Topology of slices through the Sierpinski tetrahedron

This paper establishes a sharp topological dichotomy for slices of the Sierpinski tetrahedron, demonstrating that slices at dyadic rational heights possess finitely many connected components with infinite first Čech homology, whereas slices at non-dyadic rational heights are totally disconnected with vanishing positive-degree homology.

The Gist

Imagine you have a magical, infinitely detailed 3D object called the Sierpiński Tetrahedron. Think of it like a giant, hollow pyramid made of smaller and smaller hollow pyramids, repeating forever. It's a fractal: if you zoom in on any part of it, you see the same pattern again and again.

Now, imagine you have a giant, invisible laser cutter that can slice this object horizontally at any height, from the very bottom (0) to the very top (1). The question the authors ask is: "What does the slice look like?"

You might expect the slice to be a messy, complicated shape. But the authors discovered something fascinating: the shape of the slice depends entirely on how you describe the height number you chose.

Here is the breakdown of their discovery, using simple analogies:

1. The Two Types of Heights

The authors divide all possible heights into two camps based on how you write the number in binary (using only 0s and 1s, like a computer code).

Camp A: The "Dyadic" Heights (The Clean Cuts)

These are heights that can be written as a fraction with a power of 2 in the denominator (like 1/2, 3/8, 5/16). In binary, these numbers eventually stop or repeat in a very specific way (ending in ...01 or ...10).

• The Analogy: Imagine slicing a loaf of bread at a mark that lines up perfectly with the grain.
• The Result: When you slice at these heights, the resulting shape is connected. It looks like a collection of "Sierpiński Gaskets" (which are flat, 2D versions of the pyramid, looking like a triangle with holes in it).
  • It has a finite number of separate pieces (like 3 or 9 distinct islands).
  • However, each piece is full of holes (infinite loops), so it has "infinite complexity" in its structure.
  • Topologically: It's a solid, connected shape with a lot of tunnels.

Camp B: The "Non-Dyadic" Heights (The Chaotic Cuts)

These are heights that cannot be written as a simple fraction with a power of 2 (like 1/3, 1/π, or 0.1010010001...). In binary, these numbers go on forever without settling into a simple pattern.

• The Analogy: Imagine slicing the loaf of bread at a mark that falls between every single grain, hitting only the empty air.
• The Result: When you slice at these heights, the object doesn't just break apart; it shatters into dust.
  • The slice is totally disconnected. It's not a shape; it's a cloud of infinitely many individual points, with no two points touching each other.
  • Topologically: It has no holes, no lines, and no connections. It is just a scattered collection of points.

2. The "Binary Code" Connection

Why does this happen? The authors realized that the Sierpiński Tetrahedron is built by a set of rules that follow a binary code (0s and 1s).

• If your height number has a "clean" binary code (ending in a specific pattern), the slicing rules align perfectly, leaving you with whole, connected pieces.
• If your height number has a "messy," infinite binary code, the slicing rules keep shifting back and forth, preventing any pieces from ever connecting. It's like trying to build a bridge where the planks keep moving away from each other every time you try to lay one down.

3. The "Dust" vs. "Islands" Summary

The paper proves a sharp "dichotomy" (a split into two distinct categories):

If the height is...	The Slice Looks Like...	Topological "Vibe"
Dyadic Rational (e.g., 0.5, 0.75)	Islands of Swiss Cheese (Connected shapes with infinite holes)	Connected but complex.
Non-Dyadic (e.g., 1/3, 0.10101...)	Galaxy of Dust (Infinite points, none touching)	Totally Disconnected.

4. Why Does This Matter?

You might ask, "Who cares about slicing a math pyramid?"

The authors are using this specific shape to test a new way of measuring the "shape" of complex, broken objects. They are using a tool called Čech Homology (a fancy mathematical way of counting holes and connections).

• The Big Picture: This research helps mathematicians understand how complex shapes behave when you cut them. It shows that for fractals, the "smoothness" of a cut isn't about the physical knife; it's about the mathematical nature of the number you choose.
• The Takeaway: Even in a world of infinite complexity, there are strict rules. If you pick the "right" numbers, you get structure. If you pick the "wrong" numbers, you get chaos.

In a Nutshell

The Sierpiński Tetrahedron is a shape that is incredibly sensitive to how you slice it.

• Slice it at "nice" numbers, and you get connected, hole-filled islands.
• Slice it at "messy" numbers, and you get dust.

The authors mapped out exactly which numbers give you islands and which give you dust, proving that the topology (the shape's connectivity) changes abruptly based on the binary code of the height.

Technical Summary

Here is a detailed technical summary of the paper "Topology of Slices Through the Sierpiński Tetrahedron" by Yuto Nakajima and Takayuki Watanabe.

1. Problem Statement

The paper investigates the topological properties of slices (level sets) of the Sierpiński tetrahedron, a classic fractal generated by an Iterated Function System (IFS) of similitudes in $\mathbb{R}^3$.

While the dimension of fractal slices has been extensively studied (e.g., Marstrand-type theorems, Moran constructions), the topological structure of these slices remains less understood. Specifically, the authors aim to characterize the connectivity and homological properties of the slice $J_c = \{x \in J : p_z(x) = c\}$ for every height $c \in [0, 1]$, where $p_z$ is the projection onto the third coordinate.

A central difficulty addressed is that while the Sierpiński tetrahedron $J$ is the limit set of a standard IFS, its slices $J_c$ are generally not limit sets of standard IFSs. This necessitates a more generalized framework to analyze their topology.

2. Methodology

The authors employ a combination of Non-Autonomous Iterated Function Systems (NIFS) and Čech (co)homology theory.

A. Non-Autonomous IFS (NIFS) Framework

To represent the slices, the authors generalize the standard IFS concept.

• Definition: A NIFS is a sequence of collections of contracting maps $(\Phi^{(j)})_{j=1}^\infty$. Unlike standard IFSs, the set of maps can change at each iteration step.
• Application to Slices: For a given height $c$, the slice $J_c$ is realized as the limit set of a specific NIFS $\Phi_c$. The structure of this NIFS is determined by the binary expansion of $c$, denoted as $(a_j(c))_{j=1}^\infty$.
  • If $a_j(c) = 0$, the system uses a single map.
  • If $a_j(c) = 1$, the system uses three maps.
• Reduction: The authors prove that the slice $J_c$ is homeomorphic to the limit set of a projected NIFS $\Psi_c$ acting on the unit square $[0, 1]^2$. This reduces the 3D problem to a 2D analysis.

B. Čech (Co)homology and Nerves

To capture refined topological features (beyond simple connectivity), the authors utilize Čech homology ($\check{H}_q$) and Čech cohomology ($\check{H}^q$).

• Nerve Construction: They construct a sequence of simplicial complexes (nerves), denoted $N_{1,k}$, based on the coverings of the limit set induced by the NIFS at step $k$.
• Inverse/Direct Limits: The homology of the slice is defined as the inverse limit of the homology groups of these nerves:
  $$\check{H}_q(J_c) \cong \varprojlim_k H_q(N_{1,k})$$
  This framework, previously established by Sumi and extended by the authors in prior work, allows for the calculation of homology ranks for non-self-similar limit sets.

3. Key Contributions and Results

The paper establishes a sharp dichotomy in the topology of the slices based on whether the height $c$ is a dyadic rational or a non-dyadic rational.

A. The Dichotomy (Main Theorem A)

Let $c \in [0, 1]$.

1. Case 1: $c$ is a Dyadic Rational

   • Structure: The slice $J_c$ is a finite disjoint union of copies of the Sierpiński gasket (the 2D Sierpiński triangle).
   • Connectivity: It has a finite number of connected components ($r \ge 1$).
   • Homology:
     • $\check{H}_0(J_c) \cong \mathbb{Z}^r$ (finite rank).
     • $\check{H}_1(J_c)$ has infinite rank (indicating infinitely many "holes" or loops).
     • $\check{H}_q(J_c) = 0$ for all $q \ge 2$.
   • Rank Formula: If the binary expansion of $c$ is $a_1 \dots a_n 01$ and $\ell$ is the count of zeros in the prefix, then $\text{rank } \check{H}_0(J_c) = 3^{n-\ell}$.

2. Case 2: $c$ is a Non-Dyadic Rational

   • Structure: The slice $J_c$ is totally disconnected (a Cantor-like set).
   • Homology:
     • $\check{H}_q(J_c) = 0$ for all $q \ge 1$.
     • The zeroth homology reflects the infinite number of components, but higher homology vanishes.

B. Growth Rates and Dimensions (Main Theorem B & Corollaries)

The authors analyze the asymptotic growth of the rank of the homology groups of the nerves $N_{1,n}$:

• Dyadic Case: The growth rate of the first homology rank is $\lim_{n \to \infty} \frac{1}{n} \log \text{rank } H_1(N_{1,n}) = \log 3$.
• Non-Dyadic Case: The rank of the zeroth homology is determined by the binary digits:
  $$\log \text{rank } H_0(N_{1,n}) = \sum_{j=1}^n a_j(c) \log 3$$
• Dimensional Implications:
  • For Lebesgue almost every $c$, the limit of the normalized log-rank is $\frac{1}{2} \log 3$, implying a specific average dimension.
  • The paper derives formulas for the Hausdorff, Box, and Packing dimensions of the slices for non-dyadic $c$, linking them directly to the frequency of 1s in the binary expansion of $c$.

C. Generalization to Higher Dimensions

The results are extended to the $d$-dimensional Sierpiński gasket ($d \ge 3$).

• The dichotomy holds: Dyadic slices are unions of $(d-1)$-dimensional gaskets; non-dyadic slices are totally disconnected.
• The rank of the zeroth homology for dyadic $c$ becomes $d^{n-\ell}$.
• The growth rate for the first homology becomes $\log d$.

4. Significance

1. Topological Classification: This work provides the first rigorous topological classification of slices of the Sierpiński tetrahedron, moving beyond dimensional analysis to characterize connectivity and "holes."
2. Methodological Advancement: It successfully applies the theory of Non-Autonomous IFSs to solve slicing problems where standard self-similarity is lost. This bridges the gap between the geometry of fractals and the dynamics of non-stationary systems.
3. Dichotomy Insight: The result highlights a fundamental difference in fractal geometry: the "rational" (dyadic) heights preserve the fractal's self-similar structure (gaskets), while "irrational" (non-dyadic) heights destroy connectivity entirely, resulting in totally disconnected sets.
4. Complement Topology: Using Alexander Duality, the authors deduce the homological structure of the complement of the slices in $\mathbb{R}^d$, showing that the topology of the void space is also dictated by the arithmetic nature of the slice height.

In summary, Nakajima and Watanabe demonstrate that the arithmetic properties of the slicing parameter $c$ (specifically its binary expansion) strictly dictate the topological complexity of the resulting fractal slice, creating a clear boundary between structured, hole-rich sets and totally disconnected dust.