I don't think that's a buffer problem. I'd recommend a non-linear fit and check the accuracy&error range at the concentrations of interest, and to dilute the samples if required.

Thanks Mr. Schechinger,

I want to get linearity up to 150 mg/L with single point calibration(linear fit).

Wolfgang Schechinger Finally, i got the 150 linearity with single point calibration. Thanks for your reply.