Vendor column: Mining mineral analysis using near-infrared technology: Don't laugh until you've tried it!

Skip to Navigation


  • Published: Apr 30, 2012
  • Author: Daniel Shiley
  • Channels: Infrared Spectroscopy
thumbnail image: Vendor column: Mining mineral analysis using near-infrared technology: Don't laugh until you've tried it!

Daniel Shiley, Senior Application Chemist, SummitCAL Solutions Team, ASD Inc.

Potential applications for mining exploration and control ore processing

Figure 1. Calcite and dolomite.

Mineral analysis by near-infrared (NIR) analysis is not well known, even though spectral information on minerals has existed for years. This is primarily because NIR users typically have focused on the elemental minerals rather than geologic mineral species. Researchers using near-infrared for minerals have applied a remote sensing approach rather than a chemometric-based approach, and unfortunately these two groups seldom talk to one another. By bringing together NIR, software and chemometric modeling, advanced technology applications are being created for diverse industries, including mineral analysis in mining exploration and control ore processing.

During my career, I have had the opportunity to create calibrations to measure grain quality attributes for some very large global companies. I created and managed networks of more than 50 NIR systems and the net value of the products tested in these very large programs was between $15-75 million each year. However, this value pales in comparison to the value of the materials produced in the mining sector. Data on the largest of the mines that I am working with (2007 data) shows that it produced more than $10 billion of copper per year! We can measure minerals that would otherwise increase expense of operations or that cause problems with extraction of metal. If only one percent savings was achieved by using NIR technology, this would mean a net value of $100 million per year. A savings of only one-half percent would result in a savings of $50 million. By properly identifying the mineralogy, mine throughput can be increased and costs reduced to help achieve this type of result. This of course would be nothing to laugh at.

In 2007, ASD exhibited at the NIR2007 meeting in Umea, Sweden. Although we had been working with mineral modeling for many years, it was at this conference that we began to really promote quantitative mineral analysis. While our demonstration piqued interest, the most common first reaction was, "Don't you know you can't measure minerals with NIR?"

I once believed this also. As I began my spectroscopy career almost 18 years ago, one of the first things that I learned was that it was not possible to measure minerals using near-infrared. In food and feed products, typically the term "mineral" is referring to elemental minerals, such as how much iron, potassium or zinc is present in a product. Of course when we measure typical organic materials, we are measuring organic bonds. Typically these are OH, CH, SH and NH bonds. When we examine typical organic materials that contain elemental minerals we do not see spectral response from these elemental minerals. So even today, in spite of all the improvements in NIR spectrometers, we cannot accurately predict elemental concentration using NIR across a range of background materials.

However, when we discuss geologic minerals, this is something quite different. In the case of mining-related minerals we see very distinct, different and diagnostic near-infrared spectra produced by various mineral species. This is not new information. Since the 1960s, there has been a growing body of knowledge relating spectral measurements to mineral species. In fact, NASA planned the lunar landing sites through the use of telescopic reflectance spectroscopy of the lunar surface.

The spectral features that we see in rock are overtones of the fundamental absorption features in the mid-infrared. These spectral features are the result of electronic and vibrational processes in mineral lattices. Electronic processes include crystal field effects, charge transfer, color centers and conduction band transitions.

The position of wavelength features varies dependent upon the substitution of elements in the mineral lattice. For example, a substitution of Mg for Ca in calcite results in a shift of wavelength in the 2300 nm region and is a leading indicator that can be used to track the level of dolomitization of the limestone.

Appreciating Mineral Spectra

Figure 2. Kaolinite and muscovite spectra.

Several publicly available spectral libraries are available free from various agencies. I would encourage anyone with a bit of curiosity to download some spectra to get an appreciation of the uniqueness of the spectral characteristics of minerals. The USGS has a database that contains over 1300 spectra available at This library contains minerals, man-made materials, vegetation and other materials. Jet Propulsion Laboratory has ASTER Spectral library version 2.0 available at This library contains more than 2,400 spectra of minerals, vegetation and man-made materials. The ASTER spectral library includes data from three other spectral libraries: the Johns Hopkins University (JHU) Spectral Library, the Jet Propulsion Laboratory (JPL) Spectral Library, and the United States Geological Survey (USGS - Reston) Spectral Library.

Additionally, several commercially available programs that contain their own libraries could also be purchased. These include TSG (The Spectral Geologist), from AusSpec International, Ltd., and SPECMIN from Spectral International, Inc.

Some minerals do not have good diagnostic features, so not all mineral species can be identified using near-infrared. But the main pathfinder minerals that can be used to identify an area of mineralization (a zone that has the target metal present) can be identified and also many of the minerals that cause issues with extraction of metal can be successfully measured using NIR. So NIR analysis has utility in both mining exploration and for control of ore processing.

The Established Art of Remote Sensing

Figure 3. USGS wide area assessment map.

Remote sensing projects have utilized the knowledge of mineral spectra for nearly 40 years. This information has been primarily used to create wide area assessment maps. In this usage, an aerial or space-borne sensor is used to collect spectra of the area as it overflies the area of interest, and then a field crew uses a portable spectrometer with the same wavelength range as the overflight sensor to collect ground truth spectra in this area. Ground truth spectra are simply spectra that are collected using the ground-based portable spectrometer that are used as the key to create a wide area assessment map that describes what is included in the area. Typically, this approach does not create a quantitative map, but can create a spectral abundance or semi-quantitative map that can be used to identify areas of alteration that are useful for mining exploration projects. Figure 3 contains a wide area assessment map from USGS showing various minerals and alterations in a false-color image based on NIR.

Increasing the Speed of Mineralogy Determination

In normal practice, mines determine mineralogy using XRD, QEMSCAN or various chemical reference assays. These reference methods can produce quantitative mineralogy values, but they are labor intensive, expensive and require extensive sample preparation. Worse still, it may take anywhere from days to months to obtain the results. So typically, mineralogy is not used to control a process.

NIR technology is now being used to produce some understanding of mineralogy on a near real-time basis. There are multiple approaches to determination of mineralogy using NIR. The most common use is to utilize a library and match the target spectrum to the library, which can provide an understanding of the spectrally predominant minerals present in the sample. However, because these programs assume a linear combination of spectra, they cannot account for variation in absorption coefficients in the minerals. However, this approach can produce a map or understanding of the alteration within the rock samples. As previously mentioned TSG and SPECMIN are commercially available products that are used to provide mineral identification information typically used for mining exploration. Both of these programs can be used to identify the presence of certain mineral species, but neither produces a true quantitative value for the minerals. While certain minerals have very diagnostic features that indicate the presence of the mineral, others have very little near-infrared signal or their signal is buried beneath other mineral features. These programs utilize a library of known mineral spectra and either use an algorithm such as spectral angle mapper or a spectral unmixing technique to identify the minerals that are contributing most to the spectrum of the sample that is measured. The benefits of such programs are that they can be used to immediately produce results straight out of the box and that useful information regarding alterations can be obtained even in remote locations. The major drawbacks to these approaches are that the operator training must be very high, the values are qualitative or semi-quantitative and minor components of the rock are typically not well identified.

Quantitative Mineralogy through Chemometrics

The second option for mineralogy determination by near-infrared is through the use of a multivariate calibration model to produce a true quantitative value for the rock sample. In this scenario, we use standard chemometric techniques to produce a quantitative model for each mineral. Although the creation of the model is much more expensive than the use of the library-based approach and the development of the model requires a high amount of skill, the interpretation and use of the near-infrared solution becomes much simpler and does not require a highly trained operator for routine analysis. We will discuss this approach in greater detail in a future article.

You, too, Can be an NIR Believer!

Through our work with NIR on mining-related projects we have been able to conclusively show that many minerals can be successfully identified using NIR technology. These minerals are the pathfinders that exploration geologists can utilize to help them identify areas of mineralization and can also be used to help provide more efficiency and better control of the ore processing plant. Although our application of NIR technology is not in the 'typical' area of organic analysis, this is a well-documented use of NIR technology. Try it yourself and you will be a believer, too!


  1. Clark, R. N., G. A. Swayze, K. E. Livo, R. F. Kokaly, S. J. Sutley, J. B. Dalton, R. R. McDougal, and C. A. Gent, 2003, Imaging spectroscopy: Earth and planetary remote sensing with the USGS Tetracorder and expert systems, J. Geophys. Res., 108(E12), 5131.

  2. Goetz, A. F. H., B. Curtiss and D. A., Shiley, 2009, Rapid gangue mineral concentration measurement over conveyors by NIR reflectance spectroscopy, Minerals Engineering, 22 (2009), 490- 499.


Article by Daniel Shiley

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.

Social Links

Share This Links

Bookmark and Share


Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Copyright Information

Interested in separation science? Visit our sister site

Copyright © 2017 John Wiley & Sons, Inc. All Rights Reserved