RPA was purchased from Yonggang Decoction Piece Factory Co., Ltd. (Bozhou, China). The standard substances including paeoniflorin, benzoylpaeoniflorin, gallic acid, tannic acid, 2′-hydroxy-4′-mehoxyacetophenone, and D-(+)-glucose were HPLC grade and bought from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). The water used in this study was deionized and prepared by the PSDK-C system (Beijing, China).

As shown in Figure 1, the sunlight-dried RPA pieces were crushed into powder and sieved (120 mesh). Afterwards the handmade platform (7.5 × 8.0 × 0.1 cm) was filled with RPA and the DBD board was placed at 0.1 cm height from the platform. The RPA samples were treated with DBD plasma at three doses of 60 s, 120 s, and 180 s and recorded as DP-60, DP-120, and DP-180, respectively. The untreated RPA, also regarded as control, was coded as DP-0.

As explained above, the extracted content of each bioactive compound in RPA was measured according to the conventional methods (include high performance liquid chromatography and ultraviolet spectrophotometer) in Chinese Pharmacopeia 2020, with some modifications. Paeoniflorin: 0.5 g of RPA was immersed in 80 mL of methanol solution, assisted by ultrasonic extraction for 30 min, and cooled to room temperature. Then the extraction was supplemented to 100 mL with methanol. Finally, the combined solution was filtered with a 0.45-μm microporous membrane. The extraction was separated on a Shim-pack VP-ODS C18 chromatographic column (4.6 × 150 mm, 5 μm; Shimadzu, Japan). The mobile phase was acetonitrile-0.1% phosphate buffer (PBS) at a stable flow rate of 1.0 mL/min and a split ratio of 14:86. The temperature of the column was controlled at 30°C and the injection samples volume was set at 10 μL. The absorbance values were obtained at 230 nm and calculated using Chromeleon (c) Dionex software. Benzoylpaeoniflorin and 2′-hydroxy-4′-methoxyacetophenone: The extractions were prepared by the same procedures as paeoniflorin, but the mobile phase was changed to methanol-water (45:55). Accompanied by the same flow rate, column temperature, and injection volume, the contents of 2′-hydroxy-4′-methoxyacetophenone and benzoylpaeoniflorin were detected at 230 and 274 nm, respectively. Tannic acid and gallic acid: 0.1 g of RPA was immersed in 100 mL of deionized water for 30 min, shaken in a constant temperature bath at 60°C for another 30 min, and cooled to room temperature. Finally, the absorbance values of tannic acid and gallic acid were obtained at 276 and 263 nm, respectively, and the contents of the two bioactive compounds were calculated using standard curves (Y_tannicacid = 0.0398 X−0.00865, R = 0.9999; Y_gallicacid = 0.0451 X +0.00297, R = 0.9998). Polysaccharide: 0.5 g of RPA was immersed in 50 mL of deionized water, accompanied by heating reflux extraction for 2 h. The extracted solution was centrifuged at 7,000 rpm for 10 min and 2 mL of supernatant was mixed with 10 mL of absolute ethanol. After being centrifuged at 7,000 rpm for 20 min, the sediment was washed with 8 mL of 80% ethanol twice. Finally, the washed sediment was dissolved with 50 mL of hot deionized water and the absorbance value and content were obtained at 488 nm (Y_{polysaccharide} = 0.006289 X +0.0447, R = 0.9999). All assays were conducted at least three times and the values were recorded as mean ± standard deviations (std) mg/g. To evaluate the effects of DBD plasma treatment on the extraction efficiency of RPA, the percentage of content of each compound was obtained by the following equation,

where m_p and m₀ represent the content of DBD plasma-treated and untreated RPA samples, respectively.

The water and oil holding capacities of RPA were studied according to Nawaz et al., with some modifications (24). A total of 0.2 g of RPA powder was mixed with 5 mL of water or oil (Luhua peanut oil, Shandong Luhua Group Co. Ltd, Laiyang, Shandong, China), the value was calculated by the following equation after being vortexed for 1 min and left to stand in the dark for 24 h,

where m₁ and m₀ represent the weight of soaked and dried RPA samples, respectively.

The surface morphology of the four RPA samples was observed using a Scanning Electron Microscope (GeminiSEM 500, Germany), and the specific surface area and pore change were determined by a Tristar II 3020M analyzer (Micromeritics, USA). The FTIR spectra of the four RPA samples were recorded on a Fourier transform infrared spectrometer (Thermo Nicolet 8,700, USA). Each sample was blended with KBr at a 1:20 ratio and scanned from 4,000 to 400 cm⁻¹ wavenumbers. The XPS spectra were recorded on a Thermo ESCALAB250Xi spectrometer (Thermo Fisher Scientific Co. Ltd, UK) with an excitation source of monochromatized Al Kα (hv = 1486.6 eV) and a pass energy of 30 eV. The values of binding energies were calibrated with the C 1s peak of contaminant carbon at 284.80 eV. The thermostability of RPA was characterized using a thermal gravimetric analyzer (TGA Q5000iR, USA) at a heating procedure of 5°C/min from 25 to 600°C under a nitrogen atmosphere.

All assays were conducted at least three time with the values recorded as mean ± std. Significance of differences in data was determined by IBM SPSS Statistics 21 (International Business Machines Corporation, USA), and p < 0.05, p < 0.01, and p < 0.001 represented a significant, highly significant, and extremely significant difference, respectively, as compared with control (DP-0). All figures were plotted by Origin 8.5 and the XPS results were analyzed using XPSPEAK4.1 (Raymund W.M. Kwok, The Chinese University of Hong Kong, Hong Kong, China).

As shown in Table 1, the paeoniflorin content of RPA in the DP-180 group exerted a significant reduction (p < 0.05) of 2.1% as compared with the DP-0 group, while the content in the DP-60 and DP-120 groups showed no significant (p > 0.05) changes. The benzoylpaeoniflorin and 2′-hydroxy-4′-methoxyacetophenone contents of RPA in the DP-60, DP-120, and DP-180 groups all showed a non-Significantly difference (p > 0.05) as compared with the DP-0 group. In the DP-60 and DP-120 groups, tannic acid and polysaccharide contents had non-Significant (p > 0.05) changes as compared with the DP-0 group, however, the content of that in the DP-180 group presented a highly significant (p < 0.001) increase of 24.6 and 12.0%, respectively. Finally, the gallic acid contents in the DP-60 and DP-120 groups showed a highly significant (p < 0.01) reduction of 3.5 and 3.4% respectively, while the content of that in the DP-180 group exhibited a highly significant (p < 0.001) reduction of 5.4% as compared with the DP-0 group.

As shown in Figure 2, non-Significant (p > 0.05) changes were observed in the water and oil holding capacity between the DP-0 group with DP-60, DP-120, and DP-180 groups, while the water holding capacity of RPA in the DP-120 and DP-180 groups further improved as compared with the DP-0 group.

As shown in Figures 3A,B, the untreated RPA had a rough surface and various small particles. However, it can be seen in Figures 3C,D, the RPA in the DP-60 group had some shallow gaps and a smoother plane than the DP-0 group. In the Figures 3E,F, the surface morphologies of RPA in the DP-120 group was obviously scratched and had small pores, while the RPA in the DP-180 group showed intensive cracks and smaller pores (Figures 3G,H).

As can be seen in Table 2, the BET specific surface area of RPA in the DP-0 group was 0.4634 m²/g. The area of that in the DP-60 group reduced to 0.0421 m²/g, while it increased to 1.3328 m²/g in the DP-120 group. With a longer plasma treatment time, the area obtained a significant improvement of 7.7209 m²/g in the DP-180 group. As shown in Figures 4A,B, the pore diameter of RPA in the DP-0 group was located in the range of 0-2000 Å and most in 0-500 Å, while each pore occupied a lower pore volume. In the DP-60 group, the pore diameter of RPA was reduced to two ranges around 1,000 and 20 Å, while the proportion of 1000 Å particles increased and the 20 Å grains decreased. In the DP-120 and DP-180 groups, the RPA particles were further scaled into smaller size, and the particles in the DP-120 group were uniformly distributed in the range of 0–400 Å at a lower rate. Contrarily, the particles in the DP-180 group were intensively strewn in 0–150 Å and presented a higher pore volume as compared with the DP-120 group. It can be seen in Figures 4C,D, the RPA in the DP-0 group was most distributed in the range of 0–1000 Å and showed a low pore area, while the particles in the DP-60 group displayed a much lower pore area in the same range. In agreement with the results of pore volume, the RPA in the DP-120 and DP-180 groups was densely distributed in the range of 0–150 Å and showed a higher pore area, and the RPA in the DP-180 presented an extremely higher pore area than the DP-120 group.

As shown in Figure 5, the infrared spectra of RPA in the DP-0, DP-60, DP-120, and DP-180 groups were obtained. The obvious peak around 3,440 cm⁻¹ was related to the stretching vibration of the -OH group in the polymer. A peak near to 2,930 cm⁻¹ belonged to the asymmetric stretching vibrations of the -CH group on the benzene ring. The peak at 1,630 cm⁻¹ was assigned to the stretching vibrations of -C=O in the aromatic ring framework. The stronger peak in the wavenumber of 1,318 cm⁻¹ represented the CaC₂O4 molecules. The peaks around 1,150 cm⁻¹ and 1,020 cm⁻¹ displayed the characteristic spectra of ether bond (=C-O-C/C-O-C). The peaks in the range of 700–500 cm⁻¹ in the four RPA samples were mainly attributed to the external bending vibration of the C-H groups connected with the benzene ring.

As shown in Figure 6A, there were three stronger peaks around 285, 400, and 530 eV and they were mainly attributed as the binding energy of C 1s, N 1s and O 1s, respectively. The XPS spectra were calibrated with the C 1s peak of contaminant carbon at 284.8 eV. Figure 6B shows three standard peaks of C 1s around 284.7, 286.2, and 288 eV which were attributed as the binding energy of C-H/C-C, C-O/C-N, and O-C=O, respectively. It also can be seen that three peaks showed a small shift of 0.5–1.0 eV to left in the ranges of 286.2–286.3 eV and 287.9–288.4 eV. As compared with the DP-0 group, the peak intensity of O-C=O was increased with the increment of treatment time from 60 to 180 s. Figure 6C showed a normal peak near 400 eV which was related to the binding energy of C-NH₂, and the intensity also presented the same increasing trend with plasma exposure duration. Figure 6D shows two obvious peaks at 531.25 and 532.75 eV corresponding to C=O and C-O, respectively. The intensity of C=O of RPA in the DP-60, DP-120, and DP-180 groups was stronger than that in the DP-0 group with an increasing trend over plasma treatment time, while the intensity of C-O in the DP-60, DP-120, and DP-180 groups was weaker than in the DP-0 group. As shown in Table 3, the oxygen atom content of RPA in the DP-60, DP-120, and DP-180 groups was 27.09, 27.27, and 27.61%, respectively, which was higher than the DP-0 group of 26.72%.

The TG and DTG characteristic curves of the four RPA samples are shown in Figure 7A. The four RPA samples displayed similar weight loss regions in the range of 26.30–141.98°C (region I; Figure 7B), 141.98–237.73°C (region II; Figure 7C), and 237.73–556.82°C (region III; Figure 7D), and presented about 6.34%, 9.47%, and 59.98% of weight loss, respectively. In both region I and region II, the weight loss of the RPA samples in the DP-60, DP-120, and DP-180 groups was further decreased as compared with the DP-0 group and the decreased value was related to the plasma treatment time.