Cyclic voltammetry (CV): how it works and how to read a voltammogram

A potentiostat controls the potential of the working electrode relative to a reference electrode and passes current through a counter electrode. In a CV experiment it scans that potential linearly from an initial value past the formal potential of the analyte to a switching potential, then reverses and scans back, usually to the starting value, while recording the current (Rafiee et al. 2024). If the solution contains the reduced form of a couple, the forward (anodic) scan oxidises it and the reverse (cathodic) scan reduces the product back (Rafiee et al. 2024).

The experiment is run in an unstirred, quiescent solution, so diffusion is the main way the analyte reaches the electrode. As the potential passes the formal potential the analyte near the surface is consumed, the diffusion layer grows and the current peaks and then decays, which gives the classic "duck" shape (Rafiee et al. 2024). The scan rate sets the time scale: the peak current rises with the square root of the scan rate for a diffusing species, as described by the Randles–Ševčík equation (Rafiee et al. 2024), which at 25 °C reads

iₚ = (2.69 × 10⁵) n³ᐟ² A D¹ᐟ² C v¹ᐟ²

with n the number of electrons, A the electrode area in cm², D the diffusion coefficient in cm² s⁻¹, C the concentration in mol cm⁻³ and v the scan rate in V s⁻¹ (Rafiee et al. 2024). For a species adsorbed on the electrode the current instead grows linearly with scan rate (Venton & Cao 2020). A one-way sweep without the return scan is linear sweep voltammetry (LSV), the standard way to record polarisation curves of electrocatalysts (Sharma et al. 2026).

Reference electrodes and the RHE scale. Aqueous work uses Ag/AgCl, Hg/HgO, Hg/Hg₂SO₄ or Hg/Hg₂Cl₂ references chosen by the pH of the electrolyte (Sharma et al. 2026). Because the thermodynamic potentials of the hydrogen and oxygen evolution reactions shift with pH, electrocatalysis papers convert every potential to the reversible hydrogen electrode (RHE) with the Nernst equation (Sharma et al. 2026):

E(RHE) = E(ref) + E°(ref) + 0.059 × pH

where 0.059 V is the Nernst factor at 298 K and E°(ref) is the standard potential of the reference, 0.0983 V for Hg/HgO in saturated 1 M NaOH (Sharma et al. 2026). In 1 M KOH (pH 14) a reading against Hg/HgO therefore moves up by about 0.924 V (calculated from 0.0983 + 0.059 × 14). On the RHE scale the OER has E° = 1.23 V and the HER 0.00 V, and the overpotential is the measured, iR-corrected potential minus 1.23 V (Sondermann et al. 2025). The free LabCanvas RHE calculator does this conversion for common reference electrodes and any pH.

• Peak potentials. A CV has an anodic and a cathodic peak potential; for well-behaved couples their midpoint (E½) coincides with the formal potential E°′ (Rafiee et al. 2024).
• Peak separation and reversibility. A reversible, diffusion-controlled couple shows a peak separation of about 59/n mV (Venton & Cao 2020). Wider separations mean slower electron transfer: dopamine at glassy carbon gives 70 mV at conventional scan rates, so it is not reversible, and at 400 V s⁻¹ the separation reaches about 750 mV (Venton & Cao 2020). A surface-bound (adsorbed) species ideally gives 0 mV separation and symmetric peaks (Venton & Cao 2020).
• Scan-rate test. Record CVs at several scan rates: current proportional to √v points to diffusion, current proportional to v to adsorption (Rafiee et al. 2024; Venton & Cao 2020). Dividing the current by √v lets CVs at different scan rates be overlaid, which shows mechanistic changes clearly (Rafiee et al. 2024).
• Coupled chemistry. If the electrogenerated species reacts (an EC mechanism) the return peak shrinks, so the reverse-to-forward peak-current ratio falls below one, and it rises again at faster scans (Rafiee et al. 2024).
• Catalysis. When a mediator or electrocatalyst is regenerated by a substrate (EC′), the forward current grows into a sigmoidal wave with a scan-rate-independent plateau whose height reflects the catalytic turnover rate, and the return peak disappears (Rafiee et al. 2024).
• Electrocatalyst CVs. Ni-based OER catalysts show a Ni²⁺/Ni³⁺ oxidation peak around 1.35–1.45 V vs RHE before oxygen evolution starts (Beglau et al. 2023), and the overpotential at 10 mA cm⁻² is read from the LSV (Sharma et al. 2026).
• Tafel plots. Plotting overpotential against log current density gives η = a + b log j; the slope b, in mV dec⁻¹, is the extra overpotential needed to raise the current tenfold (Beglau et al. 2023; Sharma et al. 2026). The LabCanvas Tafel slope calculator fits the linear region of your own η–log j data.

• Molecular solutions. Record molecular analytes in an unstirred solution so that mass transport is by diffusion (Rafiee et al. 2024).
• Remove oxygen. Purging the electrolyte with nitrogen for about 10 min before measuring removes dissolved O₂ (Beglau et al. 2023; Sondermann et al. 2025).
• Catalyst inks. A typical OER protocol drops 10 µL of catalyst ink onto a rotating disc electrode of 0.196 cm² (0.25 mg cm⁻²), rotates it at 1600 rpm and pre-cycles until the response is steady (Beglau et al. 2023); powders can also be coated onto carbon paper and activated by 10 cycles between 1.0 and 1.7 V vs RHE at 100 mV s⁻¹ (Sondermann et al. 2025).
• Loading and substrate. Report the mass loading: for NiO on carbon cloth the overpotential fell as loading rose from 0.2 to 0.6 mg cm⁻² (Sharma et al. 2026). The support matters too: nickel foam gives far higher currents than carbon cloth and adds its own Ni redox peak (Sharma et al. 2026).
• Uncompensated resistance. Measure the electrolyte resistance Rₛ by EIS, for example at open-circuit potential, and correct the potential for iR drop (Beglau et al. 2023; Sharma et al. 2026).

Redox thermodynamics. Redox potentials from CV are quick to measure and correlate with thermodynamic quantities such as substituent effects and catalyst reactivity, which is why they guide the design of mediators and catalysts (Rafiee et al. 2024).

Electrocatalysts for water splitting. LSV and CV give onset potentials, overpotentials at a set current density, Tafel slopes and stability over 1000 or more cycles (Sharma et al. 2026). Metal–organic frameworks are widely tested this way as OER (pre-)catalysts: a NiFe MOF/carbon nanotube composite reached 301 mV at 10 mA cm⁻² with a Tafel slope of 58 mV dec⁻¹ against 354 mV and 91 mV dec⁻¹ for RuO₂ (Sondermann et al. 2025), and Ni(Fe) MOFs transform during OER into Ni(OH)₂, NiOOH and FeOOH phases (Beglau et al. 2023). For the frameworks themselves see our pages on HKUST-1, ZIF-8 and UiO-66.

Active surface area and active sites. CVs in a potential window without faradaic reactions measure the double-layer capacitance: half the slope of Δj against scan rate gives Cdl, and dividing by a specific capacitance (0.04 mF cm⁻² in 1 M KOH, 0.035 mF cm⁻² in 1 M H₂SO₄) gives the ECSA (Sharma et al. 2026). For Pt, hydrogen underpotential deposition or CO stripping (420 µC cm⁻²) counts surface sites, and for transition-metal catalysts integrating a redox peak counts the metal sites that switch oxidation state (Sharma et al. 2026). Compare the ECSA with the gas-adsorption area from BET, and the surface oxidation states with XPS.

CV and electrochemical impedance spectroscopy are usually run in the same cell; both are part of our guide to characterisation techniques.

• Scanning too fast for activity metrics. Faster LSV scans add capacitive current and make catalysts look better: on carbon cloth the overpotential at 10 mA cm⁻² dropped from 796 mV at 1 mV s⁻¹ to 647 mV at 25 mV s⁻¹, an 18.71% error, so the authors recommend ≤2 mV s⁻¹ (Sharma et al. 2026). Charging current grows linearly with scan rate (Venton & Cao 2020).
• Trusting automatic iR compensation. Instrument-applied compensation did not match manual 100% correction, and on an electrode with Rₛ = 7 Ω full compensation distorted the curve, so only about 20% could be applied (Sharma et al. 2026). Report data with and without correction.
• Counting redox current as catalysis. A metal oxidation peak on the forward scan inflates OER current; the reverse scan gives activity free of it (Sharma et al. 2026).
• Over-reading Tafel slopes. Slopes from LSV or CV include double-layer and metal-redox current, different methods gave 59 to 82 mV dec⁻¹ for the same nickel foam, and steady-state chronoamperometry or chronopotentiometry is preferred (Sharma et al. 2026). The value also depends on the fitted range, and a Tafel slope alone does not prove one catalyst better than another (Sondermann et al. 2025).
• Mixing potential scales. Quote the reference electrode and convert to RHE with the electrolyte pH before comparing with other work (Sharma et al. 2026).
• ECSA window. Cdl depends on which non-faradaic window is chosen, and Ni-based hydroxides are not conductive below the Ni²⁺ oxidation, so impedance at OER potentials can be more reliable (Sharma et al. 2026).
• Stability from CV alone. 1000 CV cycles show on–off stability, but chronoamperometry or chronopotentiometry is preferred for claiming practical durability (Sharma et al. 2026).

The LabCanvas figure editor Library has the potentiostat and three-electrode cell as an editable object, including a cutaway view of the cell that shows the working, reference and counter electrodes and the current path, plus an editable data plot for the voltammogram. Draw the forward and reverse scans with arrows, label the anodic and cathodic peaks, E½ and ΔEp, and state the reference electrode or RHE on the potential axis. Overlay CVs at several scan rates or before and after cycling, then export the panel for a methods figure, SI or graphical abstract.