The spatial distribution of second-order nonlinearity in thermally poled optical fibers was characterized by second-harmonic microscopy. The second-order nonlinearity was found to be confined to a thin layer close to the anode surface and progressed further into the silica as the poling time increased. Position uncertainty of the anode metal wire was observed to have an effect, as the nonlinear layers were found not always symmetrically located around the nearest points between the anode and cathode. Optical microscopy results were obtained on etched poled fiber cross-sections and compared with those from second-harmonic microscopy.
© 2005 Optical Society of America
Vitreous silica is an isotropic material that lacks any even-order nonlinearity. The discovery of inducing second-order nonlinearity (SON) of the order of 1 pm/V in fused silica by the thermal poling method opens up the prospect of realizing electro-optic switching, modulation, and harmonic generation in silica-based materials , and has attracted extensive research efforts [2–4]. An important topic in this field is the characterization of the spatial distribution of the induced SON profile. This information is both scientifically and technically important, as it can provide much insight to the mechanism of SON formation and offers the possibility of better designing optical waveguides in which the SON is to be induced. The most frequently exploited technique is the Maker fringe method, from which the width of the SON region can be deduced . However the spatial resolution of this method is limited by the coherence length, about 24 µm, for an excitation source at 1.06 µm. Some modifications to the Maker fringe method have been suggested to improve the spatial resolution of this technique [5–7]. Another common technique requires chemical etching of the SON region with hydrofluoric (HF) acid [8,9]. The variance in the etching rate has been suggested to be an indication of the onset of the frozen-in electric field, which is believed to act on the intrinsic third-order nonlinearity of the silica to give the final induced SON. An improved method using simultaneous monitoring of the second-harmonic (SH) signal and the etched away thickness of the poled sample was subsequently proposed in an attempt to fully reconstruct the SON spatial distribution . However this removal of the silica under the anode, which contains the ions that create the field, may also alter the internal frozen-in field. Therefore, the reconstructed SON may not be a good representation of the initial SON distribution.
Recently we reported the characterization of the SON profile in thermally poled fused silica with SH microscopy with sub-micron spatial resolution . This technique measures the true spatial distribution of the induced SON no matter what the mechanism is. Here we apply this technique and present detailed results on the characterization of SON profiles in thermally poled silica fibers. Two different fibers are used. Results from SH microscopy are also compared with those from the HF etching technique.
2. Optical fibers and SH microscopy
Two silica optical fibers fabricated by the modified chemical vapor deposition method with Ge-doped cores were used in this experiment. One is a D-shaped fiber, 260 µm in diameter. There is a hole, 75 µm in diameter, in the cladding of the D fiber. The distance between the flat surface and the nearest edge of the hole is 30 µm. The fiber core, 4 µm in diameter, is 4 µm (edge-to-edge) away from the fiber flat. It should be stressed that our aim in this paper is to characterize SON profiles rather than to optimize the overlap between the SON and the fiber core (although ultimately this is important). Therefore, for convenience, the D fiber was poled with its flat in contact with the top surface of a grounded Al heater serving as the cathode. The other fiber is a twin-hole fiber. The diameters of the fiber and the holes are 230 and 94 µm, respectively. The edge-to-edge spacing between the two holes is ~16 µm. The fiber core, located between the two holes, is 4 µm in diameter, and 2 µm edge-to-edge to the anode hole and 10 µm to the cathode hole. Aluminum (Al) wires (diameter 50 µm) were inserted in the holes as electrodes.
In the experiment, a poling voltage of 3.5 kV and temperature of 250°C were chosen. Poling times varied from 10 to 90 min. To prepare samples for the SH microscopy, the poled fibers, with Al wires removed, were either embedded in epoxy inside a glass capillary or sandwiched between two supporting thin glass plates. Short sections of 1–2 mm were then cut perpendicular to the fiber axis and polished on both sides to an optical finish.
The SH microscopy was conducted with an inverted Leica DMIRBE microscope equipped with a Leica TCS2MP confocal system and Coherent Verdi-Mira tunable pulsed titanium sapphire laser. An excitation wavelength of 830 nm was used, with pulses in the 100–200 fs range. The microscope was also equipped with dual photomultiplier transmitted light detectors, with a 505 DCLP dichroic mirror dividing the detectable spectrum (380–680 nm) at 505 nm between the two channels. The detecting spectrum range of the longer wavelength channel (channel 1) is from 505 to 650 nm. This channel will therefore receive two-photon fluorescence over a wide range. In some cases we also used it to detect a transmitted non-confocal image of the sample using 543 nm helium-neon laser light. The shorter wavelength channel (channel 2) was fitted with a 415/10 nm narrow bandpass filter to receive only the second harmonic signal and reject any two-photon excited fluorescence. A×20 plan-fluorite objective (NA 0.5) was used, with a dry/oil-immersion NA 0.93/1.4 condenser used dry. The spatial resolution of the SH microscopy was estimated to be around 0.6 µm. The excitation laser beam was mainly linearly polarized along a direction 15° counter-clockwise relative to the horizontal direction of the micrographs.
3. Results and discussions
3.1 Typical SH microscopy results
A typical SH microscopy result is shown in Fig. 1. The D fiber was poled for 30 min. The epoxy in which the D fiber was embedded generated fluorescence and appears bright in the micrograph, while the D fiber itself appears darker. It can be seen that the induced SON, the bright area in Fig. 1(b), is distributed in a thin layer beneath the anode surface. At about the center of the SON layer, the SH intensity appears weaker, presumably because the frozen-in field there is almost perpendicular to the polarization direction of the excitation laser. Note that the SON layer is not symmetrical around the line joining the closest points between the anode hole and the flat. This will be discussed in the following subsection.
The spatial profile of the SON layer can be obtained by doing a line-scan across the SON layer with starting and ending points in the epoxy and fiber, respectively. For every point along the scanning line, the detected SH intensity is related to the SON χ (2) through P(2ω)∝(P(ω)χ (2))2. Therefore, the SON profile can be obtained by taking the square root of the measured SH signal. The result is shown in Fig. 2. The SON profile is not symmetric but sharper on the side closer to the anode and tails more slowly into the bulk of the silica fiber. This is expected, as the frozen-in field is generally believed to be formed between an externally injected hydrogenated species such as H+ or H3O+ and the immobile negative charges left behind in the so-called depletion region after the mobile cations (mainly alkali ions like Na+) have migrated to the cathode . As the mobility of the hydrogenated species is 3 to 4 orders of magnitude lower than that of Na+, in the short poling time, they are distributed only in a very thin layer beneath the anode. Suppose the density of the negative charges in the depletion region is N, then the theory-predicted SON profile should be featured by a step jump to its peak value at the hydrogenated species layer, followed by a linear decrease deeper into the sample with a slope of eN/ε, where ε is the dielectric permittivity of the silica glass. The experimental result is in accordance with this trend. It is also consistent with the result reported by others through simultaneously monitoring of both the SH signal and the removed thickness of the poled sample in HF . The full width at half maximum (FWHM) of this profile was measured to be around 3.2 µm.
The position of the SON layer relative to the anode surface can also be deduced from the line-scan results. An interesting point is to see how the SON layer shifts away from the anode with poling time. To do this, we poled a few samples for different time periods. The result is shown in Fig. 3. The SON layer moves further into the bulk of the fiber when the poling time increases. This can be explained by the migration of the less mobile hydrogenated species to the cathode driven by the strong internal electric field existing within the depletion region.
3.2 Effect of the uncertainty in the position of the metal wire on the SON profile
It has been briefly mentioned that the SON layer is not always symmetrically distributed around the line joining the closest points between the anode hole and the flat. Our hypothesis is that this is due to varying position of the metal wire electrode in the anode hole caused by the curving of the Al wire, as the wire (50 µm in diameter) is smaller than the hole. In the process of inserting the Al wire into the hole, the wire inevitably curved slightly and was not always located in the exact geometric center of the hole. Such configuration will result in the possibility that the maximum electric field is formed elsewhere from the closest region between the anode hole and the flat. In some cases we even observed a ring shaped SON profile, as shown in Fig. 4. Because it is desirable to achieve optimum overlap between the SON and the fiber core, this uncertainty will need to be reduced to manufacture repeatable devices. Choosing a metal wire with a diameter closely matching that of the hole or even using a liquid metal wire may help solve this problem .
3.3 Comparison between SH and HF etching results
The etching rate of silica glass in HF is affected by the presence of electric field and this has been exploited to characterize the SON region both in bulk silica and optical fibers [9,14,15]. As the SH microscopy presents the real visualization of the SON profile, it is interesting and important to compare the results from these two techniques.
To do this, a poled fiber sample was first characterized with SH microscopy. Then it was etched in 10 % (in volume) HF acid for 3 min and observed with an optical microscope (OM). Results are shown in Fig. 5. In Fig. 5(a) is the SH overlay image from channels 1 and 2. The green line across the SON layer is the scanning line for profile quantification. For a proper comparison, the sample on the specimen stage of the OM was so adjusted as to achieve the same position relative to the micrograph as in SH microscopy. Careful comparison between Figs. 5(a) and 5(b) shows that the results about the SON distribution from the two techniques are in excellent agreement. But of course, the SH microscopy can provide more information about the SON profile. It can be also seen that, there is a discontinuity on the etched layer, which can also be vaguely seen in the corresponding section of the SON layer from the SH microscopy, where the SH signal is very weak due to the perpendicularity between the frozen-in electric field and the polarization of the input fundamental laser beam.
3.4 SON in thermally poled twin-hole fibers
As well as D fiber, we also thermally poled twin-hole fibers. Typical results are shown in Fig. 6. Again the fluorescence from the epoxy has been used to obtain the optical image of the sample geometry. The SON layer is found, again, only around the anode hole, and not always symmetrically distributed between the two holes. The FWHM and distance to the anode edge of the SON layer were measured and found to be around 4.3 and 4.0 µm. These results on twin-hole fiber are consistent with those on D fiber.
We have investigated the SON profile in detail in thermally poled optical fibers. The SON layer was only found beneath the anode and was 3–5 microns wide. Its position relative to the anode edge was poling time dependant and its time evolution has been characterized in the D fiber. The position of the metal wire in the anode hole has an influence on the SON profile, often causing it to be asymmetrically distributed along the line connecting the closest points between the anode and cathode. Comparison of the results for SON distribution between the SH and chemical HF etching shows very good agreement between the two techniques in locating the SON layer position.
The authors would like to thank Ellie Kable of the Electron Microscope Unit at the University of Sydney for her assistance in SH microscopy.
References and links
3. D. Wong, W. Xu, S. Fleming, M. Janos, and K.M. Lo, “Frozen-in electrical field in thermally poled fibers,” Opt. Fiber Technol. 5, 235–241 (1999). [CrossRef]
4. J. Arentoft, M. Kristensen, K. Pedersen, S.I. Bozhevolnyi, and P. Shi, “Poling of silica with silver-containing electrodes,” Electron. Lett. 36, 1635–1636 (2000). [CrossRef]
5. C. Corbari, O. Deparis, B.G. Klappauf, and P.G. Kazansky, “Practical technique for measurement of second-order nonlinearity in poled glass,” Electron. Lett. 39, 197–198 (2003). [CrossRef]
6. A. Ozcan, M.J.F. Digonnet, and G.S. Kino, “Inverse Fourier transform technique to determine second-order optical nonlinearity spatial profiles,” Appl. Phys. Lett. 82, 1362–1364 (2003). [CrossRef]
7. A. Ozcan, M.J.F. Digonnet, and G.S. Kino, “Improved technique to determine second-order optical nonlinearity profiles using two different samples,” Appl. Phys. Lett. 84, 681–683 (2004). [CrossRef]
9. T.G. Alley and S.R.J. Brueck, “Visualization of the nonlinear optical space-charge region of bulk thermally poled fused-silica glass,” Opt. Lett. 23, 1170–1172 (1998). [CrossRef]
10. A. Kudlinski, Y. Quiquempois, M. Lelek, H. Zeghlache, and G. Martinelli, “Complete characterization of the nonlinear spatial distribution induced in poled silica glass with a submicron resolution,” Appl. Phys. Lett. 83, 3623–3625 (2003). [CrossRef]
11. H. An, S. Fleming, and G. Cox, “Visualization of second-order nonlinear layer in thermally poled fused silica glass,” Appl. Phys. Lett. 85, 5819–5821 (2004). [CrossRef]
12. T.G. Alley, S.R.J. Brueck, and R.A. Myers, “Space charge dynamics in thermally poled fused silica,” J. Non-Cryst. Solids 242, 165–176 (1998). [CrossRef]
13. M. Fokine, L.E. Nilsson, Å. Claesson, D. Berlemont, L. Kjellberg, L. Krummenacher, and W. Margulis, “Integrated fiber Mach-Zehnder interferometer for electro-optic switching,” Opt. Lett. 27, 1643–1645 (2002). [CrossRef]
14. P. Blazkiewicz, W. Xu, D. Wong, and S. Fleming, “Mechanism for thermal poling in twin-hole silicate fibers,” J. Opt. Soc. Am. B 19, 870–874 (2002) [CrossRef]
15. N. Myrén, H. Olsson, L. Norin, N. Sjödin, P. Helander, J. Svennebrink, and W. Margulis, “Wide wedge-shaped depletion region in thermally poled fiber with alloy electrodes,” Opt. Express 12, 6093–6099 (2004), http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-25-6093. [CrossRef] [PubMed]