Geometry Control of Source/Drain Electrodes in Organic Field-Effect

materials

Article Geometry Control of Source/Drain Electrodes in Organic Field-Eﬀect Transistors by Electrohydrodynamic Inkjet Printing

Piotr Sleczkowski 1,* , Michal Borkowski 1, Hanna Zajaczkowska 1 , Jacek Ulanski 1, Wojciech Pisula 1,2 and Tomasz Marszalek 1,2,*

1 Department of Molecular Physics, Faculty of Chemistry, Lodz University of Technology, Zeromskiego 116, 90-924 Lodz, Poland; [email protected] (M.B.); [email protected] (H.Z.); [email protected] (J.U.); [email protected] (W.P.) 2 Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany * Correspondence: [email protected] (P.S.); [email protected] (T.M.)

Received: 9 October 2020; Accepted: 2 November 2020; Published: 5 November 2020

Abstract: In this work we study the influence of dielectric surface and process parameters on the geometry and electrical properties of silver electrodes obtained by electrohydrodynamic inkjet printing. The cross-section and thickness of printed silver tracks are optimized to achieve a high conductivity. Silver overprints with cross-section larger than 4 µm2 and thickness larger than 90 nm exhibit the lowest resistivity. To fabricate electrodes in the desired geometry, a sufficient volume of ink is distributed on the surface by applying appropriate voltage amplitude. Single and multilayer overprints are incorporated as bottom contacts in bottom gate organic field-effect transistors (OFETs) with a semiconducting polymer as active layer. The multilayer electrodes result in significantly higher electrical parameters than single layer contacts, confirming the importance of a careful design of the printed tracks for reliable device performance. The results provide important design guidelines for precise fabrication of electrodes in electronic devices by electrohydrodynamic inkjet printing.

Keywords: printed electronics; electrohydrodynamic inkjet printing; organic ﬁeld-eﬀect transistors

1. Introduction Micro- and nanofabrication of functional structures on surfaces is of the utmost importance for the development of numerous existing and emerging technologies in electronics [1,2], optics [3,4], medicine [5] and pharmaceuticals [6] among others [7]. The electronic and optoelectronic industries typically exploit lithographic methods to continuously increase the density of integrated circuits by size reduction of the components [8]. The pitch of (opto)electronic elements is precisely controlled down to the 100 nm scale by optical lithography [9]. However, the emergence of new research areas, especially at the interface with organic and biological materials, requires the development of fabrication methods employing conditions that are less demanding in terms of chemical and thermal processing. The subtractive character of conventional lithographic methods can cause, for example, degradation of materials under extreme conditions of temperature or pressure. To circumvent these constraints, alternative approaches based on additive fabrication are continuously being developed [10,11]. Inkjet printing is one of the most widely used additive fabrication methods which enables deposition of material of interest on various types of surface, including ﬂexible and stretchable substrates [12]. This is particularly important from the perspective of development of wearable devices that can integrate seamlessly with the human body as e.g., sensors for health monitoring [13]. Inkjet printing is a non-contact method that relies on the generation of droplets from the aperture of the printing nozzle [14].

Materials 2020, 13, 4974; doi:10.3390/ma13214974 www.mdpi.com/journal/materials Materials 2020, 13, 4974 2 of 16

Most often it is realized with the use of pressure pulses generated by thermal or piezoelectric means [15]. The movement control of the printing nozzle (or the sample), together with the possibility of delivering the discrete amount of ink, allows pre-defined layouts of materials to be created. The achievable resolution of the inkjet printed patterns is determined by the volume of droplet ejected from the nozzle [14,15]. In the case of thermal/piezoelectric printheads the picoliter size droplets are deposited on the surface, resulting in printed features in the range of tens of micrometers [16]. Since capillary forces become more pronounced as the characteristic length shrinks, pushing out the droplet becomes more difficult as the nozzle orifice becomes smaller [17]. An alternative to pressure-driven droplet ejection is the use of the electric field which pulls the liquid out of the nozzle. Since the generation of an electrohydrodynamic jet is less dependent on the nozzle diameter, especially in the cone-jet mode where a Taylor cone is formed, the electrohydrodynamically induced flows offer an efficient way of improving the printing resolution with respect to conventional inkjet printing [18]. Over the past decade, electrohydrodynamic (e-jet) printing has attracted growing interest, especially in the areas of printed electronics [19], sensors [20], supercapacitors [21], photonic devices [22] and 3D printing [23]. Using e-jet printing for the fabrication of electronic components offers many advantages, particularly in the elaboration of source and drain electrodes for field-effect transistors (FETs), for which a precise control of shape is indispensable for ensuring reliable device performance. Regardless the droplet generation method, the first step to realize electrodes relies on the control of the deposited droplet morphology that promotes their coalescence and results in the formation of continuous lines [24]. Various strategies for tailoring the deposition morphologies include the suppressing (or utilizing) of the coffee-ring effect [25] and using patterned substrates [26]. Surface wettability influences significantly the feature size and cross-sectional profile [27] and determines the uniformity of the printed lines [28]. Besides the basic requirements of low resistance, good contact with the semiconductor and uniform morphology of printed lines are also particularly crucial for bottom contact organic FETs (OFETs). The morphology of the contact is important due to channel length variations resulting from lateral imperfections of the printed lines, while the ordering of the semiconductor molecules strongly depends on the surface topography of the electrode. Some of the principal printed line behaviours deviating from the ideal (uniform mode) are scalloped, bulging and stacked coins, whereby the morphology is altered by the drop spacing and substrate temperature [29]. The combination of drop spacing and substrate temperature reduces the coffee-ring effect and results in this way in an improved performance of inkjet-printed Ag source/drain electrodes in OFETs [30]. In the case of water-based inks, the Ag electrode profile can be tuned by controlling the ambient humidity [31]. Besides temperature and humidity, the modification of the surface energy of the substrate by self-assembled monolayers [32], ultraviolet (UV)-ozone [33] or plasma treatment [34] is established as an alternative way for reducing or eliminating the coffee-ring effect. Other groups have examined several parameters of metallic nanoparticle inks in view of their influence on the overprint geometry, including the solvent [33], particle size [35] and concentration [36], ink viscosity [37] and conductivity [38]. Apart from a proper ink and substrate, the printing parameters offer a wide range of adjustment of the line geometry. Since e-jet printing is based on electrohydrodynamically induced flows, most common process parameters are connected to the electric field applied for the ink deposition. These parameters include the voltage bias between the nozzle and the printing table, the amplitude and frequency of the voltage pulse, and on the distance between nozzle and substrate. Previous studies reported the printing of continuous Ag lines on silicon wafers [39] and glass substrates [40], with an overprint linewidth directly proportional to the amplitude of the voltage pulse. The influence of the pulse frequency was also examined in various regimes [41] and on various substrates, including highly insulating substrates [42]. Finally, a drastic change of linewidth and morphology was reported, while changing the standoff height, i.e., the nozzle–substrate distance [43]. Mantysalo et al. have conducted an extensive analysis of the effects of printing parameters on the Ag-nanoparticle ink overprints by developing a statistical model for droplet width as a function of four main parameters, such as print height, bias voltage, peak voltage and frequency [44]. Despite intensive research on the electrode deposition by the e-jet, Materials 2020, 13, 4974 3 of 16 the implementation of optimized overprints as conductive lines in electronic devices is still scarce. Development of new fabrication procedures and improvement of the electrical resistance without realization and analysis of the entire electronic device does not result in solving the most common and critical problems encountered [45,46]. In the case of OFETs, crucial aspects include contact resistance and the deposition of the source and drain electrodes [47]. In this work, we demonstrate fine geometry control of silver overprints printed by an ultraprecise e-jet. We investigate the influence of printing parameters on the linewidth and cross-section of sintered silver overprints and correlate the electrode morphology to their resistivity. The process parameters are optimized for single and multilayer overprints by performing detailed analysis of their geometrical features. To relate their properties with the device performance and finally to prove their potential in practical applications, the overprints are incorporated in OFETs as source and drain electrodes.

2. Materials and Methods

2.1. Silver Nanopaste Ink A silver nanoparticle (5–10 nm diameter) ink DGP 40LT-15C containing 31.04 wt.% Ag in triethylene glycol monoethyl ether (TGME) (Advanced Nano Products Co., Ltd., Sejong, Korea) was used for the inkjet printing of the conductive tracks and their further investigation as potential use in OFETs. The density, dynamic viscosity and surface tension of DGP 40LT-15C at 25 ◦C are 1.45 g/mL, 13.7cP and 34.65 mN/m, respectively. Conductivity of DGP 40LT-15C is 8.3–9.1 MS/m [48]. During all experiments the temperature of the DGP ink was kept at 25 ◦C.

2.2. Sample Preparation

Substrates consisted of heavily doped n-type Si wafers with a 300 nm thick SiO2 dielectric layer 2 (capacitance of 11 nF cm− ) purchased from EL-CAT Inc. (Ridgefield Park, NJ, USA). Different preparation procedures of SiO2 surface modification were used (Table1). All substrates were cleaned in a sequence of sonication baths in acetone and isopropyl alcohol (iPrOH). In addition, some samples were cleaned in chloroform (CHCl3), or treated with oxygen plasma (Plasma) for 5 min (Atto, Diener Electronic GmbH & Co KG, Ebhausen, Germany). Two other batches of SiO2 samples were, after sonication in isopropyl alcohol, chemically modified by either trichloro(octyl)silane (OTS) or hexamethyldisilazane (HMDS) monolayers. Prior to chemisorption of OTS or HMDS, the SiO2 substrates were cleaned in piranha solution. In the case of OTS functionalization, the substrates were immersed in µM solution of OTS in a 1:4 v/v mixture of chloroform and n-hexane. The substrates were kept in the OTS solution for 30 min and then rinsed with chloroform in an ultrasonic cleaner and dried with nitrogen. In the case of HMDS functionalization, the substrates were kept in HMDS vapor inside a closed chamber at a temperature 135 ◦C for 12 h, then rinsed with chloroform in an ultrasonic cleaner and finally dried with nitrogen.

Table 1. Summary of H2O and DGP contact angles (CA) for different surface treatments Si/SiO2 substrates.

Surface Treatment CA(H2O) CA(DGP)

1 Plasma 32◦ 25◦ 2 CHCl3 63◦ 32◦ 3 iPrOH 62◦ 34◦ 4 OTS 82◦ 42◦ 5 HMDS 96◦ 51◦

Acetone, isopropyl alcohol and chloroform were purchased from Avantor Performance Materials Poland S. A. (Gliwice, Poland) and used as received. Trichloro(octyl)silane, n-hexane and hexamethyldisilazane were purchased from Sigma Aldrich and used as received. Materials 2020, 13, 4974 4 of 16

2.3. Contact Angle Measurements Materials 2020, 13, x FOR PEER REVIEW 4 of 16 The static contact angle was determined using an OCA 15EC Goniometer from Data Physics Instruments2.3. Contact GmbH, Angle Filderstadt,MeasurementsGermany. Measurements were performed on SiO2 substrates with various surface treatments, as described in Section 2.2. For each sample the contact angle values were The static contact angle was determined using an OCA 15EC Goniometer from Data Physics recorded at four different locations of the substrate and averaged. The contact angle of water was also Instruments GmbH, Filderstadt, Germany. Measurements were performed on SiO2 substrates with measuredvarious for surface all types treatments, of substrates as described tested duringin Sectio thisn 2.2. study For each as a reference.sample the Thecontact droplet angle volume values was 5 µL forwere the recorded tests with at four both different liquids locations and the of accuracy the substrate of the and measurements averaged. The was contact0.5 angle◦. of water ± was also measured for all types of substrates tested during this study as a reference. The droplet 2.4. Inkjetvolume Printing was 5 μL for the tests with both liquids and the accuracy of the measurements was ±0.5°. The printing was performed using the SIJ-S050 SuperInkjet printer (SIJ Technology, Inc., Tsukuba, 2.4. Inkjet Printing Japan). SIJ-S050 is an electrohydrodynamic printer, which uses an electric field between the printer stage andThe the printing printing was nozzle performed (3–4 µ usingm inner the aperture)SIJ-S050 SuperInkjet to deposit printer small (~fL(SIJ Technology, range) ink volumes.Inc., In orderTsukuba, to eject Japan). the droplet SIJ-S050 from is an theelectrohydrodynami capillary nozzlec anprinter, electric which pulse uses in an the electric desired field waveform between was the printer stage and the printing nozzle (3–4 μm inner aperture) to deposit small (~fL range) ink generated. Triangular waveform was used throughout the reported experiments. The amplitude volumes. In order to eject the droplet from the capillary nozzle an electric pulse in the desired and frequency of the pulse were modulated by the operator, and were chosen as basic control waveform was generated. Triangular waveform was used throughout the reported experiments. The parametersamplitude in thisand study.frequency During of thethe pulse printing were modula processted the by nozzlethe operator, remained and were static chosen and the as basic substrate was movedcontrol byparameters a computer-controlled in this study. During moving the stage printing at constant process the speed nozzle of 1 remained mm/s. The static printing and the height (the substrate-nozzlesubstrate was moved distance) by a wascomputer-controlled kept constant at moving around stage 50 µm. at Theconstant optimal speed electric of 1 mm/s. field toThe induce ejectionprinting of the height DGP (the ink substrate-nozzle was set at B = 200distance) V, where was Bkeptstands constant for theat around constant 50 biasμm. voltageThe optimal between nozzleelectric and samplefield to induce stage. ejection In addition, of the DGP the amplitudeink was set at (A B) and= 200peak V, where of the B stands alternating for the currentconstant (AC) voltagebias was voltage set between between 225 nozzle V and and 300 sample V, to controlstage. In the addition, ejected inkthe volume.amplitude Applying (A) and Apeak> 300 of the V in the studiedalternating frequency current regime (AC) (i.e., voltage between was setf =between100 Hz 225 and V andf = 1000300 V, Hz) to control resulted the in ejected ejection ink ofvolume. a too high Applying A > 300 V in the studied frequency regime (i.e., between f = 100 Hz and f = 1000 Hz) DGP ink volume preventing from fine control of the overprints shape. But applying too low A did not resulted in ejection of a too high DGP ink volume preventing from fine control of the overprints lead to ejection of any ink. shape. But applying too low A did not lead to ejection of any ink. GeometricalGeometrical and and electrical electrical characterization characterization of the overprints overprints was was performed performed after after the thermal the thermal annealingannealing/sintering/sintering process. process. In orderIn order to sinterto sinter the the deposited deposited ink ink and and to to remove remove thethe TGMETGME solvent and the cappingand the agents,capping theagents, overprints the overprints were heated were heated at 150 ◦atC 150 for 1°C h. for In 1 the h. caseIn the of case multiple of multiple overprints (n > 1),overprints the thermal (n > 1), treatment the thermal was treatment performed was afterperforme thed last after layer the last was layer printed. was printed.

2.5. Geometrical2.5. Geometrical Characterization Characterization of of the the Overprints Overprints WidthWidth (W) of(W the) of overprints the overprints was was estimated estimated from from the microscopythe microscopy images images recorded recorded with with the the Omano OM349POmano optical OM349P microscope optical microscope (Microscope (Microscope LLC, Roanoak, LLC, Roanoak, VA, USA) VA, and USA) from and the from surface the profilometry surface profilometry studies. Thickness of overprints was estimated with the DektakXT stylus profiler studies. Thickness of overprints was estimated with the DektakXT stylus profiler (Bruker Ltd., Coventry, (Bruker Ltd., Coventry, UK). The maximum height (hmax) represents the arithmetic average of UK). The maximum height (h ) represents the arithmetic average of maximum height of three maximum height of three differentmax scans of the sample. Figure 1a depicts a typical sample in top differentview scans with ofboth the sample.inputs (s, Figure n) and1a outputs depicts a(W typical, L) geometrical sample in parameters. top view with Figure both 1b inputs shows ( s ,ann) and outputsexemplary (W, L) surface geometrical profile parameters. recorded by DektakXT, Figure1b showsunderlining an exemplary W, L and hmax surface. The cross-section profile recorded area by DektakXT,of the underliningoverprints wasW ,furtherL and husedmax. for The calculatio cross-sectionns of the area resistivity of the overprints from the wastwo-point further probe used for calculationsmeasurements. of the resistivity from the two-point probe measurements.

FigureFigure 1. (a 1.) Schematic(a) Schematic top top view view representationrepresentation of of DGP DGP overprints overprints with with various various interline interline spacing spacing(s) (s) andand di differentﬀerent number number ofof printingprinting passes ( (nn).). (b (b) )Exemplary Exemplary cross-section cross-section profile proﬁle underlining underlining the the geometricalgeometrical parameters parametersW, WL,and L andhmax hmax..

Materials 2020, 13, 4974 5 of 16

2.6. Resistivity Measurements of the Overprints Resistivity (ρ) was estimated from the resistance measurements and geometrical parameters of the overprints (length and cross-section). Resistance was measured using a two-point probe method with a Keithley 2634B source meter (Keithley Instruments, Inc., Cleveland, OH, USA). The resistivity represents the average of 10 diﬀerent overprints.

2.7. Organic Field-Eﬀect Transistors Donor-acceptor polymer poly[2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole-alt-5,5-(2,5-di (thien-2-yl)thieno[3,2-b]thiophene)] (DPP-DTT) was purchased from Ossila Ltd., Sheffield, UK. DPP-DTT was chosen due to its high stability in ambient conditions. It was dissolved in chloroform (spectroscopic grade) and stirred at 50 ◦C for 1 h. The solution was filtered with a standard polyvinylidene fluoride (PVDF) syringe filter (40 µm). The Si/SiO2 substrates with printed silver electrodes were functionalized with HMDS prior to the deposition of semiconductor polymer film. DPP-DTT thin ﬁlms were fabricated by means of spin coating at 3000 RPM for 30 s and angular acceleration of 3000 RPM/s. Afterwards, the samples were annealed at 200 ◦C for 15 min inside a glovebox (N2 atmosphere), to remove the remaining solvent. OFET measurements were performed by using a Keithley 2634B source meter. Output and transfer characteristics were measured in ambient atmosphere in the VDS and V range from 10 V to 40 V. The charge carrier mobility for holes was derived from the transfer G − characteristics in the saturation regime using standard equations described elsewhere [49]. Five to 10 OFETs were tested for each sample.

3. Results

3.1. Role of Surface Energy One of the crucial factors for printing is the wettability of the ink on the printed surface, because this influences the quality of the overprints and the overall process reliability. Since the printed structures are created by ink droplets, contact angle measurements are an important tool to evaluate the ink–substrate interactions. Fabrication of conductive tracks of sufficiently low electrical resistivity requires continuous overprints with homogenous shapes. To ensure reliability of the printed tracks, straight edges and uniform cross-section of the electrodes are required, especially for a well-defined constant channel length (L) between source and drain contacts of an OFET. Moreover, a smooth shape of the electrodes is important to minimize variations of the interfacial morphology at the metal/semiconductor interface in bottom contact OFETs. Both factors can result in a non-homogenous distribution of the electric field, which can reduce the device performance [50]. To control the quality of the overprints, the influence of surface energy on the DGP wettability was firstly studied. In all cases, sessile droplets deposited on various surface modifications revealed symmetric and regular shapes. The contact angles of H2O and DGP for different surface treatments are summarized in Table1. The contact angle for DGP shows a similar trend as the water reference. The values for the TGME-based DGP ink are smaller, indicating slightly improved wettability in comparison to water. Since TGME is a polar solvent, hydrophobic surfaces do not provide good wettability for the ink. OTS- and HMDS-modified surfaces exhibit the largest contact angle of 42◦ and 51◦, respectively. The substrates washed with CHCl3 and iPrOH reveal smaller values of 32◦ and 34◦, respectively. The smallest contact angle of 25◦ was found for the plasma-treated hydrophilic surface leading to the best wettability of the DGP ink. The effect of surface modification on the morphology of the printed DGP tracks is shown in Figure2, where multiple lines of single DGP overprints are presented on differently modified surfaces. A very high contact angle of DGP on HMDS results in dewetting of the ink and lack of continuous overprints. In the case of OTS modification, the DGP ink forms continuous lines, however, with saw-like edges. We postulate a limit in contact angle between 40◦ and 50◦ for the wettability of the DGP ink. This is supported by previous studies which reported isolated ink drops and no continuous lines of DGP on polydimethylsiloxane (PDMS) substrates at a contact angle of 55.7◦, Materials 2020, 13, 4974 6 of 16 Materials 2020, 13, x FOR PEER REVIEW 6 of 16 regardlesscontact angle the dropof 55.7°, spacing regardless [51]. By the contrast, drop spacing if the contact [51]. By angle contrast, of DGP if isthe low contact enough, angle as inof theDGP case is oflow polyethylenenaphthalate enough, as in the case(PEN) of polyethylenenaphth film (~27◦), wettingalate of the(PEN) substrate film (~27°), allows wetting uniform of overprints the substrate to be obtainedallows uniform [52]. overprints to be obtained [52].

Figure 2. Optical microscope images of DGP overprints on SiO surface with diﬀerent types of Figure 2. Optical microscope images of DGP overprints on SiO2 surface with different types of treatment. Scale bar equal 100 µm. treatment. Scale bar equal 100 μm.

The overprints of DGP on surfaces treated with CHCl3 and iPrOH appear comparable to each The overprints of DGP on surfaces treated with CHCl3 and iPrOH appear comparable to each other, in agreement with their close contact angles. Nevertheless, both cases also exhibit irregular other, in agreement with their close contact angles. Nevertheless, both cases also exhibit irregular shapes with edges moderately better defined than on OTS. Contrary to the other presented surfaces, shapes with edges moderately better defined than on OTS. Contrary to the other presented surfaces, the DGP overprints on plasma-treated substrates reveal linear shapes of the edges. Treatment of the DGP overprints on plasma-treated substrates reveal linear shapes of the edges. Treatment of SiO2 SiO surfaces with plasma does not only clean the surface of adsorbed contaminants [53], but also surfaces2 with plasma does not only clean the surface of adsorbed contaminants [53], but also provides an activation of the surface. This activation can lead to the modification of surface morphology provides an activation of the surface. This activation can lead to the modification of surface (i.e., roughness) [54], or can result in more uniform wetting properties, manifested by homogenous morphology (i.e., roughness) [54], or can result in more uniform wetting properties, manifested by spreading of the ink, like in the case of DGP. From all studied samples, the plasma-treated SiO is the homogenous spreading of the ink, like in the case of DGP. From all studied samples,2 the most suitable for inkjet printing of DGP electrodes and, therefore, was chosen for further studies in plasma-treated SiO2 is the most suitable for inkjet printing of DGP electrodes and, therefore, was this work. chosen for further studies in this work. 3.2. Influence of Printing Parameters on Geometry and Resistivity of the Overprints 3.2. Influence of Printing Parameters on Geometry and Resistivity of the Overprints In the previous section, the influence of the surface energy on the ink wettability was examined. In thisIn section the previous we investigate section, the the influence effect of e-jetof the printing surface parameters energy on the related ink wettability to the electric was field examined. on the printingIn this section quality we to investigate optimize thethe finaleffect electrode of e-jet prin shape.ting parameters A pair of L-shaped related to DGP the electric lines was field printed on the asprinting source quality and drain to optimize electrodes the in final OFETs, electrode with theshape. active A pair channel of L-shaped area between DGP lines the vertical was printed parts ofas printedsource and lines. drain electrodes in OFETs, with the active channel area between the vertical parts of printed lines. 3.2.1. Amplitude of the Voltage Applied 3.2.1. Amplitude of the Voltage Applied The print resolution during inkjet printing is determined by the ink volume ejected from the nozzle.The In print e-jet printing,resolution the during droplet inkjet volume printing depends is determined on the strength by the of theink electricvolume field ejected (E) actingfrom the on thenozzle. meniscus In e-jet at printing, the tip of the the droplet nozzle [volume19]. The de strengthpends on of thethe electricstrength field of the depends electric on field the amplitude(E) acting on of thethe voltagemeniscus applied at the (tipA) of and the on nozzle the distance [19]. The between strength nozzle of the and electric substrate. field depends Throughout on the the amplitude presented studiesof the voltage the nozzle–substrate applied (A) distanceand on wasthe keptdistance constant, between while nozzle the influence and substrate. of the voltage Throughout applied the on thepresented geometry studies of the the overprints nozzle–substrate was investigated. distance was Figure kept3 constant,shows optical while microscopythe influence images of the ofvoltage four overprintsapplied on withthe geometry different values of the ofoverprintsA, equal towas 225 investigated. V, 250 V, 275 Figure V and 3003 shows V, respectively. optical microscopy The base voltageimages (bias,of fourB) wasoverprints kept constant with different at B = 200 values V. The of overprints A, equal revealto 225 continuousV, 250 V, linear275 V shapesand 300 with V, increasingrespectively. width The (Wbase) for voltage increased (bias,A ,B which) was iskept confirmed constant by at surface B = 200 profilometry V. The overprints measurements reveal (cf.continuous cross-section linear scans shapes in Figurewith increasing3b). width (W) for increased A, which is confirmed by surface profilometryNot surprisingly, measurements the larger (cf. volumecross-section of deposited scans in ink Figure is reflected 3b). in an increased maximum height of theNot overprints, surprisingly,hmax. Morethe larger detailed volume analysis of deposi of the profilested ink revealsis reflected a rounded in an shape increased of the maximum overprint upperheight partof the which overprints, evolves h withmax. More the volume detailed of theanalysis deposited of the ink. profiles A small reveals amount a rounded of the deposited shape of inkthe resultsoverprint in aupper symmetric part cross-sectionwhich evolves and with similar the shapevolume of of the the overprints deposited (A ink.= 225 A V, small Figure amount3b). With of anthe increaseddepositedA inkand results larger in volume a symmetric of the depositedcross-section ink, and the cross-sectionsimilar shape of of both the linesoverprints differ and(A = the 225 line V, shapeFigure becomes 3b). With unsymmetrical an increased (e.g.,A andA =larger275V, volume Figure 3ofb). the In particular,deposited theink, point the cross-section of maximum of height both islines shifted differ from and the the center line shape in the becomes direction unsymmetrical away from the prospective (e.g., A = 275 channel V, Figure area. 3b). A possible In particular, reason the of point of maximum height is shifted from the center in the direction away from the prospective channel area. A possible reason of this distorted shape of the contact lines is an unsymmetrical

Materials 2020, 13, 4974 7 of 16

Materials 2020, 13, x FOR PEER REVIEW 7 of 16 thisMaterials distorted 2020, shape13, x FOR of PEER the contactREVIEW lines is an unsymmetrical wetting in two lateral directions related7 of 16 to wettingthe uneven in two charging lateral of directions the substrate. related The to uneventhe uneven charge charging can originate of the substrate. from the presenceThe uneven of another charge wetting in two lateral directions related to the uneven charging of the substrate. The uneven charge canconductive originate line from at thethe closepresen proximity.ce of another conductive line at the close proximity. can originate from the presence of another conductive line at the close proximity.

Figure 3. (a) Optical microscope images of DGP printed lines on plasma-treated SiO2 surface with FigureFigure 3. 3.(a )( aOptical) Optical microscope microscope images images of of DGP DGP printedprinted lineslines on plasma-treated SiOSiO2 2surface surface with with different amplitudes (A). (b) Cross-section profiles of the respective overprints. B = 200 V, f = 500 Hz, differentdifferent amplitudes amplitudes (A ().A ).(b ()b Cross-section) Cross-section profiles profiles ofof thethe respectiverespective overprints.overprints. B B = = 200 200 V, V, f f= =500 500 Hz, Hz, s = 100 µm. Scale bar equal 200 µm. s =s 100 = 100 μm. μm. Scale Scale bar bar equal equal 200 200 μ μm.m. A linear relation between the applied voltage and linewidth (W) as well as channel length (L) is A Alinear linear relation relation between between the the applied applied voltage voltage andand linewidthlinewidth (W)) asas wellwell as as channel channel length length (L ()L is) is found, as depicted by Figure4a. The linewidth of the overprint W is controlled from 37 2 µm for found,found, as asdepicted depicted by by Figure Figure 4a. 4a. The The linewidth linewidth of of thethe overprintoverprint W isis controlledcontrolled from from 37 37 ± ± 2 2±μ μmm for for A A A = 225 V, to 71 2 µm for A = 300 V. Figure4b shows that the maximum height of the overprint ( h ) maxmax = 225= 225 V, V,to to71 71 ± 2± μ2 mμm for for A A = =300 300 V. V. Figure Figure 4b 4b shows shows thatthat thethe maximum heightheight of of the the overprint overprint (h (hmax) ) increases also linearly with A. Thus, we conclude that it is possible to control not only the width, but in increasesincreases also also linearly linearly with with A A. Thus,. Thus, we we conclude conclude thatthat itit isis possiblepossible toto controlcontrol not not only only the the width, width, but but addition the thickness of DGP printed lines. By modulation of the amplitude voltage in the regime of in inaddition addition the the thickness thickness of of DGP DGP printed printed lines. lines. By By modulationmodulation of the amplitudeamplitude voltage voltage in in the the regime regime A fromof A from 225 V 225 to 300V to V, 300 a significant V, a significant change change of hmax offrom hmax 60from2 60 nm ± to2 nm 94 to2 nm,94 ± is2 achieved.nm, is achieved. Overall, of A from 225 V to 300 V, a significant change of hmax from± 60 ± 2 nm± to 94 ± 2 nm, is achieved. theOverall, voltage the amplitude voltage amplitude is determined is determined as a crucial as parameter a crucial parameter for the control for the of control geometrical of geometrical features of Overall, the voltage amplitude is determined as a crucial parameter for the control of geometrical thefeatures overprints. of the overprints. features of the overprints.

FigureFigure 4. 4.Inﬂuence Influence of of the the amplitude amplitude ofof thethe appliedapplied voltage ( (AA)) on on the the geometrical geometrical features features of of the the Figureoverprintsoverprints 4. Influence on on (a) width(a) ofwidth (theW) and(amplitudeW) edge-to-edgeand edge-to-edge of the distance, applied distance, deﬁningvoltage defining the(A) channelon the the channel geometrical length (lengthL), and features (bL),) maximumand of ( bthe) overprintsheightmaximum of the on height overprints (a) width of the( h (overprintsmaxW) ).andB = edge-to-edge 200(hmax V,). fB == 500200 distance, V, Hz, f =s 500= 100 definingHz,µ m.s = 100 the μm. channel length (L), and (b)

maximum height of the overprints (hmax). B = 200 V, f = 500 Hz, s = 100 μm. 3.2.2.3.2.2. Frequency Frequency of of the the Voltage Voltage Waveform Waveform 3.2.2.Another FrequencyAnother printing printing of the parameterVoltage parameter Waveform which which significantlysignificantly influencesinfluences the the distribution distribution of of the the printed printed ink ink is is the frequency of the applied electric field (f ). A triangular shape of the waveform was used throughout theAnother frequency printing of the parameter applied electricwhich significantlyfield (f). A triangularinfluences shapethe distribution of the waveform of the printed was used ink is the presented studies. It is known from previous e-jet studies that the frequency of droplet ejection thethroughout frequency theof presentedthe applied studies. electric It isfield known (f). frAom triangular previous shapee-jet studies of the that waveform the frequency was used of throughoutdroplet ejection the presented depends studies.not only It onis knownthe freque fromncy previous of the voltage e-jet studies waveform, that butthe frequencyalso on the of droplet ejection depends not only on the frequency of the voltage waveform, but also on the

Materials 2020, 13, 4974 8 of 16

Materials 2020, 13, x FOR PEER REVIEW 8 of 16 depends not only on the frequency of the voltage waveform, but also on the viscosity and electrical conductivityviscosity and of electrical the ink [ 38conductivity]. To study theof sizethe distributionink [38]. To of study the DGP the overprintssize distribution by the changeof the ofDGP the frequencyoverprints ofby the the applied change electricof the frequency field, the viscosityof the applied of DGP electric ink was field, kept the constant viscosity since of DGP the printingink was kept constant since the printing was performed at ambient conditions (room temperature (RT) = 23 was performed at ambient conditions (room temperature (RT) = 23 ◦C, humidity = 40%). Figure5 shows°C, humidity the influence = 40%). of Figure the frequency 5 shows of the the influence triangular of voltage the frequency change of for theA = triangular300 V on thevoltage width change of the singlefor A = layer 300 V overprint on the width and on of the the resulting single layer channel overprint length. and For onf = the100 resulting Hz, the width channel equals length. 35 For2 µ fm, = ± and100 Hz, the the channel width length equals for 35 the± 2 neighboringμm, and the overprintchannel length at s = for100 theµm, neighboring equals 65 overprint2 µm. For at slarger = 100 ± frequencies,μm, equals 65 i.e., ± f2= μ500m. For Hz orlargerf = 1000frequencies, Hz, the widthi.e., f = of 500 the Hz overprints or f = 1000 decreases Hz, the to 31width2 µofm the or ± 26overprints2 µm, decreases respectively. to 31 The ± 2 relative μm or reduction26 ± 2 μm, of respectively. the overprint The width relative is 11% reduction and 26%, of the in the overprint case of ± fwidth= 500 is Hz 11% and andf = 26%,1000 in Hz, the representing case of f = 500 significant Hz and f narrowing.= 1000 Hz, representing Simultaneously, significant frequency-driven narrowing. decreasingSimultaneously, of the frequency-driven overprint width resultsdecreasing in the of increase the overprint of the channelwidth results length in to theL = increase69 2 µ mof andthe channel length to L = 69 ± 2 μm and L = 74 ± 2 μm, for intermediate (f = 500 Hz) and high (±f = 1000 Hz) L = 74 2 µm, for intermediate (f = 500 Hz) and high (f = 1000 Hz) f, respectively. This confirms f, respectively.± This confirms the possibility of controlling the geometrical features of the overprints the possibility of controlling the geometrical features of the overprints by the frequency adjustment. by the frequency adjustment. However, it is clear that the influence of the frequency in the tested However, it is clear that the influence of the frequency in the tested regime is minor compared to the regime is minor compared to the magnitude of geometrical changes induced by the amplitude of the magnitude of geometrical changes induced by the amplitude of the waveform. This observation is in waveform. This observation is in agreement with previous report describing the influence of the four agreement with previous report describing the influence of the four main e-jet print parameters (A, B, f main e-jet print parameters (A, B, f and print height) on the droplet size. By using statistical analysis and print height) on the droplet size. By using statistical analysis Mantysalo et al. have shown that the Mantysalo et al. have shown that the A has the strongest effect on the width of the droplets [44]. A has the strongest effect on the width of the droplets [44].

Figure 5. (a) Influence of the frequency of the voltage waveform (f ) on the linewidth (W) of the Figure 5. (a) Influence of the frequency of the voltage waveform (f) on the linewidth (W) of the overprints and on the resulting channel length (L), for s = 100 µm. (b) Optical microscope images for overprints and on the resulting channel length (L), for s = 100 μm. (b) Optical microscope images for f f = 100 Hz, f = 500 Hz and f = 1000 Hz, respectively. A = 300 V, B = 200 V, s = 100 µm. Scale bar = 100 Hz, f = 500 Hz and f = 1000 Hz, respectively. A = 300 V, B = 200 V, s = 100 μm. Scale bar equal 50 equal 50 µm. μm. 3.2.3. Summary of the Geometrical Features of the Overprints 3.2.3. Summary of the Geometrical Features of the Overprints It is now well established that the printing parameters A and f, bear a significant influence on the geometryIt is now (wellW and establishedhmax) of thethat overprints. the printing Since parameters the cross-sectional A and f, bear features a significant strongly influence depend on the printinggeometry parameters (W and hmax (Figure) of the3), overprints. a more general Since picture the cross-se needsctional to be drawn. features Width strongly and maximumdepend on heightthe printing are two parameters geometrical (Figure parameters 3), a more describing general picture the maximum needs to lateral be drawn. and Width horizontal and sizemaximum of the overprints,height are two but ingeometrical order to determine parameters correct describing resistivities the ofmaximum the printed lateral DGP and lines horizontal we need to size establish of the a relationoverprints, between but in the order printing to determine parameters correct and the resistivities cross-section of of the the printed overprints. DGP Figure lines6 summarizeswe need to theestablish relation a relation between between the amplitude the printing of the parameters waveform and the widthcross-section (Figure of6a), the the overprints. maximum Figure height 6 (Figuresummarizes6b) and the the relation cross-section between (Figure the amplitude6c) of the respective of the waveform overprints, and for the di widthfferent (Figuref. 6a), the maximum height (Figure 6b) and the cross-section (Figure 6c) of the respective overprints, for different f.

Materials 2020, 13, x FOR PEER REVIEW 9 of 16 Materials 2020, 13, 4974 9 of 16 Materials 2020, 13, x FOR PEER REVIEW 9 of 16

Figure 6. Summary of the geometrical features of the overprints fabricated with different A and f. (a) Figure 6. Summary of the geometrical features of the overprints fabricated with different A and f. FigureWidth, 6.(b Summary) maximum of heightthe geometrical and (c) cross-section features of the of overprintsthe overprints fabricated increases with for different the higher A and voltage f. (a) (a) Width, (b) maximum height and (c) cross-section of the overprints increases for the higher Width,amplitude. (b) maximum height and (c) cross-section of the overprints increases for the higher voltage voltage amplitude. amplitude. TheThe data data were were collected collected from from diff erentdifferent samples, sample includings, including the overprints the overprints fabricated fabricated with di ffwitherent differentThe frequenciesdata were andcollected interline from distances different s, betweensamples, 100 including and 180 μthem. overprintsThere is a clear fabricated rising trendwith frequencies and interline distances s, between 100 and 180 µm. There is a clear rising trend for each of differentfor each offrequencies the three geometricaland interline features distances (W s/,h betweenmax/cross-section) 100 and 180as A μ m.is increased. There is a Theclear large rising sample trend the three geometrical features (W/hmax/cross-section) as A is increased. The large sample size realized forsize each realized of the by three testing geometrical multiple features input variables(W/hmax/cross-section) (A, f and s) assuggests A is increased. that by Thechoosing large sampleproper by testing multiple input variables (A, f and s) suggests that by choosing proper printing parameters a sizeprinting realized parameters by testing a desired multiple geomet inputry ofvariables overprints (A can, f and be printed. s) suggests that by choosing proper desired geometry of overprints can be printed. printing parameters a desired geometry of overprints can be printed. 3.2.4.3.2.4. ResistivityResistivity ofof thethe OverprintsOverprints 3.2.4. Resistivity of the Overprints SinceSince printedprinted silversilver trackstracks shouldshould serve as the el electrodesectrodes in in electronic electronic devices, devices, their their resistivity resistivity Since printed silver tracks should serve as the electrodes in electronic devices, their resistivity waswas determined determined by by two-point two-point probe probe measurements. measuremen Thets. resistivityThe resistivity was calculated was calculated from the from resistance the wasresistance determined and the by cross-section two-point ofprobe the overprintsmeasuremen. Thets. lengthThe resistivity of the measured was calculated element fromwas keptthe and the cross-section of the overprints. The length of the measured element was kept constant at resistanceconstant at and 1 mm. the Thecross-section results of ofthe the resistivity overprints measurements. The length are of summarizedthe measured in element Figure 7.was Prior kept to 1 mm. The results of the resistivity measurements are summarized in Figure7. Prior to the resistivity constantthe resistivity at 1 mm. calculation The results the ofresistance the resistivity was corrected measurements by the are contact summarized resistance. in Figure Linear 7. tracks Prior toof calculation the resistance was corrected by the contact resistance. Linear tracks of different lengths were thedifferent resistivity lengths calculation were fabricated the resistance and their was resistcorrectedance wasby the extrapolated contact resistance. to the length Linear equal tracks to of0 fabricated and their resistance was extrapolated to the length equal to 0 (Figure S1, Supplementary different(Figure S1,lengths Supplementary were fabricated Materials). and their The resist coancentact was resistance extrapolated from to thethe lengthtwo-point equal probe to 0 Materials). The contact resistance from the two-point probe measurements was estimated to be (Figuremeasurements S1, Supplementary was estimated toMaterials). be 9.2 ± 0.4 The Ω. Thiscontact value resistancewas taken intofrom account the two-pointas an error probevalue 9.2 0.4 Ω. This value was taken into account as an error value with respect to the total resistance, measurementswith± respect to wasthe totalestimated resistance, to be and9.2 ± it 0.4 was Ω .treated This value as an was error taken of the into calculated account as resistivity an error ofvalue the and it was treated as an error of the calculated resistivity of the single overprints (shown in Figure7a,b withsingle respect overprints to the (shown total resistance, in Figure 7a,band asit waserror treated bars). as an error of the calculated resistivity of the as error bars). single overprints (shown in Figure 7a,b as error bars).

Figure 7. Summary of the resistivity measurements for DGP printed single layer overprints as a Figure 7. Summary of the resistivity measurements for DGP printed single layer overprints as a function of their cross-section (a) and their maximum height (b) for diﬀerent A. The blue dashed lines Figurefunction 7. of Summary their cross-section of the resistivity (a) and measurementstheir maximum for height DGP (printedb) for different single layer A. The overprints blue dashed as a represent the resistivity reported by the ink supplier. functionlines represent of their the cross-section resistivity reported (a) and bytheir the maximumink supplier. height (b) for different A. The blue dashed Alllines single represent layer the overprints resistivity revealedreported by aresistivity the ink supplier. (ρ) in the range of 25–250 µΩ cm, i.e., larger than All single layer overprints revealed a resistivity (ρ) in the range of 25–250 ·μΩ∙cm, i.e., larger 12 µΩ cm reported by the ink manufacturer, which is highlighted by blue dashed lines in Figure7. than All·12 singleμΩ∙cm layer reported overprints by the revealed ink manufacturer, a resistivity which (ρ) in isthe highlighted range of 25–250 by blue μΩ∙ dashedcm, i.e., lines larger in As noticed from Figure7a, the overprints with smaller cross-section tend to show larger ρ, in comparison thanFigure 12 7. μΩ∙ As cmnoticed reported from byFigure the 7a,ink the manufacturer, overprints with which smaller is highlighted cross-section by tendblue to dashed show largerlines inρ, to the lines of larger cross-section. Overprints with cross-section smaller than 4 µm2, reveal a wide Figure 7. As noticed from Figure 7a, the overprints with smaller cross-section tend to show larger ρ,

Materials 2020, 13, 4974 10 of 16 spread in resistivity, while for cross-sections above 4 µm2 the values remain almost constant at around 40–50 µΩ cm. From Figure7a, we can conclude that a voltage amplitude above 275 V is needed for · single layer overprints of low resistivity. The minimum resistivity of 25.5 4.9 µΩ cm for the single ± · layer overprints corresponds well with the previously reported resistivity of single DGP overprints on glass after similar thermal sintering [55]. In Figure7b, the resistivity is presented as a function of maximum height of the overprint. For hmax above 90 nm, the resistivity is constant at around 50 µΩ cm, while below 90 nm the data-points are · strongly spread. Analogically to the cross-section, A of at least 300 V is required to fabricate overprints with maximum height of at least 90 nm. This means that hmax is an alternative geometrical feature which enables us to monitor the resistivity and quality of the overprints. By performing the analysis of several process parameters we established a relation between the geometry and resistivity of the printed single layer overprints. Two distinct regimes for the overprint resistance are evident and suﬃciently large A is required to maximize the electrical characteristics which depend on the cross-section or maximum height of the overprints. In the next section we will demonstrate how to further optimize the electrical performance of the printed silver lines by applying multilayer printing.

3.3. Optimization of the Number of Overprints Since we have demonstrated that the electrical resistance of the overprints increases significantly when their cross-section or hmax drops below the respective threshold, it appears particularly important to fabricate overprints with sufficiently large geometrical features. It was shown in previous sections that A is the most powerful parameter for modulating the overprint size. However, simply increasing the amplitude and thus the electric field needs to take into account the quality of the overprints. The linear shape of the overprint edges might be seriously disturbed by a momentum of the droplet impact with the substrate that is too high. Multilayer printing is an alternative way of deposition of a sufficient amount of conductive ink, and the fabrication of overprints with appropriate cross-section and height. To achieve precise geometry control of the overprints, we have performed systematic studies of multilayer printing of the DGP ink. Figure8a shows the optical microscopy images of single ( n = 1), double (n = 2) and triple overprints (n = 3). The values of A, B and f were kept constant at 250 V, 200 V and 100 Hz, respectively. The overprints reveal continuous regular shapes with increasing width (W) for increased n, which is confirmed by cross-section scans shown in Figure8b. The width of single, double and triple overprints was 50 2 µm, 75 2 µm and 82 2 µm, respectively. Similar to applying larger amplitude, performing ± ± ± the additional printing cycle results in larger hmax of the overprint. The maximum height of single, double and triple overprints was 76 2 nm, 100 2 nm and 117 2 nm, respectively. Comparing ± ± ± these values to previously established threshold of 90 nm we expect that the resistivity should drop when comparing overprint n = 1 with n = 2. This is further supported by the fact that the cross-section for single and double overprints is estimated from the profilometry scans to be equal 3.4 µm2 and 6.2 µm2, respectively. Figure9a shows the results of the resistivity measurements, for a series of single and multilayer overprints, as a function of overprint cross-section. Resistivities of the overprints discussed in Figure8 are represented by red data-points. In agreement with the assumptions based on geometrical parameters W and hmax, the resistivity of overprints drops for increased n, namely from 104 µΩ cm · for n = 1, to 37 µΩ cm for n = 2 (circle and square data-points in Figure9a). However, even though · the third layer (n = 3) results in a noticeable increase of W and hmax, the resistivity remains almost unchanged at 36 µΩ cm (triangle data-points in Figure9a). The addition of the second layer of · ink decreases the resistivity by a factor of 3, while printing the third layer does not improve the electrical properties significantly. This behavior may be also encountered when analyzing other series of overprints presented in Figure9, for both smaller ( A = 225 V) and larger (A = 275 V) amounts of ink deposited. Despite significant difference in resistivity of single and double layer DGP overprints, Materials 2020, 13, 4974 11 of 16

Materials 2020, 13, x FOR PEER REVIEW 11 of 16 Materials 2020, 13, x FOR PEER REVIEW 11 of 16 their surface proﬁlometry studies (Figure S2, Supplementary Materials) did not reveal diﬀerences in theirtheir topographytopography onon aa 11 mm mm length. length. TheThe thicknessthickness ofof bothboth overprintsoverprints isis similarly similarly uniformuniform alongalong thethe overprintoverprint without without abrupt abrupt variations. Therefore, Therefore, fa fabricationbrication of of proper proper conductive conductive paths paths of of DGP DGP on plasma-treated Si/SiO2 wafer requires deposition of a double layer and printing of subsequent layers plasma-treated SiSi/SiO/SiO2 wafer requires deposition of a double layer and printing of subsequent layers onlyonly slightly improvesimproves thethe electricalelectrical properties properties of of the thethe overprints. overprints.overprints. Figure FigureFigure9b 9b9b summarizes summarizessummarizes resistivities resistivitiesresistivities of ofthe the printed printed silver silver lines lines versus versus the the number number of of layers layers deposited. deposited. ContraryContrary toto thethe resistivityresistivity of single layerlayer overprints overprints whichwhichwhich varyvaryvary from fromfrom tens tenstens to toto hundreds hundredshundreds of ofofµ μΩ∙Ω cmcm (as (as shown shown before before in in Figure Figure7), 7), printing printing of · ofsuccessive successive layers layers almost almost always always ensures ensures more more constant constant values values at ρat< ρ 50 < 50µΩ μΩ∙cm.cm. ·

Figure 8. (a) Optical microscope images of DGP overprints on plasma-treated SiO2 surface with Figure 8. ((aa)) Optical Optical microscope microscope images images of DGP overprints on plasma-treatedplasma-treated SiOSiO2 surface with differentdiﬀerent numbernumber ofof printedprinted layerslayers ( n(n).).). ( (b(b))) Cross-section Cross-sectionCross-section proﬁles profilesprofiles of ofof the thethe respective respectiverespective overprints. overprints.overprints.A A= 250== 250250 V, V,B = B 200 == 200200 V, V,V,f = ff 100== 100100 Hz, Hz,Hz,s = ss =100= 100100µ m.μm. Scale Scale bar bar equal equal 200 200µ m.μm.

Figure 9. a Figure 9. SummarySummary of of the the resistivity resistivity measurements measurements of single of andsingle multilayer and multilayer overprints. overprints. ( ) Two-point (a)) probe resistivity as a function of the overprint cross-section for different A.(b) Resistivity presented as Two-point probe resistivity as a function of thethe overprintoverprint cross-sectioncross-section forfor differentdifferent A.. ((b)) ResistivityResistivity a function of number of printed layers (n). The blue dashed lines represent the resistivity reported by presented as a function of number of printed layers (n).). TheThe blueblue dasheddashed lineslines representrepresent thethe the ink supplier. B = 200 V, f = 100 Hz. resistivityresistivity reportedreported byby thethe inkink supplier.supplier. B == 200200 V,V, ff == 100100 Hz.Hz. The ultimate step consisted of the application of the DGP overprints as source and drain electrodes The ultimate step consisted of the application of the DGP overprints as source and drain in bottom contact/bottom gate OFETs. The resistivity change between single and multilayer electrodes electrodes in bottom contact/bottom gate OFETs. TheThe resistivityresistivity changechange betweenbetween singlesingle andand was reflected also in the electrical performance of devices containing a similar active layer. A stable multilayer electrodes was reflected also in the electricaltrical performanceperformance ofof devicesdevices containingcontaining aa similarsimilar solution-processed conjugated polymer with high hole mobility (DPP-DTT, [56]) was used as the model active layer. A stable solution-processed conjugated polymer with high hole mobility (DPP-DTT, system in these studies. Figure 10 shows typical transfer and output characteristics of OFETs based on [56])[56]) waswas usedused asas thethe modelmodel systemsystem inin thesethese studies.studies. FigureFigure 1010 showsshows typicaltypical transfertransfer andand outputoutput spin-coated DPP-DTT comprising electrodes with a different number of printed DGP layers. characteristics of OFETs based on spin-coated DPP-DTT comprising electrodes with a different number of printed DGP layers.

Materials 2020, 13, 4974 12 of 16 Materials 2020, 13, x FOR PEER REVIEW 12 of 16

Figure 10.10. (a(a) )Transfer Transfer and and ( (bb)) output output characteristics of of bottom contactcontact/bottom/bottom gate organic field-efield-effectffect transistorstransistors (OFETs)(OFETs) withwith DPP-DTTDPP-DTT (poly[2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole(poly[2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole -alt-5,5-(2,5-di(thien-2-yl)thieno[3,2-b]thiophene)])-alt-5,5-(2,5-di(thien-2-yl)thieno[3,2-b]thiophene)]) as as active layer, comprising single ((n == 1), doubledouble (n = 2) and triple (n = 3) overprints as electrodes. V = 40 V for transfer characteristics. (n = 2) and triple (n = 3) overprints as electrodes. VSDSD = −40− V for transfer characteristics. Transfer characteristics of OFETs exhibit a typical unipolar p-channel behavior (Figure 10a). Transfer characteristics of OFETs exhibit a typical unipolar p-channel behavior (Figure 10a). However, the I-V curve of devices comprising single layer electrodes differ strongly from the ones with However, the I-V curve of devices comprising single layer electrodes differ strongly from the ones double or triple layer electrodes. The ISD at VSD = 40 V is almost two orders of magnitude higher for with double or triple layer electrodes. The ISD at− VSD = −40 V is almost two orders of magnitude multilayer electrodes than for single layer ones. The average hole mobility extracted from OFETs with higher for multilayer electrodes than for3 single2 1 layer1 ones. The average hole mobility extracted from single layer electrodes is µn1 = 4.4 10− cm V− s− , and−3 increases2 −1 by−1 an order of magnitude for devices OFETs with single layer electrodes× is µn12 = 4.42 ×1 101 cm V s , and increases2 2 1 by1 an order of with multilayer electrodes to µn2 = 3.3 10− cm V− s− and µn3 = 4.0 10− cm V− s− . The charge magnitude for devices with multilayer× electrodes to µn2 = 3.3 × 10−2 cm×2 V−1 s−1 and µn3 = 4.0 × 10−2 cm2 carrier−1 −1 mobilities for bottom contact/bottom gate OFETs comprising multilayer electrodes are slightly V s . The charge carrier mobilities for bottom contact/bottom gate OFETs comprising2 multilayer2 1 1 better than devices with thermally evaporated Ag electrodes for which µAg = 1.0 10− cm V− s− electrodes are slightly better than devices with thermally evaporated Ag electrodes× for which µAg = (Figure S3, Supplementary Materials). However, either printed or evaporated silver-based electrodes 1.0 × 10−2 cm2 V−1 s−1 (Figure S3, Supplementary Materials). However, either printed or evaporated reveal mobilities smaller by a factor of 2–3 in comparison with bottom contact/bottom gate OFETs silver-based electrodes reveal mobilities smaller by a factor of 2–3 in comparison with bottom comprising Au electrodes reported previously [57]. The reason is most probably related to improperly contact/bottom gate OFETs comprising Au electrodes reported previously [57]. The reason is most aligned HOMO of DPP-DTT ( 5.2 eV, [56]) and workfunction of printed DGP lines ( 4.24 eV, [52]). probably related to improperly− aligned HOMO of DPP-DTT (−5.2 eV, [56]) and workfunction− of In agreement with the significantly lower ISD for n = 1, the on/off current ratio (ION/IOFF) for OFETs printed DGP lines (−4.24 eV, [52]). In agreement with the significantly lower ISD for n = 1, the on/off comprising single layer electrodes reveals smaller values than for devices with multilayer electrodes. ON OFF current ratio (I /I ) for OFETs comprising single layer electrodes reveals3 smaller values than for While the former exhibits ION/IOFF inferior to 100, for the latter it exceeds 10 . In addition, the threshold devices with multilayer electrodes. While the former exhibits ION/IOFF inferior to 100, for the latter it voltage Vth shifts from 13 V for OFET with n = 1 to around 5 V ( 4 V), for n = 2 (n = 3). exceeds 103. In addition,− the threshold voltage Vth shifts from− −13 V− for OFET with n = 1 to around −5 Overall, we find that the charge carrier transport in thin films of DPP-DTT is limited by the injection V (−4 V), for n = 2 (n = 3). of charge carriers in OFETs comprising single layer electrodes. This is further manifested in Figure 10b, Overall, we find that the charge carrier transport in thin films of DPP-DTT is limited by the which shows the output characteristics of OFETs with electrodes composed of a different number of injection of charge carriers in OFETs comprising single layer electrodes. This is further manifested in DGP layers. The devices with double and triple DGP layers (n = 2 and n = 3, respectively) exhibit a Figure 10b, which shows the output characteristics of OFETs with electrodes composed of a different clear linear region starting at V = 0 V, and a successive saturation region at higher negative voltage. number of DGP layers. The devicesSD with double and triple DGP layers (n = 2 and n = 3, respectively) By contrast, the source-drain current for OFETs with single layer electrodes (Figure S4, Supplementary exhibit a clear linear region starting at VSD = 0 V, and a successive saturation region at higher Materials) does not cross the abscissa at the origin and a superlinear behavior at low source-drain negative voltage. By contrast, the source-drain current for OFETs with single layer electrodes (Figure voltage is observed. This behavior is attributed to contact resistance between the source/drain electrodes S4, Supplementary Materials) does not cross the abscissa at the origin and a superlinear behavior at and the semiconducting layer [58]. In addition, the output characteristics for single layer electrodes low source-drain voltage is observed. This behavior is attributed to contact resistance between the do not exhibit a clear saturation region in the investigated voltage range, and source-drain currents source/drain electrodes and the semiconducting layer [58]. In addition, the output characteristics for are two orders of magnitude smaller than for multilayer electrode devices. Altogether, we notice a single layer electrodes do not exhibit a clear saturation region in the investigated voltage range, and qualitative and quantitative difference in the I-V characteristics of OFETs comprising single layer and source-drain currents are two orders of magnitude smaller than for multilayer electrode devices. multilayer electrodes, represented by a large difference in all key performance parameters, µ, V and Altogether, we notice a qualitative and quantitative difference in the I-V characteristics of OFETsth I /I . Addition of the second layer leads to a great improvement in the electrical performance of comprisingON OFF single layer and multilayer electrodes, represented by a large difference in all key the OFETs. The ink wetting on the plasma-treated substrate leads to the low thickness and finally to performance parameters, µ, Vth and ION/IOFF. Addition of the second layer leads to a great high resistivity of single-layer overprints. The percolation path for silver overprints with thickness improvement in the electrical performance of the OFETs. The ink wetting on the plasma-treated smaller than the estimated h threshold of 90 nm hinders an efficient injection of charge carriers substrate leads to the low thicknessmax and finally to high resistivity of single-layer overprints. The from the electrode to the active layer. percolation path for silver overprints with thickness smaller than the estimated hmax threshold of 90 nm hinders an efficient injection of charge carriers from the electrode to the active layer.

Materials 2020, 13, 4974 13 of 16

4. Conclusions In this work we have optimized the printing parameters for the realization of conductive tracks, either by applying sufficiently large voltage amplitude or by deposition of multilayers. The influence of several printing parameters on the geometrical features of the silver overprints was investigated, confirming that the amount of deposited ink is crucial for the reliability of electrodes. The precise deposition of the DGP ink was achieved by combining the extraordinary jetting capabilities of the electrohydrodynamic printer with the adjusted wettability of the substrate. A detailed exploration of the relation between geometry and resistivity of printed tracks enabled us to establish threshold values for the cross-section and maximum height, above which resistivity dropped significantly. Finally, it was shown that the geometry-dependent properties of the printed electrodes as source/drain bottom contacts in OFETs determine the device performance. A large contact resistance was observed for single layer DGP OFETs. Key performance parameters for OFETs (µ, Vth and ION/IOFF) differed by orders of magnitude for devices comprising single layer and multilayer electrodes. Control over the overprint geometry ensures an efficient injection of charge carriers from contacts to the active semiconducting film. The results presented here further deepen knowledge about the precise fabrication of electrodes, combining surface science and device engineering. In the near future, we plan to further optimize the fabrication of printed OFETs by including interlayers and performing patterning.

Supplementary Materials: The following are available online at http://www.mdpi.com/1996-1944/13/21/4974/s1, Figure S1: The length-dependent resistance measurements for estimation of the contact resistance; Figure S2: Surface profile of single (n = 1, top) and double (n = 2, bottom) DGP overprints; Figure S3: (a) Transfer and (b) output characteristics of bottom contact/bottom gate OFETs with DPP-DTT as active layer, comprising thermally evaporated Ag electrodes. VSD = 40 V for transfer characteristics; Figure S4: Output characteristics of bottom contact/bottom gate OFETs with DPP-DTT− as active layer, comprising single (n = 1), double (n = 2) and triple (n = 3) overprints as electrodes. Author Contributions: Research design, sample fabrication, experiments and data analysis, writing of original draft and editing, P.S.; manuscript review and editing, M.B., H.Z., J.U., W.P., T.M.; supervision, W.P., T.M.; funding acquisition, W.P., T.M. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by the Foundation for Polish Science under the European Regional Development Fund (POIR.04.04.00-00-3ED8/17-01). W.P. acknowledges National Science Centre, Poland, through the grant UMO-2015/18/E/ST3/00322. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Suh, Y.H.; Shin, D.W.; Chun, Y.T. Micro-to-nanometer patterning of solution-based materials for electronics and optoelectronics. RSC Adv. 2019, 9, 38085–38104. [CrossRef] 2. Birkholz, M.; Mai, A.; Wenger, C.; Meliani, C.; Scholz, R. Technology modules from micro- and nano-electronics for the life sciences. Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 2016, 8, 355–377. [CrossRef] 3. Zhang, L.; Tang, Y.; Tong, L. Micro-/nanoﬁber optics: Merging photonics and material science on nanoscale for advanced sensing technology. iScience 2020, 23, 100810. [CrossRef] 4. Rodriguez, S. Redeﬁning microfabrication of high-precision optics: How two-photon grayscale lithography improves quality and throughput of printing microparts. PhotonicsViews 2020, 17, 36–39. [CrossRef] 5. Alvarez, M.M.; Aizenberg, J.; Analoui, M.; Andrews, A.M.; Bisker, G.; Boyden, E.S.; Kamm, R.D.; Karp, J.M.; Mooney, D.J.; Oklu, R.; et al. Emerging trends in micro- and nanoscale technologies in medicine: From basic discoveries to translation. ACS Nano 2017, 11, 5195–5214. [CrossRef] 6. Betancourt, T.; Brannon-Peppas, L. Micro- and nanofabrication methods in nanotechnological medical and pharmaceutical devices. Int. J. Nanomed. 2006, 1, 483–495. [CrossRef] 7. Imboden, M.; Bishop, D. Top-down nanomanufacturing. Phys. Today 2014, 67, 45–50. [CrossRef] 8. Garner, C.M. Lithography for enabling advances in integrated circuits and devices. Philos. Trans. Soc. A 2012, 370, 4015–4041. [CrossRef] 9. Okazaki, S. Resolution limits of optical lithography. J. Vac. Sci. Technol. B 1991, 9, 2829–2833. [CrossRef] Materials 2020, 13, 4974 14 of 16

10. Stringer, J.; Althagathi, T.M.; Tse, C.W.C.; Ta, V.D.; Shephard, J.D.; Esenturk, E.; Connaughton, C.; Wasley, T.J.; Li, J.; Kay, R.W.; et al. Integration of additive manufacturing and inkjet printed electronics: A potential route to parts with embedded multifunctionality. Manuf. Rev. 2016, 3, 12. [CrossRef] 11. Zolfaghari, A.; Chen, T.; Yi, A.Y. Additive manufacturing of precision optics at micro and nanoscale. Int. J. Extrem. Manuf. 2019, 1, 12005. [CrossRef] 12. Matsui, H.; Takeda, Y.; Tokito, S. Flexible and printed organic transistors: From materials to integrated circuits. Org. Electron. 2019, 75, 105432. [CrossRef] 13. Gao, M.; Li, L.; Song, Y. Inkjet printing wearable electronic devices. J. Mater. Chem. C 2017, 5, 2971–2993. [CrossRef] 14. Teichler, A.; Perelaer, J.; Schubert, U.S. Inkjet printing of organic electronics—comparison of deposition techniques and state-of-the-art developments. J. Mater. Chem. C 2013, 1, 1910–1925. [CrossRef] 15. Singh, M.; Haverinen, H.M.; Dhagat, P.; Jabbour, G.E. Inkjet printing—process and its applications. Adv. Mater. 2010, 22, 673–685. [CrossRef] 16. Kuang, M.; Wang, L.; Song, Y. Controllable printing droplets for high-resolution patterns. Adv. Mater. 2014, 26, 6950–6958. [CrossRef][PubMed] 17. Basaran, O.A. Small-scale free surface flows with breakup: Drop formation and emerging applications. AIChE J. 2002, 48, 1842–1848. [CrossRef] 18. Sutanto, E.; Shigeta, K.; Kim, Y.K.; Graf, P.G.; Hoelzle, D.J.; Barton, K.L.; Alleyne, A.G.; Ferreira, P.M.; Rogers, J.A. A multimaterial electrohydrodynamic jet (E-jet) printing system. J. Micromech. Microeng. 2012, 22, 45008. [CrossRef] 19. Park, J.U.; Hardy, M.; Kang, S.J.; Barton, K.; Adair, K.; Mukhopadhyay, D.K.; Lee, C.Y.; Strano, M.S.; Alleyne, A.G.; Georgiadis, J.G.; et al. High-resolution electrohydrodynamic jet printing. Nat. Mater. 2007, 6, 782–789. [CrossRef] 20. Wilkinson, N.J.; Kay, R.W.; Harris, R.A. Electrohydrodynamic and aerosol jet printing for the copatterning of polydimethylsiloxane and graphene platelet inks. Adv. Mater. Technol. 2020, 5, 2000148. [CrossRef] 21. Lee, K.H.; Lee, S.S.; Ahn, D.B.; Lee, J.; Byun, D.; Lee, S.Y. Ultrahigh areal number density solid-state on-chip microsupercapacitors via electrohydrodynamic jet printing. Sci. Adv. 2020, 6, eaaz1692. [CrossRef] 22. Brossard, F.S.F.; Pecunia, V.; Ramsay, A.J.; Griffiths, J.P.; Hugues, M.; Sirringhaus, H. Inkjet-printed nanocavities on a photonic crystal template. Adv. Mater. 2017, 29, 1704425. [CrossRef][PubMed] 23. Hales, S.; Tokita, E.; Neupane, R.; Ghosh, U.; Elder, B.; Wirthlin, D.; Kong, Y.L. 3D printed nanomaterial-based electronic, biomedical, and bioelectronic devices. Nanotechnology 2020, 31, 172001. [CrossRef] 24. Sun, J.; Bao, B.; He, M.; Zhou, H.; Song, Y. Recent advances in controlling the depositing morphologies of inkjet droplets. ACS Appl. Mater. Interfaces 2015, 7, 28086–28099. [CrossRef] 25. Mampallil, D.; Burak Eral, H. A review on suppression and utilization of the coffee-ring effect. Adv. Colloid Interface Sci. 2018, 252, 38–54. [CrossRef] 26. Tian, D.; Song, Y.; Jiang, L. Patterning of controllable surface wettability for printing techniques. Chem. Soc. Rev. 2013, 42, 5184–5209. [CrossRef] 27. Vafaei, S.; Tuck, C.; Ashcroft, I.; Wildman, R. Surface microstructuring to modify wettability for 3D printing of nano-filled inks. Chem. Eng. Res. Des. 2016, 109, 414–420. [CrossRef] 28. Moon, Y.J.; Kang, H.; Lee, S.H.; Kang, K.; Cho, Y.J.; Hwang, J.Y.; Moon, S.J. Effect of contact angle and drop spacing on the bulging frequency of inkjet-printed silver lines on FC-coated glass. J. Mech. Sci. Technol. 2014, 28, 1441–1448. [CrossRef] 29. Soltman, D.; Subramanian, V. Inkjet-printed line morphologies and temperature control of the coffee ring effect. Langmuir 2008, 24, 2224–2231. [CrossRef][PubMed] 30. Liu, C.F.; Lin, Y.; Lai, W.Y.; Huang, W. Improved performance of inkjet-printed Ag source/drain electrodes for organic thin-film transistors by overcoming the coffee ring effects. AIP Adv. 2017, 7, 115008. [CrossRef] 31. Fukuda, K.; Sekine, T.; Kumaki, D.; Tokito, S. Profile control of inkjet printed silver electrodes and their application to organic transistors. ACS Appl. Mater. Interfaces 2013, 5, 3916–3920. [CrossRef] 32. Prasetyo, F.D.; Yudistira, H.T.; Nguyen, V.D.; Byun, D. Ag dot morphologies printed using electrohydrodynamic (EHD) jet printing based on a drop-on-demand (DOD) operation. J. Micromech. Microeng. 2013, 23, 95028. [CrossRef] 33. Shin, K.Y.; Lee, S.H.; Oh, J.H. Solvent and substrate effects on inkjet-printed dots and lines of silver nanoparticle colloids. J. Micromech. Microeng. 2011, 21, 045012. [CrossRef] Materials 2020, 13, 4974 15 of 16

34. Chiolerio, A.; Rivolo, P.; Porro, S.; Stassi, S.; Ricciardi, S.; Mandracci, P.; Canavese, G.; Bejtka, K.; Pirri, C.F. Inkjet-printed PEDOT:PSS electrodes on plasma-modified PDMS nanocomposites: Quantifying plasma treatment hardness. RSC Adv. 2014, 4, 51477–51485. [CrossRef] 35. Huang, Y.; Jiang, L.; Li, B.; Premaratne, P.; Jiang, S.; Qin, H. Study effects of particle size in metal nanoink for electrohydrodynamic inkjet printing through analysis of droplet impact behaviors. J. Manuf. Process. 2020, 56, 1270–1276. [CrossRef] 36. Cui, Z.; Han, Y.; Huang, Q.; Dong, J.; Zhu, Y. Electrohydrodynamic printing of silver nanowires for flexible and stretchable electronics. Nanoscale 2018, 10, 6806–6811. [CrossRef] 37. Yu, J.H.; Kim, S.Y.; Hwang, J. Effect of viscosity of silver nanoparticle suspension on conductive line patterned by electrohydrodynamic jet printing. Appl. Phys. A 2007, 89, 157–159. [CrossRef] 38. An, S.; Lee, M.W.; Kim, N.Y.; Lee, C.; Al-Deyab, S.S.; James, S.C.; Yoon, S.S. Effect of viscosity, electrical conductivity, and surface tension on direct-current-pulsed drop-on-demand electrohydrodynamic printing frequency. Appl. Phys. Lett. 2014, 105, 214102. [CrossRef] 39. Wang, K.; Stark, J.P.W. Direct fabrication of electrically functional microstructures by fully voltage-controlled electrohydrodynamic jet printing of silver nano-ink. Appl. Phys. A 2010, 99, 763–766. [CrossRef] 40. Qin, H.; Wei, C.; Dong, J.; Lee, Y.S. Direct printing and electrical characterization of conductive micro-silver tracks by alternating current-pulse modulated electrohydrodynamic jet printing. J. Manuf. Sci. Eng. 2017, 139, 021008. [CrossRef] 41. Yuan, X.; Xiong, Z. High frequency pulsed electrohydrodynamic printing with controllable fine droplets. J. Micromech. Microeng. 2018, 28, 095008. [CrossRef] 42. Wei, C.; Qin, H.; Ramirez-Iglesias, N.A.; Chiu, C.P.; Lee, Y.S.; Dong, J. High-resolution ac-pulse modulated electrohydrodynamic jet printing on highly insulating substrates. J. Micromech. Microeng. 2014, 24, 045010. [CrossRef] 43. Park, J.; Park, J.W.; Nasrabadi, A.M.; Hwang, J. Methodology to set up nozzle-to-substrate gap for high resolution electrohydrodynamic jet printing. Appl. Phys. Lett. 2016, 109, 134104. [CrossRef] 44. Laurila, M.M.; Khorramdel, B.; Dastpak, A.; Mantysalo, M. Statistical analysis of E-jet print parameter effects on Ag-nanoparticle ink droplet size. J. Micromech. Microeng. 2017, 27, 095005. [CrossRef] 45. Guo, X.; Puniredd, S.R.; Baumgarten, M.; Pisula, W.; Mullen, K. Rational design of benzotrithiophene- diketopyrrolopyrrole-containing donor-acceptor polymers for improved charge carrier transport. Adv. Mater. 2013, 25, 5467–5472. [CrossRef] 46. Tsao, H.N.; Pisula, W.; Liu, Z.; Osikowicz, W.; Salaneck, W.R.; Mullen, K. From ambi- to unipolar behavior in discotic dye field-effect transistors. Adv. Mater. 2008, 20, 2715–2719. [CrossRef] 47. Moon, S.J.; Robin, M.; Wenlin, K.; Yann, M.; Bae, B.S.; Mohammed-Brahim, T.; Jacques, E.; Harnois, M. Morphological impact of insulator on inkjet-printed transistor. Flex. Print. Electron. 2017, 2, 035008. [CrossRef] 48. Certificate of Analysis Issued by Advanced Nano Products, Co. Ltd. Available online: http://www.anapro. com/eng/product/silver_inkjet_ink.html (accessed on 20 September 2020). 49. Horowitz, G. Organic thin film transistors: From theory to real devices. J. Mater. Res. 2004, 19, 1946–1962. [CrossRef] 50. Hamadani, B.H.; Richter, C.A.; Gundlach, D.J.; Kline, R.J.; McCulloch, I.; Heeney, M. Influence of source-drain electric field on mobility and charge transport in organic field-effect transistors. J. Appl. Phys. 2007, 102, 044503. [CrossRef] 51. Kim, Y.; Ren, X.; Kim, J.W.; Noh, H. Direct inkjet printing of micro-scale silver electrodes on polydimethylsiloxane (PDMS) microchip. J. Micromech. Microeng. 2014, 24, 115010. [CrossRef] 52. Mitra, D.; Mitra, K.Y.; Dzhagan, V.; Pillai, N.; Zahn, D.R.T.; Baumann, R.R. Workfunction and conductivity of inkjet-printed silver layers: Effect of inks and post-treatments. J. Electron. Mater. 2018, 47, 2135–2142. [CrossRef] 53. Mashayekhi, M.; Winchester, L.; Evans, L.; Pease, T.; Laurila, M.M.; Mantysalo, M.; Ogier, S.; Teres, L.; Carrabina, J. Evaluation of aerosol, superfine inkjet, and photolithography printing techniques for metallization of application specific printed electronic circuits. IEEE Trans. Electron Devices 2016, 63, 1246–1253. [CrossRef] 54. Alam, A.U.; Howlader, M.M.R.; Deen, M.J. The effects of oxygen plasma and humidity on surface roughness, water contact angle and hardness of silicon, silicon dioxide and glass. J. Micromech. Microeng. 2014, 24, 035010. [CrossRef] Materials 2020, 13, 4974 16 of 16

55. Ryu, K.; Moon, Y.J.; Park, K.; Hwang, J.Y.; Moon, S.J. Electrical property and surface morphology of silver nanoparticles after thermal sintering. J. Electron. Mater. 2016, 45, 312–321. [CrossRef] 56. Li, J.; Zhao, Y.; Tan, H.S.; Guo, Y.; Di, C.A.; Yu, G.; Liu, Y.; Lin, M.; Lim, S.H.; Zhou, Y.; et al. A stable solution-processed polymer semiconductor with record high-mobility for printed transistors. Sci. Rep. 2012, 2, 754. [CrossRef] 57. Xi, Y.; Wolf, C.M.; Pozzo, L.D. Self-assembly of donor-acceptor conjugated polymers induced by miscible ‘poor’ solvents. Soft Matter 2019, 15, 1799–1812. [CrossRef] 58. Mazur, L.; Castiglione, A.; Ocytko, K.; Kameche, F.; Macabies, R.; Ainsebaa, A.; Kreher, D.; Heinrich, B.; Donnio, B.; Sanaur, S.; et al. Charge carrier mobility study of a mesogenic thienothiophene derivative in bulk and thin ﬁlms. Org. Electron. 2014, 15, 943–953. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional aﬃliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).