Does UDCA Have Polymorphs? What Milling Actually Changes

Ask a procurement team what keeps them up at night about crystal form and you will usually hear about polymorphs: the wrong form shipping, the form changing in storage, a batch that behaves differently on the tablet press. On ursodeoxycholic acid, that worry is largely misplaced. The real crystal-form risk on this material is not polymorphism. It is partial loss of crystallinity during milling and micronisation, and no pharmacopoeial test will catch it.

That distinction matters because it decides which question you ask your supplier, and whether their answer can be verified at all.

Why the polymorph question is the wrong first question

UDCA is supplied as a single crystalline form, and the industry treats that as a design requirement rather than as an accident. The clearest evidence is in the patent literature: one programme run by a European pharmaceutical company is titled, in plain terms, optimised synthesis of pure, non-polymorphic, crystalline bile acids with defined particle size. Read that title twice. Being non-polymorphic is the goal, and particle size is written into the same sentence as the crystal criterion.

That programme screened the solid state thoroughly, and its screening table is worth looking at for a different reason than the one you might expect: the forms it found — solvates, mixtures, forms that reverted to the starting form within days at ambient conditions — concern a different bile acid derivative, not UDCA itself. The lesson from the table is not that UDCA has hidden forms. It is that when a route throws off crystalline variety, the industry's response is to eliminate it deliberately, and it succeeds.

So the practical answer to "does UDCA have polymorphs" is: not in any way that routinely shows up in commerce. Writing a polymorph limit into a purchase specification for this material buys you a test you will never be able to make fail.

Why milling is the step that changes the solid

Particle size reduction is where the solid state actually gets disturbed. Milling and micronisation deliver mechanical energy into the crystal lattice, and a fraction of the material can be converted from the crystalline to the amorphous state. The longer and more aggressive the milling, the larger that fraction can become.

The consequence is a batch that satisfies every monograph test and still behaves differently on the press. Amorphous content is more soluble, less stable and more prone to moisture uptake than the crystalline form, so a batch with a higher amorphous fraction can show up as a compression problem, a dissolution difference or a shorter effective shelf life, without a single out-of-specification result to point at.

Test	What it detects	On a routine certificate	Comment
X-ray powder diffraction	Loss of crystallinity; amorphous halo; new reflections	No	The direct measurement, and the one that costs least to add
Differential scanning calorimetry	Melting behaviour, and an exothermic event if amorphous content recrystallises on heating	No	Useful as a production-control fingerprint, not as a rejection test
Melting range	Gross identity and purity	Usually yes, if specified	Wide range can indicate a disturbed lattice, but a narrow range does not prove full crystallinity
Assay, related substances, rotation	Chemical identity and impurities	Yes	Chemically silent on solid-state change

Notice the shape of that table. Every test on a standard certificate sits in the bottom row. The two tests that would tell you what happened to the lattice are the two that are almost never on the certificate.

Where this connects to the particle size conversation

Particle size and crystallinity are two outputs of the same step, which is why our note on particle size distribution and D90 treats the specification as an instruction to the mill rather than as a property of the chemical. When you specify a D90, you are also specifying how much mechanical energy will be put into the crystals. A tighter D90 can mean a harder milling regime, and a harder milling regime can mean a larger amorphous fraction.

That is a design trade-off, not a defect. It is fine as long as both sides know it exists. It stops being fine when a buyer specifies a finer powder to solve a blending problem and later finds a compression problem, with a clean certificate sitting in the file.

Polymorph, solvate and hydrate are three different words

The conversation gets easier once the vocabulary is separated, because the three cases carry different risks and call for different tests.

• A polymorph is the same molecule packed in a different lattice, with nothing else inside the structure. It is a genuine alternative solid form, and the transition between two polymorphs is a change of internal arrangement.
• A solvate or a hydrate has a second species built into the lattice — solvent or water respectively. The older literature calls these pseudopolymorphs, a name that has caused more confusion than it prevented, because a hydrate is not a polymorph at all.
• An amorphous form has no long-range lattice order. It is not a crystal form that failed; it is a different physical state, with different solubility, stability and mechanical behaviour.

For UDCA the practical value of the distinction is this: the material can take water into its lattice under the wrong storage conditions without any polymorphic transition occurring. That is a hydrate question, and it shows up first as a drying-loss figure that has drifted rather than as a diffraction pattern that has changed. The solid-state data package is what separates the two cases, and asking for it in the right form is the whole point of the request.

What a solid-state data package should contain

A supplier who can genuinely answer the crystallinity question already holds these measurements. Asking for them is not an unusual request; it is a first-article request, and it is the same package a formulation team would want before running a compression trial.

Measurement	What it establishes	Sensitivity in practice
XRPD against a reference diffractogram	The crystalline form, and whether it has changed	An amorphous fraction becomes visible in the pattern at roughly a few per cent
DSC thermogram	Melting behaviour, and an exotherm where amorphous material recrystallises on heating	Comparable in order of magnitude, and instrument-dependent
Gravimetric vapour sorption	Water uptake against humidity, and the humidity at which a form change begins	Reaches amorphous fractions below the diffraction threshold, which is why it is worth asking for on a milled batch
Particle size distribution by laser diffraction	The milling outcome, and therefore the mechanical energy delivered to the lattice	Not a crystallinity measurement, but the variable that predicts the crystallinity risk

The last row is the one that ties the package together. Particle size is what the customer specified, and it is also the process variable that decides how much energy the mill put into the crystals. A supplier who reports a D90 without a solid-state measurement has reported the input to the risk and not the outcome.

How to get this under control

1. Stop specifying polymorph limits for UDCA. Replace that requirement with something measurable, and free up the negotiation room for the two things that actually move: particle size and crystallinity.
2. Ask what milling route produced the batch you are being quoted. Jet milled, ball milled and sieved material are three different solid-state histories, even when the D90 matches.
3. Ask for XRPD or DSC data on any first article, and on any batch where the particle size specification changed. One measurement at the point of change covers the risk; routine testing of every batch is unnecessary.
4. If the material is going onto a press, evaluate the actual batch, not the specification. Compression behaviour follows the solid state, and the solid state is a process output. Our note on D90 and tabletability covers where the two meet.
5. Keep the moisture number in the same conversation. An amorphous fraction takes up water faster, so a rising drying-loss figure between shipments can be the first visible sign of a crystallinity change. The distinction between drying loss and water content is what tells you which direction to look.

The general rule worth carrying elsewhere

Pharmacopoeial monographs — among them the European Pharmacopoeia text for this substance and the USP ursodiol monograph — test chemistry. They were written to control identity, purity and content, and they do that well. They were not written to control the solid state, and for a crystalline API that is milled to order, the solid state is exactly where a supplier's process variables land.

That is not an argument against the monograph. It is an argument for knowing which of your risks the monograph covers and which ones you have to cover yourself with a letter, a data request, or a first-article evaluation. On this material, crystal form is a chemistry question with a settled answer. Crystallinity is a process question with an unsettled one, and it will not announce itself.

Frequently asked questions

Is UDCA polymorphic?

Not in any way that routinely appears in commercial supply. The material is produced and supplied as a single crystalline form, and the patent literature shows the industry treating non-polymorphic crystallinity as a design goal rather than as a lucky outcome. Screening programmes that do find crystalline variety in this family describe other bile acid derivatives.

Why does the pharmacopoeia not test for crystal form?

Because the monograph exists to control identity, purity and content, and crystal form is not required to establish those on this material. The melting range gives a coarse check on the lattice, but there is no pharmacopoeial test for the amorphous fraction, which is why a conforming certificate can coexist with a disturbed solid state.

Can milling change a batch that passed release testing?

Yes. Milling and micronisation put mechanical energy into the lattice and can convert a fraction of the material to the amorphous state. Because no routine release test measures that fraction, the batch can be released legitimately and still differ in compression behaviour from an earlier batch with the same chemical results.

Which tests detect amorphous content?

X-ray powder diffraction, which shows the loss of crystallinity as a diffuse halo, and differential scanning calorimetry, where amorphous content can appear as an exothermic recrystallisation event on heating. Neither is a standard release test, and either can be requested as a first-article measurement.

Does crystal form affect tablet compression?

The amorphous fraction does, because it changes plasticity, solubility and moisture behaviour relative to the crystalline material. Two batches with identical assay and impurity results can compress differently if their solid-state histories differ.

What should I specify if I need a defined crystal form?

For UDCA, specify the particle size distribution you need and ask for the solid-state data that shows how it was produced. A polymorph limit on this material is a requirement that cannot be meaningfully tested, and it displaces attention from the two variables that genuinely move the outcome.

Where this leaves you

The crystal-form question on this material has a short answer, and the answer is that it is the wrong question. The right one is what happened to the lattice between the crystalliser and your drum, and the honest response from any supplier is a first-article solid-state measurement rather than an assurance. Judge the answer on whether the measurement exists.

What we release against, and how a batch's particle size and solid-state history are documented, is set out on our ursodeoxycholic acid page. If you want solid-state data discussed against your own process requirements before an order, that is a legitimate request and you can make it through a quotation request.

This article describes solid-state properties, particle engineering and pharmacopoeial testing for quality, formulation and procurement professionals. It states no therapeutic use, dose or indication for any substance, and no order quantity or lead time. Confirm all pharmacopoeial requirements against the edition in force.