1: \begin{abstract}

2: The gradient discretisation method is a generic framework that is applicable to a number of schemes

3: for diffusion equations, and provides in particular generic error estimates in $L^2$ and $H^1$-like

4: norms. In this paper, we establish an improved $L^2$ error estimate for gradient schemes.

5: This estimate is applied to a family of gradient schemes, namely, the Hybrid Mimetic

6: Mixed (HMM) schemes, and provides an $\mathcal O(h^2)$ super-convergence rate

7: in $L^2$ norm. Finally, the link between HMM and Two-Point Flux Approximation (TPFA)

8: finite volume schemes is exploited to partially answer a long-standing conjecture

9: on the super-convergence of TPFA schemes.

10: \end{abstract}