In photonics since 1999. Benelux technical support, answer within one working day.
Choosing a spectral range: VIS or NIR
Silicon detectors work to about 1000 nm and are inexpensive; InGaAs covers roughly 900 to 1700 nm and costs more. The choice is decided by where your sample has features. Colour and fluorescence live in the visible; water, fats and polymer overtones live in the near infrared.
THE SHORT ANSWER
Silicon detectors work to about 1000 nm and are inexpensive; InGaAs covers roughly 900 to 1700 nm and costs more. The choice is decided by where your sample has features. Colour and fluorescence live in the visible; water, fats and polymer overtones live in the near infrared.
| Silicon range | 400-1000 nm |
|---|---|
| InGaAs range | 900-1700 nm |
| Typical VIS resolution | 1,5 nm |
| Typical NIR resolution | 6 nm |
| VIS applications | Colour, fluorescence, LED characterisation |
| NIR applications | Moisture, fat, polymer identification |
Spectrometer selection is not really a specification question. It is a question about your sample.
Where the information is
Electronic transitions and colour live in the visible. If you are measuring what something looks like, characterising an LED, or working with fluorescence or absorbance in solution, a silicon-detector instrument covering roughly 400 to 1000 nm is the right tool and the cheaper one.
Molecular overtone and combination bands live in the near infrared. Water, fats, proteins and the C-H bonds that distinguish one polymer from another all have features between 900 and 1700 nm. That range needs an InGaAs detector.
Resolution is not the figure of merit
NIR bands are broad and heavily overlapping. Six nanometres of resolution is entirely adequate because there is nothing narrower to resolve, and buying finer resolution costs you throughput and therefore signal to noise. In the visible, narrow emission lines do exist and resolution matters more.
Calibration is the real work
For quantitative NIR the instrument is the easy part. Because the bands overlap, concentration is extracted by a model built from samples you have measured by another method. Budget for building and maintaining that model, and for the fact that it will not transfer cleanly to a different sample presentation.
The unglamorous limit
In most industrial measurements, repeatability of sample presentation dominates the error budget. A fixed stand-off jig frequently improves results more than a better spectrometer.
When to use it
- Colour measurement, LED and light source characterisation: VIS
- Moisture, fat or protein content in food and agriculture: NIR
- Polymer and plastics identification, including recyclate sorting: NIR
- Fluorescence and absorbance in solution: VIS
What to watch out for
- Fluorescence from additives can bury a weak Raman signal, which sometimes makes NIR absorption the better route
- NIR bands are broad and overlapping: quantitative work needs calibration against known samples
- Higher resolution is not automatically better; it costs light and therefore signal to noise
- Sample presentation repeatability usually limits the measurement long before the instrument does
Still deciding?
Tell us the wavelength, the aperture and what you are measuring. We answer with a part number, a price and a lead time, usually the same working day.