Professional Documents
Culture Documents
net/publication/235582095
CITATIONS READS
16 417
5 authors, including:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Alexandre de Jesus on 10 March 2018.
Analytical Note
a r t i c l e i n f o a b s t r a c t
Available online 1 April 2012 In this work, a direct sampling graphite furnace atomic absorption spectrometry method has been developed
for the determination of arsenic and cadmium in crude oil samples. The samples were weighed directly on
Keywords: the solid sampling platforms and introduced into the graphite tube for analysis. The chemical modifier used
Arsenic for both analytes was a mixture of 0.1% Pd + 0.06% Mg + 0.06% Triton X-100. Pyrolysis and atomization
Cadmium curves were obtained for both analytes using standards and samples. Calibration curves with aqueous
Direct sampling
standards could be used for both analytes. The limits of detection obtained were 5.1 μg kg− 1 for arsenic and
Graphite furnace atomic absorption
spectrometry
0.2 μg kg− 1 for cadmium, calculated for the maximum amount of sample that can be analyzed (8 mg and
10 mg) for arsenic and cadmium, respectively. Relative standard deviations lower than 20% were obtained.
For validation purposes, a calibration curve was constructed with the SRM 1634c and aqueous standards for
arsenic and the results obtained for several crude oil samples were in agreement according to paired t-test.
The result obtained for the determination of arsenic in the SRM against aqueous standards was also in
agreement with the certificate value. As there is no crude oil or similar reference material available with a
certified value for cadmium, a digestion in an open vessel under reflux using a “cold finger” was adopted for
validation purposes. The use of paired t-test showed that the results obtained by direct sampling and
digestion were in agreement at a 95% confidence level. Recovery tests were carried out with inorganic and
organic standards and the results were between 88% and 109%. The proposed method is simple, fast and
reliable, being appropriated for routine analysis.
© 2012 Elsevier B.V. All rights reserved.
0584-8547/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.sab.2012.03.010
A. de Jesus et al. / Spectrochimica Acta Part B 71-72 (2012) 86–91 87
The spike/recovery tests with inorganic standard were carried out Pyrolysis curves for As were accomplished with the sample OB-32
injecting 20 μL of the modifier solution onto the platform and drying and with 1.0 ng of As in aqueous standard. As can be seen in Fig. 1a
as described above. After cooling, the sample was weighed (masses the best pyrolysis temperature was 1200 °C for both, aqueous
up to 5 mg) onto the SS platform and 10 μL of aqueous standard was standard and crude oil sample. For the sample OB-32, reliable
added on top of the sample and the complete temperature program measurement of analyte signals was only possible with pyrolysis
was executed. The samples spiked with As were OB-32, OB-12 and temperatures of 600 °C or higher due to excessively high background
SRM 1634c and those with Cd were OB-10, 0B-26 and 0B-27. As it was absorption at lower temperatures as the matrix could not be
found just one sample containing Cd above the detection limit (OB- eliminated efficiently. The same was observed by Becker et al. [18]
10), three other samples (OB-41, OB-46 and OB-47) were spiked with for As in gasoline microemulsions. A pyrolysis temperature of 1200 °C
Cd organic standard in order to obtain concentrations around 30, 10 was used for all further experiments with solutions and crude oil
and 20 μg kg − 1, respectively. In this procedure, the organic standard samples. The best atomization temperature, based on the sensitivity
was added to an exact amount of sample and after homogenization and peak shape was 2400 °C.
according to the procedure described above, it was taken to the Some typical atomization and background signals are shown in
digestion and direct sampling analysis. Fig. 2. In order to improve the absorbance signal, given a better peak
As there is no certified reference material for Cd in petrochemical shape, and decrease the background signal for As, the use of a “Cool
samples, the sample digestion was carried out in order to validate the down” step before the atomization was necessary, as shown in
proposed method for determination of Cd in crude oil. The samples Table 1. As can be seen in Fig. 2b and d a more symmetric and
were digested in a block under reflux using a “cold finger”. A mass of sensitive absorbance signal either for the standard or the sample (OB-
0.25 g of samples, was weighed directly into an open glass tube and 32) was obtained using this step. An increase in sensitivity using a
6.0 mL of concentrated nitric acid was gradually added. The mixture cool-down step was also reported by Husáková et al. [19] for the
was heated up to 150 °C and then kept for 24 h. After this 2.0 mL of direct determination of As in beer by GF AAS. Fig. 2b and d also shows
hydrogen peroxide (30% v/v) was added to complete the digestion. the similarity of the atomization signals for As obtained for aqueous
The digestion products were completed up to 15 mL with ultrapure standard and for crude oil sample by DS-GF AAS. The background
water. All samples were digested in triplicates. signal was low, demonstrating the efficient removal of the crude oil
matrix prior to the atomization stage.
3. Results and discussion For Cd, pyrolysis curves were established for 20 pg of Cd in
aqueous standard and the crude oil sample OB-10. Based on the
3.1. Modifiers and temperature program results shown in Fig. 1b for the crude oil sample, the pyrolysis
temperature of 750 °C was chosen. The best atomization temperature,
The strategy for weighing crude oil samples, the optimization of based on the sensitivity and peak shape was 1700 °C. Fig. 2e and f
the drying stage of the temperature program, as well as the use of shows the absorbance signals for Cd in aqueous standard and in the
Triton X-100 in the conventional chemical modifier, have been sample OB-10 using Pd/Mg as the chemical modifier. The background
established in previous works [12,14]. The pyrolysis and atomization
steps of the temperature program have been optimized for direct
determination of As and Cd crude oil using crude oil samples and
SRM. Among the chemical modifiers that have been proposed for the
determination of As in petroleum products as microemulsion are Pd
[16,17] and Pd/Mg [18] as conventional chemical modifiers added in
solution and, Ir [16,17] and Ir/W [17] as permanent modifiers. In this
work the following chemical modifiers were investigated: the
mixtures of 0.1% Pd; 0.1% Pd + 0.06% Mg (20 μL) for both analytes;
and the permanent chemical modifier 600 μg Ir for As.
As the procedure of depositing the modifier solution onto the SS
platform and drying it before each sample has been weighed on the
platform prior to its introduction into the graphite furnace obviously
complicates the analytical procedure; thus the use of 600 μg of
iridium as permanent modifier has been investigated first. The first
results with crude oil samples were very promising as the back-
ground signal was lower and the maximum pyrolysis temperature
(1500 °C) was higher than that obtained with the conventional
modifiers investigated. However, the sensitivity was lower for the
aqueous standard and for crude oil samples and the result obtained
was about 60% of the certified value for As in the reference material
SRM 1634c. Lower sensitivity for As in diesel, gasoline and naphtha
microemulsions with 600, 900 and 1200 μg of Ir was also related by
Brandão et al. [16]. Based on this discussion we concluded that the
permanent modifier, at least under the present conditions, was not
efficient and further investigation is necessary if it is to be used for the
direct determination of As in crude oil with GF AAS. The conventional
chemical modifiers have therefore been investigated. Considering the
conventional chemical modifiers tested, the best absorbance signals
and reduction in the background signal were obtained with the Fig. 1. Pyrolysis curves using DS-GF AAS for: (a) As in standard solution and crude oil
mixture of 0.1% Pd + 0.06% Mg + 0.06% Triton X-100. A preliminary sample OB-32 (absorbance signal normalized for 5.0 mg of the sample), atomization
temperature=2400 °C; (b) Cd in standard solution and crude oil sample OB-10 (absorbance
study of the amount of the modifier showed that the best signal for signal normalized for 2.0 mg of the sample), atomization temperature=1700 °C. Chemical
the samples was obtained with 20 μL of the modifier solution, modifier: 0.1% Pd+0.06% Mg+0.06% Triton X-100. Error bars refer to the SD of 3 and 6
containing 20 μg of Pd and 12 μg of Mg. consecutive measurements of the standards and crude oil samples, respectively.
A. de Jesus et al. / Spectrochimica Acta Part B 71-72 (2012) 86–91 89
Fig. 2. Absorbance signals for As and Cd using DS-GF AAS. Full lines represent the atomic absorption (AA) and broken lines represent the background (BG) signal. (a) 2 ng As
without cool down; (b) 2 ng As with cool down; (c) sample OB-32 without cool down; (d) sample OB-32 with cool down; (e) 50 pg of Cd; (f) sample OB-10. The temperature
programs are given in Table 1. Chemical modifier: 0.1% Pd + 0.06% Mg + 0.06% Triton X-100.
signal was negligible, demonstrating the efficient removal of the calibration with the SRM 1634c, as proposed by Kurfürst for solid
crude oil matrix in this stage. It should be emphasized that the use of sampling analysis [20], was also performed for As. The slopes of the
a cool-down step was not necessary to obtain a well defined peak calibration curve for the aqueous standard and with the SRM 1634c
shape and decrease the background signal for this element. The use of were similar confirming that aqueous standards can be used for
this step did not increase the sensitivity as well. Therefore it was not
used for Cd in order to simplify and shorten the temperature program. Table 2
The correlation between sample mass and integrated absorbance Analytical figures of merit obtained for the determination of arsenic and cadmium
was evaluated using the crude oil sample OB-32 (with low viscosity) using DS-GF AAS.
and the more viscous SRM 1634c for As, the crude oil sample OB-10 Parameter As Cd
for Cd, and the conditions optimized above. The sample mass has
Linear regression with A = 0.2249 m (ng) + 0.0028 A = 0.0043 m (pg)+ 0.0340
been varied between 3.0 and 10.0 mg. The correlation coefficients of aqueous standard
the curves were r ≅ 0.98 for As for a sample mass up to 8 mg, and Linear regression with A = 0.2592 m (ng) + 0.0029 –
r ≅ 0.99 for Cd for a sample mass up to 10 mg. SRM 1634c
Correlation coefficient 0.9971a/0.9843b 0.9992
Characteristic mass/pg 17 0.7
3.2. Figures of merit
LODc (n = 10)/μg kg− 1 5.1 0.2
LOQc (n = 10)/μg kg− 1 17 0.7
The figures of merit obtained for As and Cd by DS-GF AAS are shown a
Aqueous standard.
in Table 2. Calibration curves were established using a blank and five b
SRM 1634c.
calibration solutions in the concentration range of 25–200 μg L − 1 c
Based on the technique “zero mass response” [20], considering 8.0 mg for As and
(0.25–2.0 ng) for As and from 1 to 5 μg L − 1 (0.01–0.05 ng) for Cd. A 10 mg for Cd.
90 A. de Jesus et al. / Spectrochimica Acta Part B 71-72 (2012) 86–91
Table 3 standard was in good agreement with the certified value. Using a
Content of arsenic in crude oil samples obtained by DS-GF AAS. paired Student's t-test there was no significant difference between
Sample Calibration against aqueous standard Calibration against the SRM 1634c the results obtained for the samples by both methods on a 95%
mg kg− 1 (n=6)a mg kg− 1 (n=6)a confidence level. The agreement between the results confirms that
OB-03 0.104 ± 0.014 0.088 ± 0.012 aqueous calibration standards can be used for the determination of As
OB-10 0.059 ± 0.008 0.051 ± 0.007 in crude oil samples with the proposed technique.
OB-12 0.051 ± 0.006 0.044 ± 0.006 Table 4 summarizes the results obtained for the determination of
OB-13 0.120 ± 0.011 0.105 ± 0.010
Cd in four crude oil samples using a calibration by DS-GF AAS. As
OB-25 0.051 ± 0.008 0.044 ± 0.007
OB-32 0.055 ± 0.006 0.069 ± 0.005 there is no crude oil or similar reference material available with a
OB-85 0.214 ± 0.012 0.201 ± 0.031 certified value for Cd, it became necessary to use an independent
SRM 0.147 ± 0.017 – technique for validation (digestion under reflux using “cold finger”).
1634cb As stated before, between all samples analyzed in this work, it was
a
SD of six consecutive measurements. found that just one sample contained Cd above the detection limit
b
Certified value for As in SRM 1634c: 0.1426 ± 0.0064 mg kg− 1. (OB-10). Thus three other samples were spiked with different
amounts of Cd organic standard and homogenized as described in
Section 2.3. After homogenization, these samples were taken to the
calibration. The characteristic mass values based on aqueous standard digestion and direct sampling analysis. The results obtained for OB-10
obtained for arsenic and cadmium by DS-GF AAS are in good agreement and for the three spiked samples by DS-GF AAS did not differ
with those reported in the literature [21]. The limits of detection and significantly from the values obtained by the digestion procedure
quantification have been calculated for the maximum sample mass according to the paired Student's t-test at the 95% confidence level.
investigated for each element, i.e., 8 mg for As and 10 mg for Cd. The The recovery of the organic standard added in these samples is also
blank measurements were carried out according to the ‘zero mass presented in Table 4. The results showed that the matrix of the crude
response’ technique [20], introducing repeatedly a solid sampling oil samples does not have a significant influence on the sensitivity of
platform, containing only the modifier, followed by a regular atomiza- the organic standards added.
tion cycle. The LOD obtained in this work are 2 times better than those In addition, a recovery test for As and Cd using inorganic standards
reported by Stigter [7] for digested crude oil samples (10 and was also performed. Three samples were spiked with 2 ng of
0.4 μg kg− 1, for As and Cd, respectively). inorganic arsenic and 50 pg of inorganic cadmium. The recoveries
The precision of the DS-GF AAS method for As was calculated from varied from 91 to 98% for As, and 96 to 109% for Cd. These results
six consecutive measurements of seven real samples and expressed as showed that the matrix of the crude oil samples does not have a
the relative standard deviation (RSD), varying between 6% and 17%. significant influence on the sensitivity of the inorganic standards
The precision for Cd was calculated from six consecutive measure- added.
ments of one real sample (OB-10) and three enriched (spiked) The overall results are satisfactory and show that the proposed
samples and values varied from 2 to 20%. These values are higher than method has provided accurate results although it is very simple and it
the values reported in direct crude oil analysis for Ni and V [12], uses aqueous standards for calibration.
elements that are present in much higher concentrations, but similar
to values reported for Fe, Cu [13] and Pb [14]. Considering the 4. Conclusion
complexity of the crude oil matrix and the direct sampling, that
avoids errors of manipulation, the RSD found in this work has been The method developed in this work provided accurate results for
considered acceptable. In this sense, the reproducibility of the the determination of arsenic and cadmium in crude oil samples. The
digestion of crude oil for Cd, expressed as the RSD of the three feasibility of using calibration against aqueous standards, a conven-
aliquots digested of the same sample, varied between 1% and 19%, tional Pd/Mg modifier and the minimum sample manipulation makes
values that are similar to those obtained with DS-GF AAS. The lifetime this procedure appropriate for routine applications. The analysis of
of the graphite tubes was around 300 and 450 heating cycles for As crude oil by DS-GF AAS has the clear advantage of providing fast
and Cd respectively, and around 200 cycles for graphite platforms, for results, as no sample preparation at all is involved, besides the
both elements. homogenization of the samples, avoiding the use of organic solvents
and other reagents, which might be considered a contribution to the
3.3. Analytical results green chemistry. It might be expected that this technique could be
extended to the determination of other trace elements in crude oil.
The results obtained for the determination of As in seven crude oil
samples and in the SRM 1634c by DS-GF AAS, using a calibration with Acknowledgment
aqueous standard, are summarized in Table 3. For comparison the
samples were also analyzed using a calibration with the SRM 1634c as The authors are grateful to Conselho Nacional de Desenvolvimento
proposed by Kurfürst for solid sampling analysis [20]. The result Científico e Tecnológico — CNPq, (Projeto Universal — Processo 478271/
obtained for the SRM 1634c with the calibration against aqueous 2007-7 and 476448/2009-3) and FAPERGS/CNPq (Programa PRONEX —
Table 4
Content of cadmium in crude oil samples obtained by GF AAS.
process no. 10/0012-7) for the financial support. A.J. and A.V.Z. have [11] G.P. Brandão, R.C. Campos, E.V.R. Castro, H.C. Jesus, Direct determination of nickel
in petroleum by solid sampling-graphite furnace atomic absorption spectrome-
scholarships from CAPES and M.M.S. and M.G.R.V. have scholarships try, Anal. Bioanal. Chem. 386 (2006) 2249–2253.
from CNPq. The authors would like to thank Tamara S. Castilhos and [12] M.M. Silva, I.C.F. Damin, M.G.R. Vale, B. Welz, Feasibility of using solid sampling
Renan B. Scartazzini for their careful experimental work. We also thank graphite furnace atomic absorption spectrometry for speciation analysis of
volatile and non-volatile compounds of nickel and vanadium in crude oil, Talanta
Petrobras for providing the crude oil samples. 71 (2007) 1877–1885.
[13] G.P. Brandão, R.C. Campos, E.V.R. Castro, H.C. Jesus, Determination of copper, iron
References and vanadium in petroleum by direct sampling electrothermal atomic absorption
spectrometry, Spectrochim. Acta Part B 62 (2007) 962–969.
[1] M.D. Lewan, J.B. Maynard, Factors controlling enrichment of vanadium and nickel [14] I.C.F. Damin, M.B. Dessuy, T.S. Castilhos, M.M. Silva, M.G.R. Vale, B. Welz, D.A.
in the bitumen of organic sedimentary rocks, Geochim. Cosmochim. Acta 46 Katskov, Comparison of direct sampling and emulsion analysis using a filter
(1982) 2547–2560. furnace for the determination of lead in crude oil by graphite furnace atomic
[2] C. Hardaway, J. Sneddon, J.N. Beck, Determination of metals in crude oils by absorption spectrometry, Spectrochim. Acta Part B 64 (2009) 530–536.
atomic spectroscopy, Anal. Lett. 37 (2004) 2881–2899. [15] I.M. Dittert, J.S.A. Silva, R.G.O. Araujo, A.J. Curtius, B. Welz, H. Becker-Ross, Direct and
[3] I.A. Silva, R.C. Campos, A.J. Curtius, S.M. Sella, Determination of lead and copper in simultaneous determination of Cr and Fe in crude oil using high-resolution
kerosene by electrothermal atomic absorption spectrometry: stabilization of continuum source graphite furnace atomic absorption spectrometry, Spectrochim.
metals in organic media by a three-component solution, J. Anal. At. Spectrom. Acta Part B 64 (2009) 537–543.
8 (1993) 749–754. [16] G.P. Brandão, R.C. Campos, A.S. Luna, E.V.R. de Castro, H.C. de Jesus, Determination
[4] R. Poreba, P. Gac, M. Poreba, J. Antonowicz-Juchniewicza, R. Andrzejak, Relation of arsenic in diesel, gasoline and naphtha by graphite furnace atomic absorption
between occupational exposure to lead, cadmium, arsenic and concentration of spectrometry using microemulsion medium for sample stabilization, Anal.
cystatin C, Toxicology 283 (2011) 88–95. Bioanal. Chem. 385 (2006) 1562–1569.
[5] M.G.A. Korn, D.S.S. dos Santos, B. Welz, M.G.R. Vale, A.P. Teixeira, D.C. Lima, S.L.C. [17] R.J. Cassella, B.A. Barbosa, R.E. Santelli, A.T. Rangel, Direct determination of arsenic
Ferreira, Atomic spectrometric methods for the determination of metals and and antimony in naphtha by electrothermal atomic absorption spectrometry
metalloids in automotive fuels — a review, Talanta 73 (2007) 1–11. with microemulsion sample introduction and iridium permanent modifier, Anal.
[6] M.M. Silva, M.G.R. Vale, I.C.F. Damin, M.B. Dessuy, A. de Jesus, Direct sampling and Bioanal. Chem. 379 (2004) 66–71.
emulsion analysis for trace metal determination in crude oil by graphite furnace [18] E. Becker, R.T. Rampazzo, M.B. Dessuy, M.G.R. Vale, M.M. da Silva, B. Welz, D.A.
atomic spectrometry, Hydrocarbon World (Online) 6 (2011) 54–57. Katskov, Direct determination of arsenic in petroleum derivatives by graphite
[7] J.B. Stigter, H.P.M. de Haan, R. Guicherit, C.P.A. Dekkers, M.L. Daane, Determina- furnace atomic absorption spectrometry: a comparison between filter and
tion of cadmium, zinc, copper, chromium and arsenic in crude oil cargoes, platform atomizer, Spectrochim. Acta Part B 66 (2011) 345–351.
Environ. Pollut. 107 (2000) 451–464. [19] L. Husáková, T. Cernohorsky, J. Srámková, L. Vavrusová, Direct determination of
[8] B.K. Puri, K.J. Irgolic, Determination of arsenic in crude petroleum and liquid arsenic in beer by electrothermal atomic absorption spectrometry with deuterium
hydrocarbons, Environ. Geochem. Health 11 (1989) 95–99. background correction (D2-ET-AAS), Food Chem. 105 (2007) 286–292.
[9] M.G.R. Vale, N. Oleszczuk, W.N.L. Santos, Current status of direct solid sampling [20] U. Kurfürst, Solid Sample Analysis, first ed. Springer, Berlin, 1998.
for electrothermal atomic absorption spectrometry—a critical review of the [21] B. Welz, M. Sperling, Atomic Absorption Spectrometry, third ed. Wiley-VCH,
development between 1995 and 2005, Appl. Spectrosc. Rev. 41 (2006) 377–400. Weinheim, 1999.
[10] B. Welz, D.L.G. Borges, M.G.R. Vale, U. Heitmann, Progress in direct solid sampling
analysis using line source and high-resolution continuum source electrothermal
atomic absorption spectrometry, Anal. Bioanal. Chem. 389 (2007) 2085–2095.