Acids and bases in organic chemistry

Intuition [Beginner]

You already know that some molecules donate protons (acids) and others accept them (bases). In organic chemistry, the same idea applies but the scale shifts. The pKa values that matter range from about 5 (carboxylic acids) to about 50 (alkanes). This is a vastly wider range than the inorganic acids you see in general chemistry, and the reasons for the differences are specific to molecular structure.

The central question is: what makes one molecule a better acid than another? The answer always reduces to one thing: the stability of the conjugate base. When a molecule loses a proton ($\text{HA}\to {\text{A}}^{-}+{\text{H}}^{+}$), the resulting anion ${\text{A}}^{-}$ either tolerates its negative charge well (stable conjugate base, strong acid) or does not (unstable conjugate base, weak acid).

Two structural features stabilise negative charge on a conjugate base. First, resonance delocalisation: if the negative charge can spread across several atoms instead of sitting on one, the anion is more stable. Second, inductive electron withdrawal: if electronegative atoms near the negative charge pull electron density toward themselves through sigma bonds, the charge is partially dissipated.

Acetic acid (${\text{CH}}_3\text{COOH}$, pKa 4.76) is a much stronger acid than ethanol (${\text{CH}}_3{\text{CH}}_2\text{OH}$, pKa ~16) by about 11 pKa units. That is a factor of $10^{11}$ in the equilibrium constant. The difference comes from resonance in acetate: the negative charge on the conjugate base ${\text{CH}}_3{\text{COO}}^{-}$ is shared equally between two oxygen atoms. Ethoxide (${\text{CH}}_3{\text{CH}}_2{\text{O}}^{-}$) has no such delocalisation -- the charge sits entirely on one oxygen.

Visual [Beginner]

Picture two molecules losing a proton and compare what happens to the negative charge.

Ethanol losing a proton. The O-H bond breaks. The oxygen keeps both electrons. The resulting ethoxide ion ${\text{CH}}_3{\text{CH}}_2{\text{O}}^{-}$ has a full negative charge on one oxygen. That charge has nowhere to go -- the adjacent carbon and hydrogens do not provide delocalisation pathways.

Acetic acid losing a proton. The O-H bond breaks. The resulting acetate ion ${\text{CH}}_3{\text{COO}}^{-}$ has the negative charge on one oxygen, but the adjacent C=O double bond can shift its electrons. The double bond becomes a single bond, and a new pi bond forms to the negatively charged oxygen. The result: the negative charge is shared equally between both oxygens. Both C-O bonds in acetate have identical length (halfway between single and double), a fact confirmed by X-ray crystallography.

The more resonance structures a conjugate base has, the more stable it is, and the stronger the parent acid.

Worked example [Beginner]

Why is acetic acid (pKa 4.76) more acidic than ethanol (pKa ~16)?

Step 1. Write both dissociation reactions.

Acetic acid: ${\text{CH}}_3\text{COOH}\rightleftharpoons {\text{CH}}_3{\text{COO}}^{-}+{\text{H}}^{+}$

Ethanol: ${\text{CH}}_3{\text{CH}}_2\text{OH}\rightleftharpoons {\text{CH}}_3{\text{CH}}_2{\text{O}}^{-}+{\text{H}}^{+}$

Step 2. Draw the conjugate bases.

Acetate (${\text{CH}}_3{\text{COO}}^{-}$): the negative charge on one oxygen is delocalised to the second oxygen through resonance. Two equivalent resonance structures. The actual structure is a resonance hybrid with the charge spread over both oxygens.

Ethoxide (${\text{CH}}_3{\text{CH}}_2{\text{O}}^{-}$): the negative charge is localised on one oxygen. No resonance structures.

Step 3. Compare stabilities.

Acetate is more stable because the negative charge is delocalised over two electronegative oxygen atoms instead of being concentrated on one. A more stable conjugate base means the equilibrium lies further to the right (more dissociation), which means a lower pKa.

Step 4. Quantify.

$\Delta \text{pKa}=16-4.76=11.24$. The equilibrium constant for acetic acid dissociation is $10^{11.24}\approx 1.7\times 10^{11}$ times larger than for ethanol. Resonance stabilisation of the conjugate base accounts for this enormous difference.

The inductive effect of the carbonyl group also contributes: the C=O bond pulls electron density away from the O-H bond through the carbon skeleton, weakening the O-H bond and facilitating proton loss. But resonance is the dominant factor -- the 11-unit pKa difference is far larger than what induction alone produces (a typical inductive effect from one carbonyl is 2-3 pKa units).

Check your understanding [Beginner]

Formal definition [Intermediate+]

The Brønsted-Lowry definition of acids and bases extends directly to organic molecules: an acid is a proton donor, a base is a proton acceptor. The strength of an acid is quantified by its acid dissociation constant $K_a$ and its pKa:

$$K_a=\frac{[{\text{A}}^{-}][{\text{H}}^{+}]}{[\text{HA}]},\text{p}{K}_a=-{\log }_{10}{K}_a.$$

Lower pKa = stronger acid. The pKa scale for organic compounds spans approximately 50 units: from about $-10$ (superacids like triflic acid) to about $+50$ (alkane C-H bonds). Each unit represents a factor of 10 in $K_a$.

Conjugate-base stability and acid strength. The thermodynamic cycle for acid dissociation in solution is:

$${\text{HA}}_{(aq)}⟶{\text{H}}_{(aq)}^{+}+{\text{A}}_{(aq)}^{-},\Delta G^{\circ }=-RT\ln K_a.$$

The free energy of dissociation is dominated by two terms: the bond-dissociation energy of the H-A bond (unfavourable) and the solvation energy of ${\text{A}}^{-}$ (favourable). Anything that stabilises ${\text{A}}^{-}$ lowers $\Delta G^{\circ }$ and increases $K_a$ (lowers pKa). The three principal mechanisms for conjugate-base stabilisation are:

1. Resonance delocalisation. The negative charge on ${\text{A}}^{-}$ is distributed over two or more atoms via pi-system delocalisation. Each equivalent resonance structure contributes to the hybrid, and the energy lowering is proportional to the number of significant contributing structures. Quantitative measure: the resonance stabilisation energy is estimated from the difference between the observed pKa and the pKa predicted by induction alone.

For carboxylic acids vs alcohols, the resonance stabilisation of carboxylate (${\text{RCOO}}^{-}$) vs alkoxide (${\text{RO}}^{-}$) accounts for approximately 10 pKa units of the observed 11-unit difference.

2. Inductive (field) effects. Electron-withdrawing groups (EWGs) near the site of negative charge stabilise the anion by pulling electron density through sigma bonds or through space (field effect). The magnitude of the inductive effect decays rapidly with distance: each additional carbon between the EWG and the charge attenuates the effect by roughly a factor of 2-3. The effect is additive: two equivalent EWGs produce approximately twice the stabilisation of one.

Quantitative framework: the Hammett equation (see Master tier) parameterises substituent effects on acidity as $\log (K/K_0)=\rho \sigma$, where $\sigma$ is the substituent constant (positive for electron-withdrawing, negative for electron-donating) and $\rho$ is the reaction constant (sensitivity to substituent effects).

3. Hybridisation effects. An sp-hybridised carbon has 50% s-character (vs 33% for sp${}^2$, 25% for sp${}^3$). Higher s-character means the electrons are held closer to the nucleus, making the carbon more electronegative. This stabilises adjacent negative charge. Consequently, C-H acidity increases with s-character: sp C-H (pKa ~25, terminal alkynes) > sp${}^2$ C-H (pKa ~44, alkenes) > sp${}^3$ C-H (pKa ~50, alkanes).

Equilibrium in organic reactions. Many organic reactions are acid-base equilibria. Predicting the direction of equilibrium requires comparing the pKa values of the acid on each side:

$${\text{HA}}_1+{\text{A}}_2^{-}\rightleftharpoons {\text{A}}_1^{-}+{\text{HA}}_2.$$

The equilibrium favours the side with the weaker acid (higher pKa). If $\text{p}{K}_a({\text{HA}}_1)<\text{p}{K}_a({\text{HA}}_2)$, the equilibrium lies to the right. The equilibrium constant is $K=10^{\text{p}{K}_a({\text{HA}}_2)-\text{p}{K}_a({\text{HA}}_1)}$.

Counterexamples to common slips

• "Resonance always makes an acid stronger." Resonance in the conjugate base makes an acid stronger. Resonance in the undissociated acid (e.g., intramolecular hydrogen bonding that stabilises the acid form) makes it weaker. The effect depends on which form is more stabilised.

• "Electronegativity alone explains acid strength." Electronegativity of the atom bearing the negative charge is one factor, but within a row of the periodic table the trend is not monotonic: O-H acids (pKa ~15-16 for alcohols) are weaker than C-H acids of sp carbons (pKa ~25) despite oxygen being more electronegative. Resonance and hybridisation override the electronegativity trend in these cases.

• "All functional groups have a single pKa." Polyprotic acids have multiple pKa values (one per dissociable proton). Phosphoric acid (${\text{H}}_3{\text{PO}}_4$) has three pKa values (2.15, 7.20, 12.35). Each successive dissociation is harder because the conjugate base is increasingly negatively charged.

Key theorem with proof [Intermediate+]

Proposition (Equilibrium direction from pKa comparison). For the proton-transfer equilibrium ${\text{HA}}_1+{\text{A}}_2^{-}\rightleftharpoons {\text{A}}_1^{-}+{\text{HA}}_2$ in dilute aqueous solution at 25 ${}^{\circ }$C, the equilibrium constant is $K_{eq}=K_{a1}/K_{a2}=10^{\text{p}{K}_a({\text{HA}}_2)-\text{p}{K}_a({\text{HA}}_1)}$.

Proof. Write the two acid-dissociation equilibria and their constants:

$$K_{a1}=\frac{[{\text{A}}_1^{-}][{\text{H}}^{+}]}{[{\text{HA}}_1]},K_{a2}=\frac{[{\text{A}}_2^{-}][{\text{H}}^{+}]}{[{\text{HA}}_2]}.$$

The proton-transfer equilibrium is obtained by adding the forward dissociation of ${\text{HA}}_1$ and the reverse association of ${\text{A}}_2^{-}$ with ${\text{H}}^{+}$:

$${\text{HA}}_1\rightleftharpoons {\text{A}}_1^{-}+{\text{H}}^{+}(K_{a1})$$ $${\text{H}}^{+}+{\text{A}}_2^{-}\rightleftharpoons {\text{HA}}_2(1/K_{a2})$$

Adding: ${\text{HA}}_1+{\text{A}}_2^{-}\rightleftharpoons {\text{A}}_1^{-}+{\text{HA}}_2$. The equilibrium constant for the sum is the product:

$$K_{eq}=K_{a1}\times \frac{1}{K_{a2}}=\frac{K_{a1}}{K_{a2}}.$$

Taking negative logarithms: ${\log }_{10}{K}_{eq}={\log }_{10}{K}_{a1}-{\log }_{10}{K}_{a2}=\text{p}{K}_a({\text{HA}}_2)-\text{p}{K}_a({\text{HA}}_1)$.

So $K_{eq}=10^{\Delta \text{pKa}}$ where $\Delta \text{pKa}=\text{p}{K}_a({\text{HA}}_2)-\text{p}{K}_a({\text{HA}}_1)$. When ${\text{HA}}_1$ is the stronger acid ($\text{p}{K}_a({\text{HA}}_1)<\text{p}{K}_a({\text{HA}}_2)$), $\Delta \text{pKa}>0$ and $K_{eq}>1$: the equilibrium favours the products (weaker acid side). $■$

Corollary. A proton transfer from an acid of pKa $a$ to the conjugate base of an acid of pKa $b$ proceeds essentially to completion when $b-a>4$ (i.e., $K_{eq}>10^4$). This is the practical rule for predicting whether a given base will deprotonate a given acid in the laboratory.

Bridge. The pKa-comparison proposition builds toward 15.04.02 pending where leaving-group ability is ranked by the pKa of the conjugate acid of the leaving group, and appears again in 15.07.01 where the acidity of alpha-hydrogens next to carbonyl groups determines enolate formation. The foundational reason is that proton transfer is the simplest bond-breaking and bond-making event in organic chemistry, and this is exactly the thermodynamic machinery that predicts its direction. Putting these together, every acid-base equilibrium in organic synthesis -- from choosing a base for deprotonation to predicting the products of a proton-transfer step in a multistep mechanism -- reduces to a single pKa comparison.

Exercises [Intermediate+]

The Hammett equation and linear free-energy relationships [Master]

The Hammett equation provides a quantitative framework for predicting the effect of substituents on acid-base equilibria and reaction rates. For the ionisation of substituted benzoic acids ^{[Hammett 1937]}:

$${\log }_{10}\left(\frac{K}{K_0}\right)=\rho \sigma$$

where $K$ is the equilibrium constant for the substituted acid, $K_0$ for the unsubstituted (benzoic acid, $\text{p}{K}_a=4.20$), $\sigma$ is the substituent constant (positive for electron-withdrawing groups, negative for electron-donating), and $\rho$ is the reaction constant ($\rho =1.00$ for benzoic acid ionisation by definition).

Selected $\sigma$ values: ${\text{-NMe}}_2$ (-0.83), ${\text{-OCH}}_3$ (-0.27), ${\text{-CH}}_3$ (-0.17), H (0.00), $\text{-F}$ (+0.06), $\text{-Cl}$ (+0.23), ${\text{-CF}}_3$ (+0.54), $\text{-CN}$ (+0.66), ${\text{-NO}}_2$ (+0.78).

The Hammett equation is a linear free-energy relationship (LFER): it states that $\Delta \Delta G$ (the change in free-energy change caused by a substituent) is proportional to $\sigma$. Since $\Delta G^{\circ }=-RT\ln K$, the relation $\log (K/K_0)=\rho \sigma$ translates to $\Delta \Delta G^{\circ }=-2.303RT\rho \sigma$. At 298 K, $2.303RT=5.71$ kJ/mol, so one unit of $\sigma$ corresponds to 5.71 kJ/mol of free-energy perturbation when $\rho =1$.

The slope $\rho$ measures the sensitivity of the reaction to substituent effects. For acid-base reactions, $\rho$ is always positive because electron-withdrawing groups stabilise the negative charge on the conjugate base. For reactions where a positive charge develops (e.g., SN1 ionisation in 15.04.02 pending), $\rho$ is negative because electron-donating groups stabilise the positive charge. The absolute magnitude $|\rho |$ measures how closely the transition state resembles the ionised product: $|\rho |\approx 1$ means full charge development; $|\rho |\approx 0$ means the reaction is insensitive to electronic effects.

The standard Hammett $\sigma$ constants are derived from meta-substituted benzoic acids, where the substituent communicates with the reactive centre primarily through inductive and field effects, with minimal direct resonance interaction. This choice is deliberate: meta positions on a benzene ring do not allow conjugation with the reaction site, so ${\sigma }_{\text{meta}}$ isolates the inductive component of the substituent effect.

When the substituent can interact with the reactive centre through resonance, the standard $\sigma$ values fail. Two extended scales correct for this. The Brown-Okamoto ${\sigma }^{+}$ constants ^{[Brown Okamoto 1958]} account for direct resonance donation from the substituent to a cationic centre. For substituents that are strong resonance donors (e.g., $p$-OCH${}_3$, ${\sigma }^{+}=-0.78$ vs $\sigma =-0.27$), ${\sigma }^{+}$ is much more negative than $\sigma$ because the para-methoxy group can stabilise a developing positive charge by direct mesomeric donation. The ${\sigma }^{+}$ scale is the appropriate choice for reactions involving electrophilic aromatic substitution or carbocation formation.

The ${\sigma }^{-}$ constants account for direct resonance withdrawal from an anionic centre. For $p$-NO${}_2$, ${\sigma }^{-}=+1.27$ vs $\sigma =+0.78$: the para-nitro group withdraws electron density by direct resonance from a developing negative charge far more effectively than the meta constant suggests. The ${\sigma }^{-}$ scale applies to phenolate ionisation and other reactions that develop negative charge conjugated to the substituent.

The Yukawa-Tsuno equation introduces a resonance correction parameter $r$ that interpolates between $\sigma$ and ${\sigma }^{+}$:

$${\log }_{10}(K/K_0)=\rho [\sigma +r({\sigma }^{+}-\sigma )].$$

When $r=0$, the equation reduces to the standard Hammett relation (no resonance correction). When $r=1$, full ${\sigma }^{+}$ weighting applies. Intermediate $r$ values ($0<r<1$) arise when the transition state has partial positive-charge development at the site conjugated to the substituent. The Yukawa-Tsuno parameter $r$ is thus a quantitative measure of the degree of resonance demand at the transition state.

For acid-base equilibria specifically, the $r$ value is near zero for meta-substituted systems (no direct resonance to the reaction site) and near 0.5-0.7 for para-substituted phenols and anilines, where the conjugate base can communicate with the substituent through the aromatic ring. The ability to quantify the resonance contribution separately from the inductive contribution is one of the principal achievements of the extended Hammett framework.

The Taft equation extends the LFER approach to aliphatic systems where steric effects are substantial. Taft ^{[Taft 1952]} separated polar (${\sigma }^{\ast }$) and steric ($E_s$) contributions by comparing acid-catalysed and base-catalysed hydrolysis rates of esters. The acid-catalysed rates are sensitive primarily to steric effects (the tetrahedral transition state resembles the tetrahedral ester in charge distribution), while the base-catalysed rates are sensitive to both steric and polar effects. The difference isolates the polar component. Selected $E_s$ values: H (0.00), CH${}_3$ (0.00), C${}_2$H${}_5$ ($-0.07$), $i$-Pr ($-0.47$), $t$-Bu ($-1.54$). The large negative $E_s$ for tert-butyl quantifies the severe steric inhibition of approach to the reaction centre.

Acid-base catalysis [Master]

Many organic reactions are catalysed by acids or bases. The rate of a Brønsted-acid-catalysed reaction depends on the pKa of the catalyst through the Brønsted catalysis law ^{[Bronsted 1923]}:

$${\log }_{10}{k}_{\text{HA}}=\alpha \text{p}{K}_a(\text{HA})+C,$$

where $k_{\text{HA}}$ is the rate constant for the catalysed reaction, $\alpha$ is the Brønsted coefficient ($0<\alpha <1$), and $C$ is a constant. A large $\alpha$ (close to 1) indicates that proton transfer is nearly complete in the transition state; a small $\alpha$ (close to 0) indicates the proton is barely transferred. The Brønsted coefficient is a structural probe of the transition state, analogous to the Hammett $\rho$ value: it measures the degree of proton transfer at the rate-determining transition state.

Specific vs general acid catalysis. Specific acid catalysis operates through the solvated proton ${\text{H}}_3{\text{O}}^{+}$ alone: the rate depends on pH but not on the identity or concentration of the buffer. The mechanism involves rapid pre-equilibrium protonation of the substrate followed by rate-determining reaction of the protonated intermediate. General acid catalysis operates through the undissociated acid molecule: the rate depends on buffer concentration at constant pH because the acid molecule participates directly in the transition state. The two are distinguished experimentally by varying buffer concentration while holding pH constant.

In specific acid catalysis, the mechanism is:

$$\text{S}+{\text{H}}_3{\text{O}}^{+}\stackrel{\text{fast}}{\to }{\text{SH}}^{+}\stackrel{k(\text{slow})}{\to }\text{products},$$

so the rate law is $\text{rate}=k[{\text{SH}}^{+}]\approx k(K_{\text{prot}}[\text{S}][{\text{H}}_3{\text{O}}^{+}])$, proportional to $[{\text{H}}_3{\text{O}}^{+}]$ alone.

In general acid catalysis, the acid molecule is a co-reactant in the rate-determining step:

$$\text{S}+\text{HA}\stackrel{k(\text{slow})}{\to }\text{products}+{\text{A}}^{-},$$

and the rate depends on $[\text{HA}]$. Increasing the buffer concentration at fixed pH increases $[\text{HA}]$ and increases the rate -- the diagnostic signature.

Brønsted base catalysis follows an analogous law with coefficient $\beta$:

$${\log }_{10}{k}_{\text{B}}=-\beta \text{p}{K}_a({\text{BH}}^{+})+C^{\prime }.$$

The coefficient $\beta$ measures the degree of proton removal from the substrate at the transition state. For a concerted proton transfer, $\alpha +\beta \approx 1$ (the proton is being transferred from catalyst to substrate or vice versa).

A classic application is the hydrolysis of acetals and ketals. Acetal hydrolysis proceeds by specific acid catalysis: the rate depends only on pH and is independent of buffer identity, indicating that the protonation of the acetal oxygen is fast and reversible, followed by rate-determining C-O bond cleavage of the protonated intermediate. In contrast, the hydrolysis of orthoesters shows general acid catalysis: the undissociated acid participates in the transition state, protonating the substrate and assisting leaving-group departure simultaneously.

Enzymatic acid-base catalysis represents the most sophisticated manifestation of these principles. Serine proteases (chymotrypsin, trypsin, elastase) employ a catalytic triad of Ser-His-Asp residues. The mechanism involves two half-reactions, each requiring acid-base catalysis. In the acylation step, histidine acts as a general base, abstracting the proton from serine's hydroxyl to generate the alkoxide nucleophile that attacks the peptide carbonyl. In the deacylation step, histidine acts as a general acid, protonating the leaving-group oxygen as the tetrahedral intermediate collapses. The aspartate orients and stabilises the protonated histidine through a hydrogen-bond network.

The $\alpha$-chymotrypsin mechanism has been studied with Brønsted-type analysis using substrate analogues and mutant enzymes. The measured Brønsted coefficients ($\alpha \approx 0.2$--$0.3$ for the acylation step) indicate that proton transfer from serine to histidine is only partial at the transition state: the serine oxygen has begun to lose its proton but the proton is still closer to serine than to histidine. This is the structural information that the Brønsted coefficient encodes: the degree of proton progress along the transfer coordinate at the rate-determining transition state.

Quantitative prediction of pKa [Master]

Predicting pKa from molecular structure is one of the central quantitative problems in physical organic chemistry. The Hammett equation handles substituted aromatics, but many molecules lack an aromatic reference scaffold. The Taft equation and its descendants extend quantitative pKa prediction to aliphatic and alicyclic systems.

The Taft inductive parameter ${\sigma }_I$. Taft separated the inductive and resonance components of substituent effects by defining ${\sigma }_I$ (inductive) and ${\sigma }_R$ (resonance) constants. For a substituent X:

$${\sigma }_{\text{para}}={\sigma }_I+{\sigma }_R,{\sigma }_{\text{meta}}\approx {\sigma }_I+0.33{\sigma }_R.$$

The meta position transmits roughly one-third of the resonance effect relative to para, because the meta connectivity does not allow direct conjugation with the reaction site. By measuring both ${\sigma }_{\text{meta}}$ and ${\sigma }_{\text{para}}$ for a range of reactions, ${\sigma }_I$ and ${\sigma }_R$ can be extracted. Selected values: for $-{\text{NO}}_2$, ${\sigma }_I=+0.58$, ${\sigma }_R=+0.19$ (both inductive and resonance withdrawal); for $-{\text{OCH}}_3$, ${\sigma }_I=+0.30$, ${\sigma }_R=$ $-0.55$ (inductive withdrawal but resonance donation).

The dual-substituent-parameter (DSP) equation uses both:

$$\log (K/K_0)={\rho }_I{\sigma }_I+{\rho }_R{\sigma }_R.$$

For a reaction where the transition state develops negative charge, both ${\rho }_I$ and ${\rho }_R$ are positive (electron-withdrawing substituents accelerate by both pathways). The ratio ${\rho }_R/{\rho }_I$ measures the relative importance of resonance vs inductive transmission to the reaction site -- a structural fingerprint of the transition state.

For aliphatic acids, the inductive effect dominates and the Hammett equation in its standard form is unreliable (no aromatic ring to define the $\sigma$ scale). The Taft equation for aliphatic acidity is:

$$\text{p}{K}_a=\text{p}{K}_{a,0}+{\rho }^{\ast }{\sigma }^{\ast },$$

where ${\sigma }^{\ast }$ is the aliphatic substituent constant and ${\rho }^{\ast }$ is the aliphatic reaction constant. Selected ${\sigma }^{\ast }$ values: H (0.49), CH${}_3$ (0.00), C${}_2$H${}_5$ ($-0.10$), $i$-Pr ($-0.19$), $t$-Bu ($-0.30$), CH${}_3$CO ($+0.60$), CN ($+1.30$). The negative values for alkyl groups reflect their electron-donating character in aliphatic systems (hyperconjugation). The positive values for electron-withdrawing groups are comparable in magnitude to the aromatic $\sigma$ constants.

Computational pKa prediction. The thermodynamic cycle for aqueous pKa connects the gas-phase deprotonation energy to the solution-phase value through solvation free energies:

$$\Delta G_{\text{aq}}^{\circ }(\text{HA}\to {\text{A}}^{-}+{\text{H}}^{+})=\Delta G_{\text{gas}}^{\circ }(\text{HA}\to {\text{A}}^{-}+{\text{H}}^{+})+\Delta G_{\text{solv}}({\text{A}}^{-})-\Delta G_{\text{solv}}(\text{HA})-\Delta G_{\text{solv}}({\text{H}}^{+}).$$

The gas-phase deprotonation energy $\Delta G_{\text{gas}}^{\circ }$ is computed directly from density-functional theory (DFT) by optimising the geometries of HA and ${\text{A}}^{-}$ in the gas phase and computing their electronic energies with a thermal correction. Modern functionals (B3LYP, M06-2X, $\omega$B97X-D) give gas-phase deprotonation energies accurate to 4--8 kJ/mol ($\sim$1--2 pKa units) for small organic molecules.

The solvation free energies are the bottleneck. Implicit solvation models (PCM, COSMO, SMD) compute $\Delta G_{\text{solv}}$ by surrounding the solute with a continuum dielectric. For neutral molecules, implicit models are accurate to 2--4 kJ/mol. For anions, the accuracy degrades to 4--10 kJ/mol because the strong electric field of the anion demands more accurate treatment of the solute-solvent interface. The proton solvation free energy $\Delta G_{\text{solv}}({\text{H}}^{+})$ is taken as an experimental value ($-265.9$ kcal/mol, based on the cluster-pair extrapolation method).

The net accuracy of DFT + implicit-solvent pKa prediction is typically 1--2 pKa units for molecules without strong intramolecular hydrogen bonding or ion-pairing effects. For molecules where explicit solvation matters (e.g., carboxylic acids where the first solvation shell strongly stabilises the anion), microsolvation clusters (adding 2--4 explicit water molecules before the continuum) improve accuracy to within 0.5--1 pKa unit.

Linear-regression and machine-learning approaches achieve higher accuracy for specific classes of compounds. The Marvin pKa plugin, ACD/pKa, and similar tools use databases of 10,000+ experimental pKa values to train fragment-based or graph-convolution models. These methods predict pKa to within 0.3--0.5 units for drug-like molecules within their training domain, but extrapolate poorly to novel scaffolds. The thermodynamic-cycle approach from DFT is more general but less precise; the two approaches are complementary.

Borderline acid-base phenomena [Master]

The pKa values discussed at Intermediate tier cover the range from strong organic acids (carboxylic acids, pKa ~5) to moderately weak acids (phenols, pKa ~10; alcohols, pKa ~16). But organic chemistry involves acids far outside this range, and the acid-base framework extends in instructive ways at both extremes.

Carbon acids and the pKa > 20 regime. Molecules whose acidic proton is on carbon rather than oxygen or nitrogen are called carbon acids. Their pKa values range from ~9 (beta-diketones) to ~50 (alkanes). The enormous range reflects the degree of stabilisation of the resulting carbanion. The pKa of acetone ($\sim$19.2) is far lower than that of ethane ($\sim$50) because the conjugate base of acetone (the enolate) is stabilised by resonance delocalisation of the negative charge onto the electronegative oxygen. Terminal alkynes (pKa ~25) owe their relative acidity to the high s-character (50%) of the sp-hybridised carbon, which holds the electron pair in the conjugate base closer to the nucleus.

The carbon-acid regime matters because many important synthetic transformations require deprotonation of C-H bonds. Enolate formation from ketones (pKa ~20), alkylation of terminal alkynes (pKa ~25), and deprotonation of malonate esters (pKa ~13) all depend on choosing a base strong enough to drive the equilibrium. The practical rule ($\Delta \text{pKa}>4$ for quantitative conversion) dictates that deprotonating a ketone (pKa ~20) requires a base whose conjugate acid has pKa > 24. Lithium diisopropylamide (LDA, conjugate-acid pKa ~36) satisfies this with a 16-unit margin, driving $K_{eq}=10^{16}$.

Kinetic vs thermodynamic acidity. For many carbon acids, the rate of proton transfer does not correlate with the thermodynamic pKa. Proton transfer to or from carbon involves rehybridisation of the carbon (e.g., sp${}^3$ $\to$ sp${}^2$ for enolate formation), which has a significant activation barrier unrelated to the equilibrium constant. Nitroalkanes illustrate this dramatically: nitromethane (pKa 10.2) is thermodynamically as acidic as phenol, but the rate of proton transfer from nitromethane to hydroxide is slow enough to measure on a laboratory timescale, while proton transfer from phenol to hydroxide is essentially diffusion-controlled ($k\sim 10^{10}{\text{M}}^{-1}{\text{s}}^{-1}$).

The discrepancy is captured by the Brønsted coefficients $\alpha$ and $\beta$ for the proton-transfer reaction. When $\alpha =\beta =0.5$, the transition state is symmetric and the rate correlates smoothly with pKa. When $\alpha$ or $\beta$ deviate strongly from 0.5, the transition state is asymmetric (the proton is much closer to one partner), and rate-pKa correlations break down. Carbon acids typically have $\beta <0.5$ (the proton is closer to carbon than to the base at the transition state), reflecting the high intrinsic barrier to carbanion formation. This is the physical-organic basis for the observation that thermodynamic acidity does not always predict kinetic reactivity.

Kinetic isotope effects on proton transfer. Replacing hydrogen with deuterium at the acidic position slows proton transfer because the zero-point energy of the X-H bond is higher than that of the X-D bond, and this difference is partially retained in the transition state. The primary kinetic isotope effect KIE $=k_H/k_D$ ranges from 2--3 for most proton transfers to values as high as 7--8 for symmetric or near-symmetric proton transfers where the transition state has the proton equally shared between donor and acceptor.

The Bell-Evans-Polanyi model treats the KIE as a function of the proton position at the transition state. For a very early transition state (proton still on the donor), the X-H and X-D stretching frequencies are barely perturbed and the KIE is near unity. For a symmetric transition state (proton halfway between donor and acceptor), the stretching mode converts to a low-frequency translational mode with much less ZPE, maximising the KIE. The maximum KIE ($\sim$7--10 at room temperature) is predicted for thermoneutral proton transfer ($\Delta \text{pKa}=0$), where the transition state is most symmetric.

This isotope effect is exploited in mechanistic studies: a large KIE ($>2$) is evidence that the C-H(D) bond is being broken in the rate-determining step. A small KIE ($\sim 1$) means proton transfer occurs after the rate-determining step or is not involved. Enzyme mechanisms are routinely probed by measuring the KIE on deuterated substrates; the magnitude of the KIE constrains the geometry of the transition state.

Superacids and the Hammett acidity function. The pH scale becomes undefined in concentrated acid solutions because the activity of water deviates strongly from unity and the concentration of free ${\text{H}}_3{\text{O}}^{+}$ no longer represents the effective proton-donating power of the medium. The Hammett acidity function $H_0$ extends the acidity scale into the superacid regime by using indicator bases that are neutral molecules (not hydroxide ions):

$$H_0=\text{p}{K}_a({\text{BH}}^{+})+{\log }_{10}\left(\frac{[\text{B}]}{[{\text{BH}}^{+}]}\right).$$

A series of weakly basic indicators (nitroanilines, nitroanisoles) with known pKa values of their conjugate acids is used to establish $H_0$ for strongly acidic media. Pure sulfuric acid has $H_0=-12$; magic acid (${\text{FSO}}_3{\text{H-SbF}}_5$) has $H_0\approx -23$. At these acidities, even alkanes can be protonated: the carbocation chemistry that normally requires reactive intermediates at normal acid strengths becomes thermodynamically accessible.

The concept of superacidity connects this unit to carbocation chemistry in 15.04.02 pending. In a superacid medium, the tertiary butyl cation ($t$-Bu${}^{+}$), which is a reactive intermediate with nanosecond lifetime in normal solvents, becomes a stable species observable by NMR. The Olah carbocation studies ^{[Olah 1972]}, conducted in antimony pentafluoride / fluorosulfonic acid mixtures, provided direct spectroscopic evidence for carbocation structures that had previously been inferred only from kinetic and stereochemical data. The bridge is that the thermodynamic stability of carbocations is the mirror image of conjugate-base stability in acid-base chemistry: the same electronic effects (induction, hyperconjugation, resonance) that stabilise anions destabilise cations, and vice versa.

Advanced results [Master]

Theorem 1 (Hammett correlation for phenol ionisation). The ionisation of para-substituted phenols follows $\log (K_X/K_H)=\rho {\sigma }^{-}$ with $\rho \approx +2.2$ in water at 25 ${}^{\circ }$C, using the ${\sigma }^{-}$ scale to account for direct resonance withdrawal from the anionic phenoxide oxygen. The large $\rho$ value (more than double that for benzoic acid ionisation) indicates that phenolate ionisation is more sensitive to substituent effects than benzoate ionisation, because the phenolate negative charge is closer to the ring and communicates more directly with the substituent.

Theorem 2 (Brønsted $\alpha$ as a TS structural probe). For a Brønsted-acid-catalysed reaction with coefficient $\alpha$, the proton is transferred by a fraction $\alpha$ from the acid to the substrate at the transition state. When $\alpha \approx 0$, the transition state resembles the reactants (proton entirely on the catalyst); when $\alpha \approx 1$, it resembles the products (proton fully transferred). This is the Marcus-theory interpretation: $\alpha$ is the position of the transition state along the proton-transfer coordinate.

Theorem 3 (Curtin-Hammett principle for acid-base equilibria). When two conformers of a molecule have different pKa values, the product of deprotonation is determined not by the relative populations of the conformers but by the relative energies of the transition states for proton removal from each conformer. The more rapidly deprotonated conformer determines the product, regardless of whether it is the major conformer at equilibrium.

Theorem 4 (Effective molarity in intramolecular acid-base catalysis). Intramolecular general acid catalysis (where the acid group and the reaction site are in the same molecule) is accelerated relative to the intermolecular analogue by an effective molarity (EM) of 10--1000 M, depending on ring size. The EM measures the entropic advantage of having the catalytic group pre-organised near the reaction site.

Theorem 5 (Aromaticity and anomalous acidity). Cyclopentadiene (pKa ~16) is 34 pKa units more acidic than a typical sp${}^3$ C-H (pKa ~50) because its conjugate base (the cyclopentadienyl anion) has 6 pi electrons in a planar, cyclic, fully conjugated 5-membered ring, satisfying Hückel's $(4n+2)$ rule with $n=0$. This aromatic stabilisation of the conjugate base is the most dramatic example of structure-driven acid enhancement in organic chemistry, exceeding even the resonance stabilisation of carboxylate.

Theorem 6 (Gas-phase vs solution-phase acidity inversion). In the gas phase, the acidity order of halogenated acetic acids is ${\text{FCH}}_2\text{COOH}>{\text{ClCH}}_2\text{COOH}>{\text{BrCH}}_2\text{COOH}>{\text{ICH}}_2\text{COOH}$, following the electronegativity of the halogen. In aqueous solution, the order reverses to ${\text{ICH}}_2\text{COOH}>{\text{BrCH}}_2\text{COOH}>{\text{ClCH}}_2\text{COOH}>{\text{FCH}}_2\text{COOH}$ because the larger, more polarisable halides interact more favourably with the solvent through dispersion forces, stabilising the conjugate base in solution.

Synthesis. The foundational reason that the Hammett equation works is that substituent effects on free energy are approximately additive and transferable across reactions, and this is exactly the property that identifies acid-base chemistry as a thermodynamic yardstick for all of organic reactivity. The central insight is that pKa encodes the full electronic structure of a molecule in a single number, and the Taft and Yukawa-Tsuno extensions decompose that encoding into inductive, resonance, and steric contributions. Putting these together with the Brønsted catalysis law, the degree of proton transfer at a transition state ($\alpha$ or $\beta$) becomes a structural probe connecting equilibrium acid-base thermodynamics to kinetic reactivity. The bridge is between the equilibrium pKa values treated at Intermediate tier and the transition-state structures probed by KIE, Brønsted coefficients, and Hammett $\rho$ values at Master tier. This pattern recurs across all of physical organic chemistry: the free-energy perturbation from a substituent is the same quantity whether it appears in an equilibrium constant, a rate constant, or a spectral shift, and the various LFERs (Hammett, Brønsted, Taft, Grunwald-Winstein) are different projections of the same underlying thermodynamic variable. The generalises from aromatic systems to aliphatic and heteroatom systems through the Taft and DSP frameworks, completing the quantitative apparatus for pKa prediction from molecular structure.

Full proof set [Master]

Proposition 1 (Hammett $\rho$ from thermodynamic cycle). For a reaction $\text{HA}\to {\text{A}}^{-}+{\text{H}}^{+}$, the Hammett equation $\log (K_X/K_H)=\rho \sigma$ is equivalent to $\Delta \Delta G_X^{\circ }=-2.303RT\rho \sigma$, where $\Delta \Delta G_X^{\circ }$ is the change in $\Delta G^{\circ }$ caused by substituent X relative to H.

Proof. From $\Delta G^{\circ }=-RT\ln K$, the free energy for the unsubstituted acid is $\Delta G_H^{\circ }=-RT\ln K_H$ and for the substituted acid $\Delta G_X^{\circ }=-RT\ln K_X$. Their difference is:

$$\Delta \Delta G_X^{\circ }=\Delta G_X^{\circ }-\Delta G_H^{\circ }=-RT(\ln K_X-\ln K_H)=-RT\cdot 2.303{\log }_{10}(K_X/K_H).$$

Substituting the Hammett equation $\log (K_X/K_H)=\rho \sigma$:

$$\Delta \Delta G_X^{\circ }=-2.303RT\rho \sigma .$$

At 298 K, $2.303RT=5.706$ kJ/mol. One unit of $\sigma$ perturbs the free energy by $5.706\rho$ kJ/mol. For benzoic acid ionisation ($\rho =1$), a para-nitro group ($\sigma =0.78$) stabilises the anion by $5.706\times 0.78=4.45$ kJ/mol, corresponding to a pKa shift of $0.78$ units. $\square$

Proposition 2 (Marcus-theory interpretation of the Brønsted coefficient). For a proton-transfer reaction $\text{HA}+{\text{B}}^{-}\to {\text{A}}^{-}+\text{HB}$, if the intrinsic barrier $\Delta G_0^{‡}$ is independent of the identity of HA, then the Brønsted coefficient $\alpha =\partial (\Delta G^{‡})/\partial (\Delta G^{\circ })$ is given by $\alpha =1/2+\Delta G^{\circ }/(4\Delta G_0^{‡})$.

Proof. The Marcus theory expression for the activation barrier of proton transfer is:

$$\Delta G^{‡}=\Delta G_0^{‡}+\frac{\Delta G^{\circ }}{2}+\frac{(\Delta G^{\circ }{)}^2}{16\Delta G_0^{‡}}.$$

Differentiating with respect to $\Delta G^{\circ }$:

$$\frac{\partial \Delta G^{‡}}{\partial \Delta G^{\circ }}=\frac{1}{2}+\frac{2\Delta G^{\circ }}{16\Delta G_0^{‡}}=\frac{1}{2}+\frac{\Delta G^{\circ }}{8\Delta G_0^{‡}}.$$

The Brønsted coefficient $\alpha$ relates to the logarithmic derivative: $\alpha =-\partial (\log k)/\partial (\text{p}{K}_a)=(1/2.303RT)\partial \Delta G^{‡}/\partial \Delta G^{\circ }$. The factor $2.303RT$ cancels in the ratio, so $\alpha =\partial \Delta G^{‡}/\partial \Delta G^{\circ }=1/2+\Delta G^{\circ }/(8\Delta G_0^{‡})$. For a thermoneutral reaction ($\Delta G^{\circ }=0$), $\alpha =0.5$ (symmetric transition state). For a very exothermic reaction ($\Delta G^{\circ }\ll 0$), $\alpha <0.5$ (early transition state, proton closer to the donor). For a very endothermic reaction ($\Delta G^{\circ }\gg 0$), $\alpha >0.5$ (late transition state, proton closer to the acceptor). $\square$

Connections [Master]

• General chemistry acid-base 14.10.01 (pending). Supplies the Brønsted-Lowry definition, the $K_a$/pKa formalism, and the concept of conjugate acid-base pairs. This unit extends that framework to organic functional groups and adds the structural rationalisation of pKa trends through resonance, induction, and hybridisation.

• SN1 vs SN2 substitution mechanisms 15.04.02 pending. Leaving-group ability in nucleophilic substitution is ranked by the pKa of the conjugate acid of the leaving group: $p{K}_a(\text{HI})=-10$ makes iodide an excellent leaving group, while $p{K}_a({\text{H}}_2\text{O})=15.7$ makes water a moderate one. The acid-base framework developed here builds toward the leaving-group rankings used in 15.04.02 pending.

• Electrophilic addition 15.05.01 (pending). Protonation of an alkene is the first step of electrophilic addition. The regiochemistry depends on which carbon of the double bond is better able to stabilise positive charge, an argument parallel to conjugate-base stability in acid-base chemistry.

• Carbonyl chemistry 15.07.01 (pending). The acidity of alpha-hydrogens next to carbonyl groups (pKa ~20) and the use of strong bases (LDA) to generate enolates are direct applications of the acid-base principles here. The Hammett equation for substituted acetophenones quantifies the substituent effect on alpha-C-H acidity.

• Amino acids and protein chemistry 15.12.01 pending (pending). The pKa values of amino acid side chains (aspartate ~3.9, glutamate ~4.3, lysine ~10.5, arginine ~12.5, histidine ~6.0) determine protein charge states, enzyme catalytic mechanisms, and electrophoretic mobility. The isoelectric point of a protein is computed from these pKa values using the equilibrium framework established here.

• Enzyme mechanism 15.14.01 pending. General acid-base catalysis by enzyme active-site residues (His, Asp, Glu, Cys) is a direct application of the proton-transfer and pKa analysis developed here. The pKa values of amino acid side chains determine which residue can act as a general acid or base at physiological pH, and the pKa perturbation by the protein environment extends the same structural effects — resonance, induction, hybridisation — that rationalise organic pKa trends.

Historical & philosophical context [Master]

Johannes Brønsted ^{[Bronsted 1923]} introduced the proton-transfer definition of acids and bases in 1923 (Rec. Trav. Chim. Pays-Bas 42, 718--728), independently of Thomas Lowry who published the same concept that year. The Brønsted-Lowry definition replaced the Arrhenius framework (limited to aqueous H${}^{+}$ and OH${}^{-}$) with a general proton-donor/proton-acceptor model applicable to all solvents. Gilbert Lewis proposed the broader electron-pair-donor/acceptor definition in the same year (Valence and the Structure of Atoms and Molecules, 1923).

The pKa scale was established by Soren Sorensen's introduction of pH in 1909 and extended to acid dissociation constants through the work of Lawrence Henderson and Karl Hasselbalch. The Henderson-Hasselbalch equation $\text{pH}=\text{p}{K}_a+\log ([{\text{A}}^{-}]/[\text{HA}])$ remains the standard relationship for buffer chemistry.

Louis Hammett introduced the sigma-rho framework in 1935--1937. The key paper ^{[Hammett 1937]} (J. Am. Chem. Soc. 59, 96--103) demonstrated that $\log (K/K_0)$ for the ionisation of meta- and para-substituted benzoic acids correlates linearly with $\log (K/K_0)$ for other reactions of the same substituted benzenes. The Hammett equation established that substituent effects are transferable across reactions, providing the quantitative foundation for physical organic chemistry.

Robert Taft ^{[Taft 1952]} extended the LFER framework to aliphatic systems in 1952 (J. Am. Chem. Soc. 74, 3120--3128), separating polar and steric effects through the dual analysis of acid- and base-catalysed ester hydrolysis rates. The Taft equation opened quantitative pKa prediction to non-aromatic molecules. George Olah's superacid studies in the 1960s--1970s (J. Am. Chem. Soc. 94, 808--820, 1972) demonstrated that carbocations -- previously inferred reactive intermediates -- could be observed as stable species in media with Hammett acidity function $H_0<-15$.

Bibliography [Master]

@article{Bronsted1923,
  author = {Br{\o}nsted, J. N.},
  title = {Einige Bemerkungen {\"u}ber den Begriff der S{\"a}uren und Basen},
  journal = {Rec. Trav. Chim. Pays-Bas},
  volume = {42},
  year = {1923},
  pages = {718--728}
}

@article{Hammett1937,
  author = {Hammett, L. P.},
  title = {The Effect of Structure upon the Reactions of Organic Compounds. {B}enzene Derivatives},
  journal = {J. Am. Chem. Soc.},
  volume = {59},
  year = {1937},
  pages = {96--103}
}

@article{Taft1952,
  author = {Taft, R. W.},
  title = {Polar and Steric Constants for Comparative Reactivity Studies},
  journal = {J. Am. Chem. Soc.},
  volume = {74},
  year = {1952},
  pages = {3120--3128}
}

@article{Hansch1991,
  author = {Hansch, C. and Leo, A. and Taft, R. W.},
  title = {A Survey of {H}ammett Substituent Constants and Resonance and Field Parameters},
  journal = {Chem. Rev.},
  volume = {91},
  year = {1991},
  pages = {165--195}
}

@article{Olah1972,
  author = {Olah, G. A.},
  title = {Stable Carbocations. {T}he \textit{tert}-Butyl Cation},
  journal = {J. Am. Chem. Soc.},
  volume = {94},
  year = {1972},
  pages = {808--820}
}

@book{Hammett1940,
  author = {Hammett, L. P.},
  title = {Physical Organic Chemistry},
  publisher = {McGraw-Hill},
  year = {1940}
}

@book{Bell1973,
  author = {Bell, R. P.},
  title = {The Proton in Chemistry},
  publisher = {Chapman and Hall},
  edition = {2nd},
  year = {1973}
}

@book{Clayden2012,
  author = {Clayden, J. and Greeves, N. and Warren, S.},
  title = {Organic Chemistry},
  publisher = {Oxford University Press},
  edition = {2nd},
  year = {2012}
}

@book{Anslyn2006,
  author = {Anslyn, E. V. and Dougherty, D. A.},
  title = {Modern Physical Organic Chemistry},
  publisher = {University Science Books},
  year = {2006}
}

@book{March2013,
  author = {Smith, M. B.},
  title = {March's Advanced Organic Chemistry},
  publisher = {Wiley},
  edition = {7th},
  year = {2013}
}