Methods for gas phase selective etching of silicon over silicon-germanium layers

Abstract

Methods for selectively etching Si relative to SiGe are provided. The methods utilize a sulfur-containing additive gas in combination with a nitrogen-containing gas and a fluorine-containing gas under certain controlled operating conditions and process sequences. The methods can occur in a cyclic manner including purging after each etch step, until the desired thickness of the Si layer is obtained.

Description

METHODS FOR GAS PHASE SELECTIVE ETCHING OF SILICON OVER SILICON-GERMANIUM LAYERSField of Invention

[0001] This invention relates to selective etching of silicon layers on a wafer in connection with semiconductor fabrication. More particularly, the invention pertains to novel methods for improving the selectivity ratio of silicon layers relative to silicongermanium layers during semiconductor fabrication that is not possible with conventional etchant techniques.Background of the Invention

[0002] Amorphous silicon (a-Si), polycrystalline silicon (poly-Si) and single crystal Si, in conjunction with silicon-germanium (“SiGe”) are widely used in the semiconductor industry for different applications. For sub-20 nm technology nodes, the integration of SiGe as a source-drain material has shown potential to improve electrical performance for transistors. Various transistors can be formed on a substrate.

[0003] In scaling beyond 5 nm technology nodes, the semiconductor device industry is moving away from FinFet transistor structures towards so-called gate all-around (GAA) device architectures. While the gate in the Fin FET transistor surrounds the channel on three sides, the GAA surrounds the SiGe channel on all four sides to allow better control of the transistor switch. One fundamental requirement for GAA implementation is the formation of SiGe and Si nanowires (NW) or nanosheets (NS). Specifically, horizontally-stacked silicon nanowires (HNW), and horizontally-stacked nanosheets (HNS) are being investigated. HNS has been identified as the next generation device structure at 3 nm node and sub-3nm nodes.

[0004] In order to successfully manufacture HNS structures, advanced methodologies are required to address the challenges and meet the demands for aggressive patterning that continue to reduce the size of the critical features. The ability to successfully integrate patterning schemes with highly selective etch processes is critical to reliable pattern transfer. Pattern transfer can be carried out as follows. After the circuitpattern is formed, a protective layer (e.g., photo-sensitive material patterned using optical lithography, a mechanically imprinted patterned layer or a direct self-assembled layer) is used to mask certain regions of the semiconductor substrate, while other regions of the semiconductor substrate remain exposed. The remaining exposed regions allow the transfer of the circuit pattern onto an underlying layer of the substrate by utilizing a dry etching process. SiGe HNS for GAAFET architectures are considered the next generation devices as the SiGe channel enables greater hole mobility. However, the ability to fabricate the HNS features requires an etch process with high selectivity of Si relative to SiGe. The etch selectivity is defined as the ratio of etch rate for two different surfaces when exposed to an etchant gas. For example, an etch selectivity of 100: 1 for Si to SiGe under exposure to a certain etchant gas means that the etch rate for the Si surface is 100 times higher than a SiGe surface for a particular etch process.

[0005] Conventional etching techniques such as direct plasma etch, plasma atomic layer etching (ALE), or ion milling are not capable of achieving a higher etch selectivity ratios of Si to SiGe required for creating SiGe HNS features in a GAAFET. By way of example, for three stacked SiGe NS’s, there are four Si sacrificial layers. The removal of the Si sacrificial layers without SiGe NS damage is required and represents a critical process step in GAA HNS in a GAAFET devices. Silicon within stacked layers must be etched laterally selective to SiGe. The undercut or cavity Si etch requires high precision and high selectivity isotropic chemical etching not possible by conventional etch techniques.

[0006] Other traditional etchant methods involve oxidative wet etch methods that are commonly seen in selective etching of Si over SiGe in the micro-electromechanical systems (MEMS) industry and in early GAA HNS research. However, generally speaking, the semiconductor industry prefers not to utilize a wet-based etchant method. Instead, the industry desires a selective dry chemical etch for better Si HNS release quality and easier high-volume manufacturing processes.

[0007] Despite current developmental activities, there remains a continued demand for higher etch selectivity to improve GAA nanostructure performance and increasemanufacturing productivity. There is currently an unmet demand for improved etch selectivity of Si to SiGe.Summary of the Invention

[0008] In one aspect, a method of using an additive composition to improve selectivity of dry etching a substrate, comprising the steps of: positioning a substrate in an etch chamber, said substrate comprising silicon-germanium layers and silicon layers; and selectively etching the silicon layers over the silicon-germanium layers by (A) introducing a passivating composition of a nitrogen-containing gas and the additive composition comprising sulfur-containing gas into the etch chamber towards an exposed surface of the substrate to form a passivated silicon-germanium layer of the substrate; followed by (B) introducing an etchant composition comprising a fluorine-containing gas, the nitrogencontaining gas and the additive composition comprising sulfur-containing gas into the etch chamber to etch at least a portion of the silicon layer of the substrate without substantial etching of the passivated silicon-germanium layer.

[0009] In a second aspect, a method of using an etchant composition to improve selectivity of dry etching a substrate, comprising the steps of: providing a substrate in an etch chamber, said substrate comprising silicon-germanium layers and silicon layers; and selectively etching silicon layers with the use of the etchant composition comprising an additive of a sulfur-containing gas while using a passivating gas composition to create and maintain a passivating layer on the silicon-germanium layer.

[0010] In a third aspect, a method of performing etching using an additive composition to improve selectivity of dry etching a substrate, comprising the steps of: positioning a substrate in an etch chamber, said substrate comprising silicon-germanium layers and silicon layers; performing a first etch cycle, comprising: (A) either (i) coflowing a nitrogen-containing gas and a sulfur-containing gas into the etch chamber towards an exposed surface of the substrate or (ii) sequentially flowing the nitrogencontaining gas and then flowing the sulfur-containing gas towards the exposed surface; (B) forming a passivated silicon-germanium layer on the substrate; followed by (C) introducing a fluorine-containing gas, the nitrogen-containing gas and the sulfur-containing gas into the etch chamber to etch at least a portion of the silicon layer of the substrate without substantial etching of the passivated silicon-germanium layer; and (D) introducing a purge gas into the etch chamber to remove etch by-products created from the etch; and initiating an optional second cycle by re-establishing the passivated silicongermanium layer followed by introducing the fluorine-containing gas, the nitrogencontaining gas and the sulfur-containing gas into the etch chamber to etch at least a portion of the silicon layer of the substrate without substantial etching of the passivated silicongermanium layer.Brief Description of the Drawings

[0011] Fig. 1a illustrates a representative Si / SiGe structured film to be selectively etched by the methods of the present invention;

[0012] Fig. 1b illustrates the resultant Si / SiGe structured film of a GAAFET device that is formed by removal of Si films without removal of SiGe films from Fig. 1a in accordance with the methods of the present invention;

[0013] Fig. 2a illustrates a process for a conventional etch process utilized to selectively etch Si over SiGe;

[0014] Fig. 2b illustrates a process for an etch process utilized to selectively etch Si over SiGe in accordance with the principles of the present invention;

[0015] Fig. 3 shows a representative schematic of the etching system of the present invention;

[0016] Figs. 4a-4d show mass spectrometer results indicating the type of etch byproducts produced during the etch process of Comparative Example 1 and Example 1;

[0017] Fig. 5a illustrates a representative schematic of Si / SiGe structured films produced by the conventional etch process of Fig. 2a;

[0018] Fig. 5b illustrates a representative schematic of Si / SiGe structured films produced by the etch process of Fig. 2b in accordance with the principles of the present invention;

[0019] Figs. 6a, 6b and 6c show selective Si / SiGe etching recipes for a continuous etch process; a 2-cyle etch process; and a 5-cycle etch process, respectively;

[0020] Figs. 7a and 7b show etch performance results using the continuous etch process recipe of Fig. 6a;

[0021] Figs. 7c and 7d show etch performance results using the cyclic etch process of Fig. 6b and Fig. 6c, respectively;

[0022] Fig. 8a shows a 5-cycle etch recipe utilizing FFS as an additive gas with argon purging in between the etch cycles;

[0023] Fig.8b shows a 5-cycle etch recipe utilizing H2S as an additive gas without argon purging in between the etch cycles;

[0024] Fig. 9a shows etch performance results utilizing the process of Fig. 8a;

[0025] Fig. 9b shows etch performance results utilizing the process of Fig. 8b;

[0026] Fig. 10a shows a process recipe whereby H2S is separately fed into an etch chamber before the passivation step of co-flowing NH3 with H2S;

[0027] Fig. 10b shows a process recipe whereby NH3 is separately fed into an etch chamber before the passivation step of co-flowing NH3 with H2S;

[0028] Fig. 11a shows the etch performance results utilizing the process recipe of Fig. 10a;

[0029] Fig. 11b shows the etch performance results utilizing the process recipe of Fig. 10b;

[0030] Fig. 12 shows a gas flow sequence utilizing HF as a replacement for H2S;

[0031] Fig. 13 shows etch performance results utilizing the process of Fig. 12;

[0032] Fig. 14 shows a process recipe in which H2S, NH3 and F2 are fed separately into an etch chamber; and

[0033] Fig. 15 shows the etch performance results utilizing the process recipe of Fig. 14.Detailed Description of the Invention

[0034] As will be described, the present invention offers methods for performing selective dry etching of a substrate. Si can be preferentially etched relative to SiGe by using the methods of the present invention. The higher selectivity ratios of Si to SiGe of the present invention allow fabrication of GAA device architectures.

[0035] The terms “conduit” and “conduit flow network” as used herein and throughout means tube, pipe, hose, manifold and any other suitable structure that is sufficient to create one or more flow paths and / or allow the passage of gas species.

[0036] The terms “connected” and “operably connected” as used herein and throughout means a direct or indirect connection between two or more components by way of conventional piping and assembly, including, but not limited to valves and conduit, unless specified otherwise, so as to enable fluid, mechanical, chemical and / or electrical communication between the two or more components.

[0037] The term “dry etch” as used herein and throughout means a thermal process without plasma.

[0038] The terms “etch selectivity” and “selective etching” and “selectivity” and “selectivity ratio”, as used herein and throughout may be used interchangeably to mean the ratio of etch rate for two different surfaces when exposed to an etchant process, where etch rate is typically measured in units of nanometers per minute.

[0039] The terms “etcher” and “etch chamber” and “chamber” may each be used interchangeably herein and throughout and are intended to have the same meaning.

[0040] The terms “film” and “layer” as used herein and throughout may be used interchangeably and are intended to mean a continuous, semi -continuous or non-continuous deposition of a certain material along a substrate that may be composed of a single layer or one or more layers.

[0041] Unless specified otherwise, all of the passivating species and etchant species and inert purge and carrier materials described herein and throughout are intended to be in the gas phase.

[0042] SiGe” or “SiGe alloy” or “silicon-germanium alloy” as used herein and throughout may be used interchangeably to mean an alloy of silicon and germanium with varying compositional ratios of Si to Ge.

[0043] Si” or “silicon” as used herein and throughout may be used interchangeably; “Ge” or “germanium” as used herein and throughout may be used interchangeably; “Ar” or “argon” as used herein and throughout may be used interchangeably; “F2” or “fluorine” as used herein and throughout may be usedinterchangeably; “NH3” or ammonia as used herein and throughout may be used interchangeably; and “H2S” or “ hydrogen sulfide” as used herein and throughout may be used interchangeably.

[0044] “Substrate” as used herein and throughout is intended to mean any portion of a semiconductor or other electronics device such as a semiconductor wafer, or one or more layers on or overlying a base substrate structure.

[0045] The term “source” as used herein and throughout includes, but is not limited to, cylinders, dewars, bottles, and bulk or microbulk tanks.

[0046] The term “pure” as used herein and throughout means 98 vol% or greater purity of a particular species.

[0047] The term “non-passivated surface” as used herein and throughout is intended to mean a reactive or non-reactive surface that has a tendency to be removed by an etchant at a greater rate compared to a passivated surface. The term “passivated surface” as used herein and throughout is intended to mean a reactive or non-reactive surface that becomes a protective layer which has a tendency to be resistant to the etchant or to be removed by the etchant at a substantially lower rate in comparison to a non-passivated surface.

[0048] All concentrations herein and throughout are expressed as a volume percentage, unless specified otherwise.

[0049] Where a range of values describes a parameter (e.g., physical properties, measured variables or dimensions), all sub-ranges, point values within that range and endpoints defining a range are explicitly disclosed therein.

[0050] The designation “(a seem NH / b seem Ar / c seem F2 / d seem H2S)” is intended to mean that NH3, Ar, F2 and H2S are co-flowed at their respective flow rates of a, b, c and d seem, where a, b, c and d are intended to designate certain flow rates as specified in the examples. It should be understood that similar designations are utilized to refer to other combinations of gases flowing at certain flowrates.

[0051] The embodiments as described herein are intended to refer to epitaxial grown SiGe and Si structures. However, the present invention can be applicable to other growth modes.

[0052] The embodiments as described below are by way of example only, and the invention is not limited to the embodiments illustrated in the drawings. It should be understood that the drawings are not to scale and in certain instances details have been omitted, which are not necessary for an understanding of the embodiments, such as the etch chamber process schematic and conventional details of fabrication and assembly of the Si / SiGe GAAFET. It should also be understood that the exact structural configurations of GAA device architectures are not drawn to scale, and certain features are intentionally omitted in each of the drawings to better illustrate various aspects of the etchant processes in accordance with the principles of the present invention.

[0053] The embodiments are described with reference to the drawings in which similar elements are referred to by like numerals. The relationship and functioning of the various elements of the embodiments are better understood by the following detailed description. The detailed description contemplates the features, aspects and embodiments in various permutations and combinations, as being within the scope of the disclosure. The disclosure may therefore be specified as comprising, consisting or consisting essentially of, any of such combinations and permutations of these specific features, aspects, and embodiments, or a selected one or ones thereof.

[0054] In one embodiment of the present invention, an improved dry etch process is provided for selectively etching Si layers relative to SiGe layers on a substrate. The process is used to produce a GAAFET structure. By way of non-limiting example, Fig. 1a shows the starting Si / SiGe structured film to be selectively etched by the methods of the present invention. The Si / SiGe structured film may be composed of any number of alternating layers of Si and SiGe. The composition of each SiGe layer may vary with respect to Ge content from 0.1 atomic % to 90 atomic %. The Si / SiGe structured film may contain layers of other materials such as silicon oxide or silicon nitride coming into contact with either SiGe, Si, or both. By implementing the methods of the present invention, the Si layers can be selectively etched in a substantially uniform manner while preserving theintegrity of the SiGe layers and other layers, thereby creating the structured film of Fig. 1b. In comparison to conventional methods, the present invention can etch higher amounts of Si films while reducing, minimizing, or eliminating the amount of SiGe film that is removed, thereby creating a significantly increased selectivity ratio of Si / SiGe that has not been previously achievable.

[0055] The process flow schematic of Fig. 2 may be utilized to perform Si selective etching. Fig. 3 shows a representative etch system. The etch system includes an etch chamber 106, which has a platter 107 and platter support 108 onto which the substrate deposited with a Si / SiGe structured film is positioned. The etch chamber 106 is operably connected to a gas supply system, which includes Ar source 101, NHs source 102, H2S source 103, and aFz / Ar source 104. The Ar source 101 supplies Ar containing gas, the NH3 source 102 supplies NH3 containing gas, the H2S source 103 supplies H2S containing gas and the F2 / Ar gas source 104 supplies a F2 containing gas with argon. It should be understood that the term “containing gas” as used herein and throughout means that each gas source can be a pure gas or a mixture of two or more gases containing a target gas. By non-limiting example, the F2 containing gas can be pure F2, or a mixture of F2 with another gas, with the F2 being the target gas and having a gas concentration ranging from 1 vol% to 50 vol%, preferably 10 vol% to 30 vol%, and more preferably 15% to 25 vol% with the balance being the other gas or gases. Each gas source 101, 102, 103 and 104 is connected to a dedicated mass flow controller 205, 206, 207 and 208, respectively, to regulate flow of the gasses therefrom to the etch chamber 106. Each mass flow controller 205, 206, 207 and 208 is preferably installed in a dedicated supply conduit. Pre-mixing of H2S with F2 and H2S with NH3 necessitates a controlled, potentially sequential approach, as undesirable reactions can occur when H2S contacts F2 or NH3 prior to delivery of these species to the substrate. In this regard, and as will be explained, the present invention controls the specific conditions of delivery, including the degree of pre-mixing of H2S with NH3 and F2 prior to contacting the substrate with the Si / SiGe structured film deposited thereon.

[0056] The etch chamber 106 is connected to a recirculator 111 for cooling the platter 107 and platter support 108 so as to maintain the desired temperature of the substrate (not shown) that is positioned on the platter 107 during the etch process.Pressure transducer 112 is used to measure the pressure level in the etch chamber 106 while dry vacuum pump 115 is configured to evacuate the interior of the etch chamber 106 to operating pressure levels. Pressure control valve 114 is shown installed in an exhaust conduit and is configured to control the internal pressure of the etch chamber 106 by opening and closing as needed for the dry vacuum pump 115 to establish desired pressure levels inside the etch chamber 106. An in-situ mass spectrometer 113 is connected to etch chamber 106 and is configured to receive etch by-products from the internal volume of the etch chamber 106. Thermocouple 109 and thermocouple display 110 allow temperature to be read from the platter 107.

[0057] One embodiment of a dry etch process which uses the etch system of Fig. 3 will now be described. One or more substrates (not shown in Fig. 3) are positioned on the platter 107, which is connected to platter support 108. The substrate has Si layers and SiGe layers similar to that shown in Fig. la to create a structured film. With pressure control valve 114 in the open position, dry vacuum pump 115 is activated to evacuate the interior of etch chamber 106 until pressure levels of 0.001 to 100 Torr, preferably 1 to 50 Torr and more preferably 5 to 25 Torr are achieved. The platter 107 and platter support 108 are temperature adjusted to achieve a temperature of 60 to 150 deg C, preferably 70 to 90 deg C and more preferably 75 to 85 deg C. All of the gas species in the invention to be discussed hereinbelow are preferably pure or a mixture of pure species.

[0058] Having established the proper conditions for the operating temperature and operating pressure which are thereafter maintained in the controlled range, Ar gas is introduced from Ar source 101. The Ar gas is regulated at a predetermined flow rate using mass flow controller 205. The Ar gas enters into the top of the etch chamber 106 to purge the interior of the etch chamber 106. With the Ar gas preferably continuing to flow through the etch chamber 106 at substantially the same predetermined flow rate, NH3 gas and H2S gas are supplied from their respective sources 102 and 103 through corresponding mass flow controllers 206 and 207 and preferably co-flowed into the etch chamber 106 as shown in Fig. 3. Excess pre-mixing of H2S and NH3 upstream of the inlet to the etch chamber 106 is sufficiently reduced, minimized or eliminated to avoid risk of deleterious reaction of the NH3 and H2S, which can generate reaction products that render inoperable the passivation.A predetermined flow rate ratio of H2S gas to NH3 gas is created and maintained in a range between about 1:1000 to 100:1, preferably between about 1:100 to 10:1 and more preferably between about 1: 10 to 1:1. The exact flow rate ratio of H2S to NH3 is selected to optimize creation of relatively high etch rates of Si without compromising selectivity of Si over SiGe. Generally speaking, the inventors have observed an insufficient flow rate of H2S relative to NH3 may can create unacceptably low etch rates of Si over SiGe, whereas an excess flow rate of H2S relative to NH3 can adversely cause both Si and SiGe to be etched when F2 and H2S are subsequently introduced onto the Si / SiGe structured film. Upon entering etch chamber 106, the H2S gas and the NH3 gas flow towards an exposed surface of the substrate to form passivated SiGe layers. The Si layers remains as a non-passivated surface. The step of passivation occurs for a duration that does not exceed a predetermined upper limit. Exceeding the predetermined upper limit can cause H2S to irreparably damage the Si / SiGe structured film. In a non-limiting example, the step of passivation can occur for greater than 0 seconds to less than 600 seconds, preferably less than 100 seconds and more preferably less than 50 seconds.

[0059] After passivating the SiGe layers, and while preferably maintaining the same regulated flow rates of H2S, NH3 and Ar gases from mass flow controllers 207, 206 and 205, respectively, a mixture of F2 / Ar is introduced from source 104 through mass flow controller 208. The H2S gas and the NH3 gas continue to flow through the etch chamber 106 at their predetermined flow rate ratio. Next, the mixture of F2 / Ar is regulated by its mass flow controller 208 and then is directed into the etch chamber 106. Excess pre-mixing of H2S and F2 upstream of the inlet to the etch chamber 106 is sufficiently reduced, minimized or eliminated to avoid risk of deleterious reaction of the H2S and F2 species, which can generate deleterious reaction products that render inoperable the etching. The mixture of F2 / Ar flows at a flow rate such that a predetermined flow rate ratio of H2S to F2 is created and maintained in a range between about 1: 1000 to 100: 1, preferably between about 1: 100 to 10: 1, and more preferably between about 1:10 to 1:1. The H2S and F2 / Ar gas mixture are co-flowed and enter the etch chamber 106 where they flow towards the substrate to etch the Si layers without substantially etching the passivated SiGe layers. The selective etching of Si to SiGe occurs at a selectivity ratio of at least about 10: 1, preferably at least about 1000: 1 and morepreferably at least about 2200: 1. The H2S serves as an additive gas that in combination with F2 / Ar improves the etch profile of the Si layer while preserving the SiGe layers, and increases the Si etch rate over SiGe (as demonstrated in Examples 1-5). The etch rate of the Si in the Si / SiGe structured film is at least about 1 nm / min, preferably at least about 100 nm / min and more preferably at least about 250 nm / min. Additionally, the H2S in combination with the F2 / Ar enables the increased Si selective etching to occur with a reduction, minimization, or elimination of the formation of deleterious residues on the surface of the substrate by reducing the reactions of certain nitrogen-based precursors with fluorine-based precursors on the Si surface as will be explained in the Examples. The presence of the residues can facilitate etching of SiGe. Hence, the reduction, minimization or elimination of residue formation by the present invention enables higher Si selective etch rates relative to SiGe than previously possible with conventional processes more prone to residue pickup on the wafer surface. The residue may be composed of ammonium fluorosilicate (AFS), ammonium fluorogerminate (AFG), ammonium sulfide, ammonium bisulfide, or a combination of thereof. The conventional processes and formation of the residue is disclosed in U. S. Patent Pub. No. 2023 / 0260802, which is incorporated herein by reference in its entirety for all purposes.

[0060] The step of etching occurs for a duration that does not exceed a predetermined upper limit whereby H2S can irreparably damage the Si / SiGe structured film. The step of etching can occur for greater than 0 seconds to less than 600 seconds, preferably less than 200 seconds and more preferably less than 100 seconds. Having determined the required amount of selective Si etch over SiGe has occurred, the flow of H2S, NH3 and Ar / F2 is stopped while the Ar gas continues to preferably flow from source 101 at the same flow rate as regulated by its mass flow controller 205 to remove (i.e., purge) from the etch chamber 106 the etch by-products generated from the Si selective etching.

[0061] After completion of the purging, a second cycle for etching is preferably initiated by re-establishing the passivated SiGe layer. The predetermined flow rate ratio of H2S gas to NH3 gas is created and maintained at a certain flow rate ratio that is within the ranges mentioned hereinabove. Depending on the specific surface profile created in thefirst etch cycle, the exact parameters utilized in the second passivation cycle may be adjusted but remain within the prescribed ranges disclosed hereinabove. Subsequently, the F2 / Ar gas mixture is added to the co-flow of NH3 gas and H2S gas. The mixture of species are fed into the etch chamber to etch during the second cycle at least a portion of the silicon layers of the substrate without substantial etching of the passivated silicongermanium layer. Exact parameters for the second cycle of etching may be adjusted but remain within the prescribed ranges disclosed hereinabove as needed to ensure optimal selective Si etching over SiGe. The passivation and etching can be continued in this cyclic manner until a predetermined amount of the silicon layers have been selectively etched over SiGe to create the resultant structure and profile of Figure lb.

[0062] The ratio of flow rate of H2S to NH3 is within the predetermined range of the present invention as disclosed hereinabove. Failure to maintain the ratio of flow rate of H2S to NH3 within the predetermined range can negatively impact the effectiveness of the passivation layer, whereby a flow rate ratio of H2S to NH3 below the lower limit can adversely result in unacceptably low etch rates of Si layers and a flow rate ratio of H2S to NH3 above the upper limit can adversely result in a loss of Si selectivity as both Si and SiGe have a tendency to be etched by F2 / H2S.

[0063] Additionally, the ratio of flow rate of H2S to F2 is within the predetermined range of the present invention as disclosed hereinabove. Failure to maintain the ratio of flow rate of H2S to F2 within the predetermined range can negatively impact etch performance including creating lower Si etch rates, lower Si selectivity ratios, and a poorly defined etch profile (e.g., rounded edges, pinholes, and non-uniform removal of Si layers) of the resultant Si / SiGe structured film.

[0064] In another embodiment for the passivation step, as an alternative to solely co-flowing the NH3 gas with the H2S gas, sequential flow may occur whereby NH3 gas may be initially introduced into the etch chamber 106 followed by H2S with NH3, as validated in Example 5. After performing the passivation step, the F2 / Ar can be fed into the etch chamber 106, whereby the ratio of flow rate of H2S to NH3 and the ratio of flow rate of H2S to F2 remain within their predetermined ranges to ensure optimal passivation and etch performance, respectively.

[0065] The present invention contains numerous restrictions on the viable sequence of introducing the various gases into the etch chamber, as the inventors recognize and appreciate that the sequence of H2S in combination with the other gases during passivation and etch can, at least in part, impact overall process performance. In particular, the present invention excludes flowing H2S prior to flowing NH3 with H2S during passivation as the inventors observed that no etching occurs (as demonstrated in Comparative Example 5). Additionally, H2S cannot be exposed separately from NH3 and from F2 throughout the entire process, as the inventors have observed that such a process results in SiGe being etched with no Si etched (as demonstrated in Comparative Example 7). Furthermore, in some instances, depending on exact operating conditions, introducing H2S before NH3 may undesirably functionalize the passivation of the entire structure of the Si, thereby preventing subsequent etching of the Si with the etchant gas composition. Still further, the H2S, NH3 and F2 / Ar cannot be substantially co-flowed in a single step, as the inventors have discovered the entire Si / SiGe structured film can potentially degrade. Hence, after establishing flow of Ar purge gas, the Si / SiGe structured film must be initially passivated with a co-flow of H2S and NH3 or by sequentially feeding NH3 prior to feeding H2S as in Example 5, whereby the ratio of flow rate of H2S to NH3 is within the required range disclosed hereinabove, and further wherein, during the step of passivation, F2 / Ar is not flowed into the etch chamber 106.

[0066] The present invention can optimize the Si etch over SiGe of a wafer having Si layers that coexist and alternate with SiGe layers by operating under a combination of controlled conditions. In a preferred embodiment, a method for performing selective Si etching over SiGe may be performed by feeding Ar gas at a constant flow rate for a duration of 60 to 120 seconds to purge contaminants from the etch chamber. With flow of Ar preferably maintained and at the same flow rate, H2S and NH3 are co-flowed at a flow rate ratio of H2S to NH3 ranging from 0.1 to 0.5 for a duration of 1 to 20 seconds at a temperature of 60 to 120 deg C and a pressure of 1 to 10 Torr to effectively passivate the SiGe layers. Subsequently, while maintaining the same flow rates of H2S and NH3 and the Ar, a mixture of 20 vol% F2 / Ar is introduced into the etch chamber at a flow rate that establishes a flow rate ratio of H2S to F2 of 0.1 to 0.5 for a duration 1 to 20 seconds at thetemperature of 60 to 120 deg C and the pressure of 1 to 10 Torr. Si is selectively etched over SiGe at a selective ratio of 2200: 1 or higher. After the desired amount of Si material has been etched, the flows of H2S, NH3 and F2 / Ar are stopped, and Ar continues to flow at substantially the same flow rate as initially fed to the chamber to enable removal (i.e., purge) of the etch by-products generated during the Si selective etch step and any other impurities. The passivation, etching and purging steps may be repeated one or more times.

[0067] Because the process is temperature sensitive, maintaining the temperature within the prescribed temperature range enables optimal etch performance, whereby residue formation can be reduced, minimized or avoided, and a sharp edge profile (e.g., Fig. lb and Fig. 5b) can be obtained. While the residue deposited on the substrate surface can be heated to decompose and evaporate, the etch profile can be adversely impacted and residue has a tendency to remain deposited on the substrate surface. Hence, the present invention serves to eliminate any formation of the residue in the first instance, thereby avoiding the need to take steps to remove said residues, where such steps may risk damaging the Si / SiGe structure. The ability of the present invention to etch a Si layer that is devoid of residues and characterized by a sharp profile (i.e., not exhibiting rounded edges, square shaped profile and devoid of pinholes or recesses) provides a significant benefit for fabrication of various semiconductor devices.

[0068] Although the embodiments have been described with NH3 and H2S as the gases for passivation of the SiGe surface, it should be understood that other suitable types of nitrogen-containing gases and sulfur-containing gases can be utilized. Suitable nonlimiting examples of nitrogen-containing gases include N2H4, NF3, NHF2, NH2F, N2F4, (NH4)F, (NH4)HF2, CH3NH2, (CH3)2NH, (CH3)IN, (CnH2n+2)xNH3-x, where n>2 and x=l, 2, 3 and any combination thereof. Suitable non-limiting examples of sulfur-containing gases include H2S, H2S2, SO2, SO3, COS, CS2, SF4, SF6, SOF2, SO2F2, SOF4, (NH4)2S, (NH4)HS, (CH3)2S, CH3SH and any combination thereof. The nitrogen-containing gases preferably comprise a purity level of 98 vol% or higher. The sulfur-containing gases preferably comprise a purity level of 98 vol% or higher.

[0069] As an alternative to using pure H2S, a mixture of H2S with a suitable inert gas (e.g., Ar) may be utilized. In another embodiment, as an alternative to utilizing asulfur-containing gas, the additive gas may include a suitable hydride, including, without limitation PH3, GeH4 and / or SiH4.

[0070] Other types of fluorine-containing gas mixtures besides F2 / Ar may be used, including, without limitation, C1F, CIF3, CIFs, BrF, BrFs, BrFs, IF, IF3, IFs, IF7, OF2, XeF2, XeF4 and XeFe. Additionally, other suitable types of inert gases for purging contaminants and / or acting as a carrier gas besides Ar may be used, including, without limitation, He, Ne, Kr, Xe, and N2. By way of non-limiting example, the fluorine-containing gas may comprise various amounts of fluorine of 1%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99%.

[0071] While all of the embodiments disclose methods for selective etching of Si to SiGe, the principles of the present invention can also be applied to the selective etching of Si over various other semiconductor compounds including but not limited to SiN, SiO2, SiC, Si3N4 or Si alloyed with other elements besides Ge.

[0072] The following Examples demonstrate the benefits of introducing H2S as an additive gas during the selective Si etching over SiGe under controlled operating conditions and certain sequences. However, the following Examples are not to be construed as limiting the invention. All tests were performed utilizing a Linde etch test tool (i.e., etcher) created by the inventors and located at the Linde Technology Center in Tonawanda, NY.Example 1 (Invention performed with process recipe of Fig.2B; Calculation of etch rates and Si / SiGe selectivity etch ratio)

[0073] A dry etch process for selectively etching Si over SiGe was carried out as shown in Figure 3. The etcher contained an electrostatic chuck onto which an 8-inch diameter of stainless-steel substrate was mounted. Different materials of 2cm x 2cm sized test coupons were positioned onto the stainless-steel substrate for selective etching. The test coupons utilized were SiN, Poly-Si, SiGe, silicon oxide (SiO2) blankets and Si / SiGe structured film samples as representatively shown in Figure 1.

[0074] The etching was performed under conditions of 10 Torr of pressure in the etcher and a substrate temperature of 80 deg. C. In accordance with the process recipe of Figure 2b, a sequence of gases was fed into the etcher, flowed over the test coupons and thenexited from the etcher. The sequence of gases that flowed into the etcher is shown in Figure 2b. First, Ar by itself was introduced into the etcher at a flow rate of 250 sccm for 120 seconds to purge any contaminants from the internal volume of the etcher. The Ar continued to flow throughout the process at a substantially constant flow rate of 250 seem. Next, 2 sccm of H2S and 10 sccm of NH3 were introduced into the etcher for 10 seconds. The flow rates of NH3 and H2S were within the established predetermined flow rate ratios of the H2S gas to the NH3 gas required for the present invention. This gas composition (NH3 / H2S) functionalized the surface of the test coupons. Next, as the 2 sccm of H2S; 10 sccm of NH3; and 250 sccm of Ar continued to flow through the etcher, 84 sccm of 20 vol% F2 gas and 80 vol% Ar (designated as 20 vol% F2 / Ar) were introduced into the etcher to flow over the test coupons and then exit from the etcher. The flow rates of H2S and F2 were within the established predetermined flow rate ratio of H2S gas to F2 gas required for the present invention. The 20 vol% F2 / Ar in combination with NH3, H2S and Ar selectively etched the Poly-Si test coupon and the Si layers of the Si / SiGe structured samples. There was minimal etching of the SiGe layers of the Si / SiGe structured sample. The duration of the flow of this composition of gases was for 10 seconds. After 10 seconds of etching, the flow of H2S, NH3 and 20 vol% F2 / Ar was stopped. The Ar continued to flow at 250 sccm for 60 seconds after completion of the Si etch to purge the etch by-products created from the etching.

[0075] The final thicknesses of each of the test coupons was measured. The difference between the final thickness and the initial thicknesses for each of the SiN, SiO2, SiGe and Poly Si test coupons, and Si in the Si / Ge structured film test coupons divided by the 10 second etch duration was the calculated etch rate. Table 1 shows the calculated etch rates for the following test coupon blankets: SiN, SiCh, SiGe, Poly-Si, and Si in Si / SiGe structured film. Table 1 shows that the Ar / NH3 / H2S / F2 process of Fig. 2b removed material from the (i) SiN test coupon at an etch rate of 1.3 nm / min; (ii) SiO2 test coupon at an etch rate of 3.7 nm / min; (iii) SiGe test coupon at an etch rate of 0.1 nm / min; (iv) Poly-Si test coupon at an etch rate of 222.3 nm / min; and (v) Si film from the Si / SiGe structured film test coupon at an etch rate of 251.4 nm / min. To calculate the selectivity ratio of Si relative to SiGe, the etch rate of Poly-Si was divided by the etch rate of SiGe. The etch rate and selectivity results for Ar / NH3 / H2S / F2 as the process gases are shown in the second row ofTable 1. Given the short etching time of less than 60 seconds, the confidence in calculated etch rate for blanket coupons (e.g., SiN, SiO2, and SiGe) is fair because of the error associated with thickness measurements (ellipsometry), especially when the change in thickness is less than Inm. More distinct results (with high selectivity of poly Si / SiGe) should be attainable in the further study by employing a longer etch duration.Table 1Etch rate (ER) of SiN, SiO2, SiGe and poly Si films (blanket) and selectivity (Si ER / SiGe ER) data using Ar / NH3 / F2 (Fig. 2a) and Ar / NH3 / H2S / F2 (Fig. 2b) etch processes.Comparative Example 1 (Prior Art); Calculation of etch rates and Si / SiGe selectivity etch ratio

[0076] A conventional dry etch process for attempting to evaluate the effectiveness of selective etching Si over SiGe was carried out as substantially shown in Figure 3, but without the H2S additive gas source 103. The etcher contained an electrostatic chuck onto which an 8-inch diameter of stainless-steel substrate was mounted. Different materials of 2cm x 2cm sized test coupons were positioned onto the stainless-steel substrate for selective etching. The test coupons utilized were SiN, Poly-Si, SiGe, silicon oxide (SiO2) blankets and Si / SiGe structured film samples as representatively shown in Figure 1.

[0077] The etching was performed under conditions of 10 Torr of pressure in the etcher and a substrate temperature of 80 deg. C. In accordance with the process recipe of Figure 2a, a sequence of gases was fed into the etcher, flowed over the test coupons and then exited from the etcher. The sequence of gases that flowed into the etcher is shown in Figure 2a. First, Ar by itself was introduced into the etcher at a flow rate of 250 sccm for120 seconds to purge any contaminants from the internal volume of the etcher. The Ar continued to flow throughout the process at a substantially constant flow rate of 250 sccm. Next, 10 sccm of NH3 was introduced into the etcher for 10 seconds to passivate the SiGe test coupon; SiN test coupon; SiGe in the Si / SiGe structured sample and the SiCh blankets. As the 10 sccm of NH3 and 250 sccm of Ar continued to flow through the etcher, 84 sccm of 20 vol% F2 gas and 80 vol% Ar (designated as 20 vol% F2 / Ar) was introduced into the etcher to flow over the test coupons and then exit from the etcher. The 20 vol% F2 / Ar in combination with NH3, and Ar etched the Poly-Si test coupon and the Si layers of the Si / SiGe structured samples. There was minimal etching of the Si layers of the Si / SiGe structured sample. The duration of the flow of this composition of gases was for 40 seconds. After 40 seconds of etching, the flow of NH3 and 20 vol% F2 / Ar was stopped. The Ar continued to flow at 250 sccm for 60 seconds after completion of the Si etch to purge the etch by-products created from the etching.

[0078] The final thicknesses of each of the test coupons was measured. The difference between the final thickness and the initial thicknesses for each of the SiN, SiO2, SiGe, Poly Si and Si in the Si / SiGe test coupons divided by the 40 seconds etch duration was the calculated etch rate. The etch rate of Si was significantly less compared to that of Example 1. Specifically, Table 1 shows that the Ar / NH3 / F2 process of Fig. 2a removed material from the (i) SiN test coupon at an etch rate of 0.2 nm / min; (ii) SiO2 test coupon at an etch rate of 0.4 nm / min; (iii) SiGe test coupon at an etch rate of 3.7 nm / min; (iv) Poly-Si test coupon at an etch rate of 117 nm / min; and (v) Si from the Si / SiGe structured film test coupon at an etch rate of 12 nm / min. To calculate the selectivity ratio of Si relative to SiGe, the etch rate of Poly-Si was divided by the etch rate of SiGe. The etch rate and selectivity results for Ar / NH3 / F2 as the process gas are shown in the first row of Table 1.

[0079] The results indicate the selectivity for Poly Si / SiGe was significantly less than that of Example 1. Example 1 exhibited a selectivity ratio for Poly Si / SiGe greater than 70x that achieved by the Ar / NH3 / F2 process of Fig. 2a. Even though the Ar / NH3 / F2 process of Fig. 2a had an etch duration that was 4 times the duration of Example 1, less material of the Poly Si was etched. The results validate that the addition of H2S gas inExample 1 increased the Si etch rate and increased the etch selectivity ratio for Poly Si / SiGe compared to Comparative Example 1, which did not utilize H2S as an additive gas.Example 2 (Invention) Post-Etch Analysis

[0080] During the dry etch process of Example 1, a portion of the etch by-products created from the etching was fed into in-situ mass spectrometer shown in Figure 3. Similarly, during the conventional etch processes of Comparative Example 1, a portion of the etch by-products created from the etching was fed into in-situ mass spectrometer connected to the etch chamber. Figures 4a-4d show the results of the gaseous fragments collected from in-situ mass spectrometer in the etch processes of Comparative Example 1 and Example 1. Higher partial pressures corresponded to a higher number of gaseous fragments and vice versa. A comparison of the partial pressures of the various gaseous fragments of NH3+ (Fig. 4a); NHF2+(Fig. 4b); NF3+(Fig. 4c); and SiFa+(Fig. 4d) validate that the etch process of Example 1 (Ar / NH3 / H2S / F2) produced significantly less ammonia-based by-products (ammonium bifluoride (NHF2+) and ammonium tetrafluoride (NF3+) than that of Comparative Example 1 (Ar / NH3 / F2). Figure 4(d) shows a higher SiFs+peak in the Ar / NH3 / H2S / F2 process of Example 1 than the Ar / NH3 / F2 process of Comparative Example 1. Without being bound by any theory, the greater amount of SiFs+gaseous fragments in Example 1 are indicative of more F2 molecules that were available to react with the Si to form SiFs+due to less F2 converted into AFS residue on the Si surface when H2S was added as part of the etch process in Example 1. On the contrary, the lower partial pressure of SiF3+in Comparative Example 1 is indicative of more AFS residue formed on the Si surface, which made available less F2 molecules to react with the Si to form SiF3+gaseous fragments.

[0081] The conventional etch process of Ar / NH3 / F2 of Comparative Example 1 generated a higher amount of NH3+ gaseous fragments (Fig. 4a); higher amount of NHF2+ gaseous fragments (Fig. 4b); and a higher amount of NF3+ gaseous fragments (Fig. 4c) than the etch process of Ar / NH3 / H2S / F2 of Example 1. Without being bound by any theory, the higher NH3+partial pressure is indicative of faster diffusion of the NH3 gases in the etch chamber when the conventional process of Comparative Example 1 was carried out. As a result, Comparative Example 1 produced insufficient NH3 passivation onto the Si / SiGesurface, as evident by Figs. 4b and 4c showing higher partial pressures of NHF2+and NF3 gaseous fragments, respectively, generated from gaseous reactions between NH3 and F2. On the contrary, significantly less NHF2+and NF3 gaseous fragments were detected in Example 1. These mass spectrometer results of Figs. 4a-4d validate that the present invention as carried out in Example 2 significantly minimized formation of AFS residues during the selective etching of Si over SiGe.

[0082] The fifth test coupons of Si / SiGe structured film from Comparative Example 1 and Example 1 were visually observed with the results summarized in Figs. 5a and 5b, respectively. Fig. 5a illustrates a representative image of the Si / SiGe structured film that was produced by the conventional etch process of Fig. 2a; and Fig. 5b illustrates a representative image of the Si / SiGe structured film produced by the etch process of Fig. 2b in accordance with the principles of the present invention. The edges of Fig. 5a showed only a small amount of an average of 8 nm of Si was laterally etched away. Additionally, Fig. 5a shows the edge profile was round shaped due to unetched Si film at the Si / SiGe interface. Additionally, the edge profile was recessed and contained pin holes. This etch profile represented detrimental features that can cause severe defects such as irregular epitaxial growth of source / drain materials and electrical shorts between gate and source / drain, thereby potentially rendering poor performance in the resultant GAAFET device. On the contrary, remarkable etch profile improvement was observed in Fig. 5b with greater than an average of 40 nm of Si laterally etched away. The edge geometry was optimized as substantially uniform and square shaped. Additionally, there was a substantial absence of pinholes or recesses. The etch depth (i.e., the amount of material removed along the depth of the silicon layer) as indicated by the numbers in Fig. 5b was substantially uniform from the edge at the Si / SiGe interface to the middle of the Si layer for each of the Si layers that was laterally etched, thereby indicating that a substantially uniform amount of Si material was laterally etched at each of the Si layers of the Si / SiGe pattern. In summary, the etch profile of Fig. 5b was superior to that of Fig. 5a.Comparative Example 2 (Ar / F2 / H2S etch performance)

[0083] A dry etch process for attempting to evaluate the effectiveness of selective etching Si over SiGe was carried out as substantially shown in Figure 3 and described inComparative Example 1, but without the NH3 gas source. Si blanket, SiGe blanket, and Si / SiGe structured film test coupons were used in this test. The etching was performed under conditions of 10 Torr of pressure in the etcher and a substrate temperature of 80 deg C. First, Ar by itself was introduced into the etcher at a flow rate of 250 sccm for 120 seconds to purge any contaminants from the internal volume of the etcher. The Ar continued to flow throughout the process at a substantially constant flow rate of 250 seem. Next, 2 seem of H2S was introduced into the etcher for 10 seconds. Next, as the 2 seem of H2S and 250 seem of Ar continued to flow through the etcher, 84 seem of 20 vol% F2 gas and 80 vol% Ar (designated as 20 vol% F2 / Ar) were introduced into the etcher to flow over the test coupons and then exit from the etcher for a duration of 40 seconds. The SiGe film (thickness of 37 nm) in the blanket coupon and the structured test coupon were completely etched away when using the Ar / F2 / H2S mixture, confirming that H2S did not function as a passivation gas for the SiGe film. Almost no etching of the Si film occurred under this condition. The results indicated that NH3 passivation was essential for selectively etching Si / SiGe film using H2S / F2. H2S was not able to serve as a replacement for NH3 given that all of the SiGe film was etched away. Rather, H2S is required to be an addition to Ar / NH3 / F2.Comparative Example 3 (Continuous etch, Figs. 6a, 7a and 7b)

[0084] A dry etch process using a continuous etch process as shown by the recipe of Fig. 6a was carried out utilizing H2S as an additive gas in accordance with the principles of the present invention. All operating parameters for pressure, temperature, and gas flow rates for passivation and etch were the same as previously used in Example 1. The Ar continued to flow at 250 sccm for 60 seconds after completion of the Si etch to purge the etch by-products created from the etching. The Si / SiGe structured film test coupons were used in this continuous etching process.

[0085] Fig. 7a showed an etch amount (EA) of 41.9 nm of the Si film in the Si / SiGe structured film when a relatively short etch time 10 seconds was used in step 3 of the process recipe of Fig. 6a, which indicated adequate Si selectivity over SiGe, but an unacceptably low etch rate of the Si. Figure 7b showed a Si / SiGe structural loss (etched away) when a relatively longer etch time of 20 seconds was used in step 3 of the processrecipe of Fig. 6A. Both results demonstrated that a continuous etch process was unacceptable.Example 3 (Invention with cyclic etching and purge: Figs. 6b, 6c, 7c and 7d)

[0086] A dry etch process using a cyclic etch process as shown by the recipe of Fig. 6b and Fig. 6c was carried out utilizing H2S as an additive gas in accordance with the principles of the present invention. The same operating parameters for pressure, temperature, and gas flow rates for passivation and etch of Example 1 were utilized. Fig.6b utilized 2 cycles. Fig. 6c utilized 5 cycles. Si / SiGe structured film test coupons were used in this cyclic etching process.

[0087] In Fig. 6b, steps 1-3 were kept the same as in Fig. 6a of Comparative Example 3, except for including an additional step 4 for Ar purging performed to remove etch by-products in between the cycles. Two cycles were run from steps 2 to 4. Fig. 7c showed the etch result obtained using the 2-cycle method depicted in Fig. 6b. By employing the 2-cycle etching process, the Si / SiGe structural loss could be effectively controlled, thereby resulting in an intense EA of 71.7 nm of the Si film.

[0088] To judge the impact of the cyclic etching process, the exposure time in step 3 was shortened from 10 seconds to 4 seconds, and 5 cycles were performed from steps 2 to 4, as shown in Fig. 6c. This resulted in a more intense EA of 89 nm compared to the 2-cycle etching of Fig. 7c. When comparing scenarios of Fig. 7b, 7c, and 7d, where the aggregate exposure time in step 3 remained constant (20 seconds), the cyclic approach significantly improved the EA of Si films for short exposure times with many etch cycles, whereby the etch process of Fig. 6c produced the best results as shown in Fig. 7d. Fig. 7d showed the highest Si etch rate (deepest Si EA) while maintaining high Si selectivity over SiGe with a uniform etch profile. This test revealed that passivation time and etching time must be optimally allocated with a higher relative number of passivation and etch cycles with each etch cycle conducted at a relatively lower etch time to achieve a deeper EA with Si selectivity, suitable for GAAFET applications.Comparative Example 4 (Cyclic etching without purging in between the cycles; Figs.8b and 9b)

[0089] A dry etch process using a 5-cycle etch process without purging in between the cycles as shown by the recipe of Fig. 8b was carried out utilizing the H2S as an additive gas. Si / SiGe structured film test coupons were used in this cyclic etching process. The same operating parameters for pressure, temperature, and gas flow rates for passivation and etch of Example 1 were utilized.

[0090] The passivation step is shown at step 2, which utilized a gas composition of Ar, NH3 and H2S for 10 seconds. Next, the etch step occurred as shown at step 3, which utilized Ar, NH3, H2S and F2 for 4 seconds. The passivation followed by the etch was repeated 4 times for a total of 5 cycles. In step 4, the Ar was flowed at 250 seem for 60 seconds after completion of the cyclic etching to purge the etch by-products created from the etching. However, this one-time purge step was not enough to remove all etch byproducts accumulated during the test because no Ar was purged in between the etch cycles.

[0091] The corresponding etch characteristics of the Si / SiGe structured film test coupon created by the process recipe of Fig. 8b is shown in Fig. 9b. The EA of Si film was 35.4 nm, which was significantly less than that achieved by Example 4, discussed below.

[0092] The inventors collected gaseous fragments from the test in a mass-spectrometer. The partial pressure versus time profile indicated a lack of distinct F2+peaks for each of the 5 cycles.Example 4 (Invention with cyclic etching and purging in between the cycles; Figs. 8a and 9a)

[0093] A dry etch process using a 5-cycle etch process with purging as shown by the recipe of Fig. 8a was carried out utilizing H2S as an additive gas in accordance with the principles of the present invention. Si / SiGe structured film test coupons were used in this cyclic etching process. The same operating parameters for pressure, temperature, and gas flow rates for passivation and etch of Example 1 were utilized.

[0094] The passivation step is shown at step 2, which utilized a gas composition of Ar, NH3 and H2S for 10 seconds. Next, the etch step occurred as shown at step 3, which utilized Ar, NH3, H2S and F2 for 4 seconds. A purge with Ar gas at step 4 was performedfor 20 seconds in each cycle. Steps 2, 3 and 4 were repeated 4 times for a total of 5 cycles. Etch by-products did not accumulate during the test as a result of the Ar purge gas removing the etch by-products from the etch chamber after completion of each etch step 3. In step 5, the Ar continued to flow at 250 seem for 60 seconds after completion of the cyclic etching process to purge the etch by-products created from the etching.

[0095] The corresponding etch characteristics of the Si / SiGe structured film test coupon created by the process recipe of Fig. 8a is shown in Fig. 9a. The EA of Si film was 86.6 nm, which was significantly more than that achieved by Comparative Example 4.

[0096] The inventors collected gaseous fragments from the test in an in-situ mass-spectrometer. The partial pressure versus time profile indicated stable and distinct F2+peaks for each of the 5 cycles. The absence of these peaks in Comparative Example 4 without purging may have contributed to slower migration of the F2+fragments, which yielded less EA of Si film.Comparative Example 5 (H2S leading; Figs. 10a and 11a)

[0097] A dry etch process was carried out utilizing the process recipe of Fig. 10a whereby H2S was separately fed into an etch chamber before the passivation step of coflowing NH3 with H2S. Si / SiGe structured film test coupons were used. Ar was fed for 120 seconds at 250 seem, followed by Ar and H2S (250 seem Ar / 2 seem H2S) for 10 seconds, Ar, NH3 and FES (250 seem Ar / lOsccm NH3 / 2 seem EES) for 10 seconds and lastly Ar, NH3, EES and F2 (250 seem Ar / 10 seem NEE / 2 seem EES / 84 seem F2) for 10 seconds. In step 5, the Ar continued to flow at 250 seem for 60 seconds after completion of the etching process to purge the etch by-products created from the etching.

[0098] The corresponding etch characteristics of the Si / SiGe structured film test coupons created by the process recipe of Fig. 10a is shown in Fig. Ila. No etching of the Si film or the SiGe films was observed as shown in Fig. Ila.

[0099] The inventors collected gaseous fragments from the test in an in-situ mass-spectrometer. The partial pressure versus time profile of the various species showed a relatively lower partial pressure of NH3+gaseous fragments, which suggests that the preexposure of H2S may have functionalized the Si / SiGe surface to prevent incoming NH3+from adhering to the Si / SiGe surface. Instead, the NH3 in step 3 of Fig. 10a reacted withH2S to generate other solid or gaseous etch resistant by-products not detected by the in-situ mass spectrometer. A relatively higher partial pressure of SFX+(x = 3, 4, 5) was detected, which verified that the F2 and H2S reacted to generate a large amount of SFX+(x = 3, 4, 5), where formation of SF4 or SFe is believed to have slowed down or prevented the etching of the Si film.Example 5 (NH3 leading; Figs. 10b and lib)

[0100] A dry etch process was carried out utilizing the process recipe of Fig. 10b whereby NH3 was separately fed into an etch chamber before the passivation step of coflowing NH3 with H2S. Si / SiGe structured film test coupons were used. Ar was fed for 120 seconds, followed by flowing 250 seem of Ar and 10 seem of NH3 (Ar / NH3) for 10 seconds, and then Ar, NH3 and H2S (250 seem Ar / lOsccm NH3 / 2 seem H2S) for 10 seconds and lastly Ar, NH3, H2S and F2 (250 seem Ar / 10 seem NH3 / 2 seem H2S / 84 seem F2) for 10 seconds. In step 5, the Ar continued to flow at 250 seem for 60 seconds after completion of the etching process to purge the etch by-products created from the etching.

[0101] The corresponding etch characteristics of the Si / SiGe structured film test coupons created by the process recipe of Fig. 10b is shown in Fig. 1 lb. The EA of Si film was 20.8 nm.

[0102] The inventors collected gaseous fragments from the test in an in-situ mass-spectrometer. The partial pressure versus time profile of the various species showed a relatively lower partial pressure for SFx; and higher partial pressures for NH3 and NHFX+fragments. Relatively high partial pressures for NHF+and NHF2+revealed a higher possibility of AFS residue formation as a result of higher NH3 exposure time in the etcher compared to the process recipe utilized for Figure 2b.Comparative Example 6 (Replacement of H2S with HF; Figs. 12 and 13)

[0103] A dry etch process was carried out in which H2S was replaced with HF as the additive gas. Si / SiGe structured film test coupons were used. The process recipe that was carried out is shown in Fig. 12. The process involved flowing Ar gas at 250 seem for 120 seconds, followed by flowing Ar, NH3 and HF (250 seem Ar / 10 seem NH3 / 4 seem HF) for 10 seconds and lastly Ar, NH3, HF and F2 (250 seem Ar / 10 seem NH3 / 4 seemHF / 84 seem F2) for 10 seconds. The Ar continued to flow at 250 seem for 60 seconds after completion of the etching process to purge the etch by-products created from the etching.

[0104] The corresponding etch characteristics of the Si / SiGe structured film test coupons created by the process recipe of Fig. 12 is shown in Fig. 13. The use of HF as the additive gas resulted in poor etch performance. Only the Si / SiGe interface was etched deeply at an EA of 60 nm. The middle layer of the Si film was etched slowly at an EA of less than 5 nm. This test demonstrates that H2S is not interchangeable with other known etch species, such as HF, and that H2S serves as an additive gas to Ar / NH3 / F2 that achieves superior etch performance.Comparative Example 7 (Feeding separately H2S, NH3 and F2; Figs. 14 and 15)

[0105] A dry etch process was carried out in which H2S, NH3 and F2 were fed individually into the etch chamber to evaluate etch performance. The process recipe is shown in Figure 14, where the process involved flowing Ar gas at 250 seem for 120 seconds, followed by Ar and H2S (250 seem Ar / 2 seem H2S) for 10 seconds, Ar and NH3 (250 seem Ar / 10 seem NH3) for 10 seconds and lastly Ar and F2 (250 seem Ar / 84 seem F2) for 10 seconds. Si / SiGe structured film test coupons were used. The Ar continued to flow at 250 seem for 60 seconds after completion of the etching process to purge the etch byproducts created from the etching.

[0106] The corresponding etch characteristics of the Si / SiGe structured film test coupons created by the process recipe of Fig. 14 is shown in Fig. 15. 10.4 nm of SiGe was undesirably etched, while no Si was etched. The test validated that all of the gases cannot be exposed separately.

[0107] While it has been shown and described what is considered to be certain embodiments of the invention, it will, of course, be understood that various modifications and changes in form or detail can readily be made without departing from the spirit and scope of the invention. It is, therefore, intended that this invention is not limited to the exact form and detail herein shown and described, nor to anything less than the whole of the invention herein disclosed and hereinafter claimed.

Claims

CLAIMS1. A method of using an additive composition to improve selectivity of dry etching a substrate, comprising the steps of:positioning a substrate in an etch chamber, said substrate comprising silicongermanium layers and silicon layers; andselectively etching the silicon layers over the silicon-germanium layers by(A) introducing a passivating composition of a nitrogen-containing gas and the additive composition comprising sulfur-containing gas into the etch chamber towards an exposed surface of the substrate to form a passivated silicongermanium layer of the substrate; followed by(B) introducing an etchant composition comprising a fluorine-containing gas, the nitrogen-containing gas and the additive composition comprising sulfur- containing gas into the etch chamber to etch at least a portion of the silicon layer of the substrate without substantial etching of the passivated silicon-germanium layer.

2. The method of claim 1, further comprising performing two or more cycles until a predetermined amount of the silicon layer is selectively etched by the etchant composition without substantial etching of the passivated silicon-germanium layer, whereby each cycle comprises repeating the step of passivating the silicon-germanium layers and selectively etching the silicon layers.

3. The method of claim 1, wherein each of the steps of passivating and etching occurs for greater than 0 to less than 600 seconds.

4. The method of claim 1, wherein a flow rate ratio of the sulfur-containing gas to the fluorine-containing gas is between about 1:1000 to 100:1.

5. The method of claim 1, wherein a flow rate ratio of the sulfur containing gas to the nitrogen-containing gas is between about 1:1000 to 100:1.

6. The method of claim 1, further comprising introducing the nitrogen-containing gas with the sulfur-containing gas or introducing the nitrogen gas before introducing the sulfur-containing gas.

7. The method of claim 1, further comprising the step of etching at least a portion of the silicon layer at an etch rate of at least about 1 nm / min to selectively etch silicon to silicon-germanium at a selectivity ratio of at least about 10:1.

8. The method of claim 1, further comprising introducing a purge gas to purge the etch chamber before performing the step of passivating, followed by maintaining the flow of the purge gas during the selective etch and thereafter to remove by-products generated from the selective etching of the silicon layer from an interior of the etch chamber.

9. The method of claim 1, further comprising operating the etch chamber in a range of 0.001 to 100 Torr and a temperature in a range of 60 to 150 deg C during the selective etching.

10. The method of claim 5, wherein each of the ammonia-containing gas, the hydrogen sulfide gas and the fluorine-containing gas is fed from corresponding separate sources.

11. The method of claim 1, wherein the nitrogen-containing gas is selected from the group consisting of NH3, N2H4, NF3, NHF2, NH2F, N2F4, (NH4)F, (NH4)HF2, CH3NH2, (CH3)2NH, (CH3)3N, (CnH2n+2)xNH3-x, where n>2 and x=l, 2, 3 and any combination thereof.

12. The method of claim 1, wherein the sulfur-containing gas is selected from the group consisting of H2S, H2S2, SO2, SO3, COS, CS2, SF4, SF6, SOF2, SO2F2, SOF4, (NH4)2S, (NH4)HS, (CH3)2S, CH3SH and any combination thereof.

13. A method of using an etchant composition to improve selectivity of dry etching a substrate, comprising the steps of:providing a substrate in an etch chamber, said substrate comprising silicongermanium layers and silicon layers; andselectively etching silicon layers with the use of the etchant composition comprising an additive of a sulfur-containing gas while using a passivating gas composition to create and maintain a passivating layer on the silicon-germanium layer.

14. The method of claim 13, wherein the passivating layer is formed by flowing a passivating gas composition comprising nitrogen-containing and the additive of sulfur-containing gas into the etch chamber towards an exposed surface of the silicon-germanium layer.

15. The method of claim 13, wherein the substrate is a gate all-around (GAA) device architecture.

16. The method of claim 13, further comprising performing the step of selectively etching the silicon layer over the silicon germanium layer at a selectivity ratio of 10: 1 or more.

17. The method of claim 13, further comprising purging etch by-products from an interior of the etch chamber followed by repeating the steps of passivating and selectively etching the silicon layers.

18. The method of claim 13, wherein the etchant composition is formed prior to entry to the etchant chamber.

19. The method of claim 13, wherein the etchant composition is formed in the etchant chamber.

20. The method of claim 13, wherein the additive of the sulfur-containing gas has a purity of 98 vol% or higher.

21. The method of claim 13, wherein the additive of the sulfur-containing gas is premixed with an inert gas before entering the etchant chamber.

22. The method of claim 13, wherein the passivating gas composition is formed in the etchant chamber or prior to entry to the etch chamber by (i) co-flowing the additive of the sulfur-containing gas with the nitrogen-containing gas or (ii) sequentially feeding the nitrogen-containing gas followed by feeding the additive of the sulfur-containing gas with the nitrogen-containing gas.

23. A method of performing etching using an additive composition to improve selectivity of dry etching a substrate, comprising the steps of:positioning a substrate in an etch chamber, said substrate comprising silicongermanium layers and silicon layers;performing a first etch cycle, comprising:(A) either (i) co-flowing a nitrogen-containing gas and a sulfur-containing gas into the etch chamber towards an exposed surface of the substrate or (ii) sequentially flowing the nitrogen-containing gas and then flowing the sulfur-containing gas towards the exposed surface;(B) forming a passivated silicon-germanium layer on the substrate; followed by (C) introducing a fluorine-containing gas, the nitrogen-containing gas and the sulfur-containing gas into the etch chamber to etch at least a portion of the silicon layer of the substrate without substantial etching of the passivated silicon-germanium layer; and (D) introducing a purge gas into the etch chamber to remove etch by-products created from the etch; andinitiating an optional second cycle by re-establishing the passivated silicon-germanium layer followed by introducing the fluorine-containing gas, the nitrogencontaining gas and the sulfur-containing gas into the etch chamber to etch at least a portionof the silicon layer of the substrate without substantial etching of the passivated silicongermanium layer.

24. The method of claim 23, wherein a ratio of the flow rate of the sulfur-containing gas to the nitrogen-containing gas ranges from about 1: 10 to 1:1.

25. The method of claim 23, wherein a ratio of the flow rate of sulfur-containing gas to the fluorine-containing gas ranges from about 1:10 to 1:1.

Images