Summary Ruben Sharpe

The focus of this thesis has been the control of mass transport in microcontact printing (μCP). The archetypal system thereof is the transfer of alkanethiol patterns onto a gold substrate. Studies of the formation of patterned and non-patterned alkanethiol self-assembled monolayers (SAMs) on gold provide vital clues to the dynamic behavior of alkanethiol inks during the microcontact printing process (Chapter 2). The progressive evolution of distinguishable adlayer phases upon an increase in surface coverage indicates the mobility of the ink across the gold. Dense monolayer structures are markedly less mobile and therefore microcontact printed patterns of dense SAMs are stable enough for post-processing without having to take into account any mobility after the initial formation of the pattern. For practical purposes, morphology changes of dense SAM patterns constitute of incorporation of ink at their growing periphery and are, therefore, limited by the supply of ink molecules. For μCP, this entails specifically that the relevant mass transport processes occur during the time of contact between the inked stamp and the substrate.

In μCP, the stamp properties are intricately linked with the overall mass transport processes. The stamp material provides a medium in which the monolayer assembly process takes place, whereas its surface allows for some preorganization. It provides an ink source, and provides mechanical stability to the scaffold pattern, which in principle defines the monolayer pattern. The topographical layout of the stamp also determines local differences in the pathways for ink to eventually assemble on the substrate’s surface, which has consequences for the spreading behavior (i.e. mass transport to non-contact areas).

Poly(dimethylsiloxane) (PDMS), the most widely used stamp material in μCP, was shown to also affect the printing process because it can induce contamination (Chapter 3). The extent of PDMS-induced contamination is found to be highly dependent on the nature of the used ink. Contamination was observed to be of only minor consequence when printing with n-octadecanethiol (ODT), an apolar alkanethiol ink. A markedly higher degree of contamination was found upon printing with 16-mercaptohexadecanoic acid (MHDA), which is, barring a polar headgroup, very similar to ODT. From the here described experiments, it is strongly suggested that the contaminating species themselves are relatively polar in nature, the PDMS being apolar notwithstanding.

The ability of the stamp surface, in some cases, to pre-organize an ink comes into play when discussing some peculiarities in the spreading behavior of MHDA (Chapter 4). Spreading of MHDA was only observed beyond some threshold inking concentration. At lower concentrations, however, the observed contact time-independent feature size increased with increasing concentration. For these concentrations, the edges of the stamp features were also clearly observed to dominate the ink transfer, while the region were ink was transferred increased with increasing concentration. More or less for the same concentration where spreading became observable, there was no longer a clear dominance of the stamp feature’s edges. These observations were rationalized by assuming that only MHDA that was present on the stamp’s surface contributed to the spreading process, that the amount of ink on its surface constituted a finite reservoir and that MHDA was preferentially pre-organized at the edges of the stamp features.

The notion of the importance of a stamp’s surface in the ink transfer process was exploited to selectively block, or expedite, ink transfer by means of a chemical surface modification of the stamp (Chapter 5). With this approach, the recessed regions of a relief stamp, which serve the purpose of ink barriers, could be replaced by chemical barriers. Since the incorporation of a relief in a slab of PDMS has adverse effects on its mechanical stability, and since void regions, as opposed to chemical barriers, cannot be redesigned for specific purposes, this replacement greatly expanded the applicability of μCP. As a proof of concept, barriers were created by the local oxidation of a flat piece of PDMS, with a subsequent modification with a 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS) monolayer. Patterns with feature sizes as small a 500 nm were printed and faithfully reproduced by selective etching. Etch contrast was observed even for inks and conditions for which no contrast is to be expected when relief stamps are used.

Elaborating on the concepts of local blocking of ink transport, or conversely, locally assembling ink for transfer, other options for controlling ink transport were explored (Chapter 6). These options consisted of, one: applying a barrier layer to an otherwise conventional relief patterned stamp so that only the contribution of transport from the stamp surface comes into play (Section 6.3); and two: providing a template monolayer pattern on the substrate surface that gathers, but not permanently binds the ink, thereby increasing the transfer probability from the gas phase to the areas of the substrate bordering these template patterns (Section 6.4). Using the first method, it was found that the conveyance of etch protection (which is a measure of monolayer quality) could be limited to the edges of the contact areas only. Unlike the edge dominance effect, as described in Chapter 4, ink was found to spread from the edges outward, and it was thus found that the feature sizes could be controlled by varying the contact time. Using the second method, similar, process time-dependent, structures could be created. In this case, the mass transport contribution from the gas phase was isolated from the other transport pathways. Controlling a gas phase is conceptually simpler than overseeing all the possible ink/stamp interactions. It offers a convenient route for fabrication of sub-micrometer sized structures without having to take into account the, previously discussed, various stamp-related issues. The smallest feature size thus far produced, in this manner, was ±750 nm but process optimization may easily decrease the obtainable feature sizes.

An important application of μCP is the use of a patterned SAM as an etch resist. The etch-resistant properties of alkanethiol SAMs are based on steric hindrance and their wettability properties. Because of their thinness, the quality of the SAM must be extremely good, and this entails mobility to allow for molecular rearrangements and therefore the inherent ability to spread. Another concept of barrier formation is the use of electrostatic interactions to inhibit etching, which is explored in Chapter 7. Oxidized gold, which may be expected to be negatively charged in alkaline, aqueous environment, was found to inhibit etching in a standard etch bath, which requires negatively charged species to approach the surface. In this way a clear contrast in etch rate between oxidized and non-oxidized gold could be observed. Patterning of substrates was demonstrated by μCP of a reductive ink on uniformly oxidized gold surfaces. Advantages of this approach include the fact that the barrier consists of a multilayer and therefore the barrier properties are averaged in the direction perpendicular to the substrate’s surface, which decreases the sensitivity to local defects. Furthermore, the gold-oxide barrier appeared easy to remove, and its removal may often be unnecessary because of chemical similarities between oxidized, and non-oxidized gold regions. Conceptually, in view of controlling mass transport, other strategies, such as e.g. electrical field-induced patterning, may be used to create a pattern of oxidized and non-oxidized regions that may eliminate the need for an ink altogether.

In conclusion, mass transfer in μCP proceeds via many pathways, in which the stamp plays an important role. It was found to be possible to selectively open or close some of these pathways. This offers possibilities for optimization of the microcontact printing process to suit specific problems in specific applications.