Micro Energy Harvesters - An Alternative Source of Renewable Energy

by Professor Khalil Najafi

Edwar Romero1, Tzeno Galchev2, Erkan Aktakka2, Niloufar Ghafouri2, Hanseup Kim2, Michael Neuman1, Khalil Najafi2,3 and Robert Warrington1,3
1 Michigan Technological University
2 University of Michigan Department of Electrical Engineering and Computer Science
3 Wireless Integrated Micro Systems Engineering Research Center (WIMS ERC)
Corresponding author: najafi@umich.edu

Topics Covered

Available Power
     Mechanical Motion Sources
     Solar and Thermal Sources
Transduction Techniques
     Electromagnetic Generators
     Piezoelectric Generators
     Multi-Mode Energy Generators


The increasing number of autonomous miniature electronic devices brings with it the problem of an adequate, reliable power supply. Micropower environmental energy harvesting generators offer an alternative source of renewable energy. These power supplies can help to support environmental, wearable or surgically implantable microsystems. They can assist or even replace batteries in some applications.

Harvesting energy from various environmental sources has been an area of research focus at the Wireless Integrated Micro Systems Engineering Research Center (WIMS). Micro energy harvesters based on piezoelectric, electromagnetic, thermoelectric techniques from both vibration and heat sources are being developed. Micro batteries are also being studied.

A review of research on energy conversion will be presented, including optimization of battery-powered systems for biological implants, energy scavenging from transpiration, a micro thermoelectric generator for microsystems, multimode energy scavenging from the environment, MEMS-based energy harvesting for low frequency vibrations, and MEMS-based mechanical energy scavenger for flying insects

Some ongoing challenges remain before these scavengers can be adopted on a commercial scale. These include: 1) Miniaturization of generators; 2) improving the available energy density, 3) increasing the efficiency of environmental energy coupling to micro harvesters, 4) developing high efficiency power rectification and energy storage, and 5) developing suitable device packaging for long-term reliability. Some recent results in these areas will be presented.


Remote-controlled microsystems have been limited by battery lifetime and battery size. Batteries are typically the dominant component in terms of size at the micro scale. Energy harvesting shows promise as an alternative for powering these devices. Energy generation from vibration or motion, solar light, and temperature changes has been established as a commercial viable alternative on human-powered flashlights, solar calculators, and thermal-powered wristwatches.

Micropower harvesters are targeted to applications where deployment of battery-based devices or when battery replacement is difficult, costly or impossible. Remote sensing locations, embedded structural monitoring, tracking of shipping containers, pacemakers, and humanimplants are among those applications where battery-operated devices have limitations.

Harvesting energy from environmental sources has been an area of research over the last decade, and a research focus at the Wireless Integrated Micro Systems Engineering Research Center (WIMS). WIMS is developing piezoelectric, electromagnetic, and thermoelectric techniques for harvesting energy from vibration and heat sources. Power management and micro batteries are also being studied. A review of research on energy conversion is presented, including recent WIMS projects.

Available Power

Mechanical Motion Sources

Mechanical motion is an energy source that has attracted wide attention for energy harvesting. This can be done either actively or passively. Passive generators use inertial mechanisms, such as proof masses attached to machines or even human bodies. These inertial generators use the proof mass displacement and a transduction mechanism for power generation.

A common design consists of a proof mass (m) attached to a moving host through a spring-like joint. The electrical generator typically damps the motion of the inertial mass. The power available for a linear displacement movement at resonance is

P max elect = (1/4)(a2/ω) mQ      (1)

The limiting factors are three, the ratio of acceleration-squared-to-frequency (ASTF) factor (a2/ω), the proof mass (m) and the quality factor (Q). The first one is an input source constraint, and the second and third one being a design constraint. From Table 1, the ASTF factor can be as low as 0.001, for machine vibrations, to values as high as 3, for human walking.

Table 1. ASTF factor1,2

Vibration Source

Acceleration (m/s2)

Frequency (Hz)

ASTF (a2/ω)

Car engine compartments




Base of 3-axis machine tool




Blender casing




Clothes dryer




Car instrument panel








Walking (head acceleration)




The maximum power delivered to an electrical load is half of what is available (P max elect = Pavailable/2)3. Rearranging for volumetric power density, where m=ρV,

(P max elect / V) = (1/4)(a2/ω) ρQ      (2)

Plotting (2) using Table 1 data, selecting Q-factors ranging from 1-1000, and assuming a proof mass density of 10 g/cm3 (for simplicity, and similar to molybdenum), gives the graph of Figure 1 which represents the maximum power that can be transferred to the electrical load.

Figure 1 helps to visualize untapped areas for energy harvesting. The use of human body movements, represented by high ASTF factors and low Q-factors, opens the possibility of human-based energy harvesting at levels comparable to those reached by machine-based devices (low ASTF factors and Q>100).

Figure 1. Maximum power available

Solar and Thermal Sources

Solar cells or photovoltaic (PV) energy generators can convert solar energy into electricity employing the photoelectric effect. Mono-crystalline silicon (mono-c-Si), polycrystalline silicon (poly-c-Si), and amorphous silicon (a-Si) are the dominant materials for PV generation. PV cells can produce up to 100 W/m2 (with 10% of efficiency and a light intensity of 1000 W/m2). Cells made of a-Si will produce less than that because of its lower efficiency (5 - 7%). Typical efficiencies of commercial cells are about 13-16% for mono-c-Si and 12-14% for poly-c-Si4. The efficiency of solar cells decreases logarithmically with the light intensity.

Thermoelectric generators (TEG) produce electricity based on the Seebeck effect. This is the generation of electricity due to temperature differences on two different metals forming a loop. Typical conversion efficiency for these systems is well below 10%. Power outputs up to 340 mW/cm2 for ΔT = 200°C at 4.5% efficiency, and modest values of 13µW/cm2 at ΔT = 1°C for the Citizen TEG wristwatch have been reported4.

Transduction Techniques

Electromagnetic Generators

Energy harvesting from electromagnetic transduction is based on the induced voltage on a coil by a moving magnet, or a fixed magnet and a moving coil. The amount of power generated depends on the strength of the magnetic field, the number of turns of the coil, and the change of the magnetic flux density through the coil due to the external input movement. A common scenario is a moving magnet attached to a beam or spring. The magnet by itself typically acts as the proof mass. The opposing magnetic field generated by currents in the coil will damp the magnet movement while supplying energy. Table 2 summarizes the findings of inertial electromagnetic energy harvesters.

A preliminary prototype of a MEMS-based energy harvester for low frequency vibrations at WIMS is composed of discrete NdFeB magnets on an oscillating mass with a gear-shaped multi-layer coil fabricated using photolithography5. The prototype has produced 2 µWrms of power at 2.5 Hz. A test for human-based energy generation generated 7.4 mVrms under no-load conditions when placed close to the knee while walking. Higher power outputs are expected for optimized prototypes.

Table 2. Electromagnetic energy harvesters5


Vol. (cm3)

Freq. (Hz)

Max Power (µW)

Power Density (µW/cm3)


























Ferro Solutions





Piezoelectric Generators

Energy generation from piezoelectric transduction is based on the generated voltage when a piezoelectric material is subject to a mechanical deformation. Piezoelectric generators are typically shaped as cantilever beams, membranes or other structures. An applied external or inertial force produces the deformation needed to generate energy. Table 3 summarizes the different approaches being studied for piezoelectric energy generation.

Preliminary work on a MEMS-based piezoelectric mechanical energy scavenger for flying insects at WIMS has been used for energy generation from flying beetles. Piezoelectric cantilever beams glued on the back of beetles start vibrating when they are hit by the wing strokes. Prototypes tested had provided up to 11.5 µW for an 11 mm3 device at 92 Hz. One hundred fifteen µW of power can be expected from beetle wing strokes6.

Table 3. Piezoelectric energy harvesters5


Vol. (cm3)

Freq. (Hz)

Max Power (µW)

Power Density (µW/cm3)






UC Berkeley





Nebraska U.





K.U. Leuven















Multi-Mode Energy Generators

Multi-mode energy scavenging from the environment is a WIMS project that seeks to develop a power generation unit that can scavenge energy from different sources including vibration, heat or solar energy. In addition to harvesting energy from those sources novel approaches on energy generation are being developed, such as transpiration-based, micro thermoelectric for flying insects, and frequency up-conversion.

A project on energy scavenging from transpiration is an approach employing evaporation at room temperature. Flow induced by evaporation on micro-fluidic channels drives gas bubbles through capacitor plates generating energy. A high power density is expected from this project7. A micro thermoelectric generator for microsystems is reported to scavenge energy from flying beetles. Power of 10-15 µW is expected to be generated when implanted on the back of a beetle, with a power density close to 200 µW/cm2 and a ΔT=11 °C8.

A frequency-up conversion scheme is intended to capture environmental low-frequency vibration (less than 100 Hz) to activate a high-resonant frequency structure (over 1 kHz). Development on mechanical frequency up-converters is expected to show a 23% increased efficiency, and an energy density increase from 14.5 mW/cm3 (low-frequency device) to 17.8 mW/cm3 (up-conversion device)9.


There are some limitations that have challenged energy harvesting at the microscale, such as the efficiency of the energy generation and energy density, DC-rectification, energy storage and management, manufacturing, longevity and packaging. An overview of these ongoing challenges is discussed in the following paragraphs in order to have a better understanding of them.


The efficiency is represented as the ratio of generated power to the available power. Although the power produced is presented, and the available power is defined by (1), not all the terms of Eq. (1) are reported to estimate the efficiency. Coupling coefficients up to 0.6-0.8 have been presented as possible limits for each of those transduction techniques3. In addition, projects under development at WIMS, such as the frequency up-conversion approach9, can help to improve the efficiency of the system.

Electromagnetic power harvesters over 1 mW and under 1 cm3 for machine vibrations around 100 Hz, and harvesters producing over 100 µW and under 1 cm3 for human-based activities can be expected in the near future. Power levels from piezoelectric energy harvesters close to 10 mW at 1 Hz and power densities of 1 mW/cm3 at 92 Hz were presented in Table 3. Piezoelectric power harvesters over 1 mW and under 1 cm3 for machine-based vibration around 100 Hz can be expected further on.


One of the limitations for electromagnetic transducers is the permanent magnet (PM) fabrication. MEMS-compatible processes do not yield PMs with the same characteristics as the bulk magnets. Typical fabrication techniques of sputtering and electroplating produce thin-film layers (<10µm) although thick-film (100-800µm) deposition at low temperatures has been studied10. Patterning of a PM has been demonstrated for mm-sized magnetic pole patterns, either using coils or soft-magnetic arrangements to induce magnetization10,11.

Another limitation for rotational electromagnetic generators is the need of low-friction bearings suitable for MEMS fabrication processes. Micro-ball bearings, rotating pivots, and magnetic bearings can be possible alternatives. Energy harvesters operated in vacuum have fewer losses associated with air damping, but create the need for special packaging to maintain the working environment (bio-implanted, structure-embedded, exposure to harsh environment).


Due to the nature of the energy generation, the output is a time-variant AC signal. Thus, DC rectification and voltage regulation are needed for most electronic applications. Circuitry should account for rectification, regulation, control, and storage of the energy produced. Most of the energy harvesters employed bridge rectification circuits. But the forward-bias of the diodes can still be high for the low-voltage output of some devices. In this case, voltage multipliers or transformers have been used to increase the voltage levels.

Active electronics can overcome some of the previous limitations, but a balance between their energy consumption and the energy produced should be taken into account. Optimization of power supplies for wireless integrated microsystems is also under study at WIMS. A micromachined battery for hybrid-power supplies that is compatible with MEMS fabrication processes was developed and the results have been used to design and optimize the power source for the WIMS implantable intraocular pressure sensor, and the WIMS cochlear implant12.


Energy harvesting is a growing research area that slowly has been evolving to become commercialized products, from hand-cranking radios and shake-driven flashlights to wireless monitoring applications.

Photovoltaic energy generation yields a high power output (10 mW/cm2) and it is a proven technology that can be implemented at the MEMS-Scale. Thermo electric generation is dependent on temperature gradients, tenths of µW/cm2 can be obtained from a modest ΔT=1 °C. Piezoelectric energy generation offers a simple approach for harvesting energy from motion or vibrations. The simplicity of these generators makes them well suited for MEMS fabrication and even nano applications. Energy density up to 1 mW/cm3 has been reported. Electromagnetic energy generation is a well-established transduction technique, but at MEMS-scale permanent magnets and printed-coils become less efficient. Although technology is evolving, they do not appear to be as simple to fabricate as piezoelectric generators. Commercial devices have shown a high power output (~10 mW), mm-sized devices have shown up to 3 mW, and smaller devices are on the order of tenths to hundreds of µW.

Multi-mode energy generation is an approach where power is produced from several environmental sources. It can take the best of the above transduction technologies according to the available energy sources. All the above transduction techniques prove that the technology is maturing at a rapid rate for powering portable, embedded, implantable or wireless devices. Although limitations on the technology still exist, the future looks promising for widespread applications.


1. S Roundy. On the effectiveness of vibration-based Energy Harvesting. Intelligent Mat. Systems and Struc., J, V. 16, No. 10. Oct. 2005, pp. 809-823.
2. E Hirasaki, ST Moore, T Raphan, and B Cohen. Effects of walking velocity on vertical head and boy movements during locomotion. Exp. brain research, 1999, 127(2). pp.117-30.
3. N G Stephen. On energy harvesting from ambient vibration. J. of sound and vibration, V. 293, No.1-2. May 2006, pp. 409-425.
4. S F J Flipsen. Alternative power sources for portables and wearables. Delft University of Technology, 2005, 90 p.
5. E Romero. MEMS-Based Energy Harvesting for Low-Frequency Vibrations. Unpublished doctoral dissertation, Michigan Technological University. 2009.
6. E E Aktakka, H Kim, M Atashbar, and K Najafi. Mechanical Energy Harvesting from flying insects. Solid State Sens., Act., and Microsys. Workshop, Jun 2008, pp 382-383.
7. R Borno, J Steinmeyer, and M Maharbiz. Energy Scavenging From Transpiration. (Project description available from http://www.wimserc.org). May 2008.
8. N Ghafouri, H Kim, and K Najafi. A Micro Thermoelectric Generator for Microsystems. (Project description available from http://www.wimserc.org). May 2008.
9. T Galchev, H Kim, M Atashbar, and K Najafi. Multi-Mode Energy Scavenging from the Environment. Unpublished manuscript, University of Michigan, 2008.
10. B Pawlowski, S Schwarzer, A Rahmig, and J Topfer. NdFeB thickfilms prepared by tape casting. J. of Magnetism and Magnetic Mat. V. 265, 2003, pp. 337-344
11. N Achotte, P A Gilles, O Cugat, J Delamare, P Gaud, C Dieppedale. Planar Brushless Magnetic Micromotors. J. Microelect. Sys. V. 15, N. 4, Aug. 2006, pp. 1001-1014.
12. F Albano and A M Sastry. Design and Optimization of Power Supplies for Wireless Integrated MicroSystems. (Project description available from http://www.wimserc.org). May 2008.

Presented at COMS 2008, Mexico

Copyright AZoNano.com, MANCEF.org

Date Added: Jun 8, 2010 | Updated: Jul 15, 2013
Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback