Contents

    

Interactions of n-hexadecane with 52100 steel surface under friction conditions

Monika Makowskaa and Czeslaw Kajdasb and Marian Gradkowskia
aInstytut Technologii Eksploatacji, PL-26-600 Radom, ul. Pulaskiego 6/10
bPolitechnika Warszawska, PL-09-400 Pock, ul. Łukasiewicza 17

Abstract

Straight-chain aliphatic hydrocarbons, particularly n-hexadecane, are used as reference fluids in research concerning the AW/EP effectiveness of triboactive additives. It is assumed, however, that under boundary lubrication conditions these apparently non-reactive hydrocarbons might influence the tribochemistry of the additives because aliphatic compounds also undergo chemical reactions. Both the thermochemical and tribochemical oxidation of aliphatic hydrocarbons lead to the formation of triboactive products (e.g. alcohols, aldehydes and carboxylic acids). Interactions of these compounds with rubbing surfaces were examined using GC/MS, XPS and EDS techniques. It has been found that carboxylic acids react with the iron surface generating salts or complex compounds.

KEY WORDS: tribochemistry, n-hexadecane, bulk lubricant analysis, surface analysis


Published in: Tribology Letters, v.13, No. 2 p.65-70

 

1. Introduction

Tribochemical reactions of mineral lubricant components with freshly exposed metal surfaces lead to the formation of protective layers. They prevent negative effects of metal surface asperities reducing wear of rubbing elements. Sources of tribochemistry can be conventional chemical reactions occuring in the bulk lubricant at the contact zone, as well as mechanically and thermally induced reactions at the metal asperities [1]. Investigations of tribochemical changes of mineral lubricating oils generally are limited to additives leaving the hydrocarbon base fluid out of account. However, hydrocarbons-influenced by the processes occurring under friction conditions-can undergo reactions which lead to the formation of chemical active compounds [2]. Sufficient evidence exists in the literature [3] to indicate that components of mineral lubricating oils (especially additives) can react with the metal surfaces of sliding elements. This is also due to unsaturated organic additives [4]. On the other hand, the chemical structures of the reaction products and mechanisms explaining compound formation are still poorly known. Generally, apart from metal oxide carboxylic acid salts (soaps), alkoxides, complex compounds and/or organometallic compounds are taken into account. Some papers on this subject also suggest the formation of friction polymers.
There are very diverse opinions on the chemical reactions that take place on sliding surfaces and type of products, even on the assumption that the same chemical compound was used as a lubricant. Sources of these diversities are the different conditions used to provide tribological tests and analyse sliding surfaces, as well as various interpretations of obtained results. More often it is difficult to avoid problems related to the limited amount of sample, separation of formed compounds from a solution, lack of suitably sensitive apparatus or experimental procedures.
Considering the above-discussed issues, the primary objective of this paper is to study in great detail the tribochemical changes of n-hexadecane, which is widely used as a model reference fluid in testing tribological additives for both metal and ceramic substrates.

2. Experimental procedures

Lubricant. N-hexadecane (>99%) was used as a lubricant without any additional purification.

Tribological tests. To perform tribological tests under boundary lubrication conditions, a T-11 ball-on-disk tester was used. Elements of the friction pair (balls and disks) were made from 52100 bearing steel, 60HRC, Ra = 0.3 µm. Before each run, balls and disks were cleaned ultrasonically in n-hexane for 15 minutes.
All tests were performed under the following operating conditions: load 9.81N; sliding velocity 0.25m/s; sliding distance 500 m and 3000 m; ambient temperature 20±2°C. After setting up the device and the test parameters, n-hexadecane (2 cm3) was placed in the disk-holding cup. Disks after tests were again washed with n-hexane and stored above silica gel until further analysis.
Products of the tribochemical changes of n-hexadecane were investigated ex situ using GC/MS to determine reaction products forming in the bulk lubricant as well as EDS and XPS techniques to investigate the chemistry of the products generated on the steel surface.

Gas chromatography/mass spectrometry. GC/MS analysis was performed by an HP 5890 Series II Gas Chromatograph coupled with an HP 5972 Series Mass Selective Detector equipped with an autosampler. It was performed under following conditions: injection 1 µl (without solvent), column HP PONA (50m, f 0.2mm), carrier gas He 6.0, temperature 60 ® 260 °C at 8°C/min, pressure 100 ® 200 kPa at 4kPa/min, flow rate 0.4ml/min. To interpret the mass spectra, the DataBase, Wiley 275.L commercial library was used.

Fourier transform infrared microspectrometry. To investigate the organic layer on the wear tracks, an i-series PE Fourier Transform Infrared Microspectrophotometer (FTIRM) was used. Reflection spectra were recorded in the range of 4000 to 700 cm -1 with a resolution of 4 cm -1 (100 scans at each point). All spectra were corrected by subtracting spurious bands originating from carbon dioxide, near 2350 cm -1, as well as smoothing by the Savitsky-Golay method and multipoint normalization of the baseline. The mathematical processing of spectra shows no influence on their appearance.

Energy dispersion spectroscopy. EDS analysis was performed under the following conditions: accelerating voltage 10 kV, pressure 10 -3 Pa and take-off angle 25°. Measurements were directed at the investigation of changes of elementary composition within the wear scar compared to the surface beyond the contact region.

X-ray photoelectron spectroscopy. Organic films layered in the wear track on the disks were subjected to XPS analysis using a Phi-5702 Multifunctional X-Ray Photoelectron Spectrometer with MgKa monochromatic exciting radiation, at a constant power of 250 W and 5A1. aperture. The reference binding energy of carbon for the Is line was 284.6 eV. The energy resolution of the high-resolution spectra was ±0.2eV. The Fe2p, C1s and O1s profiles were recorded on f 0.8mm diameter region at a constant pass energy of 29.35eV.

Figure 1: Part of the gas chromatogram of used fi-hexadecane after the 3000m sliding distance under boundary lubrication conditions in the 1-11 Tribometer.

3. Results and discussion

It is important to know what product types are formed under friction conditions as a result of n-hexadecane reactions. To determine the chemical changes of the bulk lubricant coming from the disk-holding cup, the used lubricant was subject to a gas chromatographic separation. A representative part of the recorded chromatogram is depicted in figure 1.
It has been found that, in the chromatogram, there are groups of peaks at constant intervals of retention time. This indicates the presence of separated compounds relating to the same homologous series [5]. Application of gas chromatography coupled with mass spectrometry (GS-MS) allows clear characteristics of the compounds separated from the used n-hexadecane to be obtained by means of the corresponding m/e peaks and their abundances. Figure 2 shows an example of a spectrum relating to the chromatographic peak at 9.01 min. The fragmentation pattern is characterized by clusters of peaks and the corresponding peaks of each cluster are 14 (CH2) mass units apart. In this instance probably ionisation caused a strong fragmentation of the molecules. Therefore identification of the molecular ion peak was impossible. It has been supposed that the mass spectrum presented in figure 2 is not related to straight-chain saturated hydrocarbons. The molecular ion peak of these compounds is usually present [6]. The undetectable molecular ion peak caused difficulty in determination of the compound's molecular mass. Therefore interpretation of the mass spectra has been based on empirical principles and results of comparisons with a commercial library.
On the basis of the above interpretation it has been found that the mass spectrum presented in figure 2 corresponds to an aldehyde. The good diagnostic peak at m/e = 29 is due to the hydrocarbon C2H5+ ion and formyl [HC=0]+ ions. In aldehydes containing at least four atoms of carbon in the chain, cleavage of the C-Cb bond next to the C=0 group (with hydrogen rearrangement) occurs to give a major peak at m/e = 44, 58 or 72
Figure 2: Mass spectrum corresponding to the chromatographic peak at 9.01 min (in Figure 1).

depending on the a-substituents. Increment of the carbon chain length causes a gradual loss of the molecular ion peak; in the butanal spectrum it may be missing [7]. Simultaneously, the spectrum becomes similar to straight-chain hydrocarbons (peaks at m/e = 29, 43, 57, 71...) and the relative intensities of peaks decrease with increasing m/e values. This is evident in figure 2. On the basis of a detailed interpretation of the spectra, it has been concluded that the spectrum corresponds ton-pentanal.

The next predominant chromatographic peaks (see figure 1) correspond to aldehydes (from the same homologous series) of successive increasing masses by 14 mass units. Less intense chromatographic peaks accompanying the major peaks indicate the presence of alcohols (before the aldehyde peak) and carboxylic acids (over the aldehyde peak). All n-hexadecane reaction products are potential triboactive compounds which can affect or react with iron atoms on the steel surface. In order to investigate such a possibility, wear tracks were analyzed by FTIRM, EDS and XPS techniques.
The FTIRM findings [8] suggest the formation of either carboxylic acid salts, unsaturated compounds or complex compounds under friction conditions in the presence of pure n-hexadecane. Their characteristic infrared adsorption peaks appear at around 3300, 1650 and 1550 cm -1. Figure 3 presents the full spectra and result of a 1548 cm -1 band deconvolution. Similar signals were identified by the authors of paper [9] during investigations of tribochemical and thermochemical reactions of stearic acid on copper surfaces. They recognized that the IR bands in region of 1550 cm -1 are assigned to surface reaction products leading to salt and/or complex compound formation.
The participation of nitrogen cannot also be excluded in the chemical structures of the products. In previous papers [8, 10], the influence of nitrogen from the air on reactions proceeding in the friction contact was not investigated. Despite the lower chemisorption activity of nitrogen (about three times) in comparison to oxygen [11], this aspect cannot be omitted, considering the approximate sizes of nitrogen (0.71 Ĺ) and oxygen atoms (0.74 Ĺ). To check if nitrogen is incorporated into the tribochemical reaction products, the elemental composition in the wear track surface was determined by EDS analysis. The spectra recorded on a steel surface before and after n-hexadecane lubricated friction test under air (78% N) are presented in figure 4.
In the wear track, intense signals are noticeably originating from iron (705 eV, 403 eV), carbon (282 eV) and oxygen (523 eV). In this range of energy, signals also appeared in the EDS spectrum of steel before the friction test. However, the contribution of carbon (figure 4b) has increased relative to the initial surface (figure 4a). This provides evidence of organic compound formation on the steel surface during friction. No signal originating from nitrogen in range of 392 eV has been observed. In this connection, the presence of nitrogen-containing compounds on sliding surfaces is probably impossible. The author of work [11]-based on Auger electron spectroscopy-has ascertained that, in the formation of surface layers on steel lubricated during friction with mineral oil (without any additives), there is a much higher content of oxygen (43—47%) than nitrogen (1.63—3.7%). Such significant differences in the content of both elements was explained by the higher reactivity of oxygen.
Results obtained by FTIRM and EDS lead to the finding that products of tribochemical changes of n-hexadecane are carboxylic acid salts or complex compounds. Since the technique used cannot be sensitive enough to analyse monolayers, the sliding surfaces have been examined by means of XPS.
Figure 5 presents high-resolution spectra of iron 2p, oxygen 1s and carbon 1s photo-electrons as elements entering into the composition of n-hexadecane tribochemical reaction products. The XPS spectrum of nitrogen shows no significant signal. The spectrum of Fe2p photoelectrons has a peak at 725 eV (beyond the interpretative iron 2p region) and at about 711 eV assigned to iron compound formation (elemental iron binding energy is 706.4 eV [12]). The last
(a)
(b)
Figure 3: FTIRM spectrum of tribochemical products of«-hexadecane: (a) whole range spectrum and (b) result of l548 cm -1 adsorption hand deconvolution.

one is typical for an Fe-0 bond, for example in such compounds as oxides, hydroperoxides and carboxylates[13].

In the oxygen region a multiple peak at ~532eV has been identified. The low binding energy of ~530eV is the characteristic location for oxide oxygen, but also corresponds to oxygen in organic compounds. It can be iron carboxylate [13] or complex compounds. The peak at ~536eV offers no significant information because it is located beyond the interpretative oxygen region. The
(a)
(b)
Figure 4: Results of disk surface EDS analysis: (a) before and (b) after the friction test lubricated with n-hexadecane.

C1s photoelectron spectrum contains peaks at ~288.5eV assigned to the C=0 bond and at ~285eV assigned to the C-0 bond in carboxylates [13,14]. This was also confirmed by XPS analysis of a standard compound (iron acetate).

4. Conclusions

Using T-11 pin-on-disk apparatus, tribochemical changes of n-hexadecane under boundary lubrication conditions were investigated in detail. The research considered both (i) the chemical transformation of the bulk lubricant and (ii) the chemistry of products generated in wear tracks. Gas chromatography/mass spectrometry analysis of n-hexadecane after friction tests show that, during the friction process, the following compounds are formed: aldehydes, alcohols and carboxylic acids. This finding clearly indicates that tribochemical reactions cause significant changes of the apparently non-reactive paraffin hydrocarbon. The reaction process is initiated by the frictional energy. These tribochemical processes lead to the formation of homologous series of the found compounds having smaller molecular weights than n-hexadecane.
Diverse instrumental surface analyses (FTIR, EDS, XPS) of the wear tracks allowed the chemistry of the formed reaction products in the sliding contact area to be determined. The obtained analytical data indicate the possible presence of carboxylates in the analysed deposits.

Acknowledgements

The authors express their appreciation to Mr. Wiktor Wesolowski for GC/MS analysis.
(a)
(b)
(c)
Figure 5: XPS spectrum of: (a) Fe2p, (b) O1s and (c) C1s recorded on tribochemical products of n-hexadecane layered on steel surface.

This work was financed by the Polish Comittee for Scientific Research under grant number 3 T09B 017 18.
    

REFERENCES

[1]
S.M. Hsu et al. in: Proc. of International Tribology Conference (Nagasaki, 2000) p. 25.
[2]
S.M.Hsu and E.E. Klaus, ASLE Trans. 22 (1979) 135.
[3]
A. Wachal, Zagadnienia Eksploatacji Maszyn, 3 (1983) 325.
[4]
C. Kajdas, J. Molenda, M. Makowska and M. Grądkowski, in: Proc. of Symposium on Lubricating Materials and Trihochemistry (Lanzhou, 1998) p. 83.
[5]
J. Molenda, M. Makowska and M. Grądkowski, Problemy Eksploatacji 4 (2000) 167 (in Polish).
[6]
R.M. Silverstein and G.C. Bassler, Spectrometric Identification of Organic Compounds. (Silverstein, Bassler and Morril, New York, 1981).
[7]
W. Zielinski and A. Rajca, Spectroscopic Methods for Organic Compound Identification (WNT, Warszawa, 1995) (in Polish).
[8]
M. Makowska, M. Grądkowski and C. Kajdas, Tribologia 3 (2000) 449 (in Polish).
[9]
Z.S. Hu, S.M. Hsu and P.S. Wang, Trib. Trans. 35 (1992) 189.
[10]
C. Kajdas, M. Makowska and M. Grądkowski, in: Proc. of Nordic Symposium on Tribology (Porvoo, 2000) p. 594.
[11]
W. Czupryk. Tribologia 6 (2000) 1085 (in Polish).
[12]
B. Cavdar and K.C. Ludema, Wear 148 (1991) 329.
[13]
W. Liu, S. Chen, Q. Xue and Z. Zhang, in: Proc. of Symposium on Lubricating Materials and Tribochemistry (Lanzhou, 1998) p. 159.
[14]
L.M. Ng, E. Lyth, M.V. Zeiler and D.L. Boyd, Langmuir 11 (1995) 127.