Cyclic voltammetry

What a CV measures

In cyclic voltammetry the potential of the working electrode is swept in a straight line from a start potential to a switching potential and back, while the current is recorded. The rate of the sweep, the scan rate v, is usually between 1 mV/s and 1 V/s. Plotting current against potential gives the voltammogram.

A CV is often the first experiment on a new system because it shows quickly where a species is oxidised or reduced, whether the process can be reversed, and whether it is limited by diffusion, by electron transfer or by something adsorbed on the surface.

Reading the duck-shaped curve

Take a solution that contains only the reduced form R of a couple, with the potential starting well below the formal potential E⁰′. As the scan moves positive, R near the electrode starts to oxidise to O and the current rises. Close to E⁰′ the surface concentration of R drops quickly. Soon R cannot reach the electrode fast enough, because it has to diffuse from farther and farther away, so the current passes through a peak and then decays. On the way back, the O that has piled up near the surface is reduced, giving the cathodic peak.

The four numbers to read off are the peak potentials E_pa and E_pc, and the peak currents i_pa and i_pc. For the return peak, the current is measured from the extrapolated decay of the forward scan, not from zero, which is easy to get wrong.

The Randles–Sevcik equation

For a reversible couple with diffusion to a flat electrode, the peak current follows the Randles–Sevcik equation.

Worked example. For 1 mM ferrocyanide (C = 1 × 10⁻⁶ mol/cm³, D ≈ 6.5 × 10⁻⁶ cm²/s) at a 3 mm glassy carbon disk (A = 0.0707 cm²) and 100 mV/s, i_p ≈ 2.69 × 10⁵ × 0.0707 × 2.55 × 10⁻³ × 10⁻⁶ × 0.316 ≈ 15 µA.

Two things follow. The peak current is proportional to concentration, which is the basis of many sensors. And it grows with the square root of the scan rate. So a plot of i_p against √v should be a straight line through the origin, and its slope gives D if n, A and C are known. A slope of 0.5 on a log i_p against log v plot points to a diffusing species, and a slope of 1 to a species adsorbed on the surface, whose peak current grows in direct proportion to v.

A smaller version of the full simulator. Try doubling the concentration, then making the scan four times faster, and compare the peak with the Randles–Sevcik value.

Reversible, quasi-reversible and slow reactions

"Reversible" here means that electron transfer is fast enough to keep the surface at Nernst equilibrium all through the scan. Whether that holds depends on the standard rate constant k⁰ compared with how fast we scan. Matsuda and Ayabe expressed this with the dimensionless parameter Λ = k⁰ / √(D f v), where f = nF/RT.

The same couple can look reversible at 10 mV/s and quasi-reversible at 1 V/s. Nicholson (1965) used this to obtain k⁰ from how ΔE_p grows with scan rate, a method still widely used. In the simulator, choose Medium and run a scan-rate study to see it.

Charging current and resistance

Two effects that are not part of the redox reaction change every real CV.

• Charging current. The electrode–electrolyte interface behaves partly like a capacitor. Sweeping the potential charges it with a current i_c = C_dl·A·v, which offsets the whole curve into a box shape. It grows with v while the Faradaic peak grows only with √v, so at high scan rates the charging current takes over. Lesson 3 turns this effect into a way to estimate the electrochemically active surface area.
• Uncompensated resistance. The ohmic drop i·R_u means the surface sees less than the applied potential. Peaks move apart and flatten, which looks like slow kinetics. As a rough guide, the drop at the peak should stay below a few millivolts if you want to read kinetics from ΔE_p. Use a supporting electrolyte of at least 0.1 M, keep the reference close and use iR compensation where it is reliable.

A good CV in practice

1. Prepare the electrode. Polish glassy carbon with 0.05 µm alumina slurry on a polishing cloth, in a figure-eight motion for about 30 seconds, then rinse well. Sonicate briefly in water if alumina sticks to the surface.
2. Check it. Record a CV of 1 mM ferricyanide or ferrocyanide in 1 M KCl at 100 mV/s. A ΔE_p close to 60 to 70 mV suggests a clean, active surface. A much larger value suggests the surface needs another polish.
3. Remove oxygen by bubbling N₂ or Ar for 10 to 15 minutes when you work at negative potentials, because dissolved O₂ is reduced and adds a wave of its own.
4. Choose the window from the open-circuit potential, wide enough to see both peaks, but inside the region where the solvent and electrolyte are stable.
5. Record several cycles. The first cycle often differs from the rest. Report which cycle you show.
6. Write down everything that sets the curve shape: electrode material and area, electrolyte and concentration, scan rate, potential window, cycle number, reference and any iR compensation.

Check your understanding

References and further reading

1. N. Elgrishi, K. J. Rountree, B. D. McCarthy, E. S. Rountree, T. T. Eisenhart, J. L. Dempsey, A practical beginner's guide to cyclic voltammetry, J. Chem. Educ. 95 (2018) 197–206. doi:10.1021/acs.jchemed.7b00361
2. A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and Applications, 2nd ed., Wiley, 2001, chapter 6.
3. R. S. Nicholson, Theory and application of cyclic voltammetry for measurement of electrode reaction kinetics, Anal. Chem. 37 (1965) 1351–1355. doi:10.1021/ac60230a016
4. H. Matsuda, Y. Ayabe, Zur Theorie der Randles-Sevčikschen Kathodenstrahl-Polarographie, Z. Elektrochem. 59 (1955) 494–503.
5. Analytical Sciences Digital Library, Electrochemistry modules, LibreTexts Chemistry (electrode polishing and ferricyanide test).