Near infrared

NIR for routine analysis of food and agricultural products

By Richard Mills,
A definition of the technology as applied in food production

Near Infrared (NIR) analysis is a spectroscopic technique that makes use of the naturally occurring electromagnetic spectrum.
The NIR region is the area of the spectrum defined by wavelengths between 700nm and 2500nm. Near Infrared is an accurate and rapid analysis method that is well suited for quantitative determination of the major constituents in most types of food and agricultural products.
The NIR advantage
The overall advantages of using NIR analysis is that it provides rapid analysis data for better decision making in food and agri production processes. Compared to traditional analysis methods it requires little or no sample preparation and no chemicals or consumables.  It is non-destructive, operator friendly, fast (30-60 seconds), reliable and precise.

How it works
The working principle can be defined as follows:

  • Near Infrared light is directed onto a sample
  • The light is modified according to the composition of the sample and this modified light is detected (see transmission and reflectance below)
  • The spectral modifications are converted to information regarding the composition  of the sample
  • These conversion algorithms are called ”calibrations” 

Transmission and reflectance
An infrared spectrum can be obtained by passing infrared light through a sample and determining what fraction is absorbed by the sample (transmission). Alternatively, light can be reflected from the sample and the absorption properties can be extracted from the reflected light (reflectance). NIR reflectance is a subject in itself that will not be covered in detail here. 

NIR reflectance and transmittance methods can be chosen according to the analysis job, for example, transmittance is good for measuring cheese to obtain a representative measurement throughout the sample. For homogenous samples such as milk powder, reflectance is ideal. 

For both methods, the NIR wavelength range is an important consideration. For instance, a Short Wave NIR range (850 – 1050nm) gives good sample penetration with NIR transmission.

NIR Transmission and reflectance

An infrared spectrum can be obtained by passing infrared light through a sample (transmission) or the light can be reflected from the sample (reflectance). 

What’s your wavelength?

Any discussion about NIR wavelength must start with the spectrometer that converts the infrared light into a usable signal. A common consideration across all spectrometer technologies is the wavelength range measured in nanometers (nanometers nm or reciprocal centimeters cm-1). 

Without getting too deep into a technical discussion it can be said that certain wavelengths are better for measuring certain samples and parameters than others. The wavelength required to measure protein in grain for example is different from that required to measure amino acids in feed ingredients and so on.  Another consideration is the signal to noise ratio offered by different instruments at certain wavelengths as this indicates the quality of the signal generated.

NIR wavelength

The electromagnetic spectrum is divided into several regions where each region stands for a specific kind of molecular transition. For instance, Xrays have wavelengths of only a few nm and are very harmful, because they break chemical bonds and ionize molecules. In comparison, near-infrared (780 and 2500 nm) is not harmful. Molecules simply absorb infrared light according to the nature of the sample. 

Monochromator, FT-NIR and DDA 

For quantitative measurements of feed, a so-called scanning grating monochromator instrument is a proven choice. It is ideal for quantitative measurements across a broad spectrum of applications with a broad wavelength range for a wide range of parameters including those such as colour in fish food or similar requiring the visible region in addition to the NIR region. Alternatively, when using NIR transmis¬sion for measuring inhomogeneous samples such as grain, it is an advantage to use the Short Wave NIR range (850 – 1050nm) where the light penetration is good and the premium signal to noise ratio offered by a scanning grating monochromator is essential. 

On the minus side, an internal wavelength standard is required to achieve good wavelength accuracy. However, like all other aspects of NIR technology, scanning grating instruments are constantly evolving and many are using this in¬ternal wavelength standard to demonstrate good wavelength accuracy. 

For qualitative measurements in the laboratory where narrow instrument bandwidth is needed, FT-NIR technology has advantages. It is ideal for pin-point qualitative measurements of substances with narrow absorption bands or for some quantitative measurement ap¬plications for samples having closely spaced narrow absorption bands. Plus, it is possible to adjust the resolution to obtain the best tradeoff between wavelength resolution and the signal to noise ratio. 


Disadvantages include a lower signal/noise ratio than a monochromator instrument, particularly at short wavelengths and the omitted visible wavelength range (below 850 nm). FT NIR is also a vibration-sensitive technology and the design must take this into account for applications in a production environment.

On the subject of production, the use of NIR analysis instruments close to the production line or even for continuous measurements of material in the production line is increasingly widespread. For these measurements, fixed grating Detector Diode Array (DDA) is the best option. It is ideal for robust and vibration-tolerant instruments and the simultaneous measurement of the full spectrum also makes it tolerant to sample movement. On the downside, a tradeoff between wavelength range, resolution and signal to noise ratio must be made. 

NIR and FTIR are, in principle, very similar. The main difference is that FTIR works across longer wavelengths. Additionally, many special optical components and materials are used with FTIR. The benefit of the longer wavelengths is that more specific chemical information is typically obtained from the samples.
A comparison of NIR and FTIR is provided in this video in relation to applications in the dairy industry. Called ‘Directions in dairy analysis’, the video includes interviews with experts from FOSS who explain the technology and the considerations to be made when selecting an infrared analytical instrument. 

What is the difference between FTIR and NIR for in-line milk standardisation?
FOSS chemometrician Per Waaben explains the difference as well as the benefits of each technology.

The six elements of a good NIR instrument

  1. Optical performance across the required wavelength 
    A broad wavelength allows you to test many things from moisture, fat and protein to more demand¬ing parameters such as amino acids, ash or fibers.  However the same high performance across the whole wavelength is necessary to ensure reliable results. 

  2. Flexible and easy to use. Increasingly, NIR is being used in production environments where production staff need to perform tests quickly and easily with a low risk of human error. The right sample presentation options, for example, with no or little sample preparation and functionality such as touch screen operation are essential features of a modern NIR solution.

  3. 100% calibration compatibility. A new solution should be 100% backwards compatible with earlier instruments so that it is easy to leverage existing calibration data using straightforward migration paths without any loss of performance. 

  4. Factory standardised. Every instrument that leaves the factory should be hardware standardised for light intensity, bandwidth and wavelength precision to ensure complete consistency between instruments. Furthermore, once an instrument is up and running, in-built measurement standards are needed to control its performance and ensure that no deviations occur over time. This keeps continuous control of consistency between instruments and makes it easy to add any new instruments to your operation. Multiple in¬struments can easily use the same calibrations without any modifications.

  5. Robustness for the analysis environment. The NIR instrument should be designed for the user environment, for example, if it is for use in a feed plant, it should be able to withstand the harsh production conditions. IP65 certification indicates that it will withstand humidity, dust, vibra¬tions and temperature fluctuations

  6. True network capability for effective instrument management. True networking is much more than an internet connection. It should allow you to look inside the instrument to check performance and make adjustments or updates to calibrations. This is a must to allow remote management and surveillance of multiple instruments from a single location. All updates and calibrations can be performed centrally for improved instrument performance and convenience. Not only does this save you time, but it has also been shown to significantly reduce costs 


A typical NIR platform for a range of applications in food and agri production

  • Scanning Monochromator
  • Wavelength Range- 400-2500 nm
  • Operating temperature and humidity: 5-40°C (41-104°F ) and  <93%RH 

True networking is much more than an internet connection. It should allow you to look inside the instrument to check performance and make adjustments or updates to calibrations. This is a must to allow remote management and surveillance of multiple instruments from a single location. All updates and calibrations can be performed centrally for improved instrument performance and convenience. Not only does this save you time, but it has also been shown to significantly reduce costs. 


In-line NIR, the new frontier
A recent trend is the use of NIR directly in-line in the process stream to give continuous measurements every few seconds. This gives operators a much sharper picture of fluctuations in the process flow than with a bench-top instrument that might be used only once an hour or so. Operators can then spot trends in protein, moisture and other parameters and take timely action accordingly to meet production targets.  Inline NIR is already an established concept in the dairy industry for production of products such as butter and soft cheese where close control of moisture is essential both for quality and production economy.  See a video interview with a user of inline NIR here.

General Manager of Lake Country Dairy, John Peterson describes how ProFoss in-line cheese analyzer offered even more benefits than they had hoped for.
“It gave us so much more control over the process it actually had more benefit than we had originally thought,” says Peterson.
“We were able to optimize our moisture level to really give customers the value that everybody is looking for.”

FOSS products using NIR – some examples:

FoodScan™ Dairy Analyser for analysing cheese, whey powder, butter and yoghurt. It measures a variety of parameters with a minimum of sample preparation and delivers results in just 50 seconds.

NIRS™ DS2500 for near infrared analysis of feed, providing top performance and accuracy across a broad wavelength range of 400 to 2500 nm.

Infratec™ NOVA tests multiple parameters (moisture, protein, oil, starch, etc.) in a broad range of grain and oilseed commodities. True networking and identical instruments reduce instrument management work required for consistent test results throughout grain receival networks.

ProFoss™ in-line process analysis solution offers a range of applications for accurate monitoring of dairy production, from butter to cheese and dairy powders.



Read about other technologies