Alcohols, Phenols and Ethers

Aspirin is made by acetylating salicylic acid, itself made from phenol. Wine and beer exist because yeast's own enzymes convert sugar into ethanol. Diethyl ether was, for decades, the standard surgical anaesthetic, until its slow action and unpleasant recovery period saw it replaced. All three stories trace back to the same starting point: a hydrogen atom on a hydrocarbon, replaced by an oxygen-containing group. Replace it with a hydroxyl group, -OH, attached to an ordinary sp3 alkyl carbon, and the result is an alcohol; attach that same -OH group directly to an aromatic ring's sp2 carbon instead, and the result is a phenol, a genuinely different compound class despite the shared -OH; replace a hydrogen with an alkoxy or aryloxy group, R-O- or Ar-O-, instead of a bare hydroxyl, and the result is an ether, oxygen now bonded to two carbon groups with no O-H bond left at all. This one small structural choice, where exactly the oxygen sits and what it's still bonded to, turns out to separate three compound classes with dramatically different acidity, reactivity, and even physical behaviour, and this chapter follows all three all the way through.

Alcohols and phenols both sort first by count: monohydric (one -OH group), dihydric (two), trihydric (three), or polyhydric beyond that. Monohydric alcohols sort further by exactly which carbon holds the -OH: primary, secondary or tertiary, the same vocabulary already established for haloalkanes and carbocations, while allylic alcohols place the -OH one carbon from a double bond and benzylic alcohols place it one carbon from an aromatic ring, both still on an sp3 carbon. A vinylic alcohol, by contrast, puts the -OH directly on an sp2, doubly-bonded carbon, a genuinely rarer and less stable arrangement. Ethers sort by symmetry instead: a simple, or symmetrical, ether carries the same alkyl or aryl group on both sides of the oxygen, like diethyl ether; a mixed, or unsymmetrical, ether carries two different groups instead, like ethyl methyl ether. Every one of these classification questions matters later in the chapter, since an alcohol's own reactivity, and even which reaction mechanism it follows, depends directly on whether it is primary, secondary or tertiary.

In an alcohol, oxygen's own two lone electron pairs repel the bonding pairs more strongly than bonding pairs repel each other, so the C-O-H angle sits slightly below the ideal tetrahedral angle of 109.5 degrees. In phenol, that same -OH group sits on an sp2 hybridised ring carbon instead, and the C-O bond measurably shortens to 136 pm, noticeably less than an alcohol's own C-O bond, for two compounding reasons: the oxygen's lone pair partly delocalises into the aromatic ring, giving the bond partial double-bond character, and the sp2 carbon's own greater s-character holds the shared electrons more tightly. In an ether, finally, the two bulky R groups attached to the same oxygen repel each other more than a single hydrogen ever could, so the C-O-C angle actually widens slightly above the tetrahedral angle instead, the exact opposite shift alcohols show, for the exact opposite structural reason.

An alkene reacting with water under acid catalysis adds -OH exactly where Markovnikov's rule predicts, through a genuine three-step mechanism: the acid first protonates the alkene to form the more stable possible carbocation, water then attacks that carbocation as a nucleophile, and a final deprotonation releases the alcohol. Hydroboration-oxidation reaches the opposite regiochemical outcome from the same starting alkene: diborane first adds across the double bond with boron landing on the less hindered carbon, and hydrogen peroxide under basic conditions then swaps that boron cleanly for a hydroxyl, delivering water as if by anti-Markovnikov addition, the reverse of the acid-catalysed route. Herbert C. Brown, who first reported hydroboration in 1959, shared the 1979 Nobel Prize in Chemistry for this work, and the method remains prized precisely because it delivers the anti-Markovnikov alcohol in excellent, reliable yield, a regiochemical outcome the acid-catalysed route simply cannot reach.

Aldehydes and ketones reduce cleanly to alcohols, either by catalytic hydrogenation over finely divided platinum, palladium or nickel, or by treatment with sodium borohydride or lithium aluminium hydride instead; aldehydes give primary alcohols this way, ketones give secondary alcohols. Carboxylic acids and esters reduce too, though only the strong, expensive reducing agent lithium aluminium hydride manages the acid directly, which is why the industrial route instead converts the acid to its ester first, then reduces that ester by ordinary catalytic hydrogenation. The Grignard reagent, already met as a haloalkane's own reaction product, supplies a third, genuinely versatile route: its carbanion-like carbon attacks a carbonyl's own electrophilic carbon in a nucleophilic addition, and hydrolysis of the resulting adduct releases the alcohol. What makes this route so useful is the range of outcomes a single Grignard reagent can reach, simply by choosing the carbonyl partner: reacting with methanal gives a primary alcohol, reacting with any other aldehyde gives a secondary alcohol, and reacting with a ketone gives a tertiary alcohol, one reagent, three distinct product classes, decided entirely by what it is added to.

Chlorobenzene, fused with sodium hydroxide at a punishing 623 K and 320 atmospheres, converts to sodium phenoxide, which acidification then turns into phenol, a route that recalls exactly how stubbornly a haloarene resists nucleophilic substitution under ordinary conditions. A second route starts by sulphonating benzene with oleum, then converting the resulting benzenesulphonic acid to sodium phenoxide by heating with molten sodium hydroxide, before acidification again releases phenol. A third route runs through the same diazonium salt chemistry already established for making haloarenes: an aromatic primary amine, treated with nitrous acid at 273 to 278 K, forms a diazonium salt, and warming that salt with water hydrolyses it straight to phenol, releasing nitrogen gas as the diazonium group departs. The fourth route, and by far the most industrially important, starts from cumene, isopropylbenzene: air oxidises cumene to cumene hydroperoxide, and dilute acid then converts that hydroperoxide to phenol and acetone together, acetone arriving as a genuinely useful by-product rather than mere waste, which is exactly why most of the world's commercial phenol is made this way today.

Boiling point rises with carbon count in both alcohols and phenols, thanks to growing van der Waals forces, and falls with branching in alcohols specifically, exactly the same surface-area effect already established for alkanes and haloalkanes. What makes alcohols and phenols genuinely stand out, though, is how much higher they boil than a hydrocarbon, ether, or haloalkane of comparable molecular mass: ethanol and propane, nearly identical in mass, differ enormously in boiling point, and methoxymethane, an ether of that same mass, boils at a point squarely between the two. The reason is intermolecular hydrogen bonding, the -OH group in one molecule reaching out to hydrogen-bond with the -OH of its neighbour, a genuinely strong attraction entirely absent in ethers and hydrocarbons, which must rely on weaker van der Waals forces alone. That very same hydrogen-bonding, this time formed with water molecules rather than with each other, is exactly why the lower alcohols dissolve in water in all proportions, a solubility that steadily fades as the alkyl or aryl group grows larger and more hydrophobic, eventually overwhelming what the hydrogen bond alone can compensate for.

Sodium, potassium and even aluminium metal react with both alcohols and phenols, releasing hydrogen gas and forming an alkoxide or phenoxide salt, and phenol goes one step further, also reacting with aqueous sodium hydroxide itself to form sodium phenoxide, something an ordinary alcohol cannot do, an early hint at how much more acidic phenol really is. Both reactions confirm that alcohols and phenols are genuine Bronsted acids, capable of donating a proton to a sufficiently strong base, and alcohols are Bronsted bases too, their oxygen's own lone pairs able to accept a proton in turn. Within the alcohol family, acid strength tracks the polarity of the O-H bond directly: an electron-releasing alkyl group pushes extra electron density onto oxygen, weakening that polarity and reducing acid strength, so bulkier, more electron-rich alcohols are measurably weaker acids than simpler ones. Water sits above all of them, though; reacting water with an alkoxide ion shows water donating its own proton back, proof that water is a stronger acid, and the alkoxide correspondingly a stronger base, than any simple alcohol.

Phenol's own -OH sits on an sp2 hybridised, more electronegative ring carbon, which pulls electron density away from oxygen and increases the O-H bond's polarity, favouring ionisation more than an alcohol's own O-H ever would. The deeper reason phenol ionises so much more readily, though, lies in what happens after the proton leaves: an alkoxide ion keeps its negative charge stranded entirely on one oxygen, while a phenoxide ion spreads that same charge out across the ring by resonance, several genuinely distinct resonance structures sharing the burden, and a charge spread across many atoms is always more stable than one concentrated on a single atom. Substituents on the ring tune this effect further: an electron-withdrawing group like -NO2 pulls even more electron density away and stabilises the phenoxide ion's resonance structures further still, but only from the ortho or para position, where its own resonance structure can actually reach the negative charge directly, exactly the same ortho/para-versus-meta logic already established for haloarene reactivity; an electron-releasing alkyl group does the opposite, so cresols are measurably less acidic than phenol itself. The resulting numbers are dramatic: nitrophenols carry a pKa near 7, phenol itself sits near 10, and ethanol trails at nearly 16, meaning phenol is roughly a million times more acidic than ethanol despite both carrying the exact same -OH group.

Alcohols and phenols both react with carboxylic acids, acid chlorides and acid anhydrides to form esters, a reaction called esterification, and acetylation of salicylic acid this way produces aspirin itself. The acid chloride route needs a base like pyridine alongside it, to neutralise the hydrogen chloride the reaction releases and pull the equilibrium toward the ester; the carboxylic acid route needs a trace of concentrated sulphuric acid instead, and since that reaction is reversible, water is removed as it forms to keep pulling it forward. A separate family of reactions, though, breaks the carbon-oxygen bond itself rather than the O-H bond, and these happen only in alcohols, not phenols, except in one specific case involving zinc dust. Treated with concentrated hydrochloric acid and zinc chloride together, the three alcohol classes reveal themselves by how fast they turn cloudy: a tertiary alcohol produces turbidity immediately, a secondary alcohol more slowly, and a primary alcohol barely at all at room temperature, the Lucas test, a genuinely useful diagnostic built entirely on the same tertiary-fastest reactivity order already established for haloalkane-forming reactions in general.

Heated with a protic acid, an alcohol loses a molecule of water entirely and forms an alkene, through a three-step mechanism mirroring exactly what a haloalkane's own E1 elimination looks like: protonation of the -OH itself, loss of water to form a carbocation, the genuinely slow, rate-determining step, and finally loss of a proton from an adjacent carbon to form the alkene. Because a carbocation's own stability decides how readily that middle step happens, dehydration follows the reactivity order tertiary faster than secondary faster than primary, tertiary alcohols dehydrating under noticeably milder conditions than ethanol's own demanding 443 K. Oxidation tells a genuinely different story, one that depends entirely on how many hydrogens sit on the alcohol's own carbon: a primary alcohol oxidises first to an aldehyde, using a controlled reagent like pyridinium chlorochromate, and further still to a carboxylic acid under a stronger oxidiser like acidified potassium permanganate; a secondary alcohol oxidises only as far as a ketone; a tertiary alcohol, with no hydrogen left on its own carbon for the reaction to remove, resists oxidation entirely under ordinary conditions, though genuinely forcing conditions can cleave its carbon skeleton apart instead. This same oxidation ladder plays out inside the human body too, with serious consequences: the body oxidises methanol first to methanal and then to methanoic acid, and it is that acid, not the methanol itself, that causes the blindness and death associated with methanol poisoning, a fact put to direct clinical use, since deliberately infusing a poisoned patient with diluted ethanol keeps the relevant enzyme busy processing the safer alcohol instead, buying the kidneys enough time to clear the methanol unchanged.

The -OH group's own lone pair delocalises into the aromatic ring, activating phenol strongly toward electrophilic substitution and directing incoming groups specifically to the ortho and para positions, exactly the resonance logic already established for phenoxide's own stability. Dilute nitric acid at a cool 298 K gives a mixture of ortho- and para-nitrophenol, separable by steam distillation, since the ortho isomer's own intramolecular hydrogen bond makes it steam-volatile while the para isomer's intermolecular hydrogen bonding, linking whole molecules together, makes it far less so; concentrated nitric acid instead drives all the way to 2,4,6-trinitrophenol, better known as picric acid. Bromination needs no Lewis-acid catalyst at all, unlike ordinary benzene, since the -OH group's own activating effect polarises the bromine molecule on its own, giving mostly a mono-bromophenol in a low-polarity solvent at low temperature, or a white precipitate of 2,4,6-tribromophenol in bromine water instead. Beyond simple substitution, phenoxide ion, formed by treating phenol with sodium hydroxide, is even more reactive still, reactive enough to undergo electrophilic substitution with carbon dioxide itself, a weak electrophile ordinary phenol could never engage, forming ortho-hydroxybenzoic acid, or salicylic acid, in Kolbe's reaction; phenol treated instead with chloroform and sodium hydroxide installs a -CHO group at the ortho position via the Reimer-Tiemann reaction, forming salicylaldehyde. Zinc dust reduces phenol straight to benzene, and air alone slowly oxidises phenol to a conjugated diketone, benzoquinone, and its darker relatives.

Methanol, once called wood spirit and made by destructive distillation of wood, is today produced by catalytic hydrogenation of carbon monoxide over a zinc oxide-chromium oxide catalyst under high pressure and temperature; it is a colourless, highly poisonous liquid boiling at 337 K, used chiefly as a solvent and as the starting material for making formaldehyde. Ethanol's story runs through biology instead: the invertase enzyme first splits cane or fruit sugar into glucose and fructose, and the zymase enzyme found in yeast then ferments both sugars into ethanol, entirely in the absence of air, releasing carbon dioxide as it goes, and fermentation stalls on its own once the alcohol concentration passes roughly 14 percent, since zymase's own activity is inhibited past that point; letting air into the mixture instead lets atmospheric oxygen oxidise the ethanol to ethanoic acid, spoiling the drink's taste entirely. Ethanol acts on the central nervous system, impairing judgement and lowering inhibitions at moderate levels and risking loss of consciousness or fatal respiratory depression at higher ones, and commercial ethanol is deliberately denatured, mixed with copper sulphate for colour and foul-smelling pyridine, to render it undrinkable while still leaving it useful as an industrial solvent; increasingly, large-scale ethanol is instead made synthetically, by direct hydration of ethene.

An alcohol under acid-catalysed dehydration can form either an alkene or an ether depending purely on temperature: ethanol heated with sulphuric acid at 443 K gives ethene, but held instead at a cooler 413 K gives ethoxyethane, the ether forming through a nucleophilic bimolecular attack of one alcohol molecule on a second, already-protonated alcohol molecule. This route only really works for unhindered, primary alkyl groups; bulkier secondary or tertiary alcohols favour elimination so strongly that the corresponding ether essentially never forms. The Williamson synthesis supplies a far more general laboratory route instead, an alkyl halide reacting with a sodium alkoxide, or a sodium phenoxide for an aryl ether, in a straightforward SN2 attack that works cleanly for both symmetrical and unsymmetrical ethers alike. Here too, though, elimination lurks as a genuine competitor, since an alkoxide is not just a nucleophile but a strong base as well: a primary alkyl halide gives good ether yields, but a tertiary alkyl halide gives no ether at all, elimination winning outright and delivering only the alkene, exactly the same substitution-versus-elimination contest already established for haloalkanes in general, simply replayed here with an alkoxide standing in as both the nucleophile and the base.

An ether's own C-O bonds are polarised, giving it a genuine net dipole moment, but that polarity barely touches its physical behaviour: ether boiling points sit close to an alkane of comparable mass, since there is no O-H bond available for hydrogen bonding between ether molecules themselves, and dramatically below a comparable alcohol's own boiling point; ether solubility in water, though, resembles an alcohol's own, since the ether's own oxygen lone pairs can still hydrogen-bond with water molecules even without an O-H group of its own to offer. Chemically, ethers are the least reactive of all the functional groups met so far, their C-O bond breaking only under genuinely forcing conditions: excess concentrated hydrogen halide, in the reactivity order HI greater than HBr greater than HCl, protonates the ether oxygen first, and the halide ion then attacks, following an SN2 pathway on a primary or secondary carbon or an SN1 pathway through a carbocation when a tertiary group is involved, exactly the same mechanism split already established for haloalkanes' own substitution chemistry. An alkyl aryl ether always cleaves on the alkyl side specifically, never the aryl side, since the aryl-oxygen bond carries its own partial double-bond character and greater strength, precisely as already established for phenol's own shortened C-O bond; anisole, treated with excess hydrogen iodide, therefore gives phenol and methyl iodide, never iodobenzene. And despite this general unreactivity, an ether's own alkoxy group activates its aromatic ring toward electrophilic substitution and directs the incoming group to the ortho and para positions, exactly the same resonance behaviour phenol itself shows, letting anisole undergo halogenation, nitration and Friedel-Crafts reactions all without difficulty.