P. Balk

Lp-movpe Growth and Optical Characterization of Galnp/gaas Heterostructures: Interfaces, Quantum Wells and Quantum Wires

Lattice matched Ga0.5In0.5P/GaAs single interfaces, single quantum wells (QW) and finally quantum wires on mesa-like selective GaAs were grown by LP-MOVPE. The photoluminescence (PL) of GaInP/GaAs QW shows an anomalous emission band of high intensity below the GaAs band gap. The effect of the growth temperature, AsH3 partial pressure, gas switching sequence and growth interruption times on these new PL features was investigated. The formation of GaInPAs layers at the GaInP-to-GaAs interface, due to the substitution of P to As, is responsible for the anomaly in the luminescence. The stability of the GaInP/GaAs interfaces was checked after growth by rapid thermal annealing (RTA). Intermixing at the GaInP/GaAs heterointerfaces was also considered. The introduction of a 1 nm thick GaP barrier between the GaInP and GaAs layer was enough to suppress the GaInPAs intermediate layers and large area quantum wells and wires were successfully grown.

Composition of selectively grown InxGa1−xAs structures from locally resolved Raman spectroscopy

The composition of InxGa1−xAs stripes of different width, selectively grown by low pressure MOVPE or MOMBE, was determined from the eigenfrequency of the Ga-As lattice vibration using Raman spectroscopy. In contrast to MOMBE, structures grown by MOVPE show a strong enhancement of the In content with decreasing width and increasing distances between the stripes. Comparative studies on MOVPE grown InxGa1−xP structures indicate a behavior similar to that observed in the case of InxGa1−xAs.

Control of selective area growth of InP

Selective growth of InP by low pressure metalorganic vapor phase epitaxy (LP MOVPE) on (100) InP substrates was studied using SiO2 masks. The role of the mask geometry on the shape of the selectively grown structures was investigated. The rate and the cross section of the patterns depends on deposition temperature, V/III ratio, growth velocity and total pressure. The experimental findings can be explained by a simple model assuming orientation dependent growth rates and surface migration. Hall data (μ77 = 65,000 cm2/V·s) show the excellent material quality of the selectivity grown structures.

Gaseous dopant sources in MOMBE/CBE

Abstract In this paper we will review the results obtained when using gaseous dopant sources in the MOMBE of GaAs and InP. The main focus will be on the decomposition behavior of the dopant sources and the dopant incorporation process. Due to the fact that the molecules have to be decomposed on the growing surface the overall process of dopant incorporation is more complex than that for elemental dopants. Based on the presented findings the suitability of a number of dopant sources will be evaluated. The properties of the dopant atoms themselves (diffusion, surface segregation, maximum obtainable doping level) are summarized.

CARS in situ diagnostics in MOVPE: The thermal decomposition of AsH3 and PH3

Abstract Coherent anti-Stokes Raman Scattering (CARS) is used to study the thermal decomposition of AsH 3 and PH 3 in a horizontal reactor for metalorganic vapor phase epitaxy (MOVPE) of III/V compound semiconductors. Both in situ diagnostics and ex situ sampling experiments are considered. The decomposition temperatures obtained from in situ experiments are much lower (300 to 400 K) than those measured by ex situ techniques. The ex situ data compare well to previously reported observations. The results indicate that the thermal decomposition of AsH 3 and PH 3 is complex involving gas phase and surface reactions. The large difference in decomposition behavior between the two measuring techniques also demonstrates the importance of in situ diagnostics and the difficulty in ex situ monitoring of MOVPE species concentrations. No change in AsH 3 or PH 3 decomposition behavior is observeed with the addition of TMG or TMI. Molecular hydrogen is detected as a decomposition product, however in concentrations which are lower than expected from complete hydride decomposition.

Substituted arsines as As sources in MOMBE

Abstract We have for the first time used in the MOMBE (metal organic molecular beam epitaxy) of GaAs gaseous arsenic sources which do not require precracking in the gas inlet tube. The partially substituted arsines tBAsH 2 and PhAsH 2 decompose on the hot substrate surface and thus allow GaAs deposition. Using TEG as the Ga source layers grown from PhAsH 2 were weakly p-type (4×10 15 cm -3 ) and showed narrow excitonic transitions in the low temperature photoluminescence spectrum. For both PhAsH 2 and tBAsH 2 we could not determine the occurrence of intrinsic (e.g. carbon) impurity uptake on a level much higher than that achieved with AsH 3 . However, the extrinsic purity of these starting materials needs further improvement. Whereas growth from uncracked tBAsH 2 and PhAsH 2 is possible, the trialkyl TMAs, similar to AsH 3 , is too stable to be used without precracking.

Physical properties of non-pyrophoric group III precursors for MOVPE

Abstract Non-pyrophoric organometallic compounds of aluminum, gallium and indium have been developed for safer application in metalorganic vapor phase epitaxy (MOVPE) of III–V semiconductor layers. They represent molecules which are inter- or intramolecular coordinatively saturated. Their physical properties can be easily varied by simple structure variations. Thus MOVPE precursors are available which are liquid at room temperature and which have already proved to be suitable for use with conventional as well as with new organic group V sources. Synthesis and epitaxy are discussed.

Deposition by LP-MOVPE in the Ga-In-As-P System on Differently Oriented Substrates

Abstract We have studied the MOVPE of GaInAs and its binary constituents GaAs and InP and its binary constituents GaP and InP on InP, and on GaAs (100), (011), (111)A, (111)B oriented unmasked planar substrates in order to gain an understanding of the deposition on the bottom plane and on side wall facets of trenches in selective embedded growth. Temperature (940 K) and total pressure (20 hPa) were selected with this purpose in mind. It is shown that the behaviour of the binaries provides an understanding of the factors determining the deposition of the ternaries and that insights for selective growth on partially masked substrates can be gained from these data. In contrast to InP and InAs growth, the rates for the Ga binaries GaAs and GaP and the ternaries depend on the substrate orientation. The variation of the rates of the ternaries is very similar to that of the binary Ga constituents. More importantly, also the composition (Ga content) of the ternaries appears to track with this parameter and varies rather widely, particularly for GaInAs. In the case of selective embedded deposition the differences in rates between the different surfaces tend to be enhanced compared to those on unmasked substrates due to migration effects.

Selective embedded growth of GaInAs by low pressure MOVPE

The study reported on in this paper is concerned with selective low pressure (20 hPa) MOVPE of GaInAs and GaInAs/InP heterostructures at 940 K in etched recesses in partially masked substrates. Different growth behavior is obtained for the single layers of the ternary and binary/ternary heterostructures. Moreover, the shape of the recesses, which is related to the etching procedure, also affects the cross section of the deposited stripes. Proper choice of the shape of the recesses permits the growth of planar stripes. The feasibility of in-situ doping of the GaInAs stripes using H2S or DEZn is demonstrated.

MOMBE of InAs on GaAs

Abstract We present results of a study on MOMBE growth of InAs on GaAs (100), using precracked arsine and two different alkyls (TMI, TEI). Despite the 7% lattice mismatch, mirrorlike layers with good crystalline and excellent electrical properties were achieved, as shown by RBS and Hall measurements. At growth temperatures above 810 K the carrier concentration ( n = (2−3)×10 16 cm -3 ) did not depend on the nature of the precursor. At low temperatures the carrier concentration remained unchanged for TEI, whereas for TMI a significant increase of background doping level to n = 3×10 18 cm -3 was observed. The assumption that this is caused by carbon incorporation due to the strength of the In-methyl bond is supported by SIMS measurements.