Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Chemical recycling of amine/epoxy composites at atmospheric pressure
(USC Thesis Other)
Chemical recycling of amine/epoxy composites at atmospheric pressure
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
CHEMICAL RECYCLING OF AMINE/EPOXY COMPOSITES AT
ATMOSPHERIC PRESSURE
By
Yijia Ma
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Chemical Engineering)
May 2019
Copyright 2019 Yijia Ma
ii
Acknowledgements
Graduate school and my time at USC have been a challenging yet fulfilling journey. The
expertise, knowledge, experience that I gained here have been invaluable.
My deepest thanks goes to Prof. Steven Nutt, my Ph.D. advisor, for his tremendous support
and guidance throughout the completion of my degree. He taught me to do research, write and
present my work, and think from a broader perspective. He also gave me great opportunities to
challenge myself and grow. I started with almost no research experience and have become an
increasingly independent scholar under his supervision.
I would also like to thank Prof. Travis Williams, who taught me chemistry and catalysis. I
have greatly benefited from his knowledge and many years of close collaboration, through which I
have been learning the chemist’s mindset and new perspectives of my research. His suggestion and
input have been invaluable towards the completion of my study.
I am also grateful to have the support from my colleagues at the USC M.C. Gill
Composites Center, who were always willing to help when I needed a hand. I would particularly
like to thank Prof. Timotei Centea and Dr. Daniel Kim who provided enormous support to my
research. They helped me get started on my projects, took time to discuss the details of my work
and provided me with valuable feedback to my papers. I would not have gone this far without their
expertise and encouragement.
This work was made possible with the financial support from National Science Foundation
and the University of Southern California, as well as the generous material/testing support from
several companies, including Huntsman Corporation, Cytec Industries, Airtech International, ELG
Carbon Fibre, and Shimadzu. I would like to acknowledge several industrial experts, Pete George
from Boeing and Dr. Jack Boyd from Axiom Materials, for their helpful suggestions and continued
support during my study.
iii
Lastly, I would like to thank my family and friends, for believing in me and always being
there for me throughout these years. None of this would have been possible without their love and
support.
iv
Table of Contents
Acknowledgements ............................................................................................................................ ii
Table of Contents .............................................................................................................................. iv
List of Figures ................................................................................................................................. viii
List of Tables .................................................................................................................................. xiv
Abstract ............................................................................................................................................ xv
CHAPTER 1. Introduction ................................................................................................................. 1
1.1. Challenges for Composites ...................................................................................................... 3
1.1.1. Effective Recycling ........................................................................................................... 3
1.1.2. Sustainable Manufacturing ................................................................................................ 6
1.2. Objectives and Structure .......................................................................................................... 6
CHAPTER 2. Composites Recycling – Current Technologies ....................................................... 10
2.1. Mechanical Recycling ........................................................................................................... 12
2.2. Thermal Processing ............................................................................................................... 13
2.3. Chemical Recycling ............................................................................................................... 15
2.3.1. Recycling of Composites in Near- and Supercritical Fluids ........................................... 16
2.3.2. Recycling of Composites at Atmospheric Pressure ........................................................ 18
CHAPTER 3. Chemical Treatment for Dissolution of Amine-cured Epoxies at Atmospheric
Pressure ............................................................................................................................................ 20
3.1. Introduction ........................................................................................................................... 20
3.2. Experiments ........................................................................................................................... 22
3.2.1. Resin Formulation ........................................................................................................... 22
3.2.2. Thermal Analysis ............................................................................................................ 24
3.2.3. Amine-cured Epoxy Dissolution ..................................................................................... 24
v
3.2.4. Nuclear Magnetic Resonance Spectroscopy ................................................................... 26
3.2.5. Test Matrix ...................................................................................................................... 26
3.3. Results and Discussion .......................................................................................................... 28
3.3.1. Bi-functional Epoxy Dissolution ..................................................................................... 28
3.3.2. Effect of Diffusion .......................................................................................................... 29
3.3.3. Tri- and Tetra-functional Epoxy Dissolution .................................................................. 30
3.3.4. Cleavable Sites ................................................................................................................ 33
3.4. Conclusions ........................................................................................................................... 42
CHAPTER 4. Chemical Treatment for Recycling of Amine/Epoxy Composites at Atmospheric
Pressure ............................................................................................................................................ 44
4.1. Introduction ........................................................................................................................... 44
4.2. Experiments ........................................................................................................................... 46
4.2.1. Resin Formulation ........................................................................................................... 46
4.2.2. Thermal Analysis ............................................................................................................ 47
4.2.3. Composite Fabrication .................................................................................................... 48
4.2.4. Composite Recycling ...................................................................................................... 50
4.2.5. Fiber Characterization ..................................................................................................... 51
4.2.6. Test Matrix ...................................................................................................................... 53
4.3. Results and Discussion .......................................................................................................... 55
4.3.1. Recycling Methods: Depolymerization vs Acid Digestion ............................................. 55
4.3.2. Effect of Fiber Bed Architecture ..................................................................................... 59
4.3.3. Effect of Matrix Functionality ......................................................................................... 60
4.3.4. Effect of Reinforcement Type ......................................................................................... 61
4.3.5. Strategies to Accelerate Diffusion ................................................................................... 65
vi
4.3.5.1. Pre-treatment ............................................................................................................. 65
4.3.5.2. Shredded Materials ................................................................................................... 68
4.3.6. Catalyst Development ..................................................................................................... 69
4.4. Conclusions ........................................................................................................................... 71
CHAPTER 5. Recovery and Reuse of Acid Digested Amine/Epoxy-based Composite Matrices .. 74
5.1. Introduction ........................................................................................................................... 74
5.2. Experiments ........................................................................................................................... 77
5.2.1. Epoxy Formulation for Acid Digestion ........................................................................... 77
5.2.2. Oxidative Acid Digestion ................................................................................................ 77
5.2.3. Reuse of Decomposed Matrix Residues .......................................................................... 78
5.2.4. Characterization .............................................................................................................. 80
5.3. Results and Discussion .......................................................................................................... 80
5.3.1. Recovered Epoxies .......................................................................................................... 80
5.3.2. Reuse Approach I: As Accelerators for Anhydride/Epoxy Formulations ....................... 82
5.3.3. Reuse Approach II: As Fillers for Epoxy Resins ............................................................ 89
5.4. Conclusions ........................................................................................................................... 91
CHAPTER 6. Defect Reduction Strategies for the Manufacture of Contoured Laminates Using
Vacuum Bag-Only Prepregs ............................................................................................................ 94
6.1. Introduction ........................................................................................................................... 94
6.1.1. Objectives and Structure ................................................................................................. 98
6.2. Methods ................................................................................................................................. 98
6.2.1. Mold Design .................................................................................................................... 99
6.2.2. Materials ........................................................................................................................ 100
6.2.3. Cure ............................................................................................................................... 100
vii
6.2.4. Test Matrix .................................................................................................................... 102
6.2.5. Quality Analysis ............................................................................................................ 102
6.3. Results and Discussion ........................................................................................................ 104
6.3.1. General Defects in Sharp Corner Laminates ................................................................. 104
6.3.2. Effect of Intermediate Debulk Method ......................................................................... 110
6.3.3. Effect of Mold Corner Radius ....................................................................................... 110
6.3.4. Pressure Strip Application ............................................................................................. 112
6.4. Conclusions ......................................................................................................................... 115
CHAPTER 7. Conclusions and Future Work ................................................................................ 117
7.1. Conclusions ......................................................................................................................... 117
7.2. Broader Implication ............................................................................................................. 118
7.3. Future Work ......................................................................................................................... 119
7.3.1. Improved Recycling Conditions for Composite ............................................................ 119
7.3.2. Prepreg Recycling ......................................................................................................... 122
7.3.3. Recyclates Reuse ........................................................................................................... 123
7.3.4. Scale-up Potential .......................................................................................................... 124
Appendix A. Supplementary Materials for Neat Epoxy Dissolution ............................................ 125
Appendix B. Supplementary Materials for Composites Recycling ............................................... 131
Appendix C. Supplementary Materials for Decomposed Polymers Reuse ................................... 134
References ...................................................................................................................................... 139
viii
List of Figures
Figure 1-1. Global carbon fiber market volume share by application, 2016 (%) [5] ........................ 2
Figure 1-2. Closing the CFRP lifecycle loop: carbon fiber reclaimed from end-of-life sources are
repurposed for use in automotive applications [8] ............................................................................. 4
Figure 1-3. Recovering near-virgin quality fibers and useful polymers from amine/epoxy
composites using atmospheric pressure and moderate temperature .................................................. 8
Figure 2-1. A plane graveyard in in a Californian desert (Picture: Jassen Todorov/Caters) ........... 11
Figure 2-2. Schematic of cross-section of Wittmann mechanical recycling machine [19] ............. 12
Figure 2-3. Fluidised bed process schematic [22] ........................................................................... 14
Figure 2-4. The Thermolyze
system in Forst, Germany used for the bulk pyrolysis of the studied
composite materials [24] .................................................................................................................. 15
Figure 2-5. Process flow of grinding / subcritical water hydrolysis / inorganic materials separation
[32] ................................................................................................................................................... 17
Figure 3-1. Epoxy monomers: bi-(a), tri-(b), and tetra-(c) functional epoxies, and amine curing
agents: (d) 3,3’-DDS, and (e) M-DEA (red and blue represent oxygen (O) and nitrogen (N),
respectively) ..................................................................................................................................... 23
Figure 3-2. Effect of A/E ratio on a) dissolution time and b) Tg’s for 3,3’-DDS-cured epoxies ..... 29
Figure 3-3. Effect of pre-treatment on epoxy dissolution time ........................................................ 30
Figure 3-4. Effect of epoxy functionality on a) dissolution time and b) Tg values and numbers of
crosslinked bonds for 3,3’-DDS-cured epoxies ............................................................................... 32
Figure 3-5. Possible cleavable sites in amine-cured epoxies ........................................................... 34
Figure 3-6. Dissolution times for M-DEA-cured epoxies: a) effect of A/E ratio (bi-functional
epoxy), b) effect of epoxy functionality (A/E = 40%) ..................................................................... 35
ix
Figure 3-7. Qualitative a) 1D and b) 2D HNMR spectrum analyses of dissolution products (26 h
depolymerization) of tri-functional epoxy in DMSO at room temperature (32 scans) .................... 36
Figure 3-8. Quantitative 1D HNMR spectrum analyses of dissolution products of tri-functional .. 38
Figure 3-9. Molecular structure in epoxy monomers and amine-cured epoxies .............................. 39
Figure 3-10. Assignment of 1D HNMR full spectrum chemical shift of dissolution products of tri-
functional epoxy in DMSO at room temperature (32 scans) ........................................................... 42
Figure 4-1. Resin formulations for composite matrices: (a) bi-functional epoxy, (b) tri-functional
epoxy, (c) tetra-functional epoxy, (d) amine curing agent, (e) crosslinked bi-functional epoxy
matrix, (f) selected epoxy matrices for composite fabrication ........................................................ 48
Figure 4-2. Composite fabrication: a) vacuum bag-only manufacturing setup, b) polished section of
cured laminates ................................................................................................................................ 49
Figure 4-3. Sample mounting sheet with red dash lines indicating the cut positions ...................... 52
Figure 4-4. Effect of thickness on epoxy matrix (DDS/Bi =100%) dissolution rate ....................... 56
Figure 4-5. XPS spectra and SEM images of virgin and recovered carbon fibers: (a) XPS survey
spectrum, and C1s spectra and SEM images for (b) virgin carbon fibers, (c) carbon fibers
recovered from depolymerization, (d) carbon fibers recovered from acid digestion ...................... 58
Figure 4-6. Effect of fiber bed architecture on epoxy matrix (DDS/Bi =100%) dissolution rate .... 60
Figure 4-7. Effect of epoxy functionality on matrix dissolution rate .............................................. 61
Figure 4-8. Effect of fiber reinforcement type on epoxy matrix (DDS/Bi =100%) dissolution rate 63
Figure 4-9. EDS spectra of epoxy residues (DDS/Bi= 100%) on recovered fibers: a) CF twill
weave, acid digestion, b) GF UD, acid digestion, c) GF plain weave, acid digestion, d) GF plain
weave, depolymerization ................................................................................................................. 64
x
Figure 4-10. Effect of pre-treatment on matrix (DDS/Bi= 100%) dissolution rate via acid digestion
(a), cross-sectional images of laminates (b) before and (c) after pre-treatment, and carbon fiber
fabrics recovered from (d) an 8-ply laminate using acid digestion with pre-treatment ................... 66
Figure 4-11. Carbon fiber fabrics recovered from commercial composites (5320-1/8HS): (a)
separated plies, (b) recovered fabrics with toughener residue, (c) clean fabric after DMSO wash . 67
Figure 4-12. Matrix dissolution rate in shredded composite waste ................................................. 69
Figure 4-13. Effect of catalysts on resin dissolution time ............................................................... 71
Figure 5-1. Amine-base epoxy formulation: (a) bi-functional epoxy (DGEBA), (d) tri-functional
epoxy, (c) tetra-functional epoxy, (d) 3,3’-DDS, (e) crosslinked bi-functional epoxy matrix ........ 76
Figure 5-2. Recovered matrix residues from (a) solutions after (b) precipitation and (c) granulation
.......................................................................................................................................................... 79
Figure 5-3. Anhydride-based epoxy formulation: (a) diglycidyl ether of bisphenol A (DGEBA), (b)
methyltetrahydrophthalic anhydride (MTHPA), (c) crosslinked bi-functional epoxy matrix ......... 79
Figure 5-4. HNMR spectra of recovered matrix residues after (a) 1 h reaction and (b) 4 h reaction,
and (c) DGEBA epoxy monomer .................................................................................................... 81
Figure 5-5. Reuse of recovered matrix residues as an accelerator for an anhydride-based bi-
functional epoxy formulation: (a) curing reactions, (b) Tg values of cured epoxies ........................ 84
Figure 5-6. Neat anhydride/epoxy specimens cured with a commercial accelerator (top) and with
recovered matrix from amine/epoxy matrices (bottom) .................................................................. 86
Figure 5-7. Flexural modulus (a) and stress-strain curves (b) for anhydride/epoxy samples cured
with a commercial accelerator and recovered matrices ................................................................... 87
Figure 5-8. FTIR spectra of DGEBA/ matrix residue systems after a cure cycle. .......................... 88
Figure 6-1. Schematic illustration of mold design and bagging configuration ................................ 99
xi
Figure 6-2. Thickness measurement locations for: (a) concave corner laminate, (b) convex corner
laminate, (c) concave corner laminate manufactured with a pressure strip ................................... 103
Figure 6-3. Laminates made of prepreg A: (a) void content in flange region, concave corner region
and convex corner region; (b) thickness variation ......................................................................... 105
Figure 6-4. Thickness variation of laminates made of prepreg B .................................................. 106
Figure 6-5. Defects in corner laminates: (a) concave corner, (b) convex corner ........................... 107
Figure 6-6. Overall thickness variation and fiber bed thickness variation in prepreg B laminates: (a)
four-ply laminates, (b) eight-ply laminates .................................................................................... 108
Figure 6-7. Concave corner resin length in laminates made of prepreg B .................................... 109
Figure 6-8. Thickness variation of laminates manufactured with and without debulking ............. 111
Figure 6-9. Thickness variation of laminates manufactured with various tool corner curvatures: (a)
concave corner laminates, (b) convex corner laminates ................................................................ 112
Figure 6-10. Thickness variation of laminates manufactured without and with a pressure strip .. 113
Figure 6-11. Typical microscopic images of laminates manufactured with a pressure strip: (a)
laminate fillet at the end of the pressure strip, (b) void distribution .............................................. 114
Figure 7-1. Fibers with residue (a) and matrix (b) recovered from 110 °C cured prepreg by DMSO
wash, clean fibers (c) and matrix (d) recovered after applying aerobic digestion to sample (c). .. 121
Figure 7-2. Clean fibers (a) and matrix (b) recovered from expired prepregs ............................... 123
Figure A-1. Comparison of Tg’s of epoxies cured using DDS and M-DEA with variations in (a)
amine/epoxy ratio and (b) epoxy functionality .............................................................................. 125
Figure A-2. 1D HNMR spectrum of benzyl alcohol in DMSO at room temperature (32 scans) .. 126
Figure A-3. 1D HNMR spectrum of 3,3’-DDS (3,3’-diamonodiphenyl sulfone) in DMSO at room
temperature (32 scans) ................................................................................................................... 126
xii
Figure A-4. 1D HNMR spectrum of bi-functional epoxy monomer (DGEBA, diglycidyl ether of
bisphenol A) in DMSO at room temperature (32 scans) ............................................................... 127
Figure A-5. 1D HNMR spectrum of tri-functional epoxy monomer (triglycidyl of para-
aminophenol) in DMSO at room temperature (32 scans) .............................................................. 127
Figure A-6. 1D HNMR spectrum of tetra-functional epoxy monomer (tetraglycidyl-4,4'-
methylenebisbenzenamine) in DMSO at room temperature (32 scans) ........................................ 128
Figure A-7. Analytical results of MALDI-TOF/MS of the dissolution products of 3,3’-DDS-cured
tri-functional epoxy: (a) depolymerization (34 h), (b) acid digestion (6 h) ................................... 128
Figure A-8. Analytical results of GC/MS of the dissolution products of 3,3’-DDS-cured bi-
functional epoxy: (a) mass chromatogram of dissolution products, (b) mass spectrum of product
bisphenol A .................................................................................................................................... 129
Figure B-1. (a) EDS spectrum of the residue on carbon fiber surface after depolymerization, (b)
EDS line scan crossing the residue. Phosphorus (P) signal in the spectra confirms the source of the
residue to be catalyst K3PO4. ......................................................................................................... 131
Figure B-2. SEM images of virgin (a) and recovered (b) glass fiber via acid digestion. Sizing was
removed after reaction. No defect or residue were observed on fiber surfaces. ........................... 131
Figure B-3. SEM images of the recovered carbon fibers from a laminate made of 5320-1/8HS via
acid digestion with pre-treatment. No defect or residue were observed on fiber surfaces. ........... 132
Figure B-4. SEM images of recovered carbon fiber from commercial composite waste (ELG
Carbon Fibre Ltd.) via acid digestion. No defect or residue were observed on fiber surfaces. ..... 132
Figure B-5. Recovered carbon fiber from shredded composite waste: a) and b) before acid
digestion, c) and d) after acid digestion ......................................................................................... 133
Figure C-1. Effect of commercial accelerators on curing reactions of anhydride-based (a) bi-, (b)
tri- and (c) tetra-functional epoxies ................................................................................................ 134
xiii
Figure C-2. Effect of commercial accelerator concentration on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies ........................................................................... 134
Figure C-3. Effect of recovered matrix residue concentration on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies (without a commercial accelerator) .................. 135
Figure C-4. Comparison of (a) curing reactions and (b) Tg values for anhydride-based bi-functional
epoxies with matrix residues without and with ScCl3 catalyst ...................................................... 135
Figure C-5. Effect of recovered matrix residue state on (a) curing reactions and (b) Tg values of
anhydride-based bi-functional epoxies .......................................................................................... 136
Figure C-6. Effect of recovered matrix residue functionality on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies ........................................................................... 136
Figure C-7. (a) Curing reactions and (b) Tg values of DGEBA/recovered matrix residue systems
........................................................................................................................................................ 137
Figure C-8. Full FTIR spectra of DGEBA/recovered matrix residue systems after a cure cycle. . 137
Figure C-9. Effect of recovered matrix concentration and state on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies (with accelerators) ............................................ 138
Figure C-10. Effect of recovered matrix concentration and state on (a) curing reactions and (b) Tg
values of amine-based bi-functional epoxies ................................................................................. 138
xiv
List of Tables
Table 1-1. Comparing virgin and recycled carbon fibers [9] ............................................................. 5
Table 3-1. Test matrix for amine-cured epoxy dissolution .............................................................. 27
Table 3-2. Possible cleavable crosslinked units in amine-cured epoxies with A/E = 40% ............. 40
Table 4-1. Test matrix ...................................................................................................................... 54
Table 4-2. Percentages of functional groups on carbon fiber surface. ............................................. 57
Table 4-3. Tensile strength and modulus of recovered carbon fibers from laminates with 8 plies . 59
Table 5-1. Assessment of using recovered amine/epoxy matrices as an accelerator for anhydride-
based bi-functional epoxy formulations ........................................................................................... 83
Table 5-2. Assessment of using recovered amine/epoxy matrix residue as fillers for anhydride-
(with a commercial accelerator) and amine-based bi-functional epoxy formulations ..................... 90
Table 6-1. Test matrix for out-of-autoclave processing ................................................................. 101
Table A-1. Bond dissociation energies (BDEs) for C-O, C-N and C-S bonds (boldface =
dissociated group) [76] .................................................................................................................. 130
xv
Abstract
Because of the increasing demand for lightweight structures in aerospace, automotive, and
wind energy industries, the global market size for carbon fiber polymer composites is anticipated
to reach $35 billion by 2020. The increasing use of composites poses significant environmental
problems because most end-of-life composite waste is not recovered and/or reused due to a lack of
viable recycling technologies. This issue is especially true for thermoset composites that undergo
irreversible cure reactions. Thermoset composites have largely resisted attempts to recycle because
the crosslinked polymer matrices cannot be easily separated from the fiber reinforcements. At the
present juncture, composite recyclability is essential to the sustainability of the growing composite
industry. Without a robust and effective method to recycle composites and complete the material
life-cycle, these materials will not be able to compete with steel and aluminum in mass market
applications, for which recycling rates are already high.
Herein, the recycling project aim to develop an effective chemical digestion method for
recycling of highly crosslinked amine/epoxy matrices using moderate conditions (atmospheric
pressure and moderate temperature) and safe chemicals. These features are critical to recover near-
virgin quality fibers and useful chemical components from composites with different crosslink
densities, but have not been reported to date.
Recycling is practical only if the recycling process itself can be scaled up to industrial
scale, in which fast epoxy dissolution rate is one crucial criterion. Therefore, this thesis first
focuses on neat polymer dissolution of amine-cured epoxies. Two chemical treatment methods -
depolymerization and acid digestion (both at atmospheric pressure) - were employed to dissolve
amine-cured epoxy formulations. Both depolymerization and acid digestion were shown to be
effective dissolution processes for all amine/epoxy samples that encompassed variations in
amine/epoxy stoichiometric ratio (A/E ratio), epoxy monomer functionality, and amine curing
xvi
agent type. The relationship between epoxy properties and dissolution rate was determined, and the
key parameters affecting thermoset matrix dissolution were identified. The dissolution rate was
controlled by both the chemical reaction and diffusion rates. The components of the chemical
solutions after epoxy dissolution were analyzed and identified, and protocols to quantitatively track
the products after dissolution were developed. The two major cleavable sites during epoxy
dissolution were the C-N and C-O bonds, and the aromatic structures of the epoxies were
preserved.
Based on the knowledge learned from neat epoxy dissolution, this thesis continues to
investigate the viability of recycling actual composites. Depolymerization and acid digestion were
applied to amine/epoxy composites, including composites produced from lab-made and aerospace
prepregs and commercial composite waste. Findings indicated that acid digestion was more
effective for highly crosslinked amine/epoxy composites than depolymerization. Furthermore,
digestion occurred via reaction steps of oxygen atom transfer to the aniline groups and then bond
cleavage, resulting in recovery of near-virgin quality fibers at faster dissolution rates and lower
temperatures. The relationship among epoxy functionality, fiber bed architecture, fiber
reinforcement, laminate thickness, and matrix dissolution rate were investigated, and key
parameters affecting the dissolution rate were identified. Two strategies to enhance the diffusion
rate – pre-treatment and mechanical shredding – were evaluated, and both were effective. Polymer
matrices in pre-treated and shredded composites were homogeneously decomposed in 1 h.
Prospective catalytic conditions were screened to accelerate the chemical reaction rate for acid
digestion. The most effective catalyst was ScCl3, which reduced ~ 30% of the time required by
acid digestion with no catalyst.
Current chemical recycling of thermoset composites has been focused largely on
recovering high-value carbon fibers with property retention. However, recovery and reuse of
xvii
decomposed polymer matrix residues is rarely considered, despite the fact that doing so constitutes
an essential component of a sustainable approach to the problem. Therefore, this thesis also
investigates the viability of recovery and reuse of the decomposed amine/epoxy residue after acid
digestion of the matrix, effectively closing the recycling loop. Findings indicated that polymer
matrix residues recovered from acid digestion solutions via neutralization and evaporation
contained molecular components of the epoxies in which aromatic regions were preserved. The
recovered matrix residues were blended into virgin resin formulations and two approaches were
evaluate for potential reuse. Approach I utilized the matrix residue as an accelerator for a virgin
anhydride/epoxy formulation that contained no accelerator and thus could not be self-catalyzed.
This study discovered that adding matrix residue produced catalytic effects on the curing reaction.
In general, anhydride/epoxy samples blended with recovered matrix residues and cured were
homogenous and exhibited thermal and mechanical properties comparable to specimens cured with
a commercial accelerator. Approach II deployed the matrix residue as a filler for both virgin
anhydride- (with a commercial accelerator) and amine-based epoxies to produce blended
formulations. In such cases, blended formulations yielded acceptable retention of thermal
properties, provided the fraction of matrix residue added was < 20 wt%.
Besides developing effective chemical recycling method for thermoset composites, the
other focus of the thesis is sustainable manufacturing. Complex structures manufactured using
low-pressure vacuum bag-only (VBO) prepreg processing are more susceptible to defects than flat
laminates due to complex compaction conditions present at sharp corners. Consequently, effective
defect mitigation strategies are required to produce structural parts. This study investigated the
relationships between laminate properties, processing conditions, mold designs and part quality in
order to develop science-based guidelines for the manufacture of complex parts. Generic laminates
consisting of a central corner and two flanges were fabricated in a multi-part study that considered
xviii
variation in corner angle and local curvature radius, the applied pressure during layup and cure,
and the prepreg material and laminate thickness. The manufactured parts were analyzed in terms of
microstructural fiber bed and resin distribution, thickness variation, and void content. The results
indicated that defects observed in corner laminates were influenced by both mold design and
processing conditions, and that optimal combinations of these factors can mitigate defects and
improve quality.
1
CHAPTER 1. Introduction
Carbon fiber-reinforced polymer (CFRP) composites comprise lightweight structures with
high specific mechanical properties, excellent fatigue life, thermal and chemical stability. As a
result, CFRPs have attracted interests as structural materials and their market is expected to expand
at a considerable pace over the next few years. CFRPs first emerged in aviation sector in the 1980s
and they are traditionally used in the aerospace and defense industry. Currently, carbon fibers have
made inroads into commercial applications such as automotive parts, wind turbine blades and
energy storage tanks because of the reduced price for carbon fibers and the high demand for
lightweight structures. The global market size for carbon fiber-reinforced polymer composites is
anticipated to reach 146,000 ton/yr in 2020, consuming 89,000 tons of carbon fibers [1].
The use of CFRPs in automotive shows the greatest growth potential, driven by the
industry’s demand for lightweight components, greater fuel efficiency and lower emissions than
structures made from tradition materials such as steel and aluminium. Generally, composites can
achieve 60% weight saving compared to steel and 40% with respect to aluminum [2]. For example,
Ford Motor Company demonstrated that a prototype carbon fiber hood weighs more than 50% less
than a standard steel version. According to the U.S. Department of Energy, reducing a vehicle’s
weight by 10% can improve its fuel efficiency by 6-8% [3]. Another analysis suggests that a 10%
reduction in vehicle weight is associated with an 8% reduction of CO2 emissions [4]. Weight
reduction can also improve the freight efficiency of heavy-duty vehicles. For example, carbon fiber
composite trailers manufactured by Hexagon Lincoln can deliver more than double the amount of
hydrogen that can be carried in traditional steel tube trailers.
2
Figure 1-1. Global carbon fiber market volume share by application, 2016 (%) [5]
CFRPs are also increasingly used in wind turbine blades as they grow larger and more
powerful for economic reasons. According to the Global Wind Energy Council (GWEC), since the
1980s, wind turbine blades have increased eightfold, surpassing 60 meters in length, to harvest
more energy as a green solution to meet global energy demand [6]. Lightweight CFRPs that have
high strength and stiffness, as well as great fatigue and corrosion resistance, are far superior
compared to traditional aluminum and steel parts.
The increasing use of CFRPs has raised concerned about the environmental impacts of this
material, including effective recycling and sustainable manufacturing. Due to the lack of effective
recycling method, most composite waste is not recycled/treated and ends up in landfills. Without a
robust and effective method to recycle composite and complete the material cycle, composite
materials will face difficulty replacing metals in aerospace, automotive, energy and other
industries. Furthermore, CFRPs are traditionally manufactured using a pressurized autoclave
environment to suppress defects, which requires high energy and infrastructural cost and limits the
3
use of CFRPs in sectors requiring mass production. An effective alternative must be developed to
autoclave cure.
1.1. Challenges for Composites
1.1.1. Effective Recycling
Because of the increasing demand for lightweight structures in aerospace, automotive, and
wind energy industries, a major portion of the carbon fibers produced will be consumed in the
production of airplanes and automobiles. The lifespan of a commercial airplane (20-30 years), and
the average life expectancy of a new car (less than 10 years), pose needs and challenges for
recycling. Carbon fibers from end-of-life composites retain properties nearly equivalent to virgin
fibers, yet few are recovered and/or reused due to a lack of viable recycling technologies. This
absence of recovery/recycling is especially true for thermoset composites that undergo irreversible
cure reactions.
Current industrial scale recycling focuses on two physical technologies: mechanical
recycling and pyrolysis. Mechanical recycling produces fillers or reinforcements for lower grade
composites, but the fibers are not separated from the matrix, and little value is recovered. Pyrolysis,
on the other hand, employs heat to decompose the polymer matrices, thus separating the fibers.
However, pyrolysis degrades fibers and leaves char residues on fiber surfaces. Fibers recovered
from pyrolysis are generally short and are thus deployed in molding compounds. Thus, most of the
value of the continuous fibers/fabrics is lost.
4
Figure 1-2. Closing the CFRP lifecycle loop: carbon fiber reclaimed from end-of-life sources are
repurposed for use in automotive applications [8]
Unfortunately, a truly sustainable recycling method must recover valuable recyclates and
fibers, have scale-up potential, and be economically feasible, and no such method has been
demonstrated to date. Thus, only 2% of composites-related companies are active recyclers [7]. For
example, MIT-RCF (now Carbon Conversion) aims to close the CFRP lifecycle loop using
pyrolysis. Carbon fiber reclaimed from end-of-life sources are repurposed for use in automotive
composites applications [8]. Boeing has been working with a number of third-party technology
firms for the past several years on the recycling of aerospace grade composites [9].
Chemical recycling is widely viewed as a promising waste management strategy for end-
of-life thermoset composites. Relatively low temperatures are required compared to pyrolysis, and
clean fibers can be recovered with nearly full retention of mechanical properties. In addition, high-
value products from the chemical feedstock can also be recovered after additional treatment. Over
the past 20 years, chemical recycling has drawn research attention, and discoveries are being made
5
in recycling processes, reaction mechanism identification, catalyst development, and resin
chemistry.
To date, chemical recycling methods have not been employed on an industrial scale
recycling due to 1) the high capital costs associated with high temperature and high pressure
conditions, 2) the slow recycling rates and corrosive reaction media, and/or 3) a lack of post-
treatment methods for chemical solutions after recycling. There is an urgent need for effective low-
pressure methods that require less corrosive reaction media and are readily scalable, which
motivates us to investigate and develop chemical recycling methods that can be performed under
atmospheric pressure and low temperature (< 200 °C) for composites with amine-cured epoxy
matrices (the most widely used polymer matrix in advanced composites [10]).
With an effective recycling method to recover high quality fiber arrays/fabrics, carbon
fibers – both virgin and recycled fibers – will achieve more widespread adoption in production of
future commercial products. For example, Dell Technologies says that it uses approximately 1
million pounds of recycled carbon fiber in 2017, more than double the volume from the previous
two years [11]. Boeing estimates that recycling carbon fiber can be done at approximately 70% of
the cost and using less than 5% of the electricity required to make new carbon fiber (Table 1-1)
[9].
Table 1-1. Comparing virgin and recycled carbon fibers [9]
6
1.1.2. Sustainable Manufacturing
High performance composites parts often require autoclave processing which bring not
only high energy and infrastructural cost, but also the size limit problem. Out-of-autoclave (OoA)
prepreg processing using vacuum bag only (VBO) consolidation was developed as an alternative to
autoclave processing to reduce energy costs and infrastructural investments. While this process can
easily produce void-free flat parts, the precise manufacturing of complex geometries is challenging
because of the relatively low consolidation pressure, which makes it difficult to control flow- and
gas- induced porosity. Thus OoA-VBO laminates with corner geometries are more likely to
contain defects such as macro- and micro- voids, corner thickness variations, and micro-cracks
compared to flat laminates, reducing part quality and mechanical performance. Feasible strategies
must be developed to address these issues. Furthermore, the influence of geometry, material, and
processing on the mechanisms of defect formation also need to be better understood.
1.2. Objectives and Structure
The thesis covers two projects: Project I chemical recycling for amine/epoxy composites at
atmospheric pressure and Project II sustainable manufacturing for composites with complex
shapes, but focuses primarily on Project I. The polymer matrix for chemical recycling is amine-
cured epoxies, the most widely used polymer matrix in advanced composites. Unlike industrial-
scale physical recycling methods that can accommodate composites with different matrices and
fibers by a single process, chemical recycling, especially performed at atmospheric pressure must
be tuned to specific matrix formulations. To develop an effective chemical recycling method
requires a thorough understanding of the current research status, critical issues that must be
addressed, correlations between material properties and recyclability, rate-controlling factors for
recycling, reaction mechanism identification and catalyst development. Project II discusses OoA
7
prepreg processing for composite parts with different corner geometries. Defects in the parts are
identified and categorized, and feasible strategies are developed to address these issues. Details of
each Chapter are outlined below.
Chapter 2 presents the current and future markets for CFRPs and explain the justification
for composites recycling, including environmental legislations and economic drivers. The
challenges for thermoset composites recycling are described, followed by the limitations of current
industrial-scale recycling technologies that includes mechanical recycling and thermal processing,
and the development of chemical recycling that has drawn recent research attention. Chemical
recycling technologies that employ high pressure/temperature systems, as well as atmospheric
pressure/moderate temperature systems are discussed.
Chapters 3-5 lay out the main body of the thesis, presenting a comprehensive study of
chemical recycling of amine/epoxy composites at atmospheric pressure and moderate temperature
to recovers both high quality fibers and useful polymer products (Figure 1-3) [12-19]. Chapter 3
focuses on neat polymer dissolution of amine-cured epoxies [12]. Complete and effective
dissolution of the epoxy matrix is essential to separate the fiber reinforcement from the cured
composite, and must be evaluated prior to recycling actual composite parts. A parametric study of
the epoxy dissolution rate is investigated that varies in chemical recycling methods and the nature
of the polymer matrix (amine curing agent amount/type and epoxy functionality). Rate-controlling
factors for epoxy dissolution are identified and the decomposed epoxies are characterized.
8
Figure 1-3. Recovering near-virgin quality fibers and useful polymers from amine/epoxy
composites using atmospheric pressure and moderate temperature
Chapter 4 explores the viability of applying the recycling methods studied in Chapter 3 to
amine/epoxy composites with various matrix crosslink density and fiber reinforcement type
(carbon and glass fibers) [13-14]. Clean fibers in their original fibric forms are recovered using a
two-step process that first physically permeabilize the composites without disarranging the fibers
embedded in the polymer matrix, and second, chemically dissolve the matrix into the solvent under
mild conditions. Surface and mechanical property characterization are performed for these
recovered fibers. Furthermore, reaction mechanisms are identified and target catalysts to accelerate
each reaction step are developed.
Chapter 5 address environmental concerns that arise from the use of reagents in chemical
recycling that must be properly post-treated to prevent pollution [15]. Decomposed polymers are
recovered from the solution via neutralization and precipitation, and then reused in virgin epoxy
formulations as an accelerator or a filler. Data show that the decomposed polymer residue has
catalytic effects on the curing reaction of anhydride/epoxy systems, generating cured epoxies with
thermal and mechanical properties comparable to epoxies cured with a commercial accelerator.
9
Chapter 6 diverts the research focus to sustainable manufacturing of composites with
complex shapes using OoA prepreg processing [20]. It aims to expand the existing knowledge of
OoA processing by clarifying the compaction mechanisms that lead to defects in complex shape
laminates, and to form the basis for practical guidelines for the selection of mold designs and
processing parameters that mitigate defects and improve part quality.
Chapter 7 highlights the key takeaways form the work presented in this thesis and discusses
the broader impaction of these results and discoveries. The technology gaps between current lab-
scale systems and future industrial-scale operations for chemical recycling are identified, providing
insights to encourage the increased effectiveness in recycling. Lately, suggested future work for
the recycling project is outlined.
10
CHAPTER 2. Composites Recycling – Current Technologies
Thermoset composites exhibit superior mechanical properties, good resistance to heat and
corrosion, low shrinkage upon cure, and low moisture absorption. Most of these attributes derive
from the thermoset nature of the matrix. After cure, thermoset matrices have cross-linked, three-
dimensional structures and are generally insoluble and inert under mild-to-severe conditions [21-
22]. Thus, separating the fiber reinforcements from matrices poses challenges for composite
recycling.
Historically and currently, for lack of effective composites recycling methods, CFRPs at
end-of-life are commonly disposed in landfills without treatment or incinerated as energy sources.
Landfilling is a relatively cheap and easy way to dispose of CFRPs. But it is the least preferred
disposal route as CFRPs are non-biodegradable and have infinite life in the environment.
Incineration is another waste disposal route, which requires a great amount of heat and generates
polluting emissions. After incineration, most of the material converts to ash, fiber residue, and
other solid waste, which is sent to landfills as well. Both landfill and incineration are not
environmentally friendly, and most of the economic value of the materials is not recycled. A report
from Boeing points out that without an effective airplane recycling program, operators are unaware
of the value of recovered materials from the retired airplane parked in the desert (Figure 2-1).
Boeing believes that airplanes could be recycled in a way that offered both economic advantages to
operators and environmental benefits [9].
11
Figure 2-1. A plane graveyard in in a Californian desert (Picture: Jassen Todorov/Caters)
Composite recycling is gaining more attention also because the environmental legislation
becomes stricter, and many countries are expected to enact laws regulating composite waste in the
coming years, urging the composites industry to seek effective recycling solutions to the waste
problem. European legislation has been passed forbidding landfill disposal of CFRPs. The End of
Life Vehicle Directive (2000/53/EC) requires re-use and recovery of at least 95% of the average
weight of a car by 2015 [23]. China has also formulated the Technical Policy for the Recovery of
Automobile Products to push forward the development of the system of discarding and recovery of
automobile products in China [24]. Its goal is to reach the recoverability rate of 95% of all
automobiles, of which the recyclability rate of materials shall not be less than 85% from 2017.
Currently, three routes of recycling methods have been extensively studied: mechanical
grinding, thermal processing, and chemical treatment [25-26]. An ideal composite recycling
approach should feature ease of recovery, process robustness, recovery of fibers with near-virgin
quality, low overall cost, high throughput, low gaseous emissions, and little hazardous waste. All
these features may not be present in a real process because of multiple problems concerning
12
material properties of thermoset composites, such as good thermal and chemical stability.
However, some of the necessary attributes for effective recycling are being incorporated into
composite recycling methods, as outlined next.
2.1. Mechanical Recycling
Mechanical grinding has been investigated for both glass and carbon fiber-reinforced
composites, although most reported research has dealt with glass fiber composites [27-29].
Mechanical grinding involves the use of cutting, crushing or other mechanical processes to reduce
the CFRPs into smaller pieces (Figure 2-2). After grinding, typically the finer fractions are
powders that contain a higher proportion of the polymer matrix and the coarser fractions tend to be
particles with a higher fiber content. Finer pieces are separated from the coarser ones as they have
different applications. Resin-rich powders can be burned as energy source, while fibers of various
lengths with resin can be used as fillers or reinforcement for lower grade composites.
Figure 2-2. Schematic of cross-section of Wittmann mechanical recycling machine [28]
13
The process of mechanical grinding is simple, but clean fibers of sufficient length generally
cannot be recovered, and little value is recovered from the composites waste. Also, dust polluting
can occur during mechanical grinding and cause negative impacts on the environment. Lastly,
there is not the same demand for grinded composite as for materials such as steel and aluminum
because the incorporation level of filler or reinforcement material is very limited, making
mechanical grinding unfavorable to the recycling industry due to the low value of the recovered
products compared to the facility and energy costs incurred.
2.2. Thermal Processing
Polymer matrices are organic materials and almost all of them can be burned as a source of
energy if the temperature is high enough. Thermal processing relies on heat to degrade polymer
matrices into lower molecular weight components. Those products can be volatilized into gases,
after which fibers can be separated and recovered. The decomposed lower molecular weight
organic components can be burned in a combustion chamber, allowing the energy to be reused as
supplementary energy source for thermal processing. Two thermal processing method are being
used currently, including fluidized-bed process and pyrolysis [30-33]. Fluidized-bed process is
generally fluidized by hot air to enable rapid heating of the materials, and pyrolysis, which
operates in presence or in absence of oxygen. Fluidized-bed process requires hot air passing
through a solid particle substance (silica sand) at typical fluidizing velocities to cause the solid
composites materials to behave as a fluid, which brings tremendous damage to the fibers in
addition to the high temperature [31].
14
Figure 2-3. Fluidised bed process schematic [31]
Pyrolysis operates in absence of oxygen to minimize the formation of char on the fiber
surface and is now the most widely used recycling process for industrial scale recycling (Figure 2-
4) [32-33]. However, pyrolysis generally requires high temperatures (450 to 700 °C) depending on
the properties of the matrix, which can lead to fiber degradation and entangling. Reported data
showed that during pyrolysis of CF composites at 500 °C, there was negligible degradation of fiber
tensile strength. However, pyrolysis above 600 °C caused severe oxidation of the carbon fiber and
the tensile strength of the fibers declined by > 30% [32]. Furthermore, thermally degraded matrix
(generally epoxy) often leaves residual char on fiber surfaces, even in the absence of oxygen, and
fiber purity was less than the desired level of 99.5% [30], which reduces the interfacial bond
strength between recycled fibers and new matrix systems.
15
Figure 2-4. The Thermolyze
system in Forst, Germany used for the bulk pyrolysis of the studied
composite materials [33]
Lastly, high temperature also raises the cost for recycling a material that is considered a
waste product. Fibers recovered from thermal processing are generally chopped into short fibers
for use in molding compounds and other applications, and most of the value of the continuous
fibers/fabrics is lost. Due to the limited market demand for recycled short fibers, most composite
materials are not being recycled today. Consequently, recent effort has focused on developing
chemical recycling approaches that recover residue-free fibers with high tensile strength retention.
2.3. Chemical Recycling
Chemical recycling is a relatively new recycling method that uses a wide range of solvents
and catalysts as reactive medium to decompose polymer matrix into oligomers or monomers under
supercritical or atmospheric pressure. Relatively low temperature compared to pyrolysis is
required to decompose the resin matrix. By separating the resin matrix from the fiber
reinforcement using chemical treatment, carbon fiber can be recovered from the waste in its intact
form without char residues, as well as valuable resins from the chemical feedstock after additional
16
treatment. Thus, most of the economic value of the materials can be recovered. In chemical
treatment, it is possible to adjust the properties of the reaction medium and control the reaction by
changing solvent, catalyst type, temperature and pressure. Thus, among all recycling methods,
chemical treatment has received the most research attention. After chemical recycling, carbon
fibers from end-of-life composites are clean (99.5% purity) [30] and are generally able to retain
unchanged elastic modulus and 90% of the ultimate tensile strength [34-35], making the recovered
carbon fiber possible to be reused in new composites parts. As a result, the price of carbon fibers
will be further reduced, making carbon fiber more viable and affordable to the manufactures,
leading to tremendous fuel consumption saving and greenhouse gas emission reduction.
2.3.1. Recycling of Composites in Near- and Supercritical Fluids
Traditionally, supercritical fluid has been used for recycling due to the unique interactions
achieved between gas and liquid phases in the reaction medium, especially the high mass transfer
coefficient and diffusivity. High temperature (> 250 °C) and high pressure (> 5 MPa) are required
to reach near- and supercritical conditions due to hydrogen bonds in the solution [26,34]. Although
specific reaction conditions depend on the nature of the matrix, a supercritical fluid process that
performs at a high temperature are closer to a thermal process than a chemical recycling process.
Commonly used solvents include water [36-41], methanol [42-45], ethanol [45], propanol
[45-48]. For example, Okajima et al. attempted recycling of bisphenol A type epoxies cured with
phthalic anhydride using supercritical methanol at 250-350 °C and 10 MPa [42]. They reported
that the crosslinked structure of the epoxy decomposed and dissolved in supercritical methanol at >
270 °C within 1 h. When the temperature decreased to 250 °C, the epoxy dissolution time
increased to 2 h. Hyde et al. used supercritical propanol as the solvent fluid to remove bisphenol
A-type epoxies cured with amines from composites, using conditions above 450 °C and above 5
17
MPa [46]. Results showed little damage to the fibers, and the recovered fibers retained 95% of the
tensile strength of virgin fibers, despite the severe conditions. Solvent mixtures have also been
investigated, including acetone/water [49], ethanol/water [50], benzyl alcohol/water [51], and
ketone/water [52] systems.
Alkaline catalysts, such as sodium hydroxide (NaOH) and potassium hydroxide (KOH), are
often added to accelerate the reaction rate and reduce the reaction temperatures [53-56]. In
addition, various metal salts were employed as catalysts to cleave the C–N bonds in epoxy
matrices, including AlCl3, FeCl3, ZnCl2, CnCl2, MgCl2 [57-59]. Liu et al. investigated a chemical
recycling method for CFRPs with high crosslink density (Tg = ~210 °C) using a ZnCl2/ethanol
system at 180-220 °C [59]. The CFRPs used were manufacturing scraps, and the main composition
of the matrix was amine-cured epoxy. Results showed that a matrix degradation degree of ~90%
could be achieved.
Figure 2-5. Process flow of grinding / subcritical water hydrolysis / inorganic materials separation
[41]
18
However, the aggressive conditions inevitably led to fiber degradation, and the tensile
strength of recovered fibers was reduced by 1-15% [36]. Furthermore, under supercritical
conditions, reactors made of special alloys are required to withstand the high pressures and
chemical corrosion. Thus, the capital costs associated with supercritical chemical treatment are
generally high. Consequently, chemical treatment that can be performed under atmospheric
pressure and low temperature (< 200 °C) has emerged as a research focus in thermoset composite
recycling.
2.3.2. Recycling of Composites at Atmospheric Pressure
Chemical treatment methods for recycling thermoset composites at atmospheric pressure
have been reported, often using an acidic medium, such as nitric [60-62] or sulfuric acid [63-64].
However, strong acid solutions pose environmental hazards, especially nitric acid that is a highly
toxic acid and must be disposed of in a proper manner. Thus, atmospheric chemical treatment
methods that require less corrosive reaction media and are easier to handle have drawn attention
recently. For example, Hitachi Chemical reported a solvolytic depolymerization method conducted
at atmospheric pressure for anhydride/epoxy systems, using chemical solutions comprised of alkali
metal salt (catalyst) and high boiling point alcohol (solvent) [65-67]. The method was
demonstrated for anhydride-cured matrices, and thus, the viability of the process for amine-cured
epoxies, which are widely used in high performance composites, is unlikely. Amine/epoxy
formulations lack ester groups for transesterification, a reaction essential to the Hitachi technology.
Thus, this approach cannot be expected to effectively dissolve amine/epoxy matrices.
Chemical dissolution of amine-cured epoxy formulations requires a different approach.
Oxidative treatment methods using a mixed solution of hydrogen peroxide and organic solvents,
such as N,N-dimethylformamide [68-69] and mild acid [70-71], were reported to dissolve
19
amine/epoxy matrices. Xu et al. investigated an oxidative treatment, using a mixed solution of
hydrogen peroxide and N,N-dimethylformamide as the reaction medium to dissolve amine/epoxy
matrices with an epoxy value of 0.54–0.57 (bi-functional resin) [68]. Results showed that clean
fibers were recovered, and the recovered fibers retained > 95% tensile strength. Most chemical
recycling methods focused on amine-cured bisphenol A-type epoxies, which featured relatively
low crosslink density and glass transition temperature (Tg). However, the dissolution of
amine/epoxy matrices with higher crosslink density is more challenging, and doing so in
atmospheric pressure environment has not been considered.
The long-term goal of the recycling project is to develop a catalytic method for cleavage of
highly crosslinked amine/epoxy matrices using moderate conditions (atmospheric pressure and
moderate temperature) and safe chemicals. These features are critical to practical, large-scale
composite recycling to recover long continuous fibers and useful chemical components from
aerospace, and high- and low-grade industrial composites, but have not been reported to date.
20
CHAPTER 3. Chemical Treatment for Dissolution of Amine-cured Epoxies at
Atmospheric Pressure
3.1. Introduction
Because of the absence of effective methods to separate carbon fibers from thermoset
matrices, end-of-life composites are commonly 1) disposed in landfills without treatment, 2)
incinerated for energy, because polymer matrices can be burned at sufficiently high temperature
[21-22], or 3) mechanically ground to produce fillers or reinforcement for lower grade composites
[27-29]. After incineration, most of the material converts to ash, fiber residue, and other solid
waste, which is sent to landfills as well. Mechanical grinding has been investigated for both glass
and carbon fiber-reinforced composites, although most reported research has dealt with glass fiber
composites. The process of mechanical grinding is simple, but clean fibers of sufficient length
generally cannot be recovered, and little value is recovered from the composites waste. Effective
recycling methods are thus needed to make composite materials more sustainable.
Pyrolysis is now the most widely used recycling process for industrial scale recycling.
Pyrolysis relies on heat to degrade polymer matrices into lower molecular weight components.
However, high temperature thermal processing causes severe oxidation of the carbon fiber and the
tensile strength of the fibers declined by > 30% [32]. Furthermore, thermally degraded matrix
(generally epoxy) often leaves residual char on fiber surfaces, even in the absence of oxygen,
which reduces the interfacial bond strength between recycled fibers and new matrix systems.
In chemical treatment, a wide range of solvents are used as a reactive medium with a
catalyst or reagent to decompose the matrix into oligomers or monomers, under supercritical or
atmospheric pressure. Relatively low temperature is required (compared to pyrolysis) for matrix
dissolution, and clean, intact, carbon fibers can be recovered which retain 100% of the elastic
21
modulus and 95% of the ultimate tensile strength compared to virgin carbon fibers [26]. Valuable
products from the chemical feedstock can also be reused after additional treatment. Consequently,
among all recycling approaches for thermoset composites, most research attention has focused on
chemical recycling. For example, supercritical fluids have been evaluated for recycling due to the
high mass transfer coefficient and diffusivity [34]. High temperature and high pressure were used
in experiments to recover fibers from composites, which inevitably led to significant fiber
degradation. Furthermore, the high capital costs associated with supercritical recycling raises
challenges for scale-up.
The present study focuses on chemical treatment methods at (1) atmospheric pressure and
(2) low temperature (< 200 °C) for amine-cured epoxy formulations with highly crosslinked
networks. These features are critical to practical, large-scale composite recycling, but have not
been reported to date. The investigation focuses on amine-cured epoxies, which is the most widely
used polymer matrix in high-performance composites. Two methods - depolymerization and acid
digestion - are applied to amine-cured epoxies with high crosslink densities. Data show both
depolymerization and acid digestion routes can achieve 100% resin dissolution in all amine/epoxy
samples. Correlations between epoxy formulation and dissolution reveal that the dissolution rate is
influenced not only by the chemical reaction rate, but also greatly by the diffusion rate, which has
not been reported previously. Protocols are developed using NMR spectroscopy to characterize the
components of the chemical solutions. This study shows that the two major cleavable sites during
epoxy dissolution are C-N and C-O bonds. During dissolution, cured polymers are first cleaved
into molecular chains with various lengths, which further react to yield oligomers or monomers.
These discoveries of molecular level mechanisms of dissolution provide insights to inform further
studies on amine/epoxy composite recycling optimization and development of large-scale
recycling processes.
22
3.2. Experiments
Amine/epoxy resins were formulated (without toughening agents or other additives) to
analyze cured epoxy dissolution by chemical treatment. The resin formations, relatively simple
compared to commercial epoxy formulations, provided less complex crosslinked structures,
allowing identification of the chemical reactions involved in dissolution. Thermal analyses,
including thermogravimetric analysis (TGA, TA Instruments Q5000 IR) and modulated
differential scanning calorimetry (MDSC, TA Instruments Q2000), were used to measure the
thermal properties of the epoxies. Cured samples were subjected to (a) depolymerization and (b)
acid digestion. One- and two-dimensional proton NMR spectroscopy (1D and 2D NMR) was used
to identify the components of the chemical solutions after dissolution.
3.2.1. Resin Formulation
Resins were formulated using three types of epoxy monomers that varied in epoxy
functionality. The first, bi-functional epoxy, was a diglycidyl ether of bisphenol A (DGEBA,
Araldite GY 6010, Huntsman Corporation). DGEBA was a medium viscosity liquid epoxy with a
chemical structure as shown in Figure 3-1 (a). The second was tri-functional epoxy, a triglycidyl of
para-aminophenol, with medium functionality and low viscosity (Araldite MY 0510, Huntsman
Corporation, Figure 3-1 (b)). The third was tetra-functional epoxy - tetraglycidyl-4,4'-
methylenebisbenzenamine - that featured high functionality and yielded systems with good thermal
stability and mechanical performance (Figure 3-1 (c)). Epoxy equivalent weight (EEW) was used
to calculate the A/E ratio for amine curing agents and epoxy resins. The EEWs of bi-, tri- and
tetra-functional epoxy monomers were 187, 101 and 113 g/eq, respectively. 1D HNMR spectra of
bi-, tri- and tetra-functional epoxy monomers are shown in Supplementary Figure A-4, Figure A-5
and Figure A-6, respectively.
23
Two types of tetra-functional, amine-based curing agent were selected to react with the
epoxy monomers. Type I curing agent was diamine 3,3’-diaminodiphenyl sulfone (3,3’-DDS,
Aradur 9719-1, Huntsman Corporation), a high-performance curing agent shown in Figure 3-1 (d).
Type II was 4,4’-methylenebis (2,6-diethylaniline) (M-DEA, Sigma-Aldrich), a component often
used as curing agent or chain extender (Figure 3-1 (e)). Amine hydrogen equivalent weight
(AHEW) was used in A/E ratio calculations, and the AHEWs of 3,3’-DDS and M-DEA were 63
and 78 g/eq, respectively. A 1D HNMR spectrum of 3,3’-DDS is shown in Supplementary Figure
A-3
During resin formulation, specific amounts of epoxy monomer and curing agent were
mixed at room temperature in clean aluminum cans until fully homogenized. The mixing A/E ratio
varied from 40% to 100%, providing a wide range of crosslinking densities. The mixture was then
heated to 120 °C in a pre-heated convection oven to further improve the mixing quality, yielding a
clear homogenous mixture.
Figure 3-1. Epoxy monomers: bi-(a), tri-(b), and tetra-(c) functional epoxies, and amine curing
agents: (d) 3,3’-DDS, and (e) M-DEA (red and blue represent oxygen (O) and nitrogen (N),
respectively)
24
3.2.2. Thermal Analysis
TGA tests were used to measure the degradation temperature of the epoxy samples. First,
TGA was performed on samples to obtain a correlation of sample weight loss as a function of
temperature. In each TGA test, a dynamic ramp was applied from 30 to
400 °C at a heating rate of
1.5 °C/min. From the TGA data, polymer degradation temperature was defined as the temperature
at which 5% sample weight loss was achieved. The degradation temperature set an upper limit for
further MDSC tests. All samples began to degrade at ~300 °C.
MDSC tests were performed on each sample to measure the glass transition temperature
(Tg) of the cured sample. For each MDSC measurement, a ramp was applied from -60 to 300 °C at
a constant rate of 1.5 °C/min with ±0.5 °C/minute modulation. After cure, samples were heated
from -60 to 300 °C at a rate of 10 °C/min with ±0.5 °C/minute temperature modulation. The Tg of
the cured epoxy was obtained from the inflection point of the last reversible heat flow signal
during the ramp cycle.
3.2.3. Amine-cured Epoxy Dissolution
Epoxy dissolution samples were prepared by curing specific amounts of resin in an oven
via the same temperature ramp at 1.5 °C/min to 250 °C, followed by a dwell for 0.5 hour. MDSC
measurements were performed again on oven-cured samples to confirm that no residual
exothermic reaction peaks existed and that the Tg’s of the oven cured samples were comparable to
that of the MDSC-cured samples. Cured samples were then dissolved by depolymerization and
acid digestion. Epoxy dissolution time was defined as the time required to achieve complete
dissolution in solvent (determined by visual observation), which marked the end of each
experiment. The standard deviation of dissolution time for samples with the same formulation was
25
below 5%, so one sample was tested for each condition. For consistency, sample concentration
used in all experiments was 10 mg/mL.
Approach I: Depolymerization. Depolymerization, (more accurately described as alkali
digestion, based on the reaction medium), was performed using a supersaturated solution of 100
mL benzyl alcohol (solvent, Sigma-Aldrich) and 7 g tripotassium phosphate (catalyst, Sigma-
Aldrich). Thus, a three-neck round-bottom flask (1 L) containing the depolymerization solution
and a weighed sample (1 g) was fluxed at 200 °C in an oil bath. Magnetic stir bars were placed in
both the oil bath and the flask to homogenize the heat and concentration distribution. Nitrogen
flow was employed to create an inert environment for the reaction to keep the catalyst in a reactive
form. After reaction, the homogenized chemical solution was stored in a glass bottle for further
analysis.
Approach II: Acid Digestion. Acid digestion (more accurately described as oxidative
digestion) was performed using a solution of 100 mL glacial acetic acid (solvent, Sigma-Aldrich)
and 10 mL hydrogen peroxide solution (oxidant, 30% (w/w) in H2O, Sigma-Aldrich). The same
setup as depolymerization was used in acid digestion. The resultant mixture and a weighed sample
(1 g) were refluxed at 110 °C, and additional hydrogen peroxide solution (30%, 5 mL) was added
to the flask every hour. No nitrogen flow was used for acid digestion.
Pre-treatment. Pre-treatment was employed as a strategy to permeabilize (swell) the cured
sample before dissolution, so that the rate-limiting effect of solvent diffusion could be reduced or
eliminated. During pre-treatment, cured samples were placed in benzyl alcohol at 200 °C for 3
hours, which was more than sufficient for the sample size (1 g). After pre-treatment, the epoxy
samples were softened and expanded compared to their original size. The soaking pre-treatment
allowed the solvent to penetrate the crosslinked network, enabling reactant molecules to
subsequently reach the cleavable bonds more easily, thus increasing dissolution rates.
26
3.2.4. Nuclear Magnetic Resonance Spectroscopy
NMR spectroscopy (Varian Mercury 400) is useful for characterizing small molecules and
polymers. 1D HNMR is the most sensitive NMR technique, well-suited to polymer analysis,
because the large macro-molecules in polymers are generally tangled and have repeat units. 2D
HNMR can provide information complementary to 1D HNMR in a manner that is more easily
interpreted, and is often used to confirm data from 1D HNMR. Chemical solutions from
depolymerization and acid digestion were analyzed in 1D and 2D HNMR. The liquid mixture was
first purified by removing solid impurities after letting the mixture standing overnight. Next, the
chemical solution (benzyl alcohol) was evaporated, and the dried sample was dissolved in
deuterated dimethyl sulfoxide until a clear solution was achieved for NMR analysis.
3.2.5. Test Matrix
Table 3-1 summarizes the five sets of dissolution experiments performed in this study. In
Set I, the viability of depolymerization and acid digestion for epoxy dissolution was investigated
using 3,3’-DDS-cured bi-functional samples. The selected A/E ratios promoted different reaction
kinetics and crosslinking densities in cured epoxies, thus affecting those properties related to
crosslinked network morphology. To investigate the effect of A/E ratio on the recyclability of
amine-cured epoxy, samples at A/E = 40%, 60%, 80%, and 100% were tested. In Set II, the effect
of diffusion was analyzed by applying pre-treatment to 3,3’-DDS-cured bi-functional epoxies at
A/E = 60%, 80%, and 100% before acid digestion. The Tg values for those samples were similar to
or greater than the reaction temperature of acid digestion, and the effect of diffusion was expected
to be non-negligible. In Set III, highly crosslinked samples (produced from 3,3’-DDS-cured tri-
and tetra-functional epoxies) were subjected to depolymerization and acid digestion to evaluate the
effect of crosslink density. Sets IV and V were used to determine the cleavable sites during epoxy
27
dissolution. By replacing 3,3’-DDS with M-DEA without sulfone functional groups, Sets IV and V
revealed the effects of sulfone functional groups on dissolution by depolymerization and acid
digestion.
Table 3-1. Test matrix for amine-cured epoxy dissolution
Set Amine Epoxy A/E Ratio Chemical Treatment
I 3,3’-DDS Bi-functional 40%, 60%, 80%, 100% Approach I and Approach II
II 3,3’-DDS Bi-functional 60%, 80%, 100% Pre-treatment + Approach II
III 3,3’-DDS
Tri-functional 40% Approach I and Approach II
Tetra-functional 40% Approach I and Approach II
IV M-DEA Bi-functional 40%, 60%, 80%, 100% Approach I and Approach II
V M-DEA
Tri-functional 40% Approach I and Approach II
Tetra-functional 40% Approach I and Approach II
28
3.3. Results and Discussion
3.3.1. Bi-functional Epoxy Dissolution
Figure 3-2 (a) shows that as the A/E ratio increases from 40% to 100%, the dissolution time
increases for both depolymerization and acid digestion processes. For a given A/E ratio (40%,
60%, and 80%), dissolution times for acid digestion are less than those for depolymerization,
indicating a faster chemical reaction rate for acid digestion. However, for bi-functional samples
with A/E = 100%, the dissolution times for depolymerization and acid digestion are comparable.
To explain this finding, the Tg’s of the cured epoxies are analyzed (Figure 3-2 (b)). As the
A/E ratio (in bi-functional samples) increases from 40% to 100%, the Tg’s of epoxies cured using
the same cure cycle increases from 60 to 160 °C in a quasi-linear manner. The Tg is the
temperature where the free volume available for chain movement has reached a minimum value,
and the internal mobility of the polymer chains starts to change markedly. Thus, the material
physically changes between the glassy state and the rubbery state. The increase in Tg represents the
decrease in free volume and the decrease in freedom of polymer chains to achieve different
physical conformations. For cured epoxy at A/E = 100%, the Tg (160 °C) is greater than the
reaction temperature for acid digestion (110 °C), so the mobility of the polymer chains is reduced
compared to samples with lower A/E ratios. The low mobility of the polymer chains, as well as the
decreased free volume, limits the diffusion of the solvent and reactant molecules into the
crosslinked network, and thus longer dissolution time is required. For depolymerization, the
reaction temperature (200 °C) is greater than the Tg’s of the samples, and the dissolution rate is
mainly limited by the chemical reaction rate of depolymerization.
29
Figure 3-2. Effect of A/E ratio on a) dissolution time and b) Tg’s for 3,3’-DDS-cured epoxies
I conclude from the above observations that both depolymerization and acid digestion are
effective for amine-cured epoxy dissolution. The dissolution rate is influenced by the chemical
reaction rate, and also by the diffusion rate. When the reaction temperature is less than the Tg of
the epoxy, the dissolution rate is mainly determined by the diffusion rate. However, in other cases,
the chemical reaction rate is the deciding factor, and the chemical reaction itself is faster for acid
digestion than for depolymerization.
3.3.2. Effect of Diffusion
To evaluate the effect of diffusion on dissolution rate, 3,3’-DDS-cured bi-functional
samples with A/E = 60%, 80%, 100% are pre-treated in benzyl alcohol. Figure 3-3 shows that pre-
treatment reduces the dissolution times for all the above formulations, even for the sample with
A/E = 60% that exhibits a Tg (95 °C) lower than the reaction temperature (110 °C). This finding
indicates that the limit from diffusion exists in all cured epoxies, though the extent may vary. The
sample with A/E = 100% shows the greatest reduction due to the highest crosslink density
compared to the other two amine/epoxy formulations. This observation supports the assertion that
30
the particularly long time required for dissolution of the sample with A/E = 100% (Figure 3-2 (a))
is due to slow diffusion.
Pre-treatment experiments demonstrate that the diffusion rate plays a key role in
determining epoxy dissolution rate if the reaction temperature is less than or comparable to the Tg’s
of the cured samples. Choice of appropriate pre-treatment solvents to better permeabilize the cured
epoxies with heavily crosslinked network can effectively reduce/overcome the limit from
diffusion.
3.3.3. Tri- and Tetra-functional Epoxy Dissolution
Tri- and tetra-functional epoxy monomers show end groups similar to bi-functional
monomers (Figure 3-1), but differ in molecular structure. Higher functionality resins also promote
more heavily crosslinked networks, which further limits diffusion of the solvent molecule. In this
section, polymerization and acid digestion of 3,3’-DDS cured tri- and tetra-functional samples with
A/E = 40% are analyzed. The results are compared to the data for bi-functional samples (A/E =
40%) to gain a more complete understanding of amine/epoxy dissolution.
Figure 3-3. Effect of pre-treatment on epoxy dissolution time
31
Figure 3-4 (a) shows the dissolution times of 3,3’-DDS cured bi-, tri-, and tetra-functional
epoxies with A/E = 40%, indicating that all samples can be dissolved via depolymerization and
acid digestion. Acid digestion again requires shorter dissolution time than depolymerization, which
is attributed to the faster reaction rate and is consistent with previous discussion. The MALDI-
TOF/MS data in Supplementary Figure A-7 show that the maximum molecular weight of the
dissolution products for tri-functional epoxy is 834 at 34 h into depolymerization, and 620 at 6 h
into acid digestion, confirming that the chemical reaction rate for acid digestion rate is faster than
depolymerization. For both recycling methods, the dissolution times for bi-functional epoxies are
less than 4 hours. But for tri- and tetra-functional epoxies, the dissolution times are more than ten
times longer than for bi-functional samples. This observation can also be attributed to the slower
diffusion rate in tri- and tetra-functional epoxies.
Figure 3-4 (b) shows the Tg values for bi-, tri- and tetra-functional epoxies prepared using
the same cure cycle, which are 50, 190, and 210 °C, respectively. For the bi-functional sample, the
Tg is lower than the temperature for both depolymerization (200 °C) and acid digestion (110 °C),
so the dissolution rate is determined primarily by the chemical reaction rate. However, for tri- and
tetra-functional samples, the corresponding Tg values are greater than the reaction temperature for
acid digestion, and are comparable to the reaction temperature for depolymerization. In addition,
tri- and tetra-functional epoxies contain more heavily crosslinked networks than bi-functional
epoxies. Thus, diffusion is rate-limiting for tri- and tetra-functional epoxy dissolution in both
chemical treatment methods, and the resultant dissolution rate is slower than for bi-functional
epoxies.
32
Figure 3-4. Effect of epoxy functionality on a) dissolution time and b) Tg values and numbers of
crosslinked bonds for 3,3’-DDS-cured epoxies
While cured tri-functional epoxies exhibit Tg values lower than tetra-functional epoxies
(Figure 3-4 (b)), signifying a less heavily crosslinked network, they require longer dissolution
times than tetra-functional epoxies for both depolymerization and acid digestion (Figure 3-4 (a)).
The finding can be explained as follows. There are three main types of reactions occurring during
the curing process of amine/epoxy resins (Figure 3-5) [72-75]. Reactions I and II are the primary
(
o
1) and secondary (
o
2) amine reactions with the epoxy groups, respectively. Reaction III is an
etherification reaction of the pendant hydroxyl groups formed during Reaction I with the epoxy
groups. The etherification reaction is reported to be negligible at low temperature, becoming
important only above 150 °C [72]. Reaction rate constants for Reaction II (
o
2 epoxy-amine) and
Reaction III (epoxy-hydroxyl) are reportedly equal and are about one tenth the rate constant of
Reaction I (
o
1 epoxy-amine) [73]. This study assumes that the number of C-O bonds formed
during cure is proportional to the number of C-N bonds, and the number of total crosslinked bonds
can be expressed as equation (1) to explain the difference in dissolution times for tri- and tetra-
functional epoxies.
Number of crosslinked bonds = (A/E ratio)/(EEW+A/E ratio ´ AHEW) (1)
33
where:
Number of crosslinked bonds: number of C-N bonds in crosslinking (mol/g)
A/E Ratio: Amine/Epoxy stoichiometric ratio (%)
EEW: Epoxy equivalent weight (g/mol)
AHEW: Amine hydrogen equivalent weight (g/mol)
The tri-functional epoxy monomer has a lower EEW (101 g/mol) than the tetra-functional
epoxy monomer (113 g/mol), and thus more crosslinked bonds exist in cured tri-functional
samples with the same weight (1 g). Due to the slow diffusion rate, samples with more crosslinked
bonds take longer to cleave and fully dissolve, so longer dissolution times are required for tri-
functional epoxies than for tetra-functional epoxies.
3.3.4. Cleavable Sites
While the reaction mechanism for dissolution of anhydride-cured epoxies is triggered by
transesterification [65-67], the mechanism for amine-cured epoxies is fundamentally different and
was not well understood. Chemical attack on the polymer matrix should occur at polarized carbon-
heteroatom linkages [76]. This study envisages three groups of possible cleavable sites in
amine/epoxy samples (Figure 3-5): (1) carbon-amine nitrogen bonds (C-N bonds) from Reactions I
and II [54, 61, 76], (2) ether bonds from Reaction III (C-O bonds) [76-77], and (3) sulfone
functional groups (SO2 groups) within the curing agents 3,3’-DDS [76]. To identify the cleavable
sites for depolymerization and acid digestion, this section analyzes the three possible cleavable
sites separately.
Sulfone functional groups. Cleavable Site I (SO2 groups) was investigated first. To evaluate
the effect of SO2 groups, M-DEA (Figure 3-1 (e)), which contains no sulfone functional groups,
was used as the amine curing agent (instead of 3,3’-DDS). Formulated resins were characterized,
34
cured and subjected to both depolymerization and acid digestion. Comparison of the Tg values of
M-DEA and DDS cured epoxies is provided in Supplementary Figure A-1. Figure 3-6 (a) shows
the dissolution times for bi-functional epoxies with A/E = 40%, 60%, 80% and 100%. The data
indicate that all M-DEA cured bi-functional epoxies are completely dissolved by both methods.
M-DEA-cured tri- and tetra-functional epoxies with A/E = 40% are also subjected to
depolymerization and acid digestion. Figure 3-6 (b) shows the dissolution times for tri- and tetra-
functional samples with A/E = 40%, which confirms that depolymerization and acid digestion
processes dissolve cured epoxies with high functionality that contain no sulfone functional groups.
The data demonstrate that the absence of SO2 groups does not affect the dissolution properties, and
the SO2 groups are not the dominant cleavable sites for depolymerization or acid digestion.
Figure 3-5. Possible cleavable sites in amine-cured epoxies
35
Figure 3-6. Dissolution times for M-DEA-cured epoxies: a) effect of A/E ratio (bi-functional
epoxy), b) effect of epoxy functionality (A/E = 40%)
C-N and C-O bonds. Possible cleavable Sites II (C-N bonds) and III (C-O bonds) were
considered next. 1D and 2D HNMR were performed to analyze the chemical components after
epoxy dissolution. Signals from the 4-6 ppm region were analyzed, as multiple signals in the
aromatic region produced overlapping peaks. Figure 3-7 (a) shows four major peaks in the 4-6
ppm region from the depolymerized solution of tri-functional epoxies. The peak at 4.47 ppm
corresponds to the secondary carbon hydrogen of the benzyl alcohol solvent residue, and the peaks
at 4.51, 4.57 and 5.75 ppm correspond to the epoxy components in the chemical solutions.
2D HNMR (gCOSY, Figure 3-7 (b)) shows J coupling between peaks 4.57 and 5.75 ppm.
Depolymerization is characterized by breaking the polymer chain backbone, so that after
dissolution, the products are similar to the parent material, yet the crosslinked structures are still
distinguishable [76]. Therefore, the peaks at 4.57 and 5.75 ppm correspond to the protons of the
unreacted crosslinked units. The peak at 4.51 ppm does not show any correlation to other peaks,
indicating that the peak corresponds to protons in the cleaved bonds.
36
Figure 3-7. Qualitative a) 1D and b) 2D HNMR spectrum analyses of dissolution products (26 h
depolymerization) of tri-functional epoxy in DMSO at room temperature (32 scans)
To confirm the assertion above, the NMR data were analyzed quantitatively. The plot in
Figure 3-8 (a) shows that as reaction time increases, the peak intensity at 4.51 ppm increases,
using the peak at 4.57 ppm as a reference. Integrating the area under the peak, which represents the
number of the corresponding protons, the number of protons at 4.51 ppm increases with time.
Cleaved bonds increase as reaction continues, indicating that the peak at 4.51 ppm corresponds to
protons in the cleaved bonds. Figure 3-8 (b) shows the ratio of the protons at 5.75 and 4.57 ppm.
The ratio remains constant (~ 0.25) during depolymerization for all samples, providing further
37
evidence that the two correlated peaks (5.75 and 4.57 ppm) correspond to the crosslinked
structures in the cured epoxies.
To identify the specific structure of the crosslinking, details of the crosslinked network are
analyzed. Recall that in bi- and tetra-functional epoxy monomers, epoxy functional groups are
attached to oxygen atoms and nitrogen atoms, respectively (see Figure 3-1). Tri-functional epoxy
monomers contain epoxy functional groups attached to both oxygen and nitrogen atoms, and thus
can be considered as a mixture of bi- and tetra-functional epoxy monomers. This assertion is
supported by the NMR spectra of epoxy monomers shown in Supplementary Figure A-4, Figure
A-5 and Figure A-6 At A/E = 40%, an excess of epoxy groups exists in the resin relative to amine
curing agents. Thus, both Reactions II and III occur at later stages of cure, before the epoxy groups
are depleted to form C-N and C-O bonds. Figure 3-9 color marks the crosslinking of cured bi- and
tetra-functional epoxies based on epoxy monomer end groups (orange and blue represent the
crosslinked units from epoxy-amine and epoxy-hydroxyl reactions, respectively). Note that a
similar crosslinked unit structure (Figure 3-9 (a)) exists in all systems, regardless of epoxy
functionality (bi-, tri, and tetra-functional epoxy), curing agent type (3,3’-DDS and M-DEA), or
reaction type (Reactions I, II and III).
38
Figure 3-8. Quantitative 1D HNMR spectrum analyses of dissolution products of tri-functional
This crosslinked unit contains three carbon atoms in the middle (one 3
o
carbon and two 2
o
carbons), with an oxygen atom attached to the 3
o
carbon atom, and one oxygen (or nitrogen) atom
attached to each of the 2
o
carbon atoms. The oxygen atom attached to the 3
o
carbon atom can be
bonded to hydroxyl groups, due to the slow etherification reaction at temperature < 150 °C. The
ratio of the hydroxyl groups attached to the 3
o
carbon hydrogens and the 2
o
carbon hydrogens is
1:4, which is consistent with the constant proton ratio at 5.75 and 4.57 ppm. Thus, the peaks at
5.75 and 4.57 ppm correspond to the hydroxyl groups attached to the 3
o
carbon hydrogens and to
the 2
o
carbon hydrogens, respectively.
Table 3-2 summarizes the possible cleavable crosslinked units in cured bi- and tetra-
functional epoxies, based on the reoccurring crosslinked units from Reactions I, II and III in
(Figure 3-9 (a)). The functionalities (f’s) of these cleavable units vary from 3 – 5. In bi-functional
epoxies, some crosslinked units form in Reaction III (shaded gray) that contain only C-O bonds (f
= 3). Without breaking the C-O bonds, the shaded units for bi-functional epoxies remain intact (f =
3), and a less extensively crosslinked network is formed (even if the C-N bonds are cleaved),
preventing complete dissolution. Previous experiments have shown that complete dissolution can
39
be achieved for all cured bi-functional epoxies via depolymerization and acid digestion, so the C-O
bonds must be cleaved during both processes.
In cured tetra-functional epoxies as shown in (Table 3-2), some of the crosslinked units
formed in Reactions I and II (shaded gray) contain two nitrogen atoms and one oxygen atom (f =
5). By breaking only the C-O bonds in such crosslinked units, the shaded units in tetra-functional
epoxies remain crosslinked by the C-N bonds (f = 4), and complete dissolution is impossible.
Experiments demonstrated that cured tetra-functional epoxies were completely dissolved via both
depolymerization and acid digestion. Thus, the C-N bonds must be cleaved to some extent during
dissolution via depolymerization and acid digestion. The finding demonstrates that C-N bonds are
also one type of cleavable site during epoxy dissolution.
Figure 3-9. Molecular structure in epoxy monomers and amine-cured epoxies
40
Table 3-2. Possible cleavable crosslinked units in amine-cured epoxies with A/E = 40%
Reactions I and II: Epoxy-Amine Reaction III: Epoxy-OH
Cured Bi-functional Epoxy
Cured Tetra-functional Epoxy
Bond dissociation energy (BDE), also commonly referred to as bond energy, is the energy
required to break the bond homolytically under standard conditions, which can be used as a
measure of covalent bond strength [78]. Note that the bond strength is only applicable to radical
cleavage mechanism, not to ionic cleavage mechanism that is involved in chemically digestion
epoxy matrix. Although the BDE measurement conditions differ from our chemical dissolution
approaches, the BDE values can be used as references to determine relative strength of bonds. This
thesis used experimental hemolytic BDE data, collected and organized by Luo, in the ideal gas
state at standard pressure and at a reference temperature of 298.15 K, to evaluate the bond strength
in our epoxy crosslinking [78]. Associated BDE values are summarized in Supplementary Table
A-1. There are three possible types of ether bond cleavage (R-O-R’, boldface = dissociated atom)
in the cured epoxies: 1) R, R’= alkyl substituents (CH3–OCH3, 351.9 J/mol), 2) R = alkyl
substituent, R’= aryl substitute (CH3–OC6H5, 263.2 J/mol), 3) R = aryl substitute, R’ = alkyl
substituent (C6H5–OCH3, 418.8 J/mol). For C-N bonds, there are two possible types of bond
cleavage (R-NH-R’ or (R-N-R’R’’, boldface = dissociated atom): 1) R = alkyl substituent, R’,
R’’= aryl substitutes (CH3–NHC6H5, 298.7 J/mol), 3) R = aryl substitute, R’, R’’ = alkyl
substituent (C6H5–NHCH3, 420.9 J/mol).
41
The above data show that C-O and C-N bonds with R = alkyl substituent and R’= aryl
substitute have the lowest BDE, indicating that the bonds between O/N atoms and the secondary
carbons in crosslinked units (-C6H5O-CH2-, -(C6H5)2N-CH2-), as shown in Figure 3-9 (a), are
more likely to be broken during dissolution. For C-S bonds in DDS, there is only one possible type
of bond cleavage: (C6H5– SO2C6H5). A similar structure (C6H5– SO2CH3) shows a BDE of 344.3
J/mol, which is ~50 J/mol greater than the BDEs required by C-O and C-N bonds, and thus less
like to be broken during the reaction. GC-MS data in Supplementary Figure A-8 confirming the
existence of the bisphenol A structure in the dissolution products, indicating that the aromatic
structures of the epoxies were preserved during dissolution, and the C-O and C-N bonds were
selectively cleaved.
I conclude that during dissolution, polymer chains with various lengths are first separated
from the heavily crosslinked network. As the reaction continues and more polymer chains are
digested, complete resin dissolution occurs gradually. The dissolved polymer chains continue to
react to yield progressively shorter oligomers or even monomers (Supplementary Figure A-7 and
Figure A-8). Data show that the C-O and C-N bonds are the cleavable sites for depolymerization
and acid digestion, where -C6H5O-CH2- and -(C6H5)2N-CH2- (boldface = dissociated atom) are
the most possible cleavable positions (Figure 3-9 (a)). The assignment of the full spectrum is
shown in Figure 3-10.
42
Figure 3-10. Assignment of 1D HNMR full spectrum chemical shift of dissolution products of tri-
functional epoxy in DMSO at room temperature (32 scans)
3.4. Conclusions
This chapter has demonstrated key aspects of amine-cured epoxy dissolution at
atmospheric pressure using two chemical treatment methods: (a) depolymerization and (b) acid
digestion. The insights provided are essential to developing an effective chemical treatment
method for recycling amine/epoxy composites that is also practical on an industrial scale. To my
knowledge, such a process has not been reported in the literature. This chapter has also
demonstrated that both depolymerization and acid digestion are effective processes for dissolving
amine-cured formulations. Correlations between epoxy properties and dissolution rate were
identified, and data showed that the rates of chemical reaction and diffusion were the two factors
controlling the rate of dissolution. When the reaction temperature was less than the Tg of the epoxy,
the dissolution rate was determined primarily by the diffusion rate, while in other cases, the
chemical reaction was the rate-controlling factor. These findings indicate that to accelerate the
dissolution rate, we must understand not only chemical reactions and catalytic conditions that can
43
cleave bonds more efficiently, but also more effective ways to permeabilize heavily crosslinked
networks to accelerate diffusion.
This chapter has also demonstrated protocols for using NMR spectroscopy to qualitatively
and quantitatively track the products from chemical solutions after epoxy dissolution. The
molecular-level studies clarified the process of dissolution and provide a basis for developing
future strategies to employ catalysis. Data showed that similar crosslinked units existed in all
amine-cured epoxies with different epoxy functionality and type of curing agent. Target cleavable
sites - C-O and C-N bonds - were identified for screening catalysts to achieve faster reaction rates
of bond cleavage. -C6H5O-CH2- and -(C6H5)2N-CH2- (boldface = dissociated atom) were the most
possible cleavable positions. The aromatic structures of the epoxies were preserved during
dissolution, and the C-O and C-N bonds were selectively cleaved.
The viability of depolymerization and acid digestion processes of amine-cured epoxies with
high crosslink density (at atmospheric pressure) affords opportunity to recover and recycle high-
value fibers from composites, as well as parts of the epoxies. Presently, prospective catalytic and
permeabilization conditions are being evaluated and screened. By developing effective catalysts
and permeabilization agents, I expect faster reaction rates will be possible, sufficient to meet
industrial demand. Studies of depolymerization and acid digestion in amine/epoxy composites will
be undertaken to understand the effects of chemical treatment on recovered fiber quality. These
efforts will furnish additional insight and more comprehensive understanding of composite
recycling processes, and possibly yield a practical and scalable solution for recycling amine/epoxy
composites at atmospheric pressure.
44
CHAPTER 4. Chemical Treatment for Recycling of Amine/Epoxy Composites
at Atmospheric Pressure
4.1. Introduction
Due to the absence of effective recycling methods, most composite waste is not
recycled/treated but is sent to landfills. Without a robust and effective method to recycle
composites and complete the material life cycle, composite replacement of metallic parts will be
limited in automotive, sporting goods, and energy industries. Consequently, effective recycling
methods for CFRPs are required to reduce environmental impact and facilitate entry into sectors
requiring mass production. Our previous study has demonstrated that chemical treatments at
atmospheric pressure – depolymerization (benzyl alcohol/K3PO4 at 200 °C) and acid digestion
(acetic acid/H2O2 at 110 °C) – are both effective for amine-cured neat epoxy with high crosslink
density. The epoxy dissolution rate was controlled by both the chemical reaction rate and the
diffusion rate. The reaction mechanism for oxidative degradation of amine-cured epoxies was
investigated, revealing that reaction occurred by oxygen atom transfer to the linking aniline
groups, and the polymers were then cleaved by an elimination and hydrolysis sequence [14].
The objective of this work is to demonstrate and evaluate chemical treatment methods for
amine/epoxy composites at atmospheric pressure that can effectively separate fibers and epoxy
matrix to near-virgin quality fibers and potentially useful polymer components. A parametric study
is performed that encompasses variations in chemical treatment method (depolymerization and
acid digestion), matrix properties (epoxy monomer functionality and amine/epoxy molar ratio
(A/E)), fiber bed architecture (2 ´ 2 twill weave and unidirectional (UD) fibers), and fiber type
(carbon fiber (CF) and glass fiber (GF)). Key factors affecting the recycling process are identified
and modified to accelerate the separation process.
45
Recycling of epoxy composites presents a daunting challenge because the epoxies are
highly cross-linked, three-dimensional structures and are insoluble and under mild conditions [21-
22]. Consequently, most efforts to recycle epoxy composites have largely focused on recovering
the higher value component, the carbon fibers, discarding the polymer matrix. Currently, there are
two physical recycling approaches practiced on industrial scales: mechanical grinding [27-29] and
thermal processing [30-33]. The main advantage of physical recycling methods is that composite
materials with various matrices and fiber reinforcements can be accommodated by a single
process, and tuning these processes to particular matrix formulations is not necessary. However,
neither method recovers any of the matrix, and both methods have inherent drawbacks, as outlined
in Sections 2.1 and 2.2. Chemical treatment methods for thermoset composites at atmospheric
pressure have been reported, often using strong acids and focusing on low-grade industrial
composites.
This section reports methods for depolymerization and acid digestion on various highly
crosslinked amine/epoxy composites and compare with similar methods applied to neat epoxy.
Composites were fabricated using select epoxy formulations and fiber reinforcements, and
subsequently subjected to chemical treatments to identify correlations between epoxy matrix
functionality, laminate thickness, fiber bed architecture, fiber type and dissolution rates. Results
show that both depolymerization and acid digestion achieve complete matrix dissolution and fibers
are recovered. Acid digestion is shown to be a more effective approach for amine/epoxy composite
recycling, because it offers faster chemical reaction rates at lower reaction temperature and enables
recovery of residue-free fibers. Furthermore, two strategies to improve the matrix dissolution rate
– pre-treatment and mechanical shredding – were employed using both laboratory and commercial
amine/epoxy composites. Both strategies effectively accelerate the diffusion rate and recover clean
46
fibers. Finally, a catalytic strategy to accelerate the reaction rate was developed, reducing the
reaction time by 30%.
4.2. Experiments
Amine/epoxy resins were formulated (without toughening agents or other additives) and
characterized using thermal analysis. A total of ten fiber-reinforced composite laminates were
fabricated using prepregs that consist of the resin formulations with variations in crosslink density.
Samples of composites produced from lab-made prepregs, composites produced from commercial
aerospace prepregs (Cycom 5320-1/8HS) and commercial composite waste (provided by ELG
Carbon Fibre Ltd.), were subjected to chemical treatment to evaluate speed and effectiveness of
dissolution for recycling. Recovered fibers were examined by scanning electron microscopy (SEM)
and X-ray photoelectron spectroscopy (XPS) to determine surface quality. Residue on the fibers
was analyzed using energy dispersive X-ray spectroscopy (EDS) to determine chemical
composition.
4.2.1. Resin Formulation
Epoxy resins were formulated using three types of epoxy monomers that varied in epoxy
functionality: (1) bi-functional epoxy (diglycidyl ether of bisphenol A (DGEBA), Araldite GY
6010, Huntsman, Figure 4-1 (a)), (2) tri-functional epoxy (triglycidyl of para-aminophenol,
Araldite MY 0510, Huntsman, Figure 4-1 (b)), and (3) tetra-functional epoxy (tetraglycidyl-4,4'-
methylenebisbenzenamine, Araldite MY 721, Huntsman, Figure 4-1 (c)). A tetra-functional, high-
performance amine-based curing agent, diamine 3,3’-diaminodiphenyl sulfone (3,3’-DDS,
Aradur®9719-1, Huntsman, Figure 4-1 (d), was selected to react with the epoxy monomers). The
structure of crosslinked bi-functional epoxy from the primary and secondary amine reactions with
the epoxy groups is shown in Figure 4-1 (e). During resin formulation, specific amounts of epoxy
47
monomer and curing agent were mixed at room temperature in clean aluminum cans until fully
homogenized. The mixing ratio of amine/epoxy (A/E) were varied from 40% to 100% (molar),
providing a wide range of crosslink densities. The mixture was then heated to 120 °C in a pre-
heated convection oven to further improve the mixing quality, yielding a clear homogenous
mixture.
4.2.2. Thermal Analysis
Thermal analyses, including thermogravimetric analysis (TGA, TA Instruments Q5000 IR)
and modulated differential scanning calorimetry (MDSC, TA Instruments Q2000), were used to
measure the thermal properties of the epoxies. TGA data showed that all resin formulations began
to degrade at ~300 °C. The degradation temperature set an upper limit for further MDSC tests.
MDSC tests were performed on each sample to measure the glass transition temperature (Tg) of the
cured sample. For each MDSC measurement, a ramp was applied from -60 to 300 °C at a constant
rate of 1.5 °C/min with ±0.5 °C/minute modulation. After cure, samples were heated from -60 to
300°C at a rate of 10 °C/min with ±0.5 °C/minute temperature modulation. The Tg of the cured
epoxy was obtained from the inflection point of the last reversible heat flow signal during the ramp
cycle.
48
Figure 4-1. Resin formulations for composite matrices: (a) bi-functional epoxy, (b) tri-functional
epoxy, (c) tetra-functional epoxy, (d) amine curing agent, (e) crosslinked bi-functional epoxy
matrix, (f) selected epoxy matrices for composite fabrication
Figure 4-1 (f) shows that as the A/E ratio increased from 40% to 100%, the Tg values of bi-
functional epoxies increased from 60 to 160 °C due to the increase of the crosslink density. For tri-
and tetra-functional, the Tg values initially increased as the A/E ratio increased because of the
higher crosslink density, but then decreased due to the lower miscibility of 3,3’-DDS in tri- and
tetra-functional epoxy monomers compared to bi-functional epoxy monomers. The highest Tg
values of bi-, tri- and tetra-functional epoxies were 160 °C (DDS/Bi = 100%), 194 °C (DDS/Tri =
60%), and 217 °C (DDS/Tetra = 60%), respectively. These three formulations were selected as the
epoxy matrices for composite fabrication.
4.2.3. Composite Fabrication
Composites were fabricated using the formulated epoxies and four types of fiber
reinforcement: 2 ´ 2 twill weave CF fabric (3K, 135 g/m
2
, FibreGlast), UD CF (12K, 305-325
g/m
2
, FibreGlast), plain weave GF fabric (323 g/m
2
, FibreGlast), stitched UD GF (955 g/m
2
,
49
FibreGlast). To fabricate composites, resin films (230 ´ 230 mm) were first produced by spreading
resin onto release films on a heated plate (50 °C). Dry fabrics (230 ´ 230 mm) were secured at all
four edges with tape to preserve the fiber orientations. The dry fabric sheets and the resin films
were stacked together with one resin film on each side of the fabric sheets. The stack was then hot
pressed (Wabash) at 50 °C and 100 kPa for 2 minutes. The final size of all prepregs excluding the
tape edge was 200 ´ 200 mm, with a resin content of 45 ± 2%.
Using the prepregs above, composites laminates were fabricated via vacuum bag-only
(VBO) processing. Bag edges were sealed to prevent resin bleeding, as shown in Figure 4-2 (a).
The same cure cycles were used for all laminates, consisting of a 1.5 °C/min ramp rate, a 120 °C
dwell of three hours, another 1.5 °C/min ramp rate, and a 180°C dwell of three hours. Final resin
content in all laminates was 45 ± 2%. Cross-sectional images (Figure 4-2 (b)) revealed that carbon
fibers were fully impregnated by epoxies, and the void content was < 2%. Laminates were also
produced using commercial aerospace grade VBO prepreg (Cycom 5320-1 T650-35 3K 8HS 36%)
using VBO processing.
Figure 4-2. Composite fabrication: a) vacuum bag-only manufacturing setup, b) polished section of
cured laminates
50
4.2.4. Composite Recycling
Three types of composite materials were investigated: composites produced from lab-made
prepregs, composites produced from commercial prepregs (Cycom 5320-1/8HS), and shredded
composite waste (provided by ELG Carbon Fibre Ltd.). The cured laminates were cut into 20 ´
100 mm coupons and subjected to chemical treatment. The thicknesses of the laminates fabricated
from twill weave CF fabrics were 0.5 mm (2 plies), 1 mm (4 plies) and 2 mm (8 plies). The
shredded composite waste was subjected to chemical treatment as received. Matrix dissolution
time was used to assess the effectiveness of the chemical treatment, and was defined as the time
required to fully dissolve the matrix and separate residue-free carbon fiber bundles (determined
visually). Test coupons were cut from the same laminate, and the standard deviation of the
dissolution time for samples with the same formulation was below 10%. After chemical treatment,
the recovered fibers were rinsed in acetone (~30 mL, 5 min) and water (~30 mL, 5 min) until no
residue was observed. Clean fibers were oven-dried for further analysis.
Depolymerization. Depolymerization, (more accurately described as alkali digestion, based
on the reaction medium), was performed using a supersaturated solution of 100 mL benzyl alcohol
(solvent, Sigma-Aldrich) and 7 g tripotassium phosphate (catalyst, Sigma-Aldrich). Thus, a three-
neck, round-bottom flask (1 L) containing the depolymerization solution and weighed composites
was fluxed at 200 °C in an oil bath. A magnetic stir bar was placed in the flask to homogenize the
catalyst concentration distribution. The frequency of rotation was set at 100 rpm to minimize
mechanical distortion of the fibers. Nitrogen flow was employed to create an inert environment for
the reaction and to keep the catalyst in a reactive form.
Acid digestion. Acid digestion (more accurately described as oxidative digestion) was
performed using a solution of 100 mL glacial acetic acid (solvent, EMD Millipore) and 10 mL
hydrogen peroxide solution (oxidant, 30% (w/w) in H2O, EMD Millipore). The same setup as
51
depolymerization was used for acid digestion, except that magnetic stir bars were not placed in the
flask. The resultant mixture and a composite coupon were refluxed at 110 °C, and additional
hydrogen peroxide solution (5 mL) was added to the flask every hour. No nitrogen flow was used
for acid digestion.
Pre-treatment before acid digestion. An important element of this recycling method was
the use of a two-step process with specific chemical substances that first physically permeabilized
(swelled) the composites without disrupting the fiber weave in the polymer matrix, and second,
chemically dissolved the matrix into the solvent, thus recovering near virgin-quality fibers in
woven form. After pre-treatment, the rate-limiting effect of solvent diffusion was reduced or
eliminated. During pre-treatment, laminates were placed in benzyl alcohol at 200 °C for 4 hours.
The pre-treated laminate was then subjected to acid digestion.
4.2.5. Fiber Characterization
Scanning Electron Microscopy (SEM, JEOL JSM 7001)) was used to examine the surface
quality of recovered fibers after chemical treatments. Non-conductive GFs were sputter coated
with a conductive platinum layer before imaging to prevent charging of the specimen. CFs were
imaged without coating. Energy-dispersive X-ray spectroscopy (EDS, JEOL JSM 7001) provided
qualitative indications of the elements present in the sample, and was used to analyze the chemical
composition of epoxy residues on the fibers. Observation conditions of accelerate voltages of 15-
20 kV and working distances of 15-20 mm were used. Fiber surface functional groups were
analyzed using X-ray photoelectron spectroscopy (XPS, Kratos Axis Ultra DLD). Survey spectra
in the range of 0-1200 eV were perform for each sample, followed by high resolution scans over
the C1s range. Curve-fitting of the C1s spectra was performed in CasaXPS software using
Gaussian-Lorentzian curves with constraints on position.
52
Figure 4-3. Sample mounting sheet with red dash lines indicating the cut positions
The strength and elastic modulus of recycled and virgin carbon fibers were measured in a
single fiber tester (AGS-X, Shimadzu) to determine if the mechanical properties of the recycled
carbon fibers were affected by chemical treatment according to British Standard 11566 (1996)
[79]. Sample mounting sheets were cut into 35 × 100 mm strips with a gauge length window of 50
mm in the center. Bundles of carbon fiber filaments were cut to 100 mm in length. Individual
filaments were mounted across the windowed area of the sample mounting sheets and pinned using
tapes at either end (Figure 4-3). A 10 N load cell was used, and the sample was loaded in tension at
a rate of 5 mm/min until break. The break force was recorded and the elastic modulus calculated.
The elastic modulus was calculated using the data points between 0.02 N and 0.05 N for each
curve. Ten filaments were tested for each sample.
53
4.2.6. Test Matrix
Table 4-1 summarizes the four sets of recycling experiments performed. In Set I, the
viability of depolymerization and acid digestion for recycling composites produced from lab-made
prepregs (epoxy DDS/Bi = 100%) was evaluated. The matrix dissolution rate and the properties of
the composites, including laminate thickness (2, 4, 8 plies), fiber bed architecture (2 ´2 twill
weave and UD fibers) and the effect of diffusion, were analyzed and correlated. Advantages and
drawbacks of both methods were determined. In Set II, the effect of epoxy functionality on matrix
dissolution rate was evaluated by subjecting laminates with tri- (DDS/Tri = 60%) and tetra-
(DDS/Tetra = 60%) functional matrices to acid digestion. Tri- and tetra-functional epoxy matrices
featured Tg’s comparable to high-performance epoxy matrices, and thus provided indications of the
suitability of acid digestion for recycling aerospace composites. In Set III, the effect of fiber
reinforcement type was investigated, by replacing CF reinforcements with GFs. Thus, Set III
revealed difference in matrix dissolution rate during acid digestion and depolymerization for CF
and GF composites. Set IV was used to determine the viability of acid digestion for commercial
composite recycling by using two types of composite materials: composites produced from
commercial prepregs and shredded composite waste.
54
Table 4-1. Test matrix
Set Reinforcement Matrix Method Pre-treatment No. of Plies
I
CF/Twill Weave DDS/Bi=100% Acid Digestion Yes and No 2, 4, and 8
CF/Twill Weave DDS/Bi=100% Depolymerization No 2, 4, and 8
CF/UD DDS/Bi=100% Acid Digestion No 2, and 4
II
CF/Twill Weave DDS/Tri=60% Acid Digestion No 4
CF/Twill Weave DDS/Tetra=60% Acid Digestion No 4
III
GF/UD DDS/Bi=100% Acid Digestion Yes and No 1 and 2
GF/UD DDS/Bi=100% Depolymerization No 2
GF/Plain Weave DDS/Bi=100% Acid Digestion Yes and No 4
GF/Plain Weave DDS/Bi=100% Depolymerization No 4
IV
CF/8 Harness Satin 5320-1 Acid Digestion Yes 8
CF/Shredded epoxy Acid Digestion No /
55
4.3. Results and Discussion
4.3.1. Recycling Methods: Depolymerization vs Acid Digestion
For depolymerization, Figure 4-4 shows that the matrix dissolution time increased slowly
in a quasi-linear manner as the number of plies in the laminate increased. The reaction temperature
for depolymerization (200 °C) was greater than the matrix Tg values (DDS/Bi = 100%, Tg = 160
°C), so the matrix dissolution rate was limited primarily by the chemical reaction rate. In contrast,
for acid digestion, the matrix dissolution time increased sharply with number of plies (Figure 4-4).
The time required for dissolution in a 4-ply laminate was twice the time required for a 2-ply
laminate, indicating that the dissolution rate for acid digestion was diffusion-limited. Note that the
reaction temperature (110 °C) in this case was lower than the matrix Tg’s values (160 °C). Epoxy
residue on fibers was typically present between central plies, and not within fiber tows, indicating
that diffusion was rate-limiting in the chemical treatment. Furthermore, the matrix dissolution time
required for an 8-ply laminate was more than 4 times that of a 2-ply laminate (Figure 4-4). The
extended dissolution time was due to the increased water content introduced by H2O2 from both
the H2O2 aqueous solution (major) and the chemical reaction (minor). As water content increased,
the reaction temperature fell below 110 °C, reducing the chemical reaction rate. In thin laminates
(2 and 4 plies), water content had a negligible effect, and acid digestion yielded a faster overall
dissolution rate than depolymerization.
56
Figure 4-4. Effect of thickness on epoxy matrix (DDS/Bi =100%) dissolution rate
Similar trends were reported previously for neat epoxy dissolution [12]. When the reaction
temperature was less than the epoxy Tg, the dissolution rate depended primarily on the diffusion
rate. However, in other cases, the chemical reaction rate was the rate-limiting factor, and the
chemical reaction was faster for acid digestion than for depolymerization. Figure 4-4 shows a
comparison of the dissolution time for a composite (4 plies, ~1 g resin content) and neat epoxy (~1
g) of the same thickness. The matrix dissolution time was slightly greater than that in neat epoxy
for both depolymerization and acid digestion, indicating that fiber reinforcement retarded the
dissolution rate.
After recycling, the quality of recovered carbon fibers was evaluated. The main criteria for
recovered fibers evaluation are fiber surface quality and single fiber strength and modulus. Figure
4-5 compares the surface properties of virgin fibers and recovered fibers after depolymerization
and acid digestion. The SEM images reveal that the virgin fiber surfaces were smoother than those
of recovered fibers because chemical treatment removed fiber sizing. (Sizing is a mixture of
chemicals applied to optimize the surface chemistry.) No significant defects or residues were
observed on the recovered fibers after acid digestion. Fibers recovered after depolymerization
57
showed occasional residue, and EDS spectra (Supplementary Figure B-1) revealed that the residue
was primarily undissolved catalyst (K3PO4). To identify the surface functional groups of recovered
carbon fibers, XPS spectra of virgin fibers and recovered fibers were analyzed, as shown in Figure
4-5. All survey scans (Figure 4-5 (a)) showed three main peaks: carbon (284.6 eV), nitrogen (400.5
eV) and oxygen (532.0 eV). In the C1s high resolution spectra (Figure 4-5 (b), (c) and (d)), the
peaks at ~ 284.8, 286.4 and 288.9 eV were assigned to C-C, C-OH and COOH groups,
respectively. Percentages of those functional groups on carbon fiber surface are shown in Table 4-
2. I observe that the percentages of the C-OH groups decrease (compared to virgin carbon fiber)
for carbon fibers recovered via both depolymerization and acid digestion. This finding indicates
removal of sizing from fiber surfaces, which is consistent with the SEM images shown in Figure 4-
5. Furthermore, the percentage of the COOH group increases in fibers recovered via acid digestion
(Figure 4-5 (d)) due to the strong oxidative environment during the reaction.
Table 4-2. Percentages of functional groups on carbon fiber surface.
Samples C-C [%] C-OH [%] COOH [%]
Virgin carbon fibers 66.5 32.0 1.5
Fibers recovered via depolymerization 77.8 20.8 1.4
Fibers recovered via acid digestion 75.5 17.7 6.9
58
Figure 4-5. XPS spectra and SEM images of virgin and recovered carbon fibers: (a) XPS survey
spectrum, and C1s spectra and SEM images for (b) virgin carbon fibers, (c) carbon fibers
recovered from depolymerization, (d) carbon fibers recovered from acid digestion
Single fiber tensile tests were performed on recovered carbon fibers, and calculated
strength and modulus values were compared to those of virgin fibers. Table 4-3 shows that the
tensile strength of recycled fibers is 15% less than the de-sized virgin fibers. Recovered fibers
from depolymerization and acid digestion retain 98% and 85% of the tensile strength, respectively.
The lower tensile strength of recovered fibers from acid digestion is partly due to the longer
reaction time (48 h) than polymerization (8 h). With improved diffusion rate and reduced reaction
time, the tensile strength of the fibers recovered form acid digestion are expected to improve.
Recovered fibers from both alkali digestion and oxidative digestion retain 98% of the tensile
modulus compared to de-sized virgin carbon fibers.
59
Table 4-3. Tensile strength and modulus of recovered carbon fibers from laminates with 8 plies
Samples Strength (MPa) Modulus (GPa)
De-sized virgin fiber 2693.87 ± 490.62 136.17 ± 6.70
Recycled fibers: alkali digestion 2586.10 ± 692.40 135.09 ± 9.97
Recycled fibers: oxidative digestion 2175.07 ± 559.12 133.40 ± 5.27
The observations described above demonstrate that both depolymerization and acid
digestion are potentially effective methods for recycling of amine/epoxy composites. In
depolymerization, the matrix dissolution rate depends primarily on the chemical reaction rate,
while for acid digestion, diffusion is rate-controlling. Acid digestion appears to be more effective
than depolymerization for recycling amine/epoxy composites for five reasons: 1) faster reaction
rate, because the linking aniline groups create pathways for oxygen atom transfer from H2O2, and
the polymer can then be cleaved by elimination, 2) lower reaction temperature, 3) recovery of
residue and defect-free fibers, 4) robust and convenient setup (depolymerization requires N2
environment), and 5) clean chemical feedstock after reaction (depolymerization uses K3PO4 as a
catalyst, which has low solubility in benzyl alcohol and leaves solid residues in the solution after
reaction). Therefore, I henceforth focus on acid digestion.
4.3.2. Effect of Fiber Bed Architecture
To evaluate the effects of fiber bed architecture on matrix dissolution rate, laminates were
fabricated using twill weave (135 g/m
2
) and UD (305-325 g/m
2
) fibers (the same epoxy matrix:
DDS/Bi = 100%). (The most common fiber arrangements used in aerospace composites are UD
and fabric.) Figure 4-6 shows that as laminate thickness increased, dissolution time increased in
quasi-linear fashion for twill weave laminates. The data for UD laminates (dissolution time and
laminate thickness) followed the same trend observed for twill weave laminates, indicating that
60
fiber bed architecture had negligible effect on dissolution time or diffusion rate, and that laminate
thickness was a more important parameter in determining the matrix dissolution rate.
4.3.3. Effect of Matrix Functionality
In this section, acid digestion of tri- and tetra-functional composites is investigated. Note
that tri- and tetra-functional matrices have more heavily crosslinked networks compared to bi-
functional epoxies, and this limits the diffusion of solvent molecules. The results are compared to
the data for bi-functional matrix composites to gain a more complete understanding of
amine/epoxy matrix dissolution. Figure 4-7 shows the dissolution times of bi- (Tg = 160 °C), tri-
(Tg = 194 °C), and tetra- (Tg = 217 °C) functional composites. All sample matrices dissolved in <
10 h via acid digestion, resulting in recovery of clean carbon fibers. Tri- and tetra-functional
composites required comparable dissolution times, as they contained similar levels of crosslinked
bonds and Tg values [12]. For bi-functional composites, the dissolution time was shorter than for
tri- and tetra-functional composites. The shorter dissolution time was attributed to a more rapid
diffusion rate in the bi-functional matrix, which exhibited lower Tg and crosslink density.
Figure 4-6. Effect of fiber bed architecture on epoxy matrix (DDS/Bi =100%) dissolution rate
61
Figure 4-7. Effect of epoxy functionality on matrix dissolution rate
Composite materials in aerospace parts typically require thermally stable matrices that can
tolerate high service temperatures. The Tg requirement for aerospace resin systems is generally >
160 °C. For example, the selected aerospace grade VBO prepreg in this study, 5320-1/8HS,
exhibits a Tg = 197
°C after post-cure. The Tg values of the tri- and tera-functional epoxy matrices
used in the composites produced from lab-made prepregs were comparable to that of 5320-1/8HS.
The ability to fully dissolve tri- and tetra-functional matrices demonstrates that acid digestion is a
potentially viable approach for recycling of aerospace grade amine/epoxy composites. This
assertion will be explored further in section 4.3.5.
4.3.4. Effect of Reinforcement Type
Although CF is by far the most widely used fiber in high-performance applications, the cost
of carbon fiber is generally 10-50 times that of E-glass fiber. Consequently, GF is the most
commonly used reinforcement overall in industry. Results presented thus far have shown that acid
digestion is effective for matrix dissolution in CF composites. In this section, I seek to understand
the viability of acid digestion for glass fiber composites. Two types of E-glass fiber reinforcement
62
were selected: UD GFs and plain weave GF fabrics. The same epoxy matrix (DDS/Bi = 100%)
was used for both fiber types.
Figure 4-8 shows matrix dissolution time as a function of laminate thickness for CF and GF
composites. For laminates with similar thickness (~1 mm), the matrix dissolution time for GF
composites was more than twice that of CF composites. Thicker GF laminates (~1.5 mm) were not
fully dissolved even after 70 h (Figure 4-8), and matrix residue remained. The residue on GFs was
rust red, while the residue on CFs before reaching complete dissolution was eggshell. To verify
the above observation, a different type of GF reinforcement (plain weave GF fabrics) was used.
Full dissolution of the epoxy matrix in the second GF laminate (~1.4 mm) also was not achieved
after 40 h. Previous recycling studies using nitric acid showed that the time required to reach the
same degradation degree was longer in GF composites than in CF composites (using the same
reaction temperature and acid concentration) [59-60], and these findings were consistent with our
observations. Furthermore, GF composites (UD and plain weave, ~1.5 mm) were also subjected to
depolymerization and results were compared with the data from acid digestion. GFs were
recovered via depolymerization in ~8 h, which was comparable to CF composites, indicating that
unlike acid digestion, depolymerization was effective for GF composites. To understand the
relative ineffectiveness of acid digestion for GF composites, the chemical compositions of the
residues were analyzed using EDS.
63
Figure 4-8. Effect of fiber reinforcement type on epoxy matrix (DDS/Bi =100%) dissolution rate
Figure 4-9 (a) shows an EDS spectrum acquired from residue from CF composites
collected before reaching full dissolution via acid digestion. The three most prominent elements in
the spectrum are carbon (C), oxygen (O), and sulfur (S), all of which are typical for amine-cured
epoxies. In contrast, (b) and (c) reveal that the elements present in the residues on the GFs also
contained silicon (Si), aluminum (Al), and calcium (Ca). These three elements, Si, Al and Ca, are
related to the components in GFs (aluminum oxide (Al2O3), calcium oxide (CaO) and silicon
dioxide (SiO2)), indicating partial decomposition of GFs during acid digestion. Supplementary
Figure B-2 shows that the recovered GFs had no significant defects, and thus the extent of fiber
degradation was low. Figure 4-9 (d) shows the presence of phosphorus (P) and potassium (K) in
the residue on GFs after depolymerization, elements attributed to catalyst (K3PO4) contamination
of the fibers. Other elements (C, O, and S) were identical to the residue on CFs, indicating that
there was no degradation of the GFs in depolymerization, which matched the observation that
clean GFs were recovered from depolymerization.
64
Figure 4-9. EDS spectra of epoxy residues (DDS/Bi= 100%) on recovered fibers: a) CF twill
weave, acid digestion, b) GF UD, acid digestion, c) GF plain weave, acid digestion, d) GF plain
weave, depolymerization
I can conclude from these observations that during acid digestion, GFs are generally more
susceptible to reaction than CFs. The degradation of GFs can lead to changes in chemical
composition of the epoxy matrix, which prevents the matrix from reaching full dissolution. The
reaction condition of depolymerization has negligible effects on both CFs and GFs, and the
dissolution behavior of CF and GF composites are nearly identical.
65
4.3.5. Strategies to Accelerate Diffusion
The results described above demonstrate that acid digestion is a suitable candidate for
recycling of amine/epoxy matrix composites - it presents a relatively fast reaction rate at a
moderate reaction temperature (110 °C) and yields clean fibers. However, the slow diffusion rate
in heavily crosslinked matrices limits the overall matrix dissolution rate. In the following sections,
two strategies to increase the diffusion rate and accelerate the process - pre-treatment and
mechanical shredding - are explored.
4.3.5.1. Pre-treatment
The underlying purpose of pre-treatment is to accelerate dissolution by first immersing the
composite in appropriate solvents before the chemical reaction. During the immersion process, the
solvent penetrates the crosslinked network, enabling reactant molecules to reach cleavable bonds
more easily, thus reducing/eliminating the rate-limiting effect of diffusion. Different solvents were
evaluated, including benzyl alcohol, xylene, diethylene glycol dimethyl ether (DGDME), and
diethylene glycol methyl ether (DGME). Trials determined that the most promising pre-treatment
solvent was benzyl alcohol (solvent) at 40 °C above the matrix Tg, for 1.5 hours per millimeter
thickness of the laminate. The solvent pre-treatment could be applied repeatedly, as no chemical
reaction occurred during this step. In the second step, pre-treated composites were subjected to
chemical reaction (acid digestion) to dissolve the epoxy matrix and separate clean fibers from the
matrix.
66
Figure 4-10. Effect of pre-treatment on matrix (DDS/Bi= 100%) dissolution rate via acid digestion
(a), cross-sectional images of laminates (b) before and (c) after pre-treatment, and carbon fiber
fabrics recovered from (d) an 8-ply laminate using acid digestion with pre-treatment
Figure 4-10 (b) and (c) show that after pre-treatment at 200 °C, the laminate (amine/epoxy
DDS/Bi= 100%, Tg = 160 °C) swelled (weight increase = 100%) and expanded in thickness
(thickness increase = 120%). By applying pre-treatment prior to acid digestion, the dissolution
time reduced to 1 h for all laminates, regardless of thickness (Figure 4-10 (a)), confirming that pre-
treatment effectively removed the diffusion limit in composites. No mechanical stirring was
required for the pre-treatment step or for the subsequent chemical reaction step, and the fiber
arrays remained organized. Figure 4-10 (d) shows the fabrics recovered from an 8-ply laminates.
Recall that without pre-treatment, the chemical reaction on solid composites was heterogeneous,
initiating at the surface and progressing to the laminate center. The long dissolution time distorted
and tangled the fibers, adversely affecting fiber properties. In contrast, pre-treatment before acid
digestion homogeneously decomposed and dissolved the polymer matrix, preserving the fiber
weave and minimizing fiber damage from abrasion. GF composites were also subjected to acid
67
digestion after pre-treatment in benzyl alcohol. GF composite matrices reached full dissolution in 3
h, confirming the assertion that the residue on GF fibers (Figure 4-8 and Figure 4-9) arose from GF
degradation after acid digestion (40+ h).
Pre-treatment was also applied to an 8-ply laminate fabricated from commercial prepreg
(Cycom 5320-1/8HS). Prior experience dictated a pre-treatment temperature of 240 °C (the Tg =
197 °C for the cured 5320-1/8HS laminate). However, because the boiling point of benzyl alcohol
is 205 °C at atmospheric pressure, the pre-treatment was performed in a closed apparatus at 220 °C
and 5 atm. After pre-treatment, the laminate was subjected to acid digestion, and all 8 plies
separated in 48 h (Figure 4-11 (a)). The dissolution time required will likely be much shorter if the
laminate is pre-treated at a higher temperature.
Figure 4-11. Carbon fiber fabrics recovered from commercial composites (5320-1/8HS): (a)
separated plies, (b) recovered fabrics with toughener residue, (c) clean fabric after DMSO wash
68
Images of the recovered CF fabrics (Figure 4-11 (b)) showed a white surface residue,
which was likely attributed to polyethersulfone (PES), a commonly used toughener that is stable
during acid digestion. Both 1D HNMR and FTIR spectra from the PES residue are shown in
Supplementary Figure B-3 (a) and (b), respectively. PES can be dissolved in amide solvents [80],
such as dimethylformamide (DMF), N-Methyl-2-pyrrolidone (NMP), dimethylacetamide (DMA),
as well as other organic solvents, such as dimethyl sulfoxide (DMSO). Figure 4-11 (c) shows the
clean fabric after rinsing in DMSO. SEM images of the CFs recovered from 5320-1/8HS laminates
are shown in Supplementary Figure B-4.
4.3.5.2. Shredded Materials
Pyrolysis of composites at end-of-life is commonly preceded in practice by mechanical
shredding of waste into small pieces (ELG Carbon Fibre Ltd). Shredding introduces matrix cracks
and accelerates pyrolysis, but can also have benefits to chemical treatment approaches, reducing
diffusion distances required for reactions. To investigate the effect of shredding on matrix
dissolution rate, acid digestion was performed on shredded composite waste. The material was a
blend of longer fiber tapes impregnated with matrix, clusters of loose fibers, and fine powders
generated during shredding (Supplementary Figure B-5 (a) and (b)). The as-received shredded
material was subjected to acid digestion for 12 h. After rinsing in acetone and water, clean fibers
were recovered (Supplementary Figure B-5 (c) and (d). The recovered CFs showed no obvious
defects or residue.
69
Figure 4-12. Matrix dissolution rate in shredded composite waste
To determine the minimum time required to reach full dissolution in shredded materials,
reaction times from 0.5 - 12 h were evaluated. Figure 4-12 shows that using solvent (acetic acid)
only, the sample weight loss after 2 h was ~20%, which was attributed to dissolution of fine
powders into the solvent. After adding H2O2, the sample weight loss was comparable (~37%)
regardless of reaction time, indicating that full matrix dissolution was achieved in as little as 0.5 h.
This observation confirms that the limit of diffusion in shredded composites is negligible, and acid
digestion is a viable approach for recycling of shredded composite waste. Recovered short carbon
fibers are suitable for a wide range of industrial applications and processes, including injection
molding and molding compounds. Short fibers can also be manufactured into non-woven mats,
which offer enhanced drapeability and formability required for complex geometries.
4.3.6. Catalyst Development
This work has shown that acid digestion work for all amine/epoxy with various crosslink
density and resin functionality, but the resin dissolution rate must be further reduced, because end-
of life parts from industries have much greater dimensions than the laminates used in our
70
experiments and the recycling time is expected to prolong. Therefore, both reaction rate and
diffusion rate need to be investigated and accelerated.
This section seeks to develop effective and sustainable catalysts that can significantly
increase the chemical reaction to meet industrial demand. Epoxies with A/E = 40% (Tg = 50 °C)
were selected as the low Tg makes it possible to remove the diffusion limit at the reaction
temperature (110 °C) during acid digestion. Two potential rate-limiting steps during the chemical
reaction – oxygen atom transfer (OAT) and elimination – were proposed and catalysts were
screened for each step. First, two OAT catalysts were studied, including methyltrioxorhenium
(MTO) and ascorbic acid (VC). Results show that the OAT catalysts bring no significant improve
to the reaction rate, indicating that the OAT step is not the key rate-limiting step for acid digestion
(Figure 9). Then, elimination catalysts – scandium chloride (ScCl3) and aluminum chloride (AlCl3)
– were applied to acid digestion. Results reveal that ScCl3 reduced ~30% of the dissolution time
compared to acid digestion conditions without catalysts, confirming that elimination is a more
critical rate limiting step than OAT for acid digestion. However, AlCl3 doesn’t show significant
improvement to the epoxy dissolution rate.
I conclude from the above observation that the use of elimination catalysts can accelerate
the reaction rate. ScCl3 is the most effective catalyst to date. Epoxies with higher A/E ratios
(100%, Tg = 190 °C) and carbon fiber-reinforced composites with the same epoxy matrix were also
subjected to acid digestion with the above catalysts. Unfortunately, no significant improvement in
dissolution rate was observed, which was attributed the slow diffusion rate in the heavily
crosslinked epoxies.
71
Figure 4-13. Effect of catalysts on resin dissolution time
4.4. Conclusions
The results presented demonstrate key aspects of matrix dissolution for both CF and GF
composites (with amine-cured epoxy) using two chemical treatment methods at atmospheric
pressure: (a) depolymerization (benzyl alcohol/K3PO4 at 200 °C) and (b) acid digestion (acetic
acid/H2O2 at 110 °C). Both depolymerization and acid digestion processes dissolved amine/epoxy
matrices. However, acid digestion was deemed more suitable and practical for amine/epoxy
composite recycling because of 1) faster chemical reaction rate at 2) lower reaction temperature,
with 3) the recovery of residue and defect-free fibers. The reaction mechanism of acid digestion
(oxidative digestion) for amine-cured epoxies occurred by oxygen atom transfer to the linking
aniline groups followed by bond cleavage via elimination. Furthermore, correlations between
epoxy matrix functionality, fiber bed architecture, type of fiber reinforcement, and dissolution rate
were identified. Acid digestion was an effective method for dissolving heavily crosslinked epoxy
matrices, and fiber bed architecture had a negligible effect on dissolution rate. For GF composites,
however, acid digestion caused fiber degradation.
72
Experiments also demonstrated that the rates of chemical reaction and diffusion were the
two factors controlling matrix dissolution rate, and the diffusion limit was most significant in
matrices with high crosslink density when using acid digestion. The findings indicated that the
dissolution rate was accelerated by permeabilizing the heavily crosslinked networks to accelerate
diffusion. Two strategies to accelerate the diffusion rate – pre-treatment and mechanical shredding
– were investigated, and aerospace-grade amine/epoxy composites and commercial shredded
composite waste were evaluated. Both strategies accelerated the diffusion rate, resulting in
recovery of clean fibers. Furthermore, pre-treated polymer matrix was homogeneously
decomposed and dissolved into solution, preserving the fibers in fabric form with negligible fiber
damage. Finally, data confirmed that catalytic conditions could improve the chemical reaction rate
for acid digestion. The most effective catalyst to date was ScCl3, which reduced ~ 30% of the time
required by acid digestion with no catalyst.
The viability of acid digestion with pre-treatment of amine/epoxy composite at atmospheric
pressure affords opportunity to recover and recycle high-value fibers from composites, while
preserving useful parts of epoxy molecules. However, evaluation of this solution revealed three
drawbacks. First, H2O2 is unstainable – it is an expensive and explosive reagent that cannot be
deployed on an industrial scale. Second, the oxidative degradation reaction in H2O2 is not
selective, and consequently, the dissolution products have little value and are thus difficult to
reuse. We are currently working on acid digestion methods to recover and reuse polymers as
additives for new epoxy formulations. Lastly, the pre-treatment temperature to remove diffusion
limit in composite with high Tg exceeds the boiling point of the pre-treatment solvent (benzyl
alcohol), which brings safety issues and adds infrastructural costs. Current efforts are devoted to
seeking safer alternatives to H2O2 oxidation that can effectively recover fibers and valuable
polymer from amine/epoxy composites, as well as permeabilization methods that requires lower
73
temperature. The insights provided in this study has furnished additional insights and a more
comprehensive understanding of amine/epoxy composite recycling processes. Development of
effective catalysts and permeabilization methods may yield faster and sustainable reactions,
possibly sufficient to meet present and future needs. Doing so will yield a more sustainable
solution for recycling amine/epoxy composites at atmospheric pressure on an industrial scale.
74
CHAPTER 5. Recovery and Reuse of Acid Digested Amine/Epoxy-based
Composite Matrices
5.1. Introduction
The increasing use of carbon fiber-reinforced polymer (CFRP) composites poses severe
environmental problems. Presently, most end-of-life CFRPs are sent to landfills because there is no
process technology for sustainable recycling. Current composite recycling methods focus on
recovery and reuse of the carbon fibers (only), typically by incorporating into molding compounds
or stitched mats for second-use applications [81-84]. While polymer matrices in CFRPs can be of
substantial value, they are far more difficult to recover than fibers, and few studies have attempted
to develop reuse approaches for matrices. The absence of viable reuse routes for polymer matrices
generates new sources of waste after recycling/re-use of carbon fibers, especially for chemical
recycling methods that require the use of chemical reagents.
Previous studies demonstrated recovery of near virgin-quality carbon fiber fabrics/arrays
from amine-cured epoxy composites using oxidative acid digestion at atmospheric pressure [12-
19]. The reaction mechanism for acid digestion was identified, and target catalysts were evaluated
to accelerate the reaction rate [14]. Chemical recycling enables recovery of high-quality carbon
fibers, although the chemical solutions generated during the process yield a mixture of solvents,
catalysts, and decomposed matrix residues. This mixture must be addressed to prevent adverse
environmental impact and make the process sustainable. In the present study, I focus on recovering
the decomposed matrix residues from chemical solutions after acid digestion, and demonstrating
routes for reusing the matrix residues in virgin resin formulations, effectively closing the recycling
loop.
75
Efforts to develop strategies for reuse of decomposed polymer matrix residue after
chemical recycling of CFRPs have been reported in the literature. For example, Okajima et al
studied mixtures of decomposed polymer matrices and virgin epoxy resins [42]. The matrices for
recycling and for mixing with matrix residues were both anhydride-based epoxies, with
triethylamine as the promoter. They reported that the strength of the blended polymers decreased
linearly as the ratio of matrix residue increased. The addition of the decomposed matrix product
had to be reduced to < 20% to retain 80% flexural strength. Secondly, Liu et al investigated mixing
decomposed polymers from amine- and anhydride-cured epoxy composites in anhydride/epoxy
systems (with an accelerator) [59,85]. They reported that the glass transition temperature (Tg) and
flexural strength of the mixed epoxy first increased as the amount of the decomposed matrix
loading increased, then decreased. The flexural strength of mixed resins with < 15 wt%
decomposed amine/epoxy matrix loadings and with < 40 wt% decomposed anhydride/epoxy
matrix loadings were greater than or similar to virgin polymer samples. Similar trends were
reported by Dang et al when mixing the neutralized matrix extracts from amine-cured epoxy
composites in anhydride-cured epoxies [60]. The flexural strength of the blended resins was
evaluated for different decomposed matrix contents, typically at 5-30 wt%.
When decomposed matrix residues are used as fillers in virgin resin formulations, the
resulting polymers have shown acceptable property levels, provided the mixing ratios are suitably
low. However, the matrix residues recovered from CFRPs are not reused as an essential
component of resins. Thus, to increase sustainability, reuse approaches for matrix residues must be
further explored and developed. Moreover, the reaction mechanism for oxidative acid digestion
differs from the methods reported in previous studies, leading to recovery of matrix residues with
distinct molecular structures and functional groups. Consequently, the matrix residue reuse
protocols for acid digestion warrant further evaluation and adjustment.
76
This study focuses on recovering the decomposed matrix residues after oxidative acid
digestion at atmospheric pressure, and demonstrating the feasibility of reusing the matrix residue
as value-added components in second-use epoxy resin formulations. The decomposed polymer
matrix residues preserve bisphenol A structures from the acid digestion solution and can be
recovered via neutralization and precipitation. Two potential reuse approaches are investigated,
including (a) reuse as an accelerator and (b) as a filler for virgin resin formulations. Data show that
the matrix residue can effectively catalyze the curing reaction of an anhydride/epoxy formulation
(without an accelerator). The flexural properties of anhydride-based epoxies samples cured with
the matrix residue are comparable to samples cured with the commercial accelerator. Data also
show that the matrix residue (< 20 wt%) can be implemented as fillers for both anhydride-cured
epoxies (with a commercial accelerator) and amine-cured epoxies, yielding blended polymers with
acceptable retention of thermal properties.
Figure 5-1. Amine-base epoxy formulation: (a) bi-functional epoxy (DGEBA), (d) tri-functional
epoxy, (c) tetra-functional epoxy, (d) 3,3’-DDS, (e) crosslinked bi-functional epoxy matrix
77
5.2. Experiments
5.2.1. Epoxy Formulation for Acid Digestion
Three types of epoxy monomers, each with different epoxy functionality, were selected for
oxidative acid digestion: (a) bi-functional epoxy (diglycidyl ether of bisphenol A (DGEBA),
Figure 5-1 (a), Huntsman), (b) tri-functional epoxy (triglycidyl of para-aminophenol, Figure 5-1
(b), Huntsman), and (c) tetra-functional epoxy (tetraglycidyl-4,4'-methylenebisbenzenamine,
Figure 5-1 (c), Huntsman). An amine curing agent (diamine 3,3’-diaminodiphenyl sulfone (3,3’-
DDS), Figure 5-1 (d), Huntsman) was selected to react with the epoxy monomers. The molar
mixing ratio for amine hydrogen/epoxy groups was 40% (1:7.5 by weight). Note that decomposed
matrices from different amine ratios had similar structures. The mixture was heated to 120 °C for
mixing until a visually clear liquid was achieved. Neat epoxies were cured in a convection oven
using the same cure cycle, consisting of a 1.5 °C/min ramp to 250 °C, follow by a dwell for 0.5
hour. Figure 5-1 (e) shows the two main types of reactions occurring during cure of amine/epoxy
resins, including amine reactions with the epoxy groups, and an etherification reaction of the
pendant hydroxyl groups formed during amine-epoxy reaction with the epoxy groups [70-71].
5.2.2. Oxidative Acid Digestion
Oxidative acid digestion was performed at 110 °C using a solution of 100 mL glacial acetic
acid (EMD Millipore) and 5 mL H2O2 solution (30% (w/w) in H2O, EMD Millipore). Additional
H2O2 solution (5 mL) was added to the flask every hour. The reaction times for bi-functional
epoxy samples were 1 h (full dissolution) and 4 h (over-reacted). The reaction time to reach full
dissolution was 8 h for tri- and tetra-functional epoxies. After reaction, the chemical feedstock
(Figure 5-2 (a)) was neutralized using a 10% sodium hydroxide aqueous solution. Viscous liquid
78
residues (Figure 5-2 (b)) were subsequently obtained through precipitation. The recovered viscous
residues contained traces of acetic acid, which were removed by granulation and washing in water,
yielding solid powder residues (Figure 5-2 (c). The matrix residues were further washed in water
and dried at room temperature overnight for reuse. The matrix recovery rate was ~25% by weight.
Note that matrix residues could also be recovered from solution by evaporating all the acetic acid
solvent. However, the matrix recovered via evaporating contained much more acetic acid solvent
than matrix recovered via neutralization and evaporation, and was difficult to reuse in virgin resin
formulations.
5.2.3. Reuse of Decomposed Matrix Residues
For reuse of recovered matrix residue, specific amounts of the materials were added to
formulated virgin bi-functional epoxies and manually mixed at 100 °C until homogenized. (Bi-
functional epoxy resin formulations exhibit lower highly crosslink densities and reduced
performance levels compared to tri- and tetra-functional epoxies, and thus are more suitable for
potential reuse applications.) Three types of virgin epoxy resins were used. Type I was anhydride-
based (Figure 5-3 (b)) bi-functional epoxies (DGEBA) without a commercial accelerator. The
molar ratio for anhydride/epoxy groups was 100% (9: 10 by weight), and the crosslinked structure
of anhydride-cured epoxy is shown in Figure 5-3 (c). Type II was an anhydride-cured bi-functional
epoxies with a commercial accelerator (1-methylimidazole, Huntsman), while Type III was a DDS
cured bi-functional epoxies (amine hydrogen/epoxy ratio = 100%, Scheme 1 (e).).
79
Figure 5-2. Recovered matrix residues from (a) solutions after (b) precipitation and (c) granulation
Bi-functional epoxy/anhydride systems without an accelerator (such as Type I) generally
have low reactivity, and thus the curing must be carried out in the presence of Lewis bases (such as
tertiary amines) to catalyze the reaction [86-87]. Material recovered from amine-cured epoxies
contain tertiary amines, and potentially these can be used as accelerators for the aforementioned
anhydride-cured bi-functional epoxies (Type I). In contrast, tri- and tetra-functional
epoxy/anhydride systems also contain tertiary amines in the epoxy monomers, and thus are self-
catalyzed during cure (Supplementary Figure C-1). The mixing ratios of recovered matrix/virgin
epoxies range from 1-20 wt% (of the epoxy part of the resin). Recovered matrix residues were
added as fillers to Types II and III at mixing ratios of 10-20 wt% of the resin and blended.
Figure 5-3. Anhydride-based epoxy formulation: (a) diglycidyl ether of bisphenol A (DGEBA), (b)
methyltetrahydrophthalic anhydride (MTHPA), (c) crosslinked bi-functional epoxy matrix
80
5.2.4. Characterization
Modulated differential scanning calorimetry (MDSC, TA Instruments Q2000) was
performed to analyze the curing reaction and the glass transition temperature (Tg) of epoxy
samples after cure. Nuclear magnetic resonance spectroscopy (NMR, Varian Mercury 400) was
used to determine the chemical structures of the recovered polymers after chemical treatment. The
dried sample was dissolved in deuterated dimethyl sulfoxide (DMSO, EMD Millipore) for HNMR
analysis. Fourier-transform infrared spectroscopy (FTIR, Nicolet 4700) was used to monitor the
crosslinking progress by observing of concentration of oxirane rings. Flexural strength and
modulus of virgin matrix and recovered matrix samples were measured by four-point bending at
room temperature (Instron 5567) with a 500 N load cell following ASTM D6272-17 [88]. The
epoxy sheets were cured using a cure cycle consisting of a 1.5 °C/min ramp rate, a 120 °C dwell of
three hours, another 1.5 °C/min ramp rate, and a post-cure at 180 °C for three hours. The testing
specimen dimensions were 110 ´ 12.7 ´ 3.2 mm. Five specimens were tested for each sample.
5.3. Results and Discussion
5.3.1. Recovered Epoxies
Figure 5-4 compares HNMR spectra from bi-functional epoxy monomer (DGEBA) with
matrix residues recovered via acid digestion at 1h and 4h. The HNMR spectra of DGEBA (Figure
5-4 (c)) and the matrix residue after 1h of reaction (Figure 5-4 (a)) show three peaks occurring at
the same chemical shift positions. The two peaks at ~7 ppm correspond to the aromatic carbon
hydrogens of the benzene rings in DGEBA, while the peak at 1.5 ppm corresponds to the methyl
groups between the benzene rings, indicating that the aromatic regions of the epoxies were
preserved. The GC-MS data reported previously [12] also showed the existence of the bisphenol A
81
structure in the dissolution products, confirming that the aromatic structures of the epoxies were
preserved during dissolution. The peaks at ~ 4 ppm correspond to the 2
o
and 3
o
carbon hydrogens
between the epoxy and the amine curing agent, and the unreacted hydroxyl groups attached to
these carbons [12]. The recovered matrix after 4h of reaction (Figure 5-4 (b)) showed no peak at
the aromatic region, indicating that the decomposed matrix continued to react in the solution to
yield progressively smaller organic components, which were of little value and were also difficult
to recover. This finding has significance, as efficient recycling/pretreatment is required to retain
valuable components of the matrix [12].
Figure 5-4. HNMR spectra of recovered matrix residues after (a) 1 h reaction and (b) 4 h reaction,
and (c) DGEBA epoxy monomer
82
I conclude from these observations that after acid digestion, the decomposed matrix
residues can be recovered (via precipitation from the solution) in both viscous liquid and solid
powder states. The recovered organics contain the aromatic structures of the epoxies, tertiary
amines from the amine curing agent, and hydroxyl groups from epoxy ring opening during cure.
Thus, these components can be added to virgin epoxy formulations as accelerators and/or fillers for
reuse. However, longer reaction times during acid digestion progressively decompose the matrix
into smaller organics, reducing the value of the recovered components. Therefore, sustainable
chemical recycling must not only reclaim high quality fibers, but also recover useful matrix
components for second-use.
5.3.2. Reuse Approach I: As Accelerators for Anhydride/Epoxy Formulations
Table 5-1 summarizes the reactivity of the matrix residues recovered from amine-cured
epoxies when used as an accelerator for anhydride-based bi-functional epoxy formulations
(without a commercial accelerator). Recovered matrix residues in both viscous liquid and solid
powder states were tested at select mixing ratios. (The weight percent of accelerators in this
section is based on the epoxy part of the resin unless otherwise stated. The mass of 10 wt% of the
epoxy part of the resin is equivalent to ~5 wt% of the resin.) In the absence of accelerators, no
exothermic chemical reaction occurred, as expected. The Tg of the epoxy with no accelerator was -
40 °C after the cure cycle, confirming that the resin remained uncured. When 1 wt% commercial
accelerator (1-methylimidazole) was added, the peak of the curing reaction occurred at 118.9 °C,
and the total heat of the reaction was 311.1 J/g, yielding a cured epoxy with Tg = 126.8 °C. Note
that the peak of the reaction decreased as the concentration of the accelerator increased from 1% to
3% (Supplementary Figure C-2).
83
Table 5-1. Assessment of using recovered amine/epoxy matrices as an accelerator for anhydride-
based bi-functional epoxy formulations
Accelerator
Peak of Reaction
[°C]
Heat of
reaction [J/g]
Cured Tg
[°C] Material State Ratio
a
None / 0 / / -40
Commercial Liquid 1% 118.9 311.1 126.8
Bi-
b
matrix residue Liquid
e
1% 196.4 149.1 121.5
Bi- matrix residue Liquid 5% 182.1 209.1 113.1
Bi- matrix residue Liquid 10% 170.9 222.3 114.7
Bi- matrix residue Solid
f
10% 183.6 178.7 104.7
Bi- matrix residue Liquid 15% 155.3 268.5 112.0
Bi- matrix residue Liquid 20% 159.9 218.3 103.6
Bi- matrix residue Solid 20% 179.7 179.9 100.1
Tri-
c
matrix residue Solid 10% 144.9 117.3 111.8
Tetra-
d
matrix residue Solid 10% 141.1 263.1 111.0
a
Mixing ratios are based on the epoxy part of the resin by weight. Recommended mixing ratios of
commercial accelerators for anhydride-based epoxy are 0.5-2%,
b
DDS cured bi-functional epoxy
(Figure 5-1 (a)),
c
DDS cured tri-functional epoxy (Figure 5-1 (b)),
d
DDS cured tetra-functional
epoxy (Scheme 1(c)),
e
recovered viscous liquid residues after precipitation (Figure 5-2 (b)),
f
recovered solid powder residues after granulation (Figure 5-2 (c))
84
When the commercial accelerator was replaced with recovered bi-functional viscous
residues, the curing reaction could also be initiated. As the amount of the matrix residue increased,
the peaks of reaction decreased and the heat of reaction increased, because a greater concentration
of tertiary amines provided more reactive sites (Supplementary Figure C-3 (a)). This behavior was
consistent with findings from the commercial accelerator. However, the Tg of the cured epoxy first
increased as the amount of the matrix residue increased, then decreased. The initial increase in Tg
was attributed to the higher crosslinked density resulting from an increase in tertiary amine sites.
The subsequent decrease in Tg occurred when the mixing ratio reached 20 wt%, which was
attributed to a greater concentration of non-reactive small molecules and short-chain polymers
(Supplementary Figure C-3 (b)).
Figure 5-5. Reuse of recovered matrix residues as an accelerator for an anhydride-based bi-
functional epoxy formulation: (a) curing reactions, (b) Tg values of cured epoxies
85
Based on the peaks of reaction, heats of reaction, and cured Tg values, 10 wt% (of the
epoxy part of the resin) was the optimal mixing ratio for reusing matrix residues as an accelerator
in virgin anhydride-based epoxies. Figure 5-5 shows the representative heat flows and Tg curves
for anhydride/epoxy samples without an accelerator, with 1 wt% commercial accelerator, and 10
wt% matrix residue. To verify the repeatability of these data, three anhydride/epoxy samples with
10 wt% matrix residue loading were prepared and analyzed. The peaks of reaction, heats of
reaction, and cured Tg values were 170.5 ± 0.9 °C, 255.8 ± 29.2 J/g and 117.2 ± 7.5 °C,
respectively, indicating that the findings were repeatable within normal measurement bounds.
Previous studies reported that ScCl3 effectively accelerated the chemical reaction for acid
digestion of amine-cured epoxies [14]. In this study, the effectiveness of recovered matrix residues
with and without ScCl3 catalyst was compared (Supplementary Figure C-4). The heat flow curves
during cure for anhydride-based bi-functional epoxies with matrix residues with and without ScCl3
catalyst overlapped, and the cured Tg values were comparable, indicating that the existence of
metal salt catalysts during acid digestion had negligible effect on the curing reaction of epoxy
blends containing matrix residues.
The reactivities of recovered matrix in both viscous liquid and solid powder states were
also compared (Supplementary Figure C-5). The curing reaction with solid powders occurred at a
higher temperature than with viscous liquids, and the cured Tg was lower, indicating that solid
powder residues were not as effective at catalyzing the reaction as viscous liquid residues. This
observation can be understood in terms of the lower miscibility of the solid powders and the virgin
epoxies, as well as the relatively low molecular weights resulting from mechanical grinding
compared to the viscous liquids.
86
Figure 5-6. Neat anhydride/epoxy specimens cured with a commercial accelerator (top) and with
recovered matrix from amine/epoxy matrices (bottom)
Finally, I investigated the reactivities of the components recovered from tri- and tetra-
functional epoxy matrices. Tri- and tetra-functional epoxies contained tertiary amines in the epoxy
monomers (Figure 5-1), in addition to the amine-epoxy linkers formed during cure, leading to a
greater concentration of reactive sites in matrices. The increased concentration of reactive sites can
be expected to result in faster reaction rates than bi-functional epoxies. Results showed that the
peaks of reaction for recovered bi-, tri-, tetra-functional matrix residues (all at 10 wt%) were 170.9
°C, 144.9 °C and 141.1 °C, respectively, confirming that tri- and tetra-functional matrix residues
were more effective in initiating the curing reaction than bi-functional matrices (Supplementary
Figure C-6).
To evaluate the flexural properties of the cured epoxies, neat epoxy specimens were
prepared for four-point bending tests. Figure 5-6 shows anhydride-cured bi-functional epoxy
specimens cured with a commercial accelerator (top) and with recovered bi-functional matrix
residue (bottom). Both cured specimens were homogenous and void-free. None of the bend test
samples ruptured within the 5% strain limit required by the ASTM D6272 standard. The Tg of the
specimen cured with matrix residues was ~10 °C lower than the Tg of the specimen cured with the
commercial accelerator because of the presence of small molecular components in the recovered
87
matrix, which was consistent with previous DSC data. However, the flexural modulus of the
specimens cured with matrix residues was slightly greater than specimens with the commercial
accelerator, as shown in Figure 5-7. This phenomenon can be explained by a structural change in
the polymer caused by reaction between hydroxyl groups in recovered matrices and epoxy
monomers, introducing fragments from amine-cured epoxies into anhydride-cured epoxies.
Aromatic amine-cured epoxies generally exhibit a greater storage modulus than anhydride-cured
epoxies (using the same epoxy monomer) because of the higher functionality of amine, which
yields a higher crosslinking density [89], as well as the rigid aromatic structures from the amine
curing agents.
Figure 5-7. Flexural modulus (a) and stress-strain curves (b) for anhydride/epoxy samples cured
with a commercial accelerator and recovered matrices
88
The chemical reactions between bi-functional epoxy monomer (DGEBA, Figure 5-1 (a))
and the recovered matrix residue were detected via MDSC (Supplementary Figure C-7). At 40
wt% (of the resin) mixing ratio for the recovered matrix, the reaction took place at 174.3 °C,
yielding a partially cured epoxy with Tg = 26.4 °C. The peaks of reaction for DGEBA/recovered
matrices and anhydride-cured DGEBA catalyzed by recovered matrices were similar, indicating
that the two types of reactions occurred simultaneously during cure, and supporting our
explanation. Figure 5-8 shows the FTIR spectra of DGEBA cured with the matrix residues. The
peak at 915 cm
-1
was assigned to C-O stretching vibration in the oxirane ring [90], which became
weaker as the mixing ratio of the recovered matrix increased from 20 to 40 wt% (of the resin),
confirming that the matrix residues reacted with epoxy groups. The full FTIR spectra are shown in
Figure S8.
Figure 5-8. FTIR spectra of DGEBA/ matrix residue systems after a cure cycle.
89
The observations presented above demonstrate that the matrix components recovered from
oxidative acid digestion can be used as accelerators for those anhydride-based epoxy formulations
that are not self-catalyzed during cure. A mixing ratio of 10 wt% (of the epoxy part of the resin)
matrix residue provides accelerators for curing anhydride-based epoxies and yields the lowest
reaction temperature and highest cured Tg. Epoxy samples cured with recovered matrix residues
and commercial accelerators are both homogeneous and void-free. The Tg and flexural modulus of
the epoxy specimen cured with matrix residues are comparable to that of specimens cured with a
commercial accelerator.
5.3.3. Reuse Approach II: As Fillers for Epoxy Resins
In this section, recovered matrix components were explored as filler additives to virgin
epoxies and evaluated by measuring heats of reaction and Tg values. Both anhydride- (with a
commercial accelerator) and amine-based epoxy formulations were investigated. Recovered matrix
residues in viscous liquid and solid powder states were blended into virgin resin formulations at
mixing ratios of 10 and 20 wt%, which were greater than the mixing ratios for reuse as
accelerators. (In this section, the weight percent for matrix residue is based on the resin unless
otherwise stated.) Table 5-2 summarizes the properties of anhydride- and amine-based epoxies to
which matrix residues have been added as a filler.
For anhydride-cured epoxies with a commercial accelerator, I observed that both the heats
of reaction and the Tg values decreased as the mixing ratio of matrix residue increased. This trend
was consistent for both viscous liquid and solid powder residues (Supplementary Figure C-9)
because of the non-reactive small molecules and short-chain polymers in the matrix residues. The
viscous liquid residue blended into the virgin resin more quickly than the solid powders, but the Tg
values of epoxies cured with viscous liquids were less than those of epoxies cured with solid
powders, and the difference increased as the content of the matrix residue increased. This
90
phenomenon was ascribed to the higher acetic acid content in viscous liquids than in solid
powders. Acetic acid expedited the blending process, but also interfered with the curing reaction,
especially for amine-based epoxy formulations that contained alkali curing agent (discussed
below), thus reducing the Tg after cure. Note that acetic acid in viscous liquid matrix residue had
negligible effect on the reaction when used as an accelerator, because the matrix residue loading
required as an accelerator was low.
Table 5-2. Assessment of using recovered amine/epoxy matrix residue as fillers for anhydride-
(with a commercial accelerator) and amine-based bi-functional epoxy formulations
Virgin resins for
reuse
Fillers Peak of
reaction [°C]
Heat of
reaction [J/g]
Cured Tg
[°C] State Ratio
b
Anhydride/bi-
epoxy
a
/ 0 118.9 311.1 126.8
Liquid
c
10% 119.4 220.0 102.5
20% 117.7 170.4 77.2
Solid
d
10% 121.1 213.1 110.4
20% 122.0 208.5 99.2
Amine/bi- epoxy
/ 0 157.4 310.7 160.9
Liquid
10% 137.7 286.2 121.7
20% 142.1 191.9 93.9
Solid
10% 146.1 228.7 146.9
20% 139.0 254.2 136.0
a
With 1 wt% commercial accelerator (of the epoxy part of the resin),
b
mixing ratios are based on
the the resin by weight,
c
recovered viscose liquid residues after precipitation (Figure 5-2 (b)),
d
recovered solid powder residues after granulation (Figure 5-2 (c))
91
The trends observed for amine-cured epoxies resembled those observed for anhydride-
cured epoxies (Supplementary Figure C-10). The Tg values for amine-cured epoxies decreased as
the ratio of the matrix residue increased, and the Tg values for epoxies with solid powders were
greater than those with viscous liquids. The solid matrix loading in both anhydride- and amine-
based epoxies must be < 20 wt% to retain Tg values for high temperature service capability (90-120
°C). At recovered solid matrix contents greater than 20 wt%, inadequate mixing between the
recovered matrix and the resin greatly reduced the mechanical properties. The strength of the
mixed resin decreased linearly as the content of recovered matrix as a filler increased, based on
prior work [42, 59-60, 85]. The addition of 20 wt% recovered matrix residue resulted in ~80%
retention of flexural strength.
5.4. Conclusions
This chapter has investigated the viability of recovering decomposed polymer matrix
residues from amine-cured epoxies by acid digestion, and then reusing those matrix residues in
virgin epoxy formulations (both anhydride- and amine- based). Useful matrix residues can be
recovered via neutralization and evaporation of the chemical solutions produced by oxidative acid
digestion, and aromatic structures of epoxies are preserved during oxidative peroxide digestion. I
discovered that the recovered matrix residue has catalytic effects on the curing reaction of
anhydride/epoxy formulations (without an accelerator) and cannot otherwise be self-catalyzed.
Furthermore, the thermal and flexural properties of anhydride-based epoxies samples cured with
recovered matrix are comparable to samples cured with a commercial accelerator. The recovered
matrix residue can also be reused as fillers in virgin anhydride- (with a commercial accelerator)
and amine-based epoxy resins at higher mixing ratios. I conclude that the recovered matrix residue
is most suitable for reuse in anhydride-based epoxy formulations because (1) reuse as an
accelerator for the anhydride/epoxy reaction captures more value than filler additives, and (2) the
92
acid solvent trace in the matrix residue is not compatible with amine curing agents. Note that
amine-cured epoxies are the most widely used polymer matrix in high-performance composites.
Further studies are needed to increase the compatibility of the matrix residues from acid digestion
and amine-based epoxy formulations.
The reuse potential of matrix residues deconstructed by oxidative acid digestion at
atmospheric pressure affords opportunity to preserve useful parts of epoxy matrices in addition to
carbon fibers, and to thus close the recycling loop. Moreover, the value recovered from matrices
compensates for costs associated with acid digestion, rendering the process more cost effective for
end-of-life recycling of composite materials, and reducing environmental impact.
Recyclability is essential to the sustainability of the growing composite industry. Without a
robust and effective method to recycle composites and complete the material life-cycle, composite
materials face challenges to replace steel and aluminum parts, for which recycling rates are already
high. For example, the overall end-of-life recycling rates for steel and aluminum are presently >
85% and 50%, respectively, and the average recycling rate for steel in automobiles is close to
100% [91-92]. In contrast, the current recycling rate for end-of-life CFRPs is < 5% in industry
[81], and only ~2% of composites-related companies are active recyclers [7]. Clearly, more
effective recycling technologies are required to increase the end-of-life recycling rate of CFRPs.
An effective recycling process must yield valuable recyclates, has potential for scale-up,
and be economically feasible. Unfortunately, the current oxidative acid digestion does not fully
satisfy these requisites. Although it effectively recovers near-virgin quality carbon fiber from
composites, the chemical bond cleavage is not selective in the presence of H2O2, and thus the
recovered value from the polymer matrix is limited. Moreover, H2O2 is an expensive and explosive
reagent that cannot be deployed on an industrial scale. Consequently, our current efforts are
devoted to seeking safer oxygen source alternatives to H2O2 for oxidative acid digestion that can
93
be deployed to recover high quality fibers and more valuable polymers from amine/epoxy
composites, as well as to developing more effective catalysts to yield faster reactions, possibly
sufficient to meet present and future needs. Until a more viable recycling process is developed for
CFRPs, the use of composites to substitute for traditional aluminum and steel parts will be limited,
particularly in mass market applications such as automotive and infrastructure.
94
CHAPTER 6. Defect Reduction Strategies for the Manufacture of Contoured
Laminates Using Vacuum Bag-Only Prepregs
6.1. Introduction
Out-of-autoclave (OoA) prepreg processing has been developed as an effective alternative
to autoclave cure. While autoclaves use a pressurized environment to suppress voids and mitigate
geometric variability [93], OoA processing consists of curing specially designed prepregs using
vacuum bag-only (VBO) method, and can provide technical, economic and environmental benefits
compared to conventional manufacturing. OoA processing can be used to produce flat laminates
with autoclave-level microstructural and geometric quality without difficulty. However, the limited
compaction pressure available during manufacturing and the distinctive characteristics of VBO
prepregs render the production of parts with complex geometries challenging.
The two defects most commonly observed during composites manufacturing are
microstructural voids and dimensional non-uniformity. Void content is a primary concern due to
its effect on mechanical properties [94-96], and the mechanisms of void formation in prepreg
laminates have been studied extensively [96]. It is now widely accepted that gas-induced voids can
be caused by absorbed moisture, residual solvents within the resin, and air entrapped between and
within prepreg plies [96-97]. Furthermore, flow-induced voids can arise if the resin matrix does
not fully saturate the fiber bed during cure. However, in-situ monitoring of void formation within
laminates remains difficult to perform, and the exact mechanisms that govern void nucleation,
growth and migration are not entirely understood. Process-induced deformation such as thickness
variation and laminate warpage are similarly complex, as they can be caused by microstructural
features, process conditions, and phenomena such as cure shrinkage, anisotropic thermal
expansion/contraction, and tool-part interactions.
95
During autoclave cure, defects are typically suppressed by increasing the compaction
pressure. Conversely, for low-pressure VBO prepreg cure, the compaction is primarily governed
by a partially impregnated prepreg microstructure that features pervasive and highly permeable dry
regions within the fiber tows [98]. These “evacuation channels” allow entrapped air or vaporized
moisture to migrate towards the laminate boundaries and escape into the bag during the early
stages of cure process. They are subsequently infiltrated by nearby resin to form a fully saturated
microstructure. Despite these characteristics, higher void contents have been associated with
elevated levels of dissolved moisture [97, 99], excessive out-time [100], and low-temperature cure
cycles [96] that inhibit resin flow into the dry evacuation channels. Such defect sources are
relevant to both flat and complex geometries. Fortunately, their effects may be minimized by
rigorous process control during material handling, layup and cure. Due to their partial
impregnation, VBO prepreg plies have a higher bulk factor (ratio of initial thickness to final
thickness) than comparable autoclave prepregs [101]. Consequently, they must undergo greater
compaction during cure, and may be more difficult to conform to non-flat geometries.
Investigators have identified several issues that arise specifically during the manufacturing
of complex shape laminates. Ply layup over curved surfaces can pose difficulties due to the
limited drapeability of the fiber bed. Furthermore, at corners, the pressure distribution between the
mold-side and bag-side surfaces of thick laminates can be uneven, creating non-uniform
compaction conditions [102]. This effect is particularly pronounced for prepreg product forms with
high bulk factors. Prepreg and consumable materials can bridge over concave molds or wrinkle
over convex tooling [103], introducing in-plane stresses and adversely reducing the compaction
pressure over mold corners. Finally, large deformations can affect the initial prepreg
microstructure, and hence alter its original permeability. While void content remains a major
concern, commonly observed microstructural defects in complex shape laminates also include ply
96
wrinkling, corner thickening or thinning for parts manufactured on concave or convex tools,
respectively, and, in extreme cases, delamination [103]. Moreover, resin cure shrinkage and
anisotropic thermal expansion can induce residual stresses in angle laminates during autoclave
processing, leading to dimensional changes such as spring-in at corners and warpage in flat
sections [104-105].
Resin flow and fiber bed compaction are critical governing factors for defect formation in
complex shape laminates. Percolation flow, during which the resin transport occurs relative to a
compacting fiber bed, and shear flow, during which the fiber bed and resin are considered an
inextensible viscous fluid, are the two major mechanisms of flow behavior during cure [106]. The
compaction of angle laminates involves the interaction of both mechanisms [103, 107]. In addition
to resin flow and fiber bed compaction, studies also show that during autoclave processing, the lay-
up method has a significant effect on the thickness uniformity of cured laminates, while the corner
curvature radius, flange length and laminate thickness have less effect on the thickness uniformity
of the cured quasi-isotropic laminates than on the [90°]n laminates [108]. Within this context, the
low compaction pressure available during VBO cure and the distinctive flow and compaction
characteristics of VBO prepregs are likely to further complicate the fabrication of complex parts.
Studies on the OoA/VBO manufacturing of laminates with corners or other sharp
curvatures have been reported. Laminates manufactured over concave (female) tools exhibited
resin accumulation and corner thickening, whereas corner thinning was observed for convex
(male) tooling, in agreement with their autoclave counterparts [109]. Lightfoot et al. [110]
suggested a mechanism of wrinkling formation in L-shape laminates governed by the shear force
generated by mismatches of coefficient of thermal expansion between tool and laminate, as well as
ply slippage during compaction, and experiments based on this mechanism were able to minimize
wrinkles and severe in-plane misalignments. Brillant and Hubert [102,111] studied the effect of
97
different bagging configurations on thickness uniformity and void content in L-shaped concave
and convex laminates with various laminate thicknesses and corner radii. Their results showed that
void content and thickness variation could be greatly influenced by lay-up strategies. Levy et al.
[112] proposed an analytical corner compaction model for L-shaped laminates that accounted for
local pressure difference and interplay friction, and data from the model agreed with experiments.
Cauberghs and Hubert [113] focused on the effect of Z-shaped tight corner geometry with one
convex corner between two concave corners and the levels of connectivity required between ply
drop-offs and the vacuum source. They showed that resin accumulation in concave corner regions
was associated with consumable bridging, and low void content could be achieved by connecting
the dropped plies to the vacuum source.
More complex OoA/VBO composite structures were also evaluated in previous studies. For
example, Grunenfelder et al. [114] studied the void content in embedded doublers that consisted of
a local build-up of additional plies in regions of laminates and hat-stiffeners. Results showed that
voids within the embedded plies could be reduced by fostering air evacuation through prepreg
plies impregnated on only one side and longer vacuum hold times prior to cure. Finally, Hughes
and Hubert [115] investigated how quality changes with part size and complexity scale-up. They
observed that while the microstructural quality of small-scale samples was indicative of what were
observed in samples extracted from a large-scale part, the physical and manufacturing parameters
governing part quality could interact to increase void content and render industrial-scale
manufacturing challenging.
The literature cited above clarified several phenomena governing the manufacturing of
complex shapes, and confirmed that deviations from a flat geometry introduce additional defect
sources during VBO processing. However, the fundamental aspects of VBO prepreg compaction,
which are affected by local regions with complex geometry, remain unclear. Moreover, most
98
studies have focused on L-shaped or C-shaped laminates with 90° angles, and have not analyzed
the effects of other geometries, or of increasing geometric complexity. Finally, effective strategies
to limit defect levels and ensure part quality remain elusive, but highly sought after.
6.1.1. Objectives and Structure
Several factors are typically considered during studies that seek to optimize processing.
Some, including the prepreg format and ply stacking sequence, are “intrinsic” to the part and
cannot be easily adjusted due to material availability, design constraints, and desired mechanical
properties. Others, such as processing conditions and tool design, are “extrinsic” and can be altered
to maximize quality if the relevant scientific relationships are understood. This chapter seeks to
identify the effect of key processing conditions and tool characteristics (extrinsic factors) on part
quality for several prepreg and laminate combinations (intrinsic factors). I focused on a generic
part geometry consisting of a corner located between two flanges. First, I studied the effect of
increasing geometry complexity by varying tool corner angles for two prepreg materials and
various thicknesses. Then, I investigated three feasible process modifications: variation to local
corner radii, the use of intermediate vacuum hold prior to oven cure, and the application of
pressure intensifiers at corners.
6.2. Methods
Corner laminates were manufactured using VBO prepreg processing. Two extrinsic factors,
mold design and pressure application methodology, were considered in detail. These factors are
specifically relevant to non-flat geometries, and can be adjusted feasibly within a production
environment. In addition, several intrinsic factors, including prepreg product and laminate
thickness, were also varied.
99
6.2.1. Mold Design
The typical tooling used in this study is shown schematically in Figure 6-1. A total of six
tools were used to manufacture parts. Each corner angle was associated with two tools, which
featured sharp corners or corners with finite radii of curvature. Each tool included both concave
and convex corner regions, allowing concave and corner laminates to be manufactured
simultaneously. Specific corner angles (30°, 45°, 60°) were chosen to evaluate the effect of
increasing geometric complexity on laminate quality, while various corner radii (0 mm, 6.35 mm,
9.53 mm, 12.7 mm) were selected to verify the effect of local geometric discontinuity. The tools
were machined from a single billet of aluminum to avoid leaks, and surfaces were fine-polished.
Figure 6-1. Schematic illustration of mold design and bagging configuration
100
6.2.2. Materials
Experiments were carried out using two kinds of prepregs designed for OoA processing.
The first, denoted “prepreg A,” consisted of a toughened epoxy resin (Cycom 5320) and a five-
harness satin (5HS) carbon fiber fabric (T650-35 3K). The second, “prepreg B,” was comprised of
a later-generation OoA resin (Cycom 5320-1) and an eight-harness satin (8HS) carbon fiber fabric
(T650-35 3K). Both fabrics possessed the same areal weight (375 g/m2). The two resins can, and
were, cured at the same temperature. However, the resin of prepreg B was less reactive, exhibiting
a longer out-life (30 days at ambient conditions, versus 21 days for prepreg A), thus requiring a
longer cure at a given temperature. Note that the freezer storage time of prepreg A exceeded the 12
months freezer storage time stated by the manufacturer, resulting in reduced tack and potentially
other resin degradation phenomena, whereas prepreg B underwent negligible storage time, and was
therefore comparatively pristine. Both materials were used within the study because, as described
below, prepreg A was more prone to voids and consequently highlighted several trends.
6.2.3. Cure
The laminates were manufactured using traditional OoA/VBO prepreg processing
protocols. Figure 6-1 shows details of the layup configuration. Prepreg plies measuring 76.2 mm
(width) by 63.5 mm (flange) by 63.5 mm (flange) were laid up on non-perforated release film in
the 0o direction (along the laminate length). Edge-breathing dams, which provided paths for air
evacuation while preventing resin bleed, were placed along the laminate perimeter. A second non-
perforated release film was placed on top of the laminate. Two layers of breather were applied
evenly on the top of the parts within the bag. In some cases, pressure strips were used to intensify
the compaction pressure locally, at the corners. The vacuum bag was finally set, with excess bag
material located at the corner region to prevent bridging. After a vacuum bag was assembled and
101
vacuum was drawn with a stand-alone vacuum pump (Busch R5), a leak test was performed by
measuring the pressure inside the vacuum bag using a pressure gauge to ensure that sufficient
vacuum was drawn in the bag and no leak was present. First, the vacuum pump was disconnected
from the bag. Then, the gauge was used to ensure that the vacuum level within the bag did not
decrease by more than 1” Hg (3.3 kPa) in five minutes, as recommended by the manufacturer. The
laminates were cured in a convection oven (Thermal Product Solutions Blue M). Manufacturer
recommended cure cycles were used in all laminates, consisting of a four-hour room temperature
vacuum hold prior to cure, a 1.5 °C/min ramp rate, and a 121 °C dwell of two hours for prepreg A
or three hours for prepreg B.
Table 6-1. Test matrix for out-of-autoclave processing
Set Material Debulk Pressure
Strip
Tool
Shape
Corner
Radius/mm
Corner
Angle
No. of
Plies
I Prepreg A None None Concave 0 30
o
, 45
o
, 60
o
4, 8
None None Convex 0 30
o
, 45
o
, 60
o
4, 8
II Prepreg B None None Concave 0 30
o
, 45
o
, 60
o
4, 8
None None Convex 0 30
o
, 45
o
, 60
o
4, 8
III Prepreg B Yes None Concave 0 30
o
, 45
o
, 60
o
8
Yes None Convex 0 30
o
, 45
o
, 60
o
8
IV Prepreg B None None Concave 9.53, 12.7 30
o
, 45
o
, 60
o
8
None None Convex 6.35, 9.53 30
o
, 45
o
, 60
o
8
V Prepreg B None Yes Concave 0, 9.53, 12.7 30
o
, 45
o
, 60
o
8
102
6.2.4. Test Matrix
The five sets of manufactured parts are summarized in Table 6-1. First, Sets I and II studied
the effect of laminate thickness (four and eight plies), geometric complexity (corner angles of 30°,
45° and 60°) and corner type (concave and convex). These laminates comprised a baseline dataset
that was used to evaluate the effect of further process modifications. In Set III, the influence of
intermediate debulking was assessed by applying a five-minute vacuum hold after laying down
every 2 plies. In Set IV, the effect of mold corner curvature was analyzed, using corner radii of
9.53 mm and 12.7 mm for concave corners and 6.35 mm and 9.53 mm for convex corners. Finally,
in Set V, the local pressures at the concave corners were intensified by the application of pressure
strips. One laminate was manufactured for each test point.
6.2.5. Quality Analysis
Two samples were cut along the laminate length from each laminate for analysis. To avoid
edge effects, samples were taken no closer than 10 mm from the end. The two inward-facing cross-
sectional areas were polished to 2400 grit abrasive using a metallographic grinder/polisher
(Buehler MetaServ). A digital stereo microscope (Keyence VHX-600) was used to capture
micrographs over the entire cross-sectional area at a 100 ´ magnification. These individual
micrographs were stitched into a complete image of the cross-section. For convenience, the
complete image was divided into a corner region and two flange regions.
103
Figure 6-2. Thickness measurement locations for: (a) concave corner laminate, (b) convex corner
laminate, (c) concave corner laminate manufactured with a pressure strip
Laminate quality was analyzed on the basis of void content and thickness variation, two
frequently used defect metrics that can be accurately quantified via light microscopy. The void
content was measured using image analysis software (ImageJ) by converting the stitched color
image to 8-bit binary image, selecting the void regions, applying a binary threshold, and
calculating the ratio of void area to total cross-sectional area within the region of interest. The void
contents within the flange and corner regions were calculated individually for comparison. To
quantify the thickness variability, nine measurement locations were selected, with three located in
the left flange, three in the corner, and three in the right flange (Figure 6-2). The thickness at each
individual location (xi) was then measured from the micrographs. The coefficient of variation
(CoV) of the thickness, shown in equation (2), was used to compute a single numerical metric for
the thickness non-uniformity of the entire laminate (n is the number of measurement locations).
Other microstructural features, such as resin accumulation, were analyzed individually as
described in the following section.
104
(1)
(2)
6.3. Results and Discussion
6.3.1. General Defects in Sharp Corner Laminates
The two main defects observed in laminates were thickness variation and void content,
both of which were concentrated in corner regions. Figure 6-3 (a) shows void contents in laminates
made of prepreg A as a function of laminate thickness in flange regions, concave corner regions
and convex corner regions, respectively. The average void contents over the entire cross-sections
were < 2% in all prepreg A laminates. However, void contents were < 0.5% in laminates made of
prepreg B, which was much less than those in the corresponding prepreg A samples. The relatively
high void content in prepreg A laminate was attributed to the prepreg age, which resulted in
reduced tack and potentially compromised the ability for resin to flow and eliminate entrapped air.
For prepreg A, the void content increased with laminate thickness in a quasi-linear manner in all
three regions, but the defect level increased more markedly in the concave corner regions than in
the convex corner regions, where behavior was similar to that of the flange areas. In contrast, the
void content in prepreg B was negligible, and no clear trend was observed.
x=(x
1
+x
2
+x
3
+x
4
+x
5
+x
6
+x
7
+x
8
+x
9
)/9
Coefficient of Variation=
1
x
(x−x)
2
∑
(n−1)
105
Figure 6-3. Laminates made of prepreg A: (a) void content in flange region, concave corner region
and convex corner region; (b) thickness variation
Figure 6-3 (b) shows the coefficient of variation (CoV) for thickness of concave and
convex laminates made of prepreg A. The thickness variation in concave parts was generally
greater than those in convex parts. Furthermore, in the concave laminates, for a given number of
plies, the CoV increased linearly with corner angle. However, this trend was not consistent in
convex corner parts. For a given corner angle, the CoV decreased as the laminate thickness
increases in both mold types. Similar thickness variation behavior was observed in prepreg B
laminates, as shown in Figure 6-4, indicating that prepreg age affected the amount of voids but not
the variation in laminate thickness, and that mechanisms governing laminate dimensional
uniformity were therefore generally applicable. Laminates with a wider flange on one side (127
mm (flange 1) versus 63.5 mm (flange 2)) were also manufactured to evaluate the effect of flange
width, and results showed that flange width had no significant effect on thickness variation of
laminate or on fiber distribution.
106
Figure 6-4. Thickness variation of laminates made of prepreg B
Figure 6-5 shows representative micrographs of laminates manufactured over concave and
convex molds. Microstructural defects were observed, including voids, corner thickening and
corner thinning, and resin accumulation. Generally, more defects were observed in the concave
molded laminates than in their convex counterparts. In the concave corner, resin was accumulated
beyond the fiber bed, resulting in greater thickness changes in corner regions, while curvature radii
of the fiber bed were consistent through the thickness. In the convex corner, no resin accumulation
was observed, although the radius of curvature of the fiber on the mold side was significantly less
than others on the bag side, leading to convex corner thinning and increased thickness variation.
107
Figure 6-5. Defects in corner laminates: (a) concave corner, (b) convex corner
To identify the fundamental causes of thickness non-uniformity, microscopic inspection
was used to measure the CoV of the fiber bed thickness, independent of the overall laminate
thickness. Figure 6-5 (a) shows how the measurement of both thickness values was performed at a
given location. Figure 6-6 compares the overall thickness variation measured from such
micrographs to that of the fiber bed thickness in prepreg B laminates. Figure 6-6 (a) and (b)
display graphs of four-ply and eight-ply data, respectively. In concave corner laminates, the CoV
of the fiber bed was significantly less than that of the laminate thickness, but remained greater than
that of the flanges. This difference, along with observations of consistent fiber curvature radius
through the thickness, suggested that thickness variation in concave laminates was caused mainly
by resin accumulation at the concave corner. The stiffness of the fiber bed and its inability to
conform to the sharp mold radius during low pressure processing was a contributing but minor
cause. Conversely, in convex corner laminates, the two thickness data sets were generally
comparable. Locally, the fiber bundles nearest the mold had a curvature radius of 0.8 mm, which
was much less than that of the far-field plies (4 mm). This specific difference implied that in
convex laminates, corner thinning arose due to increased compaction pressure at the sharp corner
region but was limited by the average stiffness of the fiber bed during low-pressure processing.
108
To support this assertion, the length of the accumulated resin region in concave corners was
measured directly from micrographs (as shown previously in Figure 6-5 (a)). Figure 6-7 shows the
concave corner resin length as a function of corner angle in prepreg B laminates. The resin length
generally increased with corner angle, just as the CoV of the overall thickness increased,
confirming that thickness variation in concave laminates was generated primarily by resin
accumulation.
Figure 6-6. Overall thickness variation and fiber bed thickness variation in prepreg B laminates: (a)
four-ply laminates, (b) eight-ply laminates
109
From the results described above, I concluded that defect formation mechanisms in
concave corner laminates were more prone than their convex counterparts to parameters such as
corner angle and laminate thickness. In concave parts, both the inability of the fiber bed to
conform to the tool under low compaction pressure, and the consequent resin accumulation,
adversely influenced the laminate dimensional uniformity. Furthermore, low compaction force also
increased the likelihood of void formation. Conversely, the stiffness of fiber beds in convex
corners prevented corner thinning during low-pressure processing, thus reducing thickness
variation. The dataset presented above forms a benchmark against which further process
modifications were evaluated.
Figure 6-7. Concave corner resin length in laminates made of prepreg B
110
6.3.2. Effect of Intermediate Debulk Method
Intermediate debulking can improve compaction and reduce void content in laminates by
ensuring that individual prepreg plies conform to substrate layers during layup, and by allowing
entrapped air pockets to escape more readily. Laminates consisting of eight plies of prepreg B
were cured on both concave and convex tools, with vacuum compaction applied every two plies
for 5 minutes during layup. Microscopic inspection of polished sections revealed that most of these
laminates were void-free. This represented a marked improvement from laminates manufactured
without intermediate debulking, which exhibited voids.
Figure 6-8 shows a comparison of thickness variation in laminates with select corner angles
processed with and without debulking. The CoV was not strongly affected by debulking, and no
clear trend was observed across all corner angles. The inconsistent and negligible nature of the
effect was confirmed by comparing the corner resin length in concave laminate, which was also
insensitive to debulking. These results indicated that the debulking steps used in this study
contributed to the removal of entrapped air and reduced the void content, but did not affect the
conformation of the prepreg laminate to the tool.
6.3.3. Effect of Mold Corner Radius
Eight-ply laminates of prepreg were manufactured on molds with varying corner radii.
Because concave corner laminates were shown to be more prone to defects than convex corner
laminates, larger corner radii (9.53 mm and 12.7 mm) were selected for the concave tooling than
for the convex tooling (6.35 mm and 9.53 mm). No debulking was carried out.
111
Figure 6-8. Thickness variation of laminates manufactured with and without debulking
Inspection of polished sections revealed that resin accumulation within the corner regions
was almost eliminated in all cases. Figure 6-9 (a) and (b) show the CoVs in concave and convex
laminates versus mold radius. In concave laminates, the CoV decreased with increasing corner
radius. Moreover, the difference in thickness uniformity between corner angles was also greatly
reduced. Molds with large corner radii required the fiber bed to deform less, and therefore
prevented resin accumulation. In concave laminates fabricated on molds with radii of 12.7 mm, the
CoV decreased close to that of the flanges (about 0.02). In convex laminates, larger corner radii
also decreased the thickness variation, but the effect was less pronounced than in concave corner
parts, since only the prepreg plies closest to the tool were affected.
These observations indicated that increasing the radius of curvature of the mold was an
effective and practical approach to reduce (or control) thickness variation at corners in low-
pressure VBO processing, especially when concave tooling was used. In addition, increasing the
corner curvature could also counteract the detrimental effects of increasing geometric complexity,
allowing parts with large corner angles to attain the same dimensional consistency as those with
smaller angles.
112
Figure 6-9. Thickness variation of laminates manufactured with various tool corner curvatures: (a)
concave corner laminates, (b) convex corner laminates
6.3.4. Pressure Strip Application
The pressure distribution conditions at a concave corner can be improved by using a
pressure strip (a silicone insert) placed between the bag-side release film and the breather. The
strip concentrates the pressure at the corner region, while potentially mitigating the effects of
consumable bridging. Laminates comprised of eight plies of prepreg B prepreg were fabricated
using pressure strips over molds with sharp corner radii (as in Sets I and II) and corner radii of
9.53 mm and 12.7 mm (as in Set IV). During layup, the pressure strip material was carefully
shaped into the form of the concave mold. Debulking was not performed.
113
Figure 6-10. Thickness variation of laminates manufactured without and with a pressure strip
Figure 6-10 presents the thickness variation of laminates produced with and without
pressure strips. For molds with corner radii of 0 mm and 9.53 mm, thickness variation was greatly
reduced by deployment of pressures strips, indicating that the insert redistributed pressure and
concentrated compaction force at corner regions, increasing the extent of conformation of the fiber
bed to the tool surface. However, the pressure strip did not affect laminates produced using the
mold with a corner radius of 12.7 mm, since the tool geometry already reduced the CoV close to
flange levels.
These results indicated that the application of pressure strips was an effective method for
reducing thickness variation in laminates produced over concave molds with relatively low corner
curvatures (0 mm and 9.53 mm). They are therefore well-suited to situations in which the part
geometry is tightly prescribed and cannot be locally adjusted by changing the mold geometry.
While the use of pressure strips effectively reduced CoVs in concave laminates, poorly
compacted areas (or laminate fillets) were sometimes observed at the boundary of the region of
application, as shown in Figure 6-11 (a). These defects were sensitive to the manner in which the
pressure strip was placed, and could be eliminated by carefully designed layup. However, they
114
reduced the robustness of the process. In addition, the use of pressure strips was also associated
with concentrated void distributions along the pressure strip, as shown in Figure 6-11 (b). These
voids were larger than those observed in previous laminates, and exhibited elongated ellipsoidal
shapes. Increasing the room-temperature vacuum hold time from four hours to eight hours
eliminated the voids. These defects can be attributed to the modified pressure distribution caused
by the pressure strip. The fillets corresponded to the edge of the strip, which experienced a local
and steep reduction in compaction pressure. The concentrated pressure caused by the strip could
also possibly decrease the air permeability of the prepreg, reducing the rate of air evacuation and,
for an equivalent ambient evacuation time, causing higher void contents. The longer room
temperature vacuum hold allowed more time for the evacuation of entrapped air, and consequently
decreased the void content.
Figure 6-11. Typical microscopic images of laminates manufactured with a pressure strip: (a)
laminate fillet at the end of the pressure strip, (b) void distribution
115
6.4. Conclusions
This chapter investigated several aspects of the OoA manufacture of VBO prepreg
laminates with corner geometries. First, concave and convex corner parts were manufactured for
various corner angles and laminate thickness levels and from two prepreg materials. Polished
sections revealed that for both concave and convex parts, defects were concentrated in corner
regions, and increasing geometric complexity led to a reduction in quality. However, defects levels
were always greater in concave laminates than in convex laminates. Microstructural data indicated
that two different mechanisms caused the dimensional non-uniformities. For concave laminates,
corner thickening was primarily associated with resin accumulation due to reduced compaction
pressure and the inability of the fiber bed to conform to the geometry of the tooling. Conversely, in
convex laminates, corner thinning was caused by uneven pressure distribution but counteracted by
the stiffness of the fiber bed. The void content was primarily affected by the prepreg age, with the
aged material exhibiting greater void content. However, void content was also associated with the
compaction quality, with greater laminate thicknesses leading to higher voids levels.
Based on these observations, three defect reduction strategies were considered:
intermediate debulking, variation in the mold corner radius, and the use of a pressure strip
intensifier. Debulking was an effective means of reducing void content in laminates, providing
evidence that air entrapped between plies during layup was a key source of voids. Increasing the
mold corner radius improved quality by allowing the fiber bed to conform to the tool and reducing
the potential for resin accumulation. The pressure strip effectively concentrated pressure in
concave corners, and improved the quality of laminates fabricated on molds with tight corner radii,
although air evacuation was partially impeded.
The results presented here can be used as guidelines for reducing defects and optimizing a
manufacturing cycle. For example, for a given laminate layup and geometry (corner angle), a
116
concave molding set-up may be converted to a convex one to reduce defect levels. The results
further indicated that this conversion is particularly important when transitioning from an
autoclave to an OoA environment (e.g., vacuum bag). If concave tooling is specifically required,
the local mold geometry can be modified to increase the corner radius and reduce resin
accumulation. Moreover, a pressure strip can be used to further control local defects. Interestingly,
the effects of increasing geometric complexity can be counteracted by relatively simple
modifications to the mold corner radius or by using a pressure strip, enabling the fabrication of
more aggressive geometries. Finally, for a generic set of prepreg, laminate, and molding
conditions, these results provide approximate predictions for laminate quality.
The shift from autoclave to non-autoclave manufacturing methods necessarily eliminates
the safeguards associated with elevated pressure. As a result, the successful expansion of OoA
manufacturing requires the development of optimized and robust processes, as well as more
careful attention to protocols. The present study reveals some of the fundamental phenomena
associated with low-pressure processing of corner laminates, and provides a basis for developing
effective manufacturing guidelines. The existing literature suggests that size and complexity scale-
up may also complicate other phenomena that govern quality, including air evacuation.
Furthermore, such scale-up may also lead to non-ideal temperature and pressure distributions
within the vacuum-bagged part. As a result, the insights developed here must be combined with
further research on factors affecting prepreg compaction, and with the development of more
accurate predictive simulations for compaction, cure, and microstructural quality.
117
CHAPTER 7. Conclusions and Future Work
7.1. Conclusions
The results presented in this work have demonstrated key aspects for recycling of
amine/epoxy composites using depolymerization and acid digestion at atmospheric pressure.
Results show that both depolymerization and acid digestion are effective for amine/epoxy
composite recycling. Generally, the increase in epoxy functionality and laminate thickness results
in longer epoxy dissolution time. Both recycling methods have negligible effects on recovered
fiber surface topography, however, acid digestion shows faster chemical reaction rate with
recovery of cleaner carbon fibers than depolymerization, indicating that acid digestion is more
suitable for amine-cured epoxy matrix dissolution.
Furthermore, rate-determining factors for acid digestion are identified and specific
strategies are developed to increase the recycling rate. This study discovers that the dissolution rate
of acid digestion is controlled primarily by diffusion, followed by reaction rate. To improve the
epoxy dissolution rate, two strategies – pre-treatment and shredding – are investigated, and both
are effective. This study also identifies the reaction mechanism for acid digestion and developed
specific catalytic conditions that improve the chemical reaction rate for acid digestion. By
understanding the mechanism of the chemical reaction and applying effective catalysts, the
crosslinked chemical bonds can be cleaved selectively, thus enabling the recovery and reuse of
useful polymer products. Moreover, the decomposed polymer components are recovered from the
chemical feedstock and reused in virgin epoxy resin formulations, thus closing the recycling loop
and possibly leading to an optimized solution to amine/epoxy composites recycling under
atmospheric pressure.
118
Lastly, several aspects of the OoA manufacture of VBO prepreg laminates with corner
geometries are investigated, and defects observed in the parts are identified and categorized. Three
defect reduction strategies are developed, including intermediate debulking, variation in the mold
corner radius, and the use of a pressure strip intensifier. These results provide guidelines for
reducing defects and optimizing a manufacturing cycle for parts with complex geometries,
supporting the successful expansion of OoA prepreg manufacturing.
7.2. Broader Implication
To achieve a high recycling rate for end-of-life CFRPs, an effective process technology is
required, and three factors must be considered. First, a process must yield valuable recyclates after
recycling. Current industrial scale pyrolysis recovers only short carbon fibers at end-of-life with
reduced mechanical properties compared to virgin carbon fibers. Those short fibers are typically
incorporated into bulk molding compounds (BMCs) or stitched mats, both of which embody much
lower value than CFRPs with continuous carbon fiber reinforcement, and thus are unlikely to
eligible for subsequent recycling. Note that metals can be recycled repeatedly with negligible loss
of material properties. Future recycling technologies should aim to recover near-virgin quality
continuous carbon fibers (without loss of organization), as well as high-value matrix residues.
Second, the recycling process must be economically feasible. In addition to diverting
composites from landfills to comply with environmental legislation, composite recycling should
include a reduction in cost/energy associated with producing virgin products to attract more
recycling interests. For example, recycled aluminum and steel require 95% and 56% less energy,
respectively [91]. The recovery and reuse of high-value matrix residues after chemical recycling
not only closes the recycling loop, but also potentially reduce the overall cost of the process.
Finally, potential for scale-up must be viable. Although valuable recyclates can be
recovered via chemical recycling, current lab-scale processes require high pressure or unsafe
119
reagents (corrosive or explosive). These conditions cannot be deployed readily in industry and
must be modified/replaced. Mild recycling technologies also recover more value from end-of-life
composites (high quality fibers and useful matrix residues), enabling the economic feasibility of
recycling. The short production turnaround of chemical recycling is also a critical factor for entry
into industrial scale recycling. This need is especially acute for advanced composites, which
embody high intrinsic value but are difficult to recycle because of high crosslink densities.
Although this study has found conditions for recovery of near-virgin quality carbon fibers
and useful decomposed polymers from amine/epoxy composites using oxidative acid digestion,
many optimization challenges remain because the current digestion condition uses hydrogen
peroxide that is an expensive and explosive reagent and cannot be deployed on an industrial scale.
The detailed reaction mechanism identification for oxidative acid digestion underlies the
background for further improvements through which safe replacements for hydrogen peroxide are
being investigated, as well as target catalysts that yield faster reaction rates. Doing so will yield a
more sustainable recycling solution for end-of-life amine/epoxy composites, as well as for
production scrap waste, at atmospheric pressure on an industrial scale.
7.3. Future Work
7.3.1. Improved Recycling Conditions for Composite
The viability of acid digestion for amine-cured epoxies with high crosslink density affords
opportunity to recover and recycle both carbon fibers and useful epoxy parts from composites.
However, a truly effective recycling process must not only yield valuable recyclates, but also has
potential for scale-up and be economically feasible. Unfortunately, H2O2 used in the current
oxidative acid digestion is an expensive and explosive reagent that cannot be deployed on an
industrial scale. Furthermore, the chemical bonds cleavage is not selectively in the presence of
120
H2O2, and thus the recovered value from the polymer matrix is limited. Safer oxygen source
alternatives to H2O2 must be developed to achieve a truly effective recycling method.
The future efforts will be devoted to seeking safer oxygen source alternatives to H2O2 for
oxidative acid digestion that can be deployed to recover high quality fibers and more valuable
polymers from amine/epoxy composites, as well as to developing more effective catalysts to yield
faster reactions. Our previous study on oxidative peroxide condition has confirmed that the
mechanism proceeds by oxygen atom transfer (OAT), followed by elimination, and catalysts
capable of accelerating either the OAT or elimination step (or both) can accelerate the degradation
of the composite matrices. Herein, future work aims to develop an aerobic digestion method in
which we substitute the peroxide in oxidative digestion with a safe oxygen source and use catalysts
to accelerate the two reaction steps.
First, the aerobic digestion condition is applied to small molecule models of amine-cured
epoxy matrices. Prospective catalysts for the two reaction steps are being screened and evaluated,
and digested molecules are being characterized. Then, the aerobic digestion conditions are applied
to amine-cured neat epoxies to evaluate the effectiveness aerobic digestion on crosslinked
materials. Digested epoxies will be characterized by NMR, GC-MS and MALDI-MS. Furthermore,
the aerobic digestion condition will be applied to commercial prepregs that have reach
manufacture’s stated out-life and cured composites fabricated from lab-made and commercial
prepregs. No corrosive or toxic reagent are used for aerobic digestion and moderate reaction
condition will be applied, resulting in a sustainable and cost-efficient method. The selectively
decomposed polymer components via catalytic aerobic digestion will also retain high values and
can be reused for new applications. Preliminary data show that clean carbon fiber fabrics can be
recovered from partially cured aerospace prepregs (Figure 7-1), however, the current aerobic
121
digestion process is not fast enough to meet industrial demand and more work is needed to
increase the chemical reaction rate.
Moreover, the previous study has also demonstrated that the dissolution rate for a high
crosslinked polymer matrix is controlled by both the rates of chemical reaction and diffusion.
Therefore. I have evaluated two strategies to increase the diffusion rate during recycling, and both
are effective. However, the current strategies involve the use of either high temperatures that
increase the cost of recycling or mechanical forces that reduce the value of continuous fibers.
Therefore, we need to continue to investigate more effective methods to improve the diffusion rate
for polymer matrices with high crosslink densities. By developing effective catalysts and
permeabilization agents, I expect faster reaction rates will be possible, sufficient to meet industrial
demand.
Figure 7-1. Fibers with residue (a) and matrix (b) recovered from 110 °C cured prepreg by DMSO
wash, clean fibers (c) and matrix (d) recovered after applying aerobic digestion to sample (c).
122
7.3.2. Prepreg Recycling
Another technical obstacle to the continued growth and acceptance of composites has
recently drawn our attention – prepreg waste. Unlike composite waste that will become a problem
in 20-30 years, prepreg waste, however, is being continuously produced and needs an immediate
solution. Boeing and Airbus each is generating as much as 1 million pounds of carbon fiber
prepreg waste each year from 787 and A350 XWB production [115]. Prepreg scrap rate in some
manufacturing processes can be 40% of the material used. If we include the entire supply chain for
these planes and other industries, the total can be over 4 million pounds per year. Currently,
prepreg waste is cured and subsequently milled to use as fillers, or cut into small-size chips and
compression molded into bulk molding compounds [116]. Little value of the continuous
fibers/fabrics is recovered and market demand for those products is limited. Recycling of prepreg
waste has been reported, which uses a thermal process at 500 °C to recover the fibers [117-118].
The recovered fibers, however, retain < 70% tensile strength, and most of the value of the resin
matrix is lost.
Consequently, future work should also aim to tackle the issue of prepreg recycling and
come up with a technically and economically feasible solution for the industry to manage the scrap
prepreg stream. Preliminary data show that clean carbon fiber fabrics and matrices can be
recovered from expired aerospace prepregs (Figure 7-2) by DMSO wash. If the expired prepreg
has achieved a significant degree of cure, the prepreg waste recycling will be an extension of the
previously discussed aerobic digestion for cured composite wastes. Although the resin components
on the prepreg waste exist in various state of cure, they contain chemical structures similar to cured
polymer matrices but in a much less crosslinked state. Therefore, expired prepregs are expected to
be an easier target for aerobic digestion.
123
Figure 7-2. Clean fibers (a) and matrix (b) recovered from expired prepregs
7.3.3. Recyclates Reuse
Through effective chemical recycling, near-virgin quality carbon fibers in the original
fabric/array forms could be recovered, as well as useful polymer matrix. However, reuse
approaches for those recyclates are recycling process dependent, thus current reuse strategies are
not applicable to our methods. For example, current industrial scale recycling for CFRPs focused
on pyrolysis. ELG Carbon Fibre uses pyrolysis to recover carbon fibers from composite materials.
They can process 2,000 metric tonnes of waste and generate 1,000 metric tonnes of reclaimed
carbon fibers per year using a patented pyrolysis process [81]. However, pyrolysis produces
chopped, milled or pelletized fiber products, which are used as fillers or made into rolls of chopped
fiber mats. These limitations will be overcome by our recycling method that aim to recover
continuous carbon fiber fabrics. Reuse application for continuous fibers must be developed upon
successful aerobic digestion.
Moreover, this chemical mixture after recycling must also be addressed to prevent adverse
environmental impact and make the process sustainable. Reuse of decomposed polymer matrix
residue after chemical recycling have been reported in the literature, however, the bond cleavage
from our study differs from the methods reported in previous studies, leading to recovery of
124
decomposed matrices with distinct molecular structures and functional groups. Consequently, the
matrix reuse protocols need further evaluation and adjustment.
7.3.4. Scale-up Potential
Future work for the recycling project should also look into the possibility of processing real
end-of-life parts from industry that are typically large in dimensions and have complex structures.
Doing so is expected to be more difficult than recycling lab-scale parts for two reasons. First,
composites must be sorted and separated into scraps before recycling. Industrial composite parts
can consist of multiple components in addition to fibers and matrices. For example, sandwich
structures are commonly used in composite parts to further reduce the weight while maintaining
the mechanical strength. Sections other than fiber reinforcements and polymer matrices must be
removed before recycling because the chemical process is designed only for removing the polymer
matrix from the fibers. Second, industrial parts with large and complex structures require more
effective recycling methods. Recycling of industrial parts might raise questions for the effectively
of the current lab-scale recycling methods because the diffusion rate is expected to be slower in
larger parts. Components that cannot be easily removed from the composites, such as tougheners,
paints, and fillings, will also have negative effects on the recycling process.
Lastly, a cost analysis should be performed as the recycling technology maturity gets close
to the scale-up stage. Boeing pointed out that as soon as the retired airplane is designated for scrap,
it will immediately lose as much as 75% of its value [9]. If it is cheaper to leave the end-of-life
parts in the landfill than to recycle them, the economic incentive for the industry to recycle
composites will be limited.
125
Appendix A. Supplementary Materials for Neat Epoxy Dissolution
A.1. Glass Transition Temperature
Glass transition temperatures (Tg’s) of epoxies cured using 3,3’-DDS (3,3’-
diamonodiphenyl sulfone) and M-DEA (4,4’-methylenebis (2,6-diethylaniline)) were measured
from MDSC (modulated differential scanning calorimetry, TA Instruments Q2000). Effects of
amine/epoxy stoichiometric ratio and epoxy functionality on Tg’s were evaluated.
Figure A-1. Comparison of Tg’s of epoxies cured using DDS and M-DEA with variations in (a)
amine/epoxy ratio and (b) epoxy functionality
A.2. Proton NMR spectroscopy
1D HNMR spectra of benzyl alcohol, 3,3’-DDS (3,3’-diamonodiphenyl sulfone), bi-
functional epoxy monomer (DGEBA, diglycidyl ether of bisphenol A), tri-functional epoxy
monomer (triglycidyl of para-aminophenol), and tetra-functional epoxy monomer (tetraglycidyl-
4,4'-methylenebisbenzenamine) were analyzed [119-122].
126
Figure A-2. 1D HNMR spectrum of benzyl alcohol in DMSO at room temperature (32 scans)
Figure A-3. 1D HNMR spectrum of 3,3’-DDS (3,3’-diamonodiphenyl sulfone) in DMSO at room
temperature (32 scans)
127
Figure A-4. 1D HNMR spectrum of bi-functional epoxy monomer (DGEBA, diglycidyl ether of
bisphenol A) in DMSO at room temperature (32 scans)
Figure A-5. 1D HNMR spectrum of tri-functional epoxy monomer (triglycidyl of para-
aminophenol) in DMSO at room temperature (32 scans)
128
Figure A-6. 1D HNMR spectrum of tetra-functional epoxy monomer (tetraglycidyl-4,4'-
methylenebisbenzenamine) in DMSO at room temperature (32 scans)
A.3. MALDI-TOF Mass Spectrometry
Mass spectra of 3,3’-DDS cured tri-functional epoxy dissolution products from
depolymerization and acid digestion were analyzed using MALDI-TOF MS.
Figure A-7. Analytical results of MALDI-TOF/MS of the dissolution products of 3,3’-DDS-cured
tri-functional epoxy: (a) depolymerization (34 h), (b) acid digestion (6 h)
129
A.4. Gas Chromatography-Mass Spectrometry (GC-MS)
Mass spectra of 3,3’-DDS cured bi-functional epoxy dissolution products from
depolymerization and acid digestion were analyzed using GC-MS. The mass spectrum in Figure
S8 (b) was compared with literature [123]. Bisphenol A structure was distinguishable from the
spectrum, indicating that the aromatic structures of the epoxies were preserved during dissolution.
Figure A-8. Analytical results of GC/MS of the dissolution products of 3,3’-DDS-cured bi-
functional epoxy: (a) mass chromatogram of dissolution products, (b) mass spectrum of product
bisphenol A
130
A.5. Bond dissociation energies (BDEs)
Bond dissociation energies (BDEs) for C-O, C-N and C-S bonds were discussed to
determine the cleavable site in the crosslinking of cured epoxies.
Table A-1. Bond dissociation energies (BDEs) for C-O, C-N and C-S bonds (boldface =
dissociated group) [78]
Bond BDE (kJ/mol) Bond BDE (kJ/mol)
CH3–OCH3
351.9 ± 4.2
CH3–NHC6H5
298.7 ± 8.4
CH3–OC2H5
349.8 ± 4.2
CH3–N(CH3)C6H5
296.2 ± 8.4
CH3–OnC3H7
354.8 ± 6.3
C2H5–NHC6H5
296.6 ± 8.4
CH3–OC6H5
263.2 ± 4.2
CH3NH–C6H5
420.9 ± 10.5
C2H5–OC6H5
269.0 ± 4.8
(CH3)2N–C6H5
389.9 ± 10.5
CH3O–C6H5
418.8 ± 5.9
CH3SO2–C6H5
344.3 ± 8.4
C2H5O–C6H5
416.7 ± 5.4
131
Appendix B. Supplementary Materials for Composites Recycling
B.1. Energy-dispersive X-ray Spectroscopy (EDS)
Figure B-1. (a) EDS spectrum of the residue on carbon fiber surface after depolymerization, (b)
EDS line scan crossing the residue. Phosphorus (P) signal in the spectra confirms the source of the
residue to be catalyst K3PO4.
B.2. Scanning Electron Microscopy (SEM) Images
Figure B-2. SEM images of virgin (a) and recovered (b) glass fiber via acid digestion. Sizing was
removed after reaction. No defect or residue were observed on fiber surfaces.
132
Figure B-3. SEM images of the recovered carbon fibers from a laminate made of 5320-1/8HS via
acid digestion with pre-treatment. No defect or residue were observed on fiber surfaces.
Figure B-4. SEM images of recovered carbon fiber from commercial composite waste (ELG
Carbon Fibre Ltd.) via acid digestion. No defect or residue were observed on fiber surfaces.
133
B.3. Shredded Composites Waste Before and After Acid Digestion
Figure B-5. Recovered carbon fiber from shredded composite waste: a) and b) before acid
digestion, c) and d) after acid digestion
134
Appendix C. Supplementary Materials for Decomposed Polymers Reuse
Figure C-1. Effect of commercial accelerators on curing reactions of anhydride-based (a) bi-, (b)
tri- and (c) tetra-functional epoxies
Figure C-2. Effect of commercial accelerator concentration on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies
135
Figure C-3. Effect of recovered matrix residue concentration on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies (without a commercial accelerator)
Figure C-4. Comparison of (a) curing reactions and (b) Tg values for anhydride-based bi-functional
epoxies with matrix residues without and with ScCl3 catalyst
136
Figure C-5. Effect of recovered matrix residue state on (a) curing reactions and (b) Tg values of
anhydride-based bi-functional epoxies
Figure C-6. Effect of recovered matrix residue functionality on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies
137
Figure C-7. (a) Curing reactions and (b) Tg values of DGEBA/recovered matrix residue systems
Figure C-8. Full FTIR spectra of DGEBA/recovered matrix residue systems after a cure cycle.
138
Figure C-9. Effect of recovered matrix concentration and state on (a) curing reactions and (b) Tg
values of anhydride-based bi-functional epoxies (with accelerators)
Figure C-10. Effect of recovered matrix concentration and state on (a) curing reactions and (b) Tg
values of amine-based bi-functional epoxies
139
References
1. M. Holmes, Global carbon fibre market remains on upward trend, Reinf. Plast. 58 (2014)
38–45. doi:10.1016/S0034-3617(14)70251-6.
2. M. Holmes, High volume composites for the automotive challenge, Reinf. Plast. 61 (2017)
294–298. doi:10.1016/j.repl.2017.03.005.
3. W.J. Joost, Reducing Vehicle Weight and Improving U.S. Energy Efficiency Using
Integrated Computational Materials Engineering, Integr. Comput. Mater. Eng. 64 (2012).
doi:10.1007/s11837-012-0424-z.
4. C. Air, R. Board, N. Lutsey, Review of technical literature and trends related to automobile
mass-reduction technology, (2010).
5. Carbon Fiber Market Analysis by Raw Material.
https://www.grandviewresearch.com/industry-analysis/carbon-fiber-market-analysis [last
accessed Dec 2018]
6. Global Wind Report 2015, Global Wind Energy Council, (2015).
7. S. Black, Composites recycling is gaining traction, CompositesWorld
https://www.compositesworld.com/blog/post/composites-recycling-is-gaining-traction [last
accessed Nov 2018]
8. Recycled carbon fiber update: Closing the CFRP lifecycle loop.
https://www.compositesworld.com/articles/recycled-carbon-fiber-update-closing-the-cfrp-
lifecycle-loop [last accessed Dec 2018]
9. W. Carberry, Airplane recycling efforts benefit Boeing operators. Aero Quarterly QTR 04
(2008).
10. P.D. Mangalgiri, Composite materials for aerospace applications, Bull. Mater. Sci. 22 (1999)
657–664. doi:10.1007/BF02749982.
140
11. Dell Uses 1 Million Pounds of Recycled Carbon Fiber This Year.
http://www.themoldingblog.com/2017/12/06/dell-uses-1-million-pounds-of-recycled-carbon-
fiber-this-year/ [last accessed Dec 2018]
12. Y. Ma, D. Kim, S.R. Nutt, Chemical treatment for dissolution of amine-cured epoxies at
atmospheric pressure, Polym. Degrad. Stab. 146 (2017) 240–249.
doi:10.1016/j.polymdegradstab.2017.10.014.
13. Y. Ma, S.R. Nutt, Chemical treatment for recycling of amine/epoxy composites at
atmospheric pressure, Polym. Degrad. Stab. 153 (2018).
doi:10.1016/j.polymdegradstab.2018.05.011.
14. C.A. Navarro, E.A. Kedzie, Y. Ma, K.H. Michael, S.R. Nutt, T.J. Williams, Mechanism and
catalysis of oxidative degradation of fiber-reinforced epoxy composites, Top. Catal. (2018)
1–6. doi:10.1007/s11244-018-0917-2.
15. Y. Ma, C.A. Navarro, T.J. Williams, S.R. Nutt, Recovery and reuse of acid digested
amine/epoxy-based composite matrices, in preparation. (2018)
16. Y. Ma, D. Kim, S.R. Nutt, A parametric study of the recyclability of carbon fiber reinforced
polymers. Proc of the Society for the Advancement of Material and Process Engineering
(SAMPE) Technical Conference, Long Beach, CA, May 23–26, (2016).
17. Y. Ma, D. Kim, T.J. Williams, S.R. Nutt, Recycling of carbon fiber composites using
chemical treatment: reaction characterization and optimization. Proc of the 2017 Society for
the Advancement of Material and Process Engineering (SAMPE) Technical Conference,
Seattle, WA, May 22–25, (2017).
18. Y. Ma, T.J. Williams, S.R. Nutt, Recycling of amine/epoxy composites using chemical
treatment at atmospheric pressure. Proc of the 2018 Society for the Advancement of Material
141
and Process Engineering (SAMPE) Technical Conference, Long Beach, CA, May 21-24,
(2018).
19. Y. Ma, D. Kim, M-S Wu, L. Bai, B. Qing, J. He, X. Li, S.R. Nutt, B. Liang, processing and
performance of recyclable composites for wind turbine blades, Proc of the 2017 TechConnect
World Innovation Conference, Washington D.C., May 14-17, (2017).
20. Y. Ma, T. Centea, S.R. Nutt, Defect reduction strategies for the manufacture of contoured
laminates using vacuum BAG-only prepregs, Polym. Compos. 38 (2017) 2016–2025.
doi:10.1002/pc.23773.
21. H. Lee, K. Neville. Handbook of Epoxy Resins, McGraw-Hill, (1967).
22. W.L. Hawkins, Polymer Degradation and Stabilization, Springer-Verlag, (1984).
doi:10.1007/978-3-642-69376-2. ISBN:0387128514.
23. Directive 2000/53/EC on End of Life Vehicles, European, (2000).
24. The Technical Policy for the Recovery of Automobile Products, China, (2006).
25. S.J. Pickering, Recycling technologies for thermoset composite materials—current status,
Compos. Part A Appl. Sci. Manuf. 37 (2006) 1206–1215.
doi:10.1016/j.compositesa.2005.05.030.
26. G. Oliveux, L.O. Dandy, G.A. Leeke, Current status of recycling of fibre reinforced polymers:
Review of technologies, reuse and resulting properties, Prog. Mater. Sci. 72 (2015) 61–99.
doi:10.1016/j.pmatsci.2015.01.004.
27. J. Palmer, O.R. Ghita, L. Savage, K.E. Evans, Successful closed-loop recycling of thermoset
composites, Compos. Part A Appl. Sci. Manuf. 40 (2009) 490–498.
doi:10.1016/j.compositesa.2009.02.002.
142
28. J. Howarth, S.S.R. Mareddy, P.T. Mativenga, Energy intensity and environmental analysis of
mechanical recycling of carbon fi bre composite, J. Clean. Prod. 81 (2014) 46–50.
doi:10.1016/j.jclepro.2014.06.023.
29. A.C.M. Castro, M.C.S. Ribeiro, J. Santos, J.P. Meixedo, F.J.G. Silva, A. Fiúza, M.L. Dinis,
M.R. Alvim, Sustainable waste recycling solution for the glass fibre reinforced polymer
composite materials industry, Constr. Build. Mater. 45 (2013) 87–94.
doi:10.1016/j.conbuildmat.2013.03.092.
30. J.M. Gosau, T.F. Wesley, R.E. Allred, Integrated composite recycling process, Proc of the
38th Society for the Advancement of Material and Process Engineering (SAMPE) Technical
Conference, Dallas, TX, November 7–9, (2006).
31. S.J. Pickering, T.A. Turner, F. Meng, C.N. Morris, J.P. Heil, K.H. Wong, S. Melendi,
Developments in the fluidised bed process for fibre recovery from thermoset composites,
Proc of the Composites and Advanced Materials Expo (CAMX) Conference Proceedings.
Dallas, TX, October 26–29, (2015)
32. K. Ushikoshi, N. Komatsu, M. Sugino, Recycling of CFRP by pyrolysis method, Journal-
Society of Materials Science Japan. 44 (1995) 428–428.
33. R.S. Ginder, S. Ozcan, Controlled pyrolysis: a case study of scalable glass and carbon fiber
composite recycling technology, Proc of the Composites and Advanced Materials Expo
(CAMX) Conference Proceedings. Dallas, TX, October 15–18, (2018)
34. C. Morin, A. Loppinet-Serani, F. Cansell, C. Aymonier, Near- and supercritical solvolysis of
carbon fibre reinforced polymers (CFRPs) for recycling carbon fibers as a valuable resource:
State of the art, J. Supercrit. Fluids. 66 (2012) 232–240. doi:10.1016/j.supflu.2012.02.001.
35. G. Jiang, S.J. Pickering, E.H. Lester, T.A. Turner, K.H. Wong, N.A. Warrior,
Characterisation of carbon fibres recycled from carbon fibre / epoxy resin composites using
143
supercritical n-propanol, Compos. Sci. Technol. 69 (2009) 192–198.
doi:10.1016/j.compscitech.2008.10.007.
36. R. Piñero-Hernanz, C. Dodds, J. Hyde, J. García-Serna, M. Poliakoff, E. Lester, M.J. Cocero,
S. Kingman, S. Pickering, K.H. Wong, Chemical recycling of carbon fibre reinforced
composites in nearcritical and supercritical water, Compos. Part A Appl. Sci. Manuf. 39
(2008) 454–461. doi:10.1016/j.compositesa.2008.01.001.
37. G. Oliveux, L.O. Dandy, G.A. Leeke, Degradation of a model epoxy resin by solvolysis
routes, Polym. Degrad. Stab. 118 (2015) 96–103.
doi:10.1016/j.polymdegradstab.2015.04.016.
38. Y. Bai, Z. Wang, L. Feng, Chemical recycling of carbon fibers reinforced epoxy resin
composites in oxygen in supercritical water, Mater. Des. 31 (2010) 999–1002.
doi:10.1016/j.matdes.2009.07.057.
39. K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Degradation of crosslinked unsaturated
polyesters in sub-critical water, Polym. Degrad. Stab. 92 (2007) 317–322.
doi:10.1016/j.polymdegradstab.2006.10.008.
40. L. Yuyan, L. Li, W. Songquan, L. Yuting, The Decomposition of Bisphenol A Diglycidyl
Ether / Alicyclic Amine Hardened Networks in Near-critical Water, Polym. Polym. Compos.
20 (2012) 809–816.
41. T. Nakagawa, S. Matsugi, S. Hirota, T. Miyazaki, K. Shibata, T. Itoh, H. Yano, N.
Yabunouchi, Y. Yasuda, M. Hidaka, J. Matsui, Enhanced and Horizontal Recycling
Technology of FRP using Sub-Critical Water (2), J. of network polym, Japan, 30 (2009) 83-
90.
144
42. I. Okajima, M. Hiramatsu, Y. Shimamura, T. Awaya, T. Sako, Chemical recycling of carbon
fiber reinforced plastic using supercritical methanol, J. Supercrit. Fluids. 91 (2014) 68–76.
doi:10.1016/j.supflu.2014.04.011.
43. A. Kamimura, Y. Akinari, T. Watanabe, K. Yamada, F. Tomonaga, Efficient chemical
recycling of waste fiber-reinforced plastics: use of reduced amounts of
dimethylaminopyridine and activated charcoal for the purification of recovered monomer, J
Mater Cycles Waste Manag. 12 (2010) 93–97. doi:10.1007/s10163-010-0276-y.
44. H. Jie, H. Ke, Q. Wenjie, Z. Zibin, Process analysis of depolymerization polybutylene
terephthalate in supercritical methanol, 91 (2006) 2527–2531.
doi:10.1016/j.polymdegradstab.2006.01.014.
45. R. Piñero-Hernanz, J. García-Serna, C. Dodds, J. Hyde, M. Poliakoff, M.J. Cocero, S.
Kingman, S. Pickering, E. Lester, Chemical recycling of carbon fibre composites using
alcohols under subcritical and supercritical conditions, J. Supercrit. Fluids. 46 (2008) 83–92.
doi:10.1016/j.supflu.2008.02.008.
46. J.R. Hyde, E. Lester, S. Kingman, S. Pickering, K.H. Wong, Supercritical propanol, a
possible route to composite carbon fibre recovery: A viability study, Compos. Part A Appl.
Sci. Manuf. 37 (2006) 2171–2175. doi:10.1016/j.compositesa.2005.12.006.
47. H. Yan, C. Lu, D. Jing, X. Hou, Chemical degradation of TGDDM / DDS epoxy resin in
supercritical 1-propanol : Promotion effect of hydrogenation on thermolysis, Polym. Degrad.
Stab. 98 (2013) 2571–2582. doi:10.1016/j.polymdegradstab.2013.09.026.
48. H. Yan, C. Lu, D. Jing, C. Chang, N. Liu, X. Hou, Recycling of carbon fibers in epoxy resin
composites using supercritical 1-propanol, New Carbon Mater. 31 (2016) 46–54.
doi:10.1016/S1872-5805(16)60004-5.
145
49. H.U. Sokoli, J. Beauson, M.E. Simonsen, A. Fraisse, P. Brøndsted, E.G. Søgaard, Optimized
process for recovery of glass- and carbon fibers with retained mechanical properties by means
of near- and supercritical fluids, J. Supercrit. Fluids. 124 (2017) 80–89.
doi:10.1016/j.supflu.2017.01.013.
50. L. Henry, A. Schneller, J. Doer, W.M. Mueller, C. Aymonier, S. Horn, Semi-continuous flow
recycling method for carbon fibre reinforced thermoset polymers by near- and supercritical
solvolysis, 133 (2016) 264–274. doi:10.1016/j.polymdegradstab.2016.09.002.
51. R. Morales, I. Mitsuru, Carbon fiber recovery using water and benzyl alcohol in subcritical
and supercritical conditions for chemical recycling of thermoset composite materials, (2015)
369–379. doi:10.1007/s10163-014-0252-z.
52. I. Okajima, K. Watanabe, S. Haramiishi, M. Nakamura, Recycling of carbon fiber reinforced
plastic containing amine-cured epoxy resin using supercritical and subcritical fluids, J.
Supercrit. Fluids. 119 (2017) 44–51. doi:10.1016/j.supflu.2016.08.015.
53. fJ.A. Onwudili, E. Yildirir, P.T. Williams, Catalytic Hydrothermal Degradation of Carbon
Reinforced Plastic Wastes for Carbon Fibre and Chemical Feedstock Recovery, (2015) 87–
93. doi:10.1007/s12649-013-9204-4.
54. Y. Liu, J. Liu, Z. Jiang, T. Tang, Chemical recycling of carbon fibre reinforced epoxy resin
composites in subcritical water: Synergistic effect of phenol and KOH on the decomposition
efficiency, Polym. Degrad. Stab. 97 (2012) 214–220.
doi:10.1016/j.polymdegradstab.2011.12.028.
55. J. Liu, K. Wang, L. Ma, T. Tang, Insight into the role of potassium hydroxide for accelerating
the degradation of anhydride-cured epoxy resin in subcritical methanol, J. Supercrit. Fluids.
107 (2016) 605–611. doi:10.1016/j.supflu.2015.07.022.
146
56. C.C. Knight, C. Zeng, C. Zhang, B. Wang, C.C. Knight, C. Zeng, C. Zhang, B. Wang,
Recycling of woven carbon-fibre-reinforced polymer composites using supercritical water
Recycling of woven carbon-fibre-reinforced polymer composites using supercritical water,
Environ. Technol. ISSN. 33 (2012) 639–644. doi:10.1080/09593330.2011.586732.
57. Y. Wang, X. Cui, H. Ge, Y. Yang, Y. Wang, C. Zhang, J. Li, T. Deng, Z. Qin, X. Hou,
Chemical recycling of carbon fiber reinforced epoxy resin composites via selective cleavage
of the carbon–nitrogen bond, ACS Sustain. Chem. Eng. 3 (2015) 3332–3337.
doi:10.1021/acssuschemeng.5b00949.
58. T. Deng, Y. Liu, X. Cui, Y. Yang, S. Jia, Y. Wang, C. Lu, D. Li, R. Cai, X. Hou, Cleavage of
C–N bonds in carbon fiber/epoxy resin composites, Green Chem. 17 (2015) 2141–2145.
doi:10.1039/C4GC02512A.
59. T. Liu, M. Zhang, X. Guo, C. Liu, T. Liu, J. Xin, J. Zhang, Mild chemical recycling of
aerospace fiber/epoxy composite wastes and utilization of the decomposed resin, Polym.
Degrad. Stab. 139 (2017) 20–27. doi:10.1016/j.polymdegradstab.2017.03.017.
60. W. Dang, M. Kubouchi, H. Sembokuya, K. Tsuda, Chemical recycling of glass fiber
reinforced epoxy resin cured with amine using nitric acid, Polymer (Guildf). 46 (2005) 1905–
1912. doi:10.1016/j.polymer.2004.12.035.
61. W Dang, M. Kubouchi, S. Yamamoto, H. Sembokuya, K. Tsuda, An approach to chemical
recycling of epoxy resin cured with amine and nitric acid., Polymer (Guildf). 43 (2002)
2953–2958. doi: 10.1016/S0032-3861(02)00100-3.
62. Y. Liu, L. Meng, Y. Huang, L. Lixin, Method of Recovering the Fibrous Fraction of Glass /
Epoxy Composites, J. Reinf. Plast. Compos. 25 (2006) 1525–1533.
doi:10.1177/0731684406066748.
147
63. P. Feraboli, H. Kawakami, B. Wade, F. Gasco, L. DeOto, A. Masini, Recyclability and
reutilization of carbon fiber fabric/epoxy composites, J. Compos. Mater. 46 (2011) 1459–
1473. doi:10.1177/0021998311420604.
64. J.D. Tanks, M. Kubouchi, Y. Arao, Influence of network structure on the degradation of poly
( ether ) amine-cured epoxy resins by inorganic acid, Polym. Degrad. Stab. 157 (2018) 153–
159. doi:10.1016/j.polymdegradstab.2018.10.011.
65. T. Iwaya, S. Tokuno, M. Sasaki, M. Goto, K. Shibata, Recycling of fiber reinforced plastics
using depolymerization by solvothermal reaction with catalyst, J. Mater. Sci. 43 (2008)
2452–2456. doi:10.1007/s10853-007-2017-8.
66. M. Nakagawa et al., Assessment of effects of carbon fibers recovered from CFRP with
depolymerization under ordinary pressure on the environmental (LCA), Proc of the 60th
Society of Polymer Science Japan (SPSJ) Annual Meeting, Japan, May 25–27, 6 (2011) 2088.
67. M. Nakagawa, K. Kasuga, K. Aoyagi, K. Ishihara, Y. Ikeda, CFRP recycling technology
using depolymerization under ordinary pressure, Proc of the American Society for
Composites (ASC) 29th Technical Conference, San Diego, CA, Sep 8–10, (2014).
68. P. Xu, J. Li, J. Ding, Chemical recycling of carbon fibre/epoxy composites in a mixed
solution of peroxide hydrogen and N,N-dimethylformamide, Compos. Sci. Technol. 82
(2013) 54–59. doi:10.1016/j.compscitech.2013.04.002.
69. J. Li, P. Xu, Y. Zhu, J. Ding, L. Xue, Y. Wang, A promising strategy for chemical recycling
of carbon fiber / thermoset composites : self-accelerating decomposition in a mild oxidative
system, Green Chem. 14 (2012) 3260–3263. doi:10.1039/c2gc36294e.
70. M. Das, S. Varughese, A Novel Sonochemical Approach for Enhanced Recovery of Carbon
Fiber from CFRP Waste Using Mild Acid-Peroxide Mixture, ACS Sustain. Chem. Eng. 4
(2016) 2080–2087. doi:10.1021/acssuschemeng.5b01497.
148
71. M. Das, R. Chacko, S. Varughese, An Efficient Method of Recycling of CFRP Waste Using
Peracetic Acid, ACS Sustain. Chem. Eng. 6 (2018) 1564−1571 Research.
doi:10.1021/acssuschemeng.7b01456.
72. L. Chiao, Mechanistic reaction kinetics of 4, 4’-diaminodiphenyl sulfone cured tetraglycidyl-
4, 4’-diaminodiphenylmethane epoxy resins, Macromolecules. 23 (1990) 1286–1290.
doi:10.1021/ma00207a010.
73. C.C. Riccardi, R.J.J. Williams, A kinetic scheme for an epoxy-amine reaction with
simultaneous etherification, J. Appl. Polym. Sci. 32 (1986) 3445–3456.
doi:10.1002/app.1986.070320208.
74. C.J. de Bakker, N.A. St John, G.A. George, Simultaneous differential scanning calorimetry
and near-infra-red analysis of the curing of tetraglycidyldiaminodiphenylmethane with
diaminodiphenylsulphone, Polymer (Guildf). 34 (1993) 716–725. doi:10.1016/0032-
3861(93)90353-C.
75. R.J. Morgan, E.T. Mones, The cure reactions, network structure, and mechanical response of
diaminophenyl sulfone ‐ cured tetraglycidyl 4,4” diaminodiphenyl methane epoxies, App.
Polym. Sci. 33 (1987) 999–102. doi: 10.1002/app.1987.070330401.
76. P.A. Schweitzer, Corrosion Engineering Handbook, CRC Press, (1996). ISBN:0824797074.
77. S.L. Buchwalter, L.L. Kosbar, Cleavable epoxy resins: design for disassembly of a
thermoset., J. Polym. Sci. Part A Polym. Chem. 34 (1996) 249–260. doi:10.1002/(SICI)1099-
0518(19960130)34:2<249::AID-POLA11>3.0.CO;2-Q.
78. Y.-R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC Press, 2007.
doi:10.1201/9781420007282.
79. British Standard ISO 11566, Carbon fibre-Determination of the tensile properties of single-
filament specimens, (1996)
149
80. J.L. Zhou, C. Cheng, H. Zhang, Z.Y. Sun, S. Zhu, M.H. Yu, Dissolution behaviour of
polyethersulfone in diglycidyl ether of bisphenol-A epoxy resins, IOP Conf. Ser. Mater. Sci.
Eng. 213 (2017) 12038. doi:10.1088/1757-899X/213/1/012038.
81. M. Holmes, Recycled carbon fiber composites become a reality, Reinf. Plast. 62 (2018) 148–
153. doi:10.1016/j.repl.2017.11.012.
82. S. Pimenta, S.T. Pinho, The influence of micromechanical properties and reinforcement
architecture on the mechanical response of recycled composites, Compos. Part A Appl. Sci.
Manuf. 56 (2014) 213–225. doi:10.1016/j.compositesa.2013.10.013.
83. M. Dauguet, O. Mantaux, N. Perry, Y.F. Zhao, Recycling of CFRP for high value
applications: Effect of sizing removal and environmental analysis of the Super Critical Fluid
Solvolysis, Procedia CIRP. 29 (2015) 734–739. doi:10.1016/j.procir.2015.02.064.
84. G. Oliveux, J.L. Bailleul, A. Gillet, O. Mantaux, G.A. Leeke, Recovery and reuse of
discontinuous carbon fibres by solvolysis: Realignment and properties of remanufactured
materials, Compos. Sci. Technol. 139 (2017) 99–108.
doi:10.1016/j.compscitech.2016.11.001.
85. T. Liu, X. Guo, W. Liu, C. Hao, L. Wang, W.C. Hiscox, C. Liu, C. Jin, J. Xin, J. Zhang,
Selective cleavage of ester linkages of anhydride-cured epoxy using a benign method and
reuse of the decomposed polymer in new epoxy preparation, Green Chem. 19 (2017) 4364–
4372. doi:10.1039/c7gc01737e.
86. L. Matějka, J. Lövy, S. Pokorný, K. Bouchal, K. Dušek, Curing epoxy resins with anhydrides.
Model reactions and reaction mechanism, J. Polym. Sci. Polym. Chem. Ed. 21 (1983) 2873–
2885. doi:10.1002/pol.1983.170211003.
87. F. Kolář, J. Svítilová, Kinetics and mechanism of curing epoxy/anhydride systems, Acta
Geodyn. Geomater. 4 (2007) 85–92.
150
88. ASTM D6272-17, Standard test methods for flexural properties of unreinforced and
reinforced plastics and electrical insulating materials by Four-Point Bending, (2017).
89. W.-F.A. Su, K.C. Chen, S.Y. Tseng, Effects of chemical structure changes on thermal,
mechanical, and crystalline properties of rigid rod epoxy resins, J. Appl. Polym. Sci. 78
(2000) 446–451. doi:10.1002/1097-4628(20001010)78:2<446::AID-APP250>3.0.CO;2-W.
90. S.T. Cholake, M.R. Mada, R.K. Singh Raman, Y. Bai, X.L. Zhao, S. Rizkalla, S.
Bandyopadhyay, Quantitative analysis of curing mechanisms of epoxy resin by mid- and
near-fourier transform infra red spectroscopy, Def. Sci. J. 64 (2014) 314–321.
doi:10.14429/dsj.64.7326.
91. R. Leblanc, About metal recycling: an introduction to scrap metal recycling.
https://www.thebalancesmb.com/about-metal-recycling-2877921 [last accessed Nov 2018]
92. T.E. Graedel, J. Allwood, J.P. Birat, M. Buchert, C. Hagelüken, B.K. Reck, S.F. Sibley, G.
Sonnemann, What do we know about metal recycling rates?, J. Ind. Ecol. 15 (2011) 355–366.
doi:10.1111/j.1530-9290.2011.00342.x.
93. Campbell Jr, F.C., 2011. Manufacturing technology for aerospace structural materials.
Elsevier.
94. S.F.M. de Almeida, Z.D.S.N. Neto, Effect of void content on the strength of composite
laminates, Compos. Struct. 28 (1994) 139–148. doi:10.1016/0263-8223(94)90044-2.
95. K.J. Bowles, S. Frimpong, Void effects on the interlaminar shear strength of unidirectional
graphite-fiber-reinforced composites, (1992) 1487–1509. doi:10.1177/002199839202601006.
96. T. Centea, L.K. Grunenfelder, S.R. Nutt, A review of out-of-autoclave prepregs – Material
properties, process phenomena, and manufacturing considerations, Compos. Part A Appl. Sci.
Manuf. 70 (2015) 132–154. doi:10.1016/j.compositesa.2014.09.029.
151
97. J.P. Anderson, M.C. Altan, Formation of Voids in Composite Laminates : Coupled Effect of
Moisture Content and Processing Pressure, Polym. Compos. 36 (2015) 376–384.
doi:10.1002/pc.22952.
98. T. Centea, P. Hubert, Measuring the impregnation of an out-of-autoclave prepreg by micro-
CT, Compos. Sci. Technol. 71 (2011) 593–599. doi:10.1016/j.compscitech.2010.12.009.
99. L.K. Grunenfelder, S.R. Nutt, Void formation in composite prepregs – Effect of dissolved
moisture, Compos. Sci. Technol. 70 (2010) 2304–2309.
doi:10.1016/j.compscitech.2010.09.009.
100. L.K. Grunenfelder, T. Centea, P. Hubert, S.R. Nutt, Effect of room-temperature out-time on
tow impregnation in an out-of-autoclave prepreg, Compos. Part A Appl. Sci. Manuf. 45
(2013) 119–126. doi:10.1016/j.compositesa.2012.10.001.
101. T. Centea, P. Hubert, Out-of-autoclave prepreg consolidation under deficient pressure
conditions, J. Compos. Mater. 48 (2014) 2033–2045. doi:10.1177/0021998313494101.
102. M. Brillant, P. Hubert, Out-of-autoclave manufacturing of complex shaped composite
laminates, eThesis, McGill University, Department of Mechanical Engineering, 2011.
103. P. Hubert, A. Poursartip, Aspects of the Compaction of Composite Angle Laminates :, J.
Compos. Mater. 35 (1999) 2–26. doi:10.1106/X8D7-PR9V-U6F2-0JEK.
104. C. Albert, G. Fernlund, Spring-in and warpage of angled composite laminates, Compos. Sci.
Technol. 62 (2002) 1895–1912. doi:10.1016/S0266-3538(02)00105-7.
105. G. Fernlund, N. Rahman, R. Courdji, M. Bresslauer, A. Poursartip, K. Willden, K. Nelson,
Experimental and numerical study of the effect of cure cycle , tool surface , geometry , and
lay-up on the dimensional ® delity of autoclave-processed composite parts, Compos. Part A
Appl. Sci. Manuf. 33 (2002) 341–351. doi:10.1016/S1359-835X(01)00123-3.
152
106. P. Hubert, A. Poursartip, A Review of Flow and Compaction Modelling Relevant to
Thermoset Matrix Laminate Processing.pdf, J. Reinf. Plast. Compos. 17 (1998) 286–318.
doi:10.1177/073168449801700402.
107. P. Hubert, R. Vaziri, A. Poursartip, A two-dimensional flow model for the process simulation
of complex shape composite laminates, 26 (1999) 1–26. doi:https://doi-
org.libproxy2.usc.edu/10.1002/(SICI)1097-0207(19990110)44:1<1::AID-
NME481>3.0.CO;2-K.
108. Y. Li, M. Li, Y. Gu, Z. Zhang, Numerical and Experimental Study on the Effect of Lay-Up
Type and Structural Elements on Thickness Uniformity of L-Shaped Laminates, Appl.
Compos. Mater. 16 (2009) 101–115. doi:10.1007/s10443-009-9080-z.
109. Y. Ma, T. Centea, G. Nilakantan, S.R. Nutt, Vacuum Bag Only Processing of Complex
Shapes: Effect of Corner Angle, Material Properties and Process Conditions. Proc of 29th
Tech. Conf. Am. Soc. Compos. US-Japan Conf. Compos. Mater. D-30 Meeting, La Jolla,
CA, Sept. 8-10, (2014).
110. J.S. Lightfoot, M.R. Wisnom, K. Potter, Composit es : Part A A new mechanism for the
formation of ply wrinkles due to shear between plies, Compos. Part A. 49 (2013) 139–147.
doi:10.1016/j.compositesa.2013.03.002.
111. M. Brillant, P. Hubert, Out-of-Autoclave Processing of Complex Shape Laminates. Proc of
the 2010 Society for the Advancement of Material and Process Engineering (SAMPE)
Technical Conference, Seattle, WA, May 17-20 (2010).
112. A. Levy, J. Stadlin, H. Pascal, Corner Consolidation in Vacuum Bag Only Processing of Out-
of-Autoclave Composite Prepregs Laminates. Proc of the 2014 Society for the Advancement
of Material and Process Engineering (SAMPE) Technical Conference, Seattle, WA, June 2-5,
(2014).
153
113. L.K. Grunenfelder, C. Fisher, C. Cabble, S. Thomas, S. R. Nutt, Defect Control in Out-of-
Autoclave Manufacturing of Structural Elements. Proc of the 2012 Society for the
Advancement of Material and Process Engineering (SAMPE) Technical Conference,
Charleston, SC, Oct 22-25 (2012).
114. S.M. Hughes, P. Hubert, Out-of-Autoclave Prepreg Processing: Effect of Integrated
Geometric Effects on Part Quality. Proc of the 2013 Society for the Advancement of Material
and Process Engineering (SAMPE) Technical Conference, Wichita, Oct 21-24 (2013).
115. J. Sloanf, Composites recycling becomes a necessity.
https://www.compositesworld.com/articles/composites-recycling-becomes-a-necessity [last
accessed Dec 2018]
116. G. Nilakantan, S. Nutt, Reuse and upcycling of aerospace prepreg scrap and waste, Reinf.
Plast. 59 (2015) 44–51. doi:10.1016/j.repl.2014.12.070.
117. F.A. López, O. Rodríguez, F. José, I. García-díaz, T.A. Centeno, J.L. García-fierro, C.
González, Journal of Analytical and Applied Pyrolysis Recovery of carbon fibres by the
thermolysis and gasification of waste prepreg, J. Anal. Appl. Pyrolysis. 104 (2013) 675–683.
doi:10.1016/j.jaap.2013.04.012.
118. L. Giorgini, T. Benelli, L. Mazzocchetti, C. Leonardi, G. Zattini, G. Minak, E. Dolcini, M.
Cavazzoni, I. Montanari, C. Tosi, Recovery of carbon fibers from cured and uncured carbon
fiber reinforced composites wastes and their use as feedstock for a new composite
production, Polym. Compos. 36 (2015) 1084–1095. doi:10.1002/pc.23440.
119. X.H. Zhang, F. Liu, S. Chen, G.R. Qi, Novel flame retardant thermosets from nitrogen-
containing and phosphorus-containing epoxy resins cured with dicyandiamide, J. Appl.
Polym. Sci. 106 (2007) 2391–2397. doi:10.1002/app.26698.
154
120. R. Duarah, N. Karak, B. Voit, T. Heinze, N. Karak, A starch based sustainable tough
hyperbranched epoxy thermoset, RSC Adv. 5 (2015) 64456–64465.
doi:10.1039/C5RA09955B.
121. K. Hatada, T. Kitayama, NMR spectroscopy of polymers, Springer-Verlag, 2005.
122. A.J. Brandolini, D.D. Hills, NMR spectra of polymers and polymer additives, Marcel Dekker,
2000.
123. D.D.A. Azevedo, S. Lacorte, P. Viana, D. Barceló, Occurrence of Nonylphenol and
Bisphenol-A in Surface Waters from Portugal, J. Braz. Chem. Soc. 12 (2001) 532–537.
doi:10.1590/S0103-50532001000400015.
Abstract (if available)
Abstract
Because of the increasing demand for lightweight structures in aerospace, automotive, and wind energy industries, the global market size for carbon fiber polymer composites is anticipated to reach $35 billion by 2020. The increasing use of composites poses significant environmental problems because most end-of-life composite waste is not recovered and/or reused due to a lack of viable recycling technologies. This issue is especially true for thermoset composites that undergo irreversible cure reactions. Thermoset composites have largely resisted attempts to recycle because the crosslinked polymer matrices cannot be easily separated from the fiber reinforcements. At the present juncture, composite recyclability is essential to the sustainability of the growing composite industry. Without a robust and effective method to recycle composites and complete the material life-cycle, these materials will not be able to compete with steel and aluminum in mass market applications, for which recycling rates are already high. ❧ Herein, the recycling project aim to develop an effective chemical digestion method for recycling of highly crosslinked amine/epoxy matrices using moderate conditions (atmospheric pressure and moderate temperature) and safe chemicals. These features are critical to recover near-virgin quality fibers and useful chemical components from composites with different crosslink densities, but have not been reported to date. ❧ Recycling is practical only if the recycling process itself can be scaled up to industrial scale, in which fast epoxy dissolution rate is one crucial criterion. Therefore, this thesis first focuses on neat polymer dissolution of amine-cured epoxies. Two chemical treatment methods—depolymerization and acid digestion (both at atmospheric pressure)—were employed to dissolve amine-cured epoxy formulations. Both depolymerization and acid digestion were shown to be effective dissolution processes for all amine/epoxy samples that encompassed variations in amine/epoxy stoichiometric ratio (A/E ratio), epoxy monomer functionality, and amine curing agent type. The relationship between epoxy properties and dissolution rate was determined, and the key parameters affecting thermoset matrix dissolution were identified. The dissolution rate was controlled by both the chemical reaction and diffusion rates. The components of the chemical solutions after epoxy dissolution were analyzed and identified, and protocols to quantitatively track the products after dissolution were developed. The two major cleavable sites during epoxy dissolution were the C-N and C-O bonds, and the aromatic structures of the epoxies were preserved. ❧ Based on the knowledge learned from neat epoxy dissolution, this thesis continues to investigate the viability of recycling actual composites. Depolymerization and acid digestion were applied to amine/epoxy composites, including composites produced from lab-made and aerospace prepregs and commercial composite waste. Findings indicated that acid digestion was more effective for highly crosslinked amine/epoxy composites than depolymerization. Furthermore, digestion occurred via reaction steps of oxygen atom transfer to the aniline groups and then bond cleavage, resulting in recovery of near-virgin quality fibers at faster dissolution rates and lower temperatures. The relationship among epoxy functionality, fiber bed architecture, fiber reinforcement, laminate thickness, and matrix dissolution rate were investigated, and key parameters affecting the dissolution rate were identified. Two strategies to enhance the diffusion rate—pre-treatment and mechanical shredding—ere evaluated, and both were effective. Polymer matrices in pre-treated and shredded composites were homogeneously decomposed in 1 h. Prospective catalytic conditions were screened to accelerate the chemical reaction rate for acid digestion. The most effective catalyst was ScCl₃, which reduced ∼ 30% of the time required by acid digestion with no catalyst. ❧ Current chemical recycling of thermoset composites has been focused largely on recovering high-value carbon fibers with property retention. However, recovery and reuse of decomposed polymer matrix residues is rarely considered, despite the fact that doing so constitutes an essential component of a sustainable approach to the problem. Therefore, this thesis also investigates the viability of recovery and reuse of the decomposed amine/epoxy residue after acid digestion of the matrix, effectively closing the recycling loop. Findings indicated that polymer matrix residues recovered from acid digestion solutions via neutralization and evaporation contained molecular components of the epoxies in which aromatic regions were preserved. The recovered matrix residues were blended into virgin resin formulations and two approaches were evaluate for potential reuse. Approach I utilized the matrix residue as an accelerator for a virgin anhydride/epoxy formulation that contained no accelerator and thus could not be self-catalyzed. This study discovered that adding matrix residue produced catalytic effects on the curing reaction. In general, anhydride/epoxy samples blended with recovered matrix residues and cured were homogenous and exhibited thermal and mechanical properties comparable to specimens cured with a commercial accelerator. Approach II deployed the matrix residue as a filler for both virgin anhydride- (with a commercial accelerator) and amine-based epoxies to produce blended formulations. In such cases, blended formulations yielded acceptable retention of thermal properties, provided the fraction of matrix residue added was < 20 wt%. ❧ Besides developing effective chemical recycling method for thermoset composites, the other focus of the thesis is sustainable manufacturing. Complex structures manufactured using low-pressure vacuum bag-only (VBO) prepreg processing are more susceptible to defects than flat laminates due to complex compaction conditions present at sharp corners. Consequently, effective defect mitigation strategies are required to produce structural parts. This study investigated the relationships between laminate properties, processing conditions, mold designs and part quality in order to develop science-based guidelines for the manufacture of complex parts. Generic laminates consisting of a central corner and two flanges were fabricated in a multi-part study that considered variation in corner angle and local curvature radius, the applied pressure during layup and cure, and the prepreg material and laminate thickness. The manufactured parts were analyzed in terms of microstructural fiber bed and resin distribution, thickness variation, and void content. The results indicated that defects observed in corner laminates were influenced by both mold design and processing conditions, and that optimal combinations of these factors can mitigate defects and improve quality.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Characterization, process analysis, and recycling of a benzoxazine-epoxy resin for structural composites
PDF
Sustainable manufacturing of out-of-autoclave (OoA) prepreg: processing challenges
PDF
Chemical depolymerization of amine-epoxy cured carbon fiber reinforced polymer composites and their re-use
PDF
Fabrication and analysis of prepregs with discontinuous resin patterning for robust out-of-autoclave manufacturing
PDF
In situ process monitoring for modeling and defect mitigation in composites processing
PDF
Defect control in vacuum bag only processing of composite prepregs
PDF
Polymer flow for manufacturing fiber reinforced polymer composites
PDF
Processing and properties of phenylethynyl-terminated PMDA-type asymmetric polyimide and composites
PDF
Efficient manufacturing and repair of out-of-autoclave prepreg composites
PDF
Developing efficient methods for the manufacture and analysis of composites structures
PDF
Material and process development and optimization for efficient manufacturing of polymer composites
PDF
Processing, thermal-oxidative stability and thermal cyclic fatigue of phenylethynyl-terminated polyimides
PDF
Fabrication of polymer films on liquid substrates via initiated chemical vapor deposition: controlling morphology and composition
PDF
Environmental effects on the hybrid glass fiber/carbon fiber composites
PDF
Vapor phase deposition of dense and porous polymer coatings and membranes for increased sustainability and practical applications
PDF
In situ process analysis for defect control during composites manufacturing
PDF
Fatigue and fracture of pultruded composite rods developed for overhead conductor cables
PDF
Engineering chimeric antigen receptor-directed immune cells for enhanced antitumor efficacy in solid tumors
PDF
Sound transmission through acoustic metamaterials and prepreg processing science
PDF
Comprehensive uncertainty quantification in composites manufacturing processes
Asset Metadata
Creator
Ma, Yijia
(author)
Core Title
Chemical recycling of amine/epoxy composites at atmospheric pressure
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
02/20/2019
Defense Date
01/11/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
carbon fiber,chemical treatment,composites,epoxy,OAI-PMH Harvest,Recycling,reuse
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Nutt, Steven (
committee chair
), Centea, Timotei (
committee member
), Malmstadt, Noah (
committee member
), Williams, Travis (
committee member
)
Creator Email
yijia.ma26@gmail.com,yijia@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-125492
Unique identifier
UC11675301
Identifier
etd-MaYijia-7100.pdf (filename),usctheses-c89-125492 (legacy record id)
Legacy Identifier
etd-MaYijia-7100.pdf
Dmrecord
125492
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Ma, Yijia
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
carbon fiber
chemical treatment
composites
epoxy
reuse