You are on page 1of 3

In the Laboratory

Was the Driver Drunk? An Instrumental Methods Experiment W


for the Determination of Blood Alcohol Content
Jennifer L. Zabzdyr and Sheri J. Lillard*
Department of Chemistry, University of California, Riverside, Riverside, CA 92521; *sheri.lillard@ucr.edu

Rationale routinely accepted technique for such a determination is gas


chromatography (GC) (12–15). Since the experiment was de-
The combination of forensic science and instrumental signed around existing laboratory equipment, the students
analysis is a very effective approach for teaching analytical will use GC with flame ionization detection (FID) for analysis.
chemistry (1). A major goal in revitalizing our instrumental FID has the additional advantage that water vapor does not
methods laboratory is to add forensic-based experiments to interfere with it (16 ). Qualitative analysis is accomplished
spark student motivation and enthusiasm for the course by comparing the retention time of ethanol in a standard with
material. A course on forensic analytical chemistry has been the retention times of peaks that occur in the sample. Both
taught by Brewer et al. at the University of South Carolina external and internal calibration methods are used to quan-
with the similar goal of increasing student participation in the titate the ethanol content in a (simulated) serum sample from
chemistry lecture (2). We want the students to understand a suspected drunk driver.
the implications of an incorrect determination (a suspect The significance of this new experiment is to teach students
could go to jail), thereby motivating correct analytical pro- that experiments are performed for a reason and that careful
cedures and careful interpretation of the data. Our students laboratory technique is extremely important. When placed
seemed to enjoy the added responsibility of using their in the context of a drunk-driving scenario with emphasis that
measurements as the basis for important forensic-based an innocent person could go to jail (or a guilty person could be
conclusions instead of simply calculating a possibly mean- set free) if sloppy experiments are performed, these obvious
ingless number. statements become clearer.
Experiments of a forensic nature can be easy to incor-
porate into existing laboratory formats using available instru- Experimental Procedure
ments. These experiments can be performed at all levels, as
evidenced by an introductory experiment in the identification Reagents
of arson accelerants for general chemistry students (3). In our Ethanol and n-propanol were obtained from Fisher
instrumental methods laboratory we have added forensic Scientific (Fairlawn, NJ) and bovine serum was obtained
experiments based on topics included in L. J. Kaplan’s course from Sigma (St. Louis, MO). Deionized water (resistance
for non-science majors, Chemistry and Crime: From Sherlock ≥18 MΩ) was obtained from a Milli-Q system (Bedford, MA).
Holmes to Modern Forensic Science (4, 5). Because our
course is taught on the quarter system (ca. 27 lectures, 50 Preparation of Stock Solutions
min each), time does not permit us to cover some of the A stock solution of ethanol (10 mg/mL in deionized
interesting topics taught by others (e.g., court testimony) that water) and two stock solutions of n-propanol (1:50 dilutions
are not directly related to chemical analysis. in both deionized water and 10 mg/mL ethanol) were prepared
The specific objective of this new experiment, designed for student use.
for a 4-hour laboratory period, is to perform qualitative and
quantitative analysis of blood alcohol content (BAC). BAC Preparation of Standards
is typically determined through analysis of breath (6 ) or blood Five ethanol solutions (0.2, 0.5, 0.8, 1.0, and 1.5 mg/mL)
(7, 8). The Breathalyzer, in particular, is a common method were prepared in 100-mL volumetric flasks using 10 mg/mL
used in student experiments for estimation of BAC (9–11). ethanol. Ten milliliters of the n-propanol/water stock solution
In our experiment students measure the alcohol level in a (internal standard) was added to each flask before diluting
simulated serum sample and convert the value to BAC. The to the mark with water. Tenfold dilutions of the n-propanol/

JChemEd.chem.wisc.edu • Vol. 78 No. 9 September 2001 • Journal of Chemical Education 1225


In the Laboratory

water and n-propanol/ethanol stock solutions were also prepared.


Fifteen milliliters was transferred immediately to a 20-mL
vial and sealed with a screw-top septum cap. To facilitate
transfer of volatile components into the vapor phase, each
vial was allowed to equilibrate at room temperature for at
least 15 min before the sample was withdrawn.
Preparation of Serum Unknowns
The teaching assistant prepared serum unknowns by
adding 850 to 990 µL of the 10 mg/mL ethanol solution to
10-mL volumetric flasks and diluting to the mark with bovine
serum, giving BAC values ranging from 0.0746 to 0.0868.
The students prepared serum unknowns for analysis by add-
ing 1.0 mL of n-propanol/water stock solution to a 10-mL
volumetric flask and diluting to the mark with the serum
unknown. After careful mixing, exactly 7.5 mL of serum
unknown was transferred from the volumetric flask to a
10-mL vial and sealed with a screw-top septum cap. A serum
blank (i.e., without ethanol) was prepared by adding 7.5 mL of
bovine serum to a 10-mL vial and sealing it with a screw-top
septum cap. Both the unknown and the blank were allowed
to equilibrate in the vials at room temperature for at least
15 min prior to analysis.
Figure 1. Chromatogram of 1.5 mg/mL ethanol standard.
Equipment
A Shimadzu GC-14APsF gas chromatograph/flame
ionization detector is used in this experiment. Chromatograms
are plotted with a chart recorder. The column is a 2-meter ×
1
⁄8-in. stainless steel column with a stationary phase of 10% 10000

Carbowax 1500 on Chromosorb W 60–80 mesh support.


A helium mobile phase (20–40 mL/min) is used. Hydrogen 8000
y = 5988.6x − 95.857
(30 mL/min) and compressed air (240 mL/min) are detector R 2 = .9952
Peak Area

gases. Injection, column, and detector temperatures are 175, 6000

80, and 200 °C, respectively. Headspace analysis is performed


by withdrawing and injecting exactly 500 µL of vapor from the 4000

vials using a 1-mL Hamilton Gastight syringe. The syringe


is flushed with air between runs to prevent cross-contamination. 2000

Hazards 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

There are no significant hazards associated with this 0.8


experiment. We recommend the use of latex gloves and
eye protection.
0.6 y = 0.4044x + 0.0007
R 2 = .9998
Area Ratio

Results
0.4

Qualitative Analysis
A representative chromatogram obtained by the students 0.2

is shown in Figure 1. Typical retention times for n-propanol and


ethanol ranged from 0.60 to 0.75 min and 0.38 to 0.48 min, 0.0
respectively. Owing to this variation, the relative retention 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

time (RRT) method was used for positive identification. The Ethanol Concentration / (mg mL᎑1)
RRT of an unknown peak is the ratio of the retention time Figure 2. Representative calibration curves used to quantitate ethanol
of the unknown peak to the retention time of an internal in serum samples: Top: External standardization using ethanol peak
standard. The average RRT value of ethanol (with the n-pro- area. Bottom: Internal standardization. Area ratio refers to ethanol
panol internal standard) was 0.637 ± 0.008 (n = 26), based peak area divided by n-propanol peak area. Two injections were
on student data. performed for each standard.

1226 Journal of Chemical Education • Vol. 78 No. 9 September 2001 • JChemEd.chem.wisc.edu


In the Laboratory

Quantitation legal limit of 0.080 by a relatively small margin. If they were


Ethanol is quantitated in the serum unknown using both real samples, even a seemingly small error by a technician
external and internal standardization methods. Examples of could have severe consequences. This is impressed upon the
external and internal calibration curves (using 0.2 to 1.5 mg/ students, many of whom may soon be working in a com-
mL ethanol standards) are shown in Figure 2. When students mercial laboratory. To achieve accurate results, students must
performed experiments carefully, both calibration curves typi- be meticulous in their laboratory practices. Careful technique
cally yielded R 2 > .98. can result in experimental BAC values within 5% of the actual
values; sloppy sample preparation or analysis is evidenced by
Determination of BAC BAC values that deviate as much as 50% from the actual
BAC is defined as grams of ethanol per 100 mL of blood. values. Overall, students found the experiment interesting
Since dealing with blood presents serious safety issues, and a vast improvement over other GC experiments they had
especially in an undergraduate teaching laboratory, bovine performed in the past.
serum is used. Alcohol does not partition equally between
red blood cells and blood serum owing to differences in water W
Supplemental Material
content (17). Therefore, to determine BAC using serum instead
of blood, a conversion factor (17) is required: Notes for the instructor and detailed instructions for
students are available in this issue of JCE Online.
serum ethanol concn (g ethanol/100 mL serum)
BAC =
1.14 Literature Cited
Compiled student data show that this experiment provides 1. Gaensslen, R. E.; Kubic, T. A.; Deslo, P. J.; Lee, H. C. J. Chem.
an accurate method for quantitation of alcohol in serum Educ. 1985, 62, 1058–1060.
samples. Using external standardization, the differences between 2. Brewer, W. E.; Lambert, S. J.; Morgan, S. L.; Goode, S. R.
actual BAC values and student values ranged from 0.48 to Teaching Forensic Analytical Chemistry; ChemConf ’98, On-
50.33%, averaging 15.13 ± 13.57% (n = 31). Nineteen per- Line Conference on Chemical Education 1998; http://
cent of the students were within 5% of the actual BAC and www.inform.umd.edu/EdRes/Topic/Chemistry/ChemConference/
50% were within 10% of the BAC. The results from internal ChemConf98/forensic/ChemConf.htm (accessed May 2001).
standardization are even better. Differences between actual and 3. Elderd, D. M.; Kildahl, N. K.; Berka, L. H. J. Chem. Educ.
student BAC values ranged from 0.44 to 40.68%, averaging 1996, 73, 675–677.
9.61 ± 9.37%. Using this method of quantitation, 33% of 4. Kaplan, L. J. http://www.williams.edu/Chemistry/lkaplan/
the students came within 5% of the actual value and 53% 113des.html (accessed May 2001).
came within 10% of the actual value. 5. Kaplan, L. J. Crime Lab. Digest 1992, 19, 107–132.
6. Labianca, D. A. J. Chem. Educ. 1990, 67, 259–261.
Conclusions 7. Berkebile, J. M. J. Chem. Educ. 1954, 31, 380–382.
8. Labianca, D. A. J. Chem. Educ. 1992, 69, 628–632.
Gas chromatography is a common method in commercial 9. Vitz, E.; Chan, H. J. Chem. Educ. 1995, 72, 920–925.
and forensic laboratories for determination of BAC in suspected 10. Timmer, W. C. J. Chem. Educ. 1986, 63, 897–898.
drunk drivers. Our experiment is designed for upper-division 11. Anderson, J. M. J. Chem. Educ. 1990, 67, 263.
chemistry students and serves several purposes. The students 12. CG Labs, Inc.: Concord, NH; http://www.cglabs.com/
gain experience using headspace GC as an analytical tool for cgl_pro.htm (accessed May 2001).
qualitative and quantitative analysis. The experiment also gives 13. Macchia, T.; Mancinelli, R.; Gentili, S.; Lugaresi, E. C.;
students an opportunity to compare external and internal Raponi, A.; Taggi, F. J. Anal. Toxicol. 1995, 19, 241–246.
methods of standardization. Although experiments that illustrate 14. Schuberth, J. J. Chromatogr. Sci. 1996, 34, 314–319.
such methodology are common, the novelty in this experiment 15. Baselt, R. C.; Cravey, R. H. Disposition of Toxic Drugs and
is that it goes one step further. It shows how these techniques Chemicals in Man, 3rd ed; Year Book Medical Publishers: Chi-
are used outside the classroom in real laboratories and in a cago, 1989; pp 322–326.
situation that is relevant to the students, especially those who 16. Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instru-
consume alcoholic beverages. mental Analysis, 5th ed; Harcourt Brace: Philadelphia, PA,
Students also learn that poor laboratory technique could 1998; p 707.
have penalties far worse than bad grades. The serum unknowns 17. Payne, J. P.; Hill, D. W.; Wood, D. G. L. Nature 1968, 217,
are designed to have BAC values that bracket the California 963–964.

JChemEd.chem.wisc.edu • Vol. 78 No. 9 September 2001 • Journal of Chemical Education 1227

You might also like