Proton-Conducting Polymer-Coated Carbon Nanofiber Mats for Pt-Anodes of High-Temperature Polymer-Electrolyte Membrane Fuel Cell

High-temperature polymer-electrolyte membrane fuel cells (HT-PEM FC) are a very important type of fuel cell since they operate at 150–200 °C, allowing the use of hydrogen contaminated with CO. However, the need to improve stability and other properties of gas diffusion electrodes still hinders their distribution. Anodes based on a mat (self-supporting entire non-woven nanofiber material) of carbon nanofibers (CNF) were prepared by the electrospinning method from a polyacrylonitrile solution followed by thermal stabilization and pyrolysis of the mat. To improve their proton conductivity, Zr salt was introduced into the electrospinning solution. As a result, after subsequent deposition of Pt-nanoparticles, Zr-containing composite anodes were obtained. To improve the proton conductivity of the nanofiber surface of the composite anode and reach HT-PEMFC better performance, dilute solutions of Nafion®, a polymer of intrinsic microporosity (PIM-1) and N-ethyl phosphonated polybenzimidazole (PBI-OPhT-P) were used to coat the CNF surface for the first time. These anodes were studied by electron microscopy and tested in membrane-electrode assembly for H2/air HT-PEMFC. The use of CNF anodes coated with PBI-OPhT-P has been shown to improve the HT-PEMFC performance.


Introduction
The search for new and the development of existing electrocatalytic materials are very important for progress in the field of alternative energy, particularly fuel cells (FC) [1,2]. In the context of global decarbonization, renewable energy sources, compared with intermittent and unstable power generation, may eliminate the temporal and spatial gap between energy consumption by end-users and its availability; fuel cells can generate power flexibly at any time, as long as hydrogen supply is sufficient [3,4]. The high-temperature polymerelectrolyte membrane (HT-PEM) fuel cells [5,6] ( Figure S1) possess many advantages, the most significant of which is the ability to operate with CO-contaminated hydrogen [7,8]. Polybenzimidazole (PBI)-type membranes doped with phosphoric acid are used for this type of FC and are capable of conducting protons without moisture [9][10][11]. This allows the polymer-electrolyte membrane to operate at elevated temperatures of 150-200 • C [12,13]. However, to increase its efficiency, further improvement of the main components of the

Electrospinning
Composite PAN-based nanofibers were obtained by the needle-free electrospinning method from a free surface, according to the Nanospider TM technology. The electrospinning process was performed on a NS Lab Nanospider TM setup from Elmarco (Liberec, Czech Republic) at a relative humidity of 8%, at voltage of 69 kV, with a distance between electrodes of 190 mm from the electrospinning polymer solution, which contained 3.25 g of PAN (Mw 150·10 3 Da), 0.1 g of UM-76 carbon black (~3 wt.% relative to PAN) and 0.03 g of zirconium (IV) chloride well-dispersed in 50 mL of N,N-dimethylformamide (DMF) in an ultrasonic bath for 3 h. As a result, PAN/UM/Zr composite nanofibers were obtained in the form of a mat.

Stabilization, Zinc Deposition and Pyrolysis
PAN/UM/Zr nanofiber mat was stabilized (oxidized) at 350 °C in air for 2 h in a Binder MDL 115 heating chamber (Tuttlingen, Germany) to make the material suitable for further pyrolysis. The resulting material (PAN/UM/Zr/Ni-350) was immersed in 0.5 wt.% solution of zinc nitrate in water/ethanol (1:3 v/v) for ~24 h and dried at 100 °C for 2 h. Then, the sample was pyrolyzed at 1000 °C for 2 h under vacuum at a heating rate of 3 °C min −1 using a Carbolite (CTF 12/80/700) vacuum oven (Hope Valley, UK) equipped with a Eurotherm 3216 series controller. As a result, the shape of the CNF mat (PAN/UM/Zr-350/Zn-1000) was saved.

Platinum Deposition
Platinum was deposited onto the CNF and polymer-coated CNF mats, with a surface area of 6.76 cm 2 separately for each mat, in 10 mL of aqueous solution containing the calculated amount of H 2 [PtCl 6 ]·6H 2 O as a source of platinum and 0.5 g of formic acid as a reducing agent for 3 days to obtain Pt/CNF, Pt/PIM-1/CNF, Pt/PBI-O-PhT-P/CNF and Pt/Nafion/CNF electrocatalysts with a Pt concentration of 1.2 mg Pt cm −2 . The resulting electrocatalysts were thoroughly washed with distilled water and dried at 100 • C for 2 h under vacuum.

Inverted Platinum and Polymer Deposition
For comparison, the proton-conducting polymers were applied to already platinated samples of Pt/CNF in the same way as described in Section 2.1.3. As a result, the polymercoated electrocatalysts PIM-1/Pt/CNF, PBI-O-PhT-P/Pt/CNF and Nafion/Pt/CNF were also obtained.

Elemental Analysis and Electrical Conductivity
The elemental analysis data were determined using an Elementar vario MICRO cube C,H,N-analyzer (Langenselbold, Germany) equipped with a thermal desorption column.
The in-plane electrical conductivities of the CNF mat (PAN/UM/Zr-350/Zn-1000) were found with a RLC E7-8 setup (Minsk, Belarus) equipped with a four-point probe. The sample thickness, which is necessary for the electrical conductivity measurements, was determined with an ElektroPhysik eXacto thickness gage (Cologne, Germany). The elemental analysis and electrical conductivity of the CNF are shown in Table 1.

Morphological Characterization
The structure of the composite еlectrospun polyacrylonitrile-based CNF was investigated by the methods of scanning electron microscopy (SEM) using a FEI Scios microscope (FEI, Hillsboro, OR, USA), transmission electron microscopy (TEM), high-resolution transmission electron microscopy (HR TEM), scanning transmission electron microscopy with a high-angle annular dark-field detector (HAADF STEM) and energy-dispersive X-ray spectroscopy (EDX) elemental mapping using a Thermo Fisher Scientific Osiris (Waltham, MA, USA) equipped with a high-angle annular dark field (HAADF) detector and Super-X EDX detection system based on Silicon Drift Detector (SDD) technology. Electron microscope images were analyzed using Digital Micrograph (GMS 3, Gatan, Pleasanton, CA, USA), TIA (TIA 16, Siemens AG, Munich, Germany), Esprit (Esprit 2, Bruker, Billerica, MA, USA) and JEMS software (P. Stadelmann JEMS-EMS Java version 2004 EPFL, Lausanne, Switzerland). For electron microscopy studies, the samples of CNF were well-dispersed in acetone to separate the fibers using an ultrasonic bath for 20-30 min. Then, the obtained suspensions were introduced onto copper lacey carbon grids.

Electrochemical Characterization
For the anode tests, the membrane-electrode assemblies (MEA) were prepared with a working area of 5 cm 2 . The MEA were placed in a standard Arbin Instruments testing cell (College Station, TX, USA) with two graphite flow field plates. The HT-PEM fuel cell was operated with a typical Celtec ® -P 1000 MEA cathode [70] and the CNF-based anodes developed in this study. The PBI-OPhT membrane, cross-linked with zirconium acetylacetonate and doped with phosphoric acid (350-400%, up to 25 molecules of ophosphoric acid per polymer unit), which was developed earlier by our group [71][72][73], was used for the MEA. Fuel cell operation was carried out at 160 and 180 • C. The anode was supplied with hydrogen obtained through electrolysis from a GVCh-6 hydrogen generator (Khimelektronika, Moscow, Russia) at a rate of 200 mL min −1 and the cathode was supplied with atmospheric air at a rate of 1000 mL min −1 without additional humidification. The polarization curves were obtained on a P-150X Potentiostat-galvanostat (Electrochemical Instruments, Chernogolovka, Russia). For voltammetry measurements, the fuel cell voltage was scanned at a rate of 10 mV s −1 in the cell voltage range 0.95-0.02 V.
The measurements of the HT-PEMFC membrane through plane resistance (mΩ cm 2 ) were performed by the method of electrochemical impedance spectroscopy (EIS). The EIS experiments were performed on a SmartStat PS-250 Potentiostat-galvanostat (Electrochemical Instruments, Chernogolovka, Russia) at 0.4 A cm −2 , applying a sinusoidal current with an amplitude of 50 mA in the frequency range 50 kHz-0.1 Hz.

Hydrogen Crossover Measurements
Hydrogen crossover through the membrane at the operating temperatures of the HT-PEM fuel cell was measured by the method of linear sweep voltammetry by supplying hydrogen to the anode and argon (99.998% purity) to the cathode; the gases were supplied at an ambient pressure with flow rates of 50 mL min −1 . Finally, the open-circuit voltage reached a steady-state value of about 120 mV, and the voltage was swept slowly at a rate of 1 mV s −1 to 500 mV; the current of hydrogen oxidation was recorded. Hydrogen penetrated the membrane, then appeared in the cathode catalyst layer and was completely oxidized. The current value measured under these conditions is assumed to be equal to the flow of hydrogen diffusing through the membrane. The current density corresponding to the crossover of hydrogen through the membrane was quantified at 350 mV. Higher electrode potentials were avoided to prevent the oxidation of platinum.

N 2 Adsorption
Nitrogen adsorption isotherms were obtained on a 3P Micro 200 Surface Area and Pore Size Analyzer (3P Instruments, Odelzhausen, Germany) at 77 K in the pressure range 10 −3 -1 bar. The Brunauer-Emmett-Teller (BET) equation was applied to the N 2 adsorption isotherm data according to the Rouquerol criteria [74]. N 2 cross-sectional area and adsorbed N 2 density were taken as 0.162 nm 2 and 0.808 g mL −1 , respectively.

Electron Microscopy
Three proton-conducting polymers, PIM-1, PBI-O-PhT-P and Nafion ® (Figure 1), may provide additional proton conductivity when applied to the CNF. We expect this improvement because PIM-1 protonation ability is well-known [55]. Nafion ® [56][57][58][59][60] is another well-known proton-conducting polymer [2]. The N-ethyl phosphonated PBItype polymer (PBI-O-PhT-P), deposited on the CNF of the anode, introduces additional phosphonic functional groups, which can improve proton conductivity. Earlier, we showed that a proton-conducting membrane can be prepared from PBI-O-PhT-P (also abbreviated as PEPBI-O-PH or PhEPBI-O-PhT) [67]; its proton conductivity was described in detail [76]. At the same time, it could remain permeable to hydrogen under the operation conditions of HT-PEMFC due to the peculiarities of its structure, as explained before. Platinum nanoparticles were deposited onto polymer-coated carbon nanofibers (polymer/CNF) to obtain electrodes of general formula Pt/polymer/CNF. The Pt/polymer/CNF structures were studied by the method of electron microscopy.
SEM images of the CNF structure with the deposited PIM-1 are shown in Figure 2.
PhEPBI-O-PhT) [67]; its proton conductivity was described in detail [76]. At the same time, it could remain permeable to hydrogen under the operation conditions of HT-PEMFC due to the peculiarities of its structure, as explained before. Platinum nanoparticles were deposited onto polymer-coated carbon nanofibers (polymer/CNF) to obtain electrodes of general formula Pt/polymer/CNF. The Pt/polymer/CNF structures were studied by the method of electron microscopy. SEM images of the CNF structure with the deposited PIM-1 are shown in Figure 2. Platinum was deposited on the PIM/CNF in an aqueous solution of H2[PtCl6], using HCOOH as a reducing agent. The general SEM image ( Figure 2a) indicates that the fibers located on the surface of the mat are covered with a uniform layer of platinum. However, the CNF lying in the depth are covered with platinum "islands" or have no coating at all (Figure 2b,c). At the same time, porous accumulations of platinum several microns in size are observed on the CNF surface ( Figure 2d). Most likely, they are located in a polymer shell. Figure 3 shows a bright field TEM image of a single Pt island that is "immersed" in the polymer.
The presence of a polymer shell on the CNF is also indicated by a selected area electron diffraction (SAED) pattern with a strong background and individual bright signals of platinum (Figure 3a-c). The island of platinum is a polycrystalline agglomerate with individual nanocrystals of 5-10 nm. In addition, individual platinum particles of rounded shape with a size of 2-10 nm are clearly observed on the fibers.
When Pt is deposited on Nafion/CNF, the formation of a Pt "skin" with a width of up to 120 nm is observed in the upper layers of the mat. Due to the large thickness, the coating peels off and breaks on many fibers, which is clearly seen in the TEM and SEM images (  Platinum was deposited on the PIM/CNF in an aqueous solution of H 2 [PtCl 6 ], using HCOOH as a reducing agent. The general SEM image (Figure 2a) indicates that the fibers located on the surface of the mat are covered with a uniform layer of platinum. However, the CNF lying in the depth are covered with platinum "islands" or have no coating at all (Figure 2b,c). At the same time, porous accumulations of platinum several microns in size are observed on the CNF surface ( Figure 2d). Most likely, they are located in a polymer shell. Figure 3 shows a bright field TEM image of a single Pt island that is "immersed" in the polymer.  The presence of a polymer shell on the CNF is also indicated by a selected area electron diffraction (SAED) pattern with a strong background and individual bright signals of platinum (Figure 3a-c). The island of platinum is a polycrystalline agglomerate with individual nanocrystals of 5-10 nm. In addition, individual platinum particles of rounded shape with a size of 2-10 nm are clearly observed on the fibers.
When Pt is deposited on Nafion/CNF, the formation of a Pt "skin" with a width of up to 120 nm is observed in the upper layers of the mat. Due to the large thickness, the coating peels off and breaks on many fibers, which is clearly seen in the TEM and SEM images (Figures 4 and 5).  In contrast to the Pt/PIM/CNF and Pt/Nafion/CNF samples, the polymer coating in t Pt/PBI-O-PhT-P/CNF sample is less uniformly distributed over the surface. It can be se that the CNF are coated with the polymer, while the platinum possesses the shape of island In some places, the polymer layer is absent (shown by arrows in Figure 6b). TEM studi (Figure 7) show that the morphology of platinum particles differs from those indicat above: Pt nanocrystals possess an acicular shape. The length of some Pt needles can be u to 20 nm long. At the same time, particles of rounded shape are present. The needle-shape particles placed directly on the nanofiber with the corresponding SAED image are show in Figure 7a,b. The HAADF STEM image, element distribution maps and correspondin Comparison with uncoated Pt/CNF (prepared without a polymer coating) showed that platinum is uniformly distributed on the fiber surface throughout the entire depth of the mat, with a predominance of the acicular morphology of platinum (Figures 8 and 9).  In contrast to the Pt/PIM/CNF and Pt/Nafion/CNF samples, the polymer coating in the Pt/PBI-O-PhT-P/CNF sample is less uniformly distributed over the surface. It can be seen that the CNF are coated with the polymer, while the platinum possesses the shape of islands. In some places, the polymer layer is absent (shown by arrows in Figure 6b). TEM studies (Figure 7) show that the morphology of platinum particles differs from those indicated above: Pt nanocrystals possess an acicular shape. The length of some Pt needles can be up to 20 nm long. At the same time, particles of rounded shape are present. The needle-shaped particles placed directly on the nanofiber with the corresponding SAED image are shown in Figure 7a     Comparison with uncoated Pt/CNF (prepared without a polymer coating) showed that platinum is uniformly distributed on the fiber surface throughout the entire depth of the mat, with a predominance of the acicular morphology of platinum (Figures 8 and 9).

Adsorption Studies
The initial non-platinated CNF (PAN/UM/Zr-350/Zn-1000) sample was studied by the N 2 and CO 2 adsorption methods to determine its specific surface area and specific volume. The data from the Brunauer-Emmett-Teller (BET) method applied to the N 2 adsorption isotherm (77 K) show quite low values (15 m 2 /g). Possibly, it is due to the inaccessibility of micropores to nitrogen due to kinetic difficulties at such a low temperature (77 K). Therefore, to overcome the kinetic difficulties, an alternative method of the CO 2 adsorption (273 K) was applied [77]. The CO 2 adsorption isotherm data were used to find the micropore specific volume (V), adsorption energy (E) and pore width (D) values by the DR method and the micropore specific volume (V) and specific surface area (S) values by the NL-DFT and GCMC methods ( Table 2).
The differences in values obtained by different methods are typical, since different theoretical approaches and considerations are applied to different methods. The adsorption isotherm and NL-DFT pore-size distribution data are shown in Figure 10. K) was applied [77]. The CO2 adsorption isotherm data were used to find the micropore specific volume (V), adsorption energy (E) and pore width (D) values by the DR method and the micropore specific volume (V) and specific surface area (S) values by the NL-DFT and GCMC methods ( Table 2). The differences in values obtained by different methods are typical, since different theoretical approaches and considerations are applied to different methods. The adsorption isotherm and NL-DFT pore-size distribution data are shown in Figure 10. The presented data on microporosity (Table 2, Figure 10) are typical for microporous materials and suggest the possibility to use the material as a support for Pt nanoparticle catalyst in HT-PEM FC.

HT-PEM Fuel Cell Performance
The performance of the HT-PEM FC MEA with different Pt/polymer/CNF anodes when platinum is deposited on polymer-coated CNF is shown in Figure 11. The presented data on microporosity (Table 2, Figure 10) are typical for microporous materials and suggest the possibility to use the material as a support for Pt nanoparticle catalyst in HT-PEM FC.

HT-PEM Fuel Cell Performance
The performance of the HT-PEM FC MEA with different Pt/polymer/CNF anodes when platinum is deposited on polymer-coated CNF is shown in Figure 11. The polarization curves and power density data for the Pt/polymer/CNF are compared with the Pt/CNF sample without polymer coating in the HT-PEM FC operation at 160 and 180 °C. As seen in Figure 10, Pt/PBI-O-PhT-P/CNF shows higher performance compared with Pt/CNF. At the same time, the application of CNF coated with Nafion ® and PIM-1 as supports for platinum nanoparticles leads to much lower HT-PEM FC performance. It can be related to a less uniform polymer coating on the CNF in the case of PBI-O-PhT-P compared with Nafion ® or PIM-1, as shown above by electron microscopy. As a result, more platinum nanoparticles possess an acicular shape and grow directly on the CNF surface in- The polarization curves and power density data for the Pt/polymer/CNF are compared with the Pt/CNF sample without polymer coating in the HT-PEM FC operation at 160 and 180 • C. As seen in Figure 10, Pt/PBI-O-PhT-P/CNF shows higher performance compared with Pt/CNF. At the same time, the application of CNF coated with Nafion ® and PIM-1 as supports for platinum nanoparticles leads to much lower HT-PEM FC performance. It can be related to a less uniform polymer coating on the CNF in the case of PBI-O-PhT-P compared with Nafion ® or PIM-1, as shown above by electron microscopy. As a result, more platinum nanoparticles possess an acicular shape and grow directly on the CNF surface instead of being located on the polymer layer, which improves electron transfer during HT-PEM FC operation.
The inverted anode structure, when the polymer was applied on the already prepared CNF with deposited platinum (Pt/CNF), was also studied. The polarization curves and power density data for these samples of general formula polymer/Pt/CNF for HT-PEM FC operation at 160 and 180 • C are shown in Figure 12.
The polarization curves and power density data for the Pt/polymer/CNF are compared with the Pt/CNF sample without polymer coating in the HT-PEM FC operation at 160 and 180 °C. As seen in Figure 10, Pt/PBI-O-PhT-P/CNF shows higher performance compared with Pt/CNF. At the same time, the application of CNF coated with Nafion ® and PIM-1 as supports for platinum nanoparticles leads to much lower HT-PEM FC performance. It can be related to a less uniform polymer coating on the CNF in the case of PBI-O-PhT-P compared with Nafion ® or PIM-1, as shown above by electron microscopy. As a result, more platinum nanoparticles possess an acicular shape and grow directly on the CNF surface instead of being located on the polymer layer, which improves electron transfer during HT-PEM FC operation.
The inverted anode structure, when the polymer was applied on the already prepared CNF with deposited platinum (Pt/CNF), was also studied. The polarization curves and power density data for these samples of general formula polymer/Pt/CNF for HT-PEM FC operation at 160 and 180 °C are shown in Figure 12. As can be seen from Figure 12, the HT-PEMFC performance for the anode coated with PBI-O-PhT-P (PBI-O-PhT-P/Pt/CNF) is still higher than for the uncoated Pt/CNF sample. As can be seen from Figure 12, the HT-PEMFC performance for the anode coated with PBI-O-PhT-P (PBI-O-PhT-P/Pt/CNF) is still higher than for the uncoated Pt/CNF sample. The application of Nafion ® and PIM-1 onto the already prepared Pt/CNF resulted in a very slight decrease in HT-PEMFC performance, making its performance similar to the uncoated Pt/CNF sample.
Data on the membrane resistance and hydrogen crossover through the membrane during HT-PEMFC operation at 160 and 180 • C, obtained by the electrochemical impedance spectroscopy and linear sweep voltammetry methods, are given in Table 3. As can be seen from Table 3, the membrane resistance at 160 • C is in the range 60-90 mΩ cm −2 and, at 180 • C, is in the range 65-90 mΩ cm −2 . These values are typical for the PBI-O-PhT membrane and do not differ significantly, since the same PBI-O-PhT membrane was used in all studied MEA. At the same time, the higher resistance for samples with Nafion ® (91.2 mΩ cm −2 for Pt/Nafion/CNF) may be related to a more uniform deposition of Nafion ® on the CNF surface, according to electron microscopy data, which slightly worsens the electrical conductivity. The hydrogen crossover through the membrane at 160 and 180 • C ( Figure S2) corresponds to the typical PBI-O-PhT membrane operation in HT-PEMFC and does not exceed 3-5 mA cm −2 .

Discussion
In this part, a discussion of the main findings and the revealed general observations is provided for three proton-conducting polymers (Figure 1) applied on CNF ( Table 1). The polymers applied onto the CNF are PIM-1 (Figures 2 and 3), Nafion ® (Figures 4 and 5) and PBI-O-PhT-P (Figures 6 and 7). The data is compared with the pristine samples without polymer deposition (Figures 8 and 9).
The application of proton-conducting polymers on carbon nanofiber materials in order to increase the proton conductivity of electrodes is possible in two ways. The first possibility is to apply the polymer onto the CNF before platinum deposition. In this case, platinum nanoparticles would be open to gases, but the electron transfer between platinum nanoparticles and carbon material may be disturbed. Such a situation can be observed using electron microscopy methods when platinum nanoparticles are "immersed" into a polymer layer. Another possibility is to deposit platinum nanoparticles directly onto the CNF, followed by the deposition of a proton-conducting polymer. In this case, electron transfer between platinum nanoparticles and carbon material can be ensured. At the same time, after polymer deposition, platinum can be excessively covered with the polymer, which may prevent good contact of platinum nanoparticles with gases, particularly hydrogen. One of the best ways to evaluate the effectiveness of the presented approach is to perform HT-PEM fuel cell tests where polymer-coated CNF with deposited Pt nanoparticles are used as electrodes, particularly anodes. The presented data for the microporosity studies ( Figure 10, Table 2) are typical for microporous PAN-based CNF and suggest the possibility of using the material as a support for a Pt nanoparticle catalyst in HT-PEM FC.
According to the HT-PEM FC performance data presented in Figure 11, when Pt is deposited on polymer-coated CNF, the HT-PEM FC performance depends on the nature of the studied polymer. In the case of Nafion ® and PIM-1, the HT-PEM FC performance is significantly lower compared with the case of the uncoated Pt/CNF. At the same time, in the case of PBI-O-PhT-P, the performance is higher than in the case of the uncoated Pt/CNF. This can be explained by a less uniform PBI-O-PhT-P coating on the CNF, which leaves some parts of the CNF uncoated (as shown in Figure 6), compared with Nafion ® and PIM-1 coatings, which appear more uniform (Figures 2 and 4). A more uniform polymer coating results in poorer contacts between Pt nanoparticles and CNF and, presumably, affects the distribution of Pt nanoparticles, which also looks irregular compared with the Pt/CNF sample (Figures 8 and 9). A less uniform PBI-O-PhT-P coating on the CNF results in a significant number of places uncoated with polymer. It allows platinum nanoparticles to grow directly on the carbon surface of the CNF, however, very close to the proton-conducting polymer necessary for proton transfer. Apparently, it results in a better distribution of Pt nanoparticles and an acicular morphology of Pt nanoparticles, which has been shown in the past to be more active than spherical. The uncoated Pt/CNF also possesses a uniform distribution of platinum predominantly acicular in shape (Figures 8 and 9). Obviously, these features, combined with higher proton transfer due to polymer deposition, can improve the HT-PEM FC performance for Pt/PBI-O-PhT-P/CNF, as seen from the data in Figure 11. Thus, the HT-PEMFC performance order depending on the polymer applied onto the anode is as follows: PBI-O-PhT-P > polymer-uncoated > PIM-1 > Nafion ® The inverted method of polymer deposition, when the polymer is applied to the already Pt-decorated CNF, leads to a very different situation. Interestingly, the order of the HT-PEMFC performance depending on the polymer remains the same, but the values are much closer to each other and to the uncoated Pt/CNF sample. This can be explained by the Pt nanoparticles of acicular morphology already formed on the CNF surface with a uniform distribution, which were used for the proton-conducting polymer deposition. The application of PIM-1 and Nafion ® reduces the HT-PEM FC performance, slightly reducing H 2 transport to platinum, but not significantly, compared to the uncoated Pt/CNF sample ( Figure 12, Table 3). At the same time, the application of PBI-O-PhT-P onto Pt/CNF leads to an improvement in the HT-PEM FC performance, presumably due to better proton conductivity in the anode without preventing H 2 transport to platinum nanoparticles ( Figure 12, Table 3).
The electrochemical impedance spectroscopy and linear sweep voltammetry provide the membrane resistance and hydrogen gas crossover through the membrane values in the assembled HT-PEMFC at 160 and 180 • C. The values are similar and correspond to the typical PBI-O-PhT membrane. However, the slightly higher resistance values in the case of Nafion ® may reflect some disruption in electrical conductivity and result in inferior HT-PEM FC performance compared with other polymers and the uncoated sample.

Conclusions
The study shows that HT-PEMFC performance depends on the proton-conducting polymer applied to the anode. The order of positive effect of the polymer on HT-PEM fuel cell performance is as follows: PBI-O-PhT-P > polymer-uncoated sample > PIM-1 > Nafion ® This is true for both cases: direct deposition of a proton-conducting polymer onto the CNF followed by deposition of Pt nanoparticles and deposition of a polymer onto already prepared Pt-decorated CNF. However, for the second case of polymer deposition, the values are much closer to each other. The EIS data correspond to these findings.
As shown by electron microscopy, the difference in performance can be explained by how the polymer coats the CNF and by platinum morphology. The obtained data show the advantage of the non-uniform coating of CNF by PBI-O-PhT-P.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.