Use of F-based gate etch to passivate the high-k/metal gate stack for deep submicron transistor technologies

[0001] This application claims the benefit under 35 U.S.C. §119(e) of U.S. provisional application Ser. No. U.S. 60/879,704 filed Jan. 10, 2007, the disclosure of which is hereby expressly incorporated by reference in its entirety and is hereby expressly made a portion of this application.

[0002] The preferred embodiments are related to the field of gate patterning in planar and multiple-gate devices. More specifically the preferred embodiments are related to the simultaneous incorporation of Fluorine into the gate stack during the patterning of said gate stack. The preferred embodiments are further related to the creation of semiconductor devices with increased lifetime and significant Bias Temperature instability (BTI) improvement. The preferred embodiments are especially beneficial for short channel devices.

[0003] The introduction of high-k and metal gate materials in scaled CMOS technologies (45 nm node and beyond) faces significant difficulties due to severe threshold voltage instability and performance degradation of the devices. These problems are related to the high amount of bulk defects and interface states in the high-k/metal gate stack. In addition, preferential lateral re-oxidation of the gate dielectric is a significant problem, especially for short channel devices. Proper passivation of the dielectric is crucial in order to overcome these problems.

[0004] Passivation of Hf based high-k dielectric materials by Fluorine as possible means to reduce the number of interface and bulk defects is an attractive alternative to H passivation, and was the subject of several recent studies. The reason for this is that Hf and Si bonds formed with F are much stronger compared to Hf—H and Si—H bonds. The strength of the formed bonds allows the passivation to be maintained even throughout the standard high temperature CMOS processing for deep submicron devices, and results in a more robust defect passivation that allows better withstanding of the normal device operation conditions, resulting in improved BTI behavior.

[0005] The conventional method for Fluorine introduction is by means of implantation techniques. However, several implantations during the gate stack deposition are needed in order to passivate the entire dielectric. The incorporation of Fluorine in an HfO₂ comprising high-k stack using implantation techniques is described by Seo et al (IEDM Technical Digest, p. 647-650, 2005). However, implanting Fluorine during the gate stack deposition may lead to uncontrolled oxide re-growth as described by Mogul et al (IEEE Trans. Electron Dev., pp. 388-394, 1997).

[0006] As a conclusion there is still a need to find a more attractive method to achieve Fluorine incorporation (passivation) in the high-k dielectric material.

[0007] According to preferred embodiments, a method is provided to increase the lifetime and performance of a semiconductor device comprising a high-k material such as HfO₂ by incorporation of Fluorine in the high-k material and more specifically to incorporate or replace the Si—H and Hf—H bonds (especially at the interface of the silicon substrate and HfO₂) by Si—F and Hf—F bonds.

[0008] According to preferred embodiments a novel method for introducing Fluorine in the high-k dielectric is disclosed.

[0009] The novel method relies on the introduction of highly reactive F ions and radicals during the metal gate etch.

[0010] Furthermore, according to preferred embodiments a cost-effective method to introduce Fluorine in the high-k dielectric, requiring no extra processing steps and still offering all the benefits of Fluorine passivation is disclosed.

[0011] According to preferred embodiments a method for fluorine passivation of a dielectric layer (or for introducing fluorine in a dielectric layer) having a high-k (i.e. a dielectric constant higher than that of SiO₂, k>k_SiO2) is disclosed, said method comprising the steps of:

• • [0012] providing gate stack layers comprising at least (or consisting of) a channel layer (or a substrate, e.g. Si, SOI), a high-k dielectric layer, and a metal gate layer,
  • [0013] then exposing said gate stack layers to a fluorine comprising plasma which does not contain carbon, and then,
  • [0014] submitting said exposed gate stack layers to a thermal treatment, wherein the temperature is higher than 700° C.

[0015] According to preferred embodiments, the fluorine comprising plasma is a plasma that does not contain carbon compounds to avoid carbon deposition onto the sidewalls of the gate structure during patterning (etching) of the gate structure.

[0016] According to preferred embodiments, exposing said gate stack layers to said fluorine comprising plasma is performed during the patterning of said gate stack layers.

[0017] Preferably, in a preferred embodiment, exposing said gate stack layers to said fluorine comprising plasma is performed simultaneously with or (immediately) after the dry etching of said gate stack layers.

[0018] More particularly, said step of dry etching of said gate stack layers and said step of exposing said gate stack layers to said fluorine comprising plasma are performed consecutively.

[0019] According to preferred embodiments, said step of exposing said gate stack layers to said fluorine comprising plasma consists of a dry etching of said metal gate using said fluorine comprising plasma.

[0020] Preferably, said fluorine comprising plasma is a SF₆ comprising plasma.

[0021] More preferably, said fluorine comprising plasma further comprises O₂.

[0022] More preferably, said fluorine comprising plasma further comprises N₂.

[0023] According to preferred embodiments, said fluorine comprising plasma comprises (or consists of) a mixture of SF₆ and O₂.

[0024] According to preferred embodiments, said fluorine comprising plasma comprises (or consists of) a mixture of SF₆ and N₂.

[0025] According to preferred embodiments, said gate stack layers are exposed to a BCl₃ comprising plasma prior to exposing said gate stack layers to said fluorine comprising plasma.

[0026] More particularly, said metal gate is etched, preferably partially etched, using a BCl₃ comprising plasma prior to exposing said gate stack layers to said fluorine comprising plasma.

[0027] More preferably, the remaining metal gate layer, corresponding preferably to 10% to 50% of the initial thickness of said metal gate layer, is etched using said fluorine comprising plasma. In other words, said exposure to said BCl₃ comprising plasma is stopped when 10% to 50% of the thickness of said metal gate layer is left, and then, during said subsequently exposure to said fluorine comprising plasma, said 10% to 50% of the thickness of said metal gate layer is further etched by said fluorine comprising plasma.

[0028] More preferably, said BCl₃ comprising plasma further comprises Cl₂.

[0029] According to preferred embodiments, said gate stack layers further comprise a capping layer onto the metal gate layer, such as poly-Si.

[0030] According to preferred embodiments, said metal gate layer is a layer comprising of (or consisting of) TiN, TaN, TiN/TaN, TaC, WN₂, TiW, WSi₂, MoSi₂, Mo, Ti, or W.

[0031] More preferably, said metal gate layer is a layer comprising of (or consisting of) TiN, TaN, TiN/TaN, or W.

[0032] Preferably, said metal gate layer is a dual metal gate.

[0033] Preferably, said dielectric layer is an Hf comprising layer.

[0034] More preferably, said Hf comprising layer is layer comprising of (or consisting of) HfO₂.

[0035] More preferably, said Hf comprising layer is a layer comprising of (or consisting of) hafnium silicon oxide (HfSi_xO_y).

[0036] Preferably, said channel layer comprises at least (or, preferably, consists of) Si.

[0037] Preferably, said channel layer comprises at least (or, preferably, consists of) Ge.

[0038] Preferably, said channel layer comprises at least (or, preferably, consists of) a group III/V element.

[0039] According to preferred embodiments, said thermal treatment is performed for at least 2 minutes.

[0040] According to preferred embodiments, said thermal treatment step is performed during a Selective Epitaxial growth (SEG) process, or during a Rapid Thermal Anneal (RTA) step.

[0041] More preferably, said thermal treatment step is performed during a Selective Epitaxial growth (SEG) process.

[0042] According to preferred embodiments, the method as described above is used for the manufacture of a semiconductor device.

[0043] More preferably, said semiconductor device is a multiple gate device (e.g. FinFET), said channel layer can be a fin.

[0044] According to preferred embodiments, a semiconductor device obtainable by the method described in preferred embodiments is obtained.

[0045] More preferably, said semiconductor device is a multiple gate device, such as a FinFET.

[0046] The term “high-k” used in the context of the preferred embodiments refers to its well-recognized meaning in the field of semiconductors.

[0047] More particularly, the term “high-k” refers to a dielectric constant higher than that of SiO₂, k>k_SiO2.

[0064] In relation to the drawings the preferred embodiments are described very briefly and in general in the following. It is apparent, however, that a person skilled in the art knows how to perform the etch process in detail and compose an etch recipe.

[0065] It is the goal of the preferred embodiments to increase the lifetime and performance of a semiconductor device comprising a high-k material such as HfO₂ by incorporation of Fluorine in the high-k material and more specifically to incorporate or replace the Si—H and Hf—H bonds (especially at the interface of the silicon substrate and HfO₂) by Si—F and Hf—F bonds.

[0066] It is the goal of the preferred embodiments to propose a novel method of introducing Fluorine in the high-k dielectric. The novel method relies on the introduction of highly reactive F ions and radicals during the metal gate etch.

[0067] It is further a goal of the preferred embodiments to propose a cost-effective method to introduce Fluorine in the high-k dielectric, requiring no extra processing steps and still offering all the benefits of Fluorine passivation.

[0068] According to preferred embodiments, a method is provided for patterning a gate stack in a semiconductor device.

[0069] Said semiconductor device can be either a planar or a multiple gate semiconductor device.

[0070] Said gate stack comprises at least a metal gate layer and a high-k dielectric layer.

[0071] Optionally said gate stack can further comprise a capping layer onto the metal gate layer (e.g. poly-Si).

[0072] During the patterning of at least one of the layers of said gate stack the high-k dielectric layer is fluorinated.

[0073] The method starts with the step of providing a stack of layers of at least a gate dielectric layer and on top of said gate dielectric layer a metal gate layer and underneath said gate dielectric layer a channel layer.

[0074] In a preferred embodiment, said metal gate layer can be a layer comprising of (or consisting of) TiN, TaN, TiN/TaN, TaC, WN₂, TiW, WSi₂, MoSi₂, Mo, Ti, or W.

[0075] According to preferred embodiments the metal gate layer is a TiN, TaN, TiN/TaN, or a W comprising (or consisting of) layer.

[0076] According to preferred embodiments the high-k dielectric layer is an Hf comprising layer such as HfO₂, or hafnium silicon oxide (HfSi_xO_y).

[0077] According to preferred embodiments the channel layer is selected from at least one of a Si, Ge or a group III/V element.

[0078] In case the semiconductor device is a multiple gate device (e.g. FinFET); said channel layer can be a fin.

[0079] The method of the preferred embodiments proceeds with the patterning of the gate stack layers by means of dry etching and performing during the patterning at least one dry etching step and/or exposure step using a Fluorine comprising plasma such that Fluorine is incorporated into the gate stack layers.

[0080] According to preferred embodiments the Fluorine comprising plasma is a SF₆ comprising plasma such as SF₆/O₂.

[0081] According to preferred embodiments the Fluorine comprising plasma is a SF₆ comprising plasma such as SF₆/N2.

[0082] After patterning the gate stack, a thermal treatment is preferably performed to activate the Fluorine in the gate stack layers such that it is diffusing in the high-k dielectric layer and stable F-bonds are realized at the interfaces between the gate stack layers.

[0083] According to preferred embodiments the thermal treatment step is performed at temperatures above 700° C.

[0084] According to preferred embodiments the thermal treatment step is performed at temperatures above 700° C. for several minutes (or for at least 2 minutes).

[0085] According to preferred embodiments the thermal treatment step is performed during a Selective Epitaxial (SEG) growth process. This is especially attractive for non-planar devices such as multiple gate devices (MugFET, FinFET, or the like) whereby a standard SEG process step is used for fin regrowth. During the SEG the Fluorine is able to diffuse and bond to the high-k dielectric (e.g. to form Hf—F bonds at the interfaces).

[0086] Furthermore, the use of the method described above is provided for etching a gate stack comprising at least one metal gate and a high-k dielectric layer such that Fluorine is incorporated into the gate stack layers and stable F-bonds are realized at the interfaces between the gate stack layers.

[0087] Also a semiconductor device obtainable by a method according to the preferred embodiments above is provided.

[0088] In a first aspect, a method is provided for patterning a gate stack in a semiconductor device, said gate stack comprising at least a metal gate layer and a high-k dielectric layer and whereby during the patterning of at least one of the layers of said gate stack the high-k dielectric layer is fluorinated such that the dangling bonds are passivated, said method comprising at least the steps of: providing a stack of layers of at least a gate dielectric layer and on top of said gate dielectric layer a metal gate layer and underneath said gate dielectric layer a channel layer, patterning the gate stack layers by means of dry etching and performing during the patterning at least one dry etching step and/or exposure step using a Fluorine comprising plasma such that Fluorine is incorporated into the gate stack layers, and performing a thermal treatment step to activate the Fluorine in the gate stack layers such that it is diffusing in the high-k dielectric layer and stable F-bonds are realized at the interfaces between the gate stack layers.

[0089] In an embodiment of the first aspect, said semiconductor device is a planar or a multiple gate semiconductor device.

[0090] In a preferred embodiment, said metal gate layer can be a layer comprising of (or consisting of) TiN, TaN, TiN/TaN, TaC, WN₂, TiW, WSi₂, MoSi₂, Mo, Ti, or W.

[0091] In an embodiment of the first aspect, said metal gate layer is a TiN, TaN, TiN/TaN, or a W comprising (or consisting of) layer.

[0092] In an embodiment of the first aspect, said high-k dielectric layer is an Hf comprising layer such as HfO₂, or hafnium silicon oxide (HfSi_xO_y).

[0093] In an embodiment of the first aspect, said channel layer is selected from at least one of a Si, Ge or a group III/V element.

[0094] In an embodiment of the first aspect, said Fluorine comprising plasma is a SF₆ comprising plasma such as SF₆/O₂.

[0095] In an embodiment of the first aspect, said Fluorine comprising plasma is a SF₆ comprising plasma such as SF₆/N₂.

[0096] In an embodiment of the first aspect, said thermal treatment step is performed at temperatures above 700° C.

[0097] In an embodiment of the first aspect, said thermal treatment step is performed at temperatures above 700° C. for several minutes (or for at least 2 minutes).

[0098] In an embodiment of the first aspect, said thermal treatment step is performed during a Selective Epitaxial growth process.

[0099] In an embodiment of the first aspect, said gate stack further comprises a capping layer onto the metal gate layer (e.g. poly-Si).

[0100] Use is provided of a method according to any of the above aspects or embodiments for etching a gate stack comprising at least one metal gate and a high-k dielectric layer such that Fluorine is incorporated into the gate stack layers and stable F-bonds are realized at the interfaces between the gate stack layers.

[0101] A semiconductor device is provided obtainable by a method according to any of the above aspects or embodiments.

[0102] The following description illustrates methods to passivate a high-k dielectric layer by incorporation of Fluorine during the patterning of the gate stack layers. It will be appreciated that there are numerous variations and modifications possible.

[0103] The preferred embodiments relate to a new, effective and cost-efficient method of introducing Fluorine into a stack comprising high-k dielectric materials, said high-k comprising stack being part of a planar or multi-gate semiconductor device (MuGFET).

[0104] The method of the preferred embodiments is further resulting in a significant improvement in both Negative and Positive Bias Temperature Instabilities (NBTI and PBTI).

[0105] According to preferred embodiments, said high-k dielectric material or gate dielectric is preferably an Hf-based material such as HfO₂. Said high-k comprising stack further comprises a gate electrode, said gate electrode is preferably a metal gate, a preferred example of said metal gate is TiN, TiN/TaN or TaN but the preferred embodiments are not limited to these metal gate layers only.

[0106] The method of the preferred embodiments uses a Fluorine comprising plasma (e.g. an SF₆ based metal gate etch chemistry) for the introduction of F into the high-k dielectric, which after a thermal treatment within the standard process flow, results in excellent F passivation at the interfaces of the gate stack layers.

[0107] The method of the preferred embodiments has the advantage over existing methods in prior art by using a Fluorine based (e.g. SF₆-based) plasma exposure during the patterning of the gate stack for Fluorine introduction such that no extra Fluorine implantations or treatments are required. In addition to a significant Bias Temperature Instability (BTI) improvement, for the first time, a better V_th control and increased drive current on MuGFET devices is thereby achieved.

[0108] Using the method of the preferred embodiments makes it possible to introduce highly reactive Fluorine ions and radicals during the patterning (in a dry etch chamber) into the high-k dielectric. The method is further characterized as a cost-effective method, requiring no extra processing steps and still offering all the benefits of Fluorine passivation.

[0109] The incorporation of Fluorine into the gate dielectric using the method of the preferred embodiments is further characterized as an in-line processing step whereby the Fluorine based (e.g. SF₆-based) gate stack etch chemistry not only etches the gate stack (e.g. metal gate layer) but simultaneously incorporates Fluorine in the high-k dielectric layer such that the necessary amount of Fluorine needed to passivate the dielectric layer at the interface is achieved and no additional processing steps are needed to incorporate Fluorine into the high-k dielectric layer.

[0110] In a preferred embodiment, the incorporation of Fluorine into the gate stack (more particularly into the high-k dielectric layer) is performed during the dry etching of the metal gate layer using a Fluorine comprising plasma. Said plasma is preferably a SF₆ comprising plasma such as SF₆/O₂. The Fluorine is preferably such that enough Fluorine is incorporated into the gate stack. Said Fluorine is accumulated at the sidewalls of the gate stack but can be further transported/diffused by means of a thermal treatment step performed after the patterning.

[0111] In another and also preferred embodiment, the incorporation of Fluorine into the gate stack (more particularly into the high-k dielectric layer) is performed during the dry etching of the metal gate layer. The patterning of the metal gate layer is performed by means of at least two steps whereby the first step comprises a BCl₃ comprising plasma (e.g. BCl₃/Cl₂) to etch the bulk of the metal gate layer and a second step using a Fluorine comprising plasma (e.g. SF₆/O₂) to etch the remaining metal gate layer and to incorporate Fluorine into the gate stack.

[0112] If needed, a supplemental Fluorine plasma exposure can be performed. This extra Fluorine exposure is preferably performed immediately after performing the dry etching of the gate stack layers (in the dry etching chamber). Said Fluorine treatment is preferably performed in a dry etching plasma using a Fluorine based chemistry (plasma) without applying a substrate bias (no etching).

[0113] In yet another and also preferred embodiment, the incorporation of Fluorine into the gate stack (more particularly into the high-k dielectric layer) is performed after the patterning of the gate stack by exposure to a Fluorine comprising plasma.

[0114] FIG. 1 illustrates at the left side part of the gate stack 5 of a prior art device (also referred to as reference device) after patterning said gate stack using no Fluorine comprising chemistry (using a BCl₃/Cl₂ chemistry to pattern the metal gate).

[0115] The gate stack shown comprises a metal gate layer (here TiN) 3 with a high-k layer 2 underneath (here HfO₂).

[0116] Underneath the high-k layer 2 a Si (substrate) 1 is situated.

[0117] Also the interface 4 between the Si bulk and the high-k dielectric layer is indicated.

[0118] In the prior art device, Si—H and Hf—H bonds are present in the stack after patterning.

[0119] On the right side of FIG. 1, the gate stack after patterning according to the preferred embodiments is shown whereby a Fluorine comprising plasma (e.g. SF₆/O₂) is used to pattern the gate stack such that Fluorine is incorporated in the high-k dielectric 2.

[0120] After applying a thermal treatment step the Fluorine is diffused throughout the bulk of the high-k layer 2 such that passivation of the (dangling bonds) high-k layer is realized (especially at the interfaces 4).

[0121] Si—F and Hf—F bonds are indicated in the gate stack.

[0122] Compared to existing methods to achieve Fluorine incorporation in the high-k dielectric layer such as ion implantation, the method of the preferred embodiments is a gentle method which avoids damages to the high-k dielectric such as uncontrolled oxide re-growth.

[0123] The in-line incorporation of Fluorine during the metal gate etch needs to be carefully balanced in such a way that at the same time optimal passivation of the dangling bonds can be achieved.

[0124] FIG. 2 shows a gate stack comprising TiN as metal gate layer 12 and HfO₂ as high-k dielectric 13 in a FinFET device (Si fins 15 shown).

[0125] The Figure further illustrates the Fluorine diffusion and accumulation in the hard mask 10 during the etch process.

[0126] Further diffusion of Fluorine species towards the (high-k) dielectric 13 interfaces can take place during the thermal treatment step.

[0127] The thermal treatment step is preferably performed at temperatures above 700° C.

[0128] The thermal treatment step is preferably performed at temperatures above 700° C. for several minutes (at least 2 minutes).

[0129] According to preferred embodiments, the thermal treatment step can be performed during a Selective Epitaxial growth process.

[0130] Furthermore the preferred embodiments provide a semiconductor device obtainable by the preferred methods described above.

[0131] MuGFET devices have been fabricated on standard [100]/[110] (top surface) SOI substrates with fin height H_fin=60 nm and fin widths down to 20 nm.

[0132] After fin patterning, the devices received a H₂ anneal to smoothen the sidewalls and round the corners. The gate stack consists of a 5 nm ALD TiN layer on top of a 2 nm HfO₂/1 nm SiO₂ gate dielectric.

[0133] 100 nm poly is used as a capping layer.

[0134] After gate stack deposition, an oxide hard mask (HM) is used to pattern the poly-silicon, stopping on the TiN layer.

[0135] For the TiN etch, BCl₃/Cl₂ was used for the reference devices.

[0136] For the other devices, SF₆/O₂ is used as gate-etch chemistry. The benefits of this chemistry are that it etches the TiN layer and at the same time introduces the necessary amount of F needed to passivate the interfaces of the high-k dielectric.

[0137] This method was demonstrated both for planar and MuGFET devices.

[0138] After gate patterning, the extensions were implanted (high angle), and 45 nm recessed RTCVD nitride spacers were formed.

[0139] Next, Si selective epitaxial growth (SEG) films were grown at 810° C. on nMOS S/D regions, and SiGe SEG was grown on pMOS at 750° C.

[0140] SEG in MuGFET is a key processing step in order to reduce the intrinsically high S/D series resistance.

[0141] The thermal budget of this processing step is crucial for the effectiveness of the SF₆ chemistry for F passivation.

[0142] The SEG thermal budget allows further diffusion of the F and aids Si—F bond formation.

[0143] Next, the gate HM was removed and HDD implantations were performed. Dopants were activated with a 1050° C. spike anneal and NiSi was used as a salicide.

[0144] The metal gate etch processing is schematically depicted shown in FIG. 2. It is an isotropic etch and as consequence it has the same impact on all sides of the device.

[0145] During the TiN etch, F diffuses and accumulates in the hard mask (HM).

[0146] The thermal budget of the subsequent SEG step (750-810° C.) assists further diffusion of the accumulated F towards the dielectric interfaces. It also provides the necessary activation energy for passivating the interface states and/or for substituting the H atoms in the existing Si—H/Hf—H weaker bonds.

[0147] FIG. 3 shows the nFET and pFET saturation V_th versus gate length (LG).

[0148] An anomalous V_th behavior is seen for the reference devices (using no SF₆ chemistry during the metal gate etch), a roll-up for nMOS and enhanced roll-off for pMOS). This behavior is typically attributed to process damage generated at the gate edges during device fabrication.

[0149] The SF₆/O₂ devices do not exhibit this behavior indicating that this chemistry neutralizes the gate-edge traps and/or prevents re-oxidation at the edges.

[0150] This is again demonstrated in FIG. 4 where the normalized gate leakage current (JG) is shown as function of gate length (LG).

[0151] Due to re-oxidation and damage at the gate edges, the normalized gate leakage (JG) is decreased as function of LG. However, the effect is significantly reduced for the devices using a SF₆ based chemistry during metal gate etch.

[0152] The improved I_on-I_off behavior of nFET and pFET with SF₆ is shown in FIGS. 5A and 5B.

[0153] Interrupted stress (0.5 s delay) BTI measurements were performed at stress VG in the |1-2.2| V range and temperatures from 125 to 175° C.

[0154] The studied MuGFET devices (W_fin≦H_fin) have dimensions W_fin=20 nm, H_fin=60 nm, N_fins=30 (various LG). Planar devices (W_fin>>H_fin) have LG=W_fin=1 μm.

[0155] Si—F and Hf—F bonds have higher dissociation energies than Si—H and Hf—H, and are not broken during stress conditions similar or higher to typical operating conditions, such as temperature of 125-175° C. and electric fields (E_ox) of 3-8 MV/cm (see FIG. 6).

[0156] FIG. 7 shows that the E_ox, distribution in MuGFET and planar SOI devices is similar. As a consequence, the same E_ox, evaluation method can be used in both cases.

[0157] FIG. 8 shows the NBTI lifetime improvement for devices having used a SF₆/O₂ etch chemistry during metal gate etch (and additional selective epitaxial growth (SEG) as described above at 810° C.), as compared to devices having used a BCl₃/Cl₂ without Fluorine (and additional selective epitaxial growth (SEG) as described above at 810° C.).

[0158] This is shown for both planar and MuGFET devices.

[0159] The effective area of planar devices is dominated by the top surface [100], while in MuGFET devices it is dominated by the sidewalls [110] (whereby A=N_finsL_gate(W_fin+2H_fin).

[0160] This improvement demonstrates that F presence at the interface is beneficial for both [100] and [110] Si surface orientations.

[0161] The higher lifetime observed in the case of planar devices, compared to the MuGFET devices is consistent with the higher amount of Si bonds in the [110] surface orientation, compared to [100] (see FIG. 6).

[0162] The BTI time dependence slopes vary from 0.2 to 0.25 (see FIG. 7). In this case the Reaction-Diffusion model suggests that the degradation mechanism is governed by H diffusion from the interface, rather than by hole trapping.

[0163] This suggestion is also confirmed by the activation energies (0.16-0.2 eV) shown in FIG. 10.

[0164] The reference and F passivated devices exhibit similar degradation mechanisms, governed by hydrogenic diffusion from the interface. This is an indication that during stress only the weaker Si—H/Hf—H bonds dissociate and the lifetime improvement in the case of F passivated devices is due to the presence of stable Si—F/Hf—F bonds at the interface.

[0165] Since a detrimental effect of F on the dielectric is known, low frequency 1/f noise measurements were performed to further investigate the dielectric quality.

[0166] FIG. 11 shows the drain current noise spectral density SID/ID2 and the scaled input referred noise SVG of the pMOS planar and MuGFET reference and F passivated devices.

[0167] No additional dielectric degradation due to F is observed.

[0168] The lower noise spectral density in the case of planar devices is consistent with the difference in the dominant surface orientation.

[0169] Similar results are obtained for PBTI and 1/f noise in nMOS devices (see FIGS. 12 and 13).

[0170] In order to investigate the role of the SEG thermal budget, nMOS devices with the same etch chemistries as described in the examples above (the reference using a BCl₃/Cl₂ chemistry (no Fluorine) and the chemistry according to the preferred embodiments comprising a SF₆/O₂ chemistry) but without SEG were fabricated. In this case the hard mask (HM) was removed without applying a thermal budget beforehand.

[0171] FIG. 14 shows no significant difference in lifetimes between reference and SF₆-etch devices in this case. Also, no difference is observed in the 1/f behavior and activation energies.

[0172] This confirms that the SEG thermal budget is crucial for F diffusion and bonding.

[0173] The lack of ΔV_th(LG) dependence in FIG. 15 suggests isotropic F diffusion from the top as well as from the sidewalls/gate-edges of the devices and that it is not LG dependent.

[0174] All references cited herein are incorporated herein by reference in their entirety. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and/or take precedence over any such contradictory material.

[0175] The term “comprising” as used herein is synonymous with “including,” “containing,” or “characterized by,” and is inclusive or open-ended and does not exclude additional, unrecited elements or method steps.

[0176] All numbers expressing quantities of ingredients, reaction conditions, and so forth used in the specification and claims are to be understood as being modified in all instances by the term “about.” Accordingly, unless indicated to the contrary, the numerical parameters set forth in the specification and attached claims are approximations that may vary depending upon the desired properties sought to be obtained by the present invention. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the scope of the claims, each numerical parameter should be construed in light of the number of significant digits and ordinary rounding approaches.

[0177] The above description discloses several methods and materials of the present invention. This invention is susceptible to modifications in the methods and materials, as well as alterations in the fabrication methods and equipment. Such modifications will become apparent to those skilled in the art from a consideration of this disclosure or practice of the invention disclosed herein. Consequently, it is not intended that this invention be limited to the specific embodiments disclosed herein, but that it cover all modifications and alternatives coming within the true scope and spirit of the invention as embodied in the attached claims.