Method for depositing silicon nitride film

Abstract

A method for depositing a silicon nitride layer on a stack is provided. The method includes providing an atomic layer deposition comprising a plurality of cycles, wherein each cycle includes feeding the stack with a silicon-containing precursor by providing a silicon-containing precursor gas, providing an N2 plasma conversion, and providing an H2 plasma conversion.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of priority to U.S. Application No. 62 / 748,085, filed on October 19, 2018, which is incorporated herein by reference for all purposes. Background Art

[0003] The present disclosure relates to methods of depositing silicon nitride films. More particularly, the present disclosure relates to depositing silicon nitride films using atomic layer deposition. Summary of the invention

[0004] To achieve the aforementioned objectives and in accordance with the purposes of the present disclosure, a method for depositing a silicon nitride layer on a stack is provided. The method includes providing an atomic layer deposition comprising a plurality of cycles, wherein each cycle includes feeding the stack with a silicon-containing precursor by providing a silicon-containing precursor gas, providing N2 plasma conversion, and providing H2 plasma conversion.

[0005] In another implementation, an apparatus for depositing atomic layer deposits of silicon nitride on a stack is provided. A processing chamber is provided. A substrate support is within the processing chamber. A gas inlet provides gas into the processing chamber. A gas source provides gas to the gas inlet, wherein the gas source includes a silicon-containing precursor source, an N2 gas source, and an H2 gas source. An exhaust pump is provided for pumping gas from the processing chamber. An electrode provides RF power in the processing chamber. At least one power supply powers the electrode. A controller is controllably connected to the gas source and the at least one power supply. The controller includes at least one processor and a computer readable medium. The computer readable medium includes computer code for deposition, the deposition being performed by depositing an atomic layer of silicon nitride on the stack by providing a plurality of cycles, wherein each of the plurality of cycles includes feeding the stack with a silicon-containing precursor by providing a silicon-containing precursor gas from a silicon-containing precursor source, providing an N2 plasma conversion, and providing an H2 plasma conversion.

[0006] These and other features of the present disclosure will be described in more detail below in the detailed description of the disclosure and in conjunction with the following figures. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The present disclosure is illustrated by way of example and not limitation in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:

[0008] Figure 1 is a high-level flow chart of an implementation scheme.

[0009] Figure 2A and Figure 2B is a schematic cross-sectional view of a stack processed according to one embodiment.

[0010] Figure 3 is a schematic diagram of a processing chamber that may be used in one embodiment.

[0011] Figure 4 is a schematic diagram of a computer system that can be used to implement one embodiment. DETAILED DESCRIPTION

[0012] The present disclosure will now be described in detail with reference to some exemplary embodiments of the present disclosure as shown in the accompanying drawings. In the following description, many specific details are set forth in order to provide a thorough understanding of the present disclosure. However, it will be apparent to those skilled in the art that the present disclosure may be practiced without some or all of these specific details. In other cases, known processing steps and / or structures are not described in detail to avoid making the present disclosure unclear unnecessarily.

[0013] Conformal plasma enhanced atomic layer deposition (PEALD) of silicon nitride (SiN) films deposited between 150°C and 450°C faces the challenge of achieving low wet etch rate (WER) targets while simultaneously achieving high growth rates. Since halide-based silicon precursors are the only known ALD SiN silicon precursors that can achieve 100% conformality in this temperature range, films with both low WER and high growth rates typically require some combination of nitrogen (N2) and ammonia (NH3) in the nitride conversion plasma. N2 plasma results in lower WER, while NH3 plasma results in higher growth rate and conformality. Unfortunately, the use of NH3 can lead to undesirable side effects. The high reactivity of NH3 can lead to upstream particle generation and metal corrosion when NH3 reacts with residual halide-based silicon precursors. These reactions result in excessive particle generation and release of volatile halide-based acid byproducts that can attack metal chamber components. Removal of NH3 from the conversion plasma results in low growth rates (< / cycle), and long radio frequency (RF) on-times are required to obtain fully conformal membranes (>8 seconds). Both of these effects result in very low flux (tpt), which makes the membranes commercially unattractive.

[0014] Current low WER, high growth rate conformal ALD SiN processes typically use a combination of N2 and NH3 plasmas to convert Si-X (X = halide or hydrogen atom) ligands to -Si-N- bonds. Issues associated with NH3 are high particle generation and high chamber etch rates. For packaging applications, NH3-plasmas are known to promote high elemental losses when depositing on sensitive chalcogenide materials. Additionally, typically, NH3 plasmas promote the incorporation of large amounts of -NH xpart, which results in a higher WER in dilute hydrofluoric acid (dHF).

[0015] Example

[0016] For ease of understanding, Figure 1 is a high level flow diagram of one embodiment. A stack is provided (step 104). Figure 2A is a schematic cross-sectional view of a stack 200 in one embodiment. The stack 200 may include a substrate 204 having a structure 208 formed over the substrate 204. In this embodiment, the stack 200 forms a phase change random access memory (PRAM).

[0017] A SiN layer is deposited over the stack 200 by a SiN ALD process (step 108). In this embodiment, the SiN ALD process is a SiN PEALD process. The SiN layer is used to encapsulate PRAM. In this embodiment, the ALD process includes multiple cycles. Each cycle has a step of feeding the stack with a silicon-containing precursor gas (step 112). In this embodiment, the feeding of the stack includes heating the stack to a temperature of about 400° C. In this embodiment, an example of a silicon-containing precursor is diiodosilane (H2SiI2) vapor. Between about 0.5 and 1.5 seconds, the feeding is completed and the flow of the silicon-containing precursor gas is stopped. In this embodiment, the feeding of the stack (step 112) is a plasma-free or plasma-free process. By ensuring that the adsorption of the silicon-containing species is self-limiting, the plasma-free dose can achieve a more conformal saturated half-reaction than the method using plasma.

[0018] After the dosing (step 112) is completed and the flow of the silicon-containing precursor gas is stopped, a first purge (step 116) is provided to purge the silicon-containing precursor gas. In this embodiment, the first purge (step 116) includes flowing a first purge gas. In this embodiment, the first purge gas is N2. In this embodiment, the first purge is a plasma-free or plasma-free process. The process is plasma-free to prevent the formation of a silicon-containing plasma. After about 0.25 seconds, the first purge (step 116) is stopped.

[0019] After the first purge (step 116) is completed, N2 plasma conversion is provided (step 120). In this embodiment, the flow of N2 continues during the first purge (step 116), however, the N2 plasma conversion provides a plasma. In this embodiment, sufficient RF power is provided to convert the N2 gas to N2 plasma. No bias is applied. The stack is exposed to the N2 plasma to provide silicon nitride conversion. By using a hydrogen-free conversion, the film is almost free of NH xbond, which results in a lower WER in dilute HF acid. The N2 plasma conversion is provided for between 3 and 5 seconds. After the N2 plasma conversion (step 120) is completed, the flow of N2 is stopped.

[0020] In this embodiment, after the N2 plasma conversion is completed (step 120) and the flow of N2 gas is stopped, an optional second purge may be provided (step 124). In this embodiment, a second purge gas Ar may be provided. In this embodiment, RF power is continued to maintain the plasma.

[0021] After the second purge is completed (step 124), a hydrogen (H2) plasma conversion is provided (step 128). In this embodiment, H2 gas flows. The H2 gas is converted to H2 gas plasma. In this embodiment, the RF power provided during the H2 plasma conversion (step 128) is higher than the RF power provided during the N2 plasma conversion (step 120). No bias is applied. The stack is exposed to the H2 plasma to hydrogenate the surface, making it more reactive to incoming halide molecules. The H2 plasma conversion is provided for between 1 and 2 seconds. After the H2 plasma conversion (step 128) is completed, the flow of H2 is stopped.

[0022] In this embodiment, after the H2 plasma conversion is completed (step 128) and the flow of H2 gas is stopped, a third purge is provided (step 132). In this embodiment, a third purge gas N2 is provided. In this embodiment, the RF power is stopped to extinguish the plasma. In this embodiment, a third purge is provided for between about 0.25 and 0.75 seconds. After the third purge is completed, the flow of the third purge gas is stopped and the cycle is complete. The cycle returns to Si feeding (step 112) again, and the process performs multiple cycles.

[0023] Figure 2B is a schematic cross-sectional view of the stack 200 after a SiN layer 212 has been conformally deposited on the stack by a SiN ALD process (step 108). The PEALD process provides a highly conformal SiN layer 212.

[0024] The resulting SiN layer 212 can be formed with high throughput. Various embodiments eliminate NH3 from the process and still have a conformal ALD SiN film with the same high growth rate as NH3 plasma, the same low WER as N2 plasma, low particles and no upstream metal corrosion. This is accomplished by inserting an H2 plasma step after the initial N2 plasma step. Note that this is not just a mixture of N2 and H2 in the plasma, but a discrete H2 plasma step with its own optimized flow, pressure and power, which is accomplished in a way that very high flux is achieved with the high growth rate and low WER described above. The short and dilute H2 plasma is expected to be less reactive to the PRAM chalcogenide material and allow the formation of a dual encapsulation layer where a high-flux N2 / H2 ALD SiN process is deposited on top of a protective layer of N2-only ALD SiN. This hybrid stack will have higher flux and better oxidation resistance than an equivalent thickness of N2-only film.

[0025] Various embodiments provide an NH3-free method that meets low WER [<4 angstroms per minute in 200:1 diluted hydrofluoric acid (dHF)] by inserting a second H2 plasma conversion step after the initial N2 plasma conversion step. ] and high growth rate (> / cycle) requirements. The advantage of H2 is that it is less thermally reactive with halide precursors than NH3. This means there is no upstream particle generation in the gas distribution chamber (GDP). In addition, there is no upstream release of volatile hydrogen iodide (HI) and hydrogen chloride (HCl) byproducts that cause metal corrosion. The H2 plasma converts to adding enough reactive ligands to the surface of the growing SiN film to promote the next cycle of silicon precursor adsorption without incorporating excessive -NHx bonds that lead to high wet etch rates.

[0026] What is surprising about this ALD SiN deposition process is that the embodiments can still achieve high conformality and high growth rates with much less reactive nitrogen source than NH3. In addition, the intra-film particle additives have been reduced by 2-3 orders of magnitude. Various embodiments enable optimization of sequential N2 and H2 plasma steps to produce best-in-class film properties and productivity. Elimination of NH3 can also lead to a significant reduction in tool costs, as it is expected that less aggressive metal reduction strategies can be employed and the need for expensive dual plenum designs for separating NH3 from residual silicon precursors is eliminated.

[0027] In some embodiments, the second purge (step 124) can be dispensed with. In such an embodiment, the N2 gas is first flowed. After the N2 gas begins to flow and while the N2 gas continues to flow, the H2 gas flows so that the H2 gas and the N2 gas flow together for a period of time. In another embodiment, the flow of the N2 gas is stopped before the flow of the H2 begins. However, if there is no second purge (step 124), the remaining N2 gas is mixed with the H2 gas in some way. Exempting the second purge (step 124) reduces the processing time and increases the throughput.

[0028] In various embodiments, the silicon-containing precursor is at least one of the silicon halides. In various embodiments, silicon iodide, such as diiodosilane (H2SiI2), triiodosilane (HSiI3), iodosilane (H3SiI) or tetraiodosilane (SiI4), will provide a preferred silicon-containing precursor. In other embodiments, the silicon halides may be silicon chloride or silicon bromide. Preferably, in addition to the silicon halides having at least one halogen connected to at least one silicon atom, the silicon halides may also have one or more hydrogen atoms connected to at least one silicon atom. In other embodiments, the silicon halides may be at least one halogen connected to at least one silicon atom and at least one atom of another element connected to at least one silicon atom. These other silicon-containing precursors may be amino, amide or alkyl groups that can be used as leaving groups, crosslinking promoters or dopants. In various embodiments, the substrate is heated to a temperature of at least 150°C during the feeding (step 112). More specifically, the substrate is heated to a temperature between 200°C and 400°C. The heating provides activation energy to the silicon-containing precursor so that the silicon-containing precursor forms a saturated conformal monolayer.

[0029] Figure 3300 is a schematic diagram of a processing chamber that can be used in one embodiment. In one or more embodiments, the processing chamber 300 includes a gas distribution plate 306 providing a gas inlet and a substrate support 308, which are in a chamber 349 surrounded by a chamber wall 352. In the chamber 349, the stack 200 is positioned above the substrate support 308. The edge ring 309 surrounds the substrate support 308. The gas source 310 is connected to the chamber 349 through the gas distribution plate 306. In this embodiment, the gas source 310 includes a silicon-containing precursor source 311, an N2 gas source 312, and an H2 gas source 316. The support temperature controller 350 is connected to the heater 351 to heat the substrate support 308. The radio frequency (RF) source 330 provides RF power to the upper electrode. In this embodiment, the upper electrode is the gas distribution plate 306. In an exemplary embodiment, 400 hertz (kHz), 13.56 megahertz (MHz) and optional 2MHz, 27MHz power supplies constitute the RF source 330. In this embodiment, the substrate support 308 is grounded. In this embodiment, one generator is provided for each frequency. In other embodiments, the generators may be in separate RF sources, or separate RF generators may be connected to different electrodes. For example, the upper electrode may have inner and outer electrodes connected to different RF sources. Other arrangements of RF sources and electrodes may be used in other embodiments. A controller 335 is controllably connected to the RF source 330, the exhaust pump 320, and the gas source 310. An example of such a chamber is the Striker manufactured by Lam Research Corporation of Fremont, California. TM system.

[0030] Figure 4 4 is a high-level block diagram of a computer system 400 showing a controller 335 suitable for use in implementing the embodiments. The computer system can have many physical forms ranging from integrated circuits, printed circuit boards, and small handheld devices to large supercomputers. The computer system 400 includes one or more processors 402, and may also include an electronic display device 404 (for displaying graphics, text, and other data), a main memory 406 [e.g., a random access memory (RAM)], a storage device 408 (e.g., a hard drive), a removable storage device 410 (e.g., an optical drive), a user interface device 412 (e.g., a keyboard, a touch screen, a keypad, a mouse or other pointing device, etc.), and a communication interface 414 (e.g., a wireless network interface). The communication interface 414 allows software and data to be transmitted between the computer system 400 and external devices via a link. The system may also include a communication infrastructure 416 (e.g., a communication bus, a crossbar, or a network) connected to the aforementioned devices / modules.

[0031] The information transmitted via the communication interface 414 may be in the form of a signal, such as an electronic, electromagnetic, optical or other signal, which can be received by the communication interface 414 via a communication link that carries the signal and can be implemented using wires or cables, optical fibers, telephone lines, cellular phone links, radio frequency links and / or other communication channels. Using such a communication interface, it is expected that one or more processors 402 can receive information from a network during the execution of the above-mentioned method steps, or can output information to the network. In addition, the method implementation can be executed only on a processor, or can be executed on a network such as the Internet with a remote processor that shares a portion of the processing.

[0032] The term "non-transitory computer-readable medium" is generally used to refer to media such as main memory, secondary memory, removable memory, and storage devices such as hard disks, flash memory, disk drive memory, CD-ROMs, and other forms of persistent memory, and should not be construed to encompass transient subject matter such as carrier waves or signals. Examples of computer code include machine code (such as that generated by a compiler), and files containing high-level code that are executed by a computer using an interpreter. A computer-readable medium may also be computer code transmitted by a computer data signal embodied in a carrier wave and representing a sequence of instructions that can be executed by a processor.

[0033] Although the present disclosure has been described according to several preferred embodiments, there are still changes, modifications, substitutions and various substitution equivalents that fall within the scope of the present disclosure. It should also be noted that there are many alternatives to the methods and devices of the present disclosure. Therefore, it is intended that the following appended claims be interpreted as including all such changes, modifications, substitutions and various substitution equivalents that fall within the true spirit and scope of the present disclosure.

Claims

1. A method for depositing a silicon nitride layer on a stack, the method comprising providing an atomic layer deposition comprising a plurality of cycles, wherein each cycle comprises: feeding the stack with a silicon halide-containing precursor by providing a silicon halide-containing precursor gas to form a silicon precursor deposit; providing N2 plasma conversion after feeding the stack to convert the silicon precursor deposit into a silicon nitride deposit; as well as An H 2 plasma conversion is provided after providing the N 2 plasma conversion to form a hydrogenated surface of the silicon nitride deposit, wherein subsequent charging of the stack deposits a silicon precursor on the hydrogenated surface.

2. The method of claim 1, wherein: During said charging of said stack, said stack is maintained at a temperature in the range of 200°C to 400°C.

3. The method of claim 1, wherein: The charge to the stack is plasma-free.

4. The method of claim 1, wherein: Each cycle contains no NH3.

5. The method of claim 1, wherein: Providing the N2 plasma conversion comprises: flowing N2 gas that does not contain hydrogen; Converting the N2 gas into a hydrogen-free N2 plasma; exposing the stack to the N 2 plasma; and Stop the flow of the N2 gas.

6. The method of claim 5, wherein: Providing the H2 plasma conversion comprises: allowing H2 gas to flow; converting the H2 gas into H2 plasma; exposing the stack to the H 2 plasma; and Stop the flow of the H2 gas.

7. The method of claim 1, wherein: The silicon halide-containing precursor is silicon iodide.

8. The method of claim 1, wherein: The silicon halide-containing precursor is silicon chloride or silicon bromide.

9. The method of claim 1, further comprising purging the H2 plasma formed by the H2 plasma conversion after providing the H2 plasma conversion and before charging the stack.

10. The method of claim 1, further comprising purging the silicon-containing precursor gas after charging the stack and before providing the N2 plasma conversion.

11. An apparatus for depositing silicon nitride atomic layer deposition on a stack, the apparatus comprising: a processing chamber; a substrate support within the processing chamber; a gas inlet for providing gas into the processing chamber; a gas source for providing the gas to the gas inlet, wherein the gas source comprises: a silicon-containing precursor source; N2 gas source; and H2 gas source; an exhaust pump for pumping gas from the processing chamber; an electrode for providing RF power in the processing chamber; at least one power source for supplying power to the electrodes; and a controller controllably connected to the gas source and the at least one power source, wherein the controller comprises: at least one processor; and A computer readable medium comprising: Computer code for depositing by atomically layer depositing a silicon nitride layer on a stack by providing a plurality of cycles, wherein each cycle of the plurality of cycles comprises: feeding the stack with a silicon-containing halide precursor by providing a silicon-containing halide precursor gas from the silicon-containing precursor source to form a silicon precursor deposit; providing N2 plasma conversion after feeding the stack to convert the silicon precursor deposit into a silicon nitride deposit; and An H 2 plasma conversion is provided after providing the N 2 plasma conversion to form a hydrogenated surface of the silicon nitride deposit, wherein subsequent charging of the stack deposits a silicon precursor on the hydrogenated surface.

12. The apparatus of claim 11, further comprising a heater for heating the substrate support.

13. The apparatus of claim 12, wherein: The computer readable medium also includes computer code for heating the substrate support to a temperature in the range of 200°C to 400°C.

14. The device of claim 11, wherein: The gas source is NH3-free.

15. The apparatus of claim 11, wherein: Providing the N2 plasma conversion comprises: flowing N2 gas that does not contain hydrogen; Converting the N2 gas into a hydrogen-free N2 plasma; exposing the stack to the N 2 plasma; and Stop the flow of the N2 gas.

16. The device of claim 15, wherein: Providing the H2 plasma conversion comprises: allowing H2 gas to flow; converting the H2 gas into H2 plasma; exposing the stack to the H 2 plasma; and Stop the flow of the H2 gas.

17. The apparatus of claim 11, wherein: Each cycle of the plurality of cycles comprises: After charging the stack and before providing the N2 plasma conversion, purging the silicon-containing precursor gas; and After providing the H 2 plasma conversion and before charging the stack, the H 2 plasma is purged.

Images