We demonstrate integration of GaAs-AIGaAs multiple quantum well modulators to silicon CMOS circuitry via flipchip solder-bonding followed by substrate removal. We obtain $J.5~0 device yield for 32 x 32 arrays of devices with 15 micron solder pads. We show operation of a simple circuit composed of a modulator and a CMOS transistor.

No abstract

Electroless gold plating in an alkaline bath containing potassium borohydride as the reducing agent has been found to produce tightly adherent plated films on GaAs, AlxGai-xAs, GaP, and InP. In this process, the substrate is activated in an acidified solution of palladium chloride prior to plating. The same process applied to Si or Ge produces poorly adherent films. A transmission electron microscope study of activated GaAs and Si surfaces revealed markedly different morphology and distribution of Pd grains on these two substrates. The adhesion of the subsequently plated electroless gold is discussed in terms of these observations and a model involving atomic interactions at interfaces.An electroless (autocatalytic) gold plating process utilizing potassium borohydride as the reducing agent has been previously described (1) and used for plating on noble metals such as Pd, Pt, Rh, or Au, or on active metals such as Cu and its alloys (2). The noble metals catalyze the plating reaction, and this fact has led to a method of activating certain metallic surfaces, such as tantalum. An activator solution for that purpose was developed by DeAngelo, et al. (3), using acidified palladium chloride containing HF and acetic acid. The HF removes oxide from the surface, allowing activation of the freshly cleaned surface in situ. We have found that treatment in similar solutions will induce electroless gold plating on a variety of III-V compound crystals. Comparisons of the electroless gold films (about 1 #m thick) plated by this method with evaporated films of Cr-Au or Ti-Pt-Au indicate that the electroless gold is more tightly adherent, more reproducible, and less expensive to prepare than the evaporated films. These films pass the "Scotch tape test" and cannot be removed by rubbing or scratching with a metal point.It is shown in this paper that the activation with the PdCl2-containing solution involves the reduction of Pd 2+ ions to metallic Pd by galvanic displacement. With this activation mechanism one can easily imagine that the overall adhesion of the subsequently plated electroless gold layer will be influenced by the adhesion of the catalyst to the substrate. In fact, there are substrates to which electroless gold adheres poorly. An example is silicon activated by the same process as is used for GaAs. In order to obtain insight into the difference in adhesion on different semiconductor substrates, the morphology and distribution of the Pd catalyst on the substrates were studied using a transmission electron microscope (TEM). The results will be discussed in relation to the mechanism of catalyst deposition. A model involving atomic interactions at various interfaces is presented to explain the observed strong adherence of electroless gold to the III-V compounds. Solution Preparation and Plating ProcedureIn the procedures described here, the plating bath was held constant. The bath composition is given in Appendix 1 (2). We have found that it is convenient to prepare a 5• concentrated solution for storage. It is important t...

No abstract

No abstract

No abstract

No abstract

The effect of 300-keV proton bombardment and subsequent annealing on the optical absorption and electrical resistivity of bulk p-type (p =2X 10 18 and 1.4 X 10 19 cm-~ GaAs has been studied. Proton doses were in the range 10 13 _10 17 cm-'. It is found that bombardment-induced optical absorption increases monotonically, but sublinearly, with proton dose. The shape of the optical transmission spectrum indicates that bombardment creates a distribution of energy levels extending into the forbidden gap. Activation energies for annealing of the optical absorption have been determined from isothermal annealing data and range from '" 1.5 to '" 3.4 eV, which indicates that at least two kinds of defects are involved. Current-voltage measurements show that the average electrical resistivity of the bombarded layers goes through a maximum at '" 2.5 X lOs n cm at a proton dose of '" 3 X lOIS cm-'. It is shown that annealing can eliminate the bombardment-induced optical absorption while still retaining a high electrical resistivity. The optimum annealing time and temperature is a function of the proton dose. From these results a set of useful conditions for the proton-bombardment fabrication of stripe-geometry GaAs lasers is determined. Many qualitative similarities exist between the results presented here and those obtained in GaP, which are presented in the following paper.(at 9360 A) for the two regions are deSignated 'To and 'Tb'