Contributed by Paul Stransky, CEF, Paul Stransky Associates
23 Aug 1995
ABSTRACT
Eleven papers have been reviewed covering the time period 1986-1991. They represent a mix of highly theoretical to those of a more commercial nature. In general they deal with development of models, application of these models to formulate appropriate plating solutions, and explanations of solution components and their functions.
The purpose of this paper is to try to put the information in a more practical context for platers, who do not have to formulate plating solutions, but have to use them. To help develop an understanding of how copper plating solutions developed for high aspect ratio through holes (HARTH) differ from other plating solutions and what the limitations are if used for multiple plating tasks. For the researcher it summarizes many, but not all of the papers on the subject. However, the references cited in these papers represent significant works on the subject.
REVIEWS
Carano (1) discusses increasing need to plate HARTHs, the effects of throwing power (primary and secondary current distribution), the role of addition agents, solution agitation, other factors such as anode / cathode placement, and ratio,as well as good plating techniques. He concludes process windows will be narrower, and that there has to be a synergism between components of the plating system (chemical, mechanical, electrical).
Amadi (2) discusses primarily plating solution chemistry aspects of HARTH plating. He examines the effects of sulfuric acid, copper concentration, addition agents, and current density on the ratio of surface to hole wall plating thickness. He finds low Copper, high acid, with sulfuric acid:Cu ratio of greater than 20:1, and low current density necessary for maximum uniformity.
Glantz (3) and Hazlow (4) discuss forced flooding and ultrasonics as techniques to insure adequate solution exchange in small holes. This work (3) is primarily about de-smear. He uses a model of laminar flow and kinematic viscosity.
Fisher, Sonnenberg and Bernards (5) write about fluid dynamics and electrochemistry. Fundamental aspects of fluid dynamics are discussed using a capillary tube as a model taking contact angle between process solution and substrate into consideration as well as surface tension. This is followed by a model for looking at mass transfer within a hole during plating, and determining the number of hole volume turnovers needed for a 0.1% depletion rate of copper in the hole not to be exceeded. Electrochemical factors such as voltage drop within a hole are discussed and a model is developed taking Ohms law and hole geometry into consideration. Experimental work done using a test cell to develop a copper plating solution for HARTHs is described. Results included optimization of metal, sulfuric acid, and organic addition agents. Experimental results were then confirmed with actual drilled boards. In addition general guidelines for equipment, chemistry, and control are also given. Some of the conclusions are:
1. Wetting is not an issue by the time you get to electroplating, because of all the previous preparation steps.
2. Pressure requirements for forced solution flow through holes are minimal and easily met by conventional board agitation.
3. Ohmic potential drop is a very important factor in obtaining uniform plating thickness throughout length of hole. It is the limiting factor determining plating rate.
4. Sulfuric acid/Cu ratio more important than just conductivity re; throwing power (center thickness/surface thickness). Throwing power increases with increasing acid/metal ratio (decreasing metal content).
5. Organic additives (carriers, brighteners, levelers) work to increase the current density or plating rate that can be maintained with satisfactory throwing power.
D'Ambrisi et.al. (6) approach the subject of small diameter holes (HAR) from the prospect of the entire process. From drilling to electroless plating in this paper. The potential for excessive heat generation and subsequent smear during drilling is discussed. The importance of adequate de-smear, and resulting wettable wall surface is addressed. Different de-smear methods (Sulfuric acid, chromic acid, plasma, alkaline permanganate) are covered, with alkaline permanganate recommended. Several steps of the electroless plating process are explained (conditioning, acceleration, plating). Solution flow with respect to providing appropriate mass transfer of plating solution components (Cu, NaOH, HCHO) is discussed. The model presented is based on laminar flow through a circular cross section and Fick's First Law of Diffusion. The authors conclude that alkaline de-smear renders hole walls hydrophilic (wettable), increases surface area by 6x improving adhesion, and aiding small H2 gas bubble nucleation during electroless plating.
D'Ambrisi et.al. (7), in a second paper follow up with a discussion of electrolytic and full build electroless copper plating. Problems with ohmic resistance being the limiting factor in uniform HARTH plating are discussed. Full Build electroless copper is suggested as a way around the problem. Conclusions include:
1. Mass transport not problem, methods of solution agitation can rectify this.
2. Ohmic resistance is the main problem in achieving uniform plating thickness, with hole length being more important than diameter.
3. Aspect ratios greater than 10:1 can not be plated uniformly.
Yung and Romankiw (8) used a gap cell which represented a modified version of a Hull cell equipped for solution flow, and plexiglass drilled boards to study the acid copper PTH process. They were concerned with both plating on the surface and hole wall. Plating solutions with and without additives were used. Mathematical models are developed for electrode kinetics, and mass transfer based on electrochemical engineering principles. Experimental results are correlated with mathematical models. Three distinct plating regions and mechanisms are described, at extremely low current densities (region I) charge transfer dominates and additives can improve current distribution. At somewhat higher current densities (region II) mixed control prevails, both charge transfer and ohmic resistance, and additives are still effective. In region III at higher current densities ohmic resistance dominates and additives are no longer effective. The authors conclude:
1. Ohmic resistance rather than mass transport is the controlling factor in plating HARTHs at normal current densities.
2. Even with suitable agitation and flow to achieve high mass transfer, high current densities will still give non uniform HARTH wall thicknesses, the model will predict this.
3. Agitation must be balanced on surface and in hole.
4. When ohmic resistance dominates process, its difficult to plate HARTHs above 10:1 without lowering current density to 20-40 mA/cm2 (19-38 ASF).
5. Possible schemes to improve copper distribution inside the holes include increasing bath temperature and conductivity, as well as reducing both hole diameter and length.
Hazelbeck (9) develops a non-dimensional mathematical model which includes effects of convective, diffusive, and ohmic transport in plating through holes with additives to the bath. The model which incorporates Laminar flow is used to examine effects of additives and other process variables on deposit uniformity. There are three important dimensionless parameters. The Thiele modulus is a measure of the ratio of reaction rate to diffusion rate. The inverse dimensionless conductivity measures the ratio of diffusion rate to electrical migration rate. The Peclet number provides a measure of the ratio of convection rate to diffusion rate. Amongst his conclusions are:
1. Uniformity of plating correlated to the three dimensionless parameters.
2. Deposition uniformity can be improved by increasing conductivity when electrodeposition is ohmically controlled.
3. Forced convection in HARTHs is important in improving deposition rate while maintaining uniformity.
4. Additives that decrease cathodic charge transfer coefficient of electrodeposition significantly improve HARTH uniformity.
Hazelbeck, and Talbot (10) review development of their two dimensional model and other models. The two dimensional model is used to examine the effects of plating variables on through hole thickness uniformity in an acid copper plating bath with and without additives. The general model is used to determine conditions needed to achieve ohmic limited plating necessary for HARTH uniformity. Discussion of various models of plating under limiting conditions such as convective-diffusion, coupled convective-diffusion / ohmic resistance and ohmic resistance are presented. Conclusions include:
1. Uniformity of plating correlated to three dimensionless parameters as described in earlier work above (9).
2. Increased solution conductivity and forced convection in the hole are important in improving deposition rate while maintaining uniformity.
3. A smaller through hole can be plated easier than a larger one of the same aspect ratio.
4. Plating additives that reduce charge transfer can significantly improve HARTH plating uniformity, and with flow enhancement reasonable plating rates of 10-40 mA/cm2 (9-38ASF) can be used. Likewise higher plating rates can be achieved for low aspect ratio holes .
5. When using unidirectional flow solution conductivity must be increased or charge transfer reduced by additives.
6. At very high Peclet numbers plating rate is ohmic limited with uniform plating regardless of flow direction.
7. Uniformity of deposition can be improved by increased solution conductivity, use of additives to reduce charge transfer, reducing hole diameter and length while maintaining the same aspect ratio, and lowering current density.
Hazelbeck, and Talbot (11) a model is developed for PTH when two counter electrodes and solution flow reversal is used, and another model for unidirectional flow with only a downstream counter electrode. This work follows from the authors earlier work in(9) and (10) above with the following conclusions.
1. The downstream process allows for higher current densities than with the other configurations while still maintaining plating uniformity.The reason being low metal ion concentration and low ohmic losses at the down stream end of hole with high metal ion concentration and large ohmic losses at the upstream end .
2. Drawback to the downstream process is that it would require holes to be plated separately from surface lines, however it is promising for applications other than PTHs.
DISCUSSION
Some common themes come through from the various investigators:
1. Ohmic resistance within the HARTH is the major limiting factor in obtaining uniform plating.
2. Ohmic resistance is minimized by increasing plating solution conductivity which is accomplished by increasing the ratio of sulfuric acid to copper metal and / or increasing temperature.
3. The use of additives in plating solutions that minimize charge transfer also is very important to hole wall uniformity and the ratio of surface to wall thickness.
4. Plating solution flow is important, with the usual air sparging and panel movement being sufficient to avoid mass transport problems.
Well, what does this mean to the plater or plating engineer actually doing the work? We usually do not formulate plating solutions, leaving that to our suppliers and at this point in time there are many commercial plating solutions available for HARTH. It may be helpful to understand how commercial plating solutions available for various plating tasks differ from each other.
Table 1 contains a listing of data taken from product literature for 4 different commercial plating solutions (12,13,14,15) designed for different plating tasks. Obviously there is no reference to specific additives since they are proprietary, but for this exercise it is assumed they are appropriate for the task.
The first represents a "standard" copper plating bath for non electronic use (16).
Baths 2 and 3 represent the "high throw" acid copper used for many years for PTHs.
Bath 4 is representative of baths developed for HARTH plating.
Table 1: Comparison Of Properties From Acid Copper Plating Baths
Property Bath 1 Bath 2 Bath 3 Bath 4 CuSO4 (g/L) 255 75 68 25 Cu (g/L) 57 19 17 6 H2SO4 (g/l) 60 188 173 225 H2SO4:Cu 1:1 10:1 10:1 38:1 Current Density 30-60 1-80 20-40 10-20 Range (A/ft2) Notes: 1. Techni Copper-U 2. Copper Gleam PCM 3. CUBATH M 4. Electroposit 1000
References:
1. M. Carano, P.C.Fab, 12,34,(1986)
2. S.I. Amadi, P.C.Fab, 10,85,(1987)
3. E.J.Glantz, P.C.Fab, 2,60,(1989)
4. R.E. Hazlow, P.C.Fab, 2,78,(1989)
5. Fisher,Sonnenberg and Bernards, P.C.Fab, 4,39,(1989)
6. D'Ambrisi et al., P.C.Fab, 4,78,(1989)
7. D'Ambrisi et al., P.C.Fab, 8,30,(1989)
8. E.K.Yung,L.T.Romankiw, J.Electrochem.Soc., 3,756,(1989)
9. D.A.Hazelbeck, J.Electrochem.Soc., 4,215C,(1989)
10. D.A.Hazelbeck,J.B.Talbot, J.Electrochem.Soc., 7,1985,(1991)
11. D.A.Hazelbeck,J.B.Talbot, J.Electrochem.Soc., 7,1998,(1991)
12. Technic TECHNI COPPER-U product data sheet
13. LeaRonal Copper Gleam PCM technical bulletin 30013
14. SelRex CUBATH M product data sheet
15. Shipley Electroposit 1000 product data sheet
16. A.Sato,R.Barauskas, Metal Finishing Guidebook,Hackensack N.J.,1989; p. 216
