Device fabrication in high-index 3D photonic crystals
O M Roche1, J Scrimgeour1, J S King2, D N Sharp2, C F Blanford3, E
Graugnard2, R G Denning3, C J Summers2 and A J Turberfield1
1
University of Oxford, Department of Physics, Oxford OX1 3PU, UK.
2
Georgia Institute of Technology, School of Materials Science and
Engineering, Atlanta, GA, 30332, USA.
3
University of Oxford, Inorganic Chemistry Laboratory, Oxford OX1 3QR, UK.
a.turberfield1@physics.ox.ac.uk
We demonstrate two key steps towards the creation of integrated optical
devices based on waveguides and microcavities operating within a complete
photonic band gap: infiltration of a holographically-defined polymeric 3D
photonic crystal template with high-index dielectric by Atomic Layer Deposition
(ALD) and creation of localised structural defects embedded in, and in
registration with, a 3D photonic crystal by direct two-photon laser writing.
Structural and optical characterisation of TiO2 photonic crystals produced by
infiltration and removal of the polymer template demonstrates the high quality
of the negative replica. Structural characterisation of photonic crystals with
embedded defects shows a faithful rendering of the designed structure in the
developed polymer photonic crystal. The combination of these three
techniques (HL, two-photon writing and ALD) maps out a clear route to device
fabrication in high-index 3D photonic crystals.
1. Introduction
A photonic crystal is a dielectric microstructure with periodic variation in refractive index.
Photonic crystals can control photons much like the way in which semiconductors control
electrons. They can exhibit a photonic band gap, a range of frequencies within which no
electromagnetic modes are allowed to propagate[1,2]. Microcavities[3] and waveguides[4]
operating within the optical bandgap of a photonic crystal have the potential to create
integrated optical devices capable of all-optical signal processing[5]. The demonstration of a
practical technology for optical microfabrication of three-dimensional (3D) photonic crystal
devices is a vital step in the development of these applications. It is a formidable materials
processing challenge which must be capable not only of sub-micron pattern definition in 3D
and precise local modification of the wavelength-scale microstructure but also of the transfer
of this pattern into a homogeneous dielectric with an appropriately high refractive index.
Two new optical-lithography techniques, holographic lithography[6] and two-photon laser
writing[7-9], can be combined to create a rapid and flexible method for the definition of
photonic crystal device structures in photoresist[10]. High-index inverted replicas of
holographically defined photonic crystals can be created by atomic layer deposition (ALD) of
TiO2 within the polymeric template[11]. These three techniques can be combined to create an
integrated technique for the production of optical devices.
2. Holographic lithography
Holographic lithography is a fast, flexible method for the fabrication of 3D photonic crystals
with the sub-micron periodicity required for optical and near-infrared wavelength operation[6].
An interference pattern generated at the intersection of four coherent laser beams is used to
expose a film of photoresist typically 20μm thick. Highly exposed photoresist is rendered
insoluble while unexposed areas are dissolved away leaving a polymeric photonic crystal
with interconnected air voids. We use a negative, epoxy-based photoresist, SU-8, containing
a photoacid generator (PAG). Absorption of a UV photon by the PAG liberates a hydrogen
ion. Acid-catalysed polymerisation occurs when the film is heated in a post-exposure bake.
Structural characterisation of the resulting films in a scanning electron microscope (SEM)
show that holographic lithography allows the creation of large, well-defined 3D photonic
crystals which are defect-free over hundreds of lattice constants, see figure 1.
Figure 1 SEM image of a developed epoxy photonic crystal
showing a (111) top surface and a (11 1 ) cleavage plane.
3. Atomic layer deposition
Polymeric resists for 3D microfabrication have refractive indices, n, in the range 1.4-1.6,
which is too low for most photonic crystal device applications. Devices based on waveguides
and microcavities embedded within a photonic crystal[5] are designed to operate at
frequencies within a complete (omnidirectional) photonic band gap in order to suppress
radiative loss[3]; to create a complete photonic band gap, even in an optimized air-dielectric
structure, a refractive contrast of at least 1.9 is necessary[12-14]. This limitation may be
overcome by using the polymeric structure as a template for infiltration to create a negative
or (via a two-stage process) positive replica in a high-index dielectric. The infiltration
technique adopted must be capable of filling a multiply connected network of sub-micron
channels to produce a dense and homogeneous network of transparent amorphous or
nanocrystalline material, extending uniformly through a layer many cells deep. It must also
be compatible with the low thermal decomposition temperature (~300oC) of the epoxy resist
used for holographic lithography[6]. We have explored the potential of ALD[15], a method that
has been used successfully for the formation of single- and multi-component inverse opals[1620]
. ALD utilises sequential reactant pulses, separated by inert gas purges, to force surfacelimited reactions, thereby allowing the growth of dense, highly conformal films with submonolayer control.
Amorphous TiO2, the material that we have chosen, is deposited using alternating pulses
of TiCl4 and water vapour[21]. ALD-deposited amorphous TiO2 is dense and homogeneous,
with n  2.3 and low absorption down to UV wavelengths. SEM images of the resulting
structure are shown in figure 2. These clearly show the successful formation of a uniform,
high-quality TiO2/air photonic crystal and complete removal of the polymer template.
Figure 2 (a) Top surface, (111) plane, of the inverse replica after removal of the polymer,
showing uniform and conformal coating. (b) Image of cleavage plane, (11 1 ) plane, of inverse
replica showing uniform and homogeneous infiltration throughout the depth of the film.
Optical properties of the polymer template and inverse replica were obtained via normal
incidence transmission and reflectivity measurements. Figure 3a shows reflectivity and
transmission spectra of the polymer template. A peak in reflectivity and a corresponding dip
in transmission were observed at ~1485nm and 1430nm, respectively, close to the calculated
position of the fundamental (111) stop band; a strong reflectivity peak and transmission dip at
790nm correspond to the second-order stop band (calculated stop bands for an infinite
photonic crystal are indicated by shading). Figure 3b shows the optical properties of the final
TiO2 photonic crystal. As expected, the stop bands have shifted to longer wavelengths (1600
and 828nm) in excellent agreement with band-structure calculations. Evidence for the optical
quality of the inverted photonic crystal is provided by the increase in peak reflectivity from
20% for the template to 65% for the TiO2/air structure and by the consistent agreement
between observed stop-band positions and the predictions of band structure calculations.
Figure 3 Microreflectivity (R) and microtransmission (T) for (a) a holographically defined
polymer template and (b) a TiO2 photonic crystal after the template has been removed. The
shaded bands indicate the calculated position and width of the fundamental and second order
(111) photonic band gaps.
4. Two-photon writing
Incorporation of passive devices in 3D photonic crystals can be achieved by using twophoton absorption of a laser writing beam. 3D photonic-device structures must be aligned
and registered with the surrounding photonic crystal with a tolerance that is much smaller
than the operating wavelength and comparable to the overlay accuracy of current
semiconductor fabrication[22]. By imaging the latent photoacid pattern created by a
holographic exposure, it is possible to align subsequent two-photon exposures to create
device structures embedded in, and in registration with the 3D photonic-crystal lattice. Acidcatalysed polymerisation of a holographic exposure does not occur until the film is heated in
a post-exposure bake. By incorporating a fluorescent, acid-sensitive dye in the photoresist
and using a confocal fluorescence microscope we can produce a 3D map of the photoacid
distribution[23], which can be used to align the two-photon device-writing process. Figure 4
shows confocal fluorescence images of an undeveloped, holographically patterned film of
photoresist. They correspond closely to the SEM image of the developed structure, figure 1,
demonstrating that acid indicator fluorescence from an undeveloped film provides a latent
image of the developed structure[23] with sufficient resolution to align the subsequent twophoton exposure step.
Figure 4 (a) (111) and (b) (211) confocal fluorescent sections through an undeveloped,
holographically patterned film of photoresist; an indicator dye reveals the distribution of
photoacid, a latent image of the undeveloped structure. The section in (a) was recorded
at a depth of 3μm below the top surface. Scale bar is 5μm.
Figure 5 shows a test structure in the form of a Mach-Zehnder interferometer at a depth of
5μm in a 12μm film. Waveguides were created by scanning the focused writing beam along
rows of “atoms” in a (111) plane. After development the embedded device was imaged using
a confocal microscope by back-filling air spaces with a fluorescent dye, rhodamine B,
dissolved in 1-ethoxy naphthalene to match the refractive index of the epoxy, see figure 5c.
This confirms that all stages in the fabrication process – holographic patterning, registration,
direct two-photon laser writing, and the development of the composite device structure –
worked as designed.
Figure 5 (a) stack of confocal (111) sections
through the indicator fluorescence intensity in
the unpolymerised film. Excellent registration
between the waveguides and the underlying
photonic
crystal
lattice
is
observed.
Comparison of confocal sections through
fluorescence from the indicator dye (b) and
the infiltrated dye (c) showing the test
structure embedded within the photonic
crystal host.
5. Conclusion
In summary, we have shown that holographic lithography can be combined with direct twophoton laser writing to produce photonic crystals containing precisely localised, truly 3D
structural modifications. This process has sufficient resolution to create embedded
waveguide and micro-cavity structures by selectively modifying individual unit cells in the
bulk of a 3D photonic crystal. We also show that holographically generated photonic crystals
can be used as templates for infiltration of high-index dielectrics by atomic layer deposition.
Therefore, by incorporating engineered defects in the template it will be possible to create
device structures within a photonic crystal exhibiting a complete photonic band gap.
Acknowledgements: This work was supported by the UK Engineering and Physical
Sciences Research Council, by the U.S. Army Research Office under MURI contract
DAAD19-01-10603, and by EC-funded Network of Excellence PHOREMOST
(FP6/2003/IST/2-511616). OMR acknowledges the support of the Scatcherd Science
Foundation.
References
[1] E. Yablonovitch, Phys. Rev. Lett., 1987, 58, 2059.
[2] S. John, Phys. Rev. Lett. 1987, 58, 2486.
[3] E. Yablonovitch, T. J. Gmitter, R. M. Meade, A. M. Rappe, K. D. Brommer, J. D.
Joannopoulos, Phys. Rev. Lett., 1991, 67, 3380.
[4] A. Mekis, J. C. Chen, I. Kurland, S. Fan, P. R. Villeneuve, J. D. Joannopoulos, Phys.
Rev. Lett., 1996, 77, 3787.
[5] M. Soljacic, J. D. Joannopoulos, Nat. Mater., 2004, 3, 211.
[6] M. Campbell, D. N. Sharp, M. T. Harrison, R. G. Denning, A. J. Turberfield, Nature,
2000, 404, 53.
[7] B. H. Cumpston, S. P. Anonthevel, S. Barlow, D. L. Dyer, J. E. Ehrlich, L. L. Erskine, A.
A. Heikel, S. M. Keubler, I.-Y. S. Lee, D. McCord-Maughon, J. Qin, H. Rockel, M
Rumi, X.-L. Wu, S. R. Marder, J. W. Perry, Nature, 1991, 398, 51.
[8] H.-B. Sun, S. Matsuo, H. Misawa, Appl. Phys. Lett., 1999, 74, 786.
[9] M. Deubel, G. von Freymann, M. Wegener, S. Pereira, K. Busch, C. M. Soukoulis, Nat.
Mater., 2004, 3, 444.
[10] J. Scrimgeour, D. N. Sharp, C. F. Blanford, O. M. Roche, R. G. Denning, A. J.
Turberfield, Adv. Mater., 2006 , 18, 1557.
[11] J. S. King, E. Graugnard, O. M. Roche, D. N. Sharp, J. Scrimgeour, R. G. Denning, A.
J. Turberfield, C. J. Summers, Adv. Mater., 2006, 18, 1561.
[12] C. T. Chan, K. M. Ho, C. M. Soukoulis, Europhys. Lett., 1991, 16, 563.
[13] M. Maldovan, A. M. Urbas, N. Yufa, W. C. Carter, E. L. Thomas, Phys. Rev. B:
Condens. Matter Mater. Phys., 2002, 65, 165123.
[14] D. N. Sharp, A. J. Turberfield, R. G. Denning, Phys. Rev. B: Condens. Matter Mater.
Phys., 2003, 68, 205102.
[15] M. Ritala, M. Leskela, in Handbook of Thin Film Materials, vol. I (Ed: H. S. Nalwa),
Harcourt Academic, London, UK 2002, Ch. 2.
M. Ritala, Appl. Surf. Sci., 1997, 112, 223.
[16] J. S. King, E. Graugnard, C. J. Summers, Adv. Mater., 2005, 17, 1010.
[17] J. S. King, D. Heineman, E. Graugnard, C. J. Summers, Appl. Surf. Sci., 2005, 244,
511.
[18] A. Rugge, J. S. Becker, R. G. Gordon, S. H. Tolbert, Nano Lett., 2003, 3, 1293.
[19] J. S. King, C. W. Neff, C. J. Summers, W. Park, S. Blomquist, E. Forsythe, D. Morton,
Appl. Phys. Lett., 2003, 83, 2566.
[20] M. Scherrer, X. Wu, A. Yamilov, H. Cao, R. P. H. Chang, Appl.Phys. Lett., 2005, 86,
151113.
[21] J. Aarik, A. Aidla, T. Uustare, V. Sammelselg, J. Cryst. Growth, 1995, 148, 268.
M. Ritala, M. Lesmela, E. Nyuanen, P. Soininen, L. Niinisto, Thin Solid Films, 1993,
225, 288.
[22] The International Technology Roadmap for Semiconductors, 2003 edition,
Semiconductor Industry Association, San Jose, CA 2003. Available from URL
http://public.itrs.net (last accessed: March 20,2006).
[23] S. J. Bukofsky, G. D. Feke, Q. Wu, R. D. Grober, P. M. Dentinger, J. W. Taylor, Appl. Phys.
Lett., 1998, 73, 408.