Abstract
We report optical signal amplification in a solid-state dye-doped polymer with a rib waveguide structure. A 625 nm pulsed signal and a collinear 575 nm pump are coupled into a 1 μm × 120 μm poly(methyl methacrylate) waveguide doped with 1% by weight Rhodamine 640 dye. Depending on the signal intensity, a maximum optical gain in the 21-26 dB range is obtained from a 1.2-cm-long device, accompanied by a signal-to-noise ratio in the 9-16 dB range.
| Original language | English |
|---|---|
| Article number | 3 |
| Pages (from-to) | 5137-5139 |
| Number of pages | 3 |
| Journal | Applied Physics Letters |
| Volume | 85 |
| Issue number | 22 |
| DOIs | |
| Publication status | Published - 29 Nov 2004 |
Funding
Reilly M. A. a),b) Marinelli C. Morgan C. N. Penty R. V. White I. H. Ultrafast Photonics Collaboration , Centre for Photonic Systems, Engineering Department, University of Cambridge, Trumpington Street, Cambridge CB2 1PZ, United Kingdom Ramon M. Ariu M. Xia R. Bradley D. D. C. a),c) Ultrafast Photonics Collaboration , Blackett Laboratory, Imperial College of Science, Technology and Medicine, Prince Consort Road, London SW7 2BZ, United Kingdom a) Authors to whom correspondence should be addressed. b) Electronic mail: [email protected] c) Electronic mail: [email protected] 29 11 2004 85 22 5137 5139 21 06 2004 03 10 2004 2004-12-01T15:13:57 2004 American Institute of Physics 0003-6951/2004/85(22)/5137/3/ $22.00 We report optical signal amplification in a solid-state dye-doped polymer with a rib waveguide structure. A 625 nm pulsed signal and a collinear 575 nm pump are coupled into a 1 μ m × 120 μ m poly(methyl methacrylate) waveguide doped with 1% by weight Rhodamine 640 dye. Depending on the signal intensity, a maximum optical gain in the 21 – 26 dB range is obtained from a 1.2 - cm -long device, accompanied by a signal-to-noise ratio in the 9 – 16 dB range. Polymers have many potential advantages over inorganic materials for fabricating active and passive optical components because they are cheap, flexible, strong, and can be patterned using ultraviolet writing and hot embossing techniques, or spin-coated onto pre-patterned substrates. 1 Many polymers can be transparent across the low-loss windows of typical plastic optical fiber (POF) at 570 and 640 nm , making them attractive for use in waveguide devices that can be readily deployed with POF. In this context, polymer based optical amplifiers could play a key role in achieving signal amplification in plastic fiber networks. In addition, polymer based gain blocks could be integrated into passive components to compensate for coupling, scattering, and absorption losses. In recent years, optical signal amplification in the visible wavelength range has been demonstrated using conjugated polymers in solution, 2 and amplified spontaneous emission (ASE) has been observed in a range of thin conjugated polymer films. 3 At the same time, doped polymers, where optically active dopants are suspended in polymer to create an active gain region, 4 have also attracted much interest, as they allow the gain spectrum to be tuned across a wide range of visible wavelengths and potentially into the near-IR region. At wavelengths above 1 μ m , where conventional silica-based fiber optic systems operate, doping of polymer waveguides with rare earth metals has been used to generate novel optical gain media, 5 and gains of 13 dB have recently been achieved. 6 In the visible range, optical amplification has been achieved using dye-doped polymer optical fibers, 7,8 and ASE excitation has been observed in thin dye-doped polymer films. 9 In this letter we report an optically pumped, dye-doped polymer amplifier with a rib waveguide structure. The dependence of the amplifier gain on pump and signal energies, as well as the signal-to-noise ratio performance are investigated by facet-coupling collinear pump and signal beams into the waveguide. The active waveguide material used in this study is PMMA doped with Rhodamine 640 dye, which is dissolved in the polymer at 1% by weight. In order to evaluate the material characteristics and choose optimum operating conditions, the absorption, photoluminescence and ASE spectra are initially measured for a 450 - nm -thick sample. These measurements, shown in Fig. 1 , suggest an optimum pump wavelength of 575 nm , corresponding to the absorption peak, and a signal wavelength of 625 nm to minimize absorption in unpumped portions of the medium and to maximize the gain. The polymer waveguide structure illustrated in Fig. 2 (inset) is obtained by spin-coating the dye-doped PMMA onto a silica substrate that has previously been patterned with a 120 - μ m -wide and 10 - μ m -deep rib waveguide structure using reactive ion etching. The optical properties of this device are investigated with a Q -switched Nd:YAG laser, used to pump a type-II BBO optical parametric oscillator delivering 10 ns pulses at 575 nm with a repetition rate of 10 Hz . The resultant light beam is split in two: one half used to pump a solution of Rhodamine 640 dye suspended in methanol ( 500 μ g ∕ ml ) to generate the 625 nm signal beam; the other half to act as the 575 nm pump beam for the waveguide amplifier under test. Neutral density filters are used to independently vary the power of the signal and pump beams, which are then recombined and coupled into the cleaved facet of the silica waveguide using a × 40 microscope objective lens. The light in the silica waveguide then couples into the higher refractive index polymer waveguide above. The near field of the polymer waveguide output is imaged using a × 16 lens and passed into a CCD camera for spectral intensity analysis (a pinhole is placed in front of the camera to eliminate the contribution of any scattered light). The variation of optical gain with signal energy, E signal , is shown in Fig. 2 . Pump energy is maintained at 14 μ J for this measurement after preliminary characterization showed this to be a suitable energy for signal amplification. Gain is calculated from the intensity ratio of the amplified to the unamplified signal at 625 nm , after subtraction of the ASE contribution. It should be stressed that the signal energies presented in Fig. 2 are the values measured at the 625 nm source and not the actual signal energies coupled into the amplifier waveguide; no adjustment has been made for coupling losses. The plot in Fig. 2 is therefore shifted toward higher signal energies. However, this does not affect the overall trend, or the theoretical fit, which clearly indicate a decrease in optical gain as the signal beam energy rises high enough to begin depleting the excited state. 4 The curve in Fig. 2 is a theoretical fit to the data obtained using Eq. (1) 10 for the variation in gain, G , with E signal : G = ( E s ∕ E signal ) ln [ 1 + G 0 ( e ( E signal ∕ E s ) − 1 ) ] , (1) where G 0 is the gain of the device when the amplitude of the signal beam is small, known as the small-signal gain, and E s is the energy of the signal beam in the waveguide required to saturate the amplifier. From the fit we find that G 0 = 19 dB and E s = 0.83 μ J , though it should be noted that this E s value does not take account of waveguide coupling losses, estimated to be ∼ 10 dB . Next we investigate the dependence of the optical gain on pump energy. The trend observed in Fig. 2 prompts the choice of an optimum E signal value of 4 nJ and a higher value of 40 nJ for comparison. Figure 3 (top graph) presents the measured gain variation with increasing pump energy. An improvement in optical gain is observed up to E pump ∼ 60 μ J , reaching a maximum gain of approximately 26 dB , followed by gain saturation. Once again, the pump energies in Fig. 3 (top graph) represent the values at the 575 nm source, prior to coupling into the waveguide, and are not corrected for coupling losses. To complete the set of amplifier characterizations, the device signal-to-noise ratio (SNR) is assessed at the same two values of signal energy used in Fig. 3 (top graph), and in the pump energy range where the gain is > 3 dB . We calculate the SNR by integrating the emission spectra of the pumped amplifier over a 6 nm window, centered at 625 nm , and taking the ratio of the full peak area to the underlying ASE area. Figure 3 (bottom graph) indicates a decrease in SNR with increasing pump energy and optical gain. A SNR between 7 and 15 dB is observed for E signal = 4 nJ . Increasing E signal to 40 nJ results in a 6 – 7 dB improvement in SNR over the entire pump energy range, although this is accompanied by a decrease in gain. For the investigated E signal range, a 60 μ J pump energy produces an optical gain between 21 and 26 dB , and a corresponding SNR between 9 and 16 dB . The amplifier emission is monitored during all of the measurements to ensure that a broadening of the ASE spectrum, indicative of photo-bleaching, does not occur at high pump energies. Therefore, in the absence of photo-bleaching, we may estimate an upper limit of ⩽ 10 μ m for the absorption length of the pump in the waveguide, using a measured loss coefficient of α = 1320 cm − 1 at 575 nm . This suggests that most of the optical gain occurs in a small region at the beginning of the 1.2 cm waveguide, before the pump beam has been significantly absorbed, and therefore any reduction of signal beam absorption that might occur in this region when the pump beam is present has a negligible ( < 0.1 dB ) contribution to the gain estimate. Furthermore, these results suggest that a yet more compact device is possible. A larger region of the sample could be pumped, and therefore the gain could be enhanced, by use of transverse pumping, but a collinear arrangement is attractive for optical communication applications where compact device layout is a priority. Tests carried out on the material in liquid form have also suggested that gain performance might be further improved by fine tuning the concentration of dye suspended in the polymer. In addition, the dye used in this experiment allows a tunable wavelength window of 40 – 50 nm , though this range may be more limited in polymer composite form. By incorporating different dyes in the PMMA matrix the active range could be further extended. In conclusion, we have demonstrated a solid-state, dye-doped rib waveguide polymer amplifier, formed by a 1 μ m × 120 μ m PMMA stripe doped with Rhodamine 640. Depending on the signal intensity, a maximum gain in the 21 – 26 dB range at 625 nm has been achieved in a 1.2 - cm -long device operating with collinear optical pump and signal beams. The device exhibits a promising signal-to-noise ratio in the 9 – 16 dB range and has the potential for tuning in a 40 – 50 nm wavelength window with the same dye, and throughout the visible spectrum using other dyes. Even better gain performance and a more compact device geometry can be envisaged using a shorter waveguide and optimum dye concentration, leading to a compact, easy to fabricate, high gain block for applications in the 640 nm wavelength region. The authors are grateful to the UK Engineering & Physical Sciences Research Council (EPSRC) for financial support (Ultrafast Photonics IRC Grant No. GR/R 55078). FIG. 1. Absorption, photoluminescence, and ASE spectra for a 450 - nm -thick sample of PMMA doped with Rhodamine 640 dye. FIG. 2. Variation in optical gain with signal energy for a dye-doped polymer waveguide. The pump energy was kept constant at 14 μ J . Inset: Cross section of the polymer waveguide spin-coated on a patterned silica substrate. The silica substrate is formed of two layers; the top (waveguiding) layer is produced using flame hydrolysis deposition with Ge doping to achieve a larger refractive index than the lower substrate layer (produced by wet oxide growth without doping), which ensures that light is confined to the waveguiding layer. FIG. 3. Dependence of the optical gain and signal-to-noise ratio (SNR) on pump energy for two values of signal energy.
ASJC Scopus subject areas
- Physics and Astronomy (miscellaneous)