Method for detecting a thinning of the semiconductor substrate of an integrated circuit from its back face and corresponding integrated circuit

Abstract

The thinning of a semiconductor substrate of an integrated circuit from a back face is detected using the measurement of a physical quantity representative of the resistance between the ends of two electrically-conducting contacts situated at an interface between an insulating region and an underlying substrate region. The two electrically-conducting contacts extend through the insulating region to reach the underlying substrate region.

Claims

1. A method, comprising: forming a doped region in a semiconductor substrate at a location underneath an insulation region which is also located in the semiconductor substrate; forming transistor components in and on the semiconductor substrate; depositing an etch-stop layer covering protruding parts of the transistor components; depositing a pre-metallization dielectric layer over the etch-stop layer; etching through the pre-metallization dielectric layer and etch-stop layer to form: a first opening and a second opening, wherein the first and second openings further extend through the insulation region to reach the doped region; and third openings that reach terminals of the transistor components; and filling the first, second and third openings with a metal material.

2. The method of claim 1, further comprising connecting the metal material filled first and second openings to first and second inputs of a circuit configured to measure a physical quantity representative of a resistance of said doped region between first and second ends of the metal material filed first and second openings situated at the doped region.

3. The method of claim 1, wherein etching comprises using simultaneous and identical etch operations to form the first, second and third openings.

4. The method of claim 1, wherein the semiconductor substrate comprises silicon, and wherein the etch operations comprise a final etch step, selective with respect to the silicon and to a metal silicide, that is configured to etch the etch-stop layer, said final etch step being a timed etch process further etching material of the insulating region, wherein an etch time of the timed etch process is determined as a function of a height of the insulating region.

5. The method of claim 1, wherein the semiconductor substrate is doped with a first conductivity type and the underlying substrate region is doped with a second conductivity type opposite the first conductivity type.

6. The method of claim 1, wherein forming transistor components comprises forming a metal silicide for said terminals of the transistor components.

7. The method of claim 6, wherein etching comprises performing an etch step which is selective with respect to the semiconductor substrate and the metal silicide, wherein said etch step etches through the etch-stop layer and the insulating region.

8. The method of claim 7, wherein said etch step is a timed etch for etching through the insulating region, wherein an etch time of the timed etch process is determined as a function of a height of the insulating region.

9. The method of claim 7, wherein said etch step is performed following an etch operation made to pass through the pre-metallization dielectric layer.

10. The method of claim 9, wherein said etch operation comprises a first etch configured to pass only partially through the pre-metallization dielectric layer followed by a second etch to finish passing through the pre-metallization dielectric layer.

11. The method of claim 10, wherein the second etch polymerizes sides of the opening extending through the pre-metallization dielectric layer.

12. The method of claim 10, wherein the first etch is a plasma etch which uses CH.sub.2F.sub.2 as process gas and wherein the second etch is a plasma etch which uses C.sub.4F.sub.6 as process gas.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Other advantages and features of the invention will become apparent upon examining the detailed description of non-limiting embodiments and their implementation, and from the appended drawings in which:

(2) FIGS. 1 and 2 are schematic illustrations of various embodiments and implementations.

DETAILED DESCRIPTION

(3) In FIG. 1, the reference CI denotes an integrated circuit comprising a semiconductor substrate SB, for example with a P type conductivity, comprising at least one insulating region RIS, for example of the shallow trench isolation (STI) type which, in the example illustrated here, is situated on top of a well CS with an N type of conductivity.

(4) The top face (or front face) FS of the substrate is covered with an etch-stop layer 1 (CESL layer) generally of silicon nitride SiN. This layer 1 is covered with a dielectric layer 2, commonly denoted by those skilled in the art under the acronym PMD (pre-metallization dielectric), which separates the etch-stop layer 1 from the first metallization level M1 of the interconnection part of the integrated circuit, commonly denoted by those skilled in the art under the acronym BEOL (for Back End Of Line).

(5) In order to be able to detect a potential thinning of the substrate SB from its back face FA, opposite to its top or front face FS, the integrated circuit IC comprises a detector DT here comprising two electrically-conducting contacts C1, C2 extending through the dielectric layer 2, the etch-stop layer 1 and the insulating region RIS.

(6) The two contacts C1 and C2 respectively have two first ends EX11 and EX21 situated at the interface between the insulating region RIS and the underlying substrate region, here the well CS.

(7) The two contacts C1 and C2 also respectively comprise two second ends EX12 and EX22, opposite to the first ends, situated at the interface between the dielectric layer 2 and the first metallization level M1.

(8) These two second ends EX12 and EX22 are in contact with two metal tracks PST1 and PST2 of the metallization level M1 which are connected to an electrical circuit 3.

(9) Although not indispensable, this electrical circuit 3 is preferably incorporated within the integrated circuit IC.

(10) The electrical circuit 3 comprises here, by way of non-limiting example, a comparator 31 whose non-inverting input is connected to a voltage divider bridge 30 and whose inverting input is connected to the metal track PST2 and hence to the contact C2.

(11) The other metal track PST1, and hence the other contact C1, is connected to a supply voltage, here ground GND.

(12) The comparator 31 compares the voltage present on the metal track PST2 with the reference voltage supplied by the voltage divider 30 and delivers a signal S whose value is representative of the fact that the voltage present on the metal track PST2 is lower or otherwise than the reference voltage.

(13) Moreover, the voltage PST2 is a quantity representative of the current flowing in the resistive channel formed by the two contacts C1 and C2 and the underlying substrate region CS, and notably of the resistance of this underlying substrate region.

(14) If the substrate is not thinned, the resistance between the two first ends EX11 and EX21 is low, for example of the order of 10 , for a distance L between the two contacts, for example, of the order of 0.8 microns and a width W equal to 0.8 m (for a 90 nanometer technology).

(15) In contrast, if an attacker thins the substrate SB so as to come very near to, or even reach, the insulating region RIS, then the resistance between the two first ends EX11 and EX21 increases significantly (to reach for example a value of 20 ), which then causes an increase in the voltage on the track PST2 and the switching of the comparator 31, the signal S then being representative of a thinning of the substrate.

(16) It will be evident that, in this case, processing means, for example a logic circuit, not shown here, can inhibit the operation of the integrated circuit.

(17) Reference is now more particularly made to FIG. 2 in order to describe one embodiment of the contacts C1 and C2.

(18) FIG. 2 shows schematically other components of the integrated circuit such as, by way of non-limiting example, two transistors T1 and T2.

(19) The transistor T1 is a transistor with a dual gate region P1 and P2, such as those used in non-volatile memories for example of the FLASH or EEPROM type.

(20) The first gate region P1 is isolated from the substrate by a first gate oxide OX1 and the two gate regions P1 and P2 are mutually isolated by a second gate oxide OX2.

(21) The transistor T2 is a conventional transistor whose gate region P1 is separated from the substrate by a gate oxide OX3.

(22) The source, drain and gate regions of these transistors conventionally comprise, close to their surface, regions of metal silicide (silicided regions) ZS1, ZS2, ZS3, ZS4 and ZS5.

(23) Certain of these silicided regions are intended to be contacted by additional electrically-conducting contacts, for example the silicided regions ZS3, ZS4 and ZS5.

(24) FIG. 2 shows the orifices ORD1, ORD2 and ORD3 that will be filled by one or more electrically-conductive materials, for example tungsten, so as to form the aforementioned three additional electrically-conducting contacts, together with the two orifices OR1 and OR2 intended to be filled by the same electrically-conducting metal, so as to form the two electrically-conducting contacts C1 and C2.

(25) These various orifices result from etch steps here comprising four plasma etch operations, GV1, GV2, GV3 and GV4 having conventional characteristics notably in terms of process gases employed.

(26) Conventionally, the dielectric layer 2 is covered with an antireflective layer, generally known by those skilled in the art under the acronym BARC. This antireflective layer is under a layer of photoresist which undergoes a photolithographic step and an exposure to light in such a manner as to define the locations of the various orifices ORD1-ORD3 and OR1-OR2.

(27) Subsequently, after development of the resist, the first etch GV1, which is conventionally a plasma etch, is carried out so as to remove the portion of antireflective layer situated within the holes of the resist.

(28) By way of non-limiting example, in a 90 nanometer technology, CF.sub.4 may be used as a process gas at a pressure of around 80 millitorr.

(29) Subsequently, a second etch GV2 is carried out which will allow a first part of the dielectric layer 2 to be etched.

(30) This second etch GV2 is a fairly aggressive plasma etch which uses for example CH.sub.2F.sub.2 as process gas at a pressure of 100 millitorr.

(31) However, this aggressive etch produces a barrel effect in the orifices; in other words, the more it is etched, the bigger the diameter of the orifice will get.

(32) It is for this reason that this second etch GV2 is interrupted after a chosen time to be replaced by a third etch GV3 which will not only etch the remainder of the dielectric layer 2 but will also polymerize the sides of the orifice in such a manner as to eventually obtain a virtually cylindrical orifice.

(33) By way of non-limiting example, C.sub.4F.sub.6 could be chosen for such a third plasma etch GV3 at a pressure of around 45 millitorr.

(34) When these etch operations are finished, the various orifices open out onto the etch-stop layer 1.

(35) A fourth plasma etch GV4 is then carried out so as to etch the layer 1 in order to open out onto the silicided regions ZS4, ZS5 and ZS3.

(36) By way of non-limiting example, this time CHF.sub.3 can be used as process gas at a pressure of around 120 millitorr.

(37) This fourth etch GV4 is a timed etch which also allows, as illustrated in FIG. 2, the insulating material, for example silicon silicide, of the insulating region RIS to be etched.

(38) The etch time depends on the height h of the insulating region and those skilled in the art will know how to adjust the etch time as a function of the characteristics of the etch in such a manner that the orifices OR1 and OR2 reach the underlying substrate region CS.

(39) Moreover, this lengthening of the etch time has virtually no impact on the silicided regions ZS3, ZS4 and ZS5 because this etch chemistry is selective with respect to the metal silicide and to the silicon.

(40) For this reason, the formation of the two contacts C1 and C2 will only have required a local modification of the contacts mask and an increase in the time of the etch GV4 with respect to a conventional etch GV4.