Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Power from nano-engineered wood

By chemically treating wood it is possible to fabricate a nanofluidic device for generating electricity from the harvesting of ubiquitous low-grade heat.

Climate change concerns are forcing the scientific and industrial communities to look at ways to sustain economic productivity while minimizing the use of traditional non-renewable fossil fuels for energy generation. In addition, there is also a significant push towards implementing energy-efficient technologies to reduce the carbon footprint. Even so, in modern life there is an enormous quantity of waste heat that is generated. To put this into perspective, a 2016 report1 notes that the US alone consumed about 97.3 quads of energy, where 1 quad is 1.055 × 1018 J, and 68.4 quads of this was wasted as heat. Based on this analysis as well as a US EPA study2, we estimate about 50 quads of energy is annually rejected as low-grade waste heat, where low-grade is defined as a temperature below 150 °C. The annual amount of low-grade waste heat lost is equivalent to the energy produced by 1,700 average-size 1-GW nuclear reactors, which can be understood by noting that only about 100 nuclear reactors operate in the US. Global low-grade waste heat is estimated to be about four to five times that of the ~50 quads of the US. Various technologies can be used to harness this low-grade waste heat, such as the organic Rankine cycle3, the Kalina cycle4, direct solid-state heat-to-electric conversion using thermoelectric materials5, and pyroelectric conversion methods6. The fundamental challenge in all these approaches is that device-level heat-to-electric conversion efficiencies using heat sources below 150 °C need to be significantly improved, and the manufacturing processes scaled up.

An approach to develop thermal harvesting devices at large scale is to use materials that are cheap, plentiful and easy to process. This would favour naturally occurring organic materials. Now, writing in Nature Materials, a study by Tian Li et al.7 reports that it is possible to process naturally occurring cellulose, from trees, to prepare a thermal harvesting device that potentially has a good efficiency for electricity generation.

Li et al. report that it is possible to take the wood of a common tree, in this case American basswood, remove sections of the trunk, extract the heart wood (Fig. 1a–c) and use the intrinsic nanostructured membrane channels as the basis for a nanofluidic device (Fig. 1d). To do this, Li et al. first performed a delignification, removing hemicellulose and lignin. The remaining cellulosic nanofibres are 30 nm wide, with each nanofibre formed of elementary nanofibrils separated by 2 nm. These nanofibres then undergo a further oxidation treatment, using 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), whereupon NaOH electrolyte is then infiltrated between parallel stacked molecular cellulose chains spaced 0.7 nm apart. The hydroxyl groups on cellulose have a tendency to dissociate into negatively charged O groups. After processing, the hydroxyl groups of the cellulose nanofibres are converted to carboxyl groups, leading to a negative surface charge density (Fig. 1e). The resultant electrostatic field surrounding the cellulosic nanofibres triggers surface-charge-governed ion transport along the fibre direction, by providing a disparity that in turn facilitates easier transport of Na+ ions towards the cold end relative to the OH ions, while diffusing from the hot end (Fig. 1f,g). Thus the infiltrated woody material offers a unique mobility difference between the Na+ and OH ions under a thermal gradient, facilitating electric power generation from thermodiffusion, akin to the Soret effect8. The uniqueness of the work of Li et al. is in the use of wood as a starting material and its nanochannels for charge selectivity during fluid transport across a thermal gradient. This temporal disparity in ion transport and the resultant potential they produce can be observed to cause a capacitive-like power discharge in an external electrical load (Fig. 1h).

Fig. 1: Processing wood to form a nanofluidic thermal harvesting device.

ac, A common tree trunk is sectioned into discs, with heart wood then extracted. d, Processed wood section showing the cellulosic membrane that exhibits a hierarchically aligned structure. e, The nanostructured, negatively charged channels are infiltrated with electrolyte. f, The microscopic distribution of positive (Na+) and negative (OH) ions is uniform in the infiltrated nanochannels. g,h, The application of a thermal gradient preferentially causes the positive ions to drift towards the cold side, causing a transient output voltage V that varies with time t and can be discharged through an external electric load.

Using this approach Li et al. can generate a high differential thermal voltage of 24 mV K–1, four times better than previous reports, with a temperature gradient of ~5 °C. With ideal electrical contacts, from the measured thermal voltages and estimated heat flow, they calculate a heat-to-electric conversion efficiency of ~1% during a capacitive discharge. This efficiency, although small, represents about two thirds of the thermodynamically limited Carnot efficiency, which is exciting for a wood-based material. Most thermal energy harvesting devices, assuming they can utilize small gradients like 5 °C in the first place, offer ~0.1% efficiency.

While this initial demonstration is exciting, further work needs to be done to fully develop the concept. The device needs to generate steady-state power with efficient electrical contacts in the presence of heat flow and reactive chemical agents like NaOH. To achieve steady-state power output, one needs to implement innovative heat cycling (utilizing mechanical methods) across the membrane, or carry out redox reactions8 at the ends of the electrochemical cell. There will be energy requirements for the mechanical heat cycling or redox processes, and these have to be optimized not to reduce the overall heat-to-electric efficiency. The ion-selective nanomembranes under a thermal gradient could also be relevant for fuel cells, batteries and epidermal electronics.

For any thermal harvesting technology, technical and economic feasibility at a system level are important. In particular, can the device use both gas and liquid waste heat streams? Can the approach be extended to larger temperature gradients, beyond the 5 °C shown in this work, for higher efficiencies? What is the scalability and cost of the delignification and TEMPO steps in producing uniform woody nanochannel membranes and that of any required heat exchangers? Are the nanostructured cellulosic materials structurally and chemically stable over long periods of time in the presence of heat and electrolyte? These are questions worth addressing, since this approach appears to offer an efficient way to harness the ubiquitous low-temperature waste heat in this world.


  1. 1.

    Estimated U.S. Energy Consumption in 2016 (LLNL, 2017);

  2. 2.

    Waste Heat to Power Systems (EPA, 2012);

  3. 3.

    Bao, J. & Zhao, L. Renew. Sustain. Energy Rev. 24, 325–342 (2013).

    CAS  Article  Google Scholar 

  4. 4.

    Knudsen, T., Clausen, L. R., Haglind, F. & Modi, A. Energy Procedia 57, 361–370 (2014).

    Article  Google Scholar 

  5. 5.

    Venkatasubramanian, R., Siivola, E., Colpitts, T. & O’Quinn, B. Nature 413, 597–602 (2001).

    CAS  Article  Google Scholar 

  6. 6.

    Zhu, H., Pruvost, S., Cottinet, P. & Guyomar, D. Appl. Phys. Lett. 98, 222901 (2011).

    Article  Google Scholar 

  7. 7.

    Li, T. et al. Nat. Mater. (2019).

    CAS  Article  Google Scholar 

  8. 8.

    Dietzel, M. & Hardt, S. Phys. Rev. Lett. 116, 225901 (2016).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Rama Venkatasubramanian.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Venkatasubramanian, R. Power from nano-engineered wood. Nat. Mater. 18, 536–537 (2019).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing