ABSTRACT 
 
 
Title of Document: CHARACTERIZATION OF FR-4 PRINTED 
CIRCUIT BOARD LAMINATES BEFORE 
AND AFTER EXPOSURE TO LEAD-FREE 
SOLDERING CONDITIONS 
  
 Ravikumar Sanapala, Master of Science, 2008 
Directed By: Chair Professor, Michael G. Pecht, Department 
of Mechanical Engineering 
 
 
 The transition to lead-free soldering of printed circuit boards (PCBs) using 
solder alloys such as Sn/Ag/Cu has resulted in higher temperature exposures during 
assembly compared with traditional eutectic Sn/Pb solders. The knowledge of 
possible variations in the PCB laminate material properties due to the soldering 
conditions is an essential input in the selection of appropriate laminates. 
 An experimental study was conducted to investigate the effects of lead-free 
processing on key thermomechanical, physical, and chemical properties of a range of 
FR-4 PCB laminate materials. The laminate material properties were measured as per 
the IPC/UL test methods before and after subjecting to multiple lead-free soldering 
cycles. 
 The effect of lead-free soldering conditions was observed in some of the 
material types and the variations in properties were related to the material 
constituents. Fourier transform infrared (FTIR) spectroscopy and combinatorial 
property analysis were performed to investigate the material-level transformations 
due to soldering exposures. 
 
  
 
 
 
 
 
 
CHARACTERIZATION OF FR-4 PRINTED CIRCUIT BOARD LAMINATES 
BEFORE AND AFTER EXPOSURE TO LEAD-FREE SOLDERING 
CONDITIONS 
 
 
 
By 
 
Ravikumar Sanapala 
 
 
 
 
 
Thesis submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park, in partial fulfillment 
of the requirements for the degree of 
Master of Science, 
Mechanical Engineering. 
August, 2008 
 
 
 
 
 
 
 
 
 
Advisory Committee: 
Dr. Michael Pecht, Chair Professor of Mechanical Engineering 
Dr. Patrick McCluskey, Associate Professor, Mechanical Engineering 
Dr. Peter Sandborn, Associate Professor, Mechanical Engineering 
Dr. Diganta Das, Faculty Research Scientist, Mechanical Engineering 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
? Copyright by 
 
Ravikumar Sanapala 
 
2008 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ii 
 
Dedication 
I wish to dedicate this thesis to my parents for their unconditional love and support 
through out my life 
 
 iii 
 
Acknowledgements 
 I would like to express my sincere gratitude to my advisor, Prof. Michael 
Pecht, for giving me the opportunity to be involved in this experimental study, and for 
all his support and guidance. I am also very much indebted to Bhanu Sood for his 
consistent supervision and productive discussion on my work. I also thank  
Dr. Diganta Das for his helpful insights on giving directions to my thesis and constant 
guidance throughout my stay at University of Maryland. I wish to extend my thanks 
to Dr. Patrick McCluskey and Dr. Peter Sandborn for serving on my thesis committee 
and reviewing my work. 
 I want to personally thank Dr Michael Azarian for his valuable inputs related 
to my thesis. I would like to thank all the CALCE EPS consortium members for their 
support in performance of this study. I also appreciate and acknowledge the inputs 
from Louis Lin of NanYa Plastics Corporation, Taiwan, and Joe Beers of Gold 
Circuit Electronics, USA, during the course of my thesis. 
 Lastly, I wish to express my thanks to all my friends and faculty at CALCE 
for making it an enjoyable place to work. 
 
 iv 
 
Table of Contents 
 
Dedication..................................................................................................................... ii 
Acknowledgements......................................................................................................iii 
Table of Contents......................................................................................................... iv 
List of Tables ............................................................................................................... vi 
List of Figures............................................................................................................. vii 
Chapter 1: Introduction................................................................................................. 1 
1.1 Overview of PCB fabrication........................................................................ 1 
1.2 Literature Review.......................................................................................... 3 
1.3 Objectives of Thesis...................................................................................... 6 
1.4 Overview of Thesis....................................................................................... 6 
Chapter 2: Constituents of FR-4 Laminates.................................................................. 8 
2.1 Reinforcement............................................................................................... 8 
2.2 Resin system ................................................................................................. 9 
2.2.1 Curing agents ...................................................................................... 10 
2.2.2 Flame retardants.................................................................................. 12 
2.2.3 Fillers .................................................................................................. 13 
2.2.4 Accelerators ........................................................................................ 13 
Chapter 3: Test Materials and Exposure Conditions .................................................. 14 
3.1 Test materials.............................................................................................. 14 
3.2 Material construction .................................................................................. 17 
3.3 Exposure conditions.................................................................................... 18 
Chapter 4: Material Properties and Test Methods ...................................................... 20 
4.1 Glass transition temperature (T
g
)................................................................ 20 
4.2 Coefficient of thermal expansion (CTE)..................................................... 22 
4.3 Decomposition temperature (T
d
)................................................................. 25 
4.4 Time-to-delamination (T-260).................................................................... 26 
4.5 Water absorption......................................................................................... 28 
4.6 Flammability............................................................................................... 28 
Chapter 5: Results and Discussion.............................................................................. 31 
5.1 Glass transition temperature (T
g
)................................................................ 31 
5.2 Coefficient of thermal expansion (CTE)..................................................... 33 
5.3 Decomposition temperature (T
d
)................................................................. 36 
5.4 Time-to-delamination (T-260).................................................................... 40 
5.5 Water absorption......................................................................................... 41 
5.6 Flammability............................................................................................... 42 
Chapter 6:  Analysis of Results................................................................................... 45 
6.1 Fourier transform infrared (FTIR) spectroscopy analysis .......................... 45 
6.1.1 Introduction......................................................................................... 45 
6.1.2 Analysis............................................................................................... 47 
6.2 Combinatorial property analysis................................................................. 52 
Chapter 7: Summary and Conclusions........................................................................ 55 
Chapter 8: Contributions............................................................................................. 57 
 
 v 
 
Appendix I: Datasheet values for the material properties studied .............................. 58 
Appendix II: FTIR spectra of control and 6X reflowed samples of material B ......... 59 
Bibliography ............................................................................................................... 60 
 
 
 vi 
 
List of Tables 
 
Table 1: Some of the common PCB material types [4] ................................................ 2 
Table 2: Typical constituents of FR-4 laminates .......................................................... 8 
Table 3: Typical constituents of E-Glass [4] ................................................................ 9 
Table 4: Laminate material classification................................................................... 15 
Table 5: EDS analysis of laminates ............................................................................ 16 
Table 6: Material properties and test methods............................................................ 20 
Table 7: The T
g
 comparison between DSC and TMA methods ................................. 33 
Table 8: Control in-plane (warp and fill) CTE measurement results.......................... 35 
Table 9: Post exposure in-plane (warp) CTE measurement results............................ 35 
Table 10: Post exposure in-plane (fill) CTE measurement results ............................. 36 
Table 11: Typical functional groups and wave numbers present in epoxy systems... 48 
Table 12: Summary of variations in properties for materials B and H....................... 52 
 
 vii 
 
List of Figures 
 
Figure 1: Typical steps involved in FR-4 printed circuit assembly fabrication............ 2 
Figure 2: Typical glass-weave styles used in PCBs...................................................... 9 
Figure 3: Formation of DGEBA [4]............................................................................ 10 
Figure 4: Tetra-functional and multi-functional epoxy monomers [4]....................... 10 
Figure 5: Typical structures of DICY and phenolic-cured resin systems [15] ........... 11 
Figure 6: Chemical structure of Tetrabromobisphenol A (TBBPA) .......................... 12 
Figure 7: Chemical structure of DOPO (9, 10-Dihydro-9-oxa-10-
phosphophenanthren-10-oxide) .................................................................................. 13 
Figure 8: Chemical structure of Imidazole ................................................................. 13 
Figure 9: ESEM picture showing filler particles in material B .................................. 16 
Figure 10: Flow of test materials ................................................................................ 17 
Figure 11: 6-ply laminate structure............................................................................. 18 
Figure 12: 12 layered fabricated board structure ........................................................ 18 
Figure 13: Lead-free reflow profile used for the exposures ....................................... 19 
Figure 14: Perkin Elmer DSC..................................................................................... 21 
Figure 15: Glass transition temperature measurement plot (material J)..................... 22 
Figure 16: Perkin Elmer TMA.................................................................................... 23 
Figure 17: Out-of-plane CTE measurement plot (material B).................................... 23 
Figure 18: In-plane CTE (warp) measurement plot (material B) ............................... 24 
Figure 19: In-plane CTE (fill) measurement plot (material B)................................... 24 
Figure 20: Shimadzu TGA 50..................................................................................... 25 
Figure 21: Decomposition temperature measurement plot (material F)..................... 26 
Figure 22: T-260 measurement plot (material B) ....................................................... 27 
Figure 23: Typical PCB combustion steps [31].......................................................... 29 
Figure 24: Test set up for flammability measurement................................................ 30 
Figure 25:  Effect of lead-free soldering exposures on T
g 
of the laminates................ 32 
Figure 26: Effect of lead-free soldering exposures on out-of-plane CTE of the 
laminates ..................................................................................................................... 34 
Figure 27: Effect of lead-free soldering exposures on T
d
 of the laminates ................ 36 
Figure 28: T
d
 comparison between materials A and J ................................................ 37 
Figure 29: T
d
 comparison between materials A and G1............................................. 37 
Figure 30: Effect of type of curing agent on T
d
 of the laminates ............................... 38 
Figure 31: T
d
 comparison between materials C1 and C2 ........................................... 38 
Figure 32: Effect of presence of fillers on T
d
 of the laminates................................... 39 
Figure 33: T
d
 comparison between materials A and B ............................................... 39 
Figure 34: T
d
 comparison between materials H and I................................................. 40 
Figure 35: Effect of type of flame retardant on T
d
 of the laminates........................... 40 
Figure 36: Effect of lead-free soldering exposures on T-260 of the laminates .......... 41 
Figure 37: Effect of lead-free soldering exposures on water absorption.................... 42 
Figure 38: Effect of lead-free soldering exposures on flammability of the laminates 
(total burning time) ..................................................................................................... 43 
Figure 39: Effect of lead-free soldering exposures on flammability of the laminates 
(Average burning time)............................................................................................... 43 
Figure 40: Common molecular bending movements [38] .......................................... 46 
 
 viii 
 
Figure 41: Typical functional groups present in the epoxy laminate systems............ 50 
Figure 42: FTIR spectrum of control sample with functional groups identified 
(material H)................................................................................................................. 51 
Figure 43: FTIR spectra of a control and 6X reflowed sample (material H).............. 51 
Figure 44: Effect of baking on the T
g
 of laminates..................................................... 54 
 
1  
Chapter 1: Introduction 
Printed circuit boards (PCBs) are the baseline in electronic packaging upon which 
electronic components are formed into electronic systems. The directive on 
Restriction of certain hazardous substances (RoHS) [1], and the Waste electrical and 
electronic equipment (WEEE) directive [2] have led to the adoption of lead-free 
soldering conditions in the PCB assembly process. The drive towards lead-free 
electronics has resulted in the use of lead-free solders in the PCB assembly. Common 
lead-free solder alloys such as Sn/Ag/Cu and Sn/Ag typically require a peak reflow 
temperature increase of 30-40
o
C for longer time periods compared to eutectic Sn/Pb 
reflow soldering conditions [3]. Rework and repair of assembled circuit boards also 
contribute to additional high temperature exposures. These exposures can alter the 
circuit board laminate material properties thereby creating a shift in the expected 
reliability of the PCB and the entire electronic assembly. The dependence of the 
laminate properties on the material constituents combined with the possible variations 
due to lead-free soldering exposures has not yet been extensively investigated.  
1.1 Overview of PCB fabrication 
The basic building blocks of a PCB are composites of resin and reinforcement. A 
wide variety of resin and reinforcement types that are commonly used in the PCB 
industry are listed in Table 1. The nomenclature shown for each type of material 
corresponds to the grades developed by National Electrical Manufacturers 
Association (NEMA). 
 
2  
Table 1: Some of the common PCB material types [4] 
Nomenclature Reinforcement Resin Flame retardant 
FR-2 Cotton paper Phenolic Yes 
FR-3 Cotton paper Epoxy Yes 
FR-4 Woven glass Epoxy Yes 
CEM-1 Cotton paper/woven glass Epoxy Yes 
CEM-2 Cotton paper/woven glass Epoxy No 
CEM-3 Woven glass/matte glass Epoxy Yes 
 
FR-4 PCB is a composite of epoxy resin with woven fiberglass reinforcement and it 
is the most widely used printed circuit board (PCB) material. The steps involved in 
the fabrication of FR-4 printed circuit assembly (PCA) are shown in Figure 1. 
Glass Raw Materials
(e.g., Silica, Calcium oxide, Aluminum oxide)
Copper Foil
Glass Fiber Production
(Melting, Formation, Coating/Binders)
Glass Fabric 
(e.g, 1080, 2116, 7628 weave styles)
Prepregs 
(B-stage resin)
Cores (Laminates)
(C-stage resin)
Resin Raw Materials
(Petrochemicals)
Epoxy Resin Production 
(A-stage resin)
Additives
?Curing agents
?Flame 
Retardants
? Fillers
? Accelerators
Printed Circuit Board
? Single/double sided
? Multi-layered 
Oxide Coatings
Drilling
Plating Materials
Solder Mask
Fluxes, Etchants, 
Cleaners
Coupling 
Agent
Electronic 
Components
Reflow/Wave 
Soldering
Assembly Process
Printed Circuit Assembly
 
Figure 1: Typical steps involved in FR-4 printed circuit assembly fabrication 
Glass raw materials are melted in a furnace and extruded to form fiberglass 
filaments that are combined into strands of multiple fiber yarn. Yarns are then weaved 
to form fiberglass cloth. A coupling agent, typically an organosilane, is coated onto 
the fabric to improve the adhesion between organic resin and inorganic glass. Resin is 
obtained from processing the petrochemicals and in its pure (uncured) form is called 
 
3  
A-stage resin. Additives such as curing agents, flame retardants, fillers, and 
accelerators are added to the resin to tailor the performance of the board. 
A prepreg is fabricated from a glass fabric impregnated with the semi-cured (B-
stage) epoxy resin. Multiple prepregs are thermally pressed to obtain a core or 
laminate (C-stage resin). Copper foil is then typically electrodeposited to obtain a 
copper clad laminate. Several prepregs and cores (with copper cladding etched as per 
the circuit requirements) are stacked together under temperature and pressure 
conditions to fabricate a multi-layered PCB. Through-holes and micro-via 
interconnects are drilled in the PCB as per the application specific design data and 
then plated with copper. Solder mask is applied on the board surface exposing the 
areas to be soldered. Flux is applied at regions where the electronic components are to 
be soldered. The boards are then subjected to reflow and/or wave soldering process 
depending upon the type of components (surface mount or through-hole) to obtain the 
printed circuit assembly. 
1.2 Literature Review 
Very few studies were conducted to characterize FR-4 laminate materials and 
assess the variations in material properties due to multiple lead-free soldering profile 
exposures. Bergum [5] characterized some of the laminate materials (low T
g  
DICY & 
non-DICY-cured, mid T
g
 halogen-free, and high T
g
 DICY & non-DICY-cured) by 
analyzing the material properties such as glass transition temperature (T
g
), coefficient 
of thermal expansion (CTE), decomposition temperature (T
d
), and flexural modulus 
combined with their variations due to simulated thermal excursions. The thermal 
exposures were carried out by standard analytical techniques, such as differential 
 
4  
scanning calorimetry (DSC), thermomechanical analysis (TMA), dynamic 
mechanical analysis (DMA), and thermogravimetric analysis (TGA) to simulate the 
thermal excursions experienced by the laminates during assembly process. Multiple 
cycles were repeated on the same sample at peak temperatures of 235
o
C and 260
o
C to 
observe the variations in properties. 
Kelley, et al., ([6], [7]) reported the effect of thermal exposures associated with 
typical Sn/Pb and Sn/Ag/Cu reflow conditions on epoxy-based laminates by assessing 
the decomposition temperature. The laminates were exposed to multiple thermal 
cycles in TGA with peak temperatures of 235
o
C and 260
o
C and a change in the 
percentage weight loss of the sample was monitored. The study illustrated the 
possibility of rapid thermal degradation in the epoxy systems of conventional high T
g
 
materials with repeated cycling at peak temperature of 260
o
C. The authors also 
demonstrated the usage of a methodology in the selection of lead-free compatible 
laminates for appropriate applications. The methodology was developed based on the 
PCB design features such as overall thickness, number of layers, copper foil weight, 
aspect ratio, resin content and process conditions such as number of reflow cycles. 
The study was focused on four types of laminate materials with different 
combinations of T
g
 and T
d
. 
Kelley [8] also illustrated the differences in four types of laminate materials 
(combination of low T
g
, high T
g
, low T
d
, and high T
d
) by assessing the properties of 
T
g
, T
d
, thermal expansion and time-to-delamination (T260/T288). The study also 
reported the evaluation of reliability of multilayered PCBs fabricated out of the four 
materials. Survivability of the boards with multiple lead-free soldering exposures was 
 
5  
assessed. Interconnect stress test (IST) was performed before and after exposure of 
the boards to multiple reflow cycles to estimate the effect of exposures on the 
reliability of PCBs. 
Ehrler [9]  in his study investigated the response of two types of DICY-cured epoxy 
laminate materials to lead-free reflow cycles by assessing the reliability with repeated 
reflow tests and thermal cycling tests. He demonstrated that the DICY-cured 
materials tested were not able to withstand the thermal stresses during lead-free 
soldering exposures 
Christiansen et al., [10] reported the comparative analysis on different types of 
epoxy-based laminate materials (high T
g
, mid T
g
 and low T
g
; halogen-free and 
halogenated; DICY and non-DICY cured) by measuring the properties of T
g
, CTE, 
T
d
, and T260/T288. 
Valette et al., [11] have also followed the approach of characterizing epoxy 
laminates based on material properties such as T
g
, CTE, T
d
, dielectric constant and 
assessed the performance of certain laminates that have higher thermal stability and 
lower dielectric constant in comparison to conventional FR-4 laminates. 
Pecht et al., [12] characterized a range of PCB laminate materials (low T
g
 FR-4, 
high T
g
 FR-4, polyimide, cyanate ester, and bismaleimide triazine) based on the 
moisture absorption capability at various relative humidity and temperature levels.  
Qi et al., [13] assessed the effect of type of PCB material on the durability of solder 
joints by characterizing certain types of laminate materials (low T
g
 FR-4, high T
g
 FR-
4, and polyimide). 
 
6  
1.3 Objectives of Thesis 
Based on the literature review, there have been few attempted studies that were 
conducted on characterizing certain types of FR-4 laminates and assessing the impact 
of lead-free soldering assembly conditions on reliability of printed circuit boards. To 
date, no comprehensive report is available on the effect of lead-free soldering 
exposures on laminate material properties. Furthermore, a wide variety of laminate 
types that are commercialized as ?Lead-free process compatible? are available 
recently and selection of appropriate laminates has been a challenge for the electronic 
industry. An insight into the laminate material constituents and variations in their 
material properties due to lead-free soldering exposures is essential in the selection of 
appropriate laminate materials. The broad objective of this thesis is to characterize a 
wide range of commercially available FR-4 PCB laminate materials and investigate 
the effects of lead-free processing on the thermomechanical, physical, and chemical 
properties. The analysis is aimed at correlating the properties to the material 
constituents of laminates. 
1.4 Overview of Thesis 
The constituents of FR-4 laminates are discussed in Chapter 2. The test materials 
and construction is presented in Chapter 3. Lead-free soldering exposures namely, 
three reflow cycles, six reflow cycles, and combination of one wave and two reflow 
cycles, considered for this study are also discussed in Chapter 3. In Chapter 4, the 
definitions of the material properties and the measurement procedures as per IPC /UL 
test standards are discussed. Results and the discussion on the effect of lead-free 
 
7  
soldering exposures on the material properties are presented in Chapter 5. Chapter 6 
consists of discussion on Fourier transform infrared spectroscopy (FTIR) analysis and 
combinatorial material property analysis to investigate the potential material-level 
transformations due to the soldering exposures. Summary and conclusions of the 
research are listed in Chapter 7. Research contributions are discussed in Chapter 8. 
 
8  
Chapter 2: Constituents of FR-4 Laminates 
The typical constituents of a FR-4 laminate are listed in Table 2 and are discussed 
in detail in the subsequent sections. Each of these constituents is important in its own, 
and in combination they determine the properties of the laminates. 
Table 2: Typical constituents of FR-4 laminates 
Constituent Major function(s) Example material(s) 
Reinforcement 
Provides mechanical strength and electrical 
properties 
Woven glass (E-grade) fiber 
Coupling agent 
Bonds inorganic glass with organic resin 
and transfers stresses across the matrix 
Organosilanes 
Resin Acts as a binder and load transferring agent Epoxy (DGEBA) 
Curing agent 
Enhances linear/cross polymerization in the 
resin 
Dicyandiamide (DICY), 
Phenol novolac (phenolic) 
Flame retardant Reduces flammability of the material 
Halogenated (TBBPA) or 
Halogen-free (Phosphorous 
compounds) 
Fillers Reduces thermal expansion Silica 
Accelerators 
Increases reaction rate, reduces curing 
temperature, controls cross-link density 
Imidazole, 
Organophosphine 
2.1 Reinforcement 
The woven glass (generally E-grade) fiber cloth acts as reinforcement for the 
laminate, primarily providing mechanical strength and electrical properties. Glass 
fabric is woven with two sets of fiber yarns (fibers are combined into strands of 
multiple fiber yarn). Warp yarn fibers lie in the machine direction of the fabric while 
those of fill yarn lie perpendicular to the warp direction. The typical constituents of 
glass (E-grade) used in FR-4 laminates are shown in Table 3. The type of glass-weave 
style is defined by the parameters such as glass fiber bundle diameter, number of fiber 
bundles, and linear density of the fabric ([4], [14]). The difference in the fabric styles 
that are commonly used in PCBs (1080, 2116 and 7628 weave styles) is represented 
 
9  
in Figure 2. The order of decreasing fabric density and thickness of the fabric among 
the glass-weave styles shown is 7628>2116>1080. 
Table 3: Typical constituents of E-Glass [4] 
Constituent Composition (%) 
Silicon dioxide (SiO
2
) 
Calcium oxide (CaO
2
) 
Aluminum oxide (Al
2
O
3
) 
Boron oxide (B
2
O
3
) 
Sodium oxide (Na
2
O) + Potassium oxide (K
2
O) 
Magnesium oxide (MgO) 
Iron oxide (Fe
2
O
3
) 
Titanium oxide (TiO
2
) 
Fluorides 
52-56 
16-25 
12-16 
5-10 
0-2 
0-5 
0.05-0.4 
0-0.8 
0-1 
 
 
                      1080 weave                                    2116 weave                          7628 weave 
Figure 2: Typical glass-weave styles used in PCBs 
2.2 Resin system 
The resin system of a FR-4 laminate primarily consists of bi, tetra or multi-
functional epoxy groups. Resin is derived from the reaction of Bisphenol-A with 
Epichlorohydrin which creates ?Diglycidyl Ether of Bisphenol A? called DGEBA, 
also referred as Oxirane (shown in Figure 3). The epoxy groups present in DGEBA 
react in subsequent resin polymerization and result in curing of the resin system. 
Higher cross-linking in the cured system is achieved by the use of epoxy monomers 
with more than two epoxy functional groups per molecule (shown in Figure 4).  
 
10  
 
Figure 3: Formation of DGEBA [4] 
 
 
Figure 4: Tetra-functional and multi-functional epoxy monomers [4] 
Additives such as curing agents, flame retardants, fillers, and accelerators are added 
to the resin system to improve the performance of laminate. These are added to the A-
stage resin before prepreg fabrication. 
2.2.1 Curing agents 
Curing agents are added to the resin system to enhance cross-linking thereby 
thermosetting the composite structure. They are usually aliphatic or cycloaliphatic 
amines and polyamines or amides. Most widely used curing agents in the PCB 
industry are dicyandiamide (commonly known as DICY) and phenol novolac 
(phenolic). The chemical formulae of DICY and phenolic curing agents and typical 
molecular structures of the respective cured resin systems are shown in Figure 5. 
Epoxide groupEpoxide group
 
11  
 
 
 
 
Dicyandiamide (DICY) DICY-cured epoxy system 
 
 
 
Phenol novolac (phenolic) Phenolic-cured epoxy system 
 
Figure 5: Typical structures of DICY and phenolic-cured resin systems [15] 
The properties of DICY and phenolic cured systems differ because of the inherent 
differences in their molecular structures. DICY and its cured epoxy systems are linear 
aliphatic molecules compared to the aromatic structure of phenolic and its cured 
epoxy systems. This makes the phenolic-cured systems to be more thermally stable 
than DICY-cured systems. DICY-cured systems are more hydrophilic than phenolic-
cured systems due to the presence of highly polar carbamidine-carbamide bond in 
their chemical structure. Also, at higher temperatures the strong polar nitrogen atom 
present in the DICY-cured system can destabilize the brominated epoxy resin 
resulting in the production of corrosive bromide ions. Overall, phenolic-cured resin 
systems offer better thermal resistance, chemical resistance, humidity resistance and 
improved mechanical properties but poor processability (e.g., drilling) compared to 
DICY-cured systems [15]. 
H
2
NCNHC?N
NH
==
OH
CH
2
OH
CH
2
n
OH
 
12  
2.2.2 Flame retardants 
Flame retardants are added to the resin system to reduce flammability of the 
laminate material. Tetrabromobisphenol-A (TBBPA) is the most commonly used 
halogenated (brominated) flame retardant for epoxy resin systems. The chemical 
structure of TBBPA is shown in Figure 6. Brominated flame retardants decompose 
during ignition, and retard combustion by trapping radicals generated from resins and 
by forming gas-phase barriers against oxygen [16]. 
 
Figure 6: Chemical structure of Tetrabromobisphenol A (TBBPA) 
 
Halogen-free flame retardant epoxy systems are gaining importance recently 
because of the shift in market trends to halogen-free products due to the public 
consciousness of the hazardous halogenated products, the industrial end user 
initiatives and the environmental legislations. Phosphorous compounds (such as 
Organophosphate esters) and metal hydroxides (such as Aluminum hydroxide and 
Magnesium hydroxide) are some of the commonly used halogen-free flame 
retardants. Organophosphate esters work by forming flame retarding glass-like 
barriers on the resin surface during ignition thereby cutting off oxygen necessary for 
combustion ([17], [18], [19]). The chemical structure of one of the most commonly 
used organophosphate ester (DOPO) is shown in Figure 7. Metal hydroxides retard 
flames by absorbing the heat during ignition, when added in large amounts [20]. 
 
13  
.  
Figure 7: Chemical structure of DOPO (9, 10-Dihydro-9-oxa-10-
phosphophenanthren-10-oxide) 
2.2.3 Fillers 
Fillers are added to the resin system for specific performance requirements such as 
lowering out-of-plane coefficient of thermal expansion (CTE) and to prevent barrel 
cracks in plated through holes. Fillers are also added to enhance the flame retardancy 
to meet UL 94V-0 flammability rating and to reduce material costs (as they replace 
resin). Typical fillers used in FR-4 PCBs are Silica and Aluminum silicate. 
2.2.4 Accelerators 
Accelerators are added to the resin system to increase the curing rate, reduce curing 
temperature and control cross-link density. Imidazole is one of the commonly used 
accelerators in FR-4 PCBs. The chemical structure of Imidazole is shown in Figure 8. 
NN
CH
2
CH
3
 
Figure 8: Chemical structure of Imidazole 
 
14  
Chapter 3: Test Materials and Exposure Conditions 
Laminate material classification, material construction and lead-free soldering 
exposure conditions used for the study are discussed in the following subsections. 
3.1 Test materials 
Fourteen commercially available FR-4 PCB laminates from two suppliers (I and II) 
were used in this study. The laminate materials used for the study were classified as 
shown in Table 4. Suppliers were chosen from different geographic locations with 
suppliers I and II from Taiwan and Japan respectively. Laminates were broadly 
categorized on the basis of their glass transition temperatures (T
g
) as high-T
g
 
(T
g
>165
o
C), mid-T
g
 (140
o
C<T
g
<165
o
C), and low-T
g
 (T
g
<140
o
C) materials. Under 
each T
g
 category, laminates were grouped based on the type of curing agent 
(dicyandiamide [DICY] or phenol novolac [phenolic]), presence or absence of fillers, 
and type of flame retardants (halogenated or halogen-free). Laminates marketed for 
the high frequency applications (materials E and K) were also considered for the 
study. Coupling agents and accelerators were not controlled in the laminate materials. 
Laminates amongst the pairs C1/C2, D1/D2 and G1/G2 differ only in the presence 
of fillers. The difference between C1/C2 and G1/G2 pairs is the ratio of high T
g
 resin. 
Material pair C1/C2 has higher percentage of tetrafunctional high T
g
 resin than that of 
G1/G2. The material type C2 also has a higher percentage of fillers than G2. The rest 
of the formulations are same amongst C1/C2 and G1/G2.  
 
15  
Table 4: Laminate material classification 
Material classification 
Supplier 
Material 
ID 
Glass transition 
temperature (T
g
) 
Curing 
agent 
Fillers Halogen-free 
I A No No 
I B 
DICY 
Yes Yes 
I C1 No No 
I C2 Yes No 
II D1 No No 
II D2 Yes No 
II E 
High T
g
  
(T
g
 > 165
o
C) 
Phenolic 
Yes Yes 
I F DICY Yes Yes 
I G1 No No 
I G2 Yes No 
II H Yes No 
II I 
Mid range T
g
 
(140
o
C<T
g
<165
o
C) Phenolic 
Yes Yes 
I J No No 
I K 
Low T
g
 
(T
g
 <140
o
C) 
DICY 
Yes No 
 
Electro dispersive spectroscopy (EDS) analysis  of the laminates (shown in Table 
5) revealed that the flame retardant used in all the halogen-free materials is a 
phosphorous based compound; whereas halogenated materials have brominated 
compounds. Also, the EDS analysis showed that the choice of type of fillers and 
flame retardants in the laminates varies from supplier to supplier. Halogenated 
laminates C2, G2 from supplier I and D2, H from supplier II contain silicon based 
fillers (probably SiO
2
). Halogen-free laminates B, F from supplier I contain silicon 
and aluminum based fillers (probably SiO
2
, Al(OH)
3
), whereas only laminate E from 
supplier II contain silicon and aluminum based fillers. Laminate I do not contain 
silicon based fillers. Environmental scanning electron microscopy (ESEM) analysis 
on the laminates with fillers (shown in Figure 9 for material B) revealed the 
microstructure of the filler particles.  
 
 
16  
Table 5: EDS analysis of laminates 
Material classification Inorganic elements detected 
Supplier 
Material 
ID 
Curing 
agent 
Fillers 
Halogen- 
free 
Br P Si Al 
I A DICY No No 9 - - - 
I B DICY Yes Yes - 9 9 9 
I C1 Phenolic No No 9 - - - 
I C2 Phenolic Yes No 9 - 9 - 
II D1 Phenolic No No 9 - - - 
II D2 Phenolic Yes No 9 - 9 - 
II E Phenolic Yes Yes - 9 9 9 
I F DICY Yes Yes - 9 9 9 
I G1 Phenolic No No 9 - - - 
I G2 Phenolic Yes No 9 - 9 - 
II H Phenolic Yes No 9 - 9 - 
II I Phenolic Yes Yes - 9 - 9 
I J DICY No No 9 - - - 
I K DICY Yes No 9 - - - 
 
 
Figure 9: ESEM picture showing filler particles in material B 
Filler 
particles 
 
17  
3.2 Material construction 
The logistics involved in obtaining the materials for the study is shown in Figure 
10. Laminates (cores) from supplier I and II with a nominal thickness of 1.2 mm and 
1 oz (0.036 mm) copper cladding were considered for evaluating most of the 
properties. 
Board fabricator
CALCE
CALCE
3X Reflow 6X Reflow 2X Reflow+1X-WaveControl
Cores and Prepregs
Cores
(Laminates)
Properties Measurement at CALCE
Supplier I Supplier II
Fabricated boards
 
 
Figure 10: Flow of test materials 
Laminates from supplier I have 6-plies of 7628 glass weave style with a resin 
content of 41%, and those from supplier II have 6-plies of 7629 glass weave style. 
Some of the property measurements require fabricated boards which were made at 
Gold circuit electronics, Taiwan. Fabricated boards consist of 12-layered stack up 
(nominal thickness of 2.5 mm, resin content: 53%) with alternative layers of cores 
and prepregs of 1080, 2116 and 7628 weave styles. Fabricated boards were available 
only for materials A, B, C1, C2, G1, J and they span the range of T
g
, curing agents, 
flame retardants and fillers. The 6-ply structure of the cores and the 12-layered 
structure of the fabricated boards are represented in Figure 11. 
 
18  
 
 
 
 
 
 
 
 
X
Z
Y
 
Prepreg: 2ply 1080 weave
Core: 2ply 2116 weave with  
1 oz copper clad on two sides
Prepreg: 2ply 2116 weave
Core: 2ply 7628 weave with  
1 oz copper clad on two sides
Prepreg: 2ply 2116 weave
Core: 2ply 2116 weave with  
1 oz copper clad on two sides
Prepreg: 2ply 1080 weave
Copper foil: 1 oz thick
Copper foil: 1 oz thick
0.034 mm
0.130 mm
0.268 mm
0.210 mm
0.418 mm
0.210 mm
0.268 mm
0.130 mm
0.034 mm
0.130 mm Prepreg: 2ply 1080 weave
Core: 2ply 2116 weave with  
1 oz copper clad on two sides
0.268 mm
Core: 2ply 2116 weave with  
1 oz copper clad on two sides
0.268 mm
Prepreg: 2ply 1080 weave0.130 mm
 
Figure 11: 6-ply laminate 
structure 
 
Figure 12: 12 layered fabricated board 
structure 
The boards were designed as per the guidelines listed in IPC-4121 document [21]. 
The designed boards have a symmetric structure with 7628 weave style core at the 
center with 2116 and 1080 weave style cores and prepregs spreading outwards. The 
presence of finer weave fabrics with higher resin content on the outer surface and 
coarse weave fabrics with lower resin content on the inner layers also ensures better 
chemical resistance and resistance to measling (condition in which internal glass 
fibers are separated from the resin at the weave intersection) [21]. The inner layers 
consist of staggered circular copper etch pattern and the outer layers of copper do not 
have any etch pattern. 
3.3 Exposure conditions 
Laminate samples from all the material types were divided into four lots. The first 
lot refers to control and represents the samples as-received from the laminate 
suppliers. The second lot was exposed to three reflow cycles (3X-R), the third lot to 
six reflow cycles (6X-R), and the fourth lot to a combination of two reflow cycles and 
 
19  
one wave soldering cycle (2X-R+1X-W). The lead-free reflow test profile (measured 
using thermocouples placed at different locations on the first test board) is shown in 
Figure 13 and meets the IPC/JEDEC-J-STD-020D [22] recommended lead-free 
reflow profiles. For the lead-free wave soldering exposures, samples were passed 
through two preheat zones maintained at 175
o
C and 199
o
C respectively, followed by 
a lead-free solder (Sn96.5/Ag3.0/Cu0.5) wave at 295
o
C. 
0
50
100
150
200
250
300
0 100 200 300 400 500 600
Time (seconds)
T
e
m
p
er
a
t
u
re (
d
e
g
r
ees
 C
)
Liquidus temperature: 217
o
C
Time above liquidus: 105 seconds
Peak temperature: 243
o
C
T
e
m
p
er
a
t
u
re (
d
e
g
r
ees
 C
)
 
Figure 13: Lead-free reflow profile used for the exposures 
 
20  
Chapter 4: Material Properties and Test Methods 
 
The key thermomechanical, physical, and chemical properties of the PCB laminates 
considered for evaluation and the corresponding test methods are listed in Table 6. 
Previous studies ([5], [6], [8], [10]) have identified the properties of T
g
, CTE, T
d
, and 
T-260 as some of the primary metrics in assessing the lead-free process compatibility 
of laminates. 
Table 6: Material properties and test methods 
Property Unit Test method Equipment 
Glass transition temperature (T
g
) o
C 
IPC-TM-650 
2.4.25 
Differential scanning calorimeter (DSC) 
Coefficient of thermal expansion 
(CTE) (out-of-plane, in-plane) 
ppm/
o
C 
IPC-TM-650 
2.4.24 
Thermo mechanical analyzer (TMA) 
Decomposition temperature (T
d
) o
C 
IPC-TM-650 
2.4.24.6 
Thermo gravimetric analyzer (TGA) 
Time-to-delamination (T-260) min 
IPC-TM-650 
2.4.24.1 
Thermo mechanical analyzer (TMA) 
Water absorption % 
IPC-TM-650 
2.6.2.1 
Micro-balance 
Flammability sec 
UL 94  
(V-0) 
Bunsen burner and gas supply  
4.1 Glass transition temperature (T
g
) 
Glass transition temperature (T
g
) of a resin system is the temperature at which 
material transforms from rigid and glass like state to a rubbery and compliant state 
due to the reversible breakage of Van der Waals bonds between the polymer 
molecular chains. The T
g
 of a FR-4 laminate system primarily depends on the type of 
the epoxy resin and its percentage composition. Material properties such as 
 
21  
coefficient of thermal expansion, Young?s modulus, heat capacity, and dielectric 
constant undergo a change around T
g
.  
The T
g
 of laminates was measured using Perkin-Elmer differential scanning 
calorimeter (Pyris 1 DSC) as per IPC-TM-650 2.4.25 test method [23]. The test 
equipment with the sample and reference containers is shown in  Figure 14. 
Two samples weighing between 15-30 mg were used for the measurements. Copper 
cladding was etched from the samples using sodium persulphate solution. Samples 
were then baked at 105
o
C for 2 hours and cooled to room temperature in a dessicator 
prior to the measurements. Samples were subjected to a temperature scan of 25?C to 
220?C at a ramp rate of 20
o
C/min, and T
g 
was identified as the midpoint of step 
transition in the DSC measurement plot. The representative glass transition 
temperature measurement plot is shown in     Figure 15.  
Sample 
container
Reference 
container
Display
 
 Figure 14: Perkin Elmer DSC 
 
 
 
 
   Sample and reference containers 
 
 
22  
27
28
29
30
31
32
50 70 90 110 130 150 170 190 210
Temperature (degrees C)
He
a
t
 f
l
o
w
 
(m
W
)
Glass transition temperature = 137.6
o
C
He
a
t
 f
l
o
w
 
(m
W
)
 
    Figure 15: Glass transition temperature measurement plot (material J) 
4.2 Coefficient of thermal expansion (CTE) 
Coefficient of thermal expansion (CTE) of a laminate is the fractional change of 
linear dimensions with temperature. Out-of-plane CTE influences failure mechanisms 
such as barrel cracking and delamination, whereas in-plane CTE in warp and fill 
directions affects shear failures of solder joints. 
CTE of laminate materials in out-of-plane and in-plane directions were measured 
using Perkin-Elmer thermomechanical analyzer (Pyris TMA 7) as per IPC-TM-650 
2.4.24 test method [24]. The test equipment used for the measurement is shown in 
Figure 16. 
Two samples of 7 mm x 7 mm size were used for the measurements. Copper 
cladding was etched and the samples were baked at 105
o
C for 2 hours followed by 
cooling to room temperature in a dessicator prior to the measurements. Samples were 
subjected to a temperature scan of 25
o
C to 250
o
C at 10
o
C/minute ramp rate for out-of-
plane CTE measurements, and 25
o
C to 220
o
C at 10
o
C/minute ramp rate for in-plane 
CTE measurements. The T
g
 was identified as the temperature at which slope of the 
 
23  
TMA measurement plot change, and CTE was measured below and above T
g
. The 
representative out-of-plane and in-plane CTE (warp/fill) measurement plots are 
shown in Figure 17 to Figure 19.  
 
Figure 16: Perkin Elmer TMA 
 
 
 
   
      Sample placement 
 
 
1.225
1.230
1.235
1.240
1.245
1.250
1.255
1.260
25 40 55 70 85 100 115 130 145 160 175 190 205 220 235 250
Temperature (degrees C)
Le
n
g
t
h
 (
m
m
)
CTE below T
g 
= 41.1 ppm/
o
C
CTE above T
g 
= 153.8 ppm/
o
C
Le
n
g
t
h
 (
m
m
)
 
  Figure 17: Out-of-plane CTE measurement plot (material B) 
Glass probe 
Furnace 
Sample 
 
24  
  
7.472
7.476
7.480
7.484
7.488
7.492
25 40 55 70 85 100 115 130 145 160 175 190 205 220
Temperature (degrees C)
Le
n
g
t
h
 (
m
m
)
CTE below T
g
= 12.8 ppm/
o
C
CTE above T
g
= 3.7 ppm/
o
C
Le
n
g
t
h
 (
m
m
)
 
  Figure 18: In-plane CTE (warp) measurement plot (material B) 
       
7.580
7.584
7.588
7.592
7.596
7.600
7.604
25 40 55 70 85 100 115 130 145 160 175 190 205 220
Temperature (degrees C)
Le
n
g
t
h
 (
m
m
)
CTE below T
g
= 14.5 ppm/
o
C
CTE above T
g
= 9.3 ppm/
o
C
Le
n
g
t
h
 (
m
m
)
 
Figure 19: In-plane CTE (fill) measurement plot (material B) 
The CTE of laminate materials in all the three directions (out-of-plane, warp and 
fill) is different. In-plane CTE (warp and fill) is less than out-of-Plane CTE as the 
former is dominated by the glass-fabric (typical CTE of E-glass: 5.4 ppm/
o
C [25]), 
whereas the latter is dominated by the epoxy resin (typical CTE of epoxy resin: 63 
ppm/
o
C [25]). Also, epoxy resin behaving like an incompressible fluid is primarily 
constrained along the in-plane directions by the glass fabric [26]. 
In-plane CTE above T
g
 is lower compared to that of below T
g
 because of the 
significant reduction of Young's modulus of resin above T
g
 (below T
g
 Young?s 
modulus of resin: 2.4 GPa, above T
g
 Young?s modulus of resin: 0.07 GPa) [26]. Also, 
 
25  
epoxy acting as incompressible liquid when expanding along out-of-plane direction 
shrinks in the warp and fill directions. As the expansion rate along out-of-plane 
significantly increases above T
g
, this may result in a reduction of above T
g
 in-plane 
CTE. 
Difference between above T
g
 in-plane CTE and below T
g
 in-plane CTE along warp 
direction is higher than that along fill direction [27]. Changes in the tension and 
spacing of the warp and fill fiber bundles during weaving result in varying degrees of 
flexibility and hence different properties along the two directions [14]. 
4.3 Decomposition temperature (T
d
) 
Decomposition temperature (T
d
) is the temperature at which a resin system 
irreversibly undergoes physical and chemical degradation with thermal destruction of 
cross-links, resulting in weight loss of the material.  
The T
d
 of laminates was measured using Shimadzu thermogravimetric analyzer 
(Shimadzu TGA 50) as per IPC-TM-650 2.4.24.6 test method [28]. The equipment 
used for the measurement is shown in       Figure 20. 
 
      Figure 20: Shimadzu TGA 50 
Furnace 
Weight balance 
mechanism 
 
26  
One test specimen from each material type weighing between 8-20 mg was used for 
the measurement. Copper cladding was etched and the samples were baked at 110
o
C 
for 24 hours followed by cooling to room temperature in a dessicator prior to the 
measurement. Sample was subjected to a temperature scan of 25
o
C to 400
o
C at a 
ramp rate of 10
o
C/minute in an inert nitrogen atmosphere (purged at 50 ml/minute). 
The change in weight of the sample was obtained as a function of temperature, and T
d
 
was recorded at 2% and 5% weight loss (compared to sample weight at 50
o
C as per 
the test method). The representative decomposition temperature measurement plot is 
shown in Figure 21.  
 
75
80
85
90
95
100
0 50 100 150 200 250 300 350 400
Temperature (degrees C)
% w
e
i
g
h
t
T
d
at 2% weight loss
T
d
at 5% weight loss
% w
e
i
g
h
t
 
Figure 21: Decomposition temperature measurement plot (material F) 
4.4 Time-to-delamination (T-260) 
Time-to-delamination is the time taken by a laminate material to delaminate 
(separation between layers of prepregs and copper clad cores in a multilayered 
structure), when exposed to a constant temperature. Due to lead-free soldering 
process reaching peak temperatures of around 260
o
C, time-to-delamination measured 
at 260
o
C (T-260) is being used as a metric for assessing lead-free process 
 
27  
compatibility of laminates. It is a measure of the ability of a laminate material to 
withstand multiple soldering cycles without delamination. 
The T-260 of fabricated boards was measured using Perkin-Elmer 
thermomechanical analyzer (Pyris TMA 7) as per IPC-TM-650 2.4.24.1 test method 
[29]. Two samples of 7 mm x 7 mm size were used for the measurements. Samples 
were baked at 105
o
C for 2 hours and then cooled to room temperature in a dessicator 
prior to the measurements. Samples were then subjected to a temperature scan of 
25
o
C to 260
o
C at 10
o
C/minute ramp rate and held at 260
o
C until an irreversible 
change in thickness (assumed to be caused by layer to layer delamination) of the 
sample was observed. The test was terminated at 60 minutes for the laminate 
materials that did not delaminate until then. Time-to-delamination was determined as 
the time between onset of isotherm (260
o
C) and the onset of delamination. The 
representative T-260 measurement plot is shown in Figure 22.  
2.28
2.30
2.32
2.34
2.36
2.38
2.40
2.42
0 5 10 15 20 25 30 35
Time (minutes)
T
h
i
c
kne
s
s
 (
m
m
)
3.6 minutes
Onset of 
isotherm 
(260
o
C)
Onset of 
delamination
T
h
i
c
kne
s
s
 (
m
m
)
 
Figure 22: T-260 measurement plot (material B) 
 
 
28  
4.5 Water absorption 
Water absorption is a measure of the amount of water absorbed by laminate 
materials when immersed in distilled water for 24 hours at room temperature. Water 
absorption of the laminates was measured as per IPC-TM-650 2.6.2.1 test method 
[30].  
Three samples of 50 mm x 50 mm size were used for the measurements. Copper 
cladding was etched and the samples were baked at 105
o
C for 1 hour followed by 
cooling to room temperature in a dessicator prior to the measurements. Samples were 
weighed using microbalance before and after immersion in distilled water for 24 
hours and percentage water absorption was calculated by the following expression:. 
 
4.6 Flammability 
Flammability is a measure of the material?s tendency to extinguish the flame once 
the specimen has been ignited and separated from the flame source. FR-4 PCB 
laminates, being plastics, are inherently flammable and flame retardants are added to 
the resin system in order to enhance their self-extinguishing property. Typical steps 
involved in a PCB combustion process are shown in Figure 23.  
The flammability of laminates was measured as per UL 94 test method [32]. Five 
specimens per material type were cut into sizes of 127 mm x 12.7 mm. Copper 
cladding was etched and the samples were conditioned at 25
o
C, 50% relative 
humidity for 48 hours prior to testing.  
100
 weightdConditione
 weightdConditioneWet weight
absorptionWater % ?
?
=
 
29  
Transfer of thermal energy to the laminate
Pyrolysis 
(Thermal degradation of solid matrix, in 
the absence of oxygen, to release 
combustible gases)
100
o
C to 250
o
C
Functional group eliminations from chain 
ends (e.g., water molecules); PWB 
discoloration
250
o
C to 500
o
C
Scission of high energy bonds leading to 
production of flammable monomers
Formation of flammable mixture 
(combustible gases + oxygen) 
Ignition occurs if the flammable mixture is 
heated above its ignition temperature (heat is 
transferred from the laminate to the mixture 
through radiation/convection)  
Flame propagates as new fuel becomes available 
and as degradation is accelerated by the added 
thermal energy from the flame
When the energy source is removed, 
flame alone doesn?t supply sufficient 
heat to maintain thermal Pyrolysis
Flame retardants act to extinguish the flame
Gas-phase combustion consists of free 
radical chain reactions involving 
oxidation of hydrocarbons
Non-pyrolyzed hydrocarbons:
liberated as smoke, and condense as soot 
beyond flame front
Non-volatalized hydrocarbons:
These are high molecular weight polymer 
fragments and remain as char.
Above 500
o
C
Aromatic, condensed ring systems, either 
intrinsic to the polymer or formed as a 
result of recombination of fragments, will 
decompose
 
Figure 23: Typical PCB combustion steps [31] 
 
Samples were ignited using a bunsen burner by a 19 mm high blue flame (as shown 
in Figure 24). Gas supply to the burner was adjusted to maintain the height. Flame 
was placed below the sample for 10 seconds and withdrawn. If active combustion 
ceases prior to specimen being completely consumed, samples were placed again for 
10 seconds and withdrawn. Following parameters were noted: duration of the 
specimen burning after the first flame application, duration of the specimen burning 
after the second flame application, and if any specimen burns up the holding clamp 
on any ignition. Total specimen burning time was then calculated by summing the 
burning times in both flame applications. 
 
30  
19 mm
9.5 mm
Laminate
Bunsen 
burner
127 mm
12.7 mm
Clamp
6.4 mm
 
Figure 24: Test set up for flammability measurements 
The test results were compared against UL 94 V-0 flammability criteria, according 
to which, the time of burning per specimen after either flame application should not 
be greater than 10 seconds, and total burning time for the 10 flame applications for 
each set of 5 specimens should not be greater than 50 seconds. 
 
31  
Chapter 5: Results and Discussion 
In all the properties considered, pre-exposure measurement results were used to 
characterize the laminate materials. The laminate material data sheet values are listed 
in Appendix I. In the following subsections, the effects of laminate material 
constituents (such as curing agent, fillers, and flame retardant) in relationship to 
changes in the material properties due to lead-free soldering exposures are discussed. 
5.1 Glass transition temperature (T
g
) 
The pre and post-exposure T
g
 measurement results were grouped by material (e.g., 
A, B, C1) as shown in Figure 25. Under each material group the results of four sets of 
data are represented. The first set of data points shown in each group corresponds to 
the results of control samples, the second set to 3X reflowed samples, the third set to 
6X reflowed samples, and the fourth set to the samples that were exposed to a 
combination of 2X reflow and 1X wave soldering cycles. This pattern of data 
representation is followed in all the properties except time-to-delamination and 
flammability. The materials that underwent a mean-to-mean variation of greater than 
5
o
C in T
g
 after the exposures are highlighted in Figure 25. 
The control sample results show to have a similar T
g 
range for the laminates, 
irrespective of the type of curing agent, the type of flame retardant, and the presence 
of fillers. For example, materials A to E have a T
g 
range of 165
o
C to 180
o
C even 
though they differ by the material constituents. The T
g
 of a laminate system primarily 
depends on the type of epoxy (bi, tetra or multi-functional) and its percentage 
 
32  
composition. Higher cross-linking density in the multi-functional epoxy systems 
compared to their bi-functional counterparts results in a higher T
g
. 
130
135
140
145
150
155
160
165
170
175
180
185
G
l
as
s
 t
r
an
s
i
t
i
on
 t
e
m
p
erat
u
re
 (
d
e
g
ree
s
 
C
)
AC1D1D2C2BG1G2HIJK
High T
g
Mid T
g Low T
g
FE
G
l
as
s
 t
r
an
s
i
t
i
on
 t
e
m
p
erat
u
re
 (
d
e
g
ree
s
 
C
)
 
Figure 25:  Effect of lead-free soldering exposures on T
g 
of the laminates 
 
The post-exposure results show that five out of seven high T
g 
and four out of five 
mid T
g 
materials have relatively stable T
g 
with a variation of less than 5
o
C. All of 
these are phenolic-cured materials. The materials (A, B, F, J and K) that underwent a 
mean-to-mean variation in T
g 
of greater than 5
o
C are all DICY-cured. A decrease in 
T
g 
was observed in high and mid T
g 
DICY-cured materials, whereas an increase in T
g
 
was observed in low T
g 
DICY-cured materials. The highest variation in T
g
 (a 
reduction of about 20
o
C from control) was observed in the high T
g
 DICY-cured 
halogen-free material (B). 
The T
g
 of laminates was also obtained from the coefficient of thermal expansion 
measurement plots by TMA. The T
g 
comparison between DSC and TMA is shown in 
Table 7. The difference in T
g
 values between DSC and TMA methods is due to the 
difference in measurement method. In DSC, the midpoint of transition in heat 
 
33  
capacity is taken as T
g
, whereas in TMA the point at which slopes of CTE plot 
change is taken as T
g
. 
Table 7: The T
g
 comparison between DSC and TMA methods 
Control 3X Reflow 6X Reflow 2X-R+1X-W 
# Datasheet 
DSC TMA DSC TMA DSC TMA DSC TMA 
A 
165-175 
(DSC) 
170.1 172.0 166.3 171.8 163.2 163.7 167.5 159.8 
B 
165-175 
(DSC) 
170.3 160.0 159.7 158.0 154.9 154.9 151.1 151.3 
C1 
165-175 
(DSC) 
170.1 167.6 168.0 163.8 166.6 163.9 170.4 165.8 
C2 
165-175 
(DSC) 
168.5 167.1 163.8 166.3 164.2 168.5 164.7 161.1 
D1 
173-183 
(DSC) 
174.4 152.9 176.6 162.0 178.9 163.8 178.8 165.3 
D2 
170-175 
(DSC) 
171.1 163.8 166.9 161.8 169.5 163.9 167.8 158.6 
E 
170-180 
(TMA) 
172.2 170.3 172.8 167.4 175.4 165.7 174.1 165.4 
F 
145-155 
(DSC) 
152.2 157.1 152.8 151.5 143.3 147.7 137.9 145.4 
G1 
150-160 
(DSC) 
150.3 156.3 149.0 149.8 152.1 152.1 147.7 153.0 
G2 
145-155 
(DSC) 
147.4 156.0 148.6 153.4 147.8 151.7 146.9 148.7 
H 
135-145 
(DSC) 
148.1 149.2 151.7 146.9 145.7 148.6 149.2 146.8 
I 
140-150 
(TMA) 
152.0 138.1 153.2 138.1 150.8 134.5 152.4 137.2 
J 
135-145 
(TMA) 
135.4 137.2 138.0 144.2 140.3 141.2 140.9 137.0 
K 
175-185 
(DMA) 
134.2 136.2 140.2 139.5 138.0 139.8 139.2 136.7 
  
5.2 Coefficient of thermal expansion (CTE) 
The out-of-plane CTE (below and above T
g
) measurement results are shown in 
Figure 26. The materials that underwent a mean-to-mean variation of greater than 
15% in (above T
g
) CTE after the exposures are highlighted in the figure. 
 
34  
20
40
60
80
100
120
140
160
180
200
220
240
260
C
T
E
 (
p
p
m
/
d
eg
ree
s
 C
)
Above Tg
Below Tg
AC1D1D2C2BG1G2HIJK
High T
g
Mid T
g
Low T
g
FE
C
T
E
 (
p
p
m
/
d
eg
ree
s
 C
)
 
Figure 26: Effect of lead-free soldering exposures on out-of-plane CTE of the 
laminates 
 
The control results show that high T
g 
material (A) has lower CTE compared to a 
low T
g
 material (J) with similar constituents (DICY-cured with halogenated flame 
retardant). This is because of the higher cross-linking density in the epoxy resin 
system of high T
g 
materials that resists the thermal expansion of the laminate. DICY 
and phenolic-cured materials have similar CTE range (A vs. C1). Filled materials 
have lower CTE values compared to unfilled materials (C2 vs. C1, D2 vs. D1, G2 vs. 
G1), as fillers replace the epoxy in filled materials. Halogen-free materials have lower 
CTE values than halogenated materials (B vs. A, I vs. H). This could be because of 
the combined effect of presence of fillers and material chemistry of halogen-free 
laminates that typically result in lower CTE values than the halogenated laminates 
[10]. 
The post-exposure results show a reduction of out-of-plane CTE in most of the 
materials, with a highest reduction of approximately 25% observed in the above T
g
 
 
35  
CTE of material B. The reduction in CTE could be due to further curing of the resin 
system, resulting in an increase in the cross-linking density. 
The control in-plane (warp and fill) CTE measurements were performed on all the 
materials and the results are shown in Table 8. The post-exposure measurements were 
performed on six materials A, B, C1, F, G2 and J (shown in Table 9 and Table 10) as 
similar CTE values were observed in all the materials irrespective of the type of 
curing agent, flame retardant and presence of fillers. Post-exposure in-plane CTE 
results did not show noticeable variation. 
Table 8: Control in-plane (warp and fill) CTE measurement results 
Material classification Warp Fill 
Supplier 
Material 
ID 
Curing 
agent 
Fillers 
Halogen- 
free 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
I A DICY No No 11.8?0.2 1.1?0.1 15.4?0.5 7.9?1.4 
I B DICY Yes Yes 12.2?0.7 2.7?1.0 14.8?0.3 9.8?0.5 
I C1 Phenolic No No 13.4?1.5 4.3?1.1 16.3?0.5 12.2?0.7 
I C2 Phenolic Yes No 12.6?0.5 3.9?0.7 14.6?0.5 11.6?0.5 
II D1 Phenolic No No 14.7?2.3 3.3?0.6 16.9?1.5 11.3?1.0 
II D2 Phenolic Yes No 13.4?0.9 3.9?1.1 18.9?1.4 13.4?0.8 
II E Phenolic Yes Yes 15.2?1.5 3.2?0.6 16.2?0.8 9.5?0.5 
I F DICY Yes Yes 12.1?0.8 4.9?1.0 16.8?1.9 10.6?0.5 
I G1 Phenolic No No 13.1?0.7 2.0?0.2 15.2?0.8 10.4?1.2 
I G2 Phenolic Yes No 15.2?2.7 3.5?2.5 14.6?0.4 10.9?0.8 
II H Phenolic Yes No 14.1?0.8 4.2?0.2 18.8?1.2 13.5?2.4 
II I Phenolic Yes Yes 17.9?1.2 3.1?0.9 17.8?1.2 7.9?0.3 
I J DICY No No 12.1?0.5 2.2?1.1 14.9?0.5 10.2?1.8 
I K DICY Yes No 13.0?1.1 2.8?0.4 15.7?0.3 10.8?0.1 
 
Table 9: Post exposure in-plane (warp) CTE measurement results 
Control 3X Reflow 6X Reflow 2X R+ 1X wave 
Material 
ID 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
A 11.8?0.2 1.1?0.1 14.1?0.1 1.4?0.2 13.7?1.5 0.9?0.1 15.3?1.3 1.8?0.5 
B 12.2?0.7 2.7?1.0 13.5?1.1 3.4?0.8 12.9?3.1 2.1?0.6 14.8?2.7 3.2?1.5 
C1 13.4?1.5 4.3?1.1 15.7?1.1 3.5?0.9 13.6?2.7 4.5?1.7 15.2?1.9 3.8?0.9 
F 12.1?0.8 4.9?1.0 15.9?1.8 2.2?1.9 13.6?2.4 1.8?0.7 11.2?1.5 2.4?0.5 
G2 15.2?2.7 3.5?2.5 13.3?1.3 4.2?1.1 16.3?2.1 3.8?0.9 12.3?0.8 2.3?1.1 
J 12.1?0.5 2.2?1.1 14.9?2.3 4.9?2.8 13.6?1.7 2.6?1.3 12.7?1.9 4.5?2.1 
 
36  
Table 10: Post exposure in-plane (fill) CTE measurement results 
Control 3X Reflow 6X Reflow 2X R+ 1X wave 
Material 
ID 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
A 15.4?0.5 7.9?1.4 14.2?1.1 9.1?0.7 16.3?1.6 8.7?0.7 15.9?1.1 8.6?0.4 
B 14.8?0.3 9.8?0.5 16.8?1.2 8.6?1.3 15.1?0.9 9.1?1.4 15.2?0.4 8.4?0.9 
C1 16.3?0.5 12.2?0.7 15.7?0.2 11.8?0.2 15.1?1.6 10.6?0.9 15.3?1.2 11.1?1.0 
F 16.8?1.9 10.6?0.5 15.1?0.8 9.8?0.3 15.9?1.1 9.1?1.5 16.7?0.7 11.6?0.2 
G2 14.6?0.4 10.9?0.8 15.7?1.5 10.1?1.2 13.1?0.9 8.8?0.8 15.1?1.0 9.9?1.1 
J 14.9?0.5 10.2?1.8 13.1?1.2 9.1?0.7 14.1?1.9 10.1?1.5 16.8?1.4 11.2?0.5 
 
5.3 Decomposition temperature (T
d
) 
The decomposition temperature measurement results corresponding to 2% and 5% 
weight loss are plotted in Figure 27. Lead-free soldering exposures did not show 
noticeable variation in T
d
 (>10
o
C), and hence no materials are highlighted in the 
figure. 
310
320
330
340
350
360
370
380
390
D
e
com
p
o
s
i
t
i
o
n
 
t
e
m
p
er
at
u
r
e
 (
d
e
g
re
es
 C
)
Td (2% weight loss)
Td (5% weight loss)
A
C1 D1 D2C2
BG1G2HIJK
High T
g
Mid T
g
Low T
g
F
E
D
e
com
p
o
s
i
t
i
o
n
 
t
e
m
p
er
at
u
r
e
 (
d
e
g
re
es
 C
)
 
Figure 27: Effect of lead-free soldering exposures on T
d
 of the laminates 
 
The control measurement results show that low T
g 
material (J) has higher T
d 
compared to high T
g 
material (A) with similar constituents [8]. The comparison of 
degradation behavior between materials A and J is plotted in Figure 28.  
 
37  
70
75
80
85
90
95
100
105
110
25 75 125 175 225 275 325 375
Temperature (degrees C)
% We
i
g
ht
High Tg (A)
Low Tg (J)
Low T
g 
(J)
High T
g 
(A)% We
i
g
ht
% We
i
g
ht
 
Figure 28: T
d
 comparison between materials A and J 
 Amongst the halogenated materials, phenolic-cured materials (G1, C1) have higher 
T
d
 compared to their DICY-cured counterparts (A, J). The lower decomposition 
temperatures in DICY-cured epoxy systems could be attributed to the presence of 
linear aliphatic molecular bonds with amine linkages, compared to more thermally 
stable aromatic bonds with ether linkages of phenolic-cured systems ([10], [15]) 
These observations are shown in Figure 29 and Figure 30. 
70
75
80
85
90
95
100
105
110
25 75 125 175 225 275 325 375
Temperature (degrees C)
% W
e
i
g
ht
DICY cured (A)
Phenolic cured (G1)
DICY cured (A)
Phenolic cured (G1)
% W
e
i
g
ht
% W
e
i
g
ht
 
Figure 29: T
d
 comparison between materials A and G1 
 
38  
High Tg
High Tg
Low Tg
Low Tg
High Tg
Mid Tg
Mid Tg
High Tg
290
300
310
320
330
340
350
360
370
380
D
eco
m
p
o
s
i
t
i
o
n
 
t
e
m
p
era
t
u
re (
d
eg
re
es
 C
)
 
J
A
G1 C1
J
A
G1 C1
2% weight loss 5% weight loss
Dicy cured
Phenolic cured
D
eco
m
p
o
s
i
t
i
o
n
 
t
e
m
p
era
t
u
re (
d
eg
re
es
 C
)
 
 
Figure 30: Effect of type of curing agent on T
d
 of the laminates 
Laminates with fillers have lower T
d 
compared to their counterparts without fillers 
(C2 vs. C1, D2 vs. D1, G2 vs. G1). Inorganic fillers such as alumina or silica 
accelerate the thermal decomposition process by lowering the activation energy 
required for decomposition, thereby acting as catalysts ([33], [34], [35]). These 
observations are shown in Figure 31 and Figure 32. 
70
75
80
85
90
95
100
105
110
25 75 125 175 225 275 325 375
Temperature (degrees C)
% W
e
i
g
h
t
No fillers (C1)
Fillers present (C2)
Fillers present (C2)
No fillers (C1)
% W
e
i
g
h
t
% W
e
i
g
h
t
 
Figure 31: T
d
 comparison between materials C1 and C2 
  
 
 
39  
No fillers
Fillers present
310
320
330
340
350
360
370
D
eco
m
p
o
s
i
t
i
o
n
 t
e
m
p
e
r
at
u
re (
d
e
g
re
es 
C
)
High T
g
Midrange T
g
Low T
g
C1
C2
G1
G2
J
K
D
eco
m
p
o
s
i
t
i
o
n
 t
e
m
p
e
r
at
u
re (
d
e
g
re
es 
C
)
 
  Figure 32: Effect of presence of fillers on T
d
 of the laminates 
Halogen-free material that is DICY-cured (B) could withstand higher temperatures 
before 2% and 5% weight degradation compared to the halogenated DICY-cured 
material (A) (shown in Figure 33). On the contrary, halogen-free material that is 
phenolic-cured (I) has lower T
d
 compared to its halogenated counterpart (H) (shown 
in Figure 34). Irrespective of the type of curing agent, halogenated resin systems (A, 
H) underwent degradation from 2% to 5% within a narrow temperature range, which 
was not observed in halogen-free systems (B, I) (shown in Figure 34). 
70
75
80
85
90
95
100
105
110
25 75 125 175 225 275 325 375
Temperature (degrees C)
% W
e
i
g
h
t
DICY cured halogenated (A)
DICY cured halogen-free (B)
DICY cured halogenated (A)
DICY cured halogen-free (B)% W
e
i
g
h
t
 
Figure 33: T
d
 comparison between materials A and B 
 
40  
70
75
80
85
90
95
100
105
110
25 75 125 175 225 275 325 375
Temperature (degrees C)
% W
e
i
g
h
t
Phenolic cured halogenated (H)
Phenolic cured halogen-free (I)
Phenolic cured halogenated (H)
Phenolic cured halogen-free (I)
% W
e
i
g
h
t
% W
e
i
g
h
t
 
Figure 34: T
d
 comparison between materials H and I 
A
Dicy
A
Dicy
H
PhenolicH
Phenolic
B
Dicy
B
Dicy
I
Phenolic
I 
Phenolic
290
300
310
320
330
340
350
360
370
380
D
e
co
m
p
o
s
i
t
i
o
n
 t
e
m
p
era
t
u
re (
d
egrees
 C
)
Halogen-free
Halogenated
2% weight loss 5% weight loss
D
e
co
m
p
o
s
i
t
i
o
n
 t
e
m
p
era
t
u
re (
d
egrees
 C
)
 
Figure 35: Effect of type of flame retardant on T
d
 of the laminates 
The post-exposure results show a maximum variation of 7
o
C in decomposition 
temperature of the laminates. The effect of material constituents such as curing 
agents, fillers, and flame retardants on the decomposition temperatures for the control 
samples remained the same after the exposures. 
5.4 Time-to-delamination (T-260) 
Time-to-delamination was measured for fabricated boards on a subset of materials 
i.e., A, B, C1, C2, G1, and J and the results are shown in Figure 36. The order of data 
representation under each material is the same as that of T
g
 results. 
 
41  
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
T
i
me
-to
-d
e
l
a
m
i
n
a
t
i
o
n
 (mi
n
u
t
e
s
)
Control
3X-R
6X-R
2X-R+1X-W
AC1C2BG1J
High T
g
Mid T
g
Low T
g
Time-to-delamination longer than 60 minutes
T
i
me
-to
-d
e
l
a
m
i
n
a
t
i
o
n
 (mi
n
u
t
e
s
)
 
Figure 36: Effect of lead-free soldering exposures on T-260 of the laminates 
The control results show that three materials A, B and J delaminated within 20 
minutes whereas materials C1, C2 and G1 did not delaminate for 60 minutes. 
Material J has higher T-260 than that of A and B. Thus, DICY-cured materials have 
lower time-to-delamination compared to phenolic-cured materials irrespective of their 
T
g
. Low T
g 
DICY-cured materials have higher T-260 compared to their high T
g 
counterparts. The effect of type of flame retardant and presence of fillers is not as 
prominent as that of curing agent. 
Lead-free soldering exposures tend to lower the time-to-delamination of materials 
A, B, and J all of which are DICY-cured materials. Materials C1, C2 and G1 which 
are phenolic-cured did not delaminate below 60 minutes even after exposures. 
5.5 Water absorption 
The pre and post-exposure water absorption measurement results are shown in 
Figure 37. Average and the range of results from the average of three measurements 
are plotted. Materials with a variation of greater than 25% in water absorption from 
control are shown within separate boxes in Figure 37. 
 
42  
0.020
0.040
0.060
0.080
0.100
0.120
0.140
0.160
W
a
t
er a
b
s
o
r
p
t
i
on
 
(
%
)
A C1 D1 D2C2BG1G2HIJK
High T
g
Mid T
g
Low T
g
FE
W
a
t
er a
b
s
o
r
p
t
i
on
 
(
%
)
 
Figure 37: Effect of lead-free soldering exposures on water absorption 
The control sample results show that high T
g
 materials (A, B) have higher water 
absorption compared to low T
g
 materials (J, K) with similar constituents. This could 
be attributed to the availability of higher free volume due to higher cross-linking 
density in the high T
g 
materials compared to low T
g 
materials [36]. DICY-cured 
epoxy systems are more hygroscopic compared to phenolic-cured systems (A vs. C1, 
F vs. I) because of the presence of highly polar bonds in the former compared to the 
later [15]. The effect of presence of fillers and type of flame retardant is not as 
prominent as that of type of curing agent. 
The post-exposure results showed an increase in water absorption due to lead-free 
soldering exposures for most of the materials. The material type with highest increase 
in water absorption value (55%) after 6X reflow exposure is a low T
g
 DICY-cured 
halogenated material (J). 
5.6 Flammability 
The pre and post-exposure flammability measurement results are shown in Figure 
38 and Figure 39. Total burning time for the 10 flame applications for each set of 5 
 
43  
specimens is plotted in Figure 38, whereas average burning time and one standard 
deviation is plotted in Figure 39. The UL 94 V-0 flammability criteria i.e., total 
burning time for the 10 flame applications for each set of 5 specimens should not be 
greater than 50 seconds, is shown in Figure 38.  
0
10
20
30
40
50
60
70
80
T
o
ta
l
 b
u
r
n
in
g
 ti
me
 (s
e
c
o
n
d
s
)
A C1 D1 D2C2BG1G2HIJK
High T
g
Mid T
g
Low T
g
FE
UL 94 V-0 criteria
T
o
ta
l
 b
u
r
n
in
g
 ti
me
 (s
e
c
o
n
d
s
)
 
Figure 38: Effect of lead-free soldering exposures on flammability of the 
laminates (total burning time) 
0.00
5.00
10.00
15.00
20.00
A
v
e
r
age
 bur
ni
ng t
i
m
e
 (
s
e
c
o
nds
)
A C1 D1 D2C2BG1G2HIJK
High T
g
Mid T
g
Low T
g
F
E
A
v
e
r
age
 bur
ni
ng t
i
m
e
 (
s
e
c
o
nds
)
 
Figure 39: Effect of lead-free soldering exposures on flammability of the 
laminates (Average burning time) 
The control sample results show that UL 94 V-0 flammability criteria are satisfied 
by all the samples. The effect of material constituents on the flammability of the 
laminates is more prominent in Figure 38 compared to Figure 39. Low T
g
 material (J) 
 
44  
has higher flammability compared to high T
g
 material (A) with similar constituents. 
DICY-cured materials (A, B, F, J, K) have higher burning times compared to 
phenolic-cured systems irrespective of T
g
, type of flame retardant and presence of 
fillers. This could be because of the highly aromatic structure of phenolic-cured 
systems that can produce a substantial amount of char which forms a boundary layer 
between the flame front and the combustible material reducing the flame propagation 
[31]. Also, the flammable mixture of phenolic-cured systems could still contain 
higher energy bonds compared to that of DICY-cured systems. The effects of 
presence of fillers and type of flame retardant are not as prominent as that of the 
curing agent. 
The post-exposure measurement results show that samples still meet UL 94 V-0 
flammability criteria after lead-free soldering exposures with a common trend of 
increase in combustion times observed in most of the materials. 
 
45  
Chapter 6:  Analysis of Results 
 
The high temperature exposures associated with lead-free soldering resulted in a 
noticeable variation in the laminate material properties. The exposures could have 
possibly resulted in the cleavage of bonds at the chain ends of the epoxy system 
leading to an ?enhanced-cure? structure; or resulted in the cleavage of bonds in 
polymer backbone leading to a ?degraded? structure. The former mechanism results in 
an increase in the cross-linking density of the epoxy matrix whereas the later results 
in a change in the material structure due to loss of certain functional groups [37], both 
of which can potentially change the material properties of laminates. Fourier 
transform infrared spectroscopy (FTIR) analysis was performed to verify the possible 
degradation in the epoxy structure, and combinatorial analysis of material properties 
is performed to verify the enhanced-cure structure. 
6.1 Fourier transform infrared (FTIR) spectroscopy analysis 
6.1.1 Introduction 
Fourier transform infrared (FTIR) spectroscopy is a tool that provides structural 
information about the presence of certain functional groups in a material. 
Electromagnetic radiation in the infrared (IR) region is directed at the sample which 
absorbs (or transmits) the infrared radiation at different frequencies and produces a 
unique spectra based on the frequencies at which it absorbs (or transmits) IR, and the 
intensity of the absorption (or transmission). The absorption (or transmission) is 
shown as a peak in the FTIR spectrum. The wave number corresponding to the peak 
 
46  
is an indication of the energy absorbed (or transmitted). Each molecule has its own 
distinct quantized vibrational energy level, and hence the wave number which are 
related by the following expression. 
 ?E
vib
 = hc ? , h-Planck?s constant; c- speed of light, ? - wave number 
Each wave number depends on several factors such as mass of the atoms, force 
constants of the bonds, and geometry of the molecules. The absorbed energy of 
infrared radiation is converted into atomic bond vibrations. Different types of 
molecular bending movements are shown in Figure 40. 
 
Figure 40: Common molecular bending movements [38] 
The atomic vibrations are classified as either stretching or bending. Stretching 
vibrations are categorized as being either symmetrical (movement in the same 
direction) or asymmetrical (movement in opposite directions). Bending vibrations are 
classified as scissoring, rocking, twisting, or wagging. 
 
47  
6.1.2 Analysis 
FTIR analysis is performed on the laminates to identify the key functional groups 
in the material systems and investigate their possible changes due to lead-free 
soldering exposures, and correlate the compositional changes to the variation in 
material properties. Samples from control and 6X reflowed lots of materials B and H 
that underwent significant variation in the properties were selected for the analysis. 
Mattson FTIR spectrometer with Galaxy Microscope attachment was used for the 
analysis. A small piece of the material was taken from the sample using a stainless 
steel scalpel and mounted on an IR-reflective brass plate. IR beam was passed 
through the sample and then reflected off of the brass plate and passed back through 
the sample where it was detected and analyzed. Each sample was scanned 32 times 
with 4 cm
-1
 resolution to obtain the spectrum. 
FTIR library was prepared from the literature with typical wave numbers 
corresponding to the functional groups present in epoxy systems and is shown in 
Table 11. Due to the nature of various types of atomic vibrations, a functional group 
appears with different wave numbers in the library. From the knowledge of basic 
chemical structures present in the epoxy systems, the functional groups present in the 
laminate materials were identified. The wave numbers corresponding to the 
functional groups are listed in Figure 41 based on the prepared FTIR library. 
 
 
 
 
48  
Table 11: Typical functional groups and wave numbers present in epoxy systems 
Absorption 
(cm
-1
) 
Functional 
groups 
Origin Reference 
3500-3400 -OH or -NH
2
 
-OH from opening of epoxy ring and -NH
2
 from 
DICY curing agent 
[39] 
3350-3200 -NH
2
 and -NH 
From BAP, TAM (phosphorus containing-amine 
curing agents) cured epoxy systems 
[17] 
3410 phenyl-OH 
From DHPDOPO (phosphorous based flame 
retardant) 
[40]  
3323 phenyl-OH From BPHPPO (phosphorous based flame retardant) [41] 
2900 -CH From epoxy resin [42]  
2980-2850 -CH 
From WSR-HPPE (epoxy resin with phosphorus 
flame retardant) 
[43]  
2384 P-H stretch 
From DOPO (monomer for DHPDOPO-phosphorous 
based flame retardant) 
[40]  
2180 -C?N (Nitrile) From DICY-cured epoxy [44]  
1740 -C=O (Carbonyl) From DICY-cured epoxy [44]  
1728, 1721 -C=O 
From organo phosphate esters, (phosphorus 
containing copoly esters (PET-co-PEPPs)) 
[18]  
1723 -C=O 
From WSR-HPPE (epoxy resin with phosphorus 
based flame retardant) 
[43]  
1660 
-C=N  
(Imine stretch) 
From DICY-cured epoxy [44] 
1650 -C=N  From DICY-cured epoxy [44] 
1640 -C=O 
From BAP, TAM (phosphorus containing-amine 
curing agents) cured epoxy systems 
[17] 
1630, 1565 
-N-H  
(Amine bend) 
From DICY-cured epoxy [44] 
1604, 1492 Aromatic C-C 
From BPHPPO-containing epoxy resin (phosphorus 
based flame retardant system) 
[41] 
1594 Aromatic C-C From DDS curing agent [39] 
1591 phenyl-P 
From DHPDOPO (phosphorous based flame 
retardant) 
[40] 
1591, 1298, 
715 
phenyl-NH
2
 From epoxy resin [42]
1509 Aromatic C-C From DGEBA (unbrominated epoxy resin) [39] 
1506 Aromatic C-C From epoxy resin [42] 
1479 Aromatic C-C 
From BPHPPO (phosphorus containing flame 
retardant) 
[41] 
1466 Aromatic C-C From DGEBTBA (brominated epoxy) [39] 
1461, 1350 phenyl-P 
From BGPPO (phenyl phosphine) resin with DDS 
curing agent 
[45] 
1457 phenyl-P 
From DGEBA with DHPDOPO and BPA 
(phosphorus containing epoxy system) 
[40] 
1363 -CH
3
 
From DGEBA with DHPDOPO and BPA 
(phosphorus containing epoxy system) 
[40] 
1296 P=O 
From WSR-HPPE (epoxy resin with phosphorus 
flame retardant) 
[43] 
 
49  
Absorption 
(cm
-1
) 
Functional 
groups 
Origin Reference 
1289 P=O 
From BPHPPO-containing epoxy resin  
(phosphorus based flame retardant epoxy system) 
[41] 
1283 P=O 
From BGPPO (phenyl phosphine) resin with DDS 
curing agent 
[45] 
1276 P=O 
From BPHPPO  
(phosphorus based flame retardant) 
[41] 
1276 S=O From DDS curing agent [39] 
1272 phenly-O From DGEBTBA (brominated epoxy) [39] 
1263, 995 phenyl-C-P 
From BPHPPO-containing epoxy resin  
(phosphorus based flame retardant epoxy system) 
[41] 
1249, 1228 P=O 
From organo phosphate esters (phosphorus 
containing copoly esters (PET-co-PEPPs)) 
[18] 
1246 phenyl-O From DGEBA [39] 
1246 phenyl-O 
From DGEBA with DHPDOPO and BPA 
(phoshporus containing epoxy system) 
[40] 
1234, 979 
phenyl-O-P 
stretch 
From BPHPPO  
(phosphorus based flame retardant) 
[41] 
1236, 817 -C-O-C- From epoxy resin [42] 
1195 P=O From DHPDOPO flame retardant [40] 
1184 P=O 
From DGEBA with DHPDOPO and BPA 
(phosphorus containing epoxy system) 
[40] 
1176, 1131 P-O-C 
From organo phosphate esters (phosphorus 
containing copoly esters (PET-co-PEPPs)) 
[18] 
1066 phenyl-Br From DGEBTBA (brominated epoxy) [39] 
1036 phenyl-O-C From DGEBA (unbrominated epoxy) [39] 
1034 P-O-C 
From WSR-HPPE  
(phosphorus based flame retardant epoxy system) 
[43] 
1001 phenyl-O-C DGEBTBA (brominated epoxy) [39] 
970 P-O 
From organo-phosphate esters (phosphorus 
containing copoly esters (PET-co-PEPPs)) 
[18] 
926, 755 phenyl-O-P From DHPDOPO flame retardant [40] 
915 
CH
2
-O-CH
2
 
(epoxide ring) 
From epoxy resin monomer [44] 
915 
CH
2
-O-CH
2 
(epoxide ring) 
From DGEBA with DHPDOPO and BPA 
(phosphorus containing epoxy system) 
[40] 
909 CH
2
-O-CH
2
 
From BPHPPO-containing epoxy resin  
(phosphorus based flame retardant system) 
[41] 
739 phenyl-H DGEBTBA (brominated epoxy) [39] 
 
 
 
 
 
50  
1. Resin (DGEBA) 
Functional 
groups 
Absorption 
(cm
-1
) 
-C-OH 3500-3400 
-C-H 3000-2850 
 
1600-1450 
phenyl-O-C 1040-1000 
 
915-905 
 
2. Curing agent 
 
DICY curing agent Phenolic curing agent 
 
Functional 
groups 
Absorption 
(cm
-1
) 
Functional 
groups 
Absorption 
(cm
-1
) 
 
-NH- 3350-3200 
 
-C=N- 1650 
Functional groups listed 
in 1 (Resin) also exist in 
phenolic curing agent 
 
 
3. Flame retardant (FR) 
Bromine based FR Phosphorus based FR 
Functional 
groups 
Absorption 
(cm
-1
) 
Functional groups 
Absorption  
(cm
-1
) 
phenyl-P 1590, 1460, 1350 
phenyl-Br 1066 
P=O 1300-1180 
 
Figure 41: Typical functional groups present in the epoxy laminate systems 
Complex molecular structure of the laminates combined with the overlap of bond 
energies made it difficult to interpret the spectra by locating the exact wave numbers 
corresponding to the functional groups. The peaks corresponding to key functional 
groups were identified in the FTIR spectra of material H (control sample) (shown in 
Figure 42). Material H is a mid T
g
 phenolic-cured brominated flame retardant system 
with fillers and hence the following functional groups exist in the material: OH: 
3500?3400 cm
-1
, CH: 3000?2850 cm
-1
,
 
Benzene ring: 1600?1450 cm
-1
, Phenyl-Br: 
1066 cm
-1
. The spectra corresponding to the control and 6X reflowed samples of 
 
51  
material H is shown in Figure 43 for illustration. The spectra of material B is shown 
in Appendix II. 
0.10
0.20
0.30
0.40
0.50
0.60
0.70
800 1300 1800 2300 2800 3300 3800
Wave number (cm-1)
Ab
s
o
r
p
t
i
o
n
Ab
s
o
r
p
t
i
o
n
 
Figure 42: FTIR spectrum of control sample with functional groups identified 
(material H) 
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
800 1300 1800 2300 2800 3300 3800
Wave number (cm-1)
Ab
s
o
r
p
t
i
o
n
Control
6X Reflowed
Ab
s
o
r
p
t
i
o
n
Ab
s
o
r
p
t
i
o
n
 
Figure 43: FTIR spectra of a control and 6X reflowed sample (material H) 
FTIR results did not show noticeable change in the wave numbers corresponding to 
the functional groups between control and 6X reflowed samples (shape of the 
spectrum is retained). This indicates that the variation in material properties could 
possibly be attributed to an increase in cross-linking of the resin system than to the 
degradation of the polymer network. The results of decomposition temperature 
 
52  
measurements also compliment this as no noticeable variation in T
d
 of the materials 
was observed due to lead-free soldering exposures. 
6.2 Combinatorial property analysis 
A combinatorial property analysis to investigate the underlying mechanisms for the 
variations in material properties is illustrated for materials B and H. B is a high T
g
 
DICY-cured halogen-free material with fillers, whereas H is a mid T
g
 phenolic-cured 
halogenated material with fillers. The illustration is focused on the material properties 
listed in Table 12 and the magnitude of variation shown in the table is between the 
results of control and 6X reflowed samples. The analysis is aimed at correlating the 
variations in material properties to the assumed enhanced-cure structure. 
Table 12: Summary of variations in properties for materials B and H 
Property Material ID 
Observed 
variation 
Magnitude of 
variation 
Glass transition temperature (T
g
) B Decrease >5
o
C 
Coefficient of thermal expansion  
(CTE, out-of-plane) 
B, H Decrease >15% 
Water absorption B, H Increase >25% 
 
The reduction of CTE observed in the 6X reflowed samples of materials B and H 
could be attributed to the increased cross-linking density of the epoxy resin in the 
assumed enhanced-cure structure. Water absorption in the epoxy matrix occurs due to 
the availability of free-volume in the laminate and the affinity of water molecules to 
the polar sites present in the polymer ([36], [46], [47]). An increase in cross-linking 
density of the resin system enhances the free-volume of the polymer structure making 
the laminates more hydrophilic, resulting in an increase in water absorption ([36], 
 
53  
[48], [49], [50]). This could be the cause of increase in water absorption after 
exposures observed in materials B and H. 
The T
g
 of a polymer is dependent on several parameters such as molecular weight, 
degree of cross-linking, and entrapped plasticizers [51]. Water typically acts as a 
plasticizer in the epoxy systems, resulting in a reversible reduction of T
g
 ([52], [53], 
[54], [55]). As discussed earlier, the laminates exposed to soldering conditions are 
more susceptible to moisture absorption. A decrease of T
g
 observed in material B 
could be the result of the plasticizing effect of moisture. The samples might have 
retained the moisture even after the preconditioning (2 hours of baking at 105
o
C) 
performed prior to T
g
 measurement as per the IPC test method. Material B is a DICY-
cured system, and hence has highly polar bonds compared to phenolic-cured material 
H. Also, B is a high T
g
 material having high cross-linking density compared to mid T
g
 
material H. The combined effect of high polarity and the presence of greater free-
volume in the structure could have resulted in a higher moisture absorption in 
material B, leading to a reduction of T
g. 
To further verify the hypothesis, T
g
 of the 6X reflowed samples of material B was 
measured after baking at 110
o
C for 24 hours. It was found that the T
g
 reduction 
between control and 6X reflowed samples decreased from 15
o
C to 8
o
C, thus proving 
that the increase in hydrophilic nature of the material results in a degradation of T
g
. 
The T
g
 of material F (mid T
g
 DICY-cured halogen-free) was also measured after 
baking at 110
o
C for 24 hours, and a decrease in the reduction of T
g
 from 9
o
C to 4
o
C 
between control and 6X reflowed samples was observed. These observations are 
represented in Figure 44. The incomplete recovery of the T
g
 after baking suggest a 
 
54  
possibility for a secondary mechanism in the high and mid T
g
 DICY-cured halogen-
free systems contributing to the reduction of T
g
 due to lead-free soldering exposures. 
130
135
140
145
150
155
160
165
170
175
G
l
a
s
s
 
t
r
a
n
s
i
t
i
o
n
 t
e
m
p
era
t
u
re (
d
eg
rees
 C
)
Control
6X
6X-Baked
Material B Material F
G
l
a
s
s
 
t
r
a
n
s
i
t
i
o
n
 t
e
m
p
era
t
u
re (
d
eg
rees
 C
)
 
Figure 44: Effect of baking on the T
g
 of laminates 
 
55  
Chapter 7: Summary and Conclusions 
 
Selection of PCB laminates compatible with lead-free processes is primarily based 
on their material properties, and is also impacted by factors such as application 
environment, cost, reliability, regulatory compliance, material sources, and 
availability. The laminate properties are determined by the constituents such as type 
of epoxy, curing agents, fillers, and flame retardants present in the material. 
In the materials studied: 
? High T
g
 laminates have lower out-of-plane CTE and flammability compared 
to low T
g 
materials. Low T
g 
laminates, on the other hand, have higher T
d
, T-
260 and lower water absorption compared to high T
g 
materials with similar 
constituents.  
? Although DICY and phenolic-cured laminates can have similar T
g
 and out-
of-plane CTE; a higher T
d
, T-260 and lower water absorption, flammability 
was observed in the phenolic-cured materials compared to similar DICY-
cured counterparts. 
? The presence of fillers lowers the out-of-plane CTE of the laminates, 
whereas the T
g
, T
d
, T-260, water absorption, and flammability does not have 
a strong dependence on fillers. 
? Halogen-free and halogenated materials can have similar T
g
,
 
T-260, water 
absorption and flammability, whereas lower out-of-plane CTE was observed 
in halogen-free materials compared to halogenated materials. Also, halogen-
 
56  
free material that is DICY-cured has higher T
d
 compared to the halogenated 
DICY-cured material. On the contrary, halogen-free material that is 
phenolic-cured has lower T
d
 compared to its halogenated counterpart. 
The high temperature exposures associated with lead-free soldering assembly 
conditions result in variations in the material properties of certain FR-4 laminate 
material types. The exposures tend to lower the T
g
, out-of-plane CTE, and T-260 of 
the laminate materials. An increase in water absorption and flammability was 
observed in most of the laminates due to exposures. The exposures did not affect the 
laminate materials to an extent of changing their decomposition temperatures (T
d
). 
The variation in material properties due to lead-free soldering exposures is attributed 
to the degree of cross-linking and to the extent of water absorption in the exposed 
laminates. 
 
57  
Chapter 8: Contributions 
? This is the first published study demonstrating the effects of lead-free 
soldering exposures on key thermomechanical, physical, and chemical 
properties of FR-4 laminates. The analysis of characterization provides a 
guideline for the selection of laminates for appropriate applications. 
? The laminate material types that are most affected by lead-free soldering 
exposures are illustrated. 
? Based on the observations, the following recommendations have been made: 
? Laminate manufacturers should conduct in-house qualification tests 
on the laminates to assess the variations in material properties. 
Corrective actions should be taken by tailoring the material 
constituents and/or laminate fabrication process conditions for 
achieving thermally stable laminates.  
? Electronic product manufacturers should gather the information 
about material constituents from the laminate suppliers, and consider 
the extent of variations in material properties due to lead-free 
soldering exposures before making a decision on the selection of 
appropriate laminates. 
 
 
58  
Appendix I: Datasheet values for the material properties studied 
CTE (ppm/
o
C) 
[out-of-plane] 
CTE  (ppm/
o
C) 
[in-plane: warp] 
CTE  (ppm/
o
C) 
[in-plane: fill] 
Supplier # T
g
 (
o
C) 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
Below 
T
g
 
Above 
T
g
 
I A 165-175 (DSC) 50-70 200-300 N/A N/A N/A N/A 
I B 165-175 (DSC) 30-50 200-230 N/A N/A N/A N/A 
I C1 165-175 (DSC) 40-60 270-300 N/A N/A N/A N/A 
I C2 165-175 (DSC) 40-60 250-270 N/A N/A N/A N/A 
II D1 173-183 (DSC) 50-60 200-300 12-15 N/A 14-17 N/A 
II D2 170-175 (DSC) 20-30 130-160 12-14 N/A 12-14 N/A 
II E 170-180 (TMA) 40-50 280-310 15-18 N/A 15-18 N/A 
I F 145-155 (DSC) 30-50 200-230 N/A N/A N/A N/A 
I G1 150-160 (DSC) 40-60 270-300 N/A N/A N/A N/A 
I G2 145-155 (DSC) 40-60 250-270 N/A N/A N/A N/A 
II H 135-145 (DSC) 35-45 180-240 12-15 N/A 14-17 N/A 
II I 140-150 (TMA) 40-55 150-220 13-16 N/A 16-19 N/A 
I J 135-145 (TMA) 50-70 250-350 N/A N/A N/A N/A 
I K 175-185 (DMA) 50-70 250-350 N/A N/A N/A N/A 
 
T
d
 (
o
C) 
Supplier # 
2% weight loss 5% weight loss 
Water 
absorption 
Flammability 
I A N/A 310 0.20-0.30 UL 94 V-0 
I B N/A 350 0.20-0.30 UL 94 V-0 
I C1 N/A 351 0.20-0.30 UL 94 V-0 
I C2 N/A 351 0.05-0.10 UL 94 V-0 
II D1 N/A N/A N/A UL 94 V-0 
II D2 N/A N/A N/A UL 94 V-0 
II E N/A 350-370 N/A UL 94 V-0 
I F N/A 350 0.20-0.30 UL 94 V-0 
I G1 N/A 350 0.20-0.30 UL 94 V-0 
I G2 N/A 350 0.05-0.10 UL 94 V-0 
II H N/A 340-360 N/A UL 94 V-0 
II I N/A N/A N/A UL 94 V-0 
I J N/A 325 0.05-0.10 UL 94 V-0 
I K N/A 350 0.15-0.18 UL 94 V-0 
 
* The data sheet values are for a specific construction (type of glass weave style, 
number of glass plies, resin content, and nominal thickness) of the laminate material 
which is different from that of the laminate materials considered for this study. 
 
59  
Appendix II: FTIR spectra of control and 6X reflowed samples of 
material B 
FTIR spectra of material B (control)
0.10
0.20
0.30
0.40
0.50
0.60
0.70
800 1300 1800 2300 2800 3300 3800
Wave number (cm-1)
Ab
s
o
r
p
t
i
o
n
 
FTIR spectra comparison between control and 6X reflow samples of 
material B
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
800 1300 1800 2300 2800 3300 3800
Wave number (cm-1)
Ab
s
o
r
p
t
i
o
n
Control
6X Reflowed
Ab
s
o
r
p
t
i
o
n
 
 
60  
 Bibliography 
 
[1] The European Parliament and the Council of the European Union, ?Directive 
2002/95/EC on the restriction of the use of certain hazardous substances (RoHS) 
in electrical and electronic equipment?, Official Journal of the European Union, 
pp. L. 37/19-37/23, February, 2003. 
[2] The European Parliament and the Council of the European Union, ?Directive 
2002/95/EC on the waste electrical and electronic equipment (WEEE)?, Official 
Journal of the European Union, pp. L. 37/24-37/38, February, 2003. 
[3] Ganesan S., and Pecht M., ?Lead-free electronics?, IEEE Press, Wiley-
Interscience, A. John Wiley and Sons, Inc., New Jersey, USA, 2006. 
[4] Coombs C., ?Printed circuit handbook?, McGraw Hill, fifth edition, 2001. 
[5] Bergum E., ?Thermal analysis of base materials through assembly: Can current 
analytical techniques predict and characterize difference in laminate 
performance prior to exposure to thermal excursions during assembly??, Printed 
Circuit Design & Manufacture, September 2003. 
[6] Kelley E., and Bergum E., ?Laminate material selection for RoHS assembly, 
Part 1?, Printed Circuit Design & Manufacture, pp. 30-34, November 2006. 
[7] Kelley E., and Bergum E., ?Laminate material selection for RoHS assembly, 
Part 2?, Printed Circuit Design & Manufacture, pp. 30-37, December 2006. 
[8] Kelley E., ?An assessment of the impact of lead-free assembly processes on 
base material and PCB reliability,? Proceedings of IPC APEX Conference, pp. 
S16-2-1, 2004. 
[9] Ehrler S., ?The compatibility of epoxy-based printed circuit boards with lead-
free assembly?, Circuit World, vol. 31, no. 4, pp.3-13, 2005. 
[10] Christiansen W., Shirrell D., Aguirre B., and Wilkins J., ?Thermal stability of 
electrical grade laminates based on epoxy resins?, Proceedings of IPC Printed 
Circuits EXPO, Anaheim, CA, pp. S03-1-1-S03-1-7, 2001. 
 
61  
[11] Valette L., and Wiechmann R., ?High-performance substrate from new epoxy 
resin and enhanced copper foil?, Circuit World, vol. 30, no. 4, pp. 20-26, 2004. 
[12] Pecht M., Ardebili H., Shukla A., Hagge J., and Jennings D., ?Moisture ingress 
into organic laminates?, IEEE Transactions on Components and Packaging 
Technology, vol. 22, no. 1, pp. 104-110, March 1999. 
[13] Qi H., Ganesan S., Wu J., and Pecht M., ?Effects of printed circuit board 
materials on lead-free interconnect durability?, 5th International Conference on 
Polymers and Adhesives in Microelectronics and Photonics, Wroclaw, Poland, 
pp. 140-144, October 23-26, 2005. 
[14] Sottos R., Ockers M., and Swindeman M., ?Thermoelastic properties of plain 
weave composites for multilayer circuit board applications?, Journal of 
Electronic Packaging, Transactions of the ASME, vol. 121, no. 1, pp. 37-44, 
1999. 
[15] Peng Y., Qi X., and Chrisafides C., ?The influence of curing systems on 
epoxide-based PCB laminate performance?, Circuit World, vol. 31, no. 4, pp. 
14-20, 2005. 
[16] Camino G., and Costa L., ?Performance and mechanisms of fire retardants in 
polymers?A review?, Polymer Degradation Stability, vol. 20, no. 3-4, pp. 271-
294, 1988. 
[17] Jain P., Choudhary V., and Varma I., ?Flame retarding epoxies with 
phosphorous?, Journal of Macromolecular Science-Polymer Reviews, vol. 42, 
no. 2, pp. 139-183, 2002. 
[18] Chang, J., Sheen Y., Chang R., and Chang F., ?The thermal degradation of 
phosphorous-containing co polyesters?, Polymer Degradation and Stability, vol. 
54, no. 2-3, pp. 365-371, 1996. 
[19] Liu W., Varley R., and Simon G., ?Understanding the decomposition and fire 
performance processes in phosphorus and nanomodified high performance 
epoxy resins and composites?,  Polymer, vol. 48, no. 8, pp. 2345-2354, 2007. 
 
62  
[20] Iji M, and Kiuchi Y., ?Flame resistant glass-epoxy printed wiring boards with 
no halogen or phosphorous compounds?, Journal of Materials Science: 
Materials in Electronics, vol. 15, pp. 175-182, 2004. 
[21] IPC-4121, ?Guidelines for selecting core constructions for multilayer printed 
wiring board applications?, Bonnockburn, IL, January 2000. 
[22] IPC/JEDEC J-STD-020D, ?Moisture/reflow sensitivity classification for non-
hermetic solid state surface mount devices?, August 2007. 
[23] IPC-TM-650 2.4.25, ?Glass transition temperature and cure factor by DSC?, 
The Institute for Interconnecting and Packaging Electronic Circuits, 
Northbrook, IL, December 1994. 
[24] IPC-TM-650 2.4.24, ?Glass transition temperature and z-axis thermal expansion 
by TMA?, The Institute for Interconnecting and Packaging Electronic Circuits, 
Northbrook, IL, December 1994. 
[25] Reinhart J., ?Engineering materials handbook?, vol. 1, Composites, ASM 
International, Metals Park, OH, 1987. 
[26] Wu T, Guo Y., and Chen W., "Thermal-mechanical strain characterization for 
printed wiring boards," IBM Journal of Research and Development, vol. 37, no. 
5, pp. 621-634, 1993. 
[27] Yuan J., and Falanga L., ?The in-plane thermal expansion of glass fabric 
reinforced epoxy laminates?, Journal of Reinforced Plastics and Composites, 
vol. 12, pp. 489-496, 1993. 
[28] IPC-TM-650 2.4.24.6, ?Decomposition of laminate material using TGA?, The 
Institute for Interconnecting and Packaging Electronic Circuits,, Bannockburn, 
IL, April 2006. 
[29] IPC-TM-650 2.4.24.1, ?Time-to-delamination (by TMA method)?, The Institute 
for Interconnecting and Packaging Electronic Circuits, Northbrook, IL, 
December 1994. 
 
63  
[30] IPC-TM-650 2.6.2.1A, ?Water absorption, metal clad plastic laminates?, The 
Institute for Interconnecting and Packaging Electronic Circuits, Northbrook, 
IL, May 1986. 
[31] Lambert, R, ?The impact of materials on the flammability of printed wiring 
board products?, 43rd Proceedings of Electronic Components and Technology 
Conference, pp. 134 - 142, June 1993. 
[32] UL 94, ?Tests for flammability of plastic materials for parts in devices and 
appliances?, Underwriters Laboratories, June 1990. 
[33] Paterson-Jones J., Percy V., Giles R., and Stephen A., ?The thermal degradation 
of model compounds of amine-cured epoxide resins. II. The thermal 
degradation of 1,3-diphenoxypropan-2-ol and 1,3-diphenoxypropene?, Journal 
of Applied Polymer Science, vol. 17, no. 6, pp. 1877-1887, 1973. 
[34] Brito Z., and Sanchez G., ?Influence of metallic fillers on the thermal and 
mechanical behavior in composites of epoxy matrix?, Composite Structures, 
vol. 48, no. 1-3, pp. 79-81, 2000. 
[35] Sanchez G., Brito Z., Mujica V., and Perdomo G., ?The thermal behavior of 
cured epoxy-resins: The influence of metallic fillers?, Polymer Degradation 
Stability, vol. 40, no. 1, pp. 109-114, 1993. 
[36] Diamant Y., Marom G., and Broutman L., ?The effect of network structure on 
moisture absorption of epoxy resins?, Journal of Applied Polymer Science
,
 vol. 
26, pp. 3015-3025, 1981. 
[37] Levchik S., and Weil E., ?Thermal decomposition, combustion and flame-
retardancy of epoxy resins-a review of the recent literature?, Polymer 
International, vol. 53, pp. 1901-1929, 2004. 
[38] http://www.ptli.com/testlopedia/tests/FTIR-E168andE1252-more.asp (accessed 
on 03/05/2008) 
[39] Luda M., Balabanovic A., and Camino G., ?Thermal decomposition of fire 
retardant brominated epoxy resins?, Journal of Analytical and Applied 
Pyrolysis, vol. 65, pp. 25-40, 2002. 
 
64  
[40] Wang X., and Zhang Q., ?Synthesis, Characterization, and cure properties of 
phosphorus-containing epoxy resins for flame retardance?, European Polymer 
Journal, vol. 40, pp. 385-395, 2004. 
[41] Ren H., Su J., Wu B., and Zhou Q., ?Synthesis and properties of a phosphorus-
containing flame retardant epoxy resin based on bis-phenoxy (3-hydroxy) 
phenyl phosphine oxide?, Polymer Degradation and Stability, vol. 92, pp. 956-
961, 2007. 
[42] Ogi K., ?Influence of thermal history on transverse cracking in a carbon fiber 
reinforced epoxy composite?, Advanced Composite Materials, vol. 11, no. 3, pp. 
265-275, 2003. 
[43] Wang Q., and Shi W., ?Kinetics study of thermal decomposition of epoxy resins 
containing flame retardant components?, Polymer Degradation and Stability, 
vol. 91, pp. 1747-1754, 2006. 
[44] Gundjian M., and Cole K., ?Effect of copper on the curing and structure of dicy-
containing epoxy composite system?, Journal of Applied Polymer Science, vol. 
75, pp. 1458-1473, 2000. 
[45] Liu Y., Hsiue G., Lan C., and Chiu Y., ?Phosphorus-containing epoxy for flame 
retardance: IV. Kinetics and mechanism of thermal degradation?, Polymer 
Degradation and Stability, vol. 56, pp. 291-299, 1997. 
[46] Marsh L., Lasky R., Seraphim D., and Springer G., ?Moisture solubility and 
diffusion in epoxy and epoxy-glass composites?, IBM Journal of Research and 
Development, vol. 28, no. 6, pp. 655-661, 1984. 
[47] Maggana C., and Pissis P., ?Water sorption and diffusion studies in an epoxy 
resin system?, Journal of Polymer Science: Part B: Polymer Physics, vol. 37, 
no. 11, pp. 1165-1182, 1999. 
[48] Ko M., and Kim M., ?Effect of postmold curing on plastic IC package 
reliability?, Journal of Applied Polymer Science, vol. 69, no. 11, pp. 2187-2193, 
1998. 
 
65  
[49] Gonon P., Sylvestre A., Teysseyre J., and Prior C., ?Combined effects of 
humidity and thermal stress on the dielectric properties of epoxy-silica 
composites?, Materials Science and Engineering B , vol. 83, no. 1-3, pp. 158-
164, June 2001. 
[50] Aronhime M., Peng X., and Gillham J., ?Effect of time-temperature path of cure 
on the water absorption of high T
g
 epoxy resins,? Journal of Applied Polymer 
Science, vol. 32, pp. 3589-3626, 1986. 
[51] Wondraczek K., Adams J., and Fuhrmann J., ?Effect of thermal degradation on 
glass transition temperature of PMMA?, Macromolecular Chemistry and 
Physics, vol. 205, no. 14, pp. 1858-1862, 2004. 
[52] Sala G., ?Composition degradation due to fluid absorption?, Composites Part B: 
Engineering, vol. 31, no. 5, pp. 357-373, July 2000. 
[53] Smith C., ?Water absorption in glass fiber-epoxide resin laminates?, Circuit 
World, vol. 14, no. 3, pp. 22 -26, 1988. 
[54] Marsh L., Lasky R., Seraphim D., and Springer G., ?Moisture solubility and 
diffusion in epoxy and epoxy-glass composites?, IBM Journal of Research and 
Development, vol. 28, no. 6, pp. 655-661, 1984. 
[55] Flor G., Campari-Vigano G., and Feduzi R., ?A thermal study on moisture 
absorption by epoxy composites?, Journal of Thermal Analysis and 
Calorimetry, vol. 35, no. 7, pp. 2255-2264, November, 1989.