Ultrafast Phenomena and Optical Communications Group

A brief description is given of research activities on Ultrafast Phenomena and Optical Communications currently in progress at Unicamp. These activities cover new glass and polymer materials for photonics, nonlinear optics and femtosecond processes studies on these materials, photonic devices such as Erbium Doped Fiber Amplifiers, and applications in high bit rate optical communications.

The purpose of this article is to introduce our Group on Ultrafast Phenomena and Optical Communications and describe our objectives and current research activities. We hope to facilitate the identification of common interests for international collaborations. We are investigating new concepts and technologies with potential applications in optical communications.

Our group was formed in 1975 to develop optical fibers for the government agency TELEBRAS (Telecomunicações do Brasil). This development was successful and today most of the optical fiber produced in Brazil use our technology. Along the 20 years of our existence, various of our developments and "know how" have been transferred to Brazilian industries, such as Optical Fiber Couplers, characterization procedures, and Erbium Doped Fiber Amplifiers. Our general objective has been to participate of the Brazilian effort for the independent development of optical communications and photonics, providing specialized human resources at all levels from technicians to PhD's, laboratories for special tests and measurements, and monitoring of emerging technologies.

A very simplified description of our activities is given in this paper. Further information can be obtained through our web home page at http://www/ifi.unicamp.br/~gfurco.

Our research program involves four principal activities: Optical communications, New materials, Nonlinear optics, and Femtosecond phenomena. In the New Materials area, we develop glasses doped with quantum dots (CdTe, PbTe) for all optical switches, as well as special glasses (Telurites, Chalcogenides, Niobates) doped with rare earth ions (Er³⁺, Yb³⁺, Pr³⁺) for optical amplifiers. These glasses can accommodate high doping levels of rare earth's and are compatible with fabrication processes of optical fibers or planar waveguide integrated optical amplifiers and lasers.

In the field of Optical Communications field we collaborate with TELEBRAS in studies of high bit rate systems (> 2.5 Gb/s), impact of optical nonlinearities in fiber communications, and components for 1.5 µm systems such as Erbium Doped Fiber Amplifiers (EDFA) and soliton laser sources. In this field we also monitor advances in optical communications technology to recommend strategies for future upgrades of Brazilian optical links.

Presently we are developing novel and efficient methods of characterization of Erbium Doped Fibers (EDF) for optical amplifiers and lasers. An example of an accurate method for measuring the saturation power of EDF's is given in fig. 1. The objectives of these activities are (1) to support design optimization of the EDF's tailored for specific applications (low noise preamplifiers, high output power boosters, etc.) and (2) to provide fully automated EDF characterization procedures for specifications and quality control, as a support for the Brazilian industries that acquired the EDFA technology from TELEBRAS.

Other developments under progress in our group are new laser configurations operating at 1.5 µm. This activities include laboratory explorations of Gain Switched laser diodes as inexpensive devices for high bit rate systems, externally modulated lasers for high quality soliton sources, and high power Cavity Dumped harmonically mode-locked EDF lasers for nonlinear optics experiments.

We also perform fiber loop tests for laboratory simulations of long distance transmissions at high bit rates using EDFA. This activity is directed to evaluate the impact of nonlinear optical effects in different types of fibers and the performance of components such as modulators, filters, isolators, high power transmitters and booster amplifiers in long distance fiber communications, and to explore new transmission concepts such as solitons.

Fig. 1. Fast and nondestructive method for characterization of the intrinsic saturation power of Erbium Doped Fibers (EDF) developed at our group. A flat top pulse is launched into the EDF and the output pulse shape is measured with a digitizing oscilloscope (an actually measured pulse is represented in this figure). The linear (P₁) and saturated (P₂) transmitted outputs are simultaneously displayed in the scope; from these values and the initial time derivative, the intrinsic saturation power is determined by using P_sat = (P₂ - P₁)/[td(lnP₁)/dt - ln(P₂/ P₁)] (t is Erbium excited state lifetime).

____________________________________________________________________________

In the field of Nonlinear Optics our activities are directed to answer the very basic question of what makes a given optical material more nonlinear than others and identify materials for photonic switching applications. We investigate (experimentally and theoretically) the nonlinear optical properties of new materials such as Semiconductor Doped Glasses (SDG) and Organic Polymers in their regions of high transparency.

We measure the nonlinear refractive index (n₂) and the two photon absorption (TPA) coefficient (b) which determine, respectively, the real and imaginary parts of the third order optical susceptibility c⁽³⁾. For all optical switching applications we qualify the materials at a given wavelength (l) according to the two photon figure of merit F = n₂/lb and other qualifiers such as the linear losses, switching power, group velocity dispersion (GVD), response time, optical quality, photo-stability, and so on. A candidate material for a nonlinear directional coupler device (which is of most interest for optical communications) must have |F| > 2.

In SDG's the nonlinearity depends on quantum confinement and we are presently investigating how F depends on the size and transition frequencies of the quantum dots (QD). For a given material, larger crystallites exhibit less quantum confinement but larger n2. In the limit of crystallite radius >> exciton Bohr radius (or 7 nm for CdTe) the SDG behaves like the bulk semiconductor, with a large figure of merit for photon energies below but close to the band gap (the dominant nonlinearity being the optical Stark shift effect) or close to the TPA gap (two photon enhancement of c⁽³⁾). An advantage of SDG over bulk semiconductors is that the GVD is much smaller in the first, thus introducing less pulse broadening and chirp. In the region of c⁽³⁾ dominated by the optical Stark shift, the linear losses and GVD are large and in the region below the two photon gap, n₂ is too small. Smaller crystallites, on the other hand, the strong quantum confinement concentrates the oscillator strength of the QD transitions in narrow spectral regions and, if size dispersion can be controlled, one can get close to the resonances without increasing linear or TPA losses.

In commercially available SDG's the electronic third order susceptibility is often masked by the presence of traps and photo-darkening effects which dominates the nonlinear response and can even change the sign of the measured nonlinear refractive index (see Figure 1). These SDG's show weak quantum confinement. In the case of our SDG of CdTe, the presence of traps can be reduced by special treatments. This material, with strong quantum confinement, has the largest n₂ of all SDG's in the transparency region so far reported.

Another property of CdTe SDG is that it exhibits a lifetime of the QD excited state which is fast (sub-picosecond) and depends on the size of the nanocrystal. For some QD radii the relaxation is extremely fast (360 fs lifetime for 2 nm radius) and complete (no background or slow process after the main decay).7 We believe that this material is very attractive as a fast saturable absorber for laser pulse shaping and photonic switching using saturable absorption applications.

For photonic switching based on saturable absorption with applications in optical communications we have recently developed quantum dots of PbTe, with QD transitions occurring in the 1.5 µm region, where optical fibers present the minimum loss and where the Erbium doped fiber amplifiers operate. This development opens new possibilities for photonic switching.

Fig. 2. The two photon figure of merit (from measurements of n₂ and b at l = 1 µm) for several commercially available SDG color filters before and after partial photo-darkening. Data points are organized as a function of the parameter hn/E_g, where E_g is the optical band gap energy and hn is photon energy of the measurement laser. Labels close to the data points are trade names from Corning (CS) or Schott (RG and GG).

____________________________________________________________________________

Polymers doped with conjugated chromophores exhibit a nonlinear refractive index in the transparency region which is even larger than those of SDG's. In this case n₂ scales with the concentration of the chromophores. By physically doping PMMA (Polymethyl Methacrylate) polymer films with DR1 (Disperse Red 1 dye), for example, we measured (at l = 1.06 µm) n₂ = 3.2´10^-12 cm2/W at the maximum concentration. By chemical doping (functionalized or side chain co-polymerization) of the same chromophore in the same MMA (Methyl Methacrylate) monomer, we measured n₂ = 5.3´10^-12 cm²/W. The larger n₂ is due to a larger concentration of chromophores in the chemically doped polymer, which is not limited, as in the physically doped case, by the solubility of the dye. Our studies indicate that Organics exhibit larger n₂ and larger figure of merit (F) than SDG's but they have less photo-stability and the optical quality is lower.

Still another activity of our group in collaboration with TELEBRAS is the study of electro-optical properties of poled polymer films. Polymers doped with optically nonlinear chromophores are attractive for electro-optical switching devices at high bit rates, as an alternative to Lithium Niobate devices. We develop films and planar optical waveguides of MMA (Methyl Metha Acrylate) copolarized with pendant groups or crosslinked chromophores which are poled in an electric field during the fabrication processes. We study the time decay of the molecular orientation of chromophores, which is generally attributed to thermal and free volume relaxation. We have recently observed, in azo type chromophores, light induced changes in molecular orientation through a processes which is most likely due to photoisomerization. From these measurements we concluded that common laboratory light intensity levels provides a significant contribution to the total orientational relaxation rate.

In the field of Ultrafast Phenomena we use time resolved spectroscopic techniques with 10 fs laser pulses to investigate the relaxation dynamics of quantum dots and bulk semiconductors. From hole burning measurements with femtosecond resolution and excitation spectra (which are QD size selective) we are able to determine the energy levels of our CdTe quantum dots. An advantage of using femtosecond pulses is that we can time resolve the different contributions to the optical nonlinearities of materials. In particular, we can discriminate the fast electronic response from slower thermal or trap contributions. For example, using femtosecond pulses we clearly observe the optical Stark effect in SDG's and thus measure the QD transition dipole moment which determines the nonlinear refractive index near the band gap.

Our work is supported by the Brazilian agencies TELEBRAS, PADCT, RHAE, FAPESP, CAPES, FAEP, FINEP and CNPq.