We perform a detailed semi-analytical analysis of the electroweak phase transition (EWPT) property in NMSSM, which serves as a good benchmark model in which the 126 GeV Higgs mixes with a singlet. In this case, a strongly first order electroweak phase transition (SFOEWPT) is achieved by the tree-level effects and the phase transition strength γ c is determined by the vacua energy gap at T = 0. We make an anatomy of the energy gap at both tree-level and loop-level and extract out a dimensionless phase transition parameter R κ ≡ 4κv s /A κ , which can replace A κ in the parameterization and affect the light CP odd and even Higgs spectra. We find that SFOEWPT only occurs in R κ ∼ −1 and positive R κ O(10), which in the non-PQ limit case would prefer either a relatively light CP odd or CP even Higgs boson ∼ (60, 100) GeV, therefore serves as a smoking gun signal and requires new search strategies at the LHC.

Given the fact that the relatively light Higgsino mass $\mu$ favored in natural supersymmetry usually results in a sizable scattering cross section between the neutralino dark matter and the nucleon, we study the impact of the recently updated direct detection bounds from LUX experiment, including both Spin Independent (SI) and Spin Dependent (SD) measurements, on the parameter space of natural Next-to-Minimal Supersymmetric Standard Model (nNMSSM). Different from the common impression that the SI bound is stronger than the SD one, we find that the SD bound is complementary to the SI bound and in some cases much more powerful than the latter in limiting the nNMSSM scenarios. After considering the LUX results, nNMSSM is severely limited, e.g. for the peculiar scenarios of the NMSSM where the next-to-lightest CP-even Higgs corresponds to the $125 {\rm GeV}$ Higgs boson discovered at the LHC, the samples obtained in our random scan are excluded by more than $85\%$. By contrast, the monojet search at the LHC Run-I can not exclude any sample of nNMSSM. We also investigate the current status of nNMSSM and conclude that, although the parameter points with low fine tuning are still attainable, they are distributed in some isolated parameter islands which are difficult to get. Future dark matter direct search experiments such as XENON-1T will provide a better test of nNMSSM.Comment: 20 pages, 10 figures, meet JHEP published versio

Natural Next-to-Minimal Supersymmetric Standard Model (nNMSSM) is featured by predicting one CP-even Higgs boson satisfying m h 1 120 GeV and Higgsinos lighter than about 300 GeV, and consequently the cross section for DM-nucleon scattering in this scenario is usually quite large. We study the diphoton signal of the light Higgs boson in nNMSSM by considering the tight constraints from the latest LUX and PandaX-II experiments, and we conclude that the optimal value of the signal rate at 8 TeV LHC is greatly reduced in comparison with earlier predictions. For example, previous studies indicated that the rate may exceed 120 fb for m h 1 80 GeV, while it is at most 25 fb if the lightest neutralino in the scenario is fully responsible for the measured DM relic density. We also investigate the case of m h 1 98 GeV which is hinted by the excesses of the LEP analysis on Zbb signal and the CMS analysis on the diphoton signal. We conclude that nNMSSM can explain simultaneously the excesses at 1σ level without violating any known constraints.

In the Next-to-Minimal Supersymmetric Standard Model (NMSSM), all singlet-dominated particles including one neutralino, one CP-odd Higgs boson and one CP-even Higgs boson can be simultaneously lighter than about 100 GeV. Consequently, dark matter (DM) in the NMSSM can annihilate into multiple final states to explain the galactic center gamma-ray excess (GCE). In this work we take into account the foreground and background uncertainties for the GCE and investigate these explanations. We carry out a sophisticated scan over the NMSSM parameter space by considering various experimental constraints such as the Higgs data, B-physics observables, DM relic density, LUX experiment and the dSphs constraints. Then for each surviving parameter point we perform a fit to the GCE spectrum by using the correlation matrix that incorporates both the statistical and systematic uncertainties of the measured excess. After examining the properties of the obtained GCE solutions, we conclude that the GCE can be well explained by the pure annihilationsχ 0