Introduction
Elemental analysis plays an important role in many aspects of life today.
Industries producing or processing raw materials require reliable quality control of both their base materials and their finished products, and in many cases also use spectral analysis instruments for monitoring their processes. Research and development departments require flexible analytical techniques to handle their constantly changing requirements. Additionally, waste and waste water also need to be checked for compliance with national regulations before being deposited or released into the environment.
The oil industry uses elemental analysis not only to monitor the production of their fuels, oils and additives themselves, but also to study the effectiveness of their products by analyzing wear metal content and additive consumption in used oils. The latter is also of high interest for people and companies who monitor the health of their well-oiled machines, e.g., turbines for energy production or motors in cargo ships and other large vehicles.
In agriculture, elemental analysis is commonly used for verifying the state of the soil in order to determine both type and amount of fertilizer required for improving quality and yield of the harvest. The finished food products themselves need to be checked for toxic elements, too. Other healthcare related applications include the monitoring of drinking water, analyzing toxic elements in medical products and examining the release of toxic metals and metallic allergens from toys and clothing.
Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) has become the leading technology for routine analysis of liquid samples as well as materials that can be easily turned into a liquid form by dissolution or digestion. Its origin lies in so-called spectroscopes and spectrographs that allow visual evaluation of spectral lines on a film, which required much experience and time in order to obtain reliable results. The developments in electronics and data processing allowed instruments to appear on the market in the middle of the 1970s which helped overcome these obstacles and enable the routine use of optical emission spectroscopy in laboratories.
During the last decades, ICP-OES has seen dramatic improvement. While the first commercially available instruments relied on time consuming sequential measurements or had limited availability of emission lines due to the use of Photomultiplier Tubes (PMT), today's systems are able to capture wide spectral areas simultaneously in a short time thanks to modern CCD technology.
The following chapters present, an easy-to-understand introduction into the physics and the technology of an optical emission spectrometer, so that you will know a little more about this “black box”.
History
Through the observations of Isaac Newton and Christiaan Huygens, it was already clear in the 17th century that light was a special phenomenon that behaved in a manner that was apparently contradictory to the ideas accepted at that time. Newton discovered that sunlight could be dispersed by a glass prism into different colors that merged into each other (continuous spectrum). A second prism unites the colors back into white light.
Based on his observations, Newton postulated a particle character of light. In turn, Huygens discovered phenomena in the dispersion of light (diffraction, reinforcements and cancellations) that were more easily comparable to waves.
In 1887, Heinrich Hertz proved that visible light is a section of the electromagnetic spectrum, thus having a frequency (ν, in 1/s) and a wavelength (λ, in m or nm). These values can be converted into each other with the help of the constant c (speed of light): ν=c/λ
Optical Emission Spectroscopy almost always works with the wavelength λ in nm (= 10-9 m) as unit.
Around 1860, Robert Wilhelm Bunsen and Gustav Robert Kirchhoff observed that chemical flames became colored when certain salts are introduced into the flame. The prime example is NaCl, table salt, which emits (radiates) the intensive orange-yellow colored spectral lines (λ = 589.0 nm and λ = 589.6 nm) of the Na atom.
Joseph von Fraunhofer discovered lines in the spectrum of sunlight that had an interrupted spectrum (thus the term spectral lines). He proved that these “missing locations” belonged to certain chemical elements, such as hydrogen or helium. Today, this phenomenon is called an absorption spectrum, i.e., atoms absorb light of very specific wavelengths in the light spectrum. This was the foundation of spectroscopy: the proof of chemical elements by means of evaluating light spectra. Today, we could not imagine chemistry or astronomy without spectroscopy.
Quantum theory was established at the beginning of the 20th century through the work of Werner Karl Heisenberg and Albert Einstein, which provided an explanation for the apparently contradictory behavior of light. The term wave-particle duality means that light, on one hand, has properties of particles and, on the other hand, it has properties of waves.
As we will see, spectrometers take advantage of these two properties in a classical way in order to disperse light (diffraction at the grating) and to measure the intensity (photoelectric effect).
Structure of Matter
To understand the process of emission or absorption of light, knowledge of the structure of matter is necessary. For this purpose, we use a very model-like concept (Bohr’s atomic model) to clarify the principle.
All matter is built from atoms. An atom is the smallest unit, and it can only be divided further through special measures, such as nuclear fission. Currently, there are over 115 different types of atoms, which are called the chemical elements. Examples are carbon (C), iron (Fe), chromium (Cr), and silicon (Si). Groups of the same or different atoms, which are combined due to chemical reactions, are called molecules. Examples here are CH4 (methane, a bond of carbon and hydrogen) or H2 (hydrogen, a bond of two H atoms). In the optical emission spectroscopy (OES) discussed here, we analyze the chemical elements, i.e., molecular compounds are resolved down to the atomic level.
Structure of Atoms
Atoms are made of the atom nucleus and electrons, which, in the model, circle the nucleus in different orbits (“shells”), like the planets circling the sun. The nucleus is made of electrically positively charged protons and electrically neutral neutrons. The electrons are electrically negatively charged. From the outside, an atom appears to be electrically neutral, i.e., the number of protons and electrons is equal.
An element is identified by the number of protons in the nucleus (“atomic number”). The number of neutrons, however, can vary for a given element, especially for larger atoms. These atoms with the same number of protons but different numbers of neutrons are referred to as isotopes of the corresponding element. Some Isotopes are stable while others are not stable and decay corresponding to the law of radioactive decay with a certain half-life.
Nearly the entire mass of an atom is concentrated in the nucleus, the electrons play almost no role in this value. This fact is important to understand the high mobility of electrons, which function practically only as charge carriers and thus assume a central role in many processes, such as current conduction, heat conduction, and chemical bonding.
For OES, only the electron processes in the outer shell (“optical electrons”) play a role. In contrast, X ray radiation, with its much higher energy, also reaches electrons of the inner shell.
Periodic Table of Elements
All chemical elements are represented in the so-called periodic table of elements. This arrangement follows a certain system, which considers the arrangement of the electrons in the outer shells. This is because these electrons ensure that different atoms with the same outer electron configuration (elements that are arranged in a column with each other) behave chemically very similar.
Ionization and Recombination
Positively or negatively charged atoms are called ions. This process (“ionization”) occurs through the loss or addition of electrons, resulting in the formation of cations or anions, respectively.
Free ions are very reactive, they rapidly form salts or neutral molecules by attracting ions of the opposite charge or, in case of cations, free electrons. This process is called recombination.
Excitation and Relaxation
When the outermost electron is raised to a shell that is farther away from the nucleus (but still belongs to the same atom or ion), the atom or ion enters a so-called excited state.
Because an exited state is higher in terms of energy than the atomic ground state, it is not stable. Thus, the excited atom (or ion) attempts to change back into the ground state. The energy difference between the excited and the ground state is radiated in the form of light.
Both excitation and ionization of atoms requires energy from the outside. This can happen in the form of light radiation, heat, or also electrical energy.
The quantum theory of atoms states that the energy gaps of the electron shells cannot have arbitrary values, but instead are different by very definite energy quanta. As a consequence, only fixed (discrete) energy amounts can be absorbed, which are a multiple of the natural constant h (“Plank’s constant”). If there is no sufficient energy, then even if one continues to radiate there will be no excitation. The same also applies for relaxation: only very certain energy amounts can be emitted.
Because each type of atom is combined differently from the subatomic components, there are quite a few energy levels at which energy can be absorbed or emitted, but each level belongs to a very certain type of atom. This means that if one knows the energy of the absorbed or emitted radiation, then one knows the type of atom and thus the chemical element.
For practical analysis, the wavelengths usually employed lie in the Ultra-violet and Visible regions of the optical spectrum, between about 130 nm and 780 nm. Due to of the large number of transitions possible in most elements, their emission profiles (spectra) can be very complex. Over 4000 different emission wavelengths (referred to by spectroscopists as emission lines) have been identified for iron alone. Multielement spectra can be very complex indeed, and ionized atoms give rise to emission spectra of their own, to further complicate the issue.
Kirchhoff’s Law
This law summarizes the previous statements again:
Atoms and ions can only absorb the same energy that they also emit. This means that they absorb light of the same wavelength that they also emit.
Atomic absorption and emission are equal but opposite processes.
Planck’s Law
Light is nothing other than a form of energy and can be described as an electromagnetic wave phenomenon very simply through its characteristic values of frequency (ν) and wavelength (λ).
One can see the relationship of energy and frequency or wavelength purely qualitatively from the formula: The greater the energy the higher the frequency and the smaller the wavelength. If one looks at the electromagnetic spectrum, it is obvious that the energy of infrared radiation is never sufficient to excite atoms. Only the bonding links in molecules are being excited, which then leads to heating. Energy levels in the range of visible light are the first values sufficient for elements which can be easily excited (Na, K, Li), but the most important range lies in the ultraviolet (UV) range.
Boltzmann Distribution
To be able to observe an emission from a particle (atom or ion in the case of optical emission spectroscopy) which changes from one state to another, the starting state has to be present. The probability of finding a particle in a specific state is described by the Boltzmann distribution.
The Plasma
On earth, matter is typically encountered in one of three states: solid, liquid or gas. Plasma is the fourth fundamental state of matter. It is obtained by ionizing a gas through either heating to very high temperatures or by exposing it to a high voltage, e.g., an electric spark. The result is a mixture of positively charged particles and free electrons, making the plasma electrically conductive and susceptible by electromagnetic forces. Once the heat or energy input stops, however, the charged particles recombine and return to a neutral gas state.
Spectroscopy
Basic Setup
Based on the principles described above, all optical spectrometers are fundamentally working in the same way. They have a source to irradiate or excite the sample, and a detector to analyze the light. They either compare the light directly emitted from the source to the light remaining after the same light has passed through the sample (absorption spectroscopy), or they analyze the light emitted from the sample after excitation from a source (emission spectroscopy).
Light Dispersion and Detection
The light needs to be detected and converted into a spectrum. While some spectroscopic techniques (e.g., Infrared spectroscopy) use interferometers and then apply a Fourier transformation to convert the signal obtained by the detector into a spectrum, the more common approach is to first disperse the light into a spectrum which is then measured using suitable detectors.
The Grating
For modern spectrometers, diffraction gratings are used nearly exclusively for spectral dispersion. In contrast to prisms, there is less absorption of wavelengths, because the light does not pass through a medium (e.g., glass).
The effect takes advantage of diffraction at the slits, where the elementary waves at the edges exhibit certain path differences according to the diffraction angle. Therefore, overlapping regions result with cancellations (minimum) and reinforcements (maximum). Now, for the grating, many individual slits are etched and vapor-deposited next to each other on a carrier layer, so that the diffracted light is reflected.
The distance between two slit centers is called the grating constant. Today, typical gratings have 1800 - 3600 lines/mm. The more lines (the smaller the grating constant), the stronger the light is diffracted and the brighter are the maximums.
The diffraction at the grating leads to the appearance of several maximums that are symmetrical to the grating normal. The central maximum is the directly reflected light ray and is referred to as the light of 0 order. The subsequent maximums to the left and right are light of the 1st, 2nd, 3rd, etc. order. As the light intensity quickly decreases with increasing order, one normally only uses light of the 1st order which provides sufficiently high light intensities. The physical properties of the grating are described by the grating equation:
n*λ=G (sinα+sinβ)
G = grating constant; α = incident light angle; β = reflected light angle
CCD Detectors
Modern CCD detectors (Charge Coupled Device) have replaced Photomultiplier tubes (PMT) in many areas of optical spectroscopy. These detectors produce no currents, but instead electrical charges. They operate without a high-voltage power supply and are produced in large numbers as line array detectors (the light-sensitive pixels are arranged next to each other) or as a two-dimensional array (the pixels are arranged as a two-dimensional plane).
Advantages of CCD detectors:
- Relatively low price
- Small size
- Requires no special power supply
- Robust
- Analytical flexibility, because a large wavelength range can be covered
- Simplified optical design since lower number of components are needed
ICP-OES
As indicated by its name, Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) is a technique that uses a plasma as a source and relies on optical emission for analysis. However, unlike many other spectrometers, the sample is not simply placed in-between source and detector.
ICP-OES is mainly used for liquid samples, which first need to be turned into an aerosol (“nebulization”) and then are injected into the plasma. Solid samples can be directly analyzed if a means of vaporization is available, e.g., laser ablation (LA-ICP-OES) or electrothermic vaporization (ETV).
The high temperatures in the plasma are sufficient to break down the sample into atoms and provide the energy for ionization and excitation.
Structure of an Emission Spectrometer
For practical analysis with OES, several essential components must be provided:
- An energy source to atomize the test sample and excite the atoms. In an ICP-OES Spectrometer this is an Inductively Coupled Plasma operating at several thousand K. The plasma is created in a torch assembly, normally made of quartz, and is powered by a radio-frequency generator.
- A sampling system to introduce the sample into the plasma. Various types of pumped nebulizers are used to convert the sample into a fine mist spray (aerosol), usually coupled with a spray chamber that removes large droplets that could pass into the plasma and generate signal noise and instability.
- High resolution optics to observe the emissions from the plasma and to separate and isolate the specific emitted wavelengths for the elements to be measured.
- A detector system to measure the intensity of the light emissions.
- Electronics to acquire the detector signals and to control the functions of the spectrometer.
- A computer with software for calculation and display of emission spectra and concentration values.
The Excitation (Plasma)
A plasma based on the noble gas Argon (Ar) is commonly used as excitation source in ICP-OES. An electric spark is used to create the initial ions and free electrons in the plasma, which is then continuously maintained by inductively coupling radio frequency energy into it. Therefore, it is referred to as an Inductively Coupled Plasma.
The sample is then introduced into the plasma as an aerosol, resulting in an excitation of the sample atoms and ions and consequently emission of element-specific radiation as described above.
The plasma and its emissions can be observed in two ways: either from the side (“side-on plasma”, SOP) or from the tip (“end-on plasma”, EOP), commonly also referred to as radial and axial plasma observation, respectively.
Axial Plasma Observation
In axial plasma observation, light from the entire central channel is sampled. This provides higher sensitivity because light from the complete emission zone is used. It is, however, also strongly influenced by effects in the excitation zone, and the recombination zone in front of the plasma has to be removed from the optical path. Therefore, this plasma observation is less suitable for samples with a high amount of total dissolved solids (TDS) and organic solutions.
One problematic effect observable only in axial plasma observation is the so-called EIE (“Easily Ionizable Elements”) effect, which can result in wrong determination of alkaline elements. This led to the development of so-called “Twin Interface” or “dual view” instruments with an added secondary light observation path which is used to also view the plasma in a radial fashion. While this technique does increase the time required for a measurement, it is considered preferable to the alternative: the addition of a highly concentrated ionization buffer with its side effects of increasing the matrix load and possible addition of contamination to the samples.
Radial Plasma Observation
The radial plasma observation provides higher stability. Only a part of the emission zone is used for analysis, but since matrix effects are at their lowest in this area, radial plasma observation has a much higher matrix tolerance and is far better suited for organic solutions.
While this reduced viewing volume does result in a somewhat lower sensitivity, it conversely also means an extended dynamic range ideal for the determination of major compounds.
The Generator
The energy needed to maintain the plasma is provided by a radio frequency generator. Radiation of RF frequencies is strictly limited by international regulation, and only certain frequencies may be used in instrumentation without comprehensive and expensive shielding. The frequencies most commonly used for ICP-OES are 27.12 MHz and 40.68 MHz.
This RF power is inductively coupled into the plasma and thereby making the plasma itself part of the oscillating circuit. Thus, when a sample is introduced it can change the load of the plasma, potentially introducing instability into the system and hence degrading the precision of the analysis. This is particularly pronounced when samples containing high TDS or organics are being analyzed.
There are two possible ways of compensating for this load change in the plasma: one is to fix the generator frequency and match the impedance to the plasma, and the other is to vary the frequency to compensate for the change in load. Each of the permitted frequencies is at the center of a narrow frequency band within which the frequency can change and remain legal. Therefore, the necessary adjustments must be fast enough to react to sudden changes in the plasma load and prevent the plasma from extinguishing while also ensuring that the frequency stays within the legal boundaries. The bandwidth permitted at 27.12 MHz is considerably wider - nearly ten times - than that at 40.68 MHz. For this reason, instruments operating at 27.12 MHz can handle high salt concentrations better than those running at 40.68 MHz.
Sample Introduction Systems
The diversity of applications that can be done by ICP-OES results in a variety of sample introduction systems available on the market. While they all have the same goal – turning the liquid sample into an aerosol to be injected into the plasma – each has its own advantages and disadvantages which makes it better for specific sample types. Most sample introduction systems consist of two components: the nebulizer and the spray chamber.
The nebulizer creates the primary, polydisperse aerosol, i.e., it contains droplets of various sizes. The majority of commonly used nebulizers are pneumatic nebulizers. The sample liquid is brought close to a small opening through which a gas at high pressure expands, thus forming a gas jet that draws the liquid into itself and breaks it up into droplets.
The spray chamber is responsible for separating the aerosol droplets by size. Small droplets can pass through the chamber and enter the plasma, while bigger droplets hit the wall of the chamber and are subsequently transferred to the waste.
Concentric Nebulizer
Concentric nebulizers consist of two capillaries mounted in a concentric fashion. The small inner capillary is used for the sample liquid, while the gas flows through the larger outer capillary. They produce a very fine aerosol but can be easily clogged if the liquid contains particles.
Concentric nebulizers are available in both glass and PTFE variants.
Cross Flow Nebulizer
The cross-flow nebulizer is named after the fact that gas and sample flow meet each other at right angles. As the sample capillary is larger than the one used in concentric nebulizers, this nebulizer is better suited for samples that may contain small particles or have a high amount of total dissolved solids. The aerosol, however, is not as fine as the one obtained from concentric nebulizers.
Parallel Path Nebulizer
The two capillaries in a parallel path nebulizer are not mounted one inside the other, but rather next to each other. This design allows for a larger sample capillary while maintaining the outer form factor of a concentric nebulizer.
V-Groove Nebulizer
Modern V-groove nebulizers are built in a way comparable to parallel path nebulizers, getting their name from the V-shaped tip. Both the small gas capillary and the huge sample capillary have their openings in the groove of the V, with the sample capillary placed at some distance from the gas capillary. The sample flows down the groove and hits the gas jet at a right angle, which makes the nebulization more comparable to a cross flow. Thanks to the large diameter of the sample capillary, this nebulizer can handle even larger particles. The droplets in the aerosol are typically larger than those created by the other nebulizers, and it must be mounted in the right orientation to function properly.
Scott Spray Chamber
The Scott type spray chamber, also known as double pass spray chamber, consists of two concentric tubes with the aerosol being injected into the inner tube. For a droplet to reach the plasma it has to move all the way to the other end of the chamber and then back again through the outer tube.
The benefit of this spray chamber is a very constant creation of aerosol for the plasma, but due to the large inner surface, wash-out effects are more common than for the smaller cyclonic chambers.
This type of spray chamber is typically used in combination with a cross flow nebulizer, but through the use of an adapter it can also be used with the other nebulizers mentioned above.
Cyclonic Spray Chamber
These spray chambers have a round, remotely sphere-like design. The aerosol is injected through an off-center/tangential opening at the equator of the chamber from where it moves around the inside in a cyclonic fashion. Small droplets can exit the chamber through a small opening at the top while larger droplets run down the wall into the drain.
Cyclonic spray chambers are available with and without an additional center tube. They are made from either glass or PTFE and can be used with all nebulizers mentioned above except for the cross flow.
The Optical System
The optical system is the heart of the spectrometer and definitively determines its performance. Modern ICP-OES spectrometers are capable of simultaneously measuring a wide wavelength range, commonly utilizing one of two types of optics: Echelle and Paschen-Runge.
Echelle Optics
A combination of an Echelle Grating and a second, orthogonally mounted dispersive element (a prism or another grating, the “Cross Disperser”) forms the core of an Echelle optic. The resulting Echellogram consists of multiple lines, each line containing a certain wavelength segment obtained from one of the diffraction orders created by the first grating. A two-dimensional CCD detector is required to obtain the resulting spectra.
The main advantages of an Echelle optic are its small size and a good spectral resolution in the UV region (200 – 250 nm). However, at higher wavelengths the resolution increases noticeably while at lower wavelengths the absorption of the transmission optical elements that are required to build this type of optic reduces the light throughput.
As this type of optic uses light of higher diffraction orders (and thus lower intensity), its CCD detector requires strong cooling (-30°C and lower) to reduce the electronic noise level.
The Paschen-Runge Mounting
At the end of the 19th century, Henry August Rowland created concave gratings, which were then used by Friedrich Paschen and Carl Runge at the beginning of the 20th century to build the first of this type of optical systems. The concave reflection grating not only disperses the light but also focuses it on a circular path, provided all optical components are mounted on that circle and the focal length of the grating equals the diameter of the circle. The photo plates used in the initial spectrometers were first replaced with photomultiplier tubes (PMT) and later with linear CCD detectors.
This optical concept uses only the intensive light of 1st diffraction order obtained from one or more gratings. As no transmissive optical elements are used inside the optic, few light is lost and low wavelengths (down to 130 nm) become accessible if either the optical system is evacuated or it is purged or filled with a gas that does not absorb in these spectral regions (typically Argon or Nitrogen).
Due to the high light intensities, cooling of the CCD detectors is not required, but thermal stabilization at moderate temperatures improves performance. Furthermore, the spectrum created by a single grating provides a rather uniform resolution across a larger spectral range. Optimized designs like the SPECTRO’s ORCA optic (Optimized Rowland Circle Alignment) make use of multiple gratings to increase the spectral coverage of the optic while still avoiding components that strongly absorb light.
The resolution of the optic is determined by two factors: the grating constant and the focal length of the optic. As a result, high resolution optics utilizing the Paschen-Runge mount tend to be larger compared to the Echelle design.
Computer and Software
Without a meaningful evaluation, our spectrometer can only measure amounts of light in a defined wavelength range. How does a light-quantity measurement instrument become an analysis system?
Optical emission spectroscopy is a so-called relative analytical technique, i.e., one first has to calibrate the instrument with samples of known content of analytes to establish a correlation between the concentration in the sample and the light emitted at various wavelengths. Only then one can proceed to analyze unknown samples.
Today, thanks to modern, powerful computer systems and software, the capabilities of an analytical system go far beyond just evaluating light intensities and storing results. Complete spectra of both calibration standards and samples can be stored and offer far-reaching possibilities for post processing when necessary. Spectral evaluation offers various possibilities for subtraction of the plasma’s background radiation, inter element correction for overlapping emission lines of different elements, line switching to cover large concentration ranges for an element, etc.
Automation is another application field that has become increasingly important, from simple autosamplers which provide the analytical instrument with one solution after the other to highly complex robots that handle everything from sample preparation and dilution to transport of the liquid into the instrument. When combined with such a system, modern analytical software can be used to create a complex control logic which regularly checks the state of the instrument by measuring control standard and automatically re-calibrating the system and re-measuring samples if necessary.
Calibration
As already mentioned before, only the calibration turns a spectrometer into an analytical instrument. The basic principle of a calibration is simple: known samples are measured, and the light amounts of different elements are recorded. The correlation between concentration and emitted light is used to calculate calibration functions for all wavelengths used in an application. Even though in ICP OES these calibration functions are typically linear, it is common practice to make sure to use multiple points for the calibration of each element to cover the complete concentration range that can be expected in the samples.
When samples enter the plasma, the available energy is used to vaporize, dissociate, partially ionize and excite everything in the sample. As the generator keeps the total energy in the plasma at a constant level, the energy available for the excitation of an analyte is lower in a sample that contains a high amount of matrix than in one consisting of merely water. In order to obtain correct results for samples containing high amounts of TDS, it is therefore important to matrix-match the standards to the samples, and analyzing various types of samples often requires multiple calibrations.