m2 to
m2. The
m2 lasers had a threshold current of 
mA, threshold voltage of 
V and slope resistance of 
.
The results obtained in these experiments are not particular to these devices as rectangular 980 nm devices with deposited top mirrors have displayed very similar behaviour [77]. The devices have even current injection across the facet as the spontaneous emission intensity does not vary with position. This result is not fully understood because we observed current-crowding around the edges of other similarly sized devices. One structural difference between the devices displaying current-crowding effects and the devices studied here is that the current confining layers were close to the active region in the former. It is also conceivable that differences between the vertical and lateral conductivities of the two sets of devices plays a role. The homogeneity of the current injection may be important for pattern formation in providing translational symmetry to the laser and, perhaps significantly, this homogeneous injection was also provided in the devices used in Ref. [77]. Because positive detuning is such a significant parameter for pattern formation, it is important that it be determined whether this is possible here. To this end, a sample of the wafer was mounted on a cryogenic probe station and the threshold current measured as a function of heatsink temperature. The results are presented in figure. At zero detuning the laser will have its minimum threshold and this occurs at around 300 K. This measurement is compromised however, because the injection current will heat the active region, altering the detuning for a constant heatsink temperature. However, it has been shown [64] that the active region temperature rise is on the order of 10oC so figure suggests that a region of positive detuning should be available near room temperature with this sample.
