Gauge invariant effective potential

We show that the long-standing problem of gauge dependence of the effective potential arises due to the factorization of the determinant of operators, which is invalid when we take the zeta-regularized trace of the operators. We show by correcting for this assumption by computing the multiplicative anomaly, the gauge-dependent terms of the effective potential cancel. We also show that in two- and odd-dimensional noncompact spacetime manifolds where the multiplicative anomaly term is zero, the standard calculation of one-loop effective potential gives a gauge-independent result. These results are in support of our claim that the multiplicative anomaly may play a crucial role in removing the gauge dependence in the effective potential in the four-dimensional noncompact manifold. Noting the nontrivial aspects of this anomaly computation for a generic scenario, we propose the heat-kernel method to compute the effective potential where this anomaly emerges as a total derivative, thus redundant. We explicitly show how one can calculate the gauge independent, effective action and the Coleman-Weinberg effective potential by employing the heat-kernel method. Based on this result, we advocate the heat-kernel expansion as the most straightforward method, as it naturally deals with the matrix elliptic operator for the calculation of manifestly gauge independent, effective actions compared to other conventional methods.