Search

A semi-empirical model is compared with measurements to establish limiting factors in the performance of p-modulation doped InAs quantum dot (QD) lasers. Fitted absorption spectra allow identification of supposed factors and comparison of multiple samples isolates their origin, providing insights for future laser design.

A systematic analysis of the performance of VCSELs, fabricated with a decreasing number of structural elements, is used to assess the complexity of fabrication (and therefore time) required to obtain sufficient information on epitaxial wafer suitability. Initially, sub-mA threshold current VCSEL devices are produced on AlGaAs-based material, designed for 940 nm emission, using processing methods widely employed in industry. From there, stripped-back Quick Fabrication (QF) devices, based on a bridge-mesa design, are fabricated and this negates the need for benzocyclcobutane (BCB) planarisation. Devices are produced with three variations on the QF design, to characterise the impact on laser performance from removing time-consuming process steps, including wet thermal oxidation and mechanical lapping used to reduce substrate thickness. An increase in threshold current of 1.5 mA for oxidised QF devices, relative to the standard VCSELs, and a further increase of 1.9 mA for unoxidised QF devices are observed, which is a result of leakage current. The tuning of the emission wavelength with current increases by ~0.1 nm/mA for a VCSEL with a 16 μm diameter mesa when the substrate is unlapped, which is ascribed to the increased thermal resistance. Generally, relative to the standard VCSELs, the QF methods employed do not significantly impact the threshold lasing wavelength and the differences in mean wavelengths of the device types that are observed are attributed to variation in cavity resonance with spatial position across the wafer, as determined by photovoltage spectroscopy measurements.

High-volume low-cost production of vertical cavity surface emitting lasers (VCSELs) will allow their exploitation in new commodity markets. We report the successful scaling up from research level fabrication to produce oxide confined VCSELs across a whole 150mm wafer. On-wafer light-current-voltage (L-I-V) and spectral measurements are analyzed to determine the cross-wafer variations in threshold current, threshold current densities and emission wavelength, which is compared with reflectivity measurements taken immediately after growth. We examine the dependence of VCSEL performance on fabrication parameters over a range of device dimensions to assess whether variation arises from non-uniformity of the epitaxial material or wafer processing.

Coupled-cavity lasers have attracted wide attention in the past, in particular for telecommunication applications where their wavelength tunability and ability for side mode suppression are desirable. The inherent sensitivity of these devices to changes in the optical coupling has also led to their proposed use in optical sensing systems. Small changes to the refractive index of the coupler section can lead to shifts in the resonance frequency of the laser. Here we present an alternative approach to coupled-cavity sensing that exploits changes to the imaginary part of the refractive index of the coupler. An optical loss, introduced to the cavity by the passage of micro-particles, influences the optical loss of the lasing mode and changes the threshold gain requirement of the laser. The sub-linear nature of the gain-current density characteristics of the quantum confined gain medium amplifies this effect, producing an even larger perturbation in output power. We demonstrate this sensing mechanism using a monolithic coupled-cavity particle detector with on-chip capillary fill microfluidics and an in-line photo-detector section for photo-voltage transduction. Both laser and detector are pulsed allowing for a time-resolved measurement to be taken.

The potential of the device is demonstrated through the results of a micro-bead counting experiment. A 0.5 µl sample volume containing 10 µm polystyrene micro-beads in Dl water is deposited into an on-chip inlet reservoir from where it flows, under capillary action, through a buried flow cell. A laser on one side of the cell is forward biased and the opposing laser is operated as a photodiode (Fig. 2(a)). Both are pulsed to provide sub-µs time-resolution of bead transit ‘events’ with a 30 mV noise floor that affords a large dynamic range of over 1.2 V (Fig. 2(b)). After passing through the flow cell the sample fluid is drawn into a spiral patterned exit reservoir that provides sufficient pull through to sustain a continuous flow for over 30 s with flow rate of > 4mm/s.