Engineering the thermal profile of a polymer by means of conducting microwires provides a path towards versatile tunable optical elements.
Modern photonic applications call for versatile optical devices that can spatially modulate light while dynamically changing their operating parameters on-demand. For instance, tunable lenses are used in three-dimensional (3D) imaging1 while more sophisticated spatial light modulators (SLMs) fuel applications such as optical tweezers to manipulate microparticles2,3, optogenetics to investigate functional neuronal activity4, beam shaping for 3D printing5, high-resolution microscopy at the subcellular level6 and astronomical imaging through the atmosphere7.
Reflective devices create spatial optical path modulation by shaping the topography of a mirror, producing spatially varying delays. Deformable mirrors8 are composed of a reflective membrane selectively distorted by actuators that enable correction of low-order aberrations, namely deviations from the ideal wave exiting an imaging system9. These systems have fast response times but typically a limited number of degrees of freedom, and hence they are attractive for astronomical or ophthalmic imaging where low-order aberrations are corrected in real time through a process called adaptive optics. Other mechanical mirrors, such as the digital light projector10, provide a large number of pixels but only binary (on–off) operation.
Light wavefronts can also be modulated by selectively changing the optical path length through a propagation medium, namely by controlling the delay light experiences when crossing a material. A variable optical path length is typically achieved by changing the thickness of the device — such as in a spherical lens — or by spatially modifying the index of refraction using materials such as liquid crystals (LCs).
Commonly used LC SLMs are usually composed of a LC layer on a silicon substrate and include circuitry to address each pixel of a 2D matrix. The electric field across the LC layer tilts the, otherwise aligned, molecules, creating a change in the effective index of refraction11. Unfortunately, the polarization direction of the incident light has to be aligned with the direction of the LC molecules for optimal performance, which limits operation to linearly polarized light, if intricate manipulations of the polarization state are to be avoided. Another issue is the limited amount of optical path differences LCs can provide and the strong dependence on the colour (wavelength). Other physical properties that are used to locally change the refractive index via external changes include the acousto-optic, electro-optic or magneto-optic effects.
Now, reporting in Nature Photonics, a team based in Barcelona and Paris propose and demonstrate the use of thermal effects to create exquisitely controlled local refractive index changes, leading to generally non-symmetric micro-optical devices12. Instead of shaping the material, they alter the refractive index by taking advantage of its change as a function of temperature. While the concept is not totally new, what is remarkable is the precision of the temperature control across the material slab to create localized variations in the refractive index.
Pascal Berto and colleagues pattern microwires by design over the surface of a polydimethylsiloxane layer while a voltage is applied to create the desired heat map12. Each heat map produces a specific index distribution and, in turn, a corresponding spatial light modulation associated with a particular wavefront. The physical phenomenon involved, which is known as the thermo-optic effect, is simple. It describes the changes in refractive index with temperature, an effect that has been widely used in optoelectronics to tune the properties of semiconductor resonators and lasers13.
Berto and colleagues generate an accurate model of the thermo-optic effect, enabling rigorous designs. Towards this goal, they implement an optimization process that generates a microheater design based on a target refractive index profile and wavefront. Following this process, they design and validate various optical elements, including non-circular symmetric demonstrations contrasting the traditional symmetric optical devices12. The wavefront accuracy is typically better than a fraction of a wavelength while the focal length error is under 2%.
A striking advantage of the technique is the possibility to tune a given device by changing the applied voltage and hence producing a variable heat map. The particular device demonstrations include conducting wire microheaters made of either gold or indium tin oxide (ITO). The wires produced are 50-nm thick and less than a micrometre wide. Remarkably, these wirings introduce negligible distortion in the light wavefronts. However, the gold version presents increased scattering and absorption as compared with the transparent ITO version that, in turn, requires a higher applied voltage. Interestingly, the smaller the devices, the faster their response time, reaching the sub-millisecond level for microlenses on the order of 10 μm in diameter. For comparison, these would be faster than most LC devices but still slower than microelectromechanical or acousto-optic systems.
The flexibility to engineer thermal profiles that produce free-form index of refraction variations is a key advantage of the technique by Berto and colleagues. Free-form optics is a term generally used to describe optical elements without translational or rotational symmetry. Nevertheless, there are still limitations to the refractive index profiles that are possible to generate following Berto and colleagues’ work, in part due to the constraints imposed by the heat distributions and possible wiring width and topologies. It is worth emphasizing that while non-symmetric optics can be implemented in LC SLMs or deformable mirrors, the former are typically pixelated and polarization-dependent while the latter have a limited number of degrees of freedom.
Among the implementations presented as a proof of concept by Berto and colleagues, tunable lens arrays, as shown in Fig. 1, and axicons to generate Bessel-like beams as well as an astigmatic lens demonstrate the flexibility of the process. Furthermore, the authors show that the lenses present low dispersion, similar to that of traditional glass lenses. Therefore, this research provides an attractive avenue to tunable, versatile, polarization-independent and spectrally broadband devices. Future work should provide a path to investigate the extent to which the process can be scaled up. Such devices would be appealing for applications such as microscopy, adaptive optics or machine vision.
Fahrbach, F. O., Voigt, F. F., Schmid, B., Helmchen, F. & Huisken, J. Opt. Express 21, 21010–21026 (2013).
Reicherter, M., Haist, T., Wagemann, E. U. & Tiziani, H. J. Opt. Lett. 24, 608–610 (1999).
Grier, D. G. Nature 424, 810–816 (2003).
Nikolenko, V. et al. Front. Neural Circuits https://doi.org/10.3389/neuro.04.005.2008 (2008).
Lu, Y., Mapili, G., Suhali, G., Chen, S. & Roy, K. J. Biomed. Mater. Res. Part A 77, 396–405 (2006).
Pavani, S. R. P. et al. Proc. Natl Acad. Sci. USA 106, 2995–2999 (2009).
Tyson, R. K. Principles of Adaptive Optics 4th edn (CRC Press, 2015).
Bifano, T. G., Perreault, J., Mali, R. K. & Horenstein, M. J. IEEE J. Sel. Top. Quant. Electron. 5, 83–89 (1999).
Goodman, J. W. Introduction to Fourier Optics 4th edn (W. H. Freeman, 2017).
Younse, J. M. IEEE Spectrum 30, 27–31 (1993).
Saleh, B. E. A. & Teich, M. C. Fundamentals of Photonics 2nd edn (Wiley, 2012).
Berto, P. et al. Nat. Photon. https://doi.org/10.1038/s41566-019-0486-3 (2019).
Gu, L., Jiang, W., Chen, X. & Chen, R. T. IEEE Photon. Technol. Lett. 19, 342–344 (2007).
About this article
Cite this article
Piestun, R. Free-form micro-optical elements heat up. Nat. Photonics 13, 583–584 (2019). https://doi.org/10.1038/s41566-019-0515-2