A Novel Forming Method of Traditional Chinese Medicine Dispersible Tablets to Achieve Rapid Disintegration Based on the Powder Modification Principle

Slow disintegration and poor solubility are common problems facing the dispersible tablets of Traditional Chinese Medicine (TCM). In an early study, the research group found that co-grinding of extracts and silica could achieve a rapid disintegration effect, though the mechanism of this effect was not thoroughly elucidated. In this study, Yuanhu Zhitong dispersible tablets (YZDT) were selected as a model drug to explore the mechanism of rapid disintegration and dissolution. First, eight types of silica were used to prepare modified YZDT, and their disintegration time and amount of dissolution within 5 min were measured. Next, the powder properties of eight types of silica were investigated. By correlation analysis, it was found that the average pore size and density of silica were closely related to the effect of promoting disintegration. It was determined that the co-grinding of silica and extracts provided high porosity for the raw material drug, and its abundant narrow channels provided a strong static pressure for water penetration to achieve a rapid disintegration effect. Meanwhile, it was found that the addition of silica had a certain effect on promoting dissolution. Our results provide a highly operational approach for improving the disintegration and dissolution of TCM dispersible tablets. Meanwhile, this approach is also beneficial for establishing a high-quality evaluation index for silica.

SCIENTIFIC RePoRTS | (2018) 8:10319 | DOI: 10.1038/s41598-018-28734-x disintegration time of YZDT, and different silica had different disintegration effects. Therefore, exploring how the different types of silica promote disintegration is the key step for inducing TCM dispersible tablets to disperse.
Powder properties of eight types of silica. In Table 3, d(0.5) indicates the powder that particle size was accounted for 50% of the total 24 . The particle size of M-5 was the smallest, while US had the minimum value. M-5 had the minimum bulk density (BD), real density (RD), and average pore size (APS), while precipitated silica (PS) had the maximal BD, RD, APS and minimum compression rate (COM). The specific surface area of hydrophilic silica (M-5, A200, A380) was significantly smaller than other types. It can be concluded from Table 3 that the angle of repose (AR) was variable among all silicas, which led to differences in flowability. MCM-22 was the worst, and the best was M-5. It could be concluded that silica produced by different manufacturers exhibited different powder properties. Figure 2 shows that the isotherms of eight types of silica belong to Type IV adsorption isotherms. The typical characteristics of nitrogen adsorption curves are that the adsorption-desorption curves suddenly increase in the high-pressure zone, where the P/P 0 value is approximately 0.8~1, and the hysteresis loop appears. These phenomena are related to the capillary condensation phenomenon of mesoporous materials (diameter was between 2 and 50 nm). The conclusion could be drawn from the above results that the silica was a type of fine powder with a microporous structure.   Fig. 3, and the content of the five components is listed in Table 4. These results showed that almost all of the five components dissolved relatively faster from A 200 modified YZDT, while the PS was slower. It was more clear that the amount of dissolution in STA was considerably lower than other silica modified samples, even less than 50%. This finding indicated that the addition of silica played an important role in the dissolution of active components.
Correlation Analysis Test. The disintegration and dissolution data were analyzed, as shown in Fig. 4. As the disintegration time increased, the content of dissolution generally showed a gradually decreasing trend. Clearly, Fig. 4 shows that the disintegration of the STA was the slowest, and the dissolution effect was the most unsatisfactory. Therefore, it could be concluded that the promotion of disintegration could also be conducive to the dissolution of active components.
To explore the micromechanism of silica in promoting the disintegration and drug release of dispersible tablets, the research team analyzed the correlation between disintegration time and powder properties. The analytical method was simple linear correlation and was performed using Origin 9.0 software. The correlation analysis results are shown in Fig. 5. According to the R value, it was clear that d(0.5), AR, CS, BET, and Langmuir were almost not correlated with disintegration time, and the highest one was APS with the R value of 0.6785. The R value of BD and RD also reached 0.6484 and 0.6573. These results suggested that the aiding effect of silica on YZDT might be related to its APS and density.
Results of water absorption. The water absorption device is shown in Fig. 6A. The water absorption data of eight kinds of silica was analyzed by Origin 9.0 software, as shown in Fig. 6B. In Fig. 6B, M-5 and A200 had a strong water absorption capacity, while US, MCM-22 and PS showed weak effects. In addition, R972 and R974 are hydrophobic silicas and showed no absorption of water. There was a certain relationship between water absorption and promotion of disintegration effect. The stronger the water absorption was, the shorter the disintegration time was.

Discussion
According to our research, APS and density have been demonstrated to be the key factors affecting the rapid disintegration of dispersible tablets. The smaller APS and density were, the stronger the disintegration performance was. The mechanism of this phenomenon is shown in Fig. 7. The co-grinding of silica and extracts provided high porosity for the raw material drug, and its rich narrow channel provided a strong static pressure for water penetration across a high viscosity water barrier on the surface of the tablets [25][26][27] . Based on the Laplace equation, the static pressure of water is inversely proportional to the APS. The smaller the APS was, the stronger the water absorption power was, and the faster the water absorption was, which was also the microscopic mechanism by which the silica channels affected the disintegration of the dispersible tablets. Meanwhile, the density of silica also had a great effect on rapid disintegration. The lower the density was, the higher the quantity of silica particles per unit mass was, and the more able the particles were to fully contact and disperse with the extracts, which was conducive to reducing the cohesion of the extracts and forming a micro-environment that is beneficial to disintegration. Meanwhile, the addition of silica also served to reduce the use of a portion of the disintegrants. On the one hand, the elastic recovery was effectively controlled; on the other hand, the cheap silica could reduce production costs. This study formed a new technology for the preparation of dispersible tablets such that the rapid disintegration of TCM dispersible tablets was no longer dependent upon the role of disintegrants, greatly simplifying the difficulty of screening excipients.
As a medicinal excipient, the quality evaluation indexes of silica were various in different countries' pharmacopoeia, primarily including pH, chloride, sulfate, iron salt, heavy metal, ash, moisture, content and other indicators. These indicators effectively ensure the purity and safety of the silica. However, due to different sources and manufacturers, the silica was made by different production methods and different process parameters, which led to discrepant physical properties. It was unreasonable to measure the application of silica as a pharmaceutical excipient with these indicators alone. Therefore, a high-quality evaluation index of silica needed to be established. The macroscopic nature was closely related to the microstructure. In this paper, we showed that the APS and density were the key indicators for evaluating silica as an auxiliary disintegrant. The establishment of the standard range on APS and density could effectively determine the auxiliary disintegration function, which was the significant value of this study.   According to the Laplace equation, the smaller the APS of the silica was, the stronger the water absorption was. At the same time, the water absorption of modified YZDT was verified in this paper.

Conclusion
Silica is a type of auxiliary disintegrant with strong water absorption, nonexpansion and high security. The average pore size and density were the key influencing factors to evaluate disintegrative performance. In addition, the co-grinding of silica and extracts could significantly speed the disintegration of dispersible tablets. This approach was conducive to the dissolution of active components. Meanwhile, the cost of production was somewhat reduced. This approach will be beneficial to form a new pattern for the preparation of TCM dispersible tablets.

Preparation and evaluation of dispersible tablets. Preparation of extracts. Angelica and Vinegar
Corydalis were respectively crushed into coarse powder by HX-200 (Yongkang Xi'an Hardware Factory, China). Next, 588 g of Vinegar Corydalis coarse powder and 445 g of Angelica coarse powder were extracted twice with 5-fold volume of ethanol with a purity of 60% for 1 h each. The extracts was filtered through a qualitative filter paper to yield the sample solution. Next, the filtrate was concentrated at 81 °C followed by freeze-drying, and grinding to powder 17 .

Preparation of YZDT.
To study the correlation between the properties of silica and disintegration time of dispersible tablets, 5% silica was ground with extracts until fully modified. At the same time, 5% soluble starch was In the corresponding mode, pseudo-molecule-ions were specified as the parent ion to start second-order fragment-MS determination. According to the structural information of second-order mass spectrometry, chemical components in YZDT were identified.  Powder properties of silica. The powder properties of silica were determined, including particle size distribution, specific surface area and porosity, density and compression and angle of repose.
Particle size distribution. A little sample was placed on the sample groove, and the Scirocco 2000 dry dispersion method was adopted to determine the distribution of particle size by Marvern MS2000 laser particle size analyzer (Marvern Ltd, USA) 24 .
Specific surface area and porosity. The powder was added to a hemisphere of the sample tube and placed in a Smart Prep065 pretreatment system with N 2 blowing to constant mass under heating at 200 °C. Next, the BET method was adopted to measure specific surface area and porosity and was tested at 55 points of each sample by TriStar 3000 automated surface area and porosity analyzer (Micromeritics Ltd., USA) 29,30 .
Density and compression. Sample was added to a 20 mL cylinder, and the volume was approximately 10 mL. At the same time, precision volume (V1) and weight (m1) were recorded. Loose density (ρ1) was calculated as formula (1). Next, the cylinder was beaten 60 times per minute for 5 minutes. In addition, the vibratory volume (V 2 ) was recorded to figure out vibratory density (ρ 2 , listed in formula 2). Compressibility (COM) was the ability to reduce the volume of powder under pressure (listed in formula 3). 2 Angle of repose. The double funnel method was used to measure the angle of repose in closed and independent conditions. The radius (r) and height (h) of cone were accurately measured. The angle of repose (θ) was calculated as formula (4). MS Conditions. Samples were analyzed by an Agilent 1260 high-performance liquid chromatograph and Agilent 6460C triple-quadrupole tandem mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) using an Agilent Technologies Zorbax Eclipse XDB-C18 column (4.6 mm × 150 mm, 5 μm). The column temperature was 35 °C and 10 μL of the sample solution was injected into the system. The mobile phase was composed of (A) 0.1% aqueous formic acid in water and (B) acetonitrile using a gradient program of 100-60% B for 0~1 min, 60-10% B for 1~3 min, 10-0% B for 3~5 min, and 0-100% B for 5~10 min with a mobile flow rate of 1 mL/min 31 . Mass spectrometric scans were obtained by electrospray ionization (ESI) in positive-ion mode with a scanning interval 100-1000 m/z. The main parameters for MS were set as follows: gas temperature, 300 °C; gas flow, 11 L/min; nebulizer, 35 psig; capillary voltage, 4000 V; atomizer pressure 15 psi (1 psi = 6.895Kpa). MS parameters and MRM transitions of each analyte are shown in Table 5.
Measurement of dissolution. The dissolution study was performed by placing the YZDT in 200 mL of ultra-pure water that had been treated by ultrasound for 30 minutes, using the small cup method at 100 rpm and 37 ± 0.5 °C. Dissolution medium (2 mL) was withdrawn at 5, 10, 20, 30 and 40 min through a 0.22 μm microporous membrane 32 . The sample was analyzed by LC-MS (Agilent 1260 high performance liquid chromatography, Agilent, USA; Agilent 6460 C Triple Quadrupole LC/MS with ESI ion source, Agilent, USA), each YZDT was paralleled six times and five components were determined.
Correlation analysis between disintegration time and powder properties. Horizontal axis showed the powder properties of each silica, and vertical axis showed the results of disintegration time. The simple correlation analysis was adopted to test the correlation coefficients using Origin 9.0 software. The larger the correlation coefficient R value was, the greater the influence of powder properties was.
Water absorption capacity verification. The water absorption device consists of a sintered glass funnel and a 10 mL pipette. The funnel and pipette were connected by a rubber tube to control the sintered glass funnel, which was 38 cm away from the table, 37.5 cm from the head of the pipette to the table, meaning that the pipette was filled with water and the funnel filter plate was fully wetted without water leakage. A filter paper with a small opening in the middle was placed in the sintered glass funnel, and a hollow glass tube with smooth inner circular wall and a diameter of 2.7 cm was placed under the opening. A certain amount of silica was poured into the hollow glass tube and stacked into a loose powder heap with the same thickness and diameter. At the same time, the change of the liquid level in the pipette was captured by a digital camera. The temperature was controlled at 18~22 °C and the humidity was 40~60%. When the water was saturated, the record was stopped. The video was introduced into the computer to read the data of liquid surface changing with time. Each sample was repeated 3 times 14 .
Statistical analysis. All the values were expressed as means ± SD. The results were analyzed by one-way analysis of variance (ANOVA) using SPSS22.0 software (SPSS Inc., USA). p < 0.05 was considered to be statistically significant.