Tribology International 66 (2013) 259–264

Contents lists available at SciVerse ScienceDirect

Tribology International

0301-67http://d

n CorrUnivers

E-m

journal homepage: www.elsevier.com/locate/triboint

Study on the tribological properties of plasma nitrided bearing steelunder lubrication with borate ester additive

Song Wang a, Wen Yue a,b,n, Zhiqiang Fu a, Chengbiao Wang a, Xingliang Li a, Jiajun Liu c

a School of Engineering and Technology, China University of Geosciences (Beijing), Beijing 100083, Chinab Key Laboratory on Deep Geodrilling Technology of the Ministry of Land and Resources, China University of Geosciences (Beijing), Beijing 100083, Chinac Department of Mechanical Engineering, Tsinghua University, Beijing 100084, China

a r t i c l e i n f o

Article history:Received 3 April 2013Received in revised form27 May 2013Accepted 4 June 2013Available online 11 June 2013

Keywords:Boundary lubricationLubricant additivesTribochemistryIon implantation

9X/$ - see front matter & 2013 Elsevier Ltd. Ax.doi.org/10.1016/j.triboint.2013.06.004

esponding author at: School of Engineerity of Geosciences (Beijing), Beijing 100083, Cail address: cugbyw@163.com (W. Yue).

a b s t r a c t

The nitrided layer on 52100 steel was prepared by plasma nitriding treatment. The tribological behaviorswere examined under lubrication with borate ester containing nitrogen (B-N). The nitrided steelexhibited a lower friction coefficient and smaller wear scar diameter compared with the untreatedsteel. When 1.25 wt% B-N was used, the friction coefficient and wear scar diameter of nitrided steel werereduced by 34% and 45%, respectively. The main mechanism of this effect was attributed to the fact thatthe B-N additive can produce a tribofilm containing higher content of BN on the nitrided surface.

& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

For wind energy equipments, severe working conditions, suchas high loading, unsteady operation, system vibration, systemmisalignment and exposure to the extreme environment, lead tothe reduced durability and efficiency of some key components[1,2]. Improving the reliability of the devices is important to lowerthe cost of wind energy. Plasma nitriding is a widely used chemicalheat treatment in mechanical equipments. This method canproduce a compact and high hardness nitrided layer on the steelsurface with improved fatigue, wear and corrosion resistance ofwork pieces [3]. Plasma nitriding is an effective way of fulfillingthe requirements of the materials used for key components inwind energy equipments. In addition, an appropriate selection oflubrication additives is also important to the wind energy devicesfacing very harsh working conditions [4].

However, the majority of the lubrication additives were studiedfor Fe based material without surface modification, which isinconsistent with the real application. Whether the additives forFe based material are still effective for the modified surface isworth studying. Some previous works about the tribochemicalinteractions between nitrided steel surface and base oils, like alkylnaphthalene, ionic liquid and liquid paraffin, and some additives,have been conducted [5–7]. It showed that the nitriding treatmentcan make large contribution to improving the friction and wearbehaviors effectively [8–10].

ll rights reserved.

ing and Technology, Chinahina. Tel.: +86 10 82320255.

In order to accommodate the working condition of the windenergy equipment exposed in open-air, the lubricating oil shouldbe environment-friendly. In other words, the additives should notcontain S and P elements and other harmful components in themolecular formula. As a kind of potential eco-friendly lubricationadditive, B and N containing additives are of significant interestbecause of the desired intrinsic properties, including environ-ment-acceptability, absence of S and P in the molecular formula,good anti-wear and friction-reducing properties and good hydro-lytic stability [11–13]. Zheng et al. [14] synthesized several kinds ofN-containing borate esters as additives. They found that borateesters with a stable five- or six-member ring possessed goodhydrolytic stability and friction-reducing ability, and exhibitedgood extreme pressure and anti-wear properties as well.

Greco et al. [15] found that the electrochemical boride surfaceand the nano-colloidal boron nitride lubricant additive showed anexcellent synergistic effect in the wind turbine gearbox. Whetherthe nitrided surface lubricated with N-containing borate esteradditives can also show such a good effect? In order to find theanswer, a study on the tribological properties of plasma nitrided52100 steel under lubrication with borate ester additives wascarried out in this work. A new idea for improving the durabilityand efficiency of wind energy devices is proposed.

2. Experimental details

AISI 52100 steel balls (with a diameter of 12.7 mm, hardnessof 770 HV, and surface roughness Ra 0.025 mm) were nitrided in apulsed plasma nitriding furnace (LDM2-25). The treatment

S. Wang et al. / Tribology International 66 (2013) 259–264260

proceeded at a temperature of 520 1C with a voltage of 700 V for atotal duration of 5 h. NH3 is used as the source gas with a pressureof 670 Pa. Then the samples were cooled in vacuum to ambienttemperature.

The friction and wear tests were carried out on an MS-10JRfour-ball friction and a wear tester with a ball-to-ball contactconfiguration. The upper sample was a rotating ball fixed on thespindle, and the lower samples were three fixed balls. The fourballs used in each test were selected as the same material. Thefriction and wear tests were conducted at a condition of load of392 N (corresponding to the Hertz mean contact stress of2.293 GPa), linear speed of 0.461 m/s, and test duration of60 min. The minimum film thickness was determined using theDowson and Hamrock minimum film thickness equation for anelastohydrodynamic point contact [16]. The lambda ratios (calcu-lated minimum film thickness/initial composite surface root meansquare roughness) for these tests were calculated as 0.812 and 0.13for the untreated and nitrided surfaces, respectively. They are notthe same but both are less than 1. Therefore, the lubrications onuntreated and nitrided surfaces are in the boundary regime. Thefriction force and friction coefficient were measured by a forcesensor and recorded by a computer. The wear scars were mea-sured by a 15 J-type microscope. The tests were replicated at leastthree times. A good repeatability for the friction coefficients andwear scars in the whole process of the test was recorded and theresults were averaged.

The base oil used was synthetic oil polyalphaolefin (PAO)whose kinematic viscosity was 16.7 mm2 s−1 at 40 1C. The mainchemical composition of borate ester (B-N) is B 1.0 wt% and N2.6 wt%. The density is 0.98 g cm−3 at room temperature. Twoseries of tests were carried out under lubrication of PAO with B-Nadditive. One was on the untreated surface, and another was onthe nitrided surface. B-N's concentrations (mass fraction) in thetested oil were 0.0%, 0.25%, 0.5%, 0.75%, 1.0%, 1.5% and 2.0%.

The microhardness of the sample was measured by an MH-6microhardness tester, and the applied load was 1.96 N withloading time of 5 s. The phase composition of the plasma nitridedlayer was determined with a D-max/2550 X-ray diffractometer(XRD) in glancing angle geometry, using Cu Kα radiation as theexcitation source. After the wear test, the samples were ultra-sonically washed with petroleum ether and alcohol (the volumeradio 3:1). A JSM-6460LV scanning electron microscope (SEM) wasutilized to observe the wear scar morphologies of the balls. Thechemical states of tribofilms were investigated using a PHIQuantera X-ray photoelectron spectroscope (XPS). The instrumentemployed a high-power rotating anode and monochromatised AlKα X-ray source. The tested sample surface was sputtered about5 nm in depth by Ar ion (sputtering depth on the standard SiO2

sample). The binding energy of 284.6 eV for contaminated C wasused as a reference for charge correction.

3. Results and discussion

3.1. Characterization

Fig. 1 shows the SEM images of the untreated surface and thenitrided surface. It can be clearly seen that the spherical particleswere piled up closely with a lot of fine micro-pores distributed onthe nitrided surface (Fig. 1b), which resulted in a higher roughnessRa 0.16 mm. However, none was found on the untreated surface(Fig. 1a).

Fig. 2 shows the variation of the hardness of nitrided layer withthe depth. The hardness of nitrided samples increased to 900–1000 HV with a distance of about 40 mm from the surface, andthen decreased gradually to 500–600 HV when the distance from

the surface reaches 40–140 mm. When the distance from thesurface is larger than 140 mm, the hardness remained at alower value.

From Fig. 3, it was found that the compositions of the nitridedsample were mainly γ′-Fe4N, ε-Fe2−3N and CrN, while those of theuntreated sample were martensite and austenite.

3.2. Friction and wear behavior

Fig. 4 shows the variations of friction coefficients of theuntreated and nitrided surfaces with different concentrations ofB-N. It indicated that the friction coefficients of the nitridedsurface are always lower than those of the untreated surfaceunder the same lubrication condition. The variation of the frictioncoefficients along with the increase of additive contents is irre-gular. Only when lubricated with 1.0 wt% and 1.25 wt% B-N, thefriction coefficients of nitrided surface presented a sharp decreaseto about 0.07, much lower than other concentrations. It mayattribute to a friction reducing tribofilm formed on the nitridedsurface lubricated with the proper B-N content.

Fig. 5 shows the variation of wear scar diameters with differentconcentrations of B-N. It can be seen that wear scar diameters onboth surfaces were almost the same on untreated and nitridedsurfaces in the lubrication of PAO. This results from the fact thatthe hardness of mating pairs was the same in each test [17]. Thewear scar diameters of the untreated surface were almostunchanged for all concentrations, while those of nitrided surfacewere gradually reduced with the increase of B-N percentage. Theminimum wear scar diameter reached 0.4 mm at the concentra-tion of 1.25%, 45% less than that on the untreated surface. Whenthe concentrations of the B-N continue to rise, the wear scardiameter on the nitrided surface increased again, but was stillsmaller than that on the substrate surface. It can be consideredthat there existed certainly a beneficial synergistic effect betweenthe nitrided 52100 steel surfaces and B-N additive, which led tothe improvement of wear-resistance through forming a protectivetribofilm.

3.3. SEM analysis

In order to understand the difference of the friction-reducingand anti-wear mechanism of the B-N additives on the nitrided andthe untreated surfaces, the wear surfaces were analyzed by SEM.Fig. 6 shows the SEM morphologies of the worn surfaces lubri-cated with PAO and 1.25 wt% B-N. So many deep furrows can befound on the surfaces of untreated and nitrided samples under thelubrication of PAO. It is obvious that the nitrided surface shows asmoother surface than the untreated ones under the same lubrica-tion conditions. The furrows become obviously shallower as theaddition of the additives. Compared with some corrosive pittingon the untreated surface lubricated with 1.25% B-N, the nitridedsurface lubricated with 1.25% B-N shows the smoothest wearsurface. This might be attributed to the excellent synergistic effectof borate ester and a higher hardness of the nitrided surface.

3.4. XPS analysis

The wear surfaces of the nitrided 52100 steel lubricated withdifferent lubricants were analyzed by XPS so as to acquire moreinformation about the tribochemical reaction involved during thesliding process, as shown in Fig. 7. Table 1 exhibits the elementalcomposition, determined by XPS quantitative analysis, of thetribofilm on the wear surface lubricated with 1.25% B-N. Thebinding energy values and quantifications of all the elements onthe surface and the main components of the reference compounds[18] are listed in Table 2. It can be found that 9.45% B and 1.80%

Fig. 1. SEM morphologies of the sample surfaces, (a) untreated surface and (b) nitrided surface.

0 20 40 60 80 100 120 140 160 180 200 220

500

600

700

800

900

1000

Mic

roha

rdne

ss (H

V)

Distance from surface ( m)

Fig. 2. The variation of hardness with the depth below surface of nitrided steel.

30 40 50 60 70 80

Untreated steel

Rel

ativ

e in

tens

ity

--Martensite--Austenite

Nitrided steel -- Fe4N-- Fe2-3N--CrN

2 (degree)

Fig. 3. X-ray diffraction patterns of untreated and nitrided steel surface.

0.04

0.05

0.06

0.07

0.08

0.09

0.10

0.11

0.12

0.13

Fric

tion

Coe

ffic

ient

Time (min)

PAO PAO+0.25% B-N

PAO+0.50% B-N PAO+0.75% B-N

PAO+1.00% B-N PAO+1.25% B-N

PAO+1.50% B-N PAO +2.00% B-N

0 10 20 30 40 50 60

0 10 20 30 40 50 600.04

0.05

0.06

0.07

0.08

0.09

0.10

0.11

0.12

0.13

Fric

tion

Coe

ffic

ient

Time (min)

PAO PAO+0.25% B-N PAO+0.50% B-N PAO+0.75% B-N PAO+1.00% B-N PAO+1.25% B-N PAO+1.50% B-N PAO+2.00% B-N

Fig. 4. Variation of friction coefficients with different concentrations of B-N,(a) untreaed surfaces and (b) nitrided surfaces.

S. Wang et al. / Tribology International 66 (2013) 259–264 261

N were detected on the nitrided surface, while only 1.13% B and0.47% N were present on the untreated surface. The binding energyof B 1s at 189.4–192 eV was corresponding to BNxO(1−x), BN, andB2O3. The binding energy at 190.4 eV and 191.3 eV was corre-sponding to BN. And also, some different compounds were foundon the two surfaces. On the untreated surface, the binding energyof C 1s corresponded to C at 284.6 eV and two organics at 285.4 eVand 287.8 eV. While on the nitrided surface, in addition to C at284.6 eV and organic at 287.8 eV, Fe3C and another organic weredetected. Oxygen element existed in three chemical states on theuntreated surface, namely Fe2O3 at 530.1 eV, FeOOH at 531.2 eVand BNxO(1−x) at 532.6 eV. Additionally, another peak at 529.1 eVassigned to Fe3O4 was only found at the nitrided surface. The Fe2p

signals exhibit two peaks, 2p1/2 and 2p3/2, due to spin–orbitsplitting. The binding energy of the 2p1/2 peak was lower than the2p3/2. The two peaks meet an area ratio of 1:2, and only the moreprominent 2p3/2 signal was used for quantification. The mainchemical states of Fe on the wear surfaces can be observed: FeOand Fe2O3 corresponding to 709.4 and 710.7 eV, as well as FeOOHat 711.8 eV. Fe (II)-satellite at 715.1 eV has only been detected onthe untreated surface [19].

The most significant difference between the untreated surfaceand the nitrided surface was the content of different elements andcompounds. It implies that different tribochemical reactions couldoccur on different surfaces. This results in more excellent friction-reduction and anti-wear effect for the nitrided sample than theuntreated surface.

S. Wang et al. / Tribology International 66 (2013) 259–264262

3.5. Discussion

In this study, the friction reduction and anti-wear properties onthe untreated surface were compared with those on the nitridedsurface. The results proved that the latter exhibited an obviouslybetter performance. It implies that different mechanisms mightoccur on the different surfaces. One reason is that the nitridedlayer could effectively increase the hardness of steel and improveits wear resistance and friction reduction. According to the aboveresults in the wear test, the friction coefficient and the wear scardiameter of the nitrided surface are slightly lower than those ofthe untreated surface under the lubrication of PAO.

Furthermore, the tribofilm formed on the wear surface couldplay a more important role in improving the wear resistance and

0.0%

0.5%

0.75%

1.0%

1.5%

2.0%

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Wea

r sc

ar d

iam

eter

(mm

)

B-N concentration

Untreated surface

Nitrided surface

1.25%

0.25%

Fig. 5. Variation of wear scar diameter of ball with different concentrations of B-Nlubrication.

100

100

Fig. 6. SEM images of the worn surfaces lubricated by PAO containing B-N. (a) Untreatsurface lubricated with 1.25% B-N and (d) nitrided surface lubricated with 1.25% B-N.

friction reduction under boundary lubrication. Much higher ele-ment concentrations of B and N were detected on the nitridedsurface than that on the untreated surface. It implies that morefriction reducing and anti-wear chemical compositions wereproduced on the nitrided surface. The XPS analyses revealed thatonly few B2O3 was found on the untreated surface. While on thenitrided rubbed surface, B element was found in three differentchemical states. The formed boron-containing compounds weremainly organics, B2O3 and BN. BN possesses a lower shearingstress by its hexagonal structure and is widely used as a solidlubricant [16,20]. The other reaction products such as B2O3 andorganics presented in this study possessed also the good wearresistance and friction reduction as reported in other literatures[14,21].

The content of B on the untreated surface was much lower thanthat on the nitrided surface even with the optimal concentrationof B-N. Therefore, the reductions of wear scar diameter andfriction coefficient on the untreated surface were not obvious withthe increase of additive content. But on the nitrided surface, whenthe concentration of B-N is low, the BN formed on the contactregion is insufficient to separate the counterpart surfaces. How-ever, with the increase of B-N concentration, more BN and otherlubricant compounds were produced on the frictional surfaces.This can sufficiently prevent the metallic asperities of bothsurfaces from directly contacting, thus the friction and weardecrease. While the concentration further increases, too muchcompounds were delivered to the contact area and lead tocoagulation at the interface of rubbing-pair. This in turn causesthe increase of friction force and unstable operation, even the localdamage of tribofilm, and thus results in higher friction and wear.

A beneficially synergistic effect between the plasma nitridedAISI 52100 steel and nitrogen-containing borate ester was pre-sented, as the electrochemical boride surface and the nano-colloidalboron nitride lubricant additive performed in Greco's [15] work.

100

100

ed surface lubricated by PAO, (b) nitrided surface lubricated by PAO, (c) untreated

Rel

ativ

e In

tens

ity

B1s 1.25% B-NNitrided surface

Binding energy (eV)

Binding energy (eV)

Binding energy (eV)

1.25% B-NUntreated surface

Nitrided surface0.50% B-N

Nitrided surface0.50% B-N

Rea

tive

Inte

nsity

N-B%52.1s1C

Nitrided surface

1.25% B-NUntreated surface

Nitrided surface

Rel

ativ

e In

tens

ity

N1s 1.25% B-NNitrided surface

1.25% B-NUntreated surface

0.5% B-N

Rel

ativ

e In

tens

ity

O1s 1.25% B-NNitrided surface

Untreated surface

Bending energy (eV)

1.25% B-N

0.5% B-NNitrided surface

183 186 189 192 195 198 201 276 279 282 285 288 291 294

390 395 400 405 525 530 535 540

705 710 715 720 725 730 735

1.25 % B-NUntreated surface

0.5% B-NNitrided surface

1.25 % B-NNitrided surface

Rel

ativ

e In

tens

ity

Binding Energy (eV)

Fig. 7. XPS spectra of different elements on the untreated and nitrided surfaces under lubrication with B-N.

Table 1Element contents on wear scars of untreated and nitrided surfaces.

Atomic content B (%) C (%) N (%) O (%) Fe (%)

1.25% Untreated surface 1.13 65.80 0.47 29.85 2.760.5% Nitrided surface 4.70 46.15 2.05 41.34 5.771.25% Nitrided surface 9.45 48.79 1.80 37.21 2.75

Table 2Electron binding energies and possible chemical states of elements on the wornsurfaces of untreated and nitrided samples. (Fe(II)sat is the satellite of Fe (II)).

Elements Untreated surface Nitrided surface

B.E.(eV)

FWHM(eV)

Compound B.E.(eV)

FWHM(eV)

Compound

B 1s 189.4 1.5 BNxO(1−x)

190.4 1.5 BN191.3 1.5 BN192 1.5 B2O3

C 1s 283 1.5 Organic284.6 1.5 C 283.6 1.5 Fe3C285.4 1.5 Organic 284.6 1.5 C287.8 1.5 Organic 287.8 1.5 Organic

S. Wang et al. / Tribology International 66 (2013) 259–264 263

The increase of the surface hardness and the formation of aneffective tribofilm lead the excellent tribological properties together.The combination of surface treatment and lubricant additive wouldbe a good application in a wind turbine drive train to accommodatesevere operating conditions and mitigate surface originated failure.

N 1s 397.5 2.3 BN399.2 2.3 Organic

O 1s 529.1 1.7 Fe3O4

530.1 1.7 Fe2O3 530.1 1.7 Fe2O3

531.2 1.7 FeOOH 531.2 1.7 FeOOH532.6 1.7 BNxO(1−x) 532.6 1.7 BNxO(1−x)

Fe 2p 709.4 2.5 FeO 709.4 2.5 FeO

4. Conclusions

The main conclusions can be drawn from this research asfollows:

710.7 2.5 Fe2O3 710.7 2.5 Fe2O3

711.8 2.5 FeOOH 711.8 2.5 FeOOH715.1 2.5 Fe(II)sat

(1) A beneficially synergistic effect between the plasma nitrided

AISI 52100 steel and nitrogen-containing borate ester was

S. Wang et al. / Tribology International 66 (2013) 259–264264

presented. The usage of PAO with 1.25 wt% B-N additivereduced the friction coefficient by 38% and wear scar diameterby 45% for nitrided surface, compared with those on theuntreated surface.

(2)

The higher hardness of nitrided surface and the formation ofhexagonal BN and B2O3 in the tribofilm play an important rolein reducing the friction and wear of bearing steel.

Acknowledgments

The authors would like to thank the Beijing Natural ScienceFoundation (3132023), the National Natural Science Foundation ofChina (51005218 and 51275494), the Fundamental Research Fundsfor the Central Universities and the Tribology Science Fund of StateKey Laboratory of Tribology Tsinghua University (SKLTKF11B04)for their financial support to this research, and also thank engineerHaipeng Huang from R&D Center of SINOPEC Lubricant Company(Beijing) for providing the base oil and additives.

References

[1] Joselin Herbert GM, Iniyan S, Sreevalsan E, et al. A review of wind energytechnologies. Renewable and Sustainable Energy Reviews 2007;11:1117–45.

[2] Blau PJ, Walker LR, Xu H, et al. Wear analysis of wind turbine gearboxbearings; 2010.

[3] Neville A, Morin A, Haque T, et al. Compatibility between tribological surfacesand lubricant additives—how friction and wear reduction can be controlledby surface/lube synergies. Tribology International 2007;40:1680–95.

[4] Laine E, Olver A, Beveridge T. Effect of lubricants on micropitting and wear.Tribology International 2008;41:1049–55.

[5] Xia YQ, Wang SJ, Zhou F, et al. Tribological properties of plasma nitridedstainless steel against SAE52100 steel under ionic liquid lubrication condition.Tribology International 2006;39:635–40.

[6] Xia YQ, Zhou F, Sasaki SY, et al. Remarkable friction stabilization of AISI 52100steel by plasma nitriding under lubrication of alkyl naphthalene. Wear2010;268:917–23.

[7] Li XL, Yue W, Wang CB, et al. Comparing tribological behaviors of plasmanitrided and untreated bearing steel under lubrication with phosphor andsulfur-free organotungsten additive. Tribology International 2012;51:47–53.

[8] Ma YS, Liu JJ, Zheng LQ. The synergistic effects of EP and AW additives withoxynitrided surface of steel. Tribology International 1995;28:329–34.

[9] Ma YS, Wu YS, Gu ZQ, et al. Lubricating mechanisms of sulfurized olefin onoxynitrided steel surface under boundary lubrication condition. Wear1996;194:174–7.

[10] Ma YS, Liu JJ, Wu YS, et al. The effect of oxy-nitrided steel surface onimproving the lubricating performance of tricresyl phosphate. Wear1997;210:287–90.

[11] Yao JB, Dong JX. Improvement of hydrolytic stability of borate esters used aslubricant additives. Lubrication Engineering 1995;51:475–9.

[12] Shen GQ, Zheng Z, Wan Y, et al. Synergistic lubricating effects of borate esterwith heterocyclic compound. Wear 2000;246:55–8.

[13] Yao JB. Antiwear function and mechanism of borate containing nitrogen.Tribology International 1997;30:387–9.

[14] Zheng Z, Shen GQ, Wan Y, et al. Synthesis, hydrolytic stability and tribologicalproperties of novel borate esters containing nitrogen as lubricant additives.Wear 1998;222:135–44.

[15] Greco A, Mistry K, Sista V, et al. Friction and wear behavior of boron basedsurface treatment and nano-particle lubricant additives for wind turbinegearbox applications. Wear 2011;271:1754–60.

[16] Stachowiak GW, Bachelor AW. Engineering tribology. 3rd ed. UK: ElsevierButterworth-Heinemann; 2001.

[17] Leśniewski T, Krawiec S. The effect of ball hardness on four-ball wear testresults. Wear 2008;264:662–70.

[18] ⟨http://srdata.nist.gov/xps/⟩.[19] Yue W, Gao XC, Wang CB, et al. Synergistic effects between plasma-nitrided

AISI 52100 steel and Zinc dialkyldithiophosphate additive under boundarylubrication. Tribology Transactions 2012;55:278–87.

[20] Liu GL, Hu YQ, Sun X, et al. The study on performance of the borate estercontaining nitrogen as lubricant additive. Lubrication Engineering2004;2:114–6.

[21] Qiao YL, Liu WM, Qi SK, et al. The tribochemical mechanism of the boratemodified by N-containing compound as oil additive. Wear 1998;215:165–9.