Thank you Shaban Ahmed Ali Abdel-Raheem Sir thats big help

Are you using the GC? If it is, you may have matrix effect. If you use LC/MS, you should see matrix suppression. If you use standard in solvent, you will see low recovery on the LC/MS and high recovery in the GC method.

You can also use Solid Phase extractor (SPE)

 I am using UPLC, the recoveries are very high for pesticides. 

How you construct the calibration curve, in solvent or in matrix?  Easy check if it is matrix related. Prepare matrix (with no pesticide) to the final stage after cleanup and concentration or whatever, then spike small volume (high concentration) of pesticide and inject to LC/MS. Do the same with solvent. If you have no matrix issue (enhancement or suppression), then the response "must" be the same because you spike at the same amount. This will rule out the matrix related issue and thne you can  focus on something else. If you have matrix related (enhancement in this case), then you know that it is not your extract or cleanup or your protocol. Most of the time you should see matrix suppression in LC/MS. However, I have analyzed diquat in matrix, and see enhancement. You never know what you are gonna get. It is like a box of chocolate.

Thanks a ton sir I will work on your advice and will discuss it with you.

Dear,

what is your recovery%,?

if it in range 70-120% you can accepted but if it high

er than recovery range u must review the blank sample and calbriation curve and. Culcuting CV

Sir its more than 200% and I adopted QuEChERS tech. for sample cleanup and preparation. My matrix is agricultural Soil. Why do we need to clean the soil samples with PSA as given in many papers I am not getting it but there's higher recovery rate problem arise with adsorbents florisil and silica. I think PSA is best suited for biological material. I have lack of funds therefore cannot afford PSA right now and my work is suffering so i have to develop some way out with cleanup problem. Sir kindly suggest.

I would assume you are using GC with some kind of detectors (ECD, NPD, or MS) and that is why you have high recovery when you quantify the peak using standard in solvent. Do you use standard in solvent to build the calibration curve? If it is the case, you need to use a matrix-matched standard. Make blank soil extract and add standard to make a curve or one-point calibration and see how it goes. Easy check is to make standard in both solvent and matrix blank to see the responses. Ideally, you should have the same peak height. However, in reality, you will see response of standard in matrix higher than in solvent. The analyte breaks down easily in the hot injector port you you don't have matrix to protect.