Two-Nucleon Systems from Effective Field Theory

I. INTRODUCTION

• Low-energy nuclear physics is suited to effective field theory because nucleon energies lie below the complex hadron spectrum, enabling controlled expansions with error estimates.Chiral perturbation theory illustrates how unknown short-distance physics can be parametrized with few parameters and organized by power counting.Source: §I · p. 2
• Large S-wave scattering lengths make short-distance NN physics nonperturbative and finely tuned, exposing inconsistencies in Weinberg’s proposed power counting.At a given order, required counterterms can correspond to operators assigned higher order.Source: §I · p. 2
• The paper develops a different expansion, explains PDS subtraction and renormalization-group methods, and extends power counting to pion-inclusive NN scattering.It also provides subleading analytic spin-triplet amplitudes in the 3S1−3D1 channel.Source: §I · pp. 2–3

II. EFFECTIVE FIELD THEORY FOR NONRELATIVISTIC SCATTERING: A TOY EXAMPLE

• The toy theory replaces short-range interactions with contact operators in a derivative expansion and organizes its divergent diagrams so amplitudes expand in p/Λ.The construction connects power counting to subtraction and renormalization, with PDS suited to systems having large scattering lengths.Source: §II · p. 3
• The toy-model effective field theory generalizes beyond effective-range expansions because it can incorporate chiral symmetry and relativistic effects in realistic NN systems.Phenomenological NN potentials fit phase shifts accurately but lack systematic error estimates and are less suited to inelastic, three-nucleon, and many-body calculations.Source: §I · p. 2, §II · p. 3
• For nonperturbative coupling η∼1, expanding in p/Λ can remain accurate even when a Born expansion in η is inappropriate.The toy-model discussion assumes η∼1 and motivates the low-energy expansion as the relevant perturbative organization.Source: §II · pp. 3–4

A. The momentum expansion for a scattering length of natural size

• When |a|∼1/Λ and |r_n|∼1/Λ, the amplitude has a regular momentum expansion whose coefficients can be matched to low-energy scattering data.The first terms receive contributions from tree and successive bubble diagrams, with derivative operators entering at higher orders.Source: A. The momentum expansion for a… · pp. 4, 6
• The bubble-chain diagrams arise from local contact operators and are evaluated using dimensional regularization in the S-wave channel.The formulation uses on-shell tree-level amplitudes as internal vertices when summing bubble diagrams.Source: Figure 1 · p. 5, A. The momentum expansion for a… · p. 5
• In minimal subtraction, each C_2n∇^2n vertex counts as p^2n and each loop contributes one power of p, yielding a simple power counting.The couplings are independent of the subtraction point because the relevant integrals have no D=4 poles.Source: A. The momentum expansion for a… · pp. 5–6
• The leading effective-theory couplings are determined by comparing the calculated amplitude with the effective-range expansion and fitting to scattering data.The first two couplings are obtained from the corresponding low-energy coefficients.Source: A. The momentum expansion for a… · p. 6
• Even for η∼1 and a possible bound state below threshold, the calculation remains perturbative when the amplitude has no poles for |p|≲Λ.With an appropriate subtraction scheme, the Taylor expansion in p/Λ corresponds to a perturbative sum of Feynman graphs.Source: A. The momentum expansion for a… · p. 6

Large-scattering-length expansion

• Retaining ap to all orders replaces the ordinary momentum expansion when |a| ≫ 1/Λ, because the near-threshold interaction makes the naive series break down at p ≳ 1/|a|.The effective-theory expansion is organized in powers of p/Λ rather than expanding the amplitude in powers of p alone.Source: B. The momentum expansion for large… · p. 6

Subtraction-scheme limitation

• The MS bubble sum lacks a justified expansion parameter for |ap| > 1, making individual graphs grow even though their sum is small and leaving omitted contributions uncontrolled.The paper identifies this as a failure of minimal subtraction power counting rather than of dimensional regularization itself.Source: B. The momentum expansion for large… · p. 7

PDS power counting

• In the PDS scheme, C2n(µ) scales as 1/p^(n+1) for µ ∼ p, so C0 bubble diagrams generate the leading p^-1 amplitude while derivative insertions supply higher powers of p.The resulting coupling relations match the low-energy parameters a and rn, with the first contributions represented by local-operator graphs in Fig. 2.Source: B. The momentum expansion for large… · p. 8, Figure 2 · p. 9

Renormalization-group analysis

• RG equations follow from requiring µ-independence of the amplitude, while boundary conditions at µ = 0 set the couplings from scattering data and reproduce the solutions for C0(µ) and C2(µ).The corresponding β-functions can also be obtained from the one-loop graphs shown in Fig. 3.Source: B. The momentum expansion for large… · pp. 9–10

RG solutions

• For µ ≫ |1/a|, C2n(µ) ∝ µ^−(n+1), and quasi-fixed-point behavior makes the leading higher couplings depend primarily on lower-dimensional parameters such as a and r0.The leading O(µ^−3) part of C4 depends on r0, while its subleading O(µ^−2) part is proportional to r1.Source: B. The momentum expansion for large… · p. 11

Connection to dibaryon methods

• Resumming the leading C2n behavior produces an s-channel dibaryon description equivalent at µ = 0, while r1-dependent corrections enter at O(p2).The paper states that this dibaryon treatment is as good as, but no better than, the effective-field-theory calculation through O(p1).Source: B. The momentum expansion for large… · p. 11

Naive scaling

• Naive Luke-Manohar scaling identifies a contact interaction as irrelevant in D = 4 because it vanishes as v → 0, but that result applies to perturbative underlying theories.The rescaled variables include E = E/Mv^2, P = p/Mv, and ˆµ = µ/Mv.Source: C. Luke-Manohar velocity scaling · pp. 11–12

Velocity scaling with large scattering length

• For a large scattering length, the four-fermion interaction remains relevant near the UV fixed point reached as a → ∞, explaining its strong low-momentum effect.It becomes irrelevant again only for v ≲ 1/Ma, where perturbation theory is justified.Source: C. Luke-Manohar velocity scaling · p. 12

• The EFT goal is a systematic expansion for computing low-momentum NN phase shifts, estimating errors order by order, and incorporating relativistic and inelastic effects.The perturbative framework may also be useful for many-body systems where potential-model calculations are computationally difficult.Source: §III · p. 12
• Realistic NN scattering requires explicit pion fields, spin and isospin degrees of freedom, and relativistic corrections beyond the toy model.Source: §III · p. 12
• For p ≲ mπ, the expansion improves as the pion mass decreases, identifying a QCD-parameter region where the results can hold with high accuracy.The analysis does not take the pion mass so small that amπ is also small.Source: §III · p. 13
• The chiral expansion parameter for NN scattering is empirically about 30%, supporting fairly rapid convergence for momenta p ∼ mπ.Although ηπ = mπ/ΛNN = 0.46, the OPE Yukawa potential produces phase shifts no larger than approximately 0.25 radians at the physical value.Source: §III · p. 13

The chiral expansion

• The chiral EFT contains nucleons, pions, and contact interactions constrained by chiral symmetry, with two independent zero-derivative operators for the 1S0 and 3S1 channels.At the working order it also includes seven independent two-derivative contact interactions and quark-mass insertions.Source: A. The chiral expansion · pp. 13–14
• Large 1S0 and 3S1 scattering lengths require enhanced contact interactions, with C0 bubble diagrams producing the leading O(Q^-1) amplitude.At subleading O(Q^0), single pion exchange, C2, and D2 insertions are dressed to all orders by C0.Source: A. The chiral expansion · pp. 14, 16
• Single pion exchange enters at the same order as C2p2 or D2mπ^2 in the 1S0 and 3S1 channels and can therefore be treated perturbatively.The scaling changes at sufficiently high momentum because pion exchange and contact interactions become equally important.Source: A. The chiral expansion · pp. 14–15
• The power counting is expected to fail completely for momenta of order ΛNN, where one-pion exchange and a C0 insertion become equally important.Source: A. The chiral expansion · p. 15
• In P waves and higher partial waves, the leading amplitude is O(1) single-pion exchange, while two-pion exchange contributes at O(p) and C2 operators at O(p^2).The leading single-pion-exchange approximation is stated to account well for observed phase shifts.Source: A. The chiral expansion · pp. 15–16
• At O(Q^1), the expansion includes two-pion exchange, four-derivative operators, and related higher-order contributions, while exact OPE iteration without matching contact terms is inconsistent.Omitting the higher-derivative operators required for renormalization amounts to modeling short-distance physics.Source: A. The chiral expansion · p. 16

Subleading amplitude

• At O(Q0), the 1S0 amplitude includes five contributions from derivative and quark-mass operators, pion exchange, and their renormalized graphs.The subleading graphs are dressed by the leading C0 interaction and require PDS subtraction and counterterms.Source: B. The 1S0 channel to subleading order · pp. 17–18

Consistency

• The amplitude produces an S-matrix that is unitary up to the computed order, with S†S = 1 + O(Qk+2) when calculated through O(Qk).This order-by-order unitarity follows from the expansion of the amplitude and phase shift.Source: B. The 1S0 channel to subleading order · p. 18

Fit to scattering data

• C0(mπ) = −3.63 fm2, D2(mπ) ≡0, and C2(mπ) = 2.92 fm4 give scattering length and effective range consistent with experiment.About 43% of the effective range comes from C2(mπ).Source: B. The 1S0 channel to subleading order · pp. 18–19
• Fitting the phase shift to Nijmegen data for p ≤200 MeV gives excellent agreement at relatively large momenta.The alternative fit treats the relevant subleading contact parameters as free and is shown in Fig. 7.Source: B. The 1S0 channel to subleading order · p. 19

Validity range

• For p > 100 MeV, |A0/A−1| exceeds about 0.5, making neglected terms of order (A0/A−1)2 difficult to justify.Differences between the two fitting procedures are attributed to higher-order effects.Source: B. The 1S0 channel to subleading order · p. 19

Power counting

• Single potential pion exchange dominates 3D1 scattering at O(Q0), while operators connecting 3D1 and 3S1 contain two derivatives and are renormalized on the S-wave side.Pure 3D1 contact operators first contribute at O(Q4) and are neglected.Source: C. The 3S1 −3D1 channels to subleading… · p. 19

Parameter ambiguity

• At the working order, C0 and D2 are not independently determined because only their linear combination enters the phase shift.Higher-order N†NN†Nππ interactions proportional to D2 distinguish them.Source: C. The 3S1 −3D1 channels to subleading… · p. 19

Leading order

• At leading O(Q−1), only the 3S1 partial wave contributes, while A−1[02] = A−1[20] = A−1[22] = 0.The S–D mixing and D-wave observables therefore start at higher order.Source: C. The 3S1 −3D1 channels to subleading… · pp. 20–21

3S1 fit

• The 3S1 phase shift is fitted using the scattering length a3 = 5.423 ± 0.005 fm and deuteron binding energy ED = 2.224644 ± 0.000034 MeV.At this order, D2(mπ) can be set to zero while C0 and C2 are fitted to these observables.Source: C. The 3S1 −3D1 channels to subleading… · p. 22

Mixing predictions

• The mixing parameter ε1 and D-wave phase shift δ2 begin at O(Q0) and have no free parameters once C0 is fixed from δ0.Their predictions are compared with the Nijmegen partial-wave analysis in Fig. 8.Source: C. The 3S1 −3D1 channels to subleading… · pp. 20, 22

Higher partial waves

• An operator connecting angular momenta L and L′ requires at least L+L′ spatial derivatives, setting its power-counting order.S-wave-to-S-wave scattering is treated separately, while mixed and nonscalar channels follow derivative-based rules.Source: D. Higher partial waves · p. 22
• Pion exchange contributes at O(Q0), making it the leading contribution to all non-S-wave-to-S-wave scattering amplitudes.The figure caption identifies the 3S1−3D1 phase shifts and mixing parameter used to illustrate the channel analysis.Source: D. Higher partial waves · p. 22, Figure 8 · p. 23

Radiation pions

• Potential pion exchange contributes at O(Q0), whereas radiation pions associated with the pion pole enter at O(Q1).Radiation-pion effects arise from s-, t-, and u-channel exchanges.Source: E. Radiation pions and operator mixing · p. 23

Operator mixing

• Virtual s-channel radiation pions mix operators in different spin channels because intermediate NNπ states can rescatter in different isospin and angular-momentum states.The example in Fig. 9 uses an insertion of (N†σaN)2 to define Γ1.Source: E. Radiation pions and operator mixing · p. 23, Figure 9 · p. 24

Renormalization-group effects

• The leading radiation-pion beta-function contributions produce mixing between S-wave spin-singlet and spin-triplet operators.The contribution is obtained after evaluating the loop integral and including irreducible graphs and wavefunction renormalization.Source: E. Radiation pions and operator mixing · p. 24

F. Relativistic effects

• Relativistic corrections arise from higher-dimension operators inserted perturbatively into nucleon-interaction graphs.The crossed-circle denotes these insertions in Fig. 10.Source: F. Relativistic effects · p. 24, Figure 10 · p. 25
• A single insertion of N†∇4N contributes at O(Q1) and modifies the lowest-order energy relation with a p4/8M3 term.The corrected relation is E/2 = p2/2M + p4/8M3 + . . ..Source: F. Relativistic effects · pp. 24–25
• These corrections are expected to be small because they are suppressed by the nucleon mass rather than ΛNN.Source: F. Relativistic effects · p. 25

IV. OVERVIEW

• The expansion uses a subtraction scheme that tracks power-law divergences and renormalization-group analysis to identify operator contributions consistently.This addresses delicate short-distance cancellations associated with large scattering lengths.Source: §IV · p. 25
• NN scattering calculations through subleading O(Q0) show quite good low-energy agreement with experiment, including a parameter-free 3S1−3D1 mixing parameter ε1 calculation.The paper also treats higher partial waves, relativistic effects, and virtual radiative pions.Source: §IV · p. 25
• The paper extends power counting to higher partial waves, where pion exchange dominates in the Born approximation.Source: §IV · p. 25
• Extending the expansion above ΛNN and applying it to systems with three or more nucleons remain theoretical challenges.The techniques are also described as applicable to radiative capture and electromagnetic moments of the deuteron.Source: §IV · p. 25