Method of electroless plating

[57] ABSTRACT

The tendency of an electroless plating bath having a considerable amount of oxygen dissolved therein to “skip-plate” on catalytic surfaces of objects immersed therein, is inhibited by the removal of at least most of the oxygen initially present in the bath.

Claims, No Drawings

3,900,599

METHOD OF ELECTROLESS PLATING

BACKGROUND OF THE INVENTION

So-called “electroless” or “autocatalytic” plating baths usually contain a salt of the metal being deposited, a reducing agent for the metal salt, a chelating or complexing agent and a pH adjustor. If the surface of the object being plated is not one which is already catalytic for the deposition of the metal, a film of a suitable catalyst must be deposited on the surface before it is subjected to the plating bath. And, if the surface is on a body made of a substance such as a glass or a plastic, which is usually not capable of causing nuclei of the catalyst to deposit thereon, the surface must first be properly treated so that it will be receptive to the catalyst.

In the present state of the art, two procedures for sensitizing and catalyzing are widely used. One of these is immersion in an acidic stannous chloride solution followed by immersion in an acidic palladium chloride solution. The second method is immersion in a solution composed of palladium chloride and excess acidic stannous chloride, followed by an immersion in an accelerator solution.

During the plating operation, the plating bath is usually agitated to keep the composition of the bath uniform and to dislodge hydrogen bubbles that tend to form on the surface of the object being plated. Agitation is often carried out by bubbling air through the bath. It has been believed that the presence of oxygen improves bath stability.

However, in making printed circuits by an additive process which includes depositing copper electrolessly on a coated phenolic laminate which contains many through-holes, it has been found that many plating skips occur around the edges of the holes especially when the bath is freshly prepared. Such a product is not acceptable. It has now been determined that the defective plating is due to the presence of dissolved oxygen which is normally present when the bath has been standing exposed to air. The oxygen concentration present in a bath that has been exposed to air for some time will be referred to as the “equilibrium concentration.” It would be of advantage to have a simple and economical way to modify the plating bath or the process of using it so that the number of plating skips which occur due to dissolved oxygen is either considerably reduced or entirely eliminated.

The present method is an improved process of using “freshly” prepared or other electroless copper or other metal electroless plating baths containing considerable dissolved oxygen, that eliminates electroless plating skips on catalytic substrates. The improvement comprises reducing the oxygen content which may be present in the bath when it is first made up, or when it has been standing unused for a period preferably by removing it with nitrogen or other non-oxidizing gas which is inert with respect to the catalyst and to the plating solution.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The invention will be described in connection with manufacturing printed circuit boards by a process that includes both electroless plating and subsequent electro-deposition, but the present invention is of significance only with respect to the electroless plating step.

It should be understood that the invention is not limited to use of the substrates mentioned in the following examples, nor is it limited to any specific electroless plating composition. The substrate may be an epoxyfiberglass laminate, for example, glass, ceramic or a synthetic resin of any type.

EXAMPLE 1

Phenolic laminate boares are given an adhesive coating of uncured thermosetting synthetic resin and the coated surface is etched, sensitized and activated as described in U.S. Pat. No. 3,704,208. The adhesive coating composition may be a polyvinyl butyral phenolic mixture such as Pittsburgh Plate Glass Company’s E835, for example.

After application of the resinous coating composition, the coated panel is air dried and heated in an oven at about 300° F for 4—6 minutes to drive off the solvents and any free moisture.

After the panel has cooled, it is punched or drilled in accordance with the desired through hole configuration. In the present example, each panel is drilled with 180 holes. Alternatively, the holes may be formed before the coating is applied.

Next, the panel is passed through a cold water spray and abraded on both surfaces by rotating brushes. After the abrading step, the panel is given another cold water spray rinse.

After the coated panel is abraded, it is spray etched with a solution containing about 10% by volume of nitric acid and about 5% by volume of hydrochloric acid, in deionized water. The spraying temperature is about 130° F and the spraying time is about 2 minutes. At the conclusion of the etching treatment, the panel is rinsed in hot water.

Next, the panel is treated with a strong oxidizing conditioner which may be of the chromic acid type such as Enthone’s Enplate 470, which has had its activity raised by additions of chromium trioxide or a metal chromate. Following treatment with the conditioner, the panel is thoroughly rinsed.

After the oxidizing treatment, the panel is treated with a sensitizer solution. Although the sensitizer may be of conventional type such as stannous chloride acidified with hydrochloric acid, it is preferred that the following solution be used.

SnCl₂ 0.13 molar

HC1 0.47 molar

SnCL, (Aged) 6.5 X 10^-3 molar

NaCl 3.0 molar

The SnCl₄ is added from a 0.5 molar stock solution which was prior aged for about 1 week at room temperature. The aging process improves the wetting properties of the sensitizer solution and increases the density of the catalytic sites. A more complete description of the SnCl₄ aging process is formed in U.S. Pat. No. 3,666,527 issued May 30, 1972. A description of how the SnCl₄ is believed to be modified by the aging process can be found in an article entitled “Some Aspects of the Chemistry of Tin Sensitizing Solutions,” Feldstein et al, J. Electrochem. Soc., Vol. 119, p. 1482 (1972). The sodium chloride increases both the wettability and the stability of the solution.

3,900,599

The sensitized panel surface is then activated by treating it with an aqueous solution containing 0.2 g PdCl₂ and 1 ml concentrated HC1 per liter of solution. This deposits a discontinuous film of palladium nuclei on the surface. The palladium catalyzes the electroless deposition of metal that follows.

The panel is then electrolessly plated with copper. First, an aqueous plating bath is made up with the following composition:

PDTA Na, (Propylenediamine 0.072 molar tetraacetic acid, the tetrasodium salt)

CuSO₄.5H₂O J⁵ g/*!^ter

HCHO 20 ml/hter

CH₃OH ⁴⁰ ml/liter

NaOH ⁸ g/liter

NaCN ⁶ PP^m “TMN” a non-ionic wetting agent 3 mg/liter (manufactured by Union Carbide)

Temperature of the bath is 22° C.

Before the bath is used, nitrogen or other inert gas is bubbled through it at a rate of 2 cu. ft./hour (using 15 gallons of solution) for at least 30 minutes. The gas is bubbled throughout the entire volume of the solution. This removes a large proportion of the oxygen content of the bath. Besides nitrogen, some other suitable gases are argon, helium, neon and hydrogen.

The treated panel is then immersed in the plating bath for 10 minutes so that a thin continuous layer of copper deposits on both faces, including the through holes.

In order to compare the effect on the electroless plating process of a bath saturated with nitrogen with a bath having an equilibrium concentration of oxygen dissolved therein, 4 pairs of panels were plated. The oxygen content of the oxygen-containing bath was that due to normal exposure of the surface of the bath to air. One panel of each pair was plated in the above bath which had not had nitrogen bubbled through it and which therefore had considerable oxygen present due to dissolved air. The other member of each pair was plated in the same bath which had nitrogen bubbled throughout for 30, 60, 90 or 120 minutes. The four panels which were plated in the bath with no nitrogen treatment had plating skips (around the through holes) of from 15 to 22 per panel. Since there were 180 holes in each panel, this meant that from 8.3 to 12% of the through hole areas were improperly plated. In contrast, the panels that were plated in the baths pre-treated with nitrogen gas had no plating skips at all.

Since it is highly desirable that electroless plating baths remain stable over considerable periods of time and since past experience had shown that any modification of an electroless plating bath that reduced the number of plating skips also tended to make the bath unstable, a stability comparison test was run. In this case stability was determined by observing the time it took a plating bath to either become turbid or to plate on the bottom of the container under conditions of accelerated usage. The accelerated test utilized a number of pairs of 2¼ inch X 3¼ inch adhesively coated phenolic laminate panels having no through holes. These panels were treated by abrading, conditioning, sensitizing and activating as described above.

The electroless plating bath used was exactly as given above except that the amount of NaOH was 10 g/liter and the bath temperature was 25° C. Two 500 ml portions of this bath were taken. One of these portions was treated by bubbling nitrogen gas throughout it for 1 hour. The other bath portion was given no nitrogen treatment.

To begin the test, one of a first pair of panels was immersed in 400 ml of the nitrogen-treated solution and the other member of the pair was immersed in 400 ml of the untreated solution. Both panels were replaced every 16-17 minutes with fresh panels. To keep the baths in chemical balance, the baths were periodically analyzed for copper sulfate, sodium hydroxide and formaldehyde and replenished as required. In both baths the time that elapsed before the baths began to become turbid was 225 minutes and the time for plating to begin on the bottom of the container was 237 minutes. Thus, it was concluded that pre-treatment of the bath with nitrogen does not affect bath stability.

EXAMPLE 2

Using the same type of coated phenolic laminate panels, the same oxidizing conditioner, the same sensitizer and the same activator, as in Example 1, a number of panels having 180 through holes each were plated in the same freshly prepared electroless plating bath previously described, where the bath was given a 60 minute nitrogen pre-treatment before use. However, some of the panels were plated in a bath which was slightly agitated with air during the plating process and some were plated in a bath that was considerably air agitated. There were a few plating skips on the panels with slight air agitation and many skips on the panels with considerable air agitation. On the panels which were plated with no air agitation, there were no plating skips.

EXAMPLE 3

In order to show the effect of dissolved oxygen content on the ability of an electroless copper bath to produce “skip-free” plating on phenolic board, the following bath was employed at 20° C:

CuSO₄.5H₂O 15 g/liter

PDTA Na₄ (40% solution) 60 ml/liter

NaOH 8 g/liter

H₂CO (37%) 40 ml/liter

CH₃OH 40 ml/liter

NaCN 6 mg/liter

TMN 12 mg/liter

Phenolic boards, 6 inch x 2¼ inch, were prepared for plating using conventional preparatory techniques. The boards were sensitized for 1.0 minute at room temperature in a solution of 0.125 molar SnCl₂.2H₂O in 0.45 molar HC1, rinsed at 25° C for 1 minute in flowing tap water, activated in a solution of 0.50 g/1 PdCl₂ in 5 X 10^-3 molar HC1, again rinsed and plated in the above bath for 3.0 minutes.

The oxygen content of the bath was altered by bubbling nitrogen throughout the bath, the oxygen content being indicated in the following table.

Board

<td> No.</td><td> Oxygen Content*</td><td> % Skips</td><td> Visual Observations</td>

<td> 1</td><td> Equilibrium concentration when the bath is in contact with air</td><td> - 100%</td><td> Many large bare areas</td>

<td> 2</td><td> 85% of equilibrium concentration</td><td> &gt;75%</td><td> Very spotty</td>

<td> 3</td><td> 74% of equilibrium</td><td> about</td><td> Spotty</td>

3,900,599

-Continued

-Continued

<td> Board No.</td><td> Oxygen Content*</td><td> % Skips</td><td> Visual Observations</td>

<td></td><td> concentration</td><td> 25%</td><td> 5</td>

<td> D</td><td> 57% of equilibrium</td><td> about</td><td> Some spottiness</td>

<td></td><td> concentration</td><td> 1%</td><td></td>

<td> 5</td><td> 28% of equilibrium</td><td> about</td><td> Good coverage</td>

<td></td><td> concentration</td><td> 0%</td><td></td>

<td> 6</td><td> 14% of equilibrium</td><td> 0%</td><td> Good coverage</td>

<td></td><td> concentration</td><td></td><td> 10</td>

•As measured using galvanic-type dissolved oxygen probe made by Precision Scientific.

Electroless Plating Bath

<td> No.</td><td> Oxygen Content</td><td> through holes</td><td> Remarks</td>

<td> 1</td><td> Equilibrium concentration</td><td> -100</td><td> Many large bare areas</td>

<td> 2</td><td> 34% of equilibrium concentration</td><td> -50</td><td> Few large bare areas</td>

<td> 3</td><td> 5% equilibrium concentration</td><td> 0</td><td> Good coverage</td>

EXAMPLE 6

The adverse effect of dissolved oxygen in “freshly” prepared electroless plating baths decreases as the ac- 15 tivity of the bath increases. Activity of a bath can be increased by raising the temperature of operation or increasing the concentration of bath ingredients such as the metal salt or the reducing agent. However, as activity of a bath is increased, the bath becomes increasingly 20 unstable; therefore, it is necessary to compromise on the various factors associated with these baths and it usually proves to be most advantageous to sacrifice some degree of activity in favor of increased stability.

In order to show what effect (if any) changes in con- 25 centration of the various reagents and changes in reagents have on plating results obtained using the present method, the following examples are presented.

EXAMPLE 4 30

Electroless Plating Bath

<td colspan="3"> NiSO₄.6H₂O Na4P₂O₇.10H₂O nh₄oh Dimethylamine borane</td><td> 25 g/liter 50 g/liter 15 ml/liter 0.25 g/liter</td>

<td> Board No.</td><td> Oxygen Content</td><td> % skips around through holes</td><td> Remarks</td>

<td> 1</td><td> Equilibrium concentration</td><td> -100</td><td> Many bare areas</td>

<td> 2</td><td> 53% of equilibrium concentration</td><td> -75</td><td> Few bare areas</td>

<td> 3</td><td> 5% equilibrium concentration</td><td> 0</td><td> Good coverage</td>

EXAMPLE 7

Electroless Plating Bath

CuSO₄.5H₂O

PDTA Na, (40% solution)

NaOH

H₂CO (37%)

NaCN g/liter ml/liter 35 g/liter ml/liter mg/liter

Phenolic laminate boards provided with through 40 holes and an adhesive coating and pre-treated as described above were plated for 3 minutes. Except for Board No. 1 (the control) nitrogen gas was bubbled throughout the plating bath prior to the plating step to lower the oxygen content as indicated below. 45

Electroless Plating Bath

<td colspan="2"> NiSO₄.6H₂O Na₄P₂O₇.10H₂O NH₄OH NaH₂PO₂.H₂O</td><td colspan="2"> 25 g/liter 50 g/liter 15 ml/liter 10 g/liter</td>

<td> Board No.</td><td> Oxygen Content</td><td> % skips around through holes</td><td> Remarks</td>

<td> 1</td><td> Equilibrium concentration</td><td> -100</td><td> Very poor coverage</td>

<td> 2</td><td> 40% of equilibrium concentration</td><td> -50</td><td> Small bare areas</td>

<td> 3</td><td> 3% of equilibrium concentration</td><td> 0</td><td> Good coverage</td>

<td rowspan="2"> Board No.</td><td rowspan="2"> Oxygen Content</td><td colspan="2"> % skips around</td>

<td> through holes</td><td> Remarks</td>

<td> 1</td><td> Equilibrium concentration</td><td> ~100</td><td> Many bare areas</td>

<td> 2</td><td> 63% of equilibrium concentration</td><td> -25</td><td> Few bare areas</td>

<td> 3</td><td> 3% of equilibrium concentration</td><td> 0</td><td> Good coverage</td>

EXAMPLE 5

Electroless Plating Bath

<td> CuSO₄.5H₂O Ethylenediamine tetraacetic acid (EDTA) NaOH NaCN H₂CO (37%)</td><td> 15 g/liter 0.1 M to give pH of 13.0 24 mg/liter 20 ml/liter</td>

<td> Board</td><td> % skips around</td>

Although in all of these baths, the oxygen content was lowered after the bath was prepared, it can also be removed from the water prior to mixing in the other ingredients.

If a bath stands around for a time after it has been prepared and de-oxygenated, it should be tested for oxygen content just prior to use to determine whether or ⁵⁵ not the oxygen content has risen to such an extent that plating quality may be affected.