AA or ICP – Which Do You Choose?

Application Note
Inductively Coupled Plasma-Optical Emission Spectrometers

Author
Geoff Tyler

Introduction
For many analysts Atomic Absorption Spectrometry (AAS) is a well established and understood technique. However, even though Inductively Coupled Plasma Emission Spectrometry (ICP-ES) instrumentation has been commercially available for over a decade, the technique has proven to be more complex. This article discusses the main differences between the two techniques.

AAS Versus ICP
The basic difference between the two techniques is that one relies upon an atomic absorption process while the other is an atomic/ionic emission spectroscopic technique. The next essential difference is the means by which the atomic or ionic species are generated. A combustion flame or graphite furnace is typically used for AA while ICP-ES uses a plasma.

B. Sample Throughput In ICP-ES. Ti. The typical maximum temperature for an air/acetylene flame is 2300 °C while for nitrous oxide acetylene. Al) are better by ICP than by flame AAS. nitrogen. Detection Limits The comparison of detection limits in Table 1 highlights the following differences: • • • • Furnace AA detection limits are generally better in all cases where the element can be atomized. and the halogens (for example. Bi and Ge. For example. effectively increasing the working day from 8 hours to 24 hours. its detection limit by ICP is more than three orders of magnitude better. An ICP-ES instrument or graphite furnace AA can be left to run overnight as no combustible gases are involved. copper can be measured at the 324. Se. While it is possible to determine phosphorous by AAS. Surveys have shown that most analysts are interested in 6–15 elements per sample and choose to pump the sample (which increases washout times) to improve precision and accuracy by minimizing viscosity effects. the rate at which samples may be determined depends on the type of instrument: both simultaneous and sequential ICP spectrometers are available. extrapolation of two point calibrations can be accurately used to achieve orders of magnitude above the top standard. then purged with the inert plasma gas. This compares to a linear dynamic range of typically 103 for AAS. the monochromator must first be evacuated. Linear Dynamic Range The inductively coupled plasma is doughnut shaped (with a “hollow” center).75 nm wavelength from its detection limit of about 0. N and halogens by ICP-ES can only be achieved if a vacuum monochromator.002 ppm to over 200 ppm. with purged transfer optics. providing maximum flexibility of choice of element and analytical wavelength. the cross over point where sequential ICP becomes faster than AAS is approximately 6 elements/sample for routine analysis. The regions refer to those seen when a Yttrium solution is introduced. I. The sample aerosol enters the base of the plasma via the injector tube. K) are generally better by flame AAS than by ICP. Na. A plasma used for emission spectrometry. Non metals such as sulfur. Figure 1. with flame equilibration time. Br) can only be determined by ICP. flame AAS is faster. Cl. The “optical thinness” of the ICP results in little self absorption and is the main reason for the large linear dynamic range of about 105. Sb. To detect the primary aluminium wavelength at 167.000 K can be reached in an argon plasma. program recall and monochromator condition changes. Hg. is used. Detection limits for refractory elements (for example. Unattended Operation Flame AAS cannot be left completely unattended for safety reasons. 2 .08 nm. However. Note that a continuous flow vapor generation accessory can be used with either ICP-ES or AAS for improved detection limits for As. V. In ICP. carbon. Simultaneous ICP spectrometers demonstrate an advantage in analytical speed over sequential ICP spectrometers when more than 6 elements/sample are measured. Detection limits for Group I elements (for example.73 nm by purging the monochromator. If a "one off" sample is presented for a few elements. it is 2900 °C. Optimum detection of non metals such as S. The optics must be purged to exclude atmospheric oxygen and eliminating its absorption. Temperatures as high as 10.Sulfur can be measured at 180. Most ICP spectrometers purchased are the sequential type.

but recalibration is required over a longer period. Spectral The optical requirements of AAS are fairly simple.5 nm resolution with only iron. A refractor scan is then performed over a smaller wavelength region in order to identify and locate the peak position. The chemical bonds that still exist at below 3000 °C are completely ruptured at above 6000 °C. In ICP-ES. Spectral resolutions of 0. This method offers better precision. especially with graphite furnace AA.010 nm or better are required to resolve nearby interfering lines from the atomic and ionic analytical emission signals of interest. inter-element correction may be used to minimize the spectral interference. The high temperatures reached in a plasma eliminate chemical interferences. Ionization The ICP contains a large number of free electrons. Most elements require 0. and an alternative line may not be available. The lamp itself only emits a few spectral lines. ICP tends to be more susceptible to such interference because of the smaller droplet size required and lower transport efficiency. For AAS a precision of 0. but may be minimized with chemical modifiers. One technique used to eliminate backlash in the grating drive mechanism of ICP spectrometers is by scanning and measuring at the same time. This method of measurement can be termed as “measurement on the move” and effectively results in poor short term precision. these effects can be minimized by optimizing the plasma viewing height. be overcome by selecting a different elemental wavelength with similar detection limits. but precisions of 2–5% are not uncommon over an 8 hour period without recalibration. Li or K can be added to both samples and standards to minimize this effect. With ICP-ES the short term precision is typically 0. nickel and cobalt of the common elements requiring 0. An ionization buffer such as Cs. so ionization interferences for most applications are virtually nonexistent. Finally the refractor plate is repositioned “at the peak” where the replicate measurements are then performed. Ionization interferences can be encountered when determining elements in matrices that contain very high concentrations of Group I elements (for example.3–2%.1–1% is typical for the short term. Na & K). the elements and the wavelengths which may be determined are fixed at the time of purchase. Ionization interferences may also be found in AAS. Precision Precision can be termed short term (or within-run) and long term (over a period of one day).Interferences Chemical Chemical interferences are relatively common in AA. 3 . Spectral interference in sequential ICP spectrometers can. such as. A more recent method drives the grating to a wavelength near the analytical peak. The monochromator only needs to distinguish a spectral line emitted from the hollow cathode lamp from other nearby lines. ICP-ES is almost free from chemical interferences. when measuring certain Group II elements in a nitrous oxide flame. With simultaneous ICP spectrometers. However. the rich spectra present in the plasma means that there is a greater possibility of spectral interference. In this case.2 nm or better. in most cases. Physical These interferences relate to the different properties of various samples and can affect sample transport and droplet formation. which accounts (for the most part) for the better detection limits achieved for refractory elements.

271 381.3 5 1.716 455. Refractories.0 240.5* 9* 150* *Modifier used to obtain these results.595 249.55 0.068 167.075 0.042 223.950 Flame type Air N2O N2O Air N2O N2O N2O Air MSR % 97 100 86 94 70 100 64 88 – 94 87 – – 98 100 58 84 100 100 100 97 80 – 100 – 69 El Ag Al As Au B Ba Be Bi Br C Ca Cd Ce Cl Co Cr Cs Cu Dy Er Eu Fe Ga Gd Ge Hf Hg 422. **20 µL injection ***The Characteristic Masses listed were determined in aqueous solution using maximum heating rate in argon with zero gas flow during atomization.7 357.3 – 2% Less than 5% 103 0.06 0.8 553.45 0.22 43 0.5 9 Element Silver Aluminium Arsenic Gold Boron Barium Beryllium Bismuth Bromine Carbon Calcium Cadmium Cerium Chlorine Cobalt Chromium Cesium Copper Dysprosium Erbium Europium Iron Gallium Gadolinium Germanium Hafnium Mercury Ag Al As Au B Ba Be Bi Br C Ca Cd Ce Cl Co Cr Cs Cu Dy Er Eu Fe Ga Gd Ge Hf Hg AA λ (nm) 328.5 0.7 421.7 10 10 100000 50 50 20 30 600 500 300 50 800 20000 1000 10000 1500 1 2 100000 5 6 4 3 30 50 1.773 455.6 234.5 6.5 13 4 8.3 1.754 353. K Volatile elements Pb.7 228.5 6 100 2000 200 2000 200 N2O Air N2O Air N2O Air Air N2O N2O N2O Air Air N2O N2O N2O Air 0. S Halogens Sample throughput Linear dynamic range Precision Short term Long term (over 8 hrs) Interferences Spectral Chemical Ionization Operating costs Combustible gases Table 1.03 0.1 ICP λ (nm) 328.2 12 6000 65 0.6 0.8 520.2 1.4 368.5 – 5% Many Virtually none Minimal High No Virtually none Some Some Low Yes Minimal Many Minimal Relatively high No Guide to ICP/AAS Analytical Values ICP Detection limit µg/L 3 1.8 459.7 242. Al V.403 313.1 307.1 324.AAS v ICP – A quick guide ICP-OES Detection limits Best for : Refractories Non metals P. Ti Flame AAS Best for : Group I metals Na.340 247.U.940 417.03 1.4 265.366 228.2 4.1 – 1% 102 0.5 12 5. Ba.665 228.035 0.4 855* 17 0. S. Zn Rare Earths Furnace AAS Best for : All elements except : B.5 2.967 259.081 188.4 248.8 249.3 0.3 2.985 267.25 0.061 163.01 0.3 193.856 393.7 0.616 267.9 852.5 11 6 45 100 25 1.3 294.3 253.531 324.660 725.2* 0.85 0.1 309.206 342.5 7. for example P.45 7. 4 .5 200000 5 4 3200 2 0.025 0.9 223.247 265.5 Flame AA Characteristic Detection conc limit µg/L µg/L 30 800 500 100 8000 200 15 200 2 30 300 10 500 20 1 50 Zeeman Furnace AA Characteristic*** conc** Mass µg/L pg 0.5 1.23 0. Best if more than Best if less than Slow (typically 6 elements/sample 6 elements/sample 4 mins/element) 105 0.170 337.07 0.W.5 4.5 0.118 264.3 0.802 418.141 184. B.21 0.2 400.7 5 10* 4.

35 6.0 1000 Flame type N2O Air Air Air N2O Air N2O Air Air N2O Air N2O N2O Air N2O N2O Air Air N2O Air Air N2O Air Air Air N2O N2O N2O N2O N2O N2O N2O N2O Air N2O Air N2O N2O N2O N2O N2O N2O Air N2O Zeeman Furnace AA Characteristic*** conc** Mass µg/L pg MSR % – El Ho I In Ir K La Li Lu Mg Mn Mo N Na Nb Nd Ni Os P Pb Pd Pr Pt Rb Re Rh Ru S Sb Sc Se Si Sm Sn Sr Ta Tb Te Th Ti Tl Tm U V W Y Yb Zn Zr 0.9 780.0 285.02 0.3 2 20 0.6 16 1.8 20 35 11 5 5.2 196.771 268.517 350.9 492.9 1.6 391.5 255.5 313. Element Holmium Iodine Indium Iridium Potassium Lanthanum Guide to ICP/AAS Analytical Values (continued) ICP Detection AA ICP limit λ (nm) λ (nm) µg/L Ho I In Ir K La Li Lu Mg Mn Mo N Na Nb Nd Ni Os P Pb Pd Pr Pt Rb Re Rh Ru S Sb Sc Se Si Sm Sn Sr Ta Tb Te Th Ti Tl Tm U V W Y Yb Zn Zr 410.937 213.75 0. 5 .5 20 18 0.05 0.542 279.1 100 97 90 – 49 – 75 92 96 92 – – 98 – 69 92 100 – 82 90 – 95 100 – 96 – 92 100 – 93 94 – 90 93 – 100 63 – – 79 – – 97 92 – Lithium Lutetium Magnesium Manganese Molybdenum Nitrogen Sodium Niobium Neodymium Nickel Osmium Phosphorous Lead Palladium Praseodymium Platinum Rubidium Rhenium Rhodium Ruthenium Sulphur Antimony Scandium Selenium Silicon Samarium Tin Strontium Tantalum Terbium Tellurium Thorium Titanium Thallium Thulium Uranium Vanadium Tungsten Yttrium Ytterbium Zinc Zirconium 0.9 217.35 0.05 0.2 2000 1000 10 100 40000 10 10 10000 100 10 1000 5 100 40 50 500 300 1000 100 2 2000 700 30 100 20 20 40000 100 1000 200 4 1.489 267.2 279.609 224.2 0.434 242.958 309.0 251.716 334.5 8.7 235.5 432.030 174.6 217.1 670.1 0.3 364.75 0.353 340.8 213.Table 1.585 177.9 213.225 231.0* 135 0.5 0.3 0.220 385.945 780.8 358.15 0.856 339.18 0.3 589.0075 3 0.1 410.5 0.924 346.949 407.0 244.6 0.4 0.8 2200* 5.525 343.6 7 0.0 290.490 379.995 309.941 351.5 0.604 225.9 208.5 10 0.5 18 2 17 0.734 217.418 401.5 232.1 0.2 0.9 766.876 180.75 4.384 196.6 429.5 349.917 214.5 5 18 14 7 0. ** 20 µL injection *** The Characteristic Masses listed were determined in aqueous solution using maximum heating rate in argon with zero gas flow during atomization.02 0.4 4 0.3 4 50 000 1 4 2 5.281 274.3 276.5 318.8 0.4 37 5 7 15 0.6 70 1 8 15 10 14* 15 10* 2 3.43 3.1 345.1 265.610 202.600 178.030 328.458 417.0 346.28 0.5 550.7 271.268 766.4 303.5 9* 50 15 1.8 371.8 336.9 360.01 0.709 371.03 0.198 0.02 9 5 27 17 0.24 110 0.2 0.2 398.005 7.5 60 18 3.939 265.581 361.553 257.784 261.311 239.5 Flame AA Characteristic Detection conc limit µg/L µg/L 700 150 800 7 40000 20 7000 3 20 300 3 20000 6000 70 1000 120000 100 50 20000 1000 50 8000 100 400 300 300 1000 1500 6000 700 40 10000 7000 200 1000 200 300 100000 700 5000 2000 60 8 9000 40 40 500 3 2000 2 300 0.499 220.1 22 0.611 442.7 0.5 0.026 251.272 588.8 495.276 325.478 670.45 2.1 343.0 334.5 460.023 227.15 *Modifier used to obtain these results.7 214.

effluents. Can the user afford an automated AAS or ICP-ES. the chemist must define both present and future analytical requirements.com/chem www. where the samples may change in viscosity from sample to sample. that both an ICP-ES and a furnace AAS will be necessary to meet the analytical requirements. performance.Analytical Requirements Before deciding which technique is appropriate. soils.com/chem Agilent shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing. What expertise do the operators have? How much money is available to purchase or lease costs/month? Cost of ownership and running costs. • • • For More Information For more information on our products and services. © Agilent Technologies. visit our Web site at www. For Deuterium Furnace systems. That is: • • • • • • • Number of samples/week? What matrices need to be analyzed? For example. bronzes.agilent. Sometimes the answer is further complicated by the fact that neither flame AAS nor ICP-ES will satisfy all requirements. battery acid analysis. for example. as many do. You may find. or is a simple AAS sufficient? The answers to these questions will help you to decide which is the preferred technique. 1991 Printed in the USA November 1. or use of this material. Inc. descriptions. and specifications in this publication are subject to change without notice. the equivalent Characteristic Concentration and Characteristic Mass is easily calculated using the following conversion: CMn = CMz X MSR (%)/100 CCn = CCz X MSR (%)/100 where: CMn = Characteristic Mass for Deuterium Furnace Systems CMz = Characteristic Mass for Zeeman Furnace Systems (from Table 1) MSR = Magnetic Sensitivity Ratio (as % from Table 1) CCn = CCz = Characteristic Concentration for Deuterium Furnace Systems Characteristic Concentration for Zeeman Furnace Systems (from Table 1).agilent. Information. How many elements need to be determined for each sample type? What are the typical sample volumes? What elements need to be determined? What concentration ranges are present in the matrices? Would an Internal Standard be useful? For example. steels.. 2010 ICPES003 .

Sign up to vote on this title
UsefulNot useful