The problems you have encountered may have resulted from:
- incomplete conversion of the hydrogensulfate
- sulfuric acid present in your impure (hydrogensulfate still present) pyrosulfate, which may lead to oxidation of the acetate on high temperatures
- not using an inert diluent, e.g. glacial acetic acid
Btw., heating of a hydrogensulfate with an acetate is a well
known, very old procedure for obtaining glacial acetic acid
Anyway, the problem seems to be to obtain the pyrosulfate dry and uncontaminated with sulfuric acid. After digging in the patent literature, I found the following:
Translated from Patent DE40696, entitled "Process for the synthesis of pyrosulfates of alkali metals as well as of ammonium":
....The first method of production
(the method of making pyrosulfates from the corresponding hydrogensulfates by strong heating) has the big disadvantage, that at the high temperature of brown glowing heat, some part of the already formed pyrosulfate is decomposed, producing sulfur trioxide.
....The inventor now has found out that hydrosulfates, especially these of the alkalis and ammonium, or equimolar mixtures of neutral sulfates with sulfuric acid
(very interesting) can be transformed under elimination of water into pyrosulfates at relatively low temperatures (260°-320°C)
without even forming a trace of sulfur trioxide when the heating is done under vacuum.
....The preparation goes as follows:
For the synthesis of sodium pyrosulfate, 240kg of sodium hydrogensulfate or a mixture of 142kg sodium sulfate with 98kg of sulfuric acid monohydrate in the form of 66°B sulfuric acid
(this corresponds to sulfuric acid of 96%), are brought into a retort made from cast iron, equipped with an agitator. The retort is fixed to a pump and is heated to 260°C while the apparatus is being evacuated. The mixture is finally heated to 300°-320°C within 5 to 8 hours, while the mass is agitated. From 260°-280°C (at a vacuum of 500-600mm Hg) the main part of the reaction formed water distills off, and the mixture is only heated to 320°C to make sure that the reaction has been completed. The remaining melt is cast into plates, it consists of pure sodium pyrosulfate.
In the same manner the pyrosulfates of other alkali metals as well as the pyrosulfate of ammonium can be produced. It is hereby suitable to install a condenser with receiver between retort and vacuum pump. Between condenser and receiver a glass tube should be installed. This enables the operator to see when the reaction has finished, because no more water will be observed in the glass tube then.
....(italics mine)