Michael’s blog post on laboratory practicals (and a long day of manuscript revisions) got me thinking about laboratory practicals. I particularly liked the distinction between ‘science’ and ‘sciencing’ that he highlighted, the difference between theoretical science and the practice of science, or between lectures and labs. In as much as I like any binary system, I like this to a point. One of the things that irritate me about laboratory practicals is the piling on of stuff. The experiment needs to teach laboratory technique, have a real world context to be ‘interesting’, have a data analysis component, a write up component, a laboratory diary component, develop awareness of safety, and relate to theory in a lecture last week. And all of this in a constrained period of time dictated by timetabling. It’s like the most hideous episode of Master Chef known.
Would you really train a chef by saying ‘in three hours I want you to serve the only chocolate mousse you’ll make this year to three restaurant critics and by the way your future employment might depend on it’? Why do we train chemists by saying ‘in three hours I want you to do the only column chromatography this year and I will assess you heavily on it’? It is perfectly legitimate for trainee chefs to spend hours practicing knife skills before deploying them in a high stakes environment like making French Onion Soup.
In my ideal laboratory course, there would be a distinction between skills and ‘capstone experiments’ that draw in theory (and relate to science from the lectures) and are assessed.
In the image above, I have mapped out many of the techniques needed for a student to successfully complete a synthetic inorganic experiment, preparing and characterising some cobalt complexes. In my ideal scenario, students would have 5 weeks to achieve this. In weeks 4 and 5 they would work on the capstone experiment: the cobalt complex synthesis. In the first 3 weeks students would reflect on their laboratory skills and practice the specific techniques they feel they need to develop. Ideally, through a full laboratory course, these techniques would be split between ‘new’ and ‘revision’.
For example. Student Q has just completed the ‘synthesis of an organic reaction product’ capstone experiment and over 5 weeks mastered stirring a reaction at room temperature, handling conc. acid, characterisation by IR and NMR, weighing and measuring and simple recrystallisation. They did pretty well on the capstone experiment and feel pretty confident in the skills so far. They make a plan to develop the new skills:
Week 1 – Preparation of standard solutions, operation of UV-Vis Spectrometer, UV-Vis analysis and Beer-Lambert Plots. They do this using copper sulfate. They also use one solution to perform conductivity. The Beer-Lambert Plot seems straightforward and was recently covered in lectures so they take that away to do.
Week 2 – Complex recrystallisation which includes stirring at temperature, vacuum filtration, and filtering when hot. They do this using aspirin (or other simple organic compound) with some sand in it.
Week 3 – Use of Gouy balance, measurement and calculation of magnetic moment. They do this using the iron standard and a sample compound. They allow more time to work through the calculations in lab class. They get their Beer-Lambert plot for week 1 checked, and use the last part of the lab to fill out the safety forms and read the capstone background. They identify the types of literature they need to find to compare their results to when they get them and start a literature search.
Week 4 and 5- they synthesis a cobalt (II) complex and characterise it. They produce a laboratory report that contains an introduction (relationship to background theory demonstrated), experimental, results and discussion (Beer-Lambert plot, conductivity, magnetic moment, relate values to literature), conclusion and references.
The advantages of this approach would be sufficient time to repeat techniques if they are done wrong. The chemicals used for the ‘training’ phase are cheap and low hazard thus reducing the cost of the laboratory session without compromising the goal of the session. There are also significant gains in safety. It also allows a focus on developing experimental skills without being unduly bogged down in the context of an experiment until they are more confident in using their skills to produce a result. This could be merged with an approach that involves students producing videos of techniques or other methods of recording their training such as badges (for example: http://michaelseery.com/home/index.php/2017/01/peer-review-and-digital-badge-project/)
With apologies for the abysmal use of pictures containing text. These two images show 5 capstone experiments being carried out over 24 weeks with a gradual decrease in training time as the students become more skilled, and a shift from lab report writing in class (semester 1) to doing it in own time. The lab report writing weeks in semester 1 would permit sufficient time to introduce students to style and form (perhaps getting them to bring a draft to the session then work on it – I always feel support is better at the editing and improving stage of writing lab reports rather than before they’ve had a go). Yes I suggest that 5 full lab reports are written but I think students should produce lab reports across the various major sub-disciplines of chemistry to get the full range of experience of presenting experimental data.
The final experiment in semester 2 involves purification by column chromatography and using NMR to identify products that are not know to the student. This idea comes from my general frustration at the ‘I followed the recipe so obviously it is the right stuff’ thinking, and also the reluctance to be critical of data obtained. I’m being more transparent than I’m sometimes minded to be – occasionally I feel students should be given a prep that on face value produces product X but in reality produces product Y and they have sufficiently different NMR data that they should be able to figure it out. Of course, this is getting on for a trick question and you do still have to ensure your experiments stand up to the inevitable Google trawl for the answers (I know they do this because I’ve seen the hits this blog gets for this post when we run the experiment: http://wp.me/p3BLsC-6s)