Moving from Lab to Line

TSH's picture


I was hoping to get some advice on moving a process from the lab to the line. I would like to move an NIR reflectance measurement to the line where my sample is falling into a bin.  I am unsure of the challenges associated with such a sample configuration and the type of suitable instrumentation.

I am curious if anyone has any experience acquiring measurements of samples in free-fall and what challenges were faced (calibration transfer between different sampling methods)?  As for instrumentation, my thought is that an imager may be a better choice in order to integrate a larger surface area.

Thank you for your time.


hwswien's picture

Dear Thad,

you may want to contact Dr. Sven Borchert ([email protected]) who has gathered pharmaceutical industrial experience on this sampling issue.

Kind regards,

Heinz W. Siesler


ianm's picture

What is the sample? Is it grain, or something else?

(posted on behalf of Phil Williams)

gabiruth's picture


First question - is it free fall in the form of a continuous flow or just drop after drop?

There is a whole world of difference between the two conditions.

If it is a free continuous stream - can you define the approximate diameter of the falling "column"?

Is it very opaque and does the surface of the "column" reflective of light like a "mirror" or is it non reflective?

Maybe based on that we can progress.


Gabi Levin



TSH's picture

Thank you for the comments.  To answer your questions the sample is a natural product (grains - not transparent) and can be considered a continuous stream (basically we are dumping the sample out of one bin and into another at a constant rate). The sample column is several inches thick with some variability and several feet wide. 

rogerjm's picture

Dear Thad,

measuring spectra from moving particles is not a very challenging issue, provided that :

- the stream is optically thick enough ; that seems to be the case here

- each wavelength is measured for the same object. That means that you can use only "true snapshot" technology, as diode array devices. You cannot use moving gratings nor Fourier transform spectrometers.

Good luck,

Jean-Michel ROGER


JM Roger
Irstea, Montpellier France

jcg2000's picture

And you can also the acousto-optic-tunable-filter based NIR. Brimrose has it.


In any case, during method development, you might want to evalauet method robustness by adjusting the grain flow rate.

Jerry Jin

gabiruth's picture

With the Brimrose you can scan from 1100 to 2100 300 times in about 11 seconds - this will average a very large number of grains - however - with all instruments, be it diode array, Brimrose AOTF, you will need to maintain a distance of about 65 mm from the instrument's front window to the stream - if this can be done and the stream isn't "moving" around back and forth - and you have a fixed distance - it can be done.

It will be much better if it could be arranged for part of the stream to fall on a slanted surface into which we will install a sapphire window and look through the window - this way the distance of the grains from the instrument is always fixed - bit it is possible to do it the other way also, if the distance can be kept constant.

Can you give us information on the shape of the "pipe"? Is it it round, rectangular? Dimensions?




Gabi Levin


[email protected]

mike's picture

I have published two patents on my area of interest including using online NIR for bulk material analyses.  One patent concerns scanning and reporting analyses in real time for materials on a moving conveyor belt.  The other patent, which follows the original by two years, is for analysis of bulk materials in a closed air slide or pneumatic "pipe" to address fine materials transported by this means for normal online to storage or loading.  If you are interested, I can send you references on these patents.  Although I no longer commercialize these systems, I would be happy to put you in touch with one company that manufactures, sells, and supports systems based on these patents.   You can contact me via the forum, or privately at [email protected].


Good luck,




ianm's picture

Hello TSH

Your needs could be addressed by a NIRS diode-array instrument, such as the instrumentation offered by Perten Instruments. They offer diode array instrumentation, and have extensive experience on whole-grain on-line applications. They are located in Springfield, IL. You will need calibrations for whatever constituents that you wish to measure, based on on-line measurement. Sample presentation to the instrument is an important factor, the instrument company will be able to assist you with this.  

I will be happy to answer any further questions, if you like.

(posted on behalf of Phil Williams)

TJ7's picture

Dear all,

As a continuation of the topic of in line control: does someone have an experience implementing it for herbs&spices (to measure things like moisture, fat, volatile oil, etc.)?

shileyda's picture


There are several comments about the type of spectrometer system that you should where I do not agree with their recommendations.  Since you are not trying to identify the individual grains in the flowing stream, virtually any spectrometer type can be used because you are trying to obtain an average representing the entire grain transfer. So this allows array systems, but also scanning systems, imaging systems, LVF, AOTF and likely some I am missing here. On a flowing stream or moving conveyor you can consider that your sample is actually infinitely large, thus a single  "snap shot" is actually not important or even appropriate.

russell's picture

I agree with the approach suggested by Gabi Levin regarding directing the stream to a window to assure consistant sample presentation.

I also agree with the comment that the type of NIR instrument required cannot be assumed from the information provided.  That choice would be determined by the needs of the measurement.

Codita71's picture

Is it possible to use NIRS diode array intruments at endof a dryer belt to determine moisture in the final poduct?

Is it affected by difference in color?