Part VI, Chapter 7

Effective Field Theory

The modern framework: Integrating out heavy fields

The EFT Philosophy

Effective Field Theory (EFT) is the modern way to think about all of physics. The key insight: at low energies E << M, you don't need to know the details of heavy physics at scale M - its effects are captured by a few parameters!

This explains why QFT works despite our ignorance of Planck-scale physics, and provides a systematic way to organize corrections from unknown high-energy physics.

7.1 The EFT Paradigm

The central idea: Separate physics into energy scales and write an effective Lagrangianappropriate for each scale:

$$\mathcal{L}_{\text{eff}} = \mathcal{L}_{\text{light}} + \sum_{d>4} \frac{C_i^{(d)}}{\Lambda^{d-4}} \mathcal{O}_i^{(d)}$$

where:

ℒ_light: renormalizable Lagrangian of light fields
𝒪_i^(d): operators of mass dimension d > 4 (irrelevant operators)
C_i: Wilson coefficients encoding heavy physics effects
Λ: cutoff scale (mass of heavy particles)

💡Why EFT works: Decoupling

At energies E << Λ, heavy particles can't be produced (not enough energy). They only contribute virtually in loops. These virtual effects are suppressed by powers of E/Λ.

We systematically expand in E/Λ. The leading term (d=4, renormalizable) captures most physics. Higher-dimension operators (d=5,6,...) are increasingly suppressed corrections.

7.2 Integrating Out Heavy Fields

Integrating out means doing the path integral over heavy field degrees of freedom to obtain an effective action for light fields:

$$e^{iS_{\text{eff}}[\phi_L]} = \int \mathcal{D}\phi_H \, e^{iS[\phi_L, \phi_H]}$$

where φ_L are light fields (E << M) and φ_H are heavy fields (mass ~ M).

Simple Example: Heavy Scalar

Consider a theory with light field ϕ (mass m) and heavy field Φ (mass M >> m) coupled by:

$$\mathcal{L} = \frac{1}{2}(\partial\phi)^2 - \frac{m^2}{2}\phi^2 + \frac{1}{2}(\partial\Phi)^2 - \frac{M^2}{2}\Phi^2 - \frac{\lambda}{2}\phi^2\Phi$$

Solving for Φ at tree level: (□ + M²)Φ = -λφ²/2, giving:

$$\Phi \approx -\frac{\lambda}{2M^2}\phi^2 + O(1/M^4)$$

Substituting back into the Lagrangian:

$$\mathcal{L}_{\text{eff}} = \frac{1}{2}(\partial\phi)^2 - \frac{m^2}{2}\phi^2 + \frac{\lambda^2}{8M^2}\phi^4 + O(1/M^4)$$

The heavy field Φ generates an effective φ⁴ interaction suppressed by 1/M²!

7.3 Classic Example: Fermi Theory of Weak Interactions

Before the electroweak theory, Fermi described β-decay with a 4-fermion interaction:

$$\mathcal{L}_{\text{Fermi}} = \frac{G_F}{\sqrt{2}}(\bar{\nu}\gamma^\mu(1-\gamma^5)e)(\bar{u}\gamma_\mu(1-\gamma^5)d)$$

where G_F ≈ 1.166 × 10⁻⁵ GeV⁻² is the Fermi constant. This is a dimension-6 operator(non-renormalizable)!

Resolution: W Boson Exchange

Fermi theory is the low-energy effective theory of electroweak interactions. At high energies E ~ M_W, we resolve the W boson propagator:

$$\frac{-ig_w^2}{q^2 - M_W^2} \xrightarrow{q^2 \ll M_W^2} \frac{g_w^2}{M_W^2} \equiv \frac{G_F}{\sqrt{2}}$$

The Fermi constant is related to the W mass and weak coupling:

$$G_F = \frac{g_w^2}{4\sqrt{2}M_W^2} \approx \frac{1}{2v^2}$$

where v ≈ 246 GeV is the Higgs VEV. Fermi theory works perfectly for E << M_W ≈ 80 GeV!

7.4 Matching: Connecting Theories at Different Scales

Matching is the procedure to determine Wilson coefficients C_i by requiring that physical observables are the same in both the full and effective theories at the matching scale μ ~ Λ:

Matching Condition:

$$\langle \mathcal{O} \rangle_{\text{full theory}}(\mu) = \langle \mathcal{O} \rangle_{\text{EFT}}(\mu)$$

Calculate the same physical process in both theories and equate them at μ ~ M (the heavy mass scale).

Example: QCD → Chiral Perturbation Theory

At energies E << 1 GeV (below Λ_QCD), quarks and gluons confine into hadrons. The effective degrees of freedom are pions (Goldstone bosons).

Chiral Perturbation Theory (χPT) is the EFT of QCD at low energies:

$$\mathcal{L}_{\chi PT} = \frac{f_\pi^2}{4}\text{Tr}[\partial_\mu U^\dagger \partial^\mu U] + \cdots$$

where U = exp(iπ^a t^a/f_π) contains pion fields. The cutoff is Λ_χPT ~ 1 GeV. Corrections are organized in powers of p/Λ_χPT where p is the pion momentum.

7.5 Power Counting and Organization

In an EFT, we systematically expand in the small parameter ε = E/Λ. Operators are organized by their mass dimension:

Dimension 4 (Renormalizable)

No suppression, dominant at all energies. Examples: (∂φ)², m²φ², λφ⁴, ψ̄iγ·∂ψ, F²

Dimension 5

Suppressed by ~ 1/Λ. Example: (1/Λ)ℓℓHH (neutrino mass after Higgs VEV)

Dimension 6

Suppressed by ~ 1/Λ². Examples: 4-fermion ops, (1/Λ²)(D_μH)†(D^μH)|H|²

Dimension 8+

Suppressed by ~ 1/Λ⁴ or more. Negligible at low energies.

Predictivity of EFT:

At energy E, corrections from dimension-d operators are ~ (E/Λ)^(d-4). For E/Λ = 0.1, dim-6 ops contribute ~ 1%, dim-8 ~ 0.01%. By keeping only a finite number of operators, we makecontrolled predictions with calculable uncertainties!

7.6 The Standard Model EFT (SMEFT)

The most important modern application: treating the Standard Model as an EFT valid below some scale Λ_BSM (beyond the SM). The effective Lagrangian is:

$$\mathcal{L}_{\text{SMEFT}} = \mathcal{L}_{\text{SM}} + \sum_{i} \frac{C_i^{(5)}}{\Lambda}\mathcal{O}_i^{(5)} + \sum_{i} \frac{C_i^{(6)}}{\Lambda^2}\mathcal{O}_i^{(6)} + \cdots$$

Dimension-5: Weinberg Operator (Neutrino Mass)

The only dim-5 operator allowed by SM gauge symmetry:

$$\mathcal{O}_5 = \frac{1}{\Lambda}(\overline{L^c}\tilde{H})(\tilde{H}^T L)$$

After H gets VEV v ≈ 246 GeV, this gives Majorana neutrino mass m_ν ~ v²/Λ. For m_ν ~ 0.1 eV, we need Λ ~ 10¹⁴ GeV (seesaw scale)!

Dimension-6 Operators

There are 59 independent dim-6 operators (Warsaw basis). Examples:

4-fermion: (q̄γ_μq)(ℓ̄γ^μℓ) - affects precision electroweak tests
Triple gauge: H†D_μH W^μν B^μν - anomalous gauge couplings
Higgs: (H†H)³ - affects Higgs self-coupling

LHC searches for SMEFT operators constrain Λ > few TeV for most operators.

7.7 Decoupling Theorem

Appelquist-Carazzone decoupling theorem: Heavy particles with mass M decouple from low-energy physics (E << M). Their effects are suppressed by powers of E/M.

Statement of Decoupling:

In a renormalizable theory, the only remnants of heavy particles at E << M are:

Threshold corrections to light-field parameters (one-time shift)
Non-renormalizable operators suppressed by powers of 1/M

Exception: Massless particles (like photons, gluons) never decouple - they're always relevant at all energy scales!

7.8 EFT Philosophy and Naturalness

The EFT Worldview:

Separation of scales: Physics at scale E only depends on degrees of freedom with mass ≲ E. Heavier stuff decouples.
Ignorance is OK: We don't need a "final theory" to make predictions. Low-energy physics is largely independent of UV details.
Systematic expansion: Corrections from unknown high-energy physics are organized in powers of E/Λ. This is calculable and predictive!
All theories are effective: Even the Standard Model is an EFT valid below some scale (Planck scale, GUT scale, or new physics scale).

Naturalness and Hierarchy Problems:

The Higgs mass hierarchy problem: quantum corrections drive m_H² ~ Λ². For Λ ~ M_Planck, we'd expect m_H ~ 10¹⁹ GeV, but we measure 125 GeV!

This suggests either: (1) New physics at Λ ~ TeV that protects Higgs mass (SUSY, composite Higgs), (2) Fine-tuning, or (3) Anthropic selection. Still an open question!

7.9 More EFT Examples

Heavy Quark Effective Theory (HQET)

For b/c quarks in hadrons: m_Q >> Λ_QCD. Expand in 1/m_Q. Simplifies B-meson decays, CKM matrix elements. Crucial for flavor physics.

Soft-Collinear Effective Theory (SCET)

For jet physics: separate hard, collinear, and soft modes. Resums large logs systematically. Used for LHC jet cross sections.

Non-Relativistic QED (NRQED)

For bound states like hydrogen: v << c. Systematic expansion in v/c explains fine/hyperfine structure. Predicts Lamb shift, g-2 corrections.

🎯 Key Takeaways

EFT: Systematic expansion in E/Λ for physics below cutoff Λ
Integrating out heavy fields generates higher-dimension operators
ℒ_eff = ℒ_ren + Σ (C_i/Λ^(d-4)) 𝒪_i^(d) (d > 4)
Fermi theory is EFT of electroweak theory with Λ = M_W
Matching: connect full and effective theories at μ ~ Λ
Power counting: dim-d ops suppressed by (E/Λ)^(d-4)
SMEFT: SM + dim-5,6,... operators parameterize BSM physics
Decoupling theorem: heavy particles decouple at low energies
All theories are effective - valid below some cutoff
EFT philosophy: ignorance of UV is OK, make controlled predictions!

🎓 Congratulations!

You've completed Part VI: Renormalization Theory! You now understand:

How UV divergences arise in loop diagrams
Regularization methods (dimensional regularization, MS-bar)
Counterterms and renormalization of parameters
Renormalization conditions and schemes
The renormalization group and beta functions
Running couplings in QED and QCD (asymptotic freedom!)
Effective field theory as the modern framework

These are the most profound ideas in quantum field theory. With this understanding, you're ready for advanced topics and modern research!

← Running Couplings Continue to Part VII →