Functionalization and Chemical Linking of Reduced Graphene Oxide or Graphite onto Polyurethane and the Impact on the Tensile Strength and Sheet Resistance of Polymer Composites

Abstract

Reduced graphene oxide (RGO) or graphite is functionalized with hydroxyl groups for linking to the sides of polyurethane (PU) chains. Blended PU with RGO or graphite is prepared as a control for comparison. The PU composites are compared with respect to their spectroscopic, thermal, mechanical, shape memory, and sheet resistance properties. Scanning electron microscopy images demonstrate the good distribution of functionalized graphene oxide (FGO) or functionalized graphite (FG) particles on the inner surface of the PU. The linking of FGO or FG onto PU does not significantly affect the thermal behavior or shape memory properties but sharply improves the tensile strength of the PU composites without a noticeable decrease in tensile strain. The shape recovery of PU composites remains at approximately 90%, regardless of the FGO or FG content. The FG-linked PU composites exhibit a sharp decrease in sheet resistance as the FG content increases, whereas the sheet resistance of the FGO-linked PU composites does not decrease with increasing FGO content. The control PU composites with blended RGO or graphite show significant reductions in their sheet resistance. Considering the ease of functionalization of the graphite surface and the significant improvement in tensile strength, linking FG onto PU is advantageous for the development of PU composites with low sheet resistance.

1. Introduction

Fillers such as metals,^1,2 carbon nanotubes,^3,4 graphene, and graphite have been used to improve the electrical conductivity of polymer composites^5,6. Graphite has long been used as a conductive filler due to its high electrical conductivity, scalability for bulk applications, abundance and low cost for applications such as electrically conductive engineering plastic, ⁷ conductive protective coatings, ⁸ conductive foams, ⁹ and radio wave-absorbing materials ¹⁰ . Expandable graphite has recently been used as an electrically conductive filler in polymer composites due to the uniform dispersion of graphite ¹¹ . Graphene is prepared from graphite by oxidation and exfoliation and improves the electrical conductivity, mechanical strength, and gas barrier properties of the polymer composites ¹² . However, the even distribution of graphene particles in polymer composites requires control of the interface compatibility between the graphene and the polymer matrix. Recently, graphene, which is a conductive functional filler, was chemically linked to a polymer matrix via click chemistry ¹³ . The linking of functional materials, such as graphene or graphite, onto PU can impart various characteristics to polymer composites and extend their potential applications. Polymer composites are usually prepared by melt mixing or solvent blending, but the chemical linking of graphene or graphite to the polymer may provide secure bonding and a uniform distribution of graphene or graphite in the polymer matrix. However, the chemical linking of graphene or graphite onto polymer is scarcely reported due to the poor reactivity of graphene or graphite particles with polymer surface and the potential problems in large scale processing steps^14–16. PU is selected as the polymer matrix in this investigation because of its exceptional mechanical and chemical properties, such as high tensile strain and impact resistance, easy modification, and resistance to oxidation^17–19. The modification of PU with functional side groups has been used to develop characteristics such as low-temperature flexibility, ²⁰ biocompatibility,^21,22 hydrophilicity, ²³ pH sensitivity, ²⁴ and antimicrobial activity. ²⁵ Previously, we have linked RGO ²⁶ and graphite ²⁷ onto PU using the PU modification method. However, the electric conductivity of RGO-linked PU was missing and the tensile strength was low compared with the above functionalized PUs. In addition, the graphite content in the graphite-linked PU was excessive as a polymer composite. Therefore, the RGO- or graphite-linked PUs should be investigated again with the redesigned structures. In this investigation, RGO or graphite is functionalized with hydroxyl groups and chemically linked to PU as a filler through allophanate bonding to attain an even distribution and secure binding of RGO or graphite onto PU. The tensile mechanical properties of PU composites are compared with plain PU because the linked RGO or graphite particles may reduce the tensile properties, as has been observed in some polymer composites ²⁸ . Graphene or graphite is known for its electrical conductivity, so the impact of RGO or graphite on the sheet resistance of the PU composites is one of the main focuses of this investigation. PU composites with low sheet resistance have a few advantages over metals, such as easy molding, corrosion and impact resistance, and elasticity, and therefore have potential applications in electromagnetic wave reflection, electrostatic discharge, and high-voltage cables.

2. Experimental Method

2.1 Materials

Poly(tetramethylene ether)glycol (PTMG, M_n∼2,000 g/ mol, Sigma-Aldrich, St. Louis, MO, USA), 4,4'-methylenebis (phenyl isocyanate) (MDI, Junsei Chemical, Tokyo, Japan), and 1,4-butanediol (BD, Junsei) were dried under high vacuum (0.1 Torr) overnight before use. The synthetic graphite powder (soft black hexagonal crystals, particle size <20 μm, density = 2.2 g/cm³, mp = 4,489°C), potassium permanganate, sodium nitrate, and hydrazine hydrate were obtained from Sigma-Aldrich. Isopentylnitrite (TCI, Tokyo, Japan), 2-(4-aminophenyl) ethanol (Alfa Aesar, Ward Hill, MA, USA), and hydrogen peroxide (Junsei Chemical) were used as received. Dimethylformamide (DMF), diethyl ether, and acetonitrile were obtained from Duksan Pure Chemicals (Ansan, Korea), and DMF was distilled over CaH₂ before use.

2.2 Preparation of RGO

Graphite (10.0 g) was added to a mixture of sulfuric acid (300 mL) and sodium nitrate (5.0 g) at 0°C, and the mixture was magnetically stirred for 10 min. Potassium permanganate (30.0 g) was added slowly to the mixture, and the mixture was stirred for 30 min at 0°C. The mixture was sonicated for 1 h using a sonicator (Powersonic 410, Hwashin Tech, Korea) and magnetically stirred for 3 h at 25°C. The mixture was then diluted with a solution of distilled water (1.2 L) and hydrogen peroxide (300 mL), stirred for 1 h at 25°C, filtered by aspirator suction through filter paper (Whatman 42), and dried at room temperature for one day to obtain graphene oxide (GO) powder. The GO (5.0 g) was dispersed in distilled water (1.67 L), reduced by magnetic stirring with hydrazine hydrate (1.92 mL) for 24 h at 80°C, and filtered to obtain the RGO powder.

2.3 Functionalization of the RGO or Graphite Surface

A solution of 2-(4-aminophenyl)ethanol (23.0 g, 167 mmol) dissolved in 300 mL of acetonitrile and 60 mL of a 1.00 M aqueous HCl solution was stirred at 25°C for 30 min. After the addition of 20.0 g of RGO or graphite, the mixture was stirred at 25°C for 1 h. A solution of isopentylnitrite (19.5 g, 167 mmol) in 100 mL of acetonitrile was added dropwise to the above mixture via an addition funnel with vigorous stirring, and the mixture was stirred at 60°C for 12 h. The mixture was filtered through filter paper (Whatman 42) and washed with DMF until the filtrate was clear. The FGO or FG was washed with diethyl ether (100 mL x 3) to remove the remaining DMF and then dried in a vacuum oven at 60°C for 3 days.

2.4 PU Composite Synthesis

MDI (MDI-1, 5.00 g, 20.0 mmol) was added to PTMG (40.0 g, ca. 20.0 mmol) in a 500 mL four-neck flask equipped with a mechanical stirrer, condenser, temperature-controlled heating mantle, and nitrogen inlet. The mixture was allowed to react for 3 h at 50°C to prepare the prepolymer. BD (2.70 g, 30.0 mmol) was dissolved in 10 mL of DMF and added to the prepolymer. The mixture was stirred in the presence of BD for 1 h before MDI (MDI-2, 7.50 g, 30.0 mmol) was added to the reaction mixture. After the addition of MDI-2, the reaction was continued for 1 h with the slow addition of 10 mL of DMF via a dropping funnel to prevent a sudden increase in viscosity. MDI (MDI-3), as specified in Table 1 , was added to the reaction mixture, and the reaction continued at 50°C under nitrogen flow for 40 min. The MDI-3 was used as an agent to link the FGO or FG onto PU via an allophanate reaction. The FGO or FG was dispersed in 40 mL of DMF, added to the PU reaction mixtures described above, and stirred for 2 h. The PU product was precipitated in distilled water (1.5 L) to terminate the linking step onto PU. The resulting PU was cut into pieces and was thoroughly washed by magnetic stirring in distilled water (1.5 L x 3) and ethanol (1 L x 2) to remove the remaining reactants and free FGO or FG. The final PU product was suction-filtered and dried in an oven (60°C) for 3 days. The control series (the CG1 and CG2 series in Table 1 ) was prepared using the same method but using unfunctionalized RGO or graphite instead of FGO or FG to prevent linking the FGO or FG onto the PU. The electrophilic substitution of the diazotized phenylethyl hydroxyl group on the RGO or graphite surface and the linking of FGO or FG to PU are shown in Figure 1 . Specimens for the mechanical and shape memory tests were prepared according to ASTM D638 from the sheet prepared using the solvent casting method. Specifically, a solution of PU in DMF was slowly evaporated at 60°C for 60 h to obtain a 0.5-mm thick PU sheet. The film thickness was measured using a digital caliper (Mitutoyo CD-15CPX, Tokyo, Japan) as an average of five points. In Table 1 , the G1 and G2 series signify the FGO- and FG-linked PUs, and the CG1 and CG2 series belong to the control series of G1 and G2, respectively.

Figure 1

(a) Structure of plain PU, (b) hydroxyl functionalization on the RGO or graphite surface, and (c) linking of FGO or FG onto PU

Table 1

Composition of the PU composites

Sample Code	Reactant (mmol)					FGO a	FG b
	MDI-1	PTMG	MDI-2	BD	MDI-3c	(wt%)	(wt%)
L	20	20	30	30	–	–	–
G1-5	20	20	30	30	4	0.5	–
G1-10	20	20	30	30	4	1.0	–
G1-15	20	20	30	30	4	1.5	–
G1-20	20	20	30	30	4	2.0	–
G1-25	20	20	30	30	4	2.5	–
CG1-10	20	20	30	30	–	1.0	–
CG1-30	20	20	30	30	–	3.0	–
CG1-50	20	20	30	30	–	5.0	–
G2-10	20	20	30	30	4	–	1.0
G2-20	20	20	30	30	4	–	2.0
G2-30	20	20	30	30	4	–	3.0
G2-40	20	20	30	30	4	–	4.0
G2-50	20	20	30	30	4	–	5.0
CG2-30	20	20	30	30	–	–	3.0
CG2-40	20	20	30	30	–	–	4.0
CG2-50	20	20	30	30	–	–	5.0

a

The CG1 series used RGO.

b

The CG2 series used graphite.

2.5 Cross-Link Density

A specimen with dimensions of 20 × 20 × 1 mm and a known weight (m₁) was swollen in 50 mL of toluene in a closed-cap bottle for 24 h. The swollen weight of the specimen (m₂) was measured after quickly removing the toluene on the polymer surface with a tissue. The swollen specimen was dried at room temperature for one week and then weighed to obtain the dry mass (m₃). The solvent volume (V_s) of the swollen specimen, averaged over five swelling experiments, was calculated using the weight difference between the swollen (m₂) and dry (m₃) states and the solvent density (0.8699 g/cm³). The volume of the polymer (V) in its dry state was calculated by dividing the polymer dry weight (m₃) by the polymer density. The volume fraction of the polymer in the swollen state (v₁) was calculated using the equation V_p/(V_s + V_p). The cross-link density was derived using the following method. Specifically, the interaction parameter ($\chi$) between the solvent and PU was determined from equation (1) ²⁹ .

$$\chi ={\left({\delta }_1-{\delta }_2\right)}^2{V}_1/RT$$

(1)

δ₁ and δ₂ =

solubility parameters of the solvent and polymer, respectively

V₁ =

molar volume of the solvent (106.3 cm³/mol)

R =

gas constant (8.31 MPa·cm³·K⁻¹·mol⁻¹)

T =

absolute temperature (298 K)

The solubility parameters of toluene (δ₁) and PU (δ₂) were 18.2 and 20.5 (MPa)^1/2, respectively^30,31. The cross-link density was calculated from the Flory-Rehner equation (2):

$$-[\text{In}\left(1-v_2\right)+v_2+\chi v_2{}^2]=V_{1}n[v_2{}^{1/3}-1/2v_2]$$

(2)

v₂ =

volume fraction of the polymer in the swollen mass

χ =

interaction parameter

n =

cross-link density

2.6 Characterization

A Fourier-transform infrared (FT-IR) spectrophotometer (JASCO 300E, Tokyo, Japan) with attenuated total reflectance capability was used to measure the infrared spectra: 25 scans were performed with a resolution of 4 cm⁻¹ and a 2 mm/s scan speed. A differential scanning calorimeter (DSC, TA instruments DSC-Q20, New Castle, DE, USA) was used to obtain the calorimetry data from heating and cooling scans at a rate of 10°C/min between −50 and 250°C. After melting at 250°C for 5 min and cooling quickly to −50°C, the phase transition of 5 mg of specimen was monitored while warming to 250°C at 10°C/min. The dynamic mechanical properties were measured to determine the phase transition at a very low temperature using a dynamic mechanical analyzer (DMA, Triton TTDMA, Lincolnshire, UK), in which the storage modulus and loss modulus were scanned in tension mode between −150 and 100°C at 10 Hz. Scanning electron microscopy (SEM, Hitachi S-4300SE, Tokyo, Japan) images were obtained of fractured samples that had been coated with platinum (thickness of 12 nm) to observe the inner surface; samples were fractured by freezing in liquid nitrogen, followed by fracturing steps. The electrical sheet resistance of the PU composite film was measured using a four-point probe system (CMT Series, Chang Min Tech, Sungnam, Korea) at room temperature.

2.7 Tensile and Shape Memory Tests

The tensile mechanical properties were measured on a universal testing machine (UTM, LR10K, Lloyd Materials Testing, West Sussex, UK) according to the ASTM D638 standard using a 20 mm gauge length, a 10 mm/min crosshead speed, and a 500 N load cell. The same UTM equipped with a temperature-controlled chamber was used for the shape memory test. A specimen of length L₀ was drawn to 2L₀ at 45°C for 2 min and held at 45°C for 5 min. Afterwards, the specimen was cooled to −25°C with liquid nitrogen for 10 min, and the shrunken length (L₁) was measured after the upper grip was released. The shape retention (%) was calculated from L₀ and L₁ using equation (3). The specimen was heated to 45°C for 10 min, and the length (L₂) was measured. The shape recovery (%) was calculated from 2L₀ and L₂ using equation (4).

$$\text{Shape retention}=(L_1-L_0)\times 100/L_0(\%)$$

(3)

$$\text{Shape retention}=(2L_0-L_2)\times 100/L_0(\%)$$

(4)

3. Results and Discussion

3.1 Synthesis of the PU Composite

The FGO or FG is chemically linked to PU instead of blending to improve the binding and distribution of RGO or graphite in PU and the interface compatibility with PU. The surface of RGO or graphite is functionalized with groups that can react with PU for the chemical linking. The functionalization proceeds by electrophilic substitution between the π bonds of RGO or graphite and the diazonium salt form of 2-(4-aminophenyl) ethanol. The diazotization of 2-(4-aminophenyl) ethanol is conducted with isopentylnitrite instead of sodium nitrite to improve the solubility in acetonitrile. RGO is prepared from graphite through successive oxidation and reduction steps and then functionalized to obtain phenylethyl hydroxyl groups on its surface, whereas graphite is directly functionalized using the same method to produce hydroxyl groups on its surface. The graphite oxidation step may lead to the breakdown of carbon-carbon bonds but, after reduction can provide, hydroxyl groups other than the functionalized hydroxyl groups ³² . The FGO or FG is then used for linking onto PU in the subsequent linking step. PU is composed of hard (MDI and BD) and soft (PTMG) segments, for which the addition of MDI is divided into MDI-1 and MDI-2 to prevent side reactions between MDI molecules. The carbamate bond in the hard segment is the site of linking FGO or FG onto PU using MDI-3 as an activating reagent following the reported methods^23,33-35. The steps of linking FGO or FG onto PU are shown in Figure 1 . The G1 and G2 series listed in Table 1 contain FGO or FG chemically linked onto PU, respectively, whereas the RGO or graphite in the control CG1 and CG2 series is not chemically linked onto PU. The FGO content in the G1 series is gradually increased from 0.5 wt% to 2.5 wt%, whereas the RGO content in the control CG1 series is increased to 5.0 wt% to maximize the effect of the RGO on PU. However, the FG content in G2 series increases from 1.0 wt% to 5.0 wt% for both G2 and CG2 series. The inner surfaces of G1-25, CG1-50, G2-50, and CG2-50 are compared with the inner surfaces of the L form based on the SEM micrographs in Figure 2 . The FG particles (G2-50) are larger than the FGO particles (G1-25), and the FG particles in G2-50 are well dispersed in the PU matrix. However, the FGO particles in G1-25 are not evenly distributed, which may affect the sheet resistance of the PU composite. The RGO or graphite particles in the control PU composites (CG1-50 and CG2-50) are equally as well distributed as in G2-50.

Figure 2

SEM images (x 3,000) of the fractured inner surfaces of L, G1-25, CG1-50, G2-50, and CG2-50

3.2 Cross-Link Density

The cross-link density of PU composites was determined using the polymer swelling method, in which the swelling of PU is inversely related to the degree of cross-linking ( Figure 3 ). The cross-link density of the G1 series increases compared with that of the L form as the FGO content increases, whereas the CG1 series exhibits a lower cross-link density even at higher RGO contents. The cross-link density of the G2 series increases at low graphite contents but decreases as the graphite content increases further. However, the CG2 series exhibits a notably lower cross-link density compared with the G2 series. The lower cross-link densities of the CG1 and CG2 series, compared with the G1 and G2 series, are due to the lack of cross-linking by the RGO or graphite ³⁶ .

Figure 3

Cross-link density profiles of the (a) G1 series and (b) G2 series

3.3 IR Analysis

The IR spectra of the FGO and FG in Figures 4(a) and 4(b) show that new O–H stretching peaks appear at approximately 3400 cm⁻¹ for both FGO and FG. The OH groups on FGO or FG react with the MDI-activated PUs. In Figures 4(c) and 4(d), the IR spectra of the G1 and G2 series are compared with linear PU (L): N–H stretching peak at 3314 cm⁻¹, C–H stretching peaks at 2935 cm⁻¹ (aromatic) and 2835 cm⁻¹ (aliphatic), aromatic C–C ring stretching peak at 1595 and 1413 cm⁻¹, O–H bending peak at 1362 cm⁻¹, C–N stretching peak at 1309 cm⁻¹, and C–O stretching peak at 1220 cm⁻¹ are commonly observed for both the G1 and G2 series as well as the L form. The C=O stretching peaks at approximately 1700 and 1729 cm⁻¹ are used to compare the intermolecular interactions, such as hydrogen bonding and dipole-dipole interactions, between the hard segments, for which free carbonyl groups appear at a lower wavenumber (1700 cm⁻¹) than bonded carbonyl groups (1729 cm⁻¹)^37,38. The relative intensity difference between the bonded carbonyl and free carbonyl stretches provides the degree of phase separation (DPS). The DPS indicates a change in molecular interaction between the hard segments and is calculated using the equation (DPS = Area (bonded)/(Area (bonded) + Area (free))), where Area (bonded) and Area (free) represent the absorbance peak areas at 1695–1722 cm⁻¹ and 1724–1742 cm⁻¹, respectively. The overlapped carbonyl peaks are attributed to bonded and free carbonyl peaks, and their peak areas were compared in Figures 4(e) and 4(f). The DPS of G1 series increases with the increase of FGO but the DPS of G2 series decreases with the increase of FG compared to the DPS of L form: 73.5% for L increases to 86.8% for G1-5 and 82.1% for G1-25 and decreases to 67.6% for G2-10 and 68.1% for G2-50. The control PU composites also show similar trends: 85.2% for CG1-50 and 67.8% for CG2-50. The DPS results indicate that the molecular interactions between carbonyl groups increases for the G1 series and decreases for the G2 series, suggesting that the FG particles reduce the molecular interactions more than FGO particles. Therefore, the IR results confirm that the OH groups are secured onto the FGO or FG and that the molecular interactions between carbonyl groups are interrupted by the FG particles.

Figure 4

IR spectra of (a) GO, FGO, (b) graphite, FG, (c) the G1 series and (d) the G2 series, and the resolution of the overlapped carbonyl peaks of (e) L and (f) G2-50

3.4 Thermal Analysis

The melting temperature (T_m) of the soft segment and the glass transition temperature (T_g) of the PU composite were investigated by DSC and DMA, respectively. The melting peak of the soft segment is observed between 18 and 21°C in the second heating scan in Figures 5(a) and 5(b), and the T_m, T_g, and ΔH for the melting are compared in Table 2 . T_m is used as a reference temperature for the shape memory test in a later section. The T_ms of the G1 and G2 series are not different from that of the L form (20.9°C): 19.1°C for G1-15, 19.1°C for G1-25, 19.4°C for CG1-50, 18.6°C for G2-30, 19.5°C for G2-50, and 19.2°C for CG2-50. The ΔHs of the G1 and G2 series are lower than that of L (34.1 J/g) but did not change with the FGO or FG content: the ΔH of 30.0 J/g for G1-5 and 26.6 J/g for G2-10 changed to 29.5 J/g for G1-25 and 16.6 J/g for G2-50. The ΔHs of the control series are not different from those of the G1 and G2 series: 30.5 J/g for CG1-30 and 28.0 J/g for CG2-30. The data on the enthalpy change of crystallization (ΔH_c) during the cooling scan are summarized in Table 2 and show a similar trend to that of ΔH_m ( Figures 5(c) and 5(d)). The ΔH_c of the G1 and G2 series decreases slightly compared to that of the L form (-26.3 J/g), and the decrease is more obvious for the G2 series than the G1 series: −14.8 J/g for G1-15, −16.1 J/g for G1-25, −7.9 J/g for G2-30, and −13.3 J/g for G2-50. Therefore, the FGO or FG linked onto PU does not significantly affect the soft segment alignment and the energy of soft segment melting.

Figure 5

DSC heating scans of the (a) G1 series and (b) G2 series, and cooling scans of the (c) G1 series and (d) G2 series

Table 2

Comparison of the thermal properties of the soft segment of the PU composites

Sample Code	Tm (°C)	ΔHm (J/g)	ΔHc (J/g)	Tg (°C) a
L	20.9	34.1	-26.3	-65.3
G1-5	19.0	30.0	-14.5	-63.5
G1-10	19.3	30.8	-18.5	-63.8
G1-15	19.1	30.6	-14.8	-63.4
G1-20	19.0	30.5	-16.3	-62.5
G1-25	19.1	29.5	-16.1	-63.3
CG1-10	19.1	29.5	-16.2	-64.1
CG1-30	18.6	28.7	-13.3	-64.8
CG1-50	19.4	30.5	-17.1	-64.6
G2-10	19.2	26.6	-9.8	-64.2
G2-20	18.9	24.7	-5.5	-64.6
G2-30	18.6	24.7	-7.9	-63.8
G2-40	19.6	27.5	-10.0	-63.7
G2-50	19.5	28.2	-13.3	-64.6
CG2-30	19.5	28.8	-14.5	-66.1
CG2-40	19.0	28.3	-12.8	-65.8
CG2-50	19.2	28.0	-14.3	-66.3

a

T_g was determined from the loss modulus plot.

The glass transition temperature (T_g) of the soft segment was investigated by DMA, in which the storage modulus and loss modulus were monitored between −150 and 100°C ( Figure 6 ). A rapid decrease in the storage modulus and the appearance of a loss modulus peak are observed at approximately −65°C, indicating the glass transition of the soft segment. The T_g data in Table 2 are based on the loss modulus peak. The T_gs of the G1 and G2 series are expected to increase because the linked FGO or FG hinders the rotation of the PU chains during the glass transition. However, the T_g data of the G1 and G2 series remain constant with an increase in FGO or FG: the T_g of the L form (-65.3°C) changes to −63.3°C for G1-25, −64.6°C for CG1-50, −64.6°C for G2-50, and −66.3°C for CG2-50. The storage modulus and loss modulus show crystallization peaks at approximately −23°C after the glass transition, and these crystallization peaks disappear as the FGO or FG content increases. Therefore, T_g is not affected by the FGO or FG content, but the soft segment crystallization is decreased by the linked FGO or FG. Therefore, the thermal analysis results suggest that T_m, T_g, and ΔH_m are not affected by the FGO or FG content.

Figure 6

Storage modulus plots of the (a) G1 series and (b) G2 series, and loss modulus plots of the (c) G1 series and (d) G2 series

3.5 Tensile Mechanical Properties

The maximum stresses of the G1 and G2 series sharply increase compared to that of the L form: 38.9 MPa for G1-5, 41.2 MPa for G1-15, 51.7 MPa for G2-10, and 44.8 MPa for G2-30 compared to 15.2 MPa for linear PU (L) ( Figures 7(a) and 7(b)). However, the maximum stresses of the CG1 and CG2 series do not increase as much as those of the G1 and G2 series: the maximum stresses of CG1-50 and CG2-50 increase to only 11.2 MPa and 15.8 MPa, respectively. The maximum stress results suggest that the linked FGO or FG acts as a cross-linker if connected to multiple PU chains. The maximum stress does not proportionally increase with the FGO or FG content, suggesting that cross-linking does not proceed well at a high FGO or FG content due to the reduced mobility of the cross-linked PU chains. G1-20 and G2-40 have maximum stresses that are as high as the previously reported PUs with high tensile strength^39-41. The strain at break of the G1 and G2 series is similar to that of the L form: 1534% for G1-15, 1779% for G1-25, 1373% for G2-30, and 1715% for G2-50, compared to 1671% for the L form ( Figures 7(c) and 7(d)). The strain at break of the CG1 and CG2 series is comparable to that of the G1 and G2 series: 1411% for CG1-25 and 1411% for CG2-50, which are slightly less than the values for G1-25 and G2-50. The tensile test results indicate that the linking of FGO or FG sharply increases the tensile strength of the G1 and G2 series and does not decrease the tensile strain.

Figure 7

Maximum stress profiles of the (a) G1 series and (b) G2 series, and strain-at-break profiles of the (c) G1 series and (d) G2 series

3.6 Shape Memory Effect

The shape recovery and retention results of the G1 and G2 series are compared with those of linear PU (L) in Figure 8 . The soft segment T_m is used as the reference temperature for the shape memory test because the soft segment T_m is close to room temperature and shape recovery at room temperature is useful for PU applications^42,43. The shape recovery of the G1 series remains above 90% with the increase in the FGO content, and that of the G2 series is close to 90% for the entire range of FG contents: 98.3% for L changes to 94.3% for G1-25 and 89.6% for G2-50. The shape recovery of the CG1 and CG2 series also remains close to 90%:89.0% for CG1-50 and 87.9% for CG2-50. The shape retention decreases slightly after linking FGO or FG compared to the L form (89.8%): the shape retention of G1-5 (90.3%) and G2-10 (89.5%) changes to 85.3% for G1-25 and 95.3% for G2-50. The shape retention of the CG1 and CG2 series increases with the blending of RGO or graphite: 96.8% for CG1-50 and 100% for CG2-50. Therefore, the linking of FGO or FG slightly decreases the shape recovery and retention compared to linear PU because the linked FGO or FG hinders the stretching and shrinking of the PU chain that are necessary for the high shape memory effect.

Figure 8

Shape recovery profiles of the (a) G1 series and (b) G2 series, and shape retention profiles of the (c) G1 series and (d) G2 series

3.7 Sheet Resistance

The sheet resistance results of the FGO- or FG-linked PU composites are shown in Figure 9 . The sheet resistance of linear PU (1.27×10 ¹³ ω/sq) is very high due to the insulating nature of PU. The sheet resistances of the G1 series do not notably decrease with an increasing FGO content: 6.60×10 ¹² ω/sq for G1-5 changes to 4.36×10 ¹² ω/sq for G1-15 and 2.04×10 ¹² ω/sq for G1-25. However, the sheet resistances of CG1 series sharply decrease with an increasing RGO content: 1.82×10 ⁸ ω/sq for CG1-30 and 2.07×10 ⁶ ω/sq for CG1-50. The sheet resistances of the G2 series rapidly decrease with the increase in the FG content: 4.09×10 ¹² ω/sq for G2-10 decreases to 1.15×10 ⁹ ω/sq for G2-30 and 6.96×10 ³ ω/sq for G2-50. The sheet resistances of the CG2 series also exhibit a sharp decrease with an increasing graphite content: 3.24×10 ⁴ ω/sq for CG2-30 and 7.62×10 ³ ω/sq for CG2-50. Compared to the plain PU, G2, CG1, and CG2 series exhibit a significant decrease in sheet resistance, whereas G1 series does not show a noticeable decrease in sheet resistance. The uneven distribution of the FG particles in the G1 series could be responsible for its poor sheet resistance. The low sheet resistance of these PU composites is comparable to that of previous polymer composites exhibiting low sheet resistance^44-47. Considering the ease of functionalization of the graphite surface and the significant improvement in the tensile strength of G2 series, the linking of FG onto PU is beneficial for the development of PU composites with low sheet resistance. The linking of functionalized particles onto PU, instead of blending, is advantageous for simultaneously improving various properties of polymer composites.

Figure 9

Sheet resistance of the (a) G1 series and (b) G2 series

4. Conclusions

The PU composites linked with FGO or FG are compared with respect to their spectroscopic, thermal, mechanical, shape memory, and sheet resistance properties. Graphene/graphite particle surfaces are functionalized with hydroxyl groups for linking to the sides of PU chains. Blended PU composites with plain RGO or graphite are also prepared as control groups for comparison. The SEM images demonstrate the even distribution of FGO or FG on the inner surface of the PU, except in the G1 series. The linking of FGO or FG onto PU does not significantly affect the thermal behavior or shape memory properties but sharply improves the tensile strength of PU (287% for the G1 series and 344% for the G2 series) without a noticeable decrease in tensile strain. The shape recovery characteristics of G1 and G2 series remain constant at approximately 90% regardless of the FGO or FG content. G2, CG1, and CG2 series exhibit a sharp decrease in sheet resistance as the FGO or FG content increases, whereas the sheet resistance of G1 series does not decrease with an increasing FGO content. The control PU composites prepared with blended RGO or graphite (CG1 and CG2 series, respectively) unexpectedly show reduced sheet resistance. Considering the ease of functionalization of the graphite surface and the significant improvement in tensile strength of G2 series, the linking of FG onto PU is beneficial for the development of PU composites with low sheet resistance.