Miniature temperature sensors take the heat

Fibre optic sensors really do shine. No, I don't just mean when you pass light down one end and see it come out of the other, I mean they can be made to perform some pretty clever tasks, often triumphing over their electronic rivals. A fibre sensor fabricated by splicing together three different types of optical fibre and capable of measuring temperatures up to 1000°C, has recently been demonstrated by scientists in South Korea^†.

It Ain't 'Alf Hot Mum

Glass optical fibres have long been favoured as the means for interrogating sensors because of their immunity to electromagnetic interference, their inert properties and for their ability to operate at extreme temperatures. The weakness in the system, however, is often the sensing device itself, whether it is located along the length of the fibre or is positioned at the end. For example, if the sensor is made from materials that are prone to corrosion, are metallic, or bonded together using glue that will change the sensor's physical characteristics under vacuum or at temperatures within the operating range of the fibre itself, it's going to be the weak link in the chain. One of the best solutions is therefore to make the sensor from the fibre material itself, and to make it an intrinsic part of the fibre without glues or external parts.

A fibre temperature sensor which could operate at similar extremes was discussed in a previous post. It was intrinsic to the fibre, used no external parts and was fabricated by using a femtosecond laser to carve out a small cavity in the fibre to form a Fabry-Pérot interferometer. By making open cavities, this method is particularly well suited to chemical sensing applications, but the laser machining adds significant complexity to the fabrication of the sensor. This latest work, which uses fusion splicing technology to construct the sensor, demonstrates what is arguably a much simpler way of building an optical fibre temperature sensor.

Fusion splicing is a way of bonding the end of one fibre onto the end of another using the heat from an electric arc. The method results in greater light transmission between the two fibres than would be achieved by physically bringing the fibres together inside a connector. Connectors of this type typically contain an index-matching fluid to minimise the light lost to reflection at the fibre surfaces, but the preferred choice for permanently coupling two fibres is fusion splicing, a well established technology which is used routinely in the telecommunications industry.

The sensor was comprised of three fibre types: a photonic crystal fibre (PCF) which carried the incident light from a broadband light emitting diode to the sensor and guided the reflected light back to an optical spectrum analyser; a short section of hollow core optical fibre (HOF) fused onto the end of the PCF; and a short length of single-mode fibre (SMF) which was fused onto that (see diagram below).

Construction of fibre sensor comprising three different fibre types fused together

Sensors could be made with lengths as short as 580 μm, this being the combined lengths of the HOF and SMF sections. And since single-mode fibre is typically 125 μm in diameter when stripped down to the cladding layer, these sensors can be made very small.

Both the HOF and SMF sections act as Fabry-Pérot cavities in which light reflects back and forth producing interference patterns in the spectrum of light reflected back down the PCF and recorded by the optical spectrum analyser. The interference pattern from each cavity was seen as a series of consecutive low and high intensity peaks as a function of wavelength. Careful analysis of the combined interference patterns revealed the effect of temperature change on the sensor.

A demonstration was conducted in which a series of spectra were recorded as the sensor temperature was varied between 50° and 1000° C; in steps of 50°. The sensor was sensitive to changes in temperature in two ways: by the thermo-optic effect, which is caused by a change in refractive index with temperature; and by thermal expansion, whereby the length of the Fabry-Pérot cavity changes with temperature. Both effects cause a change in the optical path length taken by the light resonating in the cavities.

The temperature sensitivity of the sensor was largely due to the thermo-optic effect in the SMF, which is attributed to the fact that the dopant material in its core has a high thermo-optic coefficient. The HOF exhibited no thermo-optic sensitivity due to its air-filled core. Similarly, the PCF, which did not have a doped core, had negligible sensitivity to temperature.

The authors did not specifically address their reasons for using the 70 μm length HOF, referring to it as an "auxiliary" Fabry-Pérot cavity; the actual sensing element being the SMF. As an experimental device, however, it allowed the researchers to discern the comparative magnitudes of the thermo-optic and thermal expansion effects. As a working device, it is possible that the HOF would be used to improve the visibility of the light interference produced in the SMF optical cavity, by increasing the reflection at the front surface to the cavity.

The appeal of this sensor is in its construction, which relies on standard fusion splicing technology, and in the fact that it uses only optical fibre with no additional parts, making it well suited to high temperature applications. As a very small sensor, it could also enable temperature measurements to be made with pinpoint spatial accuracy. One might conceive of a design in which an array of such sensors are spread across a surface, enabling multiple temperature measurements to be made within a compact area.
 
 
† Hae Young Choi, Kwan Seob Park, Seong Jun Park, Un-Chul Paek, Byeong Ha Lee, Eun Seo Choi (2008). Miniature fiber-optic high temperature sensor based on a hybrid structured Fabry–Perot interferometer Optics Letters, 33 (21) DOI: 10.1364/OL.33.002455