ARTICLE pubs.acs.org/jchemeduc
Fostering Multirepresentational Levels of Chemical Concepts: A Framework To Develop Educational Software Guilherme A. Marson*,† and Bayardo B. Torres‡ †
Department of Chemistry and ‡Department of Biochemistry, University of S~ao Paulo, S~ao Paulo 05508-000, Brazil ABSTRACT:
This work presents a convenient framework for developing interactive chemical education software to facilitate the integration of macroscopic, microscopic, and symbolic dimensions of chemical concepts—specifically, via the development of software for gel permeation chromatography. The instructional role of the software was evaluated in a study involving 237 undergraduates. The results suggest that the software fostered transition from lower- to higher-order cognitive reasoning. Common misconceptions regarding microscopic phenomena were also detected and addressed. KEYWORDS: First-Year Undergraduate/General, Interdisciplinary/Multidisciplinary, Computer-Based Learning, MultimediaBased Learning, Chromatography, Student-Centered Learning, Undergraduate Research
’ INTRODUCTION A central goal in science education is to promote scientific literacy.1,2 Scientific literacy places science education into a larger framework, composed of four major perspectives: conceptual (scientific knowledge); epistemological (the nature of scientific knowledge); social (the interaction between science, culture, and society); and learning and cognition (the way people learn). From these four viewpoints, a variety of initiatives to increase scientific literacy can be proposed and effectively carried out. These initiatives may facilitate concrete actions at different levels of the educational system: curriculum development; preservice and in-service teacher education, assessment, and evaluation standards; learning resources and didactic materials; trips to science museums and other informal learning spaces; and more involved school community activities. Key issues in the effectiveness of those initiatives include conveying scientific information and promoting significant learning of the associated concepts. Several studies3 8 have indicated that difficulties in learning chemistry arise from the fact that chemical concepts are presented in different representational modes: macroscopic; microscopic (or semimicroscopic); and symbolic. Significant learning requires the incorporation of all of these dimensions. Science educators have proposed a variety of strategies to integrate these modes, often resorting to visual analogies that play a central role not only in chemical education but also in the way chemists elaborate, express, and convey scientific concepts.9 16 The current digital information era offers a plethora of multimedia-based educational resources, such as 3D molecular renderers, tutorials, and simulators.17 23 Several studies have indicated Copyright r 2011 American Chemical Society and Division of Chemical Education, Inc.
that appropriate choices for visual analogies and interface interactivities can facilitate the integration of the representational modes of chemical concepts. There is also evidence that the instructional potential of educational software is enhanced during student-centered activities, which foster group work, peer interaction, and collaborative learning.24 33 In this paper, we present a useful framework for developing software that addresses the integration of macroscopic, microscopic, and symbolic representational dimensions of chemical concepts. This framework served as a guide for developing the software packages called Principles of Gel Permeation Chromatography34 and Applications of Gel Permeation Chromatography.35 This study also evaluates the usefulness of that software for various users.
’ FRAMEWORK FOR THE DEVELOPMENT OF THE SOFTWARE The proposed framework for the development of the software entails four major steps: 1. selecting chemical content 2. designing software structure and content distribution 3. integrating the macroscopic, microscopic, and symbolic modes 4. setting the exploration mode These steps follow the protocol specifications that were used to develop the software on gel permeation chromatography (GPC). Published: October 05, 2011 1616
dx.doi.org/10.1021/ed100819u | J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education
ARTICLE
Table 1. Software Structure and Content Distribution Relative to Representational Modes
Selecting the Chemical Content
To illustrate our framework, this paper uses GPC as a paradigm because of the importance of GPC to chemical knowledge and its particular conceptual characteristics.36,37 Chromatography is ubiquitous in the curriculum of science courses offered in higher education. Rapid advancements in chromatography require students to understand the underlying physical chemical principles that enable separation: namely, the concept of dynamic binding equilibrium. Students must also understand experimental procedures and outcomes, and be able to integrate their graphic representations and molecular models. Another reason for developing chromatography software stems from the fact that experiments using the traditional technique are highly time-consuming.38 40 Appropriate use of software simulation would allow students more time for planning in-depth experiments and analyzing data. For this software’s simulations, we chose to use three cases of the system hemoglobin and methylene blue; this included a pure sample of each component, as well as a mixture of both. Our decision was based on three factors: (i) the distinct colors of the components; (ii) their 200-fold molar mass difference; and (iii) the possibility to visually follow their separation in a lab class. Designing the Software Structure and Content Distribution
The instructional software presents two independent modules: Principles and Applications. The Principles module presents the essential concepts of GPC separation and then requires
students to use low-order cognitive skills to grasp those concepts. The Applications module presents the same concepts in the context of specific applications of the technique: students perform a virtual determination of molar mass and a deduction of the affinity equilibrium constant via a chromatogram—and this mobilizes high-order cognitive skills. Most of the software’s content is conveyed by images and animations. A minimal amount of information is delivered as text. Therefore, students have to infer the chemical concepts by actively exploring visual resources. The Principles module has three sections. The first section, Real Lab, shows pictures of the experimental apparatus and illustrates the major steps of chromatographic separation of a liquid mixture of methylene blue and hemoglobin. Virtual Lab, the second section, presents animations that simulate the same processes shown in the Real Lab section. The third section, Microscopic Model, shows an animation of the separation phenomenon at the microscopic level. Further details of the software’s structure and content distribution are outlined in Table 1. Integrating Macroscopic, Microscopic, and Symbolic Modes
The Principles module software is intended to present and facilitate a sound understanding of the main concepts of GPC separation, which is the process in which substances with different molar masses are separated as their molecules interact differently with the stationary phase during elution. 1617
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education
ARTICLE
fundamental to integrating macroscopic and graphical representations of chemical concepts. Molecular Model, the final section of the Principles module software, displays microscopic representations of the processes that were shown in the previous sections (Figure 1C). The microscopic dimension is connected to the symbolic and macroscopic dimensions by adopting an appropriate color code to represent methylene blue and hemoglobin throughout the software (samples in the photographs, chromatogram lines, and molecules). Further representational integration was achieved by adjusting the diffusional range of the sample spots in all representations of the column adopted in the software. Particle motion and distribution in the Microscopic Model section are proportional to the corresponding sample spots in the animated column of the Virtual Lab section, which, in turn, is directly related to those pictured in the Real Lab section. Special attention must be given to the fact that by coloring microscopic particles with the same color of the substances one might raise or reinforce animistic misconceptions on the particulate nature of matter.41 44 Therefore, it is necessary to make clear for the students that the animation is a dynamic representation of a model, not a picture of the events occurring at the molecular scale. Setting the Exploration Mode
Figure 1. Sections within the Principles module of the GPC software. A: Actual lab. B: Virtual lab. C: Microscopic model.
The content was organized in terms of the technique’s experimental elements: samples (the components to be separated); column (the separation medium); and elution (the separation process). These contents are featured in each of the three sections of the Principles module. Visual representations were created to facilitate symbolic transposition among the multiple representational modes. Therefore, all analogies served as both translational and scaffolding elements to simultaneously convey and integrate the three sections of the Principles module. Real Lab delivers photographs that are macroscopic, qualitative, and concrete representations of the lab apparatus and the chromatographic run (Figure 1A). Virtual Lab simulates the same processes as realistic, nonschematic animations of the elution process in order to link the photographic representation to the simulation output, which is more symbolic (Figure 1B). Additionally, in the Virtual Lab, another representation of the separation is introduced: the chromatogram. As a graphic expression of the process, the chromatogram adds a quantitative dimension to the previous qualitative information. To foster the transition from the qualitative (column) to the quantitative (chromatogram), fraction tubes were also displayed in this section. Tube filling is synchronized to sample elution and pointwise plotting of the chromatogram, so that the color intensity of each tube is proportional to the fraction absorbance. This way, the array of tubes portrays the time dimension of the separation while also reinforcing the connection between symbolic quantitative (chromatogram) and macroscopic qualitative (column) representations. Therefore, in the Virtual Lab section, both the images and the pace of the computer animations were
The software’s interactive nature34,35 was designed to address two important issues. First, it allows students to explore the various concepts at their own pace. Second, it mimics the key steps required to perform GPC experiments: sample injection onto the column, elution and fraction collection, and chromatogram analysis. Additionally, the interface was designed to allow students to focus on learning the concepts rather than learning to operate the software. This was achieved by employing well-known interface elements and behaviors, which are widely used on the Internet and in other media already familiar to the students.
’ EVALUATION OF THE INSTRUCTIONAL ROLE OF THE SOFTWARE The software34,35 was tested by 237 students from different courses and universities; use of the software was one of the activities incorporated into the students’ regular biochemistry courses (specifically, it was introduced when studying the topic of protein purification methods). One group of students explored only the Principles software, and the other group explored both the Principles and the Applications software. Students were organized in groups of three people per computer. Groups were given the task of answering a set of questions using the software as their only source of information. They could spend no more than two hours on the activity. The two groups of students who participated in this study are described below. Subset 1: Students Exploring Principles of GPC
Subset 1 (S1) was composed of students with different backgrounds in chemistry: first-year biology students, first-year nutrition students, and second-year pharmacy students. The students in biology and nutrition had not completed university-level courses in general chemistry and analytical chemistry. In general, nutrition students are often less acquainted with chemical concepts, and biology students are usually more science-oriented than students enrolled in nutrition courses. Subset 1 was composed of students from these courses: 1618
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education
ARTICLE
Table 2. Distribution of Likert Categories for Different Audiences student responses, in % S1 statements for response I: Instructions given within the software made it easy to use.
II: The software contains sufficient information to understand the technique.
III: Drawings and symbols present in the software are familiar and representative.
IV: The navigation of the software is appropriate.
V: The previous usage of the software facilitates the execution of the technique in the lab.
a
categoriesa
Nut, n = 36
Bio, n = 45
S2 Ph1D, n = 54
Ph2D, n = 45
Ph2N, n = 57
TA
11
31
48
0
0
A
44
69
52
82
84
M
20
0
0
18
0
D
25
0
0
0
16
TD
0
0
0
0
0
TA
5
42
17
30
20
A
45
28
52
40
60
M D
30 20
30 0
8 23
10 20
0 20
TD
0
0
0
0
0
TA
0
40
52
40
60
A
70
60
48
60
40
M
30
0
0
0
0
D
0
0
0
0
0
TD
0
0
0
0
0
TA A
22 58
46 29
42 41
30 50
21 79
M
10
25
13
10
0
D
10
0
4
10
0
TD
0
0
0
0
0
TA
39
50
42
40
20
A
56
50
50
40
60
M
5
0
2
10
20
D TD
0 0
0 0
6 0
8 2
0 0
TA: totally agree. A: agree. M: maybe. D: disagree. TD: totally disagree.
• Biology (Bio): 45 first-year biology students from the State University of Campinas • Nutrition (Nut): 36 first-year nutrition students from the University of S~ao Paulo (USP) • Pharmacy (Ph1D): 54 second-year pharmacy students (day course) from USP Subset 2: Students Exploring Principles and Applications of GPC
Subset 2 (S2) was composed of second-year pharmacy students who are typically interested in chemistry and have already completed general chemistry and analytical chemistry courses. Students in subset 2 were from these courses: • Pharmacy (Ph2D): 45 second-year pharmacy students (day course) from USP • Pharmacy (Ph2N): 57 second-year pharmacy students (night course) from USP
Students’ performances were assessed by their scores on three multiple-choice questions (Q1, Q2, and Q3 in Table 3); these questions probed for information about the chemical principle of GPC separation at increasing cognitive levels.46 49 Question 1 required students to understand the principle of GPC separation at a low-order cognitive level. The correct answer (molecular size) can be directly inferred from the Principles software. Questions 2 and 3 deal with the same concepts, but at higher-order cognitive levels. Answering Q2 and Q3 requires students to do the following: (i) interpret the information present in the software; (ii) understand the concept the information demonstrates; and (iii) apply the concept in two hypothetical, limit situations (a fully porous and a fully nonporous stationary gel phase). It is worth noting that answering Q1 correctly required only partial integration of the representational modes, whereas Q2 and Q3 could be answered correctly only if the three modes were fully integrated in a student’s mind.
Data Generation
Students’ opinions about the software were registered using a Likert scale45 for each of the following features: interface guidance and visuals; navigation; chemical information; and usefulness as a preparatory activity for an actual experiment (Table 2). In addition, students’ spontaneous comments and instructors’ observations were also incorporated into the data set.
’ RESULTS AND DISCUSSION Table 2 shows data concerning students’ impressions of the software. Although students in both subsets displayed a positive perception of the software overall, some differences in opinion are apparent, especially regarding the distribution of categories 1619
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education
ARTICLE
Table 3. Distribution of Some Answers among Different Groups of Students
for statements I and II (e.g., totally agree, TA, and agree, A). Statement I (“Instructions ... to use ...”) refers to the software interface, and statement II (“... information to understand ...”) refers to the chemical content in the interface. These elements of the software are indissociable. However, it is reasonable to assume that discerning statement I (interface: to use) from statement II (chemical concepts: to understand) is an indicator of the learners’ acquaintance with the chemical concepts involved. It is also related to the way learners use information to accomplish an assignment: either by understanding information or by simply transferring it from the source to the answer sheet. In this sense, it is noteworthy that the evaluation of statements I and II by the nutrition students in S1 differs significantly from evaluations by students in other programs. First, the occurrence of categories TA and A in statement I is remarkably lower for nutrition students (55%) compared to the others (80 100%). Furthermore, the distribution of responses to statements I (use) and II (understand) is quite similar among the nutrition students, but quite different relative to the other students. Because the results for statements III (visual representation) and IV (navigation) are similar and mostly positive for all students, it is reasonable to attribute the singularities in the answers for statements I and II to differences in students’ academic backgrounds and not to differences in their visual and digital skills.
Table 4. Percentage of Correct Choices for Q1 among Those with Correct Choices for Q2, Q3, and Both Q2 and Q3 S1 students
Q2
Q3
Q2 and Q3
Nut
100
100
100
Bio
100
100
100
70
80
100
Ph1D
The instructional role of the software can be discussed based on the performance of S1 and S2 students in questions involving GPC concepts at low- and high-order cognitive skills, shown in Table 3. A first glance at the data presented in Table 3 indicates that both S1 (Principles) and S2 (Principles and Applications) students scored equally on Q1, which measures comprehension of GPC concepts at a lower-order cognitive level. The majority of the students in all audiences opted for the correct answer. It is also remarkable in Table 3 that S2 students displayed a significantly different distribution profile of answers for Q2 and Q3 (which measure chemical concepts at higher-order levels). S2 student options for the correct answers were considerably higher than those of S1 students. As one could presume from the fact that S2 students also explored the Applications software, a considerable number of these individuals were able to deal with 1620
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education limit situations more accurately. This interpretation indicates that exploring the Principles software alone facilitates learning concepts at low-order cognitive levels and, at least, scaffolds partial representational integration. Comparing results from S1 and S2 students suggests that further exploration of the Applications software presumably fosters learning concepts at high-order cognitive levels. This suggests that using both Principles and Applications software enhanced representational integration. Assuming that higher-order cognitive reasoning comprehends lower-order cognitive reasoning as well, it is expected that students’ performance on Q2 and Q3 would be somehow related to their performance on Q1. This assumption was investigated by analyzing answers to Q1 for each of the subpopulations of students who chose correct answers to either or both Q2 and Q3 (Table 4). The data presented in Table 4 reveal that the vast majority of S1 students who chose the correct answer for Q2 and Q3 also chose the correct answer in Q1. This finding indicates that it is very unlikely that the results presented so far in Tables 2 and 3 actually result from random choices. Observing the performance of S1 students on Q2 (Table 3), it is quite interesting to notice that they exhibit a very similar profile of answers: ∼45% chose the correct answer (C); option A was the most common incorrect choice for the entire S1 subset; and other choices were almost equally distributed. Notice that these similarities occur despite the different backgrounds of the audiences in S1. Two possible explanations for these similarities could pertain. One possibility is that, for most students, chemistry backgrounds did not significantly interfere with their performance on Q2 and Q3. Another explanation might be that students tended to choose the chromatogram (answer D) displaying an elution order inverse to the one shown in the software, indicating that they took the limit situation (nonpermeable gel phase) for the opposite situation. S1 results for Q3 are more heterogeneous among students in different programs. This might be because the limited situation proposed in Q3 (fully porous gel phase) is more difficult to visualize than the one in Q2; in Q2, a rigid sphere analogy can be used to visualize a nonporous gel matrix. In addition to the answer profiles, instructors’ observations were recorded on students’ spontaneous comments. These data were compiled together with students’ written notes. The most frequent comments fell into at least one of these categories: 1. “More content is needed to answer questions 2 and 3.” Many students in both audiences expected the software to display the same situations that were presented in Q2 and Q3, instead of trying to infer them from the concepts in the software. 2. “Lack of previous knowledge.” Some of the students complained that their previous knowledge was insufficient to fully understand the software. This sort of comment was more common among nutrition students. 3. “The separation is based on molar mass because hemoglobin, which is heavier, falls faster through the column.” Comments like this were made by a few students in all audiences. The nature of these comments is very similar to animist ideas about the particulate nature of matter. In this case, students seem to consider gravity, and not elution, as the driving force in the size-based separation. These and other misconceptions were clarified in a group discussion activity following the software exploration session.
ARTICLE
’ CONCLUSIONS The study reported in this paper suggests that the proposed framework might be a valuable resource to create software integrating macroscopic, microscopic, and symbolic representation modes of chemical concepts. In the case of GPC, the use of both Principles and Application software was associated with a low-order to high-order cognitive improvement in learning, which is indicative of representational integration. A clear picture of the role of the software in the learners’ process of knowledge construction remains elusive and requires further research to improve the software development framework. We believe that the development approach discussed in this paper is applicable to several other topics in chemistry.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT We are thankful to FAPESP, http://www.fapesp.br (accessed Sep 2011), for funding the research, to the Institute of Chemistry of the University of S~ao Paulo, and to the Institute of Biology at the State University of Campinas. ’ REFERENCES (1) Hodson, D. Sci. Educ. 1999, 83, 775–796. (2) Hodson, D. Int. J. Sci. Educ. 2003, 25, 645–670. (3) Herron, D.; Nurrenbern, S. C. J. Chem. Educ. 1999, 76, 1353– 1361. (4) Treagust, D. F.; Chittleborough, G.; Mamiala, T. L. Int. J. Sci. Educ. 2003, 25, 1353–1368. (5) Chittleborough, G.; Treagust, D. F. Res. Sci. Educ. 2008, 38, 463–482. (6) Chandrasegaran, A. L.; Treagust, D. F.; Mocerino, M. J. Chem. Educ. 2009, 86, 1433–1436. (7) Adadan, E.; Irving, K. E.; Trundle, K. C. Int. J. Sci. Educ. 2009, 31, 1743–1775. (8) Hand, B.; Choi, A. Res. Sci. Educ. 2010, 40, 29–44. (9) Johnstone, A. H. J. Comput. Assist. Learn. 1991, 7, 75–83. (10) Kozma, R. B.; Russell, J. J. Res. Sci. Teach. 1997, 34, 949–968. (11) Wu, H.-K. Sci. Educ. 2003, 87, 868–91. (12) Pedersen, J. E.; Yerric, R. Y. J. Sci. Teach. Educ. 2000, 11, 131–153. (13) Cook, M. P. Sci. Educ. 2006, 90, 1073–1091. (14) Rappoport, L. T.; Ashkenazi, G. Int. J. Sci. Educ. 2008, 30, 1585– 1603. (15) Gilbert, J. K. Visualization: An Emergent Field of Practice and Enquiry in Science Education. In Visualization: Theory and Practice in Science Education; Gilbert, J. K., Reiner, M., Nakhle, M., Eds.; Springer: New York, 2008; pp 3 24. (16) Ferreira, C.; Arroio, A. Probl. Educ. 21st Century 2009, 16, 48–53. (17) Burke, K. A.; Greenbowe, T. J.; Windschitl, M. A. J. Chem. Educ. 1998, 75, 1658–1661. (18) Richardson, D. C.; Richardson, J. S. Biochem. Mol. Biol. Educ. 2002, 30, 21–26. (19) Honey, D. W.; Cox, J. R. Biochem. Mol. Biol. Educ. 2003, 31, 356–362. (20) Pavkovic, S. F. J. Chem. Educ. 2005, 82, 167. (21) Charistos, N. D.; Tsipis, C. A.; Sigalas, M. P. J. Chem. Educ. 2005, 82, 1741. (22) Burkholder, P. R.; Purser, G. H.; Cole, R. S. J. Chem. Educ. 2008, 85, 1071–1077. (23) Dalgarno, B.; Bishop, A. G.; Adlong, W.; Bedgood, D. R., Jr. Comput. Educ. 2009, 53, 853–865. 1621
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
Journal of Chemical Education
ARTICLE
(24) Wu, H. K.; Krajcik, J. S.; Soloway, E. J. Res. Sci. Teach. 2001, 38, 821–842. (25) Kozma, R. B. J. Res. Technol. Educ. 2003, 36, 1–14. (26) Kozma, R. B. Learn. Instr. 2003, 13, 205–226. (27) Galembeck, E.; Macedo, D. V.; Torres, B. B. Biochem. Mol. Biol. Educ. 2003, 31, 204–208. (28) Snir, J.; Smith, C. L.; Raz, G. Sci. Educ. 2003, 87, 794–830. (29) Habraken, C. L. J. Sci. Educ. Technol. 2004, 13, 89–94. (30) Kearney, M. Res. Sci. Educ. 2004, 34, 427–453. (31) Barak, M.; Dori, Y. D. Sci. Educ. 2005, 89, 117–139. (32) Costa M. J.; Galembeck E.; Marson G. A.; Torres, B. B. PLoS Comput. Biol. 2008, 4, e1000035; 10.1371/journal.pcbi.1000035. (33) Frailich, M.; Kesner, M.; Hofstein, A. J. Res. Sci. Teach. 2009, 46, 289–310. (34) Marson, G. A.; Torres, B. B. J. Chem. Educ. 2006, 83, 1567 1568; 10.1021/ed083p1567.2. See the Web-based, interactive Flash animation, Principles of Gel Permeation Chromatography, at http://www.jce.divched.org/JCEDLib/WebWare/collection/reviewed/ JCE2006p1567_2WW/index.html (accessed Sep 2011). (35) Marson, G. A.; Torres, B. B. Experimental Methods in Biochemistry: Interactive Software. http://www.iq.usp.br/bayardo/softwares/ english/metexbioq/index.html (accessed Sep 2011). (36) Hagel, L. J. Chromatogr., A 1993, 648, 19–25. (37) Shi, Y.; Xiang, R.; Horvath, C.; Wilkins, J. A. J. Chromatogr., A 2004, 1053, 27–36. (38) Haddad, P. R.; Shaw, M. J.; Madden, J. E.; Dicinoski, G. W. J. Chem. Educ. 2004, 81, 1293–1298. (39) Stone, D. C. J. Chem. Educ. 2007, 84, 1488–1496. (40) Shalliker, R. A.; Kayillo, S.; Dennis, G. R. J. Chem. Educ. 2008, 85, 1265–1268. (41) Williamson, V. M.; Abraham, M. R. J. Res. Sci. Teach. 1995, 32, 521–534. (42) Yang, E.-M.; Greenbowe, T. J.; Andre, T. J. Chem. Educ. 2004, 81, 587. (43) Papageorgiou, G.; Johnson, P.; Fotiades, F. Res. Sci. Technol. Educ. 2008, 26, 165–183. (44) Levy, S. T.; Wilensky, U. J. Sci. Educ. Technol. 2009, 18, 243–254. (45) Likert, R. Arch. Psychol. 1932, 140, 1–55. (46) Zoller, U.; Tsaparlis, G. Res. Sci. Educ. 1997, 27, 117–130. (47) Zoller, U. J. Res. Sci. Teach. 1999, 36, 583–596. (48) Zoller, U. Int. J. Sci. Educ. 2002, 24, 185–203. (49) Oliver-Hoyo, M.; Justice, J. J. Coll. Sci. Teach. 2008, 37, 62–67.
1622
dx.doi.org/10.1021/ed100819u |J. Chem. Educ. 2011, 88, 1616–1622
