2349
J. Sep. Sci. 2014, 37, 2349–2356
Hongmin Wei Jinjuan Yang Hanqi Zhang Yuhua Shi College of Chemistry, Jilin University, Changchun, P.R. China Received March 22, 2014 Revised May 20, 2014 Accepted June 9, 2014
Research Article
Ultrasonic nebulization extraction assisted dispersive liquid–liquid microextraction followed by gas chromatography for the simultaneous determination of six parabens in cosmetic products A simple, rapid, and efficient method of ultrasonic nebulization extraction assisted dispersive liquid–liquid microextraction was developed for the simultaneous determination of six parabens in cosmetic products. The analysis was carried out by gas chromatography. Water was used as the dispersive solvent instead of traditional organic disperser. The experimental factors affecting the extraction yield, such as the extraction solvent and volume, extraction time, dispersive solvent and volume, ionic strength, and centrifuging condition were studied and optimized in detail. The limit of detections for the target analytes were in the range of 2.0–9.5 g/g. Good linear ranges were obtained with the coefficients ranging from 0.9934 to 0.9969. The proposed method was successfully applied to the analysis of six parabens in 16 cosmetic products. The recoveries of the target analytes in real samples ranged from 81.9 to 108.7%, and the relative standard deviations were 99.5%. Supporting Information Fig. S1 illustrates the molecular structures of parabens. The reagents including n-pentanol, n-hexanol, n-octanol, n-octane, and sodium chloride (NaCl) were purchased from Guangfu Chemical Institute (Tianjin, China). The microsyringe was obtained from Anting Corporation (Shanghai, China). The centrifuge was purchased from Xiangyi Instruments Corporation (Hunan, China). The high-purity water was obtained by Milli-Q water purification system (Millipore, USA). Stock standard solutions of the target analytes (1.0 mg/mL) were prepared separately in methanol and stored in volumetric flask. Standard solution at the concentration of 100 g/mL was obtained by appropriate dilution of the stock standard solutions with methanol. The above solutions were stored in the dark at 4⬚C. Different cosmetic products including face masks, moisture cream, face cream, hair conditioner, hand cream, and glycerin were purchased from a local supermarket in Changchun (Jilin, China) and kept at room temperature. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.2 Instruments The UNE-DLLME system was assembled in our laboratory and the schematic diagram was given in Fig. 1. An ultrasonic humidifier (Beijing Branson Ultrasound, China) working at 1.7 MHz with a maximum output power of 35 W was used as the ultrasonic source of energy. An in-house designed glass flask (100 mL, 5 cm id) was used as the extraction vessel. The bottom port of the extraction vessel with the same size as that of the piezocrystal was sealed by polyvinyl chloride film. The space between bottom port and piezocrystal was filled with coupling water. The extraction solvent and dispersive solvent was added by the other port of the extraction vessel.
2.3 UNE-DLLME Fifty milligrams of cosmetic sample, 200 L of n-octanol, and 3 mL of water were added into the extraction vessel. When the ultrasonic nebulizer at the power of 35 W was switched on, the “ultrasonic fountain” sprayed out from the bottom of the extraction vessel with the assistance of ultrasonic energy. The cosmetic sample and n-octanol were dispersed evenly with the help of the ultrasonic fountain to obtain the aerosol. After 5 min, the ultrasonic nebulizer was switched off and the aerosol disappeared gradually. Then the mixed solution in the extraction vessel was transferred into a 10 mL centrifuge tube. The extraction vessel was washed twice with 1 mL of water and then the washings were combined. One gram of NaCl was added to the mixed solution and the centrifuge tube was shaken to dissolve NaCl. And then the mixed solution turned clear and the organic phase was gathered on the surface of aqueous phase. The mixed solution was centrifuged for 5 min at 4000 rpm. After that, the mixed solution was divided into different layers, the residue that was not soluble in n-octanol, and water was present between the organic phase and aqueous phase. The organic phase was removed through dipping by a glass capillary tube (100 mm length and 1.0 mm id), and then an amount of 0.3 L was withdrawn by a microsyringe and injected into GC. The experimental procedure is shown in Fig. 1. The peak areas of the target analytes were used to evaluate the extraction efficiency.
2.4 GC analysis The GC analysis was carried out on HP4890 (Hewlett– Packard, Palo Alto, USA) with flame ionization detector. As the six target analytes have similar chemical structure, a capillary column with high separation efficiency and reasonable polarity should be selected. The separation was performed on a CP-SIL 24CB fused silica capillary column (25 m × 0.53 mm, 1.0 m film thickness). Considering the properties of parabens and the complex matrix of cosmetics, the GC conditions were programmed as follows: initial oven temperature was 150⬚C held for 2 min, programmed to 220⬚C www.jss-journal.com
Gas Chromatography
J. Sep. Sci. 2014, 37, 2349–2356
2351
Figure 1. The schematic diagrams of UNE-DLLME system and the experiment procedure.
Figure 2. (A) Effect of extraction solvent. Extraction solvent volume, 200 L; extraction time, 5 min; dispersive solvent and volume, 3 mL of water; ionic strength, 1.0 g NaCl; centrifuging condition, 5 min at 4000 rpm. (B) Effect of extraction solvent volume. Extraction solvent, n-octanol; extraction time, 5 min; dispersive solvent and volume, 3 mL of water; ionic strength, 1.0 g NaCl; centrifuging condition, 5 min at 4000 rpm.
at 25⬚C/min, where it was held for 5 min, and finally programmed to 240⬚C at 20⬚C/min, where it was held for 5 min. The temperatures of injector and detector were set at 260 and 240⬚C, respectively. High purity nitrogen (99.999%) was used as the carrier gas at a flow rate of 1.0 mL/min. The injection volume was 0.3 L at a splitless mode.
3 Results and discussion 3.1 Optimization of UNE-DLLME Several operation parameters including the types and volume of extraction solvent and dispersive solvent, extraction time, ion strength, and centrifuging condition were optimized. The optimization experiments were carried out on a real cosmetic sample spiked each paraben at 200 g/g, and the sample is free of parabens. All experiments were performed in triplicate.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.1.1 Extraction solvent In DLLME, the extraction solvent should have the characteristics of insoluble in water, good chromatography behavior, and high extraction capability for the target compounds [25]. Considering the above factors and the highly toxic nature of chlorinated extraction solvent, four low-density organic solvents including n-pentanol, n-hexanol, n-octanol, and n-octane were investigated. The extraction efficiency obtained with the extraction solvents are shown in Fig. 2A. The results show that the extraction efficiency obtained with n-octanol was higher than those obtained with n-pentanol and n-hexanol. Moreover, as the polarity of n-octanol was higher than n-octane, the relatively high polarity of Mep and Etp in the six parabens cannot be extracted in n-octane because their polarities are not similar to that of n-octane. The other four parabens can be extracted in n-octane, but the extraction efficiency was lower. Finally, n-octanol was selected as the extraction solvent.
www.jss-journal.com
2352
J. Sep. Sci. 2014, 37, 2349–2356
H. Wei et al.
Figure 3. The chromatograms of (A) face masks (FM3) and (B) face cream (FC1). In all cases, the lower chromatograms belong to real samples and the upper chromatograms belong to the samples spiked each paraben at 200 g/g.
3.1.2 Extraction solvent volume The effect of extraction solvent volume on the extraction efficiency is shown in Fig. 2B. It can be found that the peak areas of the target analytes decrease with the increase of extraction solvent volume from 200 to 500 L. It might be the reason that the organic phase volume increased with the increase of extraction solvent volume, resulting in a decreased concentration of the target analytes. The peak areas of the target analytes in some real samples might be overloaded when 100 L of extraction solvent was chosen. In fact, parabens in cosmetic products are not trace levels in which the legal
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
limits is 0.4% for single ester and 0.8% for the mixtures in cosmetics [10]. Therefore, 200 L of n-octanol was used in the work. 3.1.3 Extraction time The extraction time ranging from 3 to 20 min was investigated. The results are shown in Supporting Information Fig. S2. It can be seen that the peak areas of the target analytes increase from 3 to 5 min, and the signals remain almost constant after 5 min. It illustrated that the dynamic distribution equilibrium between the target analytes and extraction
www.jss-journal.com
Gas Chromatography
J. Sep. Sci. 2014, 37, 2349–2356
2353
Table 1. Calibration curves and detection limits
Analytes
Linear range (g/g)
Regression equationa)
Correlation coefficient (r)
LOD (g/g)
Mep Etp Prp iBup Bup iAmp
26–1017 28–1012 29–1019 15–1001 30–1022 38–1038
A = 5201C − 60 290 A = 6232C − 505 88 A = 7012C − 100 472 A = 6791C − 105 831 A = 7517C − 118 864 A = 7759C − 117 519
0.9969 0.9967 0.9966 0.9938 0.9962 0.9934
5.7 5.5 2.0 2.8 2.2 9.5
a) A and C are the peak areas and concentration of the analytes, respectively.
solvent was not reached when it was
J. Sep. Sci. 2014, 37, 2349–2356
Hongmin Wei Jinjuan Yang Hanqi Zhang Yuhua Shi College of Chemistry, Jilin University, Changchun, P.R. China Received March 22, 2014 Revised May 20, 2014 Accepted June 9, 2014
Research Article
Ultrasonic nebulization extraction assisted dispersive liquid–liquid microextraction followed by gas chromatography for the simultaneous determination of six parabens in cosmetic products A simple, rapid, and efficient method of ultrasonic nebulization extraction assisted dispersive liquid–liquid microextraction was developed for the simultaneous determination of six parabens in cosmetic products. The analysis was carried out by gas chromatography. Water was used as the dispersive solvent instead of traditional organic disperser. The experimental factors affecting the extraction yield, such as the extraction solvent and volume, extraction time, dispersive solvent and volume, ionic strength, and centrifuging condition were studied and optimized in detail. The limit of detections for the target analytes were in the range of 2.0–9.5 g/g. Good linear ranges were obtained with the coefficients ranging from 0.9934 to 0.9969. The proposed method was successfully applied to the analysis of six parabens in 16 cosmetic products. The recoveries of the target analytes in real samples ranged from 81.9 to 108.7%, and the relative standard deviations were 99.5%. Supporting Information Fig. S1 illustrates the molecular structures of parabens. The reagents including n-pentanol, n-hexanol, n-octanol, n-octane, and sodium chloride (NaCl) were purchased from Guangfu Chemical Institute (Tianjin, China). The microsyringe was obtained from Anting Corporation (Shanghai, China). The centrifuge was purchased from Xiangyi Instruments Corporation (Hunan, China). The high-purity water was obtained by Milli-Q water purification system (Millipore, USA). Stock standard solutions of the target analytes (1.0 mg/mL) were prepared separately in methanol and stored in volumetric flask. Standard solution at the concentration of 100 g/mL was obtained by appropriate dilution of the stock standard solutions with methanol. The above solutions were stored in the dark at 4⬚C. Different cosmetic products including face masks, moisture cream, face cream, hair conditioner, hand cream, and glycerin were purchased from a local supermarket in Changchun (Jilin, China) and kept at room temperature. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.2 Instruments The UNE-DLLME system was assembled in our laboratory and the schematic diagram was given in Fig. 1. An ultrasonic humidifier (Beijing Branson Ultrasound, China) working at 1.7 MHz with a maximum output power of 35 W was used as the ultrasonic source of energy. An in-house designed glass flask (100 mL, 5 cm id) was used as the extraction vessel. The bottom port of the extraction vessel with the same size as that of the piezocrystal was sealed by polyvinyl chloride film. The space between bottom port and piezocrystal was filled with coupling water. The extraction solvent and dispersive solvent was added by the other port of the extraction vessel.
2.3 UNE-DLLME Fifty milligrams of cosmetic sample, 200 L of n-octanol, and 3 mL of water were added into the extraction vessel. When the ultrasonic nebulizer at the power of 35 W was switched on, the “ultrasonic fountain” sprayed out from the bottom of the extraction vessel with the assistance of ultrasonic energy. The cosmetic sample and n-octanol were dispersed evenly with the help of the ultrasonic fountain to obtain the aerosol. After 5 min, the ultrasonic nebulizer was switched off and the aerosol disappeared gradually. Then the mixed solution in the extraction vessel was transferred into a 10 mL centrifuge tube. The extraction vessel was washed twice with 1 mL of water and then the washings were combined. One gram of NaCl was added to the mixed solution and the centrifuge tube was shaken to dissolve NaCl. And then the mixed solution turned clear and the organic phase was gathered on the surface of aqueous phase. The mixed solution was centrifuged for 5 min at 4000 rpm. After that, the mixed solution was divided into different layers, the residue that was not soluble in n-octanol, and water was present between the organic phase and aqueous phase. The organic phase was removed through dipping by a glass capillary tube (100 mm length and 1.0 mm id), and then an amount of 0.3 L was withdrawn by a microsyringe and injected into GC. The experimental procedure is shown in Fig. 1. The peak areas of the target analytes were used to evaluate the extraction efficiency.
2.4 GC analysis The GC analysis was carried out on HP4890 (Hewlett– Packard, Palo Alto, USA) with flame ionization detector. As the six target analytes have similar chemical structure, a capillary column with high separation efficiency and reasonable polarity should be selected. The separation was performed on a CP-SIL 24CB fused silica capillary column (25 m × 0.53 mm, 1.0 m film thickness). Considering the properties of parabens and the complex matrix of cosmetics, the GC conditions were programmed as follows: initial oven temperature was 150⬚C held for 2 min, programmed to 220⬚C www.jss-journal.com
Gas Chromatography
J. Sep. Sci. 2014, 37, 2349–2356
2351
Figure 1. The schematic diagrams of UNE-DLLME system and the experiment procedure.
Figure 2. (A) Effect of extraction solvent. Extraction solvent volume, 200 L; extraction time, 5 min; dispersive solvent and volume, 3 mL of water; ionic strength, 1.0 g NaCl; centrifuging condition, 5 min at 4000 rpm. (B) Effect of extraction solvent volume. Extraction solvent, n-octanol; extraction time, 5 min; dispersive solvent and volume, 3 mL of water; ionic strength, 1.0 g NaCl; centrifuging condition, 5 min at 4000 rpm.
at 25⬚C/min, where it was held for 5 min, and finally programmed to 240⬚C at 20⬚C/min, where it was held for 5 min. The temperatures of injector and detector were set at 260 and 240⬚C, respectively. High purity nitrogen (99.999%) was used as the carrier gas at a flow rate of 1.0 mL/min. The injection volume was 0.3 L at a splitless mode.
3 Results and discussion 3.1 Optimization of UNE-DLLME Several operation parameters including the types and volume of extraction solvent and dispersive solvent, extraction time, ion strength, and centrifuging condition were optimized. The optimization experiments were carried out on a real cosmetic sample spiked each paraben at 200 g/g, and the sample is free of parabens. All experiments were performed in triplicate.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.1.1 Extraction solvent In DLLME, the extraction solvent should have the characteristics of insoluble in water, good chromatography behavior, and high extraction capability for the target compounds [25]. Considering the above factors and the highly toxic nature of chlorinated extraction solvent, four low-density organic solvents including n-pentanol, n-hexanol, n-octanol, and n-octane were investigated. The extraction efficiency obtained with the extraction solvents are shown in Fig. 2A. The results show that the extraction efficiency obtained with n-octanol was higher than those obtained with n-pentanol and n-hexanol. Moreover, as the polarity of n-octanol was higher than n-octane, the relatively high polarity of Mep and Etp in the six parabens cannot be extracted in n-octane because their polarities are not similar to that of n-octane. The other four parabens can be extracted in n-octane, but the extraction efficiency was lower. Finally, n-octanol was selected as the extraction solvent.
www.jss-journal.com
2352
J. Sep. Sci. 2014, 37, 2349–2356
H. Wei et al.
Figure 3. The chromatograms of (A) face masks (FM3) and (B) face cream (FC1). In all cases, the lower chromatograms belong to real samples and the upper chromatograms belong to the samples spiked each paraben at 200 g/g.
3.1.2 Extraction solvent volume The effect of extraction solvent volume on the extraction efficiency is shown in Fig. 2B. It can be found that the peak areas of the target analytes decrease with the increase of extraction solvent volume from 200 to 500 L. It might be the reason that the organic phase volume increased with the increase of extraction solvent volume, resulting in a decreased concentration of the target analytes. The peak areas of the target analytes in some real samples might be overloaded when 100 L of extraction solvent was chosen. In fact, parabens in cosmetic products are not trace levels in which the legal
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
limits is 0.4% for single ester and 0.8% for the mixtures in cosmetics [10]. Therefore, 200 L of n-octanol was used in the work. 3.1.3 Extraction time The extraction time ranging from 3 to 20 min was investigated. The results are shown in Supporting Information Fig. S2. It can be seen that the peak areas of the target analytes increase from 3 to 5 min, and the signals remain almost constant after 5 min. It illustrated that the dynamic distribution equilibrium between the target analytes and extraction
www.jss-journal.com
Gas Chromatography
J. Sep. Sci. 2014, 37, 2349–2356
2353
Table 1. Calibration curves and detection limits
Analytes
Linear range (g/g)
Regression equationa)
Correlation coefficient (r)
LOD (g/g)
Mep Etp Prp iBup Bup iAmp
26–1017 28–1012 29–1019 15–1001 30–1022 38–1038
A = 5201C − 60 290 A = 6232C − 505 88 A = 7012C − 100 472 A = 6791C − 105 831 A = 7517C − 118 864 A = 7759C − 117 519
0.9969 0.9967 0.9966 0.9938 0.9962 0.9934
5.7 5.5 2.0 2.8 2.2 9.5
a) A and C are the peak areas and concentration of the analytes, respectively.
solvent was not reached when it was
