Predicting Redox Behavior from Diagrams

Frost-diagram reasoning underlies practical water-treatment chemistry: chlorine disinfection relies on hypochlorite being a convex, kinetically-slowed-but-thermodynamically-doomed species that nonetheless persists long enough to oxidize pathogens before disproportionating, a balance directly visible on the chlorine oxyanion diagram worked above. Battery and fuel-cell engineers use the same geometric logic when screening candidate redox couples: a steep, well-separated slope between two accessible oxidation states of a transition metal (vanadium redox flow batteries being the classic example, exploiting V(II)/V(III) and V(IV)/V(V) couples that are both concave and well-separated) signals a couple that will cycle cleanly without disproportionating mid-charge. In analytical chemistry, permanganate and dichromate titrimetry both rely on being at the extreme oxidizing edge of a steep Frost segment — the same geometric feature (steep rightward rise with no point beyond it) that predicts these species are powerful, kinetically accessible oxidants for redox titrations. Corrosion scientists extend the same convexity logic into two dimensions via Pourbaix Diagrams, where the stability field of a metal or its oxide is bounded by lines whose slopes are still literally reduction potentials, generalizing the one-dimensional Frost geometry to include pH dependence. Geochemists use Frost-style reasoning to explain why certain manganese and iron oxidation states dominate in different depths of soils and sediments (redox stratification), while biochemists invoke the same logic to explain why certain metalloenzyme cofactors (e.g., manganese in photosystem II, iron-sulfur clusters) cycle between two specific, diagram-predicted concave oxidation states rather than drifting toward disproportionation products during catalytic turnover. Even the question of why water itself is redox-stable over a specific potential window — a question this same geometric machinery answers directly once oxygen and hydrogen are added to the same kind of diagram — is a direct extension of the ideas here, covered fully under Redox Stability of Water.

Ask the tutor: How would the manganese Frost diagram change shape if we redrew it for basic solution instead of acidic, and would Mn²⁺ still be the global minimum?

A Frost diagram is built by choosing a reference state (usually the free element, $n=0$, at $nE^\circ = 0$ by convention) and plotting every known oxidation state of an element along the horizontal axis, with the vertical axis equal to $n$ times the standard reduction potential for reducing that species all the way down to the element. Because $nE^\circ$ is proportional to $-\Delta G^\circ$ for the reduction to the element, the vertical axis is literally a free-energy axis in volt-equivalents. This is why the whole geometric machinery works: differences in height are differences in free energy, and the chemist's usual intuition about "downhill" reactions being favorable translates directly into "downhill on the diagram." Consider three consecutive oxidation states $A$ (highest, oxidation number $n_1$), $B$ (middle, $n_2$), and $C$ (lowest, $n_3$), all of the same element. There are two ways to get from $A$ to $C$: directly, in one two-electron (or more) step, or via $B$, in two separate steps. The disproportionation of $B$ is

(coefficients depend on electron counts, but the essential comparison is always free energy of $B$ against the weighted average free energy of $A$ and $C$ at the same electron count). Because the Frost diagram plots free energy directly, this comparison is just: is point $B$ above or below the straight line (chord) connecting points $A$ and $C$? If $B$ lies above the chord, then converting $B$ into a mixture of $A$ and $C$ lowers the total free energy — disproportionation is spontaneous. If $B$ lies below the chord, $B$ is at lower free energy than any $A$/$C$ mixture at the same average oxidation state, so $B$ is stable and, if you started with a mixture of $A$ and $C$, comproportionation to $B$ would be favorable instead. This three-point test is completely local: it only involves the species immediately adjacent to $B$ in oxidation state (assuming those are the only species chemically accessible; some intermediate integer oxidation states may not correspond to any known compound and are simply skipped over, using the next known species as the relevant neighbor). Convexity of the whole curve — bulging upward — signals instability everywhere it occurs; concavity — sagging into a valley — signals stability. A diagram that is everywhere concave (convex downward, like a bowl) has no disproportionation anywhere and every intermediate state is a genuine local minimum relative to its neighbors, though only one of them is the global minimum.

The other geometric feature that matters is slope, because slope between two adjacent points is exactly the standard reduction potential connecting them:

A steep positive rise moving left to right (i.e., a large positive slope) means the higher oxidation state is a powerful oxidizer of the lower one — it wants to grab electrons and fall down the diagram. A steep drop (large negative slope reading left to right, meaning the reduced form sits much higher) means the lower oxidation state is a powerful reducing agent. Comparisons between an element's diagram and the water-stability lines (fixed slopes for the $\text{O}_2/\text{H}_2\text{O}$ and $\text{H}^+/\text{H}_2$ couples) then predict whether a given species can even persist in aqueous solution — the natural bridge toward Redox Stability of Water and ultimately full Pourbaix Diagrams.

The Frost-diagram vertical coordinate for a species with oxidation number $n$, referenced to the element ($n=0$), is defined by

where:

• $\Delta G^\circ_{n \to 0}$ is the standard Gibbs energy for reducing the species all the way to the free element,
• $F$ is the Faraday constant,
• the negative sign converts a (usually negative, favorable) free energy into a positive-going potential axis when reduction is favorable.

Reading: this is just $\Delta G^\circ = -nFE^\circ$ rearranged and summed if several successive steps are needed to reach the element; consecutive $E^\circ$ values are additive in free energy (in $nE^\circ$ units) even though they are not directly additive as raw potentials. The slope between two adjacent points is the reduction potential of the couple linking them:

where:

• $n_1 > n_2$ are the two oxidation states being compared,
• $(nE^\circ)_1, (nE^\circ)_2$ are their respective Frost-diagram heights,
• the result $E^\circ_{n_1/n_2}$ is the standard potential for reducing $n_1$ to $n_2$.

Reading: any two points, adjacent or not, define a chord whose slope is a bona fide reduction potential for the (possibly multi-electron, possibly non-elementary) couple between them; this is what lets you read off $E^\circ$ for any pair of oxidation states directly from the graph. The disproportionation/comproportionation criterion for a species $B$ (oxidation state $n_2$) flanked by $A$ (state $n_1$) and $C$ (state $n_3$), with $n_1>n_2>n_3$, is

where $y_{\text{chord}}(n_2)$ is the height of the straight line from $A$ to $C$ evaluated at $n_2$; more usefully this collapses to the sign rule

where:

• $y_{\text{chord}}(n_2)$ is interpolated linearly between the $A$ and $C$ points,
• $(n_2E^\circ)_B$ is $B$'s actual plotted height,
• the comparison direction (above chord ⇒ unstable) follows because higher free energy (higher on the plot, since the axis is $-\Delta G/F$ scaled — wait, check sign) — properly, since $nE^\circ = -\Delta G/F$, a higher point means more negative $\Delta G^\circ_{n\to0}$, i.e. lower free energy for the elemental-reduction pathway, but the comparison that matters is relative to the chord, not absolute height, and that relative comparison is what the sign rule captures correctly regardless of the overall axis orientation.

Reading: you never need the messy general formula — the geometric picture (above chord = unstable, below chord = stable) is exactly equivalent to it and is how the diagram is used in practice.

Example 1 (foundational): reading manganese's Frost diagram in acid. The standard Frost data for manganese in 1 M acid (approximate, from tabulated $E^\circ$ values) give the following $nE^\circ$ values relative to Mn(0):

(Values are the standard tabulated ones; small variations exist between sources but the qualitative shape is robust.) Reading the shape: Mn²⁺ at $-2.36$ V is the deepest point on the whole diagram — by the global-minimum argument, Mn²⁺ is the thermodynamically preferred manganese species in acidic aqueous solution, consistent with everyday observation that Mn²⁺ salts are the stable, shelf-bottle form of manganese. Check Mn³⁺ for disproportionation: its neighbors are Mn²⁺ ($n=2$, $y=-2.36$) and MnO₂ ($n=4$, $y=+0.10$). The chord height at $n=3$ is the simple average (since $n=3$ is the midpoint of $n=2$ and $n=4$):

Mn³⁺'s actual height is $-0.85$ V, which is above $-1.13$ V. By the sign rule, Mn³⁺ is convex and disproportionates:

This matches the well-known experimental instability of Mn³⁺(aq) outside of complexing media that stabilize it (e.g., pyrophosphate). Check the far right: MnO₄²⁻ to MnO₄⁻ has slope $E^\circ = 0.56$ V, and MnO₄⁻ sits at the far right with nothing above it — this steep upward-right trajectory identifies MnO₄⁻ (permanganate) as a strong oxidizing agent, exactly as used in redox titrimetry.

Limiting-case sanity check: if the Mn³⁺ point had instead fallen exactly on the chord (hypothetically $y = -1.13$ V), $\Delta G^\circ_{\text{disp}}$ would be exactly zero — Mn³⁺ would be in equilibrium with the Mn²⁺/MnO₂ mixture at standard conditions, neither favoring disproportionation nor comproportionation. This equal-height case is the borderline that the sign rule correctly assigns to $K_{\text{eq}}=1$.

Example 2 (applied): chlorine oxyanions and the "why does bleach work" question. Chlorine in base has accessible states Cl⁻ ($n=-1$), Cl₂ ($n=0$, reference for the elemental convention shifted appropriately), ClO⁻ (hypochlorite, $n=+1$), ClO₂⁻ ($n=+3$), ClO₃⁻ ($n=+5$), ClO₄⁻ ($n=+7$). Using standard basic-solution potentials (approximate, in volts vs. NHE, referenced so that Cl⁻ sits at a convenient baseline): Take $E^\circ(\text{ClO}^-/\text{Cl}^-)$ (net 2-electron, base) $\approx +0.89$ V and $E^\circ(\text{Cl}_2/\text{Cl}^-) \approx +1.36$ V (this one is medium-independent since it doesn't involve OH⁻ directly), and $E^\circ(\text{ClO}^-/\text{Cl}_2)\approx+0.42$ V. Setting Cl⁻ at $n=-1$, $y=-1\times E^\circ(\text{Cl}^-/\text{Cl}^-)$ is not meaningful directly; instead build the diagram anchored at Cl₂ ($n=0$, $y=0$ by convention) and work outward:

• Cl⁻ ($n=-1$): $y = -1\times(+1.36) = -1.36$ V (oxidizing Cl⁻ to Cl₂ costs $+1.36$ V, so reducing Cl₂ to Cl⁻ releases that; the sign convention for $n<0$ side is $y = n E^\circ_{\to \text{element}}$, giving $-1.36$).
• Cl₂ ($n=0$): $y=0$.
• ClO⁻ ($n=+1$): using $E^\circ(\text{ClO}^-/\text{Cl}_2)=+0.42$ (this is the potential for reducing ClO⁻ to Cl₂, per electron since it's a 1-electron step per Cl), $y = 0 + 1\times0.42 = +0.42$ V.
• ClO₂⁻ ($n=+3$): with $E^\circ(\text{ClO}_2^-/\text{ClO}^-)\approx+0.66$ V (2-electron step), $y = 0.42 + 2\times0.66 = 1.74$ V.
• ClO₃⁻ ($n=+5$): with $E^\circ(\text{ClO}_3^-/\text{ClO}_2^-)\approx+0.33$ V (2-electron), $y=1.74+2\times0.33=2.40$ V.
• ClO₄⁻ ($n=+7$): with $E^\circ(\text{ClO}_4^-/\text{ClO}_3^-)\approx+0.36$ V (2-electron), $y=2.40+2\times0.36=3.12$ V.

Check ClO⁻ for disproportionation using neighbors Cl₂ ($n=0,y=0$) and ClO₂⁻ ($n=3,y=1.74$). Chord at $n=1$:

ClO⁻'s actual height is $0.42$ V, which is below the chord ($0.42 < 0.58$), so by the sign rule ClO⁻ is (barely) concave here relative to this particular pair and resists disproportionation into Cl₂ and ClO₂⁻ specifically. But the chemically relevant comparison for household bleach chemistry is ClO⁻ against Cl⁻ and ClO₃⁻ (the actual observed disproportionation products), a wider three-point test: Chord between Cl⁻ ($n=-1,y=-1.36$) and ClO₃⁻ ($n=5,y=2.40$) evaluated at $n=1$:

ClO⁻ sits at $+0.42$ V, well above this chord, so ClO⁻ is convex and unstable with respect to disproportionation into Cl⁻ and ClO₃⁻:

This is exactly the reaction that limits shelf-life of bleach and is dramatically accelerated by heat and by molybdate/certain transition-metal catalysts — a textbook case where the thermodynamic verdict (unstable, confirmed by the diagram) is correct, but kinetics keeps bottled bleach usable for months.

Limiting-case check: if instead ClO⁻ had been measured to sit exactly on the wide Cl⁻–ClO₃⁻ chord, disproportionation would be at equilibrium ($K=1$) rather than strongly favorable — illustrating that the size of the vertical gap above the chord, not just its sign, sets how far $K_{\text{eq}}$ deviates from unity via $\Delta G^\circ = -RT\ln K$.

Example 3 (exam/competition level): comparing two hypothetical elements from bare Frost data, no chemical names given. An exam question gives you only numbers: element X has accessible oxidation states $n = 1, 2, 3, 4$ with $nE^\circ = 0.40, 0.30, 1.50, 1.20$ V respectively (state $n=0$, the element, fixed at $y=0$). Determine (a) which state is thermodynamically most stable, (b) which states disproportionate, and (c) the strongest oxidizing and reducing couples present. Step (a) — global minimum. Compare all five points: $n=0\ (0.00)$, $n=1\ (0.40)$, $n=2\ (0.30)$, $n=3\ (1.50)$, $n=4\ (1.20)$. The lowest value is $y=0.00$ at $n=0$ — the free element is the thermodynamically preferred form. This is an important reminder that the "most stable" oxidation state is not always some exotic intermediate; sometimes it really is $n=0$ (as, for instance, with noble metals like gold, whose Frost diagram is famously "all uphill" from the element). Step (b) — check every interior point for convexity. For $n=1$ (flanked by $n=0,y=0$ and $n=2,y=0.30$): chord at $n=1$ is

Actual $y=0.40 > 0.15$: convex, so $n=1$ disproportionates into $n=0$ and $n=2$. For $n=2$ (flanked by $n=1,y=0.40$ and $n=3,y=1.50$): chord at $n=2$:

Actual $y=0.30 < 0.95$: concave, so $n=2$ is stable against disproportionation into $n=1,3$ — comproportionation of $n=1$ and $n=3$ toward $n=2$ would instead be favorable if both were present together. For $n=3$ (flanked by $n=2,y=0.30$ and $n=4,y=1.20$): chord at $n=3$:

Actual $y=1.50 > 0.75$: strongly convex, so $n=3$ disproportionates sharply into $n=2$ and $n=4$. Step (c) — strongest oxidizer/reducer from slopes. The steepest rising segment (largest positive slope reading left to right) is $n=2\to n=3$: slope $= (1.50-0.30)/(3-2) = 1.20$ V — this identifies the $n{=}3/n{=}2$ couple as the strongest oxidizing couple on the diagram (the $n=3$ species is eager to grab electrons and fall to $n=2$). The steepest falling segment moving left to right identifies the strongest reductant; here $n=3\to n=4$ has slope $(1.20-1.50)/(4-3) = -0.30$ V, and $n=1\to n=2$ has slope $(0.30-0.40)/(2-1)=-0.10$ V — the more negative one, $n=3\to n=4$ at $-0.30$ V, means $n=4$ (if you approach it by reduction from higher, or think of it as its own couple with $n=3$) has $E^\circ(n{=}4/n{=}3) = -0.30$ V, i.e., $n=3$ is a moderately strong reducing agent toward $n=4$. The single strongest reductant overall, however, is best read off the leftmost segment $n=0\to n=1$: slope $=(0.40-0)/(1-0)=0.40$ V, meaning $E^\circ(n{=}1/n{=}0)=0.40$ V — actually this says the element (n=0) is a reasonably strong reductant toward forming $n=1$ (positive $E^\circ$ for that couple means the reduction $n{=}1\to n{=}0$ is spontaneous, so the element is willing to be oxidized, i.e., acts as reductant, whenever $n=1$ is around as oxidant). Reporting both the oxidizer and reductant identifications together gives the full qualitative redox map of element X purely from the four numbers supplied, with no reference compound names needed at all.