Electrolyte-gated organic field-effect transistors for

Electrolyte-gated organic field-effect transistors for sensing applications F. Buth, D. Kumar, M. Stutzmann, and J. A. Garrido Citation: Appl. Phys. Lett. 98, 153302 (2011); doi: 10.1063/1.3581882 View online: http://dx.doi.org/10.1063/1.3581882 View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v98/i15 Published by the AIP Publishing LLC.

Additional information on Appl. Phys. Lett. Journal Homepage: http://apl.aip.org/ Journal Information: http://apl.aip.org/about/about_the_journal Top downloads: http://apl.aip.org/features/most_downloaded Information for Authors: http://apl.aip.org/authors

Downloaded 26 Sep 2013 to 221.130.23.116. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions

APPLIED PHYSICS LETTERS 98, 153302 共2011兲

Electrolyte-gated organic field-effect transistors for sensing applications F. Buth, D. Kumar, M. Stutzmann, and J. A. Garridoa兲 Walter Schottky Institut, Technische Universität München, Am Coulombwall 4, 85748 Garching, Germany

共Received 4 February 2011; accepted 31 March 2011; published online 15 April 2011兲 We report on the electrolytic gating of ␣-sexithiophene thin film transistors, in which the organic semiconductor is in direct contact with an electrolyte. Due to the large capacitance of the electrical double layer at the electrolyte/semiconductor interface, modulation of the channel conductivity via an electrical field effect is achieved at low voltages. The transistors are stable for several hours and are sensitive to variations in the pH resulting from a pH-dependent surface charge, which modulates the threshold voltage. The response to different ion concentrations is described by the influence of the ions on the mobility and an electrostatic screening effect. © 2011 American Institute of Physics. 关doi:10.1063/1.3581882兴

a兲

Electronic mail: [email protected].

0003-6951/2011/98共15兲/153302/3/$30.00

200 nm兲 contacts. Prior to the deposition of the organic semiconductor, the contacts were covered with a chemically stable photoresist 共SU-8兲, leaving only a 20 ␮m long area of gold exposed at the channel 关see Fig. 1共a兲兴. 80 nm thin ␣6T films 共Sigma Aldrich兲 were deposited at a background pressure of around 10−8 mbar with a deposition rate of 0.6 Å/min and a substrate temperature of 45 ° C. For measurements, the samples were mounted onto a ceramic chip carrier and all remaining exposed metal parts were covered with silicone rubber. Unless otherwise specified, all measurements were performed in an aqueous electrolyte containing 10 mM K-based phosphate buffer saline 共PBS兲, adjusted to an ionic strength of 50 mM with KCl. The gate voltage Usg was applied between a Ag/AgCl reference electrode and the source contact of the transistor. In the following, Usg will always be given with respect to the Ag/AgCl reference electrode. Prior to the transistor characterization, the ␣6T/ electrolyte interface was investigated by electrochemical impedance spectroscopy in a three-electrode configuration. For this purpose, ␣6T was evaporated onto a polycrystalline gold electrode. As a reference, a bare gold electrode was characterized. Figure 1共b兲 shows the frequency dependence of the effective capacitance and the phase of the impedance. The effective capacitance was calculated assuming a simple equivalent circuit, consisting of a capacitor in parallel with a resistor. As indicated by a phase around 80°, the impedance is mainly of capacitive nature for frequencies below 103 Hz. The measured double-layer capacitances, CAu ⬇ 20 ␮F / cm2 for the Au-reference and C6T = 2 – 8 ␮F / cm2 for ␣6T, agree well with literature values for gold and hydrophobic semiconductors in contact with aqueous electrolytes.15,16 The Ag/AgCl reference electrode Usg

electrolyte

sexithiophene substrate Ti/Au contacts

(b)

Au reference

-20

10-5

-40 6T on Au

10-6

-60

resist

-80 10-7

10-1

Uds

0

Phase (°)

(a)

Ceff (F/cm2)

Because of their potential as disposable sensors in health care and food monitoring applications, organic materials have recently been studied as the active element in chemical or biological sensors.1–5 It was shown that, even if the organic layer is in direct contact with an electrolyte, organic field effect transistors 共OFETs兲 are stable and very sensitive to analytes in the media.4,6 However, it is crucial that these devices can be operated at low voltages due to limitations imposed by the electrolytic environment. To this end, high-k dielectrics,7 ultrathin organic layers,8 or electrolytes9–11 have been investigated as substitutes for the conventional oxide gate insulators. The high electrical double layer 共EDL兲 capacitance formed at an electrolyte/organic semiconductor interface enables the operation of devices below 1 V. This working principle is also commonly used in solution-gated field-effect transistors 共SGFETs兲 with inorganic semiconductors like carbon nanotubes,12 diamond,13 and graphene.14 In all these cases the active material is inert at the potentials applied across the electrolyte/semiconductor interface. In the case of organic semiconductors, electrochemical doping can lead to a modulation of the bulk conductivity of the organic film.2 However, modulation of the conductivity via a pure electric field effect is favorable under certain conditions, as recently demonstrated for rubrene single crystals and spin-coated poly共3-hexylthiophene兲 films using pure water as the electrolyte.9 In contrast to the chemical doping, the electrical field across the EDL modulates only the conductivity in close proximity to the interface with the electrolyte. Such devices can be used to detect changes in the chemical and charge composition of the electrolyte, enabling chemical, and biosensing. In order to investigate the suitability of organic semiconductors for SGFETs, we chose ␣-sexithiophene 共␣6T兲 since it was previously shown to be stable in an aqueous environment.6 The ␣6T SGFETs were fabricated on oxygen-terminated diamond substrates 共Diamond Detectors Ltd兲 with a roughness below 0.2 nm in order to ensure good crystal growth and avoid any parasitic Faradaic currents from the substrate. Transistors with channel dimensions 共length⫻ width兲 of 20 ⫻ 200 ␮m2 were defined using conventional photolithography, followed by the thermal evaporation of Ti/Au 共20 nm/

101

103

105

Frequency (Hz)

FIG. 1. 共a兲 Transistor layout. 共a兲 Effective capacitance extracted from impedance spectroscopy measurements on an electrolyte/␣6T/Au stack compared to the capacitance of an electrolyte/Au reference. The frequencydependent phase of the impedance is also shown.

98, 153302-1

© 2011 American Institute of Physics


(a)

Appl. Phys. Lett. 98, 153302 共2011兲

Buth et al.

(b)

Usg= 0.3 V to 0.6 V

Uds=-0.4V

20

(a)

-30

Ids/Ids(pH = 5.4)

Ids/Ids(off)

Ids (nA)

-20

-10

10

5

0 -0.2

-0.4

-0.6

1

60

120

180

high capacitance of the EDL results in a large transconductance and, thus, enables the operation of the device at low voltages. The capacitive behavior at low frequencies was confirmed by cyclic voltammetry measurements 共data not shown兲, which exhibit no Faradaic peaks in the potential region below 0.6 V versus Ag/AgCl. Thus, electrochemical oxidation of the organic semiconductor can be excluded at these voltages. The literature value for the oxidation of solid films of ␣6T is 0.85 V versus Ag/AgCl.17 The characteristics of the device, as shown in Fig. 2共a兲, exhibit the typical behavior of a FET, with a clear saturation regime at higher drain-source voltages and a linear region, where the drain-source current can be described by the equation

冋

册

1 2 ␮CW 共Usg − Ut兲Uds − Uds , L 2

共1兲

where ␮ is the field-effect mobility of carriers in the thin film, Uds the drain-source voltage, Ut the threshold voltage, and C the double layer capacitance. Using Eq. 共1兲, we can estimate field-effect mobilities of up to 2.2⫻ 10−2 cm2 / V s for our devices. Roberts et al.6 reported a mobility of 1.8 ⫻ 10−1 cm2 / V s for bottom-gated ␣6T transistors immersed in water. This difference might be caused by the rougher ␣6T/electrolyte interface at which the conductive channel is formed in our case. The comparably low mobility was also observed for water-gated devices in the work of Kergoat et al.9 Even though the organic film was directly exposed to the electrolyte, the devices are quite stable, as shown in Fig. 2共b兲. The drain-source current decreased by roughly 10% during continuous cycling of the source-gate voltage for more than 3 h, which is enough for the envisioned application as disposable sensing devices. Figure 3共a兲 shows the dependence of the drain-source current on the electrolyte pH, revealing a decrease in Ids with decreasing pH. No sign of hysteresis was observed when subsequently increasing the pH again. At pH values higher than 7 the devices are unstable, the reason of which is currently under investigation. This experiment was conducted in a 10 mM PBS-buffered solution with the ionic strength adjusted to 100 mM with KCl. This high ionic strength was at least one order of magnitude higher than any change in ionic strength due to the addition of the 0.2 M HCl or KOH solutions used to adjust the pH. The resulting pH sensitivity 共change in gate voltage against pH兲 was around 9 mV/pH versus Ag/AgCl. As shown in Fig. 3共b兲 this decrease in Ids is

pH7 pH5

Uds= -0.1 V

pH2

-5

0.6

2

4

pH

Time (min)

FIG. 2. 共a兲 Transistor characteristics of an electrolyte-gated ␣6T OFET. 共b兲 Recording of the drain-current during continuous cycling of the device between its on 共Usg = 0.6 V兲- and off-state 共Usg = 0.3 V兲.

lin = Ids

-10

Usg= 0.6 V

0.8

0.4

0

Uds (V)

(b)

Uds= -0.1 V 1.0

15

0 0.0

1.2

Ids (nA)

153302-2

6

0

0.3

0.4

0.5

0.6

0.7

Usg(V)

FIG. 3. 共a兲 Dependence of the drain-source current on the pH of the electrolyte measured for decreasing 共solid symbols兲 and increasing pH 共open squares兲. The solid line shows the current simulated using the amphifunctional model with the following parameters: C0 = 2 ␮F / cm2, C1 = 20 ␮F / cm2, the surface site density used is 3 ⫻ 1013 cm−2. The acidity and alkalinity constants were 1 ⫻ 10−4 and 1 ⫻ 10−3, respectively. 共b兲 Drain current vs gate voltage measured in electrolytic solutions with different pH. The lines show fits to the linear region.

due to a shift in threshold voltage and not due to a decreasing mobility, which would result in a variation in the slope of the Ids − Usg curves. This can be interpreted considering the effect of a pH-dependent surface charge, rather than by the introduction of traps due to diffusion of hydronium or hydroxide ions into the grain boundaries. The latter mechanism would have an influence on the mobility since the transport in OFETs depends on the distribution of traps at the grain boundaries.18 The change in surface charge occurs either due to protonation or deprotonation of groups at the surface, for example, at the sulfur atom,19 or due to the specific adsorption of OH− or H3O+ ions onto the surface. Interestingly, a similar low pH sensitivity of 15 mV/pH was reported for diamond devices and attributed to preferential adsorption of water ions.20The pH-dependent surface charge of conductive electrodes in an electrolyte has previously been described with the so-called amphifunctional model.21 In this model, the surface charge is balanced by the electronic charge in the active layer and the diffuse charge as described by the Grahame equation.15 This can be idealized by three planar surfaces, forming the inner layer capacitance C0 and the Helmholtz capacitance C1. The applied potential drops over these two capacitances and the diffuse layer. In the original model, the surface charge results from the protonation or deprotonation of amphoteric groups at the surface; however, the presence of such groups at the ␣6T layer is not evident.The mathematical treatment of the specific adsorption of the dissociated water ions is the same. Furthermore, the preferential absorption of hydroxide ions, giving rise to the required negatively charged surface, is commonly observed on organic surfaces.22 The parameters used for the simulation are given in the caption of Fig. 3. A fit to our data using this model is given by the solid line in Fig. 3共a兲. In order to assess the sensitivity of the device to the ionic strength of the electrolyte, the drain-source current was recorded while the salt concentration of the electrolyte was increased stepwise. During the whole measurement, the pH was kept constant at 7 with the help of a 5 mM 4-共2hydroxyethyl兲-1-piperazineethanesulfonic acid 共HEPES兲 buffer. Figure 4共a兲 summarizes the results of such measurements for different monovalent salts. After the measurements, the original conductance could be recovered by rinsing the device with de-ionized water. As can be seen in Fig.


153302-3

Appl. Phys. Lett. 98, 153302 共2011兲

Buth et al.

(a)

(b) KCl NaCl KBr

0.8

0.6

Uds=-0.1 V -4

Ids (nA)

Ids/Ids(c = 0 mM)

1.0

pH=5.5

no salt 1 mM 10 mM 500 mM

-2

0.4 0.1

1

10

100

1000

Salt Concentration (mM)

0

0.3

0.4

0.5

0.6

0.7

0.8

Usg(V)

FIG. 4. 共a兲 Normalized drain-source current vs ion concentration for several monovalent salts. The solid line shows a simulated curve using a screening model 共assuming a surface charge of −10 ␮C / cm2兲. 共a兲 Drain current vs gate voltage measured in electrolytic solutions with varying KCl concentration, at constant pH. The lines show fits to the linear region.

4共b兲, the decrease in current upon increasing salt concentration cannot simply be explained by a shift in the threshold voltage but is the combined result of an increase in mobility and a shift of Ut to higher voltages. Besides the increase in mobility the subthreshold swing decreases with increasing ionic strength 共from 240 mV/dec with no salt to 180 mV/dec with 1 mM KCl兲,23 which hints to the passivation of trap states at the interface with the active layer by electrolyte ions. This effect is higher at low salt concentrations, while at higher salt concentration a shift of Ut to higher voltages is the main reason for the decrease in Ids. Previous experiments on the ion sensitivity of diamond SGFETs were interpreted in terms of the screening of a pH-dependent surface charge by ions in the diffuse layer.24 Härtl et al.24 explained the decrease in charge upon increasing ionic strength with a model similar to the amphifunctional model. In their model, at higher ionic strength the electrolyte ions screen the surface charge more effectively by increasing the charge in the diffuse layer close to the interface. We have used this screening model to describe the response of our device, and the result is shown in Fig. 4共a兲 by the solid line. Due to the 5 mM HEPES background, the calculated response at low salt concentrations is too low while at higher salt concentrations the screening model is adequate to explain the threshold voltage shift. The observed high response at low salt concentrations results from the competing interplay between the decrease in the subthreshold swing and the increase in the field effect mobility. In conclusion, we could show that polycrystalline 共␣6T兲 thin film transistors can be operated via a field effect with an aqueous electrolytic gate. The large EDL capacitance formed at the organic semiconductor/electrolyte interface enables the low-voltage operation of the devices. The OFETs exhibit a pH sensitivity, which has been attributed to the change in the surface charge, either due to protonation or deprotonation reactions or due to the specific adsorption of the dissociated water ions. Moreover, the current through the channel was sensitive to changes in the ionic strength of the electrolyte—at low salt concentrations this effect is tenta-

tively attributed to the passivation of trap states while at higher concentrations the current variation can be explained by the electrostatic screening of the pH-dependent surface charge. Furthermore, we have shown that these devices, even in direct contact with the aqueous electrolyte, are stable over hours. This, together with a facile tuning of the sensitivity by, for instance, introducing suitable functional groups in the organic semiconductor, demonstrates the potential of OFETs for chemical and biochemical sensing applications. This work is funded by the Nanosystems Initiative Munich 共NIM兲 and the graduate school for Complex Interfaces 共CompInt兲 of the Technical University Munich. 1

C. Bartic, B. Palan, A. Campitelli, and G. Borghs, Sens. Actuators B 83, 115 共2002兲. 2 J. Mabeck and G. Malliaras, Anal. Bioanal. Chem. 384, 343 共2006兲. 3 Organic Semiconductors in Sensor Applications, edited by D. A. Bernards, R. M. Owens, and G. G. Malliaras 共Springer, Berlin, Heidelberg, 2008兲. 4 M. E. Roberts, S. C. B. Mannsfeld, N. Queraltó, C. Reese, J. Locklin, W. Knoll, and Z. Bao, Proc. Natl. Acad. Sci. U.S.A. 105, 12134 共2008兲. 5 T. Someya, A. Dodabalapur, J. Huang, K. C. See, and H. E. Katz, Adv. Mater. 共Weinheim, Ger.兲 22, 3799 共2010兲. 6 M. E. Roberts, S. C. B. Mannsfeld, M. L. Tang, and Z. Bao, Chem. Mater. 20, 7332 共2008兲. 7 B. Stadlober, M. Zirkl, M. Beutl, G. Leising, S. Bauer-Gogonea, and S. Bauer, Appl. Phys. Lett. 86, 242902 共2005兲. 8 E. C. P. Smits, S. G. J. Mathijssen, P. A. van Hal, S. Setayesh, T. C. T. Geuns, K. A. H. A. Mutsaers, E. Cantatore, H. J. Wondergem, O. Werzer, R. Resel, M. Kemerink, S. Kirchmeyer, A. M. Muzafarov, S. A. Ponomarenko, B. de Boer, P. W. M. Blom, and D. M. de Leeuw, Nature 共London兲 455, 956 共2008兲. 9 L. Kergoat, L. Herlogsson, D. Braga, B. Piro, M.-C. Pham, X. Crispin, M. Berggren, and G. Horowitz, Adv. Mater. 共Weinheim, Ger.兲 22, 2565 共2010兲. 10 H. Shimotani, H. Asanuma, and Y. Iwasa, Jpn. J. Appl. Phys., Part 1 46, 3613 共2007兲. 11 M. J. Panzer, C. R. Newman, and C. D. Frisbie, Appl. Phys. Lett. 86, 103503 共2005兲. 12 J. Kong, N. R. Franklin, C. Zhou, M. G. Chapline, S. Peng, K. Cho, and H. Dai, Science 287, 622 共2000兲. 13 H. Kawarada, Y. Araki, T. Sakai, T. Ogawa, and H. Umezawa, Phys. Status Solidi A 185, 79 共2001兲. 14 M. Dankerl, M. V. Hauf, A. Lippert, L. H. Hess, S. Birner, I. D. Sharp, A. Mahmood, P. Mallet, J.-Y. Veuillen, M. Stutzmann, and J. A. Garrido, Adv. Funct. Mater. 20, 3117 共2010兲. 15 A. Bard and I. Faulkner, Electrochemical Methods: Fundamentals and Applications 共Wiley, New York, 2001兲. 16 J. A. Garrido, S. Nowy, A. Härtl, and M. Stutzmann, Langmuir 24, 3897 共2008兲. 17 K. Meerholz and J. Heinze, Electrochim. Acta 41, 1839 共1996兲. 18 S. Verlaak, V. Arkhipov, and P. Heremans, Appl. Phys. Lett. 82, 745 共2003兲. 19 M. Can, F. Sevin, and A. Yildiz, Appl. Surf. Sci. 210, 338 共2003兲. 20 M. Dankerl, A. Reitinger, M. Stutzmann, and J. A. Garrido, Phys. Status Solidi 共RRL兲 2, 31 共2008兲. 21 J. Duval, J. Lyklema, J. M. Kleijn, and H. P. van Leeuwen, Langmuir 17, 7573 共2001兲. 22 R. Zimmermann, U. Freudenberg, R. Schweiß, D. Küttner, and C. Werner, Curr. Opin. Colloid Interface Sci. 15, 196 共2010兲. 23 See supplementary material at http://dx.doi.org/10.1063/1.3581882 for a semilogarithmic plot of the transfer characteristic in Fig. 4共b兲. 24 A. Härtl, J. A. Garrido, S. Nowy, R. Zimmermann, C. Werner, D. Horinek, R. Netz, and M. Stutzmann, J. Am. Chem. Soc. 129, 1287 共2007兲.
