Method for improving the bonding properties of microstructured substrates, and devices prepared with this method

Abstract

A method for treating the surface of a polymeric substrate, including the following steps: providing a first polymeric substrate; contracting at least one part of one face of the first substrate with some liquid solvent system, the liquid solvent system containing at least a first volatile compound and at least a second compound having a low molecular weight and able to swell and/or soften the polymeric material forming the face; letting at least the volatile compound to evaporate from the face of the first substrate and; contacting the so-treated face of first substrate with a third material.

Description

BACKGROUND OF THE INVENTION[0001]Year after year, microfluidic devices appear more clearly as a valuable alternative to conventional systems for numerous applications. Several commercial systems are already on the market, and their potential applications increase steadily. There are, however, a number of application that seem to resist this tend. For the detection of point mutations in DNA by heteroduplex analysis, for instance, earlier studies demonstrated that the resolution depends critically on the length of the capillary used in separation, because the difference in mobility between two duplex DNA fragments with the same length and a single bp mismatch is very minute. Recently, innovative matrices could increase this difference, and allow highly reliable separations in bench-top DNA sequencers, but this performance could be achieved in 50 cm long capillaries only. In order to transpose this protocol to microchip fort, without compromising the resolution obtained with 50 cm long...

AI Technical Summary

Benefits of technology

[0115]The advantage of the invention with regards to this mode of embossing is that it facilitates the deformation of the surface of the face of the substrate in contact with the master at a lower temperature, at which the other side of the substrate is not yet deformable, or less deformable. This allows to perform embossing at a lower temperature, to shorten the embossing time, and also to reduce the constraints left in the substrate after embossing. These constraints are known to lead to potential problems in further use, such a stress cracking.

[0116]The invention is particularly advantageous in combination to continuous embossing processes known as roll embossing or band embossing.

[0117]In this process, the substrate to be embossed, provided as a film, is kept in tension and pressed onto a hot roll bearing the negative microstructures to be transferred to said substrate. Roll embossing of microstructures is described e.g. in U.S. Pat. No. 6,375,871 to Bentsen. The contact between the substrate and the embossing tool is limited in time, so that the deformation of the substrate must be fast enough, and said substrate must be brought to a relatively fluid state. But then, the substrate would not be able to withstand the tension necessary for its continuous transport U.S. Pat. No. 6,375,871 proposes a solution to this, called extrusion embossing, in which the substrate is delivered in the liquid state between the master and a supporting film, that remains unmelted at the temperature of embossing. This method, however, leads to a composite film, which may not be suitable for many applications. First, it limits the number of materials usable, due to problems of adhesion and compatibility between the substrate and the supporting film. Also, because of differences in dilatation coefficients, composite films prepared this way will tend to deform upon temperature changes. US patent application 2005 / 0029708 to Coyle proposes a solution to this problem, in which the material for making the substrate is provided in a melt state between one roll at a temperature above the material's glass transition temperature, bearing the microstructures to be replicated, and a roll at a temperature lower than the material's glass transition temperature, inducing the formation of a film during the embossing process. The tuning of this process is delicate, however, it is not compatible with all materials and in practice it often also requires a supporting second film. Finally the strong temperature gradient across the substrate induces strong residual stresses, that can lead to stress cracking and other defects in the roll. Finally, the situation of continuous embossing can be improved by band embossing, which allows a longer time of contact between the substrate and the master, but the above problems remain.

[0118]With the process according to the invention, in contrast, the plasticity of the material can be tuned to vary continuously across the substrate, allowing for a plastically deformable upper layer, that will replicate the microstructures on the master, whereas the bulk of the substrate remains non-deformable and keeps its shape.

[0119]Thus, in a preferred embodiment, the invention is combined with a further step involving continuous embossing, and particularly roll embossing or band embossing, of a substrate.

[0120]According to a preferred embodiment, embossing is performed at a temperature below the glass transition of said first substrate.

Problems solved by technology

However, the presence of urns in such geometries introduces band dispersion (“racetrack effect”), as described in Paegel, et al.

Koutney, et at, (Anal. Chem. 2000, 72, 3388-3391) describe a glass-based DNA sequencing chip with a 40 cm long straight separation channel, which solves this problem of channel length However, this device requires complex lithographic steps (e.g. specialized spin-coating, direct UV-laser writing, wet chemical etching and thermal bonding).

This makes the construction and operation costs of such chips extremely high Also these large microchannel arrays are difficult to manipulate.

The length available in a CD format is still insufficient for high resolutions sequencing, however.

Laboratory-scale replication methods (e.g. mold casting of elastomers or hot-embossing of thermoplastics using a press) are not convenient either for fabricating long straight channels.

In contrast with hot embossing or injection molding, in which the substrate is fully enclosed in a container, and can thus be raised above its glass transition for an arbitrary length of time in order to allow for an accurate reproduction of the microstructure of the mold, continuous microfabrication processes based on lamination raise specific and difficult problems.

This, however, makes the fabrication complex, and restricts the number of materials that can be used.

It also raises problem of adhesion between the substrate and the supporting layer.

However, a thin flexible substrate cannot be raised globally above its glass transition and kept under tension, since it would deform or even break No example of microstructures made by this way were presented in the above patent.

This requires, however, that the molding device be at least partially transparent to said radiation, which is not convenient for industrial processes.

This method, however, is limited to very thin films, thus requiring an additional supporting case, and it is also limited to relatively large microstructures, of order 100 μm.

However, this photocurable resin requires serial processing and relatively long curing and developing, rendering it inadequate for cost-effective mass production.

However, PDMS also suffers from certain disadvantages, such as, swelling in organic solvents (thus limiting the range of microfluidic applications), low mechanical strength (leading to sagging of high-aspect ratio structures in the device) and unstable surface treatments.

Oxidized PDMS becomes hydrophobic in air within 30 minutes, thereby not being able to prevent non-specific adsorption of molecules on the surface of the device.

This method, however, does not lead to a treatment that is very stable in time, and it leads to imperfect surface treatment.

However, due its chemical inertness, this polymer is not easily amenable to surface treatment and bonding.

More generally, bonding materials presenting microstructures or nanostructures without altering these structures remains a challenge.

Another challenge in the fabrication of embedded microstructures such as microchannels, is the closing of microchannels.

A disadvantage of this method, however, is that the lateral walls of the microchannel are of a different chemical nature as the top and bottom walls.

All of the above methods for bonding two substrates in order to create an embedded microchannel, however, share the inconvenient, that they cannot lead to a microchannel with uniform surface properties around its perimeter.

This is very detrimental to numerous applications, in particular those involving the transport in the microchannels, of species that tend to adsorb on the microchannel surfaces.

This is also very disadvantageous to electrophoretic separation methods, or more generally to electokinetic transport, because differences in surface properties lead to inhomogeneous electroosmosis, which in turn lead to dispersion

It is thus very difficult to keep microstructures intact: if the substrate carrying microstructures is the one with the lowest glass transition, the structures will tend to collapse during bonding.

In contrast, if the layer with the lowest glass transition is a planar cover substrate, it will tend to flow into the microstructures, and also lead to an alteration of the wanted microchannel characteristics.

This problem is particularly serious for the fabrication of thin-film systems.

Tho preparation of scaled, thin microfluidic systems is disclosed in US 2005-0089449, but the microstructures prepared were rather large, and the systems prepared this way do not present uniform surface properties all around the microchannel perimeter.

Images