Linear Fitting Analysis Research Articles

Quantification and Identification of Components in Solution Mixtures from 1D Proton NMR Spectra Using Singular Value Decomposition

One-dimensional proton NMR spectra of complex solutions provide rich molecular information, but limited chemical shift dispersion creates peak overlap that often leads to difficulty in peak identification and analyte quantification. Modern high-field NMR spectrometers provide high digital resolution with improved peak dispersion. We took advantage of these spectral qualities and developed a quantification method based on linear least-squares fitting using singular value decomposition (SVD). The linear least-squares fitting of a mixture spectrum was performed on the basis of reference spectra from individual small-molecule analytes. Each spectrum contained an internal quantitative reference (e.g., DSS-d6 or other suitable small molecules) by which the intensity of the spectrum was scaled. Normalization of the spectrum facilitated quantification based on peak intensity using linear least-squares fitting analysis. This methodology provided quantification of individual analytes as well as chemical identification. The analysis of small-molecule analytes over a wide concentration range indicated the accuracy and reproducibility of the SVD-based quantification. To account for the contribution from residual protein, lipid or polysaccharide in solution, a reference spectrum showing the macromolecules or aggregates was obtained using a diffusion-edited 1D proton NMR analysis. We demonstrated this approach with a mixture of small-molecule analytes in the presence of macromolecules (e.g., protein). The results suggested that this approach should be applicable to the quantification and identification of small-molecule analytes in complex biological samples.

Measurement of Atmospheric Formaldehyde and Monoaromatic Hydrocarbons using Differential Optical Absorption Spectroscopy during Winter and Summer Intensive Periods in Seoul, Korea

A Long-Path DOAS (LP-DOAS) developed by GIST was used to measure simultaneously atmospheric formaldehyde, monoaromatic hydrocarbons, and other trace gases over 740 and 2000 m beam paths at an urban site in Seoul, Korea during two field campaigns in February and August 2003. The mean concentrations of formaldehyde, benzene, toluene, m-xylene and p-xylene were measured to be 6.32 (± 2.41), 0.78 (± 0.32), 3.32 (± 1.87), 0.52 (± 0.24), and 0.38 (± 0.18) ppbv during the two measurement periods. Based on the analysis of the ratio of the OH⋅ -NO2 to OH⋅ -VOCs rate constants, it was possible to draw a conclusion that OH⋅ reacted preferentially with NO2 during the two measurement periods. Diurnal variation of formaldehyde was not typical of photochemically generated components during the two measurement periods in Seoul. Results from the linear regression fit analysis and diurnal variation of the measured species support that formaldehyde measured could be from primary sources (vehicle exhaust and heating burners) in winter and from primary and secondary sources (photochemical process) in summer. Monoaromatic hydrocarbons were likely emitted from the primary sources.

Study of acid hydrolysis of bromazepam

Quantitative study of an acid hydrolysis of bromazepam, which in acidic solution simultaneously undergoes reversible 4,5-azomethine bond cleavage and protolytic reactions, was performed (t = 25 °C). When an equilibrium is established, four species are simultaneously present in solution. Two approaches, both based on theoretically derived dependence between the absorbance (A) and the solution acidity ([H3O+]), were used to determine appropriate equilibrium constants. In the first, the least-squares method was used to fit experimentally obtained (A, [H3O+]) value pairs to a derived function. With values obtained as fitting parameters, the probability distribution for equilibrium constants was calculated with the virtual experiments performed using Monte Carlo simulations. In the second one, spectrophotometric analysis indicated an existence of two distinct pH intervals, both with three dominant species involved. Spectral data were used to calculate equilibrium constants by a linear curve fitting analysis. Values obtained for equilibrium constants from the numerical and the experimental approaches are in good accordance and were used to calculate distribution of all the species as a function of pcH. Key words: bromazepam, acid hydrolysis, equilibrium constants.

Application of target factor analysis and linear least squares fitting to extracting chemical information from Auger depth profiles of a Mo/Si thin multilayer system

Linear least square fitting (LLS) analysis was applied to Auger depth profile data to verify the chemical information which was extracted by target factor analysis (TFA) from a Si(2.5 nm)/Mo(5 nm) thin multilayer sample. The chemical state assignment was confirmed by measurement with external standard samples and each spectrum was used as the basis spectra for LLS. The profiles from the Si LMM and Mo MNN peaks constructed by LLS showed the clear existence of silicide layers with different thicknesses less than 1 nm at each interface, which was confirmed by TFA and transmission electron microscopy cross section. The usefulness and problems of combining LLS with TFA for characterization of this kind of complicated multilayer sample are discussed.

Initial clinical experience with a new dual sensor SSIR pacemaker controlled by body activity and minute ventilation.

Fourteen patients were implanted with a single chamber dual sensor pacemaker (Legend Plus) that measures minute ventilation (VE) via variations in impedance between a bipolar lead and the pacemaker case, and activity via a piezoelectric crystal bonded to the pacemaker case. Chronotropic incompetent patients were exercised on a treadmill and a bicycle in dual sensor mode. Activity only indicated pacing rate was measured using a strap-on pacemaker. Both implanted and strap-on pacemakers were adjusted to yield a steady-state pacing rate of 100 beats/min during hall walk. Pacing rate, VE, and oxygen uptake (VO2) were measured continuously. Linear curve fit analysis slopes for plots of VE versus pacing rate during exercise (1.33-1.49) compared favorably to values reported in normals. Peak pacing rates achieved for treadmill and bicycle testing for dual sensor mode were higher than activity mode alone. Slopes of heart rate to VE or VO2 were not significantly different (P < 0.05) for dual sensor mode in contrast to activity alone. In conclusion, the Legend Plus dual sensor rate adaptive pacing therapy delivered pacing rates more proportional to VE and VO2 under different types of exercise than rates indicated by a strap-on pacemaker in activity mode.

ENHANCING THE INTERPRETATION OF X-RAY PHOTOELECTRON AND AUGER ELECTRON SPECTRA USING NUMERICAL METHODS

X-ray Photoelectron Spectroscopy (XPS) and Auger Electron Spectroscopy (AES) have long been used as a surface analytical methods for the determination of composition and chemistry of the elements in a material. Peak overlap or subtle changes in chemistry may make the chemical state analysis of the elements in a material difficult. By using a variety of numerical methods, including linear least squares fitting and target factor analysis, it is possible to enhance the analysis and interpretation of surface analysis data