0% found this document useful (0 votes)
3 views

Design_and_Characterisation_of_a_Non-Contact_Flexible_Sensor_Array_for_Electric_Potential_Imaging_Applications

This document presents the design and characterization of a novel flexible sensor array for non-contact electric potential imaging, achieving spatial resolution down to 20 µm. The array consists of five flexible probes that enhance scanning speed and reduce mechanical damage compared to traditional rigid probes. Applications demonstrated include imaging latent fingerprints, surface topography of conductive samples, and visualizing dielectric variations.

Uploaded by

Varun kulkarni
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
3 views

Design_and_Characterisation_of_a_Non-Contact_Flexible_Sensor_Array_for_Electric_Potential_Imaging_Applications

This document presents the design and characterization of a novel flexible sensor array for non-contact electric potential imaging, achieving spatial resolution down to 20 µm. The array consists of five flexible probes that enhance scanning speed and reduce mechanical damage compared to traditional rigid probes. Applications demonstrated include imaging latent fingerprints, surface topography of conductive samples, and visualizing dielectric variations.

Uploaded by

Varun kulkarni
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

26328 IEEE SENSORS JOURNAL, VOL. 21, NO.

23, DECEMBER 1, 2021

Design and Characterisation of a Non-Contact


Flexible Sensor Array for Electric
Potential Imaging Applications
Arash Pouryazdan , Member, IEEE, Júlio C. Costa , Graduate Student Member, IEEE,
Leonardo Garcia-Garcia, Member, IEEE, Pasindu Lugoda, Member, IEEE, Robert J. Prance,
Helen Prance, and Niko Münzenrieder , Senior Member, IEEE

Abstract —Capacitive non-contact imaging of electric fields


and potentials with micro-metre resolution can provide
relevant insights into material characterisation, structural
analysis, electrostatic charge imaging and bio-sensing appli-
cations. However, scanning electric potential microscopes
have been confined to rigid and single-probe devices, making
them slow, prone to mechanical damage and complex to
fabricate. In this work, we present the design and characteri-
sation of a novel 5-element flexible array of electric potential
probes with spatial resolution down to 20 µm to speed up
the scanning time. This was achieved by combining flexible
thin-film probes for active guarding and shielding with state-
of-the art discrete conditioning circuits. The potential of this
approach is showcased by using the fabricated array to image latent fingerprints deposited on an insulating surface by
contact electrification, obtain the surface topography of conductive samples and to visualise local dielectric variations.
Index Terms — Flexible electronics, capacitive sensors, electric potential imaging, sensor array.

I. I NTRODUCTION niques. Contact measurements require resistive connection


to the sample under test. Common instruments to measure
N ON-CONTACT measurements of static and low-
frequency dynamic electric potential and electric fields
have many applications in science [2] and engineer-
AC and DC potential differences are voltmeters, oscilloscopes
and amplifiers. In contrast, there are various capacitive sensors
ing, including non-invasive material characterisation [3], available to measure electric potentials. A key advantage of
electro-physiology [4] and context activity recognition [5], [6]. capacitive sensing is that measurements can be made without
Measurements of electric potential (voltage) can be divided any physical contact. To capacitively measure electric poten-
into two broad categories of contact and non-contact tech- tials and electric fields, voltage amplifiers with virtually infi-
nite input impedance are required. This avoids any disturbance
Manuscript received December 1, 2020; revised February 10, 2021; to the field under test and results in a sufficient signal to noise
accepted February 11, 2021. Date of publication March 8, 2021; date
of current version November 30, 2021. This work was supported by ratio [7].
Engineering and Physical Sciences Research Council (EPSRC), in part When measuring electric potential with micro-metre spatial
by Global Challenges Research Fund (GCRF), and in part by National resolution, the required input impedance is beyond the capa-
Institute for Health Research (NIHR) under Contract EP/R013837/1.
This article was presented at the 2020 IEEE FLEPS Conference titled bilities of conventional measurement systems. Distortion-free
Flexible Array of Sensing Probes for Non-Contact SEPM Applications. measurements require e.g. highly sophisticated MEMS based
The associate editor coordinating the review of this article and approv- technologies [8]. Hence, most methods to measure elec-
ing it for publication was Prof. Arokia Nathan. (Corresponding author:
Arash Pouryazdan.) tric potential with micro-scale or better resolution require
Arash Pouryazdan, Júlio C. Costa, Leonardo Garcia-Garcia, Pasindu some sort of physical contact to the sample e.g. contact
Lugoda, Robert J. Prance, and Helen Prance are with the Flexible Elec- Atomic Force Microscopy (AFM) [9] or Scanning Capacitance
tronics Laboratory, Sensor Technology Research Centre, University of
Sussex, Brighton BN1 9QT, U.K. (e-mail: [email protected]). Microscopy (SCM) [10]. However, one approach to overcome
Niko Münzenrieder is with the Flexible Electronics Laboratory, Sensor this limitation is the use of the capacitive Electric Potential
Technology Research Centre, University of Sussex Falmer, Brighton Sensor (EPS) [7]. The ultra-high input resistance of EPS
BN19QT, U.K., and also with the Faculty of Science and Technology,
Free University of Bozen-Bolzano, 39100 Bozen, Italy. (≈0.1 T) along with the ultra-low input capacitance
Digital Object Identifier 10.1109/JSEN.2021.3064276 (≈0.3 fF) enables the measurements of AC potentials with

1558-1748 © 2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
POURYAZDAN et al.: DESIGN AND CHARACTERISATION OF NON-CONTACT FLEXIBLE SENSOR ARRAY 26329

Fig. 2. Arrays of flexible sensing probes fabricated for electric potential


microscopy. (a) Micro-graph of the fabricated polyimide substrate with
5-element SEPM probes and the layer schematic diagram of the probes.
Ti/Au tracks are deposited on a polyimide substrate and subsequently
stacked on top of each other. (b) Photograph of the 40 µm pitch flexible
probes stacked and aligned to provide either active guarding or ground
shielding. (c) Single layer of 5-element flexible probes with 20 µm pitch
Fig. 1. Concept of electric potential microscopy using a flexible probe spacing.
array. (a) Imaging of electrostatic charge distribution on insulation sub-
strates to reveal latent fingerprints. (b) Micro-scale surface topography
imaging of conductive samples. (c) Spatial imaging of dielectric samples.
charged insulating surfaces, the SEPM array is able to image
the electrostatic charge distribution of the sample (Figure 1a).
frequencies√from 50 mHz to 330 MHz and noise floors as low The EPS array can also be used to obtain the surface topogra-
as 3.5nV/ Hz [6], [11], [12]. It is also possible to measure phy of conductive samples and to measure the local variation
DC potentials with microscopic resolution using EPS [13]. of relative permittivity in dielectric samples as shown in
The capabilities of EPS technology to image electric fields Figure 1b and Figure 1c respectively. To realise this array a
have been previously demonstrated by integrating such sen- hybrid approach is taken where the flexible sensing probes
sors into non-contact scanning electric potential microscopes are interfaced to the front-end electronics using conventional
(SEPM). SEPM can be applied to a plethora of applications off-the-shelf flexible printed circuit connectors (FPC). These
in non-destructive material characterisation [3], electrostatic connectors are commonly used as a board-to-board connector
charge imaging [14], measurement of propagation delay on for thin flexible copper PCBs in space constraint application
printed circuit boards [15] and localised electro-physiological such as smart rings or virtual reality headsets. The advantages
sensing [16]. The key characteristic of EPS that allows for such of this approach are manifold: 1) Contrarily to rigid substrates
measurements is the use positive feedback techniques, active such as glass or silicon, these flexible probes will not damage
guarding of the sensing probes and special biasing network the substrate in case of contact and are more durable due to
that makes it DC stable. Thus far, only conventional rigid their flexibility; 2) The process used to fabricate these probes
probes have been used to realize SEPM [17]. Additionally, are large-area compatible and relatively low-cost; 3) The
the scanning process of SEPM can take many hours to image a inherent insulating property and low electric permittivity of
large area using a single sensing probe. In this context, guarded flexible polymer substrates minimises cross-talk; 4) Finally,
arrays with micro-scale spatial resolution are highly desirable we used the low thickness and flexibility of polyimide foil
to reduce the scanning time and cross-talk between the sensing substrates to easily produce overlaying conductive tracks that
elements. Previously, an eight-element array of EPS was act as effective shielding or guarding structures, which satisfied
developed using semi-rigid coaxial cables with 2 mm spatial the aforementioned requirements of EPS sensing technology.
resolution [15]. Beardsmore-Rust et. al. [18] also developed an Multiple 5-element sensing arrays with pitch spacing of 40 μm
array of EPS for macro-scale electrostatic charge imaging of and 20 μm were fabricated. This work demonstrates that
insulating materials with a spatial resolution of 1 cm. Both of flexible thin-film technology can be readily used to fabricate
these works lacked the adequate spatial resolution for SEPM EPS sensors with micro metre resolution. This is an innova-
applications. However, the advancement of flexible thin-film tive approach to realise inexpensive, bespoke and large-area
electronics has opened up the possibility of fabricating active capacitively coupled sensor arrays [20]–[26].
and passive sensor arrays on flexible substrates [1], [19]–[26].

II. A PPROACH III. D ESIGN AND FABRICATION


Here, we present the design, fabrication, characterisation The sensor system was realized using hybrid approach
and various operation modes of the flexible 5-element array employing flexible thin-film technology and rigid surface
of EPS sensors for SEPM applications. We provide example mount (SMD) circuitry.
data for various operation modes and discuss the advantages
and shortfalls of the flexible sensor array in comparison to the
traditional single probe SEPM. A. Flexible Sensor Arrays
Figure 1 demonstrates the concepts of using the flexible Figure 2a shows an optical micrograph of the presented
array in three different operation modes. When scanning flexible array. The sensing probe array was fabricated by
Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
26330 IEEE SENSORS JOURNAL, VOL. 21, NO. 23, DECEMBER 1, 2021

depositing a 10 nm Titanium adhesion layer followed by


a 60 nm Gold layer on a 50 μm thick polyimide substrate
through e-beam evaporation. Gold was chosen since it does not
easily oxidise and provides both high conductivity and duc-
tility. In addition, 60 nm Gold thin-films were used since this
thickness provides a good trade-off between cost, conductivity
and connector stability. These metal layers were then patterned
using standard photolithography and lift-off. A schematic of
the layers structure is illustrated in Figure 2a. The sensing
probe array is then shielded or guarded by stacking and
aligning another identical thin-film structure on top of the
sensing probe array. The two thin-film arrays are then bonded
together by wrapping a Kapton tape around them. Using these
methods, arrays of 5 independent sensing electrodes with
pitch sizes of 40 μm and 20 μm were fabricated on one
substrate. These reduced pitch sizes were possible given the
polyimide’s low surface roughness (< 4 nm), good electric
insulating properties and compatibility with photolithography
techniques. The arrays can be guarded/shielded or un-shielded
as shown in Figure 2b and Figure 2c respectively. The spatial
resolution is defined by the width of each conductive sensing
element on the flexible probe array and the pitch spacing
between them.

B. Conditioning Circuit
Fig. 3. Assembly of 5-element SEPM array. (a) Photograph of the
The flexible sensing probes are conditioned using a high conditioning circuit used to read out the signals from the flexible probes.
impedance circuit made from discreet components on a 1 mm (b) Assembly of the flexible sensing array interfaced to the front-end
thick FR4 printed circuit board. Figure 3a shows an image of conditioning circuits using FPC connectors. All scale bars are 5 mm.
the conditioning circuit board for all five sensing elements.
Figure 3b shows a close-up of a commercially available FPC
connector with 600 μm pitch spacing used to interface the
flexible sensing probes and their corresponding flexible guard-
ing or shielding structure to the conditioning circuits. The
complete assembly results in a 5-element SEPM sensor array
enclosed in a grounded metal box that acts like a Faraday cage.
Figure 4a shows an equivalent circuit schematic diagram of
the EPS sensor technology. Ccoupling represents the capacitive
coupling between each sensing probe and the sample under test
and is defined by the sensing probe area and its separation to
the sample. Cin and Rin represent the internal input capaci-
tance and the input resistance of the EPS conditioning circuit.
Cin forms a capacitive voltage divider network with Ccoupling
and Rin forms a first order high-pass filter with the coupling
capacitance, both of which can decrease sensitivity. By use of
positive feedback the input impedance (Z in ) is enhanced which Fig. 4. Equivalent circuit representation of a single conditioned sensor.
in turn increases the sensitivity. High sensitivity is essential (a) Schematic diagram of the EPS sensor technology. Ccoupling is the
to achieve high spatial resolution, since the sensed signal capacitive coupling between the sensing probe and a sample under test.
Cin and Rin represent the input capacitance and the input resistance of
strength is inversely proportional to the size of the probes the EPS conditioning circuit. (b) Active guarding and (c) Ground shielding
and sample to sensor separation. Here the higher sensitivity of the sensing probes.
is attained by neutralising any parasitic capacitance between
the high impedance sensing probes to ground. Rin can also element in an array of sensors as discussed in the next section.
be artificially increased to improve the lower cut-off point of When the electric potential to be measured is large, active
the frequency response. Depending on the sample under test, guarding is replaced in favor of shielding at the cost of lower
either active guarding or passive shielding is implemented. sensitivity.
Active guarding is a form of positive feedback. The active
guarding topology is shown in Figure 4b, whereas the shield- C. Positioning System
ing topology is shown in Figure 4c. While guarding increases The complete sensor assembly is mounted on a com-
the sensitivity, it can also increase cross-talk between each puter controlled XYZ positioning system for characterisation

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
POURYAZDAN et al.: DESIGN AND CHARACTERISATION OF NON-CONTACT FLEXIBLE SENSOR ARRAY 26331

Fig. 5. Voltage noise spectral density of one of five input condition


circuits. obtained using the HP 3562A. Above√ the 1/f corner frequency
of 40 Hz, the noise level is flat at 4.8 nV/ Hz refereed to input.

and scanning electric potential measurements. The three-axis


positioning system used here is a re-purposed PCB milling
machine which has a resolution of 6.2 μm in X and Y axes.
It has an increment of 625 nm in Z-axis for precise adjustment
of the sensing probe to sample separation. All three axes
are driven by open-loop stepper motors with 1.8◦ rotational Fig. 6. Frequency response of the sensor array. (a) Each element in the
angle configured to rotate at quarter-steps using Toshiba sensor array was DC coupled to the spectrum analyser. (b) Spectrum with
TB67S109 chopper-type ICs. The motor drivers are connected the array AC coupled to the spectrum analyser through a 1 pF capacitor.
to a micro controller unit (MCU) which is connected to
are designed to have a DC voltage gain of 23 V V−1 (27.3 dB).
a PC. The MCU handles positioning commands received from
The higher noise levels at lower frequencies can be attributed
a bespoke LabView virtual instrument interface on the PC.
to the vibration in the sensing probes and cabling connecting
The LabView VI is also responsible for acquiring, visualising
the sensor array to the signal analyser.
and storing the digitised sensor outputs on the PC. Before
However, the sensors have different frequency response
the sensor outputs are digitised using a National Instrument
when AC coupled (Figure 6b). This is due to the change in
NI-6210 USB acquisition device the signals are filtered using
layout on both the flexible sensor array and on the sensor
4th order Butter-worth high-pass and low-pass filters to reduce
conditioning PCB. Specifically, sensor 1 and 5 differ to the
scanning artefacts and any other extrinsic noise.
rest of the sensing elements. Sensors 1 and 5 are both routed
on the edge of the flexible probe structure and have twice the
IV. P ERFORMANCE AND C HARACTERISATION
pitch spacing (1.2 mm) to their adjacent sensing elements at
The fully assembled sensor system was characterised under
the FPC connector end. The variation in spacing and layout
ambient conditions in an un-shielded environment using a
changes the capacitive coupling of the high impedance sensing
HP 3562A dynamic signal analyser. The frequency response,
traces to their surroundings and resulted in different sensitivity
spatial resolution and cross-talk between each sensing element
levels. Furthermore, the sensors have lower voltage gains or
within the array were evaluated. The noise floor of the
sensitivity when AC coupled. This is due to the capacitive
front-end conditioning circuit is the most important parameter
voltage divider network formed between CCoupling and Cin .
in non-contact SEPM as it defines the minimum detectable
The noise spikes around 50 Hz and its harmonics are unwanted
signal. Figure 5 shows the voltage noise spectral density of
interference pick-up from nearby 240 mains power. 6b also
one of the conditioning circuits when the input was shorted to
shows that the sensors have a lower cut-off point of ≈2 Hz
ground through a 1 µF capacitor. The 1/f corner √ is at 40 Hz. and an upper bandwidth limit of ≈100 kHz.
Above this frequency the noise level is 4.8 nV/ H z refereed
to input. B. Cross-Talk
Two methods were used to characterise the cross-talk
A. Frequency Response between each element of the sensor array, namely, spatial
The frequency response for all five sensors was measured cross-talk and spectral cross-talk as described below.
one at a time as shown in Figure 6. The measurement was 1) Spatial Cross-Talk: Figure 7 shows an schematic of the
carried out with each sensor being once resistively (DC) measurement setup and the results used to quantify the spatial
coupled and another time capacitively (AC) coupled (through cross-talk for various shielding and guarding configurations.
a calibrated 1 pF capacitor) to the source generator of the A 15 μm insulated wire was placed on a grounded copper
dynamic signal analyser. As shown in Figure 6a all five plane. A 10 kHz AC voltage signal with an amplitude of 1 V
sensors have similar responses when DC coupled. The sensors was applied between the wire and the copper plane. The sensor

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
26332 IEEE SENSORS JOURNAL, VOL. 21, NO. 23, DECEMBER 1, 2021

Fig. 7. Spatial cross-talk characterisation of the five-element array. (a) Schematic of the setup utilised to study the cross-talk between the micro-scale
probes. (b) 20 µm pitch ground shielded probe. (c) 40 µm pitch single layer un-shielded probe. (d) 40 µm stacked guarded electrode with maximum
active guarding and (e) 40 µm stacked guarded electrode with minimum active guarding.

array was then moved over this wire horizontally at a constant


height of 100 μm above the wire while the outputs of all
five sensors were recorded. This is illustrated in Figure 7a.
Multiple guarding and shielding configurations were tested and
the results from each is shown in Figures 7b to 7e. As shown,
ground shielding or no shielding at all were the most effective
configurations at minimising cross-talk (Figures 7b and 7c).
However, ground shielding reduces sensitivity, while lack of
shielding introduces interference from unwanted electric field
sources such as the mains 50 Hz signal. For the guarded double
layer structure shown in Figure 7d, the cross-talk increases
with the amplitude of the guard signal. While active guarding
can increase sensitivity, above a certain level the guard signal
also increases the cross-talk beyond a usable condition. This
can be attributed to capacitive coupling of the guard signals
from the adjacent sensors through the PCB and the Polyimide
flexible probe. To maintain both a high sensitivity and guarding
against unwanted noise sources, the active guard amplitude
was reduced to its minimum (Figure 7e).
2) Spectral Cross-Talk: Another approach to characterise the
cross-talk is to apply different frequencies to each sensing
element and compare the amount of spectral leakage in other
sensors. This is also illustrated in Figure 8a. The sources
were fed to the sensor using the same Polyimide structure
as the sensing probe. Figure 8b shows the isolation between Fig. 8. Spectral cross-talk characterisation. (a) Schematic diagram of
each sensing channel in a matrix form. Figure 8c shows the the measurement setup. (b) Cross-talk isolation between each sensing
element. (c) Sensor array resistively coupled to the sources and (d) sen-
measurements for the sensing probes resistively coupled to sor array capacitively coupled to the sources.
the AC signal sources. As expected, the spectral leakage to
the adjacent sensors is almost negligible in this case as the the isolation between the channels is adequate as to distinguish
sensors are driven by a low-impedance source. However, when the designated signal emitting from the sources under each
the sensing elements were separated by a distance of 100 sensing element. The isolation between channels is visualised
μm from the source (i.e. AC or capacitively coupled) the with the aid of a confusion matrix shown in Figure 8b. The
spectral leakage to the adjacent sensors increases (Figure 8d). orthogonal of the matrix shows the isolation between each
This is due to the unwanted capacitive coupling of the high sensor to the rest of the array. The highest isolation is between
impedance sensing probes to the adjacent sources. Here, the two furthest sensing probes as expected.
the active guarding configuration shown in Figure 7e was
chosen to carry out this measurement. Figure 8d shows that the V. R ESULTS AND D ISCUSSION
three sensing probes in the middle with 0.6 mm pitch spacing There are various possible operation modes for SEPM.
have more cross-talk in comparison to the two outer probes All modes, are classified as capacitive measurements as the
which have double the pitch spacing at the FPC end. However, sensing array is capacitively coupled to the sample under test.
Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
POURYAZDAN et al.: DESIGN AND CHARACTERISATION OF NON-CONTACT FLEXIBLE SENSOR ARRAY 26333

also deposited on a 75 μm PTFE (Polytetrafluoroethylene) film


by contact electrification. Figure 10b shows the electrostatic
image of the fingerprint obtained using a single sensor while
Figure 10c shows the image measured using the presented
flexible 5-element sensor array. The two micrographs are
comparable, except for the vertical scanning scar lines visible
in Figure 10c. The lines appear since the sensors do not have
the exact same performance. To compensate for the difference
in AC sensitivity of the sensors, the outputs of the sensors
were normalised. The full scan of the fingerprint using the
five-element sensor array reduced the measurement time by
a factor of five from 25 min to 5 min. The reduced scanning
Fig. 9. Various possible operation modes for SEPM. (a) Imaging time together with the close match between optical and SEPM
of spatial electrostatic charge distribution on dielectric materials. imaging have significant benefits for forensic applications.
(b) Surface topography measurement of conductive samples. (c) Spatial
imaging of dielectric variation within a surface.
B. Surface Topography
This makes the measurements both electrically and mechani- Another operation mode of SEPM is imaging of the surface
cally non-invasive; a property which is highly desirable for topography of conductive samples. This is an active measure-
any sensing technology. The measurement modes can be ment as an AC signal is applied to the sample under test.
broadly divided into the two categories of passive and active. Figure 9b shows the schematic diagram for this measurement
The measurements are called passive when no excitation mode. The amplitude of the excitation signal measured by
signal is applied to the sample and active when an AC the sensing probes varies with the change in distance to the
excitation voltage is used. In both modes of operation the sample due to the samples feature. This is noted as d in the
input impedance is maintained substantially high, to avoid schematic. To confirm the functionality of the presented array
any significant disturbance to the electric field under test. for SEPM based surface topography, the surface of a 20 pence
Figure 9 illustrates the concept for three different operation coin (Figure 10d) was mapped. A 1 V AC voltage signal
modes of imaging of spatial electrostatic charge distribution of 10 kHz was applied to the coin and a grounded copper plane
on dielectric material; Surface topography measurement of below it separated by an insulator (paper). The ground surface
conductive samples; Spatial imaging of dielectric variation below the paper ensures that the coin is not at an arbitrary
within a surface prompt by measuring change in local relative potential. Hence the sensed amplitude of the excitation signal
permittivity. Each one of these modes is described below. will vary with the features of the sample under test. In other
words, the variation in the measured signal is due to the
A. Electrostatic Charge Distribution separation between the sample features and the sensor tip or
the capacitive coupling between the two. Figure 10e shows the
Spatial imaging of electrostatic charge distribution is clas-
results obtained using the presented flexible sensor array while
sified as a passive operation mode. Figure 9a shows the
Figure 10f shows the single ridged sensing probe counterpart.
principles of this operation mode. The sensing probes are
Similar to the electrostatic charge mapping, the measurement
scanned over a charged surface at a constant sample to probe
time was reduced by a factor of five by use of the presented
separation of ≈ 150 μm. The presence of electrostatic charges
5-element sensor array for the surface topography, without
induces a voltage on the conductive sensing probe of the
significant impact on the image quality.
array. The voltage measured is proportional to the density
of these electro-static charges. These voltage are read by
the high impedance signal conditioning front-end circuitry. C. Dielectric Variation
The amplitude of the signals are proportional to the electric Scanning electric potential microscopy can also be used to
charge and the frequency of this time varying signal is a image the variation in relative permittivity (r ) of dielectric
function of the scanning speed. A detailed description of this surfaces. The schematic diagram for this mode is shown in
measurement mode is described elsewhere [1], [14]. Even Figure 9c. This is another active mode of operation where an
though the Polyimide substrate of the sensor array can collect AC excitation voltage is applied to a plane under the sample.
electro-static surface charges, these charges have no influence Here, a genuine plastic five pounds banknote was used as
on the measurement since they move with the array itself. an example of a composite dielectric surface (Figure 10g).
Furthermore, before the measurement is performed, the probes Various inks and materials utilised to produce this banknote
are de-ionised using a bench-top zero Volt 5kV air de-ioniser have different dielectric properties. By measuring the local
device. An immediate application for measuring spatial elec- variation in the amplitude of the 1 V 10 kHz AC excitation
trostatic charge distribution is the imaging of latent fingerprint voltage traveling through the bank note, an electrical image
deposited on insulating materials due to contact electrification of the note was produced based on its varied local dielectric
of two dissimilar materials. This was previously shown using properties. The permittivity of each material in the bank note
a single ridge sensing probe [27]–[29]. Figure 10a shows an alters the electric potential travelling though it and hence the
optical image of a fingerprint. The very same fingerprint was local amplitude of the sensed excitation signal will vary at

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
26334 IEEE SENSORS JOURNAL, VOL. 21, NO. 23, DECEMBER 1, 2021

Fig. 10. Electric potential imaging using the flexible sensor array. (a) Optical image of a fingerprint. Electrostatic imaging of the fingerprint using
(b) A single probe and (c) the presented 5-element flexible array. (d) Optical image of a 20 pence coin used as an example of a conductive sample
and (e) the corresponding surface topography of the coin acquired using the 5-element flexible array. (f) single probe measurement of the 20p coin.
(g) Optical photo a 5 Pounds plastic bank note used as a dielectric sample. (h) The corresponding SEPM imaging of the bank note using a single
probe and (i) using the 5-element sensor array.

each measured point. Increasing the amplitude of the excitation active guarding of the sensing probes. For applications such
signal can compensate for thicker samples. The presented as imaging of electrostatic charge distribution, where large
flexible sensing probe array was used to image the marked voltages (kV) are measured, a simple ground shield provides
area of the banknote (Figure 10i). The same area was also adequate cross-talk protection at the cost of reduced sensitivity.
measured using a single ridged probe image for comparison Guarding should only be used when a high sensitivity is
purposes. This is show in Figure 10h. Compared to the single required as this method of guarding increases cross-talk inter-
rigid sensing probe, the image produced by the array has ference. Due to the large-area compatible fabrication process
lower spatial resolution. This is mainly due to the fact that of the presented flexible and mechanically durable sensor
the Polyimide based sensor array has to be separated from array, these can be fabricated inexpensively on bio-compatible
the surface of the dielectric sample further to avoid physical polymeric substrates. The use of FPC connectors enables swift
contact; as any contact between the two dielectric samples can reconfiguration of the guarding/shielding as well as replacing
cause electrostatic charging of the sample due to contact elec- the probes to achieve various spatial resolutions. In the future,
trification. Electrostatic charges can induce large voltage on the we envision stacking multiple 1D arrays of these sensors to
sensing probes an interfere with the amplitude of the excitation achieve a 2D array. Additionally the size and pitch spacing
signal being measured. The increased in separation between of the sensing probes can be reduced to achieve a higher
the sample and the sensor array also reduces the capacitive spatial resolution. The presented technique can be extended
coupling between the two which in turn degrades the signal to take advantage of the bendable and stackable nature of
to noise ratio. Similar to the other measurement modes, the use the thin-film probes to implement a two dimensional array
of the 5-element array has reduced the measurement time by suitable for SEPM applications. With further advancement in
a factor of five but this time at the cost of reduced sensitivity thin-film transistor technology, and the use of innovative and
and spatial resolution. efficient readout typologies [30] it will be possible to integrate
a complete SEPM sensing system including the front-end
VI. C ONCLUSION conditioning circuitry on a single flexible substrate.
We presented a proof of concept and confirmed the feasi-
bility of using thin-film technology and conventional silicon
R EFERENCES
based active components to realize a microscopic array of
electric potential sensors for SEPM application. The use of [1] A. Pouryazdan, J. C. Costa, P. Lugoda, R. J. Prance, H. Prance, and
N. Münzenrieder, “Flexible micro-scale sensor array for non-contact
the 5-element sensor array reduced the measurement time by a electric potential imaging,” in Proc. IEEE Int. Conf. Flexible Printable
factor of five. This technique can be scaled up to have a higher Sensors Syst. (FLEPS), Aug. 2020, pp. 1–4.
number of sensing elements in an array. Here, we fabricated [2] K. Kamiya et al., “Characterization of edge radial electric field structures
in the large helical device and their viability for determining the location
a five-element array merely as a proof of concept. We found of the plasma boundary,” Nucl. Fusion, vol. 53, no. 1, Jan. 2013,
that by stacking thin-film structures it is possible to achieve Art. no. 013003.

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
POURYAZDAN et al.: DESIGN AND CHARACTERISATION OF NON-CONTACT FLEXIBLE SENSOR ARRAY 26335

[3] S. Mukherjee, P. Watson, and R. J. Prance, “Non-contact measurement of [27] P. Watson, R. Prance, H. Prance, and S. Beardsmore-Rust, “Imaging the
local conductivity variations in carbon fibre based composite materials,” time sequence of latent electrostatic fingerprints,” Proc. SPIE, vol. 7838,
J. Phys., Conf. Ser., vol. 307, Aug. 2011, Art. no. 012034. Oct. 2010, Art. no. 783803.
[4] N. Steinhausen, H. Prance, and R. Prance, “Human computer interface [28] J. Lowell and A. Rose-Innes, “Contact electrification,” Adv. Phys.,
based on eye movement (EOG) signals,” in Proc. AAATE, 2011, pp. 1–6. vol. 29, no. 6, pp. 947–1023, Dec. 1980.
[5] T. Grosse-Puppendahl et al., “Platypus: Indoor localization and identi- [29] D. J. Lacks and R. Mohan Sankaran, “Contact electrification of insu-
fication through sensing of electric potential changes in human bodies,” lating materials,” J. Phys. D, Appl. Phys., vol. 44, no. 45, Oct. 2011,
in Proc. 14th Annu. Int. Conf. Mobile Syst., Appl., Services, Jun. 2016, Art. no. 453001.
pp. 17–30. [30] J. C. Costa, A. Pouryazdan, R. J. Prance, H. Prance, and
[6] H. Prance, P. Watson, R. J. Prance, and S. T. Beardsmore-Rust, “Position N. Münzenrieder, “Flexible bootstrapped cascode system with feedback
and movement sensing at metre standoff distances using ambient electric for capacitive through-substrate electric potential measurements with a
field,” Meas. Sci. Technol., vol. 23, no. 11, 2012, Art. no. 115101. 55 db relative gain,” in IEDM Tech. Dig., Dec. 2020, pp. 1–4.
[7] R. Prance and C. Harland, “Electric potential sensor,” U.S. Patent
8 054 061, Nov. 8, 2011.
[8] A. Kainz et al., “Distortion-free measurement of electric field strength
with a mems sensor,” Nature Electron., vol. 1, no. 1, p. 68, 2018.
[9] S. Decossas, J. J. Marchand, and G. Brémond, “Electrical character- Arash Pouryazdan (Member, IEEE) received
isation of local electronic properties of self-assembled semiconductor the B.Eng. degree in electrical and electronic
nanostructures using AFM,” Phys. E, Low-Dimensional Syst. Nanos- engineering from the University of Sussex, U.K.,
truct., vol. 23, nos. 3–4, pp. 396–400, Jul. 2004. in 2013. He completed his industry funded
[10] C. Y. Nakakura, D. L. Hetherington, M. R. Shaneyfelt, P. J. Shea, and Ph.D. in engineering at the Sensor Technology
A. N. Erickson, “Observation of metal–oxide–semiconductor transistor Research Centre at the same institution in 2018.
operation using scanning capacitance microscopy,” Appl. Phys. Lett., His key research interests include non-contact
vol. 75, no. 15, pp. 2319–2321, Oct. 1999. electric potential sensing and microscopy, elec-
[11] A. Pouryazdan, “Non-contact single-ended electric potential measure- trostatic charge imaging, near-field capacitive
ments; techniques and applications,” Ph.D. dissertation, Univ. Sussex, communication, novel flexible sensors, wearable
Brighton, U.K., 2019. electronics, and innovative acoustic interfaces.
[12] A. Pouryazdan, J. C. Costa, R. J. Prance, H. Prance, and
N. Munzenrieder, “Non-contact long range AC voltage measurement,”
in Proc. IEEE SENSORS, Oct. 2019, pp. 4–16.
[13] A. Pouryazdan, J. C. Costa, F. Spina, R. J. Prance, H. Prance, and
N. Munzenrieder, “Non-contact measurement of DC potentials with
Júlio C. Costa (Graduate Student Member,
applications in static charge imaging,” in Proc. IEEE SENSORS,
IEEE) received the M.Sc. degree in micro and
Oct. 2020, pp. 1–4.
nanotechnologies from the New University of
[14] P. Watson, R. J. Prance, S. T. Beardsmore-Rust, and H. Prance, “Imaging Lisbon, Lisbon, Portugal, in 2015. He is currently
electrostatic fingerprints with implications for a forensic timeline,” pursuing the Ph.D. degree in flexible electronics
Forensic Sci. Int., vol. 209, nos. 1–3, pp. e41–e45, Jun. 2011. with the University of Sussex, U.K. His current
[15] W. Gebrial, R. J. Prance, T. D. Clark, C. J. Harland, H. Prance, and research interests include flexible transistors, cir-
M. Everitt, “Noninvasive imaging of signals in digital circuits,” Rev. cuits, and sensor systems.
Sci. Instrum., vol. 73, no. 3, pp. 1293–1298, Mar. 2002.
[16] C. Harland, T. Clark, N. Peters, M. J. Everitt, and P. Stiffell, “A compact
electric potential sensor array for the acquisition and reconstruction of
the 7-lead electrocardiogram without electrical charge contact with the
skin,” Physiol. Meas., vol. 26, no. 6, p. 939, 2005.
[17] A. Clippingdale, R. Prance, T. Clark, and F. Brouers, “Non-invasive
dielectric measurements with the scanning potential microscope,”
J. Phys. D, Appl. Phys., vol. 27, no. 11, p. 2426, 1994. Leonardo Garcia-Garcia (Member, IEEE)
[18] S. T. Beardsmore-Rust, P. Watson, R. J. Prance, C. J. Harland, and received the Ph.D. degree in mechanical
H. Prance, “Imaging of charge spatial density on insulating materials,” engineering from the University of Manchester,
Meas. Sci. Technol., vol. 20, no. 9, Sep. 2009, Art. no. 095711. U.K. He is a Research Fellow in Wearable
[19] J. C. Costa, F. Spina, P. Lugoda, L. Garcia-Garcia, D. Roggen, and Electronics and Smart Textiles with the Sensor
N. Münzenrieder, “Flexible sensors—From materials to applications,” Technology Research Centre, University of
Technologies, vol. 7, no. 2, p. 35, 2019. Sussex, U.K. His research interests include
[20] E. Pritchard, M. Mahfouz, B. Evans, S. Eliza, and M. Haider, “Flexible flexible sensors, human monitoring, wearable,
capacitive sensors for high resolution pressure measurement,” in Proc. and biomedical application.
IEEE Sensors, Oct. 2008, pp. 1484–1487.
[21] C. Metzger et al., “Flexible-foam-based capacitive sensor arrays for
object detection at low cost,” Appl. Phys. Lett., vol. 92, no. 1, 2008,
Art. no. 013506.
[22] Y.-J. Yang et al., “An integrated flexible temperature and tactile sensing
array using PI-copper films,” Sens. Actuators A, Phys., vol. 143, no. 1, Pasindu Lugoda (Member, IEEE) received
pp. 143–153, May 2008. the degree in electronic and communication
[23] M.-Y. Cheng, X.-H. Huang, C.-W. Ma, and Y.-J. Yang, “A flexible engineering from the University of Nottingham
capacitive tactile sensing array with floating electrodes,” J. Micromech. Malaysia Campus in 2012, and the Ph.D. degree
Microeng., vol. 19, no. 11, Nov. 2009, Art. no. 115001. in electronic textiles from the Advanced Textiles
[24] T. Someya, T. Sekitani, S. Iba, Y. Kato, H. Kawaguchi, and T. Sakurai, Research Group (ATRG), Nottingham Trent Uni-
“A large-area, flexible pressure sensor matrix with organic field-effect versity, in 2019. Thereafter, he worked as a Tex-
transistors for artificial skin applications,” Proc. Nat. Acad. Sci. USA, tile Development Engineer with MAS Holdings
vol. 101, no. 27, pp. 9966–9970, Jul. 2004. from 2012 to 2014. He was then appointed as
[25] F. Jiang, G.-B. Lee, Y.-C. Tai, and C.-M. Ho, “A flexible micromachine- a Research Fellow with the Sensor Technology
based shear-stress sensor array and its application to separation-point Research Centre, University of Sussex, develop-
detection,” Sens. Actuators A, Phys., vol. 79, no. 3, pp. 194–203, ing shape sensing textiles for orthotic applications, from 2018 to 2020.
Feb. 2000. Since 2020, he rejoined ATRG as a Research Fellow developing auto-
[26] H. Fang et al., “Capacitively coupled arrays of multiplexed flexible sili- mated systems to produce electronic yarns. His main scientific interests
con transistors for long-term cardiac electrophysiology,” Nature Biomed. are electronic textiles, wearable and flexible electronics, sensors, and
Eng., vol. 1, no. 3, p. 0038, Mar. 2017. medical devices.

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.
26336 IEEE SENSORS JOURNAL, VOL. 21, NO. 23, DECEMBER 1, 2021

Robert J. Prance received the B.Sc. degree from the University of Niko Münzenrieder (Senior Member, IEEE)
Liverpool, the M.Sc. degree from Lancaster University, and the D.Phil. received the Diploma degree in physics from the
degree from the University of Sussex. He is a Professor of Sensor Technical University Munich, Munich, Germany,
Technology with the School of Engineering and Design, University of in 2008, and the Ph.D. degree in electrical engi-
Sussex. His career started in low temperature physics and during this neering from ETH Zurich, Zürich, Switzerland,
time he developed expertise in electromagnetic sensors. This led to the in 2013. From 2015 to 2019, he was a (Senior)
invention of a novel electric potential sensor with generic measurement Lecturer with the Sensor Technology Research
capabilities. An application of this to electrophysiology received recog- Centre, University of Sussex, Brighton, U.K. He is
nition in 2002 with the Institute of Physics Best Paper Award. He has currently an Honorary Senior Lecturer at Sussex
published extensively in the scientific literature with some 158 journal and an Associate Professor for Physics of Mat-
publications, and 146 contributions to conferences and the media. ter with the Free University of Bozen-Bolzano,
He has directed a broad range of funded projects, both highly specific Bolzano, Italy, where he is working on flexible and stretchable thin-film
and interdisciplinary spanning, medicine, geophysics, materials testing, electronics, smart textiles, and wearable systems.
neuroscience, security, forensics and resonance imaging, funded from
a mixture of research council grants, government agency contracts and
industrial funding. He is actively engaged in the commercialization of the
IP generated by his group with the support of the University. He is a
member of the Institute of Physics and a Chartered Physicist.

Helen Prance received the B.Sc. degree in physics from the Univer-
sity of Birmingham in 1977, the M.Sc. degree in bio-medical physics
and bioengineering from the University of Aberdeen in 1978, and the
Ph.D. degree in experimental physics from the University of Sussex.
She is an Emeritus Professor (retired) and the former Head of the
Department of Engineering and Design, University of Sussex. Her
thesis topic was macroscopic quantum behavior in superconducting
quantum interference devices (SQUIDs). Following her appointment to
the Faculty of the School of Engineering, her research also encompassed
electromagnetic sensors. She holds two international patents in sensor
technology. She also has expertise in ultra-low noise amplifiers and
receivers from radio frequency to microwave, and cryogenically cooled
electronic systems. She is a member of the Sensor Technology Research
Centre, specializing in the applications of a novel electric potential sensor
to electrophysiology, human–machine interfacing, movement tracking,
and sensor arrays, and the Institute of Physics and a Chartered Physicist.

Authorized licensed use limited to: J.R.D. Tata Memorial Library Indian Institute of Science Bengaluru. Downloaded on February 01,2024 at 12:36:46 UTC from IEEE Xplore. Restrictions apply.

You might also like