Fast underwater bonding to polycarbonate using photoinitiated cyanoacrylate

ARTICLE IN PRESS International Journal of Adhesion & Adhesives 30 (2010) 208–213 Contents lists available at ScienceDirect International Journal of Adhesion & Adhesives journal homepage: www.elsevier.com/locate/ijadhadh Fast underwater bonding to polycarbonate using photoinitiated cyanoacrylate William E. Cloete, Walter W. Focke à Institute of Applied Materials, Department of Chemical Engineering, University of Pretoria, Lynnwood Road, Pretoria 0001, South Africa a r t i c l e in f o a b s t r a c t Article history: Accepted 13 January 2010 Available online 25 January 2010 Rapid underwater bonding of clear polycarbonate to metal or plastic substrates at temperatures approaching 0 1C was studied. Bonding was achieved within minutes using ethyl 2-cyanoacrylate gel cured using the photoinitiator (dibenzoylferrocene) with a blue-LED light source. The optimum initiator concentration varied from 0.3% to 0.1 wt % for adhesive films 0.5 to 1.2 mm thick, respectively. The polymerisation rate shows a negative temperature dependence making it highly suitable for cold environments. The ultimate shear strength of the bonds was temperature independent and ranged from 1 MPa for metallic to 5 MPa for plastic substrates, respectively. & 2010 Elsevier Ltd. All rights reserved. Keywords: Cyanoacrylates Destructive testing Thermal analysis Cure/hardening Plastics 1. Introduction Neat cyanoacrylate reacts rapidly when it comes in contact with water. It is therefore not generally used for underwater bonding. However, the cured resin is not much affected by exposure to water. Consequently, this adhesive finds interesting underwater applications including tagging of sea mammals [1] and mussels or scallops [2] and fixing coral to rocks underwater [3,4]. Numerous other applications are described in the patent literature [5–7]. This study considered the use of a commercial ethyl cyanoacrylate-based adhesive for underwater bonding of clear polycarbonate sheets to other substrates for use with a proprietary applicator system. The key idea of the applicator is to use the adhesive itself to rapidly displace any water present between the two substrates to be bonded. This is achieved by extruding the adhesive from a central application hole. The water displacement is best achieved using a high-viscosity adhesive. Suitable thickening agent, e.g. fumed silica, can be used to impart the required consistency. Previous studies [5] showed that the strongest bonds are achieved with thin adhesive layers. Thicker adhesive sections take longer time to cure and sometimes full cure is not achieved. Since the present application requires rapid bonding of a clear polycarbonate sheet to flat underwater substrates, photoinitiated cure was a definite option. Photoinitiated curing of acrylate adhesives is well established [8–10] and commercial systems are available [11]. The advantage is that photoinitiated cure can be initiated on demand and that thicker adhesive sections polymerise to completion. Thus this study considered the addition of an à Corresponding author. Tel.: + 27 012 420 2588; fax: + 27 012 420 2516. E-mail address: walter.focke@up.ac.za (W.W. Focke). 0143-7496/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijadhadh.2010.01.002 anionic photoinitiator to the cyanoacrylate adhesives to act as the primary cure initiator. Ferrocene is a transition metal complex that has been the subject of numerous investigations [12]. The photochemical characteristics of ferrocene and several of its derivatives have been studied exhaustively [13–15]. Kutal and Yamaguchi [16] identified dibenzoylferrocene as a preferred anionic photoinitiator for cyanoacrylates. It is insoluble in water and has a maximum absorption peak at 485 nm. This closely matches the 470 nm blue LED light sources used in underwater camphorquinone initiated acrylate systems [8–10]. This light source was therefore selected for used in the present study. 2. Materials and methods 2.1. Adhesives and substrates A commercial ethyl cyanoacrylate gel (Loctite 454) was used without modification. Dibenzoylferrocene was supplied by Sigma Aldrich Chemicals and used without further purification as a solution in hexane. Sea water was simulated using the aquarium product ‘‘Ocean Fish’’ supplied by Prodac. All other chemicals used were obtained from Sigma Aldrich. Metal sheets (hot-rolled mild steel sheet (12% carbon), stainless steel 304, and aluminium) were cut into squares measuring 10  100  100 mm in accordance with ASTM D 4501. The bonding surfaces of the different metal substrates were sanded with 100 grit sandpaper and then cleaned with a lint-free cloth soaked in isopropyl alcohol. The sulphuric-acid anodised aluminium was used as supplied. All thermoplastic sheets were supplied by Maizeys. The fibreglassreinforced polyester (FRP) sheet was cast using glass fibre and polyester resin supplied by Plastocure. ARTICLE IN PRESS W.E. Cloete, W.W. Focke / International Journal of Adhesion & Adhesives 30 (2010) 208–213 2.2. Light sources and calibration The light source unit was custom-designed and consisted of 144 LED’s in a 12  12 array. Nichia Corporation (Model: NSPB 500S) 5 mm blue ultra-bright LED’s were used. The whole system was sealed inside an acrylic box to allow underwater illumination of the substrates. The light source was powered by a Vanson Deluxe Universal Regulated DC Power Supply (Model RC-1200). The output was set at 12 V DC and the overall current measured was 750 mA. The light source was calibrated by the National Metrology Institute of South Africa (NMISA) by comparing the spectral irradiance of the blue light LED array against the spectral irradiance of a standard lamp, traceable to the national measuring standard for spectral irradiance. In this investigation the adhesive was illuminated with the light source placed at a distance of 50 mm away. This distance was chosen to ensure overlapping of the individual LED light beams and to allow sufficient space to fit the experimental equipment between the light source and the adhesive samples. All adhesive samples were irradiated at an effective light intensity of 5 mW/cm2. 2.3. Photodifferential scanning calorimetry (Photo-DSC) Isothermal photopolymerisation studies were performed on a Perkin-Elmer DSC-7 differential scanning calorimeter and analyzed using Pyris software. An indium standard was used for calibration. The DSC head was modified with a single polycarbonate window covering both the sample and the reference cells as described by Pappas [17]. The polycarbonate window was regularly replaced because, over time, vapours from the cyanoacrylate samples affected its transparency. The blue LED light source was positioned at a distance of 50 mm above the measuring pans. The light source was switched on and off using a timer. Cylindrical aluminium sample pans were used with depths of 0.5, 0.8, and 1.2 mm to control the sample thickness. Nitrogen was used as the purge gas. The instrument was allowed to stabilize at every set isothermal temperature before photocuring experiments commenced. 2.4. Tensile testing The shear strength of the adhesive bonds between the rigid substrates was measured according to the ASTM D4501 shear block testing standard on an Instron 4303 tensile tester. The load cell had a maximum capacity of 25 kN in tension. Shear strength values reported here are averages of at least five replicates in accordance with the ASTM specification. 2.5. Bonding process The application process is a vital factor that determines the ultimate strength of underwater bonds. In this study the top substrate was always a clear polycarbonate sheet. This allowed facile illumination of the adhesive sandwiched between the two substrates. A 19 mm diameter pencil ring was drawn in the centre of the polycarbonate sheet used as the top substrate. A quantity of 0.50 g ( 70.01 g) of adhesive was placed on this demarcated area. This was done to keep the surface area of the adhesive, exposed to water, constant. The bottom substrate was first submerged in water. It was left in the water bath for at least 5 min to allow temperature equilibration. The top substrate was then submersed into the water bath and pressed firmly onto the bottom substrate. This action caused the adhesive to flow radially outward and cover the entire bond area. Next, the light source was submersed. 209 It was positioned at a distance of 50 mm away from the bond line and switched on to initiate cure. It should be noted that rapid polymerisation of the cyanoacrylate resin ensues as soon as it comes in contact with water. However, this gives rise to a protective skin layer that acts as a barrier for further water ingress by diffusion [5,6]. Thus only the adhesive’s outer surface (inside the 19 mm ring on the top substrate) is instantly affected by the exposure to water. This part of the cyanoacrylate adhesive, cured by the reaction with water, assumes a white colour. This contrasts with the orange tint of cyanoacrylate cured by the blue light radiation. Since the whitecoloured skin represents already polymerised material, it does not contribute to the bonding of the bottom substrate. When the two sheets are pressed together, the protective skin ruptures and adhesive is squeezed out. The uncured adhesive displaces the water as it travels radially outwards between the two substrates. Only this part of the surface is responsible for the measured bond strength. The non-bonded water-cured skin region trapped between the two substrates had an average diameter of 21 mm. This inactive bond area (346 mm2) was subtracted from the total bond area (2 500 mm2) to give, on average, an active bond area of 2154 mm2 for calculating the shear bond strength. The average bond line thickness was measured at 0.2 mm. The effect of illumination time and water temperature on the ultimate bond shear strength was determined using polycarbonate as top and bottom substrates. The sheets were bonded in potable water. The effect of illumination time was determined using a water temperature of 15 1C. Next, the illumination time was kept constant at 1 min and the water temperature varied from 1.5 to 40 1C. Bond strength measurements to various other substrate materials were done on samples bonded in 15 1C potable water as well as in simulated sea water. The different materials tested are categorised as metals (mild steel, aluminium, anodised aluminium, stainless steel 304) and other polymers including ABS (Acrylonitrilebutadiene-styrene), poly(vinyl chloride) (PVC), poly(methyl methacrylate) (PMMA), polycarbonate (PC), and fibreglass-reinforced polyester (FRP). Substrate preparation was the same in each case and an illumination time of 1 min was used throughout. The effect of the length of time for which the adhesive was exposed to water before bonding studied by delaying the pressing together of the sheets for predetermined times. These tests were performed in potable and in artificial sea water at a temperature of 15 1C. In all cases the actual shear bond strengths were measured at room temperature. The averages of at least five replications are reported. 3. Results and discussion 3.1. Thermal analysis 3.1.1. Experimental problems with Photo-DSC The cure reaction of cyanoacrylates is highly exothermic and calorimetric techniques seem ideal to study the polymerisation kinetics. Cyanoacrylates are among the most reactive monomers to be examined kinetically. Pepper and co-workers [18,19] studied their cure kinetics and the mechanism of polymerisation. Despite the fact that they employed carefully controlled experimental conditions, problems were experienced with respect to reproducibility. Similar problems were experienced in this study where the polymerisation of cyanoacrylate adhesives was followed using photo-DSC. 3.1.2. Effect of photoinitiator concentration It is conventional to assume that the DSC-measured heat flux is proportional to the rate of polymerisation. The amount of ARTICLE IN PRESS 210 W.E. Cloete, W.W. Focke / International Journal of Adhesion & Adhesives 30 (2010) 208–213 0.5 240 210 -0.5 0.50 mm -1.5 Cure time, t95 (s) Heat flow (W/g) 180 0.05% PI 0.10% PI -2.5 0.20% PI 0.30% PI -3.5 Light on at 30 seconds 0.40% PI 0.80 mm 150 1.20 mm 120 90 60 0.50% PI -4.5 30 Cure temperature 15ºC -5.5 0 0 50 100 150 Time (s) 200 250 0 0.1 0.2 0.3 0.4 0.5 Concentration Initiator (% m/m) 0.6 0.7 Fig. 1. Effect of the photoinitiator concentration on the isothermal cure exotherms for Loctite 454. Fig. 2. Effect of photoinitiator concentration and adhesive film thickness on the cure time (t95). photoinitiator (PI) present affects the rate of polymerisation. Thus experiments were conducted varying both the concentration of dibenzoylferrocene and the adhesive film thickness. Since the idea was to simulate actual underwater bonding conditions, relatively thick films (0.5–1.2 mm) were tested in order to determine the optimum photoinitiator concentration. The photoinitiated DSC cure studies revealed complex cure behaviour. The measured cure rates were not sufficiently repeatable to allow a proper kinetic analysis of the data. Consequently, the cure time was simply characterized by the time required to attain 95% of the full conversion time (t95). For a given experiment, this was calculated as the time point where 95% of the total heat release, due to the reaction exotherm, was reached. The optimum photoinitiator concentration was associated with the shortest cure time. Fig. 1 shows DSC exotherms measured at a film thickness of 0.8 mm. The peak in the exotherm corresponds to the maximum rate of polymerisation. Fig. 1 reveals that the maximum cure rate first increases and then decreases with increasing photoinitiator concentration. This counterintuitive behaviour can be rationalized as follows: Cure initiation is hampered by the acidic stabilizer. This means that the stabilizer neutralization reaction competes with the cure initiation reaction. Consequently, it takes an even longer time to reach the maximum cure rate when the concentration of initiator is low. At intermediate photoinitiator concentrations the maximum cure rate is reached in a very short time. However, at higher concentrations the exotherm ‘‘tails out’’, i.e. takes a longer time to return to the base line. This is attributed to excessive light absorption by the outer layers hampering photoinitiation of the deeper layers, the so-called ‘‘the inner filter effect’’. Similar exotherm trends were observed for experiments conducted with adhesive film thicknesses of 0.5 and 1.2 mm. Nevertheless, the overall heat of reaction was independent of initiator concentration and film thickness and amounted to 26274 J/g. Fig. 2 shows the effect of initiator concentration and film thickness, at a temperature of 20 1C, on the cure time. As mentioned, this was taken as the time to reach 95% conversion (t95). Generally and unsurprisingly, the thicker the film, the longer is the cure time (t95). The cure time curves obtained by a plot against the initiator concentration show shallow minima. The minimum cure time and corresponding initiator concentration increases with increasing film thickness in accordance with expectations considering the ‘‘inner-filter’’ effect [20]. The optimum concentration of dibenzoylferrocene for the polymerisation of films of 0.5, 0.8, and 1.2 mm are approximately 0.28, 0.2, and 0.14 wt %, respectively. In the current underwater adhesion tests, the average bond line thickness was observed to be slightly less than 0.2 mm. This means that a photoinitiator concentration of 0.3 wt % or a little higher would probably provide the fastest cure. It was, however, decided to standardize on 0.2 wt % photoinitiator in all further testing because real-life underwater surfaces are generally rather rough and one would therefore expect that thicker adhesive lines may be encountered. 3.1.3. Effect of photopolymerisation temperature Isothermal photopolymerisation experiments were performed at temperatures of À10, 20 and 50 1C. Owing to the poor reproducibility of the data, each curve in Fig. 3 actually represents the average of ten separately measured exotherms. The exotherms in Fig. 3 show that the cure advances faster at lower reaction temperatures in agreement with previous studies dealing with ferrocene derivatives [21,22], or aliphatic amines and pyridine derivatives [19] as initiators. These results indicate that this adhesive system will cure faster in cold water conditions. This gives it an advantage over epoxybased underwater adhesives which often fail to cure properly under such cold conditions. If it is assumed that the 95% conversion state is controlled by a single reaction with its associated Arrhenius activation energy, the effective cure time (in seconds) for 0.80 mm thick adhesive containing 0.20% initiator is given by the expression t95 ¼ 1495eÀ782=T ð1Þ 3.2. Bond strength results 3.2.1. Illumination time Fig. 4 shows the effect of illumination time on ultimate bond shear strength obtained in potable water at 15 1C with polycarbonate as both top and bottom substrate. The bond strength increases with increased illumination time but reaches ARTICLE IN PRESS W.E. Cloete, W.W. Focke / International Journal of Adhesion & Adhesives 30 (2010) 208–213 1.0 Steel Light source on at 30 s 0.0 Al anodized -1.0 Heat flow (W/g) 211 Aluminium SS 304 -2.0 -10ºC ABS 20ºC -3.0 PVC 50ºC -4.0 PMMA 0.20 wt % PI Film thickness 0.8 mm -5.0 -6.0 PC FRP 0 50 100 150 Time (s) 200 250 0 Fig. 3. Averaged DSC cure exotherms obtained at different temperatures. 2 4 6 Shear bond strength (MPa) 8 Fig. 5. Shear strength for polycarbonate bonded to other substrate materials. 7 Shear bond strength (MPa) 6 5 4 3 2 Shear bond strength without illumination (kept in position for 60 s and then tested) 1 0 0 25 50 75 Illumination time (s) 100 125 Fig. 4. Effect of illumination time on the shear strength of polycarbonatee bonded to polycarbonate. a plateau value beyond 50 s. An advantage offered by the cyanoacrylates over standard acrylic adhesives [8–10] is that a useful level of bond strength develops even in the absence of illumination. 3.2.2. Substrate material Fig. 5 shows the shear bond strengths achieved for bonds of polycarbonate to other substrates. It is clear that the adhesive bonds to polymeric substrates were much stronger than those to metals. For mild steel and aluminium the failure mode was adhesive failure at the metal surface. Visual inspection revealed the presence of patchy moisture films on the bared metal. This implies that the adhesive was unable to completely and effectively remove water from the metal surface during the bonding process. Low shear bond strengths of 0.86 and 0.49 MPa were recorded for mild steel and aluminium, respectively. Bonding was more effective with the anodised aluminium averaging 1.27 MPa, i.e. more than double the strength obtained with untreated aluminium. In this case the failure mode was mixed. It occurred mainly between the anodised layer and the metal surface, revealing the shiny metal surface underneath the grey anodised layer. In minor parts cohesive failure in the adhesive layer was also observed. Kinloch [23] also reported this type of failure mode for aluminium surfaces anodised with sulphuric acid. Adhesion to stainless steel 304 was somewhat better than that to anodised aluminium. The failure mode was mixed with small amounts of adhesive present on the stainless steel surface after bond failure. All the bonds made to polymers showed varying degrees of cohesive failure, except for the FRP which showed massive substrate failure. The average shear bond strength to polymers was in the order of 5 MPa. Drain et al. [24,25] found that cyanoacrylates dissolve polycarbonate to form a type of ‘‘solvent’’ welded interface that is able to withstand moisture over long periods. Whether cyanoacrylate adhesives have the ability to do the same to other polymers like ABS, PVC, PMMA, and FRP is not clear, but very high bond strengths are measured on these materials. The large difference in shear bond strength between metallic and polymeric materials could be attributed to their respective hydrophilicities. Metallic surfaces tend to be more hydrophilic and polymeric surfaces more hydrophobic. It is more difficult for the cyanoacrylate adhesive to completely displace the water from immersed metal surfaces and ensure the good initial contact necessary for the development of strong bonds. 3.2.3. Effect of temperature Fig. 6 shows that varying the water temperatures between 1.5 and 40 1C did not significantly influence the bond strength. Although photo-DSC data revealed negative temperature dependence, i.e. a lower rate of cure at high temperatures, this did not affect the ultimate bond strength obtained at 40 1C. It can therefore be concluded that the 1-min illumination time on a bond of 0.2 mm thick adhesive layer cured at 40 1C was sufficient to achieve the full bonding. ARTICLE IN PRESS 212 W.E. Cloete, W.W. Focke / International Journal of Adhesion & Adhesives 30 (2010) 208–213 5 environments. The reduction in bond strength is due to a combination of two factors: the reduction in the amount of adhesive available for bonding and therefore a smaller bonding area, and the increase in bond line thickness due to the PECA skin which becomes thicker over time causing the bond line to be thicker. 4 4. Conclusions 3 Clear polycarbonate sheets can be bonded rapidly to underwater substrates using the commercial cyanoacrylate adhesive Loctite 454 spiked with the photoinitiator dibenzoylferrocene. In this regard the application process is critical. First a bead of the viscous adhesive is placed on the polycarbonate plate. On submersion a protective skin forms on the adhesive’s surface owing to water-initiated polymerisation. When the plate is pressed against the underwater substrate, the squeezing action causes unreacted adhesive to ooze out from the broken protective skin. It spreads radially outward to fill the gap between the two substrates. Simultaneously it displaces the water initially present. Full cure, throughout the adhesive layer, is guaranteed by photochemical polymerisation. Initiation is triggered by illumination with a suitable blue light source, e.g. blue LED’s (wavelength: 467 nm). More specifically, a one minute exposure to this blue light with an intensity of ca. 5 mW/cm2 was sufficient to reach the full ultimate bond strength of adhesive films with a thickness of 0.2 mm when the resin contained 0.2 wt % benzoylferrocene. Bond strength development was insensitive to the water temperature in the range of 1 – 40 1C. Best adhesion performance, with the shear strength exceeding 5 MPa, was obtained when bonding one polycarbonate sheet to another. Strong bonds were also obtained when using other plastics but bonds to metal substrates were substantially weaker. Bond strengths were ca. 1 and 2 MPa to anodised aluminium and stainless steel, respectively. Finally, care should be taken to minimize the exposure time to water to avoid loss of bonding strength. 7 Shear bond strength (MPa) 6 2 1 0 0 10 20 Temperature (°C) 30 40 Fig. 6. Effect of temperature on shear bond strength. 6 Potable Water Shear bond strength (MPa) 5 Sea Water 4 3 2 Acknowledgements 1 Financial and technical support from the Council for Scientific and Industrial Research is acknowledged with gratitude. 0 0 10 20 30 40 Work time (min) 50 60 Fig. 7. Effect of underwater work time on shear bond strength. 3.2.4. Effect of water exposure time Unlike acrylate-based adhesives, cyanoacrylates react when coming in contact with water. On submersion, the outer layer of the cyanoacrylate adhesive (clear with a light orange tint) rapidly polymerises to form a white protective barrier skin. This skin is permeable and water ingress via diffusion will slowly cause polymerisation of the remaining adhesive on the inside until all of it has been converted. Preliminary measurements indicate that in room temperature potable water the polymerised cyanoacrylate skin grew to a thickness of about 1.5 mm in one hour. It is therefore imperative that the bond should be made as soon as possible after submersion, i.e. contact with water. Fig. 7 shows the effect of water exposure time on the shear bond strength. It shows a linear decline in bond strength with water exposure time. The bond strength decays slightly faster in artificial sea water, possibly due to the pH being higher than that of potable water. This accords with the findings of Katti and Krishnamurti [26] who found that alkyl cyanoacrylates polymerise faster in higher pH References [1] Mate B, Mesecar R, Lagerquist B. The evolution of satellite-monitored radio tags for large whales: one laboratory’s experience. Deep-Sea Res II 2007; 54(3–4):224–47. ´ [2] Lemarie DP, Smith DR, Villella RF, Weller DA. Evaluation of tag types and adhesives for marking freshwater mussels (Mollusca: Unionidae). J Shellfish Res 2000;19(1):247–50. [3] Ross KA, Thorpe JP, Norton TA, Brand AR. An assessment of some methods for tagging the Great Scallop, Pecten Maximus. J Mar Biol Assoc UK 2001;81(6): 975–7. [4] http://www.garf.org/bestglue.html, 5 November 2006. [5] Card SW, West JR. Underwater Bonding of Surface Conforming Material. US Patent 4,793,887, assigned to Morton Thiokol, Inc, Chicago, USA, 1988a. [6] Card SW, West JR. System for Underwater and Cold Temperature Bonding. US Patent 4,793,888, assigned to Morton Thiokol, Inc, Chicago, USA, 1988b. [7] Leonard F, Brandes G. Underwater Adhesive. US Patent 3,896,077, assignee unknown, USA, 1975. [8] Dolez P, Love B. Adhesive bonding as an alternative for underwater structural repair. Paper presented at the Eleventh International Offshore and Polar Engineering Conference, Stavanger, Norway, 17–22 June 2001. [9] Dolez P, Marek M, Love BJ. Photopolymerizable acrylic resin: effect of curing time and temperature. J Appl Polym Sci 2001;82:546–54. [10] Dolez PI, Williams C, Goff A, Love BJ. Properties of photopolymerisable acrylic adhesive for underwater bonding. J Soc Underw Technol Lond 2003;25(4): 199–208. [11] North Sea Resin, /http://www.northsearesins.com/S (accessed 8 June 2007). [12] Yamaguchi Y, Palmer BJ, Kutal C, Wakamatsu T, Yang DB. Ferrocenes as anionic photoinitiators. Macromolecules 1998;31(15):5155–7. ARTICLE IN PRESS W.E. Cloete, W.W. Focke / International Journal of Adhesion & Adhesives 30 (2010) 208–213 [13] Palmer BJ, Kutal C. A new photoinitiator for anionic polymerization. Macromolecules 1995;28(4):1328–9. [14] Yamaguchi Y, Kutal C. Efficient photodissociation of anions from benzoylfunctionalized ferrocene complexes. Inorg Chem 1999;38(21):4861–7. [15] Yamaguchi Y, Kutal C. Benzoyl-substituted ferrocenes, an attractive new class of anionic photoinitiators. Macromolecules 2000;33:1152–6. [16] Kutal CR, Yamaguchi Y. Substituted Benzoylferrocene Anionic Photoinitiators. US Patent 6,127,445, assigned to The University of Georgia Research Foundation Inc., Georgia, USA, 2000. [17] Pappas SP. Radiation curing – science and technology. New York and London: Plenum Press; 1992. [18] Pepper DC. Anionic and zwitterionic polymerisation of a-cyanoacrylates. J Polym Sci Polym Symp 1978;62:65–77. [19] Pepper DC. Kinetics and mechanisms of zwitterionic polymerizations of alkyl cyanoacrylates. Polym J 1980;12(9):629–37.  [20] Mauguiere-Guyonnet F, Burget D, Fouassier JP. On the inhibiting effect of phenolic compounds in the photopolymerization of acrylates under highintensity and polychromatic UV/Visible lights. J Applied Polym Sci 2007;103: 3285–9. 213 [21] Chan WY, Lough AJ, Manners I. Photoreactivity and photopolymerization of silicon-bridged [1] ferrocenophanes in the presence of terpyridine initiators: unprecedented cleavage of both iron–cyclopentadienyl bonds in the presence of chlorosilanes. Chem Eur J 20078867–76. [22] Tanabe M, Vandermeulen GWM, Chan WY, Cyr PW, Vanderark L, Rider DA, et al. Photocontrolled living polymerizations. Nat Mater 2006;5:467–70. [23] Kinloch AJ. Developments in adhesives – 2. London and New York: Applied Science Publishers; 1981. [24] Drain KF, Guthrie J, Leung CL, Martin FR, Otterburn MS. The effect of moisture on the strength of steel-steel cyanoacrylate adhesive bonds. J Adhes 1984;17: 71–82. [25] Drain KF, Guthrie J, Leung CL, Martin FR, Otterburn MS. The effect of moisture on the strength of polycarbonate-cyanoacrylate adhesive bonds. Int J Adhes Adhes 1985;5(3):133–6. [26] Katti D, Krishnamurti N. Anionic polymerization of alkyl cyanoacrylates: in vitro model studies for in vivo applications. J Appl Polym Sci 1999;74: 336–44.