The innovation was to develop a novel, programmable, multi-wavelength, light source for IR gas sensors. The technology is seen an enabler for growing business in the gas sensor applications provided by the industrial beneficiaries. The technology development was divided into three phases: Phase I, Multi-wavelength source in lab validation (TRL 4), Phase II 3500-nm multi-wavelength source in lab validation (TRL 4) and Phase III 3500-nm multi-wavelength source industrial validation (Secondary option 2550nm source industrial validation) (TRL 5). The development phases were modified based on the requirements reported in WP1 in such a way that Phase II represents an intermediate step targeting to wavelengths 2.5 … 2.7 µm requested by the end-users. Phase III represents the final target at 3500nm still including the 2550nm target as a backup.
The workflow was organized into seven workpackages. In the beginning, the specifications for the light source and gas sensor demonstrators were defined by the end-user considering the technological constraints provided by the research partners. Then the basic components, including Mid-IR SLED devices, Si PIC filters and Mid-IR lenses, were designed, processed and tested in respective workpackages. Components were packaged and integrated in order to be used in end-user setups. Finally, the gas sensing system based on the use of the components were tested at end-user laboratories.
We demonstrated a 2.65µm single-mode SLED with 300-nm bandwidth, 3µm multimode SLED with ~550nm bandwidth and 44-nm tuning range with an external cavity tuneable laser using 2.65µm gain element. The sources are applicable for gas sensing. We demonstrated gas sensing with the 2.65 µm SLED using a 0.5 m absorption length. Commercial sensors use typically 4 µm absorption band and ~5 to 15 cm absorption length. The absorption band at 2.65 µm is 100 times weaker than at 4 µm wavelength. Our demonstrator, however, shows that a 10-fold improvement in the signal-to-noise ratio is reachable at 2.65 µm when using sophisticate detection techniques and SLED.
Echelle grating PIC utilizing low-loss Euler bends was demonstrated potentially providing narrow band selective gas sensing. Different waveguide designs were demonstrated proving specificity for different gases.
The spectral bandwidth of a MOEMS filter at 2.65 µm is on the order of 40 nm. With the Echelle grating narrow spectral bandwidth is achievable. However, it was difficult to measure the bandwidth at 2.65 µm wavelength. The measured band-width of ~5-nm was limited by the resolution of the grating spectrometer. At 1550 nm narrow-band Echelle filters were demonstrated with 1-nm bandwidth.
We estimate 35 k€ price for a multi-gas instrument using discrete DFB lasers; 1 k€ is reachable for a sensor based on wide-band SLED. It is not possible to estimate the cost of the integrated SED-PIC light source due to the difficulties met in the packaging. However, when using wafer-level testing and packaging techniques, the target price 300€/unit should be reachable.
Within the frame of the project, we were able to successfully build and test the following gas measurement setups/demonstrators: 1) C1-C5 detection (Primary application), 2) Acetylene/Ethylene/Ethane (Primary application), 3) H2O detection (Secondary application) and 4) CO2 detection (Secondary application). The overall performance of the test analyzers was satisfactory. Fully unattended operation was demonstrated for the CO2 and C1-C5 analyzer.
