Acetonitrile-water mixtures are a unique model system for implementing a new family of coherent two-dimensional (2D) vibrational spectroscopy-doubly vibrationally enhanced (DOVE) four wave mixing (FWM) spectroscopy for structure analysis of hydrogen-bonded systems. These mixtures often have a heterogeneous microstructure that consists of acetonitrile (CH3CN) or water clusters. There are two types of CH3CN environments in mixtures-free CH3CN species and hydrogen-bonded CH3CN species. To design new DOVE-FWM experiments and interpret DOVE-FWM spectra, it is crucial to have a complete structure information regarding this model system. In this work, we have used a highly sensitive spectroscopic method, 2D infrared (IR) correlation spectroscopy for probing structure information of the mixtures (a 50:50 mol % CH3CN-H2O solution, a 50:50 mol % CH3CN-D2O solution and a 38:62 mol % CH3CN-D2O solution) under temperature perturbation. In addition to two cross-peaks at 2253 and 2259 cm(-1), corresponding to the Cequivalent toN stretch of free acetonitrile molecules and hydrogen-bonded acetonitrile molecules, respectively, the 2D correlation spectra show a new feature at 2256 cm(-1) which has not been observed by conventional IR and Raman spectra in these mixtures. The second derivatives of the IR spectra also show this new feature whose intensity increases with temperature, whereas the features at 2253 and 2259 cm(-1) become weaker. Correspondingly, a new feature appears at about 3168 cm(-1) for the combination band of the Cequivalent toN stretch and the C-C stretch, whereas the combination band of the free CH3CN species is centered at 3164 cm(-1), and 3171 cm(-1) for the hydrogen-bonded CH3CN species. The new feature may be related to an intermediate microstructure between clusters of acetonitrile molecules and hydrogen-bonded acetonitrile molecules associated with clusters of water molecules. The results will provide useful structural and spectral information for the design of new 2D DOVE-FWM experiments and the molecular structure modeling of the mixtures.