Synthesis of poly(S-r-BTT) copolymers.
Thiols can be used as efficient hydrogen atom donors and chain transfer agents and are highly reactive with elemental sulfur.19,20 Accordingly, thiol compounds are commonly used for the synthesis of polymeric polysulfides. The interactions of sulfur with thiol-based compounds have been reported in previous studies. These studies synthesized polymer solutions via a condensation reaction between the thiol monomer and sulfur in a toluene/carbon disulfide mixture.20,21,22,23 In contrast, in this study, the molten sulfur acts as a comonomer, as well as the reaction solvent itself.8
BTT has three thiol groups with high reactivity, promoted by the conjugated nature of the benzyl ring, which is very advantageous for eliminating protons in a condensation reaction with elemental sulfur.22,2324 The loss of hydrogen as H2S during inverse vulcanization via condensation reaction with elemental sulfur and BTT results in direct S-S bonding, which results in excellent optical properties due to the high sulfur content of the copolymer.
We first confirmed the reactivity between sulfur and the thiol cross-linker through an inverse vulcanization reaction using 1,3-benzenedithiol (BDT). The differential scanning calorimetry (DSC) curve for the poly(S-r-BDT) copolymer (70 wt% sulfur and 30 wt% BDT) showed no evidence of unreacted sulfur, however, the copolymer had a low glass transition temperature (Tg = − 12.38°C) and lacked shape persistence (Supplementary Fig. 9). The inverse vulcanization of sulfur and BTT formed a highly cross-linked network. (Supplementary Fig. 1). The poly(S-r-BTT) copolymers showed properties that were clearly different from sulfur copolymers based on 1,3-diisopropenylbenzene (DIB) or divinylbenzene (DVB) (Fig. 1).
Inverse vulcanized poly(S-r-BTT) copolymers were successfully prepared with a sulfur content of 80 ~ 50 wt% (see Supplementary information for details). The sulfur and BTT mixture was reacted in an oil bath at 185°C, producing vigorous H2S gas. The presence of H2S gas could be observed by the bubbling of the reaction mixture (Supplementary Fig. 2). All experimental procedures were carried out in a fume hood and required special attention by the operator. After 1 hour of reaction, the cured poly(S-r-BTT) copolymers were obtained as a bright yellow, transparent glassy material. The obtained copolymers were named S’X’-BTT’Y’ according to the sulfur and BTT content, where ‘X’ is the wt% of sulfur and ‘Y’ is the wt% of BTT (as was the case with DIB and DVB).
Fourier transform infrared (FT-IR) spectra of the poly(S-r-BTT) copolymers showed the disappearance of S-H stretching (2500–2600 cm–1 range)23 and the introduction of an aromatic ring of BTT (1566–1386 cm–1 range). In addition, new characteristic peaks of S-S bonds stretching (541–478 cm–1 range) appeared in the poly(S-r-BTT) copolymers.25,26 This suggests successful polymerization between sulfur and BTT (Fig. 2A). X-ray diffractometer (XRD) measurements of the poly(S-r-BTT) copolymers with various sulfur contents showed no evidence of crystalline of unreacted sulfur or BTT monomer. As with the FT-IR analysis, this indicated successful polymerization of the sulfur and BTT (Fig. 2B). Crystallization of the unreacted sulfur, which was not observed at sulfur contents of 80 ~ 50 wt%, was clearly observed in the S90-BTT10. Indeed, after the S90-BTT10 was stored for one day at room temperature it showed unreacted sulfur on the DSC and XRD curves. The unreacted sulfur was also confirmed visually (Supplementary Figs. 11 and 12).
Subsequently, various characterizations were carried out on the poly(S-r-BTT) copolymers with sulfur contents of 80 ~ 50 wt%.
Thermal properties of the poly(S-r-BTT) copolymers
The inverse vulcanization of BTT and elemental sulfur resulted in sulfur copolymers with improved thermal properties. The thermal stability of the poly(S-r-BTT) copolymers was enhanced as BTT content increased, as demonstrated by Thermogravimetric analysis (TGA), which indicated an increase in char yield and decomposition temperature (Fig. 2D). All of the poly(S-r-BTT) copolymers were thermally stable until about 200°C without weight loss. The S50-BTT50 with the highest BTT content showed excellent thermal stability with 5% weight loss occurring from about 369°C.
The formation of a highly cross-linked structure was also confirmed by the solvent resistance of the poly(S-r-BTT) copolymers. The S70-BTT30 was completely insoluble in organic solvents, such as dichloromethane (DCM) and tetrahydrofuran (THF) at room temperature, and maintained its original shape and weight without any decomposition reaction or swelling in THF to 50°C for 1 hour (Supplementary Fig. 15).
We also observed that the glass transition temperatures (Tg) of the poly(S-r-BTT) copolymers, determined by the 2nd DSC heating cycle, increased gradually with increasing BTT content (Tg from 14.85 to 100.14°C). As with the XRD patterns, there were no sulfur related peaks observed for sulfur contents of 80 ~ 50 wt%. The Tg values of the poly(S-r-BTT) copolymers, defined as the tan delta of dynamic mechanical analysis (DMA) (Tg from 40.83 to 118.25°C), showed a higher value that the Tg values observed with DSC. However, the Tg range showed reasonably consistent trends (Fig. 2E and 2F) and allowed for a composition with overall good processability. The thermal properties of poly(S-r-BTT) copolymers are summarized in Supplementary Table 6. In the MD simulations, the Tg values of the S-BTT models based on the various contents of sulfur and BTT were 46.15, 54.55, and 95.15°C for the S70-BTT30, S60-BTT40, and S50-BTT50 models, respectively (Supplementary Fig. 16). The Tg values predicted by the MD simulations were matched well with that of poly(S-r-BTT) copolymers observed with DSC (Fig. 2G). The Tg of the S70-BTT30 model was lower than those of other models due to the flexible polymer chain induced by the high sulfur content. The Tg of the S-BTT model gradually increased with increasing BTT content. This result was related to the density of the S-BTT model. At the same temperature, the total volume of the S50-BTT50 model decreased compared to the S70-BTT30 model (Supplementary Fig. 16). This result indicates that the polymer chains in the S50-BTT50 model were densely packed in the simulation cell. The increase in density leads to an increase in the rigidity of the polymer chains. Therefore, the S50-BTT50 model had the highest Tg due to the increase in the rigidity of the flexible polymer chains, induced by the high BTT content. The high Tg of the S50-BTT50 model leads to the improvement in thermal stability.
Optical Properties Of The Poly(S-r-btt) Copolymers
Previous researchers have advised that for more meaningful IR imaging, lenses at least 1 mm thick should be used, since even conventional vinyl cross-linker and aromatic ring based polymers exhibit some IR transparency.9,17
Bulk poly(S-r-BTT) copolymers were readily prepared into lenses with a diameter of 25 mm and a thickness of about 1 mm by hot-pressing at 185°C and 20 MPa (see Supplementary information for details). Despite the highly cross-linked structure of the poly(S-r-BTT) copolymers with high thermal properties, the presence of large quantities of dynamic S–S bonds allowed their easy and fast processing.
The processed S-BTT lenses were sufficiently transparent without any bubbles or pores (Supplementary Figs. 3 and 4). The interior of this transparent yellow lens was characterized by X-ray microscopy (XRM). The 3D image of S70-BTT30 lens showed a clean interior without pores or defects (Supplementary Fig. 5). We also prepared about 1 mm thick lenses of widely used optical polymeric materials (e.g., polyethylene (PE), cyclic olefin polymer (COP) and poly(methyl methacrylate) (PMMA)) as well as inverse vulcanization polymers (e.g., S70-DIB30 and S70-DVB30) for a relative evaluation of IR transmission (%) (Supplementary Figs. 6 and 7). A detailed description of the lenses’ fabrication can be found in the Supplementary Information. The average thickness of the polymer lenses is given in Supplementary Fig. 8 and Supplementary Table 1.
Another advantage of the high sulfur content in the poly(S-r-BTT) copolymers was revealed in the optical properties of the S-BTT lenses. One of the optical properties of sulfur polymers synthesized via the inverse vulcanization is their high refractive index (RI) due to the high sulfur content of these materials.7,27 The S-BTT lenses exhibited an ultrahigh RI with overall sulfur content (80 ~ 50 wt%). The RI values of the S-BTT lenses are summarized in Supplementary Tables 8 and 9. After inverse vulcanization, the alkyl chain moieties remained in the S70-DIB30 and S70-DVB30 copolymers. In contrast, the S-BTT copolymers consist of an aromatic ring and S-S bonds without alkyl chain moieties. Because of these structural features, the RI of the S-BTT lenses dramatically improved compared to the S70-DIB30 and S70-DVB30 lenses (Fig. 3A). These results were supported by the high sulfur content of the S-BTT, as determined by the elemental analysis described below. The RI values of the S-BTT lenses increased with increasing sulfur content. The highest RI, ranging from n = 2.00 to 1.94 in the range of 637–1549 nm was observed for the S80-BTT20 lens with the highest sulfur content (80 wt%). Moreover, S50-BTT50 lens with the lowest sulfur content (50 wt%) also retained a high RI ranging from n = 1.95 to 1.89 in the same wavelength range.
To understand this high RI values, we determined the optimized structures for the two types of model compounds based on a linear sulfur oligomeric structure (Ph-S16-Ph, M1, S content: 76.89 wt%) and a cross-linked structure ((PhS6)3Ph, M2, S content: 65.32 wt%) using DFT calculations (Supplementary Fig. 18), and calculated the theoretical RI according to the Lorentz-Lorenz equation.33 The calculation results indicated a higher RI value in M2 of the cross-linked structure, despite the low sulfur content. Basically, the high sulfur content created high molar reflection, but the results suggest that higher RI value was obtained when cross-linking occurred between the aromatic structures. For this reason, the BTT vulcanization system showed a higher RI value than the DIB system.
In addition, it was confirmed that the calculated RI values from M1 and M2, which possessed structures similar to S70-BTT30 in sulfur content, agreed relatively well with the experimental values obtained (Fig. 3B and Supplementary Table 5).
The FT-IR spectra of the polymer lenses were measured five times to ensure the accuracy and reliability of the experimental results. No significant difference was observed between the five measurement results, which shows that all of the polymer lenses were homogeneously fabricated through hot-pressing (Supplementary Fig. 23–33 and Supplementary table 10–18). We calculated the average IR transmission (Tavg) (%) of each of the mid-wave infrared (MWIR) and the long-wave infrared (LWIR) regions from the FT-IR spectra, and the results are summarized in Fig. 3C and Supplementary Table 19. The MWIR region was applied in the range of 3–5 µm (3300 − 2000 cm− 1) and the LWIR region was applied in the range of 7–14 µm (1420 − 710 cm− 1). The S80-BTT20 lens with the highest sulfur content showed the highest IR transmittance, a Tavg (MWIR) of 54.12% and Tavg (LWIR) of 32.38%. As the sulfur content decreased, the Tavg of the S-BTT lenses decreased. However, the Tavg of the S50-BTT50 lens with the lowest sulfur content was significantly higher than the S70-DIB30 and S70-DVB30 lenses (Tavg (MWIR) of 33.59% and Tavg (LWIR) of 13.93%).
This trend was also confirmed in the DFT calculations using the M1 and M2 model compounds. In both cases, absorption peaks derived from organic components other than C-S and S-S were expected in the IR region, and the molar extinction coefficient was very small, with a maximum of 300 M/cm− 1 compared to general organic materials (generally 103 or more, reaching 104 M/cm− 1). In particular, the peak position of the calculation agreed very well with the experimental absorption of the S-BTT series, which had a conjugated M1 and M2 structure. This indicated that the S-BTT structure was as expected, and at the same time had very low IR absorption due to the structural effect (Fig. 3D and Supplementary Figs. 19 and 20). Accordingly, the FT-IR spectra of the approximately 1 mm thick S-BTT lenses showed numerous windows of transparency over the entire IR region. In contrast to the S70-DIB30 and S70-DVB30 lenses, which exhibited broad absorption in the 3.3–3.5 µm region, corresponding to C–H bond stretching, the S70-BTT30 lens showed over 60% transmission windows in the MWIR region (Fig. 3E).
Surprisingly, the S70-BTT30 lens had numerous transmission windows of over 60%, even in the LWIR region. On the other hand, the LWIR transmission (%) of the previously studied S70-DIB30 and S70-DVB30 lenses of the same thickness approached zero, which means theses lenses showed virtually no LWIR transparency (Fig. 3F). The fingerprint region of the S70-BTT30 copolymer, which is simpler than those of the S70-DIB30 and S70-DVB30 copolymers, contributed to the high IR transmittance (Supplementary Figs. 36 and 37).
Wholly aliphatic organic polymers such as PE or COP were used as thin LWIR-barrier coatings due to the simple fingerprint region.28,29 We measured the IR transmission (%) of 1 mm thick transparent polymer PMMA, PE and COP lenses. Interestingly, our experimental results showed that the S-BTT lenses had excellent IR transmittance, better than the organic polymers based on aliphatic units.
The S-BTT lenses exhibited ultrahigh RI and excellent IR transmittance compared to other widely used optical polymeric materials (e.g., PE, COP and PMMA) and previously studied inverse vulcanization polymers (Supplementary Figs. 38 and 39 and Supplementary table 20). To the best of our knowledge, our study is the first example of an IR transmissive polymeric material with excellent transmittance in both the MWIR and LWIR regions (Supplementary Figs. 40 and 41). The outstanding IR transmittance quality of the S-BTT lenses was not affected by processing conditions such as temperature or time (Supplementary Fig. 42–53 and Supplementary table 21–28). Additionally, even after being stored at room temperature for one month, the lenses maintained excellent IR transmittance in both the MWIR and LWIR regions without significant degradation (Supplementary Figs. 54–63 and Supplementary Tables 29–35).
We have explained that the structure of the poly(S-r-BTT) copolymers consisted only of an aromatic ring and S-S bonds. These structural features of the poly(S-r-BTT) copolymers were positive for IR transmittance as well as RI values. Elemental analyses of the sulfur copolymers were then carried out. The results showed the poly(S-r-BTT) copolymers had relatively high sulfur (S) content and low carbon (C) and hydrogen (H) content. Furthermore, the S50-BTT50 with the highest BTT feed ratio showed lower carbon content and higher sulfur content than S70-DIB30 and S70-DVB30 (Supplementary Fig. 64 and Supplementary Table 36). These results further indicate the structural features of the poly(S-r-BTT) copolymers are composed of numerous S-S chains centered on an aromatic ring.
The high S/C atomic ratio of the poly(S-r-BTT) copolymers, calculated from the elemental analysis results, contributed to the enhanced optical properties, particularly due to the specific properties of the S-S bonds (e.g., mostly IR inactive and high assigned n) (Supplementary Fig. 65 and Supplementary Table 37). These findings confirmed the S90-BTT10 lens exhibited unprecedentedly high IR transmittance with a Tavg (MWIR) of 60.48% and Tavg (LWIR) of 33.26% (Supplementary Fig. 13 and Supplementary Tables 3 and 4), but an IR imaging experiment was not considered because of the presence of unreacted sulfur.
IR imaging experiments.
We carried out IR imaging experiments to demonstrate that e.g., poly(S-r-BTT) copolymers are sufficiently transparent for IR applications. For comparison, IR visualization was also attempted through the PMMA lens and the previously studied S70-DIB30 and S70-DVB30 lenses. MWIR and LWIR images were taken through the 1 mm thick polymer lenses previously described (Fig. 4E) and used for the IR transmittance measurements, and the IR imaging setup is shown in Fig. 4A and 4B.
For the MWIR imaging experiments of the USAF 1951 target, a thermal emission microscope with a transmission range of 3–5 µm was used. The transparent plastic PMMA failed to visualize the USAF target due to its low Tavg (MWIR) (2.53%), while the S70-DIB30 and S70-DVB30 lenses were capable of imaging in the MWIR region, with results consistent with previous studies. A decrease in the Tavg (MWIR) was observed as the sulfur contents of the poly(S-r-BTT) copolymers decreased, however, high-quality MWIR imaging was achieved for the overall sulfur content of 80 ~ 50 wt%. The S80-BTT20 lens with the highest Tavg (MWIR) (54.12%) provided the clearest image, and the S50-BTT50 lens with the lowest sulfur content (50 wt%) also provided a clear image with a Tavg (MWIR) of over 33% (Fig. 4F). The obtained MWIR images were reasonably consistent with FT-IR analysis trend.
For LWIR imaging experiments with various patterned PMMA masks, an FLIR camera with the transmission range of 7.5–13 µm was used. The S70-DIB30, S70-DVB30 and PMMA lenses, which were capable of imaging in the MWIR region, clearly showed opaque properties in the LWIR region, while the S-BTT lenses showed excellent transparency in the LWIR region. As with the MWIR imaging experiments, a decrease in Tavg (LWIR) was observed as the sulfur contents of the poly(S-r-BTT) copolymers decreased, however, high-quality LWIR imaging was achieved for overall sulfur contents of 80 ~ 50 wt%. The S80-BTT20 lens with the highest Tavg (LWIR) (32.47%) provided the clearest image. The obtained LWIR images were of a comparable quality to those obtained through a Ge lens with an expensive inorganic transmissive material. Moreover, the S70-BTT30 lens successfully captured a human finger (Fig. 4D). The S50-BTT50 lens with the Tavg (LWIR) of over 13% was also sufficiently transparent in the LWIR region (Fig. 4G). The obtained LWIR images were reasonably consistent with the FT-IR analysis trend. In an additional near-IR imaging experiment, the S70-BTT30 lens successfully captured female subjects (Fig. 4C).