Characterization of phosphorus deactivation in epitaxial Si:P and surpassing 1021 cm−3 carrier concentration

Abstract

Deactivation behavior of epitaxial phosphorus-doped silicon, Si:P, was analyzed using a combination of Hall effect measurement and high-resolution X-ray diffraction. Processing at two different temperatures provided the ability to determine when the deactivation mechanism transitions from thermally-driven to primarily dopant-density driven. For < 500 °C growth temperature, a minimum resistivity of 0.258 ± 0.003 mΩ-cm was observed at (2.88 ± 0.02)% total P with active carrier concentration, n, of (7.18 ± 0.09) × 10²⁰ cm⁻³ equating to (50.1 ± 1.9)% active dopant. In contrast, maximum n = (7.76 ± 0.09) × 10²⁰ cm⁻³ was seen at (4.32 ± 0.02)% P. For < 400 °C growth temperature, a minimum resistivity of 0.214 ± 0.002 mΩ-cm was measured at (2.97 ± 0.02)% P with n = (1.01 ± 0.01) × 10²¹ cm⁻³ which is (67.9 ± 1.7)% active dopant. Maximum n = (1.09 ± 0.02) × 10²¹ cm⁻³ was at (4.11 ± 0.02)% P, again different than resistivity minimum. Lower temperature processing pushed the active dopant amount for maximum n from (35.7 ± 1.6)% to (53.0 ± 1.9)%, at nearly identical %P, meaning deactivation is indeed inhibited. This benefit appears to dissipate at ∼ 5.5 % P, based on converging resistivity and active dopant dependences, indicating the onset of dopant-density driven deactivation. Growth at < 400 °C also provided near-unity (96.0 ± 2.2)% active dopant [where n = (7.51 ± 0.07) × 10²⁰ cm⁻³] at (1.57 ± 0.02)% P giving 0.247 ± 0.003 mΩ-cm resistivity. Both ionized and non-ionized (inactive) dopants were shown to influence mobility, with the latter coming on at ∼ 50 % inactive dopant and causing a maximum decrease of ∼ 7 cm²/V-s with the %P range considered here. Threshold and maximum were both independent of growth temperature. These results have implications on Si:P optimization for contact versus for bulk resistivity as well as lowering both resistivities overall.