A convergent expansion about mean field theory: I. The expansion

We give a convergent expansion for nearly Gaussian quantum field theory in the multiphase region. The expansion combines (1) an expansion in phase boundaries, (2) a cluster expansion, and (3) a perturbation expansion to isolate dominant behavior. We study in detail the ground state of the $\text{P}$(φ)₂ = (λφ⁴ ? φ² ? μφ)₂ model, with ∥ μ ∥ ? λ² ? 1. The ground state is close to the classical free field, obtained by replacing $\text{P}$(φ) by the quadratic mean field polynomial $\text{P}$_c(φ), tangent to $\text{P}$ at a global minimum. Selecting one minimum gives a pure phase (ergodic ground state) satisfying the Wightman-Osterwalder-Schrader axioms with a positive mass. We also establish analyticity in λ for μ = 0 in the sector ∥ Im λ ∥ < ? Re λ ? 1, for ? ? 1.     

Conformal field theories near a boundary in general dimensions

The implications of restricted conformal invariance under conformal transformations preserving a plane boundary are discussed for general dimensions d. Calculations of the universal function of a conformal invariant ξ which appears in the two-point function of scalar operators in conformally invariant theories with a plane boundary are undertaken to first order in the ge = 4 − d expansion for the operator φ² in φ⁴ theory. The form for the associated functions of ξ for the two-point functions for the basic field φ^α and the auxiliary field λ in the N → ∞ limit of the O(N) nonlinear sigma model for any d in the range 2 < d < 4 are also rederived. These results are obtained by integrating the two-point functions over planes parallel to the boundary, defining a restricted two-point function which may be obtained more simply. Assuming conformal invariance this transformation can be inverted to recover the full two-point function. Consistency of the results is checked by considering the limit d → 4 and also by analysis of the operator product expansions for φ^αφ^β and λλ. Using this method the form of the two-point function for the energy-momentum tensor in the conformal O(N) model with a plane boundary is also found. General results for the sum of the contributions of all derivative operators appearing in the operator product expansion, and also in a corresponding boundary operator expansion, to the two-point functions are also derived making essential use of conformal invariance.     

Rotational diffusion of a tracer colloid particle: I. Short time orientational correlations

We extend the generalized Smoluchowski equation to descrbe the diffusional relaxation of position and orientation in a suspension of interacting spherical colloid particles. Considering a tracer particle which interacts with other particles through spherically symmetric pair potentials and with an external field we obtain a cluster expansion representation of the orientational time correlation functions for the tracer. The one and two body cluster contributions are studied explicitly at short times. Working to first order in volume fraction φ we show that the initial slope of the time correlation functions is described by a modified diffusion coefficient D^r = D^r₀(1 −C^rφ) where C^r is a number determined by hydrodynamic and potential interactions. We evaluate C^r numerically for spheres with slip-stick hydrodynamic boundary conditions and also for permeable spheres.     

Spontaneous symmetry breaking in the present and early universe

Spontaneous symmetry breaking in λφ⁴ theory is formulated in terms of the operator φ², and in a manner which requires no specific expectation value to be assigned to φ. At the one-loop order of perturbation theory, a renormalized effective action for a field ζ, linearly related to φ², is obtained as a gradient expansion. Potential advantages of this formulation in applications to phase transitions in the early universe are discussed. They include the possibilities (i) of obtaining a well-defined semiclassical equation of motion, and (ii) of following the evolution of a field theory from an initial symmetrical high temperature state without the introduction, ad hoc, of regions in which 〈φ〉 ≠ 0.     

Dimensional renormalization of scalar field theory in curved space-time

The regularization and renormalization of an interacting scalar field φ in a curved spacetime background is performed by the method of continuation to n dimensions. In addition to the familiar counter terms of the flat-space theory, c-number, “vacuum” counter terms must also be introduced. These involve zero, first, and second powers of the Reimann curvature tensor R_αβψδ. Moreover, the renormalizability of the theory requires that the Lagrange function couple φ² to the curvature scalar R with a coupling constant η. The coupling η must obey an inhomogeneous renormalization group equation, but otherwise it is an arbitrary, free parameter. All the counter terms obey renormalization group equations which determine the complete structure of these quantities in terms of the residues of their simple poles in n ? 4. The coefficient functions of the counter terms determine the construction of φ² and φ⁴ in terms of renormalized composite operators 1, [φ²], and [φ⁴]. Two of the counter terms vanish in conformally flat space-time. The others may be computed from the theory in purely flat space-time. They are determined, in a rather intricate fashion, by the additive renormalizations for two-point functions of [φ²] and [φ⁴] in Minkowski space-time. In particular, using this method, we compute the leading divergence of the R² interaction which is of fifth order in the coupling constant λ.     

The φ4 lattice field theory as an asymptotic expansion about the Ising limit

For a d-dimensional φ⁴ lattice field theory consisting of N spins, an asymptotic expansion of expectations about the Ising limit is established in inverse powers of the bare coupling constant λ. In the thermodynamic limit (N → ∞), the expansion is expected to be valid in the noncritical region of the Ising system.     

The :φ2: Field in theP(φ)2 model

Euclidean Field Theory techniques are used to study the Schwinger functions and characteristic function of the :φ²: field in evenP(φ)₂ models. The infinite volume limit is obtained for Half-Dirichlet boundary conditions by means of correlation inequalities. Analytic continuation yields Lorentz invariant Wightman functions. It is shown that, in the infinite volume limit, <:φ(x)²:>≧0 for both the Half and the Full-Dirichlet (λφ⁴)₂ model. This result also holds for a finite volume with periodic boundary conditions.     

A phase cell cluster expansion for a hierarchical Ø 3 4 model

The formalism developed in a previous paper is applied to yield a phase cell cluster expansion for a hierarchical ø ₃ ⁴ model. The field is expanded into modes with specific renormalization group scaling properties. The present cluster expansion for a vacuum expectation value is formally the natural factorization of each term in the perturbation expansion into the contribution of modes connected to the variables in the expectation via interactions, and that of the complementary set. The expectation value is thus realized as a sum of contributions due tofinite subsets of the modes. We emphasize the following additional features:

1. Partitions of unity are not used.
2. There areessentially no cut-offs.
3. The expansion is developed directly, without an initial need to prove an ultraviolet stability bound, the most difficult part of the traditional approach.

Our main interest in the present phase cell cluster expansion is founded in the belief that it may be the right vehicle for proving the existence of a nontrivial four-dimensional field theory.     

A non-perturbative analysis of symmetry breaking in two-dimensional φ4 theory using periodic field methods

We describe the generalization of spherical field theory to other modal expansion methods. The main approach remains the same, to reduce a d-dimensional field theory into a set of coupled one-dimensional systems. The method we discuss here uses an expansion with respect to periodic-box modes. We apply the method to φ⁴ theory in two dimensions and compute the critical coupling and critical exponents. We compare with lattice results and predictions via universality and the two-dimensional Ising model.     

Field-theoretic treatment of the 2d planar model with random p-fold symmetry-breaking field

It is shown that the effects of a random p-fold symmetry-breaking field on the two-dimensional planar model can be studied systematically using field-theoretic methods within the context of the replica scheme. In the absence of vortices we show that the model is renormalizable in a double expansion in a temperature difference variable δ and in g₀²: the strength of the disorder. The spin—spin and spin—glass correlation functions are calculated and their behavior under the renormalization group is analyzed. The spin—glass correlation function exhibits power-law behavior at any temperature but the power law is gaussian at high temperatures and controlled by a non-zero field fixed line at low temperature. Although our results agree substantially with previous work we can now point out situations in which the expansion breaks down and higher order terms must be retained. Adding vortices we can show that for $\text{p > 2}\text{2}\text{an}\text{XY}$ phase exists at intermediate values of temperature and for small values of g₀² for which our expansion is valid, vortices are irrelevant perturbations. However, in the low-temperature phase our expansion breaks down and we can no longer conclude that vortices are irrelevant. The nature of the low-temperature phase remains unresolved.     

Mass spectrum of the two dimensional λφ4−1/4φ2−μφ quantum field model

It is shown thatr-particle irreducible kernels in the two-dimensional λφ⁴?1/4φ²?μφ quantum field theory have (r+1)-particle decay for |μ|≦λ²?1. As a consequence there is an upper mass gap and, in the subspace of two-particle states, a bound state. The proof extends Spencer's expansion [20] to handle fluctuations between the two wells of the classical potential. A new method for resumming the low temperature cluster expansion is introduced.     

Generalised vector dominance: Off-diagonal terms and asymptotic SU(4)

Asymptotic SU(4) symmetry is imposed on generalised vector dominance models with off-diagonal terms, relating the magnitudes of these terms in the various sectors V = ?, ω, φ, ψ. While the structure functions are modified, their ratios remain similar to those of the diagonal model. Strangeness and charm production are suppressed for low Q² < M_φ², M_ψ² respectively.     

Thermodynamic properties of the φ4 lattice field theory near the Ising limit

For a d-dimensional φ⁴ lattice field theory consisting of N spins with nearest-neighbor interactions, the partition function is transformed for large bare coupling constant λ into an Ising-like system with additional neighbor interactions. For d = 2 a mean field approximation is then used to estimate the difference in critical temperature between the lattice φ⁴ field theory and its Ising limit (λ = ∞). Expansions are obtained for the susceptibility and specific heat. The critical exponents are shown to be identical to the Ising exponents.     

Mean work functions effective for negative-ionic, electronic and positive-ionic emissions from polycrystalline surfaces

To elucidate the thermionic property of polycrystalline surfaces, a further study is made on the mean work functions (φ⁻, φ^e and φ⁺) effective for negative-ionic, electronic and positive-ionic emissions. Comparison between theoretical analyses and experimental data yields the conclusions as follows. (1) The equation of φ⁻ = φ^e holds always with both mono- and polycrystalline surfaces. (2) The relation of φ⁻ = φ^e < φ⁺ applies to polycrystalline surfaces because they bear the thermionic contrast (Δφ^* ≡ φ⁺ − φ^e > 0). (3) The value of Δφ^* ranges from ∼0.4 to 0.9 eV depending upon the surface species of polycrystalline metals (e.g., W, Re and Pt), whilst Δφ^* = 0 for monocrystalline surfaces. (4) When the degree of monocrystallization (δ_m) is less than ∼50%, the theoretical value of Δφ^* is virtually independent of δ_m and agrees well with experimental data, nearly the same within ±0.1 eV among the so-called “polycrystalline” surfaces of W. (5) As δ_m increases beyond ∼80 up to 100%, Δφ^* decreases rapidly down to 0 eV, showing again a good agreement between theory and experiment. (6) Our theoretical model is valid in evaluating the effective mean work functions, irrespective of the range of δ_m.     

Theoretical evaluation of the effective work functions for positive-ionic and electronic emissions from polycrystalline metal surfaces

Effective work functions (φ⁺ and φ^e) for positive-ionic and electronic emissions from polycrystalline metals of Nb, Mo, Ta, W and Ir are calculated according to our theoretical model by using those published data on both fractional surface area (F_i) and local work function (φ_i) of each metal surface composed of several patchy faces (1, 2, …, i). Comparison between the theoretical values thus obtained and those experimental data published to date yields the conclusions as follows. (1) With a slight error of less than ∼0.1 eV, the value of φ^e calculated with each of the metals is in fair or good agreement with that determined by experiment. (2) Such agreement is found also with φ⁺ for W. (3) In a typical case of W, where the degree of monocrystallization (δ_m) corresponding to the largest among the values of F_i is less than ∼0.5, the thermionic contrast (Δφ* ≡ φ⁺ − φ^e) is found again to be nearly equal to both theoretical and experimental values reported previously. (4) Each of the five metals shows that Δφ* at δ_m = 0.68-0.95 is smaller than Δφ* at δ_m < 0.5. (5) This result strongly supports our theoretical prediction that Δφ* decreases gradually to zero as δ_m increases beyond ∼0.5 up to ∼1. (6) Particularly, such a surface which has δ_m ≥ 0.96 exhibits Δφ* ≈ 0, apparently equivalent to the so-called “monocrystalline surface (δ_m = 1)”. These results lead to the conclusion that our theoretical model is valid for evaluating the effective work functions probably with a slight error of less than ∼0.1 eV, irrespective of both the surface species and the range of δ_m. In addition, our simple model makes it possible to analyze the mechanism of change in φ⁺ and φ^e according to the change in surface characters of both φ_i and F_i.     

The quantum kinks of two-dimensional theories of the φ4 type

We apply a recently established method of kink quantization, to two-dimensional theories of the φ⁴ type possessing either a Z(4) × O(N) or an SU(N) symmetry and containing N complex scalar fields. Correlation functions of kinks are estimated through a 1/N expansion. Quantum kinks are interpolated by a local field, whenever a broken symmetry phase occurs. These kinks are massless. This result holds up to all orders in 1/N.     

The random-walk representation of classical spin systems and correlation inequalities

We use the random-walk representation to prove the first few of a new family of correlation inequalities for ferromagnetic ?⁴ lattice models. These inequalities state that the finite partial sums of the propagator-resummed perturbation expansion for the 4-point function form an alternating set of rigorous upper and lower bounds for the exact 4-point function. Generalizations to 2n-point functions are also given. A simple construction of the continuum ? _d ⁴ quantum field theory (d<4), based on these inequalities, is described in a companion paper.     

Vacuum instability in scalar field theories

The class of scalar field theories with interaction gφ^2N?1, are studied using the semi-classical approximation. The imaginary part of the vertex functions generated by tunnelling out of the metastable ground state is calculated to first order. Using this result, the leading asymptotic behaviour of the renormalisation group β function for φ³ field theory is obtained in six dimensions. The validity of this result is discussed in view of the extra singularities which appear when the theory is just renormalisable. The structure of the perturbation expansion for n component φ³ theory is also discussed, and cases in which these theories yield perturbation expansions which are Borel summable, are pointed out.     

Analysis of the lattice,strong coupling series for g0φ4 field theory in d dimensions

We analyze the renormalized strong coupling series for lattice g₀φ⁴ field theory which is a double series in x = M⁴a⁴/g₀a_4?dandy = 1/M²a², where M is the renormalized mass, a the lattice spacing, g₀ the bare coupling constant and d the dimension of space-time. We extrapolate to large y for fixed x by using a Padé-like extrapolation technique. We study the dimensionless renormalized coupling constant G/M^4?d and find that as we approach the continuum(x → 0, y → ∞) the entire spectrum of g₀ from zero to infinity can be studied. Our results for d = 1,2,3,4 based on a series in y up to y⁵ and in x up to x³ show that for fixed lattice spacing a, G/M^4?d is a monotonic function of g₀ ranging from zero at g₀ = 0 to a maximum at g₀ = ∞. Using the high temperature expansion results, we have also derived 9 terms in y on 8 lattices of dimension 1,2,3 and 4 for the linear term in x, and studied this series to see if one can see a breakdown in this monotonic behavior of G for large y. The analysis of this latter series is inconclusive.