User login


You are here

Journal Club Theme of August 2011: Energy Harvesting Using Soft Materials

Adrian S. J. Koh's picture

Energy harvesting is the process of converting energy that will otherwise be dissipated into the ambient environment, into useful energy to do work.  I shall focus this discussion on motion-based energy harvesting.  Motion-based energy harvesting is the process of converting dissipated mechanical energy into electrical energy.  Sources of mechanical energy include the ocean waves, wind, human motion, vehicular traffic, and vibrations in buildings and bridges.  This source of energy is ubiquitous and pervasive, and yet, it is one of the least developed energy harvesting technology.

Traditional methods of motion-based energy harvesting adopt piezoelectric, electrostatic MEMS and electromagnetic (EM) generators as the media of conversion.  These techniques are incapable of packing a huge amount of mechanical energy as they are either too stiff (piezos), or too compliant (electrostatic MEMS and EM generators).  This results in poor yield-to-size energy conversion ratio.  Typical piezoelectric systems are able to produce a yield in the region of uJ (micro-joules), with an energy conversion efficiency of less than 10%.  EM generators are able to produce energy at a much larger scale, but it requires an enormous system to do so.  Typical yield of EM generators is in the region of mJ/g of the system, with a conversion efficiency of about 20%.

I shall focus on a technology using dielectric elastomers as energy harvesters.  This technology was proposed by Pelrine et. al., in 2001.  Dielectric elastomers are thin membrane of polymers sandwiched between compliant electrodes.  When subject to a voltage through the thickness of the membrane, the elastomer thins down and expands in area.  It works as an actuator.  On the other hand, when a pre-stretched and pre-charged elastomer is mechanically relaxed in the open-circuit condition, voltage across the electrodes may be boosted.  It works as a generator.  In Pelrine’s work (2001), they used a thin membrane of polyacrylate, very-high-bond (VHB) adhesive tape (manufactured by 3M), and sandwiched the membrane between compliant electrodes made from carbon grease.  They managed to produce a voltage boost of five times the input voltage, and computed a potential energy conversion capacity of 400 mJ/g.

Inspired by their work, me and my coworkers at Harvard and Johannes-Kepler University in Linz, Austria, performed theoretical calculations to estimate the maximum amount of energy that may be converted using a dielectric elastomer generator (DEG).  We assume that a DEG will cease to produce useful energy if one or more of the following occurs:  Rupture, electrical breakdown, electromechanical instability and loss of tension.  Using equilibrium thermodynamics, we estimated that VHB has the potential to produce 1.7 J/g of energy, and that natural rubber produces a comparable amount at 1.3 J/g.  We expect the actual yield of VHB to be very poor due to the large mechanical and electrical dissipation in that material.  On the other hand, natural rubber looks to be a promising material due to its low dissipation, and high resistance to electrical breakdown.  We further proposed that the energy of conversion scales linearly with the dielectric constant, and to the square of the dielectric strength.  Our analysis shows that a material with a dielectric constant of 6.0, and dielectric strength of 100 MV/m, is capable of converting 1.0 J/g of energy when cycled at 100% strain.  This amount of energy conversion, if realized, will be orders of magnitude higher than current technologies.

There is much work left to be done in this field.  The mechanisms of dissipation in dielectric elastomers are not well-understood.  This hampers the evaluation, design and optimization of energy harvesting systems using DEGs.  Furthermore, experiments conducted on dielectric elastomer as generators are limited to only VHB and silicone elastomers.  There is a wealth of materials out there waiting to be discovered.  Finally, the durability of elastomer materials may be a crucial factor in determining if this technology takes off or not.  Fracture and fatigue of elastomers must be better understood.


Pelrine et. al., Proc. SPIE 4329, pp. 148–156, 2001.
S. J. A. Koh, X. Zhao & Z. Suo, Appl. Phys. Lett. 94, 262902, 2009.
S. J. A. Koh, C. Keplinger, T. Li, S. Bauer and Z. Suo, IEEE/ASME Trans. Mech. 16, 33–41, 2010.
T. McKay et. al., Smart Mater. Struct. 19, 055025, 2010.
T. McKay et. al., Appl. Phys. Lett. 97, 062911, 2010.


Xiaodong Li's picture

Adrian, this is a very interesting topic. I would like to know how biological soft materials do energy harvesting and how to store it? Do you know any papers? Thanks.

Adrian S. J. Koh's picture

Hi Xiaodong,

Soft materials are capable of converting mechanical to electrical energy in several ways, I shall just touch on the conversion process of a deformable capacitor (also known as a dielectric elastomer).  A dielectric elastomer (DE) consists of a thin membrane of polymer (for instance, rubber), sandwiched between compliant electrodes (for instance carbon grease).  The DE is first pre-stretched and pre-charged with a small electric field.  After which, the DE is mechanically relaxed.  When it is relaxed in the open-circuit condition, the electrodes are separated, separating the unlike charges, and squeezing the like charges closer together.  This action increases the potential difference between the electrodes, thereby boosting the voltage.  When it is relaxed in the closed-circuit condition, the charges are pumped to an external circuit, creating a current that powers a electrical load.

The generated electrical energy may be used to directly power a load, or be stored in a battery or capacitor.

The paper by Pelrine et. al. addresses the basic mechanisms of energy conversion (first attached paper in original post), my papers illustrate that the energy density of conversion for soft materials may be orders of magnitude higher than existing technologies like piezoelectrics and EM generators (next two papers).  The final two papers by a New Zealand Group (headed by Prof. Iain Anderson) introduces a creative circuit design that allows a DE generator to boost voltage from Volts to kilo-Volts.

Interesting topic here, Adrian.

As you pointed out, most experimental studies have focused on VHB and silicone, and neither material is designed specifically for DEGs. Dissipative processes such as viscoelasticity and current leakage have been shown to affect the performance of these DEs and thus limit their application. Viscous losses reduce the useful mechanical input work while current leakage may result in a lower voltage boost across the generator. In particular, current leakage in dielectrics is a complex phenomenon; for example, the nature of the conduction mechanisms appears to be elusive in many cases.

Understanding the impact of these dissipative mechanisms on the performance of a DEG is a challenging issue.

Adrian S. J. Koh's picture

Yes Keith, I do agree with your observations.
 Dissipation mechanisms are very complex topics, especially for deformable
polymers.  Added to the complexity is the mechanisms are vastly varied
between polymers of different molecular structures, and also dependent on
various ambient factors like temperature, humidity and chemical composition.
 Its non-determinstic behavior is compounded by the fact that input
variables like mechanical stress, electric field, excitation frequency and
material properties may further modify their dissipation characteristics.

Having said that, we really seek like-minded researchers who
are interested in electromechanical dissipation.  Only through better
understanding of these effects, that we may design materials that minimize
them, thereby improving the conversion efficiency by a quantum leap.




Liu ZhuangJian's picture

Hi Adrain,

Thank you for posting this interesting thread.  The dielectric elastomer is great promise as actuator materials in converting mechanical to electrical energy. 

One quick question, is there any research work about the energy of conversion is strain rate or loading rate dependence in using DEG? As you said, Motion-based energy harvesting is the process of converting dissipated mechanical energy into electrical energy.  Sources of mechanical energy include the ocean waves, wind, human motion, vehicular traffic, and vibrations in buildings and bridges.  This source of energy is dynamic loading, the materials properties could be changed when the strain is higher.

Adrian S. J. Koh's picture

Dear Zhuangjian,

Let us begin by discussing the dissipative mechanisms in dielectric elastomers.

Dielectric elastomers are
electromechanically-coupled systems that dissipates energy in two major ways -
mechanically and electrically.  Subject to a mechanical force, the
deformation relaxes to a equilibrium state after some time tau_v.  This
process is known as viscoelastic relxation, and tau_v is the viscoelastic
relaxation time.  Subject to an electric field, the dipoles relaxes and
orientates towards the direction of the field after some time tau_e.  This
process is known as dielectric relaxation, and tau_e is the dielectric
relaxation time.  Furthermore, if the electric field is sustained long
enough or if the electric field is sufficiently high, charges may begin to leak
through the dielectric.  This process is known as current leakage, and the
product of the resistivity and capacitance of the DE gives some kind of
"RC" time constant, or the characteristic time where current leakage
builds up to a stable magnitude.

Some experiments on typical DE materials like VHB
acrylic elastomer and silicone elastomers have shown that tau_v is in the order
of 10^2 seconds, tau_e is in the 10^-6 seconds, and RC time is about 10^3
seconds.  This will determine the rate of operation whereby loss will be
significant or may be avoided.

Experiments specifically performed to determine
dissipative processes in DE are limited, yet essential to provide guidance for
optimal operation of DE actuators and generators.  I certainly hope more
work can be done in this aspect.





R. Palakodeti & M. R. Kessler,  Mater. Lett. 60, 3437-3440, 2006.  (On DEA viscoelasticity, efficiency with dependence on prestrain & frequency)

J. S. Plante & S. Dubowsky, Sens. Actuators A 137, 96-109, 2007.  (On major dissipative mechanisms - viscoelasticity & current leakage, experiments & modelling)

T. A. Gisby, S. Q. Xie, E. P. Calius & I. A. Anderson, EAPAD Proc. SPIE 7642, 764213, 2010.  (An experiment to measure leakage current with applied field)

X. Zhao, S. J. A. Koh & Z. Suo, Int. J. Appl. Mech. 3, 203-217, 2011.  (A theory of viscoelastic dissipation in DEs)



Liu ZhuangJian's picture

Hi Adrian,


Tang Lihua's picture

Hi, Adrian

interesting topic. I currently works on harvesting vibration energy via piezoelectric transduction. I first saw your idea in this year's SPIE NDE conference at San Deigo. seems the mechanism is similar to electrostatic generator.

Is there any commercialized elastomer transducer with flexible and durable electrodes avaliable in the market. or you just made the electrodes yourself in the labratory. I have interest to try this new material for energy harvesting purpose.


Adrian S. J. Koh's picture

Hi Lihua,

There are no commercialized electrodes for DE transducers per se, but there are commercialized DE actuators.  These actuators generally makes use of carbon grease as  compliant electrodes.  There are other candidates like metallic powder, and carbon nanotubes.

We really welcome newcomers to this field, and any interesting ideas should be actively shared between researchers.  Do feel free to share your thoughts on this topic, if you have any.




Cai Shengqiang's picture

Dear Adrian,

Can you comment on the merits of harvesting energy with soft materials in general?

Subscribe to Comments for "Journal Club Theme of August 2011: Energy Harvesting Using Soft Materials"

Recent comments

More comments


Subscribe to Syndicate