Research Highlights


Optical sorting in microchannels

Single-beam optical traps or optical tweezers are laser-based techniques for manipulating micron-sized objects. Such techniques are based on the principle that small objects can be trapped in the waist of a strongly focused laser beam. The ‘trap’ basically results from the fact that objects in the focus of a laser beam experience a strong restoring force if they try to leave the high intensity volume. In a similar way, optical fields can be used to ‘guide’ or ‘deflect’ particles. Researchers at the University of St Andrews and Illinois Wesleyan University have used these ideas to develop an interesting new technique for separating or ‘sorting’ microscopic particles such as biological cells.

Instead of using a single, focused laser beam to hold or trap a particle, this new kind of optical sorter exploits the interaction of microscopic objects with an extended three-dimensional optical lattice. Selective manipulation of different particles arises for a number of reasons. Firstly the optical lattice is dynamically reconfigurable. Secondly, the degree of interaction with the optical field depends on the optical polarizability of the particle. Consequently, if a fluid containing particles is pumped through the lattice, variation of the flow rate and lattice parameters can be used to selectively deflect and separate specific particle types (the extent of deflection being dependant on the size, shape and optical properties of the particle type).

To demonstrate this idea, the researchers use a five-beam interference pattern to sort co-running particles within microchannel flows. When a laminar flow of mixed particles passes through the optical lattice selected particles are deflected from their original trajectories whilst others remain unaffected and continue in the direction of fluid flow. The lattice is made by passing an infra-red laser beam through a diffractive beam splitter and recombining the output. Chemical discrimination is demonstrated by passing a streaming mixture of 2 μm diameter silica and polymer spheres through an optical lattice. Even at flow velocities in excess of 35 μm s−1, a deflection angle of 45° and a sorting efficiency of at least 96% is achievable. Interestingly, particle throughput in the prototype system was superior to similar separations in microfabricated fluorescence-activated cell sorters. The same approach is also shown to efficiently separate 2 μm protein microcapsules from a polydisperse particle stream.

Since the technique is non-invasive, accurate and reconfigurable it is well-suited for medical applications. Indeed, the authors have already demonstrated the separation of red and white blood cells within laminar flow streams and hope to apply the technique to high-throughput discrimination of cancerous and non-cancerous cells.

Nature, 2003, 426, 421.


Solvent compatibility of PDMS

Soft lithography has become popular over the past five years as a rapid, flexible and low-cost route to the creation of micron-sized features on planar substrates. Elastomeric siloxane polymers such as poly(dimethylsiloxane) (PDMS) are easily molded, optically transparent, durable, cheap, non-toxic and stable over a wide temperature range.

PDMS can be cast against a positive relief template to form microfluidic structures with high aspect ratios by simply pouring a mixture of the elastomer precursor and a curing agent over a template. After (low temperature) curing the structured polymer is peeled away from the template and an enclosed fluidic structure created by contacting the elastomer with a planar surface. Clearly, such fabrication processes are simple to implement and obviate the need for expensive cleanroom facilities. PDMS microstructures have found considerable application in the field of bioanalysis for a number of additional reasons. These include, rapid prototyping, low fabrication costs and good chemical compatibility with most biological fluids.

Nonetheless, microfluidic systems have been shown to be highly flexible and efficient tools in other fields such as small-molecule and nanomaterial synthesis. In these applications a greater diversity of reaction conditions and solvent systems is typical, which presents problems when using siloxane-based materials. For example, a particular problem when using PDMS is that it swells when contacted with non-polar solvents. To further understand how PDMS interacts with different solvent types George Whitesides and co-workers at Harvard University have recently assessed the compatibility of PDMS with a wide range of organic solvents. The authors considered three aspects of ‘compatibility’: the swelling of PDMS in a solvent, the solubility of solutes in PDMS and the dissolution of PDMS oligomers in a solvent. Of these parameters, the swelling of PDMS was found to have the greatest influence of compatibility. Solvents such as water, dimethyl sulfoxide, ethylene glycol, acetonitrile and glycerol were found to cause minimal swelling, whereas solvents such as pentane, triethylamine, xylenes, chloroform and tetrahydrofuran were shown to induce significant swelling of PDMS. Using this information, the authors demonstrate the feasibility of performing organic chemistry in PDMS microstructures, using a Diels–Alder reaction. In addition, highly swelling solvents were shown to be useful in extracting contaminants from bulk PDMS and modifying surface properties.

Due to the low cost and simple fabrication methods associated with PDMS microstructures such reference data should prove highly useful in assessing new applications for this versatile material.

Analytical Chemistry, 2003, 75, 6544.


Microfluidic modelling of malarial infection

Severe malaria by Plasmodium falciparum is responsible for over a million deaths worldwide per year. The severity of this infection is dependent on capillary blockage by infected cells in organs such as the brain, kidney and liver. Healthy red blood cells are highly deformable structures whereas infected erythrocytes lose this deformability and significantly contribute to capillary blockage. To understand and model the behaviour of single infected-erythrocytes in capillary-like environments Daniel Chiu and colleagues at the University of Washington, Seattle have used elastomeric microchannels to mimic capillaries between 2 and 8 μm in diameter. Suspensions of infected red blood cells were motivated through constricted channels at flow speeds imitating in vivo flow rates (∼100–500 μm s−1).

Motivation of 8 μm diameter healthy erythrocytes through these PDMS microchannels demonstrated minimal adherence to channel walls and easy passage through channels as narrow as 2 μm (due to a high degree of flexibility). In addition, cells in an early stage of infection show broadly similar behaviour with a relatively high degree of deformability. However, cells in early trophozoite, late trophozoite and schizont stages of infection show an increasing reduction in deformability. For example, some cells were not able to pass through channel restrictions less than 4 μm and others that could were unable to recover their shape on a short timescale. Other interesting observations include the fact that when a normal erythrocyte approaches a blockage formed by infected cells it does not become a static part of the cellular structure but winds its way through the blockage and out the other side (Fig. 1). This simple observation provides a better understanding of why exchange transfusion is known to temporarily alleviate some of the effects of severe malaria.


Video sequence showing the passage of a normal red blood cell (arrow) through a blockage formed by infected cells at a constriction in a PDMS microchannel: (a) a normal erythrocyte flows smoothly through the main channel, (b) the normal cell makes its way through the blockage towards the 6 μm constriction and (c) the normal cell exits the blockage and moves downstream. (Adapted with permission. Copyright 2003, National Academy of Sciences, USA).
Fig. 1 Video sequence showing the passage of a normal red blood cell (arrow) through a blockage formed by infected cells at a constriction in a PDMS microchannel: (a) a normal erythrocyte flows smoothly through the main channel, (b) the normal cell makes its way through the blockage towards the 6 μm constriction and (c) the normal cell exits the blockage and moves downstream. (Adapted with permission. Copyright 2003, National Academy of Sciences, USA).

This simple approach to studying infected erythrocytes in capillary-like environments has much potential for modelling malarial infection and for screening certain antimalarial drugs. Furthermore, the authors intend to refine their test structures by modification of both surface properties and elasticity of the PDMS microchannels.

Proceedings of National Academy of Sciences, 2003, 100, 14618.


MALDI mass spectrometry on a CD

The large-scale and comprehensive analysis of proteins expressed in biological systems is crucial in functionally assigning genomic sequences. Due to the inherent complexity of protein mixtures, multidimensional separations provide the best route to discrimination of individual proteins. The standard technique used in proteome analysis is two-dimensional gel electrophoresis (2DE). The first dimension involves an isoelectric focusing (IEF) step that discriminates proteins on the basis of their isoelectric point. This is then followed by a second-dimension separation (through an SDS polyacrylamide gel) that yields migration times proportional to protein size. Proteins resolved by 2DE can subsequently be identified through MS analysis. The normal approach to MS analysis involves enzymatic digestion of the protein of interest (trypsin) followed by MALDI-TOF-MS for peptide mass fingerprinting. In addition, ESI-MS may be used to analyse selected peptide fragments to generate peptide sequence information.

The large dynamic range of protein expression ensures that many proteins will exist in low numbers within a given sample. This means that minimal loss of material during sample preparation is an important issue when performing MS analysis. Tomas Bergman and co-workers at Gyros AB and the Karolinska Institutet have addressed this issue by describing an integrated, microfluidic device that performs sample pre-concentration, desalting, elution and crystallization before MALDI analysis. The conceptual operation of a single device is illustrated in Fig. 2. Here a crude sample is introduced at (A) and motivated onto a 10 nL reverse-phase chromatographic column by spinning the microfluidic disk. Peptides are retained and concentrated on the column whilst salts and polar components are washed away. Peptides are then eluted from the column and transported to a 200 × 400 μm MALDI target area (C) where solvent is evaporated and peptides crystallized. The microfluidic CD developed by the Swedish researchers allows 96 samples to be processed in a parallel and automated fashion. The entire CD is then inserted into a MALDI mass spectrometer for peptide mass fingerprinting. Initial results demonstrate sensitivities down to 50 amol, high-throughput sample processing and good-reproducibility. Importantly, integration of several processing operations on a monolithic substrate minimizes surface contacts and thus reduces sample losses. This crucial benefit should prove immensely useful when characterizing highly complex protein mixtures where many proteins exist at reduced concentrations.


Microfluidic structure for on-CD processing of samples of up to 1 μL (96 structures per CD). An individual microstructure element is shown separately for the three consecutive steps of sample application (A), washing/elution (B), and cocrystallization with MALDI matrix (C). The dilute and salt-containing crude sample (grey) is applied at (A) onto the 10 nL reversed-phase column (white). The washing and elution/matrix solutions (black) are applied via a common distribution channel at (B). Cocrystallization of the concentrated and desalted sample with MALDI matrix occurs at (C) in a 200 × 400 μm target area accessible to the laser beam of the MALDI instrument as shown by a photograph of the crystals in the desorption area. (Adapted with permission. Copyright 2003, The American Chemical Society).
Fig. 2 Microfluidic structure for on-CD processing of samples of up to 1 μL (96 structures per CD). An individual microstructure element is shown separately for the three consecutive steps of sample application (A), washing/elution (B), and cocrystallization with MALDI matrix (C). The dilute and salt-containing crude sample (grey) is applied at (A) onto the 10 nL reversed-phase column (white). The washing and elution/matrix solutions (black) are applied via a common distribution channel at (B). Cocrystallization of the concentrated and desalted sample with MALDI matrix occurs at (C) in a 200 × 400 μm target area accessible to the laser beam of the MALDI instrument as shown by a photograph of the crystals in the desorption area. (Adapted with permission. Copyright 2003, The American Chemical Society).

Analytical Chemistry, 2004, 76, ASAP Article.

Andrew J. de Mello


This journal is © The Royal Society of Chemistry 2004
Click here to see how this site uses Cookies. View our privacy policy here.