High performance epoxy composites cured with ionic liquids

High performance epoxy composites cured with ionic liquids

G Model JIEC-2568; No. of Pages 7 Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx Contents lists available at ScienceDirect Jour...

805KB Sizes 1 Downloads 97 Views

G Model

JIEC-2568; No. of Pages 7 Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

High performance epoxy composites cured with ionic liquids Honorata Ma˛ka a, Tadeusz Spychaj a,*, Marek Zenker b a b

West Pomeranian University of Technology, Polymer Institute, ul. Pulaskiego 10, 70-322 Szczecin, Poland West Pomeranian University of Technology, Department of Electrotechnology and Diagnostic, ul. Sikorskiego 37, 70-313 Szczecin, Poland

A R T I C L E I N F O

Article history: Received 15 September 2014 Received in revised form 8 June 2015 Accepted 27 June 2015 Available online xxx Keywords: Epoxy resin Ionic liquid Carbon nanotubes Curing reaction Thermomechanical and conductive properties

A B S T R A C T

Two ionic liquids (ILs) with the same dicyanamide anion and various cation types: imidazolium and phosphonium as catalytic agents of epoxy resin have been compared. Neat epoxy materials cured with imidazolium IL exhibited above 45 days latency, high thermal performance, i.e. glass transition temperature (above 190 8C), low tan d (0.17), and stability (395 8C/5% weight loss), whereas with phosphonium IL – pot life above 70 days, similar tan d value and high transparency (85%). Carbon nanotubes modified epoxy composites cured with imidazolium dicyanamide IL showed bulk electrical resistivity in a range 107–103 Vm for carbon filler content 0.0625–0.25 wt.%. These features designate developed materials for various engineering applications. ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Introduction The ionic liquids (ILs) can be used as catalytic epoxy resin crosslinkers solely [1–7] or in combination with other conventional hardeners, such as polyamines [3,8–11] or anhydrides [12,13]. Epoxy compositions with ILs exhibited latency at room temperature which could be regulated in a rather wide range from a few days (e.g. 3–4, for systems with 1-butyl-3-methylimidazolium thiocyanate, [BMIM]SCN, [7] or phosphonium phosphinate derivative [6,14]) up to ca. 50 days (imidazolium ILs with chloride or tetrafluoroborate anions [4,5]) when used in similar weight ratio to epoxy resin (3 parts per 100 parts of resin, 3 phr). Besides the mentioned latent character, to important advantages of ILs as epoxy resin curing agents belong low loading (usually 3–9 phr), fast and easy miscibility as well as homogeneity with liquid epoxy resin (before and after curing process) resulting in technological feasibility. In addition, ILs can play multifunctional role in epoxy material acting as polymer matrix modifier [8,15] or carbon nanofiller dispersing agent [9–12,14] as well. Until today, imidazolium type ILs have been more often applied in epoxy resin systems than phosphonium ones [14]. However, the reasons why one might be interested in a phosphonium IL, even in industrial process, include its availability and cost. Phosphonium salts are manufactured on multiton scale [16,17]. Recently,

* Corresponding author. Tel.: +48 914494684; fax: +48 914494247. E-mail address: [email protected] (T. Spychaj).

phosphonium ILs have been tested as epoxy resin crosslinkers used independently [6,14] or in combination with aromatic polyamine [18]. Soares at al. [18] published limited DSC and FTIR results on epoxy resin curing process with some imidazolium (1,3bis octadecylimidazolium iodide) and phosphonium ILs (octadecyltriphenyl phosphonium hexafluorophosphate). However, according to our best knowledge, up to now, no systematic study on simultaneous an influence of the imidazolium and phosphonium ILs on epoxy resin curing process and properties of the relevant materials has been reported. On the other side it seemed that known features of epoxy materials cured with ILs could be improved by careful selection of IL type and concentration as well as by introduction of modifying ingredient, e.g. carbon nanotubes in low loading to obtain high performance epoxy materials. Epoxy materials filled with carbon nanotubes (CNT) can be used for various engineering applications, like electrical conductive adhesives [12], antistatic or corrosion resistant coatings [19,20], electromagnetic shielding materials [21] or with thermally improved features [22,23]. Considering polymeric materials exhibiting enhanced electrical features the amount of filler is a critical aspect allowing to change material properties from insulator to conductor. Percolation threshold depends on many factors, such as: aspect ratio, allignment, functionalization, processing, polymer type etc. [24]. In this work two ILs with the same counteranion: imidazolium type, i.e. 1-ethyl-3-methylimidazolium dicyanamide, [EMIM]N(CN)2, and phosphonium derivative, i.e. trihexyltetradecyl phosphonium dicyanamide, [THTDP]N(CN)2, were used as

http://dx.doi.org/10.1016/j.jiec.2015.06.023 1226-086X/ß 2015 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

the curing catalysts to obtain epoxy materials (also CNT modified) with enhanced performance. The study had two main goals: (i) scientific: to compare curing characteristics of epoxy systems depending on the ILs cation type, (ii) to enhance technological and utility properties of the epoxy compositions and composites for industrial usage. Experimental Materials Epoxy resin: bisphenol A-based low molecular weight Epidian 6 (E6): epoxy equivalent 185 g (viscosity 18,000 mPa s at 23 8C), Organika Sarzyna, Poland was used. Trihexyltetradecyl phosphonium dicyanamide (95%), Sigma–Aldrich, and 1-ethyl-3-methylimidazolium dicyanamide (95%) from Iolitec Ionic Liquid Technologies GmbH, Heilbronn, Germany were used as epoxy resin curing catalysts (Table 1). Multiwall carbon nanotubes (CNT), Nanocyl NC7000, with specific surface 250–300 m2/g, average length 1.5 mm, average diameter 9.5 nm, carbon content 90 wt.% (Nanocyl, Belgium) were applied for epoxy composite preparation when [EMIM]N(CN)2 was used as epoxy crosslinking catalyst. Preparation of epoxy compositions, CNT dispersions and epoxy composites The neat epoxy compositions were prepared by mixing epoxy resin with [EMIM]N(CN)2 or [THTDP]N(CN)2 at ambient temperature. The IL content was 1, 3, 6 and 9 wt. parts/100 wt. parts epoxy resin (phr). CNT (0.0625–1.0 wt.% in relation to epoxy resin) were dispersed in epoxy resin by ultrasonication for 2 h (amplitude 50%, frequency 50 Hz, UP 200S, Hielscher GmbH, Germany). Next, [EMIM]N(CN)2 was added into the epoxy composition with CNT and the systems mixed manually for 10 min. Constant amount of [EMIM]N(CN)2 (3 phr) was adjusted in all epoxy compositions and composites modified with CNT. Eventually, the epoxy compositions were cured in Teflon mold at 160 8C for 2 h. The resultant samples, i.e. compositions and composites were used for further investigations.

The glass transition temperatures (Tg), tan d values and storage moduli were determined using dynamic mechanical thermal analysis (DMTA Q – 800, TA Instruments) with dual cantilever, at heating rate of 3 8C/min from 30 to 250 8C, frequency 1 Hz. Thermogravimetric analysis (TGA Q-500, TA Instruments) was performed using platinum pan under 25 mL/min air flow, in temperature range 40–800 8C, at a heating rate of 10 8C/min. The volume electrical resistance of cured composites with various CNT content was tested at room temperature in accordance with IEC 93:1980 and ASTM D 257-99 using Keithley Instruments, Inc., USA, with a set of electrodes (Keithley 8009). Thermal conductivity of disk shaped composite samples was measured at room temperature using TPS 2500S apparatus (Hot Disk AB Company) with 7577S sensors. Results and discussion Pot life of epoxy compositions Pot life of epoxy compositions with applied ILs was controlled during storage at ambient temperature by rheometric measurements. Relevant dependences are shown in Fig. 1. The both E6/IL systems exhibited prolonged storage time >45 days and >70 days when [EMIM]N(CN)2 and [THTDP]N(CN)2 was applied, respectively. That feature was to some extent dependent on IL/epoxy resin weight ratio. Increasing IL amount in epoxy composition resulted

(a)

1-Ethyl-3-methylimidazolium dicyanamide, [EMIM]N(CN)2

4000

3000

2000

1000

0 0

20

40

60

80

100

120

Storage time (days)

(b) 7000

E6/[THTDP]N(CN)2_3 E6/[THTDP]N(CN)2_6 E6/[THTDP]N(CN)2_9

6000

5000

Viscosity (Pas)

Table 1 Ionic liquids used as epoxy curing catalysts.

E6/[EMIM]N(CN)2 _1 E6/[EMIM]N(CN)2 _3 E6/[EMIM]N(CN)2 _9

5000

Methods The viscosity of epoxy compositions with CNT was determined using stress rheometer (Rheometric Scientific, USA) at room temperature, a plate–plate system, f = 40 mm, a gap of 1 mm. The pot life for epoxy resin/IL compositions at ambient temperature was determined on a basis of viscosity measurements during storage at 23  2 8C using stress rheometer. The curing process of epoxy compositions was investigated using differential scanning calorimeter DSC Q-100 (TA Instruments, USA), at a heating rate of 5 8C/min in the temperature range of 30–300 8C and rheometer at the same measurement schedule.

7000

6000

Viscosity (Pas)

2

4000

3000

2000

1000

Trihexyltetradecylphosphonium dicyanamide, [THTDP]N(CN)2

0 0

20

40

60

80

100

120

Storage time (days) Fig. 1. Viscosity change of epoxy compositions during storage at room temperature: (a) E6/[EMIM]N(CN)2, (b) E6/[THTDP]N(CN)2 systems.

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

in shortening the pot life. Considered influence was more pronounced in case of E6/([EMIM]N(CN)2 system. We reported room temperature storage times for similar epoxy system containing [BMIM]N(CN)2 [4]. Some differences with the length of imidazolium alkyl chain (C4 to C2) were observed. The relevant pot lives increased for compositions containing [EMIM]N(CN)2 – from 30 to 80 days, with decreasing ILs content in epoxy system. The values of that parameter for epoxy systems with phosphonium IL bearing dicyanamide anion differed dramatically in comparison to those with phosphinate anion reported recently [6,14]. Relevant results for the latter epoxy systems were: 4–24 days (10 ! 2.5 phr IL [6]) or 2–7 days (9 ! 3 phr [14]), whereas pot life for E6/[THTDP]N(CN)2 was 70 ! 120 days (Fig. 1B). The only reason of the observed differences could be various stability of analogous salts which correlated with anion hydrophobicity, since this was a measure for the H-bonding capacity and hence nucleophilicity [25,26]. Even when both discussed ILs with [THTDP] cation were hydrophobic that containing phosphinate anion exhibited higher water tolerance as compared to [THTDP]N(CN)2 (maximum water capacitance 20.6% in comparison with 3.1%, respectively) [27]. As a result, phosphinate IL could decompose easier at room open atmosphere and cause epoxy crosslinking. Also thermal stability of phosphonium IL with phosphinate anion was lower than for IL with dicyanamide anion [27]. Epoxy resin curing with ILs at elevated temperatures Curing process of epoxy compositions was followed by rheometric and DSC techniques during heating above 40 8C. In

3

Fig. 2 cumulative results of viscosity and heat flow changes measured at dynamic mode (5 8C/min) were presented. Some characteristic parameters determined from rheometric and DSC runs, i.e. onset viscosity or onset of exothermal effect, temperature at exothermal peak and reaction enthalpy, were collected in Table 2. Viscosity jumps on rheometric curves evidencing gelation of epoxy systems with imidazolium IL began in temperature range: 183 ! 163 8C (IL content 1 ! 9 phr). The epoxy systems containing phosphonium IL exhibited higher range of steep viscosity increase: 206 ! 174 8C, (IL content 3 ! 9 phr). It might be also seen that the lowest [THTDP]N(CN)2 content (i.e. 3 phr) was not enough to reach composition viscosity about 106 Pa s, likewise for other systems shown in Fig. 2A and C. DSC thermograms for E6/[EMIM]N(CN)2 compositions were bimodal with the first maximum in temperature range Tmax = 136 ! 133 8C and the other 180 ! 164 8C. Epoxy compositions with phosphonium IL exhibited unimodal thermograms with Tmax = 180 ! 170 8C. Bimodal DSC thermograms were observed also when other imidazolium ILs with dicyanamide anion for epoxy resin curing have been used [2,4]. Considering available literature data on DSC thermograms of epoxy resin cured with imidazolium ILs [1–5,18] it could be concluded that for the first exothermal maximum (alkyl)imidazolium cation, and/or its thermal decomposition products and for the other counteranion type could be responsible. There is a simple relation between IL thermal stability and its activity as epoxy resin curing agent: the most labile catalysts cause resin gelation at lower temperature. The following order of 1-ethyl-3-methylimidazolium ILs with various anions follows from literature: BF4 (450 8C) > I

Fig. 2. Rheometric curves and DSC thermograms of epoxy compositions heated with ionic liquids: (a) and (b) with [EMIM]N(CN)2, (c) and (d) with [THTDP]N(CN)2 (heating rate: 5 8C/min).

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

4

Table 2 Curing parameters for epoxy compositions with ionic liquids. Composition acronym

E6/[EMIM]N(CN)2_1 E6/[EMIM]N(CN)2_3 E6/[EMIM]N(CN)2_9 E6/[THTDP]N(CN)2_3 E6/[THTDP]N(CN)2_6 E6/[THTDP]N(CN)2_9

Rheometry

DSC

To (8C)

To (8C)

Tmax (8C)

DH (J/g)

183 168 163 206 182 174

128/150 124/150 121/139 164 159 151

136/180 133/169 133/164 180 180 170

22/569 20/488 50/507 106 315 388

(303 8C) > Cl (285 8C) > N(CN)2 (275 8C) > SCN (226 8C) [26,28]. These data (based on TGA experiments performed under N2 atmosphere and a heating rate 10/min [26]) correlated qualitatively with the observed Tmax values of the second exothermal peaks of investigated epoxy curing process. The results for epoxy systems cured with [THTDP]N(CN)2 could be compared with those containing [THTDP] phosphinate [14]. The dicyanamide IL was less reactive as gelation began at temperature 50–60 8C higher than with phosphinate IL. Similarly, Tmax values were about 50 8C higher for [THTDP]N(CN)2 than for [THTDP] phosphinate used as catalytic curing agents. These results correlate with the thermal stability of applied ILs, i.e. 395 and 340 8C, respectively determined by TG measurements [27]. The main heat effect during epoxy resin curing with imidazolium dicyanamide was observed in the second exotherm of DSC thermograms. Relevant DH values were 569 ! 507 J/g (i.e. 10–20 times higher than for the first exotherm) for 1 ! 9 phr IL, respectively. Lower exothermic effect was noted when phosphonium dicyanamide was applied (106 ! 388 J/g, respectively). Thermomechanical and thermal properties of epoxy materials cured with ILs DMTA measurements showed that the glass transition temperatures of epoxy materials were strongly dependent on epoxy resin/IL ratios (Table 3). The highest value of Tg was noted for [EMIM]N(CN)2 – with 3 phr of IL (196 8C). Related tan d value was as low as 0.17, whereas storage modulus was 1708 MPa. The value of Tg was the highest and tan d the lowest for bisphenol A-based epoxy materials cured with ionic liquids reported so far. Moreover, Tg value of 196 8C was on the top level of that parameter determined for any bisphenol A-based epoxy systems [29]. The E6/[EMIM]N(CN)2 samples with the lowest (1 phr) and the highest IL amount were either not effectively crosslinked {E6/ [EMIM]N(CN)2_1} or over-plasticized by IL {E6/[EMIM]N(CN)2_9} and exhibited substantially lower Tg values and higher tan d than the system containing 3 phr IL. Similarly, the epoxy material with the lowest [THTDP]N(CN)2 content (3 phr) was not crosslinked effectively (Tg = 53 8C) what was expected on a basis of rheometric measurements (Fig. 2C). However, both epoxy materials cured with higher ratios of phosphonium dicyanamide IL showed high Tg

values (170 and 166 8C), low tan d (0.17 and 0.14) and comparable storage modulus values (1874 and 1887 MPa). The glass transition temperatures and tan d values are related to crosslink density of the systems. The crosslink density of the investigated materials was estimated on a basis of equation ne = Er/ 3RTr (where: Er was storage modulus at ‘‘rubbery’’ state, i.e. at Tr =Tg + 30, R – was universal gas constant) [7]. Higher Tg (and lower tan d) were noted for denser crosslinking. So, the observed Tg and tan d values correlated with the highest crosslink density for the samples E6/[EMIM]N(CN)2_3 (12 885 mol/m3) and E6/ [THTDP]N(CN)2_6 and 9 (14,856 and 15,908 mol/m3). The thermogravimetric results for epoxy materials cured with imidazolium IL {[EMIM]N(CN)2_3} and phosphonium IL {[THTDP]N(CN)2_6} for investigated materials, i.e. temperatures of 5 and 10% mass loses, were presented in Table 3. The both epoxy materials with medium amount of ILs exhibited the highest temperatures of 5 and 10% mass losses. Thermal stability of imidazolium dicyanamide cured materials was higher (395 and 421 8C, respectively), than those with phosphonium IL (377 and 405 8C, respectively). Pronounced differences between TG curves might be observed in temperature range 320–420 8C (Fig. 3). Generally, one can expect that incorporation of phosphonium IL should improve thermal stability of epoxy material as compared to that cured with imidazolium IL [10]. One reason for better thermomechanical and thermal stability of epoxy materials cured with imidazolium dicyanamide could be possibility of forming isocyanurate rings from products of 1-ethyl-3-methylimidazolium dicyanamide decomposition. Taking into account literature reports on thermal decomposition of alkylimidazolium dicyanamides [30,31] as well as the possibility of cyano groups condensation [32] we proposed the mechanism of [EMIM]N(CN)2 decomposition and isocyanurate rings formation (Fig. 4). Kroon et al. [30] on a basis of quantum chemical calculations predicted that the main 1alkyl-3-methylimidazolium dicyanamide thermal degradation products could be alkylimidazole and methylated dicyanamide. Liang et al. [31] found that although the decomposition temperatures for [BMIM]N(CN)2 were about 200 8C, the isothermal TGA data clearly showed significant mass loss at temperatures substantially below that observed by scanning TGA. Moreover, it is worth to notice that phosphonium dicyanamide cured epoxy materials exhibited high thermomechanical properties being highly transparent materials even after processing at 160 8C. In Fig. 5 the images of epoxy materials samples cured with [THTDP]N(CN)2_6 and [EMIM]N(CN)2_3 as well as transparencies of the both samples are presented. It could be seen that transparency of the former at 550 nm was ca. 85% whereas the latter was black opaque. Epoxy compositions and composites filled with CNT On a basis of thermomechanical and thermal stability results as well as considering dispersing ability of [EMIM]N(CN)2 toward carbon nanofillers [10] the epoxy compositions and composite

Table 3 Results of DMTA and TGA measurements for epoxy materials cured with imidazolium and phosphonium ionic liquids. Composition acronym

E6/[EMIM]N(CN)2_1 E6/[EMIM]N(CN)2_3 E6/[EMIM]N(CN)2_9 E6/[THTDP]N(CN)2_3 E6/[THTDP]N(CN)2_6 E6/[THTDP]N(CN)2_9

DMTA

TGA

Tg (8C)

Tan delta

Storage modulus (MPa)

Crosslink density (mol/m3)

T5% (8C)

T10% (8C)

110 196 149 53 170 166

0.43 0.17 0.55 1.53 0.17 0.14

2524 1708 1788 310 1874 1887

2889 12,885 4563 282 14,586 15,908

374 395 357 276 377 369

409 421 399 306 405 398

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

5

However, rather high concentration of the IL (17.6 phr relative to epoxy) for CNT dispersion and 4,40 -diaminodicyclohexylmethane as the main curing agent has been applied. In our work, CNTs first were dispersed in a proper amount of epoxy resin via sonication and then IL has been introduced into the system. Constant amount of IL (3 phr) was kept whereas CNTs content was changed. The viscosities of epoxy resin with changing CNT content 0.0625–1.0 wt.% were controlled via rheometric measurements at ambient temperature. The results were collected in Table 4. It could be seen that even very low CNT loading caused important viscosity increase: ca. 60%, and ca. 100% for 0.0625 wt.% and 0.125 wt.%, respectively. The highest CNT content (i.e. 1 wt.%) resulted in more than 13 times viscosity increase (18,000 ! ca. 244,000 mPa s). These results are in accordance with other reports [14,33]. Curing of epoxy resin/CNT/[EMIM]N(CN)2 systems followed by rheometry exhibited some differences in comparison to [EMIM]N(CN)2_3 neat system (Fig. 6). For the latter system viscosity decrease could be observed while heated from 40 ! 100 8C (6 ! 0.15 Pa s), whereas those modified with nanofiller exhibited viscosity on stable levels, i.e. 2.2  101 (0.25 wt.% CNT), 1.2  102 (0.50 wt.% CNT) and 4.7  102 Pa s (1.00 wt.% CNT) in the same temperature range. Slight and irregular shift of rheometric curves to higher temperature values was found in relation to the epoxy neat system. Relevant onset temperature differences DT were in a range up to 5 8C (Fig. 6). The highest temperature shift was observed for epoxy composition with the lowest CNT amount (0.25 wt.%) investigated.

Fig. 3. Thermogravimetric curves of cured epoxy materials with 3 phr imidazolium IL (E6/([EMIM]N(CN)2_3) and 6 phr of phosphonium IL ([THTDP]N(CN)2_6).

materials basing on [EMIM]N(CN)2_3 system and modified with CNTs have been prepared. Palmese et al. [10] have found very low percolation threshold for epoxy materials modified with single wall carbon nanotubes (8.6  105 volume fraction) when [EMIM]N(CN)2 was used as CNT dispersing medium and dispergation performed using 3-role mill.

N

N +

N

C

C N

reverse Menschutkin

N

+

N

N

N

C

C

N

N [EMIM]N(CN) dealkylation SN 2

fragmentation

N

N

+

R

N N C

C

N

N

+

C

N N

N C N H3C N N

N

H3C

C

C N

N

N

N

N

C

N

N

N N

N

CH3

N

C N

N

N

N

N

C

N CH 3

N N

C N

N

Fig. 4. Scheme of possible 1-ethyl-3-methylimidazolium dicyanamide thermal decomposition with creation of isocyanurate rings.

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

6

Table 4 Viscosity of epoxy resin with various CNT content. Composition acronym

Viscosity, (Pa s) at 23 8C

E6 E6/CNT_0.063 E6/CNT_0.125 E6/CNT_0.25 E6/CNT_0.50 E6/CNT_1.00

18.0  0.4 28.6  0.7 36.1  0.5 38.3  0.1 105.8  0.5 243.8  0.8

All CNT-filled composites exhibited lower glass transition temperatures, however, difference between Tg values for the composite with 0.25 wt.% CNT and reference neat material was especially pronounced (184 and 196 8C, respectively) – Table 5. This downshift was significantly higher than an error of Tg determination (ca. 1 8C). It is reported that Tg behavior is dependent on the filler shape [34,35]. The physical hindrance of CNT in epoxy system might be a major factor impairing the mobility of the active groups of rather large epoxy resin molecules, thus leading to lower curing degree. Therefore, increased distance between chains in polymer network resulted in lowered intermolecular forces and observed decrease Tg and crosslink density (Table 5). Similarly, Barrau et al. [35] observed Tg decrease in epoxy nanocomposites with CNT (content below 0.1 wt.%) corresponding to electrical percolation threshold (i.e. in a region close to that found in this work). An increase of that parameter with higher CNT amount could be observed. Also some increase of tan d and decrease of storage moduli as compared to reference epoxy neat material was registered (Table 5). Moreover, substantial crosslinking density decrease was similar to recently reported results where CNTs were used for modification of epoxy materials cured with phoshonium phosphinate IL [14]. Thermal resistance of epoxy composites with CNT was slightly lower than that for neat epoxy material cured with [EMIM]N(CN)2_3. Because of excellent thermal conductivity of individual CNT (3000 W/mK) [36] it was often employed in polymer composites for enhancing the heat transfer rate. However, literature results showed that thermal conductivities of epoxy composites modified with CNT increased almost linearly with its content; no thermal percolation threshold was noted [37,38]. Reported thermal conductivity of neat crosslinked epoxy materials changed in a range 0.13–0.24 W/mK [37–39]. The observed differences could be caused by different substrate types and other experimental conditions as well as measuring techniques. The basic value of thermal conductivity for [EMIM]N(CN)2_3 was placed in lower region of the range mentioned above (i.e. 0.16 W/mK, Fig. 7). Our results of thermal conductivity for epoxy composites modified with CNTs changed in a range 0.16–0.28 W/mK. Introduction of CNT resulted in about 22%, 51% and 79% thermal conductivity increase of modified epoxy materials with nanofiller content 0.25 wt.%, 0.5 wt.%, and 1 wt.%, respectively (Fig. 7). Moreover, the trend of thermal conductivity change with CNT load increase was almost linear, i.e. such as reported elsewhere [37–39].

Fig. 5. Comparison of general view (a) left: E6/[THTDP]N(CN)2_6, and right: E6/ [EMIM]N(CN)2_3 and transparency (b) of epoxy castings.

Fig. 6. Rheoviscometric curves of epoxy compositions with various contents of CNT (0.25–1.00 wt.%) during curing process (heating rate: 5 8C/min).

Table 5 Results of DMTA, TGA measurements for epoxy composite materials cured with imidazolium dicyanamide and modified with CNT. Composition acronym

E6/[EMIM]N(CN)2_3 E6/CNT_0.25/[EMIM]N(CN)2_3 E6/CNT_0.50/[EMIM]N(CN)2_3 E6/CNT_1.00/[EMIM]N(CN)2_3

DMTA

TGA

Tg (8C)

Tan delta

Storage modulus (MPa)

Crosslink density (mol/m3)

T5% (8C)

T10% (8C)

196 184 195 193

0.17 0.28 0.21 0.24

1708 1751 1374 1381

12,885 6966 7184 6683

395 386 388 383

421 412 419 419

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023

G Model

JIEC-2568; No. of Pages 7 H. Ma˛ka et al. / Journal of Industrial and Engineering Chemistry xxx (2015) xxx–xxx

7

of the developed materials can be underlined: (i) good technological features (high latency), (ii) high thermomechanical and thermal properties, (iii) range of electrical resistance: between electrostatic and electromagnetic interference materials. These properties could designate investigated epoxy materials for various engineering purposes, especially where high thermal and enhanced electrical properties are required. References

Fig. 7. Electrical volume resistivity and thermal conductivity of epoxy nanocomposites with CNT as a function of nanofiller content.

The results of electrical bulk resistivity evaluation revealed (Fig. 7) that the percolation threshold was below 0.0625 wt.% CNT. This CNT content caused 6 orders decrease of electrical resistivity whereas 0.25 wt.% resulted in 10 orders decrease as compared to the reference neat epoxy material. With respect to surface electrical resistivity polymer-based coats could be divided into three categories: (i) antistatic (1011–1014 V), static dissipative (105–1011 V), and (iii) conductive (102–105 V) [38]. On the other side, Rohatekar et al. [39] considering advanced epoxy composites for aerospace application classified these materials into other 3 classes showing properties: (i) electrostatic dissipative (ca. 5  106–5  103 V m), (ii) electromagnetic interference (ca. 5  103–5  101 V m), and (iii) lightning strike percolation (>5  101 V m). According to the classifications above the developed epoxy composites with CNTs could belong to conductive [40] or electrostatic/electromagnetic interference materials [39]. Conclusions Comparison of epoxy resin curing characteristics with the two types of dicyanamide ILs with various cations: imidazolium and phosphonium was performed. In case of phosphonium ILs an important influence of anion type was revealed: N(CN)2 allowed to prolong pot life to >70 days; for comparison, pot life for investigated earlier IL with phosphinate anion was ca. 4 days. On the other hand, the epoxy compositions containing IL with imidazolium cation exhibited pot life >45 days. The epoxy materials cured with [THTDP]N(CN)2 showed higher transparency (85%) in comparison to that with [EMIM]N(CN)2 (black opaque). The epoxy neat and CNT modified composites cured with the latter IL showed high Tg (above 190 8C) and very low tan d values (0.17) as well as high mass losses temperature (386–395 8C/5 wt.%). This features were more beneficial than in cases of other IL-catalyzed epoxy materials. Nanocomposites with CNT exhibited relatively high range of bulk electrical resistivity 2.7  107–1.5  103 V m at low nanofiller content (0.0625–0.25 wt.%). The following findings

[1] K. Kowalczyk, T. Spychaj, Polimery (Warsaw) 48 (2003) 833. [2] A.M. Rahmathullah, A. Jeyarajasingam, B. Merritt, M. Van Landingham, S.H. McKnight, G.R. Palmese, Macromolecules 42 (2009) 3219. [3] B.G. Soares, S. Livi, J. Duchert-Rumeau, J.F. Gerard, Macromol. Mater. Eng. 296 (2011) 826. [4] H. Ma˛ka, T. Spychaj, R. Pilawka, Ind. Eng. Chem. Res. 51 (2012) 5197. [5] H. Ma˛ka, T. Spychaj, Polimery (Warsaw) 57 (2012) 34. [6] A.A. Silva, S. Livi, D.B. Netto, B.G. Soares, J. Duchet, J.F. Gerard, Polymer 54 (2013) 2123. [7] H. Ma˛ka, T. Spychaj, K. Kowalczyk, J. Appl. Polym. Sci. 131 (2014) 40401. [8] J. Sanes, F.J. Carrion, M.D. Bermudez, Wear 268 (2010) 1295. [9] B. Guo, J. Wan, Y. Lei, D. Jia, J. Phys. D: Appl. Phys. 42 (2009) 145307. [10] J.A. Throckmorton, A.L. Watters, X. Geng, G.R. Palmese, Compos. Sci. Technol. 86 (2013) 38. [11] N. Hameed, N.V. Salim, T.L. Hanley, M. Sona, B.L. Fox, Q. Guo, Phys. Chem. Chem. Phys. 15 (2013) 11696. [12] X. Zhang, H. Sun, C. Yang, K. Zhang, M.M.F. Yuen, S. Yang, RSC Adv. 3 (2013) 1916. [13] M.S. Fedoseev, M.S. Gruzdev, L.F. Derzhavinskaya, Int. J. Polym. Sci. 2014 (2014), ID 607341. [14] H. Ma˛ka, T. Spychaj, R. Pilawka, Express Polym. Lett. 8 (2014) 723. [15] J. Sanes, F.J. Carrion-Vilches, M.D. Bermudez, e-Polymer 5 (2007) 1. [16] S.Y. Yang, F. Cicoira, R. Byrne, F. Benito-Lopez, D. Diamond, R.M. Owens, G.G. Malliaras, Chem. Commun. 46 (2010) 7972. [17] C.J. Bradaric, A. Downard, C. Kennedy, A.J. Robertson, Y. Zhou, Green Chem. 5 (2003) 143. [18] B.G. Soares, S. Livi, J. Duchert-Rumeau, J.F. Gerard, Polymer 53 (2012) 60. [19] B.M. Sangermano, S. Pegel, P. Potschke, B. Voit, Macromol. Rapid Commun. 29 (2008) 396. [20] H.R. Jeon, J.H. Pork, M.Y. Shon, J. Ind. Eng. Chem. 19 (2013) 849. [21] A.S. Hong, H.N. Nguyen, H.T. Bui, A.T. Tran, V.A. Duong, V.B. Nguyen, Adv. Nat. Sci. Nanosci. Nanotechnol. 4 (2013) 025012 (5 pp.). [22] T. Kuilla, S. Bhandra, D. Yao, Progr. Polym. Sci. 35 (2010) 1350. [23] B. Li, W.H. Zhong, J. Mater. Sci. 4 (2011) 5595. [24] G. Mittal, V. Dhand, K.Y. Rhee, S.J. Park, W.R. Lee, J. Ind. Eng. Chem. 21 (2015) 11. [25] J.G. Huddleston, A.E. Visser, W.M. Reichert, H.D. Willauer, G.A. Broker, R.D. Rogers, Green Chem. 3 (2001) 156. [26] C. Maton, N. De Vos, C.V. Stevens, Chem. Soc. Rev. 42 (2013) 5963. [27] K.J. Fraser, D.R. MacFarlane, Aust. J. Chem. 62 (2009) 309. [28] P.S. Kulkarni, L.C. Branco, J.G. Crespo, C.A.M. Afonso, Chem. Eur. J. 13 (2007) 8470. [29] R.S. Bauer, L.S. Corley, Reference Book for Composites Technology, Technomic. Publ. Co. Inc., Lancaster, USA, 1989, pp. 17–48 (Chapter 2). [30] M.C. Kroon, W. Buijs, C.J. Peters, G.J. Witkamp, Thermochim. Acta 465 (2007) 40. [31] R. Liang, M. Yang, X. Xuan, Chin. J. Chem. Eng. 18 (2010) 736. [32] N. Fechler, Salts as Highly Diverse Porogens: Functional Ionic Liquid-Derived Carbons and Carbon-Based Composites for Energy-Related Applications, (Ph.D. thesis), Max-Planck Institut fu¨r Kolloid und Grenzfla¨chenforschung, Potsdam, 2012. [33] S.S. Rahatekar, K. Koziol, S.A. Butler, J.A. Elliott, M.S.P. Shaffer, M.R. Mackley, A.H. Windle, J. Rheol. 50 (2006) 599. [34] K. Tao, S. Yang, J.C. Grunlan, Y.S. Kim, B. Dang, Y. Deng, R.L. Thomas, B.L. Wilson, X. Wei, J. Appl. Polym. Sci. 102 (2006) 5248. [35] S. Barrau, P. Demont, C. Maraval, A. Bernes, C. Lacabanne, Macromol. Rapid Commun. 26 (2005) 390. [36] F. Wu, X. He, Y. Zeng, H.M. Cheng, Appl. Phys. A 85 (2006) 25. [37] Ch. Kuo, H.M. Huang, J. Therm. Anal. Calorim. 103 (2011) 533. [38] M.M. Gallego, R. Verdejo, M. Khayet, J.M. Ortiz de Zarate, M. Essalhi, M.A. LopezManchado, Nanoscale Res. Lett. 6 (2011) 610. [39] M. Russ, S.S. Rahatekar, K. Koziol, B. Farmer, H.X. Peng, Comp. Sci. Technol. 81 (2013) 42. [40] M. Berkei, Chemanager Europe 7–8 (2011) 10, www.chemanager-online.com/en/ tags/coatings.

Please cite this article in press as: H. Ma˛ka, et al., J. Ind. Eng. Chem. (2015), http://dx.doi.org/10.1016/j.jiec.2015.06.023