ICP-MS?

Plasma at the entry of the sampler

Plasma at the entry of the sampler. Image courtesy of:

http://iramis.cea.fr/en/Phocea/Vie_des_labos/Ast/ast_sstechnique.php?id_ast=886

What is ICP-Q-MS?

Inductively coupled plasma mass spectrometry (ICP-MS) is the fastest growing multi-element
trace element analysis technique with more than 8000 ICP-MS instruments installed worldwide
(Thomas 2008), most of which are quadrupole systems. In general, “ICP-MS” without further
description implies a quadrupole-based mass analyzer system (ICP-Q-MS), but there are also
multicollector (ICP-MC-MS) and time-of-flight (ICP-TOF-MS) instruments. ICP-Q-MS is principally
used for rapid, precise and accurate trace (<1000 ppm) element determinations in liquid and
solid samples, but other applications include isotopic determinations and speciation studies. The
power of modern ICP-Q-MS resides in its ability to rapidly measure trace elements at very low
detection limits (to sub parts per trillion levels) as well as minor and major elements (at parts
per million levels) in the same analytical run on suitably diluted samples. The technique was
commercially developed in the early 1980s and because of superior detection capabilities has
superseded many other analytical methods including: atomic absorption, optical emission
spectroscopy, and ICP atomic emission spectroscopy (Wolf, 2005). Major advancements required
for routine elemental analysis by ICP-Q-MS included development of: (1) a stable plasma source
capable of ionizing most elements of the periodic table, (2) an interface system to transfer ions
from the corrosive high temperature (6000-10,000°K- atmospheric pressure (760 torr)
environment of the Ar plasma into the mass spectrometer analyzer region at room temperature
and low pressure (10-5 torr), and (3) technology to remove interferences, notably
collision/reaction cells (CRCs).

Figure 1. Inductively coupled plasma mass spectrometers (ICP-MSs), such as the Agilent 7500ce
shown here, can rapidly, accurately, and precisely measure elemental concentrations for most of
the periodic table in liquid and solid samples.

Figure 1. Inductively coupled plasma mass spectrometers (ICP-MSs), such as the Agilent 7500ce
shown here, can rapidly, accurately, and precisely measure elemental concentrations for most of
the periodic table in liquid and solid samples.

Several integrated hardware components (most hidden from direct view) comprise a modern
ICP-Q-MS instrument as shown cryptically above (Fig 1) and schematically below (Fig 2):
Sample introduction system

Excitation source for ionization (ICP)

Ion transport system operating from atmospheric pressure to 10-6 torr (Interface/Ion Lens
Optics)

Collision-reaction cell

Mass separation device (quadrupole)

Ion detector

Figure 2. Schematic view of typical hardware components of an ICP-Q-MS for solution mode
sample introduction. For analysis of solids, a laser ablation system (not shown) replaces the
roles of the autosampler, peristaltic pump, nebulizer, and spray chamber.

Figure 2. Schematic view of typical hardware components of an ICP-Q-MS for solution mode
sample introduction. For analysis of solids, a laser ablation system (not shown) replaces the roles
of the autosampler, peristaltic pump, nebulizer, and spray chamber.

Fundamental principles of ICP-Q-MS

An ICP-Q-MS uses an inductively coupled Ar plasma (described below) as an excitation source to

ionize the sample and a quadrupole mass spectrometer as an analyzer to separate and
selectively transmit sample (analyte) ions of a single mass-to-charge ratio (m/z) to the detector.
During this process, sample ions rapidly undergo large temperature (6000K-to-room temp) and
pressure (760 to 10-6 torr) reductions. In essence, ions are systematically transferred from the
plasma to the detector in a highly controlled electrostatic field within a dynamically increasing
vacuum. A computer controls all hardware components, including peripherals involved in
automated sample introduction, as well as data acquisition, storage, display, and processing.

Sample Introduction: The sample is introduced as either a liquid, typically via a peristaltic pump,
nebulizer, and spray chamber, or as a solid particulate aerosol, as generated by a separate laser
ablation system. For liquids, the nebulizer converts the sample into a fine aerosol mist using a
stream of Ar gas. The finer aerosol droplets, optimal for efficient ionization in the ICP, are
separated from larger droplets in an enclosed spray chamber, a process that also dampens
sample flow variations introduced by the peristaltic pump. For in situ sampling of solids using a
laser ablation system (not pictured above), a high energy pulsed photon source interacts (or
couples) with the sample in an ablation cell (chamber), ideally breaking atomic bonds at the
sample surface before heating occurs. The laser induced stress and resulting photomechanical
fracturing ejects fine solid particulates into the ablation plume, which are transported through
the ablation cell by a carrier gas (usually He or Ar) and mix with Ar carrier gas flowing to the ICP
torch. Regardless of the sample introduction system used, the aerosol produced (liquid or dry
particulate) similarly flows through the central injector tube of the ICP torch.

Ionization Source: The inductively coupled plasma (ICP) is a highly ionized gas generated by the
interaction of a strong magnetic field with Ar gas flowing through a quartz glass torch. The
magnetic field is generated by 27-40 MHz radio frequency transmitted from a (water- or Ar-)
cooled copper induction coil that encircles the distal (open) end of the torch. When a high
voltage spark is applied, the flowing Ar gas ionizes to form a high temperature (~6000-10,000° K)
plasma that becomes self-sustaining. The ICP torch typically consists of three nested concentric
tubes, through which Ar flows at different rates. The outermost tube (Main Plasma Gas: Ar flow
~15 L/min) provides most of the plasma-forming Ar and prevents direct contact of the plasma
with the torch walls. The medial tube (Auxiliary Gas: Ar flow ~ 0.75 L/min) controls the spacing
of the plasma relative to the open (front) end of the torch. The central tube (Nebulizer Gas: Ar
flow ~ 1 L/min) transfers the sample aerosol into the central region of the plasma (6500°K),
where it is rapidly (virtually instantaneously; ~2 ms, Houk 1986) dried, dissociated, atomized and
ionized.

Interface Region: The sample ions next pass through the plasma-vacuum interface region
situated between two conical disks (cones), usually made of Ni or Pt, each containing a small (≤
1mm) central orifice; the outer “sampler” cone with a larger (~2.5x) orifice compared to the
inner “skimmer” cone. A pressure reduction (to 1-3 torr) in the interface region between the
cones, maintained with a mechanical roughing pump, causes the incoming ion beam to
dramatically expand behind the sampler cone such that the orifice of the second skimmer cone
consistently passes on an ion distribution that is representative of the central analyte-rich
portion of the plasma. The interface vacuum also serves to remove most Ar atoms at this point.

Ion Lens Optics & Collision/Reaction Cell: The sampled ions exiting the interface region next pass
through ion lens optics, which consist of one or more positively charged (typically stainless steel)
lenses. Differential voltages applied to the lens components achieve two purposes with minimal
mass discrimination: (1) separation of positively charged ions from undesired neutral species and
photons, and (2) focusing and collimation (narrowing) of the ion beam. The resulting positively
charged ion stream then passes through the collision-reaction cell, which may be used to
remove or highly attenuate polyatomic interferences (CRC, described separately below), to the
quadrupole mass analyzer. The action of one or more turbomolecular pumps achieves further
pressure reductions of ~10-3 – 10-4 torr for the ion lens optics and ~10-5 – 10-6 torr for the CRC
and quadrupole mass analyzer.

Quadrupole: The quadrupole mass analyzer (described in more detail below) filters out non-
analyte, matrix, and interfering ions, allowing only desired analyte ions of a single mass-to-
charge ratio (m/z) to be transmitted to the detector. As its name conveys, a quadrupole consists
of four parallel (conductive) rods, often made of gold-coated ceramic or molybdenum.
Radiofrequency (RF) and DC potentials applied to each set of opposing rods produce a dynamic
(hyperbolic) electrostatic field within the space between the rods. The electrostatic field, which
controls the trajectories of ions entering the quadrupole as a function of their inertial mass and
charge, can be rapidly changed and accurately optimized in order to allow only analyte ions of a
specified m/z to pass entirely through the quadrupole and reach the detector. By changing the
applied RF/DC voltages, the quadrupole can either scan iteratively across the entire amu range
(continuous scanning) or jump selectively from target m/z to target m/z (peak hopping) in order
to measure elemental ion abundances in a sample.

Ion Detector: The detector (usually an electron multiplier) receives and amplifies an ion signal
that is proportional to concentration, as determined through calibration with one or more
reference materials of known composition. Dual-stage discrete dynode electron multipliers are
often employed because of their effective wide dynamic range. Positive ions passing from the
quadrupole are attracted to the first negative dynode of the detector, whereupon impact
secondary electrons are emitted from the dynode surface. These are then attracted to the
positively charged second dynode, where more electrons are emitted. The electron emission
process continues in this manner throughout the entire cascade of individual dynodes (usually
12-24), resulting in a geometrically amplified signal. Integrated circuitry for this detector,
involving simultaneous measurement of signal intensity at mid- and final-dynode positions, can
obtain a wide linear dynamic detection range spanning eight orders of magnitude, while also
protecting the detector by excluding extreme high intensity signals. Low intensity signals (< 2M
cps) are measured as counts per second at the lower dynode position (pulse mode). High
intensity signals are measured from the final dynode as a voltages (analog mode), which are
subsequently normalized to cps equivalents from pulse-to-analog conversion factors (previously
optimized during tuning of the instrument). The dual-stage functionality of this detector
achieves a linear detection range extending from ppt to 100s of ppm within a single multi-
element analysis, commonly obviating the need for additional dilution.

How does ICP-Q-MS work?

I. Making Singly Charged (+1) Ions

Energy achieved by the Ar ICP is sufficient to cause the majority of sample atoms passing
through it to exceed their first, but not second, ionization potentials (Fig 3). The significance of
this is two-fold: (1) elements from most of the periodic table will be ionized to a +1 state, and (2)
because the ions generated will principally differ by mass, not charge, they can be focused and
separated on the basis of their inertial masses within an electrostatic field.

Figure 3. Energetics of progressive ionization of 56Fe. The Ar plasma temperature is sufficient to

convert the vast majority of iron atoms into singly charged species (remove a single outer shell
electron), with negligible production of doubly charged species (remove two outer shell
electrons).

Figure 3. Energetics of progressive ionization of 56Fe. The Ar plasma temperature is sufficient to

convert the vast majority of iron atoms into singly charged species (remove a single outer shell
electron), with negligible production of doubly charged species (remove two outer shell
electrons).

It is important to realize that although most elements are substantially ionized in the high
temperature Ar plasma (Fig 4, Houk 1986; Douglas and Tanner, 1998), some elements having
first ionization potentials approaching or exceeding that of Ar (15.76 eV) can have much lower
ionization efficiencies. Quantitative ion measurements require both consistent +1 ionization and
adequate sensitivity. Arsenic (9.81 eV), selenium (9.75 eV), and phosphorous (10.49 eV), for
example, have fairly low first ionization efficiencies on the order of 52, 33 and 33%, respectively,
which are still sufficient for robust quantitative determinations. By comparison, sulfur (10.36 eV)
and fluorine (17.42 eV) have respective first ionization efficiencies of only 14% and 0.00091%, for
which quantitative determinations are impractical or impossible. With well-optimized plasma
conditions, doubly charged ions for most elements occur in negligible amounts, with worst case
maxima typically on the order of 1-3% for elements having lowest second ionization potentials
(e.g., Ba2+).

Figure 4. Elemental first ionization efficiencies (as percents) calculated for a Tion of 7500K and
electron density (ne) of 1 x 1015/cm3 (modified after Houk 1986).

Figure 4. Elemental first ionization efficiencies (as percents) calculated for a Tion of 7500°K and
electron density (ne) of 1 x 1015/cm3 (modified after Houk 1986).

II. Filtering Ions According to Mass-to-Charge (m/z) Ratio

ICP-Q-MS differs from other ICP-MS systems due to the employment of a quadrupole mass filter,
through which the ion beam is directed before reaching the detector. It is within the dynamic
electrostatic field of the quadrupole that ion flight stability as a function of inertial mass and
charge is exploited to selectively transfer analyte ions of a single m/z to the ion detection
system. The essential physical processes at work are electrostatic attraction/repulsion and
inertia, with flight paths of “smaller” (less massive) ions more sensitive to electrostatic variations
compared to “bigger” (more massive) ions. Within the quadupole electrostatic field, ion paths
become spiral. Unstable flight paths result in ion collisions with one of the quadrupole rods,
whereupon they are removed by neutralization. Ion paths either adjust sufficiently (maintain
sufficiently narrow spirals) to pass through the length of the quadrupole, or obtain terminal
collision trajectories (spiral paths too wide). The electrostatic field is highly controlled so that the
m/z filtering process obtains a resolution less than 1 amu (typically ~0.7 amu). The quadrupole
mass filtering process is extremely fast. For example, the quadrupole can scan over the 1-200
amu range in ≤ 20ms, or switch repeatedly from peak to peak in jumps of 1.8-25 µs (Houk 1986,
Steel and Henchman 1998). Employing signal averaging techniques, the entire amu range (2-260
amu) can be scanned 100 times in about 3 minutes to obtain a robust full spectral scan.

Quadrupole Mass Filter In More Detail

A quadrupole consists of four parallel poles, or rods, two that obtain a net negative charge and
two that obtain a net positive charge (Figs 5-6). The polarity and strength of the electromagnetic
field achieved by each set of opposing rods results from the simultaneous application of DC and
RF (AC) voltages, which can be increased or decreased, but are always maintained in a fixed ratio
(typically on the order of 1:6). The applied DC voltages are constant but of opposite polarity –
one rod set positive, the other negative. Meanwhile, the applied RF current alternates between
positive and negative polarity in the range of 2-3 MHz, with the polarity alternations maintained
exactly out-of-phase between the opposing rod sets (Fig 6). In other words, while one set of
opposing rods has an applied positive RF potential, the other set has an equally applied but
negative RF potential. It is when the magnitude of the RF polarity field strength exceeds that of
the oppositely applied DC polarity field strength that a given set of opposing rods obtains its
effective net polarity, which then affects ion paths as a function of their inertial masses. Ion
trajectories are ultimately governed by both sets of rods, but are more easily understood in
terms of the local electrostatic field between a given rod set. In this simplification, one set acts
as a high-pass filter (high masses stable through the quad). The other acts as a low-pass filter
(low masses stable through the quad).

Figure 5 (Left). View of front end of a quadrupole mass analyzer removed from an Agilent 7500ce
ICP-Q-MS showing hyperbolic cross section Mo rods. Black wires on the external housing supply
opposing rod sets with applied RF and DC potentials. Rods are ~20 cm in length. Figure 6 (Right).
From Steel and Henchman (1998): Upper diagram shows basic configuration of a quadrupole and
possible (spiral) ion flight paths for a given RF/DC. Lower diagram shows how the voltage
potentials applied to opposing rod sets vary through one RF cycle. U is the unchanging DC
potential; V is the variable RF (AC) potential.

Figure 5 (Left). View of front end of a quadrupole mass analyzer removed from an Agilent 7500ce
ICP-Q-MS showing hyperbolic cross section Mo rods. Black wires on the external housing supply
opposing rod sets with applied RF and DC potentials. Rods are ~20 cm in length. Figure 6 (Right).
From Steel and Henchman (1998): Upper diagram shows basic configuration of a quadrupole and
possible (spiral) ion flight paths for a given RF/DC. Lower diagram shows how the voltage
potentials applied to opposing rod sets vary through one RF cycle. U is the unchanging DC
potential; V is the variable RF (AC) potential.

High Pass Condition: When opposing rods obtain an overriding positive polarity field (Fig 6, rods
2 and 4 above), a massive ion passes through the entire length of quadrupole, whereas a less
massive ion is sufficiently attracted by the negative pulses of the applied RF such that its
trajectory ultimately collides with a rod.

Low Pass Condition: When the opposing rods obtain an overriding negative polarity field (Fig 6,
rods 1 and 3 above) a large ion is initially accelerated toward a negative pole and subsequent
positive pulses of the applied RF current are insufficient to thwart an eventual
collision/neutralization trajectory. Simultaneously, smaller ions are sufficiently attracted by the
alternating positive and negative RF polarity pulses to maintain a stable path through the length
of the quadrupole.

Because the two sets of rods operate simultaneously and the RF component on each set is out-
of-phase, the resulting magnetic field changes dynamically with each successive RF cycle such
that ion trajectories actually spiral as they move through the quadrupole. The opposing rods also
simultaneously act as a high and low pass filter with respect to a single m/z, as specified by the
imposed DC/RF current. The number of RF cycles affecting ion trajectories depends on the
length of the quadrupole rods, the frequency of AC, and the mass (velocity) of the ion. For
example, a massive ion (e.g., 200 amu) with flight time of 25 µs through an 11-cm long
quadrupole, experiences about 50 RF cycles (Steel and Henchman, 1998). The specific ratio of DC
to RF current is optimized, during routine (daily) calibration of the instrument, to achieve
resolution below 1 amu (usually ~0.7 amu) while centered at peak ion sensitivity throughout the
entire amu range (2-260).

III. Removal of Interferences

Mass interferences on a given mass-to-charge-ratio (m/z) are possible due to the presence of
isobars (e.g. 204Hg, 204Pb), polyatomic/molecular species (e.g., 40Ar16O vs 56Fe, 40Ar:40Ar vs.
80Se) formed by various recombinations of sample, matrix and Ar ions in cooler parts of the
plasma, and doubly charged species (e.g., 138Ba2+ vs. 69Ga+), also formed in the plasma. A key
reason why modern ICP-Q-MS is so effective for quantitative elemental analysis is that these
interferences can now be eliminated or substantially reduced by a combination of optimizing
plasma characteristics, selection of alternative uninterfered isotopes, use of interference
equations, and/or use of collision-reaction cell (CRC) technology.

CRC technology is an especially significant advancement whereby polyatomic interferences are

reduced or eliminated by passing the ion beam, prior to reaching the quadrupole mass filter,
through a cell that can be pressurized with either an inert collision gas (e.g. He) or a reactive gas
(e.g., H2, NH3, O2). The polyatomic interferences are attenuated or eliminated by kinetic energy
discrimination (larger polyatomics undergo more energy dampening collisions than analyte ions,
lose energy, and thus do not pass through the cell) or interact with a reactive gas to form species
with m/z different from that of the analyte ions. Multiple modes, using collision, reaction, or
neither (no gas), can be used during an analytical session to achieve virtually interference free
results on analytes throughout most of the periodic table.

ICP-Q-MS Applications

The main purpose of Q-ICP-MS is to determine high quality (precise and accurate) elemental
concentrations of various materials. Owing to its high sample throughput capabilities and wide
range of detection and elemental coverage, ICP-Q-MS is routinely used for the elemental analysis
of geological, environmental, chemical, biological, industrial, semi-conductor, metallurgical,
petrochemical, medical, forensic, nuclear, and archaeological samples.

A few examples of materials analyzed by category include:

Geological – trace elements in rocks, mineral separates, sediments, soils, soil leachates, fossils

Environmental – heavy metals/toxins in surface and ground waters (fresh to saline), sediment
extracts, waste, fish bone and tissue, bivalve carbonate/tissue, vegetation digests

Medical – trace metal contents in bone/tissue/food/urine/blood

Biological – metal contents of proteins/enzymes

Archaeological – trace elements in bone, ceramic, or other artifacts

Strengths and Weaknesses of ICP-Q-MS

Strengths

Wide elemental coverage – most elements with 1st ionization potentials below that of Ar can be
analyzed by ICP-Q-MS

High sensitivity

Wide analytical working range for solution mode – up to nine (9) orders of magnitude in a single
acquisition; from single ppt to 100s of ppm

Fast quantitative and semi-quantitative analysis times – with high speed scanning quadrupole
analyzer, a full elemental suite in a sample can be analyzed in about four (4) minutes

Good precision and accuracy (~1-3% for solution mode; ~1-10% for laser mode)

Good control of interferences (isobaric, polyatomic, doubly-charged species) through advances

in plasma optimization, ion focusing, and collision-reaction cell technology

Liquid and solid sample introduction systems

Relatively small sample volumes required (typically 1-3 mL for solution mode)

Capable of isotope ratio measurement

Excellent chromatographic detector – to determine chemical form (speciation) in a sample

Limitations

Limited total dissolved solids (TDS) tolerance, which can affect instrument stability if deposits
accumulate on nebulizer, cones, or lens components, as well as ionization consistency (matrix
effects) in the plasma compared to lower TDS samples/standards

. . . . . . . . . . However, nebullizer type (Babington) and high matrix introduction systems can
greatly extend tolerance.

Impractical to generate high precision (low RSD) data for brief transient peaks (e.g., vs. ICP-TOF-
MS).

Impractical to generate high precision isotope ratios (e.g., vs. TIMS or ICP-MC-MS) due to: (1)
“flicker” noise associated with sample introduction, including fluctuations associated with the
peristaltic pump, nebulizer, and plasma, and “shot noise” associated with undesirable photons,
electrons, and ions hitting the detector; and (2) the fact that a scanning quadrupole cannot
simultaneously measure internal standard and all analyte isotopes for optimal drift
compensation. The latter limits precision compared to multicollector systems (by a factor of
~10).

Analysis speed and sample throughput can be limited by quadupole scanning rates (a
consideration mainly for precision of isotope ratio measurements or measurement of very fast
transient signals) and wash-out time required for rinsing between samples and introduction of
new samples

Spectral resolution (0.7-1.0 amu) for ICP-Q-MS systems are inadequate for discrete
measurement of elements prone to Ar-, solvent- and/or sample-based spectral
interferences . . . . . . . . . However, collision-reaction cell systems render this problem as minor
for trace element determinations.

Destructive technique

ICP-Q-MS Users Guide – Sample Collection and Preparation

No Particles: Samples for solution mode ICP-Q-MS should be particulate free in order to prevent
clogging of sample tubing and the nebulizer capillaries (particularly for concentric and micro-
concentric types). Samples containing solids (even nanoparticles) will not ionize the same way as
liquid samples, and thus can negatively affect quantitative analysis. Particles may be removed by
filtration or by centrifugation and decanting.

Dilute Samples in 1-2% Nitric Acid: Sample solutions should ideally share a matrix similar to
calibration standards. Commercially available single and multielement standards often have
matrices comprised of 1-5% nitric acid. Acidification prevents dissolved species from readily
sorbing onto tubing and container walls. Nitric acid is preferred because it generates fewer
polyatomic interferences compared to other acids.

Dilute Samples to <<< 0.2 wt% (2000 ppm; preferably ≤200 ppm) Total Dissolved Solids: It is
important to limit total dissolved sample contents in order to minimize matrix effects that can
affect the consistency of ionization in the plasma as well as focusing and collimation of the ion
beam beyond the interface region of the ICP-Q-MS. Dilution is also important to ensure that
measurement conditions remain consistent throughout the analytical session. Analysis of high
TDS samples can rapidly deposit coatings onto sampler and skimmer cones that change orifice
shapes and flow dynamics. Such deposits can also constitute a contamination source if partially
ionized during subsequent analyses.

Plan Quality Control in Advance: Quality control statistics typically derive from analyses of
method blanks, matrix spikes, and certified reference materials of known concentration. Method
(or procedural) blanks, sample processed identically to real samples except that no sample is
actually added, provide a measure of all processing-related contamination (airborne,
reagents/acids, labware, personal) that could potentially contribute to measured concentrations.
Matrix spikes (discussed further below) provide a means of evaluating the extent to which
quantitative results may suffer as a result of matrix differences between samples and calibrations
standards. Quality control standards are often analyzed periodically throughout a run of
unknowns. Accuracy and precision obtainable by the data acquisition method can be evaluated
from recoveries and standard deviations relative to certified concentrations. Reference standards
should be independently sourced from stocks used for calibration standards.

ICP-Q-MS Data Collection, Results and Presentation

ICP-Q-MS raw data, in the form of counts per second (cps) may be collected by continuous
scanning of the quadrupole over all or portions of the AMU range, or by discrete peak-hopping
for each m/z of interest. Continuous scanning mode produces a spectra of signal intensity (y)
versus mass (x) that is may be used to evaluate signal peak shapes (to evaluate abundance
sensitivity) or for qualitative identification of elements or interferences in a sample. Elemental
identities are confirmed when all measured isotopes of an element have relative abundances
matching natural isotopic abundances of an element. Peak-hopping mode is used for
quantitative and semi-quantitative measurements of a sequence of pre-selected analyte
isotopes, and involves ion signal measurement at each m/z peak position over a dwell time
sufficient to obtain robust (low RSD) counting statistics (often 10 to 100 ms).

Detection Limits: Limits of detection have various operational definitions, which are particularly
mandated for purposes of quality assurance for data acquired for regulatory agencies.
Calculations are typically based on the level of statistical variance (standard deviation) obtained
on a population (n) of replicate measurements of the analytical blank, assuming a gaussian
distribution and specified uncertainty level (e.g., 95% confidence interval). The student t-test
provides a corresponding factor for multiplying statistical variance and deriving a quantitative
estimate of the minimal detectable concentration (e.g., at the 95% confidence level) above
background. The t-factor is close to 2.3 for a population of 10 replicates, and often detection
limits are conservatively assigned as 3 x stdev.

External Calibration: Quantitative elemental determinations by ICP-Q-MS typically involve the

conversion of signal intensities (cps) into concentration units based on extrapolations from
measurements of an analytical blank and one or more calibration standards of known
concentration that are obtained in the same analytical run from identical operating parameters.
The blank and external standards typically define highly linear (r ≥ 0.9999) regressions for each
analyte, thereby relating signal intensity to concentration. The extrapolation process is highly
accurate, provided matrix differences between unknowns and calibration standards are
insufficient to variably affect ionization efficiency in the plasma and subsequent extraction of the
ion beam. The conversion from cps intensities to concentrations, extrapolated from regressions
on calibration standards, is typically performed automatically by data acquisition software,
usually after drift compensation (see below).

Accuracy: Accuracy is usually evaluated from analysis of independently sourced quality control
standards of known concentration, in matrices similar to the unknowns and calibration
standards. Accuracies (and precisions) for most elements are typically on the order of 1-3% for
solution mode and 1-10% for laser ablation.

Precision: Precision, or degree of analytical variability, is estimated by variance (standard

deviation) associated with replicate analyses of standards and unknowns. in-run precision
corresponds to consecutive replicates. Intra-run precision corresponds to variance associated
with the population of all replicates interspersed within an analytical run. Extra-run precision
corresponds to variance for replicate populations in different analytical runs. Often some level of
precision is built into analytical methods for data acquisition. For example the quadrupole may
acquire data for multiple dwell time intervals, then report the mean and standard deviation of
the measured intervals. The latter, once converted to concentration equivalents, provides a
measure of quantitative uncertainty associated with the sample analysis.

Matrix Effects: The possibility of non-quantitative matrix effects for unknowns can be evaluated
from measurement of unknowns relative to an aliquot that has been spiked with known
concentrations of analytes. Matrix effects are most likely for high TDS samples, so if spiked
analyte recoveries are close to 100% on highest concentration unknowns, matrix effects are
unlikely for lower concentration unknowns. Non-quantitative spike recoveries indicate matrix
effects, which may be reduced to tolerable levels by further dilution (to better matrix match
calibration standards) or by using the method of standard additions for calibration.

Standard Addition: When unknown matrix incompatibilities cannot be reduced by dilution, for
example when analyte concentrations occur at very low concentrations in complex samples (e.g.,
trace metals in digested tissue or brines), aliquots of unknowns can be spiked with known
amounts of the analyte(s) of interest and measured relative to the unspiked unknown. The small
spike additions have negligible effects on the original sample matrix so measurement of analyte
signals occurs with the same overall efficiency. The resulting measured intensities define a linear
relationship corresponding to the concentration of spike additions that extrapolates to a
negative concentration for the unspiked unknown (“blank”). The absolute value of this
concentration corresponds to the analyte concentration in the unspiked unknown.

Isotope Dilution: Isotope dilution, wherein a sample is spiked with an artificially enriched isotope
solution of known composition and then all analyte isotope are measured, provides another
means for quantitative extrapolation of concentration. By this approach, spike introduced
changes to the expected natural abundance distribution are used to correct for analyte
concentration in the unspiked sample.

Drift Compensation: Modern ICP-Q-MS systems drift slowly during extended analytical runs as
integrated components associated with interface and ion lens system change in terms of their
geometries (deposits on cones) and electrical properties (lenses, quadrupole), often lowering
signal intensities. These temporal variations require correction in order to ensure high quality
quantitative data throughout the analytical session. Drift correction is most often achieved by
adding single concentration internal standard analytes to all samples and standards prior to
measurement, often by continous online mixing of an internal standard solution during sample
introduction. Internal standards are ideally uninterfered isotopic species that do not occur in
samples or standards. They are selected to span the measured portion of the AMU spectrum
and have first ionization potentials comparable to the analyte elements. Some commonly
employed internal standard isotopes include 9Be, 45Sc, 89Y, 103Rh, 115In and 209Bi. Sample
analyte intensities are normalized according to the factor necessary to restore an assigned
internal standard analyte intensity to its initial value at the start of the analytical run (e.g., during
measurement of the calibration blank). Data acquisition software typically performs drift
corrections automatically prior to reporting intensities and corresponding concentrations.

Data Filtering: Corrected analyte concentrations need to be evaluated relative to detection limits
and the range of the calibration standards. Concentrations below detection limits can only be
reported as such. Measured concentrations should ideally be bracketed by calibration standards
and those exceeding the high concentration linear range may not be quantitative.

Compensation for Dilution Factor: Samples typically need some level of dilution to reduce TDS
levels appropriately and bring analyte concentration into the practical measurement range (ppt
to 10’s of ppm). The measured concentrations need to be multiplied by the appropriate dilution
factor to obtain analyte concentrations in the undiluted sample.
Literature

Douglas, D.J. and Tanner, S.D., 1998. Fundamental considerations in ICPMS. In: Montaser A. ed.:
Inductively coupled plasma mass spectrometry. Wiley-WCH, New York, 615-679.

Hongsen Niu, R.S. Houk, 1996. Fundamental aspects of ion extraction in inductively coupled
plasma mass spectrometry, Spectrochimica Acta Part B: Atomic Spectroscopy 51, 779-815.

Houk, R.S., 1986. Mass spectrometry of inductively coupled plasmas. Analytical Chemistry 58,
95-105A.

Ma, FM and Taylor, S., 1996. Simulation of ion trajectories through the mass filter of a
quadrupole mass spectrometer. IEE Proc.-Sci. Meas. Technol. 143, 71-76.

Steel, C. and Henchman, M., 1998. Understanding the Quadrupole Mass Filter through
Computer Simulation. Journal of Chemical Education 75, 1049-1054.

Thomas, R., 2008. Practical guide to ICP-MS – a tutorial for beginners (2nd Ed) CRC Press, Boca
Raton, Florida. 347p.

Wolf, R., 2005. http://minerals.cr.usgs.gov/icpms/intro.html

Related Links

A Beginner’s Guide to ICP-MS – 14 article series by Robert Thomas in Spectroscopy Online

A Beginner’s Guide to ICP-MS

The Sample Introduction System

The Plasma Source

The Interface Region

The Ion Focusing System

The Mass Analyzer

Mass Separation Devices – Double-Focusing Magnetic-Sector Technology

Mass Analyzers – Time-of-Flight Technology

Mass Analyzers – Collision/Reaction Cell Technology

Detectors

Peak Measurement Protocol

A Review of Interferences

Sampling Accessories

Sampling Accessories, Part II