Flexible and Printed Electronics PAPER • OPEN ACCESS Impact of contact overlap on transconductance and noise in organic electrochemical transistors To cite this article: Anastasios G Polyravas et al 2019 Flex. Print. Electron. 4 044003 View the article online for updates and enhancements. This content was downloaded from IP address 95.169.233.50 on 19/06/2020 at 17:03 Flex. Print. Electron. 4 (2019) 044003 https://doi.org/10.1088/2058-8585/ab4dc4 PAPER Impact of contact overlap on transconductance and noise in organic OPEN ACCESS electrochemical transistors RECEIVED 21 July 2019 Anastasios G Polyravas1, Vincenzo F Curto1, Nathan Schaefer2, Andrea Bonaccini Calia2, REVISED 4 October 2019 Anton Guimera-Brunet3,4, Jose A Garrido2,5 and George G Malliaras1 1 Electrical Engineering Division, Department of Engineering, University of Cambridge, 9 JJ Thomson Ave, Cambridge CB3 0FA, United ACCEPTED FOR PUBLICATION 15 October 2019 Kingdom 2 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC, Barcelona Institute of Science and Technology, Campus UAB, PUBLISHED 1 November 2019 Bellaterra, Barcelona, Spain 3 Institut de Microelectronica de Barcelona, IMB-CNM (CSIC), E-08193, Bellaterra, Barcelona, Spain 4 CIBER-BBN, Networking Center on Bioengineering, Biomaterials and Nanomedicine, Barcelona, Spain Original content from this 5 ICREA, Pg. Lluís Companys 23, E-08010 Barcelona, Spain work may be used under the terms of the Creative E-mail: gm603@cam.ac.uk Commons Attribution 3.0 licence. Keywords: organic electrochemical transistors, transconductance, noise Any further distribution of Supplementary material for this article is available online this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Abstract Organic electrochemical transistors (OECTs) from poly(3,4-ethylenedioxythiophene) doped with polystyrene sulfonate (PEDOT:PSS) are used as amplifying transducers for bioelectronics. Although the impact on performance of device geometry parameters such as channel area and thickness has been widely explored, the overlap between the semiconductor film and the source and drain contacts has not been considered. Here we vary this overlap and explore its impact on transconductance and noise. We show that increasing contact overlap does not alter the magnitude of the steady-state transconductance but it does decreases the cut-off frequency. Noise was found to be independent of contact overlap and to vary according to the charge noise model. The results show that high-quality contacts can be established in PEDOT:PSS OECTs with minimal overlap. 1. Introduction The figure-of-merit that characterises the ability of transistors to transduce a signal is the transconduc- Organic electrochemical transistors (OECTs) are tance gm, defined as the first derivative of the drain attracting a great deal of attention for applications in current Id over the gate voltage Vg. OECTs have been electronics and bioelectronics [1]. Their operation shown to exhibit a very high transconductance, due to relies on the injection of ions from an electrolyte into the volumetric changes in the conductivity of the the volume of a semiconducting polymer. Poly(3,4- channel [9]. A great deal of work has focused on ethylenedioxythiophene) doped with polystyrene sul- understanding how to optimise transconductance by fonate (PEDOT:PSS), a p-type degenerately doped tuning the dimensions of the channel [1, 10, 11] and semiconductor, has become the material of choice for the properties of the semiconductor [1, 12]. A second OECTs due to its commercial availability, stability and parameter that is important in signal transduction is biocompatibility [2]. Small metal cations can be read- the noise produced by the transistor. Flicker noise, ily injected in this material from a variety of electro- which decreases with frequency as 1/f, has been shown lytes under the influence of an applied gate voltage, to be dominant at low frequencies in a variety of tran- leading to large changes in the drain current. PEDOT: sistor technologies [13–17], including OECTs [18]. PSS OECTs have been used to detect a variety of Fluctuations in the hole density inside the channel are chemical and biological analytes [3], to transduce assumed to be responsible for this behaviour [18]. electrophysiological activity [4, 5], to develop 3D tissue An important parameter that has not yet been models [6], and also to fabricate large-area circuitry [7] explored in OECTs is the overlap between the polymer and neuromorphic devices [8]. film and the source and drain contacts. This overlap is usually determined by the registration of the © 2019 IOP Publishing Ltd Flex. Print. Electron. 4 (2019) 044003 A G Polyravas et al Figure 1. (a) Schematic of an OECT and digram of the circuit used for the noise measurements. (b) Optical micrograph of OECTs with 10% (top) and 70% (bottom) contact overlap. patterning technique employed to define the top layer. 4500 rpm for 30 s and was post baked at 110 °C for two In thin film transistors, such design parameters are minutes. The samples were exposed to UV light, and known to affect performance through changes in con- developed for 6 min using a MIF 726 developer. Reac- tact resistance and overlap capacitance [19]. Here we tive ion etching (Oxford 80 Plasmalab plus) opened explore how this overlap affects transconductance and the window for the deposition of the PEDOT:PSS. noise in PEDOT:PSS OECTs. 19 ml of Clevios PH1000 (Heraeus), 1 ml of ethylene glycol, 50 μl of dodecyl benzene sulfonic acid (DBSA, Sigma Aldrich) and 0.2 ml of (3-glycidyloxypropyl)tri- 2. Materials and methods methoxysilane (GOPS, Sigma Aldrich) were mixed and sonicated for 15 min The surface was activated 2.1. Device fabrication and the PEDOT:PSS dispersion was spun at 3000 rpm The fabrication of OECTs was based on the parylene for 30 s. The samples were annealed at 110 °C for peel-off processes published previously [20] and 1 min and the sacrificial layer of parylene C was peeled included the deposition of gold, parylene C and off. Finally, the samples were hard-baked at 130 °C for PEDOT:PSS films. Glass slides (Knittel Glass) with an hour and immersed into DI water overnight to dimensions of 26×76 mm2 were used as substrates. remove any excess compounds. They were cleaned by an ultrasonic bath in a dilute solution of Micro-90 (Cole Parmer) in deionised water (DI) for 20 min, rinsed with DI water, and sonicated 2.2. Device characterization again inside an acetone—isopropyl alcohol solution An optical microscope (Nikon Eclipse LV100ND) was (90%–10%) for another 20 min. To pattern the metal used to verify that the target dimensions were within pads a negative photoresist, AZ nLOF2035 (Micro- 2 μm of the target values. A Bruker Dimension Icon Chemicals GmbH), was spun on the glass substrate at AFM was used to verify that film morphology does not 4000 rpm for 30 s, baked at 110 °C for 1 min and change with contact overlap (figure S1 in supplemen- exposed to UV light using a mask aligner (Karl Suss tary information is available online at stacks.iop.org/ MA/BA6). The samples were then baked at 110 °C for FPE/4/044003/mmedia). A Dektak XT stylus profil- 1 min and developed using a MIF 726 developer. They ometer (Bruker) was used to measure the thickness of were activated for 2 min using an O2 plasma cleaner the PEDOT:PSS. Transistors were characterised using (Diener Electronic Femto) and placed inside an a 0.01 M phosphate buffered saline (PBS) solution and e-beam evaporator (Kurt J Lesker PVD-75) for metal- a Ag/AgCl gate electrode (World Precision Instru- lization. A thin layer of Ti (5 nm) was deposited to aid ments). Electrical measurements were performed adhesion, followed by 100 nm of gold. An overnight inside a Faraday cage. A semiconductor device analyser lift-off, using Technistrip NI555 (MicroChemicals (Keysight B1500A) was employed to measure the GmbH) led to the final metal patterns. output and transfer curves. A delay of at least 100 ms The samples were activated again for 2 min and was applied between sourcing voltage and measuring submerged in an ethanol solution of the adhesion pro- current to ensure that the latter reached steady state. moter 3-(trimethoxysilyl)propyl methacrylate (Sigma The frequency response of the devices was measured Aldrich A-174) for 30 s, then hard-baked on a hotplate with two NI-PXI-4071 digital multimeters controlled at 70 °C for an hour. Subsequently, they were placed in by LabVIEW. The OECTs were biased at a constant Vd a parylene coater (SCS) and 2 μm of parylene C was while the gate was biased with a series of sine waves of deposited. An anti-adhesive layer of Micro-90 in DI 50 mV amplitude and frequency between 1 Hz and water (2% v/v soap solution) was spun at 1000 rpm, 20 kHz. Noise was measured using a custom-built and a second (sacrificial) layer of parylene C was transimpedance amplifier with 10 k gain [21], as deposited. A layer of photoresist AZ9260 (Micro- shown in figure 1(a). The output was split into DC Chemicals GmbH) was then spun on the samples at (<0.1 Hz) and AC (0.1 Hz<frequency<5 kHz) 2 Flex. Print. Electron. 4 (2019) 044003 A G Polyravas et al Figure 2. (a) Transconductance as a function of Vg for OECTs with different contact overlap at Vd=−0.6 V. Lines are guides to the eye. (b) Corresponding normalised transconductance versus frequency. components and the latter was amplified by an with performance that is suitable for cutaneous additional factor of 100 and recorded with a data electrophysiology [11]. The overlap between the acquisition system (National Instruments USB-6363). PEDOT:PSS film and the source and drain contacts An FFT transformation was performed to calculate the was varied to 5, 15, 25 and 35 μm respectively, leading power spectral density of the drain current fluctua- to 10%, 30%, 50% and 70% overlap between PEDOT: tions SId. Customised software in MATLAB removed PSS film and contact, respectively. Optical micro- the 50 Hz noise and its harmonics, and Sid was graphs of OECTs with 10% and 70% contact overlap smoothed with the smooth() function. Figures S2(a) are shown in figure 1(b). The OECTs operated in and (b) in supplementary information show the effects depletion mode (figure S3 in supplementary informa- of filtering and smoothing on the data. tion), consistent with previously published data [22]. 3. Results and discussion 3.2. Impact on transconductance Figure 2(a) shows the dependence of the transconduc- 3.1. Contact overlap tance gm on Vg for OECTs with different contact In order to investigate the impact of contact overlap on overlaps. These four OECTs were fabricated on the device performance, we fabricated arrays of OECTs same glass slide and were equally spaced from the with a fixed source and drain electrode geometry and centre, to ensure similar PEDOT:PSS film thickness. varied the dimensions of the PEDOT:PSS film span- The data shows no systematic variation in gm as a ning the Au contacts, as seen in figure 1(b). The OECTs function of contact overlap. This is not surprising, as had a channel length L=50 μm (defined by the all devices have the same nominal channel dimen- distance between the source and drain contacts), a sions. It means that a high-quality ohmic contact is width W=50 μm (defined by the width of the achieved at both the source and the drain, even for the PEDOT:PSS film) and channel thickness d∼100 nm OECT with the smallest overlap. (defined by the thickness of the PEDOT:PSS film). The frequency dependence of gm for the four tran- These dimensions were chosen as they lead to devices sistors with different contact overlaps is shown in 3 Flex. Print. Electron. 4 (2019) 044003 A G Polyravas et al Figure 3. (a) Relative noise at 10 Hz versus Vg for OECTs with different contact overlaps at Vd=−0.5 V. Lines are guides to the eye. (b) Relative noise at 10 Hz versus g2m/I2d for OECTs with 30% contact overlap and different values of Vg and Vd=−0.5 V. The line is a fit to the charge noise model. figure 2(b). A systematic decrease in the cut-off fre- devices and bias conditions. Figure 3(a) shows the quency fc=gm,max/√2 is observed as the overlap relative noise at 10 Hz as a function of Vg, for different increases: The OECT with 70% contact overlap is values of contact overlap. The value of 10 Hz was ∼300 Hz slower than the one with 10% contact over- selected as it corresponds to the strongest brain lap. The cut-off frequency of transistors depends on rhythm (alpha wave) in encephalography [26]. The the RC time constant of the gate circuit [23]. In data shows no systematic variation in noise as a OECTs, the capacitance C arises from the charging of function of contact overlap. This is consistent with the the polymer film with ions [24]. The data indicates fact that device performance, as seen by the steady- that the capacitance that is important here is the one state transconductance, does not vary with contact associated with the entire volume of the PEDOT:PSS overlap. It means that the relevant feature that film, rather than with just the area between the source determines noise is the channel length and not the and drain contact. At the same time, the resistance R length of the PEDOT:PSS film. depends on the perimeter of the PEDOT:PSS film [25]. Figure 3(b) shows that relative noise scales with A calculation based on these arguments (see supple- g2m/I2d. This is consistent with the charge noise model mentary information) predicts a change in fc of [27, 28], according to which the measured noise in the 287 Hz, in close agreement to the measured value. drain current originates from fluctuations in the hole density inside the channel. Noise can therefore be 3.3. Impact on noise referred to voltage fluctuations at the gate according to Consistent with previous measurements in electrolyte Sid=g2m · SVg, where SVg is the gate-referred noise gated transistors, the power spectral density of the [27, 28]. The line is a fit to the model yielding drain current fluctuations SId was found to follow a 1/f SVg=12.7·10–14 V2 Hz−1. This value, corrected for dependence (figure S4 in supplementary information) channel dimensions, compares favourably to previous [14, 17, 18]. The normalised power spectral density measurements in PEDOT:PSS OECTs [18] and elec- SId/I2d, also referred to as the relative noise, is often trolyte-gated graphene transistors [17]. Additional used in literature to compare data from different 4 Flex. Print. Electron. 4 (2019) 044003 A G Polyravas et al characterisation data are shown in figures S5 and S6 in References supplementary information. [1] Rivnay J, Inal S, Salleo A, Owens R M, Berggren M and Malliaras G G 2018 Organic electrochemical transistors Nat. 3.4. Discussion Rev. Mater. 3 17086 Simple fabrication, biocompatibility and high trans- [2] Mantione D et al 2017 Poly(3,4-ethylenedioxythiophene) conductance make OECTs excellent candidates for (PEDOT) derivatives: innovative conductive polymers for bioelectronics Polymers 9 354 recording of biological signals. They are particularly [3] Lin P and Yan F 2012 Organic thin-film transistors for well suited for recording cutaneous electrophysiology chemical and biological sensing Adv. Mater. 24 34–51 [29], as these signals typically occur below 100 Hz and [4] Khodagholy D et al 2013 In vivo recordings of brain activity using organic transistors Nat. Commun. 4 1575 have low amplitudes. The latter range from mV for [5] Campana A, Cramer T, Simon D T, Berggren M and signals from muscles to tens of μV for brain activity Biscarini F 2014 Electrocardiographic recording with recordings. According to the change noise model, the conformable organic electrochemical transistor fabricated on noise Vrms during a recording with a bandpass filter set resorbable bioscaffold Adv. Mater. 26 3874–8 between 1 and 100 Hz can be calculated by: [6] Pitsalidis C, Ferro M P, Iandolo D, Tzounis L, Inal S and Owens R M 2018 Transistor in a tube: a route to three- 100 Hz dimensional bioelectronics Sci. Adv. 4 eaat4253 ò1 Hz S Id · df [7] Sun H et al 2018 Complementary logic circuits based on high- Vrms = . (1) performance n-type organic electrochemical transistors Adv. gm 2 Mater. 30 1704916 Accordingly, we find Vrms=3.4 μV. This value is [8] Y van de B, Melianas A, Keene S T, Malliaras G and Salleo A 2018 Organic electronics for neuromorphic computing Nat. more than adequate for enabling recordings of cuta- Electron. 1 386 neous electrophysiology with high signal-to-noise [9] Khodagholy D et al 2013 High transconductance organic ratio [29, 30]. electrochemical transistors Nat. Commun. 4 2133 Previous studies showed that the trade-off [10] Rivnay J et al 2013 Organic electrochemical transistors with maximum transconductance at zero gate bias Adv. Mater. 25 between gain and bandwidth can be navigated in 7010–4 OECTs by adjusting channel thickness: increasing [11] Rivnay J et al 2015 High-performance transistors for channel thickness increases transconductance but bioelectronics through tuning of channel thickness Sci. Adv. 1 decreases the cut-off frequency [11]. An additional e1400251 [12] Inal S, Malliaras G G and Rivnay J 2017 Benchmarking organic design rule identified here calls for minimizing contact mixed conductors for transistors Nat. Commun. 8 1767 overlap, as the latter only serves to make the OECT [13] Klaassen F M 1971 Characterization of low 1/f noise in MOS slower and does not affect steady-state transconduc- transistors IEEE Trans. Electron Devices 18 887–91 tance and noise. [14] Sharf T, Kevek J W, DeBorde T, Wardini J L and Minot E D 2012 Origins of charge noise in carbon nanotube field-effect transistor biosensors Nano Lett. 12 6380–4 [15] Clément N, Nishiguchi K, Dufreche J F, Guerin D, 4. Conclusions Fujiwara A and Vuillaume D 2011 A silicon nanowire ion- sensitive field-effect transistor with elementary charge In this work, we investigated how overlap between the sensitivity Appl. Phys. Lett. 98 014104 PEDOT:PSS film and the source and drain contacts [16] Rhayem J, Rigaud D, Valenza M, Szydlo N and Lebrun H 1999 1/f Noise in amorphous silicon thin film transistors: effect of affects OECT performance. We find that while steady- scaling down Solid-State Electron. 43 713–21 state transconductance and noise remain the same, the [17] Heller I et al 2010 Charge noise in graphene transistors Nano cut-off frequency decreases with increasing contact Lett. 10 1563–7 overlap. This calls for minimising this overlap to [18] Stoop R L, Thodkar K, Sessolo M, Bolink H J, Schönenberger C and Calame M 2017 Charge noise in organic achieve the best performing devices. electrochemical transistors Phys. Rev. Appl. 7 014009 [19] Münzenrieder N et al 2014 Contact resistance and overlapping capacitance in flexible sub-micron long oxide thin-film Acknowledgments transistors for above 100 MHz operation Appl. Phys. Lett. 105 263504 This work was funded by the European Union’s [20] Sessolo M et al 2013 Easy-to-fabricate conducting polymer Horizon 2020 research and innovation programme microelectrode arrays Adv. Mater. 25 2135–9 [21] Hébert C et al 2018 Flexible graphene solution-gated field- under grant agreement no. 732032 (BrainCom). We effect transistors: efficient transducers for micro- thank Ioannis Prattis and Prof. Rachel Oliver for electrocorticography Adv. Funct. Mater. 28 1703976 fruitful discussions and assistance with the AFM [22] Bernards D A and Malliaras G G 2007 Steady-state and measurements. transient behavior of organic electrochemical transistors Adv. Funct. Mater. 17 3538–44 [23] Sze S M and Ng K K 2006 Physics of Semiconductor Devices. 3rd ORCID iDs edn (New York: Wiley) 832 p [24] Proctor C M, Rivnay J and Malliaras G G 2016 Understanding volumetric capacitance in conducting polymers J. Polym. Sci., George G Malliaras https://orcid.org/0000-0002- Part B: Polym. Phys. 54 1433–6 4582-8501 [25] Koutsouras D A, Gkoupidenis P, Stolz C, Subramanian V, Malliaras G G and Martin D C 2017 Impedance spectroscopy of spin-cast and electrochemically deposited PEDOT:pss films 5 Flex. Print. Electron. 4 (2019) 044003 A G Polyravas et al on microfabricated electrodes with various areas [28] Bedner K et al 2014 Investigation of the dominant 1/f noise ChemElectroChem 4 2321–7 source in silicon nanowire sensors Sensors Actuators B 191 [26] Buzsaki G 2011 Rhythms of the Brain 1 edn (Oxford: Oxford 270–5 University Press) p 464 [29] Leleux P et al 2015 Organic electrochemical transistors for [27] Tersoff J 2007 Low-frequency noise in nanoscale ballistic clinical applications Adv. Healthcare Mater. 4 142–7 transistors Nano Lett. 7 194–8 [30] Thakor N V 2013 Translating the brain-machine interface Sci. Transl. Med. 5 210ps17 6