Article pubs.acs.org/EF
Determination of Silicon in Vegetable Oil and Biodiesel by HighResolution Continuum Source Flame Atomic Absorption Spectrometry Using Sample Dilution with Xylene Lígia C. C. de Oliveira,† Mariana A. Vieira,*,‡ Anderson S. Ribeiro,‡ Meibel T. Lisboa,‡ Rodrigo A. Gonçalves,† and Reinaldo C. de Campos†,§ †
Departamento de Química, Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio), 22453-900, Rio de Janeiro (RJ), Brazil Laboratório de Metrologia Química, Programa de Pós-Graduaçaõ em Química, Centro de Ciências Químicas, Farmacêuticas e de Alimentos,Universidade Federal de Pelotas, 96160-000 Capão do Leão (RS), Brazil
‡
ABSTRACT: In this paper, a simple procedure for the determination of Si in vegetable oils and biodiesel of different origins by high-resolution continuum source flame atomic absorption spectrometry is proposed. Samples were diluted with xylene, and external calibration was performed with calibration solutions prepared with xylene and mineral base oil used for viscosity matching. The accuracy was assessed using recovery tests and comparing the results for a set of vegetable oils and biodiesel to those determined by inductively coupled plasma−optical emission spectroscopy (ICP−OES). Recoveries at 3.0 and 5.0 mg L−1 levels using organic Si standard-spiked solutions ranged from 93 to 113%. The limit of detection at the optimized conditions in the original sample was approximately 0.60 mg kg−1, which is adequate for the analysis of vegetable oils and biodiesel according to the Brazilian legislation.
■
INTRODUCTION Most of the energy consumed worldwide comes from oil, coal, and natural gas. These sources are limited, and estimates lead to scenarios in which they may come to exhaustion in the near future. Their use is also the main cause of CO2 release into the atmosphere, contributing to the increase of the greenhouse effect and global warming. Therefore, the search for alternative energy sources turned out to be mandatory. In this context, biodiesel appears as an alternative in the replacement of regular diesel in compression ignition engines.1−5 Biodiesel is a mixture of fatty acid esters derived from the transesterification of triglycerides present in vegetable oils and fats with short-chain alcohols, which also results in glycerol as a co-product. The transesterification is performed in the presence of catalysts, and homogeneous alkaline catalysts are the most used because they are cost-effective and more efficient, promoting high yields.1,5,6 Today, the use of heterogeneous catalysis solves many of these problems associated with homogeneous catalysts.7 Brazilian Law number 11.097/2005 imposes the mandatory addition of 2% biodiesel to diesel (B2) from January 2008 onward, along with a progressive biodiesel increment reaching B5 to be implemented before 2013. In parallel, the National Energy Policy Council (CNPE) from its resolution number 2/ 2008 imposed the addition of 3% biodiesel to diesel (mixture B3) from July 1, 2009 onward. This measure, while also reducing the contribution of mineral diesel in the Brazilian energy matrix, ensures the biodiesel market and production. These deliberations are part of the National Biodiesel Program created alongside the laws.8 Raw materials available for biodiesel in Brazil are extremely diverse, namely, castor beans, babassu, peanuts, Jatropha curcas L., sunflower, cotton, and soybean. All of these vegetables can produce their respective © 2012 American Chemical Society
vegetable oil that can be processed into biodiesel. Also, a wellestablished cattle meat productive chain assures the availability of large amounts of animal fat, a raw material for biodiesel production as well.9 The diversity of feedstock compels the development of advanced techniques to ensure the quality of the resulting biodiesel fuel.9−15 Vegetable oils used in the production of biodiesel can assimilate several metals from soil and water along the vegetable growing cycle by having contact with the metallic tubing and walls of the transport or storage systems. Arsenic, for example, may be present because it is absorbed from the soil by the plant, or it may be incorporated during the industrial production and storage processes. Fertilizers and insecticides are also a potential As source.16 Organometallic compounds are also added to fuels to improve their physical or burn characteristics. For example, silicon compounds, such as siloxanes and manganese, are used as an antifoaming agent and a burn improver, respectively. Silicon concentrations in biofuel feedstock crops have a critical role in combustion processes. During combustion, if this element is present as silicon oxide, it can react and form slag deposits and corrosive compounds and promote air pollution.17−20 Still regarding silicon, specifications for this element in biodiesel are under discussion, and the expected limit will probably be close to 5 mg kg−1 threshold. The legislation in Brazil established the maximum concentration of silicon together with aluminum used only for diesel, and the maximum value expected is 80 mg kg−1.21 Received: August 1, 2012 Revised: October 3, 2012 Published: October 3, 2012 7041
dx.doi.org/10.1021/ef3012867 | Energy Fuels 2012, 26, 7041−7044
Energy & Fuels
■
Analytical procedures that can detect silicon in biodiesel and vegetable oils are necessary. Thus far, no recommended methodology is described, and only one analytical study has been found in the literature. Woli et al.17 determined total silicon concentrations in plant biomass samples by dry-ashing plant tissue in a muffle furnace, followed by alkaline fusion and then colorimetric analysis. An important paper was published by Ferreira et al.22 dealing with the direct determination of silicon in petroleum products, such as naphtha. High-resolution continuum source flame atomic absorption spectrometry (HR-CS FAAS) is a recent breakthrough in relation to the conventional line source AAS because of its enhanced background correction capabilities, the visibility of the spectral environment around the analytical line, enhanced dynamic range, and improved limits of detection. It is also possible to make sequential analysis because it is especially advantageous if a multi-elemental analysis is necessary, as happens in biodiesel characterization. Thus, the objective of the present study is to use the procedure of sample preparation based on the ABNT NBR 15556:200823 for the determination of silicon in vegetable oils and biodiesel samples by HR-CS FAAS.
■
Article
RESULTS AND DISCUSSION
Optimization of the Instrumental Parameters. Among the most important non-spectral variables to be evaluated in the flame atomic absorption technique are flame stoichiometry, defined by the flow of the flame gases, the sample aspiration rate, and the observation height in the flame. In the case of aqueous solutions, these variables are well-defined; however, in the case of organic solutions, their values can differ greatly. Initially, the need to work with a more oxidizing flame was evaluated to compensate for the effect of the fuel used. The minimum aspiration rate was used to compensate for the effect caused because of the introduction of the solvent into the flame. The choice of the observation of the height was made manually by verifying which height generated the maximum absorbance. Given that the procedure present in the Brazilian norm23 for the determination of Na, K, Ca, and Mg indicates the dilution of the sample with xylene and because we sought to use the same sample solution in the determination of Si, this same dilution was chosen in the present study. For external calibration, as occurs for other procedures in which sample pretreatment consists only of a dilution with an organic solvent, it is necessary to adjust the viscosity of the calibration solutions. On the basis of studies reported by Oliveira et al.15 a 120 g L−1 mineral oil solution in xylene was chosen as the diluting medium, because this way the calibration solutions showed a final viscosity similar to those of the diluted samples. Another important point to be considered is the stability of the solutions in xylene. Metal solutions in organic medium are unstable, and the analytical signal tends to decrease with time. Thus, because of this lack of stability, it is important to analyze the calibration and sample solutions immediately after their preparation, making the use of the whole automatic sampler carrousel difficult.15,24 Calibration Studies. Because vegetable oil and biodiesel of different sources present different composition (and consequently different properties, such as viscosity and density) as observed earlier,16 the matrix influence was investigated by observing the slopes of analyte addition curves with a set of vegetable oil and biodiesel samples. Also, to define the calibration mode, the slope of an external calibration curve was compared to those obtained from analyte addition curves in the presence of the samples. No significant differences (t test; p < 0.05) were observed between these slopes. Table 1 shows the slope values. Possible spectral interferences were also investigated, because the equipment allows for the visibility of the spectral environment around the analytical line, and such interferences were found to be absent. Figures of Merit. The dynamic range of the calibration curve should be selected on the basis of the range of analyte concentrations (maximum permissible concentration of 5 mg L−1) that should be expected; however, this curve was found to be linear up to, at least, 50 mg L−1. On the basis of the calibration curve, the characteristic concentration, limit of detection (LOD), and linear determination coefficient were obtained. The instrumental LOD was calculated as 3 times the standard deviation of 10 measurements of the blank divided by the slope of the respective calibration curve. In the case of the LOD in the original sample, the dilution factor (10 times) was taken into account, which certainly underestimated the uncertainty associated with this step. However, because the uncertainty associated with the dilution step is usually much lower than that of the instrumental measurement, it may be
EXPERIMENTAL SECTION
Instrumentation. Silicon measurements were conducted with a continuum source atomic absorption spectrometer model ContrAA 300 (Analytik Jena, Jena, Germany) with a flame atomization system. This equipment is provided with a xenon short-arc lamp, operating in hot-spot mode as the continuum radiation source (190−850 nm), an Echelle double monochromator, and a charge-coupled device (CCD) line detector. The analytical wavelength of 251.611 nm was adopted with a simultaneous evaluation of 200 pixels corresponding to a spectral environment of approximately ±0.2 nm around the center pixel. The atomization was performed using an acetylene−nitrous oxide flame formed at gas flows of 280 and 613 L h−1, respectively. The aspiration rate was 1.1 L min−1, adjusted for all samples and calibration solutions, and the read time was 5 s. All measurements were made in absorbance. Samples were weighed in an Ohaus Adventurer AR 2140 (Pine Brook, NJ) analytical balance with an accuracy of 0.0001 g. For comparison studies, an inductively coupled plasma optical emission spectrometer model Plasma P1000 (PerkinElmer, Norwalk, CT) was used. The analytical emission ionic line chosen was 251.611 nm. Materials, Reagents, Solutions, and Samples. Acetylene and nitrous oxide (both 99.99%) were provided by Linde (Rio de Janeiro, Brazil). All reagents were of analytical reagent grade. Analyticalreagent-grade xylene and mineral oil (viscosity = 75 mm2 s−1) were purchased from Vetec, Rio de Janeiro, Brazil. A silicon base oil standard of 1000 mg kg−1 (Conostan, Houston, TX) was used. A silicon stock standard solution of 100 mg L −1 was prepared using xylene as a diluent. Calibration solutions were obtained from adequate microvolume dilutions of this solution in a 120.0 g L−1 mineral oil in xylene solution for viscosity adjustment. The 120.0 g L−1 solution of oil in xylene was prepared diluting 120.0 g of mineral oil to 1000 mL in a volumetric flask with xylene. Eight types of vegetable oil samples from different origins were analyzed, namely, soybean (crude, milled, and degummed), castor, corn, sunflower, cotton, and canola. Nine biodiesel samples from different oil sources were also analyzed, namely, soybean, sunflower, algae, frying, peanut, castor, J. curcas L., cotton, and canola. Analytical Procedure. Approximately 1.0 g of vegetable oil or biodiesel was accurately weighed in a 15 mL polypropylene flask and diluted to 10 mL with xylene. Calibration solutions were prepared from the intermediary (100.0 mg L−1) Si organometallic solution and adequately diluted with the stock solution of oil in xylene (120.0 g L−1) to match the viscosity of these solutions with those of the diluted samples.15 7042
dx.doi.org/10.1021/ef3012867 | Energy Fuels 2012, 26, 7041−7044
Energy & Fuels
Article
Table 3. Measureda Concentrations of Organic Si (mg L−1; n = 3) by HR-CS FAAS in Vegetable Oil and Biodiesel after the Addition of 1, 3, and 5 mg L−1 to the Original Sample
Table 1. Slopes and Intercepts of the External and Analyte Addition Si Calibration Curves in Vegetable Oil and Biodiesel Samples addition curves soybean (crude) soybean (degummed) sunflower canola corn cotton soybean castor peanut sunflower canola xylene + mineral oil
1 mg L−1
slope (±sd) + intercept (±sd) Vegetable Oil 0.0058 (±0.0003)x 0.0063 (±0.0002)x 0.0062 (±0.0002)x 0.0061 (±0.0001)x 0.0061 (±0.0007)x 0.0062 (±0.0002)x Biodiesel 0.0061 (±0.0002)x 0.0061 (±0.0003)x 0.0062 (±0.0002)x 0.0062 (±0.0002)x 0.0066 (±0.0002)x 0.0066 (±0.0003)x
+ + + + + +
0.0002 0.0002 0.0001 0.0015 0.0012 0.0004
(±0.0009) (±0.0005) (±0.0001) (±0.0003) (±0.0002) (±0.0006)
soybeanb (crude) soybeanc soybean sunflowerd corn castor
+ + + + +
0.0019 0.0017 0.0015 0.0023 0.0021
(±0.0003) (±0.0009) (±0.0007) (±0.0005) (±0.0007)
soybeanb sunflower peanut J. curcas L. canola cotton a
5 mg L−1
3.3 3.4 3.2 3.3 3.1 3.4
± ± ± ± ± ±
0.1 0.1 0.3 0.2 0.5 0.4
5.7 5.6 5.2 5.8 4.9 5.5
± ± ± ± ± ±
0.2 0.1 0.2 0.4 0.3 0.2
3.4 3.3 2.9 3.2 3.2 2.8
± ± ± ± ± ±
0.3 0.2 0.4 0.6 0.4 0.1
5.2 5.1 5.5 5.6 5.3 5.1
± ± ± ± ± ±
0.3 0.2 0.4 0.4 0.3 0.3
External calibration as described in the text. bCrude. cDegummed. Milled.
d
considered negligible in this level of concentration. Table 2 presents the figures of merit obtained.
Table 4. Measured Concentrations of Si (mg kg−1; n = 5) Obtained for Vegetable Oil and Biodiesel Sample Solution by HR-CS FAASa and ICP−OESb
Table 2. Figures of Merit linear range (mg L−1) characteristic concentration (mg L−1) correlation coefficient (R2) reproducibility (%) instrumental LOD (mg L−1) sample LOD (mg kg−1)
Vegetable Oil 1.2 ± 0.1 1.1 ± 0.1 0.95 ± 0.12 1.1 ± 0.1 0.92 ± 0.19 1.2 ± 0.1 Biodiesel 1.1 ± 0.2 0.99 ± 0.13 0.94 ± 0.07 1.1 ± 0.1 1.0 ± 0.1 1.2 ± 0.1
3 mg L−1
1.0−50.0 0.63
