ABSTRACT
In classical calibration, the statistically uncertain variable y is regressed on the error-free variable x for a number of known samples, and the results are used to estimate the x value (x0) for an unknown sample from its measured y value (y0). It has long been known that inverse calibration--regression of x on y for the same data--is more efficient in its prediction of x0 from y0 than the seemingly more appropriate classical procedure, over large ranges of the controlled variable x. In the present work, theoretical expressions and Monte Carlo calculations are used to illustrate that the comparison favors the inverse procedure even more for small calibration data sets than for the large sets that have been emphasized in previous studies.
ABSTRACT
The D' (2 (3)P2) left and right arrow A'(2 (3)Pi) transition in ClF is studied in emission, in fluorescence, and through OODR excitation experiments. The emission spectrum is excited in a flowing Tesla discharge of ClF in half an atmosphere of Ar and is photographed at high resolution. The excitation and fluorescence spectra are obtained by pumping through perturbed levels of the B (0(+) 3Pi) valence state. The emission data yield a detailed characterization of v' levels 0 and 1 and v" levels 3-8. The excitation spectra extend the v' range to v' = 16 and put both states on an absolute energy scale. The low-resolution fluorescence spectra span v" = 1-17, the upper limit of which lies within approximately 20 cm-1 of dissociation. Key spectroscopic parameters (cm-1) from the study, valid for 35Cl19F, are: T'e = 55 253, omega'e = 365.22, omegaex'e = 2.005, T"e = 18 257, omega"e = 363.53, omegaex"e = 8.30, B'e = 0.21790, a'e = 0.00175, B"e = 0.33412, a"e = 0.00631, R'e = 2.5069 Å, R"e = 2.024 Å. A near-dissociation fit of the D' --> A' emission and fluorescence data locates the first dissociation limit at 21 495(5) cm-1, which is lower than the accepted value, but by just one standard error. Copyright 1999 Academic Press.
ABSTRACT
The A(3Pi1) state of ClF is studied (1) in emission from the beta(1(3)P2) ion-pair state in a high-pressure discharge, (2) from its role in perturbations of the B(3Pi0(+)) and A'(3Pi2) states, and (3) in direct absorption from the X(1Sigma+) ground state. The last of these is obtained via a UV-labeling technique and represents the first observation of the A <-- X transition. The emission spectrum yields extensive assignments for the 0-4, 0-5, and 1-4 bands of beta --> A, with both e and f lines identified. The well-known perturbation in v = 7 of the B state is attributed to interaction with v = 9 of A. A newly identified perturbation in B(6) involves coupling with A(8). The known perturbation in B(8) may involve A(11), but other assignments cannot be ruled out. Approximate potential curves are derived for all three valence states and used to compute the vibrational factors for the perturbations, permitting extraction of the electronic coupling strengths. The B-A coupling elements are within approximately 10% of expectations for rotation-electronic coupling within a Hund's case (a) 3Pi manifold, but the strength of the perturbation between A(5) and A'(6) is only about 30% of predictions for this model. Copyright 1999 Academic Press.
ABSTRACT
The intensity calibration of photographic emulsions is examined from the standpoint of nonlinear leastsquares fits to analytic expressions. A simple but widely neglected expression for computing the propagated statistical error is used to examine the error in the relative exposure as determined from the fitted calibration function. The statistical error is also compared with the dependence on the choice of fit function, or model error. In the present applications, which replicate those obtainable routinely in emission spectroscopy, the relative intensity is obtainable to better than 10% over a 2-order-of-magnitude change in intensity.
ABSTRACT
A method for studying the pressure and temperature dependence of rare gas halide and halogen emissions is described. The tesla discharge utilized in this method is characterized in detail with respect to its temporal behavior and its radiant output. The light is found to be produced in moderately intense pulses as short as 0.5 micros in duration. Average dc radiant powers for strong emitters can exceed 5 mW. The method appears suitable for the study of most emitters that can be produced by electron beam excitation.
