I think it depends on the thickness and interaction happening with your film and the redox couple. Another possibility could be that while oxidizing it is generating some intermediates that are reducing and adding up with the oxidation (in case if you did it simultaneously one after another). You have to remember that if it the scan rate is different which is affecting. One question I have in my mind (if you have changed the scan rate and did the whole experiment on one/same film) after finishing up is their any dissociation happening onto the film? like thickness change? or else? any changes of the electrolyte color?

Thanks for you answer Md Saleh Akram Bhuiyan , i used the same electrodes and the same solution, i just had changed the scan rate, and no color was changed.

i am wonder because i did the same experiment on another screen printed electrode (working electrode fictionalized with carbon nano tube) and the behavior was liner (relation between (scan rate)^0.5 and delta(V) ) with 0.99 regression.

Well in that case I could say the system is not stable. Something is definitely reacting with the WE and that's why you see redox peak and the WE is also reacting with the buffer that's why the current level is changing or shifting. may be there are something else with the carbon as a material while you screen print that are reacting. Please have some study on the screen printed carbon electrode while you prepare them, if there is other material you incorporate with the precursor solution that might interact when you do that experiment. You can run the same experiment separate with fresh setup with different scan rate. And see if there is stiff those changes like increasing or shifting is appearing. also do some thickness test before and after the experiment to have idea. or do ATR-FTIR/AFM kind of things. They will let you know if something is changing after the experiment . Thanks

It is normal the shifting negative potential with increasing of the scan rate,

you see some examples in:

Article Impedimetric sensor for tyramine based on gold nanoparticle ...

Article Tyrosinase based amperometric biosensor for determination of...

Article Gold nanoparticle decorated multiwalled carbon nanotube modified electrodes for the electrochemical determination of theophylline

However, it is well-know, that Ferricyanide has a strong tendence of adsorb on electrode surface, try to apply a cleaning-step using CV between each cycle, maybe you will find a shifting less-acentuatedthan than you have obtained.

For each swept rate, you needs to wait until you get a steady voltammogram. Is that what you are showing?

It's indeed important to consider only steady state voltammograms but from the figure I assume this is the case. The negative shift is not to dramatic, but one explanation could involve the hydration ferrocyanide which may be a limiting factor at the screen printed electrode. Ferro(hex)cyanide strongly coordinates water (about 6 molecules). Consequently you can think of ferrocyanide as complex with water, when there is less water ferrocyanide becomes more "free" and the reduction potential shifts to more negative potentials, as you can verify by including water in the Nernst equation. This can be further demonstrated by using mixed solvents of varying composition (e.g. water/acetonitrile), the redox potential will then shift to more negative values as a function of increasing acetonitrile concentration. So some solvent residues (from the printing), or the hydrophobicity of the electrode can explain this slight negative shift.

Best regards, Mike