OK so it isn’t a new radical by any means, but we have discovered silylene in an electric discharge of H3SiOSiH3. Check out the wavelength-resolved emission spectrum! During WRE experiments, we fix the laser wavelength at some absorption feature of the radicals, then scan a monochromator from the laser frequency to lower frequencies to determine what frequencies of light are being emitted after absorption. The difference between the fixed laser frequency and the emitted frequency corresponds to a vibrational frequency in one or more of the radicals (usually only one), allowing us to identify radicals based on their calculated vibrational frequencies.
The plot above shows peaks (vibrational levels) near 1000 1/cm and 2000 1/cm. The calculated frequencies of silylene, SiH2, match up uncannily with these experimental peaks. The peak at zero, by the way, is unabsorbed, scattered laser light. What baffles me is the insane signal to noise ratio of this graph, which is very high compared to many WRE spectra we take. Silylene must reeeeeally love emitting light.