International Journal of Adhesion & Adhesives 42 (2013) 11–20

Contents lists available at SciVerse ScienceDirect

International Journal of Adhesion & Adhesives

0143-74

http://d

n Corr

E-m

journal homepage: www.elsevier.com/locate/ijadhadh

Synthesis and characterization of polyurethane sealants containing rosinintended for sealing defect in annulus for disc regeneration

Pilar Carbonell-Blasco a, Jose Miguel Martın-Martınez a, Iulian Vasile Antoniac b,n

a Adhesion and Adhesives Laboratory, University of Alicante, 03080 Alicante, Spainb Politehnica University of Bucharest, Department of Materials Science and Physical Metallurgy, Faculty of Materials Science and Engineering, 313 Independentei Street,

Building J, 060032 Bucharest, Romania

a r t i c l e i n f o

Article history:

Accepted 28 November 2012Thermoplastic polyurethanes containing rosin or mixtures of rosin and 1,4 butane diol in the chain

extender were proposed as potential sealants for defects in disc regeneration surgery. The polyurethane

Available online 10 December 2012

Keywords:

Polyurethane

Peel

Lap-shear

Thermal analysis

Viscoelasticity

96/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.ijadhadh.2012.11.011

esponding author. Tel.: þ40 744369288.

ail address: antoniac.iulian@gmail.com (I.V. A

a b s t r a c t

sealants were prepared by using the prepolymer method and different mixtures of rosin and 1,4 butane

diol were used as chain extenders. The existence of one carboxylic moiety in the rosin structure allowed

the reaction with the isocyanate end groups in the prepolymer during polyurethane synthesis, creating

additional urethane-amide hard segments. The polyurethanes were characterized by ATR-IR spectro-

scopy, differential scanning calorimetry (DSC), dynamic mechanical thermal analysis (DMA), thermal

gravimetric analysis (TGA), scanning electron microscopy, X-ray diffraction and laser confocal micro-

scopy. The adhesion of the polyurethane sealants were tested by T-peel test of leather/polyurethane

sealant/leather joints and by single lap-shear tests of aluminium/polyurethane sealant/aluminium

joints. Depending on the rosin content in the chain extender the structure of the polyurethanes was

different, i.e. more urethane and urethane-amide hard segments were created up to 50 eq% rosin in the

chain extender and separation of domains was prevailing in the polyurethanes with higher rosin

content. Furthermore, the addition of rosin caused an increase in the length of the polymer chains but a

decrease in the storage modulus was produced (particularly in the polyurethane containing 50 eq%

rosin), likely due to the bulky structure of the rosin as compared to the linear structure of 1,4 butane

diol, allowing the separation of the linear polyurethane chains. On the other hand, the melting of the

soft segments in the polyurethanes started at 40–57 1C and the addition of more than 50 eq% rosin in

the chain extender decreased the melting enthalpy. Moreover, the crystallinity of the polyurethanes

containing up to 50 eq% rosin showed lower number and smaller spherulites. Finally, the peel strength

increased in the joints made with the polyurethane sealants containing rosin whereas the adhesive

shear strength decreased when the polyurethane sealant contained 50 eq% rosin or less.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In the framework of the FP7 European project DISC REGEN-ERATION a novel non-invasive surgical approach is currentlyunder development [1]. The project is intended to solve the backpain problems caused by herniated disc proposing theregeneration of the intervertebral disc by means of a newhydrogel injection. Because of the injection of the hydrogel inthe disc, the defect made in the annulus pulposus should beeffectively seal to avoid hydrogel losses. It has been shown[2,3] that one limitation of the current disc regenerationtherapy is the lack of an effective sealing system of the defect inthe disc as an efficient sealing system has not been developed yet.

ll rights reserved.

ntoniac).

ln this study, new thermoplastic polyurethane sealants (PU) areproposed as potential sealants for the defect in disc regenerationtherapy.

Polyurethanes are among the most versatile materials that canbe used for construction of medical devices including hollowfibres membranes for haemodialysis, balloons, drug-eluting stentcoatings, and central-venous catheter probe covers, among other.They are tough, biocompatible, and hemocompatible [4,5]. Theycan be strong elastomers or rigid plastics, and they can beprocessed by extrusion, injection moulding, film blowing, solutiondipping, and two-part liquid moulding.

Melt-processable, or thermoplastic, polyurethanes, are usedextensively in medical devices. Thermoplastic polyurethanes arelong-chain linear polymers without cross-links. Their linearstructure allows the polyurethane to be melted to form parts ofdevices that can be resolidified. Thermoplastic polyurethanes areoften used to produce catheters (over-the-needle catheters,

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–2012

central-venous access catheters, multilumen catheters), in asym-metric membranes manufacture for dialysis, and in wound dres-sings used to make a covering that is impermeable to fluids andbacteria but allows moisture to permeate. Some polyurethaneemulsions intended for catheter device were synthesized fromurethane prepolymer extended in water using L-lysine, ethylene-diamine, and their mixture as chain extenders [6]. These emul-sions exhibited satisfactory freeze/thaw stability and the filmscast from the emulsions possessed excellent mechanical proper-ties. Although the swelling ratios of the polyurethanes immersedin water for 24 h were different, the polyurethane obtained withL-lysine possessed the smallest weight loss after hydrolysis.On the other hand, all polyurethanes showed antiblood coagulationin-vitro.

Polyurethanes are manufactured by reacting an isocyanate, amacroglycol or polyol, and a chain extender. Thermoplasticpolyurethanes are fully reacted materials, i.e. they do not containunreacted NCO moieties, and they are block copolymers with –(AB)n

– structure where A is the hard segment and B is the soft segment.The hard segments, formed by reaction of the isocyanate groupswith the chain extender, have low molecular weight and they arerigid and highly polar; on the other hand, the soft segmentscorrespond to long hydrocarbon chains of macroglycol, and theyare flexible and non-polar [7]. The segmented structure of thethermoplastic polyurethanes allows phase separation into hardand soft domains which determines the polyurethane properties.Thermoplastic polyurethane structure is made of long linearchains that interact by physical forces which can be destroyedby heating and/or by solvents.

Although thermoplastic polyurethanes have excellent proper-ties, in general, they lack of sufficient immediate adhesion at bodytemperature, i.e. 37 1C, to allow immediate attach of the sealantinto the defect and maintain it in the location of application.Immediate adhesion can be reached in thermoplastic polyur-ethanes by heating at temperature higher than 50 1C which isnot feasible in the sealing of the disc defect [8]. In this study, theaddition of rosin into the thermoplastic polyurethane sealantstructure is proposed, to impart immediate adhesion at bodytemperature.

Rosin is a natural tackifier obtained from pine exudates that iscommonly added in the formulation of surgical tapes to imparthigh initial adhesion and tack, and it is also used as an additive inglazing food applications. It has been shown elsewhere [9] thatthe addition of rosin to formulated thermoplastic polyurethaneadhesive was not successful due to poor miscibility causing phaseseparation. Therefore, rosin has to be incorporated within thepolyurethane structure during thermoplastic polyurethane synth-esis. Chemical composition of rosin consists in terpene structurebased on abietic, neoabietic and palustric acids, all of them havingone carboxylic functionality (Fig. 1). The existence of onecarboxylic moiety in the rosin structure allows the reaction withthe isocyanate moieties in the prepolymer during polyurethanesynthesis, anchoring the rosin molecules at the extreme of thepolyurethane chains.

Fig. 1. Chemical composition of rosin.

The synthesis of polyurethane foams containing rosin formedical applications has been reported earlier [10]. Rosin-basedpolyester polyols were synthesized by reacting rosin–maleicanhydride adduct, diethylene glycol, and ethylene glycol withand without adding adipic acid and phthalic anhydride, in thepresence of catalyst. Rigid polyurethane foams were preparedwith these rosin-based polyols and it was shown that they hadhigher compression strength, and similar dimensional stability at100 1C than those prepared with polyols without rosin derivative.Furthermore, the rosin-modified polyurethane foams exhibitedeven lower thermal conductivity and much higher activationenergies during pyrolysis.

There is little literature [11,12] dealing with the addition ofrosin during the synthesis of thermoplastic polyurethanes. Aran-Aıs et al. [11] reported the characterization of thermoplasticpolyurethanes (TPUs) prepared with the prepolymer method byusing chain extenders containing different mixtures (0–50 eq%) ofrosin and 1,4 butane diol. The TPUs were used as raw materials toprepare solvent-based polyurethane adhesives. The addition ofrosin contributed to the production of two types of hard seg-ments, producing an increase in the average molecular weight, anincrease in the viscosity, and improved rheological properties.On the other hand, the immediate adhesion to plasticized PVC inall joints was improved if the TPU contained rosin. Aran-Aıs et al.[12] also reported the synthesis and characterization of TPUscontaining different hard/soft segment ratios (1.05–1.4) andsynthesized with a mixture of 50 eq% rosin and 50 eq% 1,4 butanediol as chain extender. The addition of rosin increased the averagemolecular weight, more markedly in the TPUs containing higherhard/soft segment ratios, but the elastic and viscous modulidecreased. An increase in the hard/soft segment ratio of the TPUsretarded the kinetics of crystallization and increased the immedi-ate adhesion to PVC but the final adhesion decreased.

The structure of polyurethane elastomers containing rosin wasstudied by Sanchez-Adsuar et al. [13]. Rosin was used either as anadditive, mixed in the TPU solutions, or as a reactant in the chain-extension step of polymer synthesis. Rosin as an additive did notmarkedly change the polymer properties. On the contrary, the useof rosin in the chain-extension step lead to sharp increase ofviscosity and molar mass as well as improved rheological propertiesand changes in morphology as the crystalline regions were moreaffected than the amorphous ones. It was concluded that rosinmodified the organization of both the hard and the soft segments ofthe polymers. However, addition of rosin did not improve the lowtack of TPU, although as chain extender or co-chain extender(together with 1,4 butane diol) rosin allowed development ofsignificant initial adhesive strength.

Therefore, in this study thermoplastic polyurethanes contain-ing different amounts of rosin were prepared for potential use asdisc defect sealant. The synthesis procedure was modified fromthe previous methods existing in the literature and much widerrange of rosin–1,4 butane diol mixtures was tested in order toprovide an adequate performance at 37 1C. The TPUs preparedwere characterized by using several experimental techniques,including adhesion measurements.

2. Experimental

2.1. Materials

The synthesis of the polyurethane sealants was carried out byusing the prepolymer method (Fig. 2). The hard to soft segmentratio (NCO/OH) used was 1.05. The prepolymer was prepared byreacting the polyol and the isocyanate in a 500 ml glass flaskunder inert atmosphere, with continuous mechanical stirring.

Fig. 2. Scheme of the polyurethane sealants synthesis.

Table 1Nomenclature of the polyurethanes prepared with

different rosin content in the chain extender.

Polyurethane Amount of rosin (eq%)

100BD 0

75BDþ25RR 25

50BDþ50RR 50

25BDþ75RR 75

100RR 100

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–20 13

First, the solid isocyanate (4,40-diphenylmethanediisocyanate,98 wt% purity, supplied by Sigma-Aldrich) was introduced intothe reactor, and once it was melted, the polyol (polyadipate of1,4-butanediol—Mw¼2500 Da, supplied by Synthesia EspanolaS.A., Barcelona, Spain) was added under continuous stirring andallowed to react. The rate of reaction was checked by titrationwith dibutylamine according to the European standard EN1242:1998. Once the desired NCO/OH ratio was reached, thechain extender was added and allowed to react for 5 min.Polyurethane annealing was performed for 18 h at 65 1C. Severalchain extenders were used: 1,4 butane diol (99 wt% purity,supplied by Sigma-Aldrich, Barcelona Spain), rosin (supplied byLa Union Resinera Espanola, Madrid, Spain), and several mixturesof 1,4 butane diol and rosin. Table 1 shows the nomenclature ofthe thermoplastic polyurethanes and their composition. The mainproperties of rosin are molecular weight of 462 Da, softeningpoint of 63 1C and acid number of 145 mg KOH/g.

For most of the experiments, polyurethanes films were pre-pared by placing solutions of 20 wt% solid polyurethane in methylethyl ketone in Teflon mould, allowing the removal of the solventat room temperature for 1 week.

2.2. Experimental techniques

2.2.1. Infrared spectroscopy (ATR-IR spectroscopy)

The ATR-IR spectra of the polyurethane films were obtained ina Bruker Tensor 27 spectrometer (Bruker Optik GMbH, Ettlingen,Germany), provided with Fourier transform analysis (FTIR).The measurements were carried out by using the attenuated totalreflectance (ATR) technique, and the Golden Gate single reflectiondiamond was used as prism. An incidence angle of the laser beamof 451 was used and 60 scans were averaged with a resolution of4 cm�1.

2.2.2. Molecular weight

The molecular weight and molecular weight distribution of thepolyurethane films were determined by gel permeation chroma-tography (GPC) performed in GPC GL1100 (Agilent TechnologiesEspana S.L, Las Rozas, Spain) instrument equipped with HT4 and

HT2 columns and refraction index detector. The eluent wasdimethylformamide (DMF), the flow rate was 1 ml/min, theoperation temperature was set to 25 1C and the equipment wascalibrated with polystyrene standards in DMF.

2.2.3. Dynamical mechanical thermal analysis (DMA)

The viscoelastic properties of the polyurethanes were mea-sured in a TA DMA Q800 instrument (TA Instruments, New Castle,DE, USA), using the single cantilever bending mode. The experi-ments were carried out by heating the sample from �100 1C to80 1C, using a heating rate of 5 1C/min, a frequency of 1 Hz and astrain of 0.5%. All experiments were carried under nitrogenatmosphere (flow rate: 100 ml/min).

2.2.4. Differential Scanning Calorimetry (DSC)

DSC experiments were carried out in a TA DSC Q100 instru-ment (TA Instruments, New Castle, DE, USA). Aluminium panscontaining 10–15 mg of sample were heated from �70 1C to 80 1Cunder nitrogen atmosphere (flow rate: 50 ml/min). The heatingrate was 10 1C/min. The first heating run was carried out toremove the thermal history of the samples. From the secondheating run, the glass transition temperature (Tg) and the meltingenthalpy (DHm) and melting temperature (Tm) of the polyur-ethanes were obtained.

2.2.5. Thermal gravimetric analysis (TGA)

TGA studies were carried out in thermal gravimetric analyserTA Q-500 (TA instruments, New Castle, DE, USA). About 20 mg ofpolyurethane were heated under nitrogen flow (100 ml/min) from30 1C to 500 1C using a heating rate of 10 1C/min.

2.2.6. Scanning electron microscopy (SEM)

A Philips XL30 ESEM-FEG (Eindoven, The Netherlands) wasused to analyse the topography of the polyurethanes. The poly-urethane films were gold coated and the electron beam energywas 20 kV.

2.2.7. Confocal laser microscopy

A Leica TCS SP2 microscope (Heidelberg, Germany) was usedto analyse the crystallinity (i.e. spherulites formation) and thesegmented structure of the polyurethanes. One drop of polyur-ethane solution in MEK was placed on glass microscope slide(dimensions of 76 mm�26 mm) and then covered by a smallglass cover slide. The samples were dried at room temperature for72 h before analysis.

2.2.8. X-ray diffraction

A JSO-Debyeflex 2002 instrument (Ahrensburg, Germany) wasused to quantify the crystallinity of polyurethane films; a coppercathode and a nickel filter were used. A scanning of 2y anglesbetween 51 and 901 was carried out.

2.2.9. Adhesion measurements

Because during performance of the polyurethane sealant in thedisc defect both peel and shear stresses will be present, adhesionmeasurements were carried out by means of T-peel and single-lapshear tests.

Adhesive strength (under peeling stresses) of the polyurethanesealants was obtained from T-peel test of leather/polyurethanesealant/leather joints. Calf leather pieces of dimensions 100 mm�20 mm�2 mm (supplied by Iresa Consultores, Elda, Spain) wereused to produce the joints. Leather was selected as substratebecause of its collagen nature which is somewhat similar tothat of the annulus pulposus in the disc defect. Before applyingthe polyurethane solution in MEK, the leather was previously

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–2014

roughened with a scouring machine (about 0.5 mm of leather wasremoved) to expose the corium of the leather to the surface.Approximately 0.45 ml polyurethane solution was applied ontoeach leather piece for filling the external porosity and 5 min lateradditional 0.92 ml polyurethane solution was applied on theleather pieces allowing the solvent to evaporate for 30 min underopen air. A polyurethane thickness of about 50 mm was obtainedon each leather strip. After solvent evaporation the polyurethanefilms on the two leather strips were heated suddenly at 80 1Cfor 10 s under infrared radiation (reactivation process). Theleather strips were immediately placed in contact and apressure of 0.4 MPa was applied for 10 s to achieve a suitablejoint. The T-peel strength was measured 5 min and 24 h afterjoint formation in Instron 4411 universal testing machine(Instron, Buckinghamshire, UK) by using a cross-head speedof 100 mm/min. The values obtained were the average of fivereplicates.

2954

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

Abs

orba

nce

(a.u

.)

3000

Wavenumbe

100BD

100RR

Fig. 3. (a) IR spectrum of the rosin. (b) IR spectra of 100BD (1,4 butane diol

The adhesive strength (under shear stresses) of polyurethanewas evaluated from single lap-shear test of aluminium/polyur-ethane sealant/aluminium joint. Aluminium 5754 test samplesof dimensions of 30 mm�150 mm�1.7 mm were used. Thissubstrate was selected because of in these joints adhesive failureis always obtained, i.e. after adhesion tests the adhesive remainon both sides of the failed joints; in this way, the intrinsicadhesion of the sealant can be obtained. Before applying thepolyurethane solution in MEK, the aluminium surface was rough-ened with Scotch Brites cloth followed by isopropanol wiping,allowing the solvent to evaporate for 30 min under open air. Then,0.3 g polyurethane solution in MEK was applied on an area of30�30 mm2 of one of the aluminium pieces to be joined. The otheraluminium piece was immediately placed in contact and a pressureof 0.1 MPa was applied for 72 h to achieve a suitable joint. The shearstrength was measured 72 h after joint formation in Instron 8516universal testing machine (Instron, Buckinghamshire, UK) by using a

734

958

1064

1164

1220

1258

13691416

15311598

1726

10002000

r(cm-1)

as chain extender) and 100RR (rosin as chain extender) polyurethanes.

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–20 15

cross-head speed of 10.2 m/min. The values obtained were theaverage of five replicates.

3. Results and discussion

Fig. 3a shows the ATR-IR spectrum of the rosin. Typical bandsof the carboxyl groups are found at 1258, 1695, 2342 and

Table 2Assignment of the main IR bands in the ATR-IR spectra of the polyurethanes

prepared with different rosin contents in the chain extender.

Wavenumber (cm�1) Assignment Group assignment

3339 nst N–H (associated) Urethane, urethane-amide

2954, 2872 nst C–H in CH2 Methylene

1726 nst C¼O Urethane (free)

1630 nst –CO–NH– Urethane, urethane-amide

1598 nst –CO–NH– Urethane-amide

1531 vst C–N, d N–H Urethane

1416 nst (sim) C–N Secondary amide

1258 nst (as) N–CO–O, nst C–O–C Urethane, polyol

1164, 1064, 958 nst C–O–C Polyol

734 d C–N Secondary amide

nst: stretching; nst (sim): symmetric stretching; nst (as): asymmetric stretching; d:

skeletal bending.

+

-CO

Fig. 4. Scheme of the reaction of the prepolymer with rosin and 1,4-butanediol sho

(adapted from reference [13]).

2361 cm�1; furthermore, the bands of terpene structure at 2868and 2930 cm�1 and of double CQC bond at 3073 cm�1 can bedistinguished. On the other hand, the chemical structure of thepolyurethanes was analysed by ATR-IR spectroscopy. Fig. 3bshows as typical examples the ATR-IR spectra of the polyur-ethanes obtained with 1.4 butane diol only (100BD) and withrosin only (100RR) in the chain extender. The assignment of themost characteristic IR bands in the polyurethanes is given inTable 2.

It has been shown earlier [13] that the incorporation of rosininto the structure of the thermoplastic polyurethane producedtwo types of hard segments, urethane hard segments producedby reaction of the isocyanate and the polyol, and by reaction ofthe prepolymer with 1,4-butanediol, and urethane-amide hard

2

wing the creation of urethane-amide and urethane hard segments, respectively

Table 3Molecular weights of the polyurethanes prepared with different rosin contents in

the chain extender.

Polyurethane Mn (Da) Mw (Da) Mz (Da)

100BD 98,000 184,000 338,000

75BDþ25RR 95,000 186,000 372,000

50BDþ50RR 98,000 277,000 1,211,000

25BDþ75RR 102,000 217,000 502,000

100RR 107,000 246,000 724,000

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–2016

segments produced by reaction of the prepolymer with thecarboxylic acid functionality of the rosin. Fig. 4 shows a schemeof the formation of both types of hard segments.

The ATR-IR spectra of the polyurethanes in Fig. 3b show thecharacteristic bands of the urethane at 3339 cm�1 (stretching ofNH group), 1726 cm�1 (stretching of CQO group), 1531 cm�1

(stretching of C–N group) and 1258 cm�1 (bending of N–C–Ogroup); the band at 3339 cm�1 may also correspond to theurethane-amide in the ATR-IR spectrum of 100RR polyurethaneand the band that should appear in urethane-amide at around1630 cm�1 cannot be distinguished because of the strong inten-sity of the CQO band. Furthermore, the bands at 2954, 2872 and1416 cm�1 correspond of the C–H groups of methylene and thebands of the polyol (polyadipate of 1,4 butane diol) appear at1258, 1220, 1164, 1064 and 958 cm�1. The bands of urethane-amide at 1598, 1416 and 734 cm�1 in the ATR-IR spectrum of100RR polyurethane cannot be distinguished properly due totheir low relative content in the polyurethane structure. On theother hand, the ATR-IR spectra of the 100BD and 100RR poly-urethanes are similar and the typical intense bands of rosincannot be distinguished, likely indicating the incorporation ofrosin into the polyurethane structure.

The molecular weights of the polyurethanes are given inTable 3. The average in number molecular weight (Mn) increasesslightly by increasing the amount of rosin in the chain extender,indicating that the number of polymer chains is not changingmuch by using 1,4 butane diol only, rosin only or mixtures of1,4-butanediol and rosin as chain extenders. However, the averagein weight molecular weight (Mw) increases by adding more than50 eq% rosin in the chain extender, the highest value corresponds tothe 50BDþ50RR polyurethane. Similarly, the average Z molecular

75BD+25RR

50BD+50RR100RR

100BD

25BD+75RR

10

100

1000

10000

1.0E5

Log

E´ -

MP

a

Temperature (°C)

75BD+25RR 50BD+50RR

25BD+75RR

100RR

100BD

0.0

0.1

0.2

0.3

Tan

Del

ta

-100 -80 -60 -40 -20 0 20 40 60 80

-100 -80 -60 -40 -20 0 20 40 60 80Temperature (°C)

Fig. 5. Variation of a storage modulus and b tan delta as a function of temperature

for the polyurethanes prepared with different amounts of rosin in the chain

extender.

weight (Mz) increases as the Mw values do. Therefore, the use ofmixtures of 1,4 butane diol and rosin as chain extenders causes anincrease in the length mainly but less in the number of polymerchains, particularly in the 50BDþ50RR polyurethane. As catalystwas not used during polyurethane synthesis, it seems that themixture of 50 eq% rosin and 50 eq% 1,4 butane diol as chainextender shows slower kinetics of polymerization that justifiesthe increase in Mw and Mz values.

The viscoelastic properties of the polyurethanes were obtainedby using DMA. Fig. 5a shows the variation of the storage modulusas a function of temperature for the thermoplastic polyurethanes.In the glassy region, the storage modulus of the polyurethanedecreases, in general, by adding rosin in the chain extender, likelydue to the bulky structure of the rosin as compared to the linearstructure of 1,4 butane diol, allowing more separation of thelinear polyurethane chains. In the rubbery plateau, a similar trendis found. On the other hand, the melting of the soft segments inthe polyurethanes starts at 40–57 1C and the addition of morethan 50 eq% rosin in the chain extender displaces the starting ofthe melting at higher temperature as compared to 100BD poly-urethane (without rosin in the chain extender). It is interestingthat the melting temperature of the 100RR polyurethane (withonly rosin as chain extender) is similar to that of 100BD (withonly 1,4 butane diol as chain extender), indicating that the use ofmixtures of rosin and 1,4 butane diol as chain extenders changesthe polyurethane structure differently.

Fig. 5b shows the variation of tan delta as a function of thetemperature for the thermoplastic polyurethanes. The maximumin the tan delta peak corresponds to the glass transition of thepolyurethane and the melting of the soft segments can be noticedby the sudden increase in tan delta values at high temperature.

-25

-20

-15

-10

Tg(°

C)

Rosin content in chain extender (eq%)

0.1

0.12

0.14

0.16

0.18

0.2

0.22

0.24

0 20 40 60 80 100 120

0 20 40 60 80 100 120

tan

delta

Rosin content in chain extender (eq%)

Fig. 6. Variation of the maximum of tan delta and glass transition temperature

(Tg) of the polyurethanes as a function of the rosin content in the chain extender.

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–20 17

The melting of the soft segments in the polyurethanes preparedwith more than 50 eq% rosin in the chain extender is displaced tohigher temperature and the melting is less marked than for theother polyurethanes; therefore, the degree of phase separation inthe polyurethane changes by using mixtures of rosin and 1,4butane diol as chain extenders. Fig. 6 shows the variation of themaximum in tan delta curves and the glass transition tempera-ture of the soft segments in the polyurethanes as a function ofthe rosin content in the chain extender. The lower the value of themaximum tan delta, the stronger is the interaction betweenthe polymer chains, and therefore the addition of rosin in the

100BD

75BD+25RR

50BD+50RR

25BD+75RR

100RR

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Hea

t Flo

w (W

/g)

Temperature (°C)

EXO

44

45

46

47

48

49

50

51

52

53

Tm (

°C)

Rosin content in chain extender (eq%)

15

20

25

30

35

40

-80 -60 -40 -20 0 20 40 60 80

0 20 40 60 80 100 120

0 20 40 60 80 100 120

ΔHm

(J/g

)

Rosin content in chain extender (eq%)

Fig. 7. (a) DSC thermograms of the polyurethanes prepared with different

amounts of rosin in the chain extender. Second heating run. (b) Variation of the

melting temperature (Tm) and melting enthalpy (DHm) of the polyurethanes as a

function of the rosin content in the chain extender.

chain extender decreases the interactions between the polymerchains more importantly for rosin contents lower than 50 eq%.Furthermore, the glass transition temperature of the polyurethanesdecreases by adding 25–50 eq% rosin in the chain extender, but it issimilar to that of 100BD in the polyurethanes prepared with chainextender having higher rosin content. Therefore, depending on therosin content in the chain extender, the structure and the degree ofphase separation in the polyurethanes varies, i.e. the addition of25–50 eq% rosin produces greater degree of phase separation andgreater influence on the viscoelastic properties of the polyurethanesthan the addition of 75–100 eq% rosin does.

The structure of the polyurethanes was also studied with DSC.Fig. 7a shows the DSC thermograms of the polyurethanes corre-sponding to the second heating run and they show the glasstransition of the soft segments at low temperature (around�50 1C) and the melting of the soft segments at 45–52 1C. Coldcrystallization only appears in 50BDþ50RR and 100RR polyur-ethanes indicating that the more rosin in the chain extenderfavours the crystallization of the polyurethanes. The meltingtemperature and the melting enthalpy of the polyurethanesdecrease more when 25–50 eq% rosin is added in the chainextender (Fig. 7b). The higher is the rosin content in the chainextender, the lower is the melting temperature, particularly inthe 50BDþ50RR polyurethane (Fig. 7b). The trend in meltingtemperature of the polyurethanes differs from that obtained inDMA experiments, likely due to the different geometries and thedynamic stresses applied to the polyurethane films in DMA andDSC experiments. However, the evolution of the melting processas a function of the rosin content in the chain extender is similarwhen it is measured with DSC and DMA.

The thermal degradation of the hard and soft segments and thethermal stability of the polyurethanes were measured with TGA.It has been established [14] that the hard segments decompose atlower temperature than the soft segments, and furthermoreurethane decomposes at lower temperature than urea hardsegments. All polyurethanes show relatively similar thermalstability, showing the main decomposition at 404–409 1C(Table 4) which is due to the decomposition of the soft segments.In fact the polyadipate of 1,4 butane diol decomposes at 401 1C(data not shown here). The 100BD (1,4 butane diol as chainextender) and 100RR (rosin as chain extender) polyurethanesshow only the decomposition of the soft segments but thepolyurethanes prepared with mixtures of 1,4 butane diol androsin chain extenders exhibit one additional decomposition at357–369 1C due to the new hard segments produced in thepolyurethanes (Table 4). The temperature and weight loss ofthe hard segments of the polyurethanes increase by increasing therosin content in the chain extender up to 50 eq%; the temperatureand weight loss for the 25BDþ75RR polyurethane are similar tothat for 50BDþ50RR polyurethane (Table 4). Therefore, the use ofmixtures of 1,4 butane diol and rosin chain extenders creates newand higher amount of hard segments in the polyurethane, thehighest content corresponds to the 50BDþ50RR polyurethane.Consequently, the degree of phase separation in the polyurethanes

Table 4Main decompositions in the TGA thermograms of the polyurethanes prepared

with different amounts of rosin in the chain extender.

Polyurethane T1

( %oC)

Weight loss1

(wt%)T2

( %oC)

Weight loss2

(wt%)Residue

100BD – – 409 99 1

75BDþ25RR 357 12 404 84 4

50BDþ50RR 369 21 405 76 3

25BDþ75RR 367 21 406 77 2

100RR – – 405 98 2

100BD 75BD+25RR 50BD+50RR

25BD+75RR 100RR

Fig. 8. SEM micrographs of the polyurethanes prepared with different amounts of rosin in the chain extender.

50µm

75BD+25RR100BD 50BD+50RR

25BD+75RR 100RR

Fig. 9. Confocal laser micrographs of the polyurethanes prepared with different amounts of rosin in the chain extender.

Table 5Mean diameter of the spherulites in the polyur-

ethanes prepared with different amounts of rosin

in the chain extender.

Polyurethane Spherulite diameter (lm)

100BD 32

75BDþ25RR 18

50BDþ50RR 15

25BDþ75RR 14

100RR 27

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–2018

prepared with mixtures of 1,4 butane diol and rosin chain extendersis more noticeable than in the 100BD an 100 RR polyurethanes.

The topography of the polyurethanes was analysed by SEM.Fig. 8 shows the spherulites in 100BD polyurethane, and theaddition of 1,4 butane diol and rosin mixtures in chain extendersproduces a gradual reduction in the number and size of thespherulites up to 50 eq% rosin. The 25BDþ75RR polyurethaneshows phase separation, and domains of 50–60 mm can bedistinguished. However, the 100RR polyurethane shows a morehomogeneous surface without domain separation. Therefore,depending on the rosin content in the chain extender, different

A

A

C

A

C

0

1

2

3

4

5

Peel

Str

engt

h (k

N/m

)

Rosin content in chain extender (eq%)

A

24h

5min

0 20 40 60 80 100 120

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–20 19

topography and crystallinity in the thermoplastic polyurethanesare produced.

The cristallinity of polyurethanes was evaluated by bothconfocal laser microscopy using polarized light and X-ray diffrac-tion. Fig. 9 shows the confocal laser micrographs of the polyur-ethanes in which the spherulites can be more clearlydistinguished with respect to the SEM micrographs. The micro-graphs of Fig. 9 show the same type and size (27–32 mm) ofspherulites in the 100BD and 100RR polyurethanes (Table 5), andthe distribution of the spherulites is quite homogeneous. How-ever, the addition of 1,4 butane diol and rosin mixtures in chainextenders reduces the size (14–18 mm) and changes the distribu-tion of the spherulites. On the other hand, the topography of thepolyurethanes changes depending on the rosin content in thechain extender. Whereas, the topography of 25BDþ75RR issimilar to that of 100BD, the one for 50BDþ50RR is morecompact and that for 25BDþ75RR shows differentiated domains.The spherulites in thermoplastic polyurethanes are soft segmentsand therefore, the increase in the hard segment content and thehigher molecular size of rosin with respect to 1,4 butane diolfavour the degree of phase separation that should also affect thecrystallinity.

X-ray diffractograms of the thermoplastic polyurethanes showtwo main diffraction peaks at 211 and 241 that have been ascribedto the polyadipate crystallites (i.e. soft segments) in thermoplasticpolyurethanes [15]. The intensity of the diffraction peaks inthermoplastic polyurethanes is related to their crystallinity.Fig. 10 shows the variation of the intensity of the X-ray diffractionpeak at 2y¼241 of the polyurethanes as a function of the rosincontent in the chain extender. The highest crystallinity corre-sponds to the polyurethane without rosin in the chain extender,i.e. 100BD, and the addition of up to 50 eq% rosin causes amoderate decrease in crystallinity. The 25BDþ75RR polyurethaneshows almost similar crystallinity than 100BD and it is higherthan for the other polyurethanes prepared with mixtures of rosinand 1,4 butane diol chain extenders. The lowest crystallinitycorresponds to the 100RR polyurethane that was prepared withrosin chain extender. Although in general the trend in crystal-linity of the polyurethanes prepared with mixtures of 1,4 butanediol and rosin chain extenders is similar when measured withconfocal laser microscopy and X-ray diffraction, the variationamong the polyurethanes is different because of the differentbackgrounds of the experimental techniques used in this study.100RR shows higher crystallinity when measured by confocallaser microscopy than by X-ray diffraction, and the reason shouldbe further investigated.

100

150

200

250

300

350

400

450

500

550

600

0 20 40 60 80 100 120

Inte

nsity

at 2

θ at

24°

(a.u

.)

Rosin content in chain extender (eq%)

Fig. 10. Variation of the intensity of the X-ray diffraction peak at 2y¼241 of the

polyurethanes as a function of the rosin content in the chain extender.

Finally, because adhesion to the disc defect is the targetedapplication of the thermoplastic polyurethane sealants preparedwith mixtures of 1,4 butane diol and rosin chain extenders, thatproperty was measured by using two different tests, T-peel andsingle lap-shear tests. The adhesive strength under peel stresseswas obtained from T-peel test of leather/polyurethane sealant/leather joints and it was measured 5 min and 24 h after jointformation (Fig. 11a). The initial peel strength (measured 5 minafter joint formation) in the joint made with 100BD is low andinsufficient to maintain the sealant in the application site whensubmitted to peel stresses; an adhesion failure of the sealant tothe leather is obtained. However, the addition of up to 50 wt%rosin in the chain extender increases two fold the initial adhesionand changes the locus of failure to cohesive failure in theadhesive, i.e. the cohesion of the sealant determines its adhesionto leather. On the other hand, the highest initial peel strength isproduced in the joint made with the polyurethane sealant pre-pared with rosin chain extender and a failure in the adhesive isproduced: the lowest adhesive strength corresponds to the jointobtained with 25BDþ75RR polyurethane that shows an adhesionfailure. Similar trends in adhesive strength values and loci offailure were obtained in the adhesion measured 24 h after jointformation, indicating that the adhesion of all sealants is reachedin a short time after joint formation, i.e. the sealant can bepotentially used for sealing the disc defect, 50BDþ50RR and100RR particularly.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 20 40 60 80 100 120

Lap

Shea

r (M

Pa)

Rosin content in chain extender(eq%)

A

A

A

C

A

Fig. 11. Variation of (a) the peel strength values of leather/polyurethane sealant/

leather joints, and (b) the single lap-shear strength values of aluminium/poly-

urethane sealant/aluminium joints, as a function of the amount of rosin in the

chain extender. Locus of failure: C cohesion failure in the adhesive; A adhesion

failure to leather or aluminium.

P. Carbonell-Blasco et al. / International Journal of Adhesion & Adhesives 42 (2013) 11–2020

The variations in the peel strength of the joints made withpolyurethanes prepared with chain extenders having different rosincontents can be related to their structure. Thus, the crystallinitydecrease and the hard segment content increase in the polyur-ethanes containing up to 50 eq% rosin, and therefore the resistanceagainst peel stresses should be higher. The 25BDþ75RR polyur-ethane is an exception as phase separation is produced, but neitherits high crystallinity nor its high hard segment content are deter-mining its peel adhesion. The polyurethane prepared with rosinchain extender has somewhat similar structure than the one madewith 1,4 butane diol chain extender, but its soft segment contentand degree of phase separation is higher, as well as the molecularweight is higher too; therefore, high elastomeric properties can beexpected and higher adhesive strength against peel stresses too.

The adhesive strength of the polyurethane sealants undershear stresses was evaluated from single lap-shear test of alumi-nium/polyurethane sealant/aluminium joints, 72 h after jointformation. Fig. 11b shows the variation of the shear strengthvalues of the aluminium/polyurethane sealant/aluminium jointsas a function of the rosin content in the chain extender. In general,the trend in the shear strength values is opposite to the oneprovided by peel strength values because of the greater influenceof the crystallinity of the polyurethanes in their resistance againstshear stresses; the higher hard segments content improves morethe resistance of the polyurethane to shear stresses. According toFig. 11b, the increase in the rosin content in the chain extender upto 50 eq% decreases the shear strength, and for the joints madewith 25BDþ75RR and 100RR higher shear strengths are obtained(although they are lower than in the polyurethane prepared with1,4 butane diol chain extender).

4. Conclusions

The addition of rosin as part of the chain extender in thesynthesis of thermoplastic polyurethanes contributed to createnew urethane-amide hard segments. Depending on the rosincontent in the chain extender, different structures, degrees ofphase separation and some unexpected properties were obtained.The best balance in properties was obtained in the thermoplasticpolyurethane prepared with 50 eq% rosin and 50 eq% 1,4 butanediol chain extender. The incorporation of rosin caused theseparation of the linear chains in the polyurethane leading toreduced crystallinity and size of the spherulites, higher average inweight molecular weights were obtained and lower meltingenthalpy was measured. However, the structure of the polyur-ethane containing 75 eq% rosin and 25 eq% 1,4 butane diol chainextender was dominated by the formation of domains mainly andits properties were unexpected. Finally, improved initial adhesion

was obtained in joints made with the thermoplastic polyurethanesealants making the polyurethanes prepared with mixtures ofrosin and 1,4 butane diol chain extenders very promising forsealing the defect in disc regeneration surgery.

Acknowledgements

Financial support was provided by European Commission FP7Project ‘‘Disc Regeneration’’ (Grant no. NMP3-LA-2008-213904),MINECO (MAT2010-19904 and PET2008-0264 projects), andGeneralitat Valenciana (ACOMP2011/254 project). Authors thankSynthesia Espanola S.A. (Barcelona, Spain) for providing the polyoland La Union Resinera Espanola S.A. (Madrid, Spain) for providingthe rosin used in this study.

References

[1] DISC REGENERATION Project. European Commission. Contract no. NMP3-LA-2008-213904.

[2] Hui JHP, Azur M, Lee EH. Stem cell therapy in orthopaedic surgery: currentstatus and ethical considerations. Malays Orthop J 2009;3(1):4–12.

[3] Brisby H, Tao H, Ma DD, Diwan AD. Cell therapy for disc degeneration:potentials and pitfalls. Orthop Clin North Am 2004;35:85–93.

[4] Seifalian AM, Salacinski HJ, Tiwari A, Edwards A, Bowald S, Hamilton G. Invivo biostability of a poly(carbonate-urea)urethane graft. Biomaterials2003;24(14):2549–57.

[5] Hsu S-H, Sun S-H, Chen D-Ch. Improved retention of endothelial cells seededon polyurethane small-diameter vascular grafts modified by a recombinantRGD-containing protein. Artif Organs 2003;27(12):1068–78.

[6] Chen H, Jiang X, He L, Zhang T, Xu M, Yu X. Novel biocompatible waterbornepolyurethane using L-lysine as an extender. J Appl Polym Sci 2002;84:2474–80.

[7] Szycher M, Siciliano AA, Reed AM. Biostable polyurethane elastomers. MedDevice Technol 1992;3(10):42–51.

[8] Kosinski LER, Rajendran GP, Reitz RR. US Patent 6,203,901. 2001.[9] Aran-Aıs F, Torro-Palau AM, Orgiles-Barcelo AC, Martın-Martınez JM. Addi-

tion of rosin acid during thermoplastic polyurethane synthesis to improve itsimmediate adhesion to PVC. lnt J Adhes Adhes 2005;25:31–8.

[10] Jin JF, Chen YL, Wang DN, Hu ChP, Zhu S, Vanoverloop L, et al. Structures andphysical properties of rigid polyurethane foam prepared with rosin-basedpolyol. J Appl Polym Sci 2002;84(3):598–604.

[11] Aran-Aıs F, Torro-Palau AM, Orgiles-Barcelo AC, Martın-Martınez JM. Synth-esis and characterization of new thermoplastic polyurethane adhesivescontaining rosin resin as an internal tackifier. J Adhes Sci Technol2000;14:1557–73.

[12] Aran-Aıs F, Torro-Palau AM, Orgiles-Barcelo AC, Martın-Martinez JM. Char-acterization of thermoplastic polyurethane adhesives with different hard/softsegment ratios containing rosin as an internal tackifier. J Adhes Sci Technol2002;16(11):1431–48.

[13] Sanchez-Adsuar MS, Papon E, Vlllenave J-J. Thermoplastic polyurethaneelastomers chemically modified by rosin. J Appl Polym Sci 2001;82:3402–8.

[14] Valero MF, Pulido JE, Hernandez JC, Posada JA, Ramirez A, Chaeng Z.Preparation and properties of polyurethanes based on castor oil chemicallymodified with yucca starch glycoside. J Elastom Plast 2009;41:223–44.

[15] Kovacevic V, Smit I, Hace D, Suceska M, Mudri I, Bravar M. Role of thepolyurethane component in the adhesive composition on the hydrolyticstability of the adhesive. Int J Adhes Adhes 1993;13:126–36.