FROM THE ACS MEETING
As computation and modeling carve an ever-deeper niche in the field of chemistry, the chemists of tomorrow need to learn these tools and techniques today.
In addition to imparting to their students traditional wet laboratory skills, educators need to produce computer-savvy graduates who are ready to design molecules, model spectra, or analyze reams of data in silico.
"It's increasingly important for students to integrate fully the use of computers in learning, as well as doing, chemistry," said
Theresa Julia Zielinski, chemistry professor at Monmouth University, West Long Branch, N.J. At a symposium Zielinski chaired at the ACS meeting in Boston last month, sponsored by the Computers in Chemistry Division, speakers demonstrated that the computer-as-educational-tool is far more than a means for automating quizzes and homework.
Rather, computers can be sophisticated learning aids, helping students visualize complicated three-dimensional concepts such as potential energy surfaces or molecular docking. They can facilitate long-distance collaboration. They're also invaluable feedback tools, allowing educators to evaluate how well their students are learning.
Students today are also unprecedentedly computer savvy. "They're used to interacting on the computer now," said symposium speaker James B. Foresman, chemistry professor at York College of Pennsylvania. "It's more natural for them to do that than to start up an instrument or other piece of equipment," he said. When students enter a chemistry class, many of them can already do some calculations within programs like Netscape. "They can then gradually work into more advanced things," Foresman said.
And in fact, computer-based teaching techniques and tools have "dramatically improved--within the last year, even," thanks to factors such as increased computer power and more sophisticated software, Foresman said. What that can mean for students is to "be able to visualize at the molecular level or be able to take a spectrum obtained experimentally and see what each peak means in terms of the electrons in a molecule," Foresman said.
Both Foresman and Augustus W. Fountain III, director of the photonics research center at the U.S. Military Academy, West Point, gave symposium attendees an overview of educational computer programs used in chemistry classrooms.
West Point has a history of keeping on top of technology: Each student has been issued a computer since the 1980s, and classrooms have networked workstations and Web access.
"We continue to keep a finger on the pulse of what's going on," Fountain said. "As different types of technology come up, we try to get our foot in the door early and test it."
For example, the CPS (Classroom Performance System) is an online service that allows students to, among other things, take in-classroom quizzes anonymously. The researchers tested CPS in eight of 46 chemistry classes. The system allowed instructors to see how well students were understanding certain concepts, but there was "no dramatic improvement in student preparation or performance," Fountain said.
Another online system, WebCT, lets students submit homework electronically as well as take multiple-choice quizzes. Students get instant feedback on their performance, and instructors can see "how many people completed the homework, what answers students missed, and what they did well in. They can then discuss the information in class," he said.
The disadvantage of such systems, however, Fountain said, is that any interruption in the service, such as a downed server, prevents students from getting access to their work. Additionally, students can fall into the habit of guessing answers rather than doing the work.
Troy A. Wolfskill and David M. Hanson, chemistry professors at the State University of New York, Stony Brook, are trying to circumvent the pitfalls of guessing and pattern recognition that plague some learning systems. For the past five years, they've been developing a program called LUCID (Learning and Understanding through Computer-based Interactive Discovery), which, they say, helps students actually understand rather than just memorize concepts.
"Part of the trouble is that chemistry textbooks have so much information, the students are overwhelmed--they don't know what to do," Hanson said.
LUCID's questions, which require understanding to answer, help students focus on what's important, according to Wolfskill and Hanson. "For example, if they learned something by reading it in a textbook, we might ask them about the concept by asking them to draw a picture," Wolfskill said.
LUCID also uses a network, where students can get answers to questions and then evaluate them with their peers. "So essentially, they can get a sense of how people are agreeing or disagreeing on ideas," Wolfskill said.
THE IMMEDIATE feedback is a plus as well, they said. "Students hate to do problems and then wait a week to find out what the answer is supposed to be," Hanson said.
In the early days, however, when working with computer-based learning, instant feedback functions just gave the right answer. People soon learned, however, that students could plug in any old answer to get the right one.
So Wolfskill devised a strategy whereby the computer divulges only the student's correct responses. "If a student gets the wrong answer, the computer says 'you did this or that right, but you have the wrong answer,' " Hanson said.
"I want students to use a textbook," Wolfskill added. Providing them with a sense of what they've done correctly allows students to go to their text for suggestions to learn what needs to be done to get the right answer.
An unexpected plus turned up when students who used LUCID came back for follow-up interviews, Hanson and Wolfskill said. They found that the students who had later interviews spent time studying the material on their own, apparently motivated by curiosity generated through using the program. "That's really a fringe benefit I didn't expect," Hanson said.
Katherine I. Barnhard and John W. Moore, chemistry professors at the University of Wisconsin, Madison, are also trying to steer students toward problem solving. Barnhard described their interactive problem-analysis aid for online homework assignments: Rather than give students an answer to a problem, the program asks a series of questions designed to prompt the student to think about the problem.
"It's increasingly important for students to integrate fully the use of computers in learning, as well as doing, chemistry."
EDUCATORS ARE also taking advantage of computers' unique ability to foster collaborations. PCOL (Physical Chemistry Online), a system developed by Zielinski, chemistry professor Marcy H. Towns at Ball State University in Muncie, Ind., and their colleagues, allows students to do chemistry projects online with students at other schools. "From a teaching perspective, that's one of the biggest advantages," Towns said.
Tackling subjects from hair dyes to the thermodynamics of bungee jumping, chemistry students from 22 institutions since 1996 have participated in the program, analyzing data and discussing problems with the help of faculty, who assume an online facilitator role.
"When you design online modules where collaboration is key, you've got to include tasks that encourage discussion," Towns said. Students need to analyze data or do a synthesis for which they have to compare data and draw conclusions, or do evaluations in which they look at data and use it to answer questions or solve a problem, he said. The questions must be hard enough that students need to talk about them with each other to solve them.
"Very seldom in industrial chemistry do you make a measurement and say 'Here's the result, that's all there is to it,' " Towns noted. For example, drug researchers will need to work with the Food & Drug Administration or other agencies. "There's a whole lot more to problem-solving than going and getting a measurement," she said.
"The projects also mimic what they will have to do as graduate students," Towns added. "Chemistry and chemical engineering don't work in a vacuum. You have to be able to communicate effectively."
Towns and her colleagues recently analyzed data from PCOL discussion boards and student surveys. The results, Towns said, highlight the important role of faculty facilitators, as well as show that the program is reaching students across the board. "We know we're creating modules for everybody, not just males or females or a particular ethnic group," Towns said. They're now working on a faculty handbook as well as a CD.
Today's undergraduates are also being introduced to molecular modeling. Michelle M. Francl, chemistry professor at Bryn Mawr College, in Pennsylvania, designed a new course--mathematical modeling of chemical phenomena--which fills a niche for the college's large contingent of math and chemistry majors, who have a strong interest in applied math. "We were trying to develop a course to serve them both," Francl said.
Francl starts her students out with a random-walk problem; they talk about the phenomenon and review different variations, such as self-avoiding or nonreversing random walks, Francl said. The students also look for examples of these models in published papers. "It's useful for them to know that an answer might be in another discipline's literature," Francl said. "Stuff is moving--does it matter whether it's molecules or elk?"
The students then write programs, attempting to reproduce published answers. Students with results closest to those in the literature win prizes: highly coveted items such as a beaker mug or trinkets brought home from ACS meetings. "They're as bad as second-graders" winning a prize in a contest, Francl joked.
Solving the problem in different ways also brings students face to face with the reality of real-world computer use: Some approaches might get better answers but might take much longer, and would therefore be much more expensive.
FOR GRADUATE STUDENTS gearing up to enter the computer- and bioscience-driven industrial world, bioinformatics programs are springing up across the country. Masayuki Shibata, biomedical informatics professor at the University of Medicine & Dentistry of New Jersey's School of Health Related Professions, in Newark, described his institution's program in biomedical informatics.
Computers' three-dimensional graphics capabilities also bring a much-needed perspective to the learning process. For example, potential energy surfaces are particularly difficult for students to visualize, so Alexander Grushow, chemistry professor at Rider University, in Lawrenceville, N.J., developed a program called PTRJ to make that task easier.
PTRJ, published in the Journal of Chemical Education [77, 1527 (2000)], began as a "really old, clunky program" used while Grushow was a graduate student at the University of Minnesota, he says. As computer technology improved, Grushow made the program increasingly sophisticated, adding a graphic interface.
With PTRJ, a student chooses a state for the H + H2 system, then calculates a classical reaction trajectory to end up with the final state of the system. An animation of three atoms at the bottom of the screen accompanies the program's plotting of the trajectory path.
"The student needs to answer questions such as, Does a reaction occur? What is the energy distribution among the product species? How did the initial conditions affect the outcome of this reaction?" Grushow said.
"This visualization, in direct conjunction with the plot of the trajectory path, allows students to quickly become expert at visualizing the motions of the atoms" on the potential energy surface, Grushow explained in his paper.
THE CHEMISTRY DEPARTMENT at Pacific University, in Forest Grove, Ore., has used Fujitsu's CAChe semiempirical modeling software for more than a decade. Chemistry professor James O. Currie and undergraduate student Crispin Wong created a laboratory workbook, "Teaching with CAChe," designed to help students with general chemistry.
Spurred on by the workbook, George D. Purvis III, vice president for the CAChe product line, and University of Florida chemistry professor Nigel G. J. Richards decided to design a teaching unit for proteins as well. "This is an area where software is undergoing a lot of changes, because the research taking place is related to the life sciences," Purvis said. "So it became obvious we needed to fill that gap in teaching exercises to deal with something related to biochemistry."
Students are presented with a problem: Design a better antituberculosis drug. Mycobacterium tuberculosis produces a purine nucleoside phosphorylase (PNP), inhibition of which could render the bacillus inactive. But drug designers need to avoid creating molecules that could also inhibit human PNP. With help from the program, students can visualize and analyze PNP and design docking ligands.
Indeed, computer teaching aids for more complicated molecules and chemical systems are the wave of the future, Foresman said. Eventually, computers will help students learn about and model systems that include hundreds of molecules in solution, or even solid phases. "You need that to understand catalysis and surfaces," he noted. "Those are the areas where you really want students to see the whole picture."
Chemical & Engineering News