Factorial design of electrolyte systems for the separation of fatty acids by capillary electrophoresis.

In this work, a capillary zone electrophoretic methodology using UV indirect detection (224 nm) for the analysis of fatty acids (FAs) in saponified oils is proposed. The electrolyte consisted of a 5 mmol l(-1) phosphate buffer, pH 7. containing 4 mmol l(-1) sodium dodecylbenzenesulfonate (SDBS) as chromophore, 4 mmol l(-1) dimethyl-beta-cyclodextrin and 45% acetonitrile (ACN). The composition of the electrolyte was optimized by a 2(3) factorial design with triplicate at the central point. The design established practical concentration boundaries for SDBS and ACN. In a defined concentration range of 2-4 l(-1), SDBS can certainly be used as a chromophore for indirect detection without imparting excessive baseline noise. For ACN, a suitable interval of 45-55% was found to enhance FAs solubilization without overflowing the system with bubble formation and current interruption. Additionally, the design revealed the importance of dimethyl-beta-cyclodextrin in the resolution of difficult pairs and its function as a solubilizing agent for long chain FAs. At the optimized conditions, nine FAs from C10 to C20, including mono- di- and tri-unsaturated C18 fatty acids were baseline separated in less than 10 min. The proposed method was applied to the separation of FAs in edible oils and polyunsaturated fatty acid enriched margarine. Additionally, spectral monitoring at 206 nm was used to confirm peak identity in the samples.

A multivariate statistical analysis of the composition of rainwater near Cubatão, SP, Brazil.

With the aim of estimating the chemical composition of bulk precipitation (wet + dry) in Cubatão, rainwater samples were collected at several localities in the Cubatão region from May 1984 to October 1985. The levels of some inorganic ions (Na+, Ca2+, K+, Mg2+, Cl-, SO4(2-), NH4+, PO4(3-)) were measured by atomic absorption spectrophotometry, turbidimetry and titration analysis. Correlation and Fisher discrimination indices, principal component and varimax loading and scores were determined in the multivariate statistical data treatment. The results showed that ionic concentrations in rainwater from Vila Parisi are significantly higher than those of Cubatão Centro, Santos and Serra do Mar. The ionic compositions of the Vila Parisi rainwater seem to be mainly determined by local anthropogenic activities (industrial pollution). Besides the influence of this factor on water quality, there is also a natural oceanic contribution involving Cl- and Na+ at the other locations.

The infrared fundamental intensities and polar tensor of allene.

The polar tensor of allene was calculated from the infrared fundamental band intensities of C3H4 and C3D4. The ambiguities in the signs of the dipole moment derivatives with respect to their normal coordinates were resolved by comparison of tensor elements with ab initio calculations at the B3LYP, MP2(FC) and CCD(FC) levels with a 6/311 + + G(3d,3p) basis set. The results are similar to those previously obtained by Koga and co-workers except for the choice of an average of two sign combinations for the E symmetry elements. The values of the mean dipole moment derivatives for the sp and sp2 carbon atoms obtained in this work, 0.032 and -0.133 e, respectively, are in good agreement with the CCD(FC)/6-311 + + G(3d,3p), 0.061 and -0.128 e, and MP2(FC)/6-311 + + G(3d,3p), 0.072 and -0.153 e, theoretical results. The mean dipole moment derivatives are shown to be consistent with potential models relating 1s electron ionization energies and atomic charges.

The infrared fundamental intensities and polar tensor of CH3NC.

The molecular force field and polar tensor of methyl isocyanide have been determined from its gas phase vibrational frequencies and infrared intensities. Quantum chemical results from MP2(FC), B3LYP and quadratic configuration interaction calculation including single and double substitutions procedures using a 6-311 + +G(3d,3p) basis set have been used to determine the signs of the dipole moment derivatives with respect to the normal coordinates as well as estimate individual fundamental intensities of the overlapped v1-v5 and v3-v6 band systems. Principal component graphical representations of the A1 and E symmetry polar tensor elements were useful in determining preferred sets of tensor elements. The mean dipole moment derivative (GAPT charge) of the methyl carbon in CH3NC, 0.347 e, is between the corresponding values in CH3CN, 0.110 e, and CH3F, 0.541 e. The mean dipole moment derivatives obtained here indicate the correct 1s methyl carbon ionization energy as 293.35 eV which is 0.98 eV higher than the corresponding ionization energy of the terminal atom.

The infrared vibrational intensities and polar tensors of HFCO and DFCO.

The infrared vibrational intensities of HFCO and DFCO have been calculated at the B3LYP/cc-pVTZ, MP2(FC)/6-311++G(3d,3p) and QCISD(FU)/aug-cc-pVTZ levels. All calculations predict the isotopomers to have identical intensity sums, within about 1 km mol(-1). This is in contrast with experimental intensity sum results reported in the literature. Dipole moment derivative directions calculated by the three methods are in excellent agreement for all in-plane normal coordinates. All the theoretical polar tensor elements are also in good agreement with each other having standard deviations varying between 0.003 and 0.043 e. The oxygen and fluorine atoms have negative mean derivatives (approximately -0.6 e), whereas the carbon mean derivative is very positive (approximately +1.1e) and the hydrogen one is almost zero (approximately +0.03 e). The HFCO theoretical intensity sums calculated by all three methods as well as their carbon and oxygen mean dipole moment derivatives are in good agreement with those estimated from the experimental intensities and atomic mean derivatives of H2CO and F2CO.

The infrared fundamental intensities and polar tensor of CF4.

Atomic polar tensors of carbon tetrafluoride are calculated from experimental fundamental infrared intensities measured by several research groups. Quantum chemical calculations using a 6-311 + + G(3d, 3p) basis set at the Hartree-Fock, Möller-Plesset 2 and Density Functional Theory (B3LYP) levels are used to resolve the sign ambiguities of the dipole moment derivatives. The resulting carbon mean dipole moment derivative, pC = 2.051 e, is in excellent agreement with values estimated by a MP2/6-311 + + G(3d, 3p) theoretical calculation, 2.040 e, and by an empirical electronegativity model, 2.016 e. The pC value determined here is also in excellent agreement with the one obtained from the CF4 1s carbon ionization energy using a simple potential model, 2.059 e. Crawford's G intensity sum rule applied to the fundamental intensities of CH4, CH3F, CH2F2 and CHF3 results in a prediction of a 1249 km mol(-1) intensity sum for CF4 in good agreement with the experimental values of 1328 +/- 37.9, 1208.0 +/- 54.4 and 1194.8 +/- 7.4 km mol(-1) reported in the literature.

The infrared intensities and polar tensors of the fluorochloromethanes.

The polar tensors of CF3Cl, CF2Cl2 and CFCl3 have been calculated using recent measurements of their gas phase infrared fundamental intensities. The polar tensors obtained for CF2Cl2 and CFCl3 are in very good agreement with those obtained previously since the more recent experimental intensity results are in good agreement with those reported earlier. For CF2Cl2 rhoC = +1.626, rhoF = -0.577 and rhoCl = -0.26e whereas rhoC = +1.369, 0.478 and rhoCl = 0.297e for CFCl3. However, two sets of significantly different mean dipole moment derivatives are obtained from the experimentally measured intensities of CF3Cl reported by two different laboratories. On the other hand, the differences in the mean derivatives of these two sets are not large enough so that results from electronegativity models, potential models for core ionization energies and quantum chemical calculations at the Moller-Plesset 2 and B3LYP density functional levels are sufficient to indicate which set is the correct one. As such average values of rhoC = +1.907+/-0.178e, rhoF = -0.590+/-0.056e and rhoCl = -0.139+/-0.013e obtained from both sets of polar tensor elements are recommended for the CF33Cl mean dipole moment derivatives.

Statistical mixture design development of digestion methods for Oyster tissue using inductively coupled plasma optical emission spectrometry for the determination of metallic ions.

The quantitative determination of chemical elements in organic or biological samples is an important analytical problem. Normally the elements to be determined in the organic matrix must be transformed into a simple inorganic form. A digestion method by heating on a block digestor has been developed for the determination of Al, As, Ba, Ca, Cd, Co, Cu, Fe, Mg, Mn, V and Zn in Oyster tissue by ICP OES. A simplex centroid statistical mixture design has been used to study the effects of changing HNO(3), HCl and H(2)O(2) reagent proportions on the digestion of these samples. Response surface and principal component analyses show that the species Ca, Cd, Cu, Fe, Mg, Mn and Zn have very similar analytical tendencies under this experiment. By means of mixture modeling maximum recoveries for these ions were predicted using 19%, 18% and 63% of the HCl, HNO(3) and H(2)O(2) pseudocomponent mixtures, respectively. This corresponds to 21.4%, 30.8% and 47.8% of the HCl, HNO(3) and H(2)O(2) commercial solutions. Furthermore the As, Co and V ions present large recoveries for these mixtures as well. The Al and Ba ion recoveries are seen to be independent of the mixture proportions. The analysis of Oyster tissue reference material (SRM 1566b - NIST) under optimized conditions at the selected wavelengths resulted in ion recoveries between 90% and 100%.

Box-Behnken design: an alternative for the optimization of analytical methods.

The present paper describes fundamentals, advantages and limitations of the Box-Behnken design (BBD) for the optimization of analytical methods. It establishes also a comparison between this design and composite central, three-level full factorial and Doehlert designs. A detailed study on factors and responses involved during the optimization of analytical systems is also presented. Functions developed for calculation of multiple responses are discussed, including the desirability function, which was proposed by Derringer and Suich in 1980. Concept and evaluation of robustness of analytical methods are also discussed. Finally, descriptions of applications of this technique for optimization of analytical methods are presented.