Fizzy Extraction of Volatile and Semivolatile ... - ACS Publications

Subscriber access provided by Northern Illinois University

Article

Fizzy Extraction of Volatile and Semi-Volatile Compounds into the Gas Phase Cheng-Hao Chang, and Pawel L Urban Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b02074 • Publication Date (Web): 09 Aug 2016 Downloaded from http://pubs.acs.org on August 9, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Fizzy Extraction of Volatile and Semi-Volatile Compounds into the Gas Phase

Cheng-Hao Chang, Pawel L. Urban*

Department of Applied Chemistry, National Chiao Tung University 1001 University Rd., Hsinchu, 300, Taiwan

Word count: 4264 – 153 + 500 (Table 1) + 250 (Figure 1) + 250 (Figure 2) + 500 (Figure 3) + 250 (Figure 4) + 250 (Figure 5) = 6111

* Corresponding author: P.L. Urban ([email protected])

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

ABSTRACT

Extraction of volatile and semi-volatile compounds from liquid matrices with high yields, and transferring the extracts to detectors in real time, is challenging. Common extraction procedures involve heating the samples to release the analytes to the gas phase, and – in some cases – trapping the gas-phase analytes into sorbents or containers. Here, we propose a new method for fast extraction of volatile and semivolatile compounds from liquid matrices. This method involves dissolution of a carrier gas in the liquid sample by applying a moderate overpressure (∼ 150 kPa), and stirring the sample. An abrupt decompression of the extraction chamber leads to effervescence. In this step, many bubbles are instantly formed in the sample matrix. The dissolved carrier gas as well as dissolved volatiles are liberated into the headspace of the extraction chamber within a short period of time (few seconds). The gaseous effluent of the extraction chamber is immediately transferred to the on-line detector; in this case, an atmospheric pressure chemical ionization interface of a triple quadrupole mass spectrometer. The fast release of the gas-phase extract gives rise to a high signal recorded by the detector; several times higher than the signal recorded during direct infusion of headspace vapors without fizzy extraction. This feature provides the means to detect and quantify analytes present in solutions in a short period of time. Here we show that fizzy extraction is suitable for analysis of volatile/semi-volatile compounds present in various samples, including those containing complex matrices.


Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION

The technology behind carbonation of water dates back to the 18th century when Joseph Priestley developed and tested an apparatus for infusing carbon dioxide into water.1 According to the Henry’s Law,2 at fixed conditions, partial vapor pressure (Pi) of a volatile component (i) is proportional to its quantity (Qi) in the solution: ܲ୧ = ‫ܭ‬୧ ܳ୧

(eq. 1)

where Ki is a constant. The solubility of a gas in a liquid solvent (e.g. water) can be controlled by setting its partial pressure. On the industrial scale, the process of carbonation is implemented during the production of fizzy drinks. Carbon dioxide gas is injected to the beverage at an elevated pressure. Shaking and the turbulence caused by gas injection speeds up the dissolution process. The liquid/gas mixture remains stable as long as the elevated pressure is maintained in the bottle. When the bottle is opened, some of the dissolved carbon dioxide is immediately released. A characteristic noise foretells what is to come. The bubbling liquid rises toward the bottle’s neck. If one shakes a half-full bottle of soda (or better – stores it in a warm place – not difficult when going out for an outdoor trip in the summer) and opens it again, bubbling is more vigorous. The burgeoning bubbles promote the release of the aroma compounds, contributing to the organoleptic properties of the fizzy drink.3 Extraction is one of the most common, but also one of the most challenging, steps in analytical protocols. Type and quality of extraction directly influences the entire analysis. There exist a number of extraction techniques, for example: liquid-liquid extraction; solidphase extraction; liquid-gas extraction; solid-phase microextraction; supercritical fluid



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

extraction (SFE); and Soxhlet extraction. These techniques have various strengths in terms of selectivity, sensitivity, reproducibility, and throughput. However, the remaining problems with the extraction of volatile and semi-volatile compounds dissolved in liquid matrices include: (i) the requirement to heat up the samples to volatilize the solutes (if their vapor pressures are too low); (ii) necessity to use solid-phase sorbents to move the volatile analytes from the sample to the detector; (iii) considerable share of (manual) mechanical operations; (iv) restricted selectivity of different extraction modes. Thus, there is a need to search for new ways to separate volatile/semi-volatile compounds from complex matrices prior to detection using common gas-phase techniques (e.g., mass spectrometry (MS), gas chromatography). These new ways should provide high extraction efficiency to warrant high analytical sensitivity. On the other hand, the non-volatile matrix components should not be extracted from complex matrices to prevent spectral/chromatographic interferences, and to minimize contamination of the detection systems (e.g., separation columns, ion optics). Here we propose a new extraction scheme, which takes advantage of the effervescence phenomenon induced by sudden decompression of gas/liquid solution. This scheme involves dissolution of a carrier gas in the liquid sample by applying a pressurized gas and stirring. An abrupt decompression of the extraction chamber leads to effervescence. In that step, many bubbles are suddenly formed in the sample. The dissolved carrier gas, as well as the dissolved volatiles, are liberated into the headspace of the extraction chamber within a short period of time (Figure 1).


Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


EXPERIMENTAL SECTION

Samples Caffeine, ethanol (GC grade, >99.8%), ethyl acetate, ethyl octanoate, hexyl acetate, octyl acetate, and limonene were purchased from Sigma-Aldrich (St. Louis, MO, USA). Water (chromatography grade) was from Merck Millipore (Darmstadt, Germany). Fresh lime juice was prepared by squeezing half of lime fruit (Citrus latifolia) by hand. Lime peel was cut by knife, dried at 60 ºC for 4 h. After cooling, the dry pieces were ground manually in a ceramic mortar. A small amount of the resulting powder (~ 1.1 g) was mixed with 10 mL ethanol. The mixture was shaken by hand, and left to equilibrate for ~ 1 h. The liquid phase was diluted with water and ethanol (95:5, v/v) prior to subsequent fizzy extraction.

Fizzy extraction system The extraction system was assembled on an aluminum breadboard (30 × 45 cm; Newport, Irvine, CA, USA). The extraction cell (20-mL screw top headspace glass vial with septum cap; Thermo Fisher Scientific, Waltham, MA, USA) was installed in a protective casing with an acrylic glass observation window (Figure 2). The vial’s septum was fitted with two pieces of tubing: (i) to supply carbon dioxide; and (ii) to transfer the gaseous extract to the detector (Figure 1A). A small direct current motor (∼ 3 V, RF-3020; KwongWah, Taichung, Taiwan) with an attached flexible polymer spindle (length: ∼ 30 mm; OD: ∼ 2 mm) was installed under the cap’s septum to stir the sample during carbonation and decompression. Carbon dioxide



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

carrier gas was supplied from a cylinder at a pressure ∼ 150 kPa (unless otherwise noted). A normally closed solenoid valve (Valve 1; UWS-08; Thai Xin Machinery, Kaohsiung, Taiwan) was used to control the flow of the carrier gas. A normally closed pinch valve (Valve 2; P/N 390NC12150; Asco, Florham Park, NJ, USA) was used to control the flow of gas out of the extraction cell (during detection and decompression). The valves were actuated with aid of a relay board controlled by a Uno32 microcontroller board (chipKIT; Digilent, Pullman, WA, USA). The extraction program was written in C language (Mpide 0150 software), and uploaded to the microcontroller over USB. Table 1 lists the main steps in the extraction routine. All the tests were carried out at room temperature (∼ 25-28 °C). Due to application of elevated pressure, precautions were taken to minimize the risk of injury. Two-layer acrylic shield was installed in front of the extraction chamber. Safety glasses were always worn. A carbon dioxide meter was placed in the proximity of the experimental setup to make sure that the carbon dioxide concentration in the ambient air is always below 1000 ppm.

Mass spectrometry The extraction chamber was coupled via Valve 2 and polytetrafluoroethylene tubing (length: 42 cm; OD: 1.5 mm; ID: 1.0 mm) with the atmospheric pressure chemical ionization (APCI) interface (Duis; Shimadzu, Tokyo, Japan) of a triple quadrupole mass spectrometer (LCMS8030; Shimadzu). In most experiments, the mass spectrometer was operated in the positiveion Q3 full-spectrum scan mode. The voltage applied to the APCI emitter was 4.5 kV, while the temperature of the interface was set to 400 °C. Unless noted otherwise, mass spectra were recorded for the m/z range 50-200. Extracted ion currents were exported to ASCII files, and


Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


treated in the Origin software (version 9.0; OriginLab, Northampton, MA, USA). Peak areas were measured by HVL curve fitting and integration in PeakFit software (version 4.12; SeaSolve Software, San Jose, CA, USA).

RESULTS AND DISCUSSION

Proof of concept Fizzy extraction comprises five stages (Table 1, Figure 1B): 1) Recording detector baseline. 2) Flushing the headspace of liquid sample in the sample chamber with a carrier gas (here: carbon dioxide). 3) Pressurizing the carrier gas (such as carbon dioxide) inside the sample chamber. The sample can be stirred during pressurization in order to accelerate dissolution of the carrier gas in the sample matrix. 4) Equilibration of the system (to stop flow of carrier gas). 5) Abrupt depressurization of the sample chamber, and transferring the extracted gasphase analyte species toward a detector by setting a pressure difference (sample chamber headspace vs. detector). In this step, the carrier gas is released from the sample matrix, extracting the analyte species dissolved in that matrix. The extracted sample can be any liquid capable of solubilizing the carrier gas used in the extraction (such as carbon dioxide), and amenable to effervescence.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

To demonstrate operation of fizzy extraction, a test sample was loaded into a 20-mL vial (extraction chamber; Figures 1A and 1B). During the extraction procedure, carbon dioxide was pumped to the headspace of the extraction chamber for 60 s in order to remove the residues of ambient air and volatile contaminants (Figure 2 and Movie S1). In this stage, Valve 1 was open. Subsequently, the gas outlet of the extraction chamber was closed (by Valve 2), and the pressure inside the extraction chamber was allowed to rise up to ∼ 1.5 bar (150 kPa) above the atmospheric pressure. The spindle motor (installed inside the extraction chamber) was switched on to facilitate dissolution of the carrier gas in the sample matrix during 60 s. Next, the supply of the carrier gas was cut off (by Valve 1), and the extraction chamber was depressurized by opening the gas outlet of the extraction chamber (by Valve 2). The gaseous extract was transferred to the APCI-QQQ-MS system. During the first 2 s from opening the pinch valve, the gaseous extract moved to the detector due to the pressure difference between the headspace of the extraction chamber and the detector. In order to remove the residues of the gaseous extract from the extraction chamber, Valve 1 was opened again (for 28 s). Thus, the carrier gas pushed the extracted analytes toward the APCI-MS detection system.

Extraction performance Importantly, when the depressurization started, numerous gas bubbles were observed in the sample, revealing the effervescence phenomenon (Figure 2). Immediately after opening Valve 2 at the start of the depressurization, a signal was recorded (cf. Figure 1C). Various samples produced signals confirming extraction of the main volatile components (Figure 3). They


Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


included standard solutions of ethyl octanoate, hexyl acetate, octyl acetate, limonene; as well as lime juice and lime peel extract. The identity of the limonene signal (m/z 137, when analyzing real samples) was confirmed by MS/MS analyses (Figure S1). The intra-day repeatability ranged from 6 to 19% (satisfactory for a prototype; Table S1). The recovery of a model analyte (limonene) was estimated by infusing a liquid sample before and after fizzy extraction to APCI-MS by syringe pump. The intensity of a limonene ion fragment (at the m/z 81) recorded during multiple reaction monitoring decreased by 33% (n = 3; Figure S2). This value is an estimate of the fizzy extraction recovery (one cycle). Because only a fraction of the analyte is extracted in one cycle, it is possible to re-extract the same sample. However, the amount of the extracted analyte would decrease in every cycle. In one experiment, it was found that the intensity of the recorded ion signal correlates with the concentration of the analyte molecule present in the liquid sample (either in full spectrum scan or multiple reaction monitoring; Figure S3). This result suggests that fizzy extraction can be used in quantitative analyses. The highest concentration of limonene presented in this plot is 6 mM because of the limited solubility of limonene in 5 vol. % ethanol. Pressure of the carrier gas and pressurization (saturation) time have some influence on the extraction efficiency (Figures 4 and 5). The former effect is directly related to the Henry’s Law, while the latter effect is related to the dissolution kinetics. The decrease of analyte peak area at pressurization times longer than 60 s is likely due to the formation of carbonic acid in the aqueous sample matrix, and decrease of pH:4 COଶ ሺaqሻ + Hଶ O ↔ H ା + HCOି ଷ

(eq. 2)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

At low pH, the solubility of CO2 (responsible for bubble formation) is lower than at neutral or high pH. The pressurization time may need to be adjusted when extracting analytes from different matrices because pH and main matrix components influence solubility of carbon dioxide. For example, the pH of water:ethanol (95:5, v/v) mixture dropped from 4.2 to 3.8 following 60-s pressurization at 150 kPa CO2 and depressurization. Please note that the pH value immediately before depressurization might be lower but it could not be measured due to technical constraints imposed by the experimental setup. Moreover, decreasing sample volume from 10 to 9 and 8 mL led to a decrease of analyte peak area from 2.6 × 104 to 2.3 × 104, and then 2.0 × 104 a.u., respectively (sample: 0.6 mM limonene in water:ethanol 95:5, v/v; signal m/z 137). This suggests that fizzy extraction occurs in the entire sample volume, not only at the liquid meniscus.

Considerations regarding the extraction mechanism An important feature of the proposed extraction scheme is that the analytes are directly extracted to the gas phase, and transferred via gas phase to a detector operating in the gas phase. In fact, in one experiment (using a solution of fluorescein; Figure S4), we confirmed that no liquid droplets (which might be sprayed out from the sample during effervescence) were transferred from the extraction chamber to the extract tubing outlet (normally coupled with the APCI source). The release of analyte species into the gas phase occurs in a very short period of time, as a burst (Figures 1B and 3); typically, within few seconds although the exact time of release depends on the sample and operating conditions. The burst of concentrated analyte species provides instantaneous response of the detector. Thus, fizzy extraction


Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


confines analyte species in a short time interval, what can be regarded as an analogy to preconcentration (confinement into small volume). Carbon dioxide is generally considered to be a non-polar gas.5 Thus, similarly to SFE, it may promote the release of non-polar compounds. In fact, in the case of a relatively polar analyte (ethyl acetate), we observed negligible extraction (Figure S5). However, such a low extraction yield can also be due to the fact that ethyl acetate is highly volatile (Table S2), and the equilibrium between the gas-phase ethyl acetate and the dissolved ethyl acetate is rapidly established. When extracting ethyl acetate to the gas phase, fizzy extraction cannot increase the MS signal, which is already high when scavenging headspace vapors without prior extraction (Figure S5, < 1.2 min). On the other hand, caffeine was not detected during fizzy extraction (Figure S6); probably because it has very low vapor pressure (Table S2), and it cannot efficiently be transferred from the liquid phase to the gas phase in the course of effervescence. Although the non-volatile species are not detected at the outlet of the extraction chamber, they may still be released in the form of tiny droplets produced by bubble bursting, as discussed in the previous work (e.g.6,7). In the current setup (Figure 1A), the flow of carbon dioxide may not be strong enough to carry these droplets to the detector. Nonetheless, formation of such droplets may still contribute to extraction of volatile compounds to the gas phase. Moreover, when many bubbles are formed in the sample, scavenging of surface-active compounds by the bubble surface may also occur.7 Overall, compounds with medium vapor pressure are most amenable to fizzy extraction (Table S2). However, in this study focused on fizzy extraction, we did not notice any obvious correlation of signal enhancement (due to extraction) with surface tensions and solubilities (in



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

water) of the extracted compounds. We suggest that formation of numerous bubbles during effervescence drastically increases the surface area of the liquid-gas interface, so that the transfer of volatile and semi-volatile species to the gas-phase can occur rapidly. Moreover, effervescence produces high turbulence within the sample, what may increase mass transfer of analytes toward and across liquid-gas boundary. Similarly to boiling, during effervescence, evaporation of the species occurs in the entire sample volume. However, temperature is not increased, so decomposition of thermally labile analytes can be prevented. The process of analyte release from the liquid phase into the gas phase can be compared to co-precipitationbased preconcentration methods,8 in which small amounts of analyte species precipitate from the liquid phase to the solid phase. In fizzy extraction, small amounts of volatile and semivolatile species “co-precipitate” with the carrier gas into the gas phase, producing bubbles. As expected, stirring the sample during saturation with carbon dioxide slightly increased the analyte signal (Figure S7) because it increased the contact area between the sample and the carrier gas. However, according to other control experiment, stirring alone (without saturating the sample with carbon dioxide) did not produce a prominent analyte signal (Figure S8).

Distinctive features Contrary to SFE,9 fizzy extraction uses relatively low pressures of carrier gas (typically, ∼ 150 kPa). Thus, the extraction device is relatively inexpensive and easy to operate. However, unlike SFE, fizzy extraction is limited to liquid samples with moderate viscosity. Fizzy extraction bears some similarity to sparging methods.10-13 Sparging extraction entails passing gas continuously through the liquid column. To ensure high extraction yields, relatively long


Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


columns and large sample volumes need to be used. Such extractions are time-consuming.11 Moreover, affinity of the analyte to that of the extracting gas in sparging extraction must be high. The effluent gas may need to be trapped prior to analysis.10 In principle, if no trap is used, the gaseous extracts can be transferred directly to the detector.13 However, removing the trapping step can compromise sensitivity – considering that only a tiny amount of analyte can be extracted within the detection cycle of the detector (typically,

Fizzy Extraction of Volatile and Semivolatile ... - ACS Publications

Recommend Documents