Metal hard mask for precise mandrel adjustment

Abstract

Metal hard mask for precise adjustment of mandrels. A method for forming mandrels for use in pitch doubling processes is provided, in which a metal hard mask is inserted between a mandrel material layer and a soft mask. Inserting the metal hard mask allows for easier pattern transfer to the mandrel material layer, avoiding many of the problems encountered during multiple patterning steps. Inserting the metal hard mask results in the formation of square mandrels with flat tops due to its resistance to etching and the ability to wet strip the metal hard mask. Adjusting the metal hard mask prior to pattern transfer to the underlying mandrel material layer can provide smaller or larger hard mask patterns than patterning without such adjustment. This method can also be used to protect downstream non-mandrel processes where selectivity is critical.

Description

[Background technology]

[0001] FIELD OF THE INVENTION The present invention relates to semiconductor technology, and more particularly to a method for forming mandrels used in forming semiconductor structures.

[0002] As the area available for semiconductor devices continues to scale down and shrink, engineers are faced with the challenge of how to meet market demands for ever-increasing device density. One technique for sub-80 nm pitch patterning is to achieve doubling of pattern density through a technique called sidewall image transfer (SIT), also known as sidewall spacer image transfer. In a conventional SIT process, mandrels are first formed on at least one material layer that needs to be patterned, followed by SIT spacer formation and then mandrel removal. The SIT spacers are used to pattern at least one material layer underneath the SIT spacers. After patterning the at least one material layer, the SIT spacers are removed, exposing the patterned structure made of the at least one material layer. In SIT and other pitch-doubled patterning processes, a soft mask is used in the formation of the mandrels. Mandrels formed using only a soft mask as a pattern mask are unreliable, making it difficult to form mandrels with a tight pitch. Summary of the Invention

[0003] A method for forming mandrels for use in pitch-doubling processes, such as self-aligned litho-etch-litho-etch (SALELE), is provided, in which a metal hard mask is inserted between the mandrel material layer and the soft mask. Inserting the metal hard mask allows for easier pattern transfer to the mandrel material layer, avoiding many of the problems encountered during multiple patterning steps. Inserting the metal hard mask results in the formation of square mandrels with flat tops due to its resistance to etching and the ability to wet-strip the metal hard mask. Conditioning the metal hard mask prior to pattern transfer to the underlying mandrel material layer can provide hard mask patterns that are smaller or larger than patterns without such conditioning. This method can also be used to protect downstream non-mandrel processes where selectivity is critical.

[0004] In one aspect of the invention, a method of forming mandrels is provided. In one embodiment, the method includes forming a plurality of metal hard masks on top of a mandrel material layer overlying at least one material layer, and patterning the mandrel material layer using the plurality of metal hard masks as an etch mask to provide a plurality of mandrels having a pitch of less than 36 nm.

[0005] In some embodiments, at least one layer of material is present in a back-end of line (BEOL) region of a semiconductor structure, and thus the methods of the present invention can be used to pattern back-end structures, such as, for example, conductive structures (lines or vias) and / or dielectric material structures.

[0006] In some embodiments, at least one layer of material (in the BEOL region or outside the BEOL region) is a dielectric material, a conductive material, or any combination thereof.

[0007] In some embodiments, the mandrel pitch is between 26 nm and 35 nm, resulting in closely spaced mandrels.

[0008] In some embodiments, each of the metal hard masks has a first lateral width, and the method further comprises adjusting the first lateral width to a second lateral width different from the first lateral width. In one example, the first lateral width is greater than the second lateral width, and the adjusting includes reducing the first lateral width using an etching process that removes portions of each of the metal hard masks. In another example, the first lateral width is less than the second lateral width, and the adjusting includes forming metal spacers along sidewalls of each of the metal hard masks.

[0009] In some embodiments, one or more mandrel base masking layers are formed between the mandrel material layer and the at least one material layer, and the one or more mandrel masking layers are used to pattern the at least one material layer and provide a smooth landing surface for the mandrel material layer.

[0010] In some embodiments, forming the plurality of metal hard masks includes forming a mandrel-patterning material stack including a metal hard mask layer and a soft mask material layer on the mandrel material layer, forming a patterned photoresist having a mandrel pattern on top of the mandrel-patterning material stack, and transferring the mandrel pattern into the soft mask material layer and the metal hard mask material layer of the mandrel-patterning material stack by etching to provide the plurality of soft masks and the plurality of metal hard masks, respectively. In embodiments, each soft mask is removed before patterning the mandrel material layer.

[0011] In some embodiments, the mandrel patterning material stack further includes a dielectric oxide layer disposed between the metal hard mask material layer and the soft mask material layer, the dielectric oxide layer transforming into a plurality of dielectric oxide masks upon mandrel pattern transfer, the presence of which can improve defects during lithographic rework.

[0012] In some embodiments, the metal hardmask layer is comprised of titanium nitride, the dielectric oxide layer is comprised of silicon dioxide, the softmask material layer is comprised of an organic planarization layer material, and the mandrel material layer is comprised of silicon nitride.

[0013] In an embodiment, the method further comprises forming mandrel pitch-doubling spacers along the sidewalls of each of the mandrels and each of the metal hard masks.

[0014] In another embodiment, a method includes forming a plurality of metal hard masks on top of a mandrel material layer overlying at least one material layer, the material layer being a dielectric material, a conductive material, or any combination thereof, present in a back-end of line (BEOL) region of a semiconductor structure, and patterning the mandrel material layer using the metal hard mask as an etch mask to provide a plurality of mandrels having a pitch between 26 nm and 64 nm.

[0015] In yet another embodiment, a method includes forming a plurality of metal hard masks having a first lateral width on top of a mandrel material layer overlying at least one material layer, reducing the first lateral width of each of the metal hard masks to a second lateral width by etching, and patterning the mandrel material layer using the metal hard mask having the second lateral width as an etch mask to provide a plurality of mandrels.

[0016] In yet a further embodiment, a method includes forming a plurality of metal hard masks having a first lateral width on top of a mandrel material layer overlying at least one material layer, increasing the first lateral width of each of the metal hard masks to a second lateral width by forming metal spacers along sidewalls of each of the metal hard masks, and patterning the mandrel material layer using the metal hard mask having the second lateral width as an etch mask to provide a plurality of mandrels. [Brief explanation of the drawings]

[0017] [Figure 1A] 1A-1D are cross-sectional views illustrating basic processing steps that may be used in one embodiment of the present invention. [Figure 1B] 1A-1D are cross-sectional views illustrating basic processing steps that may be used in one embodiment of the present invention. [Figure 1C] 1A-1D are cross-sectional views illustrating basic processing steps that may be used in one embodiment of the present invention. [Figure 1D] 1A-1D are cross-sectional views illustrating basic processing steps that may be used in one embodiment of the present invention.

[0018] [Figure 2A] 1A-1C are cross-sectional views illustrating basic processing steps that may be used in another embodiment of the present invention. [Figure 2B] 1A-1C are cross-sectional views illustrating basic processing steps that may be used in another embodiment of the present invention. [Figure 2C] 1A-1C are cross-sectional views illustrating basic processing steps that may be used in another embodiment of the present invention.

[0019] [Figure 3A] 10A-10C are cross-sectional views illustrating basic processing steps that may be used in yet another embodiment of the present invention. [Figure 3B] 10A-10C are cross-sectional views illustrating basic processing steps that may be used in yet another embodiment of the present invention. [Figure 3C]10A-10C are cross-sectional views illustrating basic processing steps that may be used in yet another embodiment of the present invention. [Figure 3D] 10A-10C are cross-sectional views illustrating basic processing steps that may be used in yet another embodiment of the present invention.

[0020] [Figure 4A] 1 is a cross-sectional view illustrating one advantage that can be obtained using the method of the present invention. [Figure 4B] 1 is a cross-sectional view illustrating one advantage that can be obtained using the method of the present invention. [Figure 4C] 1 is a cross-sectional view illustrating one advantage that can be obtained using the method of the present invention. [Figure 4D] 1 is a cross-sectional view illustrating one advantage that can be obtained using the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0021] The present invention will now be described in more detail by reference to the following discussion and the drawings that accompany this application. It should be noted that the drawings herein are provided for illustrative purposes only, and therefore the drawings are not drawn to scale. It should also be noted that like and corresponding elements are referred to by like reference numerals.

[0022] In the following description, numerous specific details are set forth, such as particular structures, components, materials, dimensions, processing steps, and techniques, to provide an understanding of various embodiments of the present invention. However, it will be understood by those skilled in the art that various embodiments of the present invention may be practiced without these specific details. In other instances, well-known structures or processing steps have not been described in detail to avoid obscuring the present invention.

[0023] When an element, such as a layer, region, or substrate, is referred to as being "on" or "over" another element, it will be understood that the element can be directly on the other element, or that intervening elements may be present. In contrast, when an element is referred to as being "directly on" or "directly over" another element, there are no intervening elements present. When an element is referred to as being "beneath" or "under" another element, it will be understood that the element can be directly below or directly underneath the other element, or that intervening elements may be present. In contrast, when an element is referred to as being "directly beneath" or "directly under" another element, there are no intervening elements present.

[0024] At aggressive pitches, conventional soft masks (i.e., organic planarization layers (OPLs)) used for multi-patterned mandrel formation can suffer from aspect ratio-induced flop-over, limiting mask height. Furthermore, the etch selectivity of soft mask materials is poor, necessitating alternative masking materials that are easily removed but better retained. Furthermore, the tight process window for pattern transfer becomes more challenging at smaller pitches, making it difficult to obtain the desired critical dimension (CD) within such a window. A flexible method for adjusting the CD without introducing defects is difficult to achieve with standard dry etching.

[0025] In some embodiments, the present invention avoids the above problems by inserting a metal hard mask into a pitch-doubling process, such as SALELE, for the formation of mandrels with a sub-36 nm pitch. The insertion of a metal hard mask stack allows for easier pattern transfer and avoids many of the problems encountered during multiple patterning steps. Due to its resistance to etching and the ability to strip the metal hard mask, square mandrels with a flat top layer can be formed. Furthermore, the present invention allows for precise control of the size of the mandrels formed by adjusting the metal hard mask pattern to be smaller or larger than the original metal hard mask pattern without such adjustment. The method of the present invention can also be used to protect downstream non-mandrel processes where selectivity is critical. These and other aspects and advantages of the method of the present invention are now described in more detail.

[0026] Referring first to FIGS. 1A-1D, basic processing steps that can be used in one embodiment of the present invention are illustrated. In particular, FIGS. 1A-1D illustrate a method according to the present invention for forming mandrels having a base pattern. The term "base pattern" refers to a pattern formed without adjusting the size (i.e., lateral width) of the metal hard mask used. In particular, as shown in FIG. 1A, the method begins with forming a structure including at least one material layer 10L that needs to be patterned. The structure may also include one or more optional mandrel-based masking layers. In FIG. 1A, a first mandrel-based masking layer 12L and a second mandrel-based masking layer 14L are shown by way of example. The structure further includes a mandrel material layer 16L that overlies the at least one material layer 10L; as shown in the embodiment shown in FIG. 1A, the optional mandrel-based masking layers 12L, 14L are disposed between the mandrel material layer 16L and the at least one material layer 10L. The structure still further includes a mandrel patterning material stack disposed on the mandrel material layer 16L. The mandrel patterning material stack includes a metal hard mask layer 18L, an optional dielectric oxide layer 20L, and a soft mask material layer 22L. The structure still further includes a patterned photoresist 24 disposed on top of the mandrel patterning material stack.

[0027] At least one material layer 10L that may be utilized in the present invention typically, although not always, resides in the back-end of the line (BEOL) region of a semiconductor structure. As known to those skilled in the art, the back-end of the line (BEOL) region is the area of a semiconductor structure where wiring structures are formed. Such wiring structures provide interconnections to individual devices present in the front-end of the line (FEOL). One metal layer 10L that may be utilized in the present invention may be composed of a dielectric material, a conductive material, or any combination thereof, including a multi-layer stack including a dielectric material and a conductive material.

[0028] Dielectric materials that can be used as at least one material layer 10L can include any material with electrically insulating properties, including, but not limited to, silicon dioxide, silicon nitride, undoped silicate glass (USG), fluorosilicate glass (FSG), borophosphosilicate glass (BPSG), spin-on low-k dielectric materials, chemical vapor deposition (CVD) low-k dielectric materials, or any combination thereof. The term "low-k" as used throughout this application refers to a dielectric material having a dielectric constant less than 4.0. All dielectric constants referred to herein are measured in a vacuum unless otherwise specified. Exemplary low-k dielectric materials that can be used include, but are not limited to, silsesquioxane, C-doped oxides (i.e., organosilicates) containing atoms of Si, C, O, and H, thermosetting polyarylene ethers, or multilayers thereof. The term "polyarylene" is used herein to refer to aryl or inertly substituted aryl moieties linked together by bonds, fused rings, or inert linking groups (e.g., oxygen, sulfur, sulfone, sulfoxide, carbonyl, and the like). When a dielectric material is utilized for at least one material layer 10L, the dielectric material can be formed using a deposition process including, for example, chemical vapor deposition (CVD), plasma-enhanced chemical vapor deposition (PECVD), evaporation, or spin-on coating.

[0029] Conductive materials that can be used as the at least one material layer 10L include conductive metals, conductive metal alloys, or any combination thereof, including multilayer stacks. Examples of conductive metals that can be used to provide the at least one material layer 10L include, but are not limited to, copper (Cu), aluminum (Al), tungsten (W), ruthenium (Ru), cobalt (Co), or iridium (Ir). Examples of conductive metal alloys that can be used to provide the at least one material layer 10L include, but are not limited to, Cu-Al alloys, Co-Ru alloys, or W-Co alloys. When a conductive material is used as the at least one material layer 10L, the conductive material can be formed using a deposition process such as CVD, PECVD, physical vapor deposition (PVD), sputtering, chemical solution deposition, or plating. In one example, the at least one material layer 10L includes a layer of Ru formed on one of the dielectric materials, such as silicon dioxide, as described above.

[0030] The at least one material layer 10L may have a variety of thicknesses depending on the material used to provide the at least one material layer 10L and the method used to form it. In one example, when the at least one material layer 10L is composed of a layer of Ru, the at least one material layer 10L may have a thickness of 10 nm to 100 nm.

[0031] The mandrel-based masking layer(s) (i.e., masking layers disposed below the mandrel material layer 16L) that may be used in the present invention may include, but are not limited to, a dielectric hard mask material, a metallic hard mask material, and / or a landing material layer. Illustrative examples of dielectric hard mask materials include, but are not limited to, silicon nitride or silicon oxynitride. Illustrative examples of metallic hard mask materials include TiN, TiO x, Ru, or W. Illustrative examples of landing material layers include, but are not limited to, amorphous silicon (a-Si), silicon oxide, TiN, nitrides, or conductive metals. In one example, a first mandrel-based masking layer 12L and a second mandrel-based masking layer 14L are utilized. In such an embodiment, the first mandrel-based masking layer 12L is composed of a compositionally different material than the second mandrel-based masking layer 14L. In one example, the first mandrel-based masking layer 12L is composed of TiN, and the second mandrel-based masking layer 14L is composed of a-Si. Such a mandrel-based masking layer combination is particularly suitable for use in patterning at least one material layer 10A composed of a conductive metal or conductive metal alloy, as described above, to provide patterned conductive structures (e.g., Ru-containing structures) having a square, flat top layer.

[0032] The one or more mandrel-based masking layers may be formed using a deposition process, including, but not limited to, CVD, PECVD, PVD, spin-on coating, evaporation, sputtering, chemical solution deposition, plating, or atomic layer deposition (ALD). The one or more mandrel-based masking layers may have various thicknesses depending on the materials used to provide the one or more mandrel-based masking layers and the method used in forming them. In one example, the one or more mandrel-based masking layers may have a thickness of 5 nm to 40 nm. In one specific example, when the first mandrel-based masking layer 12L is composed of TiN and the second mandrel-based masking layer 14L is composed of a-Si, the first mandrel-based masking layer 12L may have a thickness of 5 nm to 20 nm, while the second mandrel-based masking layer 14L may have a thickness of 5 nm to 20 nm.

[0033] The mandrel material layer 16L is a sacrificial material typically composed of a dielectric material such as silicon nitride or silicon oxide. Other types of sacrificial materials, such as a-Si, aC, or organic layers, can be used as the mandrel material layer 16L. The mandrel material layer 16L can be formed using a deposition process, including, but not limited to, CVD, PECVD, PVD, or ALD. The mandrel material layer 16L can have a thickness of 20 nm to 80 nm, more typically 40 nm to 60 nm, although other thicknesses are contemplated and can be used as the thickness of the mandrel material layer 16L.

[0034] The metal hard mask layer 18L of the mandrel patterning material stack is TiN, TiO x The metal hard mask material 18L may include, but is not limited to, TiN, Ru, or W. In one example, TiN is used as the metal hard mask material 18L to pattern the mandrel material layer 16L, which is composed of silicon nitride. The metal hard mask layer 18L typically forms a direct interface with the mandrel material layer 16L. The metal hard mask layer 18L may be formed using a deposition process, including, but not limited to, CVD, PECVD, PVD, ALD, sputtering, or plating. The metal hard mask layer 18L may have a thickness of 10 nm to 25 nm, although other thicknesses are contemplated and may be used as the thickness of the metal hard mask layer 18L.

[0035] In some embodiments, the mandrel patterning material stack further includes a dielectric oxide layer 20L. In other embodiments, the dielectric oxide layer 20L is omitted. The dielectric oxide layer 20L is typically used to protect the metal hard mask layer 18L during rework. In embodiments, the dielectric oxide layer 20L forms an interface with the metal hard mask layer 18L. The dielectric oxide layer 20L is typically composed of silicon dioxide, although other dielectric oxide materials, such as nitrides, SiOCN, or SiCN, can be used to provide the dielectric oxide layer 20L. The dielectric oxide layer 20L can be formed using a deposition process, including, but not limited to, CVD, PECVD, ALD, or PVD. The dielectric oxide layer 20L can have a thickness of 10 nm to 25 nm, although other thicknesses are contemplated and can be used as the thickness of the dielectric oxide layer 20L.

[0036] The soft mask material layer 22L of the mandrel patterning material stack is composed of an organic material, such as an organic planarization layer (OPL) material. In embodiments, the soft mask material layer 22L forms an interface with the dielectric oxide layer 20L or the metal hard mask layer 18L if the dielectric oxide layer 20L is omitted. The soft mask material layer 22L may be formed using a deposition process, including, but not limited to, CVD, PECVD, ALD, spin-on coating, or evaporation. The soft mask material layer 22L may have a thickness of 40 nm to 60 nm, although other thicknesses are contemplated and may be used as the thickness of the soft mask material layer 22L.

[0037] The patterned photoresist 24 disposed on top of the mandrel patterning material stack is composed of any photoresist material, including, for example, a negative-tone photoresist material, a positive-tone photoresist material, or a hybrid photoresist material including a combination of negative-tone and positive-tone photoresist materials. The patterned photoresist 24 can be formed by lithography, which includes depositing at least a photoresist material on top of the soft mask material layer 22L, exposing the deposited photoresist material to a pattern of radiation (herein, the pattern of radiation forming the mandrel pattern), and developing the exposed photoresist material into the deposited photoresist material. In embodiments, one or more photolithographic masking layers, such as a bottom antireflective coating (BARC), can be formed between the soft mask material layer 22L and the deposited photoresist material. For clarity, the one or more photolithographic masking layers are not shown in the drawings of this application.

[0038] 1B, the structure of FIG. 1A is shown after transferring the pattern provided by the patterned photoresist 34 (i.e., the mandrel pattern) into the mandrel patterning material stack, i.e., soft mask material layer 22L, optional dielectric oxide layer 20L, and metal hard mask material layer 18L, to provide a plurality of mandrel masks, each consisting of a remaining portion of soft mask material layer 22L, a remaining location of optional dielectric oxide layer 20L, and a remaining portion of metal hard mask layer 18L. The remaining portion of soft mask material layer 22L may be referred to herein as a soft mask 22, the remaining portion of optional dielectric oxide layer 20L may be referred to herein as a dielectric oxide mask 20, and the remaining portion of metal hard mask material layer 18 may be referred to herein as a metal hard mask 18. Each mandrel mask has a mandrel pattern, a portion of which will be later used to pattern mandrel material layer 16L. Each mandrel mask typically has a lateral width, measured from one sidewall to the opposite sidewall, of between 6 nm and 16 nm.

[0039] Within each mandrel mask, the soft mask 22, optional dielectric oxide mask 20, and metal hard mask 18 typically have sidewalls that are vertically aligned with one another. Each mandrel mask typically has sidewalls that are oriented perpendicular to the horizontal top surface of the mandrel material layer 16L; however, some inward or outward tapering of the mandrel masks may occur. The number of mandrel masks formed can vary and is not limited to the exemplary embodiment shown in FIG. 1B in which two spaced-apart mandrel masks are formed.

[0040] The pattern transfer shown in FIG. 1B can be performed using one or more etching processes. The one or more etching processes stop on the top surface of the mandrel material layer 16L. The one or more etching processes include dry etching, wet chemical etching, or a combination thereof. Dry etching can include, but is not limited to, reactive ion etching (RIE), ion beam etching (IBE), or plasma etching. In one example, the one or more etching processes used to provide the structure shown in FIG. 1B include an RIE process that utilizes an ashing step (N2 / H2 or O2), an oxide etch (CF4), and a TiN etch (Cl2).

[0041] 1B, the patterned photoresist 24 is removed from the structure using a conventional resist removal process, such as a resist strip process. The patterned photoresist 24 may be removed any time after the pattern has been transferred into the soft mask material layer 22L.

[0042] 1C, the structure shown in FIG. 1B is shown after the soft mask 22 has been removed from each mandrel mask, leaving behind the metal hard mask 18 and, if present, the dielectric oxide mask 20. Removing the soft mask 22 from each mandrel mask may be performed by ashing or any other technique that selectively removes the soft mask 22 from the mandrel mask. This step is performed to reduce the aspect ratio and prevent flop-over of lines from less robust layers. In some embodiments, removing the soft mask 22 can be omitted, and the mandrel mask shown in FIG. 1B can be used to pattern the mandrel material layer 16L.

[0043] Referring now to FIG. 1D, the structure shown in FIG. 1C is shown after a pattern (i.e., a mandrel pattern) has been transferred to mandrel material layer 16L. Transferring the pattern to mandrel material layer 16L provides multiple mandrels 16 (two are shown in FIG. 1D by way of example) on at least one material layer 10L. Transferring the pattern to mandrel material layer 16L includes one or more etching processes (dry etching and / or chemical etching, as defined above) that stop on one or more mandrel-based masking layers (e.g., second mandrel-based masking layer 14L shown in FIG. 1D) or, if no mandrel-based masking layers are present, on at least one material layer 10L. After or before patterning mandrel material layer 16L into mandrels 16, dielectric oxide mask 20 can be removed using a material removal process that selectively removes the dielectric oxide that provides dielectric oxide mask 20.

[0044] The mandrels 16 formed using the metal hard mask 18 are square and have a flat top surface after removal of the metal hard mask 18. The mandrels 16 formed herein have sidewalls that are substantially perpendicular to the horizontal top surface of the at least one material layer 10L. The term "substantially perpendicular" is used herein to indicate that the sidewalls of the mandrels 16 are 90°±10% relative to the top horizontal surface of the at least one material layer 10L. In some embodiments, the pitch of the mandrels 16 is between 26 nm and 64 nm. In some embodiments, the pitch of the mandrels 16 is less than 36 nm, more typically between 26 nm and 35 nm. The pitch is measured from a point on one of the mandrels 16 to the same point on the nearest neighboring mandrel 16. In the present application, the pitch can be measured from the center of two laterally adjacent mandrels 16, as shown in FIG. 1D.

[0045] The mandrels 16 formed using the metal hard mask 18 do not suffer from inductive flop-over, which can limit the mask height. No movement or collapse of the metal hard mask 18 is observed. Furthermore, the metal hard mask 18 is easier to remove using a material removal process that selectively removes the metal that provides the metal hard mask 18, and the metal hard mask 18 is better retained during patterning of the mandrel material layer 16L.

[0046] After patterning the mandrel material layer 16L, conventional pitch-doubling processing steps are then performed, including forming mandrel pitch-doubled spacers (not shown) on the sidewalls of each of the mandrels 16, removing the mandrels 16 after the mandrel pitch-doubled spacer formation, and then patterning the underlying at least one material layer 10L using the mandrel pitch-doubled spacers as an etch mask. In embodiments of the present invention, the metal hard mask 18 can be removed before or after mandrel pitch-doubled spacer formation using a material removal process that selectively removes the metal hard mask material. In some embodiments, removal of the metal hard mask 18 occurs after mandrel pitch-doubled spacer formation to provide added protection during downstream pitch-doubling processing.

[0047] Reference is now made to Figures 2A-2C, which illustrate basic processing steps that may be used in another embodiment of the present invention. In this embodiment shown in Figures 2A-2C, the size, i.e., lateral width, of the metal hard mask 18 can be adjusted to provide a metal hard mask with a smaller size than that shown in Figures 1A-1D. The smaller size metal hard mask then provides mandrels 16 with a smaller size, i.e., lateral width. Figures 2A-2C include similar and corresponding elements to those shown in Figures 1A-1D, and those similar and corresponding elements are referenced using similar reference numerals.

[0048] Referring first to FIG. 2A , a structure is shown that includes at least one material layer 10L, one or more optional mandrel-based masking layers, including a first mandrel-based masking layer 12L and a second mandrel-based masking layer 14L, a mandrel material layer 16L, and a mandrel mask, each mandrel mask including a metal hard mask 18 and an optional dielectric oxide mask 20. Because the structure shown in FIG. 2A is identical to that shown in FIG. 1C , the processing steps previously described to provide the structure shown in FIG. 1C can now be used to provide the structure shown in FIG. 2A . At this point in the application, each metal hard mask 18 and each optional dielectric oxide mask 20 has a first size, i.e., a first lateral width. In one example, the first lateral width of each metal hard mask 18 and each optional dielectric oxide mask 20 is between 10 nm and 20 nm. In some cases, this first size needs to be adjusted to a smaller size to enable patterning of smaller-sized mandrels 16.

[0049] Referring now to FIG. 2B , the structure of FIG. 2A is shown after the size, i.e., first lateral width, of each metal hard mask 18 and each optional dielectric oxide mask 20 has been reduced. In FIG. 2B , the dotted lines represent the first size, i.e., first lateral width, of each metal hard mask 18 and each optional dielectric oxide mask 20. Each metal hard mask 18 remaining after this reduction process may be referred to as a reduced-size metal hard mask 18P, while each dielectric oxide mask 20 remaining after this reduction process may be referred to as a reduced-size dielectric oxide mask 20P. This reduction step may be performed using a wet etching process that removes a portion of each metal hard mask 18 and, if present, each dielectric oxide mask 20. In one example, the wet etching process includes diluted hydrofluoric acid (HF). In one embodiment, the diluted HF includes a ratio of water to HF of approximately 300:1. The wet etching process removes each metal hard mask 18 and, if present, each dielectric oxide mask 20 laterally inward from the original sidewalls to provide a reduced-size metal hard mask 18P and a reduced-size dielectric oxide mask 20P shown in Figure 2B. Each reduced-size metal hard mask 18P and each reduced-size dielectric oxide mask 20P has a second lateral width that is smaller than the aforementioned first lateral width.

[0050] Referring now to FIG. 2C, the structure of FIG. 2B is shown after transferring a mandrel mask pattern (i.e., a reduced-size mandrel pattern) including a reduced-size metal hard mask 18P and a reduced-size dielectric oxide mask 20P to provide mandrels 16. This pattern transfer step is identical to that shown above in FIG. 1D. Note that the structure shown in FIG. 2C is after each reduced-size dielectric oxide mask 20P has been removed. The mandrels 16 have a smaller size, i.e., lateral width, compared to the size, i.e., lateral width, of the mandrels 16 shown in FIG. 1D. The pitch of the mandrels 16 in this embodiment is within the ranges described above for previous embodiments of the invention.

[0051] After patterning the mandrel material layer 16L, conventional pitch-doubling processing steps are then performed, including forming mandrel pitch-doubling spacers (not shown) on the sidewalls of each of the mandrels 16, removing the mandrels 16 after the mandrel pitch-doubling spacers are formed, and then patterning the underlying at least one material layer 10L using the mandrel pitch-doubling spacers as an etch mask. In embodiments of the present invention, the reduced-size metal hard mask 18P can be removed before or after mandrel pitch-doubling spacer formation using a material removal process that selectively removes the metal hard mask material. In some embodiments, removal of the reduced-size metal hard mask 18P occurs after mandrel pitch-doubling spacer formation to provide additional protection during downstream pitch-doubling processing.

[0052] Referring now to Figures 3A-3D, basic processing steps that can be used in yet another embodiment of the present invention are shown. In this embodiment shown in Figures 3A-3D, the size, i.e., lateral width, of each metal hard mask 18 can be adjusted to provide a metal hard mask with a larger size than that shown in Figures 1A-1D. The larger size metal hard mask then provides a mandrel 16 with a larger size, i.e., a second lateral width. Figures 3A-3C include similar and corresponding elements to those shown in Figures 1A-1D. These similar and corresponding elements are referenced using similar reference numerals. In this embodiment, the larger size metal hard mask includes a combination of a metal hard mask 18 and a metal spacer 26 as shown in Figure 3C.

[0053] Referring first to FIG. 3A, a structure is shown that includes at least one material layer 10L, one or more optional mandrel-based masking layers, including a first mandrel-based masking layer 12L and a second mandrel-based masking layer 14L, a mandrel material layer 16L, and a mandrel mask, each mandrel mask including a metal hard mask 18 and an optional dielectric oxide mask 20. Because the structure shown in FIG. 3A is identical to that shown in FIG. 1C, the processing steps previously described to provide the structure shown in FIG. 1C can now be used to provide the structure shown in FIG. 3A. At this point in the application, each metal hard mask 18 and each optional dielectric oxide mask 20 has a first size, i.e., a first lateral width. In one example, the first lateral width of each metal hard mask 18 and each optional dielectric oxide mask 20 is between 10 nm and 20 nm. In some cases, this first size needs to be adjusted to a larger size to enable patterning of larger-sized mandrels 16.

[0054] Referring now to FIG. 3B, the structure of FIG. 3B is shown after forming a metal-containing material layer 26L on the structure shown in FIG. 3A. The metal-containing material layer 26L is composed of one of the metals previously described for the metal hard mask layer 18L. In some embodiments, the metal-containing layer 26L is composed of a metal that is compositionally identical to the metal that provides the metal hard mask layer 18L. In other embodiments, the metal-containing layer 26L is compositionally different from the metal that provides the metal hard mask layer 18L. The metal-containing layer 26L can be formed by a deposition process, including, but not limited to, CVD, PECVD, PVD, or ALD. In some embodiments, the metal-containing layer 26L is a conformal layer. By "conformal layer," we mean a material layer that has a vertical thickness along horizontal surfaces that is substantially the same (i.e., within ±5%) as its lateral thickness along vertical surfaces. The metal-containing layer 26L can have a thickness ranging from 1 nm to 20 nm, although smaller and larger thicknesses can also be utilized. As shown in FIG. 3B, a metal-containing layer 26L is formed on each optional dielectric oxide mask 20, each metal hard mask 18, and the physically exposed surfaces of the mandrel material layer 16L.

[0055] 3C, the structure of FIG. 3B is shown after a directional etching process has been performed to remove the metal-containing layer 26L from all horizontal surfaces of the structure, leaving portions of the metal-containing layer 26L along the sidewalls of each metal hard mask 18 and, if present, each dielectric oxide mask 20. The remaining portions of the metal-containing layer 26L may be referred to as metal spacers 26. The metal spacers 26 function to increase the size, i.e., lateral width, of each metal hard mask 18, such that the combination of the metal spacers 26 and the metal hard mask 18 has an increased lateral (i.e., second) width compared to the first lateral width of each metal hard mask 18 without the metal spacers 26. This increased lateral (i.e., second) width provided by the combination of the metal spacers 26 and the metal hard mask 18 is used to increase the size of the mandrel pattern that is subsequently transferred to the mandrel material layer 16L. In one embodiment, the directional etching includes an RIE that selectively lands on the mandrel material layer 16L. The metal spacers 26 are generally I-shaped and have top surfaces that are flush with the top surface of the dielectric oxide spacers 20, or the top surface of the metal hard mask 18 if the dielectric oxide mask 20 is not present.

[0056] Referring now to FIG. 3D, the structure of FIG. 3C is shown after transferring the pattern of each metal hard mask 18 / metal spacer 26 combination (i.e., the increased size mandrel pattern) to provide the mandrels 16. This patterning transfer step is identical to that shown above in FIG. 1D. Note that the structure shown in FIG. 3D is after the dielectric oxide mask 20 and metal spacers 26 have been removed. The mandrels 16 have an increased size, i.e., lateral width, compared to the size, i.e., lateral width, of the mandrels 16 shown in FIG. 1D. The pitch of the mandrels 16 in this embodiment is within the ranges described above for previous embodiments of the invention.

[0057] After patterning the mandrel material layer 16L, conventional pitch-doubling processing steps are then performed, including forming mandrel pitch-doubled spacers (not shown) on the sidewalls of each of the mandrels 16, removing the mandrels 16 after the mandrel pitch-doubled spacers are formed, and then patterning the underlying at least one material layer 10L using the mandrel pitch-doubled spacers as an etch mask. In embodiments of the present invention, each metal hard mask 18 can be removed before or after mandrel pitch-doubled spacer formation using a material removal process that selectively removes the metal hard mask material. In some embodiments, removal of each metal hard mask 18 occurs after mandrel pitch-doubled spacer formation to provide additional protection during downstream pitch-doubling processing.

[0058] Referring now to Figures 4A-4D, one advantage that can be obtained using the method of the present invention is illustrated. This advantage, illustrated in these figures, provides protection for downstream pitch doubling processes, and this example is particularly applicable for SALELE. In conventional SALELE pitch doubling processes, a non-mandrel etch is performed after the mandrel pitch doubling spacers are formed along the sidewalls of the mandrels 16, allowing subsequently formed dielectric material present between the mandrels 16 to be removed from the structure. In conventional SALELE pitch doubling processes, this non-mandrel etch can remove and damage portions of the mandrels and mandrel pitch doubling spacers. Such damage affects the mandrel height and the pattern transferred to the underlying material layer that needs to be patterned. The presence of a metal hard mask 18 on top of the mandrels 16 mitigates this problem. Figures 4A-4D include similar and corresponding elements to those shown in Figures 1A-1D, and these similar and corresponding elements are referenced using similar reference numerals.

[0059] Referring first to Figure 4A, there is shown a structure including at least one material layer 10L, one or more optional mandrel base masking layers including first mandrel base masking layer 12L and second mandrel base masking layer 14L, mandrel material layer 16L, and mandrel masks, each mandrel mask including a metal hard mask 18 and an optional dielectric oxide mask 20. Because the structure shown in Figure 4A is identical to that shown in Figure 1C, the processing steps previously described used to provide the structure shown in Figure 1C can now be used to provide the structure shown in Figure 4A.

[0060] 4B, the structure shown in FIG. 4A is shown after forming a mandrel pitch doubled spacer material layer 30L on the structure shown in FIG. 4A. The mandrel pitch doubled spacer material layer 30L may be, for example, TiN, TiO x The mandrel pitch-doubled spacer material layer 30L is composed of a pitch-doubled spacer material, such as a metal, oxide, or nitride. In embodiments, the mandrel pitch-doubled spacer material layer 30L is compositionally different from at least the mandrels 16. The mandrel pitch-doubled spacer material layer 30L can be formed by a deposition process, including, but not limited to, CVD, PECVD, PVD, or ALD. In some embodiments, the mandrel pitch-doubled spacer material layer 30L is a conformal layer as defined above. The mandrel pitch-doubled spacer material layer 30L can have a thickness ranging from 1 nm to 20 nm, although smaller and larger thicknesses can also be used. As shown in FIG. 4B, the mandrel pitch-doubled spacer material layer 30L is formed on the metal hard mask 18, the mandrels 16, and the top mandrel-based masking layer, or on the physically exposed surface of at least one material layer if no mandrel-based masking is used.

[0061] Referring now to Figure 4C, the structure of Figure 4C is shown after a directional etching process has been performed to remove mandrel pitch doubled spacer material layer 30L from all horizontal surfaces of the structure, leaving portions of mandrel pitch doubled spacer material layer 30L along the sidewalls of each metal hard mask 18 and each mandrel 16. The remaining portions of mandrel pitch doubled spacer material layer 30L may be referred to as mandrel pitch doubled spacers 30. The mandrel pitch doubled spacers 30 are used to pattern at least one material layer 10L after mandrel removal. The mandrel pitch doubled spacers 30 are generally I-shaped, and their top surfaces would be flush with the top surfaces of each metal hard mask 18 if the dielectric oxide mask 20 were not present.

[0062] 4D, the structure is shown after forming a dielectric material layer 32 and patterned photoresist 34, followed by etching away the dielectric material layer 32 disposed between two adjacent mandrels 16. The dielectric material layer 32 may include one of the dielectric materials described above for the at least one material layer 10L. The dielectric material layer 32 may be formed utilizing one of the deposition processes described above when forming the at least one material layer 10L comprised of a dielectric material. Following the deposition of the dielectric material is a planarization and / or etch-back process to provide the dielectric material layer 32 with a top surface that is coplanar with the top surfaces of the mandrel pitch doubled spacers 30 and each of the mandrels 16.

[0063] A patterned photoresist 34 is then formed by the aforementioned lithography. The patterned photoresist 34 has openings that physically expose the dielectric material layer 32 disposed between adjacent mandrels 16. The exposed dielectric material layer 32 disposed between adjacent mandrels 16 is then removed using an etch such as RIE. The presence of the hard mask 32 mitigates any loss of the mandrels 16 and, in some embodiments where metal is used as the mandrel pitch doubled spacers 30, mitigates any loss of the mandrel pitch doubled spacers 30.

[0064] While the present invention has been particularly shown and described with respect to preferred embodiments thereof, it will be understood by those skilled in the art that the foregoing and other changes in form and detail may be made therein without departing from the scope of the invention. It is therefore intended that the present invention not be limited to the exact forms and details described and illustrated, but fall within the scope of the appended claims.

[0065] In a preferred embodiment of the invention described herein, a method of forming a structure is provided, the method comprising: forming a plurality of metal hard masks on top of a mandrel material layer overlying at least one material layer, the mandrel material layer being a dielectric material, a conductive material, or any combination thereof, present in a back-end interconnect (BEOL) region of a semiconductor structure; and patterning the mandrel material layer using the metal hard mask as an etch mask to provide a plurality of mandrels having a pitch between 26 nm and 64 nm. Preferably, each of the metal hard masks has a first lateral width, and the method further comprises adjusting the first lateral width to a second lateral width different from the first lateral width. Preferably, the first lateral width is greater than the second lateral width, and the adjusting step includes reducing the first lateral width using an etching process that removes portions of each of the metal hard masks. Preferably, the first lateral width is less than the second lateral width, and the adjusting step includes forming metal spacers along sidewalls of each of the metal hard masks. The method may further comprise forming one or more mandrel-based masking layers between the mandrel material layer and the at least one material layer. Preferably, in some embodiments, forming the metal hard mask comprises forming a mandrel-patterning material stack on the mandrel material layer, the mandrel-patterning material stack including a metal hard mask layer and a soft mask material layer, forming a patterned photoresist having a mandrel pattern on top of the mandrel-patterning material stack, and transferring the mandrel pattern to the soft mask material layer and the metal hard mask layer of the mandrel-patterning material stack by etching to provide a plurality of soft masks and a plurality of metal hard masks, respectively. Preferably, each of the soft masks is removed before patterning the mandrel material layer. Preferably, the mandrel-patterning material stack further comprises a dielectric oxide layer disposed between the metal hard mask layer and the soft mask material layer, the dielectric oxide layer transforming into the plurality of dielectric oxide masks upon transfer of the mandrel pattern.Preferably, the metal hard mask layer is comprised of titanium nitride, the dielectric oxide layer is comprised of silicon dioxide, the soft mask material layer is comprised of an organic planarization layer material, and the mandrel material layer is comprised of silicon nitride. The method may further comprise forming mandrel pitch double spacers along sidewalls of each of the mandrels and each of the metal hard masks.

[0066] In another preferred embodiment of the invention described herein, a method for forming a semiconductor structure is provided, the method comprising: forming a plurality of metal hard masks having a first lateral width on top of a mandrel material layer overlying at least one material layer; reducing the first lateral width of each of the metal hard masks to a second lateral width by etching; and providing a plurality of mandrels by patterning the mandrel material layer utilizing the metal hard masks having the second lateral width as an etch mask. In yet another preferred embodiment of the invention described herein, a method for forming a semiconductor structure is provided, the method comprising: forming a plurality of metal hard masks having a first lateral width on top of a mandrel material layer overlying at least one material layer; increasing the first lateral width of each of the metal hard masks to a second lateral width by forming metal spacers along sidewalls of each of the metal hard masks; and providing a plurality of mandrels by patterning the mandrel material layer utilizing the metal hard masks having the second lateral width as an etch mask.

Claims

1. forming a plurality of metal hard masks on top of the mandrel material layer overlying the at least one material layer; and patterning the mandrel material layer using the metal hard mask as an etch mask to provide a plurality of mandrels having a pitch of less than 36 nm.

1. A method of forming a structure, comprising:

2. 10. The method of claim 1, wherein the at least one layer of material is in a back-end-of-line (BEOL) region of a semiconductor structure.

3. The method of claim 1 , wherein the at least one material layer is a dielectric material, a conductive material, or any combination thereof.

4. The method of claim 1 , wherein the pitch of the mandrels is between 26 nm and 35 nm.

5. 10. The method of claim 1, wherein each of the metal hard masks has a first lateral width, the method further comprising adjusting the first lateral width to a second lateral width different from the first lateral width.

6. 6. The method of claim 5, wherein the first lateral width is greater than the second lateral width, and wherein the adjusting step comprises reducing the first lateral width using an etching process that removes respective portions of the metal hard mask.

7. 6. The method of claim 5, wherein the first lateral width is less than the second lateral width, and the adjusting step includes forming metal spacers along sidewalls of each of the metal hard masks.

8. The method of claim 1 , further comprising forming one or more mandrel base masking layers between the mandrel material layer and the at least one material layer.

9. The step of forming the plurality of metal hard masks comprises: forming a mandrel patterning material stack over the mandrel material layer, the mandrel patterning material stack including a metal hard mask layer and a soft mask material layer; forming a patterned photoresist having a mandrel pattern on top of the mandrel-patterned material stack; and transferring the mandrel pattern into the soft mask material layer and the metal hard mask layer of the mandrel patterning material stack by etching to provide a plurality of soft masks and a plurality of metal hard masks, respectively. The method of claim 1 , comprising:

10. The method of claim 9 , wherein the soft mask is removed prior to the patterning of the mandrel material layer.

11. 10. The method of claim 9, wherein the mandrel patterning material stack further comprises a dielectric oxide layer disposed between the metal hard mask layer and the soft mask material layer, the dielectric oxide layer transforming into a plurality of dielectric oxide masks during the transfer of the mandrel pattern.

12. 12. The method of claim 11 , wherein the metal hard mask layer is comprised of titanium nitride, the dielectric oxide layer is comprised of silicon dioxide, the soft mask material layer is comprised of an organic planarization layer material, and the mandrel material layer is comprised of silicon nitride.

13. 10. The method of claim 1, further comprising forming mandrel pitch-doubling spacers along a sidewall of each of the mandrels and each of the metal hard masks.

Images