Strength in Numbers: Bending Thermodynamics to Get Voltage for Free

Building an energy conversion material that circumvents the Second Law of Thermodynamics is revealed to be feasible.

(University of Cologne, Max Planck Institute for Polymer Research) – Let’s try a thought experiment. Imagine a box with the properties of a black body, so as to absorb all light from without and within. Inside the box are a lightbulb and an electric generator, the latter capable of converting ambient heat into voltage, with which it powers the lightbulb. Imagine you close the box. Can the lightbulb keep giving light indefinitely? Is there any way you can tell, from outside the black box? Does the generator break Thermodynamics? Can you say with certainty, if you can’t tell from outside the box?  

The Best Case Scenario

A fundamental question exists, in the field of energy research, of the degree to which improvements in efficiency and renewability can actually extend.

Thermoelectric materials are among the best energy solutions in terms of efficiency/footprint ratio, as they constitute very simple machines very directly converting a thermal differential into voltage, with no immediate waste products. However, their working principle also puts a heavy limitation to their use, as the necessity for thermal differentials restricts them to environments where temperature varies enough within close distances, a condition both generally undesired for human habitats, and often requiring to be artificially kept.

If we were to eliminate the necessity for temperature differentials, however, and produce an equivalent of thermoelectrics capable of producing voltage out of uniform heat, this would constitute a new ideal benchmark for clean energy harvesting.

If of easy enough fabrication, industrial and consumer-grade applications of this sort of technology would be virtually unlimited, as it would present all the advantages of thermoelectrics such as the absence of waste, while being safe from all drawbacks, such as the necessity for significant local thermal differentials and the eventual consumption of the differentials themselves.

How much of a benchmark would this technology be? Let’s follow the white rabbit.

With unstructured ambient heat as the only fuel needed, the immediate consequence is that any cell based on this technology would produce voltage indefinitely as long as the environment around it were thermalized. Since most environments inhabited by humans are reasonably thermalized to begin with, the technology would represent the closest solution to literally free energy. Concerning electric output, there is no reason to think that optimization couldn’t deliver efficiencies comparable with thermoelectrics, with the added crucial advantage of providing energy for an unlimited amount of time. Indeed, the only way in which energy from such systems couldn’t be called free would be the need for the user to purchase the generator in the first place. But that, of course, would be a blessing for the market.

In both cities and smaller settlements, arrays built out of this technology could be buried next to natural heat sources to provide easily harvested geothermal power. If integrated within the chassis or the engine of vehicles, the technology could recycle the heat from both motor and solar irradiation. Wearable and portable electronics could be made to run on body heat, ranging from smartphones that recharge by body contact to pacemakers that never stop as long as the body is alive and warm. In space travel, where every bit of energy is precious, the technology could make a difference by providing a free source of illumination and other basic functions, crucial for comfort to the human crew, as well as recycling power back into the machinery. In warzones or emergency situations, cells could quickly be set to provide much needed power.

A pipe dream? The answer, of course, lies within the rules of nature.

The Confines of Improvement

As we need to move within the laws of Physics, further optimizations to our energy solutions are faced with more and more constraints of a qualitative rather than quantitative nature. The Shockley-Queisser efficiency limit[1] is a prime example concerning photovoltaics, as is the deeper and more universal constraint of the Second Law of Thermodynamics. The Second Law in particular holds special relevance, as its inherent requirements are what binds most technologies to the base necessity of consumption.

There are two main ways to formulate the Second Law of Thermodynamics, both equivalent to each other: the first is that a physical system cannot decrease its own degree of disorder, a.k.a. its entropy; the second, that any machine needs at least two thermal reservoirs – one hot, one cold – in order to perform work. The second formulation is a phenomenological expression of the first: if a machine could perform work out of a single thermal reservoir, it would effectively be converting chaotic energy (random thermal motion) into a more ordered form of energy (the work performed), thus lowering the entropy of the system. So if a machine must draw upon chaos to perform work, the chaos must at least possess a degree of order by consisting of two reservoirs, distinct in position and temperature, to sooner or later be reduced to a uniform, lukewarm and at last exhausted system. Motion goes from hot to cold. Motion makes the machine work, then motion stops. The machine stops. Consumption wins.

The People v. Feynman

While the First Law of Thermodynamics is a rock-solid foundation and something that everyone can easily agree upon – “Energy cannot be created or destroyed” – the Second Law has something of a more slippery quality, subject to interesting boundary conditions, the primary of which being the intrinsic requirement of working over large ensembles of particles. This by necessity binds it to either characteristic length scales observed, or thresholds in the number of particles involved. Already in early stages of Thermodynamics studies the implication was discussed that, if one were able to control particles at the local scale, a way could be theoretically found to escape the confines of the Second Law. Indeed, such personalities as James Clerk Maxwell and Richard Feynman have entertained thought experiments in this direction, leading to such concepts as the entirely theoretical Maxwell’s demon[2], and the less philosophical Brownian ratchet[3]. Feynman, however, also offered a thought counter-argument on why the ratchet could not work in practice, which did much to dampen the debate.

However, as the topic moved away from the pure speculation of the Sixties and into the systematic investigation of the nanoscopic, it gradually became evident that systems that work outside the Second Law not only exist in nature, but they are actually ubiquitous in the biological world at the nanoscale, now known as Brownian motors.

Brownian Motors

Existing at the nanometric scale, the asymmetric machines known as Brownian motors can produce work out of random Brownian excitation, thanks to their size making them sensitive to the environment on the granular level, where thermal fluctuations dominate over thermal averages. Under these conditions, the very concept of heat reservoirs begins to break down, and conventional Thermodynamics give way to the more lenient non-equilibrium Thermodynamics, allowing for the boon of temporary fluctuations in entropy within given intervals in time and space. Under random, granular Brownian motion, the asymmetric structure of the tiny Brownian motors is pushed to perform a tiny amount of work in a direction determined by their asymmetry. And as this mechanism is inherently advantageous, it’s not a huge surprise that biological Brownian motors are pervasive in nature, the muscle protein Myosin[4] being a well-known example, propelled to climb Actin filaments by random interactions with a thermalized surrounding environment[5].

Riding the tiger of this knowledge, attempts have been made in very recent years at theorising possible ways to artificially circumvent the Second Law [6,7], as well as both theoretical[8] and practical[9,10] investigations into actively engineering and extending the conditions under which Brownian motors operate to larger systems.

The case can be made, at this point, for the virtues of macroscopic electric generators sharing the non-equilibrium Thermodynamics privileges of Brownian motors. If attained, such machines would work as two-states systems, where the natural high-entropy state of unstructured heat, as provided by the environment, would be converted by the material into a temporary lower-entropy state, such as direct voltage yielded by the generator or polarity stored in a battery; and finally reverted to disordered heat and the original high-entropy state once the potential were used up. We would then possess a generator presenting null variation in entropy over long enough time intervals, which would constitute an actual application of our best-case-scenario heat-recycling technology.

Aggregating the Nanomachines

Imagine now an artificial, but still conventionally nanometric Brownian motor – or, rather, imagine a nanometric Brownian electric generator, where thermal excitation doesn’t cause the motor to move itself in a given direction, but rather to move electrons within its structure in a given direction. Like Myosin takes tiny steps along Actin filaments, so does our nanometric generator produce a nanovoltaic electric potential. (Video of Myosin walking seen on

Imagine you develop a method to reliably produce a billion copies of our nanometric generator, and to closely pack them together so as to form a compact material. Each sensitive to heat at the granular level, they each yield their own minuscule electromotive force. Imagine, finally, that you also develop a method to deposit the billion nanometric generators in place so that they are not randomly pressed together, but rather ordered within the material so as to face a common direction.

What happens at this point? If your answer involves the billion nanometric contributions summing up to a macroscopic one, you would be right.

The Plot Thickens

This is not a thought experiment.

The author would like, at this point, to show his cards and announce the successful production and characterization of a single-reservoir thermal-to-voltaic conversion material as described, to obtain what is effectively a first example of Brownian motor – or, in this case, electric generator – existing on the macroscale. A patent for it has recently been filed, and peer-reviewed publication is currently underway, covering working details and fabrication method of the system.

The system was produced in the form of a composite material, in which an active phase was constituted by rectified nanometric Brownian generators and a second, inert phase served as scaffolding for the active phase. Upon characterization it was observed that, indeed, in such a rectified system the individual nanovoltaic contributions do add to each other, rather than cancelling each other out as it would be expected in the case of a non-rectified system. The observable result is a macroscopic electromotive force, determined by both the efficiency of a single nanometric generator and the number of copies packed within the material. The obtained material is macroscopic, but its structure is sensitive to heat on the nanometric, granular scale, and is thus capable of converting thermal motion from single, macroscopically uniform reservoirs into usable voltage. In other words, thanks to its internal structure the material can work according to non-equilibrium Thermodynamics on a larger scale than previously observed in natural systems.

The picture above provides a conceptual schematic of the material, where an individual nanometric Brownian generator is shown as an asymmetric object in the magnification on the top left, and where the common orientation of the ensemble of motors is also represented. For ease of representation, the scaffolding phase is not represented in the picture. While the picture is of course not a reflection of the actual structure of the material, the observed internal structure of prototypes produced by the author has shown itself to reflect the general principle.

Thermodynamically, usage of the material can be seen as an entropy fluctuation of macroscopic, rather than nanoscopic, magnitude, but the underlying physical principles remain the same. In the end, over a long enough time, the produced voltage is used and entropy returns to the system, and thus the laws of Physics stay safe and sound. To the universe it matters little whether an entropy fluctuation lasts a nanosecond or a week, but to humans the fluctuation is in this case large enough that we can engineer it as an affordable, virtually unlimited thermal-to-voltaic mean of power conversion. Below, a scheme of the fluctuation.

The author estimates that commercial-grade implementation of this technology can be feasible within two years time.

One More for the Road

On the phenomenological side, the newly observed single-reservoir electromotive effect can be seen as a static electromotive effect, function of constant and uniform temperature rather than temperature gradients in space, as in thermoelectrics, or in time, as in pyroelectrics. To keep with the heat-inspired naming convention, the author likes to refer to the new effect as Igneoelectric Effect, from ignis, Latin for fire.

So what happens when you find out that the best case scenario is not only possible, but actually within reach? The author can speculate, but is pretty much convinced that these are interesting times.


An outcome of the University of Cologne and a former Max Planck Society member, Dr. Riccardo Raccis is an independent technology developer in the field of nanocomposites. He welcomes discussion on this topic, its implications, fundamental questions, and commercialization. He can be reached at

Pin It on Pinterest