In this work, a design of experiments (DoE) was used to study the acetalization of glycerol with acetone using carbon dioxide as switchable catalyst. Experimental tests were carried out according to a fractional factorial design, with 2 levels and 4 factors (24-3), in combination with a response surface method (RSM), to study the significance of temperature, initial pressure of CO2, glycerol/acetone molar ratio and reaction time and then optimize the performance of the CO2 in this type of reaction. The response variable was the glycerol conversion that was analyzed by GC-MS. Temperature, reaction time and initial pressure of CO2 were statistically significant factors by Analysis of Variance (ANOVA). Molar ratio of glycerol/acetone was an insignificant factor in this reaction. An empirical model was derived, and experimentally validated, to predict the conversion of glycerol. The maximum conversions obtained by the model and experimentally were 60.36% and 60.60% at the optimal conditions defined as 118 oC, 42 bar and 5 h reaction time. It was done the ketalization with doped glycerin with several impurities (methanol, water and sodium chloride) to study its effects in this reaction. The addition of water and methanol decreases until 34% in glycerol conversion, while the addition of sodium chloride increases until 14% this variable. It was done reactions with glycerin from biodiesel production with NaOH and Ba(OH)2 and it was reached conversions of 6% and 20%, respectively.