🔬 optics

Transition Metal Dichalcogenides Multijunction Solar Cells Toward the Multicolor Limit

This paper establishes a unified thermodynamic framework to evaluate the efficiency limits of transfer-printed transition metal dichalcogenide (TMD) multijunction solar cells, revealing that while a 50-junction device could theoretically reach 84.5%, the realistic TMD bandgap window (1.0–2.1 eV) caps performance at approximately 63.4% due to photon losses outside the absorption range, while also identifying a practical 5-junction design and quantifying the impact of optical and radiative constraints.

Original authors: Seungwoo Lee

Published 2026-05-04

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Seungwoo Lee

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 the Sun as a massive, chaotic orchestra playing every note of the musical scale at once, from the deepest bass to the highest soprano. A traditional solar cell is like a single musician who can only hear and respond to one specific note. If the orchestra plays a high note, the musician ignores it (wasting energy). If it plays a low note, the musician tries to play it but gets confused and loses energy in the process. This is why standard solar cells have a "ceiling" on how much energy they can capture.

This paper, written by Seungwoo Lee, explores a new way to build solar cells using a special family of materials called Transition Metal Dichalcogenides (TMDs). Think of TMDs as "Lego bricks" for light. Unlike traditional solar materials that must be grown together perfectly (like trying to stack mismatched puzzle pieces), TMDs are like sticky notes. You can peel them off one surface and stick them onto another without worrying about whether they fit perfectly. This allows scientists to stack dozens of layers on top of each other, creating a "multijunction" solar cell.

Here is the simple breakdown of what the paper discovers:

1. The "Staircase" of Light

Instead of one big solar cell, imagine building a staircase where each step is tuned to catch a different color of light.

The top step catches the high-energy blue light.
The middle steps catch the green and yellow light.
The bottom step catches the low-energy red and infrared light.

The paper asks: "If we stack as many steps as we want, how efficient can we get?"

2. The "Goldilocks" Zone (The Bandgap Window)

The researchers found a catch. While you could theoretically stack 50 steps to catch every single note the Sun plays, TMD materials only exist in a specific "Goldilocks" range of energy (between 1.0 and 2.1 electron volts). They can't easily make steps for the very deep bass (low energy) or the very high soprano (high energy).

Because of this limitation, the paper shows that adding more than 5 steps doesn't help much.

The Analogy: Imagine you are trying to fill a bucket with water using cups of different sizes. If you only have cups that hold between 1 and 2 liters, adding a 50th cup won't help you fill the bucket any faster once you've already used 5 cups. The extra cups just sit there doing nothing because the "water" (sunlight) outside that size range is already spilling over the sides.
The Result: With these specific materials, stacking 50 layers only gets you about 63.4% efficiency, whereas a theoretical "perfect" material could reach 84.5%. The extra layers don't add value because the materials simply can't catch the light outside their specific range.

3. The "Echo Chamber" Problem (Luminescence)

When a solar cell works, it also glows a little bit (it emits light). In a stack of layers, this light can bounce around.

The Problem: If the top layer glows downward, the layer below it might get confused. It absorbs the light but loses some energy in the process (like a hot potato being passed around).
The Solution: The paper suggests that instead of just stacking more layers, we need to be smarter about how we manage this "glow." We need to stop the light from escaping upward (which is wasted energy) and guide it carefully downward.

4. The "One-Way Street" Idea (Non-Reciprocity)

The paper also tests a futuristic idea: What if we could build a solar cell where light is allowed to go down but not back up?

The Analogy: Imagine a hallway with a one-way door. You can walk in, but you can't walk back out. This prevents the "echo" from wasting energy.
The Finding: For a single solar cell, this trick doesn't help. But for a stack of many layers, it could boost efficiency by a few more percentage points. However, building such a "one-way" device is extremely difficult and remains a theoretical concept for now.

5. The Thickness Trap

Finally, the paper warns against making these layers too thin.

The Analogy: Think of a TMD layer like a very thin net. If the net is too thin, the "fish" (photons) swim right through it without getting caught. Even though TMDs are naturally thin, the paper shows that to catch enough light, you need to stack them to a moderate thickness (tens to hundreds of nanometers) or use special mirrors to bounce the light back and forth so it gets caught.

The Bottom Line

The paper concludes that for these specific "Lego brick" solar materials, quality is more important than quantity.

Don't try to build a 50-layer tower; it won't work better than a 5-layer tower.
Focus on building a high-quality 5-layer stack where every layer is perfect, the light is trapped effectively, and the materials are stable.
This approach brings us closer to the theoretical maximum efficiency for these materials, which is a significant step forward for the future of solar energy.

Technical Summary: Transition Metal Dichalcogenides Multijunction Solar Cells Toward the Multicolor Limit

Problem Statement
Conventional multijunction photovoltaics are constrained by lattice matching and thermal expansion mismatches, which limit the number of junctions ( $N$ ) that can be epitaxially stacked. While van der Waals (vdW) semiconductors, particularly Transition Metal Dichalcogenides (TMDs), offer a pathway to transfer-printed, lattice-mismatch-free stacking of arbitrary junction numbers, the theoretical efficiency limits for such systems remain under-examined under realistic material constraints. Specifically, it is unclear how the limited bandgap window of TMDs (typically 1.0–2.1 eV) affects the approach to the multicolor limit, and how thermodynamic penalties—such as entropy loss from upward emission, luminescent coupling, and finite external radiative efficiency (ERE)—scale with $N$ in these stacks. Furthermore, the interplay between ultrathin absorber thickness, excitonic absorption, and nanophotonic light trapping in achieving high efficiency requires a unified thermodynamic treatment.

Methodology
The authors develop a unified thermodynamic framework for multijunction photovoltaic devices tailored to transfer-printed TMD stacks. Key methodological components include:

Detailed-Balance Framework: The model extends the Shockley–Queisser (SQ) limit to an unlimited-junction, split-spectrum multijunction system. It utilizes dynamic programming (DP) to optimize bandgap ladders ( $E_{g,1} > E_{g,2} > \dots > E_{g,N}$ ) for both unconstrained bandgaps and a conservative TMD window (1.0–2.1 eV).
Reciprocity and Emission Channels: The model enforces reciprocity between absorption and emission. It explicitly accounts for two-sided emission (upward and downward) in stacks lacking intermediate mirrors. Downward emission is treated as a source of luminescent coupling ( $J_{LC}$ ) to lower junctions, while upward emission is quantified as an entropy-loss proxy.
Finite Absorptance and Thickness Constraints: Moving beyond step-absorber models, the framework incorporates thickness-dependent absorptance ( $A(E; t)$ ) based on excitonic absorption coefficients (including A/B exciton peaks and Urbach tails). A normalized broadband enhancement function $g(t)$ is introduced to model nanophotonic light trapping, acknowledging that the geometric-optics Yablonovitch limit ( $4n^2$ ) is only approachable in optically thick regimes.
Loss Accounting: The framework distinguishes between thermodynamic limits and implementation-level derating. It models penalties from finite ERE, parasitic absorption, series resistance, and interface recombination as separate factors applied to the ideal thermodynamic ceiling.
Nonreciprocal Benchmarking: To assess the potential of nonreciprocal optics, the authors apply an idealized nonreciprocal multijunction model (suppressing upward emission) to the reciprocity-optimized bandgap ladders, providing a conservative estimate of the maximum theoretical gain.

Key Contributions

Unified Thermodynamic Framework: A comprehensive model that defines device bandgap-window constraints, optical boundary conditions, and entropy/luminescence penalties for transfer-printed TMD stacks.
Quantification of the Bandgap Window Constraint: The study demonstrates that a realistic TMD bandgap window (1.0–2.1 eV) imposes a strict efficiency plateau, distinct from the unconstrained multicolor limit.
Power-Budget Decomposition: The introduction of "upward-emitted luminescence power" as a direct entropy-loss proxy and a separate metric for "coupling-induced thermalization," clarifying the trade-offs between current boosting via coupling and voltage losses.
Thickness-Dependent Light Trapping: A physically motivated proxy linking absorber thickness to broadband light-trapping efficiency, highlighting that ultrathin films cannot achieve unity absorptance without significant optical path enhancement.
Material Mapping: A practical mapping of the optimal $N=5$ bandgap ladder to specific vdW/TMD candidates (e.g., WS $_2$ , MoS $_2$ , MoSe $_2$ , WSe $_2$ , MoTe $_2$ ), identifying the bottom cell (near 1.0 eV) as the critical bottleneck.

Results

Efficiency Plateau: Under full concentration, unconstrained multijunctions approach ~~84.5% efficiency at $N=50$ . In contrast, the conservative TMD window (1.0–2.1 eV) plateaus near 63.4% at $N=50$ . The majority of the gain is realized by $N=5$ (~~61.5%), indicating diminishing returns for adding junctions beyond this point within the TMD window.
Dominance of Optical and Radiative Quality: For $N > 5$ in the TMD window, efficiency is no longer limited by spectral partitioning but by photons outside the bandgap window, radiative quality (ERE), and optical losses.
Optimal Ladder: The DP optimizer identifies an experimentally achievable $N=5$ ladder with bandgaps of approximately (2.10, 1.78, 1.50, 1.24, 1.00) eV.
ERE Sensitivity: Efficiency degrades logarithmically with ERE. A stack with $N=5$ and ERE $= 10^{-2}$ loses approximately 6 absolute percentage points compared to ERE $= 1$ under full concentration, emphasizing that high radiative quality is required in every subcell, not just the top one.
Thickness Effects: Finite absorptance in ultrathin TMDs significantly reduces both short-circuit current ( $J_{sc}$ ) and open-circuit voltage ( $V_{oc}$ ) due to reciprocity. Achieving high efficiency requires moderate thicknesses (tens to hundreds of nanometers) or strong light trapping, as single-pass absorption is insufficient.
Luminescent Coupling: In reciprocal stacks without intermediate mirrors, downward emission creates coupling currents but also introduces thermalization losses in lower junctions, particularly when bandgap spacings are small.
Nonreciprocal Gain: Applying an idealized nonreciprocal model to the TMD-optimized ladder yields a modest efficiency increase (reaching ~67.7% at $N=50$ vs. ~63.4% for reciprocal), primarily by suppressing upward entropy loss. However, for single junctions, nonreciprocity provides no benefit, consistent with previous findings.

Significance and Claims
The paper claims to provide a rigorous thermodynamic ceiling for transfer-printed TMD multijunction solar cells, moving beyond idealized assumptions to include realistic material constraints. The authors assert that:

$N \approx 5$ is the Practical Target: Within the conservative TMD bandgap window, pushing for extremely large $N$ offers diminishing returns. The immediate experimental objective should be realizing high-quality, optically isolated stacks with $N \le 5$ rather than maximizing junction count.
System-Level Requirements: High efficiency requires high ERE in every subcell, precise control of emission channels (via intermediate mirrors or angular filters) to manage entropy and coupling, and co-designed photonic structures to overcome thickness limitations.
Nonreciprocity as an Asymptotic Strategy: While nonreciprocal optics can theoretically improve multijunction efficiency by reducing entropy loss, the gain is marginal for single junctions and requires complex implementation. The primary path to high efficiency lies in optimizing reciprocal designs with high-quality materials and optics.
Design Rules: The work establishes that interface defects, exciton collection efficiency, parasitic absorption, and series resistance must be treated as explicit derating factors applied to the thermodynamic limit, rather than hidden within the ideal efficiency calculation.

The study concludes that while TMDs offer a unique integration pathway, approaching the multicolor limit requires overcoming specific thermodynamic penalties related to the limited bandgap window and the optical constraints of ultrathin absorbers, rather than simply stacking an infinite number of junctions.