rollefson_1940#
R. Rollefson and R. Havens, Phys. Rev. 57, 710 (1940)
Source record · Catalogue · Review register
Source kind: measurement
Recorded technique: Prism method: the change of deviation produced by a change of pressure of the gas in a hollow prism, in the infrared spectrometer of Rollefson and Rollefson’s HCl work (Phys. Rev. 48, 779 (1935)), with a monochromator and a gas spectrometer. The slit image is located by a new compensated thermopile of two 10 mm x 1 mm receivers connected in opposition, on a comparator carriage: ‘successive determinations of the image position would check to within 0.3 micron’ in the high-energy regions, and ‘for all the wave-lengths less than 11 u, variations of more than 2 microns were rare’ (p. 710). Methane from sodium acetate and soda lime, purified by repeated distillation; runs with methane from a modified Grignard process ‘agreed with the old to within 0.5 x 10^-6 for the absolute value of n’ (p. 711). Absorption spectra were taken throughout to avoid wavelengths where the gas absorbs; the 2.3 um overtone region is left out as not quantitative (p. 712). Changing the slit widths ‘had no effect’ (p. 712). The measurement pressures and temperature are not stated.
Materials in structured entries: CH4
Dataset quantities: refractivity
Contents: 3 datasets; 0 formulas.
Curation notes#
State: ‘the values of the index of refraction which were obtained were reduced to standard conditions (0 degrees, 760 mm) by means of coefficient of expansion and pv data given in the International Critical Tables (alpha = 0.003683; the pv correction was at most about 1/4 percent)’ (p. 712): a real-gas reduction, state 1 atm, 0 degrees C, density real. alpha = 0.003683 is methane’s real expansion coefficient (the ideal gas’s is 0.003661).
Values: Table I (p. 713), ‘averages for several runs’; runs that did not cover the whole range were shifted to agree in absolute value with the average of the runs that did; ‘the largest shift was 1.1 x 10^-6 in n, and most of the curves were not shifted at all’ (p. 712). Wavelengths in microns as printed, to 0.01-0.001 um; the paper states neither their medium nor their accuracy (the prism’s calibration is not described). Taken as vacuum wavelengths: an air wavelength differs by 2.7e-4 relative, which on the authors’ own two-band curve moves no value by more than 0.08 x 10^-6 (at 8.36 um, on the 7.7 um band’s wing), 44 of the 49 by less than 0.05.
Uncertainty: not stated as such. The per-point uncertainty given is the rms of the paper’s own (n - 1)_EXP - (n - 1)_CALC over the 38 rows below 11 um whose band constants are not in parentheses, 0.28 x 10^-6 (the paper’s scatter about its two-band model; largest -0.8 at 8.95 um). The eleven rows whose band constants are parenthesised, ‘very near the absorption bands’ (3.000, 6.790, 6.940, 7.000 um), where ‘the spread of the band due to rotation begins to have an appreciable effect’, and beyond 11 um, where ‘the energy available in this region is very small and the points consequently scatter rather badly’ (pp. 713-714), scatter by 1.7 x 10^-6 rms about the model (up to 4.3 at 14.00 um): they carry 0.4 % more (relative_rows). The authors’ own statement on the level: their values lie 4 x 10^-6, ‘or about 1 percent in the absolute value of n - 1’, above the extrapolation of Friberg’s visible and ultraviolet data, and ‘even if the 1 percent difference which remains were due entirely to systematic error in the present experiments or impurity of the gas used’ (pp. 711-712); the same apparatus gave HCl and CO within 1/4 percent of the Cuthbertsons’ extrapolated values (p. 711). The scale uncertainty is that 1 %, taken as one sigma (systematic_relative); purity was checked by vapour pressure, fractional distillation and absorption spectra, ‘corroborative evidence and not conclusive proof of purity’ (p. 711). Near the bands the prism method can be biased: ‘unsymmetrical absorption of the slit image causes an apparent shift of the image’ (p. 712).
Derived values, checked (not extracted): - Delta(n - 1) = |EXP - EXT| for all 49 rows, at the printed precision (the extraction script checks it). - The calculated column is (n - 1)_EXT + K3/(nu3^2 - nu^2) + K4/(nu4^2 - nu^2) in units of 10^-6, nu = 1/lambda in units of 1000 cm^-1, with the averages K3 = 16.50 (nu3 = 3019.6 cm^-1) and K4 = 7.50 (nu4 = 1304 cm^-1) (p. 713): reproduced within 0.013 x 10^-6 at the 20 rows printed to 0.01 below 6.6 um, 0.15 at 6.79-7.00 um and 0.10 beyond 8 um (printed to 0.1). Each printed band constant is reproduced as (EXP - EXT - the other band’s term) x |nu_i^2 - nu^2| within 0.10. - K3/nu3^2 = 1.81 and K4/nu4^2 = 4.41, n_inf - 1 = 435.9 + 1.81 + 4.41 = 442.1 and n_inf^2 - 1 = 884.4 x 10^-6, and the 12.4 x 10^-6 or 16 percent of the gap to Sanger’s epsilon - 1 = 960 x 10^-6 (pp. 716-717): reproduced. The table’s K is the constant for n - 1; the text’s Delta(n^2 - 1) = K/(nu_i^2 - nu^2) (p. 712) needs 2K, which is what the effective-charge calculation uses: with 2K and nu in s^-1 the printed epsilon_3 = 0.562 x 10^-10 and epsilon_4 = 0.277 x 10^-10 e.s.u. (p. 715) reproduce to 1 %. - The present work’s values quoted against Koch at 6.557 and 8.678 um, 426.35 and 458.10 (p. 712), are Table I interpolated linearly (426.34, 458.11); the differences 31.75 and Koch’s 30.86 reproduce.
Not extracted: the extrapolated column (Friberg, Diss. Lund (1933), quoted by T. Larsen, Z. Phys. 105, 166 (1937); an extrapolated model of other data, adjusted vertically, 435.9 x 10^-6 beyond 7 um), the differences and band constants (derived), Fig. 2 (the same values plotted), n_inf - 1 = 442.1 x 10^-6 (derived from the extrapolation) and Sanger’s epsilon - 1 (no reference or state given). refractiveindex.info’s ‘Rollefson’ entry (data/external) is Table I’s experimental column at 0 degrees C, 760 torr: it agrees at all 49 rows.
Recorded bibliographic identity status: VALIDATED_WITH_WARNINGS. This concerns identification of the source; it does not establish the scientific correctness of its contents.
Datasets#
rollefson_1940/experiment_CH4#
{
"id": "experiment_CH4",
"material": "CH4",
"file": "CH4/rollefson_1940__experiment_CH4.csv",
"table": "tbl_d3b491ff89fa",
"quantity": "refractivity",
"scale": 1e-06,
"abscissa_unit": "um",
"wavelength_medium": "vacuum",
"state": "1atm_0C",
"density": "real",
"use": "primary",
"uncertainty": {
"absolute": 0.28,
"level": "rms",
"systematic_relative": 0.01,
"relative_rows": {
"3.000": 0.004,
"6.790": 0.004,
"6.940": 0.004,
"7.000": 0.004,
"11.46": 0.004,
"11.89": 0.004,
"12.36": 0.004,
"13.20": 0.004,
"13.60": 0.004,
"14.00": 0.004,
"14.8": 0.004
}
}
}
Table I, (n - 1)_EXP x 10^6, 1.68-14.8 um, the average of several runs. Wavelength medium not stated (see the source notes).
rollefson_1940/calc_CH4#
{
"id": "calc_CH4",
"material": "CH4",
"file": "CH4/rollefson_1940__calc_CH4.csv",
"table": "tbl_d3b491ff89fa",
"quantity": "refractivity",
"scale": 1e-06,
"abscissa_unit": "um",
"wavelength_medium": "vacuum",
"state": "1atm_0C",
"density": "real",
"use": "check"
}
Table I, (n - 1)_CALC x 10^6: Friberg’s extrapolation plus the authors’ two infrared bands with the averaged K3 = 16.50 and K4 = 7.50 (reproduced; see the source notes). Not independent of experiment_CH4.
rollefson_1940/koch_1909_CH4#
{
"id": "koch_1909_CH4",
"material": "CH4",
"table": "text",
"quantity": "refractivity",
"scale": 1e-06,
"abscissa_unit": "um",
"wavelength_medium": "vacuum",
"state": "1atm_0C",
"density": "unknown",
"use": "reproduced",
"reproduced_from": [
"J. Koch, Nova Acta Soc. Upsala 2, No. 5 (1909)"
]
}
2 inline points in the source record.
p. 712: Koch’s (n - 1) x 10^6 with an interferometer and residual rays of 6.557 and 8.678 um, either side of the 7.7 um band. Koch’s state and reduction are not given here; Rollefson and Havens’s 0 degrees C, 760 mm is assumed. They lie 1.7 % and 1.8 % below this work’s interpolated values, ‘in the same direction on the absolute values’; the difference between the two wavelengths, 30.86 against 31.75, is ‘satisfactory agreement’. Koch 1909 is not in the library.
Formulas#
No formulas are recorded.
Corrections#
No dataset errata are recorded in this source record.
Formula corrections, where recorded, appear with their formulas above. Curation notes may also discuss corrections or unresolved interpretation. Consult the review register for the checks and decisions.