A simple preparation method and characterization of epoxy reinforced microporous phenolic open-cell sound absorbent foam

Abstract

A series of micro-porous phenolic open-cell sound absorbent foams reinforced by epoxy resin were fabricated by a physical foaming method. Compound emulsifiers consisting of anionic and non-ionic surfactants were physically blended at high speed with modified resols, foaming agent and a mixed acid curing agent. The viscosity, surface tension, and gel permeation chromatography (GPC) properties were characterized. The fabricated foams were characterized with regard to their pore size distribution, water absorption, sound absorption, and mechanical and thermal properties. Surface tension and porosity analysis demonstrated that the anionic surfactant, sodium dodecyl sulfonate (SDS), promoted the formation of homogeneous micro-pores. Additionally the open cell porosity reached up to 90%. The open-celled pore structure and cell size distribution revealed by scanning electron microscopy (SEM) demonstrated that more homogeneous and smaller open cells existed with an increasing dosage of epoxy. The sound absorption properties were also increased significantly by adding the epoxy resin, even reaching about 95%. The micro-pore structure and cell distribution controlled by the content of epoxy resin and the ratio of SDS/Tween-80 greatly improved the mechanical strength of the phenol–formaldehyde resin (PF).

---

Introduction

Open-celled microcellular foam polymer materials, because of their unique interconnected micro-pore structure, possess a wide range of properties well suited for many micro-porous material applications.¹ The open-celled structure allows particles, fluids, sound, and gas to flow through the material, resulting in a product that can be used as a filter or membrane, catalyst supports, drug delivery materials and so on.^1,2

Currently, studies on porous functional polymer materials are gradually becoming a research hotspot. Up to now, there exist four main ways to prepare open cell plastics and foams: thermally induced phase separation, use of a supercritical fluid, monomer polymerization and precipitation with a compressed fluid anti-solvent.^3,4 Besides, previous researchers have also developed new techniques (such as block copolymer and polymer nanometer-composite foaming methods) to obtain polymers with open cell structures.^5,6 In addition, over the past few decades, the numerous investigations into PF have mainly focussed on closed cell insulation materials. There are few studies about open cell phenolic foam which is produced from resols at medium-low temperature under atmospheric pressure, using a simple physical foaming method. Indeed, merely two methods have been reported for the preparation of open cell phenolic foams so far: supercritical fluid foaming from novolacs⁷ and open cell carbon foams prepared using phenolic foam as a precursor.⁸ Therefore, it is necessary to invent new practical and simple methods to manufacture open cell polymers with excellent performances.

As for epoxy resin, it also possesses many excellent advantages of low shrinkage, chemical resistance and superior mechanical properties.^9–11 Furthermore, it is acknowledged that the epoxy group can react with the hydroxyl groups in resol via etherification.¹⁰ Given the excellent mechanical performance of epoxy resin, and the reaction mechanism, epoxy resin of adaptive viscosity is taken into consideration as a reinforced functional agent, because it is the fine open-celled microcellular structure of phenolic foam combined with the performance of epoxy that can make the composite foam possess many mutual excellent properties including low toxicity, smoke and flame retardancy, high strength and dimensional stability, and chemical resistance. The prepared foams with superior compressive strength and flexural properties possess a considerably porous structure, good properties of filtration and water permeability, and low production cost. These combined performances and low production cost of open cell phenolic foam make it very versatile in some applications, such as: sand control in oilfields, chemical filtration and separation, sewage treatment, microfiltration membranes, adsorption materials and so on. Thus, the prospective applications of open cell phenolic foams are promising. In addition, in the last several decades, the use and variety of available specialized sound-absorbing materials has increased greatly.¹² There are many studies on acoustic insulation materials, such as polyurethane and polyethylene terephthalate foams, and a lot of sound-absorbing foam is green and recyclable.^13–15

In our study, we developed a new simple physical foaming approach to prepare an epoxy modified micro-porous phenolic open-cell foam with resols, differing from previous modifications of the direct physical blending foaming method. It should be noted that we innovated the use of different anionic and nonionic compound surfactants¹⁶ as our emulsifiers, and a mixed acid curing catalyst. Hence phenol–formaldehyde resol modified by epoxy resin was synthesized, and the reaction between epoxy and resol was catalyzed with a suitable sodium hydroxide solution. In addition, we also mixed epoxy resin with pure resol directly, followed by foaming with our method for comparison, only to find that it failed to foam with an open porous structure. A series of pre-polymers with different dosages of epoxy resin were synthesized. And the viscosity of the modified resols was measured at different shear rates of 0.6 rpm, 1.5 rpm, 3.0 rpm and 6.0 rpm. The influence of different ionic surfactants on the emulsification effect was characterized by surface tension tests. Gel time measurements were carried out to assess the curing reaction activity and thermal behavior of the resol resins. Gel permeation chromatography (GPC) was used in the determination of average molar masses and the polydispersity index, as a function of the degree of condensation of the resin. Finally, open-celled microcellular foams with various dosages of epoxy were prepared by a simple physical foaming method. The bulk density, solid content of the resins and thermal conductivity were measured. The open cell porosity and water absorption were both tested. The microstructure of the cells and the physical phase distribution of modified resol were observed by SEM. Moreover, the compressive and flexural properties of open-celled microcellular phenolic foam were characterized.

Experimental

Materials

Phenol (≥99%), paraformaldehyde (≥95%) and sodium hydroxide (analytical grade) were provided by Chengdu Kelong Chemical Regents Factory, China. Hydrochloric acid, phosphoric acid, toluene-p-sulfonic acid, acetic acid, Tween-80 (polyoxyethylene sorbitan monooleate), sodium dodecyl benzene sulfonate (SDBS), sodium dodecyl sulfate (SDS′), and sodium dodecyl sulfonate (SDS) (all were analytical grade) were also purchased from Chengdu Kelong Chemical Regents Factory, China. Cyclopentane (≥95%) was provided by Chengdu Aike Chemical Regents Factory, China. Diglycidyl ether of bisphenol-A (DGEBA) type epoxy resin (trademark: E-51:WSR618) was purchased from Nantong Xingchen Synthetic Material Co., Ltd (epoxy value = 0.48–0.54 eq. per 100 g; viscosity (40 °C) ≤ 2500 mPa s).

Synthesis of the modified resol

In this study, the modified resols were condensed in a 250 mL flask equipped with a stirrer, a external cooling condenser and internal heating units. Epoxy modified resol in this study was prepared by the polymerization of phenol and formaldehyde (in a mole ratio of R = [F]/[P] = 1.8) with a given amount of modifier epoxy resin at 80 °C for 3 h, using sodium hydroxide solution as a catalyst.^17,18 The detailed synthesis procedure was as follows: firstly, a given amount of molten phenol and 20% sodium hydroxide solution (0.123 mol) were added into a three-necked flask and stirred for 15 min. Then, a certain amount of epoxy was added into the mixed solution with stirring. When the temperature reached 80 °C, the paraformaldehyde, divided into four equivalent parts, was added every 10 minutes. The process lasted for 3 hours and the resol-type pre-polymer was obtained.^17–19 The mechanism of chemical modification of epoxy-modified phenol and resol is shown in Scheme 1.

Scheme 1 The synthesis route and mechanism of epoxy-modified phenol and resol.

Preparation of high strength micro-porous phenolic open-cell foam

The choice of surfactants, foaming agents and curing agent directly affects the resulting open-cell phenolic foam. It is particularly important to select a suitable surfactant, foaming agent and curing agent. Anionic surfactants are conducive to the formation of an open-cell structure, while non-ionic surfactants are conducive to the formation of a closed cell structure. Two surfactants used as a complex mixture are effective to form phenolic foam with an open-cell structure of high porosity. The selection of the foaming agent is important in the course of microporous phenolic open-cell foam formation, because you must give full consideration to its boiling point and its resin matrix compatibility. Liquid foaming agent dispersion degree, vaporization rate, and speed of escape from the resin are very important for the formation of an open cell structure. After comparison of dichloromethane, petroleum ether, cyclopentane and other types of foaming agents, it was found that the choice of cyclopentane with a boiling point of 49.5 °C produced a good open cell structure and excellent performance foam. So at last cyclopentane was finalized as the foaming agent. The curing reaction between active methylol groups for crosslinking of the resole resin needs acid as a catalyst. Acid can accelerate the condensation reaction of the resin molecules, and the exothermic heat of reaction prompts a sharp foaming agent vaporization, and gas overflow of the resin droplets and forming foaming resin, and the resin is cured. The catalyst is also a resin curing agent. Selection of a curing agent, and the amount of curing agent used, should follow the principle of ensuring the resin curing and foaming rates match, and the curing reaction proceeds at a relatively low temperature. When a strong acid acts as a curing agent, the curing crosslinking rate is faster than the rate of foaming, and the acid is highly corrosive of the mold. Using a strong acid and a medium-strong acid as a mixed curing agent not only can reduce the curing rate to match the foaming rate, but also reduce mold corrosion. By comparison of hydrochloric acid, phosphoric acid, methyl sulfonic acid, etc. as different curing agents, it was found that by mixing hydrochloric acid and phosphoric acid, the foaming speed and curing speed can be controlled in the appropriate range, which can achieve a better performance of the microporous phenolic open-cell foam.

Micro-porous phenolic open-cell foam samples were fabricated using our proprietary technology. The foaming formulation was typically composed of epoxy modified resol (100 Phr), a certain ratio of the anionic (SDS, SDBS, SDS′) and nonionic (Tween-80) surfactants as compound emulsifiers (4 Phr), mixed acid catalyst (10 Phr) and an appropriate amount of cyclopentane. Under our preferred foaming conditions, open cell phenolic foam was prepared by homogeneously blending modified phenolic resol, the compound emulsifier, and the foaming agent cyclopentane successively with a high-speed mechanical mixer. Next the mixture was mixed with the compound acid and stirred quickly for 35 s. Finally the obtained viscous mixture was poured into a preheated foaming mold immediately and cured at 70 °C for about 30 minutes, followed by foaming. Samples were cut precisely and used for mechanical testing.

Characterization and measurements

Viscosity tests on the epoxy modified resols were carried out with a NDJ-8S digital viscometer (Shanghai Jinhai Instrument Co. China) at 25 ± 0.1 °C, using the 2nd rotor, at different shear rates.²⁰

Surface tension was investigated with a surface tension instrument (Krüss K100, Germany) at 25 ± 0.1 °C. Resol/anionic and nonionic surfactant mixed solutions of a certain equal concentration were also prepared for testing.

The resin was characterized by GPC (HLC-8320 GPC gel permeation chromatography, Tosoh Corporation, Japan) with a series chromatographic column (TSK ge super HZM-M 6.0 × 150 mm and TSK gel SuperHZ3000 6.0 × 150 mm). The resins were dissolved in filtered tetrahydro-furan (THF; 20 mg mL⁻¹), which was also used as an eluent. The sample size was 10 μL, and the eluent rate was 0.6 mL min⁻¹. The measurements were made at 40 °C. The retention times were recorded. A professional software was used to calculate the number-average molecular weight (M_n) and weight-average molecular weight (M_w), as well as the polydispersity.

The micro-porous morphologies of the samples prepared were characterized with an INSPECTF scanning electron microscope (FEI Holand). Pore size and distribution of these samples were observed from the section surfaces of foam specimens. Samples were carefully cut from the freshly peeled flat surfaces using a razor blade. Gold sputtering onto the sample surface was used to impart electrical conductivity. Also the dispersion of the physical phase in the resin matrix of modified resol and pure resol mixed with equal content of epoxy were observed by an INSPECTF scanning electron microscope (FEI Holand). Resol samples were fully dried under vacuum. The distribution of the phase in resol was also observed from the section surface. The statistical analysis of cell size and distribution is conducted through Image Pro Plus 6.0 software.^21,22

The thermal conductivity of these samples was measured by using a thermal constant analyzer (Hot-Disk 2500-OT) at 293 K. The size of all the samples was about 30 mm × 30 mm × 20 mm.

The bulk density was measured according to ISO 845:2006. Samples were prepared in 50 mm squares of 5 mm thick. The result was calculated as follows:

ρ = m/v

where ρ is the apparent density (kg m⁻³), m is the mass of sample (kg), and v is the volume of the sample (m³).

The water absorption of the open-celled phenolic foams was investigated in accordance with ISO 2896:2001. The porosity of the open cell PF was also tested with an immersion method according to GB10799-89. Test specimens were used with dimensions of 50 × 50 × 20 mm³. The specimens were soaked completely in distilled water for 24 h, and the surface water of the foam was wiped dry with absorbent paper, before monitoring of the weight changes of the samples in air and water to measure their water absorption and porosity. Parallel tests were made at least three times. Determination of the water absorption and porosity of the foams was done using the following formulas:²³

Water absorption (%) = [(m₂ − m₁)/m₁] × 100%

where m₁ is the mass of dry phenolic foam, and m₂ is the mass of soaked phenolic foam after 24 h, and

Porosity = [(G′₂ − G′₁)ρ_L]/[(G′₂ − G₃ + G₄)ρ_me]

where G′₁ is the weight of the soaked saturated phenolic foam in the air, G′₂ is the weight of the soaked phenolic foam after 24 h, G₃ is the total weight of the soaked phenolic foam, trays and weight after 24 h in the water, G₄ is the weight of the trays and weight in the water, ρ_L is the density of water, and ρ_me is the density of the saturated medium.

The water permeability test was done to measure the volume of permeated water to evaluate its permeable effect. The weight of the sample before and after was measured on a balance. The thickness of the open cell phenolic foam was 20 mm. Samples were laid on top of a 50 mL beaker for 15 min after water was dropped on the upper of the foams.

The sound absorption properties of the samples were characterized with an SW466-type wave tube provided by the Sound Power Technology Co., Ltd Beijing. Foam samples were cut into cylinders with a diameter of 78 mm and a thickness of about 20 mm. The test frequency range was from 400 to 2500 Hz.

The compressive tests and bending properties were carried out according to ISO 844:244 and ISO 1209-2:2004, respectively, using an (AGS-J 10KN) Universal Testing Machine under ambient conditions. At least five specimens of each material were tested to obtain average values. The foam was compressed at a controlled rate of compression by 10% of the initial thickness at a time, until it was compressed to 85% of the initial thickness. The bending span was 100 ± 1 mm. The density of the foams was calculated according to the dimensions of the test species.²⁴

Results and discussion

Viscosity analysis

Fig. 1 presents the relationship between the viscosity of the pre-polymer resols and the content of epoxy modifier, measured at a shear rate of 3.0 rpm. With the dosage of epoxy increasing, the viscosity of the pre-polymer decreased at a content of 5 wt% and then gradually increased from 5 wt% to 15 wt%.

Fig. 1 Viscosity of the pre-polymer measured at a 3.0 rpm shear rate with different epoxy content.

There are two main factors affecting the viscosity: molecular weight and molecular weight distribution. The relative molecular weight of epoxy resin is higher than that of the pre-polymer. So the introduction of the higher molecular weight epoxy resin will lead to an increase of viscosity. With the dosage of epoxy increased, the molecular weight distribution also changed, leading to a change in the viscosity of the resin system. Furthermore, the increase of viscosity in a certain range is also beneficial for inhibiting the diffusion of the surfactant molecules, which causes an increase of the surface elastic effects, resulting in an improvement in the stability of the bubbles. What is more, there are still other factors accounting for the decreasing trend of viscosity at an epoxy content of 0–5 wt%. The graft structure side chain, which is a result of the reaction occurring between the hydroxyl group, methylol and epoxy group via etherification, can make the mean square radius of gyration larger and make the molecular motion space larger, resulting in the decrease of viscosity of the modified resol. Additionally, an appropriate amount of modifier epoxy may also break the intermolecular hydrogen bonds between the polymer molecules, leading to a reduction in viscosity too. The reduction in viscosity indicated that the addition of a suitable dosage of modifier epoxy resin could reduce the viscosity of the pre-polymers effectively, affording an optimal foaming viscosity for preparing open-celled porous phenolic foam compared to pure resol. Once further increased, excessive unreacted epoxy in the matrix resin increased systemic viscosity.

In addition, different added amounts of modifier epoxy showed that the residual phenol and formaldehyde of resol were not the same, too. The results presented in Fig. 2 show that the more epoxy that was added, the less residual phenol and formaldehyde was obtained. Besides, we also conducted a contrast trail (10% content of epoxy mixed with pure resol directly). Its residual phenol was 2.60%, which was not the same as with 10 wt% epoxy modified resol. Also when the content of epoxy reached 10 wt%, residual formaldehyde was at its lowest level. Compared to pure resol, residual phenol and formaldehyde of modified resol reduced in various degrees. It is also indirectly demonstrated that phenol or methylol of resol reacted with the epoxy group. This also proved that epoxy modified resol indeed decreased the content of free phenol and residual formaldehyde.

Fig. 2 The amount of residual formaldehyde and methylol index in modified resols of various epoxy content.

Surface tension analysis

There is a particularly complex procedure consisting of nucleation, growth, stability and solidification of bubbles during the foaming progress. In the foaming process bubbles have gone through four different stages: bubble nucleation, bubble growth, bubble collision, and bubble combination. The effect of SDS as an anionic surfactant is to promote post-foaming bubble collision and merging. In fact, during the later stages of the foaming progress, the anionic surfactant (SDS, an electrolyte) might break the resin droplets encapsulating cyclopentane and act as a demulsifier to make vaporizing cyclopentane escape outward gently from the resin droplets. An SDS increase can significantly improve the porosity and the number of micropores. Moreover, the series of changes are closely associated with the surface properties of the phenolic resin solution and the emulsification of surfactants. Therefore, it is necessary to investigate the surface tension of resol emulsified with different percentages of compound surfactants to reveal the impact of surfactants on the surface properties of the resin solution. The surface tension curve of resol emulsified with different SDS content is shown in Fig. 3. The surface tension of resol decreases first with a content of SDS of between 0% and 45%, and then gradually increases with the increasing dosage of SDS. In the range of 45–55%, the surface tension increases slowly. However, the surface tension increases sharply when the content of SDS exceeds 55%. This phenomenon can be explained by the following formula:

ΔF = vΔA

where ΔF is the change in the free-energy of the system, ΔA is the change in the total interfacial area of bubbles in the system, and v is the surface tension of the system.

Fig. 3 The influence of SDS content on surface tension.

According to the above formula, when surface tension v decreases, it is bound to increase the surface area of the system ΔA to make the free energy of the system maintain a constant value. And once ΔA increased, large numbers of bubbles could be formed and grow rapidly. At the same time, the reduction of systemic surface tension can also prevent coalescence of the generated bubbles and stabilize these bubbles. When the dosage of SDS exceeds 55%, bubbles may start to collide and connect, and when the content of SDS exceeds 67.5%, excessive SDS is insufficiently dissolved in the resin for emulsification, which affects the stability of the system. Therefore, we could assume that initially, SDS played a role in making the bubbles combined and interconnected in the resin system, until the amount of SDS increased to a certain extent (67.5%), and the function of SDS changed. At this point, it had some negative impacts on the stability of Tween-80 in the system. At the same time, Table 1 also records the influence of different anionic emulsifiers on surface tension; under the equal conditions of anionic emulsifiers/modified resol, the surface tension of resol emulsified with SDS remained the lowest. Compared to other anionic emulsifiers, SDS could be more compatible with resol to make the surface tension of the system decrease so as to stabilize the bubbles. That is the reason why we chose it as the anionic emulsifier. Besides, judging from Table 1, although modification with epoxy resin improved the surface tension of the system a little, once the content of epoxy reached 10%, its surface tension became lower than other percentages of epoxy. This is ascribed to the different increase of the systematic viscosity.

Table 1 Surface tension of different anionic emulsifiers^a

Samples	60% SDBS/40% Tween-80	60% SDS/40% Tween-80	60% SDS′/40% Tween-80
a SDBS-sodium dodecyl benzene sulfonate; SDS′-sodium dodecyl sulfate; SDS-sodium dodecyl sulfonate.
Surface tension (mN m^−1)	30.47	29.60	30.71

---

Samples	Pure resol	5% epoxy-resol	10% epoxy-resol	15% epoxy-resol
Surface tension (mN m^−1)	24.59	30.81	29.47	30.70

---

Molar mass distributions and molecular weight

The effect of the degree of condensation of the resol solution on the number-average molecular weight (M_n) and weight-average molecular weight (M_w), and the polydispersity index (PD) of the resin was determined using GPC. From Fig. 4 and Table 2, the GPC curve of 10% content epoxy modified resol was slightly different from that of pure resol. The proportion of large molecules was higher. Both the M_n and M_w values of modified resol (2) were slightly higher than those of pure resol (1), and even the polydispersity value of (2) was lower than that of pure resol, indicating that modified resol exhibited a narrow molecular weight distribution. Also with regard to the viscosity curve, the viscosity of modified resol indeed increased as the masses (M_n and M_w) increased. Predictably, the M_n and M_w values of the resin indicate polymer condensation reactions and the formation of methylene bridge structures between the phenolic derivatives, thus confirming the results of analyses previously described.

Fig. 4 GPC curves of the resins: (1) the pure resol; (2) the epoxy-modified resin.

Table 2 Average molar mass (M_n and M_w), and polydispersity index (PD) values of pure resol and 10% content of epoxy E51 modified resol

Sample	Number-average molecular weight Mn (g mol^−1)	Weight-average molecular weight Mw (g mol^−1)	Polydispersity index, PD
Pure resol	254	405	1.594
Epoxy-modified resol	307	432	1.408

---

Cell morphology and distribution in open cell phenolic foam

The cellular structure of the foam samples was studied by SEM. Fig. 5 shows the cell morphology of the foam samples emulsified with 67.5% SDS and 32.5% Tween-80 in four parts, with an increasing amount of modifier epoxy resin from 0 wt% to 15 wt% from top left to bottom right. The cell size and the effect of the addition of epoxy on the microstructure of open cell phenolic foams are examined. As is shown in Fig. 5, with the increasing addition of epoxy, the cell size of the phenolic foams generally decreases; the distribution of pores turns uniform and concentrated. Besides, the micro-pores in foams are interconnected with each other, and the wall structure of the pores also exhibits some interconnected smaller pores. By comparison, such micro-porous structure of phenolic foams showed a great contrast with that of pure open cell phenolic foam (the pore structure was at the level of low open porosity and inhomogeneous). However, with the increase of dosage of epoxy, the pore size turns smaller, and the distribution of pores turns more concentrated and homogeneous.

Fig. 5 SEM images (100 μm) of open-celled phenolic foams with different epoxy content (1: 0 wt%, 2: 5 wt%, 3: 10 wt%, 4: 15 wt%).

Fig. 6 is a pore size distribution of phenolic foams with different epoxy content. It is obvious that the pore size distributions of the four open-celled phenolic foams are substantially similar, with the aperture concentrated in an area of 200 μm or less. The number of pores also rapidly increases accompanied by the reduction in pore size, reaching a maximum at a size of less than 50 μm. In addition, the number of phenolic foam pores significantly increased as the dosage of epoxy reached 15 wt%, which is consistent with the SEM image.

Fig. 6 The cell distribution of open-celled phenolic foams with different epoxy content (0 wt%, 5 wt%, 10 wt%, and 15 wt%).

When the content of epoxy is 10 wt%, a large number of open cells with moderate pore size exist, showing a more uniform distribution. The smallest sized cells with an inhomogeneous pore distribution exist in the foams with a dosage of epoxy of up to 15 wt%. When the content of epoxy reaches 5 wt%, not only do the micro-pores of the interconnected structure differ in size, but also the distribution is relatively un-concentrated compared to that of 10 wt% epoxy modified foam. As for pure open cell PF, open cells exist less, with an inhomogeneous pore distribution. However, compared to pure PF, the micro-pores in modified foams are still more numerous and relatively concentrated. This is because the anionic surfactant, sodium dodecyl sulfonate (SDS), which was more compatible with resol, could promote the formation of open cells and produce good homogeneity and moderate micro-pores. There are two factors leading to the existence of various pore sizes and distributions. Firstly, the increasing viscosity of the pre-polymer contributes a lot to the effect of motion of the molecules and the combination of bubbles. High viscosity makes the motion of the molecules difficult and also delays coalescence of the generated bubbles, which means a decrease of the cure rate and an imbalance between the curing and foaming rates, resulting in an impact on the combination rate of adjacent cells during the foaming process. Second, because of the steric hindrance, the difference in the curing rate between molecules with grafted side chains and those without the side chains causes the difference in cell size and distribution too. Therefore, the results above demonstrate that with increase of epoxy in the range of 0 wt% and 10 wt%, the cell size turns more moderate and homogeneous, which is attributed to the increasing combination of cells. Besides, when the dosage of epoxy is up to 15 wt%, excessive epoxy caused some effect on the uniformity of cells.

The physics of phase equilibrium in resols were investigated to verify the successful modification with epoxy. As is observed from Fig. 7, chemical modification of resol resulted in a more uniform and single phase, with no dispersed phase. However, under the same conditions pure resol, through physical blending with equal content of epoxy directly, presented a disordered dispersed phase, because unreacted epoxy resin droplets were distributed in the resin matrix randomly, causing emerging separated multi-phases with an irregular dispersion. From phase analysis, we reached the conclusion that modified resol could demonstrate a more uniform single phase. The comparison of phases also proved the successful modification with epoxy.

Fig. 7 SEM micrograph of modified resol with 10% content of epoxy (1); pure resol physically mixed with 10% content of epoxy directly (2).

Thermal conductivity

The effects of epoxy content on the thermal conductivity of open cell phenolic foams are shown in Fig. 8. The interconnected open-celled pore structure reflected in Fig. 5 provided effective thermal conduction channels. The more open-celled pores and higher open cell porosity, the larger the thermal conductivity would be. In the case of thermal conduction and water absorption in particular, cell geometry, cell size, the shape of the openings between cells and cell wall thickness are some of the most influential structural characteristics that control the maximum amount of water absorption and the thermal conductivity.

Fig. 8 The thermal conductivity curves of open-celled phenolic foams emulsified equally with 65% SDS and 35% Tween-80 in 4 parts and modified with different epoxy content.

Judging from Fig. 8, with the dosage of epoxy increasing, the thermal conductivity first increases up to 5 wt% content of epoxy and then decreases from 5–15 wt%. However, the thermal conductivity of epoxy modified PF is still higher than that of pure PF. From Fig. 8, it can be clearly seen that open cell phenolic foam modified with 5 wt% content of epoxy had the highest thermal conductivity, and the highest open cell porosity could reach up to 91.209%. Open-celled phenolic foams also had higher thermal conductivity than closed cell phenolic foam (0.02 W m⁻¹ K⁻¹ in thermal conductivity and 2 in water absorption). By experiments, we found that SDS could be more compatible with resol compared to other anionic surfactants. The formation of open pores in PF could be ascribed to the fact that anionic surfactants promoted the formation of an open-celled porous structure in the foam. In fact, during later bubble collision and combination progress, the anionic surfactant (SDS, an electrolyte) might break the resin droplets encapsulating cyclopentane and act as a demulsifier to make vaporizing cyclopentane escape outward gently from the resin droplets, resulting in an inter-connected open-celled pore structure.

Water absorption, porosity and water permeability of phenolic foam

For the sake of verifying the open porous structure of PF, a convincing experiment was carried out by testing whether it is pervious. We also used a water absorption test to certify its open cell structure. From Fig. 9, it can be clearly seen that open cell phenolic foam emulsified with an anionic surfactant (SDS) had a higher water absorption than that merely emulsified with a non-ionic (Tween-80) surfactant. Besides, Table 3 and Fig. 10 present the influence of the related factors on water absorption, porosity and water permeability. It is acknowledged that water absorption increases with the improvement of open porosity, and water permeability is closely linked with the pore size and distribution. The water absorption capability of a porous medium is also directly related to fluid flow through the medium, which is affected by the volume and shape of the pore structure. The formation of open micro-pores in PF and the difference in water absorption of open-celled and closed foams could be ascribed to the fact that anionic surfactants promoted the formation of an open cell structure in the foam. And the specific mechanism had already discussed in the previous paragraphs.

Fig. 9 The relationship, in modified foams, between the water absorption and the content of epoxy ((a) emulsified by 100% Tween-80; (b) emulsified by 67.5% SDS and 32.5% Tween-80).

Table 3 Porosity and water absorption data for different ratios of hydrochloric/phosphoric acid in volume, and different amounts of foaming agent (the content of epoxy was 15 wt%)

Ratio of hydrochloric/phosphoric acid in the volume (10 parts curing agent)	Porosity of open cell PF	Water absorption
6 : 4	85.658%	0.538
7 : 3	88.430%	2.918
8 : 2	88.019%	2.008
9 : 1	84.237%	2.631
10 : 0	81.830%	1.953

---

Amount of foaming agent (in equal ratio 9 : 1 of mixed acid 10 parts curing agent)	Porosity of open cell PF	Water absorption
8%	83.063%	2.422
10%	88.430%	2.918
12%	82.665%	2.079

---

Fig. 10 Effect of different SDS content on water absorption and permeability.

From Fig. 10, we could also reach the conclusion that the water permeability gradually decreased with increasing content of SDS in the compound surfactants, owing to the pore structure becoming smaller. The smaller the pore size, the more water will be retained in the micro-pores. It also indirectly indicates the relationship between pore size and content of SDS: the pore size became smaller with the increasing ratio of SDS. Furthermore, as the SDS content increased, the water absorption also increased until the content of SDS reached 67.5%. Further increasing the content of SDS caused the adjacent cells to collapse and rupture. With the increasing numbers of ruptured bubbles, its water absorption decreased instead. Thus we concluded that the higher the amount of SDS, the smaller the pore diameter of the cells in the obtained PF, and more water would be retained inside the pores, resulting in higher water absorption. Therefore, this phenomenon was attributed to two factors: one was the impact of the amount of anionic surfactant (SDS) on the pore structure; the other was the mutual emulsification and demulsification effects of the SDS anionic emulsifier. However, once the proportion of SDS further exceeded 67.5%, the cells would collapse to a large extent, owing to insolubilization of excessive SDS in the resol and the impact on the foam curing process. Noticeably, in the curing process, the curing rate with a single strong acid catalyst was too fast; our method used a mixture of a strong acid (hydrochloric acid) and a medium-strong acid (phosphoric acid) as a compound curing agent with the aim of controlling and adjusting the curing rate so that it could suitably match the foaming rate. Under such suitable conditions, foaming-gas could escape outward from the resin droplets smoothly, leading to the cells in the foam being interconnected with each other and forming a massive micro-porous structure. What is more, because of the reduced amount of hydrochloric acid, this also could decrease the corrosion of the mold in the foaming process. By comparison of Tables 3 and 4, we can conclude that there are three factors affecting the porosity of PF: the ratio of the hydrochloric/phosphoric acids, the amount of foaming agent and the content of epoxy. We found that when the ratio of hydrochloric/phosphoric acid in the volume was 7 : 3, and the amount of blowing agent and epoxy content were 10 wt%, the open cell porosity reached the maximum value of 88.430%. In addition, we also chose the preferred different mixed acid type for comparison by water absorption and open cell porosity.

Table 4 Porosity, bulk density and water absorption data for different amounts of epoxy resin

Sample no.	Pure PF	5% epoxy-PF	10% epoxy-PF	15% epoxy-PF
Water absorption	2.0272	2.218	2.251	2.918
Open cell porosity	89.123%	91.209%	90.235%	88.430%
Bulk density (kg m^−3)	111.79	157.39	223.37	235.51

---

Sound absorption properties

Fig. 11 shows the experimental and theoretical sound absorption coefficients of the specimens. The results are averages of four samples. To comprehensively assess the sound absorption performance of the different samples, the average sound absorption is defined as the average of the sound absorption coefficient determined at octave bands with center frequencies of 400, 500, 800, 1000, 1250, 1600 and 2500 Hz. The average absorption coefficient determines the overall acoustic performance of the porous material and is shown in Fig. 11. It is obvious that the sound absorption coefficient generally increases with frequency. The absorption coefficient of pure phenolic foam is the lowest, and the sound absorption coefficient is not more than 35% in the range of 400–2500 Hz. Meanwhile, with the dosage of epoxy resin increasing, the sound absorption coefficient increased significantly; when the loading of epoxy reaches 10 wt%, the acoustic sound absorption can reach 95% at high frequencies. According to the theory previously reported, the main mechanism of sound absorption can be summed up by resonance absorption and interface dissipation. When a sound wave propagates in the microcellular foam, sound travels through the porous material, and sound energy is converted to thermal energy through friction between the air and the inner walls of the pores. This friction is largely captured by air flow resistivity, resulting in the sound attenuation, and thus achieves the purpose of weakening the reflected acoustic and sound absorption. As for the variation of the sound absorption coefficient, this is ascribed to the cell size and distribution, and open-cell porosity. With regard to the porosity and SEM analysis, it clearly verifies the conclusion: the higher the open-cell porosity and the more uniform the cell distribution, the higher the sound absorption coefficient obtained. Besides, the sound absorption coefficient of phenolic open-cell foam is also related to the pore size, so it will actually increase along with the pore size decreasing.

Fig. 11 The sound absorption coefficient of phenolic open-cell foams with different epoxy content (0 wt%, 5 wt%, 10 wt%, and 15 wt%).

Mechanical properties

Fig. 12 shows the compressive and flexural strengths of open cell phenolic foam modified with different loadings of epoxy resin. For PF/epoxy composite systems, it is obvious that epoxy resin can significantly improve the mechanical strength of open cell phenolic foam. In Fig. 12, with increasing dosage of epoxy, the compressive strength increases at first and then ceases to increase, and then increases progressively gently when the addition of epoxy changes from 10 wt% to 15 wt%. At the same time, the flexural strength of PF/epoxy increases gradually as the content of epoxy increases. When the epoxy loading reached 10 wt%, the modified open cell phenolic foam composite showed higher mechanical strengths (0.498 MPa in compressive strength, 0.78 MPa in flexural strength), compared to the pure phenolic foam (0.09 MPa in compressive strength, 0.21 MPa in flexural strength), indicating that the incorporation of a suitable amount of epoxy resin could improve the mechanical properties of the epoxy/PF composite systems. As for the improvement in mechanical properties, apart from the influence of the microporous structure of the foam, we can also assume that increased crosslink density (cause by increasing the epoxy concentration) causes the network strands to become shorter and have a chance to increase their strength.²⁵

Fig. 12 Compressive and flexural strengths of epoxy/PF systems modified with different content of epoxy with equal ratio of SDS = 67.5%, Tween-80 = 32.5% in 4 parts.

It is well known that the strength of open cell phenolic foam depends on the foam structure, such as the pore size, the uniform distribution of pore structure, the toughness of the cell wall and the ratio of open cells. During compressive loading, open pores help scatter and absorb the stress and energy for keeping the integrity of the framework, which may ensure that the material has a high strength. Furthermore, the addition of epoxy could make the open pore structure more uniform in distribution and smaller in size. Also, it can be seen in Fig. 12 that the strength increases little when the epoxy content increases above 10 wt%, indicating that there was an optimum addition quantity of modifier to improve the flexural and compressive strengths of open cell phenolic foam in our experiments. What’s more, when the content of epoxy resin exceeds 15%, the resin viscosity is too high, resulting in lower liquidity, which will cause a considerable increase in the non-uniform cell structure and a gradual increase of the collapse of the holes, resulting in a reduction in the mechanical properties. Besides, comparing Table 4 and Fig. 12 indicated that the bulk density of open cell PF increases continuously with the increasing dosage of epoxy resin, but the mechanical strengths did not always increase. Additionally there was an optimum addition quantity of modifier to improve the flexural and compressive strengths of open cell phenolic foam in our experiments. So we can draw the following conclusions: the amount of epoxy was responsible for higher mechanical strengths but not the foam density; in addition, the influence of the content of the epoxy resin was affected by the open cell porosity, the size and the distribution of the micropores.

Conclusion

In this work, a simple and effective route for obtaining high strength micro-porous phenolic open-cell foam modified by epoxy resin with a simple physical foaming method has been described. We successfully synthesized modified resols with different loadings of epoxy. The effects of epoxy on the micro-structure of open-cell porous phenolic foams and the properties of the pre-polymers were studied. The porous structure, water absorption, mechanical properties and so on of the open-cell phenolic foams and the curing behavior, molar weight, and viscosity of the resols were characterized.

The M_n and M_w values of modified resol with a narrow molecular weight distribution increased more than that of pure resol. The addition of epoxy increased the viscosity of the pre-polymer which was obvious when the dosage of epoxy was over 5 wt%. Surface tension analysis also revealed the emulsification effect of anionic surfactants, i.e. SDS could be more compatible with resol to make the surface tension of the system decrease so as to stabilize the bubbles. Besides, the micro-pore structure and distribution inferred from the water absorption, porosity, water permeability and morphology of the micro-pores, demonstrate that when the ratio of SDS in the four part compound emulsifier is in the range of 65–67.5%, the amount of foaming agent cyclopentane reaches 10%, the epoxy resin content reaches 10%, and the mixture of hydrochloric/phosphoric acids has a ratio of 7 : 3 in volume, the pore structure is more homogenously dispersed with a moderate size. Water absorption, sound absorption properties and thermal conductivity also increase with the improvement of open cell porosity. The conductivity of 5 wt% epoxy modified phenolic foam exhibits the highest thermal conductivity and open cell porosity (91.209%), and when the content of epoxy is 15 wt%, the highest water absorption reached up to 2.918. With the dosage of epoxy resin increasing, the sound absorption coefficient increased significantly; even the acoustic sound absorption can reach 95% at a higher frequency. Besides, through the SEM results it was shown that the pore size and distribution of foams would turn smaller and more homogenous with an increasing content of epoxy resin. As to the mechanical properties, the compression and bending results show that the incorporation of 10% content of epoxy can improve the bending strength of the epoxy/PF system at maximum extent by 2.71, and increase the compressive mechanical performance by 4.53. This is ascribed to the fact that open micro-pores help scatter and absorb the stress and energy for keeping the integrity of the framework, which may ensure that the material has a high strength.

References

1. Z. A. Van and J. J. Williams, Polymer, 1986, 5, 123–140.
2. D. Doermann and J. F. Sacadura, J. Heat Transfer, 1996, 118, 88–93.
3. J. H. Auber and R. L. Clough, Polymer, 1985, 26, 2047–2054.
4. P. W. R. Raj, M. Sasthav and H. M. Cheung, J. Appl. Polym. Sci., 1993, 47, 499–513.
5. V. K. Rangari, T. A. Hassan, Y. X. Zhou, H. Mahfuz, S. Jeelani and B. C. Prorok, J. Appl. Polym. Sci., 2007, 103, 308–314.
6. J.-N. Yang and P. Li, J. Reinf. Plast. Compos., 2015, 34, 534–546.
7. S.-W. Lei, Q. Guo, D.-Q. Zhang, J.-L. Shi, L. Liu and X.-H. Wei, J. Appl. Polym. Sci., 2010, 117, 3543–3550.
8. S.-W. Lei, Q. Guo, J.-L. Shi and L. Liu, Carbon, 2010, 48, 2644–2646.
9. D. Puglia, L. B. Manfredi, A. Vazquez and J. M. Kenny, Polym. Degrad. Stab., 2001, 73, 521–527.
10. M. L. Auad, L.-H. Zhao, H.-B. Shen, S. R. Nutt and U. Sorathia, J. Appl. Polym. Sci., 2007, 104, 1399–1407.
11. J. P. Pascault and R. J. J. Williams, Epoxy Resin, WILY-VCH Verlag GmbH & Co. KGaA, 2010.
12. P. A. Jorge and J. C. Malcolm, J. Sound Vib., 2010, 44, 12–17.
13. R. D. Rey, J. Alba, P. A. Jorge and V. J. Sanchis, Appl. Acoust., 2012, 73, 6–7.
14. X. Sagartzazu, N. L. Hervella and J. M. Pagalday, Arch. Comput. Meth. Eng., 2008, 15, 311–342.
15. J.-H. Lin, C.-M. Lin, C.-C. Huang, C.-C. Lin, C. T. Hsieh and Y.-C. Liao, J. Compos. Mater., 2011, 45, 1355–1362.
16. Pool field Yihong JP Pat., 5876432, 1983.
17. S. A. Kumar and K. S. Ahmed, J. Reinf. Plast. Compos., 2013, 32, 1370–1379.
18. J. Song, Y. Zhang and Y.-J. Liu, RSC Adv., 2015, 5, 203734–220740.
19. A. Aierbe, J. M. Cheverria and C. Riccardi, Polymer, 2002, 43, 2239–2243.
20. B. D. Park, B. Riedl, E. W. Hsu and S. Jack, Polymer, 1999, 40, 1689–1699.
21. S. R. Dhakate, K. M. Subhedar and B. P. Singh, RSC Adv., 2015, 5, 43036–43057.
22. R. Lin, L. Fang, X. Li, Y. Xi, S. Zhang and P. Sun, Polym. J., 2006, 38, 178–183.
23. M.-C. Li, X. Ge and U. R. Cho, Macromol. Res., 2013, 21, 793–800.
24. Y. Yang and J. He, Polym. Compos., 2014 DOI:10.1002/pc.23066.
25. X. Ge, M. Li and U. R. Cho, Polym. Compos., 2014 DOI:10.1002/pc.23080.

---