Currency Converter

Monthly Specials

Security Info

Making Better Contacts
by AMTECH

A Guide to Surface Mount Technology

Printer Friendly Page

Introduction
Powder and Alloys
Particle Size and Consistency
Solder Cream Chemistry
Printing
Reflow
Ordering and Technical Support

AMTECH's thoughts on No-Lead Solder

Introduction

Beginning with chip capacitors and resistors on hybrid substrates in the 1960's, the process of Surface Mount Technology (SMT) has become the state of the art in electronics assembly. By placing the components on the board rather than through the board, PCB manufacturers realize significant production benefits and cost savings. Reduced component size and lead length increase circuit density, resulting in more applications per unit area or weight of board, which leads to lower production costs.

In the beginning, switching to SMT was an option for PCB manufacturers; the concern over component availability and questions over cost savings delayed the conversion for many assemblers. Today, SMT is not an option but a necessity, since many components are no longer available in a through-hole package.

The shift to SMT has focused a great deal of attention on solder joints. Now, in addition to providing the electrical pathway from the device to the board, the joint must also provide the mechanical connection. As the device sizes and lead pitches decrease, with a corresponding increase in lead count, the mechanical and electrical properties of the solder joint are critical.

Defect-free assembly demands choosing the proper materials and equipment to provide maximum control of the process.


Powders and Alloys

To make the best cream, you start with the best powder. Only the highest quality solder powder is used in AMTECH products - solder powder manufactured by Advanced Metals Technology, Inc. Solder powder quality is determined by purity of the alloy, particle size, shape and distribution, oxide levels and lot-to-lot consistency. All of the alloys manufactured by AMT exceed the applicable IPC, QQS and J Standard specifications. AMT's proprietary processes allow the tightest control over size, shape and oxide levels in the industry.

The degree of sphericity is related to the oxide levels on the powder - generally, the more irregulars that exist, the higher the oxide level. Solder alloys cannot exist in free-flowing powder form without a thin oxide coat, but through a controlled production and classification process, AMT consistently produces the lowest oxide powder available.

One of the properties of finer size powder is that, as the particle diameter decreases, the ratio of surface area (and therefore oxide level) to mass or volume increases. However, a major contributor to the higher oxide levels typically found on smaller particles is the extensive processing most other manufacturers need to perform to produce these finer sizes. Excessive processing thickens the oxide layer. Through advanced technology, AMT makes ultra fine pitch powder available with less concern over oxide levels.

Advanced technology means working to develop better, stronger and safer alloys for the electronics industry. Our recent work with Hughes Aircraft has led to the development of a Fatigue Resistant Solder that can prolong the life of solder joints exposed to cyclic stress. Ongoing research and development projects will continue to bring the latest in soldering technology to the industry.


Partical Size and Consistency

Classification of powders has been a much disputed area. A good deal of the controversy centers on the tools available for analysis. Until the advent of computers, determining the particle size distribution was carried out through the use of test sieves or screens. This technique was fairly quick and easily carried out by personnel with little training. With particle size distributions that ranged from 150?- 15?(-100/+625 mesh) the technique was adequate. However, the demand for more restrictive size distributions pointed out the lack of precision and resolution of test sieves. The development of supplemental sizing techniques (e.g., laser diffraction, image analysis, and/or volumetric resistivity modification) permits more precise characterization of powder distribution. However, the labeling of a given powder distribution is still tied to sieve designations. Unfortunately, these designations can be misleading since test sieves have defined tolerances from sieve to sieve as well as tolerances in the openings of a given sieve. It is much more prudent to specify a given distribution by stating the actual particle diameters required.

Saying that a powder contains particles between 45?and 25?(-325/+500 mesh) still does not define the distribution. Two powder samples may have the same average particle size (whether the average is calculated on a weight percent basis or population basis) but have widely different properties. Therefore, to properly specify a powder, it is necessary to recognize that it is a population of particles, and it is necessary to characterize that population in terms of the distribution of parameters that are important to the particular application.

The demand for finer pitch printing and the development of no-clean pastes require proper characterization of the powders used in solder pastes and creams. Considering Stokes Law (where F is the force on a spherical particle of radius a and velocity v in a medium of viscosity ) for the behavior of particles in a viscous fluid, it is seen that variations of size distribution in a solder cream will have a significant impact on the rheology with a corresponding effect on the paste's printing characteristics. Secondly, with no-clean creams, it is imperative that solder balling be nonexistent. This requirement demands that the printed solder cream come to the reflow temperature uniformly. This can only occur if the individual solder particles are heated at the same rate, which requires that they have nearly the same specific area. Hence, the size distribution must be precisely specified, and specification must be uni-modal with a small standard deviation. Under these conditions the cream will have a consistent rheology and the oxide content of the powder will be tightly controlled.

A beneficial side effect will be metallurgical uniformity in the powder particles themselves, which will promote a metallurgically sound solder joint. Of course, these specifications will preclude irregular particles and multi-modal distributions.


Solder Cream Chemistry

Solder cream (sometimes refered to as paste) is made up of the powdered alloy and the flux binder system. As the name implies, one of the system's duties is to act as a flux.

Lead finishes are normally covered by thin films of tarnish, which can be described as two layers, differentiated by the way they are bound to a surface. Located directly on the metallic substrate and chemically bound are the layers of oxide, sulfide and carbonate, as well as products from any preceding production steps. On top of this layer is a physically bound absorption layer. Absorptive forces attract residues including water, gases and residues from preceding reactions, which will collect on the first layer. Therefore, the specific requirements for a flux are:

• The dissolution of the outer absorption layers.
• The displacement of the chemically bound reaction layer.
• The dissolution of some of the substrate molecules to enable the formation of intermetallic compounds.

The flux binder system has to perform many additional tasks during the life of the cream. It must:

• Suspend the powder so it does not separate in the container.
• Protect the powder from oxidation without reacting with the surface at storage temperatures - to promote a long shelf life.
• Give the cream thixotropic properties to aid in the printing and releasing from the stencil.
• Provide enough tack to allow for processing time between printing and component placement and between the placement and reflow.
• Perform the fluxing actions to promote coalescence of the molten solder.
• Protect the freshly cleaned surfaces from re-oxidation prior to reflow.
• Leave a residue that can be easily cleaned in the selected process, or if a leave-on, be non-conductive, non-corrosive and not interfere with or detract from visual or mechanical testing of the assembly.

RMA creams were the major type used for SMT applications. Rosin, along with the proper solvent, activator system and rheology modifiers, was able to meet all of the above criteria. The residue could be left on most assemblies without concern or could be easily removed with the then readily available chlorofluorocarbons (CFCs).

The adoption of the Montreal Protocol has phased out the production and use of CFCs with total elimination not too far away. The elimination of CFCs as a cleaning option presented a problem to manufacturers: customers had become spoiled with sparkling clean assemblies and did not want to settle for less. Drop-in replacements for CFCs were not perfected, so new formulations were required for customers who still demanded cleaned assemblies; water-washable assemblies were thereupon developed. The other alternative was to modify the RMAs to leave a less tacky and less noticeable, cosmetically acceptable residue. Today, there are two main options for post reflow residues: cleaning them off or leaving them on. Presently, a different flux binder must be chosen for each option.

The benefits of cleaning include better looking boards, easier visual and bed-of-nails inspection, better adhesion for conformal coatings and removal of residues and contamination from other manufacturing processes. Cleaning also allows the use of more aggressive fluxes to widen the process window, with less concern over the solderability of boards and components. Another concern has been raised over the effect residues will have on high frequency circuits.

The first water-washable creams had very short working lives, measured in minutes, and also left residues that had to be removed almost immediately to avoid corrosion problems. Developments in formulation technology have resulted in significant improvements in the succeeding generations of water-washable formulas. Today, formulations are available which offer excellent working life and activity, with reduced concern for corrosiveness prior to cleaning.

The main benefits of leave-on formulations are that they save manufacturing operations, saving on equipment and the associated expenses of labor, maintenance, power, chemicals and waste treatment. Many new leave-on formulations offer RMA activity and process window in a no-clean product. Newer formulations also have reduced residue levels, so the cosmetic problem also decreases. The value of visual inspection has raised concerns in that potentially damaging rework is being done on joints that were merely cosmetically unacceptable, and defective joints often pass. Also, as board complexity increases, the practicality and reliability of visual inspection decreases. With the increased use of J leads and BGAs, unaided inspection is impossible; the assembler must make the leap of faith that if his process is in control, the joints will be acceptable.

In developing a leave-on formulation, the ultimate goal is no residue. The drive for less residue is accomplished in two ways. The first is to increase the percent metal. A small increase in weight percent makes a large difference in volume percent. The second way is to reduce the total non-volatiles in the flux. When one or both of these modifications are used, all of the cream parameters are affected. In the very low residue formulations, the flux in no longer able to prevent re-oxidation of the cleaned surfaces prior to reflow. To combat this situation, many ovens are being designed with a nitrogen option. The nitrogen displaces the oxygen and reduces both the potential for re-oxidation and the charring of any remaining residue. The user must decide if the added expense in nitrogen and process modifications is justified for his particular product.

The R&D team at AMTECH is continually developing formulations that will give manufacturers the widest possible process window. These ideal formulations will allow for a manufacturer's shift changes or other down time, without the cream drying out or absorbing moisture. They would also have activation robust enough to compensate for poor solderability of parts or boards, while remaining benign under any conditions the product may see. It would also not require any modification or additions to the present state of manufacturing, such as inert atmosphere. Each new formulation is evidence of our commitment to this goal.


Printing

The primary method for en masse depositing of solder cream on a circuit board is through the printing process. The increased complexity of board design has led to tighter tolerances for the printing process. This progression has resulted in the near-total shift from screens to stencils and the increasing popularity of polymer-coated metal blades.

The consistent deposition of solder cream is the first step in controlling the SMT process. Having the proper amount of cream will enable the formation of a joint with the proper fillet geometry, which will determine the thermal and mechanical properties of the joint. Too little solder cream can result in opens or mechanically and metallurgically weak joints. Too much cream can lead to bridging. Also, excess solder will make the joint less compliant and more prone to cracks due to component/substrate thermal coefficient of expansion (TCE) mismatch.

The printing process involves using a squeegee to roll the cream across the stencil surface. The cream fills the apertures corresponding to the pads, and releases onto the pads as the stencil is separated from the board. Many factors combine to determine the success of the printing process. These factors include the printer itself, the stencils or screens, the squeegees, the operators and their training, the environmental conditions, the board characteristics and the solder cream. Optimizing these is the key to printing consistently.

AMTECH has found excellent results in the lab and in the field using 4-6 mil metal stencils with metal blades in the on-contact printing mode. Squeegee pressure should be just enough to wipe the stencil clean, with speed in the 10-50 mm/sec range.

The design, quality and accuracy of the stencil becomes increasingly important; as the openings and spacings decrease, tolerances become much tighter. Recent improvements in the fabrication of stencils have resulted in less variation in apertures, less roughness in the walls, and more uniform prints.

An often-overlooked cause of print variation is the preparation of the bare boards. Uneven tinning on fine pitch pads can lead to voids or insufficient solder. This is one reason for the increased interest in organic solderability preservatives (OSPs) as an alternate to hot air solder leveling (HASL).

Solder cream factors for printing include powder size and shape, percent metal and viscosity of the cream. For applications down to 25 mil, AMTECH recommends our -325/+500 mesh powder. For 20 mil and below, the -400/+500 mesh powder gives increased resolution without excessive fines. For stencil printing the range of 89-91 percent metal is recommended, as well as a viscosity range of 180-340 Malcom (700-1400 Kcps Brookfield).

The conventional instrument for viscosity measurement has been the spindle type viscometer. AMTECH has found the spiral pump viscometer to be more repeatable and more representative of the shear actually experienced in the printing process. We currently test at 5, 10 and 20 rmp, with the 10 rpm reading being the reported measurement, also expressed in kcps.


Reflow

Soldering can be defined as the joining of two metals by material heated above its melting point but below the melting points of the materials to be joined. The bond is formed in one of two ways: by the formation of intermetallic compounds, which is an irreversible chemical process, or by diffusion or absorption, which is a physical process.

When joining 63Sn/37Pb and other high tin alloys with copper, two intermetallic compounds are formed. On the copper side is Cu₃Sn and on the solder side, the relatively rough and irregular Cu₆Sn₅. The total thickness of the intermetallic layers (IL) is usually 0.5-0.7 µm. The intermetallic compounds of copper and tin form crystalline grains, the structure of which is determined by the length and intensity of the thermal interaction. Short reaction times form fine equiaxed grains, which promote good solderability and solder joint strength. Long reaction times result in coarse grains, and a thick intermetallic layer. A thick IL gives poor solderability and poor joint strength, both in t = o shear and long-term reliability. The thickness of the IL depends on the temperature but will continue to grow even at ambient temperatures (which on the absolute (°K) scale approaches 60% of the 63/37 eutectic point). This is particularly important when parts or boards are solder coated or pre-tinned. Upon prolonged or improper storage, these ILs can grow through the surface, severely affecting the solderability.

The majority of parts and boards used today come to assemblers pre-tinned with the ILs already established. However, alternative lead finishes and passivated copper pads are becoming more common. These new lead finishes rely on the solubility of the metal in liquid Sn-Pb to form the bond. For some metals, e.g., Pd, Pt and Ni, the dissolution rate is very slow, requiring temperatures above the normal soldering temperatures and/or dwell times much longer than needed for normal soldering. The result is a joint that does not appear like the traditional solder fillet, causing concern during visual inspection.

The objective of the reflow process is to achieve high quality solder joints on all of the component leads of a particular assembly, and do to this consistently. The process involves heating the leads, pads and cream above the melting point of the alloy so that the solder on the leads and pads, and in the cream, reflows into a homogeneous fillet. Consistency in the process depends on the ability to control the application of heat and the variation of heat both across the board and from board to board. This controlled heating is called the profile. The typical profile includes a preheat, drying or soak, and reflow or spike zone. The preheat zone brings the assembly up to temperature uniformly, generally at a rate of 2?/em>C/second or less. This will minimize the potential for thermal shock on the components due to varying heat capacities. The preheat zone also begins the volatilization of some of the solvents added to the cream for printing and releasing. The second zone continues the drying process to prevent out-gassing and possible spattering of the cream. This zone, sometimes called the soak zone, is also where the flux begins to remove the oxides from the surfaces of the leads, pads and the powder itself. The resins and/or higher boiling solvents remain as a cover to prevent the reoxidation that would readily occur at the elevated temperatures. In the reflow or spike zone, the temperature is quickly raised 20-40?/em>C above the melting point of the alloy. It is here that the solder wets the surfaces and forms the intermetallic bonds. The period of time above reflow is called the dwell time, typically 30-60 seconds. The dwell should be long enough to allow for all of the joints to reach temperature and form the bonds. Too long a dwell time can lead to excessive intermetallic formation. Both of the intermetallics are brittle, and if they make up a large portion of the fillet, can lead to premature failure of the joint.

Howard Electronic Instruments, Inc.
974 SE Pioneer Rd
El Dorado, KS 67042
1-800-394-1984 Toll Free (USA and Canada)
01-316-321-2800 International Phone
316-321-2803 Fax
Contact Us
Our Shop Safe Guarantee

Copyright © 1996 - 2005 Howard Electronic Instruments, Inc. All rights reserved.
Revised: February 02, 2005.