Open Access
Add to BibSonomyAbstract
We present a non-mechanical microiris based on two complementary electrochromic (EC) materials, namely viologens and phenozines, with an almost neutral spectral behavior. Measurements concerning the spectral light transmission, modulation transfer function, and response time validate that the optical performance of the EC-iris is comparable to that of a classical blade iris. The time constant is limited due to diffusive mass transport of the molecules, but can be reduced by a short voltage pulse. The current controlled transmission of the EC-material renders the individual control of each iris segment without crosstalk possible, allowing its usage as tunable spatial filter.
© 2015 Optical Society of America
1. Introduction
In the modern world, camera systems have become an everyday tool with numerous applications. With the integration of adaptive optical functionality and electronic signal processing, even more advanced opportunities for miniaturized optical systems have emerged. They range from stealth security to high performance smartphone cameras or even medical imaging equipment used for endoscopy or surgery assistance [1]. However, it is challenging to integrate all optical components required for high quality imaging into miniaturized camera systems. In specific, there is no technology available for the implementation of a non-mechanical miniature variable aperture that combines low volume requirements and electronic control with both low power consumption as well as low voltage levels.
Recently, promising approaches for small variable aperture devices have been proposed. All major concepts rely on the motion of a strongly absorbing material into the path of light [2–4]. Our concept differs from those, because we employ EC-materials which inherently change their optical properties when an external electrochemical potential is applied. This concept reduces the space and energy requirements drastically, supporting its use in small battery powered systems. So far, EC-materials are commercially used in auto-dimming mirrors and tunable windows [5, 6]. They have also been proposed for the use in optical applications, but up to now the performance was insufficient compared to conventional devices [7–9, 11]. In this paper, we present an EC-iris matching the optical requirements of real applications. The involved EC-materials were specifically chosen to achieve almost neutral optical absorption characteristics with the capability to control the amount of absorption by the applied current.
2. Design and fabrication
In general, EC-devices rely on an electrochemical cell set-up as schematically shown in Fig. 1 (a). The electrolyte with the dissolved EC-molecules is sandwiched between two parallel aligned glass substrates each carrying a thin ITO layer as translucent and conductive electrode. The realization of neutral color in an EC-transmissive device requires the combination of at least two redox-active EC-molecules with their specific absorption behavior. The coloring efficiency is optimized if these molecules are chosen in such a way that they establish the two complementary half cell redox reactions. A current flow through the cell causes both types of molecules to color: one on oxidation and the other on reduction at the individual ITO electrodes. The colored molecular species meet at the center of the cell. They recombine by intermolecular charge transfer, which causes the bleaching of both molecular species. The complete redox cycle is highly reversible. A variable steady state absorption level is achieved feeding an appropriate constant current through the cell.
Fig. 1 Working principle of the EC-cell based on the two complementary EC-molecules phenazine (Pz) and viologene (Vio). a) The spatially separated redox reactions at the individual electrodes with the involved charge transfer. b) and c) redox states and corresponding colors of phenazine and viologene, respectively.
The specific EC-molecules used in our devices, are 5,10-dihydro-5,10-dimethylphenazines (phenazine) and 1,1′-dibenzyl-4,4′-bipyridinium dihexafluorophosphate (viologen) dissolved in propylene carbonate. Phenazine (ABCR) was used without further purification. The viologenes were synthesized by dissolving dibenzyl viologen dichloride (Sigma-Aldrich) in a minimum amount of water and treated with a saturated solution of NH4PF6, followed by collecting and drying in high vacuum. Figure 1 (b) and 1 (c) visualize the variation of the molecules electronic configuration imposed by charge transfer and the impact on their optical absorption behavior. Both molecules are almost transparent in their natural state. The phenazines and viologenes turn green on oxidation and violet on reduction, respectively.
The substrates used in our devices are ITO coated glass slides with a resistivity of 70–100 Ω/sq (Sigma Aldrich). They were cleaned by an ultrasonic bath in acetone, isopropanol, and deionized water. To establish two iris levels, both ITO films were laterally structured to form two concentric rings with an additional conducting path that allows individual external control of that segment. The exploded view of the device is shown Fig. 2 (a). The structuring of the ITO-electrodes and the encapsulation of the device exploiting the dry filmresist (Ordyl 355; Elga Europe) were reported earlier [7]. The assembled device was filled under vacuum with the electrolyte containing the EC-molecules. For airtight sealing a polymer (Crystalbond 590) was used to close the filling channel.
Fig. 2 a) Exploded view of the EC-iris device. b) Optical images of all possible switching states of a two-level iris (electrolyte thickness 275 μm).
3. Results
3.1. Optical transmission
Optical images of the four possible switching states of the two-level iris are shown in Fig. 2 (b). A maximal current of 1.2 mA @ 2 V was fed to each segment to saturate the absorption. We conducted additional experiments to investigate the optical transmission as function of the applied current for an unstructured EC-device. The electrolyte layer thickness was 275 μm realized by stacking five layers of Ordyl 355 as spacer between the electrodes. Illumination was performed by a halogen lamp (Schott Kl 1500). The transmitted light was captured with an integrating sphere and finally detected with a spectrometer (OceanOptics PlasCalc 2000). The spectra of both the transparent and the opaque state were recorded and are shown in Fig. 3 (a).
Fig. 3 Transmission spectra of an unstructured EC cell: a) transparent (j=0 mA) and saturated opaque state (j=1.2 mA) for a sample with a 275 μm electrolyte layer. b) Transmission obtained integrating the spectral data in the wavelength range from 400 to 750 nm for various control currents. The sample used had an electrolyte layer thickness of 165 μm.
The neutral state shows an almost uniform transmission of over 80 % for wavelengths larger than 500 nm and a moderate transmission of about 60 % for shorter wavelengths. However, this includes approximately 16.7 % Fresnel reflection losses @500 nm (from the air/glass (nG = 1.517), glass/ITO (nITO = 2), ITO/electrolyte (nE = 1.421), electrolyte/ITO, ITO/glass, and glass/air interfaces) as well as approximately 3–4 % losses owing to the intrinsic absorption of the electrolyte and the two ITO layers. Further absorption losses from the uncolored molecules are in the order of 2–3 %. These values depend on the wavelength. The overall spectral absorption of the device is a superposition of the absorption of the colored layers next to the ITO electrodes as is obvious from Fig. 1. The two molecular species were chosen in such a way that their spectral absorption complement each other to establish an almost neutral absorption behavior over the entire optical range. A maximum variation in absorption on switching was obtained at a wavelength of about 600 nm where the transmission changed from 83 % to about 1.7 %, which is almost a factor of 50. To further investigate the ability to control transmission, the recorded intensity through an unstructured cell with a 165 μm thick electrolyte, was integrated over the wavelength range from 400 to 750 nm and plotted against the applied current j [Fig. 3 (b)]. In accordance to Lambert-Beer’s law established in linear absorption theory, it was found that the transmitted intensity I depends exponentially on the applied current: ln(I(j)/I0) ∝ − j with I0 the intensity in the neutral state. The number of absorbing molecules is corresponding to the number of injected electrons and thus linearly depending on the current.
3.2. Switching time
Measurements conducted with a HeNe laser (@633 nm) on an unstructured cell with a 275 μm electrolyte thickness driven with 1 V revealed a coloring and bleaching time of 10 s and about 70 s, respectively. Coloring of the device is driven by the applied voltage and is thus much faster than the bleaching process, which is purely driven by mass diffusion transport of the colored molecules towards the center of the EC-cell. Fig 4 (a) shows the transmission upon bleaching of unstructured devices with different electrolyte layer thicknesses. The bleaching time depends on the square of the thickness, as is expected for diffusive mass transport. To estimate the diffusion coefficient D of the cell, a simple theoretical model was utilized. Imposed by the symmetry of the EC-cell, we considered only one half cell filled with colored viologenes and analog the other half cell with phenazines assuming that the diffusion coefficient of both molecular species are identical and that the molecules neutralize at the center plane of the cell. Then the time dependent spatial concentration of the viologenes in its half cell is given by [10]:
Under the assumption that 10 % total colored molecules correspond to 90 % of total transmission through the device the diffusion coefficient can be determined from the formula to be: D = d2/5t, where d is the distance from the electrode to the center of the cell (half thickness) and t is the bleaching time required to regain 90 % of the initial transmission. Using the data presented in Fig 4 (a) we approximate the diffusion coefficient to be D = 0.52 · 10−10 m2/s. This result can be used to find an optimum electrolyte layer thickness concerning maximal absorption and switching speed of the device.Fig. 4 a) Time dependence of the transmission of four EC devices with different electrolyte layer thickness, when a voltage of 1 V is applied. b) Transients of the transmission through an EC-iris. After the outer iris ring is colored by an applied voltage of 1 V for 120 s, bleaching is supported by a 1 V counter-pulse with different pulse duration.
The bleaching time was reduced feeding a short counter voltage pulse to the device. All molecules in the vicinity of the electrode surface are bleached rapidly. That technique can be used to quickly increase the transmission up to almost 90 % of total bleaching. The molecules that are not close to the electrodes are still bleached by diffusion to the center of the device. The effectiveness of this procedure depends on the length of the pulse. A measurement of the transmission through an EC-Iris, where one segment is initially colored and then bleached with different pulse length is shown in Fig. 4 (b). The results show a large improvement on bleaching times for a 7 s long pulse, while very short or long pulses only lead to a small improvement. Long pulses result in a coloring of the device in the opposite direction and thus decrease the transmission.
3.3. Modulation transfer function
Finally, we investigated the effect of the iris on the image quality by measuring the modulation transfer function (MTF). The measurements were conducted by imaging an USAF 1951 target on a commercial camera sensor (Canon 300 D). The target was illuminated with a halogen lamb, where the spectrum was narrowed to a wavelength of (633 ±5) nm. Images with different iris diameters were taken with either the EC-iris or a classic iris for comparison. Figure 5 shows the contrast against the spatial frequency of the target.
Fig. 5 Comparison of MTF measurements for different aperture diameters (3, 6, 10 mm) of a classical and an EC-iris.
The classical iris works as almost ideal reference. The figure shows, that the overall performance of both irides are comparable. However, the EC-iris has two limitations compared to the classical one, which become obvious in the measurements. For the 10 mm aperture diameter, the overall contrast with the EC-iris is lower because the transparent areas are slightly scattering light. In the 3 mm opening on the other hand, the contrast with the EC-iris is somewhat higher than desired, because the large opaque areas are slightly transmitting light. The measurement with a 6 mm aperture diameter shows both effects, with a crossing point at medium frequencies.
To improve the MTF of the EC-iris in future designs, it will be important to decrease the transmittance in the opaque state by using higher performing materials. Furthermore it is necessary to increase the transmittance in the transparent state by including anti-reflex layers and matching the refractive index of the electrolyte and the electrode materials.
4. Conclusion
A high contrast micro-iris based on two complementary EC-molecules has been presented with a virtually neutral spectral behavior. The absorption of each circular segment can be individually tuned by the applied current, leading to the iris functionality with apodisation capability. The difference in the coloring and bleaching time constant was attributed to the mass diffusion transport inherent to the bleaching process. By involving counter voltage pulses it was possible to reduce the overall time constant by neutralizing the redox species in the vicinity of both electrodes. Experiments revealed that the MTF of the EC-iris almost matches the MTF of the classic iris despite some structural limitations. while requiering less space and offering full control over the transmitted intensity. The low requirements in space, voltage, and energy, together with an adequate optical performance, makes the EC-iris an ideal solution for miniaturized optics.
Acknowledgments
We thank the German Research Foundation (DFG) for their financial support through Priority Program 1337 ‘Active Micro-optics’. We also thank the Nano Structuring Center (NSC) for their technical support.
References and links
1. G. Iddan, G. Meron, A. Glukhovsky, and P. Swain, “Wireless capsule endoscopy,” Nature 405(6785), 417 (2000). [CrossRef] [PubMed]
2. H. Yu, G. Zhou, Y. Du, X. Mu, and F.S. Chau, “Mems-based tunable iris diaphragm,” J. Microelectromech. Syst. 21(5),1136–1145 (2012). [CrossRef]
3. P. Müller, R. Feuerstein, and H. Zappe, “Integrated optofluidic iris,” J. Microelectromech. Syst. 21(5), 1156–1164 (2012). [CrossRef]
4. C. Kimmle, U. Schmittat, C. Doering, and H. Fouckhardt, “Compact dynamic microfluidic iris for active optics,” Microelec. Eng. 88(8), 1772–1774 (2011). [CrossRef]
5. A. Kraft and M. Rottmann, “Properties, performance and current status of the laminated electrochromic glass of gesimat,” Sol. Energ. Mat. Sol. Cells 93(12), 2088–2092 (2009). [CrossRef]
6. Gentex Corporation, www.gentex.com.
7. T. Deutschmann and E. Oesterschulze, “Micro-structured electrochromic device based on poly(3,4-ethylenedioxythiophene),” J. Micromech. Microeng. 23(6), 065032 (2013). [CrossRef]
8. T. Deutschmann and E. Oesterschulze, “Integrated electrochromic iris device for low power and space-limited applications,” J. Opt. 16(7), 075301 (2014). [CrossRef]
9. J.H. Kang, S.M. Paek, S.J. Hwang, and J.H. Choy, “Optical iris application of electrochromic thin films,” Electrochem. Commun. 10(11), 1785–1787 (2008). [CrossRef]
10. H.S. Carslaw and I. Jaeger, Conduction of Heat in Solids (Oxford University, 1959).
11. D. Pätz, T. Deutschmann, E. Oesterschulze, and S. Sinzinger, “Depth of focus analysis of optical systems using tunable aperture stops with a moderate level of absorption,” Appl. Opt. 53(28), 6508–6512 (2014). [CrossRef] [PubMed]
