2.1

Stimuli-responsive hydrogels

Hydrogels are a class of cross-linked polymers that can hold large volumes of water and have been investigated extensively in biomedical and microfluidic applications due to their degree of controllable swelling/shrinking. Stimuli-responsive swelling and de-swelling can result from changes in temperature, pH level, electromagnetic fields (to induce temperature changes), or ionic strength. Stimuli-responsive hydrogels make excellent actuators and have often been utilized to construct microfluidic valves or pumps in microfluidic systems.

The hydrogel PNIPAAm is a thermo-sensitive polymer that exhibits a reversible phase transition from a swollen hydrated state to a dehydrated state in the presence of a temperature change⁴. It is also pH-responsive⁵, and is the most well-researched hydrogel for microfluidics and biomedical applications. Figure 2⁶ illustrates a PNIPAAm hydrogel sample that shows the transition from transparent to opaque during a change from room temperature to its phase transition temperature. This change in volume with temperature is substantial, and can be used for a wide variety of actuation mechanisms. PNIPAAm has also been employed on polymer surfaces (polystyrene) for cell capture using hydrophobicity control, and for coating of magnetic and gold nanoparticles, or on substrate surfaces, to aggregate particles or hold biomolecules. Figure 3 shows the chemical structure of the monomers and cross-linker used for synthesis of this particular hydrogel⁶.

Figure 2.

PNIPAAm hydrogel in: a) cooled, swollen state; b) heated, un-swollen state⁶.

Figure 3.

PNIPAAM hydrogel monomer and cross-linker⁶.

Examples of other hydrogels of interest to biomedical MEMS and microfluidics are those based on poly(acrylic acid) sodium salt (PAAS)⁷, poly(HEMA-co-DMAEMA) hydrogel⁸, and the electroactive hydrogel 4-hydroxybutyl acrylate (4-HBA)⁹. Much research continues in both the microfluidics and MEMS communities using an ever-increasing list of stimuli-responsive hydrogels, and includes many investigations in hydrogel patterning including micromolding¹⁰ and other polymer-compatible methods.

2.2

Composite polymer materials

Polymer materials are typically insulating, non-conducting, and have low heat transduction. In microfluidics and biomedical instrumentation, they typically serve as the passive microfluidic channels and chambers, and/or the insulating substrate material. Many different polymers are used for microfluidics and biomedical microdevices and systems, and include a wide variety of thermosetting and thermoplastic polymers². A thermosetting polymer is one that irreversibly cures, and covers materials such as epoxy-based materials, including SU-8 photopatternable epoxy; as well as polyimide, and polydimethylsiloxane (PDMS) and other thermosetting silicones, that are used widely in MEMS and microfluidics. Thermoplastic polymers turn to liquid when heated and freeze to a glassy state when cooled. Thermoplastic polymers are widely used in commercial microfluidics, and include materials such as polystyrene, polycarbonate, polymethlymethacrylate (PMMA), cyclic olefin polymer (COP) and polyimide. Polyurethane is a widely known thermoplastic elastomer.

One way in which polymer materials can be rendered functional is through embedding nanoparticles or microparticles into a polymer which forms the base, or matrix, of the functional composite polymer (CP) material. For example, conductive nanoparticles such as silver spheres or flakes may be employed to create conductive nanocomposite polymers (C-NCPs), NdFeB-based microparticles may be employed to create magnetic composite polymers (M-CP), and carbon nanotubes (CNTs) may be employed for C-NCPs that have a strong sensitivity to pressure or polymer swelling due to analyte in a microfluidic sample. Examples of nanoparticles are metal or carbon nano-sized flakes, spheres, cylinders, needles or tubes with at least one dimension typically less than 0.1 μm in average size. Microparticles may be, e.g., metal or alloy flakes or spheres typically 1 to 10 μm in average size. Metal-based particles include silver, gold, aluminum, tungsten, copper, platinum, iron and its oxides, nickel, and rare earth particles based on SmCo or NdFeB, Particle shapes include flakes, rods, spheres, and random assortments that may contain all of these shapes¹¹. Early work on carbon-based nanoparticles involved the utilization of carbon black to make conductive polymers for employment as electrodes or sensors, or for improved polymer mechanical properties¹². CNTs, which are now regularly used for C-NCPs, usually fall into two categories: single walled CNTs (SWCNTs) and multi-walled CNTs (MWNCTs). They may also be semiconducting, metallic, or most typically will contain a mixture of both types in a given sample¹³.

The properties of a CP strongly depend on the volume fraction of particles in the polymer matrix, known as the fill factor. Weight percent (wt-%) is equal to the weight of solute divided by the weight of solution multiplied by 100%, and is the most typical method in which fill factor is specified (volume percent, or expression in fractional amounts, are also used). The filler density can have a large effect on mechanical and other properties of the composite polymer, e.g., the material will typically become stiffer (higher modulus of elasticity) with increasing fill factor, which is an important consideration if employing the CP as an actuator or actuator electrode that must undergo large deformations. The wt-% of nanoparticles in a C-NCP is very important to its electrical behavior due the existence of a percolation threshold, which is the point at which the first continuous chain of linked conductive filler particles is present, and can be explained using percolation theory^14,15. The percolation threshold is a function of both polymer and nanoparticle properties, with aspect ratio of the particles (ratio of length to diameter of nanoparticle) being a very important consideration, with high aspect ratio nanoparticles typically resulting in lower thresholds. Despite the importance of aspect ratio, most percolation theory was historically derived using spherical particle models. Figure 4 shows an example of a typical reversed “S-shaped” curve for resistivity versus wt-% of nanoparticle filler for a C-NCP (COOH-functionalized MWCNT/PDMS C-NCP)¹⁶. It is observed that as concentration of nanoparticles increases, the percolation paths via the conductive particles are set up with the result that the resistivity decreases sharply near approximately 1.5 wt-% (the percolation threshold, which can be defined as the maximum change of slope of the curve). The small size and large aspect ratio of many nanoparticles, such as CNTs and silver rods, are generally helpful in lowering the percolation threshold. Percolation threshold generally increases as particle size decreases¹⁷. However, the degree of convolution and entanglement in very high aspect ratio fibers significantly complicates predictions: CNTs may have aspect ratios from 100-10000, and very low percolation thresholds (e.g., less than 5 wt-%) as a result¹⁸.

Figure 4.

A typical (reversed) “S-shaped” curve for resistivity versus wt-% of nanoparticle filler in C-NCP (adapted from¹⁶).

Wt-% is also important to the properties of M-CPs, with higher wt-% typically resulting in an increase in magnetic force important for, e.g., actuator applications. Many materials exhibit ferromagnetic behavior and possess high magnetic susceptibility, which indicates a strong attraction to magnetic fields. Soft magnetic materials are those such as Fe, Ni, and Co, which can be readily magnetized but lose their magnetization easily. Hard magnetic materials can produce stronger magnets than soft magnetic materials, and include materials such as hard ferrites (e.g., strontium ferrite), and materials based on rare earth elements, such as neodymium (Nd) or samarium (Sm). Nanoparticles consisting of soft magnetic materials are readily available, while hard magnetic materials are typically microparticles¹⁹. However, a lower wt-% of hard magnetic particles typically results in higher magnetic attraction to magnetic fields. Hard magnetic materials may also be permanently magnetized in high (e.g., 2T) fields, which is an important consideration if bi-directional actuation is desired using a single, reversible-field electromagnet.

Micropatterning of CPs typically follows whatever microfabrication method has been established for the particular base polymer in the microfluidics and MEMS fields. CP microfabrication methods include photopatterning of SU-8 epoxy, PDMS, and some thermoplastic polymers; soft-lithography of PDMS; and other micromolding/casting methods, often against a sacrificial mold, of a wide variety of both thermosetting and thermoplastic polymers. Other microfabrication methods under investigation include ink-jet printing, screen-printing, microcontact printing, nano-imprinting, embossing, and post-fabrication filling with nanoparticles^11,19.

2.3

Other smart and functional polymer materials

Many other smart and functional polymer materials have been investigated for biomedical MEMS and microfluidics applications. A short list of such polymers includes the following:

Many other EAPs and other smart and functional polymers exist, e.g., various ionic gels, but have not been fully exploited for use in microfluidics and bioMEMS, although they may be under intense investigation for other applications in, e.g., non-biomedical MEMS and robotics.

3.1

Flow control

In any microfluidic system, flow control is required in order to move a sample between operations and perform fluidic manipulations on the sample such as mixing it with other reagents, which can be difficult in microfluidic systems as fluid flow is laminar. Microfluidic flow control typically employs microfluidic devices such as valves, pumps, and other devices that alter channel or chamber geometry; mixers; filters; hydrodynamic steering of flow streams; and interconnect between microfluidic channel-containing devices.

Stimuli-responsive hydrogels have been used extensive for flow control in microfluidic systems, especially as the basis for microfluidic valve or pump actuation schemes^30,31. One of two different mechanisms is usually employed for valve actuation: use of the hydrogel to actuate a flexible polymer membrane, or diaphragm; and use of the hydrogel as a plug that is in contact with the sample fluid and can shrink or swell in response to heat or pH level. Membrane-type actuators are also applicable to microfluidic pumps, as the most common type of microfluidic pump involves actuating a membrane with input and output check valves (or diffuser/nozzle input and output structures) so that actuation of the pump results in flow primarily in one direction. Figure 5^6,10 shows a plug-type thermally-responsive PNIPAAm hydrogel valve that is different from other hydrogel-based valves in that all of its components, including its heater element, are mechanically flexible. Microvalve technology for mechanically flexible polymer microsystems remains a largely unexplored area; however, not only are mechanically flexible polymer microvalves generally applicable to polymer microsystems, they further enable microfluidics to be applied to flexible wearable and implantable microsystems, or those located in flexible packaging such as textiles³ or contact lenses³². While the copper metal on a flexible printed circuit board can be employed for the flexible heater element¹⁰, an even better solution, from the standpoint of materials integration and fabrication simplicity, is to employ flexible C-NCP heater elements doped with metallic or carbon-based nanoparticles, such as tungsten⁶. PDMS can be employed as the base polymer for all other components of the valve, including microfluidic channels, heater element, and electronic routing to the heater, or between heaters when an array of valves is used.

Figure 5.

Mechanically-flexible plug-type PNIPAAm hydrogel valve with W-doped PDMS C-NCP, or flexible PCB metal, flexible heater element: a) schematic of valve showing subcomponents of device; b,c) hydrogel plug microfabrication via micromolding; d) normally-closed valve operation with heater off; e) open valve with heater on showing fluid flow from left to right (adapted from¹⁰).

Many different M-CP actuators have been proposed in recent years for microfluidic valves and pumps, with most typically falling into one of four categories based on actuation structure: membrane, beam, cilia, or rotary¹⁹. Many membrane-type actuators intended for use in microfluidic valves and pumps have been proposed. In one example, membranes ranging in size from 1 to 8 mm in diameter were fabricated from PDMS with 75 wt-% 5 μm (Nd0.7Ce0.3)10.5Fe83.9B5.6²⁰ rare-earth magnetic particle filler³³. Deflections up to 1 mm under a 70 mT oscillating electromagnetic field were demonstrated, as well as deflections of 100 μm under fields as small as 6.2 mT for larger membranes. This work was followed up with a valve designed to employ the actuator, resulting most recently in demonstration of pairs of fully operation microfluidic valves with fast and efficient flow switching under magnetic fields up to 80 mT³⁴. Biomimetic cilia consist of small cantilevers or “hairs” that can be actuated in a way that mimics the cilia that line the inner surfaces of the trachea of mammals to expel mucus from the lungs. M-CPs make excellent cilia actuators as high-aspect ratio structures can be formed in highly flexible polymers and easily controlled by global or local magnetic fields. M-CP cilia have been demonstrated for microfluidic mixing, fluid manipulation, and for manipulation of small objects. Figure 6³⁵ shows an example of microfabricated cilia structures for microfluidic mixing that was fabricated using the same materials as the membrane-type actuators mentioned previously. Due to the high magnetization level of the composite polymer, miniature electromagnets were employed to provide relatively small magnetic fields of 7mT to actuate the cilia microstructures over a very wide motion range. Iron-doped magnetic nanocomposite polymers have also been employed as microfluidic interconnect structures³⁶. A recent review¹⁹ provides a more thorough discussion of employment of M-CPs in microfluidics.

Figure 6.

Magnetic composite polymer cilia: a) six cilia at rest; b) four cilia of different heights under a 7mT oscillating field³⁵.

Many other smart materials have been employed for microfluidic flow control, including EAPs for microfluidic valves, pumps, and control of microfluidic channel geometry. The most typical EAPs employed are DEAs^21,22 and PVDF²³ for membrane-type actuators. ER and GER²⁴, and stimuli-responsive shape memory polymers²⁰ have also been employed to realize valve operations and microfluidic channel re-organization toward smart microfluidic systems. Furthermore, C-NCPs have been employed as GER actuator electrodes²⁵, as an example of a functional CP combined with a stimuli-responsive polymer material.

3.2

Surface control

PNIPAAm hydrogel has been extensively used for modification of surfaces of substrates and nanoparticles for assays and diagnostics, as well as control of hydrophobicity for biological cell culture, and controlled absorption and release of biomolecules^27,37. For example, when PNIPAAm is coated on magnetic nanoparticles and gold nanoparticles, it can act to aggregate the different nanoparticles together when the temperature is raised²⁷. Temperature-responsive and photosensitive wettability has also been demonstrated for microfluidic channel surfaces, for defining areas for biological cell immobilization^27,29.

3.3

Sensors and electrodes

C-NCPs have been employed for microfluidic applications such as sensor electrodes, direct analyte detection, and as flow sensors. Figure 7³⁸ shows C-NCP for use as electronic routing and electrodes using a silicone base material (PDMS) doped with Ag. Such sensor electrodes can be chloridized to result in Ag/AgCl electrodes which are useful for microfluidic and biomedical MEMS applications. C-NCP electrodes have also been employed as flow sensors, where the electrodes are operated as heaters and the heat conduction away from the electrodes due to passing fluid flow can be monitored³⁹.

Figure 7.

C-NCP electrodes fabricated via micromolding from PDMS filled with Ag nanoparticles³⁸. Substrate has been twisted to show flexibility.

When a C-NCP matrix containing CNTs is exposed to an analyte, the reaction between the molecules and the C-NCP matrix material changes the physical properties and dimensions of the polymer, causing it to swell; this physical distance change between CNTs has been suggested as the mechanism of resistance increase. Such sensors have been demonstrated for detection of biohazardous solvent vapor, using functionalized CNTs embedded in a PMMA matrix⁴⁰. Other example sensors detect various sorts of biohazardous gases and vapors using primarily carbon black or CNTs in a wide variety of different thermoplastic and thermosetting polymer base materials.

Other stimuli-responsive materials such as hydrogels and halochromic polymer materials have also been employed for pH level sensors and for biosensors⁴¹ where the stimulus is provided by the quantity that is to be measured, and is an active and intense area of research.