Atomic Absorption Spectroscopy, Principles and Applications

A short history of light

One can go as far back as the 17^th century when Sir Isaac Newton discovered that white light breaks up into its constituent spectral colors when directed through a glass prism.¹ From this work, he developed the corpuscular theory of light (the fact that light consists of particles), as opposed to only having a wave-like nature, which opened the door to several discoveries almost two centuries later.
 

Alan Walsh,² a Lancashire-born physicist was working in his garden on a Sunday morning somewhere in the early 1950s when an idea that would solve a huge analytical chemistry puzzle popped up in his mind: how to accurately measure small concentrations of metallic elements by spectroscopy. The normal procedure in spectroscopy was to vaporize an element and measure the emission spectra, but this technique was flawed and produced inaccurate results. Walsh decided to measure absorption, not emission. By teatime on Monday morning, he showed that it could be done. It took him several more years to convince manufacturers to use atomic absorption spectroscopy (AAS) for the detection of metals, but he eventually succeeded. Today, most analytical laboratories will boast at least one atomic absorption spectrophotometer.

What is atomic absorption spectroscopy?

AAS is an analytical technique used to determine the concentration of metal atoms/ions in a sample. Metals make up around 75% of the earth’s chemical elements. In some cases, metal content in a material is desirable, but metals can also be contaminants (poisons). Therefore, measuring metal content is critical in many different applications, which we will explore later in this article. Suffice to say for now that it finds a purpose in quality control, toxicology and environmental testing to name a few.

What is the principle of atomic absorption spectroscopy?

The basic principles of AAS can be expressed as follows. Firstly, all atoms or ions can absorb light at specific, unique wavelengths. When a sample containing copper (Cu) and nickel (Ni), for example, is exposed to light at the characteristic wavelength of Cu, then only the Cu atoms or ions will absorb this light. The amount of light absorbed at this wavelength is directly proportional to the concentration of the absorbing ions or atoms.

The electrons within an atom exist at various energy levels. When the atom is exposed to its own unique wavelength, it can absorb the energy (photons) and electrons move from a ground state to excited states. The radiant energy absorbed by the electrons is directly related to the transition that occurs during this process. Furthermore, since the electronic structure of every element is unique, the radiation absorbed represents a unique property of each individual element and it can be measured.

An atomic absorption spectrometer uses these basic principles and applies them in practical quantitative analysis. A typical atomic absorption spectrometer consists of four main components: the light source, the atomization system, the monochromator and the detection system (Figure 1).
 

Figure 1: Schematic diagram of a typical atomic absorption spectrometer.

In a typical experiment, the sample, either liquid or solid, is atomized in either a flame or a graphite furnace. The free atoms are then exposed to light, typically produced by a hollow-cathode lamp, and undergo electronic transitions from the ground state to excited electronic states. The light produced by the lamp is emitted from excited atoms of the same element that is to be determined, therefore the radiation energy corresponds directly to the wavelength absorbed by the atomized sample. A monochromator is placed between the sample and the detector to reduce background interference. From here, the detector measures the intensity of the beam of light and converts it to absorption data.

While solid samples can be used for AAS this analysis is usually restricted to the more expensive graphite furnaces where the sample can be heated by controlled electrical heating as opposed to a direct flame. Also, AAS is normally only used to analyze metal atoms. The main reason for this is that metals have narrow, bright and clear single emission and absorption lines.

Atomizing techniques - flame atomic absorption spectroscopy (FAAS)

FAAS is mainly used to determine the concentration of metals in solution in parts per million (ppm) or parts per billion (ppb) ranges. The metal ions are nebulized as a fine spray into a high-temperature flame where they are reduced to their atoms and subsequently absorb light from an element-specific hollow cathode lamp.
 

Figure 2: The atomization process in FAAS.

While this method has proved to be a robust technique for routine metal determinations, it does have drawbacks. Firstly, it has limited sensitivity because of the spectral noise created by the flame. This aspect has, however, been improved as the technology has evolved. The main drawback is that one can only measure one metal at a time and as different lamps are required for each element, the lamp must be changed each time you want to analyze for something different. Also, a large part of the sample is lost in the flame (up to 90%) in FAAS, further influencing the sensitivity.

Atomizing techniques - graphite furnace atomic absorption spectroscopy (GFAAS)

In GFAAS, a type of electrothermal atomization, a sample is placed in a hollow graphite tube which is heated until the sample is completely vaporized. GFAAS is much more sensitive than FAAS and can detect very low concentrations of metals (less than 1 ppb) in smaller samples. Using electricity to heat the narrow graphite tube ensures that all of the sample is atomized in a period of a few milliseconds to seconds. The absorption of the atomic vapor is then measured in the region immediately above the heated surface. Naturally, the detection unit does not have to contend with spectral noise, leading to improved sensitivity.

Figure 3: A typical graphite tube atomization process. Credit: Mertmetin96, recreated under the Creative Commons Attribution-Share Alike 4.0 International license.

Atomizing techniques – specialized techniques

In glow-discharge atomizing systems, an atomized vapor is produced which can be swept into a cell for absorption detection. Glow-discharge cells can be used as an accessory to most FAAS systems. The solid-state sample is introduced onto a cathode after which high energy argon (Ar) ions, generated by an electrical current running from the anode to the cathode, are used to bombard and eject atoms into the path of radiation. The process is called sputtering.

For this technique to work, samples must be electrical conductors or must be mixed with a conductor such as finely ground graphite or Cu.³ The detection limits are in the low parts-per-million range for solid samples.

Hydride-generating atomizers are mostly used for the analysis of heavy metal samples and even other elements like arsenic (As), tin (Sn), selenium (Se) and bismuth (Bi). Samples are diluted and acidified before being mixed with sodium borohydride (NaBH₄). A metal-hydride is generated and transferred to the atomization chamber by an inert gas like Ar. Here, the sample is introduced to a flame or furnace to produce the free metal atoms, ready for detection.

Mercury (Hg) is the only metal that does not atomize well in a flame or furnace. To analyze Hg, a special technique called, cold-vapor atomization, is employed. The Hg sample is acidified and reduced before it is swept through by an inert gas. The absorption of the gas is then determined.

Radiation sources in AAS and signal detection

There are two main sources of radiation available to AAS, namely, line source (LS) and continuum source (CS). CS is typically produced by deuterium lamps and emits light over a broad range of wavelengths, while LS, on the other hand, emits radiation at specific wavelengths, normally produced by a hollow cathode lamp.

To improve detection limits, monochromators are used to select the same wavelength of light absorbed by the sample and to exclude other wavelengths. This ensures that only the element of interest will be detected. The detector converts the light signal into an electrical signal proportional to the light intensity. Older equipment used photomultipliers for detection purposes, but nowadays they are replaced by charge-coupled device (CCD) detectors. The resolution per pixel in a CCD array detector is about 1.5 pm, small enough to use continuum light. Furthermore, if a CCD detector uses 200 pixels, for example, each of them makes its own measurement of absorbance or integrated absorbance, meaning that the equipment has 200 independent detectors providing better signal-to-noise ratios.

Until the late 1990s, LS spectrometers were the only devices used in AAS. Walsh had already realized in the 1950s that the main difficulty of atomic absorption was "that the relations between absorption and concentration depend on the resolution of the spectrograph, and on whether one measures peak absorption or total absorption as given by the area under the absorption/wavelength curve".⁴ He, therefore, excluded the use of continuum light sources (sources employing a wide range of wavelengths) because doing so would need a resolution of about 2 pm, something that could not be attained at the time.

In the meantime, many researchers saw the potential advantages of using a continuum radiation source. One such advantage would be that more than one analyte could be detected simultaneously. This led to the development of all kinds of radiation sources, mono- and polychromators, detectors and evaluation principles for CS AAS. It was only in 1996 that a research group from Germany proposed a completely new spectrometer concept.⁵ They used a xenon (Xe) short-arc lamp, a high-resolution double monochromator and a linear CCD array detector. In the early 2000s the first high-resolution-CS AAS (HR-CS AAS) spectrometer was manufactured by a German company.

HR-CS GFAAS has limitations, mainly because each element needs different atomization temperatures. In fast, sequential determination using HR-CS FAAS however, an unlimited number of analytes are available (atomized simultaneously) and can be detected by simply changing the wavelength to move from one analyte to the next. The equipment also makes it possible to adjust flame composition, stoichiometry and burner height within about a second. Previously, in LS AAS, these adjustments took up a long time, especially when several analytes were determined one after another.

There are several ways to do background corrections in LS AAS, most of which have limitations due to the use of a photomultiplier tube or solid-state detectors which integrate over the spectral range transmitted by the exit slit. In contrast, the software coupled to a CCD detector in HR-CS AAS automatically makes corrections on both sides of the analytical line. Any changes in radiation intensity are corrected automatically to the baseline, leading to extremely low noise levels.⁶

Interpreting an atomic absorption spectrometric output

The interpretation of outputs in AAS is quite simple and follows Beer’s law, namely that absorbance is directly proportional to concentration. This means that the analyte’s concentration correlates with the electrical output received from the detector. One of the ways to determine the unknown concentration of an analyte is to use several solutions of known concentrations to calibrate the instrument. The curve shows radiation (absorbance) versus concentration and once the sample is measured, the concentration value could be obtained from the calibration curve.

In the analysis of lead (Pb) for example, light at 283.3 nm, corresponding to one of the spectral lines of Pb, is passed through the flame containing the sample. The beam contains light at 283.3 nm. The Pb atoms absorb light, and excitation of their electrons occurs. The absorption value is obtained and the unknown concentration of Pb in the sample can be read from the calibration curve. These days, all of this is done by the computer, but it is important to understand the principle of a calibration curve. In Figure 4 below, there are four absorbance values of analyte solutions with known concentrations. The absorbance of the sample is measured and the concentration can be deduced from the graph.
 

Figure 4: A typical calibration curve. 

In Figure 5, a typical readout of absorbance versus wavelength is shown from analysis for Sn, cadmium (Cd) and iron (Fe) simultaneously, for example from a digested food sample. The spectral lines of the three elements, as indicated by the wavelength values, are close to each other which allows for simultaneous detection in this case.

Figure 5: Example of Sn, Fe and Cd detection in a food sample. 

Background correction

When obtaining absorption spectra, the detector might pick up signals from other particles in the flame, leading to background interference. This does not mean that the obtained spectrum is not representative of the sample, it simply results in a loss of spectral detail, such as peak broadening and the appearance of peaks in places where the sample does not absorb. Proper background correction techniques can minimize these deviations and enhance the signal from the analyte.

1) Koirtyohann and Pickett⁷ developed the first automatic background technique using a combination of a CS, such as a deuterium lamp, and a hollow cathode lamp (single LS). The radiation passing through the instrument alternates between the deuterium continuum and the analyte source and then subtracts background absorption from the total absorption measured with the hollow cathode lamp. This method has flaws though, the main one being the fact that deuterium is an ultraviolet source which limits the wavelength range available to the analyst.

2) Smith and Hieftje⁸ introduced a background correction method based on the high and normal current pulsing of a hollow cathode lamp. The total absorbance, which includes interferences, is obtained when the normal current is in operation and the background is obtained with the high-current pulse. This technique only works well with volatile elements. In addition, it can only be used for FAAS and the continuous pulsing decreases the lifetime of the hollow cathode lamps.

3) The Zeeman-effect background correction method, which uses an alternating magnetic field to produce background versus sample data, is mostly used in LS GFAAS and has developed substantially over the years.^9, 10 However, it does have limitations when the sample contains another metal, different from the analyte, with spectral lines close to the analyte wavelength, like Ni and Fe for example.

4) In HR-CS AAS, in contrast, the software of the instrument automatically selects correction pixels on both sides of the analytical line, which do not show any absorption lines. Any increase or decrease of the radiation intensity that is the same for all correction pixels will then be corrected automatically to the baseline. This means that the signal output is already corrected for lamp noise and any continuous background. This is also in contrast to LS AAS, where background corrections contribute significantly to the baseline noise (at least a factor two or more increase in noise). This in turn will influence the precision, limit of detection (LOD) and the limit of quantification (LOQ). Interested readers are directed to the review article of Resano¹¹ et al. for more information.

What are the applications of atomic absorption spectroscopy?

AAS finds wide application in fields that vary from mining to pharmaceuticals, environmental control and agriculture. Most heavy metals are toxic and should be avoided as far as possible. If you ever had to use an antibiotic, chances are that the quality control process to ensure that the drug is free from the catalysts like palladium or platinum used to make them was performed by an AAS.¹² Similarly, the food, cannabis and health supplement industries make use of AAS to ensure that their products are safe for consumption.^13, 14, 15 In mining, a lot of focus is on the recovery of precious metals like gold from old mine heaps. With the help of AAS, the amount of gold can be quantified to determine whether it would be profitable to extract it.¹⁶ The analysis of drinking water is probably one of the most important applications of AAS, especially in places where the environment is not properly cared for.¹⁷

References

1.    Filho HJI, Salazar RF, Capri M, Neto AC, de Alcantara MAK, Peixoto AL. State of the art trends in atomic absorption spectrometry. Intechopen. Published 2011. doi:10.5772/26076
2.    Hannaford P. Alan Walsh 1916-1998. Australian Academy of Science. Available at: https://www.science.org.au/fellowship/fellows/biographical-memoirs/alan-walsh-1916-1998. Published 2000. Accessed November 5, 2021.
3.    Winchester MR, Marcus K. Glow discharge sputter atomization for atomic absorption analysis of nonconducting powders. App. Spec.. 1988;42(6):941-944. doi:10.1366/0003702884430335
4.    Walsh, A. Atomic Absorption Spectroscopy – stagnant or pregnant? Anal. Chem.. 1974;46(8):698A-708A. doi:10.1021/ac60344a720
5.    Beker-Ross, H.; Florek, S.; Heitmann, U.; Weisse, R. Influence of the spectral bandwidth of the spectrometer on the sensitivity using continuum source AAS.  Fresenius' J. Anal. Chem. 1996, 355, 300. doi:10.1007/s0021663550300 
6.    Welz B, Vale MGR, Pereira ER, Castilho INB, Dessuy MB. Continuum source absorption spectrometry: past, present and future aspects – a critical review. J. Braz. Chem. Soc. 2014;25(5):799-821. doi:10.5935/0103-5053.20140053 
7.    Koirtyohann SR, Pickett EE. Background corrections in long path atomic absorption spectrometry. Anal. Chem. 1965;37(4):601-603. doi:10.1021/ac60223a048
8. Smith SB, Hieftje G. A new background-correction method for atomic absorption spectrometry. App. Spec.. 1983;37(5):419-424. doi:10.1016/j.jpba.2010.11.030
9. Gleisner H, Eichardt K, Welz B. Optimization of analytical performance of a graphite furnace atomic absorption spectrometer with Zeeman-effect background correction using variable magnetic field strength. Spec. Acta Part B: Atomic Spectroscopy. 2003;58(9):1663-1667. doi:10.1016/S0584-8547(03)00141-1

10. de Loos-Vollebregt MTC, de Galan L. Theory of Zeeman atomic absorption spectrometry. Spectrochimica Acta Part B: Atomic Spec.. 1978;33(8):495-511. doi:10.1016/0584-8547(78)80058-5

11. Resano M, Aramendia M, Belarra MA. High-resolution continuum source grpahite furnace atomic absorption spectrometryfor direct analysis of silid samples and complex materials: a tutorial review. J. Anal. At. Spectrom. 2014;29:2229-2250. doi:10.1039/c4ja00176a

12. Lewen N. The use of atomic spectroscopy in the pharmaceutical industry for the detection of trace elements in pharmaceuticals. J. Pharm. Biomed. Anal. 2011;55(4):653-651. doi:10.3109/03639045.2010.509726

13. Borges AR, Bazanella DN, Duarte AT. et al. Development of a method for the sequential determination of cadmium and chromium from the same sample aliquot of yerba mate using high-resolution continuum source graphite furnace atomic absorption spectrometry. Microchem. J. 2017;130:116-121. doi:10.1016/j.microc.2016.08.010

14. Choi S, Yarlagadda P, Abdullah-Al-Wadud M. Sequence determination of Cd and Pb in honey by imcomplete digestion-high resolution continuum source graphite furnace atomic absorption spectrometry. App. Mech. and Ma.. 2014;511-512:22-27. doi:10.4028/www.scientific.net/AMM.511-512.22

15. Krawczyk M. Determination of macro and trace elements in multivitamin dietary supplements by high resolution continuum source graphite furnace atomic absorption spectrometry with slurry sampling. J. Pharm. Biomed. Anal. 2014;88:377-384. doi:10.1016/j.jpba.2013.09.016

16. Fouad HK, Elrakaiby RM, Hashim MD. The application of flame atomic absorption spectrometry for gold determination in some of its bearing rocks. Am. J. Anal. Chem. 2015;6(5):411-421. doi:10.4236/ajac.2015.65040 

17. Behari JR, Prakash R. Determination of total arsenic content in water by atomic absorption spectroscopy (AAS) using vapour generation assembly (VGA). Chemosphere. 2006;63(1):17-21. doi:10.1016/j.chemosphere.2005.07.073