[Paper Review] Comment on 'Spin Ice: Magnetic Excitations without Monopole Signature Using $μ$SR' (arXiv:1110.0877)

S. T. Bramwell, S. R. Giblin|arXiv (Cornell University)|Nov 17, 2011

Advanced Condensed Matter Physics3 citations

TL;DR

This paper defends the interpretation of muon spin relaxation (μSR) data in dysprosium titanate spin ice as evidence for magnetic monopole-like excitations, arguing that the observed depolarization rate λ scales with monopole density x via ν ∝ x, consistent with a Coulomb gas model. The authors refute claims that their experiment is flawed or misinterpreted, asserting that the signal originates from interior muons and is reproducible, with corrections for external muons improving agreement with the Wien effect theory.

ABSTRACT

Dunsiger and co-workers have criticised our paper on magnetic monopole dynamics spin ice [S. T. Bramwell et al., Nature, 461 956 (2009)]. We consider their criticisms and results but show that these do not warrant any revision of our earlier conclusions. In contrast, we point out that their conclusion of a `new' type of spin dynamics in spin ice is inconsistent with established facts.

Motivation & Objective

• To refute claims by Dunsiger et al. that the μSR experiment in Dy2Ti2O7 is flawed in design, execution, and theoretical interpretation.
• To demonstrate that the observed muon spin depolarization rate λ is consistent with a monopole Coulomb gas model, with λ ∝ x (monopole density) and ν ∝ x via dynamical scaling (z = 2).
• To show that the signal arises from interior muons near the sample, not exterior ones, and that corrections for distant muons improve agreement with the Wien effect theory.
• To confirm the reproducibility of the original experiment and counter claims that the data collapse is unverifiable or misattributed.
• To argue that alternative explanations—such as random crystal fields or persistent spin dynamics—are unlikely and inconsistent with existing ac-susceptibility and specific heat data.

Proposed method

• Analysis of μSR data from Dy2Ti2O7, distinguishing between a dominant fast-dephasing component (δB ≫ B₀) and a minority component (δB ≪ B₀) with λ(B₀, T) that collapses to reveal monopole charge Q.
• Application of dynamical scaling with z = 2, leading to ν ∝ l_D^(-z) ∝ x, where l_D is the Debye length and ν is the magnetic relaxation rate.
• Use of the relation λ = Δ₀(ν/ν₀)^y with y ≈ 1, showing λ ∝ ν ∝ x, and arguing that dimensional analysis supports this scaling despite D et al.'s objections.
• Separation of near-field (interior) and far-field (exterior) muon contributions using experimental geometry and sample mounting techniques.
• Correction of data for muons beyond 0.25 mm from the sample, which reduces anomalous Q rise at T > 0.3 K and improves agreement with Wien effect theory.
• Reproducibility checks via sample rotation (implanting into Ag backing) and use of non-spin ice materials, both of which eliminate the signal, confirming its origin in spin ice.

Experimental results

Research questions

• RQ1 Is the observed μSR depolarization rate λ in Dy2Ti2O7 consistent with a monopole Coulomb gas model, and does it scale with monopole density x?
• RQ2 Can the minority component of μSR data, with λ ∝ x, be reliably interpreted as evidence for magnetic monopoles, despite claims of theoretical flaws?
• RQ3 Do exterior muons in the silver backing plate dominate the signal, or is the signal primarily from interior muons near the sample?
• RQ4 Is the observed data collapse and monopole charge inference reproducible, and do alternative sample mounts or non-spin ice materials eliminate the signal?
• RQ5 Do alternative explanations—such as random crystal fields or persistent spin dynamics—provide a better fit than the monopole interpretation, especially given consistency with ac-susceptibility and specific heat?

Key findings

• The μSR depolarization rate λ scales with monopole density x via λ ∝ ν ∝ x, consistent with a Coulomb gas model and dynamical scaling with z = 2.
• Corrections for muons beyond 0.25 mm from the sample eliminate the anomalous rise in monopole charge Q at T > 0.3 K, improving agreement with the Wien effect theory.
• The signal disappears when muons are implanted directly into the silver backing plate or when non-spin ice samples are used, confirming the signal originates from the spin ice material.
• The experiment is reproducible: another group confirmed the result in a different spin ice system (Lees et al., 2011), and the authors have reproduced the signal under controlled conditions.
• The observed μSR relaxation rate in weak longitudinal field closely tracks ac-susceptibility data, implying a significant monopole contribution, contrary to D et al.'s claim of no monopole signal.
• Dunsiger et al.'s claims of systematic and statistical errors in the original experiment are invalid, as their own measurements suffer from poor thermal contact, strong damping, and low statistics, which would dominate any signal.