Search

Electron spin resonance (ESR) study on 2H-polytype synthetic MoS2 revealed the N acceptor dopants as being characterized by a spectrum of axial symmetry [g∥ = 2.032(2); g⊥ = 2.270(2)], typical for a hole-type center in MoS2. The N impurities substitute for S sites, with a density of ∼2.3 × 1017 cm−3, which accounts for the overall p-type doping. With respect to measurements for the applied magnetic field directed along the c-axis, the signal consists of a 14N primary hyperfine triplet of splitting constant A∥ = 14.7 ± 0.2 G superimposed on a correlated Gaussian single central line of peak-to-peak width ΔBpp = 15.3 ± 0.5 G, the latter making up only ∼26% of the total signal intensity. The current work extends on these results through extensive monitoring of the temperature (T) dependence of salient ESR parameters and studying the impact of thermal treatment. ESR signal saturation studies indicate a N acceptor spin–lattice relaxation time T1 (4.2 K) ≈ 3 × 10−4 s, notably different from the much smaller As acceptor’s T1 in geological MoS2. Concerning the thermal stability of the dopant, the N acceptor is found to be drastically passivated when exposed to H2 at ∼500 °C. Yet, subsequent reactivation attempts in vacuum at temperatures up to 740 °C appear unsuccessful, urging great caution with conventional forming gas treatments at T ≳ 500 °C. Combination of careful K- and Q-band ESR monitoring of the T-dependent signal intensity resulted in the consolidation of the N dopant as a shallow acceptor of activation energy Ea = 45 ± 7 meV. The consolidated results establish N as a promising candidate for stable covalently bonded p-type doping of MoS2 layers intended for application in novel nanoelectronic devices. [ABSTRACT FROM AUTHOR]

In extending on recent electron spin resonance (ESR) work which has revealed the N acceptor (N substituting for S site) in 2H-polytype bulk synthetic MoS2, the dopant is extensively analyzed in terms of its frequency, temperature ( T), and magnetic field B angular dependent ESR spectral characteristics. For B ∥c-axis, the multi-frequency analysis confirms the ESR spectrum as being composed of a 14N hyperfine (hf) triplet with hf splitting constant A∥ = 14.7 ± 0.2 G ( B ∥c-axis) and making up ≈74% of the total spectrum intensity, superimposed on a central line centered at about equal g-value [ g∥ = 2.032(2)]. The presence of the latter signal, points to some non-uniformity in dopant distribution, that is, clustering, with about ≈26% of the total N response not originating from N incorporated in the preferred 'isolated' dopant configuration. Angular dependent measurements reveal distinct anisotropy of the hf matrix, whereas ESR probing over a wide T-range exposes drastic signal broadening with increasing T above ≈150 K. Detailed study of the N acceptor signal intensity versus T at Q-band reveals an activation energy Ea = 50 ± 10 meV, herewith consolidating the value reported initially. Besides unveiling the S-site substitutional N impurity as an appropriate p-type dopant for MoS2, the total of the ESR work establishes a basic frame of the N acceptor ESR characteristics, giving way for further in-depth theoretical perusal. [ABSTRACT FROM AUTHOR]

Al 2 O 3 on Si is known to form an ultrathin interfacial SiO 2 during deposition and subsequent annealing, which creates a negative fixed charge ( Q fix ) that enables field-effect passivation and low surface recombination velocities in Si solar cells. Various concepts were suggested to explain the origin of this negative Q fix . In this study, we investigate Al-O monolayers (MLs) from atomic layer deposition (ALD) sandwiched between deliberately grown/deposited SiO 2 films. We show that the Al atoms have an ultralow diffusion coefficient (∼4 × 10 -18 cm 2 /s at 1000 °C), are deposited at a constant rate of ∼5 × 10 14 Al atoms/(cm 2 cycle) from the first ALD cycle, and are tetrahedral O-coordinated because the adjacent SiO 2 imprints its tetrahedral near-order and bond length into the Al-O MLs. By variation in the tunnel-SiO 2 thickness and the number of Al-O MLs, we demonstrate that the tetrahedral coordination alone is not sufficient for the formation of Q fix but that a SiO 2 /Al 2 O 3 interface within a tunneling distance from the substrate must be present. The Al-induced acceptor states at these interfaces have energy levels slightly below the Si valence band edge and require charging by electrons from either the Si substrate or from Si/SiO 2 dangling bonds to create a negative Q fix . Hence, tunneling imposes limitations for the SiO 2 and Al 2 O 3 layer thicknesses. In addition, Coulomb repulsion between the charged acceptor states results in an optimum number of Al-O MLs, i.e., separation of both interfaces. We achieve maximum negative Q fix of ∼5 × 10 12 cm -2 (comparable to thick ALD-Al 2 O 3 on Si) with ∼1.7 nm tunnel-SiO 2 and just seven ALD-Al 2 O 3 cycles (∼8 Å) after optimized annealing at 850 °C for 30 s. The findings are discussed in the context of a passivating, hole-selective tunnel contact for high-efficiency Si solar cells.