🔬 materials science

First principles band structure of interacting phosphorus and boron/aluminum $δ$ -doped layers in silicon

Using Density Functional Theory, this study reveals that interacting phosphorus and boron/aluminum $\delta$ -doped layers in silicon mimic intrinsic silicon at separations under 1 nm due to potential cancellation, while behaving as independent p-n junctions at larger distances with enhanced tunneling probabilities.

Original authors: Quinn T. Campbell, Andrew D. Baczewski, Shashank Misra, Evan M. Anderson

Published 2026-02-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Quinn T. Campbell, Andrew D. Baczewski, Shashank Misra, Evan M. Anderson

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine silicon, the material that makes up the brains of our computers, as a vast, quiet city. Usually, this city is empty and neutral. But to make it work, engineers need to add "citizens" with specific jobs: some are donors (like Phosphorus) who give away extra electrons (negative charge), and some are acceptors (like Boron or Aluminum) who are hungry for electrons (positive charge).

For a long time, scientists could only build these cities with one type of citizen at a time. But thanks to a new technology called Atomic Precision Advanced Manufacturing (APAM), we can now place these citizens with single-atom precision. We can build a single, ultra-thin line of donors, and right next to it, a single line of acceptors. These lines are called $\delta$ -doped layers (delta layers).

This paper is like a simulation of what happens when you build two of these lines facing each other in the silicon city, separated by a tiny gap. The researchers wanted to know: Do these two lines talk to each other? Do they cancel each other out? Or do they work together to create something new?

Here is the story of their findings, broken down into simple concepts:

1. The "Hug" vs. The "High Five" (Distance Matters)

The most important discovery is that the distance between these two lines changes everything.

The "Hug" (Less than 1 nanometer apart):
Imagine the donor line and the acceptor line are so close they are practically hugging. Because they are so close, their electric fields overlap and cancel each other out.
- The Result: The silicon city forgets it has any special citizens. It acts just like intrinsic silicon (pure, boring silicon). The "magic" of the extra electrons and the hungry holes disappears. It's like two people shouting opposite things at the same time; the noise cancels out, and you hear silence.
The "High Five" (More than 1 nanometer apart):
Now, imagine you pull them apart just a tiny bit—about the width of a few atoms (1 to 10 nanometers). They are no longer hugging; they are giving a high five.
- The Result: They stop canceling each other out. Now, you have a distinct "Negative Zone" and a distinct "Positive Zone." This creates a structure that acts like a diode (a one-way valve for electricity), but with a very special, ultra-thin gap in the middle.

2. The "Ghost" Tunnel

One of the coolest parts of the paper is about tunneling. In the quantum world, particles can sometimes "teleport" through barriers they shouldn't be able to cross.

The Analogy: Imagine a hill (a barrier) that is too high for a ball to roll over. In normal silicon, the ball would get stuck. But in this quantum world, the ball can sometimes "tunnel" through the hill as if it were a ghost.
The Discovery: The researchers found that when these two $\delta$ -layers interact, they actually lower the hill or make the tunnel easier to cross compared to a standard silicon barrier.
Why it matters: This means electrons can jump between the donor and acceptor layers much more easily than we expected. This could lead to super-fast, ultra-efficient electronic switches that are much smaller than anything we have today.

3. Boron vs. Aluminum: The "Messy" vs. The "Clean"

The paper also compared using Boron (a common acceptor) versus Aluminum (a newer option).

Boron is the "Messy Neighbor": When Boron is placed in the silicon city, it's a bit clumsy. It pushes the surrounding silicon atoms around, creating a lot of "stress" and ripples. This makes the electric fields a bit fuzzy and harder to define.
Aluminum is the "Clean Neighbor": Aluminum fits in more neatly. It doesn't push the neighbors around as much. This results in cleaner, sharper electric fields.
The Takeaway: While both work, Aluminum might be better for building very precise, high-performance devices because it causes less "traffic" and confusion in the silicon city.

4. The Big Picture: Why Should We Care?

This research is a blueprint for the future of computing.

Smaller is Better: Current computer chips are hitting a limit on how small they can get. By stacking these atomic layers, we can build devices that are only a few atoms thick.
New Types of Computers: This could lead to a new generation of electronics that are faster, use less power, and could even be used for quantum computing (computers that solve problems normal computers can't).
The "Intrinsic" Gap: The ability to turn the silicon "off" (make it act like pure silicon) by bringing the layers close together, and "on" (make it act like a diode) by pulling them apart, gives engineers a new "switch" to play with.

Summary

Think of this paper as a guidebook for building atomic-scale Lego structures.

If you stack the pieces too close, they cancel out and become invisible.
If you space them just right, they create a powerful, one-way electrical highway.
And the "traffic" (electrons) can move through this highway much faster than we thought possible, thanks to quantum tunneling.

This isn't just theory; it's a roadmap for building the next generation of super-computers right inside the silicon chips we use every day.

1. Problem Statement

Atomic Precision Advanced Manufacturing (APAM) allows for the placement of dopant atoms in silicon with single-atom precision, creating ultra-thin $\delta$ -doped layers. While significant progress has been made in creating $n$ -type (phosphorus) and $p$ -type (boron/aluminum) $\delta$ -layers independently, there is a lack of understanding regarding the electronic structure of interacting bipolar $\delta$ -layers (donor and acceptor pairs) within the same silicon lattice.

Key questions addressed include:

How do the electronic structures of opposing polarity $\delta$ -layers interact at varying separation distances?
Does the interaction create a direct bandgap semiconductor, or does it mimic intrinsic silicon?
How does the separation distance affect charge transfer mechanisms, specifically tunneling, between the layers?

2. Methodology

The authors employed Density Functional Theory (DFT) to simulate the electronic properties of interacting $\delta$ -doped layers.

System Configuration:
- Substrate: $2\times1$ Si(100) slabs using a supercell approach.
- Doping: Two $\delta$ -layers placed at 1/4 monolayer coverage. One layer is Phosphorus (P, donor), and the other is either Boron (B) or Aluminum (Al, acceptors).
- Separation Distances: Simulations were run for separations ranging from 0.1 nm to 10 nm.
- Boundary Conditions: Periodic boundary conditions were used, creating an infinite array of $\delta$ -layer pairs with a 21.8 nm spacing between images (sufficient to converge image effects).
Computational Details:
- Software: QUANTUM ESPRESSO package.
- Exchange-Correlation Functional: SCAN (Strongly Constrained and Appropriately Normed) meta-GGA. This was chosen as a compromise between the high cost of hybrid functionals and the inaccuracy of standard semilocal functionals, providing better bandgap estimates for semiconductors.
- Parameters: Kinetic energy cutoffs of 680 eV (orbitals) and 2721 eV (charge density).
- Analysis: Calculations included band structures, Local Density of States (LDOS), and planar-averaged electrostatic potentials.
Tunneling Calculation:
- The tunneling probability $T(E)$ was estimated using the WKB approximation based on the local potentials ( $V(z)$ ) extracted from the DFT LDOS calculations.
- Effective mass ( $m^*$ ) was assumed to be $0.19m_e$ (standard transverse mass in Si).

3. Key Contributions

First-Principles Interaction Study: This is one of the first DFT studies to model the interaction between both donor and acceptor $\delta$ -layers simultaneously, moving beyond single-layer analyses.
Distance-Dependent Behavior: The study establishes a critical threshold (~1 nm) separating two distinct regimes of electronic behavior: strong mutual compensation vs. independent junction behavior.
Material Comparison: It provides a comparative analysis between Boron and Aluminum as acceptors, revealing that Aluminum induces less lattice strain and results in less suppression of the $\delta$ -layer bands compared to Boron.
Tunneling Enhancement: The paper quantifies tunneling probabilities in these structures, demonstrating that $\delta$ -layer interactions enhance tunneling compared to a standard intrinsic silicon barrier.

4. Key Results

A. Electronic Structure vs. Separation Distance

The interaction between the P and B/Al layers is highly dependent on the separation distance ( $d$ ):

$d < 1$ nm (Strong Compensation):
- The donor and acceptor potentials overlap significantly and largely cancel each other out.
- The electronic structure closely mimics intrinsic silicon.
- The effective bandgap is reduced (e.g., from 0.98 eV to 0.86 eV at 0.1 nm), but the material does not become a direct gap semiconductor; the Fermi level remains within the gap, and charge carriers are screened.
- The LDOS shows adjacent triangular potentials rather than distinct impurity bands.
$d > 1$ nm (Independent Behavior):
- The layers behave as independent entities, effectively forming a p-n junction with an intrinsic silicon layer acting as the depletion region.
- Distinct impurity bands appear: a donor band near the Conduction Band Minimum (CBM) and an acceptor band near the Valence Band Maximum (VBM).
- The energy separation between the donor and acceptor peaks decreases as distance increases (e.g., 0.45 eV at 1 nm, 0.36 eV at 2 nm).
- At 10 nm, the bands overlap in energy, creating a structure akin to an ultra-short p-n junction with high precision unattainable by ion implantation.

B. Boron vs. Aluminum

Boron (B): Induces significant stress in the surrounding silicon lattice, leading to strong atomic displacement. This results in a "noisier" potential landscape and stronger suppression of the $\delta$ -layer induced bands at close range.
Aluminum (Al): Induces less lattice stress. Consequently, Al-P structures show less suppression of the $\delta$ -induced bands compared to B-P structures, resulting in cleaner and more distinct peaks in the LDOS even at smaller separations.

C. Carrier Transport and Tunneling

Mechanism: Charge transfer between spatially separated layers (without external bias) occurs primarily via tunneling.
Probability: The calculated tunneling probability for B-P and Al-P layers separated by 1–2 nm is significantly higher than what would be expected for a standard 1.1 eV triangular barrier in intrinsic silicon.
- Example: At 1 nm separation, $T \approx 0.141$ ; at 2 nm, $T \approx 0.0305$ .
Implication: The interaction between $\delta$ -layers creates a potential landscape that facilitates easier electron tunneling than traditional junctions, suggesting enhanced transport capabilities for bipolar devices.

D. Density of States (DOS)

The DOS at the Fermi level increases as the separation distance increases.
At small separations, the DOS is low (intrinsic-like).
At larger separations ( $>1$ nm), the Fermi level moves into the valence bands (for the acceptor side), leading to weak metallic-like behavior where electrons can move easily between bands.

5. Significance and Future Outlook

Device Design: These findings provide a theoretical foundation for designing silicon-based bipolar devices and superlattices using APAM. The ability to tune the effective bandgap and tunneling rates by controlling the inter-layer distance offers a new degree of freedom for silicon electronics.
Beyond Traditional Diodes: The study suggests that interacting $\delta$ -layers can achieve outcomes (enhanced tunneling, specific band alignments) that are impossible with traditional ion-implanted p-n junctions.
Experimental Validation: The authors propose that Angle-Resolved Photoemission Spectroscopy (ARPES) could be used to validate these band structures.
Challenges: Future work must address the experimental difficulty of growing defect-free intrinsic silicon between the layers and the potential for dopant segregation during epitaxy. The study also suggests extending this research to other Group-IV materials (SiGe, Ge, GeSn) to explore strain effects.

In conclusion, the paper demonstrates that interacting donor and acceptor $\delta$ -layers in silicon can be tuned from an intrinsic-like state to a functional p-n junction simply by varying the atomic-scale separation, with significant implications for quantum computing and advanced silicon electronics.