Electron Diffraction: The Missing Link Between SC-XRD and XRPD

Fri, 27.02.2026

X-ray diffraction has been the backbone of solid-form characterization in pharmaceutical development for over a century. Two methods dominate the toolkit, and they are fundamentally different in what they measure, what they deliver, and where they fall short. Understanding both – and the gap between them – is the starting point for understanding where electron diffraction fits.

Single-crystal XRD: complete structural information, one crystal at a time

Single-crystal X-ray diffraction (SC-XRD) delivers the most complete structural picture available. By directing an X-ray beam at a single crystal and collecting diffraction data from multiple orientations, it yields the full three-dimensional crystal structure: unit cell parameters, space group, and atomic positions. This is the definitive answer to what polymorph you have, how the molecule packs, and what the conformation is.

The constraint is physical. SC-XRD requires a crystal large enough to generate sufficient diffraction signal – typically tens to hundreds of micrometres. A compound that forms only nanocrystals, or whose crystal size has been reduced by milling or spray drying, cannot be mounted and measured. When crystals are too small, the structural question simply goes unanswered.

X-ray powder diffraction: bulk phase characterization, but with limitations

X-ray powder diffraction (XRPD) takes a different approach entirely. It measures the combined diffraction signal from millions of crystals in a bulk powder – the material as it actually exists in development and manufacturing. The output is a characteristic diffraction pattern that identifies the dominant crystalline phases present across the whole sample. XRPD is the standard tool for confirming solid form in a manufactured batch, monitoring form changes during process development, and supporting regulatory submissions. For regulatory requirements relating to XRPD in pharmaceutical development, see the ICH Q6A guideline.

In principle, crystal structure determination is also possible from powder diffraction data. Methods exist to index the pattern, determine the unit cell, and solve and refine a structure from the resulting data. In practice, this is technically demanding. It requires high-quality data, a pure sample, and considerable expertise in powder crystallography. The process is sensitive to peak overlap – a consequence of the one-dimensional nature of a powder pattern compared to the three-dimensional data available from a single crystal – and ambiguities in indexing and structure solution multiply when data quality is imperfect.

These challenges become substantially harder when the sample is not phase pure. If a minority crystalline phase is present alongside the dominant form, its reflections overlap with those of the majority phase, complicating unit cell determination and making structure solution from the powder pattern unreliable or impossible without additional information. The practical limit is compounded by a concentration issue: phases present at below approximately 1–5% by mass are often not reliably detected at all, and the exact limit of detection for a specific minority phase in a specific matrix is not known in advance and must be empirically established. A trace polymorph at that level may be present and consequential – for stability, bioavailability, or intellectual property – while remaining invisible to the bulk measurement.

The gap between the two methods

Put simply: SC-XRD delivers full structural information but requires large, well-formed single crystals. XRPD works with the powders that pharmaceutical development actually produces but provides bulk-level phase data without single-crystal structural resolution, structure solution from powder is only feasible under favorable conditions, and minority phases may fall below detection limits. For many compounds under routine development conditions, that gap creates no problem. But for a nanocrystalline API, a compound with suspected trace polymorphism, or a material where XRPD yields ambiguous or incomplete results, the gap matters.

3D electron diffraction: structural resolution applied to powder-scale samples

This is where electron diffraction enters – not as a replacement for either X-ray method, but as the technique that connects the two.

Electrons interact with matter roughly 1,000 times more strongly than X-rays of comparable energy. In 3D electron diffraction (3D-ED), a focused electron beam is directed at an individual crystal in a standard powder sample – typically 50–1000 nm thick – while the sample is continuously rotated. Diffraction data from multiple orientations of that single crystal yields the same output as SC-XRD: unit cell parameters, space group, and atomic positions [1]. The full crystal structure, solved from a crystal too small to mount in an X-ray diffractometer, from the same powder you would submit for XRPD.

That combination – single-crystal structural resolution, applied to the nanocrystalline and powder materials that XRPD works with – is what opens capabilities that neither X-ray method can offer alone.

How can this become useful for the pharmaceutical industry?

Because 3D-ED delivers single-crystal structural resolution from standard powder samples, it opens several applications that are either inaccessible or unreliable with X-ray methods alone. The following are the core areas where ELDICO’s CRO service adds value in pharmaceutical development.

Many active pharmaceutical ingredients form only nanocrystals during synthesis, spray drying, or milling – crystals too small to mount for SC-XRD. 3D-ED solves the full structure directly from these materials, providing unit cell parameters, space group, and atomic positions without requiring a larger crystal.

Example: dabigatran etexilate mesylate (the active ingredient in Pradaxa) exists in multiple polymorphic forms. Two anhydrous forms and a hemihydrate were characterized in full using 3D-ED – including complete crystal structures of all three room-temperature-accessible forms – directly from nanocrystalline powder, without the need for large single crystals. The study provided the structural basis for selecting anhydrous form I for development, despite it being metastable, based on its superior bulk processing properties and demonstrated long-term stability [4].

• Applicable at any stage where crystal growth to X-ray size is not feasible
• Deliverable: CIF structure file, data quality report, publication-ready figures on request
• Typical turnaround: 2-4 weeks from sample receipt
• Supports pharmaceutical form selection, CMC package development, and IP landscape mapping
• Deliverable: form identification report, comparative structural analysis, recommendations for further study
• Applicable to process impurity identification, degradation product characterization, and raw material qualification
• Detection not limited by concentration – limited only by the number of crystals surveyed
• Relevant for amorphous solid dispersions, spray-dried intermediates, and milled materials
• Provides structural identity of nanocrystalline domains, not just detection

Different crystal forms of the same molecule can differ substantially in solubility, stability, and bioavailability. Identifying all relevant forms – and understanding their structures – is a core requirement of pharmaceutical solid-form development and IP protection. 3D-ED characterizes each form individually at the structural level, including forms present as minor components that XRPD may not resolve.

Example: in a collaboration with Ardena, 3D-ED was applied to characterize multiple solid forms of a pharmaceutical compound where XRPD patterns could not be resolved independently. Electron diffraction identified each form at single-crystal resolution, providing structural data that supported form selection and CMC decision-making [5]. 

Because 3D-ED operates crystal by crystal, concentration in the bulk sample is not a limiting factor. A minority crystalline phase – a trace polymorph, a process impurity, a degradant – that falls below XRPD’s detection limit is structurally accessible if a crystal of that phase can be located and measured. Automated data collection increases the number of crystals surveyed, which directly increases the probability of detecting and characterizing rare phases [2].

Example: automated 3D-ED experiments on a commercial sample of dapsone – an antimicrobial API with five previously known polymorphic forms – surveyed approximately 650 individual crystals in 24 hours. Around 3% of crystals returned an unknown unit cell. Structure solution confirmed a new polymorphic form (Form VI) of Dapsone, crystallizing in the monoclinic P2(1) space group with four independent molecules in the asymmetric unit – a form not predicted by prior computational screening [6]. In a separate study, 3D-ED was used to solve the crystal structure of ritonavir form 4, a metastable polymorph that had resisted characterization since its discovery and could not be isolated as a single crystal of sufficient size for SC-XRD. The structure was solved by combining 3D-ED data with crystal structure prediction and experimental XRPD, revealing a disordered motif and confirming the form is thermodynamically less stable than forms 1 and 2 [7].

Materials that appear amorphous or poorly crystalline by XRPD may contain discrete nanocrystalline domains that are structurally ordered at the local level. This distinction matters in formulation and stability work: the presence of nanocrystalline nuclei in an ostensibly amorphous matrix can be a precursor to crystallization on storage, with direct consequences for bioavailability and shelf life.

An example is a case study by ELDICO Scientific and RCPE, which compared melt quenching and ball milling as preparation techniques for amorphous APIs, using Carvedilol as a model compound. Using automated electron diffraction crystal mapping, ~100 particles per sample were analyzed on a TEM grid. The melt-quenched sample proved fully amorphous, while the ball-milled sample still contained residual crystallinity, detectable and matchable to the Carvedilol unit cell. This not only demonstrated that ball milling time needed to be extended, but also showcased how sensitive electron diffraction crystal mapping is in detecting trace crystalline phases — critical for ensuring product quality and regulatory compliance.  [2].

A complete workflow

XRPD remains the right tool for bulk phase characterization. SC-XRD remains the gold standard where large single crystals are available. 3D-ED adds the capability that connects them: structural information from the powders and nanocrystalline materials that conventional X-ray methods cannot fully address. The three methods are complementary, and a complete solid-form characterization strategy for a compound with meaningful polymorphic risk will typically draw on more than one.

If characterization is incomplete

 ELDICO’s CRO service provides easy access to 3D electron diffraction without the need for in-house instrumentation. If you’re working with a nanocrystalline API, a compound where SC-XRD has not been achievable, XRPD results are ambiguous, or minority phase detection is a concern — ELDICO is you partner. A free feasibility testing is available, making it straightforward to explore whether 3D electron diffraction is the right solution for your sample.

We’re happy to discuss the sample and whether electron diffraction is the appropriate next step before any project commitment.

Contact us: sales@eldico.ch

References

[1] Simoncic, P., Romeijn, E., Hovestreydt, E., Steinfeld, G., Santiso-Quiñones, G. & Merkelbach, J. Electron crystallography and dedicated electron-diffraction instrumentation. Acta Cryst. E79, 410–422 (2023). https://doi.org/10.1107/S2056989023003109

[2] Merkelbach, J., Stam, D., Jandl, C., Kushwah, V., Paudel, A. & Simoncic, P. Crystal mapping using the ELDICO ED-1 electron diffractometer. Imaging & Microscopy (2023). https://analyticalscience.wiley.com/do/10.1002/was.0004000393

[3] Holzwarth, U. & Gibson, N. The Scherrer equation versus the ‘Debye–Scherrer equation’. Nat. Nanotechnol. 6, 534 (2011). https://doi.org/10.1038/nnano.2011.145

[4] Sieger, P., Werthmann, U. & Saouane, S. The polymorph landscape of dabigatran etexilate mesylate: Taking the challenge to bring a metastable polymorph to market. European Journal of Pharmaceutical Sciences, 186, 106447 (2023). https://doi.org/10.1016/j.ejps.2023.106447

[5] ELDICO Scientific AG & Ardena. Application note: solid form characterisation by 3D electron diffraction. Available on request from ELDICO Scientific.

[6] ELDICO Scientific AG. New polymorphic form of dapsone found with automated electron diffraction (microED) experiments on a commercial sample. Application Note AN18 (2025). https://5756737.fs1.hubspotusercontent-na1.net/hubfs/5756737/application%20notes/AN18%20-%20ELDICO-AutomatedED-Dapsone.pdf

[7] Iuzzolino, L., Kelly, A.W., Chaudhry, M.T., Jandl, C., Stam, D. et al. Predicting the ritonavir crisis by revisiting the polymorph landscape with crystal structure prediction and form 4 structure solution. Communications Chemistry, 8, 404 (2025). https://doi.org/10.1038/s42004-025-01814-6