TW466593B - CVD TiN plug formation from titanium halide precursors - Google Patents

Abstract

A method of depositing high quality titanium nitride (TiN) films and filling small contacts having high aspect ratio features using TiN. The method uses a CVD process with titanium tetraiodide (TiI4) as a precursor material. The method allows TiN films with a thickness greater than about 0.3 μm to be deposited without cracking. For sufficiently high TiN deposition rates and sufficiently low TiN resistivities the preferred process temperature is at least about 500 DEG C. The method varies process pressure to obtain a seamless TiN plug fill in high aspect ratio structures.

Description

4 6 659 3 V. Description of the invention (υ Name Cangming Scope The invention relates to the formation of printed circuit boards, and in particular the chemical vapor deposition of hafnium halide titanium nitride films. BACKGROUND OF THE INVENTION Integrated circuit boards are provided in Signal transmission path in electronic design. The integrated circuit board (ic) in the design is composed of many active transistors contained in the silicon substrate of the semiconductor substrate. To increase the IC capacity, one of the The active transistor and another active transistor in the Shi Xi bottom layer of the substrate make a large number of interactive connections with the metal "network cable". The circuit board is interactively connected through the hole, track or groove cut into the substrate, which is generally known as Intermetallic connection. In fact, the special metal connection point that is in contact with the bottom layer of silicon is known as the contact point. The rest of the hole, channel or trench is filled with a conductive substance called a contact plunger. As the density of transistors continues to increase, the formation of For higher-level integrated circuit boards, the diameter of the contact plunger must be reduced to allow for increased amounts of interconnections, multilayer metallization, and higher The aspect ratio aspect. The diameter greater than about 0.16 microns is typically filled in the following manner. First, a titanium (Ti) liner of about 100 angstroms is deposited by CVD or PVD = the Ti layer is reinforced to the underlying silicon substrate for electrical connection Point. Next, a titanium nitride (T i N) liner of about 500 angstroms is deposited on the Ti layer. The deposition can be either CVD or PVD, and it is better to have low pressure CVD (LPCVD) because only LPC VD will provide conformality, which is defined as the ability to indeed reproduce the surface morphology of the underlying substrate, which must cover the bottom and side walls of a submicron structure with a high aspect ratio. The Ti layer functions as a metal diffusion barrier To avoid the corrosion and etching of fluorine (F) during the subsequent deposition of T i from tungsten (ff) from WF6. T1 N also has an adhesion layer that becomes W

4 6 659 3 V. Description of the invention (2) Function, because W will not adhere to metal oxides. "Although T i N provides an excellent contact barrier, T i N must have the ability to become an effective barrier. Thickness (about 50 0 Angstroms). If the T i N thickness is less than about 500 Angstroms, the T i metal will diffuse into the silicon. The remainder was then filled with CVD deposited W. W is used because of its low resistivity and reliability when forming a contact plunger. The W layer provides a low-resistance region, which is important in current IC conductance. The contact plunger surface is then etched or polished. The resulting planarized surface is required for excellent metal interaction and is therefore required for excellent IC functionality. The transistor density continues to increase, and the components continue to become smaller, that is, having a diameter of 0.25 microns or less. The diameter of the contact plunger must be reduced to allow for an increased number of interconnections. However, with respect to the way having a diameter of less than about 0.16 micrometers, the resistance of the metal layer of the contact plunger is dominated by the T i N diffusion barrier layer. Because the T i N barrier layer must be maintained at about 500 angstroms in terms of its sound performance as a diffusion barrier layer, the contact plunger portion filled with W will then be reduced. For example, a structure with a diameter of 0.15 micrometers may have a W film or only about 300 Angstroms in the center of the plunger. Therefore, the effective plunger resistance will be dominated by a higher resistivity TiN, More importantly, the interface resistance between TiN and W layers. Subsequent filling of the contact plunger with W will provide an additional procedural step of the overall resistivity of the contact plunger without significant effects. In addition, the formation of the IC can be eliminated Processing steps, and filling the track with only the T i N contact plunger can increase the manufacturing efficiency, even more than TiN and W. Therefore, it is necessary to form the TiN contact plunger by CVD and eliminate the contact in the IC formation function. Method to point the W layer of the plunger. However, the T i N film deposited by CVD has a relatively high stress. The film with high stress has a high strength internal distribution force or a composition of forces to resist the body of the film

4 6 659 3 V. Description of the invention (3) The product or shape changes when the membrane is subjected to external force. High stress limits the maximum film thickness that can be deposited. The maximum thickness of a TiN film typically deposited by CVD is about 800 Angstroms, which is larger than the well-known first layer oxide. TiN films with a thickness exceeding about 800 Angstroms begin to crack due to internal stresses in the films. The discontinuous microcracking that occurs on the surface of the TiN film and is defined as a surface material is large enough to increase the resistance of the film, thus causing non-optimal performance of the IC. Although the absolute thickness of the W layer can be changed according to the size of the track to be filled, the relative thickness is about 80% of the track diameter. This is because the deposited film not only fills the volume with the contact plunger, but also must be filled in the "depression" of the contact plunger. The "depression" is defined as the cavitation in T i N, which It is formed during the filling of the track to deposit more T i N on the tip of the plunger to eliminate pits and create a sealing layer. Therefore, for a component of 0.25 micrometers, a TiN film having a thickness of 2000 Angstroms (0.8 x 2500 Angstroms) is required. Regarding good plunger filling, these film thicknesses with continuous, thorough shape retention and seamlessness are also important. A conformal film is a film that does reproduce the surface morphology of the underlying substrate. Seamless film is a non-cracking film. U.S. Patent Application Serial No. 08/964, 532, assigned to Tokyo Electron Limited and incorporated herein by reference in its entirety, discloses a track filled with a TiN plunger and a TiN layer seal in a substrate. The CVD method. The existing TiN deposition using the TiCl4 precursor CVD method does not have these features. With regard to the TiCl4 precursor, when the thickness exceeds about 5 0-8 0 0 Angstroms, the film will crack uniformly. Cracking is unacceptable because it prevents the film from adhering to the underlying layer, causing the film to fall π and thus endangering subsequent processing. Cracking also increases the expected resistivity of the plunger.

O: \ 64 \ 64000.PTD Page 7 4 6 659 3 V. Description of the invention ¢ 4) Therefore, there is a need for a method of filling a high aspect ratio with a high-quality conformal contact plunger of Τ N film, Does not crack. This film eliminates W deposition steps and therefore reduces the number of process steps required to fill the contacts. This can show a significant reduction in design and manufacturing costs. SUMMARY OF THE INVENTION Finally and in accordance with the principles of the present invention, it is disclosed a method for eliminating the tungsten (W) deposition step by using a high aspect ratio of the T i N plunger. The T i N plunger was deposited with C V D from titanium iodide (T i I) precursor. The preferred T i I precursor is titanium tetraiodide (T i 14) and is deposited with thermal C V D. The present invention also focuses on a method of completely filling a TiN layer deposited in CVD with a high aspect ratio having a diameter of less than about 0.16 microns. The present invention also focuses on a method of contacting a plunger formed in a 1C channel by CVD of a TiN layer provided by a Ti 14 precursor. The track is a track having a high aspect ratio with a diameter of less than about 0.16 microns. The T i N film filling the contact plunger according to the present invention has a morphology of 100% in the lower layer. A film with 100% conformality is advantageous because it does reproduce the surface morphology of the underlying substrate, which allows for the best IC function. Therefore, this method is suitable for completely filling high aspect ratio parts. Another benefit of this method is that it eliminates the separate processing steps for depositing W, while saving time and money. This method also eliminates the problem of sticking the W layer to T i N. These and other objects and advantages of the present invention will be readily apparent from the accompanying drawings and descriptions thereof. BRIEF DESCRIPTION OF THE FIGURES Figure 1 is a diagram of a chemical vapor deposition (CVD) apparatus.

O: \ 64 \ 64000.PTD Page 8 Correction 4 6 65 9 3 Correction No. 89107861 V. #Explanation IB) 圊 2 is the stress in a titanium nitride (TiN) film based on titanium halide (TiX). Its graphics. FIG. 3 is a SEM photograph of a TiN film mainly composed of titanium iodide (Til4). 4A and 4B are transmission electron micrographs of a T i N film dominated by T i halides. Figure 5 is a SEM photograph of a Ti: N deposited 10: 1 aspect ratio structure with CVD deposited at a pressure of 1.5 Torr. Figure 6 is a SEM photograph of a Ti: N deposited 10: 1 aspect ratio structure with CVD deposited at 1.0 Torr. Fig. 7 is a graph of contact resistance data filled with the present invention and T i, T i N and a tungsten plunger. Brief description of the symbols 10 indicates a chemical vapor deposition (C VD) system; 11 indicates a CVD reaction chamber; 12 indicates a precursor transport system; 13 indicates a precursor gas source; 1 4 indicates a gas outlet; 15 indicates a Metering system; 16 means a gas inlet; 20 means a vacuum chamber; 21 means a vacuum pressure chamber wall; 22 means a substrate support or receiver; 23 means a semiconductor wafer; 24 means a vacuum pump; 25 means a Inlet gas source;

O: \ 64 \ 64000.ptc Page 9 2001.08. 23. 009 4 6 65 9 3 Case No. 89107861_qp Year of the Moon Rev. V. Description of Invention (6) 26 indicates a gas source; 27 indicates an inert gas source; 28 Indicates a spray head; 30 indicates a sealed evaporation gas; 31 indicates a cylindrical evaporation chamber; 32 indicates a vertical orientation axis; 33 indicates a cylindrical wall; 34 indicates an inner surface; 35 indicates a flat circular closed bottom 36 indicates a lid; 37 indicates a flange ring; 38 indicates a high-temperature tolerance vacuum-compatible metal seal; 39 indicates a source of carrier gas; 40 indicates a precursor block; 4 1 indicates a vacuum pump; 42 Indicates a surface; 4 4 indicates a heater; 45 indicates a single controlled heater; 46 indicates an annular recessed air gap; 4 7 indicates a concentric outer aluminum wall or can; 48 indicates a silicone Foam insulation ring layer; 50 represents a transfer tube; 51 represents a baffle; 52 represents a round hole; 53 represents a control valve;

O: \ 64 \ 64000.ptc Page 9a 2001.08. 23.010 4 6 89107861 V. Description of the invention (7) 54 represents a shut-off valve; 55, 56, 57 and 58 represent a pressure sensor: 60 represents a control is' And g I represents an immutable memory. _Detailed description In chemical vapor deposition (C VD) 'or using thermal energy or electrical energy to activate gas precursors. Once activated, the 'gas precursor will chemically react to form a film. The preferred C VD method is exemplified in Figure 1, and the device and APPARATUS AND METHOD FOR DELIVERY OF VAPOR FROM SOLID SOURCES TO A CVD CHAMBER) is the subject of a co-trial application and the entire text of which is incorporated herein by reference. The chemical vapor deposition (CVD) system 10 includes a CVD reaction chamber 11 and a precursor transfer system 12. The reaction is carried out in a reaction chamber to convert a halogenated ammonium compound, for example, Til, into a barrier film of TiN. The precursor delivery system 12 includes a precursor gas source having a gas outlet 14

〇: \ 64 \ 64〇〇〇, ptc page 9b 2001. 08. 23.011 j η 65 9 3 case number 89107861 11] 0 years > ^ said amendment _ V. Description of the invention (8) 1 3 It is transferred into the CVD reaction chamber U via a metering system 15 having a gas inlet 16. Source 13 generates precursor gases from individual T i I compounds (preferably T i 14), for example, T 1 I vapor. The compound is in a solid state at standard temperature and pressure. The precursor source is maintained (preferably) in a manner that controls heating at a temperature that will produce the expected precursor vapor pressure. It is preferred that the vapor pressure is such that it is sufficient to transfer precursor vapor to the reaction chamber 11 without using a carrier gas. The metering system 15 maintains the precursor gas from the source 13 and the vapour flows into the reaction chamber 11 at a rate sufficient to be maintained in the reaction chamber and capable of performing the CVD method economically. The reaction chamber 11 is a generally known CVD reactor and includes a vacuum chamber 20 bounded by a vacuum pressure chamber wall 21. A support substrate, such as a substrate support or acceptor 22 of a semiconductor wafer 23, is placed in the chamber 20. The chamber 20 is maintained under a vacuum suitable for the CVD reaction performance of depositing a radon, such as a TiN barrier layer, on the semiconductor wafer substrate 23. The preferred pressure range is 0.2-5.0 Torr. In order to control the vacuum pump 24 and the delivery system 12 and also include a reducing gas source 26 such as argon (H2), nitrogen (N2) or ammonia (NH3) used in carrying out the T i reduction reaction, and such as argon (Ar) Or an inert gas source 27 such as a gas such as helium (He), the inlet gas source 25, and the like are maintained in a vacuum. The gas from the source 25 is introduced into the chamber 20 through an ejection head 28 located at one end of the chamber 20 opposite the substrate 23 (usually parallel to and toward the substrate 23). The precursor gas source 13 includes a sealed evaporator 30 including a cylindrical evaporation chamber 31 having a vertical orientation axis 32. A cylindrical wall 33 made of a high temperature tolerance and a non-corrosive substance (such as alloy INCONEL 600) is the boundary of the chamber 31. The inner surface 34 is highly polished and smooth. The wall 33 has a flat circle

O: \ 64 \ 64000.ptc Page 10 2001.08. 23.012 4 6 65 9 | No. 89107861 «^ leather month y said correction -------- '7 V. Description of the invention (9) Shaped closed bottom 3 5 and the opening top 'seal it with a lid 36 of the same thermal tolerance and non-corrosive substance as the wall 33. The outlet 14 of the source 13 is located in the lid 36. When using high temperature, 'if it is T i 14 or T a B r 5', the cap 3 6 is sealed into a flange ring 37, and a vacuum-compatible metal seal 3 8 (such as a C-shaped nickel tube) HELICOFLEX, which consists of an INCONEL coil spring, integrates a flange ring to the top of the wall 33. With regard to TaCl5 and TaFs, the lid can be sealed with a well-known elastomer 0-ring seal 38. Connected to the container 31 via the lid 36 is a source 39 of carrier gas, preferably an inert gas (such as He or Ar). The source 13 includes a precursor material at the bottom of the container 31, such as T i I 'T i 14 Preferably, it is charged into the container 31 in a solid state at a standard temperature and pressure. Til vapor is charged into the container 31 in such a manner that the Til solid block is sealed therein. Supply Tii as a precursor block 40 and supply it and place it on the bottom of the container 31, and heat it here to heat it to a liquid state. As long as the generated vapor pressure is within an acceptable range. Where the precursor block 40 is liquid 'then the vapor will be above the liquid block 40 height. Because the wall 33 is a vertical cylinder, if the > 111 block 40 is a liquid, its surface area will remain fixed regardless of the consumption height of T i I. Gas and steam can be evacuated from the container 31 by a vacuum pump 41. The transport system 12 is not limited to the direct delivery of precursors 40. Two alternatives can be used to transfer precursors 40 and gas from the source 载 can be introduced into the container 31. Zhao =: Such a gas can be a gas (Η) Or an inert gas such as helium or argon Ur). Where the plant gas is used, it can be introduced into the container # ί distributed on the entire top surface of the precursor block 40, or can be! • 、 桃 ^ = 益 3 1 中, 'to facilitate the passage of the precursor block 4 〇 Penetrate from container 3 1 bottom 3 5 '> to reach the precursor block 40 and expose the carrier with the maximum surface area

11 pages

2001.08. 23.013 丄 6 659 each number 89] Chi Ying and the year p. 4th article Zheng _ V. Description of the invention (10) j There is another alternative way of gas L is to evaporate the liquid in the container 31. These alternatives, however, add undesirable particles and do not provide a controlled transfer speed achieved by direct delivery of precursors, i.e. without carrier gas transfer. Therefore, it is better to use the direct delivery precursor. In order to maintain the temperature of the precursor 40 in the f container 31, the bottom 3 5 of the wall 3 3 is maintained in heat transfer with the heater 44 to maintain the precursor 40 at a controlled temperature so as to be greater than its solubility. The point is better, it will produce a vapor pressure greater than about 3 Torr in the absence of a carrier gas (ie, a direct delivery system), and will produce a lower vapor pressure when using a carrier gas, as about! Care. The exact vapor pressure depends on other variations, such as the amount of carrier gas and the surface area of the substrate 23. At Ti I. In direct transfer systems, especially Ti, such temperatures are in the range of about 180 ° C to 190 ° C. The temperature should not be so high as to cause the gas of the ejection head 28 t to react prematurely or otherwise before contacting the wafer 23. For the purpose of the embodiment, a temperature of 180 ° C is taken as the control temperature for heating the container 31, 35. This temperature is suitable to produce the expected Roh pressure with the Tii4 precursor. This temperature is provided at the bottom 35 of the container 31 to prevent the precursors from condensing on the wall 33 and the lid 36 of the container 31, and the lid is maintained at a lower temperature than the bottom of the wall 33 by a controlled heater 45 in thermal contact with the outside of the lid 36 The heater 44 of 35 is more than two temperatures, for example, i90 ° C. An air gap 4 6 is received around the side of the chamber wall 33 with an annular recessed air space contained between the chamber wall 33 and a concentric outer aluminum wall or tank 4 7. The tank 47 is then surrounded by a ring-shaped layer of silicone foam insulation 48. In the temperature range between 180 ° C and 190. (: And the expected embodiment t where the pressure is greater than about 3 Torr (preferably greater than 5 Torr), the temperature maintained in the arrangement will maintain the steam in a container bounded by the lid 36, the side of the wall 33, and the surface 40 of the precursor block 40

O: \ 64 \ 64000.ptc 2001.08. 23.014 Page 12 4 6 6 5 9 3 V. Description of the invention (9) Within 31 volumes. The temperature suitable for maintaining the desired pressure will vary depending on the precursor material that mainly contains the compound that is the group or the iS group. The vapor flow metering system 15 includes a transfer tube 50 of at least 1/2 inch diameter or at least 10 mm inner diameter, and is preferably larger so as to provide no pressure drop at the expected flow rate, It has at least about 2 to 40 standard cubic centimeters per minute. The transfer tube 50 extends from the precursor gas source i3 (connected to the outlet 14 at its upstream end) to the reaction chamber (connected to the inlet 16 at its downstream end). It is also preferable to heat the entire transfer pipe length from the evaporator outlet 14 to the reactor inlet 16 and the ejection head 28 of the reactor chamber 20 at an evaporation temperature greater than that of the precursor substance 40, for example, to 19 5 ° C. A baffle 51 located in the center of the circular hole 52 is provided in the transfer tube 50 and preferably has a diameter of about 0.089 inches. Use the control valve 5 3 to adjust the pressure drop from Table 1 5 6 to Table 2 5 7. The pressure drop through the circular hole 5 2 and into the reaction chamber 11 after the control valve 53 is greater than about 10 mTorr and will be proportional to the flow velocity. A shut-off valve 5 4 is provided in a line 50 between the outlet 14 of the evaporator 13 and the control valve 53 to close the evaporator 13 container 3 1. A pressure sensor 5 5-5 8 is provided in the system 10 to provide a message to the controller 60 for controlling the system 10, which includes a precursor for controlling the self-conveying system 15 to the chamber 20 of the CVD reaction chamber 11 Gas flow speed. The pressure sensor includes a sensor 5 5 connected to a transfer pipe 50 between the outlet 14 of the evaporator 13 and the closing valve 54, to monitor the pressure in the evaporation vessel 31. Connect the pressure sensor 5 6 to the transfer pipe 50 between the control valve 53 and the baffle 51 to monitor the upstream pressure of the circular hole 52, and connect the pressure sensor 57 to the baffle 51 and A transfer tube 50 between the reactor inlets 16 to monitor downstream of the round hole 52

O: \ 64 \ 64000.PTD Page 13 466593 V. Description of the invention (ίο) Pressure. Another pressure sensor 58 is connected to the chamber 20 of the reaction chamber to monitor the pressure in the CVD chamber 20. The controller 60 responding to the pressure sensed by the sensors 5 5-5 8 reaches the precursor vapor flow entering the CVD chamber 2 1 of the reaction chamber 1 1, especially the inductor 56 and the pressure drop that determines the pressure drop across the circular hole 57. In the condition that the precursor vapor flow through the circular hole 52 is not resistant to flow, the actual pressure vapor flow through the transfer tube 52 is the effect of the pressure monitored by the pressure sensors 56 and 57, and can be freely rounded. The ratio of the pressure measured by the sensor 56 on the upstream side of the hole 5 2 to the pressure measured by the sensor 57 on the downstream side of the circular hole 5 2 is determined. In the condition that the precursor vapor flow through the circular hole 5 2 is resistant to flow, the actual pressure vapor flow through the transfer tube 52 is only the effect of the pressure monitored by the pressure sensor 57. In any instance, the controller 60, described in terms of processing conditions, can determine the presence or absence of anti-flow. When measuring with the controller 60 *, the flow velocity of the precursor gas can be measured via the calculated controller 60. Preferably, the true flow velocity of the precursor gas is calculated from the corrected flow rate data stored in the lookup table or extended range table stored in the non-volatile memory 6 1 obtained by the controller 60. When measuring the true flow velocity of precursor gas, use the closed coil feedback control of one or several variable circular hole control valve 53, reduce the CVD chamber pressure of inert gas from sources 26, 27 through the evacuation pump 24 or control Or, by controlling the vapor pressure of the precursor gas in the chamber 31 controlled by the heaters 44, 45, the desired flow velocity can be maintained. Ti 14 with a purity of 99.9% is widely used. It is a purple-black solid at room temperature (18 ° C-2 2 ° C), has a melting point of about 150 ° C, and is humidity sensitive

O: \ 64 \ 64000.PTD P.14 4 6 G 5 Q ^ _. 0 Case number 89107861 彳. Year corpse month day correction _ 'y V. Description of invention (13). According to the display of Fig. 1, the solid Ti 14 precursor substance 40 is sealed in a cylindrical anticorrosive metal container 31 which maximizes the surface area of the precursor substance. The vapor from or Ti 14 is transferred to the reaction chamber 11 by a highly conductive transfer system, that is, no carrier gas is used. The reaction chamber 11 is heated to a temperature of at least about 1 ° C to avoid vapor concentration or by-product deposition. In order to precisely control the thickness of the deposited film, it is desirable not to use a carrier gas. In order to obtain a sufficient vapor pressure greater than about 3 Torr (preferably greater than 5 Torr), the solid Ti 14 precursor is heated to a temperature range of about 180 ° C-190 ° C to complete the process. The Ti 14 vapor is controlled to be sent directly into the reaction chamber 11. This pressure needs to be maintained to maintain a fixed pressure drop across the circular hole defined in the south conductive transmission system, while at the same time transmitting up to about 50 seem of Til4 precursor to the processing chamber under the operation of about 0.1-2.0 Torr. The temperature to obtain this pressure is 185 ° C in the case of Ti 14. Utilizing parallel plate RF discharge, the electrode activated here is a gas transfer jet, and the receiver or step of the wafer or substrate 23 is RF grounded. The Ti 14 vapor is combined with an ammonia (NH3) -containing process gas on a substrate that has been heated to a temperature between about 300 ° C and 500 ° C. Argon (Ar), nitrogen (N2), hydrogen (H2), and helium (He) can be used as a process gas or alone or in combination. The requirements for TiN film deposition from the precursor of Ti 14 are as follows. The deposition temperature must be lower than about 6 50 ° C to maintain the integrity of the underlying material on the substrate or wafer 2 3. The deposition rate must be greater than about 300 Angstroms per minute to provide an acceptable output. The chamber pressure can be changed to obtain the desired film thickness. For example, with NH3 of 3000 seem and Til4 of 25 seem (without carrier gas) at a wafer temperature of about 5 50 ° C, a pressure of about 2.0 Torr will

O: \ 64 \ 64000.ptc Page 15 2001.08. 27.017 4 6 659 3 V. Description of the invention (12) The deposition rate in the range of about 50 to 60 angstroms per minute is obtained. In these same cases, a deposition rate of about Angstroms per minute will be obtained with a force of about U Torr and a deposition rate of about 15 Angstroms per minute at a pressure of about 1.0 Torr must be less than about 1x1 〇ia dyne, square centimeters and having a resistivity greater than about 8 八 ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ ◊ 貘 电阻 The resistivity to less than about 25Q micro-ohm hi A 2 ^. The film should exhibit 100% conformality in a 鬲 aspect ratio structure. The high aspect ratio structures used herein have greater than about 8.0 and at most: ί includes structures having an aspect ratio of 10_0 or even higher. ># 有 ί, 孔 或 沟 etc. Subsequent deposition of films, such as aluminum (a υ film is not Λ). There must be a small amount of impurities in the film, ideally a small amount, such as He e ° Λ Finally, a commercially reasonable processing gas must be used, such as He , Ar, η2, and the heart. Ϋ Δ is the best range for CVD ... film deposition, in brackets ^ \ conditions. Slin stands for standard liters per minute and W / cm2 疋 female thousand cubic centimeters Watt.

O: \ 64 \ 64000.PTD

Table 2 400 ° C-650 ° C (55 ° r) V is only +1 speed ---- trace 300 angstroms / minute (300 angstroms of waist JJ 1-15X109 dyne / cm2 ", Dyne / Cm²) Non-cracked film thickness> 4000 Angstroms sg — 85 * 400 microohm centimeters (< 200 microohm centimeters) 0.5-5 atomic% (< 2 atomic%) required processing gas ...- ----_ ~ 5-40 seem Til4 (25 seem T1I4) Page 16 466593 V. Description of the invention (13) According to the present invention, the TiN film deposited by the CVD of Ti 14 precursor meets all the expected standards. Cracking occurred on Shenji A 1 layer, and showed 100% shape retention in parts with aspect ratios even greater than 10.0. Higher deposition temperature results in lower resistivity, lower residual iodine concentration, and higher deposition rate without sacrificing shape retention and cracking limits. A TiN film deposited using the Ti4 precursor according to the present invention has a larger cracking limit than a TiN film deposited using other tantalum halide precursors, such as TiC14. Figure 2 compares the cracking stress of TiN films with Ti 14 and 1 ^ (: 14 as the main component. The circle represents the T i N film deposited from the T i C 14 precursor at 5 80 ° C. The triangle represents 5 TiN film deposited from Ti 14 precursor at 50 ° C. Arrows indicate the points where cracking was found in the TiCl4 film. Rapidly reduced film stress (measured in dyne / cm2) and slightly reduced film thickness (in angstroms) (Measurement) corresponds to spreading cracking. According to the display of FIG. 2, the film based on TiC14 exhibits a rapidly decreasing film stress for a thickness of less than about 100 Angstroms. At a thickness of more than about 600 Angstroms Spreading cracking was observed on scanning electron microscopy (SEM) of a T i N film predominantly deposited with T i C 14 on the film. T i having a thickness greater than 4 0 0 Angstroms was observed. No evidence of cracking was found at any of the joints in the 14 film. Figure 3 Scanning electron micrograph of a Ti 14 film mainly composed of Ti 14 without cracking and deposited in accordance with the present invention and filled with trenches (SE M). Deposited on the silicon dioxide layer 62-2 0 0 Angstrom T i N film 60. Compared with the ΊΊN film deposited with the TiCl4 precursor, the use of Ti 14 For the larger cracking limit of the T i N film deposited by the precursor, one reason is that

O: \ 64 \ 64000.PTD Page 17 4 6 6 5 9 3 V. Description of the invention (14) The particle size is essentially small. Compared with a film composed of larger T i N particles, a film composed of a matrix of small particles may exhibit crack propagation. Fig. 4 shows the electron transfer of Ti Ni film deposited by CVD according to the present invention using the precursors based on T i 14 and i C 丨 4 deposited at 5 50 ° C and 5 80 ° C, respectively. Microphotograph (TEM). According to the display in Fig. 4, the film mainly composed of D14 has substantially smaller particles. The possible reasons for the smaller particles to invade the T i N film, which is dominated by T i 14 The processing pressure in the structure of the aspect ratio component is related. These conditions are as follows: a temperature of 5 50 ° C, a flow of N H3 of 3 s 1 m, and a flow of 1 ^ 14 of 2 5 sccro. Fig. 5 is an SEM photograph of a TiN filled 10: 1 aspect ratio structure deposited by CVD at a pressure of 1.5 Torr. FIG. 6 is a SEM photograph of a Ti: N filled 10: 1 aspect ratio structure deposited by CVD at 1.0 Torr. In a high aspect ratio structure, it is necessary to form a good plunger without an M-key hole in a very highly saturated method. When the T i N formed in the track contacting the plunger does not completely fill the track, a "keyhole effect" will occur, leaving an area called the keyhole without T i N. When the track has a substantially vertical wall, That is, the effect occurs when the wall is substantially perpendicular to the base. In channels with inclined walls, the keyhole effect can be eliminated. It was found that these structures need to be filled with a processing pressure reduced to less than about 1.5 Torr while maintaining about 3 s 1 m NH3 and 2 5 seem Til4 gas flow. Good plunger filling may be at a lower flow rate, but it will only sacrifice the deposition rate. Conversely, higher deposition rates may have a higher rate than in Table 2. The stated flow rates are high. Figure 7 shows the contact resistance data of the electrical test structure. According to the present invention,

O: \ 64 \ 64000.PTD Page 18 466593 V. Description of the invention (15) ^ Ti and TiN plugs (filled circles) from Ti I * precursor deposition and the typical Ti / Tin with Ti I precursor deposition and Comparison of W plunger filled plungers (hollow circles). The contact size is 0 _ 3 microns and the aspect ratio is 4: 1. According to the display in Fig. 7, replacing general Ti, TiN and W layers with τ soil and TiN to fill the contact plunger will result in the same contact resistance. This implies that reducing the metal interface more compensates for the overall increase in material resistance from W to T, N. Therefore, the plunger filling method with T i * is a feasible method to replace the current W plunger in the formation of the IC design. The contact resistance using the present invention is the same as a standard W plunger. It is worth noting that, overall, the overall resistance of TiN is 15 to 20 times greater than W, which has a resistivity of 10 micro-ohm centimeters, which emphasizes the advantages of a single method and reduces the number of interfaces in the plunger. The films deposited by the method of the present invention show the importance of characterizing the formation of IC. The film has a resistivity sufficiently low for low cross-link resistance, a deposition rate sufficient for output considerations (greater than 100 Angstroms / minute), and a film having a thickness greater than 0.3 micron without cracking. The method of the present invention can be used to fill components with a diameter as small as 0.15 microns and an aspect ratio greater than 10: 1. Of course, the specific embodiments of the present invention shown and described in the application are only specific embodiments for the inventors skilled in the art as a reference, and are not intended to be limiting. For example, Ta films can be deposited by PECVD, and T a Nx films can be deposited by thermal CVD, PECVD, and plasma treatment. For example, PECVD OF Ta FILMS FROM TANTALUM HALIDE PRECURSORS for Ta thin films from halogenated precursors. THERMAL CVD OF TaN FILMS FROM TANTALUM HALIDE PRECURSORS

O: \ 64 \ 64000.PTD Page 19 466593 V. Description of the invention (16) CVD (PLASMA ENHANCED CVD OF TaN FILMS FROM TANTALUM HAUDE PRECURSORS) and electrothermal treatment thermal CVD (PLASMA) TREATED THERMAL CVD OF TaN FILMS FROM TANTALUM HALIDE PRECURSORS), all of which were invented and designated by Tokyo Electronics Co., Ltd. by Hautala and Westendorp, and are jointly reviewed applications filed on the same day as the application for this patent, and in particular It is incorporated herein by reference in its entirety. According to another embodiment, T i N from a halide neoprene precursor can be deposited by CVD for congestion formation, as disclosed in the co-trial application under the title, which was invented and designated by Tokyo Electronics Co., Ltd. filed an application on the same day as this patent application, and the entire contents thereof are specifically incorporated herein by reference. Furthermore, T a / T a Nx bilayer can be deposited by CVD, and TaNx can be used for plunger filling according to the present invention, such as the integration of CVD Ta and TaNx films from tantalum halide precursors (INTEGRATION OF CVD Ta AND TaNx FILMS FROM TANTALUM HALIDE PRECURSORS) and CVD TaNx PLUG FORMATION FROM TANTALUM HALIDE PRECURSORS order, both of which were invented by Hautala and Westendorp and designated by Tokyo Electron Limited The company co-examined the application with the patent application, and specifically incorporated the entire text of this application for reference. Therefore, various changes, improvements, or replacements of these specific embodiments can be made or sub-classified without departing from the spirit of the present invention and the scope of the following patent applications.

O: \ 64 \ 64000.PTD Page 20

Claims (1)

4 6 659 3 Case No. 89107861 Rev. 6 of the 4th year of the 'Amendment Patent Application Scope 1. A method of filling a component in a substrate having a titanium (Ti) film deposited therein, which includes heating a T i _ compound precursor to heat A temperature sufficient to vaporize the precursor to provide the precursor vapor into a reaction chamber containing the substrate, and then combine the vapor with a nitrogen-containing processing gas to deposit the TiN by chemical vapor deposition (CVD) Titanium halide (TiN). 2. The method according to item 1 of the scope of patent application, wherein the component has an aspect ratio greater than 8.0. Wherein the part has a large precursor to make the halogenated condensate therein, the substrate is heated therein to about 5-40 seem among which about 1 0 0-6 0 0 angstroms thereof f the nitrogen-containing gas wherein the ammonia gas has 1 in which This method is thermal 3. The method according to item 1 of the patent application range is about 0.16 microns in diameter. 4. The method according to item 1 of the scope of patent application is titanium tetraiodide (Ti 14). 5. According to the method in the scope of patent application No. 4 to a temperature of about 4 0 ° C-6 5 0 ° C. 6. Provide the precursor to the speed of the scope according to the method of scope 4 of the patent application. 7. Deposition the TiN at a rate in the range of 4 per minute of the method of patent application. 8. The method according to item 1 of the scope of patent application is ammonia. 9. The method according to item 8 of the scope of patent application. The flow velocity in the range of about 3 s 1 m. 10. Method according to item 1 of the scope of patent application CVD method. 1 1. A conformation provided in a high aspect ratio component of a substrate in a reaction chamber O: \ 64 \ 64000.ptc Page 丨 2001. 08. 23. 024 '4 6 659 t No. 89107861 _year t month Qiao Yue _ Amendment 6. Application for patent scope seamless titanium nitride (T 1 N) column A method of plugging, comprising a pressure in the chamber of less than 1.5 Torr, while providing titanium tetraiodide to the chamber at a temperature sufficient to evaporate the precursor, and then applying the vapor with about 5-40 seem The TiN is deposited by a chemical vapor deposition method for a combination of process gas containing nitrogen under a flowing range. 12.-A integrated circuit substrate comprising a high aspect ratio component filled with a shape-preserving seamless T film deposited on a Ti film derived from a Ti halide precursor, the T i Ν film being capable of supporting A stress in the range of about 1-1 5 X 1 09 dyne / cm 2 and a resistance in the range of about 8 5-4 0 microohms. 13. The substrate according to item 12 of the application, wherein the TiN film has a thickness of at least about 4000 angstroms. Wherein the TiN film has the T i _ compound wherein the T i halide 1. The substrate according to item 12 of the patent application scope is less than about 2 atomic% impurities. 1 5. The substrate precursor according to item 12 of the scope of patent application is titanium tetraiodide. 16. The substrate compound precursor according to item 12 of the scope of the patent application is heated to provide sufficient evaporation to the reaction chamber, and then the vapor is combined with a processing gas containing nitrogen, and the substrate is chemically vapor-deposited on the substrate. This TiN is deposited on. O: \ 64 \ 640GO.ptc Page 2 2001.08. 23. 025