Introduction

Many of the theoretical and experimental papers published so far concerned the study of the relationship between the complex structure and function of elastin due to the different behavior of this protein in healthy and pathological tissues. Review articles [1,2,3] show that the biochemical structure of elastin contains hydrophobic domains consisting of nonpolar residues of glycine, alanine, valine and neutral proline, and cross-linking domains composed of polar lysine and alanine or lysine and proline residues. This structure is also enhanced by the thermodynamic mechanism associated with the change of entropy of hydrophobic domains as a function of temperature [4, 5], for example in the physiological range of 30–40 °C. In turn, an intensive study of the thermal [6,7,8], mechanical [9, 10] and dielectric properties [11,12,13] of elastin helps in the evaluation of the physiological response of the living tissue containing this protein to imposed external parameters, such as temperature, humidity and mechanical or electrical forces. The results of dielectric and mechanical measurements of elastin are particularly important, due to the use of similar physical terms in the analysis of the molecular mechanism, such as storage and energy loss, which are related to the tangent loss, as well as the relaxation time. Therefore, a comparative analysis of data obtained using both methods may be useful in the use of elastin as a biomaterial for tissue engineering [14,15,16,17].

The aim of this work is to study molecular dynamics of elastin in the frequency range from 100 Hz to 100 kHz at temperatures from 22 to 200 °C. Thus, expected dielectric results will provide information on the polarization and conduction processes occurring in the elastin–water system below and above the glass transition temperature near 190 °C. Our article expands the temperature range of elastin dielectric properties previously reported by others [12, 13] for different levels of hydration, and for relaxation processes below the glass transition temperature, significantly shifted to lower temperatures compared to a temperature of about 200 °C for this dry protein.

Experimental

Dielectric measurements were made at normal air pressure for elastin samples formed into pellets, obtained from the bovine neck ligament as a powder (Sigma Chemicals). These experiments for wet and dry samples were carried out using the HIOKI 3522-50 LCR impedance analyzer in the frequency (f) range from 500 Hz to 100 kHz at temperatures (T) from 22 to 200 °C. Each sample was placed between two silver electrodes with a surface area of S = 78 mm2 and spacing of d = 1 mm. A sample of wet elastin, which was air-dried at room temperature (RT) at a relative humidity of about 70%, was continuously heated from RT to 200 °C at a rate of about 1 °C min−1. To obtain a dry sample, it was kept at 150 °C for about 1 h [18] and then cooled to RT and heated exactly like a wet sample. After the water removal procedure, the loss of mass in the elastin was ~ 8% of the original mass at RT before measurements corresponding to the content of loosely bound water in wet samples.

The values of the relative permittivity, dielectric loss and conductivity of these samples were calculated from \( \varepsilon{'} = Cd/\varepsilon_{\text{o}} S \), \( \varepsilon{''} = d/\omega \varepsilon_{\text{o}} RS \) and \( \sigma = \omega \varepsilon_{\text{o}} \varepsilon{''} \), respectively, where the electrical resistance (R) and capacitance (C) are related to the electrode and the bulk materials [19], \( \varepsilon_{\text{o}} \) is the permittivity of a vacuum \( \left( {\varepsilon_{\text{o}} = 8.854\;{\text{pF}}\;{\text{m}}^{ - 1} } \right) \) and ω is the angular frequency (ω = 2πf).

Results and discussion

Figure 1 shows the temperature dependence of relative permittivity \( (\varepsilon^{{\prime }} ) \) and dielectric loss \( (\varepsilon^{{\prime \prime }} ) \) for wet and dry elastin samples at 5 and 50 kHz, i.e., for low and high frequency in the measuring range of the applied electric field. The value \( \varepsilon^{{\prime }} \) is a measure of the energy stored as a result of relaxation processes of elastin molecules, and the value \( \varepsilon^{{\prime \prime }} \) is a measure of energy dissipated as heat, resulting from the movement of charge carriers between relaxation sites. In the full temperature range, the effect of water on the dielectric properties of elastin is clearly illustrated by the higher values of \( \varepsilon^{{\prime }} \) and \( \varepsilon^{{\prime \prime }} \) for wet samples than those for dry samples. The wet elastin curves show clear maximum \( \varepsilon^{{\prime }} \) values of approximately 50 °C, which are attributed to the thermal decomposition of the water absorbed by the sample from the environment. Similar behavior was found for freeze-dried elastin using the modular differential scanning calorimetry (MDSC) technique [7]. Above 50 °C, curves \( \varepsilon^{{\prime }} \) and \( \varepsilon^{{\prime \prime }} \) for wet samples drop significantly due to diffusion of water from the samples to about 150 °C, and then at higher temperatures near the glass transition temperature Tg (~ 185 °C), reduction \( \varepsilon^{{\prime }} \) is accompanied by a rapid increase in value \( \varepsilon^{{\prime \prime }} \). This observed thermal transition for elastin below 200 °C is supported by other authors [10, 13], who also indicate that with the increase in water content in this material, the Tg value decreases. For dry elastin, as shown in Fig. 1a, values of \( \varepsilon^{{\prime }} \) are substantially unchanged to about 150 °C, and the corresponding values of \( \varepsilon^{{\prime \prime }} \) (Fig. 1b) tend to decrease. These results indicate that the accumulation of protons in polar locations on the surface of elastin molecules predominates over the mechanism of conduction of these protons, which is hindered by excess hydrophobic amino acids [4,5,6]. As in the case of wet samples, in which the temperature rises to 200 °C, the curves for dry samples also show a peak \( \varepsilon^{{\prime }} \) in Tg and increases in \( \varepsilon^{{\prime \prime }} \). To compare the electrical conduction mechanism for wet and dry elastin at 5 and 50 kHz, we presented in Fig. 2 the Arrhenius plot of logarithmic conductivity (σ) versus reciprocal of temperature (T). The results for wet and dry samples at each temperature show higher σ at 50 kHz than those at 5 kHz, because the density of fluctuating protons along the surface of molecules, as well as between these molecules, is greater in response to a shorter period (t) of the electric field used (t = f−1). In the case of wet elastin, slight changes in the proton conductivity (σ) with a positive inclination up to 50 °C do not require activation energy (ΔH) according to the relationship, σ = σο exp (− ΔH/RT), where σο is a preconditioning factor and R is gas constant. This proton conductivity behavior of the wet sample indicating the negative ΔH values is a consequence of the proton transport on the surface of the elastin molecules between the aggregates of water molecules associated with hydrogen around the hydrophobic amino acid residues. In addition, as shown in Fig. 1a, the movement of water clusters in a wet sample is supported by an increase in the value of \( \varepsilon^{{\prime }} \) to 50 °C. Because water protons and aggregation of water molecules belong to the same environment, the contribution of ΔH in the polarization and conduction mechanisms during the decomposition of water does not occur. For comparison, our previous collagen (Col) studies [20] indicate that the proton conduction processes associated with water decomposition require ΔH to break the hydrogen bonds formed by this water with polar groups on the surface of the Col molecule. A further increase in temperature to about 150 °C for a wet elastin sample (Fig. 2) also reduces the proton conductivity, confirming the lack of ΔH during the diffusion of water. In contrast, the proton conductivity values for dry elastin decrease over the entire temperature range of 20–150 °C. This means that the number of free protons released during relaxation of the polar groups decreases due to the increased accumulation of these charge carriers on the surface of the elastin molecules. This situation affects the intermolecular polarization observed as a gradual increase in the \( \varepsilon^{{\prime }} \) values illustrated in Fig. 1a. Then, up to 200 °C (Fig. 2), all curves of both wet and dry samples show a significant increase in the slope relative to T−1. This behavior is reflected in changes in the activation energy below and above Tg. Below Tg, ΔH is responsible for the breaking of hydrogen bonds formed between the molecules of bound water and the hydrophilic polar groups of the main elastin chains [12]. Thus, the results of ΔH at 5 and 50 kHz for the wet elastin are 92 and 34 kJ mol−1, and the corresponding ΔH values for the dry elastin are 42 and 11 kJ mol−1. However, for these values of ΔH, the rotation of the main chain is difficult below Tg. Therefore, only above Tg at 5 and 50 kHz, the main chains are driven by an increase in ΔH to 132 and 72 kJ mol−1 for wet elastin, and up to 97 and 35 kJ mol−1 for dry elastin.

Fig. 1
figure 1

Temperature dependencies of a\( \varepsilon^{{\prime }} \) and b\( \varepsilon^{{\prime \prime }} \) for wet and dry elastin at 5 and 50 kHz

Full size image
Fig. 2
figure 2

Arrhenius plots of log σ for wet and dry elastin at 5 and 50 kHz

Full size image

Figure 3 shows the dielectric spectra \( \varepsilon^{{\prime }} \) and \( \varepsilon^{{\prime \prime }} \) for wet and dry samples at 35 and 200 °C depending on the physiological conditions and after the end of the glass transition process. The \( \varepsilon^{{\prime }} \) lots for wet and dry at 35 °C show distinct three dispersion regions: below 3 kHz, in the range of 3–20 kHz and above 20 kHz, and the corresponding plots \( \varepsilon^{{\prime \prime }} \) also show three separate relaxations around 2, 10 and 60 kHz. As the temperature rises, the slope of \( \varepsilon^{{\prime }} \) and \( \varepsilon^{{\prime \prime }} \) plots increases to about 10 kHz, which leads to relaxation disappearance at 2 and 10 kHz. In the case of dry elastin, relaxation at 2 and 10 kHz is attributed to the local movement of the polar groups of surface macromolecules and also enhanced by the movement of water clusters around the hydrophobic groups of wet elastin. However, at an increasing temperature of 35–200 °C, the maximum \( \varepsilon^{{\prime \prime }} \) at 60 kHz for wet and dry elastin exhibits a slight change, in contrast to the decreasing \( \varepsilon^{{\prime }} \), especially for the wet sample. The peak \( \varepsilon^{{\prime }} \) at 60 kHz implies a symmetrical distribution of the relaxation time (τ) around the mean value τ ~ 3 μs, resulting from the movement of small polar groups in the chains of the main elastin molecules.

Fig. 3
figure 3

Frequency dependencies of a log \( \varepsilon^{{\prime }} \) and b log \( \varepsilon^{{\prime \prime }} \) for wet and dry elastin at of 35 and 200 °C

Full size image

Conclusions

The results of this study indicate that dielectric behavior in elastin below the glass transition temperature (Tg) is attributed to the polarization and conduction mechanisms resulting from the interaction between water and polar groups. However, above Tg, both of these mechanisms are attributed to the relaxation of the main elastin chain. In the physiological temperature range, the frequency dependences of the dielectric loss of elastin significantly indicate three separate relaxations around 2, 10 and 60 kHz, the first two of which are associated with polar groups of surface macromolecules, and with water clusters around hydrophobic groups, and the latter is assigned to small polar groups of main elastin molecules. The results of these measurements using external parameters, such as temperature and humidity, provide the ability to detect and diagnose the electrical activity of human tissues containing elastin fibers, for example during hypothermic or atherosclerotic artery process.