In the Laboratory
Determining the Percent Water in Organic Solvents Using the Zwitterionic Dimroth–Reichardt Betaine ET-30 Dye
W
An Industrially Relevant Application of a Previously Published Laboratory Experiment Mark F. Vitha Department of Chemistry, Drake University, 2507 University Avenue, Des Moines, IA 50311;
[email protected] 0.75
0.50
Absorbance
Recently in this Journal, Deng and Acree described an experiment in which the solvatochromic shift of the dye ET-30 (see structure below) is used to teach the fundamental basis of UV–vis absorbance spectroscopy (1). In this experiment, ET-30 is dissolved in a number of solvents of varying polarity and hydrogen bonding ability. Students record the wavelength of maximum absorption and are asked to explain the striking changes in λmax. As Deng and Acree discuss, the changes are due to differential solvation of the ground and excited states arising from the differences in the polarities of the various solvents, as depicted in Figure 1 in ref 1. The ability of the solvent to donate hydrogen bonds also affects the λmax of ET-30: hydrogen bond donating solvents stabilize the ground state more than they stabilize the excited state. Excellent discussions of solvatochromism and the spectroscopy of ET-30 can be found in refs 2–4.
Pure acetone 0.3% water
0.25
0.5% water 1.0% water 0 500
600
700
800
Wavelength / nm Figure 1. Sample spectra of ET-30 in acetone and in 0.3%, 0.5%, and 1.0% water–acetone solution. All spectra are scaled to give a maximum absorbance of 0.6 AU. +
N
O−
Ground-state structure of ET-30 2,6-diphenyl-4-(2,4,6-triphenyl-N-pyridino)phenolate
Using ET-30 in a Quantitative Experiment with Industrial Application I have used the experiment described by Deng and Acree in essentially the same format in the laboratory component of a senior-level instrumental analysis course, but I have also added a related experiment to illustrate to students how the application of theory can be used to solve relevant industrial problems. In this additional experiment, the phenomenon of solvatochromism is used to determine the percent water in organic solvents (5–7). The ability to measure the quantity of water in organic solvents is important because water can interfere with desired reactions, lead to unwanted by-products, and change the product distribution. Thus, the removal of water from solvents
370
and the measurement of the quantity of water in a solvent are important topics to teach undergraduate students. The experiment described below allows these topics to be discussed, provides a relevant industrial example of the complications caused by the presence of water in organic solvents, and demonstrates a simple instrumental method for the approximate determination of the percent water in a solvent.
Determination of the Percent Water in Organic Solvents Before determining the percent water in an organic solvent, students observe the sizable wavelength shift of ET-30 in several pure solvents, as described in ref 1. To determine the percent water in an organic solvent, they then create standard solutions of 0.0–1.0% (v/v) water in the chosen solvent (both acetone and acetonitrile have been used successfully), add ET-30 (absorbances are kept below 0.8), and measure the λmax of ET-30 in these solutions. Because ET-30 is so sensitive to the polarity and hydrogen-bond-donating ability of its chemical environment, the added water induces measurable (and visible) changes in λmax. Students create a calibration curve of the energy of transition, E T(30), versus percent water and use it to determine the water content in an unknown solution. Values of E T(30)
Journal of Chemical Education • Vol. 78 No. 3 March 2001 • JChemEd.chem.wisc.edu
In the Laboratory
are calculated from λmax (in nm) using eq 1 and have units of kcal/mol (5).
28,590 λmax
(1)
Since the goal of the experiment is to provide a quantitative application of solvatochromism, rather than to determine the absolute water content of a solvent, heroic efforts are not made to rigorously dry the solvent being studied. The solvent is dried using a typical adsorbent such as calcium sulfate. Since it is unlikely that all the water is removed, what is actually being determined is the percent added water (i.e., the water content above that of the “dried” solvent). While it is difficult to rigorously dry acetone, it has been reported that HPLC-grade acetonitrile contains just 0.017% (w/w) water even after the bottle has been open for three weeks (8). In some ways the incomplete removal of water is a limitation of the experiment when acetone is used, but it does allow for the discussion of the drying of solvents, and thus extends the educational opportunities afforded by the experiment. This limitation could potentially be turned into a separate laboratory experiment in which multiple methods of drying solvents are tested by first recording the λmax of ET-30 in the “wet” solvent, and then recording the λmax of ET-30 in samples of the solvent that have been subjected to a variety of drying agents such as molecular sieves, calcium sulfate, or a variety of other techniques. The method that yields the largest shift compared to the “wet” solvent could then be categorized as the best of the methods tested (although an absolute value of the percent water could not be determined if the solvent were not rigorously dried). I have not personally tried this experiment, I but see it as a logical extension of the laboratory analysis described in this article. Students use a scanning spectrophotometer to determine the percent water in an organic solvent. Pedagogically, this affords the opportunity to introduce them to the improved wavelength resolution that allows finer differentiation of the smaller shifts observed here compared with those in the bulk solvent experiments. The determination of the percent water in acetone can also be accomplished using a diode array spectrophotometer with 2-nm resolution because the shift in λmax of ET-30, unlike that of most other solvatochromic dyes, is quite large, changing by approximately 40 nm between pure acetone and acetone with 1% added water. Sample spectra of ET-30 in “dry” acetone and acetone with 0.3%, 0.5%, and 1.0% water, obtained using a diode array spectrophotometer, are shown in Figure 1 to provide a sense of the total shift observed. In general, plots of E T(30) versus percent water in organic solvents are not linear (5, 6 ). The students verify this experimentally by measuring the λmax in solutions of higher percent water, such as 10% and 50%, and adding these points to their graphs in the 0.0–1.0% region. In the limited range of 0.0–1.0% water, however, a linear approximation can be applied (see Fig. 2). It is useful to explicitly discuss with the students the possible origins of the nonlinearity of the curve over a wide range of percent water and the linear approximation in a limited region. These discussions increase students’ awareness of the fact that not all physical processes are linear and can help reduce the frustration borne of a lack of familiarity with nonlinear calibration curves.
ET / (kcal/mol)
ET =
44.5
44.0
43.5
43.0
42.5 0
0.2
0.4
0.6
0.8
1.0
Percent water in acetone (v/v) Figure 2. Calibration curve of E T (30) versus the percent water in acetone (v/v): • experimental data points; ––– best straight line determined by linear regression analysis.
Relevant Industrial Applications To make the determination of the percent water in an organic solvent relevant to the students, a case study from Procter & Gamble’s workshop “Professional Analytical Chemists in Industry—A Short Course in Problem Solving for Undergraduate Students” (9) is used.1 The case is titled “What Caused the Drums to Bulge?” (10). It describes a situation in which drums of an ethoxylated alcohol were bulging owing to the generation of an explosive gas inside the drums. Several hypotheses about the identity and the origin of the gas are given, along with the relevant chemistry. A series of common analyses (both instrumental and wet chemical) are described and it is ultimately discovered that hydrogen was causing the drums to bulge. The cause of the hydrogen gas production was determined to be a combination of excessive water in the drums (tested by Karl Fisher titrations) and a somewhat low pH for the raw material (9.5–10.5), under which conditions the water was reduced to hydrogen by sodium borohydride, an antioxidant/stabilizer added to the ethoxylated alcohol. As part of the laboratory report, students are required to list and briefly describe all the analytical tests performed, as well as the purpose, results, and conclusions drawn from each test. They are also expected to comment on the connections between the case study and the determination of percent water they performed in the laboratory—an obvious connection but one that solidifies the relevance of the laboratory work. Finally, students are asked to make the link between the quantitative aspect of this experiment and the theoretical basis of UV–vis absorbance spectroscopy explored in the first half of the experiment (the part described in ref 1). Hazards Acetone (CAS #67-64-1) is a flammable solvent. ET-30 (Reichardt’s Dye, CAS #10081-39-7) is an irritant whose chemical, physical, and toxicological properties have not been thoroughly investigated (information obtained from the MSDS). Calcium sulfate (CAS #7778-18-9) is an irritant.
JChemEd.chem.wisc.edu • Vol. 78 No. 3 March 2001 • Journal of Chemical Education
371
In the Laboratory
Conclusion
W
Overall, by combining the experiment described in ref 1 with the experiment described here, students learn
A student handout for this experiment and the Procter & Gamble case study are available as supplemental material in this issue of JCE Online.
1. the fundamental basis of UV–vis spectroscopy, 2. the chemical importance of dry solvents for organic reactions, 3. a quantitative application of solvatochromic shifts, 4. the differences between diode array and scanning spectrophotometers, 5. a variety of instrumental and wet-chemical techniques used in industrial analytical laboratories, and 6. a scientific approach for providing solutions to analytical problems.
Thus this experimentally simple laboratory exercise, which can be completed in a single 4-hour laboratory period, provides a wide variety of educational opportunities and directly connects fundamental theory with applications of the theory— a connection sometimes not made by students on their own. Acknowledgments I thank the Procter & Gamble course instructors, especially Alan Ullman, for their helpful comments regarding this manuscript, their permission to discuss the case study, and their permission to post their material on JCE Online. I also thank the students from Drake University and the University of Minnesota–Duluth who have performed this experiment and offered suggestions for its improvement.
372
Supplemental Material
Note 1. Information about the Procter & Gamble workshop can be found in reference 9 and at their Web site: http://www.pg.com/ about/rnd/ashort.htm.
Literature Cited 1. Deng, T.; Acree, W. E. Jr. J. Chem. Educ. 1999, 76, 1555. 2. Kamlet, M. J.; Abboud, J. L.; Taft, R. W. J. Am. Chem. Soc. 1977, 99, 6027. 3. Reichardt, C. Solvents and Solvent Effects in Organic Chemistry; VCH: Weinheim, 1988; pp 285–311. 4. Carr, P. W. Microchem. J. 1993, 48, 4. 5. Langhals, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 724. 6. Johnson, B. P.; Gabrielsen, B.; Matulenko, M.; Dorsey, J. G.; Reichardt, C. Anal. Lett. 1986, 19, 939. 7. Kumoi, S.; Oyama, K.; Yano, T.; Kobayashi, H.; Ueno, K. Talanta 1970, 17, 319. 8. Dunn, B. C.; Ochrymowycz, L. A.; Rorabacher, D. B. Inorg. Chem. 1995, 34, 1954. 9. Wilkinson, S. Chem. Eng. News 1999, 77, 39. 10. Thorpe, T. M. Anal. Chem. 1984, 56, 603A.
Journal of Chemical Education • Vol. 78 No. 3 March 2001 • JChemEd.chem.wisc.edu
