Microscale Techniques For The Organic Laboratory
I keep seeing people try to run standard organic procedures on a small scale and end up with maybe 10% yield because they didn't actually understand what they were doing differently. The equipment changes, but more importantly your approach to handling materials changes. You're not just using smaller stuff. You're fighting a different set of physical problems. The core idea is running reactions and purifications with quantities in the milligram range instead of the gram range. I've used scales down to about 50 milligrams consistently, sometimes lower if I have enough starting material and the reaction is reliable. The reason isn't just saving money on reagents. It's about handling things like toxic or explosive intermediates where you don't want large quantities sitting around. It's also faster for scouting new reactions before committing to a bigger scale. The setup is fairly standard. A typical bench will have micro centrifuge tubes, Craig tubes, Pasteur pipettes with the tips pulled slightly, small hot plates, and a decent analytical balance that reads to 0.1 milligrams at minimum. For reactions, I use 5-milliliter round-bottom flasks with magnetic stirring bars cut in half or specialty micro stir bars. Boiling chips are basically irrelevant at this scale. You rely on good stirring and sometimes a slight positive pressure of inert gas to manage bumping. Melting point apparatus works the same, just fill the capillary tube less. Capillaries for microscale MP are the same size as standard ones, but you pack them tighter because there's less sample.
Workup is where people fall apart. At gram scale you can dump your reaction into a separatory funnel and not think about it. At milligram scale, extraction with a micro centrifuge tube and a few hundred microliters of solvent is your only real option. The phase separation is slower. You need to be patient and let the layers actually separate instead of pipetting too early. I usually spin the tube briefly to help the phases come apart cleanly. Transferring between vessels means using a pipette or a microspatula, and you lose material with every transfer. That's just the math. If you do three transfers, expect to lose a meaningful fraction of your product. I ran a Grignard reaction last year on about 80 milligrams of aryl bromide and tried to isolate the product by standard aqueous workup. The yield was abysmal because the product was partitioning into the aqueous layer more than I expected. I couldn't see it properly through the centrifuge tube. The workaround was to acidify the aqueous layer, extract with a minimal amount of dichloromethane, then dry and concentrate in the same microtube instead of transferring to a separate flask. It saved maybe 30% of the product that would have been stuck to glass surfaces or left behind in a larger vessel. The lesson was simple: minimize transfers at this scale. Do everything in one container if you can. Purification on this scale is usually flash chromatography on a tiny silica gel column or preparative TLC. A standard 2-centimeter diameter column packed with about 3 grams of silica will handle 100 to 200 milligrams of crude material without much trouble. You elute with just a few milliliters of solvent and collect fractions in micro tubes. The problem is visualizing where your product is. At this scale, a UV lamp is basically mandatory. Spots are small and faint. Developing plates with iodine works too but it's less precise for quantifying where your compound actually is.
Recrystallization follows the same logic. You dissolve the crude solid in the minimum amount of hot solvent in a microtube or small flask, then cool it slowly. The yields are often higher than at larger scale because the solubility dynamics work more in your favor when you're controlling the volume precisely. But you can also overshoot and lose everything to the mother liquor if you use too much solvent. I keep a log of how much solvent each compound needs at this scale. It saves time on the next run. There are some hard limits to this approach. Scales below about 10 milligrams become extremely difficult to handle accurately. Weighing errors dominate. Transferring between vessels destroys your yield. Reaction monitoring by TLC becomes unreliable because the spot intensity is too low. At that point you're better off using flow chemistry or a different analytical approach altogether. Also, some reactions simply don't work well at low concentration because the kinetics change. Bimolecular reactions slow down, and side reactions that are negligible at higher concentration can become dominant when you dilute everything. Microscale NMR is another topic. Shigemi tubes and other micro NMR inserts exist but they require specialized equipment and the signal-to-noise ratio drops significantly. Most people just run a standard 5-millimeter tube with a microdrop of sample in deuterated solvent. It works fine if you have at least 10 to 20 milligrams of pure material.
Get the Full Details
The biggest practical advice I can give is to map out every transfer step before you start and plan to combine steps whenever possible. Each additional vessel interaction is a source of product loss. Use glassware with minimal surface area. Prefer vials over flasks for storage and intermediate steps. And calibrate your balance regularly because at this scale a 0.5-milligram error is a significant percentage of your starting material.
