CHAPTER 1: Biomimicry and Materials in Medicine

Past generations have attempted to tame, harness and conquer nature, and bring her to her knees. Such efforts have transformed over the years into a wiser, enthusiastic pursuit to unravel her secrets and copy the “engineering” afforded by nature. Implicit in this is the recognition that literally billions of years of evolution have resulted in the refining of biological structures at the macro- and nano-scales.

A living cell that grows and multiplies looks much like a robot that has learned to build itself. We are still not sure exactly how this machine was first assembled and programmed. In comparison, strident announcements with respect to scientists “creating life” in the lab are only lame self-promoting marketing strategies. The reality is that there is a failure to acknowledge the fact that the inception of such efforts were initiated with existing living cells before any major manipulations were made inside the cells. There is no clear definition of life; we even struggle to ascertain the difference between a living cell and a dead one. All the chemicals and physical properties are essentially identical, with the most remarkable feature being the ineffable force that controls those 100 000 chemical reactions in each second, in each living cell. For that matter, we are required to recognize that we do not even have a real definition of reality, only hypotheses which contradict the Copenhagen interpretation because of the need of an observer, explicitly the multiverse theory,¹  Bohm’s implicate order,²  Everett’s parallel universes,³  and orchestrated quantum coherence.⁴  Troublesome as it is, we do not have a definition of time either, since the vibrations of the cesium atoms and quartz crystals only measure something that does not exist when we dream, for example; something that expands exasperatedly when we are bored and contracts when we are in the flow.

Leaving such big questions under the rug of modern science, where they belong, we turn to the somewhat more uncomplicated quest of mimicking nature by creating synthetic materials capable of both detecting biochemical and biophysical changes at the molecular level that reflect sensory mechanisms. Additionally, synthetic materials can replace natural biological components for medical corrective procedures and/or healing. Such activities are often characterized as being incorporated in the relatively new field of “biomimicry”, a word coined by Schmitt⁵  some years ago.

New materials which have features that can be modified under specific conditions are pushing forward the frontiers of scientific and technological capabilities. Synthetic materials designed to change their size or shape when exposed to heat, change from a liquid to a solid in magnetic fields, or dramatically change their volume, viscosity, conductivity, work function, etc., will allow new and interesting applications to emerge, not only for sensing and detection purposes, but also for everyday use. Smart materials with piezoelectric, magneto-rheostatic, electro-rheostatic, and memory hysteresis are already used in cars, coffee pots, glasses, or in space missions. Relevant to this chapter, intelligent biomaterials that respond to biological signals show great promise in regenerative medicine, diagnostics, and drug delivery.

Common examples include piezoelectric materials which produce a charge under mechanical stress and, conversely, contact or expand when a voltage is applied.⁶  In shape-memory alloys and polymers, large deformations can be induced and recovered by temperature changes or stress changes. This pseudoelasticity is the result of martensitic phase changes.⁷  In magnetostrictive materials, changes in shape can be instigated by magnetic fields; on the other hand, these materials also exhibit changes in magnetization under mechanical stress.⁸  Magnetic-shape memory alloys change their shape when under a magnetic field, and pH-sensitive polymers change in volume when exposed to alterations in the pH of the solution in which they are immersed. Halochromic materials change their color as a result of change in acidity; they are very suitable detection materials for applications such as corrosion detection.⁹  Chromogenic materials change color as a response to electrical, thermal, or optical changes.¹⁰  When an electric voltage is applied, electrochromic materials change their color or opacity, as in liquid crystal displays.¹⁰  Thermochromic materials change in color depending on temperature changes and photochromic materials respond to light changes (used in light-sensitive sunglasses which darken when exposed to bright sunlight).¹⁰  Photomechanical materials change shape when exposed to light. Dielectric elastomers produce large strains under external electrical fields.¹¹  Magnetocaloric materials change reversibly in temperature under a changing magnetic field.¹²  Thermoelectric materials convert temperature differences into electricity.¹³  Such active materials are ideal for sensor devices and for such applications as resorbable bioceramics, adaptive bioglasses, biomimetic polymers and gels, active nanoparticles, smart textiles, and active optical fibers. Manipulation of material properties at the atomic and molecular scale is leading to self-assembling materials, nanolithography, DNA-based technologies (such as DNA computing), nano- and micro-engineered devices for diagnostics, pharmaceuticals, therapies, drug delivery systems, biocompatible implants and prostheses, and bio-functional systems.

We now turn to the employment of new materials in a plethora of applications in biology and medicine.

A shape memory transformation was first observed in 1932 in an alloy of gold and cadmium, and then later in brass in 1938.¹⁴  The shape memory effect was seen in the gold–cadmium alloy in 1951, but this was of little use. Some 10 years later, in 1962, an equiatomic alloy of titanium and nickel was found to exhibit a significant shape memory effect and Nitinol (so named because it is made from nickel and titanium and its properties were discovered at the Naval Ordinance Laboratories) has become the most common shape memory alloy (SMA).¹⁴  Other SMAs include those based on copper (in particular CuZnAl), NiAl, and FeMnSi, although it should be noted that the NiW alloy has by far the most superior properties.¹⁴ 

Metal alloys of nickel and titanium have the remarkable properties of remembering the shape by undergoing a phase change in which atoms are shifting their position in response to a specific stimulus, such as temperature or stress. The phase change temperature can be tuned by varying the ratio of nickel to titanium. The resulting structure has a highly symmetrical cube structure called austenite at temperatures above the phase change temperature and a much less symmetrical structure at temperatures below the phase change temperature, called the martensite state. In the latter state the material is very elastic, but the austenite state is rigid. The modification of the crystal lattice during the transformation can be carefully controlled. The shape change may exhibit itself as either an expansion or contraction. The transformation temperature can be tuned to within a couple of degrees by changing the alloy composition. Nitinol can be made with a transformation temperature anywhere between −100 °C and +100 °C, which makes it very versatile.

The memory mechanism is based on the thermal energy acquired by the sample through heating, providing the energy necessary for the atoms to return to their original positions so the sample regains its original shape. Springs made of such alloys return to their original shape in warm water or in streams of hot air. Nitinol and superelastic materials made of shape memory alloys of gold–cadmium, copper–aluminum–nickel, copper–zinc–aluminum, and iron–manganese–silicon are used today in medical applications, aerospace, and the leisure industry, namely in vascular stents, for anchors attaching tendons to bones, medical guidewires, medical guidepins, root canal fillings, bendable surgical tools, cardiac catheters, orthodontic wires, flexible eyeglass frames, etc.¹⁵ 

Still many years away is the use of SMAs as artificial muscles, i.e. simulating the expansion and contraction of human muscles. This process will utilize a piece of SMA wire in place of a muscle on the finger of a robotic hand. When it is heated, by passing an electrical current through it, the material expands and extends the joint; on cooling, the wire contracts again, flexing the finger. In reality this is incredibly difficult to achieve since complex software and surrounding systems are also required. NASA has been researching the use of SMA muscles in robots which walk, fly, and swim.

SMA tubes can be used as couplings for connecting two tubes. The coupling diameter is made slightly smaller than the tubes it is to join. The coupling is deformed such that it slips over the tube ends and the temperature changed to activate the memory. The coupling tube shrinks to hold the two ends together but can never fully transform so it exerts a constant force on the joined tubes.

In addition to the shape memory effect, SMAs are also known to be very flexible or superelastic, which arises from the structure of the martensite. This property of SMAs has also been exploited in mobile phone aerials, spectacle frames, and underwires in bras. The kink resistance of the wires makes them useful in percutaneous angioplasty, a surgical procedure requiring catheters which need to remain straight as they are passed through the body. Nitinol can be bent significantly further than stainless steel without suffering permanent deformation.¹⁶ 

Ceramics are more chemically stable and inert than metals. They can be classified into three distinct categories:

1. Oxides such alumina, beryllia, zirconia, and ceria

2. Non-oxides such as carbide, boride, nitride, and silicide

3. Composite materials such as particular reinforced fibers or combinations of oxides and non-oxides

Ceramics are used in numerous technical applications. Since ceramics are so much more biocompatible when used inside the human body, they play a major role as synthetic biomaterials that can be part of the systems which treat or replace a living tissue or lost function.¹⁷  Depending on the particular application, the major requirements beside biocompatibility are resistance to abrasion and wear, fatigue, strength, durability, and resistance to corrosion, especially when they are used as an implant material. The most common materials are alumina, zirconia, bioglass, hydroxyapatite, and tricalcium phosphate. Not only they are inert, but they are also resorbable. They can dissolve and integrate actively in physiological processes such as bone healing. Hydroxyapatite can be actually found in the human body, in teeth and bones. The synthetic material is commonly used as a filler to replace amputated bone or as a coating to promote bone ingrowth into prosthetic implants. Natural coral skeletons can be transformed into hydroxyapatite at high temperatures. Their porous structure encourages the rapid ingrowths of bones. The high-temperature treatment destroys any organic molecules such as proteins, thus preventing immune rejection.

Zirconia is used on the artificial femoral heads employed in hip replacements.¹⁷  It gives strength to the structure, so its dimensions can be smaller and less traumatic for the patient. It is also used in shoulder, knee, spinal implants, and phalangeal joints.¹⁷  In dentistry, it is used with increased frequency for crowns, bridges, and implant abutments. Crystal zirconia is a modern dental ceramic replacement for the metal substructures used under porcelain crowns and bridges. It is translucent, thus giving the overlaid porcelain a brighter and more natural look. It is biocompatible and, unlike amalgams and metal alloys, does not generate adverse reactions or allergies. It is virtually unbreakable, so the dental work can last for a lifetime.¹⁸ 

Finally, a particularly interesting application of ceramic materials is their use in the treatment of cancer, through hyperthermia and radiotherapy.¹⁹  In the quest to avoid the devastating effects of chemotherapy, glass microspheres are inserted into the tumor using a catheter and the radiation is focused on the tumor, similar to brachytherapy for prostate cancer. This causes minimal damage to the surrounding tissue. It is a simple treatment that can be performed on an outpatient basis.

Smart polymers are used for industrial purposes, in medicine, sports, and agriculture due to their inert or bioactive properties. Biodegradability is also a great advantage of such polymers. High-performance polyethylene is used in medicine for total or partial joint replacement of hip, knee, or intervertebral implants due to its high impact strength given by extremely long chains.²⁰  It is highly resistant to corrosion, has very low moisture absorption and a very low friction coefficient. It is self-lubricating and highly resistant to abrasion, more resistant than carbon steel.

Hydrogels, networks of hydrophilic polymer chains in colloidal gels, with water as the dispersion medium, are highly absorbent materials; they possess a high degree of flexibility, very similar to natural tissue, due to their significant water content.²¹  Smart gels contain fluids, usually water in a matrix of large, complex polymers. These polymers are unique in that they respond to stimuli in an advanced way. Types of stimuli that affect smart gels are physical and chemical factors. Temperature, light, electric forces, magnetic forces, and mechanical forces are types of physical interactions on the gel that will precipitate a reaction. Chemical stimuli are usually pH changes or solvent exchanges. The reaction of the smart gel is always an expansion or contraction within milliseconds upon stimulation. When a gel swells, it absorbs additional fluid into it. When it deflates, it expels this fluid out of its membrane. The expansion and contraction are usually caused by a change in the polymer; the stimulus alters the polymer by making it more or less hydrophilic. For example, a significant pH decrease will neutralize ions in the gel, precipitating the polymers to be less hydrophilic and causing the gel to contract.

The effects of such synthetic gels are greatly aided by using nanoparticles.²²  While microparticles usually allow the gel to function properly, smaller particles at the nanoscale increase intended effects dramatically. A great example of this is in the use of ferromagnetic particles. Ferromagnets are tiny particles that act as little bar magnets; applying a magnetic field on a smart gel encourages the ferromagnets to move.²²  This movement raises the temperature of the gel and consequently causes the gel to expand. While microparticles of iron still allow the gel to expand, nanoparticles make the gel more responsive to the magnetic field.

Gels are used as scaffolds in tissue engineering to support living cells for tissue repair, as coatings of wells for cell cultures. Smart hydrogels use their environmental sensitivity to detect changes in pH, temperature, or concentration.²³  They are also used in drug delivery systems, as biosensors, contact lenses (silicone hydrogels), EEG and ECG electrodes, and for dressings used in the healing of burns or hard-to-heal wounds.²⁴ 

Applications of smart gels permeate into various fields, including both medical and industrial. The two main applications for smart gels are in artificial muscle fabrication and drug release.²⁴  In drug release, a smart gel containing the desired water-soluble drug is injected into the patient. After receiving a certain stimulus (usually temperature or pH), a hydrogel will expand by allowing the water and salt in the blood to enter the gel. The drug will be released from the gel in the desired environment. This concept can be used to release drugs to attack tumors or aid specific areas of the body (i.e. eye drops for the eye). This concept is beneficial because the area, duration, and speed of the release can be better controlled with a smart gel. Nanoparticles will help this medical technology by increasing the effectiveness of the gel by increasing the surface area of its constituents.

Developments with respect to the electrical properties of smart gels could result in the future production of artificial muscles.²⁵  When an electric field is applied to certain types of gel, there is an asymmetric charge distribution within the gel. This asymmetry yields different rates of expansion throughout the gel layer. In fact, in some cases, one end might contract while the other expands. Asymmetry can also be created by producing heterogeneous gels with different rates of expansion throughout the gel. As a result of the electric field and asymmetry, the gel bends. This bending is significant because it mimics the role of muscles in the body which respond to electrical signals sent from the brain by creating mechanical energy. With more development, these gels could become prosthetic muscles for patients.

Poly(methyl methacrylate) is a transparent thermoplastic polymer often used as a shatter-resistant alternative to glass.²⁶  It has a good degree of biocompatibility and can be used for the replacement of intraocular lenses in the eye. In orthopedic surgery it is used as bone cement to affix implants or remodel lost bone. It is used for dentures and in dental fillings, and in cosmetic surgery in the form of suspended microspheres injected under the skin to permanently reduce scars.

Poly(glycolic acid) is a biodegradable, thermoplastic polymer with high initial tensile strength, smooth passage through tissue, easy to handle, excellent knot tying properties, and is commonly used for synthetic absorbable sutures, intracutaneous closures, implantable devices, tissue engineering scaffolds, bioabsorbable screws, and in abdominal and thoracic surgeries.²⁷  Scientists are also reporting the development and successful initial testing of the first practical “smart” material that may supply the missing link in efforts to use a form of light that can penetrate four inches into the human body. Near-infrared (NIR) light (which is just beyond what humans can see) penetrates through the skin and almost four inches into the body, with great potential for diagnosing and treating diseases. Low-power NIR does not damage body tissues as it passes. Figure 1.1 shows the new polymer, which has potential for use in diagnosing diseases and in engineering new human tissues in the laboratory.²⁸ 

Intelligent biomaterials that can respond to a biological signal show revolutionary promise in regenerative medicine, diagnostics, and drug delivery. Enzyme-responsive materials have the potential to detect, respond to, and ultimately repair biological processes.²⁹  For example, the materials could be used in medical devices that release drugs on receiving a biological signal from a cell. Enzyme-responsive materials change their properties when triggered by specific enzymes. Such materials can form a gel in response to the catalytic action of a protease enzyme and can be used as an injectable cell-scaffold that gels when triggered by tissue fluid enzymes. The flow of molecules into and out of polymer particles can be controlled by very specific enzyme switches – the first steps in making truly bio-responsive materials. An essential goal is to mimic the in vivo feedback mechanisms that control secretory endocrine enzyme activity.

Smart advanced synthetic materials are combining the sensing activity with actuating activity, so they can be used in implementing systems that respond to different chemical changes, can learn to detect certain patterns by using neural networks and learning algorithms, then react accordingly.

In order to understand the brain mechanisms and function, the last frontier of science and technology, subtle molecular processes must be detected and monitored using instruments of the same nanometric and atomic scale. The human brain has been called “a computer made of meat” by Marvin Minsky, an expert in artificial intelligence.³⁰  This computer apparently decides, chemically and genetically, how we feel, how we grow, live, and die. The only problem is that we are not sure how its unthinking atoms create our thoughts and experiences. Moreover, we are constantly reminded that it has quite a limited power, comparative to the sophisticated computers which can find solutions to astronomically complex mathematical problems at impressive processing speed. We cannot erase from our memory the embarrassment of Deep Blue, the chess-playing machine, beating Gary Kasparov, the reigning world chess champion, in that historic six-game match back in 1997. Deep Blue was making its decisions based on 200 million chess positions stored in its huge libraries of strategies, looking 20 moves ahead, while Kasparov could analyze just a few positions each second, relying more on intuition and experience than on processor power. The major difference in intelligence is that the brain can adapt to any changes in the rules of a game, while a computer is not able to adapt at all, unless it is programmed according to the new rules. Despite superhuman ability in chess, Deep Blue was not intelligent.

The brilliant mathematician Alan Turing, who delineated the foundation of computer science even before the first computer was built, defined the triumph of artificial intelligence as referring to the moment when an impartial judge will not be able to discern who the real person is and who the computer in a textual dialogue. Feeding a computer millions of bits of information and stuffing its huge memory space with enough rules to help it find a perfectly matching answer, a dialogue between a human and a machine now sounds like a normal “human” conversation. Then again, is a computer passing the Turing test really intelligent? There is no better test to define intelligence. Recent attempts in image recognition refer to the fact that, even if a computer can use a camera to detect an image, only an intelligent being can interpret what it represents, can analyze it, and reason about it. So far a computer can successfully make the difference between a cat and a dog, something that a toddler can do. No computer comes close to the richness of human perception.

Still, the “brain-as-computer” metaphor is an oversimplification at best. Conventional computers consist of transistors implementing a series of Boolean logical operations at very high speed. Brains function in a PARALLEL manner, doing thousands of operations simultaneously. The death of a neuron will not affect the brain, while if a single transistor is destroyed the whole operation of the computer is affected. A computer must be designed and programmed, while the brain comes “factory installed”, capable of plasticity, regeneration and repair, of thinking and learning due to its dynamic thousand trillion synapses. Like a transistor, a neuron can send an excitation signal to thousands of other neurons. However, the architecture of the brain includes chemical modulation at synapses and inhibitory circuits that can change the nuances of the incoming signals in a way so complex and sophisticated that cannot be paralleled by any electronic devices.

Above all, computers cannot assign MEANING to their programmed processes, as the eminent mathematician and computer scientist John von Neumann famously remarked. On the other hand, human mind is all about meaning. It creates meaning even when none is present.

The big question is: will technology will ever be able to reproduce the human brain or is the brain something impossible to replicate? A brain does much more than applying specific algorithms to a set of data. It is surely the greatest mystery that science has yet to discover.

Quantum dots are small (5–8 nm) inorganic compounds made of semiconductor or metallic materials with well-defined quantum states and electronic structure.³¹  They are composed of a metal core (cadmium, selenium, or cadmium telluride), a zinc sulfate shell, and an outer coating functionalized using bioactive molecules (Figure 1.2). Fluorescent quantum dots are used to visualizing molecular processes in neuron cells using fluorescent microscopy methods.³¹  Small changes in the radius translate into distinct color changes so they can be used to replace bulky organic fluorophores, which interfere with the molecular structure of the object of investigation. The design of these nanostructures is based on the ability to control plasmonic behavior in metallic nanoparticles, quantum size effects in semiconductor heterostructures with designed asymmetries, and nanoparticles with implanted dopants possessing sharp emission spectra. Their small size allows large and specific energy jumps between the energy band gaps of excited electrons or electron–hole pairs.

Inorganic nanoparticle optical probes can be tuned to match the photon energy requirements of the various excitation and detection systems. Unlike organic optical probes, they are photochemically robust during extended interrogation. For neuroscience studies, nanoparticles are combined with organic nanostructures for biofunctionalization, to attach them within neural cells configurations. The structure of semiconducting nanoparticles enables the generation of excitons, very sensitive to the external electric field. This sensitivity can turn these nanoparticles into reporters with externally modulated fluorescence intensity spectra. They may be combined with selective molecular binding moieties to confer sensitivity to changes in local neurotransmitter concentrations. Quantum dots can be used as local optical reporters for neuroscience, for visualizing dynamic molecular processes in neurons and glia on a large time scale, starting from seconds to many minutes, and on the small size scale of the synaptic cleft (20 nm) of neuron–neuron interactions or intracellular processes.³¹  Owing to their intrinsic voltage sensitivity, they could be used directly as optical readouts of membrane potential. These reporters must be embedded into neural membranes (thickness ∼2 nm) and react to local electric fields as well as local chemical environments. Functionalization with specific proteins make quantum dots capable of tracking receptors and functional responses in neurons (e.g. to glycine, nerve growth factor, glutamate, etc.).

Recent work using tools from atomic physics has shown that optically manipulated color centers in diamond provide exceptionally sensitive magnetic and electric field probes at sub-100 nm distances. Diamond is uniquely suited for studies of biological systems because it is chemically inert, cytocompatible, and ideal for coupling to biological molecules.

Nanomaterials that can provide nanoscale topographical features have become popular materials, as culture substrates with nanoscale features have significantly different effects on neuronal adhesion and growth. Vertical nanowires were shown to selectively promote neuronal adhesion and guide neurite growth even without the use of a cell-adhesive coating.³¹  Micropatterned islands of tangled carbon nanotubes also showed similar spontaneous adhesion and growth effects. Guided neuronal growth was reported on various nanotopographical substrates made of nanomesh carbon nanotubes, electrospun nanofibers, or patterned poly(urethane acrylate).³¹ 

One-dimensional structures such as nanotubes and nanowires may be used for highly local electrical measurements, for the delivery of photons to specific locations, and for the local release or collection of chemicals. These types of nanoparticles could be used alone, or combined with conventional organic chromophores, as they have been shown to greatly enhance optical signals, acting as an “antenna” for the light. Indeed, membrane-bound, antibody-linked gold nanoparticles have been already used for site-specific measurements of membrane potential. Traditional organic chromophores suffer from several drawbacks: they are large and can morphologically or chemically perturb the cellular environment. They can also bleach, that is, become ineffectual after exposure to light. Nanoparticles, by contrast, can be coated with a passivation layer or specialized shell that limits direct interaction with the surrounding media, which greatly minimizes bleaching and, inside the cell, the generation of reactive oxygen species. Present challenges include developing inorganic nanoparticles with enhanced voltage sensitivity, and orchestrating plasmonic enhancement of existing optical reporters. It should also be feasible to develop inorganic nanoparticles with voltage sensitivity, verify their biocompatibility, and identify and validate routes to targeted delivery. New organic nonlinear voltage probes and multifunctional nanoprobes that can locally report not only voltage but also chemical species (e.g. calcium, neurotransmitters, etc.), local temperature, or ionic environment are the focus for present research efforts.

Semiconductor nanowires can detect specific intracellular biomolecules, perform small-molecule drug screening, detect intracellular signaling, and also deliver drugs and genetic material into the cell.³¹  This nanotube spearing necessitates an oscillating magnetic field to spear the nanotubes, followed by a static field to drive them inside the cells. Figure 1.3 shows how these nanowires are not fatal to the cell. The cell remains functional for a few days and can even differentiate from stem cells. The mechanism of cellular engulfment of the nanorods and subsequent normal function remains to be explored.

Silicon nanowires have been implemented with either field-effect transistor-type active sensors or metal nanoelectrodes for in vitro neural sensors. The Lieber group at Harvard University has reported silicon nanowire field-effect transistor (NW-FET) arrays.³¹  They showed that simultaneous recordings from the axon and dendrites of a single neuron were possible with NW-FET arrays. In addition, neural signals ranging from 0.3 to 3 mV were recorded from neural circuits in brain slices using a NW-FET array. NW-FET is a promising sensor that can provide sufficient sensitivity with unprecedented spatial selectivity.

Field-effect transistors (FETs) can also record electric potentials inside cells. Their performance does not depend on electrode impedance and they can be made much smaller than micropipettes and microelectrodes. FET arrays are better suited for multiplexed measurements. SiO₂ nanotubes synthetically integrated on top of a nanoscale FET penetrates the cell membrane, bringing the cell cytosol (Figure 1.4) into contact with the FET, which is then able to record the intracellular transmembrane potential.

A branched intracellular nanotube FET (BIT-FET) possess a bandwidth high enough to record fast action potentials even when the nanotube diameter is decreased to 3 nm, a length scale well below that accessible with other methods. Studies show that a stable and tight seal forms between the nanotube and cell membrane, not killing the cells because the diameter of the tubes is so small. Multiple BIT-FETs can record multiplexed intracellular signals from both single neurons and networks of neurons.

In the case of metal nanoelectrodes, the Park group at Harvard University developed a vertical silicon nanowire array with individual nanowires 150 nm thick and 3 μm high.³¹  Several nanowires were grouped (2 μm spacing) to cover a single neuron, and an array of grouped nanowires was used to interrogate a small neural circuit. A high signal-to-noise ratio on the order of 100 was achieved, with the measured signal amplitude on the order of a few mV.

Carbon nanostructures are potential candidates to develop neural prostheses due to their similar nanoscale dimensions to neurites, as well as their unique electrical and mechanical properties.¹¹  When being used as a scaffold, they are able to repair injured nerves and even long gaps in severed nerves, by stimulating the healing of the severed ends in a nerve. Figure 1.5 shows the basic structures used as carbon scaffolds. These structures can be coated with thin layers of polymers in order to decrease the formation of glial scar tissue and to provide suitable sites for cell adhesion and proliferation.

Furthermore, the structures can be functionalized in order to improve biocompatibility by decreasing the toxicity. Figure 1.6 shows different solutions implemented in order to explore the controversial aspect of carbon nanotube toxicity, which manifests itself by decreasing cell viability through blocking ion channels in the membrane, increase of cell oxidative stress, and reduction of cell adhesion or induction of apoptosis.

Recent improvements with respect to prosthetic medical devices have been associated with the growth of carbon nanotubes on biocompatible Pt used as a catalyst. The new material shows less cellular degeneration due to oxidative stress because the main product at a Pt catalyst from oxygen reduction is water. Guiding axon regeneration through tubes made of materials such as chitosan, a biocompatible and biodegradable natural material, can provide, as a gel sponge, a suitable scaffold for nerve regeneration.

If the electrical response of neurons to applied voltages has been studied extensively, the mechanical response has been largely ignored; such research could advance knowledge of cellular function and physiology, especially in the area of axon elongation and dendrite formation. Using piezoelectric nanoribbons made of PbZr or Ti_1−xO₃ it was found that the cells deflect by 1 nm when 120 mV is applied to the cell membrane.¹¹  Such depolarization induces changes in the membrane tension so it is accommodating the stimulus by equalizing the overall pressure across the membrane through a process resembling converse flexoelectricity.

Figure 1.7 shows piezoelectric nanoribbons suspended over a trench as nanobeams to maximize deflection. The use of an underlying substrate of transparent MgO as well as transparent indium tin oxide (ITO) electrodes facilitates backside chip visualization during electrophysiology measurements. The electrodes are electrically isolated by a coating of SiN_x to ensure no cross-signal response when the chip is placed into solution. PC12 cells, a rat pheochromocytoma cell line that acquires many of the characteristics of sympathetic neurons when treated with nerve growth factor (NGF), were used. PC12 cells were cultured on the piezoelectric chip, and those cells located on the nanobeam arrays were patch-clamped with a standard glass electrode for membrane voltage stimulation.

In general terms, it is not clear how neurons interact with nanostructures, why they continue to function impaled by nanospears from all directions, and how they heal when confined in nanoscaffolds, in closing the gap between severed nerves ends. The reports on the toxicity of nanomaterials are extremely controversial. However, if any such undesirable effect can be minimized and these new materials can contribute to brain regeneration and repair, the outcome might be a positive one for the future. Like the miniaturized submarine crew in the 1966 movie Fantastic Voyage, smart nanodevices might one day be sent beyond the blood/brain barrier (BBB) to perform lifesaving surgery inside the brain tissue and destroy cancerous brain tumors with extreme precision and in a minimally invasive mode, by delivering the necessary drug, then leaving the body in a harmless way. It would represent a less expensive and far more effective treatment because only the target cells will be affected, without the side effects of today’s chemotherapies. Such an approach to medicine has the potential to transform health care for everybody.

Ever since the first isolation of free-standing graphene sheets were described in 2004, this two-dimensional (2D) carbon crystal has been highly anticipated to provide unique and new opportunities for sensor applications.³¹  Significant potential for application of the material in various novel sensors has been demonstrated. Useful features include the exceptional electrical properties of graphene (extremely high carrier mobility and capacity), electrochemical properties (high electron transfer rate), optical properties (excellent ability to quench fluorescence), structural properties (one-atom thickness and extremely high surface-to-volume ratio), and its mechanical properties (outstanding robustness and flexibility).

Graphene nanostructures exhibiting such excellent properties are very suitable for the fabrication of channel structures in field-effect transistors (FETs), which are typically used as electronic sensor devices to detect biomolecules and neural cell activity. This involves the incorporation of graphene into FETs, via insertion of a material with superior sensing properties in a structure of high sensitivity, simple device configuration, low cost, high miniaturization, and capable of real-time detection. A typical FET consists of a semiconducting channel between two metal electrodes, the drain and source electrodes, through which the current is injected and collected. Varying the gate potential through a thin dielectric layer, typically 300 nm SiO₂, can capacitively modulate the conductance of the channel. In a typical p-type metal oxide semiconductor field-effect transistor (MOSFET), the negative gate potential leads to the accumulation of holes (majority charge carries), resulting in an increase of the channel conductance, while the positive gate potential leads to the depletion of holes and hence a decrease of the conductance. In the case of the electronic sensor, the adsorption of molecules on the surface of the semiconducting channel either changes its local surface potential or directly dopes the channel, resulting in change of the FET conductance. This makes the FET a promising sensing device with easily adaptable configuration, high sensitivity, and real-time capability, providing, again, that the nonspecific absorption problem is solved using smart chemistry to prevent the fouling of the surface when the device is exposed to complex media such as human serum or blood.

In some cases, the gate electrode is removed, in order to simplify the device structure, to form a chemoresistor. Such a configuration is suitable for the fabrication of graphene-based sensors on polymer substrates for flexible electronic applications.³¹  Despite the lack of modulation by the gate potential, the working principle of the chemoresistor is the same as a normal FET sensor. Figure 1.8a shows a typical gas sensing system, where the channel is directly exposed to a target gas. The adsorption of gas molecules results in the doping of the semiconducting channel, leading to a conductance change of the FET device. The charge transfer from the adsorbed gas molecules to the semiconducting channel is the dominant mechanism for the current response, which is similar to carbon nanotube-based gas sensors.

In order to detect biospecies, the graphene FETs should operate in an aqueous environment. As shown in Figure 1.8b, the graphene channel is usually immersed in a flow or sensing chamber, which is used to confine the solution. The drain and source electrodes are electrically insulated to prevent current leakage from ionic conduction. Different insulators, including poly(dimethylsiloxane) (PDMS)/silicone rubber, SiO₂ thin film, SU8 passivation, and silicone rubber, are used in different device structures. The gate electrode, usually Ag/AgCl or Pt, is immersed in the solution. The gate potential is applied through the thin electric double layer capacitance formed at the channel/solution interface. The double-layer thickness (or Debye length) is determined by the ionic strength in the solution, typically within 1 nm. Normally, the solution-gate FET is over two orders of magnitude more sensitive than the typical backgate FET.

Two major sensing mechanisms have been proposed for graphene-based biosensors in solution, i.e. the electrostatic gating effect and the doping effect. The gating effect suggests that the charged molecules adsorbed on graphene act as an additional gating capacitance, which alters the conductance of the graphene channel. On the other hand, the doping effect suggests a direct charge transfer between the adsorbed molecules and the grapheme channel, similar to gas sensing. In a real case, the actual sensing mechanism might be a combination of both mechanisms, or involve more complicated mechanisms.

The detection of living cells is more challenging as the interaction between the graphene channel and living cell membranes is much more complicated. Graphene offers a improved opportunity to study the cell/nanomaterial interface since its 2D structure provides a homogeneous contact with the 2D cell membrane. A solution-gate FET to investigate living cell behavior is shown in Figure 1.9. The devices were fabricated based on large-scale micropatterned reduced graphene oxide (rGO) thin films with a thickness of 1–3 nm. Living neuron cells (PC12) were directly cultured on the rGO channel to obtain an intimate contact. The rGO FET was able to detect the adsorption of hormonal catecholamine molecules and those secreted from the living PC12 cells with a high signal-to-noise ratio. Moreover, the rGO FET can be fabricated on the flexible PET substrate and functions well during bending, which might be useful in complicated in vivo biosensing.

Collections of individual animals, birds, and fish sometimes act in precisely coordinated ways, as if moved by a group mind in perfectly choreographed movements. Herds of animals, flocks of birds, or schools of fish seem to exhibit intelligent group behavior. This “swarm intelligence” is used in the new “swarm theory” with interesting practical applications in artificial intelligence, to optimize the distribution of aircraft to the gates, move masses of passengers in airports as efficiently as possible, as well as route trucks in a smooth operating transportation system.

To explain bird flocking, fish schooling, beehives, or ant and locust colonies, scientists observed that in such large groups, neighbor-to-neighbor changes take less than 15 ms. Originated by single individuals somewhere in the group, such changes of direction propagate as a rapid wave throughout the group. When trying to mimic such behavior using artificially simulated graphic units, three main conditions were incorporated in the computer program:

1. Do not crowd nearby units

2. Fly in the average direction of nearby units

3. Stay close to nearby units

After running the program, the graphic result was a striking life-like movement. The conclusion: there is no coordinating command that has to propagate through the entire flock of birds or school of fish; the emergent behavior is the result of individual optical cues received by the members of the group.

Bio-inspired computation can be used in modeling cancer and bacterial growth, for testing dedicated algorithms for global optimization. Further applications mimicking insect swarms are the swarms of robots performing dangerous tasks such as mine sweeping, risky search-and-rescue missions, or even exploring the surface of Mars.³² 

“Living” machines based on advanced materials and sophisticated designs would have new characteristics such as fault tolerance, self-repair, self-replication, reproduction, evolution, adaptation, and learning. Such artificial life would be based on evolutionary computation algorithms and artificial or living neural networks, the former being interfaced with computing devices, for use in applications dangerous for humans (to control unmanned fighter planes, for example). Their interaction with the environment would be based on complex sensor systems such as computing tissues mimicking the human skin. Computing tissues can be used to design novel man/computer interfaces, intelligent and adaptive prostheses, intelligent doors, floors, walls, black boards, displays, etc. Artificial systems capable to grow, adapt, and reproduce in hardware are imagined and designed by scientists working in artificial intelligence. Figure 1.10 shows the fundamental element of a sensing computing tissue. It consists of a transparent touch-sensitive element, a LED color display, and a reconfigurable chip, a new fine-grained field-programmable gate array (FPGA). This element of the reconfigurable computing tissue consists of an input, an output, and the FPGA computing unit, organized in three hierarchical levels as shown in Figure 1.10. The inputs might include temperature sensors, force sensors, cameras, etc., and the outputs might include microphones, motors, speakers, displays, etc.

Parallel systems made of high numbers of such miniature elements would be able to compute in parallel. Such configurations, based on smart materials already employed in display technology and organic electronics, confirm the new trend towards intelligent interactive systems.³³ 

The nature of the interaction of material surfaces with biological fluids such as urine, blood, plasma, and serum is crucial in terms of both the biocompatibility of medical implants and biosensor devices, whether the latter are implanted or used in the clinical laboratory. Serious medical problems can arise from deleterious effects of materials on blood proteins and cells. For example, formation of thrombotic emboli upon contact of blood with materials employed in bypass circuitry during extracorporeal circulation (e.g. in cardiopulmonary bypass and renal dialysis procedures) is a well-recognized clinical issue.³⁴  Indeed, it is now known that micro-clots produced in such apparatus can result in cognitive disability in later years for patients subjected to dialysis and bypass surgery.^35,36  With regard to implantable biosensor technology, the lack of a closed-loop system with incorporation of a glucose device for operation with an artificial pancreas system has become a legendary research problem. Biosensors do not even figure prominently in the standard clinical biochemistry laboratory, largely because of the biological fluid/surface interaction problem.³⁷ 

A key aspect of negative effects caused by biologically foreign materials is the spontaneous adsorption of blood-based proteins on substrate surfaces.³⁸  This process is often referred to as “fouling” by those concerned with implant biocompatibility, whereas the analogous effect with regard to sensor technology is called “nonspecific adsorption” (NSA). The distinction between these two terms, however, is somewhat artificial since there is a great deal of common ground in their physical chemistry. Not surprisingly, tremendous research effort has been devoted over many years towards the development of protein-resistant surfaces in order to ameliorate the effects of deleterious interfacial interactions. The many strategies employed for this purpose, involving both coatings and covalently-bound layers, have been reviewed comprehensively by Blaszykowski et al.³⁹  Of the vast array of surface-modifying agents used in this field, peptides, poly(ethylene oxide)s/glycols, and zwitterionic sulfo- and carboxybetaines have been prominent. Despite the massive attention apparently paid to attempts to avoid fouling, it is fair to say that by far the majority of the effort has been on surface interactions with solutions containing single protein species. Far less research has been conducted on important biological matrices such as blood and serum. Here, we briefly describe a strategy originating from our own laboratory that does function in such complex media.

As alluded to above, the devices employed for hospital renal dialysis and coronary bypass surgery, often termed bypass circuitry, are generally composed of polymers such as polycarbonate (PC), polymethacrylate, and polyesters. In our own research we have been concerned with the covalent surface modification of PC using silanization chemistry. We report a new mono(ethylene glycol) (MEG-OH) silane adlayer coating for PC (Figure 1.11) that displays excellent anti-thrombogenicity, far exceeding that of the bare substrate.⁴⁰ 

Following thorough surface characterization by X-ray photoelectron spectroscopy, the treated PC surfaces were tested for anti-thrombogenicity in a real-time fashion with whole human blood labeled with fluorescent dye. Thrombus formation was visualized with an inverted fluorescence microscope. The MEG-OH surface displayed a 97% reduction in thrombus events when compared to the response of the bare polymer (Figure 1.12).

The mechanism responsible for this sort of dramatic result is still under debate, but it appears from neutron reflectometry experiments that interstitial water in the adlayer is at least partly involved.⁴¹  An important aspect of this chemistry is not simply that a water “barrier” prevents protein adsorption at surfaces, but that the adlayer-instigated structure of the water plays a key role. Circling back to the beginning refrain of this chapter, it is highly noteworthy that structured water is also considered to be a crucial component of the behavior of the head-group zone at the surface of biological membranes.⁴² 

The emerging field of biomimicry tells us that the most sophisticated technologies which involve future smart materials and sensing devices will be ones which emulate biological systems. There are already many examples of technological advances being proposed and developed that provide evidence for this. In the world of medicine, multifunctional materials will have the capability to detect, select, and execute specific “intelligent” functions in response to changes in the local environment, which is obviously crucial in terms of maintaining homeostasis. Their ability to recognize, analyze, and discriminate will be based on superior properties of self-repair, self-multiplication, self-degradation, self-learning, and artificial homeostasis. Although much remains to be accomplished, not least obviation of the serious biocompatibility issue referred to above, it is clear that smart materials science offers huge potential for future treatment of a plethora of disease and injury conditions. Maybe one day we will see a true “bionic man”!