Why will paraformaldehyde not dissolve in unaltered seawater without added sodium hydroxide?
Paraformaldehyde is a white solid formed by combination of large numbers of formaldehyde molecules in an aqueous solution: a polymer. Formaldehyde, HCHO, is a gas and strictly speaking it doesn't exist in aqueous solution because it tacks on a water molecule to form methylene hydrate, which is HO-CH2-OH. This is the active ingredient of fixatives. Methylene hydrate molecules just love one another, and join together (eliminating H2O, so I suppose it's really the original formaldehyde carbon atoms that are so affectionate) to make polymers of all sizes. In commercial formalin (37-40% HCHO by weight) the polymer molecules are small enough to stay in solution. In paraformaldehyde they are big enough to be insoluble.
Manufacturers add some methanol to formalin. This retards the formation of large polymer molecules (see Recommended Reading if you want to know why). Probably the methanol doesn't affect fixative properties when diluted, though some people in the late 1950s claimed that it did. If you buy paraformaldehyde, you can depolymerize it yourself and get a solution of "formaldehyde" (actually methylene hydrate) that doesn't contain any methanol.
From what I've said so far, _Please Take Note!_ it follows that there is no such thing as a "2% (or any other %) paraformaldehyde" solution. Paraformaldehyde is a high polymer, and its molecules are too big to dissolve in water, alcohol or anything else.
You have to depolymerize paraformaldehyde to get it to "dissolve" and form a formaldehyde (really methylene hydrate) solution. The depolymerization is a reaction of the polymer with water: a hydrolysis. It needs hydroxide ions (OH-) as a catalyst, and also some heat to get the job done in reasonable time. In the making of ordinary phosphate-buffered formaldehyde from paraformaldehyde, the usual procedure is to heat the PF with the dibasic sodium phosphate component of the buffer. This contains enough OH- ions to catalyse the hydrolysis and depolymerization. You add the acidic part of the buffer (sodium or potassium dihydrogen phosphate) when the solution has become transparent. This occurs when the temperature reaches about 60 C. It should not be necessary to go any hotter than that.
In the earliest recommended fixatives that started with paraformaldehyde, a few drops of sodium hydroxide were added to a heated suspension of paraformaldehyde in water or saline. This hardly affected the pH of the final solution.
My supervisor (who has been trained in histology, unlike myself!!) said that in most of my staining and fixative methods can have the phosphate buffer component replaced by seawater with no problems as seawater is a buffer, at the right osmolarity for fish tissue. Is this the case?
I don't know how good a buffer sea water is, but it's unlikely to be as robust as 0.1M phosphate. In a fixative the osmolarity is more important than the pH, but for a slowly acting agent like formaldehyde or a slowly penetrating one like osmium tetroxide, the solvent should be as similar as possible to the extracellular fluids of whatever you're fixing. If the formaldehyde (takes hours to do its stuff) is mixed with more rapidly acting fixative agents (alcohol, mercuric chloride, picric acid etc., which act as soon as they reach the cells), the osmolarity is less important, and most such mixtures are acidic too. The formaldehyde does its cross-linking after the proteins have been insolubilized by the coagulant components.
For formaldehyde chemistry: Walker, JF 1964. Formaldehyde. 2nd ed. New York: Reinhold; London: Chapman & Hall.
For how formaldehyde works: Pearse, AGE: Histochemistry, Theoretical and Applied. Any edition of this book should be OK. There's also lots of erudite discussion in Baker, JR (1958) Principles of Biological Microtechnique. London: Methuen, which is a great classic in the field. For some stuff on the slowness of formaldehyde fixation and importance of an isotonic buffer: Paljarvi,L, Garcia,JH & Kalimo,H 1979. Histochem. J. 11, 267-269; Schook, P 1980. Acta morph. Neerl.-Scand. 18: 31-45. See also some of MA Hayat's books on techniques for electron microscopy, which discuss the subject thoroughly.
John A. Kiernan,
Department of Anatomy & Cell Biology,
The University of Western Ontario,
LONDON, Canada N6A 5C1