1 Introduction

Zinc dialkyldithiophosphate (ZDDP) is a well-known additive used in lubricating oils because of its multifunctional anti-wear (AW), extreme pressure (EP) and antioxidant properties. Nowadays, the harmful effect of ZDDP molecule on catalytic converter limits its use as an anti-wear additive for Internal Combustion Engine (ICE) oil. New lubricants with good tribological performances, i.e., exhibiting low friction and low wear, are needed regarding environmental limitations (Norm euros VI). The idea is to reduce the levels of phosphorus and sulphur, specific elements contained in the ZDDP molecule, at the origin of the damage of catalytic converters.

Two options are currently investigated:

  • the development of systems completely different from ZDDP molecules using for example nanoparticles [1, 2].

  • the development of additives with chemical composition close to ZDDP. The objective is to have the best anti-wear protection while limiting the content in phosphorus and sulphur [3–6] in the molecule.

This work focuses on the second option. The understanding of ZDDP anti-wear mechanism is important before going further with modified molecules. A sum up of ZDDP tribochemistry of tribofilm generation is so reported in the following.

The literature provides three main mechanisms explaining the decomposition of ZDDP additive molecule within the lubricant under test conditions. This degradation can be thermal [7–9], hydrolytic [10] or oxidative [11] thanks to hydroperoxides and peroxidic radicals present in the lubricant. Martin [12] estimates that the decomposition of ZDDP is mainly thermo oxidative when temperature exceeds 100 °C.

During this degradation process, ZDDP molecules and their decomposition products are adsorbed physically or chemically on metal surfaces. The deposited film is called ‘thermal film’, and it is further modified under rubbing conditions to generate a protective layer called ‘tribofilm’.

Basically, ZDDP anti-wear additive decreases the wear in a contact running under mixed or boundary lubrication conditions thanks to this tribofilm generation (50–100 nm thick) on rubbing surfaces [13–15]. Surface analyses such as the X-ray Photoelectron Spectroscopy (XPS) showed that the bulk ZDDP tribofilms are mainly composed of a mixed zinc and iron short chain (ortho or pyro) phosphate glass with iron sulphides precipitates [16, 17]. The phosphate chains are longer [17–19] on the top of the tribofilm than in its bulk [20]. Recently, Zhou et al. [21] assumed the presence of ultrapolyphosphate in the outer layer. The ZDDP thermal films have similar composition to ZDDP tribofilms, but consisting mainly of a thinner outer layer of polyphosphate (≈10 nm thick) grading to pyro- or orthophosphate in the bulk [7].

The structural evolution of tribofilm material (i.e. zinc polyphosphate) during tribological solicitation is important to clarify for a better understanding of ZDDP anti-wear mechanism. To investigate structural modification of the material thanks to the effect of hydrostatic pressure, zinc polyphosphate was compressed in Diamond Anvil Cell coupled with in situ Raman or EXAFS analyses [22, 23]. Simulations by quantum chemistry were also carried out [24]. Results suggest a change of coordination number of metallic cation and no polymerisation of phosphate chains. The effect of phosphate glass parameters on their mechanical properties (influence of metallic cations nature, presence of hydroxyl group on phosphate molecules etc…) was also investigated [25, 26].

To insure a strong adhesion of the tribofilm to the substrate and to ‘digest’ iron oxide wear particles, a tribochemical mechanism was reported in literature [12] proposing a reaction between zinc metaphosphate (representative of the top of the thermal film) and iron oxide (representative of the native iron oxide layer). During this reaction, a shortening of phosphate chain length is proposed and was confirmed by friction test on metaphosphate glass [27].

Recent studies [5, 28] showed that the use of zinc orthophosphate powders (crystalline grains of a few microns diameter) is an interesting alternative for anti-wear organic additives. Moreover, this material is free from sulphur and is very close in composition to the main part of the final ZDDP tribofilm. Furthermore, it was showed that the use of zinc orthophosphate powders as an anti-wear additive has the advantage of being effective at the first cycles of friction and at 25 °C [5].

In our study, the additive is close to zinc orthophosphate powder in terms of chemical composition and it has the advantage of being soluble in the base oil thanks to alkyl groups in the molecule. Then, the use of zinc phosphate additive (ZP) is expected to facilitate the formation of a phosphate base tribofilm even at ‘low’ temperature and low concentration because it can avoid the detrimental induction period. Actually, the thermal degradation, which is necessary for the activation of ZDDP molecule through the generation of degradation products, is not necessary using directly ZP additive since it is already the final tribofilm material which is directly introduced in the contact.

The aim of this work is so to compare the anti-wear property of ZP additive in comparison with zinc dithiophosphate (ZDDP).

The first part will relate to the comparison of the anti-wear behaviour of these two molecules at 25 and 100 °C. The second part will deal with the effect of the sliding speed on wear at 25 °C. Finally, a study of the behaviour of ZP tribochemical reaction will be carried out by coupling friction tests with surfaces analyses (AES) at different experiment durations. The role of the concentration is investigated.

2 Materials and Methods

2.1 Lubricants

Three lubricants were tested:

  • A mineral base oil of group III noted BO in the following.

  • A mixture of mineral base oil BO and zinc di-2-ethyl-hexyl dithiophosphate additive (containing 800 ppm of phosphorus) noted ZDDP in the following (Fig. 1a).

    Fig. 1
    figure 1

    ZDDP and ZP molecules

    Full size image

  • A mixture of mineral base oil BO and zinc di-2-ethyl-hexyl orthophosphate (containing 800 ppm of phosphorus) noted ZP in the following (Fig. 1b). It is reminded that the ZP molecule does not contain any sulphur.

2.2 Materials

The balls used were 12.7 mm in radius and 30 nm (RaB) in roughness. The cylinders employed were 6 mm in length and 6 mm in diameter. The rectangular flats measured 10 × 8 × 2 mm3. All these specimens are made of AISI 52100 steel. This iron alloy contains 97 wt% of Fe, 1.45 wt% of Cr, 1.04 wt% of C, 0.35 wt% of Mn and 0.23 wt% of Si. The cylinder and flat specimens were polished using diamond slurry with, respectively, 3 and 1 μm grains. The roughness after polishing of cylinder (RaC) and flat (RaF) are, respectively, 50 and 12 nm.

2.3 Methods

2.3.1 Tribological Parameters

Friction experiments were carried out using a home-made reciprocating cylinder (or ball)-on-flat tribometer [29]. First, experiments were performed in the ball-on-flat configuration to characterise wear behaviour of lubricants. Second, the cylinder-on-flat configuration was used to perform XPS surface analyses because the wear track is larger than the size of the XPS probe. The tribofilms generated under cylinder-on-flat configuration were homogeneous all over the track and of same morphology (patchy) as tribofilms generated under ball-on-flat configuration. It was so considered that tribofilms obtained in both cases were similar in composition (confirmed by XPS analyses not shown here) and morphology. Because the perfect alignment of a cylinder on a flat is difficult, ball-on-flat configuration was more convenient for wear measurements.

The influence of temperature and sliding speed on tribofilm formation was investigated. We choose temperatures similar to those encountered in an Internal Combustion Engine such as starting in ambient condition (25 °C) and in steady-state operation (100 °C).

The ball slides reciprocally on a fixed flat with a frequency of 7 Hz and a stroke length of 7 mm. The applied load for each test is 50 N corresponding to a maximum Hertzian pressure of 928 MPa. For the cylinder-on-flat test, the load was adjusted to obtain the same maximum Hertzian pressure as for the ball-on-flat experiment. The tests were repeated at least twice for each lubricant. The friction coefficient was measured all over the test. In the following, the average of all friction coefficient values measured for one test is reported. Standard deviation is calculated from the different repeated test values. The wear scar diameter on the ball was measured by optical microscopy. The wear tracks obtained were homogenous. The EHL film thickness and lambda ratio are calculated using the Hamrock Dowson formula [30] regarding various sliding speeds and values are reported in Table 1.

Table 1 Physical properties and rheological parameters of the mineral base oil (BO) in different test conditions
Full size table

2.3.2 Surface Analyses

Before any analyse, samples were degreased by rinsing in n-heptane several times in ultrasonic bath.

The tribofilms formed on the flat (using the cylinder-on-flat configuration) were analysed by Auger Electron Spectroscopy (AES) and X-Ray Photoelectron spectroscopy (XPS). Surface analyses were performed under a pressure of 10−7 Pa in the analytical chamber. These techniques provide very surface sensitive information. The depth sensibility is different from one element to another but it is considered to be less than or equal to 10 nm.

The AES analyses were performed using a FEG electron gun 1000 (Thermo Scientific) −5 keV. The electron spot size is about 1 μm, and the lateral resolution is also about 1 μm. For XPS analyses, a monochromator X-ray AlKα source was used. The X-ray probe size (rectangular) is around 1300 μm2. The emission angle is 90° with respect to the horizontal of the sample. The detection is made by the ESCALAB 220i (Thermo Scientific) spectrometer. The spectrometer is calibrated in energy to the 4f7/2 electronic level of gold (Binding energy: 84.0 eV). In a typical XPS analysis, a survey scan is carried out first in order to identify the different elements present in the sample. Then, high-resolution spectra of selected peaks (characteristics of each element) are performed. The deconvolution of these peaks allowed an identification of the different chemical species. Acquisition conditions for the survey spectra were as the following: pass energy of 100 eV, dwell time of 500 ms and step size of 1.0 eV. Concerning acquisition parameters for high-resolution spectra, they were slightly different: pass energy of 20 eV, dwell time of 500 ms and step size of 0.1 eV. The binding energy of carbon (C1s ~ at 284.8 eV) is used as a reference for any charge correction.

Special attention has been paid for fitting P2p, S2p (for ZDDP), O1s, Fe2p3 and Zn2p3/2 photopeaks. CasaXPS [31] software was used for performing the curve fitting procedures on AES and XPS spectra. For XPS, a Shirley background was used and the Lorentzian/Gaussian ratio (L/G) was fixed at 60%.

2.3.3 Transmission Electron Microscopy (TEM)

We used a JEOL 2010F TEM operating with 200 kV accelerating voltage and equipped with an Energy Dispersive X-ray spectrometer (EDX). The cross sections of the near-surface regions of the flat were obtained by the Focused Ion Beam (FIB) method. Before milling, platinum and tungsten layers were deposited on the worn track to preserve the surface from damage due to nano-machining with Ga+ ion beam.

3 Results

3.1 Friction and Wear

First, let us examine the friction and anti-wear performances at low temperature (25 °C). The average friction coefficients measured during tests with ZDDP and ZP lubricants at 25 °C are shown in Table 2. Figure 2 illustrates the curve of friction coefficient versus time for the three different lubricants (BO, ZDDP and ZP) at 25 °C and 100 mm/s.

Table 2 Friction coefficient for both the lubricant (ZDDP and ZP) at 25 and 100 °C with sliding speed of 25 and 100 mm/s
Full size table
Fig. 2
figure 2

Friction coefficient curves as a function of time at 25 °C with sliding speed of 100 mm/s under 0.9 GPa of Hertzian pressure for BO, ZDDP and ZP lubricants

Full size image

Wear scar diameters were measured on the balls for the three lubricants: BO, ZDDP and ZP, and the tests were carried out at 25 °C (room temperature) with sliding speeds of 100, 50 and 25 mm/s, respectively.

At 25 °C and 100 mm/s, friction coefficient (Table 2) obtained with ZDDP lubricant (0.119 ± 0.001) is approximately equal to the one obtained with ZP lubricant (0.117 ± 0.001). Concerning the anti-wear efficiency at 100 mm/s, Fig. 3 and Table 3 show the ball wear track diameters obtained after ball-on-flat test. As we can see, at room temperature, there is no significant difference between the anti-wear properties for the three lubricants. This is attributed to a predominant EHL/mixed lubrication regime at this temperature (see elevated film thickness and lambda ratio in Table 1). To increase contact severity, we divided the sliding speed by two (50 mm/s) but the sliding distance (360 m) was kept the same as for the experiment at 100 mm/s. Figure 3 and Table 3 also present the wear results obtained at 50 mm/s. They indicate a slightly better anti-wear effect for ZP lubricant. We further increased the severity by using a sliding speed of 25 mm/s (always with the same sliding distance of the ball). Figure 3 and Table 3 illustrate too the wear results obtained for ZDDP and ZP lubricants at 25 °C and 25 mm/s. It can be noticed that wear is much higher with BO (813 ± 22 μm) than with ZDDP (448 ± 90 μm) and ZP (344 ± 38 μm) lubricants due to the occurrence of the boundary regime. Concerning the friction, there is no noticeable difference in friction coefficient (~0.12) between ZDDP and ZP lubricants. However, this small sliding speed allowed us to discriminate clearly anti-wear performances with the ZDDP and ZP additives at room temperature. The overall results indicate a much better anti-wear performance for the ZP lubricant at low temperature and low speed in the boundary regime. Furthermore, optical observations of flat wear tracks for ZDDP and ZP lubricants (25 mm/s—25 °C) presented in Fig. 4 suggest also a better anti-wear behaviour of ZP additive as the number of tribofilm pads is higher for ZP than for ZDDP lubricant.

Fig. 3
figure 3

Ball wear track diameters after ball-on-flat tests at 25 °C with sliding speed of 25, 50 and 100 mm/s under 0.9 GPa of Hertzian maximum pressure for lubricants BO, ZDDP and ZP

Full size image
Table 3 Ball wear track diameters after ball-on-flat test at 25 °C with sliding speed of 25, 50 and 100 mm/s under 0.9 GPa of Hertzian maximum pressure for lubricants BO, ZDDP and ZP
Full size table
Fig. 4
figure 4

Optical images of flat wear tracks (Ball-on-Flat configuration) obtained at 25 and 100 °C with sliding speed of 25 and 100 mm/s under 0.9 GPa of Hertzian maximum pressure with ZDDP and ZP lubricants

Full size image

At 100 °C and 100 mm/s of sliding speed, compared with BO (685 ± 168 μm), wear on the balls for ZDDP (342 ± 8 μm) and ZP (418 ± 59 μm) lubricants drastically decreases (Table 4). The friction coefficients obtained with the three lubricants are summarised in Table 2. The optical images of flat wear tracks obtained with ZDDP and ZP lubricants at 100 mm/s are represented in Fig. 4. They show the presence of coloured patchy tribofilms typical of anti-wear action of this kind of P-containing additives. Looking at calculated EHL film thickness (11.8 nm) and lambda ratio (0.27) at 100 °C, we can assume that the lubrication regime at 100 mm/s sliding speed is predominantly boundary. The anti-wear performances of ZP (418 ± 59 μm) and ZDDP (342 ± 8 μm) lubricants are close, although ZDDP molecule exhibits a slightly better anti-wear behaviour Fig. 5.

Table 4 Ball wear track diameters after ball-on-flat tests at 100 °C with sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure for lubricants BO, ZDDP and ZP
Full size table
Fig. 5
figure 5

Ball wear track diameters after ball-on-flat tests at 100 °C with sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure for lubricants BO, ZDDP and ZP

Full size image

3.2 AES and XPS Analyses

Let us examine surface chemistry of tribofilm at low temperature (25 °C) and low sliding speed (25 mm/s) where ZP was found much better than ZDDP (Fig. 3). Phosphorus, sulphur (detected only for ZDDP) and zinc are found in AES spectra performed on ZDDP and ZP tribofilms (Fig. 6). Iron is detected in the case of ZDDP only. Oxygen is mainly in oxide form (peak OKLL ~ 512 eV) in the ZDDP tribofilm and in a phosphate form (peak OKLL ~ 507 eV) for the ZP case. ZDDP tribofilm at room temperature consists of a mixture of zinc and iron phosphate with probably iron oxide and metallic sulphides [32]. On the other hand, the ZP tribofilm is made of zinc phosphate only.

Fig. 6
figure 6

Auger Spectra of ZDDP and ZP tribofilms obtained at 25 °C with a sliding speed of 25 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image

XPS spectra carried out on ZDDP and ZP tribofilms obtained at room temperature and 25 mm/s (Fig. 7) also display phosphorus, sulphur (with ZDDP) and zinc. The O1s peaks from ZP and ZDDP tribofilms show two contributions indicating that oxygen is involved mainly in phosphate form (531.6 eV (P–O) and 533.2 eV (P–O–P)) and another contribution is attributed to oxide form (529.6 eV) [17, 32]. However, this last contribution is found in very small amount and is negligible considering the fact that it is close to the detection limit.

Fig. 7
figure 7

XPS spectra of ZDDP and ZP tribofilms obtained at 25 °C with a sliding speed of 25 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image

Finally, the ZDDP and ZP tribofilms also consists mainly of a mixture of zinc and iron phosphate, with sulphide (162.3 eV) in case of ZDDP tribofilms. Iron oxide is also detected but AES and XPS results are not in total agreement in the case of ZDDP tribofilm. A strong iron oxide contribution was clearly found on the AES analysis but this was not so obvious on the XPS analyses. As the analysed area with AES technique (≈1 μm2) is much smaller than with XPS (≈1300 μm2), this difference is attributed to the local tribofilm heterogeneity.

Auger spectra of ZDDP and ZP tribofilms obtained at 100 mm/s and at 100 °C are shown in Fig. 8. The characteristic elements of the additives are detected in both ZP and ZDDP tribofilms: phosphorus, sulphur (detected only for ZDDP) and zinc. Oxygen is clearly in the phosphate chemical form (peak OKLL ~ 506 eV). No iron is detected at the top of the two tribofilms. The results indicate that both ZDDP and ZP tribofilms are made of zinc phosphate (probably with some metallic sulphides for ZDDP).

Fig. 8
figure 8

Auger Spectra of ZDDP and ZP tribofilms obtained at 100 °C with a sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image

XPS analyses were performed in the same tribofilm locations as for AES analyses. The advantage of XPS is to provide semi-quantitative elementary analysis. Figure 9 shows the general survey (SG) and O1s spectra of ZDDP and ZP tribofilms obtained at 100 °C and 100 mm/s. The O1s peak from ZDDP tribofilm shows two contributions indicating that oxygen is involved mainly in phosphate form (531.6 eV (P–O) and 533.2 eV (P–O–P)) and another contribution attributed to oxide form (529.6 eV). However, this last contribution is found in very small amount (about 1.6 at % cf. Table 5) and is negligible considering the fact that it is close to the detection limit. Two small contributions of oxygen linked to carbon are also detected at 531.6 eV (C=O) and 533.2 (C–O) at same positions as phosphate peaks. Taking into account semi-quantification of corresponding carbon peak deconvolution, these two contributions are expected to be of few atomic percentages. The O1s peak from ZP tribofilm shows the contributions of oxygen in phosphate form (531.6 eV (P–O) and 533.2 eV (P–O–P)) with no oxide form. S2p binding energy from ZDDP additive corresponds to metallic sulphides (ZnS, FeS, FeS2…) [16, 17] and is detected only in ZDDP tribofilm. The Table 5 shows the quantification of component detected on ZDDP and ZP tribofilms.

Fig. 9
figure 9

XPS spectra of ZDDP and ZP tribofilms obtained at 100 °C with a sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image
Table 5 XPS quantification (% at) of ZDDP and ZP tribofilms obtained at 100 °C with a sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure
Full size table

The overall results of AES and XPS studies clearly show that ZDDP and ZP tribofilms formed at 100 °C and 100 mm/s consists of a zinc phosphate (with sulphide (162.3 eV) in the case of the ZDDP).

Figure 10 shows the TEM images of FIB cross sections for ZDDP and ZP tribofilms obtained at 100 °C and 100 mm/s. The tribofilms formed on steel is about 60 nm thick for both additives. The EDS spectra carried out on both ZDDP and ZP tribofilms confirm the elemental composition previously obtained by AES.

Fig. 10
figure 10

TEM observations (a1 et b1) and EDX spectra (a2 and b2) of the FIB cross section of ZDDP (a) and ZP (b) tribofilms obtained at 100 °C with a sliding speed of 100 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image

3.3 Wear Behaviour of ZDDP and ZP Molecule for Various Sliding Distances

Some additional tribological experiments (ball-on-flat) were performed at various sliding distance to study wear behaviour of ZDDP and ZP molecules during the tribofilm formation. The study was carried at 25 °C—25 mm/s (Fig. 11a) in the low-temperature regime where ZP was found more efficient and at 100 °C and 100 mm/s (Fig. 11b). Figure 11 presents wear results obtained at various sliding distance for the two additives at 25 °C—25 mm/s and at 100 °C—100 mm/s. Figure 11a presents wear results obtained at various test durations (4, 20, 120 and 240 min) corresponding, respectively, to different sliding distances (6, 30, 180 and 360 m) for the two additives at 25 °C and 25 mm/s. For ZP and ZDDP additives, the wear obtained at the beginning of the experiment (sliding distance = 6 m) is very small and close to the calculated Hertzian diameter. It is the same for ZP even after 360 m of sliding. However, for ZDDP, we observe a wear increase after 360 m of sliding. The Auger analyses of the wear scar show the presence of additive elements after 6 m of sliding in ZDDP and ZP tribofilms (Fig. 12), although iron is detected in each case. After 360 m of sliding, iron is detected in the ZDDP tribofilm but not for ZP. Moreover, friction tests were performed for 1–60 min at 100 mm/s of sliding speed and 100 °C. Experiment durations were adjusted in order to have same sliding distances (6 and 360 m) as for the experiment at 25 mm/s. For ZP and ZDDP additives, the wear obtained at the beginning of the experiment (sliding distance = 6 m) is small and close to Hertzian diameter. This wear increases slightly for ZDDP after 360 m of sliding. For ZP additives, same feature is found at the beginning of tests but wear is found to increase a little more after 360 m of sliding.

Fig. 11
figure 11

Ball wear track diameters after ball-on-flat tests at 25 and 100 °C with different sliding distance for ZDDP and ZP lubricants

Full size image
Fig. 12
figure 12

Auger spectra obtained on ZDDP and ZP tribofilms after 6 and 360 m of sliding at 25 °C with a sliding speed of 25 mm/s under 0.9 GPa of Hertzian maximum pressure

Full size image

3.4 Effects of ZP and ZDDP Concentrations

We also focused on the effect of ZDDP and ZP concentrations on their anti-wear efficiency at room temperature and 25 mm/s of sliding speed for 240 min. Several dilutions were made by mixing a certain volume of base oil with a volume of ZDDP or ZP lubricants (25 and 75% of base oil corresponding, respectively, to 600 and 200 ppm of P, respectively). Figure 13 shows wear scar diameters on balls at the three concentrations and at room temperature. As can be seen for the most dilute solution (75% BO—200 ppm of P), the wear scar diameter obtained with ZDDP lubricant (~358 ± 0.1 μm) is significantly higher than with ZP lubricant (~321 ± 3.0 μm). The comparison of ZDDP and ZP lubricants at low concentration gives evidence for the better anti-wear property of ZP at room temperature.

Fig. 13
figure 13

Ball wear track diameters after ball-on-flat tests at 25 °C with sliding speed of 25 mm/s under 0.9 GPa of Hertzian maximum pressures for lubricant with different additives concentration (200 and 600 ppm of phosphorus)

Full size image

4 Discussion

At 100 °C and 100 mm/s for 60 min, both ZP and ZDDP molecules exhibit anti-wear capabilities (ZDDP molecule is slightly better) as well as similar tribofilms compositions. The absence of sulphur in ZP molecule does not inhibit the tribofilm formation. The origin of anti-wear capabilities of such phosphorus-based additives is probably the same for the two molecules. It is important to notice that we did not check the efficiency of these additives in the EP regime. Different mechanisms are proposed in literature related to specific tribochemical reaction pathways [12, 19] or specific tribofilm material modification under solicitations [24].

In the first case (tribochemical reactions), it is proposed that ZDDP molecule and its degradation products react under boundary conditions with native iron oxide of steel surfaces to form mixed zinc and iron phosphate glass [12]. Thanks to this tribochemical pathway, the tribofilm adheres well on metal surfaces. Additionally, any iron oxide particle trapped in the contact will lose its abrasive character when being digested in the phosphate tribofilm. In case of more severe lubrication conditions (extreme pressure), the tribofilm could not stay in the contact and a different tribochemical reaction occurs between iron metal and sulphur species. Metallic sulphides are generated in the contact. This last reaction is explaining the extreme pressure capabilities of ZDDP molecule.

Concerning the second mechanism at the origin of anti-wear capabilities of such material, it is related to interfacial material modification (zinc phosphate) under solicitations. A change of zinc atoms coordination number is proposed under hydrostatic pressure [24] and could so contribute to the modification of tribofilm mechanical properties under solicitations. The effect of shearing was not investigated.

In our experimental conditions, as the tribofilms generated with both additives are made of phosphate materials, same kind of anti-wear mechanisms can be proposed for both molecules. The tribochemical reaction of polyphosphate with native iron oxide could explain the adhesion of the tribofilm on the substrate and the loss of abrasive character of wear particles. Although our experiments do not allow any conclusion about structural changes during the tribological solicitations, same kind of modifications are expected for ZP and ZDDP tribofilms. Concerning extreme pressure conditions, no ZP activity is expected as there is no sulphur in the molecule.

At 25 °C and 25 mm/s for 360 m of sliding, tribological tests results show that ZP molecule exhibits a better anti-wear behaviour than ZDDP. At lower concentration (200 ppm of phosphorus—25 °C—25 mm/s—240 min), ZP additive shows an even better anti-wear behaviour than ZDDP. As it was proposed first (§1), the ZP molecule is closed in composition to tribofilm final material, so ZP molecule does not need to follow the thermo-oxidative degradation pathway of ZDDP molecule, avoiding high wear rate during the detrimental induction period of ZDDP. This makes the ZP molecule more reactive and more efficient than ZDDP at 25 °C, 25 mm/s and 360 m of sliding.

To conclude, ZP molecule has an advantage compared to ZDDP in terms of anti-wear capabilities at 25 °C. At 100 °C, it is ZDDP molecule that exhibits a better anti-wear behaviour. A mixture of both molecules is so an interesting option to get a good compromise in terms of anti-wear capabilities in the range of 25 and 100 °C with the same amount of phosphorus and with a small amount of sulphur atoms in the lubricant.

5 Conclusion

The comparison of ZP anti-wear performance with ZDDP allows us to conclude that

  • although the anti-wear efficiency of these two molecules is found at 100 mm/s of sliding speed and 100 °C, ZDDP exhibits a slightly better anti-wear behaviour than ZP at this temperature. For both additives, tribofilms are mainly made of zinc and iron phosphate.

  • at room temperature and 25 mm/s of sliding speed, ZP is a better anti-wear additive

  • at room temperature and 25 mm/s of sliding speed, ZP is able to protect steel surfaces from wear even at 200 ppm of phosphorus. In these conditions, ZDDP is not so active.

Taking into account all these data, we show that ZP is an interesting anti-wear additive for the lubrication of Internal Combustion Engines at ambient temperature, which is the characteristic of cold engine start. In steady-state conditions, ZDDP molecule is more efficient than ZP. The combination of both additives (keeping a small amount of P) is a good option to optimise the anti-wear capabilities of engine lubricants. The efficiency of these additives in the EP regime was not studied. However, the loss of extreme pressure properties and antioxidant properties with ZP molecule is expected and requires the addition of other molecules in the lubricant completely formulated.