C/S analyzers refers to products that measure Carbon and Sulfur in inorganic materials. Alternative denomination is Elemental Analyzers even if only 2 elements are measured. ISO standards talk about Infrared Absorption Method after combustion. HORIBA uses the acronym EMIA, which stands for Element Material Infrared Analyzer.
O/N/H analyzers refers to products that measure Oxygen, Nitrogen, and Hydrogen in inorganic materials. Depending on the model, 1, 2 or 3 gases are measured. Alternative denomination is also Elemental Analyzers. ISO standards describe the principle as «Inert Gas Fusion Technique. HORIBA uses the acronym EMGA, which stands for Element Material Gas Analyzer.
C/S & O/N/H analyzers both measure inorganic materials. They are often found together in metallurgical laboratories notably, as the elements they measure are critical for materials’ performance and life time and not possible to get with alternative techniques at the speed and the level of sensitivity and accuracy required.
HORIBA is the world leader in gas analysis for automotive and environment.
The EMIA & EMGA series were developed in the 80’s by HORIBA, first as a response to demands of Japanese steel manufacturers. By combining its gas analysis core technology with advanced electric furnaces, HORIBA has been able to provide high end instruments for metallurgy that ensures ease of use and superior performance.
C/S & O/N/H analyzers measure inorganic solid samples that usually are present in the shape of small blocks, chips, powders or granulates. For instance, with a C/S analyzer, a ceramic crucible is typically filled with a small block (often 1g) of sample. Combustion facilitators are added (W, Sn) and the components are heated to very high temperatures under an O2 flow leading to the melting of the sample. The C and S from the sample combine with O2 to generate gases (CO, CO2, SO2) that are measured. The principle of C/S analysis is therefore an oxidation.
Even from this schematic description of the principle, several key points can be highlighted.
First the combustion needs to be complete. If not the information collected could not be fully related to the material analyzed. Optimum recipes that assure full and complete combustion are available for the different materials. Second the sample needs to be weighed. 1g of sample is a typical targeted value for C/S, well adapted to the furnace and melting conditions and usually representative enough for a given material but 0.95 g or 1.05g will be equally melted. In order to correlate the total gas amount measured to the mass concentration of C and S in the sample, the sample weight must be checked prior to the analysis using a high precision balance. On the other hand, decreasing the sample weight is sometimes needed when high concentrations are present in order to avoid detector saturation or incomplete combustion.
Third when the amounts to measure are extremely low (a few ppms of C for instance in steel) sources of contamination need to be carefully eliminated. Crucibles, for instance, are often pre-burned prior to the analysis to remove residual C. However about 20ppm of C may originate from crucibles that have not been cleaned, and this is of course significant when low concentrations have to be measured (or higher concentrations but with smaller amount of material). Similarly, surface cleaning of the sample is done prior to punching, and the carrier gas may have to be purified.
Finally the combustion could generate unwanted gases (such as water) and dust that need to be carefully eliminated to avoid contamination of the analysis. The C/S instruments from HORIBA for instance, have advanced patented features that optimize dust collection and greatly minimize and facilitate the cleaning.
Typical applications include of course metallurgy (ferrous and non ferrous materials), but such instruments are also used to control catalysts, solders, rare earths, Li ion batteries, electrical components, semiconductors, glasses etc.
Throughput is of utmost importance for C/S and N/O/H. In steel plants, 24 hours operation is the rule and for each melt a full analysis must be done in less than 5 minutes – including sample preparation. Therefore the instruments must be readily available at any time and capable to provide accurate determinations in typically less than 2 minutes.
Automation is also often the rule and multiple dimensions of automation are proposed from autosamplers allowing batches of samples to be measured to fully automatic labs that can run unattended.
Multiple techniques provide elemental information for inorganic materials including Spark Optical Emission, Glow Discharge Optical Emission, XRF, ICP OES etc.
XRF cannot be compared with C/S and O/N/H analyzers. XRF does not measure N and O and EDXRF cannot measure C. (For WDXRF only high levels of C – above % level - can be measured). Similarly ICP OES that offers excellent detection limits in liquids does not measure C/S or O/N/H.
The high end Spark Optical Emission Instruments have made significant progress in the last 10 years and now represent a serious competitive technique to C/S & O/N/H analyzers with the advantage that all elements together with C/S and O/N can be measured by Spark. However, Spark instruments still do not properly measure H and for both low levels and high levels of C/S or N/O the performance of the top line C/S and N/O/H analyzers are better and do correspond to the requirements for advanced materials.
Instruments primarily dedicated to the measurement of organic materials can possibly also be used for inorganic ones. But the measurement techniques are not certified which is mandatory in industry and, compared to C/S & N/O/H analyzers, the performance is more modest and the measurement complexity is higher.
So in summary we can say that there is a significant market for C/S & O/N/H instruments for which the driving forces are the sensitivity and the repeatability of the units, together with the ease of operation and maintenance.
As seen before, the C/S and N/O/H instruments combine electric furnaces and detection systems. For detection, 2 techniques are used (IR absorption and TCD) depending on the element and the sensitivity required.
The above table relates to the most commonly found instruments. However for C and S, another technique featuring a resistance furnace is also available.
We will describe later the differences of principles between the different furnaces. Let us first focus on the principles of detection.
The main components of an NDIR sensor are an infrared light source, a sample chamber (cell), an optical filter and an infrared detector.
The gas in the sample chamber causes absorption of specific wavelengths according to the Beer–Lambert law, and the attenuation of these wavelengths is measured by the detector to determine the gas concentration.
Log (I0/I) = e x c x L
I0: Incident light intensity
I: Absorbed light intensity
e: Molar extinction coefficient
c: Molar concentration
L: Light pass length
Fig 6: Expression of the Beer-Lambert law
The detector has an optical filter in front of it that eliminates all light except the wavelength that the selected gas molecules can absorb. The IR signal from the source is usually chopped or modulated so that thermal background signals can be offset from the desired signal. IR detectors are used for CO2, CO, SO2 and H20 (when H is to be measured by IR). For CO2, two detectors are used with two optimum filters to measure with the highest accuracy, both low and high amounts.
Example: Measurement of CO
NDIR is a core technology of HORIBA used for Motor Exhaust Gas Analyzers, Ambient NOx Monitors and Stack Gas Analyzers. This gives the capability to handle unique developments: direct CO detection for instance, is difficult and only HORIBA can process it within its C/S analyzers. In addition, in-house manufacturing by skilled engineers assures high quality and reliability.
TCD is used for N and H (when H is measured by TCD). The following table shows the thermal conductivity of N2 and H2 and the ones used as carrier gases: He or Ar.
The principle of operation requires that the measured gas and the carrier gas have very different conductivities for optimum sensitivity. This is the reason why He is necessary for N2 detection. (If Ar is used the sensitivity will be 100 less and therefore generally not useful for the applications. In addition, this would necessitate a major instrumental change).
Wheatstone bridge TCD
The reference cell is filled with the carrier gas only, the sample cell will see a change when the combustion takes place and the measured gas (N2 or H2) is carried and introduced.
TCDs have no selectivity – only a change in resistivity is measured. If more than one gas is introduced in the cell (for instances CO2 or H20 together with N2) the measurement will be affected by all. This is why the other gases must be carefully filtered prior to TCD detection to assure that only N2 (or H2 if measured by TCD) is introduced.
H can be measured by IR (after conversion to H2O) or directly by TCD.
The sensitivity by using TCD is better by a factor of at least 10 which explains why the TCD is the technique of choice for low H detection in the most demanding applications
Fragilisation by Hydrogen
Let us now review the different furnaces present on the C/S and O/N/H instruments.
The principle of the induction furnace is somewhat similar to what is available in ICP. The coil surrounds the ceramic crucible in which the sample and the accelerators (to facilitate the combustion) are placed. High temperature is reached (over 2300°C). When the measurement is finished the crucible is disposed and a new measurement can be done. The crucible that has been burned is hot but the rest of the chamber is not.
The exact temperature of the sample is not known with an induction furnace, however it is possible to control the applied current and change it during an analysis cycle to optimize the measurement.
The following example illustrates this function and shows a steel sample polluted with base oil. The red curve
“Temperature” rather refers to the applied current; 2 levels are set up. At a low level, only the C from the oil pollution is released, then at a higher level, the steel melts and C/S from the steel are released.
This example is of course an extreme case and usually for the highest accuracy sample pre-cleaning prior to measurement is applied.
For similar reasons, when operating a C/S analyzer to measure low C, the crucibles need to be free of contamination and therefore have to be pre-burned and kept in a protective vessel before usage. Pre-burning the crucibles is generally done using a dedicated unit or a muffle furnace.
In a resistance furnace the exact temperature at the sample stage is controllable and can be changed during a measurement. The resistance furnace is horizontal and the maximum temperature that it can reach is 1450°C. The sample is directly placed in a ceramic boat in the middle of the furnace where the temperature is uniform. Usually no accelerators are used.
In order to minimize the set up time, the resistance furnace is maintained at high temperature between 2 measurements operation therefore requires more care than in the case of an induction furnace.
In a resistance furnace, the sample can be slowly burned and programmable temperature curves can be set up allowing for instances to precisely separate contributions of the surface and of the bulk of a material. Applications include toner for copy machines or the measurement of free carbon in Silicon Carbide (the melting point of SiC is 2830°C so only the non binded C will come out there).
A C/S with a resistance furnace requires manual operation where the instruments with induction furnaces can be more easily automated.
In the N/O/H analyzers the sample is placed in a graphite crucible that is sandwiched between the 2 electrodes of the impulse furnace. A very high current is generated and maintained (by a series of voltage pulses hence the name). The crucible conducts electricity but does not melt (Carbon sublimes at 5800°K only – a temperature higher than the melting temperature of all metals including W), however the sample inside the crucible and the eventual fluxes used will completely melt.
The Oxygen present in the sample will combine (this is a reduction) with Carbon from the crucible and generates CO that will be carried together with N2 and H2 also coming from the sample.
The output current of the impulse furnace is monitored and it can possibly be changed during an analysis sequence to optimize some results (also separating surface contamination from bulk content for instances).
Impulse furnaces are demanding in energy and usually require dedicated power lines.
The capability of a C/S analyzer to measure CO simplifies the instrument; CO as a result of combustion is always present (about 1% compared to CO2 if combustion is fine) and it therefore contributes to the total C amount.
With CO2 only detectors, CO needs to be oxidized into CO2 prior to the measurement which requires an additional oxidation agent (catalyst) that unfortunately also converts SO2 into SO3 (a serious pollutant which has to be trapped) – leading to more parts to be maintained and regularly changed. If SO3 and water are not correctly trapped (by lack of maintenance), they will combine into sulfuric acid that may induce serious corrosion issues on the optical filters and in the tubes.
Measurement principle on instruments without CO detection Capability
HORIBA has patented several developments around the furnace first to avoid dust to deposit at unwanted places and facilitate the dust collection without the need of brushes and second to remove the water adsorption by properly locating and heating the dust filters.
This lead to a generation of instruments where no brushing or vacuum cleaner activation is needed between measurements and where fast and easy maintenance needs to be done after only every 200 samples.
Efficient dust removal (before measurement (left), and after 200 measurements (right))
This long “Time Between Maintenance” (TBM) is crucial for C/S routine operation. Demonstrating videos built in the software show in details the procedures for cleaning the source and replacing the consumables.
Here is the schematic of a 3 gas analyzer N, O and H (with H converted to H20 and measured by IR). Versions for N/O only also exist (model HORIBA EMGA 920) in which the H20 detector is absent.
Principle of a N/O/H an
The main components of the O/N/H analyzers beyond the furnace and the detectors are purifier and filters. Most are similar to the ones found in C/S analyzers (previouslydescribed).
CuO heated, acts as an oxidizing agent for CO and H2 The main components of the O/N/H analyzers beyond the furnace and the detectors are purifier and filters. Most are similar to the ones found in C/S analyzers (previouslydescribed).
CuO heated, acts as an oxidizing agent for CO and H2
Dual loading principle
The principle of a “H only” analyzer is shown below. The carrier gas here is Argon which has a thermal conductivity that very much differs from Hydrogen.
The furnace is the same as in the O/N/H analyzers, simply longer crucibles are usually used for H. The detector is a TCD. Purifiers and filters have their classical role. However an additional device is required just before the TCD detector – namely a separation column, similar to the ones used in gas chromatography equipment.
The different filters could not remove Nitrogen and so, without separation of Hydrogen and Nitrogen, both gases would simultaneously arrive to the TCD and the result would therefore be biased. With the introduction of a GC column, the detection of the two gases is not simultaneous any longer: the light Hydrogen will travel faster and therefore will be detected first followed by N, allowing software separation of the two signals.
A last feature that could also possibly be found for the other elements but which is of special interest for H analysis is the gas dosing technique.
There are actually few good solid reference samples for H measurement and therefore the calibration of the detector could be done by bypassing the furnace stage and introducing in the carrier gas a known amount of Hydrogen from an external gas bottle (usually a very small one). As Hydrogen gas manipulation (even within small bottles) might be restricted in some labs, calibration is sometimes done with He (instead of H2) using known correction factors.
C/S and N/O/H analyzers are routinely found in industries. Whether you measure N and O in steels, Oxygen in Barium Titanate (BaTiO3), electronic Cu grade or Li batteries electrodes, Carbon or Nitrogen in Tantalum, H in Titanium, or other products the manufacturers provide recipes for optimum performances.
Attention to details (sample preparation, accurate weighting, proper selection and dosage of accelerators and regular maintenance) is extremely important to get optimum and stable results.
Automation is often considered with C/S and N/O/H analyzers to optimize the workflow and minimize the random errors.
Several levels of automation can be found such as automatic dispensers of fluxes and crucibles in N/O/H analyzers or autosamplers that manipulate the crucibles in the C/S analyzers to fully automatic systems able to prepare, weigh, measure the samples and send results to a central Laboratory Information Management System (LIMS) within a predefined time frame.
The presence of halogen elements (F, Cl, Br, I) in the samples to be analyzed may lead to issues as these elements could combine with H compounds from the combustion and generate acids potentially harmful for the detectors and the tubings. Br was often found in electronic boards, F is present in some glasses or CaF2 coatings etc.
It is therefore crucial to trap these compounds at an early stage before their negative effect affects the sensitive components of the instruments.
The software can be operated on any modern computer including touch screen. The user interface includes Analysis help with best recipes for operation, Maintenance help with built-in videos and Trouble Shooting help if a warning occur.
The extraction curves are shown and stored together with the elemental values.
Optimum operating procedures can be set up and stored for each application, such as ramping the temperature, as in the following example done on a O/N/H analyzer.
C/S and O/N/H analyzers are crucially important in many industries, as the elements they measure have a major influence on the materials’ properties and cannot be analyzed by other techniques with the speed, the sensitivity and the accuracy provided by the combustion techniques. Let us finish by a quote that applies to the hard work provided by our instruments: I had an extremely busy day converting oxygen into carbon dioxide
• User manuals and software manuals for EMIA and EMGA, HORIBA
• Steels, edited by Wei Sha, Springer 2013
• Hydrogen Storage Technology, edited by L. Klebanoff, CRC Press 2013