# Fileset [MurataPhysRevAppl2024.pdf](https://mdr.nims.go.jp/filesets/3b9fbb25-f12a-42ec-ae5a-7de644acf6a3/download) ## Creator Keito Murata, Gyo Kitahara, Satoru Inoue, Toshiki Higashino, Satoshi Matsuoka, [Shunto Arai](https://orcid.org/0000-0002-0055-3006), Reiji Kumai, Tatsuo Hasegawa ## Rights [Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/) ## Other metadata [Stability of ternary interfaces and its effects on ideal switching characteristics in inverted coplanar organic transistors](https://mdr.nims.go.jp/datasets/9e5e0cf5-fb10-4370-854f-bba28898ba61) ## Fulltext Stability of ternary interfaces and its effects on ideal switching characteristics in inverted coplanar organic transistorsPHYSICAL REVIEW APPLIED 21, 024005 (2024)Stability of ternary interfaces and its effects on ideal switching characteristics ininverted coplanar organic transistorsKeito Murata ,1,* Gyo Kitahara ,1 Satoru Inoue ,1 Toshiki Higashino ,2 Satoshi Matsuoka ,1Shunto Arai ,1,† Reiji Kumai ,3 and Tatsuo Hasegawa 11Department of Applied Physics, The University of Tokyo, Tokyo 113-8656, Japan2National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba 305-8565, Japan3Photon Factory, Institute of Materials Structure Science, High Energy Accelerator Research Organization(KEK), Tsukuba 305-0801, Japan (Received 3 May 2023; revised 18 July 2023; accepted 22 December 2023; published 2 February 2024)Inverted coplanar or bottom-gate bottom-contact (BGBC)-type thin-film transistors (TFTs) presentseveral advantages for the manufacture and application of organic TFTs, although serious difficultiesare encountered when trying to achieve sufficiently high performance. Recently, it was demonstratedthat both high mobility and ideal ON-OFF switching are attainable in BGBC-type printed organic TFTswith highly clean semiconductor-gate dielectric interfaces. However, an unknown channel materialdependence in the device performance is found. Here, we show that the stability of semiconduc-tor/metal/dielectric ternary interfaces is a crucial factor in the operation of BGBC-type organic TFTs. Wefabricate single-crystal organic semiconductor (OSC) films with various numbers of layers using two dif-ferent materials (phenyl/alkyl-substituted benzothieno[3,2-b]benzothiophene and phenyl/alkyl-substitutedbenzothieno[3,2-b]naphtho[2,3-b]thiophene) on highly lyophobic Cytop gate dielectric surfaces. Thetransfer characteristics exhibit notable time-dependent degradation, which clearly depends on the material,layer number, and encapsulation. Kelvin-probe force microscopy measurements reveal that the degrada-tion is ascribed to contact resistance at the source electrodes, while it can be more suppressed in multilayer(two or more layers) OSCs. Atomic force microscopy and in-plane x-ray diffraction profiles present signsof the transformation in single molecular bilayer OSCs laid on the electrodes. The results suggest theimportance of the quality of the OSC layer at ternary interfaces, providing a clue for improving theperformance of BGBC-type organic TFTs.DOI: 10.1103/PhysRevApplied.21.024005I. INTRODUCTIONFor wide application of organic thin-film transistors(TFTs), improvements are required in the performanceindexes of electrical characteristics, such as device mobil-ity, switching sharpness, operation voltage, and opera-tion stability. It is expected that the parasitic resistanceof the semiconductor bulk arising between the channeland the source or drain electrodes can be reduced byadopting coplanar [bottom-gate bottom-contact (BGBC) ortop-gate top-contact] TFTs, which are in principle more*murata@hsgw.t.u-tokyo.ac.jp†Present address: National Institute for Materials Science(NIMS), Tsukuba, Ibaraki 305-0044, Japan.Published by the American Physical Society under the terms ofthe Creative Commons Attribution 4.0 International license. Fur-ther distribution of this work must maintain attribution to theauthor(s) and the published article’s title, journal citation, andDOI.advantageous than staggered [bottom-gate top-contact(BGTC) or top-gate bottom-contact] TFTs. This is sim-ply because vertical transport across the channel layer isnecessary in the latter [1]. Additionally, inverted copla-nar or BGBC-type TFTs are more promising for practicaluses, because this type of device should allow a variety ofmanufacturing processes to be used for the stacking andpatterning of electrode and/or dielectric components with-out having adverse effects on the organic semiconductor(OSC) channel layer [2–4].Meanwhile, it is known to be quite difficult to obtainBGBC-type organic TFTs that exhibit sufficiently highperformance, as compared to BGTC-type organic TFTsthat are utilized as standard. It has been pointed outfor vacuum-deposited polycrystalline BGBC-type organicTFTs that the crystal growth mode of the OSC layer onan electrode surface or edge is distinct from that on agate dielectric surface, which eventually limits the car-rier injection and device performance [2,5–8]. Further-more, there are often serious unknown difficulties when2331-7019/24/21(2)/024005(11) 024005-1 Published by the American Physical Societyhttps://orcid.org/0000-0002-4757-9940https://orcid.org/0000-0003-3046-3332https://orcid.org/0000-0002-1943-1594https://orcid.org/0000-0002-9227-8207https://orcid.org/0000-0003-3357-8282https://orcid.org/0000-0002-0055-3006https://orcid.org/0000-0002-5320-0028https://orcid.org/0000-0001-5187-7433https://crossmark.crossref.org/dialog/?doi=10.1103/PhysRevApplied.21.024005&domain=pdf&date_stamp=2024-02-02http://dx.doi.org/10.1103/PhysRevApplied.21.024005https://creativecommons.org/licenses/by/4.0/KEITO MURATA et al. PHYS. REV. APPLIED 21, 024005 (2024)trying to achieve high performance in BGBC-type organicTFTs composed of either polycrystalline or single-crystalOSC films fabricated either by solution or vacuum pro-cessing [9–11]. One problem is that the origin of theperformance deterioration is not understood, although afew recent studies have reported high device performancein BGBC-type organic TFTs [3,12–14]. The followingaspects have been proposed as the origin of the deteriora-tion: (1) the applied electric field from the gate does notwork effectively for carrier injection [15–17]; (2) spacecharge accumulates at the conduction path [18–20]; and (3)other effects [4,7,21–26]. Thus far, a way of constructingideal BGBC-type organic TFTs has not been clarified orestablished.Recently, it was reported that both high mobility andideal ON-OFF switching are attainable in solution-processedBGBC-type organic TFTs [10]. A notable feature is thatthe devices exhibit extremely sharp switching character-istics close to the theoretical limit, which is ascribed tothe formation of highly clean semiconductor-gate dielec-tric interfaces with the use of the highly lyophobic Cytop.The production of such interfaces was enabled by adopt-ing the extended meniscus-guided (EMG) coating tech-nique. However, it was also shown that the device mobil-ity strongly depends on the OSC channel material. Themobility value is more than 10 times lower than that ofBGTC-type TFTs for phenyl/alkyl-substituted benzothieno[3,2-b]benzothiophene (Ph-BTBT-Cn) [27–29], while acomparably high value is obtained for phenyl/alkyl-substituted benzothieno[3,2-b]naphtho[2,3-b]thiophene(Ph-BTNT-Cn) [30–32]. [The molecular structures ofPh-BTBT-Cn and Ph-BTNT-Cn are shown in Fig.1(a)]. We consider that the observed channel materialdependence is associated with the unknown problemat the semiconductor-metal-dielectric ternary interfaces,and thus offers a unique opportunity to clarify why manyBGBC-type TFTs fail to operate.In this study, we investigate the operations of sharp-switching BGBC-type organic TFTs based on both Ph-BTBT-Cn and Ph-BTNT-Cn. We focus on the dependenceof the device characteristics on the number of layers ofsingle-crystal OSC films and on their temporal variations.We find that the devices composed of single molecularbilayer films of Ph-BTBT-Cn exhibit rapid and notabledegradation of the device characteristics, while the degra-dation is evidently suppressed by using multilayer (twoor more layers) films, Ph-BTNT-Cn films, or by encap-sulation. We use Kelvin-probe force microscopy (KPFM)to investigate the electric potential variation along thechannel, and we also investigate the OSC film mor-phology around the ternary interfaces using atomic forcemicroscopy (AFM) and x-ray diffraction (XRD) mea-surements. Based on the results, we discuss the originof the degradation in the device characteristics, and theimportant roles of the ternary interfaces in BGBC-typeorganic TFTs.(a) (d)(b)(e)(c)FIG. 1. Single-crystal TFTs with various numbers of layers of OSC films. (a) Chemical structures of Ph-BTBT-Cn and Ph-BTNT-Cn.(b) Schematic cross section of TFTs. (c) Optical image of various numbers of layers of OSC films with a mixed solution of Ph-BTBT-Cn. The red dashed line shows the boundary between regions with uniform thickness of OSC films. The black dashed line rectangleshows one of the TFTs with a uniform layer number. (d) Transfer characteristics of four layers of Ph-BTBT-Cn TFTs. The mobilityvalue is estimated as 9.1 cm2 V–1 s–1. The SS value is extracted by the linear fit of current slope in the range from |I d| = 10−11.5 to10−10 A. (e) Layer-number dependence of the device mobility in the linear regime several days after device fabrication.024005-2STABILITY OF TERNARY INTERFACES . . . PHYS. REV. APPLIED 21, 024005 (2024)II. EXPERIMENTA. Manufacturing BGBC-type TFTs with differentnumbers of layersFigure 1(b) schematically shows a cross section of theBGBC-type organic TFTs that we produce. We use heav-ily p-doped silicon plates with a 100-nm-thick surfaceSiO2 layer as substrates. The Cytop layer is spin-coated (at2000 rpm for 60 s) using a solution of Cytop (CTL-809M;AGC Inc., Japan) diluted with the solvent (CT-Solv.180;AGC Inc., Japan) at a ratio of 1:4. The Cytop layer is thencured in a vacuum oven (at 50 °C for 10 min at 0.08 MPa,80 °C for 15 min at 0.02 MPa, and 180 °C for 60 min at0.02 MPa). To form the source and drain electrodes and U-shaped metal guide required for the EMG coating (see Sup-plemental Material Fig. S1 [33]), 0.5-nm-thick Cr and 25-nm-thick Au are vacuum deposited through a shadow maskon top of the Cytop layer. The electrode surface is modifiedby chemical vapor treatment with pentafluorobenzenethiol(PFBT) (>95.0%; Tokyo Chemical Industry Co., Ltd.,Japan).The OSC materials are synthesized according to theliterature [30,34]. It is reported that the use of a mixed solu-tion of OSCs with different alkyl chain lengths effectivelysuppresses the multiple stacking of molecular bilayer films[10,35]. Thus, we use a mixture of Ph-BTBT-C10 andPh-BTBT-C14 for fabricating Ph-BTBT-Cn films, and amixture of Ph-BTNT-C10 and Ph-BTNT-C12 for fabricat-ing Ph-BTNT-Cn films. The respective compounds aredissolved in chlorobenzene (Wako Special Grade; FujifilmWako Pure Chemical Corp., Japan) at 0.05 or 0.1 wt %.The solutions are mixed at different mixing ratios tocontrol the layer-number thickness of single-crystal OSCfilms by the moderate suppression of the multiple stack-ing of bilayer units. For example, we utilize a volumeratio of 9:1 (Ph-BTBT-C10:Ph-BTBT-C14) to fabricatesingle molecular bilayer films (hereafter referred to assingle-layer films), and a ratio of 97:3 to obtain filmswith several layers (hereafter referred to as multilayerfilms), as presented in Fig. S1 in the Supplemental Mate-rial [33]. Finally, the EMG coating technique is usedto obtain uniform larger-area single-crystal OSC filmson the Cytop surface under ambient pressure and tem-perature. The fabricated films are dried under ambienttemperature.The TFT channel is isolated from unnecessary partsby removing the outside OSC films with uniform thick-ness using a micromanipulator (Axis Pro; Micro Sup-port Co., Ltd., Japan), as shown in Fig. S2 in theSupplemental Material [33]. Optical microscope imagesand crossed-Nicols micrographs are collected with adigital microscope (VHX-6000; Keyence Corp., Japan).The layer number of the channel semiconductor layeris determined using AFM (VN-8010; Keyence Corp.,Japan).B. Device and thin-film characterizationThe electrical characteristics are measured in a glovebox filled with N2 gas using a semiconductor parameteranalyzer (E5270A; Agilent Technologies Inc., USA). Thechannel length L is 100 μm. The channel width W variesfrom device to device, ranging from 100 to 800 μm, asthe areas with uniform thickness are selected from thesingle-crystal OSC films [see Figs. 1(c) and S2 in the Sup-plemental Material [33]]. The gate capacitance is estimatedas 22 nF cm−2, composed of the series capacitance ofCytop and SiO2, with relative permittivity 2.0 and 3.9 andthickness 30 and 100 nm. To encapsulate the OSC chan-nel layer of the TFTs, we use another Cytop layer that isspin-coated with a solution diluted with solvent to a ratioof 1:1 with solvent on top of the devices and heated at atemperature below the glass transition point (108 °C). Thethickness of the encapsulating film is about 200 nm.KPFM measurements and observation of the surfacemorphology of the OSC films are conducted with MFP-3D (Asylum Research, USA). The in-plane XRD mea-surements are conducted by a thin-film diffractometer(SOR-SmartLab; Rigaku Co., Ltd., Japan) installed inthe beamline BL-7C at the Photon Factory, High-EnergyAccelerator Research Organization (KEK). Diffractionintensity is recorded using a scintillation counter.III. RESULTS AND DISCUSSIONA. Material and layer-number dependence of devicemobilityWe successfully obtain OSC films with various layernumbers on top of the Cytop layer and source and drainelectrodes for both Ph-BTBT-Cn and Ph-BTNT-Cn, aspresented in Figs. 1(c) and S2 in the Supplemental Mate-rial [33]. The OSC films are stacked only in bilayerunits within very large single-domain crystal films, asconfirmed by AFM measurements and crossed-Nicolspolarized microscope observation. The BGBC-type TFTswith different layer-number thicknesses are fabricated byextracting the OSC films with uniform thickness areas, asshown in Fig. 1(c). We find that almost all devices exhibitturn-on voltage (Von) around 0 V, sharp switching withthe subthreshold swing (SS) value of 60–80 mV dec−1,and negligible hysteresis, regardless of the thickness orthe OSC material, as shown in Fig. 1(d). We observe cleardependence of the ON current on the thickness of the OSCfilms, associated with the difference in the device mobility.Figure 1(e) summarizes layer-number dependence of thedevice mobility for both Ph-BTBT-Cn and Ph-BTNT-CnTFTs, measured a few days after device fabrication. Thesingle-layer TFTs of Ph-BTNT-Cn exhibit higher mobil-ity than those of Ph-BTBT-Cn, as consistent with a pre-vious report [10]. Importantly, multilayer TFTs exhibitmuch higher mobility than single-layer TFTs based on024005-3KEITO MURATA et al. PHYS. REV. APPLIED 21, 024005 (2024)the same OSCs. Note that the difference owing to thelayer-number dependence is much larger than the differ-ence expected from the in-plane anisotropy of the crystalfilms [35]. It is worth mentioning that the multilayer TFTsof Ph-BTBT-Cn exhibit a higher mobility (average valueof 6 cm2 V−1 s−1) than those of Ph-BTNT-Cn (averagevalue of 4 cm2 V−1 s−1), for which the material depen-dence is reversed compared with the case of single-layerTFTs. Nonetheless, the material dependence of the mul-tilayer TFTs is rather similar to that of the BGTC-typesingle-crystal TFTs [27–31].B. Time-dependent TFT characteristics and effects ofencapsulationFigures 2(a) and S3 in the Supplemental Material [33]show the temporal variation in the transfer characteris-tics of the respective TFTs with different layer numbersor OSC materials, measured repeatedly after device fabri-cation. No voltage is applied between the measurements.The characteristics exhibit a strong time dependence, anddepend on the layer number and materials, as seen in Fig.2(b). Single-layer TFTs composed of Ph-BTBT-Cn exhibitmobility at 4 cm2 V−1 s−1 immediately after device fabri-cation, whereas a gradual but significant degradation of theON current is observed within 200 h. In contrast, the mobil-ity of multilayer (three layers) TFTs is much more stable,with the initial high mobility of 8 cm2 V−1 s−1 being main-tained for up to 200 h. The distribution of mobility valuesis estimated at 2–9 cm2 V−1 s−1 for both the single-layerTFTs measured immediately after device fabrication andthe multilayer TFTs (see also Fig. S4 in the SupplementalMaterial [33]). Previous studies have reported a mobility ofover 10 cm2 V−1 s−1 for Ph-BTBT-C10 [12,27–29], whichis somewhat larger but close to the value observed in thisstudy.For Ph-BTNT-Cn, the speed of degradation of the devicemobility in single-layer TFTs varies greatly from deviceto device, as shown in Fig. S5 in the SupplementalMaterial [33]. The degradation of some devices proceedsmuch more slowly than for Ph-BTBT-Cn. The resultsdemonstrate that the observed layer-number and materialdependence is associated with the difference in the speedof degradation. In contrast to the time-dependent degrada-tion of mobility, we find that the turn-on voltage, switchingsharpness, and hysteresis exhibit a high temporal stabil-ity; low-voltage operation below 3 V with a small SS(a)(b) (c)FIG. 2. Transfer characteristics ofTFTs with various OSC materials andlayer numbers. All devices are storedunder nitrogen atmosphere. (a) Transfercurves in the linear regime measured atvarious times after device fabrication. Theschematic of each device structure abovethe gate dielectric layer is also shown. (b)Temporal variation in the device mobilityextracted from the transfer curves. (c)Temporal variation in the SS value of thesingle-layer Ph-BTNT-Cn TFTs.024005-4STABILITY OF TERNARY INTERFACES . . . PHYS. REV. APPLIED 21, 024005 (2024)value of less than 70 mV dec−1 [Fig. 2(c)] is achieved.Even in the case of single-layer Ph-BTBT-Cn, the SS valuefor most of the devices remains less than 70 mV dec−1for up to 150 h. The results lead us to conclude that thevaried time-dependent degradation of mobility in BGBC-type TFTs causes the apparent difference of the devicemobility shown in Fig. 1(e), whereas the subthresholdcharacteristics are kept stable.We also investigate the atmospheric effects on the tem-poral variation in the TFT characteristics. The single-layerPh-BTBT-Cn TFTs are subjected to various atmospheres(air, N2 gas, and vacuum), and the temporal variation inthe device characteristics are measured. We also measurethe temporal variation for TFTs encapsulated by Cytoplayers [Fig. 3(a)]. We did not observe any damage onthe semiconductor layer due to the encapsulation process[see Fig. S6(a) in the Supplemental Material [33]]. Wefind that the degradation occurs at any atmospheric con-ditions up to 200 h, but not in the case of the encapsulatedTFTs, as shown in Figs. 3(b) and S6 in the SupplementalMaterial [33]. The encapsulated TFTs exhibit exception-ally high stability while maintaining the initial mobility.Furthermore, we also observe an interesting feature in thatthe degradation of mobility is gradually saturated in airafter 100 h, whereas this characteristic is further degradedin nitrogen gas for both Ph-BTBT-Cn and Ph-BTNT-Cn,as seen in Figs. 3(c) and S7 in the Supplemental Mate-rial [33]. In addition, drain or gate bias stress betweeneach measurement does not affect degradation speed inthe device characteristics (see Fig. S8 in the Supplemen-tal Material [33]). Note that these results contrast with thegeneral understanding that the degradation of TFT charac-teristics is attributed to the effects of exposure to ambientair and bias stress [36,37].C. Potential drop at ternary interfaces in BGBC-typeTFTsTo investigate the origin of the degradation in devicecharacteristics, we measure the local surface electricpotential distribution along the channels using KPFM,which allows us to discriminate and compare the channeland contact resistance [38–40]. First, we use single-layerPh-BTBT-Cn TFTs and single-layer Ph-BTNT-Cn TFTsin the KPFM measurements. Figure 4 shows the poten-tial profiles through the channel area in the linear regime(Vd= −0.5 V, Vg=−3 V). In the respective measurements,we confirm that the drain potential is 0.5 V (=|Vd|) lowerthan the source potential, while the potential profiles arechanged considerably by the change in the gate voltageVg, as seen in Fig. S9 in the Supplemental Material [33].We observe that the electric potential decreases graduallyalong the channel in both the single-layer Ph-BTBT-CnTFTs and Ph-BTNT-Cn TFTs immediately after devicefabrication, as seen in Fig. 4(a). After 78 h, a potential(a)(b) (c)FIG. 3. Transfer characteristics of single-layer Ph-BTBT-CnTFTs under various atmospheres. (a) Device structures abovethe gate dielectric layer. (b),(c) Temporal variation in the devicemobility. TFTs stored in a laboratory, in a glove box filledwith N2 gas, and in a chamber (10−4 Pa) are shown as “Air”,“Nitrogen,” and “Vacuum”, respectively.drop around the boundary of the source electrode andchannel increase significantly and the potential in the chan-nel become almost flat in the single-layer Ph-BTBT-CnTFT. The single-layer Ph-BTNT-Cn TFT also exhibits asimilar potential drop at the source electrode edge after79 h, although the drop is clearly smaller than that of thesingle-layer Ph-BTBT-Cn TFT. The results indicate thatthe effective drain bias applied to the channel decreaseswith time in the single-layer TFTs, which proceeds morerapidly in Ph-BTBT-Cn TFTs than in Ph-BTNT-Cn TFTs.To investigate the layer-number dependence of thepotential distribution, we conduct KPFM measurementsof the single-layer and multilayer (four layers) Ph-BTBT-Cn TFTs that exhibit mobility of 0.02 cm2 V−1 s−1 and6.0 cm2 V−1 s−1, respectively. Figure 4(b) shows thepotential profiles for single-layer and multilayer TFTsbased on Ph-BTBT-Cn and Ph-BTNT-Cn. We observe alarge and a small potential drop around the edges of thesource and the drain electrodes, respectively, in the case ofthe single-layer Ph-BTBT-Cn TFT. In contrast, the poten-tial drop is negligibly small in the case of the multilayerPh-BTBT-Cn TFT, where the drain potential uniformlyand gradually decreases along the channel. As seen on thetwo-dimensional maps of the surface potential shown inFig. S10 in the Supplemental Material [33], these potentialdrops around the contacts are uniformly observed in thesingle-layer TFTs but not in multilayer TFTs. A similartrend is also observed in the saturation regime (Vd=−3 V,Vg= −3 V) (see Fig. S9 in the Supplemental Material[33]).The potential drop observed at the electrode edge in thesingle-layer Ph-BTNT-Cn TFTs is not considerable, butis slightly larger than that in the multilayer (two layers)024005-5KEITO MURATA et al. PHYS. REV. APPLIED 21, 024005 (2024)(a) (b)FIG. 4. Surface potential distribution in the linear regime (Vd=−0.5 V, Vg=−3 V) along the TFT channel. (a) Time dependenceof surface potential of specific (upper) single-layer Ph-BTBT-Cn and (lower) single-layer Ph-BTNT-Cn TFTs. (b) Layer-numberdependence of surface potential of (upper) Ph-BTBT-Cn and (lower) Ph-BTNT-Cn TFTs, respectively. Ph-BTBT-Cn and single-layerPh-BTNT-Cn TFTs are measured within 20–40 h. The multilayer Ph-BTNT-Cn TFTs are measured at approximately 250 h. “S” and“D” correspond to source and drain electrodes, respectively. Each dashed line represents the boundary between the electrode and thechannel.TFTs. The layer-number dependence is basically simi-lar between Ph-BTBT-Cn and Ph-BTNT-Cn, whereas thepotential drop in Ph-BTNT-Cn is much smaller than thatin Ph-BTBT-Cn. These features are consistent with thetrend of the device mobility; the TFTs with higher mobil-ity exhibit a smaller potential drop. Thus, we concludethat the degradation of the device mobility depending onlayer number and material is attributed to the contactresistance. This means that the contact resistance shouldincrease with time, especially in single-layer Ph-BTBT-Cnand Ph-BTNT-Cn TFTs.To evaluate the sheet resistance of the OSC chan-nels separately from the contact resistance, we also applythe transfer line method (TLM) for the single-layer andmultilayer Ph-BTBT-Cn TFTs. The results are shown inFig. S11 in the Supplemental Material [33]. From theresults, we find that the slope in the TLM plot (corre-sponding to the sheet resistance) remains constant withtime, whereas the intercept at L = 0 μm (correspondingto the contact resistance) increases gradually with time.We evaluate the sheet resistance as 3.0−3.3 M�/sq forthe single layer and as 1.8−1.9 M�/sq for the multi-layer. The contact resistance increases more rapidly in thesingle layer than in the multilayer. All the results indi-cate that the mobility degradation is mainly caused by thetime-dependent contact resistance, and not by the sheetresistance.D. Transformation of OSC films on metal electrodesTo further investigate the origin of the degradation,we conduct a structural characterization of OSC filmsdeposited both on Cytop and PFBT-modified Au electrodesurfaces. Figures 5(a)–5(f) and S12 in the SupplementalMaterial [33] show the surface topography map measuredby AFM for single-layer and multilayer (five layers) filmsof Ph-BTBT-Cn. We find that the images exhibit a rela-tively rough surface morphology in the case of a singlelayer on electrodes, whereas the surfaces remain flat andhomogeneous for single-layer films on Cytop and multi-layer films on both Cytop and electrodes. The morphologyof the single-layer film on electrodes is composed of a rela-tively rough “base region” [observed as gray-colored areasin Figs. 5(a) and 5(e)], accompanied by “raised particles,”which are granular areas with a certain thickness (observedas white-colored areas in the same figures). The raised par-ticles are 10–200 nm in lateral size and the step heightfrom the base region is estimated at approximately 5 nm, asseen in Fig. 5(g). Interestingly, the step height is close tothe thickness of a single molecular bilayer (5.3 nm). Thelateral size and number of these raised particles increasewith time, as seen in Fig. 5(h). This feature contrasts withthe fairly stable surface morphology for a single layer onCytop. It is most probable that the additional bilayer areagradually grows on top of the base region. On the otherhand, we investigate whether the base region is a surface of024005-6STABILITY OF TERNARY INTERFACES . . . PHYS. REV. APPLIED 21, 024005 (2024)(a) (b)(c) (d)(e) (f)(g) (h)(i)FIG. 5. Morphology of (a),(c),(e) single-layerand (b),(d),(f) multilayer (five layers) of Ph-BTBT-Cn films (a),(b) around contact edge,(c),(d) on Cytop, and (e),(f) on electrodes. Imagesof the single-layer film show the temporal varia-tion of a specific area. Those of the multilayer filmare obtained more than one month after devicefabrication. Color scale is common for each row.(g) Height profiles of the blue dashed lines onthe right-hand image of (e). The black dashedline at 5.3 nm corresponds to the thickness of asingle layer. (h) Statistical distribution extractedfrom (e). The average height of the base regioncorresponds to 0 nm. (i) Schematic illustration ofstructure of single-layer films. Single-layer film ispartially defective, disordered, and stacked.an OSC layer or a bare electrode by using crossed-Nicolsmicroscope observations and polarized absorption mea-surements. We observe a clear optical anisotropy owing toPh-BTBT-Cn over the entire electrode surface with identi-cal peak energy at about 3.3 eV [41], as seen in Figs. S13and S14 in the Supplemental Material [33]. Therefore, it isassumed that the base region on the electrode is not a bareelectrode surface but instead involves a single layer of Ph-BTBT-Cn with retained crystalline anisotropy but possiblya more disordered arrangement, as shown in Fig. 5(i).We conduct in-plane XRD measurements more than100 h after device fabrication to investigate thestructural difference between Ph-BTBT-Cn on electrodesand on Cytop. Single-domain crystal films are fabricatedand extracted for the XRD measurements, as shown inTable I and Fig. 6. We first search for (020) in-planediffractions, and then scan the ϕ angle with 2θ fixed ata specific angle corresponding to each (020) diffraction.Sharp Bragg reflections are observed for all the films, indi-cating that the Ph-BTBT-Cn films are composed mainly ofa single domain with aligned lattice spacing even for thesingle layer on electrodes (see also Fig. S15 in the Sup-plemental Material [33]). However, the peak position ofthe scattering vector is clearly different for the single layer024005-7KEITO MURATA et al. PHYS. REV. APPLIED 21, 024005 (2024)(a) (b)(c)FIG. 6. In-plane x-ray diffraction (XRD) measurements of Ph-BTBT-Cn films. (a) XRD profile of single layer on PFBT/Aucollected by scanning the ϕ angle with 2θ fixed at 19.88° corre-sponding to the (020) diffraction. (b) XRD profile of single layeron PFBT/Au and on Cytop collected by scanning the 2θ anglewith ϕ fixed at the optimal angle. q is the scattering vector. Theinset is the layered-herringbone packing motif of Ph-BTBT-C10.(c) Crossed-Nicols micrograph of the single layer on PFBT/Au.The yellow dashed line shows a single-domain crystal film, asused for XRD measurements.on electrodes compared to that on Cytop, as presented inFig. 6(b). The b-axis length of the single layer on elec-trodes is estimated at 7.98 Å, which is slightly but clearlylonger than that of the bulk crystal (7.76 Å) [42], whereasthe value is almost the same for the single layer on Cytop(7.77 Å) and for the multilayer on electrodes (7.70 Å).Additionally, in ϕ-angle scanning, the peak width of thesingle layer on electrodes is wider than that on Cytop, indi-cating that the crystallinity of the single layer on electrodesis worse than that on Cytop.Based on the results of the AFM and XRD measure-ments, we conclude that the single-layer films of Ph-BTBT-Cn should undergo a gradual but qualitative changeon electrodes, even though the single-crystalline natureseems to be preserved. We consider that the gradual changein the OSC single layer should be responsible for thecontact resistance as proved by KPFM, and also for thetime-dependent degradation of the device mobility.TABLE I. b-axis length extracted from 2θ scan of XRD. Thevalue of bulk crystal is referenced from the literature [42].Single layeron CytopSinglelayer onPFBT/AuMultilayeronPFBT/AuBulkcrystalb-axislength(Å)7.77 7.98 7.70 7.76E. On the material and layer-number dependence ofTFT characteristicsBased on the experimental results presented thus far, wediscuss the origin of the material and layer-number depen-dence of the TFT characteristics. We first point out that thepossible strain of the OSC layer due to the step of thickelectrodes (25 nm) can be excluded for the observed per-formance deterioration in the BGBC-type organic TFTsused in this study when we consider the uniform poten-tial drop at the contact along the channel width direction(see Fig. S10 in the Supplemental Material [33]) and thegradual and reproducible degradation of the device mobil-ity (see Fig. S5 in the Supplemental Material [33]). Wenote that the height slope at the electrode edge should havea slight angle, considering the edge width of about 1 μmagainst the edge height of 25 nm.Thus, it is reasonable to consider that the degradationof the device characteristics should be associated with thegradual and qualitative change in the ultrathin single layeron the PFBT/Au electrode surfaces caused by the rel-atively high surface energy or surface roughness of theelectrodes. It is most probable that the material and layer-number dependence of device mobility is related to thedifference in fragility or robustness of the OSC layer on theelectrodes, which causes the increase in the contact resis-tance at the semiconductor-metal-dielectric ternary inter-face and eventually causes the time-dependent degradationof the device characteristics. The qualitative change in thesingle layer of Ph-BTBT-Cn on electrodes is accompa-nied by a slight variation in the lattice constants while thecrystalline anisotropy is maintained. Although the detailsof the molecular arrangement on the electrodes is not yetclear, we conjecture that the as-grown film having a qual-ity that differs from that of the bulk crystal is the originfor the motion of OSC molecules and eventually causesthe gradual film transformation, including the growth ofraised particles on top of the base region. Although mor-phological or structural evolution has been observed invacuum-evaporated and spin-coated OSC films [43,44],it is worth noting that a similar phenomenon occurs insingle-crystalline films growing at the air-liquid interfaceof the meniscus [45–47]. As additional evidence of thequalitative change in the single layer, we confirm that thedegradation is accelerated by heating single-layer TFTsat temperatures lower than the temperature of the liquid-crystalline transition of Ph-BTBT-Cn or for the depositionof PFBT (see Fig. S16 in the Supplemental Material [33])[30]. This result is consistent with our discussion con-sidering that the film transformation is expected to beaccelerated by the thermal motion of OSC molecules.The degradation proceeds more slowly in some devicesof the single-layer Ph-BTNT-Cn TFTs than in the single-layer Ph-BTBT-Cn TFTs (see Fig. S5 in the SupplementalMaterial [33]). This feature indicates the more robustnature of the single layer of Ph-BTNT-Cn compared with024005-8STABILITY OF TERNARY INTERFACES . . . PHYS. REV. APPLIED 21, 024005 (2024)that of Ph-BTBT-Cn, where the larger fused ring num-ber is likely to be associated with the stability of theOSC layer. Additionally, significant temporal stability isobtained in the multilayer TFTs, and a similarly highstability is achieved in the encapsulated TFTs. We canconsider that the bottom single bilayer, which is locatedclosest to the electrodes and Cytop layer, works as theactual channel layer, and that the additional OSC layeror the encapsulating layer should effectively suppress thequalitative change on electrodes by restricting the motionof molecules in the bottom single bilayer. These argumentsimply that the carrier conduction path is mostly restrictedto the bottom single bilayer, which is consistent with theresults shown in Fig. S4(a) in the Supplemental Mate-rial [33]: single-layer and multilayer TFTs exhibit similarhighest mobility values (9 cm2 V−1 s−1). Furthermore,the effects of the atmosphere on the device characteristicspresent a rather peculiar behavior but can be understood interms of the above scenario; the degradation in air becomesgradually saturated, whereas the degradation proceeds fur-ther in nitrogen gas or in vacuum, as shown in Figs. 3(c)and S7 in the Supplemental Material [33]. We considerthat in the former case the absorption of some atmosphericmolecules on the OSC layer or electrodes should restrictthe motion of OSC molecules like an encapsulating layer,and would eventually suppress the qualitative change inthe single layer.Finally, we discuss the observed difference in the effecton the device mobility and the subthreshold characteristics;the device mobility exhibits severe time-dependent degra-dation, whereas the temporal variation in the value of SSand Von is negligibly small. Similar features are observedin all the devices used in this study. Theoretically, the SSvalue is correlated with the density of carrier trap states atthe semiconductor-gate dielectric interface and in the semi-conductor bulk [48,49]. Therefore, the observed invariantfeature of the SS value implies the highly stable nature ofthe semiconductor-gate dielectric interface in the channelregion. Regarding the effect of the contact resistance on SS,some studies have reported that the contact resistance andSS were improved simultaneously by the insertion of themetal oxide layer [50], whereas other studies have reportedthat atmospheric exposure of the source and drain elec-trodes or adoption of different gate dielectrics causes thechange in the contact resistance without affecting the SSvalue [21,51]. All these results clearly demonstrate thatthe origin of the contact resistance can be clearly separatedfrom that of the SS values in BGBC-type TFTs when weutilize an extremely clean semiconductor-gate dielectricinterface with the highly lyophobic gate dielectric layer.IV. CONCLUSIONWe successfully reveal the origin of the considerablechannel material dependence of the device mobilityobserved in extremely sharp-switching BGBC-type organicTFTs based on solution-processed, single-crystalline, andultrathin OSC layers of Ph-BTBT-Cn and Ph-BTNT-Cn.We find that the device mobility crucially depends onthe layer-number thickness of the OSC layers, and thatthe difference in the device mobility originates from thetime-dependent degradation of the OSC layers on the elec-trodes. In particular, the single-layer Ph-BTBT-Cn TFTsexhibit severe time-dependent degradation, and this fea-ture can be ascribed to the gradual and qualitative changein the ultrathin OSC layer on the electrode surfaces asobserved by AFM and XRD measurements, although thesingle-crystalline nature of the single layer of Ph-BTBT-Cnis maintained as confirmed by crossed-Nicols microscopeobservation, polarized absorption spectra, and XRD mea-surements. The gradual and qualitative change in the OSClayer on the electrodes would be responsible for the contactresistance at the semiconductor-metal-dielectric ternaryinterface as observed by KPFM, which eventually causesthe time-dependent degradation of the device mobility inBGBC-type TFTs. The observed qualitative transforma-tion can be ascribed to the fragile nature of the ultrathinOSC layer on the electrodes, which depends on the OSCmaterial. We also find that the degradation can be sup-pressed by using multilayer OSC films or encapsulatingthe channel OSC layer.Our results demonstrate that the stability of the ternaryinterface is quite important in realizing the high deviceperformance of BGBC-type organic TFTs as the mostpromising device configuration towards practical deviceapplications. Understanding the stability of the ultrathinOSC layer at the ternary interface and the correlation ofthe component OSC molecules is crucial for attaining highperformance and printed electronics applications.ACKNOWLEDGMENTSWe are grateful to Dr. Mitsuhiro Ikawa for his help in thedevice fabrication (University of Tokyo). The synchrotronx-ray experiments were performed with the approvalof the Photon Factory Program Advisory Committee(2020S2-001). This work was supported by the Japan Sci-ence and Technology Agency (JST) CREST Grant No.JPMJCR18J2 and by the Japan Society for the Promotionof Science (JSPS) KAKENHI Grants No. 21H04651, No.20H05867, No. 20J10479, No. 21K05209, No. 21H05234,No. 22H01933, and MEXT Leading Initiative for Excel-lent Young Researchers Grant No. JPMXS0320220012.[1] C. R. Kagan and P. Andry, Thin-Film Transistors (CRCPress, London, 2003).[2] I. Kymissis, C. D. Dimitrakopoulos, and S. Purushothaman,High-performance bottom electrode organic thin-film tran-sistors, IEEE Trans. Electron Devices 48, 1060 (2001).024005-9https://doi.org/10.1109/16.925226KEITO MURATA et al. PHYS. REV. APPLIED 21, 024005 (2024)[3] J. W. Borchert, U. Zschieschang, F. Letzkus, M. Giorgio, R.T. Weitz, M. Caironi, J. N. Burghartz, S. Ludwigs, and H.Klauk, Flexible low-voltage high-frequency organic thin-film transistors, Sci. Adv. 6, eaaz5156 (2020).[4] X. Ye, X. Zhao, S. Wang, Z. Wei, G. Lv, Y. Yang, Y. Tong,Q. Tang, and Y. Liu, Blurred electrode for low contact resis-tance in coplanar organic transistors, ACS Nano 15, 1155(2021).[5] D. J. Gundlach, L. Zhou, J. A. Nichols, T. N. Jackson, P.V. Necliudov, and M. S. Shur, An experimental study ofcontact effects in organic thin film transistors, J. Appl. Phys.100, 024509 (2006).[6] P. Cosseddu and A. Bonfiglio, A comparison between bot-tom contact and top contact all organic field effect transis-tors assembled by soft lithography, Thin Solid Films 515,7551 (2007).[7] J. W. Borchert, R. T. Weitz, S. Ludwigs, and H. Klauk, Acritical outlook for the pursuit of lower contact resistancein organic transistors, Adv. Mater. 34, 2104075 (2022).[8] R. Miyata, S. Inoue, K. Nakajima, and T. Hasegawa, Insu-lating polymer blend organic thin-film transistors basedon bilayer-type alkylated benzothieno[3,2-b]naphtho[2,3-b]thiophene, ACS Appl. Mater. Interfaces 14, 17719(2022).[9] T. Kadoya, O. Pitayatanakul, and T. Mori, Suppression ofaccess resistance using carbon electrodes in organic transis-tors based on alkyl-substituted thienoacene, Org. Electron.21, 106 (2015).[10] G. Kitahara, S. Inoue, T. Higashino, M. Ikawa, T. Hayashi,S. Matsuoka, S. Arai, and T. Hasegawa, Meniscus-controlled printing of single-crystal interfaces showingextremely sharp switching transistor operation, Sci. Adv. 6,eabc8847 (2020).[11] P. Cosseddu and A. Bonfiglio, Soft lithography fabricationof all-organic bottom-contact and top-contact field effecttransistors, Appl. Phys. Lett. 88, 023506 (2006).[12] H. Iino, T. Usui, and J. I. Hanna, Liquid crystals for organicthin-film transistors, Nat. Commun. 6, 6828 (2015).[13] J. W. Borchert, B. Peng, F. Letzkus, J. N. Burghartz, P.K. L. Chan, K. Zojer, S. Ludwigs, and H. Klauk, Smallcontact resistance and high-frequency operation of flexi-ble low-voltage inverted coplanar organic transistors, Nat.Commun. 10, 1119 (2019).[14] A. Bilgaiyan, S. Cho, M. Abiko, K. Watanabe, andM. Mizukami, Solution processed organic transistorson polymeric gate dielectric with mobility exceeding15 cm2 V−1 s−1, Phys. Status Solidi RRL 14, 2000156(2020).[15] L. Bürgi, R. H. Friend, and H. Sirringhaus, Close look atcharge carrier injection in polymer field-effect transistors,J. Appl. Phys. 94, 6129 (2003).[16] C. H. Shim, F. Maruoka, and R. Hattori, Structural analysison organic thin-film transistor with device simulation, IEEETrans. Electron Devices 57, 195 (2010).[17] T. J. Richards and H. Sirringhaus, Analysis of the con-tact resistance in staggered, top-gate organic field-effecttransistors, J. Appl. Phys. 102, 094510 (2007).[18] T. Richards and H. Sirringhaus, Bias-stress induced contactand channel degradation in staggered and coplanar organicfield-effect transistors, Appl. Phys. Lett. 92, 023512 (2008).[19] M. Koehler, I. Biaggio, and M. G. E. Da Luz, Resolving thecontact voltage dilemma in organic field effect transistors,Phys. Rev. B 78, 155312 (2008).[20] L. Merces, R. F. de Oliveira, H. L. Gomes, and C. C.Bof Bufon, The role of the electrode configuration onthe electrical properties of small-molecule semiconductorthin-films, Org. Electron. 49, 107 (2017).[21] T. Zimmerling and B. Batlogg, Improving charge injectionin high-mobility rubrene crystals: From contact-limited tochannel-dominated transistors, J. Appl. Phys. 115, 164511(2014).[22] Y. Xu, H. Sun, and Y. Y. Noh, Schottky barrier in organictransistors, IEEE Trans. Electron Devices 64, 1932 (2017).[23] M. Gruber, E. Zojer, F. Schürrer, and K. Zojer, Impact ofmaterials versus geometric parameters on the contact resis-tance in organic thin-film transistors, Adv. Funct. Mater. 23,2941 (2013).[24] K. Zojer, E. Zojer, A. F. Fernandez, and M. Gruber, Impactof the capacitance of the dielectric on the contact resistanceof organic thin-film transistors, Phys. Rev. Appl. 4, 044002(2015).[25] S. W. Jo, S. Cho, and C. H. Kim, Key factors affecting con-tact resistance in coplanar organic thin-film transistors, J.Phys. D: Appl. Phys. 55, 405101 (2022).[26] K. Nomoto, N. Hirai, N. Yoneya, N. Kawashima, M.Noda, M. Wada, and J. Kasahara, A high-performanceshort-channel bottom-contact OTFT and its application toAM-TN-LCD, IEEE Trans. Electron Devices 52, 1519(2005).[27] J. M. Cho and T. Mori, Low-temperature band transport andimpact of contact resistance in organic field-effect transis-tors based on single-crystal films of Ph-BTBT-C10, Phys.Rev. Appl. 5, 064017 (2016).[28] T. Hamai, S. Arai, H. Minemawari, S. Inoue, R. Kumai,and T. Hasegawa, Tunneling and origin of large accessresistance in layered-crystal organic transistors, Phys. Rev.Appl. 8, 054011 (2017).[29] J. Zeng, D. He, J. Qiao, Y. Li, L. Sun, W. Li, J. Xie, S.Gao, L. Pan, P. Wang, et al., Ultralow contact resistance inorganic transistors via orbital hybridization, Nat. Commun.14, 324 (2023).[30] S. Inoue, S. Shinamura, Y. Sadamitsu, S. Arai, S. Hori-uchi, M. Yoneya, K. Takimiya, and T. Hasegawa, Extendedand modulated thienothiophenes for thermally durableand solution-processable organic semiconductors, Chem.Mater. 30, 5050 (2018).[31] T. Higashino, S. Inoue, Y. Sadamitsu, S. Arai, S. Horiuchi,and T. Hasegawa, Bilayer-type layered herringbone pack-ing in 3-n-octyl-9-phenyl-benzothieno[3,2-b]naphtho[2, 3-b]thiophene, Chem. Lett. 48, 453 (2019).[32] T. Hamai, S. Inoue, S. Arai, and T. Hasegawa, Trap-statesuppression and band-like transport in bilayer-type organicsemiconductor ultrathin single crystals, Phys. Rev. Mater.4, 074601 (2020).[33] See Supplemental Material at http://link.aps.org/supplemental/10.1103/PhysRevApplied.21.024005 for additionalexperimental results.[34] S. Inoue, H. Minemawari, J. Tsutsumi, M. Chikamatsu, T.Yamada, S. Horiuchi, M. Tanaka, R. Kumai, M. Yoneya,and T. Hasegawa, Effects of substituted alkyl chain length024005-10https://doi.org/10.1126/sciadv.aaz5156https://doi.org/10.1021/acsnano.0c08122https://doi.org/10.1063/1.2215132https://doi.org/10.1016/j.tsf.2006.11.182https://doi.org/10.1002/adma.202104075https://doi.org/10.1021/acsami.2c01676https://doi.org/10.1016/j.orgel.2015.02.024https://doi.org/10.1126/sciadv.abc8847https://doi.org/10.1063/1.2166487https://doi.org/10.1038/ncomms7828https://doi.org/10.1038/s41467-019-09119-8https://doi.org/10.1002/pssr.202000156https://doi.org/10.1063/1.1613369https://doi.org/10.1109/TED.2009.2035540https://doi.org/10.1063/1.2804288https://doi.org/10.1063/1.2825584https://doi.org/10.1103/PhysRevB.78.153312https://doi.org/10.1016/j.orgel.2017.06.041https://doi.org/10.1063/1.4873635https://doi.org/10.1109/TED.2017.2650216https://doi.org/10.1002/adfm.201203250https://doi.org/10.1103/PhysRevApplied.4.044002https://doi.org/10.1088/1361-6463/ac8124https://doi.org/10.1103/PhysRevApplied.5.064017https://doi.org/10.1103/PhysRevApplied.8.054011https://doi.org/10.1038/s41467-023-36006-0https://doi.org/10.1021/acs.chemmater.8b01339https://doi.org/10.1246/cl.181038https://doi.org/10.1103/PhysRevMaterials.4.074601http://link.aps.org/supplemental/10.1103/PhysRevApplied.21.024005STABILITY OF TERNARY INTERFACES . . . PHYS. REV. APPLIED 21, 024005 (2024)on solution-processable layered organic semiconductorcrystals, Chem. Mater. 27, 3809 (2015).[35] S. Arai, S. Inoue, T. Hamai, R. Kumai, and T. Hasegawa,Semiconductive single molecular bilayers realized usinggeometrical frustration, Adv. Mater. 30, 1707256 (2018).[36] S. G. J. Mathijssen, M. Cölle, H. Gomes, E. C. P. Smits,B. De Boer, I. McCulloch, P. A. Bobbert, and D. M. DeLeeuw, Dynamics of threshold voltage shifts in organic andamorphous silicon field-effect transistors, Adv. Mater. 19,2785 (2007).[37] U. Zschieschang, F. Ante, T. Yamamoto, K. Takimiya,H. Kuwabara, M. Ikeda, T. Sekitani, T. Someya, K.Kern, and H. Klauk, Flexible low-voltage organic transis-tors and circuits based on a high-mobility organic semi-conductor with good air stability, Adv. Mater. 22, 982(2010).[38] L. Bürgi, H. Sirringhaus, and R. H. Friend, Noncon-tact potentiometry of polymer field-effect transistors, Appl.Phys. Lett. 80, 2913 (2002).[39] Y. Yamagishi, K. Noda, K. Kobayashi, and H. Yamada,Interlayer resistance and edge-specific charging in lay-ered molecular crystals revealed by Kelvin-probe forcemicroscopy, J. Phys. Chem. C 119, 3006 (2015).[40] Y. Hu, G. Li, W. Peng, and Z. Chen, Comparing the gatedependence of contact resistance and channel resistance inorganic field-effect transistors for understanding the mobil-ity overestimation issue, IEEE Electron Device Lett. 39,421 (2018).[41] S. Arai, K. Morita, J. Tsutsumi, S. Inoue, M. Tanaka,and T. Hasegawa, Layered-herringbone polymorphs andalkyl-chain ordering in molecular bilayer organic semi-conductors, Adv. Funct. Mater. 30, 1906406 (2020).[42] H. Minemawari, J. Tsutsumi, S. Inoue, T. Yamada, R.Kumai, and T. Hasegawa, Crystal structure of asymmet-ric organic semiconductor 7-decyl-2-phenyl[1]benzothieno[3,2-b][1]benzothiophene, Appl. Phys. Express 7, 091601(2014).[43] X. Chen, Z. Wang, J. Qi, Y. Hu, Y. Huang, S. Sun, Y. Sun,W. Gong, L. Luo, L. Zhang, et al., Balancing the film strainof organic semiconductors for ultrastable organic transis-tors with a five-year lifetime, Nat. Commun. 13, 1480(2022).[44] T. Oka, N. Shioya, T. Shimoaka, and T. Hasegawa, Struc-tural rearrangement of organic semiconductor moleculeswith an asymmetric shape in thin films, J. Phys. Chem. C127, 7560 (2023).[45] G. Giri, R. Li, D. M. Smilgies, E. Q. Li, Y. Diao, K.M. Lenn, M. Chiu, D. W. Lin, R. Allen, J. Reinspach,et al., One-dimensional self-confinement promotes poly-morph selection in large-area organic semiconductor thinfilms, Nat. Commun. 5, 3573 (2014).[46] M. Yoneya, H. Minemawari, T. Yamada, and T. Hasegawa,Interface-mediated self-assembly in inkjet printing ofsingle-crystal organic semiconductor films, J. Phys. Chem.C 121, 8796 (2017).[47] M. R. Niazi, R. Li, M. Abdelsamie, K. Zhao, D. H.Anjum, M. M. Payne, J. Anthony, D. M. Smilgies,and A. Amassian, Contact-induced nucleation in high-performance bottom-contact organic thin film transistorsmanufactured by large-area compatible solution processing,Adv. Funct. Mater. 26, 2371 (2016).[48] S. Lee, S. Jeon, and A. Nathan, Modeling sub-thresholdcurrent-voltage characteristics in thin film transistors,J. Disp. Technol. 9, 883 (2013).[49] B. Blülle, R. Häusermann, and B. Batlogg, Approach-ing the trap-free limit in organic single-crystal field-effecttransistors, Phys. Rev. Appl. 1, 034006 (2014).[50] M. Kano, T. Minari, and K. Tsukagoshi, Improvement ofsubthreshold current transport by contact interface modifi-cation in p-type organic field-effect transistors, Appl. Phys.Lett. 94, 143304 (2009).[51] G. Kitahara, M. Ikawa, S. Matsuoka, S. Arai, and T.Hasegawa, Approaching trap-minimized polymer thin-filmtransistors, Adv. Funct. Mater. 31, 2105933 (2021).024005-11https://doi.org/10.1021/acs.chemmater.5b00810https://doi.org/10.1002/adma.201707256https://doi.org/10.1002/adma.200602798https://doi.org/10.1002/adma.200902740https://doi.org/10.1063/1.1470702https://doi.org/10.1021/acs.jpcc.5b00611https://doi.org/10.1109/LED.2018.2798288https://doi.org/10.1002/adfm.201906406https://doi.org/10.7567/APEX.7.091601https://doi.org/10.1038/s41467-022-29221-8https://doi.org/10.1021/acs.jpcc.3c01003https://doi.org/10.1038/ncomms4573https://doi.org/10.1021/acs.jpcc.7b02143https://doi.org/10.1002/adfm.201502428https://doi.org/10.1109/JDT.2013.2256878https://doi.org/10.1103/PhysRevApplied.1.034006https://doi.org/10.1063/1.3115826https://doi.org/10.1002/adfm.202105933 I. INTRODUCTION II. EXPERIMENT A. Manufacturing BGBC-type TFTs with different numbers of layers B. Device and thin-film characterization III. RESULTS AND DISCUSSION A. Material and layer-number dependence of device mobility B. Time-dependent TFT characteristics and effects of encapsulation C. Potential drop at ternary interfaces in BGBC-type TFTs D. Transformation of OSC films on metal electrodes E. On the material and layer-number dependence of TFT characteristics IV. CONCLUSION ACKNOWLEDGMENTS . References<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /All /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile () /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Warning /CompatibilityLevel 1.4 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /LeaveColorUnchanged /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 5 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness false /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages false /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Average /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Average /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Average /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /PDFX1a:2003 ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError false /PDFXTrimBoxToMediaBoxOffset [ 33.84000 33.84000 33.84000 33.84000 ] /PDFXSetBleedBoxToMediaBox false /PDFXBleedBoxToTrimBoxOffset [ 9.00000 9.00000 9.00000 9.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /CreateJDFFile false /Description << /ARA /BGR /CHS /CHT /CZE /DAN /DEU /ESP /ETI /FRA /GRE /HEB /HRV /HUN /ITA /JPN /KOR /LTH /LVI /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken voor kwaliteitsafdrukken op desktopprinters en proofers. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR /POL /PTB /RUM /RUS /SKY /SLV /SUO /SVE /TUR /UKR /ENU (Use these settings to create Adobe PDF documents for quality printing on desktop printers and proofers. Created PDF documents can be opened with Acrobat and Adobe Reader 5.0 and later.) >> /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ << /AsReaderSpreads false /CropImagesToFrames false /ErrorControl /WarnAndContinue /FlattenerIgnoreSpreadOverrides false /IncludeGuidesGrids false /IncludeNonPrinting false /IncludeSlug false /Namespace [ (Adobe) (InDesign) (4.0) ] /OmitPlacedBitmaps false /OmitPlacedEPS false /OmitPlacedPDF false /SimulateOverprint /Legacy >> << /AddBleedMarks true /AddColorBars false /AddCropMarks true /AddPageInfo true /AddRegMarks false /BleedOffset [ 9 9 9 9 ] /ConvertColors /NoConversion /DestinationProfileName () /DestinationProfileSelector /NA /Downsample16BitImages true /FlattenerPreset << /PresetSelector /MediumResolution >> /FormElements false /GenerateStructure true /IncludeBookmarks true /IncludeHyperlinks true /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 6 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed false >> << /AllowImageBreaks true /AllowTableBreaks true /ExpandPage false /HonorBaseURL true /HonorRolloverEffect false /IgnoreHTMLPageBreaks false /IncludeHeaderFooter false /MarginOffset [ 0 0 0 0 ] /MetadataAuthor () /MetadataKeywords () /MetadataSubject () /MetadataTitle () /MetricPageSize [ 0 0 ] /MetricUnit /inch /MobileCompatible 0 /Namespace [ (Adobe) (GoLive) (8.0) ] /OpenZoomToHTMLFontSize false /PageOrientation /Portrait /RemoveBackground false /ShrinkContent true /TreatColorsAs /MainMonitorColors /UseEmbeddedProfiles false /UseHTMLTitleAsMetadata true >> ]>> setdistillerparams<< /HWResolution [2400 2400] /PageSize [612.000 792.000]>> setpagedevice