熊本大学学術リポジトリ

Kumamoto University Repository System

Title Recycling of Carbon Fiber and Epoxy Resin from

Carbon Fiber Reinforced Plastics

Author(s) 柴田, 勝司

Citation

Issue date 2014-03-25

Type Thesis or Dissertation

URL http://hdl.handle.net/2298/31435

Right

1

Recycling of Carbon Fiber and

Epoxy Resin from Carbon Fiber

Reinforced Plastics

March 2014

Katsuji Shibata

Graduate School of Science and Technology

KUMAMOTO UNIVERSITY

2

Recycling of Carbon Fiber and Epoxy Resin from Carbon Fiber Reinforced Plastics

(炭素繊維複合材料から回収した炭素繊維及びエポキシ樹脂のリサイクル技術の開発)

Abstract（within 1600 words）

1. Introduction

Since fiber reinforced plastics (FRP) have high strength and durability in lightness, these have

been widely used for sporting goods, bathtubs, automobiles, railway vehicles, boats, etc.

However, because thermosetting resins (thermosets) used for FRP neither melt nor dissolve,

recycling of FRP is very difficult. We have developed the technology to recover materials from

FRP by dissolving the resins. FRP is classified by the types of the reinforced fibers. GFRP using

glass fiber has the highest market share. Next, there are a lot of CFRP that uses carbon fiber.

The common chemical recycling technologies of thermosets are pyrolysis and solvolysis. They

are often carried out with supercritical fluids. Pyrolyses damage fiber and filler because of high

temperature. Supercritical fluid methods require high-pressure vessels that are usually expensive.

Solvolyses are carried out under mild conditions. But they need long treatment time without

appropriate catalysts and solvents, or require crushers and grinders.

2. Previous Works

Printed wiring board (PWB) wastes are composite of metals, glass fiber, paper, fillers, devices

and plastics. In studying reactions of epoxy resins for glass-epoxy laminates, we found

depolymerization of brominated epoxy resins. We tried to apply this reaction to recycling of

PWB, then found the effective catalysts for this depolymerization. Finally, we applied these

solutions to PWB wastes made of brominated epoxy-glass cloth laminate, obtaining separated

recyclates such as electronic devices, glass cloth, copper foil depolymerized epoxy resin.

In the next stage, we applied the method of dissolving resins under ordinary pressure to GFRP.

In this method, tripotassium phosphate (K3PO4) is used as a catalyst and benzyl alcohol (BZA)

as a solvent at 190˚C under ordinary pressure. These chemicals are both food additives.

Depolymerized resins, fibers, fillers and metals are separated and recovered by this method. The

processing time is from 5 to 20 hours. This method needs no preprocessing such as grinding or

crushing. We treated helmets made of GFRP treated by the method then obtained GF, filler and

depolymerized UP. GF recovered from GFRP is 20 mm or more in length. Nonwoven fabric of

the recovered GF can be made by the dry process or the wet process. In the dry process, a

carding machine which is usually used to manufacture futon cotton is useful. In the wet process

papermaking machine can be used. The tensile strength of recycled GFRP that uses nonwoven

3

fabric was lower than new one by 30 %. But, it is applicable to some uses which do not require

high strength.

AFRP is used for aircraft materials and concrete reinforcement. AF is difficult to separate from

AFRP molded with EP. Since AF decomposes at about 500 ˚C, EP cannot be removed from

AFRP by pyrolysis or supercritical fluids. The method of dissolving resin under ordinary

pressure is the only method to recover AF from AFRP. Aramid rods made of AFRP are used for

concrete telephone poles. They were treated for 25 h at 190 ˚C, then AF was recovered. With

recovered AF, nonwoven fabric is made by the same dry process as GF and CF.

3. Recovering CF and EP from CFRP

Epoxy resin cured with acid anhydride is mainly in CFRP matrix. This is applied to golf club,

tennis racket and aerospace parts, but is difficult to recycle. We estimated that cured epoxy resin

can be depolymerized by transesterification with mono alcohols, and then be dissolved in

solvent. Various CFRP products could be dissolved by depolymerization under ordinary

pressure. For example, tennis rackets and badminton rackets were dissolved in 8 hours.

The nonwoven fabrics were produced by the dry or wet process similar to GF from the

recovered cotton-like CF. We produced the recovered CF nonwoven fabric using a carding

machine.

4. Recycling of Recovered CF

The investigated materials are non-woven fabrics of CF (rCF) recovered by our method from

depolymerized CFRP-moldings and tennis rackets. We compared the properties of the

non-woven fabric of our recovered CF by depolymerization with those of the commercially

available TORAYCA Mat and the recovered CF by pyrolysis process. The resin used was

Bisphenol-A type EP cured with anhydride.

We evaluated the tensile properties and the surface of CF monofilaments. The tensile

properties of CF monofilaments were measured in accordance with JIS R 7606. The surfaces of

the rCFs were not damaged and clean as TORAYCA.

We produced non-woven fabric of rCFs by using a carding machine. The recycled CFRP

(rCFRP), which is composed of non-woven fabric and Bisphenol-A type EP cured with

anhydride, was manufactured by compression molding. The volume content of fiber (Vf) was

between 0 % and 40 %.

The thicknesses of CFRP-m, Racket and TORAYCA increase with increasing Vf when Vf is

larger than 30 %. We speculated that when the density of non-woven fabric is low, the

4

compression molding does not proceed properly due to the bulkiness of the fabric. On the

contrary, in the case of Takayasu, which has high density, the thickness of CFRP was almost

constant even at the high Vf of 40 %. Looking at the cross-section, some voids were observed in

CFRP-m and Racket of 40 % Vf. We considered the cause of the void was that EP couldn’t be

impregnated properly due to bulkiness of the fabric which hinders the compression of the resin

and fabric. In the CFRP of Takayasu and TORAYCA, voids were not observed. So, it may be

necessary to improve the density of the fabric to enhance the moldability of rCFRP. We

dissolved rCFRP and evaluated the CF length varying Vf from 10 % to 40 % after compression

molding. The length of CF of 40 % Vf was shortened by about 1 mm～2 mm comparing with

the length of about 30 mm at Vf of 10 % in CFRP-m, Racket and TORAYCA. We thought that

the breaking of fiber may be caused by the increased contact point of fibers of high Vf. On the

other hand, the ratio of cut fiber was low in Takayasu.

We evaluated mechanical properties of rCFRP using non-woven rCF fabrics following JIS K

7073 (tensile properties), JIS K 7074 (flexural properties) and JIS K 7077 (Charpy impact

strength). The mechanical properties increased with increasing Vf, but after reaching a maxima,

decreased with further increase in Vf. We thought that the cause of the non-linear relation was

the imperfect molding and increased ratio of shortened fibers.

5. Recycling of Recovered EP

Acid anhydride (Ah) cured EP (EP/Ah) usually used as the matrix resin of CFRP has ester

bond by the reaction between epoxy group and anhydride. So, we investigated the

depolymerization of EP/Ah under ordinary pressure and found that EP/Ah easily dissolves into

BZA in standard conditions, finally recovering the carbon fiber.

We used model compounds of EP/Ah for analysis. The model epoxy was phenylglycidylether

(PGE) and the model Ahs were succinic anhydride (ScAh), cis-1, 2-cyclohexanedicarboxylic

anhydride (ChAh), cis-4-cyclohexene-1, 2-dicarboxylic anhydride (CheAh). The model

compounds consisting of PGE and Ahs were introduced in a test tube. The reactive mixture was

heated at 80 ˚C for 1.0 h followed by heating at 150 ˚C for 1.0 h. In order to analyze the

depolymerization, the model compounds, BZA and K3PO4 were introduced in a test tube, then

were heated at 180 ˚C for 10.0 h. For identification of products, we used 1H NMR and 13C

NMR. Model compounds were dissolved in d6-dimethylsulfoxide (d6-DMSO). The model

compounds were synthesized stoichiometrically by the reaction between PGE and Ahs with

catalyst. While, the depolymerization of PGE/Ah model compounds, we found that

dibenzylester and bisdiol are formed by the depolymerization of the thermosets under ordinary

5

pressure. The depolymerization mechanism of PGE/Ah was the transesterification with BZA, so

we concluded that the depolymerization were mainly transesterification reaction accelerated

with K3PO4.

We have developed the depolymerization of thermosets by using subcritical fluids. Insoluble or

slightly soluble thermosets like Am cured EP was possible to be depolymerized in a short time

by this method.

In order to analyze the mechanism of EP/Am depolymerization, bisphenol A diglycidylether

(DGBPA) was used as the model matrix. Isophrondiamine (IPDA) and

2,4,6-tris(dimetylaminomethyl) phenol (DMAmP) were used as the curing agent and catalyst.

The selected model compounds of Ams were dicyclohexylamine (DChAm),

N,N-dicyclohexylamine (DChMeAm), dibenzylamine (DBzAm) and tribenzylamine (TBzAm).

DGBPA, IPDA and DMAmP of the ratio of 100:25:2 were mixed, then were heated at 80 ˚C for

0.5 h. The additional heating was 150 ˚C for 1.0 h. The obtained cured EP of about 5 mm thick

was cut into 10 mm × 15 mm for further testing. A piece of a model matrix, BZA and K3PO4

were introduced into tube bomb reactor, then heated at 250 ˚C to 325 ˚C for 1.0 h to 4.0 h. To

depolymerize Am compounds, BZA and K3PO4 were introduced in the reactor, then heated at

280 ˚C for 6.0 h.

The test pieces dissolved perfectly after 4.0 h at 325 ˚C. It was difficult to analyze the

depolymerization products of EP which has cross-linked three-dimensional networks. To

identify the structure of products, Am compounds and depolymerization products were

measured with NMR. The cleavage of the C-N bond didn’t occur in alicyclic Am by the

depolymerization reaction. On the other hand, we found that there was the possibility of

cleaving the C-N bond in aromatic Am compounds with the depolymerization reaction by

subcritical fluids. Because IPDA is alicyclic Am，the cleavage of the ether linkage (C-O-C) or

transetherification may occur.

6

Table of Contents

Abstract

1. Introduction --------------------------------------------------------------------------------------- 7

2. Previous Works ----------------------------------------------------------------------------------- 9

2.1 Recycling of Printed Wiring Board (PWB) -------------------------------------------------- 9

2.2 Recycling of Fiber Reinforced Plastics (FRP) ---------------------------------------------- 16

3. Recovering CF and EP from CFRP ------------------------------------------------------------- 21

4. Recycling of Recovered CF ---------------------------------------------------------------------22

4.1 Recovered CF Nonwoven Fabric --------------------------------------------------------------22

4.2 Recycled CFRP ----------------------------------------------------------------------------------23

5. Recycling of Recovered EP ----------------------------------------------------------------------33

5.1 Anhydride Cured EP ----------------------------------------------------------------------------33

5.2 Amine Cured EP ---------------------------------------------------------------------------------49

6. Conclusion ------------------------------------------------------------------------------------------61

References

7

1. Introduction

1.1 Background

Since fiber reinforced plastics (FRP) have high strength and durability in lightness, these have

been widely used for sporting goods, bathtubs, automobiles, railway vehicles, boats, etc.

However, because thermosetting resins (thermosets) used for FRP neither melt nor dissolve,

recycling of FRP is very difficult. We have developed the technology to recover materials

materials from FRP by dissolving the resins. FRP is classified by the types of the reinforced

fibers. GFRP using glass fiber has the highest market share. Next, there are a lot of CFRP that

uses carbon fiber. AFRP using aramid fiber is used in an architectural field. These FRPs can be

recycled by the method of dissolving thermosets under ordinary pressure.

1.2 Comparison with Conventional Technologies

The common chemical recycling technologies of thermosets are mainly pyrolysis and solvolysis (Table 1) .

Japan Carbon Fiber Manufacturers Association (JCMA) started working on CFRP recycling in

2006 1). JCMA made a CFRP pyrolysis plant for recovering milled CF at Omuta City, Fukuoka

Prefecture, that are taken over by Toray Industries, Inc., Toho Tenax Co., Ltd. and Mitsubishi

Rayon Co., Ltd..from 2013.

Okajima et al. 2), 3) developed CFRP recycling to decompose EP with supercritical methanol or

acetone at 250 ˚C for 2 h in a 5 L pressurized vessel. The decomposed products are obtained by

the ester bond cleaved selectively.

Gas phase Vegitable oil Liquid phase Glycolysis Our method

Temperature 250～900℃ 350℃ 180～400℃ 200～440℃ 150～250℃ 〈200℃

Pressure closed,ordinary ordinary 2～22MPa ordinary～

2MPa ordinary～20MPa ordinary

Solvent no vegetable oil water,alchol,phenol

hydrogen donorsolvent glycol alcohol

Catalyst no no no salt acid or alkali salt

Grinding size ＜10mm ＜5mm ＜1mm ＜5mm ＜1mm -

Recyclate gas,oil oil monomer monomer,oil Oligomer Oligomer

MethodPyrolysis Supercritical

FluidSolvolysis

Table 1 Comparison of chemical recycling methods of thermosets

8

M. Goto et al. 4), 5) at Kumamoto University investigated recycling of CFRP with subcritical

benzyl alcohol at 300 ˚C-400 ˚C。This method needed lower pressure than the methods with

other supercritical or subcritical fluids.

S. J. Pickering et al. 6), 7) at University of Nottingham have developed the method of recovering

CF from CFRP with supercritical propanol. This method needs lower temperature than one with

supercritical water.

M. Kubouchi et al. 8) at Tokyo Institute of Technology have investigated dissolving amine

cured EP with nitric acid at 80 ˚C for 100 h. The cured EP that consisted of 25 % recovered EP

and 75 % fresh EP with anhydride as curing agent had higher glass transition temperature and

higher bending strength than that of all fresh EP.

Mizuguchi et al. 9) at Shinshu University have developed the FRP recycling technology on the

basis of the thermal activation of semiconductors (TASC). TASC enables to decompose

polymer matrices of FRP into H2O and CO2 in 10-20 min at about 400-500 ˚C, yielding CF or

GF without damage.

Pyrolyses damage fiber and filler because of high temperature and oxygen. Supercritical fluid

methods require high-pressure vessels that are usually expensive. Solvolyses are carried out

under mild conditions. But they need long treatment time without appropriate catalysts and

solvents, or require crushers and grinders. We regard cost efficiency as most important to put a

recycling technology to practical use; consequently we have been searching for a simple

recycling process.

9

2. Previous Works

2.1 Recycling of Printed Wiring Board (PWB) 9)-14)

Printed wiring board (PWB) wastes are composite of metals, glass fiber, paper, fillers, devices

and plastics. These components consist of various kinds of raw materials. Moreover, plastics

used in PWB often contain halogens for flame retardancy. As they are difficult to separate into

recyclable materials and to incinerate, they are mainly landfilled. Laminates used for PWB are

divided broadly into paper-phenol laminates and glass-epoxy laminates. The former is used

more than the latter, but the latter has been increasing recently for IT instruments. In studying

reactions of epoxy resins for glass-epoxy laminates, we found a depolymerization of brominated

epoxy resins or brominated polyhydroxyethers made from bifunctional epoxy resins and

bifunctional phenols. We tried to apply this reaction to glass-epoxy laminates recycling,

especially for the separation of materials used in PWB. First, we studied the mechanism of these

polymers' degradation, then we clarified the mechanism of depolymerization into raw materials,

epoxy resins and phenols. Then we found the effective catalysts for this depolymerization.

Finally, we applied this technology to PWB wastes.

2.1.1 Experimental

(1) Materials

Chemicals used as solvents and catalysts are shown in Table 2.

Chemicals abbr.　Phenylglycidylether PGE　Tribromophenol TBPh　N,N-Dimethylformamide-d7 DMF-d7　Cyclohexanone CHON　Diethylene glycol monomethyl ether DGMM　N-methyl-2-pyrrolidone NMP　Potassium hydroxide KOH　Potassium iodide KI　Potassium bromide KBr　Potassium chloride KCl　Sodium hydrogen carbonate NaHCO3　Sodium iodide NaI　Sodium bromide NaBr　Sodium chloride NaCl　Lithium hydroxide LiOH　Lithium iodide LiI　Lithium bromide LiBr　Lithium chloride LiCl

Table 2 Chemicals used as solvents and catalysts

10

All the chemicals used were reagent grade supplied by Kanto Chemical Co., Inc. Japan.

Brominated epoxy polymer model compound 1-(2,4,6-tribromo-phenoxy)- 3-phenoxy-

2-propanol (TBPP) was prepared in our laboratory by reaction of phenylglycidylether with

2,4,6-tribromophenol.

(2) Laminates Preparation

Laminates No. 1 and No. 2 shown in Table 3 were commercial products by Hitachi Chemical

Co., Ltd. Japan. Laminate No. 3 was prepared from brominated epoxy resin, brominated

polyphenol and glass cloth with molding press.

(3) NMR Analysis

13C-NMR spectra were measured at room temperature on a Brucker 400 MHz NMR

spectrometer. Previously NMR spectra of TBPP solution in DMF-d7 were measured . The

samples were heated at 120 °C for 4.0 h and measured every hour.

(4) Molecular Weight Determination

Weight-average molecular weights of polymers were determined with a gel-permeation

chromatograph supplied by Tosoh Co. Ltd. Japan. Standard polystyrenes used for calibration

and columns of styrene gel were also from Tosoh Co. Ltd. Japan.

(5) Solubility Measurement

Laminates were in the form of 10 mm X 30 mm test pieces. Each had a thickness of about

0.2 mm. The test pieces were weighed and put into the solutions at 60-140 °C for 1.0-4.0 h.

The pieces were taken out from the solution, then dried and weighed again. Solubility of

laminates was calculated by the next equation.

2.1.2 Results and Discussion

(1) Model Compound Decomposition

The decomposition of 1-(2,4,6-tribromo-phenoxy)-3-phenoxy-2-propanol (TBPP) as a model

Composite No.1 No.2 No.3commercial commercial prepared

Curing Agents Polyphenol Dicy PolyphenolBr contents 15% 20% 28%FiberResin contentsCuring Temp.

E-glass　cloth47%

170℃／90min

Table 3 Laminates

11

compound of brominated epoxy polymer was observed with LiOH as a catalyst at 120 °C for

4.0 h. In Table 4, 13C-NMR spectrum data of reactant and reaction products are shown. Before

heating, there were only signals of TBPP. Upon heating, TBPP signals gradually weakened, and

new signals appeared.

The new signals are found to be ones of phenylglycidylether (PGE) and lithium

2,4,6-tribromophenolate (TBPh-OLi), with a small amount of by-product, 1-phenoxy-

2,3-propandiol (PPDO). TBPh-OLi will be able to change into TBPh by reacting with water.

PGE and TBPh are the raw materials of TBPP, so this reaction seems to be a reversible reaction.

TBPP decomposition is described schematically in Figure 1.

(2) Depolymerization of Brominated Epoxy Polymer

As TBPP can be cleaved at the oxygen atom linked with a brominated benzene ring, epoxy

polymers can also be cleaved at the ether linkage. Depolymerization of brominated epoxy

polymer at several concentrations of LiOH as a catalyst are given in Figure 2. The molecular

weight (Mw) of brominated epoxy polymer did not change without a catalyst, but increasing the

concentration of a catalyst, Mw of the polymer decreased more rapidly. Moreover, the

ppm 171.9 161.0 158.7 158.5 158.1 152.2 151.9 134.8 134.5 132.3 129.4 120.7 120.6 120.5 120.2 118.80h 51 27 95 100 39

1.0h 21 45 16 34 6 78 17 72 108 43 8 37 122.0h 24 44 19 29 68 12 83 117 33 33 104.0h 6 29 11 31 24 31 5 60 12 83 111 28 34 12 10

118.6 117.2 116.3 114.8 114.4 114.2 95.2 82.3 74.0 69.9 69.2 69.0 68.7 67.9 62.8 60.5 49.7 43.758 28 120 34 38 5287 38 7 52 120 22 31 5 29 7 22 42 24 4 40 4172 28 5 52 120 15 33 5 22 15 25 45 23 39 3559 16 5 82 120 14 26 7 23 9 11 26 29 29 7 4 39 31

O CH 2 CH CH 2 O

OH

Br

B r

B rTBPP

158.5

69.0

120.5117.2

74.0 67.9

152.2

118.6 134.8114.4129.4

O CH 2 CH CH 2O

PGE

120.7

43.7

134.5 114.2

49.768.7

158.1

O CH 2 CH CH 2

OH OHPPDO

120.2

69.9

129.4 120.7

69.262.8

158.7

Br

B r

B r

L i O

TBPh-OLi

161.0 95.2

132.3114.8

Table 4 13C-NMR spectra of TBPP decomposition (in DMF-d7, at 120 ˚C)

Figure 1 Scheme of TBPP decomposition

12

concentration of a catalyst decided the

Mw at the end points of the

depolymerized polymers. Mw of this

particular polymer decreased from

332,000 to 35,000 with 0.10 mol catalyst

at 120 °C for 6.0 hr.

(3) Solubility Measurement

PWB wastes from electronic

instruments were mainly made of

brominated epoxy polymers, glass cloths

and copper circuits. If epoxy polymers

dissolve into some solution, glass and

copper easily separate. In the

mechanism of brominated epoxy

depolymerization, bromine content (Br

content) influences the rate of the

depolymerization.

shows a relationship between bromine

content and solubility of laminates. As Br

content increased, solubility of laminates

also increased. 28% Br content laminate

has high solubility, compared to lower Br

content laminates. But 20% Br content

laminates are most common now, so that

type was used next.

The laminate solubility in the solutions of

alkali metal compounds as catalysts are

shown in Figure .

It shows KOH, NaHCO3 and LiCl are

better catalysts for the solubility of 20% Br

content laminates. Not only alkali metal

hydroxide, but alkali metal salts are also

0

10

20

30

40

0 2 4 6

Mw

(*10

00)

Heating time (h)

Catalyst : LiOH

0mol

0.02mol

1.00mol

0.20mol0.10mol

0.04mol

0

20

40

60

80

100

15 20 25 30Br Content (%)

Solu

bilit

y (%

)

140℃

60℃

100℃

　

Figure 2 Depolymerization of brominated epoxy polymers with various amount of LiOH as catalyst

Figure 3 Relationship between bromine content and solubility of brominated epoxy laminates

13

useful. Without a catalyst, the solubility value was as low as 1%.

The effect of combinations of catalysts and solvents are shown in Figure .

0 5 10 15 20 25

LiCl

LiBr

LiI

NaCl

NaBr

NaI

NaHCO3

KCl

KBr

KI

KOH

No Catalyst

Solubility (%)

Cata

lysts

0 5 10 15 20 25

LiI

LiCl

KOH

No Catalyst

Solubility (%)

Cat

alys

ts

NMPDGMM

CHON

Figure 4 Solubility's of brominated epoxy laminates with various alkali metal compounds as catalysts

Figure 5 Solubility of brominated epoxy laminates with combination of catalysts and solvents

14

NMP is a good solvent for KOH catalyst, and CHON is good for LiI.

Figure 6 shows the relationship between temperatures and the solubility of brominated epoxy

laminates. In these investigations KOH in NMP and LiI in CHON were used as catalysts in

solutions. KOH / NMP was able to dissolve the laminates more rapidly than LiI / CHON. The

solubility in both solutions are much higher over 140 °C than at lower temperatures.

Figure 7 Relationship between catalyst concentration and solubility of brominated epoxy

laminates shows the relationship between concentrations of catalysts and the laminate solubility.

This chart indicates the difference in the solubility between two solutions. The solubility for

KOH / NMP increased, along with an increase in concentration. Whereas, LiI / CHON had the

maximum solubility at the concentration of 3.0 mol /1000 g solvent.

2.1.3 Application to PWB Waste

Finally, we applied these solutions to PWB wastes made of brominated epoxy-glass cloth

laminate. Figure 8 shows separated recyclates from a PWB waste with electronic devices. The

recyclates are electronic devices, glass cloth, copper foil depolymerized epoxy polymer..

0

10

20

30

40

50

100 110 120 130 140 150 160 Temperature (deg C)

Solubility (%)

KOH/NMP

LiI/CHON

0

10

20

30

40

50

0.0 1.0 2.0 3.0 4.0 5.0 6.0 Concentration (mol/1000 g solvent)

Solubility LiI/CHON

KOH/NMP

Figure 4 Relationship between temperatureand solubility of brominated epoxy laminates

Figure 5 Relationship between catalyst concentration and solubility of brominated epoxy laminates

15

Separated circuits chiefly made of copper

Separated devices on glass cloth PWB waste with electronic devices

Separated glass cloth Separated electronic devices

Depolymerized brominated epoxy polymer solution in NMP with LiCl as a catalyst

Figure 6 Recyclates from PWB waste with electronic devices

16

2.2 Recycling of Fiber Reinforced Plastics (FRP) 15)-23)

2.2.1 The method of dissolving resins under ordinary pressure

In this method transesterification reaction is used to dissolve the resins of FRP at 190˚C under

ordinary pressure. Tripotassium phosphate (K3PO4) is used as a catalyst and benzyl alcohol

(BZA) as a solvent. These chemicals are both food additives. Depolymerized resins, fibers,

fillers and metals are separated and recovered by this method. The processing time is from 5.0 h

to 20.0 h, depending on the thickness, the kind of resins and manufacturing process of FRP. If

FRP is previously ground, the processing time can be shortened remarkably, but the recovered

GF is also shortened resulting in the low reinforcement for FRP. This method needs neither

preprocessing such as grinding nor an autoclave, and so it is more economical than other

recycling methos. In addition, as the method is dust free, there is no risk of pneumoconiosis and

dust explosion. Thus, this method is advantageous in the health and safety.

2.2.2 Glass fiber composite material (GFRP)

GFRP is mainly made from glass fiber (GF), unsaturated polyester resin (UP) and filler. It is

necessary to separate these materials in order to recycle GFRP. Each material can be separated

and recovered by the method of depolymerizing and dissolving UP under ordinary pressure.

Figure 9 shows helmets made of GFRP treated by the method with the passage of time.

17

GF recovered from GFRP is 20 mm or more in length. However, the recovered GF is a form of

cotton, which is not applicable to GFRP, while nonwoven fabric of the recovered GF made by

the dry process or the wet process can be applicable to FRP. In the dry process, a carding

machine which is usually used to manufacturing futon cotton is useful. In the wet process

papermaking machine can be used. The recovered GF nonwoven fabric made with dry process

is shown in Figure 10.

Before treatment After 0.5 h

After 2.5 h After 5.0 h

Figure 7 FRP helmets treated by the dissolving method

18

The tensile strength of recycled GFRP that uses the nonwoven fabric was lower than new one

by 30 %. But, it is applicable to some uses which do not require high strength.

2.2.3 Aramid fiber composite materials (AFRP) 14)

Aramid fiber (AF) has two kind of chemical structures which are para-aramid and meta-aramid.

Para-aramid which has high strength and high modulus is used for tire cords, conveyer belts,

protective clothing, etc. While meta-aramid which has flame retardancy and heat resistance is

used for filters, wire coverings, and flame retardant clothing, etc., it is hardly used as composite

materials. As for AFRP, para-aramid is chiefly used, and there are a lot of use of aircraft

materials and concrete reinforcement. AFRP molded with EP is difficult to separate AF. Since

AF decomposes at about 500 ˚ C, EP cannot be removed from AFRP by pyrolysis.

Supercritical fluids also degrade AF remarkably, and are not applicable for the recovering. The

method of dissolving resin under ordinary pressure is the only method that is announced to be

able to recover AF from AFRP.

Aramid rods made of AFRP are used for concrete telephone poles. They were treated for 25.0 h

at 190 ˚C, then AF was recovered. The recovered AF is shown in Figure 11.

Figure 8 Recovered GF nonwoven fabric

19

With recovered AF, nonwoven fabric is made by the same dry process as GF and CF. AF

nonwoven fabric currently made with a carding machine is shown in Figure 12.

Before treatment After 20.0 h

Before carding Recovered AF nonwoven fabric

Figure 9 An aramide rod made of AFRP treated by the dissolving method

Figure 10 Recovered AF nonwoven fabric with a carding machine

20

2.2.4 Pilot Plant for FRP Recycling

We have constructed the pilot plant for FRP recycling that can be used also for other

composites recycling. The FRP recycling process is shown in Figure 13. The pilot plant for FRP

recycling is shown in Figure 14, that was constructed with subsidy from the Ministry of

Economy, Trade and Industry of Japan.

FRP FRP Wastes

Solid Solid

Recovered Fiber

Recovered Filler

Recovered Resin

Washing Tank

Washing Tank

Treating Tank

Filtration Drying Distillation

Figure 11 FRP recycling process by the dissolving method

Figure 12 The pilot plant for FRP recycling by the dissolving method

21

3. Recovering CF and EP from CFRP

3.1 Method of Dissolving EP under Ordinary Pressure 24)-26)

Epoxy resin (EP) cured with acid anhydride is mainly used for a matrix of CFRP. This is

applied to golf club, tennis racket and aerospace parts, but is difficult to recycle. We estimate

that cured epoxy resin can be depolymerized by transesterification with mono alcohols, and then

be dissolved into a solvent.

3.2 Depolymerization of CFRP

Various CFRP products could be dissolved by the depolymerization under ordinary pressure.

For example, tennis rackets and badminton rackets made of CFRP were dissolved in 8.0 h

respectively (Figure 15)

CFRP badminton rackets

CFRP tennis rackets

Figure 13 CFRP rackets treated by the dissolving method

22

4. Recycling of Recovered CF 27)-42)

4.1 Recovered CF Nonwoven Fabric

The nonwoven fabrics using the recovered cotton-like CF were produced by the dry or wet

process as same as GF nonwoven fabrics of recovered GF (Figure 15). The recovered CF

nonwoven fabric currently produced with a carding machine is shown in Figure 16.

Before carding Recovered CF nonwoven fabric

Figure 14 Recovered CF nonwoven fabric with a carding machine

23

4.2 Recycled CFRP

4.2.1 Materials

The investigated materials are non-woven fabrics of CF (rCF) recovered by our method from

depolymerized CFRP-moldings and tennis rackets. The dissolving process of tennis racket is

shown in Figure 17. We compared the properties of the non-woven fabric of our recovered CF

by depolymerization with those of the commercially available TORAYCA Mat (TORAY

Industries Inc.) and the recovered CF by pyrolysis process (Takayasu Co., Ltd.). The EP used

was Bisphenol-A type EP cured with anhydride. Table 5 shows materials in this study.

Figure 15 Recovered CFs from a racket and a molding

24

4.2.2 Evaluation of CF monofilaments

We evaluated the tensile properties and the surface of CF monofilaments. The tensile

properties of CF monofilaments were measured in accordance with JIS R 7606. Table 6

characterizes tensile strength (T), tensile modulus (E) and fiber diameters (φ) of CF

monofilaments, comparing with the Takayasu and TORAYCA.

The rCFs had tensile properties close to the Takayasu and TORAYCA. The differences of

tensile strength and tensile modulus among CFs seemed to be due to the differences of the

original fibers properties.

Figure 18 shows the images scanning electron microscopic (SEM) images of rCFs, Takayasu

and TORAYCA. As obviously shown, the surfaces of the rCFs and Takayasu were not damaged

and clean as TORAYCA.

Item Product name Content Abbreviation ManufactuerNon-woven fabric Recycle CF Recovered from depolymerized CFRP-m this study

non-woven fabric (1) CFRP-moldingsRecycle CF Recovered from depolymerized Racket this studynon-woven fabric (2) Tennis racketRecycle CF Recovered from Pyrolysis Takayasu Takayasu Co.,Ltdnon-woven fabric (3) processTORAYCA Mat Commercial product TORAYCA TORAY Industies Inc.

Resin jER 828 Bisphenol-A type EP - Mitsubishi Chemical Corp.HN-2200 Anhydride type hardner - Hitachi Chemical Co., Ltd2E4MZ-CN Hardening accelerator - Shikoku Chemicals Corp.

Type of CF T (MPa) E (GPa) φ(μm)CFRP-m 4393.4 303.0 7.10Racket 3199.5 188.2 6.51Takayasu 3458.7 300.8 6.91TORAYCA 3198.4 152.3 6.94

Table 5 Materials

Table 6 Comparison of tensile properties of CFs

25

CFRP-m Racket Takayasu TORAYCA

4.2.3 Comparison with density of CF non-woven fabric

We produced non-woven fabric of rCFs by using a carding machine. Table 7 show the

densities and the bulkiness of non-woven fabrics (Vf = 40 %). As shown, the density of rCF

non-woven fabric is lower than those of Takayasu and TORAYCA.

0.00651 mm 0.00710 mm 0.00691 mm 0.00694 mm

Type of Thickness Fabric weight Density

non-woven fabrics (μm) (g/m2) (g/m3)CFRP-m 10 140 140

Racket 10 140 140

Takayasu 10 1000 1000

TORAYCA Mat 1 30 300

Table 7 Comparison of bulkiness of non-woven fabric

Figure 16 SEM images of rCFs (5000×)

26

4.2.4 Molding “Recycled CFRP”

The Recycled CFRP (rCFRP) composed of non-woven fabric (Figure 19) and Bisphenol-A

type EP cured with anhydride was manufactured by compression molding. The condition of

compression molding and molding process of rCFRP are shown in Table 8 and Figure 20. The

volume content of fiber (Vf) was between 0 % and 40 %.

Impregnation of EP resin Appearance of rCFRP

Item ConditionsPressure 12 MPaTemperature 150 ℃Time 2 h

EP resin

Table 8 Conditions of Compression Molding

Figure 17 Non-woven fabric of rCF

Figure 18 Molding process of rCFRP

27

4.2.5 Moldability of rCFRP

The dependence of CFRP thickness on the volume content of fiber (Vf) is shown in Figure 21.

As shown, the thicknesses of CFRP-m, Racket and TORAYCA increase with increasing Vf

when Vf is larger than 30 %.

We speculated that when the density of non-woven fabric is low, the bulkiness of the fabric

increases with the increase of Vf, and the compression molding does not proceed properly due

to the bulkiness of the fabric. Thus thickness of CFRP has increased with increasing Vf. On the

contrary, in the case of Takayasu whose density is high, the thickness of CFRP was almost

constant even at the high Vf of 40 %.

0

1

2

3

4

5

0 10 20 30 40

CFR

P Th

ickn

ess (

mm

)

Vf (vol%)

●：Takayasu

■：TORAYCA

◆：Racket

▲：CFRP-m

Figure 19 Relationship between Vf and thickness of rCFRPs

28

Table 9 Cross-section observation of rCFRPs

Item CFRP-m Racket Takayasu TORAYCA

Vf=10%

Vf=40%

As shown in the Table 9, some voids were observed in CFRP-m and Racket of 40 % Vf. We

considered the cause of the void was that EP couldn’t be impregnated properly due to bulkiness

of the fabric which hinders the compression of the resin and fabric. In the CFRP of Takayasu

and TORAYCA, voids didn’t occur. So, it may be necessary to improve the density of the fabric

to enhance the moldability of rCFRP.

4.2.6 CF length of rCFRP

We dissolved rCFRP and evaluated the change of the CF length varying Vf from 10 % to 40 %.

The optical microscopy images of CF of rCFRP are shown in Table 10.

Table 10 length of rCF

Item CFRP-m Racket Takayasu TORAYCA

Vf=10 %

0

Vf=40 %

0

void

500 μm

500 μm

29

As shown, the length of CF of 40 % Vf was shortened about 1 mm～2 mm comparing with the

length of about 30 mm at Vf of 10 % in CFRP-m, Racket and TORAYCA. We thought that the

breaking of fiber may be caused by the increased contact point of fibers of high Vf. On the other

hand, the ratio of cut fiber was low in Takayasu.

4.2.7 Mechanical properties of rCFRP

We evaluated mechanical properties of rCFRP using non-woven rCF fabrics following JIS K

7073(tensile properties), JIS K 7074(flexural properties) and JIS K 7077(Charpy impact

strength). The results are shown in Table 11 and Figure 20-Figure 22.

As shown, the mechanical properties increased with increasing Vf, but after showing the

maxima and decreased with the increasing Vf. We thought that the cause of the non-linear

relation was the imperfect molding and increased ratio of shortened fibers.

Table 11 The results of mechanical properties of rCFRP Item Unit CFRP-m Racket Pyrolysis TORAYCA

Fiber form -Matrix resin -Tensile　strength MPa 146 97 145 146Tensile　modulus GPa 5.8 6.7 5.5 5.6Flexural strength MPa 300 200 250 250Flexural modulus GPa 20 17 23 21Charpy impact strength kJ/m2 26 14 15 16

CF non-woven fabricAnhydride cured Bisphenol-A type EP

30

0

50

100

150

200

0 5 10 15 20 25 30 35 40

Tens

il st

reng

th (M

Pa)

Volume content of fiber (vol%)

0

1

2

3

4

5

6

7

8

0 5 10 15 20 25 30 35 40

Tens

il m

odul

us(G

Pa)

Volume content of fiber (vol%)

Figure 20 Tensile properties of rCFRPs

●：Takayasu

■：TORAYCA

◆：Racket

▲：CFRP-m

●：Takayasu

■：TORAYCA

◆：Racket

▲：CFRP-m

31

0

50

100

150

200

250

300

350

0 5 10 15 20 25 30 35 40

Flex

ural

stre

ngth

(MPa

)

Volume content of fiber (vol%)

0

5

10

15

20

25

0 5 10 15 20 25 30 35 40

Flex

ural

mod

ulus

(GPa

)

Volume content of fiber (vol%)

Figure 21 Flexural properties of rCFRPs

●：Takayasu

■：TORAYCA

◆：Racket

▲：CFRP-m

●：Takayasu ■：TORAYCA

◆：Racket

▲：CFRP-m

32

0

5

10

15

20

25

30

0 10 20 30 40 50

Cha

rpy

impa

ct st

reng

th （

ｋJ/

m2 ）

Volume content of fiber (vol%)

Figure 22 Charpy impact strength of rCFRPs

4.2.8 Conclusion

The mechanical properties of rCF and rCFRP using rCFs recovered from waste CFRP was

investigated.

・The mechanical and surface properties of the recovered CF monofilament was as the same as

the fresh one.

・The mechanical properties of rCFRP increased with the increasing volume content of fiber, but

there were maxima and did not have a linear relation to the volume content of CF.

・We thought that the causes of the non-linear relation were both imperfect molding and increase

in the ratio of shortened fibers.

・The mechanical properties of rCFRP is equal to that of Takayasu and TORAYCA.

●：Takayasu ■：TORAYCA

◆：Racket

▲：CFRP-m

33

5. Recycling of Recovered EP

5.1 Anhydride Cured EP 43)-46)

Acid anhydride (Ah) cured EP (EP/Ah) usually used as the matrix resin of CFRP has ester

bond by the reaction between epoxy group and anhydride. So, we investigated the

depolymerization of EP/Ah under ordinary pressure and found that EP/Ah easily dissolves into

BZA in standard conditions, finally recovering carbon fiber (Figure 25) .

5.1.1 Experimental

5.1.1.1 Materials

The model epoxy was a commercial phenylglycidylether (PGE). The Ahs were succinic anhydride

(ScAh), cis-1, 2-cyclohexanedicarboxylic anhydride (ChAh), cis-4-cyclohexene-1, 2-dicarboxylic

anhydride (CheAh) purchased from Tokyo Chemical Industry Co., Ltd.

The hardening accelerator was 2,4,6-tris (dimethylaminomethyl) phenol purchased from Tokyo

Chemical Industry Co., Ltd. K3PO4 and BZA were commercially available from Taiyo Chemical

Industry Co., Ltd and Sun Chemical Co., Ltd, respectively.

5.1.1.2 Synthesis of a model compounds The model compounds that consist of PGE and Ahs were introduced in a test tube. The

OOO

O

O

O

O+

OO

OH

OO

O

OO

O

HO

O OH

O

O

O

O

O OHO

O

transesterification

prepolymersmonomers

Epoxy resin Acid anhydride

Figure 23 Depolymerization of epoxy resin cured with acid anhydride

34

reactive mixture was heated at 80 ˚C for 1.0 h followed by the heating at 150 ˚C for 1.0 h. An oil bath was used to heat the mixture in a test tube.

5.1.1.3 Depolymerization of a model compounds

The model compounds, BZA and K3PO4 were introduced in a test tube. The reactive mixture was heated at 180 ˚C for 10.0 h in an oil bath. 5.1.1.4 Characterization

1H NMR and 13C NMR spectra were measured at a room temperature with a JEOL AL-400

(400 MHz) NMR spectrometer. Model compounds were dissolved in d6-dimethylsulfoxide

(d6-DMSO).

In a suffix of signals, alphabet shows the degree of signal intensity （s: strong, m: middle, w:

weak） and numerals show number of signals.

"Spectra Database for Organic Compounds SD'S" was used to assign the NMR spectrum. This

is a free site organized by National Institute of Advanced Industrial Science and Technology

(AIST), Japan.

5.1.2 Results and discussion

5.1.2.1 Polymerization reaction of model compounds

The model compounds were synthesized stoichiometrically by the reaction between PGE and

Ahs with a catalyst. NMR signal assignment of PGE, ScAh, ChAh and CheAh are shown in

Figure 26, and the estimated model polymerization reaction between PGE and Ah is shown in

Figure 27.

O CH2 CH CH2

O

PGE

121.44

43.95，2.85/2.71

6.84～7.30

69.13 4.29/3.83

49.96，3.31

159.04

115.25 130.16

ScAh O

O

O

CH

HC

170.65 28.34，3.011

O

CC

CC

HHHH

H O

OC

C173.16

40.43，3.14/3.17 24.86，1.81/1.90

ChAh

22.29，1.50/1.53

CC

CC

HH

H

O

C

CO

OH

CheAh

174.26

24.49

2.29/2.34

40.09

5.98

128.31，3.52

Figure 24 Assignment of PGE, ScAh, ChAh and CheAh in NMR (Italic: 1H NMR, Bold: 13C NMR)

35

5.1.2.2 Characterization of the polymerization products

(1) PGE/ScAh

The NMR signal assignment of PGE/ScAh is shown in Figure 26 and NMR signals of

polymerization product, PGE and ScAh are shown in Table 12 and Table 13.

The disappearance of epoxide signals of PGE at both σ 2.72 ppm and σ 2.86 ppm of 1H NMR

and at both σ 44.54 ppm and σ 50.10 ppm of 13C NMR indicates the cleavage of epoxides and

progress of polymerization reaction. 1H NMR signal at σ 3.39 ppm and 13C NMR signal at σ 66.77 ppm may suggest the possibility

of unreacted materials or polymerization reaction between PGEs.

PGE/Ah

COCH2CH

O

CH2O R C

O

O

C

O O

OCH2CHCH2

O

OCR

O CH2 CHO

CH2

OC

OCO

R＋

PGE

Ahs

CH2 CH2R =

Figure 25 Estimated model polymerization reaction between PGE and Ah

C

O

OCH2CH2 C

O

O CH2

O CH2

OO

CH CH2 O C CH2 CH2 C

O

OCH2CH1

(a) (b) (c) (d)

(e)

2 3

4

5 6

7 8

9

Figure 26 Assignment of PGE/ScAh in NMR( alphabet :1H, numeral :13C )

36

Table 12 Signal assignment and chemical shifts in 1H NMR

1H NMR σ(ppm)PGE/ScAh PGE ScAh

(a ) 2.58s12.722.86 2.90

3.39s1 3.32(d ) 3.93w1 3.94(d ) 4.02w2(b ) 4.12m1 4.17(b) 4.29m1(c) 5.30m1(e ) 6.94s3 6.94(e ) 7.27m1 7.25

Assignment

Table 13 Signal assignment and chemical shifts in 13C NMR

13C NMR σ(ppm)PGE/ScAh PGE ScAh

1 28.54m3 28.3844.5450.10

4 62.16m13 65.91m2

66.77w15 69.68m2 68.707 114.58s2 114.649 120.85m3 121.188 129.52s1 129.50

6 158.02m2 158.522 171.61m5 170.65

Assignment

The NMR signal assignment of PGE/ScAh is shown in Figure 29 and NMR signals of a

polymerization product, PGE and ChAh are shown in Table 14 and Table 15.

37

Table 14 Signal assignment and chemical shifts in 1H NMR

CHCH C O CH2 CH CH2 O

O

O CH2

OO

CH CH2 O C CH CH C

O

C

O O

11 1

(a)

(b) (c)

(f)

(g)

3

4 5 8

9 10 6

7

(d) (e)

2

Figure 27 Assignment of PGE/ChAh in NMR( alphabet :1H, numeral :13C )

1H NMR σ(ppm)PGE/ChAh PGE ChAh

(b ) 1.32m11.511.52

(b ) 1.67m1(c ) 1.84m1 1.84(c ) 1.92w1 1.91

2.722.86

3.153.16

(a ) 3.34s6 3.32(f ) 3.85w3 3.94(f ) 4.06w1(d ) 4.24w1 4.17(d ) 4.30w4(e ) 5.26w1(g ) 6.95m4 6.94(g ) 7.30m5 7.25

Assignment

38

Table 15 Signal assignment and chemical shifts in 13C NMR

13C NMR σ(ppm)PGE/ChAh PGE ChAh

3 22.88w1 21.942 24.52w1 23.701 41.72w2 40.43

44.5450.10

6 62.01w15 65.94w1

66.81m17 69.01m3 68.709 114.43s2 114.6411 120.86m3 121.1810 129.52s2 129.50

8 158.24m3 158.524 172.29m2 173.16

Assignment

The disappearance of epoxide signals of PGE at both σ 2.72 ppm and σ 2.86 ppm of 1H NMR

and at both σ 44.54 ppm and σ 50.10 ppm of 13C NMR indicates the cleavage of epoxides and

progress of polymerization reaction. 1H NMR signal at part of σ 3.34 ppm and 13C NMR signal

at σ 66.81 ppm may suggest the possibility of unreacted materials or polymerization reaction

between PGEs.

39

(2) PGE/CheAh

The NMR signal assignment of PGE/CheAh is shown in Figure 30. And NMR signals of a polymerization product, PGE and CheAh are shown in Table 16 and Table 17. Table 16 Signal assignment and chemical shifts in 1H NMR

1H NMR σ(ppm)PGE/CheAh PGE CheAh

(b ) 2.27m1 2.26(b ) 2.36m1 2.40

2.722.86

(a ) 3.02m13.36s1 3.32 3.52

(f ) 3.89w1 3.94(f ) 4.03m1(d ) 4.22w1 4.17(d ) 4.30w2(e ) 5.24w1(c ) 5.56m1 5.95(g ) 6.88m1 6.94(g ) 7.24m1 7.25

Assignment

40

Table 17 Signal assignment and chemical shifts in 13C NMR

13C NMR σ(ppm)PGE/CheAh PGE CheAh

2 25.24w1 24.491 40.13m1 40.09

44.5450.10

6 62.12w15 65.86w1

66.50w17 69.50w1 68.709 114.46s1 114.6411 120.61m2 121.183 124.86m110 129.45s1 129.50 128.318 157.97m2 158.524 171.97w2 174.26

Assignment

The disappearance of epoxides signals from PGE at both σ 2.72 ppm and σ 2.86 ppm of 1H

NMR and at both σ 44.50 ppm and σ 50.10 ppm of 13C NMR indicated the cleavage of epoxides

and progress of polymerization reaction. σ 3.36 ppm of 1H NMR signal and σ 66.50 ppm of 13C

NMR signal may suggest the possibility of unreacted materials or polymerization reaction

between PGEs.

CHCH C

O

O CH2

O

O

C

O CH2

OO

CH CH2 O C CH CH C

O

CH CH2 O 11

(a)

(b)

(e) (d)

(g)

2

3

4 5 8

9 10

(c)

7

6

(f)

1

Figure 28 Assignment of PGE/CheAh in NMR( alphabet :1H, numeral :13C )

41

5.1.2.3 Proposed depolymerization mechanism

In Figure 31, we showed the proposed mechanism of the depolymerization of thermosets under

ordinary pressure by the BZA. As shown, the most probable mechanism is thought to be

transesterification.

Following the proposed mechanism shown in Scheme 2, the depolymerization products with

BZA may be dibenzyl ester compounds from the Ah and 3-phenoxy-1, 2-propanediol from PGE.

5.1.2.4 Characterization of the depolymerization products

(1)Depolymerization of PGE/ScAh (DEP(PGE/ScAh))

The NMR signals of DEP(PGE/ScAh), PhPD and DbSuc are shown in Table 18 and Table 19.

NMR signal assignment of DEP(PGE/ScAh) is shown in Figure 32. Spectra always contain

signals of BZA as a solvent.

Dibenzyl ester compounds

CH2CH2 O C R C O

O O

＋ O CH2 CH

OH

CH2 OH

3-phenoxy-1, 2-propanediol

Transesterification

COCH2CH

O

CH2O R C

O

O

C

O O

OCH2CHCH2

O

OCR

PGE/Ahs

CH2 CH2R =

Figure 29 Estimated model depolymerization reaction of PGE/Ah

42

Table 18 Signal assignment and chemical shifts in 1H NMR 1H NMR σ(ppm)

DEP(PGE/ScAh) PhPD DbSuc BZA

2.50w5 2.24

(i ) 2.63m3 2.69

(h ) 3.39w1 3.49

(d ) 3.47w2 3.69

(d ) 3.85w6 3.78

(f ) 3.94w2 3.83

(g ) 4.01w2 3.96

(e ) 4.12w1 4.07

4.51m1 4.64

(j ) 5.19w1 5.11

(a ) 6.93m12 6.87

(c ) 6.94

(b ) 7.23w9 7.24 7.25

(k) 7.32s20 7.32 7.35

Assignment

43

Table 19 Signal assignment and chemical shifts in 13C NMR 13C NMR σ(ppm)

DEP(PGE/ScAh) PhPD DbSuc BZA

14 28.66m1 29.13

7 62.92s2 63.73

65.42m2 65.16

12 66.77w1 66.46

5 69.45w2 69.05

6 69.99w1 70.62

3 114.43m3 114.62

1 120.39m2 121.28

126.61s3 126.92

8 127.78m2 127.39 127.54

9 128.03w2 128.19

10 128.24w2 128.52 128.46

2 129.44m2 129.55

11 136.06m1 135.81

142.53m1 140.81

4 158.77w2 158.47

13 171.88w2 171.94

Assignment

12 9 10

O CH2 CH

OH

CH2 OH(a)

(b) (c) (d) (e)

(f)

(g)(h) (i) 1

2 3

4 5 6 7

PhPD

DbSuc (j)

CH2CH2 O C

O

C

O

OCH2 CH2

(k) (l) 8 11

13 14

Figure 30 Assignment of depolymerization products in NMR

44

As a result of NMR analyses, the signals of DEP(PGE/ScAh) consisted with those of PhPD

and DbSuc. So we thought that DEP(PGE/ScAh) consists of PhPD and DbSuc.

(2) Depolymerization of PGE/ChAh(DEP(PGE/ChAh))

The NMR signals of DEP(PGE/ChAh) are shown in Table 20 and Table 21. NMR signal

assignment of DEP(PGE/ChAh) is shown in Figure 33.

Table 20 Signal assignment and chemical shifts in 1H NMR

1H NMR σ(ppm)DEP(PGE/ChAh) PhPD DbHp BZA

(l ) 1.26w1 1.51(m ) 1.34w1 1.52(j ) 1.66w1 1.84(k ) 1.85w1 1.91

2.49m5 2.24(i ) 2.60w1 3.15(i ) 2.80w1 3.16(h) 3.43w1 3.49(d ) 3.50w2 3.69(d ) 3.78(f ) 3.80w1 3.83(g ) 3.90w1 3.96(e ) 4.07w3 4.07

4.52s1 4.64(n ) 5.21m1 5.11(a ) 6.92m5 6.87(c ) 6.94(b ) 7.23m9 7.24 7.25(o ) 7.31s14 7.32 7.35

Assignment

45

Table 21 Signal assignment and chemical shifts in 13C NMR

13C NMR σ(ppm)DEP(PGE/ChAh) PhPD DbHp BZA

16 23.15w1 21.9415 24.61w1 23.70

25.77w128.48w1

14 41.71w1 40.4344.33w1

7 62.86s1 63.7365.45w1 65.1666.01w1

12 66.35w1 66.465 69.48w1 69.056 70.04w2 70.623 114.45m1 114.621 120.64w3 121.288 126.63s3 127.39 126.929 127.80m2 128.19 127.54

10 128.40m1 128.52 128.462 129.48m1 129.55

11 136.11w2 135.81142.56s1 140.81

4 158.52w3 158.4713 174.01w2 171.94

Assignment

PhPD

11 8 14

16

(k) (l)

O CH2 CH

OH

CH2 OH1

(f)

(a)

(b) (c) (d) (e) (g)(h)

(i)

2 3

4 5 6 7

DbHp

CH2CH2 O C

O

CH CH C

O

O

9 12

10 13 15 (o), (p)

(m), (n)

(j)

Figure 31 Assignment of depolymerization products in NMR

46

As a result of NMR analyses, the signals of DEP(PGE/ChAh) consisted with those of PhPD

and DbHp. So we thought that DEP(PGE/ChAh) consisted of PhPD and DbHp. On the other

hand, signals of σ 25.77 ppm, σ 28.48 ppm, σ 44.33 ppm and σ 66.01 ppm of 13C NMR were not

assigned to PhPD and DbHp, suggesting the existence of the side reaction products.

(3)Depolymerization of PGE/CheAh (DEP(PGE/CheAh)

The NMR signals of DEP(PGE/CheAh), PhPD and DbTp are shown in Table 22 and Table 23.

NMR signal assignment of DEP(PGE/CheAh) is shown in Figure 34.

Table 22 Signal assignment and chemical shifts in 1H NMR

1H NMR σ(ppm)DEP(PGE/CheAh) PhPD DbTp BZA

(m ) 2.29w1 2.26(m ) 2.38w1 2.40

2.49w3 2.243.03w1

(h ) 3.40w1 3.49(d ) 3.47w2 3.69 3.52(d ) 3.78(f ) 3.80w1 3.83(g ) 3.92w1 3.96(e ) 4.07w2 4.07

4.51s1 4.64(m ) 5.20w1 5.11(l ) 5.61w3 5.95(a ) 6.91m2 6.87(c ) 6.94(b ) 7.23m12 7.24 7.25(o ) 7.31s9 7.32 7.35

Assignment

47

Table 23 Signal assignment and chemical shifts in 13C NMR

13C NMR σ(ppm)DEP(PGE/CheAh) PhPD DbTp BZA

15 25.43w1 24.4914 40.13m1 40.097 62.93s2 63.73

65.68w3 65.1612 66.79w1 66.465 69.46w1 69.056 70.00w1 70.623 114.48w1 114.621 120.62w3 121.28

16 124.85w1 126.68126.61s3 126.92

8 127.76m2 127.39 127.549 128.04w1 128.19 128.46

10 128.36m2 128.522 129.46w1 129.55

11 136.00w1 135.81142.53m1 140.81

4 158.01w2 158.4713 172.01w1 171.94

Assignment

PhPD

DbTp

O CH2 CH

OH

CH2 OH1

(f)

(a)

(b) (c) (d) (e) (g)(h)

(i)

2 3

4 5 6 7

CH2CH2 O C

O

CH CH C

O

O

10

11 13

(o)

15

16 (j)

(k) (l)

(m), (n) 9

8 12

14

Figure 32 Assignment of depolymerization products in NMR

48

As a result of NMR analyses, the signals of DEP (PGE/CheAh) consisted with those of PhPD

and DbTp. So we thought that DEP (PGE/CheAh) consisted of PhPD and DbTp. On the other

hand, signals of σ 3.03 ppm of 1H NMR was not assigned to PhPD and DbTp, suggesting the

existence of the side reaction products.

5.1.2.5 Conclusions

In this work the chemical recycling of the epoxy resin cured with Ahs were investigated.

Synthesis of a model compounds enable to study the polymerization and depolymerization

reaction.

The polymerization and depolymerization products were almost identified by NMR analyses.

From the NMR analyses of depolymerization of PGE/Ah model compounds, we found that

dibenzylesters and bisdiols are formed by the depolymerization of Ahs reacted with model EP

under ordinary pressure. The depolymerization mechanism of PGE/Ah was the

transesterification with BZA, so we estimated that the depolymerization reaction of EP cured

with Ah under ordinary pressure occured mainly transesterification reaction accelerated with

K3PO4 as a catalyst.

49

5.2 Amine Cured EP 47)-50) We have developed the depolymerization of thermosets under ordinary pressure. This method

uses tripotassium phosphate (K3PO4) as a catalyst and benzyl alcohol (BZA) as a solvent to depolymerize anhydride cured epoxy resin (EP) and unsaturated polyester resin (UP). It is possible to recycle thermosets based on transesterification. Although depolymerization of resins takes more time, our method is also applicable to depolymerize amine (Am) cured EP often used for CFRP. Besides we have developed the depolymerization of thermosets by using subcritical fluids. Insoluble or slightly soluble thermosets like Am cured EP was possible to be depolymerized in a short time by this method. Am cured EP doesn’t have ester bond, so we investigated the depolymerization mechanism of Am cured EP. 5.2.1 Experimental

5.2.1.1 Materials

Bisphenol A diglycidylether (DGBPA), the epoxy resin of epoxy equivalent weight (EEW)

170-175, was used as the model matrix. Isophrondiamine (IPDA) and

2,4,6-tris(dimetylaminomethyl) phenol (DMAmP) were used as the curing agent and catalyst.

In order to analyze the depolymerization mechanism, we investigated the reactions of secondary

and tertiary Am as the model compound of Am cured EP. The selected Ams were

Dicyclohexylamine (DChAm), N,N-Dicyclohexylamine (DChMeAm), Dibenzylamine

(DBzAm) and Tribenzylamine (TBzAm).

5.2.1.2 Synthesis of a model matrix

DGBPA, IPDA and DMAmP of the ratio of 100 : 25 : 2 were mixed in the aluminum plates.

This reactive mixture was heated at 80 ˚C for 0.5 h. The additional heating was 150 ˚C for 1.0 h.

The obtained cured EP of about 5 mm thick was cut into small pieces (size: 10 mm × 15 mm )

for further testing.

5.2.1.3 Depolymerization of a model matrix and model Am compounds

Piece of a model matrix, BZA and K3PO4 were introduced into a tube bomb reactor. The

reaction temperatures ranged from 250 ˚C to 325 ˚C in an inert gas oven and reaction time was

from 1.0 h to 4.0 h.

Am compound (DChAm, DChMeAm, DBzAm, TBzAm), BZA and K3PO4 were introduced in

the reactor, and then the reactor was heated at 280 ˚C for 6.0 h in an inert gas oven.

50

5.2.1.4 Characterization 1H and 13C NMR spectra of test pieces in DMSO and CDCl3 were measured with JEOL AL

400 operating at 200 MHz sing.

5.2.2 Results and discussion

5.2.2.1 Depolymerization of a model matrix

The solubility of the test pieces were measured changing the temperature from 250 ˚C to 325

˚C. The solubility (%) was calculated by the next equation.

Solubility (%) = (A-B)*100/A

A (g):Mass of a test piece before treatment

B (g):Mass of a test piece after treatment

The results are shown in Table 24 and Figure 33, Figure 36 shows the temperature profile. The

test pieces before / after the treatment were shown in Figure 37 and Figure 36 respectively.

Table 24 Solubility of test pieces

Treatment time (h) 0.0 1.0 2.0 3.0 4.0Treatment temperatuer (℃) 0.0 250.0 275.0 300.0 325.0Solubility (％) 0.0 0.0 2.6 15.3 100.0

51

0.0

20.0

40.0

60.0

80.0

100.0

120.0

0.0 1.0 2.0 3.0 4.0 5.0

Solu

bilit

y (％

)

Treatment time (h)

Figure 33 Solubility of test pieces

Figure 34 Temperature profile of subcritical fluids process

250 ˚C 275 ˚C

300 ˚C

325 ˚C

Figure 35 A test piece before treatment

52

Figure 36 Test pieces after treatment

As shown in Table 24 and Figure 35, the test pieces dissolved perfectly after 4.0 h at 325 ˚C. It

was difficult to analyze the depolymerization products of EP cured with Am which has

cross-linked three-dimensional networks. So we analyzed the depolymerization mechanism by

using secondary and tertiary Am as model compounds of Am cured EP.

5.2.2.2 Depolymerization of model Am compounds

To identify the structure of products, Am compounds and depolymerization products were

characterized by NMR. In a suffix of signals, the alphabet shows the degree of signal intensity

and numerals show the number of signals. The results of NMR analyses are as follows.

5.2.2.3 Alicyclic Am compounds

The NMR signals of before and after the depolymerization reaction of DChAm and

DChMeAm are shown in Table 25 and Table 26. The attributions of NMR signals are shown in

Figure 39.

1.0 h 2.0 h 3.0 h 4.0 h

53

Table 25 The 1H and 13C chemical shifts in NMR spectra of DChAm (σ, ppm)

Attribution 1H NMRBefore After BZA

(j ) 0.7w1 0.7w1(g ) 1.03s10 1.00m8(f ) 1.15s10 1.15m8(e ) 1.24s10 1.22m5(d ) 1.61s8 1.60m8(c ) 1.71s5 1.70m5(b ) 1.86s6 1.81m4

2.14s1 2.24s11(a ) 2.55s9 2.52m7

3.42w4 4.64s1 4.64s15 7.26m5 7.25m13 7.33s7 7.34s17

Attribution 13C NMRBefore After BZA

4 25.31s5 25.19s43 26.27s3 26.02s12 34.26s4 33.96s41 52.93m3 52.90m1

64.60m1 65.14s2 126.82s1 126.92s1 127.23s1 127.54s1 128.30s5 128.46s1 129.00w1 141.30m1 140.81s1

54

Table 26 The 1H and 13C chemical shifts in NMR spectra of DChMeAm (σ, ppm)

Attribution 1H NMRBefore After BZA

(j ) 1.06m9 1.06w6(g ) 1.16m8 1.15m5(f ) 1.24s8 1.21s5(e ) 1.61m12 1.58m4(d ) 1.77s4 1.62m3(c ) 1.86s1 1.78s5(b ) 2.26s7 2.17s1 2.24s11(a ) 2.48m12 2.47w9

3.42w44.62s1 4.64s15

7.26m12 7.25m137.33s3 7.34s17

Attribution 13C NMRBefore After BZA

5 25.45s2 24.27s24 26.26s4 26.16s33 30.50s5 30.02m32 32.74m1 32.67m11 59.44s4 59.10s2

64.71s1 65.14s2126.82s1 126.92s1127.24s3 127.54s1128.50s5 128.46s1141.18m1 140.81s1

CC C

CC

C

NH

HH

H

H

H

H

H

H

H

HH

C

C

CC

CC

C

C

CC

CC

CC C

CC

C

N

CH3

H

HH

H

H

H H

H

HH

H

4 (b)

(c)

(d)

(e) (f)

(g)

(j) 1 2 5 3

1 2 3 4

(a) (j)

(c)

(d)

(e)

(f)

(g) (b) (a)

Figure 37 Attribution of NMR Signals of DChAm and DChMeAm

55

As shown in the Tables, there were no difference between NMR signals of DChAm and

DChMeAm before and after the depolymerization treatment. So, we considered that the

depolymerization reaction by subcritical fluids didn’t occur in alicyclic Ams.

5.2.2.4 Aromatic amine compounds

The NMR signals of before and after the depolymerization reactions of DBzAm and TBzAm

are shown in Table 27 and Table 28. The attributions of NMR signals of DBzAm and TBzAm

are shown in Figure 40. The attributions of NMR signals and chemical shifts of toluene,

benzaldehyde and benzylamine are shown in Figure 39.

56

Table 27 The 1H and 13C chemical shifts in NMR spectra of DBzAm (σ, ppm)

Attribution 1H NMRBefore After BZA Toluene Benzaldehyde

(c ) 1.64m1 2.33m1 2.24s11 2.32 3.54w1 3.42w4

(b ) 3.78s15 3.73s1 4.58s1 4.77w2 4.64s15

(a ) 7.22m15 7.18w4 7.00 7.27m23 7.25m13

(a ) 7.32s13 7.38s7 7.34s17 7.38 7.48w5 7.51 7.74w4 7.61 7.82w2 7.87 8.34w1 9.94w1 10.00

Attribution 13C NMRBefore After BZA Toluene Benzaldehyde

21.34w1 21.415 53.06s3 52.88s3

57.80w164.82s2 65.14s2

125.50w2 125.384 126.60s2 126.75s4 126.92s13 127.73s3 127.47s4 127.54s12 128.28s7 128.52s11 128.46s1 128.28 128.98

130.75m3 129.09 129.68134.37w1 134.43135.86w1 136.47137.74w2 137.83139.79m3

1 140.25m1 141.04w1 140.81s1162.35w2192.41w1 192.28

57

Table 28 The 1H and 13C chemical shifts in NMR spectra of TBzAm (σ, ppm)

Attribution 1H NMRBefore After BZA DBzAm Toluene Benzaldehyde

1.79w1 1.64m1 2.13w1 2.33w1 2.24s11 2.32

(d ) 3.63s13 3.54s1 3.42w4 3.73w4 3.78s15 4.61s4 4.64s15

(c ) 7.09m7 7.14w7 7.00(b ) 7.22s6 7.25m14 7.25m13 7.22m15(a ) 7.40s4 7.33s10 7.34s17 7.32s13 7.38

7.47m2 7.51 7.58w3 7.61 7.83w4 7.87 9.95w1 10.00

Attribution 13C NMRBefore After BZA DBzAm Toluene Benzaldehyde

21.38w1 21.4152.88s1 53.06s3

5 57.90s3 57.83s165.14s1 65.14s2

125.24w1 125.384 126.82m1 126.89s3 126.92s1 126.60s23 127.45s2 127.40s4 127.54s1 127.73s32 128.54s7 128.45s6 128.46s1 128.28s7 128.28 128.98

129.69w1 129.09 129.68134.41w1 134.43

137.83 136.471 139.61m1 139.57m1

140.83m1 140.81s1 140.25m1192.45w1 192.28

58

Toluene Benzaldehyde Benzylamine

Figure 39 Attribution of NMR signals (σ, ppm) of Toluene, BZA and Benzylamine As shown in the Tables, the NMR spectra excepting the signals of DBzAm, TBzAm and BZA

correspond to that of toluene and benzaldehyde. And as for TBzAm, 1H NMR signals at 1.79

ppm and 3.73 ppm and 13C NMR signal at 52.88 ppm are attributed to DBzAm’s -NH- and

-CH2-. Therefore we considered DBzAm came from the depolymerized TBzAm.

We anticipated that benzylamine may be generated by the depolymerization of DBzAm but

NMR signals of –NH2 and -CH2- were not observed in Table 28. Both 1H NMR signal at 8.34

ppm and 13C NMR signal at 162.4 ppm in the table were close to the imino group signals of

N-Benzylideneaniline shown in Figure 42. So, those signals were presumed to be attributed to

the imino group.

7.61

46.4

C

C C

C

CC

C

HH

H

H H

H

H

H

125.4

21.4

129.1 128.3

137.8

2.34

7.00～7.38

C

C C

C

CC

C

HH

H

H H

H

O

134.4

129.7 129.0

192.3 126.4

7.51 7.87

10.00 C

C C

C

CC

C

H

H

H

H

H

H

H

N

H

H143.3

127.0 128.5

126.7

7.07～7.49

3.84

1.52

C

C C

C

CC

CH2

HH

H H

H C

C C

C

CC

CH2NH

(a)

5 1

2 3 4

(c) (b)

C

C C

C

CC

C

C C

C

CC

CH2 N CH2

CH2

CC

CC

C

C

H

H

H

H

H

1 2 3

4 5

(a)

(b)

Figure 38 Attribution of NMR signals of DBzAm and TBzAm

59

The possible depolymerization mechanism of DBzAm and TBzAm with subcritical fluids is

shown in Figure 43 and Figure 44.

In the case of DBzAm，the reaction started with a homolysis (1). BZA reacted with benzyl

radical to form dibenzyl ether (2). The hydrogen atom of methylene was released from

benzylamino radical to give imine compound (3). Depolymerization reaction produced

benzaldehyde and toluene from dibenzyl ether (4).

Imine compound

CH NH

CH2 NH CH2

DBzAm C-N cleavage Benzyl radical Benzylamino radical

CH2 ＋ CH2ENH

CH2 O CH2CH2

BZA Dibenzyl ether

＋ HO CH2

CH2 NH

7.19～7.37

C

C C

C

CC

C

H

H

H

H

H

H

N C

C C

C

CC

H

H

H

H

H

7.46～7.89

8.42 131.3

128.7 128.7 129.0

C

C C

C

CC

CH2 N C

C C

C

CC

129.0

160.2

125.8

152.0 136.1

Figure 40 NMR signal attribution of N-Benzylideneaniline (σ, ppm)

Figure 41 Depolymerization mechanism of DBzAm

60

In the case of TBzAm, the reaction also started with a homolysis and then forms DBzAm (5).

BZA reacted with benzyl radical to form dibenzyl ether (6). Depolymerization reaction

produced benzaldehyde and toluene from dibenzyl ether (7).

5.2.2.5 Conclusion

In order to analyze the depolymerization mechanism, we investigated the reactions of

secondary and tertiary Am as model compounds of Am cured EP and conducted the

depolymerization reaction. We thought that the cleavage of the C-N bond was occurred by the

depolymerization reaction with subcritical fluids. As a result, the cleavage of the C-N bond

didn’t occur in alicyclic Am by the depolymerization reaction. On the other hand, we found that

there was the possibility of cleaving the C-N bond in aromatic Am compounds with the

depolymerization reaction by subcritical fluids.

Because IPDA is alicyclic Am，the cleavage of the ether linkage (C-O-C) or transetherification

may occur. We are going to study the depolymerization mechanism of Am compound having

hydroxyl group in the future.

CH2 O CH2

Dibenzyl ether

CH2 N CH2

CH2

CH2 NH CH2

C-N cleavage CH2

CH2 O CH2 CH3CHO ＋

Toluene Benzaldehyde

CH2 ＋ HO CH2

BZA

TBzAm Benzyl radical

Figure 42 Depolymerization mechanism of TBzAm

61

6. Conclusion

6.1 Recycling of CF

The mechanical properties of rCF and rCFRP using rCFs recovered from waste CFRP were

excellent. The mechanical properties of rCFRP increased with the increasing volume content of

fiber, but there were maxima and did not have a linear relation to the volume content of CF. We

thought that the causes of the non-linear relation were both imperfect molding and increase in

the ratio of shortened fibers.

6.2 Recycling of EP

We also investigated the chemical recycling of EP cured with Ahs and Ams using model

compounds. From the NMR analyses of depolymerization of PGE/Ah model compounds, we

found that dibenzylester and bisdiol are formed by the depolymerization of the thermosets under

ordinary pressure. The depolymerization mechanism of PGE/Ah was the transesterification with

BZA, so we concluded that the depolymerization were mainly transesterification reaction

accelerated with K3PO4.

While the investigation of EP/Am model compounds, the cleavage of the C-N bond didn’t

occur in alicyclic Am by the depolymerization reaction. On the other hand, we found that there

was the possibility of cleaving the C-N bond in aromatic Am compounds with the

depolymerization reaction by subcritical fluids. Because IPDA is alicyclic Am，the cleavage of

the ether linkage (C-O-C) or transetherification may occur. We are going to study the

depolymerization mechanism of Am compound having hydroxyl group in the future.

62

References

1) 山藤家嗣, "炭素繊維の環境負荷性能とリサイクル，土木学会平成 20 年度全国大会研

究討論会 研－05 資料，pp.10-11

(http://www.jsce.or.jp/committee/fukugou/zenkoku/2008.pdf) (2008)

2) Idzumi Okajima, Kaori Watanabe, and Takeshi Sako, "Chemical Recycling of Carbon Fiber

Reinforced Plastic with Supercritical Alcohol", Journal of Advanced Research in Physics

vol. 3, no. 2, 021211 (2012)

3) 岡島いづみ, 佐古猛, "炭素繊維強化プラスチックのリサイクル, 工業材料, vol. 56, pp.

70-72 (2008)

4) Motonobu Goto, "Chemical recycling of plastics using sub- and supercritical fluids", The

Journal of Supercritical Fluids, vol. 47, pp. 500-507 (2009)

5) 後藤元信, "超臨界・亜臨界流体を利用した繊維強化プラスチックのリサイクル", vol. 65，

pp. 62-66 (2009)

6) Guozhan Jiang, Stephen J. Pickering, Edward H. Lester, Nick A. Warrior,

"Characterization of carbon fibres recycled from carbon fibre/epoxy resin composites using

supercritical n-propanol", Composites Science And Technology, vol. 69, no. 2, pp. 192-198

(2009)

7) Guozhan Jiang, Stephen J. Pickering, Edward H. Lester, Nick A. Warrior, "Decomposition

of Epoxy Resin in Supercritical Isopropanol", Industrial Engineering Chemistry Research,

vol. 49 no. 10, pp. 4535–4541 (2010)

8) Dang Weirong, Masatoshi Kubouchi, Hideki Sembokuya, Ken Tsuda, Kazuyoshi Arai,

"Decomposition of Amine Cured Epoxy Resin in Nitric Acid for Recycling" , Proc. 13th

Intl. Conf. on Compos. Materials, ID-1340 (2001)

9) Jin Mizuguchi, Yuichro Tsukada, Hiroo Takahashi, "Recovery and Characterization of

Reinforcing Fibers from Fiber Reinforced Plastics by Thermal Activation of Oxide

Semiconductors", Materials Transactions, vol. 54, no. 3, pp. 384 - 391 (2013)

10) Katsuji Shibata, "Depolymerization of a Brominated Polyhydroxyether with Alkali Metal

Compounds as Catalysts", Preprints of 7th SPSJ International Polymer Conference,

pp. 133 (1999)

11) Katsuji Shibata, Hiroshi Shimizu, "Solubility of Glass-Brominated Epoxy PWB in the

Solutions of Alkali Metal Compounds and Organic Solvents", Electronics Goes Green

2000+ Proceedings, vol. 1, pp. 391-396 (2000)

63

12) 柴田勝司, "常圧溶解法によるプリント配線板リサイクル技術", エレクトロニクス実装学会

誌，vol. 11，no. 6，pp. 408-412 (2008)

13) 柴田勝司, "プリント配線板並びに FRP のリサイクルと GSC", 化学と教育，vol. 56, no. 7，

pp. 348-349 (2008)

14) 柴田勝司，前川一誠，池田ゆかり，廣瀬祐子，海津朋宏，伊豆名具己, "常圧溶解法によ

るモールドコイル変圧器のリサイクル技術", 第 56 回高分子学会年次大会予稿集, vol. 56,

no. 1, pp. 2352 (2007)

15) 柴田勝司, "常圧溶解法を用いた FRP リサイクル技術", 強化プラスチックス，vol. 59，

no. 7，pp. 224-229 (2013)

16) 柴田勝司，伊澤弘行，松尾亜矢子, "リサイクルを目的とした不飽和ポリエステル樹脂を主

成分とする複合材料の溶解性", 第 50 回高分子討論会予稿集，vol. 50，no. 14, p. 3981

(2001)

17) Katsuji Shibata, Hiroyuki Izawa, Kazonobu Maekawa, Mitsuru Iwai, "Properties of Fiber

Reinforced Plastic (FRP)", IUPAC Polymer Conference Preprints, pp. 640 (2002)

18) Katsuji Shibata, Kazunobu Maekawa, Masahito Kitajima, "Composites Recycling Using

Depolymerizing Thermosets under Ordinary Pressure", The Ninth Japan International

SAMPE Symposium Preprint, pp. 38-43 (2005)

19) 柴田勝司, "常圧溶解法によるポリエステル，エポキシの解重合", 高分子，vol. 57, no. 5，

pp. 365 (2008)

20) 柴田勝司, "FRP のリサイクル技術", ネットワークポリマー，vol.28，no.4，pp.247-255

(2007)

21) 柴田勝司, "常圧溶解法による熱硬化性樹脂複合材料のリサイクル技術", 日本接着学会

誌，vol.42，no.4，pp.153-157 (2006)

22) 柴田勝司，栗谷弘之，池田ゆかり，廣瀬祐子，島田勝, "常圧溶解法によるアラミドロッド

からのアラミド繊維の回収", 第 57 回高分子討論会予稿集，vol.57, no.2, pp.5202 (2008)

23) Katsuji Shibata, Hiroyuki Izawa, Ayako Matsuo, U. S. Patent 6,780,894, "Treatment liquid

for cured unsaturated polyester resin and treatment method thereof" (2004)

24) Katsuji Shibata, Hiroshi Shimizu, Ayako Iwamaru, Takeshi Horiuchi, U. S. Patent

6,962,628, "Method of treating epoxy resin-cured product (2005)

25) Katsuji Shibata, "FRP recycling technology by dissolving resins under ordinary pressure",

JEC Composites Magazine, No. 66, pp. 50-52 (2011)

64

26) 柴田勝司，前川一誠，池田ゆかり，廣瀬祐子，平澤秀典, "常圧溶解法によるテニスラケッ

トからの炭素繊維の回収", 第 56 回高分子学会年次大会予稿集, vol. 56, no. 1, pp. 2351

(2007)

27) 柴田勝司, 前川一誠, 北嶋正人, "回収炭素繊維を用いた CFRP の製法", 第 17 回廃棄

物学会研究発表会講演論文集，pp. 544 (2006)

28) 柴田勝司，前川一誠，北嶋正人, "回収炭素繊維を使用した CFRP の特性", 第 55 回高

分子討論会予稿集，vol. 55, no. 2, pp. 5525 (2006)

29) 柴田勝司, "常圧溶解法による CFRP リサイクル技術", 廃棄物資源循環学会誌，vol. 24,

no. 5, pp. 358-363 (2013)

30) 中川光俊，栗谷弘之，柴田勝司, "炭素繊維不織布を用いた CFRP の機械的性質", 第

60 回高分子学会年次大会予稿集，vol. 60, no. 1, pp. 2048 (2011)

31) 前川一誠，柴田勝司，栗谷弘之，中川光俊, "常圧溶解法を用いて CFRP から回収した炭

素繊維の環境影響評価（LCA）", 第 60 回高分子学会年次大会予稿集，vol. 60, no. 1,

pp.2088 (2011)

32) Tomoko Iwaya, Shinpei Tokuno, Mitsuru Sasaki, Motonobu Goto, Katsuji Shibata,

"Recycling of fiber reinforced plastics using depolymerization by solvothermal reaction

with catalyst“, J. Mater. Sci., vol. 43, no. 7, pp. 2452-2456 (2008)

33) 中川光俊，柴田勝司，西河裕, "回収炭素繊維不織布を用いた CFRP の機械的性質",

第 62 回高分子学会年次大会予稿集，vol. 62，no. 1，pp. 2048 (2013)

34) 柴田勝司, "常圧溶解法による CFRP リサイクル技術", プラスチックスエージ，vol. 59,

no.12, pp. 69-72 (2013)

35) Mitsutoshi Nakagawa, Katsuji Shibata, Hiroshi Nishikawa, "Characterization of CFRP

using recovered carbon fibers from waste CFRP", 4th International Symposium on Fiber

Recycling, proceedings, I-5 (2013)

36) 高瀬諒人編，”次世代自動車(EV・HV)に向けた自動車材料の樹脂化による車体軽量

化”，第 2 章，第 16 節，”CFRP のリサイクル技術～常圧溶解法を中心として～”，

pp.223-228，技術情報協会 (2013)

37) サイエンス＆テクノロジー社編，”CFRP/CFRTP の加工技術と性能評価”，第 7 章 第 3

節，”CFRP の常圧溶解法によるリサイクル技術”，pp. 234-241，サイエンス＆テクノロジー

社 (2012)

38) 福田博，邉吾一，末益博志監修，”新版 複合材料・技術総覧”，第 7 章 第 7 節，”複合

材料のリサイクル”，pp. 829-837，産業技術サービスセンター (2011)

65

39) エポキシ樹脂技術協会編，“総説エポキシ樹脂 最近の進歩Ⅰ”，第 6 章 第 1 節，”エポ

キシ樹脂複合材料のリサイクル技術”, pp. 195-201，エポキシ樹脂技術協会 (2009)

40) 鈴木淳史編，”エコマテリアルハンドブック”，第 5 章 第 2 節 第 2 項，”廃ガラス繊維系

複合材料を利用した複合材料とその特長”, pp. 280-282，丸善 (2006)

41) 柴田勝司，清水浩，松尾亜矢子，堀内猛, 特許第 4967885 号, "エポキシ樹脂硬化物のリ

サイクル方法" (2012)

42) 柴田勝司，清水浩，堀内猛，松尾亜矢子, 特許第 4765202 号, "エポキシ樹脂硬化物の

処理溶液、これを用いた処理方法および処理生成物" (2011)

43) 柴田勝司，福澤寿代，前川一誠，保木淳子, 特許第 4686991 号, "炭素材料／酸無水物

硬化エポキシ樹脂複合材料の分離方法" (2011)

44) 柴田勝司，池田ゆかり，廣瀬祐子，キタイン アルマンド，佐々木満, "酸無水物硬化エポ

キシ樹脂の加アルコール分解", 第 62 回高分子学会年次大会予稿集，vol. 62，no. 1，

pp.2053 (2013)

45) Katsuji Shibata, Mitsutoshi Nakagawa, Armando T. Quitain, Mitsuru Sasaki, "CFRP

recycling using depolymerization of acid anhydride cured epoxy resin", 9th International

Conference on Composite Materials, proceedings, pp. 8511-8518 (2013)

46) 中川光俊，春日圭一，柴田勝司，畔田博文, "モデル化合物を用いた酸無水物硬化エポ

キシ樹脂の解重合反応解析", 第 63 回ネットワークポリマー講演討論会講演要旨集，

pp.117 (2013)

47) 柴田勝司，廣瀬祐子，池田ゆかり，キタイン アルマンド，佐々木満, "アミン硬化エポキシ

樹脂の加アルコール分解", 第 62 回高分子学会年次大会予稿集，vol. 62，no. 1，

pp.2054 (2013)

48) Katsuji Shibata, Mitsutoshi Nakagawa, Armando T. Quitain, Mitsuru Sasaki, "CFRP

recycling using depolymerization of amine cured epoxy resin", 4th International

Symposium on Fiber Recycling, proceedings, O-10 (2013)

49) 春日圭一，中川光俊，柴田勝司，畔田博文, "モデル化合物を用いたアミン硬化エポキシ

樹脂の解重合反応解析", 第 62 回高分子討論会予稿集，vol. 62，no. 2，pp. 4967 (2013)

50) 柴田勝司, 中川光俊, "常圧溶解法による CFRP リサイクル技術", 日立化成テクニカルレ

ポート, vol. 56, pp. 6-11 (2013)