Study on Anhydrous Proton Conduction in Imidazole–Collagen Composite

Abstract

Recently, hydrogen-fuel cells have attracted attention as an environmentally friendly next-generation energy device. Very recently, biomaterials such as collagen and chitin have realized proton conductivity via water bridges under humidity condition, and the fabrication of fuel cells using biomaterials is possible. However, the fuel cell electrolyte via water has demerits, such as the complication of fuel cell instruments and the operating temperature limit. Therefore, fuel cell electrolytes without humidified conditions are desired. In the present work, we have synthesized an anhydrous proton conductor using imidazole and collagen, which are biomaterials, and investigated the anhydrous proton conductivity in imidazole–collagen composites. It was found that an imidazole–collagen composite is a high-proton conductor above 10⁻³ S/m and above 200 °C without the humidified condition compared with other anhydrous bio-proton conductors such as the hydroxyapatite–collagen composite. Moreover, the motional narrowing of the ¹H-NMR line width reveals that the proton conductivity is realized in the temperature region from 120 to 200 °C. In addition, the DTA measurement and the impedance analyses reveal that the imidazole–collagen composite film undergoes the phase transition at 120 °C. Furthermore, the proton conductivity in the imidazole–collagen composite strongly depends on n, which is the imidazole concentration per collagen molecule and takes a maximum at n = 2.0. In addition, the proton conductivity perpendicular to the collagen fiber is approximately ten times higher than that parallel to the collagen fiber. From these results, it can be deduced that the proton conductivity in the imidazole–collagen composite is caused by breaking and rearranging the hydrogen bonds of the collagen side chain with the imidazole molecule formed between the collagen fibers.

1. Introduction

As is well known, fuel cells are attracting attention as a next-generation environmentally friendly energy source because of their ability to generate energy at high efficiency using a simple reaction that produces water from hydrogen and oxygen. The hydrogen fuel cell, which is one of the fuel cells that can operate in low-temperature regions, is most widely useful as a power source of mobile devices and automobiles [1,2,3,4,5]. In the hydrogen fuel cell, a proton conductor is necessary as the electrolyte, and the development of inexpensive and environmentally friendly proton conductors is desired.

In recent years, tissue-derived biomaterials such as nucleic acids, sugars, and proteins produced by living organisms have attracted attention as the next generation of environmentally friendly materials because they decompose naturally [6,7,8,9,10,11,12,13,14]. These biomaterials are often found in discarded items such as fish scales, crab and shrimp shells, tree pieces, etc., and are available in large quantities and at low cost. Moreover, it is known that these biomaterials have proton-conducting properties [8,9,10,11,12,15,16]. Fibrous biomaterials such as collagen and chitin exhibit proton conductivity by forming water cross-links in their structures under high-humidity conditions [17]. It has also been reported that fuel cells can be fabricated using these biomaterials as electrolytes [6,7,8,9,18]. In addition, it is known that ion channels, which are membrane proteins, can rapidly transport sodium and potassium ions and protons through pores in the structure. We have recently demonstrated that ion channel membranes extracted from the squid axon are proton-conductivity membranes [9]. These electrolytes belong to the proton-exchange membrane fuel cell (PEM fuel cell).

The PEM fuel cell has one more hurdle to overcome, namely the realization of no humidification. Anhydrous proton conductors have attracted particular attention because they can be used in temperature ranges below 0 °C and above 100 °C, simplifying the structure of fuel cells. However, there are only a few materials that can achieve proton transport under non-humidified conditions [19,20,21,22,23,24]. For example, it has been reported that the solid-acid fuel-cell electrolytes CsHXO₄, M₃H(XO₄)₂, and CsH₂PO₄ (M: K, Rb, Cs, Tl; X: S, Se), become anhydrous fuel cell electrolytes at around 200 °C [25,26,27,28,29,30,31]. The proton conductivity in these materials is realized by breaking and rearranging hydrogen bonds formed in the crystal in a non-humidified condition. It is also known that imidazole-doped materials and imidazolium salt composites are non-humidified proton conductors [3,32,33,34,35,36]. Imidazole is a five-membered ring aromatic molecule with two nitrogen atoms. The proton in imidazole has two equivalent positions in the vicinity of two annular nitrogen atoms, and proton transfer can be achieved by rotating the imidazole’s five-membered ring. However, there are very few reports concerning anhydrous bio-proton conductor doping imidazole to biomaterials. The only anhydrous proton conductor using biomolecules is the composite of hydroxyapatite (HAp), which is a major component of bone [37,38]. Very recently, we synthesized the Hap–collagen composite and successfully showed that the HAp-collagen composite becomes an anhydrous proton conductor in the temperature range from 110 °C to 200 °C [39].

Figure 1 shows the temperature dependence of proton conductivity and the proton-conductivity pathway in the HAp-collagen complex. Proton conductivity appears above 110 °C in Hap–collagen composites, as shown in Figure 1a. Based on the anisotropic proton conductivity shown in Figure 1a, the anhydrous proton conductivity is deduced to be achieved by utilizing the breaking and rearrangement of hydrogen bonds between hydroxyl groups in the structure of the Hap–collagen composite (Figure 1b). However, as shown in Figure 1a, low-proton conductivity is a weakness of the Hap–collagen composite. In order to find an anhydrous bio-proton conductor with a higher proton conductivity, we recently prepared an imidazole–collagen composite film, investigated its proton conductivity, and found that this composite film exhibits higher anhydrous proton conductivity. In the present paper, we report these results in the imidazole–collagen anhydrous bio-proton conductor. These results will lead to the development of new environmentally friendly anhydrous proton conductors.

2. Materials and Methods

2.1. Preparation of Collagen Membrane and Imidazole–Collagen Composites

For the preparation of collagen film, the artificial collagen with the sequence of glycine (Gly)-hydroxyproline (Hyp)-proline (Pro) (polypeptide-6: UNIQS. Co., Ltd., Tokyo, Japan) was used in the present work. The collagen membrane was prepared by casting 300 μL of collagen (0.5% polypeptide-6 collagen solution) onto a Teflon substrate and then drying in a desiccator at 45 °C for 24 h. The imidazole–collagen composite membranes were prepared by dissolving imidazole powder in a collagen solution and then casting 300 μL of the obtained mixture onto a Teflon substrate and drying in a desiccator at 45 °C for 24 h. In Figure 2, we show photographs of the prepared collagen membrane and the imidazole–collagen composites of n = 2.0 and n = 4.0, where n is the imidazole molecular concentration per collagen molecule (Gly-Pro-Hyp). The membranes of collagen and imidazole–collagen composites were successfully prepared, as shown in Figure 2. Figure 3 exhibits micrographs for the collagen membrane and the imidazole–collagen composite (n = 2.0) at 25 °C and 200 °C. As shown in Figure 2a–d, it is evident that there are no changes in the shape of the collagen film and the imidazole–collagen composite until 200 °C. These results mean that the collagen film and the imidazole–collagen composite film do not melt until 200 °C.

2.2. FT-IR Measurement

The measurements of FT-IR spectra in imidazole, collagen membranes, and the imidazole–collagen complex were carried out in the wavenumber from 500 cm⁻¹ to 4000 cm⁻¹ by an FT-IR spectrometer (iS-5, Thermo Fisher Scientific, Waltham, MA, USA). The lyophilization process was also performed to dehydrate the samples. During the measurements, the samples were put in the vacuum conditions of 5.33 KPa using a dry vacuum pump (DA-20D: ULVAC. Co., Ltd., Kanagawa, Japan).

2.3. Impedance Measurements

The impedance measurement in the imidazole, collagen, and imidazole–collagen membrane was carried out using the precision LCR meter (E4980A: Agilent technology, Santa Clara, CA, USA). The impedance measurements were carried out with the frequencies from 1kH to 1MHz in the temperature from 100 °C to 210 °C. The impedance in the imidazole–collagen composite was measured parallel to the collagen-main chain and normal to the collagen-main chain. The samples were set in the dry vacuum tube of 5.33 KPa with a dry vacuum pump (DA-20D: ULVAC. Co., Ltd., Japan). The gold electrodes were deposited on the specimens. The thickness of the collagen and imidazole–collagen composite membranes used in the impedance measurement is adjusted to 20 μm.

2.4. DTA Measurement

The DTA curve was measured in the temperature region from room temperature to 200 °C using a homemade DTA instrument, which was constructed with a digital multimeter (Keithley 2000, Keithley instruments) and a computer. The temperature control and data acquisition were carried out using a computer. The alumina powder was used as the standard sample, and the cupper-constantan thermocouple was used to measure temperature. In the DTA measurements, the powdered specimen of the collagen membrane or the imidazole–collagen composite was used, and the specimens were sufficiently dehydrated in a vacuum at 150 °C.

2.5. Measurement of ¹H-NMR

¹H-NMR spectra in the imidazole–collagen composite (n = 2.0) were measured by a homemade pulse-NMR spectrometer. The pulse-NMR spectrometer constructed from a Multi-Function Synthesizer (WAVE FACTORY, Kanagawa, Japan), a pulse generator (N146-4746AM, Thamway, Shizuoka, Japan), a Multi-Function Generator (WAVE FACTORY, Kanagawa, Japan), and an amplifier (Thamway, Shizuoka, Japan) for the pulse transmitter and receiver (Thamway, Shizuoka, Japan). We sealed the sample in the glass tube of 10 mm diameter and observed ¹H-NMR absorption lines at a resonance frequency of 9.979 MHz from 80 °C to 200 °C.

3. Results and Discussion

Figure 4a–c show the FT-IR spectra in the imidazole, collagen film, and imidazole–collagen composite film of n = 2.0, respectively. Here, n is the imidazole molecular concentration per collagen molecule (Gly-Pro-Hyp). The absorbances for the collagen film are observed at 1403 cm⁻¹, 1453 cm⁻¹, 1553 cm⁻¹, and 1639 cm⁻¹, as shown in Figure 4b. These absorbances are specific to collagen film. For example, the absorbances at 1639 cm⁻¹, 1553 cm⁻¹, and 1453 cm⁻¹ are known as those of C=O stretching motion, N-H stretching motion, and C-N vibration motion, respectively [40,41]. It is noted that the absorbances at 752 cm⁻¹, and 1063 cm⁻¹, which cannot be observed in the collagen film, are observed with the slight peak shift in the imidazole–collagen composite in Figure 4c. These absorbances correspond to the specific ones in imidazole and derive from C-H out-of-plane bending [36]. In addition, it is also noted that the absorbances corresponding to the ring bending of imidazole observed at around 615 cm⁻¹ and 662 cm⁻¹ are reduced by forming the imidazole–collagen composite film. This result indicates that the ring bending is suspected by binding between imidazole and collagen. It is deduced that imidazole is introduced into collagen and that the imidazole–collagen composite is formed.

Figure 5a,b show the DTA curves in the collagen and the imidazole–collagen composite film of n = 2.0, respectively. We can see that the thermal anomaly cannot be observed in the collagen in the temperature range from 30 °C to 230 °C, as shown in Figure 5a. On the contrary, as shown in Figure 5b, the endothermic peak is observed at around 120 °C with increasing temperature in the imidazole–collagen composite. It is known that the phase transition by melting of imidazole occurs around 90 °C. As shown in Figure 5b, the endothermic peak at approximately 90 °C cannot be observed in the imidazole–collagen composite film. In addition, in the photograph of the imidazole–collagen composite film observed at 120 °C, there is no change in the imidazole–collagen composite film. These facts indicate that the endothermic peak at around 120 °C is not caused by the melting of the imidazole–collagen composite but by the phase transition of the imidazole–collagen composite.

Figure 6 shows the frequency dependence of AC electrical conductivity σ_AC at 160 °C in the collagen membrane and the imidazole–collagen composite of n = 2.0. As shown in Figure 6, σ_AC in the imidazole–collagen composite becomes 10³ times higher than that in the collagen membrane.

These results indicate that the synthesis of collagen and imidazole yields a new electric-conduction pathway that cannot be observed in the collagen film or imidazole.

In order to investigate the proton dynamics in the imidazole–collagen composite film, we measured the ¹H-NMR absorbance line.

Figure 7 shows the results of ¹H-NMR in the imidazole–collagen composite (n = 2.0). ¹H-NMR absorbance lines, which were broad at 100 °C, become sharper with increasing temperature, as shown in Figure 7a. This result indicates the motional narrowing due to proton motion. That is, protons migrate in the imidazole–collagen composite. In Figure 7b, we show the temperature dependence of the line width ΔH of the NMR absorption line. As shown in Figure 7b, ΔH decreases with increasing temperature until around 110 °C, shows a small peak at around 120 °C, and monotonously decreases with increasing temperature again. It is speculated that the anomalous behavior at around 120 °C results from the phase transition of the imidazole–collagen composite film. From the temperature dependence of ΔH, we can obtain the temperature dependence of the correlation time τ_c concerning proton motion. The correlation time τ is calculated from the equation [42]

$${\tau }_c=\frac{2\pi }{\alpha \gamma \Delta H}tan\left[\frac{\pi }{2}\frac{∆H^2-B^2}{C^2-B^2}\right],$$

(1)

where C and B are the NMR line widths before and after motional narrowing, respectively, and γ and α are the gyromagnetic ratio of the proton and the constant of (8ln2)⁻¹, respectively. Figure 7c shows the temperature dependence of τ_c calculated using Equation (1). As shown in Figure 7c, τ_c decreases by increasing temperature. Furthermore, it is noted that the slope of the temperature dependence of τ_c changes at around 120 °C.

This result means that the correlation time concerning proton motion changes at the phase transition of 120 °C, indicating that proton transfer becomes even faster above 120 °C. It is deduced that the phase transition at around 120 °C is caused by the appearance of the rotational motion of imidazole molecules in the imidazole–collagen composite. In addition, the activation energy for proton motion can be obtained from the temperature dependence of τ_c. The activation energy for microscopic proton motion above 120 °C in the imidazole–collagen composite is calculated to be 0.72 eV. It is deduced that the electrical conductivity in the imidazole–collagen composite results from proton transfer with an activation energy of 0.72 eV. This result indicates that the electrical conductivity in Figure 6 results from proton conduction.

Next, we investigated the relation between the molecular dynamics and proton conductivity in the imidazole–collagen composite in detail using the impedance measurement. In Figure 8a, we show the frequency dependence of AC proton conductivity in the imidazole–collagen composite (n = 2.0). The value of σ_AC increases with increasing frequency in all temperature regions, as shown in Figure 8a. When the imidazole–collagen proton conductor is described with the parallel equivalent circuit of resistance R and capacitance C, σ_AC obeys the equation

σ_AC = σ₀ + ωε₀ε″,

(2)

where σ₀ is DC proton conductivity derived from resistance R, and ε″ is the imaginary component of the dielectric constant. In addition, ε₀ and ω are the dielectric constant in a vacuum and the angular frequency, respectively. We show the frequency dependence of σ_AC calculated with Equation (2) as a dashed line in Figure 8a. We can clearly see that the frequency dependence of the measured σ_AC does not agree with that of the calculated one. This result indicates that the equivalent circuit of the imidazole–collagen composite cannot be expressed by the parallel equivalent circuit of R and C. As is well known, biomaterials exhibit the non-Debye dielectric dispersion derived from α relaxation [43,44]. Considering this fact, we need to consider that the AC proton conductivity in the imidazole–collagen composite includes dielectric dispersion. The AC conductivity, which includes dielectric dispersion, is expressed as the following equation,

$$\begin{array}{c}{\sigma }_{\mathrm{AC}}={\sigma }_0-\mathrm{Im}\left[\omega {\epsilon }_0{\epsilon }_{\infty }+\frac{\omega {\epsilon }_0\left({\epsilon }_s-{\epsilon }_{\infty }\right)}{1+{\left(j\omega \tau \right)}^{\beta }}\right].\\ ={\sigma }_0+\frac{\omega {\epsilon }_0\left({\epsilon }_s-{\epsilon }_{\infty }\right){\left(\omega \tau \right)}^{\beta }\sin \left(\frac{\pi }{2}\beta \right)}{{\left(1+{\left(\omega \tau \right)}^{\beta }\cos \left(\frac{\pi }{2}\beta \right)\right)}^2+{\left({\left(\omega \tau \right)}^{\beta }\sin \left(\frac{\pi }{2}\right)\right)}^2}\end{array}$$

(3)

Here, ε_∞ and ε_s are high-frequency (unrelaxed) permittivity and the static dielectric constant in the imidazole–collagen composite, respectively. The symbols of β and τ are the degree of multi-dispersion and the relaxation time for the dielectric dispersion, respectively. Using Equation (2), the frequency dependence of σ_AC in the imidazole–collagen composite can be calculated. The results are displayed with the solid line in Figure 8a. The values of fitting parameters are listed in

Table 1. List of parameters using the fitting of AC proton conductivity.

T (°C)	σ0 (S/m)	εs-ε∞	τ (s)	β
100	1.4 × 10−7	40	4.0 × 10−2	0.26
115	6.0 × 10−7	90	8.0 × 10−2	0.29
120	1.1 × 10−6	165	1.5 × 10−1	0.31
160	6.8 × 10−5	85	6.0 × 10−5	0.41
190	5.3 × 10−4	80	5.0 × 10−6	0.45

. The calculated frequency dependences are in excellent agreement with the measured ones, as shown in Figure 8a. These results indicate that dielectric dispersion exists in the imidazole–collagen composite. In addition, we can obtain DC proton conductivity from the analyses of Equation (2).

In Figure 8b, we show the temperature dependence of direct-current proton conductivity σ₀ in the imidazole–collagen composite. The proton conductivity log σ₀ increases with the increase in temperature and is proportional to 1/T above 120 °C, as shown in Figure 8b. From this result, we can obtain the activation energy for proton conductivity. It was found from this result that the activation energy for proton conductivity above 120 °C is 0.98 eV. These values are close to the activation energy (0.72 eV) obtained from NMR measurement, although these values are slightly larger than those with the activation energy obtained from NMR measurement. Considering that the activation energy obtained from NMR proton correlation time derives from local proton transfer, it is speculated that proton transfer in the macroscopic proton conduction is prevented by sites with slightly higher activation energy existing in the long proton-conduction pathway.

Figure 8c shows the temperature changes in the dielectric relaxation time τ and the static dielectric constant ε_s-ε_∞ calculated using Equation (3). The dielectric relaxation time τ becomes long at around 120 °C and becomes shorter with increasing temperature, as shown in Figure 8c. The static dielectric constant ε_s-ε_∞ peaks at around 120 °C and increases with increasing temperature. As is well known, in the materials that undergo a phase transition, the critical slowing down is often observed due to the strong correlation of electric dipole moment in the vicinity of the phase transition. In this case, the dielectric constant increases, and the dielectric relaxation time becomes long at around the phase transition temperature.

Considering these facts, the anomalous behaviors of ε_s-ε_∞ and τ at around 120 °C are caused by the critical slowing down in the imidazole–collagen composite. It is noted that the relaxation time of the free flip-flop motion of imidazole in the imidazole crystal is reported to be approximately 10⁻⁵ s [34].

This relaxation time is close to that at around 170 °C in Figure 8c. Based on these results, the dielectric relaxation obtained in the present work is closely related to the flip-flop motion of the imidazole molecule in the imidazole–collagen composite. In addition, it is also noted that the relaxation time of the flip-flop motion of imidazole is close to that of proton correlation time τ_c. It is speculated that the flip-flop motion of imidazole plays an important role in the realization of proton conductivity in the imidazole–collagen composite.

Next, in order to investigate the relation between proton conductivity and the amount of imidazole molecules, we prepared the imidazole–collagen composite films with the various imidazole concentration and measured the proton conductivity and FT-IR spectra. The imidazole–collagen composite films prepared are n = 0, 0.2, 1.0, 1.2, 1.4, 1.6, 2.0, 3.0, and 4.0. Figure 9a,b show the FT-IR spectra of the imidazole–collagen composite for the various imidazole concentration. The absorbances at 1403cm⁻¹, 1453 cm⁻¹, 1553 cm⁻¹, 1639 cm⁻¹, and 3347 cm⁻¹ decreases with increasing n shown in Figure 9a,b. These absorbances at 1403cm ⁻¹, 1453 cm⁻¹, 1553 cm⁻¹, 1639 cm⁻¹, and 3347 cm⁻¹ correspond to the absorbance derived from C-N stretching, C-N stretching, N-H stretching, C-N vibration, C=O stretching, and O-H stretching in collagen, and therefore, the decrease in the intensity of absorbances with increasing n indicates that the imidazole molecules are bonded with the collagen molecules [45]. Figure 9c shows the imidazole concentration n dependence of DC conductivity σ₀ and the activation energy ΔE_a. Here, σ₀ is obtained from Equation (3), and ΔE_a is obtained from the temperature dependence of σ₀. As shown in Figure 9c, σ₀ increases with increasing n until n = 2.0 and decreases at n > 2. The activation energy takes a minimum at n = 2.0. In this way, the proton conductivity and the activation energy strongly depend on the imidazole concentration. That is, these results indicate that the imidazole concentration is closely related to the proton conductivity in the imidazole–collagen composite film. Furthermore, it is also noted that σ₀ and ΔE_a exhibit steep change at n = 1.0. It is deduced that the proton-conduction pathway begins to form at n = 1.0 and connects above n > 1. In addition, we can see that, at n > 2, σ₀ decreases and ΔE_a increases. These phenomena would be caused by the inhibition of macroscopic proton conduction due to the excessive imidazole molecules. It is known that the imidazole crystal undergoes proton conduction by breaking and rearranging hydrogen bonds in the imidazole molecule, accompanied by the continuous reorientation motion of the imidazole molecule [3,32,34,35,36]. From the FT-IR spectra, it is confirmed that O-H, C-N, N-H, and C=O groups of collagen and N-H groups of the imidazole molecule are connected by the hydrogen bond in the imidazole–collagen composite. Considering that hydrogen bonds are weak and can rearrange above 100 °C, the breaking and rearrangement of hydrogen bonds between collagen and imidazole molecules yield the proton conduction in the imidazole–collagen composite film. On the basis of these results, we show the schematic figure concerning the proton conduction in the imidazole–collagen composite film in Figure 10.

Next, we shall discuss the proton conductivity in the imidazole–collagen composite film using the schematic structure. From the FT-IR results, it was found that imidazole binds to the carbonyl oxygen (C=O), amide nitrogen (N-H), and the hydroxyl group (O-H) of hydroxyproline in collagen. Therefore, a schematic diagram of the imidazole–collagen composite at n = 2.0 based on these results is shown in Figure 10. As shown in Figure 10a, two imidazole molecules are bound to a collagen molecule at the binding site obtained by FT-IR measurement. Thus, the imidazole molecules form a hydrogen bond with the collagen side chain and exist around the collagen main-chain axis. It is easy to predict that the hydrogen bonds between imidazole and collagen will be continuously broken and rearranged, considering the rotation of imidazole above the phase transition at 120 °C, as obtained by DTA, impedance analysis, and NMR. In Figure 10b, we illustrate the proton transfer in the imidazole–collagen composite film. Here, the thick orange line in Figure 10b shows the proton-conduction pathways. As shown in Figure 10b, it can be seen that the protons are transferred to another collagen chain due to the breaking and rearrangement of the hydrogen bonds by the imidazole rotates. Thus, it seems that the imidazole–collagen composite exhibits proton transport mainly in the direction perpendicular to the main collagen chain. Based on this structure, the proton conduction’s dependence on imidazole concentration n as shown in Figure 9c can be interpreted. At n = 1.0, the conduction path connecting imidazole and collagen begins to form, and proton conductivity increases. At n = 2.0, two pathways in Figure 10b (orange thick line) are formed, and the conductivity is inferred to increase. On the other hand, for n > 2, by further bonding of imidazole, the rotational motion of the imidazole is inhibited, and the breaking and rearrangement of hydrogen bonds are suppressed, leading to a decrease in conductivity. In this way, the imidazole molecule connects between collagen fibers and the rotation of the imidazole yields proton conduction with the breaking and rearrangement of hydrogen bonds between collagen and imidazole. Considering that the imidazole exists at a position perpendicular to the collagen fiber direction, it is noted that, in this case, proton conductivity perpendicular to the collagen fiber becomes high compared with that parallel to the collagen fiber. Therefore, we have investigated the anisotropy of proton conductivity.

Figure 11 shows the temperature dependence of proton conductivities parallel and perpendicular to collagen fiber in the imidazole–collagen composite film (n = 2.0). As shown in Figure 11, proton conductivity perpendicular to the collagen fiber direction is approximately ten times higher than that along the collagen fiber.

These results are consistent with our proton-conduction model in Figure 10. Further experiments, such as the quasi-elastic neutron diffraction, will be planned for the microscopic verification of the model speculated in the present work. These results will appear in future issues.

4. Conclusions

In the present study, we prepared a new bio-proton conductor using imidazole and collagen and investigated the proton conductivity in the imidazole–collagen composite with FT-IR, ¹H-NMR, and impedance measurement. It was found that the imidazole–collagen composite becomes a proton conductor and is approximately 10⁻³ S/m at 200 °C. This value is 10⁴ or higher than other bio-proton conductors of the hydroxyapatite film and the hydroxyapatite–collagen composite film. Moreover, from the motional narrowing of the ¹H-NMR line, the imidazole–collagen composite film becomes a proton conductor above 120 °C. In addition, the activation energy for the proton correlation time τ_c of a proton motion is estimated to be 0.72 eV above 120 °C. Furthermore, from the DTA measurement and the impedance analyses, the imidazole–collagen composite undergoes the phase transition at 120 °C. The relaxation time of the dielectric dispersion by the flip-flop motion of the imidazole is estimated to be ~10⁻⁵ s at around 170 °C. This value is close to that of the proton correlation time. It is speculated that the flip-flop motion of the imidazole plays an important role in the realization of proton conductivity in the imidazole–collagen composite. By preparing the imidazole–collagen composites with the different imidazole concentrations n, we obtain the result that proton conductivity in the imidazole–collagen composite abruptly increases at around n = 1.0, reaches a maximum at n = 2.0, and decreases by increasing n thereafter. These results indicate that the proton-conduction pathway begins to connect at n = 1.0, and the path is entirely constructed at n = 2.0. From these results, it is deduced that the rotation of the imidazole molecule bonded between the collagen fibers and imidazole molecules causes the breaking and rearrangement of hydrogen bonds and that proton conductivity yields in the imidazole–collagen composite. This model is consistent with the anisotropy of proton conductivity in the imidazole–collagen composite film. These results in this study are expected to be helpful for the development of new non-humidified proton conductors based on biogenic materials.