorption calculated by using Eq. (2) at the same temperature. Therefore, this figure shows that the variation of the bandgap energy as a function of temperature on the MgGa2Se4 epilayer obtained from PC and absorption measurements is in good agreement. In addition, this shows that the PC measurement is a useful method for the bandgap determination of the MgGa2Se4 epilayer. In fact, the absorption experiment has been known to be inaccurate for obtaining the bandgap energy because of the difficulty in defining the position of the absorption edge. This discrepancy comes from the fact that the bandgap energy is obtained by fitting the experimental spectra to theoretical models. The PC measurement can directly give the energy corresponding to the bandgap by measuring the PC peak position. As shown in Fig. 9, the energy gap variations of the absorption, peaks A, B, and C, show nonlinear relationships. The temperature dependence of the bandgap energy is well expressed

Fig. 8. Plots of (ahn)2 versus photon energy for different temperatures.

Fig. 6. PC response of the photoconductive MgGa2Se4 epilayer obtained at temperatures ranging from 10 to 293 K.

Fig. 7. Optical absorption spectra obtained in the temperature range of 10 K to 293 K.

Fig. 9. Variation of the bandgap energy as a function of temperature on the MgGa2Se4 epilayer obtained from PC and absorption measurements.

146

S.H. You et al. / Journal of Crystal Growth 361 (2012) 142–146

by the Varshni’s formula [13,14] 2

Eg ðTÞ ¼ Eg ð0ÞaT =ðT þbÞ,

ð3Þ

where a and b are constants. When a and b are taken to be 8.87  10  4 eV/K and 251 K, respectively, the curve plotted by Eq. (3) closely fits the experimental values, as shown in Fig. 9. In addition, Eg(0) is the bandgap energy at 0 K. Therefore, it is estimated to be 2.3412, 2.5047, and 2.5967 eV for the transitions corresponding to peaks A, B, and C, respectively. The bandgap energy fitted by Eq. (3) was 2.2012 eV at 293 K. Through the optical absorption measurement using the bulk crystal, Kim et al. [4] reported that the values of the room-temperature bandgap energy, Eg(0), a, and b were 2.20 eV, 2.34 eV, 8.79  10  4 eV/K, and 250 K, respectively. Our obtained parameters agree well with these values. Thus, a splitting of the valence band into three subbands restricted by the selection rule was investigated through the PC spectroscopy measurement. Consequently, three subbands in the valence band of G4(z), G5(x), and G5(y) from G1(s) were extracted out to be 2.3412, 2.5047, and 2.5967 eV, respectively.

4. Conclusions The MgGa2Se4 epilayers were successfully grown on GaAs substrates through the HWE method. The optimum temperatures for the growth turned out to be 400 1C for the substrate and 610 1C for the source. Thus, the grown epilayer was accumulated along the o116 4 direction onto the GaAs(100) substrate. From the Hall effect measurement, the mobility was determined to be 264 cm2/Vs at 293 K. At a high temperature range (T4150 K), it tended to decrease as a function of T  3/2 by increasing the temperature, and increase as a function of T3/2 at the lowtemperature range (To100 K). Also, from the relation between the reciprocal temperature and the carrier concentration, we extracted out that the shallow donor level located at 61.8 meV below the edge of the conduction band. In the PC measurement, we observed the A, B, and C peaks corresponding to 529.9 (2.3398 eV), 495.2 (2.5037 eV), and 477.6 nm (2.5960 eV) at

10 K, respectively. The three peaks of A, B, and C were caused by the band-to-band transitions from the valence band state of G4(z), G5(x), and G5(y) to the conduction band state of G1(s), respectively. At a temperature of absolute zero, they were found to be 2.3412, 2.5047, and 2.5967 eV, respectively. Also, the bandgap energies obtained from PC and absorption measurements were in good agreement at each temperature. Consequently, the optical bandgap variation was well interpreted by using Varshni’s formula: Eg(T)¼ 2.3412 aT2/(Tþ b), where a and b were 8.87  10  4 eV/K and 251 K, respectively.

Acknowledgment This study was supported by research funds from Chosun University, 2011.

References [1] S.I. Radautsan, A.N. Georgobiani, I.M. Tiginyanu, Progress in Crystal Growth and Characterization 10 (1984) 403. [2] O. Vigil, O. Calzadilla, D. Seuret, J. Vidal, Solar Energy Materials 10 (1984) 139. [3] E. Fotin, F. Raga, Solid State Communications 14 (1974) 847. [4] Hyung-Gon Kim, Wha-Tek Kim, Yong-Geun Kim, Physical Review B 38 (1988) 9469. [5] Hyung-Gon Kim, Chan-Dae Kim, Wha-Tek Kim, Hong-Lee Park, Hyun-Nam Kim, Solid State Communications 72 (1989) 905. [6] N.V. Joshi, Photoconductivity: Art, Science, and Technology, Marcel Dekker, New York, 1990. [7] J.P. Faurie, C. Hsu, S. Sivananthan, X. Chu, Surface Science 168 (1986) 473. [8] S.M. Sze, Semiconductor Devices Physics and Technology, Wiley, New York, 1985. [9] Hyung-Gon Kim, Kwang-Ho Park, Bok-Nam Park, Han-Jo Lim, Suk-Ki Min, Hong-Lee Park, Wha-Tek Kim, Japanese Journal of Applied Physics 32 (Suppl. 3) (1993) 476. [10] R.H. Bube, Photoconductivity of Solids, Wiley, New York, 1969. [11] S.H. You, K.J. Hong, T.S. Jeong, C.J. Youn, J.S. Park, D.C. Shin, J.D. Moon, Journal of Crystal Growth 256 (2003) 116. [12] C.C. Klick, Physical Review 89 (1953) 274. [13] Y.P. Varshni, Physica 34 (1967) 149. [14] K.J. Hong, T.S. Jeong, C.J. Yoon, Y.J. Shin, Journal of Crystal Growth 218 (2000) 19.