Is There Only One Vacuum? Asymptotic Boundary, Vacuum Energy, and the Limit of QFT in dS Spacetime

Xinmu Zong, Chunyu Huang · 2025

Read Is There Only One Vacuum? Asymptotic Boundary, Vacuum Energy, and the Limit of QFT in dS Spacetime by Xinmu Zong, Chunyu Huang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-22084.txt.

Introduction

Xinmu Zong and Chunyu Huang question basic assumptions about empty space in quantum field theory in a 2025 paper. Textbooks build scattering theory on a single vacuum state with particles flying in from infinity. Expanding universes like ours have no such quiet infinity, which strains the whole setup.

The authors propose splitting the theory's machinery in two. The path integral governs a bounded interaction region, while the state space only encodes the free boundary conditions. Vacuum energy loops get a fresh thermodynamic reading under this scheme. Schemes that cancel vacuum energy are reexamined for their real motivation. The paper asks openly whether quantum field theory applies, as built, to de Sitter-like spacetimes at all.

Catalog record

CREATORXinmu Zong, Chunyu Huang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE15 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-22084

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Is There Only One Vacuum? Asymptotic Boundary, Vacuum Energy, and the Limit of QFT in dS Spacetime Xinmu Zong, Chunyu Huang Abstract This work proposes a reconstruction of the quantum field theory (QFT) scattering framework: the path integral governs an interaction kernel region, while the Hilbert space encodes asymptotic free boundary conditions. We critically reexamine foundational assumptions and standard observables of QFT, raising the possibility of a unique vacuum. We also introduce a new thermodynamic perspective on vacuum loop contributions. The motivation behind certain vacuum energy cancellation schemes is reconsidered, and the applicability of QFT to spacetimes without asymptotic flatness, such as de Sitter (dS) space, is questioned at a fundamental level. The framework of QFT scattering In QFT, scattering can be calculated under the following framework: a whole spacetime is divided into two parts: A kernel bulk where interaction may happen, with the path integral traveling all the way through it, and, an asymptotic boundary where a free theory can apply, with the Hilbert space defined there as a boundary condition. t t=+\infty t=-\infty \ket{0} \ket{p} \bra{p'} \bra{0} interaction region Figure 1: A hybrid framework. Instead of defining a Hilbert space in the interaction region and then constructing interaction picture’s states and operators, take the path integral as an axiom which describes time evolution whether or not involving interaction 1 1 1 In interaction theories like QCD, the Hilbert space may even be ill-defined, while the path integral works well. . The practice of placing the path integral in a more fundamental position when calculating interaction processes is also consistent with Haag’s theorem Haag1955 , which challenges the validity of the interaction picture in quantum field theory. The Hilbert space, however, still plays a role in the asymptotic free region 2 2 2 Here, asymptotic free refers to the condition that interactions vanish at spatial or temporal infinity, not to the asymptotic freedom in QCD. ; there, particle states are defined through the limiting process as boundary conditions. Think of how LSZ reduction is deduced Srednicki . The LSZ reduction formula \frac{\langle p^{\prime}\dots|p\dots\rangle}{\langle 0|0\rangle}=\int i(2\pi)^{-3/2}e^{ipx}(-\partial^{2}+m^{2})\dots\int^{\prime}i(2\pi)^{-3/2}e^{-ip^{\prime}x^{\prime}}(-\partial^{\prime 2}+m^{2})\dots\frac{\langle 0|T\phi^{\prime}\dots\phi\dots|0\rangle}{\langle 0|0\rangle} (1) with the n -point Green function / correlation function ( n\geq 0 ) \frac{\langle 0|T\phi(x_{1})\dots\phi(x_{n})|0\rangle}{\langle 0|0\rangle}=\frac{\int[D\phi]\,\phi(x_{1})\dots\phi(x_{n})e^{iS}}{\int[D\phi]\,e^{iS}} (2) build a bridge between path integral language and Hilbert space language, bulk integral behavior and boundary condition behavior, interaction involving theory and asymptotic free theory. The role of vacuum amplitude LSZ reduction has some precondition including that the ground state is unique and separated from excited states, so \langle 0|0\rangle=1 , which can be regarded as a normalization condition. Instead of setting \langle 0|0\rangle=1 , one can always divide a \langle 0|0\rangle to normalize like what is done above. And one can write \langle 0|0\rangle=\int[D\phi]\,e^{iS} (defined up to a factor). This path integral is divergent even in free theory. Think about a k -space loop \circ , or its exponential, \exp(\circ) , a loop is divergent, so \int[D\phi]\,e^{iS} is divergent. If the factor is nonzero, \langle 0|0\rangle=\infty ; if the factor is 0 or say \frac{1}{\infty} , \langle 0|0\rangle can be any number between 0 and \infty . In this sense, the value of \langle 0|0\rangle is not decided, and there is a physics reason. That is, \langle 0|0\rangle is not observable. One can imagine a scattering amplitude from incoming particles to outgoing particles, but who can imagine a scattering amplitude from vacuum to vacuum? \langle 0|0\rangle is a condition, not an observable. Up to this point, \langle 0|0\rangle can be any number. But in practical QFT calculations, one can pretend that \langle 0|0\rangle=1 as long as one omits all the vacuum bubbles ( \exp(\circ) in free theory, \exp(8\ldots) in \phi^{4} theory). In fact, \langle 0|0\rangle=1 serves not only as a normalization condition, but also as a renormalization condition. Here are some optional renormalization conditions: • \langle 0|0\rangle=1\ \Longleftrightarrow omit vacuum bubble • \langle 0|\phi|0\rangle=0\ \Longleftrightarrow omit tadpole, and not write linear counterterm Y\phi in Lagrangian density \mathcal{L} • \langle p|\phi|0\rangle=(2\pi)^{-3/2}e^{-ipx}\ \Longleftrightarrow OS scheme and \langle 0|0\rangle=1 , \langle 0|\phi|0\rangle=0 , \langle p|\phi|0\rangle=(2\pi)^{-3/2}e^{-ipx} is exactly what one wants an asymptotic free vacuum state to act like when deducing LSZ reduction formula. Renormalization technique also includes conditions on vertex functions and coupling constants. Although it is often said that renormalization has something to do with the fact that interaction alters the vacuum structure, one does not need such an interaction vacuum state formalism, just like one does not need Hilbert space language in an interaction region. One only needs the asymptotic free vacuum, |0\rangle 3 3 3 If Spontaneous symmetry breaking effect exists, the vacuum will transition from unstable to stable but remain asymptotic free. . Renormalization has everything to do with rescaling states and operators to connect the interaction region and asymptotic free region. Despite the renormalization condition \langle 0|0\rangle=1 , in principle, \langle 0|0\rangle=\int[D\phi]\,e^{iS(\phi)} (defined up to a factor) can still be any number if choose other renormalization scheme. Instead of writing a specific factor, say \langle 0|0\rangle can be set arbitrarily. Thermodynamics of vacuum Although unnormalized \langle 0|0\rangle is not an observable, it could have some deeper meaning other than a (re)normalization condition to be set. Considering Euclidean path integral with zero temperature limit: Z=\langle 0|0\rangle=\int[D\phi]\,e^{iS}=\int[D\phi]\,e^{-S_{E}} (3) and thermodynamics: F=-\frac{1}{\beta}\ln Z,\qquad F=\langle H\rangle-\frac{1}{\beta}S_{\text{TH}} (4) We found an equation: \langle H\rangle-\frac{1}{\beta}S_{\text{TH}}=-\frac{1}{\beta}\ln\langle 0|0\rangle\qquad(\beta\to\infty) (5) which seems to give (if S_{\text{TH}}-\ln\langle 0|0\rangle is not \infty ): \langle H\rangle=0,\qquad S_{\text{TH}}=\ln\langle 0|0\rangle (6) If set \langle 0|0\rangle=\infty , then thermal entropy S_{\text{TH}}=\infty . This means, if vacuum bubble is taken into account, they will contribute to vacuum entropy. If set \langle 0|0\rangle=1 , then thermal entropy S_{\text{TH}}=0=S_{\text{EN}} . This may seem like a trivial conclusion—that the vacuum has zero thermal entropy—and that the unitary condition \langle 0|0\rangle=1 corresponds to that the whole system has zero entanglement entropy S_{\text{EN}}=0 . However, this conclusion is not so trivial considering it still holds when interaction occurs somewhere. There is no information loss from an initial asymptotic free vacuum through an interaction zone to a final asymptotic in QFT framework. It is interesting to notice that thermal entropy is related to the path integral, as shown in this equation and its deduction, and the fact that the direction of thermal entropy increase is the direction of time, and the path integral cannot reverse the time arrow. Although notation \phi is used, the deduction is based on path integral, so our equation is not limited to scalar field. The vacuum energy \langle H\rangle comes to be zero, and we don’t see any boson and fermion energy cancellation here. In conventional quantum field theory, vacuum energy is often said to receive divergent contributions from vacuum bubble diagrams, which motivates mechanisms such as supersymmetry to cancel bosonic and fermionic zero-point energies. However, according to the equation we derived above, the vacuum energy expectation value is zero regardless of whether vacuum loops are included or subtracted. The underlying divergence from vacuum loops contributes only as an additive constant to the thermal entropy, not to the vacuum energy itself. Inclusion of gravity From now on, QFT in curved spacetime will be discussed. It is also said that gravitation changes the structure of vacuum. However, this need not be the case. First, gravitation is not a

Read the complete text