This paper describes the development of a surface-imaging process for a positive-tone silylated, dry-developed bilayer resist which has 0.2 μm resolution and an aspect ratio of 4.5 using deep-UV (248 nm) exposure. The many processing variables such as thermal treatment parameters, silylation conditions, and etching conditions were examined to determine their effects on lithographic performance in terms of resolution, feature size linearity, focus latitude, and sensitivity. Critical to the success of the process are: the bilayer structure which restricts diffusion of the Si, the use of a disilane reagent to increase the Si content of the masking layer, limiting migration of photogenerated acid by the appropriate choice of softbake and post-exposure bake temperatures, initial etching with an Ar/Cl2 mixture to remove the thin layer of silylated resist in the exposed areas, and employing CO2 instead of O2 as the etching gas to eliminate lateral etching of the features. With this process we have obtained good critical dimension linearity down to 0.25 μm for bright-field and dark-field lines and spaces as well as isolated lines and isolated spaces. The dose required is ∼75 mJ/cm2 and the dose latitude is ±6%. Focus latitude is at least ±0.4 μm. We also observe no environmental effects on sensitivity or resolution.

A selective deposition of W(Zn) metallization, for formation of diffused ohmic contacts onto InP-based material was realized by means of rapid thermal, low pressure metalorganic chemical vapor deposition (RT-LPMOCVD). The W(Zn) layers were deposited using a reactive gas mixture that contained diethylzinc (DEZn), WF6, H2 and Ar, at temperatures of 450 to 550°C and pressures in the range of 1–5 torr. Uniform andcontinuous layers of W(Zn), 30 to 120 nm thick, were obtained. These layers contained Zn at concentrations higher than 1×l018 cm−3, which was subsequentially in-diffused into the underlying semiconductor layer to form highly doped semiconductor layers as thick as 0.2μm. As a result, the specific contact resistance of the W(Zn)/ In0.53Ga0.47 As contact was reduced to minimum value of 5×l0−6 Ω.cm2. The W(Zn) film were found to be mechanically stable with a small compressive stress of 5.10−8 dyne. cm−2, and dry etch rates of up to 90 nm.min.

Tantalum nitride films were reactive sputter deposited onto chemical vapor deposited (CVD)-diamond self-standing thick layers, to be used as resistors for microelectronic applications. The TaN films had excellent morphology and were very stable through heating cycles at temperatures up to 400 °C for a few hours. Post-deposition sintering of the films at temperatures up to 300 °C stabilized the film resistance at values in the range of 75–85 Ω. The deposited film was later patterned with photoresist and dry etched, at rates of up to 70 nm min−1 and the resulting features served as masks for further self-aligned etching processes of the underlying CVD-diamond layer.

High purity Al films have been deposited on TiN and in situ evaporated Cu films on Si in a cold wall low pressure metalorganic chemical vapor deposition reactor using trimethylamine alane. TiN and Cu serve as nucleation layers for Al deposition to replace the commonly used but corrosive TiCl4. Al films deposited on TiN at 250 °C exhibit highly textured (111) orientation, whereas those grown on Cu at 150–350 °C are randomly oriented with some preferential growth in the (100) direction. These films are compared with films grown on Si and SiO2 using TiCl4 activation. Cu serves the additional purpose of diffusing into the Al during deposition to improve the electromigration resistance of the films. Electrical resistivities of the Al films on TiN are 4–6 μΩ cm, roughly twice that of bulk Al. The variation of film morphology with deposition temperature, substrate, and precursor pressure is discussed.

It is shown for the first time that carbon behaves predominantly as an acceptor in InGaAs and AlInAs under co-implantation conditions. The co-implanted ion, regardless of species, acts to create vacant lattice sites for occupation by the carbon. Implantation of 40 keV carbon ions alone at doses between 5×1012 and 5×1014 cm−2 followed by annealing in the range 600–950 °C for 10 s does not produce any measurable electrical activity in either material. In InGaAs, carbon implantation at 5×1014 cm−2 produced net acceptor activations of 20, 11, or 6% for Ga, Ar, or As co-implantation, respectively, at the same doses after 700 °C, 10 s anneals. Similar results were obtained for AlInAs after annealing at 900 °C. The diffusion coefficient for carbon is estimated from secondary-ion mass spectrometry measurements to be less than 3.3×10−14 cm2 s−1 at 800 °C in both materials.

This paper compares Hall-effect measurements combined with rapid thermal annealing (RTA) and low-temperature photoluminescence (4-K PL) as characterization techniques for the optimization of the growth of pseudomorphic InGaAs channel modulation-doped field-effect-transistor structures. The Hall-effect measurements with the RTA were used to determine the optimized growth temperature. 4-K PL was used to determine the quality of the InGaAs quantum well and the In mole fraction.

Recent advances in heterostructure bipolar transistor (HBT) technology have created a need for p-type doping at levels ≥1020 cm-3. We have achieved p-type doping levels as high as 5×1020 cm-3 using C, which is introduced through the use of trimethylgallium (TMG) during metalorganic molecular beam epitaxy (MOMBE) growth of GaAs. By utilizing the atomic planar doping method, we have also been able to grow C-doped spikes with hole concentrations as high as 7×1019 cm-3, with a full width at half maximum of ∼50 Å at 300 K. This level is among the highest reported for planar doping. By switching out the TMG, and switching in the triethylgallium (TEG) to continue to growth of C-free GaAs, we have grown sandwich-type structures with C levels of 1020 cm-3, which fall off within 210 Å to C levels of <1017 cm-3. High temperature annealing of such structures reveals a C diffusion coefficient of <10-16 cm2 s-1 at 950°C, in agreement with other reports. The electrical properties of layers annealed at high temperatures appear to be influenced by the presence of strain arising from the high C concentration. X-ray diffraction patterns of 3 μm layers doped in excess of 1020 cm-3 show a lattice constant which corresponds roughly to that calculated by assuming a Vegard law mixture of GaAs and 0.7% GaC. Preliminary results of C-doping of InGaAs will also be discussed. Finally, the usefulness of carbon doping has been demonstrated in ohmic contact formation, Schottky barrier height enhancement in MESFETs and as the base layer in HBTs.

The growth of thin (50–100 Å), C or Zn δ-doped layers on n-type GaAs is shown to yield large enhancements in the effective Schottky barrier height (ΦB) of TiPtAu contacts subsequently deposited on the material. The incorporation of a single C δ-doped layer (p=1.5×1020 cm−3, 50 Å wide) within 100 Å of the surface leads to a barrier height of 0.93 eV, a significant increase over the value for a control sample (0.76 eV). The use of two sequential δ-doped layers leads to an apparent barrier height in excess of the GaAs band gap (ΦB=1.67 eV). This appears to be consistent with the predictions of a unified defect model. Zinc δ doping (p∼3×1018 cm−3) in a similar fashion produces barrier heights of 0.81 eV for one spike and 0.95 eV for two spikes.

We describe reverse bias diode leakage and physical analysis (secondary ion mass spectroscopy, Rutherford backscattering and transmission electron microscopy) data from shallow p+/n junctions made by implanting various doses of BF2 into polycrystalline Si and out-diffusing at various temperature/time conditions, into the underlying (100) Si substrate. The polycrystalline Si is cobalt disilicided to provide the first level of metallization. Subsequent metallization consists of electroless plated Co followed by sputtered Al. The minimum process specifications giving good junction quality (reverse bias diode leakage current density ≤10 nA/cm2 at 10 V) are 5×1015 cm−2 BF2 outdiffused at 900 °C, 30 min with a junction depth 1200 Å below the polycrystalline Si. At these conditions, most of the BF2 (>90%) uniformly redistributes in the polycrystalline Si. There is a slight increase in B concentration towards the polycrystalline Si/Si interface which is characterized by a porous, <50-Å thin silicon oxide interface. The polycrystalline Si/Si interface is also characterized by a peaked fluorine concentration. The lower boron dose diffusing into the Si substrate may be used to explain the observation that these p+/n diodes are ∼10 times leakier than similar n+/p diodes made by outdiffusing As or P implanted from the polycrystalline Si diffusion source.

In the metalorganic chemical vapor deposition of GaAs there is increasing interest in replacing arsine with a less toxic arsenic source. However, GaAs films grown with metalorganic arsenic reactants usually contain significantly higher levels of carbon than films grown with arsine. Using 50% isotopically enriched13C trimethylarsenic (TMAs), we report the first direct evidence that the methyl groups from TMAs are a major source of the carbon observed in the GaAs films. The measured13C concentration in these films was 5 x 1016 cm.3 Conversely, incorporation of13C was not detected when 99%13C-enriched methane was added to the source gases during growth of GaAs with arsine in place of the13C-TMAs.

The stability of the electrical and structural properties of GaAs directly deposited on Si by metalorganic chemical vapor deposition is examined. Extended annealing at 900 °C leads to substantial diffusion of Si across the heterointerface while under the same conditions there is no significant motion of Si incorporated as a dopant into the GaAs surface region. The degree of enhancement of Si diffusion ranges from a factor of ∼250 for 0.5-μm-thick GaAs films to ∼5 for 4-μm-thick films. The annealing time and GaAs layer thickness dependence of Si diffusivity near the interface is consistent with a defect-modulated mechanism. A large fraction of this mobile Si is electrically inactive.