1. Introduction
Mercury is recognised as one of the most hazardous environmental pollutants. It is released into
the environment by anthropogenic and natural sources such as: volcanoes; industrial runoffs from
contaminated soils; as well as gold and ore mining. According to the U.S. Environmental Protection
Agency (US EPA), concentrations of total mercury (T-Hg) in biological samples are usually less than
0.1 mg/kg [1], while levels in sediment vary depending on the state of pollution of the area under study,
the proper assessment of which necessitates the analysis of a large number of samples. Thus, there is
a need for a robust, validated and inexpensive analytical method. Moreover, these methods should
match the levels of total Hg that are hopefully, expected to be low.
The most frequently used protocols to determine total Hg in some biological samples employ
atomic absorption spectrometry [2] and inductively coupled plasma mass spectrometry (ICP-MS) [3].
However, these methods require an additional sample preparation step that, significantly increases theanalysis time and may be a source of contamination. Furthermore, there are some issues encountered
when ICP-MS is used in the determination of total Hg. The use of the mercury analyser instrument
model-AULA-254 is one of the recent alternatives to perform the T-Hg determination. Its uses include
the evaluation of various environmental samples such as water, soils, sludge, and the analysis of
foodstuffs and biota.
The analysis principle is based on a continuous flow method, where the mercury within the
sample is converted into its elemental state with the aid of a reducing agent. The mercury is then
stripped and carried to an optical cell, where the determination is performed by UV absorption at
a wavelength of 253.7 nm. This technique is more commonly known as cold vapour atomic absorption
spectrometry (CV AAS) [4].
In general, sensitive analytical methods are required for validation of traces of mercury in sediment
and biota, since its concentration may be too low to be analysed by conventional methods.
The study aimed to develop and validate a robust method used to prepare and analyse various
environmental samples. The novelty of the proposed method lies in its capability to estimate T-Hg
content in a variety of environmental samples (water, sediment, and biota) following set sample
preparation steps.
The quality of the proposed method has been checked by performing a complete method
validation using set International Union of Pure and Applied Chemistry (IUPAC) criteria and
employing two CRMs, its application to real samples has been illustrated by processing several
environmental samples from the marine environment.
2. Material and Methods
2.1. Apparatus, Chemicals, and Reagents
Cold vapour atomic absorption spectrometry was utilised for T-Hg measurement,
using an AULA-254 Gold (Mercury instruments GmbH Analytical Technologies, Karlsfeld, Germany),
equipped with an electrodeless low-pressure mercury UV-absorption source set at a wavelength
of 253.7 nm and running AULA-254 software (AULA-WIN, TM based) [5] for data processing.
Argon 99.999% (Buzwair Scientific and Technical Gases, Doha, Qatar) and high-quality type 2 deionised
water (~18 MW cm) from Thermos Barnstead system (GENPURE UV-TOC Life Technologies Ltd.
Paisley, UK) were used. Other reagents such as Suprapur nitric acid (65%), sulfuric acid (98%),
hydrofluoric acid (40%), an oxidising reagent (Potassium dichromate, EMSURE® ACS), potassium
permanganate (EMSURE® ACS, Reag), hydroxylamine hydrochloride (99.999% trace metals basis),
and tin-II-chloride ( 99.99% trace metals basis) were all obtained from Merck (Boston, MA, USA).
Certified reference material (CRM) NIST-2976 (mussel tissue (trace elements and methylmercury),
certified value: 0.061 0.0036 mg/kg) were sourced from the National Institute of Standards and
Technology (Gaithersburg, MD, USA). CRM PACS-3 (Marine Sediment Reference Material for Trace
Metals and other Constituents), certified value: 2.98 0.36 (mg/kg) was purchased from the National
Research Council Canada (Ottawa, ON, Canada). Mercuric nitrate standard was used as a stock
calibration standard (1001 2 pg/mL, 2%, nitric acid in low TOC water (<50 ppb)).
2.2. Sample Preparation
The Environmental Science Centre has adapted the following procedure in preparing samples for
T-Hg analysis. Approximately 0.1 0.05 g of freeze-dried and homogenised sample was transferred
into a 50 mL Teflon tube, 1 mL HF was added together with 5 mL HNO3, and 3 mL H2SO4
(Suprapur, Merck) to decompose and release matrix-bound mercury. The sample was then digested
using a hot block set at 125 C for 12 h with the cap closed to reduce mercury loss by volatilisation.
After complete digestion, indicated by a clear solution, the tube was left to cool to room temperature
after which 1 mL of resulting mixture was transferred into a 50 mL measuring flask together with 2 mL
potassium dichromate, and the final volume made up to the mark with reagent water. This constitutesthe sample solution. Other protocols employ different sample preparation techniques that may require
different sample weights, utilise different digestion acids and other digestion apparatus [6,7].
2.3. Sample Analysis
Ten mL of the prepared sample was introduced into the AULA-254 for analysis. The process is
automated in which the sample is oxidised, heated, reduced and finally analysed for T-Hg.
Moisture content determination of the environmental sample matrices was also carried out by
a drying procedure (105 2 C for 48 h) using a validated and ventilated oven. Ten samples were
used to calculate the mean moisture content with minimum sample mass 1 g [8,9].
3. Results and Discussion
3.1. Method Validation
The method was validated according to the IUPAC guideline, and Eurachem Guides,
which include characterisation of selectivity, trueness, recovery, precision (repeatability & intermediate
precision), limit of detection, limit of quantitation, linearity, range, and ruggedness [10,11].
The uncertainty of measurement was also calculated. The validation process was performed by
analysing two different CRM matrices (PACS-3 and NIST-2976).
3.1.1. Selectivity
Selectivity is the degree to which a method can quantitatively select a distinct analyte in the
presence of other analytes that may interfere with the analyte under investigation. Matrix effect was
investigated by analyzing two CRMs; each has a different matrix (muscle tissue and marine sediment).
Interferences can result from anions or matrix. The main interference anion is chloride and sulfide will
be absorbed in the same wavelength of mercury causing a positive interference. The matrix effects can
be of two types; mask effects or background effects. Chloride and sulfide anions can be eliminated by
adding potassium permanganate [12,13].
Two distinct peaks were observed from the CRMs previously described in Section 3.1, for low
and high levels of T-Hg with no significant interference from other constituents within the CRMs with
recovery > 90%. Figure 1 shows a typical of mercury at high (A) and low (B) levels (7 and 0.01 g/L
respectively) and a blank (C) with low noise and low background (absorbance < 0.001). The only T-Hg
peak appears at the same retention time (~60 s) in the, from two prepared CRM samples containing
other constituents.
3.1.2. Trueness
Trueness relates to the systematic error of a measurement system and to how the mean value of
replicate measurements are close to the true value. Eight measurements from two CRMs were analysed
for bias; Table 1 shows the results of the trueness test on different days. The data obtained shows
a relative bias of 4%, and .9% for NIST2976 and PACS3 respectively (Table 2).
Table 1. The mean of eight replicate results of two different matrices (biota and marine sediment
certified reference materials (CRMs)).
CRM n NIST-2976 Ref. Vale PACS-3 Ref. Vale
Mean value (mg/Kg) 8 0.058 0.002 0.061 0.004 2.87 0.15 2.98 0.36