Revisiting the S-matrix approach to the open superstring low energy effective lagrangian
Luiz Antonio Barreiro, Ricardo Medina
TL;DR
The paper revisits the S-matrix program for the open superstring, showing that spacetime supersymmetry imposes a constraint that eliminates the need for high-point amplitudes to determine the bosonic part of the low-energy effective action. By exploiting the absence of $(\zeta\cdot k)^N$ terms in N-point amplitudes, the authors reduce the number of independent coefficients and determine the OSLEEL terms up to ${\alpha'}^4$ using mainly the 4-point amplitude, with results consistent with BPS-based methods. They explicitly derive the OSLEEL up to ${\alpha'}^3$ and, remarkably, obtain the full ${\alpha'}^4$ bosonic sector, including nontrivial tensor structures, in a non-perturbative way with respect to higher-point amplitudes. The approach hints at a deep link between open string amplitudes and the NS-NS sector of type II theories via KLT, suggesting a path to fully determine the closed-string low-energy action from tree-level data.
Abstract
The conventional S-matrix approach to the (tree level) open string low energy effective lagrangian assumes that, in order to obtain all its bosonic ${α'}^N$ order terms, it is necessary to know the open string (tree level) $(N+2)$-point amplitude of massless bosons, at least expanded at that order in $α'$. In this work we clarify that the previous claim is indeed valid for the bosonic open string, but for the supersymmetric one the situation is much more better than that: there are constraints in the kinematical bosonic terms of the amplitude (probably due to Spacetime Supersymmetry) such that a much lower open superstring $n$-point amplitude is needed to find all the ${α'}^N$ order terms. In this `revisited' S-matrix approach we have checked that, at least up to ${α'}^4$ order, using these kinematical constraints and only the known open superstring 4-point amplitude, it is possible to determine all the bosonic terms of the low energy effective lagrangian. The sort of results that we obtain seem to agree completely with the ones achieved by the method of BPS configurations, proposed about ten years ago. By means of the KLT relations, our results can be mapped to the NS-NS sector of the low energy effective lagrangian of the type II string theories implying that there one can also find kinematical constraints in the $N$-point amplitudes and that important informations can be inferred, at least up to ${α'}^4$ order, by only using the (tree level) 4-point amplitude.
