An ICP-MS Expedition Through the Elements – Part 1

Showing results for 
Search instead for 
Did you mean: 

An ICP-MS Expedition Through the Elements – Part 1

Team TFS
Team TFS

The Periodic Table is one of the most familiar creations in the world of chemistry.  Since Dmitry Mendeleev first proposed that different elements could be related together in rows and columns according to similarities in their properties in 1869, the Periodic Table has grown to include various stable elements that were unknown in Mendeleev’s time, as well as a kaleidoscope of exotic short-lived nuclides produced in particle accelerator experiments.

periodictableAs the Periodic Table has evolved, so have techniques for measuring the elements within it. From the early days of atomic absorption spectrometry (AAS) in the 1950s to the present generation of sensitive and fast inductively coupled plasma optical emission and mass spectrometers (ICP-OES and ICP-MS), detection and quantification of metals and semi-metals has expanded to encompass every area of analytical science, from geology and environmental analysis to clinical and life science research.

With this increase in the utilization of elemental analysis, instrumentation has come a demand for ever-higher sensitivity and ever-lower detection limits.  Leading the charge to meet these requirements has been ICP-MS, but the pursuit of these goals has uncovered a number of challenges for this technique, including interference, contamination, and sample-to-sample washout issues.  I have previously discussed these issues in detail in my blog post series ‘How to Improve Your ICP-MS Analysis’ parts 1, 2 and 3.

Building on these earlier articles, in this blog post series, I will take you on a journey of exploration through the Periodic Table, from the perspective of ICP-MS.  I’ll discuss how this technique can be applied to the members of the various Periodic Table groups in terms of what parameters work best for them, what pitfalls you need to watch out for and how you can avoid these problems and successfully achieve good quality elemental analysis in your laboratory.

Our expedition starts with the elements of Group 1 – the alkali metals.


A short history of the alkali metals1

The stable alkali metals (Group 1 of the Periodic Table) consist of lithium (Li), sodium (Na), potassium (K), rubidium (Rb) and caesium (Cs), with the unstable francium (Fr) completing the group.

The first member of the group to be identified was potassium, discovered by the English chemist, Sir Humphrey Davy, in 1807.  Davy extracted the element from caustic potash (potassium hydroxide) using electrolysis - potassium was actually the first metal to be isolated using this technique.  This element derives its name from the Arabic word ‘al-qali’ meaning alkali, with the symbol K coming from the Latin word, ‘kalium.’

It was also Davy who discovered sodium later in 1807, again using electrolysis, but this time of molten sodium hydroxide.  The origin of the name comes from the Latin word for sodium carbonate, ‘natrium.’

Lithium was detected by the Swedish chemist Johan August Arfvedson in 1817, but actually first isolated by William Thomas Brande and Sir Humphrey Davy through the electrolysis of lithium oxide (Li2O) in 1821.  The element derives its name from the Greek word ‘lithos’ meaning ‘stone’ because it was found in the mineral petalite (LiAlSi4O10).

Caesium (US, cesium) was discovered by Robert Bunsen and Gustav Kirchhoff, in Heidelberg, Germany, in 1860 during their analysis of mineral water samples from Durkheim using flame spectroscopy. Their identification was based upon two distinctive bright blue lines in the spectrum of Cs, leading them to coin the name caesium, from the Latin ‘caesius’ meaning ‘heavenly blue.’

It was also Bunsen and Kirchhoff who discovered rubidium in 1861 during their analysis of the mineral lepidolite using flame spectroscopy. The bright red lines in its emission spectrum led them to choose a name derived from the Latin word ‘rubidus’, meaning ‘deep red.’

The last member of Group 1 is the unstable element francium (Fr), discovered by Marguerite Perey at the Curie Institute in Paris in 1939 when she was researching the radioactive decay of actinium-227. This element takes its name from the country where it was discovered, France. Francium's most stable isotope, 223Fr, has a half-life of around 22 minutes. It decays into radium-223 through beta decay or into astatine-219 through alpha decay.

These elements are all easily ionized in an argon ICP and generate large populations of singly charged ions.

Table 1 shows some key properties of these elements with regard to their analysis by ICP-MS.

Element Natural isotopes (abundance, in %) First ionization potential (eV)
Li 6Li (7.6), 7Li (92.4) 5.4
Na 23Na (100) 5.1
K 39K (93.3), 40K (0.01), 41K (6.7) 4.3
Rb 85Rb (72.2), 87Rb (27.8) 4.2
Cs 133Cs (100) 3.9

Table 1.  Key properties for the Group 1 elements for ICP-MS analysis
(preferred isotope for analysis in bold).
Although the Group 1 elements are easily ionized in the ICP-MS plasma, their relative sensitivity varies because despite the fact that the amount of available atoms to ionize per gram of each element decreases with increasing mass of the element (1 g of Li contains more atoms than 1 g of caesium) the effect of space charge repulsion in the region between the sampler and skimmer cone and to an extent behind the skimmer cone reduces the signal transmission of lighter ions relative to heavier ones.  The result is that the sensitivity (in counts per second per ng mL-1) for Li is actually considerably lower than the sensitivity for Cs.  Despite this, parts per trillion detection (ppt or μg L-1) limits can still easily be achieved for Li, mainly because, in almost all analyses, it isn’t affected by interferences.  This brings us neatly to the subject of interferences, together with some other concerns to take into account when analyzing the alkali metals.


Interferences and other considerations for Group 1 elements in ICP-MS.

 As mentioned above, lithium, at masses 6 and 7, is not generally affected by interferences in ICP-MS, mainly because being a light element, it doesn’t suffer from polyatomic species overlap.  The more abundant 7Li is generally selected for analysis, but 6Li has value as a low mass internal standard (enriched 6Li standards are available for this purpose).  Lithium is quite a sticky element and can cause sample-to-sample washout memory effects, so if you encounter samples with higher than expected Li concentrations you may need to increase the rinse time between samples or use mixed acid wash solutions (e.g. 2% (v/v) HNO3 plus 2% (v/v) HCl).  Finally, Li deposited on the interface cone surfaces can increase the background of this element.  If you need to measure Li at ultralow concentrations in low matrix samples, using cool plasma operation (operating the plasma at lower power) dramatically reduces the background enabling sub-ppt detection limits.

Sodium, mono-isotopic at mass 23, can be affected by 7Li16O+ interference in the presence of high amounts of Li, such as are encountered in Li metaborate fusion samples, but in these samples, the Na concentration is usually much higher than the LiO+ interference, so this isn’t generally a problem.  Doubly charged 46Ti and 47Ti isotopes generate interference on mass 23 and mass 23.5 respectively and can cause trouble for Na if the Ti concentration is high.  In this case, with single quadrupole ICP-MS (such as the Thermo Scientific iCAP RQ ICP-MS), mathematical interference correction would be required.  With triple quadrupole ICP-MS (such as the Thermo Scientific iCAP TQ ICP-MS), the selection of oxygen or ammonia collision/reaction cell gas will attenuate Ti2+ while reacting far less (if at all) with Na.  High resolution, magnetic sector (HR-ICP-MS) instruments, such as the Thermo Scientific Element 2, can easily spatially resolve Ti2+ interference from Na.  Sodium washes out well between samples so memory effects are not usually an issue.  However, as Na is ubiquitous in the environment, contamination is one of the biggest problems when measuring it.  For low-level Na determination, rigorous cleaning procedures (as described in this blog post are required.  Like Li, Na also adheres to the surface of the interface cones and elevates the instrument background.  Also like Li, cool plasma operation reduces the Na background to levels that allow sub-ppt detection limits.

Potassium, at masses 39, 40 and 41, lies right in the heart of the mass range of the biggest source of interference in ICP – argon.  Argon has three natural isotopes, 36Ar (0.3%), 38Ar (0.06%) and 40Ar (99.6%) and interference from Ar+ in the plasma is huge, especially for mass 40.  In addition, the presence of water and acids in the plasma (from the samples being analyzed) leads to the formation of ArH+ which interferes with both 39K and 41K.  The intensity of ArH+ is much higher on mass 41 than on mass 39, whereas the abundance of 39K is much higher than 41K, so in practice 39K is the preferred isotope to measure.  In environmental, food/beverage and geological analysis, the presence of ArH+ interference on 39K is not a problem as this interference causes a false blank in the low ppb range whereas the samples contain K in the ppm range (even after sample dilution).  For those sample types, quadrupole instruments can be operated in standard mode (i.e. without the collision cell activated) when analyzing K.  When this element must be quantified at ultra-trace levels, such as in semiconductor analysis applications, cool plasma operation again dramatically reduces the Ar+ and ArH+ signals.  Triple quadrupole systems operated with H2 cell gas also reduce Ar+ and ArH+, enabling low ppt detection limits.  HR-ICP-MS, while able to resolve 39K from 38Ar1H+, also provides enhanced detection of K when its resolving power is combined with cool plasma operation.  Like Na, K is easy to rinse out between samples but it can adhere to the interface cones as well. As with Li and Na, this source of background for K can be alleviated using cool plasma operation.

Rubidium, at masses 85 and 87 is fairly straightforward to measure.  The 87Rb isotope is less abundant and is interfered by 87Sr, so for direct quantitation purposes, 85Rb is preferred.  There is potential for GaO+ and SrH+ interference on both Rb isotopes in the presence of high concentrations of Ga and Sr, but this is generally not a concern.  However, if these interferences are present, use of triple quadrupole ICP-MS with O2 as the collision/reaction cell gas certainly helps alleviate the SrH+ problem (by mass shifting the interference away from the Rb isotopes as SrOH+) problem and may also help with GaO+ interference (I have not tested this approach at the time of writing this article).

If you need to measure Rb isotope ratios, using triple quadrupole ICP-MS with O2 as the cell gas is also an elegant way of resolving the 87Sr interference.  Using O2, Sr can be mass-shifted away from 87Rb as 87Sr16O+.   This approach also allows Sr isotope ratios to be measured free of Rb interference, via the mass-shifted SrO+ signals.  HR-ICP-MS can resolve GaO+ and SrO+ interferences but cannot resolve 87Rb from 87Sr.  The presence of Kr as an impurity in the Ar plasma gas supply can potentially produce KrH+ interference on both Rb isotopes, but this interference is usually negligible.

Caesium, mono-isotopic at mass 133, lies in the mass region of xenon (Xe) which, like krypton, can be present in the argon supply as a trace contaminant.  Xenon has isotopes at mass 132 and 134 but not at 133 so it doesn’t directly interfere with 133Cs.  However, Xe can, like Kr, generate a hydride interference (XeH+) on mass 133, but as Xe contamination in the argon plasma gas is usually very low, this interference can generally be disregarded.  There is the potential for 117Sn16O+ interference to appear on 133Cs but this is only a consideration in samples that are very high in Sn.  In this case, the mathematical correction would be required for single quadrupole instruments.  With triple quadrupole ICP-MS, there is the option to test different reaction gases (e.g. O2, NH3, N2O, etc.) to evaluate if the SnO+ interference can be removed.  As powerful a technique as it is, the resolving power of HR-ICP-MS is not high enough to separate 117Sn16O+ or 116Sn17O+ interference from 133Cs, so mathematical interference correction would be required.

Well, that concludes our first expedition through the elements.  I look forward to you joining me on my next excursion which will take us on an ICP-MS journey through Group 2 of the Periodic Table, the alkaline earth elements.



1 See for further information.


Additional Resources

Take a look at our trace elemental analysis instrumentation page to explore our range of AA, ICP-OES, and ICP-MS instruments.

Visit our Food and BeveragePharmaceutical and Environmental pages to learn about our trace elemental solutions for the applications that are relevant to you.

Subscribe to one of our Community pages to receive informative and useful content by e-mail for the application area most relevant to you!