I have conducted the above test for Water, methanol and ethanol extracts of herbal plant leaves. MWCO of the dialysis membrane I have used was 12-14KD. The problem I face was plant color also diffused to the outer solution through the membrane. I am using a colorimetric method (DNS) to detect the final glucose concentration in the outer solution. Therefore, this plant color interferes with my results.

I can not use a control (plant+water inside the membrane) to avoid the effect of plant color, because that control has different concentration gradient with the outer solution.

Can anyone give me a solution for counteract this problem?