What is your sintering temperature? If the sintering temperature is higher than 950°C, you may do this: after polishing the sample, youd can make an annealing at arround 900°C for 5 h. Please find as attached file my publication about spark plasma sintering.
Is the surface quality is important for you or no? If not, you can make annealing directly without polishing and after take image with "SEM" for microstructural investigation.
Otherwise, for a perfect polishing you need a polishing machine and SiC papers. If the surface quality is not a priority, you may use only a SiC paper to remove graphite on the sample surface. If your sample is an oxide material, the dark color is due to the oxygen vacancies. For more information you can read my paper. To recover white color, you have to anneal your sample....but in your case, it will be complicated because the sintering time is very low (650°C). If you anneal your sample at 900°C/5h, the ceramic grains will abnormally grow. You can try to anneal at 600°C for long time (around 10h) to recover the oxygen composition.
Is it necessary to remove the carbon coating? If yes, you can partially remove the carbon coating by polishing in different diamond suspensions. After, you must make a annealing in air to remove the residual carbon. Non-oxide materials are burned out below 800ºC to avoid oxidation. It is very difficult to establish a procedure of thermal attack, seen that the microstructure has fine grain when is densified by SPS. In this case, you can make the microstructural characterization on the fracture surface.
You may also consider a chemical method for etching graphite, if (not a small if...) your substrate can resist such a strong treatment. While considering this approach you can check another discussion held at this forum: https://www.researchgate.net/post/Chemical_etchant_or_thermal_etching_condition_for_graphite
I use the graphite foil to protect the graphite die, due to the fact that the samples stick to the die after the SPS treatment. Does the spray prevent the sticking too Dariusz?.
Rebeca, yes, spray protect die and punches. Additionally samples have smooth surface and removing sample from die is very simple. But if you use graphite foil, best way to remove it from samples is sandblasting (e.g. sand or corundum) sample using a sandblasting cabin.