March 2015
Spotlight Summary by Daniel Poitras
Spectral design of temperature-invariant narrow bandpass filters for the mid-infrared
Many properties of optical devices change with temperature: for example, laser light output shifts and broadens in wavelength, sensitivity and selectivity of sensors drop, efficiency varies. Even devices that are considered passive, like optical filters, are affected by temperature changes, particularly narrowband filters that see their peak shift and reduce in height. Controlling the temperature, often to a low value, is a frequent remedy for preventing any detrimental thermal effects on active and passive components.
But we live in an age of miniaturization, where the size of devices is reduced, and their constituent components are integrated in a single package, if not on a single chip, so that many of these devices can work in parallel (modern high-definition displays and detectors are examples). The energy, cost and logistics required to cool such integrated systems can become significant (look at data centers), and the incentive to develop cooling-free integrated devices is high. It is thus not a surprise that a large effort goes to the creation of cooling-free light sources and sensors, and it becomes equally important that the passive devices are also temperature insensitive.
The paper by Stolberg-Rohr and Hawkins describes a way to design and fabricate temperature-insensitive narrowband filters for mid-infrared applications. In that wavelength range, it is not rare to use cryogenic temperature to maintain good performance of sensors and filters, but the recent advances in cooling-free sensors now forces the development of narrowband filters that can equally perform without cooling.
The effect of the temperature on optical materials used in optical coatings has been relatively well known for years: a thermal expansion that affects the thicknesses, and a generally more prominent variation of the refractive indices (thermo-optic effect) that affects both the optical thicknesses (refractive index times thickness) and the reflection coefficients at interfaces. Most optical materials have a positive thermo-optic effect (refractive index increases with temperature), and materials with a negative thermo-optic effect are rare (practically limited so far to some organic materials, phases of TiO2 and lead-based materials). Previous attempts to design temperature-invariant narrowband filters consisted of a balanced ratio of materials with positive and negative thermo-optic effects, based on thermo-optic effects measured from single-layer, bulk material measurements.
Stolberg-Rohr and Hawkins have refined the methodology for predicting the thermal properties of narrowband filters. They first observed that bulk values of the thermo-optics effect may not always be useful to predict the behavior of materials in thin-film multilayer structures. Also, and most importantly, they observed a large variation of the temperature sensitivity of narrowband filters with different design structures (i.e. low- vs. high-index cavities, single vs. multiple cavities, cavities with different optical thicknesses), and were able to develop a model for predicting the behavior of a filter based on its design type. To validate their findings, Stolberg-Rohr and Hawkins had the brilliant idea of using a large variety of narrowband filters previously manufactured for various space applications, as well as a few more filters made specifically for this study. Their measurements confirmed the usefulness of their approach: as predicted by their model, some filters showed a high-stability in performance from 20°C to 200°C, on a wavelength range as large as 9 microns! In my view, the methodology described in this work should be applied in the future to other types of optical filters and other optical materials. In addition, their model could help to design experiments for better characterizing the thermo-optic effect in thin films embedded in multilayer systems.
You must log in to add comments.
But we live in an age of miniaturization, where the size of devices is reduced, and their constituent components are integrated in a single package, if not on a single chip, so that many of these devices can work in parallel (modern high-definition displays and detectors are examples). The energy, cost and logistics required to cool such integrated systems can become significant (look at data centers), and the incentive to develop cooling-free integrated devices is high. It is thus not a surprise that a large effort goes to the creation of cooling-free light sources and sensors, and it becomes equally important that the passive devices are also temperature insensitive.
The paper by Stolberg-Rohr and Hawkins describes a way to design and fabricate temperature-insensitive narrowband filters for mid-infrared applications. In that wavelength range, it is not rare to use cryogenic temperature to maintain good performance of sensors and filters, but the recent advances in cooling-free sensors now forces the development of narrowband filters that can equally perform without cooling.
The effect of the temperature on optical materials used in optical coatings has been relatively well known for years: a thermal expansion that affects the thicknesses, and a generally more prominent variation of the refractive indices (thermo-optic effect) that affects both the optical thicknesses (refractive index times thickness) and the reflection coefficients at interfaces. Most optical materials have a positive thermo-optic effect (refractive index increases with temperature), and materials with a negative thermo-optic effect are rare (practically limited so far to some organic materials, phases of TiO2 and lead-based materials). Previous attempts to design temperature-invariant narrowband filters consisted of a balanced ratio of materials with positive and negative thermo-optic effects, based on thermo-optic effects measured from single-layer, bulk material measurements.
Stolberg-Rohr and Hawkins have refined the methodology for predicting the thermal properties of narrowband filters. They first observed that bulk values of the thermo-optics effect may not always be useful to predict the behavior of materials in thin-film multilayer structures. Also, and most importantly, they observed a large variation of the temperature sensitivity of narrowband filters with different design structures (i.e. low- vs. high-index cavities, single vs. multiple cavities, cavities with different optical thicknesses), and were able to develop a model for predicting the behavior of a filter based on its design type. To validate their findings, Stolberg-Rohr and Hawkins had the brilliant idea of using a large variety of narrowband filters previously manufactured for various space applications, as well as a few more filters made specifically for this study. Their measurements confirmed the usefulness of their approach: as predicted by their model, some filters showed a high-stability in performance from 20°C to 200°C, on a wavelength range as large as 9 microns! In my view, the methodology described in this work should be applied in the future to other types of optical filters and other optical materials. In addition, their model could help to design experiments for better characterizing the thermo-optic effect in thin films embedded in multilayer systems.
Add Comment
You must log in to add comments.
Article Information
Spectral design of temperature-invariant narrow bandpass filters for the mid-infrared
Thomine Stolberg-Rohr and Gary J. Hawkins
Opt. Express 23(1) 580-596 (2015) View: Abstract | HTML | PDF