14.3 Soldering Fundamentals, Flux Chemistries & Intermetallic Formation

14.2 Soldering Fundamentals, Flux Chemistries & Intermetallic Formation

Quick Answer: Soldering is a non-fusion joining process utilizing filler metals with a liquidus temperature ≤ 450°C (840°F). Adhesion is achieved through surface wetting, capillary pull, and the formation of a microscopic Intermetallic Compound (IMC) boundary layer (e.g., Cu6Sn5 and Cu3Sn). The global transition from eutectic tin-lead (63Sn-37Pb, melting at 183°C) to lead-free alloys under RoHS has standardized SAC305 (96.5Sn-3.0Ag-0.5Cu, melting at 217 - 220°C). While an initial IMC layer of 1 - 2 μm is essential to achieve a metallurgical bond, excessive thermal aging drives solid-state diffusion that thickens the brittle IMC layer and induces Kirkendall voiding at the Cu/Cu3Sn interface, leading to catastrophic brittle drop-shock failure.

---

Thermodynamic & Metallurgical Foundations of Soldering

Under AWS A3.0 and IPC/J-STD-001/004, soldering is categorized as a low-temperature joining discipline operating below the brazing boundary line:

1. Thermal Criterion: Liquidus of the solder alloy must be ≤ 450°C (840°F).
2. Substrate State: The substrate metal remains in the solid state throughout processing. The joint is created by dissolution of the solid base metal into the liquid solder, forming an intermediate alloy phase at the boundary.
3. Capillary Spreading & Wetting: Similar to brazing, the liquid solder must wet the faying surfaces and be drawn into clearances ranging from 0.05 to 0.15 mm (0.002 to 0.006 in) by capillary force.

Wetting Thermodynamics and Spread Factor

Good soldering requires rapid wetting characterized by a low contact angle θ:

• θ ≤ 30°: Optimal wetting for high-reliability aerospace and medical electronics.
• 30° < θ ≤ 60°: Acceptable industrial wetting.
• 60° < θ ≤ 90°: Marginal wetting (borderline non-wetting or partial de-wetting).
• θ > 90°: Total non-wetting; rejected under IPC-A-610 standards.

The quantitative spreading efficiency is measured by the Solder Spread Factor (S):

S = ((D - H) / D) * 100%

Where:

• D = Diameter of a sphere having a volume equal to the solder sample tested (mm)
• H = Height of the spread solder droplet after reflow (mm)
• A spread factor S ≥ 90% indicates outstanding wetting and flux activity.

---

Soldering Flux Chemistries & Classifications

Flux is mandatory in soldering because low thermal energy (<300°C) cannot thermally dissociate base metal oxides. Flux cleans surfaces, depresses liquid surface tension, and prevents re-oxidation during reflow.

Traditional Flux Classifications (U.S. Mil-F-14256)

1. R (Rosin): Natural pine stump extract consisting primarily of abietic acid. Highly unreactive at room temperature; activates weakly at reflow temperatures (>130°C). Safe, non-corrosive residue, but possesses negligible cleaning power on tarnished surfaces.
2. RMA (Rosin Mildly Activated): Rosin formulated with small quantities of organic activators (e.g., amine hydrochlorides). Industry standard for military and aerospace assemblies requiring high reliability where post-soldering residue is non-conductive.
3. RA (Rosin Activated): Rosin fortified with aggressive ionic halide activators. Excellent cleaning on oxidized surfaces, but leaves corrosive, conductive residues. Mandates 100% solvent or aqueous cleaning.
4. Water-Soluble (WS / Organic Acid): Formulated from organic acids (citric, lactic, glutamic). Extremely active, but residues are hygroscopic and corrosive; requires thorough deionized water washing.
5. Inorganic Acid (IN): Formulated with zinc chloride (ZnCl2), ammonium chloride (NH4Cl), or hydrochloric acid (HCl). STRICTLY PROHIBITED IN ELECTRICAL/ELECTRONIC SOLDERING. Used exclusively in structural sheet metal, galvanized iron, and copper plumbing.

Modern IPC/J-STD-004 Classification System

The contemporary electronics and welding engineering industry classifies fluxes under IPC/J-STD-004 using a 4-character code:

[Flux Base Type] + [Activity Level] + [Halide Content Flag]

   FLUX BASE TYPE                ACTIVITY LEVEL             HALIDE CONTENT
   RO = Rosin                    L = Low (No wetting test)   0 = Halide-Free (<0.05%)
   RE = Resin                    M = Moderate               1 = Halide Present (≥0.05%)
   OR = Organic (Water-Soluble)   H = High
   IN = Inorganic Acid

The Dendrite / CAF Failure Mode: When organic acid fluxes (OR) or high-halide fluxes (ROH1) leave unwashed ionic residues on circuit assemblies operating under voltage in humid environments, an electrochemical reaction initiates. Metal cations (Sn2+, Pb2+, Cu2+) dissolve at the anode, migrate along moisture pathways, and plate out at the cathode as needle-like conductive whiskers (dendrites). This phenomenon, known as Conductive Anodic Filament (CAF) growth, produces sudden catastrophic short circuits.

---

Lead-Free Solders & Solder Alloys

For nearly a century, the electronics industry relied on eutectic tin-lead alloy (63 wt% Sn - 37 wt% Pb), which possesses a sharply defined eutectic melting point of 183°C (361°F). The European Union RoHS (Restriction of Hazardous Substances) Directive and global environmental mandates eliminated lead (Pb) due to its neurotoxicity and bioaccumulation in municipal groundwater.

SAC305 Physical Metallurgy

SAC305 represents the global manufacturing standard. Its microstructure consists of a primary β-Sn dendrite matrix interspersed with two finely dispersed eutectic intermetallic phases:

1. Ag3Sn Needle-like Precipitates: Silver additions form hard sub-micron Ag3Sn precipitates that pin dislocation motion, providing elevated yield strength, creep resistance, and cyclic fatigue resistance.
2. Cu6Sn5 Fine Particles: Copper additions saturate the molten solder, suppressing the rapid dissolution of thin copper surface-mount trace pads into the liquid solder puddle.

Thermal Impact of the Lead-Free Transition: Moving from Sn63-Pb37 (T_melt = 183°C) to SAC305 (T_melt = 217 - 220°C) forced industrial reflow temperatures to rise from 210 - 215°C to 235 - 255°C. This 35°C increase dramatically increases thermal stress on electronic packages, accelerates popcorning in plastic ICs, and quadruples the diffusion rate of brittle intermetallic compound layers.

---

Intermetallic Compound (IMC) Formation & Kinetics

True metallurgical adhesion in soldering does not occur by simple mechanical interlocking. It is driven by chemical reaction between the active tin (Sn) atoms in the molten solder and the substrate metal (e.g., Cu or Ni), precipitating a continuous layer of Intermetallic Compounds (IMCs).

              CROSS-SECTION OF SOLDER-COPPER INTERACTION

       Bulk Solder Matrix (β-Sn + Ag3Sn precipitates)
   ====================================================
       Cu6Sn5 (η-phase, Scallop-shaped morphology)
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
       Cu3Sn (ε-phase, Planar layer)
   ----------------------------------------------------
       Kirkendall Voids (Planar micro-cavities)  oo o oo
   ----------------------------------------------------
       Copper Substrate (Base Metal Pad)

Phase Evolution at the Copper-Solder Interface

1. Liquid-Solid Reaction (Reflow Stage): When molten tin contacts solid copper at 240°C, copper dissolves rapidly. Upon reaching saturation, the η-phase (Cu6Sn5) nucleates and grows as characteristic rounded "scallops" protruding into the liquid solder:

   6 Cu + 5 Sn → Cu6Sn5
   

2. Solid-State Aging Reaction (Service Stage): During prolonged thermal exposure or operational heating, solid copper diffuses into the Cu6Sn5 layer, precipitating a secondary, planar ε-phase (Cu3Sn) directly between the copper substrate and the Cu6Sn5 layer:

   Cu6Sn5 + 9 Cu → 5 Cu3Sn
   

The Dual Nature of Intermetallics

• Benign Initial Bond: An initial IMC layer thickness of 1.0 to 2.0 μm is proof of sound metallurgical wetting and provides excellent bond strength.
• Embrittlement Threshold: IMCs are stoichiometric intermetallic compounds possessing complex crystalline lattices. They are inherently hard, highly brittle, and possess near-zero fracture toughness (K_Ic ≈ 1.0 - 1.5 MPa·m^0.5) compared to the ductile tin matrix (K_Ic > 15 MPa·m^0.5). When the total IMC thickness exceeds 4.0 - 5.0 μm, any mechanical strain, thermal cycling, or drop shock concentrates stress at the IMC interface, causing catastrophic cleavage fracture.

Solid-State Growth Kinetics

The thickening of the interfacial intermetallic layer during thermal aging is governed by Fickian diffusion kinetics modeled by the parabolic rate law:

x(t) = x_0 + sqrt(k * t)
k = D_0 * exp(-Q / (R * T))

Where:

• x(t) = Total IMC thickness at time t (m)
• x_0 = Initial as-reflowed IMC thickness (typically 1.0 - 1.5 μm)
• k = Temperature-dependent parabolic growth coefficient (m²/s)
• Q = Apparent activation energy for diffusion (e.g., ≈ 70 - 90 kJ/mol for Cu-Sn)
• R = Universal gas constant (8.314 J/(mol·K))
• T = Absolute temperature (K)

---

Kirkendall Voiding, Gold Embrittlement & Thermomechanical Fatigue

The Kirkendall Effect & Voiding Mechanics

When a solder joint undergoes thermal aging (>100°C), atoms diffuse across the Cu / Cu3Sn / Cu6Sn5 boundaries. Crucially, the intrinsic diffusion rate of copper atoms diffusing outward into the solder is significantly faster than the inward diffusion of tin atoms into the copper substrate:

J_Cu >> J_Sn

To compensate for this unequal mass transport, vacancies must diffuse in the opposite direction—toward the copper substrate. When the flux of vacancies exceeds the annihilation capacity of local grain boundary dislocations, the vacancies supersaturate and coalesce into planar micro-voids, termed Kirkendall voids, concentrated within the thin Cu3Sn layer and along the Cu/Cu3Sn interface.

Under sudden mechanical shock (such as dropping an electronic device onto a hard surface), these continuous arrays of planar Kirkendall voids act as pre-existing Griffith microcracks. The joint unzips instantly with zero ductile plastic deformation, severing the electrical connection.

Gold Embrittlement

In high-reliability electronics, printed circuit boards and lead frames are often plated with Electroless Nickel Immersion Gold (ENIG) to prevent oxidation of the nickel underlayer. Gold dissolves into molten solder at a staggering rate (>1 μm/s).

• The 3% Rule: If the concentration of gold dissolved in a solder joint exceeds 3.0 wt%, gold will not stay in solid solution upon cooling. It precipitates as massive, acicular (needle-like) AuSn4 intermetallic plates scattered throughout the bulk solder.
• These brittle plates create planes of weakness that shear under minimal thermal stress, causing catastrophic joint embrittlement.

Room-Temperature Creep & Homologous Temperature

A critical, frequently misunderstood physical characteristic of solder is its tendency to undergo continuous plastic creep under ambient conditions. The homologous temperature (T_H) is defined as:

T_H = T_service / T_melt    (in Kelvin)

Creep deformation mechanisms (dislocation climb and grain boundary sliding) activate when T_H ≥ 0.4 - 0.5. For SAC305 (T_melt = 217°C = 490 K), standard ambient room temperature (20°C = 293 K) yields:

T_H = 293 K / 490 K = 0.60

Engineering Reality: Solder is operating in the high-temperature creep regime at room temperature. Solder joints under sustained mechanical tension or shear will continuously deform, relax, and eventually rupture over time without any increase in ambient temperature.

Thermomechanical Cyclic Fatigue: The Coffin-Manson Relation

In electronic packages, the silicon chip (α_Si ≈ 2.6 × 10^-6 / K), copper leadframe (α_Cu ≈ 16.5 × 10^-6 / K), and FR-4 epoxy laminate (α_FR4 ≈ 15 - 18 × 10^-6 / K) possess starkly different thermal expansion coefficients. Operational power cycles (ΔT) force cyclic shear strains (Δγ) into the solder joint:

Δγ = (L * Δα * ΔT) / (2 * h_s)

Where L is the distance to neutral point (DNP) from the chip center, Δα is the CTE mismatch, and h_s is the solder joint standoff height. The fatigue life N_f (cycles to failure) is modeled by the modified Coffin-Manson relationship:

N_f = C * (Δε_p)^(-m) * f^n * exp(Q / (R * T_max))

Where Δε_p is plastic strain range, f is cyclic frequency, and m is the Coffin-Manson fatigue exponent (typically 1.5 - 2.5).

---

Comprehensive Worked Engineering Example: Solid-State IMC Growth and Fatigue Shear Strain

Problem Statement

A high-reliability Ball Grid Array (BGA) package utilizes SAC305 solder balls (h_s = 0.40 mm standoff height) attached to copper PCB pads. The initial as-reflowed intermetallic layer (Cu6Sn5) has a thickness of x_0 = 1.20 μm.

1. The assembly operates in an automotive under-hood control unit at a sustained temperature of 125°C (398.15 K). The parabolic growth rate coefficient for total IMC (Cu6Sn5 + Cu3Sn) at this temperature is k = 3.60 × 10^-18 m²/s. Calculate the total IMC layer thickness after 2,500 hours of operation, and determine if it exceeds the critical 4.0 μm embrittlement threshold.
2. The distance from the center of the silicon die to the outermost corner solder ball is L/2 = 6.0 mm. The thermal expansion coefficient of the silicon die is α_die = 3.0 × 10^-6 / K and the FR-4 PCB is α_board = 17.0 × 10^-6 / K. During engine operation, the package thermal cycle spans ΔT = 80 K (from 25°C to 105°C). Calculate the cyclic engineering shear strain range (Δγ) imposed on the corner solder joint.

Step-by-Step Engineering Solution

Part 1: Intermetallic Growth Calculation

• Convert exposure time to seconds:

  t = 2,500 hr * 3,600 s/hr = 9.00 × 10^6 s
  

• Calculate the diffusional growth increment:

  Δx = sqrt(k * t) = sqrt((3.60 × 10^-18 m²/s) * (9.00 × 10^6 s))
  Δx = sqrt(3.24 × 10^-11 m²) = 5.692 × 10^-6 m = 5.69 μm
  

• Compute the total final IMC thickness:

  x_total = x_0 + Δx = 1.20 μm + 5.69 μm = 6.89 μm
  

• Engineering Evaluation: The final IMC thickness of 6.89 μm far exceeds the 4.0 μm critical embrittlement threshold. Extensive Kirkendall voiding will occur within the Cu3Sn sub-layer, rendering the corner ball vulnerable to shock failure.

Part 2: Thermomechanical Cyclic Shear Strain Calculation

• Differential thermal expansion:

  Δα = α_board - α_die = (17.0 - 3.0) × 10^-6 / K = 14.0 × 10^-6 / K
  

• Relative thermal displacement at corner joint (Δu):

  Δu = (L/2) * Δα * ΔT = (6.0 mm) * (14.0 × 10^-6 / K) * (80 K)
  Δu = 6.0 mm * 0.00112 = 0.00672 mm = 6.72 μm
  

• Imposed cyclic engineering shear strain (Δγ):

  Δγ = Δu / h_s = 0.00672 mm / 0.40 mm = 0.0168 = 1.68%
  

• Engineering Evaluation: A plastic shear strain range of 1.68% per thermal cycle represents severe cyclic loading. In conjunction with a brittle 6.89 μm IMC boundary, early thermomechanical fatigue cracking via grain boundary sliding and interface unzipping is guaranteed unless an underfill encapsulant is applied.

---

Industrial Scenarios & Certified Welding Engineer Exam Pitfalls

Real-World Field Disaster Scenario

An aerospace contractor manufactured marine navigational guidance boards using an Electroless Nickel Immersion Gold (ENIG) finish. During drop-impact qualification testing, 40% of the surface-mount Ball Grid Arrays spontaneously detached. Initial failure analysis falsely blamed insufficient soldering flux. Cross-sectional transmission electron microscopy (TEM) revealed widespread "Black Pad" syndrome combined with hyper-active Kirkendall voiding. During the immersion gold displacement process, excessive phosphorus was rejected from the electroless nickel bath, creating an ultra-brittle, hyper-corroded Ni3P "black pad" surface layer. When soldered, molten SAC305 formed discontinuous Ni3Sn4 crystals over a heavily micro-cavitated nickel-phosphorus interface. The entire batch of 350 guidance assemblies had to be scrapped at a cost of $1.85 million.

Common Exam Traps

Exam Trap 1: Soldering Temperature Limit A standard CWEng question asks: "An engineer specifies an alloy with a liquidus of 455°C for low-temperature electronic component attachment. Under AWS terminology, what is the process classification?" Because 455°C > 450°C (840°F), the process is legally brazing, NOT soldering! Confusing 450°C with 500°C or 400°C is a routine candidate pitfall.

Exam Trap 2: Solder Creep Threshold Candidates frequently assume metals only creep under severe furnace heating (>500°C). For solders, because their absolute melting point is low (T_m ≈ 456 - 500 K), room temperature (293 K) represents a homologous temperature T_H > 0.60. Solder creeps under its own residual assembly stress at room temperature.

Exam Trap 3: Flux Residues on Electronics Selecting an inorganic acid flux (such as zinc chloride or hydrochloric acid) for electrical work is an immediate exam failure. Inorganic acid fluxes leave ionic, hygroscopic salts that conduct electricity and rapidly corrode micro-traces through conductive anodic filament (CAF) mechanisms.