Determination of Aldehydes and Ketones in Fuel Ethanol by High-Performance Liquid Chromatography with Electrochemical Detection 2006, 63, 45–51
A. A. Saczk&, L. L. Okumura, M. F. de Oliveira, M. V. B. Zanoni, N. R. Stradiotto UNESP – Instituto de Quı´mica, C.P. 355, 14801-970, Araraquara S.P., Brazil; E-Mail:
[email protected]
Received: 22 July 2005 / Revised: 22 November 2005 / Accepted: 28 November 2005 Online publication: 3 January 2006
Abstract A new methodology was developed for analysis of aldehydes and ketones in fuel ethanol by high-performance liquid chromatography (HPLC) coupled to electrochemical detection. The electrochemical oxidation of 5-hydroxymethylfurfural, 2-furfuraldehyde, butyraldehyde, acetone and methyl ethyl ketone derivatized with 2,4-dinitrophenylhydrazine (DNPH) at glassy carbon electrode present a well defined wave at +0.94 V; +0.99 V; +1.29 V; +1.15 V and +1.18 V, respectively which are the basis for its determination on electrochemical detector. The carbonyl compounds derivatized were separated by a reverse-phase column under isocratic conditions with a mobile phase containing a binary mixture of methanol / LiClO4(aq) at a concentration of 1.0 · 10)3 mol L)1 (80:20 v/v) and a flow-rate of 1.1mL min)1 . The optimum potential for the electrochemical detection of aldehydes-DNPH and ketones-DNPH was +1.0 V vs. Ag/AgCl. The analytical curve of aldehydes-DNPH and ketones-DNPH presented linearity over the range 5.0 to 400.0 ng mL)1, with detection limits of 1.7 to 2.0 ng mL)1 and quantification limits from 5.0 to 6.2 ng mL)1, using injection volume of 20 lL. The proposed methodology was simple, low time-consuming (15 min/analysis) and presented analytical recovery higher than 95%.
Keywords Column liquid chromatography Electrochemical detection Aldehydes and ketones in fuel ethanol Dinitrophenylhydrazones
Introduction Ethanol produced from sugarcane has been an important fuel additive since 1913, primarily to insulate the sugar industry from international market fluctuations. In 1975, its production was stimulated by Brazil’s government, with a special program assigned Proa´lcool, in
Original DOI: 10.1365/s10337-005-0698-1 0009-5893/06/01
order to minimizing costs of imported petroleum derivates. Initially, a ‘‘gasohol’’ mixture (22% of anidrous ethanol and 78% of gasoline) was implanted and furthermore the pure form of ethanol was employed as an alternative fuel for national automotive vehicles, leading to several economical and ecological advantages. Today, the annual production of
Brazilian fuel ethanol is around 14 billion m3/year, which classifies it as one of the world’s biggest producers [1]. In spite of being less pollutant than petroleum derivatives, the fuel ethanol also contains contaminants, as example inorganic species [2–3] (such as: zinc, copper, iron, nickel, etc.) and organic compounds [4–5] (such as: aldehydes, ketones, esters, acid organic, and others), which can arise from production process, transport or fuel storage [6–8]. Usually, carbonyl compounds, higher alcohols, esters and organic acids are sub-products generated during alcoholic fermentation. Nevertheless, the quality of final alcoholic products is highly influenced by the presence of aldehydes and ketones [9], and their evaluation requires sensitive analytical methodologies. Classical methods for determination of carbonyl compounds are based on colorimetric measurements [10]. These methods allow determining only the average concentration expressed as sum of all carbonyl species. In last decade the chromatographic methods, mainly high performance liquid chromatography (HPLC) became very popular for the determination of carbonyl species. HPLC with electrochemical detection (ED) is an option to be considered for relatively high polar compounds with redox properties, since high sensitivity and selectivity can be achieved, operating at potential where the analyte can be reduced or oxidized. Taking into consideration that direct determination of carbonyl compounds by
Chromatographia 2006, 63, January (No. 1/2) 2006 Friedr. Vieweg & Sohn/GWV Fachverlage GmbH
45
HPLC methods are not possible, since there is no achievement of level detection, the most methods described in the literature [11–16] for its finality are based on their derivatization reactions. As an example, the most common procedure involves de derivatization of carbonyl compounds with 2,4-dinitrophenylhydrazine (DNPH) [17–19]. Nevertheless, there is no mention of any analytical methodology using HPLC coupled to electrochemical detector (HPLC-ED) on application in fuel ethanol as matrix. The aim of the present work was to develop a new methodology for analysis of aldehydes and ketones such as 5-hydroxymethylfurfural (5-HMF), 2-furfuraldehyde, butyraldehyde, acetone and methyl ethyl ketone, by HPLC-ED using derivatization reaction with DNPH. In addition, the electrochemical behavior of these derivatized compounds was evaluated, and a simple, rapid and sensitive method for quantification of these carbonyl compounds in fuel ethanol samples by HPLCED is proposed.
Experimental Reagents The 2-furfuraldehyde, acetone and metyl ethyl ketone (J.T. Baker – Phillipsburg, USA), 5-HMF and butyraldehyde (Acros Organics – New Jersey, USA) and all solvents used such as methanol, ethanol and acetonitrila (Mallinckrodt – Xalostoc, Mexico) were of HPLC grade. Demineralized water was obtained from a Milli-Q Water System (Millipore, CA, USA). The 2,4-dinitrophenylhydrazine (Merck – Darmstadt, Germany) was purified by three successive recrystallizations from methanol and H3PO4 and LiClO4, LiCl (Aldrich Chemical Company - Milwaukee, USA).
Voltammetric Parameters All the voltammetric measurements were performed using a potentiostat (Autolab, model PGSTAT 30) and a voltammetric cell containing three electrodes, a glassy carbon disk electrode (Methrom), with geometric area 0.71 cm2 was used as working electrode, a platinum wire used as counter electrode and an Ag/AgCl (KCl
46
3.0 mol L)1) used as reference electrode (Methrom). The working electrode was polished using 0.05 lm alumina slurry and rinsed with demineralized water. The optimised voltammetric parameters were: equilibration time of 15 s and scan rate of 50 mV s)1. A methanol solution containing LiClO4 0.1 mol L)1 was used as supporting electrode.
Chromatographic Parameters A ProStar Varian HPLC apparatus was used under isocratic conditions with the mobile phase containing a mixture of methanol / LiClO4(aq) 1.0 · 10)3 mol L)1 (80:20 v/v) and a flow rate of 1.1 mL min)1. Nevertheless, others supporting electrolyte: phosphate buffer solution 1 mM (pH 5 e 7), LiCl(aq) solution 1 mM, acetate buffer solution 1 mM (pH 5) were tested during the optimization of the method. An injection valve of Rheodyne, model 7725 with a 20 lL sample loop was used. A commercial reversed-phase column Shimadzu Shim-pack C18 (150· 6.0 mm I.D., 5 lm) connected with a short guard column Shimadzu Shim-pack C18 (1 cm · 4.0 mm I.D., 5 lm) was employed for all measurements. The electrochemical detector (Varian ProStar 370) based on the wall-jet principle, consisted of a glassy carbon working electrode (area = 0.71 cm2), an Ag/AgCl reference electrode and platinum plate as auxiliary electrode. The optimized potential was obtained at +1.0 V.
Synthesis of Aldehydes and Ketones-2,4Dinitrophenylhydrazone (DNPH) The DNPH compound used as standard was synthesized by the known reaction of carbonyl compounds with DNPH, in acidic medium that promotes the protonation of the carbonyl group [20]. The reaction was obtained as described: DNPH (0.4 g; ca. 2 mmol) was dissolved in 2 mL concentrated phosphoric acid and 3 mL demineralized water. To this solution, the standard solutions of 1 g aldehydes and ketones (5-HMF, 2-furfuraldehyde, butyraldehyde, acetone and metyl ethyl ketone), dissolved in 15 mL ethanol absolute, was added. The reaction products were isolated and purified
twice by recrystallization from absolute ethanol. Standard solutions were prepared by dissolving weighed amounts (10 mg L)1) for each pure aldehydes and ketones-DNPH in acetonitrile. Aliquots of these solutions were diluted with suitable ethanol:water (95:5 v/v) to minimize the matrix effects. The standard solutions used for analytical curves were prepared in the concentration range of 5.0 to 400.0 ng mL)1. The purity of standard reagents was confirmed by melting point determination, elemental analysis (C, H, N), infrared spectroscopy and proton nuclear magnetic ressonance.
Sample Derivatization A 0.4% solution of DNPH was prepared by dissolving DNPH (0.4 g; ca. 2 mmol) in acetonitila (100 mL). In a volumetric flask, 1 mL of the DNPH solution, 4 mL of the sample and 50 lL of H3PO4 1.0 mol L)1 were introduced, consecutively. The yielding solution was stirred at room temperature for 30 min Samples derivatives were filtered through a 0.45 lm filter from Millipore and injection of 20 lL was carried out into the HPLC system.
Quantitative Analysis A quantitative conversion of aldehydes and ketones in fuel ethanol samples to their derivative form was guaranted by using a large excess (1.0 · 10)2 mol L)1) of DNPH [21]. The analytical curves for aldehydes-DNPH and ketones-DNPH were obtained by linear regression, plotting the peak area vs. concentrations.
Analysis of Fuel Ethanol Samples Different commercially available fuel alcohol samples (ten in total) were collected from local gas station in Araraquara, Sa˜o Paulo, Brazil. These samples were storaged in pyrex glass flasks at controlled room temperature at 20 C, and measured recently. The aldehydes and ketones contents in the samples were derivatized as described above and the analyses carried out on HPLC with electrochemical detection in triplicate.
Chromatographia 2006, 63, January (No. 1/2)
Original
Results and Discussion
Original
150
Ι/µΑ
Ι/µΑ
c
b
150
100
100
50
50
a
0 0.0
0.2
0.4
0.6
0.8
1.0
a
0 1.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
E /V vs. Ag/AgCl
E /V vs. Ag/AgCl
200
200
d
150
150
e
Ι/µΑ
Ι/µΑ
100
100
50 50 0
a
a
0 0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
E /V vs. Ag/AgCl
E /V vs. Ag/AgCl
200
200
150
150
f
Ι/µΑ
In order to establish the best potential to be used as suitable amperometric detection for analysis of aldehydes and ketones derivatives, preliminaries studies were undertaken using cyclic voltammetry on glassy carbon electrode. Fig. 1, exhibits the the voltammograms obtained at scan rate 50 m V s)1 for oxidation of 1.0 · 10)3 mol L)1 of carbonyl compounds derivatized with DNPH in methanol contend LiClO4 0.1 mol L)1. They presented well-defined anodic peak at potential of +0.94 V for the 5-HMF (Fig. 1c); +0.99 V for 2-furfuraldehyde (Fig. 1d); +1.29 V for butyraldehyde (Fig. 1e); +1.15 V for acetone (Fig. 1f) and +1.18 V for methyl ethyl ketone (Fig. 1g), vs Ag/AgClsat which can be the basis for its electroanalytical analysis. The DNPH used as derivatizing reagent is also oxidized at +0.73 V (Fig. 1b), but this value do not interfere significantly with the peak potentials of their derivatized compounds. There is no reduction wave on the reverse scan; at any scan rate investigated from 10–500 mV s)1, suggesting a mechanism with characteristics of an irreversible process. The anodic peak currents values for all species investigated showed a linear increase with the square root of the scan rate (10–500 mV s)1), following equations: ip(lA) = )0.66 + 25.63 x; r = 0.9997 for 5-HMF; ip(lA) = )0.10 + 3.37x; r = 0.9992 for 2-furfuraldehyde; ip(lA) = )0.70 + 20.11x; r = 0.9998 for butyraldehyde; ip(lA) = )0.73 + 8.72x; r = 0.9998 for acetone and ip(lA) = 0.62 + 12.34x; r = 0.9998 for ethyl methyl ketone; where x = m1/2 and m = mV s)1, indicating that the electrochemical oxidation of aldehydes-DNPH and ketonesDNPH on glassy carbon electrode is controlled by diffusion process. The results obtained are in agreement with the literature [22], which have proposed that the electrochemical oxidation of ketones-DNPH, occurs via two electron transfer to the -C=N- bond generated in the derivative and subsequent cleavage after loss of 2H+. The oxidation of DNPH reagent occurs at less positive potential due oxidation of the primary amine present as substituent in the
200
Ι/µΑ
Electrochemical Behavior of Aldehydes and Ketones2,4-Dinitrophenylhydrazone
200
100
50
g 100
50 a
0
a 0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
E /V vs. Ag/AgCl
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
E /V vs. Ag/AgCl
Fig. 1. Cyclic voltammograms obtained for oxidation of 1.0 · 10)3 mol L)1 aldehydes-DNPH and ketones-DNPH on glassy carbon electrode in methanol / LiClO4 0.1 mol L)1 (80:20 v/v). a) supporting electrolyte; b) DNPH; c) 5-HMF; d) 2-furfuraldehyde; e) butyraldehyde; f) acetone; g) methyl ethyl ketone. Scan rate= 50 mV s)1
derivatizing reactant, as known in the literature [20]. These results indicate that both derivatizing reagent (DNPH) and the aldehydes-DNPH and ketonesDNPH could be monitored by HPLCED under amperometric conditions using their oxidation on glassy carbon electrode. The chromatographic parameters were investigated and reported below.
Optimization of Mobile Phases The composition and concentration of the supporting electrolyte constitute can promotes marked influence in the response of electrochemical detectior [23– 25]. For this, several supporting electrolyte were tested and best peak-shape,
Chromatographia 2006, 63, January (No. 1/2)
lower detection limits and maximum sensibility peak areas were obtained for aldehydes-DNPH and ketones-DNPH derivatives when methanol/LiClO4 aqueous solution at concentration of 1.0 · 10)3 mol L)1 were used. This experimental condition was utilized in all further experiments. The first chromatograms recorded at isocratic condition for separation and identification of aldehydes and ketones DNPH derivatives using 75:25 v/v methanol / LiCLO4(aq) (1.0 · 10)3 mol L)1) [26], presented poor resolution. So, other conditions were investigated with the aim to optimize the influence of the mobile phase. News proportions for methanol/ LiClO4(aq) (1.0 · 10)3 mol L)1) as mobile phase changing from: 70:30 v/v; 80:20 v/v e
47
Table 1. Retention times (tR), retention factor (k) and resolutions (Rs) of aldehydes and ketones DNPH derivatives as a function of methanol:LiClO4(aq) mobile phase composition Peaks
1 2 3 4 5
70:30
75:25
80:20
85:15
tR
k
Rs
tR
k
Rs
tR
k
Rs
tR
k
Rs
9.0 16.1 16.7 27.2 28.1
8.0 15.2 15.5 26.2 27.2
– 12.0 0.7 10.0 0.8
7.8 11.1 11.7 17.2 18.0
5.8 10.1 10.8 16.2 17.0
– 6.6 1.1 6.4 1.0
6.2 9.5 10.2 13.3 13.9
5.7 8.5 9.2 12.1 12.8
– 4.8 1.3 4.5 1.1
4.6 6.1 6.6 8.0 8.4
3.6 5.1 5.6 6.8 7.0
– 4.6 1.2 3.5 0.9
1 = 5-HMF; 2 = 2-furfuraldehyde; 3 = acetone; 4 = butyraldehyde; 5 = methyl ethyl ketone
85:15 v/v were tested. Using some chromatographic parameters as retention time (tR), retention factor (k) and resolution among peaks (Rs), it was concluded that the best condition for separation of the investigated compounds were obtained 80:20 v/v methanol / LiClO4(aq) (1.0 · 10)3 mol L)1), but the resolution among peaks present values inferior than 1.5 (Table 1). Taking into consideration that the desirable resolution should be superior than 1.5 new compositions were investigated. Thus, mobile phase composition using a ternary mixture, of methanol/ acetonitrile / LiClO4(aq) (1.0 · 10)3 mol L)1) was studied, in order to define better resolution among the carbonilic derivatives. Although the use of ternary composition have shown more efficient separation for aldehydes-DNPH and ketones-DNPH, the separation times exceed 50 min, which is considered not recommended as analytical methodology. Thus, the optimum mobile phase chosen for analysis was methanol / LiClO4(aq) (1.0 · 10)3 mol L)1) in proportion, of 80:20 v/v, where chromatograms presented better separation for aldehydes and ketones-DNPH derivatives and analyses at times inferior than 15 min.
Effect of Mobile Phase Flow-Rate In order to improve the peaks resolution in the chromatogram of aldehydes and ketones-DNPH compounds, the influence of mobile phase flow-rate through the column was investigated comparing chromatograms recorded from 0.6 to 2.2 mL min)1. The flow rate can be an important parameter to optimize the analysis, since the flowrate variation leads to significant change in the column efficiency, thereby interfering in the number of theoretical plates (N) in the column. The results obtained indicate that an upper limit of flow-rate of 2.2 mL min)1 can be used, since higher flow-rates cause excessive column pressure
48
usually limited to 4000 psi. Flow-rates lesser than 0.6 mL min)1 require separation times longer than 40 min, which is considered unviable for this methodology. Nevertheless, it is important to clarify that the transportation of the analytes to the working electrode surface in the electrochemical detector occurs by convection, where therefore it well undergo electrochemical oxidation. An increase of the flow-rate will result in an increase in peak intensity on the corresponding chromatograms, but the oxidized organic compounds can usually be adsorbed at electrode surfaces. Therefore, the adsorbed material can also leads to the consequent reduction of the electrode sensibility [27] attributed due passivation effects on the electrode surface that have cumulative effect. So the increasing in flow-rate of the analyte can also promotes the corresponding increases in the amount of adsorbed material, since more analyte is reaching the electrode surface. In order to avoid these anomalies, the chromatograms were recorded using a careful polish of the working electrode surface before each increase in the flowrate, through clean electrochemistry program set changing from +1.0 V to -1.0 V during 10 s. Using this experimental condition, a flow-rate of 1.1 mL min)1 can be considered as the optimum value for the resolutions of aldehydes and ketones-DNPH derivatives in the chromatogram with excelent repeatability.
Electrochemical Detection of Aldehydes and Ketones2,4-Dinitrophenylhydrazone Electrochemical detector based on controlled potential amperometry, where the current is measured as a function of time using a controlled applied potential, have been the most useful kind of detector coupled to HPLC. Usually, the applied detector potential corresponds to the
minimum potential at which, the current reaches its limiting current plateau; assigned at Eplateau. By operating at Eplateau, maximum current response of the analyte is obtained at any time. This is especially important in trace analysis, when the system has to be optimized to guarantee higher current response for low concentrations of the analyte. The effect of the applied potential from +0.3 V to +1.4 V on the peak intensity of the chromatograms recorded for aldehydes and ketones-DNPH derivatives was investigated using methanol/ LiClO4(aq) (1,0 · 10)3 mol L)1) in proportion 80:20 v/v as mobile phase and flow-rate of l.1 mL min)1. Fig. 2, shows to the yielding curves of hydrodynamic voltammograms in function of applied potential for aldehydes and ketones DNPH derivatives investigated. The analysis of Fig. 2, indicates that higher peak intensities are observed for applied potential (Eoxd) from +1.0 to +1.3 V; therefore, values applied potential of +1.0 V was chosen as optimum to detect all derivatives which can promote most selective response and high sensitivity using the proposed method. Fig. 3 shows a typical chromatogram for standard solutions of 300 ng mL)1 of aldehydes-DNPH and ketones-DNPH via electrochemical detection using the best conditions, previously: flow-rate of 1.1 mL min)1, methanol/LiClO4(aq) (80:20 v/v), Eoxd = +1.0 V. The elutions of the examined compounds were completed in a time of 13 min during a chromatographic run. Peaks identifications were based on the retention times, which were confirmed by spiking authentic standard solutions of aldehydes-DNPH and ketones-DNPH, flowing the sequence: 5-HMF tR = 5.8 min; 2-furfuraldehyde tR = 8.6 min; acetone tR = 9.3 min; butyraldehyde tR = 12.3 min and methyl ethyl ketone tR = 13.0 min.
Analytical Curve The analytical curves for HPLC-ED based on the relation of peak area and concentration were preferred instead of peak height for aldehydes and ketonesDNPH in order to eliminate possible effects of adsorbed species on the electrode surface, which could prejudice the quantitative response [28]. The analytical curves for each aldehydes-DNPH and
Chromatographia 2006, 63, January (No. 1/2)
Original
Determination of Aldehydes and Ketones in Fuel Ethanol Samples Using the best experimental conditions previously defined, ten commercial samples of fuel ethanol collected from different gas stations of Araraquara city, were treated with DNPH, as described in the experimental section, and aliquots of 20 lL were analysed by HPLC-ED operating at +1.0 V. The Fig. 4 shows a characteristic chromatographic separation obtained for one typical commercial fuel ethanol sample, submitted to oxidation on glassy carbon electrode used as detector. The presence of a relatively high concentration of DNPH (peak 1) did not produce any visible effects on the signals of the aldehydes and ketones product generated after the reaction, which were sufficiently separated from each other. The peak 2 in Fig. 4, which high intensity is assigned as the acetaldehyde component as majority contaminant in fuel
Original
20
I /nA
15
10
5
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
E /V vs. EAg/AgCl Fig. 2. Hydrodynamic voltammograms of the standard aldehydes-DNPH and ketones-DNPH 300 ng mL)1 solution at a glassy carbon electrode. Mobile phase containing methanol/ LiClO4(aq)1.0 · 10)3 mol L)1 (80:20 v/v). Identification of graphics: —j— 5-HMH; —d—2furfuraldehyde; —m— butyraldehyde; —.— acetone; —r— methyl ethyl ketone
2
1
250 200 nA
ketones-DNPH were constructed by plotting the peak area against the concentration and a linear range was obtained from 5.0 to 400.0 ng mL)1, which parameters are shown in Table 2. The correlation coefficients obtained are close to unity (Table 2), suggesting that all the aldehyde and ketone can be easily determined from long interval of concentration with good sensibility. The detection limit (LOD) expressed as the smallest level of analyte that gives a measurable response, were evaluated by statistics treatment, as the 3:1 signal-tonoise (S/N) ratio [26]. The limit of quantitation (LOQ) was also estimated as S/N equal to 10:1 and can be defined as the smallest concentration of analyte, which gives a response that can be accurately quantified. All values of LOD and LOQ determined for aldehydes-DNPH and ketones-DNPH investigated using the electrochemical detection is also reported on Table 2. The results indicate that all the aldehydes and ketones-DNPH derivatives can be determined al low levels using the proposed method, with high repeatability with relative standard derivation lower than 4.0 to 2.0% for all measures (n = 6). The proposed method was then employed for determination of aldehydes and ketones in fuel ethanol samples.
150
4
3
5
100 50 0 2.5
5.0
7.5 Time (min)
10.0
12.5
Fig. 3. HPLC chromatogram of the 2,4-dinitrophenylhydrazones of standard aldehydes and ketones 300 ng mL)1 solution. Mobile phase: methanol / LiClO4(aq) 1.0 · 10)3 mol L)1 (80:20 v/v) with a flow-rate of 1.1 ml min)1.Electrochemical detection was at +1.0 V. Identification of peaks: 1 = 5-hydroxymethylfurfural; 2 = 2-furfuraldehyde: 3 = acetone: 4 = butyraldehyde and 5 = methyl ethyl ketone Table 2. Analytical parameters (y = a + b x) for HPLC determination of aldehyde and ketones 2,4-dinitrophenylhydrazones and limits of detection and quantification obtained for the proposed methodology DNPHs
a
b
5-HMF 2-furfuraldehyde Butyraldehyde Acetone Methyl ethyl ketone
)4495.26 )2744.12 )3751.45 2608.95 2667.89
4.50 4.45 2,59 2.04 2.69
· · · · ·
1012 1012 1012 1012 1012
r2
LOD (ng mL)1)
LOQ (ng mL)1)
0.9997 0.9999 0.9999 0.9998 0.9999
1.8 1.7 1.9 2.0 1.8
6.2 6.0 5.7 5.0 5.3
y = peak area; x = concentration; a = intercept; b = slope; r2 = correlation coefficient
ethanol sample, which were investigated previously by the authors [4]. Although, the 5-HMF and 2-furfuraldehyde have not quantified at any analysed fuel ethanol sample, the chromatograms
Chromatographia 2006, 63, January (No. 1/2)
indicates the occurrence of all other contaminants in the analysed sample, which are listed on Table 3. In Fig. 4 is possible to see the presence of acetone (peak 3); butyraldehyde (peak 4), and
49
Table 3. Concentration of aldehydes and ketone determined in fuel ehanol samples and analytical recoveries added to its samplesa Sample
1 2 3 4 5 6 7 8 9 10
Butyraldehyde
Acetone
Conc.Pres. ng mL)1
Conc.Add. ng mL)1
133.2 136.3 124.0 164.0 135.2 135.7 111.8 130.3 135.8 112.6
38.0 38.0 38.0 38.0 38.0 38.0 38.0 38.0 38.0 38.0
Recovery (%) 98±2 99±3 100±1 101±2 98±4 97±3 99±5 100±2 100±2 99±3
Methyl ethyl ketone
Conc.Pres. ng mL)1
Conc.Add. ng mL)1
143.0 144.0 137.0 171.5 147.2 150.0 129.0 159.0 166.3 135.8
40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0
Recovery (%) 100±5 103±2 99±3 97±3 95±2 100±2 102±1 100±5 103±3 97±2
Conc.Pres. ng mL)1
Conc.Add. ng mL)1
113.9 125.4 112.3 161.4 128.0 128.9 128.2 132.0 115.3 80.7
40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0
Recovery (%) 97±3 100±1 99±4 95±1 99±2 100±3 98±2 100±4 104±5 99±3
a
n =5, injections were performed in triplicate Conc. Pres. = concentration present; Conc. Add. = concentration added
1
2
75
nA
3 50
4 5
25
0
2.5
5.0
7.5
10.0
12.5
Time (min)
Acknowledgments
Fig. 4. HPLC chromatogram of the aldehydes and ketones 2,4-dinitrophenylhydrazones in fuel ethanol sample. Mobile phase: methanol / LiClO4(aq) 1.0·10)3 mol L)1 (80:20 v/v) with a flow-rate of 1.1 mL min)1. Electrochemical detection was at a glassy carbon electrode set at +1.0 V. Identification of peaks: 1 = DNPH; 2 = acetaldehyde; 3 = acetone: 4 = butyraldehyde and 5 = methyl ethyl ketone
methyl ethyl ketone (peak 5), as contaminants. They are identified in all fuel ethanol samples in a concentration range from 81–171 ng mL)1. The recovery values were evaluated for aldehydes-DNPH and ketonesDNPH using fuel ethanol samples spiked with a standard derivative solution of carbonyl compounds derivatized at a level of around 25% of the measured content and performing five assays after each addition (Table 3). The concentrations of aldehydes-DNPH and ketones-DNPH were calculated using the area of the peaks, by linear regression approach using the method of standard addition. The coefficient variation (six repetition) indicates values from 0.8 to 2.0%, suggesting that the HPLC-ED using DNPH derivatization can be a good option to
50
contents in fuel ethanol samples at low concentration without any elaborated pre-treatment. These results are important to control the alcohol oxidation process, since different aldehydes contents can be related to unsuitable condition of fermentation, transport and storage of fuel. In summary, our findings report a rapid, simple, selective, sensitive and reproducible method for quantification and identification of aldehydes and ketones in fuel ethanol samples, which represents an important economic aspect in country in development as Brazil.
determine aldehydes and ketones with good accuracy.
We gratefully acknowledge the financial support and fellowships provided by Financiadora de Estudos e Projetos (FINEP), Plano Nacional de Cieˆncia e Tecnologia do Setor Petro´leo e Ga´s Natural (CTPetro), Ageˆncia Nacional do Petro´leo (ANP) and Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP).
Conclusions The analytical procedure targeting the detection of selected organic contaminants in fuel ethanol sample. HPLC-ED presents an excellent technique to determine butyraldehyde, acetone and methyl ethyl ketone with good repeatability and accuracy. Although, the contaminants are present at relative high concentration the electrochemical detector has shown the particular advantage of higher sensibility than other methods could be very useful matrix with low levels of organics contamination. The system was suitable to determine aldehydes and ketones
References 1. www.agricultura.gov.br accessed on 10th December 2004 2. de Oliveira MF, Saczk AA, Okumura LL, Fernandes AP, Moraes M, Stradiotto NR (2004) Anal Bioanal Chem 380:135–140 3 de Oliveira MF, Saczk AA, Okumura LL, Stradiotto NR (2002) Ecl Quı´ m 27:153–160 4. Okumura LL, Saczk AA, de Oliveira MF, Zanoni MVB, Stradiotto NR (2005) Anal Sci 21:441–444 5. Okumura LL, Saczk AA, de Oliveira MF, Zanoni MVB, Stradiotto NR (2005) Anal Bioanal Chem 381:1619–1624 6. Carbonel V, Salvador A, de La Gardia M (1992) Fresenius J Anal Chem 355:529– 532.
Chromatographia 2006, 63, January (No. 1/2)
Original
7. Padilha PM, Padilha CCF, Rocha JC (1999) Anal Chem 18:299–303 8. Padilha PM, Gomes LA, Moreira JC, Filho NLD, Gushiken Y (1998) J Bras Chem Soc 9:494–498 9. Wardencki W, Orlita J, Namiesnik J. (2001) Fresenius J Anal Chem 369:661–665 10. Matthews TG, Howell TC (1981) J Air Pollut Control Assoc 31:1181–1184 11. Wells RJ (1999) J Chromatogr A 843:1–18 12. Rosenfeld JM (1999) J Chromatogr A 843:19–27 13. Vogel M, Bu¨ A, Karst U (2000) Fresenius J Anal Chem 366:781–791 14. Kiss G, Varga B, Varga Puchony Z, Gelencsar A, Kriva´csy Z, Hlavay J (1999) Talanta 48:755–762 15. Garcı´ a-Alonso S, Pe´rez-Pastor RM (1998) Anal Chim Acta 367:93–99
Original
16. Stashenko EE, Ferreira MC, Sequeda LG, Martı´ nez JR, Wong JW (1997) J Chromatogr A 779:360–369 17. Kissinger P T, Jacobs WA (1982) J Liq Chromatogr 5:669–676 18. Possanzini M, Dipalo V (1995) Chromatographia 40:134–138 19. Lehmpuhl DW, Birks JW (1996) J Chromatogr A 740:71–81 20. Shriner RC, Fugon RC, Curtin DY, Morril TC (1980) In: The Systematic Identification of Organic Compounds, (eds) John Wiley & Sons, New York 21. Nascimento RF, Marques JC, Lima Neto BS, Keukeleire DD, Franco DW (1997) J Chromatogr A 782:13–23 22. En-Chu L, Van de Mark MR (1982) J Chem Soc Chem Comm 13:1175–1178
Chromatographia 2006, 63, January (No. 1/2)
23. Lee HS, Choi S J, Lee H M, Jeong CK, Kim SB, Lee JT, Yoo SD, DeLuca. PP, Lee KC (1999) Chromatographia 50:701– 704 24. Meriluto J, Kincaid B, Smyth MR, Wasberg M (1998) J Chromatogr A 810:226230 25. Fleming LH, Reynolds NC (1984) J Liq Chromatogr 7:793–8-8 26. Coutrim MX, Nakamura LA, Collins CH (1993) Chromatographia 37:185–190 27. Snyder LR, Kirkland JJ, Glajch JL (1997) In: Practical HPLC Method Development, Second edition, (eds) John Wiley & Sons, New York 28. Albery WJ, Beck TW, Brooks WN, Fiellenz M (1981) J Electroanal Chem 125:205– 217
51
