Science and technology services for industry

Four groups of science and technology services

All four service groups share one technology core: non-destructive Raman measurement, spectral processing by deep learning models running on the device, and an explanation layer that points to the wavenumber regions supporting each conclusion.

• 1

  Active ingredient assay and content uniformity testing

  The problem

  A batch of medicine, food supplement or cosmetic product needs its true active ingredient content measured, along with the variation between individual tablets or packaged units within the same batch. Chromatography is accurate but destroys the sample, consumes solvent, and is too slow to test many points along a production line.

  How we work

  We build a Raman calibration curve for your specific sample matrix from a set of samples with known content, cross-checked against HPLC results produced by a third party. Once the calibration meets acceptance criteria, each sample unit is measured directly in tens of seconds, with no grinding and no dissolution.

  What you receive

  Content report for every sample unit; batch uniformity index with relative standard deviation; distribution plot; raw spectra and the calibration model so you can keep them in your quality records.

  Expected turnaround

  2–4 weeks for the first engagement (calibration development included); 3–5 working days for subsequent batches on the same matrix

• 2

  Rapid detection of counterfeit medicines and banned substances in cosmetics

  The problem

  Suspected counterfeits, and cosmetics adulterated with corticosteroids, hydroquinone or mercury, usually cannot be distinguished by eye or by packaging. Distributors need to screen quickly at the warehouse, and inspection bodies need a basis for deciding which samples must go for confirmatory analysis.

  How we work

  We compare the Raman fingerprint of the suspect sample against reference spectra of the genuine product and against a spectral library of commonly encountered banned substances. The classifier returns a similarity level together with a map of the wavenumber regions causing the difference, so the conclusion can be presented and defended.

  What you receive

  A screening conclusion per sample: matches reference / significantly different / signs of a banned substance detected; the list of suspected substances with the supporting spectral regions; a recommendation on which samples need confirmatory analysis.

  Expected turnaround

  5–10 working days for batches of up to 50 samples; on-site measurement is possible for larger volumes

• 3

  Pesticide residue and mycotoxin analysis

  The problem

  Export consignments rejected for exceeding residue limits or for aflatoxin contamination cost far more than testing would have. Cooperatives and exporters need screening cheap enough to run on many samples before the shipment is packed.

  How we work

  We use surface-enhanced Raman spectroscopy (SERS) with nanostructured substrates fabricated in-house to push the detection limit down to parts per billion, combined with deep learning that separates the analyte signal from the produce matrix. Positive results are then recommended for confirmation by LC-MS/MS at an accredited laboratory.

  What you receive

  Screening report per sample and per substance on the agreed list; limit of detection and limit of quantification of the measurement; comparison against the maximum residue limits of your target market; a recommendation on which samples need confirmatory analysis.

  Expected turnaround

  2–3 weeks for batches of up to 100 samples, counted from receipt of the samples and the target substance list

• 4

  Technical consulting and training

  The problem

  Many organisations already own, or plan to buy, a spectroscopy system but have not established a repeatable measurement procedure, do not know how to standardise spectral data, and have no operator experienced enough to defend the numbers before a technical panel or a customer.

  How we work

  We survey your existing equipment and requirements, design the sampling and measurement procedure, write the standard operating procedure, build an initial calibration dataset, and then train your staff directly. Training covers both how to read a result and how to recognise a spectrum that must be discarded.

  What you receive

  Consulting report with equipment configuration recommendations; a written standard operating procedure handed over to you; the initial calibration dataset; a 2–3 day training course with participant materials; technical support by email for three months after training.

  Expected turnaround

  4–8 weeks depending on scope; the training course is scheduled by agreement

These services operate on a non-profit basis

BIRALAB is a Strong Research Group of Vietnam National University, Hanoi, operating on a non-profit basis. All income from science and technology services is reinvested in research, consumables and equipment, doctoral scholarships and community activities; no profit is distributed to members in any form.

Service fees are therefore set at actual delivery cost, with no profit margin added. In return, we prioritise problems with clear scientific or public value, and we may decline requests that are purely commercial.