GiveMeED: Enabling 3DED on Conventional TEM

• The paper presents GiveMeED as a continuous-rotation electron diffraction script that automates stage movement, beam control, and metadata capture to enable high-resolution 3DED data collection.
• It bridges microscopists and crystallographers by producing data compatible with SCXRD workflows, facilitating structure determination of sub-micron crystals.
• The software reduces operator error through automation and integrates seamlessly with existing crystallographic tools, supporting comprehensive analysis via low-dose data acquisition.

GiveMeED is a practical, modifiable script for controlled three-dimensional electron diffraction (3DED) data acquisition on a conventional transmission electron microscope (TEM). It is designed to remove a central barrier to wider 3DED adoption: the need for bespoke software to coordinate continuous sample-stage rotation, uninterrupted diffraction acquisition, metadata capture, and low-dose control. In the reported implementation, GiveMeED enables a conventional JEOL TEM plus a Gatan camera to function as an electron diffractometer, and the collection of useable reflections beyond $0.8\ \text{\AA}$ makes 3DED crystallographic processing effectively routine through standard software and workflows derived from single-crystal X-ray diffraction (SCXRD) techniques (Weare et al., 14 Jul 2025).

1. Position within 3DED and the problem it addresses

3DED is used for crystal structure solution of sub-micron sized crystals that are too small for structure determination via SCXRD or PXRD. The paper situates GiveMeED within this methodological gap: 3DED is powerful, but its uptake has been limited because routine use generally requires a TEM adapted with specialized software for synchronized stage rotation and diffraction movie acquisition. GiveMeED addresses that coordination problem directly by providing a software layer for controlled continuous-rotation electron diffraction on standard TEM platforms (Weare et al., 14 Jul 2025).

The software is explicitly framed as a bridge between several technical communities. For microscopists, it provides a straightforward workflow for collecting useful diffraction data under low-dose conditions. For crystallographers, it delivers data in formats compatible with standard SCXRD-style software. For sample scientists, it enables structural analysis of very small crystals while preserving near-native morphology and composition through combination with direct imaging, energy dispersive X-ray spectroscopy (EDS), and electron energy loss spectroscopy (EELS).

A common misconception is that practical 3DED necessarily requires an instrument built around bespoke hardware or a dedicated electron diffractometer. The reported result is narrower and more precise: GiveMeED is intended for conventional TEMs, and it is stated to work with standard microscope/camera/software combinations provided that the camera supports “In-Situ” or video mode within DigitalMicrograph. This does not eliminate all instrument constraints, but it lowers the entry threshold for laboratories already operating compatible TEM and camera systems.

2. Operating principle, interface, and software structure

GiveMeED is a continuous-rotation electron diffraction acquisition script. Once a dataset is started, it automates the essential 3DED sequence: it moves the stage to the chosen start angle, unblanks the beam, starts camera acquisition, continuously rotates the stage to the end angle, blanks the beam at the end, and saves the images and metadata (Weare et al., 14 Jul 2025). The automation point is operationally important because it reduces operator error and makes the experiment repeatable.

The main user inputs include the save data path, sample name, wavelength (Lambda / nm), frame rate, notes, tilt range, start angle, and end angle. The interface also provides Go to Start, Go to End, Tilt Neutral, Start 3DED, and Abort 3DED. The reported operating sequence is correspondingly simple: insert the camera in In-Situ mode, check the save path and experiment name, confirm that the specimen stays visible over the tilt range and is at eucentric height, switch the microscope to diffraction mode, and press Start 3DED. If a problem occurs, Abort 3DED blanks the beam and stops recording.

Stage rotation is performed at eucentric height using the microscope’s lowest speed setting, reported as about $9.1^\circ\ \text{s}^{-1}$ on a JEOL 2100Plus. In the datasets described, the stage typically covered around $120^\circ$ in $12$–$14\ \text{s}$, indicating that GiveMeED is designed for practical continuous-rotation acquisition rather than stepwise tomography. Metadata capture is treated as a major feature: the script records electron wavelength, the rotational relationship between successive diffraction patterns, and other microscope-derived parameters. The metadata contents are also configurable through the log_message string.

The code structure is deliberately compartmentalized, with each process isolated into a function. That design choice is presented as a portability mechanism. The reported successful configurations are a JEOL 2100Plus with a Gatan OneView and a JEOL 2100F with a Gatan K3, and the paper states that GiveMeED should work on any TEM using a Gatan camera with In-Situ mode and DigitalMicrograph.

3. Low-dose strategy and quantitative fluence control

Dose management is a central part of the GiveMeED workflow, not an auxiliary consideration. The paper treats electron fluence quantitatively and argues that 3DED should be carried out before more dose-intensive imaging and spectroscopy if the objective is structure solution from a near-native crystal (Weare et al., 14 Jul 2025).

During crystal search, the reported strategy is to use very low flux, around ~0.01 e- Å⁻² s⁻¹, with a $50\ \mu\text{m}$ condenser aperture, spot size 5, largest convergence angle, fully spread beam, and $10^{-4}\ \text{A}$ emission current. Once a suitable crystal is found, the beam is blanked to reduce total fluence. For 3DED acquisition, the stated conditions are a $100\ \mu\text{m}$ selected area aperture, spot size 1 or 2, $10^{-8}\ \text{A}$ emission current, $600\ \text{mm}$ camera length, $9.1^\circ\ \text{s}^{-1}$0 on the OneView, and image size $9.1^\circ\ \text{s}^{-1}$1. The practical target is to record reflections to at least $9.1^\circ\ \text{s}^{-1}$2, judged in live view with the helper script AutoResRings.

The dose argument is mechanistic as well as empirical. Higher-angle reflections are reported to disappear first as dose damages the lattice, so the preservation of reflections beyond $9.1^\circ\ \text{s}^{-1}$3 becomes an operational criterion for usable crystallographic data. The cumulative-dose analysis further shows that 3DED and BF-TEM imaging contribute relatively little dose, whereas EDS and especially EELS can dominate cumulative fluence. The stated consequence is procedural: 3DED should precede spectroscopy if preserving crystallinity matters.

The Supporting Information gives the quantitative framework used for flux and fluence estimation: $9.1^\circ\ \text{s}^{-1}$4 where $9.1^\circ\ \text{s}^{-1}$5 is electron flux, $9.1^\circ\ \text{s}^{-1}$6 is beam current, $9.1^\circ\ \text{s}^{-1}$7 is beam area, $9.1^\circ\ \text{s}^{-1}$8 is the elementary charge, and $9.1^\circ\ \text{s}^{-1}$9 is beam radius. For beams larger than the camera sensor, the beam radius is estimated by

$120^\circ$0

where $120^\circ$1 is the chord length and $120^\circ$2 is the chord height. The camera gain calibration is reported as

$120^\circ$3

at $120^\circ$4, and the experimental flux conversion is

$120^\circ$5

Applied fluence is defined as

$120^\circ$6

and cumulative fluence as

$120^\circ$7

4. Integration with crystallographic software and the correlative workflow

A defining feature of GiveMeED is that it is not presented as a standalone diffraction environment but as an acquisition layer integrated into ordinary crystallographic processing. Data are recorded as .dm4 images in DigitalMicrograph In-Situ format, and the reported export pathways connect directly to established software ecosystems (Weare et al., 14 Jul 2025).

For DIALS, the stated procedure is to export a .dm4 stack and pass experimental parameters via the command line. For PETS2, the images are exported as .tif, after which a Python script creates a .pts2 project. For CrysAlisPro, the export is .dat via the Esperanto Importer; some parameters, including image size, byte type, rotation axis, and electron wavelength, must be entered manually because they are not read from headers. After reduction, the workflow uses CrysAlisPro and PETS2 for data reduction, SHELXT for structure solution, and SHELXL for refinement, all within Olex2. Additional utilities reported in the workflow are Platon ADDSYM for symmetry checking, Proffit merge for combining multiple CuPC grains in one case, and Vesta for visualization.

The full experimental sequence is given explicitly: prepare sample grid; locate candidate crystals by low-dose BF-TEM; confirm crystal morphology and diffraction contrast; set eucentric height; switch to diffraction mode; collect 3DED with GiveMeED; record complementary BF-TEM, EDS, and EELS from the same specimen volume; export and reduce the diffraction data; and solve and refine the structure with standard crystallographic software. This sequence is significant because it places diffraction, imaging, and spectroscopy in a single correlative pipeline rather than treating 3DED as an isolated measurement.

The paper also reports two companion utilities: Go2Alpha, for quickly moving the stage between two tilt angles, and AutoResRings, for live resolution visualization. Together with GiveMeED, these tools define a practical software suite for acquisition, alignment, and resolution assessment on compatible TEM platforms.

5. Demonstrated structures and empirical performance

The performance of GiveMeED is demonstrated through three case studies: paracetamol, perchlorocoronene (PCC), and copper(II) phthalocyanine (CuPC). In all three, the determined unit cell parameters and atomic connectivity are reported to match accepted literature X-ray structures, and the measurements were carried out in near-native states under controlled low dose conditions at either room or cryogenic temperatures (Weare et al., 14 Jul 2025).

Sample	Acquisition	Structural outcome
Paracetamol	$120^\circ$8; $120^\circ$9 over $12$0; mean flux $12$1; total fluence $12$2	$12$3 unique reflections; $12$4 completeness to $12$5; $12$6; $12$7; $12$8
PCC	$12$9; $14\ \text{s}$0 over $14\ \text{s}$1; mean flux $14\ \text{s}$2; total fluence $14\ \text{s}$3	$14\ \text{s}$4 unique reflections; $14\ \text{s}$5 completeness to $14\ \text{s}$6; Cmce; $14\ \text{s}$7; $14\ \text{s}$8
CuPC	Room temperature; three datasets, $14\ \text{s}$9, $50\ \mu\text{m}$0, and $50\ \mu\text{m}$1; mean flux $50\ \mu\text{m}$2; total fluence $50\ \mu\text{m}$3	$50\ \mu\text{m}$4 unique reflections; $50\ \mu\text{m}$5 completeness to $50\ \mu\text{m}$6; $50\ \mu\text{m}$7; $50\ \mu\text{m}$8; $50\ \mu\text{m}$9

The case studies are also correlative demonstrations. For paracetamol, nitrogen and oxygen were verified by EDS and EELS before refinement. For PCC, EDS and EELS confirmed chlorine and $10^{-4}\ \text{A}$0 carbon. For CuPC, EDS and EELS confirmed copper, nitrogen, and $10^{-4}\ \text{A}$1 carbon. This integration matters because the claimed outcome is not only structure solution but a comprehensive description linking morphology, composition, and crystallography.

The reported data quality threshold is pragmatic rather than idealized: once the 3DED data contain useable reflections above $10^{-4}\ \text{A}$2, routine processing with SCXRD-style pipelines becomes possible. At the same time, the paper explicitly cautions against interpreting relatively high refinement statistics as evidence of incorrect structures. The stated reasons are that dynamical scattering is not fully modeled in the kinematic refinement and that reciprocal-space coverage remains incomplete because of limited tilt range.

6. Significance, limitations, and adoption

GiveMeED is best understood as an enabling layer for practical crystallography with a conventional TEM. Its importance lies less in introducing a new diffraction geometry than in making continuous-rotation 3DED accessible, repeatable, and compatible with standard crystallographic reduction and refinement workflows (Weare et al., 14 Jul 2025).

Several limitations remain explicit. GiveMeED depends on camera support for In-Situ mode within DigitalMicrograph. The microscope stage tilt range must be verified before attempting a dataset. Some downstream software still requires manual entry of parameters. The refinement statistics remain affected by the mismatch between electron-diffraction data and kinematic small-molecule refinement models. These are not presented as failures of the script; they define the current operational boundary conditions of routine 3DED on conventional TEM platforms.

At the same time, the software is made freely available for use and modification, and its modular function-based design is intended to support adaptation to local hardware and future methods. A plausible implication is that GiveMeED lowers the barrier to entry not merely by automating acquisition, but by normalizing a reproducible workflow in which low-dose scouting, controlled diffraction, metadata capture, export, and SCXRD-style processing form a single experimental chain. In that sense, GiveMeED occupies a transitional position between bespoke electron diffraction practice and routine laboratory crystallography.