Article pubs.acs.

org/Langmuir

Superlubricity Achieved with Mixtures of Acids and Glycerol

Jinjin Li, Chenhui Zhang,* Liran Ma, Yuhong Liu, and Jianbin Luo
State Key Laboratory of Tribology, Tsinghua University, Beijing, 100084, China ABSTRACT: In this work, superlubricity between glass and Si3N4 surfaces lubricated by mixtures of acid solutions and glycerol solutions has been found by using a traditional tribometer. Ultralow friction coecients of between 0.004 and 0.006 were obtained after a running-in period. Related experiments indicate that the hydrogen ions in the mixtures play an important role in achieving superlubricity. Moreover, the ultralow friction is also closely related to the pH value of the acid and the concentration of glycerol. According to these results, the possible superlubricity mechanism has been revealed, which is attributed to a uid-hydrated water layer between the hydrogen-bonded networks of glycerol and water molecules on the positively charged surfaces.

INTRODUCTION Since the appearance of the concept of superlubricity,1 it has attracted a great amount of attention among researchers in many elds in their search for lubricants with superlubricity properties ( < 0.01),2 which would be a very eective way to partially solve energy issues. Some self-lubricating solid lubricants were rst found to have ultralow friction properties, such as diamondlike carbon lms,3,4 molybdenum disulde,5,6 and highly oriented pyrolytic graphite.7 This kind of superlubricity generally originates from incommensurate surface lattice structures, weak dispersive interlayer interactions, or coulomb repulsion at the contact,810 which always requires a special lubrication condition, such as high vacuum or nitrogen protection, to reach ultralow friction. In contrast, some waterbased materials (e.g., polymer brushes with water,11,12 ceramic materials with water,13,14 glycerol solution with boric acid or polyhydric alcohol,15,16 and some kinds of polysaccharide mucilages from plants17,18) can lead to superlubricity without special lubrication conditions. In our previous work, the superlubricity between a glass plate and a Si3N4 ball with the lubrication of phosphoric acid solution (pH 1.5) has also been obtained.19 It is found that the achievement of superlubricity is closely related to two important conditions. The rst is the positively charged surfaces induced by the attached hydrogen ions via a protonation reaction.20 The second is the hydrogen-bonded network between phosphoric acid and water formed in the contact region.19 According to this mechanism, it could be inferred that if there is a lubricant meeting these two conditions simultaneously then superlubricity would appear. To meet the rst condition, the acid solutions are the best choice. However, it is hard to nd one kind of acid (except for phosphoric acid) that also meets the second condition. To meet the second condition, polyhydroxy aqueous solutions such as glycol and glycerol are the most probably candidate because of their strong
2012 American Chemical Society

The sulfuric acid (H2SO4) solutions used in the tests were prepared by diluting pure sulfuric acid (purity >99.7%) with deionized water to dierent pH values (0, 0.5, 1, 1.5, 2, 2.5, and 3.0). Pure glycerol (purity >99.7%) was diluted with dierent amounts of deionized water (5, 10, 20, 30, 40, 50, 60, 70, 80, and 90% v/v). The mixtures were obtained by mixing the acid solutions with glycerol solutions in a volume ratio of 1:10 (the pH value of the mixtures is approximately equal to the pH value of sulfuric acid solutions plus 1, respectively). In addition, ve representative kinds of acids, including hydrochloric acid (HCl), lactic acid (C3H6O3), oxalic acid (H2C2O4), citric acid (C6H8O7), and sulfamic acid (H3NO3S) with a pH value of 1, mixed with glycerol solution (20% v/v) in a volume ratio of 1:10 (the pH value of the mixtures is approximately 2), were used for comparison. The friction tests were performed on a Universal Micro-Tribotester (UMT-2, CETR) with a rotational mode of ball-on-disk. The friction Received: November 19, 2012 Revised: December 6, 2012 Published: December 10, 2012
271
dx.doi.org/10.1021/la3046115 | Langmuir 2013, 29, 271275

hydrogen bond eect, but there are no hydrogen ions in the polyhydroxy aqueous solutions. Therefore, we propose a novel method to achieve superlubricity by mixing the acid solution and polyhydroxy aqueous solution to meet the two conditions simultaneously. To conrm whether superlubricity can be obtained by using this method, six kinds of acid solutions were mixed with the glycerol solution, and the frictional behavior lubricated by these mixtures was investigated with a traditional tribometer. With the lubrication of these mixtures, the superlubricity was obtained after a running-in period. In the present work, the eect of the concentration of acid solution and glycerol solution on the superlubricity was discussed to reveal the key factors in obtaining the superlubricity. On the basis of these experimental results, a lubrication model was proposed and a new waterbased superlubricity system was consequently established.

MATERIALS AND METHODS

Langmuir
pairs were Si3N4 ball with a diameter of 4 mm (obtained from Shanghai Research Institute of Materials, the manufacturing process is gas-protecting sintering and hot isostatic pressing) and a glass plate with a surface roughness of 5 nm (part of the common glass slide for microscopy). The lubricants were supplied between the ball and the glass plate with a volume of 20 L before the tests. The load applied to the glass plate was 3 N, corresponding to the maximum contact pressure of 700 MPa. The rotational speed of the glass plate was 180 rpm with a track radius of 4 mm, corresponding to a linear speed of 0.075 m/s. To ensure the measurement accuracy of the friction coecient, the measuring errors during tests were eliminated with the method in ref 21. All tests were performed at ambient temperature (25 C), and the relative humidity was about 4555%.

Article

lubricants between a glass plate and a Si3N4 ball. It is found that the friction behavior lubricated by these ve kinds of mixtures is similar to that lubricated by the mixture of sulfuric acid solution and glycerol solution. The ultralow friction coecients after the running-in period are shown in Table 1. It is observed that all of Table 1. Friction Coecient under the Lubrication of Mixtures of Acid Solutions (pH 1) and Glycerol Solution (20% v/v) with a Volume Ratio of 1:10 and the Friction Coecient under the Lubrication of Acid Solutions Only
HCl friction coecient with mixtures friction coecient with acids only 0.004 0.03 C3H6O3 H2C2O4 0.005 0.04 0.004 0.05 C6H8O7 0.006 0.06 H3NO3S 0.005 0.05

RESULTS AND DISCUSSION The friction coecient as a function of time, lubricated by sulfuric acid solution (pH 1) mixed with glycerol solution (20% v/v), was rst investigated, as shown in Figure 1. It was found

the mixtures mentioned above can lead to friction coecients of less than 0.01 in spite reducing to dierent acid radical ions. Therefore, it can be concluded that the hydrogen ions in the mixture play an important role in the achievement of superlubricity. To investigate further the eect of hydrogen ions on the superlubricity, the friction behavior under the lubrication of H2SO4 solutions with dierent pH values (from 0 to 3 with an interval of 0.5) mixed with glycerol solution (20% v/v) was investigated, as shown in Figure 2. It is found that only when

Figure 1. Friction coecient with time as lubricated by the mixture of sulfuric acid solution (pH 1) and glycerol solution (20% v/v) with a volume ratio of 1:10. The inset is the friction coecient with time as lubricated by sulfuric acid solution (pH 1) and glycerol solution (20% v/v) separately.

that the friction coecient is about 0.35 at the beginning of the test and then decreases rapidly to a value of 0.05 after 80 s. After that, the friction coecient decreases further to a value of 0.004 after a period of 600 s, which enters the superlubricity regime. Here, the time that elapses from the beginning of the test to the appearance of the smallest amount of friction is dened as the running-in period, which is about 600 s in this test. After the running-in period, the friction coecient remains at an ultralow value ( = 0.004) stably until the end of the test. However, if the Si3N4 ball and glass plate are lubricated by sulfuric acid solution (pH 1) or glycerol solution (20% v/v) separately, then the superlubricity ( < 0.01) cannot be obtained, as shown in the Figure 1 inset. The friction coecient lubricated by sulfuric acid solution singly reduces to 0.05 after about 130 s and then remains stable without further reduction. A similar resultlubrication by glycerol solution singlyhas also been observed. The friction coecient hardly decreases anymore when it decreases to 0.03 after about 350 s. Compared to these results, it is concluded that mixing sulfuric acid solution with glycerol solution can endow them with more excellent lubricating properties. To investigate whether the positive hydrogen ions or the negative acid radical ions in sulfuric acid solution contribute to the superlubricity for the mixture, ve other kinds of acid solutions (HCl, C3H6O3, H2C2O4, C6H8O7, and H3NO3S, pH 1), mixed with glycerol solution (20% v/v), were used as
272

Figure 2. Friction coecient with time under the lubrication of H2SO4 solutions (pH 0, 0.5, 1, 1.5, 2, 2.5, and 3) mixed with glycerol solution (20% v/v). The volume ratio of the acid solution and the glycerol solution is 1:10.

the pH value of H2SO4 solution is not more than 1 (corresponding to a pH value of 2 for the mixed solution) can the superlubricity ( < 0.01) be obtained after the runningin period. And the nal friction coecient would be greater than 0.01 if the pH value of the H2SO4 solution is greater than 1. These results suggest that the superlubricity is very dependent on the concentration of hydrogen ions in the mixture. It even has no eect on the superlubricity if the concentration of hydrogen ions in the mixture is too low. Therefore, it is inferred that there must be an interaction among hydrogen ions, glycerol, and friction surfaces in the contact region during the running-in period, which is a crucial factor in the superlubricity. It has been conrmed that the hydrogen ions in a strong acid solution (pH 2) can be adsorbed onto SiOH surfaces by a protonation reaction,22 which is represented as
SiOH + H+ SiOH 2+
(1)

dx.doi.org/10.1021/la3046115 | Langmuir 2013, 29, 271275

Langmuir where the symbol represents the surface atoms attached to the underlying bulk solid.23 Because of the SiOH bonds on both the glass surface and the Si3N4 surface, the protonation reaction in eq 1 would occur when the glycerol solution is mixed with an acid solution (pH 2 for the mixture), leading to many positively charged sites on the friction surfaces. The positive charge is neutralized by equal dissolved counterions to form the stern layer and the diuse electrical double layer, which can lead to the friction decreasing in the early stage of the test.20 As shown in Figure 2, the superlubricity appears when the pH value of H2SO4 solution is not more than 1, corresponding to a pH value of less than 2 in the mixed solution. Therefore, this indicates that the hydrogen ions adsorbed on the friction surfaces, making the surfaces electrically positive, are a precondition of superlubricity. However, superlubricity cannot be obtained if there are only hydrogen ions adsorbed on the friction surfaces, as conrmed by the friction results with the lubrication of single acid solutions. As shown in Table 1, the friction coecients are in the range of 0.03 to 0.06 when only the acid solutions are used as lubricant, which implies that the glycerol molecules also play an important role in the superlubricity. To investigate the contribution of glycerol to the superlubricity, the friction behavior under the lubrication of H2SO4 solution (pH 1) mixed with glycerol solutions (5, 10, 20, 30, 40, 50, 60, 70, 80, and 90% v/v) was investigated. As shown in Figure 3, the

Article

viscosity coecient of the glycerol solution ( = 6 109 Pa1),25 u is the average linear speed of a glass plate and ball, W is the load, k is a coecient (1), and E is the reduced Youngs modulus of the two contacting solids dened by
1 v 2 1 v2 2 1 E = 2 + E2 E1

(3)

Figure 3. Friction coecient with time under the lubrication of the mixture of H2SO4 solution (pH 1) and glycerol solutions (5, 10, 20, 30, 40, 50, 60, 70, 80, and 90% v/v) with volume ratio of 1:10 and the lm thickness under the lubrication of glycerol solutions (5, 10, 20, 30, 40, 50, 60, 70, 80, and 90% v/v) predicted by the HD equation.

superlubricity can be obtained after a running-in period only when the concentration of glycerol is lower than 40%, whereas the nal friction coecient would be greater than 0.01 if the concentration of glycerol is greater than 50%. This suggests that the superlubricity is also closely related to the concentration of glycerol in the mixture. It is obvious that the concentration of glycerol has an eect on the viscosity of the mixed solution, which can inuence the lm thickness between two sliding surfaces. According to the HamrockDowson (HD) theory,24 the lm thickness of glycerol solutions can be predicted by the following expression
G*0.53u*0.67 (1 0.61e0.73k ) (2) W *0.067 2 where Hc * = hc/R, G* = E, u* = 0u/ER, W* = W/ER , hc is the lm thickness, R is the radius of the ball, is the pressure * = 2.69 Hc
273

where vi is Poissons ratio for material i and Ei is the elasticity modulus of material i. 0 is the bulk viscosity of the glycerol solutions with dierent concentrations, which is measured at 25 C with a standard rheometer (Physica MCR301, Anton Paar). The lm thickness between two surfaces as a function of the concentration of glycerol, predicted by the HD equation, is shown in Figure 3. It is found that the lm thickness is less than 2 nm when the concentration of glycerol is lower than 40% whereas the lm thickness is more than 3 nm when the concentration of glycerol is higher than 50%. It is thought that there is no elastohydrodynamic lm between the two contacting surfaces when the concentration of glycerol is lower than 40%, which is also the precondition of superlubricity as conrmed by Figure 3. What is the origin of superlubricity when mixing an acid solution with a glycerol solution? As mentioned above, at least two conditions are required to obtain the superlubricity: the hydrogen ions adsorbed on the two surfaces to make the contact surfaces positively charged and no elastohydrodynamic lm formed between the two contact surfaces. In our previous work, it has been conrmed that the tribochemical reaction of surface protonation (1) occurs when the Si3N4 and glass surfaces slide against each other in an acid solution (pH 2), which can make the two contact surfaces positively charged (forming the Stern layer) and lead to the friction coecient decreasing to about 0.05 after a running-in period.20 This is the reason that the initial friction coecient lubricated with these mixtures decreases to 0.05 after a short test period (80 s). During the test, the free water in the mixture evaporates gradually, which leads to a reduction of the proportion of water molecules in the mixture. When the proportion of water molecules reduces to a constant (the water molecules and the glycerol molecules are in equilibrium), the lowest friction coecient appears. At the moment, it is thought that there is a stable hydrogen-bonded network between glycerol molecules and water molecules formed on the Stern layer between the two surfaces as a result of the strong hydrogen bond eect of hydroxy in glycerol molecules, just like the hydrogen-bonded network between phosphoric acid and water molecules.19 After that, the superlubricity would remain constant as long as the equilibrium state between water and glycerol is not broken. The ultralow friction also suggests that there has to be an layer easy to shear between the two contacting surfaces. Therefore, it can be inferred that a layer of hydrated water molecules is adsorbed on the hydrogen-bonded network, just as a water layer on the ice surface.26 This layer is hard to squeez out of the contact region under pressure, and it can also provide a very low shear strength because of its excellent uidity.27 The possible lubrication model is illustrated in Figure 4. The superlubricity is attributed to forming a hydrated water layer between the hydrogen-bonded networks of glycerolwater on the Stern layer. Therefore, if the concentration of hydrogen ions in the mixture is too low, then the tribochemical reaction of surface protonation induced by hydrogen ions would not occur in the
dx.doi.org/10.1021/la3046115 | Langmuir 2013, 29, 271275

Langmuir

Article

Figure 4. Schematic illustration of the possible structure between two friction surfaces. The red circles represent oxygen, and the white circles represent hydrogen.

running-in period. In this case, the positively charged surfaces would not be formed, which is the reason that the superlubricity does not appear. Moreover, if the concentration of glycerol is too high, then the elastohydrodynamic lm would be formed at the beginning of the test (Figure 3). In this case, the running-in process (rubbing action) would not occur in the early stage of the test, which is the reason that the nal friction coecient is a typical value of the elastohydrodynamic lubrication. To verify this lubrication model further, two designed experiments were carried out as follows. When the superlubricity appears, a spot of NaOH powder (0.01 g) is introduced into the lubricant (mixture of acids and glycerol) to neutralize hydrogen ions, as shown in Figure 5. It is found that the friction coecient would increase to 0.02 once the NaOH powders are introduced. After that, there is a slight increase with time, and then the value remains high until the end of the test. In addition, when the superlubricity appears, a spot of deionized water (2 L) is also introduced into the

lubricant (mixture of acids and glycerol) to destroy the hydrogen-bonded network, as shown in Figure 5. It is observed that the friction coecient would increase to 0.06 once the deionized water is introduced, which conrms that the superlubricity would disappear when the equilibrium state between water molecules and glycerol molecules is broken. However, with free water evaporating, the friction coecient can decrease gradually to 0.005 after a running-in period of 200 s. This indicates that the superlubricity can be restored once excess free water has been removed to form the stable hydrogen-bonded network. It is therefore clear that the superlubricity is closely linked to the hydrogen ions adsorbed on the friction surfaces and the hydrogen-bonded network formed on the Stern layer. However, it is hard to give the full structure of the hydrogen-bonded network at present, and a new experiment with sum frequency generation spectroscopy (SFG) is planned to study the structure in the near future. According to this lubrication model, a new water-based superlubricity system is established as follows. First, the friction surfaces should have SiOH bonds, which can adsorb hydrogen ions in the lubricant by surface protonation. Second, there are a sucient number of hydrogen ions in the lubricant to make the surfaces positively charged. Third, there are many hydroxyl groups in the lubricant that form the hydrogen-bonded network. If these three conditions are met simultaneously, then superlubricity appears; two typical examples are phosphoric acid19 and the mixtures of acids and glycerol solution as we discussed above. On the basis of this superlubricity system, we believe that more and more superlubricity materials will be found in the foreseeable future.

CONCLUSIONS We have shown that the mixture of acid solution and glycerol solution can eectively reduce the friction coecient between Si3N4 and glass to 0.004 after a running-in period. Experimental results indicate that the ultralow friction is dependent on the pH value of the acid and the concentration of glycerol. The ultralow friction is mainly attributed to forming the hydrogenbonded network of glycerol and water on the Stern layer with a uid-hydrated water layer. This nding seems to be helpful for us to identify additional kinds of water-based superlubricity materials for technological and biomedical applications.

Figure 5. Friction coecient with time after a spot of NaOH powder is introduced into the lubricant and the friction coecient with time after a spot of deionized water is introduced into the lubricant. The lubricant is the mixture of H2SO4 solution (pH 1) and glycerol solution (20% v/v) with a volume ratio of 1:10.
274

AUTHOR INFORMATION

Corresponding Author

*Tel: +86-10-62773129. E-mail: chzhang@tsinghua.edu.cn.

Notes

The authors declare no competing nancial interest.

ACKNOWLEDGMENTS This work is nancially supported by the NSFC of China (51021064 and 51222507), the Program for New Century Excellent Talents in University of the Ministry of Education of China, and the Basic Research Program of Shenzhen (0021539012100521066).

(1) Hirano, M.; Shinjo, K. Atomistic Locking and Friction. Phys. Rev. B 1990, 41, 1183711851. (2) Erdemir, A.; Martin, J.-M. Superlubiricty; Elsevier: New York, 2007.
dx.doi.org/10.1021/la3046115 | Langmuir 2013, 29, 271275

REFERENCES

Langmuir
(3) Heimberg, J. A.; Wahl, K. J.; Singer, I. L.; Erdemir, A. Superlow Friction Behavior of Diamond-Like Carbon Coatings: Time and Speed Effects. Appl. Phys. Lett. 2001, 78, 24492451. (4) Erdemir, A. Genesis of Superlow Friction and Wear in Diamondlike Carbon Films. Tribol. Int. 2004, 37, 10051012. (5) Martin, J. M.; Donnet, C.; Lemogne, T.; Epicier, T. Superlubricity of Molybdenum-Disulfide. Phys. Rev. B 1993, 48, 1058310586. (6) Chhowalla, M.; Amaratunga, G. A. J. Thin Films of FullereneLike MoS2 Nanoparticles with Ultra-Low Friction and Wear. Nature 2000, 407, 164167. (7) Dienwiebel, M.; Verhoeven, G. S.; Pradeep, N.; Frenken, J. W. M.; Heimberg, J. A.; Zandbergen, H. W. Superlubricity of Graphite. Phys. Rev. Lett. 2004, 92, 12610111261014. (8) Socoliuc, A.; Bennewitz, R.; Gnecco, E.; Meyer, E. Transition from Stick-Slip to Continuous Sliding in Atomic Friction: Entering a New Regime of Ultralow Friction. Phys. Rev. Lett. 2004, 92, 134301-1 134301-4. (9) Hirano, M.; Shinjo, K.; Kaneko, R.; Murata, Y. Observation of Superlubricity by Scanning Tunneling Microscopy. Phys. Rev. Lett. 1997, 78, 14481451. (10) Sun, C. Q.; Sun, Y.; Ni, Y. G.; Zhang, X.; Pan, J. S.; Wang, X. H.; Zhou, J.; Li, L. T.; Zheng, W. T.; Yu, S. S.; Pan, L. K.; Sun, Z. Coulomb Repulsion at the Nanometer-Sized Contact: A Force Driving Superhydrophobicity, Superfluidity, Superlubricity, and Supersolidity. J. Phys. Chem. C 2009, 113, 2000920019. (11) Raviv, U.; Giasson, S.; Kampf, N.; Gohy, J. F.; Jerome, R.; Klein, J. Lubrication by Charged Polymers. Nature 2003, 425, 163165. (12) Chen, M.; Briscoe, W. H.; Armes, S. P.; Klein, J. Lubrication at Physiological Pressures by Polyzwitterionic Brushes. Science 2009, 323, 16981701. (13) Xu, J. G.; Kato, K. Formation of Tribochemical Layer of Ceramics Sliding in Water and Its Role for Low Friction. Wear 2000, 245, 6175. (14) Chen, M.; Kato, K.; Adachi, K. Friction and Wear of Self-Mated SiC and Si3N4 Sliding in Water. Wear 2001, 250, 246255. (15) Ma, Z. Z.; Zhang, C. H.; Luo, J. B.; Lu, X. C.; Wen, S. Z. Superlubricity of a Mixed Aqueous Solution. Chin. Phys. Lett. 2011, 28, 056201-1056201-4. (16) Matta, C.; Joly-Pottuz, L.; Bouchet, M. I. D.; Martin, J. M. Superlubricity and Tribochemistry of Polyhydric Alcohols. Phys. Rev. B 2008, 78, 08543610854368. (17) Li, J. J.; Liu, Y. H.; Luo, J. B.; Liu, P. X.; Zhang, C. H. Excellent Lubricating Behavior of Brasenia schreberi Mucilage. Langmuir 2012, 28, 77977802. (18) Arad, S.; Rapoport, L.; Moshkovich, A.; van Moppes, D.; Karpasas, M.; Golan, R.; Golan, Y. Superior Biolubricant from a Species of Red Microalga. Langmuir 2006, 22, 73137317. (19) Li, J. J.; Zhang, C. H.; Luo, J. B. Superlubricity Behavior with Phosphoric Acid-Water Network Induced by Rubbing. Langmuir 2011, 27, 94139417. (20) Li, J. J.; Zhang, C. H.; Sun, L.; Lu, X.C..; Luo, J. B. Tribochemistry and Superlubricity Induced by Hydrogen Ions. Langmuir 2012, 28, 158156115823. (21) Li, J. J.; Zhang, C. H.; Sun, L.; Luo, J. B. Analysis of Measurement Inaccuracy in Superlubricity Tests. Tribol. Trans. 2012, DOI: doi: 10.1080/10402004.2012.732200. (22) Sverjensky, D. A.; Sahai, N. Theoretical Prediction of Single-Site Surface-Protonation Equilibrium Constants for Oxides and Silicates in Water. Geochim. Cosmochim. Acta 1996, 60, 37733797. (23) Sahai, N. Is Silica Really an Anomalous Oxide? Surface Acidity and Aqueous Hydrolysis Revisited. Environ. Sci. Technol. 2002, 36, 445452. (24) Dowson, D.; Higginson, G. R. Elasto-Hydrodynamic Lubrication; Pergamon Press: Oxford, U.K., 1977. (25) Joseph, G. G.; Hunt, M. L. Oblique Particle-Wall Collisions in a Liquid. J. Fluid Mech. 2004, 510, 7193. (26) Dash, J. G.; Fu, H. Y.; Wettlaufer, J. S. The Premelting of Ice and Its Environmental Consequences. Rep. Prog. Phys. 1995, 58, 115 167.
275

Article

(27) Raviv, U.; Klein, J. Fluidity of Bound Hydration Layers. Science 2002, 297, 15401543.

dx.doi.org/10.1021/la3046115 | Langmuir 2013, 29, 271275