Contamination and Cleanliness IMAPS 2010 06

5/19/2018 Contamination and Cleanliness IMAPS 2010 06

1/172

Craig Hillman

Contamination and CleanlinessDeveloping Practical Responses to aChallenging Problem

IMAPS-Nordic Gotenborg, Sweden June 8, 2010


2/172

2004 - 2007

5110 Roanoke Place, Suite 101, College Park, Maryland 20740

Phone (301) 474-0607 Fax (240) 757-0053

www.DfRSolutions.com 2004 - 2009 2

Outline

Overview of Contamination

and Cleanliness

Motivation

Trends

Drivers

Temperature

Humidity / Moisture

Voltage and Electric Field

Contamination

Sources of Contamination

Printed Board Fabrication

Fluxes / Assembly

Handling

Use Environment

Detection and Measurement

Types of Equipment and

Processes

Industry Best Practices

Contamination Limits

Mitigations

Cleaning

Conformal Coating


3/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Motivation / Trends


4/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Why Contamination and Cleanliness?

Believed to be one of theprimary drivers of fieldissues in electronics today

Induces corrosion and metal migration(electrochemical migration ECM)

Intermittent behavior lends itself tono-fault-found (NFF) returns

Driven by self-healing behavior Difficult to diagnosis

Pervasive Failure modes observed on batteries, LCDs, PCBAs, wiring,

switches, etc.

Will continue to get worse


5/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Future of Contamination / Cleanliness

Continued reductions in pitch between conductors will makefuture packaging more susceptible

Increased use of leadless packages (QFN, land grid array, etc.)results in reduction in standoff

Will reduce efficiency of cleaning, which may lead to increasedconcentration of contaminants

Increased product sales into countries with polluted and tropicalenvironments (East Asia, South Asia, etc.)

ECM occurrence very sensitive to ambient humidity conditions

Pb-Free and smaller bond pads Require more aggressive flux formulations


6/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


The Future: Reduced Spacings

Critical distance effect? Two studies by DfR staff

NaCl seeding

Conformal coating over no-clean flux residues Over 300 coupons tested (IPC-B-25)

6.25, 12.5, and 25 mil spacings; 40C/93%RH, 65C/88%RH,85C/85%RH

No ECM at 25 mil spacings

Experience based onolder designs may notbe valid

Test coupons must

have the same spacing andsame electric field as actualproduct 2 mil (48 m)BGA Substrate Traces

4 mil (100 m)PCB External Traces (low

voltage line)

7 mil (170 m)Thin Shrink Small Outline

Package (TSSOP) Leads

5 mil (120 m)Peripheral Flip Chip Solder

Bumps

SpacingParameter


7/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


QFN and Dendritic Growth

Large area, multi-I/O and low standoff can trap fluxunder the QFN

Processes using no-clean flux should be requalified Particular configuration could result in elevated weak

organic acid concentrations

Those processes not using no-clean flux will likelyexperience dendritic growth without modification ofcleaning process

Changes in water temperature Changes in saponifier

Changes to impingement jets


8/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


QFN and Dendritic Growth (cont.)

The electric field strength between adjacent conductors is astrong driver for dendritic growth

Voltage / distance

Digital technology typically has a maximum field strength of0.5 V/mil

TSSOP80 with 3.3VDC power and 16 mil pitch

Previous generation analog / power technology had a maximumfield strength of 1.6 V/mil

SOT23 with 50VDC power and 50 mil pitch

Introduction of QFN has resulted in electric fields as high as

3.5 V/mil 24VDC and 16 mil pitch


9/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


QFN and Dendritic Growth (cont.)

Some component manufacturers are aware of thisissue and separate power and ground Linear Technologies (left) has strong separation power and

ground Intersil (right) has power and ground on adjacent pins


10/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Higher Activity Flux

Reduced bond pads = smaller volumes of solder paste (less flux) Oxide thickness remains the same

Requires higher activity fluxes

Copper

Copper

Oxide

Solder

Paste

6 mil stencils

4 mil stencils


11/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Drivers

Background on ECM


12/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Failure Mode

Why do you care about excessive

contamination or insufficient cleanliness lead

to?

Electrochemical Migration

(note: not Electromigration; completely different mechanism)

Understanding the mechanism provides

insight into the drivers and appropriatemitigations


13/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Electrochemical Migration (ECM)

Defined, per IPC-TR-476A, Electrochemical

Migration: Electrically Induced Failures in

Printed Wiring Assemblies (1997), as The growth of conductive metal filaments or

dendrites on or through a printed board under the

influence of a DC voltage bias

Too narrow


14/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


What is ECM?

Alternative definition

Movement of metal through an electrolytic solution under

an applied electric field between insulated conductors

Electrochemical migration can occur on or in almost

all electronic packaging

Die surface

Epoxy encapsulant

Printed board

Passive components Etc.


15/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


ECM Mechanisms

Some ECM Mechanisms have more definitivedescriptions

Dendritic growth Descriptor for ECM along a

surface that produces adendrite morphology

Tree-like, Feather-like

Conductive anodic filaments(CAF) Descriptor for migration within

a printed circuit board (PCB)


16/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


ECM Steps

Traditional electrochemical migration involvesfour steps

Path formation Electrodissolution

Ion migration

Electrodeposition

In ECM along internal surfaces (e.g., CAF),ion migration / electrodeposition co-exist


17/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Path Formation

Physio-chemical changes necessary to initiate ECM

Different meanings for different mechanisms

Believed to be the rate-limiting step

Dendritic growth

The creation of an electrolytic solution sufficiently

conductive Driven by relative humidity, contaminants,

delamination

Conductive anodic filaments (CAF) Degradation of the epoxy/glass interface


18/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Electrodissolution

Minimum requirements Electric field and a polar liquid

Usually water

Voltage > Anodicelectromagnetic force (EMF) 0.34 volts for Cu

Electrochemical reactionusually occurs adjacent toanode 2H

20 O

2+ 4H+ + 4e-

Increased concentration of H+

reduces pH to below 7.0

Dissolution of anode material (Cu) Cu Cu+ + e- or Cu Cu2+ + 2e

Halides are not required to induce dissolution Can increases susceptibility through changes in ion formation

Reduction in pH; increases solubility of metal ions (i.e, copper)

Cu+2

H+ H2

H+

H+

Copper

e e

H+

H+

Cu2

O

CuO Cu2O

NaClNaCl

Na+ Cl-

Cl-

Na+

CuCl H2O+

Cu2O

Cl-

CuCl++ H+

H2

H+

Cl-

Cu+2Pure Water --

Cu+1Chloride --


19/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Ionic Migration

DecreasingpH

v = E = q/(6r)

v is velocity, is mobility, E is field strength, q is charge, r is ionic radius, is viscosity

Also known as electrophoresis

Migration of charged

particles through a solution

under the influence of anelectric field

Positive ions travel along the

field lines From anode to cathode

Electrons travel in reverse

direction

Velocity can be computed


20/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Electrodeposition

Once ions reach the cathode, electrochemical reactions occur:

O2 + 2H20 + 4e- 4OH -

2H20+ 2e- 2OH- + H2

Cu

+

+ e

-

Cu or Cu2+

+ 2e

-

Cu Because cathodic electrodeposition is largely diffusion controlled, rate

depends on the metal ion concentration in the aqueous medium

Production of hydroxyl ions (OH-) at cathode can reduce the rate of

electrodeposition by combining metal ions with hydroxyl ions to formhydroxides Hydroxides remove metal ions from aqueous medium through partial

precipitation

Thus, ECM rate is correlated with the solubility product of the metal ionhydroxide

Ionic contaminates, such as Cl -, can further change the ECM rate byforming alternate reaction paths and by forming additional ionic species, Example: CuCl in the case of copper and Cl-.


21/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Dendrite Propagation

Growth along preferentialcrystallographic directions

Creates tree-likeappearance

Primarily along (111) plane

Branches in the [211] and

[110] directions

(111) is a close-packed (CP)plane in FCC structures.

Most metals are FCC CP planes are slip planes


22/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Drivers

Temperature / Moisture


23/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


ECM Drivers: Electrolytic Solution

Comprised of water and ions (thin film of moisture)

Where does the water come from? Ambient moisture in the air

Evaporation of absorbed moisture (surface ECM)

Measurement (techniques) Adsorption -- Quartz crystal, Ellipsometry

Absorption -- Weight gain

Measurement (units) Adsorption -- Monolayers of moisture, or areal mass density (ng/cm2) 1 monolayer = 31 ng/cm2

Absorption -- Percent change in weight

How much water?

Very dependent upon relative humidity Relatively insensitive to temperature

Even for moisture absorption (numerous internal interfaces)(faster diffusion)


24/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Water Adsorption (Metals)

Leygraf and Graedel, 2000Lee and Staehle, 1997

Metals

2

4

6

8

10

Monolayerso

fMoisture

0 Copper

Marcus, 2002


25/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Water Absorption

0.00

0.05

0.10

0.15

0.20

0.25

0 5 10 15 20 25 30 35

Time1/2

(hour1/2

)

Moist

ureContent(%)

85oC/85%RH

40oC/93%RH

Epoxy encapsulant


26/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Monolayers of Moisture

Insulator Material

_

AnodeCathode

Most measurements performed onnon-porous solids (metals, ceramics)

Surface of interest for ECM tends

to be polymeric (solder mask, FR4,

encapsulant)


27/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Moisture Adsorption (Polymer)

Monolayers over polymerdifficult to measure

Dueling processes ofadsorption and absorption

Influenced by surfaceproperties (hydrophilic,hydrophobic), bulk properties,and porosity

Parameters rarely measuredby materials supplier

Actual number of monolayersover polymeric surface as afunction of relative humidity is

debatable Estimate of few to several

hundred

ECM possible> 80%RH

ECM likelyCondensation

ECM unlikely60 80%RH

ECM rare< 60%RH

General guidelines

Partially dependent upon ionic solubility


28/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Deliquescence

Dissolution of ionisableresidues Resistance change can be

several orders of magnitude(der Marderosian, 1977)

Each inorganic compoundhas a different critical RH

HCl contaminatedsubstrates Dilution of contaminants at

70%RH (Zamanzadeh,1989)

%RH


29/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Condensation

What is condensation?

When surface moisture becomes visible?

The amount of adsorbed moisture at 100%RH?

Definition somewhat unclear Film thickness can range (metals)

Minimum (10 monolayers of moisture)

Dew (~1,000 monolayers of moisture)

Raindrops (~10,000 monolayers of moisture) When does condensation occur?

At 100%RH

When the surface temperature is below the dew point

temperature Presence of cracks and delaminations

Hygroscopic


30/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Problems with Condensation

Mimics behavior of bulk water Large decrease in resistance

5 monolayers has conductivity two orders of magnitude lowerthan bulk water

20 monolayers has a conductivity one order of magnitudelower

Dissolution of additional surface contaminants

Rapid decrease in time to failure

Days to seconds

Changes in migration behavior

Larger distances Over alternative surfaces (e.g., conformal coating)


31/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


*From Contamination Studies Laboratory, Inc., http://www.residues.com

Elapsed time 12 sec.

Dendritic Growth during Water Drop Test


32/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Pin to pin migration over conformal coating

DendriticGrowth


33/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Cracks and Condensation

Vapor pressure of water in a crack is reduced

Condensation of liquid occurs when partial

pressure matches reduced vapor pressure

Water molecules adsorb to surface; first

layers bonding energy is comparable to

chemical bond


34/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Capillary Condensation

ln RH = -2v/RTd, where is surface tension of water (~80 dyne/cm)

v is molar volume (18 cm3/mole)

R is gas constant (107 dynecmmole-1K-1)

T is absolute temperature

d is crack opening

d

k d


35/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Cracks and ECM

Greatly reduces humidity

necessary to induce

condensation Why ECM in cracked chip

capacitors happens

relatively rapidly

Why conformal coatingmust adhere to the board

surface

Why popcorning results in

elevated leakage currents in

plastic encapsulated

microcircuits (PEMs)

H i R id


36/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Hygroscopic Residues

Certain contaminants create conditions thatincrease moisture film thickness

Increase risk of condensation Ionic and non-ionic contaminants

Examples: Polyglycols

When present, turns surface from hydrophobic(water repelling) to hydrophilic (water attracting)

Non-ionic: Not detectable using ion

chromatography or Omegameter

L L


37/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Lessons to Learn

Do not underestimate relative humidity in controlledenvironments (telecom, server, storage)

Macro environment controlled to 40-60%RH

Micro environments can easily exceed 80%RH or greater forlong periods of time

Case study

Failure of hard drive controller due to ECM internal to the epoxyencapsulant

Environmental tests by DfR Solutions staff indicated ECM wouldnot initiate below a certain percentage of moisture absorption

Equivalent to 82%RH

Drivers

Minimal power dissipation during standby Use of metal enclosure (minimal air flow)

Possible outgassing of absorbed moisture from polymeric materials

L L ( )


38/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Lessons to Learn (cont.)

Do not overestimate potential for condensation inuncontrolled environments

Temperature rise of 5C often sufficient to preventcondensation Constant power on

Excessive condensation can initiate corrosion over

conformal coating Critical focus

Cleanliness

Housing design (prevent physical water from coming into

contact with board surface)


39/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Drivers

Voltage / Electric Field

El t i Fi ld (V lt /Di t )


40/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Electric Field (Voltage/Distance)

Voltage is a primary driver in two processes

Electrodissolution (oxidation reaction)

Ion Migration Electrodissolution

Applied voltage must exceed EMF

0.13 V for Sn/Pb, 0.25 V for Ni, 0.34 V for Cu,

0.8 V for Ag, and 1.5 V for Au

Ion migration

Force on the ions, and therefore velocity, is a function of

electric field strength

El t i Fi ld St th (N i l)


41/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053www.DfRSolutions.com 2004 - 2009 41

Electric Field Strengths (Nominal)

Logic devices

Previous generation 6.4 V/mm (SO32 --

1.27 mm pitch, 5 VDC)

Current generation 20 V/mm (TSSOP80 --

0.4 mm pitch, 3.3 VDC)

Power devices

Previous generation 64 V/mm (SOT23 --


Current generation 140 V/mm (QFN


Copper traces: 240 V/mm (0.5 mm spacing, 120 VDC)

IPC-2221A allows 300 V/mm (0.1 mm spacing, 30 VDC)


42/172

2004 - 2007



Copper traces

120 VDC with 20 to 30 mil spacing (5 to 6

V/mil)

IPC allows 15V / mil (IPC-2221A)

Electric Field (cont )


43/172

2004 - 2007



Electric Field (cont.)

Prior studies have demonstrated varying effects of

electric field on ECM

DfR (seeding of IPC-B-25 coupons with NaCl) 6V / 6.25 mil = 1 V/mil (dendritic growth)

42V / 25 mil = 1.7 V/mil (no dendritic growth)

Undefined combination of electric field and spacing to

initiate ECM

Conductive anodic filament (CAF) growth models

have demonstrated conflicting behavior

V/d, V/d2, V/d3, and V/d4

E Field and Contamination


44/172

2004 - 2007



E-Field and Contamination

0.0

1.0

1.5

2.0

0.5

6 10 14 18

Halide concentration (g/in2)

ElectricField(V/mi

l)

E Field and Dendritic Growth


45/172

2004 - 2007



E-Field and Dendritic Growth

Immersion silver (Ag)plating

85C / 85%RH / 10VDC Observation

Migration only at tip ofcomb pattern

Dendrites stopped growing

Why? Maximum electric field

strength

Electric Field Strength (ImAg Migration)


46/172

2004 - 2007



Electric Field Strength (ImAg Migration)

E-field (6.25 mil spacing) E-field (12.5 mil spacing)

Emax = 3.50 V/mil Emax = 3.02 V/mil

At 85C/85RH, the critical electric field strength to initiate

silver migration is between 3.0 to 3.5 V/mil


47/172

2004 - 2007



Drivers

Contamination

Contamination


48/172

2004 - 2007



Contamination

Two concerns

Hygroscopic contaminants (discussed)

Ionisable contaminants that are soluble in water (e.g., acids,

salts)

Ionic contaminants of greatest concern

Primarily anions; especially halides (chlorides and bromides)

Chemically aggressive due to chemical structure

Very common in electronics manufacturing process

Decreases pH; few metal ions found in dendrites are soluble at midto high pH. Cu dendrites require pH less than 5 to form.

Silver(I) ions are soluble at higher pH; reason it is one of easiest toform dendrites.

Cations primarily assist in the identifying the source of anions

Example: Cl with K suggests KCl (salt from human sweat)

Sources of Contaminants


49/172

2004 - 2007



Sources of Contaminants

Printed board fabrication process

Insufficiently cured polymers

Rinse water

Fluxes

Handling

Storage and use environment

Sources of Contaminants (cont )


50/172

2004 - 2007



Sources of Contaminants (cont.)

Rinse WaterNO4

Solder FluxWeak Organic Acids

Rinse Water, Air Pollution, Papers/ PlasticsSO4

Cleaners, Red PhosphorusPO4

Teflon, KaptonFl

Printed Board (flame retardants), HASL FluxBr

Board Fab, Solder Flux, Rinse Water, HandlingCl

Possible SourcesIon

Printed Board Fabrication Process


51/172

2004 - 2007



Printed Board Fabrication Process

One of the most common source of

contaminants

Greatest use of active/aggressive chemicals Low margin business (reverse auctions)

Increasing use of no-clean assembly process (last

chance to clean)

PCB Contaminants (examples)


52/172

2004 - 2007



PCB Contaminants (examples)

Etching Chloride-based: Alkaline ammonia (ammonium chloride), cupric chloride,

ferric chloride, persulfates (sometimes formulated with mercuric chloride)

Other: Peroxide-sulfuric acid

Neutralizer Hydrochloric acid

Cleaning and degreasing Hydrochloric acid, chlorinated solvents (rare)

Photoresist stripping methylene chloride as a solvent

Oxide Sodium chlorite

Electroless plating Sodium hypochlorite (in potassium permanganate)

Palladium chlorides (catalyst)

Printed Board Fab (other examples)


53/172

2004 - 2007



b ( p )

Bromide sources

Surface processes Solder masks, marking inks, and fluxes

Flame retardant

FR-4 Epoxy has used a brominated bisphenol A (TBBA)

epoxy resin

IPC-TR-476A: Bromide in epoxy resin can diffuse to the

surface during a high temperature process such as

soldering

Source of Contaminants: Fluxes


54/172

2004 - 2007



Chemicals used for preparing metal surfaces for soldering

High molecular weight chemistries

Slightly acidic

Optimum behavior

Maximum activity during reflow; minimum activity after reflow

Difficult balancing act

Flux nomenclature Rosin only (RO)

Rosin, midly activated (RMA)

Rosin activated

Water soluble Low residue (no-clean)

Flux Chemistry


55/172

2004 - 2007



y

Two types: Rosin and Water-Soluble (WS)

No-clean can be rosin (Indium 3549-HF) or water-soluble

(Kester 977)

No-clean based on ability to pass surface-insulation

resistance (SIR) and electrochemical migration (ECM) tests

Initial classifications from military (Mil-F-14256E);primarily focused on rosin-based fluxes

Rosin/Resin (R), Rosin Mildly Activated (RMA), Rosin

Activated (RA), Rosin Super Activated (RSA)

Primarily R and RMA used for electronics assembly; someRA

J-STD-004 Flux Classification


56/172

2004 - 2007



J

J-STD-004 (cont.)


57/172

2004 - 2007



J ( )

Materials (Definition)


58/172

2004 - 2007



( )

Rosin: Derived from tree sap. Mix of abietic acid and othercompounds

Resin: Natural and synthetic compounds designed to act likerosin

Organic (aka OA): Solution of organic acids and othercompounds (hydrohalides, amines, and amides). Tend to be

water soluble since they contain no rosin

Inorganic: Solution of inorganic acids. Tend to be watersoluble since they contain no rosin. Not as common asorganic (very aggressive)

Activity levels based on Copper Mirror Test (IPC TM-650)

Flux Classification Correlation


59/172

2004 - 2007



L0 - All R, Some RMA, Some Low Solids "No-Clean"

L1 - Most RMA, Some RA

M0 - Some RA, Some Low Solids "No-Clean"

M1 - Most RA, Some RSA

H0 - Some Water Soluble

H1 - Some RSA, Most Water Soluble

Fluxes (cont.)


60/172

2004 - 2007



5% of the market uses rosin

Preferred in the early days of electronics manufacturing

Insoluble residues encapsulated contaminants

Hydrophobic surface

Quasi-conformal coating

This is decreasing (concerns regarding environmental-friendliness of solvents required to clean)

25% uses water soluble fluxes

70% of the market uses no-clean

80-90% in consumer, computer, telecom markets

This is increasing

Fine pitch, low clearance, high density greatly decreases

cleaning effectiveness Electronics industry very cost sensitive (eliminate one process,

increase throughput)

No-Clean Flux Chemistry


61/172

2004 - 2007



Activators (weak organic acids) -- 0.5 - 5%

Wetting agents (surfactants) -- 1%

Polyglycols (polyethylene glycol & polypropylene glycol),alkylenic ethoxylate, and fluorinated esters

Chealtors -- 1%

Reacts with green copper salts, formed from activators and

copper, to create clear or white copper products. Masking so operator will not think of this as a corrosion product

Solvents -- remainder (93 98%)

VOC-free is primarily water; others are IPA, methanol, etc.

Flux Residues


62/172

2004 - 2007



Residues of no-clean soldering?

Water-soluble dicarboxylic acids

Hygroscopic polyethylene glycol ethers

List of potential weak organic acids (WOAs)

Benzoic, Butyric, Formic, Lactic, Malonic, Oxalic, Propionic,

Succinic, Citric, Glutaric, Adipic, Malic

Optimum flux

Acids are neutralized after soldering process

Residual wetting agents are minimized

Other Concerns


63/172

2004 - 2007



Combinatorial effects are not well known

Anecdotal indications of unexpected deleterious

reactions between fluxes (poor understanding ofthe root-cause)

Reactions between fluxes and industrial pollutants

may also accelerate corrosive behavior

Excessive flux

Typically occur with wave, selective solder, and

hand soldering operations

Handling / Storage / Environment


64/172

2004 - 2007



Handling Salts from human contact (KCl and NaCl)

Storage Cleaning chemicals

Outgassing

Polymeric materials

Use Environment Dust

Evaporated sea water

Industrial pollutants

Handling / Sweat


65/172

2004 - 2007



Composition of dissolved salts in water

Can include other biological molecules.

Main constituents, after the solvent (water),

Chloride, sodium, potassium, calcium, magnesium, lactate,

and urea.

Chloride and sodium dominate.

To a lesser but highly variable extent, iron, copper,urocanate (and the parent molecule histidine), and other

metals, proteins, and enzymes are also present.

The main concern regarding sweat is as a source ofchloride

Handling / Sweat (cont.)


66/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


0

20

40

60

80

100

120

140

160

Raw stock After Cleaning Handling (office) Handling

(exercise)

Handling (brow)

Type of Exposure

Con

taminationExtracted(

g)

Chloride

Sodium

Potassium

Calcium

Magnesium

Lactic acid

0.141.200.360.000.3946.610.00Handling (after wiping brow)5

0.090.920.410.000.3925.630.00Handling (after exercise)4

0.101.300.410.000.4914.350.00Handling (office environment)3

0.091.070.210.000.450.470.00After polish and clean2

0.071.000.260.000.432.140.00Raw stock aluminum1

SO4(g/in2)

PO4(g/in2)

NO3(g/in2)

Br

(g/in2)

NO2(g/in2)

Cl

(g/in2)

F

(g/in2)ID

0.141.200.360.000.3946.610.00Handling (after wiping brow)5

0.090.920.410.000.3925.630.00Handling (after exercise)4

0.101.300.410.000.4914.350.00Handling (office environment)3

0.091.070.210.000.450.470.00After polish and clean2

0.071.000.260.000.432.140.00Raw stock aluminum1

SO4(g/in2)

PO4(g/in2)

NO3(g/in2)

Br

(g/in2)

NO2(g/in2)

Cl

(g/in2)

F

(g/in2)ID

Influence of Pollutants: Creepage Corrosion


67/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Recent field issues with printed circuitboards (PCBs) plated with immersionsilver Sulfur-based creepage corrosion

Failures in customer locations withelevated levels of sulfur-based gases Rubber manufacturing

Sewage/waste-water treatment plants

Vehicle exhaust fumes (exit / entrance

ramps) Petroleum refineries

Coal-generation power plants

Paper mills

Landfills

Large-scale farms Automotive modeling studios

Swamps P. Mazurkiewicz , ISTFA 2006

Creepage Corrosion Failure of ImAg


68/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Corrosion product is semi-conductive (resistance of about 1Mohm).

Resistance decreases as humidity increases.

Traces sensitive to leakage current trigger the system failure.

Visual inspection required to identify failures (most are CNDs).

Initiation of

a short

Creepage Corrosion Mechanism


69/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Exposed Cu was consumed to form coppersulfide that could cause electrical shorts.

Creepage Corrosion: Observations


70/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Galvanic corrosion

Failures at locations with thin or

absent immersion silver

Edge of solder mask, deep in platedthrough hole barrel

Copper and silver in contact

Similar mechanism observedwith electroless nickel /

immersion gold (ENIG) plating

Corrosion of copper trace

at solder mask edge

Solder Mask

Laminate

Copper

Nickel

Creepage Corrosion: Observations


71/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Failure behavior not observed duringqualification testing [Mixed Flowing Gas(MFG) Testing Class III]

Reason 1: Lack of appropriate structures in

test vehicle

Reason 2: Test conditions may not beappropriate

Strong indication that creepagemechanism requires thatone or more MFG testparameters are exceeded

Especially %RH

Hillman: >75%RH

Cullen: 93%RH ----2005030520020402752IV

200502005020510020301702IIIA

----2005020510020302752III

1002020050103105301702IIA

----20050103105302702II

------------------------I

SO2

(ppb)

NO2

(ppb)

Cl2

(ppb)

H2S

(ppb)

Temp

(C)

RH

(%)

Class

MFG Test Conditions


72/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Are existing MFG test conditions still

relevant?

Different material system (silver, not copper) Changing environment (is there more / less

pollution?)

Industrial Pollutants (Sulfur-Based)


73/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


SO2 MFG Test

100ppb, 200ppb

Average annual outdoor

2-20ppb (USA) 25-100ppb (Asia)

24 hour

~150ppb (NAAQS / Telcordia)

150-600ppb (Industrial-USA)

100-1500ppb (Asia)

May not be critical forsulfidation of silver

Rate independent of SO2

concentration

H2S MFG Test

10ppb, 100ppb, 200ppb

Average annual outdoor/indoor

0.05 to 0.8ppb 24 hour (outdoors)

8 to 100ppb (State Regs)

24 hour (indoors)

500 to 20,000 ppb

May be more critical

70000

28500

10000

4075

500

200

10

4

10

4

1.5

0.6H2S

40000

15300

10000

3800

1000

380

100

38

100

38

100 ug/m3

38 ppbSO2

Inside

industrial

Adjacent to

industrial

Urban with

heavy traffic

or industrial

RuralControlled

environment

Clean

room

IEC 60721-3-3

Test Conditions (cont.)


74/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Carbonyl Sulfide (COS) Ignored by MFG

Outdoor levels can be higher than H2S

Nominal: 0.5 0.8ppb

Elevated: 80 ppb

Can be as corrosive

as H2S

Test Conditions (Relative Humidity)


75/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Influence of %RH somewhat

contradictory

Vernon reported a critical %RH

(70-80%)

Graedel reported an increasing

corrosion rate with increasing %RH

Driven by monolayers (ml) of

moisture

ln (ml) = 2.73 p/p0 - 0.366

(p/p0 is %RH)

Rice reported no influence of %RH

Relative Humidity


76/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Validation of Rices observation

%RH levels in ceramic hybrid and thick film resistors coated withhydrophobic silicone likely low

Important differentiation by mechanism Most references investigate the tarnish aspect of sulfidation

Creepage behavior is likely very sensitive to %RH

The rough surface of a polymeric material becomesconducive to material transport once micro-condensationwithin occurs.

Filling-in of surface pores may greatly reduce the adhesion ofthe polymer surface

Allows forces created by volumetric expansion of corrosionproduct to push the growth out to an adjacent conductor

Discussion


77/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Modification of MFG test specs may be appropriate

Elimination of SO2 gases

Increase in H2S concentrations (>200 ppb)

Possible intro of COS Elimination or reduction of Cl2

Speculation that formation of

AgCl inhibits sulfidation of silver Elevated Cl2 displays parabolic

behavior

Elevated H2S displays unlimitedgrowth

Pollutants: Not Always in Industrial Settings

Drywall Sulfur Fumes Blamed for A.C. & Electrical Equipment failures


78/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Chinese Drywall Cited in Building Woes The drywall is emitting sulfur-based gases that are corroding

air-conditioner coils, computer wiring and metal picture frames.

Drywall blamed for A.C. failures Air-conditioning coils have turned black, along with wiring,

piping and even silver jewelry.

"We have definitely identified that a combination of sulfide

gases are the cause of the corrosion," said Robert P. DeMott,

managing principal of Environ.

"Foul odors reported by people living in the homes may also be

caused by the combination of sulfur gases being released from

the drywall,

Chinese drywall class action lawsuit LEE COUNTY, Fla. - The Lawsuit was filed against Knauf

Plasterboard Tianjin Co., LTD, The Knauf Group, Rothchilt

International Limited and the Banner Supply Company.

Known as "Chinese Drywall", it was manufactured oversees

and was made from waste materials. As a result, it emits

sulfur compounds that corrode copper wiring and other metals

found in homes. Copper Corroded by Sulfur flumes


79/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Drivers

Others

Insulation


80/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


The influence of insulation (migration surface) onECM is poorly quantified

Hydrophobic surfaces superior Silicone

Solder mask / FR4 epoxy selection rarely based onability to resist ECM

Exposed epoxy glass is much more hydrophilic than mostsolder mask materials

Greater concern and investigation with CAF Degradation of insulation (epoxy/glass interface) results in

path formation

Conductor Material


81/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Conductors of concern SnPb (solder)

Sn alloy (Pb-free solder, lead plating)

Copper (traces, connectors, component leads)

Silver (conductive adhesives, thick film resistor)

Nickel (capacitor electrodes, lead plating)

Gold (connector plating) Harsanyi and Inzelt established a ranking of pure metals according to propensity

to migrate Ag > Pb > Cu > Sn

Ag will migrate under temperature/humidity/bias unless special prevention Alloying with an anodically stable metal such as palladium or platinum (MLCC) Using an organic coating (immersion silver plating)

Using a metallic coating (SMT resistors)

Cu and SnPb will alternately migrate

Propensity for Pb-free solder to migrate more or less than SnPb has not beendefinitively established Disagreement over which is more migratable (Sn or Pb)

Gold requires contamination to migrate (typically chloride)

Immersion Silver (ImAg)


82/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Effect of NaCl contamination level

Minimal effect at levels below 10 g/in2

Repressed dendritic growth at levels above 10 g/in2

AgCl is less soluble in water than Ag

0 g/in2 2 g/in2 10 g/in2


83/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Drivers

Time to Failure Models

Drivers of ECM


84/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Failure models can provide a more

quantitative understanding of the

primary drivers

Time to Failure (TtF) Models for ECM


85/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Relevant TtF models must include theprimary drivers of ECM

Electric field (voltage, distance)

Temperature

Relative humidity

Contamination

Other effects to consider

Conductor material

Insulator material

Potential TtF Models


86/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Fundamental chemical behavior Arrhenius

Eyring

Experimental observations Barton-Bockris

DiGiacomo

Similar corrosion-based mechanisms Peck

Howard

CAF

Arrhenius


87/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Widely used to describe variety of chemicalreactions

Limited to temperature effects

=kT

HAtf exp

A = scaling constant

H = activation energy (eV)k = Boltzmann constant (8.62 x10-5 eV/K)

T = temperature (K)

Arrhenius (Hornung)


88/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Based on dendritic growth of silver through borosilicate glass under an

electric field Activation energy determined to be 1.15 eV

Growth rate linear dependence on electric field.

No humidity effect

Not true ECM model (dry silver migration)

Proportionality constant must be empirically determined

= proportionality constantG = electrode spacing

V = voltage

H = activation energy (eV)k = Boltzmann constant

T = temperature (K)

= kT

H

V

G

tf exp

Eyring


89/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Extension of Arrhenius equation Takes into account multiple stresses and synergy with

temperature

Recommended by IPC SIR Handbook to determine accelerationfactors for ECM

Too comprehensive?

Number of stresses undefined

Number of unknowns increases twice as fast as the number ofstresses

Stress functions undefined (natural log, exponential, linear)

+

++

++

= 21exp ST

EDS

T

CB

kT

HATtf

Barton and Bockris


90/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Based on dendritic growth of silver in an electrolytic cell

h = conductor spacing

= interfacial energy (metal/solution)R = universal gas constant

T = temperatureF = Faraday constant

D = diffusion coefficient

c = metal ion concentration

= overpotential (function of dendrite rad.)i = current density at the dendrite tip

22

8

=DcF

RThtf

as i

Barton and Bockris

C t i t f i ti t l i t ti b t t f t i t


91/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Contains a term for migrating metal ion concentration, but not for contaminantconcentration

Dual effect of temperature

Linear dependence

D = D0exp

No humidity term

Good general model for dendritic growth

Requires dendritic growth to be rate limiting step

No terms for path formation, electrodissolution, and ion transport

Influence of ionic contamination on the various parameters, such as the

interfacial energy, would have to determined theoretically or measured throughexperimentation.

Zamanzadeh [11] reviewed this model

Concluded that [This] model does not give accurate growth predictions, but gives ageneral indication of the functional dependence of growth on a variety ofelectrochemical parameters.

DiGiacomo

r 2


92/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


( )

=

v

HrkTerfce

kTVV

zFD

r

MCt

kTH

T

f

2

lnln

2

11

2

0

C = ionic concentration (Ag+)

zF= electric field,

D0

= diffusivity,

r = radius of curvature of the dendrite tip,

l/M = the critical amount of metal ions that must migrate to achieve dendritic growth,

= degree of oxidation or fractional active surface, changes from metal to metal,

VT

= determined by the critical current density for dendritic growth,

H = activation energy, k = Boltzmann's constant, T = absolute temperature,

= ln (r50/r16), sigma for a lognormal distribution

Based on migration of silver in encapsulated packages

Derived from use of the Butler-Volmer equation Relates electrode potential to current density

DiGiacomo

S i tifi i it l t d it


93/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Scientific curiosity unless current density canbe related to measurable physical factors

Current density is a function of conductance and

electric field Conductance is function of contamination and

relative humidity.

Could be the basis of describing time tofailure in terms of contamination, relativehumidity, and e-field.

Peck

( )kTEfRHA n )(


94/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


tlife = time to failure,A0 = material constant

RH = relative humidity, n = empirical constant (2.66)

Ea = activation energy, k = Boltzmann constantT = temperature, f(v) = voltage function (power law, ~1.5)

( )kTEvfRHAt an

life = exp)(0

Galvanic corrosion of aluminum bond pads in encapsulated

microcircuits

Based on Eyring model

Review of previous research

Primary environments: 85/85, 110/85 Voltages: 5 to 70 VDC

Howard

w = conductor width


95/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Based on corrosion of conductors on ceramic substrates

Derived from Faradays Law

Maybe applicable to Huawei situation

tMV

wlhndFtf

=

w = conductor width

l = conductor lengthh = conductor thickness

d = density of conductor material

F = Faraday constantM = atomic weight of conductor

V = voltage bias

= resistivity of electrolyte

t = electrolyte thickness

Conductive Anodic Filament (CAF) Models


96/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


( )( ) tt

m

n

eff

f MMMMV

Lfat >= ,

1000

ECM Time to Failure Behavior

Little interest in time to failure model for


97/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Little interest in time to failure model for

dendritic growth

Why? Under high humidity (>90%RH) testing, strong

experimental indications that rate limiting step

may be nucleation/initiation Most data suggests ECM will initiate within 72 to

120 hours under test conditions, or not at all

Additional indicators

Growth vs. Nucleation (1a)

Higher levels of stress result in failure mechanism shift


98/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Higher levels of stress result in failure mechanism shift

Prevents acceleration of mechanism

6.25 mils 6.25 mils

10 g/in2 Br + 2 g/in2 Cl 10 g/in2 Br + 50 g/in2 Cl

40C/93%RH / 10VDC / HASL Finish

Gro

un

d/A

no

de

Power

/Ca

tho

de

Growth vs. Nucleation (1b)


99/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


40C/93%RH 85C/85%RH

No-Clean Rosin-Based Flux / Urethane Conformal Coating10VDC / HASL Finish

Growth vs. Nucleation (2a)


100/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Many drivers have critical stress levels

Relative humidity for deliquescence

Voltage for electrodissolution

Contaminant levels for ECM

Time to Failure

Stress

Wide variation in timeto failure over range of

stresses not observed

for ECM

Growth vs. Nucleation (2b)


101/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Direct evidence of critical

electric field strength

Recent study performed on

a dummy microcircuits of

copper on silicon

Dendrite formation instantaneous at right conditions

(presence of oxalic acid)

0.4V at 3m spacing

1.2V at 10m spacing

Growth vs. Nucleation (3)

D d iti h l t d d


102/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Dendritic morphology suggests dependence onnucleation

6.25 mils6.25 mils

Heterogenous nucleation Homogenous nucleation

Growth vs. Nucleation (4)

Where else are dendrites observed?


103/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Where else are dendrites observed?

Solidification of liquid metal

Analogy

Copper ions in solution nucleate on the walls ofthe cathode conductor (heterogeneous)

Solidification consist of nucleation and growth

At supercooled conditions, nucleation behaviordominates

Preventing ECM

If nucleation-dependent shift should be away


104/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


If nucleation dependent, shift should be away

from predicting time to failure and towards

preventing failure

The most effective method is through tight

controls of contamination


105/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Contamination / Cleanliness

Detection, Measurement, and Requirements

How to Measure Cleanliness?

Standard ion chromatography (IC) testingIPC TM 650 M th d 2 3 28A


106/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


IPC-TM-650, Method 2.3.28A

Submerge whole board; 75 IPA / 25 DI

Updated IC IPC-TM-650, Method 2.3.28.2

Submerge whole board; 10 IPA / 90 DI (Delphi requirements)

Modified IC Use of saponifiers or alternative solvent

Submerge whole board

Localized Testing C3 from Aqueous Technologies

PCB Cleanliness Control: Industry Specs

IPC-6012B, Qualification and PerformanceS ifi ti f Ri id P i t d B d S ti 3 9


107/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Specification for Rigid Printed Boards, Section 3.9 Requires confirmation of board cleanliness before solder

resist application

When specified, requires confirmation of board cleanlinessafter solder resist or solderability plating

Board cleanliness before solder resist shall not be

greater than 10 ug/in2

of NaCl equivalent (totalionics) Based on military specifications from >30 years ago

Board cleanliness after solder resist shall meet therequirements specified by the customer

Cleanliness Control: Test Procedures

IPC-6012B specifies a Resistance of Solvent Extract


108/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


(ROSE) method Defined by IPC-TM-650 2.3.25

IPC-6012B specifies this measurement should beperformed on production boards every lot Class 1 boards: Sampling Plan 6.5

Class 2 and 3 boards: Sample Plan 4.0

Sampling plan (example) If a lot contains 500 panels of a Class 2 product, 11 panels

should be subjected to ROSE measurements forcleanliness testing

Test Procedures: Common Problems

ROSE is the least sensitive of ionic measurement techniques 5 ug/in2 detected by ROSE is equivalent to ~20 ug/in2 detected by ion


109/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


g y q g ychromatography

Equipment is not calibrated

Insufficient volume of solution is used

Insufficient surface area Panels are preferred over single boards

Cut-outs are not considered when calculating surface area

Insufficient measurement time

7 to 10 minutes is preferred

Technique Technology Equivalency Factor

ROSE Static / Unheated 1

Omega-Meter Static / Heated ~1.5

Ionograph Dynamic / Heated ~2.0Modified-ROSE, Zero-Ion, etc. Varied ~4.0 (?)

Ion Chromatography 80C for 1 hr ~4.0

Test Procedures: Best Practice

Ion Chromatography (IC) is the gold standard


110/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Some, but very few, PCB manufacturers qualify lots

based on IC results

Larger group uses IC to baseline ROSE /

Omegameter / Ionograph (R/O/I) results

Perform lot qualification with R/O/I

Periodically recalibrate with IC (every week, month, or

quarter)

Cleanliness Control: Requirements

The majority of knowledgeable OEMs


111/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


completely ignore IPC cleanliness requirements

Option 1: Requirements are based onR/O/I test results, but adjusted for lack of

sensitivity

Most companies now specify 2.5 to 7 ug/in2

Option 2: Requirements are based on IC test

results and then monitored using R/O/I

PCB Cleanliness: Moving Forward

Extensive effort to update PCB Cleanliness Standards


112/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


IPC-5701: Users Guide for Cleanliness of Unpopulated

Printed Boards (2003) IPC-5702: Guidelines for OEMs in Determining Acceptable

Levels of Cleanliness of Unpopulated Printed Boards (2007)

IPC-5703: Guidelines for Printed Board Fabricators in

Determining Acceptable Levels of Cleanliness of UnpopulatedPrinted Boards (Draft)

IPC-5704: Cleanliness Requirements for Unpopulated

Printed Boards (2010)

PCB Cleanliness (cont.)

The old 10 ug/in2 (1.56 ug/cm2) being replaced


113/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Driven by Delphi requirements

Nominal Ionic Levels

Bare printed circuit boards (PCBs)

Chl id 0 2 t 1 /i h2 ( f 0 5 t 1)


114/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Chloride: 0.2 to 1 g/inch2 (average of 0.5 to 1)

Bromide: 1.0 to 5 g/inch2 (average of 3 to 4)

Assembled board (PCBA)

Chloride: 0.2 to 1 g/inch2 (average of 0.5 to 1)

Bromide: 2.5 to 7 g/inch2 (average of 5 to 7)

Weak organic acids: 50 to 150 g/inch2 (average of 120)

Higher levels

Corrosion/ECM issues at levels above 2 (typically 5 to 10)

Corrosion/ECM issues at levels above 10 (typically 15 to 25)

Corrosion/ECM issues at levels above 200 (typically 400)

General rule

Dependent upon board materials and complexity

Cleanliness Controls: Ion Chromatography

Contamination tends to be controlled through industrialspecifications (IPC-6012, J-STD-001)


115/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Primarily based on original military specification

10 g/in2 of NaCl equivalent Calculated to result in 2 megaohm surface insulation resistance (SIR)

Not necessarily best practice

Best practice is contamination controlled through ionchromatography (IC) testing

IPC-TM-650, Method 2.3.28A

*Based on R/O/I testing

PaulsGeneral

ElectricNDCEE DoD* IPC* ACI

Chloride (g/in2) 2 3.5 4.5 6.1 6.1 10

Bromide (g/in2) 20 10 15 7.8 7.8 15

Major Appliance Manufacturer (IC)

Incoming PCB Processed PCB

ContaminantMaximum Level

(ug/in2)

Maximum Level(ug/in

2)

Upper Control Limit(ug/in

2)


116/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Ammonium


117/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


200150Total Weak Organic Acids

64Sulfate

64Phosphate

64Nitrite

64Nitrate

21Fluoride

42Chloride

1510Bromide

**Upper control limits may be considered maximum levels

for high reliability applications or products used in

uncontrolled environments


118/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Mitigation -- Cleaning

PCB Cleaning: Process Flow

At a minimum, PCB manufacturers should clean thePCB:


119/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


PCB: Immediately before the application of solder resist

Immediately after the application of any solderability plating

HASL

Electroless Nickel and Immersion Gold

Immersion Tin

Immersion Silver

Some PCB manufacturers also perform a final clean Should not substitute cleaning after solderability plating

Residues from plating operations can become more difficultto remove with any time delay

PCB Cleaning Process: Requirements

Final rinse with deionized (DI) water 2-8 M is preferred; >10 M may be too aggressive


120/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


p y gg

Distilled water is insufficient

City water is unacceptable

Potential options Use of saponifier during the cleaning process

Heated DI water is nice, but not absolutely necessary

Common problems DI water is only used if specified by the customer

DI water is turned off to reduce water and energy usage

Failure to monitor DI water at the source Failure to alarm the DI water on the manufacturing floor


121/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053

www.DfRSolutions.com 2004 - 2009121

MitigationPCBA Cleaning

When to Clean?

Very high reliability applications Medical, Military, Avionics, Industrial, Telecom


122/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Sensitive circuitry High-impedance circuits High-frequency circuits

Operation in uncontrolled environments

Use of conformal coating Concern over compatibility

Note: Some high-reliability markets have moved

away from cleaning Automotive, Enterprise, etc.

Industry Standards on PCBA Cleaning

Requirements driven by J-STD-001 Mandates 10ug/in2 (1.56 ug/cm2) for ROL0 or ROL1 (others based on limit established

by user)


123/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Section 6.1

Assemblies should be cleaned after each soldering operation so that subsequentplacement and soldering operations are not impaired by contamination

Section 8.2.2

Cleanliness testing is not required (unless specified by the customer)

SPC not required (testing should be random, but sample plan not provided)

Minimum once every production shift

If any assembly fails, the entire lot shall be evaluated and re-cleaned and a randomsample of this lot and each lot cleaned since performing the last acceptable cleanlinesstest shall be tested

Some guidance provided by two handbooks

Guidelines for Cleaning of Printed Boards and Assemblies, IPC-CH-65A (1999)

Aqueous Post Solder Cleaning Handbook, IPC-AC-62A (1996)

Critical Aspects of PCBA Cleaning

Solder Paste / Flux Chemistry

Component Selection and Board Design


124/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Component Selection and Board Design

Equipment Batch vs. Inline

Critical equipment parameters

Cleaning Solution Includes solvent, chemistry, and temperature

Process Location When to clean?

Cleanliness Requirements and Assessment

J-STD-004 Flux Classification


125/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Whats Missing?

Minimal information on cleaning

Other than no-clean


126/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Information needed

How to clean (cleaning solution, cleaning process)

What to clean (amount and type of residues)

Why no cleaning information?

Partially driven by process temperature, board design, and

board materials

Missing Information (example)

Water Soluble is not equivalent to WaterWashable


127/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Rosin-based flux residues can be removed by

water in combination with a saponifier Non-water soluble flux residues can be removed

through impingement (residue is softened byheated water and knocked off by force ofsprayed water striking the residue)

No-Clean can sometimes be cleaned

Sometimes, it can not

Component Selection / Board Design

Challenges High density


128/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Low standoff

High density Number of components per area

Number of tall / wide components

Low standoff components Chip components (0402, 0603, etc.)

Area array components (BGA, CSP, etc.)

Bottom-terminated components (LGA, QFN, etc.)

Chip Components

Some debate about worst-case Larger components (e.g., 2512) have more surface area


129/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Smaller components (0402, 0603) tend to have lowerstandoffs (1-2 mil vs. 3-4 mil)

Potential design improvements for cleaning No traces under components (most common)

Smaller bond pads (lifts up the component)

Thinner solder mask (increases standoff)

Components in parallel (prevents blocking of cleaningsolution)

Board cutouts (primarily for larger chip components)

Area Array Components

Ball grid array (BGA) and chip scalepackages (CSP)


130/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


p g ( )

Standoff and ball-to-ball spacing shrinking 1.0 mm pitch = 20-25 mil standoff

0.4 mm pitch = 10-12 mil standoff

Limited design improvements available

Solutions more process-driven

Bottom-Terminated Components

Components of greatest concern Standoff similar to chip components; surface area similar to BGA

Fast growing component packaging


131/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Fast growing component packaging

Current styles partially mitigated by thermal pad Limited area under package to clean

Future versions problematic

Dual-row QFN (current standard is single-row) Finer-pitch QFN (currently 0.4 mm)

Land grid array

Limited design improvements available

QFN and Dendritic Growth

Large area, multi-I/O and low standoff can trap fluxunder the QFN


132/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Processes using no-clean flux should be requalified Particular configuration could result in weak organic acid

concentrations above maximum (150 200 ug/in2)

Those processes not using no-clean flux will likelyexperience dendritic growth without modification ofcleaning process Changes in water temperature

Changes in saponifier Changes to impingement jets

Best Practices: PCBA Cleanliness

Confirm incoming PCB cleanliness

Clean after soldering operations


133/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


g p

Control and measure

Water quality going into process

Assembly cleanliness

Cleaning Equipment

Two options In-line


134/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Batch (dishwasher)

Batch tends to dominate

In-line primarily justified based on volumes

High-volume manufacturing tends to not clean

Ultrasonic

Rarely used on PCBAs

Batch Cleaning Equipment

Motivations for batch cleaner Insufficient volume for a dedicated line

PCBA is reasonably sized


135/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Batch cleaner manufacturers EMC Global Technologies

Aqueous Technologies

Manncorp

Austin-American

The batch cleaner should be capable of monitoring outflow

Aka, built-in cleanliness tester or resistivity monitoring

Alarm tends to be set to 4 to 7 megaohms

Closed loop can be a plus in low-cost countries

Cleaning Process and Solution

Variables Pure DI or Saponifier (type and concentration)


136/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Temperature (120F to >175F)

Nozzle pressure

Agitation (spray in air)

Time

Cleaning Process (cont.)

The majority of water cleanable solder pasteshave their residues removed by means of


137/172

2004 - 2007


Phone (301) 474-0607 Fax (240) 757-0053


Contamination and Cleanliness IMAPS 2010 06

Documents

Transcript of Contamination and Cleanliness IMAPS 2010 06