Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Study of the free carrier characteristics and surface morphology of AlGaAs/GaAs thin films deposited using MOCVD

Open Access Open Access

Abstract

Ultra-low loss optical thin films find broad applications in fields such as vertical-cavity surface-emitting lasers and optical atomic clocks. The main optical losses in AlGaAs/GaAs distributed Bragg reflectors (DBRs) prepared using metal-organic chemical vapor deposition (MOCVD) arise from absorption loss caused by free carriers within the layers and scattering loss caused by surface roughness. In this study, we fabricated AlGaAs and GaAs single-layer thin films with varying Al compositions on substrates of three crystal orientations and under different V/III ratios. The dependence of carrier concentration and surface morphology on different substrates and growth conditions was investigated. Thin films grown on substrates with three different crystal orientations exhibited three distinct growth modes (step-flow mode, SK mode, and FM mode). The impact of the V/III ratio on the growth mode was found to be complex. Higher V/III ratios resulted in poorer morphology for films grown on (100) substrates, while better morphology was observed on (211) B substrates. Furthermore, the surface morphology of films grown on (100) 15° off substrates showed less sensitivity to changes in the V/III ratio. With increasing Al composition, the carrier concentration of the films significantly increased. Elevating the V/III ratio proved effective in suppressing the incorporation of carbon, thereby reducing the carrier concentration of AlGaAs films. GaAs films exhibited a low carrier concentration at an appropriate V/III ratio. Additionally, the distinct abilities of different substrates to adsorb impurities exerted a significant impact on the carrier concentration of the films. This study demonstrates that, under optimal conditions, it is feasible to fabricate AlGaAs/GaAs Bragg mirrors with low carrier concentration and relatively small roughness on (100) 15° off substrates.

© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement

1. Introduction

The distributed Bragg reflectors (DBRs), composed of AlGaAs/GaAs materials, constitute pivotal components in optoelectronic devices, including semiconductor saturable absorber mirrors (SESAMs) [1] and vertical cavity surface emitting lasers (VCSELs). Beyond these applications, they have garnered considerable attention in diverse fields such as precision metrology, high-power lasers, and mid-infrared spectroscopy. Currently, internationally adopted technologies for fabricating AlGaAs/GaAs distributed Bragg reflectors predominantly include molecular beam epitaxy (MBE) and metal-organic chemical vapor deposition (MOCVD), both demonstrating notable potential [24]. In comparison to MBE technology, MOCVD offers advantages such as a higher growth rate and lower cost for thin film preparation. However, it also presents drawbacks, including increased absorption and scattering losses in the DBRs. The total scattering losses of DBRs primarily stem from initial substrate roughness, refractive index inhomogeneity, and local defects. Scattering imposes limitations on the energy enhancement of a system's resonant cavity, thereby reducing reflectance [5]. Effectively reducing the total scattering loss of GaAs/AlGaAs films relies on controlling epitaxial conditions to achieve a smaller surface roughness, such as modifying the partial pressure of arsine [6].While some prior studies have delved into the factors and mechanisms influencing surface morphology, establishing models to describe the formation of diverse surface morphologies, the majority of these experiments were conducted on (100) surfaces and their adjacent surfaces with misorientation angles less than 2°. Limited research exists on growing films on substrates with large misorientation angles.

The predominant absorption in the near-infrared wavelength range for DBRs is attributed to free carriers within the thin film. Among these carriers, hole carriers play a primary role in absorption [7]. At the same concentration, the absorption of holes is three times that of electrons, underscoring the significance of reducing hole carrier concentration as a key strategy to mitigate absorption losses in thin films. The principal source of hole carriers in thin films stems from the unintentional doping of carbon atoms during the film preparation process [8]. It is widely accepted that temperature, arsine partial pressure, and substrate orientation significantly influence the unintentional doping of carbon.

Consequently, in this study, MOCVD technology was employed to fabricate single-layer thin films of AlGaAs and GaAs. The investigation focused on unraveling the intricate influences of external conditions on surface morphology and carrier concentration, encompassing substrate orientation and arsine partial pressure. The overarching goal is to identify optimal epitaxial conditions that minimize both the level of carbon doping and surface roughness in the thin films.

2. Experimental preparation and testing

The experiment was conducted utilizing an AIX 2600 G3 planetary metal-organic chemical vapor deposition (MOCVD) system. In this setup, Al0.92GaAs, Al0.72GaAs, and GaAs epitaxial layers were grown on semi-insulating undoped gallium arsenide substrates, each layer having a thickness of 2 µm. Additionally, a 20 nm GaAs protective layer was deposited on the AlGaAs film to prevent oxidation, as depicted in Fig. 1. The thin film epitaxy employed trimethylgallium (TMGa) and trimethylaluminum (TMAl) as Group III source materials, along with arsine (AsH3) as the Group V source material. The chamber's internal pressure was maintained at 50 mbar, and the growth temperature was set at 700°C. The V/III ratio was selected at 50 and 100, and the deposition rate maintain in at 0.5 nm/s.

 figure: Fig. 1.

Fig. 1. schematic diagram of the film layer structure: (a) AlGaAs (b) GaAs.

Download Full Size | PDF

The chosen substrate orientations included (100) ± 0.5°, (211) B ± 0.5°, and (100) 15° ± 0.5° off towards <111> A. The substrates had a diameter of 100 ± 0.4 mm and a thickness of 625 ± 25 µm. The reflectance and transmittance of Al0.92GaAs and Al0.72GaAs thin films are provided in Fig. 2. The thickness of all batches of film samples, determined through fitting with Optichar software, falls within the range of 2000 ± 50 nm. The refractive indices of Al0.92GaAs, Al0.72GaAs, and GaAs at 1064 nm are 2.762, 2.913, and 3.327, respectively. Surface morphology and roughness were assessed through atomic force microscope (AFM) measurements.

 figure: Fig. 2.

Fig. 2. The transmittance and reflectance of thin films.

Download Full Size | PDF

3. Results and discussion

3.1 Surface topography

The six samples of Al0.72GaAs were characterized using AFM (Atomic Force Microscope), as shown in Fig. 3. Step-like morphology was observed on the Al0.72GaAs grown on (100) substrates, as shown in Fig. 3 (a) and (d). The root mean square (RMS) roughness values were 0.120nm and 0.314nm, respectively. Similar stepped morphology was also observed on the surfaces of Al0.92GaAs and GaAs grow on (100) substrate (Fig. 4). Since we did not observe any steps on the (100) substrate surface, it is likely that the step features on the sample surface originated from the pre-growth thermal cleaning annealing process [9].

 figure: Fig. 3.

Fig. 3. AFM test morphology of Al0.72GaAs: (a), (d) grown on (100) substrate, V/III ratios are 50 and 100, respectively. (b), (e) grown on (211) B substrate, V/III ratios are 50 and 100, respectively. and (c),(f) grown on (100) 15 ° off substrate, V/III ratios are 50 and 100, respectively. The measurement scale for (b) and (e) is 30um, while the rest are 3um.

Download Full Size | PDF

 figure: Fig. 4.

Fig. 4. Surface morphology of (a) GaAs V/III ratio at 50 and (b) Al0.92GaAs V/III ratio at 50.

Download Full Size | PDF

We draw a 3um long line segment along the direction perpendicular to the step edge on the surfaces of Fig. 3 (a) and (d), and extract the relationship between the step height and the length covered, as shown in Fig. 5. From the blue curve in Fig. 5, it can be observed that the height variation of the steps is within 0.5 nm, with some step heights approaching the single atomic step height of 0.28 nm. Based on this observation, it can be inferred that both the single-step flow and step bunching growth modes coexist on the surface, as shown in Fig. 3(a). The orange curve in Fig. 5 shows that the height variation of the steps exceeds 0.28 nm and forms bulges at a distance of several step widths. Therefore, Fig. 3 (d) should represent the step bunching growth mode, which leads to an increase in roughness. The transition between step-flow and step-bunching growth modes is generally believed to be related to the Ehrlich-Schwoebel (ES) barrier [10]. When there is an ES barrier on the step surface, step-flow growth occurs, while step-bunching growth occurs when there is a reverse ES barrier. It is currently unclear how the pressure of AsH3 affects this mechanism and promotes the transition of growth modes. When the AsH3 partial pressure is high (Fig. 3(d)), the step edges appear smoother and clearer compared to the case of low AsH3 partial pressure (Fig. 3(a)). This may be because, at low AsH3 partial pressure, Ga atoms not only attach to the kink sites but also to other sites on the steps, resulting in multiple step edges that are more twisted. On the other hand, at high AsH3 partial pressure, Ga atoms tend to attach more to the kink sites than to other sites on the steps, resulting in smoother step edges [6].

 figure: Fig. 5.

Fig. 5. Depicts the step height profile of Al0.72GaAs grown on (100) substrates.

Download Full Size | PDF

Al0.72GaAs grown on (211) B substrate, as shown in Fig. 3(b) and (e), exhibited three-dimensional island-like protrusions on the surface, and the density and size of these islands decreased with increasing AsH3 partial pressure. The formation of three-dimensional islands is related to the surface free energy of the substrate. The surface energy tends to increase as the substrate misorientation angle increases [11]. The misorientation angle of the (211) B substrate is approximately 35°. With such a large tilt angle, the substrate surface has a higher surface energy. When the thin film starts to deposit, the material will uniformly wet the surface to form a strained layer with atomic layer thickness. As the thickness of the thin film increases, this elastic strain energy will be released in the form of three-dimensional islands, thereby minimizing the total energy of the thin film. This growth mode is called Stranski-Krastanow mode, abbreviated as SK mode. Due to the relationship between the growth mode and the substrate surface energy, when the growth temperature increases, the substrate surface energy decreases, and the growth mode of the thin film may change. This may help us to obtain a smoother and flatter surface. Similar studies have shown that as the temperature increases, the surface protrusions on the film grown on a (111) B substrate disappear and become smoother [12].

Al0.72GaAs grown on (100) 15° off substrate, as shown in Fig. 3 (c) and (f). Due to the smaller misorientation angle compared to (211) B, the smaller surface energy prevents the occurrence of the SK growth mode and instead shows a Frank-Van der Merwe mode, abbreviated as FM growth mode [13]. In this growth mode, the material first nucleates in a two-dimensional manner on the substrate surface, and the two-dimensional nuclei then coalesce and aggregate to form layers. This process is repeated for layer-by-layer growth. This growth mode easily achieves epitaxial layers with complete crystal structure and flat outer surfaces. When the partial pressure of arsine increases, there is no significant change in the surface morphology, and the root mean square roughness slightly decreases from 0.247 nm to 0.205 nm. This indicates that the epitaxial film grown on this substrate not only has a relatively flat surface, but we can also use the V/III ratio to adjust the carrier concentration of the film without significantly affecting its surface smoothness. Therefore, it is a suitable substrate choice for epitaxial AlGaAs/GaAs DBR.

3.2 Carrier concentration of thin films

The thin film's free carrier concentration was determined through Hall testing. By varying the magnetic field strength B, the relationship between VH/I and B was obtained. The carrier concentration n was then calculated by fitting the slope KH, also known as the Hall sensitivity. During the testing, the thin film was cut into rectangular pieces of a few millimeters in size, and electrodes were attached to its surface for voltage collection. The magnetic field range for testing was -20000 Oe to 20000 Oe, and the current was controlled at 2 mA. The testing temperature was maintained at approximately 300 K.

We conducted Hall tests of Al0.72GaAs and Al0.92GaAs thin films grown on (100) substrates, with the testing curves and details regarding the carrier concentration and type illustrated in Fig. 6. Under identical growth conditions, a noticeable trend emerged: as the Al composition increased, so did the hole carrier concentration. It is commonly understood that holes in the AlGaAs epitaxial layer originate from the methyl group in the organic source trimethylaluminum (TMAl), which occupies surface arsenic sites. Since carbon acts as a donor in AlGaAs, the hole concentration serves as an indicator of the amount of carbon impurities in the epitaxial layer. The high bonding energy of the Al-C bond leads to an elevation in the background concentration of carbon in the epitaxial layer as the Al composition increases, consequently resulting in an increase in hole concentration. With an increase in the V/III ratio, corresponding to an elevation in the concentration of AsH3, more AsHx radicals generated from the decomposition of AsH3 will react with the methyl on the substrate surface. The resulting products are carried away by the gas flow in the reactor, leading to a reduction in carbon impurities in the epitaxial layer and, subsequently, a decrease in hole concentration. This observed trend holds true for epitaxial layers grown on the other two substrates as well.

 figure: Fig. 6.

Fig. 6. Hall testing of Al0.92GaAs/ Al0.72GaAs grown on (100) substrate.

Download Full Size | PDF

The carrier concentration of the Al0.72GaAs thin film, cultivated on diverse substrates, is presented in Table 1, the carbon doping dependency exhibits variation across different substrates. With the epitaxial layer grown on a (100) 15° off substrate at a V/III ratio of 100 demonstrating the lowest hole concentration, measuring only 3.43 × 1016/cm3. This reliance on substrate orientation for carbon doping can be ascribed to the distinct densities of carbon adsorption sites on surfaces with varying orientations. Additionally, it may result from the disparate abilities of substrates with different orientations to adsorb AsHx species, leading to varying degrees of carbon desorption [14].

Tables Icon

Table 1. Carrier concentration of Al0.72GaAs epitaxial layers under different substrates and V/III ratios

The behavior of the GaAs epitaxial layer differs from that of AlGaAs. Within the V/III ratio range of 0 to 100, not all carriers in the GaAs layer exhibit p-type behavior [15]. Instead, when the V/III ratio surpasses a certain threshold, the carrier polarity undergoes an inversion from p-type to n-type. This inversion is attributed to the increase in the partial pressure of arsine, resulting in a phenomenon similar to that observed in the AlGaAs epitaxial layer. Excessive decomposition of AsH3 generates hydrogen, which reacts with methyl to produce methane. This reaction rapidly desorbs and reduces the incorporation of carbon, leading to a decrease in hole carrier concentration.

However, as the V/III ratio continues to rise and surpasses the transition point, the continuous introduction of n-type impurities from AsH3 causes the electron concentration in the epitaxial layer to increase rapidly, resulting in a change in carrier type from p-type to n-type [16]. The specific value of the transition point V/III ratio is generally considered to be influenced by factors such as the purity of the gas source, temperature, and substrate orientation [15].

AlGaAs also undergoes a similar change in carrier type with varying V/III ratio, but due to the presence of aluminum, the transition point for the V/III ratio is typically much higher than that of GaAs. Consequently, this phenomenon was not observed in the current experiment. Table 2 presents the carrier concentration of GaAs epitaxial layers grown on (100) and (211) B substrates. The transition point of GaAs grown on (100) substrate is smaller compared to that grown on (211) B substrate. The inversion of carrier type occurs at a V/III ratio less than 50 for the (100) substrate; when the V/III ratio increases to 100, excessive n-type impurity doping leads to a very high electron concentration. This results in a Hall curve with an almost zero slope, making it challenging to accurately calculate the values. Conversely, for the (211) B substrate, the transition point occurs at a value between 50 and 100. The results highlight that by selecting the V/III ratio near the transition point, the grown GaAs film can exhibit a relatively low level of carrier concentration.

Tables Icon

Table 2. Carrier concentration of GaAs epitaxial layers under different substrates and V/III ratios

4. Summary and conclusions

The experiment involved depositing various Al compositions of AlGaAs and GaAs thin films on substrates oriented along three distinct crystallographic directions: (100), (211) B, and (100) 15° off toward < 111 > A. Our investigation revealed a profound influence of substrate orientation on the surface morphology and growth mode of AlGaAs epitaxial layers. Specifically, on (100) substrates, a step-flow/step-bunching growth mode was observed, contrasting with the SK growth mode on (211) B substrates, resulting in a notably rough surface. On (100) 15° off substrates, an FM growth mode was identified. The impact of AsH3 pressure on surface roughness exhibited orientation dependence, with epitaxial layers grown on (100) 15° off substrates displaying the least sensitivity to AsH3 pressure in terms of roughness.

The carrier concentration of AlGaAs epitaxial layers proved to be contingent upon substrate orientation, with the lowest carrier concentration achieved on (100) 15° off substrates. Overall, carrier concentration exhibited an upward trend with increasing Al composition, while a decrease was noted with an increase in the V/III ratio. In the case of GaAs epitaxial layers, the carrier concentration initially decreased (p-type) and subsequently increased (n-type) with an elevation in the V/III ratio. The judicious selection of the V/III ratio emerged as a critical factor for achieving the lowest background carrier concentration in GaAs epitaxial layers. Generally, under an appropriate V/III ratio, the (100) 15° off toward <111 > A substrate stands out as a suitable choice for striking a balance between low carrier concentration and high surface flatness. We believe that this work is helpful in reducing the overall scattering and absorption losses of DBRs. Of course, its significance is not limited to this, extending to various fields such as fabricate new device structures on patterned substrates and metasurface [17].

Funding

National Natural Science Foundation of China (62061136008, 61975155, 62192772).

Disclosures

The authors declare no conflicts of interest.

Data availability

The data underlying the results presented in this paper can be obtained from the authors upon reasonable request.

References

1. M. Zorn, P. Klopp, F. Saas, et al., “Semiconductor components for femtosecond semiconductor disk lasers grown by MOVPE,” J. Cryst. Growth 310(23), 5187–5190 (2008). [CrossRef]  

2. G. D. Cole, S. Groeblacher, K. Gugler, et al., “Monocrystalline AlxGa1-xAs heterostructures for high reflectivity high-Q micromechanical resonators in the megahertz regime,” Appl. Phys. Lett. 92(26), 261108 (2008). [CrossRef]  

3. G. D. Cole, Y. Bai, M. Aspelmeyer, et al., “Free-standing AlxGa1-xAs heterostructures by gas-phase etching of germanium,” Appl. Phys. Lett. 96(26), 277–850 (2010). [CrossRef]  

4. G. D. Cole, I. Wilson-Rae, M. R. Vanner, et al., “Megahertz monocrystalline optomechanical resonators with minimal dissipation,” IEEE International Conference on Micro Electro Mechanical Systems, 847–850 (2010).

5. J. E. Harvey, S. Schroeder, N. Choi, et al., “Total integrated scatter from surfaces with arbitrary roughness, correlation widths, and incident angles,” Opt. Eng. 51(1), 013402 (2012). [CrossRef]  

6. M. Kasu and T. Fukui, “Multi-atomic steps on metalorganic chemical vapor deposition-grown GaAs vicinal surfaces studied by atomic force microscopy,” Jpn. J. Appl. Phys. 31(7A), L864–866 (1992). [CrossRef]  

7. D. I. Babic, J. Piprek, K. Streubel, et al., “Design and analysis of double-fused 1.55-mu m vertical cavity lasers,” IEEE J. Quantum Electron. 33(8), 1369–1383 (1997). [CrossRef]  

8. A. V. Annapragada, S. Salim, and K. F. Jensen, “FTIR studies of organometallic surface chemistry relevant to atomic layer epitaxy,” MRS Online Proceedings Library 222, 81–86 (1991). [CrossRef]  

9. K. Hata, A. Kawazu, T. Okano, et al., “Observation of step bunching on vicinal GaAs (100) studied by scanning tunneling microscopy,” Appl. Phys. Lett. 63(12), 1625–1627 (1993). [CrossRef]  

10. R. L. Schwoebel, “Step motion on crystal surfaces II,” J. Cryst. Growth 40(2), 614–618 (1969). [CrossRef]  

11. E. S. Johnson and G. E. Legg, “Critical misorientation morphology in AlGaAs and GaAs grown by atmospheric-pressure MOCVD on misoriented substrates,” J. Cryst. Growth 88(1), 53–66 (1988). [CrossRef]  

12. K. Kato, Atsuo Kozen, Jiro Temmyo, et al., “AlGaAs epitaxial growth on (111) B substrates by metalorganic vapor-phase epitaxy,” J. Appl. Phys. 65(5), 1947–1951 (1989). [CrossRef]  

13. D C Lu and S K. Duan, Metal Organic Chemical Vapor Deposition: Fundamentals and Applications (Science Press, 2009) (in Chinese).

14. M. Kondo and T. Tanahashi, “Dependence of carbon incorporation on crystallographic orientation during metalorganic vapor phase epitaxy of GaAs and AlGaAs,” J. Cryst. Growth 145(1-4), 390–396 (1994). [CrossRef]  

15. P. D. Dapkus, K. L. Hess, T. S. Low, et al., “High purity GaAs prepared from trimethylgallium and arsine,” J. Cryst. Growth 55(1), 10–23 (1981). [CrossRef]  

16. T. Nakanisi, T. Udagawa, and K. Kamei, “Growth of high-purity GaAs epilayers by MOCVD and their applications to microwave MESFET's,” J Crystal Growth 55(1), 255–262 (1981). [CrossRef]  

17. C. Gigli and G. Leo, “All-dielectric χ(2) metasurfaces: recent progress,” Opto-Electron. Adv. 5(7), 210093 (2022). [CrossRef]  

Data availability

The data underlying the results presented in this paper can be obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1.
Fig. 1. schematic diagram of the film layer structure: (a) AlGaAs (b) GaAs.
Fig. 2.
Fig. 2. The transmittance and reflectance of thin films.
Fig. 3.
Fig. 3. AFM test morphology of Al0.72GaAs: (a), (d) grown on (100) substrate, V/III ratios are 50 and 100, respectively. (b), (e) grown on (211) B substrate, V/III ratios are 50 and 100, respectively. and (c),(f) grown on (100) 15 ° off substrate, V/III ratios are 50 and 100, respectively. The measurement scale for (b) and (e) is 30um, while the rest are 3um.
Fig. 4.
Fig. 4. Surface morphology of (a) GaAs V/III ratio at 50 and (b) Al0.92GaAs V/III ratio at 50.
Fig. 5.
Fig. 5. Depicts the step height profile of Al0.72GaAs grown on (100) substrates.
Fig. 6.
Fig. 6. Hall testing of Al0.92GaAs/ Al0.72GaAs grown on (100) substrate.

Tables (2)

Tables Icon

Table 1. Carrier concentration of Al0.72GaAs epitaxial layers under different substrates and V/III ratios

Tables Icon

Table 2. Carrier concentration of GaAs epitaxial layers under different substrates and V/III ratios

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.