Fiber-optic sensing markets are rapidly expanding. Over the last few years, extensive activity has concentrated on development and commercialization of seismic and acoustic sensing arrays for defense (intrusion detection, perimeter protection, submarine identification, hydrophones), land- and underwater-based industrial applications (oil and gas exploration, pipeline surveillance, geophysical monitoring), and civil-infrastructure monitoring (roads, bridges, water-supply systems, and power lines). Fiber-optic arrays typically consist of long (up to 10km) strings of individual sensors that can detect land vibrations or underwater acoustic signals. In typical applications, the number of sensing nodes per string can range from a few sensors for in-well down-hole acoustic devices (or several hundred for down-hole seismic sensors) to thousands of elements for ocean-bottom seismic cables and subsea streamers.1 Since each fiber sensor needs to be individually interrogated and spectrally distinguished, a large number of narrow-linewidth laser sources are needed to excite and interact with the sensor array.
New laser sources are required to improve performance and enable realization of lower-cost sensing systems. External-cavity-laser (ECL) technology is emerging as a high-performance alternative to semiconductor distributed-feedback (DFB) and fiber-laser-based solutions. In addition to improved stability to temperature variations or external vibration, these light sources have desirable characteristics including narrow linewidths and low phase noise. They currently cover the C band (1528–1568nm) but can easily be extended into the L band (1569–1610nm). They operate over a wide temperature range and comply with the stringent long-term optical-power and wavelength-stability demands of optical-sensing applications.
Figure 1. Planar-lightwave-circuit (PLC) fabrication and external-cavity-laser (ECL) integration: (a) Silicon wafer with etched PLCs, (b) diced PLCs, (c) integrated ECL, and (d) ECL packaged in 14-pin butterfly package.
ECLs were first developed in the 1990s.2 They consist of two key building blocks: a gain section—defined by a semiconductor Fabry-Perot (FP) gain chip—and a wavelength-selective feedback filter provided by either a fiber or a planar Bragg grating that forms the laser cavity and ensures single longitudinal-mode operation. The latter is also referred to as a planar-lightwave circuit (PLC). The ECL laser cavity we developed consists of an indium-phosphide gain chip coupled to a silica waveguide PLC.3 Individual PLCs with reflective Bragg gratings are fabricated on large-diameter silicon wafers using semiconductor processes (see Figure 1). The design is simple and manufacturable yet robust. It was Telcordia qualified and functions over a wide case-temperature range (from −5 to +75°C).
Fiber lasers4 used in fiber-optic sensing applications are more complex and have more components than our newly developed devices, leading to higher costs and an increased number of possible failure modes. Fiber-optic modules are typically much larger than our planar ECL implementations. Despite advances in vibration-isolation techniques, fiber lasers are also more susceptible to performance degradation caused by vibrations and operate over a more limited temperature range.
Our new planar devices offer a narrow spectral linewidth, superior relative-intensity noise compared to fiber lasers, and low phase noise. They exhibit a stable wavelength response to temperature variations of ~15pm/°C (compared to ~100pm/°C for typical semiconductor DFB lasers). The overall wavelength variation over the case-temperature range is typically less than ~20pm. Typical linewidths are <5kHz full width at half maximum (FWHM), adopting a Lorentzian model. Fiber-laser linewidths are typically in the 2–15kHz FWHM range. The measured relative-intensity noise of our new ECL is <−140dB/Hz at frequencies greater than 2kHz, with a high (>8GHz) relaxation-oscillation frequency. Fiber lasers, on the other hand, usually show a peak at several megahertz, which is a limitation for some applications.
Figure 2 compares the phase-noise performance of our planar ECLs with conventional DFB and narrow-linewidth fiber lasers. The phase noise closely approaches the performance of typical (3kHz-linewidth) fiber lasers and far exceeds that of DFB lasers. Thus, this technology offers a lower-cost and more robust alternative to conventional lasers for a wide range of distributed-sensing applications.
Figure 2. Phase-noise comparison of various laser types generally used in sensing applications. DFB: Distributed feedback, RIO: Redfern Integrated Optics.
For the first time, narrow-linewidth and low-phase/frequency-noise planar-waveguide ECLs have been developed that satisfy most interferometric-sensing performance requirements. They are manufactured in a small package, providing high stability and reliability at a lower cost than alternative technologies. Planar ECLs are suitable for most interferometric applications in demanding environmental conditions. Future work will focus on accessing additional segments of the sensing markets by producing higher output powers and further phase-noise reduction.
Redfern Integrated Optics Inc.
Santa Clara, CA
Michael Jansen is responsible for engineering and manufacturing at Redfern Integrated Optics. He spent over 24 years in the photonics industry where he has held technical (vice-president of engineering and chief technology officer) and functional (general manager and chief operational officer) executive positions.
3. L. Stolpner, S. Lee, S. Li, A. Mehnert, P. Mols, S. Siala, J. Bush, Low noise planar external cavity laser for interferometric fiber optic sensors, Proc. SPIE 7004, pp. 700457, 2008.doi:10.1117/12.786226