Double carbon 17/9/2006

Submerged waveguide constructed by the implantation of 12C ions in electrooptic crystals.

Har’el Ilan, Alexander Gumennik, Roei Fathei, and Aharon J. Agranat

Department of Applied Physics,

TheBeninSchool of Engineering and Computer Science,

TheHebrew University of Jerusalem, Jerusalem93707, Israel.

Igal Shachar and Michael Hass

Department of Particle Physics, Faculty of Physics,

Weizmann Institute of Science, Rehovot76100, ISRAEL.

Abstract:

Waveguide structures were fabricated in Potassium Lithium Tantalate Niobate crystals by the implantation of high energy 12C ions. The implantation forms an amorphous layer with a lowered index of refraction within the depth of the crystal, which serves as the cladding of the waveguides. Two amorphous layers were fabricated at 18.8 m and 25.6 m below the surface of the crystal by implantation at 30MeV and 40 MeV respectively. This formed a submerged slab waveguide sandwiched between the two amorphous layers, and two slab waveguides that were formed between the surface of the crystal and each of the amorphous layers. Coupling between those waveguides was observed and investigated, and confinement of the light in the sandwiched waveguide was demonstrated.

It is well established that the implantation of high energy light ions such as H+ and He++ in electrooptic crystals forms an amorphous layer beneath the surface of the crystal1. Such amorphous layers characterized by a lowered refractive index (RI),serve as a cladding for constructing waveguiding structures in the crystal. In particular it was shown2 that the bombardment of high energy He++ ions into aPotassium Lithium Tantalate Niobate (KLTN) crystal results with the formation of a slab waveguide beneath the surface of the crystal, with an amorphous cladding layer that hasa lowered RI up to 4%. The amorphous layer that was formed in the KLTN substrate exhibited thermal stability after annealing. Thus, the formation of lowered RI regions by selective implantation can be the basis for a fabrication method ofreliable electrooptic devices in waveguided architectures.

In this paper we report the fabrication of slab waveguidesusing the implantation of high energy Carbon ions of 12Cisotopes into KLTNcrystallinesubstrates. In particular, a slab waveguide sandwiched between two amorphous cladding layers positioned well within the depth of the substrate was fabricated by successive implantation at two energies.

The interaction between the high energy Carbonion and the mediumin which it propagates is governed by two distinctmechanisms: electronic stopping and nuclearstopping3. Upon penetrating the material, during the electronic stoppingphase, the high energyCarbonion propagates through thecrystal without diverting from its original direction andgradually slows down. At a sufficiently low energy thecross section for scattering by the lattice ions increases steeply.In this nuclear stopping phase the propagating iontears a lattice ion from its site causing it to become an interstitialion at a different site. Thus, the nuclear stopping phasegenerates a well-defined amorphous layer4,5of Frenkeldefects6 with an RI that is significantly lowerthan that of the substrate crystal.

The implantations were preformed in two substrates of KLTN. All of the samples where bombarded with a beam of high energyCarbonions generated by a tandem Van de Graff (VDG)accelerator. During the implantation the samples where kept at constant temperature of 25oC. The samples compositions and implantation data is summarized in Table1. The predicted spatial distribution of the RI was obtained by TRIM simulation and is presented in Fig. 1 for samples S1 and S2. (The simulations were preformed by TRIM version SRIM 20037.) The reflection spectrum of the TE modes of the samples was measured using the prism-coupler setup2, 8, 9 operated at.

Table 1

Sample / Substrate / Tc
[Co] / 1stEnergy
[MeV] / Dosage
[ions/cm2] / 2nd Energy
[MeV] / Dosage
[ions/cm2]
S1 / K1(-)T0.68N0.32O3:Li / -9 / 30 / / - / -
S2 / K1(-)T0.65N0.35O3:Li / 11.2 / 30 / / 40 /
S3 / K1(-)T0.68N0.32O3:Li / -9 / 40 / / 30 /

* Li concentration is estimate at 0.3% per mol.

Consider first sample S1. In this sample one amorphous layer was formed by the bombardment of Carbon ions at 30 MeV. The reflection spectrum of the TE modes propagating in the waveguide that wasformed by the implantation is shown in Fig.2a. The RI profileperpendicular to the surface was reconstructed from the spectrum by the method of Chandler and Lama10and isshown in Fig.1a. The maximumdecrease of the RI in the amorphous layer that was formed by the implantationrelative to the RI in the KLTN substrate was 3.6%. The amorphous layer was found to be 17.7 below the surface of the substrate with a width of 0.6, defined as the full width of half maximumFWHM.Note that the RI profile derived from the modes spectrum is slightly shifted from the respective profile derived from the TRIM simulation. This is attributed to the fact that the TRIM simulation uses the density and the composition of the substrate, but does not take into consideration the specifics of its crystalline structure.

Consider next sample S2. As shownin Table 1this sample was implanted with ions at 30 MeV and 40 MeV, and therefore has two amorphous layers that were formed at different depths.This leadsto a profile of RI that forms three slab waveguides: waveguide Aimmediately beneath the surface of the substrate cladded between the air and the first amorphous layer, waveguide Bsubmergedin the depth of the substrate cladded between the two amorphous layers, andwaveguide C cladded between theair and the second amorphous layer.The predicted geometry of thethree waveguides as predicted by the TRIM simulation ispresented in Fig. 1b.

It should be noted that a prism coupler can excite modes that propagate adjacent to the surface of the crystal. It was a priori expected that the reflection spectrum of the TE modes shown in Fig.2bwould be the combined spectrum of modes of waveguidesA and C, that are cladded between the surface and the first and the second amorphous layers respectively.It was also expected that the modes of waveguide A should closely resemble those of the waveguide that was formed in S1. The latter waveguide is cladded between the surface of the substrate and the amorphous layer that was formed by ions implanted at 30MeV. It is therefore expected to have the same structure as waveguide A in S2.

By comparing the spectrum of S2 to the spectrum of S1it was possible to remove the modes that belong to waveguide A. these modes are indicated by (*) in Fig. 2b. It was expected a priori that the remaining modes should be those of waveguide C. Furthermore, it was also expected that the remaining modes should be denser then the modes of waveguide A because of the latter smaller width. However, the remaining modes (that are left unmarked in Fig. 2b) are in fact spaced more sparsely than the modes of waveguideA and therefore belong to a waveguide with a smaller width. The waveguide profile that was extracted from these modes indicates that these modes belong infact to waveguide B, i.e. the waveguide that is buried in the depth of the substrate cladded between the two amorphous layers. As the prism coupler cannot excite modes that propagate in a buried waveguide,it is concluded that a coupling occurs between waveguides A and B, enabling the excitation of modes in waveguide B directly by the prism coupler. The basic mode (mode 0) of this waveguide is not excited, and modes 1÷7 are excited with an increasing magnitude due to the fact that the efficiency of the coupling between close adjacent waveguidesincreases with the mode number. The fact that the modes of waveguide C were not observedis due to shadowing of the second (deeper) amorphous layer by the first (shallower) layer.

The reconstruction of the RI profile in sample S2 is shown in Fig.1b and compared to the prediction derived bythe TRIM simulation. The first amorphous layer is formed 17.7below the surface of the substrate and has a width of 0.6 (defined as FWHM)identical to S1, the second amorphous layer was formed 26.1below the surface of the substrate with a width of 0.65.The two layers form the cladding for a buried waveguide with a width of 8.4. The peak of the relative decrease of the RI in the amorphous layers that was formed by the implantation was found to be 2.8% for the first amorphous layer and 4.8% for the second layer.

In order to substantiate further the occurrence ofcoupling between waveguides A and B, sample S3was prepared bythe implantation ofCarbon ions at 30 MeV and 40 MeV, the same energies as S2, but, with a much higher dosage: [ions/cm2] and [ions/cm2] respectively. The influence of the higher dosage is on the barrier between those waveguides. Higher dosage of the implantation causes higher decrease of the RI, and therefore a higher barrier between the waveguides. Indeed the modes spectrum of S3 (Fig. 2c),was found to be similar to that of S2 with one difference: the observed modes of the submerged waveguide started atmode 3, whereas in S2 they started at mode 1. This clearly indicates that the coupling between waveguide A and B is governed by the height of the barrier between these waveguides as predicted.

Direct observation of the light-wave as it propagates through the waveguides was obtained by illuminating the input plane of the waveguides while observing the light as it emerges from the output plane using a microscope attached to a digital imaging system.In Fig. 3a a section of the output plane of S2 is shown while the input plane was illuminated by the white light source of the microscope. The amorphous layers that were formed by the implantation are clearly seen at 17.7 m and 26.1m below the surface of the crystal. In Fig. 3b the same section of the output plane of S2 is shown while light was coupled directly into waveguide B that was formed between the two amorphous layers.The light source was a diode laser operatingthat was focused by a tapered lensed fiber into a spot of less than 5 m in diameter. It can be seen that the light propagating in the sample is confined between the two amorphous layers that function as the cladding of waveguide B.

In conclusion, carbon ions are found to be effective projectiles for the formation of cladding layers within the depth of electrooptic crystals, possessing some inherent advantages over the use of lighter ions such as H+ and He++. In general, the ratio between the width of the formed amorphous layer and the depth at which it is formed is inversely proportional to the mass of the implanted ion. For implantation processes in KLTN, it was found that for carbon implantation this ratio is approximately 3%, whereas for He it was found to be2 approximately 10%. Furthermore, a projectile ion with a higher mass, creates during the nuclear stopping phase a large number of Frenkel defects. This enables the generation of a large change in the RI with a low dosage of implantation. For example, it was found that the generation of a cladding layer with a relative RI change of 3.6% required the implantation of carbon with a dosage of 0.63·1015 cm2. Generating a relative change of 4% of the RI using He++ projectile ions2 required implantation with a dosage of 1016 cm2. Consequently, the use of carbons as the implanted projectile ion enables the formation of thin cladding layers well within the depth of the substrate using a moderate dosage, while maintaining high relative change in the index of refraction.

These results clearly indicate that the implantation of high velocity ions, and in particular carbon ions can be the basis for a generic fabrication method waveguides in a KLTN crystalline substrate. The implantation at different energies enables the formation of well defined thin cladding layers well within the depth of the substrate. The RI barrier height of these layers can be preprogrammed by governing the dosage of the implantation process. Exploiting the fact that the KLTN has a strong electrooptic effect together with the possibility of lateral control of the implantation opens the way for the construction of optoelectronic circuits with multilevel waveguiding structure.

Acknowledgement: This research was supported in part by Grant No. 2002341 from the United States-Israel Binational Science Foundation (BSF), Jerusalem, Israel.

References:

1D. Kip, Appl. Phys. B 67, 131 (1998).

2A. Gumennik, A. J. Agranat, I. Shachar, and M. Hass, Applied Physics Letters 87
(2005).

3Emanuele Rimini, Ion Implantation: Basics to Device Fabrication

(Kluwer Academic, Dordrecht, 1995), Chap. 3.

4J. Y. C. Wong, L. Zhang, G. Kakarantzas, P. D. Townsend, and P. J.

Chandler, J. Appl. Phys. 71, 49 _1992_.

5T. Pliska, D. H. Jundt, D. Fluck, and P. Gunter J. Appl. Phys. 77, 6114

(1995).

6G. H. Kinchin and R. S. Pease, Rep. Prog. Phys. 18, 1 (1955).

7J. F. Ziegler, J. P. Biersack, and U. Littmark, The Stopping and Range of

Ions in Solids (Pergamon, New York, 1985), Vol. 1.

8R. Ulrich and R. Torge, Appl. Opt. 12, 2901 (1973).

9J. Seligson, Appl. Opt. 26, 2609 (1987).

10P. J. Chandler and F. L. Lama, Opt. Acta 33, 127 (1986).

Figures


1