Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO 2 interface

summary

The gist

Die paper presents a theory for electrically detected magnetic resonance (EDMR) and near-zero field magnetoresistance (NZFMR) by generalizing Shockley Read Hall trap-assisted recombination current…

In short

The paper develops a theory using stochastic Liouville equations to model spin-selective recombination currents in Si/SiO2 MOSFETs, explaining near-zero field magnetoresistance (NZFMR) and electrically detected magnetic resonance (EDMR). It shows that a semiclassical approach for hyperfine interactions is sufficient to explain observed line shapes better than quantum models, suggesting NZFMR spectroscopy can diagnose defects.

Key concepts

Stochastic Liouville Equations
These equations are used as a mathematical framework to describe the spin density matrix of carriers in the MOSFET. They naturally account for spin-selective processes during recombination, providing a more general treatment than previous models by allowing for stochastic behavior.
Near-Zero Field Magnetoresistance (NZFMR)
NZFMR is a phenomenon observed at very low magnetic fields in these devices. The paper uses modeling to explain the line shapes of NZFMR and EDMR as they change with forward bias, linking them to spin-dependent recombination currents.
Hyperfine Interactions
These are interactions between the electron's spin and the nuclear spins of nearby atoms, like those in silicon. The study compares semiclassical (spin vectors) and quantum mechanical (spin operators) models to determine which interaction type best explains the observed NZFMR and EDMR features.
Semiclassical Approximation
The authors found that using a semiclassical approximation for the defect spin is more sensible than a full quantum model. This approach works well because defects near the interface are likely to experience many silicon hyperfine interactions simultaneously, which the semiclassical model can handle effectively.

Terminology used across episodes

This episode discusses

The paper

Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO 2 interface · Read on arXiv

Nicholas J. Harmon, James P. Ashton, Patrick M. Lenahan, Michael E. Flatte

Department of Physics, University of Evansville · Department of Engineering Science and Mechanics, Pennsylvania State University · Department of Physics and Astronomy and Optical Science and Technology Center, University of Iowa

DOI: 10.1063/5.0341503

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO 2 interface".

Mira: Here's the gist of the paper. Die paper presents a theory for electrically detected magnetic resonance (EDMR) and near-zero field magnetoresistance (NZFMR) by generalizing Shockley Read Hall trap-assisted recombination current calculations via stochastic Liouville equations, which provides a well understood semiclassical approach to recombination in Si/SiO2 MOSFETs Theoretical Framework for Recombination The authors generalize Shockley Read Hall trap-assisted recombination current calculations via stochastic Liouville equations to account for spin-selective processes in both EDMR and NZFMR The formalism is used to determine line shapes of NZFMR and EDMR as a function of forward bias They use a set of equations known as the stochastic Liouville equations which naturally account for the spin-selective processes This approach utilizes stochastic Liouville equations for a spin density matrix which allows for a more general treatment than either the theories of KSM or Rong provided in the past Modeling Spin-Dependent Recombination Current The recombination current is modeled by modifying the conventional Shockley-Read-Hall description to include ideas of pair exclusivity from KSM and the spin dependence of carrier trapping at a deep level paramagnetic defect site The maximum recombination current is given by an expression that accounts for two-step capture (into the ground state) by the trap The final expression for the maximum recombination current is Imax = q squared A cn Dst ni VF e qVF/2kBT Incorporating Hyperfine Interactions The authors analyze NZFMR and EDMR by treating both quantum and semiclassical hyperfine interactions They perform calculations of the line shape with two different models of the nuclear spin(s): semiclassical (spin vectors) and quantum mechanical (spin operators) The results from our modeling suggest that NZFMR spectroscopy may serve as a new diagnostic for defects in semiconductors and insulators, and especially at the interfaces between them Experimental Validation and Model Comparison The experimental line shape features are quantified by their peak-to-peak amplitudes, ∆pp, which are related to the maximum recombination current by ∆pp = ∂ρS/∂B0 ∗ f(VF) The semiclassical model for the defect spin is certainly sensible since each defect has a high likelihood of experiencing several Si hyperfine interactions The semiclassical approximation does a much better job of explaining the NZFMR and EDMR line shape features simultaneously when compared to the quantum model The authors demonstrate how a quantum model for the hyperfine interaction was insufficient to simultaneously describe the NZFMR and EDMR responses Instead, they utilized a semiclassical approximation for the hyperfine interaction which is reasonable considering the abundance of nuclear spins near the relevant spin pair The results here and those of Frantz et al.

Kai: suggest that the rates kS and kD are not constant but are drawn from a distribution of values Experimental Setup Details The experiments utilized a low field and frequency (5.4 mT and 151 MHz at resonance) custom-built spectrometer consisting of a Kepco BOP 50-2M bipolar power supply The planar Si/SiO2 MOSFET has a 7.5 nm thick SiO2 gate dielectric with a ≈ 41,000 µm squared channel area and a shorted source and drain region The devices were irradiated under a 0.2 V gate bias with a dose of 1 MRad The body-grounding condition corresponds to the peak recombination current and optimal signal-to-noise Hyperfine Interaction Statistics The analysis of the Pb0 defect at the Si/SiO2 interface involves calculating joint probability distributions for any number of the 13 relevant atoms to have spinful nuclei The joint probability distribution for any number of the 13 relevant atoms to have spinful nuclei is P(kd, knn,kS, kB) = Pd(kd)Pnn(knn)Pnnn,B(kB)Pnnn,S(kS), where kd ∈

0,1: , knn ∈

024: , kB ∈

03: and kS ∈

06: The defect is in one of 224 configurations — ∼ 46% of which possess at least one spinful nucleus at either the defect or its nearest/next nearest neighbors The hyperfine interactions are equivalent with its nearest neighbors when that nearest neighbor possesses nuclear spin The paper provides a sound theoretical basis for the near-zero-field magnetoresistance phenomena present at oxide-semiconductor interfaces in technologically relevant devices NZFMR is also observed The model presented is able to qualitatively explain a series of spin-dependent recombination experiments where the gate and forward biases were changed The authors attribute increase in the ratio to the increase in dc-IV current for which the experiments take place The semiclassical model for the defect spin is certainly sensible since each defect has a high likelihood of experiencing several Si hyperfine interactions Many hyperfine interactions at both spins, appears as the most likely possibility to explain both the NZFMR and EDMR relative line shape structure Future Directions The goal of this article has been to provide a sound theoretical basis for the near-zero-field magnetoresistance phenomena present at oxide-semiconductor interfaces in technologically relevant devices We expect the theoretical approach to apply for spin-dependent trap-assisted transport (SDTAT) where EDMR WS LFE ■■■■■■ The discrepancy can be explained by the following considerations that should be borne out by future in-depth analysis of the two approaches The results here and those of Frantz et al.

Mira: suggest that the rates kS and kD are not constant but are drawn from a distribution of values Acknowledgements The project or effort depicted was or is sponsored by the Department of the Defense, Defense Threat Reduction Agency under Grant HDTRA 1-18-1-0012 and Grant 1-16-0008 The content of the information does not necessarily reflect the position or the policy of the federal government, and no official endorsement should be inferred Data Availability The data that support the findings of this study are available from the corresponding author upon reasonable request The data that support the findings of this study are available from the corresponding author upon reasonable request The data that support the findings of this study are available from the corresponding author upon reasonable request References 1 J.

Kai: Wertz, Electron Spin Resonance: Elementary Theory and Practical Applications (Springer, 1970) 2 W.

Title and authors: Kai: So we’re looking at this paper titled "Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO two interface" <ref:2008.08121#pg1,Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance>. It tackles how to get a general theory for two techniques, EDMR and NZFMR, because before this, interpreting those measurements was kind of guesswork.

Mira: They are taking the existing Shockley Read Hall trap-assisted recombination current calculations and they're generalizing them by adding concepts like pair exclusivity from KSM and spin dependence in carrier trapping at a deep level paramagnetic defect site.

Lev: So they’re essentially taking an old description and layering on these new ideas about how pairs interact and how the spin affects the trapping process at that deep level defect.

Kai: This lets them determine the line shapes of NZFMR and EDMR as you change the forward bias, which is super important because it connects the electrical bias you apply to what you actually measure magnetically.

Mira: They're using these equations to map out exactly how the recombination current looks under different applied biases.

Lev: The implication here is that they are providing a way to get a detailed understanding of the physics of spin dependent recombination in Si/SiO two MOSFET systems by using this new approach when they use the stochastic Liouville equations for a spin density matrix <ref:2008.08121#pg1>.

Kai: It’s like moving from just describing what happens to just modeling how it actually evolves under different conditions.

Mira: Right, so instead of a static picture, you're getting the dynamics of the recombination process as you change those biases.

Lev: That’s exactly what the stochastic equations are good at capturing—the time evolution of the system state under perturbation.

The paper's summary: Kai: Now let’s talk about how this approach improves what came before. They aren't just doing a basic calculation; they are refining the methods themselves in a few key ways.

Mira: They focus on two main improvements to their modeling strategy. First, they incorporate the treatment of both quantum and semiclassical hyperfine interactions for lead zero centers when analyzing NZFMR and EDMR.

Lev: So instead of just picking one picture—either purely quantum or purely semiclassical—they calculate both line shapes, which lets them compare those results directly against experimental data.

Kai: That comparison is what allows them to distinguish between the two models based on how well they fit the experimental data they got back.

Mira: They also examined the possibility that a captured carrier spin experiences a hyperfine interaction from nearby atoms in addition to its own unpaired spin at the capturing defect site.

Lev: That’s a deeper level of physics because it considers interactions beyond just the primary capture event, which is where you can start to get more nuanced results.

Kai: So they are pushing the analysis past just looking at the immediate interaction and into how everything around the defect contributes to its behavior.

Mira: And by using those two things together, they managed to obtain a detailed understanding of the physics of spin dependent recombination for Si/SiO two MOSFET systems, which is what they claim as their main outcome <ref:2008.08121#pg1>.

The paper's improvements: Kai: Wrapping up this paper on "Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO two interface" <ref:2008.08121#pg1,Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance>. They've established a solid theoretical basis for the near zero field magnetoresistance phenomena in oxide semiconductor interfaces.

Mira: They’ve shown that NZFMR is observed experimentally, and their model can qualitatively explain a series of spin dependent recombination experiments where both the gate and forward biases were changed.

Kai: So for our listener, what does this mean for us in terms of practical understanding? It means that when you look at those zero field magnetic field measurements on these MOSFETs, there is a structured physical mechanism happening right at that interface.

Lev: And remember, you can’t just assume one rate for your simulations or your hardware calibration because they found that the capture rate k S and the dissociation rate k D are not constant but are drawn from some distribution of values.

Mira: And this paper sets up a path for future work, suggesting that this theoretical approach could apply to spin dependent trap assisted transport, or SDTAT.

Lev: They also pointed out that the discrepancy between different approaches can be explained by looking at those rates k S and k D from a distribution of values in future, in-depth analysis of both approaches.

Kai: So that’s what we have for this paper on "Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance from the Si/SiO two interface" <ref:2008.08121#pg1,Near-Zero-Field Spin-Dependent Recombination Current and Electrically Detected Magnetic Resonance>. Thanks for tuning in to this discussion.

Conclusion: Kai: So we've seen how they used stochastic Liouville equations to model recombination in Si/SiO two MOSFETs, focusing on near zero field magnetoresistance and electrically detected magnetic resonance from the Si/SiO two interface.

Mira: Yeah, they took that core calculation and added ideas about pair exclusivity and the spin dependence of trapping at deep level paramagnetic defects.

Lev: And they modeled both quantum and semiclassical hyperfine interactions for those lead zero centers, which is important because it lets them see how those two different approaches stack up against each other.

Kai: So, what’s the practical implication for us? It means we have a better way to predict the line shapes of NZFMR and EDMR as you change your forward bias.

Mira: It allows us to use this modeling to diagnose defects at oxide semiconductor interfaces, especially where those interfaces are important for devices.

Lev: From an engineering standpoint, they showed that even if you have many nuclear spins around the defect, the semiclassical approximation for the defect spin is actually quite sensible in explaining both NZFMR and EDMR line shapes together.

Kai: That’s a bit surprising because I thought we needed to be super careful about those quantum effects there.

Mira: They show that when you look at how they calculate things, the semiclassical approximation for the defect spin is better at explaining the data than a full quantum mechanical approach in this case.

Lev: That implies that for devices with a lot of nuclear spins near the relevant spins, we might not need to run super complex quantum error correction simulations just to get a good picture of what's happening.

Kai: Exactly, it suggests that the abundance of nuclear spins actually makes the semiclassical view more reliable for fitting these specific line shapes.

Mira: So they also found that the capture rates, k S and k D aren't just fixed numbers; they are drawn from a distribution of values.

Lev: That means when we try to build hardware simulations, we can't just use one single rate for those processes; we have to account for that variation in the rates.

Kai: It’s a big deal because it shows that the physical environment around the defect really matters for how fast things happen.

Mira: So they laid out a solid theoretical framework for understanding these spin dependent recombination phenomena in Si/SiO two MOSFETs by using this stochastic Liouville equation approach.

Lev: The next step, as they suggest, is to see if this same approach can handle spin dependent trap assisted transport where you have different types of transport mechanisms involved.

Kai: That would be a good test because it takes the groundwork they laid here and applies it to a different kind of movement in the material.

More episodes

← Home