Michael O'Neill

View Original

Reflections on 2019 Teaching: Investigative Lab Practicals

I remember exactly three things from my undergraduate lab course. 

1.     That time the demonstrator explained to me how to work out an oxidation state from a Lewis structure (of a weird anion with some S-S bonding that I’d just made);

2.     That time I somehow blew up a bath of hot silicon oil;

3.     Having friends is not just nice or useful: it is necessary.

I try to keep this in mind when I run undergraduate labs now. Students are time-poor and will likely forget most of their time with me. Perhaps a tiny interaction with them (that oxidation state explanation) will stay with them forever, so I have to try and carve out those human moments. Perhaps something will go wrong (that oil bath explosion), so I have to be ready to model good and safe lab practice. Perhaps the lab will not thrill every student, but being with friends in a nurturing atmosphere will be part of a deeper story they are building.

In terms of memories per hour spent in the lab, though, my experience gave a pretty poor yield. When I compare it to my fourth-year project (where I pretty much learned everything), it starts to seem like an active waste of time.

So, in recent years I have tried to move away from ‘recipe’ synthetic inorganic practicals and towards investigative tasks. My hope is that a deeper focus on active learning will help students learn a couple of things rather than forget a lot of things.

In my experience of leading synthetic labs, there are two big constraints.

Safety

Investigating anything requires some freedom. In a research lab, it would be common practice to – say – use a suite of ligands to react with a selection of metals to make a range of products. Each decision point multiplies the number of materials which need to be risk-assessed. Without the decision making of an experienced researcher, the experimental matrix of various reaction conditions becomes infeasible to risk-assess. When designing a molecular lab practical for a hundred undergraduates, the number of decision points has to be aggressively minimised.

But there are other kinds of investigation where the material safety is not such a profound constraint. Rates and yields can give nice numbers using the same materials (‘how does concentration affect…?’), and qualitative analysis (‘what is the product…?’) can make the investigation of a reaction look a bit more authentically like molecular research.

Task and Marking

The instruction to the students is a really delicate thing to craft, and plays out in a complex way in context. Simple ‘tell me what you made’ analysis tasks are vulnerable to charges of being cheatable in the rota system (or even the year cohort system) most departments must use to operate.

Investigations in general can also have a huge hidden student workload. The literature searching can be really hard if you aren’t used to doing it. Presenting your findings can also be arduous if the guidelines are unclear or the format long-winded.

Finally, it’s important to think about what legitimate failure might look like. Most experiments don’t work, but most undergraduate practicals do. What sort of a mark should a valiant attempt attract? Will an early failure restrict students’ access to marks associated with some final artefact like a report? How is failure presented to students before, during, and after the experiment?

Gold Nanoparticles

My first investigative lab was a 3-stage practical on gold nanoparticles. Given the bottom-up gold citrate prep for a red nanoparticle, groups (gold chloride is expensive, and groupwork is important) are then challenged to make a blue one.

Students then analyse their samples using UV-Vis and DLS (which they do with a PhD student researching nanoparticles). Comparing the (apparently contradictory) data is the core focus of the problem sheet assessment.

Monodispersity is hard to achieve for the citrate reaction, so there is safety data available for a ‘single’ gold nanoparticle product. The assessment rests on interpretation of the actual data rather than the ‘right’ data, so it’s possible to mess up the synthesis and still score well.

The practical works not smoothly, but certainly effectively. Students seem to enjoy working together to solve the blue NP problem, and respond constructively to experiments going wrong. The analysis time with an expert PhD student in the research labs seems to be a real highlight of their time in the Inorganic lab.

Solar Cells

My investigative lab this year was based around solar cells. The practical is designed to align with the research projects students do the next semester.

Using a fabrication designed for outreach, students are challenged to make a (raspberry juice) dye-sensitised solar cell. Each student selects an individual research question from a prescribed list (e.g. how does dye loading affect cell voltage? How does in-series cell stacking affect voltage?). The group then produce a poster (having been given a template) and a short reflective piece of writing (not worth too many marks).

All the materials are low-risk (outreach!), and there are no significant chemical transformations through the practical. The poster is marked on clear criteria which require the group overall to present clear analysis of data, but which do not rely on any given project ‘working’. One group used dip time as a proxy for dye loading to wrangle a graph and had a really fun section of the poster devoted to the way that raspberry seeds could obstruct the sandwiching of the glass plates.

There was a frustrating disruption to the labs this year around asbestos safety, so we were forced to shorten the lab. This meant only a few students were able to attempt the solar cell investigation. Those who did produced some really enjoyable scientific posters; the practical is flawed enough to provide ample scope for self-criticism, but robust enough to produce one or two convincing results.

Conclusions

It’s satisfying teaching in this way, but planning a new investigative lab practical is exhausting. Traditional synthetic investigations are really hard to design safely, but the nanoscale seems to have meaningful scope for materials investigation.