Rapid Quantification of Evodiamine and Rutaecarpine in Evodia Rutaecarpa（Juss.）Benth.Using Supercritical Fluid Chromatography

BAI Guo-yu，DAI Ying-ping，ZHUZHAO Jing-tao，LI Kai-lun，JIANG Wen-jia，XU Dong-yue，F(xiàn)U Qing，JIN Yu

（Engineering Research Center of Pharmaceutical Process Chemistry，Ministry of Education，School of Pharmacy，East China University of Science and Technology，Shanghai 200237，China）

Abstract：This study proved the feasibility of using supercritical fluid chromatography（SFC）in rapid quantification of evodiamine and rutaecarpine in Evodia rutaecarpa（Juss.）Benth.（E. rutaecarpa）. The two alkaloids were firstly extracted by supercritical fluid extraction（SFE）.As a result，the developed SFE conditions were as follows，sample（120-200 mesh）/cellulose acetate=1∶1，20% ethanol as modifier，40 ℃，25 MPa and the CO2 flow rate of 8 mL/min.The developed SFC method was carried out on a 2-EP column，with methanol as modifier at a flow rate of 1.4 mL/min，a column temperature of 35 ℃and a back pressure of 15 MPa.The baseline separation of evodiamine and rutaecarpine was completed within 6 min.After verification，the SFC method was confirmed to be linear with correlation coefficients（r2）of 0.9998，good precision（RSD below 0.50%）and accuracy（recoveries of 102%-109%）.Meanwhile，limits of detection（LODs）and limits of quantitation（LOQs）for evodiamine were 1.00 and 3.33 μg/mL，while LOD and LOQ for rutaecarpine were 0.95 and 3.17 μg/mL，respectively.Finally，the method was applied to the determination of contents of evodiamine and rutaecarpine in ten E.rutaecarpa samples from 4 areas.The total content of target alkaloids accounted 2.95% for Guangdong，1，0.51% for Guangdong，2，1.27% for Guizhou，3，1.00% for Hunan，4，0.93% for Hunan，5，1.81% for Jiangxi，6，0.73% for Jiangxi，7，0.58% for Jiangxi，8，0.41% for Jiangxi，9 and 0.36% for Jiangxi，10.Although the ten samples all meet the requirements for the Chinese Pharmacopoeia，the origin difference of sample quality is very obvious，and the samples from the same place also have differences.In addition，the developed method was compared with the Pharmacopoeia method（2020 Edition）.The results of content determination by the two methods were similar（1.27% by SFC vs.1.12% by Pharmacopoeia），while the time required by SFC（6 min）was significantly shorter than that by the Pharmacopoeia method（25 min）.This work showed the potential of SFC in the fast quantification of active components in traditional Chinese medicines.

Key words：supercritical fluid chromatography（SFC）；supercritical fluid extraction（SFE）；quantification；Evodia rutaecarpa（Juss.）Benth.；evodiamine；rutaecarpine

The dried fruit ofEvodia rutaecarpa（Juss.）Benth.（E. rutaecarpa）is a well-known traditional Chinese medicine（TCM）and previous phytochemical studies revealed thatE.rutaecarpamainly contained alkaloids，essential oils，limonin，carboxylic acids and flavonoids，which exhibited the effects of anti-tumor，antibacterial， anti-obesity， anti-diarrheal and anti-anoxia［1-4］. Evodiamine and rutaecarpine， two indole alkaloids（Fig. 1）ofE.rutaecarpa，are the target components in assay testing in the Chinese Pharmacopoeia（2020 Edition）［5］.It required that the content of total amounts of evodiamine and rutaecarpine should not be less than 0.15% in the driedE.rutaecarpa.Until now，a large number of studies have focused on the extraction and analysis of evodiamine and rutaecarpine.High performance liquid chromatography-ultraviolet detection（HPLC-UV）is currently the most commonly used method．It is worth noting that the further combination of HPLC and mass spectrometry（MS）improves the detection sensitivity， and ultra high performance liquid chromatography（UPLC）has also been developed to obtain more efficient separation.The advanced methods now have been used in the quantification of evodiamine and rutaecarpine in various samples including plant［6］，prescriptions［7］，supplements［8］and plasma［9］.Herein，the purpose of our research is to develop an alternative to HPLC based on supercritical fluid chromatography（SFC）for the rapid quantification of evodiamine and rutaecarpine ofE.rutaecarpa.

Fig.1 Structures of evodiamine and rutaecarpine

SFC is a newly emerging technology that uses supercritical CO2（Sc-CO2）as the main mobile phase.Due to the fact that the density of Sc-CO2is less than that of liquid and the diffusion capacity is similar to that of gas，shorter analysis time can be obtained by SFC when compared to traditional HPLC.Except this，the amounts of organic solvents used were also reduced， which was in line with the concept of green chemistry［10-12］.In SFC，the separation of basic compounds including alkaloids is also worthy of attention and sometimes there are tailing peaks due to the possible secondary ionic interactions of positively charged basic compounds with silanol groups on the surface of the stationary phase［13］.Several solutions have been reported to solve this issue，and one way is the use of specific columns such as 2-EP and DEA. The 2-EP is often used in SFC separation of basic compounds because the secondary interactions can be shielded through the hydrogen bonding and ionic interactions between silanol and nitrogen atoms of the pyridine groups［14-15］.Another way is using of the electrostatic repulsions between the basic compounds and the protonated ligand of the stationary phase.This is the case for the DEA column（pKa=9.5）［16］.In addition，a few studies have been reported for the separation of basic compounds by SFC with additives to get symmetrical peaks［17-19］.These additives are believed to provide ion pairing mechanisms that can prevent the basic compounds from the secondary interactions with the residual silanol.

Before analysis，extraction is often a necessary step.Currently available methods for extracting ofE.rutaecarpainclude leaching［20］，ultrasound［1］，reflux［6］and soxhlet extraction［21］.In recent years，the use of supercritical fluid extraction（SFE）to extract components from TCMs has gained much attention.The studies have demonstrated that SFE can rapidly and effectively extract different kinds of compounds in mild conditions and its solvating power being easily manipulated by changes in pressure and temperature［22］.The operating conditions of SFE were dull and oxygen-free which could effectively avoid the oxidation conversion and high temperature deterioration of samples［23］.The SFE method was also used to extract evodiamine and rutaecarpine from the unripe fruit ofE.rutaecarpa.The results showed that the change at time，temperature and pressure of SFE could affect the yields of the two alkaloids［24］.

The purpose of this experiment was to realize the SFC quantification of evodiamine and rutaecarpine ofE.rutaecarpain a short time.Before SFC separation，the two alkaloids were effectively extracted fromE.rutaecarpaby SFE.The parameters of SFE were optimized，including particle size，the amount of dispersant added，modifier，flow rate of Sc-CO2，extraction time，pressure and temperature.Then，a SFC method was used to separate the extracts.The analysis time would be as short as possible while the baseline separation of two alkaloids were ensured.For this reason，column，temperature，pressure and flow rate were all optimized.Finally，the SFC method was verified，including linearity，stability，precision and repeatability，and then it was applied to the content testing of evodiamine and rutaecarpine in real samples of different origins.

1 Experimental

1.1 Chemicals，reagents and plant materials

HPLC-grade methanol（MeOH） and ethanol（EtOH）were purchased from J&K（Beijing，China）.Trifluoroacetic acid（TFA）was obtained from Tedia（USA）.CO2（food grade，purity ＞99.99%） was bought from Delum（Shanghai，China）.Other reagent solutions were of analytical grade.The information of ten batches ofE.rutaecarpasamples were shown in Table 1.The standards of evodiamine and rutaecarpine， as well as cellulose acetate were all obtained from Aladdin（Shanghai）.

Table 1 The informations of E. rutaecarpa of four origins and the contents of evodiamine and rutaecarpine（n=10）

1.2 Instruments

The SFE system（Waters，Milford，MA，USA）equipped with an auto extractor，a CO2deliver pump，a modifier deliver pump，a CO2cooling device，a back pressure regulator and a collection module.Acquity UPC2system（Waters，Milford，MA，USA）equipped with a binary solvent delivery pump，an autosampler，a column manager，a photodiode array detector and a two-step（active + passive）backpressure regulator.Waters Empower 3 software was used for instrument control and data acquisition of UPC2systems.ChromScope Instrument Edition v1.20 was performed to control SFE system and realize data acquisition.

1.3 Operating parameters

Optimization of SFE conditions：the particle size of the powder was 80-120，120-200 and 200-300 meshes，the ratio of sample/cellulose acetate was 1∶0，1∶1 and 1∶2，the proportion of modifier was 15%，20% and 25%，the flow rate of Sc-CO2was 6 and 8 mL/min，the extraction time was 40，60 and 80 min，the pressure was 20，25 and 30 MPa and the temperature was 35，40 and 45 ℃.The final SFE conditions were：after crushing and sieving，1 g powder（120-200 mesh）was weighted accurately，mixed with cellulose acetate in a ratio of 1∶1（mass ratio），and placed in a 5 mL extraction kettle.20%EtOH was used as modifier，25 MPa and 40 ℃were set respectively.The dynamic extraction time was 60 min and the total flow rate was 10 mL/min（Sc-CO2：8 mL/min and EtOH：2 mL/min）. After extraction，the Sc-CO2was removed by decompression.The solution was concentrated by rotary evaporation and then EtOH was added to a constant volume.

Optimization of SFC conditions：columns were Viridis BEH 2-EP，BEH，CSH Fluoro-Phenyl，HSS C18SB，and Torus DEA，2-PIC， 1-AA， DIOL. The columns were purchased from Waters（Waters，Milford，MA，USA）and had the same dimensions and particle sizes（100 mm × 3.0 mm i. d.，1.7 μm）except HSS C18SB（100 mm × 3.0 mm i. d.，1.8 μm）.The modifiers were MeOH，EtOH and MeOH-0.01% TFA（volume ratio），temperature was 45，40 and 35 ℃，pressure was 12，14 and 15 MPa，and the flow rate was 1.0，1.2 and 1.4 mL/min.The optimized SFC conditions were：Viridis BEH 2-EP column，injection volume was 1 μL，mobile phase A was Sc-CO2and B was MeOH，elution gradient was 0-5 min，5%-8%B，5-6 min，8%B；column temperature，pressure，flow rate and detection wavelength were set to 35 ℃，15 MPa，1.4 mL/min and 220 nm.

1.4 Method validation

20 mg of evodiamine and 10 mg of rutaecarpine were accurately weighted and dissolved in EtOH to 25 mL to get stock solution.Then EtOH was added as the diluent to the stock solution in proportion to get reference solution of 0.80，0.64，0.48，0.32，0.16 and 0.04 mg/mL evodiamine and 0.40，0.32，0.24，0.16，0.08 and 0.02 mg/mL rutaecarpine.Sample solutions were obtained through the optimized SFE method.

Series concentrations of reference solutions were injected 2 times to establish calibration curves while the concentrations were taken asX，corresponding peak areas asY.The method validation of evodiamine was carried out at 220 nm，and rutaecarpine at 340 nm.The solution of lowest concentration was continuously diluted to calculate the limit of detection（LOD）and limit of quantitation（LOQ）which were defined as 3 signalto-noise（S/N）and 10，respectively.One sample solution（Guizhou）was placed at room temperature and injected at intervals of 0，2，4，6，8，12 and 24 h to evaluate the stability.In the accuracy test，50%，100% and 150% of actual content of evodiamine and rutaecarpine were calculated according to the regression equation in advance，then respective reference solution was added to the same batch of sample solution（Guizhou）in triplicate.A reference solution was taken with 6 consecutive injections within a day for the precision test.Both the intra-day and inter-day repeatability were calculated.The samples（Guizhou）were extracted by the developed SFE method 6 times a day to get intra-day repeatability， while inter-day repeatability was calculated by 2 times a day in 3 consecutive days，then they were separated by the developed SFC method.

2 Results and discussion

2.1 Optimization of SFE conditions

In this study，the extraction of evodiamine and rutaecarpine ofE.rutaecarpawas carried out by SFE.The SFE parameters were optimized in turn，including material status，percentage of modifier，flow rate，extraction time，pressure and temperature.The influence of the above parameters on the changes in peak areas of evodiamine and rutaecarpine were investigated.The state of the materials including particle size and dispersant（cellulose acetate）was firstly checked.As the particle size decreased from 80-120，120-200 to 200-300 mesh，the peak area of those two targets increased firstly and then decreased（Fig. 2A）.Smaller particles lead to larger specific surface area，which is beneficial to extraction.However，if the particle size is too small，a turbulent flow is formed on the contact surface between the sample and the solvent in the extraction kettle，bringing a negative effect on the extraction［25-26］.After testing，the 120-200 mesh was selected.Furthermore，cellulose acetate was added and it was expected to disperse samples but not adsorb the targets.As the proportion of sample and cellulose acetate was 1∶1，there was the highest value of evodiamine and rutaecarpine（Fig.2B）.Further increased cellulose acetate caused the decrease of alkaloids content.Therefore，it is necessary to add an appropriate proportion of dispersant，which was 1∶1 in this study.

The proportion of EtOH also needed to be optimized and the effects of 15%，20% and 25% EtOH were compared.From the results of Fig. 2C，the extraction with 20% EtOH is the best.When the EtOH increased to 25%，the peak areas of evodiamine and rutaecarpine decreased，indicated that more non-target components were extracted.With the increase of EtOH ratio，the polarity of the extraction environment became greater，which was beneficial to extract more polar compounds.As a result，20%EtOH was added to the Sc-CO2.

Next， the influence of flow rate of Sc-CO2was also examined.Rising from 6 mL/min to 8 mL/min，the content of two alkaloids was increased（Fig.2D）.Due to the insufficient solvent loading at lower flow rate，the targets cannot be completely extracted.8 mL/min was set.The extraction time was also varied from 60 to 40 min，the peak areas of evodiamine and rutaecarpine went down，indicating that it did not complete the full extraction of the targets in such a short time.When the time was 80 min，there was no change in the amounts of two alkaloids compared to that in 60 min.Therefore，in this experiment，when the flow rate of Sc-CO2was 8 mL/min，60 min of extraction time can ensure complete extraction.

Fig.2 The effect of SFE parameters on the changes of peak areas of evodiamine and rutaecarpineA：particle size，B：amount of cellulose acetate added，C：percentage of modifier，D：Sc-CO2 flow rate

Last，the extraction pressure and temperature were optimized.In this study，the effect of these two parameters on the extraction results was not significant.It proved that the condition of 25 MPa and 40 ℃already guaranteed the effective extraction of evodiamine and rutaecarpine in this experiment.

It can be seen that the particle size of the sample had the greatest influence on the SFE results，followed by the dispersant and modifier.In addition，relatively high flow rate of Sc-CO2and long extraction time were also necessary to ensure the full extraction of the targets.However，in this experiment，the temperature and pressure did not cause obvious changes in SFE results.The optimal parameters of SFE were as follows：1 g of the crushed sample（120-200 mesh）was weighed，stirred with 1 g of cellulose acetate and then loaded into a kettle of 5 mL.20% EtOH was used as modifier，and dynamic extraction time，the flow rate of Sc-CO2，pressure and temperature were 60 min，8 mL/min，25 MPa and 40 ℃respectively.After extraction，the Sc-CO2was removed and the sample solution was concentrated by rotary evaporation，and then EtOH was added to a constant volume.

2.2 Optimization of SFC method

In order to realize the rapid separation of evodiamine and rutaecarpine ofE.rutaecarpa，SFC columns were firstly screened， in terms of retention time， peak shape and resolution. Under the same elution conditions， the extracts ofE.rutaecarpawere separated within 10 min on all 8 columns with almost symmetrical peak shape（Fig.3）.The longest retention was obtained on the 1-AA column due to its strong π-π interactions with the two alkaloids（Fig. 3A）.Furthermore，there was the reverse elution order of rutaecarpine（peak 1）and evodiamine（peak 2）on 1-AA and C18columns（Fig.3A and 3B），which was closely related to the different chemical ligands bonded on the surface of columns［27］.Neither C18nor CSH Fluoro-Phenyl columns could effectively distinguish the two alkaloids（Fig. 3B and 3C）.On DEA and Diol columns，rutaecarpine was not separated from its surrounding peaks（Fig. 3D and 3E）.As the remaining three columns（2-EP，2-PIC and BEH），baseline separation of the two alkaloids were obtained（Fig.3F-H），among which，the 2-EP identified the largest number of peaks in the shortest time（Fig. 3F）.Therefore，the 2-EP column was selected for the subsequent SFC separation.Next，the modifier，pressure，temperature and flow rate were optimized to further shorten the analysis time on the premise of ensuring sufficient resolution between target and its surrounding peaks.

Fig.3 SFC separation of the E.rutaecarpa extract using Torus 1-AA（A），Viridis HSS C18 SB（B），Viridis CSH Fluoro-Phenyl（C），Torus DEA（D），Torus DIOL（E），Viridis BEH 2-EP（F），Torus 2-PIC（G）and Viridis BEH（H）columns conditions：mobile phase A：Sc-CO2，mobile phase B：MeOH；gradient elution：0-10 min，5%-2%B；back pressure：14 MPa；column temperature：35 ℃；flow rate：1.2 mL/min；detection wavelength：220 nm，1μL sample was injected；1. rutaecarpine，2. evodiamine

The shorter retention time of rutaecarpine and evodiamine was obtained when MeOH was as modifier，followed by EtOH.After adding 0.01% TFA，the retention time of the two alkaloids almost unchanged，and the peak shape and resolution were not improved. So，in this experiment，MeOH was used as the modifier with no additives.SFC separation can be affected by temperature and pressure，which adjusted the elution strength of the mobile phase by temperature was from changing the density of Sc-CO2.

However，when the temperature was from 45，40 to 35 ℃，almost no change in the retention behavior of two alkaloids was observed.In contrast，the effect of pressure was obvious to a certain extent. From 12，14 to 15 MPa，the retention factor（k）of evodiamine decreased from 6.3 to 5.7.So，35 ℃and 15 MPa were chosen.Last，the flow rate was optimized.

Under the conditions allowed by the instrument，the flow rate of 1.4 mL/min was selected.The retention time of the two alkaloids became shorter.At the same time，the baseline separation between the targets and the surrounding peaks was still maintained.By increasing the flow rate，the purpose of rapid separation of rutaecarpine and evodiamine was achieved.

The parameters of SFC method were as follows：1 μL sample was injected，then it was separated on the 2-EP column with the gradient elution of 0-5 min，5%-8%MeOH，5-6 min，8%MeOH.Temperature，pressure and flow rate were 35 ℃，15 MPa and 1.4 mL/min.The sample and standards were separated by the SFC method and respectively monitored at the maximum absorption wavelengths of the two alkaloids at 220 and 340 nm（Fig. 4）.Rutaecarpine（peak 1）had the retention time of 1.59 min，retention factor（k）of 2.19，tailing factor（Tf）of 1.08 and column efficiency（N）of 1.5×104，respectively，and evodiamine（peak 2）were 2.99 min，k=4.97，Tf=1.10 andN=2.2×104.

Fig.4 SFC separation of the E.rutaecarpa extract（A，B）and the mixture of standard solution（C，D）A，C：220 nm；B，D：340 nm.conditions：mobile phase A：Sc-CO2，mobile phase B：MeOH；gradient elution：0-5 min，5%-8%B，5-6 min，8%B；column：2-EP；pressure：15 MPa；column temperature：35 ℃；flow rate：1.4 mL/min，1μL sample was injected；1. rutaecarpine，2. evodiamine

2.3 Method validation and content determination

Method validation of the SFC which includes regression equation，stability，accuracy，precision and repeatability（intra-day and inter-day）was performed.Results of the linear equation were shown in Table 2 and the correlation coefficients（r2）were 0.9998.A good linearity of evodiamine was in the range of 0.04-0.80 mg/mL，and it was in 0.02-0.40 mg/mL for rutaecarpine.LOD and LOQ were 1.00 and 3.33 μg/mL for evodiamine， and 0.95 and 3.17 μg/mL for rutaecarpine，respectively.A reference solution was taken with 6 consecutive injections to calculate the precision.The RSDs of peak area were 0.35% for evodiamine and 0.31% for rutaecarpine. The samples were extracted by the developed SFE method 6 times a day for intra-day repeatability and 2 times a day in 3 consecutive days for inter-day repeatability，then they were separated by the developed SFC method.The repeatability（intra-day and interday）were 3.8% and 5.2% for evodiamine，and 4.5% and 5.8% for rutaecarpine， respectively.The sample solution was placed at room temperature and injected at intervals of 0，2，4，6，8，12 and 24 h to evaluate the stability，and the RSDs were 0.74% for evodiamine and 1.0% for rutaecarpine.50%，100%，150% of evodiamine and rutaecarpine standards were added to the same batch of sample solution in triplicate， the recoveries of evodiamine and rutaecarpine were 102%-108% and 102%-109%，respectively，with their RSDs were both less than 3.6%.The results indicated that this SFC method had good precision，accuracy and repeatability，and the samples were stable within 24 h.More importantly，it had several characteristics，including rapid separation，specific extraction and convenient sample post-processing.

Table 2 The validation of SFC method for rutaecarpine and evodiamine including linear relations，LODs，LOQs，stability，precision，accuracy and repeatability

The samples ofE.rutaecarpaof four origins were analyzed by the validated SFC method，and the content of evodiamine and rutaecarpine were calculated.According to the China Pharmacopoeia（2020 Edition），it required that the content of total amounts of evodiamine and rutaecarpine should not be less than 0.15% in the driedE.rutaecarpa.As shown in Table 1，the total contents of the two alkaloids of tenE.rutaecarpasamples were 2.95%（Guangdong，1），0.51%（Guangdong，2），1.27%（Guizhou，3），1.00%（Hunan，4），0.93%（Hunan，5），1.81%（Jiangxi，6），0.73%（Jiangxi，7），0.58%（Jiangxi，8），0.41%（Jiangxi，9）and 0.36%（Jiangxi，10），respectively.Although all the ten samples met the requirements of the Chinese Pharmacopoeia，the origin difference of sample quality was very obvious.In addition，the developed method was compared with the Pharmacopoeia method（2020 Edition）.The content results by the two methods were similar（1.27% by SFCvs. 1.12% by Pharmacopoeia）. While the time required by SFC（6 min）was significantly shorter than that by HPLC in the Pharmacopoeia method（25 min）.

3 Conclusions

The study proved the feasibility of rapid quantification of evodiamine and rutaecarpine ofE.rutaecarpaby SFC.First，the two alkaloids were effectively extracted by the optimized SFE method.The particle size of the sample had the greatest influence on the SFE results，followed by the dispersant and modifier.At the same time，the flow rate of Sc-CO2（8 mL/min）and the corresponding extraction time（60 min）ensured the complete extraction.Then，the SFC method was developed to achieve rapid separation of theE.rutaecarpaextract within 6 min.The use of 2-EP column ensured the symmetrical peak shape of evodiamine and rutaecarpine，as well as the baseline separation from the surrounding peaks.The verified SFC method was applied to content determination of the two alkaloids ofE.rutaecarpaof four origins.Although all the ten samples meet the requirements of the Chinese Pharmacopoeia，there was a significant positive correlation between sample quality and origin.