A semiconductor device includes one or more trench gates extending in a first direction in plan view, one or more first-conductivity-type regions spaced away from each other in the first direction, where the first-conductivity-type regions are shallower than the trench gates, one or more second-conductivity-type regions alternating with the first-conductivity-type regions in the first direction, where the second-conductivity-type regions are shallower than the trench gates and deeper than the first-conductivity-type regions, and a second-conductivity-type trench spacer region spaced away from the one or more trench gates, where the trench spacer region has a higher concentration than the second-conductivity-type regions. Here, the trench spacer region is positioned within the first-conductivity-type regions in plan view and closer to a back surface of the semiconductor device than the first-conductivity-type regions are.

A method of manufacturing an insulated gate bipolar transistor includes providing trenches extending from a first surface to a layer section in a semiconductor portion, introducing impurities into mesa sections between the trenches, and forming, from the introduced impurities, second portions of doped regions separated from source regions by body regions. The source regions are electrically connected to an emitter electrode. The second portions have a second mean net impurity concentration exceeding at least ten times a first mean net impurity concentration in first portions of the doped layer. The first portions extend from the body regions to the layer section, respectively.

A first laser pulse emitted from a semiconductor laser oscillator and having a first pulse width is entered onto a second surface of a semiconductor substrate in which a semiconductor device is formed on a first surface and dopants are added to a surface layer portion on the second surface side. A second laser pulse having a second pulse width less than or equal to 1/10 of the first pulse width is entered on an incident area of the first laser pulse in an overlapping manner. The relative positional relationship on a time axis between falling time of the first laser pulse and rising time of the first laser pulse is set such that the temperature of the first surface, which rises due to the incidence of the first laser pulse and the second laser pulse, does not exceed an allowable upper limit value which is predetermined.

In a method of further enhancing the performance of a narrow active cell IE type trench gate IGBT having the width of active cells narrower than that of inactive cells, it is effective to shrink the cells so that the IE effects are enhanced. However, when the cells are shrunk simply, the switching speed is reduced due to increased gate capacitance. A cell formation area of the IE type trench gate IGBT is basically composed of first linear unit cell areas having linear active cell areas, second linear unit cell areas having linear hole collector areas and linear inactive cell areas disposed therebetween.

A semiconductor device includes a first semiconductor region of a first conductivity type, a second semiconductor region having a second conductivity type, a first insulating layer on the first and second semiconductor regions, and field plate electrodes are provided in the first insulating layer at different distances from the first semiconductor layer. A first field plate electrode is at a first distance, a second field plate electrode is at a second distance greater than the first distance, and a third field plate electrode is at a distance greater than the second distance. The first through third field plate electrodes are electrically connected to each other and the third electrode is electrically connected to the second semiconductor region.

On a first wafer surface of a semiconductor wafer, a projection-depression shape is formed. On the first wafer surface, a resin member is so formed to have a resin outer peripheral end positioned away from a wafer outer peripheral end and expose the wafer outer peripheral end. By partially removing the semiconductor wafer, on a second wafer surface of the semiconductor wafer, formed is a recessed shape having a recessed-portion outer peripheral end positioned 0.5 mm or more inside from the resin outer peripheral end. After performing a processing on the second wafer surface, the resin member is removed.

An n.sup. drift layer is a parallel pn layer having an n-type region and a p-type region are alternately arranged in the direction parallel to the main surface so as to come into contact with each other, and have a width in a direction parallel to the main surface of the substrate which is less than a length in a direction perpendicular to the main surface of the substrate. A second-main-surface-side lower end portion of the p-type region has a structure in which a high-concentration lower end portion and a low-concentration lower end portion of a p-type low-concentration region are repeated at a predetermined pitch in the direction parallel to the main surface of the substrate. It is possible to provide a super junction MOS semiconductor device which can improve a trade-off relationship between turn-off loss and turn-off dv/dt and improve avalanche resistance.

An improvement is achieved in the performance of a semiconductor device. The semiconductor device includes a first trench gate electrode and second and third trench gate electrodes located on both sides of the first trench gate electrode interposed therebetween. In each of a semiconductor layer located between the first and second trench gate electrodes and the semiconductor layer located between the first and third trench gate electrodes, a plurality of p.sup.+-type semiconductor regions are formed. The p.sup.+-type semiconductor regions are arranged along the extending direction of the first trench gate electrode in plan view to be spaced apart from each other.

A method of manufacturing a semiconductor device includes: forming a lattice defect layer in a substrate having a front surface region where a bipolar element of a pn junction type is formed and a rear surface region opposing the front surface region, the lattice defect layer being formed by injecting a charged particle to a first region in the rear surface region of the substrate; forming a laminated region, in which a first conductivity type impurity region and a second conductivity type impurity region are sequentially laminated from a rear surface side of the substrate toward the first region, in a second region in the rear surface region of the substrate, the first region being positioned deeper than the second region from a rear surface of the substrate; and selectively activating the laminated region by laser annealing after the formation of the laminated region and the lattice defect layer.

To improve accuracy and shielding capabilities of impurity implantation, a method of manufacturing a semiconductor device is provided, the method including forming a first photoresist on a front surface of a semiconductor substrate, the front surface being provided with a front surface structure, forming, on the first photoresist or below a rear surface of the semiconductor substrate, a second photoresist having opposite photo-curing properties from those of the first photoresist, and implanting impurities into the semiconductor substrate using as a mask the second photoresist, which has been subjected to patterning.