Complex Formation and Stepwise Stability Constants

Complexometric titrimetry. Every EDTA titration for water hardness (Ca²⁺/Mg²�"), for trace-metal assay in pharmaceutical quality control, or for direct/back/displacement titrations of dozens of metals rests entirely on the conditional-constant machinery of Example 2: $K_f'=\alpha_{Y^{4-}}K_f$ must exceed roughly $10^{6}$–$10^{8}$ for a sharp, usable end point, and $\alpha_{Y^{4-}}$ is controlled by the analyst through pH buffering.

Masking and selectivity in analysis. When two metals are both present and only one should react (e.g., masking Fe³⁺ with fluoride while titrating Ca²⁺ with EDTA), the choice of masking ligand and pH is a direct application of comparing stepwise/overall constants for the competing metal-ligand pairs — the analytical strategy generalized further under Competing Equilibria and Conditional Constants.

Environmental and geochemical speciation. Trace-metal bioavailability and toxicity in natural waters (e.g., Cu²⁺ toxicity to aquatic organisms) depend not on total copper but on the free $\mathrm{Cu^{2+}}$ fraction, computed by exactly the $\alpha_i([\mathrm{L}])$ machinery here applied to natural organic ligands, carbonate, and hydroxide simultaneously.

Pharmaceutical and clinical chemistry. Chelation therapy for heavy-metal poisoning (e.g., EDTA for lead, DMSA for mercury/arsenic) is dosed and evaluated using stability-constant comparisons between the therapeutic chelator and the metal's endogenous ligands (proteins, thiols) — the drug must win the competition, quantitatively guaranteed by a sufficiently large conditional $K_f'$ at physiological pH.

Instrumentation and colorimetric assay. Classic spectrophotometric metal assays (Fe²⁺ with 1,10-phenanthroline, forming the intensely colored $\mathrm{Fe(phen)_3^{2+}}$) depend on driving the equilibrium essentially completely to the highest complex ($\alpha_3 \to 1$) by adding a large excess of ligand, exactly the strategy used in Example 1's silver-ammine calculation, so that absorbance reports directly on total metal without correction for partial complexation.

Materials and electroplating chemistry. Cyanide- and ammine-based plating baths (silver, gold, copper plating) are engineered around specific stepwise complexes because the free-metal-ion concentration — not the total metal concentration — controls deposition rate and grain structure at the cathode; the entire bath chemistry is a designed application of the $\alpha_i([\mathrm{L}])$ curves derived here.

:::quiz[As free ligand concentration [L] → ∞ in an n-step system, which species' fraction approaches 1?]

• Free metal, M
• The intermediate species with the largest single K_f,i
• The lowest complex, ML
• The highest complex, ML_n ✓

:::quiz[In the Cd²⁺/NH₃ example, why was CdA₂ (not CdA₄) the dominant species at [NH₃] = 0.0100 M?]

• Because β₄ is smaller than β₂
• Because the α-denominator's β₂[A]² term was numerically larger than the β₄[A]⁴ term at that specific [L] ✓
• Because CdA₄ cannot exist in aqueous solution
• Because K_f,4 is negative

Ask the tutor: How do I set up and solve the simultaneous mass-balance equations when the metal and ligand concentrations are comparable, so free ligand can't just be approximated by total ligand added?

A metal ion in aqueous solution is never truly "free" — it is solvated, typically by six water molecules occupying an octahedral coordination sphere (four for many d¹⁰ ions like Zn²⁺ or Cd²⁺ in some ligand environments, two for linear Ag⁺ complexes). Complex formation with a ligand L is fundamentally a substitution reaction: L displaces coordinated water molecules one at a time. Writing $\mathrm{M(H_2O)}_x$ simply as "M" (the convention throughout this article and in Skoog/Harris-style treatments), the ladder of substitutions is

Each step has its own equilibrium constant, the stepwise formation constant $K_{f,i}$, defined for the $i$-th step as

Because each step removes one water and adds one ligand, and because there are progressively fewer open coordination sites and progressively more electrostatic/steric crowding as $i$ increases, stepwise constants for a monodentate ligand almost always decrease monotonically: $K_{f,1} > K_{f,2} > K_{f,3} > \cdots$. This is partly a statistical effect (more ways to attach the first ligand than the last) and partly a genuine energetic effect (charge neutralization makes the metal a weaker Lewis acid after each substitution, and crowding raises the enthalpic cost of further coordination). The overall or cumulative formation constant $\beta_n$ describes the reaction of bare metal directly with $n$ ligands,

This is the identical logical move used for The Equilibrium Constant and Reaction Quotient applied to any multistep mechanism: multiplying stepwise $K$'s for a sequence of reactions that sum to an overall reaction gives the overall $K$.

Multidentate ("chelating") ligands change the picture qualitatively. EDTA (ethylenediaminetetraacetate) offers up to six donor atoms (two N, four O) and wraps around a metal in a single binding event, so instead of a ladder of four to six steps you typically see one overwhelmingly dominant 1:1 species, $\mathrm{ML}$, with a single enormous formation constant $K_f$ (often $10^{8}$–$10^{25}$ depending on the metal). This "chelate effect" — a large positive entropy gain because one EDTA molecule releases many water molecules on binding, versus six separate monodentate ligands each displacing one water with a much smaller net entropy change — is why chelating titrants dominate quantitative complexometric analysis.

The two governing relationships are the constant definitions themselves and the speciation (fractional composition) formulas built from them.

Stepwise formation constant, step $i$:

where: $[\mathrm{ML}_i]$ is the equilibrium concentration of the complex bearing $i$ ligands, $[\mathrm{L}]$ is the free (uncomplexed) ligand concentration, and $K_{f,i}$ carries units of $\mathrm{M}^{-1}$.

Overall (cumulative) formation constant:

where: $\beta_n$ has units $\mathrm{M}^{-n}$, and $\beta_0 \equiv 1$ (the "bare metal" species) by convention, so that the formula below is uniform for all species including free M.

Fractional composition (speciation) of species $\mathrm{ML}_i$:

where: $D = 1 + \beta_1[\mathrm{L}] + \beta_2[\mathrm{L}]^2 + \cdots + \beta_n[\mathrm{L}]^n$ is the common denominator (the "$\alpha$-denominator"), and $\alpha_i = [\mathrm{ML}_i]/C_M$ with $C_M$ the total (analytical) metal concentration.

Reading these equations. The stepwise constant is a local statement — it only compares two adjacent rungs of the ladder. The overall constant $\beta_n$ is a global statement connecting the fully-complexed species directly to bare metal and free ligand, exactly the way $K_{a1}K_{a2}$ connects H₂A to A²⁻ in a diprotic acid without reference to the intermediate HA⁻. The fractional composition formula is the payoff: once you know all the $\beta_i$ and the free ligand concentration, every species' fraction of total metal drops out of one polynomial, with no need to solve simultaneous equilibria by hand.

Example 1 (foundational) — Ag⁺/NH₃, two-step system. Silver forms only two ammine complexes of practical importance, with reported stepwise constants (25 °C) $K_{f,1} = 2.1\times10^{3}$ and $K_{f,2} = 8.2\times10^{3}$, giving

Suppose a solution is prepared with total silver $C_{Ag} = 3.40\times10^{-4}\ \mathrm{M}$ and free ammonia held (by large excess and buffering) at $[\mathrm{NH_3}] = 0.150\ \mathrm{M}$. Find the fraction of silver present as each species. First build the denominator polynomial with $n=2$:

Now the fractions:

Check: $2.58\times10^{-6} + 8.13\times10^{-4} + 0.99917 = 1.0000$ ✓. Essentially all silver (99.92%) is the diammine complex, as expected given the large excess of ammonia and the sizeable $\beta_2$. The actual free-silver concentration is

— a value low enough to matter for AgCl solubility calculations (the classic use of this system in qualitative analysis schemes), but not obtainable without carrying both stepwise constants through the full polynomial.

Limiting-case check: as $[\mathrm{NH_3}] \to 0$, every term but the leading "1" vanishes from $D$, so $\alpha_{Ag^+}\to 1$ and all complexed fractions vanish — correctly recovering the free-metal limit.

Example 2 (applied) — conditional stability and an EDTA-style masking calculation. A calcium ion is titrated with EDTA (Y⁴⁻ form) at pH 10.0, where the fraction of EDTA present as the fully deprotonated Y⁴⁻ species is $\alpha_{Y^{4-}} = 0.35$ (a realistic value at pH 10, since EDTA is a hexaprotic-acid–like ligand whose effective basicity strongly depends on pH — this is the Fractional Composition of Diprotic Systems machinery applied to the ligand rather than the metal). The literature absolute formation constant for CaY²⁻ is

Because only a fraction $\alpha_{Y^{4-}}$ of total EDTA is in the reactive Y⁴⁻ form at this pH, define a conditional formation constant,

This conditional constant folds the pH-dependent ligand speciation into an effective 1:1 constant, so that the titration can be treated with the simple mass-balance algebra of an ordinary 1:1 complex — a preview of the general strategy in Competing Equilibria and Conditional Constants. Now compute the free calcium concentration remaining at the equivalence point of a titration where the analytical concentration of the CaY²⁻ complex formed is $C_{CaY} = 1.00\times10^{-2}\ \mathrm{M}$ (a 1:1 stoichiometric mixture at the stoichiometric point, so free Ca²⁺ equals free total-Y). At the equivalence point, let $x = [\mathrm{Ca^{2+}}]_{\text{free}} = [\mathrm{Y}]_{\text{total, uncomplexed}}$. The conditional equilibrium is

(valid because $x \ll C_{CaY}$, checked after solving). Solving for $x$:

Confirming the approximation: $x/C_{CaY} = 7.6\times10^{-5}$, well under 1%, so dropping $x$ next to $C_{CaY}$ was justified. The corresponding pCa at the equivalence point is

This is exactly the number a real complexometric titration curve reports as the equivalence-point break — and it depends on pH through $\alpha_{Y^{4-}}$ exactly as this calculation shows: raising pH to 12 (where $\alpha_{Y^{4-}}\to$ nearly 1) sharpens the break; buffering at pH 4 (where $\alpha_{Y^{4-}}\approx 10^{-8}$) would make the conditional constant far too small to titrate calcium sharply at all.

Example 3 (exam / competition level) — full four-species Cd²⁺/NH₃ speciation and a maximum-fraction search. Cadmium forms up to four ammine complexes with literature cumulative constants (25 °C, $I=0.1$)

(equivalently $K_{f,1}=3.6\times10^2$, $K_{f,2}=211$, $K_{f,3}=4.87$, $K_{f,4}=10.5$ — note the constants do not decrease perfectly monotonically past step 3, a real feature of the Cd–NH₃ system reflecting a coordination geometry change, a useful reminder that "usually decreasing" is not "always decreasing").

(a) Find all five fractions ($\mathrm{Cd^{2+}}$, and $\mathrm{CdA}$ through $\mathrm{CdA_4}$, using A = NH₃) at $[\mathrm{NH_3}] = 1.00\times10^{-2}\ \mathrm{M}$. Build $D$:

Term by term with $[\mathrm{A}] = 1.00\times10^{-2}$:

The fractions:

Sum check: $0.0793+0.2855+0.6027+0.02935+0.00309 = 1.0000$ ✓ (rounding to four places). At this ligand concentration, CdA₂ dominates with over 60% of total cadmium — a genuine non-monotonic speciation outcome that could not have been guessed from $K_{f,1}$ alone, since $\mathrm{CdA_2}$ (not the highest complex, CdA₄) is the majority species.

(b) As a competition-level extension, find the free-ligand concentration at which $\alpha_2$ (the CdA₂ fraction) is maximized, and confirm it is a true interior maximum rather than a monotonic approach to 1. $\alpha_2 = \beta_2[\mathrm{A}]^2/D$ is maximized where $d\alpha_2/d[\mathrm{A}] = 0$. Using the quotient rule on $\alpha_2 = \beta_2x^2/D(x)$ with $x\equiv[\mathrm{A}]$,

where $D' = \beta_1 + 2\beta_2 x + 3\beta_3x^2 + 4\beta_4x^3$. Writing out $2D = xD'$ explicitly with $\beta_0=1$:

Substituting the numerical $\beta$'s and solving numerically (Newton iteration starting from $x_0 = 1.0\times10^{-2}$, the neighborhood suggested by part (a) where $\alpha_2$ was already large) gives

close to, but distinctly different from, the arbitrary $1.00\times10^{-2}$ M chosen in part (a) — confirming that $\alpha_2$ passes through a genuine interior maximum rather than climbing monotonically to unity (which only $\alpha_n$, the highest complex, ever does as $[\mathrm{L}]\to\infty$). Evaluating $D$ and $\alpha_2$ at $x^*$ shows the maximum fraction is only slightly above the 0.603 found at $x=0.0100\ \mathrm{M}$, consistent with a shallow, broad maximum rather than a sharp spike — typical behavior for an intermediate species in a stepwise ladder with comparable neighboring $\beta$ values.

The formalism of stepwise and overall stability constants is inseparable from the Danish/Swedish-trained coordination chemistry tradition, but the single figure most responsible for turning "complexes exist" into "complexes are quantifiable equilibrium ladders with measurable constants" is Jannik Bjerrum, working in Copenhagen in the 1940s under the shadow of Niels Bjerrum (his father, also a major figure in ionic equilibrium theory). Jannik Bjerrum's 1941 doctoral thesis, Metal Ammine Formation in Aqueous Solution, introduced the systematic potentiometric measurement of stepwise formation constants for a whole series of transition-metal ammine complexes — exactly the Ag⁺/NH₃ and Cd²⁺/NH₃ systems used as worked examples in this article — using pH-independent methods (since ammonia itself is a weak base) that could isolate ligand-binding equilibria from proton equilibria. His work established the "$\bar n$" function (average ligand number, essentially $\bar n = \sum i\alpha_i$) as the standard experimental route to stepwise constants: you titrate metal with ligand, measure free ligand concentration potentiometrically or otherwise, and extract $\bar n$ versus $[\mathrm{L}]$ curves whose shape directly encodes every $K_{f,i}$.