Filtering Properly: Which Half Are You Trying to Keep?

**Gravity filtration** is a cone of paper in a funnel, and the liquid runs through under its own weight. Use it when the solid is unwanted — dust, insoluble impurity, a drying agent. **Flute the paper.** A fan-folded fluted filter has several times the surface area of a plain cone and runs far faster, because the liquid is not sitting against the glass. For anything but a trivial volume this is the difference between a minute and half an hour. **Wet the paper first** with the solvent being used, and seat it against the funnel. A dry paper floats off its seat and lets liquid past. **Do not fill above the paper.** Liquid poured over the top goes straight down the outside, and this is the commonest way an otherwise careful filtration is ruined. **Use a stirring rod** to direct the pour down onto the paper rather than splashing. **For a hot filtration**, warm the funnel first and keep everything moving. Any cooling crystallises product in the paper.

A **Buchner funnel** on a filter flask with suction pulls the liquid through and leaves a compact cake. Use it when the solid is what you want. **A flat paper cut to the plate**, not a cone. It must cover every hole and not turn up the sides — a paper that is too large curls and lets solid past its edge. **Wet the paper and apply suction before adding anything.** The paper seats down onto the plate and seals. **Pour with the suction running**, and rinse the last of the crystals out of the flask with a little of the mother liquor, not with fresh solvent — fresh solvent dissolves the crop. **Wash the cake with ice-cold solvent.** Release the vacuum, add just enough solvent to cover, stir the cake gently with a spatula so the wash reaches all of it, then re-apply suction. Pouring wash straight onto a cake under suction channels through one spot and washes nothing. **Pull air through to dry**, then press the cake with a clean stopper to squeeze the last liquid out. **Use a trap** between the flask and the pump. Without one, a pressure change pulls pump oil or water back into the product — an unpleasant and entirely avoidable loss.

Filter papers differ in **pore size** and in **speed**, and the two trade against each other directly. **Coarse and fast** — for bulky precipitates and for removing a drying agent. Fine particles go straight through. **Medium** — the general-purpose choice, and right for most crystallisations. **Fine and slow** — for fine precipitates that pass through anything coarser. It can take a very long time, and it is worth it only when the product actually requires it. **Ashless paper** is made for gravimetric work: it burns to almost nothing, so a precipitate can be ignited in the paper and weighed. Expensive, and pointless for anything else. **Sintered glass** replaces paper entirely where the paper would be attacked — strong acids, strong alkali — or where paper fibres would contaminate. It is reusable and needs cleaning immediately, because a dried cake in a sinter is very difficult to remove. **Every paper holds liquid**, and that liquid contains dissolved product. On a small preparation this is a real fraction of the yield, and it is the argument for using the smallest funnel the job allows.

Filtration is not always the right answer. **Decanting** — letting the solid settle and pouring the liquid off — costs nothing, loses nothing to a filter, and is the right first move whenever the solid settles well. Follow it with a filtration of the last portion if needed. **Washing by decantation** is better than washing on a filter for a fine solid: add solvent, stir, settle, pour off, repeat. The solid is properly mixed with each wash rather than channelled through. **Centrifuging** settles what will not settle by itself, and loses nothing to paper. It is the right answer for small volumes and for very fine or gelatinous solids that blind a filter instantly. **A gelatinous precipitate should not be filtered at all** if it can be avoided. It clogs the paper in seconds and then nothing moves. Digest it — stand it warm for a while — and the particles coarsen enough to filter. This genuinely works and it is the standard fix. **Do not filter a hot saturated solution unless you have to.** Every degree it cools in the funnel is product in the paper.