Low & High scale MSSM inflation, gravitational waves and constraints from Planck

Worth a look, but read Section VII with your guard up. The mechanism is not new — Eq. (15) and the coefficients in Eqs. (18)-(21) are lifted from Mazumdar-Nadathur-Stephens (Ref. [14]), and the low-scale potential is from Refs. [10, 11]. What is new is the confrontation with the Planck 2013 data: a specific parameter point (Eq. 31), the (n_s, r) and (n_s, α_s, κ_s) planes in Figs. 4-5, and a TT spectrum in Fig. 6. That is a reasonable thing to have done, and the numbers hold up where you can check them: Eqs. (26)-(28) are internally consistent with the standard normalization, and the slow-roll machinery in the appendix is Easther-Peiris done correctly. The reheating argument is the part I actually like — the inflaton carries SM charges, so you get the right thermal bath with no dark radiation, and the T_rh bound in Eq. (26) is a real, falsifiable statement. That is the paper's best physics and it gets less space than it deserves. The soft spot is the central claim in abstract:s2. Fig. 6 is not a fit. The TT spectrum over 2 < ℓ < 2500 is set by Ω_b h², Ω_c h², H₀, τ and A_s; the paper never states which background it fed to CAMB, and the model enters only through (A_s, n_s, α_s, κ_s), all of which Eq. (30) places inside Planck's 2σ by construction. With α_s ~ -0.01 the running does essentially nothing at ℓ ~ 2500, so the §VII sentence about running 'enhancing' the high-ℓ fit is asserted, not computed. The genuine model content is Figs. 4-5, and there r is scanned from 10⁻²⁹ to 0.12 by varying H — a consistency check, not a fit. The r ≤ 0.12 in the abstract is Planck's input, not the model's output. Separately, the a_H = 2.108 in the Fig. 2 and Fig. 3 captions versus a_H = 12.650 in Eq. (31) is a real inconsistency: with c_H = 2 the inflection condition needs a_H ≈ 12.65, so the captions correspond to c_H ≈ 1/3. That is a reproducibility defect, not the load-bearing problem. And δ ~ 10⁻⁴ is severe fine-tuning that goes undiscussed. Who should read it: people already working on MSSM or inflection-point inflation, and anyone thinking about reheating without dark radiation. It deserves a referee — the program is established and the Planck comparison is a fair question — but expect the referee to demand an actual likelihood for Fig. 6 and a fix to the a_H captions.

The central claim (abstract:s2) is that the model 'matches the observed amplitude of the TT-power spectrum ... with low and high multipoles.' Section VII and Fig. 6 are the only support for this. But the CMB TT spectrum over 2 < ℓ < 2500 is overwhelmingly determined by the ΛCDM background — Ω_b h², Ω_c h², H_0, τ, and A_s — not by the detailed shape of the inflationary potential. The model enters only through (A_s, n_s, α_s, κ_s), and Eq. (30) places all four inside Planck's 2σ contours by construction (0.958 < n_s < 0.963, |α_s| ≲ 0.01, |κ_s| ≲ 0.006). The paper never states which background parameters were fed to CAMB for Fig. 6, nor does it vary them. If they are Planck's best-fit values, then Fig. 6 is essentially a standard ΛCDM TT curve and the 'match' is guaranteed for any potential yielding n_s ≈ 0.96 with negligible running — which is precisely what Eq. (30) reports. The genuine model test is the (n_s, r) plane of Fig. 4, but there r is scanned freely from 10^-29 to 0.12 by varying H, so that too is a consistency check rather than a fit. The abstract's framing ('match the observed amplitude ... with low and high multipoles') therefore overstates the evidence: what is shown is that the model's (n_s, α_s, κ_s) lie inside Planck's marginalized contours, not that the model's TT spectrum was confronted with the data. The §VII statement that running 'will enhance the power spectrum to a permissible value such that it will accurately fit Planck high-l data' is asserted, not computed; with α_s ~ -0.01 the effect at ℓ ~ 2500 is far below Planck's error bars. The a_H = 2.108 (Fig. 2, Fig. 3 captions) versus a_H = 12.650 (Eq. 31) discrepancy is a real internal inconsistency — with c_H = 2 the inflection condition a_H² = 40 c_H²(1-4δ²) requires a_H ≈ 12.65, so 2.108 corresponds to c_H ≈ 1/3 — but it is a reproducibility defect, not the load-bearing issue.

The reader's weakest_assumption correctly identifies the TT comparison as the soft spot, but frames it as 'no chi-square, no residuals, no likelihood.' I would sharpen this: the problem is not merely that a fit was not performed, but that a chi-square on Fig. 6 would not test the model either, because the TT spectrum is dominated by ΛCDM background parameters that the paper neither states nor varies. The model's only handles on the TT curve are (A_s, n_s, α_s, κ_s), and Eq. (30) already places these inside Planck's 2σ contours. So the Fig. 6 overlay is close to guaranteed to succeed for any potential with n_s ≈ 0.96 and small running, and cannot discriminate the inflection-point shape from a plain power law. This is a correctness-of-framing concern, not an internal inconsistency: the underlying slow-roll machinery (appendix, Eqs. 32-41, following Easther-Peiris [43]) is standard and the arithmetic in Eqs. (26)-(28) checks out. The reader's point (3) — that r ≤ 0.12 is Planck's input, not a model output — is correct and reinforces the same theme: the model accommodates any r from 10^-29 to 0.12 by varying H, so the 'bound' in abstract:s2 is not a prediction. The reader's point (2), the a_H = 2.108 vs 12.650 discrepancy, is a genuine internal inconsistency (2.108 is consistent with c_H ≈ 1/3, not c_H = 2), but it is a reproducibility defect rather than the load-bearing concern; it does not change the qualitative conclusion that the model can be tuned to sit inside Planck's contours. The reader's point (4), the severity of δ ~ 10^-4 fine-tuning, is a legitimate cost that the paper does not discuss, but it is a known feature of inflection-point models (Refs [10, 11]) rather than a new flaw. On balance I agree with CONDITIONAL and would not move the verdict: the paper is a legitimate consistency check of an established model against Planck 2013, but the abstract's 'match the TT-power spectrum' language claims more than Fig. 6 delivers. The concrete test above would settle whether the TT comparison has any discriminating power; if it does not, the abstract and §VII should be reworded to say the model's (n_s, α_s, κ_s) are consistent with Planck's marginalized constraints, which is what the paper actually establishes. I give credit for the CAMB-based numerical P_S(k) in Fig. 2 and the explicit slow-roll radiative-correction comparison, which are real computations; the concern is confined to how the TT overlay is interpreted, not to the underlying perturbation calculation.

If the central claim holds, the paper is a useful demonstration that MSSM inflection-point inflation survives the first Planck data release without new ingredients, and that the high-scale branch makes a falsifiable prediction in the r–n_S plane accessible to B-mode searches. The paper deserves credit for several concrete things: the reheat-temperature estimate of Eq. (26) is tied to an explicit MSSM degree-of-freedom count and to a published computation of flat-direction decay (Ref. [50]); the appendix collects the second-order slow-roll expressions in a form others can reuse; the low-scale and high-scale branches are cleanly separated by a stated criterion (m_φ vs H); and the gravitino overproduction problem at large r is acknowledged openly in §V rather than buried. The parameter scan of Eq. (29) and the r–n_S contours of Fig. 4 are the kind of falsifiable output that makes the model worth confronting with data.

The paper asks whether inflation can be driven by a flat direction of the Minimal Supersymmetric Standard Model — a gauge-invariant combination of squarks or sleptons — and still reproduce the Planck 2013 temperature anisotropy spectrum without adding new ingredients. The inflaton potential is lifted by a sixth-order superpotential term and by Hubble-induced mass and A-term corrections, and near an inflection point it becomes flat enough to support the 50 to 70 e-folds of slow roll that the observed window requires. Because the shape is controlled by essentially one tuning parameter, the authors can scan it and read off the scalar amplitude, the spectral tilt, the running, the running of the running, and the tensor-to-scalar ratio. If the argument holds, the field that drives inflation also carries Standard Model charge and decays into ordinary particles, so the universe reheats with the right thermal degrees of freedom, produces no dark radiation, and leaves the lightest supersymmetric particle as a thermal relic.

What carries the argument. The load-bearing object is the inflection point itself: the field value $φ_0$ where $V''(φ_0) = 0$, which flattens the potential into a short plateau and lets a steep-looking sixth-order potential support slow roll. How flat that plateau is depends on the tuning parameter $δ$, defined by $a_H^2/(40c_H^2) = 1 - 4δ^2$; the smaller $δ$, the flatter the plateau and the closer the predicted tilt comes to one. Everything else — the amplitude, $n_S$, $α_S$, $κ_S$, $r$ — follows from expanding the potential about $φ_0$ into the coefficients $α, β, γ, κ$ and feeding them through the slow-roll machinery with higher-order radiative corrections, then through CAMB for the angular power spectrum.

Core claim. In the high-scale regime, where the Hubble-induced terms dominate over the soft mass, the potential $V(φ) = V_0 + c_H H^2|φ|^2/2 - a_H H φ^6/(6M_{PL}^3) + |φ|^{10}/M_{PL}^6$ has an inflection point when $a_H^2/(40c_H^2) = 1 - 4δ^2$ with $|δ|$ small, located at $φ_0 = (√c_H H M_{PL}^3/10)^{1/4}$. For the benchmark point $δ ∼ 10^{-4}$, $c_H = 2$, $a_H = 12.650$, $φ_0 = 1.129×10^{16}$ GeV, the authors report $2.092 < 10^9 P_S < 2.297$, $0.958 < n_S < 0.963$, $r < 0.12$, $-0.0098 < α_S < 0.0003$ and $-0.0007 < κ_S < 0.006$, all inside the 2σ region of Planck plus WMAP-9 plus BAO. The same potential, run with CAMB and with higher-order radiative corrections to the slow-roll parameters, is plotted against the TT spectrum from $ℓ = 2$ to $ℓ = 2500$ and described as fitting both the low-multipole Sachs-Wolfe plateau and the high-multipole acoustic peaks. The low-scale branch, $m_φ ≫ H$, gives $H_* ∼ 0.1$ GeV, $φ_0 ∼ 3×10^{14}$ GeV and a tensor-to-scalar ratio around $10^{-29}$, undetectable but still compatible with the data.

The premise it rests on. The plateau exists only if $a_H^2/(40c_H^2)$ sits within roughly $4×10^{-4}$ of one, and the paper assumes the Kähler structure naturally supplies that tuning; it further assumes that overlaying the model curve on the Planck points in Fig. 6 counts as a fit, and its own benchmark parameters disagree between the captions of Figs. 2 and 3 ($a_H = 2.108$) and Eq. (31) ($a_H = 12.650$), where only the latter satisfies the stated inflection condition.

A single tuning of order $δ ∼ 10^{-4}$ in the ratio $a_H^2/(40c_H^2)$ is enough to place the model inside the Planck 2σ region for the tilt, the running, and the running of the running simultaneously.

The high-scale branch saturates at $r ≤ 0.12$, corresponding to $H_* ≤ 9.24×10^{13}$ GeV and $V_* ≤ (1.96×10^{16}\,GeV)^4$, so a B-mode detection at that level would pin the inflationary scale rather than merely bound it.

Reheating into MSSM degrees of freedom gives $T_{rh} ≤ 6.654×10^{15}\,(r_*/0.12)^{1/4}$ GeV with $g_* = 228.75$, so the model predicts no extra relativistic species and lets big bang nucleosynthesis proceed with Standard Model content alone.

At that reheating temperature the gravitino abundance stays compatible with big bang nucleosynthesis only if the gravitino is heavier than roughly 10 TeV and decays before nucleosynthesis; if the gravitino is the lightest supersymmetric particle, the high-scale branch is excluded.

The low-scale branch with $m_φ ∼ 1$ TeV gives $H_* ∼ 0.1$ GeV and $T_{rh} ∼ 3×10^8$ GeV, so the same framework covers inflationary scales from $10^{-1}$ GeV to $10^{13}$ GeV with $r$ ranging from $10^{-29}$ to $0.12$.

Because the model can slide $r$ across thirty orders of magnitude by dialing $H_*$, a future tensor detection would not test the scenario so much as locate it on the branch; the sharper discriminator is the running, which the model predicts to be tiny and slightly negative.

The $δ ∼ 10^{-4}$ tuning is the same for any flat-direction inflection model, so a statistical argument about how often a landscape or Kähler structure supplies that tuning would turn this from a fit into a prediction.

If reheating into Standard Model charges really happens within one Hubble time, the model's distinctive signature is a null result — $ΔN_{eff} = 0$ at the sensitivity of the next generation of CMB experiments — so a detection of dark radiation would remove its main selling point over string-motivated alternatives.

The quoted gravitino bound links the tensor-to-scalar ratio to supersymmetry searches: a confirmed $r$ near 0.12 combined with a light gravitino would close the high-scale branch, which is a cross-check that collider and cosmological data could jointly perform.

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review