Extended Data Figure 2: Threshold and spin-relaxation measurements.
a, Maser threshold. The peak maser output power increases linearly as a function of the optical pump power (data). Extrapolation of the linear fit (dashed line) to zero maser output power reveals a threshold optical pump power of 138 mW, which is lower than the predicted 180 mW. b, Spin–lattice relaxation time T1 as a function of laser pump power (data). The slight decrease in T1 is expected and due to an increase in temperature caused by the non-radiative (heating) processes during the NV spin-polarizing optical pump cycle. c, Spin-decoherence time T2 as a function of laser pump power (data). There is little change in T2, with a slight jump upon applying optical pumping, probably due to an increase in EPR signal amplitude and hence less error. There is subsequently a slight decrease due to temperature increase and pumping decoherence. Photo-conversion of NV− to NV0 could also be a source of decoherence. d, Power saturation broadening. The inhomogeneous spin decoherence time 
was inferred from power saturation broadening measurements of the spin resonance lines. The spectral full-width at half-maximum (FWHM) γ was measured as a function of interrogating microwave power (data). A spin decoherence rate of 
was extracted by extrapolating the square of the FWHM linewidth down to zero microwave power (dashed line). The vertical dashed line in b and c depicts the applied optical pump power of 400 mW.
