polymers

Article
Study of Non-Isothermal Crystallization of
Polydioxanone and Analysis of Morphological
Changes Occurring during Heating and
Cooling Processes
Yolanda Márquez 1,2 , Lourdes Franco 1 , Pau Turon 2 , Juan Carlos Martínez 3 and Jordi Puiggalí 1, *
1 Chemical Engineering Department, Escuela d’Enginyeria Barcelona Est (EEBE), c/ Eduard Maristany 19,
Barcelona 08930, Spain; ymlyolanda@gmail.com (Y.M.); lourdes.franco@upc.edu (L.F.)
2 B. Braun Surgical S.A., Carretera de Terrassa 121, Barcelona 08191, Spain; pau.turon@bbraun.com
3 ALBA Synchrotron Light Facility, Ctra BP 1413 km 3.3, Cerdanyola del Vallès, Barcelona 08290, Spain;
guilmar@cells.es
* Correspondence: Jordi.Puiggali@upc.edu; Tel.: +34-934-016684

Academic Editor: Naozumi Teramoto

Received: 11 August 2016; Accepted: 19 September 2016; Published: 28 September 2016

Abstract: Non-isothermal crystallization kinetics of polydioxanone (PDO), a polymer with

well-established applications as bioabsorbable monofilar suture, was investigated by Avrami, Mo, and
isoconversional methodologies. Results showed Avrami exponents appearing in a relatively narrow
range (i.e., between 3.76 and 2.77), which suggested a three-dimensional spherulitic growth and
instantaneous nucleation at high cooling rates. The nucleation mechanism changed to sporadic at low
rates, with both crystallization processes being detected in the differential scanning calorimetry (DSC)
cooling traces. Formation of crystals was hindered as the material crystallized because of a decrease in
the motion of molecular chains. Two secondary nucleation constants were derived from calorimetric
data by applying the methodology proposed by Vyazovkin and Sbirrazzuoli through the estimation
of effective activation energies. In fact, typical non-isothermal crystallization analysis based on
the determination of crystal growth by optical microscopy allowed secondary nucleation constants
of 3.07 × 105 K2 and 1.42 × 105 K2 to be estimated. Microstructure of sutures was characterized by a
stacking of lamellae perpendicularly oriented to the fiber axis and the presence of interlamellar and
interfibrillar amorphous regions. The latter became enhanced during heating treatments due to loss
of partial chain orientation and decrease of electronic density. Degradation under various pH media
revealed different macroscopic morphologies and even a distinct evolution of lamellar microstructure
during subsequent heating treatments.

Keywords: polydioxanone; surgical suture; non-isothermal crystallization; lamellar stacking;

synchrotron radiation

1. Introduction
Poly(p-dioxanone) (PDX or PDO) is a synthetic poly(ester-ether) with wide applications in the
biomedical field due to its excellent properties (e.g., biodegradability, biocompatibility, bioabsorbability,
softness, and flexibility) [1]. In relation to polyglycolide, i.e., the first polyester employed for biomedical
applications, the chemical repeat unit of PDO has an ether bond and additional methylene groups that
provide greater flexibility. In this way, PDO can be used as a monofilar surgical suture, in contrast with
the braided polyglycolide suture. PDO can be completely reabsorbed in a period close to six months,
with no significant foreign body reaction being observed in the tissues surrounding the implant.

Polymers 2016, 8, 351; doi:10.3390/polym8100351 www.mdpi.com/journal/polymers

Polymers 2016, 8, 351 2 of 18

Several companies have commercialized PDO under different trademarks (e.g., PDS II and
MonoPlus™ by Ethicon and B. Braun Surgical S.A., respectively) as a long term surgical suture that
appears as an ideal wound support for healing periods longer than four weeks. Other biomedical
applications such as bone or tissue fixation devices, fasteners, and drug delivery systems [2] should
also be considered since the polymer can be easily injection-molded.
Isothermal crystallization of PDO has been extensively studied by different techniques, including
polarized optical microscopy and differential scanning calorimetry (DSC) [3–8]. Changes in
morphological parameters have also been evaluated by small-angle X-ray scattering [9]. Several
works have focused on the influence of degradation on morphology and isothermal crystallization
behavior [10–12].
Surprisingly, only a few reports concern the crystalline structure of PDO and lead to some
conflicting results. Thus, early studies pointed to an orthorhombic unit cell containing only two
molecular segments with a tight pitch (i.e., the chain axis repeat was shortened by 52% compared to
the expected value for an extended zig-zag conformation) [13]. Subsequently, an orthorhombic unit
cell with space group P21 21 21 and parameters a = 0.970 nm, b = 0.751 nm and c (chain axis) = 0.650 nm
was postulated using X-ray and electron diffraction data [14]. The structural model was also supported
by quantum mechanical calculations that indicated a chain periodicity given by a single residue with
a TGT(-G)TT conformation and a unit cell containing four molecular segments. Crystals obtained
from solution had a variable morphology (i.e., the acute apex angle of lozenge crystals varied with the
crystallization conditions) that seems a consequence of the peculiar structure of PDO, where different
folds for adjacent chain re-entry exist [14].
It is well established that isothermal crystallization of PDO from the melt renders spherulites
with a distinctive morphology depending on the crystallization temperature. Specifically, spherulites
obtained at high temperature were double-ringed [3] whereas at low temperature they showed a
negative birefringence and the typical Maltese cross. Isothermal crystallization kinetic studies linked
the different morphologies to two crystallization regimes [3]. Ringed and negative birefringent
spherulites were observed by evaporation of concentrated formic acid solutions. It was demonstrated
that the a crystallographic axis of constitutive lamellae was oriented along the spherulite radius [14].
The knowledge of non-isothermal crystallization kinetics is essential to achieve the proper
microstructure and the required properties of a material. The control of crystallization may reduce
undesired effects during processing such as excessive anisotropy of shrinkage, warpage, and
insufficient dimensional stability [15]. Nevertheless, it should be considered that non-isothermal
crystallizations studies are usually limited to idealized situations, in which external conditions such as
the cooling rate is kept constant, although in real situations polymers are cooled down at different
rates and with high thermal gradients [16].
Non-isothermal crystallization of PDO has been scarcely studied despite the fact that
melt-processed samples (e.g., bioabsorbable surgical sutures) are obtained under non-isothermal
conditions. However, some non-isothermal melt crystallization analyses have been performed from
DSC data and by applying Ozawa [17] and Cazé [18] methodologies to evaluate Avrami exponents [4].
Values calculated by Ozawa were dependent on the temperature, whereas two different values of
the Avrami exponent were determined depending on the cooling rate [6]. Non-isothermal cold
crystallization has also been evaluated by calorimetric analysis [19] and Avrami [20,21], Ozawa [17],
and Tobin [22] methodologies. In this case, results suggested that the Avrami method was more
effective in describing non-isothermal cold crystallization kinetics, with reported values of Avrami
exponent in the 4.5–5.3 range [19].
PDO has a limited use, for example, as blow processed films due to its low crystallization rate and
melt strength, together with its high cost and relatively low thermal stability. Therefore, non-isothermal
crystallization studies have mainly focused on blends of PDO with other polymers such as poly(vinyl
alcohol) [23] and polylactide [24]. It was assumed that the crystallization rate should be enhanced in
blends. Thus, an improvement of properties and easier processing were determined.
Polymers 2016, 8, 351 3 of 18

The purpose of the present work was to perform a complete analysis of non-isothermal
crystallization considering both the overall kinetic process from DSC data and crystal growth rates
from optical microscopy data. Furthermore, morphological changes occurring during heating and
cooling processes were analyzed by real time synchrotron experiments. Dynamic diffraction data from
samples degraded under different pHs were useful in confirming a lamellar organization inside fibers.
In this way, recent studies on degraded monofilament sutures constituted by polyglycolide hard blocks
highlighted a structural organization with interlamellar and interfibrillar domains [25].

2. Experimental Section
Materials: Granulated PDO and processed PDO sutures (Monoplus™, USP 0) were kindly
supplied by B. Braun Surgical, S.A (Rubí, Spain). In order to get PDO sutures, granulated polymer
was gradually melted along the extruder barrel under an inert atmosphere applying a temperature
gradient from 140 to 190 ◦ C. The poly(p-dioxanone) polymer melt was metered by a spin pump with a
specific throughput and extruded through a spinneret, which shapes the material to a monofilament
fiber with a circular cross section. After passing through the spinneret, the fiber was quenched in a cold
water bath to stabilize it. In order to obtain good tensile properties, the fiber was drawn and relaxed
at certain drawing ratios, passing it through convection ovens. Finally, the fiber was taken up on
drums winders. Number and weight average molecular weights determined by GPC were 112,800 and
250,000 for the granulated material and 103,000 and 226,300 for the commercial threads, respectively.
Measurements: Molecular weights were estimated by size exclusion chromatography (GPC) using
a liquid chromatograph (Shimadzu, model LC-8A, Kyoto, Japan) equipped with an Empower computer
program (Waters, Mildford, MA, USA). A PL HFIP gel column (Polymer Lab) and a refractive index
detector (Shimadzu RID-10A, Kyoto, Japan) were employed. The polymer was dissolved and eluted in
1,1,1,3,3,3-hexafluoroisopropanol containing CF3 COONa (0.05 M) at a flow rate of 1 mL/min (injected
volume 100 µL, sample concentration 2.0 mg/mL). The number and weight average molecular weights
were calculated using polymethyl methacrylate standards.
Calorimetric data were obtained by differential scanning calorimetry with a TA Instruments Q100
series (TA instruments, New Castle, DE, USA) with Tzero technology and equipped with a refrigerated
cooling system (RCS, TA instruments, New Castle, DE, USA). Experiments were conducted under
a flow of dry nitrogen with a sample weight of approximately 5 mg and calibration was performed
with indium. Tzero calibration required two experiments: the first was performed without samples
while sapphire disks were used in the second. Thermal characterization was conducted following a
protocol consisting of a heating run (3 ◦ C/min), a cooling run (3 ◦ C/min) after keeping the sample
in the melt state for 5 min to wipe out the thermal history, and a subsequent heating run (3 ◦ C/min).
Non-isothermal crystallization studies were performed by cooling the previously molten samples
(5 min at 125 ◦ C) at rates varying from 30 to 1 ◦ C/min.
The spherulitic growth rate was determined by optical microscopy using a Zeiss Axioskop 40
Pol light (Zeiss, Oberkochen, Germany) polarizing microscope equipped with a Linkam temperature
control system (Tadworth, UK) configured by a THMS 600 heating and freezing stage connected to a
LNP 94 liquid nitrogen cooling system (Tadworth, UK). Spherulites were grown from homogeneous
thin films prepared from the melt. Small sections of these films were pressed or smeared between two
cover slides and inserted into the hot stage, giving rise to samples with thicknesses close to 10 µm in
all cases. Samples were kept at approximately 125 ◦ C for 5 min to eliminate sample history effects.
The radius of growing spherulites was monitored during crystallization with micrographs taken with
a Zeiss AxiosCam MRC5 digital camera (Zeiss, Oberkochen, Germany) at appropriate time intervals.
A first-order red tint plate was employed to determine the sign of spherulite birefringence under
crossed polarizers.
In vitro hydrolytic degradation assays were carried out at a physiological temperature of
◦
37 C using a pH 7.4 phosphate buffer (Sörensen medium: 19.268 g of Na2 HPO4 ·12H2 O
and 1.796 g of KH2 PO4 in 1 L of deionized water) and a pH 11 from the Universal buffer
Polymers 2016, 8, 351 4 of 18

(citrate-phosphate-borate/HCl) solution, mixing 20 mL of a stock solution with 14.7 mL of 0.1 M HCl

and distilled water up to a volume of 100 mL. The stock solution (1 L) contained 100 mL of citric acid
and 100 mL of phosphoric acid solution, each of which was equivalent to 100 mL NaOH 1 M, 3.54 g
of boric acid, and 343 mL of 1 M NaOH. Samples were kept under orbital shaking in bottles filled
with 50 mL of the degradation medium and sodium azide (0.03 wt %) to prevent microbial growth
for selected exposure times. The samples were then thoroughly rinsed with distilled water, dried to
constant weight under vacuum and stored over P4 O10 before analysis. Finally, weight retention and
molecular weight were evaluated.
Time resolved SAXS (small-angle X-ray scattering) experiments were conducted at the NCD
beamline (BL11) of the Alba synchrotron radiation light facility (Cerdanyola del Vallès, Spain).
The beam was monochromatized to a wavelength of 0.1 nm. Polymer samples were confined between
Kapton films and then held on a Linkam THMS600 hot stage (Tadworth, UK) with temperature
control within ±0.1 ◦ C. SAXS profiles were acquired during heating and cooling runs in time frames
of 20 s and rates of 10 ◦ C/min. The SAXS patterns were calibrated with diffractions of a standard
of a silver behenate sample. The diffraction profiles were normalized to the beam intensity and
corrected considering the empty sample background. The correlation function and corresponding
parameters were calculated with the CORFUNC program for Fiber Diffraction/Non-Crystalline
Diffraction provided by Collaborative Computational Project 13 (CCP13) (Chester, UK).

3. Results and Discussion

3.1. Melting and Crystallization of Poly(p-dioxanone)

Thermal behavior of PDO is rather complicated, as revealed by the DSC curves in Figure 1.
Heating traces are clearly different for commercial granulated samples crystallized from the melt and
processed surgical sutures, demonstrating the significant influence of thermal treatments on melting
behavior. Thus, two well-differentiated endothermic melting peaks at 97 ◦ C (small) and 107 ◦ C were
observed when samples crystallized from the melt state at a rate of 3 ◦ C/min. These peaks have been
largely discussed [6,7] and attributed to a typical lamellar reorganization process that leads to lamellar
thickening during heating. Commercial granulated PDO showed that the high temperature peak was
split into two equally intense peaks at 104 and 107 ◦ C. Therefore, the population of thinner crystals was
not present in the manufactured PDO form or at least they were sufficiently energetically unstable to
lead to a complete recrystallization process on heating. Probably two populations of new crystals with
practically the same thickness existed. The double peak appearing in the 104–107 ◦ C range (see label at
105 ◦ C) was also detected in the heating run of the as-processed suture, but in this case the population
of thinner crystals was highly stable and did not undergo a recrystallization process. Note the intense
and narrow peak at 98 ◦ C, which in this case was observed as a consequence of the high temperature
annealing process at which commercial sutures were submitted.
Figure 2 shows the dynamic DSC exotherms obtained by cooling the melted samples at different
rates. Crystallization peaks progressively shift to lower temperatures as the cooling rate increases,
as expected, but peaks become broader. Two different crystallization processes seemed to occur at
cooling rates equal or higher than 15 ◦ C/min, a feature that is later discussed on the basis of different
nucleation mechanisms (i.e., instantaneous and sporadic at low and high temperatures, respectively).

3.2. Non-Isothermal Kinetic Analysis of Poly(p-dioxanone) Melt Crystallization from DSC Data
The process of crystallization under non-isothermal conditions is rather complicated to be
analyzed since crystallization from the melt takes place under different degrees of supercooling,
and therefore caution should be taken when interpreting experimental results.
Polymers 2016, 8, 351 5 of 18
Polymers 2016, 8, 351 5 of 18
Polymers 2016, 8, 351 5 of 18

Figure 1. Differential scanning calorimetry (DSC) traces corresponding to heating runs of commercial
Figure 1. Differential scanning calorimetry (DSC) traces corresponding to heating runs of commercial
granulated polydioxanone
Figure 1. Differential scanning(PDO) (a) and PDO
calorimetry (DSC)sutures
traces (d), the cooling to
corresponding run of the runs
heating melted granulated
of commercial
granulated polydioxanone (PDO) (a) and PDO sutures (d), the cooling run of the melted granulated
PDO (b) and
granulated the subsequent
polydioxanone heating
(PDO) run PDO
(a) and (c). Glass
suturestransition
(d), the can be detected
cooling in melted
run of the the magnification
granulated
PDO given
(b) and the inset
in the
subsequent heatingwererun (c). Glass transition can be detected in the magnification given
PDO (b) and the in (c). All scans
subsequent heating runperformed at atransition
(c). Glass rate of 3 °C/min.
can be detected in the magnification
in thegiven
insetininthe
(c).inset
All in
scans were
(c). All performed
scans at a rate
were performed 3 ◦ C/min.
at aofrate of 3 °C/min.

Figure 2. Exothermic DSC traces performed with PDO at the indicated cooling rates. The dashed
ellipse
Figure contains the high
2. Exothermic temperature
DSC crystallization
traces performed peakat
with PDO detected at high cooling
the indicated cooling rates.
rates, whereas the
The dashed
Figure 2. Exothermic DSC traces performed with PDO at the indicated cooling rates. The dashed
dashed arrow indicates
ellipse contains the highthe evolution of
temperature the main crystallization
crystallization peak.
peak detected at high cooling rates, whereas the
ellipse contains the high temperature crystallization peak detected at high cooling rates, whereas the
dashed arrow indicates the evolution of the main crystallization peak.
dashed arrow indicates
Calorimetric the evolution
data were of the main
used to determine thecrystallization
relative degree peak.
of crystallinity at any temperature,
χ(T), Calorimetric
for all cooling rates
data by the
were usedexpression:
to determine the relative degree of crystallinity at any temperature,
χ(T), for all cooling
Calorimetric rates by
data were the to
used expression:
determine the Tc relative degree of crystallinity at any temperature,
χ(T), for all cooling rates by the expression: TT0c ( d H c / d T ) d T
χ (T ) =  T∞ ( d H c / d T ) d T (1)
χ (T ) = rTTT0c∞ (d H c / d T )d T
T
(1)
χ( T ) = rTT0
T0 ((d
0 dHH c /dT
/ d T ))ddT
T
c (1)
∞
T0 (dHc /dT )dT

where dHc is the enthalpy of crystallization released within an infinitesimal temperature range dT,
T0 denotes the initial crystallization temperature and Tc and T ∞ are the crystallization temperatures
 Polymers
Polymers 2016,2016, 8, 351
8, 351 6 of 18
6 of 18

where dHc is the enthalpy of crystallization released within an infinitesimal temperature range dT,
at time t and the
T0 denotes after completion
initial of thetemperature
crystallization crystallization
and Tprocess, respectively.
c and T∞ are Thus,temperatures
the crystallization the denominator
at
time t and
corresponds afteroverall
to the completion
enthalpyof the crystallization process,
of crystallization respectively.
for specific Thus, the
heating/cooling denominator
conditions.
corresponds
The relativetodegree
the overall enthalpy of was
of crystallinity crystallization
calculatedfor asspecific heating/cooling
a function of time by theconditions.
relationship
The relative degree of crystallinity was calculated as a function of time by the relationship
(t − (tt0−) t0)== (T − φT)/φ
(T00− T)/ (2) (2)

where 0 is the temperature at which crystallization begins (t = t0) and φ is the value of the cooling
where T0 isTthe temperature at which crystallization begins (t = t0 ) and φ is the value of the cooling rate.
rate.
Figure 3a illustrates the variation of the time-dependent degree of crystallinity, χ(t − t0 ), at
Figure 3a illustrates the variation of the time-dependent degree of crystallinity, χ(t − t 0 ), at
different cooling rates, which allows a typical Avrami analysis to be performed [20] according to
different cooling rates, which allows a typical Avrami analysis to be performed [20] according to
the equation
the equation
1 − χ(t − t0 ) = exp[−Z(t − t0 )n ], (3)
1 − χ(t − t0) = exp[−Z(t − t0)n], (3)
where Z is the rate constant and n is the Avrami exponent. A normalized rate constant, k = Z1/n , is
where Z is the rate constant and n is the Avrami exponent. A normalized rate constant, k = Z1/n, is
however more useful for comparison because its dimension (time−1−1 ) becomes independent of the
however more useful for comparison because its dimension (time ) becomes independent of the
Avrami exponent.
Avrami exponent.
Figure 3b 3b
Figure shows
showsthe
theplots
plots of log{−ln[1
of log{−ln[1 − −χ(t
− χ(t −versus
t0)]} t0 )]} versus
log(t − tlog(t − t0 ) atcooling
0) at different
different cooling
rates. A
rates.
good linearity was observed between the relative degree of crystallinities of 0.10 and 0.90, that is,0.90,
A good linearity was observed between the relative degree of crystallinities of 0.10 and
that after
is, after formation
formation of well-defined
of well-defined spherulitic
spherulitic morphologies
morphologies and and before
before occurrence
occurrence of a of a secondary
secondary
crystallization caused
crystallization byby
caused thethe
impingement
impingementof ofspherulites (seedashed
spherulites (see dashedlineslinesinin Figure
Figure 3a).3a).

Figure 3. (a) Time evolution of relative crystallinity at the indicated cooling rates for non-isothermal
Figure 3. (a) Time evolution of relative crystallinity at the indicated cooling rates for non-isothermal
crystallization of PDO; (b) Avrami analyses of non-isothermal crystallizations of PDO.
crystallization of PDO; (b) Avrami analyses of non-isothermal crystallizations of PDO.
Table 1 summarizes the main kinetic parameters calculated by the Avrami analysis. As known
Tableisothermal
from 1 summarizes thethe
studies, main kinetic parameters
normalized rate constantcalculated
was low by the
(i.e., Avrami0.58
between analysis.
× 10−3 sAs known
−1 and

from15.26
isothermal 0.58 × 10 − 3 s−1 and
× 10 s )studies,
−3 −1 the normalized
and increased rate constant
with the cooling wasexponents
rate. Avrami low (i.e.,showed
betweena moderate variation
× 10
15.26(i.e., − 3
between − 1
s )3.76 andand 2.77), with
increased withthe lowest
the values
cooling rate.being determined
Avrami exponents forshowed
high cooling rates and
a moderate the
variation
(i.e.,average
between value3.76being
andclose
2.77),to with
3.0. These exponents
the lowest are lower
values beingthan those previously
determined reported
for high coolingby Zhang
rates and
et al. [24]value
the average (i.e., 4.26–3.40)
being close and
to in
3.0.good agreement
These exponentswithare
those given
lower than bythose
Andjelic et al. [4]reported
previously for low by
crystallization rates (i.e., 3.0), although in this case a value of 1.1 was found for high
Zhang et al. [24] (i.e., 4.26–3.40) and in good agreement with those given by Andjelic et al. [4] for low crystallization
rates. Isothermal crystallization studies also indicate slightly contradictory values. Thus, minimum
crystallization rates (i.e., 3.0), although in this case a value of 1.1 was found for high crystallization rates.
changes with crystallization temperature were determined by Andjelic et al. (i.e., exponents varied
Isothermal crystallization studies also indicate slightly contradictory values. Thus, minimum changes
between 2.22 and 2.62, with 2.5 being the average value) [4], but a systematic increase (i.e., from
with crystallization temperature were determined by Andjelic et al. (i.e., exponents varied between
approximately 2 to 3.8) was also reported [6] for higher isothermal crystallization temperatures (i.e.,
2.22 and 2.62, with 2.5 being the average value) [4], but a systematic increase (i.e., from approximately
2 to 3.8) was also reported [6] for higher isothermal crystallization temperatures (i.e., from 30 to
80 ◦ C). The last behavior was interpreted as a consequence of a change from instantaneous to sporadic
nucleation as Tc was increased [6].
Polymers 2016, 8, 351 7 of 18

It can be well stated that application of the Avrami equation under non-isothermal conditions
merely corresponds to a mathematical fitting that allows appropriate values of the rate constant to
be derived [26–28]. In this case, it should be pointed out that the determined exponents may even
have a physical meaning since they suggest three-dimensional spherulitic growth and instantaneous
nucleation, as postulated from isothermal studies [4,6]. Furthermore, the sporadic nucleation detected
at high isothermal crystallization temperatures [6] is in agreement with the increase of the exponent
observed at low crystallization rates (i.e., 3.76 at 1 ◦ C/min) and supports DSC evidence of the
occurrence of two crystallization processes.

Table 1. Main non-isothermal crystallization kinetic parameters of polydioxanone (PDO) determined

by differential scanning calorimetry (DSC).

φ (◦ C/min) n Z (s−n ) k × 103 (s−1 ) τ 1/2 (s) (1/τ 1/2 ) × 103 (s−1 ) (Z/ln2)1/n × 103 (s−1 )
1 3.76 6.79 × 10−13 0.58 1,574 0.64 0.64
3 3.11 7.90 × 10−10 1.18 749 1.34 1.32
5 3.30 1.37 × 10−9 2.06 428 2.34 2.31
8 3.10 1.83 × 10−8 3.21 272 3.68 3.61
10 2.92 1.45 × 10−7 4.55 192 5.21 5.16
15 2.97 2.76 × 10−7 6.17 142 7.07 6.98
20 2.73 2.25 × 10−6 8.60 104 9.65 9.83
30 2.77 9.12 × 10−6 15.26 58 17.14 17.42

The values of the corresponding reciprocal crystallization half times (1/τ 1/2 ), calculated as the
inverse of the difference between crystallization starting time and crystallization half time, are also
given in Table 1. This parameter is a direct measure of the crystallization process, and could therefore
be used to check the accuracy of Avrami analyses, as demonstrated by the excellent agreement with
the theoretical kinetic value (i.e., 1/τ 1/2 = (Z/ln2)1/n ). In conclusion, the deduced Avrami parameters
are completely appropriate to simulate the non-isothermal crystallization process.
A kinetic equation that combines the Avrami [20] and Ozawa [17] expressions has been derived
and applied in different non-isothermal studies [29]:

log φ = log F(T) − a log(t − t0 ) (4)

where F(T) is a new kinetic parameter referring to the cooling rate which must be chosen at a unit
crystallization time when the system reaches a certain crystallinity, and a is the ratio of apparent
Avrami and Ozawa exponents.
A plot of log φ versus log (t − t0 ) yields a series of straight lines at a given value of χ(T) (Figure 4),
which suggest the validity of the combined equation for this system. Kinetic parameters can be
estimated by the intercept and slope of these lines.
Results showed that F(T) values increased with crystallinity (Table 2), which has a physical sense
because the motion of the molecular chains became slower as the material crystallized and formation
of new crystals was hindered. The values of a were almost constant between 0.91 and 0.98 although
slightly increased with the relative degree of crystallinity.
The crystallization process has non-Arrhenius behavior, and therefore a temperature-dependent
effective activation energy needs to be defined. The value corresponding to a given degree of crystallinity,
Eχ , can be determined by the Friedman isoconversional method [30]:

[dχ/dt]χ = Aexp(−Eχ /RT)f [χ], (5)

where A is a pre-exponential factor and f [χ] is the crystallization model. Values of ln[dχ/dt]χ at
different temperatures and degrees of crystallization can be obtained from crystallization experiments
performed at different cooling rates. In this way, the slopes of the linear plots of ln[dχ/dt]χ versus
1/T (Figure 5a) allow Eχ to be determined (Figure 5b). The temperature dependence of the effective
Polymers 2016, 8, 351 8 of 18

Polymers 2016, 8, 351 8 of 18

activation
Polymersenergy (Figure 5c) could finally be derived by considering also the average temperature
2016, 8, 351 8 of 18
of the effective activation energy (Figure 5c) could finally be derived by considering also the average
associated with a given conversion (Figure 5b).
temperature associated with a given conversion (Figure 5b).
of the effective activation energy (Figure 5c) could finally be derived by considering also the average
temperature associated with a given conversion (Figure 5b).

Figure 4. Plots of logφ versus log (t − t0) for non-isothermal crystallization of PDO performed at the
Figure 4. Plots of log (φ) versus log (t − t0 ) for non-isothermal crystallization of PDO performed at the
indicated crystallinities.
Figure
indicated 4. Plots of logφ versus log (t − t0) for non-isothermal crystallization of PDO performed at the
crystallinities.
indicated crystallinities.

Figure 5. (a) Plots of ln[dχ/dt]χ versus 1/T for non-isothermal crystallization of PDO at the indicated
cooling rates. Data corresponding to relative degrees of crystallinity of 0.8, 0.5 and 0.1 are represented
Figure 5. (a) Plots of ln[dχ/dt]χ versus 1/T for non-isothermal crystallization of PDO at the indicated
Figure (a) Plots
by5.blue, greenof ln[dχ/dt] χ versus 1/T
and red symbols, for non-isothermal
respectively; (b) Dependence crystallization of PDO atenergy
of the activation the indicated
of
cooling rates. Data corresponding to relative degrees of crystallinity of 0.8, 0.5 and 0.1 are represented
cooling rates. Data(●)
crystallization corresponding to relative degrees
and average temperature of crystallinity
(○) on crystallinity; of 0.8, 0.5 and
(c) Experimental 0.1 are
Eχ versus represented
T plot and
by blue, green and red symbols, respectively; (b) Dependence of the activation energy of
simulated
by blue, curves
green and redaccording
symbols, to Equation (6). Red and violet dashed lines indicate energy
the simulated curves
crystallization (●) and averagerespectively;
temperature (b) Dependence
(○) on crystallinity;of(c)
the activation
Experimental of Tcrystallization
Eχ versus plot and
for regimes III and II, respectively. Arrows indicate the expected temperatures for the maximum
( ) and average
simulated temperature
curves according (#) on crystallinity;
to Equation (6). Red and(c)violet dashed linesEindicate
Experimental χ versus theTsimulated
plot andcurves
simulated
curvescrystallizationtorates (i.e., effective
(6). Redactivation energy equal to zero).
foraccording Equation
regimes III and II, respectively. and violet
Arrows dashed
indicate lines
the indicate
expected the simulated
temperatures for curves for regimes
the maximum
III and II, respectively.
crystallization Arrows
rates (i.e., indicate
effective the expected
activation temperatures
energy equal to zero). for the maximum crystallization
The effective activation energy was negative for crystallization experiments performed from the
rates (i.e., effective activation energy equal to zero).
melt state and at low supercooling degrees (i.e., the temperature range where secondary nucleation
The effective activation energy was negative for crystallization experiments performed from the
plays a fundamental role), as shown in Figure 5c. This energy increased progressively (i.e., the
melt state and at low supercooling degrees (i.e., the temperature range where secondary nucleation
The effective activation energy was negative for crystallization experiments performed from
plays a fundamental role), as shown in Figure 5c. This energy increased progressively (i.e., the
the melt state and at low supercooling degrees (i.e., the temperature range where secondary
nucleation plays a fundamental role), as shown in Figure 5c. This energy increased progressively
Polymers 2016, 8, 351 9 of 18

(i.e., the crystallization rate increased) as the temperature decreased, as discussed at length by
Vyazovkin and Dranca [31], reflecting the expected behavior for crystallizations performed at
temperatures higher than those associated with the maximum crystallization rate.
Vyazovkin and Sbirrazzuoli proposed that crystallization parameters like the secondary nucleation
constant should be derived through the effective activation energies [31–33]:

E(T) = −RdlnG/dT −1 = U*[T2 /(T − T∞ )2 ] + Kg R[(2∆T − Tm ◦ f )/(∆T)2 f ] (6)

where G is the crystal growth rate, U* represents the activation energy characteristic of the transport
of crystallizing segments across the liquid-crystal interface, T ∞ is the temperature below which such
motion ceases, R is the gas constant, Kg is the secondary nucleation constant, ∆T is the degree of
supercooling measured as Tm − Tc (where Tm is the equilibrium melting temperature and Tc is the
crystallization temperature), and f is a correction factor accounting for the variation in the bulk melting
enthalpy per unit volume with temperature (f = 2Tc /(Tm + Tc )).
Figure 5c also compares the experimental Eχ − T plot with simulated ones using Equation (6), U*
and T∞ values of 1600 cal/mol and Tg -35 K, respectively (i.e., close to the universal values reported by
Suzuki and Kovacs [34]), the equilibrium melting temperature of 127 ◦ C, as previously determined for
PDO [3], and representative Kg values. In fact, U* and T∞ have little influence on a temperature range
that is far from the glass transition temperature. The best fit between experimental and theoretical
data was obtained considering two Kg parameters (i.e., 3.07 × 105 K2 and 1.42 × 105 K2 ), which are
in agreement with the two crystallization regimes reported for PDO from isothermal crystallization
experiments [3]. Note that the isoconversional analysis was able to detect the existence of several
crystallization regimes and also to predict two maximum crystallization rates at temperatures of 45
and 60 ◦ C, as deduced from the temperatures for each simulated curve where the effective activation
energy was zero.

Table 2. Values of kinetic parameters at a given crystallinity estimated from the combined model [35]
for non-isothermal crystallization of PDO.

χ(T) a F(T) r2
0.1 0.91 12.67 0.993
0.2 0.93 16.98 0.995
0.3 0.94 20.79 0.996
0.4 0.95 24.27 0.996
0.5 0.96 27.64 0.996
0.6 0.96 31.12 0.996
0.7 0.97 35.04 0.995
0.8 0.98 41.51 0.995
0.9 0.98 48.11 0.993

3.3. Non-Isothermal Kinetic Analysis of Poly(p-dioxanone) Melt Crystallization from Optical Microscopy Data
Non-isothermal crystallization of PDO rendered double banded spherulites with progressively
decreasing periodicity (Figure 6), which is in accordance with the continuous temperature decrease of a
non-isothermal crystallization. In fact, morphologies obtained under isothermal conditions have been
extensively studied [3,5,6], and it was assumed that interband spacing decreased when crystallization
temperature was lowered. Actually, two different banding periodicities could be detected where the
broader bands had a negative birefringence. These kinds of double bands with uneven spacings are
characteristic of spherulites having a biaxial indicatrix twisted about the optic normal [6,32,35]. PDO
spherulites were also characterized by their big size, which led to poor nucleation and slow growth rate.
Figure 6 also shows that the number of active nuclei increased during cooling, and consequently the size
and morphology of the spherulites were not identical. Logically smaller spherulites with a practically
indistinguishable double band texture were formed at lower temperatures (see yellow arrows).
Polymers 2016, 8, 351 10 of 18
Polymers 2016, 8, 351 10 of 18

Figure 6.
Figure 6. Optical
Optical micrograph
micrograph of of PDO
PDO spherulites
spherulites formed
formed during
during aa non-isothermal
non-isothermal crystallization
crystallization from
from
the melt state performed at a cooling rate of 20 C/min. Yellow arrows point to spherulites formed
the melt state performed at a cooling rate of 20 °C/min.
◦ Yellow arrows point to spherulites at
formed at
low temperatures.
low temperatures. Inset
Inset shows
shows aa micrograph
micrograph taken
taken with
with aa first-order
first-order red tint plate
red tint plate to
to determine
determine the
the
birefringence sign.
birefringence sign.

Spherulitic growth rates (G) were also determined for non-isothermal crystallization by
Spherulitic growth rates (G) were also determined for non-isothermal crystallization by measuring
measuring the change of the spherulite radius (R) with temperature (T) at a constant cooling/heating
the change of the spherulite radius (R) with temperature (T) at a constant cooling/heating rate
rate (dT/dt) [36,37]:
(dT/dt) [36,37]:
G = GdR/dt
= dR/dt== (dR/dT)
(dR/dT)× (dT/dt)
× (dT/dt) (7)
(7)
Plots showing
Plots showing the variation of
the variation of the
the spherulitic
spherulitic radius
radius with
with crystallization
crystallization temperature
temperature could
could be
be
adjusted to
adjusted to third
thirdorder
orderequations
equationswith
withgood
goodregression
regression coefficients
coefficients (i.e.,
(i.e., higher
higher thanthan 0.990)
0.990) (Figure
(Figure 7a).
7a). These
These coefficients
coefficients were were significantly
significantly better
better thanthan those
those calculatedfor
calculated forsecond
secondorder
order equations
equations and
and
remained constant
remained constant for
for higher
higher orders.
orders. Therefore,
Therefore, third order equations
third order equations were were employed
employed to to determine
determine
dR/dT as a function of the crystallization temperature. The corresponding crystal growth rate versus
dR/dT as a function of the crystallization temperature. The corresponding crystal growth rate versus
crystallization temperature
crystallization temperature curves
curves are displayed in
are displayed in Figure
Figure 7b.
7b. Note
Note thatthat data
data were obtained at
were obtained at
different cooling
different coolingrates
ratesininorder
ordertotomaximize
maximizethe thecrystallization
crystallization temperature
temperature range
range where
where radii
radii could
could be
be well measured.
well measured.
Two bell-shaped
Two bell-shaped curves
curves with
with maximums
maximums of of 45
45 and
and 60
60 ◦°C reflected the
C reflected the temperature
temperature dependence
dependence
of G,
of G, and
and therefore
therefore the
the existence
existence ofof two
two crystallization
crystallization regimes
regimes withwith different
different secondary
secondary nucleation
nucleation
constants. These were determined by the Lauritzen-Hoffman
constants. These were determined by the Lauritzen-Hoffman equation [38]: equation [38]:
G = G0exp[−U*/(R(Tc − T∞))] × exp[−Kg/(Tc(ΔT)f)] (8)
G = G0 exp[−U*/(R(Tc − T∞ ))] × exp[−Kg /(Tc (∆T)f )] (8)
where G0 is the constant pre-exponential factor and the other parameters as previously defined.
where G0 is the
Figure constant
7c shows thepre-exponential factor and
linear plots obtained the other
using U* and parameters as previously
T∞ parameters of 1600defined.
cal/mol and
Tg-35Figure 7c shows the
K, respectively. It islinear
clear plots obtained
that two using U* regimes
crystallization and T∞ defined
parameters of 1600 cal/mol
by secondary and
nucleation
T g -35 K, respectively.
constants of 3.07 × 105 K It2 is
andclear
1.42that
× 10two
5 K2 crystallization
fit the experimentalregimes defined
data. by secondary
Furthermore, regimes nucleation
III and II
constants
could be assumed 105 K2the
of 3.07 × since 1.42 × 105 K2ratio
andexperimental fit the experimental
between slopes data.
(2.16)Furthermore,
was close to regimes III and
the theoretical
II
Kgcould
III be assumed
/Kg value
II of 2. since the experimental ratio between slopes (2.16) was close to the theoretical
Kg III /K II
g value
Figure of 2.
7b also shows that the two bell-shaped curves calculated by equation 8, the estimated U*
and T∞ parameters,shows
Figure 7b also and thethatdeduced
the two bell-shaped
values of lncurvesG0 and calculated by Equation
Kg for each regime (8), the estimated
fit well with the
U* and T ∞ parameters,
experimental spherulitic and growththe data.
deduced The values
maximum G0 andrate
of lngrowth Kg for
waseach regime
found fit well
in regime III with
and
the experimental
corresponded to aspherulitic
temperature growthof 45data.
°C. TheOur maximum
observations growth
are inrate was found
agreement in regime
with the same III
and corresponded to adetermined
temperature of 45 ◦ C. Our crystallization
observations are in agreement with the same
crystallization regimes from isothermal although the nucleation constant
crystallization
becomes slightly regimes
higherdetermined from isothermal
than those previously crystallization
reported although
(i.e., 2.49 × 10 5 K2 andthe nucleation
1.19 constant
× 105 K2) [3]. Note
becomes slightly higher than those previously reported (i.e., 2.49 × 105 K2 and 1.19 × 105 K2 ) [3]. Note
also that the deduced data support the results of the calorimetric study.
also that the deduced data support the results of the calorimetric study.
Polymers 2016, 8, 351 11 of 18
Polymers 2016, 8, 351 11 of 18

Polymers 2016, 8, 351 11 of 18

Figure 7. (a) Variation in spherulite radius with temperature during heating at the indicated rates;
Figure 7. (a) Variation in spherulite radius with temperature during heating at the indicated rates;
(b) Spherulitic growth rates determined by the equations deduced for the heating runs. Theoretical
(b) Spherulitic growth rates determined by the equations deduced for the heating runs. Theoretical
curves are also drawn (dashed lines) for comparative purposes; (c) Plot of lnG + U*/R(Tc − T∞) versus
curves are also drawn (dashed lines) for comparative purposes; (c) Plot of lnG + U*/R(Tc − T ∞ ) versus
1/Tc(ΔT)f to determine the Kg secondary nucleation parameters of PDO.
1/Tc (∆T)f
Figureto7.determine theinKspherulite
(a) Variation g secondary nucleation
radius parameters
with temperature of PDO.
during heating at the indicated rates;
(b) Spherulitic
The growth
fact that the rates determined
crystallization rate isbygoverned
the equations
by twodeduced for the
different heating runs.
processes makes Theoretical
it unfeasible
to curves area also
determine drawn
single (dashedenergy
activation lines) for
forcomparative
the entire purposes;
T range. (c) Plot ofan
Instead, lnG
effective c − T∞) versus
+ U*/R(Tactivation energy to
The fact that the crystallization rate is governed by two different processes makes it unfeasible
c

(E) 1/Tc(ΔT)f toon

dependent determine
T the Kg secondary
c was evaluated by nucleation
equation 9 parameters of PDO.
[32]:
determine a single activation energy for the entire T range. Instead, an effective activation energy (E)
c
dependent
The Tc was
onfact evaluated
thatEthe
= −RdlnG/dT by−1 Equation
= U*T
crystallization rate/(T
2 −(9)
T∞)[32]:
2 + KgR[(Tm°)2 − T2−TmT]/[(Tm − T)2T]
is governed (9)
by two different processes makes it unfeasible
to determine a singleeffective
The calculated activation energy for
activation the entire
energies areTplotted
c range. in
Instead,
Figurean8 effective
and show activation energy
non-Arrhenius
−1
−RdlnG/dT
E(E)= dependent U*T2 /(Tby−equation
= evaluated
on Tc was Kg R[(Tm ◦ )2 − T2 −Tm T]/[(Tm − T)2 T]
T ∞ )2 9+[32]: (9)
behavior as expected. Different Kg values were used according to the crystallization regime. The
effective activation =energy is zero
−1 =at the2/(T
maximum
− T∞)2 + Kcrystallization rate for regimes III and II, which
The calculated E −RdlnG/dT
effective activation U*T
energies aregR[(T m°)2 − T2−TmT]/[(Tm − T)2T]
plotted in Figure 8 and show non-Arrhenius
corresponds to temperatures of 45 and 60 °C and agree again with those deduced by isoconversional
(9)

behavior as expected.
The
analysis. calculated
In each case, Different
effective
positive Kg values
activation
values are were
energies
found forused
are plottedaccording
in Figure
temperatures to the
8 and
lower crystallization
show
than the non-Arrhenius
corresponding regime.
behavior
The effective as expected.
maxima activation
because theenergyDifferent K values
is zero atrate
crystallization g were
theincreases
maximum used according to
crystallization
with the crystallization
rate for regimes
increasing temperature, regime.
III and
whereas The
II,
negative which
effective
corresponds
values areactivation
to energy
temperatures
determined is 45
of zero
for higher andat the
60 ◦maximum crystallization
C andcharacterized
temperatures agree againby withrate for deduced
those
a decrease regimes III
byand II, which
isoconversional
of the crystallization rate
corresponds
with
analysis. In eachtocase,
increasing temperatures
positiveofvalues
temperature. 45 and 60 °C and agree again with those deduced by isoconversional
are found for temperatures lower than the corresponding
analysis. In each case, positive values are found for temperatures lower than the corresponding
maxima because the crystallization rate increases with increasing temperature, whereas negative
maxima because the crystallization rate increases with increasing temperature, whereas negative
values are determined for higher temperatures characterized by a decrease of the crystallization rate
values are determined for higher temperatures characterized by a decrease of the crystallization rate
with increasing
with increasingtemperature.
temperature.

Figure 8. Dependence of effective activation energy on crystallization temperature for regimes II (■)
and III (•). Extrapolated data for regimes II and III are indicated by dotted and dashed lines,
respectively.

Figure 8. Dependence of effective activation energy on crystallization temperature for regimes II (■)
Figure 8. Dependence of effective activation energy on crystallization temperature for regimes II () and
and III (•). Extrapolated data for regimes II and III are indicated by dotted and dashed lines,
III (•). Extrapolated data for regimes II and III are indicated by dotted and dashed lines, respectively.
respectively.
Polymers 2016, 8, 351 12 of 18
Polymers 2016, 8, 351 12 of 18

3.4.
3.4. Evolution
Evolution of
of Morphologic
Morphologic Parameters
Parameters during
during Heating
Heating of
of Poly(p-dioxanone)
Poly(p-dioxanone) Samples
Samples
Figure
Figure 99shows
showsthe evolution
the of the
evolution intensity
of the of the
intensity ofpeak
the detected in SAXS
peak detected in patterns during heating
SAXS patterns during
and cooling processes of the granulated PDO sample. In the first case, a recrystallization
heating and cooling processes of the granulated PDO sample. In the first case, a recrystallization process that
led to thicker
process lamellae
that led can be
to thicker deduced
lamellae canfrom the increase
be deduced fromintheSAXS peakin
increase intensity
SAXS peakand its shift towards
intensity and its
lower values of
shift towards the scattering
lower vector
values of the (q = [4π/λ]
scattering vectorsenθ). In thesenθ).
(q = [4π/λ] second In case, a shiftcase,
the second of the peakof
a shift once
the
samples were crystallized towards higher q values and a slight decrease on its intensity
peak once samples were crystallized towards higher q values and a slight decrease on its intensity was observed.
These features These
was observed. can befeatures
explained
canconsidering
be explained that a lamellarthat
considering insertion mechanism
a lamellar insertiontook place at took
mechanism low
temperatures together with a densification of the amorphous phase (i.e., a smaller
place at low temperatures together with a densification of the amorphous phase (i.e., a smaller difference between
the densitybetween
difference of amorphous and crystalline
the density phases).
of amorphous and crystalline phases).

Figure 9.
Figure 9. Variation
Variationofofintensity (Iq(Iq
intensity
2) (full symbols)
2 ) (full andand
symbols) scattering vector
scattering (q) (empty
vector symbols)
(q) (empty of SAXS
symbols) of
(small-angle X-ray scattering) peaks observed in the diffraction profiles taken during
SAXS (small-angle X-ray scattering) peaks observed in the diffraction profiles taken during heating heating
(10 ◦°C/min)
(10 C/min) at
at room
room temperature
temperature (red)
(red) andand during
during cooling
cooling (2
(2 °C/min)
◦ C/min)from
fromthethemelt
meltstate
state(blue).
(blue).

Characteristic lamellar parameters (i.e., long period, Lγ, amorphous layer thickness, la, and
Characteristic lamellar parameters (i.e., long period, Lγ , amorphous layer thickness, la , and
crystalline lamellar thickness, lc) and crystallinity (i.e., crystallinity within the lamellar stacks,
crystalline lamellar thickness, lc ) and crystallinity (i.e., crystallinity within the lamellar stacks,
XcSAXS = lc/Lγ) were determined by the normalized one-dimensional correlation function [35], γ(r):
Xc SAXS = lc /Lγ ) were determined by the normalized one-dimensional correlation function [35], γ(r):
∞ ∞
 
2 2
γ(r) = qw∞ I (q)cos(qr)dq /w∞ q I (q)dq (10)
γ(r ) 0= q2 I (q)cos(qr )dq/ 0q2 I (q)dq (10)
where I(q) is the intensity at each value0of the scattering vector. 0

whereSAXSI(q) isdata
the were collected
intensity at eachwithin
valueaof limited angular vector.
the scattering range, with application of the Vonk’s model
[39] and
SAXS data were collected within a limited angular range,high
Porod’s law to perform extrapolations to low and withq application
values. of the Vonk’s model [39]
and Porod’s law to perform extrapolations to low and high q values. from patterns acquired during
Figure 10 illustrates representative correlation functions obtained
the heating
Figure 10 of illustrates
granulatedrepresentative
PDO. Lamellar thickening
correlation was due
functions to the increase
obtained in crystalline
from patterns acquiredlamellar
during
thickness (i.e., from 6.0 to 7.7 nm) and amorphous layer thickness
the heating of granulated PDO. Lamellar thickening was due to the increase in crystalline (i.e., from 1.5 to 1.8 nm). In this
lamellar
thickness (i.e., from 6.0 to 7.7 nm) and amorphous layer thickness (i.e., from 1.5 to 1.8 nm). In (i.e.,
way, the reordering process led to a minimum increase of the crystallinity within lamellar stacks this
from 0.80 to 0.81). However, it should be pointed out that the correlation
way, the reordering process led to a minimum increase of the crystallinity within lamellar stacks (i.e., function of the sample
heated
from 0.80uptoto0.81).
102 °C (i.e., justitbefore
However, shouldthe befirst melting
pointed peakthe
out that observed in the
correlation DSC trace)
function of theexhibited more
sample heated
defined
up to 102peaks (see
◦ C (i.e., also
just the diffraction
before patterns
the first melting peakinobserved
Figure 10) inthat weretrace)
the DSC indicative of a more
exhibited high contrast
defined
between electronic densities of amorphous and crystalline phases. Basically,
peaks (see also the diffraction patterns in Figure 10) that were indicative of a high contrast between the amorphous phase
became less dense in agreement with the maximum value detected
electronic densities of amorphous and crystalline phases. Basically, the amorphous phase became less for la (i.e., 1.9 nm) and the

minimum
dense value of with
in agreement crystallinity
the maximum(i.e., 0.78).
value Figure 10 also
detected for lashows that
(i.e., 1.9 nm)morphological
and the minimum features were
value of
completely recovered after cooling the sample and specifically the correlation
crystallinity (i.e., 0.78). Figure 10 also shows that morphological features were completely recovered functions and the X-
ray diffraction
after cooling the pattern
sample were
andidentical
specificallyto those obtained from
the correlation the initial
functions and sample.
the X-ray diffraction pattern
were identical to those obtained from the initial sample.
Polymers 2016, 8, 351 13 of 18

Figure 10.
Figure 10. (a)
(a)SAXS
SAXSpatterns
patternsofofa agranulated
granulatedPDOPDO sample
sampletaken at 25
taken andand
at 25 102102
°C during a heating
◦ C during run
a heating
performed at 10 °C/min;
◦ (b) Change in the correlation function during the heating run.
run performed at 10 C/min; (b) Change in the correlation function during the heating run. For the For the sake
of completeness
sake the pattern
of completeness and correlation
the pattern function
and correlation obtained
function at room
obtained temperature
at room afterafter
temperature cooling (10
cooling
°C/min)
(10 ◦ a previously
C/min) molten
a previously sample
molten are also
sample shown.
are also shown.

The evolution of SAXS patterns of a PDO thread on heating (Figure 11) was very different
The evolution of SAXS patterns of a PDO thread on heating (Figure 11) was very different because
because they reflect the microstructure of the processed sample.
they reflect the microstructure of the processed sample.

(a) (b)

Figure 11. (a) SAXS patterns of a PDO suture taken at representative temperatures during a heating
Figure
run 11.◦ C/min.
at 10 (a) SAXSBlue
patterns of aarrows
and red PDO suture taken
indicate at representative
meridional temperatures
and equatorial during
reflections, a heating
respectively;
run at 10 °C/min. Blue and red arrows indicate meridional and equatorial reflections, respectively;
(b) Correlation function of diffraction patterns corresponding to: the initial suture, a suture heated (b)
◦
Correlation function of diffraction patterns corresponding to: the initial suture, ◦
a
(10 C/min) just before melting and a melt crystallized (cooling rate of 10 C/min) suture atsuture heated (10
°C/min)
room just before melting and a melt crystallized (cooling rate of 10 °C/min) suture at room
temperature.
temperature.

Some observations can be made: (a) The initial pattern was characterized by a meridional reflection
Some observations can be made: (a) The initial pattern was characterized by a meridional
that indicates a stacking of lamellae perpendicularly oriented to the fiber axis. Peaks observed in the
reflection that indicates a stacking of lamellae perpendicularly oriented to the fiber axis. Peaks
corresponding correlation function were highly prominent suggesting a tie molecular arrangement
observed in the corresponding correlation function were highly prominent suggesting a tie molecular
in the dense crystalline phase. However, lc and la values (6.1 nm and 1.5 nm) were close to those
arrangement in the dense crystalline phase. However, lc and la values (6.1 nm and 1.5 nm) were close
determined for the granulated sample; (b) As the temperature was increased the interlamellar reflection
to those determined for the granulated sample; (b) As the temperature was increased the
decreased in intensity while a new perpendicular reflection appeared and progressively increased
interlamellar reflection decreased in intensity while a new perpendicular reflection appeared and
in intensity. This new equatorial reflection could be associated with the existence of interfibrillar
progressively increased in intensity. This new equatorial reflection could be associated with the
amorphous domains that correspond to the regions placed on the lateral sides of lamellae. Note that
existence of interfibrillar amorphous domains that correspond to the regions placed on the lateral
molecular chains in these domains may have had a partial orientation at the beginning of the heating
sides of lamellae. Note that molecular chains in these domains may have had a partial orientation at
(i.e., the as processed thread) but became more randomly distributed as the temperature was increased.
the beginning of the heating (i.e., the as processed thread) but became more randomly distributed as
Therefore, a decrease in electronic density of the amorphous phase and an enhancement of the intensity
the temperature was increased. Therefore, a decrease in electronic density of the amorphous phase
and an enhancement of the intensity of the SAXS reflection were derived. The correlation function of
Polymers 2016, 8, 351 14 of 18
Polymers 2016, 8, 351 14 of 18

thethe
of interfibrillar reflection
SAXS reflection wereobserved
derived. inThethe pattern taken
correlation just before
function fusion gave lreflection
of the interfibrillar c and la values (i.e.,
observed
9.7the
in nmpattern
and 2.2 nm,just
taken respectively),
before fusionwhich
gavewere
lc andclearly
la values different from
(i.e., 9.7 those2.2determined
nm and for the
nm, respectively),
interlamellar
which reflection.
were clearly differentIn any
fromcase,
thoseXdetermined
cSAXS was again close
for the to 0.81; (c)reflection.
interlamellar The pattern andcase,
In any Xc SAXS
correlation
function
was againofclose
the sample
to 0.81; after being
(c) The cooled
pattern andtocorrelation
room temperature
function fromof thethe melt after
sample state being
were similar
cooled to to
those temperature
room determined from fromthetheinitial sample
melt state wereandsimilar
indicated a similar
to those lamellar from
determined organization.
the initialNevertheless,
sample and
a slight decrease
indicated a similar of lc wasorganization.
lamellar detected (i.e., 5.3 nm as opposed
Nevertheless, to 6.1 nm)
a slight decrease of lc as
waswell as a decrease
detected (i.e., 5.3 nmof
crystallinity
as opposed towithin
6.1 nm) lamellar
as well stacks (i.e., 0.78
as a decrease as opposed within
of crystallinity to 0.80), in agreement
lamellar with0.78
stacks (i.e., theaslack of an
opposed
annealing process for the melt crystallized sample.
to 0.80), in agreement with the lack of an annealing process for the melt crystallized sample.

3.5.
3.5. Evolution of Morphologic Parameters
Parameters during
during Melt
Melt Crystallization
Crystallization of
of Poly(p-dioxanone)
Poly(p-dioxanone) Samples
Samples
Figure 12 illustrates
illustrates the
the correlation
correlation functions
functionsfrom frompatterns
patternstaken
takenduring
duringthethe cooling
cooling runrun
(2
(2 ◦ C/min) of a melted PDO sample.
°C/min) of a melted PDO sample. It Itisisclear thatLLγ,γl,al,aand
clearthat , andlcldecreased
c decreasedprogressively
progressively and
and had similar
values
values (7.6,
(7.6, 1.5,
1.5, and
and 6.1
6.1 nm)
nm) toto those
those observed
observed for
for the
the initial
initial granulated
granulated PDO
PDO sample.
sample. The decrease
decrease inin
lamellar
lamellar thickness
thickness can can be
be due to the lower value expected when crystallization temperature decreases decreases
and
and also
alsototoa lamellar
a lamellar insertion mechanism
insertion mechanism(i.e.,(i.e.,
formation
formation of thinner lamellar
of thinner crystalscrystals
lamellar betweenbetween
loosely
stacked
loosely primary lamellae).lamellae).
stacked primary Similar results were
Similar obtained
results wereatobtained
differentatcooling rates,
different but it rates,
cooling is remarkable
but it is
that morphological
remarkable parameters were
that morphological practically
parameters wereidentical
practically at room temperature,
identical as shown in Figure
at room temperature, as shown12
for the sample
in Figure 12 forcooled at 10 ◦cooled
the sample C/min.at 10 °C/min.

Figure 12.
Figure 12. Correlation functions of
Correlation functions of patterns
patterns obtained
obtained at
at the
the indicated
indicated temperatures
temperatures during
during the
the cooling
cooling
run (2
◦ °C/min) from the melt state. The correlation function of the pattern obtained at
run (2 C/min) from the melt state. The correlation function of the pattern obtained at room temperatureroom
temperature after cooling
◦ at 10 °C/min is also shown for comparative
after cooling at 10 C/min is also shown for comparative purposes. purposes.

Note that crystallization took place at lower temperatures when the cooling rate was increased,
Note that crystallization took place at lower temperatures when the cooling rate was increased,
and consequently lower lamellar thicknesses should be expected. Therefore, the invariance observed
and consequently lower lamellar thicknesses should be expected. Therefore, the invariance observed
for lc suggests a counterbalance effect derived from the enhanced insertion mechanism (i.e., decrease
for lc suggests a counterbalance effect derived from the enhanced insertion mechanism (i.e., decrease
of lamellar thickness) when the cooling rate was decreased.
of lamellar thickness) when the cooling rate was decreased.

3.6. Changes
3.6. Changes in
in Microstructure
Microstructure of
of Poly(p-dioxanone)
Poly(p-dioxanone) Degraded
Degraded Samples
Samples during
during Heating
Heating
Insights on
Insights on the
the crystalline
crystalline microstructure
microstructureofofsutures
suturescan
canbebeobtained
obtainedfollowing
following thethe
evolution of
evolution
SAXS patterns of degraded samples during a subsequent heating process [25]. To this
of SAXS patterns of degraded samples during a subsequent heating process [25]. To this end, PDO end, PDO
sutures were
sutures were exposed
exposed toto hydrolytic
hydrolytic degradation
degradation media
media at
at pHs
pHs 77 and
and 11
11 and
and at
at 37
37 ◦°C for 36
C for 36 days
days to
to
analyze samples with clear differences in their degree of hydrolysis. Specifically, M w values of 117,700
analyze samples with clear differences in their degree of hydrolysis. Specifically, Mw values of 117,700
and 83,800 g/mol were determined after exposure to pHs 7 and 11, respectively. Weight losses were
close to 3% (pH 7) and 11% (pH 11). Micrographs in Figures 13 and 14 reveal the greater
Polymers 2016, 8, 351 15 of 18

Polymers 2016, 8, 351 15 of 18

Polymers 2016, 8, 351 15 of 18
and 83,800 g/mol were determined after exposure to pHs 7 and 11, respectively. Weight losses were
morphological
close to 3% (pH changes
7) and 11% occurred
(pH 11).under basic conditions
Micrographs and
in Figures 13 specifically
and 14 reveal thethegreater
appearance of deep
morphological
morphological changes occurred under basic conditions and specifically the appearance of deep
transversal
changes occurredcracks under
that ledbasic
to narrow disksand
conditions andspecifically
tortuous suture surfaces (Figure
the appearance of deep13).transversal
This degradation
cracks
transversal cracks that led to narrow disks and tortuous suture surfaces (Figure 13). This degradation
can
that be
ledinterpreted as a consequence
to narrow disks and tortuousofsuture
greater hydrolysis
surfaces of interlamellar
(Figure amorphous
13). This degradation canregions, which
be interpreted
can be interpreted as a consequence of greater hydrolysis of interlamellar amorphous regions, which
are
as adepleted
consequence and dissolved in the medium
of greater hydrolysis [11]. The morphology
of interlamellar amorphousofregions,
sutureswhich
exposedareto neutral and
depleted pH
are depleted and dissolved in the medium [11]. The morphology of sutures exposed to neutral pH
for a relatively
dissolved in the short
medium time is quite
[11]. different because
The morphology in this
of sutures case smoother
exposed to neutral surfaces and numerous
pH for a relatively short
for a relatively short time is quite different because in this case smoother surfaces and numerous
longitudinal cracks formed
time is quite different because(Figure 14).smoother
in this case In fact, itsurfaces
has been
andreported
numerousthat hydrated,cracks
longitudinal interfibrillar
formed
longitudinal cracks formed (Figure 14). In fact, it has been reported that hydrated, interfibrillar
amorphous
(Figure 14). regions swell
In fact, it has more easily than
been reported interlamellar
that amorphous regions
hydrated, interfibrillar amorphous [12]. regions
Note that the latter
swell more
amorphous regions swell more easily than interlamellar amorphous regions [12]. Note that the latter
are constituted
easily by tie chains,
than interlamellar which connect
amorphous regionsthe lamellae
[12]. in each
Note that fibril are
the latter whereas fewer tie
constituted by chains are
tie chains,
are constituted by tie chains, which connect the lamellae in each fibril whereas fewer tie chains are
expected
which to connect
connect adjacentin
the lamellae fibrils.
each Thus, water diffusion
fibril whereas and
fewer tie formation
chains of longitudinal
are expected cracks
to connect seem
adjacent
expected to connect adjacent fibrils. Thus, water diffusion and formation of longitudinal cracks seem
to be favored.
fibrils. Thus, water diffusion and formation of longitudinal cracks seem to be favored.
to be favored.

Figure 13.
Figure 13. SAXS patterns
SAXSpatterns
13. SAXS taken
patternstaken at
takenat representative
atrepresentative temperatures of
representative temperatures
temperatures of 25
25 ◦°C (a); 102
C (a);
(a); 102 ◦°C (b)
C (b) and
(b) and 107
and 107 °C
◦ C (c)
107 °C (c)
Figure °C 102 °C (c)
during a heating
during aa heating run
heating run at
run at 10
at 10 °C/min
◦
10 °C/min of a PDO
C/minofofaaPDO suture
PDOsuture previously
suturepreviously degraded
previouslydegraded
degradedin in a
inaapHpH 11
pH11 hydrolytic
11hydrolytic medium
hydrolyticmedium
medium
during
for 36 days.
for 36
36 days. The pattern
The pattern obtained
pattern obtained at room
obtained at
at room temperature
temperature after
after cooling (10(10 ◦°C/min) and the optical
for days. The room temperature after cooling C/min) and
cooling (10 °C/min) and the
the optical
optical
micrograph
micrograph of of the
of the degraded
the degraded suture
degraded suture are
suture are shown
are shown in
shown in (d,e),
in (d,e), respectively.
(d,e), respectively.
respectively.
micrograph

Figure 14. SAXS patterns taken at representative temperatures of 25 °C (a); 102 °C (b) and 112 °C (c)
Figure 14. SAXS patterns taken at representative temperatures of 25 °C (a); 102 °C
◦ (b) and 112 °C
◦ C(c)
Figure a14.
during SAXS run
heating patterns taken atofrepresentative
at 10 °C/min a PDO suturetemperatures of 25 ◦ C in
previously degraded (a);a 102
pH 7 C (b) and 112
hydrolytic medium (c)
during a heating run at 10 °C/min
◦ of a PDO suture previously degraded in a pH 7 hydrolytic medium
during a heating run at 10 C/min of a PDO suture previously degraded in a pH 7
for 36 days. The pattern obtained at room temperature after cooling (10 °C/min) and the opticalhydrolytic medium
for
for 36
36 days. The
days. of
The pattern obtained at
at room temperature after cooling (10
(10 ◦°C/min)
C/min) and
and the
the optical
micrograph thepattern
degraded obtained
suture are room
showntemperature after cooling
in (d,e), respectively. optical
micrograph of the degraded suture are shown in (d,e), respectively.
micrograph of the degraded suture are shown in (d,e), respectively.
SAXS patterns of degraded samples showed a slight decrease of meridional interlamellar
SAXS patterns of degraded samples showed a slight decrease of meridional interlamellar
spacing,
SAXSLBpatterns
, becauseofofdegraded
the moresamples
compactshowed
structure achieved
a slight afterof
decrease scission of chains
meridional belonging
interlamellar to the
spacing,
spacing, LB, because of the more compact structure achieved after scission of chains belonging to the
amorphous
L B , because regions.
of the Therefore,
more compact the initial
structure spacing
achieved of 9.6
after nm decreased
scission of to
chains 9.2 and
belonging9.3 nm
to after
the 36 days
amorphous
amorphous regions. Therefore, the initial spacing of 9.6 nm decreased to 9.2 and 9.3 nm after 36 days
of degradation
regions. in basic
Therefore, and neutral
the initial spacing pH,
of 9.6respectively.
nm decreasedAnalysis of 9.3
to 9.2 and correlation
nm after functions
36 days of (not shown)
degradation
of degradation in basic and neutral pH, respectively. Analysis of correlation functions (not shown)
indicated that
in basic and lc andpH,
neutral la values decreased
respectively. fromof6.1
Analysis to 1.5 nm,
correlation respectively,
functions to 5.3 indicated
(not shown) and 1.3 nmthatfor the
l and
indicated that lc and la values decreased from 6.1 to 1.5 nm, respectively, to 5.3 and 1.3 nm forc the
basic pH. Logically, SAXS crystallinity increased during degradation since the main change occurred
basic pH. Logically, SAXS crystallinity increased during degradation since the main change occurred
in the amorphous layer.
in the amorphous layer.
Polymers 2016, 8, 351 16 of 18

la values decreased from 6.1 to 1.5 nm, respectively, to 5.3 and 1.3 nm for the basic pH. Logically, SAXS
crystallinity increased during degradation since the main change occurred in the amorphous layer.
Lamellar crystals in degraded samples were able to recrystallize and even reorient during
subsequent heating runs in an easier way than that observed for the initial suture. The increased
freedom resulting from scission of tie chains belonging to interfibrillar and interlamellar amorphous
regions should play a fundamental role. Diffraction patterns during subsequent heating showed clear
differences between highly and scarcely degraded samples. Thus, in the first case, an increase in
lamellar thickness, together with the appearance of an intense equatorial reflection associated with
the interfibrillar spacing previous to the disappearance of the meridional reflection, was observed.
Specifically, LB values of 13.2 and 11.7 nm were determined at temperatures close to fusion (i.e.,
102 ◦ C). This behavior was similar to that observed for the initial suture taking into account the
differences in spacings and intensities of reflections. Logically, greater spacings and intensities were
detected for degraded samples as a consequence, in the first case, of an enhanced reorganization
when tie interconnecting chains were cleaved and, in the second case, of a higher contrast between
amorphous/crystalline regions.
Figure 14 reveals a different evolution when samples were degraded in the pH 7 medium because
the great thickening of lamellae was hindered due to a still compact chain packing (note that weight
loss was minimal). Therefore, the thickened lamellae (LB = 11.4 nm) tilted with respect to the fiber axis
gave rise to a second reflection. The breadth of reorganized crystals was also clearly increased, as could
be deduced from the observed spot like reflections. Finally, Figures 13 and 14 show the reversibility of
the thermal process for degraded samples since the diffraction patterns obtained after cooling to room
temperature are identical to those obtained from the initial sample before any thermal treatment.

4. Conclusions
PDO showed complex melting and crystallization peaks because of a typical lamellar thickening
process and the existence of different nucleation mechanisms, respectively. Calorimetric analysis
of non-isothermal crystallization showed an increase of the Avrami exponent at low cooling rates,
which could be associated with a homogeneous nucleation process instead of the instantaneous
nucleation observed at high rates. Isoconversional analyses from non-isothermal calorimetric data
revealed the existence of two crystallization regimes, demonstrating the suitability of this methodology.
These regimes were well characterized by optical microscopy observations, and the non-isothermal
crystallization results were in relatively good agreement with those previously reported from
isothermal studies.
Real time SAXS profiles taken during heating and cooling processes showed the occurrence
of a lamellar reordering process and a lamellar insertion mechanism that led to an increase and a
decrease in lamellar thickness, respectively. SAXS patterns taken during heating of samples degraded
under neutral and basic pH had a different evolution that revealed the existence of interlamellar and
interfibrillar amorphous domains.

Acknowledgments: The authors acknowledge support from MINECO and FEDER (MAT2015-69547-R), and the
Generalitat de Catalunya (2014SGR188). Diffraction experiments were performed at NCD beamline at ALBA
Synchrotron with the collaboration of ALBA staff.
Author Contributions: All authors made contributions to the development of this manuscript and approved
the final manuscript. Yolanda Márquez contributed to the experimental work, Lourdes Franco to calorimetric
analyses, Juan Carlos Martínez assisted with synchrotron analyses, Pau Turon contributed to the concept and
revision of the manuscript, and Jordi Puiggalí contributed to the concept, analyses of data, and design and revision
of the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
Polymers 2016, 8, 351 17 of 18

References
1. Yang, K.J.; Wang, X.L.; Wang, Y.Z. Poly(p-dioxanone) and its copolymers. J. Macromol. Sci. Part C 2002, 42,
373–398. [CrossRef]
2. Li, X.W.; Xiao, J.; Li, X.Y.; Xiong, C.D.; Deng, X.M. Preparation of biodegradable PLA-PEG copolymer
microspheres with micron sizes. Polym. Mater. Sci. Eng. Div. 1998, 14, 20–22.
3. Sabino, M.A.; Feijoo, J.L.; Müller, A.J. Crystallisation and morphology poly(p-dioxanone).
Macromol. Chem. Phys. 2000, 201, 2687–2698. [CrossRef]
4. Andjelic, S.; Jamiolkowski, D.; McDivitt, J.; Fisher, J.; Zhou, J.; Vetrecin, R. Crystallization study on absorbable
poly(p-dioxanone) polymers by differential scanning calorimetry. J. Appl. Polym. Sci. 2001, 79, 742–759.
[CrossRef]
5. Andjelic, D.; Jamiolkowski, D.; McDivitt, J.; Fisher, J.; Zhou, J. Spherulitic growth rates and morphology of
absorbable poly(p-dioxanone) homopolymer and its copolymer by hot-stage optical microscopy. J. Polym.
Sci. Part B 2001, 39, 3073–3089. [CrossRef]
6. Sabino, M.A.; Ronca, G.; Müller, A.J. Heterogeneous nucleation and self-nucleation of poly(p-dioxanone).
J. Mater. Sci. 2000, 35, 5071–5084. [CrossRef]
7. Pezzin, A.P.T.; Alberda van Ekenstein, G.O.R.; Duek, E.A.R. Melt behavior, crystallinity and morphology of
poly(p-dioxanone). Polymer 2001, 42, 8303–8306. [CrossRef]
8. Yang, K.K.; Wang, X.L.; Wang, Y.Z.; Huang, H.X. Effects on molecular weights of bioabsorbable
poly(p-dioxanone) on its crystallization behavior. J. Appl. Polym. Sci. 2006, 100, 2331–2335. [CrossRef]
9. Andjelic, S.; Jamiolkowski, D.; McDivitt, J.; Fisher, J.; Zhou, J.; Wang, Z.; Hsiao, B.S. Time-resolved
crystallization study of absorbable polymers by synchrotron small-angle X-ray scattering. J. Polym. Sci.
Part B 2001, 39, 153–167. [CrossRef]
10. Sabino, M.A.; Feijoo, J.L.; Müller, A.J. Crystallisation and morphology of neat and degraded
poly(p-dioxanone). Polym. Degrad. Stab. 2001, 73, 541–547. [CrossRef]
11. Sabino, M.A.; Albuerne, J.; Müller, A.J.; Brisson, J.; Prud’homme, R.E. Influence of in vitro hydrolytic
degradation on the morphology and crystallization behavior of poly(p-dioxanone). Biomacromolecules 2004,
5, 358–370. [CrossRef] [PubMed]
12. Ooi, C.P.; Cameron, R.E. The hydrolytic degradation of polydioxanone (PDS II) sutures. Part II:
Micromechanisms of deformation. J. Biomed. Mater. Res. 2002, 63, 291–298.
13. Furuhashi, Y.; Nakayama, A.; Monno, T.; Kawahara, Y.; Yamane, H.; Kimura, Y.; Iwata, T. X-ray and electron
diffraction study of poly(p-dioxanone). Macromol. Rapid Commun. 2004, 25, 1943–1947. [CrossRef]
14. Gestí, S.; Lotz, B.; Casas, M.T.; Alemán, C.; Puiggalí, J. Morphology and structure of poly(p-dioxanone).
Eur. Polym. J. 2007, 43, 4662–4674. [CrossRef]
15. Spina, R.; Spekowius, M.; Dahlmann, R.; Hopmann, C. Analysis of polymer crystallization and residual
stresses in injection molded parts. Int. J. Precis. Eng. Manuf. 2014, 15, 89–96. [CrossRef]
16. Guevara-Morales, A.; Figueroa-López, U. Residual stresses in injection molded products. J. Mater. Sci. 2014,
49, 4399–4415. [CrossRef]
17. Ozawa, T. Kinetics of non-isothermal crystallization. Polymer 1971, 12, 150–158. [CrossRef]
18. Cazé, C.; Devaux, E.; Crespy, A.; Cavrot, J.P. A new method to determine the Avrami exponent by DSC
studies of non-isothermal crystallization from the molten state. Polymer 1997, 38, 497–502. [CrossRef]
19. Zeng, J.B.; Srinivansan, M.; Li, S.L.; Narayan, R.; Wang, Y.Z. Non isothermal and isothermal cold
crystallization behaviors of biodegradable poly(p-dioxanone). Ind. Eng. Chem. Res. 2011, 50, 4471–4477.
[CrossRef]
20. Avrami, M. Kinetics of phase change. I General Theory. J. Chem. Phys. 1939, 7, 1103–1112. [CrossRef]
21. Cebe, P. Non-isothermal crystallization of poly(etheretherketone) aromatic polymer composite.
Polym. Compos. 1988, 9, 271–279. [CrossRef]
22. Tobin, M.C. Theory of phase transition kinetics with growth site impingement. I. Homogeneous nucleation.
J. Polym. Sci. Part B Polym. Phys. 1974, 12, 399–406. [CrossRef]
23. Zhou, Z.X.; Wang, X.L.; Wang, Y.Z.; Yang, K.K.; Chen, S.C.; Wu, G.; Li, J. Thermal properties and
non-isothermal crystallization behavior of biodegradable poly(p-dioxanone)/poly(vinyl alcohol) blends.
Polym. Int. 2006, 55, 383–390. [CrossRef]
Polymers 2016, 8, 351 18 of 18

24. Zhang, X.; Bai, W.; Chen, D.; Xiong, C.; Pang, X. Nonisothermal crystallization behaviour of
poly(p-dioxanone) and poly(L-lactic acid) blends. Bull. Mater. Sci. 2015, 38, 517–523. [CrossRef]
25. Márquez, Y.; Martínez, J.C.; Turon, P.; Franco, L.; Puiggalí, J. Influence of pH on morphology and structure
during hydrolytic degradation of the segmented GL-b-[GL-co-TMC-co-CL]-b-GL copolymer. Fibers 2016, 3,
348–372. [CrossRef]
26. Schultz, J.M. Polymer Crystallization the Development of Crystalline Order in Thermoplastic Polymers; ACS/Oxford
University Press: Oxford, UK, 2001.
27. López, L.C.; Wilkes, G.L. Non-isothermal crystallization kinetics of poly(p-phenylene suphide). Polymer
1989, 30, 882–887. [CrossRef]
28. Privalko, V.P.; Kawai, T.; Lipatov, Y.S. Crystallization of filled Nylon 6. III Non-isothermal crystallization.
Colloid Polym. Sci. 1979, 257, 1042–1048. [CrossRef]
29. Liu, T.X.; Mo, Z.S.; Wang, S.G.; Zhang, H.F. Nonisothermal melt and cold crystallization kinetics of poly(aryl
ether ether ketone ketone). Polym. Eng. Sci. 1997, 37, 568–575. [CrossRef]
30. Friedman, H.J. Kinetics of thermal degradation of char-forming plastics from thermogravimetry. J. Polym.
Sci. Part C 1964, 6, 183–195. [CrossRef]
31. Vyazovkin, S.; Dranca, I. Isoconversional analysis of combined melt and glass crystallization data.
Macromol. Chem. Phys. 2006, 207, 20–25. [CrossRef]
32. Vyazovkin, S.; Sbirrazzouli, N. Isoconversional approach to evaluating the Hoffman-Lauritzen parameters
(U* and Kg) from overall rates of nonisothermal melt crystallization. Macromol. Rapid Commun. 2004, 25,
733–738. [CrossRef]
33. Vyazovkin, S.; Stone, J.; Sbirrazzouli, N. Hoffman-Lauritzen parameters for non-isothermal crystallization
of poly(ethylene terephthalate) and poly(ethylene oxide) melts. J. Therm. Anal. Calorim. 2005, 80, 177–180.
[CrossRef]
34. Suzuki, T.; Kovacs, A.J. Temperature dependence of spherulitic growth rate of isotactic polystyrene. A critical
comparison with the kinetic theory of surface nucleation. Polym. J. 1970, 1, 82–100. [CrossRef]
35. Vonk, C.G.; Kortleve, G. X-ray small-angle scattering of bulk polyethylene. Kolloid Z. Z. Polym. 1967, 220,
19–24. [CrossRef]
36. Di Lorenzo, M.L.; Cimmino, S.; Silvestre, C. Nonisothermal crystallization of isotactic polypropylene blended
with poly(α-pinene). 2. Growth rates. Macromolecules 2000, 33, 3828–3832. [CrossRef]
37. Di Lorenzo, L.M. Determination of spherulite growth rates of poly(L-lactic acid) using combined isothermal
and non-isothermal procedures. Polymer 2001, 42, 9441–9446. [CrossRef]
38. Lauritzen, J.I.; Hoffman, J.D. Extension of theory of growth of chain-folded polymer crystals to large
undercoolings. J. Appl. Phys. 1973, 44, 4340–4352. [CrossRef]
39. Vonk, C.G. A general computer program for the processing of small-angle X-ray scattering data.
J. Appl. Crystallogr. 1975, 8, 340–341. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).