⚛️ general relativity

Evolution of Linear Perturbations under Time-Dependent Hubble Friction I: SR-USR-SR Inflation

This paper derives accurate analytical expressions for the linear perturbation dynamics and power spectrum in SR-USR-SR inflation with instantaneous transitions, revealing that the final power spectrum's finite dip results from the cancellation between two growing modes rather than between constant and growing terms.

Original authors: Wen Li, Chao Chen

Published 2026-02-17

📖 5 min read🧠 Deep dive

CC0 1.0

Original authors: Wen Li, Chao Chen

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ 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 early universe as a giant, inflating balloon. In the standard story of the Big Bang, this balloon expands at a steady, predictable pace. Scientists call this "Slow-Roll" inflation. It's like a car cruising on a highway at a constant speed: smooth, boring, and it leaves behind a very uniform pattern of ripples (which we see today as the Cosmic Microwave Background).

But what if, for a brief moment, the driver hit the brakes and then slammed the accelerator? Or what if the car suddenly entered a section of road with zero friction, causing it to speed up uncontrollably?

This paper explores exactly that scenario. It looks at a specific type of cosmic inflation called USR (Ultra-Slow-Roll), sandwiched between two normal "Slow-Roll" phases. Think of it as a Sandwich:

Top Bun: Normal inflation (Steady cruise).
The Meat: Ultra-Slow-Roll inflation (The chaotic, frictionless middle).
Bottom Bun: Normal inflation again (Back to steady cruise).

The Problem: The "Ghost" Dip

When physicists tried to calculate what happens to the tiny quantum ripples (perturbations) during this "Meat" phase, they ran into a puzzle.

In the middle phase (USR), the universe acts like a car on ice. The usual "friction" that usually damps out ripples disappears. Instead of dying down, the ripples start to grow wildly, like a snowball rolling down a hill, getting bigger and bigger. This should create a massive spike in energy on small scales, potentially creating Primordial Black Holes (tiny black holes formed right after the Big Bang).

However, when scientists looked at the math, they saw something weird: a dip. Just before the energy spikes to its maximum, the graph of energy drops down into a valley.

Old Theory: Scientists thought this dip happened because a "steady" force canceled out a "growing" force.
The Paper's Discovery: The authors, Wen Li and Chao Chen, say, "No, that's not right." They found that the dip is actually caused by two growing forces fighting each other. Imagine two people pushing a swing in opposite directions; for a split second, they cancel each other out, and the swing stops moving, creating a dip, before one of them wins and the swing shoots up.

The Method: The "Junction" Trick

To solve this, the authors used a clever mathematical tool called the Junction Method.

Imagine the universe's history as a train journey with three distinct tracks:

Track A (Slow-Roll)
Track B (Ultra-Slow-Roll)
Track C (Slow-Roll)

The train switches tracks instantly. The "Junction Method" is like a precise set of rules for connecting the train cars at the switch points. You have to make sure the speed and position of the train match perfectly the moment it jumps from Track A to Track B, and then from B to C.

The authors realized that previous calculations were too sloppy at these "switch points." They missed some subtle terms in the math. By applying three strict rules to identify which mathematical terms matter most at every step, they built a much more accurate map of the journey.

The Results: A New Map of the Universe

Using their improved map, they derived simple formulas that describe exactly how the energy of the universe's ripples changes over time.

The Dip is Real and Finite: They proved mathematically that the dip isn't zero (as some thought) and isn't a glitch. It's a real, calculable feature caused by the "tug-of-war" between two growing modes.
The "Wiggles": Depending on when a ripple crosses the horizon (the edge of the observable universe), the final energy spectrum has "wiggles" or oscillations. It's like plucking a guitar string; the timing of when you pluck it determines the sound. These wiggles tell us exactly when the transition from normal to ultra-fast inflation happened.
Predictions for the Future: These formulas are simple enough to use but accurate enough to be tested. Future telescopes looking at the Cosmic Microwave Background (the afterglow of the Big Bang) might see these specific "dips" and "wiggles." If they do, it would be the smoking gun proof that the universe went through this "Ultra-Slow-Roll" phase.

Why Does This Matter?

This isn't just about abstract math. If the universe did this "sandwich" inflation:

It could explain the existence of Primordial Black Holes, which are candidates for Dark Matter.
It could generate Gravitational Waves (ripples in spacetime) that future detectors might hear.
It solves a mystery about why the energy spectrum has a dip, correcting a long-standing misunderstanding in the field.

In a nutshell: The authors fixed a broken map of the early universe. They showed that the "valley" in the energy graph isn't a mistake, but a signature of two growing forces canceling each other out. Their new, simple formulas allow scientists to predict exactly what we should see if we look closely enough at the oldest light in the universe.

1. Problem Statement

The paper addresses the theoretical understanding of linear curvature perturbations in a specific inflationary scenario: Slow-Roll (SR) $\to$ Ultra-Slow-Roll (USR) $\to$ Slow-Roll (SR) with instantaneous transitions.

Context: While standard SR inflation explains the scale-invariant spectrum observed in the Cosmic Microwave Background (CMB), it fails to generate the large enhancements in small-scale perturbations required for Primordial Black Hole (PBH) formation. USR inflation is the leading candidate for such enhancements.
The Gap: Previous analytical studies using the "junction method" often failed to accurately predict the dip structure (a local minimum) in the final power spectrum. Earlier literature suggested this dip arose from the cancellation between a constant mode and a negative growing mode. However, numerical simulations showed a finite (non-zero) dip, contradicting the leading-order analytical predictions which often yielded zero or incorrect scaling.
Objective: To derive accurate, systematic asymptotic expressions for the time evolution of mode functions and the resulting power spectrum across all $k$ -regimes, specifically resolving the origin of the finite dip and the oscillatory features ("wiggles") without relying on numerical simulations alone.

2. Methodology

The authors employ a rigorous analytical approach combining the Junction Method with Asymptotic Expansions of Hankel Functions.

Governing Equation: The mode function $\chi_k$ (representing curvature or field perturbations) obeys a differential equation with time-dependent Hubble friction:
$\ddot{\chi}_k + 3\tilde{H}(t)\dot{\chi}_k + \frac{k^2}{a^2}\chi_k = 0$
Under the quasi-de Sitter approximation, this maps to a Bessel equation solvable by Hankel functions $H^{(1,2)}_\nu(-k\tau)$ .
The Junction Method: The inflationary background is modeled as a sequence of phases with constant slow-roll parameters ( $\eta$ ). The transitions (SR $\to$ USR at $\tau_1$ , USR $\to$ SR at $\tau_2$ ) are treated as instantaneous. Continuity of the mode function and its derivative ( $\chi$ and $\chi'$ ) at the transition times allows for the recursive calculation of coefficients ( $A, B$ ) for Hankel functions in each phase.
Systematic Rules for Asymptotics: A core innovation of the paper is the proposal of three systematic rules to identify dominant terms during asymptotic expansions across transitions:
1. Identify dominant terms at each transition.
2. Identify dominant terms at later times.
3. Crucially: Rule 1 must be applied before Rule 2.
- Significance: Failure to follow Rule 3 leads to the incorrect neglect of sub-leading terms near transitions that become dominant later, causing errors in predicting the dip structure.
Expansion Strategy: The authors perform asymptotic expansions of Hankel functions for both superhorizon ( $-k\tau \ll 1$ ) and subhorizon ( $-k\tau \gg 1$ ) scales, retaining higher-order terms often discarded in previous literature to ensure accuracy.

3. Key Contributions

Resolution of the Dip Origin: The paper demonstrates that the finite dip in the power spectrum arises from the cancellation between two growing modes (specifically a positive $(\tau)^{-6}$ mode and a negative $(\tau)^{-3}$ mode) within the linear perturbation theory. This corrects the previous misconception that the dip resulted from a cancellation between a constant and a growing term.
Universal Asymptotic Formulas: The authors derive explicit, tractable analytical formulas for the power spectrum evolution across three distinct $k$ -regimes (Case 1, 2, and 3) defined by when the mode exits the horizon relative to the transitions.
Analytical Description of Features:
- Dip Structure: Provides an analytical formula for the dip time ( $\tau_c$ ) and dip value, showing it is finite and non-zero.
- $k^4$ Growth: Derives the conditions under which the spectrum exhibits a steep $k^4$ growth.
- Oscillations (Wiggles): Explains the origin of oscillatory features in the final spectrum, linking them to the phase of the mode at the first transition.

4. Key Results

The paper analyzes three cases based on the horizon crossing time of the mode $k$ :

Case 1 ( $k \ll 1/(-\tau_1)$ ): Horizon exits in the first SR phase.
- The power spectrum in the USR phase evolves as a sum of growing modes: $P \propto C_0 + C_3(-\tau)^{-3} + C_6(-\tau)^{-6}$ .
- The dip occurs when the negative $(-\tau)^{-3}$ term cancels the constant mode before the positive $(-\tau)^{-6}$ term takes over.
- Final Spectrum: For modes where the second transition occurs after the dip ( $-\tau_2 < -\tau_c$ ), the spectrum is suppressed. If the transition occurs before the dip, the spectrum is enhanced.
- Dip Scale: The dip position $k_{dip}$ scales as $k_{dip} \propto e^{-3N_{USR}/2}$ , and the dip value scales inversely with the peak amplitude ( $P_{dip} \propto P_{peak}^{-1/2}$ ), consistent with numerical results.
Case 2 ( $1 < -k\tau_1$ and $-k\tau_2 < 1$ ): Horizon exits during the USR phase.
- This regime yields the maximum enhancement.
- The final spectrum exhibits oscillations (wiggles) with frequency determined by the first transition time $\tau_1$ .
- Peak Enhancement: The peak value is approximately $7 e^{6N_{USR}}$ times the standard SR amplitude, roughly 7 times higher than the enhancement in Case 3.
Case 3 ( $-k\tau_1 \gg 1$ ): Horizon exits in the final SR phase.
- The mode remains subhorizon throughout the USR phase.
- The enhancement is scale-invariant ( $e^{6N_{USR}}$ ) but lacks the $k^4$ growth and dip structure seen in Case 1.

Specific Analytical Findings:

Dip Time: $\tau_c \approx - (2/5)^{1/3} k^{2/3} (-\tau_1)^{5/3}$ .
Dip Value: $P_{dip} \approx \frac{5}{2} P_{CMB} e^{-3N_{USR}}$ .
Peak Position: $k_{pk} \approx 4 k_1$ (where $k_1 = 1/(-\tau_1)$ ).
Scaling Relations: The paper confirms the relation $k_{dip} \propto P_{peak}^{-1/4}$ and $P_{dip} \propto P_{peak}^{-1/2}$ , validating previous numerical findings with analytical rigor.

5. Significance

Theoretical Validation: The work resolves a long-standing discrepancy between analytical junction methods and numerical simulations regarding the non-zero dip in USR inflation. It proves that linear perturbation theory is sufficient to explain these features if higher-order terms are handled systematically.
Predictive Power: The derived formulas are simple and tractable, allowing for rapid theoretical calculations without the need for heavy numerical integration.
Observational Constraints: The analytical expressions for the dip position, peak height, and oscillatory wiggles provide specific, testable signatures for future CMB observations and large-scale structure surveys. These features can be used to constrain the duration of the USR phase ( $N_{USR}$ ) and the timing of transitions.
Methodological Framework: The "three rules" for asymptotic expansion provide a robust framework for studying other non-attractor inflationary scenarios, including those with smooth transitions or multiple non-SR phases.

In conclusion, this paper provides a comprehensive analytical toolkit for understanding the dynamics of perturbations in SR-USR-SR inflation, correcting previous misconceptions about the dip structure and offering precise predictions for the signatures of ultra-slow-roll inflation.