- Email: [email protected]
Effect of coadsorption of water and alcohol vapor on the nanowear of silicon
Wear 332-333 (2015) 879–884
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of coadsorption of water and alcohol vapor on the nanowear of silicon L. Chen a, Y.J. Yang a, H.T. He a, S.H. Kim b, L.M. Qian a,n a b
Tribology Research Institute, National Traction Power Laboratory, Southwest Jiaotong University, Chengdu 610031, China Department of Chemical Engineering and Materials Research Institute, Pennsylvania State University, University Park, PA 16803, USA
art ic l e i nf o
a b s t r a c t
Article history: Received 15 September 2014 Received in revised form 21 January 2015 Accepted 13 February 2015
Using an environment-controlled atomic force microscope, the nanowear of silicon against SiO2 microsphere was investigated under various relative humidity (RH) and ethanol partial pressure (Pa/Psat) conditions. When RH was below 10%, there was no discernable wear of the substrate in any Pa/Psat conditions. However, when RH 410%, although the contact pressure (1.3 GPa) was far less than the yield stress of silicon material (7 GPa), the occurrence of tribochemical wear was observed on silicon surface without ethanol adsorption. When ethanol vapor was introduced, the wear of silicon was inhibited fully or partially depending on RH. At RH below 50%, there was a critical ethanol partial pressure above which tribochemical reaction was completely suppressed and no wear was generated on silicon surface. At RH above 50%, the wear of silicon could be significantly reduced, but not completely, with the increase of ethanol partial pressure. Upon addition of ethanol vapor, the frictional energy dissipation decreased and the energy barrier for tribochemical reaction could also become larger. These factors could explain the lubrication effect of the adsorbate ethanol layer in humid environment. The RH and ethanol partial pressure conditions for the lubrication of Si/SiO2 sliding pairs were identified, which can be used for optimizing the MEMS operation conditions without failure due to wear. & 2015 Elsevier B.V. All rights reserved.
Keywords: Tribochemical wear Silicon Coadsorption Water Alcohol
1. Introduction Microelectromechanical systems (MEMS) with a characteristic length of 100 nm–1 mm have been successfully employed in military and commercial fields [1–3]. Silicon has been widely used as a main construction material of MEMS due to its excellent properties in lithographic microfabrication and micromachining techniques [4,5]. As the dimensions of moving parts shrink to micro/nanometer scale, the surface force of tribological interface plays important roles owing to the large ratio of surface area to volume [6]. In this situation, the dynamic moving parts can fail because of serious tribological problems such as high friction and serious wear, which greatly restrict the reliability of MEMS [7–10]. Therefore, understanding and control of the nanowear of silicon is an important issue of concern. Silicon does not have any intrinsic lubrication properties, so it is subject to significant wear during sliding processes. When the contact pair is chemically inert (such as diamond tip), the mechanical wear of silicon surface usually occurs under high contact pressure [11,12]. However, when the contact pair is chemically reactive (such as Si or SiO2 tip), serious tribochemical wear of silicon may occur in humid air
n
Corresponding author. Tel.: þ 86 28 87600687; fax: þ 86 28 87603142. E-mail address: [email protected] (L.M. Qian).
http://dx.doi.org/10.1016/j.wear.2015.02.052 0043-1648/& 2015 Elsevier B.V. All rights reserved.
under a contact pressure far below its yield stress (7 GPa) [13–15]. Such tribochemical wear was also observed in silicon-based MEMS devices [7–10]. To prevent the wear of silicon, various methods have been attempted including the deposition of hard coating or hydrophobic organic film, such as diamond-like carbon (DLC) coating and self-assembled monolayers (SAMs), and so on [1,16–19]. However, the conformal coating of DLC inside deep gaps or trenches between high asperity objectives is difficult. The poor durability of SAMs also limits their practical applications in MEMS [20]. Recent experimental results indicated that both the mechanical wear and tribochemical wear of silicon material can be effectively prevented in an alcohol vapor environment [21,22]. This method has been successfully used for the lubrication of MEMS devices to extend their lifetime [23,24]. Previously, most studies focused on the alcohol lubrication at water-free conditions. Actually, it is difficult to completely eliminate the water vapor in the dynamic MEMS devices. The existence of water molecules would affect the adsorption state of alcohol layer and could deteriorate the alcohol lubrication [25,26]. In the case of mixed vapor of water and alcohol, the binary adsorbate layers present in a layered structure, in which alcohol is at the adsorbate/vapor interface and water is inside the adsorbate layer [25]. The thickness and structure of the binary layers would change depending on the composition of water and alcohol [25,28]. Through a MEMS side-wall tribometer, Barnette et al. [26] investigated the effect of water vapor on the
880
L. Chen et al. / Wear 332-333 (2015) 879–884
alcohol lubrication at a fixed alcohol partial pressure (Pa/ Psat ¼50%). They found that the full coverage of alcohol layer could effectively prevent the silicon wear. However, the increase of relative humidity (RH) reduced the amount of alcohol adsorbed at the interface and eventually led to the failure of MEMS device operation at RH above 30% [26]. It was reported that the coverage of adsorbate alcohol layer would be less than monolayer on silicon surface at relatively low Pa/Psat [27,28]. Therefore, it is important to verify whether the partial coverage of adsorbate alcohol layer can prevent the wear of silicon under relatively low RH and Pa/Psat conditions. Moreover, to better achieve the lubrication of siliconbased MEMS, it is essential to investigate the effect of water vapor on the alcohol vapor lubrication at various partial pressures. In this paper, the wear behavior of single-crystalline silicon rubbed against a SiO2 microsphere was investigated at various RH and ethanol partial pressure conditions. It was found that the nanowear of silicon strongly depended on the coadsorption of water and ethanol vapors. At RH below 50%, the silicon wear could be completely prevented depending on the ethanol partial pressure (Pa/Psat). However, when RH was higher than 50%, the surface damage on silicon surface was only partially suppressed under all ethanol partial pressure conditions tested in this study. The results suggest that the ethanol vapor lubrication is suitable as a protecting method in MEMS applications at RH r50% and relatively high Pa/Psat.
2. Material and methods The p-doped Si(100) wafers were purchased from MEMC Electronic Materials, Inc. By using an atomic force microscope (AFM,
SPI3800N, Seiko, Japan), the root-mean-square (RMS) roughness of the Si(100) wafers were measured to be 0.0670.01 nm over a 500 nm 500 nm area. The native oxide layer was retained on the silicon surface to simulate the real contact surface in MEMS, whose thickness was determined by Auger Electron Spectroscopy (AES) as 0.5 nm [29]. Before wear tests, the silicon wafers were ultrasonically cleaned in methanol, ethanol, and deionized water for 10 min in turn. As shown in Fig. 1, all the wear tests and topography imaging were performed by the AFM equipped with an atmosphere chamber connected an external vapor control system. Environments mixed by water and ethanol vapors were realized by a vapor control system (dashed box in Fig. 1). Before filled with mixture vapor, the AFM chamber was pumped down to 5 10 4 Torr to achieve low moisture condition. The N2 lines passed through two bottles filled with water/glass beads and ethanol/glass beads, and finally joined into a single inlet line. The partial pressures relative to the saturation pressure (P/Psat) of water or ethanol could be controlled by varying the flow rates of dry, water-saturated, and ethanol-saturated N2 gas streams through this system. More details about this method have been described in previous literature [30]. The wear tests were operated by using SiO2 spherical tips, which were purchased from Novascan Technologies, USA. Each SiO2 tip consisted of a SiO2 microsphere with a radius of 1 mm (inset in Fig. 1) attached to a silicon cantilever whose spring constant was calibrated to be in the range of 10.5–13.8 N/m [31]. During the wear tests, the sliding speed was 1 mm/s, the sliding stroke (D) was 500 nm, the applied normal load (Fn) was 3 μN and the number of sliding cycles (N) was 200. After the tests, the topography of the wear area of the silicon surface was imaged using a sharp silicon nitride probe (MLCT, Veeco, USA), which had a tip radius of 10–20 nm and a nominal spring constant of 0.1–0.3 N/m. The coadsorption of ethanol and water was measured with a Thermo Nicolet Nexus 670 infrared spectrometer with an attenuated total reflectance infrared spectroscopy (ATR-IR) setup; more details of the experiments could be found elsewhere [25,28].
3. Results 3.1. Coadsorption of water and ethanol on silicon surface
Fig. 1. The schematic illustration showing the wear tests operated in AFM with an atmosphere chamber (left part) connected an external vapor control system (right part). A SiO2 microsphere (scanning electron microscope image shown in the inset) moves horizontally on the silicon surface with a stroke D ¼ 500 nm under a normal load Fn ¼3 mN. With the atmosphere controlling system, the relative humidity can be controlled in 0–80% and the Pa/Psat in AFM chamber can be controlled in 0–100% with the error of 72%.
0.4
0.4 Approach Retraction
0.4
Approach Retraction
0.0
-0.2
0.0
-0.2
"dragged" by alcohol layer
-0.4
-0.4 0
50
100
150
Z-Piezo distance (nm)
200
0
50
100
150
Z-Piezo distance (nm)
Approach Retraction
0.2
Force (μN)
0.2
Force (μ N)
0.2
Force (μN)
The main driving force for adsorption of ethanol molecules to the silicon surface is the formation of hydrogen bonds between the surface hydroxyl and ethanol OH groups [27]. In the conditions of water and alcohol mixture vapor, the hydroxyl group of alcohol is hydrogen bonded to the water molecule underneath. The adsorbed layer of water and alcohol assumed a binary layers structure [25,28]. To confirm the coadsorption of water and ethanol on silicon surface,
200
0.0
-0.2
-0.4 0
50
100
150
Z-Piezo distance (nm)
Fig. 2. Force–distance curves of Si(100)/SiO2 pair at RH ¼30%. (a) Pa/Psat ¼ 0%, (b) Pa/Psat ¼ 20%, (c) Pa/Psat ¼ 40%.
200
L. Chen et al. / Wear 332-333 (2015) 879–884
the force–distance curves were measured at various atmosphere conditions. Fig. 2 shows the typical force–distance curves measured at 30% RH. In humid nitrogen without ethanol, the SiO2 tip separated instantaneously from the silicon surface, as shown in Fig. 2a. However, when ethanol vapor was introduced, the force–distance curves showed different behavior, as shown in Fig. 2b and c. During the approach process, the meniscus could be formed before the tip touched the substrate surface even at a distance of 5–20 nm from the surface (dotted line). During the retraction process, the meniscus was stretched due to the disjoining pressure (solid line) [32]. The difference between the force–distance curves at various RH and Pa/Psat should be attributed to the adsorbate ethanol layer formed on the outmost surface. The similar phenomenon has been observed on other films adsorbed on silicon surface, such as DNA film and polymer film [33,34]. Fig. 3 displays the ATR-IR spectra obtained at various RH and Pa/Psat conditions. The peaks centered at 3230 cm 1 and 3400 cm 1 correspond to the OH stretching vibrations of ice-like water and liquidlike water, respectively [35]. The peak at 2960 cm 1 denoted the ethyl stretching vibration of ethanol [25,28]. With the decrease of RH
Log(1/R)
Log(1/R)
Mostly due to H2O
0.003 0.002
Ethyl
0.001 0.000
and increase of Pa/Psat, the intensities of the OH stretching vibration decreased and the ethyl stretching vibrations increased. The ATR-IR measurements supported that the binary adsorbate layers of water and ethanol formed on the silicon substrate surface or SiO2 tip surface, and their thickness strongly depended on the composition of water and ethanol vapors. The results were also in accord with the measurement of sum frequency generation (SFG) vibration spectroscopy [25]. 3.2. Wear behavior of silicon under various RH and Pa/Psat conditions Fig. 4 shows the AFM images of wear scars on silicon surfaces after scratching with a SiO2 sphere tip for 200 reciprocating cycles at selected relative humidity (RH¼10%, 15%, 30% and 50%) and ethanol partial pressure (Pa/Psat ¼0%, 10%, 20%, 35% and 40%) conditions. It was found that the coadsorption of water and ethanol vapor showed a strong effect on the nanowear of silicon surface. At all RH conditions, the ethanol vapor could prevent or reduce the wear of silicon depending on its partial pressure. At 10% RH, the tribochemical wear causing material removal was observed on silicon surface without ethanol vapor (Pa/Psat ¼0%). With the increase of Pa/Psat, no discernible
0.004
0.004
0.003
0.003
OH of H2O ðanol
0.002
Log(1/R)
0.004
881
Ethyl
0.001
0.002
0.000 -0.001
-0.001
3600
3400
3200
Wavenumber
3000
3600
2800
cm-1
Ethyl
0.001
0.000
-0.001
OH of ethanol
3400
3200
Wavenumber
3000
2800
3600
cm-1
3400
3200
Wavenumber
3000
2800
cm-1
Fig. 3. ATR-IR spectra of the binary adsorbate layers of alcohol and water at the conditions of (a) Pa/Psat ¼ 5%/RH¼ 80%, (b) Pa/Psat ¼ 64%/RH ¼24% and (c) Pa/Psat ¼ 85%/RH ¼0%.
Pa/Psat = 0%
10%
20%
35%
40%
RH 10%
15%
30%
50% 400 nm Fig. 4. AFM images of wear scars on silicon surface after scratched by a SiO2 tip at selected RH and ethanol partial pressure (Pa/Psat) conditions. (a) RH ¼ 10%, (b) RH ¼ 15%, (c) RH ¼ 30%, (d) RH¼ 50%. Note the height full-scales for the images are 15 nm. Fn ¼ 3 μN, N ¼ 200 and D ¼500 nm.
882
L. Chen et al. / Wear 332-333 (2015) 879–884
12 RH 0% 10% 20% 35% 50%
10 8 6 4 2 0
RH
Wear depth d (nm)
Wear depth d (nm)
12
10
60% 70% 80%
8 6 4 2 0
0
10
20
30
40
50
60
Alcohol partial pressure Pa/Psat (%)
0 10 20 30 40 50 60 Alcohol partial pressure Pa/Psat (%)
3.3. Nanowear map of silicon surface under various RH and Pa/Psat conditions To construct a nanowear map showing the vapor condition dependence, the entire data of wear depth measured over wide ranges of RH and ethanol partial pressures were plotted in Fig. 6. At RHo10%, no wear was detected at all conditions. When RH increases from 10% to 50%, the wearless behavior of silicon could be achieved when the ethanol partial pressure was larger than the critical value. When RH was larger than 50%, the ethanol vapor lubrication was not sufficient to completely suppress silicon wear under all Pa/Psat conditions tested in this study. It was also noted that the most serious wear region located at the condition of RH ranging from 50% to 70% and Pa/Psat below 10%. In summary, the experimental results indicated that the ethanol lubrication can completely prevent the wear of the dynamic silicon parts in Si-based MEMS at relatively low RH (r50%). At high RH, the wear of silicon can only be partly prevented. It should be noted that due to the limitation of our gas flow system (Fig. 1); when the ethanol vapor pressure is increased to near-saturation, the lubrication effect is expected to further improve even at this high RH conditions.
100 (%)
Wear depth
80
0
P /P
surface damage of silicon was detected at Pa/Psat 410%. The Pa/Psat dependence of the silicon wear at RH of 15% and 30% was similar to the case at RH of 10%. There seems to be a critical ethanol partial pressure at each RH condition, below which silicon surface wore severely and above which there was no wear visible on silicon surface. As shown in Fig. 4b and c, the critical value of Pa/Psat can be identified as 20% at a RH of 15% and 40% at a RH of 30%. When the relative humidity increased to 50%, the wear of silicon resulted in grooves at Pa/Psat ranging from 0% to 40%, as shown in Fig. 4d. To quantitatively show the effect of the coadsorption of water and ethanol vapor on the wear behavior of silicon, the maximum wear depths of silicon surface under various RH and ethanol partial pressure conditions are shown in Fig. 5. Fig. 5a plots the data measured at low RH region (RH r50%). Since the tribochemical reaction of Si/SiO2 pair could not occur without water vapor [13], no surface damage was observed in dry air (RH¼0%). When the tests were performed at low RH region (10%rRHr50%), the wear depth gradually decreased below the detection limit at a critical ethanol partial pressure. Fig. 5b shows the wear depth of silicon measured at high RH region (RHZ60%). With the increase of ethanol partial pressure, the wear depth revealed the similar trend as that at low RH conditions, but it never decreased below the detection limit. For example, at 60% RH, the wear depth of silicon surface was reduced by 55% when Pa/Psat increased from 0% to 35%.
Alcohol partial pressure P /P (%)
Fig. 5. The maximum depth of wear scars on silicon surface plotted as the function of ethanol partial pressure (Pa/Psat). (a) RHr 50%, the wear depth increases with the increase in RH at given ethanol partial pressure; (b) RH Z60%, the wear depth decreases with the increase in RH. Fn ¼ 3 μN, N ¼ 200 and D ¼500 nm.
60
Wearless region
40 20
Serious wear 0 0
20
40
60
80
10 [nm]
Relative humidity RH (%)
Fig. 6. Nanowear map illustrating the wear depth of silicon surface at various RH and ethanol partial pressure (Pa/Psat) conditions. The blue area shows the wearless region of coadsorption conditions. The inset shows 75 environment conditions where data were collected.
4. Discussion 4.1. Lubrication of adsorbate ethanol layer Under the given loading conditions, the maximum contact pressure in our AFM tests is estimated to be 1.3 GPa based on the DMT model [36]. Since it is much lower than the yield stress of silicon (7 GPa) [37], the mechanical wear is not significant; thus surface wear must occur mainly through tribochemical reactions in humidity air [8,38]. During the shear process, the Si–O–Si bonding bridges may form between the two contact surfaces of Si(100)/SiO2 pair, which can further transfer the mechanical shear stress to the substrate surface and induce the dissociation of the interfacial bridge bonds or the subsurface Si–O bonds of silicon [13,39,40]. This tribochemical wear of silicon occurs only with the existence of water molecules [13]. When the alcohol vapor was introduced into the AFM chamber, the adsorbed alcohol molecules could form a lubrication layer and protect the silicon substrate [21–25]. To detect the lubrication of adsorbate alcohol layer to the sliding interfaces, the friction force of the Si(100)/SiO2 pair was measured at various RH and Pa/Psat conditions, as shown in Fig. 7. In relatively high RH without any ethanol vapor, friction force initially increased to a maximum value and then decreased with the increase of sliding cycles. This variation was accompanied by the silicon wear (Figs. 4–6), which implies some chemical and topographical changes of the sliding interface. The chemical change must include the loss of surface hydroxyl groups as well as native oxide layers [15]. Under relatively lower RH and higher Pa/Psat of ethanol conditions, the friction force was steady and low during the entire reciprocating cycles and no
L. Chen et al. / Wear 332-333 (2015) 879–884
883
2
RH=10%
Friction force Ft (µN)
1
2 RH=30% 1
3 -12 Wear rate γ (x10 m /Nm)
0.6 10% 30% 50% 70%
0.4
Wearless region 0.2
0.0 40
60
80
100 120 140 160 180 200 -12
Dissipated energy E (x 10
2
RH=50%
1 0
1
40 15 10 0 Cycle 100 (%) t sN P a/P sa
Fig. 7. Friction force vs. number of reciprocating cycles (Ft–N) curves at selected ethanol partial pressure (Pa/Psat ¼0%, 15% and 40%) and relative humidity RH ¼ (a) 10%, (b) 30% and (c) 50% conditions. The error of Ft estimated from five dependent measurements is less than 20%.
discernable wear was formed on silicon surface (Figs. 4–6). The low friction force and wearless state should be attributed to the lubrication of the adsorbate ethanol layer, which lowered the surface energy at the adsorbate/vapor interface [25,28]. The low friction force at high Pa/Psat will reduce the dissipated energy during sliding process [41]. Since the tribochemical reaction of Si/SiO2 pair is caused by the interfacial shear, its consequence (wear) must be correlated with the amount of energy dissipated through friction [22]. Fig. 8 shows the variation of wear rate on silicon substrate as a function of dissipated energy (integrated by friction force over total 200 reciprocating cycles). It was found that regardless of the variation of RH and Pa/Psat, there exists a threshold energy of 0.7 10 10 J, below which tribochemical wear does occur (gray region in Fig. 8) and above which the wear rate of silicon increases linearly with the dissipated energy. It is hypothesized that the mechanochemical reactions involve formation and dissociation of the Si–O–Si bridge bonds. The ethanol adsorbate can not only prevent the formation of Sisubstrate–O–Sitip bonded bridges during rubbing process, but also increase the energy barrier of Si–O–Si bonds dissociation on silicon surface [22]. Since the adsorption of longer chain molecule may further increase the energy barrier of Si–O–Si bonds dissociation on silicon substrate, the propanol or butanol vapor may have a better capability to resist the tribochemical wear of silicon surface [22,42,43]. 4.2. Resistance mechanism at low RH and Pa/Psat As shown in Figs. 4–6, the resistance of adsorbate ethanol layer to the tribochemical wear strongly depends on RH and Pa/Psat. The ATRIR measurement in Fig. 3a indicated that the ice-like water adsorbed on silicon surface was still prominent at high RH condition as the ethanol vapor was introduced. Compared with liquid-like water, the ice-like water with strong hydrogen bonding structure played a more dominated role in the tribochemical wear of silicon surface [13]. This might be the reason that silicon wear was just partially resisted at RHZ60% (Fig. 5b). When RH was lower than 60%, the tribochemical wear of silicon was completely suppressed when the Pa/Psat increased above a critical value. (Figs. 4–6). It was reported that the alcohol vapor could successfully prevent the silicon wear when the full
J)
Fig. 8. Correlation between the average wear rate and the total dissipated energy during 200 sliding cycles at RH of 10%, 30%, 50% and 70%. In left gray region, wear rate of silicon substrate is zero under relatively low dissipated energy.
SiO2 tip Silicon
Alcohol layer Water layer
SiO2 tip Silicon
Fig. 9. The schematics of the sphere-adsorbate-substrate contact during the sliding process in pure humidity air (a) and at relatively low Pa/Psat (b).
coverage of alcohol layer was formed on substrate surface [21,25]. In that case, due to the lubrication of adsorbate alcohol layer, the friction coefficient would be consistent with 0.2 whether the tests were operated in AFM and MEMS devices at nano/micro-scale or in pin-on-disc tribometers at macro-scale [21–25,44,45]. However, the tribochemical wear of silicon was found to be completely suppressed at relatively low Pa/Psat (o30%) where a partial coverage of ethanol adsorbate layer was expected on silicon surface (For example: Fig. 4a) [25,28]. Based on the friction results in Fig. 7a and b, the friction coefficient at Pa/Psat of 30% was estimated to be about 0.05 and 0.12 at a RH of 10% and 30%, respectively. The low friction coefficient also supported that the full coverage of ethanol adsorbate layer did not form under these conditions. Fig. 9 shows a schematic presentation of the sphere-adsorbatesubstrate contact in pure humid air and at relatively low Pa/Psat (o30%). Under high RH without alcohol, the tribochemical reaction caused serious wear of silicon [13]. However, no obvious surface damage of silicon was observed when RH was less than 10% (Fig. 6). It indicated that the tribochemical wear occurred when the amount of adsorbate water molecules were large enough to form water bridge between sliding interfaces, as shown in Fig. 9a. When the mixture vapor of water and alcohol was introduced, the hydroxyl groups of the water molecules are attached to silicon surface through hydrogen bonding and the alcohol molecules are hydrogen bonded to the water molecules forming an overlayer [25,26,28]. At relatively low Pa/ Psat (o 30%), the alcohol layer would partially cover the outmost surface, as shown in Fig. 9b [25–28]. The alcohol molecules on the top of the water layer can prevent (Fig. 4a and b) or at least significantly reduce (Fig. 4c and d) the water-mediated bridge formation across the sliding solid interfaces [28].
5. Conclusion Using an environment-controlled AFM, the effects of RH and ethanol partial pressure on the lubrication of Si/SiO2 sliding pairs were studied. The main conclusions can be summarized as follows:
884
L. Chen et al. / Wear 332-333 (2015) 879–884
(1) At RH below 50%, the tribochemical wear of silicon can be completely suppressed when the ethanol partial pressure increases above a critical value. As RH increases, the critical ethanol pressure needed for lubrication increases as well. (2) The co-adsorbed ethanol molecules decreased the frictional energy dissipation, which reduced tribochemical reaction probability. (3) The ethanol molecules on top of the water layer can prevent or at least significantly reduce the water-mediated bridge formation across the sliding solid interfaces.
Acknowledgments The authors are grateful for the financial support from the National Natural Science Foundation of China (91323302, 51375409, and 51175441). References [1] B. Bhushan, Nanotribology and nanomechanics of MEMS/NEMS and BioMEMS/ BioNEMS materials and devices, Microelectron. Eng. 84 (2007) 387–412. [2] M. Elwenspoeck, R. Wiegerink, Mechanical Microsensors, Springer, Berlin, 2001. [3] Z. Cui, Micro-nanofabrication Technologies and Application, Higher Education Press, Beijing, 2005. [4] M.A. Meitl, Z.T. Zhu, V. Kumar, K.J. Lee, X. Feng, Y.Y. Huang, I. Aadesida, R. G. Nuzzo, J.A. Rogers, Transfer printing by kinetic control of adhesion to an elastomeric stamp, Nat. Mater. 5 (2006) 33–38. [5] S. Demiri, S. Boedo, L.S. Holsen, Wear characteristics of large aspect ratio silicon microbearing systems, Wear 312 (2014) 58–69. [6] J.A. Williams, H.R. Le, Tribology and MEMS, J. Phys. D Appl. Phys 39 (2006) 201–214. [7] B. Bhushan, Tribology of industrial components and systems, in: B. Bhushan (Ed.), Modern Tribology Handbook, CRC Press, Florida, 2001. [8] K. Mizuhara, S.M. Hsu, Tribochemical reaction of oxygen and water on silicon surfaces, Tribol. Ser. 21 (1992) 323–328. [9] S. Shen, Y.G. Meng, A novel running-in method for improving life-time of bulkfabricated silicon mems devices, Tribol. Lett. 47 (2012) 273–284. [10] I.S.Y. Ku, T. Reddyhoff, A.S. Holmes, H.A. Spikes, Wear of silicon surfaces in MEMS, Wear 271 (2011) 1050–1058. [11] B.J. Yu, H.S. Dong, L.M. Qian, Y.F. Chen, J.X. Yu, Z.R. Zhou, Friction-induced nanofabrication on monocrystalline silicon, Nanotechnology 20 (2009) 465303. [12] B. Bhushan, J.N. Israelachvili, U. Landman, Nanotribology: friction, wear and lubrication at the atomic scale, Nature 374 (1995) 607–616. [13] J.X. Yu, S.H. Kim, B.j. Yu, L.M. Qian, Z.R. Zhou, Role of tribochemistry in nanowear of single-crystalline ailicon, ACS Appl. Mater. Interfaces 4 (2012) 1585–1593. [14] J.X. Yu, L. Chen, L.M. Qian, D.L. Song, Y. Cai, Investigation of humiditydependent nanotribology behaviors of Si(100)/SiO2 pair moving from stick to slip, Appl. Surf. Sci. 265 (2013) 192–200. [15] L. Chen, S.H. Kim, X.D. Wang, L.M. Qian, Running-in process of Si–SiOx/SiO2 pair at nanoscale – sharp drop of friction and wear rate during initial cycles, Friction 1 (2013) 81–91. [16] L. Chen, M.C. Yang, J.X. Yu, L.M. Qian, Z.R. Zhou, Nanofretting behaviours of ultrathin DLC coating on Si(100) substrate, Wear 271 (2011) 1980–1986. [17] R.W. Carpick, M. Salmeron, Scratching the surface: fundamental investigations of tribology with atomic force microscopy, Chem. Rev. 97 (1997) 1163–1194. [18] A. Erdemir, C. Donnet, Tribology of diamond-like carbon films: recent progress and future prospects, J. Phys. D: Appl. Phys. 39 (2006) 311–327. [19] L. Chen, M.C. Yang, C.F. Song, B.J. Yu, L.M. Qian, Is 2 nm DLC coating enough to resist the nanowear of silicon, Wear 302 (2013) 909–917. [20] S.H. Kim, D.B. Asay, M.T. Dugger, Nanotribology and MEMS, Nano Today 2 (2007) 22–29.
[21] D.J. Marchand, L. Chen, Y.G. Meng, L.M. Qian, S.H. Kim, Effects of vapor environment and counter-surface chemistry on tribochemical wear of silicon wafers, Tribol. Lett. 53 (2014) 365–372. [22] A.L. Barnette, D.B. Asay, D. Kim, B.D. Guyer, H. Lim, M.J. Janik, S.H. Kim, Experimental and density functional theory study of the tribochemical wear behavior of SiO2 in humid and alcohol vapor environments, Langmuir 25 (2009) 13052–13061. [23] D.B. Asay, M.T. Dugger, S.H. Kim, In-situ vapor-phase lubrication of MEMS, Tribol. Lett. 29 (2008) 67–74. [24] D.B. Asay, M.T. Dugger, J.A. Ohlhausen, S.H. Kim, Macro- to nanoscale wear prevention via molecular adsorption, Langmuir 24 (2007) 155–159. [25] A.L. Barnette, S.H. Kim, Coadsorption of n-propanol and water on SiO2: study of thickness, composition, and structure of binary adsorbate layer using attenuated total reflection infrared (ATR-IR) and sum frequency generation (SFG) vibration spectroscopy, J. Phys. Chem. C 116 (2012) 9909–9916. [26] A.L. Barnette, J.A. Ohlhausen, M.T. Dugger, S.H. Kim, Humidity effects on in situ vapor phase lubrication with n-pentanol, Tribol. Lett. 55 (2014) 177–186. [27] A.L. Barnette, D.B. Asay, M.J. Janik, S.H. Kim, Adsorption isotherm and orientation of alcohols on hydrophilic SiO2 under ambient conditions, J. Phys. Chem. C 113 (2009) 10632–10641. [28] A.L. Barnette, S.H. Kim, Attenuated total reflection infrared spectroscopy study of hysteresis of water and n-alcohol co-adsorption on silicon oxide, Langmuir 28 (2012) 15529–15536. [29] B.J. Yu, L.M. Qian, H.S. Dong, J.X. Yu, Z.R. Zhou, Friction-induced hillocks on monocrystalline silicon in atmosphere and in vacuum, Wear 268 (2010) 1095–1102. [30] K. Strawhecker, D.B. Asay, J. McKinney, S.H. Kim, Reduction of adhesion and friction of silicon oxide surface in the presence of n-propanol vapor in the gas phase, Tribol. Lett. 19 (2005) 17–21. [31] A. Torii, M. Sasaki, K. Hane, S. Okuma, A method for determining the spring constant of cantilevers for atomic force microscopy, Meas. Sci. Technol. 7 (1996) 179–184. [32] E. Hsiao, B.D. Veres, G.J. Tudryn, S.H. Kim, Identification of mobile species in cationic polymer lubricant layer on silicon oxide from afm and xps analyses, Langmuir 27 (2011) 6808–6813. [33] M. Wang, S.X. Cui, B.J. Yu, L.M. Qian, Effect of chain length on the conformation and friction behaviour of DNA, Sci. China Technol. Sci. 56 (2013) 2927–2933. [34] H.J. Butt, M. Kappl, H. Mueller, R. Raiteri, Steric forces measured with the atomic force microscope at various temperatures, Langmuir 15 (1999) 2559–2565. [35] D.B. Asay, S.H. Kim, Evolution of the adsorbed water layer structure on silicon oxide at room temperature, J. Phys. Chem. B 109 (2005) 16760–16763. [36] U.D. Schwarz, A generalized analytical model for the elastic deformation of an adhesive contact between a sphere and a flat surface, J. Colloid Interface Sci. 261 (2003) 99–106. [37] K.E. Petersen, Silicon as a mechanical material, Proc. of the IEEE 70 (1982) 420–457. [38] L. Chen, H.T. He, X.D. Wang, S.H. Kim, L.M. Qian, Tribology of Si/SiO2 in humid air: transition from severe chemical wear to wearless behavior at nanoscale, Langmuir 31 (2015) 149–156. [39] F. Katsuki, Single asperity tribochemical wear of silicon by atomic force microscopy, J. Mater. Res. 24 (2009) 173–178. [40] W. Maw, F. Stevens, S.C. Langford, J.T. Dickinson, Single asperity tribochemical wear of silicon nitride studied by atomic force microscopy, J. Appl. Phys. 92 (2002) 5103–5109. [41] S. Yashima, N. Takase, T. Kurokawa, J.P. Gong, Friction of hydrogels with controlled surface roughness on solid flat substrates, Soft Matter 10 (2014) 3192–3199. [42] A.J. Barthel, S.H. Kim, Lubrication by physisorbed molecules in equilibrium with vapor at ambient condition – effects of molecular structure and substrate chemistry, Langmuir 30 (2014) 6469–6478. [43] A.J. Barthel, A. Al-Azizi, N.D. Surdyka, S.H. Kim, Effects of gas or vapor adsorption on adhesion, friction, and wear of solid interfaces, Langmuir 30 (2014) 2977–2992. [44] D.B. Asay, E. Hsiao, S.H. Kim, Effects of adsorbate coverage and capillary on nano-asperity friction in atmosphere containing organic vapor, J. Appl. Phys. 110 (2011). [45] M.J. Marino, E. Hsiao, L.C. Bradley, O.L. Eryilmaz, A. Erdemir, S.H. Kim, Is ultralow friction needed to prevent wear of diamond-like carbon (DLC)? an alcohol vapor lubrication study for stainless steel/DLC interface, Tribol. Lett. 42 (2011) 285–291.
Contents lists available at ScienceDirect
Wear journal homepage: www.elsevier.com/locate/wear
Effect of coadsorption of water and alcohol vapor on the nanowear of silicon L. Chen a, Y.J. Yang a, H.T. He a, S.H. Kim b, L.M. Qian a,n a b
Tribology Research Institute, National Traction Power Laboratory, Southwest Jiaotong University, Chengdu 610031, China Department of Chemical Engineering and Materials Research Institute, Pennsylvania State University, University Park, PA 16803, USA
art ic l e i nf o
a b s t r a c t
Article history: Received 15 September 2014 Received in revised form 21 January 2015 Accepted 13 February 2015
Using an environment-controlled atomic force microscope, the nanowear of silicon against SiO2 microsphere was investigated under various relative humidity (RH) and ethanol partial pressure (Pa/Psat) conditions. When RH was below 10%, there was no discernable wear of the substrate in any Pa/Psat conditions. However, when RH 410%, although the contact pressure (1.3 GPa) was far less than the yield stress of silicon material (7 GPa), the occurrence of tribochemical wear was observed on silicon surface without ethanol adsorption. When ethanol vapor was introduced, the wear of silicon was inhibited fully or partially depending on RH. At RH below 50%, there was a critical ethanol partial pressure above which tribochemical reaction was completely suppressed and no wear was generated on silicon surface. At RH above 50%, the wear of silicon could be significantly reduced, but not completely, with the increase of ethanol partial pressure. Upon addition of ethanol vapor, the frictional energy dissipation decreased and the energy barrier for tribochemical reaction could also become larger. These factors could explain the lubrication effect of the adsorbate ethanol layer in humid environment. The RH and ethanol partial pressure conditions for the lubrication of Si/SiO2 sliding pairs were identified, which can be used for optimizing the MEMS operation conditions without failure due to wear. & 2015 Elsevier B.V. All rights reserved.
Keywords: Tribochemical wear Silicon Coadsorption Water Alcohol
1. Introduction Microelectromechanical systems (MEMS) with a characteristic length of 100 nm–1 mm have been successfully employed in military and commercial fields [1–3]. Silicon has been widely used as a main construction material of MEMS due to its excellent properties in lithographic microfabrication and micromachining techniques [4,5]. As the dimensions of moving parts shrink to micro/nanometer scale, the surface force of tribological interface plays important roles owing to the large ratio of surface area to volume [6]. In this situation, the dynamic moving parts can fail because of serious tribological problems such as high friction and serious wear, which greatly restrict the reliability of MEMS [7–10]. Therefore, understanding and control of the nanowear of silicon is an important issue of concern. Silicon does not have any intrinsic lubrication properties, so it is subject to significant wear during sliding processes. When the contact pair is chemically inert (such as diamond tip), the mechanical wear of silicon surface usually occurs under high contact pressure [11,12]. However, when the contact pair is chemically reactive (such as Si or SiO2 tip), serious tribochemical wear of silicon may occur in humid air
n
Corresponding author. Tel.: þ 86 28 87600687; fax: þ 86 28 87603142. E-mail address: [email protected] (L.M. Qian).
http://dx.doi.org/10.1016/j.wear.2015.02.052 0043-1648/& 2015 Elsevier B.V. All rights reserved.
under a contact pressure far below its yield stress (7 GPa) [13–15]. Such tribochemical wear was also observed in silicon-based MEMS devices [7–10]. To prevent the wear of silicon, various methods have been attempted including the deposition of hard coating or hydrophobic organic film, such as diamond-like carbon (DLC) coating and self-assembled monolayers (SAMs), and so on [1,16–19]. However, the conformal coating of DLC inside deep gaps or trenches between high asperity objectives is difficult. The poor durability of SAMs also limits their practical applications in MEMS [20]. Recent experimental results indicated that both the mechanical wear and tribochemical wear of silicon material can be effectively prevented in an alcohol vapor environment [21,22]. This method has been successfully used for the lubrication of MEMS devices to extend their lifetime [23,24]. Previously, most studies focused on the alcohol lubrication at water-free conditions. Actually, it is difficult to completely eliminate the water vapor in the dynamic MEMS devices. The existence of water molecules would affect the adsorption state of alcohol layer and could deteriorate the alcohol lubrication [25,26]. In the case of mixed vapor of water and alcohol, the binary adsorbate layers present in a layered structure, in which alcohol is at the adsorbate/vapor interface and water is inside the adsorbate layer [25]. The thickness and structure of the binary layers would change depending on the composition of water and alcohol [25,28]. Through a MEMS side-wall tribometer, Barnette et al. [26] investigated the effect of water vapor on the
880
L. Chen et al. / Wear 332-333 (2015) 879–884
alcohol lubrication at a fixed alcohol partial pressure (Pa/ Psat ¼50%). They found that the full coverage of alcohol layer could effectively prevent the silicon wear. However, the increase of relative humidity (RH) reduced the amount of alcohol adsorbed at the interface and eventually led to the failure of MEMS device operation at RH above 30% [26]. It was reported that the coverage of adsorbate alcohol layer would be less than monolayer on silicon surface at relatively low Pa/Psat [27,28]. Therefore, it is important to verify whether the partial coverage of adsorbate alcohol layer can prevent the wear of silicon under relatively low RH and Pa/Psat conditions. Moreover, to better achieve the lubrication of siliconbased MEMS, it is essential to investigate the effect of water vapor on the alcohol vapor lubrication at various partial pressures. In this paper, the wear behavior of single-crystalline silicon rubbed against a SiO2 microsphere was investigated at various RH and ethanol partial pressure conditions. It was found that the nanowear of silicon strongly depended on the coadsorption of water and ethanol vapors. At RH below 50%, the silicon wear could be completely prevented depending on the ethanol partial pressure (Pa/Psat). However, when RH was higher than 50%, the surface damage on silicon surface was only partially suppressed under all ethanol partial pressure conditions tested in this study. The results suggest that the ethanol vapor lubrication is suitable as a protecting method in MEMS applications at RH r50% and relatively high Pa/Psat.
2. Material and methods The p-doped Si(100) wafers were purchased from MEMC Electronic Materials, Inc. By using an atomic force microscope (AFM,
SPI3800N, Seiko, Japan), the root-mean-square (RMS) roughness of the Si(100) wafers were measured to be 0.0670.01 nm over a 500 nm 500 nm area. The native oxide layer was retained on the silicon surface to simulate the real contact surface in MEMS, whose thickness was determined by Auger Electron Spectroscopy (AES) as 0.5 nm [29]. Before wear tests, the silicon wafers were ultrasonically cleaned in methanol, ethanol, and deionized water for 10 min in turn. As shown in Fig. 1, all the wear tests and topography imaging were performed by the AFM equipped with an atmosphere chamber connected an external vapor control system. Environments mixed by water and ethanol vapors were realized by a vapor control system (dashed box in Fig. 1). Before filled with mixture vapor, the AFM chamber was pumped down to 5 10 4 Torr to achieve low moisture condition. The N2 lines passed through two bottles filled with water/glass beads and ethanol/glass beads, and finally joined into a single inlet line. The partial pressures relative to the saturation pressure (P/Psat) of water or ethanol could be controlled by varying the flow rates of dry, water-saturated, and ethanol-saturated N2 gas streams through this system. More details about this method have been described in previous literature [30]. The wear tests were operated by using SiO2 spherical tips, which were purchased from Novascan Technologies, USA. Each SiO2 tip consisted of a SiO2 microsphere with a radius of 1 mm (inset in Fig. 1) attached to a silicon cantilever whose spring constant was calibrated to be in the range of 10.5–13.8 N/m [31]. During the wear tests, the sliding speed was 1 mm/s, the sliding stroke (D) was 500 nm, the applied normal load (Fn) was 3 μN and the number of sliding cycles (N) was 200. After the tests, the topography of the wear area of the silicon surface was imaged using a sharp silicon nitride probe (MLCT, Veeco, USA), which had a tip radius of 10–20 nm and a nominal spring constant of 0.1–0.3 N/m. The coadsorption of ethanol and water was measured with a Thermo Nicolet Nexus 670 infrared spectrometer with an attenuated total reflectance infrared spectroscopy (ATR-IR) setup; more details of the experiments could be found elsewhere [25,28].
3. Results 3.1. Coadsorption of water and ethanol on silicon surface
Fig. 1. The schematic illustration showing the wear tests operated in AFM with an atmosphere chamber (left part) connected an external vapor control system (right part). A SiO2 microsphere (scanning electron microscope image shown in the inset) moves horizontally on the silicon surface with a stroke D ¼ 500 nm under a normal load Fn ¼3 mN. With the atmosphere controlling system, the relative humidity can be controlled in 0–80% and the Pa/Psat in AFM chamber can be controlled in 0–100% with the error of 72%.
0.4
0.4 Approach Retraction
0.4
Approach Retraction
0.0
-0.2
0.0
-0.2
"dragged" by alcohol layer
-0.4
-0.4 0
50
100
150
Z-Piezo distance (nm)
200
0
50
100
150
Z-Piezo distance (nm)
Approach Retraction
0.2
Force (μN)
0.2
Force (μ N)
0.2
Force (μN)
The main driving force for adsorption of ethanol molecules to the silicon surface is the formation of hydrogen bonds between the surface hydroxyl and ethanol OH groups [27]. In the conditions of water and alcohol mixture vapor, the hydroxyl group of alcohol is hydrogen bonded to the water molecule underneath. The adsorbed layer of water and alcohol assumed a binary layers structure [25,28]. To confirm the coadsorption of water and ethanol on silicon surface,
200
0.0
-0.2
-0.4 0
50
100
150
Z-Piezo distance (nm)
Fig. 2. Force–distance curves of Si(100)/SiO2 pair at RH ¼30%. (a) Pa/Psat ¼ 0%, (b) Pa/Psat ¼ 20%, (c) Pa/Psat ¼ 40%.
200
L. Chen et al. / Wear 332-333 (2015) 879–884
the force–distance curves were measured at various atmosphere conditions. Fig. 2 shows the typical force–distance curves measured at 30% RH. In humid nitrogen without ethanol, the SiO2 tip separated instantaneously from the silicon surface, as shown in Fig. 2a. However, when ethanol vapor was introduced, the force–distance curves showed different behavior, as shown in Fig. 2b and c. During the approach process, the meniscus could be formed before the tip touched the substrate surface even at a distance of 5–20 nm from the surface (dotted line). During the retraction process, the meniscus was stretched due to the disjoining pressure (solid line) [32]. The difference between the force–distance curves at various RH and Pa/Psat should be attributed to the adsorbate ethanol layer formed on the outmost surface. The similar phenomenon has been observed on other films adsorbed on silicon surface, such as DNA film and polymer film [33,34]. Fig. 3 displays the ATR-IR spectra obtained at various RH and Pa/Psat conditions. The peaks centered at 3230 cm 1 and 3400 cm 1 correspond to the OH stretching vibrations of ice-like water and liquidlike water, respectively [35]. The peak at 2960 cm 1 denoted the ethyl stretching vibration of ethanol [25,28]. With the decrease of RH
Log(1/R)
Log(1/R)
Mostly due to H2O
0.003 0.002
Ethyl
0.001 0.000
and increase of Pa/Psat, the intensities of the OH stretching vibration decreased and the ethyl stretching vibrations increased. The ATR-IR measurements supported that the binary adsorbate layers of water and ethanol formed on the silicon substrate surface or SiO2 tip surface, and their thickness strongly depended on the composition of water and ethanol vapors. The results were also in accord with the measurement of sum frequency generation (SFG) vibration spectroscopy [25]. 3.2. Wear behavior of silicon under various RH and Pa/Psat conditions Fig. 4 shows the AFM images of wear scars on silicon surfaces after scratching with a SiO2 sphere tip for 200 reciprocating cycles at selected relative humidity (RH¼10%, 15%, 30% and 50%) and ethanol partial pressure (Pa/Psat ¼0%, 10%, 20%, 35% and 40%) conditions. It was found that the coadsorption of water and ethanol vapor showed a strong effect on the nanowear of silicon surface. At all RH conditions, the ethanol vapor could prevent or reduce the wear of silicon depending on its partial pressure. At 10% RH, the tribochemical wear causing material removal was observed on silicon surface without ethanol vapor (Pa/Psat ¼0%). With the increase of Pa/Psat, no discernible
0.004
0.004
0.003
0.003
OH of H2O ðanol
0.002
Log(1/R)
0.004
881
Ethyl
0.001
0.002
0.000 -0.001
-0.001
3600
3400
3200
Wavenumber
3000
3600
2800
cm-1
Ethyl
0.001
0.000
-0.001
OH of ethanol
3400
3200
Wavenumber
3000
2800
3600
cm-1
3400
3200
Wavenumber
3000
2800
cm-1
Fig. 3. ATR-IR spectra of the binary adsorbate layers of alcohol and water at the conditions of (a) Pa/Psat ¼ 5%/RH¼ 80%, (b) Pa/Psat ¼ 64%/RH ¼24% and (c) Pa/Psat ¼ 85%/RH ¼0%.
Pa/Psat = 0%
10%
20%
35%
40%
RH 10%
15%
30%
50% 400 nm Fig. 4. AFM images of wear scars on silicon surface after scratched by a SiO2 tip at selected RH and ethanol partial pressure (Pa/Psat) conditions. (a) RH ¼ 10%, (b) RH ¼ 15%, (c) RH ¼ 30%, (d) RH¼ 50%. Note the height full-scales for the images are 15 nm. Fn ¼ 3 μN, N ¼ 200 and D ¼500 nm.
882
L. Chen et al. / Wear 332-333 (2015) 879–884
12 RH 0% 10% 20% 35% 50%
10 8 6 4 2 0
RH
Wear depth d (nm)
Wear depth d (nm)
12
10
60% 70% 80%
8 6 4 2 0
0
10
20
30
40
50
60
Alcohol partial pressure Pa/Psat (%)
0 10 20 30 40 50 60 Alcohol partial pressure Pa/Psat (%)
3.3. Nanowear map of silicon surface under various RH and Pa/Psat conditions To construct a nanowear map showing the vapor condition dependence, the entire data of wear depth measured over wide ranges of RH and ethanol partial pressures were plotted in Fig. 6. At RHo10%, no wear was detected at all conditions. When RH increases from 10% to 50%, the wearless behavior of silicon could be achieved when the ethanol partial pressure was larger than the critical value. When RH was larger than 50%, the ethanol vapor lubrication was not sufficient to completely suppress silicon wear under all Pa/Psat conditions tested in this study. It was also noted that the most serious wear region located at the condition of RH ranging from 50% to 70% and Pa/Psat below 10%. In summary, the experimental results indicated that the ethanol lubrication can completely prevent the wear of the dynamic silicon parts in Si-based MEMS at relatively low RH (r50%). At high RH, the wear of silicon can only be partly prevented. It should be noted that due to the limitation of our gas flow system (Fig. 1); when the ethanol vapor pressure is increased to near-saturation, the lubrication effect is expected to further improve even at this high RH conditions.
100 (%)
Wear depth
80
0
P /P
surface damage of silicon was detected at Pa/Psat 410%. The Pa/Psat dependence of the silicon wear at RH of 15% and 30% was similar to the case at RH of 10%. There seems to be a critical ethanol partial pressure at each RH condition, below which silicon surface wore severely and above which there was no wear visible on silicon surface. As shown in Fig. 4b and c, the critical value of Pa/Psat can be identified as 20% at a RH of 15% and 40% at a RH of 30%. When the relative humidity increased to 50%, the wear of silicon resulted in grooves at Pa/Psat ranging from 0% to 40%, as shown in Fig. 4d. To quantitatively show the effect of the coadsorption of water and ethanol vapor on the wear behavior of silicon, the maximum wear depths of silicon surface under various RH and ethanol partial pressure conditions are shown in Fig. 5. Fig. 5a plots the data measured at low RH region (RH r50%). Since the tribochemical reaction of Si/SiO2 pair could not occur without water vapor [13], no surface damage was observed in dry air (RH¼0%). When the tests were performed at low RH region (10%rRHr50%), the wear depth gradually decreased below the detection limit at a critical ethanol partial pressure. Fig. 5b shows the wear depth of silicon measured at high RH region (RHZ60%). With the increase of ethanol partial pressure, the wear depth revealed the similar trend as that at low RH conditions, but it never decreased below the detection limit. For example, at 60% RH, the wear depth of silicon surface was reduced by 55% when Pa/Psat increased from 0% to 35%.
Alcohol partial pressure P /P (%)
Fig. 5. The maximum depth of wear scars on silicon surface plotted as the function of ethanol partial pressure (Pa/Psat). (a) RHr 50%, the wear depth increases with the increase in RH at given ethanol partial pressure; (b) RH Z60%, the wear depth decreases with the increase in RH. Fn ¼ 3 μN, N ¼ 200 and D ¼500 nm.
60
Wearless region
40 20
Serious wear 0 0
20
40
60
80
10 [nm]
Relative humidity RH (%)
Fig. 6. Nanowear map illustrating the wear depth of silicon surface at various RH and ethanol partial pressure (Pa/Psat) conditions. The blue area shows the wearless region of coadsorption conditions. The inset shows 75 environment conditions where data were collected.
4. Discussion 4.1. Lubrication of adsorbate ethanol layer Under the given loading conditions, the maximum contact pressure in our AFM tests is estimated to be 1.3 GPa based on the DMT model [36]. Since it is much lower than the yield stress of silicon (7 GPa) [37], the mechanical wear is not significant; thus surface wear must occur mainly through tribochemical reactions in humidity air [8,38]. During the shear process, the Si–O–Si bonding bridges may form between the two contact surfaces of Si(100)/SiO2 pair, which can further transfer the mechanical shear stress to the substrate surface and induce the dissociation of the interfacial bridge bonds or the subsurface Si–O bonds of silicon [13,39,40]. This tribochemical wear of silicon occurs only with the existence of water molecules [13]. When the alcohol vapor was introduced into the AFM chamber, the adsorbed alcohol molecules could form a lubrication layer and protect the silicon substrate [21–25]. To detect the lubrication of adsorbate alcohol layer to the sliding interfaces, the friction force of the Si(100)/SiO2 pair was measured at various RH and Pa/Psat conditions, as shown in Fig. 7. In relatively high RH without any ethanol vapor, friction force initially increased to a maximum value and then decreased with the increase of sliding cycles. This variation was accompanied by the silicon wear (Figs. 4–6), which implies some chemical and topographical changes of the sliding interface. The chemical change must include the loss of surface hydroxyl groups as well as native oxide layers [15]. Under relatively lower RH and higher Pa/Psat of ethanol conditions, the friction force was steady and low during the entire reciprocating cycles and no
L. Chen et al. / Wear 332-333 (2015) 879–884
883
2
RH=10%
Friction force Ft (µN)
1
2 RH=30% 1
3 -12 Wear rate γ (x10 m /Nm)
0.6 10% 30% 50% 70%
0.4
Wearless region 0.2
0.0 40
60
80
100 120 140 160 180 200 -12
Dissipated energy E (x 10
2
RH=50%
1 0
1
40 15 10 0 Cycle 100 (%) t sN P a/P sa
Fig. 7. Friction force vs. number of reciprocating cycles (Ft–N) curves at selected ethanol partial pressure (Pa/Psat ¼0%, 15% and 40%) and relative humidity RH ¼ (a) 10%, (b) 30% and (c) 50% conditions. The error of Ft estimated from five dependent measurements is less than 20%.
discernable wear was formed on silicon surface (Figs. 4–6). The low friction force and wearless state should be attributed to the lubrication of the adsorbate ethanol layer, which lowered the surface energy at the adsorbate/vapor interface [25,28]. The low friction force at high Pa/Psat will reduce the dissipated energy during sliding process [41]. Since the tribochemical reaction of Si/SiO2 pair is caused by the interfacial shear, its consequence (wear) must be correlated with the amount of energy dissipated through friction [22]. Fig. 8 shows the variation of wear rate on silicon substrate as a function of dissipated energy (integrated by friction force over total 200 reciprocating cycles). It was found that regardless of the variation of RH and Pa/Psat, there exists a threshold energy of 0.7 10 10 J, below which tribochemical wear does occur (gray region in Fig. 8) and above which the wear rate of silicon increases linearly with the dissipated energy. It is hypothesized that the mechanochemical reactions involve formation and dissociation of the Si–O–Si bridge bonds. The ethanol adsorbate can not only prevent the formation of Sisubstrate–O–Sitip bonded bridges during rubbing process, but also increase the energy barrier of Si–O–Si bonds dissociation on silicon surface [22]. Since the adsorption of longer chain molecule may further increase the energy barrier of Si–O–Si bonds dissociation on silicon substrate, the propanol or butanol vapor may have a better capability to resist the tribochemical wear of silicon surface [22,42,43]. 4.2. Resistance mechanism at low RH and Pa/Psat As shown in Figs. 4–6, the resistance of adsorbate ethanol layer to the tribochemical wear strongly depends on RH and Pa/Psat. The ATRIR measurement in Fig. 3a indicated that the ice-like water adsorbed on silicon surface was still prominent at high RH condition as the ethanol vapor was introduced. Compared with liquid-like water, the ice-like water with strong hydrogen bonding structure played a more dominated role in the tribochemical wear of silicon surface [13]. This might be the reason that silicon wear was just partially resisted at RHZ60% (Fig. 5b). When RH was lower than 60%, the tribochemical wear of silicon was completely suppressed when the Pa/Psat increased above a critical value. (Figs. 4–6). It was reported that the alcohol vapor could successfully prevent the silicon wear when the full
J)
Fig. 8. Correlation between the average wear rate and the total dissipated energy during 200 sliding cycles at RH of 10%, 30%, 50% and 70%. In left gray region, wear rate of silicon substrate is zero under relatively low dissipated energy.
SiO2 tip Silicon
Alcohol layer Water layer
SiO2 tip Silicon
Fig. 9. The schematics of the sphere-adsorbate-substrate contact during the sliding process in pure humidity air (a) and at relatively low Pa/Psat (b).
coverage of alcohol layer was formed on substrate surface [21,25]. In that case, due to the lubrication of adsorbate alcohol layer, the friction coefficient would be consistent with 0.2 whether the tests were operated in AFM and MEMS devices at nano/micro-scale or in pin-on-disc tribometers at macro-scale [21–25,44,45]. However, the tribochemical wear of silicon was found to be completely suppressed at relatively low Pa/Psat (o30%) where a partial coverage of ethanol adsorbate layer was expected on silicon surface (For example: Fig. 4a) [25,28]. Based on the friction results in Fig. 7a and b, the friction coefficient at Pa/Psat of 30% was estimated to be about 0.05 and 0.12 at a RH of 10% and 30%, respectively. The low friction coefficient also supported that the full coverage of ethanol adsorbate layer did not form under these conditions. Fig. 9 shows a schematic presentation of the sphere-adsorbatesubstrate contact in pure humid air and at relatively low Pa/Psat (o30%). Under high RH without alcohol, the tribochemical reaction caused serious wear of silicon [13]. However, no obvious surface damage of silicon was observed when RH was less than 10% (Fig. 6). It indicated that the tribochemical wear occurred when the amount of adsorbate water molecules were large enough to form water bridge between sliding interfaces, as shown in Fig. 9a. When the mixture vapor of water and alcohol was introduced, the hydroxyl groups of the water molecules are attached to silicon surface through hydrogen bonding and the alcohol molecules are hydrogen bonded to the water molecules forming an overlayer [25,26,28]. At relatively low Pa/ Psat (o 30%), the alcohol layer would partially cover the outmost surface, as shown in Fig. 9b [25–28]. The alcohol molecules on the top of the water layer can prevent (Fig. 4a and b) or at least significantly reduce (Fig. 4c and d) the water-mediated bridge formation across the sliding solid interfaces [28].
5. Conclusion Using an environment-controlled AFM, the effects of RH and ethanol partial pressure on the lubrication of Si/SiO2 sliding pairs were studied. The main conclusions can be summarized as follows:
884
L. Chen et al. / Wear 332-333 (2015) 879–884
(1) At RH below 50%, the tribochemical wear of silicon can be completely suppressed when the ethanol partial pressure increases above a critical value. As RH increases, the critical ethanol pressure needed for lubrication increases as well. (2) The co-adsorbed ethanol molecules decreased the frictional energy dissipation, which reduced tribochemical reaction probability. (3) The ethanol molecules on top of the water layer can prevent or at least significantly reduce the water-mediated bridge formation across the sliding solid interfaces.
Acknowledgments The authors are grateful for the financial support from the National Natural Science Foundation of China (91323302, 51375409, and 51175441). References [1] B. Bhushan, Nanotribology and nanomechanics of MEMS/NEMS and BioMEMS/ BioNEMS materials and devices, Microelectron. Eng. 84 (2007) 387–412. [2] M. Elwenspoeck, R. Wiegerink, Mechanical Microsensors, Springer, Berlin, 2001. [3] Z. Cui, Micro-nanofabrication Technologies and Application, Higher Education Press, Beijing, 2005. [4] M.A. Meitl, Z.T. Zhu, V. Kumar, K.J. Lee, X. Feng, Y.Y. Huang, I. Aadesida, R. G. Nuzzo, J.A. Rogers, Transfer printing by kinetic control of adhesion to an elastomeric stamp, Nat. Mater. 5 (2006) 33–38. [5] S. Demiri, S. Boedo, L.S. Holsen, Wear characteristics of large aspect ratio silicon microbearing systems, Wear 312 (2014) 58–69. [6] J.A. Williams, H.R. Le, Tribology and MEMS, J. Phys. D Appl. Phys 39 (2006) 201–214. [7] B. Bhushan, Tribology of industrial components and systems, in: B. Bhushan (Ed.), Modern Tribology Handbook, CRC Press, Florida, 2001. [8] K. Mizuhara, S.M. Hsu, Tribochemical reaction of oxygen and water on silicon surfaces, Tribol. Ser. 21 (1992) 323–328. [9] S. Shen, Y.G. Meng, A novel running-in method for improving life-time of bulkfabricated silicon mems devices, Tribol. Lett. 47 (2012) 273–284. [10] I.S.Y. Ku, T. Reddyhoff, A.S. Holmes, H.A. Spikes, Wear of silicon surfaces in MEMS, Wear 271 (2011) 1050–1058. [11] B.J. Yu, H.S. Dong, L.M. Qian, Y.F. Chen, J.X. Yu, Z.R. Zhou, Friction-induced nanofabrication on monocrystalline silicon, Nanotechnology 20 (2009) 465303. [12] B. Bhushan, J.N. Israelachvili, U. Landman, Nanotribology: friction, wear and lubrication at the atomic scale, Nature 374 (1995) 607–616. [13] J.X. Yu, S.H. Kim, B.j. Yu, L.M. Qian, Z.R. Zhou, Role of tribochemistry in nanowear of single-crystalline ailicon, ACS Appl. Mater. Interfaces 4 (2012) 1585–1593. [14] J.X. Yu, L. Chen, L.M. Qian, D.L. Song, Y. Cai, Investigation of humiditydependent nanotribology behaviors of Si(100)/SiO2 pair moving from stick to slip, Appl. Surf. Sci. 265 (2013) 192–200. [15] L. Chen, S.H. Kim, X.D. Wang, L.M. Qian, Running-in process of Si–SiOx/SiO2 pair at nanoscale – sharp drop of friction and wear rate during initial cycles, Friction 1 (2013) 81–91. [16] L. Chen, M.C. Yang, J.X. Yu, L.M. Qian, Z.R. Zhou, Nanofretting behaviours of ultrathin DLC coating on Si(100) substrate, Wear 271 (2011) 1980–1986. [17] R.W. Carpick, M. Salmeron, Scratching the surface: fundamental investigations of tribology with atomic force microscopy, Chem. Rev. 97 (1997) 1163–1194. [18] A. Erdemir, C. Donnet, Tribology of diamond-like carbon films: recent progress and future prospects, J. Phys. D: Appl. Phys. 39 (2006) 311–327. [19] L. Chen, M.C. Yang, C.F. Song, B.J. Yu, L.M. Qian, Is 2 nm DLC coating enough to resist the nanowear of silicon, Wear 302 (2013) 909–917. [20] S.H. Kim, D.B. Asay, M.T. Dugger, Nanotribology and MEMS, Nano Today 2 (2007) 22–29.
[21] D.J. Marchand, L. Chen, Y.G. Meng, L.M. Qian, S.H. Kim, Effects of vapor environment and counter-surface chemistry on tribochemical wear of silicon wafers, Tribol. Lett. 53 (2014) 365–372. [22] A.L. Barnette, D.B. Asay, D. Kim, B.D. Guyer, H. Lim, M.J. Janik, S.H. Kim, Experimental and density functional theory study of the tribochemical wear behavior of SiO2 in humid and alcohol vapor environments, Langmuir 25 (2009) 13052–13061. [23] D.B. Asay, M.T. Dugger, S.H. Kim, In-situ vapor-phase lubrication of MEMS, Tribol. Lett. 29 (2008) 67–74. [24] D.B. Asay, M.T. Dugger, J.A. Ohlhausen, S.H. Kim, Macro- to nanoscale wear prevention via molecular adsorption, Langmuir 24 (2007) 155–159. [25] A.L. Barnette, S.H. Kim, Coadsorption of n-propanol and water on SiO2: study of thickness, composition, and structure of binary adsorbate layer using attenuated total reflection infrared (ATR-IR) and sum frequency generation (SFG) vibration spectroscopy, J. Phys. Chem. C 116 (2012) 9909–9916. [26] A.L. Barnette, J.A. Ohlhausen, M.T. Dugger, S.H. Kim, Humidity effects on in situ vapor phase lubrication with n-pentanol, Tribol. Lett. 55 (2014) 177–186. [27] A.L. Barnette, D.B. Asay, M.J. Janik, S.H. Kim, Adsorption isotherm and orientation of alcohols on hydrophilic SiO2 under ambient conditions, J. Phys. Chem. C 113 (2009) 10632–10641. [28] A.L. Barnette, S.H. Kim, Attenuated total reflection infrared spectroscopy study of hysteresis of water and n-alcohol co-adsorption on silicon oxide, Langmuir 28 (2012) 15529–15536. [29] B.J. Yu, L.M. Qian, H.S. Dong, J.X. Yu, Z.R. Zhou, Friction-induced hillocks on monocrystalline silicon in atmosphere and in vacuum, Wear 268 (2010) 1095–1102. [30] K. Strawhecker, D.B. Asay, J. McKinney, S.H. Kim, Reduction of adhesion and friction of silicon oxide surface in the presence of n-propanol vapor in the gas phase, Tribol. Lett. 19 (2005) 17–21. [31] A. Torii, M. Sasaki, K. Hane, S. Okuma, A method for determining the spring constant of cantilevers for atomic force microscopy, Meas. Sci. Technol. 7 (1996) 179–184. [32] E. Hsiao, B.D. Veres, G.J. Tudryn, S.H. Kim, Identification of mobile species in cationic polymer lubricant layer on silicon oxide from afm and xps analyses, Langmuir 27 (2011) 6808–6813. [33] M. Wang, S.X. Cui, B.J. Yu, L.M. Qian, Effect of chain length on the conformation and friction behaviour of DNA, Sci. China Technol. Sci. 56 (2013) 2927–2933. [34] H.J. Butt, M. Kappl, H. Mueller, R. Raiteri, Steric forces measured with the atomic force microscope at various temperatures, Langmuir 15 (1999) 2559–2565. [35] D.B. Asay, S.H. Kim, Evolution of the adsorbed water layer structure on silicon oxide at room temperature, J. Phys. Chem. B 109 (2005) 16760–16763. [36] U.D. Schwarz, A generalized analytical model for the elastic deformation of an adhesive contact between a sphere and a flat surface, J. Colloid Interface Sci. 261 (2003) 99–106. [37] K.E. Petersen, Silicon as a mechanical material, Proc. of the IEEE 70 (1982) 420–457. [38] L. Chen, H.T. He, X.D. Wang, S.H. Kim, L.M. Qian, Tribology of Si/SiO2 in humid air: transition from severe chemical wear to wearless behavior at nanoscale, Langmuir 31 (2015) 149–156. [39] F. Katsuki, Single asperity tribochemical wear of silicon by atomic force microscopy, J. Mater. Res. 24 (2009) 173–178. [40] W. Maw, F. Stevens, S.C. Langford, J.T. Dickinson, Single asperity tribochemical wear of silicon nitride studied by atomic force microscopy, J. Appl. Phys. 92 (2002) 5103–5109. [41] S. Yashima, N. Takase, T. Kurokawa, J.P. Gong, Friction of hydrogels with controlled surface roughness on solid flat substrates, Soft Matter 10 (2014) 3192–3199. [42] A.J. Barthel, S.H. Kim, Lubrication by physisorbed molecules in equilibrium with vapor at ambient condition – effects of molecular structure and substrate chemistry, Langmuir 30 (2014) 6469–6478. [43] A.J. Barthel, A. Al-Azizi, N.D. Surdyka, S.H. Kim, Effects of gas or vapor adsorption on adhesion, friction, and wear of solid interfaces, Langmuir 30 (2014) 2977–2992. [44] D.B. Asay, E. Hsiao, S.H. Kim, Effects of adsorbate coverage and capillary on nano-asperity friction in atmosphere containing organic vapor, J. Appl. Phys. 110 (2011). [45] M.J. Marino, E. Hsiao, L.C. Bradley, O.L. Eryilmaz, A. Erdemir, S.H. Kim, Is ultralow friction needed to prevent wear of diamond-like carbon (DLC)? an alcohol vapor lubrication study for stainless steel/DLC interface, Tribol. Lett. 42 (2011) 285–291.