Article pubs.acs.org/ac
Miniscale Liquid−Liquid Extraction Coupled with Full Evaporation Dynamic Headspace Extraction for the Gas Chromatography/Mass Spectrometric Analysis of Polycyclic Aromatic Hydrocarbons with 4000-to-14 000-fold Enrichment Christina Shu Min Liew,†,‡ Xiao Li,§ and Hian Kee Lee*,†,‡,∥ †
NUS Graduate School for Integrative Sciences and Engineering, National University of Singapore, Centre for Life Sciences #05-01, 28 Medical Drive, Singapore 117456, Singapore ‡ Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore § Gerstel LLP, The Alpha #02-18, 10 Science Park Road, Singapore 117684, Singapore ∥ National University of Singapore Environmental Research Institute, T-Lab Building, #02-01, 5A Engineering Drive 1, Singapore 117411, Singapore S Supporting Information *
ABSTRACT: A new sample preparation approach of combining a miniscale version of liquid−liquid extraction (LLE), termed miniscale-LLE (msLLE), with automated full evaporation dynamic headspace extraction (FEDHS) was developed. Its applicability was demonstrated in the extraction of several polycyclic aromatic hydrocarbons (PAHs) (acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, and pyrene) from aqueous samples. In the first step, msLLE was conducted with 1.75 mL of nhexane, and all of the extract was vaporized through a Tenax TA sorbent tube via a nitrogen gas flow, in the FEDHS step. Due to the stronger π−π interaction between the Tenax TA polymer and PAHs, only the latter, and not n-hexane, was adsorbed by the sorbent. This selectivity by the Tenax TA polymer allowed an effective concentration of PAHs while eliminating n-hexane by the FEDHS process. After that, thermal desorption was applied to the PAHs to channel them into a gas chromatography/mass spectrometric (GC/MS) system for analysis. Experimental parameters affecting msLLE (solvent volume and mixing duration) and FEDHS (temperature and duration) were optimized. The obtained results achieved low limits of detection (1.85−3.63 ng/L) with good linearity (r2 > 0.9989) and high enrichment factors ranging from 4200 to 14 100. The optimized settings were applied to the analysis of canal water sampled from an industrial area and tap water, and this methodology was compared to stir-bar sorptive extraction (SBSE). This innovative combined extraction−concentration approach proved to be fast, effective, and efficient in determining low concentrations of PAHs in aqueous samples.
L
results. Notwithstanding that, each has drawbacks. For example, in SDME, the drop suspended from the syringe needle can be dislodged and the droplet volume is limited.6 In SBME, the manual preparation of the solvent bar is inconvenient and is subjected to variation, especially among different users.4 In addition, there could also be potential loss of solvent from the bar during vigorous extraction.7 In the conventional DLLME technique, chlorinated solvents are used,8 which are relatively more toxic than their nonchlorinated counterparts.9 In response, there has been much attention being paid recently to the use of greener, nonhalogenated solvents in this procedure such as 1-octanol,10,11 n-hexanol,12 n-hexane,13 or
iquid−liquid extraction (LLE) has been one of the most widely used extraction procedures since its introduction in the 19th century. It continues to be very popular in contemporary analytical science, although there are concerns about its heavy consumption of organic solvents and laborintensiveness.1 In addressing the latter shortcomings, various miniaturized solvent extraction procedures, generally termed liquid-phase microextraction (LPME),2 have been developed as alternatives to LLE, including single-drop microextraction (SDME),2,3 solvent bar microextraction (SBME),4 and dispersive liquid−liquid microextraction (DLLME).5 The common features of these LPME techniques include ease of use and consumption of only small volumes of solvent (μL range). These miniaturized techniques integrate extraction and preconcentration into a single step and have been demonstrated to exhibit acceptable sensitivities and reproducible © 2016 American Chemical Society
Received: May 26, 2016 Accepted: August 18, 2016 Published: August 18, 2016 9095
DOI: 10.1021/acs.analchem.6b02056 Anal. Chem. 2016, 88, 9095−9102
Article
Analytical Chemistry
Figure 1. Schematic of msLLE-FEDHS with TDU-GC/MS. The transfer of sorbent tube to TDU for desorption into GC/MS is also automated.
toluene.14,15 Nonetheless, in both kinds of DLLME, an additional amount of solvents such as acetonitrile or acetone are required for dispersion, and not all of the extract from DLLME can be retrieved and injected for analysis, leading to a potential loss in sensitivity. Headspace analysis techniques, such as static headspace (SHS) and dynamic headspace (DHS), use vapor as an extraction phase where the analytes partition between the solid/liquid sample and the vapor space above the sample.16 SHS works by direct injection of the headspace of a sample using a gastight syringe into the gas chromatograph. On the other hand, in DHS, the headspace of a sample is heated and purged with a dual needle system using a continuous inert gas to a cryogenic trap or sorbent where the volatile analytes are immobilized.17 Another approach of performing DHS is to purge the wet sample’s headspace through the sorbent until the sample is dry, known as full evaporation DHS (FEDHS), first demonstrated by Ochiai et al.,18 based on the full evaporation technique (FET) introduced by Markelov and Guzowski.19 In this way, FEDHS can minimize the bias toward the more volatile compounds present in DHS, providing a more uniform recovery of the compounds. The use of DHS or FEDHS is already well-established in the fragrance20,21 and flavor industries.22−24 However, this FEDHS technique has not yet been applied to environmental studies due to the fact that, in this field, one usually needs to deal with trace amounts of analytes of interest in primarily milliliter volumes of aqueous samples. This would then require long purging duration to vaporize the aqueous sample to obtain sufficient sensitivity and an even longer drying step subsequently in order to eliminate moisture interference.25 For instance, Ochiai et al. performed FEDHS to detect odor compounds in 100 μL of aqueous sample in 30 min.18 The sensitivity of this method was limited by the small sample volume. This was overcome by the use of multiple adsorbent traps for a single sample on a DHS module with an additional vacuum unit and with the sample volume increased to 1 mL.26 However, the total DHS runtime was prolonged to 150 min for each sample, in order to prevent an introduction of the moisture. It is therefore desirable, when dealing with milliliter volumes of aqueous samples, to perform an extraction first before the organic solvent extract is subjected
to FEDHS so as to obtain lower limits of detection and quantification and to eliminate the need for a long purging and drying process. To enable this, an additional fast, preferably solvent-minimized, LLE step can be included before FEDHS for aqueous sample analysis. The organic extract from the LLE step is transferred to an automated DHS module for the solvent to be fully evaporated. Using a sorbent in FEDHS that has greater affinity toward the analytes of interest than the solvent allows the concentration of analytes to take place in the sorbent while the solvent breaks through it. The result is that FEDHS replaces the evaporation and reconstitution steps needed in traditional LLE with lower analyte loss. In this study, instead of microscale LLE (using μL volume of solvent), a miniversion of LLE was designed, termed miniscaleLLE (msLLE), which involved the consumption of
