Abstract Al$_2$O$_3$ is a versatile high-$\kappa$ dielectric that has excellent surface passivation properties on crystalline Si (c-Si), which are of vital importance for devices such as light emitting diodes and high-efficiency solar cells. We demonstrate both experimentally and by simulations that the surface passivation can be related to a satisfactory low interface defect density in combination with a strong field-effect passivation induced by a negative fixed charge density $Q_f$ of up to $10^{13}$ cm$^{-2}$ present in the Al$_2$O$_3$ film at the interface with the underlying Si substrate. The negative polarity of $Q_f$ in Al$_2$O$_3$ is especially beneficial for the passivation of $p$-type c-Si as the bulk minority carriers are shielded from the c-Si surface. As the level of field-effect passivation is shown to scale with $Q_f^2$, the high $Q_f$ in Al$_2$O$_3$ tolerates a higher interface defect density on c-Si compared to alternative surface passivation schemes.
Introduction: The Surface Recombination Problem
In crystalline silicon (c-Si) solar cells, the surface represents a significant "dead zone." The abrupt termination of the crystal lattice leaves behind "dangling bonds" that act as recombination centers. These traps allow electrons and holes to recombine, wasting energy that would otherwise contribute to electrical current. To mitigate these losses, researchers employ passivation layers. While traditional materials like silicon dioxide ($SiO_2$) or silicon nitride ($SiN_x$) are effective, they typically carry a positive fixed charge. This is problematic for $p$-type silicon, as positive charges attract minority electrons to the surface, potentially increasing recombination. The industry required a material providing high-quality passivation with a negative charge to protect $p$-type surfaces effectively.
Methodology
The researchers utilized plasma-assisted Atomic Layer Deposition (ALD) to grow thin amorphous $Al_2O_3$ films (6–32 nm) on silicon wafers. To dissect the underlying physics, they employed:
- Corona Charging: External charges were deposited on the $Al_2O_3$ surface to neutralize internal fixed charges, enabling precise measurement of the inherent charge density ($Q_f$).
- Sinton Lifetime Measurements: Carrier "effective lifetime" was measured to quantify passivation quality.
- Infrared Spectroscopy (FTIR): Atomic bonding changes (e.g., Si-O, Al-O) were monitored post-heat treatment.
- Numerical Simulations: The extended Shockley-Read-Hall (SRH) model was used to simulate the relationship between recombination velocity ($S_{eff}$) and charge density.
The Dual Passivation Mechanism
The study reveals that $Al_2O_3$ provides a "one-two punch" for surface protection. It combines chemical passivation (reducing dangling bonds) with powerful field-effect passivation. The latter is driven by a massive negative fixed charge ($Q_f \approx 10^{13} \text{ cm}^{-2}$) located near the interface.
High-resolution microscopy (TEM) identified a thin (~1.5 nm) $SiO_x$ interfacial layer between the silicon and the $Al_2O_3$. This layer is critical, as the negative charge—likely originating from aluminum vacancies or interstitial oxygen—is situated very close to this junction.
Scaling Laws and Performance
A key finding is the $1/Q_f^2$ scaling law: the surface recombination velocity ($S_{eff}$) scales inversely with the square of the fixed charge. This relationship implies that the high charge density in $Al_2O_3$ allows it to tolerate a higher density of physical interface defects ($D_{it}$) compared to $SiO_2$ while still maintaining superior performance.
Key results include:
- Record Lifetimes: Effective carrier lifetimes exceeding 6 ms on $n$-type silicon.
- Charge Magnitude: A quantified negative fixed charge of $-(1.3 \pm 0.1) \times 10^{13} \text{ cm}^{-2}$.
- Field-Effect Strength: Passivation up to four orders of magnitude stronger than thermal $SiO_2$.
- Efficiency: The material enabled solar cell efficiencies of up to 23.2% in contemporary trials.
Conclusion and Future Outlook
While the paper identifies the location and impact of the charge, the exact atomic-scale origin remains a subject of theoretical modeling. Future work focuses on better controlling the "unintentional" $SiO_x$ interfacial layer to optimize the balance between charge and defect density. Despite these open questions, this work established $Al_2O_3$ as a cornerstone material for high-efficiency silicon photovoltaics.
Summary updated