Mechanochemical processing of polytetrafluoroethylene with NaCl crystals

Abstract

Polytetrafluoroethylene (PTFE) is an indispensable material for society and industry. PTFE chains have a strong intermolecular force and form strongly bound crystal structures. The strong assembly causes a low surface energy and refractive index, as well as difficulty in processing and recycling. In this paper, we describe how PTFE powders were milled with NaCl crystals using a planetary ball mill. Furthermore, the milled PTFE-NaCl powders were immersed in water to remove NaCl, and filtered powders were obtained. These powders were characterized by X-ray diffraction (XRD) and infrared spectroscopy. The PTFE-NaCl powders had a lower XRD peak intensity than that of processed PTFE powders without NaCl. Even the filtered powders had a lower XRD peak intensity. These results suggest that the intermolecular assembly was loosened by the mechanochemical process with NaCl. Then, PTFE-NaCl and filtered powders were pressed to form pellets. The pellets were successfully sintered, while PTFE powders milled without NaCl were not.

Graphical Abstract

Open in new tabDownload slide

1. Introduction

Intermolecular interactions between per- and polyfluoroalkyl (R_f) compounds have never been clearly explained.¹ The interactions must be explained with reference to the forces operating in other organic compounds, such as electrostatic attraction/repulsion, hydrogen bonding, and van der Waals (VdW) forces.² The attempts to explain R_f compounds with London's dispersion force (which is one of the VdW forces) were not expected to be successful because Keesom's orientation force is dominant in R_f compounds.³ Hasegawa et al.^4,5 observed that the R_f compounds with R_f chains having serial CF₂ groups longer than 8 generate the hexagonal packing structure in Langmuir films, and infrared spectroscopy showed the strong intermolecular CF₂-CF₂ orientation. The physicochemical view based on the orientation force reveals the specific characteristics of fluoropolymers, such as hydrophobicity-oleophobicity and low refractive index.

Polytetrafluoroethylene (PTFE) is one of the most common fluoropolymers.⁶ In PTFE solids, a hexagonal crystal structure is commonly observed.⁷ The interchain interactions (intramolecular in addition to intermolecular interactions) of PTFE must be very strong. However, this leads to mechanical recycling difficulties.^8,9 Améduri and Hori⁸ reviewed reports on the mechanical recycling of PTFE, where the previous mechanical treatments of PTFE resulted in the deterioration of the mechanical properties or in a reduction of molecular weight. Figure 1 summarizes the conventional and proposed ways of recycling PTEF. In conventional discussions of PTFE recycling, the interchain interaction mechanism has rarely been considered. While mechanical recycling is preferred from the energy consumption viewpoint, chemical recycling processes such as thermolysis have long been the main topic in the field (the blue line in Fig. 1).

Here we point out that interchain interaction control is a key technology in PTFE mechanical recycling. The red lines and squares indicate the expected mechanical recycling process under interchain interaction control. As mentioned, the interaction is very strong, and aqueous and ordinary organic solvents hardly affect the interaction. Fluorocarbon solvents can exceptionally be used for that purpose, but their use in industrial processes is not recommended from the viewpoint of environmental impact and is unrealistic owing to their high cost.

Mechanochemical processes use ball mills or other milling equipment to apply high energy to powders, which not only reduces the particle diameter, but also changes the crystal structure and cause chemical reactions.¹⁰ Thus, we conceive a dry mechanochemical process of PTFE with NaCl powder because a NaCl crystal edge potentially works as a large dipole moment. NaCl is ubiquitous and friendly for possible recycling process. In this study, we will assess the applicability of a PTFE-NaCl mechanochemical process to loosen the interchain interaction. First, the mechanochemically treated PTFE-NaCl and its regenerated powders after removing NaCl are characterized and compared with those of the original PTFE and milled PTFE without NaCl. Next, we examine the possibility of processing by powder pressing and sintering.

2. Experimental

Sodium chloride (NaCl) was purchased from Fuji-Film Wako Ltd and used without further purification. PTFE was also purchased from Fuji-Film Wako Ltd (164-28791). The PTFE's average molecular weight is written as 5,000 to 20,000 in its safety data sheet. Pure water was obtained from Simplicity UV (Nihon Millipore).

As shown in Fig. 2, sample powders were milled using a planetary ball mill apparatus (Fritsch Germany, P-7) with a couple of zirconia pots having a diameter of 40 mm, a height of 38 mm, and a volume of 45 mL. In each pot, seven zirconia balls having a diameter of 15 mm were set with the sample powders.¹⁰ In this paper, the evolution and rotation speeds were set to 700 and 1400 rpm, respectively.

As a first experiment, we set the molar ratio of NaCl (formula weight, FW, 58.5) to CF₂ (FW 50) to be 1:1. Then, 1.38 g PTFE and 1.62 g NaCl were mixed and milled in the pot (PTFE-NaCl powders). Some of the PTFE-NaCl powder was immersed in 30 mL pure water. After stirring for 5 min, the dispersion was filtered, and the powders were dried in a 50 °C oven overnight. As a reference to PTFE-NaCl, 3.00 g of PTFE was milled in the pot (milled PTFE powders).

The sample powders were analyzed by X-ray diffraction (XRD; Mini Flex 600/cx, Rigaku, Cu-K_α1, a wavelength $λ=1.5418$ Å, 40 kV, 10.0 °/min scan), a scanning electron microscope (SEM; S5500, Hitachi) and FT-IR (FT/IR-4700, JASCO, with an attenuated total reflection [ATR] attachment).

The raw PTFE, milled PTFE, milled PTFE-NaCl, and filtered PTFE powders were pressed for 5 min in a hydraulic press. Mold release agent JIP637 Fluorine Type R (Ichinen Chemicals Co., Ltd) was coated on the lower rod. Initially, the pressure was set to 20 MPa, and it gradually decreased during compressing. Here, the pressure during the compression was not controlled and it typically decreased down to 17 MPa. The 2-hour-milled powders were used for the pressing experiments. After pressing, the pellets were removed from the base. In the case of the milled PTFE-NaCl pellet, it was immersed in pure water overnight to remove NaCl from the pellet.

After the pressing, the pellets were sintered on an aluminum plate in the air. The temperature was raised from 22  to 370 °C for 360 min and held for 10 min. The temperature was then lowered to 320 °C, held for 50 min, and then lowered to 22 °C for 300 min. The temporal temperature program is shown in Supplementary Fig. S1 (see Supplementary data).

3. Results and discussion

Figure 3 shows the photographs of the original PTFE powder after 8 hours of milling, the PTFE + NaCl powder, and the powder after water immersion followed by filtration (referred to as “filtered” in this paper). The lower panel of the figure shows the powder XRD patterns of these samples. The XRD pattern of the 8-hour-milled PTFE is similar to that reported in the literature.¹¹ The main (100) peak is dominant in intensity, and minor peaks such as (110), (200), (210), and (300) are also observed. Broad peaks are observed around $2θ=16∘$ and $40∘$⁠. In the pattern of the 8-hour-milled PTFE + NaCl, the main peak intensity decreases significantly, and the peaks corresponding to NaCl crystal appear instead. The decrease of the main peak suggests a low crystallinity and much amorphous content. In the pattern of the filtered sample, an intensity decrease of NaCl peaks is observed, which proves the removal of NaCl from PTFE-NaCl by immersion in water. The main peak intensity, on the other hand, redevelops significantly, but the intensity does not reach that of the original PTFE. This suggests less crystallinity and much amorphous content compared with PTFE.

To estimate the crystallinity and amorphous contents, The main sharp peak at $2θ∼18∘$ and the broad amorphous peak in the lower $2θ∼16∘$ area were fitted by two Lorentzian functions in the range $12∘≤2θ≤20∘$⁠. The (100) peak and amorphous peak areas are respectively denoted as $IC$ and $IA$⁠. The amorphous content percentage (%A) was calculated based on Ryland's method^11–14 as

Figure 4a shows the $%A$ dependence on the milling time. As is clearly seen in the milled PTFE plot, $%A$ decreases in the early stage of the milling process. The filtered sample and PTFE + NaCl have a higher $%A$⁠. These results suggest that PTFE's interchain packing is loosened by the milling with NaCl, and is partially recovered by removing NaCl, which supports our concept of the interchain interaction control.

The crystal sizes were estimated from the main PTFE peaks with the Debye–Scherrer equation:

where K is the Scherrer constant (0.89) and B is the full width of half maximum (FWHM). Figure 4b shows the average crystal size in the powder particles. During the milling process, the PTFE crystal size is mostly maintained after the initial decrease. In contrast, that of PTFE-NaCl decreases with the milling time, and that of the filtered sample remaines after removing NaCl, while $%A$ does not evolve with time. The milling might affect the polycrystal part and the refinement proceeds. NaCl crystal size in the PTFE-NaCl powder has a similar trend (see Supplementary Fig. S2 in the Supplementary data).

To clarify the distinct effect of the milling, we measured ATR-IR spectra of the powders. The IR spectra of PTFE has been well investigated,^15,16 and here we focus on the infrared absorption peaks around 630 cm⁻¹. As shown in Fig. 5a, there are twin peaks. The ratio $A625/A638$ corresponds to the density of helix reversal-defect.¹⁷ Immediately after the start of milling, the defect increases in all the sample powders, and no significant difference was observed between the samples. The timescale of the response is similar to that of $%A$ (Fig. 4a). This suggests that the mechanical energy of the milling induces structural deformation and results in crystallization in the case of PTFE and in de-crystallization in the case of PTFE-NaCl. Further discussion on IR spectra will be presented elsewhere.¹⁸

Figure 6 shows the SEM images of the raw and 8-hour-milled powders. The milling of PTFE powder results in the coagulation of powders, and the milled PTFE surface seems smooth. The milling of PTFE + NaCl results in the mixing of PTFE and NaCl, and the PTFE size seems to be kept unchanged. The filtered sample has an identical chemical composition to that of PTFE but has a different microscopic morphology. Thin string-like structures are observed. Although the molding characteristics should be elucidated in detail, the microscopic morphology of the filtered sample is similar to those of industrial fine or molding powders. This suggests that the milling of PTFE with NaCl unbinds the strong interchain assembly, and the structure is inherited in the filter sample.

Through the demonstrations detailed in the above sections, the unbinding of interchain assembly is readily proved, as expected. The result suggests that the molding ability of PTFE is recovered by the milling with the aid of NaCl. To investigate this point, the raw and milled PTFE powders were pressed to form pellets. Figure 7a-d shows photographs of raw PTFE (Fig. 7a), 2-hour-milled PTFE (Fig. 7b), 2-hour-milled PTFE-NaCl (Fig. 7c), and filtered PTFE obtained from 2-hour-milled PTFE-NaCl (Fig. 7d) powders. As clearly seen in Fig. 7a, c and d, pellets were formed successfully. In the case of milled PTFE (Fig. 7b), a stable pellet was not formed, and the pressed sample was fragile when it was released from the press base.

The water contact angles of the pellet surfaces of the 2-hour-milled PTFE-NaCl after water immersion (Fig. 7c) and of the filtered PTFE (Fig. 7d) were 131.3°±1.5° and 119.6°±1.9°, respectively. The photographs are shown in Supplementary Fig. S3 (see Supplementary data). These are significantly higher than the 106° reported for the flat PTFE surface.¹⁹ Both the surfaces showed roughness after pressing. Especially for the pellet in Fig. 7c, there were pores owing to NaCl removal. The surface roughness could lead to higher contact angles.

These samples were sintered with the temperature program explained in the experimental section and schematically shown in Supplementary Fig. S1. The results are shown in Fig. 7e-f.

The pellet of raw PTFE (Fig. 7e) melted and spread, and the sintering was unsuccessful. The fragile sample of the milled PTFE (Fig. 7f) also melted and spread. These results suggest that the interparticle binding was not sufficient to form a stable polymer solid by sintering. The results are a good indication of the reason why it is difficult to apply “mechanical recycling” to PTFE.

In contrast, pellets were successfully sintered in the case of PTFE after pressing PTFE-NaCl and washing with water (Fig. 7g) and in the case of filtered PTFE (Fig. 7h). The results suggest that interparticle binding was readily formed during the powder pressing deformation process. The main significance of the results is that the mechanochemical treatment of PTFE-NaCl can open up the possibility of PTFE mechanical recycling, as the concept is shown in Fig. 1.

The pellets’ diameter shrank to 9.2 mm (Fig. 7g) and 11.2 mm (Fig. 7h) from 13.0 mm, the diameter before sintering. In this study, the initial pressure was set at 20 MPa and decreased during the 5-min compression. The pressure and compression time were not optimized, and the porosities of the pellets may vary. In addition, we should mention the brown color in Fig. 7g and 7h. At the present stage, we have not ascertained why it was colored. These points should be clarified in future process studies.

The water contact angles of the top surfaces of the sintered pellets shown in Fig. 7g and Fig. 7h were 119.3°±4.4° and 113.8°±3.5°, respectively. The photographs are shown in Supplementary Fig. S4 (see Supplementary data). They are slightly higher than the 106° reported for the flat PTFE surface.¹⁹ The pellets shrank and became dense, but they still had roughness on the top surfaces contacting air during sintering.

Here, the mechanochemical treatment of PTFE with NaCl is reviewed to consider its mechanisms. The mechanochemical treatment might induce reversal-defects in PTFE helical chains (Fig. 5). This phenomenon appears to occur regardless of the coexistence of NaCl. In the case of PTFE milling without NaCl, the amorphous ratio decreased (Fig. 4a), and the Debye–Sherrer crystal size slightly decreased (Fig. 4b). Despite the defect introduction, micrometer-sized powder particles seemed to coalesce to form flat surface aggregates (Fig. 6). These results suggest that the interchain assembly possibly restructure by mechanochemical energy, but that this results in an increase in the macroscopic powder diameter, giving a surface inactive to molding. In the case of PTFE milling with NaCl, the NaCl crystal may potentially have a large dipole moment and might interact with the defect to prevent PTFE chain coagulation. Related effects are amorphous formation (Fig. 4a) and Deby–Scherrer crystal refinement (Fig. 4b). The powder particles are modified with NaCl, and their surfaces become active the molding process.

4. Conclusion

In this study, we proposed the concept of the mechanical recycling of PTFE by interchain interaction control and assessed the applicability of a PTFE-NaCl mechanochemical process to control the interchain interaction. The fundamental characterization suggested that mechanochemical treatment with NaCl effectively weakens the interchain binding. The milled powders were found to be suitable for press-molding and sintering, while raw PTFE and milled PTFE were not. These results can be proof-of-concept, although many points require further investigation, such as mechanochemical energy optimization, NaCl ratio optimization, NaCl removal process optimization, and so on. These process parameters may depend on each other, and we need to discuss their relationships. Physical chemistry studies on the reasons for the effectiveness of NaCl, which may be related to the quadrupole nature of its crystal, are also needed to design the right process in the future.

Extending the concept is also important. It can be applied to other widely used fluoropolymers such as perfluoroalkanes (PFA), polyvinylidene difluoride (PVDF), and Nafion. For the sustainable use of fluoropolymers, we expect the industrialization of the mechanical recycling of fluoropolymers.

Acknowledgments

The authors thank C. Watanabe of Institute of Multidisciplinary Research for Advanced Materials, Tohoku University; S. Kuboyama of PAIMS, Institute of Science Tokyo (Science Tokyo); R. Kikuchi of OFC, Science Tokyo; and Y. Watanabe of JASCO Corp. for technical assistance. We also thank Professor M. Kawano, Dr Y. Wada, A. Kashlakov, and T. Umeyama of the Department of Chemistry, Science Tokyo for their experimental and scientific support. AH thanks Dr A. Watanabe and Dr C. Watanabe of Frontier Laboratories Ltd and Professor N. Teramae of Tohoku University for valuable discussions on polymer science.

Supplementary data

Supplementary material is available at Bulletin of the Chemical Society of Japan online.

Funding

This work was supported by the Asahi Glass Foundation.

Data availability

The temperature program for the pellet sintering, NaCl crystal size data, and the contact angle photographs are available in Supplementary Materials.

References

Akihide Hibara

Akihide Hibara received his PhD from The University of Tokyo. He was appointed Research Associate at the University in 1999 and promoted to Lecturer in 2003 and to Associate Professor in 2007. He moved to the Tokyo Institute of Technology in 2013 as Associate Professor. He was appointed Professor at Tohoku University in 2016 and moved to Tokyo Institute of Technology in 2023. Tokyo Institute of Technology merged with Tokyo Medical and Dental University to become the Institute of Science Tokyo in 2024. He was the president of the Society for Chemistry and Micro-Nano Systems from 2021 to 2023.

Hasegawa Takeshi

Takeshi Hasegawa received his PhD from Kyoto University. He was appointed Assistant Professor at Kobe Pharmaceutical University in 1993 and Lecturer in 2001. After serving as Associate Professor at College of Industrial Technology, Nihon University (2003), Researcher at JST PRESTO (2004), and Associate Professor at Tokyo Institute of Technology (2006), he joined ICR, Kyoto University as Full Professor in 2011. He is a Fellow of the Society for Applied Spectroscopy (USA), following recognition of his original spectroscopic technique of “pMAIRS.” Fluorine chemistry on stratified dipole-arrays (SDA) theory is another topic of interest for him.

Junya Kano

Junya Kano is a professor at Tohoku University and currently deputy director of the Institute of Multidisciplinary Research for Advanced Materials (IMRAM). He received his doctorate in engineering from Doshisha University in 1997 and worked as a research associate, lecturer, and associate professor at Tohoku University before becoming a professor in 2012. He specializes in powder engineering and is particularly interested in establishing a field of environmental powder engineering that will contribute to global environmental conservation. He focuses on energy saving, resource saving, energy creation, and resource recycling using powder processing.

Author notes