Process Description and Engineering Parameters

Nanometer-wavelength multiple-layer alloys were initially fabricated in the late 1960s using vapor-phase evaporation to test theories on the thermodynamic stability and interdiffusion of layered materials. Although vacuum evaporation techniques have been used to form reproducible multiple-layer thin films, sputtering and electroplating techniques are currently the preferred fabrication methods. Multiple-layer alloys can be electroplated from a single bath that contains ions of all of the species to be deposited, or plating can take place sequentially by moving the electrode between multiple plating baths that contain individual species of interest. From a throughput standpoint, single bath plating techniques are preferred, although not all electrolyte chemistries and species redox properties are compatible with single-bath plating.

Multiple-layer alloy plating from a single bath typically requires that the deposition of each species in the multiple-layer pair occur at potentials that differ by more than 100 mV. The more noble component of the alloy is normally plated at the mass transfer limited rate, and the less noble component is plated under kinetics limited conditions. Figure 2 is a schematic showing deposition rate versus potential for an ideal multiple-layer alloy plating bath. The more noble species A plates readily at potentials that are negative of VA, and the less noble species B plates at potentials negative of VB. Within the potential window between VA and VB, it is possible to electroplate a film that is essentially pure in the noble component A. Conversely, the less noble component B is always codeposited with the more noble species A to form a B-A alloy. Periodically switching the potential (or current) between the region where pure A deposits and the region where B-A alloy deposits results in an A/B-A multiple-layer film. The thickness of each lamella is related to the deposition charge through Faraday's law (a knowledge of the current efficiency is also needed). Figure 2 suggests that species A is codeposited with species B at the mass transfer limited rate (Ra.mt). The mole percent of species B (in the BA alloy) is dictated largely by the potential (or current) at which deposition occurs, the strength of agitation in the cell, and the relative concentrations of reducible species A and B that are dissolved in the electrolyte. Decreasing the agitation during species B deposition serves to reduce the rate of which species A (Ra.mt) is added to the film, and increasing the concentration ratio of B to A in the electrolyte directly affects the ratio of RB.MT to RA.MT. The formulation of a multiple-layer plating bath normally begins with an electrolyte that is known to work well for plating the less noble component B at high rates. A small concentration of the more noble species A (»1% of B) is then added to the bath in an effort to minimize the amount of A that codeposits with B. Using this strategy, it is possible to make an A/B alloy in which each lamella is nearly pure (>95%).

Fig. 2 Schematic of ideal deposition rate vs. applied cathodic potential characteristics for electroplating an A/B multiple-layer alloy. The dashed lines denote deposition rates for plating each individual species, and the solid line represents the total deposition rate. Note that pure A is deposited in the potential window between the reversible potential for species A (VA) and the reversible potential for species B (VB), whereas an impure A-B

Fig. 2 Schematic of ideal deposition rate vs. applied cathodic potential characteristics for electroplating an A/B multiple-layer alloy. The dashed lines denote deposition rates for plating each individual species, and the solid line represents the total deposition rate. Note that pure A is deposited in the potential window between the reversible potential for species A (VA) and the reversible potential for species B (VB), whereas an impure A-B

alloy forms at potentials cathodic of VB.

The ideal deposition conditions represented in Fig. 2 are nearly matched in reality by some alloy systems. For example, a variety of bath chemistries and deposition conditions have been found for plating Ni/Cu multiple-layer alloys, where copper is more noble than nickel by nearly 600 mV. For the case of Ni/Cu alloy plating, one often starts with a nickel-sulfate-based electrolyte (Watts nickel bath) or a nickel sulfamate bath. Small amounts of copper sulfate (50 to 1000 ppm Cu+2) are then added to the bath. Table 1 shows typical sulfate and sulfamate plating baths for making Ni/Cu multiple-layer alloys. The deposition protocols for creating multiple-layer Ni/Cu alloys from these baths are nearly identical.

Table 1 Characteristic plating baths for making nanometer-scale Ni/Cu multiple-layer alloys

Plating bath

Typical bath composition

Temperature,

°C

Nickel sulfate

330 g/L NiSO4 ■ 6H2O, 45 g/L NiCl2 ■ 6H2O, 35 g/L H3BO4, 0.05-0.1 g/L Cu+2, 0.1 g/L coumarine, 0.1 g/L sodium dodecyl sulfate, pH » 2

» 60

Nickel sulfamate

70-90 g/L Ni, 0.3-0.9 g/L Cu (as sulfate), 30-45 g/L boric acid, 0.15 g/L sulfamate nickel anti-pit, pH » 3.5

» 50

Pulsed-potential plating of Ni/Cu multiple layers from either bath listed in Table 1 requires the copper lamella to be deposited at a potential near -0.4 V (referenced against a saturated calomel electrode, SCE) and each nickel lamella to be plated at a potential between -1.0 and -1.5 V versus SCE. Under these conditions, the copper deposition rate is controlled by the rate of Cu+2 mass transfer, and the nickel deposition rate is controlled by either electrode kinetics or by mixed mass transfer and kinetics. Copper plating proceeds with nearly 100% current efficiency, but the nickel plating has a lower current efficiency due to the simultaneous evolution of hydrogen. The electrolyte is often strongly agitated during copper plating so that higher mass transfer limited copper deposition rates are achieved. Conversely, the agitation rate is often reduced when the nickel is deposited so that the purity of the nickel layer can be increased by reducing the rate of copper codeposition. An important feature of the single-plating bath formulations described here is that the more noble species, copper, has a low deposition rate compared to nickel and therefore requires a longer duration pulse to achieve a comparable layer thickness. Reference 4 provides an example of pulsed-potential plating of Ni/Cu multiple-layer alloys from a sulfamate bath.

Pulsed-current plating is somewhat more common for making high-quality multiple-layer alloys than is pulsedpotential plating. Moreover, the best results are achieved using a triple-current-pulsing scheme that employs a brief high-current nickel deposition pulse, followed by a brief zero-current pulse and then a long-duration, low-current copper deposition pulse. The size and duration of each pulse depend on details of the electrolyte agitation, the bath formulation, and the modulation wavelength desired. The two triple-pulse-plating protocols given in Table 2 are representative of conditions that lead to coherent nanometer-wavelength Ni/Cu multiple-layer alloys (both studies used plating baths that fall in the range of the sulfamate formulations given in Table 1). The triple-pulse protocol given in Ref 3 results in a high plating rate and a high nickel alloy. The high plating rate is achieved, in part, by agitating the cell using a rotating cylinder cathode. Also contributing to the high plating rate is the formation of a high-nickel-content alloy (recall that the less noble component normally plates at high rates compared to the more noble component). The triple-pulse protocol given in Ref 5 results in lower plating rates and higher-copper-content alloys than the protocol used in Ref 3. The low overall deposition rate reported in Ref 5 is a direct result of forming a high-copper alloy using a quiescent plating bath. In the protocol given in Ref 3, the strength of agitation is modulated during each stage of current pulsing. The use and synchronization of a time-periodic agitation scheme should be carefully considered when developing a multiple-layer alloy plating protocol (Ref 6). Additional information is available in the article "Pulsed-Current Plating" in this Volume.

Table 2 Two representative triple-current-pulsing schemes for plating coherent, nanometer-scale Ni/Cu multiple-layer alloys from sulfamate electrolytes

Characteristic

triple-pulse

Bath

Growth

XNi/XCu

Reference

condition

agitation

rate (^m/h)

Nickel pulse: 90 mA/cm2 for » 0.7 s

60 rpm

3

"Rest" pulse: 0 mA/cm2 for 0.25 to 1 s

» 7

»10

Copper pulse: 1.5 mA/cm2 for » 4 s

600 rpm

Nickel pulse: 12-20 mA/cm2 for » 0.5 s

None

5

"Rest" pulse: 0 mA/cm2 for » 0.5 s

»1

» 2

Copper pulse: 0.3 mA/cm2 for » 11 s

None

Multiple-layer alloys that are plated using two baths are not limited to species with widely separated deposition potentials, nor is it necessary to modify well-established bath chemistries. However, plating a multiple layers using two baths requires the substrate cathode to be periodically moved between the baths. If transferring the cathode between baths takes substantial time, then forming nanometer-wavelength multiple-layer alloys with appreciable total film thickness becomes prohibitively slow. Reference 7 describes the use of a novel dual-bath plating cell that provides nanometer-wavelength Ni/Ni-P multiple-layer alloys at a high deposition rate (2 to 4 pm/h). The cell employs a disk-shape cathode substrate that rotates with an angular velocity of 12 to 20 rpm. The rotating substrate is exposed to each plating bath once per revolution. After each exposure to a plating bath, the cathode is mechanically wiped free of electrolyte to avoid cross-contamination of the electrolytes. A voltage divider is used with a single power supply to control the current flowing through each bath (each bath contains a separate nickel anode). Conventional nickel and nickel phosphorous bath formulations are used in the dual-bath cell to deposit each lamella of the multiple layer. Compared to single-bath strategies, the dual-bath approach substitutes a mechanically complex plating apparatus for a much simpler electrolyte formulation.

A new two-bath electroplating strategy for making multiple-layered materials with molecular-level compositional control is described in Ref 8. The technique, called electrochemical atomic layer epitaxy (ECALE), takes advantage of the underpotential deposition (UPD) phenomenon that occurs when the first monolayer (or partial monolayer) of a surface film is formed. The energetics of the first atomic layer of a deposit is often favorable compared to bulk material deposition, thus resulting in deposition at potentials that are noble of bulk deposition. The ECALE concept has been tested for Cd/Te electrodeposition. The chemistry of each bath was manipulated to provide cathodic UPD of cadmium and anodic UPD of tellurium (at potentials that did not strip the cadmium monolayer from the surface). Low-energy electron diffraction verified the formation of ordered cadmium deposits on tellurium and ordered tellurium deposits on cadmium (Ref 8). Despite the initial verification of the ECALE concept, practical applications of the technique await further development.

0 0

Post a comment