Achieving Porous Carbon Fibers Pretreated With a Swarm of CO2 Micro-Nanobubbles


 Using petroleum pitch is touted as a sustainable method to fabricate carbon fibers, yet requires further advances in inevitable pore decrease after post-treatment. In an effort to circumvent this upper limit, renewable resources are widely used in the production of commercial carbon fibers. Here we show that dissolved micro-nanobubbles of CO2 in pretreatment may aid in the pore-growth of carbon fibers during activation. The results confirm that micro-nanobubbles increase specific surface (39.21 %) and micropore (16.44 %) areas of a sample. The chemical state of the elements revealed that there were no marked impurities. The observed behaviors can be understood by the following; 1) Partial O atoms released from dissolved micro-nanobubbles may attach to surrounding CO during activation, thereafter results in a higher mass of CO2. 2) Partial O atoms may directly interact with C from unmatured crystallites and form additional CO. We further denote optimized conditions based on the derived mechanism. This study provides a new strategy for the development of highly surface carbonaceous materials, thus possibly stimulating more research on advanced performance adsorbents or electrode materials.

Lee et al. 2020). From most reports, however, there is an inevitable decrease in surface characteristics after post-treatment due to pore-blocking or introduced surface functional moieties (Fig. 1). Despite the impressive progress, surface characteristics of pitch-based carbon bers are far from satisfying compared to advanced carbon materials. To this aim, the idea of achieving high surface characteristics during activation to endure the decrease in post-treatment has been industrialized. Using CO 2 as an activation agent seems viable in most methodology-driven works, since it is facile to control, abundant, and recyclable  Here, we present a novel method for fabricating carbon bers with the aid of CO 2 micro-nanobubbles in pretreatment. Micro-nanobubbles are gas bubbles reported to have diameters on the order of micro and nanometers. Conceivably, early introduced gas bubbles may move upward by buoyancy and gradually dissolve on immersed substrates. Their existence seems paradoxical due to the large Laplace pressure inside these spherical-cap-shaped objects. Although widely used, critical commentaries have described the questionable existence of gas bubbles for decades and named as the classic Laplace pressure bubble catastrophe theory (Alheshibri et al. 2016;Lohse et al. 2015). A series of work has recently shown that tiny particles could be detected via particle tracking or scattering in transparent water. Some scholars believe that these detected particles are mainly gas bubbles, while others classify them as solid impurities (Yasui et al. 2018;Jadhav et al. 2020). Since it may beyond the scope of this letter to end the lasting debate, we gently reconcile the history of micro-nanobubbles with carbonaceous materials. Several generating strategies of gas bubbles currently exist due to the maturity of gas bubble generation techniques (Haris et al. 2020). In this work, the mechanical agitation method was adopted to generate micro-nanobubbles. The as-prepared carbon ber exhibits hierarchically interconnected micro-and macropores with high surface characteristics. By accounting for the time-dependent feedback during fabrication, we nd that the introduction of micro-nanobubbles in pretreatment and contact time of CO 2 gas in activation is consistent with the pore growth. Combined with elemental analysis, surface characteristics also quanti es the amount of chemical deposit and veri es that none of the major impurities were found in all cases. This opens a viable hypothesis of the existence of gas bubbles that provides an impetus for future environmental chemistry research.
Experimental Materials Fig. 2 shows a schematic of the fabrication method of samples. Pitch with a softening point of 300 °C was used as a precursor (see S1 for details). Melt-spinning was conducted via spinning nozzle equipment at 300 °C. A spinneret has designed with a ratio of L D -1 = 2 (length: 0.6 mm, diameter: 0.3 mm). A ber shape of the resultant was suitable for stabilization, carbonization, and activation. Before stabilization, micro-nanobubbles were introduced to the sample using a CO 2 micro-nanobubble generator for 3 hr. A generator with 40 mm and 130 mm in diameter and height, respectively, of twin protruding bubble re ners were employed for micro-nanobubble generation. The ow of CO 2 was xed at 0.2 mL min -1 with an ambient room temperature. A sample without micro-nanobubbles was also prepared as a counterpart.
Stabilization was held in air mood (2 L min -1 ) of 290 °C for 3 hr. The stabilized samples were subsequently carbonized under N 2 gas (2 L min -1 ) and heated to 880 °C (see S2 for the optimized temperature). Once the temperature has reached the suggested temperature, activation was done by adding CO 2 gas instead of N 2 gas for 3 hr and 6 hr, respectively. The resultant samples were denoted as ACF-H and BACF-H. B refers to micro-nanobubbles and H is the activation time (hr).

Methodology
Surface characteristics of the obtained samples were determined by Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods (ASAP 2460, Micrometrics). Pore size distributions were derived from N 2 adsorption measurements. Prior to the physisorption analysis, samples were heated at 300 °C for 12 hr with a vacuum state of 1.33 × 10 −3 Pa (VacPrep 061, Micromeritics). Morphologies were analyzed using eld emission scanning electron microscopy (Carl Zeiss, SIGMA) after gold sputter coating of samples for 60 s. An operating acceleration voltage was 2.00 kV. Chemical ratios were examined through X-ray photoelectron spectroscopy (AXIS NOVA, KRATOS) and energy-dispersive X-rey spectrometer (S-4300, Hitachi).

Surface characteristics
To investigate the possible structural changes due to pretreatment and activation using various forms of CO 2 , the textural properties of samples are studied (see S3 for tabulated values). Micro-nanobubblepretreated samples have typically exhibited fascinating BET speci c surface areas as plotted in Fig. 3(a).
The BET speci c surface area was 1209 m 2 g -1 after pretreatment and 3 hr of CO 2 activation, which is 39.21 % higher than the surface area of ACF-3 (735 m 2 g -1 ). Likewise, BACF-3 showed a 16.44 % increase in micropores when compared with the bare support ( Fig. 3(b)). As the activation time using CO 2 increases from 3 hr to 6 hr, the BET speci c area of ACF-3 and BACF-3 increased 7.20 % and 13.33 %, respectively. CO 2 is acknowledged as the most stable form of oxidized carbon compounds. Nevertheless, it reacts easily to form additional C-C and/or C-H bonds under a high-energy input. Given that, the use of various CO 2 forms in both pretreatment and activation may lead to the highest surface characteristics ( Fig. 3(d)). As observed in the yield curve ( Fig. 3(a)), all samples showed the loss of their total mass due to the decomposition of unmatured polymers via a sandwich of CO 2 treatment (Yoda et al. 2018).
Although BACFs possessed higher surface characteristics than ACFs, the yield curves were 10.92 % and 11.52 % lower in 3 hr and 6 hr, respectively. The upward peaks in N 2 adsorption-desorption isotherms suggest that all samples are of Type I, which is typical for microporous materials (Fig. 3(c)). The broadening of peaks shows an increasing adsorption trend and does not atten at P/P 0 between 0.8 1 .0, which veri es the multilayer adsorption. Alongside the typically characterized adsorption force in microporous materials, the van der Waals force is given on the adsorption potential within micropores. This indicates that all samples have a large micropore region and a comparably small mesopore region. N 2 volume adsorption-desorption isotherms are also in good agreement with the inert graph in Fig. 3(b).
The chemical state of elements Introduction of micro-nanobubbles and post-treatment of CO 2 (6 hr) seems to be the key factors to achieve superior surface characteristics. However, there are underlying uncertainties whether the samples are free of any impurities. Acknowledging that impurities are more likely to be attached to the samples with high BET speci c surface, ACF-6 and BACF-6 were chosen to observe the chemical state of elements. Both samples reveal porous characteristics with no visible impurities (Figs. 4(a)-(b)). According to the average chemical proportion, C and O are considered to be major elements present in both samples (Fig.  4(c) and S4). The amount of C and O was 2.91 % and 2.93 % higher in ACF-6 and BACF-6, respectively.
Evidently, C and O ratio difference is mainly due to the side chains of partial oxygen-containing functional groups (C-O, O-C=O, and C-C) introduced on the surface. Only N can be distinguished in the category of minor components (ACF-6: 1.65 % and BACF-6: 1.63 %). This is attributed to residual nitrogen from pitch.
The wide scan spectrum displayed photoelectron lines in the order of C1s (approx. 288 eV) > O1s (530 eV) > N1s (400 eV). No characteristic peaks of any impurities were noticed, suggesting that micronanobubbles are solely the major culprit of high surface characteristics.
We merely hypothesized that dissolved micro-nanobubbles at the frontline of stabilization may chemically react with the surface and aid for pore-growth ( Fig. 2(b)). However, it is still an arduous quest to theoretically understand the microchemical behavior of dissolved micro-nanobubbles without satisfactory reasons. Inspired by the two scenarios proposed with CO 2 activation, we track O atoms under the inclusion of surrounding elements (Figs. 4(d)-(e)). On the one hand, dissolved micro-nanobubbles may release partial O atoms attached to the surface as the temperature increases. Partial O atoms could form additional CO 2 with surrounding CO during activation, resulting in a higher mass of CO 2 activation (Nabais et al. 2008). Another interesting approach lies on the partial O atoms, which are not attached with surrounding CO. Partial O atoms may directly compromise with C of crystallites and result in additional CO. The activation reaction could be promoted while the unmatured carbonaceous ber consumes CO, which results in additional pore-growth (Lan et al. 2019). In summarization, dissolved micro-nanobubbles may not generate pores by themselves, but rather act as a swarm of catalyst at the beginning of stabilization. These chemical interactions made our samples bear great application potential in highperformance adsorbents or electrode materials.

Conclusion
A novel sandwich technique of CO 2 in pretreatment and activation has been explored for achieving high surface characteristics. BET and BJH analyses determine that the introduction of micro-nanobubbles in pretreatment was 29.89 % more effective for increasing BET speci c surface compared with CO 2 gas variation from 3 hr to 6 hr in activation. SEM images indicate the formation of porous surface in the order of BACF-6 > ACF-6. The estimated chemical elements were mainly C, O, and N in all samples, whereas there were no signi cant impurities. Partial O atoms from the dissolved micro-nanobubbles are believed to be the catalytic driving force for surface modi cation. On the basis of its facile and green fabrication process without the requirement of additional chemicals, great versatility in relevant elds is expected. It further lays the groundwork for utilizing CO 2 to improve the interatomic potentials of carbonaceous materials. While this letter acts as an introduction of micro-nanobubbles for the synthesis of carbonaceous materials during pretreatment, we will experimentally track partial O atoms in forthcoming work. This research did not receive any speci c grant from funding agencies in the public, commercial, or nonfor-pro t sectors.

Con ict of interest
The authors declared no con icts of interest.