UFIFT-QG-26-09
CCTP-2026-19
ITCP/2026/19
PACS numbers: 04.50.Kd, 95.35.+d, 98.62.-g
Introduction
The symmetries of cosmology are homogeneity and isotropy. We can take the background geometry to be,
d s 2 = − d t 2 + a 2 ( t ) d x ⃗ ⋅ d x ⃗ = a 2 [ − d η 2 + d x ⃗ ⋅ d x ⃗ ] ⟹ H ( t ) ≡ a ˙ a , ϵ ( t ) ≡ − H ˙ H 2 . (1) ds^2=-dt^2+a^2(t)d\vec{x}\cdot d\vec{x}=a^2[-d\eta^2+d\vec{x}\cdot d\vec{x}]\quad\Longrightarrow\quad H(t)\equiv\frac{\dot{a}}{a},\quad\epsilon(t)\equiv-\frac{\dot{H}}{H^2}. \tag*{(1)} d s 2 = − d t 2 + a 2 ( t ) d x ⋅ d x = a 2 [ − d η 2 + d x ⋅ d x ] ⟹ H ( t ) ≡ a a ˙ , ϵ ( t ) ≡ − H 2 H ˙ . (1)
Here a ( t ) a(t) a ( t ) is the scale factor, H ( t ) H(t) H ( t ) is the Hubble parameter, and ϵ ( t ) \epsilon(t) ϵ ( t ) is the first slow roll parameter.
Accelerated expansion (H > 0 H>0 H > 0 with 0 ≤ ϵ < 1 0\leq\epsilon<1 0 ≤ ϵ < 1 ) is known as inflation. Inflation alters the energy-time uncertainty relation so that massless, long wavelength virtual particles can persist indefinitely [1 ] . Particles which are both massless and not conformally invariant, such as massless, minimally coupled scalars and gravitons, experience explosive particle production during inflation. For the limiting case of de Sitter (ϵ = 0 \epsilon=0 ϵ = 0 , a ( t ) = e H t a(t)=e^{Ht} a ( t ) = e H t ) the occupation number of gravitons with a single polarization and wave vector k ⃗ \vec{k} k is,
N ( t , k ) = [ H a ( t ) 2 k ] 2 . (2) N(t,k)=\left[\frac{Ha(t)}{2k}\right]^2. \tag*{(2)} N ( t , k ) = [ 2 k H a ( t ) ] 2 . (2)
This is what produces the tensor power spectrum [2 , 3 ] , and it is natural to expect significant loop corrections as well. Because N ( t , k ) N(t,k) N ( t , k ) is negligible for the ultraviolet (k ≫ H a ( t ) k\gg Ha(t) k ≫ H a ( t ) ) and large for the infrared (k ≪ H a ( t ) k\ll Ha(t) k ≪ H a ( t ) ), we can reliably study these corrections using quantum general relativity as a low energy effective field theory in the sense of Donoghue [4 , 5 , 6 ] .
The exponential time dependence evident in expression (2) causes loop corrections that would be constant in flat space background to become time dependent in de Sitter background. One way to study this is by using the graviton 1PI (one-particle-irreducible) 2-point function − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\!^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) to quantum-correct the linearized Einstein equation,
L μ ν ρ σ κ h ρ σ ( x ) − ∫ d 4 x ′ [ μ ν Σ ρ σ ] ( x ; x ′ ) κ h ρ σ ( x ′ ) = 8 π G T μ ν ( x ) . (3) \mathcal{L}^{\mu\nu\rho\sigma}\kappa h_{\rho\sigma}(x)-\int d^4x'[\!^{\mu\nu}\Sigma^{\rho\sigma}](x;x')\kappa h_{\rho\sigma}(x')=8\pi G T^{\mu\nu}(x). \tag*{(3)} L μν ρ σ κ h ρ σ ( x ) − ∫ d 4 x ′ [ μν Σ ρ σ ] ( x ; x ′ ) κ h ρ σ ( x ′ ) = 8 π G T μν ( x ) . (3)
Here L μ ν ρ σ \mathcal{L}^{\mu\nu\rho\sigma} L μν ρ σ is the Lichnerowicz operator, the graviton field h μ ν h_{\mu\nu} h μν is defined in terms of the full metric g μ ν ≡ a 2 [ η μ ν + κ h μ ν ] g_{\mu\nu}\equiv a^2[\eta_{\mu\nu}+\kappa h_{\mu\nu}] g μν ≡ a 2 [ η μν + κ h μν ] , with κ 2 ≡ 16 π G \kappa^2\equiv16\pi G κ 2 ≡ 16 π G , and T μ ν T^{\mu\nu} T μν is the variation of the matter action with respect to h μ ν h_{\mu\nu} h μν . By setting the stress tensor to zero one can use equation (3) to work out corrections to the mode function u ( t , k ) u(t,k) u ( t , k ) of plane wave gravitational radiation,
T μ ν = 0 ⟹ h μ ν ( x ) = u ( t , k ) e i k ⃗ ⋅ x ⃗ ϵ μ ν ( k ⃗ , λ ) . (4) T^{\mu\nu}=0\quad\Longrightarrow\quad h_{\mu\nu}(x)=u(t,k)e^{i\vec{k}\cdot\vec{x}}\epsilon_{\mu\nu}(\vec{k},\lambda). \tag*{(4)} T μν = 0 ⟹ h μν ( x ) = u ( t , k ) e i k ⋅ x ϵ μν ( k , λ ) . (4)
Choosing the stress tensor to be that of a static, point mass allows one to work out corrections to the Newtonian potential Ψ ( t , r ) \Psi(t,r) Ψ ( t , r ) ,
T μ ν = M a δ 3 ( x ⃗ ) δ 0 μ δ 0 ν ⟹ κ h 00 ( x ) = − 2 Ψ ( t , r ) . (5) T^{\mu\nu}=Ma\delta^{3}(\vec{x})\delta_{0}^{\mu}\delta_{0}^{\nu}
\qquad\Longrightarrow\qquad
\kappa h_{00}(x)=-2\Psi(t,r). \tag*{(5)} T μν = M a δ 3 ( x ) δ 0 μ δ 0 ν ⟹ κ h 00 ( x ) = − 2Ψ ( t , r ) . (5)
Quantum gravity on de Sitter background is hard . Only a handful of graviton loop corrections to 1PI 2-point functions have been computed. These include:
The self-mass − i M 2 ( x ; x ′ ) -iM^{2}(x;x') − i M 2 ( x ; x ′ ) of a massless, minimally coupled scalar [7 , 8 ] ;
The self-mass of a massless, conformally coupled scalar [9 , 10 , 11 ] ;
The self-energy − i [ i Σ j ] ( x ; x ′ ) -i[{}_{i}\Sigma_{j}](x;x') − i [ i Σ j ] ( x ; x ′ ) of a massless [12 ] , or massive [13 ] , Dirac fermion; and
The vacuum polarization i [ μ ν Π ρ σ ] ( x ; x ′ ) i[{}^{\mu\nu}\Pi^{\rho\sigma}](x;x') i [ μν Π ρ σ ] ( x ; x ′ ) [14 , 15 ] .
For each of these results, which were dimensionally regulated [16 , 17 ] and fully renormalized in the sense of BPHZ (Bogoliubov and Parasiuk [18 ] , Hepp [19 ] , and Zimmermann [20 , 21 ] , the loop involved only a single graviton propagator and the 1PI 2-point function had at most two free indices. In contrast, the graviton self-energy − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) involves two graviton propagators and has four free indices. It is so complicated that the original calculation [22 ] was unregulated, in D = 4 D=4 D = 4 spacetime dimensions, and away from coincidence ( x μ ≠ x ′ μ ) (x^{\mu}\ne x'^{\mu}) ( x μ = x ′ μ ) . That suffices to check the Ward identity, but it is not enough for the effective field equation (3), which involves an integral over all x ′ μ x'^{\mu} x ′ μ , including coincidence. The purpose of this paper is to perform a dimensionally regulated and fully renormalized computation. We will also use the self-energy in equation (3) to check estimates for the leading secular dependence of u ( t , k ) u(t,k) u ( t , k ) and Ψ ( t , r ) \Psi(t,r) Ψ ( t , r ) that were obtained by applying a set of plausible rules to extend the noncoincident self-energy to a guess for the fully renormalized result [23 , 24 ] ,
u ( t , k ) ⟶ H 2 k 3 { 1 + κ 2 H 2 3 π 2 × ln 2 [ a ( t ) ] + ⋯ } . (6) u(t,k)\longrightarrow\frac{H}{\sqrt{2k^{3}}}\left\{1+\frac{\kappa^{2}H^{2}}{3\pi^{2}}\times\ln^{2}[a(t)]+\cdots\right\}. \tag*{(6)} u ( t , k ) ⟶ 2 k 3 H { 1 + 3 π 2 κ 2 H 2 × ln 2 [ a ( t )] + ⋯ } . (6)
Ψ ( t , r ) ⟶ G M a ( t ) r { 1 − κ 2 H 2 2 π 2 × ln 3 [ a ( t ) ] + ⋯ } . (7) \Psi(t,r)\longrightarrow\frac{GM}{a(t)r}\left\{1-\frac{\kappa^{2}H^{2}}{2\pi^{2}}\times\ln^{3}[a(t)]+\cdots\right\}. \tag*{(7)} Ψ ( t , r ) ⟶ a ( t ) r GM { 1 − 2 π 2 κ 2 H 2 × ln 3 [ a ( t )] + ⋯ } . (7)
In Section 2 we review the Feynman rules. The primitive, dimensionally regulated result for − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) is derived in Section 3. BPHZ renormalization is accomplished in Section 4, and the effective field equation (3) is solved for u ( t , k ) u(t,k) u ( t , k ) and Ψ ( t , r ) \Psi(t,r) Ψ ( t , r ) in Section 5. Our conclusions comprise Section 6. Some expansions for products of propagators are given in an Appendix.
Feynman Rules
The purpose of this section is to present those of the Feynman rules which are needed to compute − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma\rho\sigma](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) at 1-loop order on de Sitter background. We begin by expressing the invariant Lagrangian in terms of the graviton field and the conformally rescaled metric g ~ μ ν ≡ a − 2 × g μ ν \widetilde{g}_{\mu\nu} \equiv a^{-2} \times g_{\mu\nu} g μν ≡ a − 2 × g μν . Then the gauge fixing functional is presented, along with explicit expressions for the ghost and graviton propagators.
The invariant Lagrangian of general relativity in D D D spacetime dimensions with cosmological constant Λ \Lambda Λ is,
L i n v = [ R − ( D − 2 ) Λ ] − g 16 π G κ 2 ≡ 16 π G Λ ≡ ( D − 1 ) H 2 . (8) \mathcal{L}_{\mathrm{inv}} = \frac{[R-(D-2)\Lambda]\sqrt{-g}}{16\pi G}
\qquad
\kappa^2 \equiv16\pi G
\qquad
\Lambda\equiv(D-1)H^2 . \tag*{(8)} L inv = 16 π G [ R − ( D − 2 ) Λ ] − g κ 2 ≡ 16 π G Λ ≡ ( D − 1 ) H 2 . (8)
By judicious partial integrations one can express this as [25 ] ,
L i n v ⟶ a D − 2 − g ~ g ~ α β g ~ γ δ g ~ ρ σ { 1 2 h α ρ , γ h δ σ , β − 1 2 h α β , ρ h σ γ , δ + 1 4 h α β , ρ h γ δ , σ − 1 4 h α ρ , γ h β σ , δ } + ( D 2 − 1 ) a D − 1 H − g ~ g ~ γ δ g ~ ρ σ h ρ σ , γ h δ 0 . (9) \mathcal{L}_{\mathrm{inv}} \longrightarrow a^{D-2}\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\rho\sigma}
\left\{
\frac{1}{2}h_{\alpha\rho,\gamma}h_{\delta\sigma,\beta}
-\frac{1}{2}h_{\alpha\beta,\rho}h_{\sigma\gamma,\delta}
+\frac{1}{4}h_{\alpha\beta,\rho}h_{\gamma\delta,\sigma}
-\frac{1}{4}h_{\alpha\rho,\gamma}h_{\beta\sigma,\delta}
\right\}
+\left(\frac{D}{2}-1\right)a^{D-1}H\sqrt{-\widetilde{g}}\,\widetilde{g}^{\gamma\delta}\widetilde{g}^{\rho\sigma}h_{\rho\sigma,\gamma}h_{\delta0}. \tag*{(9)} L inv ⟶ a D − 2 − g g α β g γ δ g ρ σ { 2 1 h α ρ , γ h δ σ , β − 2 1 h α β , ρ h σ γ , δ + 4 1 h α β , ρ h γ δ , σ − 4 1 h α ρ , γ h β σ , δ } + ( 2 D − 1 ) a D − 1 H − g g γ δ g ρ σ h ρ σ , γ h δ 0 . (9)
A linearization instability precludes adding a de Sitter invariant gauge fixing function [25 , 26 ] . We instead fix the gauge by breaking spatial special conformal transformations, while preserving homogeneity, isotropy and dilatation invariance [25 , 27 ] ,
L G F = − 1 2 a D − 2 η μ ν F μ F ν F μ = η ρ σ [ h μ ρ , σ − 1 2 h ρ σ , μ + ( D − 2 ) a H h μ ρ δ σ 0 ] . (10) \mathcal{L}_{\mathrm{GF}}=-\frac{1}{2}a^{D-2}\eta^{\mu\nu}F_\mu F_\nu
\qquad
F_\mu=\eta^{\rho\sigma}\left[h_{\mu\rho,\sigma}-\frac{1}{2}h_{\rho\sigma,\mu}+(D-2)aHh_{\mu\rho}\delta^0_\sigma\right]. \tag*{(10)} L GF = − 2 1 a D − 2 η μν F μ F ν F μ = η ρ σ [ h μ ρ , σ − 2 1 h ρ σ , μ + ( D − 2 ) a H h μ ρ δ σ 0 ] . (10)
As usual, the ghost Lagrangian L g h = − a D − 2 c ~ μ δ F μ \mathcal{L}_{\mathrm{gh}}=-a^{D-2}\widetilde{c}^{\mu}\delta F_\mu L gh = − a D − 2 c μ δ F μ is formed by making a linearized coordinate transformation by the ghost field x μ → x μ + κ c μ x^\mu\rightarrow x^\mu+\kappa c^\mu x μ → x μ + κ c μ ,
L g h = − a D − 2 c ~ μ , ν [ 2 g ~ ρ ( μ ∂ ν ) + g ~ μ ν , ρ + 2 a H g ~ μ ν δ ρ 0 ] c ρ + ( a D − 2 c ~ μ ) , μ η ρ σ [ g ~ ν ρ ∂ σ + g ~ ρ σ , ν + a H g ~ ρ σ δ ν 0 ] c ν . (11) \mathcal{L}_{\mathrm{gh}}=-a^{D-2}\widetilde{c}^{\mu,\nu}
\left[2\widetilde{g}_{\rho(\mu}\partial_{\nu)}+\widetilde{g}_{\mu\nu,\rho}+2aH\widetilde{g}_{\mu\nu}\delta^0_\rho\right]c^\rho
+\left(a^{D-2}\widetilde{c}^{\mu}\right)_{,\mu}\eta^{\rho\sigma}
\left[\widetilde{g}_{\nu\rho}\partial_\sigma+\widetilde{g}_{\rho\sigma,\nu}+aH\widetilde{g}_{\rho\sigma}\delta^0_\nu\right]c^\nu. \tag*{(11)} L gh = − a D − 2 c μ , ν [ 2 g ρ ( μ ∂ ν ) + g μν , ρ + 2 a H g μν δ ρ 0 ] c ρ + ( a D − 2 c μ ) , μ η ρ σ [ g ν ρ ∂ σ + g ρ σ , ν + a H g ρ σ δ ν 0 ] c ν . (11)
In this gauge the graviton and ghost propagators can be expressed using three scalar propagators times tensor factors which are spacetime constants,
i [ μ ν Δ ρ σ ] ( x ; x ′ ) = ∑ I = A , B , C [ μ ν T ρ σ I ] × i Δ I ( x ; x ′ ) . (12) i[\mu\nu\Delta_{\rho\sigma}](x;x')=\sum_{I=A,B,C}[\mu\nu T^I_{\rho\sigma}]\times i\Delta_I(x;x'). \tag*{(12)} i [ μν Δ ρ σ ] ( x ; x ′ ) = I = A , B , C ∑ [ μν T ρ σ I ] × i Δ I ( x ; x ′ ) . (12)
i [ μ Δ ρ ] ( x ; x ′ ) = η μ ρ × i Δ A ( x ; x ′ ) − δ μ 0 δ ρ 0 × i Δ B ( x ; x ′ ) . (13) i[\mu\Delta_\rho](x;x')=\eta_{\mu\rho}\times i\Delta_A(x;x')-\delta^0_\mu\delta^0_\rho\times i\Delta_B(x;x'). \tag*{(13)} i [ μ Δ ρ ] ( x ; x ′ ) = η μ ρ × i Δ A ( x ; x ′ ) − δ μ 0 δ ρ 0 × i Δ B ( x ; x ′ ) . (13)
An overline indicates the suppression of a tensor’s temporal indices, η ˉ μ ν ≡ η μ ν + δ μ 0 δ ν 0 \bar{\eta}_{\mu\nu} \equiv\eta_{\mu\nu}+\delta^{0}_{\mu}\delta^{0}_{\nu} η ˉ μν ≡ η μν + δ μ 0 δ ν 0 . The various tensor factors in (12) are,
[ μ ν T ρ σ A ] ≡ 2 η ˉ μ ( ρ η ˉ ν ) σ − 2 D − 3 η ˉ μ ν η ˉ ρ σ (14) [{}_{\mu\nu}T^{A}_{\rho\sigma}] \equiv2\bar{\eta}_{\mu(\rho}\bar{\eta}_{\nu)\sigma}-\frac{2}{D-3}\bar{\eta}_{\mu\nu}\bar{\eta}_{\rho\sigma} \tag*{(14)} [ μν T ρ σ A ] ≡ 2 η ˉ μ ( ρ η ˉ ν ) σ − D − 3 2 η ˉ μν η ˉ ρ σ (14)
[ μ ν T ρ σ B ] ≡ − 4 δ ( μ 0 η ˉ ν ) ( ρ δ σ ) 0 (15) [{}_{\mu\nu}T^{B}_{\rho\sigma}] \equiv-4\delta^{0}_{(\mu}\bar{\eta}_{\nu)(\rho}\delta^{0}_{\sigma)} \tag*{(15)} [ μν T ρ σ B ] ≡ − 4 δ ( μ 0 η ˉ ν ) ( ρ δ σ ) 0 (15)
[ μ ν T ρ σ C ] ≡ 2 ( D − 2 ) ( D − 3 ) [ ( D − 3 ) δ μ 0 δ ν 0 + η ˉ μ ν ] [ ( D − 3 ) δ ρ 0 δ σ 0 + η ˉ ρ σ ] . [{}_{\mu\nu}T^{C}_{\rho\sigma}] \equiv\frac{2}{(D-2)(D-3)}\left[(D-3)\delta^{0}_{\mu}\delta^{0}_{\nu}+\bar{\eta}_{\mu\nu}\right]\left[(D-3)\delta^{0}_{\rho}\delta^{0}_{\sigma}+\bar{\eta}_{\rho\sigma}\right]. [ μν T ρ σ C ] ≡ ( D − 2 ) ( D − 3 ) 2 [ ( D − 3 ) δ μ 0 δ ν 0 + η ˉ μν ] [ ( D − 3 ) δ ρ 0 δ σ 0 + η ˉ ρ σ ] .
Here and henceforth, parenthesized indices are symmetrized.
The propagators i Δ I ( x ; x ′ ) i\Delta_I(x;x') i Δ I ( x ; x ′ ) correspond to minimally coupled scalars with masses M A 2 = 0 M_A^2=0 M A 2 = 0 , M B 2 = ( D − 2 ) H 2 M_B^2=(D-2)H^2 M B 2 = ( D − 2 ) H 2 and M C 2 = 2 ( D − 3 ) H 2 M_C^2=2(D-3)H^2 M C 2 = 2 ( D − 3 ) H 2 ,
[ D − M I 2 a 2 ] i Δ I ( x ; x ′ ) = i δ D ( x − x ′ ) D ≡ ∂ μ a D − 2 ∂ μ . (16) [D-M_I^2a^2]i\Delta_I(x;x')=i\delta^D(x-x') \qquad D\equiv\partial^\mu a^{D-2}\partial_\mu. \tag*{(16)} [ D − M I 2 a 2 ] i Δ I ( x ; x ′ ) = i δ D ( x − x ′ ) D ≡ ∂ μ a D − 2 ∂ μ . (16)
The three propagators depend mostly on the de Sitter invariant length function y ( x ; x ′ ) ≡ a a ′ H 2 Δ x 2 y(x;x')\equiv aa'H^2\Delta x^2 y ( x ; x ′ ) ≡ a a ′ H 2 Δ x 2 ,
i Δ A = H D − 2 ( 4 π ) D / 2 { Γ ( D 2 ) D 2 − 1 ( 4 y ) D 2 − 1 + Γ ( D 2 + 1 ) D 2 − 2 ( 4 y ) D 2 − 2 } + k [ ln ( a a ′ ) − π cot ( D π 2 ) ] + H D − 2 ( 4 π ) D / 2 ∑ n = 1 ∞ [ Γ ( n + D − 1 ) n Γ ( n + D 2 ) ( y 4 ) n − Γ ( n + D 2 + 1 ) ( n − D 2 + 2 ) Γ ( n + 2 ) ( y 4 ) n − D 2 + 2 ] , (17) \begin{aligned}
i\Delta_A={}&\frac{H^{D-2}}{(4\pi)^{D/2}}\left\{\frac{\Gamma\left(\frac{D}{2}\right)}{\frac{D}{2}-1}\left(\frac{4}{y}\right)^{\frac{D}{2}-1}+\frac{\Gamma\left(\frac{D}{2}+1\right)}{\frac{D}{2}-2}\left(\frac{4}{y}\right)^{\frac{D}{2}-2}\right\}+k\left[\ln(aa')-\pi\cot\left(\frac{D\pi}{2}\right)\right]\\
&+\frac{H^{D-2}}{(4\pi)^{D/2}}\sum_{n=1}^{\infty}\left[\frac{\Gamma(n+D-1)}{n\Gamma\left(n+\frac{D}{2}\right)}\left(\frac{y}{4}\right)^n-\frac{\Gamma\left(n+\frac{D}{2}+1\right)}{\left(n-\frac{D}{2}+2\right)\Gamma(n+2)}\left(\frac{y}{4}\right)^{n-\frac{D}{2}+2}\right], \tag*{(17)}
\end{aligned} i Δ A = ( 4 π ) D /2 H D − 2 { 2 D − 1 Γ ( 2 D ) ( y 4 ) 2 D − 1 + 2 D − 2 Γ ( 2 D + 1 ) ( y 4 ) 2 D − 2 } + k [ ln ( a a ′ ) − π cot ( 2 D π ) ] + ( 4 π ) D /2 H D − 2 n = 1 ∑ ∞ [ n Γ ( n + 2 D ) Γ ( n + D − 1 ) ( 4 y ) n − ( n − 2 D + 2 ) Γ ( n + 2 ) Γ ( n + 2 D + 1 ) ( 4 y ) n − 2 D + 2 ] , (17)
i Δ B = H D − 2 ( 4 π ) D / 2 { Γ ( D 2 ) D 2 − 1 ( 4 y ) D 2 − 1 + ∑ n = 0 ∞ [ Γ ( n + D 2 ) Γ ( n + 2 ) ( y 4 ) n − D 2 + 2 − Γ ( n + D − 2 ) Γ ( n + D 2 ) ( y 4 ) n ] } . (18) i\Delta_B=\frac{H^{D-2}}{(4\pi)^{D/2}}\left\{\frac{\Gamma\left(\frac{D}{2}\right)}{\frac{D}{2}-1}\left(\frac{4}{y}\right)^{\frac{D}{2}-1}+\sum_{n=0}^{\infty}\left[\frac{\Gamma\left(n+\frac{D}{2}\right)}{\Gamma(n+2)}\left(\frac{y}{4}\right)^{n-\frac{D}{2}+2}-\frac{\Gamma(n+D-2)}{\Gamma\left(n+\frac{D}{2}\right)}\left(\frac{y}{4}\right)^n\right]\right\}. \tag*{(18)} i Δ B = ( 4 π ) D /2 H D − 2 { 2 D − 1 Γ ( 2 D ) ( y 4 ) 2 D − 1 + n = 0 ∑ ∞ [ Γ ( n + 2 ) Γ ( n + 2 D ) ( 4 y ) n − 2 D + 2 − Γ ( n + 2 D ) Γ ( n + D − 2 ) ( 4 y ) n ] } . (18)
i Δ C = H D − 2 ( 4 π ) D / 2 { Γ ( D 2 ) D 2 − 1 ( 4 y ) D 2 − 1 − ∑ n = 0 ∞ [ ( n − D 2 + 3 ) Γ ( n + D 2 − 1 ) Γ ( n + 2 ) ( y 4 ) n − D 2 + 2 − ( n + 1 ) Γ ( n + D − 3 ) Γ ( n + D 2 ) ( y 4 ) n ] } . (19) i\Delta_C=\frac{H^{D-2}}{(4\pi)^{D/2}}\left\{\frac{\Gamma\left(\frac{D}{2}\right)}{\frac{D}{2}-1}\left(\frac{4}{y}\right)^{\frac{D}{2}-1}-\sum_{n=0}^{\infty}\left[\frac{\left(n-\frac{D}{2}+3\right)\Gamma\left(n+\frac{D}{2}-1\right)}{\Gamma(n+2)}\left(\frac{y}{4}\right)^{n-\frac{D}{2}+2}-\frac{(n+1)\Gamma(n+D-3)}{\Gamma\left(n+\frac{D}{2}\right)}\left(\frac{y}{4}\right)^n\right]\right\}. \tag*{(19)} i Δ C = ( 4 π ) D /2 H D − 2 { 2 D − 1 Γ ( 2 D ) ( y 4 ) 2 D − 1 − n = 0 ∑ ∞ [ Γ ( n + 2 ) ( n − 2 D + 3 ) Γ ( n + 2 D − 1 ) ( 4 y ) n − 2 D + 2 − Γ ( n + 2 D ) ( n + 1 ) Γ ( n + D − 3 ) ( 4 y ) n ] } . (19)
The constant k k k is,
k ≡ H D − 2 ( 4 π ) D / 2 Γ ( D − 1 ) Γ ( D 2 ) . (20) k\equiv\frac{H^{D-2}}{(4\pi)^{D/2}}\frac{\Gamma(D-1)}{\Gamma\left(\frac{D}{2}\right)}. \tag*{(20)} k ≡ ( 4 π ) D /2 H D − 2 Γ ( 2 D ) Γ ( D − 1 ) . (20)
Note that the infinite series corrections in expressions (17–19) cancel in D = 4 D=4 D = 4 spacetime dimensions. This means that the D = 4 D=4 D = 4 limits of the propagators involve only one or two terms,
lim D = 4 i Δ A ( x ; x ′ ) = 1 4 π 2 1 a a ′ Δ x 2 − H 2 8 π 2 ln ( 1 4 H 2 Δ x 2 ) , (21) \lim_{D=4}i\Delta_A(x;x')=\frac{1}{4\pi^2}\frac{1}{aa'\Delta x^2}-\frac{H^2}{8\pi^2}\ln\left(\frac{1}{4}H^2\Delta x^2\right), \tag*{(21)} D = 4 lim i Δ A ( x ; x ′ ) = 4 π 2 1 a a ′ Δ x 2 1 − 8 π 2 H 2 ln ( 4 1 H 2 Δ x 2 ) , (21)
lim D = 4 i Δ B ( x ; x ′ ) = 1 4 π 2 1 a a ′ Δ x 2 , (22) \lim_{D=4}i\Delta_B(x;x')=\frac{1}{4\pi^2}\frac{1}{aa'\Delta x^2}, \tag*{(22)} D = 4 lim i Δ B ( x ; x ′ ) = 4 π 2 1 a a ′ Δ x 2 1 , (22)
lim D = 4 i Δ C ( x ; x ′ ) = 1 4 π 2 1 a a ′ Δ x 2 . (23) \lim_{D=4}i\Delta_C(x;x')=\frac{1}{4\pi^2}\frac{1}{aa'\Delta x^2}. \tag*{(23)} D = 4 lim i Δ C ( x ; x ′ ) = 4 π 2 1 a a ′ Δ x 2 1 . (23)
Because the infinite series corrections in (17-19) vanish more and more rapidly at coincidence with increasing n n n , only a few terms ever need to be retained. The graviton propagator (12) can be expressed in a useful way,
i [ μ ν Δ ρ σ ] ( x ; x ′ ) = [ 2 η μ ( ρ η σ ) ν − 2 D − 2 η μ ν η ρ σ ] i Δ C ( x ; x ′ ) + [ μ ν T ρ σ A ] [ i Δ A ( x ; x ′ ) − i Δ C ( x ; x ′ ) ] + [ μ ν T ρ σ B ] [ i Δ B ( x ; x ′ ) − i Δ C ( x ; x ′ ) ] . (24) i[_{\mu\nu}\Delta_{\rho\sigma}](x;x') = \left[2\eta_{\mu(\rho}\eta_{\sigma)\nu} - \frac{2}{D-2}\eta_{\mu\nu}\eta_{\rho\sigma}\right]i\Delta_C(x;x') + [_{\mu\nu}T^A_{\rho\sigma}]\left[i\Delta_A(x;x')-i\Delta_C(x;x')\right] + [_{\mu\nu}T^B_{\rho\sigma}]\left[i\Delta_B(x;x')-i\Delta_C(x;x')\right]. \tag*{(24)} i [ μν Δ ρ σ ] ( x ; x ′ ) = [ 2 η μ ( ρ η σ ) ν − D − 2 2 η μν η ρ σ ] i Δ C ( x ; x ′ ) + [ μν T ρ σ A ] [ i Δ A ( x ; x ′ ) − i Δ C ( x ; x ′ ) ] + [ μν T ρ σ B ] [ i Δ B ( x ; x ′ ) − i Δ C ( x ; x ′ ) ] . (24)
One advantage of this form is that the Poincaré invariant tensor factor on the first line is the same as for the flat space limit. Another advantage is that only this term is quadratically divergent; the A A A -C C C and B B B -C C C propagator differences on the second line are only logarithmically divergent.
Some special cases occur with significant frequency. Two of these consist of diagonal and cross-diagonal contractions,
1 8 × i [ ρ σ Δ ρ σ ] − 1 4 × i [ ρ σ Δ ρ σ ] = − 1 4 D ( D − 1 ) i Δ A − 1 2 ( D − 1 ) i Δ B − 1 2 i Δ C , (25) \frac{1}{8}\times i[^{\rho\sigma}\Delta_{\rho\sigma}] - \frac{1}{4}\times i[^{\rho\sigma}\Delta_{\rho\sigma}] = -\frac{1}{4}D(D-1)i\Delta_A - \frac{1}{2}(D-1)i\Delta_B - \frac{1}{2}i\Delta_C , \tag*{(25)} 8 1 × i [ ρ σ Δ ρ σ ] − 4 1 × i [ ρ σ Δ ρ σ ] = − 4 1 D ( D − 1 ) i Δ A − 2 1 ( D − 1 ) i Δ B − 2 1 i Δ C , (25)
− 1 2 i [ ρ σ Δ μ ν ] + i [ μ ρ Δ ρ ν ] = η ˉ μ ν [ D i Δ A + i Δ B ] − δ 0 μ δ 0 ν [ ( D − 1 ) i Δ B + 2 i Δ C ] . (26) -\frac{1}{2}i[^{\rho\sigma}\Delta^{\mu\nu}] + i[^{\mu\rho}\Delta^{\nu}_{\ \rho}] = \bar{\eta}^{\mu\nu}[Di\Delta_A+i\Delta_B] - \delta^\mu_0\delta^\nu_0[(D-1)i\Delta_B+2i\Delta_C]. \tag*{(26)} − 2 1 i [ ρ σ Δ μν ] + i [ μ ρ Δ ρ ν ] = η ˉ μν [ D i Δ A + i Δ B ] − δ 0 μ δ 0 ν [( D − 1 ) i Δ B + 2 i Δ C ] . (26)
It is also interesting when one of the indices is temporal,
i [ ρ ν Δ ρ 0 ] = ( D − 1 ) δ 0 ν i Δ B + 2 ( D − 3 D − 2 ) δ 0 ν i Δ C , (27) i[^{\rho\nu}\Delta_{\rho0}] = (D-1)\delta^\nu_0 i\Delta_B + 2\left(\frac{D-3}{D-2}\right)\delta^\nu_0 i\Delta_C , \tag*{(27)} i [ ρ ν Δ ρ 0 ] = ( D − 1 ) δ 0 ν i Δ B + 2 ( D − 2 D − 3 ) δ 0 ν i Δ C , (27)
i [ μ ν Δ ν 0 ] = 2 δ 0 μ ( η ˉ μ ν ) i Δ B − [ 2 ( D − 3 D − 2 ) δ 0 μ δ 0 ν + 2 D − 2 η ˉ μ ν ] δ 0 ρ i Δ C . (28) i[^{\mu\nu}\Delta_{\nu0}] = 2\delta^\mu_0\left(\bar{\eta}^{\mu\nu}\right)i\Delta_B - \left[2\left(\frac{D-3}{D-2}\right)\delta^\mu_0\delta^\nu_0+\frac{2}{D-2}\bar{\eta}^{\mu\nu}\right]\delta^\rho_0 i\Delta_C . \tag*{(28)} i [ μν Δ ν 0 ] = 2 δ 0 μ ( η ˉ μν ) i Δ B − [ 2 ( D − 2 D − 3 ) δ 0 μ δ 0 ν + D − 2 2 η ˉ μν ] δ 0 ρ i Δ C . (28)
There is also the case when two indices are temporal,
i [ μ ν Δ ν ′ 0 ] = − η ˉ μ ν i Δ B + 2 ( D − 3 D − 2 ) δ 0 μ δ 0 ν i Δ C . (29) i[^{\mu\nu}\Delta^{{\nu'}_{0}}] = -\bar{\eta}^{\mu\nu}i\Delta_B + 2\left(\frac{D-3}{D-2}\right)\delta^\mu_0\delta^\nu_0 i\Delta_C . \tag*{(29)} i [ μν Δ ν ′ 0 ] = − η ˉ μν i Δ B + 2 ( D − 2 D − 3 ) δ 0 μ δ 0 ν i Δ C . (29)
Primitive Results
This section is the heart of the paper. In it we compute the dimensionally regulated, primitive result for the 1-loop correction to $-i3 (x;x')$. We begin with a discussion of the tensor and functional basis in terms of which the answer is expressed. The actual computation begins with the simplest diagram, which involves a single graviton 4-point vertex, then the ghost loop is presented and, finally, the graviton loop between a pair of 3-point vertices.
The Tensor & Functional Basis
The graviton self-energy $-i3 (x;x') is a second rank, bi-tensor density which is symmetric under the interchange of \muw i t h with w i t h \nu$, and also under the interchange of ρ \rho ρ with σ \sigma σ . It can be expressed as a superposition of 21 tensor differential operators [ μ ν D i ρ σ ] [{}^{\mu\nu}D_i^{\rho\sigma}] [ μν D i ρ σ ] acting on three functions f j ( x − x ′ ) f_j(x-x') f j ( x − x ′ ) of the difference between the two coordinates,
− i [ μ ν Σ ρ σ ] ( x ; x ′ ) = ∑ i = 1 21 ∑ j = 1 3 C i j ( a , a ′ , ∂ ) × [ μ ν D i ρ σ ] × f j ( x − x ′ ) . (30) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x')=\sum_{i=1}^{21}\sum_{j=1}^{3}C_{ij}(a,a',\partial)\times[{}^{\mu\nu}D_i^{\rho\sigma}]\times f_j(x-x') . \tag*{(30)} − i [ μν Σ ρ σ ] ( x ; x ′ ) = i = 1 ∑ 21 j = 1 ∑ 3 C ij ( a , a ′ , ∂ ) × [ μν D i ρ σ ] × f j ( x − x ′ ) . (30)
The 4-index tensor differential operators [ μ ν D i ρ σ ] [{}^{\mu\nu}D_i^{\rho\sigma}] [ μν D i ρ σ ] are listed in Table 1 .
Table 1. The 21 basis tensors.
Note that we only need unprimed derivatives because one can reflect primed derivatives into unprimed one when acting on a function of the difference of x μ x^\mu x μ and x ′ μ x'^\mu x ′ μ ,
∂ μ ′ f j ( x − x ′ ) = − ∂ μ f j ( x − x ′ ) . (31) \partial'_\mu f_j(x-x')=-\partial_\mu f_j(x-x') . \tag*{(31)} ∂ μ ′ f j ( x − x ′ ) = − ∂ μ f j ( x − x ′ ) . (31)
The first of the three functions is the D D D -dimensional delta function,
f 1 ( x − x ′ ) ≡ δ D ( x − x ′ ) . (32) f_1(x-x')\equiv\delta^D(x-x') . \tag*{(32)} f 1 ( x − x ′ ) ≡ δ D ( x − x ′ ) . (32)
All divergences reside in the coefficients C i 1 ( a , a ′ , ∂ ) C_{i1}(a,a',\partial) C i 1 ( a , a ′ , ∂ ) acting on this term, although these coefficients can be contain finite contributions as well. The two remaining functions are integrable in the effective field equation (3) , and depend on the conformal coordinate interval,
Δ x 2 ≡ ∥ x ⃗ − x ⃗ ′ ∥ 2 − ( ∣ η − η ′ ∣ − i ε ) 2 . (33) \Delta x^2\equiv\lVert\vec{x}-\vec{x}'\rVert^2-\left(\lvert\eta-\eta'\rvert-i\varepsilon\right)^2 . \tag*{(33)} Δ x 2 ≡ ∥ x − x ′ ∥ 2 − ( ∣ η − η ′ ∣ − i ε ) 2 . (33)
These two functions are,
f 2 ( x − x ′ ) ≡ ln 2 ( 1 4 H 2 Δ x 2 ) , f 3 ( x − x ′ ) ≡ ln ( 1 4 H 2 Δ x 2 ) . (34) f_2(x-x') \equiv\ln^2\left(\frac{1}{4}H^2\Delta x^2\right), \qquad f_3(x-x') \equiv\ln\left(\frac{1}{4}H^2\Delta x^2\right). \tag*{(34)} f 2 ( x − x ′ ) ≡ ln 2 ( 4 1 H 2 Δ x 2 ) , f 3 ( x − x ′ ) ≡ ln ( 4 1 H 2 Δ x 2 ) . (34)
Important differential relations exist between the functions,
( Δ η ∂ 2 + 2 ∂ 0 ) f 2 = − 4 ∂ 0 f 3 , Δ η ∂ 4 f 2 = − 2 ∂ 0 ∂ 2 f 3 , ( Δ η ∂ 2 + 2 ∂ 0 ) f 3 = 0 , ∂ 4 f 3 = 16 π 2 i f 1 . (35) \begin{aligned}
(\Delta\eta\partial^2+2\partial_0)f_2&=-4\partial_0f_3, \qquad\Delta\eta\partial^4f_2=-2\partial_0\partial^2f_3,\\
(\Delta\eta\partial^2+2\partial_0)f_3&=0, \qquad\partial^4f_3=16\pi^2if_1. \tag*{(35)}
\end{aligned} ( Δ η ∂ 2 + 2 ∂ 0 ) f 2 ( Δ η ∂ 2 + 2 ∂ 0 ) f 3 = − 4 ∂ 0 f 3 , Δ η ∂ 4 f 2 = − 2 ∂ 0 ∂ 2 f 3 , = 0 , ∂ 4 f 3 = 16 π 2 i f 1 . (35)
The coefficients C i 2 ( a , a ′ , ∂ ) C_{i2}(a,a',\partial) C i 2 ( a , a ′ , ∂ ) and C i 3 ( a , a ′ , ∂ ) C_{i3}(a,a',\partial) C i 3 ( a , a ′ , ∂ ) are finite and are evaluated for D = 4 D=4 D = 4 .
The factors within the various C i j ( a , a ′ , ∂ ) C_{ij}(a,a',\partial) C ij ( a , a ′ , ∂ ) have any derivatives standing to the right, so that they can be reflected (31) . As a 1PI 2-point function, the graviton self-energy must also obey Reflection Invariance,
− i [ μ ν Σ ρ σ ] ( x ; x ′ ) = − i [ ρ σ Σ μ ν ] ( x ′ ; x ) . (36) -i\left[{}^{\mu\nu}\Sigma^{\rho\sigma}\right](x;x')=-i\left[{}^{\rho\sigma}\Sigma^{\mu\nu}\right](x';x). \tag*{(36)} − i [ μν Σ ρ σ ] ( x ; x ′ ) = − i [ ρ σ Σ μν ] ( x ′ ; x ) . (36)
This implies relations between seven pairs of the coefficient functions,
C ( i + 1 ) j ( a , a ′ , ∂ ) = s i × C i j ( a ′ , a , − ∂ ) for i ∈ { 3 , 5 , 7 , 10 , 14 , 16 , 19 } . (37) C_{(i+1)j}(a,a',\partial)=s_i\times C_{ij}(a',a,-\partial)\qquad\text{for }i\in\{3,5,7,10,14,16,19\}. \tag*{(37)} C ( i + 1 ) j ( a , a ′ , ∂ ) = s i × C ij ( a ′ , a , − ∂ ) for i ∈ { 3 , 5 , 7 , 10 , 14 , 16 , 19 } . (37)
The factors s i s_i s i are ± 1 \pm1 ± 1 according to the rules,
s i = { + 1 i ∈ { 3 , 7 , 16 } − 1 i ∈ { 5 , 10 , 14 , 19 } . (38) s_i=\begin{cases}
+1 & i\in\{3,7,16\}\\
-1 & i\in\{5,10,14,19\}.
\end{cases} \tag*{(38)} s i = { + 1 − 1 i ∈ { 3 , 7 , 16 } i ∈ { 5 , 10 , 14 , 19 } . (38)
The 4-Point Contribution
The simplest 1-loop contribution to the self-energy comes from a single 4-point vertex,
− i [ μ ν Σ 4 p t ρ σ ] ( x ; x ′ ) = ⟨ Ω ∣ [ i δ 2 S δ h μ ν ( x ) δ h ρ σ ( x ′ ) ] h 2 ∣ Ω ⟩ (39) -i\left[{}^{\mu\nu}\Sigma^{\rho\sigma}_{\mathrm{4pt}}\right](x;x')=\left\langle\Omega\left|\left[\frac{i\delta^2S}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right]_{h^2}\right|\Omega\right\rangle \tag*{(39)} − i [ μν Σ 4pt ρ σ ] ( x ; x ′ ) = ⟨ Ω [ δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 S ] h 2 Ω ⟩ (39)
The subscript “h 2 h^2 h 2 ” indicates that two graviton fields are retained after the variation. The diagramatic representation of expression (39) is depicted in Figure 1 .
Figure 1. Contribution to the graviton self-energy from a single 4-point vertex (39) . Wavy lines stand for gravitons.
The 130 quadratic contributions to the second variation in expression (39) are listed in Table 4 of [28 ] , however, this formidable result simplifies considerably when the expectation value is taken. To understand this, consider the first and second variations of the first term of the invariant Lagrangian (9) ,
Table 4. Primitive local 3-point contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
L 1 ≡ 1 2 a D − 2 − g ~ g ~ α β g ~ γ δ g ~ κ λ h α κ , γ h δ λ , β . (40) \mathcal{L}_1 \equiv\frac{1}{2}a^{D-2}\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\kappa\lambda}h_{\alpha\kappa,\gamma}h_{\delta\lambda,\beta}. \tag*{(40)} L 1 ≡ 2 1 a D − 2 − g g α β g γ δ g κλ h α κ , γ h δ λ , β . (40)
The first variation results in all the fields being evaluated at x μ x^\mu x μ ,
δ S 1 δ h μ ν ( x ) = − ∂ β { a D − 2 − g ~ g ~ α β g ~ γ ( μ g ~ ν ) κ h α κ , γ } + 1 2 a D − 2 ∂ ∂ h μ ν { − g ~ g ~ α β g ~ γ δ g ~ κ λ } h α κ , γ h δ λ , β . (41) \frac{\delta S_1}{\delta h_{\mu\nu}(x)}=-\partial_\beta\left\{a^{D-2}\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma(\mu}\widetilde{g}^{\nu)\kappa}h_{\alpha\kappa,\gamma}\right\}+\frac{1}{2}a^{D-2}\frac{\partial}{\partial h_{\mu\nu}}\left\{\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\kappa\lambda}\right\}h_{\alpha\kappa,\gamma}h_{\delta\lambda,\beta}. \tag*{(41)} δ h μν ( x ) δ S 1 = − ∂ β { a D − 2 − g g α β g γ ( μ g ν ) κ h α κ , γ } + 2 1 a D − 2 ∂ h μν ∂ { − g g α β g γ δ g κλ } h α κ , γ h δ λ , β . (41)
Note the distinction between the variational derivative on the left and the ordinary derivative on the second line of the right hand side. Taking the second variation does not change the fact that all the fields are evaluated at the same point x μ x^\mu x μ ,
δ 2 S 1 δ h μ ν ( x ) δ h ρ σ ( x ′ ) = − ∂ β { a D − 2 − g ~ g ~ β ( ρ g ~ σ ) ( μ g ~ ν ) κ δ D ( x − x ′ ) } − ∂ β { a D − 2 ∂ ∂ h ρ σ [ − g ~ g ~ α β g ~ γ ( μ g ~ ν ) κ ] h α γ , κ δ D ( x − x ′ ) } + a D − 2 ∂ ∂ h μ ν { − g ~ g ~ α β g ~ γ ( ρ g ~ σ ) κ } h α γ , κ ∂ β δ D ( x − x ′ ) + 1 2 a D − 2 ∂ 2 ∂ h μ ν ∂ h ρ σ { − g ~ g ~ α β g ~ γ δ g ~ κ λ } h α γ , κ h λ δ , β δ D ( x − x ′ ) . (42) \begin{aligned}
\frac{\delta^2 S_1}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}={}&-\partial_\beta\left\{a^{D-2}\sqrt{-\widetilde{g}}\,\widetilde{g}^{\beta(\rho}\widetilde{g}^{\sigma)(\mu}\widetilde{g}^{\nu)\kappa}\delta^D(x-x')\right\}\\
&-\partial_\beta\left\{a^{D-2}\frac{\partial}{\partial h_{\rho\sigma}}\left[\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma(\mu}\widetilde{g}^{\nu)\kappa}\right]h_{\alpha\gamma,\kappa}\delta^D(x-x')\right\}\\
&+a^{D-2}\frac{\partial}{\partial h_{\mu\nu}}\left\{\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma(\rho}\widetilde{g}^{\sigma)\kappa}\right\}h_{\alpha\gamma,\kappa}\partial_\beta\delta^D(x-x')\\
&+\frac{1}{2}a^{D-2}\frac{\partial^2}{\partial h_{\mu\nu}\partial h_{\rho\sigma}}\left\{\sqrt{-\widetilde{g}}\,\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\kappa\lambda}\right\}h_{\alpha\gamma,\kappa}h_{\lambda\delta,\beta}\delta^D(x-x'). \tag*{(42)}
\end{aligned} δ h μν ( x ) δ h ρ σ ( x ′ ) δ 2 S 1 = − ∂ β { a D − 2 − g g β ( ρ g σ ) ( μ g ν ) κ δ D ( x − x ′ ) } − ∂ β { a D − 2 ∂ h ρ σ ∂ [ − g g α β g γ ( μ g ν ) κ ] h α γ , κ δ D ( x − x ′ ) } + a D − 2 ∂ h μν ∂ { − g g α β g γ ( ρ g σ ) κ } h α γ , κ ∂ β δ D ( x − x ′ ) + 2 1 a D − 2 ∂ h μν ∂ h ρ σ ∂ 2 { − g g α β g γ δ g κλ } h α γ , κ h λ δ , β δ D ( x − x ′ ) . (42)
The next step is to expand each term in (43) to quadratic order in the graviton field, and then take the expectation value, which produces coincident, sometimes differentiated, propagators.
The coincidence limits of the scalar propagators (17-19) are,
i Δ A ( x ; x ) = k [ 2 ln ( a ) − π cot ( D π 2 ) ] = − k π cot ( D π 2 ) a D − 4 + O ( D − 4 ) , (43) i\Delta_A(x;x)=k\left[2\ln(a)-\pi\cot\left(\frac{D\pi}{2}\right)\right]=-\frac{k\pi\cot\left(\frac{D\pi}{2}\right)}{a^{D-4}}+O(D-4), \tag*{(43)} i Δ A ( x ; x ) = k [ 2 ln ( a ) − π cot ( 2 D π ) ] = − a D − 4 k π cot ( 2 D π ) + O ( D − 4 ) , (43)
i Δ B ( x ; x ) = − k D − 2 = − H 2 16 π 2 + O ( D − 4 ) , (44) i\Delta_B(x;x)=-\frac{k}{D-2}=-\frac{H^2}{16\pi^2}+O(D-4), \tag*{(44)} i Δ B ( x ; x ) = − D − 2 k = − 16 π 2 H 2 + O ( D − 4 ) , (44)
i Δ C ( x ; x ) = − k ( D − 3 ) ( D − 2 ) = H 2 16 π 2 + O ( D − 4 ) . (45) i\Delta_C(x;x)=-\frac{k}{(D-3)(D-2)}=\frac{H^2}{16\pi^2}+O(D-4). \tag*{(45)} i Δ C ( x ; x ) = − ( D − 3 ) ( D − 2 ) k = 16 π 2 H 2 + O ( D − 4 ) . (45)
The coincidence limits of singly differentiated propagators are,
lim x ′ → x ∂ μ i Δ A ( x ; x ′ ) = k H a δ μ 0 = H 3 8 π 2 a δ μ 0 + O ( D − 4 ) , (46) \lim_{x'\to x}\partial_\mu i\Delta_A(x;x')=kHa\delta^0_\mu=\frac{H^3}{8\pi^2}a\delta^0_\mu+O(D-4), \tag*{(46)} x ′ → x lim ∂ μ i Δ A ( x ; x ′ ) = k H a δ μ 0 = 8 π 2 H 3 a δ μ 0 + O ( D − 4 ) , (46)
lim x ′ → x ∂ μ i Δ B ( x ; x ′ ) = 0 = lim x ′ → x ∂ μ i Δ C ( x ; x ′ ) . (47) \lim_{x'\to x}\partial_\mu i\Delta_B(x;x')=0=\lim_{x'\to x}\partial_\mu i\Delta_C(x;x'). \tag*{(47)} x ′ → x lim ∂ μ i Δ B ( x ; x ′ ) = 0 = x ′ → x lim ∂ μ i Δ C ( x ; x ′ ) . (47)
And the coincidence limits of mixed second derivatives are,
lim x ′ → x ∂ μ ∂ ν ′ i Δ A ( x ; x ′ ) = − ( D − 1 D ) k H 2 g μ ν = − 3 H 4 32 π 2 g μ ν + O ( D − 4 ) , (48) \lim_{x'\to x}\partial_\mu\partial'_\nu i\Delta_A(x;x')=-\left(\frac{D-1}{D}\right)kH^2g_{\mu\nu}=-\frac{3H^4}{32\pi^2}g_{\mu\nu}+O(D-4), \tag*{(48)} x ′ → x lim ∂ μ ∂ ν ′ i Δ A ( x ; x ′ ) = − ( D D − 1 ) k H 2 g μν = − 32 π 2 3 H 4 g μν + O ( D − 4 ) , (48)
lim x ′ → x ∂ μ ∂ ν ′ i Δ B ( x ; x ′ ) = k H 2 D g μ ν = H 4 32 π 2 g μ ν + O ( D − 4 ) , (49) \lim_{x'\to x}\partial_\mu\partial'_\nu i\Delta_B(x;x')=\frac{kH^2}{D}g_{\mu\nu}=\frac{H^4}{32\pi^2}g_{\mu\nu}+O(D-4), \tag*{(49)} x ′ → x lim ∂ μ ∂ ν ′ i Δ B ( x ; x ′ ) = D k H 2 g μν = 32 π 2 H 4 g μν + O ( D − 4 ) , (49)
lim x ′ → x ∂ μ ∂ ν ′ i Δ C ( x ; x ′ ) = − 2 k H 2 ( D − 2 ) D g μ ν = − H 4 32 π 2 g μ ν + O ( D − 4 ) . (50) \lim_{x'\to x}\partial_\mu\partial'_\nu i\Delta_C(x;x')=-\frac{2kH^2}{(D-2)D}g_{\mu\nu}=-\frac{H^4}{32\pi^2}g_{\mu\nu}+O(D-4). \tag*{(50)} x ′ → x lim ∂ μ ∂ ν ′ i Δ C ( x ; x ′ ) = − ( D − 2 ) D 2 k H 2 g μν = − 32 π 2 H 4 g μν + O ( D − 4 ) . (50)
Three facts emerge concerning these coincidence limits (44-51):
Only the undifferentiated A A A -type propagator (44) is divergent;
The only nonzero, singly differentiated propagator is ∂ μ i Δ A \partial_\mu i\Delta_A ∂ μ i Δ A in expression (47), and only the μ = 0 \mu=0 μ = 0 component is nonzero; and
Second derivatives are are proportional to the Minkowski metric,
lim x ′ → x ∂ μ ∂ ν ′ i Δ I ( x ; x ′ ) = 1 D η μ ν × lim x ′ → x ∂ ⋅ ∂ ′ i Δ I ( x ; x ′ ) . (51) \lim_{x'\to x}\partial_\mu\partial'_\nu i\Delta_I(x;x')=\frac{1}{D}\eta_{\mu\nu}\times\lim_{x'\to x}\partial\cdot\partial'i\Delta_I(x;x'). \tag*{(51)} x ′ → x lim ∂ μ ∂ ν ′ i Δ I ( x ; x ′ ) = D 1 η μν × x ′ → x lim ∂ ⋅ ∂ ′ i Δ I ( x ; x ′ ) . (51)
We compute the divergent contributions by hand. The various 4-point divergences are proportional to the constant K 4 \mathcal{K}_4 K 4 ,
K 4 ≡ H D − 4 ( 4 π ) D / 2 Γ ( D − 1 ) Γ ( D 2 ) π cot ( D π 2 ) = 1 4 π 2 H D − 4 D − 4 + Finite . (52) \mathcal{K}_4\equiv\frac{H^{D-4}}{(4\pi)^{D/2}}\frac{\Gamma(D-1)}{\Gamma\left(\frac{D}{2}\right)}\pi\cot\left(\frac{D\pi}{2}\right)=\frac{1}{4\pi^2}\frac{H^{D-4}}{D-4}+\text{Finite}. \tag*{(52)} K 4 ≡ ( 4 π ) D /2 H D − 4 Γ ( 2 D ) Γ ( D − 1 ) π cot ( 2 D π ) = 4 π 2 1 D − 4 H D − 4 + Finite . (52)
These divergences come from the undifferentiated A A A -type propagator (44),
i Δ A ( x ; x ) = − K 4 H 2 a D − 4 + O ( D − 4 ) . (53) i\Delta_A(x;x)=-\frac{\mathcal{K}_4H^2}{a^{D-4}}+O(D-4). \tag*{(53)} i Δ A ( x ; x ) = − a D − 4 K 4 H 2 + O ( D − 4 ) . (53)
Only the expectation value of the first term in (43) can produce an undifferentiated A A A -type propagator. There is such a contribution from each of the first four terms of the invariant Lagrangian (9), and their sum takes the form ∂ α { a D − 2 T α β μ ν ρ σ ∂ β i δ D ( x − x ′ ) } \partial_\alpha\{a^{D-2}T^{\alpha\beta\mu\nu\rho\sigma}\partial_\beta i\delta^D(x-x')\} ∂ α { a D − 2 T α β μν ρ σ ∂ β i δ D ( x − x ′ )} with,
T α β μ ν ρ σ = − g ~ { − g ~ α ( ρ g ~ σ ) ( μ g ~ ν ) β + 1 2 g ~ α ( ρ g ~ σ ) β g ~ μ ν + 1 2 g ~ ρ σ g ~ α ( μ g ~ ν ) β − 1 2 g ~ α β g ~ ρ σ g ~ μ ν + 1 2 g ~ α β g ~ ρ ( μ g ~ ν ) σ } . (54) T^{\alpha\beta\mu\nu\rho\sigma}=\sqrt{-\widetilde g}\left\{-\widetilde g^{\alpha(\rho}\widetilde g^{\sigma)(\mu}\widetilde g^{\nu)\beta}+\frac{1}{2}\widetilde g^{\alpha(\rho}\widetilde g^{\sigma)\beta}\widetilde g^{\mu\nu}+\frac{1}{2}\widetilde g^{\rho\sigma}\widetilde g^{\alpha(\mu}\widetilde g^{\nu)\beta}-\frac{1}{2}\widetilde g^{\alpha\beta}\widetilde g^{\rho\sigma}\widetilde g^{\mu\nu}+\frac{1}{2}\widetilde g^{\alpha\beta}\widetilde g^{\rho(\mu}\widetilde g^{\nu)\sigma}\right\}. \tag*{(54)} T α β μν ρ σ = − g { − g α ( ρ g σ ) ( μ g ν ) β + 2 1 g α ( ρ g σ ) β g μν + 2 1 g ρ σ g α ( μ g ν ) β − 2 1 g α β g ρ σ g μν + 2 1 g α β g ρ ( μ g ν ) σ } . (54)
The variation of the final term in (9) also contributes,
− ( D 2 − 1 ) ∂ α { a D − 1 H − g ~ g ~ μ ν g ~ α ( ρ δ 0 σ ) i δ D ( x − x ′ ) } + ( D 2 − 1 ) a D − 1 H − g ~ g ~ ρ σ g ~ α ( μ δ 0 ν ) ∂ α i δ D ( x − x ′ ) . (55) -\left(\frac{D}{2}-1\right)\partial_\alpha\left\{a^{D-1}H\sqrt{-\widetilde{g}}\,\widetilde{g}^{\mu\nu}\widetilde{g}^{\alpha(\rho}\delta^{\sigma)}_0 i\delta^D(x-x')\right\}
+\left(\frac{D}{2}-1\right)a^{D-1}H\sqrt{-\widetilde{g}}\,\widetilde{g}^{\rho\sigma}\widetilde{g}^{\alpha(\mu}\delta^{\nu)}_0\partial_\alpha i\delta^D(x-x'). \tag*{(55)} − ( 2 D − 1 ) ∂ α { a D − 1 H − g g μν g α ( ρ δ 0 σ ) i δ D ( x − x ′ ) } + ( 2 D − 1 ) a D − 1 H − g g ρ σ g α ( μ δ 0 ν ) ∂ α i δ D ( x − x ′ ) . (55)
The basic strategy for extracting the divergent part of the 4-point contribution is to first expand expressions (55) and (56) to second order in the graviton field, and then take the expectation values. The expansion involves the conformally transformed volume element and the inverse,
− g ~ = 1 + κ 2 h + κ 2 ( 1 8 h 2 − 1 4 h γ δ h γ δ ) + ⋯ , (56) \sqrt{-\widetilde{g}}=1+\frac{\kappa}{2}h+\kappa^2\left(\frac{1}{8}h^2-\frac{1}{4}h^{\gamma\delta}h_{\gamma\delta}\right)+\cdots, \tag*{(56)} − g = 1 + 2 κ h + κ 2 ( 8 1 h 2 − 4 1 h γ δ h γ δ ) + ⋯ , (56)
g ~ γ δ = η γ δ − κ h γ δ + κ 2 h γ ϵ h ϵ δ + ⋯ . (57) \widetilde{g}^{\gamma\delta}=\eta^{\gamma\delta}-\kappa h^{\gamma\delta}+\kappa^2h^{\gamma\epsilon}h^\delta_{\ \epsilon}+\cdots. \tag*{(57)} g γ δ = η γ δ − κ h γ δ + κ 2 h γ ϵ h ϵ δ + ⋯ . (57)
The expectation value of the κ 2 \kappa^2 κ 2 part of (56) is,
⟨ 1 8 h 2 − 1 4 h γ δ h γ δ ⟩ = − D ( D − 1 ) 2 i Δ A + Finite . (58) \left\langle\frac{1}{8}h^2-\frac{1}{4}h^{\gamma\delta}h_{\gamma\delta}\right\rangle=-\frac{D(D-1)}{2}i\Delta_A+\text{Finite}. \tag*{(58)} ⟨ 8 1 h 2 − 4 1 h γ δ h γ δ ⟩ = − 2 D ( D − 1 ) i Δ A + Finite . (58)
The total contribution from terms of this form is,
κ 2 H 2 D ( D − 1 ) K 4 8 { η μ ν ∂ ( ρ a 2 ∂ σ ) − 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν + η ρ σ ∂ μ a 2 ∂ ν − ( D − 2 ) η μ ν δ 0 ( ρ ∂ σ ) a 3 H + ( D − 2 ) a 3 H η ρ σ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) . (59) \frac{\kappa^2 H^2 D(D-1)K_4}{8}\left\{\eta^{\mu\nu}\partial_{(\rho}a^2\partial_{\sigma)}-2\eta^{\mu)(\rho}\partial_{\sigma)}a^2\partial^\nu+\eta^{\rho\sigma}\partial^\mu a^2\partial^\nu
-(D-2)\eta^{\mu\nu}\delta_0^{(\rho}\partial^{\sigma)}a^3H+(D-2)a^3H\eta^{\rho\sigma}\delta_0^{(\mu}\partial^{\nu)}\right\}i\delta^D(x-x'). \tag*{(59)} 8 κ 2 H 2 D ( D − 1 ) K 4 { η μν ∂ ( ρ a 2 ∂ σ ) − 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν + η ρ σ ∂ μ a 2 ∂ ν − ( D − 2 ) η μν δ 0 ( ρ ∂ σ ) a 3 H + ( D − 2 ) a 3 H η ρ σ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) . (59)
The order κ 2 \kappa^2 κ 2 part of (57) is always associated with the product of an order κ \kappa κ term from (56) and one from (57) ,
⟨ h γ ϵ h ϵ δ − 1 2 h γ δ ⟩ = D i Δ A η γ δ + Finite . (60) \left\langle h^{\gamma\epsilon}h_{\epsilon}^{\ \delta}-\frac{1}{2}h^{\gamma\delta}\right\rangle=Di\Delta_A\eta^{\gamma\delta}+\text{Finite}. \tag*{(60)} ⟨ h γ ϵ h ϵ δ − 2 1 h γ δ ⟩ = D i Δ A η γ δ + Finite . (60)
The total contribution from terms of this form is,
κ 2 H 2 D K 4 2 { 2 η μ ) ( ρ ∂ σ ) ‾ a 2 ∂ ν + 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν + 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν ‾ − η μ ν ∂ ρ a 2 ∂ σ − η μ ν [ ∂ ρ ‾ a 2 ∂ σ + ∂ ρ a 2 ∂ σ ‾ ] − η ρ σ ∂ μ a 2 ∂ ν − η ρ σ [ ∂ μ ‾ a 2 ∂ ν + ∂ μ a 2 ∂ ν ‾ ] + ( D − 2 ) η μ ν ‾ δ 0 ( ρ ∂ σ ) a 3 H − ( D − 2 ) a 3 H η ρ σ ‾ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) . (61) \frac{\kappa^2H^2DK_4}{2}\left\{2\eta^{\mu)(\rho}\overline{\partial^{\sigma)}}a^2\partial^\nu+2\eta^{\mu)(\rho}\partial^{\sigma)}a^2\partial^\nu+2\eta^{\mu)(\rho}\partial^{\sigma)}a^2\overline{\partial^\nu}-\eta^{\mu\nu}\partial^\rho a^2\partial^\sigma
-\eta^{\mu\nu}\left[\overline{\partial^\rho}a^2\partial^\sigma+\partial^\rho a^2\overline{\partial^\sigma}\right]-\eta^{\rho\sigma}\partial^\mu a^2\partial^\nu-\eta^{\rho\sigma}\left[\overline{\partial^\mu}a^2\partial^\nu+\partial^\mu a^2\overline{\partial^\nu}\right]
+(D-2)\overline{\eta^{\mu\nu}}\delta_0^{(\rho}\partial^{\sigma)}a^3H-(D-2)a^3H\overline{\eta^{\rho\sigma}}\delta_0^{(\mu}\partial^{\nu)}\right\}i\delta^D(x-x'). \tag*{(61)} 2 κ 2 H 2 D K 4 { 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν + 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν + 2 η μ ) ( ρ ∂ σ ) a 2 ∂ ν − η μν ∂ ρ a 2 ∂ σ − η μν [ ∂ ρ a 2 ∂ σ + ∂ ρ a 2 ∂ σ ] − η ρ σ ∂ μ a 2 ∂ ν − η ρ σ [ ∂ μ a 2 ∂ ν + ∂ μ a 2 ∂ ν ] + ( D − 2 ) η μν δ 0 ( ρ ∂ σ ) a 3 H − ( D − 2 ) a 3 H η ρ σ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) . (61)
That leaves products of two order κ \kappa κ terms from (57) . The total contribution from terms of this form is,
κ 2 H 2 K 4 D − 3 { − η μ ν [ ∂ ρ a 2 ∂ σ ‾ − ∂ ρ ‾ a 2 ∂ σ ] + η ρ σ [ ∂ μ a 2 ∂ ν ‾ − ∂ μ ‾ a 2 ∂ ν ] + 4 a 3 H δ 0 ( ρ η ‾ μ ) σ + 2 a 3 H η ‾ μ ν δ 0 ( ρ ∂ σ ) ‾ + 4 a 3 H δ 0 ( μ η ‾ ν ) ρ ∂ σ ) ‾ + 2 a 3 H η ‾ ρ σ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) + κ 2 H 2 ( D − 1 ) K 4 D − 3 { η ‾ μ ν ∂ ρ a 2 ∂ σ ‾ + 1 2 η μ ν ∂ ρ ‾ a 2 ∂ σ ‾ + η ρ σ ∂ μ ‾ a 2 ∂ ν + 1 2 η ρ σ ∂ μ ‾ a 2 ∂ ν − η ‾ μ ) ( ρ ∂ σ ) a 2 ∂ ν − η ‾ μ ) ( ρ a 2 ∂ ν − η μ ) ( ρ ∂ σ ) a 2 ∂ ν ‾ − 2 a 3 H δ 0 ( ρ η ‾ σ ) ( μ ∂ ν ) ‾ − 2 a 3 H δ 0 ( μ η ‾ ν ) ( ρ ∂ σ ) ‾ } i δ D ( x − x ′ ) . (62) \begin{aligned}
\frac{\kappa^2H^2K_4}{D-3}\left\{-\eta^{\mu\nu}\left[\partial^\rho a^2\overline{\partial^\sigma}-\overline{\partial^\rho}a^2\partial^\sigma\right]+\eta^{\rho\sigma}\left[\partial^\mu a^2\overline{\partial^\nu}-\overline{\partial^\mu}a^2\partial^\nu\right]+4a^3H\delta_0^{(\rho}\overline{\eta}^{\mu)\sigma}
+2a^3H\overline{\eta}^{\mu\nu}\delta_0^{(\rho}\overline{\partial^{\sigma)}}+4a^3H\delta_0^{(\mu}\overline{\eta}^{\nu)\rho}\overline{\partial^\sigma)}+2a^3H\overline{\eta}^{\rho\sigma}\delta_0^{(\mu}\partial^{\nu)}\right\}i\delta^D(x-x') \\
&+\frac{\kappa^2H^2(D-1)K_4}{D-3}\left\{\overline{\eta}^{\mu\nu}\partial^\rho a^2\overline{\partial^\sigma}+\frac{1}{2}\eta^{\mu\nu}\overline{\partial^\rho}a^2\overline{\partial^\sigma}+\eta^{\rho\sigma}\overline{\partial^\mu}a^2\partial^\nu+\frac{1}{2}\eta^{\rho\sigma}\overline{\partial^\mu}a^2\partial^\nu
-\overline{\eta}^{\mu)(\rho}\partial^{\sigma)}a^2\partial^\nu-\overline{\eta}^{\mu)(\rho}a^2\partial^\nu-\eta^{\mu)(\rho}\partial^{\sigma)}a^2\overline{\partial^\nu}
-2a^3H\delta_0^{(\rho}\overline{\eta}^{\sigma)(\mu}\overline{\partial^\nu)}-2a^3H\delta_0^{(\mu}\overline{\eta}^{\nu)(\rho}\overline{\partial^\sigma)}\right\}i\delta^D(x-x'). \tag*{(62)}
\end{aligned} D − 3 κ 2 H 2 K 4 { − η μν [ ∂ ρ a 2 ∂ σ − ∂ ρ a 2 ∂ σ ] + η ρ σ [ ∂ μ a 2 ∂ ν − ∂ μ a 2 ∂ ν ] + 4 a 3 H δ 0 ( ρ η μ ) σ + 2 a 3 H η μν δ 0 ( ρ ∂ σ ) + 4 a 3 H δ 0 ( μ η ν ) ρ ∂ σ ) + 2 a 3 H η ρ σ δ 0 ( μ ∂ ν ) } i δ D ( x − x ′ ) + D − 3 κ 2 H 2 ( D − 1 ) K 4 { η μν ∂ ρ a 2 ∂ σ + 2 1 η μν ∂ ρ a 2 ∂ σ + η ρ σ ∂ μ a 2 ∂ ν + 2 1 η ρ σ ∂ μ a 2 ∂ ν − η μ ) ( ρ ∂ σ ) a 2 ∂ ν − η μ ) ( ρ a 2 ∂ ν − η μ ) ( ρ ∂ σ ) a 2 ∂ ν − 2 a 3 H δ 0 ( ρ η σ ) ( μ ∂ ν ) − 2 a 3 H δ 0 ( μ η ν ) ( ρ ∂ σ ) } i δ D ( x − x ′ ) . (62)
Table 2 gives all the 4-point contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\!^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) . This includes the divergent contributions (60), (62) and (63) from undifferentiated A A A -type propagators, as well as finite contributions from the B B B -type and C C C -type propagators, and from differentiated propagators. Note that we have represented functions of D D D times K 4 K_4 K 4 as pure numbers times K 4 K_4 K 4 plus finite contributions,
Table 2. Primitive 4-point contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
f ( D ) × K 4 ⟶ f ( 4 ) × K 4 + f ′ ( 4 ) 4 π 2 . (63) f(D)\times K_4\longrightarrow f(4)\times K_4+\frac{f'(4)}{4\pi^2}. \tag*{(63)} f ( D ) × K 4 ⟶ f ( 4 ) × K 4 + 4 π 2 f ′ ( 4 ) . (63)
Only those coefficients not determined by reflection are given; the results for i = 4 , 6 , 8 , 10 , 15 , 17 i=4, 6, 8, 10, 15, 17 i = 4 , 6 , 8 , 10 , 15 , 17 and 20 20 20 follow from relations (38–39).
The Ghost Contribution
A more complicated 1-loop contribution to the self-energy comes from the ghost loop depicted in Figure 2 . The analytic expression is,
Figure 2. Contribution to the graviton self-energy from the product of two c c ˉ c\bar c c c ˉ vertices (65). Wavy lines stand for gravitons and dashed lines stand for ghosts.
− i [ Σ g h μ ν ρ σ ] ( x ; x ′ ) = ⟨ Ω ∣ i δ S g h δ h μ ν ( x ) × i δ S g h δ h ρ σ ( x ′ ) ∣ Ω ⟩ (64) -i[\Sigma_{\rm gh}^{\mu\nu\rho\sigma}](x;x')=\left\langle\Omega\left|\frac{i\delta S_{\rm gh}}{\delta h_{\mu\nu}(x)}\times\frac{i\delta S_{\rm gh}}{\delta h_{\rho\sigma}(x')}\right|\Omega\right\rangle \tag*{(64)} − i [ Σ gh μν ρ σ ] ( x ; x ′ ) = ⟨ Ω δ h μν ( x ) i δ S gh × δ h ρ σ ( x ′ ) i δ S gh Ω ⟩ (64)
Varying the ghost action (11) gives,
i δ S g h δ h μ ν = − i κ a D − 2 [ c ˉ , α ( μ c ν ) , α + i κ ∂ α [ a D − 2 c ˉ , ν ) ( μ c α ] − 2 i κ H a D − 1 c ˉ , ν ) ( μ c 0 + i κ ( a D − 2 c ˉ α ) , α c ( μ , ν ) + i κ η μ ν { − 1 2 ∂ α [ ( a D − 2 c ˉ β ) , β c α ] + a H ( a D − 2 c ˉ α ) , α c 0 } ] . (65) \frac{i\delta S_{\rm gh}}{\delta h_{\mu\nu}}=-i\kappa aD^{-2}\left[\bar c^{(\mu}_{,\alpha}c^{\nu),\alpha}+i\kappa\partial_\alpha\left[a^{D-2}\bar c^{(\mu}_{,\nu)}c^\alpha\right]-2i\kappa HaD^{-1}\bar c^{(\mu}_{,\nu)}c^0+i\kappa\left(a^{D-2}\bar c^\alpha\right)_{,\alpha}c^{(\mu,\nu)}+i\kappa\eta^{\mu\nu}\left\{-\frac{1}{2}\partial_\alpha\left[\left(a^{D-2}\bar c^\beta\right)_{,\beta}c^\alpha\right]+aH\left(a^{D-2}\bar c^\alpha\right)_{,\alpha}c^0\right\}\right]. \tag*{(65)} δ h μν i δ S gh = − iκa D − 2 [ c ˉ , α ( μ c ν ) , α + iκ ∂ α [ a D − 2 c ˉ , ν ) ( μ c α ] − 2 iκ H a D − 1 c ˉ , ν ) ( μ c 0 + iκ ( a D − 2 c ˉ α ) , α c ( μ , ν ) + iκ η μν { − 2 1 ∂ α [ ( a D − 2 c ˉ β ) , β c α ] + a H ( a D − 2 c ˉ α ) , α c 0 } ] . (65)
The various interactions are listed in Table 2 of [22 ] for D = 4 D=4 D = 4 , and for arbitrary D D D in Table 2 of [29 ] .
It is useful to reduce the contribution from the product of two vertices. Taking account of the Fermi minus sign, and suppressing the cumbersome symmetrizations, the product of the 1st vertex at x μ x^\mu x μ and the 4th vertex at x ′ μ x'^\mu x ′ μ gives,
− i κ a D − 2 ⋅ ∂ α c μ ( x ) ⋅ ∂ α c ν ( x ) × − 2 i κ H a ′ D − 1 ⋅ ∂ ρ c ˉ σ ( x ′ ) ⋅ c 0 ( x ′ ) ⟶ + 2 κ 2 a D − 2 a ′ D − 1 ∂ α [ μ Δ 0 ] ( x ; x ′ ) × ∂ α ∂ ι ′ [ σ Δ ν ] ( x ′ ; x ) . (66) -i\kappa aD^{-2}\cdot\partial^\alpha c^\mu(x)\cdot\partial_\alpha c^\nu(x)\times-2i\kappa Ha' D^{-1}\cdot\partial^\rho\bar c^\sigma(x')\cdot c^0(x')
\longrightarrow+2\kappa^2a^{D-2}a'^{D-1}\partial_\alpha\left[\mu\Delta^0\right](x;x')\times\partial_\alpha\partial'_\iota\left[\sigma\Delta^\nu\right](x';x). \tag*{(66)} − iκa D − 2 ⋅ ∂ α c μ ( x ) ⋅ ∂ α c ν ( x ) × − 2 iκ H a ′ D − 1 ⋅ ∂ ρ c ˉ σ ( x ′ ) ⋅ c 0 ( x ′ ) ⟶ + 2 κ 2 a D − 2 a ′ D − 1 ∂ α [ μ Δ 0 ] ( x ; x ′ ) × ∂ α ∂ ι ′ [ σ Δ ν ] ( x ′ ; x ) . (66)
Substituting for the ghost-anti-ghost propagator (13), and restoring the symmetrizations over μ ν \mu\nu μν and ρ σ \rho\sigma ρ σ , gives,
− 2 κ 2 H a D − 2 a ′ D − 1 ∂ α i Δ B ( x ; x ′ ) × { δ 0 ( μ η ν ) ( ρ ∂ α i Δ A ( x ′ ; x ) − δ 0 μ δ 0 ν δ 0 ( ρ ∂ α i Δ B ( x ′ ; x ) } . (67) -2\kappa^2Ha^{D-2}a'^{D-1}\partial_\alpha i\Delta_B(x;x')\times\left\{\delta^{(\mu}_{0}\eta^{\nu)(\rho}\partial_\alpha i\Delta_A(x';x)-\delta^\mu_0\delta^\nu_0\delta^{(\rho}_0\partial_\alpha i\Delta_B(x';x)\right\}. \tag*{(67)} − 2 κ 2 H a D − 2 a ′ D − 1 ∂ α i Δ B ( x ; x ′ ) × { δ 0 ( μ η ν ) ( ρ ∂ α i Δ A ( x ′ ; x ) − δ 0 μ δ 0 ν δ 0 ( ρ ∂ α i Δ B ( x ′ ; x ) } . (67)
= − 2 κ 2 H a D − 2 a ′ D − 1 ∂ α i Δ B × { δ 0 ( μ η ν ) ( ρ ∂ α i Δ A + δ 0 μ δ 0 ν δ 0 ( ρ ∂ α [ i Δ A − i Δ B ] } . (68) =-2\kappa^2Ha^{D-2}a'^{D-1}\partial_\alpha i\Delta_B\times\left\{\delta^{(\mu}_{0}\eta^{\nu)(\rho}\partial_\alpha i\Delta_A+\delta^\mu_0\delta^\nu_0\delta^{(\rho}_0\partial_\alpha[i\Delta_A-i\Delta_B]\right\}. \tag*{(68)} = − 2 κ 2 H a D − 2 a ′ D − 1 ∂ α i Δ B × { δ 0 ( μ η ν ) ( ρ ∂ α i Δ A + δ 0 μ δ 0 ν δ 0 ( ρ ∂ α [ i Δ A − i Δ B ] } . (68)
Expression (69) implies that we need expansions for two products of differentiated propagators,
( a a ′ ) D − 2 ∂ γ ∂ σ i Δ A × ∂ α i Δ B and ( a a ′ ) D − 2 ∂ γ ∂ σ [ i Δ A − i Δ B ] × ∂ α i Δ B . (69) (aa^{\prime})^{D-2}\partial^{\gamma}\partial^{\sigma}i\Delta_{A}\times\partial_{\alpha}i\Delta_{B}\quad\text{and}\quad(aa^{\prime})^{D-2}\partial^{\gamma}\partial^{\sigma}[i\Delta_{A}-i\Delta_{B}]\times\partial_{\alpha}i\Delta_{B}. \tag*{(69)} ( a a ′ ) D − 2 ∂ γ ∂ σ i Δ A × ∂ α i Δ B and ( a a ′ ) D − 2 ∂ γ ∂ σ [ i Δ A − i Δ B ] × ∂ α i Δ B . (69)
We can read these expansions off from Table 19 and Table 20 of the Appendix (which is section 7). Note first that we need only a handful of terms from the expansion (147) of each propagator,
Table 19. Local contributions from products of the form ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν i Δ J (aa')^{D-2}\partial_\mu\partial'_\rho i\Delta_I\times\partial_\nu i\Delta_J ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν i Δ J .
Table 20. Nonlocal contributions from products of the form ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν i Δ J (aa')D^{-2}\partial_\mu\partial'_\rho i\Delta_I\times\partial_\nu i\Delta_J ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν i Δ J .
i Δ A = i Δ c f + A 0 + not needed , (70) i\Delta_{A}=i\Delta_{\mathrm{cf}}+A_{0}+\text{not needed}, \tag*{(70)} i Δ A = i Δ cf + A 0 + not needed , (70)
i Δ B = i Δ c f + B 0 + not needed , (71) i\Delta_{B}=i\Delta_{\mathrm{cf}}+B_{0}+\text{not needed}, \tag*{(71)} i Δ B = i Δ cf + B 0 + not needed , (71)
i Δ A − i Δ B = A 0 + not needed , (72) i\Delta_{A}-i\Delta_{B}=A_{0}+\text{not needed}, \tag*{(72)} i Δ A − i Δ B = A 0 + not needed , (72)
where i Δ c f i\Delta_{\mathrm{cf}} i Δ cf , A 0 A_0 A 0 and B 0 B_0 B 0 are defined by expressions (147) , (148) and (150) , respectively. From Table 19 and Table 20 we read,
( a a ′ ) D − 2 ∂ α ∂ σ [ i Δ A − i Δ B ] × ∂ α i Δ B = ( a a ′ ) D − 2 ∂ α ∂ σ [ A 0 + ⋯ ] × ∂ α [ i Δ c f + ⋯ ] , (73) (aa')^{D-2}\partial^\alpha\partial^{\sigma}\left[i\Delta_A-i\Delta_B\right]\times\partial_\alpha i\Delta_B
= (aa')^{D-2}\partial^\alpha\partial^{\sigma}\left[A_0+\cdots\right]\times\partial_\alpha\left[i\Delta_{\mathrm{cf}}+\cdots\right], \tag*{(73)} ( a a ′ ) D − 2 ∂ α ∂ σ [ i Δ A − i Δ B ] × ∂ α i Δ B = ( a a ′ ) D − 2 ∂ α ∂ σ [ A 0 + ⋯ ] × ∂ α [ i Δ cf + ⋯ ] , (73)
⟶ a a ′ H 2 K 3 { − D ( D − 2 ) ( D 2 − 2 D − 2 ) 8 ( D − 1 ) ∂ σ + D ( D − 2 ) 2 8 a H δ 0 σ } i δ D ( x − x ′ ) + a a ′ H 2 [ ∂ σ − a H δ 0 σ ] ∂ 4 2 10 ⋅ π 4 ln 2 ( 1 4 H 2 Δ x 2 ) − a a ′ H 2 [ ( 5 ∂ σ − 12 a ′ H δ 0 σ ) ∂ 2 + 6 a H ∂ 0 ∂ σ ] ∂ 2 2 9 ⋅ 3 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) , (74) \begin{aligned}
&\longrightarrow aa'H^2\mathcal{K}_3\left\{-\frac{D(D-2)(D^2-2D-2)}{8(D-1)}\partial^\sigma+\frac{D(D-2)^2}{8}aH\delta^\sigma_0\right\}i\delta^D(x-x')\\
&\quad+\frac{aa'H^2[\partial^\sigma-aH\delta^\sigma_0]\partial^4}{2^{10}\cdot\pi^4}\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)\\
&\quad-\frac{aa'H^2[(5\partial^\sigma-12a'H\delta^\sigma_0)\partial^2+6aH\partial_0\partial^\sigma]\partial^2}{2^9\cdot3\cdot\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right), \tag*{(74)}
\end{aligned} ⟶ a a ′ H 2 K 3 { − 8 ( D − 1 ) D ( D − 2 ) ( D 2 − 2 D − 2 ) ∂ σ + 8 D ( D − 2 ) 2 a H δ 0 σ } i δ D ( x − x ′ ) + 2 10 ⋅ π 4 a a ′ H 2 [ ∂ σ − a H δ 0 σ ] ∂ 4 ln 2 ( 4 1 H 2 Δ x 2 ) − 2 9 ⋅ 3 ⋅ π 4 a a ′ H 2 [( 5 ∂ σ − 12 a ′ H δ 0 σ ) ∂ 2 + 6 a H ∂ 0 ∂ σ ] ∂ 2 ln ( 4 1 H 2 Δ x 2 ) , (74)
where the constant K 3 \mathcal{K}_3 K 3 is,
K 3 ≡ H D − 4 ( 4 π ) D / 2 Γ ( D 2 − 1 ) 2 ( D − 3 ) ( D − 4 ) = 1 32 π 2 H D − 4 D − 4 + F i n i t e . (75) \mathcal{K}_3\equiv\frac{H^{D-4}}{(4\pi)^{D/2}}\frac{\Gamma\left(\frac{D}{2}-1\right)}{2(D-3)(D-4)}
=\frac{1}{32\pi^2}\frac{H^{D-4}}{D-4}+\mathrm{Finite}. \tag*{(75)} K 3 ≡ ( 4 π ) D /2 H D − 4 2 ( D − 3 ) ( D − 4 ) Γ ( 2 D − 1 ) = 32 π 2 1 D − 4 H D − 4 + Finite . (75)
Putting everything together for the second part of expression (68) , and consulting Table 1 of basis tensors gives,
− 2 κ 2 a ′ H × δ 0 σ δ 0 ν δ 0 ρ ( a a ′ ) D − 2 ∂ σ ∂ α [ i Δ A − i Δ B ] × ∂ α i Δ B -2\kappa^2a'H\times\delta^\sigma_0\delta^\nu_0\delta^\rho_0(aa')^{D-2}\partial^\sigma\partial^\alpha\left[i\Delta_A-i\Delta_B\right]\times\partial_\alpha i\Delta_B − 2 κ 2 a ′ H × δ 0 σ δ 0 ν δ 0 ρ ( a a ′ ) D − 2 ∂ σ ∂ α [ i Δ A − i Δ B ] × ∂ α i Δ B
⟶ [ − 4 K 3 + 5 2 5 ⋅ π 2 ] ( a a ′ ) 2 H 4 × κ 2 [ μ ν D 13 ρ σ ] i δ D ( x − x ′ ) + [ 4 K 3 + 17 2 5 ⋅ 3 ⋅ π 2 ] a a ′ 2 H 3 × κ 2 [ μ ν D 14 ρ σ ] i δ D ( x − x ′ ) + ( a a ′ ) 2 H 4 ∂ 4 2 9 ⋅ π 4 × κ 2 [ μ ν D 13 ρ σ ] ln 2 ( 1 4 H 2 Δ x 2 ) − a a ′ 2 H 3 ∂ 4 2 9 ⋅ π 4 × κ 2 [ μ ν D 14 ρ σ ] ln 2 ( 1 4 H 2 Δ x 2 ) − a a ′ 3 H 3 ∂ 2 2 6 ⋅ π 4 × κ 2 [ μ ν D 13 ρ σ ] ln ( 1 4 H 2 Δ x 2 ) + a a ′ 2 H 3 [ 5 ∂ 4 + 6 a H ∂ 0 ∂ 2 ] 2 8 ⋅ 3 ⋅ π 4 × κ 2 [ μ ν D 14 ρ σ ] ln ( 1 4 H 2 Δ x 2 ) . (76) \begin{aligned}
\longrightarrow{}&\left[-4\mathcal{K}_3+\frac{5}{2^5\cdot\pi^2}\right](aa')^2H^4\times\kappa^2[\mu\nu D^{\rho\sigma}_{13}]i\delta^D(x-x')\\
&+\left[4\mathcal{K}_3+\frac{17}{2^5\cdot3\cdot\pi^2}\right]aa'^2H^3\times\kappa^2[\mu\nu D^{\rho\sigma}_{14}]i\delta^D(x-x')\\
&+\frac{(aa')^2H^4\partial^4}{2^9\cdot\pi^4}\times\kappa^2[\mu\nu D^{\rho\sigma}_{13}]\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)\\
&-\frac{aa'^2H^3\partial^4}{2^9\cdot\pi^4}\times\kappa^2[\mu\nu D^{\rho\sigma}_{14}]\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)\\
&-\frac{aa'^3H^3\partial^2}{2^6\cdot\pi^4}\times\kappa^2[\mu\nu D^{\rho\sigma}_{13}]\ln\left(\frac{1}{4}H^2\Delta x^2\right)\\
&+\frac{aa'^2H^3[5\partial^4+6aH\partial_0\partial^2]}{2^8\cdot3\cdot\pi^4}\times\kappa^2[\mu\nu D^{\rho\sigma}_{14}]\ln\left(\frac{1}{4}H^2\Delta x^2\right). \tag*{(76)}
\end{aligned} ⟶ [ − 4 K 3 + 2 5 ⋅ π 2 5 ] ( a a ′ ) 2 H 4 × κ 2 [ μν D 13 ρ σ ] i δ D ( x − x ′ ) + [ 4 K 3 + 2 5 ⋅ 3 ⋅ π 2 17 ] a a ′2 H 3 × κ 2 [ μν D 14 ρ σ ] i δ D ( x − x ′ ) + 2 9 ⋅ π 4 ( a a ′ ) 2 H 4 ∂ 4 × κ 2 [ μν D 13 ρ σ ] ln 2 ( 4 1 H 2 Δ x 2 ) − 2 9 ⋅ π 4 a a ′2 H 3 ∂ 4 × κ 2 [ μν D 14 ρ σ ] ln 2 ( 4 1 H 2 Δ x 2 ) − 2 6 ⋅ π 4 a a ′3 H 3 ∂ 2 × κ 2 [ μν D 13 ρ σ ] ln ( 4 1 H 2 Δ x 2 ) + 2 8 ⋅ 3 ⋅ π 4 a a ′2 H 3 [ 5 ∂ 4 + 6 a H ∂ 0 ∂ 2 ] × κ 2 [ μν D 14 ρ σ ] ln ( 4 1 H 2 Δ x 2 ) . (76)
Table 3 gives the local contributions from the ghost loop and Table 5 gives the finite, nonlocal contributions.
Table 3. Primitive local ghost loop contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x^{\prime}) − i [ μν Σ ρ σ ] ( x ; x ′ ) .
Table 5. Primitive nonlocal ghost loop contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
The 3-Point Contribution
The most complicated 1-loop contribution to the self-energy comes from the product of two 3-point vertices,
− i [ μ ν ρ σ 3 p t ] ( x ; x ′ ) = ⟨ Ω ∣ [ i δ S δ h μ ν ( x ) ] h 2 × [ i δ S δ h ρ σ ( x ′ ) ] h 2 ∣ Ω ⟩ (77) -i[\mu\nu\rho\sigma\mathord{\!}_{3\mathrm{pt}}](x;x')=\left\langle\Omega\left|\left[\frac{i\delta S}{\delta h_{\mu\nu}(x)}\right]_{h^2}\times\left[\frac{i\delta S}{\delta h_{\rho\sigma}(x')}\right]_{h^2}\right|\Omega\right\rangle \tag*{(77)} − i [ μν ρ σ 3 pt ] ( x ; x ′ ) = ⟨ Ω [ δ h μν ( x ) i δ S ] h 2 × [ δ h ρ σ ( x ′ ) i δ S ] h 2 Ω ⟩ (77)
As before, the subscript “h 2 h^2 h 2 ” indicates that two graviton fields are retained after the variation. The diagramatic representation of expression (77) is depicted in Figure 3 .
Figure 3. Contribution to the graviton self-energy from the product of two 3-point vertices (78). Wavy lines stand for gravitons.
The distinguished 3-point vertex occurring in expression (78) is,
δ S δ h μ ν = ∂ α { a D − 2 − g ~ [ − g ~ α γ g ~ δ ( μ g ~ ν ) β + 1 2 g ~ α γ g ~ β δ g ~ μ ν + 1 2 g ~ α ( μ g ~ ν ) β g ~ γ δ − 1 2 g ~ α β g ~ γ δ g ~ μ ν + 1 2 g ~ α β g ~ γ ( μ g ~ ν ) δ ] h γ δ , β } + a D − 2 ∂ ∂ h μ ν { − g ~ g ~ α β g ~ γ δ g ~ κ λ } × [ 1 2 h α γ , κ h λ δ , β − 1 2 h α β , κ h λ γ , δ + 1 4 h α β , κ h γ δ , λ − 1 4 h α γ , κ h β δ , λ ] − ( D 2 − 1 ) ∂ α { a D − 1 H − g ~ g ~ μ ν g ~ α β h β 0 } + ( D 2 − 1 ) a D − 1 H − g ~ g ~ α γ g ~ δ ( μ δ 0 ν ) h γ δ , α + ( D 2 − 1 ) a D − 1 H ∂ ∂ h μ ν { − g ~ g ~ α β g ~ γ δ } h α β , γ h δ 0 . (78) \frac{\delta S}{\delta h_{\mu\nu}}=\partial_\alpha\left\{a^{D-2}\sqrt{-\widetilde{g}}\left[-\widetilde{g}^{\alpha\gamma}\widetilde{g}^{\delta(\mu}\widetilde{g}^{\nu)\beta}+\frac{1}{2}\widetilde{g}^{\alpha\gamma}\widetilde{g}^{\beta\delta}\widetilde{g}^{\mu\nu}+\frac{1}{2}\widetilde{g}^{\alpha(\mu}\widetilde{g}^{\nu)\beta}\widetilde{g}^{\gamma\delta}-\frac{1}{2}\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\mu\nu}+\frac{1}{2}\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma(\mu}\widetilde{g}^{\nu)\delta}\right]h_{\gamma\delta,\beta}\right\}
+a^{D-2}\frac{\partial}{\partial h_{\mu\nu}}\left\{\sqrt{-\widetilde{g}}\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\widetilde{g}^{\kappa\lambda}\right\}
\times\left[\frac{1}{2}h_{\alpha\gamma,\kappa}h_{\lambda\delta,\beta}-\frac{1}{2}h_{\alpha\beta,\kappa}h_{\lambda\gamma,\delta}+\frac{1}{4}h_{\alpha\beta,\kappa}h_{\gamma\delta,\lambda}-\frac{1}{4}h_{\alpha\gamma,\kappa}h_{\beta\delta,\lambda}\right]
-\left(\frac{D}{2}-1\right)\partial_\alpha\left\{a^{D-1}H\sqrt{-\widetilde{g}}\widetilde{g}^{\mu\nu}\widetilde{g}^{\alpha\beta}h_{\beta0}\right\}
+\left(\frac{D}{2}-1\right)a^{D-1}H\sqrt{-\widetilde{g}}\widetilde{g}^{\alpha\gamma}\widetilde{g}^{\delta(\mu}\delta^{\nu)}_{0}h_{\gamma\delta,\alpha}
+\left(\frac{D}{2}-1\right)a^{D-1}H\frac{\partial}{\partial h_{\mu\nu}}\left\{\sqrt{-\widetilde{g}}\widetilde{g}^{\alpha\beta}\widetilde{g}^{\gamma\delta}\right\}h_{\alpha\beta,\gamma}h_{\delta0}. \tag*{(78)} δ h μν δ S = ∂ α { a D − 2 − g [ − g α γ g δ ( μ g ν ) β + 2 1 g α γ g β δ g μν + 2 1 g α ( μ g ν ) β g γ δ − 2 1 g α β g γ δ g μν + 2 1 g α β g γ ( μ g ν ) δ ] h γ δ , β } + a D − 2 ∂ h μν ∂ { − g g α β g γ δ g κλ } × [ 2 1 h α γ , κ h λ δ , β − 2 1 h α β , κ h λγ , δ + 4 1 h α β , κ h γ δ , λ − 4 1 h α γ , κ h β δ , λ ] − ( 2 D − 1 ) ∂ α { a D − 1 H − g g μν g α β h β 0 } + ( 2 D − 1 ) a D − 1 H − g g α γ g δ ( μ δ 0 ν ) h γ δ , α + ( 2 D − 1 ) a D − 1 H ∂ h μν ∂ { − g g α β g γ δ } h α β , γ h δ 0 . (78)
Expanding this to second order in the graviton field naively produces 45 terms, but some of these are the same so that only 42 are independent. They can be found for D = 4 D=4 D = 4 in Table 3 of [28 ] , and Table 1 of [22 ] .4
The procedure for reducing expression (78) is much the same as (65), except there are 42 2 = 1764 42^2=1764 4 2 2 = 1764 graviton vertex pairs instead of the 10 2 = 100 10^2=100 1 0 2 = 100 ghost vertex pairs. It is instructive to consider a single vertex pair,
⟨ i κ 2 ∂ α { a D − 2 η μ ν h α β ( x ) h , β ( x ) } × i ( D − 2 ) κ 2 ∂ γ ′ { a ′ D − 1 H h ρ σ ( x ′ ) h γ γ ( x ′ ) } ⟩ = − ( D − 2 4 ) κ 2 H η μ ν ∂ α ∂ γ ′ { a D − 2 a ′ D − 1 [ i Δ α β ρ σ ( x ; x ′ ) × ∂ β i Δ 0 δ ( x ; x ′ ) + i [ Δ α β Δ 0 γ δ ] ( x ; x ′ ) × ∂ β i [ Δ ρ σ ] ( x ; x ′ ) ] } . (79) \left\langle\frac{i\kappa}{2}\partial_\alpha\left\{a^{D-2}\eta^{\mu\nu}h^{\alpha\beta}(x)h_{,\beta}(x)\right\}\times\frac{i(D-2)\kappa}{2}\partial_{\gamma}'\left\{a'^{D-1}Hh^{\rho\sigma}(x')h_{\gamma}^{\ \gamma}(x')\right\}\right\rangle
=-\left(\frac{D-2}{4}\right)\kappa^2H\eta^{\mu\nu}\partial_\alpha\partial_{\gamma}'\left\{a^{D-2}a'^{D-1}\left[i\Delta^{\alpha\beta\rho\sigma}(x;x')\times\partial_\beta i\Delta_{0}^{\ \delta}(x;x')+i\left[\Delta^{\alpha\beta}\Delta_{0}^{\gamma\delta}\right](x;x')\times\partial_\beta i\left[\Delta^{\rho\sigma}\right](x;x')\right]\right\}. \tag*{(79)} ⟨ 2 iκ ∂ α { a D − 2 η μν h α β ( x ) h , β ( x ) } × 2 i ( D − 2 ) κ ∂ γ ′ { a ′ D − 1 H h ρ σ ( x ′ ) h γ γ ( x ′ ) } ⟩ = − ( 4 D − 2 ) κ 2 H η μν ∂ α ∂ γ ′ { a D − 2 a ′ D − 1 [ i Δ α β ρ σ ( x ; x ′ ) × ∂ β i Δ 0 δ ( x ; x ′ ) + i [ Δ α β Δ 0 γ δ ] ( x ; x ′ ) × ∂ β i [ Δ ρ σ ] ( x ; x ′ ) ] } . (79)
The next step is to insert the expansion (12) of the graviton propagators in terms if i Δ A i\Delta_A i Δ A , i Δ B i\Delta_B i Δ B and i Δ C i\Delta_C i Δ C , and then expand each of the scalar propagators (147) to the order we require. Because the terms inside the curly brackets of expression (80) involve just a single derivative, we see from Table 15 that
Table 15. Contributions from products of the form ( a a ′ ) D − 2 ∂ μ i Δ I × i Δ J (aa')^{D-2}\partial_\mu i\Delta_I\times i\Delta_J ( a a ′ ) D − 2 ∂ μ i Δ I × i Δ J .
only i Δ c f i\Delta_{cf} i Δ c f and A 0 A_0 A 0 are needed. The result is,
{ ⋯ } ⟶ − 4 δ 0 δ a ′ D − 2 { 2 η α ( ρ η σ ) β − 2 η α β η ρ σ D − 2 } × ( a a ′ ) D − 2 i Δ c f ∂ β i Δ c f − 4 η ρ σ a ′ D − 2 × { 2 δ 0 ( α η β ) γ − 2 η α β δ 0 γ D − 2 } × ( a a ′ ) D − 2 i Δ c f ∂ β i Δ c f − 4 δ 0 δ a ′ D − 2 { 2 η α ( ρ η σ ) β − 2 η α β η ρ σ D − 3 } × ( a a ′ ) D − 2 A 0 ∂ β i Δ c f − 4 η ρ σ a ′ D − 3 { 2 δ 0 ( α η β ) γ − 2 η α β δ 0 γ D − 3 } × ( a a ′ ) D − 2 i Δ c f ∂ β A 0 . (80) \begin{aligned}
\{\cdots\}\longrightarrow{}&-\frac{4\delta_0^\delta a'}{D-2}\left\{2\eta^{\alpha(\rho}\eta^{\sigma)\beta}-\frac{2\eta^{\alpha\beta}\eta^{\rho\sigma}}{D-2}\right\}\times(aa')^{D-2}i\Delta_{cf}\,\partial_\beta i\Delta_{cf}-\frac{4\eta^{\rho\sigma}a'}{D-2}\\
&\times\left\{2\delta_0^{(\alpha}\eta^{\beta)\gamma}-\frac{2\eta^{\alpha\beta}\delta_0^\gamma}{D-2}\right\}\times(aa')^{D-2}i\Delta_{cf}\,\partial_\beta i\Delta_{cf}-\frac{4\delta_0^\delta a'}{D-2}\left\{2\eta^{\alpha(\rho}\eta^{\sigma)\beta}-\frac{2\eta^{\alpha\beta}\eta^{\rho\sigma}}{D-3}\right\}\\
&\times(aa')^{D-2}A_0\,\partial_\beta i\Delta_{cf}-\frac{4\eta^{\rho\sigma}a'}{D-3}\left\{2\delta_0^{(\alpha}\eta^{\beta)\gamma}-\frac{2\eta^{\alpha\beta}\delta_0^\gamma}{D-3}\right\}\times(aa')^{D-2}i\Delta_{cf}\,\partial_\beta A_0 . \tag*{(80)}
\end{aligned} { ⋯ } ⟶ − D − 2 4 δ 0 δ a ′ { 2 η α ( ρ η σ ) β − D − 2 2 η α β η ρ σ } × ( a a ′ ) D − 2 i Δ c f ∂ β i Δ c f − D − 2 4 η ρ σ a ′ × { 2 δ 0 ( α η β ) γ − D − 2 2 η α β δ 0 γ } × ( a a ′ ) D − 2 i Δ c f ∂ β i Δ c f − D − 2 4 δ 0 δ a ′ { 2 η α ( ρ η σ ) β − D − 3 2 η α β η ρ σ } × ( a a ′ ) D − 2 A 0 ∂ β i Δ c f − D − 3 4 η ρ σ a ′ { 2 δ 0 ( α η β ) γ − D − 3 2 η α β δ 0 γ } × ( a a ′ ) D − 2 i Δ c f ∂ β A 0 . (80)
We then read off the relevant products from Table 15 and reduce the result to standard form. The contribution from the final term of expression (80) is,
− D − 2 4 κ 2 H η μ ν ∂ α ∂ γ ′ × − 4 η ρ σ a ′ D − 3 { 2 δ 0 ( α η β ) γ − 2 η α β δ 0 γ D − 2 } × ( a a ′ ) D − 2 i Δ c f ∂ β A 0 -\frac{D-2}{4}\kappa^2H\eta^{\mu\nu}\partial_\alpha\partial'_\gamma\times-\frac{4\eta^{\rho\sigma}a'}{D-3}\left\{2\delta_0^{(\alpha}\eta^{\beta)\gamma}-\frac{2\eta^{\alpha\beta}\delta_0^\gamma}{D-2}\right\}\times(aa')^{D-2}i\Delta_{cf}\,\partial_\beta A_0 − 4 D − 2 κ 2 H η μν ∂ α ∂ γ ′ × − D − 3 4 η ρ σ a ′ { 2 δ 0 ( α η β ) γ − D − 2 2 η α β δ 0 γ } × ( a a ′ ) D − 2 i Δ c f ∂ β A 0
⟶ 2 κ 2 H η μ ν η ρ σ ∂ α ∂ γ ′ [ δ 0 α η β γ + δ 0 β η α γ − η α β δ 0 γ ] a a ′ 2 H 2 ∂ β δ 0 2 2 7 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) , (81) \longrightarrow2\kappa^2H\eta^{\mu\nu}\eta^{\rho\sigma}\partial_\alpha\partial'_\gamma\left[\delta_0^\alpha\eta^{\beta\gamma}+\delta_0^\beta\eta^{\alpha\gamma}-\eta^{\alpha\beta}\delta_0^\gamma\right]\frac{aa'^2H^2\partial_\beta\delta_0^2}{2^7\cdot\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right), \tag*{(81)} ⟶ 2 κ 2 H η μν η ρ σ ∂ α ∂ γ ′ [ δ 0 α η β γ + δ 0 β η α γ − η α β δ 0 γ ] 2 7 ⋅ π 4 a a ′2 H 2 ∂ β δ 0 2 ln ( 4 1 H 2 Δ x 2 ) , (81)
= − κ 2 H 3 a a ′ 2 [ ∂ 0 4 + ( a − 2 a ′ ) H ∂ 4 + 4 a ′ H ∂ 2 ∂ 0 2 + 2 a a ′ H 2 ∂ 2 ∂ 0 2 ] 2 6 ⋅ π 4 × { [ μ ν D 1 ρ σ ] + [ μ ν D 3 ρ σ ] } ln ( 1 4 H 2 Δ x 2 ) . (82) =-\frac{\kappa^2H^3aa'^2\left[\partial_0^4+(a-2a')H\partial^4+4a'H\partial^2\partial_0^2+2aa'H^2\partial^2\partial_0^2\right]}{2^6\cdot\pi^4}
\times\left\{[\mu\nu D_1^{\rho\sigma}]+[\mu\nu D_3^{\rho\sigma}]\right\}\ln\left(\frac{1}{4}H^2\Delta x^2\right). \tag*{(82)} = − 2 6 ⋅ π 4 κ 2 H 3 a a ′2 [ ∂ 0 4 + ( a − 2 a ′ ) H ∂ 4 + 4 a ′ H ∂ 2 ∂ 0 2 + 2 a a ′ H 2 ∂ 2 ∂ 0 2 ] × { [ μν D 1 ρ σ ] + [ μν D 3 ρ σ ] } ln ( 4 1 H 2 Δ x 2 ) . (82)
Tables 4, 6 and 7 list the total local and nonlocal 3-point contributions from ghost and graviton loops.
Renormalization
The purpose of this section is to absorb the divergences of Tables 2 and 4 using BPHZ counterterms. The section begins with a discussion of the general principles that constrain these counterterms. The two invariant 1-loop counterterms are given next. Then discussion turns to the plethora of noninvariant counterterms. The section ends with the fully renormalized result.
General Principles
Counterterms are gauge dependent. Because we are not using the background field gauge [30 , 31 ] , and because our gauge (10) breaks de Sitter invariance [25 , 27 ] , we must consider counterterms in addition to the usual R 2 R^2 R 2 and C 2 C^2 C 2 contributions. Certain general principles constrain the BPHZ counterterms needed to renormalize the graviton self-energy at 1-loop order on de Sitter background.
They take the form of κ 2 \kappa^2 κ 2 times two graviton fields, joined by a tensor differential operator of dimension four.
The dimensionful factors consist of derivatives and factors of H H H .
The terms with four derivatives are the same on de Sitter background as those for flat space background in the same gauge.
Because our gauge respects homogeneity, isotropy and dilatation invariance, counterterms must possess the same symmetries.
Principle #4 is especially interesting. Homoegeneity and isotropy imply that h 00 h_{00} h 00 , h 0 i h_{0i} h 0 i and h i j h_{ij} h ij are distinct tensors. Dilitation invariance implies that each counterterm must possess an overall factor of a D a^D a D , with an inverse scale factor for every derivative.
The Ward Identity imposes severe restrictions on the entire self-energy − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) , both primitive and renormalized, although not necessarily on specific counterterms. We define the Ward Operator,
W α β μ ≡ δ ( α μ ∂ β ) + a H δ 0 μ η α β . (83) \mathcal{W}^{\mu}_{\alpha\beta} \equiv\delta^\mu_{(\alpha}\partial_{\beta)} + aH\delta^\mu_0\eta_{\alpha\beta} . \tag*{(83)} W α β μ ≡ δ ( α μ ∂ β ) + a H δ 0 μ η α β . (83)
A key distinction exists between matter contributions, which are annihilated by acting the Ward operator on either coordinate [32 ] ,
W α β μ × − i [ α β Σ m a t ρ σ ] ( x ; x ′ ) = 0 , W ′ γ δ ρ × − i [ μ ν Σ m a t γ δ ] ( x ; x ′ ) = 0. (84) \mathcal{W}^{\mu}_{\alpha\beta}\times-i[{}^{\alpha\beta}\Sigma^{\rho\sigma}_{\mathrm{mat}}](x;x')=0 \,,\qquad\mathcal{W}'{}^{\rho}_{\gamma\delta}\times-i[{}^{\mu\nu}\Sigma^{\gamma\delta}_{\mathrm{mat}}](x;x')=0 . \tag*{(84)} W α β μ × − i [ α β Σ mat ρ σ ] ( x ; x ′ ) = 0 , W ′ γ δ ρ × − i [ μν Σ mat γ δ ] ( x ; x ′ ) = 0. (84)
On the other hand, gravitational contributions to the graviton self-energy are only annihilated when the Ward operator acts on both coordinates [32 ] ,
W α β μ × W ′ γ δ ρ × − i [ α β Σ g r a v γ δ ] ( x ; x ′ ) = 0. (85) \mathcal{W}^{\mu}_{\alpha\beta}\times\mathcal{W}'{}^{\rho}_{\gamma\delta}\times-i[{}^{\alpha\beta}\Sigma^{\gamma\delta}_{\mathrm{grav}}](x;x')=0 . \tag*{(85)} W α β μ × W ′ γ δ ρ × − i [ α β Σ grav γ δ ] ( x ; x ′ ) = 0. (85)
This means that matter contributions are considerably more constrained than gravitational ones. In terms of structure functions, there should be four structure functions for matter contributions, but ten are allowed for gravitational contributions.
Invariant Counterterms
There are two invariant counterterms at 1-loop,
Δ L i n v = c 1 R 2 − g + c 2 C α β γ δ C α β γ δ − g . (86) \Delta\mathcal{L}_{\mathrm{inv}}=c_1R^2\sqrt{-g}+c_2C_{\alpha\beta\gamma\delta}C^{\alpha\beta\gamma\delta}\sqrt{-g} . \tag*{(86)} Δ L inv = c 1 R 2 − g + c 2 C α β γ δ C α β γ δ − g . (86)
The term proportional to c 1 c_1 c 1 is known as the “Eddington” counterterm; the term proportional to c 2 c_2 c 2 is called the “Weyl” counterterm. The D D D -dependent coefficients c 1 c_1 c 1 and c 2 c_2 c 2 are background-independent, but gauge-dependent, so they can be read off from Capper’s 1979 flat space computation of the graviton self-energy in what is the flat space limit of our gauge (10) [33 ] ,
c 1 = μ D − 4 Γ ( D 2 − 1 ) 2 10 π 2 × 9 D 4 − 43 D 3 − 4 D 2 + 212 D − 128 ( D − 1 ) 2 ( D − 3 ) ( D − 4 ) = 13 2 6 ⋅ 3 2 ⋅ π 2 μ D − 4 D − 4 + F i n i t e , (87) c_1=\frac{\mu^{D-4}\Gamma\left(\frac{D}{2}-1\right)}{2^{10}\pi^2}\times\frac{9D^4-43D^3-4D^2+212D-128}{(D-1)^2(D-3)(D-4)}=\frac{13}{2^6\cdot3^2\cdot\pi^2}\frac{\mu^{D-4}}{D-4}+\mathrm{Finite}\,, \tag*{(87)} c 1 = 2 10 π 2 μ D − 4 Γ ( 2 D − 1 ) × ( D − 1 ) 2 ( D − 3 ) ( D − 4 ) 9 D 4 − 43 D 3 − 4 D 2 + 212 D − 128 = 2 6 ⋅ 3 2 ⋅ π 2 13 D − 4 μ D − 4 + Finite , (87)
c 2 = μ D − 4 Γ ( D 2 − 1 ) 2 9 π 2 × 4 D 4 + 17 D 3 − 97 D 2 − 34 D + 16 ( D + 1 ) ( D − 1 ) ( D − 3 ) 2 ( D − 4 ) = 61 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 μ D − 4 D − 4 + F i n i t e . (88) c_2=\frac{\mu^{D-4}\Gamma\left(\frac{D}{2}-1\right)}{2^9\pi^2}\times\frac{4D^4+17D^3-97D^2-34D+16}{(D+1)(D-1)(D-3)^2(D-4)}=\frac{61}{2^6\cdot3\cdot5\cdot\pi^2}\frac{\mu^{D-4}}{D-4}+\mathrm{Finite}\,. \tag*{(88)} c 2 = 2 9 π 2 μ D − 4 Γ ( 2 D − 1 ) × ( D + 1 ) ( D − 1 ) ( D − 3 ) 2 ( D − 4 ) 4 D 4 + 17 D 3 − 97 D 2 − 34 D + 16 = 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 61 D − 4 μ D − 4 + Finite . (88)
Here μ \mu μ is the standard mass scale of dimensional regularization.
It is useful to decompose the Eddington counterterm into three parts,
R 2 = ( R − D Λ ) 2 + 2 D Λ [ R − ( D − 2 ) Λ ] + D ( D − 4 ) Λ 2 . (89) R^2=(R-D\Lambda)^2+2D\Lambda[R-(D-2)\Lambda]+D(D-4)\Lambda^2\,. \tag*{(89)} R 2 = ( R − D Λ ) 2 + 2 D Λ [ R − ( D − 2 ) Λ ] + D ( D − 4 ) Λ 2 . (89)
We label the associated sub-Lagrangians Δ L 1 a \Delta\mathcal{L}_{1a} Δ L 1 a , Δ L 1 b \Delta\mathcal{L}_{1b} Δ L 1 b and Δ L 1 c \Delta\mathcal{L}_{1c} Δ L 1 c , respectively. Evaluating their second variations at de Sitter gives the vertices,
i δ 2 Δ S 1 a δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S = 2 c 1 κ 2 F μ ν { a D − 4 F ′ ρ σ i δ D ( x − x ′ ) } , (90) \left.\frac{i\delta^2\Delta S_{1a}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}}=2c_1\kappa^2\mathcal{F}^{\mu\nu}\left\{a^{D-4}\mathcal{F}^{\prime\rho\sigma}i\delta^D(x-x')\right\}, \tag*{(90)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 1 a dS = 2 c 1 κ 2 F μν { a D − 4 F ′ ρ σ i δ D ( x − x ′ ) } , (90)
i δ 2 Δ S 1 b δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S = D ( D − 1 ) c 1 κ 2 H 2 { [ η μ ( ρ η σ ) ν − η μ ν η ρ σ ] D i δ D ( x − x ′ ) + [ 2 ∂ ( μ η ν ) ( ρ σ ) + η μ ν ∂ ρ ∂ σ + ∂ ′ μ ∂ ′ ν η ρ σ ] [ a D − 2 i δ D ( x − x ′ ) ] } , (91) \left.\frac{i\delta^2\Delta S_{1b}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}}=D(D-1)c_1\kappa^2H^2\left\{\left[\eta^{\mu(\rho}\eta^{\sigma)\nu}-\eta^{\mu\nu}\eta^{\rho\sigma}\right]\mathcal{D}i\delta^D(x-x')+\left[2\partial^{(\mu}\eta^{\nu)(\rho}\sigma)+\eta^{\mu\nu}\partial^\rho\partial^\sigma+\partial'^\mu\partial'^\nu\eta^{\rho\sigma}\right]\left[a^{D-2}i\delta^D(x-x')\right]\right\}, \tag*{(91)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 1 b dS = D ( D − 1 ) c 1 κ 2 H 2 { [ η μ ( ρ η σ ) ν − η μν η ρ σ ] D i δ D ( x − x ′ ) + [ 2 ∂ ( μ η ν ) ( ρ σ ) + η μν ∂ ρ ∂ σ + ∂ ′ μ ∂ ′ ν η ρ σ ] [ a D − 2 i δ D ( x − x ′ ) ] } , (91)
i δ 2 Δ S 1 c δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S = 1 2 D ( D − 1 ) 2 ( D − 4 ) c 1 κ 2 H 4 [ − η μ ( ρ η σ ) ν + 1 2 η μ ν η ρ σ ] [ a D i δ D ( x − x ′ ) ] . (92) \left.\frac{i\delta^2\Delta S_{1c}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}}=\frac{1}{2}D(D-1)^2(D-4)c_1\kappa^2H^4\left[-\eta^{\mu(\rho}\eta^{\sigma)\nu}+\frac{1}{2}\eta^{\mu\nu}\eta^{\rho\sigma}\right]\left[a^D i\delta^D(x-x')\right]. \tag*{(92)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 1 c dS = 2 1 D ( D − 1 ) 2 ( D − 4 ) c 1 κ 2 H 4 [ − η μ ( ρ η σ ) ν + 2 1 η μν η ρ σ ] [ a D i δ D ( x − x ′ ) ] . (92)
Here and henceforth D ≡ ∂ α a D − 2 ∂ α \mathcal{D}\equiv\partial^\alpha a^{D-2}\partial_\alpha D ≡ ∂ α a D − 2 ∂ α , and the second order differential operator F μ ν \mathcal{F}^{\mu\nu} F μν is [34 ] ,
F μ ν ≡ ∂ μ ν − η μ ν D 1 + 2 ( D − 1 ) a H δ 0 ( μ δ 0 ν ) + ( D − 1 ) ( D − 2 ) a 2 H 2 δ 0 μ δ 0 ν , (93) \mathcal{F}^{\mu\nu}\equiv\partial^{\mu\nu}-\eta^{\mu\nu}\mathcal{D}_1+2(D-1)aH\delta^{(\mu}_0\delta^{\nu)}_0+(D-1)(D-2)a^2H^2\delta^\mu_0\delta^\nu_0\,, \tag*{(93)} F μν ≡ ∂ μν − η μν D 1 + 2 ( D − 1 ) a H δ 0 ( μ δ 0 ν ) + ( D − 1 ) ( D − 2 ) a 2 H 2 δ 0 μ δ 0 ν , (93)
where we define,
D 1 ≡ ∂ 2 + ( D − 1 ) a H ∂ 0 + ( D − 1 ) a 2 H 2 . (94) \mathcal{D}_1\equiv\partial^2+(D-1)aH\partial_0+(D-1)a^2H^2\,. \tag*{(94)} D 1 ≡ ∂ 2 + ( D − 1 ) a H ∂ 0 + ( D − 1 ) a 2 H 2 . (94)
The vertex induced by the Weyl counterterm is,
i δ 2 Δ S 2 δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S = 2 c 2 κ 2 C α β γ δ μ ν { a D − 4 C α β γ δ ′ ρ σ i δ D ( x − x ′ ) } . (95) \left.\frac{i\delta^2\Delta S_2}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}}=2c_2\kappa^2\mathcal{C}^{\alpha\beta\gamma\delta\mu\nu}\left\{a^{D-4}\mathcal{C}^{\prime\rho\sigma}_{\alpha\beta\gamma\delta}i\delta^D(x-x')\right\}. \tag*{(95)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 2 dS = 2 c 2 κ 2 C α β γ δ μν { a D − 4 C α β γ δ ′ ρ σ i δ D ( x − x ′ ) } . (95)
Here the second order differential operator C α β γ δ α β γ δ μ ν \mathcal{C}_{\alpha\beta\gamma\delta}^{\phantom{\alpha\beta\gamma\delta}\mu\nu} C α β γ δ α β γ δ μν operates on the graviton field to give the linearized, conformally transformed Weyl tensor [34 , 35 ] ,
C ~ α β γ δ = C α β γ δ α β γ δ μ ν × κ h μ ν + O ( κ 2 h 2 ) . (96) \widetilde{\mathcal{C}}_{\alpha\beta\gamma\delta}=\mathcal{C}_{\alpha\beta\gamma\delta}^{\phantom{\alpha\beta\gamma\delta}\mu\nu}\times\kappa h_{\mu\nu}+O(\kappa^2h^2)\,. \tag*{(96)} C α β γ δ = C α β γ δ α β γ δ μν × κ h μν + O ( κ 2 h 2 ) . (96)
Some tedious but straightforward tensor algebra uncovers its explicit form,
C α β γ δ μ ν = D α β γ δ μ ν − η α γ D β δ μ ν − η γ β D δ α μ ν + η β δ D α γ μ ν − η δ α D γ β μ ν D − 2 + [ η α γ η β δ − η α δ η β γ ] D μ ν ( D − 1 ) ( D − 2 ) (97) C_{\alpha\beta\gamma\delta}^{\mu\nu}
= D_{\alpha\beta\gamma\delta}^{\mu\nu}
-\frac{\eta_{\alpha\gamma}D_{\beta\delta}^{\mu\nu}-\eta_{\gamma\beta}D_{\delta\alpha}^{\mu\nu}+\eta_{\beta\delta}D_{\alpha\gamma}^{\mu\nu}-\eta_{\delta\alpha}D_{\gamma\beta}^{\mu\nu}}{D-2}
+\frac{\left[\eta_{\alpha\gamma}\eta_{\beta\delta}-\eta_{\alpha\delta}\eta_{\beta\gamma}\right]D^{\mu\nu}}{(D-1)(D-2)} \tag*{(97)} C α β γ δ μν = D α β γ δ μν − D − 2 η α γ D β δ μν − η γ β D δ α μν + η β δ D α γ μν − η δ α D γ β μν + ( D − 1 ) ( D − 2 ) [ η α γ η β δ − η α δ η β γ ] D μν (97)
where we define,
D α β γ δ μ ν ≡ − 1 2 { δ α μ δ γ ν ∂ β ∂ δ − δ γ μ δ δ ν ∂ δ ∂ α + δ α μ δ δ ν ∂ α ∂ γ − δ α μ δ γ ν ∂ γ ∂ β } (98) D_{\alpha\beta\gamma\delta}^{\mu\nu}\equiv-\frac{1}{2}\left\{\delta_{\alpha}^{\mu}\delta_{\gamma}^{\nu}\partial_{\beta}\partial_{\delta}-\delta_{\gamma}^{\mu}\delta_{\delta}^{\nu}\partial_{\delta}\partial_{\alpha}+\delta_{\alpha}^{\mu}\delta_{\delta}^{\nu}\partial_{\alpha}\partial_{\gamma}-\delta_{\alpha}^{\mu}\delta_{\gamma}^{\nu}\partial_{\gamma}\partial_{\beta}\right\} \tag*{(98)} D α β γ δ μν ≡ − 2 1 { δ α μ δ γ ν ∂ β ∂ δ − δ γ μ δ δ ν ∂ δ ∂ α + δ α μ δ δ ν ∂ α ∂ γ − δ α μ δ γ ν ∂ γ ∂ β } (98)
D β δ μ ν ≡ η α γ D α β γ δ μ ν = − 1 2 { η μ ν ∂ β ∂ δ − 2 ∂ ( μ δ ( β ν ) ∂ δ ) + δ ( β μ δ δ ) ν ∂ 2 } (99) D_{\beta\delta}^{\mu\nu}\equiv\eta^{\alpha\gamma}D_{\alpha\beta\gamma\delta}^{\mu\nu}
=-\frac{1}{2}\left\{\eta^{\mu\nu}\partial_{\beta}\partial_{\delta}-2\partial^{(\mu}\delta^{\nu)}_{(\beta}\partial_{\delta)}+\delta_{(\beta}^{\mu}\delta_{\delta)}^{\nu}\partial^{2}\right\} \tag*{(99)} D β δ μν ≡ η α γ D α β γ δ μν = − 2 1 { η μν ∂ β ∂ δ − 2 ∂ ( μ δ ( β ν ) ∂ δ ) + δ ( β μ δ δ ) ν ∂ 2 } (99)
D μ ν ≡ η β δ D β δ μ ν = ∂ μ ∂ ν − η μ ν ∂ 2 (100) D^{\mu\nu}\equiv\eta^{\beta\delta}D_{\beta\delta}^{\mu\nu}
=\partial^{\mu}\partial^{\nu}-\eta^{\mu\nu}\partial^{2} \tag*{(100)} D μν ≡ η β δ D β δ μν = ∂ μ ∂ ν − η μν ∂ 2 (100)
The 2-point vertex (92) is finite, so we can take its D = 4 D=4 D = 4 limit,
i δ 2 Δ S 1 c δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S ⟶ 13 κ 2 H 4 a 4 2 5 ⋅ π 2 { 1 2 η μ ν η ρ σ − η μ ( ρ η σ ) ν } i δ 4 ( x − x ′ ) (101) \left.\frac{i\delta^{2}\Delta S_{1c}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}}
\longrightarrow\frac{13\kappa^{2}H^{4}a^{4}}{2^{5}\cdot\pi^{2}}\left\{\frac{1}{2}\eta^{\mu\nu}\eta^{\rho\sigma}-\eta^{\mu(\rho}\eta^{\sigma)\nu}\right\}i\delta^{4}(x-x') \tag*{(101)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 1 c dS ⟶ 2 5 ⋅ π 2 13 κ 2 H 4 a 4 { 2 1 η μν η ρ σ − η μ ( ρ η σ ) ν } i δ 4 ( x − x ′ ) (101)
Its contribution to the final result is given in Table 8 . Examination of the divergent 2-point vertices (90) , (91) and (95) reveals two key qualitative differences with respect to the primitive divergences of Table 2 and Table 4 :
Table 8. Coefficients of [ ∇ μ D i ρ σ ] i δ 4 ( x − x ′ ) [\nabla^{\mu}D_{i}^{\rho\sigma}]i\delta^{4}(x-x') [ ∇ μ D i ρ σ ] i δ 4 ( x − x ′ ) from the finite part of L 1 b \mathcal{L}_{1b} L 1 b and from L 1 c \mathcal{L}_{1c} L 1 c . The contributions from L 1 b \mathcal{L}_{1b} L 1 b and L 1 c \mathcal{L}_{1c} L 1 c are multiplied by 13 κ 2 2 5 ⋅ 3 ⋅ π 2 \frac{13\kappa^{2}}{2^{5}\cdot3\cdot\pi^{2}} 2 5 ⋅ 3 ⋅ π 2 13 κ 2 and 13 κ 2 2 5 ⋅ π 2 \frac{13\kappa^{2}}{2^{5}\cdot\pi^{2}} 2 5 ⋅ π 2 13 κ 2 , respectively.
Whereas the dimensional constant of the primitive divergences C 4 C_{4} C 4 and C 3 C_{3} C 3 is ( H / 2 ) D − 4 (H/2)^{D-4} ( H /2 ) D − 4 , it is μ D − 4 \mu^{D-4} μ D − 4 in the counterterms; and
Whereas the primitive divergences involve only integer powers of the scale factor, the counterterms each contain a relative factor of a D − 4 a^{D-4} a D − 4 .
This suggests that we extract the finite factors of ln ( 2 μ a / H ) \ln(2\mu a/H) ln ( 2 μ a / H ) , which constitute a very important part of the final result. For (91) this process gives,
i δ 2 Δ S 1 a δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ D S ⟶ 2 ( H 2 μ ) D − 4 C 1 κ 2 F μ ν × F ′ ρ σ i δ D ( x − x ′ ) + 13 κ 2 2 6 ⋅ 3 2 ⋅ π 2 F μ ν × F ′ ρ σ { ln ( 4 μ 2 a a ′ H 2 ) i δ 4 ( x − x ′ ) } . (102) \left.\frac{i\delta^2\Delta S_{1a}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{DS}}
\longrightarrow2\left(\frac{H}{2\mu}\right)^{D-4}C_1\kappa^2\mathcal{F}^{\mu\nu}\times\mathcal{F}'^{\rho\sigma}i\delta^D(x-x')
+\frac{13\kappa^2}{2^6\cdot3^2\cdot\pi^2}\mathcal{F}^{\mu\nu}\times\mathcal{F}'^{\rho\sigma}\left\{\ln\left(\frac{4\mu^2aa'}{H^2}\right)i\delta^4(x-x')\right\}. \tag*{(102)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 1 a DS ⟶ 2 ( 2 μ H ) D − 4 C 1 κ 2 F μν × F ′ ρ σ i δ D ( x − x ′ ) + 2 6 ⋅ 3 2 ⋅ π 2 13 κ 2 F μν × F ′ ρ σ { ln ( H 2 4 μ 2 a a ′ ) i δ 4 ( x − x ′ ) } . (102)
Extracting the scale factors from the divergent part of (96) allows us to perform the contractions on α \alpha α , β \beta β , γ \gamma γ and δ \delta δ [34 , 35 ] ,
i δ 2 Δ S 2 δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ D S ⟶ 2 ( D − 3 D − 2 ) ( H 2 μ ) D − 4 C 2 { D μ ( ρ D σ ) ν − D μ ν D ρ σ D − 1 } i δ D ( x − x ′ ) + 61 κ 2 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 C α β γ δ μ ν × C α β γ δ ′ ρ σ { ln ( 4 μ 2 a a ′ H 2 ) i δ 4 ( x − x ′ ) } . (103) \left.\frac{i\delta^2\Delta S_2}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{DS}}
\longrightarrow2\left(\frac{D-3}{D-2}\right)\left(\frac{H}{2\mu}\right)^{D-4}C_2\left\{\mathcal{D}^{\mu(\rho}\mathcal{D}^{\sigma)\nu}-\frac{\mathcal{D}^{\mu\nu}\mathcal{D}^{\rho\sigma}}{D-1}\right\}i\delta^D(x-x')
+\frac{61\kappa^2}{2^6\cdot3\cdot5\cdot\pi^2}C^{\alpha\beta\gamma\delta\mu\nu}\times C'_{\alpha\beta\gamma\delta}{}^{\rho\sigma}\left\{\ln\left(\frac{4\mu^2aa'}{H^2}\right)i\delta^4(x-x')\right\}. \tag*{(103)} δ h μν ( x ) δ h ρ σ ( x ′ ) i δ 2 Δ S 2 DS ⟶ 2 ( D − 2 D − 3 ) ( 2 μ H ) D − 4 C 2 { D μ ( ρ D σ ) ν − D − 1 D μν D ρ σ } i δ D ( x − x ′ ) + 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 C α β γ δ μν × C α β γ δ ′ ρ σ { ln ( H 2 4 μ 2 a a ′ ) i δ 4 ( x − x ′ ) } . (103)
Making a similar reduction of the Einstein counterterm (92) is complicated by the fact that it, even for D = 4 D=4 D = 4 , it possesses scale factors that must be moved to the left of any derivatives to put it in standard form. The finite part is given is the left column of Table 8 , and the middle column of Table 9 gives its divergent part. The left and right columns of Table 9 give the divergent contributions from L 1 a \mathcal{L}_{1a} L 1 a and L 2 \mathcal{L}_2 L 2 , respectively. Rather than reducing the finite parts to standard form, it is preferable to leave them in the forms given on the second lines of expressions (103) and (104).
Table 9. Coefficients of [ μ ν D i ρ σ ] i δ D ( x − x ′ ) [\mu\nu D_i^{\rho\sigma}]i\delta^{D}(x-x') [ μν D i ρ σ ] i δ D ( x − x ′ ) in the counterterms. The leftmost is multiplied by 2 ( H 2 μ ) D − 4 c 1 κ 2 2\left(\frac{H}{2\mu}\right)^{D-4}c_{1}\kappa^{2} 2 ( 2 μ H ) D − 4 c 1 κ 2 ; the middle is multiplied by D ( D − 1 ) ( H 2 μ ) D − 4 c 1 κ 2 D(D-1)\left(\frac{H}{2\mu}\right)^{D-4}c_{1}\kappa^{2} D ( D − 1 ) ( 2 μ H ) D − 4 c 1 κ 2 ; and the rightmost is multiplied by 2 ( H 2 μ ) D − 4 c 2 κ 2 2\left(\frac{H}{2\mu}\right)^{D-4}c_{2}\kappa^{2} 2 ( 2 μ H ) D − 4 c 2 κ 2 .
Noninvariant Counterterms
Although the fact that our gauge (10) breaks de Sitter invariance means that noninvariant counterterms are needed, we have still tried to build them from geometrically meaningful quantities such as curvatures and connections. Table 10 gives the complete list, including the two invariant ones. In each case the coefficient consists of a divergent part plus a finite remainder chosen to enforce the Ward identity (86). As with the invariant counterterms, each of the 17 noninvariant will give rise to factors of ln ( 2 μ a / H ) \ln(2\mu a/H) ln ( 2 μ a / H ) .
Table 10. Counterterms and their coefficients. The divergent constants are C 3 = μ D − 4 4 π D / 2 Γ ( D 2 − 1 ) ( D − 3 ) ( D − 4 ) C_3 = \frac{\mu^{D-4}}{4\pi^{D/2}} \frac{\Gamma(\frac{D}{2}-1)}{(D-3)(D-4)} C 3 = 4 π D /2 μ D − 4 ( D − 3 ) ( D − 4 ) Γ ( 2 D − 1 ) and C 4 = μ D − 4 16 π D / 2 Γ ( D − 1 ) Γ ( D 2 ) π cot ( D π 2 ) C_4 = \frac{\mu^{D-4}}{16\pi^{D/2}} \frac{\Gamma(D-1)}{\Gamma(\frac{D}{2})}\pi\cot\left(\frac{D\pi}{2}\right) C 4 = 16 π D /2 μ D − 4 Γ ( 2 D ) Γ ( D − 1 ) π cot ( 2 D π ) .
The Fully Renormalized Result
The fully renormalized result takes the same form (30) as the primitive result. In fact, the nonlocal contributions are exactly the same as the primitive nonlocal contributions given in Tables 6 and 7. The local contributions consist of the second lines of expressions (103) and (104), as well as Table 8 and 11.
Solving the Effective Field Equations
The purpose of this section is to solve the effective field equation (3) for 1-loop corrections to the graviton mode function (4) , and for the Newtonian potential (5) , in our gauge (10) . The section begins by explaining how to convert the in-out graviton self-energy computed in the previous two sections to the in-in result required to make the effective field equation real and causal. We also comment the form the effective field equation must take when working in a fixed gauge, and on how to solve it perturbatively. The next topic is solving for the 1-loop correction to the graviton mode function. The section closes with a discussion of the most significant 1-loop corrections to the Newtonian potential at late times.
The Effective Field Equation
The graviton self-energy we have derived is technically the in-out result provided by the usual Feynman rules. That is perfectly valid for asymptotic scattering problems, but it suffers from two problems when the intention is to use the effective field equations to understand the evolution of quantum corrections at finite times:
Both problems can be surmounted using the Schwinger-Keldysh formalism [36 , 37 , 38 , 39 , 40 ] . This is a diagrammatic formalism for giving true expectation values, which almost as simple as the Feynman rules. The effective field equations of the Schwinger-Keldysh formalism are both real and causal [41 , 42 , 43 ] . There are a series of simple rules for converting an in-out 1PI 2-point function to the in-in formalism [44 ] :
Every endpoint carries a ± \pm ± polarity, so that the conformal coordinate interval Δ x 2 ( x ; x ′ ) \Delta x^2(x;x') Δ x 2 ( x ; x ′ ) of the in-out formalism, which was defined in expression (33) , gives rise to in-in four intervals,
Δ x + + 2 ≡ ∣ x ⃗ − x ⃗ ′ ∣ 2 − ( ∣ η − η ′ ∣ − i ϵ ) 2 (104) \Delta x^2_{++} \equiv\lvert\vec{x}-\vec{x}'\rvert^2-\left(\lvert\eta-\eta'\rvert-i\epsilon\right)^2 \tag*{(104)} Δ x ++ 2 ≡ ∣ x − x ′ ∣ 2 − ( ∣ η − η ′ ∣ − i ϵ ) 2 (104)
Δ x + − 2 ≡ ∣ x ⃗ − x ⃗ ′ ∣ 2 − ( η − η ′ + i ϵ ) 2 (105) \Delta x^2_{+-} \equiv\lvert\vec{x}-\vec{x}'\rvert^2-\left(\eta-\eta'+i\epsilon\right)^2 \tag*{(105)} Δ x +− 2 ≡ ∣ x − x ′ ∣ 2 − ( η − η ′ + i ϵ ) 2 (105)
Δ x − + 2 ≡ ∣ x ⃗ − x ⃗ ′ ∣ 2 − ( η − η ′ − i ϵ ) 2 (106) \Delta x^2_{-+} \equiv\lvert\vec{x}-\vec{x}'\rvert^2-\left(\eta-\eta'-i\epsilon\right)^2 \tag*{(106)} Δ x −+ 2 ≡ ∣ x − x ′ ∣ 2 − ( η − η ′ − i ϵ ) 2 (106)
Δ x − − 2 ≡ ∣ x ⃗ − x ⃗ ′ ∣ 2 − ( ∣ η − η ′ ∣ + i ϵ ) 2 (107) \Delta x^2_{--} \equiv\lvert\vec{x}-\vec{x}'\rvert^2-\left(\lvert\eta-\eta'\rvert+i\epsilon\right)^2 \tag*{(107)} Δ x −− 2 ≡ ∣ x − x ′ ∣ 2 − ( ∣ η − η ′ ∣ + i ϵ ) 2 (107)
Similarly, the 1PI 2-point function of the in-out formalism gives to four in-in 1PI 2-point functions, each involving the appropriate interval.
Only the ++ and −− 1PI 2-point functions contain terms involving δ 4 ( x − x ′ ) \delta^4(x-x') δ 4 ( x − x ′ ) .
The 1PI 2-point function has a factor of − 1 -1 − 1 for each − polarity.
The graviton self-energy that goes in the the Schwinger-Keldysh effective field equation is the sum of the ++ and +− ones.
From rule 4 we see that there is no change at all to the local contributions of Table 8 and Table 11 , or to the important logarithms on the second lines of expressions (103) and (104). The only change comes to the nonlocal contributions of Table 6 and Table 7 , and the net effect is to replace the two logarithms with,
Table 6. Primitive double log contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
Table 7. Primitive single log contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
Table 11. Renormalized local contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) .
ln ( 1 4 H 2 Δ x 2 ) ⟶ 2 π i θ ( Δ η − Δ r ) ≡ f 3 S K ( x − x ′ ) (108) \ln\left(\frac{1}{4}H^2\Delta x^2\right) \longrightarrow2\pi i\theta(\Delta\eta-\Delta r) \equiv f_3^{\mathrm{SK}}(x-x') \tag*{(108)} ln ( 4 1 H 2 Δ x 2 ) ⟶ 2 π i θ ( Δ η − Δ r ) ≡ f 3 SK ( x − x ′ ) (108)
ln 2 ( 1 4 H 2 Δ x 2 ) ⟶ 4 π i θ ( Δ η − Δ r ) ln [ 1 4 H 2 ( Δ η 2 − Δ r 2 ) ] ≡ f 2 S K ( x − x ′ ) (109) \ln^2\left(\frac{1}{4}H^2\Delta x^2\right) \longrightarrow4\pi i\theta(\Delta\eta-\Delta r)\ln\left[\frac{1}{4}H^2(\Delta\eta^2-\Delta r^2)\right] \equiv f_2^{\mathrm{SK}}(x-x') \tag*{(109)} ln 2 ( 4 1 H 2 Δ x 2 ) ⟶ 4 π i θ ( Δ η − Δ r ) ln [ 4 1 H 2 ( Δ η 2 − Δ r 2 ) ] ≡ f 2 SK ( x − x ′ ) (109)
where we define the temporal and spatial intervals as,
Δ η ≡ η − η ′ , Δ r ≡ ∥ x ⃗ − x ⃗ ′ ∥ . (110) \Delta\eta\equiv\eta-\eta' \,, \qquad\Delta r \equiv\lVert\vec{x}-\vec{x}'\rVert\,. \tag*{(110)} Δ η ≡ η − η ′ , Δ r ≡ ∥ x − x ′ ∥ . (110)
The reality and causality of the Schwinger-Keldysh effective field equation are now manifest. We might define f 1 S K ( x − x ′ ) ≡ i δ 4 ( x − x ′ ) f_1^{\mathrm{SK}}(x-x') \equiv i\delta^4(x-x') f 1 SK ( x − x ′ ) ≡ i δ 4 ( x − x ′ ) .
Two more issues require discussion concerning the effective field equation (3) . The first issue derives from the fact that, right now, we only possess the 1-loop correction to the graviton self-energy. This means that the effective field equation must be solved perturbatively,
h μ ν = h μ ν ( 0 ) + h μ ν ( 1 ) + h μ ν ( 2 ) + ⋯ (111) h_{\mu\nu}=h_{\mu\nu}^{(0)}+h_{\mu\nu}^{(1)}+h_{\mu\nu}^{(2)}+\cdots \tag*{(111)} h μν = h μν ( 0 ) + h μν ( 1 ) + h μν ( 2 ) + ⋯ (111)
Here h μ ν ( 0 ) h_{\mu\nu}^{(0)} h μν ( 0 ) is the “classical” field, which obeys,
D μ ν ρ σ κ h ρ σ ( 0 ) = 8 π G T l i n μ ν . (112) \mathcal{D}^{\mu\nu\rho\sigma}\kappa h_{\rho\sigma}^{(0)}=8\pi G T_{\mathrm{lin}}^{\mu\nu}\,. \tag*{(112)} D μν ρ σ κ h ρ σ ( 0 ) = 8 π G T lin μν . (112)
The 1-loop corrections we seek obey,
D μ ν ρ σ κ h ρ σ ( 1 ) ( x ) = ∫ d 4 x ′ [ μ ν Σ ρ σ ] ( x ; x ′ ) κ h ρ σ ( 0 ) ( x ′ ) . (113) \mathcal{D}^{\mu\nu\rho\sigma}\kappa h_{\rho\sigma}^{(1)}(x)=\int d^4x'\,[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x')\kappa h_{\rho\sigma}^{(0)}(x')\,. \tag*{(113)} D μν ρ σ κ h ρ σ ( 1 ) ( x ) = ∫ d 4 x ′ [ μν Σ ρ σ ] ( x ; x ′ ) κ h ρ σ ( 0 ) ( x ′ ) . (113)
Most of the graviton self-energy is contained in Table 6 , Table 7 , Table 8 , and Table 11 . These contributions consist of a sum (30) of terms of the form,
C i j ( a , a ′ , ∂ ) × [ μ ν D i ρ σ ] × f j S K ( x − x ′ ) . (114) C_{ij}(a,a',\partial)\times[{}^{\mu\nu}D_i^{\rho\sigma}]\times f_j^{\mathrm{SK}}(x-x')\,. \tag*{(114)} C ij ( a , a ′ , ∂ ) × [ μν D i ρ σ ] × f j SK ( x − x ′ ) . (114)
In Table 12 we give the result of contracting the basis tenors [ μ ν D i ρ σ ] [{}^{\mu\nu}D_i^{\rho\sigma}] [ μν D i ρ σ ] , given in Table 1 , into h ρ σ ( x ′ ) h_{\rho\sigma}(x') h ρ σ ( x ′ ) . In some cases we have used the larger form, of an integral of (115), to reflect spatial derivatives in [ μ ν D i ρ σ ] [{}^{\mu\nu}D_i^{\rho\sigma}] [ μν D i ρ σ ] from + ∂ μ +\partial_\mu + ∂ μ to − ∂ μ ′ -\partial'_\mu − ∂ μ ′ , and then partially integrating them to act on the graviton field h ρ σ ( x ′ ) h_{\rho\sigma}(x') h ρ σ ( x ′ ) . The remaining contributions to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) come from the second lines of
Table 12. Contraction of [ μ ν D i ρ σ ] [\mu\nu D_i^{\rho\sigma}] [ μν D i ρ σ ] into h ρ σ ( x ′ ) h_{\rho\sigma}(x') h ρ σ ( x ′ ) , with some spatial derivatives reflected and partially integrated onto h ρ σ ( x ′ ) h_{\rho\sigma}(x') h ρ σ ( x ′ ) .
expressions (102) and (103) . The results can be expressed in terms of the zeroth order, linearized curvatures [46 ] ,
13 κ 2 2 6 ⋅ 3 2 ⋅ π 2 F μ ν ∫ d 4 x ′ ln ( 4 μ 2 a a ′ H 2 ) δ 4 ( x − x ′ ) × F ′ ρ σ κ h ρ σ ( x ′ ) = 13 κ 2 2 5 ⋅ 3 2 ⋅ π 2 F μ ν { ln ( 2 μ a H ) R ( 0 ) ( x ) } , (115) \begin{aligned}
\frac{13\kappa^2}{2^6\cdot3^2\cdot\pi^2}\mathcal{F}^{\mu\nu}\int d^4x'\,\ln\left(\frac{4\mu^2aa'}{H^2}\right)\delta^4(x-x')\times\mathcal{F}'^{\rho\sigma\kappa}h_{\rho\sigma}(x') \\
=\frac{13\kappa^2}{2^5\cdot3^2\cdot\pi^2}\mathcal{F}^{\mu\nu}\left\{\ln\left(\frac{2\mu a}{H}\right)R^{(0)}(x)\right\}, \tag*{(115)}
\end{aligned} 2 6 ⋅ 3 2 ⋅ π 2 13 κ 2 F μν ∫ d 4 x ′ ln ( H 2 4 μ 2 a a ′ ) δ 4 ( x − x ′ ) × F ′ ρ σ κ h ρ σ ( x ′ ) = 2 5 ⋅ 3 2 ⋅ π 2 13 κ 2 F μν { ln ( H 2 μ a ) R ( 0 ) ( x ) } , (115)
61 κ 2 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 C α β γ δ μ ν ∫ d 4 x ′ ln ( 4 μ 2 a a ′ H 2 ) δ 4 ( x − x ′ ) × C α β γ δ ′ ρ σ κ h ρ σ ( x ′ ) = − 61 κ 2 δ ρ σ 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 { ln ( 2 μ a H ) C ( 0 ) μ ν ρ σ ( x ) } . (116) \begin{aligned}
\frac{61\kappa^2}{2^6\cdot3\cdot5\cdot\pi^2}\mathcal{C}^{\alpha\beta\gamma\delta\mu\nu}\int d^4x'\,\ln\left(\frac{4\mu^2aa'}{H^2}\right)\delta^4(x-x')\times\mathcal{C}'_{\alpha\beta\gamma\delta}{}^{\rho\sigma\kappa}h_{\rho\sigma}(x') \\
=-\frac{61\kappa^2\delta_{\rho\sigma}}{2^4\cdot3\cdot5\cdot\pi^2}\left\{\ln\left(\frac{2\mu a}{H}\right)\mathcal{C}^{(0)\mu\nu\rho\sigma}(x)\right\}. \tag*{(116)}
\end{aligned} 2 6 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 C α β γ δ μν ∫ d 4 x ′ ln ( H 2 4 μ 2 a a ′ ) δ 4 ( x − x ′ ) × C α β γ δ ′ ρ σ κ h ρ σ ( x ′ ) = − 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 δ ρ σ { ln ( H 2 μ a ) C ( 0 ) μν ρ σ ( x ) } . (116)
The final issue derives from how the Ward operator (83) acts on the three terms in the effective field equation. It annihilates both the invariant kinetic operator and the linearized stress tensor,
W α β μ × D i n v α β ρ σ = 0 = W α β μ × T l i n α β . (117) \mathcal{W}^{\mu}_{\alpha\beta}\times\mathcal{D}_{\mathrm{inv}}^{\alpha\beta\rho\sigma}=0=\mathcal{W}^{\mu}_{\alpha\beta}\times\mathcal{T}_{\mathrm{lin}}^{\alpha\beta}. \tag*{(117)} W α β μ × D inv α β ρ σ = 0 = W α β μ × T lin α β . (117)
If we use the kinetic operator of the invariant theory, then consistency of the equation requires that the Ward operator also annihilate the self-energy. That is indeed the case for matter contributions to the graviton self-energy (84) , and one employs the invariant kinetic operator in solving for how these matter corrections affect gravity [35 , 45 , 46 , 47 , 48 ] . Because the Ward operator annihilates all three terms of the effective field equation, there are four relations among the ten tensor components, leading to only six independent equations.
In contrast, graviton loop corrections to − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) are only annihilated when the Ward operator acts simultaneously on both coordinates (85) . Were we to use the invariant kinetic operator, equation (113) would not be consistent because acting the Ward operator on the left hand side would give zero, whereas the right hand side would not vanish. Consistency is restored by employing the gauge-fixed kinetic operator, whose 3 + 1 3+1 3 + 1 decomposition is,
D 00 ρ σ h ρ σ = 1 4 D A ( h 00 + h k k ) − 2 a 4 H 2 h 00 , (118) \mathcal{D}^{00\rho\sigma}h_{\rho\sigma}=\frac{1}{4}D_A(h_{00}+h_{kk})-2a^4H^2h_{00}, \tag*{(118)} D 00 ρ σ h ρ σ = 4 1 D A ( h 00 + h k k ) − 2 a 4 H 2 h 00 , (118)
D 0 i ρ σ h ρ σ = − 1 2 D B h 0 i , (119) \mathcal{D}^{0i\rho\sigma}h_{\rho\sigma}=-\frac{1}{2}D_Bh_{0i}, \tag*{(119)} D 0 i ρ σ h ρ σ = − 2 1 D B h 0 i , (119)
D i j ρ σ h ρ σ = 1 2 D A [ h i j + 1 2 δ i j ( h 00 − h k k ) ] . (120) \mathcal{D}^{ij\rho\sigma}h_{\rho\sigma}=\frac{1}{2}D_A\left[h_{ij}+\frac{1}{2}\delta_{ij}(h_{00}-h_{kk})\right]. \tag*{(120)} D ij ρ σ h ρ σ = 2 1 D A [ h ij + 2 1 δ ij ( h 00 − h k k ) ] . (120)
The operators D A D_A D A and D B D_B D B are,
D A = − a 2 [ ∂ 0 2 + 2 a H ∂ 0 − ∇ 2 ] , D B = D A − 2 a 4 H 2 . (121) D_A=-a^2\left[\partial_0^2+2aH\partial_0-\nabla^2\right],\qquad D_B=D_A-2a^4H^2. \tag*{(121)} D A = − a 2 [ ∂ 0 2 + 2 a H ∂ 0 − ∇ 2 ] , D B = D A − 2 a 4 H 2 . (121)
They are, respectively, the kinetic operators for a massless, minimally coupled scalar (D A D_A D A ) and for a massless, conformally coupled scalar (D B D_B D B ). Note that all ten tensor components of the gauge-fixed effective field equation are independent.
The Graviton Mode Function
Gravitational radiation corresponds to T l i n μ ν = 0 T_{\mathrm{lin}}^{\mu\nu}=0 T lin μν = 0 and a graviton field strength of the form,
κ h ρ σ ( x ) = ϵ ρ σ ( k ⃗ ) × u ( t , k ) e i k ⃗ ⋅ x ⃗ , (122) \kappa h_{\rho\sigma}(x)=\epsilon_{\rho\sigma}(\vec{k})\times u(t,k)e^{i\vec{k}\cdot\vec{x}}, \tag*{(122)} κ h ρ σ ( x ) = ϵ ρ σ ( k ) × u ( t , k ) e i k ⋅ x , (122)
where the polarization tensor is purely spatial, transverse and traceless,
ϵ 0 μ ( k ⃗ ) = 0 , k i ϵ i j ( k ⃗ ) = 0 = ϵ i i ( k ⃗ ) . (123) \epsilon_{0\mu}(\vec{k})=0 \qquad, \qquad k_i\epsilon_{ij}(\vec{k})=0=\epsilon_{ii}(\vec{k}) . \tag*{(123)} ϵ 0 μ ( k ) = 0 , k i ϵ ij ( k ) = 0 = ϵ ii ( k ) . (123)
Acting the gauge-fixed kinetic operator on (122) gives,
D μ ν ρ σ κ h ρ σ ( x ) = ϵ μ ν ( k ⃗ ) e i k ⃗ ⋅ x ⃗ × − a 2 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( t , k ) . (124) \mathcal{D}^{\mu\nu\rho\sigma}\kappa h_{\rho\sigma}(x)=\epsilon^{\mu\nu}(\vec{k})e^{i\vec{k}\cdot\vec{x}}\times-\frac{a^2}{2}\left[\partial_0^2+2aH\partial_0+k^2\right]u(t,k) . \tag*{(124)} D μν ρ σ κ h ρ σ ( x ) = ϵ μν ( k ) e i k ⋅ x × − 2 a 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( t , k ) . (124)
Of course the mode function is expanded perturbatively,
u ( t , k ) = u ( 0 ) ( t , k ) + u ( 1 ) ( t , k ) + u ( 2 ) ( t , k ) + … (125) u(t,k)=u^{(0)}(t,k)+u^{(1)}(t,k)+u^{(2)}(t,k)+\ldots \tag*{(125)} u ( t , k ) = u ( 0 ) ( t , k ) + u ( 1 ) ( t , k ) + u ( 2 ) ( t , k ) + … (125)
The zeroth order mode function depends on time only through the scale factor a ( t ) a(t) a ( t ) ,
u ( 0 ) ( t , k ) = H 2 k 3 [ 1 − i k a H ] exp [ i k a H ] . (126) u^{(0)}(t,k)=\frac{H}{\sqrt{2k^3}}\left[1-\frac{ik}{aH}\right]\exp\left[\frac{ik}{aH}\right] . \tag*{(126)} u ( 0 ) ( t , k ) = 2 k 3 H [ 1 − a H ik ] exp [ a H ik ] . (126)
We turn now to the right hand side of equation (114). Examination of Table 12 shows that only the i = 2 i=2 i = 2 term contributes for gravitational radiation,
i ∫ d 4 x ′ ∑ i = 1 21 ∑ j = 1 3 C i j ( a , a ′ , ∂ ) [ μ ν D i ρ σ ] × f j S K ( x − x ′ ) × κ h ρ σ ( 0 ) ( x ′ ) = i ∫ d 4 x ′ ∑ j = 1 3 C 2 j ( a , a ′ , ∂ ) × f j S K ( x − x ′ ) × κ h ( 0 ) μ ν ( x ′ ) , (127) \begin{aligned}
i\int d^4x'\sum_{i=1}^{21}\sum_{j=1}^{3}C_{ij}(a,a',\partial)\left[\mu\nu D_i^{\rho\sigma}\right]\times f_j^{\mathrm{SK}}(x-x')\times\kappa h_{\rho\sigma}^{(0)}(x') \\
&= i\int d^4x'\sum_{j=1}^{3}C_{2j}(a,a',\partial)\times f_j^{\mathrm{SK}}(x-x')\times\kappa h^{(0)\mu\nu}(x'), \tag*{(127)}
\end{aligned} i ∫ d 4 x ′ i = 1 ∑ 21 j = 1 ∑ 3 C ij ( a , a ′ , ∂ ) [ μν D i ρ σ ] × f j SK ( x − x ′ ) × κ h ρ σ ( 0 ) ( x ′ ) = i ∫ d 4 x ′ j = 1 ∑ 3 C 2 j ( a , a ′ , ∂ ) × f j SK ( x − x ′ ) × κ h ( 0 ) μν ( x ′ ) , (127)
= ϵ μ ν ( k ⃗ ) e i k ⃗ ⋅ x ⃗ × ∫ d 4 x ′ ∑ j = 1 3 C 2 j ( a , a ′ , ∂ ) × i f j S K ( x − x ′ ) × u ( 0 ) ( t ′ , k ) e − i k ⃗ ⋅ Δ x ⃗ (128) =\epsilon^{\mu\nu}(\vec{k})e^{i\vec{k}\cdot\vec{x}}\times\int d^4x'\sum_{j=1}^{3}C_{2j}(a,a',\partial)\times i f_j^{\mathrm{SK}}(x-x')\times u^{(0)}(t',k)e^{-i\vec{k}\cdot\Delta\vec{x}} \tag*{(128)} = ϵ μν ( k ) e i k ⋅ x × ∫ d 4 x ′ j = 1 ∑ 3 C 2 j ( a , a ′ , ∂ ) × i f j SK ( x − x ′ ) × u ( 0 ) ( t ′ , k ) e − i k ⋅ Δ x (128)
The linearized Ricci scalar (116) vanishes for gravitational radiation, and the contribution from the Weyl tensor (117) is [46 ] ,
− 61 κ 2 ∂ ρ ∂ σ 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 { ln ( 2 μ a H ) C ( 0 ) ρ μ σ ν ( x ) } = 61 κ 2 ϵ μ ν ( k ⃗ ) e i k ⃗ ⋅ x ⃗ 2 5 ⋅ 3 ⋅ 5 ⋅ π 2 i k a 2 H 2 ∂ 0 u ( 0 ) ( t , k ) . (129) -\frac{61\kappa^2\partial_\rho\partial_\sigma}{2^4\cdot3\cdot5\cdot\pi^2}\left\{\ln\left(\frac{2\mu a}{H}\right)C^{(0)\rho\mu\sigma\nu}(x)\right\}=\frac{61\kappa^2\epsilon^{\mu\nu}(\vec{k})e^{i\vec{k}\cdot\vec{x}}}{2^5\cdot3\cdot5\cdot\pi^2}ika^2H^2\partial_0u^{(0)}(t,k). \tag*{(129)} − 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 ∂ ρ ∂ σ { ln ( H 2 μ a ) C ( 0 ) ρ μ σ ν ( x ) } = 2 5 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 ϵ μν ( k ) e i k ⋅ x ik a 2 H 2 ∂ 0 u ( 0 ) ( t , k ) . (129)
Because both sides of equation (113) are proportional to ϵ μ ν ( k ⃗ ) e i k ⃗ ⋅ x ⃗ \epsilon^{\mu\nu}(\vec{k})e^{i\vec{k}\cdot\vec{x}} ϵ μν ( k ) e i k ⋅ x , we can drop this factor and write a scalar equation for the 1-loop correction to the mode function,
− a 2 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( 1 ) ( t , k ) = 61 κ 2 2 5 ⋅ 3 ⋅ 5 ⋅ π 2 i k a 2 H 2 ∂ 0 u ( 0 ) ( t , k ) + ∫ d 4 x ′ ∑ j = 1 3 C 2 j ( a , a ′ , ∂ ) × i f j S K ( x − x ′ ) × u ( 0 ) ( t ′ , k ) e − i k ⃗ ⋅ Δ x ⃗ (130) \begin{aligned}
-\frac{a^2}{2}\left[\partial_0^2+2aH\partial_0+k^2\right]u^{(1)}(t,k)=\frac{61\kappa^2}{2^5\cdot3\cdot5\cdot\pi^2}ika^2H^2\partial_0u^{(0)}(t,k) \\
&+\int d^4x'\sum_{j=1}^{3}C_{2j}(a,a',\partial)\times if_j^{\mathrm{SK}}(x-x')\times u^{(0)}(t',k)e^{-i\vec{k}\cdot\Delta\vec{x}} \tag*{(130)}
\end{aligned} − 2 a 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( 1 ) ( t , k ) = 2 5 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 ik a 2 H 2 ∂ 0 u ( 0 ) ( t , k ) + ∫ d 4 x ′ j = 1 ∑ 3 C 2 j ( a , a ′ , ∂ ) × i f j SK ( x − x ′ ) × u ( 0 ) ( t ′ , k ) e − i k ⋅ Δ x (130)
At this point it is useful to recall the earlier computation of u ( 1 ) ( t , k ) u^{(1)}(t,k) u ( 1 ) ( t , k ) [23 ] that led to the double log enhancement (6). That analysis employed the D = 4 D=4 D = 4 result for the graviton self-energy [22 ] that agrees with the nonlocal contributions we have obtained in Table 6 and Table 7 , but which was not capable of producing reliable local results such as those we reported in Table 8 and Table 11 , and the second lines of expressions (102) and (103) . Local contributions were guessed in deriving the previous result for u ( 1 ) ( t , k ) u^{(1)}(t,k) u ( 1 ) ( t , k ) [23 ] , and we can now check those guesses.
Note that a double log enhancement of u ( 1 ) ( t , k ) u^{(1)}(t,k) u ( 1 ) ( t , k ) requires that the right hand side of (130) grow like a 4 ln ( a ) a^4\ln(a) a 4 ln ( a ) at late times. This cannot come from the local term on the first line of (130) because ∂ 0 u ( 0 ) ( t , k ) \partial_0u^{(0)}(t,k) ∂ 0 u ( 0 ) ( t , k ) falls off like 1 / a 1/a 1/ a at late times. Nor can it come from the nonlocal contributions of Table 6 and Table 7 , which are the same as for the earlier study, and were shown there to grow no faster than a 4 a^4 a 4 [23 ] . This leaves only the local contributions from the i = 2 i=2 i = 2 rows of Table 8 and Table 11 , and it is simple to show that the result from each one grows no faster than a 4 a^4 a 4 at late times. This means that the double log enhancement (6) found in the earlier calculation is wrong, and was the result of having incorrectly inferred the local parts of the graviton self-energy.
Of course it is still interesting to see whether or not there is any growth in the late time mode function. If the right hand side of (130) went like a 4 a^4 a 4 at late times, that would induce a single logarithm enhancement. This analysis is very messy because many terms contribute, so we give only the final result,
− a 2 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( 1 ) ( t , k ) ⟶ − 3041 κ 2 H 4 a 4 2 9 ⋅ 3 ⋅ π 2 × H 2 k 3 . (131) -\frac{a^2}{2}\left[\partial_0^2+2aH\partial_0+k^2\right]u^{(1)}(t,k)\longrightarrow-\frac{3041\kappa^2H^4a^4}{2^9\cdot3\cdot\pi^2}\times\frac{H}{\sqrt{2k^3}}. \tag*{(131)} − 2 a 2 [ ∂ 0 2 + 2 a H ∂ 0 + k 2 ] u ( 1 ) ( t , k ) ⟶ − 2 9 ⋅ 3 ⋅ π 2 3041 κ 2 H 4 a 4 × 2 k 3 H . (131)
It follows that our predicted late time behavior for the mode function is not (6) but rather,
u ( t , k ) ⟶ H 2 k 3 { 1 + 3041 κ 2 H 2 2 8 ⋅ 3 2 ⋅ π 2 × ln [ a ( t ) ] + ⋯ } . (132) u(t,k)\longrightarrow\frac{H}{\sqrt{2k^3}}\left\{1+\frac{3041\kappa^2H^2}{2^8\cdot3^2\cdot\pi^2}\times\ln[a(t)]+\cdots\right\}. \tag*{(132)} u ( t , k ) ⟶ 2 k 3 H { 1 + 2 8 ⋅ 3 2 ⋅ π 2 3041 κ 2 H 2 × ln [ a ( t )] + ⋯ } . (132)
Relation (133) is not quite the end of the story because it could be eliminated by what appears to be an allowable, finite renormalization. The requirements on counterterms are homogeneity, isotropy, dilatation invariance, and the Ward Identity (86). Each of those requirements is satisfied by a Lagrangian Δ L E x t r a \Delta\mathcal{L}_{\mathrm{Extra}} Δ L Extra ,
Δ L E x t r a = { − H 2 2 h α β h α β − 2 H 3 3 a h α β h 0 α , β − 2 H 3 3 a [ η α β − 2 h 00 ( η α β − 2 δ 0 α δ 0 β ) ] Γ α β 0 + 2 H 2 3 h R − 49 H 2 18 h 00 R + 16 H 2 3 a 2 h R 00 − 16 H 27 a h , 0 R } . (133) \Delta\mathcal{L}_{\mathrm{Extra}}=\left\{-\frac{H^2}{2}h^{\alpha\beta}h_{\alpha\beta}-\frac{2H^3}{3a}h^{\alpha\beta}h_{0\alpha,\beta}-\frac{2H^3}{3a}\left[\eta^{\alpha\beta}-2h_{00}\left(\eta^{\alpha\beta}-2\delta_0^\alpha\delta_0^\beta\right)\right]\Gamma^0_{\alpha\beta}+\frac{2H^2}{3}hR-\frac{49H^2}{18}h_{00}R+\frac{16H^2}{3a^2}hR_{00}-\frac{16H}{27a}h_{,0}R\right\}. \tag*{(133)} Δ L Extra = { − 2 H 2 h α β h α β − 3 a 2 H 3 h α β h 0 α , β − 3 a 2 H 3 [ η α β − 2 h 00 ( η α β − 2 δ 0 α δ 0 β ) ] Γ α β 0 + 3 2 H 2 h R − 18 49 H 2 h 00 R + 3 a 2 16 H 2 h R 00 − 27 a 16 H h , 0 R } . (133)
Table 13 gives the standard form (30) of the associated counterterm. Because the i = 2 i=2 i = 2 contribution is − ( a a ′ ) 2 H 4 × i δ 4 ( x − x ′ ) -(aa')^2H^4\times i\delta^4(x-x') − ( a a ′ ) 2 H 4 × i δ 4 ( x − x ′ ) , adding 3041 κ 2 2 9 ⋅ 3 ⋅ π 2 × Δ L E x t r a \frac{3041\kappa^2}{2^9\cdot3\cdot\pi^2}\times\Delta\mathcal{L}_{\mathrm{Extra}} 2 9 ⋅ 3 ⋅ π 2 3041 κ 2 × Δ L Extra would cancel (132), eliminating any secular enhancement of the mode function.
Table 13. Standard form (30) of the extra counterterm i δ 2 Δ S E x t r a δ h μ ν ( x ) δ h ρ σ ( x ′ ) ∣ d S \left.i\frac{\delta^2\Delta S_{\mathrm{Extra}}}{\delta h_{\mu\nu}(x)\delta h_{\rho\sigma}(x')}\right|_{\mathrm{dS}} i δ h μν ( x ) δ h ρ σ ( x ′ ) δ 2 Δ S Extra dS . Note that the cases of i = 4 , 6 , 8 , 10 , 15 i=4,6,8,10,15 i = 4 , 6 , 8 , 10 , 15 and 17 17 17 follow from reflection (38-39).
Response to a Point Mass
For a stationary point mass M M M the stress-energy is,
8 π G T l i n μ ν ( x ) = 8 π G M × a ( t ) δ 3 ( x ⃗ ) × δ 0 μ δ 0 ν . (134) 8\pi G T^{\mu\nu}_{\mathrm{lin}}(x)=8\pi GM\times a(t)\delta^3(\vec{x})\times\delta^\mu_0\delta^\nu_0 . \tag*{(134)} 8 π G T lin μν ( x ) = 8 π GM × a ( t ) δ 3 ( x ) × δ 0 μ δ 0 ν . (134)
Because we are working in a fixed gauge, four scalar potentials must be used,
κ h 00 = − 2 Ψ , κ h 0 i = − ∂ i Ω , κ h i j = − 2 δ i j Φ − 2 ∂ i ∂ j χ . (135) \kappa h_{00}=-2\Psi\,,\qquad\kappa h_{0i}=-\partial_i\Omega\,,\qquad\kappa h_{ij}=-2\delta_{ij}\Phi-2\partial_i\partial_j\chi\,. \tag*{(135)} κ h 00 = − 2Ψ , κ h 0 i = − ∂ i Ω , κ h ij = − 2 δ ij Φ − 2 ∂ i ∂ j χ . (135)
From equations (119-121) we see,
[ D 00 ρ σ + D k k ρ σ ] κ h ρ σ = − 2 D B Ψ , (136) \left[\mathcal{D}^{00\rho\sigma}+\mathcal{D}^{kk\rho\sigma}\right]\kappa h_{\rho\sigma}=-2D_B\Psi\,, \tag*{(136)} [ D 00 ρ σ + D k k ρ σ ] κ h ρ σ = − 2 D B Ψ , (136)
D 0 i ρ σ κ h ρ σ = 1 2 ∂ i D B Ω , (137) \mathcal{D}^{0i\rho\sigma}\kappa h_{\rho\sigma}=\frac{1}{2}\partial_iD_B\Omega\,, \tag*{(137)} D 0 i ρ σ κ h ρ σ = 2 1 ∂ i D B Ω , (137)
[ D k k ρ σ − 3 ∂ i ∂ j ∇ 2 D i j ρ σ ] κ h ρ σ = 2 ∇ 2 D A χ , (138) \left[\mathcal{D}^{kk\rho\sigma}-\frac{3\partial_i\partial_j}{\nabla^2}\mathcal{D}^{ij\rho\sigma}\right]\kappa h_{\rho\sigma}=2\nabla^2D_A\chi\,, \tag*{(138)} [ D k k ρ σ − ∇ 2 3 ∂ i ∂ j D ij ρ σ ] κ h ρ σ = 2 ∇ 2 D A χ , (138)
[ D k k ρ σ + ∂ i ∂ j ∇ 2 D i j ρ σ ] κ h ρ σ = 2 D A ( Φ − Ψ ) . (139) \left[\mathcal{D}^{kk\rho\sigma}+\frac{\partial_i\partial_j}{\nabla^2}\mathcal{D}^{ij\rho\sigma}\right]\kappa h_{\rho\sigma}=2D_A(\Phi-\Psi) \,. \tag*{(139)} [ D k k ρ σ + ∇ 2 ∂ i ∂ j D ij ρ σ ] κ h ρ σ = 2 D A ( Φ − Ψ ) . (139)
At zeroth order the potentials are,
Ψ ( 0 ) ( t , r ) = Φ ( 0 ) ( t , r ) = G M a ( t ) r , Ω ( 0 ) = χ ( 0 ) = 0 . (140) \Psi^{(0)}(t,r)=\Phi^{(0)}(t,r)=\frac{GM}{a(t)r}\,,\qquad\Omega^{(0)}=\chi^{(0)}=0 \,. \tag*{(140)} Ψ ( 0 ) ( t , r ) = Φ ( 0 ) ( t , r ) = a ( t ) r GM , Ω ( 0 ) = χ ( 0 ) = 0 . (140)
The procedure for finding 1-loop corrections is to consult the appropriate equation from (137-140), and then use Table 12 and relations (116-117) to compute the source. For example, the Newtonian potential obeys,
− 2 D B Ψ ( 1 ) = 13 κ 2 ( F 00 + F k k ) 2 5 ⋅ 3 3 ⋅ π 2 { ln ( 2 μ a H ) R ( 0 ) } − 61 κ 2 ∂ ρ ∂ σ 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 { ln ( 2 μ a H ) [ C ( 0 ) ρ σ 0 + C ( 0 ) ρ k σ k ] } + ∫ d 4 x ′ ∑ i = 1 21 ∑ j = 1 3 C i j ( a , a ′ , ∂ ) { [ D i 00 ρ σ ] + [ D i k k ρ σ ] } × i f j S K ( Δ x ) × κ h ρ σ ( 0 ) ( x ′ ) . (141) \begin{aligned}
-2D_B\Psi^{(1)}={}&\frac{13\kappa^2(F^{00}+F^{kk})}{2^5\cdot3^3\cdot\pi^2}\left\{\ln\left(\frac{2\mu a}{H}\right)R^{(0)}\right\}
-\frac{61\kappa^2\partial_\rho\partial_\sigma}{2^4\cdot3\cdot5\cdot\pi^2}\left\{\ln\left(\frac{2\mu a}{H}\right)\left[C^{(0)\rho\sigma0}+C^{(0)\rho k\sigma k}\right]\right\}\\
&+\int d^4x'\sum_{i=1}^{21}\sum_{j=1}^{3}C_{ij}(a,a',\partial)\left\{\left[D_i^{00\rho\sigma}\right]+\left[D_i^{kk\rho\sigma}\right]\right\}\times i f_j^{\mathrm{SK}}(\Delta x)\times\kappa h_{\rho\sigma}^{(0)}(x') . \tag*{(141)}
\end{aligned} − 2 D B Ψ ( 1 ) = 2 5 ⋅ 3 3 ⋅ π 2 13 κ 2 ( F 00 + F k k ) { ln ( H 2 μ a ) R ( 0 ) } − 2 4 ⋅ 3 ⋅ 5 ⋅ π 2 61 κ 2 ∂ ρ ∂ σ { ln ( H 2 μ a ) [ C ( 0 ) ρ σ 0 + C ( 0 ) ρ k σ k ] } + ∫ d 4 x ′ i = 1 ∑ 21 j = 1 ∑ 3 C ij ( a , a ′ , ∂ ) { [ D i 00 ρ σ ] + [ D i k k ρ σ ] } × i f j SK ( Δ x ) × κ h ρ σ ( 0 ) ( x ′ ) . (141)
The next step would be simplifying the [ D i 00 ρ σ ] × κ h ρ σ ( x ′ ) \left[D_i^{00\rho\sigma}\right]\times\kappa h_{\rho\sigma}(x') [ D i 00 ρ σ ] × κ h ρ σ ( x ′ ) and [ D i k k ρ σ ] × κ h ρ σ ( x ′ ) \left[D_i^{kk\rho\sigma}\right]\times\kappa h_{\rho\sigma}(x') [ D i k k ρ σ ] × κ h ρ σ ( x ′ ) entries of Table 12 by summing k k k on ℓ \ell ℓ and using the 0th order solution. For example, the case of i = 1 i=1 i = 1 would be,
[ D i 00 ρ σ ] × κ h ρ σ ( 0 ) ( x ′ ) + [ D i k k ρ σ ] × κ h ρ σ ( 0 ) ( x ′ ) = − h ( 0 ) ( x ′ ) + 3 h ( 0 ) ( x ′ ) = − 8 G M a ′ r ′ . (142) \left[D_i^{00\rho\sigma}\right]\times\kappa h_{\rho\sigma}^{(0)}(x')+\left[D_i^{kk\rho\sigma}\right]\times\kappa h_{\rho\sigma}^{(0)}(x')=-h^{(0)}(x')+3h^{(0)}(x')=-\frac{8GM}{a'r'}\,. \tag*{(142)} [ D i 00 ρ σ ] × κ h ρ σ ( 0 ) ( x ′ ) + [ D i k k ρ σ ] × κ h ρ σ ( 0 ) ( x ′ ) = − h ( 0 ) ( x ′ ) + 3 h ( 0 ) ( x ′ ) = − a ′ r ′ 8 GM . (142)
However, it isn’t really necessary to go through this because the results from the nonlocal contributions of Table 6 and Table 7 are identical to those of the previous calculation [23 ] , and stronger than anything the local contributions can give. The result for the Newtonian potential is,
− 2 D B Ψ ( 1 ) ⟶ − 3 κ 2 H 4 a 4 ln 2 ( a ) π 2 × G M a r ⟹ Ψ ( 1 ) ( t , r ) ⟶ − κ 2 H 2 ln 3 ( a ) 2 π 2 × G M a r . (143) -2D_B\Psi^{(1)} \longrightarrow-\frac{3\kappa^2 H^4 a^4 \ln^2(a)}{\pi^2} \times\frac{GM}{ar} \quad\Longrightarrow\quad\Psi^{(1)}(t,r) \longrightarrow-\frac{\kappa^2 H^2 \ln^3(a)}{2\pi^2} \times\frac{GM}{ar}. \tag*{(143)} − 2 D B Ψ ( 1 ) ⟶ − π 2 3 κ 2 H 4 a 4 ln 2 ( a ) × a r GM ⟹ Ψ ( 1 ) ( t , r ) ⟶ − 2 π 2 κ 2 H 2 ln 3 ( a ) × a r GM . (143)
The gravitational slip is [23 ] ,
2 D A [ Φ ( 1 ) − Ψ ( 1 ) ] ⟶ 4 κ 2 H 4 a 4 ln 2 ( a ) π 2 × G M a r 2D_A[\Phi^{(1)}-\Psi^{(1)}] \longrightarrow\frac{4\kappa^2 H^4 a^4 \ln^2(a)}{\pi^2} \times\frac{GM}{ar} 2 D A [ Φ ( 1 ) − Ψ ( 1 ) ] ⟶ π 2 4 κ 2 H 4 a 4 ln 2 ( a ) × a r GM
⟹ Φ ( 1 ) − Ψ ( 1 ) ⟶ − κ 2 H 2 ln 2 ( a ) π 2 × G M a r . (144) \Longrightarrow\Phi^{(1)}-\Psi^{(1)} \longrightarrow-\frac{\kappa^2 H^2 \ln^2(a)}{\pi^2} \times\frac{GM}{ar}. \tag*{(144)} ⟹ Φ ( 1 ) − Ψ ( 1 ) ⟶ − π 2 κ 2 H 2 ln 2 ( a ) × a r GM . (144)
None of the other potentials shows secular enhancement.
Conclusions
We have computed the 1-loop graviton contribution to the graviton self-energy on de Sitter background. Unlike contributions from matter loops [35 , 45 , 46 , 47 , 48 ] , the graviton loop contribution is gauge dependent, and our result was derived using the simplest gauge [25 , 27 ] . In § 3 we computed the primitive result in position-space, using dimensional regularization. The result consists of divergent local contributions in Table 2 and Table 4 , plus finite, nonlocal contributions in Table 6 and Table 7 . The latter agree with the old, D = 4 D=4 D = 4 computation [22 ] . In § 4 the ultraviolet divergences were absorbed using BPHZ counterterms given in Table 10 . The final, fully renormalized result consists of the nonlocal contributions in Table 6 and Table 7 , plus the local contributions of Table 8 and Table 11 , and the important local logarithms terms on the second lines of expressions (102) and (103) . This final result obeys the Ward identity (85) .
The reason for computing − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[{}^{\mu\nu}\Sigma^{\rho\sigma}](x;x^{\prime}) − i [ μν Σ ρ σ ] ( x ; x ′ ) is so that we can use it in the linearized Einstein equation (3) to derive 1-loop corrections to the graviton mode function and to the Newtonian potential. A previous analysis [23 , 24 ] obtained relations (6-7) by employing an old, D = 4 D=4 D = 4 result [22 ] that got the nonlocal part right but was not capable of reliably fixing the local part. The previous analysis was based on having made a plausible guess for the local part. Now that the full self-energy has been derived we can see that this guess was not correct. One consequence is that the graviton mode function does not experience the double log enhancement (6) , but only a single log enhancement (132) . Even that could be nulled by making what seems to be an allowable, finite renormalization involving (133) . On the other hand, the triple log enhancement of the Newtonian potential (7) was sourced entirely from the nonlocal contributions that the D = 4 D = 4 D = 4 calculation gets right. Our result (143) for the Newtonian potential is therefore unchanged, as is our result (144) for the gravitational slip.
The possible extra counterterm (133) highlights the importance of properly constraining counterterms. Even the list of 19 counterterms given in Table 10 seems excessive. The graviton self-energy can be expressed as a linear combination (30) of the 21 basis tensors [ μ ν D i ρ σ ] [\mu\nu D_i^{\rho\sigma}] [ μν D i ρ σ ] given in Table 1 . Seven of these elements are related by reflection invariance (36) , which leaves 14 algebraically independent terms. The Ward identity (85) implies relations proportional to the four tensors,
η μ ρ , ∂ μ ∂ ρ , ∂ ( μ δ 0 ρ ) , δ 0 μ δ 0 ρ . (145) \eta^{\mu\rho}, \qquad\partial^\mu\partial^\rho, \qquad\partial^{(\mu}\delta_0^{\rho)}, \qquad\delta_0^\mu\delta_0^\rho. \tag*{(145)} η μ ρ , ∂ μ ∂ ρ , ∂ ( μ δ 0 ρ ) , δ 0 μ δ 0 ρ . (145)
So one expects only ten independent coefficients, not 19. Some attention should be given to simplifying and consolidating the list of Table 10 , with the potential payoff of deciding whether or not the 1-loop correction to the graviton mode function has a logarithmic enhancement.
A major application of this work is provided by the formulae for products of differentiated propagators presented in the Appendix. These formulae can be used to compute the graviton self-energy in the gauge introduced to facilitate the nonperturbative resummation of secular effects [49 ] . This computation is vital to check the nonperturbative stochastic equations against explicit perturbative results derived in the same gauge. The same formulae from the Appendix also facilitate the purging of graviton gauge dependence from − i [ μ ν Σ ρ σ ] ( x ; x ′ ) -i[\mu\nu\Sigma^{\rho\sigma}](x;x') − i [ μν Σ ρ σ ] ( x ; x ′ ) following the technique that has been successfully applied on flat space background for the scalar self-mass [50 ] and for the vacuum polarization [51 ] , and on de Sitter background for the scalar self-mass [52 , 53 , 54 ] .
Acknowledgements
This work was partially supported by NSF grant PHY-2207514 and by the Institute for Fundamental Theory at the University of Florida.
Appendix: Products of Propagators
The purpose of this appendix is to expand products of two propagators, each with up to two derivatives, into a series of divergent and finite terms. The same procedure was applied in computing the scalar loop contribution [46 ] for the special case of doubly differentiated A A A -type propagators. Our starting point is the observation that each of the three propagators i Δ I ( x ; x ′ ) i\Delta_I(x;x') i Δ I ( x ; x ′ ) , for I = A , B , C I=A,B,C I = A , B , C can be expanded in terms of contributions which are less and less singular at x μ = x ′ μ x^\mu=x'^\mu x μ = x ′ μ ,
i Δ I ( x ; x ′ ) = i Δ c f ( x ; x ′ ) + ∑ n = 0 ∞ I n ( x ; x ′ ) . (146) i\Delta_I(x;x')=i\Delta_{\rm cf}(x;x')+\sum_{n=0}^{\infty}I_n(x;x') . \tag*{(146)} i Δ I ( x ; x ′ ) = i Δ cf ( x ; x ′ ) + n = 0 ∑ ∞ I n ( x ; x ′ ) . (146)
The most singular contribution is the propagator os a massless, conformally coupled scalar,
i Δ c f ( x ; x ′ ) = 1 4 π D / 2 Γ ( D 2 − 1 ) [ a a ′ Δ x 2 ] D 2 − 1 . (147) i\Delta_{\rm cf}(x;x')=\frac{1}{4\pi^{D/2}}\frac{\Gamma\left(\frac{D}{2}-1\right)}{[aa'\Delta x^2]^{\frac{D}{2}-1}} . \tag*{(147)} i Δ cf ( x ; x ′ ) = 4 π D /2 1 [ a a ′ Δ x 2 ] 2 D − 1 Γ ( 2 D − 1 ) . (147)
It turns out that we need only the first two of the functions I n ( x ; x ′ ) I_n(x;x') I n ( x ; x ′ ) ,
A 0 ( x ; x ′ ) = H 2 8 π D / 2 Γ ( D 2 + 1 ) D 2 − 2 1 [ a a ′ Δ x 2 ] D 2 − 2 − H D − 2 ( 4 π ) D 2 Γ ( D − 1 ) Γ ( D 2 ) [ π cot ( D π 2 ) − ln ( a a ′ ) ] , (148) A_0(x;x')=\frac{H^2}{8\pi^{D/2}}\frac{\Gamma\left(\frac{D}{2}+1\right)}{\frac{D}{2}-2}\frac{1}{[aa'\Delta x^2]^{\frac{D}{2}-2}}-\frac{H^{D-2}}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D-1)}{\Gamma\left(\frac{D}{2}\right)}\left[\pi\cot\left(\frac{D\pi}{2}\right)-\ln(aa')\right] , \tag*{(148)} A 0 ( x ; x ′ ) = 8 π D /2 H 2 2 D − 2 Γ ( 2 D + 1 ) [ a a ′ Δ x 2 ] 2 D − 2 1 − ( 4 π ) 2 D H D − 2 Γ ( 2 D ) Γ ( D − 1 ) [ π cot ( 2 D π ) − ln ( a a ′ ) ] , (148)
A 1 ( x ; x ′ ) = H 4 64 π D 2 Γ ( D 2 + 2 ) D 2 − 3 1 [ a a ′ Δ x 2 ] D 2 − 3 + H D ( 4 π ) D 2 Γ ( D ) Γ ( D 2 + 1 ) × 1 4 a a ′ Δ x 2 , (149) A_1(x;x')=\frac{H^4}{64\pi^{\frac{D}{2}}}\frac{\Gamma\left(\frac{D}{2}+2\right)}{\frac{D}{2}-3}\frac{1}{[aa'\Delta x^2]^{\frac{D}{2}-3}}+\frac{H^D}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D)}{\Gamma\left(\frac{D}{2}+1\right)}\times\frac{1}{4}aa'\Delta x^2 , \tag*{(149)} A 1 ( x ; x ′ ) = 64 π 2 D H 4 2 D − 3 Γ ( 2 D + 2 ) [ a a ′ Δ x 2 ] 2 D − 3 1 + ( 4 π ) 2 D H D Γ ( 2 D + 1 ) Γ ( D ) × 4 1 a a ′ Δ x 2 , (149)
B 0 ( x ; x ′ ) = H 2 16 π D 2 Γ ( D 2 ) [ a a ′ Δ x 2 ] D 2 − 2 − H D − 2 ( 4 π ) D 2 Γ ( D − 2 ) Γ ( D 2 ) , (150) B_0(x;x')=\frac{H^2}{16\pi^{\frac{D}{2}}}\frac{\Gamma\left(\frac{D}{2}\right)}{[aa'\Delta x^2]^{\frac{D}{2}-2}}-\frac{H^{D-2}}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D-2)}{\Gamma\left(\frac{D}{2}\right)} , \tag*{(150)} B 0 ( x ; x ′ ) = 16 π 2 D H 2 [ a a ′ Δ x 2 ] 2 D − 2 Γ ( 2 D ) − ( 4 π ) 2 D H D − 2 Γ ( 2 D ) Γ ( D − 2 ) , (150)
B 1 ( x ; x ′ ) = H 4 128 π D 2 Γ ( D 2 + 1 ) [ a a ′ Δ x 2 ] D 2 − 3 − H D ( 4 π ) D 2 Γ ( D − 1 ) Γ ( D 2 + 1 ) × 1 4 a a ′ Δ x 2 , (151) B_1(x;x')=\frac{H^4}{128\pi^{\frac{D}{2}}}\frac{\Gamma\left(\frac{D}{2}+1\right)}{[aa'\Delta x^2]^{\frac{D}{2}-3}}-\frac{H^D}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D-1)}{\Gamma\left(\frac{D}{2}+1\right)}\times\frac{1}{4}aa'\Delta x^2 , \tag*{(151)} B 1 ( x ; x ′ ) = 128 π 2 D H 4 [ a a ′ Δ x 2 ] 2 D − 3 Γ ( 2 D + 1 ) − ( 4 π ) 2 D H D Γ ( 2 D + 1 ) Γ ( D − 1 ) × 4 1 a a ′ Δ x 2 , (151)
C 0 ( x ; x ′ ) = H 2 32 π D 2 ( D − 6 ) Γ ( D 2 − 1 ) 1 [ a a ′ Δ x 2 ] D 2 − 2 + H D − 2 ( 4 π ) D 2 Γ ( D − 3 ) Γ ( D 2 ) , (152) C_0(x;x')=\frac{H^2}{32\pi^{\frac{D}{2}}}(D-6)\Gamma\left(\frac{D}{2}-1\right)\frac{1}{[aa'\Delta x^2]^{\frac{D}{2}-2}}+\frac{H^{D-2}}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D-3)}{\Gamma\left(\frac{D}{2}\right)} , \tag*{(152)} C 0 ( x ; x ′ ) = 32 π 2 D H 2 ( D − 6 ) Γ ( 2 D − 1 ) [ a a ′ Δ x 2 ] 2 D − 2 1 + ( 4 π ) 2 D H D − 2 Γ ( 2 D ) Γ ( D − 3 ) , (152)
C 1 ( x ; x ′ ) = H 4 256 π D 2 ( D − 8 ) Γ ( D 2 ) [ a a ′ Δ x 2 ] D 2 − 3 + H D ( 4 π ) D 2 Γ ( D − 2 ) Γ ( D 2 + 1 ) × 1 2 a a ′ Δ x 2 . (153) C_1(x;x')=\frac{H^4}{256\pi^{\frac{D}{2}}}(D-8)\frac{\Gamma\left(\frac{D}{2}\right)}{[aa'\Delta x^2]^{\frac{D}{2}-3}}+\frac{H^D}{(4\pi)^{\frac{D}{2}}}\frac{\Gamma(D-2)}{\Gamma\left(\frac{D}{2}+1\right)}\times\frac{1}{2}aa'\Delta x^2 . \tag*{(153)} C 1 ( x ; x ′ ) = 256 π 2 D H 4 ( D − 8 ) [ a a ′ Δ x 2 ] 2 D − 3 Γ ( 2 D ) + ( 4 π ) 2 D H D Γ ( 2 D + 1 ) Γ ( D − 2 ) × 2 1 a a ′ Δ x 2 . (153)
Note that only A 0 ( x ; x ′ ) A_0(x;x') A 0 ( x ; x ′ ) has a nonzero limit for D = 4 D=4 D = 4 ,
lim D → 4 A 0 ( x ; x ′ ) = − H 2 8 π 2 ln ( 1 4 H 2 Δ x 2 ) . (154) \lim_{D\to4}A_0(x;x')=-\frac{H^2}{8\pi^2}\ln\left(\frac{1}{4}H^2\Delta x^2\right) . \tag*{(154)} D → 4 lim A 0 ( x ; x ′ ) = − 8 π 2 H 2 ln ( 4 1 H 2 Δ x 2 ) . (154)
Our goal is to employ the graviton self-energy in the linearized Einstein equation (3) . This has two important consequences:
We can accordingly adopt the 5-step reduction procedure that was pioneered in [55 , 56 ] :
Act any derivatives on each of the two propagators and multiply them together;
Extract derivatives with respect to x μ x^\mu x μ from factors of 1 / Δ x 2 1/\Delta x^2 1/Δ x 2 until reaching a form like 1 / Δ x 2 D − 6 1/\Delta x^{2D-6} 1/Δ x 2 D − 6 that is integrable for D = 4 D=4 D = 4 ;
Isolate divergences by adding factors of zero in the form,
0 = 4 π D / 2 i δ D ( x − x ′ ) Γ ( D 2 − 1 ) − ∂ 2 1 Δ x D − 2 ; (155) 0=\frac{4\pi^{D/2}i\delta^D(x-x')}{\Gamma\left(\frac{D}{2}-1\right)}-\partial^2\frac{1}{\Delta x^{D-2}}\ ; \tag*{(155)} 0 = Γ ( 2 D − 1 ) 4 π D /2 i δ D ( x − x ′ ) − ∂ 2 Δ x D − 2 1 ; (155)
Take the limit D = 4 D=4 D = 4 on any finite, integrable terms; and
Eliminate any remaining factors of 1 / Δ x 2 1/\Delta x^2 1/Δ x 2 using the identities,
1 Δ x 2 = ∂ 2 4 ln ( 1 4 H 2 Δ x 2 ) , (156) \frac{1}{\Delta x^2}=\frac{\partial^2}{4}\ln\left(\frac{1}{4}H^2\Delta x^2\right)\ , \tag*{(156)} Δ x 2 1 = 4 ∂ 2 ln ( 4 1 H 2 Δ x 2 ) , (156)
ln ( 1 4 H 2 Δ x 2 ) Δ x 2 = ∂ 2 8 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] . (157) \frac{\ln\left(\frac{1}{4}H^2\Delta x^2\right)}{\Delta x^2}
=\frac{\partial^2}{8}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right]\ . \tag*{(157)} Δ x 2 ln ( 4 1 H 2 Δ x 2 ) = 8 ∂ 2 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] . (157)
Note that Step 1 can produce delta functions when a constant part of one propagator multiplies the derivative,
∂ μ ∂ ρ ′ 1 Δ x D − 2 = − δ μ 0 δ ρ 0 4 π D / 2 i δ D ( x − x ′ ) Γ ( D 2 − 1 ) + ( D − 2 ) [ η μ ρ Δ x D − D Δ x μ Δ x ρ Δ x D + 2 ] . (158) \partial_\mu\partial'_\rho\frac{1}{\Delta x^{D-2}}
=-\delta^0_\mu\delta^0_\rho\frac{4\pi^{D/2}i\delta^D(x-x')}{\Gamma\left(\frac{D}{2}-1\right)}
+(D-2)\left[\frac{\eta_{\mu\rho}}{\Delta x^D}-\frac{D\Delta x_\mu\Delta x_\rho}{\Delta x^{D+2}}\right]\ . \tag*{(158)} ∂ μ ∂ ρ ′ Δ x D − 2 1 = − δ μ 0 δ ρ 0 Γ ( 2 D − 1 ) 4 π D /2 i δ D ( x − x ′ ) + ( D − 2 ) [ Δ x D η μ ρ − Δ x D + 2 D Δ x μ Δ x ρ ] . (158)
However, these delta functions will always be absorbed when the derivatives are extracted in Step 2.
The factor of zero (156) called for in Step 3 typically needs to be multiplied by a dimensional factor to achieve consistent units. It will simplify our analysis by choosing this factor to be ( H / 2 ) D − 4 (H/2)^{D-4} ( H /2 ) D − 4 ,
∂ 2 D − 4 [ 1 Δ x 2 D − 6 ] = ( H 2 ) D − 4 D − 4 × 4 π D / 2 i δ D ( x − x ′ ) Γ ( D 2 − 1 ) + ∂ 2 D − 4 [ 1 Δ x 2 D − 6 − ( H 2 ) D − 4 Δ x D − 2 ] , (159) \frac{\partial^2}{D-4}\left[\frac{1}{\Delta x^{2D-6}}\right]
=\frac{\left(\frac{H}{2}\right)^{D-4}}{D-4}\times\frac{4\pi^{D/2}i\delta^D(x-x')}{\Gamma\left(\frac{D}{2}-1\right)}
+\frac{\partial^2}{D-4}\left[\frac{1}{\Delta x^{2D-6}}-\frac{\left(\frac{H}{2}\right)^{D-4}}{\Delta x^{D-2}}\right]\ , \tag*{(159)} D − 4 ∂ 2 [ Δ x 2 D − 6 1 ] = D − 4 ( 2 H ) D − 4 × Γ ( 2 D − 1 ) 4 π D /2 i δ D ( x − x ′ ) + D − 4 ∂ 2 [ Δ x 2 D − 6 1 − Δ x D − 2 ( 2 H ) D − 4 ] , (159)
= ( H 2 ) D − 4 D − 4 × 4 π D / 2 i δ D ( x − x ′ ) Γ ( D 2 − 1 ) − ∂ 2 2 [ ln ( 1 4 H 2 Δ x 2 ) Δ x 2 ] + O ( D − 4 ) , (160) =\frac{\left(\frac{H}{2}\right)^{D-4}}{D-4}\times\frac{4\pi^{D/2}i\delta^D(x-x')}{\Gamma\left(\frac{D}{2}-1\right)}
-\frac{\partial^2}{2}\left[\frac{\ln\left(\frac{1}{4}H^2\Delta x^2\right)}{\Delta x^2}\right]+O(D-4)\ , \tag*{(160)} = D − 4 ( 2 H ) D − 4 × Γ ( 2 D − 1 ) 4 π D /2 i δ D ( x − x ′ ) − 2 ∂ 2 [ Δ x 2 ln ( 4 1 H 2 Δ x 2 ) ] + O ( D − 4 ) , (160)
⟶ ( H 2 ) D − 4 D − 4 × 4 π D / 2 i δ D ( x − x ′ ) Γ ( D 2 − 1 ) − ∂ 4 16 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] . (161) \longrightarrow\frac{\left(\frac{H}{2}\right)^{D-4}}{D-4}\times\frac{4\pi^{D/2}i\delta^D(x-x')}{\Gamma\left(\frac{D}{2}-1\right)}
-\frac{\partial^4}{16}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right]\ . \tag*{(161)} ⟶ D − 4 ( 2 H ) D − 4 × Γ ( 2 D − 1 ) 4 π D /2 i δ D ( x − x ′ ) − 16 ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] . (161)
We express all divergences as functions of D D D times the constant K 3 \mathcal{K}_3 K 3 , which was defined in (75) and is repeated here,
K 3 = H D − 4 ( 4 π ) D / 2 Γ ( D 2 − 1 ) 2 ( D − 3 ) ( D − 4 ) = 1 32 π 2 H D − 4 D − 4 + F i n i t e . (162) \mathcal{K}_3 = \frac{H^{D-4}}{(4\pi)^{D/2}}\frac{\Gamma\left(\frac{D}{2}-1\right)}{2(D-3)(D-4)} = \frac{1}{32\pi^2}\frac{H^{D-4}}{D-4}+\mathrm{Finite}. \tag*{(162)} K 3 = ( 4 π ) D /2 H D − 4 2 ( D − 3 ) ( D − 4 ) Γ ( 2 D − 1 ) = 32 π 2 1 D − 4 H D − 4 + Finite . (162)
For example, the product of two conformal propagators is,
( a a ′ ) D − 2 i Δ c f × i Δ c f = Γ 2 ( D 2 − 1 ) 16 π D Δ x 2 D − 4 = Γ 2 ( D 2 − 1 ) 16 π D ∂ 2 2 ( D − 3 ) ( D − 4 ) 1 Δ x 2 D − 6 , (163) (aa')^{D-2}i\Delta_{\mathrm{cf}}\times i\Delta_{\mathrm{cf}}=\frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D\Delta x^{2D-4}}=\frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\partial^2}{2(D-3)(D-4)}\frac{1}{\Delta x^{2D-6}}, \tag*{(163)} ( a a ′ ) D − 2 i Δ cf × i Δ cf = 16 π D Δ x 2 D − 4 Γ 2 ( 2 D − 1 ) = 16 π D Γ 2 ( 2 D − 1 ) 2 ( D − 3 ) ( D − 4 ) ∂ 2 Δ x 2 D − 6 1 , (163)
⟶ 4 K 3 i δ D ( x − x ′ ) − ∂ 4 2 9 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] . (164) \longrightarrow4\mathcal{K}_3i\delta^D(x-x')-\frac{\partial^4}{2^9\cdot\pi^4}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right]. \tag*{(164)} ⟶ 4 K 3 i δ D ( x − x ′ ) − 2 9 ⋅ π 4 ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] . (164)
All products of the form ( a a ′ ) D − 2 i Δ I ( x ; x ′ ) × i Δ J ( x ; x ′ ) (aa')^{D-2}i\Delta_I(x;x')\times i\Delta_J(x;x') ( a a ′ ) D − 2 i Δ I ( x ; x ′ ) × i Δ J ( x ; x ′ ) can be read off of Table 14 .
Table 14. Contributions from products of the form ( a a ′ ) D − 2 i Δ I × i Δ J (aa')^{D-2}i\Delta_I\times i\Delta_J ( a a ′ ) D − 2 i Δ I × i Δ J .
For example, the case of I = A I=A I = A and J = A J=A J = A is,
( a a ′ ) D − 2 i Δ A × i Δ A = ( a a ′ ) D − 2 { i Δ c f + A 0 + … } { i Δ c f + A 0 + … } , (165) (aa')^{D-2}i\Delta_A\times i\Delta_A=(aa')^{D-2}\left\{i\Delta_{\mathrm{cf}}+A_0+\ldots\right\}\left\{i\Delta_{\mathrm{cf}}+A_0+\ldots\right\}, \tag*{(165)} ( a a ′ ) D − 2 i Δ A × i Δ A = ( a a ′ ) D − 2 { i Δ cf + A 0 + … } { i Δ cf + A 0 + … } , (165)
= ( a a ′ ) D − 2 { i Δ c f × i Δ c f + i Δ c f × A 0 + A 0 × i Δ c f + A 0 × A 0 + … } , = 4 K 3 i δ D ( x − x ′ ) − [ ∂ 4 + 4 H 2 a a ′ ∂ 2 − 4 H 4 ( a a ′ ) 2 ] 2 9 ⋅ π 4 ln 2 ( 1 4 H 2 Δ x 2 ) (166) \begin{aligned}
=(aa')^{D-2}\left\{i\Delta_{\mathrm{cf}}\times i\Delta_{\mathrm{cf}}+i\Delta_{\mathrm{cf}}\times A_0+A_0\times i\Delta_{\mathrm{cf}}+A_0\times A_0+\ldots\right\}, \\
&=4\mathcal{K}_3i\delta^D(x-x')-\frac{\left[\partial^4+4H^2aa'\partial^2-4H^4(aa')^2\right]}{2^9\cdot\pi^4}\ln^2\left(\frac{1}{4}H^2\Delta x^2\right) \tag*{(166)}
\end{aligned} = ( a a ′ ) D − 2 { i Δ cf × i Δ cf + i Δ cf × A 0 + A 0 × i Δ cf + A 0 × A 0 + … } , = 4 K 3 i δ D ( x − x ′ ) − 2 9 ⋅ π 4 [ ∂ 4 + 4 H 2 a a ′ ∂ 2 − 4 H 4 ( a a ′ ) 2 ] ln 2 ( 4 1 H 2 Δ x 2 ) (166)
+ [ ∂ 4 + 4 H 2 a a ′ ∂ 2 ] 2 8 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) + O ( D − 4 ) . (167) \qquad+\frac{\left[\partial^4+4H^2aa'\partial^2\right]}{2^8\cdot\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right)+O(D-4). \tag*{(167)} + 2 8 ⋅ π 4 [ ∂ 4 + 4 H 2 a a ′ ∂ 2 ] ln ( 4 1 H 2 Δ x 2 ) + O ( D − 4 ) . (167)
The higher A n A_n A n ’s are not needed because these terms are integrable, even when multiplied by i Δ c f i\Delta_{\mathrm{cf}} i Δ cf , and they vanish for D = 4 D=4 D = 4 . For the same reason, none of the B n B_n B n ’s or C n C_n C n ’s are necessary — not even B 0 B_0 B 0 or C 0 C_0 C 0 . For example, the case of I = A I=A I = A and J = B J=B J = B is,
( a a ′ ) D − 2 i Δ A × i Δ B = ( a a ′ ) D − 2 { i Δ c f + A 0 + ⋯ } { i Δ c f + ⋯ } , (168) (aa')^{D-2}i\Delta_A\times i\Delta_B=(aa')^{D-2}\left\{i\Delta_{\mathrm{cf}}+A_0+\cdots\right\}\left\{i\Delta_{\mathrm{cf}}+\cdots\right\}, \tag*{(168)} ( a a ′ ) D − 2 i Δ A × i Δ B = ( a a ′ ) D − 2 { i Δ cf + A 0 + ⋯ } { i Δ cf + ⋯ } , (168)
= ( a a ′ ) D − 2 { i Δ c f × i Δ c f + i Δ c f × A 0 + ⋯ } , (169) =(aa')^{D-2}\left\{i\Delta_{\mathrm{cf}}\times i\Delta_{\mathrm{cf}}+i\Delta_{\mathrm{cf}}\times A_0+\cdots\right\}, \tag*{(169)} = ( a a ′ ) D − 2 { i Δ cf × i Δ cf + i Δ cf × A 0 + ⋯ } , (169)
⟶ 4 K 3 i δ D ( x − x ′ ) − ∂ 4 + 2 H 2 a a ′ ∂ 2 2 9 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] . (170) \longrightarrow4K_3i\delta^D(x-x')-\frac{\partial^4+2H^2aa'\partial^2}{2^9\cdot\pi^4}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right]. \tag*{(170)} ⟶ 4 K 3 i δ D ( x − x ′ ) − 2 9 ⋅ π 4 ∂ 4 + 2 H 2 a a ′ ∂ 2 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] . (170)
Products of the form ( a a ′ ) D − 2 i Δ I × i Δ J (aa')^{D-2}i\Delta_I\times i\Delta_J ( a a ′ ) D − 2 i Δ I × i Δ J can be read off from Table 15 .
Acting ∂ μ \partial_\mu ∂ μ on i Δ I ( x ; x ′ ) i\Delta_I(x;x') i Δ I ( x ; x ′ ) produces either a factor of a H δ μ 0 aH\delta_\mu^0 a H δ μ 0 or Δ x μ / Δ 2 \Delta x_\mu/\Delta^2 Δ x μ / Δ 2 . After multiplying by ( a a ′ ) D − 2 i Δ J ( x ; x ′ ) (aa')^{D-2}i\Delta_J(x;x') ( a a ′ ) D − 2 i Δ J ( x ; x ′ ) , the first sort of factor simply multiplies a reduction involving Table 14 . The factor of Δ x μ / Δ x 2 \Delta x_\mu/\Delta x^2 Δ x μ /Δ x 2 requires that we extract a derivative using the identities,
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x 2 D − 4 ⟶ − ∂ μ ∂ 2 2 7 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) , (171) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu}{\Delta x^{2D-4}}\longrightarrow-\frac{\partial_\mu\partial^2}{2^7\cdot\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right), \tag*{(171)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D − 4 Δ x μ ⟶ − 2 7 ⋅ π 4 ∂ μ ∂ 2 ln ( 4 1 H 2 Δ x 2 ) , (171)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x 2 D − 2 ⟶ − 2 K 3 ∂ μ i δ D ( Δ x ) D − 2 + ∂ μ ∂ 4 2 11 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] , (172) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu}{\Delta x^{2D-2}}\longrightarrow-\frac{2K_3\partial_\mu i\delta^D(\Delta x)}{D-2}+\frac{\partial_\mu\partial^4}{2^{11}\cdot\pi^4}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right], \tag*{(172)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D − 2 Δ x μ ⟶ − D − 2 2 K 3 ∂ μ i δ D ( Δ x ) + 2 11 ⋅ π 4 ∂ μ ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] , (172)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x 2 D ⟶ − K 3 ∂ μ i δ D ( Δ x ) ( D − 1 ) ( D − 2 ) 2 + ∂ μ ∂ 4 2 13 ⋅ 3 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] . (173) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu}{\Delta x^{2D}}\longrightarrow-\frac{K_3\partial_\mu i\delta^D(\Delta x)}{(D-1)(D-2)^2}+\frac{\partial_\mu\partial^4}{2^{13}\cdot3\cdot\pi^4}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right]. \tag*{(173)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D Δ x μ ⟶ − ( D − 1 ) ( D − 2 ) 2 K 3 ∂ μ i δ D ( Δ x ) + 2 13 ⋅ 3 ⋅ π 4 ∂ μ ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] . (173)
Products of the form ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J (aa')^{D-2}\partial_\mu i\Delta_I\times\partial_\nu i\Delta_J ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J are given by Table 16 .
Table 16. Contributions from products of the form ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J (aa')^{D-2}\partial_{\mu}i\Delta_{I}\times\partial_{\nu}i\Delta_{J} ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J .
Note that the higher superficial degree of divergence allows B 0 B_{0} B 0 and C 0 C_{0} C 0 to contribute. Acting the two derivatives can produce the factor Δ x μ Δ x ν \Delta x_{\mu}\Delta x_{\nu} Δ x μ Δ x ν , among other things. The other factors can be reduced using Tables 14 and 15. We extract derivatives from terms involving Δ x μ Δ x ν \Delta x_{\mu}\Delta x_{\nu} Δ x μ Δ x ν using the identities,
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x 2 D − 2 ⟶ κ 3 η μ ν i δ D ( Δ x ) 2 ( D − 2 ) − η μ ν ∂ 4 2 11 ⋅ π 4 ln 2 ( 1 4 H 2 Δ x 2 ) + [ η μ ν ∂ 2 + 2 ∂ μ ∂ ν ] ∂ 2 2 10 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) (174) \frac{\Gamma^{2}\left(\frac{D}{2}-1\right)}{16\pi^{D}}\frac{\Delta x_{\mu}\Delta x_{\nu}}{\Delta x^{2D-2}}\longrightarrow\frac{\kappa_{3}\eta_{\mu\nu}i\delta^{D}(\Delta x)}{2(D-2)}-\frac{\eta_{\mu\nu}\partial^{4}}{2^{11}\cdot\pi^{4}}\ln^{2}\left(\frac{1}{4}H^{2}\Delta x^{2}\right)+\frac{[\eta_{\mu\nu}\partial^{2}+2\partial_{\mu}\partial_{\nu}]\partial^{2}}{2^{10}\cdot\pi^{4}}\ln\left(\frac{1}{4}H^{2}\Delta x^{2}\right) \tag*{(174)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D − 2 Δ x μ Δ x ν ⟶ 2 ( D − 2 ) κ 3 η μν i δ D ( Δ x ) − 2 11 ⋅ π 4 η μν ∂ 4 ln 2 ( 4 1 H 2 Δ x 2 ) + 2 10 ⋅ π 4 [ η μν ∂ 2 + 2 ∂ μ ∂ ν ] ∂ 2 ln ( 4 1 H 2 Δ x 2 ) (174)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x 2 D ⟶ κ 3 [ η μ ν ∂ 2 + ( D − 2 ) ∂ μ ∂ ν ] i δ D ( Δ x ) ( D − 1 ) ( D − 2 ) 2 − [ η μ ν ∂ 2 + 2 ∂ μ ∂ ν ] ∂ 4 2 13 ⋅ 3 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] (175) \frac{\Gamma^{2}\left(\frac{D}{2}-1\right)}{16\pi^{D}}\frac{\Delta x_{\mu}\Delta x_{\nu}}{\Delta x^{2D}}\longrightarrow\frac{\kappa_{3}[\eta_{\mu\nu}\partial^{2}+(D-2)\partial_{\mu}\partial_{\nu}]i\delta^{D}(\Delta x)}{(D-1)(D-2)^{2}}-\frac{[\eta_{\mu\nu}\partial^{2}+2\partial_{\mu}\partial_{\nu}]\partial^{4}}{2^{13}\cdot3\cdot\pi^{4}}\left[\ln^{2}\left(\frac{1}{4}H^{2}\Delta x^{2}\right)-2\ln\left(\frac{1}{4}H^{2}\Delta x^{2}\right)\right] \tag*{(175)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D Δ x μ Δ x ν ⟶ ( D − 1 ) ( D − 2 ) 2 κ 3 [ η μν ∂ 2 + ( D − 2 ) ∂ μ ∂ ν ] i δ D ( Δ x ) − 2 13 ⋅ 3 ⋅ π 4 [ η μν ∂ 2 + 2 ∂ μ ∂ ν ] ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] (175)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x 2 D + 2 ⟶ κ 3 [ η μ ν ∂ 2 + D ∂ μ ∂ ν ] ∂ 2 i δ D ( Δ x ) 2 D 2 ( D − 1 ) ( D − 2 ) 2 − [ η μ ν ∂ 2 + 4 ∂ μ ∂ ν ] ∂ 6 2 18 ⋅ 3 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] (176) \frac{\Gamma^{2}\left(\frac{D}{2}-1\right)}{16\pi^{D}}\frac{\Delta x_{\mu}\Delta x_{\nu}}{\Delta x^{2D+2}}\longrightarrow\frac{\kappa_{3}[\eta_{\mu\nu}\partial^{2}+D\partial_{\mu}\partial_{\nu}]\partial^{2}i\delta^{D}(\Delta x)}{2D^{2}(D-1)(D-2)^{2}}-\frac{[\eta_{\mu\nu}\partial^{2}+4\partial_{\mu}\partial_{\nu}]\partial^{6}}{2^{18}\cdot3\cdot\pi^{4}}\left[\ln^{2}\left(\frac{1}{4}H^{2}\Delta x^{2}\right)-2\ln\left(\frac{1}{4}H^{2}\Delta x^{2}\right)\right] \tag*{(176)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D + 2 Δ x μ Δ x ν ⟶ 2 D 2 ( D − 1 ) ( D − 2 ) 2 κ 3 [ η μν ∂ 2 + D ∂ μ ∂ ν ] ∂ 2 i δ D ( Δ x ) − 2 18 ⋅ 3 ⋅ π 4 [ η μν ∂ 2 + 4 ∂ μ ∂ ν ] ∂ 6 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] (176)
When both derivatives are primed the result follows from Table 16 by merely interchanging primed and unprimed coordinates. When only one of the two derivatives is primed the result can be read off from Table 17 .
Table 17. Contributions from products of the form ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ρ ′ i Δ J (aa')^{D-2}\partial_{\mu}i\Delta_{I}\times\partial'_{\rho}i\Delta_{J} ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ρ ′ i Δ J .
The case of two derivatives of the same coordinate acting on the one propagator times an undifferentiated propagator can be read off from earlier tables using the relation,
( a a ′ ) D − 2 ∂ μ ∂ ν i Δ I × i Δ J = ( a a ′ ) D − 2 { ∂ ν [ ∂ μ i Δ I × i Δ J ] − ∂ μ i Δ I × ∂ ν i Δ J } , (177) (aa')^{D-2}\partial_{\mu}\partial_{\nu}i\Delta_{I}\times i\Delta_{J}=(aa')^{D-2}\left\{\partial_{\nu}\left[\partial_{\mu}i\Delta_{I}\times i\Delta_{J}\right]-\partial_{\mu}i\Delta_{I}\times\partial_{\nu}i\Delta_{J}\right\}, \tag*{(177)} ( a a ′ ) D − 2 ∂ μ ∂ ν i Δ I × i Δ J = ( a a ′ ) D − 2 { ∂ ν [ ∂ μ i Δ I × i Δ J ] − ∂ μ i Δ I × ∂ ν i Δ J } , (177)
= [ ∂ ν − ( D − 2 ) a H δ ν 0 ] { ( a a ′ ) D − 2 ∂ μ i Δ I × i Δ J } − ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J . (178) =\left[\partial_{\nu}-(D-2)aH\delta^{0}_{\nu}\right]\left\{(aa')^{D-2}\partial_{\mu}i\Delta_{I}\times i\Delta_{J}\right\}-(aa')^{D-2}\partial_{\mu}i\Delta_{I}\times\partial_{\nu}i\Delta_{J}. \tag*{(178)} = [ ∂ ν − ( D − 2 ) a H δ ν 0 ] { ( a a ′ ) D − 2 ∂ μ i Δ I × i Δ J } − ( a a ′ ) D − 2 ∂ μ i Δ I × ∂ ν i Δ J . (178)
Table 18 below gives the case of a primed and unprimed derivative acting on the same propagator times an undifferentiated propagator. Note that the same extraction identities (175-177) suffice for this case.
Table 18. Contributions from products of the form ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × i Δ J (aa')^{D-2}\partial_{\mu}\partial'_{\rho}i\Delta_{I}\times i\Delta_{J} ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × i Δ J .
Tables 19 and 20 give the divergent and finite parts, respectively, of products of the form ( a a ′ ) D − 2 ∂ ρ ∂ ρ ′ i Δ I × ∂ ν i Δ J (aa')^{D-2}\partial_\rho\partial'_\rho i\Delta_I\times\partial_\nu i\Delta_J ( a a ′ ) D − 2 ∂ ρ ∂ ρ ′ i Δ I × ∂ ν i Δ J . Acting the various derivatives produces up to three factors of Δ x \Delta x Δ x . We extract derivatives from them using the relations,
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x 2 D − 2 ⟶ [ 2 ∂ μ ∂ ν ∂ ρ − 3 η ( μ ν ∂ ρ ) ∂ 2 ] 2 9 π 4 ln ( 1 4 H 2 Δ x 2 ) (179) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho}{\Delta x^{2D-2}}\longrightarrow\frac{\left[2\partial_\mu\partial_\nu\partial_\rho-3\eta_{(\mu\nu}\partial_{\rho)}\partial^2\right]}{2^9\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right) \tag*{(179)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D − 2 Δ x μ Δ x ν Δ x ρ ⟶ 2 9 π 4 [ 2 ∂ μ ∂ ν ∂ ρ − 3 η ( μν ∂ ρ ) ∂ 2 ] ln ( 4 1 H 2 Δ x 2 ) (179)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x 2 D ⟶ − 3 K 3 η ( μ ν ∂ ρ ) i δ D ( Δ x ) ( D − 1 ) ( D − 2 ) + η ( μ ν ∂ ρ ) ∂ 4 2 12 π 4 ln 2 ( 1 4 H 2 Δ x 2 ) − [ 2 ∂ μ ∂ ν ∂ ρ + 3 η ( μ ν ∂ ρ ) ∂ 2 ] ∂ 2 2 11 ⋅ 3 ⋅ π 4 ln ( 1 4 H 2 Δ x 2 ) (180) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho}{\Delta x^{2D}}\longrightarrow-\frac{3K_3\eta_{(\mu\nu}\partial_{\rho)}i\delta^D(\Delta x)}{(D-1)(D-2)}+\frac{\eta_{(\mu\nu}\partial_{\rho)}\partial^4}{2^{12}\pi^4}\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-\frac{\left[2\partial_\mu\partial_\nu\partial_\rho+3\eta_{(\mu\nu}\partial_{\rho)}\partial^2\right]\partial^2}{2^{11}\cdot3\cdot\pi^4}\ln\left(\frac{1}{4}H^2\Delta x^2\right) \tag*{(180)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D Δ x μ Δ x ν Δ x ρ ⟶ − ( D − 1 ) ( D − 2 ) 3 K 3 η ( μν ∂ ρ ) i δ D ( Δ x ) + 2 12 π 4 η ( μν ∂ ρ ) ∂ 4 ln 2 ( 4 1 H 2 Δ x 2 ) − 2 11 ⋅ 3 ⋅ π 4 [ 2 ∂ μ ∂ ν ∂ ρ + 3 η ( μν ∂ ρ ) ∂ 2 ] ∂ 2 ln ( 4 1 H 2 Δ x 2 ) (180)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x 2 D + 2 ⟶ − K 3 [ ( D − 2 ) ∂ μ ∂ ν ∂ ρ + 3 η ( μ ν ∂ ρ ) ∂ 2 ] i δ D ( Δ x ) 2 D ( D − 1 ) ( D − 2 ) 2 + [ 2 ∂ μ ∂ ν ∂ ρ + 3 η ( μ ν ∂ ρ ) ∂ 2 ] ∂ 4 2 16 ⋅ 3 ⋅ π 4 [ ln 2 ( 1 4 H 2 Δ x 2 ) − 2 ln ( 1 4 H 2 Δ x 2 ) ] (181) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho}{\Delta x^{2D+2}}\longrightarrow-\frac{K_3\left[(D-2)\partial_\mu\partial_\nu\partial_\rho+3\eta_{(\mu\nu}\partial_{\rho)}\partial^2\right]i\delta^D(\Delta x)}{2D(D-1)(D-2)^2}+\frac{\left[2\partial_\mu\partial_\nu\partial_\rho+3\eta_{(\mu\nu}\partial_{\rho)}\partial^2\right]\partial^4}{2^{16}\cdot3\cdot\pi^4}\left[\ln^2\left(\frac{1}{4}H^2\Delta x^2\right)-2\ln\left(\frac{1}{4}H^2\Delta x^2\right)\right] \tag*{(181)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D + 2 Δ x μ Δ x ν Δ x ρ ⟶ − 2 D ( D − 1 ) ( D − 2 ) 2 K 3 [ ( D − 2 ) ∂ μ ∂ ν ∂ ρ + 3 η ( μν ∂ ρ ) ∂ 2 ] i δ D ( Δ x ) + 2 16 ⋅ 3 ⋅ π 4 [ 2 ∂ μ ∂ ν ∂ ρ + 3 η ( μν ∂ ρ ) ∂ 2 ] ∂ 4 [ ln 2 ( 4 1 H 2 Δ x 2 ) − 2 ln ( 4 1 H 2 Δ x 2 ) ] (181)
Tables 21 and 22 give the divergent and finite parts, respectively, of ( a a ′ ) D − 2 ∂ ρ ∂ ρ ′ i Δ I × ∂ λ ∂ σ ′ i Δ J (aa')^{D-2}\partial_\rho\partial'_\rho i\Delta_I\times\partial_\lambda\partial'_\sigma i\Delta_J ( a a ′ ) D − 2 ∂ ρ ∂ ρ ′ i Δ I × ∂ λ ∂ σ ′ i Δ J . There are now enough derivatives that A 1 A_1 A 1 , B 1 B_1 B 1 and C 1 C_1 C 1 contribute. The four derivatives can produce up to four factors of Δ x \Delta x Δ x , from which we can extract derivatives using the relations,
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x σ Δ x 2 D ⟶ K 3 ∂ μ ∂ ν ∂ ρ ∂ σ i δ D ( Δ x ) ( D − 1 ) ( D − 2 ) − η ( μ ν η ρ σ ) ∂ 4 2 12 π 4 ln 2 ( H 2 Δ x 2 4 ) + [ 3 η ( μ ν η ρ σ ) ∂ 4 + 12 η ( μ ν ∂ ρ ∂ σ ) ∂ 2 − 4 ∂ μ ∂ ν ∂ ρ ∂ σ ] 2 11 ⋅ 3 ⋅ π 4 ln ( H 2 Δ x 2 4 ) (182) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho\Delta x_\sigma}{\Delta x^{2D}}\longrightarrow\frac{K_3\partial_\mu\partial_\nu\partial_\rho\partial_\sigma i\delta^D(\Delta x)}{(D-1)(D-2)}-\frac{\eta_{(\mu\nu}\eta_{\rho\sigma)}\partial^4}{2^{12}\pi^4}\ln^2\left(\frac{H^2\Delta x^2}{4}\right)+\frac{\left[3\eta_{(\mu\nu}\eta_{\rho\sigma)}\partial^4+12\eta_{(\mu\nu}\partial_{\rho}\partial_{\sigma)}\partial^2-4\partial_\mu\partial_\nu\partial_\rho\partial_\sigma\right]}{2^{11}\cdot3\cdot\pi^4}\ln\left(\frac{H^2\Delta x^2}{4}\right) \tag*{(182)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D Δ x μ Δ x ν Δ x ρ Δ x σ ⟶ ( D − 1 ) ( D − 2 ) K 3 ∂ μ ∂ ν ∂ ρ ∂ σ i δ D ( Δ x ) − 2 12 π 4 η ( μν η ρ σ ) ∂ 4 ln 2 ( 4 H 2 Δ x 2 ) + 2 11 ⋅ 3 ⋅ π 4 [ 3 η ( μν η ρ σ ) ∂ 4 + 12 η ( μν ∂ ρ ∂ σ ) ∂ 2 − 4 ∂ μ ∂ ν ∂ ρ ∂ σ ] ln ( 4 H 2 Δ x 2 ) (182)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x σ Δ x 2 D + 2 ⟶ K 3 [ 3 η μ ν η ρ σ ∂ 2 + 6 ( D − 2 ) η ( μ ν ∂ ρ ∂ σ ) ] i δ D ( Δ x ) 2 D ( D − 1 ) ( D − 2 ) 2 + ∂ μ ∂ ν ∂ ρ ∂ σ ∂ 2 2 13 ⋅ 3 ⋅ π 4 ln ( H 2 Δ x 2 4 ) − [ 3 η ( μ ν η ρ σ ) ∂ 2 + 12 η ( μ ν ∂ ρ ∂ σ ) ] ∂ 4 2 16 ⋅ 3 ⋅ π 4 [ ln 2 ( H 2 Δ x 2 4 ) − 2 ln ( H 2 Δ x 2 4 ) ] (183) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho\Delta x_\sigma}{\Delta x^{2D+2}}\longrightarrow\frac{K_3\left[3\eta_{\mu\nu}\eta_{\rho\sigma}\partial^2+6(D-2)\eta_{(\mu\nu}\partial_{\rho}\partial_{\sigma)}\right]i\delta^D(\Delta x)}{2D(D-1)(D-2)^2}+\frac{\partial_\mu\partial_\nu\partial_\rho\partial_\sigma\partial^2}{2^{13}\cdot3\cdot\pi^4}\ln\left(\frac{H^2\Delta x^2}{4}\right)-\frac{\left[3\eta_{(\mu\nu}\eta_{\rho\sigma)}\partial^2+12\eta_{(\mu\nu}\partial_{\rho}\partial_{\sigma)}\right]\partial^4}{2^{16}\cdot3\cdot\pi^4}\left[\ln^2\left(\frac{H^2\Delta x^2}{4}\right)-2\ln\left(\frac{H^2\Delta x^2}{4}\right)\right] \tag*{(183)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D + 2 Δ x μ Δ x ν Δ x ρ Δ x σ ⟶ 2 D ( D − 1 ) ( D − 2 ) 2 K 3 [ 3 η μν η ρ σ ∂ 2 + 6 ( D − 2 ) η ( μν ∂ ρ ∂ σ ) ] i δ D ( Δ x ) + 2 13 ⋅ 3 ⋅ π 4 ∂ μ ∂ ν ∂ ρ ∂ σ ∂ 2 ln ( 4 H 2 Δ x 2 ) − 2 16 ⋅ 3 ⋅ π 4 [ 3 η ( μν η ρ σ ) ∂ 2 + 12 η ( μν ∂ ρ ∂ σ ) ] ∂ 4 [ ln 2 ( 4 H 2 Δ x 2 ) − 2 ln ( 4 H 2 Δ x 2 ) ] (183)
Γ 2 ( D 2 − 1 ) 16 π D Δ x μ Δ x ν Δ x ρ Δ x σ Δ x 2 D + 4 ⟶ K 3 [ D ( D − 2 ) ∂ μ ∂ ν ∂ ρ ∂ σ + 6 D η ( μ ν ∂ ρ ∂ σ ) ∂ 2 + 3 η ( μ ν η ρ σ ) ∂ 4 ] i δ D ( Δ x ) 4 ( D + 1 ) D 2 ( D − 1 ) ( D − 2 ) 2 − [ 8 ∂ μ ∂ ν ∂ ρ ∂ σ + 24 η ( μ ν ∂ ρ ∂ σ ) ∂ 2 + 3 η ( μ ν η ρ σ ) ∂ 4 ] ∂ 4 2 19 ⋅ 3 ⋅ 5 ⋅ π 4 [ ln 2 ( H 2 Δ x 2 4 ) − 2 ln ( H 2 Δ x 2 4 ) ] (184) \frac{\Gamma^2\left(\frac{D}{2}-1\right)}{16\pi^D}\frac{\Delta x_\mu\Delta x_\nu\Delta x_\rho\Delta x_\sigma}{\Delta x^{2D+4}}\longrightarrow\frac{K_3\left[D(D-2)\partial_\mu\partial_\nu\partial_\rho\partial_\sigma+6D\eta_{(\mu\nu}\partial_{\rho}\partial_{\sigma)}\partial^2+3\eta_{(\mu\nu}\eta_{\rho\sigma)}\partial^4\right]i\delta^D(\Delta x)}{4(D+1)D^2(D-1)(D-2)^2}-\frac{\left[8\partial_\mu\partial_\nu\partial_\rho\partial_\sigma+24\eta_{(\mu\nu}\partial_{\rho}\partial_{\sigma)}\partial^2+3\eta_{(\mu\nu}\eta_{\rho\sigma)}\partial^4\right]\partial^4}{2^{19}\cdot3\cdot5\cdot\pi^4}\left[\ln^2\left(\frac{H^2\Delta x^2}{4}\right)-2\ln\left(\frac{H^2\Delta x^2}{4}\right)\right] \tag*{(184)} 16 π D Γ 2 ( 2 D − 1 ) Δ x 2 D + 4 Δ x μ Δ x ν Δ x ρ Δ x σ ⟶ 4 ( D + 1 ) D 2 ( D − 1 ) ( D − 2 ) 2 K 3 [ D ( D − 2 ) ∂ μ ∂ ν ∂ ρ ∂ σ + 6 D η ( μν ∂ ρ ∂ σ ) ∂ 2 + 3 η ( μν η ρ σ ) ∂ 4 ] i δ D ( Δ x ) − 2 19 ⋅ 3 ⋅ 5 ⋅ π 4 [ 8 ∂ μ ∂ ν ∂ ρ ∂ σ + 24 η ( μν ∂ ρ ∂ σ ) ∂ 2 + 3 η ( μν η ρ σ ) ∂ 4 ] ∂ 4 [ ln 2 ( 4 H 2 Δ x 2 ) − 2 ln ( 4 H 2 Δ x 2 ) ] (184)
Table 21. Local contributions from products of the form ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν ′ ∂ σ ′ i Δ J (aa')^{D-2}\partial_{\mu}\partial_{\rho'}i\Delta_{I}\times\partial_{\nu'}\partial_{\sigma'}i\Delta_{J} ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν ′ ∂ σ ′ i Δ J .
Table 22. Nonlocal contributions from products of the form ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν ∂ σ ′ i Δ J (aa')^{D-2}\partial_\mu\partial'_\rho i\Delta_I\times\partial_\nu\partial'_\sigma i\Delta_J ( a a ′ ) D − 2 ∂ μ ∂ ρ ′ i Δ I × ∂ ν ∂ σ ′ i Δ J .