We propose the application of coronagraphic techniques to the spectroscopic direct detection of exoplanets via the Doppler shift of planetary molecular lines. Even for an unresolved close-in planetary system, we show that the combination of a visible
nuller and an extreme adaptive optics system can reduce the photon noise of a main star and increase the total signal-to-noise ratio (S/N) of the molecular absorption of the exoplanetary atmosphere: it works as a spectroscopic coronagraph. Assuming a 30 m telescope, we demonstrate the benefit of these high-contrast instruments for nearby close-in planets that mimic 55 Cnc b ($0.6 lambda/D$ of the angular separation in the K band). We find that the tip-tilt error is the most crucial factor; however, low-order speckles also contribute to the noise. Assuming relatively conservative estimates for future wavefront control techniques, the spectroscopic coronagraph can increase the contrast to $ sim 50-130 $ times and enable us to obtain $sim 3-6 $ times larger S/N for warm Jupiters and Neptunes at 10 pc those without it. If the tip-tilt error can be reduced to $lesssim 0.3$ mas (rms), it gains $sim 10-30$ times larger S/N and enables us to detect warm super-Earths with an extremely large telescope. This paper demonstrates the concept of spectroscopic coronagraphy for future spectroscopic direct detection. Further studies of the selection of coronagraphs and tip-tilt sensors will extend the range of application of the spectroscopic direct detection beyond the photon collecting area limit.
