|There is a conventional wisdom that automatic devices cannot be used in teaching because there is nothing left to teach once the automation is done. Little could be further from the truth.
In our Instrumental Analysis class, Chemistry 382 and 378 lab, we use a Zymark laboratory robot to learn "analytical methods development". The robot has an ensemble of general purpose sectors on its table. These sectors perform the routine sample manipulation steps needed for analytical spectrophotometry and HPLC. The key to making a robot into a learning device is to design experiments where the objective is to program the robot for routine use in the analytical chemistry course, not just use a robot already programmed by someone else to do an analysis. In this way, what is learned is "analytical methods development". The manipulative steps can be as simple as pipetting. The creative part is linking the manipulative steps together into an analytical procedure.
| Why use a robot to teach something as intuitive and as exploratory as analytical methods development? The answer lies in what has to be known, or learned, in order to make any automation work. To automate chemistry, you first have to know and understand the chemistry you want to automate. To automate a procedure, you have to know and understand not only the steps of the procedure, but their relative criticality. To automate a complete method, you have to know and understand how the steps of the procedure interact when they are linked together.
As if this were not reason enough, consider also how important it is for a novice to understand that any complete analytical method is seldom complete from the kinetic viewpoint of the individual chemical reactions involved in it. Or, in other words, seldom any particular reaction that makes up a procedure allowed to go to 5 or 6 half-lives before another reaction intervenes and alters its progress. Thus, attempts are made to standardize everything, so that the samples, the unknowns, and the standards all experience the same degree of incompleteness in their individual chemical steps. A skilled analyst can do this. Students seldom can. It is a given for a robot. Thus, we expect a robot to be significantly precise in its methodology. Non-reproducibility then can be traced to blunders.
If the standards are good, then this high precision can reverberate into high accuracy.
| Undoubtedly the most subtle of all aspects of analytical methods development is that of "matrix effect". This is that source of determinate error that kills methods dead if they are developed with water standards and then asked to calibrate a sensor for natural mixtures with complex solvent matrices.
A robot cannot compensate for matrix effects. No analytical method as such can do that. But one that comes close is that of the "standard additions". This method uses the sample itself as the host material for the standards, and back-calculations are done from the resulting working curve to determine how much analyte must have been present originally to offset the curve from a zero blank as much as has been observed.
A robot will make up the many standard solutions needed for this method with relative ease, and, surely, with no more work for the student than any other method. This means that students learn from the very beginning that matrix effects are the rule, not the exception.
| All of these factors combine with the inherent excitement of operating a machine with hands and an arm that moves around and does human-like things under the sommand of a human-like language. This motivation is so strong that even those students who "hate computers" and initially expect analytical chemistry to be "dry as dust and heavy as a Russian dinner" are motivated to learn more of it so that they can be full-fledged participants in activity of "analytical methods devlopment".|
Want to learn more? Conyinue onto the next button and look at some of the other features of Mike Roe Chips, the St. Olaf, Zymark, Zymate II laboratory robot.