I haven't been able to find much on using bicarbonate as a restrainer. Geoffrey Crawley in the formula for FX-2 was very insistent on using potassium carbonate sesquinydrate instead of the anhydrous salt. His reasoning was that the sesquihydrate contained a small amount of bicarbonate that was necessary for the developer to work correctly. He used sodium bicarbonate directly in the formula for FX-55. The developing agent Edinol was described as being so sensitive to restrainers that sodium bicarbonate could be used instead of potassium bromide. Old carbonate developer formulas often included a small amount of potassium metabisulfite which formed bicarbonate ion in situ. Mason and Glafkides make no mention of it.