Environmental Engineering Reference
In-Depth Information
Sample Preconcentration Strategies
be mass bias-corrected, which is achieved by: (a) external
correction by measuring the (accurately known) isotope
ratio of another element, (b) internal correction by mea-
suring two additional Hg isotopes, which were added in a
known ratio to the sample of interest, or (c) bracketing the
unknown sample with a standard of known isotope com-
position. For maximum accuracy, a combination of strate-
gies a and c is most commonly used and highly advisable.
Clearly, environmental samples with low Hg concentra-
tions are still a challenge for measuring precise Hg isotope
ratios. Preconcentration is required for many matrices,
such as air and water, to obtain a suffi cient mass of Hg and
for samples of average concentration but little available
sample mass, such as zooplankton or thin slices of sedi-
ment cores. Two major strategies have emerged to deal with
this challenge. As mentioned earlier, combustion tech-
niques will volatilize Hg from larger sample masses, which
can subsequently be collected on gold traps or inline with
acidic permanganate solution (Biswas et al., 2008; Xie et
al., 2005). Similarly, Hg in large volumes of solutions can
be reduced, purged from solution, and collected as above.
Dissolved gaseous mercury can be purged from aqueous
solutions into permanganate solution, which was acidi-
fi ed with sulfuric acid (Zheng et al., 2007). My colleagues
and I have determined the Hg isotope ratio in arctic snow
by collecting 4 L of snow, which subsequently melted in
a clean room. Hg in solution was reduced by the addition
of SnCl
2
and Hg(0) purged onto gold traps in the fi eld sta-
tion. The traps were conveniently transported back to the
analytical laboratory, where Hg was thermodesorbed into
an acidic permanganate solution, which was then measured
by continuous-fl ow cold-vapor MC-ICP/MS.
EXTERNAL MASS BIAS CORRECTION
This strategy works best if an element of similar atomic
mass and chemical characteristics is available. Typically,
thallium is chosen to correct for Hg isotope mass bias and
a Tl reference material (Standard Reference Material issued
by the National Institute of Standards and Technology,
SRM NIST 997) is available with a certifi ed
205/203
Tl ratio of
2.38714. The possibility of using
208/206
Pb from concurrently
produced PbH
4
for mass bias correction of Hg ratios has also
been explored (Hintelmann and Lu, 2003). The mass bias-
corrected ratios for Hg obtained by using known Tl and Pb
isotope ratios, however, led to signifi cantly different results.
The authors concluded that the hydride generation step
itself is prone to fractionate PbH
4
and therefore is unsuit-
able for correcting measured Hg isotope ratios. Using Tl
isotopes on the other hand is also potentially problematic,
since Tl is chemically quite different from Hg. First, the ion-
ization potential of Hg (10.437 eV) is much higher than that
of Tl (6.108 eV), resulting in less effi cient ionization of Hg in
an argon plasma. Second, Hg is often introduced in gaseous
from (either as Hg(0) after cold-vapor generation or in form
of gaseous species from a GC column), while Tl species are
not volatile. Typically, a dry aerosol of Tl is produced in a
separate desolvation unit (e.g., Aridus [Sonke et al., 2008] or
Apex [Foucher and Hintelmann, 2006]) and subsequently
mixed with the Hg(0)-containing gas stream prior to reach-
ing the plasma. Detailed measurements revealed that the
mass bias factors for Hg and Tl are very similar, but not
identical. Most mass bias corrections apply an exponential
law assuming invariant mass bias factors. Hence, even those
“corrected” Hg isotope ratios are not likely to be completely
accurate, which may explain why measurements by differ-
ent laboratories and even measurements conducted in the
same laboratory, but under different instrumental condi-
tions, lead to varying absolute ratios. This is probably inevi-
table and has resulted in the adoption of the delta notation
to express Hg isotope variations among samples relative to
each other rather than in absolute terms.
Nomenclature of Mercury Isotope Fractionation
With mercury being a new and rapidly emerging fi eld of
isotope ratio determinations, there is a risk of confusion
with regard to terminology. To complicate matters, different
studies chose to express results using a bewildering array
of nomenclatures, making it often very diffi cult to com-
pare data among laboratories or to evaluate the accuracy
and precision of each study. In addition, the measurement
of extremely precise Hg isotope ratios is an exceptionally
challenging exercise and requires strict adoption of quality-
control measures. The following sections try to summarize
the state of the fi eld and to suggest a common, promising
and useful nomenclature.
Mass Bias Correction
In principle, there are two alternatives to report Hg iso-
tope ratio differences: in the form of absolute ratios or
as the deviation from a common standard. Although the
former is the traditional format used by analytical chem-
ists, it is extremely diffi cult if not impossible to determine
correct absolute ratios at the level of accuracy required
for distinguishing natural variations. Hence, geochem-
ists typically use a relative comparison using the delta
notation
DOUBLE SPIKE (INTERNAL MASS BIAS CORRECTION)
This is often the most accurate, but also the most techni-
cally challenging, mass bias-correction strategy. It requires
a total of four isotopes that must be available with the
element of interest. Two are used for the calculation of
the ratio of interest and two are added to the unknown
xxx
Hg in‰. It should be stressed that every
isotope ratio measurement by MC-ICP/MS is subject to
instrumental mass bias and will therefore deliver inac-
curate results. Hence, the initially measured ratios must
δ
