arXiv subjects

Hendrik Grundling

Publications and source records attributed to Hendrik Grundling.

At least 19 recordsLinked to original sources

arXiv2020

Crossed products of C^*-algebras for singular actions with spectrum conditions

We analyze existence of crossed product constructions of Lie group actions on C^*-algebras which are singular. These are actions where the group need not be locally compact, or the action need not be strongly continuous. In particular, we consider the case where spectrum conditions are required for the implementing unitary group in covariant representations of such actions. The existence of a crossed product construction is guaranteed by the existence of "cross representations". For one-parameter automorphism groups, we prove that the existence of cross representations is stable with respect to a large set of perturbations of the action, and we fully analyze the structure of cross representations of inner actions on von Neumann algebras. For one-parameter automorphism groups we study the cross property for covariant representations, where the generator of the implementing unitary group is positive. In particular, we find that if the Borchers-Arveson minimal implementing group is cross, then so are all other implementing groups. For higher dimensional Lie group actions, we consider a class of spectral conditions which include the ones occurring in physics, and is sensible also for non-abelian or for infinite dimensional Lie groups. We prove that the cross property of a covariant representation is fully determined by the cross property of a certain one-parameter subsystem. This greatly simplifies the analysis of the existence of cross representations, and it allows us to prove the cross property for several examples of interest to physics. We also consider non-abelian extensions of the Borchers-Arveson theorem. There is a full extension in the presence of a cyclic invariant vector, but otherwise one needs to determine the vanishing of lifting obstructions.

math.OA↗

arXiv2020

Covariant representations for possibly singular actions on C*-algebras

Singular actions on C*-algebras are automorphic group actions on C*-algebras, where the group need not be locally compact, or the action need not be strongly continuous. We study the covariant representation theory of such actions. In the usual case of strongly continuous actions of locally compact groups on C*-algebras, this is done via crossed products, but this approach is not available for singular C*-actions (this was our path in a previous paper). The literature regarding covariant representations for singular actions is already large and scattered, and in need of some consolidation. We collect in this survey a range of results in this field, mostly known. We improve some proofs and elucidate some interconnections. These include existence theorems by Borchers and Halpern, Arveson spectra, the Borchers-Arveson theorem, standard representations and Stinespring dilations as well as ground states, KMS states and ergodic states and the spatial structure of their GNS representations.

math.OA↗

arXiv2017

Full regularity for a C*-algebra of the Canonical Commutation Relations. (Erratum added)

The Weyl algebra,- the usual C*-algebra employed to model the canonical commutation relations (CCRs), has a well-known defect in that it has a large number of representations which are not regular and these cannot model physical fields. Here, we construct explicitly a C*-algebra which can reproduce the CCRs of a countably dimensional symplectic space (S,B) and such that its representation set is exactly the full set of regular representations of the CCRs. This construction uses Blackadar's version of infinite tensor products of nonunital C*-algebras, and it produces a "host algebra" (i.e. a generalised group algebra, explained below) for the σ-representation theory of the abelian group S where σ(.,.):=e^{iB(.,.)/2}. As an easy application, it then follows that for every regular representation of the Weyl algebra of (S,B) on a separable Hilbert space, there is a direct integral decomposition of it into irreducible regular representations (a known result). An Erratum for this paper is added at the end.

math.OA↗

arXiv2016

Dynamics for QCD on an infinite lattice

We prove the existence of the dynamics automorphism group for Hamiltonian QCD on an infinite lattice in R^3, and this is done in a C*-algebraic context. The existence of ground states is also obtained. Starting with the finite lattice model for Hamiltonian QCD developed by Kijowski & Rudolph, we state its field algebra and a natural representation. We then generalize this representation to the infinite lattice, and construct a Hilbert space which has represented on it all the local algebras (i.e. kinematics algebras associated with finite connected sublattices) equipped with the correct graded commutation relations. On a suitably large C*-algebra acting on this Hilbert space, and containing all the local algebras, we prove that there is a one parameter automorphism group, which is the pointwise norm limit of the local time evolutions along a sequence of finite sublattices, increasing to the full lattice. This is our global time evolution. We then take as our field algebra the C*-algebra generated by all the orbits of the local algebras w.r.t. the global time evolution. Thus the time evolution creates the field algebra. The time evolution is strongly continuous on this choice of field algebra, though not on the original larger C*-algebra. We define the gauge transformations, explain how to enforce the Gauss law constraint, show that the dynamics automorphism group descends to the algebra of physical observables and prove that gauge invariant ground states exist.

math-ph↗

arXiv2013

Quantum Systems and Resolvent Algebras

This survey article is concerned with the modeling of the kinematical structure of quantum systems in an algebraic framework which eliminates certain conceptual and computational difficulties of the conventional approaches. Relying on the Heisenberg picture it is based on the resolvents of the basic canonically conjugate operators and covers finite and infinite quantum systems. The resulting C*-algebras, the resolvent algebras, have many desirable properties. On one hand they encode specific information about the dimension of the respective quantum system and have the mathematically comfortable feature of being nuclear, and for finite dimensional systems they are even postliminal. This comes along with a surprisingly simple structure of their representations. On the other hand, they are a convenient framework for the study of interacting as well as constrained quantum systems since they allow the direct application of C*-algebraic methods which often simplify the analysis. Some pertinent facts are illustrated by instructive examples.

math-ph↗

arXiv2012

Crossed products of C*-algebras for singular actions

We consider group actions of topological groups on C*-algebras of the types which occur in many physics models. These are singular actions in the sense that they need not be strongly continuous, or the group need not be locally compact. We develop a "crossed product host" in analogy to the usual crossed product for strongly continuous actions of locally compact groups, in the sense that its representation theory is in a natural bijection with the covariant representation theory of the action. We prove a uniqueness theorem for crossed product hosts, and analyze existence conditions. We also present a number of examples where a crossed product host exists, but the usual crossed product does not. For actions where a crossed product host does not exist, we obtain a "maximal" invariant subalgebra for which a crossed product host exists. We further study the case of a discontinuous action of a locally compact group in detail.

math.OA↗

arXiv2012

QCD on an infinite lattice

We construct a mathematically well--defined framework for the kinematics of Hamiltonian QCD on an infinite lattice in $\R^3$, and it is done in a C*-algebraic context. This is based on the finite lattice model for Hamiltonian QCD developed by Kijowski, Rudolph e.a.. To extend this model to an infinite lattice, we need to take an infinite tensor product of nonunital C*-algebras, which is a nonstandard situation. We use a recent construction for such situations, developed by Grundling and Neeb. Once the field C*-algebra is constructed for the fermions and gauge bosons, we define local and global gauge transformations, and identify the Gauss law constraint. The full field algebra is the crossed product of the previous one with the local gauge transformations. The rest of the paper is concerned with enforcing the Gauss law constraint to obtain the C*-algebra of quantum observables. For this, we use the method of enforcing quantum constraints developed by Grundling and Hurst. In particular, the natural inductive limit structure of the field algebra is a central component of the analysis, and the constraint system defined by the Gauss law constraint is a system of local constraints in the sense of Grundling and Lledo. Using the techniques developed in that area, we solve the full constraint system by first solving the finite (local) systems and then combining the results appropriately. We do not consider dynamics.

math-ph↗

arXiv2012

Lie Algebras of Derivations and Resolvent Algebras

This paper analyzes the action δ of a Lie algebra X by derivations on a C*-algebra A. This action satisfies an "almost inner" property which ensures affiliation of the generators of the derivations δ with A, and is expressed in terms of corresponding pseudo-resolvents. In particular, for an abelian Lie algebra X acting on a primitive C*-algebra A, it is shown that there is a central extension of X which determines algebraic relations of the underlying pseudo- resolvents. If the Lie action δ is ergodic, i.e. the only elements of A on which all the derivations in δ_x vanish are multiples of the identity, then this extension is given by a (non-degenerate) symplectic form σ on X. Moreover, the algebra generated by the pseudo-resolvents coincides with the resolvent algebra based on the symplectic space (X, σ). Thus the resolvent algebra of the canonical commutation relations, which was recently introduced in physically motivated analyses of quantum systems, appears also naturally in the representation theory of Lie algebras of derivations acting on C*-algebras.

math-ph↗

arXiv2010

Infinite Tensor Products of C_0(R): Towards a Group Algebra for R^\infty

The construction of an infinite tensor product of the C*-algebra C_0(R) is not obvious, because it is nonunital, and it has no nonzero projection. Based on a choice of an approximate identity, we construct here an infinite tensor product of C_0(R), denoted L_V. We use this to construct (partial) group algebras for the full continuous unitary representation theory of the group R^(N) = the infinite sequences with real entries, of which only finitely many entries are nonzero. We obtain an interpretation of the Bochner-Minlos theorem in R^(N) as the pure state space decomposition of the partial group algebras which generate L_V. We analyze the representation theory of L_V, and show that there is a bijection between a natural set of representations of L_V and the continuous unitary representations of R^(N), but that there is an extra part which essentially consists of the representation theory of a multiplicative semigroup which depends on the initial choice of approximate identity.

math.OA↗

arXiv2010

Localization via Automorphisms of the CARs. Local gauge invariance

The classical matter fields are sections of a vector bundle E with base manifold M. The space L^2(E) of square integrable matter fields w.r.t. a locally Lebesgue measure on M, has an important module action of C_b^\infty(M) on it. This module action defines restriction maps and encodes the local structure of the classical fields. For the quantum context, we show that this module action defines an automorphism group on the algebra A, of the canonical anticommutation relations on L^2(E), with which we can perform the analogous localization. That is, the net structure of the CAR, A, w.r.t. appropriate subsets of M can be obtained simply from the invariance algebras of appropriate subgroups. We also identify the quantum analogues of restriction maps. As a corollary, we prove a well-known "folk theorem," that the algebra A contains only trivial gauge invariant observables w.r.t. a local gauge group acting on E.

math-ph↗

arXiv2008

The Resolvent Algebra: A New Approach to Canonical Quantum Systems

The standard C*-algebraic version of the algebra of canonical commutation relations, the Weyl algebra, frequently causes difficulties in applications since it neither admits the formulation of physically interesting dynamical laws nor does it incorporate pertinent physical observables such as (bounded functions of) the Hamiltonian. Here a novel C*-algebra of the canonical commutation relations is presented which does not suffer from such problems. It is based on the resolvents of the canonical operators and their algebraic relations. The resulting C*-algebra, the resolvent algebra, is shown to have many desirable analytic properties and the regularity structure of its representations is surprisingly simple. Moreover, the resolvent algebra is a convenient framework for applications to interacting and to constrained quantum systems, as we demonstrate by several examples.

math.OA↗

arXiv2007

Generalising Group Algebras

This paper has been withdrawn by the author, since the main result, the existence and uniqueness theorem for host algebras, Theorem 3.4, is wrong for the following reasons. In Definition 3.1 we wanted to generalise the concept of an open projection away from universal enveloping Von Neumann algebras, so that it makes sense for any Von Neumann algebra N. We defined a projection P in N as open, if N.P is the intersection of left kernels of normal states of N. Unfortunately, every projection P in N will satisfy this criterion by Theorem 3.6.11 of Pedersen (C*-algebras and their automorphism groups). Hence if N = A" for a C*-algebra A, then this definition of an open projection does not define the usual open projections (cf. Proposition 3.11.9 in Pedersen). Hence the proof of Theorem 3.4, which rests on this, fails.

math.OA↗

arXiv2006

Algebraic Supersymmetry: A case study

The treatment of supersymmetry is known to cause difficulties in the C*-algebraic framework of relativistic quantum field theory; several no-go theorems indicate that super-derivations and super-KMS functionals must be quite singular objects in a C*-algebraic setting. In order to clarify the situation, a simple supersymmetric chiral field theory of a free Fermi and Bose field defined on $\R$ is analyzed. It is shown that a meaningful C*-version of this model can be based on the tensor product of a CAR-algebra and a novel version of a CCR-algebra, the "resolvent algebra". The elements of this resolvent algebra serve as mollifiers for the super-derivation. Within this model, unbounded (yet locally bounded) graded KMS-functionals are constructed and proven to be supersymmetric. From these KMS-functionals, Chern characters are obtained by generalizing formulae of Kastler and of Jaffe, Lesniewski and Osterwalder. The characters are used to define cyclic cocycles in the sense of Connes' noncommutative geometry which are "locally entire".

math-ph↗

arXiv2004

Group Algebras for Groups which are not Locally Compact

We generalise the definition of a group algebra so that it makes sense for non-locally compact topological groups, in particular, we require that the representation theory of the group algebra is isomorphic (in the sense of Gelfand-Raikov) to the continuous representation theory of the group, or to some other important subset of representations. We prove that a group algebra if it exists, is always unique up to isomorphism. From examples, group algebras do not always exist for non-locally compact groups, but they do exist for some. We define a convolution on the dual of the Fourier-Stieltjes algebra making it into a Banach *-algebra, we prove that a group algebra if it exists, can always be embedded in this convolution algebra, and we find sufficient conditions for a subalgebra to be a group algebra. When the group is locally compact, we obtain a new characterisation of its group algebra which does not involve the Haar measure, nor behaviour of measures on compact sets.

math.OA↗

arXiv2004

Superselection in the presence of constraints

For systems which contain both superselection structure and constraints, we study compatibility between constraining and superselection. Specifically, we start with a generalisation of Doplicher-Roberts superselection theory to the case of nontrivial centre, and a set of Dirac quantum constraints and find conditions under which the superselection structures will survive constraining in some form. This involves an analysis of the restriction and factorisation of superselection structures. We develop an example for this theory, modelled on interacting QED.

math-ph↗

arXiv2004

Amenability of the Gauge Group

Let G be one of the local gauge groups C(X,U(n)), C^\infty(X,U(n)), C(X,SU(n)) or C^\infty(X,SU(n)) where X is a compact Riemannian manifold. We observe that G has a nontrivial group topology, coarser than its natural topology, w.r.t. which it is amenable, viz the relative weak topology of C(X,M(n)). This topology seems more useful than other known amenable topologies for G. We construct a simple fermionic model containing an action of G, continuous w.r.t. this amenable topology.

math-ph↗

arXiv2000

Local Quantum Constraints

We analyze the situation of a local quantum field theory with constraints, both indexed by the same set of space-time regions. In particular we find ``weak'' Haag-Kastler axioms which will ensure that the final constrained theory satisfies the usual Haag-Kastler axioms. Gupta-Bleuler electromagnetism is developed in detail as an example of a theory which satisfies the ``weak'' Haag-Kastler axioms but not the usual ones. This analysis is done by pure C*-algebraic means without employing any indefinite metric representations, and we obtain the same physical algebra and positive energy representation for it than by the usual means. The price for avoiding the indefinite metric, is the use of nonregular representations and complex valued test functions. We also exhibit the precise connection with the usual indefinite metric representation. We conclude the analysis by comparing the final physical algebra produced by a system of local constrainings with the one obtained from a single global constraining and also consider the issue of reduction by stages. For the usual spectral condition on the generators of the translation group, we also find a ``weak'' version, and show that the Gupta-Bleuler example satisfies it.

math-ph↗

arXiv2000

Host Algebras

A host algebra generalises the concept of a group algebra as follows. Let F be a unital C*-algebra, and let S_0 be a proper subset of its states within which one wants to keep the analysis (e.g. F is the group algebra of a discrete group G, and S_0 is the set of states continuous w.r.t. some nondiscrete topology of G). Then a host algebra is a C*-algebra L for which we have embeddings of F and L into a larger C*-algebra E, such that the states on L extend uniquely to F, and this extension defines a norm continuous affine bijection between S_0 and the whole state space of L. The main examples (but not the only ones) are group and covariance algebras. Here we study existence questions of a host algebra for a given pair (F,S_0), we show that if a host algebra exists, we can do integral decompositions of states in S_0 in terms of other states in S_0, and we show that if one does induction of representations via host algebras one stays within the class of representations with the right continuity properties w.r.t. S_0. Moreover, if S_0 is a folium, then up to a central algebra one can always construct a host algebra, but this central algebra can be an obstruction to the existence of a host algebra. These results should be interesting to anyone who wants to construct group algebras for general topological groups, and they are also useful for quantum physics due to some selection criteria for physically acceptable states.

math.OA↗