Note: Part 1 ran in the May issue; part 2 ran in the July issue; part 3 ran in the September issue; part 4 ran in the November issue. All five parts are included here.
If one thought that electroless copper and other metallization systems were complex, and the deep dark secrets of these systems were shrouded in black magic, this month’s discussions on electroplating will seem like brain surgery.
In this next series of columns, the intricacies of electrodeposition technology and its function of building up the thickness of copper in the through holes and on the board surface will be presented in detail. The function of the active ingredients in the electrolytic copper plating solutions will be presented. Process control limits for the various plating solution components and the effects on deposit integrity will be discussed.
In contrast to the classic initial through-hole metallization with electroless copper, the buildup of the conductive layer to a desired thickness in a through-hole is achieved by the electrodeposition of copper. It is understood that these technologies overlap, i.e., that some initial metallization processes use direct electroplating and, conversely, fully additive processes deposit copper solely with the electroless process.
The historic preference for electroplating relates to the lower cost, faster deposition rate, and the generally better metallurgical properties of the electroplated copper. Different chemical compositions of electroplating copper baths have been in use, such as alkaline pyrophosphate baths and acid cupric sulfate baths. Sulfuric acid based compositions now dominating the industry. Improvements have been concentrated on the “throwing copper” of the baths (e.g., their ability to deposit a relatively uniform layer of copper in a through-hole and on the board surface). A shift to higher acid and lower copper concentrations, as well as improved organic additive systems, (brighteners, carriers, and levelers) have demonstrated the ability to improve throwing power.
Several other electroplating processes are used in the fabrication of the printed circuit boards. Tin/lead or tin can be electroplated, and occasionally they can be deposited from an electroless chemistry to act as a metal etch resist. Other electroplated metals include nickel/gold and palladium (also deposited from immersion baths or electroless baths when deposited selectively) to form suitable surface finishes for insertion connectors or component attachment (e.g., wire bonding).
Focusing on the acid copper electroplating process, the first step is the preplate cleaning sequence of acid soak cleaner, microetch, and acid predip. The acid predip is optional and can be omitted if rinsing after microetching is sufficient. However, this only applies when using peroxide sulfuric acid micro-etchants. Use of persulfate based etchants requires an acidic predip. Critical acid cleaner variables are chemical composition, concentration, time, and temperature. Regarding the microetch process, etch depth is the critical parameter, and is influenced by the chemical composition, concentration, time, and temperature. The acid predip, typically the same acid at the same concentration that is used in the plating bath, protects the plating solution from undesirable drag-in, which might adversely affect the composition or concentration of the bath. Thus, the chemical composition of the predip and its level of contaminants are critical.
Critical parameters of the acid copper bath include electrical, mechanical, physical, and chemical variables. Starting with the electrical parameters, current density and primary current density distribution affects plating rate, thickness across the board surface, and the metallurgical properties (e.g., ductility, roughness and color) of the deposit. The current density across the plating surface depends on the rectifier capacity, related anode and cathode sizes, as well as spacing, shielding, cell efficiency, and solution conductivity. The current density distribution across the plating surface depends on several of the previously mentioned factors and, in the case of pattern plating, also on the circuit pattern. The current density in the through-hole depends on the following additional variables:
In Figure 1, throwing power is defined for purposes of this and subsequent columns. The actual mechanics of the electrodeposition process, and the influence of process parameters will be discussed in future columns. The degree of difficulty for plating through-holes depends on board thickness as well as hole diameter. For example, given a choice of two holes, both with a 10:1 aspect ratio, the preference would be to design a 100 mil thick panel with a 10 mil via diameter, versus a 200 mil thick panel with 20 mil diameter holes.
Figure 2 underscores the difficulty in maintaining throwing power as the aspect ratio increases. Figure 2 illustrates that the IR drop (or the electrical resistance through the via) increases as a squared factor, while via diameter only effects the resistance (or throwing power) in a linear fashion.
In Part 2, we will continue to explore the electrodeposition process.
There are a number of published studies on the subject of plating uniformity in through holes. In these studies, researchers developed models to test plating uniformity. With respect to plated through holes, the models attempted to predict what influence key variables had on plating uniformity. The variables include:
Basically, one has to optimize the electrolytic copper process for plating on the surface and in the through hole. Studies indicated that ohmic resistance tended to dominate the plating process with the higher aspect ratio through holes. For example, the ohmic resistance (or voltage drop) can be explained by the following model:
Where E is the ohmic resistance, J is cathode current density, K is solution resistance, d is hole diameter and L is length of hole (board thickness).
As the model shows, the thickness of the panel or length of the hole influences the difficulty of plating by a squared term. In addition, an optimal balance between agitation on the PCB surface and solution movement in the holes was required to achieve a compromise between uniform plating distribution across the panel surface and excellent throwing power in the hole. The model developed by these scientists has been verified time and time again.
Plating uniformity is a continual challenge for through-hole PCB manufacturing and is becoming more difficult with increasingly complex designs. It should be quite clear that plating uniformity is closely influenced by solution chemistry and solution agitation conditions.
Certainly, understanding the ramifications of the model is critical with respect to plating distribution, throwing power and overall plated through-hole reliability. The factors listed in the fishbone diagram below provide a framework for discussion. These and other variables will be discussed in this and subsequent columns. The fishbone diagram in Figure 1 highlight the numerous interactions that occur during the electrolytic copper plating process.
The function of the basic chemical components in an electrolytic copper bath are listed below:
In Part 3, we will look further into the electrochemistry of plating using the various mathematical models for determining plating uniformity.
In the previous column, a broad overview of copper electrodeposition was presented. In this month’s feature, the functional aspects of electrodeposition will be presented in greater detail. It is critical to obtain a fundamental understanding of the plating process, including an in depth knowledge of the various additives, organic as well as inorganic. Additionally, mechanical factors such as agitation, operating temperatures, cathode current density and anode placement have influence on the plating process.
As printed wiring board designs become increasingly more complex, with smaller diameter vias and greater panel thicknessess, the engineer faces a more difficult task with respect to optimizing plating distribution; this includes throwing power and the physical and cosmetic properties of the copper deposit. Optimizing the operation of the chemistry is a big task. The additives in the chemistry interact with the various mechanical aspects of the plating system. The rectifiers, anode to cathode relationship, filtration, agitation and plating by-product build-up influence the plating distribution and overall quality and reliability of the copper deposit as well.
Taking this a step further, achieving plating thickness uniformity is largely dependent on primary and secondary current distribution.
Primary current distribution involves the main influencers of plating uniformity; the plating cell design and geometry, anode and cathode spacing, and the size and shape of the anode and cathode. The pattern image of the PCB will also influence plating uniformity. However, the less uniform a plating pattern (according to primary current distribution), the greater the variation in the plating thickness across such a pattern. It is commonly understood that PCB patterns that are very uniform in the layout of the circuit features exhibiting a uniform plating area across the panel will plate with minimal thickness variation. However, designs that exhibit ultra-uniform circuit layout patterns are not all that common.
Certainly, this concept can be explained by looking at simple plating applications. In one such application, assume the plating is taking place on parallel plane electrodes as depicted in Figure 1. This is often referred to as the mathematical concept of equi-potential surfaces. Looking at the parallel planes in Figure 1, it is easy to see that current flowing between adjacent parallel electrodes is constant across the electrode surfaces. Thus, the length the current flow is equal.
It is easy to see that the potential between two parallel surfaces is constant across the surface. For two parallel electrodes, the current at any point on either electrode is equal. Thus, plating thicknesses would be equal. Figure 2 illustrates what would happen if the electrodes where rotated or angled with respect to one another; the equi-potential surfaces would no longer be parallel to each other.
The above figure is similar to the Hull Cell Testing concept, whereby the cathode is set at an angle from the anode to simulate varying current densities. When one of the electrodes is angled in such a fashion, the lines of current flow become more tightly spaced at the angled end. (see area AKB versus DHC in Figure 2). Due to the angle of BC, point B is closer to the AD electrode than is point C. Since the lines of current flow are always perpendicular to the to the equi-potential surface, the lines of flow become more tightly spaced at the angled end, resulting in a current density (i.e. plating thickness) variation across the BC surface.
While this explanation is somewhat a simplification, one can see that potential theory does apply to plating cell design and anode/cathode configuration. Engineers must consider anode to cathode relationships with respect to size and spacing.
In the next installment of “Positive Plating” the subject of anode to cathode spacing and secondary current distribution will be discussed in detail. In the meantime, stay positive. PCD&F
PART 4
Often, in an effort to increase throughput, anode/cathode spacing is reduced and plating uniformity suffers. Increasing the anode to cathode distance minimizes plating variations. FIGURE 1 illustrates the effect of anode to cathode distance.
Anode surface area versus cathode surface area is another important consideration. If the anodes are much larger in area than the cathode, increased current flow around the edges of the panels (cathode) will be plated to a much greater thickness than other areas, as depicted in FIGURE 2. Pattern plating of PCBs represents a similar situation.
To minimize the primary current distribution, it is recommended that anode to cathode surface area ratios should not exceed 2:1 and anode length should be three to six inches shorter than the cathode. Rack design and cathode spacing on the rack will also influence the primary current distribution.
Fortunately, secondary current comes into play through a factor known as polarization. Polarization refers to the additional potential required above the equilibrium potential to drive the deposition of the metal to be plated. When a plating cell contains copper anodes and a copper cathode (circuit board) in a plating electrolyte, an equilibrium potential exists between the anode and cathode. Basically, the potential is determined by Ohm’s Law– which is the solution resistance between the anodes and cathodes. Additional resistances arrive through the voltage required to corrode the anode and the resistance required to reduce the metal ions to metal at the cathode.
In order to improve throwing power and plating uniformity, one can increase the conductivity of the plating electrolyte and the polarization. Polarization and conductivity are both dependent on solution operating temperature, solution agitation and cathode current density.
The three types of polarization encountered in electroplating are gas polarization, concentration polarization and chemical polarization. Gas polarization results directly from oxygen and hydrogen being evolved during the plating process. By increasing solution agitation (solution movement), and more importantly solution movement uniformity, gas polarization is reduced. Chemical polarization results from a thick film forming on the anodes, which results in poor anode corrosion. Typically, one sees a thick sludge-like film form on the anode. A number of factors contribute to this issue: poor quality anodes, anode bags that are plugged with sludge that has fallen from the anodes and requires cleaning and operating parameters of the plating process that can lead to polarization of the anodes. These include too low an operating temperature, imbalance of key addition agents or a combination of too high a concentration of the metal salt in solution in combination with the acid level. Finally, concentration polarization results as the metal ions are depleted from solution close to the surface of the cathode. The concentration polarization can be altered by solution agitation.
Secondary current distribution effects are complicated. Changes in one variable may have a number of effects. One must control the anode polarization to insure that the anodes are uniformly polarized and a planar equi-potential surface is radiated from the anode. Non-uniform polarization will cause the anodes to sludge causing thickness variations from the top of the plated panel to the bottom. Solution agitation at the anodes will minimize both concentration and gas polarization. There are no guarantees that uniform anode polarization will ensure uniform plating distribution. The main consideration is plating distribution across the surface of a printed wiring board panel, and from the surface of the panel through the holes, it will vary due to resistances. By mitigating the effects of these resistances, plating distribution is improved.
The organic addition agents serve several functions with respect to acid copper electrodeposits. First and foremost, organic addition agents mediate the electrodeposition process by influencing secondary current distribution and control the physical properties of the copper deposit. The key physical properties of the deposit are tensile strength and elongation.
During the normal performance of a printed circuit board, the copper deposit will undergo flexing. An example would be a satellite orbiting the globe, being intermittently heated and cooled as the device makes its way around the earth. The copper deposit within the printed circuit board will undergo flexing, which leads to fatigue in the deposit. The result of this fatigue is cracking of the deposit within the barrel of the hole. The deposit must have sufficient tensile strength to be able to withstand the flexing that can induce fatigue cracking. It is well understood that the organic addition agents utilized in acid copper plating processes influence the physical properties of the deposit and provide improved characteristics.
Acid copper additives contain at least one material described as a brightener and a second material that is described as a high molecular weight polymeric compound. Optionally, the additive package can contain a leveling agent.
A brightener is classified as a nitrogen or sulfur containing compound, that by coulombic attraction, forms a layer on the copper (Cu) surface where it enters, together with chloride ion (Cl-), in the one electron transfer Cu++ ? Cu+ ? Cu0. Brighteners mask preferential growth sites on the planes. Even with the brightener, the growth of the plated copper shows little directional preference. The grain structure of the copper is semi amorphous and/or microcrystalline. In the absence of these brighteners, the copper deposit will have a large-grain structure with little refinement. In order for the brightener to function as it is intended, a carrier component must be present in the plating solution.
Carriers are a class of high molecular weight materials that are designed to exert a strong plating suppression effect on the high current density areas of the cathode. These polyethers are generally in the range of 6,000 to 30,000 in molecular weight. The mechanism of the suppression effect is believed to be caused by the stronger adhesion forces between the high molecular weight polyethers and the copper surface through copper complex formation and the stronger intermolecular forces between the polyether molecules. One can describe this as a smoothing effect that allows for a more uniform diffusion layer across the surface.
Leveling agents are designed to give a very strong suppression effect on the plating deposit. When the plating solution is in the proper concentration, these additives have a tendency to smooth out imperfections in the laminate foil and areas in the via that have some drill gouges. An example of surface roughness without leveling is shown in FIGURE 1. The roughness presents several issues, especially on fine pitch BGA designs and features that will have wire bonded components. This is why excellent leveling is necessary to create a smooth topology as shown in FIGURE 2. PCD&F
Michael Carano is vice president for OM Group, Inc. and can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it..