The Amazing Electret

Near-Permanent Charge Without a Voltage Supply

An electret is an electrical insulator with a near-permanent state of electric polarization. In this article, I look at some common electrets and how they are made and charged. Then I’ll show how I built an electret loudspeaker, and a triboelectric nanogenerator energy harvester that could light an LED array and power other devices.

This project started innocently enough. A friend asked why there are lots of electret microphones available, but not electret loudspeakers. This seemed like a simple question. But upon further reflection, it was not easily answered. Surely there must be electret speakers around somewhere. After all, electret technology has been around for many years. A quick search on the Web produced virtually no electret speakers! However, there were Stax electret earphones, such as the SR40. These electret earphones are quite expensive, running into the thousands of dollars, which indicated that they were probably intended for the most experienced audiophiles, persons who are interested in high fidelity sound reproduction. 

A close cousin to the electret speaker is the electrostatic loudspeaker (ESL). Typically, it uses a conductive, ultra-thin polymer film or diaphragm, suspended with a small air gap between two conductive grids. A charge is placed on the diaphragm with a high voltage DC source that must be maintained. The audio signal is placed on the grids via a step-up transformer, causing the diaphragm to vibrate via electrostatic force. Typical of these speakers is the Martin Logan ESL, ranging from $1,695 to over $7,500. ESL speakers have their own niche outside of the conventional electromagnetic speaker world, presumably too exotic for the average audio system. 

Searching for electret speakers, I discovered a burgeoning world of electret technology. My curiosity had now peaked about this subject. What was the technology behind electrets and what made them so unusual? These questions began a long path that involved learning more about electret materials, their applications, examining current technology, and reading many research papers. I found several papers describing new approaches to making electret loudspeakers. And I found a virtual avalanche of papers in the new field of TENG energy harvesters. This was getting very interesting. I finally fell down the rabbit hole, so to speak, when I learned that I could make my own electrets using PTFE material. This would allow me to experiment with an electret speaker. And it would also allow me to enter the fascinating world of microwatt energy generators powered by human motion. 

In what follows, I’ll take you through this interesting journey where these subjects will be covered in more detail. Let’s begin with an overview of electrets.

ELECTRET BACKGROUND

An electret is basically a piece of dielectric material that has a quasi-permanent electric charge. The term “quasi-permanent” means that the time periods for the decay of the charge are much longer than the time periods over which work will be performed with the electret. An electret generates internal and external electric fields, and is the electrostatic equivalent of a permanent magnet. 

History records that Michael Faraday was the first to consider the possibility of electrets in his 1839 book, Experimental Researches in Electricity. Oliver Heaviside was the next to mention electrets in 1892. Although he never made an electret, Heaviside gave a fairly complete discussion on the subject, and coined the word “electret” by combining the words “electron” and “magnet.” He defined electret as “an electrified dielectric having opposite charges on two faces,” analogous to a magnet. 

The first systematic research with electrets began in 1919, when a Japanese physicist, Mototaro Eguchi, fabricated an electret. He did this by cooling and solidifying a molten mixture of carnauba wax and resin with a little beeswax, while applying an electric field of about 10kV/cm to the mixture. This dielectric material contains polar molecules, which align themselves to the direction of the electric field, producing a permanent electrostatic “bias.” When the melt cools and solidifies, the permanent dipole moments remain locked in place and retain their orientation for a long time, ranging from several days to many years. This process is discussed in more detail by C. L. Stong [1].

In general, the electret charge may consist of “real” charges, such as surface-charge layers, space charges, or a “true” polarization; or it may be a combination of these, as shown in Figure 1. True polarization is usually a frozen-in alignment of dipoles, whereas the real charges comprise layers of trapped positive and negative carriers, often positioned at or near the two surfaces of the dielectric, respectively. Electret charges may also consist of carriers displaced within molecular or domain structures throughout the solid, resembling a true dipole polarization.

If the electret is metalized, a compensation charge may reside on the electrode, unable to cross the energy barrier between metal and dielectric. Electrets not covered by metal electrodes produce an external electrostatic field, if their polarization and real charges do not compensate each other everywhere in the dielectric. Such an electret is, in a sense, the electrostatic analog of a permanent magnet,

Classical electrets were made of thick plates of carnauba wax or similar substances. However, present electret research frequently deals with thin-film polymers, such as the Teflon materials polyfluoroethylene propylene (FEP), PTFE, or polyvinylidene fluoride (PVDF).

Typical electrets presently in use are films 0.1µm to over 1,000µm thick, such as PTFE. Electret charging is limited by internal and external breakdown. It occurs whenever the charges produce sufficiently high fields. Internal breakdown depends on the dielectric strength of the material. For PTFE, charge densities of up to 0.5µC/cm² can be stored without the danger of breakdown. Typical electret microphones made of PTFE-FEP films of 0.1µm thickness have a stored charge density on the order of 0.02µC/cm² and an air gap of about 15µm. Electret energy harvesters have larger charge densities.

A key issue in studying electret materials is to know the charge distribution in an electret and how it influences the electric field around it. A general theoretical analysis based on Gauss’s law and Kirchhoff’s law was given by G. M. Sessler [2]. It was based on a flat piece of electret sheet with dielectric constant ε, thickness d₁, a gap d₂, and vacuum permittivity ε₀, as shown in Figure 2. 

His analysis showed two major inferences. First, any charge distribution in an electret can be regarded as an “effective” surface charge, when investigating the induced external electric field around it. Assuming an arbitrary volume charge density ρ(x), then the surface charge density σ of the electret is given by: 

where d1 is the thickness of the dielectric and ρ is the volume charge density. 

The second inference is that the electric field in the electret (E1) and in the gap (E2) can be calculated using the surface charge density σ as follows: 

These equations can be helpful for calculating electric field strengths from given charge distributions. They were obtained assuming that the lateral dimensions of the electret and electrode are the same, and much larger than the thickness.

ELECTRET APPLICATIONS

Quasi-permanent electrification effects in electrets have been utilized in a wide variety of applications. These devices range from transducers and sensors to an array of different applications, including novel devices in the medical and biological areas.

The earliest electret transducers were electret microphones, first developed in Japan in 1928. But these devices were made of wax-based materials, which were poor in thermal stability. An important breakthrough came in the 1940s with the synthesis of insulating polymers such as PTFE. Their high specific resistance made them natural candidates for electrets.

A turning point for this technology was the invention of the foil electret condenser microphone by Sessler and West in 1962, while working at Bell Laboratories. Here sound waves cause a charged fluoropolymer membrane to vibrate in front of a static metal back electrode (Figure 3). Unlike earlier condenser microphones, this design needs no external bias voltage. Together with its high-quality linear performance, this property soon turned it into a microphone of choice in numerous applications.

A high percent of today’s microphones are electret microphones, being used in everyday items such as telephones, sound and music recording equipment, and hearing aids. Present annual production is estimated to be around two billion units.

Other applications of electrets include air filters for electrostatic collection of dust particles, ion chambers for measuring ionizing radiation or radon, and vibration energy harvesting.

CHARGING AND MEASURING ELECTRETS 

There are many different schemes for charging electrets, ranging from thermal charging to radiation charging. Only two methods will be discussed here: tribocharging and corona charging of thin PTFE sheets. 

Tribocharging of electrets is based on triboelectrification (TE), the charging of two dissimilar objects due to physical contact with each other. Everyone has seen examples of triboelectricity. Children generate triboelectricity by dragging their shoes on a carpet to build up a charge that allows them to shock a friend.

This method of generating electricity is simply the conversion of mechanical energy to electrical energy through friction. This motion causes electrons to be transferred from one material to another, resulting in an excess of electrons on one material and a deficiency of electrons on the other. While this form of electricity has been known for centuries, it is still the least understood. 

It is possible to place a semi-permanent charge on the surface of a thin and flexible sheet of PTFE by tribocharging. When the sheet is rubbed with ordinary newspaper, opposite-polarity charges σP and σT accumulate on the surfaces of the paper and PTFE, due to the triboelectric effect (Figure 4).

Because PTFE has a lower electron affinity than paper, charge σT is negative [3]. This charge on the PTFE sheet can then act as an electric field source. This is a proven method of creating an electret. 

Unfortunately, soon after charging, the surface potential begins to decay, until a sort of plateau is reached, after which its decay is much slower. Its permanence depends on many factors—material, temperature, humidity, and how long and strenuously it was rubbed. Its charge can last a few hours, days or years. My tribocharging of PTFE was found to last several days and in some cases months. Typically, initial surface potential was measured at 1.4kV, decaying to 1.0kV after a few hours. 

Corona charging of polymer electrets is done by depositing charge on the surface of the electret material. A corona discharge is a self-sustainable, non-disruptive electrical discharge that occurs when a sufficiently high voltage is applied between asymmetric electrodes, such as a point and a plate. Corona charging of polymers in air is a complicated process that is not completely understood [4], but it has been used successfully in many situations. 

The ionization section of the corona discharge is confined to a small region near the point where ions and excited molecules are produced, initiating movement of the air, owing to numerous collisions between the charged particles and neutral molecules. A drift region extends from the corona point to the plate, and is characterized by the presence of charge carriers of only one polarity. Depending on the corona polarity, either positive or negative ions are produced.

My corona charger uses a single point source, as shown in Figure 5. It was built from parts that were on hand. My unit operates at 7.5kV, with an air gap of about 1cm. Charging time varies from 30 seconds to 90 seconds, depending on the thickness and material type. 

Researchers have operated corona chargers from 5kV to 20kV. More sophisticated chargers use a triode element configuration and thermal heating of the polymer to increase charge permanence. Working with high voltage is dangerous, so don’t attempt to build a charger without expert advice. 

Both tribocharging and corona charging can achieve the same initial surface potential [5]. However, researchers have found that corona charged polymers decay slower, and that negatively charged polymers have greater permanence than their positive counterparts.

Teflon polymers such as PTFE and FEP can be found for sale on the Internet. Grainger and Zoro offer small quantities in various sizes and thicknesses for the experimenter. 

Electrostatic field meters are used to measure the surface potential of the electrets. A Monroe Electronics Model 256 and a Prostat PFM-711A were used here. The Monroe has a resolution of 10V, and the Prostat can measure in 1V increments. They are both chopper designs, and can be used in areas where ionized air is present.

As examples, electrostatic field measurements were made on two unusual commercial electret products: the E-PERM from Rad Elec Inc., and ClingZ from Nekoosa, which is electrically charged, printable graphic film.

E-PERM has a long-term (LT) electret ion chamber, intended for radon measurements lasting up to a year. The electret is a 1⅜” film disk housed in a 3” diameter conductive holder. It is part of a special chamber which, when open, allows ionization from radon to occur, causing positive and negative ions to form. Negative ions are attracted to the positively charged electret, reducing its voltage. Electrets are shipped with 700V, to ensure various deployments.

I found several “used” LT units on eBay. Measurements indicated +620V on unit #1 and +650V on unit #2. As an experiment, #1 was corona charged to +960V, while #2 was left alone for comparison. Measurements of the closed chambers over the next five weeks showed no voltage decline on either unit. Running the experiment again with one unit open to air might find radon radiation, if present. But time did not permit this test. 

ClingZ is described as “an adhesive-free, electrically graphic film that adheres to any dry interior surface, and is easy to position.” Nekoosa uses a patented process to charge the film during the manufacturing process. Measurements made on a clear sample found voltages from 300V to 1,000V, depending on the location on the film.

Some other common items were tribocharged, including a white plastic lid from a Cool Whip dessert topping container. It was easily charged to 1,500V. After 60 minutes, the voltage dropped to 600V, and the next day it had declined to only 90V. Another item was “Teflon Sheets for Hot Press” purchased on eBay. One sheet was tribocharged to around -3kV. After an hour, the voltage had declined to -900V, and a few hours later it was down to -50V. Clearly, in both of these cases, the initial free surface charge was responsible for the high initial voltage. 

These examples serve to illustrate the great variability of polymer electrets and their applications. The big question now is: can an electret loudspeaker be made with them?

AN ELECTRET LOUDSPEAKER

An electret speaker appears to be feasible, according to work by Yu-Chi Chen [6] on “making a thin, flexible, electret-based loudspeaker for automotive applications.” As a proof of concept, I decided to try to make my own loudspeaker.

It consisted of a stack of thin, 2⅛” x 3” copper PCBs, with a thin film of PTFE in the middle (Figure 6). The film chosen was Grainger 0.003” thick PTFE, which hopefully could be vibrated by the electrostatic forces involved.

The film needs to be metalized on one side. But not having the professional equipment needed, an alternative was used. The PTFE film was spread out and stretched on a table. Two light coatings of MG Chemicals (#841-AR) nickel conductive spray were applied to the film. After drying, the film was tensioned and attached (physically and electrically) to a copper PCB back plate with tape and a tiny amount of glue. The back plate has a large rectangular hole, allowing the film to vibrate. 

An insulated 1/16” spacer with a rectangular hole is placed above the film. An electrode made out of copper PCB forms the top. Four nylon screws in the corners hold the stack together. Wires are connected to the top and bottom copper plates for the audio signal. 

With top cover and spacer removed, the PTFE was corona charged to a surface potential of 1,400V. Electrostatic and electret loudspeakers require a boost in the audio signal, using step-up transformers. Since audio transformers were not at hand, an ordinary power transformer was used with a step-up of one to twenty.

An audio signal generator was connected to the transformer input. Amazingly, a sound was heard! It was not loud, but sufficient to be heard a distance away. While varying the frequency several things were noticed. High frequency tones were clear. But there were resonances in the lower frequency range, perhaps because of the construction. Below 160Hz, the sound was softer, probably due to the small size of the unit. But it worked! The completed unit is shown in Figure 7. 

To check voice and music, an AM/FM receiver was hastily connected in place of the generator. A baseball game was on the air. Remarkably, the announcer’s voice came through distinctly and could be heard across the room. But music was another story. Poor, low-frequency response reduced the bass sounds. It reminded me of a cheap cell phone. Over the next few months it was checked periodically and seemed to hold its sound level. The bottom line is that electret speakers are feasible, but economics are probably holding them back.

NANOGENERATOR ENERGY HARVESTERS

Harvesting energy from daily activities, such as walking, and using it to power mobile electronic devices, IoT devices, and health sensors is the dream of researchers working in the field of nanogenerators, such as piezoelectric, triboelectric, and thermoelectric nanogenerators. Compared with traditional energy harvesters, these nanogenerators can be made small and operate almost anywhere, regardless of weather conditions. It is hoped that they can provide supplementary energy to extend battery life or as independent power sources. 

The TENG, invented in 2012 by Zhong Lin Wang and others [7], is the most recent of this group. According to Wang, the TENG can be applied to harvest all kinds of mechanical energy, ranging from human motion, vibration, rotational energy, wind, ocean waves and more. Compared to other nanogenerators, TENGs have many nice properties, such as high power density and efficient low-frequency operation. 

Nevertheless, TENGs appear to be the most difficult to use directly, due to the high voltage pulsed output and high internal capacitance. Therefore, a PMS must be used to convert the raw harvested energy to a regulated form suitable for electronic devices. 

Considerable effort has been devoted to the development of a PMS since the introduction of the TENG. However, due to the TENG’s nonlinear electrical properties and capacitive behavior, the maximum energy that can be extracted by any circuit remains unknown. Thus, research is still needed to increase the energy generation of the TENG and the power output of the PMS, by looking at more effective circuit topologies. 

A parallel plate, contact and separation mode TENG will be studied here, as shown in Figure 8. As established by Niu and Wang [8], a TENG can be modeled as a series connection of a voltage source and a time-varying capacitor, as shown below. 

Here x(t) is the separation distance, S is the area of the metal electrodes, ε₀ is the permittivity of free space, σ is the tribocharge surface density, and d₀ is the effective dielectric thickness. 

The capacitance C_T takes the maximum value when the two plates are in contact and the minimum value at maximum separation. Some researchers have used this fact to design switching strategies for a PMS. Note that as the TENG is pressed and parted, the separation distance becomes a function of time, which makes it difficult to obtain analytical solutions to the above equations, except in special cases. 

A TENG ENERGY HARVESTER

A simple TENG was constructed from a large piece of copper PCB as a base electrode, a layer of charged, white, PTFE electret, and a 7cm x 10cm PCB on top (Figure 9a). When the top PCB is moved rapidly up and down on the electret, a voltage (1,000V peak to peak) is generated (Figure 9b). A 100MΩ probe was used to obtain the traces.

To use the TENG’s alternating pulsed high voltage, it must be rectified. Junwen Zhong and others [9] used a 3.5cm x 2.5cm TENG to demonstrate driving 50 blue LEDs in series, directly from a full-wave bridge (Figure 10). Their LED array had a conduction point of about 120V. By finger tapping the TENG, they were able to generate around 10µA peak current, which caused the LEDs to flash. 

One form of energy harvesting employs a store and release method, using a full-wave bridge rectifier and a large filter capacitor on the output of the bridge. While this works, it is inefficient. Nevertheless, using this arrangement in 2021, Kawaguchi and others [10] reported using a 10cm x 10cm TENG, tapping for three minutes at 4Hz, and storing the charge in a 100µF capacitor. They used a programmable unijunction transistor (PUT) to release the stored energy to run an LCD timer, which then ran for 20 seconds. If one has a source of reliable vibrations around 4-5Hz, this becomes a viable scheme.

These experiments were duplicated using my own TENG, described above. In the first case, the blue LEDs flashed nicely but were not as bright as hoped. However, in a darkened room they could easily be seen. Kawaguchi’s method also worked. After tapping for about a minute to raise the voltage on the storage capacitor, a 2N6027 PUT was triggered when the capacitor voltage reached 3V; this ran a kitchen timer for tens of seconds. So, my TENG was producing energy, albeit a small amount. Could there be a more efficient way to harvest energy from the TENG? 

An extensive analysis of TENGs was done by Niu and Wang to determine optimum matching of the device when used with a resistive load or to charge a capacitor with a bridge rectifier over multiple cycles. These parameters depend on variables such as the type and size of the TENG and the number of cycles per second.

William Harmon and others [11] confirmed this work using an LTspice simulation, and designed a self-driven PMS for TENGs, based on a buck converter. The design obviates the need for passive mechanical switches of other designs. These efforts show that, due to the capacitive internal impedance of TENGs, conventional PMS strategies are not able to extract a significant amount of energy. But by the maximum power transfer principle, a capacitive load should be used to draw maximum energy.

The TENG bridge circuit with the PMS is shown in Figure 11. High-speed switching diodes are used. Capacitor C1 receives the rectified TENG’s pulses. C1’s value is chosen to optimize the energy transfer from the TENG, around 680pF for my TENG. To capture the maximum charge, the SCR (2N5064) is not triggered until the voltage on C1 reaches close to the peak voltage. Harmon’s PMS used a 491V Zener diode to trigger the SCR, which then dumps the energy of C1 into C2 via the buck converter. My circuit uses a neon lamp (90V) in series with a 180V Zener to trigger the SCR. L1 is 6.8mH and C2 is chosen to be in the range of 0.1µF to 100µF, depending on the load to be driven.

Measurements indicated that this setup performed much better than the earlier version. The charge build-up (voltage steps) on the output capacitor C2 could easily be seen as the pulses were processed by the PMS. It was possible to generate up to 300µW into a resistive load at a frequency of 6Hz. To explore other possible uses, several different loads were tested. These included driving an LCD shutter, an E-paper display (Ynvisible Interactive Inc.), an LED array, an LCD timer and “smart film.” Photos of some of the devices driven by my TENG are shown in Figures 12-14.

A series of 32 blue LEDs is shown in Figure 12. By tapping, the LEDs are easily seen in room lighting. Their brightness was better than the earlier version without the PMS.

An operating LCD kitchen timer is shown in Figure 13. A 10µF capacitor was used for C2. After a few taps to get the voltage to 1.2V, the timer began to operate. Only 1-2Hz was needed to keep it working indefinitely.

Smart film is a liquid crystal film that uses polymer-dispersed liquid crystal (PDLC) as a display structure. It is fascinating to watch. When voltage is applied to the film, the liquid crystals align, and the film instantly becomes clear. When the power is turned off, the liquid crystals return to their normal scattered positions, rendering the film opaque.

There are several types of smart film with different power requirements. An 8cm x 10cm sample of low-power HoHo smart film (Shanghai HoHo Industry Co., Ltd) was used here. Only a few taps were needed to reach 26V, clearing it as shown in Figure 14.

FINAL THOUGHTS ON ELECTRETS 

Hopefully, you enjoyed reading about these amazing electrets and their applications. Making my own electret loudspeaker and TENG was demanding but exciting.

There are still challenges in developing TENGs. It would be wonderful to see how one could be used to power a microprocessor, charge a battery, or generate a Bluetooth signal. We are fortunate to live in a time when fascinating subjects like this can be studied. 

REFERENCES
[1] C. L. Stong, “How to make electrets.” Scientific American, Vol. 219, No. 1 (July 1968), pp. 122-132.
http://www.tuks.nl/pdf/Reference_Material/Electrets/Strong%20-%20How%20to%20make%20an%20Electret%20-%201960.pdf
[2] G.M. Sessler, Electrets, Springer-Verlag Berlin Heidelberg, 1987.
[3] Bill W. Lee and David E. Orr, “The Triboelectric Series.”
https://www.alphalabinc.com/triboelectric-series/
[4] Jose A Giacometti, “Corona charging of polymers.” Brazilian Journal of Physics, Vol. 29, No. 2, June, 1999
https://www.scielo.br/j/bjp/a/bFKy9Q7MVPf8fRRwX7CqLCN/?lang=en&format=pdf
[5] H. Mellouki, “Tribo and corona charging and charge decay on polymers plates.”
The 5th International Conference on Electrical Engineering – Boumerdes (ICEE-B)
October 29-31, 2017, Boumerdes, Algeria
https://www.researchgate.net/publication/321816804_Tribo_and_corona_charging_and_charge_decay_on_polymers_plates
[6] Yu-Chi Chen, “A thin light flexible electromechanically actuated electret-based loudspeaker for automotive applications.” IEEE Transactions On Industrial Electronics, Vol. 59, No. 11, November 2012.
[7] F.-R. Fan, Z.-Q. Tian, and Z. Lin Wang, “Flexible triboelectric generator,” Nano Energy, Vol. 1, No. 2, March, 2012, pp. 328-334.
[8] Simiao Niu and Zhong Lin Wang, “Theoretical systems of triboelectric nanogenerators.” Nano Energy, Vol. 14, May, 2015.
https://www.sciencedirect.com/science/article/pii/S2211285514002353
[9] Junwen Zhong, QizeZhong, FengruFan, YanZhang, Sihong Wang, BinHu, Zhong Lin Wang, “Finger typing driven triboelectric nanogenerator and its use for instantaneously lighting up LEDs.” Nano Energy, Vol. 2, No. 4, July, 2013.
[10] Atsushi Kawaguchi, Haruki Uchiyama, Masahiro Matsunaga and Yutaka Ohno. “Simple and highly efficient intermittent operation circuit for triboelectric nanogenerator toward wearable electronic applications.” Appl. Phys. Express Vol. 14, No. 5, April, 2021.
[11] William Harmon, David Bamboje, Hengyu Guo, Zhong Wang. “Self driven power management system for triboelectric nanogenerators.” Nano Energy, Vol. 71, 2020.

PUBLISHED IN CIRCUIT CELLAR MAGAZINE • DECEMBER 2022 #389 – Get a PDF of the issue

Sponsor this Article