POLYMER-COATED MICROELECTRONICS - COMPARISON OF BULK, SURFACE, AND INTERPHASE CONDUCTIVITIES

DC conductivity measurements were made at 22-degrees-C in three test geometries. The first study probed surface conductivity of specimens equilibrated with dry N2, and with N2 saturated with water, methanol, ethanol or isopropanol. Experiments were performed on interdigitated comb specimens having 140-mu-m wide aluminum metallization lines separated by 140-mu-m on quartz substrates. With dry N2, experimental surface conductivities ranged around the noise floor of our instrumentation, 10(-14) mho. Conductivity increased to 10(-9) - 10(-8) mho after equilibration with the polar compounds, owing to formation of absorbed surface layers. In the second study, comb specimens were coated with commercial siloxane polymers. Electrical measurements made on these samples reflect interphase conductivity, influenced both by bulk conductivity and by specific interactions within the polymer-substrate interphase region. Interphase conductivities of samples equilibrated with dry N2 were at or below the noise floor of our equipment. Equilibration with the polar compounds increased interphase conductivity to 10(-11) - 10(-9) mho, values substantially less than corresponding surface conductivities. These observations are understood in terms of favorable surface absorption of (insulator) polymer segments relative to surface absorption of the polar compounds. The third study measured bulk: i.e., volume, conductivity of dry polysiloxane and of polysiloxane equilibrated with water, methanol, ethanol, and isopropanol. Water sorption had no perceptible effect on bulk conductivity, whereas alcohol sorption increased it substantially. Sorption weight-gain measurements showed the conductivity increase parallels alcohol uptake by polysiloxane. Conductivity and solvent sorption decreased in the order isopropanol > ethanol > methanol > water. Crossplots of conductivity versus weight-gain gave preliminary evidence for a percolation threshold in these systems. Electrostatic calculations were used to predict interphase conductivities from bulk conductivity values. Agreement between calculation and experiment was generally good. Water-equilibrated specimens were a notable exception. Here, experimental interphase conductivities were some 1000 x larger than the calculated values. This points to an electrochemical mechanism, of uncertain origin, localized near the polymer-substrate boundary that produces abnormally large conductivity relative to that in the bulk.