Direct patterning of organic conductors on knitted textiles for long-term electrocardiography

Abstract

Wearable sensors are receiving a great deal of attention as they offer the potential to become a key technological tool for healthcare. In order for this potential to come to fruition, new electroactive materials endowing high performance need to be integrated with textiles. Here we present a simple and reliable technique that allows the patterning of conducting polymers on textiles. Electrodes fabricated using this technique showed a low impedance contact with human skin, were able to record high quality electrocardiograms at rest, and determine heart rate even when the wearer was in motion. This work paves the way towards imperceptible electrophysiology sensors for human health monitoring.

Full text

Direct patterning of organic conductors on knitted
textiles for long-term electrocardiography
Seiichi Takamatsu, Thomas Lonjaret, Dakota Crisp, Jean-Michel Badier,
George G. Malliaras, Esma Ismailova
To cite this version:
Seiichi Takamatsu, Thomas Lonjaret, Dakota Crisp, Jean-Michel Badier, George G. Malliaras,
et al.. Direct patterning of organic conductors on knitted textiles for long-term electrocardio-
graphy. Scientific Reports, Nature Publishing Group, 2015, 5 (15003 ), <10.1038/srep15003>.
<hal-01235717>
HAL Id: hal-01235717
https://hal-amu.archives-ouvertes.fr/hal-01235717
Submitted on 30 Nov 2015
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destin´ee au d´epˆot et `a la diffusion de documents
scientifiques de niveau recherche, publi´es ou non,
´emanant des ´etablissements d’enseignement et de
recherche fran¸cais ou ´etrangers, des laboratoires
publics ou priv´es.

1
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
www.nature.com/scientificreports
Direct patterning of organic
conductors on knitted textiles for
long-term electrocardiography
Seiichi Takamatsu1,*, Thomas Lonjaret2,3,*, Dakota Crisp2, Jean-Michel Badier4,
George G. Malliaras2 & Esma Ismailova2
Wearable sensors are receiving a great deal of attention as they oer the potential to become a key
technological tool for healthcare. In order for this potential to come to fruition, new electroactive
materials endowing high performance need to be integrated with textiles. Here we present a simple
and reliable technique that allows the patterning of conducting polymers on textiles. Electrodes
fabricated using this technique showed a low impedance contact with human skin, were able to
record high quality electrocardiograms at rest, and determine heart rate even when the wearer was
in motion. This work paves the way towards imperceptible electrophysiology sensors for human
health monitoring.
Textile-based health monitoring devices are receiving a great deal of interest for consumer and medical
applications1–6, where they are being used to monitor parameters such as blood pressure7 and heart
rhythm5,6. e main advantage of textiles as substrates for biomedical devices stems from the fact that
they establish and maintain conformal contact with the human body in a non-invasive way1–3. Indeed,
shirts1,4, gloves8, and wristbands3 outtted with sensors have been used to demonstrate the potential of
this technology. ere is an increased interest in the development of electrocardiography (ECG) elec-
trodes, to which textiles can grant the conformal skin contact which is necessary in order to accurately
detect the small electrophysiological signals of the heart4,5,9,10. Wearable ECG electrodes can enable the
remote monitoring of people at risk, signal the onset of heart disease, and help monitor physical activity
during exercise11. Despite the large interest in cutaneous electrodes, patterning conducting materials on
stretchable fabrics has been hampered by their three-dimensional nature, which makes dicult to apply
conventional patterning processes1,12–14. e choice of the pattern transfer technique is dened by the
textile’s type and its structure. e yarns in woven and non-woven fabrics are interlaced in a dense net,
resulting in very at, but mostly non-stretchable, structures15. e bers in knitted textiles are assembled
in a snake shape that can be altered by applying a mechanical force to the knit varying its design. Such
wave-form bers mimic a mechanical spring design, providing the textile with a considerable resist-
ing force when its shape is changing. Usually direct pattern transfer can be done on thin woven and
non-woven textiles, while embroidery and knitting are used to pattern thick and structured textiles.
Micro-contact, inkjet, and screen printing can be used for making conductive patterns on textiles16,12.
erefore, the controlled transfer of a pattern using these techniques on thick knitted fabrics can be
obstructed. e conductive materials need to be coated not only on the surface of the knitted structure
but also inside, providing continuous contact between the yarns during its mechanical deformation.
Micro-contact and inkjet printing allow direct pattern transfer which is usually done on thin textiles
since a small amount of the ink can be transferred at the time. e coating happens only on the top layer
of the textile however the pattern conductivity is maintained even under stretch. In the screen printing,
1National Institute of Advanced Industrial Science and Technology, Tsukuba 305-8564, Japan. 2Department of
Bioelectronics, Ecole Nationale Supérieure des Mines, CMP-EMSE, MOC, 13541 Gardanne, France. 3MicroVitae
Technologies, 13590 Meyreuil, France. 4Aix Marseille Université, INS/Inserm, UMR-S 1106, 13005 Marseille, France.
*These authors contributed equally to this work. Correspondence and requests for materials should be addressed
to E.I. (email: esma.ismailova@emse.fr)
Received: 10 May 2015
Accepted: 11 September 2015
Published: 08 October 2015
OPEN

www.nature.com/scientificreports/
2
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
additives are used to reduce the spreading of the ink (i.e., silver paste) and improve spatial resolution17,18.
e viscosity of inkjet printing inks must also be engineered for printability19. As a result, in both screen-
and inkjet printing, ink optimization has a negative impact on the nal conductivity.
Embroidery and knitting consist of using individual bers and subsequently introducing them into
the textile’s structure14,20–22. in stainless steel, copper or other metal wires are employed to sew con-
ductive patterns on textiles by embroidery. A large amount of wires is needed during these processes
to create a pattern. Usually such techniques are largely integrated in textile industries to create inter-
connections between sensors and output electronic systems. Additionally, organic electronic materials
can be also applied via die-coating, where the ber gets coated by traveling through a nozzle lled
with a conducting material and then weaved or knitted during textile manufacturing. Examples of this
approach include large area touch sensors14, electric wiring components13, and organic electrochemical
transistors15 using organic conducting material such as poly(3,4-ethylenedioxythiophene):poly(styrene
sulfonate) (PEDOT:PSS). Electrodes made of PEDOT:PSS have been successfully used in cutaneous
applications highlighting their high performance compared to commercial ones. With proven biocom-
patibility, the material has been used in in vivo studies and was shown to decrease the electrical imped-
ance over classical electrodes23. Most importantly, this commercially available polymer can be easily
modied without compromising its practicality. PEDOT’s rheological properties make it attractive for
direct integration with textiles. A key challenge, therefore, for the elaboration of biomedical devices
on textiles is the development of simple patterning techniques that allow the deposition of conducting
biocompatible materials only on a desired area, with neither the need for cutting and sewing nor for
additives that aect conductivity.
In this work, we report a simple and reliable fabrication process that allows the patterning of con-
ducting polymers on thick knitted textiles, thereby yielding wearable and conformal electronic devices
for healthcare monitoring. We applied this process to the fabrication of cutaneous electrodes using the
commercially available, high conductivity polymer PEDOT:PSS, and an ionic liquid gel that promotes
better skin contact. We evaluated the performance of these devices in electrophysiological recordings
of a human heart. e measurements showed that the textile electrodes form a low impedance contact
with skin and are able to capture the electrophysiological signal with high accuracy, even during motion.
ese results pave the way for the simple fabrication of a variety of biomedical devices on textiles.
Results
Patterning technique. e technique was inspired from the Japanese kimono dyeing process (the
Yuzen method). According to this method, a rice paste is initially painted on the textile surface to form a
stencil. e dye is subsequently applied, coating only the areas that are free of rice paste which is subse-
quently removed in water, resulting in the precise and beautiful patterns of a kimono. In Fig.1 we show
the adaptation of this technique to the patterning of PEDOT:PSS on textiles. We used polydimethylsilox-
ane (PDMS) as the stencil due to its hydrophobic nature, which can conne the aqueous PEDOT:PSS
solution, as well as due to its mechanical properties, which are a good match for the so and stretchable
structure of knitted textiles. e patterning process started by preparing a negative master made of a
polyimide (PI) lm on which the outline of the desired pattern was carved by a laser cutting machine.
PDMS was subsequently applied on this master by spin coating (step 1). e textile was then placed on
top of the PI master, and the PDMS was progressively transferred into the textile (step 2). By adjusting
PDMS viscosity (using dierent amount of the curing agent) and thickness (using dierent spin coating
speeds) it is possible to control its diusion into the textile and to replicate the master design. A short
thermal annealing was used to cure the PDMS completing the transfer step. e PI master was then
delaminated from the textile surface. Finally, the conducting polymer solution was brush-painted on the
textile and baked to dry (step 3). Contrary to the Yuzen method, in which the rice paste is removed, the
PDMS stencil remains in the textile aer the patterning process, and can be used to pattern additional
layers (see below). Typical patterning results can be seen in Fig.2. e textile used in this study was a
knitted polyester with a thickness of 300 μ m and a stretch capability up to 50% (in the knit direction).
A ower pattern with regularly spaced curved features and a lines-and-spaces test pattern are shown in
Fig.2a,b, respectively. From the latter it can be seen that the resolution is greater than 1 mm, which is
adequate for cutaneous electrodes. Additionally, the patterning on dierent textiles is shown in Suppl.
Fig. S1. ese results show that patterning of structures as ne as 0.5 mm is possible on tightly woven
textiles. e sheet resistance of the patterned PEDOT:PSS stripes was 230 Ω /sq. is value was the same
as that of a dip-coated textile, showing that there is no inuence of the patterning process on the elec-
trical properties of the conducting polymer.
Electrode fabrication. Following the described fabrication process, 1 cm2 PEDOT:PSS electrodes
were patterned on knitted polyester. Fig.2c shows the nal structure of an electrode integrated on the
polyester wristband. e deposition of the PEDOT:PSS can be followed by the deposition of additional
layers, which will also be patterned using the PDMS stencil. In this particular case we added an ionic
liquid (IL) gel, as these materials are known to help establish low-impedance cutaneous contacts24. By
adding about 45 μ L/cm2 of IL gel formulation, patterning was achieved on just the active area of the elec-
trode. e wristband containing the PEDOT:PSS/IL gel electrode was wrapped around the arm of a vol-
unteer and skin impedance was measured and compared to that of a hydrogel-assisted Ag/AgCl medical

www.nature.com/scientificreports/
3
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
electrode with a similar active area. e measurements were carried out on three dierent volunteers in
rapid succession, on the same spot. e results, shown in Fig.2d, follow the typical trend of the electrical
impedance measured on skin24, and take into account subject to subject skin condition variations, rep-
resented by condence area of the mean impedance curve. ey reveal a lower impedance for the textile
electrode (2.5 times lower at 10 Hz), paving the way for applications in cutaneous electrophysiology.
Evaluation of electrodes in ECG monitoring. e potential of the textile electrodes in biomedical
monitoring was assessed using electrocardiography recordings. ECG is a common diagnostic technique
in the clinic, and is also used to monitor heart rate during exercise. ECG measurements in dynamic
conditions are usually performed with electrodes placed on the chest, in order to reduce motion artifacts.
To evaluate the performance of the textiles electrodes in a wearable conguration, we used the limb lead
II conguration (one electrode worn on the right wrist and a second one worn on the le ankle)25 and
performed measurements on volunteers at rest and various states of movement consecutively during
3 hours. In this study, conventional medical electrodes were placed next to the textile ones for compari-
son. e results are presented in Fig.3. During a rst measurement, the volunteer was sitting at rest, to
reduce muscle movements and respiratory artifacts. is position allows to obtain high quality record-
ings that can be used to detect heart function anomalies. Both, the textile and the medical electrodes
show the typical waveform of the heart activity with similar amplitudes (Fig.3a). High resolution PQRST
complexes, corresponding to the dierent phases of polarization and depolarization of cardiac cells, can
be clearly detected with both electrodes. To compare the signal quality recorded from textile electrodes
with medical electrodes, SNR values were calculated aer ltering in the 0.5 Hz–40 Hz frequency band
(typically used in ambulatory patient monitoring). A feature recognition program rst isolated each
PQRST complexes and then found the Crest Factor, or peak-to-RMS ratio, which is the ratio between
the amplitude of the R peak26 (which is the peak with the highest amplitude) and the RMS value of the
signal all along the complex. e SNR value is averaged over 25 dierent complexes obtained with textile
and medical electrodes and found to be equal to 16.3 dB (± 0.1 dB) for both electrodes.
Evaluation of electrodes under ambulatory conditions. e evaluations were performed rst in
the movement and then during long-term experiments. Fig.3b shows recorded signals obtained while
the volunteer was standing up and moving. Movement can have a large impact on ECG recordings27,
inducing an undulating baseline that can disturb the acquisition of the PQRST complex. is is evident
in the recording obtained by the medical electrode (in red), where the R peak is barely visible. In con-
trast, the inuence of motion is signicantly lower on the recordings from the textile electrode (in blue),
Figure 1. Process ow for the patterning of PEDOT:PSS on textiles, inspired by the Japanese kimono
dying method. PDMS is rst deposited on a polyimide master dening the outline of the desired pattern.
e textile is then placed on the polyimide lm and the PDMS is progressively transferred into the bulk of
the textile. Aer a short thermal annealing, the PEDOT:PSS solution is brush-coated on the unprotected
area of the textile and dried.

www.nature.com/scientificreports/
4
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
which show richer signal content, with a well-dened R peak and even a visible T wave (the positive wave
following the R peak). We calculated that the baseline noise (low frequencies) is 13.1 dB higher for the
medical electrode. As a result, a standard algorithm for the calculation of heartbeat frequency fares con-
siderably better with recordings from the textile electrode. On a subject with a mean heartbeat of 70 bpm
(at rest or during low-level activity), Fig.3c demonstrates that R peaks are more accurately detected in
recordings obtained with the textiles electrodes for a panel of dierent types of motion.
e long-term signal stability is assessed by continuously placing two textile electrodes on a volun-
teer’s chest during 3 days. ECG recordings from these electrodes are presented in Fig.3d and demon-
strate that the signals are highly consistent during this recording time. e SNR and R-Peak amplitude
evolutions are presented in Table S1. Despite the signal variations from day to day related with the skin
hydration changes and environmental noise, their amplitude and noise level remain stable. ere was
no skin reaction to the electrodes observed aer 3 days. Moreover, the same electrodes were stored in
ambient air for 1 month and then were re-used in ECG recordings in the same setup. ese electrodes
were still able to record well dened PQRST complexes, highlighting the long-term stability of the textile
PEDOT:PSS/IL gel electrodes.
Discussion
e technique discussed here allows the patterning of conducting materials with a demonstrated reso-
lution of 0.5 mm, a value that is adequate for the majority of envisioned biomedical applications. ese
include applications requiring small cutaneous electrodes, such as neonatal care and high-density elec-
troencephalography, where the feature size is larger than several millimeters. In the described patterning
process, we were able to combine the direct patterning with relatively thick knitted textiles. We have
benetted from the scalability of this approach and we adopted it for thick structured textiles in a time
and cost ecient way. is technique makes use of industrially accepted techniques such as contact
printing and deep painting, and can be easily applied not only during textiles manufacturing but also in
post manufacturing by processing existing garments. PDMS and PEDOT:PSS materials that are used in
the patterning are fully compatible with knitted textile platform thanks to their rheological properties.
e rubber-like PDMS stencil integrated with textile conserves the mechanical freedom of its structure.
Figure 2. Patterning results. (a) Flower pattern of PEDOT:PSS on a textile (in dark blue) and (b) a “lines
and spaces” test pattern with lines of 1, 2.5, 3 and 4 mm width, with a magnied view of the 1 and 2.5 mm
lines. (c) Photograph of a textile wristband with a PEDOT:PSS electrode. e electrode has an area of 1 cm2
which is also coated with an ionic liquid (IL) gel, indicated by PEDOT/PSS + IL gel arrow. A connection
pad is also visible and indicated by PEDOT:PSS arrow. (d) Impedance spectra of a textile and a medical
electrode of similar area measured in the 1 Hz to 1 kHz frequency range in contact with skin from three
dierent volunteers (mean (straight line) and standard deviation (condence area)).

www.nature.com/scientificreports/
5
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
e viscoelastic properties of PEDOT:PSS formulation allows to achieve homogeneous coating of elastic
and exible knitted textiles, in our case, polyester.
e combination of a low contact impedance provided by the PEDOT:PSS/IL gel electrode and of a
conformable support provided by the textile is shown here to diminish the impact of motion artifacts,
which paves the way for a variety of applications, including electromyography (EMG). EMG, which mon-
itors the electrical activity of skeletal muscles, requires electrodes that can record signals during motion.
e high tolerance of textile electrodes to low frequency motion artifacts makes them well-suited for
this application. Finally, electroencephalography (EEG) is another obvious application for these textile
electrodes. EEG measurements are currently performed using electrodes mounted on casks, then lled
with gel. e monolithic fabrication of a cask with integrated electrodes will make these measurements
easier to perform. Such measurements are currently ongoing in our own lab. As textile electrodes can
be easily integrated with hats, they can be used to render EEG electrodes imperceptible to the wearer.
As a result, this can increase the acceptance of EEG in applications beyond healthcare, such as gaming
and fatigue monitoring.
Existing commercial electrodes are assisted with hydrogels recognized as biocompatible systems.
However, some people manifest cutaneous reactions to such gels aer long contact. It’s very important to
underline that the skin condition as well as the skin sensitivity dramatically vary from subject to subject.
Chemical engineers are constantly pursuing the development of novel and more tolerant chemicals that
can be used for cutaneous electrophysiology. Ionic liquid gels are fully cross-linked systems providing
solid electrolytic contact with the skin required in cutaneous electrophysiology. Generally the toxicity
of ionically charged systems originates mainly from the cation. Novel formulations of ionic liquids with
large and bulky cations that are entrapped in polymeric matrix propose stable and biocompatible gel
systems. e process used in this paper can serve as a model for integrating such systems with conduc-
tive textiles. e toxicity of this material in not fully demonstrated yet. In our experiments we have not
observed any skin irritations during electrophysiological experiments performed over a 3 day period.
e technique developed here is generic and would work with any material soluble in aqueous media,
provided that post-processing (annealing, sintering, etc.) temperatures stay within the range that the
textile can support. e ability to pattern a second layer on top of the conducting polymer paves the
way for the development of a wide variety of devices, including organic electrochemical transistors
that can be used for simple logic circuits28, endowing textiles with signal processing capabilities. It also
Figure 3. Electrodes evaluation in electrocardiography. ECG recordings performed with the PEDOT:PSS
textile electrode (in blue), and a medical Ag/AgCl electrode (in red), (a) from volunteers sitting at rest,
(b) during movement. (c) Percentage of accuracy of heartbeat detection during dierent types of activity
(seating, standing up, leg moving, arm moving, walking) with medical and textile electrodes during a 50 s
epoch. (d) ECG signal evolutions obtained with textile electrodes in permanent contact with skin over three
days. e inset shows a picture of the skin under the electrode aer 72 h. e last ECG signals were obtained
from re-used textile electrodes stored in ambient air for one month.

www.nature.com/scientificreports/
6
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
includes biosensors, which, in the simplest conguration, consist of a conducting polymer and a redox
enzyme-containing gel. Such biosensors can be used for the detection of metabolites including glucose
and lactate in sweat29, thereby providing information about blood sugar level and muscle fatigue. In
addition to applications in healthcare (along the lines of a “smart” bandage), such sensors will extend the
scope of bioelectronic textiles to areas including sports and recreation. Finally, the integration of energy
and communication modules on textiles represents an important step in the evolution of this technology.
Components such as batteries and antennas, which can be made out of conducting polymers30, can be
patterned in a straightforward way with the technique developed here.
In conclusion, we developed a technique that allows the simple patterning of conducting polymers on
knitted textiles. e technique uses a PDMS stencil to conne the spreading of the polymer to dimen-
sions as small as 0.5 mm. PEDOT:PSS electrodes fabricated this way and coated with an ionic liquid gel
showed a low impedance contact with the skin. ey were able to record high quality electrocardiograms
in clinic and ambulatory conditions, and accurately determine heart rate, even when the wearer was in
motion. Moreover, these electrodes demonstrated high performance long-term stability during 3 days of
ECG recordings and aer extended ambient storage without any special reconditioning.
Methods
Patterning. e textile used was 100% interlock knit polyester fabric from VWR International (Spec-
Wipe® 7 Wipers). e Kapton (HN100) polyimide lm with 125 μ m thickness was provided by DuPont.
Laser cutting was achieved with a Protolaser S (LPKF) to pattern the polyimide mask. e PDMS formu-
lation (RTV615, elastomer and curing agent kit) was purchased from Momentive Performance Materials
and spin-coated on top of the polyimide at 550 rpm for 18 sec. e fabric was then gently transferred to
the mask coated with PDMS. Aer 10 minutes, PDMS was fully absorbed in the textile structure which
was then cured at 100 °C during 10 minutes for the rst annealing step. e polyimide mask was removed
before the deposition of the PEDOT:PSS solution. e second step of curing was applied to the fabric
at 110 °C during 1 hour. e conducting polymer formulation consisted of 80 mL of PEDOT:PSS disper-
sion (CleviosTM PH1000, Heraeus), 20 mL of ethylene glycol (Sigma Aldrich), 40 μ L of 4-dodecylben-
zenesulfonic acid (Sigma Aldrich), and 1 mL of 3-methacryloxypropyltrimethoxysilane (Sigma Aldrich).
e ionic liquid gel consisted of a mixture of the ionic liquid 1-ethyl-3-methylimidazolium-ethyl sulfate
(Sigma-Aldrich), poly(ethylene glycol)diacrylate and the photoinitiator 2-hydroxy-2-methylpropiophe-
none at a ratio of 0.6/0.35/0.05, respectively. For the electrode of Fig. 2b we pre-coated the textile with
20 μ l/cm2 of ionic liquid, then added 25 μ L/cm2 of ionic liquid gel formulation and exposed to UV light
(a UVGL-58 handheld UV Lamp set to 365 nm) to initiate crosslinking.
Electrical characterization and ECG recordings. All volunteers provided informed signed con-
sent to participate in the study. e sheet resistance of PEDOT:PSS on the textile was measured with
a four-probe set-up: the coated fabric was placed on top of four equidistant copper electrodes, a con-
stant current source was applied between the two outer probes. e voltage drop between two inner
probes was recorded to calculate sheet resistance. Impedance was measured in a 3 electrode congura-
tion (working and counter electrodes were placed two cm away from each other on the forearm, and
the reference electrode was placed 30 cm away on the arm) using Sensor N medical grade Ag/AgCl
electrodes (Ambu) with a 0.95 cm diameter gel-assisted area as reference and counter electrodes, as pre-
viously described by our group24. e textile electrode was compared to the Sensor N electrode, placed
2 cm apart. Spectra were acquired using an Autolab potentiostat, equipped with FRA module (Metrohm
B.V.), applying sinusoidal voltage of 0.01 V. For the ECG recordings, we used TE/K50430-001 medical
grade Ag/AgCl disk electrodes (Technomed Europe) with a 2 cm diameter. Textile and medical electrodes
were placed on the wrist and the ankle of a volunteer and connected to a SandsResearch system using
EA68 or EA136 ampliers during the 3 hours of evaluation sessions. Signals were processed and ltered
using LabVIEW (National Instruments) soware with a third-order Butterworth lter (passband with
low and high cuto of 0.5 Hz and 40 Hz, respectively). To extract baseline from ECG signals we used a
wavelet approach corresponding to a low-pass lter with a cuto of 1.93 Hz). e results in Fig.3c are
based on the signals presented in Fig.3a,b and three other motion conditions processed in the same
way. e algorithm for the calculation of heartbeat (R peak) frequency is a feature extractor LabVIEW
soware (available in BioMedical Toolkit from National Instrument) with intern lters between 10 and
25 Hz. ECG signals in Fig. 3d were recorded with a portable and wireless RF-ECG2 acquisition system
(from GM3 Corporation, intern passband with low and high cuto of 0.16 and 100 Hz, respectively).
e electrodes were in continuous contact with the skin during 3 days. ECG data were collected for
30 seconds every 3 hours during 3 days.
References
1. Cherenac, . & van Pieterson, L. Smart textiles: Challenges and opportunities. J. Appl. Phys. 112, 091301 (2012).
2. Zeng, W. et al. Fiber-Based Wearable Electronics: A eview of Materials, Fabrication, Devices, and Applications. Adv. Mater. 26,
5310–5336 (2014).
3. Zheng, Y. et al. Unobtrusive Sensing and Wearable Devices for Health Informatics. IEEE Trans. Biomed. Eng. 61, 1538–1554
(2014).
4. Zhang, Z., Shen, Y., Wang, W., Wang, B. & Zheng, J. Desig n and Implementation of Sensing Shirt for Ambulatory Cardiopulmonary
Monitoring. J. Med. Biol. Eng. 31, 207–216 (2011).

www.nature.com/scientificreports/
7
SCIENTIFIC RepoRts | 5:15003 | DOI: 10.1038/srep15003
5. Cho, G., Jeong, ., Pai, M., wun, Y. & Sung, M. Performance evaluation of textile-based electrodes and motion sensors for
smart clothing. IEEE Sens. J. 11, 3183–3193 (2011).
6. Coosemans, J., Hermans, B. & Puers, . Integrating wireless ECG monitoring in textiles. Sens. and Actuators A. 130–131, 48–53
(2006).
7. Hsu, Y.P. & Young, D. J. Sin-Coupled Personal Wearable Ambulatory Pulse Wave Velocity Monitoring System Using Microelec-
tromechanical Sensors. IEEE Sens. J. 14, 3490–3497 (2014).
8. Stoppa, M. & Chiolerio, A. Wearable Electronics and Smart Textiles: A Critical eview. Sensors. 14, 11957–11992 (2014).
9. Campana, A., Cramer, T., Simon, D.T., Berggren, M. & Biscarini, F. Electrocardiographic ecording with Conformable Organic
Electrochemical Transistor Fabricated on esorbable Bioscaold. Adv. Mater. 26, 3874–3878 (2014).
10. Taji, B., Shirmohammadi, S., Groza, V. & Batin, I. Impact of Sin–Electrode Interface on Electrocardiogram Measurements
Using Conductive Textile Electrodes. IEEE Trans. Instrum. Meas. 63, 1412–1422 (2014).
11. Lobodzinsi, S. ECG patch monitors for assessment of cardiac rhythm abnormalities, Prog. Cardiovas. Dis. 56, 224–229 (2013).
12. Yang, ., Freeman, C., Torah, ., Beeby, S. & Tudor, J. Screen Printed Fabric Electrode Array for Wearable Functional Electrical
Stimulation. Sens. and Actuators A. 213, 108–115 (2014).
13. Paul, G. M. et al. A Smart Textile Based Facial EMG and EOG Computer Interface. IEEE Sens. J. 14, 393–400 (2013).
14. Taamatsu, S., Yamashita, T., Imai, T. & Itoh, T., Lightweight exible eyboard with a conductive polymer-based touch sensor
fabric. Sens. and Actuators A. 220, 153–158 (2014).
15. Castano, L. M. & Flatau, A. B. Smart fabric sensors and e-textile technologies: a review. Smart Mater. Struct. 23, 053001 (2014).
16. Cosseddu, P., Mattana, G., Orgiu, E. & Bonglio, A. Ambipolar organic eld-eect transistors on unconventional substrates. Appl.
Phys. A 95, 49–54 (2009).
17. ane, S.B., Seth, T., Phata, G.J., Amalnerar, D.P. & Ghatpande, M., Eect of inorganic binders on the properties of silver thic
lms. J. Mater Sci. Mater. El. 15,103–106 (2004).
18. Seine, T. et al. Improvement of mechanical durability on organic TFT with printed electrodes prepared from nanoparticle in.
Appl. Surf. Sci. 294, 20–23 (2014).
19. Charlot, B. et al. Micropatterning PEDOT:PSS layers. Microsyst. Technol. 19, 895–903 (2013).
20. Hamedi, M., Forcheimer, . & Inganas, O. Towards woven logic from organic electronic bres. Nat. Mater. 6, 357–362 (2007).
21. Marculescu, D. et al. Electronic textiles: a platform for pervasive computing. Proc. IEEE 91, 1995–2018 (2003).
22. Post, E., Orth, M., usso, P. & Gershenfeld, N. E-broidery: Design and fabrication of textile-based computing. IBM Syst. J. 39,
840–860 (2000).
23. Svennersten, ., Bolin, M. H., Jager, E. W. H., Berggren, M. & ichter-Dahlfors, A. Nano-ber scaold electrodes based on
PEDOT for cell stimulation. Biomaterials, 30, 6257–6264 (2009).
24. Leleux, P. et al. Ionic Liquid Gel-Assisted Electrodes for Long-Term Cutaneous ecordings. Adv.Healthcare Mater. 3, 1377–1380
(2014).
25. ossmann et al. Members of the Subcommittee on Instrumentation. ecommendations for standardization of leads and of
specications for instruments in electrocardiography and vectorcardiography, eport of Committee on Electrocardiography, American
Heart Association. 35, 583–602 (Circulation 1967).
26. Cliord, G., Azuaje, F. & McSharry, P. in Advanced Methods and Tools for ECG Analysis. Ch. 3, 70 (Artech House Inc., 2006).
27. öhler, B-U., Hennig, C. & Orglmeister, . e principles of soware QS detection. IEEE. Eng Med. Biol. Mag. 21, 42–57 (2002).
28. Nilsson, D., obinson, N., Berggren, M. & Forchheimer, . Electrochemical logic circuits. Adv. Mater. 17, 353 (2005).
29. hodagholy, D. et al. Organic electrochemical transistor incorporating an ionogel as a solid state electrolyte for lactate sensing.
J. Mater. Chem. 22, 4440–4443 (2012).
30. Berggren, M., Nilsson, D. & obinson, N. Organic materials for printed electronics, Nature Materials. 6, 3–5 (2007).
Acknowledgements
Partial funding for this work was obtained from ANR, Region PACA, and MicroVitae Technologies.
D.K was an international REU student from the University of Michigan, supported by the US National
Nanotechnology Infrastructure Network.
Author Contributions
S.T. and E.I. developed the patterning technique, E.I., T.L. and D.C. developed the IL coating, T.L., S.T.,
E.I. and J.M.B. performed the ECG measurements, T.L. and S.T. performed the ECG data analysis and
prepared Figures 1 and 3, G.G.M. and E.I. supervised the project, S.T., T.L., E.I. and G.G.M. wrote the
paper. All authors reviewed the manuscript.
Additional Information
Supplementary information accompanies this paper at http://www.nature.com/srep
Competing nancial interests: e authors declare no competing nancial interests.
How to cite this article: Takamatsu, S. et al. Direct patterning of organic conductors on knitted textiles
for long-term electrocardiography. Sci. Rep. 5, 15003; doi: 10.1038/srep15003 (2015).
is work is licensed under a Creative Commons Attribution 4.0 International License. e
images or other third party material in this article are included in the article’s Creative Com-
mons license, unless indicated otherwise in the credit line; if the material is not included under the
Creative Commons license, users will need to obtain permission from the license holder to reproduce
the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/