Industrial Preparation of Self-Lubricating Polyurethane via Direct Fluorination with Gaseous Fluorine.

Fluorination modification is an effective way to improve the friction property of a polymer; however, current preparation of fluorine-containing polymers is limited by the high cost and complex synthesis. Here, an industrial fluorination strategy with gaseous fluorine (F2) to achieve fluorine enrichment on the surface of a polyurethane (PU) sheet was demonstrated. Benefiting from the high reactivity and strong diffusion of F2 gas, the as-prepared PU sample was characterized by the rich fluorinated surface with the slightly fluorinated bulk. As a result, the fluorinated PU combined the excellent mechanical properties of pristine PU with the unique function of the fluorine-containing polymer, where the fluorinated surface may function as a robust lubricating layer, contributing to remarkable promotion of friction property while maintaining the mechanical strength unchanged. The friction coefficient decreased to 0.5 at the dry condition and 0.1 at the wet condition from the initial value of 1 and 0.45 for the pristine PU sample. The surface manipulation via direct fluorination offers great opportunities to tailor the friction property while keeping the mechanical strength unchanged.

Introduction

Benefiting from tailored hard and soft segments, polyurethane (PU) presents an excellent combination of stiffness, elasticity, and abrasion resistance and thus is widely used as a sealing material to produce transport belt, bearing, valve seals, O-rings in aircraft, and marine industry.[1−3] However, the friction coefficient is high with poor self-lubrication, so that in large sealing devices running under high frequency, huge stress followed by a sharp increase in friction heat often causes serious damage of the PU material.[4,5] Therefore, it is critical to reduce the friction coefficient for high-performance PU sealing materials.

Currently, there are two main strategies to ameliorate the tribological performance of the PU material. One common way is to incorporate self-lubricating fillers into the PU matrix.[6] Golaz investigated the effects of graphite, TiO2, MoS2, SiO2, and ZrO2 particles on the tribological properties of wax-containing PU composites and revealed that the lowest friction coefficient of 0.53 was achieved at 10 wt % ZrO2 and 15 wt % MoS2.[7] Yang utilized the strong load-carrying ability of multiwalled carbon nanotubes and the excellent lubricant capacity of MoS2 to greatly reduce the friction coefficient of the PU coating.[8] However, homogenous dispersion of lubricating fillers and good interfacial interaction with the PU matrix still remain challenges, and the introduction also can change the mechanical properties of the PU bulk. Another strategy for lowering the friction coefficient is to embed fluorine-containing groups into the PU chain by copolymerizing the fluoro-comonomer.[9] It is well-documented that the fluorine (F) atom has a small atomic radius and a strong electronegativity. When the F atoms are introduced into polymer chain, the C–F bond can completely cover the C–C main chain, providing shielding and protecting effects on the main chain. As a result, fluorine-containing polymers often have excellent thermal and chemical stabilities, antifouling behavior, and a low friction coefficient.[10,11] Tonelli reacted a fluorinated macrodiol with poly(tetramethylene glycol) and 4,4′-methylene-bis-phenylisocyanate via two-step polymerization technique to synthesize a fluoro-modified PU with perfluoropolyether blocks.[12] However, the introduction of fluorine-containing units by chemical methods involves some issues, such as time consumption and complex fabrication, limited sources and low reactivity of fluorine-containing reagents, and difficulty to control reaction conditions.[13]

In fact, the friction behaviors of a given polymer are largely determined by the surface chemical nature.[14] Therefore, it is unnecessary to fully fluorinate the polymer bulk. Instead, achieving surface fluorination should be a more efficient and economical strategy. Due to the quick diffusion and strong reactivity, direct fluorination by reacting gaseous fluorine with surface molecules is considered as an effective industrial technology and is applied widely to modify the surface nature of most materials.[15] Compared to bulk fluorination, surface fluorination proceeds at a mild condition without a complex synthesis, and the thin fluorinated layer of below 10 μm thickness can endow polymer articles with special nature such as barrier property, chemical resistance, friction property, adhesion, and printability while maintaining the bulk physical properties unchanged.[16] Jeong adopted direct fluorination to decorate the aramid fabric surface to enhance the interfacial interaction with the polymer matrix for improving the mechanical properties of polymer composites.[17] Vega-Cantú generated a fluoropolymer surface on nitrile rubber O-rings by exposure to F2 to enhance the chemical resistance to ZnBr2 brines at high temperatures and pressures.[18] Direct fluorination is also used to bring down the friction coefficient of polymer products, where the fluorinated layer can act as a shielding layer to provide additional lubrication effect.[19] Gao adopted direct fluorination to modify the fluoroelastomer to decrease the surface energy and ameliorate the friction properties, as evidenced by a substantial decrease in the friction coefficient from 1.04 of the untreated sample to 0.5 of the fluorinated sample.[20] Peyroux introduced 3.41% fluorine atoms on the surface of low-density polyethylene by exposing a reactive N2/F2 gaseous mixture, providing excellent tribological properties close to those of the reference PTFE.[21] Although direct fluorination has been applied successfully to elastomer, polyethylene and polypropylene, there is little published work on the friction reduction of the PU material as well as the fluorination mechanism. Here, we performed the direct fluorination of the PU sample through gaseous fluorine and compared the chemical characteristics near the surface with the pristine PU, probing the reaction mechanism of surface fluorination. Finally, the structure-property relationship of the fluorinated PU was investigated comprehensively. This study sufficiently demonstrated the potential of direct fluorination in friction reduction of the PU material, providing a low-cost, highly efficient surface modification technology for preparing high-performance PU materials.

Experimental Section

Materials

PU sheets with a urethane-based hard segment and a poly(ε-caprolactone) soft segment were provided kindly by Wanhua Chemical Group.

Sample Preparation

The 25 cm × 25 cm × 0.2 cm PU sheets were placed into the fluorination device (Jiangsu Rotam Boxmore Packaging Co., Ltd., China), and then the air was removed. Next, the F2/N2 gaseous mixture (20 vol % for F2) was injected, and the pressure was elevated to 500 mbar. The temperature of the fluorination process was kept constant at 50 °C for 2.5 h. The sample was named as F-PU.

Characterization

Scanning Electron Microscopy Observation

The surface and cross-sectional morphologies of the films were observed by a FEI Inspect FSEM instrument (FEI, USA) with an acceleration voltage of 5 kV. The elemental compositions were characterized by energy-dispersive X-ray spectrometry (EDS) combined with scanning electron microscopy (SEM). The view field illustration is shown in Figure .

Figure 1

View field illustration for SEM observation.

Fourier Transform Infrared Spectroscopy

The chemical compositions of the PU films were investigated in the attenuated total reflection mode using a Nicolet 20SXB FTIR spectrometer (Thermo Fisher Scientific Inc., USA). The spectra were collected from 4000 to 650 cm–1.

X-ray Photoelectron Spectroscopy

The surface nature was analyzed by an AXIS UTLTRA DLD multifunctional X-ray photoelectron spectrometer (Shimadzu/Kratos Ltd., Manchester, UK) with the excitation at 1486.8 eV from an Al K alpha.

Contact Angle

The contact angles (CAs) of the films were investigated by a Kruss DSA25 optical contact angle analyzer (Hamburg, Germany) at room temperature with 5 μL of water and ethanediol droplets.

Friction Properties

The friction coefficients were measured with a M-200 friction and wear testing machine (Guance Jingdian Instrument Equipment Co., Ltd., Beijing). The load applied was 5 kg, and the sliding speed was 50 rpm. In order to test the friction coefficient at the wet condition, a water droplet of about 0.05 mL was dripped on the sample surface, and then the friction coefficients were measured at the same condition.

Cyclic Mechanical Properties

The tests were divided into cyclic tensile, compressing, and bending with different sample dimensions (2 mm × 15 mm for tensile; 10 mm × 10 mm for compressing; 2 mm × 15 mm for bending). Then, the specimens were measured by a Boss 3220 SERIES II electronic universal testing machine (Bose, USA) with strains of 20, 10, and 100% for 500 cycles, and the frequency was 1 Hz.

Thermogravimetric Analysis

The samples were measured in the temperature range of 40–700 °C with a constant heating rate of 10 °C/min using a Q50 thermal gravimetric analyzer (TA Instruments Co. Ltd., New Castle, DE, USA).

Results and Discussion

Fluorination-Enabled Structure

The element compositions and distributions in the sample were analyzed by EDS combined with SEM. As shown in Figure , the surface of the PU sheets became slightly rougher after fluorination. Moreover, it was found that direct fluorination was an effective way to introduce F elements to the PU sample. Figure also presents the elemental distributions from EDS analysis scanned over the pristine PU surface, fluorinated PU surface, and cross section. One should note that few fluorine elements were present on the surface of the pristine PU where trace amounts of fluorine (1.89%) may originate from the signal noise. In addition to the C, N, and O elements contained in the PU sample itself, rich F elements appeared after fluorination treatment with F2 gas. Moreover, different chemical compositions between the surface and bulk were observed. The mapping images of F elements (Figure g and k) clearly revealed that the F element was mainly distributed on the surface rather than in the bulk. Compared to the high F content of 14.92 wt % on the surface, a clear decrease in F elements on the cross section (2.9 wt %) supported the preferential attachment of F element onto the surface. Considering the high reactivity and strong diffusion rate of F2 gas as well as the amorphous characteristics of the PU matrix, it should be reasonable that the bulk sample was partially fluorinated.

Figure 2

SEM morphology of the pristine PU surface (a), fluorinated PU surface (e), and cross section (i). The corresponding color-coded element maps (b,f,j), spatial distribution of fluorine atoms (c,g,k), and EDS spectra with weight proportions of different atoms (d,h,l) from EDS analysis scanned over the surface and cross section of the PU samples.

In order to reveal the surface fluorination mechanism, IR and X-ray photoelectron spectroscopy (XPS) analyses were combined to characterize the variation of the surface chemical bonds after fluorination treatment. Figure shows the IR spectra of the pristine PU sample and the fluorinated sample. For the pristine PU sample, the characteristic peaks appeared at 3334, 2951, 1725, and 1530 cm–1 corresponding to N–H stretching vibration, C–H aliphatic stretching vibration, C=O stretching vibration, and N–H bending vibration, respectively. After fluorination treatment, the characteristic peaks from the C–H stretching vibration and N–H bending vibration became weaker compared to the invariable peak at 964 cm–1 corresponding to the standard C–C stretching vibration,[22] while the new absorption peaks at 737 cm–1 from the C–F rocking and wagging vibrations were excited,[10,23] implying that the fluorine element was introduced to the PU sample. One also notes that there were multi broad peaks over 3000–3600 cm–1 corresponding to the formation of the −OH group and the accompanying hydrogen bonding with −NH groups.[24] The −OH group was related to the fluorination of the −CONH– group followed by the resulting −COF group hydrolysis.[25,26] For example, Leu investigated the −COF group formation followed by partial hydrolysis into the −COOH group in the fluorinated polyimide and poly(methyl-methacrylate) surfaces.[27]

Figure 3

IR spectra of the pristine PU sample and fluorinated PU.

The XPS measurements verified the fluorine enrichment on the PU surface upon fluorination treatment, as illuminated in Figure . The survey spectra showed that no peak assigned to the fluorine element was detected in the pristine PU sample, while the F-PU sample contained the F peak in a range of 682–692 eV. The F content in the F-PU sample was 21.7%, higher than the EDS result in Figure h. This increase essentially originated from the different test depth of XPS (10 nm) and EDS measurements (1 μm).[28] This also supported the surface fluorination with little effect on the bulk. Further, the detailed spectra in C1s, N1s, and O1s regions were shown to reveal different chemical bonding states in the PU and F-PU samples, with different profiles composed of the deconvoluted peaks (Figure b–d and f–h). For the C1s XPS spectra, one can observe two additional CHF and C–F peaks at 287.5 and 290.2 eV besides the peaks of C–C, C–N, and C=O groups at 284.8, 285.7, and 289.7 eV.[21,29] The formation of the fluorine-containing groups signified that direct fluorination introduced the F element onto the PU surface. Turning to the N1s range, a single peak at 400.3 eV corresponding to the C–N group was observed in the pristine PU sample, in good agreement with the typical XPS result of the reported PU sample.[29] After fluorination, the shift to higher energy indicated the formation of new N species, which can be attributed to strong electron-withdrawing substituents of F and O atoms attached to N atoms.[30] Accordingly, 401.6 and 403.2 eV were assigned to the N–O group and the N–F group. The formation of N–F groups suggested that F replaced H in the N–F, while the N–O group reflected partial hydrolysis of the C–N group in the amide group. The latter was confirmed by the deconvoluted O1s spectra. The pristine PU sample exhibited a single peak at 532.4 eV assigned to the COON group, while the F-PU contained two components where the high-energy peak at 533.9 eV should root from the accompanied O–H group.[30] By combining the IR and XPS results, a fluorination mechanism was proposed as following: It is well documented that direct fluorination with F2 gas was a heterogeneous reaction of gaseous fluorine with a given polymer surface based on selective substitution of H atoms with F atoms.[31] In the early stage of PU fluorination, F substitution reaction happened in the −CH2– segments away from the electron-withdrawing C=O group. Subsequently, partial fluorination at the α-C position destabilized the amide group,[32] triggering the cleavage of the C–N bond, as evidenced by the presence of N–F and O–H groups.

Figure 4

XPS spectra of the pristine PU (a–d) and fluorinated PU (e–h): survey XPS spectra (a,e) and deconvoluted spectra of C (b,f), N (c,g), and O (d,h) regions.

Fluorination-Enabled Surface Property

The introduction of C–F bonds and −COOH groups into the F-PU surface via direct fluorination modified the surface nature. Figure displays the CAs of the samples with deionized water and ethanediol. The pristine PU sample was highly hydrophobic, and the CAs of water and ethanediol were 114.7 and 90.1°. Direct fluorination reduced the two values to 51.5 and 29.8° and increased the surface energy from 19.3 mN/m of the pristine PU sample to 47.9 mN/m so that the resulting F-PU sample became hydrophilic. Different from the low surface energy and hydrophobic property of the typical fluorine-containing polymer, the hydrophilic nature was essentially attributed to the −COOH group generated by the cleavage of the C–N bond.

Figure 5

CA images of PU (a,b) and F-PU (c,d) with deionized water (a,c) and ethanediol (b,d), and the corresponding CA and surface energy (e).

Fluorination-Enabled Friction Property

By integrating individual virtues of PU and the fluorinated polymer, the resulting F-PU sample exhibited excellent comprehensive properties. The variation of friction coefficient at dry and wet conditions was evaluated. As shown in Figure a–c, the F-PU sample exhibited a lower friction coefficient than the pristine PU sample. At the dry wearing process, the pristine PU was characterized with a high friction coefficient of 1–1.5. The increase in the friction coefficient should be attributed to the friction heat generated during the tribological test. Moreover, the triboelectric property was determined by the surface nature. Inheriting the merits of the F-containing polymer, the fluorinated surface endowed the F-PU sample with excellent self-lubricating property, decreasing the friction coefficient to 0.5–07.

Figure 6

Friction behaviors of the PU and F-PU samples at dry (a) and wet conditions (b) with the corresponding friction coefficient (c); water states after cyclic sliding of the metal cylinder on the sample surface with a water drop (d).

When water was introduced into a polymer–metal sliding interface, the friction coefficient of the two samples decreased sharply, originating from the additional slippage function provided by water located at the specimen/the counterpart cylinder interface. However, the pristine PU sample was characterized by poor stability. Within the first 30 s of the test, the friction coefficient kept a low value of ∼0.5 and quickly increased to ∼1.5. The drastic fluctuation of the friction coefficient was derived from the water-repellent property initiated by the hydrophobic nature of the PU sample. In this case, the loaded water absorbed ineffectively onto the surface, resulting in a gradual decrease of water at the interface during the testing process (Figure d). Finally, the additional sliding against the counterpart cylinder conferred by water disappeared, and the value became similar to the value tested under dry conditions. On the contrary, the −COOH groups in the F-PU surface caused the hydrophobicity-to-hydrophilicity transformation, favorable to the water spread and fixation on the surface. Therefore, at the wet wearing process, the added water was immobilized tightly on the sample surface and served as a stable lubricating layer, keeping the friction coefficient at ∼0.1. As shown in Figure d, the water presented good spread at the F-PU surface after 20 cycles.

Bulk Properties of Fluorinated PU

Because direct fluorination only achieved the F enrichment on the surface, the bulk properties can be well maintained. Figure evaluates the resistances of the PU and F-PU samples to various mechanical deformations (tensile, compression, and bending stress), with the corresponding results. It can be found that although the F-PU sample exhibited excellent friction resistance, the mechanical properties of the bulk were nearly unchanged. For example, the tensile modulus was 3.5 MPa, approaching 3.3 MPa of the PU sample. The small thickness of the fluorinated layer should be responsible for the similar mechanical properties. It is worth pointing out that the F-PU sample also preserved the original fatigue performance, with stable mechanical strength in ∼500 cycles, as was presented in Figure a.

Figure 7

Cyclic tensile (a), compressing (d), and bending (g) curves of the PU and F-PU samples, with the stress–strain curve of a single cycle (b,e,h) and the corresponding stress and modulus (c,f,i).

Further, the thermal stabilities of the PU and F-PU samples were compared by thermogravimetric analysis (TGA), as shown in Figure . One can observe that the two PUs exhibited a two-stage decomposition process, where the first degradation step at 360 °C corresponds to the cleavages of the urethane bonds in hard segments, while the second step at 370–450 °C corresponds the scission of soft segments.[33] Moreover, the initial decomposition temperature and the maximum degradation rate remained similar, suggesting that the fluorinated layer was very thin, so that the thermal stability was influenced hardly.

Figure 8

TGA (a) and DTG (b) curves of the PU and F-PU samples.

Accordingly, it can be concluded that compared to the other preparation methods of fluorinated PU, this direct fluorination is featured with low energy consumption and a simple and mild process and can offer adequate strong coating of F elements onto the sample without affecting the bulk, holding promising potential in industrial fabrication of high-performance self-lubricating PU materials.

Conclusions

In this study, direct fluorination of the PU sample through F2 gas was adopted to achieve fluorine enrichment on the PU surface. As a result, a fluorinated layer was constructed by F substitution of H atom in the −CH2– segments, and the hydrophobicity-to-hydrophilicity transformation was initiated by the cleavage of the C–N bond. Benefiting from the rich fluorinated surface and water fixation, the fluorinated PU sample exhibited excellent comprehensive properties. The friction coefficient decreased to 0.5 at the dry condition and 0.1 at the wet condition from the initial value of 1 and 0.45 for the pristine PU sample, while the mechanical strength and thermal property kept unchanged. This study sufficiently demonstrated the potential of direct fluorination in friction reduction of the PU material, providing a low-cost, highly-efficient surface modification technology for preparing high-performance PU materials.