Fully Inkjet-Printed Flexible Temperature Sensors Based on Carbon ...

Available online at www.sciencedirect.com

ScienceDirect Materials Today: Proceedings 3 (2016) 739 – 745

12th International Conference on Nanosciences & Nanotechnologies & 8th International Symposium on Flexible Organic Electronics

Fully Inkjet-Printed Flexible Temperature Sensors Based on Carbon and PEDOT:PSSÕ C. Balia,b,*, A. Brandlmaiera, A. Ganstera, O. Raaba, J. Zapfa, A. Hüblerb a

Siemens AG, Corporate Technology, Research and Technology Center, CT RTC ELE SPT-DE, 81739 Munich, Germany b Institute for Print and Media Technology, Chemnitz University of Technology, 09107 Chemnitz, Germany

Abstract Printed organic sensors on plastic substrates are of great interest due to their flexibility and low-cost fabrication. We here present a fully inkjet-printed flexible Wheatstone bridge used for temperature sensing. A carbon nanoparticle ink and a mixture of poly(3,4-ethylenedioxythophene):poly(4-styrene-sulfonate) (PEDOT:PSS) and dimethyl sulfoxide (DMSO) are used as positive (PTC) and negative (NTC) thermal coefficient materials, respectively. Different carbon inks and DMSO concentrations from 0.5 to 40 wt% are evaluated in order to optimize the temperature coefficient of resistance (TCR) and the reliability of the PTC and NTC materials. The single materials are printed on PEN substrates, and subsequently encapsulated. A climate chamber is used to evaluate their reliability. The Wheatstone bridge sensors show good linearity and little hysteresis within a temperature range of 20°C to 70°C and exhibit a sensitivity of ~4 mV/°C at a bias current of 1 mA. We furthermore investigate the reproducibility of the temperature sensor fabrication in order to analyze the potential of industrial applicability of inkjet-printing. © 2015 The Authors. Published by Elsevier Ltd.

© 2016 Elsevier Ltd. All rights reserved. Selectionand andpeer-review peer-review under responsibility ofConference the Conference Committee Members of NANOTEXNOLOGY2015 (12th Selection under responsibility of the Committee Members of NANOTEXNOLOGY2015 Conference onon Nanosciences International Symposium on Flexible International (12th International Conference Nanosciences & & Nanotechnologies Nanotechnologies & & 8th8th International Symposium on Flexible OrganicOrganic Electronics)

Electronics). Keywords: Printed electronics, Organic electronics, Printed temperature sensors; Wheatstone bridge; Carbon; PEDOT:PSS; Inkjet.

Õ

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited. * Corresponding author. Tel.: +49-176-23545878; fax: +49-89-636-48555. E-mail address: [email protected]

2214-7853 © 2016 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the Conference Committee Members of NANOTEXNOLOGY2015 (12th International Conference on Nanosciences & Nanotechnologies & 8th International Symposium on Flexible Organic Electronics) doi:10.1016/j.matpr.2016.02.005

740

C. Bali et al. / Materials Today: Proceedings 3 (2016) 739 – 745

1. Introduction The interest in flexible sensors is increasing continuously due to their different applications e.g. in the field of wearable technologies such as healthcare [1–3] and robotics [4,5]. Printed electronics is proven to be a suitable technology for the realization of flexible sensors thanks to its low-cost manufacturing, rapid processing and additionally the flexibility and low weight of the printed devices. Printed temperature sensors have been realized by the use of different printing techniques such as inkjet [6–9], gravure [10], screen printing [11] and dip-coating [12]. For this purpose, solution-processed materials, which exhibit a temperature dependence of electrical resistance with positive (PTC) and negative (NTC) temperature coefficients were applied. In addition to good reliability and reproducibility of the printed temperature sensors, a good linearity of the temperature dependent resistance and a high sensitivity are generally required. For printed PTC sensors, a TCR between 0.0006/°C and 0.0021/°C was reported for inkjet-printed silver [6–8] and a higher value of 0.0030/°C was reported for inkjet-printed Cellulose Acetate Butyrate [9]. For the realization of NTC sensors, PEDOT:PSS was gravure-printed by B. Meier et al. [10] and showed a TCR of -0.0023/°C with a significant time- and temperature-dependent drift of the resistance. K. Kanao et al. [13] mixed PEDOT:PSS with carbon nanotubes (CNTs) and reported a higher absolute TCR of 0.0078/°C compared to pure PEDOT:PSS. The suitability of pure CNTs for the use as NTC-material was investigated by M. Sibinski et al. [12], who measured a comparably low TCR of -0.0013/°C with a significant hysteresis of the temperature dependent resistance. Contrarily, D.T. Britton et al. developed an NTC silicon nanoparticle based ink with a high absolute TCR of 0.0200/°C and a good linearity [14,15]. In order to enhance the sensitivity of the temperature sensors, it is also possible to combine NTC and PTC materials in a Wheatstone bridge configuration. For example, A. Aliane et al. [11] developed specific inks based on graphite-PDMS and Antimony Tin Oxide as PTC and NTC materials, respectively and screen-printed them in order to realize a Wheatstone bridge with a sensitivity of 60 mV/°C at a bias voltage V0 = 4.8 V. In this paper, we present a fully inkjet-printed temperature sensor based on a combination of NTC and PTC materials in a full Wheatstone bridge configuration (see Figure 1a). For this purpose, low-cost commercial inks are printed on flexible PEN (Polyethylene Naphthalate) substrates and encapsulated subsequently in order to optimize their reliability. The temperature dependent output voltage of the bridge showed good linearity, good repeatability and a sensitivity of ~4 mV/°C at a bias current I0 = 1mA. 2. Experimental 2.1. Materials For the deposition of the single temperature sensor materials, a piezoelectric Drop-on-Demand (DOD) inkjet printer DMP 2800 was employed. The NTC ink used for this purpose consists of a water based solution of PEDOT:PSS (CLEVIOSTM P JET OLED). In order to tune the conductivity and the TCR of the NTC ink, we mixed PEDOT:PSS with 0.3 to 40 wt% DMSO. I0

a

b

PTC

NTC

R2

R1

R4

R3

Vout

V0 NTC

PTC 5 mm

Fig. 1. Temperature sensor element in Wheatstone bridge configuration. a) Circuit design with bias current I0, bias voltage V0 and output voltage Vout. b) Wheatstone bridge printed on PEN substrate with PTC resistors (Carbon, R2 and R3) and NTC resistors (PEDOT:PSS/DMSO, R1 and R4). Electrical contact pads and connections are printed with a silver nanoparticle ink.


For the PTC ink, three commercial water-based carbon nanoparticle inks from Methode were evaluated. The contact pads and the electrical connections of the Wheatstone bridge were realized with an alcohol-based silver nanoparticle ink Cabot CCI-300 with a silver solid loading of 19 – 21 wt%. The single materials were printed on the adhesion promoting pretreated side of a PEN substrate TEONEX® Q65FA from Teijin DuPont Films. The final process step consisted of encapsulating the Wheatstone bridge with a barrier foil (moisture vapor transmission rate (VWTR) ≤ 10-3 g/m2/day) attached to an adhesive film. 2.2. Printing of the single sensor elements The single inks were filtered and filled into DMC-11610 cartridges equipped with 10 pl printing heads. 1 μm nylon filters were used for the PTC ink and 0.45 μm nylon filters were used for silver and NTC inks. All the printing steps were performed at ambient conditions with a fixed chuck temperature of 28°C. A drop spacing of 25 μm and a jetting frequency of 2 kHz were used. Voltage values of around 30 V and waveforms were adjusted before each printing step in order to optimize the jetting behavior. In order to characterize the temperature dependent behavior of the single PTC and NTC sensing elements, the materials were printed separately connected with silver electrodes with the same layout and processing parameters used for the Wheatstone bridge. 2.3. Printing of the Wheatstone bridge Figure 1b shows the printed Wheatstone bridge with R1, R4 as NTC sensor elements and R2, R3 as PTC sensor elements. The first printed layer consists of the contact pads and electrical connections, which were printed with the silver nanoparticle ink and subsequently sintered at 150°C for 60 min. The following step consisted of printing of the PEDOT:PSS-DMSO mixture cured at 150°C for 30 min. Finally, the nanoparticle carbon ink were printed and sintered at 150°C for 10 min. All the curing steps were realized in an air convection oven. 2.4. Characterization techniques The electrical characterization of the single sensor elements was realized by measuring the resistance at temperatures between 20°C and 70°C in 5°C steps for 2 consecutive temperature sweeps. The resistance values were measured at each temperature step with a Keithley 2400 SourceMeter at a bias current of 0.1 mA. For the electrical characterization of the Wheatstone bridge, a bias current of 0.1 mA or 1 mA was applied and the output voltage Vout was measured. Similarly to the characterization of the single sensor elements, 2 consecutive temperature sweeps between 20°C and 70°C in 5°C steps were performed and the output voltage of the bridge was measured at each temperature step. The reliability of the encapsulated samples was characterized by measuring the resistance within a damp heat accelerated lifetime test in a climate chamber at 65°C and 85% relative humidity (RH) for 400 or 600 hours. 3. Results and discussion 3.1. NTC sensor elements In order to optimize the NTC material regarding its conductivity, we varied the DMSO concentration in the PEDOT:PSS-DMSO mixture. As illustrated in Figure 2, adding a high boiling point solvent such as DMSO to PEDOT:PSS leads to a structural change of PEDOT:PSS. According to Q. Wei et al. [16], this change is due to a higher interaction of conductive PEDOT chains and a higher PEDOT to PSS ratio, which results in higher conductivity. The conductivity enhancement was also explained by J. Y. Kim et al. as an influence of the dielectric constant of the remaining high boiling point solvent after curing PEDOT:PSS, which induces a screening effect between PEDOT and PSS chains, resulting in a reduction of the Coulomb interaction between them [17]. A further explanation was provided by J. Ouyang et al. [18], who demonstrated a conformational change of the PEDOT-chain from benzoid to quinoid structure after adding a high boiling point solvent to PEDOT:PSS. In this paper, we use DMSO not only to change the resistivity of PEDOT:PSS but also to tune its TCR. Moreover, adding a high boiling point solvent to PEDOT:PSS reduces the coffee ring effect by inducing Marangoni

741

742


DMSO

PEDOT

+DMSO

PSS Fig. 2. Structural change of PEDOT:PSS by adding a high boiling point solvent (DMSO): Higher interchain interaction of PEDOT and higher PEDOT to PSS ratio [16].

flows [19] leading to a better uniformity of the printed structures and thus to a better control of the printing process. Figure 3a shows the relative resistance as a function of temperature for DMSO concentrations between 0.3 wt% and 40 wt%. The curves show little hysteresis for all DMSO concentrations and a good repeatability of the resistance measurements for consecutive temperature cycles. As evident, the slope and hence the absolute TCR decrease with increasing DMSO concentration. For concentrations higher than 2 wt%, the curves exhibit almost the same slope, which means that a saturation of the TCR value occurs in between 2 wt% and 40 wt%. The TCR as a function of the DMSO concentration extracted from the graphs in Figure 3a is illustrated in Figure 3b showing a variation within 0.0090/°C and 0.0025/°C for concentrations between 0.3 wt% and 2 wt%. For concentrations higher than 2 wt%, the absolute TCR saturates at ~0.0025/°C.

0.9

− −3 2–40

0

1.0

0.8

R/R

0.5

R/R0

0.7

10

20

30

40

50

T (°C) T (°C)

60

70

−4 − /°C) C

-3

−5 − −6 − − −7

TCR TC C (× 10 −8 −

0.4 0.3

0.6

−2

b

Increasing DMSO concentration [wt%]

1.0

TCR (× 10-3/°C)

a

−9 80

0

10

20

30

40

50

DMSO t.%) DMSOconcentration concentration(w (wt%)

Fig. 3. a) Normalized resistance vs. temperature for PEDOT:PSS NTC sensors for varying DMSO concentration (0.3, 0.4, 0.5, 1, 2, 3, 5, 10, 20, 30, and 40 wt%) for two consecutive temperature cycles. R0 corresponds to the resistance measured at 20°C. b) TCR vs. DMSO concentrations extracted from a.

Moreover, we optimized the reliability by encapsulating the NTC sensor elements with a barrier foil. For this purpose, we measured the resistance of encapsulated and unencapsulated samples with a DMSO concentration of 10 wt% within a damp heat accelerated lifetime test for 400 hours at 65°C and 85% RH. As illustrated in Figure 4a, the relative resistance of encapsulated samples increased by 10% compared to 130% for the sample without encapsulation. This shows that the barrier foil leads to a significant increase in stability. We furthermore evaluated the stability of the PTC sensor elements for different DMSO concentrations. To this end, the same lifetime test was realized for encapsulated NTC elements with DMSO concentrations of 0.5 wt%, 1 wt% and 10 wt% for a duration of 600 hours. The graphs in Figure 4b show a similar aging behavior for the samples with 1 wt% and 10 wt%. The samples with 0.5 wt% however exhibited a significant unstable behavior, which was also observed for further samples. Samples with low DMSO concentration tended to irreversibly turn highly resistive. Therefore, we chose a DMSO concentration of 10 wt% as a trade-off between the value of the TCR and the reliability. 3.2. PTC sensor elements The PTC sensor elements were realized by printing, curing and encapsulating resistors based on three different carbon nanoparticle inks. The relative resistance is illustrated in Figure 5a and shows a TCR of 0.0022/°C,

743

C. Bali et al. / Materials Today: Proceedings 3 (2016) 739 – 745 1.30

b

2.20

0.5 wt%

1.25

unencapsulated

2.00

1.20

1.80

0

0

1.15

1.60

R/R

R/R

10 wt%

1.10

R/R0

1.40

1.20

encapsulated

1.05

1.00

R/R0

a

1 wt%

1.00 0

100

200

300

400

0

100

200

300

400

500

600

tim e (h) (h) time

time (h) time (h)

Fig. 4. Damp heat accelerated lifetime tests at 65°C and 85% RH. Relative resistance (R0 corresponds to the resistance measured at time = 0) for a) encapsulated and unencapsulated PEDOT:PSS NTC sensors mixed with 10 wt% DMSO and b) PEDOT:PSS NTC sensors with increasing DMSO concentration (0.5, 1, and 10 wt%, two samples for each concentration are displayed).

0.0010/°C, and -0.0002/°C for inks Ink1, Ink2 and Ink3, respectively. Therefore, Ink1 and Ink2 can be utilized as PTC materials. In the following, analogously to the NTC materials, a damp heat accelerated lifetime test was performed by exposing two encapsulated samples of each ink to 65°C and 85% RH for 600 hours. The graphs in Figure 5b show a change of relative resistance after 600 hours of -8%, 19% and 30% for Ink1, Ink2 and Ink3, respectively. Therefore, we conclude that among the tested inks Ink1 exhibits the highest stability and the highest temperature sensitivity due to its comparably high TCR. a 1.15

b Ink1

1.3 1.2

1.10

0 1.1

0

Ink2

R/R0

1.00

0.95 10

Ink1

R/R R/ /R 1.0

R/R

Ink3

0.9

R/R0

1.05

Ink2

Ink3

0.8 20

30

40

50

(°C) TT (°C)

60

70

80

0

100 200 300 400 500 600

tim (h) timee(h)

Fig. 5. Evaluation of three different carbon nanoparticle PTC inks after printing and sintering at 150°C for 10 min. a) Normalized resistance R/R0 vs. temperature T for two consecutive temperature cycles. R0 corresponds to the resistance measured at 20°C. b) Normalized resistance for encapsulated carbon PTC sensors (two samples of each ink) during a damp heat accelerated lifetime test at 65°C and 85% RH. R0 corresponds to the resistance measured at time = 0.

3.3. Wheatstone bridge After the characterization of the single NTC and PTC elements, a Wheatstone bridge fabricated with the characterized materials (PEDOT:PSS with 10wt% DMSO for NTC and Ink1 for PTC) is printed on PEN, encapsulated, and characterized. The output voltage of the bridge as a function of temperature is plotted in Figure 6a for bias currents 0.1 mA and 1 mA. The graphs show good linearity, little hysteresis and sensitivities of 0.4 mV/°C and 3.9 mV/°C for bias currents of 0.1 mA and 1 mA, respectively. In order to evaluate the reproducibility of the temperature sensor fabrication, we show the output voltage of three identically produced Wheatstone bridges for two consecutive temperature sweeps at a bias current of 0.1 mA. As shown in Figure 6b, the performance of the three bridges coincide very well, which evidences a good reproducibility of the Wheatstone bridge fabrication process.

744


200.0

a

b

20

W-Bridge 3

I0 = 1 mA

150.0

10

100.0

W-Bridge 1 I0 = 0.1 mA

V (mV)

50.0

0.0 0 0

10

20

30

40

50

TT (°C) (°C)

60

70

80

V (m V) 0

V (mV)

V (m ( V)

−10 10

W-Bridge 2

20 30

40 50

60 70

80

TT (°C)

Fig. 6. a) Output voltage of the Wheatstone bridge at different bias currents of 0.1 mA and 1 mA exhibiting sensitivities of ~0.4 mV/°C and ~4 mV/°C, respectively. b) Output voltages of three identically fabricated Wheatstone bridges at a bias current of 0.1 mA.

4. Conclusion In this work, we demonstrated the fabrication of temperature sensors in Wheatstone bridge configuration by inkjet printing PEDOT:PSS-DMSO and Carbon as NTC and PTC resistors, respectively. The concentration of DMSO in the PEDOT:PSS based NTC sensor elements was optimized in order to achieve high sensitivity, good linearity and good repeatability. The results showed a TCR saturation for DMSO concentrations higher than 2 wt%, which corresponds to a value of -0.0025/°C. A concentration of 10 wt% was chosen as a trade-off between TCR and reliability determined by a damp heat accelerated lifetime test. Analogously, three nanoparticle carbon based inks were characterized by evaluating the TCR and the stability within damp heat accelerated lifetime tests. Furthermore, we evidenced that an adequate encapsulation significantly increases the reliability. The sensitivity of the Wheatstone bridge is proportional to the bias current and is equal to 3.9 mV/°C for a bias current I 0 = 1 mA. Moreover, the output voltage of the printed Wheatstone bridge vs. temperature showed a good linearity, little hysteresis. The sensor fabrication exhibits good reproducibility, which therefore demonstrates the potential of the industrial applicability of printed temperature sensors. Acknowledgements This work was realized in the context of the EU-project FLEXIBILITY (Project No. FP7-287568, www.flexibility-fp7.eu). FLEXIBILITY aims to significantly advance the competitiveness of Europe in the area of multifunctional, ultra-lightweight, ultra-thin and bendable thin film and organic large area electronic systems. References [1] A. Pantelopoulos, N. G. Bourbakis, IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews 40, 2010, pp. 1–12. [2] Y.-D. Lee, W.-Y. Chung, Sensors and Actuators B: Chemical 140, 2009, pp. 390–395. [3] S. Patel, H. Park, P. Bonato, L. Chan, M. Rodgers, Journal of NeuroEngineering and Rehabilitation 9, 2012, pp. 1–17. [4] C. Zhu, W. Sheng, IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans 41, 2011, pp. 569–573. [5] C. Zhu, W. Sheng, IEEE International Conference on Robotics and Automation, 2009, ICRA'09, 2009. [6] J. Felba, K. Nitsch, T. Piasecki, P. Paluch, A. Moscicki, A. Kinart, 9th IEEE Conference on Nanotechnology, IEEE-NANO 2009, 2009, pp. 408–411. [7] J. Courbat, Y.B. Kim, D. Briand, N. F. de Rooij, In Solid-State Sensors, Actuators and Microsystems Conference (TRANSDUCERS), 2011 16th International, IEEE, 2011, pp. 1356–1359. [8] F. Molina-Lopez, A. V. Quintero, G. Mattana, D. Briand, N.F. de Rooij, Proceedings IMCS 2012, 2012, pp. 1122–1125.

C. Bali et al. / Materials Today: Proceedings 3 (2016) 739 – 745 [9] G. Mattana, T. Kinkeldei, D. Leuenberger, C. Ataman, J.J. Ruan, F. Molina-Lopez, A. V. Quintero, G. Nisato, G. Tröster, D. Briand, N.F. de Rooij, Sensors Journal, IEEE 13, 2013, pp. 3901–3909. [10] B. Meier, L. Egermann, S. Voigt, M. Stanel, H. Kempa, A.C. Huebler, Thin Solid Films 519, 2011, pp. 6610–6612. [11] A. Aliane, V. Fischer, M. Galliari, L. Tournon, R. Gwoziecki, C. Serbutoviez, I. Chartier, R. Coppard, Microelectronics Journal 45, 2014, pp. 1621–1626. [12] M. Sibinski, M. Jakubowska, M. Sloma, Sensors 10, 2010, pp. 7934–7946. [13] K. Kanao, S. Harada, Y. Yamamoto, W. Honda, T. Arie, S. Akita, K. Takei, RSC Advances, 5, 2015, pp. 30170–30174. [14] D. T. Britton, M. Härting, Pure and applied chemistry, 78, 2006, pp. 1723–1739. [15] M. Ulrich, A. Chuvilin, B. Magunje, E.O. Jonah, H. Margit, D. T. Britton, Japanese Journal of Applied Physics 52, 05DA11, 2013. [16] Q. Wei, M. Mukaida, Y. Naitoh, T. Ishida, Advanced materials 25, 2013, pp. 2831–2836. [17] J. Y. Kim, J.H. Jung, D.E. Lee, J. Joo, Synthetic Metals 126, 2002, pp. 311–316. [18] J. Ouyang, Q. Xu, C. W. Chu, Y. Yang, G. Li, J. Shinar, Polymer 45, 2004, pp. 8443–8450. [19] J. Park, J. Moon, Langmuir 22, 2006, pp. 3506–3513.

745