Polymer 46 (2005) 5005–5011

www.elsevier.com/locate/polymer

A novel method for determination of polyester end-groups by

NMR spectroscopy
A. Richard Donovan*, Graeme Moad**
CRC for Polymers, CSIRO Molecular Science, Bag 10, Clayton South, Victoria 3169, Australia

Received 7 October 2004; received in revised form 18 April 2005; accepted 19 April 2005

Abstract
An efficient, convenient and quantitative method for characterising polyester end-groups is described. We have found that trichloroacetyl
isocyanate (TAI) reacts rapidly and quantitatively with both carboxyl [C(O)OH] and hydroxyl (OH) chain ends to form derivatives that can
be readily determined by 1H NMR spectroscopy. The TAI capped end-groups give rise to characteristic imidic NH resonances in a normally
clear region of the 1H NMR spectrum [dw10–11.5 for C(O)–O–C(O)–NH–C(O)CCl3 from C(O)OH, dw8–9 for O–C(O)–NH–C(O)CCl3
from OH]. The method has been successfully applied to quantitative determination of the end-groups of a wide variety of oligomeric
polyesters. It has also been applied to higher molecular weight polyesters including commercial, bottle grade, poly(ethylene terephthalate)
(PET) and PET based copolyesters (e.g. PETG).
Crown Copyright q 2005 Published by Elsevier Ltd. All rights reserved.

Keywords: Trichloroacetyl isocyanate; Polyester end-group determination; 1H NMR spectroscopy

1. Introduction Many analytical methods have been employed for the

analysis of polyester end-groups. These methods include
A major thrust of our recent work has been the design and Fourier transform infra-red spectroscopy (FTIR) [3–7],
preparation of polyethylene terephthalate (PET) based block matrix assisted laser desorption ionization mass spec-
copolymers with enhanced gas barrier (O2, CO2) perform- trometry (MALDI) [7–10] and high performance liquid
ance via reactive extrusion [1,2]. The method used has chromatography (HPLC) [11]. Classical titration methods
involved reaction between a low molecular weight func- [12,13] can also be applied either directly, in the case of
tional polyester and poly(ethylene terephthalate). The carboxyl end-groups, or through prior derivatisation, in the
efficiency of incorporation was found to be highly case of hydroxyl end-groups.
dependent on the end-groups of the polyester and potentially NMR spectroscopy offers information on both the
of the PET. The work necessitated the synthesis of a wide number and the type of end-groups present in a sample
variety of oligoesters that differ in monomer composition, and can be applied directly when resonances attributable to
block lengths and chain ends. Consequently, a simple, quick the end-groups are sufficiently resolved from backbone
and efficient method for quantitative determination of signals [14–17]. Thus, 13C NMR has been used to directly
polyester end-groups was required. In this communication, determine the end-groups of PET [14,18]. However, the
we report that both carboxyl and hydroxyl end-groups of technique requires long acquisition times and does not have
polyesters react rapidly and quantitatively with trichloroa- general applicability.
cetyl isocyanate (TAI) to form derivatives that can be Fox et al. [14] observed that the hydroxyl end-groups of
readily determined by 1H NMR spectroscopy. PET underwent slow trifluoroacetylation in trifluoroacetic
acid/chloroform mixtures complicating end-group determi-
nation by direct 1H or 13C NMR in that solvent mixture
* Corresponding authors. Tel.: C613 95452544; fax: C613 95452446.
** Tel.: C613 95452509; fax: C613 95452446. (CF3COOD/CDCl3 is a common NMR solvent for PET and
E-mail addresses: richard.donovan@csiro.au (A.R. Donovan), similar polyesters). Kenwright et al. [19] used this knowl-
graeme.moad@csiro.au (G. Moad). edge to develop a method for formal derivatisation of PET
0032-3861/$ - see front matter Crown Copyright q 2005 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2005.04.032
5006 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011

the diastereomeric excesses of reaction products in aldol and

related reactions. The in situ TAI-derivatisation of crude
reaction mixtures in the NMR tube afforded the diaster-
eomeric trichloroacetylcarbamates with distinct imidic NH
signals in the generally vacant d 8–9 region of the 1H NMR
spectrum. Integration of these signals provided the diaster-
Scheme 1. eomeric excess directly.
Whilst use of TAI is not uncommon in the derivatisation
of small molecules it has received little attention with regard
and subsequent end-group determination by 19F NMR to polymers. However, TAI has been used for the
analysis. A disadvantage of this method is that the determination of both primary and secondary hydroxy
derivatisation process requires relatively long reaction end-groups in poly(phenyl glycidyl ether) [27] and hydroxy
times and isolation of the product prior to analysis is terminated poly(vinyl ethers) [28] and poly(ethylene glycol)
necessary. Ma et al. [20] have reported a 19F NMR-based and related polymers [29–31].
method for determining both hydroxyl and carboxyl end-
groups in PET that involved esterification of carboxyl end-
groups with hexafluoroisopropanol and trifluoroacetylation 1.2. Carboxyl end-groups
of the hydroxyl end-groups with trifluoroacetic anhydride.
Two dimensional NMR methods can also be useful for To our knowledge, TAI-derivatisation has not previously
resolving and providing more positive identification of been applied to successfully derivatise carboxylic acids. The
signals [21]. few reports for low molecular weight acids indicate that the
Recently, derivatisation with a phosphitylation reagent reaction product is unstable and is in equilibrium with the
(2-chloro-4,4,5,5-tetramethyldioxaphospholane) and 31P starting materials and that decarboxylation occurs to
NMR has been successfully employed to determine provide the trichloroacetyl amide (3) or an anhydride (4)
hydroxyl end-groups [22] and both hydroxyl and acid end- and 1,3-bis(trichloroacetyl)urea (5) as shown in Scheme 2
groups [23,24]. Although the experiment can be carried out [32]. In accord with these literature reports, we also found
in the NMR tube and no product isolation is required, the reaction with small carboxylic acids (e.g. benzoic acid,
experimental protocol remains relatively complex and long acetic acid) was slow and the reaction products unstable.
acquisition times and an internal standard are required for However, the reaction products from acid end-groups of
quantitative results. oligo- and poly-esters were sufficiently stable to allow NMR
determination. Moreover, no changes in the NMR spectra
1.1. Hydroxyl end-groups were observed over 24 h. The spectra also did not depend on
the excess of TAI indicating that no equilibrium with
TAI has been previously used to determine hydroxyl starting materials is involved or that the equilibrium lies
groups in small molecules [25]. The reaction between entirely on the side of products (2). The proposed reaction is
alcohols and TAI proceeds as expected to yield a urethane thus analogous to that observed with alcohols. The imidic
derivative (1) as shown in Scheme 1. NH signal of 2 appears in the 1H NMR spectrum in the
Roos et al. [26] employed TAI as a probe to determine region d 10–11.5.

Scheme 2.
 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011 5007

Table 1 Table 2
 n and acid numbers for a sample of Eastman 9663 using
Correlation of M Imidic proton 1
H NMR chemical shifts of various TAI derivatised
various techniques polyesters

Titrationa NMRa,b IV Polymera Solvent COOHb d OHb d

 n (g/mol) c
M – 31,000 29,500 PET TCE-d2 10.26 8.53
CCOOH 10.57G2 13.5G2 – PETGc,d CDCl3 10.36 8.65, 8.50
(mequiv./g) P(IPA-co-HER) CDCl3 11.40 8.81
COH (mequiv./g) – 50.2G5 – P(IPA-co-NPG) CDCl3 11.23 8.51
a
P(RDOA-co- CDCl3 9.91 8.55
Titration performed by Leading Synthetics Pty. Ltd., Melbourne. Errors NPG)
estimated from variance in replicate experiments. P(RDOA-co- CDCl3 10.68 8.52
b
From integration of NMR spectrum (Fig. 2). EG)
c  n)0.68
Calculated for an IV 0.82 using the relationship IVZ7.5!10K4(M P(RDOA-co- CDCl3 10.30 8.08
[6]. CHDM)c
P(RDOA-co- CDCl3 10.58 8.60, 8.42
2. Experimental section CHD)c
P(RDOA-co- CDCl3 10.78 8.63
The 400 MHz 1H and 100.6 MHz 13C NMR spectra were HER)
P(RDOA-co- CDCl3 10.87 8.44, 8.42
obtained with a Bruker Avance AV400 spectrometer.
TMCB)c
Unless indicated otherwise spectra were recorded at room
a
temperature with CDCl3 solvent and chemical shifts are Monomer abbreviations are given in the Experimental Section.
b
Precise chemical shifts show a small dependence on the concentration
reported in ppm from tetramethylsilane. The differential
of TAI.
scanning calorimeter (DSC) used was a Mettler Toledo c
CHDM is present as a mixture of cis and trans isomers.
DSC821 equipped with a TSO801RO sample robot. Gel d
PETG is a terephthalate copolyester containing two glycols (CHDM
permeation chromatography (GPC) on oligomeric polye- w40% and EGw60%).
sters was carried out with a Waters Associates Chromato-
graph equipped with 3!Mixed C and 1 mixed E PLgel a PET homopolymer with an intrinsic viscosity of 0.82.
column (each 30 mm!7.5 mm, Polymer Laboratories). PETG used was Eastar 6763; a PET/CHDM copolyester
Tetrahydrofuran was used as eluent at 1 mL/min at 22 8C. with an intrinsic viscosity of 0.70. Two samples of P(IPA-
The columns were calibrated with narrow polydispersity co-NPG) were obtained from UCB: Crylcoat 2988 (Current
polystyrene standards (Polymer Laboratories) and number designation Crylcoat 4488-0) had M  n (polystyrene equiva-
average molecular weights (M  n) are reported as polystyrene  
lents) 6352, M w 14,020, M w/Mn 2.21 (acid value from data
equivalents. Resorcinol-O,O 0 -diacetic acid. (RDOA) and sheet of 30 mg KOH/g), Crylcoat 690 (Current designation
O,O 0 -bis(2-hydroxyethyl) resorcinol (HER) were obtained Crylcoat 4890-0) had Mn (polystyrene equivalents) 7057,
from Indspec Chemical Corporation, neopentyl glycol M  w/M
 w 21,302, M  n 3.02 (hydroxyl value from data sheet of
(NPG) from Perstorff, and butylhydroxyoxostannanne 30 mg KOH/g).
from Kynar. Trichloroacetyl isocyanate, isophthalic acid
(IPA), ethylene glycol (EG), cylclohexanediol (CHD), 2.1. Oligoester synthesis
cyclohexanedimethanol (CHDM) and tetramethylcyclobu-
tanediol (TMCB) were obtained from Aldrich. The quality The low molecular weight polyesters listed in Table 2
of each reagent was established by 1H NMR and they were were prepared using the following procedure exemplified
used without purification. The PET used was Eastman 9663; for P(RDOA-co-NPG). The synthesis and characterization
of other polyesters will be reported in a forthcoming paper.
A 1 L flange flask equipped with a mechanical stirrer and
a Dean-Stark water trap was charged with RDOA (495 g,
2.2 mol), NPG (229 g, 2.2 mol) and catalyst (butylhydrox-
yoxostannane) (45 mg, 0.01 mol%). The contents were
heated (ca. 180 8C) for 4 h and water (78 mL) was collected.
The Dean-Stark apparatus was then removed, a vacuum line
attached and heating was continued for 2 h under vacuum
(40 mm Hg). An electric diaphragm pump was then
attached and heating was continued for a further 17 h
under vacuum (2 mm Hg). The reaction mixture was then
poured onto a polished metal sheet and allowed to cool. 1H
Fig. 1. Portion of 1H NMR spectrum (256 scans) of TAI derivatised
P(RDOA-co-NPG) oligomeric polyester containing both acid and hydroxyl
NMR: dH (CDCl3), 7.16 (1H, m, arom), 6.49 (3H, m, arom),
end-groups. NMR: Mn 7213, acid number 198 mequiv./g, hydroxyl number 4.60 (4H, s, OCH2CO2), 3.91 (4H, s, CO2CH2C(CH3)2),
79 mequiv./g. 0.89 (6H, s, CH3). M  n (polystyrene equivalents) 9208, Mw
5008 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011

arom), 6.49 (3H, m, backbone arom), 4.60 (4H, s, backbone

arom-OCH2CO2), 3.91 (4H, s, backbone CO2CH2C(CH3)2),
0.89 (6H, s, backbone CH3). A portion of the spectrum is
shown in Fig. 1.
Additional scans were used to provide better signal to
noise for the higher molecular weight samples or when
higher precision was required. A summary of chemical
shifts for the TAI derivatised end groups is provided in
Table 2.

2.3. TAI derivatisation procedure and 1H NMR analysis for

poly(ethylene terephthalate)

A sample of PET (15 mg) was first rendered amorphous

by rapid thermal quenching. The sample was placed in an
open aluminium sample pan and heated to 270 8C at
10 8/min in a DSC furnace then removed from the furnace
Fig. 2. Portion of 1H NMR spectrum (1024 scans) of TAI derivatised PET
(Eastman 9663) containing both acid and hydroxyl end-groups.
by the sample robot and allowed to cool to ambient
temperature. The entire operation was carried out under
27,576, M w/M n 2.99. NMR: M  n 7213, acid number nitrogen. The amorphous material was then powdered and
198 mequiv./g, hydroxyl number 79 mequiv./g. dissolved in tetrachloroethane-1,2-d2 (CDCl2)2 (0.5 mL) at
140 8C. After dissolution the mixture was transferred to an
NMR tube, an excess of trichloroacetyl isocyanate (4 mL,
2.2. TAI derivitization procedure and 1H NMR analysis for
6.3 mg, 3.36!10K2 mmol) was added and the 1H NMR
oligoesters
spectrum recorded for 1024 scans. 1H NMR: dH ((CDCl2)2,
400 MHz)), 10.26 (1H, s, end-group NHCOOCOR), 8.53
The following procedure is typical for the oligoesters
 n 2000–10,000). (1H, s, end-group NHCOOR), 8.11 (4H, s, backbone arom
(M
H), 4.68 (4H, s, backbone OCH2). A typical NMR spectrum
A sample of P(RDOA-co-NPG) (15 mg) was dissolved in
is shown in Fig. 2.
CDCl3 (0.5 mL), after dissolution the mixture was trans-
ferred to an NMR tube and an excess of trichloroacetyl
isocyanate (4 mL, 6.3 mg, 3.36!10K2 mmol) was added
3. Results
and the 1H NMR spectrum recorded for 32 scans. 1H NMR:
dH (CDCl3) 10.50 (1H, s, end-group NHCOOCOR), 8.62
The analytical procedure involves dissolving a sample of
(1H, s, end-group NHCOOR), 7.16 (1H, m, backbone
oligoester (typically 5–10 mg) in a suitable aprotic solvent
such as CDCl3 in an NMR tube, adding a drop of TAI and
recording the 1H NMR spectrum. Reaction appears to be
essentially instantaneous and is complete within the time
taken to place the sample in the spectrometer. Any excess
TAI, being aprotic, causes no additional signals in the
spectrum. While it is desirable that samples and solvent are
dry, reaction with extraneous water yields carbon dioxide
and trichloroacetamide, neither of which give signals that
interfere in the region of interest, ie. d 8–11.5. Trichlor-
oacetamide exhibits two broad singlets that appear at d 6.0
and 6.7 (see, for example, Fig. 3). It is important to store
TAI under anhydrous conditions. It has been observed that
use of aged TAI, stored under less appropriate conditions,
can give rise to unidentified by-products that provide
extraneous peaks in the spectrum.
The method has been applied to a number of in-house
synthesised oligoesters (Table 2, Scheme 3).
The spectra of oligoesters designed stoichiometrically to
Fig. 3. Comparison of the 1H NMR spectrum (32 scans) of OH end capped be wholly acid end capped (i.e., where an excess of diacid
IPA-NPG oligomer, Crylcoat 690 (lower spectrum) with that of its TAI monomer was used in the synthesis) exhibit a single imidic
derivative (upper spectrum). hydrogen resonance at approximately d 10.5. Similarly, the
 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011 5009

Scheme 3. For acid end capped XZOH, YZRDOA, hydroxyl end capped XZNPG, YZH.

spectra of oligoesters synthesised with the design of being molecular weight calculated from the end-group concen-
alcohol end capped (i.e., excess diol used in the synthesis) tration (assuming polymer only has hydroxyl and carboxyl
show a single imidic hydrogen resonance at approximately d end-groups and neglecting the presence of cyclic oligomers)
8.5. Where more nearly equimolar quantities of diacid and was 31,350 which is consistent with a value calculated from
diol monomers were used in the synthesis both peaks were the intrinsic viscosity of 0.82 (Table 1).
observed in the spectrum. An example is provided in Fig. 1. Even though TAI has been used previously for the
The methodology thus allows for both the quantitative and quantitative determination of hydroxyl moieties in small
qualitative analysis of both hydroxyl and carboxyl polyester molecules and hydroxyl terminated polymers, it remained
end-groups by 1H NMR spectroscopy. necessary to determine that the reagent reacted quantitat-
The method has also been applied to commercial ively with both acid and hydroxyl end-groups in polymer
polyesters such as PET (Fig. 2), PETG (a PET based samples.
copolyester also containing 1,4-cyclohexanedimethanol It can be seen from Fig. 3 that there are a number of
units) and Crylcoat [low molecular weight P(IPA-co- changes in the spectra which clearly show that the reagent
NPG)]. PETG and Crylcoat are amorphous and freely has reacted quantitatively with all of the OH end-groups.
soluble in CDCl3 at room temperature. PET is insoluble in The signal (a) at d 1.0 is attributable to the geminal methyls
CDCl3. To enable dissolution, the PET sample was first of the -IPA-NPG-H end-group. Upon reaction with TAI this
rendered amorphous, by rapid thermal quenching, and then signal disappears completely and a new signal appears at d
dissolved in tetrachloroethane-d2 at 140 8C. The sample was 1.1 close to that of the geminal methyls of the -IPA-NPG-
cooled to ambient temperature before addition of TAI (no IPA- repeat unit. The signal (b) at d 3.4, due to the terminal
precipitation of PET occurred). Application of the method methylene, of the -IPA-NPG-H end-group shifts to d 4.2.
to Eastman 9663, a PET homopolymer, and analysis of the The new signal (d) at d 8.5 is attributable to the imidic
1
H NMR spectra indicated the carboxyl and hydroxyl end- hydrogen of the derivatised end-group and the integral of
groups were in a 21:79 ratio. The peaks attributed to the 0.117H is consistent with a quantitative reaction when
derivatiszed carboxyl (a) and hydroxyl (b) end-groups are compared to for example signal (b) where the methylene
discrete from the terephthalate aromatic signal at d 8.1. Due CH2 has an integral of 0.223H. These values are within
to the low concentration of end-groups in PET samples 1024 experimental error of G5%.
scans were required for adequate signal to noise. Integration Derivatisation of acid end capped IPA-NPG (Crylcoat
of the end group signals vs. the signals attributable to the 2988) with TAI yields a singlet in the NMR spectrum at wd
backbone aromatic hydrogens enabled calculation of the 11.65 (b). No other diagnostic changes are observed in the
concentration of the hydroxyl (COH) and carboxyl end- spectrum upon addition of the reagent (Fig. 4).
groups (CCOOH). The value of CCOOH was in accord with a A portion of the 1H NMR spectrum of an acid end capped
value determined by titration. The number average IPA-NPG oligomer (Crylcoat 2988) recorded in CF3-
COOD/CDCl3 is shown in Fig. 5. In this solvent system
the end-groups attributable to the H(a) of the NPG-IPA-H

Fig. 4. Portion of 1H NMR spectrum (32 scans) of TAI derivatised acid end Fig. 5. Portion of 1H NMR spectrum (32 scans) of acid end capped IPA-
capped IPA-NPG (Crylcoat 2988). NPG (Crylcoat 2988) recorded in CF3COOD/CDCl3.
5010 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011

Table 3
Comparison between various NMR techniques for end-group quantification

NMR technique Peak ratio End-group ratio Acid No mequiv./g

1 a
H NMR TAI 1:8.4G0.3 1:7.4 577
13
C NMR Cr(acac)3b 1:16.4G1.6 1:7.7 555
1
H NMR directc 1:7.6G0.3 1:7.6 562
a
See Fig. 4.
b 13
C NMR with Cr(acac)3 relaxtion agent (3%). Carboxy end-group carbonyl appears at d 169.3 while isophthalate repeat unit carbonyl appears at d 165.1 in
13
C NMR (CDCl3).
c 1
H NMR (CF3COOD/CDCl3 1:1), see Fig. 5.

end-group appear distinct from the H(b) of NPG-IPA-NPG Acknowledgements

repeat unit and can be integrated directly [2]. The result
correlates well with data obtained using the present method The authors would like to thank Kelvin Davies (VisyPak
for the same sample derivatised with TAI (Fig. 4 and Australia) for helpful discussion and Roger Mulder and Jo
Table 3). Cosgriff (CSIRO Molecular Science) for their assistance
Quantitative 13C NMR determination of the isophthalate with NMR spectroscopy.
carbonyl resonances for a sample containing added
Cr(acac)3 (to reduce relaxation times [33]) also provides a
consistent value for the end-group ratio within experimental References
error (Fig. 6 and Table 3). It can be seen from this spectrum
(Fig. 6) that even recording for 10,240 scans the signal to [1] Moad G, Groth AM, O’Shea MS, Tozer RD, WO0222705; 2002.
noise is quite poor rendering this technique less practical [2] Moad G, Groth A, O’Shea MS, Rosalie J, Tozer RD, Peeters G.
than 1H NMR based methods. Macromol Symp 2003;202:37–45.
[3] Koenig JL. Spectroscopy of polymers. New York: Elsevier; 1999.
[4] Siesler HW. J Mol Struct 1980;59:15.
[5] Hummel D, Scholl F. Atlas of polymer and plastics analysis. vol. 2.
4. Conclusions München: Carl Hanser; 1988 [part B/I].
[6] Al-AbdulRazzak S, Lofgren EA, Jabarin SA. Polym Int 2002;51:
Trichloroacetyl isocyanate reacts rapidly and quantitat- 174–82.
[7] Ward IM. Trans Faraday Soc 1957;53:1406.
ively with both acid and hydroxyl chain ends to form
[8] Nielen MWF. Mass Spectrom Rev 1999;18:309–44.
derivatives that can be readily determined by 1H NMR [9] Rader HJ, Schrepp W. Acta Polym 1998;49:272–93.
spectroscopy thus providing a convenient means of character- [10] Guittard J, Tessier M, Blais JC, Bolbach G, Rozes L, Marechal E,
ising polyester end-groups. We anticipate the method will et al. J Mass Spectrom 1996;31:1409–21.
have more general application in polymer end-group [11] Rissler K. J Chromatogr A 1997;786:85–98.
[12] Ubranski J, Czerwinski W, Janicka K, Majewska F, Zowall H. In:
determination. Recently, we have applied the method for
Cameron GG, editor. Handbook of analysis of synthetic polymers and
determination of end-groups (amino, hydroxyl) of polymers plastics. 1st ed. London: Halsted Press; 1977. p. 223–72.
formed by living radical polymerization [34–36] and [13] Price GF. In: Allen PW, editor. Techniques of polymer characteris-
polyethylene–poly(ethylene glycol) block copolymers [37]. ation. London: Butterworths Scientific; 1959. p. 209–29.
[14] Fox B, Moad G, Van Diepen G, Willing RI, Cook WD. Polymer 1997;
38:3035–43.
[15] Moad G. In: Webb GA, editor. Annual reports in NMR spectroscopy,
vol. 29. London: Academic Press; 1994. p. 287–323.
[16] Bovey FA, Mirau P. NMR of polymers. New York: Academic Press;
1996 p. 199–212.
[17] Johns SR, Rizzardo E, Solomon DH, Willing RI. Makromol Chem,
Rapid Commun 1983;4:29.
[18] Petiaud R, Waton H, Pham QT. Polymer 1992;33:3155–61.
[19] Kenwright AM, Peace SK, Richards RW, Bunn A, MacDonald WA.
Polymer 1999;40:2035–40.
[20] Ma Y, Agarwal US, Vekemans J, Sikkema DJ. Polymer 2003;44:
4429–34.
[21] Spyros A. J Appl Polym Sci 2003;88:1881–8.
[22] Chan KP, Argyropoulos DS, White DM, Yeager GW, Hay AS.
Macromolecules 1994;27:6371–5.
[23] Spyros A, Argyropoulos DS, Marchessault RH. Macromolecules
1997;30:327–9.
[24] Spyros A. J Appl Polym Sci 2002;83:1635–42.
[25] Goodlett VW. Anal Chem 1965;37:431–2.
Fig. 6. Portion of 13C NMR spectrum (10240 scans) of acid end capped [26] Roos GHP, Manickum T, Field JS, Ramesar N. S Afr J Chem 1994;47:
IPA-NPG (Crylcoat 2988) recorded in CDCl3 with 3% Cr(acac)3. 33–5.
 A.R. Donovan, G. Moad / Polymer 46 (2005) 5005–5011 5011

[27] Ronda JC, Serra A, Mantecon A, Cadiz V. Macromol Chem Phys [34] Postma A, Donovan R, Davis TP, Moad G, O’Shea M, submitted for
1994;195:3445–57. publication.
[28] Van Meirvenne D, Haucourt N, Goethals EJ. Polym Bull 1990;23: [35] Postma A, Donovan R, Davis TP, Moad G, O’Shea M. World
185–90. Polymer Congress, Macro 2004 Congress Proceedings. Paris:
[29] Stefanova R, Rankoff D, Panayotova S, Spassov SL. J Am Oil Chem IUPAC; 2004. p. 2.1–122 e-polymers http://www.e-polymers.org/
Soc 1988;65:1516–8. paris/data/L2925.pdf.
[30] Loccufier J, Vanbos M, Schacht E. Polym Bull 1991;27:201–4. [36] Moad G, Rizzardo E, Postma A, Thang SH. Polymer, in press.
[31] Devos R, Goethals EJ. Polym Bull 1986;15:547–9. [37] Moad G, Dean K, Edmond L, Kukaleva N, Li G, Mayadunne RTA,
[32] Motoki S, Saito T, Kagami H. Bull Chem Soc Jpn 1971;47:775. Pfaendner R, Schneider A, Simon G, Wermter H. Macromolecules,
[33] Braun S, Kalinowski H-O, Berger S. 150 and More basic NMR submitted for publication.
experiments: a practical course. Weinheim: Wiley-VCH; 1998.
p. 282–284.