Low-frequency electric fields at smartphone surface

Both DC and AC electric fields are endemic to smartphone devices. Capacitive touchscreen technology^1–5 utilizes projected DC electric fields to detect the presence of a human finger, which has a higher electrical conductivity than air due to its water content, causing a change in the local capacitance. AC electric fields in the hundreds of MHz to GHz are associated with cellular uplink (emitted from the phone) and downlink (received by the phone) signals.^6–8 AC electric fields in the ultra-low frequency through the low-frequency range, which spans from 300 Hz to 300 kHz, however, are not expected from operational necessity.

The electromagnetic (EM) fields originating from mobile phones have garnered significant interest in the realm of public health. The impacts of non-ionizing radiation associated with these devices are well-studied in the uplink and downlink bands with the World Health Organization’s International Agency for Research on Cancer (IARC) classifying RF signals of all sources between 30 kHz and 300 GHz as possibly carcinogenic to humans.⁹ An earlier 2002 monograph by the IARC resulted in the same classification for magnetic fields of extremely low-frequency and an inability to classify electric fields in this range.¹⁰ Due to the ubiquity of sources of these frequencies, they continue to be of high interest. The possession of mobile devices increased substantially worldwide since 2002 with almost three-quarters of the world’s population over the age of ten owning a smartphone in 2022.¹¹ A review of recent work on the impacts of extremely low-frequency radiation considers contributions from mobile phones but with an absence of consideration for the charging condition of the phones.¹² 

Mobile phones are often in a charging or plugged-in state for long periods near the device’s user, including overnight, and in some cases while the phone is in active use. Study of the AC electric field present while devices are in the charging state is needed.

Consequently, we study the AC electric field in the direct vicinity of a smartphone screen in the 0–200 kHz frequency range. While virtually no signal is observed in this range during normal operation, we find significant AC fields in this range when a phone is charging on a wall charger. In addition to the frequency content, we further investigate the spatial distribution of the electrical signal in this frequency range under the different charging conditions.

Behavior associated with charging lithium-ion batteries is also of industrial interest due to their broad application. Non-invasive techniques that measure fields generated by the batteries under different charging conditions shed light on different aspects of charging,¹³ such as the charging state¹⁴ and current distribution.¹⁵ We focus on the fields generated during charging in the complete smartphone context and on larger length scales due to the ubiquity and routine handling of the devices. We present similar measurements of a portable lithium-ion power bank.

We study the AC electric field in the vicinity of an Apple iPhone 8 under a variety of charging conditions. The iPhone 8 was chosen due to its availability and the ease of disassembly, and one system is focused on to facilitate comparison across conditions. Similar results as those presented were observed in preliminary testing with other smartphones.

The chargers used are summarized in Table I. The AC electric field is measured first with an AlphaLab Trifield TF2 meter to establish the rms magnitude and spatial distribution. This instrument is sensitive in the frequency range from 40 Hz to 100 kHz.¹⁶ A homemade 100-turn Faraday pickup coil is then used to measure the frequency content of these electric fields under various charging conditions. All wall chargers (shown in Fig. 1) are attached to a standard US outlet (120 V, 60 Hz).

The low-frequency AC field study was conducted with the iPhone 8 using the STC-A515A-Z charger and initiated with a Trifield TF2 meter in standard response mode. This tool measures the rms value of the AC electric field over the range of its sensitivity with an approximately flat response between 40 Hz and 100 kHz. The maximum signal magnitude of the instrument is 1000 V/m.¹⁶ The measurement technique is depicted in Fig. 2.

Figure 3 shows the Trifield TF2 measurements over a 60 cm × 60 cm square area centered on the iPhone 8 and the STC-A515A-Z charger. These data were generated by moving the meter to each point on a grid with 5 cm spacing and recording the electric field magnitude value. The recorded value is the average of the magnitudes measured at the indicated point with the meter oriented in four cardinal directions. The red boxes indicate the phone location. These data show that an unplugged, but on, phone [Fig. 3(a)] emanates virtually no AC electric field in this frequency range. A phone charging via the USB wall charger [Fig. 3(b)] yields a maximum field of the order of 500 V/m. The field magnitude measurements clearly indicate a change between the charging and not-charging conditions. Further insight is gained by studying the frequency distribution of this electric field.

While direct measurement of E-field frequencies is challenging,¹⁷ the oscillations of the electric field in EM radiation are coupled to oscillations in the magnetic field, which will exhibit the same frequency content. As such, by using a Faraday pickup coil and measuring the frequency content of induced voltages from the changing magnetic flux, we can determine the E-field frequencies.

For this experiment, we employed a homemade 100-turn Faraday pickup coil. The induced voltage was measured with the setup shown in Fig. 4(a) with the phone placed inside of the indicated Faraday cage and connected to a Keysight MSO-X 3024A mixed signal oscilloscope. The oscilloscope data are taken at a sampling rate of 200 kilosamples per second [Fig. 4(b)], giving a frequency resolution of 200 Hz. We plot the fast Fourier transform (FFT) of the data from 0 to 200 kHz, noting that the oscilloscope is set to AC coupling mode, which will remove most of the signal around 0 Hz [Fig. 4(c)].

We studied six different conditions with the pickup coil measuring in the 0–200 kHz frequency range: (1) a control with no phone in the Faraday enclosure, (2) phone present but not charging, (3) a phone charging from a power bank, (4) a phone charging from a generic charger, (5) a phone charging from a USB-A charger, and (6) a phone charging with an Apple USB-C charger.

Using the control for a true background subtraction is not possible due to small changes in the positions of the study elements, resulting in expected changes in the strength of each field component. However, we are able to observe the addition of the new frequency content relative to the control condition under different charging conditions. This is illustrated in Fig. 5.

In the control and phone not-charging conditions, we see significant peaks in the FFT at 175 kHz [Fig. 5(a)]. Charging the phone from the 5 V power bank [Fig. 5(b)] results in a prominent new peak around 90 kHz. In addition, note that, while charging on a 5 V power bank, no field was measured with the Trifield TF2 meter, as noted in Fig. 5(b). When various wall chargers are used to charge the phone, a wide range of frequency content is introduced [Figs. 5(c) and 5(d)].

Here, it is important to emphasize that under the conditions represented in Figs. 5(a) and 5(b), there is no accompanying AC rms electric field measured by the Trifield TF2 meter as also noted in Fig. 3(a). However, under the phone charging conditions of Fig. 5(d), there is considerable measured AC rms electric field of the order of ∼500 V/m as shown previously in Fig. 3(b). In Fig. 5(c), two conditions--no phone and not charging--have no AC rms electric field, whereas the phone charging on a USB-A charger does have this AC rms electric field.

In order to interpret the AC rms electric field results and, in particular, to investigate the origin of the various frequency components, we conducted oscilloscope measurements of the voltage output directly from the USB wall charger and also a standard power bank. We do this because the frequency content at issue did not appear under the non-charging condition. Here, the DC output signal for most of the chargers shown in Table I was tested, including the power bank. These data are provided in Fig. 6. While the USB chargers provide a nominal 5 V DC signal, high frequency components are readily observable as compared to the lower harmonic content of the power bank signal. No clear correlation between higher magnitude frequencies in the charging tests and the magnitude of the frequency content of the power signals is evident.

In order to further investigate the origin of these electric fields, we used the same pickup coil on an opened Apple iPhone 8 as shown in Fig. 7(a). With the phone charging, the pickup coil was placed on the lithium-ion battery and then on the capacitive touchscreen. The results in Fig. 7(b) indicate that the magnitude of the frequency spectra is generally larger on the lithium-ion battery side of the phone. This general magnitude comparison is appropriate in that the coil was placed in contact with and at the approximate center of both the battery and the touchscreen.

After seeing that the battery-side signal was apparently higher than the screen-side signal, the pickup coil was used to measure the power bank shown in Fig. 1. In Fig. 8, the Fourier transform for the power bank and a closed phone is plotted with those of the open phone case. In Fig. 8(a), the power bank frequency content (black) seems to be very similar to that of the lithium-ion battery side (red) of the open phone, and in Fig. 8(b), the capacitive touchscreen side of the open phone (green) behaves similarly to that of the screen-side of a closed phone. This initial study seems to indicate that most of the electric field is coming from the battery-side of the phone. The screen-side may exhibit an AC electric field due to inductive or capacitive coupling to the battery field due to the close proximity to each other in a closed phone.

We observe significant low-frequency electric fields associated with phone charging; however, no such fields are measured with a phone charging from a 5 V power bank or with an unplugged phone. Charging from different USB wall adapters resulted in broad contributions to the AC signal throughout the tested range. Measurements of the charger voltage output did not match this frequency content, suggesting generation from the phone/charger circuit. Subsequent testing further suggests that the field with its associated frequency content is being generated during charging of the lithium-ion battery within the phone.

Given that these electric field frequencies are absent while the mobile phone is charging from a power bank, we conclude that they are not a necessity of charging. This points to the opportunity for future studies directly investigating any implications for lithium-ion battery health or charging algorithms.

This study has no direct bearing on claims regarding the health impacts of electric or magnetic fields. We note that the emission of these fields from mobile devices in the charging state argues for consideration of this contribution when conducting public health studies.

The authors acknowledge funding provided by Wentworth Institute of Technology via their internal Launch Grant program. The authors also acknowledge Joe Diecidue, Doreen Cialdea, Afsaneh Ghanavati, and Kai Ren for their help with either useful discussions or help with the initial experimental setup.

J.V. and A.K. have Patent 17/803707 pending.

John Voccio: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal). Andrew Seredinski: Conceptualization (equal); Formal analysis (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal). Jiahui Song: Conceptualization (equal); Methodology (supporting); Writing – review & editing (supporting). Ali Khabari: Conceptualization (supporting); Writing – review & editing (supporting). Marina Chuery: Investigation (equal); Writing – original draft (supporting); Writing – review & editing (supporting). Hunter Oshman: Investigation (equal); Writing – original draft (supporting); Writing – review & editing (supporting). Patricia Sadde Mujica: Investigation (equal); Writing – review & editing (supporting).

The data that support the findings of this study are available from the corresponding author upon reasonable request.