UC Berkeley

IceCube has observed a diffuse astrophysical neutrino flux from $\sim$\SI{100}{TeV} to a few PeV, with an energy spectrum consistent with a single power law (SPL). In the PeV region, the spectral shape has not been precisely measured: starting track and cascade searches have a relatively small effective volume, upgoing tracks are mostly absorbed by the Earth, while the Extremely High Energy (EHE) analyses target cosmogenic neutrinos above \SI{10}{PeV}. This analysis fills the gap between \SI{1}{PeV} and \SI{10}{PeV} by selecting high-energy down-going, through-going tracks. To reject the large atmospheric muon background, two largely complementary techniques are combined. The first technique selects events with high stochasticity to reject atmospheric muon bundles whose stochastic energy losses are smoothed due to high muon multiplicity. The second technique vetoes atmospheric muons with the IceTop surface array. Using 9 years of data, we found two events in the signal region, each with relatively high signal probabilities. A joint maximum likelihood estimation is performed using this sample with the previous 9.5-year upgoing track sample to measure the neutrino spectrum, and a likelihood ratio test is performed for the single power-law (SPL) vs. SPL+cutoff hypothesis. The best-fit cutoff energy is \SI[parse-numbers = false]{1.83^{+6.92}_{-0.81}}{PeV}. The p-value of the likelihood ratio test is 0.62, and the SPL+cutoff model is not significant over SPL. High-energy astrophysical objects from four source catalogs are also searched for around the direction of the two events. A BL Lac object is found inside the 90$\%$ directional uncertainty region of one event, while the second event does not have any associated objects.