Laboratory Experiment pubs.acs.org/jchemeduc
Improving Student Understanding of Qualitative and Quantitative Analysis via GC/MS using a Rapid SPME-Based Method for Determination of Trihalomethanes in Drinking Water Shu Rong Huang and Peter T. Palmer* Department of Chemistry and Biochemistry, San Francisco State University, San Francisco, California 94132, United States S Supporting Information *
ABSTRACT: This paper describes a method for determination of trihalomethanes (THMs) in drinking water via solid-phase microextraction (SPME) GC/MS as a means to develop and improve student understanding of the use of GC/MS for qualitative and quantitative analysis. In the classroom, students are introduced to SPME, GC/MS instrumentation, and the use of MS data for compound identification and quantification through traditional lectures and in-class problem solving. In the lab, they are introduced to GC/MS tuning and method development, learn how to use the GC/MS data system, and apply their knowledge to identify and quantify THMs in tap water samples. The method for this application is relatively simple, fast, and effective, and is based on the use of autosampler, SPME headspace sampling, a GC quadrupole ion trap MS instrument, and fluorobenzene as an internal standard. Data from analyses of standards demonstrated 1 ppb LODs, good linearity in the 5−100 ppb range, RSDs of 6% or lower, and good recoveries from spikes of a tap water sample. This method as described is appropriate for environmental analysis, instrumental analysis, or quantitative analysis courses. It provides students with hands-on experience in the use of SPME-based sampling methods and GC/MS for reliable identification and accurate quantitative analysis at low ppb levels. KEYWORDS: Upper-Division Undergraduate, Qualitative Analysis, Quantitative Analysis, Laboratory Instruction, Hands-On Learning/Manipulatives, Instrumental Methods, Analytical Chemistry
T
followed by GC/MS/MS.8 Salient features here include the excellent selectivity through the use of multiple reaction monitoring (MRM), LODs in range 0.01−0.03 ppb, and recoveries in the range 95−100%. A comparison of purge and trap versus SPME for determination of VOCs via GC/MS is provided elsewhere.9 Without question, SPME-based sampling10 greatly simplifies the extraction process compared to dedicated purge and trap units or liquid−liquid extraction. The Sigma-Aldrich Web site provides detailed information on SPME fiber selection, method optimization, and publications on various applications.11 Recent J. Chem. Educ. articles describe the use of SPME for determination of bisphenol A in water12 and plant-based VOCs.13 A number of groups have evaluated SPME/GC/MS for the determination of THMs in water. Stack et al. used 20 min of manual extraction with a polydimethylsiloxane (PDMS) fiber to achieve LODs of 1−3 ppb for THMs in potable water.14 Hardee and co-workers used 30 min of manual extraction and a Carboxen/PDMS fiber to monitor CHBr3 in swimming pools but did not report LODs.15 Cardinali et al. used a CTC CombiPal autosampler, Carboxen/PDMS fiber, a
HMs (CHX3 where X is Cl or Br) are commonly found in drinking water and are a byproduct of reactions of chlorine-based disinfectant agents and bromine species with natural organic matter in water sources.1 Indeed, a J. Chem. Educ. article from 2001 describes an experiment in which students can study the formation of THMs using resorcinol as a model compound.2 Although the disinfection process reduces the spread of microorganisms in drinking water, numerous studies have shown that THMs are carcinogenic and can cause adverse reproductive and/or developmental effects.3 To balance the need for safe drinking water and minimal risk to public health, EPA established a maximum contaminant level (MCL) of 80 ppb for total THMs4 and nonenforceable maximum contaminant level goals (MCLGs) of 70, 60, 0, and 0 ppb for CHCl3, CHClBr2, CHCl2Br, and CHBr3, respectively.5 Pavon et al. provide an excellent review of various methods to monitor THMs in water.1 In regulatory laboratories, this is commonly performed using a dedicated purge and trap unit coupled to a GC/MS instrument as described in EPA Method 524.2.6 Two less conventional methods for this application are worth mentioning. In one study, THMs in water were measured using liquid−liquid extraction followed by GC/ ICP/MS.7 Key features of this method are the use of a compound-independent calibration, 0.01 ppb LODs, and recoveries in the range 81−96%. In another study, THMs were monitored in urine using liquid−liquid extraction © XXXX American Chemical Society and Division of Chemical Education, Inc.
Received: December 7, 2016 Revised: June 23, 2017
A
DOI: 10.1021/acs.jchemed.6b00939 J. Chem. Educ. XXXX, XXX, XXX−XXX
Journal of Chemical Education
Laboratory Experiment
Table 1. Summary of Analytical Figures of Merit of the Method Molecular Formula
Major Ions (m/z)
Retention Time (min)
Quan Ion (m/z)
R2 Valuea
LOD (ppb)
% RSDb
% Recovery
CHCl3 CHCl2Br CHClBr2 CHBr3 C6H5F
83, 85, 47 83, 85, 47 127, 129, 131 171, 173, 175 96, 70, 50
3.90 4.96 6.16 7.41 4.50
83 83 129 173 96
>0.99 >0.99 >0.99 >0.99 c
1 1 0.9 1 c
6 6 6 6 c
112 104 95 Not avail c
a 2 R values computed for standards in the range 5−100 ppb. b% RSDs computed from 10 replicate analyses of the 5 ppb standard. cThis compound is the internal standard, and hence, these metrics are not relevant.
water, distilled water, and then deionized water. A final rinse with HPLC-grade methanol and a drying step were deemed necessary to remove residual THMs from faucet and distilled water and lower blank response. It should be noted that the common use of solvents in most chemistry laboratories presents a challenge with respect to obtaining “clean” method blanks and low LODs. A total of six tap water samples were collected from residences in San Francisco, Alameda, Contra Costa, and Santa Clara counties, as well as samples from a high school, community college, and a university. In addition, two samples from a public pool and community college pool were collected. Samples were placed into precleaned wide-mouth 500 mL lowdensity polyethylene bottles, which were familiarized with the sample several times prior to filling to the top to minimize the headspace volume. Samples were transported to the lab, refrigerated at 4 °C, and analyzed within 2 weeks. A certified reference material (CRM) containing 100 ppm of the four THMs in methanol (Supelco P/N CRM47904) was used to prepare calibration standards ranging from 5 to 500 ppb in water. A 2000 ppm fluorobenzene in methanol CRM (Supelco P/N CRM48943) was used to prepare a 5 ppm fluorobenzene internal standard spiking solution. Dilutions were performed using micropipets, class A 100 mL volumetric flasks, and deionized water. Prior to analysis, 5 mL of the sample or standard solutions was transferred via volumetric pipet into a 10 mL headspace-compatible autosampler vial (Sigma-Aldrich P/N SU860099). Each vial was spiked with 100 μL of 5 ppm fluorobenzene standard to yield an approximate concentration of 100 ppb fluorobenzene in the vial and sealed with an autosampler-compatible magnetic cap (P/N SU860103). Although an isotopically labeled internal standard is generally preferred, fluorobenzene is less expensive and provides improved linearity and precision versus external standard based calibration. Vials were sealed with septum caps and placed in the autosampler rack. Several additional samples were prepared and analyzed for quality control purposes, including 10 replicates of a 5 ppb THM standard to assess LODs and precision, and one tap water sample spiked with 100 μL of 500 ppb THMs to assess accuracy. Headspace sampling of each vial was accomplished using a CTC-CombiPAL autosampler equipped with an 85 μm Carboxen/PDMS metal alloy core SPME fiber (SigmaAldrich). The autosampler was programmed to agitate the sample for 5 min at 30 °C, sample the headspace for 15 min, and inject the sample into a Varian 3900 GC/Saturn 2100T quadrupole ion trap MS system. The injector was equipped with a Merlin Microseal (Sigma-Aldrich) and a 0.75 mm ID SPME injection liner (Sigma-Aldrich). The injector was set to 300 °C and operated in splitless mode. A 30 m × 0.25 mm × 1 μm SLB-5 ms (Sigma-Aldrich) capillary column was programmed from 40 to 280 °C at 15 °C/min and a GC analysis
cryotrap on the head of the GC column, subambient cooling of the GC oven, and selected ion monitoring (SIM) mode to achieve LODs of 0.05−1 ppb for THMs and MTBE in tap water.16 Similarly, Antoniou et al. used a CTC CombiPal autosampler, Carboxen/PDMS fiber, and 30 min extraction times to measure chlorinated VOCs in drinking water down to 0.3−1 ppb levels.17 In a related study, Jochman et al. utilized solid-phase dynamic extraction (SPDE) to monitor THMs in groundwater.18 Unlike SPME, where the stationary phase of the fiber is coated on the outside of the metal alloy fiber support, the stationary phase of SPDE is coated on the inside of the metal alloy support, and this method requires optimization of desorption parameters to avoid peak tailing and/or peak splitting. The initial impetus for this work was to develop a GC/MS experiment for our environmental analysis course that gave students hands-on experience in performing qualitative and quantitative analysis. As described above, a number of groups have developed a variety of GC/MS based methods to monitor THMs in water. While some of these methods give outstanding selectivity and excellent LODs, they have some limitations with respect to use in an undergraduate lab class. These include a scope that does not include all four THMs,15,17,18 use of a PDMS in lieu of the more sensitive Carboxen/PDMS fiber,14,18 manual sampling14,15 which can lead to poor RSDs and broken SPME fibers, and/or use of specialized equipment (such as a dedicated purge and trap unit6 or SDPE sampling equipment,18 subambient-cooled GC,16 or an ICP-MS,7 SIM,16 or MS/MS8 system) that would be unavailable for instructional lab classes in most undergraduate programs. The method chosen for this experiment was based on our inhouse GC/ion trap MS system but can be readily adapted to quadrupole and other GC/MS systems. More importantly, the use of an autosampler enables unattended operation, provides improved precision, and gives students with more time to focus on learning and using the data system. The experiment can be completed in two 3 h lab periods, one for preparation and analysis of the standards and samples and the second for data analysis. If appropriate, the experiment can be expanded to give students experience in optimizing SPME, GC, and MS parameters. When coupled with lectures and in-class homework problems on SPME as a sample preparation method and the use of GC/MS for qualitative and quantitative analysis, this experiment provides students with a thorough, detailed, and meaningful hands-on experience in the use of GC/MS for a regulatory application.
■
METHODS More detailed information on the experiment including equipment and supplies, standard preparation, and procedures is provided in the Supporting Information. All glassware and sample containers were thoroughly cleaned with soap and B
DOI: 10.1021/acs.jchemed.6b00939 J. Chem. Educ. XXXX, XXX, XXX−XXX
Journal of Chemical Education
Laboratory Experiment
time of 16 min. MS parameters were set to EI mode, 20 μA emission current, automatic gain control with a target total ion count of 30,000, and a scan range of m/z 45−250 at 0.43 s/scan As per most EPA methods and protocols for positive identification of a compound via GC/MS, THMs in the samples were identified by simultaneous maximization of the three major ions at the known retention time for that species. For a demonstration of the importance of this concept, note that although the mass spectra of CHCl3 and CHCl2Br share the same three major ions, they have different retention times and hence can be differentiated from one another. THMs were quantified using peak area ratios (area of the base peak of each analyte divided by that of the internal standard) of major ions for each species (quan ions provided in Table 1). Peak area data were entered in Excel (Microsoft) to generate calibration curves, compute analytical figures of merit, and compute sample concentrations.
Table 2. Summary of Water Sample Analysis and Relevant EPA Limits Sample San Francisco tap water Contra Costa County tap water Santa Clara County tap water Alameda County tap water SFSU tap water Monta Vista High School tap water De Anza College pool water SF Public pool water EPA MCL EPA MCLG
■
RESULTS AND DISCUSSION Table 1 provides a summary of various analytical figures of merit for this method. The R2 values show good linearity in the calibration range 5−100 ppb, with some roll-off in response observed at 500 ppb which can be attributed to saturation of the SPME fiber at these relatively high THM levels. LODs, as assessed from 10 replicate analyses of a 5 ppb THM standard and the standard definition of the LOD as concentration at which the signal-to-noise ratio is 3:1, were all around 1 ppb range, but it should again be noted that this is dependent on the levels of THMs in the blank. Precision was significantly improved due to the use of an internal standard and an autosampler versus manual sampling, with RSDs of 6% from 10 replicate analyses of the 5 ppb standard and lower RSDs for higher level standards. It should be pointed out that replicate analyses must be performed using separate vials and not reanalysis of the same vial, as the latter shows decreased response due to removal of THMs from the SPME sampling process. Accuracy was assessed by spiking one water sample with 100 μL of 500 ppb of CHCl3, CHCl2Br, and CHClBr2, which gave recoveries well within the typical 80−120% acceptable range (no recovery study performed for CHBr3). Table 2 shows results from the application of this method to several tap and pool water samples. CHBr3 was not quantified in these samples due to this and prior work which showed nondetectable levels of CHBr3 in drinking water samples.14 Uncertainties were propagated from the uncertainties in the slope, intercept, and MS response (standard error in quan ion response). The Alameda tap water sample slightly exceeded the 80 ppb MCL for total THMs in drinking water. One of the tap water samples exceeded the 70 ppb MCGL for CHCl3, and two exceeded the 0 ppb MCGL for CHClBr2. Although the primary focus of this experiment is on drinking water, it is appropriate to ask one or more students to bring in and analyze pool water samples, compare their THM levels to those found in tap water, and hypothesize as to why these levels are higher and why the EPA limits are not relevant here. They are also required to identify references from publicly available water quality reports showing typical THM levels, and references describing EPA limits and regulations on THMs in drinking water, and to compare their results to these in their written lab reports. This SPME/GC/MS experiment is appropriate for an upper division environmental analysis or instrumental analysis laboratory course. Students gain practical experience in the use of SPME-based sampling for volatiles, GC/MS for trace
a b
[THM] (ppb)
[CHCl3] (ppb)
[CHCl2Br] (ppb)
[CHClBr2] (ppb)
5±2
5±2
NDa
ND
52 ± 3
45 ± 2
7±1
ND
38 ± 2
ND
13 ± 1
25 ± 2
94 ± 3
79 ± 3
15 ± 1
ND
11 ± 2 16 ± 2
11 ± 2 ND
ND 5±1
ND 12 ± 2
59 ± 2
59 ± 2
ND
ND
201 ± 4 80 b
201 ± 4 b 70
ND b 60
ND b 0
ND denotes not detected (concentration is less than the LOD). EPA does not provide MCL or MCLG for these species.
level analysis, and internal standard-based calibration to determine THMs in their drinking water. This experiment has been used in our senior-level environmental analysis course on five different occasions with continued refinement of the methods, accompanying lecture topics, and active learning asignments. Initial efforts were based on the use of manual injection, which was found to give poor reproducibility in the sampling process as well as the danger of students inadvertently destroying the SMPE fibers (∼$500 for a package of three fibers). Use of an autosampler greatly simplifies the analysis, enables unattended operation, and provides better control of sampling parameters such as time and temperature. Use of fluorobenzene as an internal standard significantly improved linearity, precision, and accuracy. This same method was employed with the MS instrument configured in selected ion monitoring (SIM) mode, which gave comparable LODs. Although SIM usually gives lower LODs versus full scan mode, this was not true in this case, and it should be noted that optimization of SIM parameters on a quadrupole ion trap is challenging and most likely beyond the scope of most undergraduate lab classes. After several iterations of this experiment, it was clear that students needed more focused and better preparation for this particular experiment within the larger context of environmental applications of GC/MS. Toward this end, we have made a number of improvements and additions to the lecture portion of this course that provide the necessary background and directly engage the students in relevant exercises and activities. This includes lectures on SPME that describe theory, different fibers and their scope, method development and optimization, and applications; and lectures on GC/MS which highlight instrumentation, data analysis, and applications. More importantly, this includes several in-class problems assignments, demonstrations, and hands-on activities that have greatly increased student comfort level with the challenges and pitfalls of GC/MS data analysis. One assignment asks students to predict the GC elution order of the four THMs and their relative sensitivity via the SPME method. Another assignment asks students to choose a THM, determine its average molecular weight (which is irrelevant to GC/MS data analysis), determine the nominal masses of the molecular ion as well as C
DOI: 10.1021/acs.jchemed.6b00939 J. Chem. Educ. XXXX, XXX, XXX−XXX
Journal of Chemical Education
Laboratory Experiment
of active learning materials associated with this experiment that will be posted on the Analytical Sciences Digital Library (home. asdlib.org).
other isotopic variants (an important concept in understanding mass spectra), and determine the fragmentation mechanism leading to the ion corresponding to the base peak (which relates to ion stability and structure). In the lab, students participate in a hands-on session that provides an introduction to the GC/MS hardware and software, involves them in acquiring background mass spectra and identifying major ions, assessing the mass calibration, and identifying specific parameters in the THM method (SPME exposure time and temperature, GC temperature program, m/z range). Given that many students struggle to understand the three dimensions of data represented in a GC/MS analysis, it is deemed critically important to involve students in hands-on activities to facilitate a better understanding of GC/MS data. Each student is asked to choose a peak from a chromatogram from a test mixture, obtain a background subtracted mass spectrum, perform a library search, and propose the tentative identity of the compound. For the purposes of target compound analysis and this experiment, we also have students perform hands-on activities and homework assignments to learn how a compound is positively identified (through retention time matching and simultaneous maximization of three major ions) and how quantitative analysis is achieved (on the basis of integration of the peak area of a quan ion of the compound of interest and normalizing this to the peak area of the quan ion of the internal standard). Over several years of use implementing this experiment in our environmental analysis class, the use of this multilayered approach to learning GC/MS including lectures, in-class problem solving, demonstrations, and active learning activities prior to students performing this lab experiment have led to signif icant improvement in student learning and accurate results. The majority of students are now using the appropriate logic and methods to identify and quantify a specific target compound via GC/MS. As a result, they leave the course with a much better understanding on the use of GC/MS for environmentally relevant applications.
■
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available on the ACS Publications website at DOI: 10.1021/acs.jchemed.6b00939. Experimental procedure for the students, which provides details on background information, equipment and supplies, standard and standard preparation, GC/MS setup, and data analysis (PDF) (DOCX)
■
REFERENCES
(1) Pavón, J. L. P.; Martín, S. H.; Pinto, C. G.; Cordeo, B. M. Determination of Trihalomethanes in Water Samples: A Review. Anal. Chim. Acta 2008, 629, 6−23. (2) Olson, T. M.; Gonzales, A. C.; Vasquez, V. R. Gas Chromatography Analyses for Trihalomethanes: An Experiment Illustrating Important Sources of Disinfection By-Products in Water Treatment. J. Chem. Educ. 2001, 78, 1231−1234. (3) Richardson, S.; Kimura, S. Water Analysis: Emerging Contaminants and Current Issues. Anal. Chem. 2016, 88, 546−582. (4) Table of Regulated Drinking Water Contaminants. https://www. epa.gov/ground-water-and-drinking-water/table-regulated-drinkingwater-contaminants (accessed June 2017). (5) What Are EPA’s Drinking Water Regulations for Disinfection Byproducts Like TTHMs?. https://safewater.zendesk.com/hc/en-us/ articles/212075497-2-What-are-EPA-s-drinking-water-regulations-fordisinfection-byproducts-like-TTHMs- (accessed June 2017). (6) Eichelberger, J.; Munch, J.; Bellar, T. Method 524.2 Measurement of Purgeable Organics Compounds in Water by Capillary Gas Chromatography/Mass Spectrometry; Environmental Protection Agency: Cincinnati, OH, 1992. (7) Gonzalez-Gago, A.; Marchante-Gayón, J. M.; Garcia Alonso, J. I. Determination of Trihalomethanes in Drinking Water by GC-ICP-MS Using Compound Independent Calibration with Internal Standard. J. Anal. At. Spectrom. 2007, 22, 1138−1144. (8) Charisiadis, P.; Makris, K. A Sensitive and Fast Method for Trihalomethanes in Urine Using Gas Chromatography-Triple Quadrupole Mass Spectrometry. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2014, 947, 17−22. (9) Lara-Gonzalo, A.; Sanchez-Uria, J. E.; Segovia-Garcia, E.; SanzMedel, A. Critical Comparison of Automated Purge and Trap and SPME for Routine Determination of VOCs in Drinking Waters by GC-MS. Talanta 2008, 74, 1455−1462. (10) Pawliszyn, J.; Handbook of Solid Phase Microextraction; Chemical Industry Press: New York, 2009. (11) Solid Phase Microextraction. http://www.sigmaaldrich.com/ analytical-chromatography/sample-preparation/spme.html (accessed June 2017). (12) Johnson, B. O.; Burke, F. M.; Harrison, R.; Burdette, S. Quantitative Analysis of Bisphenol A Leached from Household Plastics by Solid-Phase Microextraction and Gas Chromatography-Mass Spectrometry. J. Chem. Educ. 2012, 89, 1555−1560. (13) Van Bramer, S.; Goodrich, K. Determination of Plant Volatiles Using Solid Phase Microextraction GC-MS. J. Chem. Educ. 2015, 92, 916−919. (14) Stack, M.; Fitzgerald, G.; O'Connell, S.; James, K. Measurement of Trihalomethanes in Potable and Recreational Waters Using SolidPhase Microextraction with Gas-Chromatography-Mass Spectrometry. Chemosphere 2000, 41, 1821−1826. (15) Hardee, J.; Long, J.; Otts, J. Quantitative Measurement of Bromoform in Swimming Pool Water using SPME with GC/MS. J. Chem. Educ. 2002, 79, 633−634. (16) Cardinali, F. L.; Ashley, D. L.; Morrow, J. C.; Moll, D. M.; Blount, B. C. Measurement of Trihalomethanes and MTBE in Tap Water Using SPME-GC-MS. J. Chromatogr. Sci. 2004, 42, 200−206. (17) Antoniou, C.; Koukouraki, E.; Diamadopoulos, E. Determination of Chlorinated Volatile Organic Compounds in Water and Municipal Wastewater Using Headspace Solid-Phase MicroextractionGas Chromatography. J. Chromatogr. A 2006, 1132, 310−314. (18) Jochmann, M.; Yuan, X.; Schmidt, T. Determination of Volatile Organic Hydrocarbons in Water Samples by Solid-Phase Dynamic Extraction. Anal. Bioanal. Chem. 2007, 387, 2163−2174.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Peter T. Palmer: 0000-0002-3173-5726 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Thomas Wenzel (Bates College), colleagues, and participants in an NSF-supported workshop (DUE 1118600 and 1624898) for their feedback in the development D
DOI: 10.1021/acs.jchemed.6b00939 J. Chem. Educ. XXXX, XXX, XXX−XXX