• LIBS technology
  • Applications
  • Real-time results
  • Automation
  • How does LIBS work?
  • LIBS imaging
  • The future of LIBS

LIBS technology

LIBS (laser-induced breakdown spectroscopy) technology is at the heart of all ELEMISSION analyzers. It’s thanks to the LIBS process that ELEMISSION analyzers can detect, identify and classify or quantify the chemical composition of any material, regardless of its state (gas, liquid, solid, conductive or non-conductive). The recent success of the LIBS probe on the Curiosity Mars rover has renewed interest in this technology for many applications, particularly in geology and geochemistry. 

But what exactly is LIBS technology? In short, it’s a form of atomic emission spectroscopy involving laser-generated plasma. LIBS combines all the necessary processes for atomic spectroscopy simultaneously: sample vaporization, atomization, and excitation. A LIBS measurement is performed by forming a plasma on the sample surface and then collecting and spectrally analyzing the light emanated from this plasma. 

This document has been written to provide a knowledge base and reference text for anyone interested in learning more about LIBS technology, whether for personal interest, a source for a scientific paper, or research into potentially purchasing an analyzer. Topics will include the LIBS process, applications, advantages of this technology over other elemental analysis processes, and components of a LIBS system.

How does it work?

Every LIBS analysis starts with a laser. A high-powered light beam is generated from a source and then directed through a series of mirrors and optical lenses. The beam is then focused on the surface of the sample to be analyzed. The concentration of thermal energy at this point on the sample is sufficient to generate a plasma, which also emits light. This light is analyzed by the spectrometer after being directed through another optical apparatus not unlike the one the laser beam goes through. The spectrometer then separates the light it picks up into different wavelengths (wavelengths determine the “colour” of an electromagnetic wave). The light intensity of each wavelength is then measured by a sensor to generate an intensity-per-wavelength graph, or spectrum. It is by analyzing the characteristics of this spectrum that it becomes possible to determine the sample’s chemical composition with a high degree of accuracy and confidence. 

The LIBS process works on almost any type of sample. Metal, rock, soil, fluid, mud, gas and other samples can all be analyzed using LIBS technology. 


There are many possible LIBS technology applications for two main reasons, the most significant being the variety of sample types that can be analyzed. Whether the sample is metallic, mineral, rocky, soil, solid, liquid or even gaseous, it’s extremely likely that it can be analyzed by LIBS. The second reason lies in the large amount of information contained in a single emission spectrum obtained by LIBS (chemical composition, elemental concentration, mineralogy, etc.).

Chemical composition control

A quantitative LIBS analysis can be performed to accurately determine the concentration of one or more elements in an observed sample. This type of analysis is particularly useful for chemical composition control. In the field of metallurgy, a foundry or a steel plant can use a LIBS analyzer to differentiate, identify and control the different alloy types manufactured in the plant. This LIBS application allows the company to readjust its manufacturing process when composition deviates from the expected and to avoid errors such as delivering the wrong alloy to a customer due to mislabelling at the warehouse.

Process Analytical Technology (PAT)

LIBS can also be used for qualitative and semi-quantitative analysis. This type of analysis is more often used in applications aimed at classifying samples, rather than accurately measuring the elemental composition of the samples. Acquisition for this type of measurement requires less than one millisecond, making the process ideal for a large volume of samples. 

An ideal use for this type of application is process monitoring. Continuous LIBS analysis of an ongoing process allows real-time monitoring of the presence variation of the different elements within the process, as well as its physical properties. This data can be immediately used to keep the process extraction in the optimal zone and thus provide better control of the finished product’s average chemical composition. This technique can be compared to a continuous video of the process versus taking a limited number of snapshots. Furthermore, traditional manual sampling procedures can lead to under-sampling of the process stream and thus to an inaccurate estimate of the finished product’s chemical composition. This disadvantage can be corrected with a LIBS process analyzer directly on the production line since it can locate and identify parameters that influence product yield and quality, in addition to providing an accurate measure of the average chemical composition of the product in each batch.

Machine learning

The combination of LIBS analysis with the versatility of a deep learning algorithm opens the door to several more advanced applications of the technology.  

A semi-quantitative analysis of several samples can sort them by classification in real time. This is a very useful application for sorting metals or ores on a conveyor, whether to obtain classification statistics for a batch of analyzed material, or to be combined with a mechanical sorting system further down the conveyor to separate samples into different bins by classification. 

A qualitative analysis on a sample’s surface area can identify the different areas of interest automatically. Adding a recognition algorithm trained on a database of identifiable minerals makes it possible to infer the lithology as well as the mineralogy of the analyzed area. It’s an excellent method to determine, with a very good confidence interval, the level of different minerals contained in a sample.


There are a variety of reasons for choosing laser spectroscopy over other existing analytical methods. Here are the main advantages of using a LIBS analyzer.

Real-time results

LIBS analysis results are always available instantly; this is one of the technology’s greatest advantages. Samples sent for laboratory analysis can take several months to yield initial results. This wait is caused by several factors, including the long duration of the traditional chemical analysis method as well as the enormous demand leading to increasingly long queues. 

Additionally, because a LIBS system is much more compact than an analytical laboratory, it can be easily transported and installed at sample production sites requiring analysis (e.g. a mining site). The results can, therefore, be obtained on site instantly, requiring much less time, transportation, and communication than using analytical laboratories.


A single spectrum generated by LIBS can be used to observe the presence of any element of the periodic table. The same experimental results used to analyze a specific element or elements can, therefore, be archived and reused later to observe different elements in the future. 

LIBS is also one of the only analysis technologies sensitive to light elements. Sulphur, phosphorus, carbon, hydrogen, lithium and beryllium can all be observed by LIBS whereas the most widespread technologies are unable to do so.

Financial benefits

Acquiring most elemental composition analysis equipment requires a daunting upfront financial investment, particularly in the case of an automated analytical system. But the high speed of LIBS and its lower system manufacturing costs compared to other analytical technology make an automated LIBS instrument a very effective way to increase productivity, with a very short return on investment time. An on-site system eliminates the wait for results and the costs of outsourcing an analysis to a laboratory.  

Most LIBS automated system applications are simple to use, requiring no more than one worker with minimal training, unlike other analytical technologies requiring several workers and a nuclear licence in the case of ionizing radiation technology.

Little to no sample preparation

A proper X-ray fluorescence analysis requires the sample to be ground to uniformity, often mixed with binding agents, and then compressed or fused into lithium borate glass. Most analytical technologies require some form of preparation, but not LIBS. A sample can be analyzed as is, without preparation, and laser-material interaction will produce good results. LIBS can also easily analyze samples covered with dust, soil, water or snow. 

When necessary, physically cleaning or digging into a sample can be done directly with the LIBS laser. It’s powerful enough to clean and dig into the sample in order to reach the targeted surface. 


All steps of the LIBS process can be easily automated, making it possible to design analyzers that can operate autonomously, 24 hours a day. Automating an analysis process also minimizes or, in some cases, eliminates the risk of human error. 

An automated LIBS analyzer requires minimal monitoring and no special training or certification. It can be operated by a single worker, reducing the company’s financial and logistical burdens.


LIBS analyzers currently in use around the world have unequivocally demonstrated their durability on several occasions. One of the most striking examples is the robotic Curiosity Mars rover, launched by NASA in November 2011 as part of the Mars Science Laboratory mission. Curiosity has been on the surface of the Red Planet since August 2012 and its LIBS analysis module, dubbed ChemCam, is still working perfectly to this day, without any instrument recalibration. 

The same durability is clearly seen in the analyzers here on Earth. LIBS equipment with proper protective housing is durable and compact enough to be transported and deployed in a variety of environments. For example, a prospecting team could bring an analyzer to various drilling sites to perform on-site, same-day measurements.

Big data

The world is currently in the midst of the fourth industrial revolution, which is fundamentally characterized by “smart automation and the integration of new technology into a company’s value chain.” (

Machinery on production lines is rapidly becoming smarter, more automated and more autonomous. LIBS technology is ready for Industry 4.0, thanks to its ability to operate without interruption and with minimal supervision. 

The tremendous amount of information contained in each spectrum collected makes LIBS the ideal technology for sample digitization. For example, in the case of core analysis, elemental core composition information could be analyzed by LIBS and then archived on a server, mitigating the need to store a large number of core samples in a core-storage library. The large amount of information collected can then be used in deep learning algorithms. The more data available, the more informed are the decisions that the artificial intelligence algorithms can make.

Components of LIBS


The key component of a LIBS system is the laser, which has multiple settings, such as wavelength, energy, pulse duration, and beam quality.  

Although some wavelengths are better suited for LIBS experiments (e.g. UV), the fundamental wavelength of the Nd:YAG crystal, measuring 1064 nm, is probably the most widely used for stability and repeatability purposes as well as application constraints. Many types of lasers are used in LIBS experiments: excimer laser (gas laser), flash-lamp lasers, diode-pumped solid-state lasers, fibre lasers, and microchip lasers. Historically, flash-lamp pumped lasers were most commonly used due to their relatively affordable price. However, over the last few years, diode-pumped solid-state lasers have become increasingly affordable and have the advantage of air-cooling, which requires less maintenance. Although a LIBS plasma can be generated from a laser pulse ranging from a few femtoseconds to a microsecond, the majority of lasers use pulses that last from a few nanoseconds to a few hundred nanoseconds and is again mostly influenced by the price and maintenance of the lasers.  


Generation optics 

Although a plano-convex lens alone is sufficient to form a laser-induced plasma, there is a wide range of optical configurations that allow for LIBS analysis. Using a telescope makes it possible to have better control over the size of the focal spot and, especially, to focus at a distance. For example, ChemCam (a LIBS system that is part of the Curiosity NASA rover currently on Mars) uses a Schmitd-Cassegrain telescope that allows it to generate and collect light up to a distance of more than 7 metres. For industrial and laboratory applications, motorized lenses are often used to adjust the laser beam’s focus on the target. To sample larger surface areas, XYZ translation stages are generally used. Mounted mirrors on galvanometers can also be used to increase the analysis speed and avoid being restricted by mechanical translation stages. Once again, application is a significant deciding factor in choosing the most appropriate approach to laser-induced plasma generation. It is important to note that all of these factors may influence repeatability, reproducibility, long-term accuracy, and measurement precision. 


Collection and transmission optics 

The collection optics configuration is another component of a LIBS system that must be designed with great care, since it is this part of the equipment that collects and transmits the plasma-emitted spectral information. There is a continuum of configuration to collect the photons emitted by the laser-induced plasma. The two main categories are:

  1. transmission optics, and
  2. reflective optics

Both of which come with their own benefits and drawbacks. In general, reflective optics are known for being more achromatic than transmission optics. The use of optical coatings that have excellent reflective performance over a wide spectral range is quite common for reflective optics, given how readily available they are.  

As for transmission optics, it is more difficult to find a material with good transmission properties over a wide spectral range that also has refractive indices that are useful for lenses. Generally, in the case of LIBS, the collection can be done through a collinear or quasi-collinear configuration. And here, application is the deciding factor in choosing the most appropriate configuration for setting up a LIBS system. For example, for applications requiring the use of vacuum ultraviolet (VUV) emission lines (e.g. for detecting sulphur, phosphorus, carbon, chlorine, etc.), it is necessary to empty the optical channel by vacuum or use a VUV transparent gas such as argon, as molecular oxygen (O2) has a high absorption rate due to the charge transfer in the molecule.  


Spectrometer and detector 

There are several types of spectrometers that can be used to analyze the light emitted by the plasma in atomic emission spectroscopy. Again, the desired application is the deciding factor in choosing the most appropriate spectrometer. Spectral resolution is a key feature of a spectrometer as it will resolve potential spectral interferences, especially for applications where the matrix contains a high concentration of iron. Another important feature of the spectrometer is its bandwidth (i.e. the wavelength range needed to separate photons). For example, an echelle spectrometer has very high resolution, performs simultaneous multi-element analysis and has high bandwidth. However, the spectrometer’s dynamic range can be problematic with respect to the trace elements in a matrix with intense emission lines because they may cause detector glare and phantom lines. 

A Czerny-Turner configuration may be a good choice for these applications when trace analysis is required in a matrix with intense emission lines. The limited bandwidth of a Czerny-Turner configuration is useful for this type of application as it makes it possible to limit the range of wavelengths entering the spectrometer. However, sequential multi-element analysis can become time-consuming, tedious, and sometimes impossible, if the sample is small or if the required information is localized at the surface. Obviously, if the total cost of the system is not a major concern, it is possible to set-up several Czerny-Turner spectrometers to make simultaneous multi-element analysis possible. Paschen-Runge-type spectrometers benefit from most of the advantages of the Czerny-Turner and echelle configurations in terms of bandwidth, dynamic range, sensitivity, and multi-element analysis. However, the elements this spectrometer will analyze must be known before the spectrometer is manufactured. It is therefore usually dedicated to specific applications. Other configurations such as flat-field spectrometers, Wadsworth spectrometers, and transmission networks are also possible and have interesting options, but are much less frequent in LIBS. 

Adaptation of a fast detector that allows for time-resolved photon detection must also be considered in choosing the right spectrometer. iCCD and iCMOS cameras are the most widely used in research because of their sensitivity and flexibility to temporally resolve plasma emission. Obviously, the cost of these detectors can become a concern when manufacturing a commercial LIBS system. In addition, acquisition speed becomes very important for imaging or multi-element LIBS tomography applications, which benefit enormously from high acquisition speed. In fact, analyzing one square centimetre every 50 µm requires 40,000 measurements, which can be done in 67 minutes at 10 Hz, but in only 40 seconds at 1,000 Hz. Since most of the samples analyzed with LIBS imaging are generally homogeneous samples, it is essential to cover as large an area as possible to avoid the risk of missing the observation of certain phenomena. The new CMOS sensors are particularly suited to this kind of application because they allow for acquisition rates greater than 1,000 Hz. Once again, the deciding factor in choosing the right detector relies on the application.

How does LIBS work?

The laser focuses on the target

Using a lens, the laser pulse focuses on the surface. The laser then provides a synchronization signal that represents the start of a LIBS acquisition on a laser-induced plasma.

The laser heats the surface

The laser pulse quickly heats the surface of the target for the first 100 femtoseconds. The solid is then quickly brought to its evaporation temperature.


Matter is ablated or vaporized from the target surface by various nonlinear optical phenomena. These phenomena eventually get compromised by shielding and will completely subside by the end of the laser pulse.

Plasma creation

The ablated vapour will also absorb laser energy, heating further to cause splitting of molecules into atoms, and continues to heat until the atoms’ electrons are stripped from their orbits (ionization). Electrons absorb laser light rapidly; their absorption is proportional to wavelengths emitted to the fourth (wavelength4) (according to Mie Theory). Cascade ionization is initiated at this precise moment. Plasma (the fourth state of matter) is formed, which is defined as a mixture of highly active electrons and positively charged atomic nuclei (ions). Plasma as a whole is neutral, but is locally electrically charged and very sensitive to the action of internal and external electromagnetic fields. Plasma dynamics are complex, and have been the subject of intensive research from 1963 to now, leading to our current understanding of laser-induced plasma generation. 

Shiedling and heating

The electrons generated by the cascade ionization will heat the plasma to a sustained temperature of around 30,000 K at the end of the laser pulse. The cascade ionization can be referred to as the laser-induced plasma shielding phase. During this phase, laser energy is gradually concentrated with the increase in plasma temperature at the expense of the laser ablation phenomenon. In other words, the amount of ablated matter will decrease as the electrons shield it from the laser light. 

Plasma at the end of a laser pulse

At the end of a laser pulse, plasma temperature is at its peak. Without the external laser energy, the plasma will begin to cool from 30,000 K to ambient temperature. The lifetime of the plasma depends on several factors, including fluence (energy per unit area [J/cm²]) and especially irradiance (power per unit area [GW/cm²]), volume of plasma, ambient pressure, pulse duration, etc. In general, for pulse durations of a few nanoseconds, laser-induced plasma generated by 50 mJ with a focal spot of 500 µm can last up to 50 microseconds. Atomic and molecular recombination, on the other hand, can last up to 200 microseconds (when there are no more free electrons, by definition, it is no longer a plasma; it is a hot gas T < 3000 K). If a plasma is induced by a pulse of 1 to 2 millijoules per focal spot of approximately 30 micrometres (for example: 10 to 100 µm), the life of the plasma will rarely exceed 2 microseconds. 

Plasma after a few microseconds

In spectroscopy, it is generally preferable to wait a few hundred nanoseconds after the laser synchronization signal for the plasma temperature to reach a temperature range that is favourable (i.e. between 5,000 K and 8,000 K). At the end of the laser pulse, from a spectroscopic point of view, the generation of an intense continuum (blackbody emission pattern centred on plasma temperature) emitted by the electrons is observed. In addition, the species in the plasma (i.e. -III -IV) do not necessarily have useful information from a chemical analysis point of view. Once the analytical temperature range is reached, integration time in time-resolved spectroscopy is set to approximately 30 microseconds.


Once the plasma disappears, a crater remains, left by the laser ablation and the interaction between the plasma and the target. Furthermore, depending on the composition of the target, redistribution of condensed matter may be observed. The successive generation of several plasmas on the surface of a solid sample makes it possible to carry out a study of chemical composition in relation to depth. Thanks to LIBS imaging, it is also possible to generate a series of elementary maps that can be converted to three-dimensional tomography.

LIBS imaging

LIBS microanalysis is currently a particularly active field of application in recent literature. Until recently, sample analysis by LIBS consisted of effectively accumulating analytical signals without paying much attention to the spatial distribution of elements on the surface of the sample. It was standard practice to assume sample uniformity on a macroscopic/millimetric scale if a large enough area was sampled. However, microanalysis by LIBS reveals surprising information about the surface uniformity of a solid sample. In fact, spatial information can be used to understand several physio-chemical phenomena on the history of the sample’s forming, or any other property derived from it. In addition, LIBS tomography is a leading tool for studying the three-dimensional structure of solid samples. 

LIBS microanalysis consists of sampling the surface of a material point by point by saving an emission spectrum for each X and Y coordinate of the sample (see Figure 1). Once the surface has been scanned, a data cube containing all the obtained spectra is used to organize the information in the X and Y coordinates of the sample. It is then possible to extract the information the classical way (i.e. by extracting the net intensity of an emission line) in order to build an elementary plane corresponding to the wavelength of the selected element, as shown in Figure 1. 

Figure 1. LIBS sampling and information management for multi-element imaging.

The advantage of sampling a large surface area of a solid is that its average composition can be accurately estimated. Analysis speed is an important parameter for fast access to signals that approximate the average sample composition. Until very recently, the acquisition speed for LIBS was limited to less than 20 measurements per second (20 Hz). The work of Rifai et al. demonstrated it would be possible and useful to use an acquisition frequency of 1,000 measurements per second (1,000 Hz). This high acquisition frequency allows the study of the multi-element spatial distribution of the sample’s surface on a new scale. In addition, performing successive scanning of this sampling pattern makes it possible to penetrate the surface and build multi-elemental maps in relation to depth. Three-dimensional imaging or tomography of the sample can then be done through the assembly of these multi-elemental maps.

Outbound data: a gold mine of spectral information  

Atomic emission spectra obtained by laser-induced plasma spectroscopy have a considerable amount of information to offer. Just think of the emission lines of the same atom that can come from the atom (e.g. Z-I), or from the ions (Z-II and Z-III). “Atom A” combined with “Atom B” can form an AB molecule and emit molecular emission bands that correspond to electronic transitions in the molecule. 

Furthermore, the proportion of different species in the plasma provides information on the excitation temperature in the plasma. Without delving into the physics of the Saha-Boltzmann plots, it is possible to deduce the excitation temperature in the plasma, its electron density, and several other fundamental physical parameters. It is important to note that in the first decades of the development of the LIBS technique (1963 to 2000), scientists revealed the complexity of laser-matter interaction and particularly its non-linear aspect. They often used the fact that laser-matter interaction can generate different spectral signatures depending on the matrix, even if the elemental chemical composition was very similar—a disadvantage of the technique, often referred to as the matrix effect. 

The first publications on the application of multivariate statistics and chemometrics to LIBS (around the year 2000) reversed the negative aspect of matrix effects into a positive one, which increases the selectivity of the technique. The fact that the spectral signature of a plasma on a matrix varies, even if the elemental chemical composition is similar, turns out to be a significant advantage from a qualitative point of view. This makes it possible to distinguish changes not perceptible with other elementary technology, such as X-ray fluorescence (XRF), for example. 

From an analytical point of view, these “matrix effects” can be used to increase the selectivity of qualitative and quantitative analyses. The importance of using information-rich spectra is illustrated by the possibility of correlating the spectral signature of excipients with active agents in a pharmaceutical formulation. In atomic emission spectroscopy (AES), molecular emission bands were originally considered undesirable contributions to analytical signals in a flame (e.g. acetylene/oxygen) or even in an inductively coupled plasma (e.g. argon in an ICP torch). 

These molecular contributions (e.g. OH, CN, C2 bands, etc.) coming mainly from water (H2O, the solvent) and acetylene (C2H2, the fuel) understandably hinder chemical analysis when using these classical techniques. However, this is not the case for laser-induced breakdown spectroscopy. Given that the sample’s ablated matter becomes plasma, molecular contributions have a completely different implication from an analytical standpoint. The increase in selectivity in LIBS spectroscopy results from a combination of several direct and indirect variables, such as molecular bands, plasma temperature (which is proportional to the ratio between the ionic line and the atomic line of the same atom—Saha’s law), electron density, etc. The spectral fingerprint of a sample is unique. Indeed, even if samples come from the same sampling group, a slight variation in chemical composition from one sample to another is normal and is what makes them unique. In chemometric terms, we speak of a multivariate normal distribution (Gaussian for one variable, deformed bell for two variables, etc.). Although there are small variations between samples from the same group due to the selectivity of the LIBS technique, it is easy to distinguish a group of samples from a completely different one in terms of elemental chemical composition. The last twenty years of scientific literature is particularly rich regarding the application of chemometrics to LIBS spectroscopy.

The future of LIBS

LIBS technology isn’t new but as the years go by, new applications for it become possible and are being developed.


Outer Space

Since the equipment required to perform LIBS experiments consists of very few mobile parts, this technology is an ideal candidate for applications in outer space. The ChemCam LIBS camera on the Curiosity Rover on Mars is an excellent example of what can be done with LIBS equipment on extraterrestrial grounds. It is also conceivable that LIBS technology could be used in NEO asteroid mining exploration in the near future, given the high portability of the required equipment and the ability it has to instantaneously analyze samples in situ. 


Planet Earth 

Industrial probes 

LIBS does not only hold future possibilities in outer space. The fourth industrial revolution is currently underway on Earth and continues to bring important changes in its wake. There is a growing need for increasingly intelligent industrial probes and LIBS is the ideal technology for this type of application. It is already easy to pair automated LIBS instruments with a production line; these will become increasingly intelligent with the latest machine learning techniques. The LIBS industrial probe will be able to make informed decisions autonomously and communicate those decisions to other intelligent probes further down the line. 

Autonomous robotic probes  

When elemental analysis is carried out over a large area (e.g. soil analysis), a quick and efficient method is required, given the large amount of land that needs to be covered during sampling. The ideal solution might be some type of autonomous robotic probe capable of moving, analyzing and making autonomous decisions on where to sample next by using information from previous analyses. LIBS has a very beneficial analysis speed for this application, but for it to truly be effective, it will need to be paired with a fast and robust robotic platform, and machine learning.