Carbohydrates — Organic Chemistry Notes

Glucose is an aldehyde with an OH on almost every carbon. In water, one of those OH groups adds to the molecule’s own aldehyde and closes a ring. Almost everything that sets sugars apart follows from that one reaction: the ring itself, the two forms the ring comes in, and whether the sugar reduces Tollens’ reagent.

This section therefore applies Carbonyl Chemistry rather than starting something new. A carbohydrate, or sugar, is a polyhydroxy aldehyde or ketone: a carbon chain with an OH on nearly every carbon and one C=O.

The vocabulary, quickly

Start with glucose and fructose, drawn below as Fischer projections (the stacked crosses from Stereochemistry). Both have six carbons. Glucose carries its C=O at the end of the chain, as an aldehyde (CHO) at C1. Fructose carries it inside the chain, as a ketone at C2.

Glucose has its C=O at the end of the chain; fructose has it at C2. Below C2 the two are identical, and the shaded row is the one that sets D.

A sugar with an aldehyde is an aldose, and one with a ketone is a ketose. Add the number of carbons and you have the full class name: glucose is an aldohexose (aldehyde, six carbons), and fructose is a ketohexose.
A monosaccharide is one sugar unit. A disaccharide is two units joined together, and a polysaccharide is many.

In glucose, each of C2 to C5 carries an H, an OH and two different parts of the chain, so each one is a stereocenter. An aldohexose therefore has four stereocenters and 2⁴ = 16 stereoisomers: eight D-sugars and their eight L mirror images, with eight names between them (glucose, mannose, galactose and five more).

D and L describe the configuration at one carbon only: the stereocenter furthest from the carbonyl. In glucose and fructose that is C5. Draw the sugar as a Fischer projection with the carbonyl end at the top. If the OH on that carbon points right, the sugar is D; if it points left, the sugar is L. Almost every sugar in nature is D.

D and L are not the same as (+)/(−) or R/S. D says only where one OH sits in a Fischer projection. It says nothing about the other stereocenters or about which way the compound rotates plane-polarized light. D-glucose and D-fructose are both D, yet they rotate light in opposite directions.

The ring is a hemiacetal, and that explains nearly everything

Glucose has an OH on C5 and an aldehyde at C1. The chain is flexible, so it can curl until the C5 oxygen reaches C1. That oxygen then adds to the C=O, just as an alcohol adds to an aldehyde in Acetals & hemiacetals. The product is a cyclic hemiacetal: a ring in which C1 carries both the ring oxygen and an OH. The Haworth section below shows how the chain curls into a ring, and the section after it draws the arrows for the ring closing.

The only difference from that section is that the alcohol and the aldehyde belong to one molecule. Five- and six-membered rings close easily: they have little strain, and the OH and the C=O are on the same chain, so they meet often. For glucose the ring wins outright: over 99% of glucose in water is in ring form.

A six-membered sugar ring (five carbons and an oxygen) is a pyranose. A five-membered one (four carbons and an oxygen) is a furanose. Glucose forms a pyranose; the fructose unit in sucrose is a furanose.
The old carbonyl carbon becomes the anomeric carbon. It is the only ring carbon bonded to two oxygens: the ring oxygen and an OH.

How a ring sugar is drawn: the Haworth projection

A Haworth projection is the standard way to draw a sugar ring. It shows the ring as a flat hexagon seen nearly edge-on, like a tabletop tipped slightly toward you. The ring oxygen sits at the back right and the anomeric carbon at the right-hand corner, and the numbering runs clockwise from there. The front bond, the one nearest you, is drawn thick.

Each substituent is a vertical stick, pointing straight up or straight down from its ring carbon. A Haworth drawing never uses wedges or dashes, because the tipped ring already shows the three dimensions. This course leaves off the hydrogens on the ring carbons to keep the drawings readable.

Turning a Fischer projection into a Haworth ring takes two rules. An OH on the right in the Fischer projection points down in the ring, and one on the left points up. And in a D-sugar, the CH₂OH on C5 points up.

The second rule follows from the first. Tip the Fischer projection onto its right side and curl it into the shape of a ring: every group that was on the right now points down, the C5 OH included. That OH has to reach C1, so C5 turns about its bond to C4. The turn swings the OH into the ring and sends the CH₂OH up. Tipping the drawing and turning about a single bond change no stereocenter. The configuration is the same at every stage, just as it was when Fischer projections replaced wedges and dashes in Stereochemistry.

The same molecule at every stage. Follow the teal OH on C5: it points down after the tip, and turning C5 swings it into the ring and sends the CH₂OH up. In panels 2 and 3, the H on C5 is drawn to show where its third group goes; other hydrogens are left off.

C3 is the quick check. It is the only carbon in D-glucose whose OH sits on the left in the Fischer projection, so it is the only ring carbon apart from C1 whose OH points up. A ring with any other pattern is a different sugar.

Two anomers, α and β

Go back to C1 in the open chain. Its C=O is flat, so the C5 oxygen can add to either face. One face leaves the new C1 OH pointing down in the Haworth drawing, and the other leaves it pointing up. Closing the ring has created a new stereocenter that the open chain did not have.

The two ring forms are called anomers, and they are labeled α and β. They differ at one stereocenter only, so they are a special case of the epimers defined in Stereochemistry. In a D-sugar drawn the standard way, α has the anomeric OH pointing down, on the opposite side of the ring from the CH₂OH. β has it pointing up, on the same side as the CH₂OH. Remember β as “same side as the CH₂OH”. It still works for L-sugars and in chair drawings, where up and down on the page are less reliable.

The coral dot is C1, the anomeric carbon. The proton on the C5 oxygen ends up on the new C1 OH, so both anomers are neutral.

Worked example — drawing β-D-galactopyranose from the Fischer projection

D-galactose is D-glucose with C4 inverted. Its Fischer projection has the OH on the right at C2, the left at C3, the left at C4 and the right at C5.

Draw the empty ring first. Make it six-membered, with the ring oxygen at the back right, C1 at the right and the numbering clockwise. Ignore the substituents for now.

Place the CH₂OH. Galactose is a D-sugar, so the CH₂OH on C5 points up.

Convert C2, C3 and C4 one at a time. C2 is on the right, so its OH points down. C3 and C4 are both on the left, so both of their OH groups point up. In glucose the C4 OH points down, so that single stick is the whole difference.

Set the anomeric carbon last. β means the C1 OH is on the same side as the CH₂OH, so it points up. For α it would point down.

The order matters. Once the CH₂OH is in place, you never need to recall which way α points in the abstract. You only need to remember that β matches the CH₂OH.

Galactose drawn from its Fischer projection. Only the teal OH on C4 differs from β-D-glucose.

Mutarotation: the anomers interconvert

Dissolve crystals of pure α-D-glucose in water and measure the optical rotation. It starts at +112° and slowly falls to +52.7°. Start from pure β-D-glucose instead, and the rotation starts at +18.7° and slowly rises to the same +52.7°. This slow drift to a fixed value is called mutarotation.

It happens because the ring keeps opening and closing. The hemiacetal falls back to the open-chain aldehyde, and the aldehyde can close to either anomer. Over time the two anomers reach equilibrium, whichever one you started with.

At equilibrium, D-glucose is about 36% α and 64% β, with well under 0.1% open chain at any moment. That fraction is tiny, but mutarotation shows the open chain is there, because the anomers could not interconvert without it.

β is favored for a reason you met in Alkanes & Conformations: the chair. In the chair of β-D-glucopyranose, the anomeric OH is equatorial, and so is every other substituent: all four OH groups and the CH₂OH. β-D-glucopyranose is the only D-aldohexose that can put every substituent equatorial at once. The α anomer has the same chair except for its C1 OH, which is axial. An axial group is crowded by the two axial hydrogens on the same face of the ring, so β is the major anomer.

The coral dot is C1. Axial bonds point straight up or down; the hydrogens sit on the gray sticks.

The margin is smaller than crowding alone would predict. An electronic preference called the anomeric effect favors an electronegative group, such as OH, in the axial position on the carbon next to the ring oxygen. It works against the crowding and keeps α at about a third of the mixture.

Two traps worth naming now. Mutarotation changes one stereocenter, not five. The ring opens only at the anomeric carbon, and no C–C bond breaks, so the other four stereocenters are untouched. α-D-glucose can turn only into β-D-glucose. And an anomer is not an enantiomer. α- and β-D-glucose differ at one of five stereocenters, so they are diastereomers, with different melting points, different rotations and different solubilities.

Hemiacetal or acetal: the distinction that decides reducing sugars

Compare glucose with methyl β-D-glucopyranoside, drawn side by side below. In glucose, C1 carries the ring oxygen and an OH, so it is a hemiacetal. In the second compound that OH has become OCH₃, so C1 carries two OR groups and is an acetal. Acetals & hemiacetals showed that a hemiacetal can fall back to its carbonyl compound, while an acetal stays intact unless aqueous acid is present. In sugar chemistry, that is the whole difference between a reducing and a non-reducing sugar.

A free anomeric OH means the ring is a hemiacetal, so it can open to the aldehyde. Tollens’ and Benedict’s reagents oxidize that aldehyde, so the sugar gives a positive test. A sugar that does this is a reducing sugar, named for what it does to the reagent. Oxidizing an aldehyde shows both tests and draws the ring opening in base.
If the anomeric carbon is tied up as an acetal, the ring cannot open, no aldehyde appears, and the sugar is non-reducing. The C–O–C link from the anomeric carbon to the second alcohol is called a glycosidic bond. Both tests run in base, and a glycosidic bond is stable to base, so the ring stays shut.

The two molecules differ only in what caps the oxygen on C1 (the dotted carbon): H on the left, CH₃ on the right.

Worked example — why sucrose does not reduce and maltose does

A disaccharide is named by its glycosidic bond. α(1→4) means the bond leaves C1 of the first unit, which is α, and joins the oxygen on C4 of the second unit.

Maltose is glucose α(1→4) glucose. The bond uses the anomeric carbon of the first glucose and an ordinary OH of the second. The second glucose keeps its anomeric OH, so its ring can still open. Maltose is a reducing sugar.

Sucrose joins glucose to fructose through both anomeric carbons: C1 of glucose, which is α, to C2 of fructose, which is β. It is written glucose α(1→2)β fructose. Neither ring has an anomeric OH left, so neither can open and no aldehyde ever forms. Sucrose is non-reducing.

Lactose (galactose β(1→4) glucose) and cellobiose (glucose β(1→4) glucose) are built like maltose but with a β link. Each uses one anomeric carbon and leaves the other free, so both are reducing. Maltose, lactose and cellobiose all reduce; sucrose is the common one that does not.

The reasoning never mentions sweetness, size or source. The only question is whether a free anomeric OH is left.

Coral dots mark the anomeric carbons and coral fills the bridging oxygen. In sucrose the fructose ring is turned over, left to right, so its C2 can reach the bridge. Turning it over sends every group, OH and CH₂OH alike, to the opposite face from the usual fructose drawing.

Three reactions, all of them ones you already have

A sugar in solution is almost all ring, yet each reagent below reacts with the open-chain form. As the reagent uses up the open chain, more ring opens to replace it. That is why a reducing sugar behaves like the aldehyde it can become, even though almost none of it is an aldehyde at any moment.

Reduction. NaBH₄ reduces the open-chain aldehyde to a primary alcohol. The product is an alditol, a sugar with no carbonyl left and an OH on every carbon. D-glucose gives D-glucitol, sold as sorbitol. This is the hydride addition from Oxidation & Reduction, run on the small amount of open chain.
Oxidation. Bromine water (Br₂ in water) is a mild oxidant that leaves alcohols alone. It oxidizes the aldehyde and stops, giving an aldonic acid: D-glucose gives D-gluconic acid. A stronger oxidant, such as warm nitric acid, also oxidizes the CH₂OH at the other end and gives the diacid, an aldaric acid. Tollens’ and Benedict’s reagents do the same job as bromine water, but they are used as tests rather than to make the acid.
Glycoside formation. An alcohol and an acid catalyst turn the hemiacetal into the acetal, called a glycoside (the capped ring from the section above). Methanol and HCl give methyl α- and β-D-glucopyranoside. Warm aqueous acid hydrolyzes it back to the sugar.

Teal marks the ends that changed. Warm nitric acid can also start from glucose itself, passing through gluconic acid on the way to glucaric acid.

Tollens’ and Benedict’s tests run in base, and base lets a ketose pass them too. The figure below follows fructose. Base removes a hydrogen from C1, the carbon next to the C2 carbonyl, and the oxygen then picks up a proton from water. The product is an enediol: a C=C with an OH on each of its two carbons, the enol idea from Enolate Chemistry. Two more proton moves (panel 3) then put the C=O at C1, turning the ketose into an aldose, which the test oxidizes. So fructose gives a positive test with no aldehyde in its own structure. The rule to use is that a sugar with a free anomeric OH reduces Tollens’ reagent, ketoses included; only a capped anomeric carbon gives a negative test.

Base runs every step both ways, which is why fructose, glucose and mannose interconvert in base. Coral marks C1 and C2; teal marks the hydrogens that move.

Bromine water is not basic, so a ketose stays a ketose in it. It oxidizes aldoses and leaves ketoses alone, which makes it the reagent that tells the two apart.

The polysaccharides worth knowing

Cellulose — glucose joined β-1,4, the β(1→4) link of cellobiose, repeated. The β linkage makes a straight chain, and the chains stack into fibers. Human enzymes cannot hydrolyze it.
Starch (amylose) — glucose joined α-1,4, the α(1→4) link of maltose, repeated. The α linkage makes a helical chain, and we digest it easily.

Two units of each chain; the coral oxygen is the glycosidic link. In the lower drawing the second ring is turned over, so its ring oxygen sits at the front and its groups point the opposite way. It is still β-D-glucose.

The two polymers are made of the same monomer and differ only in the configuration at one carbon in each unit. That single stereochemical difference is why one is food and the other is wood.

What carries forward

Nothing in this section is new chemistry. It is hemiacetal and acetal formation, Fischer projections and diastereomers, applied to molecules that happen to be biological. If a carbohydrate question looks unfamiliar, find the carbonyl, or the carbon that used to be one, and ask what has added to it.