1-s2.0-s0014305714001141-main tara

7/13/2019 1-s2.0-S0014305714001141-main TARA

1/13

Tara tannins as phenolic precursors of thermosettingepoxy resins

Chahinez Aoufa,, Sofia Benyahya b, Antoine Esnoufa, Sylvain Caillol b,Bernard Boutevin b, Hlne Fulcrand a

a UMR1083, INRA, Montpellier SupAgro, University Montpellier 1, 2 place Viala, 34060 Montpellier cedex 2, Franceb UMR CNRS 5253, Institut Charles Gerhardt, 8 rue de lcole normale, 34296 Montpellier cedex 5, France

a r t i c l e i n f o

Article history:

Received 25 February 2014

Received in revised form 28 March 2014

Accepted 31 March 2014

Available online 12 April 2014

Keywords:

Tara pods

Hydrolysable tannins

Functionalisation

Epoxy polymer

a b s t r a c t

Tara pods powder was used as a phenolic source in the synthesis of thermosetting epoxy

polymer. The tannase-assisted hydrolysis of galloylquinic acids contained in tara powder

allowed the determination of the tannins hydroxyl value (13.7 mmol/g powder). Then,

galloylquinic acids were reacted with epichlorohydrin and an aqueous solution of sodium

hydroxide in the presence of benzyltriethylammonium chloride as phase transfer catalyst

(PTC). The 1D and 2D NMR analyses of glycidylated products revealed the galloylquinic

esters hydrolysis and the dimerisation of the glycidylated gallic moities. The glycidylated

derivatives of tara tannins (GETT) were cured in epoxy polymer with isophorone diamine

(IPD). The glass transition temperature (Tg= 129 C) and the thermal resistance

(Td30= 294 C) of the resulting network were determined. Preliminary results showed that

this new epoxy polymer based onGETTdisplayed interesting properties which are close to

those of the epoxy polymer formulated with commercial diglycidyl ether of bisphenol A(DGEBA).

2014 Elsevier Ltd. All rights reserved.

1. Introduction

Tannins are phenolic compounds of relative high

molecular weight. They are classified as condensed and

hydrolysable tannins. The hydrolysable tannins are readily

hydrolysed by acids, alkalis or enzymes into a sugar or a

related polyhydric alcohol (polyol) and a phenolic carbox-ylic acid [1]. Depending on the nature of the phenolic

carboxylic acid, hydrolysable tannins are subdivided into

gallotannins and ellagitannins [1b,2]. They are biosynthes-

ised by galloyltransferases as defence compounds (against

Chinese gall and Turkey gall) and are accumulated in

mesophyl cell walls[3]. The simplest hydrolysable tannins

are gallotannins that are made up of gallic esters of

glucose, shikimic acid, quinic acid and quercitol among

others [4]. These tannins are found in various plants and

trees such as chestnut, oak, sumac, and tara [5]. Tara

(Caesalpinia spinosa) is a small leguminous tree native of

Peru and widely spread in Latin America, from Venezuela

to northern Chile [6]. The fruit of tara contains approxi-

mately 65% of pods [7] and 3238% of seeds. From thepods, tara powder (100200 mesh) is obtained by simply

mechanically milling and sifting the gross powder. After-

ward, this powder is mixed with water (45 parts of its

weight) and heated at 6570 C for 3040 min. After

decantation and filtration, tara extract is obtained by

atomisation. This ecological and economical procedure

produces tara extract with high tannins content [8]. It

has been reported that 4065% of the fruit composition

ofC.spinosacorresponds to gallotannins[7]. The chemical

structures of these gallotannins were widely investigated

http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034

0014-3057/ 2014 Elsevier Ltd. All rights reserved.

Corresponding author. Tel.: +33 499612454; fax: +33 499612857.

E-mail address: [email protected](C. Aouf).

European Polymer Journal 55 (2014) 186198

Contents lists available at ScienceDirect

European Polymer Journal

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l oc a t e / e u r o p o l j
http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034mailto:[email protected]://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://www.sciencedirect.com/science/journal/00143057http://www.elsevier.com/locate/europoljhttp://www.elsevier.com/locate/europoljhttp://www.sciencedirect.com/science/journal/00143057http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034mailto:[email protected]://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://-/?-http://-/?-http://crossmark.crossref.org/dialog/?doi=10.1016/j.eurpolymj.2014.03.034&domain=pdf

7/13/2019 1-s2.0-S0014305714001141-main TARA

2/13

by Haslam et al. [9]and Horler et al. [10] who demon-

strated that the principal components of tara tannin were

based on a galloylated quinic acid structure (Scheme 1).

Thus, they differ from members of hydrolysable tannins

group which are based upon a galloylated or ellagoylated

hexose. In galloylquinic acids, not only may gallic acid moi-

eties be linked to each of the four hydroxyls carried by qui-

nic acid but these may form aryl ester(s) (depsides) with

one or more additional gallic acid moieties [9a,11]. Chains

of up to three gallic acid moieties, of which two are depsi-

dic, have been reported [9b], and a given quinic acid may

bear more than one depside chain[12]. In such structures,

depsidic linkages may be either metaor para.

Recently, Giovando et al. [13]reported the MALDI-TOF

structural analysis of a tara tannins extract. It has been

found that this extract is composed of a series of oligomers

of polygallic acid (depsidic bonds) attached by an ester link

to one quinic acid hydroxyl. Furthermore, other polygallic

chains linked to one or two repeating units such as caffeic

acid and methylated quinic, methylated gallic and methyl-

ated caffeic acids have been found in very small amounts

[13]. The same research group has also demonstrated that

tara tannins are constituted of a small amount of an ellagic

unit that is surrounded by several gallic acid units. In addi-

tion, the ellagic acid contains an ester group which could

be generated from hydrolysis reactions, during the extrac-

tion process[5].

Tara tannins find valuable applications in the food and

beverage industries to clarify and give astringency to wine,

tea, coffee, cacao, and other food. Thanks to their astrin-

gent properties and very light colour, tara tannins are

particularly appreciated in the tannery industry [14].

Moreover, the alkaline hydrolysis of tara pods yields nearly

25% of gallic acid [15]. This phenolic acid and its decarbox-

ylated form (pyrogallol) are extensively used in straining

leather and hair and also as ingredients of developer in

photography and printing inks [16]. Furthermore, they

possess a large range of biological activities, including,

antioxidant, antibacterial, antiviral, analgesic etc. [17],

allowing them to act as precursors for the commercial pro-

duction of drugs [18]. However, little attention has been

given to the exploitation of these gallotannins as substi-

tutes of phenolic compounds in thermosetting polymers

manufacturing. Indeed, in 1973, as a consequence of the

first oil crisis, Norsechem, a Norwegian paint group sub-

sidiary in Malaysia that produces phenol formaldehyde

resins, was forced to substitute 33 wt% of phenol with

chestnut tannins extract in their formulations [19]. Once

the first oil crisis had passed and the price of phenol

became affordable again, the phenol formaldehydechest-

nut tannin resin production was stopped. Recently, pheno-

lic resin wood panel adhesive based on chestnut tannins

has been developed by Spina et al. [20]. In the same way,

Garro et al. have used tara pods as starting material to pro-

duce phenol formaldehyde adhesives [14a]. Our labora-

tory, interested in developing of thermosetting epoxy

resins from bio-sourced polyphenols [21] is considering

tara tannins extract as a potential substitute of petro-

leum-based phenols in the formulation of epoxy polymers.

Indeed, epoxy resins are the prime constituents in many

adhesives, paints, coating, sealants and electronic materi-

als[22]. Nevertheless, almost 90% of the world production

of epoxy resins involves the use of non-renewable hazard-

ous compounds such as bisphenol A. Meanwhile, there is a

lack on the studies devoted to the use of tannins in the pro-

duction of epoxy resins.

Scheme 1. Supposed tara tannins chemical structure.

C. Aouf et al./ European Polymer Journal 55 (2014) 186198 187
http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

3/13

In our previous works, one of the constitutive monomer

of condensed tannins, named catechin, was reacted with

epichlorohydrin to provide the corresponding glycidylated

ether derivative. The curing of this prepolymer in the pres-

ence of reactive diluent led to a crosslinking epoxy net-

work [21c]. More recently, the alkaline-assisted allylation

of gallic acid (a building block of gallotannins) followed

by the epoxidation of the resulting double bonds by

meta-chloroperbenzoic acid, provided the tri-glycidylated

ether derivative which was involved in the formulation

of a novel bio-based epoxy thermoset [23]. Having demon-

strated the feasibility of the synthesis of bio-based epoxy

polymers from these phenolic models; the ability of natu-

ral tannin extracts (generally occurring as polymers with a

large structural diversity) to produce epoxy networks was

evaluated. For this purpose, the commercially available

tara powder (kindly supplied by Silvateam) was selected

to react with epichlorohydrin.

The goal of the work reported herein was to extrapolate

the functionalisation processes developed for phenolic

models (within our previous work) for the synthesis of

epoxy prepolymer from tara gallotannins. Then, the bio-

based prepolymer was formulated into thermosetting

epoxy polymer. Some thermal and mechanical properties

of the resulting network were compared to those of

bisphenol A based epoxy polymer. The use of a natural

phenolic extract in the formulation of epoxy resins will

pave the way for numerous industrial applications.

2. Materials and methods

2.1. Materials

Tara powder: the tara powder from tara pods (C. spinosa)

were kindly supplied by Silvateam s.p.a (Italy).

Chemicals: gallic acid (97.5%), epichlorohydrin (99.0%),

benzyltriethylammonium chloride (P98.0%), sodium

hydroxide (P98.0%), tannase from Aspergillus ficuum

(P150 U/g), isophorone diamine (IPD, P99.0%), HPLC

grade solvents (acetonitrile and methanol), formic acid

(P95.0%) were purchased from Sigma Aldrich (France).

DGEBA(D.E.R.352) (EEW = 5.66 mmol/g) (product reaction

of epichlorohydrin with bisphenol A and bisphenol F) was

supplied by Dow. UPLC water was prepared from distilled

water using a Milli-Q system (Merck-Millipore, France).

b-glucogallin (1-galloylglucose) was kindly provided by

Dr. G. Gross (ULM University, Germany). Gallic acid and

b-glucogallin were used as standards for the calibration.

Standard and tara tannins solutions: the tara tannins

solution was prepared by dissolving 5 mg of extract in

100 mL of distilled water (0.05 mg/mL). Standard solutions

of gallic acid and b-glucogallin were prepared with the

same mass concentration and injected in UPLC 7 times.

Each solution was prepared at 7 concentration levels

(initial concentration and diluted 1:2, 1:4, 1:10, 1:20,

1:40, and 1:100) to provide a range of signals suitable for

determining the molar relative response factor (MRRF).

UPLC-MS: the LC-DAD-ESI/MS consisted of an Acquity

UPLC (Waters, Milford, MA) coupled with a DAD and aBrucker Daltonics Ion trap mass spectrometer. 2 lL of

standard and gallotannins extract solutions were injected

via the autosampler onto a Nucleosil 120-3 C18 encaped

column (100 2.1 mm, 5 lm particle size, Machery-Nagel,Sweden), held at 38 C, with a constant flow rate of 550 lL/min. The mobile phase consisted of a combination of A

(H2O/HCOOH 99/1 v/v) and B (CH3CN/H2O/HCOOH 80/19/

1 v/v/v); initial 0.1% B; 05 min, 40% B linear; 58 min,

99% B linear, 89 min 0.1% B linear. The DAD was set at

280 nm (kmax of phenolic compounds). Mass spectra were

acquired using electrospray ionisation in the positive mode

and recorded in the range 701500 amu. A drying gas flow

of 12 L/min, a drying gas temperature of 200 C, a nebulizer

pressure of 44 psi, and capillary voltages of 5500 V were

used.

Concentration calculation by UPLC: it is well known that

the area of a spectral peak is proportional to the amount of

the substance that reaches the detector in LC instrument. A

response factor is obtained experimentally by analysing a

known concentration of the substance into the LC instru-

ment and measuring the area under the relevant peak.

The molar relative response factor (MRRFx

) of each pheno-

lic standard equals the area of the spectral peak (mAu)

divided by the molar concentration of this compound.

Thus, for the quantitation of phenolic compounds in gallo-

tannins extract, the molar concentration (Cx) of gallic acid

andb-glucogallin can be computed as:

Cx Ax=MRRFx

whereAxis the peak area of each phenolic compound cited

above.

NMR analyses: NMR spectra were acquired on VARIAN

Unity-Inova 500 MHz and Bruker 600 MHz liquid. All sam-

ples were dissolved in DMSO-d6. The chemical shifts were

reported on that of internal DMSO-d6

at 2.5 ppm for 1H and

39.5 ppm for 13C. Assignments of both proton and carbon

resonances, identification and structure characterisation

of products were performed using several NMR experi-

ments: homonuclear 1H and 13C 1D and 2D (COSY) and

heteronuclear 1H13C 2D (HSQC and HMBC).

2.2. Enzymatic hydrolysis of tara tannins

The tara tannins extract (0.5 mg) was dissolved in

100 mL of acetate buffer solution (200 mM, pH = 6.8).

Then, 1 mg of tannase from A. ficuum was dissolved in

1 mL of the buffered gallotannins solution and the mixture

was incubated at 37 C for 90 min. The reaction products

were analysed by UPLC-MS combined system.

2.3. General procedure for the glycidylation of gallic acid and

tara tannins extract

A 250 mL two-necked flask equipped with a condenser,

a septum cap and a magnetic stirring bar was charged with

3 g of gallic acid or tara tannins extract in epichlorohydrin

(4 M eq/OH), the suspension was heated at 100 C and

benzyltriethylammonium chloride (0.012 M eq/OH) was

added. After 1 h, the resulting solution was cooled to

30 C and the aqueous solution of NaOH 20 wt% (2 M eq/

OH) with the same previous amount of phase transfercatalyst (BnEt3NCl) were added. The mixture was stirred

188 C. Aouf et al./ European Polymer Journal 55 (2014) 186198
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

4/13

vigorously for 90 min. The organic layer was separated,

dried over MgSO4 and concentrated under vacuum. The

crude product (5.8 g from gallic acid and 4.7 g from tara

tannins extract) was purified by silica gel chromatography

using petroleum ether/ethyl acetate (PE/EA) solvent sys-

tem to yield following products:

3,4,5-Triglycidylether glycidyl benzoate 1:(PE/EA, 30/70),

colourless oil, 60% yield (10.6 mmol) from gallic acid;

530 mg/g of prepolymer from tara tannins extract. 1H

NMR (500 MHz, DMSO-d6) d= 2.622.86 (m, 8H, H1, H11

and H14), 3.293.38 (m, 4H, H2, H12 and H15), 3.903.94

(m, 3H, H130, H160), 4.08 (dd, J= 12.4, 6.3 Hz, 1H, H30),

4.26 (qd, J= 11.7, 2.9 Hz, 1H, H16), 4.44 (dd, J= 11.3,

2.1 Hz, 2H, H13), 4.63 (dd, J= 12.4, 2.5 Hz, 1H, H3), 7.30

(s, 2H, H6 and H10) ppm. 13C NMR (125 MHz, DMSO-d6)

d= 43.4 (C1), 43.6 (2C, C11), 43.8 (C14), 49.0 (C2), 49.7

(2C, C12), 50.1 (C15), 65.5 (C3), 70.0 (2C, C13), 74.0

(C16), 108.4 (2C, C6 and C10), 124.6 (C5), 141.6 (C8),

151.8 (2C, C7 and C9), 165.0 (C4) ppm. HRMS calc for

C19H23O9 [M+H]+: 395.1342, found 395.1341.

Oxiran-2-ylmethyl4-(2,3-dihydroxypropoxy)-3-(oxiran-

2-ylmethoxy)-5-(3,4,5-tris(oxiran-2-ylmethoxy)benzoyloxy)

benzoate 2: (PE/EA, 10/90 to 0/100), pale yellow oil, 23%

yield (2 mmol) from gallic acid; 188 mg/g of prepolymer

from tara powder. 1H NMR (600 MHz, DMSO-d6) d= 2.62

(dd, J= 4.5, 2.2 Hz, 1H, H4), 2.73 (m, 5H, H1, H4, H21 and

H24), 2.83 (dt, J= 9.3, 4.5 Hz, 4H, H1, H21 and H24), 3.30

(m, 1H, H5), 3.31 (m, 1H, H20), 3.33 (m, 1H, H23), 3.35

(m, 2H, H2), 3.85 (dd, J= 11.0, 5.7 Hz, 1H, H22), 3.92 (dd,

J= 11.5, 6.3 Hz, 2H, H3 and H6), 4.05 (dd,J= 12.3, 6.3 Hz,

1H, H19), 4.25 (dd, J= 11.7, 6.5 Hz, 1H, H60),4.31 (dd,

J= 11.5, 6.5 Hz, 1H, H25), 4.39 (m, 1H, H220), 4.42 (m, 1H,

H30), 4.56 (m, 1H, H250), 4.57 (d, J= 2.5 Hz, 2H, H27), 4.59

(m, 1H, H190), 4.65 (bs, 1H, H26), 7.16 (s, 1H, H15), 7.17

(s, 1H, H17), 7.27 (s, 2H, H8). 13C NMR (150 MHz, DMSO-

d6) d= 43.4 (C4), 43.5 (2C, C1), 43.6 (C24), 43.8 (C21),

49.0 (C5), 49.5 (C23), 49.6 (2C, C2), 50.1 (C20), 63.0

(C27), 64.7 (C25), 65.3 (C19), 70.0 (3C, C3 and C22), 70.7

(C26), 74.0 (C6), 106.6 (C15), 108.6 (2C, C8), 111.6 (C17),

121.2 (C16), 124.4 (C7), 137.4 (C13), 141.7 (C10), 142.9

(C12), 147.5 (C14), 151.8 (2C, C9),164.8 (2C, C18 and

C11). HRMS calc for C32H36O16 [M+H]+: 677.2003, found

677.1992.

Oxiran-2-ylmethyl 4-(3-chloro-2-(3,4,5-tris(oxiran-2-

ylmethoxy) benzoyloxy)propoxy)-3,5-bis(oxiran-2-ylmeth-

oxy)benzoate 3: (PE/EA, 10/90 to 0/100), pale yellow oil,

2% yield (0.18 mmol) from gallic acid; 209 mg/g of pre-

polymer from tara powder. 1H NMR (600 MHz, DMSO-d6)

d= 2.62 (m, 1H, H4), 2.702.77 (m, 5H, H1, H21 and

H25), 2.81 (m, 1H, H4), 2.84 (dd, J= 9.8, 5.1 Hz, 5H, H1,

H21 and H25), 3.33 (m, 5H, H2, H5 and H20), 3.34 (m,

1H, H24), 3.86 (m, 4H, H3 and H19), 3.91 (dd, J= 11.7,

6.1 Hz, 1H, H6), 4.06 (m, H, H23), 4.09 (bs, 2H, H13), 4.25

(m, 1H, H60), 4.33 (m, 4H, H30 and H190), 4.39 (m, 1H,

H14), 4.44 (m, 1H, H140), 4.60 (d, J= 12.3 Hz, 1H, H230),

5.43 (m, 1H, H12), 7.14 (s, 1H, H8), 7.16 (s, 1H, H80), 7.22

(s, 1H, H17), 7.25 (s, 1H, H170). 13C NMR (150 MHz,

DMSO-d6) d= 43.0 (C13), 43.4 (C25), 43.5 (C21), 43.6 (3C,

C1 and C21), 43.8 (C4), 49.0 (C24), 49.6 (4C, C2 and C20),

50.1 (C5), 65.5 (C23), 70.1 (4C, C3 and C19), 71.2 (C14),72.5 (C12), 74.0 (C6), 108.3 (2C, C17 and C17 0), 108.7 (2C,

C8 and C80), 124.4 (C7),124.5 (C18), 141.4 (C15), 141.9

(C10), 151.4 (2C, C16), 151.7 (2C, C9), 164.3 (C11),164.8

(C22). HRMS calc for C35H39O16Cl [M+H]+: 751.2027, found

751.2007.

2.4. Determination of the epoxy equivalent weight (EEW)

The epoxy equivalent of GETT prepolymer was deter-mined by chemical assay based on the reaction with an ali-

quot of standard pyridine hydrochloride in excess pyridine

at reflux and subsequent back titration with standard

methanolic sodium hydroxide using phenolphthalein as

indicator[24].GETT(0.21 g) was placed in a 100 mL flask.

A volume of 25 mL of 0.2 N solution of hydrochloric acid

(37%) in pyridine was added. The mixture was heated at

reflux (115 C) for 20 min. After cooling, the titration of

the excess of acid was carried out by a 0.6 N sodium

hydroxide solution in the presence of phenolphthalein (3

drops). A blank assay was performed under the same condi-

tions in the absence of pre-polymer. The epoxy equivalent

weight (EEW) was calculated by the following equation:

EEW p 10

3

A B CNaOH1

wherep is the prepolymer weight (g), A the NaOH volume

(mL) for blank, and B is the NaOH volume (mL) for

prepolymer.

2.5. General procedure for the formulation of GETT and DGEBA

The glycidylated derivatives GETTand DGEBA are liquid

and were formulated at room temperature and cured in a

silicon mould at 100 C for 12 h followed by 2 h at 250 C

to ensure complete curing of networks. The curing agentwas isophorone diamine (IPD), a cycloaliphatic diamine,

with an amine equivalent weight (AEW) of 43 g/eq (techni-

cal data sheet).GETT-IPDsystem was obtained by the for-

mulation of 2 g (16.97 mmol) of glycidyl ether of tara

tannins with 0.73 g (16.97 mmol) of IPD. The reference

DGEBA-IPD network was produced from the curing of 2 g

(11.3 mmol) of DGEBA (DER352) (EEW = 5.66 mmol/g

based on Technical data sheet) by 0.5 g (11.3 mmol) ofIPD.

2.6. Thermogravimetric analysis

TGA measurements were obtained on a TA Instruments

Q50 apparatus. The initial weight of each sample tested

was approximately 10 mg. Data were collected in nitrogen

atmosphere using a 10 C/min ramp from 20 to 580 C.

Experiments were carried out in triplicate. The tempera-

tures of 5% and 30% weight loss (Td5) and (Td30)

respectively, statistic heat-resistant index temperature

(Ts) and the percentage of residue at 575 C (%res) were

determined.

2.7. Dynamic mechanical analysis

DMA was performed with a Metravib DMA 25 using the

uniaxial stretching mode to determine the storage (E0

), andloss (E00) moduli as well as loss factory (tand) as a function

http://-/?-http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

5/13

of temperature. Samples (30 10 2 mm) were heated

from 100 C to 250 C using a heating rate of 3 C/min

in a forced convection oven using a nitrogen stream. The

sample was deformed sinusoidally with controlled strain

amplitude of 6.3 105 and 1.5 104 for GETT-IPD and

DGEBA-IPD respectively at a fixed frequency of 1 Hz.

Experiments were carried out in triplicate. Temperatures

of the relaxation processes associated with glass transition

temperatures were determined through the inflexion point

of the storage modulus E0 curve as well as the maximum

peak in tandcurves[25].

2.8. Differential scanning calorimetry

DSC measurements were performed under inert atmo-

sphere with a calorimetric NETZSCH DSC 200 F3 calibrated

with an indium standard.

The polymer was weighted in an aluminium hermetic

pan and consecutively placed in the measurement heating

cell. An empty pan was used as reference. All the samples

were heated under inert atmosphere from 50 to 250 Cat a heating rate of 10 C/min. Three runs were recorded

and the glass transition temperature (Tg) values were

measured during the second run and confirmed by a third

run. Tgs were calculated at the inflexion point of the heat

capacity jump. Experiments were carried out in triplicate.

3. Results and discussion

In order to determine the tannins composition of the

tara pods extract, the tara powder was dissolved indistilled water (0.05 mg/mL) and analysed by UPLC

DADMS (Ultra Performance Liquid Chromatography

Diode Array DetectorMass Spectrometer) combined

system. The DAD was set at kmax of phenolic compounds

(280 nm)[26]and mass spectra were acquired in positive

mode.

The UV profile and mass spectrum of the tara powder

are shown inFig. 1. The tara powder produced a relatively

complex UV chromatogram in which six oligomer groups

were identified by their pseudomolecular ions. Galloylqui-

nic acids (Q-G, m/z345), digalloylquinic acids (Q-2G, m/z

497), trigalloylquinic acids (Q-3G, m/z649), tetragalloyl-

quinic acids (Q-4G, m/z 801), pentagalloylquininc acids(Q-5G, m/z 953) and finally hexagalloylquininc acids

(Q-6G, m/z1105). All mass signals were separated from

Fig. 1. (a) UV profile at 280nm of tara tannins extract at 0.05 g/L in water and (b) ESI-MS positive ion mode spectrumof tara tannins extract in water in the3001200 Da mass range.

http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

6/13

each other by 152 amu corresponding to a galloyl unit.

Beyond Q-G derivatives, the additional galloyl moiety

may be directly attached to a quinic acid molecule or

involved in depsidic bonds.

The aimof this work was to functionalise thesegallotan-

nins by reaction with epichlorohydrin. Therefore, a deeper

characterisation of tara tannin structures to define more

accurately the linkage of each galloyl group (ester bond

with quinic acid or depsidic bond) is not of great impor-

tance. However, it is important to determine the hydroxyl

value (mmol of free OH per gram of powder) of tara tannins.

Indeed, during the glycidylation process, the hydrolysis of

both ester and depsidic bonds was observed. Moreover,

the reaction of the hydrolysis products with epichlorohy-

drin involves only the carboxyl and phenolic hydroxyls.

The less acid hydroxyl groups from hydrolysed quinic acid

and other non-phenolic components are not reactive in

the glycidylation procedure (these two aspects will be dis-

cussed more accurately in next sections). Therefore, to

establish the amount of epichlorohydrin needed for the

reaction, it is necessary to determine theamount of free car-

boxyl and phenolic hydroxyls present in the tara powder.

To determine the hydroxyl value, the enzymatic hydro-

lysis of galloyquinic acids contained in the extract was car-

ried out. Thus, tara powder was dissolved in a acetate

buffer solution (200 mM, pH = 6.8) and the hydrolysis reac-

tion was catalysed by tannase fromA. ficuum. Indeed, tan-

nin acyl hydrolase, commonly known as tannase is widely

used to catalyse the hydrolysis of ester and depside link-

ages in hydrolysable tannins releasing gallic acid and thestructure core[27].

After 90 min of incubation at 37 C (which corresponds

to the optimised reaction conditions), the hydrolysed tara

tannins produced a simple chromatogram at 280 nm in

which gallic acid was identified as a main component

along with a small amount of Q-G derivatives (Fig. 2).

Based on this chromatogram, gallic acid and Q-G deriva-

tives were quantified in our tannins extract using stan-

dards. Since quinic acid and glucose do not absorb at

280 nm, the concentration of Q-G derivatives was esti-

mated with respect to the synthetised 1-galloylglucose.

The quantification results expressed in mg of phenolic

derivatives per gram of tannin extract are summarised in

Table 1. These results indicate that the overall content of

phenolic compounds present in the tara powder accounted

for 706 mg/g, that is nearly 70% of tannins.

Additionally, the hydroxyl value corresponding to the

number of reactive OH per gram of gallotannins extract

was calculated and a value of 13.7 mmol OH/g of powder

Fig. 2. (a) UVprofile at280 nmof tannase hydrolysedtara tannins extract at0.05 g/L in water and (b) UVprofile at280 nmof gallic acid at0.05g/L inwater.

Table 1

The tara powder phenolic composition.

Phenolic

compound

[M+H]+

(m/z)

Phenolic

content (mg/g)

Hydroxyl valuea

(mmol OH/g extract)

Gallic acid 171.0 510.63 12.0

Q-G 345.0 195.11 1.7

Total 705.74 13.7

a This value was calculated based on molecular weights of gallic acid

and Q-G associated to the number of reactive hydroxyl groups

attached to each phenolic derivative.

http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

7/13

was found. This value includes the carboxylic and phenolic

hydroxyls of gallic acid and Q-G.

3.1. The functionalisation of galloylquinic acids contained in

tara powder

Due to the structural complexity of the galloylquinic

acids contained in tara powder, it seems challenging to

understand their reactivity towards epichlorohydrin and

then to characterise their functionalised products. There-

fore, gallic acid was used as a representative model for

the glycidylation reaction with epichlorohydrin.

3.1.1. Gallic acid glycidylation

The glycidylation reaction of gallic acid was carried out

according to the experimental conditions disclosed in

Tomitas invention [28]. The tetraglycidylated product 1

was obtained in 72% yield.

The chemical structure of compound 1 was easily estab-

lished by HRMS and NMR analyses (see experimental sec-

tion). Indeed, from 1H NMR spectrum, several aliphatic

signals arising from oxirane groups are identified. The

methylene and methyne ring proton signals appear in the

2.733.37 ppm range and the CH2O protons give reso-

nances signals in the 3.914.62 ppm spectral range. The

average number of oxirane groups per gallic acid ring, cal-

culated from the ratio of

1

H surface signal integrations, wasequal to 4. The hydroxyl groups (carboxylic and phenolic)

substitution occurred through a two steps mechanism

involving the ring opening of epichlorohydrin by the phe-nolate anion giving the chlorohydrin derivative. The latter

was subsequently transformed into glycidylated derivative

by the alkaline assisted intramolecular cyclisation [21a].

3.1.2. Galloylquinic acids glycidylation

By relying on the calculated hydroxyl value (13.7 mmol

OH/g of powder), the tannins extract was glycidylated

according to the reported procedure[28]. The functionali-

sation reaction yielded 3.4 g of crude product. This latter

was analysed by UPLCDADMS system (data not shown)

to give a complex UV profile at 280 nm with tree main

compounds at 3.2 min (m/z 395), 4.2 min (m/z 659) and

4.6 min (m/z731). Then, the crude product was purified

by silica gel chromatography in order to isolate (quite labo-

riously) major compounds 1, 2and 3 (Scheme 2). This puri-

fication did not allow the isolation of the products present

in small quantities.

Detailed structure characterisation of product 2 was

performed by NMR spectroscopy. The disappearance of

hydroxyl protons associated to the aromatic proton signals

displayed in the 1H NMR spectrum of 2 indicate the

presence of a symmetric aromatic ring (singlet at

7.27 ppm, 2 protons) and an asymmetric aromatic ring

(two singlets at 7.17 and 7.16 ppm, 1 proton) both substi-

tuted in their carboxylic and phenolic hydroxyls. More-over, the 1H signal surface integrations counted five

Scheme 2. Glycidylation of gallic acid and tara powder in the presence of PTC followed by alkaline treatment.

http://-/?-http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

8/13

glycidyl groups. The linkage positions of these groups on

the aromatic rings were then determined using 1H13C

long range correlations between the CH2O protons of

the glycidyl groups and quaternary carbons of the aromatic

rings as shown inFig. 3. Thus, it was found that hydroxyls

carried by C9, C10, C14 and C18 are substituted by methyl

oxirane groups. However, the CH2O protons which cor-

relate with the quaternary carbon C13 are attached to an

unshielded CH (H26). This suggests the oxirane ring

opening of this substituent and the formation of a dihydr-

oxypropoxy group. Nevertheless, this proton (H26) dis-

plays a higher chemical shift (4.65 ppm) compared to the

methyne protons of classical dihydroxy groups. Thisunshielding effect is probably due to the intramolecular

hydrogen bonding of the exchangeable hydroxyl group

linked to C26 [29]. Finally, no glycidyl nor dihydropropoxyl

substitutions were observed on OH (11) and OH (12) indi-

cating the linkage of C11 and OH (12) by a depside bond.

The postulated structure of compound 2 proposed in

Fig. 3was supported by HRMS analysis.

Using the same analytical tools, detailed structure of

product 3 has been established. NMR analyses revealed

the presence of two symmetric phenolic rings substituted

by six methyl oxirane groups in hydroxyls OH (9), OH

(10), OH (16) and OH (18) as shown in the HMBC spectrum

of3 (Fig. 4). Additionally, this 2D NMR method showed a

correlation between the carboxylic carbon C11 and meth-yne proton H12 which correlated with methylene protons

Fig. 3. A part of HMBC spectrum of compound 2 in DMSO-d6 showing correlations between methylene protons of the substituent groups and quaternary

carbons of aromatic rings.


7/13/2019 1-s2.0-S0014305714001141-main TARA

9/13

H13 and H14 (COSY spectrum is available in Supporting

data). Then, the heteronuclear correlation between C13

and H14 indicates that the two phenolic rings are linked

by an ester bond as displayed inFig. 4. Chemical shifts of

H13/C13 (4.09 ppm and 43.0 ppm respectively) suggest

that C13 is linked to a chlorine atom.

Interestingly, the formation of compound 1 as main

product demonstrates unequivocally the hydrolysis of the

initial galloylquinic acids during the functionalisation reac-

tion, releasing free gallic acid which is then fully glycidy-

lated. On the other hand, gallic acid is a tetrafunctionalmolecule bearing three phenol-type hydroxyl groups and

a carboxyl group. Then, esterification reactions would be

very probable. Thus, product 3 could be generated by the

reaction of glycidylated gallic acid with a chlorohydrine

derivative as shown inScheme 3.

Concerning product2 formation, two hypotheses could

be formulated: (i) the glycidylation of the remaining gallic

acid dimers generated by the partial hydrolysis of initial

galloylquinic acids (higher resistance of depside bonds)

and (ii) the dimerisation of free gallic acids (released by

the total hydrolysis of galloylquinic acids) through esterifi-

cation reactions followed by the glycidylation of the result-ing dimers.

Fig. 4. A part of HMBC spectrum of compound 3 in DMSO-d6showing correlations between methylene protons of the substituent groups and quaternarycarbons of aromatic rings.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

10/13

In order to verify the assumptions cited above, the gly-

cidylation reaction of tannins extract was repeated by

keeping epichlorohydrin, NaOH and PTC amounts and

replacing tara tannins by gallic acid. Thus, a large excess

of epichlorohydrin (13 M eq/OH) and higher amounts of

catalyst (0.04 M eq/OH) and NaOH (6.5 M eq/OH) were

used to perform this reaction. Surprisingly, under these

experimental conditions, gallic acid gave rise to products

1, 2 and3 in 60%, 23% and 2% yields respectively. Conse-

quently, these results support the second hypothesis inthe sense that the formation of products 1, 2 and 3 can

stem from gallic acid released by a complete hydrolysis.

Therefore, it can be deduced from these experimental

observations that, three reactions took place during the

functionalisation reaction: the hydrolysis of galloylquinic

acids, the dimerisation of the resulting free gallic acids (in

two ways) and the glycidylation of galloyl derivatives. Nev-

ertheless, the sequence by which these reactions proceed

cannot be clearly established neither the actual role played

by both phase transfer catalyst and NaOH in the process.

Despite the very interesting aspect of the mechanism

investigations, this work aimed primarily to identify the

main glycidylation products of tara tannins in order to

exploit them in the synthesis of epoxy polymers. Thus,

the prepolymer composed of the mixture of13 was for-

mulated as epoxy resin using isophorone diamine (IPD)as curing agent. The thermal resistance of the resulting

network was evaluated and its glass transition tempera-

ture (Tg) was determined by DSC. The diglycidyl ether of

bisphenol A (DGEBA) cured under the same conditions

was considered as reference.

Scheme 3. Probable mechanism of product 3 formation.

0 100 200 300 400 500 600

0

20

40

60

80

100

Weight(%)

Temperature (C)

GETT-IPD

DGEBA-IPD

Fig. 5. Thermal decomposition of GETT-IPD and DGEBA-IPD undernitrogen.

Table 2

Weight loss temperatures under N2 and DMA data of GETT-IPDand DGEBA-IPD systems.

System Td5(C) Td30(C) %Res (575 C) Tg(C)a Tb(C) Ta (C) E0 (MPa)

At 30 C AtTg+ 30 C

GETT-IPD 256 294 12 129 50 139 5.28 103 116.0

DGEBA-IPD 328 353 9 149 50 140 1.29 103 13.8

a The glass transition temperature (Tg) was determined by DSC.

-20 0 20 40 60 80 100 120 140 160 180

-0,45

-0,40

-0,35

-0,30

-0,25

-0,20

-0,15

-0,10

HeatFlow(mW/mg)

Temperature (C)

DGEBA-IPD

GETT-IPD

Tg = 149 C

Tg = 129 C

Fig. 6. DSC of GETT-IPD and DGEBA-IPD polymers.

http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

11/13

3.2. The epoxy polymers formulation

The glycidylether derivatives mixture (without silica gel

purification) obtained by reaction of tara tannins extract

with epichlorhydrin in alkaline medium is afterward

referred to as GETT. The epoxy network was built by curing

GETT with isophorone diamine (IPD) in 1:1 M ratio of

epoxy group to active H of amine [30]. To determine the

molar ratio, the epoxy equivalent weight (EEW), expressed

in mole of epoxy group per gram ofGETTwas first deter-

mined by chemical assay. The EEW value was estimatedat 8.48 103 mol/g. Based on these results, GETT-IPD

mixture was cured at 100 C for 12 h followed by 2 h at

250 C and the thermo-mechanical properties of the result-

ing network were studied. DGEBA-IPD polymer formulated

under the same conditions was used as reference.

3.3. Thermal stability of the networks

The thermal stability (under nitrogen) of the GETT-IPD

andDGEBA-IPD cured epoxy resins were investigated by

thermogravimetry analysis (TGA) from ambient tempera-

ture to 580

C. The thermograms obtained during TGAscans show the weight loss as a function of temperature

(Fig. 5). Temperatures at weight loss of 5% (Td5) and 30%

(Td30); were listed inTable 2.

The two studied epoxy resins show a continuous single

step of degradation process which starts around 256 C for

GETT-IPD and at 328 C for DGEBA-IPD (Table 2) with a

slightly weight loss around 100 C in case of the bio-based

system.

Compared to the DGEBA-IPD network, the bio-based

network, exhibits a lower thermal stability. Three

assumptions may be proposed to explain this relatively

fast degradation: (i) Unlike bisphenol A, the tara tannins

powder is a natural extract which certainly contains other

components than tannins. Indeed, the UV analysis at

280 nm allowed the identification of phenolic compounds

only (about 70% of the tara powder composition). The pres-

ence of other natural components such as carbohydrates

and lignocellulose (10%) [14a], polyalcohols, proteins,

moisture and insoluble (which combined, represent 20%

of the tara powder composition) [8,31]necessarily deter-

mines the resulting polymer behaviour. (ii) The thermal

decarboxylation of ester bonds within the prepolymer

may also be implicated [32]. (iii) The evaporation of free

quinic acid which is not involved in the glycidylation reac-

tion constitutes another possibility.

Percentages of residue under nitrogen at 575 C are

presented in Table 2. It is shown that the non-oxidative

thermal decomposition of the two studied polymers

released approximately a similar char yield. Indeed,

GETT-IPD network exhibits an increase of only 3% in char

formation upon decomposition when compared with the

DGEBA-based polymer. In both cases, the presence of aro-

matic structures increases the fire resistance of the epoxy

polymers.

3.4. Differential scanning calorimetry (DSC)

Fig. 6shows DSC thermograms of the two studied sys-

tems obtained at 10 C/min heating rate. In order to ensure

the complete crosslinking of polymers, the glass transition

temperatures were determined in the second run of DSC

analyses. Tg values are 129 C and 149 C for GETT-IPD

and DGEBA-IPD respectively (Table 2). Similarly, the

presence of non-phenolic compounds in the extract may

influence the polymer Tg. Despite the GETT-IPD lower Tg,

this polymer can keep it glassy state at high temperatures.

3.5. Thermo-mechanical properties of the networks

A plot of the storage modulus E0 and tandas a function

of temperature are reported in Figs. 7 and 8 respectively,

the values of Taand E0 are reported inTable 2.The main transition, a in the high-temperature region,

is associated with the glass transition. The highly heteroge-

neous network structure can explain the high Ta relaxa-tion spans of GETT-IPD system. The sub-transition b at

low temperature (50 C), assigned to short molecular

segment motion, hydroxyl ether groups in the particular

case of epoxy-amine is not very obvious. In contrary, a

sub-peak at the temperature from 50 to 100 C was

observed in bothE0

and tand

as well as in DSC curves whichleft-moved to the temperature range from 40 to 60 C. This

-150 -100 -50 0 50 100 150 200 250 300

-0,2

0,0

0,2

0,4

0,6

0,8

Tandelta

Temperature (C)

GETT-IPD

DGEBA-IPD

Fig. 7. Tandversus temperature of GETT-IPD and DGEBA-IPD systems.

-150 -100 -50 0 50 100 150 200 250 300

1,0

1,5

2,0

2,5

3,0

3,5

4,0

LogE'(MPa)

Temperature (C)

GETT-IPD

DGEBA-IPD

Fig. 8. Storage modulus versus temperature of GETT-IPD and DGEBA-IPD

systems.

http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

12/13

could be explained by the presence of free natural poly-

mers, which are not involved in the GETT-IPD network.

Indeed, it is important to recall that beside gallotannins,

tara powder contains polysaccharides and lignocellulose.

A recent study [33] demonstrated that some lignocellulosic

cell walls exhibit a glass transition temperature between

65 and 71 C which could be correlated to the GETT-IPD

behaviour under both DSC and DMA conditions.

The modulus E0 of GETT-IPD is higher than that of

DGEBA-IPD material. This can be explained by the mul-

tifunctionality ofGETTprepolymer inducing a more den-

sely cross-linked material compared to the DGEBA resin.

Further, the presence of aromatic rings coming from

GETT, acting as 4-functional crosslinking junctions with

non-negligible mass, greatly reduces the network mobil-

ity thus imparting a significant rigidity to the network in

the relaxed state. This constrained network structure

could also contribute to the higher glassy modulus.

However, enhanced intermolecular interactions (between

aromatics rings or H-bonding between OH groups com-

ing from the curing process and ether/carbonyl groups)

may also contribute significantly to the higher glassy

modulus.

Finally, it is important to note that several factors influ-

ence the polyphenols composition in different plants.

Although the chemical structure of the polymer building

blocks do not change, the average polymerisation degree

as well as the phenolic concentration in the powder

depends on weather conditions, the degree of maturity in

plants, extraction processes. . . [34]. Therefore, by using

another batch of tara powder, the gallic acid amount (after

hydrolysis) may change. This will certainly affect the ratio

of the glycidylated compounds 1, 2 and3. Consequently,

the number of aromatic groups and oxiran groups in the

prepolymer mixture will be modified. This will modify

the thermo-mechanical properties of the resulting resins

which strongly depend on the chemical structure of the

epoxy monomers[35].

4. Conclusions

In our on-going research program, the possibility to use

natural phenolic compounds in the synthesis of bio-based

epoxy resins is studied. Thus, the first step was to func-

tionalise phenolic models such as catechin and gallic acid

in order to evaluate their reactivity towards glycidylation.

However, these monomers do not accurately reflect the

structural complexity of natural polyphenols. Indeed, in

the plant kingdom, phenolic compounds mostly occur as

polymers with a large structural diversity. For this purpose,

we investigated the glycidylation of the tara tannins. Thus,

the reaction of tara powder with epichlorohydrin in the

presence of PTC and NaOH led to the hydrolysis of

galloylquinic acids along with the dimerisation of the

glycidylated gallic moieties through ester bonds. Then,

the glycidylated derivatives were formulated with IPD

producing a crosslinked network with good mechanical

characteristics and aTgof 129 C.

These bio-based tara tannins epoxy thermosets are

expected to be environmentally benign materials for thereplacement of petroleum-based epoxy resins.

In perspectives, a deeper characterisation of GETT-

based material, including impact strength and tensile

strength tests will be performed. In order to check the

functionalisation reproducibility, the use of different batch

of tara powder is envisioned.

Acknowledgements

The authors are grateful to Samuele Giovando

(Silvateam, Italy) for providing tara powder. Thanks also

to Pr. Antonio Pizzi (LERMAB, France) for helpful discus-

sions. We kindly acknowledge Aurlien Le Brun (UM2

Montpellier, France) and Christine Le Guernev (INRA

Montpellier, France) for their helpon NMR analyses. Thanks

to G. Gross (ULM University), Germany) and Nancy Terrier

(INRA Montpellier, France) for providingb-glucogallin.

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.eurpolymj.2014.03.034.

References

[1] (a) Hartzfeld PW, Forkner R, Hunter MD, Hagerman AE.Determination of hydrolyzable tannins (gallotannins andellagitannins) after reaction with potassium iodate. J Agric FoodChem 2002;50(7):178590;(b) Santos-Buelga C, Scalbert A. Proanthocyanidins and tannin-likecompounds nature, occurrence, dietary intake and effects onnutrition and health. J Sci Food Agric 2000;80(7):1094117;(c) Sayago-Ayerdi SG, Moreno-Hernandez CL, Montalvo-Gonzalez E,Garcia-Magana ML, Mata-Montes-de-Oca M, Torres JL, et al. MexicanAtaulfo mango (Mangifera indica L.) as a source of hydrolyzable

tannins. Analysis by MALDI-TOF/TOF MS. Food Res Int2013;51(1):18894.

[2]Schieber A, Ullrich W, Carle R. Characterization of polyphenols inmango puree concentrate by HPLC with diode array and massspectrometric detection. Innov Food Sci Emerg 2000;1(2):1616.

[3](a) Grundhofer P, Niemetz R, Schilling G, Gross GG. Biosynthesis andsubcellular distribution of hydrolyzable tannins. Phytochemistry2001;57(6):91527;(b) Tovar B, Garcia HS, Mata M. Physiology of pre-cut mango. I. ACCand ACC oxidase activity of slices subjected to osmotic dehydration.Food Res Int 2001;34(23):20715.

[4]Castaneda DM, Miguel Pombo L, Patricia Uruena C, Fredy HernandezJ, Fiorentino S. A gallotannin-rich fraction from Caesalpinia spinosa(Molina) Kuntze displays cytotoxic activity and raises sensitivity todoxorubicin in a leukemia cell line. BMC Complem Altern Med2012:12.

[5]Radebe N, Rode K, Pizzi A, Giovando S, Pasch H. MALDI-TOF-CID for

the microstructure elucidation of polymeric hydrolysable tannins. JAppl Polym Sci 2013;128(1):97107.

[6]De la Cruz Lapa P. Aprovechamiento integral y racional de la tara(Caesalpinia spinosa Caesalpinia tinctoria). Rev Inst InvestigFIGMMG 2004;7(14):6473.

[7] Chambi F, Chirinos R, Pedreschi R, Betalleluz-Pallardel I, Debaste F,Campos D. Antioxidant potential of hydrolyzed polyphenolicextracts from tara (Caesalpinia spinosa) pods. Ind Crop Prod2013;47:16875.

[8] Castell J-C, Fabregat C, Sorollal S, Jorbal M, Bacarit A, Oll L.Caesalpinia spinosa (tara): the sustainable source of tannins. J Aqeic2012;63(2):2130.

[9] (a) Haslam E, Armitage R, Bayliss GS, Gramshaw JW, Haworth RD,Jones K, et al. Gallotannins. Part III. The constitution of Chinese,turkish, sumach and tara tannins. J Chem Soc 1961:184254;(b) Haslam E, Haworth RD, Keen PC. Gallotannins. Part VII. Taragallotannin. J Chem Soc 1962:38148.

[10] Horler DF, Nursten HE. The tannins of tara.Caesalpinia spinosa(Mol.)Kuntz. J Chem Soc 1961:378692.

http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0065http://refhub.elsevier.com/S0014-3057(14)00114-1/h0065http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0070http://refhub.elsevier.com/S0014-3057(14)00114-1/h0065http://refhub.elsevier.com/S0014-3057(14)00114-1/h0065http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0060http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0055http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0050http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0045http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0040http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0035http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0030http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0025http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0020http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0015http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0010http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://refhub.elsevier.com/S0014-3057(14)00114-1/h0005http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://dx.doi.org/10.1016/j.eurpolymj.2014.03.034http://-/?-http://-/?-http://-/?-http://-/?-

7/13/2019 1-s2.0-S0014305714001141-main TARA

13/13

[11] Baratto MC, Tattini M, Galardi C, Pinelli P, Romani A, Visioli F, et al.Antioxidant activity of galloyl quinic derivatives isolated from P-lentiscus leaves. Free Radical Res 2003;37(4):40512.

[12] Nishizawa M, Yamagishi T, Dutschman GE, Parker WB, Bodner AJ,Kilkuskie RE, et al. Anti-aids agents.1. Isolation and characterizationof 4 new tetragalloylquinic acids as a new class of HIV reverse-transcriptase inhibitors from tannic-acid. J Nat Prod1989;52(4):7628.

[13] Giovando S, Pizzi A, Pasch H, Pretorius N. Structure and oligomersdistribution of commercial Tara (Caesalpina spinosa) hydrolysable

tannin. Pro Ligno 2013;9(1):2231.[14] (a) Garro JM, Riedl B, Conner AH. Analytical studies on tara tannins.

Holzforschung 1997;51(3):23543;(b) Glenz W. Renaissance of high quality vegetable-tanned leathertanned with tara tannin agent. Przegl Skorzany 1991;46(12):2779.

[15] Raquez JM, Deleglise M, Lacrampe MF, Krawczak P. Thermosetting(bio)materials derived from renewable resources: a critical review.Prog Polym Sci 2010;35(4):487509.

[16] (a) Bajpai B, Patil S. A new approach to microbial production of gallicacid. Braz J Microbiol 2008;39(4):70811;(b) Manish S, Sharma KP, Sanket K, John PJ. Purification andcharacterization of gallic acid decarboxylase from Enterobacterspp. isolated from a region in Rajasthan, India. Int J Biosci (IJB)2012;2(12):97104.

[17] Beena PS, Basheer SM, Bhat SG, Bahkali AH, Chandrasekaran M.Propyl gallate synthesis using acidophilic tannase and simultaneousproduction of tannase and gallic acid by marine Aspergillus awamori

BTMFW032. Appl Biochem Biotechnol 2011;164(5):61228.[18] Pourrat H, Regerat F, Morvan P, Pourrat A. Microbiological

production of gallic acid from Rhus-coriaria L.. Biotechnol Lett1987;9(10):7314.

[19] Kulvik E. Chestnut wood tannin extract in plywood adhesives. AdhesAge 1976;19(3):1921.

[20] Spina S, Zhou X, Segovia C, Pizzi A, Romagnoli M, Giovando S, et al.Phenolic resin wood panel adhesives based on chestnut (Castaneasativa) hydrolysable tannins. Int Wood Prod J 2013;4(2):95100.

[21] (a) Aouf C, Le Guerneve C, Caillol S, Fulcrand H. Study of the O-glycidylation of natural phenolic compounds. The relationshipbetween the phenolic structure and the reaction mechanism.Tetrahedron Lett 2012;(b) Aouf C, Lecomte J, Villeneuve P, Dubreucq E, Fulcrand H. Chemo-enzymatic functionalization of gallic and vanillic acids: synthesis ofbio-based epoxy resins prepolymers. Green Chem2012;14(8):232836;

(c) Nouailhas H, Aouf C, Le Guerneve C, Caillol S, Boutevin B,Fulcrand H. Synthesis and properties of biobased epoxy resins. Part1. Glycidylation of flavonoids by epichlorohydrin. J Polym Sci APolym Chem 2011;49(10):226170.

[22] (a) Bauer RS. Epoxy resin chemistry, advances in chemistryseries. Washington, DC: American Chemical Society; 1979;(b) May CA. Epoxy resins, chemistry and technology. New York/Basel: Marcel Dekker; 1988. p. 1.

[23] Aouf C, Nouailhas H, fache M, Caillol S, Boutevin B, Fulcrand H.Multi-functionalization of gallic acid. Synthesis of a novel bio-basedepoxy resin. Eur Polym J 2012.

[24] Lee H, Neville K. Handbook of epoxy resins. New York: Mc Graw-Hill; 1967.

[25] (a) Psomiadou E, Arvanitoyannis I, Yamamoto N. Edible films madefrom natural resources; microcrystalline cellulose (MCC),methylcellulose (MC) and corn starch and polyols. Part 2.Carbohydr Polym 1997;31:193204 [Copyright (C) 2013 AmericanChemical Society (ACS). All Rights Reserved];(b) Lopez OV, Lecot CJ, Zaritzky NE, Garcia MA. Biodegradablepackages development from starch based heat sealable films. J FoodEng 2011;105:25463 [Copyright (C) 2013 American ChemicalSociety (ACS). All Rights Reserved].

[26] Anouar EH, Gierschner J, Duroux J-L, Trouillas P. UV/Visible spectra

of natural polyphenols: a time-dependent density functional theorystudy. Food Chem 2012;131(1):7989.

[27] (a) Belmares R, Contreras-Esquivel JC, Rodriguez-Herrera R, CoronelAR, Aguilar CN. Microbial production of tannase: an enzyme withpotential use in food industry. Lebensmittel-Wiss Technol Food SciTechnol 2004;37(8):85764;(b) Bhat TK, Singh B, Sharma OP. Microbial degradation of tannins a current perspective. Biodegradation 1998;9(5):34357;(c) Sariozlu NY, Kivanc M. Isolation of gallic acid-producingmicroorganisms and their use in the production of gallic acid fromgall nuts and sumac. Afr J Biotechnol 2009;8(6):11105.

[28] Tomita H, Yonezawa K. Epoxy resin and process for preparing thesame; 1985.

[29] (a) Dolensy B, Konvalinka R, Jakubek M, Kral V. Identification ofintramolecular hydrogen bonds as the origin of malfunctioning ofmultitopic receptors. J Mol Struct 2013;1035:1248;(b) Tayyari SF, Laleh S, Zahedi-Tabrizi M, Vakili M. Structure,

vibrational assignment, and NMR spectroscopy of 1,2-bis(dichloroacetyl) cyclopentadiene. J Mol Struct 2013;1036:15160.

[30] Galy J, Sabra A, Pascault JP. Characterization of epoxy thermosettingsystems by differential scanning calorimetry. Polym Eng Sci1986;26:151423 [Copyright (C) 2012 American Chemical Society(ACS). All Rights Reserved].

[31] Luz Sanz M, Martinez-Castro I, Victoria Moreno-Arribas M.Identification of the origin of commercial enological tannins by theanalysis of monosaccharides and polyalcohols. Food Chem2008;111(3):77883.

[32] Bertero M, de la Puente G, Sedran U. Products and coke from theconversion of bio-oil acids, esters, aldehydes and ketones overequilibrium FCC catalysts. Renew Energy 2013;60:34954.

[33] Antoniow J, Maigret J, Chirtoc M, Beaugrand J, Dpole P. In: Bataille F,Flamant G, editors. Mise en vidence par micro-analyse thermique balayage (SThM) de gradients de temprature de transitionsvitreuses dans les parois lignocellulosiques. Perpignan,

France: Socit Franaise de Thermique (SFT); 2011. p. 7782.[34] (a) Cheynier V. Polyphenols in foods are more complex than often

thought. Am J Clin Nutr 2005;81(1):223S9S;(b) Morison JIL, Lawlor DW. Interactions between increasing CO2concentration and temperature on plant growth. Plant Cell Environ1999;22(6):65982.

[35] (a) Su WFA, Chen KC, Tseng SY. Effects of chemical structure changeson thermal, mechanical, and crystalline properties of rigid rod epoxyresins. J Appl Polym Sci 2000;78(2):44651;(b) Mujtaba Ellahi YG, Yang Huai. Effects of di and tetra functionalepoxy monomers structure on the morphology and the electro-optical properties of polymer-dispersed liquid crystal films. Int J SciTechnol Res 2013;2(3):12731.

http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0090http://refhub.elsevier.com/S0014-3057(14)00114-1/h0090http://refhub.elsevier.com/S0014-3057(14)00114-1/h0095http://refhub.elsevier.com/S0014-3057(14)00114-1/h0095http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0105http://refhub.elsevier.com/S0014-3057(14)00114-1/h0105http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0125http://refhub.elsevier.com/S0014-3057(14)00114-1/h0125http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0150http://refhub.elsevier.com/S0014-3057(14)00114-1/h0150http://refhub.elsevier.com/S0014-3057(14)00114-1/h0155http://refhub.elsevier.com/S0014-3057(14)00114-1/h0155http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0165http://refhub.elsevier.com/S0014-3057(14)00114-1/h0165http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0190http://refhub.elsevier.com/S0014-3057(14)00114-1/h0190http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0235http://refhub.elsevier.com/S0014-3057(14)00114-1/h0235http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0250http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0245http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0240http://refhub.elsevier.com/S0014-3057(14)00114-1/h0235http://refhub.elsevier.com/S0014-3057(14)00114-1/h0235http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0230http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0225http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0220http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0215http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0210http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0205http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0195http://refhub.elsevier.com/S0014-3057(14)00114-1/h0190http://refhub.elsevier.com/S0014-3057(14)00114-1/h0190http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0185http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0180http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0175http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0170http://refhub.elsevier.com/S0014-3057(14)00114-1/h0165http://refhub.elsevier.com/S0014-3057(14)00114-1/h0165http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0160http://refhub.elsevier.com/S0014-3057(14)00114-1/h0155http://refhub.elsevier.com/S0014-3057(14)00114-1/h0155http://refhub.elsevier.com/S0014-3057(14)00114-1/h0150http://refhub.elsevier.com/S0014-3057(14)00114-1/h0150http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0145http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0140http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0135http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0130http://refhub.elsevier.com/S0014-3057(14)00114-1/h0125http://refhub.elsevier.com/S0014-3057(14)00114-1/h0125http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0120http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0115http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0110http://refhub.elsevier.com/S0014-3057(14)00114-1/h0105http://refhub.elsevier.com/S0014-3057(14)00114-1/h0105http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0100http://refhub.elsevier.com/S0014-3057(14)00114-1/h0095http://refhub.elsevier.com/S0014-3057(14)00114-1/h0095http://refhub.elsevier.com/S0014-3057(14)00114-1/h0090http://refhub.elsevier.com/S0014-3057(14)00114-1/h0090http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0085http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0080http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075http://refhub.elsevier.com/S0014-3057(14)00114-1/h0075

1-s2.0-s0014305714001141-main tara

Documents