JOPO SCI-3

Thermal, mechanical and self-destruction properties of aluminumreinforced carbon foam

Shameel Farhan1 • Ru-Min Wang1

Published online: 22 April 2015

� Springer Science+Business Media New York 2015

Abstract Carbon foam has been developed by templating

method with phenolic resin and coal tar pitch as matrix pre-

cursor and polyurethane foam as a sacrificial template. For

improving thermal and mechanical properties, aluminum

(Al) powder with 2–8 wt% was added into the impregnation

mixture. Carbonization at 1273 K in inert environment under

the cover of coke breeze produced Al4C3 in situ and AlN by

substitution reaction. Various thermal and mechanical tests

showed a density of 0.50–0.58 g/cm3, open porosity of

64–68 %, thermal conductivity of 0.043–0.385 W/mK and a

compressive strength of 17–32 MPa for the samples con-

taining 0–6 wt%Al. Scanning electron microscope was used

to evaluate the pore morphology, distribution of Al in the

porous network and an approximate pore size distribution

which came out to be in the range of 2–200 lm. X-ray

mapping showed a homogeneous dispersion of Al with some

agglomerates into the ball shape.Carbon foamwith 8 wt%Al

showed self-destruction in 14 days after 15 days of

manufacturing. The destruction started from core and pro-

ceeded towards outer surface resulting in a core–shell effect.

X-ray diffraction analysis confirmed the destruction due to

the formation of Al4C3 and reaction with atmospheric mois-

ture forming Al(OH)3. This property can be further explored

to be applied in the field of composite tooling for advanced

carbon fiber reinforced composites. The tool will lose

strength after certain time, facilitating in easy removal from

the complex shape composites.

Keywords Carbon foam � In situ � Self-destruction �Compressive strength � Thermal conductivity �Environmental effects

1 Introduction

Carbon/aluminum composites have been receiving consid-

erable attention due to their high thermal properties, high

specific strength and modulus for application in commercial

and automobile industries [1, 2]. They offer the possibility of

thermal management applications with tailorable properties.

Graphite foams, on the other hand, are also showing a great

potential for improving the thermal management applica-

tions due to their low density, and high porosity along with

large surface area [3–5]. Googin [6] reported the first process

dedicated to controlling the structure and properties of car-

bon foams by varying the precursor material. Until now

several researchers explored a variety of applications for

carbon foam materials [7–10]. These studies pioneered me-

sophase derived graphitic foams for honeycombmaterials to

develop a highly structural material. Recently, Klett devel-

oped a new manufacturing technique to enhance the thermal

properties of carbon foams [11]. Typically, carbon foams are

prepared from organic precursors like phenolic resin, coal or

mesophase pitches [12–14]. The template carbonization

process has been regarded as the simplest and effective

method to produce carbon foam with controlled properties.

Impregnations of a precursor into the template, stabilization/

curing and template removal/conversion into carbon at high

temperature are the key controls. The most commonly used

organic precursor is a phenolic resin. Carbon foam derived

from phenolic resin can be used for the high-strength thermal

insulations in aerospace, nuclear, cryogenics and building

industry [15–17]. Coal derived and pitch based carbon foams

& Shameel Farhan

[email protected]

Ru-Min Wang

[email protected]

1 Department of Applied Chemistry, School of Science,

Northwestern Polytechnical University, Xi’an 710072, China

123

J Porous Mater (2015) 22:897–906

DOI 10.1007/s10934-015-9963-3

usually have high thermal conductivities and lower structural

strengths [18]. Recently, a hybrid carbon foam has been

prepared where both phenolic resin and pitch are used si-

multaneously using PU as a soft template [19]. Moreover

about 2 wt% Aluminum (Al) powder was added to enhance

the strength of the carbon foam. Due to hybrid nature, the

thermal properties of this foam can be modified using dif-

ferent heat treatment temperatures. Pitches are graphitizable,

which offer higher thermal conductivity when treated at

temperature[2273 K.

The aim of the present investigation is to further explore

the effect of Al powder on the thermal and mechanical

properties of hybrid resin/pitch based carbon foam. Al with

lower melting temperature (*933 K) is expected to react

in situ with the newly forming carbon matrix at the car-

bonization temperature of 1273 K. At lower amount of Al,

there was a significant increase in the thermal and me-

chanical properties. When the amount of Al in the carbon

foam increased to 8 wt%, it showed an abnormal behavior

called ‘‘self-destruction’’ in the ambient environment. The

phenomenon was pictured and probable causes were in-

vestigated in this paper. In recent years, the attempt to

manufacture larger components as a one-piece is increasing

especially in aerospace and aircraft industry. Induced

stresses due to autoclave curing of carbon fiber reinforced

composites play a key role in the performance of com-

posites. The novel carbon foam containing Al has a great

potential for using as a collapsible tooling material since it

can be tailored to be a low cost, durable and lightweight

material [20–23].

2 Experimental procedure

2.1 Raw materials

Semi-rigid PU foam (yellow color, 0.06 g/cm3 density,

0.40 mm average pore size, 8–12 % carbon contents as

determined by TGA) was purchase commercially and

washed with distilled water and dried in air. The phenolic

resin was prepared by the polymerization of industrial-

grade phenol and 37 % analytical grade formaldehyde

(mole ratio 1:1.6) at 369–373 K using a basic catalyst.

Processing details can be found in [24–27]. Pure Al (200

mesh, 99.8 wt% purity) and coal tar pitch (Wuhan China,

1.25 g/cm3 density, 358 K softening point, 398 K melting

point, 57–62 wt% volatiles, C90 wt% carbon contents)

was purchased from Beijing Jinwen Chemical Reagents

Company. Analytical grade isopropanol (purchased from

Sinopharm Chemical Reagent Beijing Co., Ltd) was used

to lower the viscosity of the phenolic resin/powdered pitch/

Al impregnating mixture.

2.2 Carbon foam preparation

Initially, coal tar pitch was crushed, ground and mixed with

Al powder in four different batches. Al to pitch weight ratio

was 10.5–32 corresponding to 2–8 wt% Al in the final cured

PU foam containing all precursors. The four batches were

ball milled in a tungsten carbide jar for about 5 h using a

QM-1SP2 planetary-type ball mill with a planetary rotation

speed of 450 rpm. The objective was to reduce the particle

size and make a close contact of Al powder with the pitch.

This fine powder mixture was added into diluted phenolic

resin and homogenized using a slow speed mixer for 1 h.

Further details of the process can be found in the Ref. [19].

Briefly the main controlling parameters were the ball mil-

ling of Al and pitch powder (-150 mesh), impregnation and

stabilization of PU foam, post curing stabilization at 523 K,

slow pyrolysis up to 1273 K in inert gas under the cover of

coke breeze and finally a slow cooling rate. In a coking

furnace, the samples were placed in a steel container and

coke breeze was poured over them. Excessive amount of

coke breeze, about ten times the volume of the samples, was

used to cover all the samples. At the first stage of the car-

bonization process, a slow heating rate of 10�/h was se-

lected until 773 K. The slow rate minimized the risk of

cracks produced due to evolution of volatiles from the

sample. Moreover, it also promoted the reaction of Al

powder with the nascent carbon and the gaseous products of

pitch/resin pyrolysis like CH4, H2, CO, CO2, and H2O etc.

forming in situ Al compounds. After 773 K, the rate was

increased to 40�/h up to 1273 K and held for 2 h. The coke

breeze, being reusable, served multiple functions other than

producing a reduced environment. It controlled the flow of

volatiles and gases, homogenized temperature distribution,

exerted a slight pressurization on the sample and preserved

heat during the cooling cycle. A slow cooling rate of 30/h

was used until 673 K. After that the furnace was shut shown

to cool naturally. Table 1 shows the various key features of

the novel carbon foam during processing and after final heat

treatment.

2.3 Characterization

The changes in geometric density were measured by the

ratio of mass of sample to the total apparent volume. The

true density was measured using a helium gas displacement

Qunatachrome Pentapycnometer. The true density is the

mass per unit volume of the material, which excludes all

the voids or pores. The porosity and the total pore volume

were calculated using the following expression:

%P ¼ qt � qaqt

� �� 100 ð1Þ

898 J Porous Mater (2015) 22:897–906

123

VT ¼ qt � qaqtqa

� �ð2Þ

Where %P is the bulk porosity, VT is the total pore volume,

qt and qa are the true and apparent densities of the samples

respectively. Thermal conductivity was measured using a

comparative cut bar instrument based on ASTM E1225

Test Method. The compressive strength was measured us-

ing the SANS CMT 5105 (100 KN) mechanical testing

machine. The vertical moving speed of the loading head

was 0.5 mm/min (corresponding to a very low strain rate of

5.5 9 10-4 s-1). The specimens were machined to

ø = 33 mm 9 50 mm size giving a length to diameter

ratio of about 1.5. In order to avoid the dust spread on the

machine, the specimens were placed in a steel pipe fitted

with a plunger. The plunger was attached to the moving

head of the machine. The specimens are shown in Fig. 1.

The morphology of the foam was examined by using a

JEOL (model JSM-6610LV) scanning electron microscope

equipped by an energy dispersive X-ray analyzer. The

phase and composition of samples were identified by X-ray

diffraction (XRD, D8 Advance, Cu-ka radiation, 0.01�wide scanning steps, and 1 s/step acquisition time). The

samples were machined with ø = 25 mm 9 1.0 mm in

size.

2.4 Comparative cut bar (ASTM E1225 test

method)

This is a simple and most widely used comparative method

for testing thermal conductivity in axial direction. The

principle of the measurement is to compare the respective

thermal gradients when steady-state heat flux is passed

through known and unknown samples. This thermal gra-

dient is inversely proportional to their thermal conduc-

tivities. A tailor made apparatus is shown in Fig. 2. The

carbon foam sample with ø = 25 mm 9 20 mm was

sandwiched between two known reference samples.

Stainless steel STS 304 samples with same size as that of

carbon foam sample and thermal conductivity of 19 W/mK

were used as reference. To eliminate heat losses in the

radial direction, rigid polyurethane sleeves with outer and

inner diameters of 150 and 25 mm respectively, were used

to cover all the samples as shown in the inset of Fig. 2. The

heat source was also insulated around the radial direction

using glass fiber reinforced housing. The room temperature

was lowered to 283 K and maintained for 24 h. The tem-

perature at the hot face was raised slowly with two holding

steps of 10 h at 300 and 310 K respectively. A cooling fan

was used on the cold face to dissipate away the heat. At

steady-state condition, the thermal conductivity of the

target sample was calculated using the following Eq. (4)

derived from Fourier’s law of heat conduction.

Q

A¼ KS

DTSL

¼ Kr

DT1 þ DT22L

� �ð3Þ

Where Q/A is the heat flux, L is the distance between two

end, Ks is the thermal conductivity and DTs is the tem-

perature gradient.

KS ¼ Kr

DT1 þ DT22DTS

� �C ð4Þ

Table 1 Key processing parameters and properties of carbon foam

Parameter Values

Weight ratio of carbon precursors

Phenolic resin: Pitch 3:1

Volumetric compression of PU foam (%) 10–15

Weight loss after curing (wt%) 6–8

Volume shrinkage after carbonization (%) 18–22

Total Weight loss after carbonization (wt%) 50–60

Fig. 1 Various carbon foam samples, a for thermal conductivity; b for compression testing and c compression test fixture

J Porous Mater (2015) 22:897–906 899

123

Where Kr is the thermal conductivity of reference sample,

DT1 and DT2 are the thermal gradients across the reference

sample 1 and reference sample 2 respectively. The sizes of

all the three samples were same. C is the correction factor

which depends on the efficiency of the apparatus and

usually within 1.0–1.02.

2.5 Self-destruction at room temperature

Two parts of carbon foam with 8 wt% Al content were

placed on the shelf at room temperature (300 K) for

30 days and pictured the degradation behavior. Cracks

appeared after 15 days and continued for another 14 days.

3 Results and discussion

3.1 General properties

Table 2 shows the general properties of the carbon foam.

Density increased with the rise in Al wt% while the open

porosity and pore volume decreased accordingly. The car-

bon foam without Al showed a very low value of thermal

conductivity. Almost 800 % increase in the thermal con-

ductivity was observed in the sample containing 6 wt% Al.

With further increase in Al contents, a sharp decreasing

trend was noticed and gain at 6 wt% reduced to half.

Powdered Al with particle size distribution (particle di-

ameter of 60–130 lm) was dispersed in the phenolic resin

along with the ground coal tar pitch. The carbon precursors

were pyrolyzed to carbon during high temperature treat-

ment. During this treatment, part of Al was converted into

in situ compounds and well mixed in the body of ligaments.

Higher thermal conductivity of Al and the most probable

conductive channel created inside the ligaments were re-

sponsible for higher values of thermal conductivity. There

was a slight drop in thermal conductivity when the amount

of Al increased to 8 wt%. The key factor of carbon/Al in-

terface affecting phonon coupling between and propagation

across the interfacial phases is aluminum carbide (Al4C3) at

the junction between the aluminum and the carbonaceous

matrix. The formation of Al4C3 lowers the phonon coupling

and propagation between the constituent materials, lower-

ing the thermal conductivity. The Al4C3 phase is brittle,

hydroscopic, and generally known to have a low thermal

conductivity. The formation of significant amount of Al4C3

decreased the thermal conductivity and stability of carbon

foam. The resulting material became very hygroscopic and

unstable which will be further explained in the following

sections. SEM pictures of the carbon foam are shown in

Fig. 3. The open cell structure is seen in all the pictures and

also evident by the porosity calculations shown in Table 2.

In this particular case, the pores are not exactly spherical or

elliptical but are of irregular, uneven and rough shape. Due

to the complexity and randomness in the pore shape, it is

difficult to obtain accurate dimensions and distribution of

pore sizes. It falls between a broad range of 2–200 lm,

although some pores in the image are seen bigger which

were broken during cutting, grinding and sample prepara-

tion. Some bigger voids of 500 lm size are also visible in

all pictures. As the Al contents increased from 2 to 8 wt%,

the pore sizes and void spaces decreased remarkably. The

specimen with 8 wt% Al seems to be more uniform and

dense.

The EDX technique was used to find the composition of

individual particle. Instead of rastering over a relatively large

inspection field, single points were also defined and ana-

lyzed. In Fig. 4d, the resulting spectra shows that, Al metal

along with carbon, nitrogen and oxygen was detected in the

inspection field. X-ray mapping of elements is also shown in

Fig. 4b. Here the position of Al element emitting charac-

teristic X-rays within an inspection field was indicated by

grey color. Al is visible in the whole inspection field with

some areas showing agglomerates into the form of balls. Al

does not wet the carbon, some particles agglomerated in the

Fig. 2 Comparative guarded

axial heat flow system;

a pictorial view with an inset

showing sample fitted in

insulated cavity and b schematic

view

900 J Porous Mater (2015) 22:897–906

123

form of spheres with outer surface either smooth or covered

with some nanoparticles and most probably some oxy car-

bides and nitrides (EDX results). Overall, the pore wall

surface in the carbon foam is rough and uneven and there are

some agglomerates of Al with sizes in the range of 1–25 lm.

The pore wall thickness is also not the same and varies be-

tween 10 and 100 lm. This was due to the multiple im-

pregnation cycles and a slight pressurization during curing of

the resin. The multiple coatings of the resin/pitch mixture

and the carbonization shrinkage transformed the open

reticulated structure of the PU foam into the coarse, rough

and uneven morphology.

3.2 Self-destruction at room temperature

Figure 5 shows the sequential destruction of carbon foam

in 14 days. The destruction was due to non-wetting con-

ditions between the monolithic phases for short contact

Table 2 Effect of Al contents on density, porosity and thermal conductivity

Al powder (wt%) Bulk density

(g/cm3)

True density

(g/cm3)

Porosity (%) Open pore volume

(cm3/g)

Thermal conductivity

(W/mK)

0 0.50 1.50 66.67 1.33 0.043

2 0.55 1.53 64.05 1.16 0.092

4 0.56 1.58 64.56 1.15 0.142

6 0.58 1.61 63.97 1.10 0.385

8 0.60 1.62 62.96 1.05 0.221

Fig. 3 SEM micrographs of carbon foam with Al; a 2 %, b 4 %, c 6 % and d 8 %

J Porous Mater (2015) 22:897–906 901

123

times, and the reactivity of carbon with liquid aluminum

[28, 29]. At 1023 K, the free energy of formation of Al4C3

is -168 kJ/mol [30]. The formation of Al4C3 at the carbon/

Al interface is often observed in carbon fiber reinforced

aluminum matrix composites infiltrated at high tem-

peratures and low cooling rates [31] In this case, the for-

mation of Al4C3 was accompanied by self-destruction of

porous structure and ultimately converting to fine to coarse

powder.

Moreover the destruction started from the inner core of

the foam and proceeded outward resulting in a core–shell

effect. To further investigate this behavior, samples were

taken from the outer and inner part of the carbon foam

before the destruction started. XRD analysis clarified this

effect by showing a vast difference in carbide content from

inner core to outer surface. Figure 6 showed a greater in-

tensity of Al4C3 peaks in the inner core sample than that in

the outer surface. Some new peaks of nitride were also

visible in the sample taken from the outer surface of the

carbon foam. It is also well known that liquid aluminum

does not wet carbon. The peaks of Al4C3 and AlN are also

visible in the sample with 6 wt% Al but with a lower in-

tensity. The surface of aluminum is covered by a thin

protective layer of aluminum oxide. The carbonization

temperature was 1273 K, which was far less than required

for the formation of AlN in the carbon foam. As shown in

the Al–C phase diagram, Al4C3 melts following the peri-

tectic reaction [32];

Al4C3 ! liquid Al - richð Þ þ graphite at 2433K: ð5Þ

Therefore, the peritectic reaction was not the deter-

mining reaction in the carbonization reaction because it

was much lower than the peritectic temperature. Instead, it

can be proposed that the AlN formation reaction was car-

ried out by the substitution reaction, where carbon atoms

were directly substituted by nitrogen atoms to form the

AlN. However the amount substitute by nitrogen atom was

very little. It has been reported that the Al4C3 substantially

worsens mechanical properties in carbon/Al composites but

a small quantity of it can enhance the interface [33, 34].

Therefore, when a lower amount of Al was used in the

manufacturing of carbon foam, the mechanical and thermal

properties improved. According to thermodynamic pre-

diction [33], carbon can react with aluminum to form

Al4C3 at any temperature under 1773 K. Since the Al is

liquid at high temperature, the diffusion of carbon cannot

be ruled out. When the liquid Al is completely surrounded

by carbon, it become saturated by the carbon. The large

carbide crystals precipitates out during slow cooling as

prevailed during this experiment. In the study of reaction in

Al–Ti–C system, the Al–C reaction occurs at 978 K pro-

vided the carbon contents are not limited. This reaction

Fig. 4 Dispersion of Al into

carbon foam; a, c SEM image

showing agglomerated particles,

b X-ray mapping of Al and

d EDX results showing oxy

carbides and nitrides

902 J Porous Mater (2015) 22:897–906

123

occurs before or soon after the melting of Al, provided that

the two reactant particles closely contact each other during

the reaction process. It is well known that the kinetics of

reactions in Al–C is a function of time and temperature.

The contact angle of pure aluminum with graphite at

1100 K decreases rapidly from 160 to 90� in about

100 min. The role of the pitch particles was also to separate

the Al powder and prevent their adjoining. The non-wet-

tability of carbon provides a suitable environment for these

liquid droplets in the suspension to undergo size changes

and adhere to newly formed carbon matrix. The higher

temperature also permits the diffusion of the molten Al

atoms into the carbon and entraps them by gradually

overcoming the restriction of van der Waals attraction from

the carbon hexagonal planes. Here in this case the Al

powder was mixed with pitch in the diluted phenolic resin.

During carbonization, the carbon precursors were con-

verting into the carbon while the Al was also near or at the

melting point (*933 K). These in situ conversions at high

temperature are the most probable cause of formation of

Al4C3. At some later period, when no direct contact exists

between the constituents, diffusion has to occur through the

already formed Al4C3. Among the possible reactions that

take place in the presence of the Al4C3 phase and moisture

are shown in following equations;

Al4C3 þ 12H2O ! 4Al OHð Þ3þ3CH4 ð6Þ

Al4C3 þ 18H2O ! 4Al OHð Þ3þ3CO2 þ 12H2 ð7Þ

Fig. 5 Self-destruction of carbon foam with 8 wt% Al in 14 days after 15 days of processing

J Porous Mater (2015) 22:897–906 903

123

Since the carbon foam was not immersed in water, but

exposed to ambient air, it is expected that the formation of

Al(OH)3 is closely related to the interaction of Al4C3 with

the surrounding moisture. As shown in Eq. (6), the reaction

of Al4C3 with water is very slow forming methane gas at

room temperature [35]. Alternatively, carbon dioxide is

formed in excessive water according to Eq. (7). This in

agreement with observations of bubbles emerging from

polished surfaces in a SiC/6061-MMC [36]. The XRD scan

of the destroyed sample taken from the inner core is shown

in Fig. 7. The peaks of Al(OH)3 are also visible confirming

the reaction shown in Eq. (6). In the current case, reaction

shown in Eq. (6) is most probable cause of destruction

because it was very slow. The net result was a decrease in

the mechanical strength of the carbon foam, manifested as

a gradual degradation of the material by the slow interac-

tion with the atmospheric moisture. It is expected that

carbon, having a smaller atomic mass than Al, is the dif-

fusing species through the Al4C3 layer, and this occurs at a

relatively slow rate. Thus, further growth of Al4C3 occurs

at an even slower rate when the amount of Al is small. This

is the most probable cause of showing no self-destruction

in carbon foam with Al B 6 wt%.

3.3 Compressive strength

Table 3 shows the data and the stress–strain curves

recorded during compression of samples with different

amounts of Al in Fig. 8. It exhibits multi-stage deformation

responses including the initial elastic region with a high

slope, a linear hardening regime, a stress-plateau region

and a graceful failure. The initial high slope corresponds to

the elastic deformation, and the beginning of the non-lin-

earity is the time when crack propagation in the pore walls

happens [37]. It can be observed that the initial high slope

and the collapse stress of the carbon foam increased with

increasing Al wt%. In addition, as clearly observed from

the deformation responses, the Al wt% has a significant

effect on the deformation behavior of the carbon foam. The

carbon foams, prepared at lower Al wt%, displayed a long

Fig. 6 XRD patterns of carbon foam sample with 6 and 8 wt% Al

(sample taken from outer surface and inner core)Fig. 7 XRD patterns of self-destructed carbon foam powder con-

taining 8 wt% Al

Table 3 Average compressive strength and modulus of the carbon

foam with 0–8 wt% Al

Sample ID Ultimate compressive

strength (MPa)

Compressive

modulus (GPa)

0 % Al 16.87 0.48

2 % Al 23.42 0.74

4 % Al 24.77 0.93

6 % Al 32.58 1.19

8 % Al 34.81 1.02

Fig. 8 Compressive stress–strain curves of carbon foams with

0–8 wt% Al

904 J Porous Mater (2015) 22:897–906

123

plateau region around their yield points, suggesting a more

flexible deformation manner. Moreover the first fracture

stress is in the range of 9–12 MPa for the carbon foam with

Al\ 4 wt%.When the carbon foam was prepared at a

higher Al wt%, it exhibited the conventional brittle fracture

with a short plateau region. The fracture stress also in-

creased to 22–23 MPa. The reason for the variation in

deformation behavior is probably attributed to the differ-

ence in the pore structures of the resulting carbon foam.

The carbon foam with 8 wt% Al presented stress–strain

curve characterized by an elastic behavior, a narrow strain

hardening region and a sudden rupture as generally

recorded for dense ceramics. The vertical downward slope

at a strain of about 0.0045, indicate the fast propagation of

micro-cracks. This could be attributed to the aggregation of

Al at high wt%.

After elastic region, the first drop in strength corresponds

to fracture stress. The maximum stress somewhat lies in

between the stress-plateau region and used to calculate the

ultimate compressive strength. Cracks initiate at the peak

stress value and the material tends to generate fragments as a

result of these cracks. The long serrated plateau region along

with strain hardening originates from the coexistence of

collapsed and uncollapsed zones, typical of brittle foam

undergoing successive pore wall fractures. It can be seen

that the compressive strength of the composites was im-

proved with an increase in additive amount of Al from 2 to

8 wt%. The porosity also decreased with increasing Al wt%

resulting in a higher values of strength. When the additive

amount of Al was 8 wt%, the compressive strength after

passing the plateau region decreased more than that of

6 wt% Al carbon foam but after started increasing to a max

value of 34.8 MPa. A possible reason for this could be due

to the densification and stabilization of carbon foam. The

fact indicates that the carbon foam was toughened after the

addition of Al. The reason for this increase is very com-

plicated. One of the explanations could be that the Al was

converted into Al4C3 and to some extent AlN that toughened

the pore walls of carbon foams. Inter diffusion between

aluminum and carbon matrix leads to the formation of an

interlayer that contributes in strengthening the ligaments.

Although excess Al4C3 is known to often lead to the dete-

rioration of mechanical properties, its potential to contribute

to matrix hardening has also been suggested.

It can also be seen in Fig. 8 that the compressive strain

of the carbon foam without Al was 1.19 %, while the

compressive strains of carbon foams with 2, 4, 6, and

8 wt% of Al were 1.01, 0.92, 0.81, and 0.69 %, respec-

tively. Carbon foam with no Al showed a lower compres-

sive strength and a higher strain. With increasing the

additive content, the compressive strength increased and

the strain decreases. Summarily, in order to improve the

compressive strength of carbon foams, the optimum

amount of Al should be around 5 wt%. Higher amount of

Al will un-stabilized the carbon foam after ageing at room

temperature.

4 Conclusions

Carbon foam was prepared from the carbonization of pitch/

phenolic resin slurry supported on a polyurethane foam tem-

plate. 2–8 wt% Al was used to improve the thermal and me-

chanical properties. The properties and structural examination

showed a density of 0.50–0.60 g/cm3, open porosity of

63–68 %, thermal conductivity of 0.043–0.385 WmK, com-

pressive strength of 17–35 MPa, pore size distribution of

2–200 lm and pore wall thickness of 10–100 lm. Carbon

foam with 8 wt% Al showed self-destruction in 14 days after

15 days of manufacturing. The destruction started from core

and proceeded towards outer surface resulting in a core–shell

effect. XRD analysis confirmed the destruction due to the

formation of Al4C3 and reaction with atmospheric moisture

formingAl(OH)3. On the outer surface, a little amount of AlN

was detected due to substitution reaction of nitrogen with

carbon. One potential application of this self-destruction is in

the field of composite tooling. The novel carbon foam con-

taining controlled amount of Al, carbonization temperature

and cooling rate can beused to fabricate a collapsiblemandrel,

which will lose strength after certain time interval facilitating

in easy and self-removal.

Conflict of interest The authors declare that they have no conflict

of interest.

References

1. T. Suzuki, H. Umehara, Pitch-based carbon fiber microstructure

and texture and compatibility with aluminum coated using che-

mical vapor deposition. Carbon 37, 47–59 (1999)

2. J.F. Silvain, C. Vincent, J.M. Heintz, N. Chandra, Novel pro-

cessing and characterization of Cu/CNF nanocomposite for high

thermal conductivity applications. Compos. Sci. Technol. 69,2474–2484 (2009)

3. Y. Chen, B. Chen, X. Shi, H. Xu, Y. Hu, Y. Yuan, Preparation of

pitch-based carbon foam using polyurethane foam template.

Carbon 45, 2132–2134 (2007)

4. N.C. Gallego, J.W. Klett, Carbon foams for thermal management.

Carbon 41, 1461–1466 (2003)

5. R. Mehta, D.P. Anderson, J.W. Hager, Graphitic open-celled

carbon foams: processing and characterization. Carbon 41,2174–2176 (2003)

6. J. Googin, J. Napier, M. Scrivner, Method for manufacturing

foam carbon products, US Patent 3,345,440 (1967)

7. F.C. Cowlard, J.C. Lewis, Vitreous carbon: as new form of car-

bon. J. Mater. Sci. 2, 507–512 (1967)

8. R. D. Klett, High temperature insulating carbonaceous material,

US Patent 3,914,392 (1975)

9. R.W. Pekala, R.W. Hopper, Low-density microcellular carbon

foams. J. Mater. Sci. 22, 1840–1844 (1987)

J Porous Mater (2015) 22:897–906 905

123

10. J. Lee, K. Sohn, T. Hyeon, Fabrication of novel mesocellular

carbon foams with uniform ultralarge mesopores. J. Am. Chem.

Soc. 123, 5146–5147 (2001)

11. J. Klett, High thermal conductivity mesophase pitched-derived

carbon foam, in Proceedings of the 1998 43rd International

SAMPE Symposium and Exhibition, Part 1 (of 2) (Anaheim, CA

31 May–4 June, 1998)

12. C. Chen, E.B. Kennel, A.H. Stiller, P.G. Stansberry, J.W. Zondlo,

Carbon foam derived from various precursors. Carbon 44,1535–1543 (2006)

13. X.W. Wu, Y.G. Liu, M.H. Fang, L.F. Mei, B.C. Luo, Preparation

and characterization of carbon foams derived from alumi-

nosilicate and phenolic resin. Carbon 49, 1782–1786 (2011)

14. H.F. Xu, H.J. Zhang, Y.D. Huang, Y. Wang, Porous carbon/silica

composite monoliths derived from resorcinol–formaldehyde/

TEOS. J. Non-Cryst. Solids 356, 971–976 (2010)

15. Y.R. Liu, B.P. Lin, D. Li, X.Q. Zhang, Y. Sun, H. Yang, Mag-

netically-separable hierarchically porous carbon monoliths with

partially graphitized structures as excellent adsorbents for dyes.

J. Porous Mat. 21, 933–938 (2014)

16. S.W. Lei, Q.G. Guo, J.L. Shi, L. Liu, Preparation of phenolic-

based carbon foam with controllable pore structure and high

compressive strength. Carbon 48, 2644–2646 (2010)

17. Y.H. Yan, X.M. Shi, J.G. Liu, T. Zhao, Y.Z. Yu, Thermosetting

resin system based on novolak and bismaleimide for resin-

transfer molding. J. Appl. Polym. Sci. 83, 1651–1657 (2002)

18. M. Calvo, R. Garcıa, A. Arenillas, I. Suarez, S.R. Moinelo,

Carbon foams from coals: a preliminary study. Fuel 84,2184–2189 (2005)

19. S. Farhan, R.M. Wang, H. Jiang, N. Ul-Haq, Preparation and

characterization of carbon foam derived from pitch and phenolic

resin using a soft templating method. J. Anal. Appl. Pyrolysis

110, 229–234 (2014)

20. A.M. Druma, M.K. Alam, C. Druma, Analysis of thermal con-

duction in carbon foams. Int. J. Therm. Sci. 43, 689–695 (2004)

21. S.R. White, Y.K. Kim, Process-induced residual stress analysis of

AS4/3501-6 composite material. Mech. Compos. Mater. Struct. 5,153–186 (1998)

22. F.C. Campbell, Manufacturing processes for advanced compos-

ites (Elsevier Advanced Technology, Amsterdam, 2004)

23. A.M. Druma, M.K. Alam, M.S. Anghelescu, C. Druma, Three

dimensional modeling of carbon foams, in ASME international

mechanical engineering congress and exposition 2005 (Orlando,

FL, 5–11 November, 2005)

24. A. Gardziella, L.A. Pilato, A. Knop, Phenolic resins: chemistry,

application, standardization, safety and ecology, 2nd edn.

(Springer, Berlin, 2000)

25. P. Byung-Dae, R. Bernard, Y.S. Kim, W.T. So, Effect of syn-

thesis parameters on thermal behavior of phenol-formaldehyde

resol resin. J. Appl. Polym. Sci. 83, 1415–1424 (2002)

26. J. Wang, H. Jiang, N. Jiang, Study on the pyrolysis of phenol-

formaldehyde (PF) resin and modified PF resin. Thermochim.

Acta 496, 136–142 (2009)

27. L.B. Manfredi, O. De La Osa, N. Galego Fernandez, A. Vazquez,

Structure-properties relationship for resols with different

formaldehyde/phenol molar ratio. Polymer 40, 3867–3875 (1999)

28. J.T. Blucher, J. Dobranszky, U. Narusawa, Aluminium double

composite structures reinforced with composite wires. Mater. Sci.

Eng., A 387–389, 867–872 (2004)

29. E. Pippel, J. Woltersdorf, M. Doktor, J. Blucher, H.P. Degischer,

Interlayer structure of carbon fibre reinforced aluminium wires.

J. Mater. Sci. 35, 2279–2289 (2000)

30. R.Y. Lin, Interface evolution in aluminum matrix composites

during fabrication. Key Eng. Mater. 104–107, 507–522 (1995)

31. M.H. Vidal-Setif, M. Lancin, C. Marhic, R. Valle, J.L. Raviart,

J.C. Daux, M. Rabinovitch, On the role of brittle interfacial

phases on the mechanical properties of carbon fibre reinforced

Al-based matrix composites. Mater. Sci. Eng., A 272, 321–333(1999)

32. C. Qiu, R. Metselaar, Solubility of carbon in liquid Al and sta-

bility of Al4C3. J. Alloys Compd. 216, 55–60 (1994)

33. S.H. Li, C.G. Chao, Effects of carbon fiber/Al interface on me-

chanical properties of carbon-fiber-reinforced aluminum-matrix

composites. Metall. Mater. Trans. A 35, 2153–2160 (2004)

34. H. Kwon, H. Kurita, M. Leparoux, A. Kawasaki, Carbon nano-

fiber reinforced aluminum matrix composite fabricated by com-

bined process of spark plasma sintering and hot extrusion.

J. Nanosci. Nanotechnol. 11, 4119–4126 (2011)

35. T.Y. Kosolapova, Carbides properties production and applica-

tions, 1st edn. (Plenum Press, New York, 1971)

36. J.K. Park, J.P. Lucas, Moisture effect on SiCP/6061 Al MMC:

dissolution of interfacial Al4C3. Scr. Mater. 37, 511–516 (1997)

37. S. Meille, M. Lombardi, J. Chevalier, L. Montanaro, Mechanical

properties of porous ceramics in compression: on the transition

between elastic, brittle, and cellular behavior. J. Europ. Ceram.

Soc. 32, 3959–3967 (2012)

906 J Porous Mater (2015) 22:897–906

123