The effect of nanocomposite synthesis and the drying procedure of graphene oxide dispersion on the polycaprolactone/graphene oxide nanocomposite properties

Abstract

In this work, we compare how the properties of polycaprolactone (PCLO) nanocomposites are affected by the method of nanocomposite synthesis and the differences in graphene oxide (GO) properties. PCLO nanocomposites with freeze-dried or hot-dried graphene oxide (FGO/HGO) were synthesized by in situ polymerization and melt mixing. The PCLO molar masses remained constant during melt mixing, but they were significantly reduced with increasing amount of GOs during in situ polymerization. Despite this fact, the in situ polymerized nanocomposites showed enhancement in Young modulus up to 45%, compared to only 10% modulus increase of melt-mixed nanocomposites. This was attributed to the higher crystallinity and the higher level of nanofiller dispersion and exfoliation in in situ polymerized nanocomposites. When comparing the effect of FGO and HGO on nanocomposites properties, the improved dispersion and ameliorated mechanical properties were observed for the former one.

Keywords

Polycaprolactone graphene oxide nanocomposites synthesis properties

Introduction

Polycaprolactone (PCLO) is an aliphatic polyester, whose biodegradability made it popular for biomedical applications.¹ Specifically, many PCLO-based products were commercialized for tissue engineering such as nerve guidance conduits, the monofilament sutures or the root canal filling materials.^2–4 However, further PCLO application potential is limited by its poor mechanical performance. This deficiency can be suppressed by addition of various nanofillers, which opens new possibilities in biomedicine and beyond. Recently, graphene-based nanofillers were reported to improve not only the mechanical properties of various polymers⁵ but also several biomedical properties of PCLO.^6–8 A combination of good biomedical and mechanical properties is promising for future medical applications. However, it is important to balance the individual properties, which are heavily influenced by synthesis procedure.

Two commonly used procedures to synthesize PCLO nanocomposites are in situ polymerization and melt mixing. An advantage of in situ polymerization is a possibility of using nanofiller functional groups to (co)initiate/catalyze polymerization, possibly leading to covalent attachment of the nanofiller to polymer matrix. This ability was reported for untreated graphene oxide (GO),^9,10 as well as for a treated one.^11,12 The covalent attachment then helped to exfoliate the GO layers in the polymer matrix and improved the mechanical properties of PCLO nanocomposites. Nanocomposites prepared by in situ polymerization were also processed to scaffolds with good biocompatibility and honeycomb-patterned films with enhanced electrical conductivity. Contrary to in situ polymerization, melt mixing is usually faster and provides nanocomposites with good biomedical and mechanical properties as well, that are often further processed by 3D printing. Nevertheless, the results obtained from different authors vary. For example, Wang et al. prepared the nanocomposite scaffolds with enhanced compression modulus and biocompatibility,¹³ whereas Scenini et al. reported the graphene nanocomposites with only a negligible change in mechanical properties.¹⁴ To conclude this section, we must emphasize that despite the large amount of research, the direct comparison of in situ polymerization and melt mixing is still missing in the literature.

The major aim of this work is to compare the properties of PCLO/GO nanocomposites prepared by in situ polymerization and melt mixing. The minor aim is to describe the influence of varying GO characteristics on the nanocomposite properties. The GOs were characterized by the transmission electron micrographs, sorption analysis and X-ray diffraction. The nanocomposite analysis focused on the relationship between mechanical and rheological properties and the level of nanofiller dispersion and exfoliation in polymer matrix.

Experimental

Materials

Graphite (325 mesh, GK Graphite Týn CR 12 995; Czech Republic). Sulfuric acid (96%, Penta s. r.o.; Czech Republic). Sodium nitrate, potassium permanganate (Lachema; Czech Republic). Hydrogen peroxide (30%, Penta s. r.o.; Czech Republic). ε-Caprolactone (97%, Sigma-Aldrich) was dried over powdered calcium hydride (Sigma Aldrich) and distilled under reduced pressure (40 Pa), water content determined by Karl-Fischer titration was less than 30 ppm. Stannous octoate (95%, Sigma-Aldrich). Toluene and methanol (p.a., Sigma-Aldrich).

Synthesis of the graphene oxide

The GO was synthesized using the modified Hummers method according to the work of Rajaura et al.¹⁵ The main steps of GO synthesis are discussed in Electronic Supporting Information (ESI). The obtained dispersion was then dried by freeze-drying or hot-drying. During the freeze-drying, the water was sublimed from the frozen sample (−70°C) at reduced pressure (50 Pa). In case of hot drying, Petri dishes with dispersion were placed in a vacuum oven (60°C) and the water was evaporated at reduced pressure (800 Pa). Both GOs were ground in the knife mill. The freeze-dried GO will be further denoted as FGO, whereas the hot-dried GO as HGO.

Synthesis of the nanocomposites

The PCLO nanocomposites were prepared by in situ polymerization and melt mixing.

All steps of the in situ polymerization were carried out under argon atmosphere. Firstly, the GOs were dispersed in caprolactone by ultrasonication to obtain a homogenous dispersion. The ultrasonication time was 30 min (FGO) or 60–120 min (HGO, the time was increased with the increasing concentration of nanofiller). Afterwards, the solution of stannous octoate (polymerization catalyst) in toluene was introduced to the GO/caprolactone dispersion; the final concentration of stannous octoate in polymerization feed was 0.5 mol%. Toluene was then removed at reduced pressure (50 Pa) and polymerization was carried out for 4 h at 120°C. Finally, the toluene residues were removed from nanocomposites by annealing at 50°C and reduced pressure (800 Pa) to constant weight.

The conditions for the synthesis of PCLO matrix used for the melt mixing were analogous to the in situ polymerization. Stannous octoate (0.5 mol%) was dispersed in caprolactone, the mixture was transferred to an open teflon mould with internal dimensions 160 × 140 × 4 mm and the polymerization was carried out for 4 h at 120°C. After demoulding, PCLO was ground to the small chips. The chips were premixed with the GOs by manual stirring and were dosed to twin-screw microextruder (HAAKE Minilab II) with co-rotating conical screws. The melt mixing was carried out for 10 min at 80°C, screw speed was set to 100 rpm for mixing and to 50 rpm for the extrusion of nanocomposites.

Characterization

The elemental composition of GOs was determined by combustion analysis using a Vario EL Cube (Elementar, Germany) and by X-ray photoelectron spectroscopy (XPS) on an ESCAProbeP (Omicron Nanotechnology Ltd.; Germany). The ESCAProbeP used an aluminium anode with energy 1486.7 eV as a source of monochromatic light. The data were analysed with CASA XPS software using Shirley background subtraction.

The interlayer distance of GO layers and the level of GO dispersion in polymer matrix were measured by X-ray diffraction (XRD) on an X´Pert³ Powder diffractometer (PANanalytical; United Kingdom) with Bragg-Brentano parafocusing geometry and a CuKα tube providing wavelength of λ = 0.154 nm. Peak profiles were obtained using HighScore 4.0 software (PANAlytical). The calculations of GO interlayer distance were based on Bragg equation (1) (d – interlayer distance, $θ$ – diffraction angle, n – positive integer, $λ$ - wavelength)

2 d s i n θ = n λ

(1)

The surface area of graphite and GOs was measured using a Coulter SA 3100 sorption analyzer, (Beckman Coulter). Before the sorption experiments, samples were degassed for 4 h at 95°C under high vacuum. Detector cooled by liquid nitrogen (−196°C) was used to evaluate the results using Kelvin and Brunauer, Emmett and Teller equations, respectively.

To obtain transmission electron microscopy (TEM) images, the powder nanofillers were ground in an agate mortar and dispersed in isopropanol. Then, 7 μl of the dispersion was dripped on a copper grid (Ted Pella Inc. mesh 300, lacey carbon) and isopropanol was evaporated. The measurements were carried out on an EFTEM 2200 FS (Jeol, Ltd.; Japan) operating at an accelerating voltage of 200 kV.

The polymer yield of nanocomposites was determined gravimetrically after extraction of low-molecular compounds, performed three times for 30 min in methanol (50 mL per 0.5 g of sample). To calculate the polymer yield, the equation (2) was used ( $y_{w}$ – polymer yield, $m_{0}$ – mass of the sample before extraction, – mass of the sample after extraction, $w_{G O}$ – mass fraction of GO in polymerization mixture)¹⁶

y_{w} = \frac{m_{1} - w_{G O} \cdot m_{0}}{m_{0} - w_{G O} \cdot m_{0}}

(2)

The average values of molar masses were determined using a chromatographic system Breeze (Waters; USA) with a refractive index detector and multi-angle laser light scattering detector miniDawn TREOS (Wyatt; USA). The separation was performed on a PSS Lux LIN M 5 μm column (7.8 × 300 mm) at 35°C in tetrahydrofuran at an elution rate of 1 mL/min using polystyrene standards for calibration.

To prepare the specimens for tensile tests and dynamic mechanical analysis (DMA), the nanocomposites were ground to chips, which were hot pressed in a mould with dimensions 60 × 50 × 0.7 mm using this procedure: (I) 5 min preheating, (II) 5 min hot-pressing at 75°C (degassing after 1 min), (III) 5 min cooling. Finally, dog bone specimens were cut from the pressed foil (distance between shoulders – 20 mm, width – 4 mm, thickness – 0.7 mm) and stored in a desiccator before the measurement. Dynamic mechanical behaviour of nanocomposites was studied using DMA DMX04T instrument (RMI; Czech Republic). The measurements were realized in bending mode in the temperature range: –110–50°C at a heating rate of 2 °C/min, frequency 1 Hz and force amplitude 500 mN. The tensile tests were carried out on an Instron 3365 (Instron; USA) at a rate of 1 mm/min until the proportional elongation exceeded 1%, then the rate was increased to 50 mm/min.

The crystalline properties of nanocomposites were obtained using differential scanning calorimeter DSC Q100 (TA Instruments; USA). The measurements were carried out in a 0–100°C range (heating – cooling – heating) at rate of 10 °C/min. To calculate the crystalline content of nanocomposites, the equation (3) was used¹⁷ ( $x_{c}$ – crystalline content, $Δ H_{m s}$ – the fusion enthalpy of the sample, $Δ H_{m_{100}}$ – the fusion enthalpy of 100% crystalline PCLO = 139 J/g, $w_{P C L O}$ – weight ratio of PCLO in nanocomposite)

x_{c} = \frac{Δ H_{m s}}{Δ H_{m_{100}}} \cdot w_{P C L O} 100

(3)

The optical micrographs of nanocomposites were obtained using a Leitz DMRXE microscope equipped with objective HC PL Fluotar (10x/0.3 PH1). Nanocomposite chips on microscopy slides were placed on Linkamcell Scientific LTS 350 plate that was heated to 80°C.

Scanning electron microscopy was carried out on TESCAN Vega 3 LMU apparatus in Univac mode.

Results and discussion

The synthesis of graphene oxide

Elemental analysis and XPS show that the choice of the drying procedure did not considerably influence the chemical composition of GOs (Table 1, Figure 1, Table ESI 1). Despite the thorough washing of both GOs with deionized water, they still contained low amount of sulfur originating from sulfuric acid used in the Hummers method. The presence of sulfur can be explained by formation of species covalently bound to the GO functional groups, which are not easily removable.¹⁸

Table 1.

The chemical composition of freeze dried graphene oxide and hot dried graphene oxide.

Sample	Elemental analysis				X-ray photoelectron spectroscopy
Sample	C (%)	H (%)	S (%)	O ^a (%)	C (%)	O (%)	S (%)	-C-C-; -C=C-(%)	-C-O- (%)	-C=O- (%)	-O-C=O- (%)
Graphite	99.2	0	0	0.8	93.3	6.7	0	90.5	9.5	0	0
FGO	48.2	2.3	1.8	47.7	69.3	29.4	1.3	63.0	26.8	6.3	3.9
HGO	49.7	2.3	1.6	46.4	67.4	31.7	0.9	37.9	37.1	20.1	4.9

^aOxygen content was calculated with the assumption that material does not contain any other element.

Figure 1.

X-ray photoelectron spectroscopy survey spectra (lefthand side) and deconvoluted C1s spectra (righthand side) of FGO (top) and HGO (bottom).

X-ray diffraction revealed almost 2.5-fold increase in interlayer distance of both GOs, which is beneficial for the dispersion in monomer or polymer matrix (Table 2). The intensity of HGO diffraction pattern increased almost seven times in contrary to FGO (Figure 2). This increase can be explained by a higher number of stacked HGO sheets, which suggests a stronger tendency of HGO to aggregate.¹⁹

Table 2.

The interlayer distances and surface areas of freeze dried graphene oxide and hot dried graphene oxide.

Sample	X-ray diffraction	Sorption analysis
Sample	Interlayer distance (nm)	Surface area (m²/g)
Graphite	0.34	6.068
FGO	0.81	6.980
HGO	0.80	0.525

Figure 2.

The X-ray diffraction patterns of FGO and HGO (top), the comparison of bulk densities of FGO and HGO (bottom). FGO: Freeze dried graphene oxide; HGO: Hot dried graphene oxide.

Sorption analysis proved that freeze-drying slightly increased the surface area of the product, whereas hot-drying decreased the surface area by an order of magnitude (Table 2). It is noteworthy that while the GOs significantly differ in their surface areas, their interlayer distances are similar. This discrepancy was resolved via TEM that indicate the presence of single sheets and multiple layers in both GOs (Figure 3). However, FGO micrographs show mostly flat surfaces with low level of folding, while HGO sheets exhibit much more folding and wrinkling. Such distortion of HGO can strongly decrease the accessibility of its surface to nitrogen atoms used in sorption analysis which can explain the difference in surface areas. We also believe that the difference in surface area is responsible for the extremely high bulk density of FGO compared to HGO (Figure 2).

Figure 3.

The transmission electron microscopy images. (a) FGO single sheet, (b) FGO multiple sheets, (c) HGO single sheet and (d) HGO multiple sheets. FGO: Freeze dried graphene oxide; HGO: Hot dried graphene oxide.

The synthesis of nanocomposites

In situ polymerization afforded nanocomposites containing both GOs in amount up to 2.5 wt%. Melt mixing was successful with FGO, but the preliminary experiment with 1 wt% of HGO resulted in its poor dispersion in the polymer matrix (Figure ESI 1). We ascribe this deficiency to the low surface area of HGO, which makes it hardly accessible for polymer chains and, therefore, unable to be well dispersed or even exfoliated. For this reason, melt mixing was used only for the synthesis of nanocomposites with FGO.

Molar masses of nanocomposites

While the increasing GO concentration had no significant effect on molar masses of melt mixed nanocomposites, it strongly decreased the molar masses of nanocomposites during in situ polymerization (Table 3). A complete explanation of this decrease is provided in ESI. Shortly, GO contains a variety of nucleophilic compounds that could initiate the polymerization of caprolactone, resulting in high amount of short chains while keeping the high polymer yield. We also note the difference in molar masses between the in situ polymerized unfilled sample and polymer matrix prepared for melt mixing. The molar masses of melt mixed sample are three times lower, although the conditions for polymerization were similar. The most likely reason of this behaviour is the absence of inert atmosphere, because the synthesis of the matrix for melt mixing was carried out in an open mould. It is also possible that the polymerization catalyst underwent degradation resulting in its reduced activity.

Table 3.

Polymer yield, molar masses and dispersity of polymer matrix from polycaprolactone/graphene oxide nanocomposites.

In situ polymerization					Melt mixing
GO type and content (wt%)^a	Polymer yield (%)	${\bar{M}}_{w}$ ^(b) (g/mol)	${\bar{M}}_{n}$ ^(c) (g/mol)	Ð^(d)	${\bar{M}}_{w}$ ^(b) (g/mol)	${\bar{M}}_{n}$ ^(c) (g/mol)	Ð^(d)	GO type and content (wt%)^a
0% GO	98.8	164 000	95 100	1.7	55 400^(e)	35 000^(e)	1.6^(e)	0% GO
0% GO	98.8	164 000	95 100	1.7	59 500^(f)	41 700^(f)	1.4^(f)	0% GO
0.1% HGO	99.1	83 000	51 800	1.6	NA ^g	NA ^g	NA ^g	0.1% HGO
0.5% HGO	98.6	46 800	28 100	1.7	NA ^g	NA ^g	NA ^g	0.5% HGO
1% HGO	98.5	39 300	24 300	1.6	NA ^g	NA ^g	NA ^g	1% HGO
2.5% HGO	96.0	24 600	15 400	1.6	NA ^g	NA ^g	NA ^g	2.5% HGO
0.1% FGO	98.7	70 500	42 800	1.6	57 000	35 500	1.6	0.1% FGO
0.5% FGO	98.6	61 500	38 600	1.6	57 700	37 200	1.6	0.5% FGO
1% FGO	97.8	35 100	22 500	1.6	57 700	36 900	1.6	1% FGO
2.5% FGO	96.5	18 900	12 700	1.5	57 600	36 500	1.5	2.5% FGO

^acontent in polymerization mixture (in situ) or mixing batch (melt mixing).

^bweight average molecular weight obtained from SEC.

^cnumber average molecular weight obtained from SEC.

^ddispersity $Ð = {\bar{M}}_{w} / {\bar{M}}_{n}$ .

^emolar masses of virgin PCLO before melt mixing.

molar masses of virgin PCLO after melt mixing.

^gNA = material not available.

Thermal properties of nanocomposites

First, we assessed the influence of synthesis methods on crystalline properties of nanocomposites measured by DSC (Table 4, Figure 4, Figure ESI 2 and 3). All in situ polymerized samples exhibited an increasing crystallinity with increasing content of both GOs. However, crystallinity of melt mixed samples remained almost unchanged and well below the values for in situ polymerized samples at higher GO concentrations. Crystallization temperatures of nanocomposites followed the trends in their crystallinity. We suggest that the stronger effect of GO concentration on the crystalline properties of in situ polymerized nanocomposites is caused by a higher number of polymer-nanofiller interactions, which results in better mechanical properties. The number of interactions is closely related to the degree of nanofiller dispersion and exfoliation in matrix, which will be further discussed below.

Table 4.

Thermal properties of nanocomposites.

In situ polymerization				Melt mixing
GO type and content (wt%)^a	T_c ^(b) (°C)	x_c^(c) 1^st/2^nd heating (%)	T_g ^(d) (°C)	T_c ^(b) (°C)	x_c^(c) 1^st/2^nd heating (%)	T_g ^(d) (°C)	GO type and content (wt%)^a
0% GO	23.7	60/41	−43	24.9^(e)	64/49^(e)	NA^g	0% GO
0% GO	23.7	60/41	−43	34.5^(f)	67/88^(f)	−41^(f)	0% GO
0.1% HGO	35.0	66/49	−41	NA^g	NA^g	NA^g	0.1% HGO
0.5 HGO	37.6	67/51	−39	NA^g	NA^g	NA^g	0.5 HGO
1% HGO	38.5	70/53	−39	NA^g	NA^g	NA^g	1% HGO
2.5% HGO	39.1	75/55	BS ^g	NA^g	NA^g	NA^g	2.5% HGO
0.1% FGO	32.0	63/50	−40	34.8	67/54	−43	0.1% FGO
0.5% FGO	36.9	68/52	−39	35.7	64/51	−41	0.5% FGO
1% FGO	39.4	71/56	−39	35.2	64/51	−41	1% FGO
2.5% FGO	38.7	77/59	BS^h	35.4	65/52	−42	2.5% FGO

^acontent in polymerization mixture (in situ) or mixing batch (melt mixing).

^bcrystallization temperature obtained from DSC (first run).

^ccrystallinity obtained from DSC (first run).

^dglass transition temperature obtained from DMA.

^evalue before melt mixing.

^fvalue after melt mixing.

^gNA = material not available.

^hBS = brittle sample, that cracked while being cut from the pressed foil, therefore, it could not have been subjected to any measurement.

Figure 4.

Differential scanning calorimeter curves of polycaprolactone nanocomposites.

Second, dynamic mechanical analysis provided an additional information about the polymer-nanofiller interactions, because nanofillers can restrain the mobility of polymer chains resulting in an increased glass transition temperature (T_g).²⁰ Indeed, we can see that the T_g of in situ polymerized nanocomposites increased by 4°C at 1 wt% of both GOs and then remained constant (Table 4). On the other hand, the melt mixed nanocomposites did not show any significant change in the T_g values. An analogy to our results can be found in the work of Liao et al.,²¹ who compared solvent blending and in situ polymerization to synthesize PMMA nanocomposites. The authors reported a higher increase in T_g values for chemical blending (in situ polymerization) than for physical blending (in solvent) and related such behaviour to the different level of interactions between the polymer and the nanofiller. This analogy suggests that in situ polymerization provides nanocomposites with a higher level of polymer-nanofiller interactions than melt mixing.

Morphology of nanocomposites

X-ray diffraction pattern of melt mixed nanocomposite with 2.5 wt% of FGO shows a peak at 12° that can be ascribed to FGO aggregate with a slightly reduced interlayer distance (Figure 5, detailed explanation in ESI). Nevertheless, in situ polymerized nanocomposites with both GOs showed only shallow peaks around 10°. This suggests that in situ polymerization pronounces the dispersion and exfoliation even in the case of HGO, the signal of which was around seven times stronger than the signal of FGO (Figure 2).

Figure 5.

X-ray diffraction patterns of virgin polycaprolactone, melt mixed (MM) and in situ polymerized (IS) nanocomposites with 2.5 wt% of freeze dried graphene oxide and hot dried graphene oxide; insets show the areas in range of x: 10–20° and y: 0–20,000 counts.

The optical microscopy indicates that although the agglomeration of GO sheets appear to some extent in all samples, in situ polymerized samples exhibit much better dispersion of GOs than melt mixed ones (Figure 6, SEM images on Figure ESI 4). This can be observed at all concentrations, although the largest difference is between 0.1 and 1 wt%. The comparison of the in situ polymerized samples suggests better dispersion for FGO samples. A very interesting feature suggesting partial exfoliation of FGO in 0.1 wt% FGO in situ polymerized nanocomposite is further described in ESI (Figure ESI 5).

Figure 6.

Optical micrographs of polycaprolactone/graphene oxide nanocomposites; red scale bar – 100 μm.

Mechanical properties

The choice of the synthesis method and the type and concentration of GO considerably influenced the mechanical properties of the nanocomposites (Figure 7, Table 5, Figure ESI 6 and 7). In case of the in situ polymerization, the increase in Young modulus was more pronounced for the FGO than for HGO nanocomposites. The melt mixed samples also showed increased tensile modulus. Nevertheless, the increase in Young modulus of melt mixed nanocomposites was not so significant as in the case of in situ polymerized samples. This difference can be clearly seen in Figure ESI 6, which shows the proportional improvement in Young modulus as compared to the unfilled sample. While the proportional improvement of in situ polymerized nanocomposites progressively increased up until 45% with increasing GO concentration, the improvement of melt mixed nanocomposites reached a plateau around 10% for all GO concentrations. Similar behaviour was also observed for the complex modulus values obtained by DMA measurements, where higher FGO concentrations even deteriorated the modulus of melt mixed nanocomposites (Table 5, Figure ESI 7). Moreover, the ductility of the polymer matrix remarkably decreased upon the addition of GOs, as was evidenced by the huge drop in elongation at break of all nanocomposites. In analogy to Young modulus, the influence of GOs on elongation was more pronounced for the in situ polymerized nanocomposites.

Figure 7.

Typical stress-strain curves of polycaprolactone nanocomposites.

Table 5.

Mechanical properties of polycaprolactone/graphene oxide nanocomposites.

In situ polymerization				Melt mixing
GO type and content (wt%)^a	Young modulus^b (MPa)	Elongation at break^c (%)	Complex modulus^d (MPa)	Young modulus^b (MPa)	Elongation at break^c (%)	Complex modulus^d (MPa)	GO type and content (wt%)
0% GO	146 ± 29	2900 ± 230	320	178 ± 3^(e)	18 ± 4^(e)	410^(e)	0% GO
0.1% HGO	152 ± 9	900 ± 70	330	NA ^f	NA ^f	NA ^f	0.1% HGO
0.5% HGO	163 ± 18	15 ± 5	420	NA ^f	NA ^f	NA ^f	0.5% HGO
1% HGO	202 ± 13	5 ± 2	460	NA ^f	NA ^f	NA ^f	1% HGO
2.5% HGO	BS ^g	BS ^g	BS ^g	NA ^f	NA ^f	NA ^f	2.5% HGO
0.1% FGO	151 ± 11	1360 ± 330	330	178 ± 10	22 ± 4	420	0.1% FGO
0.5% FGO	178 ± 11	6 ± 3	370	194 ± 26	18 ± 4	500	0.5% FGO
1% FGO	211 ± 40	2 ± 1	540	195 ± 6	18 ± 1	380	1% FGO
2.5% FGO	BS ^g	BS ^g	BS ^g	195 ± 19	11 ± 1	400	2.5% FGO

^acontent in polymerization mixture (in situ) or mixing batch (melt mixing).

^bYoung modulus obtained from tensile test.

^celongation at break obtained from tensile test.

^dcomplex modulus obtained from DMA.

^evalue after melt mixing.

^fNA = material not available.

^gBS = brittle sample, that cracked while being cut from the pressed foil, therefore, it could not have been subjected to any measurement.

To assess the influence of different variables on mechanical properties of the PCLO nanocomposites, it is crucial to take into account their thermal and morphological properties. Optical microscopy and XRD suggest that in situ polymerized nanocomposites exhibit better dispersion of GOs and higher level of exfoliation in comparison with melt mixed nanocomposites. On the one hand, this suggests that ultrasonication is much more efficient in exfoliating FGO than melt mixing, although we must consider the higher treatment time of the former method. On the other hand, the residence time of the PCLO in the microextruder can not be considerably extended due to the limited thermal stability of the PCLO. The better dispersion in FGO nanocomposites resulted in much higher number of polymer-nanofiller interactions that finally promoted the superior mechanical properties of in situ polymerized nanocomposites over melt mixed ones. We can also examine the various impact of GO type on properties of in situ polymerized nanocomposites in a similar manner. Hot-dried GO showed a 13-fold decrease in surface area and increased tendency to aggregate when compared with FGO. Therefore, in situ polymerized nanocomposites with FGO significantly surpassed those with HGO with respect to their mechanical properties.

However, the overall discussion on mechanical properties should not omit the effects of molar masses and crystallinity. We can see a notable difference between mechanical properties of unfilled samples. The Young modulus of unfilled melt mixed sample exceeded the modulus of its in situ polymerized counterpart by 22%. This might seem quite surprising, because the matrix for melt mixing had three times lower molar mass (Table 3). Such counterintuitive finding becomes reasonable when considering the crystallinity of the nanocomposites (Table 4, Figure ESI 2). Despite the difference in crystallinity between the melt mixed and in situ polymerized unfilled samples being rather small (67% melt mixing, 60% in situ), the recent findings of Eriksson et al. suggest that even such small variation can cause a significant increase in mechanical properties of PCLO with lower molar mass.²² Consequently, the bigger improvement in mechanical properties of in situ polymerized nanocomposites can be considered as a conjunction of increasing crystallinity and reinforcing effect of the GO sheets. Further comparison with literature data and corresponding discussion are provided in ESI (Table ESI 2).

Conclusion

We suggest that the in situ polymerization method favours the dispersion and exfoliation of GOs within the PCLO matrix, which subsequently boosts the crystallization of the PCLO matrix and improves the number of mutual interactions. The mechanical properties of in situ polymerized nanocomposites are then influenced not only by the presence of the GO sheets but also by the increased crystallinity of the polymer matrix. The choice of drying method of GO dispersion considerably affects the nanocomposite properties in that the freeze-drying method provides the GO with higher surface area. As a result, the FGO nanocomposites exhibited much more improved mechanical properties than HGO nanocomposites. To obtain the PCLO nanocomposites with improved mechanical properties, the application of in situ polymerization with FGO is highly recommendable.

Supplemental Material

sj-pdf-1-ppc-10.1177_09673911211068923 – Supplemental Material for The effect of nanocomposite synthesis and the drying procedure of graphene oxide dispersion on the polycaprolactone/graphene oxide nanocomposite properties

Supplemental Material, sj-pdf-1-ppc-10.1177_09673911211068923 for The effect of nanocomposite synthesis and the drying procedure of graphene oxide dispersion on the polycaprolactone/graphene oxide nanocomposite properties by Jaroslav Minář, Jan Doležal and Jiří Brožek in Polymers and Polymer Composites

Footnotes

Acknowledgements

The authors would like to thank François Tournilhac from ESPCI Paris for optical microscopy of nanocomposites and Jiří Minář for proofreading.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the specific university research (MSMT No 21-SVV/2019).

ORCID iD

Jaroslav Minář

Supplemental material

Supplemental material for this article is available online.

References

Woodruff

Hutmacher

. The return of a forgotten polymer-Polycaprolactone in the 21st century. Prog Polym Sci 2010; 35: 1217–1256.

Meek

Jansen

. Two years afterin vivoimplantation of poly(DL-lactide-ε-caprolactone) nerve guides: has the material finally resorbed? J Biomed Mater Res A 2009; 89A: 734–738.

Middleton

Tipton

. Synthetic biodegradable polymers as orthopedic devices. Biomaterials 2000; 21: 2335–2346.

Miner

Berzins

Bahcall

. A comparison of thermal properties between gutta-percha and a synthetic polymer based root canal filling material (resilon). J Endodontics 2006; 32: 683–686.

Potts

Dreyer

Bielawski

, et al. Graphene-based polymer nanocomposites. Polymer 2011; 52: 5–25.

Wang

Caetano

Ambler

, et al. Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for bone tissue engineering. Materials 2016; 9: 992.

Qian

Song

Zhao

, et al. 3D fabrication with integration molding of a graphene oxide/polycaprolactone nanoscaffold for neurite regeneration and angiogenesis. Adv Sci 2018; 5: 1700499.

Wan

Chen

. Poly(ε-caprolactone)/graphene oxide biocomposites: mechanical properties and bioactivity. Biomed Mater 2011; 6: 055010.

Hua

Kai

Inoue

. Synthesis and characterization of poly(ε-caprolactone)-graphite oxide composites. J Appl Polym Sci 2007; 106: 1880–1884.

10.

Dreyer

Jarvis

Ferreira

, et al. Graphite oxide as a carbocatalyst for the preparation of fullerene-reinforced polyester and polyamide nanocomposites. Polym Chem 2012; 3: 757.

11.

Dao

Choi

, et al. The Effect of oxidation on properties of graphene and its polycaprolactone nanocomposites. J Nanoscience Nanotechnology 2012; 12: 8420–8430.

12.

Whitby

RLD

Smith

Dichello

, et al. Cationic ring-opening polymerization of lactones onto chemically modified single layer graphene oxide. Mater Express 2014; 4: 242–246.

13.

Wang

Caetano

Chiang

W-H

, et al. Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration. Int J Bioprint 2016; 2: 95–104.

14.

Wang

Domingos

Scenini

. Advanced mechanical and thermal characterization of 3D bioextruded poly(ε-caprolactone)-based composites. Rapid Prototyping J 2018; 24; 731–738

15.

Rajaura

Srivastava

Sharma

, et al. Role of interlayer spacing and functional group on the hydrogen storage properties of graphene oxide and reduced graphene oxide. Int J Hydrogen Energ 2016; 41: 9454–9461.

16.

Malinová

Brožek

. Ethyl magnesium bromide as an efficient anionic initiator for controlled polymerization of ε-caprolactone. Polym Bull 2014; 71: 111–123.

17.

Patrício

Bártolo

. Thermal stability of PCL/PLA blends produced by physical blending process. Proced Eng 2013; 59: 292–297.

18.

Eigler

Dotzer

Hof

, et al. Sulfur species in graphene oxide. Chem-A Eur J 2013; 19: 9490–9496.

19.

. Investigation of oxygen reduction on activated carbon electrodes in alkaline solution. Carbon 2007; 45: 1296–1301.

20.

Ramanathan

Abdala

Stankovich

, et al. Functionalized graphene sheets for polymer nanocomposites. Nat Nanotechnology 2008; 3: 327–331.

21.

Liao

K-H

Aoyama

Abdala

, et al.

Does graphene change Tg of nanocomposites?

Macromolecules 2014; 47: 8311–8319.

22.

Eriksson

Peijs

Goossens

. The effect of polymer molar mass and silica nanoparticles on the rheological and mechanical properties of poly(Ε-caprolactone) nanocomposites. Nanocomposites 2018; 4: 112–126.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.33 MB