We report that it is possible to create a fiber electret by having both internal electrodes of a twin-hole fiber at the same anodic potential, i.e., without the use of a contacted cathode electrode. We find that a stronger and more temperature-stable charge distribution results when the fiber core is subjected to an external field near zero. Negative charges from the air surrounding the fiber are sufficient for the recording of an electric field across the core of the fiber that is twice stronger than when one anode and one cathode electrode are used. The enhancement in stability and in the strength of the effective χ(2) induced are a significant step towards the wider use of fibers with a second order optical nonlinearity.
© 2009 OSA
Early in school one learns that the positive and the negative electrode of the power supply must be contacted to define the potential and ensure electrical current flow. Thus, birds stand on high voltage cables, apparently unharmed. Electret formation in heated glass fibers, named thermal poling [1–3], involves applying high voltage to a pair of internal electrodes on each side of the fiber core causing the displacement of cations rendered mobile in the glass matrix by the relatively high temperature (~280 °C). With the application of high voltage (~4 kV), positive ions (e.g., Na+) migrate towards the cathode . Provided no ionic replacement occurs at the electrode, a few-microns thick depleted region is formed adjacent to the anode. A large fraction of the potential difference applied externally can be made to fall across this thin region, and the recorded field then becomes very large (Erec >108 V/m). A second-order optical nonlinearity can thus be created through χ(2) eff ~3 χ(3) Erec if the fiber is cooled with the voltage bias still on . The strength of the field recorded is proportional to the negative charge density stored in the glass. If this is increased, the nonlinear optical effect induced is made stronger. Here, we report the counter intuitive observation that two electrodes at the same high voltage positive potential are more efficient and create a more stable charge distribution than when a positive and a negative electrode are used (see Fig. 1 ). A stronger electret is thus created when the fiber core is subjected to an external field near zero. Since here only one pole is used for both internal electrodes, the process is more appropriately described by “charging”. The term “poling” is reserved for when a positive and a negative electrode are used for electret formation. The increased nonlinearity and stability of components fabricated in this way are an important step towards the wider use of electrooptical and frequency doubling fibers.
2. Charging fibers
In fibers poled for sufficiently long times it is believed that the displaced cations of high mobility are replaced by other ions of lower mobility (e.g., H3O+) [5,6]. These partly neutralize the depleted region , but negative sites remain mainly distributed along a thin buried layer (width < 1 μm) [8–10]. After poling is completed, this thin negatively charged layer attracts positive charge to the surface of holes. The recorded field then points from the anode electrode to the buried negative layer and also from the cathode electrode to the buried negative layer . Previous experiments on poled metal-filled fibers (one anode and one cathode) have showed that the depletion region is created not only between the anode and the cathode, but also between the anode and the outer surface of the fiber, which is at a floating potential (not electrically contacted and with the primary acrylate coating still on). This is reflected by the fact that etched poled fibers often show a ring [12,13] or an edged shape  around the entire hole containing the anode electrode. Calculations of the field distribution across the fiber reproduce the edged shape obtained from etching measurements in  when the fiber surface is assumed to be electrically grounded. Apparently, air supplies the required negative charge through the acrylate coating, effectively grounding the surface even without a physical electrical contact. This suggests that a fiber could be poled (charged) by establishing the required electric field between a single positive internal electrode and the non-contacted surface of the fiber.
In order to test this assumption, a piece of fiber with two holes (see photo in Fig. 1) was provided with a single AuSn electrode connected to a potential 4.3 kV. The fiber was heated for 33 minutes at 255 °C resting on the metallic surface of a hot plate and subsequently cooled to room temperature with the voltage on. The charge distribution was studied by cleaving the fiber and etching it for 45 seconds in 40% HF. The etched surface was examined with a 400 × magnification phase contrast microscope and photographed. The result, shown in Fig. 2(a) , shows a ring structure encompassing the anode electrode and the core without clear dependence of the strength of the recording field established between the metal and the fiber surface. Because one contact is deliberately missing here, this fiber could not be immediately tested for the electrooptical effect, but it is believed that the ring indicates where the net negative charge is accumulated, marking the end of the depletion region formed around the anode [12–15]. In applications where the electrodes are not required after poling, such as in electric field sensing or frequency doubling, charging with a single electrode render unnecessary some of the efforts to coat the outside of the fiber with a conductive film in order to pole long fiber lengths, as recently reported [16,17].
The lack of dependence of the recording field strength suggests, in the extreme case, that a depletion region may be formed even if two internal electrodes are given the same positive high voltage potential. If at all, electret formation would then occur with null potential difference between them, relying on two-dimensional ionic transport towards the surface of the fiber. This was tested experimentally. A fresh piece of the same fiber was provided with a pair of AuSn electrodes, both contacted to the anode of a power supply at 4.3 kV as shown in Fig. 1. The fiber was charged, etched and examined as described above, and the result shown in Fig. 2(b). The cross section now shows two ring structures, each one encompassing one electrode. There is no indication that less charge is displaced along the line between the centers of the electrodes where the applied external field is close to zero. The core is inside one of the rings, by designing the fiber with one hole nearer to the core (the holes have diameter 30 μm, the core 3.4 μm, and the edge-to-edge distance between core and holes is 5.5 µm and 9 μm, respectively). If the annular structure indicates the thin layer of negative charge created for long charging time, one could expect that the electrooptical effect experienced by the core in the charged fiber would be up to twice stronger than in conventional poling. This results from the fact that the field established across the core becomes the superposition of two electric field components of similar magnitude and same sign, each one arising from the displacement of charge from one anode.
The measurement shown in Fig. 2 is destructive. Therefore, the induced electrooptical effect was evaluated at room temperature after charging at 265 °C new pieces of fiber with two anodes at the same potential (4.7 kV) and contacting the electrodes in the conventional way, i.e., each one to one pole of a power supply. The charged fiber was used as the active arm of a fiber Mach-Zehnder interferometer (MZI), where the phase shift induced by voltage is converted to amplitude modulation. By ramping the voltage, a phase shift of several π radians can be measured. Figure 3 illustrates such a measurement. Each dot in the graph corresponds to an additional π-phase shift. The total phase excursion measured from −6.5 kV to + 6 kV was 110 π-radians. The parabolic Kerr response of the medium undergoes its minimum when the field created by the external voltage bias (Eappl) cancels the field created by the quasi-permanent charge distribution (Erec). From the plot in Fig. 3 one sees that the applied voltage needs to be as high as 11.7 ± 0.2 kV to cancel the field recorded across the core. This is accomplished with a bias potential + 4.7 kV. The effective second order nonlinear coefficient is 0.25 pm/V.
More than 30 pieces of fiber from 4 different fiber types have been charged with the two-anode procedure above. The comparison of results obtained for conventional poling for pieces of the same length and bias voltage invariably gives larger effective second-order optical nonlinearity with the new technique (typically ~1.5-2 times larger). The fiber being charged usually rests on the metal surface (hot plate) providing for the negative charges to the fiber surface. Fibers have also been wound around a metallic wire inside an oven, with similar results. Both procedures shorten the charging time and increase the reproducibility of results from sample to sample. However, fibers have also been charged free-standing in an oven far (>20 cm) from the neighborhood of any metal. In this case, the process takes much longer, but a second order nonlinearity is also induced.
The thermal stability of fibers charged with two internal anodes was studied and compared to that of fibers poled in the conventional way with one internal anode and one cathode [2,3]. It is expected that erasure of the strong recorded electric field arises from the displacement of mobile cations that return to the negative sites and cancel the charge that creates the recorded field . Isothermal measurements were carried out at 250 °C, and the results are illustrated in Fig. 4 .
Since here the mobile cations are removed from both sides of the negative charged layer, the expected stability is improved. This is indeed observed, as shown in Fig. 3. The second order nonlinear coefficient drops from ~0.15 pm/V to ~0.08 pm/V in one minute in the poled fiber, while it only drops from 0.22 pm/V to 0.12 pm/V after 15 minutes in the charged fiber. This fast component could be associated with high mobility ions, such as Na+. However, a high temperature resistant component is available in both cases, possibly associated with hydrogen-related ionic species (e.g., H3O+ ). The two curves nearly overlap for an annealing time 1000 minutes at 250 °C without external voltage bias. A residual nonlinearity ~0.04 pm/V is measured.
The results illustrated in Figs. 2 and 3 confirm that a strong electric field can be recorded even when the two internal electrodes are at the same high voltage potential. The most likely explanation encountered for this intriguing effect is the avalanche-like positive feedback mechanism behind ionic charge movement in glasses. A two-dimensional plot of the equipotentials in a fiber can be obtained from Gauss’ law before charge transport starts, assuming that the surface of the fiber is at ground potential and that both internal electrodes are at the same high voltage potential (say, 5 kV). This is shown in Fig. 5a . Because of the finite distance between the fiber’s mid-point and the surface, the potential at the fiber centre is ~500 V lower than at the electrodes. Ionic charge transport at high temperature can take place driven by this (initially relatively small) potential difference. A depletion region starts to develop around both anodes, slightly non-circular at first, modifying the uniform distribution of mobile ions assumed at t = 0. When positive ions migrate away from the electrodes, the resistivity of the glass increases very rapidly and the potential at the fiber centre drops, as illustrated in Fig. 5b. Even more charge movement and even greater potential difference between the electrodes and the fiber mid-point follow. This positive feedback results in complete depletion, as in conventional poling. The simulations show that at this stage the ionic distribution has grown from uniform (Fig. 5c) into a pair of depleted rings (Fig. 5d), as measured by etching (Fig. 2b). The simulations also showed that drift is by far more important than the diffusion for the establishment of the charge distribution observed. At long charging times, neutralization (e.g., by H3O+) is expected, but just as in conventional poling a buried layer of negative charge remains. Here, the density of negative charge that remains is twice larger than when an anode and a cathode are used in conventional poling. The simulations shown here are greatly simplified in that only one mobile carrier is considered and the core is assumed not to introduce discontinuities, but the qualitative picture seems to reproduce the main experimental findings.
5. Discussion and conclusion
A larger and more stable second order nonlinear effect is induced when optical fibers are charged instead of poled. An important additional advantage of charging samples with a single pole is a minimized risk for electrical breakdown between electrodes. This opens the doors to increasing the second order nonlinearity by increasing the amplitude of the high voltage applied. Preventing breakdown is particularly relevant in glass systems other than silica, where the third order nonlinear coefficient is large but the breakdown field strength is low. Additionally, the possibility of reducing the separation between contacts without breakdown is relevant in high-speed traveling-wave applications.
The results of our paper show that holding a single pole of a high voltage cable on an isolating floor may not be as harmless as generally assumed. The body of animals is more conductive than air, just as the optical fibers used here become more conductive than air at high temperature. Our results show that air supplies the necessary negative charges to effectively ground one part of the ionic system (in the case here, the outside of the fiber) if another part is connected to a high voltage potential (only one pole). Likewise, the upper part of the bird standing on a high voltage cable collects electrons from the atmosphere, as does our skin when holding the high voltage pole of a Van der Graaff accelerator. In this process, the skin is bombarded with electrons of several kilovolts until shielding stops the influx of new charge. Perhaps the effect of this electron bombardment on the body is negligible even in the long run. However, our results demonstrate through the use of nonlinear optics that charge movement does take place in an ionic system even when one pole is the air.
The authors acknowledge Y. Quiquempois, A. Kudlinski, G. Martinelli and F. Laurell for discussions and support. This work was partly funded by the EU FP5 - IST project GLAMOROUS and carried out within Acreo Fiber Optic Center.
References and links
3. X. C. Long, R. A. Myers, and S. R. J. Brueck, “Measurement of linear electro-optic effect in temperature/electric-field poled optical fibres,” Electron. Lett. 30(25), 2162–2163 (1994). [CrossRef]
4. D. E. Carlson, K. W. Hang, and G. F. Stockdale, “Electrode ‘polarization’ in alkali-containing glasses,” J. Am. Ceram. Soc. 55(7), 337–341 (1972). [CrossRef]
5. W. A. Lanford, K. Davis, P. Lamarche, T. Laursen, R. Groleau, and R. H. Doremus, “Hydration of soda-lime glass,” J. Non-Cryst. Solids 33(2), 249–266 (1979). [CrossRef]
6. T. J. Alley, S. R. J. Brueck, and R. A. Myers, “Space charge dynamics in thermally poled fused silica,” J. Non-Cryst. Solids 242(2-3), 165–176 (1998). [CrossRef]
7. P. G. Kazansky and P. St. J. Russel, “Thermally poled glass: frozen-in electric field or oriented dipoles?” Opt. Commun. 110(5-6), 611–614 (1994). [CrossRef]
9. 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(17), 3623–3625 (2003). [CrossRef]
10. T. G. Alley and S. R. Brueck, “Visualization of the nonlinear optical space-charge region of bulk thermally poled fused-silica glass,” Opt. Lett. 23(15), 1170–1172 (1998). [CrossRef]
11. Y. Quiquempois, A. Kudlinski, and G. Martinelli, “Zero-potential condition in thermally poled silica samples: evidence of a negative electric field outside the depletion layer,” J. Opt. Soc. Am. B 22(3), 598–604 (2005). [CrossRef]
12. W. Xu, D. Wong, and S. Fleming, “Evolution of linear electro-optic coefficients and third-order nonlinearity during prolonged negative thermal poling of silica fibre,” Electron. Lett. 35(11), 922–923 (1999). [CrossRef]
13. P. Blazkiewicz, W. Xu, D. Wong, and S. Fleming, “Mechanism for thermal poling in twin-hole silicate fibers,” J. Opt. Soc. Am. B 19(4), 870–874 (2002). [CrossRef]
17. K. Lee, P. Henry, S. Fleming, and J. L. Blows, “Drawing of Optical Fiber With Internal Co-drawn Wire and Conductive Coating and Electrooptic Modulation Demonstration,” IEEE Photon. Technol. Lett. 18(8), 914–916 (2006). [CrossRef]
18. Y. Quiquempois, A. Kudlinski, G. Martinelli, G. A. Quintero, P. M. Gouvea, I. C. S. Carvalho, and W. Margulis, “Time evolution of the second-order nonlinear distribution of poled Infrasil samples during annealing experiments,” Opt. Express 14(26), 12984–12993 (2006). [CrossRef] [PubMed]