TWI839906B - Layer forming method - Google Patents

Abstract

There is provided a method of forming a layer, comprising depositing a seed layer on the substrate; and depositing a bulk layer on the seed layer. Depositing the seed layer comprises supplying a first precursor comprising metal and halogen atoms to the substrate; and supplying a first reactant to the substrate. Depositing the bulk layer comprises supplying a second precursor comprising metal and halogen atoms to the seed layer; and, supplying a second reactant to the seed layer.

Description

Layer formation method

The present invention generally relates to a method of forming a layer on a substrate. More specifically, the present invention relates to sequentially repeating atomic layer deposition (ALD) cycles or chemical vapor deposition (CVD) processes to form at least a portion of a layer on a substrate having gaps, the gaps being generated during the fabrication of features. The layer on the substrate can be used to fabricate a semiconductor device. [CROSS-REFERENCE TO RELATED PATENT APPLICATIONS]

This application is a continuation-in-part of U.S. Nonprovisional Application No. 15/691,241, filed on August 30, 2017, entitled “LAYER FORMING METHOD,” and claims the benefit of U.S. Provisional Patent Application No. 62/607,070, filed on December 18, 2017, entitled “LAYER FORMING METHOD,” both of which are incorporated herein by reference.

In atomic layer deposition (ALD) and chemical vapor deposition (CVD), a first precursor and a first reactant are applied to a substrate that are suitable for reacting to form a desired layer on the substrate. The layer can be deposited in the gaps created during the process of fabricating features on the substrate to fill the gaps.

In ALD, a substrate is exposed to a pulse of a first precursor and a monolayer of the first precursor can be chemically adsorbed on the surface of the substrate. Surface sites can be occupied by all of the first precursor or by fragments of the first precursor. The reaction can be a chemically self-limiting reaction in that the first precursor will not adsorb on the substrate surface or will not react with the portion of the first precursor that is already adsorbed on the substrate surface. Excess first precursor is then purified by, for example, providing an inert gas and/or removing the first precursor from the reaction chamber. Subsequently, the substrate is exposed to a pulse of a first reactant, which chemically reacts with all or a fragment of the adsorbed first precursor until the reaction is complete and the surface is covered with a monolayer of reaction products.

It has been found that the quality of the deposited layer may need to be improved.

An improved method of forming a deposited layer on a substrate may be needed. Accordingly, a method of forming a layer may be provided, the method comprising: providing a substrate having a gap created during feature fabrication and depositing a seed layer on the substrate; and depositing a body layer on the seed layer. Depositing the seed layer may include: supplying a first precursor comprising metal and halogen atoms to the substrate; and supplying a first reactant to the substrate, wherein the first precursor reacts with a portion of the first reactant to form at least a portion of the seed layer. Depositing the bulk layer may include: supplying a second precursor comprising metal and halogen atoms to the seed layer; and supplying a second reactant to the seed layer, wherein the second precursor and a portion of the second reactant react to form at least a portion of the bulk layer on the seed layer. The first and second precursors may be different.

By having different first and second precursors for the seed layer and the bulk layer, the properties of the seed layer and the bulk layer can be optimized, thereby improving the quality of the overall layer. The first and second reactants can be the same and include hydrogen atoms.

In some other embodiments, a method for semiconductor processing is provided. The method includes depositing a metal layer in a gap in a substrate, thereby filling the gap.

Metal layers may be required as conductive layers in semiconductor devices. Gaps created during the fabrication of features in integrated circuit devices may be filled with metal layers. Such gaps may have a high aspect ratio because their depth is much greater than their width.

The gaps may extend vertically in a fabricated layer having a substantially horizontal top surface. Vertically extending gaps filled with metal may be used, for example, in word lines of dynamic random access memory (DRAM) type memory integrated circuits. Vertically extending gaps filled with metal may also be used, for example, in logic integrated circuits. For example, metal-filled gaps may be used as gate fills in PMOS or CMOS integrated circuits or in source/drain trench type contacts.

The gaps can also be arranged horizontally in the fabricated layer. Furthermore, the gaps can have a high aspect ratio, since their depth, in the horizontal direction, is much greater than their width. Gaps in the horizontal direction and filled with metal can be used, for example, in word lines of 3D NAND type memory integrated circuits. The gaps can also be arranged in a combination of vertical and horizontal directions.

The surface of the gap may include one type of deposited material. Alternatively, the surface of the gap may include different types of deposited materials. The surface of the gap may, for example, include aluminum oxide and/or titanium nitride. It may be difficult to deposit molybdenum on different materials in the gap when, for example, a molybdenum conductive layer may be desired in the gap. It may be desirable that the molybdenum layer can cover the entire surface of the gap and fill the entire gap. Additionally, it may be desirable that the molybdenum layer can cover the entire surface of the gap including different types of materials.

To fill the entire gap, a seed layer may be deposited in the gap and a bulk layer may be deposited on the seed layer. The seed layer may be formed by sequentially repeating a pretreatment atomic layer deposition (ALD) cycle. Alternatively, the seed layer may be formed by a chemical vapor deposition (CVD) process. The CVD process may be pulsed, where a first precursor is pulsed onto the substrate while a first reactant is continuously supplied to the substrate, or vice versa. The bulk layer may be deposited on the seed layer by sequentially repeating bulk ALD cycles. Alternatively, the bulk layer may be deposited on the seed layer by a CVD process. The CVD process may be pulsed, wherein the second precursor is supplied to the substrate in a pulse while the second reactant is continuously supplied to the substrate, or vice versa.

1a and 1b show a flow chart illustrating a method of depositing a layer according to one embodiment, wherein a seed layer may be deposited in a gap and a bulk layer may be deposited on the seed layer. A pre-treatment ALD cycle 1 for the seed layer may be as shown in FIG. 1a and a bulk ALD cycle 2 for the bulk layer may be as shown in FIG. 1b.

After providing the substrate with the gap in the reaction chamber in step 3, a first precursor comprising metal and halogen atoms can be supplied to the substrate in step 5 for a first supply period T1 (see Figure 1a). Subsequently, further supply of the first precursor to the substrate can be stopped by, for example, removing, for example purifying, a portion of the first precursor from the reaction chamber in step 7 for a first removal period R1. In addition, the cycle can include supplying 9 a first reactant to the substrate for a second supply period T2. The first precursor and a portion of the first reactant can react to form at least a portion of a seed layer on the substrate. Typically, it can take several (about 50) cycles before seed layer deposition begins. The further supply of the first reactant to the substrate may be stopped, for example, by removing, for example purging, a portion of the first reactant from the reaction chamber in step 11 for a second removal period R1.

The first precursor and the first reactant may be selected to have a suitable nucleation effect on the surface of the gap. The pre-treatment ALD cycle 1 may be repeated N times to deposit the seed layer, wherein N is selected between 100 and 1000, preferably between 200 and 800, and more preferably between 300 and 600. The seed layer may have a thickness between 1 and 20 nm, preferably between 2 and 10 nm, and more preferably between 3 and 7 nm.

After the pretreatment, ALD cycle 1 is repeated N times. A second precursor comprising metal and halogen atoms can be supplied to the substrate in step 11 by means of the main ALD cycle 2 for a third supply period T3 (see FIG. 1b ). This can be performed in the same reaction chamber as the pretreatment ALD cycle 1 of FIG. 1a or in a different reaction chamber. When the temperature requirements associated with the pretreatment cycle may be different, it may be advantageous to perform the main ALD cycle in a different reaction chamber than the pretreatment ALD cycle. Therefore, a substrate transfer may be necessary. Subsequently, the further supply of the second precursor to the substrate can be stopped, for example by removing, for example purging, a portion of the second precursor from the reaction chamber in step 13 for a third removal period R3.

In addition, the cycle may include supplying 15 the second reactant to the substrate for a fourth supply period T4. The second precursor and a portion of the second reactant may react to form at least a portion of the bulk layer on the substrate. Further supply of the second reactant to the substrate may be stopped, for example, by removing, for example, purifying a portion of the second reactant from the reaction chamber in step 17 for a fourth removal period R4. The second precursor and the second reactant may be selected to have appropriate electronic properties. For example, to have low resistivity. The molybdenum film may have a resistivity of less than 3000 μΩ-cm, or less than 1000 μΩ-cm, or less than 500 μΩ-cm, or less than 200 μΩ-cm, or less than 100 μΩ-cm, or less than 50 μΩ-cm, or less than 25 μΩ-cm, or less than 15 μΩ-cm or even less than 10 μΩ-cm.

The bulk ALD cycle 2 for the bulk layer may be repeated M times, where M is selected between 200 and 2000, preferably between 400 and 1200, and more preferably between 600 and 1000. The bulk layer may have a thickness between 1 and 100 nm, preferably between 5 and 50 nm, and more preferably between 10 and 30 nm.

The first and second precursors may include the same metal atom. The metal may be a transition metal atom. The transition metal atom may be molybdenum.

The first and second precursors may include the same halogen atom. The halogen atom may be chlorine. By having the same halogen, verification of tools and processes in the fab may be simplified because only one halogen needs to be evaluated. The first precursor may include molybdenum pentachloride (MoCl ₅ ).

During the pre-treatment ALD cycle, the processing temperature in the reaction chamber may be selected between 300 and 800° C., preferably between 400 and 700° C., and more preferably between 450 and 550° C. The vessel in which the first precursor is vaporized may be maintained between 40 and 100° C., preferably between 60 and 80° C., and more preferably at about 70° C.

The second precursor may include an additional atom that is not a metal or halogen atom. The additional atom may be a chalcogen. The chalcogen may be oxygen, sulfur, selenium or tellurium. The second precursor may include molybdenum(VI)dichloride (MoO ₂ Cl ₂ ).

During the bulk ALD cycle, the processing temperature may be between 300 and 800° C., preferably between 400 and 700° C., and more preferably between 500 and 650° C. The vessel in which the second precursor is vaporized may be maintained between 20 and 150° C., preferably between 30 and 120° C., and more preferably between 40 and 110° C.

Supplying the first and/or second precursor into the reaction chamber may take a duration T1, T3 selected between 0.1 and 10 seconds, preferably between 0.5 and 5 seconds, and more preferably between 0.8 and 2 seconds. For example, T1 may be 1 second and T3 may be 1.3 seconds. The flow rate of the first or second precursor in the reaction chamber may be selected between 50 and 1000 sccm, preferably between 100 and 500 sccm, and more preferably between 200 and 400 sccm. The pressure in the reaction chamber may be selected between 0.1 and 100 Torr, preferably between 1 and 50 Torr, and more preferably between 4 and 20 Torr.

One or both of the first and second reactants may have hydrogen atoms. At least one of the first and second reactants may include hydrogen (H ₂ ). The first and second reactants may be the same. The duration T2, T4 of supplying the first and/or second reactants to the reaction chamber may take between 0.5 and 50 seconds, preferably between 1 and 10 seconds, and more preferably between 2 and 8 seconds. The flow rate of the first or second reactant in the reaction chamber may be between 50 and 50,000 sccm, preferably between 100 and 20,000 sccm, and more preferably between 500 and 10,000 sccm.

Silane can be considered as the first and/or second reactant. The general formula of silane is _{SixH2 (x+2)} , where x is an integer 1, 2, 3 _, 4... Silane ( _SiH4 ), disilane ( _Si2H6 ) or _trisilane ( _Si3H8 ) can be suitable examples of the first and/or second reactant having hydrogen atoms.

Removing, for example, purging a portion of at least one of the first precursor, the first reactant, the second precursor, and the second reactant from the reaction chamber for a duration R1, R2, R3, or R4 can be performed between 0.5 and 50 seconds, preferably between 1 and 10, and more preferably between 2 and 8 seconds. Purging can be used after supplying the first precursor to the substrate; after supplying the first reactant to the substrate; after supplying the second precursor to the seed layer; and after supplying the second reactant to the seed layer to remove a portion of at least one of the first precursor, the first reactant, the second precursor, and the second reactant from the reaction chamber for a duration R1, R2, R3, or R4. Removal can be achieved by pumping and/or by providing a purge gas. The purge gas can be an inert gas such as nitrogen.

The method may be used in a single or batch wafer ALD apparatus. The method includes providing a substrate in a reaction chamber and a pre-treatment ALD cycle in the reaction chamber may include: supplying a first precursor onto the substrate in the reaction chamber; purging a portion of the first precursor from the reaction chamber; supplying a first reactant onto the substrate in the reaction chamber; and purging a portion of the first reactant from the reaction chamber. Additionally, the method includes providing a substrate in a reaction chamber and a main ALD cycle in the reaction chamber includes: supplying a second precursor onto the substrate in the reaction chamber; purging a portion of the second precursor from the reaction chamber; supplying a second reactant onto the substrate in the reaction chamber; and purging a portion of the second reactant from the reaction chamber.

Exemplary single wafer reactors designed specifically for performing ALD processes are available from ASM International NV (Almere, The Netherlands) under the trade names Pulsar®, Emerald®, Dragon®, and Eagle®. The method may also be performed in a batch wafer reactor, such as a vertical furnace. For example, the deposition process may be performed in an A412 ^™ vertical furnace, also available from ASM International NV. The furnace may have a processing chamber that can accommodate a load of 150 semiconductor substrates or wafers having a diameter of 300 mm.

The wafer reactor may be provided with a controller and a memory for controlling the reactor. The memory may be programmed to supply precursors and reactants to the reaction chamber according to embodiments of the present invention when executed on the controller.

Fig. 2 shows a cross section of a gap structure filled with a layer on a substrate according to one embodiment of the present invention. As shown, the gap can extend vertically and horizontally in the layer having a substantially horizontal top surface produced.

The gaps may have a high aspect ratio in that the depth in the vertical and/or horizontal directions is much greater than the width. For example, in the vertical direction, the gap has a width of 207 nm at the top, 169 nm in the middle, and 149 nm at the bottom, while the depth of the gap is much greater at 432 nm. For example, in the horizontal direction, the first gap has a width of 34 nm from the top, while the depth of the gap is much greater at 163 nm (rounded). The aspect ratio of the gap (gap depth/gap width) may be greater than about 2, greater than about 5, greater than about 10, greater than about 20, greater than about 50, greater than about 75, or in some cases even greater than about 100 or greater than about 150 or greater than about 200.

It may be noted that the aspect ratio of the gap may be difficult to determine, but in this context, the aspect ratio may be replaced by the surface enhancement ratio, which may be the ratio of the total surface area of the gap in the wafer or portion of the wafer relative to the flat surface area of the wafer or portion of the wafer. The surface enhancement ratio of the gap (surface gap/surface wafer) may be greater than about 2, greater than about 5, greater than about 10, greater than about 20, greater than about 50, greater than about 75, or even greater than about 100, or greater than about 150, or greater than about 200 in some cases.

The surface of the gap may comprise different kinds of deposited materials 19 _, 21. The surface may comprise _Al2O3 or TiN, for example.

A conformal metal layer 23 is deposited on the surface of the gap by sequentially repeating a pre-treatment ALD cycle with a first precursor to deposit a seed layer and by sequentially repeating a main body ALD cycle with a second precursor to deposit a main body layer. Details of the method used are shown in FIGS. 1a and 1b and the associated description. In some embodiments, the deposited Mo-containing film can have a step coverage of greater than about 50%, greater than about 80%, greater than about 90%, greater than about 95%, greater than about 98%, greater than about 99%.

The first and second precursors may comprise the same metal atom, such as a transition metal atom, such as molybdenum. The first and second precursors may comprise the same halogen atom, such as chlorine. The first precursor may comprise MoCl5. The second precursor may comprise another atom that is not a metal or halogen atom, such as a sulfide atom, such as oxygen. The second precursor may comprise molybdenum(VI)dichloride (MoO ₂ Cl ₂ ). The method may be performed in an atomic layer deposition apparatus. For example, the deposition processes may be performed in an EMERALD® XP ALD apparatus.

The first and second reactants are hydrogen (H ₂ ), which is supplied into the reaction chamber at a flow rate of 495 sccm for 5 seconds T2, T4. The purge gas nitrogen is used after supplying the first precursor; after supplying the first reactant; after supplying the second precursor; and after supplying the second reactant for 5 seconds R1, R2, R3 or R4.

During the pre-treatment and main ALD cycles, the processing temperature was about 550°C and the pressure was about 10 Torr. The container in which the first precursor was vaporized was about 70°C. The container in which the second precursor was vaporized was about 35°C.

A seed layer of about 4.6 nm was deposited using a pre-treatment ALD cycle for 500 cycles, and a bulk layer of about 21.4 nm was deposited using a bulk ALD cycle for 800 cycles. As shown, the molybdenum layer 23 was deposited very uniformly on the surface of the gap and had a total thickness of about 26 nm.

The orientation of the gap, whether it is horizontal or vertical, and the width of the gap do not appear to substantially affect the thickness of the layer 23. Furthermore, the material of the surface, whether it is _{Al2O3 19} or TiN 21, does not appear to affect the thickness of the layer 23. In this way, it is possible to produce metal-filled gaps with good uniformity.

The method can also be used in spatial atomic layer deposition equipment. In spatial ALD, precursors and reactants are continuously supplied in different physical sections and the substrate moves between the sections. At least two sections can be provided, in which case a half reaction can be carried out in the presence of a substrate. If a substrate is present in such a half reaction section, a monolayer can be formed from the first or second precursor. The substrate is then moved to another half reaction zone, where the ALD cycle is completed using the first or second reactant to form an ALD monolayer. Alternatively, the substrate position can be fixed and the gas supply can be moved, or a combination of the two. In order to obtain a thicker film, this process can be repeated.

According to one specific example of a spatial ALD apparatus, the method includes: placing a substrate in a reaction chamber comprising a plurality of sections, each section being separated from an adjacent section by an air curtain; supplying a first precursor to the substrate in the first section of the reaction chamber; moving the substrate surface laterally relative to the reaction chamber through the air curtain into the second section of the reaction chamber; supplying a first reactant to the substrate in the second section of the reaction chamber to form a seed layer; moving the substrate surface laterally relative to the reaction chamber through the air curtain; and repeatedly supplying the first precursor and reactant, including moving the substrate surface laterally relative to the reaction chamber, to form a seed layer.

To form a bulk layer, the method further comprises: placing a substrate in a reaction chamber comprising a plurality of sections, each section being separated from an adjacent section by an air curtain; supplying a second precursor to the substrate in a first section of the reaction chamber; moving the substrate surface laterally relative to the reaction chamber through the air curtain into a second section of the reaction chamber; supplying a second reactant to the substrate in the second section of the reaction chamber to form a bulk layer; moving the substrate surface laterally relative to the reaction chamber through the air curtain; and repeating the supply of the second precursor and reactant, including moving the substrate surface laterally relative to the reaction chamber, to form a bulk layer.

The first and second precursors may be different. The first and second reactants may be the same and include hydrogen atoms.

According to one embodiment, the seed layer may be deposited by a chemical vapor deposition (CVD) process, wherein a first precursor and a first reactant are simultaneously supplied to the substrate. The bulk layer may be deposited by a CVD process, wherein a second precursor and a second reactant are also simultaneously supplied to the substrate.

The CVD process may be a pulsed CVD process, in which the precursor is supplied to the substrate in a pulse while the reactant is supplied to the substrate continuously. This has the advantage that a higher concentration of reactant can reduce the concentration of halogens. High concentrations of halogens may damage semiconductor devices on the substrate.

For example, in a pulsed CVD process for a seed layer, the first precursor molybdenum pentachloride (MoCl5) may be provided alternately with a 1 second pulse and a 5 second purge gas flow rate. The first reactant hydrogen gas may be continuously supplied at a flow rate of 500 sccm and the substrate may be maintained at 550°C.

An exemplary single wafer reactor specifically designed for performing CVD processes is available from ASM International NV (Almere, The Netherlands) under the trade name Dragon®. The method can also be performed in a batch wafer reactor, such as a vertical furnace. For example, the deposition process can be performed in an A400 ^™ or A412 ^™ vertical furnace, also available from ASM International NV. The furnace can have a processing chamber that can accommodate 150 semiconductor substrates or wafer loads.

For manufacturing 3D NAND memory, word lines may have gaps that require low-resistivity metal fill. Existing solutions may utilize TiN as a seed layer for CVD tungsten gap fill. For current fluorine-based tungsten deposition processes, fluorine from the WF6 precursor can diffuse. A thicker (=3 nm) TiN barrier layer may be necessary to prevent fluorine diffusion and the diffused fluorine attacking the high-k Al2O3 film. However, the high resistivity of the Sn film (800 µΩ-cm at 3 nm) results in an increase in the resistivity of the TiN/W stack, which may be undesirable.

An improved method of forming a deposited layer on a substrate having low resistivity while being free of fluorine may be needed. Accordingly, a method of forming a layer may be provided, the method comprising: providing a substrate having gaps, the gaps being created during feature fabrication; depositing a seed layer on the substrate; and depositing a body layer on the seed layer. Depositing the body layer may include: supplying a second precursor comprising a transition metal such as tungsten to deposit the body layer on top of the seed layer.

The second precursor may include a halogen, such as chlorine, to deposit the bulk layer. The second precursor may be tungsten (V) chloride (WCl ₅ ) or tungsten (VI) chloride (WCl ₆ ). The bulk layer may be deposited by reacting tungsten (V) chloride (WCl ₅ ) or tungsten (VI) chloride (WCl ₆ ) with hydrogen H ₂ in an ALD or CVD mode of operation. For example, the reaction of WCl ₅ may be achieved at a temperature of 450° C. and a pressure of 40 Torr. The precursors may be provided in an ALD or CVD mode of operation.

The seed layer can be deposited by reacting a first precursor comprising molybdenum with hydrogen. The resistivity of the seed layer using molybdenum can be 107 µΩ-cm (3 nm), which is less than that of the TiN layer. In particular, good gap fill is achieved using this method for a 15 nm stack thickness (equivalent to a gap fill in a 30 nm CD structure). By depositing the host layer on top of the seed layer using tungsten (V) pentachloride (WCl ₅ ) or tungsten (VI) hexachloride (WCl ₆ ), it is possible to deposit the tungsten layer without the use of fluorine and still have a low resistivity. The seed layer precursor can include a transition metal such as molybdenum (Mo), a halogen such as chlorine (Cl) and optionally a chalcogen atom such as oxygen (O)). The driving agent before the seed layer can be, for example, pentachloride (MoCl ₅ ) or molybdenum(VI) dichloride (MoO ₂ Cl ₂ ), both of which react with hydrogen. If molybdenum pentachloride (MoCl ₅ ) is used together with molybdenum(VI) dichloride (MoO ₂ Cl ₂ ), the partial pressure of hydrogen can be reduced by 100 times.

The deposition rate of the molybdenum seed layer may be 1.2 angstroms per cycle. For comparison, the deposition rate of the TiN seed layer may be 0.6 angstroms per cycle under the same conditions. The deposition rate of the molybdenum seed layer may therefore be sufficient.

The metal deposited on the seed layer may be copper. The second precursor may include copper. The second precursor may include a halogen, such as chlorine, to deposit the main layer. The second precursor may include copper(II) chloride (CuCl ₂ ) or cuprous chloride (CuCl). The precursors may be provided in an ALD or CVD mode of operation reacting with hydrogen.

The metal deposited on the seed layer can be a transition metal or a precious metal from the following group: Ti, V, Cr, Mn, Nb, Mo, Ru, Rh, Pd, Ag, Hf, Ta, W, Re, Os, Ir and Pt. In some embodiments, the layer can include Co or Ni.

In other embodiments, the seed or host layer may contain less than about 40 atomic%, less than about 30 atomic%, less than about 20 atomic%, less than about 10 atomic%, less than about 5 atomic%, or even less than about 2 atomic% oxygen. In other embodiments, the seed or host layer may contain less than about 30 atomic%, less than about 20 atomic%, less than about 10 atomic%, less than about 5 atomic%, or less than about 2 atomic%, or even less than about 1 atomic% hydrogen. In some embodiments, the seed or host layer may contain less than about 10 atomic%, less than about 5 atomic%, less than about 1 atomic%, or even less than about 0.5 atomic% halogen atoms or chlorine. In still other embodiments, the seed or host layer may contain less than about 10 atomic%, less than about 5 atomic%, less than about 2 atomic%, less than about 1 atomic%, or even less than about 0.5 atomic% carbon. In the specific examples outlined herein, the atomic percent (atom %) concentration of an element can be determined using Rutherford backscattering (RBS).

In some embodiments of the present invention, forming a semiconductor device structure, such as a semiconductor device structure, may include forming a gate electrode structure including a molybdenum film, the gate electrode structure having an effective work function greater than about 4.9 eV, or greater than about 5.0 eV, or greater than about 5.1 eV, or greater than about 5.2 eV, or greater than about 5.3 eV, or even greater than about 5.4 eV. In some embodiments, the effective work function values provided above may be exhibited by an electrode structure including a molybdenum film having a thickness of less than about 100 angstroms, or less than about 50 angstroms, or less than about 40 angstroms, or even less than about 30 angstroms.

Those skilled in the art will appreciate that various omissions, additions, and modifications may be made to the above-described procedures and structures without departing from the scope of the present invention. It is contemplated that various combinations or sub-combinations of the specific features and aspects of the embodiments may be made and still be within the scope of the description. The various features and aspects of the disclosed embodiments may be sequentially combined or substituted with one another. All such modifications and variations are intended to be within the scope of the invention as defined by the appended claims.

1: Pre-treatment ALD cycle 2: Main ALD cycle 3: Step 5: Step 7: Step 9: Supply 11: Step 13: Step 15: Supply 17: Step 19: Deposited material 21: Deposited material 23: Conformal metal layer R1: Removal period/time R2: Time R3: Time R4: Time

These and other features, aspects and advantages of the present invention disclosed herein are described below with reference to the drawings of certain specific examples, which are intended to illustrate and not to limit the invention.

[FIGS. 1a and 1b] show a flow chart illustrating a method for depositing a layer according to one embodiment.

[FIG. 2] shows a cross section of a gap structure in which a layer is filled on a substrate according to a specific example.

Claims (19)

A method for forming a layer, comprising: providing a substrate to a reaction chamber; depositing a first layer on the substrate; and depositing a second layer on the first layer, wherein depositing the first layer comprises: supplying a first precursor comprising metal and halogen atoms to the substrate; and supplying a first reactant to the substrate, wherein the first precursor and a portion of the first reactant react to form at least a portion of the first layer; wherein depositing the second layer comprises: supplying a second precursor comprising metal and halogen atoms; and supplying a second reactant, wherein the second precursor and a portion of the second reactant react to form at least a portion of the second layer on the first layer, wherein at least one of the first reactant or the second reactant comprises silane, and wherein the first and second precursors are different. The method of claim 1, wherein the silane is represented by the general formula _{SixH (2x+2)} , wherein x is an integer greater than or equal to 1 and less than or equal to 4. The method of claim 1, wherein one of the first and second reactants comprises hydrogen (H ₂ ). The method of claim 1, wherein the first and second precursors contain the same metal atoms. The method of claim 1, wherein at least one of the first and second precursors contains a transition metal atom. The method of claim 5, wherein the transition metal atom is molybdenum. The method of claim 1, wherein the first and second precursors contain the same halogen atom. The method of claim 1, wherein the halogen atom is chlorine. The method of claim 1, wherein the first precursor comprises molybdenum pentachloride (MoCl ₅ ). The method of claim 1, wherein the second precursor comprises additional atoms that are not metal or halogen atoms. The method of claim 10, wherein the additional atom is a sulfide atom. The method of claim 11, wherein the sulfur atom is oxygen. The method of claim 11, wherein the second precursor comprises molybdenum(VI)dichloride (MoO ₂ Cl ₂ ). The method of claim 1, wherein the pressure in the reaction chamber is between 0.1 and 100 Torr. The method of claim 1, wherein the processing temperature in the reaction chamber is between 300°C and 800°C. The method of claim 1, wherein depositing the first layer comprises repeating an atomic layer deposition (ALD) cycle sequentially supplying the first precursor to the substrate and the first reactant to the substrate; and/or depositing the second layer comprises repeating an atomic layer deposition (ALD) cycle sequentially supplying the second precursor to the substrate and the second reactant to the substrate. The method of claim 1, wherein depositing at least one of the first and second layers comprises a chemical vapor deposition (CVD) process. The method of claim 5, wherein the transition metal atom is tungsten (W) or copper (Cu). The method of claim 1, wherein the second precursor comprises tungsten (W) or copper (Cu).