matsuoka_1991#
J. Matsuoka, N. Kitamura, S. Fujinaga, T. Kitaoka and H. Yamashita, “Temperature dependence of refractive index of SiO2 glass”, J. Non-Cryst. Solids 135, 86 (1991)
Source record · Catalogue · Review register
DOI: 10.1016/0022-3093(91)90447-e
Source kind: measurement
Recorded technique: Minimum deviation (Gaertner L-124 refractometer, Hg lamp, photomultiplier) on one prism of type III silica glass (Nippon Sekiei Glass Co., about 1200 ppm OH; square faces of 30 mm, 60 degree vertex) in a cryostat evacuated to 5e-6 Pa, ‘to avoid frosting of the cryostat windows at low temperatures and fluctuations in the light path due to air convection’; the refraction at the cryostat windows ‘was corrected by using the refractive index data of air’ (American Institute of Physics Handbook). Ten Hg lines, 2.27-5.21 eV (546-238 nm), seven temperatures from -165.4 to 83.3 C (p. 86).
Materials in structured entries: SiO2
Dataset quantities: refractive_index
Contents: 7 datasets; 1 formula.
Curation notes#
Journal pages 86-89 are pdf pages 1-4 (a Letter to the Editor). Identity checked against the first page. The sample maker is printed ‘Nippon Sekiei Glass Co.’; leviton_2006/leviton_2008 call it ‘Nippon Seiki Glass Company’.
MEDIUM. Absolute: ‘The absolute refractive indices at ten photon energies and seven temperatures are listed in table 1’ (p. 87); the sample was in vacuum, the windows’ refraction corrected with the index of air. ‘The values transformed to the refractive indices relative to air agree well at room temperature with the index data obtained by Wray and Neu ... and also agree well in the visible region with the data obtained by Waxler and Cleek’ (p. 87; no numbers).
PHOTON ENERGIES. The paper tabulates against photon energy (eV) and does not say how the energies were obtained. They are hc/lambda of the Hg lines’ AIR wavelengths with hc = 1239.852 eV nm (12398.52 eV A): 2.270486 eV gives 546.0734 nm, 3.063963 eV 404.6565 nm, 3.396712 eV 365.0148 nm, 3.710492 eV 334.1475 nm, the air wavelengths of those lines to 1e-3 nm, whereas the vacuum wavelengths (546.2251 nm ...) would need hc = 1240.2 eV nm (checked 2026-09-27). The true photon energies are therefore lower by the air’s refractivity, 2.8e-4 relative (2.269837 eV for the first row). The data files keep the printed energies (the loader cannot take a photon energy ‘in air’); read as photon energies they put each index at a vacuum wavelength 0.15 nm (at 546 nm) to 0.07 nm (at 238 nm) too short, which is 0.6e-5 (visible) to about 3e-5 (238 nm) in n through the dispersion. A model should use lambda_vac = vacuum(1239.852/E nm in standard air). leviton_2006’s Table 8 (Matsuoka minus CHARMS) is reproduced to its rounding only when the CHARMS formula is evaluated at hc/E read as a vacuum wavelength (see below): Leviton and Frey read the energies as printed.
ACCURACY. ‘The precision of the index measurements was Delta n = +-3 x 10^-6’ (p. 86; level not stated). The fit (Eqs. (1)-(3)) has ‘a root mean square error of 7 x 10^-6 and a maximum error of 26 x 10^-6. The main source of the errors is believed to be due to a distribution of temperatures in the prism sample that increases at low temperatures’ (pp. 87-88).
TABLE 1 STRUCTURE. The scan’s decimal points are faint and the library reads many as spaces (structural fix in the extraction script; the image shows the points). At -165.4 and -111.3 C the 4.504003 and 5.213151 eV rows print ‘-’.
dn/dT. Not tabulated. Fig. 1 shows the index and its temperature derivative at 20.5 C against photon energy, with the formula’s curves (figure, not digitised); from the formula, dn/dT at 20.5 C is 8.9e-6/K at 2.27 eV (recomputed 2026-09-27). The temperature dependence ‘is dominated by the temperature shift of the resonance energy of exciton peak in the UV region ... more than 90%’ (p. 89); the reflection peak near 10.4 eV shifted by -0.05 +- 0.01 eV from -191 to 26 C, (-2.3 +- 0.5) x 1e-4 eV/C, against db2/dT = -2.20 x 1e-4 eV/C from Table 2 at -82 C (p. 88; recomputed -2.198e-4); thermal expansion contributes about -5e-7/C (p. 89).
LEVITON AND FREY (leviton_2006 Table 8; leviton_2008, the preprint, Table 8) list ‘absolute refractive indices of Matsuoka et al. ... minus refractive indices from CHARMS’ at 0.546, 0.436, 0.405 um and 108, 162, 230, 276, 294 K (-165.4, -111.3, -43.4, 2.5, 20.5 C). Recomputed here (2026-09-27) from Table 1 and Leviton and Frey’s Table 3 formula: with the formula at lambda = 1239.852/E nm read as a vacuum wavelength, 12 of the 15 printed differences (x 1e-5) are reproduced to the rounding (0.436 um at 276 K gives -0.5, printed -1; 0.405 um at 276 and 294 K give -0.6 and -0.9, printed 0); at the true vacuum wavelengths the differences are 0.6-1.2e-5 higher (less negative): -1.6 to -1.0 at 0.546 um, -1.0 to +0.6 at 0.436 um, -0.8 to +0.6 at 0.405 um. Leviton and Frey’s finding that ‘CHARMS indices appear to be systematically higher than those of Matsuoka by 1 to 2 parts in the fifth decimal place’ holds at 0.546 um but not in the violet once Matsuoka’s wavelengths are taken as the air wavelengths they are.
FIGURES (not digitised): 1, n and dn/dT at 20.5 C against photon energy, 2.2-5.3 eV; 2, |C_ai| and |C_bi| (the contributions of each Sellmeier parameter to dn/dT) at 20.5 C, logarithmic.
Recorded bibliographic identity status: VALIDATED_WITH_WARNINGS. This concerns identification of the source; it does not establish the scientific correctness of its contents.
Datasets#
matsuoka_1991/n_m165.4C#
{
"id": "n_m165.4C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_m165.4C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 107.75,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, -165.4 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The 4.504003 and 5.213151 eV rows print ‘-’ (not measured at this temperature): 8 points. The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_m111.3C#
{
"id": "n_m111.3C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_m111.3C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 161.85,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, -111.3 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The 4.504003 and 5.213151 eV rows print ‘-’ (not measured at this temperature): 8 points. The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_m43.4C#
{
"id": "n_m43.4C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_m43.4C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 229.75,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, -43.4 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_2.5C#
{
"id": "n_2.5C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_2.5C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 275.65,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, 2.5 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_20.5C#
{
"id": "n_20.5C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_20.5C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 293.65,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, 20.5 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_45.2C#
{
"id": "n_45.2C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_45.2C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 318.35,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, 45.2 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
matsuoka_1991/n_83.3C#
{
"id": "n_83.3C",
"material": "SiO2",
"file": "SiO2/matsuoka_1991__n_83.3C.csv",
"table": "tbl_3e381ac738a4",
"quantity": "refractive_index",
"abscissa_unit": "eV",
"use": "primary",
"parameters": {
"temperature_k": 356.45,
"grade": "type III silica glass (Nippon Sekiei Glass Co.), about 1200 ppm OH",
"form": "bulk",
"technique": "minimum deviation",
"index_medium": "vacuum"
},
"uncertainty": {
"absolute": 3e-06,
"level": "not stated"
}
}
Table 1, 83.3 C: the absolute index at the printed photon energies (Hg lines, 2.27-5.21 eV). The energies are hc/lambda of the lines’ AIR wavelengths (hc = 1239.852 eV nm), not true photon energies: read as photon energies they place each point 2.8e-4 (relative) short in wavelength (see the source notes). Precision +-3e-6 (level not stated).
Formulas#
matsuoka_1991/sellmeier_temperature#
{
"id": "sellmeier_temperature",
"material": "SiO2",
"equation": "Eqs. (1)-(3), p. 87, with Table 2 (p. 88)",
"printed": "n^2 - 1 = sum_{i=1}^{3} a_i / (b_i^2 - E^2) (1); a_i = a_i0 + a_i1 T + a_i2 T^2 (2); b_i = b_i0 + b_i1 T + b_i2 T^2 (3)",
"variable": "E: photon energy in eV as the paper computes it (hc/lambda of the air wavelength, hc = 1239.852 eV nm; see the source notes); T: temperature in degrees C; n: the absolute index",
"range": "2.27-5.21 eV (546-238 nm), -165.4 to 83.3 C (the data fitted)",
"coefficients": {
"a10": "228.7018",
"a20": "46.40806",
"a30": "0.014173",
"a11": "4.93 x 10^-5",
"a21": "-3.27 x 10^-5",
"a31": "-1.704 x 10^-6",
"a12": "1.10 x 10^-7",
"a22": "-3.78 x 10^-8",
"a32": "-2.14 x 10^-9",
"b10": "18.111630",
"b20": "10.671082",
"b30": "0.125",
"b11": "9.15 x 10^-6",
"b21": "-2.9913 x 10^-4",
"b31": "<10^-5",
"b12": "7.478 x 10^-8",
"b22": "-4.8074 x 10^-7",
"b32": "<10^-8"
},
"check": "With b31 = b32 = 0, reproduces Table 1's 66 indices with an rms difference of 7.6e-6 and a largest difference of 26.4e-6 (4.995383 eV, -111.3 C), as the paper states (7e-6 and 26e-6); db2/dT at -82.5 C is -2.198e-4 eV/C (printed -2.20e-4). Checked 2026-09-27 against the printed coefficients, at the printed energies."
}
No form: the variable is a photon energy and the poles’ positions b_i are quadratic in temperature (degrees C), which none of optmat’s forms is (n2_sellmeier_temperature has polynomial pole wavelengths in K). Units as printed: a_i0 eV^2, a_i1 eV^2/C, a_i2 eV^2/C^2, b_i0 eV, b_i1 eV/C, b_i2 eV/C^2. The b_3 temperature terms are printed only as bounds (<1e-5, <1e-8). Parameters of Eq. (1) were fitted per temperature by non-linear least squares from Malitson’s (1965) values, then Eqs. (2)-(3) to them by least squares (p. 87). The decimal points of several coefficients are faint in the scan (for example ‘4 93’, ‘18 111630’); the image shows them. The three terms are read as the band-to-band transitions (b1 = 18.1 eV), the exciton near 10.4 eV (b2 = 10.67 eV) and the 1100 cm-1 lattice band (b3 = 0.125 eV) (pp. 88-89).
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.