Modular Nuclearity and Wedge-Local Constructions

Wedge-local generators do not by themselves produce compactly localized observables. For a Borchers triple, modular nuclearity applies the quarter-power of the wedge modular operator to a translated wedge algebra. If that map is nuclear, the translated inclusion is split; its relative commutant is then large enough to supply nontrivial double-cone algebras. In regular two-dimensional factorizing models this closes the gap between an input scattering function and a local net. The sinh-Gordon application is returned to the exact S-matrix bootstrap, CDD freedom, and completeness.

Required background. Factorizing S-matrices, wedge-local fields, and Borchers constructions supplies the triple; modular theory, nuclearity, and the split property supplies the implication chain; standard von Neumann algebras and Tomita–Takesaki theory supplies $\Delta$; and phase-space nuclearity and compactness maps supplies the nuclear-map definition.

Helpful background. Elasticity and factorization hypotheses fixes the scattering input, while the integrable-QFT casebook records model-specific limits.

Borchers triples and translated wedge inclusions

In two-dimensional Minkowski space, a Borchers triple $(\mathcal M,U,\Omega)$ consists of:

• a von Neumann algebra $\mathcal M$ interpreted as the right-wedge algebra;
• a positive-energy, strongly continuous translation representation $U$ with $U(x)\Omega=\Omega$;
• a standard vector $\Omega$ for $\mathcal M$;
• the inclusion $\operatorname{Ad}U(a)(\mathcal M)\subset\mathcal M$ for translations $a$ taking the right wedge into itself.

For such an $a$, set

$$\mathcal N_a=U(a)\mathcal M U(a)^* \subset\mathcal M.$$

The relative commutant $\mathcal N_a'\cap\mathcal M$ is associated with the bounded double cone between the two wedge boundaries. Wedge locality only proves the inclusion and commutation relations. It does not prove that this intersection contains anything beyond scalar multiples of the identity.

Let $\Delta$ be the modular operator of $(\mathcal M,\Omega)$. The modular nuclearity map for the inclusion is

$$\Xi_a:\mathcal M\longrightarrow\mathcal H, \qquad \Xi_a(A)=\Delta^{1/4}U(a)A\Omega.$$

Because $U(a)\Omega=\Omega$, this is the usual inclusion map $N\mapsto\Delta^{1/4}N\Omega$ written on the untranslated algebra. The domain is the Banach space $\mathcal M$ with operator norm. The unbounded operator $\Delta^{1/4}$ is applied only to vectors in the stated range; proving that the range lies in its domain is part of the estimate.

If $\Xi_a$ is nuclear with the required standardness, then $\mathcal N_a\subset\mathcal M$ is split. Consequently the relative commutant is nontrivial, $\Omega$ is cyclic for the resulting local algebra under the theorem’s hypotheses, and translated intersections define a local net. Buchholz and Lechner establish this modular-nuclearity route and its localization consequences in Buchholz and Lechner 2004, pp. 1065–1080.

Why the quarter modular power helps

For a wedge, Bisognano–Wichmann modular flow is a boost. The analytic continuation to $\Delta^{1/4}$ shifts rapidities by an imaginary amount. In factorizing models, an $n$-particle component of $\Xi_a(A)$ can therefore be represented by an analytic rapidity wave function evaluated inside a strip. The translation contributes exponential damping depending on the wedge separation, while the scattering function controls exchange relations and analytic continuation.

The proof separates the map into Hardy-space evaluation operators and trace-class one-particle factors. Summing the resulting $n$-particle nuclear norms is the hard step. Regularity of the scattering function supplies a strip wider than the physical strip, so contour shifts remain away from singularities. For some classes the Pauli-type sign $S_2(0)=-1$ improves the sum enough to handle every positive splitting distance.

This mechanism is model-specific. A formal analytic continuation of form factors, without trace-norm summability and a common strip, is not a modular nuclearity proof.

Regular sinh-Gordon scattering function

The scalar sinh-Gordon scattering function is

$$S_2(\theta) =\frac{\sinh\theta-i\sin b} {\sinh\theta+i\sin b}, \qquad 0<b<\pi.$$

It is unitary and crossing symmetric on the real rapidity axis, has no bound-state pole in the physical strip, extends boundedly to a slightly larger strip, and satisfies $S_2(0)=-1$. It therefore belongs to the regular class to which Lechner’s nuclearity estimates apply.

For every positive wedge splitting distance $s$, the map

$$\Xi_s(A)=\Delta^{1/4}U(s)A\Omega$$

is nuclear in the $S_2(0)=-1$ class. Hence the opposite translated wedge intersection contains nontrivial observables localized in the corresponding double cone, including arbitrarily small positive double cones. The large-distance theorem for general regular scattering functions and the all-distance improvement are Lechner 2008, § 5, Theorems 5.6 and 5.8, pp. 27–30 of the open manuscript.

The construction also has the prescribed factorizing scattering operator and is asymptotically complete for the constructed regular models, by a separate scattering argument. It does not prove that a path integral for the classical sinh-Gordon Lagrangian has been constructed or that every CDD-modified amplitude defines the same theory.

An independent check verifies the input identities:

$$S_2(\theta)S_2(-\theta)=1, \qquad S_2(i\pi-\theta)=S_2(\theta), \qquad S_2(0)=-1.$$

The first follows by exchanging numerator and denominator under $\theta\mapsto-\theta$; the second uses $\sinh(i\pi-\theta)=\sinh\theta$.

Adversarial test: wedge locality without nuclearity

Construct wedge-local polarization-free generators and verify the Borchers-triple relations, but omit the nuclearity estimate for $\Xi_a$. The wedge algebra, translations, vacuum, and factorizing exchange relations still exist. The relative commutant $\mathcal N_a'\cap\mathcal M$ could nevertheless be trivial.

Therefore wedge locality alone licenses a wedge-local theory, not compact localization, Reeh–Schlieder for double cones, or the split property. A candidate scattering function with singularities too close to the contour may defeat the analytic/nuclear estimate even when its on-shell unitarity and crossing identities hold.

Exercises

Verify unitarity and crossing for the sinh-Gordon scattering function.

Solution

Since $\sinh(-\theta)=-\sinh\theta$,

$$S_2(-\theta) =\frac{\sinh\theta+i\sin b}{\sinh\theta-i\sin b} =S_2(\theta)^{-1}.$$

Also $\sinh(i\pi-\theta)=\sinh\theta$, so $S_2(i\pi-\theta)=S_2(\theta)$. At $\theta=0$, the ratio is $(-i\sin b)/(i\sin b)=-1$. These identities do not prove modular nuclearity; bounded strip analyticity and trace-norm estimates are still required.

References

• Buchholz, Detlev, and Gandalf Lechner. 2004. “Modular Nuclearity and Localization.” Annales Henri Poincaré 5: 1065–1080. DOI. Open manuscript.
• Lechner, Gandalf. 2008. “Construction of Quantum Field Theories with Factorizing S-Matrices.” Communications in Mathematical Physics 277: 821–860. DOI. Open manuscript.