Synergistic Microgel-Reinforced Hydrogels as High-Performance Lubricants

The ability to create a superlubricious aqueous lubricant is important for various biological and technological applications. Here, a nonlipid biolubricant with strikingly low friction coefficients is fabricated (patented) by reinforcing a fluid-like hydrogel composed of biopolymeric nanofibrils with proteinaceous microgels, which synergistically provide superlubricity on elastomeric surfaces in comparison to any of the sole components. This two-component lubricant composed of positively charged lactoferrin microgels and negatively charged κ-carrageenan hydrogels is capable of exceeding the high lubricating performance of real human saliva in tribo tests using both smooth and textured surfaces, latter mimicking the human tongue’s wettability, topography, and compliance. The favorable electrostatic attraction between mutually oppositely charged microgels and the hydrogel reinforces the mechanical properties of the hydrogel, allowing friction reduction by combining the benefits of both viscous and hydration lubrication. The superlubricity of these microgel-reinforced hydrogels offers a unique prospect for the fabrication of biocompatible aqueous lubricants for dry-mouth therapy and/or designing of nonobesogenic nutritional technologies.

W ater forms the basis of all biological lubrication systems such as tears, saliva, and synovial fluids in humans. 1 Hydrogels and microgels are both composed of cross-linked water-swollen polymer networks, with the latter in the form of micron-sized particles. They have been extensively used to improve the rheological 2,3 and mechanical properties of biomaterials and other technological applications. 4−6 Recently, biopolymeric microgels have been reported as effective lubricants for elastomeric surfaces due to their capacity to trap water molecules, providing hydration lubrication. 2,7 Here, for the first time, we demonstrate that reinforcing a carbohydrate hydrogel with proteinaceous microgels can result in superlubricity on polydimethylsiloxane (PDMS), a commonly used material for investigation of oral processes and a novel 3D tongue-like biomimicking silicon surface. The twocomponent lubricant is composed of positively charged lactoferrin microgels (LFMs) dispersed in negatively charged κ-carrageenan hydrogels (κCHs). The synergistic effect between the components enhances the hydration lubrication, resulting in significantly lower friction coefficient values in comparison to the sole components. This patented lubricant formulation 8 is capable of exceeding the lubricity of real human saliva in orally relevant tribo-contact conditions.
All solutions are made in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES buffer) at pH 7.0. LFMs are fabricated by thermal gelation of lactoferrin protein solution followed by passing through a jet homogenizer as described in the Supporting Information, Experimental Section. Size and morphological characteristics of LFM are shown in Figure 1a. The size distribution of LFM, obtained using dynamic light scattering (DLS), is unimodal with an average hydrodynamic diameter (d H ) of 155.0 nm and a polydispersity index of 0.2. The transmission electron microscopy (TEM) image of LFMs confirms well-dispersed spheres corroborating with the DLS data. κCH is prepared by dissolving κ-carrageenan in the buffer under heating and shearing conditions. In Figure 1b, a TEM image of κCH shows typical bundles of nanofibril-shaped κCnf (κ-carrageenan nanofibrils). 9 These κCnf form an interconnected network with random distribution and orientation interacting via hydrogen bonds, resulting in gel-like structures with macroscale viscoelastic properties. 10 LFM-reinforced κCHs are fabricated by dispersing LFMs in κCH at the ratio of 0.07:1 wt./vol. Figure 1c 1 clearly shows LFM particles dispersed in the interconnected network of the κCnf-assisted hydrogel. Figure 1c 2 is the magnified image of LFMs embedded in κCH, indicating coverage of the LFMs by κCnf.
The interaction between LFM particles and κCnf is expected to produce a synergistic effect towards the lubrication capacity of the LFM-reinforced κCH. Based on these TEM images, a conceptual representation of the system is presented in Figure  1d, where a spherical cross-linked proteinaceous network represents the microgel and grey-colored rod-like flexible chains represent the nanofibrils. The visual image in Figure 1d visualizes the stretchy behavior of the LFM + κCH which is similar to the "the extensional filament-like behavior observed for a stretched drop of real human saliva".
The lubrication performance of the aforementioned lubricant compositions is first studied in rolling−sliding contacts on PDMS surfaces. Although the relevance of this conventional soft tribology test to study oral lubrication is of concern, PDMS as a model surface has been extensively used 11 in oral tribology studies. Therefore, the tests using PDMS surfaces, as a classic framework to understand the lubrication behavior, offer cross-comparison of the obtained adsorption and friction results with the literature.
In classical lubrication theories for hard-on-hard contacts, typically four lubrication regimes have been defined in tribology science, 12 which have been exploited in soft tribology. 13,14 Although a distinct transition between two successive regimes is arduous, theoretical models and modified Stribeck curves (or friction vs a modified velocity parameter) have been proven useful in qualitative determination of regimes (Supporting Information, Figure S1). 15,16 For ease of comparison in the tribo results, we use a product of lubricant viscosity and entrainment speed (ηU), which is wellestablished in tribological studies of soft surfaces. We also included the friction coefficient versus ηU for Newtonian fluids (as shown in Figure 2d) for comparison purposes. The ηU range in which the surfaces experience direct solid−solid contacts (<∼5 × 10 −5 Pa m) and the friction coefficient remains relatively constant is referred to as the "boundary regime" hereafter. Upon increasing the entrainment speed, the friction coefficient reduces sharply where the lubricant film partially carries the load; that is the "mixed regime".
The friction coefficient reduces as the contribution of the direct contact between surfaces reaches its minimum (i.e., elasto-hydrodynamic regime) and the onset of a full-film lubrication can be observed. Following transition to the full fluid-film lubrication, hydrodynamic forces bring about a lubrication regime where the contact bodies are completely separated by a lubricant film and the lubricant viscosity renders an important role (i.e., hydrodynamic regime). Figure 2a shows the friction curves for 2.0 wt % LFM, 1.1 wt % κCH, and LFM-reinforced κCH (κCH:LFM at 0.07:1 wt/ vol., containing 1.1 wt % κCH and 2.0 wt % LFM) lubricant formulations in PDMS−PDMS tribo contacts. Friction curves for LFM and κCH lubricants show nearly similar friction values across the speed range and lower values in comparison to those for the buffer. The friction curves for LFM and κCH maintain such a decreasing trend in the friction coefficient values as speed increases regardless of the concentration changes, with relatively high friction coefficients at the relatively low entrainment speeds (i.e., <0.01 ms −1 ) even at the highest concentration of original LFM and κCH (see Supporting Information, Figure S2). On the other hand, LFM-reinforced κCH shows an unprecedented superlubricity behavior with friction coefficient values as low as 1 × 10 −2 throughout the whole entrainment speed range investigated here (i.e., 1.31 × 10 −2 and 0.49 × 10 −2 at 0.005 ms −1 and 0.1 ms −1 , respectively). Interestingly, the LFM-reinforced κCH provides a superior lubricity as compared to real human saliva, demonstrating its potential as an effective biolubricant.
The mixture of untreated, i.e., nonmicrogelled, LF protein and κCH (Supporting Information, Figure S3) does not show superlubricity. This indicates the importance of using LF in microgel form in order to allow a high degree of interaction with κCH. To the best of our knowledge, such a synergistic effect to yield super low friction coefficients has been particularly achieved through synthetic aqueous lubricants 17,18 and is rarely reported for natural biopolymeric aqueous lubricants. Figure 2b shows the linear response of the elastic modulus for the hydrogel, microgels, and their mixture under shear in the frequency range of 0.1−10 Hz at a constant strain (see Supporting Information for determination of the linear regime, Figure S4). The elastic modulus of the LFM-reinforced κCH (13.26−36.63 Pa) is one order of magnitude greater than that of the sole components at equivalent concentrations (0.48− shows a relatively constant viscosity of 0.0016 Pa s which is at least two orders of magnitude lower than those of the other two fluids. The greater viscosity of LFM-reinforced κCH as compared to κCH implies the interactions between κCH and LFM (i.e., reinforcement and filling action while retaining the fluidity).
The shear thinning behavior of LFM-reinforced κCH and κCH does not reach a plateau over the shear rate examined here. Therefore, the effective tribological viscosity (η e ) is estimated using friction values in the elasto-hydrodynamic lubrication regime and de Vicente et al.'s (2005) theoretical model (see Supporting Information, Equation S1). 19 The obtained η e values are shown using dashed lines, assuming the effective Newtonian behavior of fluids in the tribological limit. 2 This brings us to evaluate the influence of viscous forces on the lubrication performance of the lubricant formulations by plotting the friction coefficient as a function of η e U ( Figure  2d). This graph sheds light on the influence of hydrodynamic viscous forces of a lubricant when compared to the performance of aqueous-based Newtonian fluid. Friction curve for κCH overlapped with Newtonian fluid almost in all regimes. This indicates that the viscous force is the main contribution to the tribological performance of κCH. Friction values for LFM are considerably lower in comparison to the values for a Newtonian fluid irrespective of the speeds. This indicates the impact of viscous and hydration action of LFM on its boundary lubrication. On the other hand, LFMreinforced κCH shows predominantly friction values which are dramatically lower than the friction coefficient values of other tested lubricants. Since the viscous behavior of LFMreinforced κCH is similar to that of κCH (Figure 2c), it is reasonable to attribute its super-low friction behavior to the hydration force which is bestowed on LFM-reinforced κCH as a result of synergistic effects between LFM and κCH. As the entrainment speed increases, the friction curve of LFMreinforced κCH overlaps with the graph for Newtonian fluid within the elasto-hydrodynamic lubrication regime, indicating that the lubrication of LFM-reinforced κCH relies completely on the high shear rate viscosity of the solution in this regime.
The gray line in Figure 2d presents the theoretical prediction of the friction coefficient in the hydrodynamic regime 19 which shows a good agreement with the experimental results. The minimum film thickness (h m ) at the onset of the hydrodynamic regime (Supporting Information, Equation S2) for LFMreinforced κCH, κCH, and the Newtonian reference is estimated to be 1514.4, 2118.2, and 1762.5 nm, respectively. 19 The film thicknesses which are greater than the diameter of microgels (an average size below 200 nm shown in Figure 1a) facilitate interposition of the microgels at the contact interface, enabling fluids to act closely to a continuum.
As for the surface adsorption properties, the quartz crystal microbalance with dissipation monitoring (QCM-D) equipped with PDMS-coated crystals is used, and the associated results are shown in Figure 2e (see Supporting Information for frequency and dissipation shift data, Figure S5 Figure S4). (c) Shear viscosity of LFM, κCH, and LFM-reinforced κCH as a function of shear rate. η e represents the effective tribological viscosity (see Supporting Information, Theory Section). (d) The friction coefficient results as a function of η e U. The gray solid line represents the estimated friction coefficient based on effective tribological viscosity (see Supporting Information, Theory Section). (e) The hydrated mass of adsorbed LFM, κCH, real human saliva, and LFM-reinforced κCH onto PDMScoated sensors obtained by QCM-D and measured using the Voigt viscoelastic model applied to 3 rd −11 th overtones (raw data of frequency and dissipation shifts of 5 th overtone are available in Supporting Information, Figure S5). The red dotted line represents the adsorption level of saliva. Human saliva was collected from a healthy young female in the morning. The subject was refrained from eating and drinking for at least 2 h before saliva collection (Ethics number: MEEC 16-046, University of Leeds, UK); the saliva was diluted with 10 mM HEPES at the ratio of 1:1 w/w and centrifuged; and the supernatant was used for the tribology measurements. Values are presented as the means ± SDs of nine readings on triplicate samples (n = 9 × 3) except for QCM-D data (n = 3 × 1).  22 Nevertheless, the high lubrication performance of LFM-reinforced κCH results from the adsorption properties of LFM and the rheological properties of κCH, the latter helping the lubricant to remain at the contact interface even under tribological shear. Further, we studied the lubrication performance of LFMreinforced κCH compositions at different ratios to better understand the synergistic interactions between the lubricant components. For this purpose, the friction coefficient results as a function of the ratio of κCH to LFM (wt/vol.) are presented in Figure 3a at 0.005 ms −1 and 0.1 ms −1 . The full data for this investigation can be found in the Supporting Information ( Figure S6). As shown in Figure 3a, the friction coefficient at both speeds increased around one order of magnitude upon decrease of the relative concentration ratio to below 0.01 wt/ vol. This indicates that a sufficient amount of κCH is required to achieve the observed superlubricity. To further elucidate this observation, ζ-potential measurements are carried out on LFM-reinforced κCH compositions. ζ-Potential results as a function of the κCH to LFM ratio are shown in Figure 3b. The ζ-potential for the sole LFM and κCH components is measured as +22.7 and −46.3 mV at pH 7.0, respectively. Oppositely charged components propel mutual electrostatic attraction, resulting in a structure observed in TEM images (Figure 1c).

ACS Macro Letters
The ζ-potential is sensitive to the κCH/LFM ratios, resulting in a decrease of electrophoretic mobility and consequently reduction in the net surface charge. As the κCH/LFM ratio increasees, the coverage of LFM by κCnf is enhancened. As a result of this, the LFM-reinforced κCH undergoes a charge reversal process, at the shear plane, gradually from net positive towards net negative values. The coverage curve, calculated from ζ-potential results ( Supporting  Information, Equation S3), 23 reaches a plateau of around 90% at a ratio of 0.07:1 wt:vol. The friction coefficient curve plateaued out approximately at the same ratio (Figure 3a). These results are in agreement with TEM observations for LFM-reinforced κCH at a ratio of 0.07:1 wt:vol (Figure 3c). These corroborate that an adequate level of coverage of LFM by the κCnf is required for LFM-reinforced κCH to deliver superlubricity.
Finally, the efficacy of the developed lubricant formulation on a tongue-emulated surface is assessed using the method developed by Andablo-Reyes et al. 24 A set of experiments are performed under orally relevant conditions with respect to speed, pressure, and surface properties. Briefly, a soft Ecoflex 00-30 was used to create silicone surfaces from a 3D-printed mold, with Young's modulus of ∼120 kPa, latter being considerably closer to the modulus of the tongue as compared to PDMS. The mold contains papillae-shaped features with appropriate size and spatial distribution of fungiform and filiform papillae, mimicking those present on the real human tongue. Span80 (0.5 wt %) was used to enhance the wettability of the Ecoflex 00-30 in some experiments. The water contact angle for the surfaces containing Span80 was measured to be 76.0°± 2.0, 24 resembling the wettability of tongue surfaces with some degree of adsorbed saliva on top. More details with reference to this testing approach can be found in a recent work by the authors. 24 Figure 4 shows friction curves for all lubricants used in this study. Panels a and b present the results for soft and textured Ecoflex 00-30 (hydrophobic) and soft Ecoflex 00-30 + Span80 (hydrophilic) materials, respectively. LFM shows poor lubrication behavior, with a slight reduction in the friction coefficient when compared to the buffer. κCH provides relatively better lubricity as compared to buffer and LFM for both surfaces at speeds below 10 −3 m/s. The lubrication performance of saliva is found to be dependent on the wettability behavior of the surfaces, particularly at speeds below 10 −3 m/s. For saliva, lower friction coefficients are observed on Ecoflex 00-30 + Span80 surfaces which surpass the lubricity of both LFM and κCH.
Results show that tribo-test conditions (contact pressure, contact geometry, and sample topography and chemistry) used with tongue-mimicking surfaces and surface properties have a large impact on the performance of the fluids (compare Figure  4 to Figure 2a). Interestingly and in agreement with the MTM results (Figure 2a), LFM-reinforced κCH outperformed the other fluids within the whole speed range measured here, with friction coefficient values an order of magnitude lower with  . Tribological performances of the lubricants obtained using tongue-mimicking tribological surfaces. LFM-reinforced κCH demonstrates the lowest friction values as compared to buffer, saliva, κCH, or LFM. Friction results obtained using soft textured (a) Ecoflex 00-30 (hydrophobic) and (b) Ecoflex 00-30 + Span80 (hydrophilic) surfaces. Human saliva was collected from a healthy young female on the morning of the testing day. The subject was refrained from eating and drinking for at least 2 h before saliva collection (Ethics number: MEEC 16-046, University of Leeds, UK). The saliva was diluted with 10 mM HEPES at the ratio of 1:1 w/w and centrifuged, and the supernatant was used for the tribology measurements. Values are presented as the means ± SDs of nine readings on triplicate samples (n = 9 × 3).

ACS Macro Letters
pubs.acs.org/macroletters Letter respect to the values for the buffer. These results prove the superlubricity of LFM-reinforced κCH for soft silicon surfaces that emulated the stiffness and topography of the human tongue with the Ecoflex 00-30 representing extreme dry mouth conditions and the modified material (Ecoflex 00-30 + Span80), representing a dry tongue with some residual saliva.
In summary, we have developed a new sophisticated nonlipidic biolubricant composed of submicron-sized lactoferrin microgels dispersed in κ-carrageenan hydrogel. The biolubricant possesses superlubricity characteristics which exceed the lubricity of real human saliva in different oral mimicking conditions. The excellent lubrication performance of this microgel-reinforced hydrogel is attributed to the synergistic effects between LFM and κCH. The synergistic effects impart superlubricity to the lubricant which is facilitated through viscous forces and surface adsorption. 25 This new hydrogel has significant potential for applications in oral care products where lubrication without lipid content is desired. For instance, dry mouth syndrome or xerostomia 26 poses a limitation to the lubrication of oral surfaces in the absence of natural saliva. In addition, intake of extra lipid for lubrication can be undesirable for these patients especially for the elderly population, where this condition is prevalent. Thus, the development of bio-inspired lubricants as an alternative to saliva, such as our LFM-reinforced κCH, is a high priority. Additionally, the present formulation can be potentially used to replicate the lubricating properties of fat content in food products, providing the possibility of decreasing caloric content, without sacrificing sensory-related attributes. 27 ■ ASSOCIATED CONTENT The manuscript was written through contributions of all authors.

Funding
The European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (Grant agreement n°757993 and 890644).