This is a hard one to balance. The trick or secret is that it makes HCl too. At least that's one way to balance it. This also doesn't necessarily accurately represent the true molecular nature of copper benzoate, which has a few different possible hydrates and potential forms and adducts. For what it's worth, I know from having done this before that the Blesser formula is basically the stoichiometric formula.
12 NH4ClO4 + Cu(CO2C6H5)2 → 6 N2 + CuCl2 + 24 H2O + 14 CO2 + 10 HCl
This is 82.2% AP and 17.8% Copper Benzoate.
A couple of questions:
Is it really correct to calculate with the formation of CuCl2 when we "know" the species emitting blue color is the excited CuCl molecule?
Do you think it would be successful in this case to balance the formulation to the production of CO instead of CO2?
I saw this was discussed on Science Madness by the user Turbosnigel in 2012 and the following reactions were also mentioned:
CuBz, water free, Burning to CO2
62 NH4ClO4 + 10 C14H10CuO4 = 148 H2O + 52 HCl + 140 CO + 31 N2 + 10 CuCl
AP 7284,38g + CuBz 3057,73g = 10342,11g
Composition uses 29,57% CuBz, and contains 6,14 mass% Cu.
CuBz, water free, Burning to CO
118 NH4ClO4 + 10 C14H10CuO4 = 232 H2O + 108 HCl + 140 CO2 + 59 N2 + 10 CuCl
AP 13863,82g + CuBz 3057,73g = 16921,55g
Composition uses 18,07% CuBz and containing 3,75 mass % Cu.
When I see the low amount of copper and the big excess of HCl in these formulations I wonder if the flame color could be even further improved by replacing some of the copper benzoate with metallic copper to increase the copper concentration in the flame?
Unfortunately copper powder is not stable together with ammonium perchlorate but I guess a little dichromate would fix that.