Comparing SN1 and SN2

Predicting whether a haloalkane will react by $S_N1$ or $S_N2$ requires weighing three factors together -- the substrate's structure, the nucleophile's strength, and the solvent -- rather than any one of them in isolation.

Substrate structure is usually the dominant factor. A methyl or primary substrate is essentially forced into $S_N2$, because it cannot form a reasonably stable carbocation and its unhindered backside is readily accessible to a nucleophile. A tertiary substrate is essentially forced into $S_N1$ (if it reacts by substitution at all), because steric bulk blocks backside attack almost completely while a tertiary carbocation is comfortably stable. Secondary substrates are the genuine borderline case, capable of either pathway depending on the nucleophile and solvent.

Nucleophile strength then tips a borderline substrate. A strong, good nucleophile in high concentration (e.g. $\text{CN}^-$ or an alkoxide) pushes a secondary substrate towards $S_N2$; a weak nucleophile present mainly as the solvent itself (e.g. water or ethanol acting as both solvent and nucleophile, a "solvolysis") favours $S_N1$.

Solvent reinforces the same choice. A polar aprotic solvent (acetone, DMSO, DMF) leaves an added nucleophile unsolvated and reactive, reinforcing $S_N2$; a polar protic solvent (water, alcohols) stabilises the ionic intermediate and transition states of $S_N1$ through hydrogen bonding, reinforcing that pathway, while simultaneously "caging" and weakening any added anionic nucleophile through solvation. …