No abstract

Record powers and power densities of 19 W from 100 µm (190 mW/µm) [1] and 18.5 W from a 90 µm aperture (205 mW/µm) [2], were reported in 2007 for devices operating in the 9xx nm range (915 to 980 nm). Powers higher than 20 W from 100 µm [3] and 21 W from 90 µm [4] have already been reported in 2008 by the same teams. These papers witness that higher power and brightness can be only obtained in longer devices. The cavity length is strongly related to another important device parameter, the d/Γ ratio [5,6], where d is the total active region width, Γ is the confinement factor. A low Γ reduces the overlap at the exit mirror between the high power generated in a long device and the active QW, which is vulnerable to degradation. It also reduces the intrinsically high gain of the active QW to the lower values required by increased lengths.In this paper we present results from two series of 940-962 nm diode lasers epi structures in which the d/Γ ratio was scaled up to 1.17 µm. The cavity length L scales directly proportional with d/Γ ratio. The experiment was conducted to find the correlation between the maximum power and the values of d/Γ and L. The main parameters used in the design are given in Table 1.

A new technology is proposed for Backside Contact (BSC) Silicon solar cells. This technology is self-aligned sincc only o n e photolithography step, without alignment, is rcquired instead of 4 to 6 in previous tcchnologies. Compared to a recently reported selfaligned process for BSC cells, this one eliminates the problcm of compensated regions, and presents the advantages of a reduction of the emitter area, a passivation of silicon surface with thermal oxide bctwccn metal fingers and a reduction of metal-silicon contact area. As a consequence, this new self-aligned technology offers great potentialities f o r t h e fabrication of high efficiency solar cells at a low cost. INTRODUCIIONThe Backside Contact (BSC) design has yet d c m o n s t r a t e d v e r y h i g h e f f i c i e n c i e s w i t h monocrystalline silicon m a t e r i a l , u n d e r highly conccntrated or unconcentrated light [ 1,2,3]. Several PV spccialists acknowledge that the BSC cell is probably thc bcst solar cell design to date for concentration application. Some of which are even thinking to apply thc BSC structure to GaAs material [4,5,6,7]. In this case, it is however not obvious that the BSC design will bring significant advantages, since this design has been proposcd in order to solve particular problems related to silicon material. The main advantages of the silicon BSC cell are due to the almost total uncoupling between thc optical optimization of the cell (essentially the front side) and the electrical optimization of the emitter rcgions on the back side of the cell. This fact implies that the series resistance of the cell is not limited by the emitter sheet resistance, nor the metal coverage fraction, that the collection efficiency is not degraded by a low lifetime doped front region (namely the front cmittcr in a conventional solar cell). It also implies that thc emitters can be optimized for a low saturation currcnt without interference with t h e collection cfficcncy. Finally, it has been demonstrated [3] that, with a well designed multilevel metallization scheme which is possible on the BSC cell, the specific series rcsistance (product of the series resistance and the cell arca) may be independent of the size of the cell.Both Interdigitated Back Contact (IBC) and PointContact (PC) structures are particular cases of the BSC dcsign and showed good results [1,2,3]. For the moment, the PC cell owns the efficiency world record, 28.5% at 15 W/cm2 AM1.5 at 25OC, for silicon material. Its advantage is due to an extremely reduced coverage fraction of the emitter area (less than 10 %), combined with an uniform distribution of the emitters in X and Y directions, in order to reduce the emitter component of the cell saturation current.The disadvantage of BSC cells, that has been pointed out, is the requirement for a sophisticated processing sequence, making use of, at least, 4 photolithography steps. This number could even be increased to 6 for double level metallization when it is necessary [3]. For this reason, some workers prefer to develop...

No abstract

No abstract