Hayden Taylor

A practical guide for the fabrication of microfluidic devices using glass and silicon

This paper describes the main protocols that are used for fabricating microfluidic devices from glass and silicon. Methods for micropatterning glass and silicon are surveyed, and their limitations are discussed. Bonding methods that can be used for joining these materials are summarized and key process parameters are indicated. The paper also outlines techniques for forming electrical connections between microfluidic devices and external circuits. A framework is proposed for the synthesis of a complete glass/silicon device fabrication flow.

Metallic glasses: viable tool materials for the production of surface microstructures in amorphous polymers by micro-hot-embossing

Metallic glasses possess unique mechanical properties which make them attractive materials for fabricating components for a variety of applications. For example, the commercial Zr-based metallic glasses possess high tensile strengths (?2.0 GPa), good fracture toughnesses (?10?50 MPa) and good wear and corrosion resistances. A particularly important characteristic of metallic glasses is their intrinsic homogeneity to the nanoscale because of the absence of grain boundaries. This characteristic, coupled with their unique mechanical properties, makes them ideal materials for fabricating micron-scale components, or high-aspect-ratio micro-patterned surfaces, which may in turn be used as dies for the hot-embossing of polymeric microfluidic devices. In this paper we consider a commercially available Zr-based metallic glass which has a glass transition temperature of Tg ? 350??C and describe the thermoplastic forming of a tool made from this material, which has the (negative) microchannel pattern for a simple microfluidic device. This tool was successfully used to produce the microchannel pattern by micro-hot-embossing of the amorphous polymers poly(methyl methacrylate) (Tg ? 115??C) and Zeonex-690R (Tg ? 136??C) above their glass transition temperatures. The metallic glass tool was found to be very robust, and it was used to produce hundreds of high-fidelity micron-scale embossed patterns without degradation or failure.

A computationally simple method for simulating the micro-embossing of thermoplastic layers

We demonstrate a highly computationally efficient approach to simulating the embossing of micrometer-scale, feature-rich patterns on to thermoplastic polymeric layers. The method employs a linear viscoelastic model for the embossed layer and computes the distribution of contact pressure between the polymeric layer and a rigid embossing stamp that is consistent with the progression of the polymer deformation. An approximation to the embossed topography of the polymeric layer is thereby generated as a function of the material being embossed, the stamps design, and the embossing processs temperature, duration and applied load. For a stamp design described with an 800 × 800 matrix of topographical heights, simulation can be completed within 30–100 s using a personal computer with an Intel Pentium 4 processor and 2 GB RAM. Our method is sufficiently fast for it to be employed iteratively for designing a pattern to be embossed or for selecting processing parameters. The viscoelastic properties of three polymeric materials—polymethylmethacrylate, polycarbonate and Zeonor 1060R, a cyclic olefin polymer—have been experimentally calibrated. For a test pattern having features with diameters of 5 µm to 90 µm, simulated and experimental topographies agree with rms errors of less than 2 µm across all processing conditions tested, with absolute topographical heights ranging up to 30 µm.

An investigation of the detrimental impact of trapped air in thermoplastic micro-embossing

In hot micro-embossing, a thermally softened polymer is forced into the cavities of a patterned stamp. If embossing is performed in normal atmospheric surroundings, it is possible for air to become trapped inside the cavities of the stamp, impeding pattern replication. We present a fast simulation technique that captures the impact of trapped air in the micro-embossing process. The technique can predict the extent of stamp cavity filling for any given stamp design and linear viscoelastic polymer properties, under the assumption that air, once trapped inside a cavity, is unable to escape. We find that the smaller the effective stiffness of the softened polymer relative to the surrounding atmospheric pressure, the more severe the impact of trapped air. Moreover, the trapping of air becomes more detrimental as the widths of stamp cavities become a larger proportion of their lateral spacing. The results of embossing experiments with two widely used thermoplastic polymers, Zeonor 1060R and polymethylmethacrylate, agree with simulation results and indicate that there is little permeation of trapped air in these materials during a typical embossing process of up to 15 min duration, carried out at between 5 and 45 °C above their glass-transition temperatures.

Bioinspired fibrillar adhesives: a review of analytical models and experimental evidence for adhesion enhancement by surface patterns

Fibrillar structures are found on the attachment pads of insects and small reptiles. These structures enable exquisite conformation to rough surfaces, increase the number of van der Waals interactions between the structure and the target surface, and thus enhance adhesion. Biomimetic adhesives replicate this effect by patterning polymer films with micron- or sub-micron-sized protrusions. Numerical contact-mechanics models as well as experimental adhesion measurements have been reported for a variety of protrusion shapes including flat, rounded, mushroom and spatula geometries. Although superior adhesion has been reported for the mushroom and spatula tip geometries, straight, flat-tipped pillars offer the potential for much simpler mass production such as by injection moulding and are thus the focus of this review. Existing models for straight, flat-tipped pillar arrays do not fully agree with reported experimental results. Analytical models are generally limited to elastic materials, and inherently assume that neighbouring pillars behave independently. For elastic pillars in close proximity, however, pillars do in fact interact mechanically, affecting adhesion. Moreover, visco- and hyper-elastic materials are often used in practice, yet dissipative effects receive little attention in analytical models. We find that no study has conclusively investigated the limit of adhesive strength achievable by fibrillar adhesives. It also remains unclear what happens to the adhesive strength as the areal density of contacting regions approaches 100%.

2D and 3D growth of carbon nanotubes on substrates, from nanometre to millimetre scales

This paper presents a suite of techniques for making Carbon Nanotube (CNT) assemblies having hierarchical two- and three-dimensional organisation: conformal films of tangled single-walled CNTs are grown on silicon microstructures, millimetre-thick films and microstructures of aligned Multi-wall CNTs (MWNTs) are grown on flat silicon substrates, large-area CNT micropatterns are fabricated by post-growth dry embossing of aligned CNT films and 3D forms of CNTs are directly grown to take the shape of microfabricated template. These growth processes use catalyst thin-films deposited by electron beam evaporation and atmospheric-pressure thermal Chemical Vapour Deposition (CVD), which is scalable from academic-level prototyping to industrial-level manufacturing.

Controlled Folding of Graphene: GraFold Printing

We have used elastomeric stamps with periodically varying adhesive properties to introduce structure and print folded graphene films. The structure of the induced folds is investigated with scanning probe techniques, high-resolution electron-microscopy, and tip-enhanced Raman spectroscopy. Furthermore, a finite element model is developed to show the fold formation process. Terahertz spectroscopy reveals induced anisotropy of carrier mobility along, and perpendicular to, the graphene folds. Graphene fold printing is a new technique which allows for significant modification of the properties of 2D materials without damaging or chemically modifying them.

A method for the accelerated simulation of micro-embossed topographies in thermoplastic polymers

Users of hot micro-embossing often wish to simulate numerically the topographies produced by the process. We have previously demonstrated a fast simulation technique that encapsulates the embossed layers viscoelastic properties using the response of its surface topography to a mechanical impulse applied at a single location. The simulated topography is the convolution of this impulse response with an iteratively found stamp–polymer contact-pressure distribution. Here, we show how the simulation speed can be radically increased by abstracting feature-rich embossing-stamp designs. The stamp is divided into a grid of regions, each characterized by feature shape, pitch and areal density. The simulation finds a contact-pressure distribution at the resolution of the grid, from which the completeness of pattern replication is predicted. For a 25 mm square device design containing microfluidic features down to 5 µm diameter, simulation can be completed within 10 s, as opposed to the 104 s expected if each stamp feature were represented individually. We verify the accuracy of our simulation procedure by comparison with embossing experiments. We also describe a way of abstracting designs at multiple levels of spatial resolution, further accelerating the simulation of patterns whose detail is contained in a small proportion of their area.

A two-level prediction model for deep reactive ion etch (DRIE)

The authors contribute a quantitative and systematic model to capture etch non-uniformity in deep reactive ion etch of MEMS devices. Non-uniformity depends on uneven distributions of ion and neutral species at the wafer level, and local consumption at the die level. An ion neutral synergism model is constructed from data obtained from several layouts of differing layout pattern densities, and is used to predict wafer level variation with an r.m.s. error below 3%. This model is combined with the die level model, which the authors have reported previously based on T. Hill et al., (2004), on a MEMS layout. The two level model is shown to enable prediction of both within die and wafer scale etch rate variation for arbitrary wafer loadings.

Effect of polymer orientation on pattern replication in a micro-hot embossing process: experiments and numerical simulation

The hot embossing process has been identified as a promising technique for fabricating micro- and nanostructures for polymer-based biological and chemical MEMS (micro electro mechanical systems). However, there has not been any investigation of the effect of polymer chain orientation in the base polymer substrate on replication during the micro-embossing process. Such effects could prove important because polymer chain orientation may develop in the polymer substrates during their production. In this investigation, it was observed that the degree and ease of microchannel replication are significantly influenced by the molecular chain orientation in injection-molded polymer substrates. Microchannels aligned along the flow direction of the polymer replicate easily compared to microchannels aligned across the flow direction of the polymer. The replication fidelity during hot embossing was investigated using a white-light confocal microscope. The anisotropy of injection-molded polymer plays a dominant role in the replication fidelity of microchannels, and the ability to model the anisotropic behavior of the material will enable understanding and prediction of the hot embossing process. Therefore, a material model that reflects the directionality was utilized to simulate the experimental embossing results obtained both along and across the flow direction of the polymer. By comparing experimental results with simulations, we observed that the model is reasonably realistic.