Chapter 6 IDENTIFICATION AND CHARACTERIZATION OF FLAVONOIDS BY HPLC AND LC-MS/MS ANALYSIS

Chapter 6 IDENTIFICATION AND CHARACTERIZATION OF FLAVONOIDS BY HPLC AND LC-MS/MS ANALYSIS

6.1 Introduction This chapter discusses the method development carried out for the identification and characterization of flavonoids on High Pressure Liquid Chromatography (HPLC) and Mass spectrometry (LC-MS). The first part of the chapter deals with an elaborate description of entire method development on HPLC for the separation and identification of flavonoid constituents along with the necessary modifications made in accordance to available literature. Second part of the chapter includes detailed description of Mass spectral studies for the characterization of flavonoid components followed by results and discussion section giving elaborate interpretation used for identification of individual compounds. 6.2 Experimental 6.2.1 Chemicals Methanol and Acetonitrile of HPLC grade were obtained from Merck Pvt. Ltd. Quercetin and Kaempferol of purity > 99.8% were obtained from Sigma Aldrich, Germany. Formic acid was of AR grade 98-100% obtained from S.D. Fine Chem. Ltd. 6.2.2 Instrumentation 6.2.2.1 HPLC Instrumentation HPLC analyses were conducted on a JASCO HPLC 1500 series connected to JASCO- MD-2010 plus multi-wavelength detector (PDA detector) as shown in Figure 6.1. LC system of JASCO -1500 series included JASCO-model PU-2080 intelligent HPLC pump equipped with JASCO-LG-1500-02-Ternary gradient unit, JASCO-DG-1580-53-3-line School of Science, SVKM s NMIMS (Deemed-to-be University) 137

degasser and JASCO-AS-1555-10 intelligent HPLC sampler having a variable (1-100µL) injection volume loop. Data obtained was analyzed and processed using the ChromNav offline processing software. The chromatographic separation was achieved using a Kromasil C18 column having dimensions of 4.6 x 250mm, particle size 5µm. Fig. 6.1: JASCO 1500 Series equipped with PDA detection system JASCO-1500 Series instrument comprised of JASCO-model-PU-1580 intelligent HPLC pump for constant flow and pressure modes. The pump works on 230V, 5A, equipped with front panel control keys to monitor pump operation, solvent flow mode selection, time program execution and display on the monitor screen. The LCD displays operating status such as, pressure readings, operating parameters and error messages. Apart from these, the other instrumental parts include, a solvent through which plunger washing solvent is drained off, purge valve allowing the solvent to flow out through drain tube, School of Science, SVKM s NMIMS (Deemed-to-be University) 138

pump head containing two chambers with inlet and outlet check valves for solvent delivery, and outlet port through which solvent from the pump is discharged. The front panel keypad controls all these functions and a rare panel controls gradient elution. In the constant flow mode, the flow rate range is 0.01 10.0mL min -1 with a flow precision of + 2% and a flow rate setting at 1.00-10.0mL min -1 increment. In another setting, the range 0.01-0.99mL min -1 increment is also available. The operating pressure ranges are 10-500 kg/cm 2 (6.10 10.0ml min -1 ). In the constant pressure mode, the pressure range is 30 500kg/cm 2 with a setting at 10kg/cm 2 and precision of + 10%. The instrument is incorporated with JASCO-AS-1550-10 intelligent HPLC sampler with variable 1-100 µl injection volume loop and a cooling facility (4 C). Fifty samples can be loaded to the sampler in a standard mode operation, while eighty four samples can be loaded in micro mode. The detector used is JASCO MD-2010 Plus Intelligent UV/VIS Multiwavelength detector. It comprises of four channels of which CH1 and CH3 calculate absorbance monitoring and CH4 calculates ratio channel output. The optical arrangement includes a single beam photometric detection with a flat-field polychromator. The wavelength range is 195-650nm and the PDA has 512 channels with a noise of + 0.7 x 10-5 AU. The light sources include deuterium and tungsten lamps. A PC-based ChromNAV, version 1.12.01 chromatographic software with LC-Net II/ADC system is used for data integration. The software is configured with a calibration format where the drug concentration can be plotted against peak height or peak area or ratio of peak height or peak area of analyte to that of the internal standard to obtain the linear regression least square fit data of analysis. 6.2.2.2 Mass Spectrophotometer instrumentation Mass spectrometry analyses for the present study were done on 3 different systems. Initial LC-MS analysis was carried out MS system 1. The next MS system used was MS School of Science, SVKM s NMIMS (Deemed-to-be University) 139

system 2 and confirmatory results were obtained on MS system 3. All the three MS systems are described below. LC-MS System 1: LC-MS analysis was carried out using a Liquid Chromatograph Mass Spectrometer make of Varian Inc. USA having model 410 Prostar Binary LC with 500 MS IT PDA detector. LC-MS System 2: LC-MS analysis was carried out using a Shimadzu Liquid Chromatograph System LC- MS 2020 with Electrospray ionization (ESI) interface, and a single quadrupole analyzer. LC system was LC-20AD equipped with a PDA detector SPD-20AC, Oven CTO-20AC. The Lab Solutions software Version 5.53 SP2 from Shimadzu Pvt. Ltd. was used for data acquisition and evaluation. LC-MS System 3: LC-ESI/MS analysis was carried out using a Shimadzu liquid chromatographic system LC-MS 8040. LC system consisted of Nexsera HPLC system equipped with LC-20AD pump along with a UV detector SPD-20AC, Oven CTO-20AC which was connected to a quadrapole mass spectrophotometer. The Lab Solutions software Version5.53 SP2 from Shimadzu Pvt. Ltd. was used for data acquisition and evaluation. 6.2.3 Preparation of Working Standards and Sample Solutions Entire HPLC method development was done using two flavonol standards - Quercetin and Kaempferol. Preparation of standard solution was done using HPLC grade solvents. Working standard of Quercetin: 10mg Quercetin dissolved in methanol and made upto 10mL volume to yield a concentration of 1mg/mL. 1mL of this stock solution was taken in a 10mL volumetric flask and further diluted with 10mL School of Science, SVKM s NMIMS (Deemed-to-be University) 140

methanol to prepare a concentration of 100µg/mL. From this 100µg/mL solution, 1ml was taken and further diluted with 10mL methanol to give a concentration of 10µg/mL. This was then used for HPLC analysis. Working standard of Kaempferol: 10mg Kaempferol dissolved in methanol and made upto 10mL volume to yield a concentration of 1mg/mL. 1ml of this stock solution was taken and diluted with 10mL methanol to prepare a concentration of 100µg/mL. From this solution, 1ml was taken and further diluted with 10mL methanol to give a concentration of 10µg/mL. This was used for HPLC analysis. Both these working standards were used for optimization of HPLC conditions. Sample solutions: Isolated bands obtained from Prep-TLC were used as samples for analysis. There were four bands obtained from Prep-TLC, Prep Sample 1 (NB- 1), Prep Sample 2 (NB-2), Prep Sample 3 (NB-3) and Prep Sample 4 (NB-4). Each of them was dissolved in a specific amount of HPLC grade methanol to obtain its stock solution of 1mg/mL These stock solutions were used further with appropriate dilutions according to the requirement of the experiment. 6.2.4 RP-HPLC Method Development RP-HPLC mode of separation was carried out for the separation of isolated Prep-TLC bands. Literature review revealed the extensive use of RP-HPLC employed in the separation of phenolic/flavonoid compounds. Reverse Phase C18 columns with lengths in the range of 150mm to 250mm, having particle size of 5μm and an internal diameter of 4.6mm were found to be commonly employed for phenolic compounds. Therefore, taking this into consideration and on the basis of the available data, the column used for the School of Science, SVKM s NMIMS (Deemed-to-be University) 141

present study was Kromasil 100 RP 5C-18 with dimensions of 250mm x 4.6µm. The optimal detection wavelength for both, phenolic and flavonoid compounds were set at 280 in accordance with earlier work of Sun et al., (2007) and Burin et al., (2011). A preliminary study on the working standards was carried out, employing the isocratic mode of elution with a mobile phase comprising of Methanol: Acetonitrile: Water [31:10:59, v/v/v] reported by Zu et al., (2006). ph of the mobile phase was found to be 5.51. This method is denoted as HPLC method 1. The UV spectrum of all the peaks was recorded in methanolic sample solutions and maximum absorption was observed in the full range from 200-650nm. Different channels were set for different wavelength using ChromNav software. The working standards, Quercetin and Kaempferol were injected individually to determine the exact retention time and their elution pattern. Thus the first chromatographic conditions set in HPLC Method 1 are as shown in Table 6.1. Reproducibility of the method was checked by injecting a standard solution in replicate (n=6). School of Science, SVKM s NMIMS (Deemed-to-be University) 142

Table 6.1: HPLC conditions for Method 1 Column Mobile Phase Flow Rate Detection Wavelength Kromasil 100 RP 5C-18 [250mm x 4.6mm, 5µm] Methanol: Acetonitrile: Water [ 31:10:59,v/v/v] ph-5.51 1.0mL/min 200-600nm [Entire Scan] 280nm Loading volume 20µL Standard Concentration Run length Quercetin and Kaempferol 10µg/mL 60 min Results obtained from HPLC method 1 suggested the need for changing HPLC parameters. Therefore, instead of direct analysis of the sample solutions, initial method development was done using working standards. HPLC method 1 was modified by changing the proportion of the mobile phase so as to enable a proper elution of standard solutions. Modified HPLC conditions set for Method 2 are shown in Table 6.2. Reproducibility of the method was checked by injecting standard solution in replicate (n=6). School of Science, SVKM s NMIMS (Deemed-to-be University) 143

Table 6.2: HPLC conditions for Method 2 Column Kromasil 100 RP 5C-18 [250mm x 4.6mm, 5µm] Mobile Phase Methanol : Acetonitrile: Water [ 34:10:56: v/v/v] ph - 6.13 Flow Rate Detection Wavelength 1.0mL/min 200-650nm [Entire Scan] 280nm Loading volume 20µL Working Standards Samples Run length Quercetin, Kaempferol (10µg/mL) Prep-TLC Bands, NB-1, NB-2 (20µg/mL) 60 min On the basis of the results obtained for working standards, Prep-TLC samples were also analyzed using HPLC Method 2. However, among the 4 isolated bands, NB-2 and NB-3 was of interest due to their TLC profile as discussed in Chapter 5. Hence preliminary HPLC analysis was done for NB-2 and NB-3 using Method 2. However this method did not bring about a separation of peaks in both the preparative samples. Hence the chromatographic conditions were slightly modified. The flow rate was lowered to 0.4mL/min to allow complete resolution of the peaks. Modified HPLC conditions set for Method 3 are shown in Table 6.3. Reproducibility of the method was checked by injecting standard solution in replicate (n=6). School of Science, SVKM s NMIMS (Deemed-to-be University) 144

Table 6.3: HPLC conditions for Method 3 Column Mobile Phase Flow Rate Detection Wavelength Kromasil 100 RP 5C-18 [250mm x 4.6mm, 5µm] Methanol : Acetonitrile: Water [ 34:10:56: v/v/v] ph - 6.13 0.4mL/min 200-650nm [Entire Scan] 280nm Loading volume 20µL Working Standards Samples Run length Kaempferol (10µg/mL) Prep-TLC Bands, NB-1, NB-2 (20µg/mL) 60min Proper separation of peaks was observed for Prep-TLC bands NB-2 and NB-3. This optimized LC method was employed for Mass Spectrometry analysis for obtaining the molecular mass for each separated peak using LC-MS system 1. Mass spectrometry analysis was done by employing ESI positive ionization mode at capillary voltage of 80V. Scanning was performed from m/z 100 to 1000. However, no mass was observed for the peak under this optimized LC method. This was thought to be due to lack of ionization of peaks which plays a crucial role in mass spectral analysis. Although the optimized mobile phase was suitable for separation of samples, it was unable to cause enough ionization. Thus the lack of proper ionization was possibly on account of absence of acid in the above mentioned optimized mobile phase. Cech et al., (2001) had earlier observed that formic acid present in mobile phase helps to improve the ionization capacity by enhancing protonation. School of Science, SVKM s NMIMS (Deemed-to-be University) 145

The mobile phase was therefore modified by the introduction of 0.01% formic acid. Another modification involved lowering of the flow rate. A low flow rate is basically employed for increasing the time for ionization. HPLC conditions set for Method 4 are shown in Table 6.4. Table 6.4: HPLC conditions for Method 4 Column Mobile Phase Flow Rate Kromasil 100 RP 5C-18 [250mm X4.6mm] Methanol : 0.01% formic acid in Acetonitrile : 0.01% formic acid in Water [ 34:10:56: v/v/v] ph 3.56 0.2mL/min Detection Wavelength 200-600nm [Channel 9] Loading volume 20µL Sample name Run length Standard Kaempferol [15µg/mL] Prep NB-1 [20µg/mL] Prep NB-2 [20µg/mL] Prep NB-3 [20µg/mL] Prep NB-4 [20µg/mL] 100min Proper separation of peaks was observed under the above set conditions for all the Prep- TLC samples. These HPLC optimized parameters were then used for further LC-MS analysis for the identification of each peak of interest from Prep sample NB-2 and Prep sample NB-3. School of Science, SVKM s NMIMS (Deemed-to-be University) 146

6.2.5 Method Development on Mass Spectrometry Preliminary mass spectrometry analysis was done using LC-MS system 2. Further confirmatory study was carried out on LC-MS system 3. 6.2.5.1 Preliminary MS analysis of Kaempferol, Prep NB-2 and Prep NB-3 Preliminary MS analysis was undertaken by employing Electrospray Ionization [ESI] in both positive and negative mode for full scan using MS system 2 (Tsimogiannis et al., 2007). HPLC conditions were same as described in Table 6.4. MS system 2 used for this study was a Shimadzu liquid chromatographic system LC-MS 2020. Conditions set on MS system 2 are shown in Table 6.5. The Lab Solutions software version 5.53 SP2 from Shimadzu Pvt. Ltd. was used for data acquisition and evaluation. Table 6.5: MS conditions set on MS system 2 Parameters Interface Voltage Working Conditions 4.50kV Desolvation Line (DL) temperature 250 C Heat block temperature 400 C Nebulizing gas flow Drying gas flow 2.0 L/min 15.0 L/min Scanning range m/z 100-1000 School of Science, SVKM s NMIMS (Deemed-to-be University) 147

6.2.5.2 LC-MS/MS analysis of Pre NB-2 and NB-3 On the basis of results obtained from these preliminary studies, further fragmentation of the parent ion was done to get the product ion scan. MS system 3 used for this study was a HPLC system Nexsera coupled to Shimadzu Liquid Chromatographic System LC-MS 8040. Positive ionization mode was used to produce fragments of parent molecule. Conditions set on MS system 3 are as tabulated in Table 6.6. Data acquisition and interpretation was done with the Lab Solutions software version 5.53 SP2 from Shimadzu Pvt. Ltd. Table 6.6: MS conditions set on MS system 3 Parameters Interface Voltage Working Conditions 4.50kV Desolvation Line (DL) temperature 250 C Heat block temperature 400 C Nebulizing gas flow Drying gas flow 2.0 L/min 15.0 L/min Scanning range m/z 100-1000 Collision energy -65.0V Compounds were identified using the data obtained for their absorbance spectra (λ max ), parent mass and fragmentation patterns by comparing with available literature of standard compounds. School of Science, SVKM s NMIMS (Deemed-to-be University) 148

6.3 Results and Discussion 6.3.1 HPLC method development Preliminary method development on HPLC was done using standard solutions of Quercetin and Kaempferol, each having a concentration of 10μg/mL. The different chromatographic conditions employed for method development have already been discussed under Section 6.2.4. In HPLC Method 1, both the working standards Quercetin and Kaempferol eluted slowly with higher retention time. Figure 6.2 and Figure 6.3 shows the HPLC chromatogram observed for Quercetin and Kaempferol, respectively. Table 6.7 and 6.8 shows the reproducibility results obtained for Quercetin and Kaempferol under HPLC Method 1. School of Science, SVKM s NMIMS (Deemed-to-be University) 149

Figure 6.2: HPLC Chromatogram of Standard Quercetin under chromatographic conditions set for HPLC Method 1 Table 6.7: Reproducibility results of Quercetin for HPLC method 1 Observation no. t R (min) Peak Area 1 21.72 768936 2 21.70 775723 3 21.72 791752 4 21.74 764521 5 21.71 773694 6 21.77 786142 Mean 21.72 776794 S.D. 0.023 10336 % C.V 0.108 1.330 School of Science, SVKM s NMIMS (Deemed-to-be University) 150

Figure 6.3: HPLC Chromatogram of Standard Kaempferol under chromatographic conditions set for HPLC Method 1 Table 6.8: Reproducibility results of Kaempferol for HPLC Method 1 Observation no. t R (min) Peak Area 1 35.20 1914662 2 35.20 1932540 3 35.18 1986890 4 35.13 1914289 5 35.21 1984075 6 35.11 1955614 Mean 35.17 1948012 S.D. 0.042 32737 % C.V 0.119 1.680 School of Science, SVKM s NMIMS (Deemed-to-be University) 151

Since the % C.V was less than 2% for the retention time as well as peak area, system suitability test passed the criteria of acceptance. Quercetin eluted with a retention time of 22 min whereas Kaempferol eluted later with a retention time of 35 min. Separation of polyphenolic compounds in reverse phase depends on the polarity of the compounds. Thus it is easy to predict the elution order of polyphenolic compounds in RP-HPLC. The elution order for separation of compounds is determined by their structural differences. Quercetin and Kaempferol are flavonol group of compounds with the same substitution pattern in the A flavone ring. Consequently, substitution in the B ring plays a major role in determining their elution order (Kulevanova et al., 2002). Quercetin contains two OH-groups at 3 and 4 position in the B-ring, whereas Kaempferol differs by having only one OH group in the 4 position. Owing to these structural differences, retention of both these compounds is affected by the number of OH-groups attached to the B-ring. Thus RP-HPLC follows the order of elution in compounds such that more polar (OH) groups are observed to have less retention on the reversed phase (non-polar), thereby eluting faster. Thus they show shorter retention time. Therefore, in a reverse phase system, the sample elution is retained for more hydrophobic compounds (Merken et al., 2000 and Kulevanova et al., 2002) Results obtained for HPLC Method 1 suggested the need for method development; hence the mobile phase composition was altered so as to bring about an early elution of the standard solutions. A study conducted by Snyder et al., (1997) indicated that the solvent system of the mobile phase has an effect on sample retention. Flavonoids are polyphenolic substances having three phenolic rings which results in stronger hydrophobic interaction with the stationary phase thereby resulting in longer retention. School of Science, SVKM s NMIMS (Deemed-to-be University) 152

Thus as a function of their molecular structure, greater amount of methanol is needed for elution of flavonoid compounds. Hence the mobile phase composition was slightly modified by changing the content of the organic solvent. Methanol content in the mobile phase was slightly increased. Figures 6.4 and 6.5 represent the chromatograms obtained for standard Quercetin and Kaempferol respectively under HPLC Method 2. Results of reproducibility for HPLC Method 2 are shown for Quercetin and Kaempferol in Tables 6.9 and 6.10, respectively. Figure 6.4: HPLC Chromatogram of Standard Quercetin under chromatographic conditions set for HPLC Method 2 School of Science, SVKM s NMIMS (Deemed-to-be University) 153

Table 6.9: HPLC Method 2: Reproducibility results of Quercetin Observation no. t R (min) Peak Area 1 16.20 735955 2 16.21 745108 3 16.21 761647 4 16.18 766480 5 16.23 746187 6 16.25 751164 Mean 16.21 751090 S.D. 0.024 11288.930 % C.V 0.149 1.503 Figure 6.5: HPLC Chromatogram of Standard Kaempferol under chromatographic conditions set for HPLC Method 2 School of Science, SVKM s NMIMS (Deemed-to-be University) 154

Table 6.10: HPLC Method 2: Reproducibility results of Kaempferol Observation no. t R (min) Peak Area 1 19.94 1235728 2 19.94 1262580 3 19.93 1238637 4 19.90 1248361 5 19.97 1251075 6 19.91 1236427 Mean 19.93 1245468 S.D. 0.025 10544.176 % C.V 0.130 0.846 With an increase in the methanol content, the retention time of Quercetin and Kaempferol changed. Retention time of Quercetin and Kaempferol decreased to 16.20 min and 19.94 min, respectively. This could be attributed to the increase in the polarity of mobile phase ultimately resulting in faster elution of retained compounds. Since the % C.V was less than 2% for retention time as well as peak area, system suitability test passed the criteria of acceptance. The same mobile phase composition was employed for analysis of preparative samples. On the basis of the TLC profile obtained for the four preparative bands, NB-2 and NB-3 were selected as discussed earlier. Figures 6.6 and 6.7 shows the chromatograms obtained for Prep NB-2 and Prep NB-3, respectively under HPLC Method 2. School of Science, SVKM s NMIMS (Deemed-to-be University) 155

Figure 6.6: HPLC Chromatogram of Prep NB-2 under chromatographic conditions set for HPLC Method 2 Figure 6.7: HPLC Chromatogram of Prep NB-3 under chromatographic conditions set for HPLC Method 2 School of Science, SVKM s NMIMS (Deemed-to-be University) 156

As observed in the above chromatograms, peaks were obtained in Prep NB-2 and NB-3. However, no proper separation of peaks was observed for both the samples. Hence, in order to achieve a proper separation of peaks, the flow rate was lowered to 0.4mL/min (HPLC method 3). Figure 6.8 shows a chromatogram obtained for Kaempferol under HPLC Method 3. Reproducibility data obtained for standard Kaempferol (n=6) under HPLC method 3 are shown in Table 6.11. Under these conditions, separation was observed for the peaks in NB-2 and NB-3, results of which are depicted in Figures 6.9 and 6.10 for NB-2 and NB-3 respectively. Reproducibility of the peaks for NB-2 and NB-3 was studied by analyzing sample in triplicate. These results are shown in Table 6.12. Figure 6.8: HPLC Chromatogram of Kaempferol under chromatographic conditions set for HPLC Method 3 School of Science, SVKM s NMIMS (Deemed-to-be University) 157

Table 6.11: Reproducibility results of HPLC Method 3 Observation no. Kaempferol t R (min) Peak Area 1 47.6 1337911 2 47.57 1359716 3 47.57 1379064 4 47.60 1407427 5 47.57 1376323 6 47.57 1352942 Mean 47.58 1368897 S.D. 0.014 24257.2 %CV 0.030 1.772 Figure 6.9: HPLC Chromatogram of NB-2 under chromatographic conditions set for HPLC Method 3 School of Science, SVKM s NMIMS (Deemed-to-be University) 158

Figure 6.10: HPLC Chromatogram of NB-3 under chromatographic conditions set for HPLC Method 3 Table 6.12: Reproducibility results of NB-2 and NB-3 for HPLC method 3 Observation no. NB-2 NB-3 t R (min) Peak Area t R (min) Peak Area 1 6.47 1239034 9.73 20546048 2 6.47 1252102 9.73 21017041 3 6.47 1223777 9.73 20789417 Mean 6.47 1238304 9.73 235890.9 S.D. 0.001 14176.59 0.001 1.1346 %RSD 0.017 1.144 0.011 1.134 Thus the HPLC conditions as given in HPLC Method 3 were optimized for separation of peaks of interest. School of Science, SVKM s NMIMS (Deemed-to-be University) 159

6.3.2 Preliminary MS analysis of Prep sample Electrospray Ionization (ESI) was employed for preliminary mass analysis employing both, the positive and negative mode. Prep NB-3 was analyzed using ESI-MS (positive ionization) under HPLC method 3. Figure 6.11 shows the ESI-MS spectrum observed for NB-3. Figure 6.11: ESI MS spectra for NB-3 under positive ionization As seen in above figure, there was no significant ionization observed at expected retention time. As discussed in earlier section of this chapter, this was thought to be due to improper ionization of peaks. Thus the optimized mobile phase was not suitable for causing ionization of molecule. This could have been due to lack of acid in mobile phase. Presence of formic acid in mobile phase has been reported to improve its ionization capacity by enhancing protonation (Cech et al., 2001) School of Science, SVKM s NMIMS (Deemed-to-be University) 160

The mobile phase was thus modified by introduction of 0.01% formic acid and flow rate was lowered to 0.2mL/min to provide more time for proper ionization. All the preparative samples were properly separated using above conditions. Figure 6.12 shows the chromatogram for Kaempferol under HPLC Method 4 and the reproducibility results for the same are shown in Table 6.13. Figures 6.13, 6.14, 6.15 and 6.16 show the chromatograms of NB-1, NB-2, NB-3 and NB-4 under HPLC Method 4. Reproducibility results for all prep samples are shown in Table 6.14. Figure 6.12: HPLC Chromatogram of Kaempferol under chromatographic conditions set for HPLC Method 4 School of Science, SVKM s NMIMS (Deemed-to-be University) 161

Table 6.13: Reproducibility results of Kaempferol for HPLC Method 4 Observation no. t R (min) Peak Area 1 85.96 2316873 2 85.93 2333868 3 85.89 2322535 4 85.88 2263640 5 85.90 2316705 6 85.90 2236010 Mean 85.91 2298271.833 S.D. 0.028 39031.969 % C.V 0.033 1.698 Figure 6.13: HPLC Chromatogram of NB-1 under chromatographic conditions set for HPLC Method 4 School of Science, SVKM s NMIMS (Deemed-to-be University) 162

Figure 6.14: HPLC Chromatogram of NB-2 under chromatographic conditions set for HPLC Method 4 Figure 6.15: HPLC Chromatogram of NB-3 under chromatographic conditions set for HPLC Method 4 School of Science, SVKM s NMIMS (Deemed-to-be University) 163

Figure 6.16: HPLC Chromatogram of NB-4 under chromatographic conditions set for HPLC Method 4 Peak purity Under this optimized HPLC conditions (Method 4), peak purity was checked using PDA for the standard Kaempferol as well as for the peaks of interest. The peak purity for kaempferol was found to be 98.9% whereas for NB-2 and NB-3 the values were 98.1% and 99.8% respectively. School of Science, SVKM s NMIMS (Deemed-to-be University) 164

Table 6.14: Reproducibility results of NB-1, NB-2, NB-3 and NB-4 for HPLC Method 4 Reading no. t R NB-1 NB-2 NB-3 NB-4 Peak t R (min) Area (min) Peak Area t R (min) Peak Area t R Peak (min) Area 1 21.37 818731 29.84 15426821 47.80 15407198 26.5 565039 2 21.37 822814 29.837 15737988 47.81 15656526 26.515 575807 3 21.36 808185 29.895 15428057 47.81 15941126 26.49 579829 Mean 21.37 816576 29.857 15530955 47.808 15668583 26.501 573558 S.D. 0.004 7548.694 0.032 179296.6 0.007 267158.1 0.012 7647.117 %RSD 0.018 0.924 0.108 1.154 0.015 1.705 0.047 1.333 % RSD observed was less than 2% which was within the acceptance criteria and thus the HPLC method also gave reproducible results. 6.3.3 Tentative identification of Preparative samples from HPLC Elution Profile and UV spectra For all the four preparative samples, tentative identification of compounds was done on the basis of their elution pattern on HPLC and their spectra obtained through PDA. Absorption maxima for each peak were compared with available literature for standard compounds. Figures 6.17 and 6.18, illustrate UV spectra obtained for Standard Kaempferol and Quercetin, respectively. School of Science, SVKM s NMIMS (Deemed-to-be University) 165

266nm 367nm Figure 6.17: UV spectra of Standard Kaempferol 253nm 370nm Figure 6.18: UV spectra of Standard Quercetin Figure 6.19 (a) and (b) shows the HPLC profile of NB-1 and its UV spectrum respectively. Figure 6.20 (a) and (b) shows the HPLC profile of NB-2 and its UV spectrum respectively. Figure 6.21 (a) and (b) shows the HPLC profile of NB-3 and its UV spectrum respectively. Figure 6.22 (a) and (b) shows the HPLC profile of NB-4 and its UV spectrum respectively. School of Science, SVKM s NMIMS (Deemed-to-be University) 166

(a) (b) Figure 6.19: (a) HPLC profile of NB-1 and (b) UV spectrum of peak (tr = 21 min) School of Science, SVKM s NMIMS (Deemed-to-be University) 167

(a) Figure 6.20: (a) HPLC profile of NB-2 and (b) UV spectrum of peak (tr = 29 min) (b) School of Science, SVKM s NMIMS (Deemed-to-be University) 168

(a) (b) Figure 6.21: (a) HPLC profile of NB-3 and (b) UV spectrum of peak (tr = 47 min) School of Science, SVKM s NMIMS (Deemed-to-be University) 169

(a) (b) Figure 6.22: (a) HPLC profile of NB-4 and (b) UV spectrum of peak (tr = 26 min) School of Science, SVKM s NMIMS (Deemed-to-be University) 170

6.3.4 Spectral analysis of NB-1 and NB-4: Spectral data obtained for NB-1 and NB-4 showed two major absorbance maxima at 245 and 325 nm and a shoulder peak at 294nm. This spectral characteristic is similar to that observed for phenolic acids (Lin and Harnly, 2010). In addition, NB-1 (t R = 21 min) and NB-4 (t R = 26 min) eluted much earlier as compared to standard Kaempferol (t R = 85 min). These elution patterns for phenolic acid and flavonol aglycones were in agreement with those observed earlier by Jianping et al., (2007) in their study on red wines wherein phenolic acids eluted much faster than flavonoids. This could be attributed to more polar nature of phenolic acids than flavonoids. Basically the structure of flavonol compounds contains a 4-keto functional group which increases the hydrophobicity of the molecule by forming a planar non-polar six-member ring; making it less polar to the solvent. Moreover, the unsaturation between positions 2 and 3 of the pyran ring on the flavonoid molecule is also known to make the compound much less polar due to a larger electron density on the oxygen atom of the 4-keto group thus explaining the higher retention time encountered for flavonols i.e. Quercetin and Kaempferol (Wulf and Nagel, 1976). Thus, on the basis of their TLC profile, HPLC elution pattern and UV spectra, NB-1 and NB-4 were tentatively identified as phenolic acids. However, further identification by MS was not carried out for these samples as main focus of the study was flavonoid group of compounds. 6.3.5 Spectral analysis of NB-2 and NB-3 Spectral data obtained for NB-2 were similar to that of Quercetin when compared with the standard compound. However, retention time observed was completely different for both. There was a possibility of this peak being glycosidic form of Quercetin as glycosylation is known to affect the elution pattern. These results are in agreement with those of Wulf and Nagel, (1976) who earlier reported that flavonoid glycosides are more School of Science, SVKM s NMIMS (Deemed-to-be University) 171

polar than flavonoid aglycones and that glycosylation results in faster elution of compounds. Quercetin glycoside eluted faster than its aglycone form and this could be due to increase in polarity of molecule by addition of sugar moiety. Spectral data obtained for NB-3 was similar to that of standard Kaempferol. However, similar effect was observed for NB-3 as observed for NB-2. Retention time of standard Kaempferol was high as compared to that of NB-3. Therefore, there was possibility of NB-3 to be present in glycosidic form of Kaempferol. The observed results were in accordance to remarks made by Spanos et al., (1990) who reported a change in the retention time of an aglycone (free molecule) esterified with a carboxylic acid or with a sugar moiety. This variation in retention time would be due to the difference in the polarity of the molecule. Therefore on the basis of this literature, it was more crucial to obtain MS data to ensure the identification of peaks as many different compounds belonging to the same family show similar scan spectra. Thus Prep NB-2 and NB-3 were subjected to MS/MS analysis for their identification. 6.3.6 Preliminary LC-MS analysis of Kaempferol, Prep NB-2 and Prep NB-3 Initial LC-ESI-MS data for standard Kaempferol, Prep NB-2 and Prep NB-3 was obtained using Shimadzu Liquid Chromatographic System LC-MS 2020. ESI-MS analysis was done using both positive and negative ionization mode. ESI-MS analysis of standard Kaempferol produced a parent ion [M-H] + at m/z 287 in positive mode and [M-H] - at m/z 285 in negative mode. Figures 6.23 (a) and (b) shows ESI-MS spectra obtained for Kaempferol in the positive and negative ionization modes respectively while Figures 6.24 (a), (b) and 6.25 (a), (b) show ESI-MS spectra obtained for NB-2 and NB-3 in the positive and negative ionization modes respectively. School of Science, SVKM s NMIMS (Deemed-to-be University) 172

Figure 6.23 (a): ESI-MS spectrum of Standard Kaempferol in Positive ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 173

Figure 6.23 (b): ESI-MS spectrum of Standard Kaempferol in Negative ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 174

Figure 6.24 (a): ESI-MS spectrum of NB-2 in Positive ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 175

Figure 6.24 (b): ESI-MS spectrum of NB-2 in Negative ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 176

Figure 6.25 (a): ESI-MS spectrum of NB-3 in Positive ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 177

Figure 6.25 (b): ESI-MS spectrum of NB-3 in Negative ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 178

6.3.7 LC-ESI-MS analysis of NB 2 LC-ESI-MS analysis of NB-2 produced a parent ion [M-H] + at m/z 465 in the positive ionization mode and [M-H] - at m/z 463 in the negative ionization mode. Hence NB-2 had a molecular weight of 464 which was same as that reported by Lin and Harnly, (2007) for Quercetin glycoside. The possibility of this compound being a quercetin glycoside was justified by its orange colour which developed in TLC and which is characteristic of Quercetin and its derivatives. UV spectra observed matched with the reported UV spectra for Quercetin glycoside (Lin and Harnly, 2007). 6.5.8 LC- ESI-MS analysis of NB-3 LC- ESI-MS analysis of NB-3 produced a parent ion [M-H] + at m/z 449 in positive ionization mode and [M-H] - at m/z 447 in negative ionization mode. Thus NB-3 had a molecular weight of 448 which on comparison with available literature was same as that reported for glycoside of Kaempferol, Quercetin and Luteolin (De Brito et al 2007; Lin and Harnly, 2007, 2010; Wan et al., 2011). However among the speculated compounds, TLC profile of NB-3 did not match with Quercetin glycoside. Moreover, comparison of UV spectra of NB-3 with luteolin showed slight variations thereby ruling out the possibility of luteolin. However, further fragmentation of aglycones was done to clarify difference in flavonol and flavone structure. 6.3.8 LC-MS/MS analysis of Prep NB-2 and Prep NB-3 The fragmentation pattern of flavonoid compounds serves as a highly diagnostic tool for identifying specific classes of compounds. This makes it possible to obtain a precise structural elucidation of the aglycone by comparison with the reported data (Fabre et al. 2001; Cuyckens et al., 2001) School of Science, SVKM s NMIMS (Deemed-to-be University) 179

The system used for obtaining the fragmentation pattern of NB-2 and NB-3 was Shimadzu LC-MS 8040. The fragmentation pattern was obtained using both the ionization modes for the precursor molecule so as to identify the aglycone moiety. However, the negative ionization mode was unable to produce characteristic fragments for each flavonoid subgroup which may be deduced from the earlier data published by Fabre et al., (2001); Hughes et al., (2001); Wu et al.,(2004) in their studies on flavonoids groups. Figure 6.26 shows the product ion scan obtained for m/z 465 in the positive ionization mode. School of Science, SVKM s NMIMS (Deemed-to-be University) 180

Figure 6.26: ESI-MS /MS spectrum of m/z 465 in Positive ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 181

As observed above in the ESI-MS/MS spectrum, the product ion scan of m/z 465, generated a ion with m/z 303 indicating the loss of 162 amu from the pseudomolecular ion [M H] + (m/z 465). This loss of 162 indicated attachment of the sugar moiety (hexose) to the aglycone part indicating that it was present in glycosidic form and not as an aglycone. The remaining aglycone part with m/z 303 indicated it to be quercetin as observed earlier by Tsimogiannis et al., (2007). Figure 6.27 explains the loss of the sugar moiety taking place from Quercetin glycoside while Figure 6.28 explains the Cleavage of Quercetin aglycone at different positions in the C-ring. In addition, Quercetin aglycone with m/z 303 produced characteristic fragments for the flavonol group. The characteristic fragments observed with m/z 257, 275, 165 and 153 were similar to the results reported by Tsimogiannis et al., (2007) in their study on phenolic compounds. They reported similar characteristic fragments for standard Quercetin in their study on the fragmentation pattern of twelve selected flavonoids. A product ion scan of quercetin aglycone m/z 303 showed the formation of an ion with m/z 257 which could be attributed to loss of water and carbon monoxide by forming [M+H-H 2 O-CO] +. Moreover, an ion was obtained at m/z 275 indicating the direct loss of carbon monoxide from the protonated aglycone through the formation of [M+H-CO] +. School of Science, SVKM s NMIMS (Deemed-to-be University) 182

H + [Aglc-H] + m/z 303 Sugar moiety m/z 162 Figure 6.27 : Cleavage of glycosidic bond to yield quercetin aglycone and the sugar moiety School of Science, SVKM s NMIMS (Deemed-to-be University) 183

[M-H] + (m/z = 303) 1,3 B + H + 0,2 A + B A C 0,2 B + 1,3 A + 0,2 A + = 165 1,3 A + = 153 Figure 6.28: Cleavage of Quercetin aglycone at different positions in the C-ring School of Science, SVKM s NMIMS (Deemed-to-be University) 184

As shown in the above Figure 6.28, formation of fragments with m/z 153 and 165 were due to C-ring cleavage of the aglycone structure. As noted earlier by Cuyckens & Claeys, (2004) and De Rijke et al., (2006), formation of fragment 153 was likely to be due to the cleavage between bond 1 and 3 which is most commonly observed and constitutes abundant fragment ion for the flavonol group. Moreover, presence of 0,2 A + fragment with m/z 165 has been exclusively assigned to the group of flavonols containing an additional hydroxyl group at 3 position which is present in the structure of Quercetin. Thus NB-2 was identified as a Quercetin glycoside on the basis of its UV spectrum, parent mass and fragmentation pattern. Figure 6.29 shows the product ion scan obtained for m/z 449 in the positive ionization mode. School of Science, SVKM s NMIMS (Deemed-to-be University) 185

Figure 6.29: ESI-MS /MS spectrum of m/z 449 in NB-3 in Positive ionization mode School of Science, SVKM s NMIMS (Deemed-to-be University) 186

As observed above, the product ion scan of m/z 449, generated an ion with m/z 287 indicating the the loss of 162 amu from the pseudomolecular ion [M H] + (m/z 449). This loss of 162 indicated attachment of the sugar moiety (hexose) to the aglycone part. The remaining aglycone part with m/z 287 indicated it to be kaempferol aglycone as noted by Tsimogiannis et al., (2007). Figure 6.30 explains the loss of the sugar moiety taking place from Kaempferol glycoside while Figure 6.31 explains the Cleavage of Kaempferol aglycone at different positions in the C-ring. Kaempferol aglycone with m/z 287 produced characteristic fragments for the flavonol group. Fragments with m/z 213,165,153 and 121 were similar to those reported by Tsimogiannis et al., (2007). They recorded similar characteristic fragments for standard Kaempferol. Among the fragments, formation of an ion with m/z 213 could be attributed to the loss of water and carbon monoxide through the formation of [M+H-H 2 O-CO] +. School of Science, SVKM s NMIMS (Deemed-to-be University) 187

[Aglc-H] + m/z 287 Sugar moiety m/z 162 Figure 6.30: Cleavage of glycosidic bond to form kaempferol aglycone and sugar moiety School of Science, SVKM s NMIMS (Deemed-to-be University) 188

[M-H] + (m/z = 287) 0,2 A + = 165 1,3 A + = 153 0,2 B + = 121 Figure 6.31 : Cleavage of Kaempferol aglycone at different positions in the C-ring School of Science, SVKM s NMIMS (Deemed-to-be University) 189

As observed above, fragment with m/z 153 and 121 indicated a cleavage between bond 1 and 3 which represents the common fragments observed for flavonoids. The fragment with m/z 153 contained the A ring whereas the fragment with m/z 121 contained the B- ring. Moreover, presence of 0,2 A + with m/z 165 confirmed that the compound belonged to the flavonol group and which is exclusively used to distinguish flavonols from other subgroups. This ion does not occur in the spectra of the other subclasses of flavonoids (De Rijke et al., 2006). This difference in formation of 165 mass at the cleavage between 0 and 2 bonds could be attributed to the presence of an additional 3-OH group which is specific to the group of flavonols. Hence, the possibility of the compound being luteolin was ruled out as it did not show the presence of fragment of m/z 165. A review paper by Pinheiro and Justino, (2010) provides supportive evidence for this fact. They also reported the presence of 0,2 A + fragment to be exclusively assigned to the group of flavonols containing an additional hydroxyl group at position 3. Thus NB-3 was identified as a Kaempferol glycoside on the basis of its UV spectrum, parent mass and fragmentation pattern. School of Science, SVKM s NMIMS (Deemed-to-be University) 190

6.4 Conclusion Important conclusions drawn from the results obtained in HPLC method development and Mass spectrometry analysis are as follows: RP-HPLC-DAD method was successfully developed for separation of preparative samples, isolated from Nyctanthes arbor-tristis L. On the basis of the results obtained from TLC profile, HPLC elution pattern and UV spectra, NB-1 and NB-4 belong to phenolic acids whereas NB-2 and NB-3 were identified as flavonols. Under optimized HPLC conditions there were differences in the HPLC elution patterns for phenolic acids and flavonoids with phenolic acids eluting faster than flavonoids. The peaks of interest showed high purity values with NB-2 having a value of 98.1% and NB-3 with a value of 99.8%. Electrospray Ionization (ESI) with positive and negative mode of ionization provided useful data of molecular mass of the parent molecule as well for the characteristic fragmentation pattern. NB-2 was found to be a Quercetin glycoside while NB-3 was a Kaempferol glycoside. School of Science, SVKM s NMIMS (Deemed-to-be University) 191