Native Oxide Mask for Zinc Diffusion in Gallium Arsenide

Experiments were performed to diffuse zinc into GaAs through anodic oxide layers of varying thickness and density. Using electrochemical profiling to determine both the electrically active zinc concentration and the diffusion depth with high resolution, the following results were found. The depth of the junction varies inversely with the thickness and the density of the oxide. However, the surface concentration appears to be independent of oxide thickness or density, attaining a value identical to that found for diffusion into a bare GaAs sample. These results demonstrate that the most significant impact of the oxide is to delay the introduction of the zinc into the GaAs lattice. In short, the anodic oxide cannot be used as either a mask or as a zinc concentration attenuator.

Section: Resultscontrasting

confidence: 62%

Section: Resultsmentioning

confidence: 98%

The Use of Anodic Oxide Films to Control the Diffusion of Zinc in GaAs

Cutlerywala

Roedel

1994

“…An experiment is now in progress to explore these possibilities. A nother class, diffused junction devices, can be optimized using well-con.trolled selective area diffusion (13) along with a diffusion damage buffer (13) as a result of the anodically grown oxides. The sectioning technique provides a fabrication technology for yet another class, mesa or quasi-planar.…”

Section: Device Applicationmentioning

confidence: 99%

“…It has been found that this native oxide acts as a diffusion mask against Zn (13). It is also a barrier against moisture and may act as a barrier to many different atomic species.…”

Section: Device Applicationmentioning

confidence: 99%

Preparation and Stabilization of Anodic Oxides on GaAs

Spitzer¹,

Schwartz²,

Weigle³

1975

Anodic oxidation of GaAs is examined and a technique for growing full surface oxides ~ 86G,0A thick is reported. Anodization in a hot, concentrated (30%) H202 electrolyte of pH ~-2 using self-anodizing metal contacts is found to yield uniform oxides with a refractive index of 1.8 and a growth rate ranging from 22 to 50 A/V. The as-grown oxide will etch in HC1 and H20. Chemical stability of this oxide is enhanced with baking in dry nitrogen and after 2 hr at 250~ H20 will no longer etch the oxide; after 1 hr at 600~ HC1, HF, HNO3, NI-I4OH will not etch the oxide but hot HsPO4 will etch the oxide. The oxide is an amorphous, vitreous film which crystallizes to ~-Ga20~ after several hours at 800~ A thin (<100A) graded, oxide-GaAs interface was found to grow at anodization potentials >100V and for heat-treatment at temperatures greater than 400~ for 1 hr. The oxide exhibits excellent adherence to the substrate, has good match with the thermal coefficient of expansion of GaAs, and forms a relatively strain-free interface. Since GaAs is consumed during anodization, this provides a technique for surface cleaning using an oxide growth-strip step. Enhanced device reliability is expected to result by utilization of the oxide growth-strip-regrowth cycle during device fabrication, since this will remove contamination from the surface. The findings here enumerated, coupled with ease of oxide growth, suggest ready implementation to a GaAs planar technology.

“…The growth of native oxides on GaAs and GaP by galvanic and anodic techniques (1-4) using 30% aqueous solutions of H202 as the electrolyte has been shown to be a useful new technology for passivation of semiconductor device surfaces (5,6), as a diffusion mask (7), and for selective area etching (2). The use of a buffered aqueous solution for the growth of an anodic oxide on GaP has recently been described (8).…”

mentioning

confidence: 99%

The Anodization of GaAs and GaP in Aqueous Solutions

Schwartz¹,

Ermanis²,

Brastad³

1976

The anodic oxidation of normalGaAs and normalGaP in properly conductivity‐ and/orpH‐adjusted water has been successfully demonstrated under both constant voltage and constant current conditions. With H3PO4 as the acidic conductivity/pH modifier, uniform, well‐controlled oxides have been grown in thepH range 2.5–3.5. The oxide thickness‐voltage relationships for both normalGaAs and normalGaP are linear, with slopes of approximately 20 and 12 Å/V, respectively. At room temperature, oxides as thick as 3600 and 2000Å can reproducibly be grown on normalGaAs and normalGaP , respectively; at 100°C, an oxide as thick as 5000Å has been grown on normalGaAs . Under basic conditions, with NH4OH as the conductivity/pH modifier, oxides have been grown in the pH range of 10–11, but dissolution of the oxide in the bath results in much poorer control than in the acidic system; this dissolution effect can be utilized more in the line of electroetching than in simple oxide formation. Anodic oxidation, with relatively little oxide dissolution, has also been accomplished in near neutral‐pH water (i.e., ∼7) with false(NH4)2HPO4 as the conductivity/pH modifier. Restricted‐area oxidation and/or etching has been demonstrated whereby a prefabricated photoresist pattern is used to define restricted areas for anodization; lines as narrow as 5μ wide are readily delineated. Anodization in aqueous solutions containing either Cl− or NO3− ions is shown to result in simple electroetching, and current densities in the range 10–20 mA/cm2 are demonstrated to be most effective for controlled electroetching. The grown oxides are soluble, in normalHCl , HNO3 , and H2SO4 , and are affected by prolonged contact with water. If properly baked, however, the oxide grown on normalGaAs shows no evidence of change in months of storage in laboratory air; the oxide grown on GaP does show evidence of moisture absorption after 1 month of equivalent storage. A preliminary analysis of some of the controlling electrical factors during anodization shows that under constant voltage conditions, the current false(Ifalse) varies asI=ABt+Cand under constant current conditions, the time derivative of the voltage is a constant.