Influence of boron clustering on the emitter quality of implanted silicon solar cells: an atom probe tomography study

Abstract

The use of ion implantation doping instead of the standard gaseous diffusion is a promising way to simplify the fabrication process of silicon solar cells. However, difficulties to form high-quality boron (B) implanted emitters are encountered when implantation doses suitable for the emitter formation are used. This is due to a more or less complete activation of Boron after thermal annealing. To have a better insight into the actual state of the B distributions, we analyze three different B emitters prepared on textured Si wafers: (1) a BCl 3 diffused emitter and two B implanted emitters (fixed dose) annealed at (2) 950°C and at (3) 1050°C (less than an hour). Our investigations are in particular based on atom probe tomography, a technique able to explore 3D atomic distribution inside a material at nanometer scale. Atom probe tomography is employed here to characterize B atomic distribution inside textured Si solar cell emitters and to quantify clustering of B atoms. Here, we show that implanted emitters annealed at 950°C present maximum clusters due to poor solubility at lower temperature and also highest emitter saturation current density (J 0e = 1000 fA/cm 2 ). Increasing the annealing temperature results in greatly improved J 0e (131 fA/cm 2 ) due to higher solubility and a consequently lower number of clusters. BCl 3 diffused emitters do not contain any B clusters and presented the best emitter quality. From our results, we conclude that clustering of B atoms is the main reason behind higher J 0e in the implanted boron emitters and hence degraded emitter quality.

Bifacial, fully screen-printed n-PERT solar cells with BF2 and B implanted emitters

“…In case of BF 2 implantation and subsequent annealing at 950°C a reduced electrical activation of boron might be accompanied by the formation of recombination active boron-Si interstitial clusters [12]. Although one might conclude that there is a correlation between the inactive boron dose and the increase in J 01 , we so far cannot exclude other recombination mechanisms, e.g.…”

“…Since solid-phase epitaxy supports the annealing of the amorphized surface, the formation of dislocation loops is much less pronounced. However, as another explanation, inactive boron paired with Si interstitials (boron interstitial clusters, BICs) could still mediate recombination after annealing at temperatures of 950°C [12]. Although the studies on BF 2 implants for the emitter preparation show promising J 0e and lifetime values [11], no results on a full solar cell level have been published so far.…”

Plasma‐immersion ion implantation: A path to lower the annealing temperature of implanted boron emitters and simplify PERT solar cell processing

“…The space between both curves corresponds to the inactive boron dose. As previously discussed, this inactive boron dose seems to be explained by the presence of BICs in the emitter . It was reported that the BICs are rapidly formed at the beginning of the annealing and once grown become immobile and hard to dissolve during further anneals .…”

“…However, the main drawback of B + implantations by the standard BLII technique remains the high annealing temperature (≥1050°C) necessary to remove the induced crystal defects as dislocation loops and boron interstitial clusters (BICs) and thus achieve low J 0e values . This high‐temperature treatment can affect the bulk charge carrier lifetime of Czochralski (Cz) wafers by the formation of oxygen‐related defects and by increasing the risk of parasitic contamination (eg, from the furnace).…”

“…11 In addition, very high-emitter quality has been reported using a beamline ion implantation (BLII) of B + ions, with emitter saturation current densities (J 0e ) lower than 40 fA/cm 2 , 12 along with promising PV conversion efficiencies up to 21.8% with screen-printed n-PERT cells. 7 However, the main drawback of B + implantations by the standard BLII technique remains the high annealing temperature (≥1050°C) necessary to remove the induced crystal defects as dislocation loops and boron interstitial clusters (BICs) 13,14 and thus achieve low J 0e values. 15 This high-temperature treatment can affect the bulk charge carrier lifetime of Czochralski (Cz) wafers by the formation of oxygen-related defects and by increasing the risk of parasitic contamination (eg, from the furnace).…”