Zwitterions and the Amphoteric Nature of Amino Acids

Dissolve solid glycine in water and something strange happens. No proton ever truly leaves the molecule to the solvent. Yet the molecule carries both a full positive and a full negative charge at once. This is the zwitterion — from the German Zwitter, "hybrid." It is the actual structure of every free amino acid at physiological pH, not merely a textbook curiosity. Understanding it resolves a paradox that trips up every beginning student. How can glycine's melting point (233 °C, like a salt) coexist with a structure drawn as H₂N–CH₂–COOH? That structure looks like it should behave like a small, volatile organic acid. The answer matters because amino acids are the monomer units of every protein. The ionization state of their α-amino and α-carboxyl groups governs everything. This includes a protein's solubility and crystallization behavior. It also includes the electrophoretic separations used daily in every biochemistry laboratory. The zwitterion is not a side detail. It is why amino acids are water-soluble, crystalline, high-melting solids with large dipole moments. It is also the starting point for the acid-base titration behavior. That behavior determines a protein's isoelectric point, its charge at any given pH. Ultimately, it shapes a protein's folding and function. The big picture is simple once stated precisely. The α-amino group (pK_a ≈ 9–10) is a weaker acid than the α-carboxyl group (pK_a ≈ 2). So at any pH between these two values, the carboxyl has already given up its proton. Meanwhile, the amino group has not yet given up its own. The molecule is simultaneously the conjugate base of the carboxylic acid and the conjugate acid of the amine. Hence it is amphoteric, from the Greek "amphoteros," meaning "both." This single fact, applied systematically, explains the shape of every amino acid titration curve. It also sets up the machinery for Titration Curves of Amino Acids. This article builds the zwitterion from first principles. It covers the two independent acid-base equilibria of an amino acid. It examines the pH-dependent population of the four possible ionization states. It derives the isoelectric point. It also explains why the un-ionized "neutral" structure is essentially never populated in water. It closes with worked numerical examples for glycine, aspartate, and histidine. These were chosen because they showcase, respectively, the simplest amphoteric case, a side-chain-carboxyl complication, and a three-pK_a system relevant to enzyme catalysis.

Key takeaways

Every free amino acid in water exists overwhelmingly as a zwitterion: –COO⁻ and –NH₃⁺ on the same molecule, net charge zero, but not because no protons moved.
The α-COOH (pK_a ≈ 2) is a much stronger acid than the α-NH₃⁺ (pK_a ≈ 9–10); at physiological pH both facts are simultaneously true, forcing the dipolar form.
"Amphoteric" means the same molecule can act as a Brønsted acid (donating H⁺ from –NH₃⁺) or a Brønsted base (accepting H⁺ at –COO⁻), depending on the pH of the surroundings.
The isoelectric point (pI) is the pH at which the average net charge is zero; for simple amino acids it is the arithmetic mean of the two flanking pK_a values.
Physical evidence for the zwitterion — high melting point, high water solubility, low solubility in nonpolar solvents, a large dipole moment (~15 D) — all point to an internal salt, not a neutral covalent molecule.
The Henderson–Hasselbalch equation applied twice (once per ionizable group) predicts the fractional populations of all four charge states at any pH.
Key idea: an amino acid's structure is not a single fixed Lewis structure — it is a pH-dependent equilibrium mixture of charged species, and near neutral pH the dominant species is dipolar, not neutral.

Where this lives today

The zwitterionic nature of amino acids is not a historical footnote — it is load-bearing machinery in modern biochemistry, pharmacology, and biotechnology. Protein purification by ion-exchange chromatography exploits exactly the pH-dependent charge behavior derived above: by choosing a running buffer pH above or below a protein's aggregate isoelectric point, the protein can be made to bind a positively or negatively charged resin, then eluted by a salt or pH gradient that shifts its net charge. Isoelectric focusing, a technique used in 2D gel electrophoresis for proteomics, separates proteins along a pH gradient purely by pI — each protein migrates until it reaches the pH at which its own net charge (summed over every ionizable side chain, not just the α-amino/α-carboxyl pair, since those are consumed in the peptide bond once amino acids polymerize) is zero, and then stops moving, because an electric field cannot exert net force on a species with zero net charge. In drug design, the zwitterionic character of many pharmaceuticals (including several β-lactam antibiotics, some fluoroquinolones, and amino-acid-derived drugs like levodopa) governs oral bioavailability, membrane permeability, and aqueous solubility, because a drug's ionization state at gastric pH (~1.5–3.5) versus intestinal pH (~6–7.4) versus blood pH (7.4) can differ dramatically, and formulation chemists must design around this using the same Henderson–Hasselbalch machinery derived here. Amino acid zwitterions are also central to the food and agricultural industries — monosodium glutamate (MSG) is the sodium salt of the glutamate zwitterion's anionic form, chosen deliberately because that ionization state is what activates the umami taste receptor. Structurally, the pK_a values of ionizable side chains (many of them shifted from their "free amino acid" values by the local protein environment — buried histidines, aspartates near metal centers, cysteines in redox-active sites) underlie the Acid-Base Chemistry of the Alpha-Amino and Alpha-Carboxyl Groups that determines enzyme catalytic mechanisms, from the histidine "shuttle" analyzed above in the chymotrypsin/serine protease catalytic triad, to the aspartate/glutamate pairs that coordinate divalent metal ions in polymerases and phosphatases. Finally, the amphoteric behavior described here is the essential prerequisite for constructing and interpreting the Titration Curves of Amino Acids, where the theoretical population fractions and pI derived in this article become directly measurable, point by point, with a pH electrode and a strong acid or base titrant.

Ask the tutor: Why does the isoelectric point calculation for lysine use its two highest pKa values instead of the two lowest, and how would you verify that using the same charge-tracking method shown for aspartate and histidine?

The physical picture — why the "textbook" neutral structure is a fiction

Draw glycine the way an organic chemistry course might: H₂N–CH₂–COOH. This structure has a carboxylic acid and a primary amine, both electrically neutral. It is a perfectly valid Lewis structure — but it is not the structure that predominates in water, in the solid state, or under physiological conditions. The reason is a matter of relative acid strength, not electron-pushing artistry. A carboxylic acid attached directly to the electron-withdrawing ammonium/amino-bearing carbon has a pK_a around 2.0–2.4 — dramatically more acidic than a typical carboxylic acid like acetic acid (pK_a 4.76), because the adjacent (protonated) amino group inductively withdraws electron density and stabilizes the developing negative charge on the carboxylate. Meanwhile, the α-amino group, once protonated to –NH₃⁺, has a pK_a around 9–10, similar to other primary aliphatic ammonium ions. Now ask: at pH 7 (roughly where a cell lives), what is the protonation state of each group? For the carboxyl (pK_a ≈ 2.3), pH 7 is nearly 5 units above the pK_a, so by Henderson–Hasselbalch the carboxylate form –COO⁻ outnumbers –COOH by a factor of $10^{7-2.3} = 10^{4.7} \approx 50{,}000$ -fold. It is, for all practical purposes, fully deprotonated. For the amino group (pK_a ≈ 9.6), pH 7 is about 2.6 units below the pK_a, so the protonated ammonium form –NH₃⁺ outnumbers the neutral amine –NH₂ by $10^{9.6-7} = 10^{2.6} \approx 400$ -fold. It is, for all practical purposes, fully protonated. Put those two conclusions together: at pH 7, the carboxyl is (–) and the amino is (+), on the same molecule. The neutral structure — carboxyl protonated and amino also protonated, or carboxyl deprotonated and amino also deprotonated — requires either two simultaneous protonation "mistakes" or none, and both of those alternative species are minority populations by many orders of magnitude. The dipolar zwitterion is simply what's left when you correctly apply two independent equilibria to the same molecule.

Figure

Glycine's four possible ionization states and their relative populations at pH 7

The zwitterion's large separation of charge over a small molecule gives it a substantial dipole moment (glycine's zwitterion has a dipole moment near 15 debye in the solid state — compare water's 1.85 D), which is precisely why amino acids behave physically like salts: high melting/decomposition points (glycine decomposes near 233 °C rather than boiling like a small organic molecule would), high solubility in water and other polar solvents, and very low solubility in nonpolar solvents like diethyl ether or hexane. Sodium chloride shows the same pattern for the same reason — ionic lattices held together by Coulombic attraction require much more energy to disrupt than molecular crystals held together by van der Waals forces.

Figure

Two independent acid dissociations layered on one molecule

The core ideas — the governing equations

The behavior of an amino acid in solution is governed by two independent Henderson–Hasselbalch equilibria, one for each ionizable group, plus the definition of the isoelectric point as the pH at which the average net charge is zero.

\text{pH} = \text{p}K_{a1} + \log_{10}\!\left(\frac{[\text{zwitterion}]}{[\text{cation}]}\right)

where: p $K_{a1}$ is the pK_a of the α-carboxyl group (–COOH ⇌ –COO⁻ + H⁺), [cation] is the fully protonated species (+1 net charge), and [zwitterion] is the dipolar species (net charge 0). This equation describes the first ionization — loss of the carboxyl proton.

\text{pH} = \text{p}K_{a2} + \log_{10}\!\left(\frac{[\text{anion}]}{[\text{zwitterion}]}\right)

where: p $K_{a2}$ is the pK_a of the α-ammonium group (–NH₃⁺ ⇌ –NH₂ + H⁺), and [anion] is the fully deprotonated species (net charge −1). This describes the second ionization — loss of the ammonium proton.

\text{pI} = \frac{1}{2}\left(\text{p}K_{a1} + \text{p}K_{a2}\right)

where: pI is the isoelectric point for a "simple" amino acid whose only ionizable groups are the α-carboxyl and α-amino (no ionizable side chain). This is the pH exactly halfway between the two flanking pK_a values that bracket the neutral (zwitterionic) species — reading it: pI sits at the midpoint of the only pH window in which the zwitterion, rather than a charged form, dominates. For amino acids with an ionizable side chain (Asp, Glu, His, Cys, Tyr, Lys, Arg), the pI is instead the average of the two pK_a values that flank the neutral, zwitterionic species specifically — not simply "the two smallest" or "the two largest" pK_a's. Identifying which two pK_a's flank the zero-net-charge species requires tracking the charge state through the full titration, done explicitly in the worked examples below.

Derivation ADeriving the isoelectric point from charge balance

Goal. Show rigorously that pI = ½(pK_{a1} + pK_{a2}) for a simple diprotic amino acid, starting only from the two dissociation equilibria and the requirement of zero net average charge.

Step 1 — define the three species and their equilibria. Let AH₂⁺ be the fully protonated cationic form, AH the neutral zwitterion, and A⁻ the fully deprotonated anion. The two acid dissociation equilibria are

K_{a1} = \frac{[\text{AH}][\text{H}^+]}{[\text{AH}_2^+]}, \qquad K_{a2} = \frac{[\text{A}^-][\text{H}^+]}{[\text{AH}]}

Step 2 — write the average net charge as a population-weighted sum. The cation carries charge +1, the zwitterion charge 0, the anion charge −1. The average charge $\bar{z}$ over the whole population is

\bar{z} = \frac{(+1)[\text{AH}_2^+] + (0)[\text{AH}] + (-1)[\text{A}^-]}{[\text{AH}_2^+]+[\text{AH}]+[\text{A}^-]}

Step 3 — impose the isoelectric condition. At the isoelectric point, by definition $\bar{z}=0$ , which (since the denominator is nonzero) requires the numerator to vanish:

[\text{AH}_2^+] = [\text{A}^-]

This is the entire physical content of the isoelectric point: it is the pH at which the cationic and anionic populations are exactly equal (both necessarily small if the two pK_a's are well separated, but equal to each other).

Step 4 — substitute the equilibrium expressions. From Step 1, solve each equilibrium for its charged species in terms of [AH]:

[\text{AH}_2^+] = \frac{[\text{AH}][\text{H}^+]}{K_{a1}}, \qquad [\text{A}^-] = \frac{K_{a2}[\text{AH}]}{[\text{H}^+]}

Step 5 — set them equal and solve for [H⁺]. Using the Step 3 condition,

\frac{[\text{AH}][\text{H}^+]}{K_{a1}} = \frac{K_{a2}[\text{AH}]}{[\text{H}^+]}

The common factor [AH] cancels (it is nonzero at any finite pH), leaving

[\text{H}^+]^2 = K_{a1}K_{a2} \quad\Longrightarrow\quad [\text{H}^+] = \sqrt{K_{a1}K_{a2}}

Step 6 — convert to pH. Taking $-\log_{10}$ of both sides,

\text{pH}_{\text{pI}} = -\log_{10}\sqrt{K_{a1}K_{a2}} = -\tfrac{1}{2}\log_{10}K_{a1} - \tfrac{1}{2}\log_{10}K_{a2} = \tfrac{1}{2}(\text{p}K_{a1}+\text{p}K_{a2})

which is exactly the claimed result. Notice the derivation never assumed the two pK_a's were close together or symmetric about anything — it holds for any values of pK_a1 and pK_a2, provided they correctly flank the neutral zwitterionic species in the titration order (pK_a1 < pI < pK_a2).

Go deeper · Derivation B · 1 minFractional population of each ionization state as a function of pH

Derivation BFractional population of each ionization state as a function of pH

Goal. Derive closed-form expressions for the fraction of molecules in each of the four charge states (cation, zwitterion, anion — for a simple amino acid) at an arbitrary pH, so that "essentially fully deprotonated" claims made in prose can be checked numerically.

Step 1 — express all species relative to the zwitterion. Rearranging the two Henderson–Hasselbalch equations from the core-ideas section,

\frac{[\text{AH}_2^+]}{[\text{AH}]} = 10^{(\text{p}K_{a1}-\text{pH})}, \qquad \frac{[\text{A}^-]}{[\text{AH}]} = 10^{(\text{pH}-\text{p}K_{a2})}

Step 2 — define the total concentration and normalize. Let $C_T = [\text{AH}_2^+]+[\text{AH}]+[\text{A}^-]$ . Dividing every term by $[\text{AH}]$ and then by $C_T/[\text{AH}]$ gives the fraction of each species. Define

\alpha = 10^{(\text{p}K_{a1}-\text{pH})}, \qquad \beta = 10^{(\text{pH}-\text{p}K_{a2})}

Then

f_{\text{cation}} = \frac{\alpha}{1+\alpha+\beta}, \qquad f_{\text{zwitterion}} = \frac{1}{1+\alpha+\beta}, \qquad f_{\text{anion}} = \frac{\beta}{1+\alpha+\beta}

Step 3 — sanity-check the limiting behavior. As pH → −∞ (extremely acidic), $\alpha \to \infty$ and $\beta \to 0$ , so $f_{\text{cation}} \to 1$ : every molecule is fully protonated, exactly as physical intuition demands. As pH → +∞, $\beta \to \infty$ and $\alpha \to 0$ , so $f_{\text{anion}} \to 1$ . At pH = pI = ½(pK_{a1}+pK_{a2}), one can show $\alpha = \beta$ (both equal $10^{(pK_{a2}-pK_{a1})/2}$ ), so the cation and anion fractions are equal — consistent with the Step 3 finding of the first derivation, $[\text{AH}_2^+]=[\text{A}^-]$ .

Step 4 — apply to glycine at pH 7 to justify the "≈50,000-fold" and "≈400-fold" claims made in prose. With pK_a1 = 2.34, pK_a2 = 9.60:

\alpha = 10^{(2.34-7)} = 10^{-4.66} \approx 2.19\times10^{-5}, \qquad \beta = 10^{(7-9.60)} = 10^{-2.60} \approx 2.51\times10^{-3}

Since both $\alpha$ and $\beta$ are ≪ 1, the denominator $1+\alpha+\beta \approx 1.00254$ , and

f_{\text{zwitterion}} \approx \frac{1}{1.00254} = 0.99747, \qquad f_{\text{cation}} \approx 2.18\times10^{-5}, \qquad f_{\text{anion}} \approx 2.50\times10^{-3}

So at pH 7, 99.75% of glycine molecules are the zwitterion, 0.25% are the anion, and only 0.0022% are the cation — confirming the qualitative claim that the zwitterion utterly dominates, while also showing the anion (fully deprotonated, both groups anionic/neutral-amine) is not quite as negligible as the cation, because pH 7 sits closer to pK_a2 than the earlier order-of-magnitude argument alone made obvious.

Worked examples

Example 1 (foundational) — Glycine: population fractions at pH 9.60, the exact pK_a2. Glycine has pK_a1 = 2.34 (α-COOH) and pK_a2 = 9.60 (α-NH₃⁺), giving pI = ½(2.34 + 9.60) = 5.97. Question: at pH exactly equal to pK_a2 (9.60), what fraction of glycine is zwitterion versus anion, and does this match the expected 50:50 split for a simple Henderson–Hasselbalch buffer pair? Using the Step 1 relations from the second derivation, with pH = pK_a2 = 9.60:

\beta = 10^{(9.60-9.60)} = 10^0 = 1

\alpha = 10^{(2.34-9.60)} = 10^{-7.26} \approx 5.5\times10^{-8}

This $\alpha$ is utterly negligible, so the denominator is $1+\alpha+\beta \approx 1+0+1 = 2$ , giving

f_{\text{zwitterion}} = \frac{1}{2} = 0.500, \qquad f_{\text{anion}} = \frac{\beta}{2} = \frac{1}{2}=0.500, \qquad f_{\text{cation}} \approx 2.8\times10^{-8}

This is exactly the expected Henderson–Hasselbalch behavior: at pH = pK_a of a given ionizable group, that group is 50% protonated and 50% deprotonated, and here it plays out as a 50:50 split between zwitterion and anion, with the cationic form (governed by the other, much lower, pK_a1) entirely negligible. This confirms the two-equilibria framework reduces correctly to ordinary single-group Henderson–Hasselbalch behavior in the appropriate limit.

Example 2 (applied) — Aspartate: identifying which two pK_a's flank the zwitterion. Aspartate has three ionizable groups: α-COOH (pK_a1 ≈ 1.88), the side-chain β-COOH (pK_a(R) ≈ 3.65), and α-NH₃⁺ (pK_a2 ≈ 9.60). This is a case where naive "average the two smallest" or "average the two biggest" rules can mislead a student — the correct rule is to track net charge through the titration. Step 1: at very low pH, all three groups are protonated: ⁺H₃N–CH(CH₂COOH)–COOH, net charge +1. Step 2: the first proton lost (lowest pK_a, 1.88) is from the α-carboxyl, since it is the most acidic group (it experiences the strongest inductive pull from the adjacent, fully protonated ammonium). Losing it gives ⁺H₃N–CH(CH₂COOH)–COO⁻, net charge 0. This is the zwitterion — but note it still has a protonated side-chain carboxyl. Step 3: the second proton lost (pK_a 3.65) is from the side-chain carboxyl, giving ⁺H₃N–CH(CH₂COO⁻)–COO⁻, net charge −1. Step 4: the third proton lost (pK_a 9.60) is from the α-ammonium, giving H₂N–CH(CH₂COO⁻)–COO⁻, net charge −2. The species with net charge exactly 0 appears between the transitions at pK_a1 = 1.88 (which converts +1 → 0) and pK_a(R) = 3.65 (which converts 0 → −1). Therefore the two pK_a's that flank the neutral zwitterion are 1.88 and 3.65 — not pK_a2 = 9.60, even though 9.60 is numerically the largest value in the list. Applying the same midpoint logic as the derivation (charge-balance argument, generalized: at the isoelectric point the +1 and −1 species must be equally populated, and here it is the +1 species and the −1 species flanking the neutral form that matter):

\text{pI(Asp)} = \frac{1}{2}\left(\text{p}K_{a1} + \text{p}K_{a(R)}\right) = \frac{1}{2}(1.88 + 3.65) = 2.77

This low pI (compared to glycine's 5.97) is exactly why aspartate and glutamate are called "acidic" amino acids: at physiological pH 7.4, aspartate sits nearly 5 pH units above its pI, so by the Henderson–Hasselbalch logic it carries a substantial net negative charge (predominantly the −1 species, with a small −2 population), consistent with its role in forming ionic/electrostatic interactions and metal coordination in enzyme active sites.

Example 3 (exam/competition level) — Histidine's imidazole side chain and its role near physiological pH. Histidine has three pK_a's: α-COOH (pK_a1 ≈ 1.82), imidazole side chain (pK_a(R) ≈ 6.00), and α-NH₃⁺ (pK_a2 ≈ 9.17). Unusually, the side-chain pK_a (6.00) sits close to physiological pH — this is precisely why histidine is a favorite general acid/general base catalyst in enzyme active sites (for example, the catalytic triad of chymotrypsin). Following the same charge-tracking method as Example 2: fully protonated cation (+2, since here both the α-ammonium and the protonated imidazolium carry positive charge, in addition to the un-ionized α-COOH) loses its most acidic proton (α-COOH, pK_a1 = 1.82) to become net +1 (still-protonated imidazolium and ammonium, deprotonated α-carboxylate). The next proton lost is from the imidazolium (pK_a(R) = 6.00), giving net charge 0 — the zwitterion, with a neutral imidazole ring and a deprotonated α-carboxylate and a still-protonated α-ammonium. The final proton lost is from the α-ammonium (pK_a2 = 9.17), giving net −1. The neutral (net charge 0) species is flanked by pK_a1 = 1.82 (below) and pK_a(R) = 6.00 (above), so:

\text{pI(His)} = \frac{1}{2}(1.82 + 6.00) = 3.91

Now compute what fraction of the imidazole side chain is protonated (imidazolium, catalytically relevant as a general acid) at physiological pH 7.4, using ordinary single-group Henderson–Hasselbalch (since this equilibrium is essentially independent of the other two, which are far away in pH):

\text{pH} = \text{p}K_{a(R)} + \log_{10}\frac{[\text{imidazole}]}{[\text{imidazolium}]}

7.4 = 6.00 + \log_{10}\frac{[\text{imidazole}]}{[\text{imidazolium}]} \quad\Longrightarrow\quad \log_{10}\frac{[\text{imidazole}]}{[\text{imidazolium}]} = 1.4

\frac{[\text{imidazole}]}{[\text{imidazolium}]} = 10^{1.4} \approx 25.1

So at pH 7.4, roughly 25 neutral imidazole molecules exist for every 1 protonated imidazolium, meaning about $1/(1+25.1) \approx 3.8\%$ of histidine side chains are protonated at any instant. This modest but non-negligible population is exactly what allows histidine, when buried in a protein active site and with its pK_a perturbed by the local environment (often shifted toward 7 by neighboring charged or polar groups), to shuttle rapidly between protonated and deprotonated forms during a catalytic cycle — accepting a proton from a substrate at one step and donating it at the next, functioning as the "swing" of many enzymatic acid-base mechanisms.

Limiting cases, scope & common traps

Scenario	What happens	Why
Simple amino acid (Gly, Ala, Val, Leu, etc.)	pI = ½(pK_a1 + pK_a2), zwitterion dominates over a wide pH plateau between the two pK_a's	Only two ionizable groups; no competition for which pair flanks charge zero
Acidic side chain (Asp, Glu)	pI is low (~2.8–3.2); computed from the two lowest pK_a's, not pK_a2	The neutral species appears early in the titration, before the α-amino group ionizes
Basic side chain (Lys, Arg, His)	pI is high (~7.6–10.8, except His); computed from the two highest pK_a's	The neutral species appears late in the titration, after the side chain has already lost its extra proton
pH ≪ lowest pK_a	Essentially 100% cationic (fully protonated)	Henderson–Hasselbalch: pH − pK_a very negative ⇒ ratio [base]/[acid] ≪ 1
pH ≫ highest pK_a	Essentially 100% anionic (fully deprotonated)	Same logic, opposite direction
pH = pI exactly	Cation and anion populated equally (both small); zwitterion (or, for tri-pK_a systems, the correct neutral species) is the dominant single form but not necessarily >99%	Charge-balance condition derived above; separation between flanking pK_a's controls how dominant the neutral form actually is
Solid state / crystal	Zwitterionic, salt-like lattice; high melting/decomposition point	Ionic (Coulombic) lattice energy, not molecular van der Waals packing
Nonpolar solvent (hexane, ether)	Very poor solubility	Charged/dipolar species cannot be solvated without polar/H-bonding solvent

Common trapCommon misconception 1: "The neutral structure H₂N–CH₂–COOH is what's really there, and the zwitterion is just a resonance form or a minor tautomer."

This gets the populations backwards. The two structures are not resonance forms of each other (resonance structures differ only in electron placement over a fixed nuclear framework; here a proton physically moves from one atom to another, and the molecule's formula and connectivity of hydrogens genuinely differ). They are constitutional tautomers/ionization states related by proton transfer, and the equilibrium overwhelmingly favors the zwitterion — by a factor of roughly $10^{\text{p}K_{a2}-\text{p}K_{a1}}$ , which for glycine is $10^{9.60-2.34}=10^{7.26}\approx 1.8\times10^7$ . The "neutral" form is real but present at only about one molecule in 18 million near neutral pH.

Common trapCommon misconception 2: "Amphoteric means 'no net charge,' so an amphoteric species must always be neutral."

Amphoteric describes behavior (the ability to act as either acid or base, depending on what it reacts with), not a fixed charge state. The zwitterion of glycine is amphoteric (its –NH₃⁺ can donate a proton to a strong base; its –COO⁻ can accept a proton from a strong acid) despite carrying two separated formal charges that happen to sum to zero. Water (H₂O) is likewise amphoteric — it can donate a proton to become OH⁻'s conjugate acid or accept one to become H₃O⁺ — without carrying any zwitterionic character at all. The two concepts (amphoteric behavior, zwitterionic structure) are related but logically distinct: amphoteric is about reactivity, zwitterionic is about structure.

The chemist behind it

The zwitterion concept and terminology entered chemistry through the work on amino acid and protein physical chemistry in the early twentieth century, most closely associated with the German physical chemist Wolfgang Ostwald and, more directly, with Niels Bjerrum, the Danish physical chemist who in 1923 formalized the "zwitterion" (dipolar ion) concept to explain the otherwise puzzling physical properties of amino acids — their salt-like melting points, their anomalously large dipole moments in solution, and the shape of their titration curves, which could not be reconciled with a simple neutral covalent structure. Bjerrum's proposal was initially controversial, since organic chemists were accustomed to thinking of carboxylic acids and amines as reacting with external acids and bases, not with each other within the same molecule — an "internal" acid-base reaction was a genuinely novel idea for organic structure at the time. The broader acid-base framework that makes the zwitterion concept rigorous — treating pK_a as a measurable equilibrium constant and using it predictively via what is now called the Henderson–Hasselbalch equation — rests on the independent contributions of Lawrence Joseph Henderson (a physiologist who in 1908 analyzed the carbonic acid/bicarbonate buffer system in blood) and Karl Albert Hasselbalch (a Danish chemist who in 1917 recast Henderson's relationship in logarithmic form using the newly introduced pH scale of Søren Sørensen). It is the fusion of Sørensen's pH concept, Henderson's mass-action buffer analysis, and Bjerrum's zwitterion structure that gives the modern quantitative picture of amino acid ionization used throughout biochemistry.

Origins · 1 minNiels Bjerrum and how this was discovered

Niels Bjerrum

1879–1958 · Danish physical chemist, University of Copenhagen

Bjerrum trained under Wilhelm Ostwald and Walther Nernst before returning to Copenhagen, where he built a career in solution thermodynamics, ionic equilibria, and the theory of strong electrolytes (his "Bjerrum length," describing the distance at which electrostatic interaction energy between two ions equals the thermal energy scale $k_BT$ , remains foundational in the theory of electrolyte solutions and polyelectrolytes, including nucleic acids). His 1923 proposal that amino acids exist as internal salts — dipolar zwitterions rather than neutral amino-carboxylic acids — explained their crystallographic and solution behavior in a way no earlier model could, and the term "zwitterion" he championed (borrowed from German, meaning "hybrid" or "hermaphrodite" ion) has remained the standard name for the species ever since.

History

- 1889 — Svante Arrhenius formalizes the ionic theory of acids and bases, providing the conceptual groundwork for treating dissolved species as discrete charged ions.

- 1908 — Lawrence Henderson analyzes the CO₂/bicarbonate buffer equilibrium in blood using mass-action equilibrium expressions.

- 1909 — Søren Sørensen introduces the pH scale at the Carlsberg Laboratory, enabling quantitative comparison of acidity across systems.

- 1917 — Karl Hasselbalch recasts Henderson's buffer equation in logarithmic (pH) form, yielding the modern Henderson–Hasselbalch equation.

- 1923 — Niels Bjerrum proposes the dipolar zwitterion structure for amino acids, resolving the puzzle of their salt-like physical properties.

Why doesn't the amino acid just release a free proton into the solution instead of transferring it internally?

It effectively does exchange protons with the solvent — the zwitterion is in fast equilibrium with both the cation (via protonation from solvent H₃O⁺) and the anion (via deprotonation to solvent). But because the α-carboxyl (pK_a ≈ 2) is a much stronger acid than the α-ammonium is as an acid (pK_a ≈ 9–10), at any pH between these two values the thermodynamically favored net outcome is that the carboxyl has given up its proton while the amino group has picked one up (from solvent, statistically, not literally the "same" proton). The "internal transfer" framing is a convenient bookkeeping shorthand for the net result of two independent solvent-mediated equilibria, not a single concerted intramolecular proton hop.

Is the zwitterion form present in the solid crystal, or only in aqueous solution?

Both. X-ray and neutron diffraction studies of glycine and other amino acid crystals show unambiguous zwitterionic geometry: C–O bond lengths in the carboxylate that are equal (consistent with a resonance-delocalized COO⁻, not a C=O/C–OH asymmetric pair), and N–H bond geometry consistent with three N–H bonds on a tetrahedral ammonium nitrogen, not two. The crystal is held together by an extended network of ionic and hydrogen-bonding interactions between zwitterions, which is exactly why amino acid crystals behave like salts — high lattice energy, high melting/decomposition temperature, and negligible vapor pressure.

Why is the isoelectric point of aspartate so much lower than that of glycine?

Aspartate has an extra ionizable carboxyl group in its side chain, which introduces an additional negative charge into the titration sequence at low pH. The neutral (zero-net-charge) species appears very early in the titration — after only two of aspartate's three acidic protons have been lost — so the two pK_a's that flank it (1.88 and 3.65) are both low. Averaging them gives pI ≈ 2.77, far below glycine's 5.97, which reflects the fact that aspartate needs a much more acidic environment before its average charge returns to zero.

Does the "amphoteric" behavior of amino acids depend on temperature?

Yes, indirectly. Every pK_a is itself a temperature-dependent equilibrium constant (via $\Delta G^\circ = -RT\ln K_a$ , and $\Delta H^\circ$ for proton dissociation from an ammonium group is typically larger in magnitude than for a carboxylic acid). Ammonium ionizations in particular shift pK_a by roughly −0.01 to −0.03 units per °C increase, meaning physiological temperature (37 °C) versus room temperature (25 °C) measurably shifts the pI of amino acids and proteins. This matters in practice: an isoelectric focusing gel run at one temperature will show a slightly different banding pattern than a calculation using room-temperature pK_a tables would predict.

How is the isoelectric point different from the pH at which the amino acid is "least soluble"?

For simple amino acids they largely coincide, and for proteins the two are often used almost interchangeably in practice, but conceptually they are distinct. The pI is defined purely electrostatically (average net charge = 0). Minimum solubility occurs at the pH where electrostatic repulsion between like-charged molecules is weakest, which allows attractive interactions (van der Waals, hydrogen bonding, hydrophobic contacts) to dominate and drive precipitation or crystallization. Because these repulsive/attractive balances are governed by the same net charge, minimum solubility usually occurs very close to the pI — which is exactly why isoelectric precipitation is a standard first step in industrial protein purification.

If pKa1 and pKa2 were closer together, would the zwitterion still dominate as completely?

No — the derivation shows the fraction of "wrong" species (cation and anion) at any pH scales with $10^{\text{p}K_{a1}-\text{pH}}$ and $10^{\text{pH}-\text{p}K_{a2}}$ , so the width of the pH window over which the zwitterion is the overwhelmingly dominant species depends directly on the gap between pK_a1 and pK_a2. Glycine's gap of 7.26 units gives an enormous dominance window. If a hypothetical amino acid had pK_a1 = 4.5 and pK_a2 = 5.5 (a gap of only 1 unit), the neutral zwitterion would never reach much above roughly 90% of the population even exactly at the pI, and the cation/anion populations would remain chemically significant across a much wider pH range — this is exactly the situation for amino acids with side-chain pK_a's close to the α-groups, or for buffer systems designed with closely-spaced pK_a's.

Why is this concept classified as "foundational" (Tier 1) rather than advanced?

Because every later quantitative treatment of amino acid and protein acid-base chemistry — titration curves, buffering capacity, isoelectric focusing, the electrostatics of enzyme active sites, the design of buffers for protein purification — reduces to correctly applying the two ideas built here: (1) two independent Henderson–Hasselbalch equilibria coexist on the same molecule, and (2) the isoelectric point is defined by a charge-balance condition, not simply "the average pKa" applied blindly. Every more advanced calculation is this same logic extended to more ionizable groups.