A peptide powder has a chemical identity and a physical form, and different instruments answer for each. Chromatography and mass spectrometry establish what the molecule is; neither says anything about whether those molecules sit on a lattice, are frozen into a glass, or lie somewhere between the two. That arrangement governs how the solid takes up water, how quickly it goes into solution, and how it changes in a sealed vial over months. The analytical characterisation of peptide products has been described as presenting challenges beyond those of conventional small-molecule drugs, with the amorphous solid-state form named among them alongside difficult-to-separate impurities and secondary structure [1].
What a diffraction pattern reports
X-ray powder diffraction reports periodicity. A crystal repeats its unit cell over long distances, and that repetition scatters X-rays into sharp Bragg peaks at angles fixed by the cell dimensions. Peak positions fingerprint the packing rather than the molecule, so two polymorphs of one compound give different patterns, and a hydrate, solvate or salt gives a different pattern again. The discrimination available between closely related forms is fine. In binary mixtures of the β and δ polymorphs of mannitol, sample rotation combined with reducing the particle size range below 125 µm halved the limits of detection and quantitation for the δ component, to 1% and 3.6% respectively [2].
Finding a small crystalline fraction inside a largely amorphous solid is a harder version of the same problem, because the crystalline intensity has to be separated from a large diffuse background. Following the isothermal crystallisation of amorphous sucrose in situ between 90 and 160 °C with synchrotron radiation and a two-dimensional detector, an estimated limit of detection of 0.2% w/w crystalline sucrose in the amorphous matrix was reached, against a value of roughly 1% w/w reported for conventional laboratory diffractometry [3]. Both numbers come from a sugar rather than a peptide, and both are specific to that pattern and that background; read them as an indication of the technique's reach, not as a specification for your sample.
The halo, and the states it can conceal
An amorphous solid has no long-range periodicity and so produces no Bragg peaks — only one or two broad diffuse maxima, the amorphous halo. The temptation is to read that halo as positive identification of a glass. Analysis of X-ray amorphous patterns from microcrystalline cellulose, indomethacin and piroxicam concluded instead that observing such a pattern may indicate amorphous, glassy or disordered nanocrystalline material in the sample [4]. Rietveld modelling of one microcrystalline cellulose grade indicated predominantly disordered crystalline cellulose Form Iβ with some amorphous contribution, and an average crystallite size of 10.9 nm [4]. Crystallites that small broaden their own reflections until those reflections merge with the background. A featureless pattern therefore licenses exactly one conclusion — no long-range order detectable at the sensitivity of this measurement — and says nothing about which disordered state you are holding.
Going further means treating the whole scattering curve as signal rather than as background. Coupling X-ray total scattering with the pair distribution function has enabled structural insight into amorphous and nanocrystalline pharmaceuticals and shows promise for fingerprinting, quantification and even modelling, in a domain where conventional powder diffractometry is of limited utility [5]. The same review notes that preparation method, along with sample history, can dictate the nature and the stability of the amorphous phase, and that the International Centre for Diffraction Data holds a collection of diffraction patterns of amorphous drugs and excipients carrying a quality mark on the data [5]. Two lots that are each described as amorphous by XRPD need not be the same glass.
Why peptide solids commonly come out amorphous
Lyophilisation tends to deliver peptides into the glassy state rather than onto a lattice, which is why the amorphous solid-state form appears on the list of characterisation problems specific to peptide products [1]. Two model peptides, Gly-Gln-Asn-Gly-Gly and Val-Tyr-Pro-Asn-Gly-Ala, lyophilised from sucrose or mannitol solutions, gave formulations that differential scanning calorimetry showed to be largely amorphous immediately after drying; the sucrose formulation remained so through storage at 50 °C and 30% relative humidity, whereas the mannitol formulation carried some ordered, crystalline-like structure from the outset and gained more of it during storage, as tracked by infrared spectroscopy and solid-state NMR [6]. In freeze-dried formulations of the model polypeptides glucagon and insulin, characterised by XRPD and DSC, crystallisation of mannitol was directly correlated with a significant reduction in the long-term stability of both polypeptides [7].
Amorphous is not inevitable. Ciclosporin A as received gave clear diffraction peaks, indicating semi-crystalline material, while modulated-temperature DSC resolved a glass transition for the same compound at 107 °C in the reversing heat-flow signal; the authors concluded that ciclosporin A may exist in a range of solid states, with the degree of crystallinity altered by processing [8].
The reason the distinction earns instrument time is that the forms behave differently in water. For several small-molecule drugs, treating the amorphous form as either an equilibrium supercooled liquid or a pseudo-equilibrium glass predicted a solubility advantage over the most stable crystalline form of between 10- and 1,600-fold; the measured advantage was usually considerably less, which was attributed to the difficulty of determining the solubility of an amorphous material under true equilibrium conditions [9]. Those compounds were not peptides, and for a peptide the direction matters more than the magnitude. The same physical instability that raises apparent solubility brings a crystallisation propensity with it [5], so the dissolution behaviour you measure on a fresh lot is not guaranteed to be the behaviour of that lot a year later.
The glass transition, and how residual water moves it
Where the glass transition sits relative to your storage temperature is the most useful single number a thermal method returns, and it is not a constant of the material. Across a wide range of amorphous and partially amorphous pharmaceutical solids, absorbed water produced a rapid initial reduction in the glass transition temperature from the dry value, followed by a gradual levelling off at higher water contents; the effect could generally be described using a simplified form of the Gordon–Taylor/Kelley–Bueche relationships derived from polymer free-volume theory. Most of those systems showed nearly ideal volume additivity and negligible tendency to interact, consistent with water plasticising in the same way as other small molecules rather than through any specific or stoichiometric interaction [10].
Water does at least two separable things to a peptide glass, and a peptide system has been used to separate them. In lyophilised poly(vinylpyrrolidone) formulations of the model hexapeptide Val-Tyr-Pro-Asn-Gly-Ala, glycerol was used as a plasticiser to vary the formulation glass transition temperature without significantly changing water content or activity. Deamidation rate rose with both moisture and glycerol content, and was strongly correlated with Tg at constant water content and activity, which implicates matrix mobility. In rubbery systems above Tg, rates appeared independent of water content and activity among formulations of similar Tg; in glassy formulations of similar Tg, by contrast, deamidation increased with water content, suggesting a solvent or medium effect distinct from plasticisation. Higher water content also changed the degradation product distribution, giving less of the cyclic imide intermediate and more of the isoAsp and Asp hexapeptides [11].
That is why a Karl Fischer result and a thermogram belong in the same record. The peptide study above measured water by coulometric Karl Fischer titration alongside thermogravimetry, DSC, infrared and solid-state NMR [6]; the companion guides in this library on Karl Fischer titration and on moisture sorption cover how that water number is obtained and why it depends on the humidity the powder last equilibrated with. A Tg quoted without the water content it was measured at is not a property of the solid.
What thermal analysis adds, and what diffraction misses
A crystalline solid melts; an amorphous one shows a step change in heat capacity at Tg. Modulated-temperature DSC separates that reversing heat-capacity step from non-reversing events overlapping it, which is how the ciclosporin A glass transition described above was assigned, supported by temperature cycling studies [8]. The same study makes the complementary point in the other direction. In ciclosporin-loaded poly(D,L-lactide-co-glycolide) microspheres, no diffraction peaks were seen, yet the glass transition corresponding to the peptide was still observed, suggesting that the peptide was present as a separate amorphous phase, with little or no apparent miscibility between it and the polymer, and scanning electron microscopy showed distinct surface islands whose size grew with loading [8]. Diffraction returned a halo; calorimetry resolved two phases.
Solid-state NMR and the local environment
Cross-polarisation magic-angle-spinning 13C NMR reports on the chemical environment of individual carbons and does not need long-range order to give resolved signals, so it separates forms that diffraction cannot. Applied to the sweetener neotame, it quantified the amount of each of two anhydrous polymorphs in binary mixtures to within 1–2%, and amorphous content in amorphous/crystalline mixtures to within 5% [12]. That work also reports the failure mode limiting every quantitative solid-form method: the crystalline standards used to prepare the mixtures were not pure crystalline forms but themselves mixtures of crystalline and amorphous material, which propagates straight into the calibration. Differences in relaxation parameters and cross-polarisation efficiency between forms have to be addressed before peak intensities are read as concentrations [12]. Solid-state NMR also reaches through a matrix: amorphous peptide has been detected non-destructively inside an intact coated-sphere formulation [1].
Vibrational spectroscopy and the order it senses
Infrared tracked the growth of ordered structure in the mannitol peptide formulation described above on storage [6], which is the mid-frequency regime doing what it does well. Low-frequency Raman works differently: it reports lattice phonon modes, which exist only where there is a lattice. Milling indomethacin and tryptophan, separately and as a mixture, and following the transformation by low-frequency Raman (10–250 cm-1), conventional Raman, FTIR and XRPD, every technique detected the transformations, but low-frequency Raman and XRPD provided less ambiguous signatures than the mid-frequency methods, with intense phonon-mode bands from the crystalline and cocrystalline samples serving as a sensitive probe of order [13].
Cake structure is not physical form
A lyophilised cake has a visible macrostructure — height, shrinkage, fissures, collapse — and this gets read as a proxy for what happened during drying. It is a weak proxy. In a systematic study on an IgG1 monoclonal antibody, a second protein and lactic dehydrogenase, collapse was induced both by varying the mannitol-to-sucrose ratio and by running an aggressive cycle against a conventional one, with the extent of collapse quantified as the decrease in specific surface area. Collapsed cakes had residual moisture levels comparable to non-collapsed lyophilisates, reconstitution times were not increased, and protein stability was not relevantly different between the two [14]. Those analytes were proteins rather than synthetic peptides, but the methodological point holds: appearance and physical form are separate measurements, and a collapsed cake is a reason to measure, not a result.
Form also changes on storage, and what diffraction follows may not be the peptide. In a lyophilised formulation of the peptide hormone secretin, XRPD together with thermal and spectroscopic methods demonstrated polymorphic transitions of mannitol and increased crystallinity in the formulations with time; in the arm held at 25 °C and 60% relative humidity, particle size measured by dynamic light scattering after reconstitution rose from about 390 nm at day 0 to above 2 µm as early as week 1, and reconstitution time went from about 20 s at day 0 to about 67 s at week 8 [15]. What XRPD followed there was the excipient, which is often the only crystalline component present and therefore the only component the technique can follow.
Sometimes diffraction sees nothing at all when something real has changed. Lots of a lyophilised recombinant human interleukin-11 formulation that differed in dissolution behaviour on reconstitution because of differing processing conditions were examined by powder XRD, DSC, inverse gas chromatography and thermally stimulated depolarised current. XRD, DSC and IGC all failed to distinguish the solid structures; only TSDC discerned differences, which it attributed to molecular mobility and micro-heterogeneity of the amorphous components [16]. The underlying differences were slight — even the cloudiest reconstituted solutions cleared within several minutes — but the analytical lesson is not. A matching diffractogram between two lots is evidence of matching long-range order, not of matching solids.
Artefacts you introduce during sample preparation
Two preparation errors dominate, and the usual remedy for the first causes the second. The first is preferred orientation. Crystallites with plate or needle habits align when packed into a flat holder, so some reflections are over-represented and others suppressed, and relative peak intensities stop corresponding to the structure. All three common mannitol polymorphs have needle-like morphology; coupling sample rotation with particle size reduction was what made those two very similar polymorphs separable at all, and in that same work sample packing and mixing errors contributed the greatest source of variation among the error terms examined [2]. Rotation is the cheap diagnostic — if intensities move when the holder spins, orientation is present.
The second is that grinding to randomise orientation can destroy the form you are trying to measure. Grinding the α, γ and δ forms of indomethacin produced, with increasing grinding time, a two-phase amorphous-plus-crystalline system that continually transformed into a completely random close-packed amorphous structure; grinding piroxicam Form II first gave a disordered phase retaining the local packing of Form II over a very short nanometre length scale, consistent with a continuous random network glass, which persisted to a critical point before transition to the fully amorphous structure [4]. How fast this happens is compound- and mixture-specific: for indomethacin only partial amorphisation had occurred after 360 minutes of milling, and the amorphisation kinetics were faster for the drug–amino acid mixture than for the components alone [13]. Hand-grinding lightly under a spatula, sieving rather than milling, and side-loading or capillary mounts are the usual compromises. Whichever you use, record it, because a diffractogram is comparable only to another one prepared the same way.
Two further handling points follow from the water story above. A hygroscopic amorphous peptide equilibrates with laboratory humidity while it sits open on the mount, so a long scan on an open holder is a measurement of a moving target; a sealed or domed holder, or a humidity-controlled stage, removes that variable. And because absorbed water depresses Tg steeply at low water contents [10], a sample that sat out during preparation may crystallise on the stage during a long exposure — a repeat scan is worth running before any pattern is called stable.
What the record should contain
A defensible solid-form record for a peptide lot pairs the diffractogram with the conditions that produced it and with at least one orthogonal method. State the geometry, radiation, scan range and step, and the mount and preparation, including whether the sample was ground and whether it was rotated. Report water content by Karl Fischer on the same lot with the same handling, and report the glass transition with the water content at which it was measured rather than on its own. If the pattern is a halo, say what that excludes rather than what it proves, and reach for calorimetry, solid-state NMR or a vibrational method to establish which disordered state you have. If a crystalline phase is present, identify whether it is the peptide or an excipient before drawing any conclusion about the peptide itself.