- Email: [email protected]
Design and operation of a modified silica gel column chromatography
Available online at www.sciencedirect.com
Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633 www.elsevier.com/locate/jcice
Design and operation of a modified silica gel column chromatography Setiyo Gunawan, Suryadi Ismadji, Yi-Hsu Ju * Department of Chemical Engineering, National Taiwan University of Science and Technology, 43 Keelung Road, Section 4, Taipei 106-07, Taiwan Received 5 February 2008; received in revised form 20 June 2008; accepted 25 June 2008
Abstract Liquid–solid chromatography (LSC) is the oldest of the various liquid chromatography methods. Despite the fact that high-performance liquid chromatography (HPLC) operation leads to better separation and analysis, classical column chromatography and thin-layer chromatography (TLC) are still widely practiced because of their convenience. In this study, a modified silica gel column chromatography was designed with the objective of reducing the amount of solvent required to achieve the same degree of separation as the classical silica gel column chromatography. The separation of squalene and fatty acid steryl esters (FASEs) from non-polar lipid fraction (NPLF) of soybean oil deodorizer distillate (SODD) was employed as a model system to test the effectiveness of this new design. Modified silica gel column chromatography process is feasible from economic point of view compare to classical silica gel column chromatography because it significantly reduces the amount of solvent and time required to achieve the same degree of separation. By employing modified silica gel column chromatography to obtain the squalene-rich fraction, the mobile phase volume and elution time required as fractions of those needed in classical silica gel column chromatography are 1/73 and 1/18, respectively. To obtain the FASEs-rich fraction, the corresponding mobile phase volume and elution time are 1/221 and 1/23, respectively of those needed in classical silica gel column chromatography. # 2008 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved. Keywords: Fatty acid steryl esters; Modified silica gel column chromatography; Soybean oil deodorizer distillate; Squalene
1. Introduction Liquid–solid chromatography (LSC) or adsorption is the oldest of the various liquid chromatography methods. Despite the fact that high-performance liquid chromatography (HPLC) operation leads to better separation and analysis, classical column chromatography and thin-layer chromatography are still widely practiced because of their convenience. The successful use of liquid chromatography for a given problem requires the right combination of operating conditions: the type of column packing and mobile phase, the length and diameter of the column, mobile phase flow rate, elution temperature, and sample size (Snyder and Kirkland, 1979). Temperature is normally not used as separation variable in LSC. Most applications are carried out at ambient temperature. A classical silica gel column chromatography was used to fractionate hydrocarbons from the unsaponifiable matter of both virgin and refined olive oil, using hexane as the eluent (Lanzo´n et al., 1994). Also, a classical silica gel column chromatography
* Corresponding author. Tel.: +886 2 27376612; fax: +886 2 27376644. E-mail address: [email protected] (Y.-H. Ju).
was used to fractionate hydrocarbons and fatty acid steryl esters (FASEs) from the non-polar lipid fraction (NPLF) obtained from crude rice bran oil (Gunawan et al., 2006) and soybean oil deodorizer distillate (SODD) (Gunawan et al., 2008). When the NPLF of SODD was introduced into a classical silica gel column chromatography to isolate squalene, squalene (95.90% purity and 93.09% recovery) was obtained in the 2nd fraction after eluting the column with 10.96 L hexane. The continued elution of column with 33.13 L hexane yielded a 3rd fraction rich in FASEs (84.04% purity and 63.17% recovery). In another study, the separation of squalene from the unsaponifiable matter of olive oil deodorizer distillate was achieved by thin-layer chromatography (TLC), using mixture of solvent as mobile phase (Bondioli et al., 1993). Classical column chromatography and TLC are laborious, time consuming (Moreda et al., 2001), and requiring the use of large amount of organic solvents (Lacaze et al., 2007). A disadvantage of TLC compared to classical column chromatography is the limited possibility of understanding quantitative analysis since exhaustive recovery of the separate spot is not sufficiently precise (Moreda et al., 2001). The advantages over classical column chromatography are that the procedure is more rapid, and the possibility of using specific reagents to immediately reveal the nature of the separated compound. Solid-phase extraction (SPE)
0368-1653/$ – see front matter # 2008 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jcice.2008.06.006
626
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Nomenclature AOCS EI FASEs FFAs GCMS h HPLC HTGC k0 LSC n NIST NPLF p S SODD SPE TLC x x¯
American oil chemists’ society electron impact fatty acid steryl esters free fatty acids gas chromatography mass spectrometry hour high-performance liquid chromatography high-temperature gas chromatography equilibrium distribution factor liquid–solid chromatography total number of experiment national institute of standards and technology non-polar lipid fraction probability standard deviation of the measures soybean oil deodorizer distillate solid-phase extraction thin-layer chromatography value of individual experiment mean value of three independent experiments
Greek symbol a significance level has been developed to isolate squalene from virgin olive oil without any chemical treatment. Rapidity, reliability, use of minimal amounts of solvent and automation are the main advantages of SPE in comparison to TLC and classical column chromatography (Giacometti et al., 2002). All the methods mentioned above are limited by small sample loading and using mixture of organic solvent in order to obtain good separations. In this study, a new silica gel column chromatography was designed with the objective of reducing the amount of solvent required to achieve the same degree of separation as the classical silica gel column chromatography. The isolation of squalene and FASEs from the NPLF of SODD was employed as a model system to test the effectiveness of this new design. The effects of parameters, such as NPLF to silica gel mass ratio, flow rate and composition of mobile phase, and elution temperature, on the separation were systematically investigated. 2. Materials and methods 2.1. Chemicals and materials Standard nonacosane, farnesene, cholesta-3,5-diene, squalene, fatty acids, a-, d-, g-tocopherol, monooleylglycerol, diolein, triolein, and tripalmitin were obtained from Sigma Chemicals Company (St. Louis, MO). Standard b-sitosterol (practical grade) was obtained from MP Biomedicals, LLC (Aurora, OH). All solvents and reagents were either of HPLC grade or analytical reagent grade and were obtained from commercial sources. SODD was donated by TTET Union Corporation (Tainan, Taiwan). TLC aluminum plates
(20 cm 20 cm 250 mm) were purchased from Merck (Darmstadt, Germany). Silica Gel (70–230 mesh) was obtained from Silicycle (Quebec, Canada). Characteristics of the gel according to the manufacturer were: particle size: 60–200 mm; ˚ ; pH: 7; water content: 6%; and specific surface pore size: 60 A 2 area: 500 m /g. 2.2. Extraction of NPLF from SODD Solid suspension was removed from SODD at 40 8C by using a 7 mm Advantec filter paper (Toyo Roshi Kaisha Ltd., Tokyo, Japan). SODD was applied to modified soxhlet extraction, under the following operation conditions: solvent = n-hexane, SODD to silica gel mass ratio = 1:3 (w/w), extraction temperature = 6 8C, and extraction time = 11 h as described by Gunawan et al. (2008). The resulting hexane extractive, named NPLF, was used for this study. 2.3. Modified silica gel column chromatography A modified version of the classical silica gel column chromatography (300 mm 4 mm i.d. glass tube) equipped with a jacket, a valve to control the flow rate of eluent and a condenser system was employed in this study. Ethyl alcohol was used as the refrigerant. This was circulated in the jacket to control the column (packing region) temperature. A schematic drawing of the modified silica gel column chromatography is shown in Fig. 1. The siphon arm for operation in soxhlet extraction mode was closed by three-way valve. Silica gel (18–60 g) was kept in a furnace at 150 8C for 1 h to remove its water content. Next, a slurry of silica gel in hexane was poured into the column that was previously half-filled with hexane. The exit of the column was plugged with 1 g cotton to retain the silica gel and a thermo couple was inserted above the cotton. The hexane was allowed to drain slightly during packing. The top surface of the silica gel was covered by 1 g cotton, and three thermocouples were set inside the elution chamber at different locations. The hexane level was lowered until at controlled level (50 and 100 mL above the cotton at the top). NPLF (3 g) was put into the column at room temperature (23 1 8C). The column was eluted with mobile phase (50 mL hexane), which was put into a 500 mL roundbottom flask at the start of the run and was heated. The mobile phase vapor traveled up the distillation arm, and flooded into the column housing the silica gel and the NPLF. The condenser ensured that any mobile phase vapor condenses and drips back down into the column. The column was slowly eluted with the mobile phase at controlled flow rate and temperature. After predetermined time, the flask that contained the desired extract was removed and replaced immediately by another flask that contained 50 mL fresh mobile phase so that the total amount of mobile phase remained the same as in the beginning of the run. After removing the mobile phase, each fraction was analyzed by TLC and high-temperature gas chromatography (HTGC). The flow rate of mobile phase vapor depends on the heat input from the heater and was predetermined (data not shown). Fig. 2 shows the flow chart of separation of squalene and FASEs from the NPLF of SODD. The 1st fraction, which
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
627
Fig. 1. A schematic drawing of the modified silica gel column chromatography. Insert: enlargement of the elution chamber and the three-way valve.
contains mostly hydrocarbons, was obtained by eluting with hexane at 4.06 0.18 mL/min and 23 1 8C for 4.5 h. The elution was stopped after the appearance of squalene peak in the GC chromatogram. The next fraction, which is almost pure squalene, was obtained after eluting with hexane at 23 8C and 17.32 0.57 mL/min for another 4.5 h. Squalene purity was chosen as an indicator for successful separation. The 3rd fraction, which contains most of the FASEs in NPLF, was obtained after eluting with a mixture of hexane/ethyl acetate (99/1, v/v) at 17.32 0.57 mL/min and 23 1 8C for 6 h. The elution was stopped after more than 95% FASEs were recovered. The remaining compounds that were still absorbed on silica gel were eluted from the column with ethyl acetate at 23 8C and 33.33 mL/min for 1 h. 2.4. Determination of free fatty acids (FFAs) contents The contents of FFAs as oleic acid were determined by the American Oil Chemists’ Society (AOCS) official method Ca 5a-40 (1997).
vapor. The FASEs and steroidal hydrocarbons spots were detected by spraying with a fresh solution of 50 mg ferric chloride in a mixture of 90 mL water, 5 mL acetic acid, and 5 mL sulfuric acid. After heating at 100 8C for 3–5 min, it was indicated by a red–violet color (Fried, 1996). The contents of squalene, FASEs, free phytosterols, tocopherols, and acylglycerols in each fraction were determined by HTGC. External standard calibration curves were obtained by using 0.2–20 mg pure standard. The chromatographic analysis was performed on a TLC plate and a Shimadzu GC-17A (Kyoto, Japan) gas chromatography equipped with a flame ionization detector (FID). Separations were carried out on a DB-5HT (5%-phenyl)-methylpolysiloxane non-polar column (15 m 0.32 mm i.d.; Agilent Tech. Palo Alto, California). Temperatures of the injector and the detector were both set at 370 8C. The temperature of the column was started at 80 8C, and was increased to 365 8C at 15 8C/min and maintained at 365 8C for 8 min. The split ratio was 1:50 using nitrogen as carrier gas with a linear velocity of 30 cm/s at 80 8C. A 20-mg sample was dissolved in 1 mL ethyl acetate, and a 1 mL sample was taken and injected into the HTGC.
2.5. Analysis by TLC and HTGC All compounds in each fraction were identified by TLC and HTGC using authentic standards. The TLC plates were developed in pure hexane. After air-drying, spots on each plate were visualized by exposing the chromatogram to iodine
2.6. Analysis by gas chromatography mass spectrometry (GCMS) Squalene peak was analyzed by a Shimadzu GCMS-QP2010 (Kyoto, Japan), equipped with a mass selective detector.
628
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
where n represents the total number of experiments. The statistical significance of parameter effects was systematically checked by the p-value method. The p-value was compared with significance level of 0.05, a. 3. Results and discussion 3.1. Modified silica gel column chromatography
Fig. 2. Flow chart showing the separation and purification of squalene and FASEs from the NPLF of SODD.
Separations were carried out on a DB-5 ms (5%-phenyl)methylpolysiloxane non-polar column (30 m 0.32 mm i.d.; Agilent Tech. Palo Alto, California). Temperatures of the injector and the detector were both set at 250 8C. The temperature of the column was started at 80 8C, and was increased to 320 8C at 15 8C/min and maintained at 320 8C for 12 min. The split ratio was 1:50 using nitrogen as carrier gas with a linear velocity of 30.6 cm/s at 80 8C. All mass spectra were acquired using the electron impact (EI) mode at 0.80 kV and an ion source temperature of 200 8C. The mass spectra were scanned in the range of m/z 30–600 at 1250 scan/s. 2.7. Statistical analysis The reliability of the results was checked by statistical analyses. The standard deviation of the measures (S) was calculated from the difference between the value of individual experiment, x, and the mean value of three independent experiments, x¯ , using the formula sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Sðx x¯ Þ2 S¼ n1
The NPLF of SODD used in this study was a yellowish liquid at room temperature. Table 1 shows its composition. Almost all present applications of LSC are limited to two adsorbent types: silica and alumina. Silica gel was used in this study because it allows higher sample loading, is less likely to promote unwanted reactions of the sample during separation and is available in a wide range of useful forms (Snyder and Kirkland, 1979). Chromatography involves the separation of components of a mixture by virtue of differences in the equilibrium distribution factor (k0 ) of the components between mobile phase and stationary phase. The k0 is equal to the ratio of concentration of a component in the stationary phase to that in the mobile phase (Hamilton and Sewell, 1982). Sample k0 values normally decrease by 2% for 1 8C increase in elution column temperature (Snyder and Kirkland, 1979). Since this new design was equipped with a heater in the bottom part, it is important to test the ability of this design to yield a uniform temperature distribution of the packing region during operation. In Fig. 3, the effects of mass ratio of NPLF to silica gel and mobile phase flow rate, and the mobile phase volume above the silica gel packing on temperature distributions in the column are shown. A mobile phase volume of 100 mL was chosen in this study because a uniform temperature distribution within the silica packing under various mobile phase flow rates was achieved. HTGC chromatograms of fractions collected from modified silica gel column chromatography at optimum conditions are shown in Fig. 4. The initial contents of squalene and FASEs are 6.22% and 12.28%, respectively (Fig. 4(a)). The 1st fraction collected in the first 4.5 h with hexane as the eluting solvent consists mostly of aliphatic, steroidal and sesquiterpene hydrocarbons and represents about 40–50% of the NPLF (Fig. 4(b)). The 2nd fraction collected in the next 4.5 h is squalene with very high purity as can be seen in Fig. 4(c). The last fraction which contains most of the FASEs in the NPLF was Table 1 Non-polar lipid fraction composition of SODD (wt.%)a Compounds
NPLF
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Othersb
6.22 0.27 12.28 2.05 2.26 0.38 0.18 0.12 21.96 2.99 2.79 0.81 53.91 4.36
a
(1)
Average of three independent experiments. Hydrocarbons, aldehydes, ketones, pesticides, herbicides, and the breakdown products of tocopherols and free phytosterols. b
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
629
Fig. 3. Temperature profile during process in modified silica gel column chromatography. Operation conditions: 100 mL mobile phase volume above the surface of the silica gel: (a) NPLF/(silica gel) = 1/10 (g/g), (b) NPLF/(silica gel) = 1/20 (g/g). 50 mL mobile phase volume above the surface of the silica gel: (c) NPLF/(silica gel) = 1/10 (g/g), (d) NPLF/(silica gel) = 1/20 (g/g). Legend: mobile phase flow rate (mL/min) at (i) 4.06, (ii) 17.32, and (iii) 33.33.
obtained after eluting the column for 6 h with mixture of hexane and ethyl acetate (99/1, v/v) (Fig. 4(d)). Substances that were still adsorbed on silica gel were collected in the washing fraction which consists mostly of FFAs as shown in Fig. 4(e). Results of TLC analyses of each fraction as described in Fig. 4 are shown in Fig. 5. 3.2. Isolation of aliphatic, steroidal, and sesquiterpene hydrocarbons Aliphatic, steroidal and sesquiterpene hydrocarbons represent about 40–50% of the NPLF of SODD. Due to such high content, efficient removal of these hydrocarbons is crucial for obtaining high purity squalene (triterpene hydrocarbon) which is the most polar hydrocarbons in the NPLF. In this study, we choose squalene as an indicator for successful separation of the 1st (aliphatic, steroidal and sesquiterpene hydrocarbons) and the 2nd fraction (squalene); parameters such as gradient elution of mobile phase, NPLF to silica gel mass ratio, elution temperature and mobile phase flow rate were adjusted until high purity (>90%) squalene with high recovery (>80%) was obtained. The major decision required in the design of a satisfactory separation is selection of the mobile phase. Hexane was
selected as the eluting solvent for the 1st and the 2nd fractions because mixture of solvent (hexane and ethyl acetate) did not result in satisfactory separation (data not shown). This is because polarities of aliphatic, steroidal, sesquiterpene, and triterpene hydrocarbons are roughly in the same range. A higher NPLF to silica gel mass ratio is desirable in column chromatography, because less adsorbent and shorter elution time are required. However, a NPLF to silica gel mass ratio larger than 1/10 did not result in successful separations (data not shown) because it was larger than the linear capacity of the column (Snyder and Kirkland, 1979). Therefore, the mass ratios employed in this study were either 1/10 or 1/20. Moreover, this result agrees with the same literature that for analytical separations, it is always preferable to work within a sample size range where k0 values are constant. In isocratic elution (chromatographic operation with constant mobile phase composition), the selection of correct column temperature and flow rate of mobile phase are important for obtaining good separation (Hamilton and Sewell, 1982). When packing region temperature was controlled at 15 and 30 8C, satisfactory separation was not obtained. A satisfactory separation was achieved at a packing region temperature of 23 8C. This may be the result of the dependence of solubility of sample in the mobile phase on temperature,
630
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Fig. 5. Results of the TLC analyses developed in pure n-hexane. (a) NPLF; (b) the 1st fraction; (c) the 2nd fraction; (d) the 3rd fraction; (e) the washing fraction.
Fig. 4. Result of the HTGC analysis. Operation conditions for the modified silica gel column chromatography: NPLF to silica gel mass ratio = 1/10, and elution temperature = 23 8C. (a) The GC chromatograms of the NPLF, (b) the 1st fraction, (c) the 2nd fraction, (d) the 3rd fraction, and (e) the washing fraction eluted from the modified silica gel column chromatography.
which in turn affects the size of sample. Snyder and Kirkland (1979) reported that mobile phase of relatively low viscosity is favored to maintain high column efficiency by setting column temperature at certain value. The five major contributions on column performance are eddy diffusion, mobile phase mass transfer, longitudinal diffusion, stagnant mobile phase mass transfer, and stationary phase mass transfer. Longitudinal diffusion is often not important, but it is significant at low mobile phase flow rates
for small particle column (Hamilton and Sewell, 1982; Snyder and Kirkland, 1979). That means the dependence of column efficiency on mobile phase flow rate is of great practical importance. Among the mobile phase flow rates that employed in this study (4.06, 17.32 and 33.33 mL/min), only 4.06 mL/min yield satisfactory separation. Since polarities of the 1st and the 2nd fractions are roughly in the same range, lower mobile phase flow rate is preferable because it will maximize the contribution of longitudinal diffusion. Hamilton and Sewell (1982) reported that in order to increase column efficiency, it is necessary to use adsorbent with small diameter particles, a small mobile phase flow rate of low viscosity and high elution temperature. Table 2 shows that a higher NPLF to silica gel mass ratio results in a higher recovery of all hydrocarbons in the 1st fraction. The p-value method was applied to determine the significance of these differences. At a lower NPLF to silica gel mass ratio (1/20), the impurity (mainly squalene) content and recovery in the 1st fraction is significantly lower ( p < 0.05) than that obtained if a higher NPLF to silica mass ratio (1/10) was used.
Table 2 Effect of NPLF/(silica gel) on the impurity (squalene) content and recovery in the 1st fractiona Compounds
Squalene FASEs Others c Amount (%) d a
NPLF/(silica gel) = 1/20 (g/g)
NPLF/(silica gel) = 1/10 (g/g)
Purity (wt.%)
Recovery (%)
0.52 0.33 0 99.48 0.33
3.37 1.87 0 81.53 0.35 43.65 0.35
b
Purity (wt.%)
Recovery (%)
1.19 0.31 0 98.81 0.31
9.43 2.13 0 86.35 3.97 47.97 2.04
Average of three independent experiments. Operation conditions: 150 mL n-hexane as mobile phase, 4.06 mL/min flow rate, 4.5 h elution time and 23 8C elution temperature. b Recovery = {(1st fraction mass (g) content of the compound in 1st fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly hydrocarbons such as aliphatic, steroidal, and sesquiterpene hydrocarbons. d Amount = (1st fraction mass (g)/NPLF mass (g)) 100.
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
631
Table 3 Effect of NPLF/(silica gel) on the purity and recovery of squalene in the 2nd fractiona Compounds
Squalene FASEs Othersc
NPLF/(silica gel) = 1/20 (g/g)
NPLF/(silica gel) = 1/10 (g/g)
Purity (wt.%)
Recovery (%)
94.96 2.14 0 5.04 2.14
93.83 0.69 0 0.61 0.25
Amount (%)d
6.06 0.06
b
Purity (wt.%)
Recovery (%)
99.36 0.13 0 0.64 0.13
81.78 0.62 0 0.07 0.02 5.18 0.06
a
Average of three independent experiments. Operation conditions: 150 mL n-hexane as mobile phase, 17.32 mL/min flow rate, 4.5 h elution time and 23 8C elution temperature. b Recovery = {(2nd fraction mass (g) content of the compound in 2nd fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly hydrocarbons such as aliphatic, steroidal, and sesquiterpene hydrocarbons. d Amount = (2nd fraction mass (g)/NPLF mass (g)) 100.
3.3. Isolation and purification of squalene The effect of NPLF/silica gel on the isolation of squalene is shown in Table 3. It can be seen that the squalene content is significantly lower ( p < 0.05) while the recovery of squalene is significantly higher ( p < 0.05) at NPLF/(silica gel) = 1/20 than those obtained at NPLF/(silica gel) = 1/10. At NPLF/ (silica gel) = 1/10, squalene with high purity and recovery can be obtained in the 2nd fraction in 4.5 h by using large mobile phase flow rate (17.32 mL/min) due to relative large difference in the polarities of squalene and FASEs. As the result, the time required is much shorter than that required to achieve comparable results (in terms of squalene purity and recovery) when classical silica gel column chromatography was used as reported by Gunawan et al. (2008). The identification of compounds in the 2nd fraction was confirmed by HTGC (Fig. 4), TLC (Fig. 5), and GCMS analyses. The GCMS fragmentation of the peak showed a molecular ion (M+) at 410 (relative intensity: 1.09%) and base peak at 69.05. Peaks were also observed at m/z at 149, 137, 121, 109, 95, 81, 55, and 41. These fragmentation patterns were quite similar to those for squalene reported in U.S. National Bureau of Standards library (Stein, 2005). These fractionations obtained in this study give better separation than previous study of Lanzo´n et al. (1994) in which the unsaponifiable matter of both virgin and refined olive oil was fractioned by classical silica gel column chromatography with a unsaponifiable matter to silica gel mass ratio of 1/75, using hexane as eluent. The fraction containing squalene was eluted after aliphatic, steroidal and sesquiterpene hydrocarbons, the purity of squalene was not reported in their study. However, judging from the chromatogram in their study, low squalene purity was obtained. 3.4. Separation and purification of FASEs Using single solvent as mobile phase is more desirable than using mixture of solvents in column chromatography because of easier recovery and reuse of solvent. When pure hexane was employed as the eluent, it was impossible to recovery most FASEs in reasonable short time in the 3rd fraction. It was found that it required 83 h to isolate FASEs with a 83.06% purity and a
97.84% recovery at 23 8C using hexane as the mobile phase. The elution time can be reduced to 20 h by increasing the elution temperature (temperature in the packing region) to 64 8C due to increasing solubility of sample in mobile phase. However, the purity of FASEs obtained was much lower (39.01% purity, 96.15% recovery). Mixture of hexane and ethyl acetate was employed as the eluent in order to successfully recover most FASEs in shorter time in the 3rd fraction. As shown in Table 4, as the polarity increases by increasing the percentage of ethyl acetate in the mobile phase from 1% to 3%, the purity of FASEs in the 3rd fraction decreases from 78.48% to 55.14%, while the corresponding recoveries varies only slightly. One of the important criteria for successful separation in this study is to recover most of the FASEs in the 3rd fraction. FASEs contents are significantly higher ( p < 0.05) when mobile phase contained 1% ethyl acetate as compared to those obtained when 2% and 3% ethyl acetate were used. However, the contents and recoveries of the FFAs and acylglycerols are not significantly different ( p > 0.05) among three studies using different percentages of ethyl acetate. These results agree with previous observation that once the squalene was completely eluted, polarity of the mobile phase must be increased to elute other more polar compounds (Lanzo´n et al., 1994). More silica gel yielded a higher adsorption area available per unit NPLF, which resulted in longer elution time with comparable separation due to a higher amount of mobile phase that was used to displace FASEs. For example, at NPLF to silica gel ratio of 1/20 and with 2% ethyl acetate in the mobile phase, the elution time to obtain the 3rd fraction (FASEs, 69.64 9.99% purity and 95.18 6.45% recovery) was significantly longer (6 h, p < 0.05) than those obtained while using a NPLF to silica mass ratio of 1/10 (2 h). As shown in Table 5, for comparable separation, both elution time and amount of solvent required are much less when using modified silica gel column chromatography as compared to classical silica gel column chromatography. By employing modified silica gel column chromatography to obtain the squalene-rich fraction, the mobile phase volume and elution time required as fractions of those needed in classical silica gel column chromatography are 1/73 and 1/18, respectively. To obtained the FASEs-rich fraction, the corresponding mobile
632
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Table 4 Effect of mobile phase polarity on the composition of 3rd fractiona Compounds
Mobile phase (hexane/ethyl acetate (v/v)) 99/1
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Others c
98/2
Purity (wt.%)
Recovery (%)
0.61 0.75 78.48 0.86 0.91 0.92 0 3.43 0.70 12.77 1.47 3.80 3.46
1.51 1.91 98.00 2.55 8.27 2.32 0 2.75 0.55 55.45 6.57 1.02 0.93
Time Amount (%) d
b
97/3
Purity (wt.%)
Recovery (%)
Purity (wt.%)
Recovery (%)
1.79 0.16 64.56 2.83 2.79 0.15 0 5.69 2.30 10.03 2.49 15.13 2.46
1.51 1.91 96.60 5.01 8.27 2.32 0 6.45 0.64 61.85 3.79 4.98 1.30
1.51 0.54 55.14 5.49 1.46 0.21 0 6.30 2.96 9.88 2.43 25.70 6.16
5.53 2.40 97.94 1.60 16.45 3.48 0 6.22 3.34 75.71 10.22 10.27 3.31
6h 15.41 0.37
2h 18.05 2.08
1h 21.59 2.22
a Average of three independent experiments. Operation conditions: NPLF/(silica gel) = 1/10 (g/g), 150 mL mobile phase volume, 17.32 mL/min flow rate, and 23 8C elution temperature. b Recovery = {(3rd fraction mass (g) content of the compound in 3rd fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly ketones and aldehydes. d Amount = (3rd fraction mass (g)/NPLF mass (g)) 100.
Table 5 Comparison method on the separation and purification of squalene and FASEs from NPLF of SODD Parameters
Modified column chromatographya
Classical column chromatographyb
Squalene fraction Purity (wt.%) Recovery (%)c Mobile phase Flow rate (mL/min) Time required (h)
99.36 0.13 81.78 0.62 150 mL hexane 17.32 4.5
95.90 93.09 10.96 L hexane 4.06 45
FASEs fraction Purity (wt.%) Recovery (%) Mobile phase Flow rate (mL/min) Time required (h)
78.48 0.86 98.00 2.55 150 mL hexane:ethyl acetate (99:1, v/v) 17.32 6
84.04 63.17 33.13 L hexane 4.06 137
a b c
Average of three independent experiments. Operation conditions: NPLF/(silica gel) = 1/10 (g/g) and 23 8C of elution temperature. Gunawan et al. (2008). Operation conditions: NPLF/(silica gel) = 1/10 (g/g) and 23 8C of elution temperature. Recovery = {(fraction mass (g)) content of the compound in fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100.
Table 6 The fractionation of NPLF obtained by modified silica gel column chromatography with an NPLF of low squalene and FASEs contents (wt.%)a Compounds
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Others a b c d e f
NPLF
1st fraction
Content (wt.%)
Content (wt.%) and recovery (%)
3.14 0.17 4.29 1.81 0.78 0.14 0.43 0.07 37.25 0.64 4.04 0.61 50.07 2.18d
b
0.30 0.15 (3.41 1.37) NDc (0) ND (0) ND (0) ND (0) ND (0) 99.70 0.15e (74.24 6.16)
2nd fraction
3rd fraction
Content (wt.%) and recovery (%)
Content (wt.%) and recovery (%)
97.37 1.26 (93.50 0.27) ND (0) ND (0) ND (0) ND (0) ND (0) 2.63 1.26f (0.16 0.02)
0.23 0.19 (1.73 0.10) 36.00 1.96 (98.35 0.32) 0.76 0.08 (16.24 1.78) 0.30 0.24 (0.34 0.28) 49.00 1.41 (23.50 2.12) 7.12 2.68 (55.46 0.76) 12.88 2.93f (4.98 1.55)
Average of three independent experiments. Recovery = {(fraction mass (g) content of the compound in fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100%. Not detected. Hydrocarbons, aldehydes, ketones, pesticides, herbicides, and the breakdown products of tocopherols and free phytosterols. Aliphatic, steroidal, and sesquiterpene hydrocarbons. Mainly aldehydes and ketones.
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
phase volume and elution time are 1/221 and 1/23, respectively of those needed in classical silica gel column chromatography. This separation was confirmed in subsequent experiments, carried out under the same conditions except with a NPLF that contains low percentage of squalene and FASEs (Table 6). Squalene content in the 2nd fraction and FASEs recoveries in the 3rd fraction in this case were not significantly different ( p > 0.05), compared to those obtained using a NPLF with high squalene and FASEs contents. Also, it was found that the 3rd fraction was richer in FFAs due to higher content of FFAs in NPLF. 4. Conclusion A new silica gel column chromatography was designed in this work. Unlike classical column chromatography, in this new design, glass tube is equipped with a jacket, a valve to control the flow rate of eluent, a condenser system and a heating mantle. A heater is placed in the bottom part and a uniform temperature in the silica gel packing is achieved by maintaining the mobile phase volume at constant value. From data obtained in this study, modified silica gel column chromatography is a more favorable mode of separation compare to classical silica gel column chromatography. The advantage of the new design presented in this work over classical silica gel column chromatography is that it significantly reduces the amount of solvent and time required to achieve the same degree of separation. Acknowledgement This work was supported by a grant (NSC 94-2214-E011004) provided by the National Science Council of Taiwan.
633
References AOCS Official Method Ca 5a-40, ‘‘Free Fatty Acids,’’ Official Methods and Recommended Practices of the AOCS, Firestone D. (Ed.), American Oil Chemists’ Society, Champaign, IL, U.S.A. (1997). Bondioli, P., C. Mariani, A. Lanzani, E. Fedeli, and A. Muller, ‘‘Squalene Recovery from Olive Oil Deodorizer Distillate,’’ J. Am. Oil Chem. Soc., 70, 763 (1993). Fried, B., Lipids, p. 704, Sherma, J. and B. Fried (Eds.), Marcel Dekker Press, New York, U.S.A. (1996). Giacometti, J., A. Milosevic, and C. Milin, ‘‘Gas Chromatographic Determination of Fatty Acids Contained in Different Lipid Classes after Their Separation by Solid-Phase Extraction,’’ J. Chromatogr. A, 976, 47 (2002). Gunawan, S., S. R. Vali, and Y. H. Ju, ‘‘Purification and Identification of Rice Bran Oil Fatty Acid Steryl and Wax esters,’’ J. Am. Oil Chem. Soc., 83, 449 (2006). Gunawan, S., N. S. Kasim, and Y. H. Ju, ‘‘Separation and Purification of Squalene from Soybean Oil Deodorize Distillate,’’ Sep. Purif. Technol., 60, 128 (2008). Hamilton, R. J. and P. A. Sewell, Introduction to High Performance Liquid Chromatography, 2nd Ed., p. 163, Chapman and Hall Ltd., London, U.K. (1982). Lacaze, J.P.C.L., L. A. Stobo, E. A. Turrell, and M. A. Quilliam, ‘‘Solid-phase Extraction and Liquid Chromatography-mass Spectrometry for The Determination of Free Fatty Acids in Shellfish,’’ J. Chromatogr. A, 1145, 51 (2007). Lanzo´n, A., T. Albi, A. Cert, and J. Gracia´n, ‘‘The Hydrocarbon Fraction of Virgin Olive Oil and Changes Resulting from Refining,’’ J. Am. Oil Chem. Soc., 71, 285 (1994). Moreda, W., M. C. Perez-Camino, and A. Cert, ‘‘Gas and Liquid Chromatography of Hydrocarbons in Edible Vegetable Oils,’’ J. Chromatogr. A, 936, 159 (2001). Snyder, L. R. and J. J. Kirkland, Introduction to Modern Liquid Chromatography, 2nd Ed., p. 321, John Wiley and Sons Inc., New York, U.S.A. (1979). Stein, S.E., ‘‘Mass Spectra,’’ NIST Chemistry WebBook, NIST Standard Reference Database Number 69, Linstrom, P. J. and W. G. Mallard (Eds.), National Institute of Standards and Technology, Gaithersburg MD, U.S.A. (2005). http://webbook.nist.gov/chemistry (accessed September 23, 2007).
Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633 www.elsevier.com/locate/jcice
Design and operation of a modified silica gel column chromatography Setiyo Gunawan, Suryadi Ismadji, Yi-Hsu Ju * Department of Chemical Engineering, National Taiwan University of Science and Technology, 43 Keelung Road, Section 4, Taipei 106-07, Taiwan Received 5 February 2008; received in revised form 20 June 2008; accepted 25 June 2008
Abstract Liquid–solid chromatography (LSC) is the oldest of the various liquid chromatography methods. Despite the fact that high-performance liquid chromatography (HPLC) operation leads to better separation and analysis, classical column chromatography and thin-layer chromatography (TLC) are still widely practiced because of their convenience. In this study, a modified silica gel column chromatography was designed with the objective of reducing the amount of solvent required to achieve the same degree of separation as the classical silica gel column chromatography. The separation of squalene and fatty acid steryl esters (FASEs) from non-polar lipid fraction (NPLF) of soybean oil deodorizer distillate (SODD) was employed as a model system to test the effectiveness of this new design. Modified silica gel column chromatography process is feasible from economic point of view compare to classical silica gel column chromatography because it significantly reduces the amount of solvent and time required to achieve the same degree of separation. By employing modified silica gel column chromatography to obtain the squalene-rich fraction, the mobile phase volume and elution time required as fractions of those needed in classical silica gel column chromatography are 1/73 and 1/18, respectively. To obtain the FASEs-rich fraction, the corresponding mobile phase volume and elution time are 1/221 and 1/23, respectively of those needed in classical silica gel column chromatography. # 2008 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved. Keywords: Fatty acid steryl esters; Modified silica gel column chromatography; Soybean oil deodorizer distillate; Squalene
1. Introduction Liquid–solid chromatography (LSC) or adsorption is the oldest of the various liquid chromatography methods. Despite the fact that high-performance liquid chromatography (HPLC) operation leads to better separation and analysis, classical column chromatography and thin-layer chromatography are still widely practiced because of their convenience. The successful use of liquid chromatography for a given problem requires the right combination of operating conditions: the type of column packing and mobile phase, the length and diameter of the column, mobile phase flow rate, elution temperature, and sample size (Snyder and Kirkland, 1979). Temperature is normally not used as separation variable in LSC. Most applications are carried out at ambient temperature. A classical silica gel column chromatography was used to fractionate hydrocarbons from the unsaponifiable matter of both virgin and refined olive oil, using hexane as the eluent (Lanzo´n et al., 1994). Also, a classical silica gel column chromatography
* Corresponding author. Tel.: +886 2 27376612; fax: +886 2 27376644. E-mail address: [email protected] (Y.-H. Ju).
was used to fractionate hydrocarbons and fatty acid steryl esters (FASEs) from the non-polar lipid fraction (NPLF) obtained from crude rice bran oil (Gunawan et al., 2006) and soybean oil deodorizer distillate (SODD) (Gunawan et al., 2008). When the NPLF of SODD was introduced into a classical silica gel column chromatography to isolate squalene, squalene (95.90% purity and 93.09% recovery) was obtained in the 2nd fraction after eluting the column with 10.96 L hexane. The continued elution of column with 33.13 L hexane yielded a 3rd fraction rich in FASEs (84.04% purity and 63.17% recovery). In another study, the separation of squalene from the unsaponifiable matter of olive oil deodorizer distillate was achieved by thin-layer chromatography (TLC), using mixture of solvent as mobile phase (Bondioli et al., 1993). Classical column chromatography and TLC are laborious, time consuming (Moreda et al., 2001), and requiring the use of large amount of organic solvents (Lacaze et al., 2007). A disadvantage of TLC compared to classical column chromatography is the limited possibility of understanding quantitative analysis since exhaustive recovery of the separate spot is not sufficiently precise (Moreda et al., 2001). The advantages over classical column chromatography are that the procedure is more rapid, and the possibility of using specific reagents to immediately reveal the nature of the separated compound. Solid-phase extraction (SPE)
0368-1653/$ – see front matter # 2008 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jcice.2008.06.006
626
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Nomenclature AOCS EI FASEs FFAs GCMS h HPLC HTGC k0 LSC n NIST NPLF p S SODD SPE TLC x x¯
American oil chemists’ society electron impact fatty acid steryl esters free fatty acids gas chromatography mass spectrometry hour high-performance liquid chromatography high-temperature gas chromatography equilibrium distribution factor liquid–solid chromatography total number of experiment national institute of standards and technology non-polar lipid fraction probability standard deviation of the measures soybean oil deodorizer distillate solid-phase extraction thin-layer chromatography value of individual experiment mean value of three independent experiments
Greek symbol a significance level has been developed to isolate squalene from virgin olive oil without any chemical treatment. Rapidity, reliability, use of minimal amounts of solvent and automation are the main advantages of SPE in comparison to TLC and classical column chromatography (Giacometti et al., 2002). All the methods mentioned above are limited by small sample loading and using mixture of organic solvent in order to obtain good separations. In this study, a new silica gel column chromatography was designed with the objective of reducing the amount of solvent required to achieve the same degree of separation as the classical silica gel column chromatography. The isolation of squalene and FASEs from the NPLF of SODD was employed as a model system to test the effectiveness of this new design. The effects of parameters, such as NPLF to silica gel mass ratio, flow rate and composition of mobile phase, and elution temperature, on the separation were systematically investigated. 2. Materials and methods 2.1. Chemicals and materials Standard nonacosane, farnesene, cholesta-3,5-diene, squalene, fatty acids, a-, d-, g-tocopherol, monooleylglycerol, diolein, triolein, and tripalmitin were obtained from Sigma Chemicals Company (St. Louis, MO). Standard b-sitosterol (practical grade) was obtained from MP Biomedicals, LLC (Aurora, OH). All solvents and reagents were either of HPLC grade or analytical reagent grade and were obtained from commercial sources. SODD was donated by TTET Union Corporation (Tainan, Taiwan). TLC aluminum plates
(20 cm 20 cm 250 mm) were purchased from Merck (Darmstadt, Germany). Silica Gel (70–230 mesh) was obtained from Silicycle (Quebec, Canada). Characteristics of the gel according to the manufacturer were: particle size: 60–200 mm; ˚ ; pH: 7; water content: 6%; and specific surface pore size: 60 A 2 area: 500 m /g. 2.2. Extraction of NPLF from SODD Solid suspension was removed from SODD at 40 8C by using a 7 mm Advantec filter paper (Toyo Roshi Kaisha Ltd., Tokyo, Japan). SODD was applied to modified soxhlet extraction, under the following operation conditions: solvent = n-hexane, SODD to silica gel mass ratio = 1:3 (w/w), extraction temperature = 6 8C, and extraction time = 11 h as described by Gunawan et al. (2008). The resulting hexane extractive, named NPLF, was used for this study. 2.3. Modified silica gel column chromatography A modified version of the classical silica gel column chromatography (300 mm 4 mm i.d. glass tube) equipped with a jacket, a valve to control the flow rate of eluent and a condenser system was employed in this study. Ethyl alcohol was used as the refrigerant. This was circulated in the jacket to control the column (packing region) temperature. A schematic drawing of the modified silica gel column chromatography is shown in Fig. 1. The siphon arm for operation in soxhlet extraction mode was closed by three-way valve. Silica gel (18–60 g) was kept in a furnace at 150 8C for 1 h to remove its water content. Next, a slurry of silica gel in hexane was poured into the column that was previously half-filled with hexane. The exit of the column was plugged with 1 g cotton to retain the silica gel and a thermo couple was inserted above the cotton. The hexane was allowed to drain slightly during packing. The top surface of the silica gel was covered by 1 g cotton, and three thermocouples were set inside the elution chamber at different locations. The hexane level was lowered until at controlled level (50 and 100 mL above the cotton at the top). NPLF (3 g) was put into the column at room temperature (23 1 8C). The column was eluted with mobile phase (50 mL hexane), which was put into a 500 mL roundbottom flask at the start of the run and was heated. The mobile phase vapor traveled up the distillation arm, and flooded into the column housing the silica gel and the NPLF. The condenser ensured that any mobile phase vapor condenses and drips back down into the column. The column was slowly eluted with the mobile phase at controlled flow rate and temperature. After predetermined time, the flask that contained the desired extract was removed and replaced immediately by another flask that contained 50 mL fresh mobile phase so that the total amount of mobile phase remained the same as in the beginning of the run. After removing the mobile phase, each fraction was analyzed by TLC and high-temperature gas chromatography (HTGC). The flow rate of mobile phase vapor depends on the heat input from the heater and was predetermined (data not shown). Fig. 2 shows the flow chart of separation of squalene and FASEs from the NPLF of SODD. The 1st fraction, which
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
627
Fig. 1. A schematic drawing of the modified silica gel column chromatography. Insert: enlargement of the elution chamber and the three-way valve.
contains mostly hydrocarbons, was obtained by eluting with hexane at 4.06 0.18 mL/min and 23 1 8C for 4.5 h. The elution was stopped after the appearance of squalene peak in the GC chromatogram. The next fraction, which is almost pure squalene, was obtained after eluting with hexane at 23 8C and 17.32 0.57 mL/min for another 4.5 h. Squalene purity was chosen as an indicator for successful separation. The 3rd fraction, which contains most of the FASEs in NPLF, was obtained after eluting with a mixture of hexane/ethyl acetate (99/1, v/v) at 17.32 0.57 mL/min and 23 1 8C for 6 h. The elution was stopped after more than 95% FASEs were recovered. The remaining compounds that were still absorbed on silica gel were eluted from the column with ethyl acetate at 23 8C and 33.33 mL/min for 1 h. 2.4. Determination of free fatty acids (FFAs) contents The contents of FFAs as oleic acid were determined by the American Oil Chemists’ Society (AOCS) official method Ca 5a-40 (1997).
vapor. The FASEs and steroidal hydrocarbons spots were detected by spraying with a fresh solution of 50 mg ferric chloride in a mixture of 90 mL water, 5 mL acetic acid, and 5 mL sulfuric acid. After heating at 100 8C for 3–5 min, it was indicated by a red–violet color (Fried, 1996). The contents of squalene, FASEs, free phytosterols, tocopherols, and acylglycerols in each fraction were determined by HTGC. External standard calibration curves were obtained by using 0.2–20 mg pure standard. The chromatographic analysis was performed on a TLC plate and a Shimadzu GC-17A (Kyoto, Japan) gas chromatography equipped with a flame ionization detector (FID). Separations were carried out on a DB-5HT (5%-phenyl)-methylpolysiloxane non-polar column (15 m 0.32 mm i.d.; Agilent Tech. Palo Alto, California). Temperatures of the injector and the detector were both set at 370 8C. The temperature of the column was started at 80 8C, and was increased to 365 8C at 15 8C/min and maintained at 365 8C for 8 min. The split ratio was 1:50 using nitrogen as carrier gas with a linear velocity of 30 cm/s at 80 8C. A 20-mg sample was dissolved in 1 mL ethyl acetate, and a 1 mL sample was taken and injected into the HTGC.
2.5. Analysis by TLC and HTGC All compounds in each fraction were identified by TLC and HTGC using authentic standards. The TLC plates were developed in pure hexane. After air-drying, spots on each plate were visualized by exposing the chromatogram to iodine
2.6. Analysis by gas chromatography mass spectrometry (GCMS) Squalene peak was analyzed by a Shimadzu GCMS-QP2010 (Kyoto, Japan), equipped with a mass selective detector.
628
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
where n represents the total number of experiments. The statistical significance of parameter effects was systematically checked by the p-value method. The p-value was compared with significance level of 0.05, a. 3. Results and discussion 3.1. Modified silica gel column chromatography
Fig. 2. Flow chart showing the separation and purification of squalene and FASEs from the NPLF of SODD.
Separations were carried out on a DB-5 ms (5%-phenyl)methylpolysiloxane non-polar column (30 m 0.32 mm i.d.; Agilent Tech. Palo Alto, California). Temperatures of the injector and the detector were both set at 250 8C. The temperature of the column was started at 80 8C, and was increased to 320 8C at 15 8C/min and maintained at 320 8C for 12 min. The split ratio was 1:50 using nitrogen as carrier gas with a linear velocity of 30.6 cm/s at 80 8C. All mass spectra were acquired using the electron impact (EI) mode at 0.80 kV and an ion source temperature of 200 8C. The mass spectra were scanned in the range of m/z 30–600 at 1250 scan/s. 2.7. Statistical analysis The reliability of the results was checked by statistical analyses. The standard deviation of the measures (S) was calculated from the difference between the value of individual experiment, x, and the mean value of three independent experiments, x¯ , using the formula sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Sðx x¯ Þ2 S¼ n1
The NPLF of SODD used in this study was a yellowish liquid at room temperature. Table 1 shows its composition. Almost all present applications of LSC are limited to two adsorbent types: silica and alumina. Silica gel was used in this study because it allows higher sample loading, is less likely to promote unwanted reactions of the sample during separation and is available in a wide range of useful forms (Snyder and Kirkland, 1979). Chromatography involves the separation of components of a mixture by virtue of differences in the equilibrium distribution factor (k0 ) of the components between mobile phase and stationary phase. The k0 is equal to the ratio of concentration of a component in the stationary phase to that in the mobile phase (Hamilton and Sewell, 1982). Sample k0 values normally decrease by 2% for 1 8C increase in elution column temperature (Snyder and Kirkland, 1979). Since this new design was equipped with a heater in the bottom part, it is important to test the ability of this design to yield a uniform temperature distribution of the packing region during operation. In Fig. 3, the effects of mass ratio of NPLF to silica gel and mobile phase flow rate, and the mobile phase volume above the silica gel packing on temperature distributions in the column are shown. A mobile phase volume of 100 mL was chosen in this study because a uniform temperature distribution within the silica packing under various mobile phase flow rates was achieved. HTGC chromatograms of fractions collected from modified silica gel column chromatography at optimum conditions are shown in Fig. 4. The initial contents of squalene and FASEs are 6.22% and 12.28%, respectively (Fig. 4(a)). The 1st fraction collected in the first 4.5 h with hexane as the eluting solvent consists mostly of aliphatic, steroidal and sesquiterpene hydrocarbons and represents about 40–50% of the NPLF (Fig. 4(b)). The 2nd fraction collected in the next 4.5 h is squalene with very high purity as can be seen in Fig. 4(c). The last fraction which contains most of the FASEs in the NPLF was Table 1 Non-polar lipid fraction composition of SODD (wt.%)a Compounds
NPLF
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Othersb
6.22 0.27 12.28 2.05 2.26 0.38 0.18 0.12 21.96 2.99 2.79 0.81 53.91 4.36
a
(1)
Average of three independent experiments. Hydrocarbons, aldehydes, ketones, pesticides, herbicides, and the breakdown products of tocopherols and free phytosterols. b
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
629
Fig. 3. Temperature profile during process in modified silica gel column chromatography. Operation conditions: 100 mL mobile phase volume above the surface of the silica gel: (a) NPLF/(silica gel) = 1/10 (g/g), (b) NPLF/(silica gel) = 1/20 (g/g). 50 mL mobile phase volume above the surface of the silica gel: (c) NPLF/(silica gel) = 1/10 (g/g), (d) NPLF/(silica gel) = 1/20 (g/g). Legend: mobile phase flow rate (mL/min) at (i) 4.06, (ii) 17.32, and (iii) 33.33.
obtained after eluting the column for 6 h with mixture of hexane and ethyl acetate (99/1, v/v) (Fig. 4(d)). Substances that were still adsorbed on silica gel were collected in the washing fraction which consists mostly of FFAs as shown in Fig. 4(e). Results of TLC analyses of each fraction as described in Fig. 4 are shown in Fig. 5. 3.2. Isolation of aliphatic, steroidal, and sesquiterpene hydrocarbons Aliphatic, steroidal and sesquiterpene hydrocarbons represent about 40–50% of the NPLF of SODD. Due to such high content, efficient removal of these hydrocarbons is crucial for obtaining high purity squalene (triterpene hydrocarbon) which is the most polar hydrocarbons in the NPLF. In this study, we choose squalene as an indicator for successful separation of the 1st (aliphatic, steroidal and sesquiterpene hydrocarbons) and the 2nd fraction (squalene); parameters such as gradient elution of mobile phase, NPLF to silica gel mass ratio, elution temperature and mobile phase flow rate were adjusted until high purity (>90%) squalene with high recovery (>80%) was obtained. The major decision required in the design of a satisfactory separation is selection of the mobile phase. Hexane was
selected as the eluting solvent for the 1st and the 2nd fractions because mixture of solvent (hexane and ethyl acetate) did not result in satisfactory separation (data not shown). This is because polarities of aliphatic, steroidal, sesquiterpene, and triterpene hydrocarbons are roughly in the same range. A higher NPLF to silica gel mass ratio is desirable in column chromatography, because less adsorbent and shorter elution time are required. However, a NPLF to silica gel mass ratio larger than 1/10 did not result in successful separations (data not shown) because it was larger than the linear capacity of the column (Snyder and Kirkland, 1979). Therefore, the mass ratios employed in this study were either 1/10 or 1/20. Moreover, this result agrees with the same literature that for analytical separations, it is always preferable to work within a sample size range where k0 values are constant. In isocratic elution (chromatographic operation with constant mobile phase composition), the selection of correct column temperature and flow rate of mobile phase are important for obtaining good separation (Hamilton and Sewell, 1982). When packing region temperature was controlled at 15 and 30 8C, satisfactory separation was not obtained. A satisfactory separation was achieved at a packing region temperature of 23 8C. This may be the result of the dependence of solubility of sample in the mobile phase on temperature,
630
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Fig. 5. Results of the TLC analyses developed in pure n-hexane. (a) NPLF; (b) the 1st fraction; (c) the 2nd fraction; (d) the 3rd fraction; (e) the washing fraction.
Fig. 4. Result of the HTGC analysis. Operation conditions for the modified silica gel column chromatography: NPLF to silica gel mass ratio = 1/10, and elution temperature = 23 8C. (a) The GC chromatograms of the NPLF, (b) the 1st fraction, (c) the 2nd fraction, (d) the 3rd fraction, and (e) the washing fraction eluted from the modified silica gel column chromatography.
which in turn affects the size of sample. Snyder and Kirkland (1979) reported that mobile phase of relatively low viscosity is favored to maintain high column efficiency by setting column temperature at certain value. The five major contributions on column performance are eddy diffusion, mobile phase mass transfer, longitudinal diffusion, stagnant mobile phase mass transfer, and stationary phase mass transfer. Longitudinal diffusion is often not important, but it is significant at low mobile phase flow rates
for small particle column (Hamilton and Sewell, 1982; Snyder and Kirkland, 1979). That means the dependence of column efficiency on mobile phase flow rate is of great practical importance. Among the mobile phase flow rates that employed in this study (4.06, 17.32 and 33.33 mL/min), only 4.06 mL/min yield satisfactory separation. Since polarities of the 1st and the 2nd fractions are roughly in the same range, lower mobile phase flow rate is preferable because it will maximize the contribution of longitudinal diffusion. Hamilton and Sewell (1982) reported that in order to increase column efficiency, it is necessary to use adsorbent with small diameter particles, a small mobile phase flow rate of low viscosity and high elution temperature. Table 2 shows that a higher NPLF to silica gel mass ratio results in a higher recovery of all hydrocarbons in the 1st fraction. The p-value method was applied to determine the significance of these differences. At a lower NPLF to silica gel mass ratio (1/20), the impurity (mainly squalene) content and recovery in the 1st fraction is significantly lower ( p < 0.05) than that obtained if a higher NPLF to silica mass ratio (1/10) was used.
Table 2 Effect of NPLF/(silica gel) on the impurity (squalene) content and recovery in the 1st fractiona Compounds
Squalene FASEs Others c Amount (%) d a
NPLF/(silica gel) = 1/20 (g/g)
NPLF/(silica gel) = 1/10 (g/g)
Purity (wt.%)
Recovery (%)
0.52 0.33 0 99.48 0.33
3.37 1.87 0 81.53 0.35 43.65 0.35
b
Purity (wt.%)
Recovery (%)
1.19 0.31 0 98.81 0.31
9.43 2.13 0 86.35 3.97 47.97 2.04
Average of three independent experiments. Operation conditions: 150 mL n-hexane as mobile phase, 4.06 mL/min flow rate, 4.5 h elution time and 23 8C elution temperature. b Recovery = {(1st fraction mass (g) content of the compound in 1st fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly hydrocarbons such as aliphatic, steroidal, and sesquiterpene hydrocarbons. d Amount = (1st fraction mass (g)/NPLF mass (g)) 100.
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
631
Table 3 Effect of NPLF/(silica gel) on the purity and recovery of squalene in the 2nd fractiona Compounds
Squalene FASEs Othersc
NPLF/(silica gel) = 1/20 (g/g)
NPLF/(silica gel) = 1/10 (g/g)
Purity (wt.%)
Recovery (%)
94.96 2.14 0 5.04 2.14
93.83 0.69 0 0.61 0.25
Amount (%)d
6.06 0.06
b
Purity (wt.%)
Recovery (%)
99.36 0.13 0 0.64 0.13
81.78 0.62 0 0.07 0.02 5.18 0.06
a
Average of three independent experiments. Operation conditions: 150 mL n-hexane as mobile phase, 17.32 mL/min flow rate, 4.5 h elution time and 23 8C elution temperature. b Recovery = {(2nd fraction mass (g) content of the compound in 2nd fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly hydrocarbons such as aliphatic, steroidal, and sesquiterpene hydrocarbons. d Amount = (2nd fraction mass (g)/NPLF mass (g)) 100.
3.3. Isolation and purification of squalene The effect of NPLF/silica gel on the isolation of squalene is shown in Table 3. It can be seen that the squalene content is significantly lower ( p < 0.05) while the recovery of squalene is significantly higher ( p < 0.05) at NPLF/(silica gel) = 1/20 than those obtained at NPLF/(silica gel) = 1/10. At NPLF/ (silica gel) = 1/10, squalene with high purity and recovery can be obtained in the 2nd fraction in 4.5 h by using large mobile phase flow rate (17.32 mL/min) due to relative large difference in the polarities of squalene and FASEs. As the result, the time required is much shorter than that required to achieve comparable results (in terms of squalene purity and recovery) when classical silica gel column chromatography was used as reported by Gunawan et al. (2008). The identification of compounds in the 2nd fraction was confirmed by HTGC (Fig. 4), TLC (Fig. 5), and GCMS analyses. The GCMS fragmentation of the peak showed a molecular ion (M+) at 410 (relative intensity: 1.09%) and base peak at 69.05. Peaks were also observed at m/z at 149, 137, 121, 109, 95, 81, 55, and 41. These fragmentation patterns were quite similar to those for squalene reported in U.S. National Bureau of Standards library (Stein, 2005). These fractionations obtained in this study give better separation than previous study of Lanzo´n et al. (1994) in which the unsaponifiable matter of both virgin and refined olive oil was fractioned by classical silica gel column chromatography with a unsaponifiable matter to silica gel mass ratio of 1/75, using hexane as eluent. The fraction containing squalene was eluted after aliphatic, steroidal and sesquiterpene hydrocarbons, the purity of squalene was not reported in their study. However, judging from the chromatogram in their study, low squalene purity was obtained. 3.4. Separation and purification of FASEs Using single solvent as mobile phase is more desirable than using mixture of solvents in column chromatography because of easier recovery and reuse of solvent. When pure hexane was employed as the eluent, it was impossible to recovery most FASEs in reasonable short time in the 3rd fraction. It was found that it required 83 h to isolate FASEs with a 83.06% purity and a
97.84% recovery at 23 8C using hexane as the mobile phase. The elution time can be reduced to 20 h by increasing the elution temperature (temperature in the packing region) to 64 8C due to increasing solubility of sample in mobile phase. However, the purity of FASEs obtained was much lower (39.01% purity, 96.15% recovery). Mixture of hexane and ethyl acetate was employed as the eluent in order to successfully recover most FASEs in shorter time in the 3rd fraction. As shown in Table 4, as the polarity increases by increasing the percentage of ethyl acetate in the mobile phase from 1% to 3%, the purity of FASEs in the 3rd fraction decreases from 78.48% to 55.14%, while the corresponding recoveries varies only slightly. One of the important criteria for successful separation in this study is to recover most of the FASEs in the 3rd fraction. FASEs contents are significantly higher ( p < 0.05) when mobile phase contained 1% ethyl acetate as compared to those obtained when 2% and 3% ethyl acetate were used. However, the contents and recoveries of the FFAs and acylglycerols are not significantly different ( p > 0.05) among three studies using different percentages of ethyl acetate. These results agree with previous observation that once the squalene was completely eluted, polarity of the mobile phase must be increased to elute other more polar compounds (Lanzo´n et al., 1994). More silica gel yielded a higher adsorption area available per unit NPLF, which resulted in longer elution time with comparable separation due to a higher amount of mobile phase that was used to displace FASEs. For example, at NPLF to silica gel ratio of 1/20 and with 2% ethyl acetate in the mobile phase, the elution time to obtain the 3rd fraction (FASEs, 69.64 9.99% purity and 95.18 6.45% recovery) was significantly longer (6 h, p < 0.05) than those obtained while using a NPLF to silica mass ratio of 1/10 (2 h). As shown in Table 5, for comparable separation, both elution time and amount of solvent required are much less when using modified silica gel column chromatography as compared to classical silica gel column chromatography. By employing modified silica gel column chromatography to obtain the squalene-rich fraction, the mobile phase volume and elution time required as fractions of those needed in classical silica gel column chromatography are 1/73 and 1/18, respectively. To obtained the FASEs-rich fraction, the corresponding mobile
632
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
Table 4 Effect of mobile phase polarity on the composition of 3rd fractiona Compounds
Mobile phase (hexane/ethyl acetate (v/v)) 99/1
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Others c
98/2
Purity (wt.%)
Recovery (%)
0.61 0.75 78.48 0.86 0.91 0.92 0 3.43 0.70 12.77 1.47 3.80 3.46
1.51 1.91 98.00 2.55 8.27 2.32 0 2.75 0.55 55.45 6.57 1.02 0.93
Time Amount (%) d
b
97/3
Purity (wt.%)
Recovery (%)
Purity (wt.%)
Recovery (%)
1.79 0.16 64.56 2.83 2.79 0.15 0 5.69 2.30 10.03 2.49 15.13 2.46
1.51 1.91 96.60 5.01 8.27 2.32 0 6.45 0.64 61.85 3.79 4.98 1.30
1.51 0.54 55.14 5.49 1.46 0.21 0 6.30 2.96 9.88 2.43 25.70 6.16
5.53 2.40 97.94 1.60 16.45 3.48 0 6.22 3.34 75.71 10.22 10.27 3.31
6h 15.41 0.37
2h 18.05 2.08
1h 21.59 2.22
a Average of three independent experiments. Operation conditions: NPLF/(silica gel) = 1/10 (g/g), 150 mL mobile phase volume, 17.32 mL/min flow rate, and 23 8C elution temperature. b Recovery = {(3rd fraction mass (g) content of the compound in 3rd fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100. c Mainly ketones and aldehydes. d Amount = (3rd fraction mass (g)/NPLF mass (g)) 100.
Table 5 Comparison method on the separation and purification of squalene and FASEs from NPLF of SODD Parameters
Modified column chromatographya
Classical column chromatographyb
Squalene fraction Purity (wt.%) Recovery (%)c Mobile phase Flow rate (mL/min) Time required (h)
99.36 0.13 81.78 0.62 150 mL hexane 17.32 4.5
95.90 93.09 10.96 L hexane 4.06 45
FASEs fraction Purity (wt.%) Recovery (%) Mobile phase Flow rate (mL/min) Time required (h)
78.48 0.86 98.00 2.55 150 mL hexane:ethyl acetate (99:1, v/v) 17.32 6
84.04 63.17 33.13 L hexane 4.06 137
a b c
Average of three independent experiments. Operation conditions: NPLF/(silica gel) = 1/10 (g/g) and 23 8C of elution temperature. Gunawan et al. (2008). Operation conditions: NPLF/(silica gel) = 1/10 (g/g) and 23 8C of elution temperature. Recovery = {(fraction mass (g)) content of the compound in fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100.
Table 6 The fractionation of NPLF obtained by modified silica gel column chromatography with an NPLF of low squalene and FASEs contents (wt.%)a Compounds
Squalene FASEs Tocopherols Free phytosterols FFAs Acylglycerols Others a b c d e f
NPLF
1st fraction
Content (wt.%)
Content (wt.%) and recovery (%)
3.14 0.17 4.29 1.81 0.78 0.14 0.43 0.07 37.25 0.64 4.04 0.61 50.07 2.18d
b
0.30 0.15 (3.41 1.37) NDc (0) ND (0) ND (0) ND (0) ND (0) 99.70 0.15e (74.24 6.16)
2nd fraction
3rd fraction
Content (wt.%) and recovery (%)
Content (wt.%) and recovery (%)
97.37 1.26 (93.50 0.27) ND (0) ND (0) ND (0) ND (0) ND (0) 2.63 1.26f (0.16 0.02)
0.23 0.19 (1.73 0.10) 36.00 1.96 (98.35 0.32) 0.76 0.08 (16.24 1.78) 0.30 0.24 (0.34 0.28) 49.00 1.41 (23.50 2.12) 7.12 2.68 (55.46 0.76) 12.88 2.93f (4.98 1.55)
Average of three independent experiments. Recovery = {(fraction mass (g) content of the compound in fraction (%))/(NPLF mass (g) content of the compound in NPLF (%))} 100%. Not detected. Hydrocarbons, aldehydes, ketones, pesticides, herbicides, and the breakdown products of tocopherols and free phytosterols. Aliphatic, steroidal, and sesquiterpene hydrocarbons. Mainly aldehydes and ketones.
S. Gunawan et al. / Journal of the Chinese Institute of Chemical Engineers 39 (2008) 625–633
phase volume and elution time are 1/221 and 1/23, respectively of those needed in classical silica gel column chromatography. This separation was confirmed in subsequent experiments, carried out under the same conditions except with a NPLF that contains low percentage of squalene and FASEs (Table 6). Squalene content in the 2nd fraction and FASEs recoveries in the 3rd fraction in this case were not significantly different ( p > 0.05), compared to those obtained using a NPLF with high squalene and FASEs contents. Also, it was found that the 3rd fraction was richer in FFAs due to higher content of FFAs in NPLF. 4. Conclusion A new silica gel column chromatography was designed in this work. Unlike classical column chromatography, in this new design, glass tube is equipped with a jacket, a valve to control the flow rate of eluent, a condenser system and a heating mantle. A heater is placed in the bottom part and a uniform temperature in the silica gel packing is achieved by maintaining the mobile phase volume at constant value. From data obtained in this study, modified silica gel column chromatography is a more favorable mode of separation compare to classical silica gel column chromatography. The advantage of the new design presented in this work over classical silica gel column chromatography is that it significantly reduces the amount of solvent and time required to achieve the same degree of separation. Acknowledgement This work was supported by a grant (NSC 94-2214-E011004) provided by the National Science Council of Taiwan.
633
References AOCS Official Method Ca 5a-40, ‘‘Free Fatty Acids,’’ Official Methods and Recommended Practices of the AOCS, Firestone D. (Ed.), American Oil Chemists’ Society, Champaign, IL, U.S.A. (1997). Bondioli, P., C. Mariani, A. Lanzani, E. Fedeli, and A. Muller, ‘‘Squalene Recovery from Olive Oil Deodorizer Distillate,’’ J. Am. Oil Chem. Soc., 70, 763 (1993). Fried, B., Lipids, p. 704, Sherma, J. and B. Fried (Eds.), Marcel Dekker Press, New York, U.S.A. (1996). Giacometti, J., A. Milosevic, and C. Milin, ‘‘Gas Chromatographic Determination of Fatty Acids Contained in Different Lipid Classes after Their Separation by Solid-Phase Extraction,’’ J. Chromatogr. A, 976, 47 (2002). Gunawan, S., S. R. Vali, and Y. H. Ju, ‘‘Purification and Identification of Rice Bran Oil Fatty Acid Steryl and Wax esters,’’ J. Am. Oil Chem. Soc., 83, 449 (2006). Gunawan, S., N. S. Kasim, and Y. H. Ju, ‘‘Separation and Purification of Squalene from Soybean Oil Deodorize Distillate,’’ Sep. Purif. Technol., 60, 128 (2008). Hamilton, R. J. and P. A. Sewell, Introduction to High Performance Liquid Chromatography, 2nd Ed., p. 163, Chapman and Hall Ltd., London, U.K. (1982). Lacaze, J.P.C.L., L. A. Stobo, E. A. Turrell, and M. A. Quilliam, ‘‘Solid-phase Extraction and Liquid Chromatography-mass Spectrometry for The Determination of Free Fatty Acids in Shellfish,’’ J. Chromatogr. A, 1145, 51 (2007). Lanzo´n, A., T. Albi, A. Cert, and J. Gracia´n, ‘‘The Hydrocarbon Fraction of Virgin Olive Oil and Changes Resulting from Refining,’’ J. Am. Oil Chem. Soc., 71, 285 (1994). Moreda, W., M. C. Perez-Camino, and A. Cert, ‘‘Gas and Liquid Chromatography of Hydrocarbons in Edible Vegetable Oils,’’ J. Chromatogr. A, 936, 159 (2001). Snyder, L. R. and J. J. Kirkland, Introduction to Modern Liquid Chromatography, 2nd Ed., p. 321, John Wiley and Sons Inc., New York, U.S.A. (1979). Stein, S.E., ‘‘Mass Spectra,’’ NIST Chemistry WebBook, NIST Standard Reference Database Number 69, Linstrom, P. J. and W. G. Mallard (Eds.), National Institute of Standards and Technology, Gaithersburg MD, U.S.A. (2005). http://webbook.nist.gov/chemistry (accessed September 23, 2007).