Patent Application
20190244869
The present invention relates to the production of MOS (metal oxide semi-conductive)-type transistors having a constrained channel. It will have an advantageous, but not limiting application in the production of such transistors on an FDSOI-type substrate.
Reducing the size and increasing the speed of microelectronic circuits require controlling, with a great precision, the control current of the transistors. The control current in particular depends on the length of the grid, on the capacitance of the grid and on the mobility of the charge carriers in the channel. Thus, various solutions have been proposed to reduce the length of the grid, to increase the capacitance of the grid and to increase the mobility of the charge carriers in the channel of the transistor.
Among the solutions explored to increase the mobility of the charge carriers, one consists of introducing a mechanical stress in the channel. More specifically, the performances of NMOS-type transistors are improved by applying a voltage stress in the channel, this voltage stress making it possible to increase the mobility of the electrons. The performances of PMOS-type transistors are improved by applying a compression stress in the channel, this stress making it possible to increase the mobility of the holes.
As illustrated in
Another manner of proceeding consists of depositing a solid plate voltage stress, then of removing this layer, only on PMOS-type transistors by carrying out conventional photolithography steps. Then, a compression stressed layer is solid plate deposited. This compression stressed layer therefore covers PMOS and NMOS transistors. The final step consists of removing this compression stressed layer only on NMOS-type transistors by carrying out conventional photolithography steps.
This production method has proved to be, in practice, relatively long and complex.
There is therefore a need, consisting of proposing a solution to facilitate the obtaining, on a plate, of a voltage stressed layer on a first type of pattern and of a compression stressed layer on another type of pattern.
Other aims, characteristics and advantages of the present invention will appear upon examining the following description and the supporting drawings. It is understood that other advantages can be incorporated.
To achieve this aim, according to an embodiment, a method for producing, preferably a first transistor is provided, surmounted at least partially on a voltage stressed layer and a second pattern, preferably a second transistor, surmounted at least partially on a compression stressed layer.
The productions of voltage and compression stressed layers comprise the following steps:
Thus, this method makes it possible to form a voltage stressed layer on a first pattern is a compression stressed layer on a second pattern. When these patterns are transistors, the stressed layers make it possible to voltage and compression stress the channel of each of the transistors improving, due to this, the mobility of the charge carriers.
This method has a considerably reduced number of steps, with respect to the conventional solutions mentioned above and illustrated in
According to an advantageous embodiment, the first and the second patterns are transistors which have different characteristics. For example, they can have different dopings. Thus, according to an example, the first pattern and the second pattern are transistors, preferably respectively of NMOS and PMOS-type.
Alternatively or in combination, the first transistor and the second transistor can have different grid oxide thicknesses.
Alternatively or in combination, the first transistor and the second transistor can have different channel thicknesses.
In the scope of the development of the present invention, it has been considered to deposit a voltage stressed layer covering both NMOS-type and PMOS-type transistors, then to protect NMOS-type transistors, then to apply an additional treatment step to invert the stress in the layer covering the PMOS transistors, this additional treatment step mainly consisting of an ultraviolet, laser or electron beam (e-Beam) treatment.
The application of an ultraviolet treatment with a power of between 300 W/m2 and 2000 W/m2, combined with a heat treatment carrying the temperature of the substrate to a temperature of a few tens even hundreds of degrees, seems to be able to make it possible to pass from a voltage stress to compression stress. However, this treatment type requires a long duration, typically from a few minutes to a few tens of minutes, which is not compatible with industrial productivity constraints.
Moreover, methods which would make an implantation intervene in the heavy ion stressed layer would have several disadvantages. In particular, this implantation would risk altering the layer underlying the stressed layer when the latter is relatively thin. The performances of the transistor would be immediately degraded. Moreover, the implantation of heavy ions would consequently alter the geometry and the dimensions of the protective blocks covering NMOS-type transistors. Typically, when these protective blocks are made of resin, the implantation of heavy ions would lead to a certain creeping of these protective blocks. This would degrade the quality of the interface between the protected stressed layer and the unprotected layer. A weak controlling of the interface between the voltage stressed layer and the compression stressed layer would result from this.
It has therefore been proved to be, that these types of solutions have numerous disadvantages.
The claimed method reduces, even removes these disadvantages.
The aims, objectives, as well as the characteristics and advantages of the invention will emerge best from the detailed description of embodiments of the latter which are illustrated by the following supporting drawings, wherein:
The drawings are given as examples, and are not limiting of the invention. They constitute schematic representations of principle intended to facilitate the understanding of the invention and are not necessarily to the scale of practical applications. In particular, the relative thicknesses of the different layers and films are not representative of reality.
Before starting a detailed review of embodiments of the invention, optional characteristics are stated below, which can possibly be used in association or alternatively:
According to an example, the voltage stressed layer and the compression stressed layer are preferably nitride-based layers and preferably made of nitride. Preferably, these layers are silicon nitride SixNy layers, x and y being whole numbers. Preferably, this is SiN, SiN3, Si2N4.
According to an example, the first zone fully covers the first pattern forming a transistor and the second zone fully covers the second pattern forming a transistor.
All the following characteristics can be generalised in the case where a transistor is a pattern.
According to an example, an electrical insulation trench separates the first transistor and the second transistor. The first and the second zones each extend until the insulation trench. Preferably, an interface defined by the first and the second zones is at least partially facing to the insulation trench. This means that the flank of the protective layer (which defines the delimitation between voltage stressed layer and compression stressed layer) is situated on the insulation trench.
According to an example, the deposition of the protective layer is done so as to cover one part at least of the first NMOS-type transistor, and preferably all of the first NMOS-type transistor, and to fully leave the second PMOS-type transistor to be uncovered.
According to an example, the first transistor and the stressed layer are shaped so as to voltage stress at least one channel of the first transistor.
According to an example, the second transistor and the stressed layer are shaped so as to compression stress at least one channel of the second transistor.
According to an example, the plasma is formed from a mixture comprising H2.
According to an example, the parameters of the plasma, in particular the type and the energy of the ions, are adjusted such that the implantation of ions reduces the thickness of the stressed layer in the second zone. This makes it possible, for example, to best control the thickness of the stressed layer of the second zone.
According to an example, the plasma is formed from a mixture comprising at least one of the following gases: HBr, NH3.
An advantage of this chemistry is that during the modification of the uncovered stressed layer, the thickness of the latter can be reduced by the bombardment of species heavier than hydrogen (for example, Br and N species) present in the plasma.
This makes it possible to adjust the thickness of the stressed layer.
This better controlling of the thickness of the stressed layer provides, for example, as an advantage, that the thickness of the stressed layers surmounting all the sources and drains, whichever the transistors, has the same thickness on the surface of the plate. During the etching of this layer to open the contacts to make electrical recoveries, the sources and drains will therefore be reached at the same time.
Moreover, this makes it possible to avoid possible short-circuits between the active zones of adjacent transistors.
According to an example, the plasma is formed from a mixture comprising at least one gas of CxHy type, x and y being whole numbers and preferably another gas, for example, helium (He) and/or argon (Ar).
An advantage of this chemistry is that during the modification of the uncovered stressed layer, the thickness of the latter can be reduced by the bombardment of species heavier than hydrogen (for example, the species He and Ar) present in the plasma.
This makes it possible to adjust the thickness of the stressed layer, with for example, the advantages mentioned above.
Moreover, this chemistry makes it possible to reduce the consumption of the protective layer, in particular when the latter is carbonaceous. Such is the case when the protective layer is a photosensitive or thermosensitive resin.
In this case, carbon atoms, coming from the dissociation of CxHy molecules, are deposited on the resin which protects it from the bombardment of other ions (argon or helium, for example) during the plasma implantation. The resin thus conserves, during all the implantation, the dimensions thereof and the geometry thereof.
This embodiment thus makes it possible to have a particularly high dimensional control at the interface between the first and the second transistors.
According to an example, the formation of the stressed layer is done such that the stressed layer has a thickness of between 2 and 50 nm (10−9 metres), preferably between 5 and 40 nm.
Advantageously, the modification step of the stressed type of the second zone of the stressed layer is carried out at a temperature less than 150° C. More specifically, the implantation is done at a temperature T≤150° C., preferably T≤130° C. and preferably T≤100° C. Preferably, T is between 10° C. and 100° C. Particularly advantageously, the method is not based on the increase in temperature of the stressed layer to relax it.
According to an example, the formation of the stressed layer comprises a deposition, preferably conform.
According to an example, the conform deposition is done by one of the following depositions: plasma-enhanced chemical vapour deposition (PECVD), low-pressure enhanced chemical vapour deposition (LPCVD), rapid thermal chemical vapour deposition (RTCVD), physical vapour deposition (PVD).
According to an example, the parameters of the implantation, in particular the energy of the hydrogen-based ions, are selected so as to modify, by implanting all the thickness of the second zone of the stressed layer.
According to an example, the parameters of the implantation, in particular the energy of the hydrogen-based ions, are selected so as to modify, by implanting the uncovered stressed layer continuously, from the surface of the stressed layer and until a depth of around 40 nm (10−9 metres).
According to an example, the implantation is done at a temperature ≤150° C., preferably 130° C.
According to an example, the method comprises, after the modification step, a step of removing the protective layer.
According to an example, the method comprises a plurality of sequences, each sequence being applied on said first and second transistors, each sequence comprising successively at least: said step of depositing a stressed layer, said step of depositing a protective layer and said modification step. Thus, the stressed layer is formed by successive depositions. These depositions are stacked on top of one another. Each sequence is applied on one same zone of the plate.
This embodiment makes it possible to also increase the precision of the modification of the stressed layer, which ultimately makes it possible to improve the precision of the stress imposed in the channel.
According to an example, at each sequence, said step of depositing a stressed layer is configured such that the thickness of the deposited stressed layer is less than 30 nm, preferably less than 20 nm and preferably of between 5 nm and 15 nm.
According to an example, the first and second transistors each have a channel formed in a semi-conductive layer, the semi-conductive layer surmounting a dielectric layer and a support substrate. The semi-conductive layer, the dielectric layer and the support substrate form a semi-conductive on insulator-type substrate.
According to an example, the first and second transistors are formed on an FDSOI-type substrate.
According to an example, the first and second transistors are formed on a PDSOI-type substrate.
According to an example, the hydrogen (H)-based ions are preferably taken from among the following ions: H, H+, H2+, H3+.
It is specified that, in the scope of the present invention, the terms “on”, “surmount”, “cover”, “underlying”, “opposite” and the equivalents thereof do not necessarily mean “in contact with”. Thus, for example, the deposition, the extension, the binding, the assembly, or the application of a first layer on a second layer, does not compulsorily mean that the two layers are directly in contact with one of the other, but means that the first layer covers at least partially the second by being either directly in contact with it, or by being separated by it, by at least one other layer or at least one other element.
A layer can moreover be composed of several sublayers of one same material or of different materials.
A substrate, a film, a material A-“based” layer, mean a substrate, a film, a layer comprising this material A only or this material A and possibly other materials, for example doping elements.
The word “dielectric” qualifies a material of which the electrical conductivity is sufficiently low in the given application to be used as an insulator.
In the scope of the present invention, an organic or organo-mineral material which could be shaped by an exposure to an electron, photon or X-ray beam, or mechanically, is qualified as a resin.
As an example, resins conventionally used in microelectronics, polystyrene (PS)-based resins, methacrylate (for example, Polymethyl methacrylate PMMA), Hydrogen silsesquioxane (HSQ), polyhydroxy styrene (PHS), etc. can be cited. The interest of using a resin is that it is easy to deposit a significant thickness of it, from several hundred nanometres to several microns.
It is specified that in the scope of the present invention, the thickness of a layer or of the substrate is measured along a direction perpendicular to the surface according to which this layer or this substrate has the maximum extension thereof. In the figures, the thickness is taken along the vertical.
When it is indicated that an element is facing to another element, this means that these two elements are both situated on one same line perpendicular to the main plane of the substrate, or on one same line oriented vertically in the figures.
In the present patent application, a doping referenced P comprises all dopings by positive charge carriers (holes), whatever the content of the doping. Thus, a doping P comprises doping P+ contents and doping P contents less than the P+ type doping. Likewise, a doping referenced N comprises all dopings by negative charge carriers, whatever the content of the doping. Thus, a doping N comprises doping N+ contents, and doping N contents less than the N+ type doping.
Conventionally, a doping referenced P+ means that it is a P-type doping (doping by positive charges) and of which the doping species content is greater than or equal to 1 atom of the doping species for less than 500 atoms of the semi-conductor and preferably for less than 10 to 100 atoms of the material forming the semi-conductive layer. Likewise, an N+ doping means that it is an N-type doping (doping by negative charges) and of which the doping species content is greater than or equal to 1 atom of the doping species for less than 500 atoms of the semi-conductor and preferably for less than 10 to 100 atoms of the material forming the semi-conductive layer.
This structure comprises a first pattern, in this example, a first transistor 100 and a second pattern, in this example, a second transistor 200. In this non-limiting example, the first transistor 100 is of NMOS-type and the second transistor 200 is of PMOS-type.
Each transistor 100, 200 comprises for example:
The active layer 130, 230 of each transistor 100, 200 surmounts an electrically insulating layer 120. Preferably, it all is supported by a support substrate 110.
Preferably, the support substrate 110, the insulating layer 120 and the active layer 130 form a substrate developed of silicon on insulator (SOI)-type, or more generally, semi-conductor on insulator-type. The insulating layer 120 is thus qualified as a buried oxide (BOX) layer. The active layer 130 is preferably very thin.
These developed substrates 100 are qualified as FDSOI (fully depleted silicon on insulator, or more generally fully depleted semi-conductor on insulator) or PDSOI (partially depleted silicon on insulator, or more generally partially depleted semi-conductor on insulator) mainly according to the thickness of the semi-conductive active layer 130, 230.
The structure illustrated in
As illustrated in
This stressed layer 190 has a thickness of a few nanometres to a few tens of nanometres. Typically, the thickness thereof is between 5 and 40 nm (10−9 metres), for example 20 nm.
The formation of the stressed layer comprises a conform deposition. It is, for example, produced by one of the following deposition techniques: plasma-enhanced chemical vapour deposition (PECVD), low-pressure enhanced chemical vapour deposition (LPCVD), rapid thermal CVD (RTCVD), physical vapour deposition (PVD), atomic layer deposition (ALD). Other types of depositions can be considered.
Preferably, the stressed layer 190 is deposited at a temperature greater than 400° C. and preferably greater than 500° C. According to an embodiment, this temperature corresponds to the temperature to which the support substrate 110 is maintained during the deposition.
Advantageously, this stressed layer 190 partially covers at least the first transistor 100 and partially at least the second transistor 200. Preferably, and as is illustrated in
The following step, illustrated in
In practice, for each transistor, the zones where the stressed layers contributing to constraining the channel of the transistor are situated around the grid of the transistor (on the top of the grid and on the flanks of the grid) as well as on the source and drain zones close to the grid.
In practice, the transistors of one same type are preferably aligned in the plane of the figures. Therefore, a protective layer 400a is deposited on a set of adjacent transistors of the same type. The stress applied on the channels of these transistors will result in the same method steps. Two transistors of different types are adjacent in a plane perpendicular to that of the figures. However, the representation illustrated in the figures makes it possible to illustrate and to explain the invention, more clearly and concisely.
Preferably, this protective layer 400a is made of resin. It is, for example, produced by solid plate depositing a photosensitive resin layer, then by defining by lithography, a protective block.
Preferably, the protective layer 400a extends until an insulation trench 300 separating the transistors 100 and 200. Typically, the protective block covering the transistors 100 has a flank 401a facing to the insulation trench 300.
The following step consists of modifying the stressed layer 190 which is not masked (i.e. the second zone 291 of the stressed layer 190) so as to invert the type of stress in this zone 291 of the stressed layer 190. However, during this step, the stress of the first zone 191 of the stressed layer 190 must not be inverted.
To this end, from a plasma, light ions can be implanted in the second zone 291 of the stressed layer. The implantation parameters 500, in particular the energy communicated to the ions, the duration and the implantation dose are provided so as to invert the type of stress in the portions modified by implantation (i.e. the second zone 291 of the stressed layer).
Preferably, this implantation 500 is done solid plate. The protective layer 400a is configured, in particular the thickness thereof and the type thereof, such that the implantation 500 does not modify the stressed layer 190 in the first zone 191.
Implanted light ions are hydrogen-based ions. The hydrogen (H)-based ions are preferably taken from among the following ions: H, H+, H2+, H3+. These ions, called “light”, have the advantage of penetrating into the stressed layer 190 without destroying it, even without damaging it.
Implantation 500 by plasma has the advantage of making it possible for a continuous implantation 500 in a volume extending from the surface of the implanted layer 190.
Advantageously, the implantation 500 is done such that the implanted ions are distributed in an uninterrupted manner from the surface of the stressed layer 190 and until a desired depth. This distribution has no discontinuity. There are therefore no layers where the implanted species are absent between the surface of the layer 190 and the maximum implantation depth.
This distribution can be homogenous. This leads to a more homogenous stress within the layer 190 than what would have been obtained with conventional implanters. Thus, the stress within the channel of the transistor 200 is best controlled.
Alternatively, this distribution can be non-homogenous. Typically, this distribution will thus decrease from the surface of the layer 190 and until said depth.
Furthermore, the use of a plasma makes it possible for an implantation 500 at shallower depths than the minimum depths that can be obtained with implanters. Thus, an implantation 500 by plasma makes it possible to effectively and relatively homogenously or at the very least continuously implant thin thicknesses. Thus, the layers are avoided being implanted and therefore altered which are situated under the stressed layer 190, even in the cases where this layer 190 is very thin.
The implantation 500 of a plasma makes it possible to typically implant thicknesses extending from the surface of the implanted layer and over a depth going from 0 nm to 100 nm. Conventional implanters, make it possible for an implantation 500 in a volume of between 30 nm and several hundred nanometres. However, conventional implanters do not make it possible to implant species between the surface of the layer to be implanted and a depth of 30 nm.
Using a plasma to modify the stressed layer 190 is therefore particularly advantageous when the latter has a low thickness, typically of between 1 and 50 nm and more generally of between 1 and 100 nm.
The modification step carried out from a plasma modifies the silicon nitride-based layer 700 continuously from the surface of this layer and over a depth of between 1 nm and 30 nm and preferably of between 1 nm and 10 nm.
Moreover, using a plasma makes it possible to implant ions with a high anisotropy and with a good precision of the implantation direction. This contributes to improving the control of the implanted zones, and therefore ultimately the compression in the channel of the transistor 200, as well as the interface between voltage stressed zone 191 and the compression stressed zone 291.
Preferably, the implantation 500 is configured so as to make the stressed layer 190 pass from a voltage stress to a compression stress. For this, a person skilled in the art will adjust, in particular, the following parameters:
A person skilled in the art knows how to adjust each of these parameters in order to invert the type of the stress in the stressed layer 190. In particular, the duration of the implantation impacts the dose in the stressed layer 190. A more significant implantation duration leads to a higher-compressed stress.
According to an embodiment, the implantation of hydrogen-based ions is done in a plasma reactor making it possible to control the energy of the ions. It can, for example, be an inductive coupling plasma (ICP) reactor, a capacitive coupling plasma (CCP) reactor, or also an immersion plasma.
Advantageously, the temperature of the plate during the implantation is less than 150° C.
As indicated above, the hydrogen-based ions come for example from an H2 plasma. This gas makes it possible to avoid the risks of destroying the stressed layer 190.
Alternatively, and as indicated above, the hydrogen-based ions can come from the following gases: NH3 and HBr. Using these gases has the advantage of damaging the stressed layer 190 during the implantation. This makes it possible to adjust the thickness of this layer 190 in the second zone 291, i.e. in this embodiment, to modify the thickness of the stressed layer covering PMOS-type transistors 200. By modifying the thickness of this layer, the stress imposed on the channel of the transistor 200 can thus be adjusted. More generally, the parameters of the plasma can be adjusted, in particular the type of ions and the energy thereof, in order to achieve this result.
Alternatively, and as indicated above, the hydrogen-based ions, can come from a CxHy-type chemistry. It is, for example, a plasma formed from the following gases: CH4 or CH4 and He. This type of gas has the same advantages as the gases NH3 and HBr mentioned above. Furthermore, this type of chemistry makes it possible to limit the consumption of resin from the protective layer 400 during the implantation 500. Indeed, when the protective layer 400a comprises carbon, as is the case when the protective layer 400a is a photosensitive or thermosensitive resin, carbon atoms, coming from the dissociation of CxHy molecules, are deposited on the resin which protects it from the bombardment of the other ions of the plasma such as argon or helium ions, for example. The resin thus conserves, during the whole implantation, the dimensions thereof and the geometry thereof.
This embodiment thus makes it possible to have a particularly high dimensional control of the carbon protective layer. Ultimately, this makes it possible to improve the dimensional control of the zone to be implanted and of the interface between the first and second transistors.
This type of chemistry thus makes it possible to conserve the geometry and the dimensions of the protective layer 400a. This results in the borders between the voltage stressed zones 191 and the compression stressed zones 291 being better controlled.
Advantageously, the implantation by plasma does not generate any creeping of the protective layer 400a or limits the creeping. Conversely, if for the step of modifying the type of stress in the uncovered zone, an exposure to an ultraviolet radiation had been resorted to, which would have required a relatively significant method time, a significant creeping of the protective layer 400a made of resin would have thus been observed. Such a creeping is represented schematically by the reference 600 in
The following step, illustrated in
Thus, on one same plate, transistors 100 covered by a voltage stressed layer 190 and other transistors 200 covered by a compression stressed layer 290 are obtained.
To finalise the structure, conventional steps can then be carried out, like for example the opening of the stressed layers 190, 290 at the level of the re-contacts 170, 180, 270, 280 to electrically connect the grid, the sources and the drain.
According to an embodiment, the stressed layer is formed by one single deposition. The modification of the second zone 291 of the stressed layer is obtained by one single step of modification by plasma implantation.
According to an alternative embodiment, the stressed layer is formed by a plurality of successive depositions. After each deposition, or at least after some of them, a step of modification by implantation is carried out. This embodiment makes it possible to have a better control of the modification of the stressed layer. Consequently, the inversion of stress and the amplitude of the stress in the second zone 291 is more specifically controlled. This embodiment also makes it possible to form thick stressed layers, making it possible, due to this, to specifically adjust the voltage in the channel of each of the transistors. For example, at each sequence, a layer of around 10 nm can be deposited, then be modified at the level of the second zone 291.
The conditions below give non-limiting examples to modify the stressed layer over a thickness of between 1 nm and a few tens of nanometres (nm), typically between 1 and 60 nm.
These modifications are made in any type of equipment making it possible to implant hydrogen ions by plasma, while controlling the energy of the implanted ions. ICP-type or CCP-type reactors can thus be resorted to.
Two examples will be given below, according to the chemistry of the plasma used.
Modification Using an H2, HBr or NH3 Type Chemistry
It will be noted that the window of the method is constrained by steps which will subsequently be carried out. In particular, the conditions mentioned above depend on the temperature which will be applied during later steps.
The conditions below give a non-limiting example to invert the stress of a silicon nitride (SiN) layer, deposited by LPCVD and having a thickness of 20 nm. More specifically, the parameters below make it possible to pass from a voltage stress to a compression stress.
Table 2 below gives the different stresses that can be obtained, after modification by implantation under the conditions of Table 1, according to:
It is thus observed, from a polarisation voltage of 400V, the voltage stress of the silicon nitride layer is converted, by implantation, into a compression stress.
Modification Using a CxHy Type Chemistry, x and y Being Whole Numbers
It will be noted, that the window of the method is broad. The conditions depend on the applications, as well as the thickness of the nitride film to be modified and to be etched. Indeed, the bombardment of ions heavier than hydrogen ions (typically helium or argon ions) leads to etching the nitride layer.
The conditions below given a non-limiting example to invert (passage from a voltage stress and a compression stress) the stress of a silicon nitride (SiN) layer having a thickness of 17 nm.
Thus, in view of the description above, it clearly appears that the proposed method makes it possible to reduce the cost and make the production of voltage and compression stressed layers more reliable on patterns of a plate, typically on transistors having different characteristics, for example NMOS and PMOS transistors.
The invention is not limited to the embodiments described above and extends to all the embodiments covered by the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 17 63190 | Dec 2017 | FR | national |
