Ultrasonic activation of inert poly(tetrafluoroethylene) enables piezocatalytic generation of reactive oxygen species

Controlled generation of reactive oxygen species (ROS) is essential in biological, chemical, and environmental fields, and piezoelectric catalysis is an emerging method to generate ROS, especially in sonodynamic therapy due to its high tissue penetrability, directed orientation, and ability to trigger in situ ROS generation. However, due to the low piezoelectric coefficient, and environmental safety and chemical stability concerns of current piezoelectric ROS catalysts, novel piezoelectric materials are urgently needed. Here, we demonstrate a method to induce polarization of inert poly(tetrafluoroethylene) (PTFE) particles ( ~ 1–5 μm) into piezoelectric electrets with a mild and convenient ultrasound process. Continued ultrasonic irradiation of the PTFE electrets generates ROS including hydroxyl radicals (•OH), superoxide (•O2−) and singlet oxygen (1O2) at rates significantly faster than previously reported piezoelectric catalysts. In summary, ultrasonic activation of inert PTFE particles is a simple method to induce permanent PTFE polarization and to piezocatalytically generate aqueous ROS that is desirable in a wide-range of applications from environmental pollution control to biomedical therapy.

This paper presents interesting results on ROS generation using ultra sound activation of piezo electric materials. The authors might include some remarks on how the ROS might be distributed into the bulk solution, considering that with a PTFE membrane the ROS is produced on its surface. This is pertinent to how this might be used in, for instance, biomedical therapy; would that involve injection of PTFE particles into the site targeted for therapy?? Using ultrasound to induce piezo-electric properties in the polymer materials, presumable leads to randomly oriented "poling" of the materials, unlike traditional poling using electric field applied at elevated temperatures. A comment related to that would be good to include. The random orientation of electrets will not, of course, affect ROS production using subsequent exposure to ultra sound. The results presented, for instance in Figure 4. includes a control where only ultrasound was used but no piezo materials. That is good, but it also requires controls WITH those materials present but without ultra sound. The results shown in figure 4 for the bacteria are not clear -the photographs of the agar plates are not of high quality. Methods section: I found a lack of detail such as the geometry of the samples/containers. For instance in the bacteria experiments, the generation of ROS would have been at the membrane surface lining the beaker. How large was this beaker in diameter. We are told it was a 100 ml beaker but this provides little insight into the distance the ROS has to diffuse to react with the bacteria in suspension. Overall, the results presented would be of high interest to many researchers interested in water remediation and biomedical treatment. There would appear to be many exciting applications, including pretreatment of feed water in membrane based water treatment plants.

Reviewer #1 (Remarks to the Author)
Authors report ultrasonic generation of reactive oxygen species from poly(tetrafluoroethylene) particle and film. They have characterized the formation of hydroxyl radicals, superoxide and singlet oxygen. They have used this approach for environmental pollution control and bacterial therapy.
Although the theme of this type piezocatalysis is important and focus of several research groups, here the authors have demonstrated that this approach can be extended to thin film of poly(tetrafluoroethylene) and for bacterial therapy. This work is important and publishable with following suggested modifications.

Response:
We greatly appreciate Reviewer 1's positive evaluation and kind recommendation in regard to publishing our manuscript in Nature communications. We have addressed all of Reviewer 1's comments and concerns in detail, and the corresponding revisions have been made in the revised Manuscript and Supplementary Information. We hope that your concerns have been eased after our detailed explanations and revisions.
General issues: Comment 1.1: Authors should estimate different radicals and compared with some of the well known piezocatalysts. In addition they should estimate different radicals and compared for particle vs film.
Response 1.1: We thank the reviewer for this comment. In the PTFE system, hydroxyl radicals (•OH), superoxide (•O 2 − ) and singlet oxygen ( 1 O 2 ) were the main reactive oxygen species (ROS).
As compared with other well-known piezocatalysts from previous research studies, we find there is no difference in the type of radical species generated between PTFE and other piezocatalysts (Nat. Commun. 2020, 11, 1328Environ. Sci. Technol. 2017, 51, 6560-6569;Chemosphere 2018Chemosphere , 193, 1143Chemosphere -1148Appl. Catal. B-Environ. 2017, 219, 550-562;Appl. Catal. B-Environ. 2020, 279, 119353). The formation of ROS is a universal reaction during piezocatalysis. The previous studies did not directly quantify the ROS concentration, thus it is difficult to compare the concentration between PTFE and other inorganic piezocatalysts. However, we did compare the piezocatalytic degradation rate of various aqueous pollutants using different piezocatalysts (see Table S1).
Obviously, the results indicate that PTFE exhibited a superior piezocatalyic activity than other piezocatalysts (i.e. significantly greater catalyst normalized degradation rate constants) in regard to organic degradation and thus ROS production.

Response:
We thank reviewer 2 for the positive evaluation of our manuscript. We have revised our manuscript according to your comments and the detailed responses are given in the comments 1 to 8.

General issues:
Comment 2.1: If the PTFE has piezoelectric property, the non-centrosymmetric structure should be demonstrated using a molecule structure. Justification should be clearly explained.

Response 2.1:
We thank the reviewer for this constructive comment. The piezoelectric mechanism between PTFE electret and classical piezoelectric materials such as ZnO and BaTiO 3 is different.
The polarization formation mechanism of polar electrets is similar to traditional piezoelectric materials. Generally, polarization of polar electrets is completed by placement of a material within an electric field at room temperature or a decreasing temperature over an appropriate range. In contrast, polarization of non-polar electrets, such as PTFE, is primarily a result of space-charges, and the charging of space-charge (or surface charge) electrets is usually achieved by injecting (or depositing) charge carriers via corona discharge, electrical breakdown radiation, light, pressure, or heat (Electrets (Topics in Applied Physics) (Springer-Verlag, Berlin, Heidelberg. New York, 1980)).
In our study, ultrasonic irradiation of aqueous solutions generates at least three elements mentioned above, which individually or in combination may be the driving force behind the polarization of the non-polar PTFE electret. Generally, as the reviewer said in the comment 2.6, the ultrasonic PTFE structural deformation could create surface defects, which results in non-polar materials having permanent charges and apparent piezoelectricity. For example, the piezoelectricity of the PTFE electret could be derived from external charge injection or introduction of structural defects rather than its specific molecule structure. In summary, the polarization mechanism of traditional piezoelectric materials and non-polar electrets differs greatly (Fig. R3), and the latter tends to have a higher piezoelectric performance.
In our study, we demonstrated a method to induce initially inert PTFE into piezoelectric electrets with a mild and convenient ultrasound process. The piezoresponse of PTFE electret was characterized by PFM. Notably, the high pressure and electric fields generated by ultrasound will create PTFE structural defects (Adv. Mater. 2020, 32, 2000006). The electret charge may consist of "real" charges, such as surface-charge layers or space charges; or it may be a "true" macroscopic polarization; or most likely it may be a combination of these two polarization processes. Specifically, when subject to an ultrasonic field, the large and transient compressive and tensile stresses on the PTFE will result in massive structural deformation that will likely create defects or even a microporous structure. Subsequently, the charges and electric fields generated during acoustic cavitation could be injected into these structural/surface defects to ultimately yield a piezoelectrically active PTFE electret (Fig. R3). Although the initial PTFE as a whole is centrosymmetric, localized regions near the chemical or physical defects may cause PTFE to be in a non-centrosymmetric phase similar to conventional piezoelectric polymers (e.g., polyvinylidenefluoride) (Cryst. Res. Technol. 1991, 26(6), 767-781).
However, we did not observe release of fluoride ions to the solution within the detection range after the PTFE was treated by ultrasound. Since the average PTFE particle size was 5 μm, it is difficult to investigate the structural changes on the atomic scale. Overall, as opposed to the natural noncentrosymmetric nature of ZnO, the piezoelectricity behavior in the PTFE electret is from deformation nonlinearity and the presence of the defect-related charges (Soft Matter 2019, 15, 262; Jpn. J. Appl. Phys. 1974, 13, 197).

Action:
In the revised version of the manuscript, we added discussion on the mechanism of PTFE exhibiting piezoelectric property under ultrasound irradiation in more detail on Page 6 as shown in following, and cited the relevant references in their proper places. not be expected for classical inorganic piezocatalysts due to their rigid structure). We agree that in the original manuscript we did not explain this mechanism in enough detail.
Please refer to response 2.1 to comment 2.1 for a more detailed explain.
Action: To avoid misleading the reader, we have deleted the corresponding sentences and relevant references ("….considering the abundance of C-F bonds in PTFE and the strong dipole of the C-F bond, the extremely high pressures and electric fields generated during ultrasonication may quasipermanently polarize localized regions of the PTFE creating a piezoelectric electret material….").
These sentences were replaced with the following explanation to assist the reader in better understanding the formation and piezoelectric properties of the PTFE electret.  Phys. 2014, 116, 066820;Adv. Mater. 2020, 32, 2000006). PTFE and other electrets may not have presented a 180° out-of-phase piezoresponse due to the following reasons: 1) The surface height can have influence on phase contrast, 2) The signal value of the measured phase contrast is a spatial average, and 3) The ultrasonic driving voltage is not large enough to achieve 180° out-of-phase inversion.

Comment 2.5:
If a parallel electric field can polarize the PTFE, its P-E curve should be presented to identify the ferroelectric properties. Does PTFE belong to piezoelectric material or ferroelectric?
Response 2.5: Many thanks for this comments. All insulating materials exhibit electrostriction and the so-called Maxwell stress effect whereby the application of electric field can deform the material (Phys. Rev. E 2014, 90, 12603), which explains why an electric field can polarize the PTFE. As the reviewer suggested, the P-E curve was measured to characterize the ferroelectric properties, if any, of the PTFE electrets produced in this study. As shown in Fig. R4, the hysteresis loop displays no saturation, thus we cannot tell whether there exist polarization switching in PTFE films because some lossy dielectric materials will display hysteresis loops similar to Fig. R4. The lack of spontaneous polarization at zero potential suggests the material is not ferroelectric. As the reviewer suggested, the PTFE aging behavior was investigated using PE curve and PFM measurements as a function of time. The result of PE loop indicates that there is no obvious change in the PTFE properties after 10 days (Fig. R5a). As the PTFE treated by ultrasound is electret material, the PE curve is similar to a dielectric material and shows no saturation typical of a ferroelectric material. In addition, since the PTFE is irradiated by ultrasound in aqueous solution before characterization, the metastable trapped charges and dipoles are reduced because they can react with water or oxygen to generate ROS. Therefore, the PTFE PE curve of displays the minimal residual polarization at the macroscopic level. Compared to Fig. 2c, the PFM results display that the localized PTFE electret polarization magnitude has decreased after 10 days (Fig. R5b), and the reduction of localized microscopic surface charges or polarized dipoles with time indicates they are metastable in nature (Fig. R5c).

Fig. R5 P-E hysteresis loops (a) and PFM (b) of polarized PTFE film after 10 days
Comment 2.8: The different pressures and electric fields should be evaluated during the ultrasonic process to compare the following parameters: ROS, degradation performance, PE curve, and piezoresponse.
Response 2.8: Thank you for the insightful comment. Actually, it is quite difficult to precisely adjust the pressure and electric field change during ultrasonic irradiation generated acoustic cavitation since the process is so transient (microseconds) and extreme (100s MPa, 100s kV/m, 1000s K, etc.). On the other hand, the ultrasound power can be easily adjusted to control the acoustic pressure field. The magnitude of the acoustic pressure waves will increase with increasing ultrasonic power and if a PTFE particle is interacting directly with a stable cavitation bubble, then as power is increased and the bubble increases in diameter, the PTFE particle may undergo a larger tensile and compressive interactions. At the suggestion of the reviewer, here we further investigated the influence of ultrasound power on ROS, organic degradation performance, PE curve, and PFM.
The ESR signals intensity for DMPO-•OH were enhanced when the ultrasound power increased from 0.5 to 2 W/cm 2 as shown in Fig. R6a below. Accordingly, the MO removal percentage increased from 8.3 ± 3.3% to 41.8 ± 1.9% (Fig. R6b). According to the equation q = d 33 T, where q and T are piezoelectric charges and the external stress, an increased acoustic amplitude 13 / 20 will increase stress T, which will induce more piezoelectric PTFE surface charges, resulting in a higher ROS generation and MO removal. However, the results of the PTFE PFM showed that increasing ultrasound power could not proportionally improve the localized polarization strength (Fig. R6c). It is probably that a low ultrasound power could sufficiently activate PTFE, but the piezocatalytic activities of activated PTFE are related to the applied pressure. In addition, the PE curve shows no obvious distinctions under different ultrasound powers, which was expected since the PTFE electret is not a ferroelectric material and the PE curve shows no residual polarization charges (Fig. R7). The relationship between ultrasound power and ROS generation, piezocatalytic activity and PTFE PFM was also investigated. The ESR signals intensity for DMPO-•OH were enhanced when the ultrasound power was increased from 0.5 to 2 W/cm 2 as shown in Supplementary Fig. 6a.
This paper presents interesting results on ROS generation using ultrasound activation of piezo electric materials.

Response:
We are grateful for this kind assessment. For real application in vivo, the PTFE nanoparticles should be considered as a nanocatalytic medicine that is injected intravenously into organisms for sonodynamic therapy. Nanoparticles, through the enhanced permeability and retention effect, preferentially accumulate in disease sites (Annu. Rev. Med. 2012, 63, 185-198). Then, the PTFE nanoparticles will catalyze ROS generation in situ for tumor eradication under the ultrasonic irradiation. Direct intratumoral injection of PTFE nanoparticles is also an effective way to target cancerous tissue and has the advantage of increasing local therapeutic concentration (Future Med. Chem. 2015, 7(12), 1503-1510. In addition to cancer therapy, sonodynamic therapy can also be applied to in vitro therapies, such as antibacterial therapy and anti-inflammatory therapy (Adv. Mater. 2019, 1901778). In this case, an activated PTFE membrane can be directly attached to the target tissue and initiate oxidative damage of pathogenic bacteria by ultrasonic irradiation.

Comment 3.2:
Using ultrasound to induce piezo-electric properties in the polymer materials, presumable leads to randomly oriented "poling" of the materials, unlike traditional poling using electric field applied at elevated temperatures. A comment related to that would be good to include.
The random orientation of electrets will not, of course, affect ROS production using subsequent exposure to ultrasound.

Response:
We thank the reviewer for this comment. In this manuscript, we demonstrated that ultrasound induced inert PTFE transformation into a piezoelectric electret. Meanwhile, we fully agree with the reviewer that piezoelectric PTFE electret will exhibit randomly oriented polarization after ultrasound irradiation. The PFM characterization in Fig. 2 confirms this interpretation. It is worth noting that the piezoelectric PTFE electret charges may consist of surface charges, and space charges. Although these charges are randomly oriented macroscopically, they can have a localized orientation on the microscopic scale (see PFM) and will inject charges into the local aqueous environment under ultrasonic irradiation. Therefore, as stated by the reviewer, the macroscopic random orientation of PTFE electrets will not affect ROS production upon exposure to ultrasound.
The piezoelectric mechanism between PTFE electret and classical piezoelectric materials such as ZnO and BaTiO 3 is different. The piezoelectricity of ZnO originates from its noncentrosymmetric nature (i.e. the ZnO crystals do not having a center of symmetry in their structure), which results in permanent electric dipoles within the material (Appl. Catal. B: Environ. 2019, 241, 256-269;Nano Today 2010, 5, 540-552). The polarization formation mechanism of polar electrets is similar to traditional piezoelectric materials. Generally, polarization of polar electrets is completed by placement of a material within an electric field at room temperature or a decreasing temperature over an appropriate range. In contrast, polarization of non-polar electrets, such as PTFE, is primarily a result of space-charges, and the charging of space-charge (or surface charge) electrets is usually achieved by injecting (or depositing) charge carriers via corona discharge, electrical breakdown radiation, light, pressure, or heat (Electrets (Topics in Applied Physics) (Springer-Verlag, Berlin, Heidelberg. New York, 1980)). In our study, ultrasonic irradiation of aqueous solutions generates at least three elements mentioned above, which individually or in combination may be the driving force behind the polarization of the non-polar PTFE electret.
Generally, the ultrasonic PTFE structural deformation could create surface defects, which results in non-polar materials having permanent charges and apparent piezoelectricity. For example, the piezoelectricity of the PTFE electret could be derived from external charge injection or introduction of structural defects rather than its specific molecule structure. In summary, the polarization mechanism of traditional piezoelectric materials and non-polar electrets differs greatly, and the latter tends to have a higher piezoelectric performance.

Action:
In the revised version of the manuscript, we added discussion on the mechanism of PTFE exhibiting piezoelectric property under ultrasound irradiation in more detail on Page 6 as shown in following, and cited the relevant references in their proper places.

Comment 3.3:
The results presented, for instance in Figure 4. includes a control where only ultrasound was used but no piezo materials. That is good, but it also requires controls WITH those materials present but without ultra sound.

Response:
We thank the reviewer for the excellent suggestion. According to the reviewer's suggestion, control experiments with piezoelectric materials in the absence of ultrasound was investigated. In the absence of ultrasound, the above-mentioned catalysts displayed negligible MO removal, indicating the piezocatalytic reaction required ultrasonic stimulation (Supplementary Fig. 5).

Comment 3.4:
The results shown in figure 4 for the bacteria are not clear -the photographs of the agar plates are not of high quality.

Response: Thanks for your suggestion.
Action: Higher quality photographs of the agar plates have been added to the revised manuscript in place of the previous low quality photographs.
Comment 3.5: Methods section: I found a lack of detail such as the geometry of the samples/containers. For instance in the bacteria experiments, the generation of ROS would have been at the membrane surface lining the beaker. How large was this beaker in diameter. We are told it was a 100 ml beaker but this provides little insight into the distance the ROS has to diffuse to react with the bacteria in suspension.

Response:
We thank the reviewer for this insightful comment. Of note is that the ultrasonic wave is not uniformly distributed in the ultrasound bath likely due to use of discrete piezoelectric elements.
We determined that the position of the beaker in the ultrasound cleaner slightly affected catalytic performance. A photograph of the pollutant degradation experiments and bacterial disinfection experiments is displayed below (Fig. R9). The beaker is suspended within the ultrasonic bath such that the water level of the solution in the beaker is the same as the water level in the ultrasonic cleaner. The inner diameter of the beaker is about 4.8 cm and the height of water in the ultrasonic bath is about 11 cm.
Additionally, we agree with the reviewer that the ROS are generated at the membrane surface and since ROS typically have an ultrashort lifetime (10 -6~1 0 -9 s) and diffusion distance ( , 1980)). In our study, ultrasonic irradiation of aqueous solutions generates at least three elements mentioned above, which individually or in combination may be the driving force behind the polarization of the non-polar PTFE electret. The ultrasonic PTFE structural deformation could create surface or bulk defects, which results in nonpolar materials having permanent surface or space charges and apparent piezoelectricity.

Action:
In the revised version of the manuscript, we added a description in the text of ultrasonic generation of PTFE piezo-charges on page 9 as displayed below. A schematic diagram of PTFE electret formation during ultrasound irradiation has also been added into Supplementary Information ( Figure S5).  Fig. 5). Response 2.4: We thank the reviewer for raising this question, which we have partially answered in Response 2.1. Polyethylene (PE) is also non-polar polymer. In Fig. 4a, we investigated the sonocatalytic activity of PE (see below); however, we found that PE particles only yielded a sonocatalytic MO degradation rate constant of 0.057 h −1 that was similar to ultrasound alone (0.053 h −1 ), and ~50 times lower than that of activated PTFE particles (2.81 h −1 ).

Meanwhile
Similar to PTFE, Polypropylene (PP) is another non-polar polymer that can be polarized into electret. PP electrets have been used in face masks for preventing respiratory infections, which have become popular during the COVID-19 pandemic (Fig. R3a). Here, we also investigated the catalytic degradation of MO by activated PP particles. The results indicated that the MO ultrasonic degradation with PP particles was greater as compared to ultrasound alone (Fig. R3b), but significantly lesser (6-fold) than with PTFE particles. Thus, a similar piezocatalytic effect is active in other non-polar polymers. However, as expected the degree of ultrasonic activation is polymer specific (PTFE >> PP >> PE), which is likely related to the polymer physical and chemistry properties e.g., the PTFE C-F bonds have a much greater polarity as compared to PP and PE C-H bonds. In summary, our discovery of PTFE piezocatalytic activity induced by ultrasonic irradiation here may only be the tip of the iceberg and the ultrasonic activation mechanism of other non-polar polymers should be further studied. Response 2.5: We thank the reviewer for this constructive suggestion. The non-activated PTFE as a whole is centrosymmetric, and the PTFE electret formation mechanism is primarily a result of the generation of space or surface charges during ultrasonic irradiation. Therefore, it is difficult to draw a specific polarized molecular structure change in PTFE (only C-F functionalization) similar to PVDF (half C-H and half C-F functionalization; Fig. R4).
The high pressures, temperatures, and electric fields generated by ultrasonic cavitation can cause large PTFE structural deformation to create permanent defects or voids (Adv. Mater. 2020, 32, 2000006). The air, oxygen, or water molecules in the voids can be ionized under the extreme ultrasonic irradiation conditions, and the positive and negative charges are deposited/injected at the ends of the voids or at the polymer surface (Fig. R2).

Fig. R4 Molecular dipole arrangement in PVDF.
The -phase has a trans-gauche-trans-gauche conformation (b), while the -phase has a planar zigzag conformation (a). The -phase PVDF, molecular dipoles are aligned in an opposite direction to each other, hence nonpolar, while in the phase the dipoles are aligned in such a fashion that self-cancellation does not occur; therefore the