首页   按字顺浏览 期刊浏览 卷期浏览 Mechanism of formation and some surface characteristics of thin polymer films formed on...
Mechanism of formation and some surface characteristics of thin polymer films formed on metal surfaces by electron bombardment

 

作者: S. Frost,  

 

期刊: Faraday Special Discussions of the Chemical Society  (RSC Available online 1972)
卷期: Volume 2, issue 1  

页码: 198-209

 

ISSN:0370-9302

 

年代: 1972

 

DOI:10.1039/S19720200198

 

出版商: RSC

 

数据来源: RSC

 

摘要:

Mechanism of Formation and Some Surface Characteristics Electron Bombardment of Thin Polymer Films Formed on Metal Surfaces by BY S. FROST, W. J. MURPHY,? M. W. ROBERTS," J. R. H. Ross AND J. H. WOOD School of Chemistry, The University of Bradford, Bradford Received 23rd June, 1972 The formation of thin films by electron bombardment has been studied and some of the important factors controlling the nature and surface characteristics of the films have been identified. A number of experimental techniques have been developed in order to obtain information on both the species likely to be involved in film growth and also the nature of the film surfaces. The surface character- istics have been monitored by contact angle studies, The possible importance of negative ions in film formation is illustrated with Si(CH&, where over 20 different negative ions were detected, the major ones being H-, C2H- and Sic;.Such results also have an important bearing on electron affinity data obtained by the magnetron method. Thin films formed from perffuorobut-Zene have surface characteristics which vary with the monomer pressure, the substrate temperature, and electron and ion energies. Information on the relative importance of the role of negative ions compared with electron and adsorbed monomer reaction is obtained. Some suggestions are discussed regarding possible surface structures inferred from wetting characteristics. Thin polymer-like films have been of interest for some years and methods of preparation have included glow discharge, u.-v.radiation and electron bombard- ment.', The object of this investigation is an understanding of the mechanism by which thin films (e g., < 1000 thick) are formed by electron-monomer interaction, the factors that control both the adhesion of the film to the substrate (in this work a metal) and the surface characteristics of the film. Although the monomer tetramethyl tetraphenyl siloxane is technologically important, we have chosen to start with struc- turally simpler, but related molecules, e.g., tetramethyl silane Si(CH& Such thin films are of interest 3* as insulators, dielectrics, computer storage elements and " chemically inert " coatings. We have adopted a number of distinct experimental approaches. Adsorption studies of the monomer (in the presence and absence of an electron beam) on a number of atomically clean metal substrates give direct information on the bonding at the metal/polymer interface and therefore are relevant to the question of adhesion.Secondly, we considered it important to investigate whether any reactions occurring at the electron emitting filament were significant in determining the nature of the polymer film. Thirdly, the surface characteristics of the thin films were monitored by determining the critical surface tension yc which Zisman 5 9 considered to reflect the molecular naure of solid surfaces. The possible variation of these surface char- acteristics with the conditions of film preparation (monomer pressure, substrate temperature, electron energy) was investigated. j. present address : Department of Chemistry, New University of Ulster, Coleraine.198FROST, MURPHY, ROBERTS, ROSS AND WOOD 199 These objectives led to the pursuit of a number of more fundamental questions, namely the nature of species resulting from molecular interaction with an electron emitting filament and the role of negative surface ionization, the significance of electron affinity values determined by the magnetron method and the significance of the critical surface energy yc of low energy polymer surfaces. EXPERIMENTAL There are a number of distinct experimental problems to be resolved, each with its special requirements. Adsorption studies were performed in a UHV (- lo-' Torr) all glass apparatus with mass-spectrometric facilities for gas analysis and for following H2/D2 exchange.' Identification of the current carriers participating in film formation required us to dis- tinguish between electrons and negative ions.A magnetron similar to that used by Page and co-workers * was constructed for this and was used for pressures in the range loF6 to N Torr. At a later stage of the work, a smaller version of the magnetron was coupled to an EAI Quadrupole mass-spectrometer which had been modified to monitor negative ions. This enabled the molecular nature of the negative ions to be determined; pressures in the range 5x to -5x Torr were used. Electrons emitted by the tungsten filament were diverted by the magnetron in order that they did not interfere with the recording of the negative-ion mass-spectrum. FIG. 1.-Details of the reaction vessel used for " polymer " film formation.Insert : Schematic representation. The thin polymer-like films formed specifically for surface studies were prepared in the reaction vessel shown in fig. 1. It consists of an optical glass plate on to which a substrate film of tungsten is evaporated from a filament ; electrical contact to the substrate (anode) is made via platinum wires. The temperature of the substrate can be varied by circulating liquid nitrogen, hot water, etc. The spherical reaction vessel had appendages for an electron gun, a Pirani gauge, facilities for introducing the monomer through a capillary leak and for200 SURFACE CHARACTERISTICS OF POLYMER FILMS fast pumping. The ultimate pressure in this system was -lo-’ Torr. The surface char- acteristics of films formed on the anode were obtained as described elsewhere for bulk solids.The “ standard conditions ” of film preparation were : electron gun (W) filament temperature 2300 K ; anode potential (V*)+ 500 V ; grid potential ( VG)+ 35 V; substrate temperature 298 K; monomer pressure in reaction vessel 5 x Torr ; time of film preparation 30 min. RESULTS AND DISCUSSION CHEMISORPTION STUDIES We have reported recently on the adsorption of tetramethyl silane Si(CH3)4 on tungsten and iron surfaces. Fragmentation of the molecule occurs on W at 293 K with the desorption of H,(g) and CH,(g), while more extensive breakdown occurs at higher temperature. With Fe, Si(CH3), is adsorbed temperatures than with W. We have good evidence dissociated surface species on W include CH3 CH3 CH3 CH3 I / H,C- i \ / Si but is less fragmented at all to suggest that at 293 K the and j The bonds formed to the surface atoms * are strong and are likely to lead to good adhesion. How these surface groupings are perturbed under polymerization con- ditions is not known, but some observations on the influence of electron interaction are relevant.When a W film surface, which had chemisorbed Si(CH3)4 at 295 K, was bombarded with electrons (100 pA at 150 V) for - 16 min, H,(g) and CH,(g) were desorbed. The number of product molecules corresponded to one per impinging electron. On the other hand, when physically adsorbed Si(CH3)4 present in muItilayers on tungsten at 77 K was bombarded with electrons, the extent of desorption of CH4(g) and H,(g) was small but extensive and efficient conversion of the monomer to “ poly- mer ” occurred.The latter is concluded from the fact that not all the Si(CH,),(g) was recovered on warming the “ bombarded layer ” to 293 K. We might therefore anticipate quite distinct surface properties with films formed under different condi- tions and we will return to this point later. Further Si(CH3)4 was adsorbed reversibly. ROLE OF NEGATIVE IONS IN FILM FORMATION That ions are involved in film growth was first suspected from current measure- ment during the interaction of Si(CH3)4 (P- to 5 x Torr) with an electron- emitting W filament. The rate of disappearance of Si(CH3)4 was close to the rate of molecular impingement with the W filament. This could clearly have been fortuitous, but its significance was emphasized when the current increased with increasing Si(CH3)4 pressure.By making use of the magnetron to distinguish between electron and ion current, we concluded that with the majority of the molecules studied, including Si(CH3)4, a high proportion of the carriers are negative ions in the pressure range Torr. These ions are accelerated to the anode (fig. 1) and presumably participate in film formation. The next question to consider was the nature of these ions ; we will not concern ourselves here with the detailed mechanism by which they form. The molecular nature of the ions leaving the W filament was investigated using the modified quadrupole mass-spectrometer coupled to the magnetron ; table 1 The nature of these current carriers is clearly important.toFROST, MURPHY, ROBERTS, ROSS AND WOOD 20 1 summarizes the major negative ions (with their relative abundances) observed with .- c .$-a r C 'C H E 0 0 -5 E 6 H tetramethyl silane. - With-perfluorobut-2-ene the F; in the respective proportions 1500 : 28 : 5 : in this molecule will become clear later. u u B H U C H 3- C H r -CH2 I r.CH- -CH-,;r.CH ;z: 2 B 2 -CHOH -CHg -COsCHg r..CHm TABLE 1 .-NEGATIVE IONS OBSERVED WITH Si(CH3)4 -2000 K main ions were F', C ; , CF- and 80. The reasons for our interest AND A w FILAMENT MAINTAINED AT ion H C CH CH2 C2H SiH SiH2 relative intensity (arb units) 760 8 23 6 690 103 48 ion SiCH SiCH2 SiCH3 c4 C4H Sicz SiC2H SiH3 1 SiC2H2 f4 c3 45 SiC3H 69 C3H 38 SiC3H2 34 C3H2 12 Si(CH3)4 140 relative intensity (arb units) 83 8 4 24 120 345 103 c..THE POLYMER SURFACE. WETTING STUDIES BASIS OF THE APPROACH Zisman et aL5* 6* lo have shown that for a homologous series of pure liquids, a plot of cos 8 against yLv (8 = contact angle, yLv the liquid surface tension) was linear with an intercept on the cos 0 = 1.0 line which Zisman designated yc and termed the critical surface tension of the solid. yc has been shown to be a useful empirical parameter related to the surface energy of the solid ys. Of particular significance to the present work was Shafrin and Zisman's correlation of yc with the molecular constitution of the solid surface (fig. 2). This " wettability spectrum " shows a distinct empirical correlation between yc and the chemical constitution of low-energy surfaces. The lowest values of yc (i.e., less wettable and higher contact angle) are202 SURFACE CHARACTERISTICS OF POLYMER FILMS obtained with surfaces comprising C and F atoms only (6 to 19 dyn cm-l) ; hydro- genation increases yc to the range 15 to 28 dyn cm-l.The next higher range of yc values is observed with surfaces containing only C and H atoms (22 to 35 dyn cm-') while higher values are obtained with C, C1, H, F; C, 0, H and C, N, 0, H surfaces. We note in particular the following : CF3 (yc = 6 dyn cm-l), CF2 (yc N 18 dyn cm-'), CH3 (yc N 24 dyn cm-l) and CH2 (yc N 30 dyn cm-l). the wetting behaviour of a number of polymer surfaces using both butanol solutions and a series of pure liquids, the yc values observed with the solutions were much lower than with the pure liquids and what is more significant, were virtually independent of the solid.This led naturally to the conclusion that yc, at least with the solutions, reflected the perturbation of the molecular nature of the surface by adsorption. This was shown convincingly to be the case when yc for any one (low energy) solid varied in a predictable manner when solutions of BuOH, MeOH and 2,2,2, trifluorethanol were used (table 2). It followed that the adsorbed molecules were likely to be oriented with the hydrocarbon groupings pointing away from the solid surface, viz : polymer- - - - -HO-CH2-CH3. Although such an orientation, with the hydrocarbon end pointing away from the surface, is not generally accepted, Ottewill and Vincent have suggested recently that both orientations may occur in the polystyrene + BuOH system.Clearly, yc obtained with solutions is not necessarily diagnostic of the molecular nature of the film surface. When we investigated TABLE 2.-yc VALUES (dyn crn-') FOR A NUMBER OF SOL^ SURFACES DETERMINED BY DIFFERENT LIQUIDS AND SOLUTIONS pure liquids BuOH EtOH MeOH 2,2,2,-trifluoroethano polystyrene 41-44 25.3 - 23.0 - polymethylmethacr ylat e 39-41 26.0 26.5 23.0 24.7 polyethylene [- CH2-CH2-], 31-32 27.5 - - 27.5 Further support for adsorption being important was obtained by combining the Gibbs adsorption isotherm with the conditions determining the equilibrium contact angle 8 to give eqn (1) where r2 is the surface excess of solute (component 2) at the interface. At low concentrations r2 is equivalent to the amount of solute adsorbed.Thus, gradients of plots of yLv cos 8 against In a (a = activity of solute) will be related to the difference in the surface excess of solute molecules at the solid-liquid and solid-vapour interfaces. For example, a variation in the contribution of adsorption at the solid/liquid and solid/ vapour interfaces with concentration would be reflected in the form of the plot. Such a variation is observed with different surfaces, and is evident in fig. 3, 4 and 5. For the present paper, however, we base our discussion mainly on the variation of Ow, the the contact angle observed with pure H20 (yLv N 72 dyn cm-l). STUDIES OF THIN FILMS BASED ON PERFLUOROBUT-2-ENE The possibility of obtaining from perfluorobut-2-ene, CF3-CF=CF-CF3, thin films whose surface properties were analogous to polytetrafluoroethylene (PTFE), yc E 18 dyn cm-l, and apparently involving CF2 and possibly CF3 surface groupings (table 3) was attractive. The physical properties of perfluorobut-Zene enabled itFROST, MURPHY, ROBERTS, ROSS AND WOOD 203 to be handled quantitatively in our mass-spectrometric and magnetron studies.Thin films were prepared using the " standard procedure " already outlined and their surface characteristics investigated by contact angle studies using various alcohol solutions and pure water. These characteristics were compared with films prepared from other related fluorocarbon " monomers " and also with those reported for fluorocarbon surfaces thought to have CF3 and CF2 groupings (table 3).The influence of such preparation conditions as monomer flux, substrate temperature and accelerating potentials were also investigated. Fig. 3(i) shows " Zisman " plots using BuOH solutions for two samples each of films formed from (a) octafluorobutane, (b) perfluorobut-Zene and (c) tetrafluoro- methane. These plots can be compared with those for films prepared from (a) methane, (b) ethane, (c) ethylene, (d) butane, (e) butadiene and (f) neopentane (fig. 3(ii)). We conclude that although the apparent yc is not particularly sensitive to the monomer, whether hydrocarbon or fluorocarbon, (not surprising in view of our adsorption model ') there are distinct differences in the manner in which cos 8 varies with yLv within the fluorocarbon series (fig. 3(i)). In all cases when Ow decreased, the apparent yc also decreased. 1.c 0.E 0, t 8 0 . 4 0.2 Table 4 summarizes Ow and yc values, the latter obtained with pure liquids, for hydrocarbon surfaces with characteristic surface groupings and also for thin films formed from various hydrocarbons. It is clear that as the degree of unsaturation of204 SURFACE CHARACTERISTICS OF POLYMER FILMS I .o 0. s 0.7 m y, 0.5 8 0.: 0. FIG. 3.-(ii) Zisman plots (as in (i)) for thin films formed from (a) methane (0), (b) ethane (01, (c) ethylene m), (d) butane (+), (e) butadiene (- - -), (f) neopentane (- - -). 0.8 0.6 y, 0.4 8 0.2 0 Q 2 0 4 0 6 0 8 0 YLvldYn cm-' c \ x 10 IV xz x 103 FIG. 4.--Influence of the variation of the monomer flux on the wetting behaviour of perfluorobutene-2- films.(a) Zisman plot, (b) a plot of [(F2)SL- (r2)sv] against the mol fraction XZ.FROST, MURPHY, ROBERTS, ROSS AND WOOD 205 8 LOOOV . \"I -6.0 1 FIG. 5.-Influence of the anode potential on (a) the Zisman plot and (b) a plot of [(r2)s~- (I',)sv] against the mol fraction X,. the surface grouping increases (i.e., CH3 to CH2 to CH) & decreases from - 110" to - 55", and yc increases from -22 dyn cm-' to - 34 dyn cm-I, reflecting an increase in surface energy. TABLE 3.-cOMPARISON OF Bw(deg) AND ')Jc (dyncrn-') FOR (U) SOME THIN FLUOROCARBON POLYMER FILMS WITH (6) " STANDARD " FLUOROCARBON SURFACES OW Yo * (a) thin films (monomer) CF4 69.0 CF,CF=CF CF3 59.0 - 35 F2 F2 I I I I 49.5 - (b) perfluoroacid monolayer (--CF,) 102.0 -6 (-CF2-) 108.0 - 18 poly tetrafluoroet h ylene * values obtained using pure liquids206 SURFACE CHARACTERISTICS OF POLYMER FILMS On this basis we generalize and suggest that a surface which exposes " unsatur- ated " groupings will be of higher energy than that exposing " saturated " groupings.We therefore accept that Ow will reflect both the molecular nature of the surface and yc, TABLE 4.-cOMPARISON OF Ow(deg) AND 'yc(dyn Cm-') FOR (a) SOME " STANDARD " HYDRO- CARBON SURFACES AND (6) THIN FILMS FORMED FROM A SERIES OF HYDROCARBON MONOMERS ow Y C * (a) standard hydrocarbon surfaces CH3 (hexatricontane) CH3 (paraffin) CH2 (polyethylene) CH2, CH (polystyrene) CH (anilene monolayer) (b) thin films (monomer) CH4 C2H6 C4HIO 110 22 108 22 94 32 87 34 53 35 74 35 71 69 - - * obtained using pure liquids and consider what influence film preparation conditions have on the magnitude of Ow and the adsorption characteristics as assessed from the application of the Gibbs adsorption isotherm.In the present paper the latter is used merely as a qualitative pointer of the adsorption behaviour. VARIATION OF MONOMER FLUX A marked variation in the wetting behaviour was observed (fig. 4) as the flux (pressure) of perfluorobut-2-ene was increased by factors of 10 and 20. 8, increased, which is in keeping with the view that increasing the flux rate encouraged the formation of a more saturated surface (see later). Analysis of the data in terms of the Gibbs- Adsorption Isotherm provides further evidence regarding variation in the surface characteristics (fig. 4). VARIATION OF ACCELERATING POTENTIALS Fig.5 shows that both the Zisman plots and the Gibbs adsorption isotherm data are influenced by the potential V, applied to the anode. As the anode potential increases from 300 to 1000 V, the contact angle Ow for pure water (yLv N 72 dyn cm-l) decreases. In other words, as V, increases the film surface becomes more wettable. These changes in Ow are also reflected in the Gibbs adsorption isotherm plotted according to eqn (1) (fig. 5). When the grid potential V, was varied, keeping the anode potential fixed at the " standard value " of 300 V, the surface characteristics of the thin films formed changed. The contact angle of water, for example, decreased as the voltage VG was decreased from 300 to 100 to 35V. This change in surface behaviour was also reflected in the Gibbs adsorption plot.SUBSTRATE TEMPERATURE Preliminary data obtained with the substrate maintained at 77 K indicate that the film exhibits (fig. 6) a very low energy surface (y,<20 dyn cm-l, see fig. 2). In thisFROST, MURPHY, ROBERTS, ROSS AND WOOD 207 case it would appear that films formed from perfluorobut-2-ene are more similar to PTFE (table 3) and could indicate that we are concerned here with the interaction of electrons with multilayers " adsorbed " on the substrate. Extensive adsorption of the monomer will not occur at 298 K and so electron/adsorbed monomer interactions will be less significant under these conditions. 298K 77K 3 FIG. 6.-The influence of substrate temperature on the surface characteristics as assessed from contact angles (0) observed with aqueous solutions of BuOH.Apparent y, values are also indicated. GENERAL CONCLUSIONS As to the mode of formation of the thin polymer films, there is clear evidence for the participation of gaseous negative ions in the growth process. Previous work has emphasized the formation of active species by electron interaction with adsorbed monomer, Haller and White arguing that growth involved a surface reaction between adsorbed monomer and active sites (positive ions) produced by electron bombard- ment. With condensed multilayers of perfluorobut-2-ene at 77 K we believe that the electron/adsorbed monomer reaction is important, but at 298 K ion-ion reactions appear to be more significant. We shall return to this point. Ionized species have of course been freely invoked in studies of polymer formation induced by electrical discharges but neither the nature nor the abundance of the ions involved has been clarified. The surface characteristics of the polymer films are clearly dependent on the nature of the monomer and ions, the monomer flux and substrate temperature. Infra-red studies did not give any evidence on their structure.We therefore have relied heavily on contact angle studies and the way in which various parameters control the wetting behaviour and presumably their molecular nature. For example, the anode potential (fig. 1) will determine the energy of the ions and electrons so that an increase in V, is likely to lead to more extensive cross-linking, giving a surface of higher energy with a smaller contact angle 8,; this is observed (fig.5). Variations in the grid potential will determine the ratio ze/& reaching the substrate as well as the total flux Ie+IN. Thus when V, decreased from 300 to 35 V the electron and negative ion currents will also decrease. We do not, however, know how the ratio varies; it will depend on such factors as the electron affinities S of the desorbed ion(s), work function q5 of the filament, etc., terms which are involved in the relationship (eqn (2)) which we have considered dN' ew-4o-l- JZ) - = AB,Eexp dt kT208 SURFACE CHARACTERISTICS OF POLYMER FILMS elsewhere.12 Although further studies to elucidate the detailed mechanism are envisaged, and also an investigation of the fate of the ions (e.g., those listed in table 1) impinging on the substrate surface, tentative mechanisms for the formation of two polymer films formed from perfluorobut-2-ene but with distinctly different surface characteristics, are given below.They illustrate our general approach rather than provide firm conclusions. In the case of film formation using perfluorobut-2-ene as monomer, an electron emitting W filament and a substrate maintained at 298 K, we suggest that the main film growth process involves negative ions (e.g., CF-) and extensive cross-linking. The surface groupings would be C4F8(g) * CF-@ 1 3 . 1 I I l l 1 I I I I I l l 11111111111111111 substrate -CF CF-CF CF-CF CF-CF -CF CF-CF unsaturated, which, according to our suggestions based on Zisman's wettability spectrum, exhibit a high surface energy.The value of Ow in this case was -59" and yo using pure liquids, -35 dyn em-l, both values in keeping with a high energy surface. On the other hand, when multilayers of C4F8 are bombarded with electrons at 77 K, no significant concentrations of gaseous negative ions are involved and we suggest that film growth occurs by electron initiation in the adsorbed monomer. This process is very efficient for polymer formation using multilayers of Si(CH3)4. I C,F,(adS) + e-, CF3-CF-CF-CF3 I F3 C-CF-CF-CF3 I llllllllllllllllllllllt t substrate Such films have characteristics approaching those of PTFE, yc using aqueous BuOH solutions was N 18 dyn cm-1 and Ow N 75". The proposed structure exhibits " satur- ated " surface groupings and according to Zisman's wettability spectrum the surface would be expected to be of " low energy ". Finally, our results indicate the need for extreme caution in determining electron affinity data by the magnetron method.* We have explored a number of molecules in addition to those reported here ; in all cases extensive fragmentation is observed. Furthermore, some of the ions are formed by molecule+electron interaction in the gas phase rather than by surface ionization. This is observed. We acknowledge generous support of this work by T.R.W. Inc. (California) through a grant (N.A.S.A. contract 7-717) to one of us (M. W. R.).FROST, MURPHY, ROBERTS, ROSS AND WOOD 209 I. Haller and P. White, J. Phys. Chem., 1963, 67, 1784. D. S. Allam and C. T. H. Stoddart, Chem. in Britain, 1965, 1,410. H. T. Mann, J. Appl. Phys., 1964,35, 2173. R. W. Christy, J. Appl. Phys., 1960,31, 1680. W. A. Zisman, Adv. Chem. Ser. (Amer. Chem. SOC.), 1964,43, 1. E. G. Shafrin and W. A. Zisman, J. Phys. Chem., 1960,64,519. M. W. Roberts and J. R. H. Ross, J.C.S. Faraday I, 1972,68,221. Interscience, N.Y. 1969). W. J. Murphy, M. W. Roberts and J. R. H. ROSS, J.C.S. Faraday I, 1972,68, 1190. * See, for example, Negative Ions and the Magnetron, F. M . Page and G. C. Goode (Wiley- lo M. K. Bernett and W. A. Zisman, J. Phys. Chem., 1959, 63, 1241. l 1 R. H. Ottewill and B. Vincent, J.C.S. Faraday I, 1972, 68, 1533. l2 M. W. Roberts and J. R. H. Ross, Chem. Comm., 1970, 1170.

 

点击下载:  PDF (763KB)



返 回