Usual calculations of the (Euclidean) effective action in quantum gravity, typically performed within the heat-kernel formalism, give rise to quartically and quadratically UV-sensitive contributions (Planck scale) to the vacuum energy $\rho_{vac}=\Lambda_{cc}/8\pi G$, where $\Lambda_{cc}$ and $G$ are the cosmological and the Newton constant, respectively. The comparison with the observed value of $\rho_{vac}$ unveils a severe naturalness problem, the strongest facet to the long-standing cosmological constant problem. Several attempts have been put forward to dispose of the aforementioned power-like UV-sensitive contributions. Among them, models with compact extra dimensions received particular attention. Usual calculations of the vacuum energy in these models lead to the automatically finite (no fine-tuning) result $\rho_{vac}\sim \mkk^4$, with $\mkk^{-1}$ the size of the compact extra dimension. In the present thesis, we show that such a result comes from a mistreatment of the asymptotics of the five-dimensional loop momentum, and as such it is incorrect. When a proper calculation is performed, previously missed UV-sensitive terms arise that do not cancel even in the presence of supersymmetry. We then discuss the relevance of these results, also in connection with the recent dark dimension (DD) proposal. Formulated in the “swampland” framework, the latter suggests that we might live in a universe with a compact extra dimension of micrometer size dictated by the observed value of the vacuum energy. The DD scenario is based on swampland conjectures in quantum gravity and phenomenological bounds, that lead to the relation $\rho_{swamp}\sim\mkk^4$, and on the corresponding result $\rho_{EFT}$ from the effective field theory limit. The results of the present thesis show that the matching between $\rho_{swamp}$ and $\rho_{EFT}$ is a non-trivial issue, and further studies are needed to understand whether the DD scenario might really be a physical reality. After this opening analysis, we continue the investigation on the UV-sensitivity of the vacuum energy considering the calculation of the one-loop effective action in four-dimensional quantum gravity. We show that the appearance in usual calculations of the aforementioned quartically and quadratically UV-sensitive terms is due to an improper treatment of the path integral measure and of the UV physical cutoff $\Lambda$ of the theory. When the diffeomorphism invariant measure proposed by Fradkin and Vilkovisky is used, and $\Lambda$ is properly introduced, the radiative correction to $\rho_{vac}$ turns out to depend only logarithmically on $\Lambda$. We then extend this analysis considering the renormalization group (RG) flow of the gravitational action. Taking for the latter the Einstein-Hilbert truncation, we derive the RG equations for the running cosmological and Newton constant. We show that, if the Fradkin-Vilkovisky (FV) measure is used and the running scale $k$ properly introduced, the beta functions are profoundly different from those of previous literature. In particular, they do not possess the non-trivial UV-attractive fixed point of the so-called asymptotic safety scenario. We then consider the theory of an interacting scalar field on a non-trivial gravitational background. We find that the issues concerning the path integral measure and the UV physical cutoff have a further important implication on another long-standing problem in quantum field theory, the Higgs naturalness problem. We show that the use of the FV measure, together with a careful introduction of the UV cutoff $\Lambda$, ensure for the radiative correction $\delta m^2$ to the mass $m^2$ of the scalar field a result free of the well-known quadratically divergent contributions. Similarly to what we find for the vacuum energy, $\delta m^2$ depends only logarithmically on $\Lambda$. It is important to stress that the above results for the vacuum energy and for the mass of scalar particles have been obtained without resorting to a supersymmetric embedding of the theory, nor to regularization schemes where power-like divergences are absent by construction. Finally, in light of the important role that the FV measure plays in the derivation of the results of the present thesis, we thoroughly investigate on its transformation properties under diffeomorphisms. This measure is sometimes claimed not to be diffeomorphism invariant due to the presence in it of non-covariant $g^{00}$ factors of the time-time component of the inverse metric. We show that such a claim is incorrect, and that, on the contrary, these $g^{00}$ factors turn out to be crucial to ensure the diffeomorphism invariance of the path integral measure, and ultimately of the effective action.
I calcoli usuali dell’azione (Euclidea) efficace in gravità quantistica, tipicamente realizzati con il formalismo heat-kernel, danno luogo a contributi quarticamente e quadraticamente sensibili all’UV (scala di Planck) nell’energia del vuoto $\rho_{vac}=\Lambda_{cc}/8\pi G$, dove $\Lambda_{cc}$ e $G$ sono rispettivamente la costante cosmologica e la constante di Newton. Il confronto con il valore osservato di $\rho_{vac}$ svela un severo problema di naturalezza, il più severo contributo all’annoso problema della costante cosmologica. Diversi tentativi sono stati realizzati per sbarazzarsi dei suddetti contributi UV. Tra questi, modelli con dimensioni extra compatte hanno ricevuto particolare attenzione. I calcoli usuali dell’energia del vuoto in questi modelli portano al risultato automaticamente finito (no fine-tuning) $\rho_{vac}\sim \mkk^4$, dove $\mkk^{-1}$ è la lunghezza caratteristica della dimensione extra compatta. Nella presente tesi si mostra che tale risultato è dovuto a un trattamento improprio del limite asintotico per il momento di loop cinque-dimensionale e come tale è incorretto. Quando il calcolo è performato in modo appropriato si generano termini fortemente sensibili all’UV, che sono persi dai calcoli usuali e che non si cancellano neanche in presenza di supersimmetria. Successivamente si discute la rilevanza di questi risultati anche in connessione con la recente proposta di una “dark dimension” (DD). Formulata nel contesto di “swampland”, questa proposta suggerisce che potremmo vivere in un universo con una dimensione extra compatta con lunghezza caratteristica dell’ordine del micrometro dettata dal valore osservato dell’energia del vuoto. Lo scenario DD è basato su congetture swampland in gravità quantistica e su limiti fenomenologici, che portano alla relazione $\rho_{swamp}\sim\mkk^4$, e sul corrispondente risultato $\rho_{EFT}$ dal limite di teoria di campo efficace. I risultati della presente tesi mostrano che la corrispondenza tra $\rho_{swamp}$ and $\rho_{EFT}$ è una questione non banale e che ulteriori studi sono necessari per capire se lo scenario DD sia (o meno) una realtà fisica. Dopo quest’analisi di apertura, si continua lo studio sulla sensibilità alla fisica UV dell’energia del vuoto considerando il calcolo dell’azione efficace a one-loop in gravità quantistica quadridimensionale. Si mostra che la comparsa nei calcoli usuali dei termini quarticamente e quadraticamente UV-sensibili menzionati sopra è dovuta a un trattamento improprio della misura del path integral e del cutoff fisico UV $\Lambda$ della teoria. Quando si usa la misura proposta da Fradkin e Vilkovisky, la quale è invariante per diffeomorfismi, e $\Lambda$ è propriamente introdotto, la correzione radiativa a $\rho_{vac}$ risulta dipendere soltanto logaritmicamente da $\Lambda$. Successivamente questa analisi viene estesa considerando il flusso del gruppo di rinormalizzazione (RG) dell’azione gravitazionale. In particolare, prendendo per quest’ultima il troncamento Einstein-Hilbert, si derivano le equazioni RG per la costante cosmologica e la costante di Newton. Si mostra che l’uso della misura di Fradkin-Vilkovisky (FV) e un’introduzione appropriata della scala running $k$ portano a delle “beta functions” che sono profondamente diverse da quelle ricavate in letteratura precedente. In particolare, esse non possiedono il punto fisso non banale UV-attrattivo del cosiddetto scenario di “asymptotic safety”. Successivamente si considera la teoria di un campo scalare interagente su un background gravitazionale non banale. Si trova che le questioni riguardanti la misura del path integral e il cutoff fisico UV hanno un’ulteriore importante implicazione su un altro annoso problema in teoria quantistica dei campi, il problema della naturalezza dell’Higgs. Si mostra che l’uso della misura di FV, insieme con un’introduzione attenta del cutoff UV $\Lambda$, assicurano per la correzione radiativa $\delta m^2$ alla massa $m^2$ del campo scalare un risultato privo dei ben noti contributi quadraticamente divergenti. Similmente a ciò che è stato trovato per l’energia del vuoto, $\delta m^2$ dipende solo logaritmicamente da $\Lambda$. È importante sottolineare che i suddetti risultati per l’energia del vuoto e per la massa di particelle scalari sono stati ottenuti senza far ricorso né a un embedding supersimmetrico della teoria né a schemi di regolarizzazione dove le divergenze a potenza sono assenti per costruzione. Infine, alla luce dell’importante ruolo che la misura di FV gioca nella derivazione dei risultati della presente tesi, vengono studiate a fondo le sue proprietà di trasformazione sotto diffeomorfismi. A volte si dice che questa misura non è invariante per via della presenza in essa di fattori non covarianti della componente tempo-tempo $g^{00}$ dell’inverso del tensore metrico. Si mostra che quest’affermazione è incorretta e che, al contrario, questi fattori di $g^{00}$ risultano essere cruciali per assicurare l’invarianza sotto diffeomorfismi della misura del path integral e, in ultimo, dell’azione efficace.
UV-sensitivity in quantum gravity and path integral measure [Sensibilità UV in gravità quantistica e la misura del path integral]
PERNACE, ARCANGELO
2026
Abstract
Usual calculations of the (Euclidean) effective action in quantum gravity, typically performed within the heat-kernel formalism, give rise to quartically and quadratically UV-sensitive contributions (Planck scale) to the vacuum energy $\rho_{vac}=\Lambda_{cc}/8\pi G$, where $\Lambda_{cc}$ and $G$ are the cosmological and the Newton constant, respectively. The comparison with the observed value of $\rho_{vac}$ unveils a severe naturalness problem, the strongest facet to the long-standing cosmological constant problem. Several attempts have been put forward to dispose of the aforementioned power-like UV-sensitive contributions. Among them, models with compact extra dimensions received particular attention. Usual calculations of the vacuum energy in these models lead to the automatically finite (no fine-tuning) result $\rho_{vac}\sim \mkk^4$, with $\mkk^{-1}$ the size of the compact extra dimension. In the present thesis, we show that such a result comes from a mistreatment of the asymptotics of the five-dimensional loop momentum, and as such it is incorrect. When a proper calculation is performed, previously missed UV-sensitive terms arise that do not cancel even in the presence of supersymmetry. We then discuss the relevance of these results, also in connection with the recent dark dimension (DD) proposal. Formulated in the “swampland” framework, the latter suggests that we might live in a universe with a compact extra dimension of micrometer size dictated by the observed value of the vacuum energy. The DD scenario is based on swampland conjectures in quantum gravity and phenomenological bounds, that lead to the relation $\rho_{swamp}\sim\mkk^4$, and on the corresponding result $\rho_{EFT}$ from the effective field theory limit. The results of the present thesis show that the matching between $\rho_{swamp}$ and $\rho_{EFT}$ is a non-trivial issue, and further studies are needed to understand whether the DD scenario might really be a physical reality. After this opening analysis, we continue the investigation on the UV-sensitivity of the vacuum energy considering the calculation of the one-loop effective action in four-dimensional quantum gravity. We show that the appearance in usual calculations of the aforementioned quartically and quadratically UV-sensitive terms is due to an improper treatment of the path integral measure and of the UV physical cutoff $\Lambda$ of the theory. When the diffeomorphism invariant measure proposed by Fradkin and Vilkovisky is used, and $\Lambda$ is properly introduced, the radiative correction to $\rho_{vac}$ turns out to depend only logarithmically on $\Lambda$. We then extend this analysis considering the renormalization group (RG) flow of the gravitational action. Taking for the latter the Einstein-Hilbert truncation, we derive the RG equations for the running cosmological and Newton constant. We show that, if the Fradkin-Vilkovisky (FV) measure is used and the running scale $k$ properly introduced, the beta functions are profoundly different from those of previous literature. In particular, they do not possess the non-trivial UV-attractive fixed point of the so-called asymptotic safety scenario. We then consider the theory of an interacting scalar field on a non-trivial gravitational background. We find that the issues concerning the path integral measure and the UV physical cutoff have a further important implication on another long-standing problem in quantum field theory, the Higgs naturalness problem. We show that the use of the FV measure, together with a careful introduction of the UV cutoff $\Lambda$, ensure for the radiative correction $\delta m^2$ to the mass $m^2$ of the scalar field a result free of the well-known quadratically divergent contributions. Similarly to what we find for the vacuum energy, $\delta m^2$ depends only logarithmically on $\Lambda$. It is important to stress that the above results for the vacuum energy and for the mass of scalar particles have been obtained without resorting to a supersymmetric embedding of the theory, nor to regularization schemes where power-like divergences are absent by construction. Finally, in light of the important role that the FV measure plays in the derivation of the results of the present thesis, we thoroughly investigate on its transformation properties under diffeomorphisms. This measure is sometimes claimed not to be diffeomorphism invariant due to the presence in it of non-covariant $g^{00}$ factors of the time-time component of the inverse metric. We show that such a claim is incorrect, and that, on the contrary, these $g^{00}$ factors turn out to be crucial to ensure the diffeomorphism invariance of the path integral measure, and ultimately of the effective action.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi-Pernace.pdf
accesso aperto
Licenza:
Tutti i diritti riservati
Dimensione
2.74 MB
Formato
Adobe PDF
|
2.74 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/360637
URN:NBN:IT:UNICT-360637