METHOD FOR FORMING PATTERNS WITH DIFFERENT CRITICAL DIMENSIONS

Abstract

METHOD FOR FORMING PATTERNS HAVING DIFFERENT CRITICAL DIMENSIONS One aspect of the invention relates to a method for forming a first group of patterns and a second group of patterns on a surface of a substrate, the first group of patterns being located in a first region of the surface of the substrate and the second group of patterns being located in a second region of the surface of the substrate, the method comprising the following steps: formation of a first sacrificial pattern and a second sacrificial pattern on the substrate, the first sacrificial pattern being located in the first region and having a first critical dimension, the second sacrificial pattern being located in the second region and having a second critical dimension strictly smaller than the first critical dimension;formation of first spacers in the first region, on flanks of the first sacrificial motif, and in the second region, on flanks of the second sacrificial motif; formation of a second spacer on one flank of each first spacer located in the second region, the formation of the second spacers including in particular a substep of conformal deposition of a second structural layer in the first and second regions, the second structural layer having a thickness equal to the difference between the first critical dimension and the second critical dimension; selective removal of the first and second sacrificial motifs from the first and second spacers.

Description

Title of the invention: METHOD FOR FORMING PATTERNS HAVING DIFFERENT CRITICAL DIMENSIONS TECHNICAL FIELD OF THE INVENTION

[0001] The technical field of the invention is that of advanced lithography techniques for the microelectronics industry. The present invention relates more particularly to a method for forming patterns on a substrate, comprising in particular the formation of spacers on the flanks of sacrificial patterns and the removal of the sacrificial patterns. TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0002] The self-aligned double structuring technique, better known by the acronym Self-Aligned Double Patterning (SADP) is an advanced lithography technique that doubles the density of patterns obtained using conventional photolithography equipment, such as a 193 nm immersion scanner.

[0003] This technique comprises five steps, illustrated by [Fig.l]: a. the formation, on a substrate 10, of sacrificial patterns 20 called “mandrels” or “guides”, typically by photolithography and etching of a sacrificial layer; b. the conformal deposition of a structural layer 30 made of a material called “spacer” on the substrate 10 and the sacrificial patterns 20; c. the anisotropic etching of the structural layer 30, so as to retain only its vertical parts, arranged on the sides of the sacrificial patterns 20 and forming spacers 40; d. removing the sacrificial motifs 20, selectively with respect to the spacers 40; e. the transfer of the spacers 40 into a surface portion of the substrate 10 (typically consisting of a hard mask layer).

[0004] A disadvantage of the SADP technique is that all the spacers 40 formed on the substrate 10 have the same critical dimension CD, defined by the thickness of the structural layer 30. However, the manufacture of an integrated circuit requires the formation of patterns having different critical dimensions.

[0005] The document [“Analysis of process characteristics of self-aligned multiple patterning”, Y. Chen et al., Microelectronic engineering, vol. 98, pp. 184-188, 2012] describes a method for simultaneously forming a first group of patterns having a first critical dimension and a second group of patterns having a second critical dimension different from the first.

[0006] [Fig.2] illustrates the main steps of this process: a. forming first mandrels 20a on a first region 10a of a substrate 10 and second mandrels 20b on a second region 10b of the substrate 10, the first mandrels 20a having a critical dimension CDi greater than the critical dimension CD2 of the second mandrels 20b; b. the formation of first spacers 41, called sacrificial spacers, on the sides of the first and second mandrels 20a-20b; c. the formation of second spacers 42, called structural spacers, on the flanks of the sacrificial spacers 41; d. removing the sacrificial spacers 41, selectively with respect to the mandrels 20a-20b and the structural spacers 42; and e. transferring the mandrels 20a-20b and the structural spacers 42 into a surface portion of the substrate 10.

[0007] The density of the patterns initially formed in the second region 10b (the second mandrels 20b) is tripled thanks to the double formation of spacers (steps b) and c)). The second region 10b is for this reason referred to as the “SATP” region, for “Self-Aligned Triple Patterning” in English. On the other hand, in the first region 10a, the structural spacers 42 between two first mandrels 20a are merged so as to obtain a doubling of the density of the patterns (region called “SADP”), and not a tripling.

[0008] This method therefore makes it possible to simultaneously form patterns of different critical dimensions, depending on the regions of the substrate. However, it does not make it possible to obtain a pitch between the patterns of the first region 10a (formed by the first mandrels 20a and the structural spacers 42) identical to the pitch between the patterns of the second region 10b (formed by the second mandrels 20b and the structural spacers 42). Furthermore, the patterns of the first region 10a are not regular in terms of critical dimension, because the first pattern and the last pattern are formed by a single structural spacer 42. Finally, the fusion of the structural spacers 42 is restrictive, because it requires perfect control of the space between the first two mandrels 20a. Summary of the invention

[0009] It is noted that there is a need to provide a method for forming a first group of patterns and a second group of patterns, the patterns of the first group having a different critical dimension from those of the second group, but the same repetition pitch.

[0010] According to the invention, this need is tended to be satisfied by providing a method for forming a first group of patterns and a second group of patterns on a surface of a substrate, the first group of patterns being located in a first region of the surface of the substrate and the second group of patterns being located in a second region of the surface of the substrate, the method comprising the following steps: • forming a first sacrificial pattern and a second sacrificial pattern on the substrate, the first sacrificial pattern being located in the first region and having a first critical dimension, the second sacrificial pattern being located in the second region and having a second critical dimension strictly less than the first critical dimension; • formation of first spacers in the first region, on flanks of the first sacrificial pattern, and in the second region, on flanks of the second sacrificial pattern, the formation of the first spacers comprising the following sub-steps: • conformal deposition of a first structural layer in the first and second regions; • anisotropic etching of the first structural layer; • forming a second spacer on a flank of each first spacer located in the second region, the formation of the second spacers comprising the following sub-steps: • conformal deposition of a second structural layer in the first and second regions, the second structural layer having a thickness equal to the difference between the first critical dimension and the second critical dimension; • anisotropic etching of the second structural layer; and • removal of the second structural layer in the first region; • removing the first and second sacrificial patterns selectively relative to the first and second spacers.

[0011] Thus, the patterns of the second group have a critical dimension greater than that of the patterns of the first group, due to the formation of the second spacers against the first spacers in the second region. The critical dimension difference is generated by a conformal deposition sub-step (of the second structural layer), which is easy to implement and can be controlled (to the nearest nm) using ALD (Atomic Layer Deposition) techniques. The pitch is preserved between the patterns of the first group and those of the second group, thanks to the fact that the critical dimension difference is found between the first and second sacrificial patterns initially formed.

[0012] Preferably, the first sacrificial pattern separates two first cavities having a third critical dimension, the second sacrificial pattern separates two second cavities having a fourth critical dimension and the first and second spacers are formed on side walls of the first and second cavities.

[0013] Advantageously, the third critical dimension and the fourth critical dimension satisfy the following relationships:

[0014] CD. = CD +2*e} J 1 1

[0015] CD4 = CD_+2xel + 2xe2

[0016] where CDi is the first critical dimension, ei is the thickness of the first structural layer, CD2 is the second critical dimension and e2 is the thickness of the second structural layer.

[0017] In first and second embodiments of the method, conformal deposition of the second structural layer is accomplished after anisotropic etching of the first structural layer.

[0018] In a third embodiment, the conformal deposition of the second structural layer follows the conformal deposition of the first structural layer and the anisotropic etching of the first structural layer follows the anisotropic etching of the second structural layer.

[0019] In addition to the characteristics which have just been mentioned in the preceding paragraphs, the method according to the invention may have one or more additional characteristics among the following, considered individually or according to all technically possible combinations: • the removal of the second structural layer in the first region is accomplished after the anisotropic etching of the second structural layer; • the removal of the second structural layer in the first region is accomplished before the anisotropic etching of the second structural layer; • the removal of the first and second sacrificial patterns is accomplished after the removal of the second structural layer in the first region; • the removal of the first and second sacrificial patterns is accomplished between the anisotropic etching of the second structural layer and the removal of the second structural layer in the first region; • removing the second structural layer in the first region comprises forming an etch mask on the second structural layer in the second region, etching the second structural layer through the etch mask, selectively with respect to the first sacrificial layer, and removing the etch mask; • the difference between the first critical dimension and the second critical dimension is greater than or equal to 4 nm, preferably between 4 nm and 70 nm. BRIEF DESCRIPTION OF THE FIGURES

[0020] Other characteristics and advantages of the invention will emerge clearly from the description given below, for information purposes only and in no way limiting, with reference to the appended figures, among which: • [Fig. 1] represents steps of a first pattern formation process, known as “SADP”; • [Fig.2] represents steps of a second method of forming patterns according to the prior art; • Figures 3A to 31 represent in schematic sectional view a first mode of implementation of the pattern formation method according to the invention; • Figures 4A to 41 represent in schematic sectional view a second mode of implementation of the pattern formation method according to the invention; and • Figures 5A to 5H represent in schematic sectional view a third mode of implementation of the pattern formation method according to the invention; and • Figures 6A to 6D represent in schematic sectional view a fourth embodiment of the pattern forming method according to the invention.

[0021] For greater clarity, identical or similar elements are identified by identical reference signs throughout the figures. DETAILED DESCRIPTION

[0022] Figures 3A-3I, 4A-4I and 5A-5H represent three embodiments of a method for forming patterns on a surface S of a substrate 10. With reference to Figures 31, 41 and 5H, this method makes it possible to obtain a first group of patterns 40a in a first region 10a of the surface S of the substrate 10 and a second group of patterns 40b in a second region 10b of the surface S of the substrate 10. The patterns 40a-40b are patterns projecting from the surface S of the substrate 10. They have for example the form of lines, these lines preferably being oriented parallel to each other. The surface S of the substrate 10 may be planar or have a topography (< 40 nm).

[0023] The substrate 10 may comprise a support layer 101, for example made of a semiconductor material such as silicon, a hard mask layer 102 arranged on the support layer 101 and an etching stop layer 103 arranged on the hard mask layer 102. The patterns 40a-40b are then arranged on the etching stop layer 103, which forms the surface S of the substrate 10.

[0024] Each of the patterns 40a of the first group is formed by a spacer and each of the patterns 40b of the second group is formed by two spacers joined to each other. The spacers are protruding patterns formed in pairs on the flanks of other patterns topographical called mandrels or guides.

[0025] The patterns 40a of the first group have a critical dimension CDA smaller than the critical dimension CDB of the patterns 40b of the second group. The term “critical dimension” refers to the smallest dimension of the patterns in a plane parallel to the surface S of the substrate 10. Here, the critical dimension of the patterns 40a-40b corresponds to their width (measured in the section plane of the figures). The critical dimensions CDA and CDb of the patterns 40a-40b are advantageously less than 40 nm. Furthermore, the difference between the critical dimension CDB of the patterns 40b of the second group and the critical dimension CDA of the patterns 40a of the first group is preferably greater than or equal to 4 nm, advantageously between 4 nm and 70 nm.

[0026] On the other hand, the patterns 40a of the first group and the patterns 40b of the second group have the same pitch P, sometimes called the repetition pitch or network pitch. The pitch P is the distance which separates the same edges (or the centers) of two consecutive patterns within the same group. The pitch P of the patterns 40a-40b is preferably less than 80 nm.

[0027] A first mode of implementation of the training method will now be described, step by step, with reference to Figures 3A-3I.

[0028] The forming method firstly comprises a step S1 aimed at forming one or more first sacrificial patterns 20a (or mandrels) in the first region 10a and one or more second sacrificial patterns 20b (or mandrels) in the second region 10b. The first sacrificial patterns 20a have a first critical dimension CDi and the second sacrificial patterns 20b have a second critical dimension CD2 strictly less than the first critical dimension CDi. The difference between the first critical dimension CDi and the second critical dimension CD2 is preferably greater than or equal to 4 nm, advantageously between 4 nm and 70 nm.

[0029] For the sake of simplification, the example of a single first sacrificial pattern 20a and a single second sacrificial pattern 20b will be taken below.

[0030] The first step S1 can be broken down into two sub-steps S1-1 and S1-2, illustrated respectively by figures 3A and 3B, themselves comprising several operations.

[0031] The first sub-step Sl-1 (see [Fig.3A]) comprises the deposition of a sacrificial layer 21 on the surface S of the substrate 10 and the formation of a first etching mask 22 on the sacrificial layer 21. The material of the sacrificial layer 21 is for example spin-on carbon (SOC), amorphous carbon (aC), amorphous silicon (a-Si), a resin, silicon nitride (SiN) or silicon dioxide (SiO2). The thickness of the sacrificial layer 21 is preferably between 50 nm and 100 nm.

[0032] The formation of the first etching mask 22 comprises a photolithography operation of a resin layer 221, and advantageously, an operation of transferring the photolithographed patterns into an underlying hard mask layer 222. A 193 nm immersion scanner can be used during the photolithography operation. The hard mask layer 222 is for example formed from a silicon anti-reflective coating (or SiARC, for “Silicon containing Anti-Reflective Coating”) or a dielectric anti-reflective coating (or DARC, for “Dielectric Anti-Reflective Coating”). SiARC has the advantage of being able to be deposited by spin-coating, like SOC (sacrificial layer 21), whereas DARC and amorphous carbon are generally deposited by plasma-enhanced chemical vapor deposition (PECVD).

[0033] The second sub-step S1-2 (see [Fig.3B]) comprises the etching of the sacrificial layer 21 through the first etching mask 22, until reaching the surface S of the substrate 10, and the removal of the etching mask 22 selectively with respect to the sacrificial layer 21 and the substrate 10. The etching of the sacrificial layer 21 preferably stops on the etching stop layer 103, for example made of titanium nitride (TiN) or silicon.

[0034] Thus, the photolithographed patterns in the resin layer 221 are transferred into the sacrificial layer 21 and form the sacrificial patterns 20a-20b. The critical dimensions of the photolithographed patterns ([Fig.3A]) can be maintained or reduced during the transfer, in order to achieve the desired values ​​of critical dimensions for the first sacrificial pattern 20a (CDi) and for the second sacrificial pattern 20b (CD2).

[0035] The method then comprises a step S2 of forming first spacers 41 in the first region 10a, on the flanks (or side walls) of the first sacrificial pattern 20a, and first spacers 41 in the second region 10b, on the flanks of the second sacrificial pattern 20b.

[0036] The formation of the first spacers 41 comprises two sub-steps S2-1 and S2-2, illustrated respectively by FIGS. 3C and 3D.

[0037] The first sub-step S2-1 (see [Fig.3C]) consists of conformally depositing a first so-called structural layer 31 in the first and second regions 10a-10b, so as to completely cover the first sacrificial pattern 20a and the second sacrificial pattern 20b (in other words, the sides and the upper face of the sacrificial patterns 20a-20b). The first layer 31 is described as “structural” because it serves to form at least part of the final patterns 40a-40b, as opposed to a so-called “sacrificial” layer which is intended to disappear completely before the end of the process.

[0038] The first structural layer 31 is thus of constant thickness ei and follows the relief formed on the surface S of the substrate 10. The thickness ei of the first structural layer 31 is preferably between 10 nm and 36 nm, for example equal to 20 nm. The material of the first structural layer 31 may be silicon dioxide (SiO2) or silicon nitride (SiN).

[0039] The first structural layer 31 is preferably deposited over the entire surface S of the substrate 10 (so-called “full wafer” deposition). The conformal deposition technique used to form the first structural layer 31 is, for example, atomic layer deposition (or ALD, for “Atomic Layer Deposition” in English), possibly plasma-assisted (PEALD, “Plasma Enhanced Atomic Layer Deposition”), or plasma-assisted chemical vapor deposition (PECVD). The maximum deposition temperature depends in particular on the material chosen to form the sacrificial patterns 20a-20b (material of the sacrificial layer 21). For example, certain SOCs impose a maximum deposition temperature of approximately 275°C, whereas amorphous carbon (deposited by PECVD) and amorphous silicon (deposited by PECVD, or even LPCVD) withstand higher temperatures.

[0040] During the second sub-step S2-2 (see [Fig.3D]), the first structural layer 31 is etched anisotropically, according to a preferential etching direction perpendicular to the surface S of the substrate 10 (or to the plane in which the substrate extends). This anisotropic etching step makes it possible to eliminate only the horizontal parts of the first structural layer 31, arranged on the surface S of the substrate 10 and on the upper faces of the sacrificial patterns 20a-20b. The vertical parts of the first structural layer 31, arranged against the sides of the sacrificial patterns 20a-20b, are preserved and constitute the first spacers 4L

[0041] The method then comprises a step S3 of forming a second spacer 42 on a flank of each first spacer 41 located in the second region 10b only. This step can be broken down into four sub-steps S3-1 to S3-4 illustrated by FIGS. 3E-3F and 3H-3I.

[0042] In this first embodiment (as in the second embodiment described below in relation to FIGS. 4A-4I), the second spacers 42 are first formed in the two regions 10a-10b (sub-steps S3-1 and S3-2), then the second spacers 42 located in the first region 10a are eliminated (sub-steps S3-3 and S3-4), to retain only the second spacers 42 located in the second region 10b.

[0043] The second spacers 42 are formed in the same manner as the first spacers 41 (described in relation to FIGS. 3C-3D), in two sub-steps S3-1 and S3-2 illustrated respectively by [Fig. 3E] and [Fig. 3F]: • S3-1: conformal deposition of a second structural layer 32 in the first and second regions 10a-10b (e.g. by ALD, PEALD or PECVD), here completely covering the sacrificial motifs 20a-20b and the first spacers 41; then • S3-2: anisotropic etching of the second structural layer 32, according to a preferential etching direction perpendicular to the surface S of the substrate 10 (or to the plane of the substrate 10).

[0044] The thickness e2 of the second structural layer 32 is constant and equal to the difference between the first critical dimension CDi (first sacrificial pattern 20a) and the second critical dimension CD2 (second sacrificial pattern 20b), i.e.:

[0045] [Math.l] e2 = CD1 - CD2

[0046] The thickness e2 of the second structural layer 32 is therefore preferably greater than or equal to 4 nm, advantageously between 4 nm and 70 nm.

[0047] The material of the second structural layer 32 may be silicon dioxide (SiO2), silicon nitride (SiN), aluminum oxide (A12O3) or titanium dioxide (TiO2). It is chosen so that it can be selectively etched relative to the material of the first structural layer 31 (and relative to the substrate 10).

[0048] During the third sub-step S3-3 represented by [Fig.3H], a second etching mask 50 is formed in the second region 10b (or outside the first region 10a), so as to protect the first and second spacers 41-42 located in this second region 10b. Then, the second spacers 42 located in the first region 10a are etched through the second etching mask 50 selectively with respect to the first spacers 41 (and with respect to the substrate 10).

[0049] The second etching mask 50 is for example formed from a two-layer stack comprising a layer of photosensitive resin arranged on a lower anti-reflective coating (or BARC, for “Bottom Anti-Reflective Coating”) (resin / BARC), or even a single layer of resin when the dimensions allow it (distance between the groups of patterns in particular).

[0050] The first and second regions 10a-10b of the surface S having dimensions much greater than the critical dimensions of the patterns 40a-40b to be formed, or even of the sacrificial patterns 20a-20b, the formation of the second etching mask 50 may comprise a conventional photolithography operation, for example using a 193 nm immersion scanner. In other words, this second photolithography operation is relaxed in terms of dimensioning. Typically, the area covered by the second etching mask 50 may have a width W of between 100 nm and 100 pm, for example equal to 1 pm.

[0051] Finally, with reference to [Fig.31], the second etching mask 50 is removed (selectively with respect to the spacers 41-42 and the substrate 10) during sub-step S3-4.

[0052] In other words, the sub-steps S3-3 and S3-4 of FIGS. 3H and 31 make it possible to remove from the first region 10a the second structural layer 32, here structured in the form of the second spacers 42, selectively with respect to the first structural layer 31, itself structured in the form of the first spacers 41.

[0053] An isotropic etching method is advantageously used to selectively etch the second structural layer 32 (or the second spacers 42) in the first region 10a. Examples of isotropic etching methods are given below, depending on the materials of the first and second structural layers 31-32.

[0054] Between the sub-step S3-2 of anisotropic etching of the second structural layer 32 and the sub-steps S3-3 and S3-4 of removing the second structural layer 32 in the first region 10a, the forming method according to the first embodiment comprises a step S4 of removing the first and second sacrificial patterns 20-20b, selectively with respect to the first and second spacers 41-42 and to the substrate 10 (see [Fig.3G]). Again, an isotropic (and selective) etching method is advantageously used. Examples include plasma etching based on O2 or H2 (N2 / H2, O2 / N2, O2, SO2 / O2, etc.) in the case of a carbon material (e.g., aC, SOC), and in the case of amorphous silicon, wet etching (e.g., tetramethylammonium hydroxide (TMAH)) or plasma etching based on HBr or Cl2 (HBr / O2 or HBr / Cl2 / O2NF3 / Hbr, etc.).

[0055] Thus, at the end of the formation process (see [Fig.31]), the patterns 40a of the first group (in the first region 10a) are formed by the first spacers 41. Their critical dimension CDA is equal to the thickness ei of the first structural layer 31.

[0056] [Math.2] CDA = e}

[0057] The patterns 40b of the second group (in the second region 10b) are each formed by the juxtaposition of a first spacer 41 and a second spacer 42. Their critical dimension CDB is equal to the sum of the thickness ei of the first structural layer 31 and the thickness e2 of the second structural layer 32.

[0058] [Math.3]

[0059] The pitch P of the patterns 40a of the first group is equal to the sum of their critical dimension CDA and the critical dimension CDi of the first sacrificial pattern 20a, but also to the sum between the thickness ei of the first structural layer 31, the thickness e2 of the second structural layer 32 and the critical dimension CD2 of the second sacrificial pattern 20b (see relation [Math. 1]), and therefore to the pitch of the patterns 40b of the second group.

[0060] [Math.4] P = CDa + CD{ = ex + CD = eA + e2 + CD2 = CDB + CD2

[0061] There are numerous combinations of materials and techniques (deposition, etching, etc.) for implementing steps S1 to S4 of the method of FIGS. 3A to 31. According to a first example, the sacrificial layer 21 (forming the sacrificial patterns 20a-20b) is made of a carbon material (e.g., aC, SOC), the first structural layer 31 (forming the first spacers 41) is made of SiO2 (can be deposited by PEALD), the second structural layer 32 (forming the second spacers 42) is made of Al2O3 or TiO2 (can be deposited by ALD) and the etch stop layer 103 of the substrate 10 is made of silicon. The second structural layer 32 is etched (in the first region 10a) isotropically and selectively with respect to the first structural layer 31 (during sub-step S3-3 of [Fig.3H]) by means of a BC13 / C12 plasma.According to a second example, the sacrificial layer 21 is made of a carbon material, the first structural layer 31 is made of SiO2, the second structural layer 32 is made of SiN and the etch stop layer 103 of the substrate 10 is made of TiN. The second structural layer 32 is etched (in the first region 10a) isotropically and selectively with respect to the first structural layer 31 by means of a plasma containing fluorine (e.g. SF6 / Hbr, NF3 / Hbr or CH3F / O2 / He) without ion bombardment.

[0062] Figures 4A to 41 represent a second mode of implementation of the pattern formation method.

[0063] This second embodiment differs from the first embodiment in that the step S4 of removing the sacrificial patterns 20a-20b (see [Fig.4I]) is performed after the sub-steps S3-3 and S3-4 of removing the second structural layer 32 (structured in the form of the second spacers 42) in the first region 10a ([Fig.4G] and [Fig.4H]). In other words, the sacrificial patterns 20a-20b are removed at the end of the patterning process.

[0064] This chronology provides two advantages. First, the sacrificial pattern 20a makes the first spacers 41 more mechanically stable, during the etching of the second spacers 42 located in the first region 10a (sub-step S3-3 of [Fig.4G]), which makes it possible to choose a wet etching without risk of collapse of the first spacers 41 (this risk being for example linked to capillary forces). Second, the formation of the second etching mask 50 is easier, because the space to be filled is smaller in the presence of the second sacrificial pattern 20b.

[0065] Furthermore, in this second embodiment, the sacrificial layer 21 (and therefore the sacrificial patterns 20a-20b) is advantageously made of an inorganic material, such as amorphous silicon. Thus, during the formation of the second etching mask 50 (sub-step S3-3 of [Fig.4G]), the resin / BARC stack (or the resin layer alone) can be removed in the first zone 10a without damaging the first sacrificial pattern 20a.

[0066] The first etching mask 22 used to delimit the sacrificial patterns 20a-20b (step Sl-1 of [Fig.4A]) may in this case comprise, in addition to the resin layer 221 and the hard mask layer 222 (e.g. SiARC), a SOC layer 223 arranged between the sacrificial layer 21 and the hard mask layer 222.

[0067] In the case of a sacrificial layer 21 made of amorphous silicon, the etch stop layer 103 of the substrate 10 is preferably made of TiN.

[0068] Beyond these differences, step S1 of forming the sacrificial patterns 20a-20b ([Fig.4A] and [Fig.4B]), step S2 of forming the first spacers 41 ([Fig.4C] and [Fig.4D]), step S3 of forming the second spacers 42 in the first region 10a ([Fig.4E], [Fig.4F], [Fig.4G] and [Fig.4H]) and step S4 of removing the sacrificial patterns 20a-20b ([Fig.41]) are accomplished in the manner described in relation to FIGS. 3A-3I.

[0069] There are many combinations of materials and techniques (deposition, etching, etc.) for implementing steps S1 to S4 of the method of FIGS. 4A to 41. According to a first example, the sacrificial layer 21 (forming the sacrificial patterns 20a-20b) is made of amorphous silicon, the first structural layer 31 (forming the first spacers 41) is made of SiN (resistant to hydrofluoric acid, HF), the second structural layer 32 (forming the second spacers 42) is made of SiO2 (can be deposited by PEALD or ALD at low temperature) and the etch stop layer 103 of the substrate 10 is made of TiN. An HF-based wet etch is performed to remove the second structural layer 32 in the first region 10a (sub-step S3-3 of [Fig.4G]).According to a second example, the sacrificial layer 21 is made of amorphous silicon, the first structural layer 31 is made of SiO2, the second structural layer 32 is made of SiN and the etch stop layer 103 of the substrate 10 is made of TiN. The second structural layer 32 is etched (in the first region 10a) isotropically and selectively with respect to the first structural layer 31 by means of a plasma containing fluorine (e.g. SF6 / Hbr or CH3F / O2 / He) without ion bombardment.

[0070] Figures 5A to 5H represent a third mode of implementation of the pattern formation method, which is distinguished from the two previous ones by a different chronology in the manufacturing steps and sub-steps.

[0071] Thus, in this third embodiment, the conformal deposition of the second structural layer 32 (sub-step S3-1 represented by [Fig.5D]) is consecutive to the conformal deposition of the first structural layer 31 (sub-step S2-1 represented by [Fig.5C]) and the anisotropic etching of the first structural layer 31 (sub-step S2-2) is consecutive to the anisotropic etching of the second structural layer 32 (sub-step S3-2, cf. [Fig.5G]).

[0072] In other words, the two structural layers 31-32 are deposited one on top of the other, advantageously using the same conformal deposition technique (PEALD, ALD and PECVD) ([Fig.5D]), then etched one after the other to delimit the first and second spacers 41-42, preferably using the same etching process, for example a fluorine-based plasma etching (e.g. CH2F2 / He / CF4 or CF4 / Ar) with ion bombardment ([Fig.5G]).

[0073] Furthermore, sub-steps S3-3 and S3-4 of step S3 (see [Fig.5E] and [Fig.5F]), for removing the second structural layer 32 in the first region 10a, are performed before the anisotropic etching of the second structural layer 32 ([Fig.5G]). In other words, the etching mask 50 is formed on the first and second structural layers 31-32 before they are structured in the form of the first spacers 41 and second spacers 42 respectively. Similarly, the second structural layer 32 is removed in the first area 10a, selectively with respect to the first structural layer 31, before it is structured in the form of the second spacers 42.

[0074] The removal of the sacrificial patterns 20a-20b (step S4 of [Fig.5H]) is accomplished after the anisotropic etching sub-steps S3-2 and S2-2, therefore after the removal of the second structural layer 32 in the first region 10a (sub-steps S3-3 and S3-4 of FIGS. 5E-5F), as in the second embodiment.

[0075] This third embodiment is the simplest to implement, since the two conformal deposition steps are consecutive and the two anisotropic etching steps are also consecutive. On the other hand, there is a risk of less good control of the critical dimension of the spacers, since the thickness of the layers to be etched anisotropically varies between the two regions 10a-10b. More precisely, in the first region 10a, the horizontal portions of the first structural layer, of thickness eb, are eliminated, whereas in the second region 10b, the horizontal portions of the first structural layer 31 and of the second structural layer 32, i.e. the total of the thicknesses ei and e2, are eliminated.

[0076] The step S1 of forming the sacrificial patterns 20a-20b ([Fig.5A] and [Fig.5B]) can be accomplished in the manner described in relation to FIGS. 3A-3B.

[0077] There are numerous combinations of materials and techniques (deposition, etching, etc.) for implementing steps S1 to S4 of the method of FIGS. 5A to 5H. According to a first example, the sacrificial layer 21 is made of a carbon material or amorphous silicon, the first structural layer 31 is made of SiO2, the second structural layer 32 is made of SiN (can be deposited by PEALD or ALD) and the etch stop layer 103 of the substrate 10 is made of TiN. Etching using a fluorine-containing plasma (e.g., SF6 / Hbr or CH3F / O2 / He) without ion bombardment is performed to remove the second structural layer 32 in the first region 10a (substep S3-3 of [Fig. 5E]). According to a second example, the sacrificial layer 21 is made of a carbon material or amorphous silicon, the first structural layer 31 is made of SiO2, the second structural layer 32 is made of SiN (can be deposited by PEALD or ALD) and the etch stop layer 103 of the substrate 10 is made of TiN. Etching using a fluorine-containing plasma (e.g., SF6 / Hbr or CH3F / O2 / He) without ion bombardment is performed to remove the second structural layer 32 in the first region 10a (substep S3-3 of [Fig. 5E]). The structural layer 31 is SiN (resistant to hydrofluoric acid, HF), the second structural layer 32 is SiO2 (can be deposited by PEALD or low temperature ALD) and the etch stop layer 103 of the substrate 10 is TiN. An HF-based wet etch is performed to remove the second structural layer 32 in the first region 10a.

[0078] The method described above in relation to Figures 3A-3I, 4A-4I and 5A-5H makes it possible to easily obtain patterns 40a-40b having a small difference in critical dimension, typically greater than or equal to 4 nm, because this difference in critical dimension corresponds to the thickness of a conformal deposition, a manufacturing step which is well controlled in the microelectronics industry. In addition, the patterns obtained are symmetrical in terms of pitch and critical dimension, unlike those obtained with the method of the prior art ([Fig.2]; SADP region).

[0079] The patterns 40a-40b formed on the surface S of the substrate 10 can then be transferred into the substrate 10, to form an integrated circuit level, for example the transistor gates or interconnection lines. In the integration example described above, the patterns 40a-40b are first transferred into the etch stop layer 103, then into the hard mask layer 102, and finally into the support layer 101.

[0080] The two groups of patterns are advantageously spaced apart by a distance D greater than or equal to the first critical dimension CDi, preferably between 44 nm and 10 pm, for example equal to 300 nm.

[0081] As illustrated in the figures, the first sacrificial pattern 20a can separate two first cavities 21a having a third critical dimension CD3 and the second sacrificial pattern 20b can separate two second cavities 21b having a fourth critical dimension CD4.

[0082] These cavities 21a-21b are formed by etching the sacrificial layer 21 during the second sub-step S1-2 of the step S1 of forming the sacrificial patterns 20A-20b.

[0083] They make it possible to form two other first spacers 41 and two other second spacers 42 in each region 10a, 10b of the surface S of the substrate 10, on side walls of the cavities 21a-21b located opposite the flanks of the sacrificial patterns 20a, 20b, and therefore to increase (from 2 to 4) the number of patterns 40a, 40b in each group. The flanks of the sacrificial patterns 20a, 20b also constitute side walls of the cavities 21a-21b.

[0084] For this, the conformal deposition of the first structural layer 31 is carried out so that it covers the bottom and the side walls of the cavities 21a-21b (but without filling them entirely) and the conformal deposition of the second structural layer 32 is carried out so that it covers the bottom and the side walls of the cavities 21a-21b (but without filling them entirely).

[0085] For the pitch P of the patterns 40a, 40b to be constant, the third critical dimension CD3 and the fourth critical dimension CD4 satisfy the following relationships:

[0086] [Math.5] CD^CD+lxe^

[0087] where CDi is the first critical dimension (first sacrificial pattern 20a) and ei is the thickness of the first structural layer 31; and

[0088] [Math.6] CD4 = CD + 2 x + 2 x e2

[0089] where CD2 is the second critical dimension (second sacrificial pattern 20b), ei is the thickness of first structural layer 31 and e2 is the thickness of second structural layer 32.

[0090] Each region 10a, 10b can contain several sacrificial patterns 20a, 20b of the same critical dimension CDi, CD2 (and separated two by two by a cavity 21a, 21b) in order to obtain a number of patterns greater than 4 in the corresponding group.

[0091] The pattern forming method can be extended to more than two groups of patterns. Figures 6A to 6D show an embodiment of the method for forming three groups of patterns (of different critical dimensions but having the same repetition pitch).

[0092] At least one third sacrificial pattern 20c is formed in a third region 10c of the substrate 10, preferably at the same time as the first and second sacrificial patterns 20a-20b (and in the manner described above). This third sacrificial pattern 20c has a critical dimension CD5 strictly smaller than the (second) critical dimension CD2 of the second sacrificial pattern 20b. The difference between the critical dimension CD5 of the third sacrificial pattern 20c and the critical dimension CD2 of the second sacrificial pattern 20b is preferably greater than or equal to 4 nm, advantageously between 4 nm and 70 nm.

[0093] In addition to the steps S1-S4 described previously, the forming method comprises a step S5 of forming a third spacer 43 on a flank of each second spacer 42 located in the third region 10c, from which it results that each pattern of the third region 10c is formed by the juxtaposition of a first spacer 41, a second spacer 42 and a third spacer 43.

[0094] The formation of the third spacers 43 comprises the sub-steps of: • S5-1: conformal deposition of a third structural layer 33 in the first, second and third regions 10a-10c, the third structural layer 43 having a thickness e3 equal to the difference between the second critical dimension CD2 of the second sacrificial pattern 20b and the critical dimension CD5 of the third sacrificial pattern 20c (cf. [Fig.6A]); • S5-2: removal of the third structural layer 33 in the first and second regions lOa-lOb (cf. [Fig.6B]); and • S5-3: anisotropic etching of the third structural layer 33 (according to a preferential direction perpendicular to the surface S of the substrate) (cf. [Fig.6D]).

[0095] The conformal deposition of the third structural layer 33 (sub-step S5-1; cf. [Fig.6A]) may be consecutive to the conformal deposition of the second structural layer 32 (sub-step S3-1), itself consecutive to the conformal deposition of the first structural layer 31 (sub-step S2-1), so that the third structural layer 33 is deposited on the first and second structural layers 31-32 before their etching in the form of spacers.

[0096] The localized removal of the third structural layer 33 preferably comprises: • the formation of a third etching mask 60 (e.g. resin / BARC) in the third region 10c (or outside the first and second regions 10a-10b), so as to protect the portions of the first, second and third structural layers 31-33 located in this third region 10c; • etching the third structural layer 33 through the third etching mask 60, selectively with respect to the second structural layer 32 and, preferably, isotropically; and • the removal of the third etching mask 60 (selectively with respect to the second structural layer 32 and the third structural layer 33).

[0097] In addition to the second region 10b, the second etching mask 50 covers the third region 10c, so as to retain the second structural layer 32 also in the third region 10c when removing the second structural layer 32 from the first region 10a (sub-step S3-3 of [Fig.6C]).

[0098] The anisotropic etchings of the third structural layer 33 (sub-step S5-3), of the second structural layer 32 (sub-step S3-2) and of the first structural layer 31 (sub-step S2-2) can follow one another in this order (see [Fig.6D]). They are preferably accomplished using the same etching method.

[0099] In an alternative implementation, the third structural layer 33 is deposited conformally (substep S5-1) and removed from the first and second regions 10a-10b (substep S5-2) before the second structural layer 32 is deposited (substep S3-1) and removed from the first region 10a (substeps S3-3 and S3-4).

[0100] Many variations and modifications of the pattern forming method will be apparent to those skilled in the art. The patterns formed may adopt geometries other than parallel lines. The patterns may in particular be rings (or ripped cylinders) or lines perpendicular to each other.

Claims

Claims

1. A method of forming a first group of patterns (40a) and a second group of patterns (40b) on a surface (S) of a substrate (10), the first group of patterns (40a) being located in a first region (10a) of the surface of the substrate and the second group of patterns (40b) being located in a second region (10b) of the surface of the substrate, the method comprising the following steps: - formation (SI) of a first sacrificial pattern (20a) and a second sacrificial pattern (20b) on the substrate (10), the first sacrificial pattern (20a) being located in the first region (10a) and having a first critical dimension (CDi), the second sacrificial pattern (20b) being located in the second region (10b) and having a second critical dimension (CD2) strictly less than the first critical dimension (CDÛ; - formation (S2) of first spacers (41) in the first region (10a), on flanks of the first sacrificial pattern (20a), and in the second region (10b), on flanks of the second sacrificial pattern (20b), the formation of the first spacers (41) comprising the following sub-steps: • conformal deposition (S2-1) of a first structural layer (31) in the first and second regions (10a-10b); • anisotropic etching (S2-2) of the first structural layer (31); - formation (S3) of a second spacer (42) on a flank of each first spacer (41) located in the second region (10b), the formation of the second spacers (42) comprising the following sub-steps: • conformal deposition (S3-1) of a second structural layer (32) in the first and second regions (10a-10b), the second structural layer (32) having a thickness (e2) equal to the difference between the first critical dimension (CDi) and the second critical dimension (CD2); • anisotropic etching (S3-2) of the second structural layer (32); and • removal (S3-3, S3-4) of the second structural layer (32) in the first region (10b); - removal (S4) of the first and second sacrificial patterns (20a-20b) selectively with respect to the first and second spacers (41-42).

2. Method according to claim 1, wherein: - the first sacrificial pattern (20a) separates two first cavities (21a) having a third critical dimension (CD3); - the second sacrificial pattern (20b) separates two second cavities (21b) having a fourth critical dimension (CD4); and - the first and second spacers (41-42) are formed on side walls of the first and second cavities (21a-21b).

3. The method of claim 2, wherein the third critical dimension (CD3) and the fourth critical dimension (CD4) satisfy the following relationships: CD3= CD+ 2xei CD4 = CD + 2 xej + 2xe2 where CDi is the first critical dimension, ei is the thickness of the first structural layer (31), CD2 is the second critical dimension, and e2 is the thickness of the second structural layer (32).

4. A method according to any one of claims 1 to 3, wherein the conformal deposition (S3-1) of the second structural layer (32) is accomplished after the anisotropic etching (S2-2) of the first structural layer (31).

5. Method according to any one of claims 1 to 3, in which the conformal deposition (S3-1) of the second structural layer (32) is consecutive to the conformal deposition (S2-1) of the first structural layer (31) and in which the anisotropic etching (S2-2) of the first structural layer (31) is consecutive to the anisotropic etching (S3-2) of the second structural layer (32).

6. A method according to any one of claims 1 to 5, wherein the removal (S3-3, S3-4) of the second structural layer (32) in the first region (10a) is accomplished after the anisotropic etching (S3-2) of the second structural layer (32).

7. A method according to any one of claims 1 to 5, wherein the removal (S3-3, S3-4) of the second structural layer (32) in the first region (10a) is accomplished before the anisotropic etching (S3-2) of the second structural layer (32).

8. A method according to any one of claims 1 to 7, wherein the removal (S4) of the first and second sacrificial patterns (20a-20b) is accomplished after the removal (S3-3, S3-4) of the second structural layer (32) in the first region (10a).

9. A method according to any one of claims 1 to 6, wherein the removal (S4) of the first and second sacrificial patterns (20a-20b) is accomplished between the anisotropic etching (S3-2) of the second structural layer and the removal (S3-3, S3-4) of the second structural layer (32) in the first region (10a).

10. The method of any one of claims 1 to 9, wherein removing the second structural layer (32) in the first region (10a) comprises forming (S3-3) an etch mask (50) on the second structural layer (32) in the second region (10b), etching (S3-3) the second structural layer (32) through the etch mask (50), selectively with respect to the first sacrificial layer (31), and removing (S3-4) the etch mask (50).

11. Method according to any one of claims 1 to 10, wherein the difference between the first critical dimension (CDi) and the second critical dimension (CD2) is greater than or equal to 4 nm, preferably between 4 nm and 70 nm.

Images