Surface modification of thermally expandable microspheres for enhanced performance of disbondable adhesive

International Journal of Adhesion & Adhesives 66 (2016) 33–40

Contents lists available at ScienceDirect

International Journal of Adhesion & Adhesives journal homepage: www.elsevier.com/locate/ijadhadh

Surface modification of thermally expandable microspheres for enhanced performance of disbondable adhesive Yuchen Lu n, James Broughton, Pat Winfield Oxford Brookes University, UK

art ic l e i nf o

a b s t r a c t

Article history: Accepted 20 October 2015 Available online 14 December 2015

This research successfully improved the compatibility between thermally expandable microspheres (TEMs) and an adhesive system for enhanced performance. TEMs were grafted with poly(glycidyl methacrylate) (PGMA) chains via atom transfer radical polymerisation (ATRP) with activators regenerated by electron transfer (ARGET) or ARGET ATRP technique. The temperature effect on the surface modification of TEMs was investigated for an optimum modification condition. Compared to adhesive incorporating unmodified TEMs, up to 15.8% increase in tensile lap shear strength and 24.0% increase in ultimate tensile strength (UTS) were achieved. Most notably, after environmental conditioning, the UTS of the adhesive system containing modified TEMs was 8.0% higher than the strength of unmodified TEMs before environmental conditioning. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Thermally expandable microspheres Adhesive disbonding Surface Treatment Epoxides

1. Introduction End-of-life vehicle (ELV) legislations and recycling issues for future multi-material vehicles necessitate the development of new joining solutions that enable rapid disassembly for automotive vehicle maintenance or recycling scenarios [1]. In the search for a disbondable adhesive system suitable for automotive applications, an adhesive system containing thermally expandable microspheres (TEMs) has been identified as one of the most promising approaches with satisfactory disbonding performance observed in previous research conducted within the Joining Technology Research Centre (JTRC) at Oxford Brookes University [2]. TEMs are micron scaled spherical particles comprised of a thermoplastic shell material (2–15 μm thickness) and an encapsulated hydrocarbon core which has a low boiling temperature. At elevated temperature, shell softening and hydrocarbon gasification allows TEMs to expand 30–80 times in volume, which can lead to fracture of the adhesive bond line. The major limitation of the TEMs approach is the poor compatibility between the TEMs and the adhesive which weakens the mechanical properties and long term durability of the original adhesive system [1,3]. A suitable surface modification technique was selected to especially improve the TEMs/adhesive compatibility n Correspondence to: Department of Mechanical Engineering and Mathematical Sciences (MEMS), Oxford Brookes University, Wheatley Campus, Wheatley, Oxford OX33 1HX, UK. Tel.: þ 44 1865 483 568. E-mail address: [email protected] (Y. Lu).

http://dx.doi.org/10.1016/j.ijadhadh.2015.12.007 0143-7496/& 2015 Elsevier Ltd. All rights reserved.

while retaining satisfactory in-service disbonding effectiveness. Surface grafting technology is a successful means of producing strong covalent bonds between TEMs and adhesive matrix. ‘Grafting to’ and ‘grafting from’ are two common approaches [4]. In the ‘grafting to’ approach, end-functionalised groups (e.g. hydroxyls, epoxides and thiols) react with a suitable surface to form a covalent bond. However, limitations such as low grafting density or difficulties in the synthesis of suitable functionalised polymers restrict its application. The ‘grafting from’ approach, in which polymer chains grow from the flat/spherical surface (Surface Initiated Polymerisation ‘SIP’), shows wider applicability. Atom transfer radical polymerisation (ATRP) has been demonstrated as a successful SIP technique in various areas including cell adhesion [5,6], drug release [7], anti-biocorrosion coatings [8], and silicon wafers [9]. First discovered in 1995 [10,11], ATRP provides a simple and controlled polymerisation solution with good control of molecular weight, structure and a high degree of chain end functionality. ATRP with activators regenerated by electron transfer or ARGET ATRP offers two major advantages over conventional ATRP technique: a) only ppm amount of catalyst is required which makes the purification or disposal process of toxic components cheaper or sometimes unnecessary, especially for industrial application [12]; b) a higher tolerance to oxygen and a catalyst that can be added in oxidatively stable state [13]. Inspired by the pioneering study conducted by Jonsson and coworkers [14], this study investigated TEMs that were modified by growing poly(glycidyl methacrylate) (PGMA) chains from the surface of TEMs via the ARGET ATRP process. Since each GMA

34

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

molecule contains one epoxide group, PGMA chains contain an amount of n epoxide groups (n denotes the degree of polymerisation). Modified TEMs containing epoxide groups can react with the hardeners during the curing process of epoxy adhesive forming covalent bonds and thereby strengthening the TEMs/ adhesive composite.

2. Experimental test programme The experimental test programme was divided into three stages: a) surface modification of TEMs; b) investigation of disbonding performance of TEMs/adhesive joints; c) investigation of bulk properties of TEMs/adhesive composite.

2.1. Materials In this research, 920DU120HEMA grade of TEMs which contains 15 wt% 2-hydroxyethyl methacrylate on the thermoplastic shell material, were kindly provided by Expancel Akzo Nobel. The particle size distribution ranges from 28 to 38 mm. This grade of TEMs starts to expand at around 125 °C and reaches the maximum expansion at around 190 °C [14]. 3M Scotch-Weld 9323, a two component amine cured epoxy adhesive typically used in automotive structural applications, was chosen as the adhesive matrix. This is a standard high strength structural adhesive with a tensile lap shear strength of 30 MPa and high environmental resistance. Information about all other chemicals used in this research was detailed in Section 2.2. Fig. 1. Schematic illustration of TEMs surface modification process.

2.2. Experimental procedures TEMs were modified at different conditions and incorporated into adhesive system for the investigation of disbonding performance and mechanical properties.

Table 1 Chemicals used during immobilisation process. Chemical name

Function

Quantities

Supplier/grade

Acid treated TEMs Triethylamine 4DMAP BIBB Dichloromethane

N/A Reducing agent Catalyst ATRP initiator Solvent

5g 5g 12.6 mg 11 g 50 ml

N/A Sigma-Aldrich ( Z99%) Sigma-Aldrich ( Z99%) Sigma-Aldrich (98%) Fisher Scientific ( Z99%)

2.2.1. Surface modification of TEMs This surface modification approach is an adaptation of the recent research conducted by Jonsson and co-workers [14]. The whole process was mainly divided into two steps, as illustrated in Fig. 1.

Table 2 Chemicals used during grafting process. Chemical name

Function

Quantities

Supplier grade

Immobilised TEMs Toluene CuBr2 GMA PMDETA L-Ascorbic acid

Macroinitiator Solvent Catalyst Grafting monomer ATRP Ligand Reducing agent

3g 30 ml 2.52 mg 32.4 g 47.1 mg 50 mg

N/A Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich Sigma-Aldrich

(98%) (99%) (97%) (99%) ( Z99%)

Fig. 3. Illustration of adhesive tensile bar specimen.

Fig. 2. Dimensions of single-lap-joint specimen.

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

Fig. 4. FTIR spectra of modified TEMs at wavelength ranges 1757–1677 cm  1.

35

Fig. 7. FTIR spectra of modified TEMs at wavelength ranges 780–720 cm  1.

Fig. 8. Molecular structure of GMA. Table 3 Wavelength ranges of major differences and main assignments. Wavelength ranges (cm  1) Main assignments 1757–1677 1325–1220

Fig. 5. FTIR spectra of modified TEMs at wavelength ranges 1325–1035 cm  1.

Fig. 6. FTIR spectra of modified TEMs at wavelength ranges 1010–810 cm  1.

1220–1035 1035–930 930–883 883–780 780–720

C¼ O stretching –CH in-plane deformation; epoxide ring vibration C–O–C stretching C–O stretching Epoxide ring vibration Epoxide ring vibration Epoxide ring vibration

a) Immobilisation process This first step was to immobilise 2-Bromoisobutyryl bromide (BIBB) onto the TEM surface forming initiation sites. Prior to immobilisation, the TEMs were treated with 10 wt% sulphuric acid solution to remove magnesium hydroxide contained in the TEMs powder. After thorough rinsing and filtering, TEMs were dried at 60 °C for 8 h. The chemical reaction involved in the immobilisation process is highly exothermic and requires extra precautions. Therefore, BIBB was first diluted with a certain amount of dichloromethane and then added dropwise to the TEM solution which contained dichloromethane as a solvent, triethylamine as a reducing agent and 4-(Dimethylamino)pyridine (4DMAP) as a catalyst. After immobilisation, TEMs were washed with dichloromethane and dried at 50 °C for 1 h. The chemical quantities used in the immobilisation process are listed in Table 1. b) Grafting process (ARGET ATRP) The grafting process was to grow the PGMA chain from the surface of TEMs using α-bromo ester as initiation sites. N,N,N0 ,N00 ,N00 Pentamethyldiethylenetriamine (PMDETA), Glycidyl methacrylate (GMA) and Copper(II) bromide was added to toluene solution in a

36

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

Fig. 9. SEM micrograph of unmodified TEMs.

Fig. 10. SEM micrograph of TEMs modified at 30 °C for 3 h.

Fig. 11. SEM micrograph of TEMs modified at 30 °C for 5 h.

rubber septa sealed container. The container was deoxygenated by bubbling nitrogen gas for 5 min. Polymerisation was initiated by adding TEMs and L-ascorbic acid and heating up to certain temperatures as specified in Section 2.2.2. After polymerisation, the TEMs were washed with dichloromethane and dried at 50 °C in an oven for 5 h. See Table 2 for chemical quantities used in this step.

2.2.2. Investigation of temperature–time effect on the surface modification of TEMs To grow PGMA chains from the surface of TEMs' while retaining their expansion properties, key parameters were carefully controlled.

The effect of temperature was investigated to optimise the ARGET ATRP system for better TEMs surface modification. TEMs were modified at three different temperatures: 23 °C, 30 °C and 37 °C. The expansion property of TEMs is mainly determined by the hydrocarbon content confined within the outer shell material. Although the shell material generally has good chemical resistance, it was shown in the literature [14] that long time immersion in certain solvents could plasticise or even degrade the shell material which would result in the leakage of hydrocarbon. Therefore, for each temperature set, TEMs were modified for 3 h and 5 h and characterised for changes. All other parameters remained unchanged. In total, six variations of modified TEMs (two time sets and

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

37

Fig. 12. SEM micrograph of TEMs modified at 37 °C for 3 h.

Fig. 13. SEM micrograph of TEMs modified at 37 °C for 5 h.

Fig. 14. SEM micrograph of bulk materials formed during TEMs modification at 37 °C for 5 h.

three temperature sets) were produced. Attenuated total reflection (ATR) FTIR spectroscopy and SEM imaging were employed for characterisation of the TEMs at this stage. 2.2.3. Investigation of disbonding performance of TEMs/adhesive joints Due to its simple configuration and wide use in standard testing (e.g. BS EN 15336:2007, BS EN 1465:2009), the single lap joint (SLJ) configuration was chosen to evaluate the modified adhesive system. A simple and well controlled method was proposed for the evaluation of disbonding performance evaluation, detailed in Section 2.3. Adhesive SLJs containing TEMs (both modified and unmodified) were examined for changes in tensile lap shear strength and residual strength after heat activation. The

effect of the TEMs concentration on disbonding performance was investigated, though not presented in this paper and the best acting TEMs concentration was found to be 20 wt%, which was adopted throughout this study [15]. 2.2.4. Investigation of bulk performance of TEMs/adhesive composite Based on the results from Section 2.2.3, the bulk performance was investigated in tensile and DMTA tests for both unconditioned and environmentally conditioned specimens. 2.3. Experimental test methods Thermally expandable microspheres (different grades and modified microspheres) were characterised using a PerkinElmer

38

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

Spectrum FTIR Spectrum 65 equipped with ATR accessory. The wavelength ranged from 600 cm  1 to 4000 cm  1. For each sample, 32 scans were obtained with a resolution of 4 cm  1. SEM imaging of the project was conducted using a Jeol 5510LV (low vacuum). Samples were gold coated using a Cressington Sputter Coater 208HR for 90 s at 0.6 kV and 18 mA. The electron beam energy was 20 kV. Tensile lap shear tests were conducted according to BS EN 1465:2009. SLJ specimens were prepared using pre-cut EN AW6082 T6 aluminium substrates. Dimensions of the SLJ are shown in Fig. 2. A Speed Mixer (DAC 150-FVR-K) was used to mix the adhesive with the TEMs at 3000 rev/min for 60 s. Bond line thickness was controlled to be around 0.2 mm using solid glass spheres. The specimens were cured at 90 °C for 90 min. An Instron 5582 tensile testing machine with 100 kN load cell was used to test the SLJs using a displacement rate of 1 mm/min. All tests were conducted at (23 74) °C and (30 75)% relative humidity. The disbonding effectiveness tests required the SLJ specimens to be heated up in the oven from room temperature to 270 °C and held for 1 min to ensure the same temperature was achieved in the bond line. 270 °C was established as the temperature to achieve optimum disbonding performance in previous research within JTRC [15]. After heat activation, test specimens were allowed to cool down to ambient temperature prior to test. The residue tensile lap shear strength was then measured. The reduction in strength, in percentage, was calculated for each formulation as a reference for the assessment of disbonding effectiveness, as described by Eq. (1). Disbonding effectiveness ¼

τSLJ  τ0SLJ τSLJ

ð1Þ

where τSLJ represents the tensile lap shear strength for SLJs before heat activation, and τ0SLJ represents the residual tensile lap shear strength for SLJs after heat activation. Bulk tensile tests were performed using the Instron 5582 in accordance with BS EN ISO 527-2:2012. Schematic illustration and dimensions of the specimens are shown in Fig. 3. Mixing and curing process followed the same procedures in SLJ specimen preparation.

DMTA was carried out with Tritec 2300 DMA using the dual cantilever configuration. The specimens were moulded and cut into rectangular prisms of 2 mmn20 mmn4 mm. Frequency was defined as 1 Hz. Glass transition temperature Tg was determined at the peak tanδ point. The environmental conditioning was designed to evaluate the adhesive performance subject to hygrothermal ageing conditions. Both bulk tensile and DMTA test specimens were immersed in deionised water at 40 °C for 14 days, followed by oven heating at 80 °C for 4 days. All specimens were conditioned at 23 °C and 30– 40% humidity for one day prior to test.

3. Results and discussion 3.1. Temperature–time effect on the surface modification of TEMs FTIR test of the modified TEMs showed major changes occurred at the 7 peaks shown in Figs. 4–7. These changes can be correlated with the molecular structure of GMA, Fig. 8. Main assignments of these peaks were listed in Table 3. The increases of the three peaks at 904 cm  1, 847 cm  1, and 761 cm  1 confirmed the presence of the grafted epoxy rings on the modified TEMs. These changes indicate that PGMA chains were successfully grafted from the surface of TEMs. The intensities of all 7 peaks generally followed the same trend: higher modification temperatures correspond to higher peak intensities. It is clear that temperature has a dramatic effect on the growth of the PGMA chains on the TEMs. This is consistent with other relevant studies regarding the temperature effect in ARGET ATRP process and can be correlated with the Arrhenius empirical equation [16,17]. It is also interesting to note that the increase of peak intensity from 30 °C to 37 °C was more substantial than from 23 °C to 30 °C. The difference of peak intensity between 23 °C 3 h and 23 °C 5 h was insignificant compared with the differences at 37 °C and 30 °C as a result of reaction time. This indicated that the rate of polymerisation was more accelerated at 37 °C. Recent research on surface initiated ARGET ATRP of PHEMAb-PDMAEMA copolymers from silicon wafers showed very similar acceleration at a temperature of 40 °C [16]. TEMs modified at 30 °C and 37 °C for 3 h and 5 h were characterised by SEM and shown in Figs. 9–14. A general trend was observed that for an increase in polymerisation temperature and time, an increasing amount of small particulate deposits was found attached to the surfaces of TEMs. For TEMs modified at 37 °C 5 h specifically, large agglomerates were observed under the microscope and more particulate deposits were found attached to the surface of TEMs. Large deposits formed during TEMs modification at 37 °C 5 h, see Figs. 13 and 14. This might be attributed to the following two reasons:

 Once a certain chain length was achieved, the polymer chains would intertwine with each other forming larger particulates;

 At a certain chain length, interchain termination also led to the Fig. 15. Photograph of bulk materials formed during surface modification of TEMs.

formation of larger particulates.

Table 4 Summary of tensile lap shear strength and disbonding effectiveness tests results.

Tensile lap shear strength (MPa) Residual strength (MPa) Disbonding effectiveness (%)

Pure adhesive

Unmodified TEMs

Modified at 30 °C 3 h

Modified at 30 °C 5 h

Modified at 37 °C 3 h

30 7 0.6 23 7 0.2 23

197 0.4 37 0.2 84

217 0.2 37 0.4 86

217 0.2 4 70.6 81

227 0.1 37 0.4 86

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

It should also be noted that the two reasons above can have a combined effect leading to large agglomerations or gel formation, as was observed for TEMs modified at 37 °C for 5 h, see Fig. 15. It is therefore suggested that 37 °C 5 h might not be a suitable polymerisation condition and an optimum chain length could be determined for the best TEMs surface modification. 3.2. Effect of TEMs surface modification on the disbonding performance of TEMs/adhesive SLJs Based on the results from FTIR and SEM, three variations of modified TEMs (30 °C 3 h, 30 °C 5 h and 37 °C 3 h) were selected and evaluated for TEMs/adhesive SLJs disbonding performance. The modified TEMs exhibited a higher tensile lap shear strength compared with the unmodified TEMs. As shown in Table 4, the greatest improvement of 15.8% was achieved using the 37 °C 3 h process. This corroborated with the changes from FTIR and SEM Table 5 Summary of bulk tensile test results. Formulation

Condition

Ultimate tensile strength (MPa)

Pure adhesive

Unconditioned Environmentally conditioned Unconditioned Environmentally conditioned Unconditioned Environmentally conditioned

417 0.2 387 0.5

Modified TEMs incorporation Unmodified TEMs incorporation

317 0.7 277 1.2 257 0.6 197 0.5

Fig. 16. Ultimate tensile strength of adhesive incorporated with 37 °C 3 h modified TEMs in comparison with pure adhesive and unmodified TEMs system.

39

characterisation. In addition, the difference of improvement between 30 °C 3 h and 37 °C 3 h variations indicated that TEMs grafted with longer PGMA chains are likely to deliver greater strength. However, as was observed during the modification process, gel formation (Fig. 15) was unavoidable when PGMA chains grew beyond a certain length. This inferred that an optimum length of PGMA chains could be determined. The changes in residual strength and disbonding effectiveness were insignificant. Compared with the TEMs/adhesive system before modification, the marginal increase in residual strength and the slight decrease in the disbonding effectiveness of TEMs/ adhesive system after modification were attributed to the reduction of TEMs' expansion property. Although TEMs generally have good chemical resistance, long time immersion in solvents during surface modification can plasticise or even degrade the TEMs' shell composition leading to hydrocarbon leakage which can affect TEMs' expansion property. It was for this reason that the adhesive system incorporated with 5 h modified TEMs (30 °C 5 h) showed the highest residual strength and lowest disbonding effectiveness Table 6 Summary of DMTA test results. Formulation

Condition

Peak tan delta Tg (°C)

Pure adhesive

Unconditioned Environmentally conditioned Unconditioned Environmentally conditioned Unconditioned Environmentally conditioned

0.583 7 0.028 0.5577 0.003

937 0.2 917 0.5

0.5617 0.008 0.4987 0.009

967 0.7 907 0.2

0.520 7 0.001 0.5007 0.003

957 0.7 907 0.7

Modified TEMs incorporation Unmodified TEMs incorporation

Fig. 18. Intensity of peak tan delta of adhesive incorporated with 37 °C 3 h modified TEMs in comparison with pure adhesive and unmodified TEMs system.

Fig. 17. Schematic illustration of covalent bonds formed between modified TEMs (grafted with PGMA chains) and adhesive.

40

Y. Lu et al. / International Journal of Adhesion & Adhesives 66 (2016) 33–40

whereas 3 h modified TEMs (30 °C 3 h and 37 °C 3 h) showed similar strength. 3.3. Effect of TEMs surface modification on the bulk properties of TEMs/adhesive composite material The improvements on bulk properties are shown in Table 5. The modified TEMs adhesive system achieved an UTS of 31 MPa which was 24.0% higher than that of the unmodified TEMs before environmental conditioning. It is very interesting to point out that the UTS of modified TEMs system post environmental conditioning was still 8.0% higher than the unmodified TEMs system before environmental conditioning, Fig. 16. It is likely that the modified TEMs grafted with long PGMA chains can extend well into the adhesive system and form covalent bonds by reacting epoxy groups with the Part A amine curing agent of the two component Scotch-Weld 9323 epoxy adhesive. A schematic illustration of adhesive incorporated with modified TEMs was shown in Fig. 17. The results from the DMTA tests in Table 6 further support the success of TEMs surface modification and indicated the strengthening mechanism of the modified TEMs system. The PGMA chains, containing numerous epoxy groups contributed to a more crosslinked network which also led to the increase in glass transition temperature. For the unconditioned specimens, the adhesive system with 37 °C 3 h modified TEMs showed a 7.9% increase in peak tan delta value (0.561) compared with the unmodified TEMs (0.520), which indicated a higher damping capacity, see Fig. 18. This could be mainly attributed to the following reasons:

 Modified TEMs had a better interaction with adhesive matrix  

which led to increased TEMs/adhesive friction and higher energy dissipation; Energy can be transferred via grafted PGMA chains and consumed by TEMs; The strong covalent bonds between TEMs and adhesive also intensified chain motions locally near the TEMs/adhesive interface local area which resulted in more adhesive chain friction and therefore a better dissipation of energy.

It is worth noticing that the introduction of PGMA chains into the adhesive system would have consumed a certain amount of the curing agent. It is therefore likely that the adhesive system with modified TEMs incorporation could achieve further improvement in mechanical performance if extra curing agents were added.

4. Conclusions It has been demonstrated that TEMs can be modified using the ARGET ATRP technique whilst retaining the equivalent expansion performance. It has also been shown that temperature can significantly accelerate the rate of polymerisation and hence the length of PGMA chains, which also increased with polymerisation time. Grafted chains, allowed growing beyond a certain length, incurred gel formation. It was suggested for future work that an optimum chain length could be determined. Surface modification of the TEMs substantially improved the compatibility with the

adhesive via the formation of strong covalent bonds. It was also shown that the modified TEMs system obtained better moisture resistance. Most notably, the ultimate tensile strength of the adhesive system containing modified TEMs after environmental conditioning was higher than the strength of adhesive system containing unmodified TEMs before environmental conditioning.

Acknowledgements The authors would like to express their sincere gratitude to Professor Fernando Audebert for his invaluable time and kind help with scanning electron microscopy at Begbroke Science Park, Oxford University. AkzoNobel Expancel was also gratefully acknowledged for the supply of TEMs samples.

References [1] Lu Y, Broughton J, Winfield P. A review of innovations in disbonding techniques for repair and recycling of automotive vehicles. Int J Adhes Adhes 2014;50:119–27. [2] McCurdy R. Disassembly methods for automotive structures (Ph.D. thesis). Oxford: Oxford Brookes University; 2011. [3] Banea MD, da Silva LFM, Carbas RJC. Debonding on command of adhesive joints for the automotive industry. Int J Adhes Adhes 2015;59:14–20. [4] Zheng G, Stover HDH. Living/controlled grafting from polymer microspheres. Chin J Polym Sci 2003;21(6):639–51. [5] Barbey R, Lavanant L, Paripovic D, Schuwer N, et al. Polymer brushes via surface-initiated controlled radical polymerisation: synthesis, characterisation, properties, and applications. Chem Rev 2009;109(11):5437–527. [6] Bencherif SA, Gao H, Srinivasan A, Siegwart DJ, Hollinger JO, Washburn NR, Matyjaszewski K. Cell-adhesive star polymers prepared by ATRP. Biomacromolecules 2009;10(7):1795–803. [7] Vasani RB, McInnes SJ, Cole MA, Jani AM, Ellis AV, Voelcker NH. Stimulusresponsiveness and drug release from porous silicon films ATRP-grafted with poly(N-isopropylacrylamide). Langmuir 2011;27(12):7843–53. [8] Yuan SJ, Pehkonen SO, Ting YP, Neoh KG, Kang ET. Inorganic–organic hybrid coatings on stainless steel by layer-by-layer deposition and surface-initiated atom-transfer-radical polymerization for combating biocorrosion. ACS Appl Mater Interfaces 2009;1(3):640–52. [9] Zhu B, Edmondson S. Polydopamine-melanin initiators for surface-initiated ATRP. Polymer 2011;52(10):2141–9. [10] Wang J, Matyjaszewski K. Controlled/"living" radical polymerization. atom transfer radical polymerization in the presence of transition-metal complexes. J Am Chem Soc 1995;117(20):5614–5. [11] Kato M, Kamigaito M, Sawamoto M, Higashimura T. Polymerisation of methyl methacrylate with the carbon tetrachloride/dichlorotris-(triphenylphosphine) ruthenium(ii)/methylaluminum Bis (2, 6-di-tert-butylphenoxide) initiating system: possibility of living radical polymerisation. Macromolecules 1995;28 (5):1721–3. [12] Shen Y, Tang H, Ding S. Catalyst separation in atom transfer radical polymerization. Prog Polym Sci 2004;29(10):1053–78. [13] Braunecker WA, Matyjaszewski K. Controlled/living radical polymerisation: features, developments, and perspectives. Prog Polym Sci 2007;32(1):93–146. [14] Jonsson M, Nyström D, Nordin O, Malmström E. Surface modification of thermally expandable microspheres by grafting poly(glycidyl methacrylate) using ARGET ATRP. Eur Polym J 2009;45(8):2374–82. [15] Lu Y. An evaluation and development of bonding technologies for rapid disassembly of automotive vehicles (Ph.D. thesis). Oxford: Oxford Brookes University; 2015. [16] Zhu B. Surface initiated polymerisation for applications in materials science (Ph.D. thesis). Loughborough: Loughborough University; 2012. [17] Tanaka K, Matyjaszewski K. Copolymerisation of (meth)acrylates with olefins using activators regenerated by electron transfer for atom transfer radical polymerisation (ARGET ATRP). Macromol Symp 2008;261(1):1–9.