Effect of polyethylenimine on properties of emulsion styrene butadiene rubber/butadiene rubber blends

Abstract

The purpose of this study was to evaluate the effect of polyethylenimine (PEI) on the properties of emulsion styrene butadiene rubber (ESBR)/butadiene rubber (BR) blends. PEI with molecular weight (M_w) of 800, 2,000, 25,000, and 270,000 g/mol with various ratios was successfully incorporated into ESBR/BR blends via rubber compounding method. Study showed almost 100% decrease in scorch time, ts₂ in lower M_w PEI (800 and 2000 g/mol), compared to higher M_w PEI (25,000 and 270,000 g/mol) in ESBR/BR blends. PEI incorporated ESBR/BR blends showed an overall improvement in terms of tensile strength. Lower M_w PEI induced better tensile strength with an optimum value at 1.22 MPa, while higher M_w PEI induced better tear strength and self-healing ability. Incorporation of 5 phr PEI 270,000 g/mol in the blend successfully achieved the optimum self-healing efficiency of 71.5% at 160°C for 1 h. Phase separation occurred between higher M_w viscous PEI with rubber matrix at crack region to promote thermal self-healing. PEI molecules were found to be accumulated and assembled on the interface of the rubber chain from scanning electron microscopy (SEM) images. It was suggested weak physical bond interaction such as Van der Waals holding both rubber chains and PEI together, without any changes in the rubber chain backbone.

Keywords

Polyethylenimine elastomer self-healing emulsion styrene butadiene rubber butadiene rubber mechanical properties rubber

Introduction

Polyethylenimine (PEI) is commonly used as adhesion promoter for various adhesive applications in the industrial area. It is a polymer consisting copolymer of amine group and two carbon aliphatic CH₂CH₂. Linear PEI (LPEI) contains secondary amines, in contrast with hyperbranched PEI (hPEI) which contains primary, secondary, and tertiary amino groups. PEI has been produced for industrial use since 1938 in Germany by Badische Anilin und Soda Fabrik AG (BASF) and Chemirad Corporation in the USA.^1,2 The PEI acts as a physical glue and there is adhesive force between the adhesive PEI and the blend surface, whereby the cohesive force acts within the PEI molecules and holds them together.

Interestingly, very less attention was placed upon self-healing on rubber using polyethylenimine (PEI), which is neither extrinsic nor intrinsic type healing mechanism.3 Wang et al.⁴ investigated the influence of hyperbranched PEI (hPEI) with M_w of 10,000 g/mol on the blending properties of bisphenol-A polycarbonate (PC) and amorphous polyamide (aPA) which were substantially immiscible. They found that the compatibility of the PC/aPA blend could be significantly improved by incorporating hPEI. They concluded that hPEI acted as a compatibilizer and it improved the miscibility between PC and aPA. Du et al.⁵ patented a type of formulation that relates to preparation of a diene-based rubber with hPEI and reinforcing filler such as carbon black and precipitated silica in tire application. PEI was used to promote sulfur vulcanization of diene-based elastomers by acting as secondary sulfur vulcanization accelerator.

Schussele et al.³ studied the self-healing in NBR blends using dispersed additives, low-viscosity hyperbranched PEIs (M_w of 800 g mol⁻¹; 2000 g mol⁻¹; hexylurea-modified PEI; phenylurea-modified PEI), at 5 and 12.5 phr, respectively. They found out the self-healing efficiency improved with the increase in elongation at break for both unmodified NBR/PEI blends, and were much better than those modified NBR/PEI blends. They deduced that the presence of phase separation of PEI in NBR is essential to promote thermal self-healing in rubber.

ESBR/BR blend is common in the rubber market to produce products with good heat, aging, and abrasion resistance at lower prices. ESBR/BR blend cured with peroxide-cured system exhibits lower creep strain and compression set than sulfur-cured ESBR/BR blend.⁶ There are wide selections of PEI with different M_w available in the market, ranging from 800 to 270,000 g/mol. It would be interesting to explore and compare the effect of four different types of M_w of PEI, which are 800, 2,000, 25,000, and 270,000 g/mol, respectively, on the rubber properties including their self-healing capability. With this self-healing capability in ESBR/BR blends, it is expected to respond to micro-crack formation and thus prolong the service life of the products.

According to the authors’ best knowledge, surprisingly there is no published work on the effect of PEI on the properties and self-healing ability in ESBR nor BR rubber. Moreover, there are no studies investigating the effect of PEI (M_w 25,000 and 270,000 g/mol) in rubber. In the present work, we developed ESBR/BR blends with different M_w of hyperbranched PEI (800, 2,000, 25,000, and 270,000 g/mol) via peroxide-cured system. By varying the content and M_w of PEI, the rheology, mechanical properties, self-healing ability, and crosslinking densities of the ESBR/BR blends were being studied.

Materials and methods

Materials

SBR 1502 (styrene content = 23.5%, ρ = 0.94 g/cm³ and Mooney viscosity= 52 MU) and KBR 01 (cis-1,4 content = ≥ 94%, ρ = 0.91 g/cm³ and Mooney viscosity= 45 MU) were purchased from Kumho Petrochemical. Toluene (purity = 99.5%, M_w = 92.14 g/mol) was purchased from Sigma Aldrich. Table 1 lists down the additives used in this study.

Table 1.

Additives used in the study.

Description	Properties	Function	Supplier
Zinc oxide (ZnO)	M_w = 81.39 g/mol	Processing aid	Centrewest Industrial Supplier, Malaysia
Zinc oxide (ZnO)	ρ= 5.61 g/cm³	Processing aid	Centrewest Industrial Supplier, Malaysia
Stearic acid (SA)	Purity ≥95%	Processing aid	Centrewest Industrial Supplier, Malaysia
Stearic acid (SA)	M_w = 284.48 g/mol	Processing aid	Centrewest Industrial Supplier, Malaysia
Dicumyl peroxide (DCP)	Purity = 98%	Vulcanizing agent	Sigma-Aldrich
Dicumyl peroxide (DCP)	M_w = 270.37 g/mol density = 1.56 g/mL.	Vulcanizing agent	Sigma-Aldrich
PEI-800, branched	Density = 1.05 g/cm³	Self-healing additive	Sigma-Aldrich
PEI-800, branched	M_w = 800 g/mol	Self-healing additive	Sigma-Aldrich
PEI-2,000, branched	Density = 1.05 g/cm³	Self-healing additive	Polysciences Inc
PEI-2,000, branched	M_w = 2000 g/mol	Self-healing additive	Polysciences Inc
PEI-25,000, branched	Density = 1.03 g/cm³	Self-healing additive	Sigma-Aldrich
PEI-25,000, branched	M_w = 25,000 g/mol	Self-healing additive	Sigma-Aldrich
PEI-270,000, branched	Density = 1.05 g/cm³	Self-healing additive	Sigma-Aldrich
PEI-270,000, branched	M_w = 270,000 g/mol	Self-healing additive	Sigma-Aldrich

Note. PEI = polyethylenimine

Formulations

ESBR/BR blends formulations are illustrated in Table 2.

Table 2.

Formulation of emulsion styrene butadiene rubber /butadiene rubber blends (unit: phr)^a

Compound	PEI-800	PEI-2000	PEI-25,000	PEI-270,000
B	—	—	—	—
BP1/5	5	—	—	—
BP1/7.5	7.5	—	—	—
BP1/10	10	—	—	—
BP1/12.5	12.5	—	—	—
BP1/15	15	—	—	—
BP2/5	—	5	—	—
BP2/7.5	—	7.5	—	—
BP2/10	—	10	—	—
BP2/12.5	—	12.5	—	—
BP2/15	—	15	—	—
BP3/5	—	—	5	—
BP3/7.5	—	—	7.5	—
BP3/10	—	—	10	—
BP3/12.5	—	—	12.5	—
BP3/15	—	—	15	—
BP4/5	—	—	—	5
BP4/7.5	—	—	—	7.5
BP4/10	—	—	—	10
BP4/12.5	—	—	—	12.5
BP4/15	—	—	—	15

Note. PEI = polyethylenimine

^aEach designation consisted of: 80 phr ESBR; 20 phr BR; 5 phr ZnO; 2 phr SA; 2.5 phr DCP.

Sample compounding and preparation

ESBR was milled and masticated on two-roll mills with roll gap set at 3–5 mm for 5 min and then kept aside. BR was then put into 2 roll mills and masticated for another 5 min. ESBR was then blended with BR for another 5 min. Subsequently, ZnO, SA, PEI, and lastly DCP were added to the compound. Each additive was added at a 5-min interval, with total compounding time of 40 min. The compound was continuously cut and flipped over on the mills to ensure good dispersion and distribution of the additives. The temperature of the compound was kept around 80°C to prevent premature vulcanization. The compounds were sheeted out at a thickness of 3 mm and kept overnight at room temperature for 12 h prior to other tests.

Rheological and mechanical tests

The unvulcanized samples were analyzed using moving die rheometer (MDR) at 175°C to determine the cure characteristic according to ASTM D2084. They were then compress molded at 175°C with a cure time of t₉₀ obtained from the MDR into round blocks for hardness analysis. Besides, it was molded into 30 cm x 30 cm rubber sheet with a thickness of 2.5 mm to be cut into tensile and tear specimens according to ASTM D412 and ASTM D624, respectively. Tensile strength, elongation at break, and tear strength (die C) were determined using universal testing machine at 50 mm/min. The average values of five specimens of each formulation were reported. Density of each molded hardness block was measured using densimeter, in accordance with ASTM D297 for the crosslinking density calculation of the ESBR/BR blends in swelling test.

Solvent swelling test

Equilibrium swelling experiment was conducted to determine the total cross-link density of the vulcanized material. A 1.0 g sample from each formulation was swollen in 80 mL toluene in the dark at room temperature (24 ± 1°C) for four to 7 days until it reaches equilibrium. The swollen samples were wiped with tissue paper to remove excessive toluene and weighed immediately. Lastly, the swollen samples were dried at 60°C until they achieved constant weight.^7,8

The weight of swollen gel, m₃, was measured followed by the dry rubber weight, m₁, after removing the solvent. Weight of the solvent in the swollen rubber, m₂, is shown in equation (1).

m_{2} = m_{3} - m_{1}

(1)

The volume fraction of the rubber in swollen state, Φ_r, was defined as below equation (2)

\begin{array}{l} Φ_{r} = (Volume of dry rubber) / (volume of dry rubber + volume of solvent in the swollen rubber) \\ = (m_{1} / ρ_{1}) / (m_{1} / ρ_{1} + m_{2} / ρ_{2}) \end{array}

(2)

m₁ = Weight of dry rubber sample.

ρ₁ = Density of the dry rubber.

m₂ = Weight of the solvent in the swollen sample.

ρ₂ = Density of toluene, 0.865 g/cm.³.

The cross-link density of vulcanized rubber without filler, the volume fraction of rubber in the swollen gel was given in equation (3) (Flory–Rehner equation)

v = [- 2 / (Φ V_{s})] [(I n (1 - Φ_{r}) + Φ_{r} + x Φ_{r}^{2}) / (Φ_{r}^{(1 / 3)} - Φ_{r} / 2)]

(3)

v = Crosslink density, mol/cm³ (total covalent and ionic cross-linking)

Φ_r = Volume fraction of rubber in equilibrium swollen sample

v_s = Mole volume of used solvent at room temperature, 106.2 cm³/mol.⁹

χ = Flory–Huggins polymer-solvent interaction parameter, 0.41.¹⁰

φ = Functionality of the cross-link point, normally assumed to be tetra-functional,

φ = 4.⁷

Self-healing test

Broken tear test specimen from each formulation was collected for self-healing test. In order to establish close contact and pressure, the two pieces were joined and compressed together in the sample cutting holder as shown in Figure 1(a) and 1(b). They were then placed in three different temperatures and durations to achieve self-healing; room temperature for 24 h, 120°C for 5 h, and 160°C for 1 h. The samples were cooled down for 12 h before conducting the tear test. The healing efficiency was evaluated from the comparison of tear strength between virgin and healed samples as shown in equation (4). Characterization tests including SEM and VMM were performed on the tear specimens including the crack area after self-healing test

S e l f - h e a l i n g e f f i c i e n c y, % = \frac{T e a r S t r e n g t h (A f t e r)}{T e a r S t r e n g t h (B e f o r e)} x 100 %

(4)

Attenuated total reflection using Fourier-transform infrared spectroscopy analysis

The chemical functionalities of the samples were analyzed by using ATR-FTIR between wavenumber of 500 and 4000 cm-1.

Scanning electron microscopy analysis

The dispersion and morphology of the virgin blend matrix were done on a scanning electron microscope of 10 kV accelerating voltage with magnification of 200x, 500x, 1000x, and 2000x. Before conducting SEM test, the samples were coated with a layer of platinum using automated platinum sputter coater. For the crack region of the healed sample, a magnification of 100x was used.

Manual video measuring machine analysis

Fracture surface of the samples with good healing ability after self-healing test was also observed under Manual Video Measuring Machine (VMM), using magnification of 1.0x. In order to get better visualization of the crack region, the microscopic images of the crack region using SEM and VMM were being compared.

Results and discussions

Rheological and mechanical properties analysis

Comparison graphs of vulcanizate properties of PEI on ESBR/BR blends are provided in Figure B1. From Table 3, it was an obvious marked difference in scorch time, s (ts₂), between lower and higher M_w PEI, and not much affected by the change in ratios of PEI. The control sample B, higher M_w BP3, and BP4 exerted much longer scorch time (≥100 s) compared to lower M_w BP1 and BP2, resulting in better processing safety as there is more handling time for operators and longer time for unvulcanized rubber to flow and fully fill the mold cavities before vulcanization takes place to avoid problem such as short-mold (non-filled) during compression molding. Interestingly, incorporating lower M_w PEI 800 and PEI 2000 in the ESBR/BR blends resulted in very fast scorch time, which was less than 60 s. It has been suggested that loss in scorch safety was primarily due to the elimination of degradative chain transfer reactions and low M_w PEI quickly reacted with the alkoxyl radical.^11,12

Table 3.

Summary of vulcanizate properties of polyethylenimine on emulsion styrene butadiene rubber /butadiene rubber blends.

Designation	Hardness, shore A	ML	MH	Torque difference, lbf·in	ts_2, s	tc_90, s	Tensile strength, MPa	Elongation at break, %	Tear strength, N/mm
B	69	0.74	27.39	27.11	100	275	0.92	38.28	2.62
BP1/5	63	0.62	16.05	15.14	37	330	1.37	41.54	2.49
BP1/7.5	58	0.51	15.16	14.37	38	281	1.05	38.74	2.40
BP1/10	58	0.49	14.75	13.91	37	314	1.14	49.00	2.44
BP1/12.5	51	0.58	9.09	8.70	36	265	1.04	43.36	2.48
BP1/15	47	1.13	5.74	4.58	43	263	1.10	46.39	2.33
BP2/5	62	0.62	26.58	26.33	37	327	1.22	37.05	2.67
BP2/7.5	62	0.52	21.51	20.34	38	320	1.18	33.32	2.45
BP2/10	60	0.55	21.64	21.00	39	319	1.00	33.76	2.88
BP2/12.5	52	0.58	16.50	15.70	40	274	0.81	35.42	3.01
BP2/15	47	0.47	12.61	12.27	43	262	1.12	41.14	2.87
BP3/5	71	0.77	29.40	28.41	122	298	1.07	37.03	3.04
BP3/7.5	73	0.71	30.26	29.63	112	317	1.12	41.38	3.39
BP3/10	75	0.73	28.42	27.71	113	303	0.96	42.22	3.05
BP3/12.5	73	0.71	26.38	25.15	105	309	0.93	42.80	3.37
BP3/15	71	0.69	24.78	23.71	109	303	1.17	42.19	2.70
BP4/5	76	0.92	30.22	29.03	126	300	1.05	26.63	3.00
BP4/7.5	78	0.83	29.88	29.79	115	317	1.10	33.32	3.04
BP4/10	75	0.81	28.38	27.31	110	305	1.00	35.31	2.96
BP4/12.5	73	0.78	27.50	26.35	116	302	1.09	44.26	3.05
BP4/15	75	0.88	27.22	26.63	114	309	1.03	44.67	2.80

It can be seen that BP1/12.5, BP1/15, BP2/12.5, and BP2/15 had shorter tc₉₀ compared to the others. The reason can be attributed to the faster reaction between the functional groups in PEI 800 and PEI 2000 and the presence of higher reactive groups (mainly double bonds and allylic hydrogens) in molecular structure of ESBR/BR blends to form cross-linking network. No significant change in tc₉₀ of rubber compound with the increasing content of PEI 25,000 and PEI 270,000 in BP3 and BP4, which was in line with their scorch time, ts₂.

Also, the trend of torque different of high M_w BP3 and BP4 was similar, increased to optimum values at 7.5phr and gradually decreased as the content increased. The torque different values dropped rapidly in BP1 and BP2 as the PEI content increased. Hardness of the rubber compound is also closely related to the torque differences. MH of BP3 and BP4 was significantly higher than BP1 and BP2, and it was supported by the fact that the hardness of BP3 and BP4 ranging from 71 to 78 shA, whereas BP1 and BP2 have hardness range of 47–63 shA. Drop in hardness indicates the decrease in crosslinking density in the rubber network in lower M_w PEI-800 and PEI-2000. Whereas for higher M_w PEI-25,000 and PEI-270,000, it might cause by the moisture-aided diffusion of the adhesive diffused into the rubber matrix during heating process,^13,14 which was followed by hardening of the compound. Thus, BP3 and BP4 had higher durometer hardness compared to control B, BP1, and BP2.

BP1, BP2, BP3, and BP4 showed an overall improvement in terms of tensile strength in Figure 2. BP1/5 compound achieved highest tensile strength of 1.37 MPa, about 48.9% increase compared to control sample B; followed by BP2/5 (1.22 MPa) with an increase of 32.6% in tensile strength. BP3 and BP4 showed slight improvement in terms of elongation at break at increasing content. However, BP1 and BP2 showed higher elongation at break at 5 phr, and fluctuated as the content increased. Vice versa, BP3 and BP4 showed better tear strength compared to BP1 and BP2, whereby BP3/7.5 achieved the optimum value of 3.39 N/mm, a 29.4% increase in tear strength compared to control B.

Interestingly, data showed that lower M_w PEI (PEI-800 and PEI-2000) induced better tensile strength, while higher M_w PEI (PEI-25,000 and PEI-270,000) induced better tear strength. The formation of Van der Waals forces between PEI molecules and the rubber molecules helped to improve the mechanical properties of the PEI incorporated ESBR/BR blends. It helped to resist flow when the stress was applied onto the rubber.

Self-healing and crosslink density analysis

Self-healing of PEI on ESBR/BR blends was investigated using three parameters: (1) room temperature (24°C) for 24 h, (2) 120°C for 5 h, and (3) 160°C for 1 h, and results are illustrated in Figure 2. By referring to Schussele et al.,³ self-healing experiment at room temperature for 24 h and 120°C for 5 h was repeated. At elevated temperature of 120°C, the duration of the test was shortened to 5 h instead of 12 h to avoid deterioration of the rubber matrix as heat aging condition that will affect the mechanical properties of the rubber. This would directly affect the result of self-healing efficiency.

In the absence of self-healing additive, control B did not exhibit any self-healing indication upon annealing and heating under three different temperatures. The remaining unreacted peroxide molecules during vulcanization would not further induce covalent crosslinking between two crack regions as control sample did not reattach after the aging test at 160°C for an hour. This is due to the irreversible covalent crosslink reaction. On the other hand, all specimens with various loadings of PEI did not possess self-healing ability at room temperature after annealing for 24 h nor 120°C for 5 h. The self-healing behavior was observed when the specimens with various loadings of PEI heated up to 160°C for 1 h under pressure. This suggested at higher temperature around 160°C, phase separation of PEI in rubber occurred to promote thermal self-healing in rubber.³

PEI is a highly branched network with a high cationic charge-density.¹⁵ Higher M_w PEI (PEI-25,000 and PEI-270,000) possessed better self-healing efficiency compared to lower M_w PEI. BP1/5 and BP2/10 had self-healing efficiency of 52.3% and 23.6%, respectively. BP4/5 successfully achieved the optimum self-healing efficiency of 71.5%. BP4/10 and BP4/15 exerted 49.4% and 65.2% self-healing efficiency, respectively. Whereas BP3 blends successfully exhibited self-healing efficiency of 25.7%, 59.8%, and 37.4%, at 5, 10, and 15 phr, respectively. This indicates that greater phase separation occurred between higher M_w viscous PEI with rubber matrix at crack region to promote thermal self-healing. The PEI molecules migrated to the surface of the crack region, and upon optimum temperature and pressure, self-healing could be achieved. From the result, higher M_w PEI provided higher adhesion strength to low surface energy rubber.

From Table 4, it was found that BP2 and BP3 increased the total crosslink density to a maximum, before decreasing as loading increased. BP4 showed 53% decrease in total crosslink density, from 0.478 x 10⁻³ mol/cm³ to 0.224 × 10⁻³ mol/cm³, as loading increased. Overall, low and high M_w PEI showed a trend of lower total crosslink density at 15 phr, which led to the drop in MH or hardness in the blends. Even though the reaction between functional group in PEI was fast at the given temperature, the amount of the peroxide decomposed decreased and thus led to lesser total crosslink density formed in the network. This was probably due to higher surface migration of PEI to the rubber surface.

Table 4.

Crosslink densities and self-healing efficiency of PEI in ESBR/BR blends.

Designation	Crosslink density, 10⁻³ mol/cm³	Self-healing efficiency,%
Designation	Crosslink density, 10⁻³ mol/cm³	24°C, 24 h	120°C, 5 h	160°C, 1 h
B	0.233	—	—	—
BP1/5	0.472	—	—	52.29
BP1/7.5	0.398	—	—	—
BP1/10	0.482	—	—	—
BP1/12.5	0.314	—	—	—
BP1/15	0.449	—	—	—
BP2/5	0.488	—	—	—
BP2/7.5	0.482	—	—	—
BP2/10	0.586	—	—	23.6
BP2/12.5	0.525	—	—	—
BP2/15	0.426	—	—	—
BP3/5	0.142	—	—	25.75
BP3/7.5	0.212	—	—	—
BP3/10	0.287	—	—	59.82
BP3/12.5	0.236	—	—	—
BP3/15	0.154	—	—	37.4
BP4/5	0.478	—	—	71.5
BP4/7.5	0.368	—	—	—
BP4/10	0.344	—	—	49.41
BP4/12.5	0.330	—	—	—
BP4/15	0.224	—	—	65.2

Characterization analysis

Attenuated total reflection using Fourier-transform infrared spectroscopy analysis

Figure 3 shows the FTIR spectrum comparison of B and BP1/5 samples. Strong ESBR/BR trans-isomer peak can be observed at 963–965 cm⁻¹, whereas cis-1,4 bond can be seen at the peak at 734 to 741 cm⁻¹ and the styrene group appeared at 757 cm⁻¹. The absorption band at 909 to 910 cm⁻¹ is the characteristic of cis-1,2 unit. Since alkanes have no functional groups, their C-H stretch bands normally appear around 2800 to 3000 cm⁻¹, whereas C-H bending normally appears at 1400 to 1500 cm⁻¹. In BP1, BP2, BP3, and BP4, the N-H bond stretch in amines (PEI) appears as weak to medium, broad band, in the range of about 3200–3600 cm⁻¹ with one spike. It can be seen that as PEI increased, the band grows stronger between 3200 to 3600 cm⁻¹, confirming the presence of secondary amine. The N-O stretch appears as weak to medium band in the range of 1556 cm⁻¹ for BP1, BP2, BP3, and BP4. This suggested the oxidation of tertiary amine with the alcohol formed from decomposition of DCP. Figure 4 shows the overall Fourier-transform infrared spectroscopy (FTIR) spectra of B, BP1, BP2, BP3, and BP4. Details of each FTIR spectrum are provided in Figure A1.

Scanning electron microscopy analysis

Scanning electron microscopy (SEM) images of the control B, BP1/5, BP3/10, BP4/5, BP4/10, and BP4/15 are shown in Figure 5 to observe the dispersion of the additives on the cross-sectional surface of the compounds. Only ESBR/BR/PEI blends with self-healing ability were tested for SEM. All samples showed uniform dispersion under 1000x. Agglomeration of PEI molecules can be seen under x5000 and x8000 in BP1/5, BP3/10, BP4/5, BP4/10, and BP4/15, confirming PEI molecules accumulated and assembled on the interface of the rubber chain. This suggested no modification in the rubber chain backbone, and ESBR/BR blends were phase separated with PEI molecules.³ Fracture surface of the samples with good healing ability was observed under VMM and SEM for better visualization of crack regions in Figure 6.

Proposed mechanism

In this study, PEI did not act as a coagent in peroxide vulcanization; therefore, only peroxide–peroxide (P–P) crosslinking reaction occurred as illustrated in Figure 7. Highly unsaturated BR and ESBR have very high crosslinking efficiency compared to other rubber due to higher concentration of allylic hydrogens;¹⁶ therefore, undesirable side reactions such as chain scission and disproportionation are unlikely to occur because these hydrogens are readily abstracted and efficiently converted to crosslinks. Kruzelak et al.¹² suggested that during peroxide crosslinking of ESBR and BR, the peroxide radicals could be reacted by hydrogen abstraction and addition reactions to the double bonds. The radicals are highly reactive and they are ready to react with more weakly bonded hydrogen (H) from the ESBR/BR blends via hydrogen abstraction. Since phase separation was examined by using SEM on thin sections of ESBR/BR blends with the PEI molecules, it is suggested only physical bond interaction such as Van der Waals holding both rubber chains and PEI together, without any changes in the rubber chain backbone, as illustrated in Figure 8.

Besides, oxidation of the tertiary amine in PEI had occurred during heating as illustrated in Figure 9. The amine attacked the OH bond from the alcohol formed during decomposition of DCP. The O⁻ in the PEI structure is weak, and elimination could occur during heating, forming a substitute hydroxylamine. This was supported by the N-O stretch observed in the FTIR spectrum earlier.

Conclusions

From the study, a marked difference was found in scorch time (ts₂) between lower (BP1 and BP2) and higher M_w PEI (BP3 and BP4) in ESBR/BR blends. BP3 and BP4 exerted much longer scorch time (≥100 s), whereby lower M_w PEI 800 and PEI 2000 in the ESBR/BR blends resulted in very fast scorch time, which were less than 60 s. BP1, BP2, BP3, and BP4 showed an overall improvement in terms of tensile strength. The formation of Van der Waals forces between PEI molecules and the rubber molecules helped to improve the mechanical properties of the PEI-incorporated ESBR/BR blends. The self-healing behavior was observed when the ESBR/BR/PEI blends were heated up to 160°C for 1 h under pressure. BP4/5 successfully achieved the optimum self-healing efficiency of 71.5%. From the SEM images, PEI molecules were found to be accumulated and assembled on the interface of the rubber chain. It is suggested only weak physical bond interaction such as Van der Waals and hydrogen bonds holding both rubber chains and PEI together, without any changes in the rubber chain backbone.

Footnotes

Acknowledgments

The authors would like to acknowledge and thank Faculty of Chemical Engineering, Universiti Teknologi Malaysia and Armstrong Electronics Malaysia for supporting the equipment in this research.

Author Contributions

“Writing—original draft preparation, Chuying, Siaw; writing—review and editing, Chuying, Siaw and Norfhairna Baharulrazi; supervision, Norfhairna Baharulrazi, Siti Hajjar Che Man and Norhayani Othman. All authors have read and agreed to the published version of the manuscript.”

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Ministry of Education, Malaysia, R.J130000.7851.5F294.

ORCID iD

Chuying Siaw

Appendix A

Figure A1.

(a) to (u) Spectra of 21 samples of control sample B and polyethylenimine incorporated emulsion styrene butadiene rubber/butadiene rubber blends. (a) B (b) BP1/5 (c) BP1/7.5 (d) Sample BP1/10 (e) Sample BP1/12.5 (f) Sample BP1/15 (g) Sample BP2/5 (h) Sample BP2/7.5 (i) Sample BP2/10 (j) Sample BP2/12.5 (k) Sample BP2/15 (l) Sample BP3/5 (m) Sample BP3/7.5 (n) Sample BP3/10 (o) Sample BP3/12.5 (p) Sample BP3/15 (q) Sample BP4/5 (r) Sample BP4/7.5 (s) Sample BP4/10 (t) Sample BP4/12.5 (u) Sample BP4/15.

Appendix B

Figure B1.

(a) to (g) Summary of vulcanizate properties of PEI on ESBR/BR blends. (a) Durometer hardness of various loadings of PEI on ESBR/BR blends. (b) Torque difference of various loadings of PEI on ESBR/BR blends. (c) Scorch time, ts₂, of various loadings of PEI on ESBR/BR blends. (d) Curing time, tc₉₀, of various loadings of PEI on ESBR/BR blends. (e) Tensile strength of various loadings of PEI on ESBR/BR blends. (f) Elongation at break, %, of various loadings of PEI on ESBR/BR blends. (g) Tear strength, N/mm, of various loadings of PEI on ESBR/BR blends. PEI = polyethylenimine; ESBR/BR = emulsion styrene butadiene rubber/butadiene rubber

References

Yemuland

Imae

. Synthesis and characterization of poly(ethylenimine) dendrimers. Colloid Polym Sci 2008; 286(6–7): 747–752.

Lungu

Diudea

Putz

Grudziński

Linear and branched PEIs (Polyethylenimines) and their property space. Int J Mol Sciva 2016; 17: 555–566.

Schussele

Nubling

Thomann

, et al. Self-healing rubbers based on NBR blends with hyperbranched polyethylenimines. Macromol Mater Eng 2012; 297: 411–419.

Wang

Yuan

Han

. Influences of hyperbranched polyethylenimine on the reactive compatibilization of polycarbonate/polyamide blends. Chin J Polym Sci 2015; 33(4): 652–660.

Mruk

Nebhani

. Rubber comprised of product of diene-based elastomer, branched polyethylenimine oligomer and reinforcing filler, and tire with component. USA: Patent, 2018. 10,005,896 B2.

Liu

. The effect of network structure on compressive fatigue. Advances in Materials Science and Engineering 2020; 2020: 1–9.

Mok

Eng

. Characterization of crosslinks in vulcanized rubbers: from simple to advanced techniques. Malays J Chem 2018; 20(1): 118–127.

Blume

Kiesewetter

. Determination of the crosslink density of tyre tread compounds by different analytical methods. Rohstoffe und anwendungen, raw materials and applications. KGK. http://www.kgk-rubberpoint.de (2019, accessed 12 October 2021).

Cao

Lin

et al. Design of self-healing supramolecular rubbers by introducing ionic cross-links into natural rubber via a controlled vulcanization. Appl Mater Inter 2016; 8(27): 17728–17737.

10.

Traeger

Castonguay

. Effect of c-radiation on the dynamic mechanical properties of styrene–butadiene rubbers. J Appl Polym Sci 1966; 10: 491–509.

11.

Henning

. The use of coagents in the radical cure of elastomers. In: Proceedings of the 56th International Wire and Cable Symposium, Lake Buena Vista, FL, 11–14 November 2007.

12.

Kruželák

Kvasničáková

Hudec

. Peroxide curing systems applied for cross-linking of rubber compounds based on SBR. Adv Ind Eng Polym Res 2020; 3(3): 120–128.

13.

Takemura

Tomita

Mizumachi

. Dynamic mechanical properties and adhesive strengths of epoxy resins modified with liquid rubber I: modification with ATBNJ. Appl Polym Sci 1985; 30(10): 4031–4043.

14.

Ratna

Banthia

. Toughened epoxy adhesive modified with acrylate based liquid Rubber. Polym Int 2000; 49(3): 281–287.

15.

Von Harpe

Petersen

, et al. Characterization of vommercially available and synthesized polyethylenimines for gene delivery. J Control Release 2000; 69(2): 309–322.

16.

Akiba

Hashim

. Vulcanization and crosslinking in elastomers. Prog Polym Sci 1997; 22(3): 475–521.