Skip to content

Organometallic Reactions: Substitution to β-Elimination

CSIR NET chemistry: substitution, oxidative addition, reductive elimination, migratory insertion, β-hydride elimination, σ-bond metathesis and ligand attack.

  • 17 min read
  • Updated
On this page · 17 sections
  1. Prerequisites
  2. The Hook: A Complex That Breathes
  3. The Master Table
  4. Ligand Substitution
  5. Oxidative Addition
  6. Reductive Elimination
  7. Migratory Insertion
  8. Elimination Reactions
  9. σ-Bond Metathesis
  10. Attack on Coordinated Ligands
  11. Memory Tricks
  12. Don’t Confuse
  13. Common Mistakes
  14. Practice
  15. Quick Revision Table
  16. Key Takeaways
  17. What’s Next

Prerequisites

The Hook: A Complex That Breathes

The Master Table

Learn this table and you can label any step in a catalytic cycle.

ReactionΔ oxidation stateΔ coordination numberΔ electron countTypical requirement
Ligand dissociation0−1−218-e complex, bulky ligands
Ligand association0+1+2≤ 16-e complex
Oxidative addition (OA)+2+2+2≤ 16 e⁻, two vacant sites, OS +2 accessible
Reductive elimination (RE)−2−2−2the two groups cis
Migratory insertion (1,1 or 1,2)0−1−2the two groups cis
β-Hydride elimination0+1+2β-H, vacant cis site, syn-coplanar M–C–C–H
α-Hydride elimination0 (carbene as neutral L; +2 if written as CR₂²⁻)+1+2 (neutral method)α-H, early metal
Oxidative coupling of two alkenes+20−2gives a metallacyclopentane
σ-Bond metathesis000d⁰ metal (no electrons for OA)
graph TD
    A["What changed at the metal?"] --> B{"Oxidation state +2?"}
    B -->|yes| C["Oxidative addition
(or alpha-elimination)"] B -->|no| D{"Oxidation state -2?"} D -->|yes| E["Reductive elimination"] D -->|no| F{"Electron count -2,
new bond within ligands?"} F -->|yes| G["Migratory insertion"] F -->|no| H{"Electron count +2,
alkene + M-H formed?"} H -->|yes| I["Beta-hydride elimination"] H -->|no| J["Substitution, sigma-bond metathesis
or attack on a ligand"]

Each pair is a microscopic reverse: OA ↔ RE, insertion ↔ elimination, association ↔ dissociation. Because each step changes the count by 0 or 2, catalytic intermediates shuttle between 16 and 18 electrons, which is Tolman’s 16/18-electron rule for catalysis, developed on the catalysis page.

Ligand Substitution

18-electron complexes: dissociative (D)

An 18-electron complex cannot accept a seventh ligand without putting electrons into an antibonding orbital, so substitution starts with loss of a ligand:

$$\text{M(CO)}_6 \xrightarrow{k_1,\ \text{slow}} \text{M(CO)}_5 + \text{CO}; \qquad \text{M(CO)}_5 + \text{L} \xrightarrow{\text{fast}} \text{M(CO)}_5\text{L}$$

Signs of a dissociative mechanism:

  • rate = \( k_1[\text{M(CO)}_6] \): first order in complex, zero order in incoming ligand L;
  • rate nearly independent of the nature of L;
  • positive ΔS‡ (two particles from one in the transition state);
  • rate increases with ligand bulk (steric relief) and for metals and ligands that bind CO less strongly.

17- and 16-electron complexes: associative (A)

A complex with fewer than 18 electrons has a vacant low-lying orbital and can bind L first.

  • 16-electron square-planar d⁸ complexes (Pt(II), Pd(II), Rh(I), Ir(I)) substitute associatively through a five-coordinate intermediate, as in classical Pt(II) chemistry.
  • 17-electron radicals substitute associatively and very fast: V(CO)₆ exchanges CO far faster than the 18-electron Cr(CO)₆, and the 17-electron Mn(CO)₅ radical is far more labile than Mn₂(CO)₁₀. One-electron oxidation or reduction of an 18-electron complex can therefore catalyse substitution (electron-transfer chain catalysis).
  • Signs: rate = \( k_2[\text{M}][\text{L}] \); rate depends on the nucleophilicity of L; negative ΔS‡.

Many 18-electron carbonyls show a two-term rate law, rate = \( (k_1 + k_2[\text{L}])[\text{M}] \): a dissociative path plus an associative path with good nucleophiles.

How an 18-electron complex can react associatively

A ligand can rearrange to free up two electrons’ worth of space without leaving:

RearrangementElectron changeExample
η⁵ → η³ ring slipCp: 5 → 3 e⁻(η⁵-indenyl)Rh(CO)₂ substitutes associatively, very much faster than CpRh(CO)₂: the indenyl effect, because slipping to η³ restores the aromaticity of the fused benzene ring
Linear → bent NO3 → 1 e⁻Co(CO)₃(NO) substitutes by an associative path
η³ → η¹ allyl3 → 1 e⁻allyl complexes
η⁶ → η⁴ arene6 → 4 e⁻arene exchange in (arene)Cr(CO)₃

Photochemical and chemical CO removal

UV light labilises CO (Cr(CO)₆ + hν in THF → Cr(CO)₅(THF), then L replaces THF). Trimethylamine N-oxide removes CO under mild conditions: Me₃NO attacks a carbonyl carbon, CO₂ and NMe₃ leave, and a vacant site remains.

Oxidative Addition

$$\text{L}_n\text{M}^{(x)} + \text{A–B} \longrightarrow \text{L}_n\text{M}^{(x+2)}(\text{A})(\text{B})$$

The metal gives two electrons to break A–B and forms two new M–A and M–B bonds.

What favours oxidative addition

  • Electron-rich metal: low oxidation state, donor ligands (PMe₃ better than PPh₃, which is better than CO).
  • Heavier metal in a group for the same reaction, especially in group 9: Ir(I) adds small molecules more readily and more irreversibly than Rh(I). Third-row metals form stronger M–L bonds and stable higher oxidation states. (Do not over-apply this: for aryl halides, Ni(0) is often more reactive than Pd(0).)
  • Coordinative unsaturation: 16-electron (or lower) complexes; an 18-electron complex must first lose a ligand.
  • A weak A–B bond and strong new M–A, M–B bonds.

Mechanisms

MechanismTypical substratesStereochemistry and evidence
Concerted (three-centre transition state)H₂, C–H, Si–H, aryl halides on Pd(0)cis addition; retention at carbon; often via a σ-complex (η²-H₂)
S_N2 (metal as nucleophile)CH₃I, benzyl and allyl halides, acyl halidessecond order; accelerated in polar solvents; large negative ΔS‡; inversion at carbon; reactivity CH₃ > 1° > 2° and I > Br > Cl; often trans addition overall
Radical (chain or non-chain)some secondary and tertiary alkyl halidesracemisation at carbon; inhibited by radical traps
Ionic (stepwise)HCl, HBr in polar mediaprotonation of metal, then halide binding

Vaska’s complex summarises all of this:

$$\textit{trans}\text{-Ir}^{\text{I}}\text{Cl(CO)(PPh}_3)_2 + \text{H}_2 \longrightarrow \textit{cis}\text{-Ir}^{\text{III}}(\text{H})_2\text{Cl(CO)(PPh}_3)_2$$

16 → 18 electrons, Ir(I) d⁸ → Ir(III) d⁶, CN 4 → 6. With O₂ it gives a reversible η²-O₂ (peroxo) adduct. In every case ν(CO) increases after addition, because the oxidised metal back-donates less.

Binuclear oxidative addition

Two metals can each be oxidised by one unit:

$$2\,[\text{Co(CN)}_5]^{3-} + \text{H}_2 \longrightarrow 2\,[\text{HCo(CN)}_5]^{3-}$$

Each cobalt goes from Co(II), 17 electrons, to Co(III), 18 electrons.

Reductive Elimination

The reverse of oxidative addition, and the product-forming step of most catalytic cycles: two cis X-type ligands couple and leave.

$$\text{L}_n\text{M}^{(x+2)}(\text{A})(\text{B}) \longrightarrow \text{L}_n\text{M}^{(x)} + \text{A–B}$$

What favours reductive elimination:

  • the two groups cis (trans groups must first isomerise);
  • high oxidation state, electron-poor metal (π-acceptor co-ligands, positive charge);
  • bulky ligands and chelating phosphines with large bite angles that push the two groups together;
  • first-row and second-row metals more than third-row (Pd more than Pt);
  • easiest for H–H and C–H, harder for C–C (the directional sp³ orbitals of two alkyl groups overlap poorly in the transition state);
  • stereochemistry at carbon is retained.

Sometimes a ligand must dissociate first (to a three-coordinate or five-coordinate intermediate), or the metal is oxidised to make elimination faster.

Migratory Insertion

An unsaturated ligand inserts into a cis M–X bond (X = H, alkyl, aryl). Two kinds:

TypeLigandProductExample
1,1-insertionCO, isocyanide, SO₂the X group and the metal end up on the same atomM–CH₃ + CO → M–C(O)CH₃ (acyl)
1,2-insertionalkene, alkyneM and X end up on adjacent atomsM–H + CH₂=CH₂ → M–CH₂CH₃

CO insertion: the CH₃Mn(CO)₅ experiments

$$\text{CH}_3\text{Mn(CO)}_5 + \text{L} \longrightarrow \textit{cis}\text{-CH}_3\text{C(O)Mn(CO)}_4\text{L}$$

Three classic results:

  1. With ¹³CO as L, the product contains the label in a terminal CO cis to the acyl, never in the acyl group. The CO that becomes the acyl is one already on the metal.
  2. The reverse reaction, decarbonylation of cis-CH₃C(O)Mn(CO)₄(¹³CO), gives CH₃Mn(CO)₅ with the label distributed cis : trans : none = 2 : 1 : 1 relative to methyl. That distribution is expected only if the methyl group migrates (in the reverse, methyl moves to a vacant cis site); if the CO moved instead, no trans-labelled product could form.
  3. Kinetics follow a pre-equilibrium:
$$\text{CH}_3\text{Mn(CO)}_5 \underset{k_{-1}}{\overset{k_1}{\rightleftharpoons}} [\text{CH}_3\text{C(O)Mn(CO)}_4] \xrightarrow{k_2[\text{L}]} \text{CH}_3\text{C(O)Mn(CO)}_4\text{L}$$$$k_{\text{obs}} = \frac{k_1 k_2[\text{L}]}{k_{-1} + k_2[\text{L}]}$$

At low [L] the rate is first order in L; at high [L] it saturates at \( k_{\text{obs}} = k_1 \). A plot of \( 1/k_{\text{obs}} \) against \( 1/[\text{L}] \) is linear with intercept \( 1/k_1 \) and slope \( k_{-1}/(k_1k_2) \).

Other features of 1,1-insertion:

  • the migrating alkyl keeps its configuration (retention);
  • Lewis acids (AlCl₃, BF₃) bound to the acyl oxygen, coordinating solvents, bulky ligands and one-electron oxidation all accelerate it;
  • CO insertion is close to thermoneutral and reversible; double CO insertion to give M–C(O)C(O)R is rare, so CO alone does not polymerise this way.

Alkene insertion

1,2-insertion of an alkene into M–H or M–C is the reverse of β-hydride elimination:

  • syn addition: M and H (or R) add to the same face of the alkene, through a four-centre planar transition state;
  • for a terminal alkene the metal usually ends up on the terminal carbon (linear alkyl), for steric reasons;
  • insertion into M–H is usually exothermic and fast; repeated insertion into M–C is chain growth in Ziegler–Natta polymerisation.

Elimination Reactions

β-Hydride elimination

$$\text{L}_n\text{M–CH}_2\text{CH}_2\text{R} \longrightarrow \text{L}_n\text{M(H)}(\eta^2\text{-CH}_2{=}\text{CHR})$$

Requirements (recap from the hydrocarbon-ligand page): a β-hydrogen, a vacant site cis to the alkyl, and a syn-coplanar M–C–C–H arrangement. Repeated β-H elimination and reinsertion with the opposite regiochemistry moves a metal along a chain (alkene isomerisation, “chain walking”).

α-Elimination and α-abstraction

High-valent early-metal alkyls without β-hydrogens lose an α-hydrogen. α-Abstraction (one alkyl removes the α-H of another, leaving as alkane) gives Schrock alkylidenes, as in the neopentyl tantalum example. α-Elimination is also invoked in the Green–Rooney view of polymerisation.

σ-Bond Metathesis

d⁰ metals (Sc(III), Y(III), lanthanides(III), Ti(IV), Zr(IV)) have no d electrons for oxidative addition. They still activate H–H and C–H bonds, through a concerted four-centre (“kite”) transition state with no change in oxidation state:

$$\text{L}_n\text{M–R} + \text{R}'\text{–H} \longrightarrow [\text{four-centre TS}] \longrightarrow \text{L}_n\text{M–R}' + \text{R–H}$$

Classic examples: Cp₂Lu–CH₃ exchanges its methyl with ¹³CH₄ in solution (Watson); Cp₂Sc–CH₃ + H₂ → Cp*₂Sc–H + CH₄ (hydrogenolysis, the chain-transfer step for d⁰ polymerisation catalysts).

Attack on Coordinated Ligands

Nucleophilic attack

Coordination to a cationic or electron-poor metal makes a ligand electrophilic.

LigandNucleophileOutcome
COOH⁻M–COOH → M–H + CO₂ (the water–gas shift step)
CORLiacyl anion → Fischer carbene after O-alkylation
COMe₃NOCO₂ + NMe₃ lost; vacant site created
Alkene on Pd(II)H₂O, ROH, amineshydroxy- or alkoxypalladation (Wacker chemistry)
η³-Allyl on Pd(II)soft carbanions (malonate), aminesallylic substitution (Tsuji–Trost); soft nucleophiles attack the allyl face anti to Pd
η⁶-Arene on Cr(CO)₃carbanionscyclohexadienyl anion, then substituted arene
Fischer carbeneamines, alkoxidessubstitution at the carbene carbon

Davies–Green–Mingos rules predict where a nucleophile attacks an 18-electron cation carrying several π-hydrocarbon ligands (under kinetic control):

  1. Even polyenes (η², η⁴, η⁶) are attacked before odd polyenyls (η³, η⁵).
  2. Open (acyclic) ligands are attacked before closed (cyclic) ones.
  3. For an even open polyene, attack is at a terminal carbon.

Electrophilic attack

  • At the metal: protonation of electron-rich complexes gives hydrides (Cp₂WH₂ + H⁺ → [Cp₂WH₃]⁺).
  • At the M–C bond: protonolysis (M–R + HX → M–X + RH) and halogenolysis (M–R + X₂ → M–X + RX); stereochemistry at carbon can be retention or inversion depending on mechanism.
  • At a ligand: Friedel–Crafts acylation of ferrocene; alkylation of the oxygen of an acyl anion.

Memory Tricks

Don’t Confuse

ThisvsThat
Oxidative addition: OS +2, two new M–X bondsLigand association: OS unchanged, one new M–L bond
Concerted OA of H₂: cis product, retentionS_N2 OA of CH₃I: trans product often, inversion at C
Migratory insertion: OS unchanged, count −2Reductive elimination: OS −2, count −2
β-H elimination: needs β-H and cis vacancyα-abstraction: needs α-H, early high-valent metal
σ-bond metathesis (d⁰, no OS change)OA followed by RE (OS +2 then −2)
Indenyl effect (η⁵ → η³, associative)Dissociative CO loss from 18-e carbonyls

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Give the changes in oxidation state and electron count when Vaska’s complex adds H₂.
  2. Which step is the microscopic reverse of 1,2-insertion of ethylene into an M–H bond?
  3. Why can Cp*₂Lu–CH₃ not activate CH₄ by oxidative addition?
Answers
  1. Ir(I) → Ir(III), +2; 16 → 18 electrons, +2.
  2. β-Hydride elimination from the ethyl complex.
  3. Lu(III) is d⁰; there are no metal d electrons to give, and Lu(V) is not accessible. It reacts by σ-bond metathesis instead.

Level 2: Part B Style

Practice Q1. The reaction CH₃Mn(CO)₅ + PPh₃ → CH₃C(O)Mn(CO)₄(PPh₃) is an example of:

(a) oxidative addition (b) migratory insertion followed by ligand association (c) reductive elimination (d) σ-bond metathesis

Answer: (b)

Methyl migrates to a cis CO (16-electron acyl intermediate, OS unchanged at Mn(I)), then PPh₃ fills the vacancy (18 electrons). No change in oxidation state, so not OA or RE.

Practice Q2. The substitution of CO in Ni(CO)₄ by PPh₃ in hydrocarbon solvent has a rate independent of [PPh₃] and a positive ΔS‡. The mechanism is:

(a) associative (b) dissociative (c) interchange with S_N2 character (d) electron-transfer catalysed

Answer: (b)

Zero order in incoming ligand and positive ΔS‡ both point to rate-determining loss of CO from the 18-electron complex.

Practice Q3. A single enantiomer of the chiral alkyl bromide CH₃CHDBr adds to an electron-rich metal centre by the S_N2 pathway. The configuration at carbon in the alkyl complex is:

(a) retained (b) inverted (c) racemised (d) unchanged because the C–Br bond is not broken

Answer: (b) inverted

In the S_N2 pathway the metal attacks carbon from the back side and bromide leaves, as in an organic S_N2 reaction: inversion at carbon. A radical pathway would racemise; a concerted three-centre addition would retain.

Practice Q4. Which complex will most readily undergo associative CO substitution?

(a) CpRh(CO)₂ (b) (η⁵-C₉H₇)Rh(CO)₂ (indenyl) (c) Cr(CO)₆ (d) Fe(CO)₅

Answer: (b)

All four are 18-electron. Indenyl slips easily from η⁵ to η³ (the benzo ring regains full aromaticity), freeing a site for the incoming ligand: the indenyl effect. Cp slips far less readily; Cr(CO)₆ and Fe(CO)₅ substitute dissociatively.

Level 3: Part C Style

Practice Q5. For the reaction CH₃Mn(CO)₅ + L → CH₃C(O)Mn(CO)₄L, the observed pseudo-first-order rate constants (illustrative data) are \( k_{\text{obs}} = 2.0 \times 10^{-4} \) s⁻¹ at [L] = 0.010 M and \( 5.0 \times 10^{-4} \) s⁻¹ at [L] = 0.100 M. Assuming the pre-equilibrium mechanism above, \( k_1 \) and \( k_{-1}/k_2 \) are:

(a) \( 6.0 \times 10^{-4} \) s⁻¹ and 0.020 M (b) \( 5.0 \times 10^{-4} \) s⁻¹ and 0.010 M (c) \( 6.0 \times 10^{-4} \) s⁻¹ and 0.050 M (d) \( 2.0 \times 10^{-2} \) s⁻¹ and 0.020 M

Answer: (a)

Linearise: \( \dfrac{1}{k_{\text{obs}}} = \dfrac{1}{k_1} + \dfrac{k_{-1}}{k_1k_2}\cdot\dfrac{1}{[\text{L}]} \).

Data points \( (1/[\text{L}],\ 1/k_{\text{obs}}) \): (100 M⁻¹, 5000 s) and (10 M⁻¹, 2000 s).

Slope \( = (5000 - 2000)/(100 - 10) = 33.3 \) M s. Intercept \( = 2000 - 33.3 \times 10 = 1667 \) s, so \( k_1 = 6.0 \times 10^{-4} \) s⁻¹.

\( k_{-1}/k_2 = \text{slope} \times k_1 = 33.3 \times 6.0 \times 10^{-4} = 0.020 \) M.

Check: \( 6.0 \times 10^{-4} \times 0.010/(0.020 + 0.010) = 2.0 \times 10^{-4} \) s⁻¹, as observed. Note that 5.0 × 10⁻⁴ s⁻¹ is not yet the saturation value.

Practice Q6. cis-CH₃C(O)Mn(CO)₄(¹³CO) is decarbonylated. Assume a cis CO (any of the four, labelled or not) dissociates with equal probability and the methyl group then migrates to the vacant site. Which statements are correct?

  1. 25% of the product carries no ¹³C label.
  2. 25% of the product has ¹³CO trans to the methyl group.
  3. If CO migrated instead of methyl, the trans-labelled product would be absent.
  4. The observation of the trans-labelled product supports alkyl migration.

(a) 1, 2, 3 and 4 (b) 1 and 2 only (c) 3 and 4 only (d) 2 and 4 only

Answer: (a) 1, 2, 3 and 4

Put the acyl on +z and ¹³CO at +x; the four cis sites are ±x, ±y.

  • Vacancy at +x (the labelled CO lost): product unlabelled, probability 1/4.
  • Vacancy at −x: methyl moves to −x, label at +x is trans to methyl, probability 1/4.
  • Vacancy at ±y: methyl at ±y, label at +x cis, probability 2/4.

Distribution cis : trans : none = 2 : 1 : 1. If instead the acyl CO moved and methyl stayed at +z, the label (at +x) could only be cis to methyl or lost: no trans product. Observing trans product supports methyl migration.

Practice Q7. Vaska’s complex reacts with CH₃I in a polar solvent. Rate = k[Ir][CH₃I], ΔS‡ is large and negative, and the reaction is faster in more polar solvents. A chiral secondary alkyl halide reacts with inversion at carbon. Which conclusions are correct?

  1. The mechanism is S_N2 oxidative addition with the metal as nucleophile.
  2. The first intermediate is likely a cationic five-coordinate Ir(III) species with iodide as the counter-ion.
  3. ν(CO) of the product is lower than that of Vaska’s complex.
  4. Replacing PPh₃ by the more basic PMe₃ should speed up the reaction.

(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 4 only
  1. Correct. Second-order kinetics, polar transition state, inversion: S_N2.
  2. Correct. Ir attacks CH₃, iodide leaves to give [IrCl(CH₃)(CO)(PPh₃)₂]⁺ I⁻, which then binds iodide.
  3. Wrong. Ir(III) back-donates less than Ir(I); ν(CO) rises after oxidative addition.
  4. Correct. A more electron-rich metal is a better nucleophile.

Quick Revision Table

StepΔOS / ΔCN / Δe⁻Key point
Dissociation0 / −1 / −218-e; rate zero order in L; ΔS‡ > 0
Association0 / +1 / +216-, 17-e; ΔS‡ < 0; ring slip, bent NO
Oxidative addition+2 / +2 / +2electron-rich, heavy metal; concerted (cis), S_N2 (inversion), radical
Reductive elimination−2 / −2 / −2cis groups; high OS, bulky ligands; retention
1,1-insertion (CO)0 / −1 / −2alkyl migrates; retention; 2 : 1 : 1 label result
1,2-insertion (alkene)0 / −1 / −2syn; M to terminal C
β-H elimination0 / +1 / +2β-H, cis vacancy, syn-coplanar
α-Elimination0* / +1 / +2early metals; alkylidenes (*+2 if carbene counted as CR₂²⁻)
σ-Bond metathesis0 / 0 / 0d⁰; four-centre TS
Nucleophilic attack—even before odd, open before closed

Key Takeaways

What’s Next