<p><span>On 2013 June 21st an eruption occurred in the active region NOAA 1177 (14S73E), </span><span>giving rise to</span> <span>a M2.9 class flare starting at 02:30 UT, a fast partial halo coronal mass ejection (CME), and a type II radio burst. The concomitant emission of solar energetic particles (SEPs) produced a significant increase in the proton fluxes measured by LET and HET aboard STEREO-B. By using stereoscopic observations in extreme ultra violet (EUV) and white light (WL) spectral intervals, we performed a 3D reconstruction of the expanding front by processing SDO/AIA, STEREO/EUVI, COR1 and COR2, and SOHO/LASCO data assuming a spheroidal model. By using the 3D reconstruction, we estimated the temporal evolution of &#952;</span><span><sub>Bn,</sub></span><span>&#160;</span><span>i.e.,</span> <span>the angle between the normal to the expanding front and the coronal magnetic field computed by the Potential-Field Source-Surface (PFSS) approximation, within 2.5 R</span><span><sub>&#664;</sub></span><span>.&#160;The front </span><span>of the CME</span><span>was found to be quasi-parallel to the magnetic field almost everywhere</span><span><sub>.</sub></span><span>&#160;Above 2.5 R</span><span><sub>&#664;</sub></span><span>, where the front was identified as a shock, we projected the 3D expanding surface </span><span>reconstructed for </span><span>different times on the ecliptic plane and</span><span>&#160;</span><span>we calculated the &#952;</span><span><sub>Bn&#160;</sub></span><span>between the normal to the front and Parker spiral arms. In this case the shock was almost perpendicular to the magnetic field (quasi-parallel shock).&#160;During the expansion the region located between the nose and the eastern flank of the shock was magnetically connected with ST-B in agreement with the significant SEP flux measured on-board this spacecraft.</span> <span>W</span><span>hile</span> <span>the shock was only marginally connected with ST-A and GOES-15. </span><span>T</span><span>he SEP release time was estimated to be 10 minutes after the Type II onset, when the shock front was already above 2.5 R</span><span><sub>&#664;</sub></span><span>&#160;with a quasi-parallel configuration. Our results are discussed in the framework of the shock acceleration scenario, even if quasi-parallel shocks are expected to have a reduced acceleration efficiency.</span></p>