Modulation Doping of Silicon using Aluminium-induced Acceptor States in Silicon Dioxide

Abstract

All electronic, optoelectronic or photovoltaic applications of silicon depend on controlling majority charge carriers via doping with impurity atoms. Nanoscale silicon is omnipresent in fundamental research (quantum dots, nanowires) but also approached in future technology nodes of the microelectronics industry. In general, silicon nanovolumes, irrespective of their intended purpose, suffer from effects that impede conventional doping due to fundamental physical principles such as out-diffusion, statistics of small numbers, quantum- or dielectric confinement. In analogy to the concept of modulation doping, originally invented for III-V semiconductors, we demonstrate a heterostructure modulation doping method for silicon. Our approach utilizes a specific acceptor state of aluminium atoms in silicon dioxide to generate holes as majority carriers in adjacent silicon. By relocating the dopants from silicon to silicon dioxide, Si nanoscale doping problems are circumvented. In addition, the concept of aluminium-induced acceptor states for passivating hole selective tunnelling contacts as required for high-efficiency photovoltaics is presented and corroborated by first carrier lifetime and tunnelling current measurements.

Room-temperature single dopant atom quantum dot transistors in silicon, formed by field-emission scanning probe lithography

“…Given the wide band gap in SiO2 (~9.7 eV), depending on whether a P impurity atom is bonded to a Si or O atom, or lies interstitially (possible, given segregation in a heavily doped material), various energy states are possible [51,52]. If the P atom bonds to a Si atom, it may act as a large ionization energy (~0.7 eV) electron donor into the SiO2 conduction band, ~4 eV above the conduction band/Fermi energy EF in the Si regions.…”

Intrinsic ultrasmall nanoscale silicon turns n-/p-type with SiO2/Si3N4-coating

“…Recently, Si modulation doping of adjacent dielectric layers based on nitrides [11] and oxides [12], in analogy to modulation doping of III–V semiconductors, were shown to be an alternative to conventional impurity doping.…”

Deactivation of silicon surface states by Al-induced acceptor states from Al–O monolayers in SiO2

“…Though there is little effect of H 2 on Q fix for up to ∼2.5 nm tunnel-SiO 2 , the reduction is up to ∼30% for thicker oxides (equivalent to a negative Q fix loss of up to 7.5 × 10 11 cm −2 ). Taking the atomic structure of the Al-induced acceptor state in SiO 2 into account (i.e., a fully O-coordinated, trivalent Al-atom replacing a Si-atom in the Si-O tetrahedrons of SiO 2 , which creates an O-DB), 11 it seems possible that the H 2 -passivation can deactivate some of the acceptor states from which the negative Q fix originates. Apparently, a typically used Si/SiO 2 passivation process (400°C, 1 h, 100% H 2 ) is (luckily) far less efficient for passivating the acceptor states than for P b -type DB-defects at the Si/SiO 2 interface.…”

“…We note that the density of Al-acceptors in SiO 2 exceeds the interface trap density considerably. 11 Half of the 6 ML Al-O thickness was added to both the tunnel-SiO 2 and the capping-SiO 2 . Figure 3 shows the E Al values as a function of tunnel-SiO 2 thickness for the unpassivated and the H 2 -passivated W-sample.…”

“…3 Examples for the latter electrons from the Si substrate and that act as hopping sites for holes. 11,12 In this contribution, the correlation between negative Q fix and minority carrier lifetimes (τ eff ) as a function of the tunnel-SiO 2 thickness and the role of H 2 -passivation are investigated. We show that the interface trap density at the Si/SiO 2 interface depends decisively on the thickness of the tunnel-SiO 2 between the wafer and the ALD Al-O MLs.…”