Multi-channel phase space with Feynman-diagram-gauge amplitudes

This paper presents a multi-channel phase space generation method enhanced by Feynman-diagram-gauge amplitudes to accurately simulate challenging high-energy lepton collider processes in the SMEFT, specifically addressing lepton-mass singularities through specialized phase-space parametrization and modifications to the HELAS library for precise evaluation of vertices at very small invariant momentum squares.

The Gist

Imagine you are trying to predict the outcome of a massive, chaotic traffic jam involving thousands of cars, but instead of cars, you are dealing with subatomic particles colliding at nearly the speed of light. This is what particle physicists do when they simulate collisions at future super-colliders (like a giant version of the Large Hadron Collider, but much bigger).

This paper is essentially a new, ultra-precise GPS and traffic simulation tool designed to handle these collisions, especially when the cars (particles) are moving so fast that they almost crash into each other head-on or fly off in a straight line.

Here is the breakdown of the paper using simple analogies:

1. The Problem: The "Gauge" Confusion

In the world of particle physics, there are different ways to calculate how particles interact. Think of these ways as different languages or dialects.

• The Old Way (Standard Gauge): Imagine trying to describe a complex dance routine using a language where the dancers constantly cancel each other's moves out in the description. To get the final picture, you have to add up thousands of tiny, opposing numbers. When the numbers get huge (which happens at high energies), the tiny errors in your calculator add up, and the final result becomes garbage. It's like trying to measure the height of a mountain by subtracting two massive numbers that are almost identical; the result is just noise.
• The New Way (Feynman-Diagram Gauge): The authors use a new "dialect" (the Feynman-diagram gauge). In this language, the "cancellations" don't happen in the math. Instead, one specific move (one Feynman diagram) usually dominates the dance. This makes the calculation stable and clean, even at extreme energies.

2. The Solution: "Single-Diagram-Enhanced" Multi-Channel Integration

The paper introduces a method called SDE MCPS. Let's break that down with an analogy:

Imagine you are a tour guide trying to show tourists (computers) around a massive, complex city (the "Phase Space" of all possible particle outcomes).

• The Old Method: You dump the tourists into the city center and tell them, "Go explore everywhere randomly." Most of the time, they wander into empty parks (unlikely outcomes) and miss the famous landmarks (rare but important physics events). The computer wastes time calculating nothing.
• The New Method (SDE MCPS): You realize that the city has specific "highways" (Feynman diagrams) that lead to the most interesting spots. You create specialized tour buses for each highway.
  • Bus A takes people specifically to the "Photon Highway."
  • Bus B takes people to the "W-Boson Highway."
  • Each bus has a map perfectly tuned to that specific road.
  • Because the map matches the road, the bus drives smoothly, and you don't waste time driving through empty fields. You get a perfect picture of the city much faster.

3. The Special Challenge: The "Forward" Leptons

The paper focuses on three specific types of collisions that are notoriously difficult to simulate at very high energies (like 100 TeV).

• The Issue: In these collisions, some particles (charged leptons, like electrons or muons) get "kicked" so hard that they fly almost straight forward, hugging the beam pipe.
• The Metaphor: Imagine throwing a pebble at a wall. Usually, it bounces off at an angle. But if you throw it perfectly parallel to the wall, it skims the surface. In physics, when a particle skims the surface (a "collinear" split), the math breaks down because the numbers get infinitely large (singularities).
• The Fix: The authors built a specialized lens for their camera. They modified the math (the "phase-space parametrization") so that instead of trying to measure the angle directly (which is hard when it's zero), they measure the distance from the wall. This allows the computer to see the "skimming" particles clearly without the math crashing.

4. The "Helas" Library Upgrade

The team also upgraded the engine that does the actual math (called the Helas library).

• The Problem: When particles are very close together or moving very fast, standard computer math loses precision. It's like trying to measure the thickness of a human hair using a ruler marked only in centimeters. You lose the "effective digits" (the fine details).
• The Fix: They added a "fifth coordinate" to their measurements. Instead of just calculating the position and speed, they explicitly stored the "invariant mass" (a specific property of the particle) as a separate number. This prevents the computer from having to do a subtraction that wipes out the important details. It's like having a dedicated "microscope" setting for the hair-thin measurements.

5. What Did They Test?

They tested their new system on three complex scenarios involving Top Quarks (the heaviest known particles) and the Higgs Boson:

1. Lepton + Antilepton $\to$ Neutrinos + Top/Anti-Top + Higgs
2. Lepton + Antilepton $\to$ Lepton + Neutrino + Top + Anti-Bottom + Higgs
3. Lepton + Antilepton $\to$ Lepton + Antilepton + Top + Anti-Top + Higgs

They simulated these collisions at energies ranging from 1 TeV up to 100 TeV (energies far beyond what we have today).

The Bottom Line

The authors have built a super-stable, high-precision simulation engine that:

1. Uses a smarter mathematical language (Feynman-diagram gauge) to avoid calculation errors.
2. Uses "specialized buses" (multi-channel integration) to efficiently find the most important physics outcomes.
3. Has a "microscope" (modified Helas code) to see particles that are flying almost straight ahead.

Why does this matter?
Future colliders (like a Muon Collider) will operate at these extreme energies. To understand what we find there—perhaps discovering new physics or proving the Standard Model wrong—we need simulations that don't crash or give wrong answers. This paper provides the toolkit to ensure that when we finally build these machines, we can trust the data we get from them.

Technical Summary

1. Problem Statement

The paper addresses the challenge of performing accurate and efficient Monte Carlo event generation for high-energy lepton collider processes (specifically at multi-TeV energies, such as future muon colliders). Two primary difficulties are identified:

• Gauge Cancellations and Numerical Instability: In standard covariant gauges (e.g., Feynman gauge for photons/gluons, unitary gauge for weak bosons), scattering amplitudes involve delicate cancellations between interfering diagrams. At high energies, these cancellations lead to a loss of numerical precision (catastrophic cancellation), making it difficult to obtain stable cross-section calculations, especially when unphysical contributions dominate before cancellation.
• Collinear and Soft Singularities: Processes involving light charged leptons ($e, \mu$) and photons exhibit severe singularities in the collinear limit (forward emission) and when virtual photon propagators become nearly on-shell ($q^2 \approx 0$). Standard phase-space parametrizations often fail to sample these regions efficiently, leading to poor convergence. Furthermore, calculating amplitudes in these regions with standard libraries (like Helas) often results in a loss of significant digits due to the subtraction of large numbers to obtain small invariant masses ($m^2 \ll E^2$).

2. Methodology

The authors propose a unified framework combining Single-Diagram-Enhanced (SDE) Multi-Channel Phase Space (MCPS) integration with Feynman-Diagram (FD) Gauge amplitudes.

A. Feynman-Diagram (FD) Gauge Amplitudes

Instead of using standard covariant gauges, the authors utilize the FD gauge (introduced in previous works by the same group).

• Key Feature: In the FD gauge, the total scattering amplitude is dominated by individual diagrams (or sets of diagrams sharing a common propagator) in their respective singular kinematic regions.
• Benefit: This eliminates the need for delicate gauge cancellations between interfering amplitudes. The ratio of the squared sum of amplitudes to the sum of squared amplitudes ($R$) remains of order unity across the phase space, making the SDE integration strategy highly effective.

B. Modular Phase-Space Parametrization

The authors developed a modular library (written in Fortran 90) to generate phase space tailored to the topology of each Feynman diagram.

• Modular Units: The library uses specific subroutines (dshat, ph2s, ph2t, ph2c, etc.) to handle different propagator structures:
  • dshat: Generates invariant mass squared ($\hat{s}$) for subsystems, handling Breit-Wigner resonances and massless propagators.
  • ph2s: Handles $s$-channel splittings.
  • ph2t: Handles $t$-channel propagators. Crucially, it uses a logarithmic variable for the inverse propagator ($\ln t'$) to cancel the singular behavior of the propagator factor in the matrix element.
  • ph2c: Handles contact interactions (multi-point vertices).
• Jacobian Matching: For each diagram channel, the integration variables are chosen such that the Jacobian factor cancels the dominant singular propagator factor ($1/q^2$ or $1/(q^2-m^2)$) in the squared matrix element $|M_k|^2$.

C. Numerical Stability Improvements

To handle the extreme kinematic regimes (multi-TeV energies with light leptons), the authors implemented specific numerical fixes:

• Five-Vector Prescription: Momentum vectors are stored as 5-component arrays $p(0:4)$, where the 5th element explicitly stores the invariant mass squared $p^2$. This avoids the loss of precision that occurs when calculating $p^2 = (p^0)^2 - |\vec{p}|^2$ for nearly on-shell particles where $p^2 \ll (p^0)^2$.
• Stable Kinematic Calculations:
  • Modified the calculation of the minimum inverse propagator ($t'_{min}$) to avoid subtracting large numbers (using algebraic rearrangement similar to Eq. 31 in the text).
  • Modified the Helas library (the standard helicity amplitude generator) to use the explicit $p^2$ value and to calculate angles in collinear limits using stable functions (e.g., atan2 for small opening angles).
  • These modifications ensure that vertices with virtualities of order $m_l^2$ (lepton mass squared) are evaluated accurately even at multi-TeV energies.

3. Key Contributions

1. Integration Framework: A robust implementation of SDE MCPS integration specifically optimized for FD-gauge amplitudes, demonstrating that this combination yields stable and efficient integration for complex multi-particle final states.
2. Modular Library: Development of a Fortran 90 library (dshat, ph2s, ph2t, etc.) that automatically constructs phase-space parametrizations based on Feynman diagram topology.
3. Helas Modifications: Critical updates to the Helas library to handle singular vertices (nearly on-shell photons and collinear lepton pairs) without numerical precision loss.
4. Process Studies: Application of this framework to three challenging processes in the Standard Model (SM) and the Standard Model Effective Field Theory (SMEFT) with a complex top-Yukawa coupling:
   • $l\bar{l} \to \nu_l \bar{\nu}_l t\bar{t}H$
   • $l\bar{l} \to l \bar{\nu}_l t\bar{b}H$
   • $l\bar{l} \to l \bar{l} t\bar{t}H$

4. Results

The authors present numerical results for center-of-mass energies ($\sqrt{s}$) ranging from 1 TeV to 100 TeV.

• Efficiency and Stability: The method achieves stable Monte Carlo convergence for all processes, even at $\sqrt{s} = 100$ TeV, where standard methods often fail due to gauge cancellations or singularities.
• Cross-Section Behavior:
  • $l\bar{l} \to \nu_l \bar{\nu}_l t\bar{t}H$: The Vector Boson Fusion (VBF) contribution dominates at high energies. The FD gauge allows for the clear observation of destructive interference between VBF and $W-l$ scattering channels, which is obscured in unitary gauge calculations.
  • $l\bar{l} \to l \bar{\nu}_l t\bar{b}H$: The cross-section shows a strong dependence on the lepton mass ($\sigma_e > \sigma_\mu$) at high energies due to $t$-channel photon exchange. The new method accurately reproduces the forward-peaked lepton distributions without kinematic cuts.
  • $l\bar{l} \to l \bar{l} t\bar{t}H$: This process involves singularities from both $t$-channel photon exchange and $s$-channel photon splitting ($\gamma^* \to l\bar{l}$). The modified Helas codes successfully handle the collinear lepton-pair splitting.
• SMEFT Sensitivity: The framework successfully isolates contributions from dimension-5 and dimension-6 operators. For instance, in the $t\bar{t}H$ production, a specific dimension-6 contact interaction ($W^+W^-t\bar{t}H$) is shown to dominate the cross-section at 100 TeV in the CP-violating SMEFT scenario.
• Interference Patterns: The study reveals that destructive interference between different diagram categories (e.g., VBF vs. $W-l$ scattering) is localized in specific kinematic regions (e.g., small invariant masses of subsystems or specific rapidity ranges).

5. Significance

• Future Collider Physics: This work provides a critical tool for precision phenomenology at future multi-TeV lepton colliders (e.g., Muon Colliders). It enables the generation of fully exclusive events with exact tree-level matrix elements in kinematic regimes where standard tools fail.
• SMEFT Analysis: The ability to accurately calculate processes with hundreds of diagrams and complex interference patterns allows for precise constraints on SMEFT operators, particularly those involving the top-quark Yukawa coupling and CP-violation.
• Methodological Advance: The combination of FD-gauge amplitudes with diagram-specific phase-space mapping and numerically stable Helas routines sets a new standard for handling singularities and gauge cancellations in high-energy physics event generation. It demonstrates that "gauge cancellation" issues can be circumvented by choosing a gauge where individual diagrams dominate, simplifying the numerical integration landscape.