[ironnica]

Hi all

I was wondering if anyone has come across something that I have. The Randles-Sevcik equation shows that peak current is proportional to the square root of scan rate. I carried out analyses in ferrocyanide (one electron transfer) with KCl. This worked perfectly, giving linearity.

Then I tried this same study with the same setup but with a very small amount of the analyte (same concentration). Originally the electrodes were sputtered onto a chip and the experiment was carried out in bulk solution, now a channel has been applied across all the electrodes.

The current rose as expected from 10mV/s to 50mV/s. 75mV/s showed a sharp decline of current, and then 100, 150 and 200 continued to decrease.

Can anyone think why this may be the case. I would imagine if the diffusion layer was being hindered by the size of the channel, that a levelling off of current would be seen, but not a decrease. An image is attached of the CVs, the current (y axis) and the voltage (x).

Would really appreciate thoughts on it. Thank you

[Borek]

Perhaps I am limited by my English, but I don't understand the geometry of the system.

[ironnica]

Hi Borek, if you mean dimensions, the channel height around the electrodes is ~40um I think (electrode height insignificant to this), the width of channel is 150um. The electrode width is 110um. Also the length of the channel is sufficiently large that it shouldn't be a problem with the length of the channel.

If you have any thoughts at all as to what could cause the relationship due to the Randles-Sevcik equation (ip = 2.69 X 10^5 n^3/2 A D^1/2 C v^1/2 ) to invert (I.e. current became inversely proportional to scan rate) I'd be very grateful.

Thank you for your response.

[Borek]

First thing that comes to mind is that Randles-Ševčik equation holds for processes controlled only by diffusion, it won't work if there is a kinetic component. The theory behind was published by Delahay (JACS 75, 1190 (1953)) and Matsuda&Ayabe (Z. Elektrochem. 59 494 (1955)).

[ironnica]

Sorry if I sound remedial here. By kinetic component I assume you mean movement as in flow (mass transport other than diffusion). I would hope that there was no movement in-channel. There was certainly no flow.

[Borek]

No, by the kinetic component I mean speed of the electrode reaction, nothing related to the solution movement. Randles-Ševčik equation assumes reaction is fast enough to be assumed it is instant, it is not always the case. Note that it is relative - some reactions are fast enough to behave reversibly at slow scan rates, but not fast enough to behave this way when the scan rate goes up.

[ironnica]

Delahay et al:
"It should be added, however, that  if  there are
kinetic complications  a  curvature  in  the  i-v^2
diagram could possibly be expected."

I had assumed this. However, what I see might be an exaggeration of a curve.


Also sorry about that misunderstanding. The thing is that the reaction proceeds (proportionally) at all tested scan rates without the channel. When the channel is applied, this behaviour is seen.

[ironnica]

I appreciate your input Borek. I will omit it for the moment as I cannot retest it just yet

[Borek]

Also sorry about that misunderstanding. The thing is that the reaction proceeds (proportionally) at all tested scan rates without the channel. When the channel is applied, this behaviour is seen.

OK, what you wrote earlier suggested you have not tested your analyte without a channel.

If memory (and logic) serves me well, faster scan rates require smaller amounts of analyte, as the diffusion layer doesn't reach that deeply into the solution. That would mean channel should be not limiting diffusion.

I have some vague ideas about what may be happening, but they are a pure speculation. Last time I dealt with these things was around 1991.