JP2009132982A - Method of manufacturing copper wiring - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a novel method of manufacturing copper wiring by which copper is embedded into a very small hole or a groove. <P>SOLUTION: In the method of manufacturing the copper wiring provided with a step of electroplating the wiring connection holes or the wiring groove of a material to be plated with copper while forcibly stirring an electrolyte containing a copper component, ≥1 wt.% water and ≥1 wt.% copper complexing agent forming a complex with univalent copper ion, exhibiting a complexing constant of higher than 1×10<SP>3</SP>with univalent copper ion and ≥0.5 g/L solubility of the complex formed with univalent copper under a use environment or copper complexing agent forming a complex with univalent copper ion, exhibiting a complexing constant of higher than 1×10<SP>3</SP>with univalent copper ion and a complexing constant of ≤1×10<SP>20</SP>with 2-valent copper ion, an embedding rate of copper into fine holes is adjusted by controlling the degree of stirring to increase the embedding rate of the copper into the fine holes. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a method of manufacturing a copper wiring by embedding copper in a wiring connection hole (via hole or contact hole) or a wiring groove (trench) by electroplating.

Semiconductor devices are provided with a number of wiring grooves (trench) for connecting elements and wiring connection holes (via holes or contact holes) for electrically connecting multilayer wirings.
Conventionally, aluminum has been used as a conductive material embedded in these wiring grooves and wiring connection holes. However, as semiconductor devices are highly integrated and miniaturized, their electrical resistivity is low instead of conventional aluminum. Copper (also referred to as low resistance), which has excellent electromigration resistance, has attracted attention and is being put to practical use.

As a copper wiring manufacturing method, grooves and holes are formed in an insulating film formed on a substrate made of a silicon wafer or the like, and a barrier metal (diffusion prevention film) and a Cu film (underlying layer for obtaining conduction) are formed thereon. After sequentially laminating the conductive film), a copper layer is formed on the surface while copper is embedded in the grooves and holes by electroplating, and then the excess copper layer is removed by chemical mechanical polishing (CMP) etc. The so-called damascene method is used.
Copper wiring formed by such electroplating (also referred to as electroplating) has a low impurity concentration in the film and a low electrical resistance, which is advantageous for increasing the speed of semiconductor devices.

  Conventionally, as an electrolytic solution for copper plating used in the formation of such copper wiring, three types of organic additives in a copper sulfate solution, that is, a carrier such as polyethylene glycol (PEG), bis (3-sulfopropyl) disulfide- An electrolytic solution in which three types of organic additives called brighteners such as 2-sodium (SPS) and levelers such as Janus Green B (JGB) and chloride ions have been used has been used.

However, the electrolyte solution containing three kinds of organic additives and chloride ions as described above needs to strictly control the concentration of each additive, and the concentration management is very difficult. It was. In particular, organic additives react easily on the electrode and decompose easily, and the concentration tends to decrease. Therefore, concentration control is extremely difficult, and the decomposition product of the organic additive makes it difficult to fill the micropores. And problems such as deterioration of film thickness uniformity. Further, it has been pointed out that carbon (C) contained in the organic additive is incorporated as an impurity in the plating film, so that the purity of the copper film is lowered and the electromigration resistance is deteriorated.
Therefore, recently, in view of such problems, development of an electrolytic solution for copper plating having a composition as simple as possible without using organic additives and additives such as chloride ions as much as possible has been promoted.

  For example, Patent Document 1 discloses a technique for embedding copper into a micropore with only a single organic compound. However, since this electrolyte is an alkaline pyrophosphate, cyan, or sulfamine system, a pH adjuster (such as phosphoric acid or potassium hydroxide) is added. In practice, several types of additives are required.

  Patent Document 2 discloses a method for realizing good embedding characteristics in micropores by adding an appropriate amount of hydrochloric acid to a copper sulfate aqueous solution. However, even in this method, if the chlorine concentration is low, embedding cannot be achieved, and if the hydrochloric acid concentration is too high, copper is easily dissolved and the film formability is lowered. However, there is a problem that chloride ions are exhausted and management of chloride ions is difficult.

  Patent Document 3 discloses a method in which a dense plating film is uniformly formed in a wiring groove or a wiring hole by appropriately controlling a duty ratio in a pulse current using a plating solution containing no additive. ing. However, this method has a problem that not only the equipment cost is high, but also current control is very difficult. Furthermore, since the diffusion layer is thinned using a pulse current, uniform precipitation in the fine pores can be expected, but there is a possibility that it cannot be embedded sufficiently.

  On the other hand, the inventor of the present invention pays attention to electroplating using an electrolytic solution containing acetonitrile, and discloses that the orientation of the plating film obtained by using such an electrolytic solution is one in which the (111) plane is preferentially oriented. (Patent Document 4). However, in the present invention, it is not specifically assumed that copper is embedded in the wiring connection hole or the wiring groove, and the effect cannot be predicted.

Special table 2003-533867 gazette JP 2002-332589 A JP-A-11-97391 JP 2007-182623 A

  In view of such a problem, the present invention has an extremely fine hole or groove (for example, a depth of 0.1 μm to 2.0 μm and an aspect ratio (depth / width) of 3 without using a plurality of additives. The present invention is intended to provide a new method for manufacturing a copper wiring that can sufficiently embed copper in a hole or groove 5.

The present invention forms a complex with a monovalent copper ion, a complexing constant with the monovalent copper ion is higher than 1 × 10 ³ , and a complex formed with a monovalent copper ion in the use environment A complexing constant with a copper complexing agent having a solubility of 0.5 g / L or more, or a monovalent copper ion, and a complexing constant with the monovalent copper ion is higher than 1 × 10 ³ , and An electrolytic solution containing 1 wt% or more of a copper complexing agent having a complexing constant with a divalent copper ion of 1 × 10 ²⁰ or less, 1 wt% or more of water, and a copper component (hereinafter referred to as “complexing agent”). We propose a method for manufacturing a copper wiring comprising a step of electroplating copper in a wiring connection hole or a wiring groove in an object to be plated while forcibly stirring the “containing electrolyte”.

  By performing electroplating using the complexing agent-containing electrolytic solution as described above, even with an electrolytic solution having a simple composition without adding a plurality of additives, extremely fine holes and grooves (for example, having a depth of It is possible to sufficiently embed copper in a hole or groove having an aspect ratio (depth / width) of 3 to 5 and 0.1 μm to 2.0 μm. Therefore, extremely fine copper wiring can be formed without strictly controlling the concentration of the electrolyte component.

Furthermore, when electroplating while forcibly stirring the complexing agent-containing electrolyte, it was found that the copper filling rate in the micropores can be adjusted by adjusting the degree of stirring. The embedding rate can be further increased.
In normal electroplating, when the electrolyte is stirred, it is known that high current efficiency close to 100% can be maintained up to a higher current density, and otherwise current efficiency decreases due to side reactions such as hydrogen generation. ing. On the other hand, in the case of electroplating using a complexing agent-containing electrolytic solution, the current efficiency is lowered when the electrolytic solution is stirred, and conversely, the current efficiency is increased by suppressing the stirring of the electrolytic solution. The facts turned out. By adjusting the agitation conditions based on this knowledge, the current efficiency in the micropores (grooves) can be improved even when copper is embedded in the micropores (grooves) with less liquid diffusion. As a result, the copper filling rate can be further increased. Furthermore, the amount of copper electrodeposited on the surface of the object to be plated can be reduced, and the processing burden for removing excess copper can be reduced.

  In addition, a copper film obtained by electroplating using an electrolytic solution containing an organic additive as in the prior art is obtained by incorporating carbon (C) contained in the organic additive as an impurity into the plating film. However, when the copper complexing agent according to the present invention, especially acetonitrile, is used, other organic additives and halogen additives are said to be reduced in purity. In addition, since acetonitrile does not remain in the copper film, the impurity concentration of the copper wiring is particularly low, and the electric resistance can be expected to be lower. Furthermore, since the orientation of copper wiring embedded with an acetonitrile-containing electrolyte tends to be preferentially oriented in the (111) plane, it can be expected that the wiring will have excellent electromigration resistance in that case. it can.

  As described above, the copper wiring formed according to the present invention can be particularly suitably used for an integrated circuit or an electronic circuit board such as a printed board.

BEST MODE FOR CARRYING OUT THE INVENTION

  As a preferred example of the embodiment of the present invention, a method for manufacturing a copper wiring will be described, but the present invention is not limited to the embodiment described below.

  Hereinafter, as an example of the embodiment of the present invention, by using a complexing agent-containing electrolytic solution, electroplating while forcibly stirring the electrolytic solution, copper is embedded in a wiring connection hole or a wiring groove to form a copper wiring. A method of manufacturing the copper wiring to be formed (referred to as “the present copper wiring manufacturing method”) will be described.

Here, a specific example according to the present copper wiring manufacturing method is introduced.
As shown in FIG. 1, an insulating film 1B such as a chemical film made of an insulating material is formed on a substrate 1A made of a silicon wafer or the like, and a groove or hole 2 is provided at a place where a wiring pattern is to be formed in the insulating film 1B. (See (A) in the figure) Next, a barrier metal film (diffusion prevention film) 3 made of Ti, Ta, W, or a nitride thereof, and a Cu base conductive film (a base conductive film for obtaining conduction) 4 is formed sequentially (see (B) in the figure), and then the electroplating of the complexing agent-containing electrolyte is performed while forcibly stirring, so that the surface of the substrate 1A is embedded while copper is embedded in the groove or hole 2. That is, the copper layer 5 is formed on the surface of the object to be plated (see (C) in the figure), and then the copper layer 5 formed on the surface of the substrate 1A is removed to expose the copper wiring 6 to form the copper wiring. (See (D) in the figure). At this time, in order to further improve the migration resistance, a metal, an oxide, or an organic substance may be laminated on the exposed copper wiring 6.
Hereinafter, although this copper wiring manufacturing method is demonstrated centering on this specific example, this copper wiring manufacturing method is not limited to this specific example.

<Wiring connection holes / wiring grooves>
Although the size of the wiring connection hole (via) or wiring groove (trench) in which copper is embedded is not particularly limited, the present copper wiring manufacturing method has, for example, a depth of 0.1 μm to 2.0 μm, and It is possible to sufficiently fill even extremely fine holes and grooves whose aspect ratio required by depth / width is 3 to 5, so that the holes have a diameter that is at least larger or shallower than that. It can be sufficiently embedded in the groove. The width of the fine hole or groove means the short diameter when the hole is a long hole, for example, and the short length when the hole is a groove.
However, the limitation of the present invention does not mean that the above-mentioned hole diameter, groove width, or aspect ratio.

For example, for holes or grooves for through electrodes such as SOC (system on chip), SiP (system in package), MEMS (mems, mechano-electric microsystem), etc., where the hole diameter or groove width is 100 μm and depth is 200 μm It can be embedded sufficiently.
Further, for example, it can be sufficiently embedded in a hole or groove of via filling plating of a printed wiring board having a hole diameter or groove width of 200 μm and a depth of 50 μm.

  The shape of the wiring connection hole (via) or wiring groove (trench) is not particularly limited. Incidentally, since it is difficult to provide a hole or groove of the same diameter from the opening to the back when the hole diameter or groove width is very fine with a hole diameter or groove width of 0.2 μm or less, usually, as shown in FIG. The cross-sectional shape is narrowed from the opening to the bottom.

<Copper>
Copper to be embedded, in other words, copper to be electroplated may be pure copper or copper containing a metal other than copper as an inevitable impurity. Further, for example, a copper alloy containing copper as a main component (at least 80% by mass or more of copper) may be used, such as a copper alloy in which a trace amount of a metal such as Co is added (for example, less than 5% by mass) for the purpose of improving corrosion resistance. Even if it is such a copper alloy, it is thought that the effect of this invention is acquired similarly to pure copper.

<Electroplating>
Next, an electroplating method in the present copper wiring manufacturing method, a method of forming the copper layer 5 on the surface of the substrate 1A while embedding copper in the groove or hole 2 will be described.

(Electrolyte)
The electrolytic solution used in the present copper wiring manufacturing method (hereinafter also referred to as “the present electrolytic solution”) is an electrolytic solution (“complexing agent”) containing a copper complexing agent of 1 wt% or more, water of 1 wt% or more, and a copper component. Containing electrolyte ").

Here, the copper complexing agent needs to form a complex with a monovalent copper ion, and the complexing constant with the monovalent copper ion needs to be larger than 1 × 10 ³ , and the upper limit value is particularly limited. Although it is not a thing, it is actually considered to be about 1 × 10 ⁴⁰ , and preferably 1 × 10 ⁴ to 1 × 10 ¹⁵ . When the complexing constant with monovalent copper ions is excessively large, the copper complex is excessively stabilized and the interelectrode voltage at the time of electroplating is increased, which is not preferable.
By the way, the complexing constant of acetonitrile with monovalent copper ions is 1 × 10 ⁴ , which is P. Kamau and RB Jordan: Inorganic Chemistry, vol. 40, Issue.
16, p. 3879 (2001).

The copper complexing agent used has a complexing constant of 1 × 10 ²⁰ or less with a divalent copper ion (including those not complexed with a divalent copper ion) or is monovalent under the use environment. The solubility of the complex formed with the copper ion must be 0.5 g / L or more, and a copper complexing agent satisfying both of these conditions is preferred.
In terms of complexing constants with divalent copper ions, it is preferably 1 × 10 ¹⁸ or less, particularly preferably 1 × 10 ¹⁶ or less.
In terms of the solubility of the complex formed with monovalent copper ions, the solubility is preferably 5 g / L or more, and particularly preferably the solubility is 10 g / L or more.

In this copper wiring manufacturing method, one of the points is to stabilize the complex formed by the copper complexing agent and the monovalent copper ion. For this purpose, the complexing constant with the monovalent copper ion is 1. It is necessary to use a copper complexing agent larger than × 10 ³ . For the same reason, when the copper complexing agent to be used also forms a complex with a divalent copper ion, it is also important that the complexing constant with the divalent copper ion is 1 × 10 ²⁰ or less. On the other hand, if the complex formed with monovalent copper ions does not have a dissolution amount of 0.5 g / L or more in the liquid used, it is considered that there is no practicality.

The above-mentioned “complexation constant (K) with monovalent or divalent copper ion” is a value defined as follows.
Successive reaction (sequential reaction: M + A → MA, MA +) in which a complex MA _n (indicating a complex of M and n A) is formed from monovalent or divalent copper ion M and copper complexing agent A a → _{MA 2,} in...), the sequential complexing constant k _n,
When k ₁ = [MA] / ([M] · [A]), k ₂ = [MA ₂ ] / ([MA] · [A]),.
The complexing constant (K) is a value represented by LogK = Logk ₁ + Logk ₂ +.
In addition, [M], [A], and [MA _n ] mean the respective concentrations (mol / L).

Although it does not specifically limit as a copper complexing agent used by this copper wiring manufacturing method, A complex with a monovalent copper ion is formed, and the complexing constant with a monovalent copper ion is higher than 1 * 10 < ³ >. A copper complexing agent having a solubility of 0.5 g / L or more in a complex formed with a monovalent copper ion in a use environment or a complex with a monovalent copper ion to form a monovalent copper Examples of copper complexing agents having complexing constants with ions of higher than 1 × 10 ³ and complexing constants with divalent copper ions of 1 × 10 ²⁰ or less include potassium pyrophosphate and cyanide. Examples thereof include sodium chloride and acetonitrile. Among them, acetonitrile is particularly preferable in consideration of complexing constants and dissolution rates with monovalent copper ions, and uniform mixing.

This electrolytic solution has a rate of dissolving copper of 0.05 to 5 mg min ⁻¹ cm ⁻² , especially 0.07 to 4 mg min ⁻¹ cm ⁻² , and in particular 0.1 to 3 mg min ⁻¹ cm ^−2. Those containing the copper complexing agent in an appropriate amount are preferred.

The above-mentioned “speed for dissolving copper” means that a copper plate is immersed in a complexing agent-containing electrolyte (temperature: 25 ° C.), and the copper plate is dissolved while stirring so that the limiting current density is 80 mA / cm ^2. It is a value obtained from the copper plate weight reduction per unit time and unit area.

  In order to obtain the effect of the present invention, it is important that the concentration of the copper complexing agent in the electrolytic solution, that is, the mixing ratio of the copper complexing agent with respect to the total amount of the copper complexing agent and water is 1 wt% or more. It is preferably 2 wt% or more, particularly 4 wt% or more. The upper limit value is not particularly limited, but is considered to be about 50 wt% in consideration of environmental load and the like, but is preferably 27 wt% from the viewpoint of securing the dissolved amount of copper sulfate.

As the copper component, for example, water-soluble copper salts such as alkaline copper cyanide, copper pyrophosphate, acidic copper borofluoride, and copper sulfate are preferable, and an aqueous copper sulfate solution containing copper sulfate and sulfuric acid is particularly preferable. These can be mixed in advance with a copper complexing agent.
The concentration of sulfuric acid added in this case can be adjusted as appropriate, but is usually 0.01 mol / L or more, and preferably 0.1 mol / L to 2 mol / L.

  As a preferred specific example of the electrolytic solution, an electrolytic solution obtained by diluting an electrolytic solution containing an aqueous copper sulfate solution and a copper complexing agent with pure water to have a desired composition concentration suitable for the purpose can be given. it can.

The electrolytic solution may or may not substantially contain halogen ions and organic additives such as carriers, brighteners, and levelers. Although it is one of the features of the present invention that the embedding rate can be sufficiently increased even if it does not substantially contain halogen ions or the above-mentioned organic additives, it is more effective if these are included. Since it can be expected to increase, it may be included.
Here, “substantially does not contain” means that it is not actively added, and when it is inevitably contained, it means that this is allowed. In terms of specific concentration, it is preferably 3 ppm or less.
In addition, additives such as brighteners, complexing agents, buffering agents, conductive agents, organic compounds (such as glue, gelatin, phenolsulfonic acid, molasses), polyhydric alcohols, and titanium may be added to the electrolytic solution. Is possible.

(Agitating the electrolyte)
When electroplating using a complexing agent-containing electrolyte, the current efficiency can be reduced by increasing the degree of stirring of the electrolyte, and conversely, the current efficiency can be increased by suppressing the stirring of the electrolyte. The surprising fact was revealed. Therefore, by adjusting the stirring conditions based on such knowledge, even when copper is embedded in micropores (grooves) with little liquid diffusion, electroplating is actively applied to the micropores (grooves). The embedding rate can be increased, and furthermore, the amount of copper electrodeposited on the surface of the object to be plated can be suppressed. The “surface of the object to be plated” does not include the surface in the fine holes (grooves) (the same applies to other cases).

The method for stirring the electrolytic solution is not particularly limited. For example, a method of stirring the electrolyte by rotating the substrate that is the object to be plated around an axis perpendicular to the substrate surface that is the plating surface, or stirring the electrolyte by vibrating the stirring tool in the electrolyte A method, a method of stirring the electrolytic solution by flowing the electrolytic solution, and other methods can be employed.
In addition, as said stirring tool, a stirring rod, a stirring blade, and the other member which can be stirred can be mentioned.

In order to increase the filling rate by positively electroplating inside the micropores (grooves), the degree of stirring of the electrolyte is controlled within a predetermined range in order to suppress the amount of copper deposited on the surface of the object to be plated Preferably, from the immersion potential using an electrolytic solution containing 0.05 mol / L copper sulfate + 0.5 mol / L sulfuric acid at 25 ° C. using a copper electrode as a counter electrode and a saturated calomel electrode as a reference electrode it is preferable to control the degree agitation of the electrolytic solution as the limiting current density when the potential scan is 10mA / cm ² ~150mA / cm ² at the cathode direction 100 mV / min, in particular 20mA / cm ² ~80mA / cm ^2, it is preferable to control the degree agitation of the electrolytic solution so as inter alia a ^{40mA / cm 2 ~80mA / cm 2} . Actually, it is preferable to measure the limit current density in advance by applying the electrode position to the manufacturing facility, and to control the degree of stirring of the electrolyte so that the predetermined limit current density is obtained.
The term “limit current density” in the present specification means the limit current density as an index indicating the degree of stirring of the electrolytic solution in the above definition.

Here, it supplements about a limiting current density.
In the present invention, the limiting current density is adopted as an index indicating the degree of stirring of the electrolytic solution. It is known that the movement limit current of a substance depends on the flow velocity, and it has been confirmed that the limit current density has a certain relationship with the degree of stirring. Therefore, the limiting current density can be evaluated as a parameter indicating the degree of stirring. For example, as shown in Test 2 to be described later, it is possible to convert the limit current density into the electrode rotation speed, or it is possible to convert the electrode rotation speed into the limit current density. The standard of the electrode area that defines the limit current density is an apparent area that does not include the inside of the hole (groove).

Since this limit current density can be obtained by plotting the relationship between the current density and the electrode potential, for example, in an electrolytic cell that actually performs embedded plating, a preliminary test under the following conditions is performed, and the current density and the electrode potential are The limit current density is obtained by plotting the above relationship, and the stirring speed in each stirring method can be adjusted so that the predetermined limit current density is obtained.
(Preliminary test conditions)
Using a copper electrode as a counter electrode and a saturated calomel electrode as a reference electrode, a potential scan at 100 mV / min from the immersion potential to the cathode in an electrolyte solution of 0.05 mol / L copper sulfate + 0.5 mol / L sulfuric acid at 25 ° C To do.

Considering the degree of stirring for each method of stirring the electrolytic solution, the method of stirring the electrolytic solution is to stir the electrolytic solution by rotating the substrate that is the object to be plated around an axis perpendicular to the surface of the substrate that is the plating surface. In the case of adopting the method, the rotation is preferably performed at 100 to 5000 rpm, more preferably 300 to 3000 rpm, and particularly preferably 600 to 3000 rpm.
In addition, when the method of stirring the electrolytic solution by vibrating the stirring tool is employed, it is preferable to vibrate the stirring tool at a frequency of 0.8 Hz to 10 Hz, particularly 1.5 Hz to 5 Hz. In particular, it is preferable to vibrate at a frequency of 2.5 Hz to 3.6 Hz.

(cathode)
The cathode used in the present copper wiring manufacturing method, that is, the material of the substrate to be plated is not particularly limited. Since the substrate material of a semiconductor device has a structure in which an insulating film such as an oxide film is usually formed on a substrate made of a silicon wafer or the like, electrical continuity cannot be obtained by itself, and electroplating cannot be performed. Therefore, it is common to form a base conductive film by laminating a conductive material such as copper on the insulating film by sputtering or other means.

(anode)
The material used as the anode, that is, the counter electrode used in the present copper wiring manufacturing method is not particularly limited. For example, in addition to copper, insoluble electrodes such as platinum and platinum-plated titanium, and other electrode plates can be exemplified, but copper is particularly preferable.

(Electrolysis temperature)
The electrolysis temperature, that is, the temperature of the electrolytic solution is not particularly limited, and may be 20 ° C. or higher. Especially, it is preferable to control so that it may become 25-45 degreeC in order to reduce manufacturing cost and evaporation of an organic component.

(Current density)
The current density is not particularly limited, but is preferably controlled to 5 mA / cm ² or more. Although the upper limit is not particularly limited, it is considered that about 500 mA / cm ² is a realistic upper limit.
More preferably, the current density is further controlled according to the purpose, and when the limit current density indicating the degree of stirring is large, it is preferable to increase the current density. For example, in order to increase the filling rate by positively electroplating the inside of the fine holes (grooves), the current density is set to 10 mA / cm ^{2 to} 200 mA in order to suppress the amount of copper deposited on the surface of the object to be plated. / cm ^2, particularly ^{20mA / cm 2 ~100mA / cm 2} , preferably controlled inter alia in ^{40mA / cm 2 ~80mA / cm 2} .

(Electrolysis time)
The electrolysis time (energization time) is not particularly limited. It is preferable to adjust appropriately according to the size and shape of the holes and grooves.

(Preferred electrolysis conditions)
In summary of the above points, as an example of preferable electrolysis conditions for suppressing the amount of copper electrodeposited on the surface of the object to be plated while increasing the filling rate by positively electroplating inside the micropores (grooves) When the concentration of copper sulfate in the electrolytic solution is 0.24 mol / L and the concentration of sulfuric acid is 1.8 mol / L, the concentration of the copper complexing agent in the electrolytic solution is 1 wt% to 27 wt%. limiting current density shown is ^{15mA / cm 2 ~80mA / cm 2} , current density can be exemplified conditions that it is ^{40mA / cm 2 ~100mA / cm 2} .

(apparatus)
The configuration of the electroplating apparatus is not particularly limited. For example, a plating tank containing an electrolytic solution is provided. The plating tank includes an electrolytic solution drain part and an electrolytic solution supply part. In the plating tank, a substrate holder for holding a substrate (for example, a semiconductor wafer) and an anode of a power source are provided. An electroplating apparatus in which an anode electrode connected to the electrode and an agitation mechanism such as an agitation rod and an agitation blade are arranged, and a rotating disk electrode apparatus configured such that the electrode can rotate to agitate the electrolyte Can do.

<Removal of excess copper layer>
Next, a method for removing an excess copper layer, that is, a method for removing a copper layer formed on a substrate surface which is a plating surface will be described.

  The method for removing the copper layer formed on the surface of the substrate includes an electropolishing method in which an object is used as an anode, an electropolishing method in which a current is passed, a surface chemical action of an abrasive (abrasive grain), or a chemical component contained in a polishing liquid. A chemical mechanical polishing method (CMP) by mechanical polishing based on relative motion between the agent and the polishing object, a method of dissolving in an etching solution, and other removal methods can be employed. These methods can be used in combination.

  Since the complexing agent-containing electrolytic solution used in the present copper wiring manufacturing method easily forms a complex with copper ions, it has an action of dissolving metallic copper and also functions as an etching solution. Therefore, at least a part of the copper layer on the substrate surface can be removed by bringing the electrolyte solution into contact with the copper layer surface formed on the substrate surface while forcibly stirring the electrolyte solution.

  Furthermore, while applying a reduction current to a to-be-plated body so that a to-be-plated body may become a cathode, and making the electrolyte solution contact the copper layer of a board | substrate surface, forcing stirring, the electroplating of copper ( Copper electrodeposition) and removal of excess copper layer (etching) can be performed simultaneously. At this time, the electrolytic solution is excellent in the electrodeposition effect of copper into the recesses as described above, but the copper layer surface of the micropores (grooves) tends to be recessed, so that the electrode in this portion The wearing effect can be preferentially advanced and flattened as a whole. Also, for example, even if the removal of the copper layer proceeds excessively and even the copper in the fine holes (grooves) is removed and the surface of the copper layer becomes concave, the fine holes (grooves) can be obtained by the excellent electrodeposition effect on the concaves. ) The electrodeposition of the part proceeds with priority and can be repaired to be flat. Therefore, it is possible to effectively remove the excess copper layer while preventing such over-etching.

At this time, the current density at the time of passing a current through the object to be plated is preferably controlled to be less than the current density at the time of electroplating, more preferably less than 40 mA / cm ² , and particularly preferably 0.1 mA / cm ^2. cm ^{2 to} 10 mA / cm ² .

  In addition, the method of removing the copper layer on the surface of the object to be plated by bringing the electrolytic solution into contact with the copper layer on the surface of the object to be plated while forcibly stirring the electrolyte is performed by chemical mechanical polishing. It can be used in combination with a method for removing the copper layer.

In addition, as a method of stirring the electrolytic solution at this time, as in the case of the above-described electroplating, a method of stirring the electrolytic solution by rotating about an axis perpendicular to the substrate surface, or a stirring tool in the electrolytic solution A method of stirring the electrolytic solution by vibrating the electrode, a method of stirring the electrolytic solution by flowing the electrolytic solution, and other stirring methods can be adopted. The stirring degree is a copper electrode as a counter electrode and a saturated calomel electrode as a reference electrode. The limiting current density is 10 mA / cm ² when the potential is scanned from the immersion potential to the cathode direction at 100 mV / min using an electrolyte containing 0.05 mol / L copper sulfate + 0.5 mol / L sulfuric acid at 25 ° C. 150 mA / cm ^2, especially preferably adjusted to about agitation of the electrolytic solution so as to ^{20mA / cm 2 ~80mA / cm 2} .
Furthermore, as a method for stirring the electrolytic solution, when a method of rotating the substrate A, which is the object to be plated, about an axis perpendicular to the surface of the substrate, which is the plating surface, is 100 to 5000 rpm, particularly 600 to It is even more preferable to rotate at 3000 rpm.
Moreover, when employ | adopting the method which vibrates a stirring tool with predetermined frequency, electrolyte solution is 0.8 Hz-10 Hz, especially 1.5 Hz-5 Hz, and especially 2.5 Hz-3.6 Hz. It is preferable to vibrate.

<Features of the obtained copper wiring>
According to this copper wiring manufacturing method, even if a plurality of additives are not substantially contained, extremely fine holes or grooves (for example, a depth of 0.1 μm to 2.0 μm and an aspect ratio (depth / width) of 3 The hole or groove which is ˜5 can be sufficiently filled with copper, and extremely fine copper wiring can be formed with excellent accuracy.

In addition, the copper wiring obtained by the present copper wiring manufacturing method has a feature of high purity. Particularly when acetonitrile is added to the electrolyte as a copper complexing agent, acetonitrile is contained in the obtained copper wiring. Therefore, a copper thin film having a low impurity concentration and a sufficiently low specific resistance can be obtained.
Furthermore, since the orientation of the copper wiring obtained by adding acetonitrile to the electrolytic solution tends to be preferentially oriented in the (111) plane, a wiring excellent in electromigration resistance can be obtained. Specifically, a crystal orientation with a relative integral intensity of (111) plane of 65% or more, preferably 70% or more, and preferably 80% or more can be obtained.
The relative integrated intensity of the (111) plane is the peak area of the (111) plane relative to the sum of the peak areas of the (111), (200), (220) and (311) planes in the X-ray diffraction chart. The ratio (%) is shown.

  Thus, the copper wiring formed by the present copper wiring manufacturing method can be effectively used for manufacturing electronic materials, for example, integrated circuits such as IC, LSI, and CPU.

(Glossary of terms)
In the present invention, “electroplating” includes all methods in which an electrolytic solution containing an ionized metal is energized to deposit the plated metal on the surface of the cathode.
In the present invention, “copper wiring” is a concept that includes both a semiconductor circuit and a printed wiring board, and includes three-dimensional wiring such as filled vias, wiring grooves, wiring connection holes, and through holes in addition to planar wiring.
In the present invention, when it is described as “X to Y” (X and Y are arbitrary numbers), it is “preferably larger than X” or “preferably Y” with the meaning of “X or more and Y or less” unless otherwise specified. It also includes the meaning of “smaller”.

  Hereinafter, although this invention is demonstrated based on a test result (equivalent to an Example), the scope of the present invention is not limited to the following test result. Tests 4 to 6 are described as test contents based on the idea of the present invention.

<Test 1>
As shown in FIG. 2, copper was embedded by electroplating into a number of grooves provided on the surface of a substrate made of a silicon wafer, using a Microcell Model I type manufactured by Yamamoto Metal Testing Co., Ltd.
As shown in FIG. 3, on the cathode, wiring grooves having a groove width of 190 nm and a depth of 700 nm are provided at intervals of 190 nm on a silicon wafer substrate (11 mm × 15 mm × 0.8 mm) having an insulating film made of silicon oxide on the surface. 185 were formed, and TaN and Cu were sequentially sputtered on the surface to form a barrier metal layer and a copper layer.
A phosphorous copper plate (8 mm × 12 mm) was used for the anode, and the distance between the electrodes was set to 25 mm.

The electrolytic solution was mixed with an aqueous copper sulfate solution and acetonitrile, and then diluted with pure water so that the water concentration was 1 wt% or more, and the Cu (II) concentration was 0.24 mol / L, and the sulfuric acid concentration was 1.8 mol / L. L and those adjusted to acetonitrile concentration of 1-27 wt% were used.
And as shown in Table 1, using a potentiostat (HA-151 manufactured by Hokuto Denko Co., Ltd.) while vibrating a stirring bar with 250 mL of electrolyte at a frequency of 0.8 Hz to 2.5 Hz, a current density of 20 while controlling the ^{~40mA / cm 2, 1.44C / cm} 2 performs electroplating at electrolyte temperature of 25 ° C., it was compared to embedding of the wiring groove for samples 1-7.

Further, as another electrolytic solution, after mixing a copper sulfate aqueous solution, acetonitrile, hydrochloric acid, polyethylene glycol (PEG, Mw: 4000) and Yanas Green B (JGB), pure water is used so that the water concentration becomes 1 wt% or more. When diluted, Cu (II) concentration 0.24 mol / L, sulfuric acid concentration 1.8 mol / L, acetonitrile concentration 14 wt%, chloride ion concentration 50 ppm, PEG concentration 300 ppm, JGB concentration 5 ppm, 3-mercapto-1-propanesulfone Using a solution prepared to a sodium oxide (MPS) concentration of 2 ppm, while stirring the stirring bar at a frequency of 2.5 Hz (limit current density 41.0 mA / cm ² ), a potentiostat (HA-made by Hokuto Denko) 151) is used to control the current density to 60 mA / cm ² , and other conditions are electroplated as described above. Then, the embedding property in the wiring trench was examined for the obtained sample 8.

(Measurement of limit current density)
As an electrolytic solution, an electrolytic solution containing 0.05 mol / L copper sulfate + 0.5 mol / L sulfuric acid at 25 ° C. was used, and at each frequency of the stirring rod, potential scanning was performed at 100 mV / min from the immersion potential to the cathode direction. The limiting current density was measured by measuring the current density and the electrode potential and plotting the relationship between the measured current density and the electrode potential.
In addition, from the viewpoint of improving the reproducibility of the measurement, acetonitrile is not added to the electrolytic solution used here, but there is no problem in terms of obtaining an index indicating the degree of stirring of the electrolytic solution.

(Evaluation of embeddability)
Cross sections of the obtained samples were observed using a focused ion beam processing observation apparatus / scanning ion microscope manufactured by SII Nano Technology, Inc., and the copper filling rate in the groove (hole or hole relative to the depth of the groove) Alternatively, the ratio (%) of the height at which copper was electrodeposited in the groove was measured, and the embedding property was evaluated according to the following criteria. Table 1 shows the average values at 12 locations.

◎: Copper filling rate in the groove is 90% or more ○: Copper filling rate in the groove is 70% or more and less than 90% Δ: Copper filling rate in the groove is 60% or more and less than 70% ×: Less than 60% copper filling in groove

(Discussion)
By electroplating using an acetonitrile-containing electrolyte, extremely fine holes and grooves (for example, a depth of 0.1 μm to 2.0 μm and an aspect ratio (depth) can be obtained without using a plurality of additives. / Width) It was confirmed that copper could be sufficiently embedded in the holes or grooves) of 3-5.
In addition, according to the acetonitrile-containing electrolyte, the composition can be made free of organic additives and halogen additives, and since acetonitrile does not remain in the copper film, the impurity concentration of the copper wiring is particularly low, and the electric resistance Can be expected to be lower.
To increase the micropores (groove) ratio embedding subjected to inside actively electroplating, to control the degree of agitation of the electrolytic solution as the limiting current density is 10mA / cm ² ~150mA / cm ² preferably, in particular ^{20mA / cm 2 ~80mA / cm 2} , is considered preferable to control the degree agitation of the electrolytic solution so as inter alia a ^{40mA / cm 2 ~80mA / cm 2} .
Further, when the method of stirring the electrolytic solution by vibrating the stirring tool is adopted, it is preferable to vibrate the stirring tool at a frequency of 0.8 Hz to 10 Hz, particularly 1.5 Hz to 5.0 Hz. In particular, it is considered preferable to vibrate at a frequency of 2.5 Hz to 3.6 Hz.

<Test 2>
The relationship between the degree of stirring and the limit current density was tested, and an equation for converting the limit current density into the electrode rotation speed was obtained.

A rotating disk electrode device (HR-201 manufactured by Hokuto Denko Co., Ltd.) without grooves or the like shown in FIG. 4 was used, a copper wire was used as a counter electrode, and a saturated calomel electrode (SCE) was used as a reference electrode. The working electrode has a 30 ° C. containing 0.5 mol / L copper sulfate and 1.0 mol / L sulfuric acid on a platinum disk electrode (HR-D2 manufactured by Hokuto Denko) having an electrode area of 0.2 cm ² . What was electrodeposited for 30 seconds at a constant potential of −0.4 V vs. SCE using the electrolytic solution was used as a working electrode.

Electroplating was performed at each rotational speed of 0 to 3000 rpm using an electrolytic solution at 25 ° C. containing 0.05 mol / L copper sulfate and 0.5 mol / L sulfuric acid.
In addition, the current density and the electrode potential at the time of scanning the potential from the immersion potential to the cathode at 100 mV / min are measured, and the limit current density is measured by plotting the relationship between the measured current density and the electrode potential. The relationship between density and electrode rotation speed is shown in FIG.

As shown in FIG. 5, a proportional relationship was recognized between the limit current density and the electrode rotation speed, and it was found that the limit current density can be converted from the limit current density to the electrode rotation speed by the following equation.
f = (− 1.37 + 0.72 × iL) ²
Where f: electrode rotation speed (rpm), iL: limit current density (mA / cm ² )

<Test 3: Etching test>
A test sample obtained by embedding copper under the same conditions as Sample 5 of Test 1 was used.
Using the apparatus of Test 1, set the sample in the same way as in Test 1, soak the sample in the etching treatment solution (25 ° C.), and the etching treatment solution at 2.5 Hz (limit current density 41.0 mA / cm ² ). While forcibly stirring, a reduction current was applied to the sample at a current density of 1 mA / cm ² or ² mA / cm ² to perform an etching treatment for 1 min to 4 min (see Table 2), and the etching effect was evaluated.

  The etching solution is mixed with an aqueous copper sulfate solution and acetonitrile, and then diluted with pure water so that the water concentration becomes 1 wt% or more, and the Cu (II) concentration is 0.24 mol / L, and the sulfuric acid concentration is 1.8 mol. / L, prepared at an acetonitrile concentration of 14 wt%.

The surface thickness reduction rate (%) and hole bottom thickness reduction rate (%) shown in Table 2 below are obtained by measuring the surface thickness (h ₀ ) and hole bottom thickness (H ₀ ) before etching in the cross section shown in FIG. The value obtained by the following formula.
Surface thickness reduction rate (%) = {(h ₀ −h) / h ₀ } × 100
Hole bottom thickness reduction rate (%) = {(H ₀ −H) / H ₀ } × 100

(Discussion)
It was found that under the conditions of 1 mA / cm ² and 2 min, the surface thickness reduction rate / hole bottom thickness reduction rate exceeded 1, and the surface layer copper was selectively etched.

<Test 4: Measurement of polarization curve>
A measurement test of a polarization curve based on the idea of the present invention will be described.

(Test method)
The electrolytic solution was prepared using copper sulfate pentahydrate (CuSO ₄ .5H ₂ O) (Wako Pure Chemical Industries, special grade), sulfuric acid (H ₂ SO ₄ ) (Wako Pure Chemical Industries, special grade) and acetonitrile (CH ₃ CN , Also referred to as “AN”) (Wako Pure Chemical Industries, special grade) and prepared with deionized water (this electrolyte is also referred to as “AN electrolyte”).
Copper concentration and sulfuric acid concentration of the electrolyte, water 1 wt% or more, and unified CuSO ₄ 0.24mol / dm ³ and ^{_{H 2 SO 4 1.8mol / dm 3}} .
Table 3 shows the composition of the three types of electrolytes used for the measurement of the polarization curve.
In the present experiment Cl in AN electrolyte solution ^- no added. In a general electrolytic solution, slime is generated without adding Cl ⁻ , but in an AN electrolytic solution, slime is not generated even without Cl ⁻ . This is considered because Cu (I) forms a stable complex with AN.

The polarization curve was measured with a rotating disk electrode device (Hokuto Denko, HR-201 and HR-202) without grooves or the like. The counter electrode was a copper wire (Niraco, 99.9 wt%), and the reference electrode was a saturated calomel electrode (SCE) (Toa DKK, HC-205C).
Note that all potentials are based on SCE (the same applies to subsequent tests).
^On a platinum disk electrode (Hokuto Denko, HR-D2) having an electrode area of 0.2 cm ² , an electrolytic solution composed of CuSO ₄ 0.5 mol / dm ³ and H ₂ SO ₄ 1.0 mol / dm ^{3 at} 30 ° C. was used. A working electrode was obtained by electrodepositing copper at a constant potential of −0.4 V for 30 seconds.
The cathodic polarization curve was measured using the three types of electrolytes shown in Table 3 at a potential scanning speed of 100 mV / min from the immersion potential to the base direction at each electrode rotation speed of 0, 500 and 2000 rpm. It was. The temperature of the electrolytic solution was 30 ° C.

(Results and discussion)
Cathodic polarization curves were measured using the typical electrolyte solution (A bath), additive-free electrolyte solution (B bath) and AN electrolyte solution (C bath) shown in Table 3 when the electrode rotation speed was changed. The result is shown in FIG.
In FIG. 7, in the bath A, a tendency was observed that the current value decreased as the electrode rotation speed increased in the potential range of 0 to −0.3V. This tendency is considered to be because the adsorption of JGB acting as an electrodeposition reaction inhibitor is due to diffusion. On the other hand, such a tendency was not observed in the B bath and the C bath.
Also, comparing FIG. 7 (b) and FIG. 7 (c), and looking at the effect of acetonitrile addition, the immersion potential is -0.076V in the C bath, compared with 0.044V in the B bath. It was a big shift to the bottom.

<Test 5: Measurement of current efficiency>
A current efficiency measurement test based on the idea of the present invention will be described.

(Test method)
The current efficiency was calculated from the ratio of the amount of electricity deposited and the amount of electricity dissolved using a rotating disk electrode device without grooves or the like. The current efficiency was calculated by assuming that the reaction progresses by the reaction shown in the formula (1) in both electrodeposition and dissolution.
Cu (II) + 2e ⁻ ⇔Cu (1)

For the electrodeposition of copper on the electrode, three types of electrolytes shown in Table 3 were used (the temperature of the electrolyte was 330 ° C.), the working electrode was a platinum electrode with an electrode area of 0.2 cm ² , and the counter electrode was a copper wire (Niraco, 99.9%), SCE was used as the reference electrode.
The current density was controlled by each electrolytic solution, and electrodeposition was performed at electrode rotation speeds of 0, 500 and 2000 rpm. At this time, the amount of electricity supplied was constant at 6.75 C / cm ² .
The amount of electrodeposited copper was calculated from the amount of electricity consumed for the anodic polarization and the dissolution reaction.
The electrode electrodeposited with copper was dissolved in a solution of CuSO ₄ 0.5 mol / dm ³ and H ₂ SO ₄ 1.0 mol / dm ³ by electrochemically dissolving copper on the platinum electrode at 0.2 V to obtain a dissolution current. The end point was when the density was less than ² mA / cm ^2, and the amount of electricity that flowed was measured with a coulomb meter.
The reason why the amount of electrodeposited copper is calculated from the amount of dissolved electricity is that the analysis time can be shortened as compared with a spectroscopic analysis method such as an inductively coupled plasma emission spectrometer (ICP).

(Results and discussion)
(1) Measurement of current efficiency The three types of electrolytes shown in Table 3 were used, and the current efficiency at various current densities was measured with a rotating disk electrode device without grooves or the like. The electrode rotation speed was 500 rpm. The results are shown in FIG.
From FIG. 8, in the A bath, the current efficiency of 80% or more was maintained in the current density region of 100 mA / cm ² or less, and the current efficiency decreased to 26.5% at 200 mA / cm ² . Similarly, in the B bath, a current efficiency of 80% or more was maintained in a current density region of 100 mA / cm ² or less, and the current efficiency was reduced to 72% at 200 mA / cm ² . These reductions in current efficiency are due to hydrogen generation as a side reaction, as can be seen from the polarization curve measurement results shown in FIG.
On the other hand, in the C bath, the current efficiency increased from 0% to 67% as the current density increased from 10 mA / cm ² to 200 mA / cm ² . As the current density increased, the slope of increase in current efficiency decreased.
There was no difference in the change in the amount of current with respect to the flow rate (electrode rotation speed) between the B bath and the C bath (FIG. 7), but it was found that there was a large difference in the current efficiency (FIG. 8).

(2) Effect of flow rate on current efficiency The effect of flow rate on the current efficiency of copper was examined.
Using three types of electrolytic solutions shown in Table 3, the current efficiency was measured when the electrode rotation speed was changed to 0 and 2000 rpm. The results when the electrode rotation speed in FIG. 8 is 500 rpm are also shown in FIG.
The A and B baths maintained high current efficiencies at all current densities from 0 to ²⁰ mA / cm ^{2 where} the experiments were conducted at electrode rotation speeds of 500 and 2000 rpm, and there was no significant difference. However, when the electrode rotation speed was 0 rpm, the current efficiency greatly decreased as the current density increased. This is because a hydrogen generation reaction occurs.
On the other hand, in the C bath, the current efficiency was remarkably low at less than 16% at electrode rotation speeds of 500 and 2000 rpm. However, when the electrode rotation speed was 0 rpm, copper electrodeposition was observed even at 5 mA / cm ² (current efficiency 25%), and when the current density was 20 mA / cm ² , the current efficiency was 63%. This is higher than 42% of the A bath without AN and 53% of the B bath under the same conditions. It was found that the AN electrolyte has a feature that the current efficiency increases as the flow rate decreases.

(2) Factors affecting current efficiency Copper electrodeposition reaction is generally expressed by the following reaction.
Cu (II) + e ⁻ → Cu (I) (2)
Cu (I) + e ⁻ → Cu (3)

In the AN electrolyte, the current efficiency was about 60% or less, and the current efficiency changed greatly depending on the current density and flow rate (FIGS. 8 and 9). That is, it is considered that reactions other than the electrodeposition of copper represented by the formulas (2) and (3) contribute to the electrodeposition using the AN electrolyte.
Since AN forms a complex with Cu (I), Cu (I) becomes stable in a sulfuric acid solution containing AN, and the following leveling reaction (reverse reaction of disproportionation reaction) occurs.
Cu + Cu (II) → 2Cu (I) (4)
Formula (4) is composed of a cathode reaction and an anode reaction, and each reaction is as follows.
Cu (II) + e ⁻ → Cu (I) (2)
Cu → Cu (I) + e ⁻ (5)

In the AN electrolyte, as shown by the formula (4), a copper dissolution reaction by Cu (II) occurs, and this is expected to cause a decrease in current efficiency. Therefore, a dissolution test was conducted to determine whether copper was dissolved in the AN electrolyte under these conditions.
A platinum disk electrode on which 6.75 C / cm ² of copper was electrodeposited was immersed in a 30 ° C. C bath at an electrode rotation speed of 500 rpm. Since platinum and copper have different immersion potentials, the immersion potential transitions preciously at the end of dissolution. After 12 minutes, the potential suddenly changed to the noble side, and the platinum electrode surface was exposed.
From the change in the immersion potential, the time until the copper was dissolved was obtained, and the average current density of the formula (5) at which the metallic copper was dissolved in Cu (I) was calculated. Its value was almost the same value as about 5 mA / cm ² of corrosion current density was determined from the intersection of 500rpm in Tafel straight and immersion potential of 4.7mA / cm ^2, and the Fig. 7 (c).
From this, the current efficiency of the AN electrolyte shown in FIG. 8 (c) was about 60% or less, and the immersion potential was shifted to the base side in FIG. 7 (c). This is thought to be due to the occurrence of the leveling reaction (equation (4)).

(4) Consideration of factors affecting current efficiency Based on the above results, the reaction model shown in FIG. 10 is assumed. The electrodeposition reaction of copper in the AN electrolyte includes mass transfer from the bulk of Cu (II) to the electrode surface (process I), reduction of Cu (II) s to Cu (I) s (process II) and Cu (I) Proceed via reduction from s to Cu (process IV). The subscript s indicates the ion species near the electrode surface.
The leveling reaction of formula (4) also occurs (process V). In FIG. 10, Cu (I) s generated by the reaction pathways II and V are diffused into the bulk (process III) or reduced to metallic copper (process IV) as shown below.
Cu (I) s → Cu (I) _b ... (Process III) ... (6)
Cu (I) _b + e ⁻ → Cu (process IV) (7)
Here, the subscript b means a solution bulk ion species. The current efficiency of copper using AN electrolyte is determined by the balance of the speeds of processes I to V in FIG.

In copper electrodeposition using AN electrolyte, when the current density is constant (deposition rate (sum of processes II and IV) is constant), the current efficiency decreases as the flow rate increases. This is probably because I is accelerated and the process V is further accelerated, so that the rate of dissolution rate of copper becomes larger than that of electrodeposition. When the flow rate is constant (the mass transfer rate of processes I and III is constant), the current efficiency decreases as the current density decreases. The deposition rate represented by processes II and IV is small. This is thought to be offset by the dissolution of the water.
From the above study, it can be seen that there are two major factors affecting the current efficiency of copper electrodeposition in the C bath. One is (b) the effect of decreasing the current efficiency as the flow velocity is large (the effect of the flow velocity), and the other is (b) the effect of decreasing the current efficiency as the current density is small (the effect of current density). It is.
The current efficiency of a general electrolyte is not affected by the flow rate or current density in a region where hydrogen generation does not occur. Even in a typical electrolyte, the current efficiency is almost 100% (FIG. 9A), and it can be said that the AN electrolyte has a specific property.

<Test 6: Electrodeposition of copper in fine pores>
The copper electrodeposition test on the fine pores based on the idea of the present invention will be described.

(Test method)
Copper embedding is a processed silicon having fine holes (diameter: 150-200 nm, depth: about 700 nm) in which a tantalum nitride layer and a tantalum barrier layer are sequentially formed and a copper seed layer of about 10 nm is formed on the outermost surface. A wafer (manufactured by Global Net) was cut into 1.5 cm × 1.1 cm and used as a cathode electrode.
A precision plating cell for silicon wafers (Yamamoto plating tester, microcell type I) was used for copper electrodeposition on the fine holes.
The silicon wafer cathode was covered with a Teflon mask provided with a 1.4 cm × 1.0 cm window so that the exposed area was 1.4 cm ² .
A copper plate was used for the anode, and the area of the anode electrode window was 0.96 cm ² .
The temperature of the electrolytic solution was 25 ° C., and the amount of the electrolytic solution was 0.25 dm ³ .
Using the three types of electrolyte shown in Table 3, the potentiostat (Hokuto Denko, HA-151) while the stirring rod was swung at about 0.8 or 2.5 Hz for the electrolyte between the anode and cathode. And electrodeposited at a predetermined current density.
A predetermined current was applied simultaneously with the immersion of the electrode in the electrolytic solution.
The amount of electricity supplied was constant at 1.44 C / cm ² .
The electrode electrode after the electrodeposition was washed with water and dried, and the cross section of the sample was observed with a focused ion beam processing observation apparatus / scanning ion microscope (FIB-SIM) (SII Nanotechnology, SMI9200).

(Results and discussion)
(1) Copper embedment in micropores Based on the results of the examinations in Tests 4 and 5, when AN electrolyte is applied to copper embedment in micropores, embeddability depends on the balance between the effects of flow velocity and current density. Will change. Therefore, copper was electrodeposited on a processed silicon wafer having fine holes, and a cross-section was observed. The result is shown in FIG.
When the C bath was used, copper was not electrodeposited at the bottom of the hole, and copper was electrodeposited only at the top (36%) of the hole.

It can be considered that the effect of the current density (a) is prioritized over the effect of the flow velocity (b) that copper is deposited only on the upper part of the hole in FIG. That is, it is considered that the current density at the bottom of the fine hole is not so large that copper can be deposited, and copper is deposited only on the top of the hole.
By the way, when 100 ppm of Cl ^- was added to the AN electrolyte, the embedding property was further deteriorated. Generally Cl ^- is accelerating dissolution of copper in the copper sulfate electrolytic solution, also accelerate the dissolution of copper in AN electrolyte solution, can be expected to copper is no longer more electrodeposited fine hole bottom portion. Therefore, the chloride ion in the plating solution is desirably less than 100 ppm.
Conversely, if the effect of the flow velocity in (b) can be dominant, the hole should be embedded. From FIG. 9C, the current density is 10 mA / cm ² , and the ratio of the amount of electrodeposited copper due to the difference in flow rate is maximized, which is convenient for embedding.

Next, copper was electrodeposited by changing the current density, and the embedding property was examined. However, since copper was not electrodeposited on the bottom of the fine hole in FIG. 11, the experiment was conducted at a current density higher than that of FIG. The result is shown in FIG.
For a current density of 10 mA / cm ² shown in FIG. 11, whereas the embedding rate from hole upper was 36%, 12 than the current density 20 mA / cm ² at 67%, the current density of 30 mA / cm ² At 70% and current density of 40 mA / cm ² , the burying rate increased with increasing current density of 76%.

Next, the effect of the flow rate on the embedding of copper in the micropores was examined.
When the flow rate is large, the current efficiency (copper electrodeposition amount) decreases. Therefore, if the flow velocity at the top of the wafer is further increased, the flow velocity difference between the inside and outside of the micropore is increased, and it is considered that electrodeposition at the bottom of the hole is more preferential than that at the top of the hole. Therefore, the rocking frequency of the stirring rod was increased from about 0.8 Hz to about 2.5 Hz at the current density shown in FIG. 12 (c), and copper was electrodeposited into the fine holes. The result is shown in FIG.
From this, it was found that by increasing the flow rate, copper electrodeposition occurred to the bottom of the hole.
The electrolyte for embedding copper in a typical micropore increases or decreases polarization by carriers, brighteners, levelers, and chloride ions, and locally controls the deposition rate to realize embedding. However, AN has been found to enable implantation with a different function than these additives.

(A)-(D) are sectional drawings explaining the manufacturing method of copper wiring in order of a process. 4 is a diagram illustrating a configuration of a cell used in Test 1. FIG. It is a figure explaining the structure of the to-be-plated body used by Test. It is a figure explaining the structure of the rotating disk electrode apparatus used by Test 2. FIG. 4 is a graph showing the relationship between the limiting current density and the electrode rotation speed as a result of Test 2. It is a figure for demonstrating the hole bottom thickness (H) and surface thickness (h) which were used for evaluation of Test 3. FIG. In Test 4, it is the graph which showed the cathode polarization curve at the time of changing an electrode rotational speed, (a) is typical electrolyte solution (A bath), (b) is additive-free electrolyte solution (B bath), ( c) is a graph when an AN electrolyte (C bath) is used. In the figure, (i), (ii), and (iii) indicate electrode rotation speed (i) 0 rpm, (ii) 500 rpm, and (iii) 2000 rpm, respectively. In Test 5, it is a graph which shows the change of the current efficiency at the time of changing an electric current density on the conditions with an electrode rotational speed of 500 rpm, and electricity supply amount 6.75C / cm < ² > constant, (a) is typical electrolyte solution ( (A bath) and (b) are graphs when an additive-free electrolyte (B bath) is used, and (c) is an AN electrolyte (C bath). In Test 5, it is a graph which shows the change of the current efficiency at the time of changing a current density for every electrode rotation speed, (a) is typical electrolyte solution (A bath), (b) is additive-free electrolyte solution (B (Bath) and (c) are graphs when an AN electrolyte (C bath) is used. It is a figure which shows the reaction model of copper precipitation in electrolyte solution. In Test 6, when the electrolytic solution was electrodeposited at 10 mA / cm ² while the stirring bar was swung at 0.8 Hz, a cross-section of the obtained sample was observed with FIB-SIM. It is a photograph at the time of using electrolyte solution (C bath). In Test 6, when the AN electrolyte (C bath) was used as the electrolyte and the electrolyte was electrodeposited at a predetermined current density while the stirring bar was swung at 0.8 Hz, the cross section of the obtained sample was a picture of observing by FIB-SIM, a photograph of (a) is 20mA / cm ^2, (b) is 20 mA / cm ^2, which is electrodeposited in (c) is 30 mA / cm ^2. In Test 6, an AN electrolyte (C bath) was used as the electrolyte, and the electrolyte was electrodeposited at a current density of 40 mA / cm ² while the stirring bar was swung at 2.5 Hz. It is the photograph at the time of observing the cross section of a sample by FIB-SIM.

Explanation of symbols

1A Substrate 1B Insulating film 2 Groove or hole 3 Barrier metal film 4 Underlying conductive film 5 Copper layer 6 Copper wiring

Claims (12)

Forms a complex with a monovalent copper ion, the complexing constant with the monovalent copper ion shows a value higher than 1 × 10 ³ , and the solubility of the complex formed with the monovalent copper ion in the use environment is 0 A copper complexing agent of 0.5 g / L or more, or a complex with a monovalent copper ion, a complexing constant with the monovalent copper ion is higher than 1 × 10 ³ , and a divalent copper While forcibly stirring an electrolytic solution containing 1 wt% or more of a copper complexing agent having a complexing constant with ions of 1 × 10 ²⁰ or less, 1 wt% or more of water, and a copper component, A method for producing a copper wiring, comprising a step of electroplating copper in a wiring connection hole or a wiring groove. 2. The method for producing a copper wiring according to claim 1, wherein an electrolytic solution containing the copper complexing agent in an amount such that the rate of dissolving copper is 0.05 to 5 mg min ⁻¹ cm ⁻² is used. .   The method for producing a copper wiring according to claim 1 or 2, wherein acetonitrile is used as the copper complexing agent. While forced stirring the electrolytic solution as the limiting current density is ^{10mA / cm 2 ~150mA / cm 2} , the production method of the copper wiring according to any one of claims 1 to 3, characterized in that electroplating.   The electrolytic solution is forcibly stirred at a speed corresponding to rotating the object to be plated at an axis of 100 to 5000 rpm about an axis perpendicular to the plating surface. Copper wiring manufacturing method.   The method for producing a copper wiring according to any one of claims 1 to 4, wherein the electrolytic solution is forcibly stirred at a speed equivalent to vibrating the stirring tool at a frequency of 0.8 Hz to 10 Hz.   The method for manufacturing a copper wiring according to claim 1, wherein a mixing ratio of the copper complexing agent to the total amount of the copper complexing agent and water is 1 to 27 wt%.   The wiring connection hole or the wiring groove has a depth of 0.1 μm to 2.0 μm, and an aspect ratio calculated by the depth / width is 3 to 5, The manufacturing method of the copper wiring of crab.   By electroplating while forcibly stirring the electrolytic solution, copper is embedded in the wiring connection hole or wiring groove and a copper layer is formed on the surface of the object to be plated. The method for producing a copper wiring according to claim 1, wherein at least part of the copper layer on the surface of the object to be plated is removed by bringing the electrolytic solution into contact with the layer.   By applying a reduction current to the object to be plated so that the object to be plated becomes a cathode and bringing the electrolytic solution into contact with the copper layer on the surface of the object while forcibly stirring the electrolytic solution, The method for manufacturing a copper wiring according to claim 9, wherein at least a part of the layer is removed.   The method for producing a copper wiring according to claim 9 or 10, wherein a method of removing at least a part of the copper layer on the surface of the object to be plated by chemical mechanical polishing is used in combination.   The electronic circuit provided with the structure formed by forming a copper wiring in a board | substrate by the manufacturing method of the copper wiring in any one of Claims 1-11.

Images