Constraining the Polarization of Gravitational Waves with the Parkes Pulsar Timing Array Second Data Release

We search for the isotropic stochastic gravitational-wave background, including the nontensorial polarizations that are allowed in general metric theories of gravity, in the Parkes Pulsar Timing Array (PPTA) second data release (DR2). We find no statistically significant evidence that the common-spectrum process reported by the PPTA collaboration has the tensor transverse, scalar transverse, vector longitudinal, or scalar longitudinal correlations in PPTA DR2. Therefore, we place a $95\%$ upper limit on the amplitude of each polarization mode, as $\mathcal{A}_{\mathrm{TT}} \lesssim 3.2\times 10^{-15}$, $\mathcal{A}_{\mathrm{ST}} \lesssim 1.8\times 10^{-15}$, $\mathcal{A}_{\mathrm{VL}}\lesssim 3.5\times 10^{-16}$ and $\mathcal{A}_{\mathrm{SL}}\lesssim 4.2\times 10^{-17}$; or, equivalently, a $95\%$ upper limit on the energy density parameter per logarithm frequency, as $\Omega_{\mathrm{GW}}^{\mathrm{TT}} \lesssim 1.4\times 10^{-8}$, $\Omega_{\mathrm{GW}}^{\mathrm{ST}} \lesssim 4.5\times 10^{-9}$, $\Omega_{\mathrm{GW}}^{\mathrm{VL}} \lesssim 1.7\times 10^{-10}$ and $\Omega_{\mathrm{GW}}^{\mathrm{SL}} \lesssim 2.4\times 10^{-12}$, at a frequency of 1/yr.