INSTRUMENT TECHNOLOGY
GramRay i MK3 connects high-performance microscopic counting and sizing with automated, selected-particle Raman analysis. DeepMorph™ supports image-based recognition and targeting; RamanMetrix® automates spectral classification. Linked particle records and pharmaceutical reference comparisons make the analytical evidence useful for real investigations.
Explore the complete analytical workflow →
High-Performance Particle Detection & Sizing
Microscopic counting and sizing are independent analytical services. Automated imaging records particle locations, counts, size distributions, calibrated 2D morphology and size-bin statistics. Concentration requires a suitable sampling method; Raman is added where chemical information is needed.
≈500 nm optical detection
The optical platform can detect and characterize particles down to approximately 500 nm under suitable measurement conditions. This optical capability is distinct from Raman identification; chemical assignments at this size are not universally available.
>99% in a defined 1 µm PS experiment
An internal experiment with 1 µm polystyrene spheres demonstrated a particle-detection and successful-targeting hit rate above 99%. This is specific to that experiment—not quantitative counting accuracy, filtration recovery or a performance guarantee for other materials and matrices.
Drift compensation supports stable automated imaging and particle targeting. Calibrated imaging, instrument checks and suitable reference measurements support reproducible operation. Performance remains dependent on the configuration, sample and controlled method.
Brightfield & transmitted light
Record particle appearance, outlines and morphology in suitable samples. Detection settings and the analytical window are selected for the matrix and study objective.
Reflected light & darkfield
Reflected-light observations and scattered-light contrast can reveal different particle features. Detection, calibration and segmentation need to be assessed for the chosen mode and substrate.
Polarized-light imaging
Polarization can add information about optical behavior and material structure. Its use depends on the installed configuration, sample and analytical question.
Fluorescence imaging
Supported fluorescence channels can help distinguish or target selected particle populations. Fluorescence is complementary optical evidence and is not, by itself, a Raman chemical assignment.
Imaging modes provide complementary optical information. The installed configuration, suitable contrast and calibration are confirmed for the sample and method.
Selected-particle Raman spectroscopy
532 nm or 785 nm excitation
Current GramRay i MK3 instruments operate with either 532 nm or 785 nm Raman excitation, depending on the installed configuration. Laser power and integration time are selected for the sample and analytical method.
DeepMorph™ — Image-based Deep Learning
DeepMorph™ uses Deep Learning on optical particle images. AECP recognition/targeting in nasal sprays and image-based API candidate targeting are established applications. Conventional selection is also available; chemical confirmation remains a separate Raman measurement.
RamanMetrix® — Automated Raman Spectral ML
RamanMetrix® is integrated into the instrument analysis workflow. Multiple established models automate Raman spectral classification, including routinely used cellulose-related classification. Additional application-specific models can be developed and integrated, with suitable evaluation and qualification or validation.
Conventional library comparison, including Pearson correlation, remains relevant for Raman identification. Automated RamanMetrix® spectral ML adds more detailed classification within defined material-reference groups. Fluorescence, weak signal or matrix interference can still leave an assignment inconclusive.
Explore DeepMorph™ image-based Deep Learning →
Optical Imaging & Morphology → DeepMorph™ image-based targeting (where useful) → Raman Spectroscopy → RamanMetrix® spectral classification (where useful) → Scientific Interpretation
Explore advanced Raman spectral fingerprinting →
Further DeepMorph™ models can be rapidly developed and adapted for suitable customer applications. DPI agglomerates, engineered-API agglomerates, spray-dried API particles and other complex structures require application-specific model development and qualification.
RamanMetrix® software is provided by Biophotonics Diagnostics GmbH.
From optical detection to chemical assignment
Counting and distribution statistics cover the optical population independently of Raman. Raman-selected particles are a subset; accepted and inconclusive or unavailable assignments describe outcomes within that subset. These are related analytical states, not four independent populations. Material-specific comparisons and chemical percentages retain their stated denominator.
Optically detected particles
Objects recorded by the imaging method within its scan area, thresholds and detection settings. These contribute to counts, size and morphology; they are not all Raman-confirmed.
Selected Raman targets
Particles chosen for further acquisition using agreed selection rules, morphology, image-based models or study priorities. A target is a measurement candidate.
Accepted chemical assignments
Measured particles whose spectral evidence supports an accepted assignment under the defined identification criteria. These can form material-specific populations.
Inconclusive or unavailable assignments
Selected targets without an accepted chemical assignment, including spectra with insufficient or ambiguous evidence and targets for which usable chemical data are unavailable. These outcomes remain documented separately from accepted assignments.
A connected particle record
The connection between a particle field, an individual particle and its Raman spectrum makes results reviewable at particle level. Identifiers, coordinates, calibrated morphology and measurement settings preserve the analytical context.
Measured application example: a particle field, an enlarged selected particle and its associated Raman spectrum. The green outline marks the selected particle. This example illustrates the link between optical and chemical evidence; it does not identify every particle in the field.
Open the full-resolution measured example →
Read the foreign-particle investigation context →
Original data and measurement settings
The documented result structure includes scan images, detected-particle tables, per-particle images and Raman spectra when acquired, and method/analysis settings. Raman-measured particles and the full detected population have separate records.
Export and reporting integration
Measurement containers can be exported with their index for review in the Index Viewer. Particle tables, images and associated spectra support the agreed project raw-data package and technically reviewed, QA-reviewed report.
Software and model versions, acquisition parameters and relevant controls accompany the analytical documentation. The agreed data package supports scientific interpretation, separate technical review and independent QA review.
How the full investigation connects evidence and independent review →
Controlled methods, meaningful results
Qualified personnel, controlled laboratory conditions, maintenance and calibration, system suitability checks, blanks and relevant references support the analytical method. Intended use, material, matrix and sampling design determine what counting, sizing and chemical assignments can establish.
SizeID.bio combines the platform with pharmaceutical manufacturing, packaging and device expertise and an extensive material-reference collection. Studies can compare unknown particles with packaging, capsule, blister, adhesive, seal or device references and customer materials. Automated spectral models and process information support probable-source investigation within the established method scope.
Quality and documented practices →
Discuss analytical services and method development →
Explore pharmaceutical reference-based particle investigations →
Discuss your analytical question
Tell us what you need to identify, characterize or compare. We will discuss the sample, available references and a suitable analytical strategy.
Back to the Technology overview →