Study of the volatilization rules of volatile oil and the sustained-release effect of volatile oil solidified by porous starch

Volatile oil from traditional Chinese medicine has various biological activities and has pharmacological activities in the central nervous system, digestive system, cardiovascular system, respiratory system, etc. These oils are widely used in clinical practice. However, the development of their clinical applications is restricted due to the disadvantages of volatile oils, such as high stimulation, high volatility and poor stability. To improve the stability of a volatile oil in the preparation process, its volatilization and stable release must be controlled. In this paper, porous starch was used as a solid carrier material, and liquid volatile oil was solidified by physical adsorption. GC–MS was used to determine the chemical constituents of the volatile oil, solidified powder and tablets, and the volatilization rules of 34 chemical constituents were analysed statistically. The solidified volatile oil/porous starch powder was characterized by XRD, TGA and DSC, and the VOCs of the volatile oil before and after solidification were analysed by portable GC–MS. Finally, the stable release of the volatile oil could be optimized by changing the porous starch ratio in the formulation. Volatilization was shown to be closely related to the peak retention time and chemical composition, which was consistent with the theory of flavour. The physical properties and chemical composition of the volatile oil did not change after curing, indicating that the adsorption of the volatile oil by porous starch was physical adsorption. In this paper, the porous starch-solidified volatile oil had a slow-release effect, and the production process is simple, easy to operate, and has high application value.

Appearance and adsorption rate of porous starch and volatile oil in a solidified cured powder. Different proportions of PS and VO were prepared and cured as a powder. The results showed that when the proportions of VO and PS were 1:5, 2:5, 3:5 and 4:5, the VO could be well absorbed, the dispersion of the powder was relatively uniform and did not stick to the walls or exhibit cohesion phenomena; the mixtures presented a powder state, and the surface of the powder was dry. When the ratio of VO to PS was 5:5, agglomeration occurred, and the prepared powder had an obvious greasy feeling, indicating that the VO proportion had exceeded the maximum adsorption capacity of the PS.
According to the appearance of the solidified PS/VO powder, the adsorption capacity of PS for VO could be preliminarily observed. On this basis, the contents of chemical components in the powders with different proportions of solidified powder were determined, and the adsorption rate of VO in different proportions of Stability study of porous starch solidified-volatile oil powder tablets. The VO:PS ratio of 3:5 was the optimal proportion of VO solidified by PS. The powder was pressed by a single stamping machine. The tablets were stored at 25 °C and sampled at different time points, and the retention and volatilization rates were calculated according to the method outlined in "Rules of volatilization of chemical constituents in the PS/ VO powders" in section (Table 2). When the solidified powder was prepared into tablets, the retention rate of d-limonene was only 31% initially, i.e., 0 h. The retention rate of d-limonene was 51% in the powder at 0 h. The reason may be that some volatile components were volatilized when the powder was exposed to the air in the www.nature.com/scientificreports/ process of pressing the tablet, so the initial retention rate of d-limonene in the tablet was low. This low initial volatile oil content increased the volatilization rate.
The peak area values after standardization were plotted over time (Fig. 3c). It can be seen from the figure that the curves of class A compounds were consistent, indicating that class A compounds exhibited a consistent www.nature.com/scientificreports/ volatilization rule. The curves of the class B components were basically consistent, indicating that the class B components showed a consistent volatilization rule. It can be concluded from the above results that the first 12 eluted compounds were are class A chemical components, all of which are olefins with small molecular weights, high volatilization curve slopes and fast volatilization speeds. The 13th chemical component (linalool), which was an alcohol, began the elution of class B components. The elution of these compounds was later than that of the class A compounds, and the class B components were alcohols, ethers, phenols, aldehydes, etc., with large molecular weights, relatively stable volatilization curve and slow volatilization speeds. From the above results, it can be concluded that the release rate of essential oil is related to the saturated vapor pressure of the compound, which is mainly affected by molecular weight and material structure. The smaller the molecular weight, the easier it is to escape; The weaker the hydrogen bond and dipole interaction between homologous compounds, the easier it is to escape from gasification. It can be seen from the figure that compounds  www.nature.com/scientificreports/ with small molecular weight and no intramolecular hydrogen bond in volatile oil have fast volatilization rate, such as pinene, limonene, Terpinene, etc. The volatilization rate of compounds with large molecular weight and intramolecular hydrogen bond is slower than the former, such as terpineol, thymol, perilla aldehyde and so on.

Rules of volatilization of chemical constituents in the PS/VO powders.
The VO:PS ratio of 3:5 was the optimal ratio for curing the volatile oil, and the resulting powder was subsequently analysed. The peak areas of the chemical constituents of the VO in the solidified powder were analysed at different time points. As conducted for the chemical constituents of the raw VO, the changes in peak areas over time of the 34 chemical constituents were analysed and plotted, and the analytical method was the same as that in Sect. 3.7.1 (Fig. 4a). Class A chemicals were still grouped together. Most of the chemical components of class B could still be grouped together, but a few of them, such as (+)-delta-cadinene, thymol methyl ether, humulene, and nerol, were relatively dispersed, which might be related to the PS carrier material (Fig. 4b). After Standardization, the peak areas of class A and class B components were plotted versus time (Fig. 4c).
The volatilization rule of chemical components in the PS/VO tablets. The solidified powder of VO:PS = 3:5 was pressed into tablets. GC-MS was used to detect the chemical composition of the VO in the tablets at different time points. The 34 peak areas of the VO chemical constituents in the tablets at different time points were standardized and analysed to generate the volatilization curves of chemical constituent peak area changes over time (Fig. 5a). The peak area values of the 34 chemical constituents were standardized by standard deviation and analysed by clustering (Fig. 5b). It can be seen from the figure that class A components and class B chemical components were grouped into one class. The normalized peak areas of the class A and class B components were plotted against time (Fig. 5c).
The chemical composition of the VO, cured powder, and tablets changes with time, and clustering analysis showed similarities in all three forms of the VO. The classification of chemical compounds was closely associated www.nature.com/scientificreports/ with the retention time. The first 12 eluted chemical compounds could be combined into a class (class A), and the 22 other compounds classed them together into class B. Second, the clustering results were closely related to the type of chemical component. Class A chemical components were all olefins, while class B chemical components were alcohols, ethers, phenols, aldehydes, etc. The volatilization rules of class A and class B components were plotted, and it was found that the slope of class A chemical component volatilization was larger than that of class B components, which was consistent with the theory of head, body and base notes in essence and fragrance theory.
In the classification of spices, spices were divided into three categories, namely, those having head notes, body notes and base notes, according to the time of volatilization and retention of their fragrance on scented paper.
Evaluation of the sustained-release effect. The volatilization rate of d-limonene in the VO, PS/VO powder and PS/VO tablets was analysed, and the volatilization curves were generated (Fig. 6a). Because the initial VO content of the tablets was lower than that of powder samples, the volatilization rate of the tablets was faster than that of the powder but still slower than that of the raw VO. With d-limonene as the index component, the aroma graphs of d-limonene in the VO, PS/VO powder and PS/ VO tablets were generated (Fig. 6b). It can be seen from the aroma graph that the VO was volatilized quickly, and the aroma graph shows a sharp peak. After the solidification of the VO in the PS, the release of the VO became slow and stable, achieving the effect of steady and slow release. Compared with that of the PS/VO tablets, the release of VO from the PS/VO powder tended to be more gradual, which was closely related to the retention rate of the initial VO when producing the powder and tablets.
XRD, TG and DSC characterization of the PS/VO powder. XRD analysis was performed on PS/VO powder with different proportions (Fig. 7a). Compared with that of PS, the PS/VO powder XRD pattern had a higher characteristic peak, and the larger the proportion of the VO was, the larger the characteristic peak, indicating that the PS absorbed the VO, and the powder surface also contained a certain amount of volatile oil, so the PS/VO powder X-ray diffraction pattern had a peak more characteristic peak of VO than of PS 18,19 . www.nature.com/scientificreports/ Thermal analysis can characterize the changes in physical and chemical properties of a material, and TGA was performed on PS/VO powders with different VO and PS proportions (Fig. 7b). It can be seen from the figure that the thermogravimetric diagram of PS before and after curing with VO was consistent, indicating that the PS had no effect on the physical and chemical properties of the VO 20 .
DSC can reflect the combined characteristics of a drug and carrier, and PS/VO powders with different PS and VO proportions were analysed via DSC (Fig. 7c). It can be seen from the figure that PS/VO powders with different VO and PS proportions all exhibited exothermic peaks at 104 °C. The larger the proportion of the VO was, the larger the exothermic peak. The above results were consistent, indicating that the PS did not change substantially after curing the VO, indicating that the mechanism by which the PS and VO were combined was physical adsorption 21 .
VOC analysis before and after curing of volatile oil. The VOCs of the VO, PS/VO powder and PS/ VO tablets in the air were measured by portable GC-MS, and according to the NIST database, the chemical components of the VO, PS/VO powder and PS/VO tablets were detected to be alpha-pinene, beta-pinene, alpha-terpinene, cymene, d-limonene, and terpinolene. In the three forms of raw VO, PS/VO powder and PS/ VO tablets, the chemical constituents of the VO that volatilized into the air were consistent, indicating that there was no chemical change after VO curing and after tablet pressing, and the adsorption mechanism of the VO by PS was physical. d-limonene accounted for more than 80% of the VO composition in dried tangerine peel and immature tangerine peel and was therefore easily detected. However, as the instrument sensitivity of the portable GC-MS is in the range of low-to-medium ppb, when other chemical components at low levels were detected, d-limonene exceeded the maximum value of detection by the instrument, and the peak area appeared as a red warning www.nature.com/scientificreports/ (Fig. 7d). The red area is the peak area of d-limonene. Therefore, this instrument could only detect high contents of alpha-pinene, beta-pinene, alpha-terpinene, cymene, d-limonene, terpinolene, etc., while other volatile components could not be detected.

Conclusion
Volatilization rules of the volatile oil are related to chemical composition and aroma theory. Based on the study of volatilization rules, the chemical constituents of the VO were grouped into two categories according to their retention times. First, the classification of compounds into these two classes was closely related to the retention time, and their classification accuracy was verified statistically. Second, the compound classification was closely related to the type of chemical component, and this was also the first study to innovatively associate volatilization rules with chemical compound type. At the same time, the volatilization rules were consistent with the aroma theory of head notes, body notes and base notes. The volatile oil cured by porous starch had optimized volatilization and stable release. The study on the stability of the solidified VO powder produced with PS and the evaluation of the slow-release effect showed that the release of the VO had changed to some extent after processing, and the release of the VO could be regulated by the PS to achieve optimal release and stability. The XRD, TGA and DSC characterizations of the VO solidified by PS showed that the physical properties of the VO did not change after solidification. The VOCs of VO before and after curing were analysed for the first time in this paper. The results showed that the volatile chemical composition of the VO did not change after curing, indicating that the adsorption of the VO by PS was a form of physical adsorption. The volatile oil of traditional Chinese medicine is highly volatile, the effective components are easy to be lost, and it is unstable to air, light and heat, which makes it difficult for the stable release and administration of traditional Chinese medicine volatile oil, affects the clinical efficacy of volatile oil preparation, and has become an important bottleneck in the development of traditional Chinese medicine volatile oil. Focusing on the key www.nature.com/scientificreports/ problems perplexing the volatile oil of traditional Chinese medicine (strong volatility and poor stability), "adsorption-desorption" curing technology enables porous carrier materials to regulate the stable release of traditional Chinese medicine volatile oil, improve the volatility and stability of volatile oil, improve the quality of traditional Chinese medicine volatile oil preparation, realize stable release and administration, and ensure the effectiveness. "Adsorption-desorption" curing technology not only solves the problem of traditional Chinese medicine volatile oil, but also innovates the key technology of traditional Chinese medicine volatile oil preparation and endows the volatile oil with modern scientific and technological connotation.

Materials and methods
Instruments and materials. The following instruments were used in this study: an RW50 top electric mixer (Shanghai Analysis), a YP2002 electronic balance (Shanghai Yongzheng Medical Instrument Co., LTD.), an Agilent 7890A/5975C gas chromatography mass spectrometer (Agilent, USA), a D8 ADVANCE X-ray diffractometer (BRUKER, Germany), a TGA4000 thermogravimetric analyser (PE), a DIAMOND DSC (US PE Company), an MS2000 laser particle size analyser ( Determination of the physical properties of porous starch. The particle size of porous starch (PS) was measured using a Malvern laser particle size meter; air was the dispersion medium, the feed vibration was 70%, and the pressure of the dispersed starch was 2.00 Bar. An appropriate amount of PS was placed in a sample tube and dried to constant weight with nitrogen at 60 °C. The surface area, porosity and pore size of PS were measured by a specific surface area meter. (1) Volatilization rate of volatile oil = initial peak area − peak area /initial peak area × 100% www.nature.com/scientificreports/ Stability of cured tablets. The cured powder with the optimal proportions the VO and PS was pressed, and the sheet weight and pressure of the single stamping machine were adjusted so that the cured powder could be pressed into tablets that met the standard requirements of medicinal tablets. The effect of the tablet form on the volatilization rate was investigated. The solidified tablets were placed at 25 °C and sampled at different time points. Four solidified tablets were weighed accurately placed in a conical flask, combined with 25 mL of anhydrous alcohol, weighed again, ultrasonicated for 10 min, cooled, weighed, and brought up to the initial weight with anhydrous alcohol. The mixture was shaken well, filtered with a 0.22 μm microporous membrane filter, and analysed by GC-MS. The retention rate (formula 3) and volatilization rate (formula 4) of the VO in the cured tablets at different time points were calculated.

Determination
Investigation of volatilization rules of the volatile oil before and after curing. The peak area of the chemical constituents of the VO changed with time, and a total of 34 chemical constituents were detected. Python software was used to generate the release curves of the 34 chemical constituents in the VO over time according to changes in their peak areas. Python software was used to standardize the standard deviations of the peak areas of all chemical components and clustered the volatilization rules of the 34 chemical components. According to the clustering results, the 34 chemical components were classified into two categories, and the curve of the peak values after standardization of these two categories versus time was plotted. The peak area of 34 chemical components in the PS cured powder and the PS cured tablets changed with time, and the analytical method of the tablets was consistent with that of the powder. The release of the VO in the solidified powder and the solidified tablet and the influence of PS on the release of the VO were investigated 25 .
Evaluation of the sustained-release effect of volatile oil before and after curing. With d-limonene as the index compound, the volatilization rate curves of the VO, solidified powder and solidified tablets versus time were plotted to evaluate the sustained-release effect. At the same time, the curve of the peak area of the d-limonene chemical composition in the VO, cured powder and cured tablet changed with time, and the axial symmetry image was included with the aroma graph of d-limonene. The change in the aroma graph before and after curing of the VO was compared to evaluate its slow-release effect.

XRD, TG and DSC characterization of volatile oil solidified by porous starch. XRD was per-
formed on a solidified PS/VO powder sample at temperatures from 10 to 80 °C. TGA was performed on the solidified powder with a heating rate of 10 °C·min −1 from 30 °C to 600 °C in nitrogen, and DSC was performed on the solidified powder with a heating rate of 10 °C·min −1 from 30 °C to 450 °C in nitrogen.
VOC analysis before and after curing of volatile oil. Portable GC-MS instruments are sensitive in the low-to-medium ppb range for the analysis of volatile organic compounds (VOCs) in the air. According to the characteristics of the instrument, the VO, solidified powder and solidified tablet samples were placed in a sealed collector and volatilized naturally, and the volatile components of the gas in the collector were determined by portable GC-MS. The detection mass range was m/z 45-250, the air intake was 40 mL, the filament delay time was 210 s, the inlet line cleaning time was 1 min, and the sample collection time was 1 min. The chemical constituents of the VO released into the air before and after curing were detected, and the changes in volatile components before and after curing were evaluated. (3) The retention rate = actual peak area of D − limonene in solidified powder/ theoretical peak area of D − limonene in solidified powder × 100% (4) The volatility rate = (initial retention rate − retention rate)/initial retention rate × 100%