Controlled Synthesis of Poly(pentafluorostyrene-ran-methyl methacrylate) Copolymers by Nitroxide Mediated Polymerization and Their Use as Dielectric Layers in Organic Thin-film Transistors

A library of statistically random pentafluorostyrene (PFS) and methyl methacrylate (MMA) copolymers with narrow molecular weight distributions was produced, using nitroxide mediated polymerization (NMP) to study the effect of polymer composition on the performance of bottom-gate top-contact organic thin-film transistors, when utilized as the dielectric medium. Contact angle measurements confirmed the ability to tune the surface properties of copolymer thin films through variation of its PFS/MMA composition, while impedance spectroscopy determined the effect of this variation on dielectric properties. Bottom-gate, top-contact copper phthalocyanine (CuPc) based organic thin-film transistors were fabricated using the random copolymers as a dielectric layer. We found that increasing the PFS content led to increased field-effect mobility, until a point after which the CuPc no longer adhered to the polymer dielectric.


Introduction
Organic thin-film transistors (OTFTs) are expected to be an integral component of next-generation electronic devices due to their beneficial qualities such as reduced manufacturing-cost, lightweight and flexibility. As charge transport in OTFTs is limited to the sub-nanometer interfacial region between the semiconducting and dielectric layer [1][2][3], it is crucial to develop these two materials in parallel to better control their interface, and ultimately engineer higher-performing devices. During fabrication, the surface chemistry of the dielectric layer in bottom-gate configuration will dictate the growth of the semiconductor film, as well as influence the structural packing, domain structures and grain boundaries of the final film [4][5][6]. Hydrophilic surfaces, such as common inorganic oxides like SiO 2 or Al 2 O 3 , can unfavorably affect the growth mode of organic semiconducting films, as well as providing an abundance of water-absorbing sites that act as a charge carrier traps, resulting in poor device performance [5,7]. Therefore, hydrophobic polymers are frequently used to shield the hydroxyl functional groups on the inorganic oxide dielectric, leading to improved organic semiconductor morphology [1,[8][9][10].
Among hydrophobic polymers, those containing fluorinated monomers, such as 2,3,4,5, 6-pentafluorostyrene (PFS), have been used to improve the surface characteristics of dielectric layers of molar compositions of PFS relative to MMA is provided in Table 1. The synthesis of PFS/MMA-20/80 is provided as an example: MAMA-SG1 (0.153 g, 0.40 mmol), SG1 (0.0177 g, 0.06 mmol), xylenes (10 g), PFS (3.26 g, 16.8 mmol), MMA (6.74 g, 67.3 mmol), were added to the reactor with a stir bar. The reaction mixture was stirred and bubbled with nitrogen for 30 min. The nitrogen purge was removed from the mixture while still maintaining an N 2 atmosphere for the entirety of the reaction. The mantle was heated to 90 • C and t = 0 min was set to when the temperature in the reactor reached 90 • C. The samples were taken periodically by syringe throughout the polymerization. After stopping the reaction and allowing it to cool, the reaction mixture was precipitated into methanol, then re-dissolved into THF and re-precipitated into hexanes, filtered and dried in a vacuum oven at 60 • C overnight to give the final product. The number-average molecular weight M n = 13.8 kg mol −1 , dispersity (M w /M n = 1. 15 and PFS molar fraction in the copolymer F PFS = 0.31).
Polymers 2020, 12, x FOR PEER REVIEW 3 of 15 to the reactor with a stir bar. The reaction mixture was stirred and bubbled with nitrogen for 30 min. The nitrogen purge was removed from the mixture while still maintaining an N2 atmosphere for the entirety of the reaction. The mantle was heated to 90 °C and t = 0 min was set to when the temperature in the reactor reached 90 °C. The samples were taken periodically by syringe throughout the polymerization. After stopping the reaction and allowing it to cool, the reaction mixture was precipitated into methanol, then re-dissolved into THF and re-precipitated into hexanes, filtered and dried in a vacuum oven at 60 °C overnight to give the final product. The number-average molecular weight ̅ = 13.8 kg mol −1 , dispersity ( ̅ ̅ ⁄ = 1.15 and PFS molar fraction in the copolymer FPFS = 0.31).

Metal-Insulator-Metal (MIM) Capacitor Fabrication
Metal-insulator-metal (MIM) capacitors were prepared in an atmospheric environment. First, glass substrates (1" × 1") were washed by sonicating in a sequence of solvents; soapy water, water, acetone, then methanol, for 5 min each, then dried by blown nitrogen. Silver electrodes (60 nm) were deposited on the prepared glass, using a shadow mask and physical vapor deposition. Then, the solution deposition of the polymer layer was achieved by spin-coating a 50 mg/mL poly(PFS-ran-MMA) in methyl ethyl ketone (MEK, 99%) solution at 2000 rpm, followed by an annealing step (150 °C, under vacuum, 1 h). The procedure was then repeated: the same fluoropolymer solution was applied again on the film-coated substrates, and annealing repeated to form the final pin-hole free

Metal-Insulator-Metal (MIM) Capacitor Fabrication
Metal-insulator-metal (MIM) capacitors were prepared in an atmospheric environment. First, glass substrates (1" × 1") were washed by sonicating in a sequence of solvents; soapy water, water, acetone, then methanol, for 5 min each, then dried by blown nitrogen. Silver electrodes (60 nm) were deposited on the prepared glass, using a shadow mask and physical vapor deposition. Then, the solution deposition of the polymer layer was achieved by spin-coating a 50 mg/mL poly(PFS-ran-MMA) in methyl ethyl ketone (MEK, 99%) solution at 2000 rpm, followed by an annealing step (150 • C, under vacuum, 1 h). The procedure was then repeated: the same fluoropolymer solution was applied again on the film-coated substrates, and annealing repeated to form the final pin-hole free insulating film of approximately 500 nm. Next, the top Ag electrodes (60 nm) were evaporated, producing the completed

Bottom-Gate Top-Contact (BGTC) Thin-Film Transistor Fabrication
Bottom-gate top-contact (BGTC) thin-film transistors were fabricated under atmospheric conditions on 20 mm × 15 mm Ossila quartz coated glass substrates. The glass substrates were cleaned by bath sonication: soapy water, distilled water, acetone then methanol, each for 5 min. Similar to the MIM fabrication, a 60 nm Ag gate was patterned on the cleaned substrates by PVD followed by spin-coating of solutions (50 mg/mL in MEK) of poly(PFS-ran-MMA) at 2000 rpm for 1.5 min, followed by thermal vacuum annealing at 150 • C for 1 h. This process was then repeated on the same polymers to give a final polymer layer with a thickness of approximately 500 nm. A 50 nm layer of copper phthalocyanine (CuPc) was then deposited on the PFS/MMA surface by PVD, using a shadow mask. The top Ag source-drain electrodes (60 nm) were then deposited on top of the polymer. This process created 20 individual transistors with a channel length of 1000 µm and a channel width of 30 µm.

Characterization
Polymerization conversion was determined by gravimetry. Polymer compositions were determined by 1 H NMR and 19 F NMR spectroscopy, and the use of an α-α-α-trifluorotoluene marker and Bruker AVANCE II 400 MHz spectrometer ( Figure S1). Polymer samples were dissolved in chloroform-d. In the 1 H NMR spectra, the integrals of the MMA peak at 3.6 ppm and the marker's peak at 7.3-7.6 ppm were compared, and in the 19 F NMR spectra, the integrals of the PFS peak at −144 to −140 ppm and the marker's peak at −63 ppm were compared. From these two ratios, the composition of poly(PFS-ran-MMA) copolymers was determined. Molecular weight charcteristics of the final copolymers were determined by gel permeation chromatography (GPC)using an Agilent 1260 Infinity at 30 • C, flowing of 1 mL·min −1 of THF, as the eluting solvent via two MZ-Gel SD plus Linear 5 µm, 300 × 8.0 mm 2 columns. Triple detection was accomplished using a multi-angle light scattering (MALS) detector (DAWN HELEOS II), differential viscometer (ViscoStar II) and a differential index detector (Optilab T-rEX). The specific refractive index increment, ∂n/∂c of the PFS homopolymer was determined via off-line, batch-mode differential refractive index (dRI) experiments. A succession of solutions of sequentially increased concentrations of polymer in THF was injected directly into the dRI detector with a syringe pump, followed by pure THF without any dissolved analyte for a baseline. The ∂n/∂c of copolymers were determined from the ∂n/∂c values of the component homopolymers and the mass fraction of the comonomer in the copolymer. For a generic poly(PFS-ran-MMA) copolymer, its ∂n/∂c is calculated from the equation [24]: ∂n where w PFS and w MMA are the mass fractions of PFS and MMA in the poly(PFS-ran-MMA) copolymers, and (∂n/∂c) PFS and (∂n/∂c) MMA are the ∂n/∂c of the pure homopolymers determined under the same experimental conditions. The glass transition temperatures (T g s) were found using differential scanning calorimetry (DSC; TA Instruments Q2000). A heating/cooling rate of 10 • C/min under a nitrogen atmosphere across a temperature span of 10 • C to 170 • C was performed twice. The T g was then found by the midpoint from the second thermogram heating cycle. Contact angle measurements were performed on a VCA Optima goniometer (AST Products Inc). Droplets (0.5 µL) of DI water were deposited from a needle, imaged and contact angle were determined using a three-point curve fitting. The thickness of poly(PFS-co-MMA) films were measured with a Bruker Dektak XT profilometer. Prepared films were scratched using a diamond tip pen three times, and step edges measured then averaged. Impedance properties of the polymers in a metal-insulator-metal (MIM) structure were measured using electrochemical impedance spectroscopy (EIS; Metrohm PGSTAT204). An equivalent circuit model consisting of a resistor and a capacitor in parallel was used to extract the effective capacitance, which was calculated from the equation: C = 1/(2π f Im(Z)), where f is the frequency and Z is the measured impedance. Measurements were conducted over a frequency range of 10 −2 -10 5 Hz with an AC amplitude of 10 mV under atmospheric conditions. Current-voltage (I-V) characteristics of transistors were measured using a Keithley 2614B, holding the gate-source voltage (V GS ) constant, then sweeping the source-drain voltage (V SD ), while measuring the source-drain current (I SD ). Voltage step increases were set with a delay of 200 ms between measurements. Each device was characterized three times, and the obtained values were averaged. Transfer curves were measured in the saturation regime and were modelled using the following equation: where L and W are the length and width of the channel, respectively; C is the capacitance density determined from the thickness and dielectric constant of the gating medium, and µ is the field-effect mobility which was calculated from the slope of trendline through the linear region of the square root I SD plotted against V GS . All measurements were conducted in the atmosphere at room temperature. and ( ⁄ ) are the ⁄ of the pure homopolymers determined under the same experimental conditions. The glass transition temperatures (Tgs) were found using differential scanning calorimetry (DSC; TA Instruments Q2000). A heating/cooling rate of 10 °C/min under a nitrogen atmosphere across a temperature span of 10 °C to 170 °C was performed twice. The Tg was then found by the midpoint from the second thermogram heating cycle. Contact angle measurements were performed on a VCA Optima goniometer (AST Products Inc). Droplets (0.5 µ L) of DI water were deposited from a needle, imaged and contact angle were determined using a three-point curve fitting. The thickness of poly(PFS-co-MMA) films were measured with a Bruker Dektak XT profilometer. Prepared films were scratched using a diamond tip pen three times, and step edges measured then averaged. Impedance properties of the polymers in a metal-insulator-metal (MIM) structure were measured using electrochemical impedance spectroscopy (EIS; Metrohm PGSTAT204). An equivalent circuit model consisting of a resistor and a capacitor in parallel was used to extract the effective capacitance, which was calculated from the equation: , where is the frequency and is the measured impedance. Measurements were conducted over a frequency range of 10 −2 -10 5 Hz with an AC amplitude of 10

Copolymer Synthesis: Kinetics & Control
Pentafluorostyrene (PFS) has been successfully homopolymerized [21] using nitroxide mediated polymerization (NMP), as well as other controlled free-radical approaches, such as ATRP and RAFT [18][19][20]25]. Methacrylic monomers are notoriously challenging to polymerize in pseudo-living fashion, due to their large equilibrium constant for propagation and unwanted side reactions [26,27]. To increase the number of living propagating chains methacrylate monomers are often copolymerized with a slowly propagating comonomer, which is more compatible with NMP, such as styrenics [28][29][30][31], acrylonitrile [32][33][34][35] or cyclic ketene acetal 2-methylene-4-phenyl-1,3-dioxolane [36,37]. PFS has been demonstrated to behave as a controlling comonomer in the NMP of various methacrylic monomers, such as methacrylic acid [22], 5-methacryloyloxy-2,6-norboranecarbolactone (NLAM) [38], oligo(ethyleneglycol) methacrylate (OEGMA) [39] and, very recently, methyl methacrylate (MMA) [40]. Delaittre and coworkers demonstrated that ≈5 mol% of PFS could control the copolymerization of MMA, leading to a high MMA-containing pseudo-living macroinitiator, suitable for the chain extension of styrene [40]. To further characterize this copolymerization at low through high PFS loadings we performed a range of copolymerizations with a target number average molecular weight (M n ) of 25 kg mol −1 in a 50 wt% solution of xylenes at 90 • C, with feed compositions ranging from 5-100 mol%, which is detailed in Table 1. The kinetics of NMP is typically characterized by k P K values, where k P and K represent the propagation rate and activation-deactivation equilibrium constants, respectively. The activation-deactivation constant (K) is a function of the propagating radicals, [P•], free nitroxide, [N•] and dormant alkoxyamine-terminated species, [P − N] concentrations. K is calculated from Equation (1) below: Numerous assumptions are made to simplify Equation (1) to produce an estimate of the important combination parameter k P K from the kinetic data. Within the initial stages of polymerization, the concentration of free nitroxide radicals is high and relatively constant. Therefore, the initial concentration of additional SG1,  (Table 1), indicating the termination by B-hydrogen chain transfer to SG1 rate is low, and the primary mechanism of irreversible termination is from homotermination [27,41]. Initially, in the polymerization, when polymer chains are of a few repeat units, homotermination is low. We can therefore assume the concentration of reversibly deactivated alkoxyamine-terminated species is equal to the initial concentration of alkoxyamine initiator ([P − N] = [MAMA-SG1] 0 ). The assumptions made are valid if polymerization exhibits pseudo-living behavior; molecular weight increases linearly with the conversion. Finally, k p K is calculated and the initial molar concentrations of SG1 and MAMA-SG1 are substituted by r to result in Equation (2): The slope of the ln(1 − X) −1 vs time plot is equal to the apparent rate constant, k p [P•]. In this study, gravimetric analysis was utilized to determine conversion, X. A representative ln(1 − X) −1 vs time plot used to calculate k p [P•] from the slope can be found in Figure 3, while all the resulting k p [P•] values can be found in Table 2. PFS/MMA copolymerizations also demonstrate a linear increase in M n as a function of X, which is consistent with the previous assumptions suggesting Equation (2) is valid ( Figure 3). As the PFS content in the feed is increased, the k p [P•] decreases (Table 2), which is the typical kinetic behavior of MMA/styrenic systems by NMP.

Reactivity Ratio Determination
The final composition of the copolymer will be used to engineer and tune material properties; therefore, it is necessary to have the ability to synthesize a copolymer of a specifically desired composition. Knowledge of the comonomer reactivity ratios is therefore required. To determine the comonomer reactivity ratios, a series of copolymerizations were performed using the same protocol used in kinetic experiments; however, these copolymerizations were purposely ended at low conversion (X < 0.10) to avoid compositional drift effects. Seven copolymers with initial PFS/MMA compositions; f PFS,0 = 0.05, 0.20, 0.35, 0.50, 0.60, 0.75 and 0.95 were copolymerized. Copolymer composition was determined by a combination of 1 H NMR and 19 F NMR spectroscopy, with the use of an α-α-α-trifluorotoluene marker, and are shown in Figure 4, where the initial molar feeds are displayed against the final copolymer compositions (X < 0.10). Reactivity ratios of the comonomers were found with the Mayo-Lewis equation (Equation (3)) [42,43]: where the final molar copolymer composition F PFS is a function of initial molar feed compositions f PFS,0 and f MMA,0 and the reactivity ratios r PFS and r MMA . ƒ PFS,0 and F PFS,10 tabulated in Table 2, were substituted into Equation (3) for a reactivity ratio determination. A non-linear least squares (NILS) fitting of the Mayo-Lewis equation to the experimentally measured copolymer compositions was performed to converge on the reactivity ratios, which are displayed in Figure 4 [44][45][46]. The fitting shows r PFS > r MMA , indicating a preferred addition of the controlling PFS comonomer over MMA suggesting the final copolymers synthesized may have a gradient composition where the initiation end is heavily comprised of the fluorinated sytyrenic, and the propagating end is more MMA rich in comparison. Earlier studies have determined a styrenics capability to maintain a pseudo-living copolymerization of methacrylates is dependent on the reactivity ratios: if the propagation of the styreneic is preferred, a pseudo-living copolymerization can be performed with MMA feeds as significant as f MMA,0 ≈ 0.95-0.99 [47]. However, if the reactivities suggest a preferred propagation of the methacrylate over styreneic, then a pseudo-living copolymerization will only be possible with feeds extremely rich in the controlling styrenic monomer, for instance, methacrylic acid (MAA) and PFS requires f PFS,0 > 0.60 to result in a pseudo-living copolymerization [22]. The results we found indicate that the copolymerization of MMA and PFS will exhibit pseudo-living copolymerization behavior over a wide range of feed compositions.
copolymerization of methacrylates is dependent on the reactivity ratios: if the propagation of the styreneic is preferred, a pseudo-living copolymerization can be performed with MMA feeds as significant as fMMA,0 ≈ 0.95-0.99 [47]. However, if the reactivities suggest a preferred propagation of the methacrylate over styreneic, then a pseudo-living copolymerization will only be possible with feeds extremely rich in the controlling styrenic monomer, for instance, methacrylic acid (MAA) and PFS requires fPFS,0 > 0.60 to result in a pseudo-living copolymerization [22]. The results we found indicate that the copolymerization of MMA and PFS will exhibit pseudo-living copolymerization behavior over a wide range of feed compositions.  Table 2 lists the experiments used for the plots.

Characterization of the final Random Copolymers
After the polymer reaction engineering was performed, a subset of materials of broad (FPFS = 0.08 -1.00) composition was selected and extensively characterized to determine relationships between the copolymer composition and material property ( Table 3). The bulk material properties were characterized by GPC and DSC. The molecular weights of the materials ranged from 19 kDa to 74 kDa, all with a relatively narrow dispersity, as expected by controlled polymerization. The glass transition temperatures, Tgs ranged from (90 to 110 °C, Figure S2).  Table 2 lists the experiments used for the plots.

Characterization of the Final Random Copolymers
After the polymer reaction engineering was performed, a subset of materials of broad (F PFS = 0.08-1.00) composition was selected and extensively characterized to determine relationships between the copolymer composition and material property ( Table 3). The bulk material properties were characterized by GPC and DSC. The molecular weights of the materials ranged from 19 kDa to 74 kDa, all with a relatively narrow dispersity, as expected by controlled polymerization. The glass transition temperatures, T g s ranged from (90 to 110 • C, Figure S2).  Contact angle measurements were performed on thin-films of the PFS/MMA copolymers, as well as blends of the respective poly(MMA) and poly(PFS) homopolymers ( Figure 5). The blended films possessed a contact angle of around 100 • , the same as the PFS homopolymer, which is due to the migration of PFS homopolymer to the surface during annealing steps, this surface segregation of fluorinated components has also been noted in polystyrene and vinyl acetate blended polymer films [15,16]. The contact angles of the copolymer films ranged linearly from 83.3 ± 2.5 to 99.7 ± 1.9 at compositions of F PFS = 0.08 to 0.89, which falls in between the respective homopolymer chracteristics ( Figure 5). These results highlight the necessity of copolymers over simple blending techniques to tune surface properties.
Contact angle measurements were performed on thin-films of the PFS/MMA copolymers, as well as blends of the respective poly(MMA) and poly(PFS) homopolymers ( Figure 5). The blended films possessed a contact angle of around 100°, the same as the PFS homopolymer, which is due to the migration of PFS homopolymer to the surface during annealing steps, this surface segregation of fluorinated components has also been noted in polystyrene and vinyl acetate blended polymer films [15,16]. The contact angles of the copolymer films ranged linearly from 83.3 ± 2.5 to 99.7 ± 1.9 at compositions of FPFS = 0.08 to 0.89, which falls in between the respective homopolymer chracteristics ( Figure 5). These results highlight the necessity of copolymers over simple blending techniques to tune surface properties. Metal-insulator-metal (MIM) capacitors were fabricated to analyze the dielectric behavior of the copolymers by electrical impedance spectroscopy (EIS). The MIM sandwich structure was fabricated by spin coating the PFS/MMA copolymers between silver electrodes, which were deposited by physical vapor deposition. Using EIS and the film thickness from profilometry, the dielectric constant was determined using the equation: Metal-insulator-metal (MIM) capacitors were fabricated to analyze the dielectric behavior of the copolymers by electrical impedance spectroscopy (EIS). The MIM sandwich structure was fabricated by spin coating the PFS/MMA copolymers between silver electrodes, which were deposited by physical vapor deposition. Using EIS and the film thickness from profilometry, the dielectric constant was determined using the equation: where C is the capacitance [F], d is the insulator thickness [m], A is the area of the electrode area [m 2 ] and ε 0 is the permittivity constant [F/m]. The dielectric constant was shown to slightly increase at lower frequencies ( Figure S3) and exhibited no voltage dependence. As shown in Figure 6, the dielectric constant of the copolymers clearly shows a linear decrease from 3.9 to 2.3, with increasing PFS content at 100 kHz. These results demonstrate that the dielectric properties of the poly(MMA) polymer can be tuned with the copolymerization of a known amount of PFS monomer.  Figure S3) and exhibited no voltage dependence. As shown in Figure 6, the dielectric constant of the copolymers clearly shows a linear decrease from 3.9 to 2.3, with increasing PFS content at 100 kHz. These results demonstrate that the dielectric properties of the poly(MMA) polymer can be tuned with the copolymerization of a known amount of PFS monomer.

Poly(PFS-ran-MMA) Copolymers as Dielectric Materials in Organic Thin-film Transistors
Bottom-gate top-contact (BGTC, Figure 7) copper phthalocyanine (CuPc) based OTFTs were fabricated using poly(PFS-ran-MMA) copolymers as a dielectric layer to identify the impact of PFS content on the hole transport mobility (µ), the on-off current (ION/OFF) and the threshold voltage (VT), and can be found in Table 4. The µ of the OTFT devices was calculated using the capacitance densities for the corresponding copolymers obtained from the dielectric constants in the previous section. The resulting OTFTs were characterized by having μ = 1 to 3 × 10 −3 cm 2 /Vs, ION/OFF 1 to 10 × 10 3 and a VT = −4 to −17 V, which is typical for CuPc based OTFT devices [48][49][50]. Characteristic output and transfer curves are shown in Figure 7. Fluorinated surfaces are electron-withdrawing, which have been shown

Poly(PFS-ran-MMA) Copolymers as Dielectric Materials in Organic Thin-Film Transistors
Bottom-gate top-contact (BGTC, Figure 7) copper phthalocyanine (CuPc) based OTFTs were fabricated using poly(PFS-ran-MMA) copolymers as a dielectric layer to identify the impact of PFS content on the hole transport mobility (µ), the on-off current (I ON/OFF ) and the threshold voltage (V T ), and can be found in Table 4. The µ of the OTFT devices was calculated using the capacitance densities for the corresponding copolymers obtained from the dielectric constants in the previous section. The resulting OTFTs were characterized by having µ = 1 to 3 × 10 −3 cm 2 /Vs, I ON/OFF 1 to 10 × 10 3 and a V T = −4 to −17 V, which is typical for CuPc based OTFT devices [48][49][50]. Characteristic output and transfer curves are shown in Figure 7. Fluorinated surfaces are electron-withdrawing, which have been shown to improve the performance of p-type semiconductors [51]. In this case and highlighted in Figure 8, the increase in the content of PFS reduces the V T , while having little effect on the µ, which is consistent with what we would expect for the addition of fluorinated dielectrics [52]. This suggests that the PFS molecules present on the dielectric surface reduce the concentration of charge trap sites at the semiconductor/dielectric interface [11,12]. For the PFS/MMA copolymers with a PFS composition greater than F PFS = 0.57 , there were few functioning devices, and the results were unreliable with large deviations. We identified that during fabrication, when the polymer was too rich in PFS content, the CuPc semiconductor would no longer form uniform films ( Figure S4). This was obvious from the lack of blue color (CuPc, Figures S4 and S5) in the films and the drop in functioning devices ( Table 4). The devices that did function when F PFS > 0.57 were when some CuPc deposited in the channel ( Figure  S5b). The microstructure of the organic semiconducting layer, especially at the dielectric interface where the conductive channel is formed, has a significant effect on the charge transport properties [53][54][55]. Therefore, we performed powder X-ray diffraction on the CuPc layer, which was deposited on the different copolymers, and the spectra can be found in the ESI ( Figure S6). In all cases, we observed a weak characteristic peak corresponding to the CuPc at 2θ = 7 • . No significant differences in diffraction intensity (or peak location) between the CuPc films that grew on different surfaces was observed. Further investigation would be necessary to probe the solid-state arrangement of these films and identify the effect of the copolymer composition on the CuPc film growth. Regardless, these results demonstrate that PFS comonomers can be introduced into poly(MMA) dielectrics to tune the surface properties, dielectric properties and the resulting OTFT properties.

Conclusions
Poly(2,3,4,5,6-pentafluorostyrene) were synthesized in large yields by Nitroxide Mediated Polymerization at 90 • C. The copolymerization exhibited pseudo-living behavior shown by the apparent polymerization rate, which follows first order kinetics with reference to monomer conversion. Additionally, the molecular weight increases in a linear relation to monomer conversion. Finally, the molecular weights experimentally measured align with the theoretical values, and are of relatively low dispersity (M w /M n ≤ 1.3). Thin-film contact angle and dielectric constant were shown to change linearly with copolymer composition. We investigated the effects of copolymer composition with respect to fluorinated content on bottom-gate top-contact CuPc OTFT performance characteristics, and found a reduction in threshold voltage with increasing fluorine content; however, materials which were too rich in fluorine content resulted in poor semiconductor adhesion and non-functional devices.

Conflicts of Interest:
The authors declare no conflict of interest.