F. Cardone

Controlled carbon doping of GaAs by metalorganic vapor phase epitaxy

The controlled incorporation of carbon has been demonstrated for the metalorganic vapor phase epitaxy of GaAs. Carbon levels between 1016 and 1019 cm−3 can be achieved under typical growth conditions by using Ga(CH3)3 and either As(CH3)3 or mixtures of As(CH3)3 and AsH3. The carbon incorporation into GaAs goes through a minimum with growth temperature at ∼650u2009°C when using Ga(CH3)3 and As(CH3)3. The controlled addition of AsH3 monotonically decreases the carbon incorporation. The high carbon levels (≳1–2×1019 cm−3), greater than the reported solid solubility, are thermally stable with a low diffusion coefficient. The GaAs:C layers exhibit a low deep level concentration, ∼1013 cm−3, with only a single midgap trap present.

Quantitative analysis of complex targets by proton‐induced x rays

We have explored the potential of proton‐induced x‐ray spectroscopy as a quantitative analytical technique to determine the composition of a homogeneous complex target. Direct production of x ray by protons in the low MeV range was considered using the binary encounter approximation and published data on the atomic number dependance of the stopping power. Also included is a correction scheme for secondary excitation processes if radiation is generated in the target capable of fluorescence excitation. The formalism of the correction scheme is closely related to the one commonly employed in electron microprobe spectroscopy. Our data indicate that the accuracy achievable in IIXS is of the order of ±5%, if one considers the two important corrections: (1) the difference in the stopping power properties of the composite target compared to the pure element target; (2) secondary excitation processes if required.

Carbon incorporation in metalorganic vapor phase epitaxy grown GaAs using CHyX4 − y, TMG and AsH3

Abstract The incorporation of carbon into GaAs grown by metalorganic vapor phase epitaxy has been studied through the addition of the halomethanes CCl 4 , CHCl 3 , CH 2 Cl 2 , CBr 4 , CI 4 , CHI 3 , CH 2 I 2 , CH 3 I and CH 4 . Growth temperatures of 600–750°C, V/III ratios of 30–100, and halomethane mole fractions of 2 × 10 -8 -1 × 10 -4 were investigated. The incorporation of carbon for all the halomethanes decreases at high V/III ratios and also decreases at high growth temperatures with an exponential Arrhenius fitting parameter of -30 to -60 kcal/mol. The carbon incorporation dependence of the chloromethanes on halomethane mole fraction conforms to a first order power function. However, carbon incorporation with CH 2 I 2 conforms to a second order power function. The trend in the halomethanes is to incorporate carbon more efficiently in more highly substituted halomethanes, following the series CH 3 X 2 X 2 3 4 throughout the range V/III ratios and growth temperatures investigated. good correlation in this series of low average bond strength to carbon incorporation is also observed. The dependence of carbon incorporation on V/III ratio and growth temperature are explained by a simple reaction scheme based on the competition at the growing interface between the decomposition of the halomethane to incorporate carbon and the desorption of the decomposing halomethane from the interface.

The control and modeling of doping profiles and transients in MOVPE growth

Abstract The accurate placement of dopants during chemical vapor deposition is complicated by many factors: growth temperature, reactor design, flow conditions, and the choice of growth and doping chemistry. Long doping transients have often been noted in structures grown using, for example, dopant precursors such as H 2 Se and Mg(C 5 H 5 ) 2 . These transients appear at the “turn-on” of a dopant source as well as the termination or “turn-off” of the source. The grown-in dopant profiles are also modified by the diffusion of the dopant during the thermal history of the structure. The major cause of these transients in the case of metal-organic doping sources, such as Mg(C 5 H 5 ) 2 , is the adsorption and desorption of the dopant precursors on the internal surfaces of the reactor. We have developed a heuristic model of this process which can describe the major features of this process. Growth conditions and dopant source characteristics are described which minimize these growth transients.

High carbon doping efficiency of bromomethanes in gas source molecular beam epitaxial growth of GaAs

Carbon tetrabromide (CBr4) and bromoform (CHBr3) have been studied as carbon doping sources for GaAs grown by gas source molecular beam epitaxy (GSMBE) with elemental Ga and thermally cracked AsH3. Hole concentrations in excess of 1×1020 cm−3 have been measured by Hall effect in both CBr4‐ and CHBr3‐doped GaAs, which agrees closely with the atomic C concentration from secondary‐ion mass spectrometry, indicating complete electrical activity of the incorporated carbon. The GaAs growth rate is unaffected by the CBr4 and CHBr3 fluxes over the range of dopant flow investigated. The efficiencies of carbon incorporation from CBr4 and CHBr3 are, respectively, 750 and 25 times that of trimethylgallium (TMG), which is commonly employed as a carbon doping source in metalorganic MBE (MOMBE). The sensitivity of carbon incorporation to varying substrate temperature and V/III ratio has been observed to be significantly reduced with CBr4 and CHBr3 from that obtained under similar growth conditions with TMG in MOMBE.

Arsenic cluster dynamics in doped GaAs

We have studied the formation of As precipitates in doped GaAs structures that were grown by molecular beam epitaxy at low substrate temperatures and subsequently annealed. We find that the As precipitates form preferentially on the n side of such fabricated GaAs pn junctions. As the coarsening process proceeds, there is a gradual increase in the amount of As in precipitates in the n‐GaAs region and a decrease in the p‐GaAs region; the depletion region between the pn junction becomes free of As precipitates. These observations can be understood qualitatively based on the charge states of the As interstitial and using thermodynamic arguments in which the crystal attempts to minimize the chemical potential during the anneal. The presence of the excess As results in a stable Be profile even to anneals of 950u2009°C. Finally, a temperature cycling technique to grow arbitrarily thick GaAs epilayers containing As precipitates was demonstrated.

Characterization of epitaxial GaAs and AlxGa1−xAs layers doped with oxygen

Intentional oxygen doping (≳1017 cm−3) of GaAs and Al0.30Ga0.70As epitaxial layers was achieved during metalorganic vapor phase epitaxy through use of an oxygen‐bearing metalorganic precursor, dimethylaluminum methoxide (CH3)2AlOCH3. The incorporation of oxygen and very low levels of Al (AlAs mole fraction <0.005) in the GaAs layers leads to the compensation of intentionally introduced Si donors. Additionally, deep levels in GaAs associated with oxygen were detected. The introduction of dimethyl aluminum methoxide during AlxGa1−xAs growth did not alter Al mode fraction or degrade the crystallinity of the ternary layers, but did incorporate high levels of oxygen which compensated Si donors. The compensation in both GaAs and Al0.30Ga0.70As indicates that high resistivity buffer layers can be grown by oxygen doping during metalorganic vapor phase epitaxy.

Doping and dopant behavior in (Al,Ga)As grown by metalorganic vapor phase epitaxy

Abstract The controlled doping of n- and p-type Al x Ga 1−x As has been studied for the dopant elements, C, Zn, Si, and Sn. Both the incorporation characteristics and the electrical properties of these dopants are reviewed and discussed fo Al x Ga 1−x As grown by th metal-organic vapor phase epitaxy (MOVPE) technique. The incorporation of Si from SiH 4 and Si 2 H 6 is dominated by heterogeneous and homogenous reactions respectively and represents the best understood of the doping systems. Zinc and carbon both possess complex dependencies on the MOVPE growth system parameters. The electrical behavior of n-Al x Ga 1− x As is dominated by the presence of the DX center. The relationship between this center and the electrical behavior of the material must be understood in order to properly characterize the doping behavior in Al x Ga 1−x As layers and structures.

The influence of hydrocarbons in MOVPE GaAs growth: Improved detection of carbon by secondary ion mass spectroscopy

Abstract The quantitative measurement of carbon concentration in films grown by MOVPE is required to elucidate both the growth reactions and the influence of reaction-by-products. Secondary ion mass spectroscopy (SIMS) can determine the total carbon content in a layer and complement measurements of the electrically active carbon. The presence of carbon compounds in the residual gas background of SIMS instrument has typically limited the carbon detection limits to ∼1017 cm−3. We have developed alternative SIMS techniques for the detection of carbon by utilizing the analyte ion, As13C−, yielding detection limits down to 1014–1015 cm−3 for 13C in GaAs. We have applied this technique to the study of the incorporation of carbon from CH4 and isotropically enriched C2H2, C2H4, and C3H6 during the MOVPE growth of GaAs.

Deep levels in p‐type GaAs grown by metalorganic vapor phase epitaxy

We report a detailed deep level transient spectroscopic study in p‐type Mg‐ and Zn‐doped GaAs epitaxial layers grown by metal‐organic vapor phase epitaxy. Dependence of deep level structures on doping concentrations and growth temperatures has been investigated. Over a wide range of growth conditions, four hole traps and an electron trap ranging in activation energy from 0.18–0.79 eV were measured in GaAs:Mg while only a single hole trap has been observed in GaAs:Zn.The presence of a certain trap and its concentration in GaAs:Mg depends mainly on the doping concentration in the layers. The total trap concentration in the GaAs:Mg decreases rapidly with doping concentration for p>4×1017 cm−3. The physical and chemical origins of several of these traps have been identified. The Mg‐doped GaAs always exhibited a greater concentration of midgap trap levels than the Zn‐doped material, regardless of dopant concentration or growth temperature. The overall defect structure and dopant incorporation characteristics i...