Environmental Engineering Reference
In-Depth Information
samples, such as Hg in air, it is probably the only realistic
approach for sample collection and analysis. However, MC-
ICP/MS instruments are optimized for acquiring steady
and continuous signals, which vary only little in intensity
for the duration of the measurement. The signal resulting
from heating up a gold trap, on the other hand, is transient
in nature—that is, the signal intensity changes continu-
ously with time. Frequently, changes in Hg isotope ratios
during peak evolution were observed during the measure-
ment (Evans et al., 2001). In addition, it is not possible to
determine the internal precision of the measurement for
an individual sample. Evaluation of transient signals is
further complicated by low signal intensities on both tails
of the peak, which leads to large variations in ratios during
the leading and trailing portions of the signal. Eventually,
the authors decided to calculate intensity-weighted isotope
ratios and achieved a precision of
necessary before the next sample (or standard) can be ana-
lyzed. Washing out the entire system with a dilute solution
of nitric acid (delivered through the sample line) for a few
minutes or until the Hg signal drops to approximately 1%
of the intensity of measured samples is normally required.
It should be noted that the most straightforward
approach of directly nebulizing an acidic sample solution is
not advisable. The sensitivity of such sample introduction
is inferior, dropping the limit of detection (LOD) by at least
a factor of 10-100. Even more disconcerting are the severe
memory effects expected when introducing high concen-
trations of ionic Hg into conventional spray chambers.
Nevertheless, a precision of
0.08‰ is achievable, when
introducing an acidic Hg solution by means of a tandem
quartz spray chamber arrangement with a perfl uoroalkoxy
(PFA) nebulizer. As much as 250 ng/mL of Hg was required
for the measurement, though (Malinovsky et al., 2008).
4-6),
which was deemed insuffi cient to detect isotope ratio
variations among coal samples. Subsequent modifi cations
aimed at extending and fl attening the peak improved the
precision and accuracy of the method and succeeded in
improving the precision of the isotope ratio measurement
to better than 0.1‰ (Xie et al., 2005).
0.5‰ RSD (n
Gas Chromatography
The next frontier in Hg isotope ratio measurements is the
determination of compound specifi c isotope ratios, specifi -
cally for methylmercury (MMHg). This requires either an
offl ine isolation of MMHg from the sample or a chromato-
graphic separation of Hg species, which is directly inter-
faced with the MC-ICP/MS. Offl ine separation techniques
must recover 100% of the MMHg in the sample (to avoid
fractionation during sample processing) and remove diva-
lent mercury [Hg(II)] quantitatively at the same time. On
the other hand, direct coupling of gas chromatography (GC)
to ICP/MS is a standard technology used routinely for quan-
titative mercury speciation, and numerous suitable sample-
processing schemes exist for measuring MMHg by GC-ICP/
MS (this topic, chapter 5). However, this technique again
generates transient signals (chromatographic peaks), which
are diffi cult to process for precise isotope ratio determina-
tions. To make matters worse, isotope ratios change con-
tinuously while the peak eluted from a packed GC column
(Dzurko et al., 2009). This apparent fractionation was not
caused by fractionation of MMHg on the column, but
instead relates to the amplifi er-detector arrangement. By
carefully optimizing the peak integration and isotope ratio
calculation algorithm, the authors achieved a precision of
0.16‰ RSD (n
Direct Injection of Gaseous Elemental Mercury
One technique collects Hg(0) vapor, which is thermode-
sorbed from gold, in the barrel of a gas-tight glass syringe for
temporal storage (Sonke et al., 2008). The syringe is then con-
nected to the nebulizer gas stream of the MC-ICP/MS. Using
a syringe pump, the Hg(0) is slowly injected at a continuous
rate, generating a steady Hg signal. Depending on the Hg con-
centration in the syringe, the rate of injection can be varied
between 0.1 and 20 mL/min to optimize the Hg signal. This
technique provided a precision of 0.24‰ (
2 SD) for
δ
202
Hg.
Continuous-Flow Cold-Vapor Generation
The need to generate a steady signal for data acquisition
has established continuous-fl ow cold-vapor generation as
the method of choice for sample introduction (Foucher
and Hintelmann, 2006; Hintelmann and Lu, 2003; Klaue
and Blum, 1999, 2000). Typically, the acidic sample is con-
tinuously mixed with a reducing solution using peristaltic
pumps and Ar gas is introduced into a gas-liquid separator.
The gaseous Hg(0) is stripped from solution and transported
to the ICP. Usually, stannous chloride is used as the reduc-
tant, but also borohydride has been evaluated (Hintelmann
and Lu, 2003).
Ideally, the Hg concentration in the sample solution
is between 2 and 30 ng/mL. At sample uptake rates of
approximately 1 mL/min, up to 10 minutes of data acquisi-
tion are required to introduce suffi cient Hg for a precise
measurement. The memory effects associated with Hg
measurements are well documented (Hintelmann, 2003).
Therefore, thorough rinsing of the cold-vapor system is
8) for the
202
Hg:
198
Hg ratio of MMHg and
0.18‰ RSD for the same ratio of inorganic Hg. Drifting
isotope ratios during peak elution were also observed when
using a capillary GC column interfaced to MC-ICP/MS
(Krupp and Donard, 2005). The authors provided a thor-
ough analysis of potential sources for this behavior, includ-
ing drift in mass bias, chromatographic fractionation, drift
in background signal, and infl uence of analyte concentra-
tion, but were unable to fi nd a satisfactory explanation. The
lower sample-loading capacity of the capillary GC tech-
nique allowed a maximum injection of only 500 pg of Hg
per measurement, resulting in an isotope ratio precision of
only ~0.5‰ RSD for MMHg standards (Epov et al., 2008).
