Experimental Verification of a Universal Operator Growth Hypothesis

This paper experimentally verifies a universal hypothesis regarding the growth of Lanczos coefficients using $^{19}$F nuclear magnetic resonance free induction decay data, confirming the theory and determining the growth parameter $\alpha$ for three crystal orientations while discussing the conditions necessary to observe a branch-point singularity in the analytic continuation of the signal.

The Gist

The Big Picture: Listening to the "Echo" of Spinning Atoms

Imagine you have a giant, perfectly organized room full of tiny, spinning tops (these are atomic nuclei in a crystal). If you give them all a little push at the same time, they start spinning in sync. But very quickly, they start bumping into each other, getting confused, and losing their rhythm.

In physics, this "losing rhythm" is called Free Induction Decay (FID). It's like the echo of a shout in a canyon that slowly fades away.

For decades, physicists have been trying to figure out exactly how that echo fades. Does it fade smoothly like a gentle slide? Or does it hit a hidden "cliff" where the math breaks down?

This paper is about two scientists, M. Engelsberg and Wilson Barros Jr., who looked at old, high-quality data from Calcium-43 crystals to answer a big question: Is there a hidden "speed limit" or a "singularity" in how these atoms lose their energy?

The Mystery: The "Entire" Function vs. The "Broken" One

To understand their discovery, let's use a Lego analogy.

1. The "Entire" Function (The Infinite Tower):
   Imagine trying to build a tower out of Legos that goes up forever. If you can build it perfectly for any height without it ever collapsing, mathematicians call this an "entire function." It's smooth, predictable, and has no weak spots.

   • The Old Belief: Many physicists thought the fading echo of the atoms was like this infinite tower. They thought if you looked at the math closely enough, it would just keep going smoothly forever.

2. The "Universal Growth Hypothesis" (The Speed Limit):
   A few years ago, a group of theorists (Parker et al.) proposed a new idea. They suggested that in complex quantum systems, "complexity" grows at a maximum speed, like a car hitting a speed limit.

   • The Prediction: If this hypothesis is true, the math describing the echo shouldn't be an infinite tower. Instead, it should hit a branch-point singularity.
   • The Metaphor: Imagine driving down a road that looks smooth, but suddenly, at a specific distance, the road turns into a cliff. You can't drive past it. The math "breaks" there. This paper is the experimental proof that the road does have a cliff.

The Experiment: Finding the Cliff

The authors took data from Calcium-43 crystals. Why Calcium?

• It's a very clean system (like a pristine pool of water with no ripples from wind).
• The atoms interact only through simple magnetic forces (dipole-dipole), making the math easier to track.

They tested three different directions for the magnetic field (like shining a flashlight from the North, East, or Top of the crystal).

The Result:
They found that the data fit a specific mathematical curve (Equation 7 in the paper) perfectly. This curve has a "cliff" (a branch-point singularity) at a specific time.

• Crucial Finding: No other mathematical curve that was "smooth forever" (an entire function) could fit the data as well. The data demanded a cliff.

This strongly supports the "Universal Growth Hypothesis." It proves that the complexity of these quantum systems grows until it hits a hard limit, creating a singularity.

The "Detectability" Puzzle: Seeing the Invisible

The paper also tackles a tricky question: How do you know there's a cliff if you can't see the edge of the map?

Imagine you are walking in the fog. You can only see 10 steps ahead.

• If the cliff is 100 steps away, you might think the path goes on forever because you can't see the edge.
• But if the cliff is 15 steps away, you might start to notice the ground getting weird before you fall.

The authors explain that to detect this "mathematical cliff," you need two things:

1. A very quiet room (High Signal-to-Noise Ratio): You need to hear the echo clearly without background static.
2. A long enough walk: You need to measure the echo for long enough to get close to the cliff, but not so long that the signal disappears into the noise.

They used a mathematical tool called Hadamard's Factorization Theorem (which is like a sophisticated way of checking if a pattern repeats perfectly). They showed that if you have good enough data, you can mathematically prove the cliff exists, even without knowing the exact formula beforehand.

The Twist: Why Some Directions Are Different

The paper found something weird about the crystal directions:

• [100] Direction: The atoms interact strongly, like a crowded dance floor. The "cliff" happens very early.
• [110] vs. [111] Direction: This was a surprise. Even though the [110] direction had stronger interactions (more crowded), the "cliff" happened later than in the [111] direction.

The Analogy:
Think of it like traffic.

• [100] is a straight highway with cars bumper-to-bumper. Traffic jams (singularities) happen instantly.
• [110] is like a one-lane road. Even though the cars are close, they can only move in a line. The "complexity" grows slowly (sub-linear), so the traffic jam takes longer to form.
• [111] is a 3D intersection. The cars can move in all directions. The chaos spreads fast, creating a jam quickly.

The Conclusion: Why This Matters

This paper is a big deal because:

1. It validates a universal theory: It proves that quantum systems have a fundamental "speed limit" on how fast they can become complex.
2. It solves a decades-old puzzle: It confirms that the fading echo of atoms isn't a smooth, infinite curve, but one that hits a hard mathematical wall.
3. It gives us a new tool: It shows scientists how to use old data and new math tricks to find these hidden "cliffs" in other quantum systems, potentially helping us understand everything from black holes to quantum computers.

In short: The authors looked at the "echo" of spinning atoms and proved that the echo doesn't fade into infinity; it hits a wall. This wall confirms that the universe has a built-in speed limit for how chaotic things can get.

Technical Summary

1. Problem Statement

The paper addresses a fundamental question in quantum dynamics regarding the analytic properties of the Nuclear Magnetic Resonance (NMR) Free Induction Decay (FID) in a rigid cubic lattice of spin-½ nuclei (specifically $^{43}\text{Ca}$ in $\text{CaF}_2$).

• The Core Question: Is the FID function, when analytically continued into the complex time plane, an entire function (convergent everywhere, implying infinite radius of convergence for its moment expansion) or does it possess singularities (poles or branch points)?
• Theoretical Context: A hypothesis proposed by Parker et al. suggests that "operator complexity" in Hamiltonian systems grows universally. This growth implies that Lanczos coefficients ($b_n$) grow linearly with $n$ ($b_n \sim \alpha n$). If true, this leads to a finite radius of convergence for the moment expansion, meaning the FID is not an entire function but possesses singularities.
• Experimental Challenge: Previous data lacked the precision or range to definitively distinguish between entire functions (like Gaussians or simple exponentials) and functions with branch-point singularities. Furthermore, calculating high-order moments from the dipolar Hamiltonian is computationally intractable.

2. Methodology

The authors utilized high-quality experimental FID data from $^{43}\text{Ca}$ in $\text{CaF}_2$ (referenced from Ref. 9), which spans two orders of magnitude in signal amplitude. The methodology involved:

• System Characteristics: The study focuses on $^{43}\text{Ca}$ (spin ½, 0.13% abundance) in a rigid cubic lattice at infinite temperature. The interactions are purely dipolar, free from quadrupolar broadening or motional effects.
• Mathematical Framework:
  • Moment Expansion: The FID is expressed as a power series of even moments ($M_{2n}$).
  • Hadamard Factorization Theorem: Used to analyze the relationship between the zeros of the FID and its analytic structure. If the FID were entire, it could be represented by a specific infinite product form.
  • Lanczos Coefficients: The authors tested the hypothesis that these coefficients grow linearly ($b_n \sim \alpha n$), which dictates the nature of the singularities.
• Fitting Strategy:
  • Instead of relying on calculated moments, the authors fitted the experimental data using an empirical function (Eq. 7) that combines an infinite product of zeros with an exponential decay term.
  • This function was designed to behave as a Gaussian at short times and an exponential at long times.
  • Comparison: They compared this "branch-point" function against a standard "entire function" model (Eq. 9, a Gaussian-modulated exponential) to see which better fits the data.
• Singularity Detection Test: The authors proposed a method using Hadamard's theorem to detect singularities in raw data without knowing the underlying function. This involves fitting polynomials to the logarithm of the FID over increasing time intervals and checking for abrupt deviations outside the fitting range.

3. Key Contributions

• Verification of Universal Growth Hypothesis: The paper provides the first strong experimental evidence supporting Parker et al.'s hypothesis that operator complexity grows linearly, leading to non-entire FID functions.
• Identification of Branch-Point Singularities: The authors demonstrate that the FID in $\text{CaF}_2$ is best described by a function with branch-point singularities in the complex plane, rather than being an entire function.
• Quantification of Growth Parameter ($\alpha$): The study successfully calculates the growth parameter $\alpha$ (related to the Lyapunov exponent of operator growth) for three distinct crystal orientations: [100], [110], and [111].
• Methodological Proposal: A novel procedure is proposed to detect analytic singularities in experimental data using polynomial least-squares fits and Hadamard's theorem, applicable even when the theoretical function is unknown.

4. Results

• Superior Fit of Branch-Point Model: The empirical function with branch-point singularities (Eq. 7) provided an excellent fit to the experimental data over nearly two orders of magnitude. In contrast, the entire function model (Eq. 9) failed to match the data as accurately.
• Calculated Parameters:
  • For the [100] orientation (strongest interaction), the singularity position $A$ was found to be 49.69 $\mu$s, corresponding to a growth parameter $\alpha \approx 0.020 \text{ MHz}$.
  • For [110] and [111], specific values for $A$ and $\alpha$ were derived (Table I in the paper), showing an inversion where the [110] orientation (higher interaction strength) has a singularity at a larger time than [111].
• Physical Interpretation of Inversion: The authors attribute the inversion in singularity positions to the spatial connectivity of the spin network:
  • [100]: Dominated by 1D-like chains (sub-linear growth), but strong interactions.
  • [111]: Dipole interactions with nearest neighbors vanish; interactions are dominated by next-nearest neighbors in an isotropic 3D distribution, leading to different connectivity and growth dynamics.
• Detectability Conditions: The simulations (Fig. 1) confirm that singularities are detectable if the signal-to-noise ratio is high enough to allow fitting intervals that extend beyond the convergence radius ($t_{fit} > t_{singularity}$) but remain below the noise floor.

5. Significance

• Validation of Quantum Chaos Theory: The results bridge the gap between abstract quantum chaos theories (operator growth) and concrete experimental NMR data, confirming that universal growth laws apply to many-body spin systems.
• Resolution of Long-standing Debate: It resolves the ambiguity regarding whether the FID is an entire function, suggesting that the "exponential decay" observed in experiments is a consequence of underlying branch-point singularities, not just a phenomenological fit.
• New Diagnostic Tool: The proposed method for detecting singularities via Hadamard factorization offers a powerful tool for analyzing complex quantum systems where exact solutions are impossible.
• Implications for Other Systems: The findings suggest that similar singularities should be observable in other systems, including quasi-1D systems (like fluorapatite), provided the experimental signal-to-noise ratio is sufficient.

In conclusion, the paper establishes that the FID in $\text{CaF}_2$ is not an entire function but possesses branch-point singularities, providing robust experimental validation for the universal hypothesis of linear operator growth in Hamiltonian systems.