High-precision thermal expansion measurements using small Fabry-Perot etalons

Abstract

Coefficient of thermal expansion (CTE) measurements using small Fabry-Perot etalons were conducted on high and low thermal expansion materials differing in CTE by a factor of nearly 400. The smallest detectable change in length was ~10^-12 m. The sample consisted of a mm-sized Fabry-Perot etalon equipped with spherical mirrors; the material-under-test served as the 2.5 mm-thick spacer between the mirrors. A heterodyne optical setup was used with one laser locked to an ~780 nm hyperfine line of Rb gas and the other locked to a resonance of the sample etalon; changes in the beat frequency between the two lasers as a function of temperature directly provided a CTE value. The measurement system was tested using the high-CTE SCHOTT optical glass N-KF9 (CTE = 9.5 ppm/K at 23 °C). Measurements conducted under reproducibility conditions using five identically-prepared N-KF9 etalons demonstrate a precision of 0.1 ppm/K; absolute values (accuracy) are within 2-sigma errors with those made using mechanical dilatometers with 100-mm long sample rods. Etalon-based CTE measurements were also made on a high-CTE (~10.5 ppm/K), proprietary glass-ceramic used for high peak-pressure electrical feedthroughs and revealed statistically significant differences among parts made under what were assumed to be identical conditions. Finally, CTE measurements were made on etalons constructed from SCHOTT's ultra-low CTE Zerodur^(R) glass-ceramic (CTE about -20 ppb/K at 50 °C for the material tested herein).