CA1271264A

CA1271264A - Adhesion promoting process for fabricating multilayer circuit boards

Info

Publication number: CA1271264A
Application number: CA000562724A
Authority: CA
Inventors: John L. Cordani
Original assignee: MacDermid Inc
Current assignee: MacDermid Inc
Priority date: 1987-08-26
Filing date: 1988-03-28
Publication date: 1990-07-03
Anticipated expiration: 2007-07-03
Also published as: EP0331684A4; JPH02500555A; US4775444A; AU1625588A; WO1989001990A1; EP0331684A1

Abstract

ADHESION PROMOTING
PROCESS FOR FABRICATING
MULTILAYER CIRCUIT BOARDS

ABSTRACT OF THE DISCLOSURE

Printed circuit boards having copper circuit patterns coated with an adhesion promoter (e.g., copper oxide, tin) are subjected to the action of an aqueous chromic acid solution prior to incorporation in a multi-layer board by interposing layers of semi-cured polymeric non-conductive material between the individual boards and laminating the assembly using heat and pressure. The chromic acid treatment serves to improve adhesion between the layers and, in particular, to overcome or minimize the "pink ring" effect which is frequently observed around holes drilled through multilayer boards.

Description

f`
~ `~ ` f `~
71;~

~)//~J /~ fd~J~7r~
~Y~Y*Æ~PROCES5 FOR FABRICATING
MULTILAYER CIRCUIT_BOARDS

Backqround Of The Invention _ _ 1. ~ield of the Invention This invention relates to an improved pxocess $or the preparation of multilayer prin~ed circuit board~ and is mor~ particularly concerned wi~h improving the adhesion between copper cixcui~ patterns and polymeric subs~ratesO

2. Descri~tion of the_Prior Art Multilayer printed circuit boards are assen~lies of several two-sided printed circuit boards bonded to each o~her through layers of s~mi-cured polymeric material by lamination using heat and pressureO The multilayer boards provide high packaging density, short conductor lengths and good reliability. At the same tlme weight and space are conserved. Accordingly, such boards are finding increased use.

The two-sided printed circuit boards fr~ which the multilayer printed circuit boards are built are yenerally fabricated from a polymeric substrate such as phenolic resins, epoxy resins, polyimides, polyesters and the like, clad on both sides with a thin layer of copper.
Printed circuits are formed on the copper layers by conventional technigues. Illustratively, a layer of photoresist is coat d over ~he copper layer, exposed imagewise and developed to yield a relief resist image on the copp~r layer. The exposed copper is etched away and the resist image i~ removed leaving the coppPr circui~
pattern exposed. The latter is then chemically trea~ed to form thereon a surface layer of copper oxide~ or other adhesion promoter as described in, for ex~mple, U.S. Patents 2,955,974; 3,177,103; 3,198,672; 3,240,662;

3,374,129; 3,481,777; 4,409,037 and 4,512,818.

Any desired number of the two-sided boards so prepared is ~hen assembled into the multilayer circutt S board by forming a stack of the boards with layers of semi-cured polymeric non-conduc~ive materlal such as epoxy resin interposed between the individual boards.
The assembly is laminated by application of heat and pressure. Typical of such laminatiny conditions are pressinq at about 300 psi to about 400 psi at a ~emperature at about 150C to about 205C for a time of up to about 4 hours. The cixcuits on the inner and outer layers are then connected by ~rilling an array of holes in the board, cle~ning the holes to remove resin smear by treatmen~ with acids and the like and thereafter plating with copper to render the sides of the holes conductiv~O

The use of a copper oxide layer as an adhesion promoter betwee~ ~he printed copper circuit pa~terns and the inter~ediate pol~meric dielectric layers o~ the multilayer board has been employed for a number o~ years.
Hswever, problems have arisen which have prompted a search for improv0d coatings. Thus total or local delamination of copper clrcuit pattern layers and polymeric layers can occur due to poor adhesion and/or the susceptibili~y of ~he copper oxide layer to dissolve when expo~ed to acidic and/or reducing solutions in later treatment of the multilayer board.

A major prohlem due to a combination of poor adhesion and the susceptibility of the copper o~ide to dissolution is that known as ~he "pinX rin~'l or "halo"
effect. When ~he multilayer ~oards are drilled in prepara~ion for making interlayer connec~ions ~y plated f 7~

through holes, a combination of stress and heat at the drill interface can cause local delamina~ion between the copper circuit pat~ern inner layers and the in~ermediate polymeric resin layers. $he copper oxide coating which is then exposed may be dissolved by the acid or like solutions employed to clean ~he resin smear from the drilled hvle prior to pla~ing the hole. Remo~al of the copper oxide coating around the edges of the copper circuit pattern adjacent ~o the through-hole reveals the underlying copper as a pink ring or halo. This causes cosmetic as well as functional problems.

A number of approaches ~o solving this widespread problem have been tried. Thus, pressing condi~ions employed in the lamination have been varied to optimize the adhesion promo~ed by the copper oxide coating. This has met with limited succes~. Thicker coatings of co~per oxide have been used in order to retard the dissolution process by sh~ar volume of the copper oxide presen~.
However, the ~hicker coatings give poor adhesion and ~herefore do not solve the pro~lem. Another approach h~s been to use a coating of tin in place of the copper oxîde layer as an adhesion promoter but the level of adhesion achieved has not been sufficient to overcome ~he problem.

It has now been found that ~he problems which have been encountered hitherto due ~o delamination, either total or local, of the copper circuit patterns and the intermediate polymeric resin layers in multilayer printed circuit boards can be overcome by a post trea~ment of the adhesion promoter layer of copper oxide, tin or like promoters~ prior to lamination of the individual ~wo-sided boards into the-mNltilayer board.

/
~:~7~

SUMMAKY OF THE INVENTION

It is an object of the invention to provide an improved process for the ~abrication o~ multilayer printed circu-t boards.

It is a further object of the invention to provide multilayer printed circuit boards having improved adhesion between, and greater resistance to delamination of, the layers therein.

It is yet another objec~ of the invention to provide multilayer printed circui~ boards which are substantially free ~rom ~ pink ring or halo effect associated with the through-holes therein.

These objects, and other object~ which will become apparent from the description which follows, are achieved by the process and articles of ~he present inve~ion.
T~e process of the invention in its broadest aspect comprises an improvement in a process for the preparation o~ a multilayer circuit board comprising the steps o~
providing a non-conductive substrate clad on each side with a copper circuit pattern layer, coating ~ach of the latter layers with an adhesion promoter, inaorporating the resulting board and one or more additional boards prepared in the same manner into a mul~ilayer board by interposing layers of semi-cured pol~meric nvn-conductive material between the boards and laminatiny t~e assembly using heat and pressure, wherein the lmpxovement comprises exposi~g the copper circuit pa~terns coated with adhesion promoter to the action of an a~ueous chromic acid solution prior to incorpora~ion of the individual boards into the multilayer board.

.

The invention also ccr..prises the multilayer printed circuit boards prepared ir. accordance with the process of the invention.

DETAILED DESCRIP~ON OF THE INVENTION

~n preparing the improved multilayer printed cir~
cuit boards in accordance with the inven~ion any of the procedures conv~ntional in ~he art for the preparation of such boards can be employed with the exception that, after the copper circuit patterns on the individual boards have been coated with an adhesion promoter, a novel step is in~roduced prior to incorporating the individual boards into ~he multilayer ~oard. The novel step oomprises exposing the copper circuit pat~erns coated with adhesion promoter to the action of an aqueous 1~ chromic acid solutio~ for a limited period of tLme.

The step in question is accomplished advant~geously by immersing the board with coated copper circuit pat~erns in the aqueous chromic acid solu~ion but it may also be accomplished by spraying the circuit patterns with the chromic acid solution or by like means. The aqueous chromic acid solution can be foxmed by dissolving the reguisite amount of chromium trioxide (also known as chromic anhydride~ in water or it can be formed in si~u by acidifying an aqueous solution of an alkali metal chro-mate or ~ichromate to a p~ of the order of about 1.5 ~o about 2.5 using a stxong mineral acid such as concent rated sulfuric acid. Advantageously the chromic acid solution, in whichever manner prepared, contains from about 0.1 to about 50 grams per liter and, preferably, from~ about 1.0 ~o about 2.0 grams per ~i~er of chromic acid. The chromic acid solution is adva~tageously a~ a temperatUx~ in th~ xange o~ about 20~C ~o about 30~C, and preferably in ~he xange of 65C to 75C, during the treatmen~ o~ the coated copper circuitryO

71f7 --6~
~he period of time for which the copper circult patterns coated with adhesion promoter are exposed to the action of the chromic acid solution can vary from as little as 10 seconds to as much as 5 minutes but is preferably in the range of about 1 to about 2 minutes.
The appropriate time of exposure in any given ins~ance is dependent upon the particular adhesion promoter employed, the temperature of the chromic acid solution, and the method by which the chromuc acid solution ls applied to the coated copper zircuit patterns. The most appropria~e time of exposure in any yiven instance can be determined readily by a process o~ trial and error.
.

When the treatment with chromic acid solution has been completed the treated board is rinsed with water and dried. The board so obtained is then employed toge~her with a plurality of similarly treated boards in the assembly o~ a multilayer board. This ~s carried out by arranging the boards in a properly sequenced and layer oriented stack. Sheets of partially cured polymeric resins (prepregs) such as epoxy resin and the like are interleaved between the boards. The boards and i~terleaved prepregs are usually pinned onto a caul plate which is used ~o hold the finished stack during the lamination process. When the stack has been completed in this manner a second caul pla~e is pinned on top. The stack and caul plates are then placed between the platenc of an appropriate press and subjected to pressing using pressures and temperatures in the ranges set forth herein above for a time generally wi~hin the range of about 1 ~o abou~ 4 hours. The resulting laminated multilayer board is generally subjected to a post-ba~e at a temperature of ~he order o~ about 150C before being trea~ed further.

In order to provide any required electrical conn~c-tions between circuits in the various layers of the multi-layer board, holes are drilled through the board in predetermined locations and the sides of the holes are subsequently plated with co~per, gen~rally by electroless copper plating, in order to provide the necessary electr~-cal contact between layers. ~owever, as set forth above, before the plating is carried out the holes have to be treated to remove resin smear left in the holes in the drilling process. This is accomplished using acldic solu~ions, for example, using dllute sulphuric acid. It is in this particular operation that the effect of the markedly improved adhesion between copper circuit patterns and resin substrate achieved by the process of the invention is manifested. Thus the adhesion between the copper circuit patter~s and the xesinous interleaved substrates is such that, unlike the multilayer boards hitherto known, no significant delamination has taken place during the drilling procedure. Accordingly the copper oxide or other adhesion promoter which is coated on the copper circuit pattern is not exposed to the action of the acid solution employed in the desmeariny of the holes. Hence, the adhesion promot~r in the layers surxounding the holes is not dissolved by the acid and n~
underlying copper is thereby exposed ~o that no pink ring effect is observed after the treatment with the desmear-ing acid solution.

As set forth above the adhesion promoters, with which the copper circuit patterns on the various boards are coated prior to the chromic acid treatment of the invention, can be any of ~hose previously employed in the artO Illustrative of such adhesion promoters are copper oxide, ~in and the like. ~ particularly advan~aqeous method of applying a layer of adhesion promoting copper oxide involves treating tbe copper circuit pat~ern~ with '7~

an oxidizing solutlon cor.:aining an alXall or alkallne earth me~al chlorite in a ~oncentration cf ~rom 50 grams per li~er to complete sa~uration and an al~ali met~1 hydroxide in a concentrat~on .of from 5 to 25 grams per liter. Appropriate condi ions for carrying ou~ this treatmen~ are described, for e~mple, in Landau U.S.
Patent 4~409,037-The multilayer printed cir~uit ~oards which areprepared i~ accordance with the invention are charac~
terized by markedly improved adhesion between the layers as evidenc~d particularly by the substantial absence of any pink ring effect associated with the through-holes in the finished boar~s.

The following examples illustrate specific emb~di-ments o the process and products of the inv~n~ion and the best mode currently Xn~wn ~o the inventors for carrying out the same but are not to be construed as limiting.

Pre~reatment In this and all fo~lowing examples the suxface of the copper specLmens employed was subjected to a standard pretreatment process as follows.
-The specimen was first cleaned by immersion in analkaline cleaning solutio~ con~aining a mixture of tri-sodium phospha~e, sodium carbsnate and a~ alXaline s~able surfac~ant ( 17~7 Soak Cleaner; MacDermid, Xncorporated,Waterbury,. Connect~cut : 8 oz. per gallon~ for 5 mimltes at 66~C. Th~ specime~ was then rinsed wi~h watsr and etched for 3 minutes in an aqueous ~;olutior~ o~ sodiwn 30 persulfa~e [G-4 MicroEtch*: ~acDenTid Incoxporated: 16 oz. p~r gallo;l~ at 21C in order ~o increase the sp~c:iflc surface ar~a. The spes:inen was ri~sed with wa~cer, then * Trade Mark 7~

with 10~ su~furic acid to remove any residues and flnally rinsed with water and dried.

Demonstration o~ Increa ed Adhesion Produced in Accordance with the Process of the Inve~

A number of pieces of copper foil (4" x 4"~, after pre~reatmen~ as described above, were csated with copper oxide by immersion for 4 minu~es in an aqueous sslutio~
containing 130 g. per liter of sodium chlorite, 25 g. per liter of sodi~m hydroxide and 5 g. per liter of trisodium phosphate at 71C, Half of the specimens were thereafter immersed for 45 seconds in a solution containing 2 g. per liter of chromium trioxide at 71C. All the specimens were finally rinsed with water and dried. One of the speclmens which had been subjected to ~he chromic acid ~reatmen~ was lmmersed in 10% sulfuric acid and it was observed that a time of 6.5 minutes of immersion was required to remove th~ copper oxide coating and leave clean copper exposed. In contrast, a spscimen which had not been sub~ected to the chromic acid treatment reguired less than 35 seconds to remove the copper oxide layer.

Multilayer boards were prepared under identical pressing conditions ~350 psi at 177C for 1 hour) from a stack of the chromic acid treated speeimens interleaved with epoxy prepregs and from a stack of the specimens not treated with chromic acid interleaved wit~ epoxy pre-pregs. The adhesion values for the layexs in the laminate were determined before and after sol~er immersion (10 seconds at 288C~ using a pull tester to measure the tension required to pull 1 inch strips of 30 - foil from the epoxy layer at a 90 angle at a ra~e of 1 inch per minu~e. The resul~s of this test were as follows.

73L;~
~10~
Treatment of Before A~ter Sold~r Copper _ ___ Solder Immersion . . .
copper oxide alone 5,8 lbs/in. 4.8 lbs/in.
copper oxide ~ chromic acid 7.1 lbstin. 5.B lbs/inO

It will be see~ from the above data that the adhesion values for the copper ~oil treated in acco~danc~ with th~
invention were significantly higher than or that which had no~ been so treated.

A number of pieces of copper foil ~4" x ~'~), after pretreatment as described in Example 1, were coated wi~h a tin layer by immersio~ for 2 minutes in a~ immersion tin solution [XD 25~2, MacDenmld Incorporated] a~ ambien~-temperature ~circa 20UC). ~alf o~ the tin coa~ed foils so prepared were additionally trea~d by immersion for 2 minutes in a solution of 2 g. of chromium trio~ide per liter of water at 71C~ All ~he spe~Lmens were finally rinsed with water and dried.
.
Using the conditions and proc~ss described in ~xample 1, multilayer boards were prepaxed by lamina~ion between interleaved epoxy prepregs using the chromic acid treated sheets in one board and the sheets not treated with chro~.ic acid in the other. Adhesion values were determined, using the pull tes~er procedure de~crib~d i~
Example 1~ both before a~d after solder immersio~. The results were as follows.

Treatment o ~sfore After Solder Cop~er Solder 3~33L5~
T~n coating alone 2.G lbs/in. 0.5 lbs/in.
Tin coating + chromic acid 5.2 lbs/in. 3.9 lbs/in.
* Trade Mark 7~

It will be seen from the above d~ta that the adheslon values for ~he copper foil treated in accordance with the invention were markedly higher than for that not ~o treated.

The procedure described in Example 1 was r~peated exactly as described except that the solution of chromium trioxide was replaced by an agueous solution of 1.5 g. of chromium trioxide and 2 g. of sodium dichromate per liter. It was found that ~he time reguired to remove the copper oxide coating fro~ the foil treated with chromic acid was 3 minutes 45 seconds whereas less than 3S
seconds was required in the case of the oil which had not been treated with chromic acid. The adh~sion values of ~he multilayer boards prepared from the chromic acid treated copper oxid~ coated foils and from the foil not treated wi~h chromic acid were as follvws.

Treatment of Before After Solder ~Per _ Solder E~@~ a___ Copper oxide alone 5.8 lbs/in. 4.8 lbs/in.
Copper oxide ~ chromic acid 6.8 lbs/in. 5.5 lbs/in.

Twelve ~12) copper clad epoXy boards ~12" x 18") having a thick~ess of 0.04 inches were subjected to the pretreatment process set forth in Example 1. Each of the boards was then immersed for 4 minut~s in an aqueous solution containing 130 g. per liter of sodium chlorite, g. per liter of sodium hydroxide and 5 g~ per liker o~
tr~sodium phosphate at 71C. Therea~er half of the pieces were immersed for 45 seconds in an aqueou~
solution of 2 g. per liter of chromium ~rioxide~ All ~he 7~ 6 pieces were ~inally rinsed and drled. Each of the pieces so obt.ained was laminated b2tween three layers o~ epoxy prepreg on ei~her side Oc ~he board with a sheet of untreated copper foil ~o~ g the outer layer of the laminate. The laminating conditions were those described in Example lo Each of the mult~layer boards so obtained was drilled with an identical pa~tern o~ through-holes each of 0.04 inches in diameter. The drilled boards were then processed through a typical commercial des~ear and plating cycle using ~he f ollowing steps.

In each of the followinq steps the drilled board was immersed in the particular solution or solve~t employed and, excep~ be~ween steps (g) and Ih~, the boards were rinscd wi~h water after each s~ep.
(a) organic hol~ conditioning solv~nt C9204-MacDermid: ~ull strength~ at 38C for 5 minutes.
(b) alkaline permanga~a~e desmear solutio~ ~9275o MacDermid: 60 g./li~re] at 74C for 15 minutes.
(c) 5~ aqueous sulfuric acid at 43C for S
,~ minutes.
(d) acidic fluoride glass etch solution ~9278:
MacDermid: 55 g./litre~ at 43C ~or 4 minutes.
(e) 10% aqueous sulfuric acid at 20C ~or 2 minutes.
- (f) alXaline hole conditioner/cleaner [9268:
MacDermid: 60 g./litre] at 66C for 5 minutes.
(g) acid persulfate microetchant [G-4: MacDermid 120 g./litre] at 35C for 2 minutes.
(h) 10% aqueous suluric acid at 20~C for 1 minut~. -(i) acidic chloride activator pre-dip [9008:
MacDennid~ a'~ 29-C for 1 minute.
l~) acidiG palladiuEI activator C~3a::~ivate 10:
MacDermid~ at 29-C ~s~r 5 minu~es.
* Trade Mark ~13-(k) acidic accelerator [9074: MaoDermid] at 29~C
for 2 minutes.
(1) alkaline elec~xoless copper bath [Macudep 52:
MacDermidl at 43C for. 30 minutss. .

Finally each board was electroplated with copper using an aoidic elect~olytic copper bath [MacuSpec:
MacDermid] ~or 54 minu~es at 29~C using 20 ASF.

At th~ end of the processing each of the boards was examined for presence of a pink ring o~ the innermos~
board by grlAding away the outer layer of copper and dielectric. It was found that those in~er boards which had been subjected to chromic acid trea~m~n~ in accordo ance wi~h the invention showed significantly smaller pink rings (average width of ring 8.3 mils) as opposed to~
those which had ~ot received the chromic acid trea~ment (avera~e width of ring 23.4 mils).

~AMP~E S

Twelve pieces o copper clad epaxy board ~12" x 18'~) having a thicknP~s of n .04 inches were scru~bed m~chanically and then subjec~ed to the pretrea~me~
descri~ed in Example 1. Eaoh of the boards was then immersed or 2 minutes in a commercial immersion ~in bath [XD-2552-T : MacDermid : full strength] at 20-C. Six of the board~ so ~reated were then immersed or 2 minute~ in an aqueous solution containi~g 2 g. per li~er of chromium trioxlde. All of ~he boards were rinsed and dried be~ore being incorporated individually in a mul~ilayer board i~
which the t~eated board was sandwiched between two lAyers each o~ 3 s~eets of epoxy preprey with copper foil shee~s on the outer surfaces~ The lamina~ing condi ions wer~
t~ose set forth in Example 1, Each of the multllayer boards so prepared was drill~d with an ide~tical pa~ern of holes o~ diameter 0.0~ inches. Piece o~ the variou~
* Trade Mark ~ ~7~

boards were then subjected to immersion in molten solder for 10 seconds at 288C. None of the boards having an inner board treated with the chromic acid suffered delamination under these conditions whereas ~wo out of three of the other boards not so treated suffered severe delamination~

Claims

WHAT IS CLAIMED IS:

1. In a process for the preparation of a multilayer printed circuit board which comprises the steps of:
providing a no-conductive substrate clad on each side with a copper circuit pattern layer;
applying a coating of adhesion promoter to the copper circuit pattern layers; and incorporating the resulting board and one or more additional boards prepared in the same manner into a multilayer board by interposing layers of semi-cured polymeric non-conductive material between said boards and laminating the assembly using heat and pressure;
the improvement which comprises exposing the copper circuit patterns coated with adhesion promoter to the action of an aqueous chromic acid solution prior to incorporation of said boards into the multilayer board.

2. A process according to claim 1 wherein the adhesion promoter is a layer of copper oxide.

3. A processing according to claim 2 wherein said copper oxide is applied by contacting the copper circuit pattern with an aqueous solution containing an alkali metal chlorite or alkaline earth metal chlorite at a concentration of from 50 grams per liter to complete saturation and sodium or potassium hydroxide at concentration of from 5 to 25 grams per liter.

4. A process according to claim 1 wherein the adhesion promoter is a layer of tin.

5. A process according to claim 1 wherein the chromic acid solution is obtained by acidifying an aqueous solu-tion of an alkali metal bichromate to a pH in the range of about 1.5 to about 2.5.

6. A process according to claim 1 wherein the chromic acid solution is at a temperature in the range of about 20°C to about 90°C.

7. A process according to claim 1 wherein the substrate with copper circuit patterns coated with adhesion promoter is immersed in the aqueous chromic acid solution for a period of about 10 seconds to about 5 minutes.

8. A process according to claim 1 wherein the chromic acid solution is applied to the copper circuit patterns coated with adhesion promoter by spraying said solution on said patterns.

9. A process according to claim 1 wherein the solution of chromic acid contains from about 0.1 to about 50 grams per liter of the said acid.

10. In a process for the preparation of a multilayer printed circuit board the steps comprising:
providing a plurality of non-conductive substrates clad on each side with a layer of copper;
applying a relief resist image of a circuit pattern to each copper layer and etching away the exposed copper;

removing the resist and applying a coating of copper oxide to the exposed copper circuit pattern;
exposing the copper oxide-coated copper cir-cuit pattern to the action of an aqueous chromic acid solution; and thereafter incorporating said plurality of boards so prepared into a multilayer board by interposing layers of semi-cured polymeric non-conductive material between said boards and laminating the assembly using heat and pressure.

11. A process according to claim 10 wherein the layer of copper oxide is applied by contacting the copper circuit pattern with an aqueous solution containing an alkali metal chlorite or alkaline earth metal chlorite at a concentration of from 50 grams per liter to complete saturation and sodium or potassium hydroxide at a concentration of from 5 to 25 grams per liter.

12. A process according to claim 10 wherein the copper oxide-coated copper circuit pattern is immersed in the chromic acid solution at about 20°C to about 90°C for a period of about 10 seconds to about 5 minutes.

13. A process according to claim 10 wherein the chromic acid solution is obtained by acidifying an aqueous solution of an alkali metal bichromate to a pH in the range of about 1.5 to about 2.5.

14. A process according to claim 10 wherein the amount of chromic acid present in the solution thereof is from about 0.1 to about 50 grams per liter.

15. A multilayer circuit board prepared in accordance with claim 1.

16. A multilayer circuit board prepared in accordance with claim 10.