Process for manufacturing electronic devices comprising high voltage MOS transistors, and electronic device thus obtained

[0002] The present invention pertains to a process for manufacturing electronic devices comprising high voltage MOS transistors, and an electronic device thus obtained.

[0003] Many electronic devices currently present on the market use both NMOS and PMOS high voltage transistors (hereinafter referred to as HV transistors) of the dualgate type (i.e., having the gate region doped with doping ionic species of the same type used for the source and drain regions) and of the drain-extension type. In these manufacturing processes, in order to achieve lengths of the channels of the HV transistors of less than 0.5 μm it is necessary to appropriately increase the doping of the substrate so as to prevent undesired effects, such as punchthrough, i.e., the undesired electrical connection between two regions having different potentials. For this reason, the region designed to house these transistors is enriched with doping ions that may confer on the region the same conductivity type as the substrate, forming a well (or tub) that has a doping level greater than the substrate.

[0004] In many devices it is, however, also indispensable to have HV transistors with a low multiplication coefficient (i.e., with a low ratio between the current flowing in the drain region and the current flowing in the substrate) and with a low body effect (i.e., a reduced increase in the threshold voltage of the transistor when the substrate biasing voltage increases), which features require a low doping of the substrate.

[0005] A commonly adopted solution to this problem is that of suitably shaping the masks so as to form so-called “NO-TUB” HV transistors formed directly in the substrate, instead of in the well or tub.

[0006] Since these transistors have a low multiplication coefficient but also a low threshold voltage, they can be used only with appropriate circuit designs to prevent unacceptable parasitic currents. On the other hand, such designs lead to an undesired complication of the circuitry associated to the HV transistors.

[0007] The disclosed embodiment of the present invention provides a new process for manufacturing electronic devices comprising HV transistors with a low multiplication coefficient, a low body effect, and a high threshold.

[0008] According to the embodiment of the present invention, a process for manufacturing electronic devices comprising HV transistors includes providing a substrate of semiconductor material with a first area having a first conductivity type and a first doping level; forming a first gate region of semiconductor material of the first transistor on the first area of the substrate; and forming in the first area of the substrate, first source and drain regions of a second conductivity type, at the sides of the gate regions, and doping the first gate region with doping species determining the first conductivity type.

[0009] In accordance with another aspect of the present invention, the foregoing process includes the step of forming a second high voltage MOS transistor in a second area of the substrate and the step of forming the second high voltage transistor includes: Forming, in the second area of the substrate, a first tub having the first conductivity type and a doping level higher than the first area of the substrate; forming a second gate region having the second conductivity type over the first tub; and forming in the first tub source and drain regions of the second conductivity laterally to the second gate region.

[0010] In accordance with yet another aspect of the present invention, the forming of the first source and drain regions and forming of the second source and drain regions is performed simultaneously.

[0011] For a better understanding of the present invention, a preferred embodiment thereof is now described, purely as a non-limiting example, with reference to the attached drawings in which:

[0012] FIG. 1 shows a cross-section through a wafer in an initial step of the manufacturing process according to the present invention;

[0013] FIG. 2 presents a top view of the wafer of FIG. 1;

[0014] FIGS. 3-9 show cross-sections similar to that of FIG. 1 during successive manufacturing steps;

[0015] FIG. 10 presents a top view of the wafer of FIG. 9;

[0016] FIGS. 11-13 show cross-sections similar to that of FIG. 9 during successive manufacturing steps;

[0017] FIG. 14 presents a top view of the wafer of FIG. 13;

[0018] FIG. 15 shows a cross-section similar to that of FIG. 11 in a subsequent manufacturing step;

[0019] FIG. 16 presents a top view of the wafer of FIG. 15;

[0020] FIG. 17 shows a cross-section similar to that of FIG. 15 in a subsequent manufacturing step;

[0021] FIG. 18 presents a top view of the wafer of FIG. 17; and

[0022] FIGS. 19-20 show cross-sections similar to that of FIG. 17 in successive manufacturing steps.

[0023] The following description refers to an embodiment of HV NMOS transistors of a traditional type (and hence with a high multiplication coefficient), HV NMOS transistors with a low multiplication coefficient and a high threshold, and EEPROM memory cells including a selection transistor and a memory transistor.

[0024] In FIG. 1, a wafer 1 comprising a substrate 2 of P-type monocrystalline silicon has undergone the steps of definition of the active areas. In detail, in a per se known manner, thick oxide regions (field oxide regions 5) have already been formed in the substrate 2, these regions delimiting between them active areas of the substrate that are to house various components of the device to be formed. In particular, FIG. 1 shows three such active areas: a first active area 6 for housing first HV transistors having a low multiplication coefficient and a high threshold (hereinafter also referred to as NO-TUB HV transistors), a second active area 7 for housing second HV transistors having a high multiplication coefficient, and an active matrix area 8 for housing EEPROM memory cells. Further active areas (not shown in the drawings) are generally provided for forming LV NMOS and PMOS transistors or HV PMOS transistors.

[0025] In detail and in a per se known manner, the active matrix area 8 defines a grid, of which FIG. 2 shows in full only the part regarding a cell, indicated by 9, having basically the shape of a T rotated through 90°.

[0026] A sacrificial oxide layer 10 extends on the surface 3 of the substrate 2.

[0027] Subsequently, as shown in FIG. 3, an HV P-well resist mask 11 is formed which covers the entire surface of the wafer 1, except for the second active area 7 and the active matrix area 8, and P-type doping species are implanted, as schematically represented by arrows 12. The implant gives rise to an HV P-type tub 13 in the second active area 7 and to a P-matrix tub 14 in the active matrix area 8. The HV P-type tub 13 and P-matrix tub 14 thus have the same conductivity type as the substrate 2, but a higher doping concentration. In contrast, the first active area 6 is not implanted to enable formation of the NO-TUB HV transistors.

[0028] After removing the HV P-well mask 11, a capacitor mask 20 is formed which covers the entire surface of the wafer 1, except for strips perpendicular to the plane of the drawing. Next, N-type doping species (e.g., phosphorus) are implanted, as represented schematically in FIG. 4 by arrows 21. N-type continuity regions 22 are then formed inside the P-matrix tub 14 providing electrical continuity between each selection transistor and the respective memory transistor of each cell.

[0029] After removing the capacitor mask 20, the wafer 1 is annealed, the sacrificial layer 10 is removed, and matrix oxidation is carried out, so as to form a matrix oxide layer 25 over the surfaces of all the active areas 6, 7, 8, as shown in FIG. 5. Next, using a matrix oxide mask 24 which covers only the wafer 1 above the P-matrix tub 14, where the gate regions of the memory transistors are to be formed, the matrix-oxide layer 25 is removed where it is uncovered to form a matrix oxide region 25din the P-matrix region 14 partially covering the continuity region 22 (FIG. 6).

[0030] After removing the matrix oxide region 24, the wafer 1 is again oxidized to form a tunnel oxide layer 26 over the entire surface 3 of the wafer 1, where the latter is exposed, and the thickness of the matrix oxide region 25dincreases. In this way, the structure of FIG. 7 is obtained.

[0031] A first polycrystalline silicon layer (poly1 layer 27) is then deposited and appropriately doped. Subsequently, an interpoly dielectric layer 31 is formed, for example comprising a triple ONO (silicon oxide - silicon nitride - silicon oxide) layer, as shown in FIG. 8.

[0032] A floating gate mask 30 is then formed, as shown in FIGS. 9 and 10. Next, the dielectric layer 31, the poly1 layer 27, and the tunnel oxide layer 26 are etched to form an interpoly dielectric region 31d, a floating gate region 27d, and a tunnel region 26d, which are aligned with respect to each other.

[0033] After removing the floating gate mask 30, an HV oxidation step is carried out, thus forming an HV gate oxide layer 34 over the entire free surface of the wafer 1 (FIG. 11). Portions of oxide 34 are formed also at the sides of the floating gate region 27dof the memory transistor, as illustrated in FIG. 11.

[0034] A second polycrystalline silicon layer (poly2 layer 43, not doped) is then deposited, as shown in FIG. 12. A dual-gate doping mask 44 is then formed and covers the active area 6, as shown in FIGS. 13 and 14, and then N-type doping ionic species are implanted, as represented schematically by arrows 47 in FIG. 13. In this way, the portions of the poly2 layer 43 over the HV P tubs 13 and P-matrix tubs 14 undergo N-type doping.

[0035] Next, using a selective doping mask 48, which covers the HV P tubs 13 and P-matrix tubs 14, P-type doping ionic species, for example boron, are implanted , as represented schematically by arrows 49 in FIG. 15. In this way, the portion of the poly2 layer 43 which extends over the active areas 6 and is to form gate regions of the NO-TUB HV transistors undergoes P-type doping so as to present the same conductivity type as the substrate 2. Thereby, the structure shown in cross-section in FIG. 15 and in top view in FIG. 16 is obtained.

[0036] Next, using an HV gate mask 56, the poly2 layer 43 is selectively removed to form a gate region 43a, which is P-type doped, of the NO-TUB HV transistor, a gate region 43b, which is N-type doped, of the TUB HV transistor, a gate region 43cof the selection transistor of the memory cell, and a control gate region 43dof the memory transistor.

[0037] Next, a re-oxidation step is carried out, thus forming an oxide layer 57 over the entire free surface of the wafer 1, in particular at the sides of the gate regions 43aand 43bof the NO-TUB HV transistor and of the TUB HV transistor, as illustrated in FIG. 19.

[0038] Subsequently, N-type doping ionic species are implanted, as represented schematically by arrows 63 in FIG. 19. In the active area 6 and in the HV P tub 13 at the two sides of the HV gate region 43aof the NO-TUB HV transistor and of the HV gate region 43bof the TUB HV transistor, N-type source and drain regions 64a, 64bare formed.

[0039] Thus, the source and drain regions 64aof the NO-TUB HV transistor are doped in an opposite way with respect to their HV gate region 43a, whereas the source and drain regions 64bof the TUB HV transistor are doped in the same way as their HV gate region 43b. At the same time, in the P-matrix tub 14 the following regions are formed: a drain region 64cof the selection transistor, on one side and self-aligned with respect to the gate region 43cof the selection transistor; and a source region 64dof the memory transistor, on the side of the latter not facing the selection transistor, aligned with the gate region 43dof the memory transistor. In addition, an implant is carried out also on the area arranged between the selection transistor and the memory transistor; however, this implant is generally performed within the continuity regions 22, which are more doped, and hence is not visible (for this reason, the corresponding area has been represented by dashed lines in FIG. 19). In the event of misalignments, however, this implant guarantees electrical continuity.

[0040] After a protective dielectric layer 72 has been formed, the final structure of FIG. 20 is obtained, showing: a NO-TUB HV NMOS transistor 80 with P-type doped gate region 43a, a TUB HV NMOS transistor 81 with N-type doped gate region 43b, and hence of a dual-gate type, and an EEPROM cell 82 including a selection transistor 83 and a memory transistor 84. Final steps are the carried out, including forming the contacts and the electrical interconnection lines, depositing a passivation layer, etc.

[0041] The method described thus enables TUB HV transistors and NO-TUB HV transistors to be formed at the same time, as well as memory cells having very different characteristics, optimizing the number of steps required.

[0042] In particular, the fact that the gate regions 43aof the NO-TUB HV transistors have the same conductivity type as the substrate 2 ensures a high work function (i.e., a high potential difference automatically present between materials because of the physical characteristics of the materials), and hence a high threshold voltage. Since the source and drain regions 64aare formed directly in the substrate 2, where a more doped tub is not present and the substrate 2 is not very doped, the NO-TUB HV transistors have a low multiplication coefficient and low body effect.

[0043] In this way, using a dual-gate process, it is also possible to obtain high threshold HV transistors, so avoiding the complex circuit solutions necessary for their use in prior art processes.

[0044] Finally, it is clear that numerous modifications and variations may be made to the method and device described and illustrated herein, all of which falling within the scope of the invention, as defined in the attached claims and the equivalents thereof. For example, it is possible to form NO-TUB HV PMOS transistors, which will consequently have an N-type doped gate region, like the substrate, and P-type doped source and drain regions. Furthermore, the order of the steps for selectively doping the gate regions 43a, 43band the control gate region 43dcan be reversed, and these regions can be doped after they have been shaped, instead of before as has been described herein, possibly adjusting the implant dose of the source and drain regions 64aof the NO-TUB HV transistors.