I gave it a thought about your question again....its not likely to be prooxidant because if you are calculating the % of scavenging of DPPH (where DPPH is 100%) it is impossible to see the prooxidans...that means that is more DPPH than you initially added (IMPOSSIBLE). Most probably the tested samples absorb at the same wave length as DPPH.
In fact, the prooxidant effect would be impossible to observe (this effect can only be observed in diferente assays, such as TBARS), unless DPPH radicals were being generated spontaneously in the solution... You should measure the absorbance of the tested solutions diluted in the same ratio with the solvent used to prepare the DPPH solution (e.g., metanol); i.e., if you prepare the reaction with 30 microliters of the tested solution and 270 microliters of DPPH solution, you must measure the absorption of mixtures containing 30 microlitters of the tested solution diluted with 270 microliters of metanol (if this was the solvent used). The values obtained in the control samples should then be subtracted to the absorbances obtained in the tested samples.
If you use UV-Vis spectrophotometer You need to mix and to measure the decrease of DPPH absorbsion exactly with a timer. The correct measurements can be obtained with Elisa reader.