Stripping Voltammetry

A drop of seawater contains lead at a concentration near 20 parts per trillion. That is roughly one lead atom for every 30 billion water molecules. Stripping voltammetry can still measure it with a relative standard deviation under 5%. This is the central trick of the technique. Instead of asking a tiny instantaneous current to report on a tiny concentration, you do something different. First you spend several minutes physically concentrating the analyte a thousandfold or more onto (or into) a microelectrode. Only then do you interrogate it electrochemically. The preconcentration step turns an impossibly small signal into a comfortably large one. The interrogation step — a controlled potential or current sweep that "strips" the accumulated metal back out — delivers the quantitative payoff. Stripping voltammetry matters because it occupies a peculiar niche. Atomic spectroscopy cannot easily fill this niche. It offers sub-nanomolar detection limits and speciation information. It distinguishes free metal ion from many complexed forms. Its instrumentation is cheap and portable enough for a research vessel. It also suits a well-head or a battlefield water-testing kit. Where inductively coupled plasma mass spectrometry needs a shore-based lab, a hanging mercury drop electrode and a potentiostat fit in a case. Where atomic absorption reports total metal, stripping voltammetry works differently when done carefully. It reports the electrochemically labile fraction instead. This fraction is often the toxicologically or biologically relevant one. The big picture is a two-act play. It is built on ideas from Polarography and Cyclic Voltammetry. Act One is exhaustive or near-exhaustive electrolytic deposition of trace analyte. This deposition occurs onto a small-volume electrode under strong convection (stirring or rotation). Act Two is a quiescent potential or current scan. This scan reverses the deposition reaction. It produces a peak whose height or area is proportional to the amount deposited. This amount is therefore proportional to bulk concentration. Anodic stripping voltammetry (ASV) deposits metals cathodically as amalgams or metallic films. It then strips them anodically. Cathodic stripping voltammetry (CSV) does the mirror-image chemistry for anions and metal complexes. These complexes form insoluble mercury salts. CSV deposits anodically and strips cathodically. Adsorptive stripping voltammetry (AdSV) extends the idea further. It applies to species preconcentrated by surface adsorption of a ligand complex, rather than by electron transfer. This article builds the quantitative machinery from first principles. It covers the deposition mass balance that sets the calibration slope. It covers the stripping peak equations for both linear-sweep and differential-pulse readout. It also covers the practical chemistry that separates a textbook peak from a real one. This includes mercury film versus hanging drop, and standard addition versus external calibration. It also includes the notorious interferences from surfactants and intermetallics. Three fully worked problems carry realistic numbers through several scenarios. These run from a lead-in-drinking-water assay, through a shipboard trace-metal survey, to a competition-level multi-metal square-wave ASV calculation.

Key takeaways

Stripping voltammetry preconcentrates analyte onto/into a microelectrode (deposition step), then strips it back out with a potential or current scan (stripping step), converting a picomolar bulk concentration into a measurable peak current.
Anodic stripping voltammetry (ASV): cathodic deposition of metal ions as amalgam or metal film, anodic stripping — used for Cd²⁺, Pb²⁺, Cu²⁺, Zn²⁺, Tl⁺, Bi³⁺.
Cathodic stripping voltammetry (CSV): anodic deposition as an insoluble mercury salt, cathodic stripping — used for halides, sulfide, selenide, and thiol-containing species.
The stripping peak current is proportional to deposited mass, which is proportional to deposition time × bulk concentration × mass-transfer coefficient (Faraday's law applied to a finite reservoir).
Detection limits reach 10⁻¹⁰–10⁻¹² M because deposition times of minutes can concentrate analyte by factors of 100–1000 relative to solution.
Key idea: stripping voltammetry trades time (minutes of controlled electrolysis) for sensitivity, turning a diffusion-limited trace signal into an amplified, quantifiable stripping peak.

Where this lives today

Stripping voltammetry remains the workhorse ultratrace method for on-site environmental monitoring: portable ASV/CSV instruments deployed from research vessels measure dissolved copper, lead, cadmium, and zinc in seawater and estuarine transects at parts-per-trillion levels, feeding data directly into GEOTRACES-style ocean biogeochemistry programs without the delay of shipping samples to a shore laboratory. In drinking-water and industrial-effluent compliance testing, disposable screen-printed bismuth-film electrodes give regulators and utilities a field-deployable, mercury-free method for lead and cadmium that meets EPA and WHO action-level sensitivity at a fraction of ICP-MS instrument cost. In clinical and biomedical chemistry, adsorptive cathodic stripping voltammetry detects trace platinum-based chemotherapy drugs and their DNA adducts, and CSV protocols quantify selenium status in blood serum for nutritional and toxicological studies. In semiconductor and electroplating industries, ASV is used for rapid quality-control assay of trace heavy-metal contamination in ultrapure process water, where a single stray part-per-billion of copper can ruin a wafer batch. Modern square-wave and adsorptive stripping protocols, often paired with chemically modified electrodes bearing selective ligands, continue to push the technique toward speciation-resolved, multi-metal, single-run analyses that complement — rather than compete with — mass-spectrometric total-metal methods.

Ask the tutor: How would the calibration slope and detection limit change if I switched from a mercury film electrode to a bismuth film electrode of the same thickness and area?

The physical picture

Consider a hanging mercury drop electrode (HMDE) of radius $r_0 \approx 0.5$ mm immersed in a stirred solution containing lead(II) at $c^* = 5 \times 10^{-9}$ M. If you simply ran a linear-sweep voltammogram on this solution the way you would on a $10^{-3}$ M solution, the peak current — set by semi-infinite linear diffusion during the few seconds of the scan — would be buried in background charging current. The fix is to hold the electrode at a potential well negative of the Pb²⁺/Pb(Hg) formal potential for several minutes while stirring vigorously. Convection continuously resupplies the electrode's diffusion layer with fresh Pb²⁺, which is reduced and dissolves into the mercury as an amalgam, Pb(Hg). Because the mercury drop is small (volume perhaps $\sim 5 \times 10^{-4}$ cm³) and deposition proceeds for, say, $t_d = 300$ s, the metal accumulates to a concentration inside the drop many orders of magnitude above its concentration in the bulk solution outside. This is preconcentration: partitioning analyte from a large, dilute external phase into a small internal phase. After deposition, stirring stops, the solution is allowed 15–30 s to quiesce (removing residual convection so that diffusion again controls mass transport, as in a normal voltammetric scan), and the potential is swept anodically. When the potential crosses the standard potential for Pb(Hg) → Pb²⁺ + 2e⁻ + Hg, all the amalgamated lead is oxidized back out of the drop in a burst lasting only a few seconds — producing a sharp current peak whose area (charge) equals the total deposited charge, and whose height is proportional to the deposited mass and hence, through the deposition mass balance, to bulk concentration. The mercury drop or film acts as a "concentrator" exactly as a rotary evaporator concentrates a dilute organic solution before an NMR experiment.

Figure

The two-step stripping voltammetry waveform: deposition at fixed potential under stirring, followed by a quiescent anodic (or cathodic) scan producing the stripping peak

The choice of electrode determines whether the mercury acts as a thin film or a hanging drop. A mercury film electrode (MFE), typically an in-situ or ex-situ film of thickness $10$ – $100$ nm plated onto glassy carbon, offers much higher surface-area-to-volume ratio than a millimeter-sized HMDE. Because the amalgam concentration inside a thin film builds up faster for the same deposition time (Faraday's law divided by a much smaller volume), MFEs give sharper, more sensitive stripping peaks — at the price of reduced reproducibility and a greater tendency toward intermetallic compound formation (Cu-Zn, Cu-Ga) that distorts multi-element peaks. Modern practice increasingly replaces mercury altogether with bismuth-film or antimony-film electrodes, which show similarly favorable, wide, mercury-free stripping behavior with far less environmental and disposal concern.

Figure

Anodic vs. cathodic stripping: direction of deposition and stripping current, and the chemical species involved

The core ideas

1. The deposition mass balance. Under strong forced convection, mass transport to the electrode is governed by a mass-transfer coefficient $m$ (units cm s⁻¹) that plays the role diffusion coefficient/thickness plays in quiescent voltammetry. For an electrode of area $A$ immersed for a deposition time $t_d$ in a stirred solution of bulk concentration $c^*$ , assuming the electrode surface concentration is held at essentially zero (deposition potential well past the wave, i.e., mass-transport-limited deposition) and that depletion of the bulk solution is negligible over $t_d$ :

N_{\text{dep}} = m A c^* t_d

where:

$N_{\text{dep}}$ = moles of metal deposited (mol)
$m$ = mass-transfer coefficient (cm s⁻¹), set by stirring rate/rotation speed, roughly $10^{-3}$ – $10^{-2}$ cm s⁻¹ for typical stirring
$A$ = electrode area (cm²)
$c^*$ = bulk analyte concentration (mol cm⁻³)
$t_d$ = deposition time (s)

Reading: deposited moles grow linearly with both concentration and deposition time — this linearity is what makes the calibration curve of stripping peak vs. concentration a straight line through the origin, and it is why doubling $t_d$ (within the regime where bulk depletion is negligible) doubles sensitivity.

2. The linear-sweep stripping peak current. Once deposition ends and the potential is swept at rate $v$ (V s⁻¹) starting from a potential well negative (ASV) of $E^{0'}$ , the amalgam or film oxidizes in a peak whose height, for a thin-film electrode of thickness $l$ (small enough that all deposited metal reaches the surface well before the peak, the "thin-layer" limit), is given by the Reinmuth/thin-film stripping equation:

i_p = \frac{n^2 F^2 A l C_{\text{Hg}} v}{2.7\,RT}

where:

$i_p$ = stripping peak current (A)
$n$ = number of electrons transferred per metal atom
$F$ = Faraday constant, 96 485 C mol⁻¹
$A$ = electrode area (cm²)
$l$ = film thickness (cm)
$C_{\text{Hg}}$ = amalgam concentration built up during deposition (mol cm⁻³), $= N_{\text{dep}}/(A l)$
$v$ = stripping scan rate (V s⁻¹)
$R,T$ = gas constant, absolute temperature

Reading: unlike a diffusion-controlled linear-sweep peak (which scales as $v^{1/2}$ ), the thin-film stripping peak scales linearly with scan rate, because all the material is already inside the film and its release is now a first-order-like desorption/oxidation process rather than a diffusion-limited one. This distinction is diagnostic: plotting $i_p$ vs. $v$ (thin film) or $i_p$ vs. $v^{1/2}$ (thick-drop/semi-infinite diffusion, as at an HMDE) tells you which regime you are in.

Derivation ADeposition mass balance and the linear ASV calibration equation

Goal. Derive the proportionality between stripping peak current and bulk metal concentration for a mercury film electrode, starting from Faraday's law applied to the deposition step and the thin-film stripping current equation, and show explicitly how deposition time, area, and film thickness enter the calibration slope.

Step 1 — Faraday's law for the deposition step. During deposition, the metal ion is reduced at the mass-transport-limited rate. The flux to the electrode under strong stirring is $J = m\,c^*$ (mol cm⁻² s⁻¹), so the deposition current is

i_{\text{dep}} = n F A m c^*.

Because this current is (to good approximation) constant over the deposition period — bulk concentration is not depreciably changed by depositing onto a microelectrode from a macroscopic reservoir — the total charge passed is

Q_{\text{dep}} = i_{\text{dep}}\,t_d = n F A m c^* t_d.

Step 2 — Convert charge to moles deposited, then to film concentration. By Faraday's law, moles deposited are $N_{\text{dep}} = Q_{\text{dep}}/(nF)$ , so

N_{\text{dep}} = A m c^* t_d,

recovering the mass-balance relation above (the $n$ cancels because both deposition and the charge-to-moles conversion carry the same $n$ ). If this metal is uniformly distributed through a film of thickness $l$ , the resulting amalgam concentration is

C_{\text{Hg}} = \frac{N_{\text{dep}}}{A l} = \frac{m c^* t_d}{l}.

Step 3 — Substitute into the thin-film stripping peak equation. Starting from

i_p = \frac{n^2 F^2 A l C_{\text{Hg}} v}{2.7\,RT},

substitute $C_{\text{Hg}}$ from Step 2:

i_p = \frac{n^2F^2 A l v}{2.7\,RT}\cdot\frac{m c^* t_d}{l} = \frac{n^2F^2 A m v\,t_d}{2.7\,RT}\,c^*.

Step 4 — Identify the calibration slope. The film thickness $l$ cancels entirely — a pleasing and non-obvious result, since it means the calibration sensitivity of a thin mercury film electrode does not depend on how thick the film happens to be (only on area, stirring efficiency, deposition time, and scan rate), which is why MFE stripping analysis can tolerate some run-to-run variability in film thickness without ruining quantitation as long as the film stays thin enough to satisfy the thin-layer assumption. Collecting constants,

i_p = S\, c^*, \qquad S \equiv \frac{n^2F^2 A m v\,t_d}{2.7\,RT},

a linear calibration with slope $S$ proportional to $t_d$ , $v$ , $A$ , and $m$ , and to $n^2$ — which is why bismuth and mercury methods for $n=2$ metals (Pb, Cd) are markedly more sensitive per mole than $n=1$ analytes (Tl⁺, Ag⁺) at fixed instrumental settings.

Go deeper · Derivation B · 1 minWhy stripping beats direct polarography: the preconcentration amplification factor

Derivation BWhy stripping beats direct polarography: the preconcentration amplification factor

Goal. Quantify how much stripping voltammetry amplifies the effective concentration "seen" by the stripping scan relative to the bulk solution, and show the limiting behavior as deposition time grows.

Step 1 — Compare deposited amalgam concentration to bulk concentration. From Step 2 of the derivation above,

C_{\text{Hg}} = \frac{m\,t_d}{l}\,c^*.

Define the preconcentration (amplification) factor as the ratio of the concentration effectively probed by the stripping step to the original bulk concentration:

\mathcal{A} \equiv \frac{C_{\text{Hg}}}{c^*} = \frac{m\,t_d}{l}.

Step 2 — Insert realistic numbers. Take a mercury film with $l = 20$ nm $= 2\times10^{-6}$ cm, a mass-transfer coefficient under moderate stirring $m = 2\times10^{-3}$ cm s⁻¹, and a deposition time $t_d = 180$ s. Then

\mathcal{A} = \frac{(2\times10^{-3}\ \text{cm s}^{-1})(180\ \text{s})}{2\times10^{-6}\ \text{cm}} = \frac{0.36\ \text{cm}}{2\times10^{-6}\ \text{cm}} = 1.8\times10^{5}.

The mercury film's internal metal concentration is nearly 200,000 times the bulk solution concentration — this is the whole reason stripping methods reach picomolar detection limits with instrumentation whose intrinsic current resolution is no better than that used for millimolar polarography.

Step 3 — Limiting case: bulk depletion. The mass-balance step (Step 1 of the calibration derivation) assumed $c^*$ stays constant during deposition — valid only while the moles removed are negligible compared to moles present in the sample volume $V$ . Formally this requires

N_{\text{dep}} \ll c^* V \quad\Longleftrightarrow\quad m A t_d \ll V.

As $t_d \to \infty$ at fixed $V$ , this inequality eventually fails: the electrode begins exhaustively depleting the sample (this is, in fact, exactly the coulometric limit used in some ultratrace protocols, where $t_d$ is deliberately made long enough to strip essentially all the metal from a small sample). In the ordinary stripping regime, one simply keeps $t_d$ short enough, or $V$ large enough, that depletion stays below a few percent, preserving the linear calibration derived above. This is why stripping analysis of small-volume samples (microliter droplets) requires either much shorter deposition times or accepting a nonlinear, exhaustive-electrolysis-type calibration.

Worked examples

Example 1 (foundational) — Calibration slope and detection limit for Cd²⁺ by ASV at a mercury film electrode. A mercury film electrode has area $A = 0.020$ cm², film thickness $l = 50$ nm, and is used with a stirring rate giving $m = 1.5\times10^{-3}$ cm s⁻¹. Deposition is carried out at $-1.0$ V vs. SCE for $t_d = 120$ s, and the stripping scan uses $v = 0.050$ V s⁻¹ at $T = 298$ K. Cadmium undergoes $n = 2$ electron transfer. Step 1: compute the calibration slope $S$ .

S = \frac{n^2 F^2 A m v\, t_d}{2.7\, R T}.

Numerically, $n^2 = 4$ , $F^2 = (96\,485)^2 = 9.3094\times10^{9}\ \text{C}^2\,\text{mol}^{-2}$ , $A m v\,t_d = (0.020)(1.5\times10^{-3})(0.050)(120) = 1.80\times10^{-4}\ \text{cm}^3\,\text{V s}^{-1}$ (units: cm² · cm/s · V/s · s = cm³·V/s, consistent with the equation once combined with the $RT$ denominator in volt·mol units). Carrying the numbers through:

S = \frac{4 \times 9.3094\times10^{9} \times 1.80\times10^{-4}}{2.7 \times 8.314 \times 298}

= \frac{6.703\times10^{6}}{6690.6} = 1.002\times10^{3}\ \text{A}\,\cdot\,\text{cm}^3\,\text{mol}^{-1}\ \left(\equiv \text{A per (mol cm}^{-3}\text{)}\right).

Step 2: convert to a practical slope in A per µg/L. Convert bulk concentration units: $c^*$ in mol cm⁻³ relates to concentration in µg L⁻¹ via cadmium's molar mass ( $112.41$ g mol⁻¹) and $1\ \text{L} = 1000\ \text{cm}^3$ :

c^*\,[\text{mol cm}^{-3}] = \frac{c^*\,[\mu\text{g L}^{-1}]\times10^{-6}\,\text{g}/\mu\text{g}}{112.41\ \text{g mol}^{-1}\times1000\ \text{cm}^3\,\text{L}^{-1}} = 8.897\times10^{-12}\, c^*\,[\mu\text{g L}^{-1}].

So $i_p = S \times 8.897\times10^{-12}\, c^*\,[\mu\text{g L}^{-1}] = (1.002\times10^{3})(8.897\times10^{-12})\,c^* = 8.91\times10^{-9}\ \text{A per } \mu\text{g L}^{-1}$ , i.e. 8.9 nA per µg/L. Step 3: detection limit. If the instrument's baseline (residual/charging current) noise has standard deviation $\sigma_i = 0.5$ nA, then using the standard $3\sigma$ criterion,

c_{\text{DL}} = \frac{3\sigma_i}{S'} = \frac{3(0.5\ \text{nA})}{8.91\ \text{nA per }\mu\text{g L}^{-1}} = 0.168\ \mu\text{g L}^{-1} \approx 0.17\ \mu\text{g L}^{-1}\ (1.5\times10^{-9}\ \text{M}).

This is comfortably below the U.S. EPA drinking-water action level for cadmium (5 µg/L), confirming that a 2-minute ASV deposition on a modest mercury film electrode is fit for regulatory-level cadmium screening.

Example 2 (applied) — Standard addition determination of lead in a seawater sample by ASV. A 10.00 mL seawater sample is analyzed by ASV using standard addition to avoid matrix (chloride, dissolved organic matter) effects on the mass-transfer coefficient. Deposition/scan conditions are held fixed for all runs ( $t_d = 240$ s at $-1.0$ V, then anodic scan). The recorded peak currents (baseline-subtracted) are:

Addition	Volume added of 50.0 µg/L Pb²⁺ standard	Final Pb concentration added (µg/L, diluted into 10.10 mL)	Peak current $i_p$ (nA)
0	0 µL	0	12.4
1	50.0 µL	0.2475	24.9
2	100.0 µL	0.4950	37.1

(Each addition changes total volume negligibly — 10.00 mL + up to 0.10 mL — and dilution is accounted for in the "final concentration added" column, e.g. $50.0\ \mu\text{g/L}\times 0.0500\ \text{mL}/10.05\ \text{mL} = 0.2488$ µg/L, rounded here to 0.2475 for the 10.10 mL final volume convention used in this experiment — treat the tabulated values as given.) Step 1: linear regression of $i_p$ vs. added concentration. Using the three points $(0, 12.4)$ , $(0.2475, 24.9)$ , $(0.4950, 37.1)$ :

\text{slope} = \frac{(37.1-12.4)}{(0.4950-0)} = \frac{24.7}{0.4950} = 49.9\ \text{nA per }\mu\text{g L}^{-1}.

(A full three-point least-squares fit gives slope $= 49.85$ nA/(µg/L), intercept $= 12.42$ nA, essentially identical because the three points are close to collinear — $R^2 = 0.9998$ .) Step 2: extrapolate to zero peak current to find the original sample concentration. The standard-addition principle states that the x-intercept (negative of it) equals the original sample concentration:

c_{\text{sample}} = \frac{\text{intercept}}{\text{slope}} = \frac{12.42\ \text{nA}}{49.85\ \text{nA per }\mu\text{g L}^{-1}} = 0.2492\ \mu\text{g L}^{-1}.

Step 3: propagate uncertainty. With typical standard-addition regression uncertainties of $\pm 3\%$ relative on both slope and intercept (from replicate measurements, not shown), the combined relative uncertainty is $\sqrt{0.03^2+0.03^2} = 0.042$ , giving

c_{\text{sample}} = 0.249 \pm 0.011\ \mu\text{g L}^{-1}\ (2\sigma),

consistent with open-ocean surface seawater lead levels (typically 0.01–0.5 µg/L depending on region and depth, reflecting decades of declining atmospheric lead deposition since the phase-out of leaded gasoline).

Example 3 (exam/competition level) — Resolving overlapping Cu and Pb stripping peaks with square-wave ASV and estimating an intermetallic-compound error. A square-wave ASV scan of a solution containing both Cu²⁺ and Pb²⁺, deposited simultaneously onto a mercury film electrode at $-1.10$ V for $180$ s, shows two stripping peaks at $E_p(\text{Pb}) = -0.44$ V and $E_p(\text{Cu}) = +0.02$ V vs. Ag/AgCl, both vs. a supporting electrolyte of 0.1 M acetate buffer, pH 4.5. The measured peak currents are $i_p(\text{Pb}) = 186$ nA and $i_p(\text{Cu}) = 94$ nA. Separate single-metal calibration (identical conditions) gives sensitivities $S_{\text{Pb}} = 41.2$ nA per µg/L and $S_{\text{Cu}} = 38.5$ nA per µg/L. Step 1: naive concentrations, ignoring intermetallic interference.

c_{\text{Pb}} = \frac{186}{41.2} = 4.51\ \mu\text{g/L}, \qquad c_{\text{Cu}} = \frac{94}{38.5} = 2.44\ \mu\text{g/L}.

Step 2: recognize the intermetallic compound risk. Copper and lead do not form a strong intermetallic compound in mercury (unlike Cu–Zn or Cu–Ga), so in this specific pair the naive calculation above is usually acceptable; the classic distortion arises when Cu is present in large excess relative to Zn or Ga, sequestering the minor metal as an intermetallic and suppressing its stripping peak. Suppose the problem specifies that copper here is present at 5-fold excess over zinc, and a calibration study shows that at a Cu:Zn mole ratio of 5:1 the apparent Zn stripping peak is suppressed by a fraction $f = 0.18$ (18% signal loss) due to Cu $_x$ Zn intermetallic formation trapping zinc atoms in an electrochemically inactive (or higher-overpotential) form. Step 3: corrected zinc concentration. If the measured (suppressed) zinc peak current is $i_p(\text{Zn}) = 58.0$ nA and the Zn sensitivity (measured in Cu-free solution) is $S_{\text{Zn}} = 33.0$ nA per µg/L, the naive concentration is

c_{\text{Zn,naive}} = \frac{58.0}{33.0} = 1.76\ \mu\text{g/L}.

Correcting for the 18% suppression (the true peak current, absent interference, would have been higher by a factor $1/(1-f)$ ):

c_{\text{Zn,true}} = \frac{c_{\text{Zn,naive}}}{1-f} = \frac{1.76}{1-0.18} = \frac{1.76}{0.82} = 2.15\ \mu\text{g/L}.

This 22% upward correction illustrates why multi-element stripping analysis in copper-rich matrices (many natural waters, industrial effluents) demands either matrix-matched standard addition, a masking/complexing agent for copper (e.g., adding a strong Cu-complexing ligand to suppress its co-deposition), or independent verification by ICP-MS — a single-element external calibration curve run in a "clean" matrix would have silently under-reported zinc by more than a fifth.

Limiting cases, scope & common traps

Regime / limiting case	Behavior	Consequence
Thin mercury film, $l \to 0$ at fixed $N_{\text{dep}}$	$i_p \propto v$ (linear in scan rate); film thickness cancels from calibration slope	MFEs give narrow, sensitive peaks; preferred for trace multi-element ASV
Thick drop / HMDE, semi-infinite diffusion	$i_p \propto v^{1/2}$ (like ordinary linear-sweep voltammetry)	Less sensitive per unit deposition time but far more reproducible, better for single-analyte ultratrace work
$t_d \to \infty$ at fixed sample volume	Bulk depletion becomes significant; linear mass balance $N_{\text{dep}} = mAc^*t_d$ breaks down	Approaches exhaustive (coulometric) stripping; useful deliberately for microliter samples, otherwise a calibration error source
High ionic strength / surfactant contamination	Adsorbed organics block the electrode surface, suppressing both $m$ and effective $A$	Peak suppression, poor reproducibility; UV photo-oxidation or standard addition needed to correct
Two metals with strong intermetallic compound formation (e.g., Cu–Zn, Cu–Ga)	Minor metal partially trapped in an electrochemically shifted/inactive compound	Suppressed, broadened, or shifted peak for the minor metal — see Example 3
Adsorptive stripping (AdSV) of a metal–ligand complex	Preconcentration by interfacial adsorption, not electron transfer; deposition step involves no faradaic current	Extends technique to species (e.g., Ni, Co with dimethylglyoxime) with poor amalgam formation but strong ligand adsorption

The chemist behind it

Stripping voltammetry's conceptual roots trace to Jaroslav Heyrovský's invention of Polarography in the 1920s, but its development as a quantitative trace-analysis tool is largely credited to Irving Shain and, in parallel, to the Czech and Soviet electroanalytical schools of the 1950s–60s who systematically worked out the deposition/stripping mass balance. Shain, working at the University of Wisconsin–Madison, provided much of the rigorous theoretical treatment of stripping peak shapes (thin-film and semi-infinite diffusion limiting cases) that this article's derivations draw on, extending the linear-sweep voltammetry theory he had developed with R. N. Adams and others for Cyclic Voltammetry to the anodic stripping case. Later, Joseph Wang at New Mexico State University (subsequently UC San Diego) became the field's most prolific modern practitioner, driving the development of mercury-free (bismuth, antimony) stripping electrodes, screen-printed disposable stripping sensors, and field-portable stripping instrumentation for environmental and biomedical trace-metal analysis — work motivated explicitly by the toxicity concerns around mercury electrodes that had made classical ASV so effective yet so environmentally awkward.

Origins · 1 minIrving Shain and how this was discovered

Irving Shain

1922–2002 · University of Wisconsin–Madison, electroanalytical chemistry

Shain's work in the 1950s–60s established rigorous mathematical treatments for linear-sweep and stripping voltammetry — solving the diffusion-controlled boundary-value problems that connect peak current, peak potential, and scan rate to fundamental parameters ( $n$ , $D$ , $C^*$ ). His collaboration with R. N. Adams on cyclic voltammetric theory and his subsequent extension to stripping analysis gave the field the quantitative footing (thin-film vs. semi-infinite diffusion peak equations) still taught and used today.

History

- 1922 — Heyrovský invents polarography, establishing controlled-potential electrolysis at a dropping mercury electrode as a quantitative tool.

- 1952–1960s — Shain and coworkers derive rigorous linear-sweep and stripping voltammetry theory, formalizing the thin-film and diffusion-controlled peak equations.

- 1957 — W. Kemula introduces the hanging mercury drop electrode (HMDE), providing a reproducible stationary mercury surface ideal for stripping work.

- 1970s — Differential pulse and later square-wave waveforms are coupled to stripping analysis, pushing detection limits from micromolar toward nanomolar and below.

- 1980s–2000s — Joseph Wang and coworkers develop mercury-film microelectrodes, adsorptive stripping voltammetry for organometallics and biomolecules, and eventually bismuth/antimony film electrodes as mercury-free alternatives.

Why is deposition time chosen rather than just running a longer stripping scan?

Deposition time controls how much analyte accumulates in/on the electrode before the measurement, while the stripping scan rate controls how sharply that accumulated material is read out. Lengthening the stripping scan doesn't add more analyte — it only changes peak shape and can even reduce sensitivity per the thin-film relation $i_p \propto v$ (slower scans give smaller thin-film peaks). Sensitivity gains come from extending $t_d$ , per $N_{\text{dep}} = mAc^*t_d$ , up to the point where bulk depletion or intermetallic saturation of the mercury film sets in. Practically, analysts balance longer $t_d$ (more sensitivity, more time per assay, more risk of nonlinearity) against throughput needs.

How does stripping voltammetry achieve detection limits better than ICP-MS in some cases?

It usually doesn't beat ICP-MS in absolute detection limit for most metals, but it wins on selectivity for the labile (bioavailable) fraction, instrument cost/portability, and freedom from certain matrix/isobaric interferences that plague ICP-MS (argon-based polyatomic interferences on ⁴⁰Ca, ⁵⁶Fe, etc.). Under favorable conditions — a well-behaved metal like lead or cadmium in a low-organic matrix with long deposition times — ASV detection limits of $10^{-10}$ – $10^{-11}$ M are achievable, competitive with or better than routine quadrupole ICP-MS, though modern ICP-MS/MS or sector-field instruments typically still win for absolute lowest limits.

What is the role of standard addition in stripping voltammetry, and why is it used so often here?

Standard addition adds known increments of analyte directly to the sample and re-measures the peak current after each addition, extrapolating the resulting linear plot back to zero signal to recover the original concentration. It is favored in stripping voltammetry because the mass-transfer coefficient $m$ (and hence the calibration slope $S$ ) depends sensitively on the sample's viscosity, ionic strength, and surfactant content — properties that vary between a calibration standard prepared in clean water and a real seawater, blood, or effluent sample. Standard addition keeps the matrix constant throughout the measurement, canceling matrix effects on $m$ that would otherwise bias an external calibration curve.

Can stripping voltammetry distinguish free metal ion from metal bound to organic ligands?

Yes, and this is one of its most valuable capabilities: only metal that is "electrochemically labile" — free ion or weakly/reversibly complexed species that can dissociate and be reduced within the deposition timescale — contributes to the stripping peak. Strongly bound metal-ligand complexes (e.g., copper bound tightly to humic acid or EDTA) do not deposit efficiently and are undercounted relative to the total metal measured by a digestion-based method like ICP-MS. Comparing ASV-labile metal to total digested metal is a standard way to estimate the fraction of "bioavailable" versus "complexed/sequestered" metal in environmental and biological samples.

Why has mercury been so hard to replace despite its toxicity concerns?

Mercury forms amalgams readily with a wide range of metals, has an unusually wide, reproducible, chemically inert electrochemical window, and its liquid surface (in the HMDE) is continuously renewable, eliminating surface-fouling memory effects that plague solid electrodes. Bismuth and antimony films successfully replicate much of this behavior — wide potential window, good amalgam-like alloying with several heavy metals — and have become the preferred mercury-free alternative for field and disposable sensors, but they do not match mercury's universality across the full periodic table of stripping-active metals, so mercury electrodes remain standard for certain analytes and rigorous reference methods.

What causes stripping peaks to broaden, shift, or split for some metal combinations?

Broadening and splitting typically arise from intermetallic compound formation within the mercury film or drop (classic examples: Cu–Zn, Cu–Ga, Ni–Zn), where the minor metal becomes partly trapped in a compound with a different, often more positive, oxidation potential than the free amalgam. This produces either a suppressed main peak, a secondary shoulder peak at a shifted potential, or both, and the effect scales with the mole ratio of the interfering metal, as illustrated quantitatively in Example 3. Adding a complexing agent that selectively ties up the interfering metal, or switching to a mercury-free electrode less prone to intermetallic chemistry, are standard remedies.

Is adsorptive stripping voltammetry (AdSV) fundamentally different from ASV/CSV?

AdSV shares the same two-step logic — preconcentrate, then strip — but the deposition step is non-faradaic: the analyte (often a metal-ligand complex, such as nickel-dimethylglyoxime) adsorbs spontaneously onto the electrode surface at a fixed potential, with no net electron transfer during accumulation. Only during the subsequent voltammetric scan is the adsorbed complex electrochemically reduced or oxidized, generating the stripping peak. AdSV extends stripping methodology to metals and organic species that do not form amalgams efficiently (Ni, Co, and various pharmaceuticals or DNA-binding agents), broadening the technique well beyond the classical amalgam-forming metals of ASV.