It should be noted that equal deposition around the circumference of the reactor is observed. Kinetic studies have revealed that the rate of polymer deposition at any point in the reactor is constant with time with all other reaction parameters being constant. Figure 6 illustrates the temperature rise in the reactor as a function both of power input and thermocouple position in the reactor. General experience shows that acceptable polymer films can be obtained outside of the reactor volume circumscribed by the coil with little or no temperature rise while operating at fairly high power levels. It has also been shown that monomeric conversion can be controlled by adjusting power inputs and monomer concentration.

At first glance, it would appear that plasma polymerization products possess a much greater thermal stability than corresponding linear polymers. Indeed, this is true after a fashion. Figure 7 shows thermograms for linear and plasma polystyrene heated in air to 700 °C. The plasma polystyrene still retains 40 percent of its initial weight whereas all of the linear polystyrene has decomposed below 400 °C. However, a closer look at these thermograms indicated that the plasma polymer actually began to lose weight initially at a considerably lower temperature than the linear polymer. Therefore, any statements about thermal stability must take into account the temperature range under consideration. The four curves of plasma polystyrene in figure 7 show the effect of polymer film thickness. It becomes apparent that the rate of escape of the decomposition products from the thicker films is a diffusion controlled process. Work currently underway is designed to determine the nature of the apparent instability of these materials. Ways and means are being sought to stabilize the plasma polymer films through quenching techniques in order to flash off any low molecular weight volatiles.

In one aspect of application, some work has been completed recently in investigating plasma deposited polymers for use as membranes. The pur

S

pose of this work was to show that adherent plasma polymer films could be deposited on smooth polymer membrane surfaces thereby altering the membrane's permeability characteristics. Unplasticized linear polyethylene membranes 1-mil (25 μm) thick, were chosen at the start of the experiments. On these membranes was deposited a 1-2 micrometer film of adherent polypropylene.

The permeability of the coated and uncoated membranes to He, Ne, Ar, O2, and CO2 was measured. The results are tabulated in table 1. In comparing the data for coated and uncoated membranes, it is plain that the thin (2 μm) film of plasma deposited polypropylene considerably reduced the permeability constants. These permeability constants are plotted against the square root of the permeating gas' molecular weight in figure 8. Both oxygen and carbon dioxide show a definite dissolution in the uncoated membrane but appear to approach Graham's law behavior with the coated membrane. The slight change in slope between helium and neon in both coated and uncoated cases indicates that some pores have been altered or plugged in the coating process.

The significance of this data must be evaluated by taking into account the two terms of the permeability constant. This constant is the product of the diffusion constant and solubility constant, P= DS. The large reduction in the permeability cannot be explained wholly by the reduction in diffusion through pores which have been altered or plugged by the plasma deposited film. It is known that plasma deposited films are highly cross-linked into a tight three-dimensional network. For example, plasma polypropylene samples have been refluxed in aliphatic and aromatic hydrocarbons, chlorinated hydrocarbons, esters, and ketones for up to a week without appreciable changes in solubility. Thus, the plasma deposited film reduces the solubility of O2 and CO2 which allows diffusion to be the predominant transport mechanism, although it is also significantly reduced. The result is to approach Graham's law.

For these reasons, plasma deposited polymer films may have certain advantages as membrane materials. It appears possible that by tailoring thickness and/or composition of the film, a membrane can be made to very nearly approach Graham's law behavior. Behavior like this could be important in those applications where either a definite ratio of two gases is desired for a particular transfer application or where a specific diffusion rate is required. However, greater film thicknesses could conceivably block gas permeation across the membrane.

A complimentary membrane study involved taking the same membrane material and subjecting it to a high wattage argon plasma. The investigation was aimed at alterations of the permeability constant by interaction of the membrane surface with the nonreactive plasma as opposed to plasma polymer depositions. Results of this work for carbon dioxide and helium permeation are shown in figure 9. Permeability was reduced but not as significantly as was the case with plasma deposited polymer films.

Since the aforementioned properties of plasma deposited polymers are markedly different from

PERMEABILITY AT O AP-CM3CM/CM Hg CM2 SEC x 10

FIGURE 9. Permeability of gases through plasma treated 1-mil polyethylene.

those of their generic polymers, an investigation into the more fundamental property of density was initiated. The densities of nine rf plasma produced polymers were determined in a gradient density column. Generally, these polymers were found to be more dense than conventional polymers produced from the same monomer species via common free radical, ionic, or condensation polymerizations. The results of the density determinations are given in table 2. A density range is given for each polymer

because the density was found to vary with the monomer concentration, the power level and the position within the reactor. For comparison, the densities of both the single crystal and the amorphous conventional linear polymer are reported. Low angle x-ray diffraction patterns of plasma polystyrene and polyethylene indicated that these materials were amorphous. Since the generic polymers of styrene and ethylene are highly crystalline, this indicates that substantial differences in structure and morphology are present. This density data has been correlated with refractive index measurements of the plasma deposited films. Further probing of these properties has resulted in preliminary data from elemental analysis that indicates oxygen and sometimes nitrogen are incorporated into the plasma deposited films. For example, plasma deposited polyethylene was found to contain from 20 to 30 weight percent oxygen. In addition, the density, refractive index and chemical composition have all been found to vary as functions of both reactor design and operation. Current investigations are aimed at determining the source of these elements and the method of their incorporation into the polymer.

5. Current Problems and Future Work

The early work of the authors involving the coating of bioelectrodes and various implantable substrates went well and yielded promising results. It appears that this early work stayed within the then undefined tolerance limits of substrate preparation, reaction parameters and coating techniques. Reports of dissimilar results from other laboratories for the same monomers regarding both changes in the polymers after formation and elucidation of different reaction mechanisms were disturbing since these phenomena had not been observed in our laboratories. However, these problems and many more began to surface once a realistic approach was

taken to the biocompatibility and use applications of these polymers. After several changes in reactor design, more sophistication in electronics and modifications to both the vacuum system and process procedures, it became apparent that experiments were being run out of control. Results which were favorable during one experiment were not favorable during another experiment, although both were run under presumably identical conditions. Most of the problems involved adhesion of the thin polymer films to the substrates. An extreme example of the differences in adhesion is shown in figure 10. Both films were deposited on glass. In micrograph A, a 2 μm plasma polypropylene film is adherent and so smooth that an artifact had to be introduced onto the surface in order to locate the surface in the SEM. Micrograph B shows cracking of another 2 μm plasma polypropylene deposited under the same conditions but in another run. A series of studies involving both the changing of reaction conditions and the effects of position within the reactor upon adhesion led to the conclusion that more reproducible results were obtained with lower power inputs and where the substrates were positioned toward the trailing edge of the plasma field. These findings only partially eliminated the adhesion problem.

It finally became necessary to curtail all routine experiments and launch a full scale investigation (basic studies have been in progress for 2 years) into the cause and effect of the various parameters upon the final properties of the polymer films. Elemental analyses of the plasma polymer films confirmed that the polymer structure contained elements not present in the original monomer. Specifically, oxygen and nitrogen were contained in the final polymer structure. Figure 11 gives some typical results. This investigation is continuing in conjunction with a more fundamental study. Additional studies are underway to determine whether