Note: Descriptions are shown in the official language in which they were submitted.
~ ` ~
:
.
SZ~
!
.'
'.'' '
- ''
BACICGROU~ID OF THE I~r~F.?~T1011
This i.nvention relates to a methad and apparatus for
I operating an electroless copper plating bath
¦ solution and specifically to a method and apparatus for
¦ operating sn electroless copper plating bath
¦ solution by utilizing the mixed potential of the bath.
¦ 1:1 ctroless copper plating baths are basically
¦¦ co~prised of a reducible co?oer salt and a reducing agent.
These plating baths are desi~ned to deposit useful copper
10 eoatin~s only on parts to be plated "lld to avoid simuitaneous
de?osition on the sides and botto~ o~ tlle platino tanlc and/or
on areas of the part:s ~hele plating i~; not desircd. Because
'.'' I -1 1
l~LlZ523 1 1
' I . ,
' I . .
¦ elcctroless copper plating baths comprise a homogeneous
2 ¦ solution of reducing agents and metal salts, it is difficult
3 1 to maintain a heterogeneous reduction to the metal only on
4 ¦ desired surfaces and to avoid homogeneous reduction which
decomposes the plating solution.
6 A~rlong the reducing agents useful in electroless
7 copper plating solutions are formaldehyde, borohydrides,
8 and amineboranes (U.S. Patent No. 3,870,526).
9 Substances called stabilizers are added to an
electroless copper plating bath solution to prevent so1ution
11 decompositiarl and to control unwanted plating. Many
12 compounds have been proposed as stabilizers, including cyanide
13 compounds such as sodium and potassium cyanide and nitriles
14 such :-s lactonitrile; 5 membered heterocyclics such as _
thiazoles and isothiazoles, e.g., 2-mercaptobenzothiazole;
16 and sulfurated potash. Otker stabilizers are disclosed in
17 U.S. Patent No. 3,607,317 (polyalkylene oxides), U.S. Patent
18 No. 3,095,309 (cyanide compounds), U.S. Patent No. 3,361,580
19 (sulfur compounds), U.S. Patent No. 2,938,805 (oxygen), BRD
Auslegenschriften 1~255,436 and 1,255,437 ~selenium
21 compounds), and French Patent Nn. 1,553,375 (mercury
2? compounds).
23 Electroless copper pl~q~ing baths may also
24 cont:ain other additives such as sequestran~s or complexing
agents, ~uctility agents, or surfac~ants. The selection of
2~ complexing agents is wèll within the ability of ~hose skilled
27 ~n ttle art. Illustrative copper ion complexing agents include
' ' , ' , ' ' ~ ' ' ~ :
.. ' ' ~
12523
' ammonia and organic complex-forming agents containing one or
.~ more of the following functional groups: primary amino group
` ~ ~-NH2), secondary amino group (~NH), tertiary amino group (~N-),
~ amino group ( =N~), carboxyl group ~ -COOH), and hydroxy group
,- ( -OH). Cahill U.S. Patent No. 2,874,072, for example, describes
complexing agents which are tartrates and salicylates used in
the presence of stabilizing amounts of carbonates. U.S. Patent
No. 3,075,856 discloses complexing agents which are ethylene-
amineacetic acids selected from the ethylenediaminetetraacetic
acid, diethylenetriaminepentaacetic acid, and 1,2-cyclohexylene-
`~ diaminetetraacetic acid. In U.S. Patent No. 2,938,805 there
.:;
is described a family of complexing agents including triethanol-
amine, ethylenediaminetetraacetic acid, sodium potassium tartrate,
ammonium hydroxide, and others. Complete details concering
the use of such complexing agents are shown in the examples
of these patents. Still more details concerning copper ion
complexing agents and their use may readily be found by those
skilled in the art by reference to standard works, for example,
William Goldie, METALLIC COATING bF PLASTICS, Volume I,
Electromechanical Publications, Limited, Middlesex, England,
1968. Also useful as complexing agents are Quadrol~, which is
the tradename for a preparation of N,N,N',N'-tetrakis-(2-hydroxy-
propyl)ethylenediamine (U.S. Patent No. 3,119,709), and Chel~
DM 41, which is the tradename for an aqueous 41% solution of
:..
the sodium salt of hydroxyethylethylenediaminetriacetic acid.
,:'
., .
, . .
',
~ 3-
~ .
.~
~. 1 ~ 52~ j -
~,. l
'~'.'-. . .
~ ~ . , ' ' .
. ~ 1 Rochelle salt, the mono-, di-, tri-, and tetrasodium
2 salts of ethylenediaminetetraacetic acid (U.S. Patents Nos.
;~ ~ 3 3,093,509 and 3,119,709), nitrilotriacetic acid and its
~- 4 alkali salts (U.S. Patent No. 3,326,700), gluconic acid (V.S.
Patent No. 3,093,509), gluconates, and triethanolamine are
6 preferred as copper ion complexing agents. Commercially .
available glucono- ~-lactone and modified
8 ethylenediamineacetates are also useful and, in certain
instances, give even better results than the pure sodium
ethylenediaminetetraacetates. One such material is
11 N-hydroxyethylethylenediaminetriacetate. Other materials
12 suitable for use as cupric complexing agents are disclosed in
13 U.S. Patents Nos. 2,996,408 and 3,075,855.
14 The duc-ility of electrolessly deposited copper can
be enhanced by having present in the bath solution an
16 extraneous ion such as an element selected frcm vanadium, i -
17 arsenic, antimony, bismuth, or mixtures thereof (U.S. Patents
18 Nos. 3,615,733 and 3,310,430). Suitable sources of vanadium
; 9 and arsenic containing ions, i.e., ductility agents, are the
oxides of such elements, as well as organic and inorganic acid
21 watcr soluble salts of such elements, e.g., the vandates and
22 arsenates of the metals of Groups IA and IIA of the Periodic
23 Table of Elements, and aiNmonium. The sodium, po~assiu~, and
24 ammonium salts are pre'erred. Sources of antimony and bismuth
containing ions are the oxides of such elements and water
26 soluble organic and inorganic acid sal~s of such elements~
27 including the sulfates~ nitrates, halides, tartrates, and the
,,., '
' _4_
:
~3LlZ6Z3 ~ ~
1 like. Other ductility agents such as sodium cyanide and the
2 polyalkylene oxides are also useful (UOS. Patents Nos.
3,095,309 and 3,607,317).
4 The performance of an electroless copper plating
bath can be improved by the addition thereto of less than
6 about 5 g/l of certain surfactants. Such surfactants include
7 organic phosphate esters and o~yethylated sodium salts, and
~ixtures thereuf. Preferred surfactants are alkylphenoxy
9 polyethoxy phosphate esters. Such a surfactant may be
obtained under the tradename Gafac ~E-610. .
11 Broadly speaking, in actual practice an electroless
12 copper plating bath will contain a copper salt in solution;
13 chemicals w"~ch act as reducing agents, fre2ing elec~rons to
14 convert copper ;ons to copper at sites where metal surfaces
or clumps of metal atoms are present as catalysts; one or more
16 sequestrants which hold the metal in solufion by inhibiting
17 the formation of insoluble metallic salts; and other components
18 which influence the plating rate and the crystal structure of .
19 the depositing metal and hence the properties of the
electrolessly plated metal. A typical electroless copper
21 plating bath may be comprised of a copper salt, a base, a re- ¦
22 duc~ng agent, a sequestrant, a ductility agent, a stabilizer9 j .
23 and a surfactant. A preferred solution for electrol~ss platin~
24 of copper may be comprised as follows: ¦
~1) about 0,02 - 0.G8 molar CuS04 or CuC12;
26 (2) NaOH or KOII in an amount such tha~ the
27 pH of the solu.ion is in ~he ranKe of
28 j about 11 to 14;
I _5_ j
23
(3) abou~ 0.01 - 3.0 molar reducing agent
such as formaldehyde or one of the
hydrides, borohydrides, or amine
boranes;
(4) about 0.02 - 0.4 molar sequestrant,
or complexing agent, such as Rochelle
salt, Na4EDTA, Quadrol~, a polyalkanol
amine, or an amino acid;
(5) about lO 9 to 10 1 g/l of a ductility
agent such as NaCN;
(6) about 10 9 to 10 3 g/l of a modifier
or stabilizer; and
(7) up to about 5 g/l of a surfactant such
as an alkylphenoxy polyethoxy phosphate
; ester.
An electroless copper plating solution contains a number of
components added intentionally, such as those mentioned ahove,
and others accumulated unavoidably, such as by-products.
Additional informztion regarding specific constit~ents
of electroless copper plating baths can be found in commonly
- assigned U.S. Patents Nos. 3,607,317, 3,615~ 737, 3, 645 9 749,
and 3,650,777.
The basic operating conditions of an electroless copper
plating bath are maintained by chemical analysis and appropriate
additions of chemical components. The plating bath has several
; variaholes, or parameters, that may be
.,
~ 5Z3
I
''';, , . .
1 measured and controlled, such as pH, copper ion concentration, ,
2 reducing agent concentration, ductility agent concentration,
3 sequestering agent concentration, stabilizer concentration,
4 temperature, plating rate, loading (total area of coppe~ being
plated as compared to the solution volume), type and degr~e of
6 agitation, ~he rate of removal of H2 (generated by the ~
reduction reaction), accumulation of by-produc~s, and trace
8 amounts of contaminants which affect the rate and quality of
pla,ing. Chemical analysis and manual additions of
replenishers are possible; however, manual maintenance of .
11 optimum conditions is difficult ~o achieve as loading and
12 other conditions change.
13 ~utomatic control of some or all of the measurable
14 parameters of an electroless copper plating bath solution is
highly desirable. Many bath parameters can now be measurPd
16 and controlled automatically. Control procedures and
17 controllers have been developed by Dr. Gunter Hernmann
18 lGalvanotechni~ 65, 950 (1974)] and others.
19 In addition to the above-mentioned para~eters, an
electrical potential of a bath solution termed the "mixed
21 potential" can also be measured. Milan Paunovic [Plating,
22 68, 1165 (1968)] has disclosed how the mixed potential l .
2 uniquely describes the condition of an electroless pla~ing
24 fiolution. Tuck~r [Amer. Electroplaters Soc., Merrimac Valley
Sec., March 10, 1972J has described the use of mixed potential
26 as a monitor of general bath conditions.
-7~
IllZ5Z3
uring the developmcnt and utilization of electroless¦
2 I copper plating baths, many attempts have been made to use the
3 ¦ mixed potential as a controlling parameter, particularly as a
4 I parameter to monitor and control stabilizer concentration or
¦ activity. The control of stabilizer activity itself is, in
; ¦ particular, inherently difficult because the quantities of.
stabilizer employed are too small for convenient analysis.
8 ¦ Only cyanide stabilizers have previously proved amenable to
¦ control. (See G. Herrmann BRD Patentschrift 2,064,822.)
: 10 ¦ In general, attempts to utilize the mixed potential .
11 ¦ to monitor and con~rol bath solution stabilization have been
12 uniformly unsuccessful because all bath parameters, not just
13 the stabilizer concentration, influence to some degree the
14 mixed potential of a bath. ~mall changes in the mixed
; potential may be due to variations in one or more bath
parame~ers that would not affect the overall bath operation.
17 I On the other hand, minute additions of stabilizers may
18 ¦ completely inhibit plating and darken and discolor previously
: 19 deposited metal, or ~ay fail to prevent spontaneous
¦ decomposition of the plating solution by uninhibited, wild
21 ¦ plating. These additions would be accompanied by wild
-~ 22 l fluctuations in mixed potential. In sum, use of the mixed ;
23 ¦ potential as ~ control parameter for stabilizer activity has .
24 l been heretofore rejected because of inability to establish a
l responsive relationship between the mixed potential and
26 stabilizer activity.
-8-
1' I . ,.
' "
:
lllZSZ~ ~ '
1 l SU~RY_OF THE INVENTION
Applicants have surprisingly discovered a novel
3 ¦ method and apparatus for operating an electroless copper
plating bath and avoiding decomposition thereof.
According to this invention, the stabilization of electroless
platihg bath solutions comprised of copper or copper and other
7 ¦ metals can be achieved by utilizing the mixed potential as the
8 ¦ control parameter. The mixed potential of the bath is
¦ measured by two electrodes~ a plating electrode and a reference
¦ electrode, and the stabilizer activity or one or more other
11 l parameters affecting the mixed potential are measured and/or
12 ¦ adjusted to maintain the mixed potential within a
¦ predetermined range. The prede~ermined range will depend on
14 ¦ several factors such as the particular plating solution or the
particular reference electrode.
16 The mixed potential can be measured by using both
17 electrodes in the bath itself. If the bath is operating
18 at high temperatures, the reference electrode will preferably
19 be located in a contiguous chamber.
21 The apparatus of this invention will be comprised of
means for measuring the mixed potential as well as means for
22 monitoring and/or controlling other factors primarily affecting ;
23 the mixed potential such as the pHJ stabilizer activity,
24 l temperature, reducing agent activity, and copper ion
~ concenera Lon.
.
BRIEF DESCRIPTION OF THE D~ WINGS
Figure 1 represents an apparatus for measuring
mixed potential and monitoring and controlling temperature,
copper ion concentration, and pH.
Figure 2 represents an apparatus for measuring
mixed potential and monitoring and controlling temperature,
copper ion concentration, pH, stabilizer activity, formaldehyde
activity and accumulation of by-products.
.~ . .
dr/~ 9a-
1~ lllZ523 1 i -
1 DESCRIPTION OF THE PREFERRED EM~ODIMENTS
2 The ~emperature nf an electroless copper plating bath
3 solution is readily measured and control~ed ~ith conventional !
4 instr~mentation. Advantageously a non-metallic temperature
sensor mount will be employed because electroless plat~ng may
6 occur on any metallic or physically rough surface.
7 Commercially available glass electrodes may be em-
8 ployed to measure pH. The pH sensing means, which must be
9 suitable for high pH solutions, may be either a combination
pH/reference electrode or a pH electrode used in conjunction
11 with a reference electrode. ~',any reference electrodes are
12 useful, particularly a standard cal~mel electrode (SCE) or 8
13 silver~silver chloride electrode. A silver/silver chloride
14 electrode or a similar electrode is suitable provided two
conditions are met: (a~ there must be essentially no flow
16 across the reference electrode junction, i.e., the ~icroporous
17 interface between the plating solution and the reference elec-
18 trode solution, such as a silver chloride saturated potassiu~
19 chloride solution, since the reference electrode solution
ma~ csuse spontaneous decompos tion of the plating solution ~t
21 extremely low concentrations; and (b) the activity of the
22 plating bath should be reduced below nor~al at the location of
23 the microporous interface to prevent plating of metal on the
24 interface.
The latter condit~on can be achieved by reducing the
26 tem~perature o~ the bath solution adj~cent to the refer~ncc
2~ electrode below the normal plating ~emperature or by
28 ¦ interposing a salt bridge, or both. This reduction in
Il -10~
-1 , . . .
~ 2~
. . . .
1 temperature may be accomplished by simply placing t~e
2 reference electrode in cold, stagnant platin~ solution set
3 "offstream.~
4 The Cu++ ion concen;ration can be determined with a
. colorimeter which measures the absorption of light of suitable
6 wavelen~th (red light). This measurement may be affected by
7 temperature and, with some sequestrænts, by pH, and is
8 therefore best made under standardized conditions or with
g sui.table corrections based on temperature or pH, or both.
The cyanide ion activity may be measured with a
11 I co~mercial specific ion (cyanide ion) eiectrode, provided ¦
12 ¦ that the solution activity is reduced so that plating does not
13 occur on the sensing electrode or the reference electrode.
14 Since this measurement is also dependent upon the pH and tem- i!
perature, measurement under standard conditions or with elec-
16 trical corrections, or both, is preferred.
The fGrmaldehyde concentration or activity may also
18 1 be me~sured by automated chemical analysis. One method of
19 measurement invcl~es (i) dilution ~ith a known volume of
liquid, (ii) addition of an excess of reagent which forms a
21 colored species in the presence of formaldehyde, and (iii) I ;
22 measurement of ~he color density by photometric or
23 1 colorimetr~c means.
24 l A second method involves addition of an amoun~ of
¦ acid (typically sulfuric acid) propGrtional to a sample stre~m
26 ¦I for the purpose of lowering the pH of the sample stream by a
27 I known amount. Sodium sulfite is then added in moderate exCCsS
5L~ 5Z3
.
.. i
and fixed volume and concentration. The sodium sulfite reacts
2 with the formaldehyde causing an increase in pH proportional
to the amount of fo~maldehyde present in the sample. The final
; pH is measured as t!~e determinant of the amount of for~aldehyde
I reacted. The amount of acid initially is set so that the final
6 ¦ pH after reaction is at a level convenient for sccurate
¦ measurement, taking into account the buffer points resulting
8 ¦ from various chemical species in the plating solution. Since
9 ¦ the final measurement is a pH value, it will be measured
¦ against a datum point such as the initial pH of the plating .
ll ¦ solution, the pH of the plating solution after the acid
12 ¦ addition, or some other pH value, as would be appreciated by
13 ¦ those skilled in the art. Control of solution pH and electri-
14 ¦ cal correction for deviation Crom a standard are important to
formaldehyde de~ermination by this second method.
16 As would also be appreciated by those skilled in the
17 art, other bath solution reagents can be similarly measured
18 and controlled.
19 The accumulation of by--products may be detected by
measurement of the specific gravity of the plating solution.
21 ¦ The mixed potential of the bath solution is
22 l basically the electrical sum of half cell reactions, the ;
23 l principal ones being the oxidation of a reducing agent and the .
24 l reduction of copper ions to copper. When two or more oxidation-
¦ reduction reactions occur simultaneously on the same conduc- ~
26 l tive surface, such as an electrode, each oxidation-reduction t
¦ system will strive to set up its own equilibrium with its o~n
~ -12-
"'' l l l
Z5~
,:: .
. ~.
:~ . 1 characteristic electrode potential. Since only one potential
2 can exist on a conductive .qurface, a steady state will be
: 3 reached where the electrode reaction occurs. This compromise
. 4 potential is the mixed poten~ial.
As a specific example, the equilibrium potential
: 6 I for O.lM copper in a 0.175M EDTA solution at pH 12.5 was.re- .
., 7 ¦ ported as -0.47V vs. SCE ~Paunovic, M., Plating 55, 1167
: 8 I (1968)].
. 9 l .
10 ¦ Cu++ ~ 2e ~ Cu
.. 11 1 . .
12 lThe equilibrium potential for formaldeh7de was re-
13 ¦ ported as -l.OlV vs. SCE in the same medium.
. 14 1 . . 1.
15 l CH20 ~ 20H ~HCOO + H20 ~ 1/2H2 ~ le
16 1
17 ¦ The mixed potential of these two reactions occurring
: I on a copper surface was the compromise potential -0.65V vs. SCE.
19 ¦ The mixed potential of an electroless copper plating
. 20 j bath is measured by use of two electrodes, a plating electrode
: 21 ¦ in active plating solution and a reference electrode located
22 ¦ in proximity to the plating electrode. The mixed potential of .
23 I the solution is de~eloped by the plating reaction, and this .
. 24 ¦ potential may be measured between the plating electrode and` 25 ¦ the reference electrode.
26 l The pla~ing elec~rode is a metallic electrode wlth
¦ a copper surface, which is cvntinuously plated with copper
-. -13-
'' l
;`~
'~, .
by the electroless copper solution. This electrode may be ,
2 made of a substrate metal such as platinum, overplated with
3 copper. Nearly any me-allic electrode immersed in an electro-
` 4 less copper plating solution is soon covered by copper~ Many
reference electrodes are useful for measuring the mixed poten-
6 tial, a calomel or silver/silver chloride electrode being
7 preferred.
8 If a silver/silver chloride reference electrode is
9 employed, it may be either a single or double junction type.
It would be advantageous to employ a double junction electrode
ll to reduce the incidence of contamination of either the plating
12 solu~ion or the electrode solution.
13 There are certain practical problems associated with
14 the measurement of mixed potential. The measurement is carried
out in active solution, and in the case of some types of
16 electroless copper plating solutions, the active solution is
17 at an elevated temperature such as about 40 to 95C. If under
18 these conditions the reference electrode is adjacent to the
l9 plating electrode, metal will usually deposit on the re.erence
electrode, rendering it inactive; if the reference electrode
21 is placed at a distance, the measurement may be unstable. A
~-22 I satisfactory measurement is achieved by using a reference
23 l electrode, preferably a double junction electrode, in a
24 ¦ separate chamber containing cooler, less agitated solution and
1 25 ¦ connected by a shor~ con~inuous liquid path to active solution ¦
26 adjacent to the plating electrode. The liquid ~ath may vary and
A 27
-14_
:' 11 1 1 .
: ,1 "
;' '
ll lllZ5Z;~ ~ I
1 may typically be 5-10 cm. The length of the path is not
particularly critical; however, excessive length should be
svoided.
4 The reference electrode chamber may be thermally
isolated from the plating electrode by a thermal barrier such
6 as an air gap or a foam spacer. The double junction electrode
7 may contain a silver/silver chloride ~unction in a chamber
8 communicating via a porous plug with a second chamber
9 containing a solution of salt which is non-reactive with the
¦ plating solution, which second chamber communicates via a .
11 second porous plug with plating solution. The plating solution
12 in the reference electrode chamber will preferably be of
13 reduced plating activity due to cooling and restricted
14 circulation.
The mixed potenti~al of a bath does not usually
16 respond in a linear manner to changes in concentrations of
17 solution components. For example, formaldehyde may have
18 greater effect in Quadrol sequestered solutions than in EDTA
19 sequestered solutions. In either solution, the mixed
potential increases, i.e., becomes more negative, initially as
21 the formaldehyde concentration increases; however, the mixed
22 potential becomes insensitive to further changes if
23 formaldehyde is present "in excessl~ so diffusion rates
24 ¦ predominate or if Cu++, ~or example, ~s not ~ncreased to
I sust~in a faster reaction rate.
In accordance with the preserl. invention~ the mixed
¦ potential of an electroless copper plating batll is to be
I
.. ~ -
. . .
, - ~ .
. ' , ~
.,, , ~ .
lilZ5Z~
1 maintained within a predetermined range by monitoring and/or
2 adjusting certain parameters. The predetermined range can
3 vary greatly and wiLl depend upon the particular bath solution
and reference electrode u;ilized. It has been found that for
commercial operations the desired mixed potential for an
6 electroless copper plating bath will be a value from about
7 -200mV to -1.5V vs. SCE and from about -155~ to -1.455V
8 vs. a silver/silver chloride electrode. The mixed potential
~ is advantageously maintained at a value from about -600nV
to -850mV vs. SCE and from about -555mV to 805mV vs. a silver/
11 silver chloride electrode.
12 Preferably, the mixed potential is maintained at a
13 value from about -630mV to -7GOmV vs. SCE and fr~ about
14 -585mV to -715mV vs. a silver/silver chloride electrode.
As one skilled in the art would appreciate, a given
16 value versus a calomel electrode translates to respective
17 specific values ~ersus other reference electrodes. Values
18 versus calomel and silver/silver chloride electrodes differ
19 by 45mV, and, for example, the value -650mV vs. SCE corres?onds¦
to -605mV vs. a silverJsilver chloride electrode. Other values
21 I vs.SCE given herein therefore correspond similarly to mcasure-
22 j ments versus silver/silver chloride ar.d other refererlce elec- ;
23 ~rodes. ¦
24 In actual practice the mixed potential will be main-
25 l tained with a range centered on the desired value, deter~ined
26 I for the specific solution e.r.ployed, prcferably from about
27 ¦ ~ 5m~ to ~ 50mV of tha~ value, mor~ preferably within ~ lOmV
~ I or ~ 25mV. For example, a specific EDT~-based electroleSS copp~
; I -16-
. ~ '
`~: ~ ..
:::
~ . ~ : l
1 plating bath having a calomel reference electrode may be
2 maintained at -675mV, - 10mV or - 25mV.
3 The m'xed rotential is the determinative or
4 controlling variable of this invention. Other bath parameters
are monitored and/or adjusted, and some of the other parameters
6 are adjusted primarily to affect the mixed potential. For
7 example, it has been found that in a bath where the mixed
potential has become more positive then desired, e.g., has
gone to about -600mV fro~ the range of about -680mV ~ 20~V,
the mixed potential can be brought back by raising the pH or
11 adding reducing agent such as formaldehyde. Similarly, where
12 the mixed potential has beco~e more negative than desired, the
13 incremental addition o~ stabilizer has returned the mixed
14 potential to the desired range.
The apparatus of this invention monitors and/or
16 adjusts several parameters of an electroless copper plating
~ bath to maintain the mixed potential within 2 certain range.
18 ~ As mentioned above, virtually all bath parameters affect the
19 mixed potential. It is desirous from a practical point of
view to monitor and/or control only those parameters having
1 the greatest effect upon the mixed potential, parameters such :
22 l as temperature, copper ion concentration~ reducing ag~nt acti-
¦ vity, stabilizer activity, pH, etc. Advantageously the app~r~
¦ tus will measure the bath mixed potential and will monitor and ¦ ¦
~! control wi~hin predeter~nined ranges the temperature, Pll,
2' ' copper ion concentration, and reducing agent concentration-
Also, the apparatus may measure the mixed potential and monitO 7
17_
,'' , ~
. ~
: ~ ~ .~
gL112~iZ3
i~ l
; i and automatically control within predetermined ranges the
? ¦ t~mperature, pH, and copper ion concentratinn. Preferably
3 1 the apparatus will measure the mixed potential, will monitor ~
¦ 4 l and automatically control the temperature, copper ion concen_ ¦
¦ tration, and pH, and will monitor and/o_ ad~ust the stabiliæer
6 ¦ or cyanide ion activity and reducing agent activity to main-
7 ¦ tain the mixed potential within a certain range. In one
embodiment, the apparatus will measure the mixed potent~al and j
' ¦ adjust the reducing agent or the stabilizer activity, or both,
¦ to maintain the mixed potential within a predetermined range,
and will monitor and autom2tically control the temperature,
12 1 copper ion concentration, pH, and cyanide activity.
13 The apparatus can sense or monitor the various bath ¦~
14 parameters in the bath itself. However, preferably a sample
stream of active solution will be drawn from the bath and that
16 stream will be analyzed to determine the values of the desired
parameters.
18 The parameters affecting the mixed potential can be
i9 ¦ measured by the instrumentation described above. The I
¦ apparatus of ~he invention will be comprised of respective
22 means according to which selected parameters are monitored
and/or controlled. A device for monitoring and autom2ticallY
23 controlling temperature, copper ion concentration, and pH will
2~ be comprised of means for measuring and controlling each of
.. ;,. ~5 / . ,'
26 / i
/ - , ~
. I -18- 1 ~
', I I ~ '
.
,, . ~
- ~ ~
: ~lZ5~3
'',' , .
l thcse parameters as well as means for measuring the mixed
2 po~ential. As shown in Fig. 1, such an apparatus is
¦ 3 comprised of pump 1 to draw a sample stream from plating tan~
4 2. The stream passes through valve 3 to chamber 4 containing
mixed potential plating electrode 5. Adjacent to chamber 4 is
' 6 another chamber, chamber 6, containing reference electrode i,
¦ 7 which in conjunction with the plating electrode S measures the
8 mixed potential of the plating solution. Chamber 4 may also
contain thermistor 8, which thermistor may alternately be
situated at position 9 on standpipe or degasser 10. The stand-
11 pipe 10 functions to penmit gas bubbles to escape from the
12 sample stream before the stream passes to analytical sensing
13 means.
14 The sample stream then passes fro~ chamber 4 through
- lS heat exchanger ll to a second heat exchanger 12 and then to
16 chamber 13 having colorimeter or copper cell 14 for measuring
; copper ion concentration. The sample stream passes from
18 chamber 13 to chamber 15 which has another thermistor 16 and
19 combination pH/reference electrode 17. The sample stream
returns to the first hea~ exchanger 11 and then to the plating
21 ~ tank 2. The second neat exchanger 12 is cooled by cooling
22 I water from an external source.
2~ ¦ Another embodiment of this invention comprised of
24 I means for measuring temperature, copper ion concentration, pH,
25 ~ ¦ cyanide ion (or stabili~er) activity, for~,aldehyde acti~lity,
¦ and accumulaticn of by-products, is shol~n in Fig. 2. Here, a ¦
27 l sample stream of pla.ing soluticn is dra~n from plating tank
. ~ 1 , . . .
' ' -
.
-~ ~ ~ -
I
: ~ SZ~
-:~ l ` . .
l l 18 by means of pump 19. The sample stream passes through valve ,
2 ¦ 20 to chamber 21 containing mixed potential plating electrode
~,- 3 ¦ 22. Adjacent to chamber 21 is chamber 23 which contains
¦ 4 ¦ reference electrode 24. Chamber 21 may also contain
,; 5 ¦ thermistor 25, which may alternately be inserted at position
, 1 6 1 26 on standpi.pe or degasser 27. ,
' l 7 ¦ From chamber 21 the sample stream passes
8 1 successively to heat exchangers 28 and 29 and then to chamber
, 9 ¦ 30, which contains copper cell 31 and cyanide ion electrode
, lO 32. The sample stream then goes to chamber 33 which contains ,
', ll a second thermistor 34 and pH/reference electrode 35. Part of
' 12 the sample stream then passes to chamber 36. Chamber 36
j 13 contains density sensor 37 for determining the accumulation of
. 14 ¦ by-products. The remainder of the sample stream passes to
¦ reagent pump 38 which pumps the plating solution to a point
16 1 where first sodium sulfite'and then air are added, and then to
18 l mixing coil 39 where the stream is contacted with an acid. The
l9 l sodium sulfite and acid are supplied from cGntainers 40 and 41,
~' ¦ respectively. The partial sample stream containing acid and
2~ 1 sulfite passes from the mixing coil 39 to chg,mber 42 where its
pH is measured by pH/reference electrode 43. An increase in :
, 22 pH will be proportional to the amount of formaldehyde present. ¦ .
,, The partial sample stream in chamber 42 wilî normall~
` pass to waste through pipe 44. However, when the stream is
26 compatible wi~h tl~e bath solution, the stream may be returned
27 l to the platin~ tank 18 either directly or in admixture with the
~ I stream from chamber 36.
'' li -20-
' 11 ' ' ~
. ', '1 , " . , .
.
:
~ I ~11Z5Z3
i; 1 The plating solution sample stream passçs from cham-
2 ber 36 to heat exchanger 28 and then back to the plating tank
3 18. The second heat exchanger 29 is cooled by externally
4 supplied water.
The apparatuses in Figs. 1 and 2 are also comprised
; 6 of means of automatically registering and utilizing the signal
7 generated by the sensors. Each of the sensors produces a
; 8 millivolt signal which is received by a poten~iometer-type
~ 9 device such as a digital adjustment potentiometer. This latte
I ~ 10 dev'ce will receive the signal produced, ascertain whether the
11 signal falls within certain limits, and will cause an appro-
12 priate response, if warranted. .
13 Similarly, the apparatus of this invention will in
., 14 general comprise means for measuring values and means for
! ' 15 reacting to th~ values measured. Typically, as above, the
~ 16 parameters will be sensed by an apparatus which will generate
-~ 18 an electrical signal, preferably in millivolts. Each signal
19 l generated will in turn be sensed by a potentiometer-type
i device into which certain limits have been incorporated. The
21 ! device will in turn cause an appropriate response if the sense
22 1 signal is outside a predetermined range to cause the parameter
- l to return to a proper value. Such a response may, for example,
I - 2~ l be the addition of a chemical solution to the bath, the par-
, ticular che~ical solution ha~ing a gurplus or lack of the
~ ,I particular paraMeter, Fo- ex~p~e, i~ ~he apparatus detected a
:; 2~ pH, the app~ratus cou~d cause the addition of a solution
7 ¦~ con~lsting essentially of alkali sol~tion. I
``` 1l ~21-
,-: !! I
-.~, ._.. . .
;'
.
.' '
125~
l Also, in one embodiment of this invention the mixed
; 2 ¦ potential signal may cause (i) the addition of stabilizer,
3 ¦ preferably sulfur containing stabilizer, if the mixed potential
4 ¦ becomes more negative than desired and/or (ii) the addition of
S I a reducing agent such as formaldehyde if the mixed potential
6 ¦ becomes more positive than desired,
j 7 1 In another emobidment of this invention where the
! 8 l formaldehyde is otherwise analyzed and controlled, such as
1 9 ¦ by the method and apparatus in Fig. 2~ variations in the mixed
potential may result in other responses. While a negative
' ll tending mixed potential may be used as a signal to add stabi-
: 12 lizer, a positive tending mixed potential signal may be used
13 as a signal to add a copper depositio~ accelerator or a formal-
; 14 dehyde oxidation catalyst.
~5
16 COMPARATIVE EXPERIM~IT A
17 An electroless copper plating bath comprised as
18 follows was employed:
, Copper sulfate lO g/l
Tetrakis (2-hydroxypropyl) 17 g/l
ethylene diamine
`~ 22 For~sldehyde (aq. 37% soln.) 15 ml/l ;
- 23 pH 12.8
24 Temperature 28C
Sodium cyanide 25 mg/l
26 Potassiu~ sulfide O.l~ ~g/l
27 2-Mercaptobenzothiazole 0.04 mg/l
22
'':'' I __ I
"
.
., . ' ' ,
,, , . : - , . . . .
., . . ~. .
. ~ .
1 ~ lllZ523
1 This plating bath was operated with cop~er sulfate
2 concentration controlled by a flow_through colorimeter and
3 copper additions added by a proportional control based on the
4 colorimeter measurem~nts. The tetrakis (2-hydroxypropyl)-
~i 5 ethylenediamine was not consumed in the plating reaction, and
, 6 thus a daily or weekly analysis and addition was suf~icient
' 7 for control. Formaldehyde in a standard aqueous 37~
-; 8 formaldehyde solution was added in proportion to the copper
'` ` g sulfate, and additional adjustments in formaldehyde were made
~ 10 by daily analysis. The pH was measured continuously using a .
; 11 glass electrode and a saturated calomel reference electrode.
12 Sodium hydroxide in &n aqueous solution was added
- 13 proportionally based on the pH signal. Sodium cyanide in an
14 ¦ aqueous solution was added in proportion to the sodium F
l hydroxide.
~ 16 ¦ Potassl~ sulfide and 2--mercaptobenzothiazole in
- 17 1 aqueous solution were added as sulfur stabilizers in
- 18 ¦ proportion to the elapsed time and the copper area plated.
19 ¦ The mixed potential was monitored between a grounding panel
~ 20 ¦ maintained in the plating bath and the saturated calomel
r,''; 21 ¦ reference electrode used for pH measurement. The mixed
~`; 22 potential measurement was a very unsteady signal, subject to .
23 spurious transient voltages which varied from -520 to -72GmV. .
24 The plating bath was difficult to control. Either
; 25 it plated copper on the walls of ~'ne tank and soMetimes
26 spontaneously decomposed or, if enough sulfur stabili~er was
27 added to avoid plat~ng the tan~, the depos~ted copper tended
-23-
l l
'1 ' "
' '. ' ' ' ' -
,~:
`'
: ''
.....
L125~3
,: I . .
:' 1- . .
` 1 ¦ to become dark and passivated, preventing further deposition
¦ on the work piece.
¦ 3 l The follo~ing examples illustrate, but do not li~it,
1 4 ¦ the present invention:
S ¦ EX~MPLE I.
6 The mixed potential measurement of Experiment A was
modified substituting a silverisilver chloride reference
electrode adjacent to the plating mixed potential electrode.
9 However, it was seen that when the reference electrode was
~ 10 immediately adjacent to the plating electrode, metal deposited
`, 11 on the reference electrode rendering it inactlve. .
; 12 The set-up was modified by using a double junction
13 silver/silver chloride reference electrode in a separate
14 chæmber, and satisfactory measuremen~ was achieved, The double
- 15 Junction reference electrode contained a cool solution and was
16 connected by a continuous liquid path with active solution
17 adjacent to the plating electrode. The electrodes were connecte
18 to a differential amplifier, and the signal from the amplifier
19 was used to monitor and control the addition of a mixture of an
aqueous solution containing potassium sulfide and 2-mercapto-
21 benzothiazole in a ratio of 10:1. The mixed solution was con-
23 trolled at -635+ lOmV by the additions. The plating bath
opera~ed unexpectcdly well~ that is, plating bright, smooth
27 copper wleh t p~ssivation or spontaneous decomposltion.
24-
.,. '
. ~- .
12S2J ~ I
. I
. l
. l
;: 1 1 ~ . .
2 ¦ The bath of Experiment A was operated in a tank
3 equipped with a circulation pump. One wall of the tank was
¦ lower than the others, at which point the plating solution
~ S ¦ spilled over. The pump circulated the solution from the
t 6 l bottom of the spillway bzck into the plating tank. The level
7 ¦ o f solution at the bottom of the spillway was 1/3 the height
8 ¦ of the tank, allowing oxygen to be absorbed from the air and
9 I allowing hydrogen, which is a reaction by-product, tD escape
¦ as the ba~h poured over the spillway. The mixed potential was .
11 ¦ -600mV, and the copper deposited was dark. ~he bath
; 12 ¦ passivated after a few hours operation.
; 13 ¦ The le~el of the solution at the bottom of the
-. 14 ¦ spillway was raised, reducing the amount of oxygen absorbed or
l hydrogen released, until the mixed potential reached -640mV. I
16 1 The level of solution at the bottom of the spillway
17 was approximately 3/4 of the height of the tank. The mixed
18 potential of the bath solution was then maintained at -635mV
19 ~ 10mV ~y controlling both tne height of the spillway and the
~0 sulfide additions, and the bath was operated continuously for
21 two weeks without passivation or spontaneous decomposition.
22 ~ This experiment showed that the mixed potential c n .
23 be used to control oxygen and hydrogen in the bath.
24 EX~MPLE III,
An electroless copper plating bath was prepared with
26 the following foL~ulation:
27 Copper sulfate 12.5 g/l
-i 2d Tetrakis (2-hydroxypropyl) 15.6 g/l
ethylenediamine
~ ; -25-
.: j
, ':
~'."
'.. " ,
`~. l
. I
1 ¦ ForQaldehyde(aq~ 37% soln.) 3.5 ml/l
¦ Sodium hydroxide to pH 12.95
¦ 3 l Sodium cyanide 10 mg/l
~1 ~ 4 ¦ Potassium sulfide 1.0 m~l
¦ 2-~ercaptobenzothiazole 0.1 mg/l
6 l Gafa -610 0.14 g/l
7 ¦ Temperature 53 C
8 Work load 2.5 sq. decimeters of
9 plating areatliter
- 10 This plating bath was operated by continuollsly .
11 monitoring the copper concentration with a flow-through
12 eolorimeter. Copper was added automatically in an aqueous
13 ¦ solution of copper sulfate based on the colorimeter
14 ¦ measurement. The p~ ~as ~.easured by a pH electrode and was
¦ controiled by the automatic addition of an equeous solution of
16 ¦ sodium hydroxide. Cyanide was replenished by adding sodium
17 l cyanide in an aqueous solution in proportion to the pH
18 ¦ controlled sodium hydroxide addition, and formaldehyde was
19 ¦ replenished by adding a standard aqueous 37% formaldellyde
l solution in proportion to the colorimetrical1y controlled
21 ¦ copper addition.
22 The mixed potential ~7as measured usin~ a double ;
23 junction silver/silver chloride reerence electrode as in
24 Example II, and an aqueous solution having potassium sulfide
and 2-mercaptobenzothia~ole in a ratio of 10:1 ~as added
26 automatically by a proportionally controlled p~np that was
27 to control the mixed potential at -650 to -655mV.
26-
..
~ ' ' . .:
:, , ' ' ' ~........ ' : .
'
52~
¦ 1 ¦ This particular type of plating bath operates at
~-~ 2 l elevsted temperatures and thus cannot tolerate a formaldehyde
3 l concentration in excess of 5 or 6 ml/l without spontaneous
, 4 I decomposition. In spontaneous decomposition, the bath
I homogeneously reduces all the copper ion to copper po~der and
6 l turns from blue (ionic copper) to a red slurry (copper dust)
7 ¦ and finally to water white (as all the copper powder settles
8 l to the bottom of the tank). Spontaneous decomposition is
9 1 accompanied by heavy foam buildup and hydrogen evolution.
Due to an electrical failure, the formaldehyde
solution valve failed to shut down, and the formaldehyde
- 12 ~ concentration inadvertently rose to 22 ml per liter. ,
13 ¦ Contrary to all previous experience, the bath did not
14 ¦ decompose but continued to operate normally because the pump
¦ for the solution of potassium sulfide and 2-mercaptobenzo-
16 1 thiazole, in response to the mixed potential signal, added
- 17 large amounts of the potassium sulfide and 2-mercaptobenzo-
18 thiazole stabilizers. After operating normally overnight for
19 a total of about sixteen hours, the reservoir of the solution
of potassium sulfide and 2-mercaptobenzcthiazole ran dry, and
21 ¦ the bath decomposed within a half hour.
22 ¦ E,YAMPLE IV
.: I . _ .
23 l The bath of Experiment A was operated with automatic
- 24 ¦ additions and the plating rate was monitored using chrono-
; 25 1 potentiometry. The bath was allowed to plate for a fixed
; 26 ¦ period of time, twenty minutes, on a palladium electrode.
27 l After the ~enty ~inutes, the electrode was transferred to a
, . .
-27- 1
.: . .~_ . ,
. .
1 copper sulfate solution and anodically stripped. The time
2 ¦ required to electrolytically strip the copper from the
¦ 3 l electrode at a fixed potential is a direct measure of the
1 4 l plating rate. The mixed potential of the bath was used to
1 control the addition of stabilizers as in Example II. The
6 plating rate as measured chronopotentiometrically varied
7 directly with the fluctuation of the mixed potential.
¦ EXAMP~E V.
~- 9 I An electroless ccpper plating bath was comprised as
¦ folLows: .
11 l Copper sulfate 0.036 moles/l
12 l Quadrol ~ 0.08 moles/l
14 ¦ Temperature 53 C
l Formaldehyde 0.05 moles/l (initially)
16 l Sodium cyanide 30 mg/l (initially)
17 Gafac RE-610 0.1 g/l
18 The copper ion concentrztion, pH, temperature, and
19 formaldehyde concentration were constantly monitored and
appropriate additions to t~e bath were made automatically when
21 so indicated. The cyanide ion concentration was measured and
22 controlled at -150nV using a calomel electrode and an Orion~ .
23 cyanide ion electrode. .
24 A~ounts of an aqueous solution of a sulfur stabi-
lizer, potassium polysulfide, were added whenever the mixed
26 potential, as ~easured using a double junction silver/silver
27 chloride reference electrode~ became more negative than 670
. .
; -~8-
:''' 1 " .
, .
:,:
"'"'
:
i25~3
.
". ,
,~ 1 Additional formaldehyde in a standard aqueous 37% fonmaldehyde
2 solution was added whenever the mixed potential became re
3 ¦ positive than -640mV. The bath operated well for about a
;t week, plating bright, smooth, unifonm copper ~7ithout plating
the walls of the containerO
6 ¦ Continuin~ chemical analysis of the bath solutio~
, 7 I showed that the formaldehyde concentration was maintained
: 8 ! between 0.01 and 0.04 moles/l, which is sufficient for normal
9 ¦ operation of this bath.
¦ EXAMPLE Vl.
11 An electroless copper plating bath was comprised as
` 12 follows:
13 Copper sulfate 0.036 moles/l
14 pH 12.9
Quadrol ~ 0.08 moles/l
16 1 Temperatu.e 53 C
17 l Formaldehyde 0.05 moles/l (initially)
18 1 ~afac RE-610 0.1 g/l
19 ¦ The copper ion concentration, pH, temperatur~, and
formaldehyde concentration were constantly monitored and
21 appropriate additions to the bath were made automatically when ;
~ 22 so indicated.
`~ 23 The mixed potential was measured by means of a
24 silver/silver chloride reference electrode. Amounts of an
¦ aqueous solution of potassium polysulfide were added whenever
26 ¦ the mixed potential became more ne~ative than -670mV.
27 ¦ Additional fonnaldehyde in a standard aqueous 37Z
_~9_
.. '
lZ51;~ I
1 ¦ formaldehyde solution was added whenever the mixed potential
2 I became more positive than -640mV. The bath operated well,
platin~ bright, smooth, uniform copper without platin~ the
i walls of the container.
Continuing analysis of the bath solution sh~ed that
6 ¦ the formaldehyde concentration was maintained between 0.01 and
7 ¦ 0.04 moles/l, which is sufficient for normal operation of this
1 8 bath. ,
9 ¦ EX~LE V~I.
¦ In order to show the control of formaldehyde alone,
11 ¦ an electroless copper plating bath comprised as follows was
12 ¦ employed:
13 Copper sulfate 0.036 moles/l
drol~ 0,05 mol~s/l ¦
16 Temperature 53C
17 l Formaldehyde 0.05 moles/l
18 1 Garfac RE-610 0.1 g/l
19 ¦ The copper ion concentration, pH, and temperature
l were constantly monitorcd, and appropriate addit-ions to the
21 ¦ bath w~re made automatically when so indicated. Additional
; 22 ¦ fo7.~aldehyde in a standard aqueous 37% formaldehyde solution
23 ¦ was added whene~-er the mi~ed potential became more positive
24 ¦ than -640mV.
¦ The bath operated well, plating bright, smooth,
¦~ 26 l unifonn copper without plating the walls of the container.
~,
~ -3~-
. - . .
:,, '
,-
lLl -5Z3
1 EXA~LE VIII
;` 2 In order to demonst~ate the control of bubbl~ng air
3 ¦ using the mixed potential, an electroless copper plating bath
4 ¦ was comprised as follows:
~; ' Copper sulfate 0.04 moles/l
A 6 Quadrol 0.06 moles/l
7 Fonmaldehyde 0.26 moles/l
8 Sodium hydroxide 0.36 moles/l
9 Sodium cyanide 0.0002 moles/l
¦ Temperature 24C
11 ¦ The mixed potential was controlled by bubbling air
12 ¦ through the bath sol~tion. Ilhen air was bubbled through the
13 l solution, the mixed potential as measurea against a calomel 5
14 electrode increased to -703mV. When the air bubbling was ~;
1 stopped, the mixed potential slowly decreased to -720mV.
16 ¦ RXAMPLE IX.
17 l An electroless coppe~ platin& employing the complexing
18 1 agent Chel DM 41, was .un. The bath solution was comprised as
` ~9 1 follows:
C~pper sulfate 7.5 g/l
21 Chel DM 41 22 ml/l
22 Formaldehyde taq. 37~/O soln.) 15 ml/l
23 Sodium hydroxide 6 g/l
24 Temperature 27C
26 Gafac RF-610 O.lg/l
"~ 27 /
-31
lZ523 1'
'
.
Initially the mixed po~ential was -720mV. vs. SCE
2 and the deposit ~as dar~. Sodium cyanide was added incre-
¦ 3 mentally until the mixed potential increased to -680~V vs. SCE,
4 a~ which point the plated copper became bright ant smooth,
i EXA~LE X.
6 The mixed potential is useful for controlling agita-
7 tion, When the solution is vigorously agitated, air is mixed
8 into the bath and the mixed potential increases; when the
9 a~itation ceases, the mixed potential decreases. A plating
bath was comprised as follows:
Copper sulfate 14.6 g/l
12 ¦ Rochelle salt 7,5 g/l
13 ¦ Sodil~ hydroxide 7.5 g/l
14 Formaldehyde (aq. 37% soln.) 38 ml/l
- 15 Temperature 26C
16 The mixed potential was -720mV vs. SCE initially,
17 wi~h no agitation. When an agitator was turned on, the mixed
18 l potential increased. The agitator was turned on and off
19 l intermittently as required to maintain tne mixed potential in
¦ the range of -640 to -680mV vs. SCE. The copper plated was
21 bright and smooth.
22 1 /
'; 23 1 /
24 ~
226 ~ /
~ I -32_
'' I
~1~ " ' , ,
, ~ ~
: