1230 Langmuir 1992,8, 1230-1231

Notes

Mechanism of Copper(I1) Reduction by Formaldehyde Studied by On-Line Mass

Spectrometry

Zenonas Jusys' and Algirdas Vagkelis

Institute of Chemistry and Chemical Technology of Lithuanian Academy of Sciences, Goitauto 9,

232600, Vilnius, Lithuania

Received April 16, 1991. In Final Form: October 22, 1991

Introduction The main reaction of the widely used electroless copper

plating process is the autocatalytic reduction of copper- (11) ions by formaldehyde. A somewhat unusual feature of this reaction is the formation of molecular hydrogen along with the metallic copper. The reaction stoichiom- etry as determined in' and confirmed by other researchers in a simplest form (taking into account neither the formation of methylene glycol and the dissociation of it in alkaline medium nor the chelation of Cu2+ in the presence of complexing agent) may be expressed as

Cu2+ + 2CH,O + 40H- - Cuo + 2HC00- + 2H,O + H, (1)

The mechanism of this reaction is assumed by most of the researchers to be an electrochemical one: two partial reactions, anodic oxidation of formaldehyde

c u

c u 2CH,O + 40H- - 2HC00- + 2H,O + H, + 2e- (2)

and cathodic reduction of Cu(I1) cu

CU'+ + 2e- - Cuo (3) are expected to proceed simultaneously on the copper surface, with electrons being transferred via metal. The electrochemical mechanism was proved experimentally in the 1960~,-~ by means of electrochemical modeling of reaction 1. Later studies showed the above electrochemical half-reactions to interact; i.e., the acceleration of cathodic copper deposition by anodic formaldehyde oxidation was found to take place together with the acceleration of the latter by copper cathodic reductiona5t6

The origin of hydrogen evolving in reactions 1 and 2 could be determined using a deuterium tracer. Using the isotope labeling, hydrogen was found to evolve entirely from the C-H bond of formaldehyde during the anodic oxidation of it on ~ i l v e r . ~ Thus, formaldehyde anodic oxidation on Cu may be expected to proceed by

2CD,O + 40H- - 2DC00- + 2H,O + D, + 2e- cu (2a) The overall reaction of copper reduction by formaldehyde in this case should be as follows:

(1) Lukes, R. M. Plating 1964, 51, 1066. (2) Saito, M. J . Met . Finish. SOC. J p n . 1965, 26, 300. (3) Vaikelis, A.; Salkauskas, M. Liet. TSR Mokslu Akad. Darb. 1967,

(4) Paunovic, M. Plating 1968, 5 5 , 1161. ( 5 ) VaHkelis, A.; JaEiauskiene, J. Elektrokhimiya 1981, 17, 1816. (6) Wiese,H.; Wei1,K. G.Ber . Bunsen-Ges.Phys. Chem. 1989,91,619. (7) Hoyer, H. 2. Naturforsch. 1949, 4a, 335.

8 4 (51), 3.

0743-746319212408-1230$03.00/0

cu Cu2+ + 2CD,O + 40H- - Cuo + 2DC00- + 2H,O + D,

( la) However, the gas evolved during electroless copper

deposition in CD2O solution was determined in the recent worke to contain more than 99 96 of HD. The mechanism was suggested,8 which explained the hydrogen formation in equal parts from formaldehyde and water. Therefore, a further investigation of electroless copper deposition is necessary. It seemed to us of great interest to apply an on-line mass spectrometric analysiss for studying this process under various conditions in the range of real elec- troless copper plating solutions, including also those of ref 8. The feasibility of differential electrochemical mass spectrometry (DEMS)lO for investigating the formalde- hyde anodic oxidation on Au electrode was demonstrated in the recent work."

Experimental Section Solutions and Materials. Two series of experimental runs

were carried out in electroless copper plating solution, containing (mol L-') CuS04 (0.04), EDTA (0.04), and CDzO (or CH20) (0.08) a t pH 12.0 or 12.5 and t o = 20 or 70 "C. One of them contained deuterated ("heavy") formaldehyde (CDzO) (deuterium content 98 mol % ) as a reducing agent, while "light" water (H20) as well as other materials including light hydrogen was used. The other contained CH20 and D20 (D content 99.8 mol 72 ) as a solvent as well as water-free salts of CuSO4 and tetrasodium ethylene- diaminetetraacetate; pH was adjusted with NaOD (D content 99 mol 5%) in this solution. In both cases paraformaldehyde was used for the solution preparation. Solutions were deaerated with Ar.

Analysis of the Evolved Gas. A mass spectrometer MI- 1201 (USSR) was used for the conducting isotopic analysis of the gas evolved, which was sucked through a porous Teflon membrane (thickness 5 pm) into the ion source of the mass spectrometer and detected on-line. Electroless copper plating was carried out on the outer side of the membrane covered with a -0.1-pm copper layer sputtered in vacuum (the geometric area of 1 cm2). Elec- troless copper plating was initiated by a cathodic switch of po- tential to -0.6 V (vs Ag/AgCl/KCl saturated electrode) for 1-2

The experimental setup used for the on-line mass spectrometric

Deuterium content was calculated by the equation

S.

gas analysis was the same as in ref 12.

I(D2+) + I(HD+)/2 I(H:) + I(HD+) + I(D:) 100 D m o l % =

where I(Hz+), I(HD+), and I(Dz+) are mass intensities of Hz+ (m/z = 2), HD+ (miz = 3), and Dz+ (m/z = 4) (PA). Other vol- atiles had been cooled with liquid nitrogen before the gases reached the ion source of the mass spectrometer.

Isotopic gas composition had been measured on 3-5 copper electrodes for 20 min, and the mean value was calculated. An accuracy of mass spectrometric analysis depended on mass intensities and was ranging from *0.2 to f0.5 mol % . The values of mass intensities presented were detected immediately after

(8) Ogura, T.; Malcolmson, M.; Fernando, Q. Langmuir 1990,6, 1709. (9) Bruckenstein, S.; Gadde, R. R. J . Am. Chem. SOC. 1971, 93, 793. (10) Wolter, 0.; Heitbaum, J. Ber. Bunsen-Ges. Phys. Chem. 1984,88,

2 . (11) Baltruschat, H.; Anastasijevic, N. A.; Beltowska-Brzezinska, M.;

Hambitzer, G.; Heitbaum, J. Ber. Bunsen-Ges. Phys. Chem. 1990, 94, 996.

(12) Jusys, 2.; Liaukonis, J . ; Vaikelis, A. J . Electroanal. Chem. 1991, 307. 87.

0 1992 American Chemical Society

Notes Langmuir, Vol. 8, No. 4, 1992 1231

written as

2CH,O + 40D- 2 2HC00- + 2D,O + H, + 2e- (2b) According to our recent datal3 obtained by DEMS

combined with isotope labeling, anodic formaldehyde oxidation on copper in Cu(I1)-free solution also results in hydrogen evolution from formaldehyde alone as could be expected from reactions 2a and 2b.

I t should be noted that anodic formaldehyde oxidation reaction 2a,b is expected to be not elementary but a complex one. The most popular reaction scheme14J5 includes a primary step of catalytic formaldehyde dehy- drogenation as a result of its dissociative adsorption on Cu and disruption of C-D (or C-H) bond with the formation of labile formyl:

(4) Recombination of hydrogen adsorbed causes the molec- ular hydrogen to evolve. The metastable formyl is anodically oxidized:

( 5 ) The data obtained are consistent with this scheme of reactions.

The data may be concluded to confirm the electro- chemical mechanism of electroless copper(I1) reduction by formaldehyde. However, they are conflicting with those of ref 8. Such a disagreement may be explained by the preparation for analyzing the samples of gas evolved,8 including the gas elution through an iron-coated alumina column and the conversion of it into water over CuO at high temperature. Possibly, deuterium exchange with ad- sorbed light water took place on the iron-coated alumina prepared apparently in a light water solution. Therefore, the data obtained may be distorted, and the mechanism proposed8 on the basis of these data is a doubtful one.

Registry No. Cu, 7440-50-8; CH20, 50-00-0; Hz, 1333-74-0;

(13) Jusys, 2.; Vaikelis, A. Submitted for publication in J . Electro-

(14) Meerakker van den, J. E. A. M. J. Appl . Electrochem. 1981,11,

(15) Buck, R. P.; Griffith, L. R. J. Electrochem. SOC. 1962,109,1005.

c u DCDO,, - Dad + DCO,,

cu DCO,, + 20H- - DCOO- + H,O + e-

CD20, 1664-98-8; DzO, 7789-20-0.

anal. Chem.

395.

Table I. Mass Intensities of I&+, HD+, and Dz+ and Isotopic Composition of Gas Evolved during Electroless

Copper Plating.

no. DH t.OC uA MA LLA mol% expt I(Hz+), I(HD+), I(Dz+), D,

~

1 12.0 20 0.04 0.34 9.30 97.8 f 0.5 2 12.0 70 0.15 1.00 24.50 97.6f 0.3 3 12.5 20 0.10 0.60 17.25 97.7 f 0.4 4 12.5 70 0.30 2.00 46.00 97.7 * 0.2 5 12.0 20 9.30 0.04 0.00 0.5 f 0.5 6 12.0 70 24.75 0.06 0.01 0.4 f 0.3 7 12.5 20 17.80 0.05 0.00 0.4 f 0.4 8 12.5 70 46.10 0.20 0.02 0.2 f 0.2 0 Solution contained (mol L-l) CuSO, (0.04), EDTA (0.04), and

formaldehyde (0.08). Experiments 1-4 were carried out in H2O solution containing CDzO (D content 98 mol 5%) and 5-8 in D20 (D content 99.8 mol % ) solution containing CHz0.

initiating the electroless plating because of membrane perme- ability changes during copper deposition.

Results and Discussion Our results showed all the amount of gas evolved during

electroless copper deposition to originate from formal- dehyde (Table I). These results were obtained in the solution containing heavy formaldehyde and light water as well as in the "mirror" system of light formaldehyde and heavy water. Isotopic gas composition was determined to be the same as that of formaldehyde used for copper(I1) reduction; i.e., no excess of water hydrogen isotope was detected in the gas as compared to formaldehyde within the measurement error.

These data are consistent with the electrochemical mech- anism of electroless copper plating. In the case when heavy formaldehyde and light water are used, reaction l a is the total reaction of anodic and cathodic half-reactions 2a and 3. In the other case, when light formaldehyde and heavy water are used, reaction 1 may be expressed as

c u Cu2+ + 2CH,O + 40D- - Cuo + 2HC00- + 2D,O + H,

(1b) and reaction of anodic formaldehyde oxidation may be