An electrifying discovery

Two scientists in Bangalore, Ajay Sood and Shankar Ghosh, make a discovery with great potential for technology application - that electricity can be generated through the flow of gases over solids.

IN today's world of science, where electronics, advanced materials and quantum theory dominate technology development, it is not often that simple school-level physics leads to some unexpected discovery with significant technology application potential. However, Ajay Sood, a Professor of Physics at the Indian Institute of Science (IISc), Bangalore, and his student, Shankar Ghosh, have achieved just that. They found that the mere flow of gases at modest speeds over the surfaces of small samples of a variety of solids generated sizable electric voltages and current.

The finding has been reported in the August 20 issue of the prestigious journal Physical Review Letters.

It was a surprising discovery, but not entirely unexpected. They were prepared for the surprise because of the equally surprising result that they discovered last year (Frontline, February 14, 2003). They found that the flow of liquids, even at low speeds ranging from 10-1 metre/second to 10-7 m/s (that is, over six orders of magnitude), through bundles of atomic-scale straw-like tubes of carbon known as nanotubes, generated tens of microvolts across the tubes in the direction of the flow of the liquid. (Nanometre means a billionth of a metre or 10-9 m. Nanotubes are tiny graphite sheets rolled into cylinders of size 1-100 nm through appropriate preparation techniques and have distinct electronic and structural properties.) This led them to the question whether the flow of gases could also generate an electric current.

Their surprise, therefore, was not so much in the finding itself as in the underlying physics, which was fundamentally different from the mechanism for current generation through the flow of liquids. It was physics that has been known for nearly two centuries. There is nothing in the nature of the experiment, including the levels of precision required in measuring the various parameters, to suggest that the discovery could not have been made, say at least 50 years ago. Only, no one thought of doing the experiment. In the case of the experiment on the flow of liquids, at least the peculiar electronic properties of nanomaterials and the one-dimensional nature of carbon nanotubes came into play.

The IISc scientists found that when common gases such as argon, nitrogen and oxygen were allowed to flow over a variety of solids, including single-walled and multi-walled carbon nanotubes (SWNTs and MWNTs), doped semiconductors and metals, with velocities in the wide range of 1-140 m/s, there was direct generation of electric current and voltage. The voltage generated, between tens and hundreds of microvolts, depended on the square of the velocity of the gas flow. However, in the case of liquids, the voltage generated varied according to the `logarithm' of the velocity, suggesting that an altogether different mechanism was in operation.

When Sood and Ghosh began to investigate the effect of gas flow, the mechanism they were looking for was the same as in the case of liquid flow, namely the direct forcing of the free charges in the nanomaterial (arising from its electronic structure and property of conductivity) by the ions in the gas (arising because of its polar nature which cause an intrinsic separation of positive and negative charges).

According to Sood, since the cost of polar gases like ozone or sulphur-hexafluoride (SF6) was prohibitive, they settled for gases like oxygen and nitrogen, which have some polarity, and planned to compare their effect with that of a completely non-polar gas like argon. The experiment, exactly analogous to the liquid experiment in which the fluid flow was horizontal over a single-walled carbon nanotube sample, showed no detectable effect in the case of all the three gases. But in trying to optimise the experimental set-up by tilting the orientation of the sample with respect to the direction of gas flow and so on, they found that there was current flow when the inclination was non-zero and that the effect was maximum at an angle of about 450. And the voltage generated, it was found, varied according to the square of the velocity of the gas .

Sood and Ghosh investigated gas flow over a variety of materials - semiconductors of both n-type (where negative electrons are the charge carriers) and p-type (where positive holes are the charge carriers), graphite, SWNTs and MWNTs, metals like copper and platinum and so on. The sample sizes were kept small, with dimensions of about 1-2 mm, in order to prevent turbulence in the gas flow, which can happen when the obstacle sizes become large. The quadratic dependence on velocity was seen in all the samples, over a wide velocity range covered in the experiments, though the rate of increase varied from material to material.

Initially, the scientists were clueless about the mechanism at work, which caused a behaviour that was distinctly different from what was observed with liquid flow. "I could not sleep that night and kept thinking about the peculiar behaviour," says Sood. "I thought perhaps some subtle fluid dynamics was at work and looked up the classic Landau and Lifshitz's Fluid Mechanics to understand fluid flow over obstacles. There was this discussion about pressure changes across the obstacle face arising from Bernoulli's principle and I began thinking along those lines. Since we were dealing with gases, this led me to the gas law and to the Seebeck effect for voltage generation," Sood explains. Now all these principles that have been invoked to understand the phenomenon - Bernoulli's principle, gas law and the Seebeck effect - all fall in the realm of school physics.

Schematic diagram of the experimental set-up for power generation by gas flow

Bernoulli's principle, named after Daniel Bernoulli (1700-82), is a statement of conservation of mechanical energy - the sum of pressure energy, potential energy and kinetic energy - in a fluid (liquid or gas). As a consequence, in fluid flow, an increase in velocity results in a pressure drop within the fluid. When an inclined surface is presented to a gas in steady flow, the leading edge of the surface where the gas meets the surface and stops, the velocity is zero (zero kinetic energy) and the pressure increases (as pressure energy must compensate for the loss in kinetic energy). As the gas flows up the surface, the velocity is non-zero (that is, it increases) and hence the pressure drops so that the total energy remains constant. That is, over the surface, the gas flow is accelerating and the pressure is reducing from the leading edge to the trailing or the far edge. This is exactly what happens when wind strikes the windshield of a car, for example. Streamlining of car shapes or an aerofoil is based on this principle and intended to reduce the drag on the vehicle.

Gas law, the second physical principle at work, states that pressure changes and temperature changes in a gas (at constant volume) have a direct relationship. So, a fall in pressure results a decrease in the temperature of the gas. The temperature change in the gas induces a temperature change in the solid surface in the direction of the gas flow. That is, there is an increase in temperature from the leading edge of the inclined surface to its far edge. This temperature difference results in a voltage difference owing to the third underlying physical principle, known as the Seebeck effect, named after T.J. Seebeck (1770-1831). The Seebeck effect states that when a thermal gradient is applied to a solid, it will be accompanied by an electric field in the opposite direction. The combined effect of the three principles is that mechanical energy is converted into thermal energy, which in turn is converted into electrical energy as a voltage and current across the sample.

The voltage developed depends on the material characteristics measured in terms of the Seebeck coefficient, which is the amount of voltage generated per degree of temperature gradient. The higher the Seebeck coefficient of the material, the greater the voltage generated for a given velocity of gas flow across the material. It is this variation in the Seebeck coefficient of the materials used that resulted in the observation that the increase in voltage with velocity showed different rates for different materials. In fact, the rate of increase was seen to be directly proportional to the Seebeck coefficients. The upshot of this observation is that to generate greater electric power from gas flow one must choose a material with a higher Seebeck coefficient.

In the experiment of Sood and Ghosh, the Seebeck coefficient varied from about 0 (zero) microVolt/degree (for platinum) to about 500 microVolt/degree (for semiconductors). Indeed, platinum showed practically no voltage generation across a large velocity range, thus confirming the understanding of the mechanism on the basis of the Seebeck effect. The maximum observed voltage generation in the Sood-Ghosh experiment was in the range of a few hundred microvolts, implying that across the samples the temperature changes were of the order of a degree. The current generated was in tens of microamperes.

For example, in the germanium sample, the voltage generated was 650 microvolts and the power generated was 43 nanowatts. Since, by the very nature of the mechanism, the effect is not restricted to the materials studied, Sood and Ghosh point out that the guiding principle to generate a higher voltage/current should be the choice of materials, such as selenium, tellurium, gallium-arsenide, oxides and electrically conducting polymers, all of which have high Seebeck coefficients.

The observed effect immediately suggests the following technology application, namely gas flow sensors to measure gas velocities from the electrical signal generated. Unlike the existing gas flow sensors, which are based on heat transfer mechanisms from an electrically heated sensor to the fluid, a device based on this newly discovered effect would be an active gas flow sensor that gives a direct electrical response to the gas flow. One of the possible applications can be in the field of aerodynamics; several local sensors could be mounted on the aircraft body or aerofoil to measure streamline velocities and the effect of drag forces. "It is possible to have a direct control over these forces and a digital means of handling them," says Sood.

As Sriram Ramaswamy, a colleague of Sood and Ghosh at the IISc, points out, one could deploy an array of these sensors to map out the flow patterns over a large region in a wind tunnel or other aeronautical settings. According to aerodynamics specialist Rama Govindarajan of the Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR), Bangalore, such flow sensors could be valuable in atmospheric sciences and meteorology and in chemical engineering applications where such tiny sensors that can give flow direction and magnitude are sought after.

The sign of the voltage generated differs between n-type semiconductors and p-type semiconductors. And the voltage is sizable enough to tell an n-type from a p-type by simply blowing on the semiconductor and measuring the voltage, says Sood.

Just as in the case of power generation induced through liquid flow, this phenomenon too suggests the tantalising futuristic possibility of large-scale power generation through the flow of gases over solids. In their paper, the duo suggests the possibility of converting gas flow energy directly into electricity without any moving parts. However, since the power generated in mm-sized solid elements is only in nanowatts, the question that naturally arises is whether the system can be scaled up. In large solids, problems of unstable gas flows and turbulence effects are likely to arise. So the strategy that the scientists suggest is that the individual small sensor elements be connected in series or parallel to scale up voltage and current generation. Also, the problem can be turned around, as Sood and Ghosh plan to do, to study turbulence itself. These sensors can be used to give point-by-point time-resolved velocity measurements in order to provide a better insight into the phenomenon of turbulence.

The curious question, however, still remains: Why was this phenomenon not discovered before? "The existence of such pressure and temperature difference was not unknown," points out Ramaswamy.

"The clever idea (of Sood and Ghosh) here is that this classical effect of fluid mechanics and thermodynamics can give rise to a voltage and an electric current along a solid surface. Not entirely surprisingly, it seems no one thought of connecting a voltmeter across an aerofoil in a wind tunnel and looked for millivolt potential differences."

While Sood and Ghosh have provided an interpretation for the phenomenon, which the observations too seem to bear out, one question still remains. Why is the mechanism that was observed in liquids not seen in gas flows? It was perhaps owing to the low polarity of the gases used and because the voltage generated by the mechanism was too small to be measured. But what if highly polar gases such as SF6 are used in the experiment? Which effect will dominate? This question will be engaging Sood and associates - notwithstanding the high cost of such gases - in the near future.