Amorphous or crystalline? X-ray powder diffraction and the solid-state character of lyophilized peptides

A lyophilized peptide vial is routinely described by its appearance — a white cake, a fluffy powder, a collapsed film — and by its chemical purity. What is described far less often is the physical form of the solid itself: whether the material is arranged in an ordered crystalline lattice, exists as a disordered amorphous glass, or is some mixture of the two. This distinction is not cosmetic. Solid-state form influences dissolution behavior, water uptake, the rates of several degradation pathways, and the reproducibility of a lot from one freeze-drying run to the next.

X-ray powder diffraction (XRPD) is the standard technique for interrogating this property. This post reviews what a diffraction pattern represents, why most lyophilized peptides produce featureless patterns, how semi-quantitative statements about order are made, and where the technique reaches its limits.

What a powder diffraction pattern measures

When a monochromatic X-ray beam strikes a sample, atoms scatter the radiation. In a crystalline solid, atoms are arranged in repeating planes separated by characteristic distances, and scattered waves reinforce one another only at specific angles that satisfy Bragg’s law, nλ = 2d sinθ. A powder contains many small crystallites in random orientations, so scanning the detector across a range of 2θ angles yields a series of sharp peaks whose positions reflect lattice spacings and whose relative intensities reflect the arrangement of atoms within the unit cell. The result is often described as a fingerprint of a crystalline phase.

An amorphous solid lacks long-range periodicity. Short-range order — the typical distances between bonded and neighboring atoms — still exists, so scattering is not absent, but it is spread across angles rather than concentrated at discrete values. The characteristic output is a broad, diffuse feature, commonly called an amorphous halo, usually centered at low-to-moderate 2θ for organic materials. A pattern that shows only a halo and a smooth baseline is consistent with a material that has no detectable crystalline domains under the measurement conditions.

For peptides, the practical consequence is that the technique answers a specific question: does this solid contain ordered domains large enough and abundant enough to produce Bragg reflections? It does not directly report on chemical identity, and it says nothing about the peptide’s secondary structure in solution.

Why lyophilized peptides are typically amorphous

Freeze-drying converts a solution into a porous solid by freezing, sublimation of ice under vacuum, and desorption of bound water. Whether the solute crystallizes during this process depends on its molecular properties and on the formulation. Peptides present several obstacles to crystallization: conformational flexibility, heterogeneous charge distribution, and the presence of closely related impurities such as deletion sequences and diastereomers that disrupt lattice formation. During freezing, increasing solute concentration in the unfrozen fraction raises viscosity sharply, which kinetically traps the peptide in a disordered state before nucleation can occur.

As a result, the large majority of synthetic peptides isolated by lyophilization yield diffraction patterns dominated by an amorphous halo. This is generally regarded as expected behavior rather than a deficiency. Amorphous solids typically dissolve faster than crystalline forms of the same compound and can be prepared with a high specific surface area. The trade-off is thermodynamic: an amorphous glass is a non-equilibrium state with higher free energy, higher molecular mobility, and a greater tendency to absorb water than a corresponding crystal.

Formulation components change the picture. Bulking agents such as mannitol and glycine are known to crystallize during freeze-drying under certain conditions, and when they do, the diffraction pattern of the cake will display sharp peaks that belong to the excipient rather than the peptide. Reading such a pattern correctly therefore requires reference patterns for each formulation component, and ideally for the individual polymorphs of crystalline excipients — mannitol alone is known to exist as multiple polymorphs and a hydrate, each with distinct reflections. Sugars used as lyoprotectants, by contrast, are chosen precisely because they tend to remain amorphous and form a glassy matrix around the peptide.

Reading a pattern: halos, peaks, and semi-quantitative statements

Interpreting an XRPD scan of a lyophilized peptide usually involves three questions. First, are there sharp reflections at all? Their presence indicates crystalline material, and matching peak positions against reference patterns can assign them to a salt, an excipient, or a residual process-related crystalline phase such as a buffer component. Second, if peaks are present, how do they compare in width? Broad peaks suggest very small crystallites or poorly ordered domains, whereas narrow peaks suggest larger, better-ordered ones. Third, how does the intensity of the crystalline contribution compare with that of the halo?

Semi-quantitative estimates of crystallinity are sometimes derived by integrating the area under crystalline peaks relative to the total scattered area, or by fitting calibration mixtures of fully amorphous and fully crystalline reference material. These approaches have well-characterized limitations. They depend on the choice of baseline and peak-fitting model, on sample preparation, and on the availability of a truly 100% crystalline and a truly 100% amorphous standard, which for a peptide may not exist. For that reason, reports commonly describe patterns in qualitative terms — “amorphous,” “amorphous with crystalline excipient reflections,” “partially crystalline” — rather than as a precise percentage.

Detection limits matter here as well. Laboratory diffractometers can generally detect crystalline phases present at a few percent by weight under favorable conditions, but small crystalline fractions within a predominantly amorphous matrix may fall below that threshold. A pattern described as amorphous is best read as “no crystalline phase detected above the method’s limit,” not as proof that order is entirely absent.

Sample handling and artifacts that influence the result

Because amorphous lyophilized solids are hygroscopic, the way a sample is prepared for diffraction can alter the very property being measured. Exposure to ambient humidity during grinding, transfer, or mounting can plasticize the glass, lower its glass transition temperature, and in some systems trigger crystallization of a component that was amorphous in the sealed vial. Studies of amorphous pharmaceutical solids have documented that brief exposure to humid air can be sufficient to change the observed pattern. Sample preparation under controlled low humidity, sealed or domed holders, and minimal mechanical grinding are therefore standard precautions. Grinding itself is a concern in the opposite direction: sufficient mechanical energy can induce disorder in a crystalline phase, lowering apparent crystallinity.

Preferred orientation is another recognized artifact. Plate-like or needle-like crystallites tend to align when packed into a holder, which distorts relative peak intensities even though peak positions remain correct. Rotating the sample or using capillary geometry reduces this effect. Finally, low-angle background from the holder, air scatter, or a glass capillary can mimic or obscure the broad amorphous feature, so a blank measurement is generally collected for comparison.

Complementary techniques and what the solid-state picture does not tell us

XRPD is rarely used in isolation. Differential scanning calorimetry identifies the glass transition and any crystallization or melting events, and it is closely linked to the collapse behavior discussed in work on glass transition and collapse temperature in lyophilized solids. Dynamic vapor sorption quantifies water uptake and can reveal humidity-induced transitions. Residual moisture by Karl Fischer titration provides the bulk water content that shapes the glass transition of the cake. Polarized light microscopy offers a quick check for birefringent crystalline domains, and infrared or Raman spectroscopy can report on secondary structure and intermolecular interactions within the solid.

Taken together, these methods connect structure to behavior. Several studies on protein and peptide solids have characterized an association between molecular mobility in the glassy state and the rate of degradation reactions such as deamidation or aggregation, though the relationship is system-specific and mobility is not the only determinant. Diffraction does not measure mobility; it reports on order. A lot that is uniformly amorphous may nonetheless differ between runs in residual moisture or in glass transition temperature, and these differences would not appear in an XRPD scan.

Interpreting solid-state data on a technical package

For a reader evaluating analytical documentation, the existence of an XRPD result raises several practical questions. Which formulation components were present, and were their reference patterns used for assignment? Was the sample handled under controlled humidity? What 2θ range and step size were used, and what is the stated detection limit for crystalline content? Was the pattern collected on the sealed-vial contents shortly after opening, or on material stored in an unspecified container? Without this context, the statement “amorphous by XRPD” is correct but incomplete.

The broader point is that physical form is a property of the whole lot history — formulation, freezing rate, drying cycle, storage humidity — rather than of the peptide sequence alone. Diffraction offers a direct, nondestructive view of one dimension of that history, with defined sensitivity and defined blind spots. Read alongside thermal, moisture, and chemical purity data, it helps describe what a lyophilized solid is, while leaving open the question of how that solid will change over time under conditions the measurement did not sample.