Effect of protein on the thermogenesis performance of natural rubber matrix

Under high-speed strain, the thermogenesis performance of natural rubber products is unstable, leading to aging and early failure of the material. The quality of rubber latex and eventually that of the final products depends among others on the protein content. We found that when the protein is almost removed, the heat generated by the vulcanized rubber increases rapidly. After adding soy protein isolate to the secondary purification rubber, the heat generation of the vulcanized rubber is reduced, and the heat generation is the lowest when the added amount is 2.5–3.0 phr, which on account of protein promotes the construction of a vulcanization network and increases the rigidity of the rubber chain, resulting in a decrease in the potential frictional behavior of the rubber chain during the curl up-extension process.


Results and discussion
Deproteinization section. Effect of protein removal on thermogenesis performance of NR matrix. The temperature rise of the sample is shown in Fig. 1. As the protein content of the NR matrix decreases, the temperature rise of the sample under cyclic stress gradually increased. It can be seen from Fig. 1a that the external temperature rises of FNR, CNR-1, CNR-2, and CNR-3 were 9.6 °C, 18.3 °C, 35.1 °C, and 57.0 °C. Among them, CNR-3 has the highest thermogenesis, which has increased by 47.4 °C compared to FNR, and its temperature still has big uptrend with time. As shows in Fig. 1b, the internal temperature rises of FNR, CNR-1, CNR-2, and CNR-3 were 20.7 °C, 32.4 °C, 39.1 °C, and 60.4 °C. Among them, CNR-3 has the highest thermogenesis, which is 39.7 °C higher than FNR. This is due to the fact that the protein in the natural latex is decomposed into amino acids after enzymatic hydrolysis, which is removed in layers after high-speed centrifugation 30 . The loss of protein will affect the vulcanization kinetics 31,32 , resulting in a weakening of the interaction force between the molecular chains 33 , an increase in the curl up-extension between the molecular chains and in heat dissipation.
Effects of protein removal on the composition and structure of NR matrix. As shown in Fig. 2a, as the number of centrifugations increases, the protein content of the sample decreases rapidly. The protein content of FNR is 2.01 m%, and CNR-3 is reduced to 0.44 m% (weight percentage). The hydrolysis reaction will degrade the bound protein into small molecule amino acids. After centrifugation, the small molecule amino acids and some free water-soluble proteins were separated from the polyisoprene chain. The more centrifugation times, the lower is the protein content in the latex. The infrared spectra of the samples of different contents of protein are shown in Fig. 2b. Because the experimental conditions are consistent, quantitative comparison can be made. The amide I vibration peak at 1630 cm −1 is weaker, which is masked by the C=C peak. Therefore, the 1540 cm −1 amide II characteristic peak was used to characterize the protein. As shown in Fig. 2d, as the number of centrifugations increases, the intensity of the 1540 cm −1 peak weakens significantly. Although enzymatic hydrolysis and centrifugation can effectively remove proteins in natural latex, centrifugal treatment may also cause the loss of other small molar mass non-rubber components (acetone soluble, water soluble and inorganic salts etc.). As the number of centrifugations increases, the intensity of these peaks (marked in the Fig. 2c,d) changed significantly, indicating that it cannot be inferred that only the protein is removed, and it may be accompanied by the loss of other non-rubber components. Therefore, centrifugation leads to an increase in thermogenesis, which cannot be attributed to protein alone.
Analysis of the effect of protein removal on thermogenesis performance. The protein removal has a great effect on the vulcanization kinetics of NR 32 . As shown in Fig. 3a, the maximum torque value (M HR ) and the positive vulcanization time (t 90 ) of the samples significantly changed. The M HR of FNR was 5.44 dN . m, and the CNR-3 was reduced to 2.58 dN m. T 90 of FNR was extended from 5.2 to 20.3 min. It shows that as the protein content decreases, vulcanization kinetics is significantly weakened. The delay of the vulcanization process leads to poor compactness of the chemical cross-linking network, and the mutual binding force between the molecular chains will also be weakened accordingly The cross-linking density of the vulcanized sample was shown in Fig. 3b. The length of the segment between the adjacent crosslinking points increases. As the crosslinking density decreases, the compactness of the vulcanization network decreases and the flexibility increases. As shown in Fig. 3c, the removal of the protein will cause the glass transition temperature of the sample to move at a low temperature (− 61.35 °C to − 63.28 °C), it shows the movement of molecular segment will become easier. It can be inferred that when loaded the same external cyclic stress, the molecular chain of CNR-3 has a lower www.nature.com/scientificreports/ rigidity and resistance to deformation than the FNR vulcanization network, and the thermogenesis by friction between the chains increases, resulting in more hysteresis energy dissipation and heat generation. Our recent research found that protein constitutes a sacrificial network embedded in the cross-linked network, which breaks preferentially upon deformation to dissipate energy 23,24 . So we can infer that when stress is loading, the preferential break of the protein network will absorb part of the energy to reduce heat generation. Moreover, the protein free in the NR matrix similar to filler will attach to the polyisoprene chain in the form of entanglement-like 22,34,35 . We have been proven that entanglement restricts the movement of molecular chains thus reduces heat generation 7 . To better understand the effect of free protein on the thermogenesis performance of NR matrix, it is necessary to explore the influence of entanglement-like on the thermogenesis in the vulcanization network. In this paper, we use a tube theory [36][37][38][39][40][41] applicable to vulcanized rubber to calculate the degree of contribution of topological entanglement in the rubber matrix to the vulcanization network modulus.
where σ M is the reduced stress; σ is the actual stress; G e is the elasticity modulus contributed by the restriction effect of entanglement points; β is an empirical parameter to describe the relationship between a deformed tube in a stretching state and an un-stretched tube in an equilibrium state. In unfilled rubbe, β is assigned 1 37 . α is the ratio of the stretched length to the original length. The value G e is the slope of the linear fit curve derived through fitting points with their X coordinate range being 0.4-0.7 in the Eq. (1a) 42 . The calculation result is shown in Fig. 4, with the protein removal, the modulus contributed by physical entanglements decreases in order, which are 0.166, 0.105, 0.066, 0.022 MPa, respectively. G c also decreased significantly, which represents elastic modulus resulting from the contribution of chemical cross-linking 43 . By combination with Fig. 1, it can be shown that the greater the number of entanglement points, the lower the thermogenesis. it can be attributed to the entanglement-like points of free protein in the NR matrix can "pinned" the molecular chains 7 , limiting the movement of molecular chains, reducing mutual friction between molecular chains and thermogenesis. Figure 5 shows the storage modulus (E') and loss factor (tanδ) of the sample under temperature scanning. Figure 5a shows that the E' of FNR is the largest. As the protein content decreases, E' decreases significantly. It shows that the interaction force between the molecular chains gradually decreases, and the strength of the www.nature.com/scientificreports/ sample network becomes weak. As shown in Fig. 5b, as the protein content decreases, the peak of tanδ moves towards low temperature (− 46.74 °C to − 54.39 °C). This can be inferred that the increased flexibility of the molecular chain is attributed to the decrease in crosslinking and entanglement points, resulting in a decrease in the restriction on the molecular chain. The reason for the inconsistency with Fig. 3c is that the sample tested by DSC is unvulcanized rubber. The removal of protein causes a significant increase in the thermogenesis of the NR matrix for two reasons:  www.nature.com/scientificreports/ 1. Protein can promote the formation of the vulcanization network, increase the compactness of the crosslinked network. The enhancement of the interaction force between molecular chains makes movement of molecular chains more difficult, and the heat generated by its friction behavior is reduced accordingly. 2. We recently observed the spatial organization of non-rubber components through high-resolution TEM and found that the protein formed a sacrificial network tangled in a cross-linked network 23,24 . According to the calculation of the tube model, it can be known that the free protein can be effect as the entanglement-like point, which can "pinning" the molecular chain and restrict the movement of the molecular chain. As shown in Fig. 6, under the same stress, the NR matrix vulcanization network with less protein content deforms more (|λ − λ 2 | > |λ − λ 1 |). The friction resistance overcome during the molecular chain curl up process became larger, and energy loss increases in the process of overcoming resistance, which increases the heat generated by the rubber matrix.
External protein validation section. In addition to protein, the NR matrix also contains other nonrubber components such as acetone soluble, water soluble and inorganic salts etc.. The centrifugal treatment of natural latex will cause the loss of other non-rubber components in rubber, and it cannot be accurately inferred whether it is the effect of protein alone. In order to more accurately investigate the effect of protein on the thermogenesis performance of NR, gradient protein was added to CNR-2 during the mixing process for verification. It can be seen from Fig. 2a that the protein removal of the sample is relatively complete. If the thermogenesis of the sample decreases with the increase of the content of protein added, the effect of protein on the thermogenesis performance of NR can be demonstrated.
Effects of external protein on the composition and structure of NR. Figure 7a is the infrared spectrum of the sulfurized sample after adding soy protein isolate. As shown in Fig. 7b, as the external protein content increases,  www.nature.com/scientificreports/ the characteristic peak of 1540 cm −1 derived from amide II is enhanced. Indicating that with the mixing and vulcanization processing, soy protein was compatible with NR matrix.
Effect of external protein on thermal performance of NR matrix. As the amount of protein added increased, the temperature rise of the sample was clearly decreased. As shown in Fig. 8a Among them, 2.5 SOY has the lowest internal temperature rise, which is 12.7 °C lower than 0.0 SOY. It can be inferred that with the addition of protein, the thermogenesis of rubber can be suppressed, but it is not better to add as much as possible. It can be seen from Fig. 8b that the internal temperature rise of 3.0SOY (29.7 °C) is higher than that of 2.5SOY (27 °C). This attributed to when a large amount of protein is added, it will lead to poor compatibility between the protein and the rubber matrix 22 .

Materials and methods
Materials. Fresh    Measurements and characterization. Heat generating test: RH-2000 N thermogenesis analyzer produced by Taiwan Gotech co. was adopt to test the temperature rise of the external and internal part of the sample. Figure 9 is the morphology of thermogenesis test cavity and samples. As shown in Fig. 9a, the temperature inductor at the bottom of the instrument can monitor external temperature changes of sample in a real-time manner. The temperature rise of the internal part of the sample will be monitored by a thermocouple. Test 3-5 times for each parallel sample, and take the curve with the best reproducibility as the result. The sample was compression curing in a cylindrical mold cavity, its specification was shown in Fig. 9b FTIR spectra analysis: TENSOR27 FTIR spectrometer produced by German Bruker co. was used to analyze the composition and structure of vulcanized rubber. The sample was cut into a thin film, and attenuated total reflection was used to obtain the spectra. The wavenumber range was 4000-650 cm −1 and scan 32 times.
Protein content: The K9860 automatic Kjeldahl meter produced by China Haineng co. was used to determine the protein content of raw rubber. The formula for calculating protein content is as follows: W Protein = W Nitrogen × 6.25.
Vulcanization curve: The QLB-D type vulcanizer produced by China Jiangdu co. was used to test the vulcanization characteristics of the rubber, and the test temperature is 145 °C. Test 3-5 times for each parallel sample, take the curve with the best reproducibility as the result.
Stress-strain test (used for tube theoretical fitting): The AI-3000 tensile tester produced by the Taiwan Gotech Company was used to determine the tensile properties of the vulcanized rubber. The sample was a dumbbell shaped strip with central dimensions of 25 mm × 6 mm × 1 mm, and the tensile rate was 500 mm/min. Test 3-5 times for each parallel sample, take the curve with the best reproducibility as the fitting data.
Determination of crosslink density: The crosslink density of vulcanized rubber was tested by using German IIC-XLDS-15HT type nuclear magnetic resonance crosslink density. The sample was a thin rectangular slice shape. The test temperature was 80 °C, the frequency was 15 MHz, and the magnetic induction intensity was 315 A/m. Test 5 times for each parallel sample, take the average as the result.
Glass transition temperature: the DSC822/400 differential scanning calorimeter produced by Switzerland Mettler-Toledo was used to test the glass transition temperature of the unvulcanized rubber. The mass of the vulcanized sample was in the range of 6-8 mg, and then put in aluminum positioning crucible. Temperature was increased from − 75 to 25 °C at a heating rate of 5 °C/min in N 2 atmosphere. www.nature.com/scientificreports/ Dynamic mechanical properties: The DMA242C/1/G dynamic thermomechanical performance analyzer produced by the German Netzsch co. was used to test the storage modulus (E′) and loss angle (tanδ) of the vulcanized sample. Temperature rise from − 120 to 20 °C at 5 °C/min heating rate, and frequency is 1 Hz. The sample was a rectangular slice with dimensions of 6 mm × 2 mm × 1 mm. Test 3-5 times for each parallel sample, and take the curve with the best reproducibility as the result.

Conclusions
There are two factors that influence the heat generation of NR by protein. Firstly, protein tuning network structure of NR from vulcanization kinetics. It promotes the construction of cross-linked networks, reduce the flexibility of molecular chains, and limit the behavior of friction between molecular chains. Secondly, combined with the analysis of the tube model, it is shown that with the removal of the protein, the number of entanglement-like points in the rubber network decreases significantly, resulting in a weakening of the "pinning" effect on the movement of the molecular chain, increasing the heat generation.