Samenvatting
Plasma enhanced chemical vapour deposition (PECVD) is a technology allowing the deposition of thin conformal dielectric barrier coatings coping with the miniaturization of consumer electronics. However, plasma deposition is a complex process with many different interactions in the plasma phase and with substrates exposed to the plasma,making a high-quality deposition of thin performant dielectric barrier coatings a challenge. These plasma deposited polymers have a special place amongst polymers due to their unique properties and deviation from the conventional polymerized starting precursors. The intention of this dissertation is the development of a halogen-free dielectric barrier coating using scientifically and industrially relevant methods for the precursor selection and for the optimization of the plasma deposition. The deposition method,centralized around the Yasuda factor, is validated by investigating the relationship between the external plasma process parameters and the internal plasma properties using a Langmuir probe as well as by a comprehensive chemical, thermal and dielectric analysis of the plasma deposited coatings.
In the first part of this work, the processing-internal plasma property relationship was investigated for an inert precursor (argon) and a reactive precursor (ethylene)by studying the effect of the following internal plasma properties:electron temperature, ionflux, ion density, Debye length and floating potential of the probe. The effect of three processing variables, the processing pressure, precursor flowrate and electrical power,on these internal properties was investigated for the inert precursor,revealing the pressure and power to have a significant impact. The pressure impacts the heating mechanism of the electrons,resulting in a significant impact on the internal plasma properties. All plasma parameters also reveal a power dependence, with increasing power an increase in electron temperature, ionflux and ion density is observed, which increases the fragmentation and ablation processes at surfaces in contact with the plasma due to the presence of higher energy electrons,while the Debye length reduces due to an increase in ion density. The precursor flowrate exhibited no significant impact on the plasma properties when other variables remained constant.Recalculation to the Yasuda factor revealed a similar dependence as observed for the power,however due to the Yasuda factor being an energy-related factor per amount of molecules,a significant impact was observed when changing the flowrate. For the inert gas and the reactive precursor,a similar dependence of the plasma properties on the Yasuda factor was observed,though different values were obtained related to the difference in nature of the precursors (atomic versus molecular). Interestingly,for both precursor an independence of the investigated process variables on the floating potential was observed,revealing the floating potential to be determined by the reactor geometry, and especially the powered electrode/ground electrode surface area ratio.
Using the precursor selection method proposed in this work,1,3-diisopropenylbenzene (DIPB) was selected and plasma deposited, yielding the Yasuda plot of plasma deposited(pd) DIPB. For three representative Yasuda factors and three process temperatures pd-DIPB films were deposited. In addition, films were deposited using intermediate CASING (Crosslinking by Activated Species of INert Gases)steps, a procedure developed prior to this work.
All deposited films were characterized using Raman and FTIR spectroscopy. The highest precursor functionality retention was observed for the lowest Yasuda factor. For higher Yasuda factors,more fragmentation occurred,resulting in functionality loss and an increase in the heterogenous character of the pd-films. At the lowest Yasuda factor and lower processing temperatures, the inclusion of unreacted precursor was observed. This unreacted material consumes high energy electrons for initiating the polymerization reaction, plasma-induced polymerization at the surface, reducing the crosslinking efficiency of CASING steps. At higher Yasuda factors,an increase in low molar mass and volatile content was observed, attributed to a more extensive fragmentation,resulting in a more heterogeneous character of the plasma-deposited films.
With MT-DSC,the residual reactivity and the glass transition temperature Tg were characterized,revealing a significant dependence of the Yasuda factor and processing temperature. With increasing Yasuda factor,an increase in the residual reactivity was found,while with increasing processing temperature the residual reactivity decreased, suggesting two major contributors to the residual reactivity: unreacted precursor and long-living radicals present in the glassy pd-film.The contribution arising from radical recombination revealed a Yasuda factor dependence and processing temperature independence.The CASING process increased the radical concentration inthe pd-DIPB films,confirming the plasma-induced polymerization efficiency of the CASINGprocess for both unreacted and plasma-deposited material. With increasing Yasuda factor,an increase of the Tgwas found due to more extensive fragmentation and increased crosslinking. At the highest Yasuda factor,the Tgcould no longer be detected. A significant impact of CASING on the Tgwas only observed for the highest processing temperature and is related to the lower amounts of unreacted precursor consuming less high energy electrons, thus enabling the crosslinking of the pd-DIPB material.
Finally, dielectric analysis using electrochemical impedance spectroscopy and breakdown strength tests were performed to characterize and investigate the crucial processing-dielectric property relationships. All plasma-deposited films exhibited dielectric constants around the estimated dielectric constant, with the lowest value being observed for higher processing temperatures and lower Yasuda factors, conditions for which precursor functionality retention is observed in combination with the reduced presence of unreacted material. Breakdown testing,being more prone to both defects and the dielectric properties of the pd-DIPB films,revealed a similar trend, the highest breakdown strengths were found for films deposited at lower Yasuda factors and higher processing temperatures,with an optimum for the middle Yasuda factor, exhibiting the least defects. The CASING process had an overall negative impact on the dielectric constant, which was attributed to an increase in the radical concentration. However, the detrimental influence of defects in the breakdown testing does impose difficulties for correlating the data obtained for different conditions,as defects include the incorporation of unreacted material(at lower Yasuda factors),and/or the incorporation of particles due to plasma-phase reactions (at higher Yasuda factors).
The industrial benchmarking with an external and internal reference was done by performing short circuit testing in different liquids. The newly developed pd-DIPB films were deposited on a PCB using the plasma conditions yielding the highest breakdown strength. The new halogen-free pd-DIPB coating was found to exhibit much higher resistance values at a 1.3 μm thickness, than the external reference, a 1.0 μm thick Parylene C coating, and even than the internal pd-fluorinated Nanofics120 S coating at 3.6 μm thick, remaining stable over time and for different liquids.
In the first part of this work, the processing-internal plasma property relationship was investigated for an inert precursor (argon) and a reactive precursor (ethylene)by studying the effect of the following internal plasma properties:electron temperature, ionflux, ion density, Debye length and floating potential of the probe. The effect of three processing variables, the processing pressure, precursor flowrate and electrical power,on these internal properties was investigated for the inert precursor,revealing the pressure and power to have a significant impact. The pressure impacts the heating mechanism of the electrons,resulting in a significant impact on the internal plasma properties. All plasma parameters also reveal a power dependence, with increasing power an increase in electron temperature, ionflux and ion density is observed, which increases the fragmentation and ablation processes at surfaces in contact with the plasma due to the presence of higher energy electrons,while the Debye length reduces due to an increase in ion density. The precursor flowrate exhibited no significant impact on the plasma properties when other variables remained constant.Recalculation to the Yasuda factor revealed a similar dependence as observed for the power,however due to the Yasuda factor being an energy-related factor per amount of molecules,a significant impact was observed when changing the flowrate. For the inert gas and the reactive precursor,a similar dependence of the plasma properties on the Yasuda factor was observed,though different values were obtained related to the difference in nature of the precursors (atomic versus molecular). Interestingly,for both precursor an independence of the investigated process variables on the floating potential was observed,revealing the floating potential to be determined by the reactor geometry, and especially the powered electrode/ground electrode surface area ratio.
Using the precursor selection method proposed in this work,1,3-diisopropenylbenzene (DIPB) was selected and plasma deposited, yielding the Yasuda plot of plasma deposited(pd) DIPB. For three representative Yasuda factors and three process temperatures pd-DIPB films were deposited. In addition, films were deposited using intermediate CASING (Crosslinking by Activated Species of INert Gases)steps, a procedure developed prior to this work.
All deposited films were characterized using Raman and FTIR spectroscopy. The highest precursor functionality retention was observed for the lowest Yasuda factor. For higher Yasuda factors,more fragmentation occurred,resulting in functionality loss and an increase in the heterogenous character of the pd-films. At the lowest Yasuda factor and lower processing temperatures, the inclusion of unreacted precursor was observed. This unreacted material consumes high energy electrons for initiating the polymerization reaction, plasma-induced polymerization at the surface, reducing the crosslinking efficiency of CASING steps. At higher Yasuda factors,an increase in low molar mass and volatile content was observed, attributed to a more extensive fragmentation,resulting in a more heterogeneous character of the plasma-deposited films.
With MT-DSC,the residual reactivity and the glass transition temperature Tg were characterized,revealing a significant dependence of the Yasuda factor and processing temperature. With increasing Yasuda factor,an increase in the residual reactivity was found,while with increasing processing temperature the residual reactivity decreased, suggesting two major contributors to the residual reactivity: unreacted precursor and long-living radicals present in the glassy pd-film.The contribution arising from radical recombination revealed a Yasuda factor dependence and processing temperature independence.The CASING process increased the radical concentration inthe pd-DIPB films,confirming the plasma-induced polymerization efficiency of the CASINGprocess for both unreacted and plasma-deposited material. With increasing Yasuda factor,an increase of the Tgwas found due to more extensive fragmentation and increased crosslinking. At the highest Yasuda factor,the Tgcould no longer be detected. A significant impact of CASING on the Tgwas only observed for the highest processing temperature and is related to the lower amounts of unreacted precursor consuming less high energy electrons, thus enabling the crosslinking of the pd-DIPB material.
Finally, dielectric analysis using electrochemical impedance spectroscopy and breakdown strength tests were performed to characterize and investigate the crucial processing-dielectric property relationships. All plasma-deposited films exhibited dielectric constants around the estimated dielectric constant, with the lowest value being observed for higher processing temperatures and lower Yasuda factors, conditions for which precursor functionality retention is observed in combination with the reduced presence of unreacted material. Breakdown testing,being more prone to both defects and the dielectric properties of the pd-DIPB films,revealed a similar trend, the highest breakdown strengths were found for films deposited at lower Yasuda factors and higher processing temperatures,with an optimum for the middle Yasuda factor, exhibiting the least defects. The CASING process had an overall negative impact on the dielectric constant, which was attributed to an increase in the radical concentration. However, the detrimental influence of defects in the breakdown testing does impose difficulties for correlating the data obtained for different conditions,as defects include the incorporation of unreacted material(at lower Yasuda factors),and/or the incorporation of particles due to plasma-phase reactions (at higher Yasuda factors).
The industrial benchmarking with an external and internal reference was done by performing short circuit testing in different liquids. The newly developed pd-DIPB films were deposited on a PCB using the plasma conditions yielding the highest breakdown strength. The new halogen-free pd-DIPB coating was found to exhibit much higher resistance values at a 1.3 μm thickness, than the external reference, a 1.0 μm thick Parylene C coating, and even than the internal pd-fluorinated Nanofics120 S coating at 3.6 μm thick, remaining stable over time and for different liquids.
Originele taal-2 | English |
---|---|
Toekennende instantie |
|
Begeleider(s)/adviseur |
|
Datum van toekenning | 21 jun 2022 |
Status | Published - 2022 |