Non-contact printing methods such as inkjet, electro hydrodynamic, and aerosol printing have attracted attention for their precise deposition of functional materials that are needed in printed electronics, optoelectronics, photonics, biotechnology, and microfluidics. In this article, we demonstrate printing of tapered optical waveguides with losses of 0.61 ± 0.26 dB/cm, with the best performing structure achieving 0.19 dB/cm. Such continuous features are indispensable for successfully printing functional patterns, but they are often corrupted by capillary forces. The proposed inkjet printing method uses these forces to align liquid bridges into continuous features, enabling the printing of smooth lines on substrates with arbitrary contact angles.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Non-contact printing methods such as inkjet, electro hydrodynamic  and aerosol  printing attracted attention for their precise deposition of functional materials which is needed in printed electronics , optoelectronics [4,5], photonics [5,6], biotechnology  and microfluidics [8,9]. State of the art multi-mode optical waveguides are typically prepared lithographically [10,11] or by reactive ion etching  and are reported to exhibit minimal losses down to 0.02 dB/cm and 0.2 dB/cm, respectively. While these techniques are well established and yield the highest efficiencies, they also consume a lot of material and can be difficult to use on fragile substrates. Non-contact printing could solve these issues, but up to now, the surface roughness achieved was not comparable to these well-established processes. Also, although jetted lines with comparable geometries have been prepared previously, optical losses have never been published .
One of the major challenges in functional printing in contrast to pixelated, graphic reproduction is the necessity of depositing continuous features . Whereas printing of tracks is an established process in electronics, there are few reports of light guides [5,6,13] or fluidic channels [8,9]. The shape requirements for the latter applications are more severe because high aspect ratios and smooth surfaces are crucial for functionality. Even a line - the simplest continuous form - is difficult to print steadily on a low energy substrate. Applied ink features change morphologically to reduce the energy of the system to a thermodynamically stable configuration and hence continuous features are bulged or split into discrete spherical caps.
A lot of work has been devoted to find conditions under which lines are quasi-stable based on contact angle and contact line , bulging dynamics [15,16], viscosity  and drop spacing [18–20]. Quasi-stable means that the line stability relies on an energy barrier or dissipative mechanism. Such energy barriers are a contact line with contact angle hysteresis or a pinned line . If not, enough activation energy is provided, the system remains in the quasi-stable configuration. For instance, dissipation due to viscosity slows down the transformation, leaving eventually enough time to dry or cure the ink in the quasi-stable configuration .
Instead of tediously tweaking parameters to hinder bulging by increasing contact angle hysteresis, viscosity or printing speed, we propose to print lines as a series of capillary bridges. This concept can be extended to form printed tapers for coupling light and introduce curvature or junctions into the light guide; elements that are essential for quantitative characterization of the deposited waveguides.
2. Results and discussion
The fabrication process is illustrated in Fig. 1(a): Discrete spherical caps are deposited and stabilized. In subsequent droplet deposition between these pinning caps, capillary forces facilitate the formation of self-aligned bridges. The resulting straight line can then be hardened. This self-alignment mechanism allows for very smooth lines, but fresh ink needs to wet the redeposited ink ideally . Further, the bridge volume must be adjusted to the distance between two caps such that each capillary bridge wets half of each pinning cap or less and separated bridges are formed. This isolation blocks capillary flow between two different capillary bridges and thereby prohibits bulging. In this case, it is also sufficient to study the fluid mechanics of a single capillary bridge without compromising the physical validity of the model.
It is worth mentioning that stable capillary bridges are not limited to span two identical pinning caps. Capillary bridges between three or more caps allow to deposit junctions and sharp edges. Figure 1(b) shows examples of simulated (Surface Evolver ) liquid bridges which can be used as primary building blocks. These basic blocks can then be assembled to build more complex structures like the ones shown in the simulations in Figs. 2(h)-2(n). Such structures are even more prone to bulging than straight lines when printed without pinning caps (Figs. 2(a)-2(g), printed onto a Teflon coated glass substrate with a static contact angle of 61 degrees). The simulations are verified by printing the same designs with the capillary bridge method Figs. 2(o)-2(u). Comparison of these structures to the ones attained with the standard direct printing method makes it obvious that the proposed method can suppress bulge formation while also increasing accuracy and uniformity of the printed structures. Such closed lines can further also be used as pinning structures for flat but thick films as shown in Fig. 2(u). Pinning at the ridge of the lines allows to have a flat top surface. Dots can be used to form a line, closed lines for films, films for 3D objects by merging.
On our substrate, typical lines are 120 μm wide and 31 μm high if several droplets are used for each pinning cap, which is preferable for process stability. Defect densities lower than 3 defects per meter have reproducibly been achieved for such lines. To estimate the competitiveness of our approach for guiding light, tapered optical waveguides with a bend radius of 5 mm are ink jet printed and characterized. A PET foil or glass slide bearing a teflonized MgF2 cladding is used as a substrate. The printed and cured acrylic polymer has a nominal refractive index of 1.52 and the cladding has one of 1.37, which allows good optical insulation.
Figure 3(a) displays the optical power attenuation along straight, defect free lines. The sliding prism method  was used to measure the optical losses of the printed waveguides after the bend (compare to photograph of the setup inset in Fig. 3(a), curved configuration avoids blinding the sensor with direct laser light). With additional blackening of the substrate, masking of other waveguides and other possible scatterers as well as background correction (measurement without prism), the outcoupled intensity could be determined using a photo-diode. The thereby performed measurements at a wavelength of 650 nm resulted in average losses of 0.61 ± 0.26 dB/cm (n = 10) for claddings thicker than the wavelength independently of the substrate (PET or glass). Performances down to 0.19 dB/cm were detected for optically smooth waveguides such as the one shown in the insert in Fig. 3(a) on the top right. Light guiding performance comparable to the established technologies mentioned above can thus be achieved.
A major factor enabling such low losses is the surface quality: Slight roughness of the waveguide for example increases losses up to 0.8 dB/cm (Fig. 3(a), inset middle). Figure 3(b) shows the optical power along a light guide with introduced bulging defects. The waveguide loses at least 0.8 dB across each defect. These measurements underline the importance of the smoothness of printed waveguides which the described method can provide. To our knowledge, losses have never been characterized for inkjet printed waveguides to date , which currently only leaves established techniques like photolithography or reactive ion etching for comparison. The reported measurements may serve as a first benchmark for printed waveguides in the future as, even without further development, our technique reaches losses close to the established alternatives. This comparison highlights it as an attractive option especially for connecting customized components for chip assemblies or fabrication of optical elements on fragile and non-planar surfaces.
The potential of the reported printing method is not limited to optical components. By using different inks, it can easily be extended to various functionalities, materials and combinations thereof. It may hence pave the way towards inexpensive fabrication of easily customizable, integrated optical chips by additive manufacturing with a single machine. The all printed device shown in Fig. 4(c) illustrates how such an integrated chip printed with the capillary bridge method could look like. It combines a microfluidic channel, a tapered light guide (still functional, arrow points at the light coupled into the blue tinted solution filling the channel) and a metallic comb capacitor, but many other functionalities are conceivable.
In conclusion, the reported approach of depositing pinning caps and subsequently connecting them with liquid bridges was shown to turn the usually hindering capillary force into the driving force for printing continuous lines. The capability of creating smooth, continuous features is utilized to form low loss optical waveguides and tapers for light coupling, two key components for optical circuits. Losses down to 0.19 dB/cm have been demonstrated, which are to our knowledge the first absolute values reported for printed waveguides so far. They are close to established methods and additional optimization of ink and substrate materials could further enhance the performance.
Using stabilized ink caps instead of e.g. patterned surfaces as pinning structures is convenient for design flexibility while still improving the print fidelity of lines and especially more complex features. Pinning caps block any meridional capillary flow, which inhibits bulging and enables the formation of sharp angles or junctions. It also simplifies the design process because only isolated capillary bridges and the static contact angle between ink and substrate need to be considered in contrast to alternative methods . Further, no material contrast exists between the pinning structures and the overprinted ink, which is crucial to minimize losses for optical applications.
The reported technique finally enables fabrication of integrated devices combining e.g. optics with microfluidics and electronics from digital blueprints, highlighting inkjet printing as an additive manufacturing platform ideal for rapid prototyping by uniting various materials and functionalities. Through its simplicity, the method can finally be adapted to multi-nozzle ink jet, electro-hydrodynamic and aerosol printers with higher resolution or throughput.
The authors declare that there are no conflicts of interest related to this article.
References and links
1. J. U. Park, M. Hardy, S. J. Kang, K. Barton, K. Adair, D. K. Mukhopadhyay, C. Y. Lee, M. S. Strano, A. G. Alleyne, J. G. Georgiadis, P. M. Ferreira, and J. A. Rogers, “High-resolution electrohydrodynamic jet printing,” Nat. Mater. 6(10), 782–789 (2007). [CrossRef] [PubMed]
2. J. Pang, J. N. Stuecker, Y. Jiang, A. J. Bhakta, E. D. Branson, P. Li, J. Cesarano 3rd, D. Sutton, P. Calvert, and C. J. Brinker, “Directed aerosol writing of ordered silica nanostructures on arbitrary surfaces with self-assembling inks,” Small 4(7), 982–989 (2008). [CrossRef] [PubMed]
5. L. Wu, Z. Dong, F. Li, H. Zhou, and Y. Song, “Emerging progress of inkjet technology in printing optical materials,” Adv. Opt. Mater. 4(12), 1915–1932 (2016). [CrossRef]
6. C. Zhang, C.-L. Zou, Y. Zhao, C.-H. Dong, C. Wei, H. Wang, Y. Liu, G.-C. Guo, J. Yao, and Y. S. Zhao, “Organic printed photonics: From microring lasers to integrated circuits,” Sci. Adv. 1(8), e1500257 (2015). [CrossRef] [PubMed]
9. W. Su, B. S. Cook, Y. Fang, and M. M. Tentzeris, “Fully inkjet-printed microfluidics: a solution to low-cost rapid three-dimensional microfluidics fabrication with numerous electrical and sensing applications,” Sci. Rep. 6(1), 35111 (2016). [CrossRef] [PubMed]
10. I. Papakonstantinou, K. Wang, D. R. Selviah, and F. A. Fernández, “Transition, radiation and propagation loss in polymer multimode waveguide bends,” Opt. Express 15(2), 669–679 (2007). [CrossRef] [PubMed]
11. R. G. Hunsperger, Integrated Optics: Theory and Technology (Springer, 2009).
12. J. Alamán, R. Alicante, J. I. Peña, and C. Sánchez-Somolinos, “Inkjet printing of functional materials for optical and photonic applications,” Materials (Basel) 9(11), 910 (2016). [CrossRef] [PubMed]
13. A. Samusjew, M. Kratzer, A. Moser, C. Teichert, K. K. Krawczyk, and T. Griesser, “Inkjet printing of soft, stretchable optical waveguides through the photopolymerization of high-profile linear patterns,” ACS Appl. Mater. Interfaces 9(5), 4941–4947 (2017). [CrossRef] [PubMed]
14. S. H. Davis, “Moving contact lines and rivulet instabilities. Part 1. The static rivulet,” J. Fluid Mech. 98(2), 225–242 (1980). [CrossRef]
16. P. C. Duineveld, “The stability of ink-jet printed lines of liquid with zero receding contact angle on a homogeneous substrate,” J. Fluid Mech. 477, 175–200 (2003). [CrossRef]
17. M. W. Lee, N. Y. Kim, S. Chandra, and S. S. Yoon, “Coalescence of sessile droplets of varying viscosities for line printing,” Int. J. Multiph. Flow 56, 138–148 (2013). [CrossRef]
18. S. Schiaffino and A. A. Sonin, “Formation and stability of liquid and molten beads on a solid surface,” J. Fluid Mech. 343, 95–110 (1997). [CrossRef]
23. K. A. Brakke, “The Surface Evolver,” Exp. Math. 1(2), 141–165 (1992). [CrossRef]