What makes a good leaving group?

Substitution and elimination reactions both begin the same way: a bond to carbon breaks, and one group leaves. Whether that step happens quickly, slowly, or not at all comes down almost entirely to how happy the departing group is to become independent. This one idea — stability of the group that leaves — unlocks most of the SN1/SN2/E1/E2 world.

The whole story in one row: stable, weak-base anions leave; strong-base anions do not.

1. A leaving group departs with the bonding electrons

When a nucleophile attacks the carbon of an alkyl halide, the carbon–halogen bond breaks heterolytically: both electrons go to the leaving group, which sails off as an anion. Nothing pushes it out — it has to be willing to hold that negative charge on its own. So the real question is never "how good is the nucleophile?" but "how comfortable is the leaving group as a free ion?"

Above, hydroxide displaces iodide from iodoethane. Iodide leaves easily because it is a large, stable, weak base. Reverse the roles and the reaction stalls: hydroxide will not spontaneously leave to be replaced by iodide, because hydroxide is a strong base that hates being a free anion.

2. The best leaving groups are the weakest bases

Stability of an anion is exactly what we mean by base strength — a weak base is a stable anion, and a stable anion is a weak base. Because every leaving group is the conjugate base of some acid, we can rank leaving-group ability directly from a pKa table. The rule is simple:

Lower pKa of the conjugate acid → weaker base → better leaving group.

The conjugate acids of iodide, bromide, and chloride are all strong acids (very negative pKa), so those halides are fine leaving groups. Water's pKa of 15.7 tells you hydroxide is a strong base — and therefore a dreadful leaving group. A quick mental test: if you have a pKa table, you barely need to memorize leaving-group ability at all.

3. Down the halogens, leaving-group ability improves: I⁻ > Br⁻ > Cl⁻ ≫ F⁻

Among the halides the trend follows size and polarizability. Iodine is huge; its negative charge is spread over a large, soft electron cloud, so iodide is very stable and leaves best. Fluoride is tiny and hard, holds its charge tightly, and is a comparatively strong base — so it is a genuinely bad leaving group despite fluorine's electronegativity.

Students often expect fluoride to leave because C–F is so polar. But polarity of the bond is not the point — stability of the departed anion is. Fluoroethane essentially does not undergo SN2, and alkyl fluorides are prized precisely because that bond stays put.

4. Sulfonates — tosylate and mesylate — are among the best leaving groups

Sulfonate esters (tosylate "OTs", mesylate "OMs", and triflate "OTf") are the workhorse leaving groups of synthesis. When the sulfonate leaves, the resulting anion carries its negative charge on oxygen and delocalizes it over three oxygens of the SO₃ group. That resonance makes it an exceptionally weak base — the conjugate acids are strong acids (TsOH has a pKa near −2.8), so these are outstanding leaving groups, often rivaling or beating iodide.

Their real value is strategic: they turn a bad leaving group into a great one. The oxygen of an alcohol is not going anywhere on its own, but esterify it as a tosylate and that same carbon is suddenly ready for substitution or elimination — with the added bonus that installing a tosylate does not touch stereochemistry at the carbon.

5. Hydroxide is terrible — so alcohols must be activated first

This is the trap that catches everyone. An alcohol looks like it should undergo SN2, but if it simply left, the leaving group would be hydroxide — a strong base with a pKa-15.7 conjugate acid. It will not go. There are two standard fixes.

Fix A — protonate it. Add strong acid and the –OH becomes –OH₂⁺ (an oxonium). Now the group that leaves is neutral water, whose conjugate acid is H₃O⁺ (pKa −1.7). Water is a fine leaving group, which is why acid-catalyzed substitutions and dehydrations of alcohols work.

Protonation converts a bad LG (OH⁻) into a good one (neutral H₂O).

Fix B — convert it to a tosylate or halide. React the alcohol with TsCl (or with SOCl₂/PBr₃ to make the halide) and you swap the awful –OH for a first-rate leaving group before you ever run the substitution.

The same logic explains the bottom of the "bad" list — hydroxide (OH⁻), alkoxide (OR⁻), amide (NH₂⁻), hydride (H⁻), and fluoride (F⁻). Every one is a strong base sitting on top of a high-pKa conjugate acid, so none of them will leave a carbon voluntarily.

6. Summary

A good leaving group is nothing more than a stable anion = a weak base = the conjugate base of a strong acid. Read leaving-group ability straight off a pKa table: the lower the conjugate acid's pKa, the better. Iodide beats bromide beats chloride; sulfonates are top-tier; water (from a protonated alcohol) is fine, but bare hydroxide, alkoxide, amide, hydride, and fluoride are all too basic to leave. When your substrate is an alcohol, activate it first — protonate it, or convert it to a tosylate or halide.