sensors

Article
Influence of the Porosity of Polymer Foams on the
Performances of Capacitive Flexible
Pressure Sensors †
Sylvie Bilent, Thi Hong Nhung Dinh , Emile Martincic and Pierre-Yves Joubert *
Centre de Nanosciences et de Nanotechnologies, CNRS, Univ. Paris-Sud, Université Paris-Saclay,
91120 Palaiseau, France; sylvie.bilent@c2n.upsaclay.fr (S.B.); thi-hong-nhung.dinh@c2n.upsaclay.fr (T.H.N.D.);
emile.martincic@c2n.upsaclay.fr (E.M.)
* Correspondence: pierre-yves.joubert@c2n.upsaclay.fr; Tel.: +33-(0)1-70-27-06-53
† This paper is an extended version of the conference paper: Porous Polymer Based Flexible Pressure Sensors for
Medical Applications. In Proceedings of the EUROSENSORS 2018, Graz, Austria, 9–12 September 2018.




Received: 9 February 2019; Accepted: 19 April 2019; Published: 26 April 2019


Abstract: This paper reports on the study of microporous polydimethylsiloxane (PDMS) foams as a
highly deformable dielectric material used in the composition of flexible capacitive pressure sensors
dedicated to wearable use. A fabrication process allowing the porosity of the foams to be adjusted
was proposed and the fabricated foams were characterized. Then, elementary capacitive pressure
sensors (15 × 15 mm2 square shaped electrodes) were elaborated with fabricated foams (5 mm or
10 mm thick) and were electromechanically characterized. Since the sensor responses under load are
strongly non-linear, a behavioral non-linear model (first order exponential) was proposed, adjusted to
the experimental data, and used to objectively estimate the sensor performances in terms of sensitivity
and measurement range. The main conclusions of this study are that the porosity of the PDMS foams
can be adjusted through the sugar:PDMS volume ratio and the size of sugar crystals used to fabricate
the foams. Additionally, the porosity of the foams significantly modified the sensor performances.
Indeed, compared to bulk PDMS sensors of the same size, the sensitivity of porous PDMS sensors
could be multiplied by a factor up to 100 (the sensitivity is 0.14 %.kPa−1 for a bulk PDMS sensor and
up to 13.7 %.kPa−1 for a porous PDMS sensor of the same dimensions), while the measurement range
was reduced from a factor of 2 to 3 (from 594 kPa for a bulk PDMS sensor down to between 255 and
177 kPa for a PDMS foam sensor of the same dimensions, according to the porosity). This study
opens the way to the design and fabrication of wearable flexible pressure sensors with adjustable
performances through the control of the porosity of the fabricated PDMS foams.

Keywords: polymer-based flexible pressure sensors; microporous PDMS foam; electromechanical

characterizations; sensor behavioral modeling; sensor sensitivity and measurement range

1. Introduction
Flexible pressure sensors have raised strong interest in non-invasive health monitoring
applications [1]. Indeed, they allow wearable devices to be developed and have numerous monitoring
applications such as plantar support monitoring [2,3], wrist radial artery pulse wave monitoring [4],
or bandage compression control in the therapy of chronic venous disorders [5]. Several flexible
pressure sensing technologies are available [1] such as piezoresistive, piezoelectric, and capacitive
sensors [6–9]. In this paper, we focused on capacitive pressure sensors because of their low sensitivity
to temperature changes and the low power consumption readout electronics that can be associated
with them. The basic principle of such sensors is to sense the deformation of a flexible polymer under
stress by means of the capacitance changes read out between the electrodes integrated within the

Sensors 2019, 19, 1968; doi:10.3390/s19091968 www.mdpi.com/journal/sensors

Sensors 2019, 19, 1968 2 of 12

polymer, that move relatively to one another when a load is applied. Since these sensors are intended
to be set in contact with the skin, they must be disposable and thus have a low production cost.
To meet these requirements, silicone based polymers such as polydimethylsiloxane (PDMS) constitute
good candidates for elaborating such sensors. Indeed, PDMS is a non-toxic material, which features
dielectric properties as well as relevant mechanical properties, allowing up to 100% deformation
without damage [10]. However, PDMS features highly non-linear deformation behavior and is hardly
compressible [10]. These drawbacks actually limit the performances of PDMS-based capacitive sensors.
Indeed, they enable a large pressure measurement range to be reached (over a 400 kPa pressure
range [2,7,11]); however, the sensitivity is rather low (in the order of 1%.kPa−1 for pressure <10 kPa,
and <0.05%.kPa−1 for higher pressure [7,12]). For this reason, some authors have developed patterned
PDMS films in order to increase the sensor sensitivity. Such patterned films enable higher deformations
of the sensor dielectric layer, hence allowing significantly higher sensitivities to be obtained (e.g., up
to 55%.kPa−1 [13,14]). However, these sensors are actually operative in low measurement ranges
(<20 kPa) where they exhibit strong non-linear behavior, in addition to being possibly more complex
to fabricate and feature hardly predictable performances [15]. For these reasons, easy to design sensors
featuring enhanced sensitivity in larger pressure ranges would be welcome to address applications
related to human motion monitoring applications [9,14] including plantar pressure monitoring which
require pressure ranges up to 300 kPa [7].
The aim of this study was to consider capacitive pressure sensors fabricated with PDMS foams,
so that high sensitivities could be reached thanks to the introduction of microporosities within the
polymer without drastically reducing the resulting measurement range. The secondary motivation
was to determine a relationship between the porosity of the fabricated PDMS foams and the sensor
performances, so that PDMS foam based sensors could be designed by adjusting the porosity of the
used foams. PDMS foams can be obtained with a simple fabrication process by mixing the polymer
with sugar particles and further dissolving them [16–18] to make microporosities appear within the
material. In this paper, we proposed a methodology allowing PDMS foams with adjusted porosity to
be fabricated, and we studied the influence of the obtained porosity on the mechanical performances of
the resulting capacitive sensors. Furthermore, since polymer-based pressure sensors feature strongly
non-linear behaviors, we proposed a behavioral model that allows for a wide variety of sensors to be
objectively and quantitatively compared in terms of sensitivity and pressure range.

2. Sensor Principle
In this paper, we considered an elementary capacitive normal pressure sensor used as a work
prototype. This was composed of two square shaped brass electrodes (l = 15 mm side length, 1 mm
thickness) placed at each face of a microporous PDMS foam used as a compressible dielectric layer
(see Figure 1). The capacitance C of the sensor is given by:

ε0 εr A
C= (1)
d
where A is the area of the electrodes; εr is the relative permittivity of the foam; and d is the distance
between the electrodes. When a load is applied to the sensor (Figure 1), the foam undergoes a mechanical
compression resulting from the deformation of the polymer and the reduction of the size of the pores.
As a result, the sensor capacitance increases because of both the reduction of the distance d and the
increase of the foam permittivity. Indeed, the value of the foam permittivity results from a combination
of the intrinsic permittivity of the constitutive polymer and the permittivity of the volume of air trapped
within the foam pores (see Section 3.2). As a result, the foam permittivity is increased when the size of
the pores is reduced by compression, since the proportion of air within the foam is reduced. Finally,
the capacitance changes under load is given by ∆C = C − C0 where C0 is the load-free capacitance of
the sensor given for the initial distance d = d0 , and the initial value of εr (d0 ) of the uncompressed foam.
Sensors 2019, 19, 1968 3 of 12
Sensors 2018, 18, x FOR PEER REVIEW 3 of 12
Sensors 2018, 18, x FOR PEER REVIEW 3 of 12

Figure 1.
Figure 1. Principle
Principle of
of operation
operation of
of the
the flexible
flexible normal
normal pressure
pressure capacitive
capacitive sensor.
sensor.
Figure 1. Principle of operation of the flexible normal pressure capacitive sensor.

3.
3. PDMS Foam
3. PDMS Foam
3.1.
3.1. Fabrication
Fabrication Process
Process
3.1. Fabrication Process
Figure
Figure 22 shows
shows the the fabrication
fabrication process
process of of the
the microporous
microporous PDMS. PDMS. PDMS
PDMS is is formed
formed by by mixing
mixing
two Figure 2 shows the fabrication process of184A) the microporous PDMS. PDMS is184B) formed by mixing
two components: silicon elastomer (Sylgard 184A) and a curing agent (Sylgard 184B) at a standard
components: silicon elastomer (Sylgard and a curing agent (Sylgard at a standard
two
ratio components: silicon elastomerand (Sylgard 184A) andmixed
a curing agent (Sylgard 184B) atfora 60 standard
ratio of
of10:1.
10:1. The
Thesilicon
siliconelastomer
elastomer andcuring curingagent agentareare mixed and
and placed
placedin in
a desiccator
a desiccator min
for 60 to
min
ratio
degas. of 10:1. The
Household silicon
powder elastomer
sucrose and
(C Hcuring
O )agent are
crystals mixed
were thenand placed
added to in
the a desiccator
PDMS in for
volume 60ratios
min
12 22 11
to degas. Household powder sucrose (C12H22O11) crystals were then added to the PDMS in volume
to degas.from
ranging Household
4 to powder
6 (sugar sucroseper (C121H 22O11) crystals were then added to the PDMS in volume
ratios ranging from 4 to 6powder)
(sugar powder) (PDMS),
per 1both volumes
(PDMS), bothbeing estimated
volumes beingusing measuring
estimated using
ratios ranging
containers. The from 4 to
preparation 6 (sugar
was powder)
poured into pera 1 (PDMS),
Petri dish both
and volumes
compacted being
to form estimated
a block using
of the
measuring containers. The preparation was poured into a Petri dish and compacted to form a block
measuring
mixture. containers.
The mixture The cured
was preparation
on a was poured
stove for 2 h into
at 90 aC.
◦ Petri dish
After and compacted
curing, the mixture towas
form a block
removed
of the mixture. The mixture was cured on a stove for 2 h at 90 °C. After curing, the mixture was
of
fromthethe
mixture.
Petri The mixture was cured on abathstovetofor 2 h at 90 °C. After curing, the mixture was
removed from dish and soaked
the Petri dish andinsoaked
a water in a water dissolve
bath to the sacrificial
dissolve sugar crystals,
the sacrificial thereby
sugar crystals,
removed
obtaining from the Petri dish
the microporous PDMSand foam.
soakedOnce in a water bath to completely
dissolve thedissolved,
sacrificialthe sugar crystals,
thereby obtaining the microporous PDMS foam.the sugar
Once thewas
sugar was completely dissolved, PDMS foam
the PDMS
thereby
was obtaining
again placed the
on microporous
the stove for 1 PDMS
h 30 at foam.
75 ◦ C Once
to the
remove sugar
the was
excess completely
water. dissolved,
Finally, the the PDMS
microporous
foam was again placed on the stove for 1 h 30 at 75 °C to remove the excess water. Finally, the
foam
PDMS was
wasagain toplaced
cutPDMS the was on the
shape stove for 1 h 30constitute
at 75 °C the to remove theelementary
excess water. Finally,
used the
microporous cutoftothe electrodes
the shape oftothe electrodes tocapacitive
constitute the capacitive sensor
elementary for
microporous
characterization PDMS was
purposes. cut to the shape of the electrodes to constitute the capacitive elementary
sensor used for characterization purposes.
sensor used for characterization purposes.
1) Prepare PDMS (1:10) 2) Add sugar (4:1 and 6:1) 3) Mix sugar with PDMS 4) Compress the mixture
1) Prepare PDMS (1:10) 2) Add sugar (4:1 and 6:1) 3) Mix sugar with PDMS 4) Compress the mixture

5) Cure the mixture 6) Dissolve the sugar 7) Dry the porous PDMS 8) Cut the microporous PDMS
5) Cure the mixture 6) Dissolve the sugar 7) Dry the porous PDMS 8) Cut the microporous PDMS
(90°C for 2h) (75°C for 1h30)
(90°C for 2h) (75°C for 1h30) Brass
Brass d0
Porous
Porous
PDMS d0
PDMSBrass
Brass
l
l
(a)
(a)

(b) (c)
(b) (c)
Figure2.2.(a)
(a) Fabrication process of the microporous PDMS(b) foams; (b) example ofpore
4:1 small
Figure
Figure 2. (a)Fabrication
Fabricationprocess of the
process ofmicroporous PDMS PDMS
the microporous foams; example
foams; (b) of 4:1 smallof
example 4:1 PDMS
small
pore sample;
foam PDMS foam sample; (c) of
(c) compressibility compressibility of
the of 4:1 smallof the of 4:1foam
pore small pore PDMS foam sample.
pore PDMS foam sample; (c) compressibility thePDMS
of 4:1 smallsample.
pore PDMS foam sample.
3.2.
3.2. Characterization
Characterization of of the
the PDMS
PDMS Foams
Foams
3.2. Characterization of the PDMS Foams
In
In order
order to
to study
study the the influence
influence of of the
the porosity
porosity onon the
the mechanical
mechanical performances
performances of of the
the fabricated
fabricated
In order to study the influence of the porosity on the mechanical performances of the fabricated
sensors,
sensors, wewe first
first adjusted
adjusted the the size
size of
of the
the used
used sugar
sugar crystals,
crystals, and
and then
then adjusted
adjusted the
the sugar:PDMS
sugar:PDMS ratioratio
sensors, we first adjusted the size of the used sugar crystals, and then adjusted the sugar:PDMS ratio
volume used to fabricate the foams. To do so, commercially available sugar
volume used to fabricate the foams. To do so, commercially available sugar crystals of unknown crystals of unknown
volume used to fabricate the foams. To do so, commercially available sugar crystals of unknown
dimensions
dimensions werewere sorted
sorted into
intothree
threesets
setsbybymeans
means ofof 500
500 µm,µm, 1000
1000 andand
µm,µm, 20002000
µm µm sieving
sieving grids.
grids. The
dimensions were sorted into three sets by means of 500 µm, 1000 µm, and 2000 µm sieving grids. The
The
size size distribution
distribution of theofcrystal
the crystal within
within each each set estimated
set was was estimated by evaluating
by evaluating the cross
the cross sectionsection of
of sugar
size distribution of the crystal within each set was estimated by evaluating the cross section of sugar
sugar particles picked up within each of the three sets, and observed by means
particles picked up within each of the three sets, and observed by means of a Keyence VHX-1000 of a Keyence VHX-1000
particles picked up within each of the three sets, and observed by means of a Keyence VHX-1000
Sensors 2019, 19, 1968 4 of 12
Sensors 2018, 18, x FOR PEER REVIEW 4 of 12

opticalmicroscope
optical microscope(around
(around100100crystals
crystalswere
wereobserved
observedinineach
eachset).
set).An
Anexample
exampleofofsugar
sugarparticles
particlesofof
smallsize
small sizeisispresented
presentedininFigure
Figure3a.
3a.

(a) (b)

(c) (d) (e)

Figure
Figure 3. 3.
(a)(a) Small
Small sugar
sugar particles
particles observed
observed with
with an an optical
optical microscope;
microscope; (b) Size distribution
(b) Size distribution of sugar
particles after the sieving procedure; (c–e) SEM images of the cross-section of microporous PDMSof
of sugar particles after the sieving procedure; (c–e) SEM images of the cross-section
microporous
foams PDMSmedium,
featuring small, foams featuring small,respectively.
and large pores, medium, and large pores, respectively.

Thesizes
The sizesofof the
the used crystals,
crystals,defined
definedasasthetheroot
rootmean
mean square
square value of the
value observed
of the crystal
observed cross
crystal
sections,
cross werewere
sections, estimated
estimated fromfrom the the
distributions
distributions of the cross
of the crosssections
sectionsof of
the
thecrystals,
crystals,presented
presentedin
inFigure
Figure3b.
3b.AsAsaaresult,
result,the the crystal
crystal sizes
sizes were
were found to be be 470
470 ±± 100
100 µmµm(small
(smallsize),
size),700
700±±230230µm µm
(mediumsize),
(medium size),and
and1100
1100±±330 330µm µm(large
(largesize).
size).Furthermore,
Furthermore,the thecross
crosssections
sectionsofofPDMS
PDMSfoam foamsamples
samples
wereobserved
were observedby bymeans
meansofofa ascanning
scanningelectron
electronmicroscope
microscope(Hitachi
(HitachiS3600N).
S3600N).Figure
Figure3c–e
3c–eshows
shows
someexamples
some examplesofofthe thepore
poredistributions
distributionsininfoamsfoamsfabricated
fabricatedwithwithsmall,
small,medium,
medium,and andlarge
largecrystals,
crystals,
respectively.The
respectively. Thesizessizesofofthe theobtained
obtainedpores
poreswere
wererather
ratherdifficult
difficulttotoaccurately
accuratelyevaluate,
evaluate,since
sincethe the
orientationof the
orientation of three-dimensional
the three-dimensional pores arepores are unknown
actually actually unknown in this two-dimensional
in this two-dimensional observation.
observation. itNevertheless,
Nevertheless, can be qualitatively it canassessed
be qualitatively assessed
that the sizes of the that the sizes
obtained pores of the obtained
correlated pores
to the sizes
ofcorrelated to the
the sacrificial sugarsizes of theused
crystals sacrificial sugarthe
to fabricate crystals
pores. used to fabricate
As a result, the pores.
the fabricated 4:1As
anda 6:1
result,
foams the
fabricated
are referred 4:1 and 6:1
as being foams are
obtained referred
with as being “medium”,
either “small”, obtained with either “small”,
or “large” “medium”,the
pores throughout or rest
“large”of
pores
this throughout the rest of this paper.
paper.
Theporosity
The porosityf of f ofthe
thefabricated
fabricatedfoamsfoamsisiscalculated
calculated from
from thethe density
density ofof
thethe PDMS
PDMS foam
foam (ρfoam
(ρfoam ) )and
and
the bulk PDMS density (ρ ) using
the bulk PDMS density (ρbulk ) using Equation (2):
bulk Equation (2):

ρ foam
ρ f oam
f =f1=−1 − , , (2)
(2)
ρbulk
ρ bulk

whereρfoam
where ρfoamisisdetermined
determinedfromfromthetheratio
ratioofofthe
themass
masstotothe
thevolume
volumeofofthe
thefabricated
fabricatedfoam, foam,and
andwhere
where
ρbulk = 33 [19]. The relative permittivity of the PDMS foam (ε ) is a combination of the relative
ρbulk = 1100 kg/m [19]. The relative permittivity of the PDMS foam (εr r) is a combination of the relative
1100 kg/m
permittivity
permittivityofofPDMS PDMS ==2.69
PDMS(ε(εPDMS 2.69[20])
[20])and
andthethe relative
relative permittivity
permittivity of of (εair(ε=air1).=The
airair 1). The resulting
resulting foam
foam relative permittivity can be theoretically estimated
relative permittivity can be theoretically estimated using [17]: using [17]:

ε r = ε air f + (1 − f )ε PDMS , (3)

Sensors 2019, 19, 1968 5 of 12

εr = εair f + (1 − f )εPDMS , (3)

where f is the foam porosity. The porosity and the expected relative permittivity of the various PDMS
foams fabricated in this study are shown in Table 1, according to the pore sizes and sugar:PDMS ratios.
In addition, the permittivity of the PDMS foam samples were experimentally estimated through the
value of the capacitance measured between two 2 mm thick copper sheet electrodes cut to the size of
the PDMS samples (5 cm diameter disk) and placed on both faces of the PDMS foam. The capacitance
was measured with a HP 4192A impedance analyzer operated at 1 MHz. For the considered samples,
the measured capacitances were found to be in the pF range and the relative permittivity of the foams
was subsequently determined using Equation (1).

Table 1. Variation of the relative permittivity of the Microporous PDMS as a function of the porosity.

Volume Ratio Sizes of Sugar

Porosity (%) Theoretical εr Experimental εr
Sugar: PDMS Particles (µm)
0:1 (bulk) No crystal 0 2.69 2.67 ± 0.06
Small: 470 ± 100 78.64 ±1.40 1.36 1.56 ± 0.01
4:1 (foam) Medium: 700 ± 230 80.36 ±1.40 1.33 1.36 ± 0.06
Large: 1100 ± 330 80.81 ± 1.40 1.32 1.34 ± 0.01
Small: 470 ± 100 81.96 ±1.40 1.30 1.30 ± 0.13
6:1 (foam) Medium: 700 ± 230 84.36 ±1.40 1.26 1.25 ± 0.07
Large: 1100 ± 330 83.53 ±1.40 1.28 1.29 ± 0.01

As observed in Table 1, the porosity of the fabricated foams significantly increased with the
sugar:PDMS ratio (e.g., from around 80% for 4:1 foams up to over 84% for 6:1 foams with medium size
pores). To a lesser extent, the foam porosity also increased with the size of the pores (e.g., from around
82% up to 83.5% for small and large pores in 6:1 PDMS foams, respectively,). Finally, as expected from
Equation (4), the permittivity of the foams decreased when the porosity of the foam increased.

4. Sensor Electromechanical Characterizations

In this section, elementary pressure sensors were fabricated as described in Figure 2 with PDMS
foams of various porosities and 1 mm thick brass electrodes were added on both sizes of the PDMS
foams for electromechanical characterization purposes. The sensors were characterized in terms of
sensitivity and measurement range thanks to a dedicated electromechanical test bench.

4.1. Measurement Setup

The used electromechanical test bench is described in Figure 4a and is composed of (i) an indenter
applying normal stress to the sensor under test through an ISEL iMC-S8 robotic arm, (ii) a FUTEK
FSH00105 110 N force sensor connected to a HP34401A voltmeter measuring the stress applied in a
0–488 kPa pressure range, and (iii) a HP4192A impedance analyzer measuring the sensor capacitance
changes at a frequency of 1 MHz. All of the measurement setup was controlled by a computer operating
under MATLAB and used for the acquisition, storage, and processing of the measurement data.
A typical sensor response curve obtained with a 6:1 small pore PDMS foam sensor, is presented in
Figure 4b (blue dots). As commonly observed with polymer based capacitive pressure sensors, the
obtained response curve is strongly non-linear, the sensitivity of the sensor is high for low stress values
and exhibits a saturation behavior for large stress values [1,2,6,11,16,18]. In the case of polymer-foam
based sensors, the sensor capacitance increases with the applied pressure under the combined effects of
the reduction of the foam thickness and the increase of the foam permittivity (see Section 2). To illustrate
these effects, the evolution of the foam permittivity with the applied pressure is presented in Figure 4c.
The permittivity was estimated from the measured capacitance C, the experimental values of d and
Equation (1). One can note that the estimated permittivity changed from the initial value of εr (d0 ) = 1.3
 expected from Equation (4), the permittivity of the foams decreased when the porosity of the foam
increased.

4. Sensor Electromechanical Characterizations

In this section, elementary pressure sensors were fabricated as described in Figure 2 with 6PDMS
Sensors 2019, 19, 1968 of 12
foams of various porosities and 1 mm thick brass electrodes were added on both sizes of the PDMS
foams for electromechanical characterization purposes. The sensors were characterized in terms of
to εr (d) ≈ and
upsensitivity 2.6 when the pressure
measurement exceeded
range thanks100
to akPa; the permittivity
dedicated of the foam
electromechanical reached
test bench.a saturation
value which was close to the permittivity of bulky PDMS (εPDMS = 2.69), meaning that the pores had
4.1. significantly
been Measurementreduced
Setup under compression, i.e., the porosity f tended toward zero in Equation (3).
The changes of d with the applied pressure are presented in Figure 4d. The distance d was estimated
The used electromechanical test bench is described in Figure 4a and is composed of (i) an
from the initial value of the unloaded sensor (d0 ) and the set positions of the robotic arm applying
indenter applying normal stress to the sensor under test through an ISEL iMC-S8 robotic arm, (ii) a
the load onto the sensor (Figure 4a). One can note that the thickness d tended to have a minimum
FUTEK FSH00105 110 N force sensor connected to a HP34401A voltmeter measuring the stress
value under high pressure loads. This is consistent with the fact that, when the pores have almost
applied in a 0–488 kPa pressure range, and (iii) a HP4192A impedance analyzer measuring the sensor
been completely “closed”, the mechanical behavior of PDMS foam tends to show the same mechanical
capacitance changes at a frequency of 1 MHz. All of the measurement setup was controlled by a
behavior as bulk PDMS, which is that of a hardly compressible polymer which features a Poisson
computer operating under MATLAB and used for the acquisition, storage, and processing of the
coefficient close to 0.49 [21].
measurement data.

Sensors 2018, 18, x FOR PEER REVIEW 6 of 12

(a) (b)

(c) (d)

Figure 4.
Figure 4. (a)
(a) Electromechanical
Electromechanical testtest bench
bench used
used to
to measure
measure the the sensor
sensor response
response curve
curve (capacitance
(capacitance
change versus applied pressure); (b) Typical sensor response curve (blue dots) for a
change versus applied pressure); (b) Typical sensor response curve (blue dots) for a sensor featuringsensor featuring a
a 6:1 small pore foam and first order exponential behavior model (black line) allowing
6:1 small pore foam and first order exponential behavior model (black line) allowing sensor sensitivity sensor
sensitivity
(S) (S) andrange
and pressure pressure
(PR)range
to be (PR) to beand
defined defined and estimated
estimated from the from the characteristic
characteristic pressurepressure
PC and
PC and
the the maximum
maximum relative relative capacitance
capacitance ∆Cmax /CΔ0C
change change (c)/C
. max 0. (c) Capacitance
Capacitance changeschanges
(black (black line)foam
line) and and
foam permittivity
permittivity changeschanges (blueversus
(blue line), line), versus applied
applied pressure,
pressure, for a sensor
for a sensor featuring
featuring a 6:1 pore
a 6:1 small smallfoam
pore
foam layer;
layer; (d) Changes
(d) Changes of theofPDMS
the PDMS
foamfoam thickness
thickness d with d with
appliedapplied pressure
pressure for afor a sensor
sensor featuring
featuring a 6:1a
6:1 small
small porepore
foam foam layer.
layer.

4.2. Sensor Behavioral

A typical sensorModel
response curve obtained with a 6:1 small pore PDMS foam sensor, is presented
in Figure 4b (blue dots). Asand
In order to objectively commonly observed
quantitatively with polymer
compare the sensorbased capacitivewe
performances, pressure
proposedsensors, the
building
obtained
an response curve
electromechanical is strongly
behavioral modelnon-linear, the sensitivity
of the pressure sensor. Theof the sensor
model used inisthis
high for was
study low astress
first
values and exhibits a saturation
order exponential model, expressed by: behavior for large stress values [1,2,6,11,16,18]. In the case of
polymer-foam based sensors, the sensor capacitance increases with the applied pressure under the
combined effects of the reduction of∆C the(Pfoam ∆Cmax  and
) = thickness theC) increase of the foam permittivity (see

1 − e−(P/P , (4)
Section 2). To illustrate these effects,Cthe
0 evolutionC0 of the foam permittivity with the applied pressure
is presented in Figure 4c. The permittivity was estimated from the measured capacitance C, the
experimental values of d and Equation (1). One can note that the estimated permittivity changed from
the initial value of εr(d0) = 1.3 up to εr(d) ≈ 2.6 when the pressure exceeded 100 kPa; the permittivity
of the foam reached a saturation value which was close to the permittivity of bulky PDMS (εPDMS =
2.69), meaning that the pores had been significantly reduced under compression, i.e., the porosity f
Sensors 2019, 19, 1968 7 of 12

where C0 is the initial capacitance value of the unloaded sensor; ∆Cmax is the maximum capacitance
change; P is the applied pressure; and Pc is the characteristic pressure at which the capacitance change
Sensors 2018, 18, x FOR PEER REVIEW 7 of 12
reaches 63% of its maximum value (Figure 4b). The choice of such a model is justified by the behavior
ofexpected
the sensor to response
be high at curves. Indeed, as
low pressure mentioned
levels, since itinincreased
Section 4.1, thethe
with capacitance
combinedchanges
effects were
of the
expected to be high at low pressure levels, since it increased with the combined
reduction of d and the increase of the foam permittivity εr(d) (Figure 4). On the other hand, effects of the reduction
at high
d and the
ofpressure increase
levels, the of the foam permittivity
capacitance change reached εr (d) a(Figure
maximum 4). On the other
value hand,
due to the at high pressure
saturation of the
levels, the capacitance change reached a maximum value due to the saturation
permittivity (Figure 4b) as well as the incompressibility of the foam featuring “closed” pores. of the permittivity
In order
(Figure 4b) as the
to measure wellrelevance
as the incompressibility
of the model, ofresponse
the foamcurves
featuringof “closed” pores.
a 4:1 small poreIn foam
order sensor
to measure
were
the relevance of the model, response curves of a 4:1 small pore foam sensor were
measured during a load cycle (load increasing from 0 to 450 kPa) and a release cycle (load decreasing measured during a
load cycle (load increasing from 0 to 450 kPa) and a release cycle (load decreasing
from 450 kPa back to 0 kPa). The model was fitted to the experimental data by the numerical from 450 kPa back to
0 adjustment
kPa). The modelof ΔCmaxwasandfitted
Pc. to the experimental
Figure 5 shows the data by the numerical
experimental adjustment
curves together withofthe ∆Cfitted
max and Pc .
model
Figure 5 shows the experimental curves together with the fitted model curves.
curves. One can note a fair agreement between the experimental and model curves fitted with a R2 One can note a fair
agreement
higher than between the experimental
0.99. Figure 5 also shows andthemodel curves
average fitted
model curve, a R2 higher
with obtained bythan 0.99. Figure
averaging 5 also
the load and
shows the average
release model curves. model curve, obtained by averaging the load and release model curves.

Figure5.5.Typical
Figure Typicalresponse
responsecurve
curvefor
fora asensor
sensorfeaturing
featuring4:1
4:1small
smallpore
porePDMS
PDMSfoam.
foam.Experimental
Experimentaldatadata
obtainedduring
obtained duringa aload
loadcycle
cycle(blue
(bluedots)
dots)and
andananadjusted
adjustedload
loadmodel
model(blue
(blueline)
line)and
anda arelease
releasecycle
cycle
(redcross)
(red cross) with
with anan adjusted
adjusted release
release model
model (red(red line).
line). Average
Average loadload
andand release
release modelmodel is represented
is represented by
by the black
the black line. line.

InInthis
thispaper,
paper,weweproposed
proposedthat thethe
that sensitivity S ofSthe
sensitivity sensor
of the can can
sensor be defined by the
be defined bytangent to theto
the tangent
origin of the sensor
the origin of the adjusted model curve,
sensor adjusted modelandcurve,
that theand
pressure measurement
that the range PR corresponded
pressure measurement range PR
tocorresponded
95% of the maximum
to 95% ofcapacitance
the maximum change. These quantities
capacitance change. Thesewerequantities
used as the sensor
were usedperformance
as the sensor
features to compare
performance theto
features different
compare sensors, and aresensors,
the different expressed
andby:are expressed by:
max
𝐶
) =∆C
 
S𝑆(%. 𝑘𝑃𝑎
%.kPa −1 =
C0 𝐶
Pc 𝑃 (5)
PR (kPa) = =
𝑃𝑅(𝑘𝑃𝑎) 3P3𝑃
c

4.3. Influence of the Foam (Sugar:PDMS) Ratio

4.3. Influence of the Foam (Sugar:PDMS) Ratio
First, the influence of the sugar:PDMS ratios on the sensor performances were studied. To do
First, fabricated
so, sensors the influencewithof 15
the×sugar:PDMS
15 mm2 brass ratios on the and
electrodes sensor performances
PDMS foams (Figurewere2)studied. Toby
featured do4:1
so,
sensors
and fabricated
6:1 ratios foamswith 15 ×pores)
(small 15 mmwere
2 brass electrodes and
considered. The PDMS foams
thickness (Figure
of the foams 2) featured
was d0 =by 4:1 and
5 mm ±
6:1 ratios foams (small pores) were considered. The thickness of the foams was
0.5 mm. The sensors were characterized using the test bench shown in Figure 4 and electromechanical d 0 = 5 mm ± 0.5 mm.
The sensors
responses werewere characterized
obtained using load using the test
and release bench
cycles: loadsshown in Figure
increasing from 04up and electromechanical
to 488 kPa were first
responses were obtained using load and release cycles: loads increasing from
applied (load), then loads decreasing from 488 to 0 kPa were applied (release). In Figure 6, one 0 up to 488 can
kPanote
were
first applied (load), then loads decreasing from 488 to 0 kPa were applied (release).
that the response curves obtained for the PDMS foam based sensor featured much higher capacitance In Figure 6, one
can note
changes thatsensors
than the response curves
fabricated withobtained
bulk PDMS.for the PDMS
This foam based
confirmed sensor
that higher featuredsignificantly
porosities much higher
capacitance changes than sensors fabricated with bulk PDMS. This confirmed
increased the sensor deformation under the load and capacitance changes. Additionally, a hysteresis that higher porosities
significantly increased the sensor deformation under the load and capacitance changes. Additionally,
a hysteresis appeared between the load and release curves. For each sensor response, the hysteresis
was estimated by differentiation between the areas measured under the load and the release curves.
It is expressed in percentage comparatively to the load curve. The hysteresis has been found to be
lower than 8.85% for bulk PDMS sensors, and lower than 6.5% for PDMS foam based sensors. If the
Sensors 2019, 19, 1968 8 of 12

appeared between the load and release curves. For each sensor response, the hysteresis was estimated
by differentiation between the areas measured under the load and the release curves. It is expressed in
percentage comparatively to the load curve. The hysteresis has been found to be lower than 8.85%
Sensors 2018, 18, x FOR PEER REVIEW 8 of 12
for bulk PDMS sensors, and lower than 6.5% for PDMS foam based sensors. If the hysteresis in all
cases was inferior to 10%, we chose to apply the proposed model (Equation (5)) on a mean response
hysteresis in all cases was inferior to 10%, we chose to apply the proposed model (Equation (5)) on a
curve obtained by averaging the load and release curves. The assessed sensitivity and pressure range
mean response curve obtained by averaging the load and release curves. The assessed sensitivity and
obtained using the proposed model are presented in Table 2 for each tested sensor. It can be observed
pressure range obtained using the proposed model are presented in Table 2 for each tested sensor. It
that the porosity increased the sensitivity S of the sensor from 6.50%.kPa−1 to 13.07%.kPa−1 for 4:1 and
can be observed that the porosity increased the sensitivity S of the sensor from 6.50%.kPa-1 to
6:1, respectively; moreover, these sensitivities became significantly higher than the sensitivity of the
13.07%.kPa-1 for 4:1 and 6:1, respectively; moreover, these sensitivities became significantly higher
bulk PDMS sensor with the same dimensions (0.14%.kPa−1 ). On the other hand, the pressure range PR
than the sensitivity of the bulk PDMS sensor with the same dimensions (0.14%.kPa-1). On the other
decreased from 594 kPa (bulk PDMS) to 243 kPa (4:1 PDMS foam) and 166 kPa (6:1 PDMS foam) when
hand, the pressure range PR decreased from 594 kPa (bulk PDMS) to 243 kPa (4:1 PDMS foam) and
the porosity increased.
166 kPa (6:1 PDMS foam) when the porosity increased.

Figure6. 6.
Figure Variation
Variation ofcapacitance
of the the capacitance according
according to thepressure,
to the applied applieddepending
pressure, on
depending on the
the sugar:PDMS
sugar:PDMS
ratio ratio
(small pores (smallfoam,
PDMS 0 = 5 mm
pores(dPDMS foam, (d0mm).
± 0.5 = 5 mm ± 0.5 mm).

Table 2. Summary of the characterization results for PDMS bulk film and porous PDMS film as a
Table 2. Summary of the characterization results for PDMS bulk film and porous PDMS
function of the sugar/PDMS ratio and pore size (d0 = 5 mm ± 0.5 mm).
film as a function of the sugar/PDMS ratio and pore size (d0 = 5 mm ± 0.5 mm).
VolumeVolume Ratio
Ratio Sugar: PDMSSizes of Sugar
Sizes of Sugar Particles PR PR (kPa)S Hysteresis
C0 (pF) C0 (pF) S (%.kPa
−1 )
-1) Hysteresis (%)
Sugar:
0:1 PDMS Particles Bulk (kPa)
1.14 ± 0.02 (%.kPa
594 0.14 (%) 8.85
0:1 Bulk Small 1.14 ± 0.020.89 ± 0.02594 243 0.14 6.50 8.85 5.60
4:1 Medium 0.62 ± 0.02 255 6.87 5.41
Small 0.89 ± 0.02 243 6.50 5.60
Large 0.54 ± 0.02 257 8.85 5.31
4:1 Medium 0.62 ± 0.02 255 6.87 5.41
Small 0.82 ± 0.02 166 12.95 6.42
Large 0.54 ± 0.02 257 8.85 5.31
6:1 Medium 0.69 ± 0.02 177 13.70 4.58
Small Large 0.82 ± 0.020.84 ± 0.02166 230 12.95 11.74 6.42 6.62
6:1 Medium 0.69 ± 0.02 177 13.70 4.58
4.4. Influence of the Foam Pore Large
Sizes 0.84 ± 0.02 230 11.74 6.62

The influence of the size of the pores constituting the PDMS foams were studied in this section.
4.4. Influence of the Foam Pore Sizes
The load and release curves of the sensors fabricated with small, medium, and large pores are
The influence
presented in Figureof the sizethe
7a for of 4:1
thefoams
pores and
constituting
in Figurethe
7bPDMS
for thefoams were In
6:1 foams. studied in this section.
both figures, one can
The
seeload
thatand release
higher curves
pore sizesofincreased
the sensorsthefabricated with small,changes
sensor capacitance medium,underand large
load.pores
For are presented
all considered
infoams,
Figureit7acan
forbetheobserved
4:1 foamsinand in Figure
Figure 8 that7bthe
forgeneral
the 6:1 foams.
tendencyIn both
for a figures, one can see that
given sugar:PDMS ratiohigher
is that
pore sizes increased
the increase the sensor
of the pore capacitance
size induced changes
the increase under
of the load.
sensor For all considered
sensitivity (e.g., for S foams, it can-1be
= 6.50 %.kPa up
observed in Figure
to 8.85%.kPa -1 8 that
for 4:1 theand
small general
largetendency for a given sugar:PDMS
foams, respectively), ratio is that
while the pressure rangestheremained
increase ofrather
the
pore size induced
unchanged. This the increase
general remarkof the sensor
does sensitivity (e.g.,
not completely to S
applyfor = 6:1
the 6.50PDMS −1 up to 8.85%.kPa−1
%.kPafoam with large pores.
for 4:1can
This small and large foams,
be attributed respectively),
to the fact that the 6:1while
large the
porepressure
PDMS foamranges remained
is rather rather
brittle and,unchanged.
due to some
loss of material, the dimensions of the used sample might not be exactly similar to those of the other
samples, so that the performances are not completely comparable to the other sensors.
Sensors 2019, 19, 1968 9 of 12

This general remark does not completely apply to the 6:1 PDMS foam with large pores. This can
be attributed to the fact that the 6:1 large pore PDMS foam is rather brittle and, due to some loss of
material, the dimensions of the used sample might not be exactly similar to those of the other samples,
so that the performances are not completely comparable to the other sensors.
Sensors 2018, 18, x FOR PEER REVIEW 9 of 12

(a) (b)
Figure 7.
Figure 7. Variation
Variation of
ofthe
thecapacitance
capacitanceaccording
accordingtotothe
the applied
applied pressure,
pressure, depending
depending on on
thethe pore
pore sizesize
for
for 4:1
(a) (a) and
4:1 and (b) PDMS
(b) 6:1 6:1 PDMS foams
foams (d50 =mm
(d0 = 5 mm ± 0.5
± 0.5 mm).mm).

Figure 8. Sensitivity
Figure 8. and pressure
Sensitivity and pressure range
range of
of the
the fabricated
fabricated sensors
sensors according the used
according the used bulk
bulk PDMS
PDMS or
or
PDMS foams. Sugar:PDMS ratios were 0:1, 4:1, and 6:1, and the pore sizes were small (S), medium (M),
PDMS foams. Sugar:PDMS ratios were 0:1, 4:1, and 6:1, and the pore sizes were small (S), medium
and large (L).
(M), and large (L).
4.5. Influence of the Foam Thickness
4.5. Influence of the Foam Thickness
In this section, sensors fabricated with 4:1 and 6:1 foams of 5 mm and 10 mm thicknesses (medium
pores)Inwere
this compared.
section, sensors fabricated
The sensor with 4:1 and
electromechanical 6:1 foams
responses areof 5 mm and
presented 10 mm
in Figure thicknesses
9. One can see
(medium pores) were compared. The sensor electromechanical responses are presented
that an increase in the thickness enhanced the sensor deformation under load, and resulted in in Figure 9.
a small
One can see that an increase in the thickness enhanced the sensor deformation under load,
increase in the sensor sensitivity, as presented in Table 3. For example, S increased from 13.7%.kPa−1 and
resulted
up to overin 15.7%.kPa
a small increase
−1 whenin d
theincreased
sensor sensitivity,
from 5 mm as to
presented
10 mm, in
forTable 3. For
medium example,
pore S increased
6:1 PDMS foams.
0
from 13.7%.kPa-1 up to over 15.7%.kPa-1 when d0 increased from 5 mm to 10 mm, for medium pore
6:1 PDMS foams.

Table 3. Summary of characterization results for PDMS foams as a function of the sugar:PDMS ratio
and pore size, for two different thicknesses (d0 = 5 mm ± 0.5 mm and d0 = 10 mm ± 0.5 mm).

Volume Ratio Sugar: Sizes of Sugar d0 PR S Hysteres

C0 (pF)
PDMS Particles (mm) (kPa) (%.kPa-1) is (%)
10 0.40 ± 0.02 221 8.32 7.22
4:1 Medium
5 0.62 ± 0.02 261 6.88 5.41
10 0.38 ± 0.02 180 15.77 5.06
6:1 Medium
5 0.69 ± 0.02 177 13.70 4.58
Sensors 2019, 19, 1968 10 of 12
Sensors 2018, 18, x FOR PEER REVIEW 10 of 12

Figure
Figure 9. 9.Variation
Variationof of
thethe capacitance
capacitance according
according to the
to the applied
applied pressure
pressure depending
depending on thickness
on the the thickness
of
ofporous
the the porous
PDMS PDMS medium
medium porespores
withwith sugar:PDMS
sugar:PDMS ratiosratios
of 4:1ofand
4:1 6:1.
and 6:1.

3. Summary of characterization results for PDMS foams as a function of the sugar:PDMS ratio
4.6.Table
Influence of the Dynamics of the Applied Load
and pore size, for two different thicknesses (d0 = 5 mm ± 0.5 mm and d0 = 10 mm ± 0.5 mm).
In this section, a 6:1 medium pore PDMS foam based sensor was characterized under load
Volumecontinuously
applied Ratio Sizes of Sugar
or discontinuously dduring the PR cycles. In
S the continuous
Hysteresis
0
C0load-and-release
(pF) mode,
Sugar: PDMS Particles (mm) (kPa) (%.kPa−1 ) (%)
the applied pressure was continuously increased or decreased during the load-and-release cycle and
the complete cycle including 10 0.40 ± 0.02 221 8.32 7.22
4:1 Mediumthe acquisition of a 100 measurement points was achieved in eight
5 0.62 ± 0.02 261 6.88 5.41
minutes. In the discontinuous mode, the applied pressure was increased or decreased step by step
with a 10 0.38 ± 0.02cycle was
10 the complete 180 achieved15.77 5.06
6:1 s pause between each step, so that
Medium in 25 min. The obtained
5 0.69 ± 0.02 177 13.70 4.58
load and release curves are presented in Figure 10, and the performances parameters are shown in
Table 4. One can see that the performances were rather unchanged whichever way the load was
4.6. Influence
applied of the Dynamics
or released. of the
However, Applied
the Loadhysteresis was reduced from 4.58% down to 2.92% when
observed
the In
load
thiswas morea 6:1
section, “slowly”
mediumapplied, i.e., the
pore PDMS material
foam has more
based sensor wastime to recoverunder
characterized its shape
load when the
applied
load is released progressively.
continuously or discontinuously during the load-and-release cycles. In the continuous mode, the
applied pressure was continuously increased or decreased during the load-and-release cycle and the
complete cycle including the acquisition of a 100 measurement points was achieved in eight minutes.
In the discontinuous mode, the applied pressure was increased or decreased step by step with a 10 s
pause between each step, so that the complete cycle was achieved in 25 min. The obtained load and
release curves are presented in Figure 10, and the performances parameters are shown in Table 4.
One can see that the performances were rather unchanged whichever way the load was applied or
released. However, the observed hysteresis was reduced from 4.58% down to 2.92% when the load
was more “slowly” applied, i.e., the material has more time to recover its shape when the load is
released progressively.

Table 4. Summary of the characterization results for the 6:1 medium pore PDMS foam based sensor for
the load applied continuously and discontinuously (d0 = 5 mm ± 0.5 mm).

Volume Ratio Sizes of Sugar PR S Hysteresis

Applied Load C0 (pF)
Sugar: PDMS Particles (kPa) (%.kPa−1 ) (%)
Continuously 177 13.70 4.58
6:1 Medium 0.69 ± 0.02
Discontinuously 174 13.66 2.92
Figure 10. Variation of the capacitance according to the applied pressure, depending on the
dynamics of the applied load (continuously or discontinuously with 10 pauses at each
measurement point) for a 6:1 medium pore, porous PDMS that was 5 mm thick.
 with a 10 s pause between each step, so that the complete cycle was achieved in 25 min. The obtained
load and release curves are presented in Figure 10, and the performances parameters are shown in
Table 4. One can see that the performances were rather unchanged whichever way the load was
applied or released. However, the observed hysteresis was reduced from 4.58% down to 2.92% when
the load was more “slowly” applied, i.e., the material has more time to recover its shape when the
Sensors 2019, 19, 1968 11 of 12
load is released progressively.

FigureFigure 10. Variation

10. Variation of the capacitance
of the capacitance accordingaccording to the
to the applied applied
pressure, pressure,ondepending
depending the dynamicson the
of thedynamics
applied load
of (continuously
the applied or discontinuously
load (continuouslywithor10discontinuously
pauses at each measurement point) for
with 10 pauses at aeach
6:1 medium pore, porous
measurement point)PDMS
for athat
6:1 was 5 mm pore,
medium thick. porous PDMS that was 5 mm thick.

5. Conclusions
In this study, PDMS foams with different porosities were obtained by controlling the size and
the concentration of sugar crystals mixed with PDMS in the fabrication process. The obtained foams
were characterized in terms of porosity, and electromechanical characterizations of capacitive PDMS
foam based pressure sensors were carried out to study the influence of the porosity on the sensor
performances. In order to quantitatively compare the sensor performances, a non-linear behavioral
model was proposed and adjusted to the sensor pressure response curves to estimate the sensor
performances in terms of sensitivity and pressure range. The main conclusion of the study is that the
porosity of the foam significantly increases the sensitivity of the sensor: it is multiplied by a factor of 100
when compared to the sensitivity of a bulk PDMS sensor of the same size and shape. Additionally, the
pressure range of PDMS foam based sensors is reduced by a factor of 2 to 3 depending on the porosity
when compared with bulk PDMS sensors. Future works will focus on the shear stress performances of
PDMS foam based sensors, so that three axes of flexible pressure sensors can be considered. To do this,
fully flexible sensors will be microfabricated with flexible thin copper electrodes by means of a transfer
of film fabrication process [11]. Further works will also focus on the dynamic behavior of PDMS foam
sensors according to the porosity and the robustness of such sensors with respect to intensive use.
This study opens the way to the design and fabrication of advanced wearable pressure sensors, the
performances of which can be adjusted to the application aim through the control of the porosity of the
fabricated foam.

Author Contributions: All authors contributed to this work and to the elaboration of the paper.
Acknowledgments: This work was done within the C2N PIMENT clean room facilities and partly supported by
the RENATECH network and the general Council of Essonne.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Trung, T.Q.; Lee, N.E. Flexible and Stretchable physical sensor integrated platforms for wearable
human-activity monitoring and personal healthcare. Adv. Mater. 2016, 28, 4338–4372. [CrossRef] [PubMed]
2. Lei, K.F.; Lee, K.F.; Lee, M.Y. Development of a flexible PDMS capacitive pressure sensor for plantar pressure
measurement. Microelectron. Eng. 2012, 99, 1–5. [CrossRef]
Sensors 2019, 19, 1968 12 of 12

3. Motha, L.; Kim, J.; Kim, W.S. Instrumented rubber insole for plantar pressure sensing. Org. Electron. Phys.
Mater. Appl. 2015, 23, 82–86. [CrossRef]
4. Schwartz, G.; Tee, B.C.-K.; Mei, J.; Appleton, A.L.; Wang, K.; Bao, Z. Flexible polymer transistors with high
pressure sensitivity for application in electronic skin and health monitoring. Nat. Commun. 2013, 4, 1859.
[CrossRef] [PubMed]
5. Khaburi, J.A.; Dehghani-Sanij, A.A.; Nelson, E.A.; Hutchinson, J. Effect of bandage thickness on interface
pressure applied by compression bandages. Med. Eng. Phys. 2012, 34, 378–385. [CrossRef] [PubMed]
6. Huang, Y.; Yuan, H.; Kan, W.; Guo, X.; Liu, C.; Liu, P. A flexible three-axial capacitive tactile sensor with
multilayered dielectric for artificial skin applications. Microsyst. Technol. 2017, 23, 1847–1852. [CrossRef]
7. Surapaneni, R.; Guo, Q.; Xie, Y.; Young, D.J.; Mastrangelo, C.H. A three-axis high-resolution capacitive tactile
imager system based on floating comb electrodes. J. Micromech. Microeng. 2013, 23, 75004. [CrossRef]
8. Viry, L.; Levi, A.; Totaro, M.; Mondini, A.; Mattoli, V.; Mazzoli, B.; Beccai, L. Flexible three-axial force sensor
for soft and highly sensitive artificial touch. Adv. Mater. 2014, 26, 2659–2664. [CrossRef] [PubMed]
9. Tao, L.; Zhang, K.-N.; Tian, H.; Wang, D.-Y.; Chen, Y.-Q.; Yang, Y.; Ren, T.-L. Graphene-Paper pressure sensor
for detecting human motions. ACS Nano 2017, 11, 8790–8795. [CrossRef] [PubMed]
10. Dinh, T.-H.-N.; Martincic, E.; Dufour-Gergam, E.; Joubert, P.-Y. Mechanical characterization of PDMS films
for the optimization of polymer based flexible capacitive pressure microsensors. J. Sens. 2017, 2017, 8235729.
[CrossRef]
11. Dinh, T.H.N.; Martincic, E.; Dufour-Gergam, E.; Joubert, P.-Y. Capacitive flexible pressure sensor:
microfabrication process and experimental characterization. Microsyst. Technol. 2016, 22, 465–471. [CrossRef]
12. Dobrzynska, J.A.; Gijs, M.A.M. Polymer-based flexible capacitive sensor for three-axial force measurements.
J. Micromech. Microeng. 2013, 23, 015009. [CrossRef]
13. Mitrakos, V.; Macintyre, L.; Denison, F.C.; Hands, P.J.W.; Desmulliez, M.P.Y. Design, manufacture and testing
of capacitive pressure sensors for low-pressure measurement ranges. Micromachines 2017, 8, 1–10. [CrossRef]
14. Xu, F.; Li, X.; Shi, Y.; Li, L.; Wang, W.; He, L.; Liu, R. Recent Developments for flexible pressure sensors: A
Review. Micromachines 2018, 9, 1–17. [CrossRef]
15. Mannsfeld, S.C.B.; Tee, B.C.-K.; Stolenberg, R.M.; Chen, C.V.H.-H.; Barman, S.; Muir, B.V.O.; Sokolov, A.N.;
Reese, C.; Bao, Z. Highly sensitive flexible pressure sensors with microstructured rubber dielectric layers.
Nat. Mater. 2010, 9, 859–864. [CrossRef] [PubMed]
16. Yoon, J.L.; Choi, K.S.; Chang, S.P. A novel means of fabricating microporous structures for the dielectric
layers of capacitive pressure sensor. Microelectron. Eng. 2017, 179, 60–66. [CrossRef]
17. Chhetry, A.; Yoon, H.; Park, J.Y. A flexible and highly sensitive capacitive pressure sensor based on conductive
fibers with a microporous dielectric for wearable electronics. J. Mater. Chem. C 2017, 5, 10068–10076. [CrossRef]
18. Kwon, D.; Lee, T.; Kim, M.S.; Kim, S.T.; Kim, S.; Park, I. Porous dielectric elastomer based ultra-sensitive
capacitive pressure sensor and its application to wearable sensing device. Transducers 2015, 2015, 299–302.
19. Mark, J.E. Polymer Data Handbook; Oxford University Press: New York, NY, USA, 1999.
20. Tsai, P.J.; Nayak, S.; Ghosh, S.; Puri, I.K. Influence of particle arrangement on the permittivity of an elastomeric
composite. AIP Adv. 2017, 7, 1–8. [CrossRef]
21. Johnston, I.D.; McCluskey, D.K.; Tan, C.K.L.; Tracey, M.C. Mechanical characterization of bulk Sylgard 184
for microfluidics and microengineering. J. Micromech. Microeng. 2014, 24, 035017. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).