Quantization of first-class constraints with structure functions (Q756814)

The paper contains the results which constitute the part of program of translating the BRST construction in terms of standard symplectic geometry. The following problem is solved. Let Q be a configuration space, \({\mathcal F}\)- regular foliation on Q describing gauge equivalent configurations. One assumes \(\bar Q=Q/{\mathcal F}\) to be a Hausdorff manifold and \(\Pi: Q\to Q/{\mathcal F}\) to be a submersion. The phase space of the real system is the cotangent bundle \(T^*\bar Q\) with standard symplectic form. This system can be obtained as a reduction from the symplectic manifold \(T^*Q\). In fact, for any vector field X on Q we define a function \(J_ X: T^*Q\to {\mathbb{R}}\) by \(T^*_ qQ\ni \beta \to J_ X(\beta)=\beta (X_ q)\in {\mathbb{R}}\) and the set of constraint functions on \(T^*Q:\) \(C=\{J^ q_ X|\) \(X\in {\mathcal F}\}\). One easily shows that this constrained system is equivalent to the one described by \(T^*\bar Q\). It has been shown [\textit{C. Duval}, \textit{J. Elhadad}, \textit{M. J. Gotay}, \textit{J. Sniatycki}, and \textit{G. M. Tuynman}, Ann. Phys. 206, No.1, 1-26 (1991)] that, in general, the standard quantization prescription (consisting in imposing the supplementary conditions \(\hat f\psi\) \(=0\), where \(\hat f\) is the quantum counterpart of a classical constraint f and \(\psi\) is the state vector) does not work. It is shown in the paper how to enlarge the original constrained system and the set of constraints in such a way that the reduced system remains \(T^*\bar Q\) and the standard quantization procedure is correct. The extended configuration space \(\tilde Q\) is defined as the dual of the foliation, \(\tilde Q={\mathcal F}^*\), the extended phase space becomes \(T^*\tilde Q\). The construction of the set of new constraints is slightly more complicated. As a corollary the problem of quantizing the system with globally independent first-class constraints with structure functions is solved.