Skip to main content
NCBI home page
3 Biotech logoLink to 3 Biotech
. 2020 Nov 10;10(12):520. doi: 10.1007/s13205-020-02523-8

A gas chromatography-flame ionization detection method for direct and rapid determination of small molecule volatile organic compounds in aqueous phase

Jia Zhou 1,2,✉, Xiaoqing Lu 1,2, Baoxia Tian 1,2, Chonglong Wang 3, Hao Shi 1,2, Chuping Luo 1, Xiangqian Li 1,2,✉
PMCID: PMC7655889  PMID: 33194524

Abstract

It is still difficult to directly detect low content of volatile organic compounds (VOCs) in water samples by gas chromatography (GC) because when water is the only solvent, it would result in the instability and poor repeatability of peak retention time and peak shape. The adverse effects of water on direct GC analysis of VOCs cannot be significantly reduced or eliminated by simply changing the detection condition of GC. However, it was found that the addition of methanol in samples to a certain final proportion, such as 50 or 75% (v/v), could greatly reduce or eliminate the adverse effects of water. By using 75% (v/v) methanol as a solvent, the standard curves of ethanol, acetic acid, acetone, and isopropanol with correlation coefficient (R2) over 0.99 were successfully plotted by gas chromatography-flame ionization detection (GC-FID) in a certain concentration range, respectively. The results showed that the retention time and peak shape stability of ethanol, acetic acid, acetone, and isopropanol in aqueous solution were greatly improved by the addition of methanol to final concertation of 75% (v/v). To verify the practical application potential of this method, the method was applied to the detection of components in isopropanol fermentation wastewater. The results showed that the method has well applicability and reliability. The key points in the application of this method were also summarized. This GC analysis method would have a wider and better application prospect in the detection of water-soluble organic matters.

Keywords: Gas chromatography, Isopropanol, Methanol, Volatile organic compounds

Introduction

In the production of the biological fermentation industry and chemical industry, some water-soluble and volatile small molecular organic compounds are often used or produced, such as common ethanol, acetic acid, acetone, and isopropanol (Chen and Hiu 1986; Cui et al. 2020; Baral et al. 2016). These substances may be discharged into the environment with sewage and cause pollution. The traditional method of water sample detection is based on chemical analysis, of which the procedure is tedious, time-consuming, and low sensitivity. High-performance liquid chromatography (HPLC) and capillary gas chromatography (GC) are relatively convenient and sensitive for the determination of many compounds. However, HPLC requires different absorption wavelengths for different chemical substances, which increases the detection time. While special equipment is often needed to separate the detection target from the water sample in GC analysis.

The direct determination of small volatile organic compounds (VOCs) in the aqueous phase by GC is often limited by detection conditions and methods. As a single solvent, water has a larger expansion coefficient and surface energy. Water with a high coefficient of gasifying expansion is easy to lead to lining overload, which directly affects the repeatability of the peak area. The surface energy of most stationary phases in the capillary column is low, which leads to poor wettability of water and seriously affects the peak shape. Many detectors, such as the electron capture detector (ECD), nitrogen and phosphorus detector (NPD), and mass spectrometry detector (MSD), are not suitable for detecting water samples. A large amount of water vapor will lead to the performance degradation of these detectors, and excessive water inflow could also extinguish the flame of the flame ionization detector (FID). When GC is applied to the detection of water samples, it is often necessary to separate the VOCs from the water phase by special equipment or methods, such as by solid-phase extraction or using headspace GC (Martins et al. 2013; Kang and Shin 2016; Drozd and Novák 1979; Zhang et al. 2018; Graffius et al. 2017; Chun et al. 2016).

Methanol is a small molecule amphoteric solvent. Studies have shown that adding substantial amounts of methanol can improve the wettability of water on the surface of hydrophobic materials (Fang et al. 2008). Besides, the low molecular weight of methanol makes a clear separation of the GC peak from that of most organics. Therefore, we added enough methanol to the water sample to change the solute properties, based on which, a direct and rapid GC analysis method for the determination of VOCs in aqueous solution was established.

Materials and methods

Chemicals and reagents

All chemical reagents used in this study were purchased from Sigma-Aldrich, including acetone (CAS No. 650501-1L), methanol (CAS No. 34860-1L-R), isopropanol (CAS No. W292907-1KG-K), acetic acid (CAS No. A6283), and ethanol (CAS No. 459836-1L). For sample preparation, standard substances were added to deionized water or methanol aqueous solution. In the test of the influence of methanol aqueous solution on GC analysis, methanol and deionized water were mixed in a volume ratio of 1:1 and 2:1 to form 50% (v/v) and 75% (v/v) methanol aqueous solutions, respectively. For the preparation of standard curves, 75% (v/v) methanol aqueous solution was taken as the standard solvent.

GC analysis

Sample analyses were performed on Shimadzu GC-2010 equipped with FID and an automatic sampler. Components were separated on a Shimadzu capillary column (30 m × 0.25 mm, 0.25 μm film thickness, the temperature range 250 °C, SH-Rtx-Wax, Cat. #221-75893-30). The GC was analyzed by programmed heating-up methods with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1. For a short-time programmed heating-up method, the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. For a long-time programmed heating-up method, the initial temperature was 40 °C maintaining for 3 min, raised to 120 °C at a rate of 15 °C/min, then to 200 °C at a rate of 10 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 2 min.

Standard curves

The standard curve of ethanol was constructed by adding 0, 1.6, 4.0 and 12 μl ethanol (purity of 99.5% and the density of 0.7893 g/ml) in 1 ml standard solvent (75% (v/v) methanol aqueous solution) to form 0, 1262, 3150 and 9460 mg/l ethanol solutions, respectively. The standard curve of acetone was constructed by adding 0, 0.4, 1.2, 4.0, and 12 μl acetone (purity of 99.9% and the density of 0.7845 g/ml) in 1 ml standard solvent to form 0, 313, 940, 3130, and 9400 mg/l acetone solutions, respectively. The standard curve of isopropanol was constructed by adding 0, 0.4, 1.2, and 4.0 μl isopropanol (purity of 99.7% and the density of 0.786 g/ml) in 1 ml standard solvent to form 0, 311, 938, and 3140 mg/l isopropanol solutions, respectively. The standard curve of acetic acid was constructed by adding 0, 0.4, 1.2, and 4.0 μl acetic acid (purity of 99% and the density of 1.049 g/ml) in 1 ml standard solvent to form 0, 420, 1260, and 4200 mg/l acetic acid solutions, respectively.

Results and discussion

Effect of water solvent on the GC-FID analysis of isopropanol

To analyze a water sample containing 311 mg/l of isopropanol, a short-time programmed heating-up GC-FID method was applied, for which the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The same isopropanol aqueous solution was repeatedly injected four times, resulting in different peak retention times, peak shape, and peak area (Fig. 1). After the first injection, a peak appeared at 1.367 min. The second and third repeated injections gave rise to the retention time of the main peak shifting to 1.400 min, and several new peaks appeared after 7 min. More peaks appeared after the fourth injection, among which the maximum peak appeared at the retention time of 7.5 min. Most of the peaks after four injections were asymmetric. It was speculated that this phenomenon is due to the short running time, resulting in a small amount of sample remaining in the capillary.

Fig. 1.

Fig. 1

Open in a new tab

Analysis of isopropanol in water solvent by GC-FID with a short-time programmed heating-up method. a–d Represent gas chromatograms of 4 times repeated injection of a water sample containing 311 mg/l of isopropanol. For a short-time programmed heating-up method, the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The GC was analyzed with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1. Arrows indicate new peaks

To determine whether the instability of these results could be reduced or eliminated, a long-time programmed heating-up GC condition was then tested, for which the initial temperature was 40 °C maintaining for 3 min, raised to 120 °C at a rate of 15 °C/min, then to 200 °C at a rate of 10 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 2 min. However, contrary to expectations, repeated injecting 6 times of isopropanol water solution at a concentration of 1555 mg/l resulted in multiple peaks, asymmetric peak shape, and non-repeatable peak retention time and peak area (Fig. 2). For example, the retention time of the first peak after 6 injections was 1.904, 1.662, 1.891, 1.410, 1.799, and 1.895 min, respectively; huge peaks appeared after 10 min of the second and sixth injections.

Fig. 2.

Fig. 2

Open in a new tab

Analysis of isopropanol by GC-FID with a long-time programmed heating-up method. a–f Represent gas chromatograms of 6 times repeated injection of a water sample containing 1555 mg/l of isopropanol. For a long-time programmed heating-up method, the initial temperature was 40 °C maintaining for 3 min, raised to 120 °C at a rate of 15 °C/min, then to 200 °C at a rate of 10 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 2 min. The GC was analyzed with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1. Arrows indicate new peaks

The occurrence of hetero peaks is probably due to the retention of isopropanol along with water molecules in capillaries, while the phenomenon of inconsistent peak retention time and peak area should be related to the fact that water can easily lead to liner overloading and the poor wettability in capillaries.

Effect of methanol addition on GC-FID analysis of isopropanol, ethanol, acetic acid, and acetone

Here, methanol was added to water samples at a final concentration of 50% (v/v) and 75% (v/v), respectively, to test the effect of methanol on the analysis of isopropanol under the short-time programmed heating-up GC condition. As expected, the phenomenon of the hetero peaks, inconsistent retention time, and the variation of peak area were no longer appeared after adding a high concentration of methanol. As shown in Fig. 3, the peak retention time of solvent methanol was 2.43 min, and the peak retention time of isopropanol was stabled at 2.78 min with symmetrical peak shape. The standard solvent of 75% methanol solution was used to test several other water-soluble compounds, including ethanol, acetic acid, and acetone. As shown in Fig. 4, the peak retention time of ethanol, acetic acid, and acetone was 2.88, 6.4 and 1.48 min, respectively. Repeated tests showed that the peak retention time and the peak shape of these compounds were stable and reproducible (data was not shown). The above results indicated that a high concentration of methanol can be utilized to correct the negative effects of water as a single solvent on GC analysis of detected substances.

Fig. 3.

Fig. 3

Open in a new tab

Effect of methanol addition on GC-FID analysis of isopropanol. The above and the following GC graphs were generated using 50% (v/v) and 75% (v/v) methanol solutions as a solvent, respectively, following the short-time programmed heating-up GC condition. For a short-time programmed heating-up method, the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The GC was analyzed with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1

Fig. 4.

Fig. 4

Open in a new tab

Effect of 75% (v/v) methanol solution as a solvent on the GC-FID analysis of isopropanol, ethanol, acetic acid, and acetone, with the short-time programmed heating-up GC condition, for which the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The GC was analyzed with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1

Construction of standard curves of ethanol, acetic acid, acetone, and isopropanol by the new GC-FID analysis method

To further test the reliability of the methanol method in the detection of ethanol, acetic acid, acetone, and isopropanol water samples, the standard curves of ethanol, acetic acid, acetone, and isopropanol were constructed by the new GC analysis method to show whether there is a linear relationship between their contents and peak areas. As shown in Fig. 5, in a concentration range of 0–10 g/l, the R2 values of the standard curves of ethanol and acetone reached 0.9979, and 0.999, respectively, while in a concentration range of 0–5 g/l, the R2 values of the standard curve of acetic acid and isopropanol reached 0.9988, indicating that their contents had a good linear correlation with the peak area in the given concentration range (Fig. 5). It was also found that, outside the above concentration ranges, a non-linear relationship between the concentration and the peak area of detected compounds was observed (data not shown), indicating that this method was reliable for the detection of ethanol, acetic acid, acetone, and isopropanol with a certain concentration range. To obtain accurate results, it is thus necessary to dilute the sample into an appropriate concentration before GC analysis.

Fig. 5.

Fig. 5

Open in a new tab

Standard curves of ethanol, acetic acid, acetone, and isopropanol. The standard curves were plotted using 75% (v/v) methanol as the solvent following the short-time programmed heating-up GC condition, for which the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The GC was tested with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1. The peak area was taken as the ordinate and the content was taken as the abscissa. Error bars represent the standard deviation of triplicates

Application of the new GC-FID analysis method in the detection of the waste liquid of isopropanol fermentation

To test the application of the new GC-FID analysis method, samples were taken from the waste liquid of isopropanol fermentation in our laboratory. For sample preparation, methanol was added to the sample at a final concentration of 75% (v/v); due to the strong water absorption of methanol and the complexity of the sample composition, the sample became turbid after adding a high concentration of methanol; before injection, insoluble matters were then removed from the sample by centrifuged at12,000 rpm for 20 min followed by filtration of the supernatant with a 0.4 μm filter membrane. The sample was repeatedly analyzed by GC with good repeatability following the short-time programmed heating-up method (repeated data not shown). As shown in Fig. 6, all the peaks have well symmetry and integrability, and the overall retention time of peak was found to be slightly earlier than that of the standard samples (Fig. 4), which suggesting that the complexity of the components in the water sample still could have potentially limited impact on the retention time of the peak, but does not affect their symmetry and repeatability. In the waste liquid of isopropanol fermentation, peaks of many unknown compounds were also detected in addition to ethanol (7.99 g/l), acetone (0.29 g/l), acetic acid (10.13 g/l), and a tiny amount of isopropanol (< 0.1 g/l).

Fig. 6.

Fig. 6

Open in a new tab

Detection of the waste liquid of isopropanol fermentation by GC-FID with a short-time programmed heating-up method. For sample preparation, methanol was added at a final concentration of 75% (v/v); before injection, insoluble matters were then removed by centrifuged at 12,000 rpm for 20 min followed by filtering the supernatant with a 0.4 μm filter membrane. For a short-time programmed heating-up method, the initial temperature was 25 °C maintaining for 3 min, raised to 120 °C at a rate of 40 °C/min, then to 230 °C at a rate of 50 °C/min maintaining for 2 min, and finally, the post-run was kept at 230 °C for 1 min. The GC was tested with an inlet temperature of 250 °C, a FID temperature of 280 °C, an injection volume of 0.2 μl, and a split ratio of 25:1

Conclusions

Here, a direct and rapid GC analysis method for the determination of VOCs in aqueous solution was established. By adding a certain proportion of methanol to the water sample, such as 50% or 75% (v/v) methanol, the adverse effects on GC analysis caused by a single water solvent could be greatly reduced, resulting in consistent retention time, peak shape, and peak area of isopropanol after repeated injections. The standard curves of ethanol, acetic acid, acetone, and isopropanol in a certain concentration range were successfully made based on this method with square values (R2) of the correlation coefficient greater than 0.99. The applicability and reliability of this GC analysis method were further well verified in the detection of isopropanol fermentation wastewater. The key points in the application of this direct and rapid GC analysis method for the determination of VOCs in aqueous solution are summarized as follows: (1) Sample preparation. A high concentration of methanol should be added to the sample. It is proposed that the final concentration should be above 50% (v/v). Before injection, if the sample becomes turbid after adding methanol, the sample should be centrifuged at high speed for 20 min and filtered with a 0.4 μm filter membrane. (2) Injection. To avoid the introduction of a large amount of water, it is necessary to reduce the injection volume as much as possible. The injection volume in this study was set at 0.2 μl with a split ratio of 25:1. (3) Replacement of quartz cotton ball in the liner. Most of the water will be absorbed by the quartz cotton ball, so it needs to be replaced regularly. (4) The initial temperature of the programmed heating-up method. The initial temperature should be as low as possible to facilitate the separation of methanol peak from that of the other small molecule organic compounds. (5) The post-runs. It is necessary to carry out a post-run program with sufficient temperature and time to avoid the residue of sample and water in a capillary column. In conclusion, this study showed that this method will have a more extensive and reliable application prospect in the direct and accurate detection of water-soluble organic matters.

Acknowledgements

This work was supported by a grant (BK20150417, BK20171266) from the Natural Science Foundation of Jiangsu Province of China; grants (21576110, 21706089) from the National Natural Science Foundation of China; a grant (17KJA180001) from Educational Commission of Jiangsu Province of China; grants (JPELBCPI2017003, JPELBCPI2015003) from Jiangsu Provincial Engineering Laboratory for Biomass Conversion and Process Integration Open Project.

Compliance with ethical standards

Conflict of interest

No conflict of interest was declared.

Contributor Information

Jia Zhou, Email: Joe50@live.cn.

Xiangqian Li, Email: lixq2002@126.com.

References

  1. Baral NR, Slutzky L, Shah A, Ezeji TC, Cornish K, Christy A. Acetone-butanol-ethanol fermentation of corn stover: current production methods, economic viability and commercial use. FEMS Microbiol Lett. 2016 doi: 10.1093/femsle/fnw033. [DOI] [PubMed] [Google Scholar]
  2. Chen JS, Hiu SF. Acetone-butanol-isopropanol production by Clostridium beijerinckii. Biotechnol Lett. 1986;8(5):371–376. doi: 10.1007/BF01040869. [DOI] [Google Scholar]
  3. Chun HJ, Poklis JL, Poklis A, Wolf CE. Development and validation of a method for alcohol analysis in brain tissue by headspace gas chromatography with flame ionization detector. J Anal Toxicol. 2016;40(8):653–658. doi: 10.1093/jat/bkw075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cui Y, He J, Yang KL, Zhou K. Aerobic acetone-butanol-isopropanol (ABI) fermentation through a co-culture of Clostridium beijerinckii G117 and recombinant Bacillus subtilis 1A1. Metab Eng Commun. 2020;11:e00137. doi: 10.1016/j.mec.2020.e00137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Drozd J, Novák J. Headspace gas analysis by gas chromatography. J Chromatogr A. 1979;165(2):141–165. doi: 10.1016/S0021-9673(00)90938-5. [DOI] [Google Scholar]
  6. Fang Y, Sun G, Cong Q. Effects of methanol on wettability of the non-smooth surface on butterfly wing. J Bionic Eng. 2008;5(2):127–133. doi: 10.1016/S1672-6529(08)60016-5. [DOI] [Google Scholar]
  7. Graffius GC, Jocher BM, Zewge D, Halsey HM, Lee G, Bernardoni F, Bu X, Hartman R, Regalado EL. Generic gas chromatography-flame ionization detection method for quantitation of volatile amines in pharmaceutical drugs and synthetic intermediates. J Chromatogr A. 2017;1518:70–77. doi: 10.1016/j.chroma.2017.08.048. [DOI] [PubMed] [Google Scholar]
  8. Kang HI, Shin HS. Determination of glutaraldehyde in water samples by headspace solid-phase microextraction and gas chromatography-mass spectrometry after derivatization with 2,2,2-trifluoroethylhydrazine. J Chromatogr A. 2016;1448:115–120. doi: 10.1016/j.chroma.2016.04.049. [DOI] [PubMed] [Google Scholar]
  9. Martins ML, Donato FF, Prestes OD, Adaime MB, Zanella R. Determination of pesticide residues and related compounds in water and industrial effluent by solid-phase extraction and gas chromatography coupled to triple quadrupole mass spectrometry. Anal Bioanal Chem. 2013;405(24):7697–7709. doi: 10.1007/s00216-013-7235-0. [DOI] [PubMed] [Google Scholar]
  10. Zhang SX, Peng R, Jiang R, Chai XS, Barnes DG. A high-throughput headspace gas chromatographic technique for the determination of nitrite content in water samples. J Chromatogr A. 2018;1538:104–107. doi: 10.1016/j.chroma.2018.01.026. [DOI] [PubMed] [Google Scholar]

Articles from 3 Biotech are provided here courtesy of Springer

RESOURCES