Synergistic effect of proton irradiation and strain on the mechanical properties of polyimide fibers

The damage behaviors of polyimide fiber after 150 keV proton irradiation and the synergistic effect of proton irradiation and strain were investigated. Changes in the mechanical properties, free radicals, element content, and element chemical state of the polyimide fiber before and after 150 keV proton irradiation were investigated. The results showed that the tensile strength and elongation at break of the material decreased significantly after proton irradiation. The synergistic effect of proton irradiation and strain weakened the reduction of mechanical properties caused by single proton irradiation. After proton irradiation and the combination of proton irradiation and strain, pyrolytic carbon free radicals were generated. According to XPS analysis, the proton-irradiated polyimide fiber underwent complex denitrification and deoxygenation reactions, and carbon enrichment appeared on the surface of the material.


Introduction
Polyimide ber has high strength and modulus and high radiation resistance, 1 which enable it to be used as lightweight cable jackets for spacecra and rockets as well as in ber-reinforced composites for space applications. As materials used in space may be subjected to substantial quantities of high-energy radiation, it is important that the response of the polyimide ber to high-energy radiation should be evaluated.
A large number of space experiments with polyimide lms have been implemented over a period of several years. The radiation sensitivity of Kapton to 3 MeV proton irradiation was investigated and the results revealed that the elongation to break, break stress, and the fracture energy of polymers decreased signicantly on radiolysis. Moreover, the elongation at break was similar in magnitude to that induced by 2 MeV electron irradiation with the same dosage. 2 The electron, proton, or both combined irradiations all induced bond break and cross-linking of polyimide molecules, while proton irradiation could break the PI bond easier than electron irradiation and then led to the formation of a graphite-like structure at the surface area of the samples. 3 Proton irradiation increased the initial friction coefficient and decreased the steady friction coefficient of polyimide. 4 The wear rate of the irradiated PI decreased in the order: electron irradiation > proton irradiation > combined irradiation. 5 The proton irradiation could also control the refractive index of the polyimide. The refractive index of the uorinated polyimide increased by about 0.21%, which was comparable to the value (0.35%) of SiO 2 glass when they were irradiated by protons at a similar uence. This result provided to be a good method for fabricating a highperformance polymer-based optical waveguide. 6 Aer proton irradiation, the surface roughness of PI increased apparently. The enrichment of carbon atom on the surface due to the bond breaking and reconstruction during the irradiation resulted in the degradation of the optical properties. 7 The electron paramagnetic resonance (EPR) results indicated that the proton irradiation induced the formation of pyrolytic carbon freeradicals with a g value of 2.0025, and the population of free radicals increased with the irradiation uence. During the post storage aer irradiation, the free-radical population decreased following the sum of exponential and linear mode with the storage time. In the meantime, the recovery of the degraded optical absorbance of the polyimide followed a similar mode to that of free radicals. [8][9][10] Compared with the case of polyimide, the free radical population of a SiO 2 /PI composite was much higher and increased faster with the irradiation uence and the nano-SiO 2 lm thickness. 11 However for the polyimide ber, there were signicant differences in the performance due to the high degree of crystallization or orientation. 12-14 Wu Dezhen conducted a detailed study on the performance changes of high-strength and highmodulus polyimide bers in high temperature, strong acid, strong alkali, and ultraviolet irradiation environments, and compared them with P84. Aer 150 h UV radiation, the mechanical properties of the ber remained above 90%, which was better than the resistance of P84. 15 Zhao Yong investigated the atomic oxygen resistance of a phosphorus-containing PI ber. The results indicated that the phosphorus-containing PI bers exhibited a denser surface morphology and lower weight loss compared with that of the pure PI bers. 16,17 However, under the irradiation of space-charged particles, the property evolution of the high-performance polyimide ber and the corresponding damage mechanism have not been reported. At the same time, the polyimide ber is usually subjected to stress/strain in the exible mechanism applications in space. This study investigated the effect of proton irradiation on the mechanical properties, free radical evolution, and element content change of the polyimide ber. On the other hand, the synergistic effect of proton irradiation and strain was also researched to simulate the environmental conditions in practical applications. 18 This research could provide a certain experimental basis for the application of polyimide ber in space.

Materials
In this study, the polyimide ber was kindly supplied by the Changchun Institute of Applied Chemistry, Chinese Academy of Sciences. The diameter of the ber was 16 AE 1.7 mm. For the convenience of testing and characterization, the long ber was cut into 50 mm lengths. The molecular structure is shown in Fig. 1. Its synthetic monomers comprised: BPDA (3,3,4,4biphenyl dianhydride), PMDA (1,2,4,5-pyromellitic dianhydride), and ODA (4,4 0 -diaminodiphenyl ether).

Test equipment and irradiation parameters
In this study, the selected proton irradiation energy was 150 keV. The equipment included a integrated irradiation simulator and space positive ion accelerator at the National Key Laboratory of Space Environment Materials Behavior and Evaluation Technology of Harbin Institute of Technology. The specic irradiation test parameters are shown in Table 1.
In order to study the synergistic effect of proton irradiation and strain on the mechanical properties of the bers, we carried out a combination of proton irradiation and strain experiments by selecting two uences of 7 Â 10 15 cm À2 and 1 Â 10 16 cm À2 under the same experimental conditions as 150 keV proton irradiation. This experiment was realized by a self-made xture with constant strain, of which the schematic diagram and physical diagram are shown in Fig. 2. The sample was clamped by the clip and xed by screws. The lower module could be kept in a stationary state, while the upper module could be raised under the action of the middle lead screw, so the sample could be elongated following the upper module. The strain of the sample could be calculated from the original and the nal distance of the two modules. Meanwhile, the constant strain of the sample could be ensured by the self-locking function of the lead screw, which does not slide downward during the upward movement and aer stopping the movement. According to the stress-strain curves of the polyimide ber before and aer proton irradiation (shown in Fig. 3), the strain applied to the sample was determined to be 5% AE 0.5%. At the same time, in order to eliminate the inuence of a single strain factor on the mechanical properties of the polyimide ber, a test with only 5% AE 0.5% strain and no irradiation was prepared as the control under the same vacuum conditions with 4 h.

Tensile measurement
The tensile tests before and aer irradiation were carried out in accordance with the ASTM D3379-75 standard, which involved using an XQ-1 ber strength meter at a tensile rate of 20 mm min À1 . Here, 10 to 20 ber monolaments were tested for each sample to obtain an average value.

Electron paramagnetic resonance analysis
The electronic paramagnetic resonance analysis was carried out with A200 electronic paramagnetic resonance spectrometer equipment produced by Bruker, Germany. A Q-band microwave source was used in this experiment. The detailed parameters used were a center magnetic eld of 3514.00 G with a sweep width of 70.00 G, microwave frequency of 100 kHz, time constant of 40.96 s, spectral line gain of 5.02 Â 10 4 , and modulation amplitude of 1 G.

X-ray photoelectron spectroscopy
The surface element composition and chemical valence state were analyzed on a PHI5300 system X-ray photoelectron spectrometer manufactured by PHI Corporation of the United States. The X-rays were generated by MgKa (1253.6 eV) radiation, and the test power was 250 W and the voltage was 13 kV. The electron binding energies of all the peaks were corrected by the binding energy of the C 1s peak of 284.6 eV. The results were tted by the Gauss-Lorentz function.

Results and discussion
3.1 Mechanical properties of the polyimide ber before and aer irradiation 3.1.1 150 keV proton irradiation. The tensile properties of the polyimide ber aer proton irradiation are shown in Table  2. Compared with the non-irradiated sample, the tensile strength and elongation at break of the ber aer 150 keV proton irradiation declined signicantly, and the degree increased with increasing the irradiation uence. Shen Zicai et al. also studied the proton-irradiated polyimide lm and showed that the tensile strength and elongation at break of the lm decreased exponentially with the increase in the proton irradiation uence. 19,20 In this test, the tensile strength decreased from 0.79 GPa for the non-irradiated samples to 0.42 GPa, which was 46.8% lower. The elongation at break of the sample decreased from 11.82% to 6.60%, representing a decrease of 44.2%. However, the modulus of the sample only decreased slightly. For the PI lms, under similar conditions, the tensile strength and elongation at break dropped by 25%. 21 This might be due to the higher specic surface area of the ber, which generated more free radicals; thus resulting in more damage, and therefore reducing the tensile strength and elongation at break.
In order to conrm that these changes were caused by the effects of irradiation rather than random uncontrollable factors, the variance analysis (ANOVA) was used to evaluate the condence level of the 150 keV proton irradiation effect on the mechanical properties of the polyimide bers.
One-way ANOVA is a data processing method to judge whether the inuence of a certain factor on the test index is signicant. By calculating the ratio of the variance S E 2 , which reects the data uctuation caused by the test itself, and the variance S A 2 , which reects the uctuations caused by the investigated factor A, F A can be obtained as per eqn (1). F A represents the ratio of the variation between the groups and the variation within the group, and is used to determine whether the level of the A factor has an inuence on the observed value. Actually, it can be used to compare the differences between the mean squares within the group and the mean squares between groups. The probability P, which reects the factor A, has no inuence on the data uctuation and can be acquired by further calculation. By comparing the P-value with the signicance index a, one can determine whether the factor A has an inuence on the uctuation of the experimental data.

RSC Advances Paper
When analyzing the tensile property data obtained in this test, the factor A is the irradiation uence, and the P-value indicates the probability that the null hypothesis that the irradiation has no effect on the tensile properties is true. Selecting a ¼ 0.05 as the signicance level, when P is less than a, the null hypothesis is rejected with a probability greater than (1 À a) 100%; that is to say, the condence level of the effect of irradiation on the tensile properties is >95%. Thus, the condence of the effect of irradiation on the different tensile properties of the polyimide ber can be derived from the P-value.
The F statistics and P-values of the mechanical properties of the polyimide ber before and aer 150 keV proton irradiation are shown in Table 2. The results showed that the P-values of the tensile strength, modulus, and elongation at break of the sample were far less than 0.05, indicating that the proton irradiation had a signicant impact on the mechanical properties of the polyimide ber. Moreover, the P-values of tensile strength and elongation at break were both 0, indicating that the effects of proton irradiation on the tensile strength and elongation at break were much more signicant than the modulus. Therefore, proton irradiation had an effect on the tensile properties of polyimide bers, and as the irradiation uence increased, the tensile strength, modulus, and elongation at break were all reduced.
3.1.2 Synergistic effect of proton irradiation and strain. Fig. 3 shows the tensile stress-strain curves for the original polyimide ber with a single applied strain, the combination of proton irradiation and strain, and single irradiation, respectively. It can be seen from the table that compared with the original sample, a single application of 5% strain caused a certain decrease in the tensile strength and elongation at break, which were further reduced aer the combination of proton irradiation and strain, indicating that the tensile properties damage of the sample aer the combination of proton irradiation and strain were not simply caused by strain. On the other hand, compared with the single irradiated sample, the Fig. 4 The five times repeated tensile stress-strain curves for each sample in Fig. 3. tensile strength and elongation at break of the sample aer the combination of proton irradiation and strain were increased, indicating that introducing strain during irradiation will weaken the damage from the irradiation to the ber. The ve times repeated tensile stress-strain curves for each sample in Fig. 3 are shown in Fig. 4. Every curve of each sample illustrated a similar variation of tensile strength and elongation at break with proton irradiation and strain.
The statistical results are more clearly shown by the tensile parameter changes listed in Table 3. For comparison, the mechanical properties of certain samples aer single 150 keV proton irradiation in Table 2 were repeated here. It can be seen from the table that compared with the tensile strength of 0.79 GPa of the original sample, it was decreased to 0.38 and 0.42 GPa aer irradiation of 7 Â 10 15 cm À2 and 1 Â 10 16 cm À2 , respectively. While, aer strain was applied, the tensile strength was only reduced to 0.72 and 0.63 GPa, respectively. Similarly, aer single irradiation, the elongation at break was decreased from 11.82% to 5.85% and 6.60%, respectively. Aer the combination of proton irradiation and strain, the elongation at break only decreased to 11.50% and 9.98%, respectively. The same study for polyimide lms also showed a similar phenomenon.
For the polyimide lms, the combination of proton irradiation and strain increased the tensile strength while it reduced the elongation at break. 21 This matter could be explained as follows. On the one hand, in the process of the combination of proton irradiation and strain, strain increased the free volume of the molecules, which accelerated the recovery of the free radical; therefore resulting in less damage than from the single irradiation. On the other hand, the irradiation-destroyed molecular chains could recombine in the direction of stretching, which enhanced the mechanical properties of the sample due to the orientation. Because the ber is usually subjected to stress/strain in space applications, the study of the synergistic effect of proton irradiation and strain can simulate the actual space application conditions. This result indicated that slight strain in the space environment reduced the damage caused by proton irradiation, which is benecial to the application of the PI ber in space.
3.2 Evolution of free radicals of the polyimide bers aer being irradiated 3.2.1 150 keV proton irradiation. Fig. 5 shows the electron paramagnetic resonance spectrum of the 150 keV protonirradiated polyimide ber. It can be seen that the g value of all the free radicals was 2.0025, which was independent of the irradiated uences. According to the previous research, there are two types of free radical with this g value. One is a pyrolytic carbon radical 22,23 and the other is a hydrocarbon-based superoxide radical. Generally, the hydrocarbon-based superoxide radical is only generated under environments with oxygen. Therefore, the free radical here was a pyrolytic carbon radical because the test proceeded in a vacuum environment. 23 Furthermore, the free radical content increased with the increase in the irradiated uences. Those results were similar to the PI lms. However, the content of the free radical was far lower than for the PI lm under the same conditions. 8 Sun's research showed that the recombination of free radicals mainly occurs on the surface. 8,10 For the ber, its specic surface area was much larger than that of the lm, so the free radicals generated by irradiation could be recombined and balanced rapidly, therefore resulting in a lower amount of free radicals.
To clarify the recovery rate of the PI ber aer irradiation, the free radical content evolution of the samples with uences of 1 Â 10 16 cm À2 and 2 Â 10 16 cm À2 as a function of storage time were plotted and the results are shown in Fig. 6. The result showed that the free radical content decreased with storage time. The evolution of the free radical content followed a sum of the linear and exponential modes, as shown in eqn (2): where N is the experimented specic free-radical population (g À1 ); N 0 is the free-radical population measured immediately  aer irradiation (g À1 ); A is a constant related to the annealable free-radical population, induced by the free-radical recombination; s is the characteristic time constant (h), which is related to the recovery rate of the annealable free radicals induced by the recombination process; k is a time constant (h À1 ) related to the effects of the surface reaction; and N residual is the residual free-radical population (g À1 ) under the testing condition. Table 4 shows the tting values of s and k of the protonirradiated polyimide ber during aging. The s of the sample with a higher proton uence was lower, where s is the characteristic time constant of the free radical recombination, which is related to the molecular structure. In higher proton uence, the degradation of the molecular chain caused by irradiation decreased the free volume of the PI molecules, which resulted in accelerated radical recombination, that is, s decreased. The k of the sample with higher proton uence was higher, which was also an indicator of the acceleration of the radical recombination. k is the time constant related to the effects of atmosphere. As mentioned above, the recombination of the free radicals occurred on the surface. 8,10 With higher proton uence, the surface roughening caused by irradiation increased the recombination area, and therefore accelerated the radical recombination.
The free radicals reected the effect of radiation on the molecular structure of the polyimide ber. The above results indicated that proton irradiation caused a large number of free radicals inside the polyimide ber, and therefore led to the damage of its molecular structure. This further resulted in a decrease in the mechanical properties, which was consistent with the tensile test result.
3.2.2 Synergistic effect of proton irradiation and strain. Fig. 7 shows the EPR spectrum of the polyimide ber aer the combination of proton irradiation and strain. The type of free radical was still g ¼ 2.0025 aer irradiation. Thus, the free radical here should be a pyrolytic carbon free radical due to the vacuum environment.
The results also showed that the free radical content for the combination of proton irradiation and strain samples was less than that in the single irradiated sample. This might be due to the increased free volumes of the polyimide molecular chain that caused by the strain, which made the radical recovery faster. In addition, the free radical content of the sample with the combination of proton irradiation and strain and with the uence of 7 Â 10 15 cm À2 was higher than that with 1 Â 10 16 cm À2 . This might because, at the same irradiation ux, the larger the irradiation uence, the longer the time it took, coupled with a faster recovery rate of the free radical under the strain, thus resulting in a lower free radical content.
At the same time, due to the lower amount of free radicals generated aer the strain application, they couldn't even be detected for a period of time aer the end of irradiation. Therefore, the annealing kinetics test for the free radicals aer the application of strain could not be performed here.   3.3 Effect of proton radiation on the chemical structure of the polyimide ber The composition ratio of each element in the polyimide ber aer 150 keV proton irradiation was analyzed, and the values are listed in Table 5. Aer proton irradiation, the proportion of each element changed, whereby the ratio of the element C increased, while the ratio of N and O decreased. The carbonization of the surface caused a brittleness of the polyimide ber, and therefore led to reductions in the tensile strength and elongation at break, which supported the results for the mechanical properties. There were four kinds of chemical bonds to C and their corresponding ratios are listed in Table 6. The analysis showed that the C-C content ratio increased aer irradiation, while the content ratios of the other bonds decreased. Among them, the content ratio of C-N decreased the most, followed by C-O, and the least decrease was for C]O. This result showed that the exible structure, such as C-O bonds, was damaged by irradiation, while the rigid structure, such as benzene rings (including C-C bonds), was not reduced. Therefore, the effect of proton irradiation on the modulus of the PI ber was very small, which supported the results for the mechanical properties. Table 7 shows the two kinds of chemical bonds to O and their corresponding ratios. The results showed that the ratio of C]O increased, while those of C-O decreased. This indicated that the reduction of C-O was more signicant than that of C] O, which was consistent with the results for C.
There was only one chemical state, C-N, for N atoms before and aer irradiation, while the ratio of N was reduced. This indicated that a denitrication reaction occurred on the surface of the sample during proton irradiation, which caused the decrease in the content of C-N and N atoms (as shown in Table  6). The N atoms might have escaped from the surface in the form of gas or small volatile molecules. According to Sun, 24 the irradiation damage process of the molecular structure of irradiated polyimides can be obtained by XPS and EPR, as shown in Fig. 8. Under proton irradiation, the internal molecular units of polyimide bers broke and free radicals were generated. Due to the low bond energy, a large number of C-N bonds were fractured at the initial stage of proton irradiation. With the increase in proton uence, C]O and C-O bonds were also fractured, and then a large number of free radicals were formed. With the further increase in proton uence, a large number of small molecules were generated, such as CO 2 , H 2 O, and N 2 . The small molecule by-products diffused to the surface of the material, and overowed from the surface to the vacuum environment, resulting in the phenomenon of material outgassing. This result was very similar with that for PI lms. 24

Conclusions
According to the above test results and analysis, the following conclusions were obtained: (1) Proton irradiation could signicantly reduce the tensile strength and elongation at break of the polyimide ber, while slightly decreased the modulus. The synergetic effect of proton irradiation and strain weakened the reduction of the mechanical properties caused by single proton irradiation, because the strain promoted the reorganization and alignment of the molecular chains during the irradiation.
(2) Aer 150 keV proton irradiation, pyrolytic carbon free radicals were generated, and the number of free radicals increased with the increasing irradiation uence, while they gradually decreased with the storage time. The decrease in the free radicals followed a sum of an exponential and linear mode. The synergetic effect of proton irradiation and strain did not change the type of free radicals formed in the polyimide ber, while it lowered their content.
(3) According to XPS analysis, the 150 keV proton-irradiated polyimide ber underwent complex denitrication and deoxygenation reactions, and carbon enrichment appeared on the surface of the material. The proportion of the C-N content decreased the most, followed by C-O, and then C]O, which decreased the least.
The results of this study can provide certain guidances for the application of polyimide bers in space. However, further detailed research is needed on the mechanism of the above changes.

Conflicts of interest
There are no conicts to declare.