Comparative evaluation of polycyanoacrylates

Acta Biomaterialia xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Full length article

Comparative evaluation of polycyanoacrylates Yoav Barkan a,b, Mira Levinman a, Ilana Veprinsky-Zuzuliya a, Tsadok Tsach b, Emmanuelle Merqioul a, Galia Blum a, Abraham J. Domb a,⇑, Arijit Basu a,⇑ a b

Institute of Drug Research, School of Pharmacy-Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem 91120, Israel Division of Identification and Forensic Sciences, Israel Police, Jerusalem, Israel

a r t i c l e

i n f o

Article history: Received 9 September 2016 Received in revised form 28 October 2016 Accepted 3 November 2016 Available online xxxx Keywords: Cyanoacrylates Alkoxy cyanoacrylate Alkyl cyanoacrylate Viscoelasticity Hydrolytic degradation

a b s t r a c t Cyanoacrylate esters (CA) and their corresponding polymers (PolyCA) are used as general and medical adhesives, biodegradable carriers for controlled drug delivery, and as agents for fingerprint development in forensic science. Most reports of cyanoacrylate are on ethyl or 2-octyl cyanoacrylate ester with little attention to other esters. It is the objective of this study to determine the differences amongst cyanoacrylate esters regarding their synthesis, chemical characterization, hydrolytic degradation, and thermal and mechanical properties. Cyanoacrylate polymers of short and long alkyls or oxy-alkyls, cyclic, and aromatic esters have been synthesized and evaluated. All monomers form polymers when exposed to triethylamine vapours possessing molecular weights in the range of 15,000–150,000 Da, where the alkyl esters form high MW polymers. A wide range of hydrolytic degradation rates has been found, as monitored by the release of formaldehyde over time. Alkoxy CAs show faster hydrolytic degradation compared to alkyl CAs. Regarding mechanical properties, CAs are classified into primarily viscous (G0 > G00 ) and primarily elastic (G00 > G0 ). Alkoxy CA polymers have a higher loss modulus (G0 ) than storage modulus (G00 ). Octyl CA polymer possess a G0  G00 (phase angle 45°) providing appropriate balance between mechanical strength and plasticity. Most alkyl CAs are compact and brittle. Alkoxy CAs show enhanced plasticity, but they lack mechanical strength. In general, the Tg for alkoxy CAs is less than alkyl CAs. Alkoxy CAs depolymerise rapidly at temperatures >200 °C. Overall, ester sidechains of CA esters strongly affect the polymer property. Statement of Significance Polycyanoacrylates are an important class of biodegradable polymers mainly used as bioadhesives. The study describes comparative evaluation of different cyanoacrylate polymers with respect to their chemistry, degradation, safety, mechanical, and thermal properties. The study forms the basis for choosing appropriate combination of cyanoacrylate esters for various biomedical uses. Moreover, this study reveals properties of a few new polycyanoacrylates for the first time. Ó 2016 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction Cyanoacrylates (CAs) are used as general and medical adhesives, drug delivery vehicles [1,2], in forensic applications [3], coatings [4], and elsewhere. Due to strong electron withdrawing nitrile and ester substituent, the vinyl carbon are extremely reactive towards Michael type 1,4-addition reactions. Consequently, these mono-

⇑ Corresponding authors at: School of Pharmacy-Faculty of Medicine, The Hebrew University of Jerusalem, Hadassah Ein Kerem Medical Center Campus, Jerusalem 91120, Israel (A.J. Domb). E-mail addresses: [email protected] (A.J. Domb), [email protected], arijit. [email protected] (A. Basu).

mers rapidly polymerize anionically upon contact with a variety of initiators, including the weakest nucleophiles like water (mostly moisture from the air) [5]. These monomers undergo almost instantaneous polymerization when applied to amine or watercontaining substances such as skin and tissues. The resulting polymer is stable and water-resistant with significant adhesive strength. These properties of CAs facilitate medical applications such as bioadhesives in surgical incisions and lacerations. They are one of most widely used medical and commercial adhesives, providing easier wound closure with enhanced cosmetic results. Being able to polymerize instantaneously, CAs are also an attractive candidate for fabricating delivery vehicles through emulsion or interfacial polymerization reactions [2]. Another feature of CAs

http://dx.doi.org/10.1016/j.actbio.2016.11.011 1742-7061/Ó 2016 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

2

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

that attracts formulation scientists is their ability to cross barriers such as the blood brain barrier [6,7] and enhance the intracellular penetration [8]. The sidechains of these CA esters determine their properties: mechanical, viscoelastic, thermal, degradation, and biocompatibility. Most reports of cyanoacrylate are focused on ethyl or 2-octyl cyanoacrylate esters with little attention to other cyanoacrylate esters. It is the objective of this study to determine the differences among cyanoacrylate esters regarding their synthesis, chemical characterization, hydrolytic degradation, and thermal and mechanical properties. Cyanoacrylate polymers of short and long alkyls or oxy-alkyls, cyclic and aromatic esters were synthesized and evaluated. As reported earlier the mechanical, thermal, degradation, and biocompatibility properties may vary widely depending on the cyanoacrylate polymer side chains [9]. In general increasing the length of alkoxy cyanoacrylates improves the flexibility. For alkyl cyanoacrylate a systematic study is still warranted. It is also important to search for newer cyanoacrylate polymers having significant flexibility and durability with acceptable biocompatibility. Newly discovered CA esters may be used as comonomers/polymers for tailoring the properties of traditional cyanoacrylates. They may also replace the additives (such as plasticizers) that are routinely added to the cyanoacrylate formulations, which generally leach out causing unwanted side effects. Only three cyanoacrylates are used for most of the commodity and medical purposes ethyl, butyl, and 2-octyl cyanoacrylate. Ethyl cyanoacrylate forms a brittle, instantaneous adhesive bond and is the glue of choice for commodity purposes. n-Butyl and 2-octyl cyanoacrylates are used mainly as medical adhesive. Polymer of n-butylCA forms a brittle, but strong and instantaneous adhesive bond. On the other hand Polymer of 2-octyl cyanoacrylate forms relatively flexible adhesive bond, but the polymerization time may be prolonged. In all commercial medical cyanoacrylate glues polymerization initiator and plasticizers are routinely used. The flexibility and durability achieved using these formulations are far from what is needed. Therefore, there is a need to identify newer more flexible cyanoacrylates that might be used by themselves or as blends. Knowledge of the mechanical properties of polyCAs is important for their use in specific applications. The only systematic study on dynamic mechanical properties were reported for methyl, ethyl, butyl, and allyl cyanoacrylate by Rooney and co-workers, in 1987 [10]. A few other studies are also reported that emphasize the effect of polymer side chains on their mechanical properties [9,11]. Thermal properties are also important to estimate the durability and mechanical properties of cyanoacrylates. It has also been shown to depend strongly on the side chains [12]. Here we report synthesis, comparative polymerization behaviour, degradation, dynamic mechanical, and thermal behaviour of various CA esters with different side-chains relevant for industrial and academic research (Table 1) [5].

2. Experimental

Table 1 Structure of the cyanoacrylate polymers used in the current study.

CN COOR Cyanoacrylates

R

1 2 3

Methoxyethyl CA Ethoxyethyl CA TEG CA

ACH2CH2AOACH3 ACH2CH2AOACH2CH3

4

peg2000 CA

5 6 7 8 9 10

Ethyl CA Isopropyl CA Butyl CA Octyl CA Ethylhexyl CA Cyclohexenyl CA

ACH2CH3 ACH(CH3)2 ACH2CH2CH2CH3 ACH(CH3)CH2CH2CH2CH2CH2CH3 ACH(CH3CH2)CH2CH2CH2CH2CH3

11 12

Phenylethyl CA Furfuryl CA

ACH2CH2Ph

13

Octadecyl CA

ACH2(CH2)16CH3

O O

O O

O O

OH O n

O

xylene) were freshly distilled and then dried using standard procedures. Determination of C, H, N was performed using the Series II CHNS/O Analyser Perkin Elmer, (MO, USA). Melting point was determined by Fischer Scientific Melting Point Apparatus, USA by capillary method. Electrospray ionization mass spectrometry (ESI MS) was recorded on a ThermoQuest, Finnegan LCQ-Duo instrument. Fourier transform infrared spectroscopy (FTIR) spectra was performed using the Smart iTR ATR sampling accessory for Nicolet iS10 spectrometer with a diamond crystal. Nuclear Magnetic Resonance (1H NMR) spectra were recorded on a Varian 300 MHz instrument using tubes with 5 mm outside diameter. 10 mg of sample was dissolved in 2 mL of deuterated chloroform as a solvent (CDCl3). UV measurements were done on a Uvikon 930 spectrophotometer using aqueous samples. Reactions were monitored by thin layer chromatography on silica gel plates. Iodine vapours and UV inspection at 254 nm and 365 nm were used for inspection of the spots. 2.2. Synthesis Synthesis of the cyanoacrylates was carried out by our earlier reported procedures [13,14]. However, there were a few modifications, especially in the purification procedures for these compounds (detailed in supporting information). Anthracene adduct of cyanoacrylic acid was esterified with different alcohols (polyethylene glycol methyl ether or triethylene glycol methyl ether) using the N,N0 -Dicyclohexylcarbodiimide/4-Dimethylaminopyridine (DMAP) chemistry in dry THF. The products were purified by crystallization from DCM and petroleum ether at 20 °C. The products were confirmed by IR and NMR spectroscopy.

2.1. Materials Reagents and deuterated solvents for NMR were purchased from Sigma-Aldrich Israel (Rehovot, Israel). Ethyl cyanoacrylate (ECA) was manufactured by Loctite Ltd. and commercialized as Hard Evidence by Lightning Co. (Jacksonville, FL). Other cyanoacrylate monomers were purchased from Chenso Inc. FL, USA. Triethylene glycol (TEG) methyl ether and polyethylene glycol (MW2000) monomethyl ether (PEG2000) CAs were synthesized in the laboratory. Other chemicals were purchased from SigmaAldrich (Rosh HaAyin, Israel). The solvents (THF, toluene, and

2.2.1. TEG CA anthracene adduct FTIR: 3323 cm1 (@CAH stretch); 2928–2851 cm1 (CAH stretch); 2238 cm1 (CN); 1749 cm1 (C@O stretch); 1459 cm1 (C@C stretch); 1245 cm1 (ester CAO stretch); 1104 cm1 (ether CAOAC stretch). 1H NMR (300 MHz, CDCl3) d 7.63–6.94 (m, 8H), 4.92 (s, 1H), 4.48–4.35 (m, 1H), 4.34–4.12 (m, 2H), 3.75–3.62 (m, 8H), 3.62–3.51 (m, 2H), 3.37 (s, 3H), 2.81 (dd, J = 13.0, 2.8 Hz, 1H), 2.20 (dd, J = 12.9, 2.6 Hz, 1H). TLC (50% ethyl acetate in hexane): anthracene adduct of cyanoacrylic acid Rf = 0.15, TEG methyl ether Rf = 0.1, cyanoacrylate adduct Rf = 0.35.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

2.2.2. PEG2000 CA anthracene adduct FTIR: 2883(CAH stretch); 1749 cm1 (C@O stretch); 1466 cm1 (C@C stretch); 1279 cm1 (ester CAO stretch); 1104 cm1 (ether CAOAC stretch). 1H NMR (300 MHz, CDCl3) d 7.56–6.98 (m, 8H), 4.91 (s, 1H), 4.42 (s, 1H), 4.22 (ddd, J = 16.1, 11.4, 7.1 Hz, 2H), 4.00–3.27 (m, 198H), 2.80 (dd, J = 13.0, 2.5 Hz, 1H), 2.19 (dd, J = 13.1, 2.4 Hz, 1H). TLC (50% ethyl acetate in hexane): anthracene adduct of cyanoacrylic acid Rf = 0.15, PEG2000 and PEG cyanoacrylate adduct do not move on TLC plate. 2.2.3. Cyanoacrylate monomers Cyanoacrylate anthracene esters (0.013 M), maleic anhydride (0.065 M), hydroquinone (20 mg), P2O5 (1 g) were refluxed in dry xylene (150 mL) under strictly inert conditions for 56 h. The reaction mixture was chilled, and unreacted maleic anhydride and maleic anhydride-anthracene adduct were filtered off (under positive N2 pressure) by repeated chilling and filtering process. The product was further purified by fractional distillation to remove unreacted maleic anhydride. The residue left behind was used without further purification. The products were confirmed by IR and NMR spectroscopy. TEG CA (compound 3, Table 1) This appeared as low straw-colour oil, isolated yield 55%, FTIR: 3110 cm1 (@CAH stretches); 2921 cm1 (CAH stretch); 2233 cm1 (CN); 1775–1743 cm1 (C@O stretch); 1452 cm1 (C@C stretch); 1237 cm1 (ester CAO stretch); 1055 cm1 (ether CAOAC stretch). 1H NMR (300 MHz, CDCl3) d 6.85–6.92 (d, J = 12.0 Hz, 2H), 4.69–4.25 (m, 1H), 3.82–3.52 (m, 10H), 3.37 (s, 3H). PEG2000 CA (compound 4, Table 1) Appeared as a yellowish brown oil, isolated yield 45%, FTIR: 3110 cm1 (@CAH stretches); 2870 cm1 (CAH stretch); 2242 cm1 (CN); 1776–1735 cm1 (C@O stretch); 1458 cm1 (C@C stretch); 1237 cm1 (ester CAO stretch); 1100 cm1 (ether CAOAC stretch). 1H NMR (300 MHz, CDCl3) d 6.85–6.95 (d, J = 12.0 Hz, 2H), 4.69–4.25 (m, 1H), 3.82–3.52 (m, 10H), 3.37 (s, 3H). 2.3. Polymerization and thin film formation PolyCAs were prepared by polymerizing the CA monomers in the following procedure. Polymerization was afforded in a vacuum desiccator, under inert atmosphere. All polymerization was carried out concomitantly for 13 monomers in triplicate. 1 mL (1 g for the solid monomers) of CA monomers were taken in a 5 mL vial (area exposed is 3.14 cm2) equipped with a magnetic stir bar (for the solid polymers no stirring is involved). First the atmosphere inside was made inert using argon and then a cotton wool (1.5 g) soaked in triethyl amine 100 lL was immediately introduced. The monomers were stirred at 60 rpm. Most of the monomers polymerize within 2 min (visually observed when the liquid monomers are solidified, stir stop time), but they were left for complete polymerization for 2 h. The completion of polymerization was assessed by 1 H NMR and GPC. The same polymers have been used for all analysis, to ensure the molecular weight remains same for all analysis. The films of the polymers were fabricated by solvent casting by spreading the synthesized polymers on a Polytetrafluoroethylene (PTFE) groove of 2 cm  5 cm  1 cm, as reported earlier [13]. In a typical procedure, 30% cyanoacrylate polymers dissolved in ethyl acetate. Once the polymers are dissolved they were uniformly spread by tilting and allowing the solvent to slowly evaporate overnight. Care was taken to avoid air bubbles. The resulting films having a thickness of 100 lM were peeled off carefully before analysis. For one PolyCA three different films having same dimensions were casted.

3

2.4. Molecular weight determination The molecular weights were determined by a gel permeation chromatography (GPC) system that consists of a Waters 1515 isocratic HPLC pump with a Waters 2410 refractive index detector, a Waters 717 plus auto sampler, and a Rheodyne (Coatati, CA) injection valve with a 20 lL-loop (Waters, MA). The samples were eluted with CHCl3 (HPLC grade) through linear Styragel HR5 column (Waters) at a flow rate of 1 mL/min. The molecular weights were determined relative to polystyrene standards (Polyscience, Warrington, PA) with a molecular-weight range of 5000– 400,000 Da using an Empower computer program. For each polyCAs three independent experiments were performed, using similar polymerization conditions (described in the earlier section). 2.5. Differential scanning calorimetry (DSC) measurements Polymer samples (4–5 mg) obtained from the previous step were weighted by micro analytical balance ±1 lg. The thermal behaviours of polyCAs were investigated using DSC Q4000 (TA Instruments, New Castle, UK). DSC thermograms were recorded for all samples by gradually heating from 20 °C to melting/ decomposition temperature (250 °C) at a rate 10 °C/min. A preheating and cooling cycle was also performed before measuring the actual thermograms from 20 °C to 150 °C. Weight loss was traced and detected by micro analytical balance. 2.6. Hydrolytic degradation Poly cyanoacrylates degrade to give formaldehyde, determined by formation of yellow coloured pyridine derivative formed by reaction with acetylacetone in the presence of ammonium acetate [13,15,16]. The hydrolytic degradation was performed as previously reported by our group. The procedure is briefly described as follows. Cyanoacrylate polymer films (same polymers as described in Section 2.3) were punched into disks of approximately 1 cm diameter weighing 10 mg. All polymeric disks were incubated in phosphate buffered saline (PBS) pH 7.4 at 37 °C in a shaker incubator in a 24 well plate. 200 lL aliquots were collected. Every time the aliquots were collected, they were replaced by an equal amount of PBS. Serial dilution was made from stock solution of formaldehyde (37 wt% in H2O). 200 lL of these solutions were taken, 1 mL of formaldehyde determination solution was added to them, and the volume was made up to 6 mL with distilled water. Formaldehyde determination solution was made by dissolving 3.8 g of ammonium acetate, 75 lL of acetic acid, and 50 lL of acetyl acetone in 25 mL of distilled water. The solutions were then warmed on a water bath at 60 °C for 10 min. Thereafter, they were rapidly cooled over ice bath and their absorbance was recorded at 410 nm. The concentration of formaldehyde released was determined and expressed as lg/mL. 2.7. Viscoelasticity behaviour Viscoelasticity behaviour was performed with an Anton Parr, MCR101 (Germany GmbH., Ostfidern) rheometer, fitted with a PP25 plate (diameter of 25 mm). All measurements were performed upon fixing the gap distance between the two plates at 1–0.5 mm, depending on the thickness of the film. Strain sweeps were performed for the strain range 0.001–100% at 25 °C, using oscillation experiments at a constant frequency of 5 rad/s. Storage modulus (G0 ) and loss modulus (G00 ) determined by the instrument were converted into their phase angles (d):

d ¼ tan1

G00 G0

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

4

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

Two types of plots were used: (i) phase angle (degrees, d) vs strain %, and (ii) stress (pa) vs strain %. The yield points were determined, which is characterized by a sharp rise in d. 2.8. Cell viability assay Similar polymer disks as those described under hydrolytic degradation were used to detertine the cell viability. Hela cells were grown at 37 °C under 5% of CO2 in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal calf serum (FCS) and 1% Penicillin-streptomycin (Biological Industries, Israel). Cells were seeded (40,000 cells/well in 24 wells plate in 1 mL) one day before the incubation with the different PolyCAs (5 mg/well). After 48 h incubation cells were fixed, and viability was measured by a quantitative colorimetric assay using Methylene Blue according to the procedure reported by Ben-Bassat et al. [17]. The data were expressed as a percentage of control; data represent the mean (± standard deviation) of triplicates. 2.9. Statistics All the monomers were polymerized in triplicate the data has been presented as average ± standard deviation (supporting information). The hydrolytic degradation and cell viability experiment were performed using identical disks in triplicate. The viscoelasticity was performed on identical polymer films in triplicate. The results are presented as average ± standard deviation, except in case of a few representative examples. 3. Results 3.1. Synthesis Cyanoacrylate esters were synthesized as previously reported (Scheme 1, Supporting Information) [13,14]. In the first step the venerable acrylate double bond of ethyl cyanoacrylate was protected by anthracene through Diels-Alder reaction. 1H NMR spectra show multiplets in the aromatic region (d 7–8 ppm), consistent with the formation of the anthracene adduct. Signals in other regions are also consistent with the structure. In the next step the ester was hydrolyzed to its corresponding acid. Mass spectrometry [MH] and FT-IR broadening of the peak 3376 cm1 suggested conversion of ester to acid. Resulting anthracene adduct-cyanoacrylic acid reacted with corresponding alcohols to form corresponding esters. The compounds were confirmed by 1H NMR appearance of new peaks in aliphatic region, compared to the starting acid. The final product was obtained by de-protection (retro Diels-Alder) of the double bond using maleic anhydride. All operations in this step (reaction, filtration, crystallization, etc.) were performed in a glove bag to prevent polymerization. The monomers were purified by repeated precipitation and filtration. Excess maleic anhydride was distilled by fractional distillation. The compounds were confirmed by 1H NMR: there was disappearance of aromatic signals due to release of anthracene and new peaks at d = 6.85 due to CH2@CA protons (Fig. 1). Table 2 shows the molecular weight and polydispersity of different polyCAs synthesized. The molecular weights of poly alkyl CA are generally higher than alkoxy CAs. Molecular weight strongly depends on the mode of polymerization. The type of initiator plays an important role in determining the molecular weight and polydispersity of the sample [18]. In the current study cyanoacrylate esters were polymerized using triethylamine vapours. In general, we observed higher dispersity for poly alkyl CAs. The general co-relation between the molecular weight distribution and CA ester side chains could not be deduced.

3.2. Polymer hydrolytic degradation PolyCA degrades in aqueous media to release formaldehyde. The rate of formaldehyde release is also proportional to the degradation of the polymer. The CA polymers undergo hydrolytic degradation through retro-Knoevenagel reaction. The hydroxyl ion performs a nucleophilic attack on PolyCA (alpha carbon to the ester), inducing release of formaldehyde and finally depolymerisation [19]. The ester side-chains of polyCAs strongly influence hydrophobicity, and the access of a hydroxyl ion into the alpha carbon. The polyCAs with hydrophilic sidechains have enhanced wettability compared to those with hydrophobic side chains. The hydrophilic PEG and alkoxy polyCAs degrade significantly faster compared to hydrophobic alkyl CAs. This observation is similar to our recently reported work in which we used PEG biscyanoacrylate crosslinker [13]. Poly PEG CA degrades the fastest, depolymerising completely within 4 days. This fast degradation is attributed to its hydrophilic side chains. The Alkyl CAs degrades significantly slower for at least 15 days. Fig. 2 shows the cumulative rate of release of formaldehyde from the PolyCAs. Table 2 describes the effect of hydrolytic degradation on molecular weight. All poly alkoxy CAs degrade to smaller fragments, and their resulting molecular weight cannot be detected. Among poly alkyl CAs ethyl, isopropyl, n-butyl, and octyl polyCAs have shown >50% drop in their molecular weights. Their change in dispersity is also significant (0.9–1.5). Poly octadecyl CA shows the least drop in molecular weight, correlating with the formaldehyde release profile. As observed by Mizrahi et al., the rate of formaldehyde release is not the only determinant of toxicity. They observed that despite slower formaldehyde release, poly octyl CA is more cytotoxic than poly alkoxyethyl CA. They attribute this discrepancy to the other degradation products of polyCAs such as poly(cyanoacrylic acid) and corresponding hydrophobic alcohols [9]. Therefore, alkoxy CA polymers release formaldehyde is enhanced, but the resultant alcohol through ester hydrolysis is safer. On the other hand, alkyl CAs release formaldehyde slowly, but the resultant alcohols are toxic. In another study, Lee et al., offer a contrasting report about in vitro and in vivo biocompatibility of poly octyl CA. They observed poly octyl CA is safe and does not cause significant cellular and DNA damage [20]. We believe knowledge of relative degradation rate of CAs is extremely important for choosing polymer adhesive blends or designing drug delivery vehicles. 3.2.1. Cell viability assay We further report on the relative safety profile. We observed that the toxicity of the cyanoacrylate correlates with the rate of release of formaldehyde (in accordance with Lee et al. [20]). All poly alkoxy CAs that showed fast hydrolytic degradation were also significantly cytotoxic in cell culture. Polymer synthesized from PEG2000 CA, however, was least cytotoxic amongst alkoxy CA despite fast degradation (see Fig. 3). Polymers synthesized from ethyl, furfuryl, and phenyl ethyl CA were cytotoxic among poly alkyl cyanoacrylates. The toxicity of Poly ethyl CA correlated with its corresponding degradation rate. Overall, we presented a preliminary relative safety profile for the polyCAs. Detailed studies that involve animal experiments should be carried out to estimate their safety profiles. 3.3. Viscoelasticity Knowledge of the viscoelastic behaviour of polymers is useful for understanding their overall physical and mechanical behaviour. Moreover, insight gained from such work should help improve the properties of existing CAs. It may also help in designing newer

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

5

Fig. 1. 1H NMR (300 MHz, CDCl3) for top: methoxy triethylene glycol CA, bottom polymer of methoxy triethylene glycol CA. Solvent peaks are marked as ⁄. All CA polymers were synthesized by polymerizing their corresponding monomer (in triplicate) using triethylamine vapours as an initiator. Samples were collected for 1H NMR and monitored for disappearance CH2@CA protons at d = 6.5.

materials or composites to tune the properties appropriately. Viscoelasticity behaviour is studied by subjecting the substance under examination through a periodic oscillatory shear. The shear may be applied at constant frequency by changing the amplitude (frequency sweep) or changing the amplitude keeping the frequency constant (amplitude sweep). Cyanoacrylates are polymers primarily used as bio-adhesives, thus undergoing a dynamic stress. Therefore, we used variable strain experiments to study their elasticity and relaxation patterns. All the polymeric samples were subjected to oscillatory shear at constant frequency and varying the rate of strain. Under these conditions the strain rate in a viscoelastic material will initially change linearly, and then fade off non-linearly to a yield point. Materials may be primarily elastic or primarily viscous. Primarily elastic materials show steep, but linear change (stress vs strain) and then break off at the yield point without much elongation. Primarily

viscous materials show a gentle stress vs strain response (less slope) and often nonlinear after the yield point. Elastic materials should have a higher storage modulus (G0 ), and viscous materials should have a higher loss modulus (G00 ). Phase angle plots reveal these properties clearly. A lower phase angle suggests the structure is more solid, and a higher phase angle suggests the structure is more soft and malleable. A steep rise from a very low phase angle suggests that structural integrity is broken. Based on the mechanical properties, polyCAs are classified into (a) those with (G00 > G0 , d > 45), and (b) those with (G0 > G00 , d > 45). 3.3.1. Cyanoacrylate polymers showing primarily viscous behaviour (G00 > G0 , d > 45) Alkoxy CA polymers (Table 1, cyanoacrylates 1–4) show a higher loss modulus than their storage modulus. Their phase angle (d) remains constant when oscillatory shear is applied. Moreover,

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

6

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

Table 2 Molecular weight and dispersity of synthesized polyCAs. Effect of hydrolytic degradation on molecular weight and dry weight for the poly cyanoacrylate esters are also shown. Sl.

PolyCAs

Initial molecular weight (Da)

Molecular weight (Da) after 15 days in PBSa

Initial PDI

PDI after 15 day in PBSd

% drop in mass after 15 daysb

Cumulative formaldehyde release after 15 days (lg/mL)e

1 2 3 4 5 6 7 8 9 10 11 12 13

Methoxy ethyl Ethoxy ethyl TEG PEG2000 Ethyl Isopropyl Phenylethyl n-Butyl Cyclohexenyl Octadecyl Octyl Tetrahydrofurfuryl Ethylhexyl

13,000 8500 15,900 145,900 135,700 138,350 40,830 99,200 47,440 55,130 108,000 39,650 110,000

Degradedc Degradedc Degradedc Degradedc 51,000 45,600 31,300 60,100 32,000 50,400 50,300 31,600 71,200

1.36 1.09 1.04 1.16 1.13 1.66 1.35 1.97 1.47 2.00 1.17 1.85 1.94

– – – – 2.34 2.98 2.11 2.90 1.89 2.23 2.66 2.77 2.31

90 79 100 100 40 34 24 51 21 15 46 31 52

153 89 167 292 33 31 18 26 16 28 30 53 3

PBS-phosphate buffer saline pH 7.4. a The samples were extracted from incubation media using chloroform and used for molecular weight determination in GPC. b The samples were extracted from incubation media using chloroform, which was evaporated under vacuum and then finally weighed. c The values obtained are beyond the range of GPC detection limit of 1000 Da. d PDI is dispersity index (Mw/Mn) determined after 15 days. e Average values of three observation, plot is presented in the supporting information. All observations are average of three independent experiments (including initial and final Mw and PDI) their standard deviations are presented in supporting information.

350 300

conc. µg/mL

250 200

Alkoxy CA

peg2000 CA methoxyethyl CA teg CA ethoxyethyl CA ethylhexy CA tetrahydro-furfuryl CA octadecyl CA phenylethyl CA cyclohexenyl CA butyl CA ethyl CA isopropyl CA

150 100

Alkyl CA

50 0 1

3

5

7

Days

9

11

13

15

Fig. 2. Rate of formaldehyde release from the cyanoacrylate polymers incubated at 37 °C with phosphate buffer saline (pH 7.4) for 15 days. Data presented as average of three experiments (n = 3), error bars represent the percentage error calculated from an average of three observations. The cumulative degradation plots after 15 days are presented in Supporting Information.

they do not show yield points, owing to their soft and flexible nature (Fig. 4). The storage modulus for polymers synthesized from ethoxy CA and methoxy CA remain constant with 100% strain. Polymers of PEG and triethylene glycol (TEG) cyanoacrylate show a sharp decrease in storage and loss modulus, suggesting the structure becomes softer when shear force is applied. The loss modulus for these polymers is always more than the storage modulus. No crossover point is observed throughout the stress process. Moreover, the phase angle is constant at 70°, so there is no yield point. All alkoxy CA show this pattern, in agreement with earlier reports in which they show enhanced plasticity

and flexibility [9]. The enhanced flexibility has been attributed to ease of rotation of the ether linkage, causing the internal plasticizing effect. Another observation is that longer side-chains are associated with less toxicity. Plasticity (expressed as elongation at break point) of hexoxyethyl cyanoacrylate is more than 2-octyl cyanoacrylate. Glycols have been used as plasticizer for many years [21–24]. These observations further support our observations. The plot of shear stress vs strain shows that methoxy methyl and ethoxy ethyl CA are linear, thus suggesting that the deformation is uniform with applied force, consistent with malleable materials. A similar pattern has also been observed for TEG and

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

7

Hela cells viability after 48h incubation with different Poly (Cyanoacrylates) 120 Viability (% of control)

100 80 60 40 20 0

cyanoacrylate esters (5mg/ml) Fig. 3. Cell viability on Hela cells with the different Poly(Cyanoacrylates). After 48 h incubation of 5 mg/well Poly(Cyanoacrylates) cell viability was measured by a quantitative colorimetric assay using Methylene Blue. The data was expressed as a percentage of control; data represent the mean (±standard deviation) of triplicates.

PEG2000 CAs, but not linearly as methoxy methyl and ethoxy ethyl CAs. These plots further consolidate that these alkoxy CA polymers are plastic and malleable. 3.3.2. Cyanoacrylate polymers showing primarily elastic behaviour (G0 > G00 , d < 45) Solid structures with storage modulus more than loss modulus deform reversibly within their elastic range. Beyond this range of stress, they deform irreversibly and their structure tends to disintegrate. The point of crossover (G0 = G00 ) is normally called the yield point and is characterized by a sharp rise in the phase angles. Most of the alkyl CA polymers reveal an average phase angle <20° within their LVE range. Poly octyl CA shows an intermediate ideal phase angle 30°, suggesting its structure is solid but with significant plasticity. Such structures have an equivalent G0  G00 throughout. We observed that the properties for octyl CA are different from all other poly alkyl CAs, having appropriate plasticity and structural integrity. Accordingly, it has been used widely as the surgical glue of choice [25,26]. All other polymers reveal low average phase angles, suggesting these structures are more solid. Poly ethyl CA is the brittle solid polymer in which the G0 , G00 crossover point comes at 0.3% strain, yield point 2000 Pa (Fig. 5). N-butyl, cyclohexenyl, tetrahydrofurfuryl, and isopropyl CAs have a significantly higher storage modulus. Poly tetrahydrofurfuryl and n-butyl CA did not survive the measured 100% strain the structure disintegrated at very low strain <1%, so we were unable to record values beyond this point. On the other hand, poly n-butyl CA can withstand shear stress at a yield point of about 15,000 Pa, but at very low strain. This suggests that the polymer is solid-compact that may withstand significant shear force, but it is not flexible. Cyclohexenyl CA survived 0.9% strain, and isopropyl CA 1% strain, at significantly low shear stress (5500 and 7500 Pa respectively) at yield point (Fig. 5). The results suggest the polymeric films are not resilient enough to withstand higher shear, nor plastic enough to undergo deformation. We present the two representative cases for poly octyl and ethyl CA (Fig. 6). Ethyl CA is the most brittle polyCA. We observed the yield point 0.3% strain.

Strain % for poly octyl (>50%), phenylethyl (20%), and ethylhexyl (25%) CAs show lower storage modulus, but tolerate significant strain before finally yielding. A brittle material breaks without significant deformation (strain) on stress. Stress vs strain plots are useful to study the plasticity of the material. The linear portion of this plot is attributed to elastic expansion. Thereafter, the curvilinear portion is the plastic region, where the material deforms irreversibly before the structural integrity is broken (yield point). We expect the adhesives to be resilient to significant stress and expand (reversible/irreversible) before breaking apart. The three polymers described above show resilience to stress (Fig. 7). Therefore, poly phenylethyl and ethylhexyl CA may be considered as a co-monomer in adhesive formulations. We observed that different CA polymers elicit different viscoelastic properties. Alkoxy cyanoacrylates are primarily viscous, and alkyl cyanoacrylates are primarily elastic. 3.4. Thermal analysis The structure of the cyanoacrylate esters affects the mechanical/viscoelasticity property of the polymer. PolyCAs are thermoplastic polymers that operate within a specific range of temperature. At higher temperature, they are theromolabile and tend to depolymerise. We recorded DSC thermograms to evaluate how the side chain alkyl group affects their thermal behaviour (Table 3). In general, the content of free volume (larger/branched alkyl side chains) should create disorder and is likely to decrease the Tg of the polymers. The DSC curves were evaluated by gradually heating the polymers from 20 to 250 °C at the rate of 10 °C/min. The Tg of the polymer depicts the transformation of a rigid amorphous state to a more flexible state. It is accompanied by a change in the heat capacity of the material. In general, the alkoxy CA presents a lower Tg, except poly-2-methoxyCA, because it has much less free volume compared to larger PEG or TEG side chains. The Tg observed for Poly(ethoxy ethyl) CA is close to that reported previously [2,11,12]. Moreover, the O---O repulsion of the alkoxy groups pushes the chains apart, creating more free volume and reducing the Tg. This observation may be correlated to the enhanced flexibility as described in the earlier sections.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

8

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

A

90 average phase angle (δ)

80 70 60 50 40 30 20 10 0 Ethoxy ethyl

methoxy ethyl

peg2000

TEG

Alkoxy Cyanoacrylate esters

B

100000

stress (pa)

10000 1000

methoxymethyl CA ethoxyethyl CA

100

PEG 2000 TEG CA

10

% strain

1 0.01

0.1

1

10

100

0.1 Fig. 4. (a) Comparison of phase angle plots for CAs whose G00 > G0 . tan d = G00 /G0 . Stress vs % strain plots for CA polymers whose G00 > G0 , viscoelasticity was recorded on three polymer films (n = 3), and data represented as average ± standard deviation, (b) stress vs strain plot for different poly alkoxy CAs (representative examples). The polyCAs are coded as presented in Table 1. TEG CA = polymer synthesized from methoxy triethylene glycol CA, PEG2000 = polymer synthesized from PEG CA (MW 2000).

The alkyl CA polymer with the largest side chain poly-2octadecyl CA has a lower Tg 37 °C. This polymer does not decompose Tm at 230 °C, confirmed by three heating-cooling cycles. The CA polymer with long chain alkyl groups is likely to be softer with significant amorphous property. Octyl, ethylhexyl, and phenyl ethyl also reveal a low Tg. Partly correlating to their flexibility as observed in dynamic mechanical analysis, those CAs with shorter side chains like n-butyl and ethyl reveal a much increased Tg. One exception is poly isopropyl CA that shows a much lower Tg 68 °C, in agreement with earlier reported observations [12]. Flexibility of polymers is often postulated to be associated with reduced Tg. We observed that substances with lower Tg are most of the time associated with more flexibility, but not for every polymer. If the softening temperature is around room or body temperature, the polymer sometime tends to be primarily viscous with no significant mechanical strength. In such cases, blends or copoly-

mers are useful to tailor correct balance between mechanical strength and plasticity. As shown earlier by Tripodo et al., blending epoxy plasticizer with ethyl cyanoacrylate makes it possible to reduce the Tg to as low as 40 °C. Yet the blends are elastic with significant mechanical properties owing to the large strong polyethyl CA strands [27]. Most of the CA polymers decompose/depolymerise at their melting temperature. In some cases almost the entire polymer is decomposed. This is a preliminary study dedicated thermo-mechanical analysis must be conducted to characterize these materials. Mechanical measurement at, above, and below Tg with variable temperature ramp is needed in the future to correlate the thermal and mechanical properties. Thermal properties are strongly dependent on molecular weight, purity, mode of polymerization, and measurement variables. The data compiled in this study are a guideline for approximating the thermal properties of polyCAs.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

9

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

i)

Average phase angle within LVE

30 25 20 15 10 5

l hy

ry

yl

ur

et

fu

en

ph

en

of dr

ra

hy

cl cy

n-

l

yl

l oh

ta oc

hy

ex

de

ex lh

oc et 2-

is

cy

yl

l ty

l ty bu

2-

op

n-

ro

E

th

yp

yl

yl

0

Fig. 5. (i) Average phase angle was calculated by plotting stress [pa] vs d, and then averaging out the linear viscoelasticity (LVE) region before yield point, where tan d = G00 /G0 obtained from plots of storage modulus (G0 ) and loss modulus (G00 ) vs stress. The plots of storage modulus (G0 ) and loss modulus (G00 ) vs stress and phase angle plots are presented in support information. The yield point (ii) stress, and (iii) strain for different alkyl CA polymers, calculated when the phase angle sharply rises on shear stress. Viscoelasticity was recorded on three polymer films (n = 3), and data represented as average ± standard deviation.

4. Discussion Poly cyanoacrylates (PolyCAs) are an important class of biopolymers, having wide range of applications as medical adhesives, drug delivery vehicles, and coatings. Here we report synthesis, comparative polymerization behaviour, degradation, dynamic mechanical and thermal behaviour of various CA esters with different sidechains (short and long alkyls or oxy-alkyls, cyclic, and aromatic). The polymers were synthesized by exposing them with traces of triethylamine vapours. All the monomers were concomitantly

A

exposed to the vapours, to maintain uniform polymerization conditions. The polymerization was initiated at the vapour/liquid interface and gradually moving to the bulk of the liquid. Through this procedure we obtained satisfactory dispersity (<2.0). The molecular weights were not affected when the surface area is increased, as demonstrated earlier [13]. As reported earlier [28] and also confirmed in our lab small traces of solvents used to dissolve the initiator affects the molecular weights and dispersity significantly. Using amine initiator in solvent initiates instantaneous polymerization the moment it touches the cyanoacrylate monomers, the polymerization does not efficiently proceed to the bulk. After repeated trials, we observed polymerization with using triethylene amine vapour provides satisfactory results. Therefore, we employed this common polymerization technique for all the other cyanoacrylate monomers. In general, poly alkyl CAs posses higher molecular weights than alkoxy CAs, also poly alkyl CAs have higher dispersity. The polar alkoxy chains may participate in chain termination resulting in lowering of molecular weight. The hydrophilic alkoxy CA also attract more water, the resulting higher concentration of initiator might have decreased their molecular weights, also this may initiate depolymerization. Different polymerization conditions may lead to different results. Therefore, it is important to polymerize all the cyanoacrylate esters concomitantly under uniform conditions. Cyanoacrylate polymerization reactions were in general exothermic as reported earlier [28–30]. We were unable to observe the end group of the initiator on NMR, suggesting zwitterion initiation, which in accordance to the earlier report [28]. The degree of polymerization were 100% in all cases as determined by 1H NMR, disappearance of the acrylate terminal protons at d = 7.00–7.08. The ester side-chains of polyCAs strongly influenced their hydrolytic degradation rates. During hydrolysis (retro Knoevenagel reaction), the access of the hydroxyl ion to the alpha carbon is severely hindered due to the presence of hydrophobic side chains. On the other hand, polyCAs with hydrophilic sidechains (poly alkoxyCAs) have enhanced wettability. The increased hydrophilic-

B

100000 10000 1000 100 10 1

0.001

0.01

0.1

Phase angle (δ)

C

1

90 80 70 60 50 40 30 20 10 0

10

100

0.001

0.01

10000000 1000000 100000 10000 1000 100 10 1 0.1 1

■- G' ▲- G'' 10

100

poly octyl CA

poly ethyl CA

1

10

100

1000

10000

100000

Stress [pa] Fig. 6. Comparison of poly octyl CA and poly ethyl CA (representative example). (i) G0 , G00 plots for poly octyl CA, (ii) G0 , G00 plots for poly ethyl CA, (iii) phase angle (d) comparison for poly ethyl CA (..j..) vs poly octyl CA (-j-).The two extreme cases of alkyl CAs one brittle (poly ethyl CA) the other malleable (poly octyl CA). Poly ethyl CA G0 , G00 crossover point comes at 0.6% strain; yield point 1790 Pa. Poly octyl CA G0 , G00 crossover point comes at 61% strain; yield point 16,200 Pa.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

10

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

100000

Stress [Pa]

10000

1000 octyl CA

100

phenylethyl CA 2ethylhexyl CA

10

1 0.001

0.01

0.1

% strain

1

10

100

Fig. 7. Stress vs % strain plots for CA polymers whose G0 > G00 , but are plastic and resilient (representative examples). The linear portion is the elastic region. Thereafter, they elongate irreversibly (plastic region) before finally yielding to stress. The three polymers are chosen with highest strain% before yield point.

Table 3 Thermal properties of Poly(cyanoacrylates). PolyCAsa

Tg (°C)

Melting (Tm) or depolymerisation temp (°C)

% weight lossb

2-Methoxy ethyl 2-Ethoxy ethyl TEG PEG2000 Ethyl Isopropyl Phenylethyl n-Butyl Cyclohexenyl Octadecyl Octyl Tetrahydrofurfuryl Ethylhexyl

81 63 49 29 132 69 55 113 130 37 57 79 143

247 254 Not observed 199 245 236 250 240 224 231 237 222 218

97 96 48 48 99 99 55 99 44 3 98 76 95

a Polymers were synthesized by anionic polymerization of CAs in a closed humid chamber at room temperature for 24 h fumed with 10 lL of triethyl amine on cotton per gram of polymer. b Weight loss was traced by weighing the sample (in the crucible) in micro analytical balance after recording the DSC thermogram (from 20 °C to melting/ decomposition temperature (250 °C) at a rate of 10 °C/min).

ity was demonstrated earlier through contact angle measurements, on hydrophobic surface (silicon rubber latex wafer). The contact angles for alkoxy CAs were observed to be significantly higher compared to 2-octyl CA. Moreover, increased side chains have shown to decrease the contact angles [9]. We observed Poly PEG CA degrades the fastest, depolymerising completely within 4 days and Poly ethyl hexyl CA degrade the slowest (insignificant degradation in 15 days). In general, poly alky CAs degrades significantly slower compared to poly alkoxy CAs. Alkoxy CAs show rapid decrease in weight and formaldehyde release, also they show sharp decline in molecular weight this can be due to very rapid degradation (possibly both surface and bulk degradation). Very small weight loss and significant molecular weight decrease of some alkyl CA suggest the bulk degradation mechanism. Hydrolytic degradation behaviour also has a significant effect on the safety of PolyCAs. It seems obvious that rate of release of formaldehyde should correlate with toxicity profile of cyanoacrylate polymers. However, a contrasting report claimed despite slower formaldehyde release, poly octyl CA is more cytotoxic than poly alkoxyethyl CA. They attributed this discrepancy to the release of corresponding hydrophobic alcohols [9]. Another study

reported on the biocompatibility of poly octyl CA, they observed poly octyl CA is safe and does not cause significant cellular and DNA damage [20]. We used the HeLa cells to study the effect of the CA polymers on cell viability. They are one of the most widely used human cell line used to asses cytotoxicity of cyanoacrylates [31–33]. In our study on cell lines, we observed that the toxicity of the cyanoacrylate correlates with the rate of release of formaldehyde (in accordance with [20]). We investigated the viscoelastic behaviour of PolyCAs films to understand their overall physical and mechanical behaviour. The PolyCAs films are subjected to a shear force while changing the amplitude keeping the frequency constant (amplitude sweep). The method has been previously used by our group in order to study the viscoelasticity behaviour of PolyCAs [13]. PolyCAs are used as bio-adhesives, which undergoes dynamic stress. Viscoelasticity experiments help us characterize their elasticity and relaxation patterns. Under these conditions, the strain rate in a viscoelastic material will initially change linearly, and then fade off non-linearly to a yield point. Primarily elastic materials show steep, but linear change (stress vs strain) and then break off at the yield point without much elongation. Primarily viscous materials show a gentle stress vs strain response (less slope) and often nonlinear after the yield point. Poly alkyl CAs are primarily elastic materials and have a higher storage modulus (G0 ). On the other hand poly alkoxy CAs are primarily soft viscous materials and have a higher loss modulus (G00 ). Most of the alkoxy CA polymers does not reveal a yield point under the measured shear. Our results are in agreement with previously reported work on alkoxy CAs [9]. The viscoelasticity studies reveal a strong correlation between nature of the side chain and the mechanical behaviours of CA polymers. Based on the mechanical properties the CA polymers may be classified (i) those G0 > G00 (primarily elastic, alky CAs) and (ii) G0 < G00 (primarily viscous, alkoxy CAs). Poly octyl CA has intermediate viscoelastic property G0  G00 . Also, the hydrophilic alkoxy CA polymers attract more water which may have also influenced the higher plasticizing effects. PolyCAs are also thermoplastic polymers that operate within a specific range of temperature. In general, the alkoxy CA showed a lower Tg. The O---O repulsion of the alkoxy groups creates more free volume and reduces the Tg. This observation may be correlated to the enhanced flexibility as observed during viscoelasticity measurements. Often flexibility of polymers is associated with reduced Tg. For most of the PolyCAs flexibility is correlated to lower Tg, but not for every polymer. Most of the CA polymers decompose/

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011

Y. Barkan et al. / Acta Biomaterialia xxx (2016) xxx–xxx

depolymerise at their melting temperature. In some cases, almost the entire polymer is decomposed. The comparative analysis of PolyCAs presented here reveals that the structure of the CA ester side chains affects the degradation, mechanical, and thermal properties significantly. Interestingly, alkoxy CA polymers have low molecular weights, enhanced hydrolytic degradation, enhanced plasticity, low glass transition and thermal stability. On the other hand, alkyl CA polymers show higher molecular weights, slow hydrolytic degradation, low plasticity, high glass transition and enhanced thermal stability. Alkoxy and PEG cyanoacrylates show fast degradation; these polymers are very soft and flexible. We believe they may be used as co-monomer blend with n-butyl or 2-octyl cyanoacrylate. The resulting co-polymer should have ideal plasticity without compromising the mechanical strength. Moreover, they may replace the use of plasticizers that may leach out and cause unwanted side effects. 5. Conclusion Overall, we present synthesis, degradation, comparative mechanical and thermal analysis of polyCAs. From this study, we may classify these polymers as soft polyCAs (G00 > G0 ) and hard polyCAs (G0 > G00 ). In general, alkoxy CAs are softer malleable polymers and alkyl CAs are harder and more brittle polymers, some of which with enhanced mechanical strength. Poly octyl CA shows an optimum mechanical property with appropriate balance of mechanical strength and plasticity. The study also compiles the thermal properties of polyCAs and degradation behaviour. Commercial CA adhesives are always blended with plasticizers, co-monomers, or crosslinkers to tailor their mechanical, thermal, and biological properties. The co-monomers form copolymers and impart balanced mechanical properties characteristic of both the monomers. This study should be the basis of choosing the correct monomer combination to tailor appropriate mechanical, thermal, or biomedical properties of CAs. Acknowledgements Arijit Basu acknowledges Planning and Budgetary Commission, Government of Israel for providing postdoctoral fellowship. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2016.11. 011. References [1] S. Gao, Y. Xu, S. Asghar, M. Chen, L. Zou, S. Eltayeb, M. Huo, Q. Ping, Y. Xiao, Polybutylcyanoacrylate nanocarriers as promising targeted drug delivery systems, J. Drug Target. 23 (2015) 481–496. [2] C. Vauthier, C. Dubernet, E. Fattal, H. Pinto-Alphandary, P. Couvreur, Poly (alkylcyanoacrylates) as biodegradable materials for biomedical applications, Adv. Drug Deliv. Rev. 55 (2003) 519–548. [3] G. Groeneveld, S. Kuijer, M. de Puit, Preparation of cyanoacrylate derivatives and comparison of dual action cyanoacrylate formulations, Sci. Justice 54 (2014) 42–48. [4] X. Du, J.S. Li, L.X. Li, P.A. Levkin, Porous poly (2-octyl cyanoacrylate): a facile one-step preparation of superhydrophobic coatings on different substrates, J. Mater. Chem. A 1 (2013) 1026–1029. [5] H.W. Coover, D.W. Dreifus, J.T. O’Connor, Cyanoacrylate Adhesives, in: I. Skeist (Ed.), Handbook of Adhesives, Springer, US, 1990, pp. 463–477.

11

[6] K. Andrieux, P. Couvreur, Polyalkylcyanoacrylate nanoparticles for delivery of drugs across the blood–brain barrier, Wiley Interdiscipl. Rev.: Nanomed. Nanobiotechnol. 1 (2009) 463–474. [7] D. Brambilla, J. Nicolas, B. Le Droumaguet, K. Andrieux, V. Marsaud, P.-O. Couraud, P. Couvreur, Design of fluorescently tagged poly (alkyl cyanoacrylate) nanoparticles for human brain endothelial cell imaging, Chem. Commun. 46 (2010) 2602–2604. [8] E. Sulheim, H. Baghirov, E. Haartman, A. Bøe, A.K. Åslund, Y. Mørch, Davies C. de Lange, Cellular uptake and intracellular degradation of poly (alkyl cyanoacrylate) nanoparticles, J. Nanobiotechnol. 14 (2016) 1. [9] B. Mizrahi, C.F. Stefanescu, C. Yang, M.W. Lawlor, D. Ko, R. Langer, D.S. Kohane, Elasticity and safety of alkoxyethyl cyanoacrylate tissue adhesives, Acta Biomater. 7 (2011) 3150–3157. [10] K.H. Cheung, J. Guthrie, M.S. Otterburn, J.M. Rooney, The dynamic mechanical properties of poly (alkyl 2-cyanoacrylates), Die Makromolekulare Chemie 188 (1987) 3041–3046. [11] Y.-C. Tseng, S.-H. Hyon, Y. Ikada, Modification of synthesis and investigation of properties for 2-cyanoacrylates, Biomaterials 11 (1990) 73–79. [12] H. Porter, Glass transition temperatures of poly (alkyl ayanoacrylates), J. Appl. Polym. Sci. 17 (1973) 3509–3514. [13] A. Basu, I. Veprinsky-Zuzulia, M. Levinman, Y. Barkan, J. Golenser, A.J. Domb, PEG-biscyanoacrylate crosslinker for octyl cyanoacrylate bioadhesive, Macromol. Rapid Commun. 37 (2016) 251–256. [14] A. Bentolila, J. Totre, I. Zozulia, M. Levin-Elad, A.J. Domb, Fluorescent cyanoacrylate monomers and polymers for fingermark development, Macromolecules 46 (2013) 4822–4828. [15] M.G. Han, S. Kim, Controlled degradation of poly (ethyl cyanoacrylate-comethyl methacrylate)(PECA-co-PMMA) copolymers, Polymer 50 (2009) 1270– 1280. [16] T. Nash, The colorimetric estimation of formaldehyde by means of the Hantzsch reaction, Biochem. J. 55 (1953) 416. [17] H. Ben-Bassat, D.V. Vardi, A. Gazit, S.N. Klaus, M. Chaouat, Z. Hartzstark, A. Levitzki, Tyrphostins suppress the growth of psoriatic keratinocytes, Exp. Dermatol. 4 (1995) 82–88. [18] K. Tanaka, A. Taura, S.-R. Ge, A. Takahara, T. Kajiyama, Molecular weight dependence of surface dynamic viscoelastic properties for the monodisperse polystyrene film, Macromolecules 29 (1996) 3040–3042. [19] W. Vezin, A. Florence, In vitro heterogeneous degradation of poly (n-alkyl a-cyanoacrylates), J. Biomed. Mater. Res. 14 (1980) 93–106. [20] Y.J. Lee, G.B. Jung, S. Choi, G. Lee, J.H. Kim, H.S. Son, H. Bae, H.K. Park, Biocompatibility of a novel cyanoacrylate based tissue adhesive: cytotoxicity and biochemical property evaluation, PLoS ONE 8 (2013) e79761. [21] Z. Gui, Y. Xu, Y. Gao, C. Lu, S. Cheng, Novel polyethylene glycol-based polyester-toughened polylactide, Mater. Lett. 71 (2012) 63–65. [22] S. Jain, A. Bajpai, Designing polyethylene glycol (PEG)–plasticized membranes of poly (vinyl alcohol-g-methyl methacrylate) and investigation of water sorption and blood compatibility behaviors, Des. Monomers Polym. 16 (2013) 436–446. [23] S.M. Martelli, J.B. Laurindo, Chicken feather keratin films plasticized with polyethylene glycol, Int. J. Polym. Mater. 61 (2012) 17–29. [24] K. Sungsanit, N. Kao, S. Bhattacharya, Properties of linear poly (lactic acid)/ polyethylene glycol blends, Polym. Eng. Sci. 52 (2012) 108–116. [25] S.Q. Tan, R. Malhotra, J.C. Allen, T.C. Tan, A pilot study of 2-octyl cyanoacrylate (DermabondTM) versus conventional suture skin closures for caesarean sections, Proc Singapore Healthcare 23 (2014) 312. [26] C.C. Sung, K. Mariwalla, Use of 2-octyl cyanoacrylate to obviate daily wound care after mohs surgery, Dermatol. Surg. 41 (2015) 294–296. [27] G. Tripodo, C. Wischke, A. Lendlein, Highly Flexible Poly (Ethyl-2Cyanoacrylate) Based Materials Obtained by Incorporation of Oligo (Ethylene Glycol) Diglycidylether, Macromol Symp: Wiley Online Library, 2011, pp. 49– 58. [28] I. Szanka, A. Szanka, J.P. Kennedy, Rubbery wound closure adhesives. II. initiators for and initiation of 2-octyl cyanoacrylate polymerization, J. Polym. Sci., Part A: Polym. Chem. 53 (2015) 1652–1659. [29] D.S. Johnston, D.C. Pepper, Polymerisation via macrozwitterions, 1. Ethyl and butyl cyanoacrylates polymerised by triethyl and triphenylphosphines. Die, Makromol. Chem. 182 (1981) 393–406. [30] D.S. Johnston, D.C. Pepper, Polymerisation via macrozwitterions, 2. Ethyl and butyl cyanoacrylates polymerised by pyridine and polyvinylpyridine. Die, Makromol. Chem. 182 (1981) 407–420. [31] L. Spangberg, H. Rodrigues, K. Langeland, Biologic effect of dental materials: 6. Effects of isobutyl cyanoacrylate on HeLa cells in vitro, Oral Surg. Oral Med. Oral Pathol. 37 (1974) 438–440. [32] M. Okabe, K. Kitagawa, T. Yoshida, C. Koike, T. Katsumoto, E. Fujihara, T. Nikaido, Application of 2-octyl-cyanoacrylate for corneal perforation and glaucoma filtering bleb leak, Clin. Ophthalmol. (Auckland, NZ) 7 (2013) 649. [33] G. Yordanov, R. Skrobanska, A. Evangelatov, Entrapment of epirubicin in poly (butyl cyanoacrylate) colloidal nanospheres by nanoprecipitation: formulation development and in vitro studies on cancer cell lines, Colloids Surf. B Biointerfaces 92 (2012) 98–105.

Please cite this article in press as: Y. Barkan et al., Comparative evaluation of polycyanoacrylates, Acta Biomater. (2016), http://dx.doi.org/10.1016/j. actbio.2016.11.011