Click here to load reader
Upload
5landers
View
7
Download
1
Embed Size (px)
DESCRIPTION
uv aging of plastics
Citation preview
POLYMERS FOR ADVANCED TECHNOLOGIES
Polym. Adv. Technol. 2005; 16: 61–66
Published online 16 November 2004 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/pat.538
A comparative study of UV aging characteristics of
poly(ethylene-co-vinyl acetate) and poly(ethylene-co-vinyl
acetate)/carbon black mixture
Mehmet Copuroglu and Murat Sen*Polymer Chemistry Division, Department of Chemistry, Hacettepe University, 06532, Beytepe, Ankara, Turkey
Received 17 May 2004; Revised 9 July 2004; Accepted 5 September 2004
In this comparative study, the effect of carbon black (CB) on the UV aging characteristics of
poly(ethylene-co-vinyl acetate) (EVA) was investigated. EVA, containing 13% vinyl acetate (VA),
and poly(ethylene-co-vinyl acetate)/carbon black mixture (EVA/CB), containing 13% VA and 1%
CB, were aged by means of UV light with a wavelength in the vicinity of 259nm, in air, up to
400hr. Sol-gel analyses were made to determine the percentage gelation of both virgin and aged
samples. FT-IR measurements were performed to follow the chemical changes which took place
in the samples during aging. Dynamic and isothermal thermogravimetry studies were performed
for determination of the thermal stabilities of virgin and aged samples.
Sol-gel analysis results showed that EVA itself has a tendency to form a gel under UV irradia-
tion. EVA/CB, however, becomes a gel to a smaller extent, comparatively, under the same condi-
tions. As a result of FT-IR measurements, some oxidation products such as ketone, lactone and
vinyl species were observed through UV ageing of EVA and EVA/CB. Thermal analysis experi-
ments exhibited that the thermal stabilities of EVA and EVA/CB decreased, to a similar extent
through UV aging. Copyright # 2004 John Wiley & Sons, Ltd.
KEYWORDS: poly(ethylene-co-vinyl acetate); carbon black; aging; degradation; FT-IR
INTRODUCTION
Restricted properties and limited use of homopolymers
alone, has given rise to exploration of composites, copoly-
mers, blends, etc. Copolymers such as poly(ethylene-co-vinyl
acetate) (EVA), ethylene-butyl acrylate, ethylene-ethyl acry-
late have wide range of usages in different industries. Among
the numerous ethylene copolymers, due to its wide range of
properties depending on its vinyl acetate (VA) content, EVA
has become one of the most useful copolymers in the trans-
portation industry as an insulator, in the electric industry as
a cable insulator, in the shoe industry as soles, and in many
other industries as an hot melt adhesive, a coating, etc. Sev-
eral pieces of work have looked at EVA itself and also its
degradation characteristics.
One of the earliest studies on UV irradiation of polymers,
was performed by Geuskens et al.1 They investigated and
compared the influences of UV light and gamma irradiation
on poly(vinyl acetate) (PVA). They estimated the products of
photolysis and radiolysis of PVA. They also proposed some
mechanisms for chemical changes undergoing during photo-
lysis and radiolysis of PVA.
Lacoste and Carlsson2 undertook a study of gamma-,
photo-, and thermally-initiated oxidation of linear low
density polyethylene (LDPE), and compared the detailed
oxidation products for different initialization types. They
observed common oxidation products, such as hydroper-
oxide, ketone, vinyl and ester groups, for all types of
initialization but with quantitative differences.
Artificial ageing studies on high density polyethylene by
UV irradiation in the presence of air was performed by
Carrasco et al.3 They observed structural modifications and
chemical changes during UV irradiation; and these were
explained via chain breaking, branching, crosslinking and
oxidation phenomena.
Recently, in order to investigate the synergism between
carbon black (CB) and different stabilisers, Pena et al.4 studied
the photodegradation of LDPE. They concluded some results
concerning synergetic and antagonistic effects, in which CB
played an important role for all combinations.
Ageing characteristics of EVA and poly(ethylene-co-vinyl
acetate)/carbon black mixture (EVA/CB) are very important
because of their usages in the fields where the products are
exposed to sunlight and/or heat. In some countries, like in
Turkey, climate has a great effect on the aging of EVA. EVA/
CB with 13% VA and 1% CB is a widely used polymer in
particular by Turkish State Railways (TCDD) due to its elastic
structure and insulation property. In our previous study,5 we
investigated the accelerated thermal aging characteristics of
Copyright # 2004 John Wiley & Sons, Ltd.
*Correspondence to: M. Sen, Polymer Chemistry Division,Department of Chemistry, Hacettepe University, 06532,Beytepe, Ankara, Turkey.E-mail: [email protected]
EVA and EVA/CB; and we concluded that CB is a very
effective stabilizer against thermal degradation of EVA at
moderate temperatures. In this comparative study; acceler-
ated UV aging characteristics of EVA (13% VA) and EVA/CB
(13% VA and 1% CB) were investigated.
EXPERIMENTAL
MaterialsEVA, containing 13% VA and of density 0.9288 kg l�1, was
supplied by Elf-Atochem Co. in the form of granules. EVA/
CB plates with contents of 13% VA and 1% CB were obtained
from Panel Co., Inc., Turkey. EVA, used in the preparation of
this EVA/CB mixture, obtained from Elf-Atochem Co.;
whereas masterbatch (PE Black 99209) from Viba Co., Italy,
in the form of 50% dispersion of CB, type SRF, in LDPE.
Xylene, used as a solvent for EVA and EVA/CB and for the
determination of gelation, was obtained from Merck.
Aging of materialsEVA granules, and EVA/CB specimens of dimensions
3.4 mm� 3.9 mm with 1.9 mm thickness were irradiated
with UV light with an intensity of 12,000 lm at a distance of
15 cm, in air, at ambient temperature; and were aged at dif-
ferent intervals of time up to 400 hr.
Analyses
Determination of percentage gelationsInorder to investigate the influence of UV light on the gela-
tions of EVA and EVA/CB, sol-gel analyses were performed.
Xylene was used as a solvent in the soxhlet extractor and
was fluxed through each sample for 14 hr. Gel percentages
were calculated gravimetrically according to the following
equation:
% Gel ¼ m=m0 � 100
where m0 and m are the masses of a sample before and after
extraction, respectively.
FT-IR studiesFT-IR studies were carried out by means of Nicolet 520 model
spectrometer. Samples were pressed in a hotplate at about
1008C for 10 sec in order to obtain film forms. Samples, cross-
linked to a high extent, were scraped to obtain tiny particles
which were then mixed with KBr.
Thermogravimetric analysis (TGA)Thermogravimetric analyses were performed by utilizing Du
Pont Instruments-Thermal Analyzer, Model 951. Dynamic
thermogravimetric studies were carried out under nitrogen
atmosphere; and 108C/min heating rate was used. Dynamic
thermogravimetric study results indicated the thermal
stabilities of virgin and aged EVA and EVA/CB. Two types
of isothermal thermogravimetric studies were performed.
One was in nitrogen atmosphere at 3508C, and was used to
determine the percentages of volatile degradation products
of virgin and aged samples. The other was under oxygen
atmosphere at 1958C; and was used to determine the
thermo-oxidative stabilities of virgin and aged EVA and
EVA/CB.
RESULTS AND DISCUSSION
Sol-gel analysesBefore investigation of chemical changes in the structure of
EVA and EVA/CB, the effect of UV aging on the gelation of
EVA and EVA/CB was investigated. The effect of UV aging
on the percentage gelations of EVA and EVA/CB are given in
Fig. 1. As shown in Fig. 1 percentage gelation increased up to
88% for EVA with UV aging time, whereas up to merely 11%
for EVA/CB.
Gelation in polymers is generally referred to as cross-
linking of macromolecules by means of covalent bonds.
Sharp increase in the percentage gelation indicates that EVA
can crosslink easily by UV irradiation. Crosslinking most
likely occurs due to combinations of the macromolecular
radicals formed during UV irradiation, as explained in the
literature for polyolefins.6–8 Since the main chains of EVA
(with a content of 13% VA) macromolecules resemble those of
polyethylene (PE), it shows similar behavior to that of PE. An
explanation proposed by Geuskens et al.1 for PVA, might also
be valid for EVA. This mechanism is shown in Scheme 1.
The extent of this mechanism is small when compared to
that of PVA, of course, due to the rare occurrence of acetate
groups on the main chains of EVA macromolecules.
In order to determine the crosslink and the chain scission
reaction yields occurring during UV aging of both EVA
and EVA/CB, the usual Charlesby–Pinner equation was
modified. Charlesby–Pinner equation,9 (sþHs¼ po/qoþ 2/
(qou2,0 D)), has been used by many researchers so far for
simultaneous determination of the crosslinking and the chain
scission reaction yields of polymers being irradiated by
ionizing radiation. In this equation, s is the sol fraction, po is
the chain scission yield (average number of main chain
scissions per monomer unit and per unit dose), qo is
crosslinking yield (proportion of monomer units crosslinked
per unit dose), u2,0 is initial weight (average degree of
polymerization), and D is irradiation dose. Although this
equation was derived from by Charlesby–Pinner for deter-
mination of crosslinking and chain scission yields for
irradiated polymers such as gamma rays or electron beam,
it is assumed that this general equation can also be used for
the simultaneously and UV crosslinked and degraded
polymers by simple replacement of dose with aging time.
Figure 1. Variation of percentage gelation with UV irradia-
tion time.
Copyright # 2004 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2005; 16: 61–66
62 M. Copuroglu and M. Sen
Replacing the UV aging time instead of irradiation dose of
ionizing radiation in the Charlesby–Pinner equation, it
becomes as follows:
sþps ¼ po=qo þ 2ðqou2;0tÞ
where t is aging time.
In this equation, u2,0 was calculated as the ratio of weight
average molecular weight (obtained from Elf-Atochem as
95,000 g mol�1) to average weight of a monomer unit
(calculated as 30.8), and was found to be 3084. Plots of the
reciprocal of aging time, 1/t, versus sþHs; yield straight
lines (with a regression coefficient, r¼ 0.972 for EVA and
0.999 for EVA/CB). and po/qo and qo values were calculated
from the intercept and the slope of the lines, respectively.
As a result of the calculations, po and qo (per hour) were
found to be 2.84� 10�3 and 8.20� 10�3, respectively, for
EVA; and 5.73� 10�2 and 3.21� 10�2, respectively, for EVA/
CB. Although the values for EVA/CB differ from those of
EVA, when they are evaluated individually, one can
conclude that EVA has a tendency to form crosslinks rather
than chain scission, since po/qo¼ 0.346. However, with a
value po/qo¼ 1.79, EVA/CB seems to be a mixture which
cannot crosslink readily upon UV irradiation. These results
are parallel with those obtained for thermally aged EVA and
EVA/CB,5 since the materials showed similar tendencies.
FT-IR studies and UV degradation mechanismsof EVA and EVA/CBEVA and EVA/CB have almost the same spectra, because CB
itself has no remarkable characteristic absorption bands and
thus, does not influence the spectrum of EVA. The FT-IR spec-
tra of virgin and UV aged EVA and those of EVA/CB are
given in Figs. 2 and 3 respectively, in the range 1950–
1550 cm�1. Main functional groups changes were observed
in this region. In order to calculate the absorbance value of
a band which overlapped with another band or bands in
the FT-IR spectra, a band separation software program was
used. Absorption at 723 cm�1 (A723) is assigned to –CH2–
CH2–CH2– sequences. It was supposed that, during UV
degradation, the amount of these methylene sequences
remains almost constant with respect to the other bands at
which significant changes occurred. Therefore, regarding
the absorption at this wavenumber as constant, changes at
any other band can be monitored.
The main UV degradation product of EVA is acetic acid
which forms due to ester elimination. Mechanism of ester
elimination is likely to be that given in our previous paper;5 so
it is anticipated that the absorption at 1740 cm�1 (A1740)
should decrease with UV ageing. The formation and the
Scheme 1.
Figure 2. FT-IR spectra in the range 1950–1550 cm�1 of
virgin and UV aged EVA. Numbers on the curves represent
UV aging time (in hours).
Figure 3. FT-IR spectra in the range 1950–1550 cm�1 of
virgin and UV aged EVA/CB. Numbers on the curves
represent UV aging time (in hours).
UV aging characteristics 63
Copyright # 2004 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2005; 16: 61–66
destruction of aldehydes (Schemes 2 and 3 in ref. 5) should be
taken into account. In the case of EVA, the decrease in the
absorption at 1740 cm�1 was observed merely up to around
50 hr in EVA (Fig. 4a). From this point on, a continuous
development of aldehydes is predominant up to around
150 hr. The ratio is almost constant between 150–400 hr. In the
case of EVA/CB (Fig. 4a), the formation of aldehydes is
predominant in the early stages of UV aging. The formation
rate of aldehydes after around 100 hr is very slow when
compared to that of EVA, which results in a decrease in the
absorbance at 1740 cm�1. It can be concluded that the
presence of CB can prevent the formation of aldehydes to
some extent. The amount of aldehydes in the virgin EVA/CB
is twice that of EVA. This might be due to some oxygen-
containing groups which could be formed in EVA/CB during
its manufacturing stages.
Ketone formation was observed in UV aged EVA and
EVA/CB. Mechanisms of ketone formation via degradation
(either non-oxidative or oxidative) were proposed in the
literature,6,10,11 Variation of ketone groups was followed by
the ratio of absorbance at 1715 cm�1 (A1715) to absorbance at
723 cm�1. As can be seen from Fig. 4(b), ketone formation in
EVA increased continuously until the end of aging. In EVA/
CB, the profile of the change is very similar to that of EVA.
Another species formed as a consequence of UV aging is
lactone. The mechanism given in Scheme 6 in ref. 5 was
proposed for lactone formation associated with a back-biting
process in the VA moieties by the acetate groups forming
methane. Lactone formation in UV aged EVA and EVA/CB
was followed by the ratio of absorbance at 1780 cm�1 (A1780)
to absorbance at 723 cm�1. The change of A1780/A723 with UV
ageing time is given in Fig. 4(c). Lactone formation increased
rapidly in the initial stage of the aging (up to around 100 hr).
This formation is almost constant between 100–250 hr. After
that step, an increase again is observed up to 400 hr.
However, the amount of formation of lactone is less than
that of ketone in every step of aging, since ketone formation is
easier than lactone formation due to the conformational
arrangements. The presence of CB prevents this formation to
a small extent; most probably by hindering the intermole-
cular back-biting processes of macromolecular segments
required for the formation of lactones.
Samples, which were subjected to UV aging (except CB
containing ones), exhibited yellowing behavior during this
procedure. It is well known that yellowing (or even darker
colors from brown to black) is a consequence of formation of
unsaturated groups.6,8 These unsaturated groups are gen-
erally vinyl, vinylidene, and trans-vinylene groups which are
responsible for coloring. Mechanisms of formations of trans-
vinylene, vinyl, and vinylidene are given in Schemes 7 and 8
in ref. 5 and also in ref. 12. The results of normalization for the
band at 910 cm�1 (A910) for which the vinyl groups formed in
UV aged EVA and EVA/CB are shown in Fig. 4(d). Vinyl
content in EVA increased continuously up to around 100 hr.
After that time, a slight decrease was observed until 400 hr
which might be due to the increase in the crosslinking ratio. In
the case of EVA/CB, however, this formation is hindered by
CB due to its radical delocalizing property. The formation of
Figure 4. Some chemical changes observed by means of FT-IR spectroscopy during UV
aging. (a) Changes occurred at 1740 cm�1 during UV aging of EVA and EVA/CB. (b) Evolution
of keto groups at 1715 cm�1 during UV aging in EVA and EVA/CB. (c) Evolution of lactone at
1780 cm�1 during UV aging in EVA and EVA/CB. (d) Vinyl formation at 910 cm�1 during UV
aging in EVA and EVA/CB.
64 M. Copuroglu and M. Sen
Copyright # 2004 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2005; 16: 61–66
vinylidene and trans-vinylene were not clearly observed due
to high noise to signal ratio. This resulted from the difficulties
in the preparation of thin films from crosslinked granules.
Influence of UV aging on the thermal stabilitiesof EVA and EVA/CBFigures 5 and 6 show the dynamic TGA thermograms of vir-
gin and UV aged EVA and EVA/CB, in nitrogen atmosphere,
respectively. As shown in Figs. 5 and 6 virgin EVA and EVA/
CB show the typical step degradation profile with the initial
stage involving acetic acid evolution and the second invol-
ving main chain degradation.10,13 Figures 5 and 6 indicate
that, when aging time increases, degradation profile changes.
Degradation at both stages becomes easier. Both first and sec-
ond degradation steps are shifted to lower temperatures in
the UV aged EVA and EVA/CB. As a consequence, CB failed
to prevent EVA from UV degradation contrary to that in the
case of thermal degradation.5
Degradation activation energies (E) of EVA and EVA/CB
were calculated by using Freeman–Carroll equation and
the data obtained from derivatives of dynamic thermograms
of UV aged EVA and EVA/CB were used in the calcula-
tions.14
� logðd%=dTÞ� log ð1 � cÞ ¼ n� E
2:3R� �ð1=TÞ� log ð1 � cÞ
whereT is the temperature in Kelvin,n is the order of reaction,
R is the gas constant which has a value of 8.314 J mol�1 K�1, c
is the conversion ratio and is equal to (m0�m)/m0, where m0
is the initial mass and m is the mass at any time.
When D(1/T)/Dlog(1� c) versus Dlog (d%/dT)/Dlog
(1� c) is plotted, the slope of this plot gives the value for E
and the intersection of ordinate indicates the order of
reaction. The c and T values were obtained from derivatives
of thermograms. The effect of UV aging time on the E values
of EVA and EVA/CB are shown in Fig. 7. TheEvalues of EVA
and EVA/CB decreased continuously with increasing UV
aging time. Oxidation and chain scission are the main reasons
for this decrease. The chain scission may occur by a seven-
membered ring transition state without acetal group
elimination. Main chain scission may also occur by the
fragmentation of the polymer alkoxy radical or as the
disproportionation of the polymer alkyl radical.8 Decreases
in the thermal stability and the E value of an oxidized
polymer and the effect of molecular weight on these
parameters have been well described in the literature by
many researchers.15–17
As can be seen from Fig. 5 due to overlapping of first and
second degradation steps, it is not easy to interpret and
determine the amounts of readily decomposing groups from
dynamic TGA thermograms. For a detailed analysis of the
first stage of degradation, isothermal TGA analyses were
performed in nitrogen atmosphere. Isothermal-nitrogen
thermograms of UV aged EVA and EVA/CB were obtained
at 3508C. This temperature is sufficient to break down all
kinds of oxygen-containing groups. Figure 8 shows the
percentages of volatile products obtained from these thermo-
grams at the end of 70 min. Fig. 8 also indicates that the
amount of volatile products increases with UV aging time.
EVA/CB has low resistance to thermo-oxidation; and the
change in EVA/CB during UV irradiation is similar to that in
EVA, contrary to thermally aged EVA/CB.5 These results are
consistent with the results of FT-IR and dynamic thermo-
gravimetry studies.
Isothermal thermogravimetry studies which were carried
out in oxygen atmosphere give important information
concerning the degradation characteristics of materials.
Isothermal thermogravimetry studies in oxygen atmosphere,
were performed at 1958C. Weight increase in the initial stage
Figure 5. Dynamic thermograms of UV aged EVA in
nitrogen atmosphere. Numbers on the curves indicate the
UV aging time (in hours).
Figure 6. Dynamic thermograms of UV aged EVA/CB in
nitrogen atmosphere. Numbers on the curves indicate the UV
aging time (in hours).
Figure 7. Variation of degradation activation energies with
UV aging time for EVA and EVA/CB.
UV aging characteristics 65
Copyright # 2004 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2005; 16: 61–66
of degradation indicates the oxygen penetration into the
polymer structure.18,19 EVA has an oxygen uptake capacity
which diminishes through UV aging time and then falls
completely to zero. This can be explained by the rapid
formation of decomposable oxygen-containing species such
as aldehydes or ketones through UV ageing. When this
formation reaches an equilibrium, uptake of oxygen stops.
Oxygen uptake capacity falls to zero in EVA/CB quicker than
that of EVA. Oxidation, undergoing during UV aging, also
makes EVA/CB relatively more easily degradable with
increasing UV irradiation time. Lower degree of crosslinking
in EVA/CB during UV aging is another reason for much
more facile decomposition. Moreover, time to reach the 10%
mass loss increased sharply after around 116 hr in EVA
(Fig. 9). This increase might be due to the higher crosslinking
of EVA macromolecules and the formation of relatively more
sustainable oxygen-containing groups such as carboxylic
acids. In EVA/CB, time period for 5% mass loss decreases
suddenly up to 24 hr, then remains unchanged (Fig. 9).
CONCLUSIONS
In order to investigate the UV aging characteristics of EVA
and EVA/CB, samples were subjected to UV irradiation at
ambient temperature and in air up to 400 hr. Sol-gel analyses
showed that EVA has a tendency to crosslink under UV irra-
diation.
FT-IR spectra of virgin and UV aged EVA and EVA/CB
showed that, some chemical changes took place in EVA and
EVA/CB during UV ageing (such as the formation of ketone,
lactone and vinyl species).
Dynamic thermograms, which were carried out in nitrogen
atmosphere, indicated that the thermal stabilities of EVA and
EVA/CB decreased during UV aging. Isothermal TGA
analyses, performed at 3508C and in nitrogen atmosphere,
indicated that the concentration of volatile products
increased during UV aging both in EVA and EVA/CB. As a
result of isothermal thermogravimetry studies, which were
carried out at 1958C and in oxygen atmosphere, the oxygen
uptake capacities of EVA and EVA/CB decreased through
UV aging time and fell to zero eventually. Time to reach the
10% mass loss in EVA increased sharply after around 116 hr,
whereas time passed for 5% mass loss in EVA/CB decreased
suddenly up to around 24 hr with UV irradiation time; then it
remained almost constant.
As a conclusion, all these studies show that, 1% CB is not a
very effective stabilizer against UV degradation of EVA at
moderate temperatures.
AcknowledgmentsThe authors would like to thank Prof. Dr Olgun Guven for the
laboratory facilities he supplied and for his encouraging and
constructive advice during this study.
REFERENCES
1. Geuskens G, Borsu M, David C. Eur. Polym. J. 1972; 8:883.
2. Lacoste J, Carlsson DJ. J. Polym. Sci. Part A: Polym. Chem.1992; 30: 493.
3. Carrasco F, Pages P, Pascual S, Colom X. Eur. Polym. J. 2001;37: 1457.
4. Pena JM, Allen NS, Edge M, Liauw CM, Valange B. Polym.Degrad. Stab. 2001; 72: 259.
5. Copuroglu M, Sen M. Polym. Adv. Technol. 2004; 15: 393.6. White JR, Turnbull A. J. Mater. Sci. 1994; 29: 584.7. Grassie N, Scott G. Polymer Degradation and Stabilisation.
Cambridge University Press: Cambridge, 1985.8. Rabek JF. Polymer Photodegradation (1st edn). Chapman &
Hall: Cambridge, 1995.9. Charlesby A. Atomic Radiation and Polymers. WNT: Warsaw,
1962.10. Allen NS, Edge M, Rodriguez M, Liauw CM, Fontan E.
Polym. Degrad. Stab. 2001; 71: 1.11. Allen NS, Chirinis-Padron A, Henman TJ. Polym. Degrad.
Stab. 1985; 13: 31.12. Ranby B, Rabek JF. Photodegradation, Photo-oxidation
and Photostabilization of Polymers. John Wiley & Sons:Chichester, 1975.
13. McGrattan BJ. Appl. Spectrosc. 1994; 48(12): 1472.14. Freeman ES, Carroll B. J. Phys. Chem. 1958; 62: 394.15. Schnabel W. Polymer Degradation. Macmillan: New York,
1981.16. Holland BJ, Hay JN. Polymer 2001; 42: 4825.17. Sen M, Basfar AA. Radiat. Phys. Chem. 1998; 52: 247.18. Uzun C. M. Sc. Thesis, 1992. Hacettepe University, Ankara,
Turkey.19. Gal O, Novakovic LJ, Markovic V, Stannett VT. Radiat. Phys.
Chem. 1983; 22: 627.
Figure 8. Percentage of volatile products against UV
irradiation time.
Figure 9. Time required to reach the 10 and 5% mass
losses for UV aged EVA and EVA/CB, respectively.
66 M. Copuroglu and M. Sen
Copyright # 2004 John Wiley & Sons, Ltd. Polym. Adv. Technol. 2005; 16: 61–66