Development of an accelerated artificial ageing method for the characterization of degradation products of antioxidants in lubricants by mass spectrometry

Abstract

The understanding of ageing mechanisms of antioxidants in base oils is indispensable for the development of improved lubricants. In this study, a novel artificial ageing method based on the application of peroxide as oxidant is presented for improved monitoring of thermo-oxidative degradation processes in combination with mass spectrometry. Model oils containing aminic and phenolic antioxidants were aged and chemical structures of their oxidation products were elucidated by ultrahigh performance liquid chromatography and electrospray ionization high resolution (Orbitrap) mass spectrometry. Additionally, synergistic mixtures of four antioxidants were investigated, because the formation of condensed molecules from amines and phenols would have a major influence on the antioxidant potential but could not be detected in the bulk lubricant.

Keywords

Tribochemistry antioxidants lubricants artificial ageing UHPLC electrospray ionization orbitrap mass spectrometry

Introduction

The primary limiting factor for the performance and the lifetime of lubricants is the ageing of the lubricants and particularly their major as well as minor components (base oil and additives). In practical applications, the main cause for the ageing is the thermo-oxidative degradation processes. Common effects of the degradation of the base oil are oil thickening due to the rising viscosity and an increased corrosion as a consequence of acidification. The applied essential additives are also degraded through the ageing. In addition, the formation of sludge and varnish deposits is induced by the polymerization and polycondensation of the degradation products of the base oil and the applied additives. These progressive effects lead finally to a functional failure of the lubricant.^1,2 This is challenging because modern design requirements for engines emphasize miniaturization (reduced lubricant volume) in combination with an increased power density. Hence, the thermal stress on lubricants has risen steadily over the last decades. Nevertheless, requirements for the lifetime of lubricants have been extended considerably.³ To increase the time interval of lubricant/oil replacements as well as to elongate the lifetime of the lubricants, they are stabilized by the addition of antioxidants. Antioxidants are a class of additives featuring certain chemical structural characteristics. These characteristics enable them to terminate the free-radical chain reaction – the base of the oxidation process. In doing so, they are depleted over time by forming their own specific degradation products. After the reservoir of antioxidants is consumed, an unrestrained thermo-oxidative degradation of the lubricant takes place, leading to a fast lubricant failure.^1,2

For the development of improved antioxidants and the (re)formulation of lubricants, a better understanding of ageing mechanisms and the synergisms of additives on the molecular level is necessary. For a profound investigation of the oxidation mechanism of antioxidants, two requirements have to be fulfilled: (i) The selection of an adequate analytical technique, able to determine detailed structural information regarding the degradation products. (ii) The development of an accelerated and reliable artificial ageing method for the lab enabling the reproducible monitoring of thermo-oxidative degradation processes.

The first requirement can be fulfilled by utilizing mass spectrometry (MS). Previous results^4,5 showed that modern MS in combination with an electrospray ionization (ESI) source for these analytes can be utilized for the identification and characterization of antioxidants in solution (e.g. in the bulk lubricant). An advantage is that most lubricant molecules (base oil) cannot be desorbed/ionized by means of ESI. Hence, their influence on the detection of the antioxidants is minimized as they are not accessible by ESI approach. MS is an analytical technique for measuring the mass-to-charge ratio (m/z) of charged analytes. The measurements enable the calculation of the molecular masses of these analytes as well as the determination of their elemental compositions.⁶ High-resolution accurate mass measurements as they are provided by Orbitrap section of the linear ion trap (LIT)-Orbitrap hybrid MS facilitate in most cases an unambiguous determination of the elemental composition. High resolution and accurate mass measurements are also especially useful to differentiate between so called isobaric ions (ions with different elemental compositions, but same nominal mass values). Therefore, characterization is enhanced considerably by this kind of instrumentation.^7–9 Under most conditions, MS enables the elucidation of the chemical structure as well. This structure elucidation can be improved by the investigation of the fragmentation of selected analyte (additive molecules and their degradation products) precursor ions.⁶ The LIT-Orbitrap MS provides in the LIT section of the instrument low energy-collision-induced dissociation (LE-CID) for fragmentation experiments (termed CID-MS²).⁹ By combining the determined elemental compositions and the elucidated structures of the compounds, the unambiguous identification of an analyte is enabled and the identification as well as characterization of unknown structural analogues like oxidative degradation products is improved.⁵ The detailed structural information obtainable renders MS indispensable for the investigation of thermo-oxidative degradation processes, the formed products and their probable reaction mechanisms.

The development of an appropriate artificial ageing method as the second requirement is the main topic of this publication. Artificial ageing methods are defined as methods to accelerate the oil degradation by applying harsh physical and chemical conditions. They are used to determine performance characteristics of lubricants like the thermo-oxidative stability. Although these methods are artificial degradations, the applied conditions should be selected to simulate oil degradation mechanisms under real engine operating conditions.¹⁰

Established standard methods like CECL-48-A00¹¹ or DIN 51352¹² degrade the lubricants under elevated temperatures (>100℃) by utilizing molecular oxygen from air as oxidant. These methods are not suitable for the detailed investigation of the thermo-oxidative degradation mechanisms. They were developed for the determination of physical oil characteristics after specific artificial degradation times. Hence, it is not possible to monitor the progress of the degradation by frequently sampling the degrading lubricant without modifying the methods. Furthermore, they require a large sample volume. However, their main disadvantages are still the long test durations of a couple of days or even longer. To decrease the test duration time, it is possible to add metallic catalysts for an acceleration of the reaction as it is used for instance in the standards DIN 51554¹³ and ASTM D.4742-08e1.¹⁴ However, the required method must not be based on the use of catalysts, because one of the long-term aims of the investigation of the thermo-oxidative degradation mechanisms also would be to study the influence of catalysts on the formation on degradation products.

An alternative for accelerating the reaction would be the use of a stronger oxidant. Hence, a novel artificial ageing method based on utilizing a selected peroxide – soluble in nonpolar solvents like lubricants (base oil) – was developed. Because no oxygen is needed, the sample can be aged under air exclusion, for instance in airtight sealed ampoules. The use of disposable glass ampoules has multiple benefits. It apparently prohibits contaminations due to the reuse of the ageing device. The complete lubricant sample is distributed to multiple ampoules. Hence, for each sampling, one fresh ampoule can be taken. This enables sampling for the monitoring of the degradation without affecting the remaining lubricant sample. The small size of the ampoules enables the parallel ageing of a multitude of samples in a drying oven. It also reduces the sampling aliquot to the necessary minimum. The amount of lubricant needed for the investigation of a complete ageing process is also reduced considerably, since only the amount really needed for the sampling is utilized in this method. Another benefit of the airtight sealed ampoules is the possibility of the investigation of the thermo-oxidative degradation mechanisms of semi- or volatiles like low-molecular mass antioxidants. This is not feasible with common artificially ageing methods. Finally, the increased oxidative power of the peroxide enables also the oxidative degradation of antioxidants which are too stable for the degradation in a practical time frame by common artificially ageing methods.

The method was developed for the investigation of the thermo-oxidative degradation and the specifically formed products of aromatic amines (e.g. dialkylated diphenylamine) and sterically hindered phenols (e.g. 4-methyl-2,6-di-tert-butyl-phenol (butylated hydroxyl toluene, BHT), two groups of antioxidants commonly applied in lubricants.¹ Often, a combination of both types of antioxidants is applied, because it increases the antioxidant effect compared to the individual antioxidants. The so-called synergism is the result of the different reactivity of the two types of antioxidants. The more reactive aromatic amine scavenges a radical and is regenerated by the slower reacting sterically hindered phenol.^1,2 This reaction is well established. The novel artificial ageing method was used to investigate if these two types of additives show additional combined reactions under thermo-oxidative conditions like, e.g. the formation of condensed molecules. This would have a major influence on the antioxidant potential of these additives.

The present study shows the benefits of the newly developed artificially ageing method in combination with the analytical abilities of high-resolution, accurate ESI MS. Together, they enable the fast investigation of thermo-oxidative mechanisms and the characterization of the formed degradation products in the bulk lubricant. Additional benefits of this new strategy were shown on the examples of a volatile antioxidant and an antioxidant with high oxidative stability. A long-term aim is the enhancement of this fast novel artificial ageing method. It should also enable the sampling of the deposits of sludge and varnish formed from thermo-oxidative degradation processes. The investigation of such samples by MS in general will further improve the understanding of ageing mechanisms on a molecular level. The analysis of the deposits of synergistic combinations of antioxidants will show if they form insoluble condensed products, which precipitate immediately.

Experimental

Chemicals and reagents

N, N′-di-sec-butyl-1,4-benzenediamine (1) was purchased from Sigma-Aldrich (St. Louis, MO, USA). 4-methyl-2,6-di-tert-butyl-phenol (butylated hydroxyl toluene, BHT) (2) and 4-[(4-hydroxy-3,5-di-tert-butyl-phenyl)methyl]-2,6-di-tert-butyl-phenol (3) were provided by Evonik DEGUSSA (Essen, Germany). 2-[2-[3-(4-hydroxy-3,5-di-tert-butyl-phenyl)propanoyloxy]ethylsulfanyl]ethyl 3-(4-hydroxy-3,5-di-tert-butyl-phenyl)propionate (4) was obtained from Ciba (Basel, Switzerland). The latter three compounds were described for use as additives in lubricants. The elemental compositions, theoretical monoisotopic molecular weights and structures of these selected antioxidants are shown in Table 1 and Figure 1, respectively.

Table 1.

List of selected antioxidants.

Compound	Chemical name	Elemental composition	Monoisotopic MWt. [Da]^a
1	N,N′-di-sec-butyl-1,4-benzenediamine	C₁₄H₂₄N₂	220.1939
2	4-methyl-2,6-di-tert-butyl-phenol	C₁₅H₂₄O	220.1827
3	4-[(4-hydroxy-3,5-di-tert-butyl-phenyl)methyl]-2,6-di-tert-butyl-phenol	C₂₉H₄₄O₂	424.3341
4	2-[2-[3-(4-hydroxy-3,5-di-tert-butyl-phenyl)propanoyloxy]ethylsulfanyl]ethyl 3-(4-hydroxy-3,5-di-tert-butyl-phenyl)propanoate	C₃₈H₅₈O₆S	642.3954

Molecular weights shown are based on theoretical calculations utilizing monoisotopic masses.

The base oils PAO-8 (group IV polyalphaolefin oil) and SN150 (group I mineral oil) were provided by OMV (Vienna, Austria). Methanol (LiChrosolv grade), 2-propanol (LiChrosolv grade), acetic acid (100%, analytical grade, p.a.), sodium acetate (p.a.), potassium acetate (p.a.), aluminum oxide active basic, diethyl disulfide (p.a.) and water (p.a.) were purchased from Merck (Darmstadt, Germany). N-hexadecane (anhydrous, ≥99%), toluene (anhydrous) and tert-butyl hydroperoxide (TBHP, 70% in H₂O) were acquired from Sigma-Aldrich (St. Louis, MO, USA). The TBHP was dehydrated by the means of a toluene/water azeotropic distillation.

Artificial ageing methods

A novel artificially ageing method for the studying of the oxidative stability of antioxidants was developed with emphasis to a short test duration. This method is termed “peroxide-ageing” method since the applied oxidant is peroxide. Most common artificial ageing methods like the standard method CEC L-48-A00¹¹ on the contrary utilize oxygen from air as oxidant. Therefore, they are termed air-flow-ageing methods in this publication. Their main disadvantages are the long test durations of many days or even longer. Both types depend on elevated temperatures (>100℃) to accelerate the degradation.

Novel peroxide-ageing method

The most evident difference of the novel peroxide-ageing method is the use of an airtight sealed glass ampoule as reactor. TBHP was applied as oxidant, because this peroxide is readily soluble in nonpolar solvents like the base oil. The dried TBHP in toluene is added to the sample before sealing of the ampoule. Preliminary results during the method development showed that best results are obtained by applying the peroxide at a molar ratio of 1:1 with prospect to the antioxidant. The samples are aged at 160℃ in an oven in the dark. The conditions of the peroxide-ageing method are shown in the first column of Table 2.

Table 2.

List of conditions for the two artificial ageing methods: The novel peroxide-ageing method and the air-flow-ageing method as example for a common ageing method.

	Peroxide-ageing method	Air flow-ageing method
Ageing device	10 mL ampoule (sealed)	500 mL three-neck glass flask equipped with a cooling device
Size of sample	7 mL/ampoule (= 49 mL for 7 samples)	300 g lubricant
Temperature	160℃	160℃
Oxidant	tert-Butyl hydroperoxide in toluene	Compressed, dried air
Addition of the oxidant	at begin of reaction (at a molar ratio of 1:1 to the antioxidant)	Continuous (flow rate = 10 L/h)
Sampling aliquot	1 ampoule (7 mL)	15 mL
Sampling time	0, 1.5, 3, 6 15, 24, (72) h	0, 4.5, 7.5, 21, 27, 32, 45, 69, 80, 96, (125, 142, 144, …) h
Test duration	min. 24 h	min. 96 h

The summarized main advantages of this newly developed method are: (a) The test duration is decreased to 24 h. Control samples after 72 h had not shown any major changes. (b) Sampling does not affect the rest of the sample, as for each sample taken a new ampoule is removed/used. (c) The sampling aliquot is reduced to the necessary minimum. (d) The volume of sample necessary for a complete ageing experiment is reduced by an order of magnitude (based on seven sample time point collections). (e) The method enables artificial ageing experiments of even volatile antioxidants. (f) Oxidative degradation of very stable antioxidants is feasible due to the increased oxidative power of the peroxide.

Air-flow-ageing method

The performance of the peroxide-ageing method was evaluated by comparison with an air-flow-ageing method published by Besser et al.¹⁰ This further developed form of the standard CEC L-48-A00 enables sampling, necessary for the monitoring of the degradation process, and the addition of contaminants throughout the process. The latter feature was not used in this study. The conditions of this method are shown in the second column of Table 2.

Artificial ageing matrix

For the development and evaluation of this new artificial peroxide-ageing method, well-defined model oils, namely oils J–M were prepared as listed in Table 3. The model oils A–C were used for the development of the method. To minimize the influence of the base oil, n-hexadecane was used as the model base oil. For the air-flow-ageing method, two sets of model oils were prepared in different base oils, thus enabling the determination of the influence of the base oil on the ageing process. The model oils D–F were prepared with a high purity polyalphaolefin PAO-8 representing a synthetic base oil, whereas the model oils G–I were prepared in the group I base oil SN150. The model oils J–M – prepared with the model base oil n-hexadecane – were used to show the advantages of the new method, since these four model oils are not feasible for a successful artificial alteration with the air-flow-ageing method.

Table 3.

Matrix of the model oils prepared for the evaluation of the two novel artificial ageing methods (A–I), for the ageing of a volatile antioxidant (J–K) and for the ageing of an antioxidant exhibiting a high antioxidant potential (L–M).

			Antioxidant
			1		2		3		4
Model oil	Artificiale aged by	Model base oil	µmol/L	% (w/w)	µmol/L	% (w/w)	µmol/L	% (w/w)	µmol/L	% (w/w)	Comment
A	Peroxide-ageing method	n-hexadecane	7	0.2	–	–	–	–	–	–	Amine
B		n-hexadecane	–	–	–	–	7	0.4	–	–	Phenol
C		n-hexadecane	3.5	0.1	–	–	3.5	0.2	–	–	Synergistic mixture
D	Air flow-ageing method	PAO-8	7	0.2	–	–	–	–	–	–	Amine
E		PAO-8	–	–	–	–	7	0.4	–	–	Phenol
F		PAO-8	3.5	0.1	–	–	3.5	0.2	–	–	Synergistic mixture
G	Air flow-ageing method	SN-150	7	0.2	–	–	–	–	–	–	Amine
H		SN-150	–	–	–	–	7	0.4	–	–	Phenol
I		SN-150	3.5	0.1	–	–	3.5	0.2	–	–	Synergistic mixture
J	Peroxide-ageing method	n-hexadecane	–	–	7	0.2	–	–	–	–	Volatile phenol
K		n-hexadecane	3.5	0.1	3.5	0.1	–	–	–	–	Synergistic mixture
L		n-hexadecane	–	–	–	–	–	–	7	0.6	Sulfide containing phenol
M		n-hexadecane	3.5	0.1	–	–	–	–	7	0.6	Synergistic mixture

The model oils A, B, D, E, G, H, J and L contain each one antioxidant, whereas the model oils C, F, I, K and M contain a mixture of an aminic and a phenolic antioxidant at a molar ratio of 1:1 for the study of the synergistic effects between the two types of antioxidants. For a better comparability, the total antioxidant content in all 12 model oils was set to 7 µmol/L. Hence, C, F, I, K and M contain 3.5 µmol/L of the amine and the phenol, respectively. The resulting weight contents of the antioxidants between 0.1 and 0.6% (w/w) are all in a common range (<1% (w/w)).

As aromatic aminic antioxidant for the model oils A, C, D, F, G, I, K and M the rather uncommon aromatic amine 1 was applied. It was used because it is obtainable as a high-purity and well-defined substance. This allows the direct identification and characterization of the oxidative degradation products by MS. Due to the fact that the most common aminic antioxidant butylated, octylated diphenylamine is only available as a complex mixture of at least four structural analogues,^5,15 it was not suitable for the development of a new method. As sterically hindered antioxidant, the diphenol 3 was applied in the model oils B, C, E, F, G and H. The more common phenol 2 could not be used for the method development, because it is too volatile and therefore evaporated in air-flow-ageing methods.¹⁶ Therefore, compound 2 was used only in the model oils J and K to show the advantage of enabling the artificially ageing of volatile compounds by the artificial peroxide-ageing method. The sulfide containing phenol 4 was applied in the model oils L and M. The oxidative stability provided by this antioxidant is too high for thermo-oxidative degradation with an artificial air-flow-ageing method within a reasonable time.¹⁶ The novel artificial peroxide-ageing method on the other hand showed the ability to degrade this compound.

Oil condition monitoring

To monitor the changes of the model lubricants during the ageing, several samples were taken and analyzed. Visual condition monitoring was applied as the simplest and most common method to obtain information about the degradation process. The relative contents (%) of antioxidants in the model oils were calculated by determination of the intensity of the absorbance bands for amines (–NH: 1515 cm⁻¹) and phenols (–OH: 3651 cm⁻¹) with the Fourier transformation infrared spectrometer (FTIR) Tensor 27 (Bruker, Billerica, MA, USA). The increase of the oxidation value – defined as absorption referenced to a one-centimeter oil layer thickness (A/cm) – as an indicator for the thermo-oxidative degradation was also monitored utilizing this instrument by determination of the absorbance for the wavenumber 1710 cm⁻¹ as described in the standard DIN 51453.¹⁷

MS

To identify and characterize the antioxidants and their specific degradation products, the samples were diluted in methanol: 2-propanol (1:1, v/v) to a final concentration of 100 ppm. High-resolution and accurate mass analyses were performed in positive and negative ion mode by means of a LTQ Orbitrap XL hybrid mass spectrometer (ThermoFisher Scientific, Bremen, Germany). It was equipped with an IonMax API ion source for ESI measurements. The following experimental conditions were applied: spray voltage, ±4.0 kV; capillary exit, tube lens, skimmer and multiple voltages tuned for maximum singly charged precursor ion transmission; transfer capillary at 275℃; sheath gas for nebulizing, nitrogen with a flow rate of 5 to 15 arbitrary units (no conversion in SI units is provided by the manufacturer); automatic gain control, on. For LE-CID fragmentation experiments (CID-MS²), the following parameters were selected: isolation and activation width, 4 Da; q_z-value, 0.25; activation time, 30 ms; collision gas, helium; collision gas pressure, 2.6 × 10⁻³ Pa; normalized collision energy, 30%.

High-resolution (HR) accurate mass spectra were obtained by the Orbitrap section at a resolution (full with at half maximum) of 30,000 at m/z 400 with a maximum injection time of 200 ms. The achieved mass accuracy was below 2 ppm (root mean square) with external calibration in all measurements.

Data acquisition and processing were accomplished with the software Xcalibur v2.0 (ThermoFisher Scientific, Bremen, Germany).

Ultrahigh performance liquid chromatography-mass spectrometry

For the separation of the isobaric degradation products of the sulfur-containing phenolic antioxidant 4 a.

Rheos Allegro ultrahigh performance liquid chromatograph (UHPLC) quaternary pump (for pressure ≤1000 bar) with a quaternary zero hysteresis constant run degasser (all from Thermo Scientiﬁc, Bremen, Germany), an HTS PAL autosampler (CTC Analytics, Zwingen, Switzerland) equipped with a six-port Cheminert injection valve (0.25 mmm ID, for pressure ≤ 1000 bar) from VICI (Schenkon, Switzerland), a ThermaSphere TS-130 column heater (Phenomex, Torrance, CA, USA) coupled to the ESI-MS was used. The system and the applied method for the separation of aromatic amines and selected phenols in positive ion mode are described in great detail (mobile phases and gradient) in a previous publication.¹⁵ Briefly, an Acquity UPLC BEH C18 column (50 × 2.1 mm column dimensions, 1.7 µm particle diameter) from Waters (Milford, MA, USA) was applied at a column temperature of 40℃. This method enables the parallel scan of high-resolution full scans and scans of low resolution MS² fragmentations of selected precursor ions. This method was modified to only dissociate the two isobaric degradation product ions (m/z 692.4) to demonstrate the separation of them.

Results and discussion

The development of the method started with the selection of a suitable oxidant and a standard base oil. The requirement for the peroxide applied as oxidant is its solubility in nonpolar solvents like lubricating oils (so-called base oil). The peroxide TBHP is well-established in the industry as oxidant for instance as starter for radical polymerizations. Furthermore, it is stable in nonpolar solvents. Because this method was developed emphasizing the production of samples for the investigation of the thermo-oxidative degradation mechanisms by means of ESI-MS, a base oil was required which would not interfere with the degradation of the antioxidant. Additionally, it should show a high oxidative stability and have no contaminations because oxidation products of hydrocarbons as well as many contaminants of them are detectable with MS. They would complicate the interpretation of the mass spectrometric data. These requirements can be fulfilled by utilizing a non-branched, saturated hydrocarbon as n-hexadecane at high purity grade as model base oil.

Preliminary tests indicated that glass ampoules are appropriate ageing container devices. When compared to flasks with ground-in glass stoppers or common headspace vials, they stay airtight even at high temperatures like 160℃ across long ageing times. This reaction temperature was adopted from the air-flow-ageing method because the examination of the temperature influence showed that it was also suitable for the peroxide-ageing method. The oxidation reaction decelerates considerably with decreasing temperatures. The peroxide even ceases to accelerate the oxidation below the self-accelerating decomposition temperature (= boiling point) of TBHP at 89℃. Preliminary tests also indicated that a molar ratio of 1:1 of oxidant to antioxidant cause the complete decomposition of the antioxidant within 24 h.

To evaluate the suitability of the peroxide-ageing method for the monitoring of the thermo-oxidative ageing mechanisms, it was compared with the air-flow method published by Besser et al.¹⁰ As described in the matrix of model oils (Table 3), n-hexadecane was used as base oil for the preparation of the model oils A–C for the ageing with the peroxide. The model oils for the air-flow-ageing method were prepared with a high purity polyalphaolefin PAO-8 representing synthetic base oils (D–F) and the group I base oil SN150 (G–I) to include the investigation of the influence of the base oil on the formation of the degradation products of the antioxidants. The model oils aged with the air-flow method created a high number of samples as a result of the long testing durations. Hence, for the identification and characterization of the degradation products in this model oils with ESI-MS samples had to be preselected based on the results of the visual monitoring and the FTIR measurements.

The additional benefits of the novel peroxide-ageing method in comparison with the air-flow method were tested on the example of the two phenolic antioxidants 2 and 4 which cannot be aged with the air-flow method. Small and volatile molecules are prone to be evaporated by the air-flow in the traditional ageing methods. After a short duration, the antioxidant and also all small oxidation products are lost. This was proven on the example of the phenolic antioxidant 2 by Besser.¹⁶ Due to the use of relative inert glass ampoules as fully closed reactors, this phenol can be aged with the peroxide-ageing method without any problems. The phenol 4 on the other hand, provides a considerably increased stability against oxygen interaction, because it features multiple moieties to react with the oxidant. This compound contains a sulfide moiety, two phenolic and two ester moieties as shown in Figure 1. Therefore, it is not possible to deplete the phenol 4 at a commonly used concentration with an artificial air-flow-ageing method within a reasonable time frame.¹⁶ Due to the high oxidative power provided by the peroxide TBHP utilized as oxidant also this phenol can be aged by means of the peroxide-ageing method.

For the investigation of the influence of the synergistic regeneration of an aminic antioxidant by the phenolic antioxidant on the degradation reactions, five synergistic mixtures in model oils (C, F, I, K and M) were aged and analyzed subsequently. In the interpretation of the mass spectrometric data of these aged samples, the search for condensed oxidation products between aminic and phenolic antioxidants was emphasized, because they would have a major influence on the antioxidant potential of these additives.

The visual condition monitoring, the monitoring of the content of antioxidants (FTIR), the monitoring of the oxidation value (FTIR) and the characterization of the formed degradation products utilizing high resolution ESI-MS for the ageing of individual model oil are summarized in Tables 4 to 6.

Table 4.

Summary of the visual monitoring of the ageing of antioxidants 1–4 in the model oils A–M.

Model oil	Anti- oxidant	Base oil	Ageing		Visual state at the start	Visually observed changes
Model oil	Anti- oxidant	Base oil	Method	Duration	Visual state at the start	Color	Clouding /sludge
A	1	n-hexadecane	peroxide-ageing	24 h	Clear orange solution	Darkened during reaction	Slightly clouding after 24 h
B	3	n-hexadecane		24 h	Clear colorless solution	Light yellow tint after 1.5 h; color intensified to yellow after 15 h	–
C	1+3	n-hexadecane		24 h	Clear orange solution	Darkened to reddish brown during 24 h	–
D	1	PAO-8	air flow-ageing	144 h	Clear orange solution	Darkened during reaction; after 80 h color change from reddish brown to yellow	Stayed clouded after color change; high amount of brownish sludge
E	3	PAO-8		144 h	Clear colorless solution	Darkened to dark yellow; after 96 h again a color change to light yellow	Stayed clouded after color change; traces of brownish sludge
F	1+3	PAO-8		144 h	Clear orange solution	Darkened to reddish brown; after 96 h again a color change to light yellow	Stayed clouded after color change; brownish sludge; less than in D
G	1	SN-150	air flow-ageing	96 h	Clear orange solution	Darkened much faster than corresponding model oils A and D (dark reddish brown after 16 h); after 64 h slightly brightening of the color	Highest amount of dark brown sludge
H	3	SN-150		379 h	Clear colorless solution	Darkened considerably slower than G (dark yellow after 16 h, dark reddish brown after 64 h); no distinct color change (consumption of 3)	Dark brown sludge; less than in G
I	1+3	SN-150		161 h	Clear orange solution	Darkened as fast as G (dark reddish brown after 16 h); similar to H no distinct color change (consumption of 1 and 3)	Dark brown sludge; less than in H (still much more than in D)
J	2	n-hexadecane	peroxide-ageing	24 h	Clear colorless solution	Light yellow tint after 15 h; color intensified to yellow tint after 24 h	–
K	1+2	n-hexadecane		24 h	Clear orange solution	Darkened to a reddish brown during 24 h	Slightly clouding after 24 h
L	4	n-hexadecane		24 h	Clear colorless solution	Light yellow tint after 3 h; color darkened to a light brownish after 15 h	Brownish sludge after 15 h
M	1+4	n-hexadecane		24 h	Clear orange solution	Darkened considerably slower than C, F, I, K (light orange brown after 24 h);	Clouding after 3 h; precipitate reddish brown sludge after 6 h

Table 5.

Summary of the monitoring of the ageing of antioxidants 1–4 in the model oils A–M by standard FTIR approach.

Model oil	Anti- oxidant	Base oil	Ageing		Content of antioxidants (FTIR: –NH: 1515 cm⁻¹; –OH: 3651 cm⁻¹)	Oxidation value based on standard DIN 51453 (FTIR: 1710 cm⁻¹)
Model oil	Anti- oxidant	Base oil	Method	Duration
A	1	n-hexadecane	peroxide-ageing	24 h	Decreased slowly to a minimum of 75% after 15 h; increased again to 80% at 24 h	Linear rise; slightly flattening after 15 h; 0.8 a/cm after 24 h
B	3	n-hexadecane		24 h	Slow decrease; declining degradation rate; 85% after 24 h	Very slow increase to 0.2 a/cm after 24 h
C	1+3	n-hexadecane		24 h	Amine 1: slow, declining decrease to 85% after 24 h Phenol 3: increased depletion in comparison to B to 55% after 24 h	gradient of the increase lies between a and b; 0.7 a/cm after 24 h
D	1	PAO-8	air flow-ageing	144 h	Fast decrease to 50% in 32 h; declining degradation rate; 15% to 10% between 96 and 144 h	Steep increase started after 80 h; 120 a/cm after 144 h
E	3	PAO-8		144 h	Degradation rate increased considerably after 45 h; 3 was not detectable anymore after 125 h	Slow increase started after 96 h; significant acceleration of oxidation after 125 h; 40 a/cm after 144 h
F	1+3	PAO-8		144 h	Amine 1: steep decrease is followed by a considerably declination of the degradation rate; 1 was not detectable anymore after 142 h phenol 3: graph similar to amine 1 instead to phenol 3 in E; 12% between 125 and 144 h	Steep increase with inclination similar to d Started after 80 h; 120 a/cm after 144 h
G	1	SN-150	air flow-ageing	96 h	Graph similar to D but faster decrease (50% in 20 h); declining degradation rate; 12% to 10% between 64 and 96 h	Steep increase started after 49 h; distinct flattening after 69 h: 30 a/cm after 96 h
H	3	SN-150		379 h	Slow almost linear decrease with a slightly flattening after 331 h; 15 A/cm after 379 h	Very slow, almost linear increase to 15 a/cm after 379 h
I	1+3	SN-150		161 h	Amine 1: graph similar to G; 50% after 20 h; 13% to 8% between 69 and 96 h phenol 3: graph similar to H but a considerably faster decrease; 30% after 161 h	Slow, almost linear increase (graph similar to h, but steeper) to 8 a/cm after 161 h
J	2	n-hexadecane	peroxide-ageing	24 h	Steady, almost linear decrease; 2 was not detectable anymore after 24 h	Slow increase at the beginning; becomes increasingly steeper; 4 a/cm after 24 h
K	1+2	n-hexadecane		24 h	Amine 1: graph similar to C; slow, declining decrease to 85% after 24 h phenol 2: graph similar to C but increased depletion to 40% after 24h	Increase to 0.6 a/cm after 6 h and then flattened; 1 a/cm after 24 h
L	4	n-hexadecane		24 h	Decreased to a minimum of 60% after 6 h; increased again to 80% at 24 h	Slow, almost linear increase to 3 a/cm after 24 h
M	1+4	n-hexadecane		24 h	Amine 1: decrease to 55% after 24 h phenol 2: graph similar to L decreased to a minimum of 80% after 6 h; increased again to 85% at 24 h	Graph similar to l; increase to 0.7 a/cm after 6 h and then flattened; 1 a/cm after 24 h

Note: The contents of the antioxidants and the oxidation value, respectively, were determined.

Table 6.

Summary of the mass spectrometric results of the ageing of antioxidants 1–4 in the model oils A–M.

Model oil	Anti- oxidant	Base oil	Ageing		Parent antioxidant	Oxidation product	Oxidation reaction	Precursor	Elemental composition of the precursor	Found m/z	Comment
Model oil	Anti- oxidant	Base oil	Method	Duration	Parent antioxidant	Oxidation product	Oxidation reaction	Precursor	Elemental composition of the precursor	Found m/z	Comment
A	1	n-hexadecane	Peroxide- ageing	24 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1931	Observable during whole ageing experiment
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.1987	Observable during whole experiment; significantly higher abundant than [1]•+
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1645	N-oxide; high abundant immediately after addition of TBHP
B	3	n-hexadecane	Peroxide- ageing	24 h	3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3664
					3			[M−H]⁻	[C₂₉H₄₄O₂−H]⁻	423.3259
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3243	Dehydrogenated phenol; traces detectable immediately after the addition of TBHP; in comparison to 3 abundance increased during ageing; Signal was abundant after 3 h and more abundant than 3 after 15 h
						3a	3-2H	[M−H]⁻	[C₂₉H₄₂O₂−H]⁻	421.3101
						3b	3-2H+O	[M-H]⁻	[C₂₉H₄₂O₃−H]⁻	437.3044	Diaryl ketone; traces detectable between 6 and 24 h
C	1+3	n-hexadecane	Peroxide- ageing	24 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1934	Most abundant positive ions during whole ageing experiment
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.1996	Most abundant positive ions during whole ageing experiment
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1647	Only traces could be detected
					3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3666
					3			[M-H]⁻	[C₂₉H₄₄O₂−H]⁻	423.3259
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3242	In comparison to 3 abundance increased in positive and negative ion mode during ageing; more abundant than 3 after 6 h
						3a	3-2H	[M-H]⁻	[C₂₉H₄₂O₂-H]⁻	421.3101
						3b	3-2H+O	[M−H]⁻	[C₂₉H₄₂O₃-H]⁻	437.3043	Traces detectable between 6 and 24 h
D	1	PAO-8	Air flow- ageing	144 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1933	Observable for the first 80 h; in sample taken after 96 h Not detectable anymore
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.1997
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1649	Low abundant signals Detectable only in sample taken after 96 h
E	3	PAO-8	Air flow- ageing	144 h	3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3669	Detectable for 96 h in positive and negative ion mode
					3			[M-H]⁻	[C₂₉H₄₄O₂-H]⁻	423.3260	Detectable for 96 h in positive and negative ion mode
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3242	Already observable directly after the start of The air-flow-ageing; detectable for 125 h
						3a	3-2H	[M-H]⁻	[C₂₉H₄₂O₂-H]⁻	421.3103
						3b	3-2H+O	[M-H]⁻	[C₂₉H₄₂O₃-H]⁻	437.3044	Similar to B; only traces were detectable, samples taken after 69 and 96 h
F	1+3	PAO-8	Air flow- ageing	144 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1934	Observable only for the first 45 h of the ageing process
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.1996	Observable only for the first 45 h of the ageing process
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1648	Detectable directly after start of the air-flow ageing (on the contrary to D); observable for 96 h
					3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3667	Detectable only for the first 32 h of the ageing process
					3			[M-H]⁻	[C₂₉H₄₄O₂-H]⁻	423.3258	Detectable only for the first 32 h of the ageing process
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3241	Observable for 125 h
						3a	3-2H	[M-H]⁻	[C₂₉H₄₂O₂-H]⁻	421.3100	Observable for 125 h
						3b	3-2H+O	[M-H]⁻	[C₂₉H₄₂O₃-H]⁻	437.3040	Not found
G	1	SN-150	Air flow- ageing	96 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1932	Detectable for only 22 h in positive and negative ion mode
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.1999	Detectable for only 22 h in positive and negative ion mode
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1647	Low abundant signals detectable only in the sample taken after 49 h
H	3	SN-150	Air flow- ageing	379 h	3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3669	Observable for the first 185 h of the ageing process
					3			[M-H]⁻	[C₂₉H₄₄O₂-H]⁻	423.3261	Observable for the first 185 h of the ageing process
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3243	Detectable during whole ageing experiment
						3a	3-2H	[M-H]⁻	[C₂₉H₄₂O₂-H]⁻	421.3104	Detectable during whole ageing experiment
						3b	3-2H+O	[M-H]⁻	[C₂₉H₄₂O₃-H]⁻	437.3043	Abundant signals Observable between 64 and 185 h
I	1+3	SN-150	Air flow- ageing	161 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1933	Traces were still detectable at the Termination of the reaction after 160 h
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.2002
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1648	Not found
					3			[M+NH₄]⁺	[C₂₉H₄₄O₂+NH₄]⁺	442.3670	Observable for the first 89 h of the ageing process
					3			[M-H]⁻	[C₂₉H₄₄O₂-H]⁻	423.3259	Observable for the first 89 h of the ageing process
						3a	3-2H	[M+H]⁺	[C₂₉H₄₂O₂+H]⁺	423.3239	Observable during whole ageing experiment
						3a	3-2H	[M-H]⁻	[C₂₉H₄₂O₂ -H]⁻	421.3104	Observable during whole ageing experiment
						3b	3-2H+O	[M-H]⁻	[C₂₉H₄₂O₃ -H]⁻	437.3043	Abundant signals Observable between 89 and 161 h
J	2	n-hexadecane	Peroxide-ageing	24 h	2			[M-H]⁻	[C₁₅H₂₄O -H]⁻	219.1747	Abundance decreases; traces detectable until 15 h
J	2	n-hexadecane	Peroxide-ageing	24 h		2a	2×2-2H	[M-H]⁻	[C₃₀H₄₆O₂ -H]⁻	437.3411	Dimerization; observable between 1.5 and 15 h of the ageing process
K	1+2	n-hexadecane	Peroxide-ageing	24 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1934	Most abundant for the first 3 h
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.2003	Became most abundant ion after 6 h
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1647	Only traces could be detected
					2			[M-H]⁻	[C₁₅H₂₄O -H]⁻	219.1747	Low abundant; observable during whole ageing experiment
						2a	2×2-2H	[M-H]⁻	[C₃₀H₄₆O₂ -H]⁻	437.3412	Not found
L	4	n-hexadecane	Peroxide-ageing	24 h	4			[M+Na]⁺	[C₃₈H₅₈O₆S+Na]⁺	665.3835	Detectable during whole ageing experiment
					4			[M-H]⁻	[C₃₈H₅₈O₆S -H]⁻	641.3871	Detectable during whole ageing experiment
						4a	4+O	[M+Na]⁺	[C₃₈H₅₈O₇S+Na]⁺	681.3608	Traces were detectable already after 3 h
						4a	4+O	[M-H]⁻	[C₃₈H₅₈O₇S -H]⁻	657.3818	Traces were detectable already after 3 h
						4b	4+2O	[M+Na]⁺	[C₃₈H₅₈O₈S+Na]⁺	697.3565	Traces of the double peak 4b and 4c were detectable already after 3 h in positive and negative ion mode
						4b	4+2O	[M-H]⁻	[C₃₈H₅₈O₈S -H]⁻	673.3771
						4c	4+S	[M+Na]⁺	[C₃₈H₅₈O₆S₂+Na]⁺	697.3570	Intermolecular rearrangement; traces of the double peak 4b and 4c were detectable already after 3 h in positive and negative ion mode
						4c	4+S	[M-H]⁻	[C₃₈H₅₈O₆S₂ -H]⁻	673.3603
						4d	4+H₂O -C₂₁H₃₄O₄S	[M-H]⁻	[C₁₇H₂₆O₃ -H]⁻	277.1803	Traces were detectable already after 3 h
						4e	4+O+H₂O-C₁₇H₂₆O₃	[M+Na]⁺	[C₂₁H₃₄O₅S+Na]⁺	421.2010	Traces were detectable already after 3 h
						4e	4+O+H₂O-C₁₇H₂₆O₃	[M-H]⁻	[C₂₁H₃₄O₅S -H]⁻	397.2005	Traces were detectable already after 3 h
						4f	4+O-C₁₇H₂₆O₃	[M+Na]⁺	[C₂₁H₃₂O₄S+Na]⁺	403.1910	Traces were detectable already after 3 h
						4f	4+O-C₁₇H₂₆O₃	[M-H]⁻	[C₂₁H₃₂O₄S -H]⁻	379.1951	Traces were detectable already after 3 h
M	1+4	n-hexadecane	Peroxide-ageing	24 h	1			[M]^•+	[C₁₄H₂₄N₂]^•+	220.1934	Most abundant; significantly decrease after 6 h
					1			[M+H]⁺	[C₁₄H₂₄N₂+H]⁺	221.2005	Most abundant; significantly decrease after 6 h
						1a	1-4H+O	[M+H]⁺	[C₁₄H₂₀N₂O+H]⁺	233.1647	Considerably increase from traces to most abundant ion after 15 h
					4			[M+Na]⁺	[C₃₈H₅₈O₆S+Na]⁺	665.3836	Abundance increases
					4			[M-H]⁻	[C₃₈H₅₈O₆S -H]⁻	641.3873
						4a	4+O	[M+Na]⁺	[C₃₈H₅₈O₇S+Na]⁺	681.3778	Abundance increases; after 15 h it decreases again
						4a	4+O	[M-H]⁻	[C₃₈H₅₈O₇S -H]⁻	657.3818	Most abundant during whole ageing experiment
						4b	4+2O	[M+Na]⁺	[C₃₈H₅₈O₈S+Na]⁺	697.3740	Traces of the double peak 4b and 4c were detectable immediately after the addition of TBHP
						4b	4+2O	[M-H]⁻	[C₃₈H₅₈O₈S -H]⁻	673.3773
						4c	4+S	[M+Na]⁺	[C₃₈H₅₈O₆S₂+Na]⁺	697.3571	Traces of the double peak 4b and 4c were detectable immediately after the addition of TBHP
						4c	4+S	[M-H]⁻	[C₃₈H₅₈O₆S₂ -H]⁻	673.3602
						4d	4+H₂O -C₂₁H₃₄O₄S	[M-H]⁻	[C₁₇H₂₆O₃ -H]⁻	277.1807	Abundant signal detectable immediately after the Addition of TBHP; slow and steady increase in comparison to 4
						4e	4+O+H₂O-C₁₇H₂₆O₃	[M+Na]⁺	[C₂₁H₃₄O₅S+Na]⁺	421.2009	Detectable immediately after the addition of TBHP
						4e	4+O+H₂O-C₁₇H₂₆O₃	[M-H]⁻	[C₂₁H₃₄O₅S -H]⁻	397.2050	Detectable immediately after the addition of TBHP
						4f	4+O-C₁₇H₂₆O₃	[M+Na]⁺	[C₂₁H₃₂O₄S+Na]⁺	403.1907	Detectable immediately after the addition of TBHP
						4f	4+O-C₁₇H₂₆O₃	[M-H]⁻	[C₂₁H₃₂O₄S -H]⁻	379.1953	Detectable immediately after the addition of TBHP

Investigation of base oils without the addition of antioxidants

For the evaluation of the peroxide-ageing method by comparing it with the air-flow-ageing method, it was necessary to age the three applied base oils also without the addition of antioxidants. All three base oils were clear and colorless at the start. For the model base oil n-hexadecane (non-branched, saturated hydrocarbon), no visual indications for an oxidation by TBHP were found during the entire peroxide-ageing. In contrast, the base oils PAO-8 (group IV synthetic oil) and SN150 (group I mineral oil) aged by the air-flow-method darkened and became cloudy over time. In both latter cases, the formation of sludge was observed after the termination of the reactions. Figure 2(a) shows exemplarily the progressing ageing of the base oil PAO-8. The mineral oil SN150 darkened much faster. Already after 16 h, it was colored dark reddish brown. It darkened further to a very dark brown, almost black color. The deposition of a considerably increased amount of sludge was observed, too.

The oxidation value of all three base oils showed a similar trend. The graph started to flatten after a steep increase at the start of the ageing reactions (data not shown). The main difference between the three base oils is the broad range of the oxidation values reached at the termination of the reaction. The observed oxidation value for PAO-8 (130 A/cm after 142 h) and SN150 (10 A/cm after 142 h) were in the normal range. The oxidation value of the n-hexadecane on the contrary was with 0.7 A/cm after 24 h extremely low. This confirms that n-hexadecane fulfilled the requirement of showing a high oxidative stability, even against the peroxide TBHP. The changes of the determined absorbance between the samples were too low for a reasonable quantitative comparison of the oxidation value of n-hexadecane with other oils.¹⁸ Only the trend of the oxidation graph is significant.

Visual monitoring

The results of the visual condition monitoring are shown in Table 4. The model oils containing the aromatic amine 1 alone (A, D and G) or in a synergistic mixture (C, F, I, K and M) were all clear and orange colored solutions at the beginning of the ageing. Without the amine, the addition of the phenols 1, 2 or 3 leads to clear and colorless model oils (B, E, H, J and L) at the beginning of the ageing processes. All model oils aged by utilizing the peroxide-ageing method (A, B, C, J, K, L and M) showed a slight darkening during the 24 h reaction time – from colorless to slight yellow or brownish for phenols (B, J and L) and from orange to orange brown for amines and synergistic mixtures (A, C, K and M). During the peroxide-ageing of these model oils, the formation of sludge was only observed for L and M, both prepared with the sulfide-containing phenol 4. The other five model oils (A, B, C, J and K) remained clear or showed only a slightly clouding after 24 h. There were no visual signs of a complete consumption of the antioxidants.

The model oil preparations with PAO-8 and aged by the air-flow method for 144 h (D–F) also darkened and additionally clouded over time (Figure 2(b)to (d)). However, all three model oils showed a distinct color change from reddish brown to yellow indicating the complete consumption of the antioxidants. The model oils stayed clouded after this color change by precipitated degradation products of the antioxidants. The following fast darkening indicated an uninhibited degradation of the model oils. The aminic antioxidant 1 (model oil D) formed a high amount, the synergistic mixture in model oil F less and the phenol 3 (model oil E) the least amount of brownish sludge.

The model oils preparations with SN150 G-H also were aged by the air-flow method. Based on the visual and FTIR monitoring, these reactions were terminated at varying times (between 96 and 379 h). These samples darkened much faster and to a darker brown tone than the corresponding model oils D–F. For the model oil G (amine 1), a color change was observable. However, it was very weak. Like their base oil, these model oils formed the highest amount of dark brown sludge observed during the investigations. Again the amine 1 (model oil G) formed a high amount, the synergistic mixture in model oil I less and the phenol 3 (model oil H) the least amount of brownish sludge. Still, the observed amount of sludge for model oil H was much higher than for the corresponding model oil E prepared in PAO-8.

Monitoring of the content of the antioxidants and the oxidation value by FTIR

The results of the monitoring by FTIR are summarized in Table 5. The decrease of the content of the amine 1 and/or the phenol 3 in the model oils A–I during the oxidation was investigated for the evaluation of the peroxide-ageing-method in comparison to the air-flow-ageing-method as well as for the investigation of the influence of the base oil quality on the thermo-oxidative degradation. The results for the synergistic mixtures C, F and I as well as additional mixtures K and M are discussed subsequently.

The content of the antioxidants in the model oils A and B prepared in n-hexadecane and aged by the peroxide-ageing method decreased only slightly (Figure 3(a) and (b)). This observation and the apparent regeneration of the amine 1 in the model oils A at the end of the reaction were taken as evidence for the formation of specific degradation products of the antioxidants still containing functional groups capable of absorbing at the bands used for quantification of amines and phenols, respectively, in FTIR. This behavior has to be considered by the structural elucidation of these specific oxidation products by MS. The oxidation value of A and B started to increase slowly at the beginning of the ageing reaction and stayed in all cases below 1 A/cm (Figure 4(a) and (b)). The immediate increase of the oxidation value, despite the application of an antioxidant, was caused by the strong oxidation power and the high concentration of the peroxide TBHP. The much steeper increase of the oxidation value was observed for the amine 1 (A) in comparison to the phenol 3 (B).

The model oils D and E (prepared in PAO-8) as well as G and H (prepared in SN15) were aged by the air-flow-ageing method. The observed alterations of the FTIR parameters of the correlated model oils of both groups were similar (Figure 3(d) to (h)). They differed considerably from the changes observed for A and B. The relative content of the aromatic amine 1 in the model oil D and G decreased fast at the beginning. Afterwards a slow reduction of the degradation rate was observed. The degradation rate of the phenol 3 in E and F did not decline during the reaction. On the contrary, in model oil E, a considerably increase of the degradation rate after 45 h was observed. The absence of interferences of the degradation products on the FTIR quantification was the result of further depletion leading to the formation of the sludge observed in the visual monitoring. Matching the visual monitoring the oxidation value started slowly to increase after the observed color change with the model oils D, E and G, whereas the model oil H showed no color change as well as no significant increase in the oxidation value (Figure 4(d) to (h)). A direct comparison of the corresponding model oils D–F (prepared in group IV oil PAO-8) and G–I (prepared in group I oil SN150) aged by the air-flow-ageing method showed that the low-quality mineral base oil SN150 decreased the lifetime of the antioxidants considerably. A comparison of the influence on the model oils A–C (prepared in synthetic model base oil n-hexadecane) aged by the peroxide-ageing method was not possible due to the considerably influence of the high oxidative power of the TBHP on the antioxidant lifetime.

The determined amount of the phenolic antioxidant 2 in the model oil J decreased steadily until it was completely consumed after 24 h (Figure 3(j)). The observed gain of the oxidation value became increasingly steeper during the ageing (Figure 4(j)).

The contents of phenol 4 in model oil L decreased for 6 h. Afterwards it increased considerably again (Figure 3(l)). This observation was explained by the hydrolysis of the ester bonds. The formation of additional carboxyl and hydroxyl groups increased the absorption corresponding to the –OH band in FTIR. The oxidation value increased almost steadily with a slight deceleration at the end of the reaction (Figure 4(l)).

Four of the five synergistic model oils C, F, I and K all showed a similar behavior. The very slow decrease of the amine 1 in the synergistic mixtures C and K as well as the significant diminished degradation rate in I and M were an evidence for a real synergistic regeneration since they were accompanied by an increased depletion of the respective phenol (2 and 3) (Figure 4(c) to (k)). The only exception was the model oil M (Figure 4(m)). Here, the degradation of amine 1 was significantly increased, while it was decreased for the phenol 4 in comparison to the correlated model oil A (amine 1.) and K (phenol 4), respectively. The gain of the oxidation value was in all five cases between the ones observed for the corresponding pure aminic and pure phenol model oils (Figure 4(c) to (m)). Again, matching the visual monitoring, the beginning of the increase of the oxidation value of model oil F was delayed until the observed color change appeared.

Characterization of the formed degradation products utilizing high-resolution mass spectrometric measurement provided by ESI-LIT-Orbitrap MS

The elemental compositions of the ions observed by ESI-MS are summarized in Table 6. The MS results generally matched to the observations of the visual and FTIR oil monitoring discussed above. After a complete consumption of the antioxidant in an ageing experiment happened, also in MS experiments neither the antioxidant nor its oxidation products could be detected anymore. It was concluded that these oxidation products had to be consumed in further reactions like degradations and polymerizations leading to the formation of (ultra-high molecular mass) sludge and varnish.

The aromatic amine 1 in the aminic model oils A, D and G as well as in the synergistic model oils C, F, I, K and M formed two distinct ions as detected by positive ion ESI-MS confirming previous results⁴: the radical cation [1]^•+ at m/z 220.1931 and the protonated molecule [1+H]⁺ at m/z 221.1987. The [M]^•+ showed generally a significant higher abundance in comparison to [1+H]⁺. The only oxidation product was observed at m/z 233.1645. It was structurally elucidated as the protonated N-oxide [1a+H]⁺ (Figure 5(a)). The chemical structure and the proposed degradation pathway are shown in Figure 6(a). This reaction involves the formation of two additional double bonds by the loss of four hydrogen atoms and the simultaneous addition of an oxygen atom. Only low abundant signals or traces of [1a+H]⁺ were detectable in the model oils C, D, G, H, K and M. In the model oil I, it was not detectable at all.

As previous results indicate,⁴ phenol 3 was detectable in the model oils B, C, E, F, G and H in positive ion mode as ammonium adduct ion [3+NH₄]⁺ at m/z 442.3666 and in negative ion mode as deprotonated molecule [3-H]⁻ at m/z 423.3259. The formation of two oxidation products was observed (Figure 5(b)). Their chemical structures and the proposed degradation pathway are shown in Figure 6(b) in detail. The first structure was characterized as dehydrogenated phenol 3a and was detectable as protonated ([3a+H]⁺ at m/z 423.3243) and deprotonated molecule ([3a-H]⁻ at m/z 421.3101) in positive and negative ion mode, respectively. It is formed rather fast since it is immediately observable – at least in traces – after the start of the peroxide-ageing as well as the air-flow-ageing. In comparison to the ions of the antioxidant, the abundance of [3a+H]⁺ and [3a-H]⁻ increased considerably during the reaction. The second oxidation product of the phenol 3 was observed at m/z 437.3044 in negative ion mode. Its chemical structure was elucidated as the deprotonated diaryl ketone [3b-H]⁻. Generally, it was found that its formation was delayed in all ageing series. Based on these findings, a slow two-step degradation reaction was proposed as shown in Figure 6(b). In a first and fast step, phenolic antioxidant 3 loses two hydrogen atoms and formed an additional double bond in the case of the dehydrogenated phenol 3a. In a second step, this intermediate is oxidized to the diaryl ketone 3b.

The influence of the quality of the base oil on the thermo-oxidative degradation was also investigated by the ageing of the model oils A–I. It was determined that the oil quality itself has no influence on the degradation pathway since matching degradation products were detected in the corresponding model oils. But on the example of the oxidation of the phenol 3, the influence of the oil quality on the degradation rate can be demonstrated. In the model oils A–E prepared in high-quality base oils (synthetic model base oil n-hexadecane; group IV synthetic oil PAO-8), only traces of the diaryl ketone 3b were detectable, whereas in the model oils G–I prepared in the low-quality group I mineral oil SN150 abundant signals of this oxidation product could be observed. This is caused by the easy oxidation of low-quality mineral oil. Their oxidation products accelerate the formation of the secondary oxidation product of 3 the diaryl ketone 3b. They also accelerate further depletion of the specific oxidation products of the antioxidants by degradation and assumed polymerization. This caused the significantly increased formation of sludge observed for G, H and I.

The phenol 2 in the model oils J and K was only detectable in negative ion mode as deprotonated molecule [2-H]⁻ at m/z 219.1747 as it was already proposed in a previous paper.⁴ It was observable – at least in traces – until 15 h of oxidative stress. Figure 6(c) shows a multistep degradation pathway for the phenol 2 as it is proposed by Vulic et al.¹⁹ A dehydrogenation is followed by a dimerization to 2a and a further double dehydrogenation. The proposed dimer 2a was detectable in the model oils J in negative ion mode as deprotonated molecule [2a-H]⁻ (m/z 437.3411) as shown in Figure 5(c). The complete depletion of the antioxidant 2 in the model oils J as well as of its degradation product 2a which also absorb at the –OH band (IR) after 15 h matched the results of the FTIR monitoring. There was no evidence about the formation of 2a in the model oil K.

In ESI-MS of the model oils L and K, the phenol 4 (for structure see Figure 1) and its degradation products could be determined in negative ion mode as [M-H]⁻ (deprotonated molecule) and in positive ion mode as [M+H]⁺ (protonated molecule), [M+NH₄]⁺ (ammoniated molecule) and [M+Na]⁺ (sodiated molecule). The first two types of cations were only of low abundance because they were suppressed due to a strong contamination of the antioxidant with sodium ions. Therfore, only sodiated molecules are discussed in positive ion mode (for structures see Figure 7).

The ions of the phenol 4 (positive ion mode [4+Na]⁺ at m/z 665.3835; negative ion mode [4-H]⁻ at m/z 641.3871) in the model oils L and M were detectable during the whole peroxide-ageing method. Six degradation products could be charcaterized and identified as shown in Figure 7. They were observable in traces already after 3 h. Three of this degradation products were the results of oxidation reactions directly at the sulfide moiety. The unambigous structure elucidation of these three oxidation products was enabled by the application of high resolution and accurate mass measurements provided by Orbitrap analyzer due to the presence of isobaric ions. The difficulties are illustrated on the example of measurements in positive ion mode as shown in Figure 8. The upper mass spectrum (Figure 8(a)) shows the sodiated adduct ion of the phenol [4+Na]⁺ ([M+23]⁺ at m/z 665.3836) and apparently two peaks at an increased m/z value of +16 Da ([4+39]⁺ at m/z 681.3778) and +32 Da ([4+55]⁺ at m/z 697.3740). A mass increment of 16 Da in the mass spectrum indicates the addition of an oxygen atom, but the combined nominal molecular weight (39) of the sodium (23) and one oxygen (16) atom is equal to the molecular weight of potassium (39). Hence, [4+O+Na]⁺ and [4+K]⁺ are isobaric ion species. Since the antioxidant 4 sample is contaminated with sodium, an additional contamination with potassium would also be possible. Figure 8(b) shows the high-resolution accurate mass measurement of an aged model oil L spiked with potassium to favor the formation of potassium adduct ions. The determined accurate masses of the two isobaric ion species were sufficiently distinct for an unambiguous differentiation between them. Therefore, the peak at m/z 681.3778 was identified as [4+O+Na]⁺ – the sodiated adduct ion of the sulfoxide 4a ([4a+Na]⁺; negative ion mode [4a-H]⁻ at m/z 657.3818). A similar challenge occurred for the mass increment of 32 Da. It could be interpreted as evidence for the addition of two oxygen atoms [4+2O+Na]⁺ as well as one sulfur atom [4+S+Na]⁺. High resolution and accuracy MS confirmed the formation of both oxidation products (double peak). Due to the fact that the m/z values of these two isobaric ions are so similar, only very high resolution (R_FWHM of approx. 100,000) and accurate measurements as shown in Figure 8(c) and (d) could do the task. Overlapping of these peaks would have shifted the determined m/z value and would therefore lead to false identifications. The determination of the accurate mass of the completely separated peaks enabled the unambiguous identification of [4+2O+Na]⁺ as the sulfone 4b (positive ion mode [4b+Na]⁺at m/z 697.3740; negative ion mode [4b-H]⁻ at m/z 673.3773) and of [4+S+Na]⁺ as the disulfide 4c (positive ion mode [4c+Na]⁺ at m/z 697.3571; negative ion mode [4c-H]⁻ at m/z 673.3602). It is proposed that this disulfide derivative was formed by dimerization of two sulfide radicals. This unstable intermediate on the other hand was formed by a cleavage of the sulfide moiety as a result of the oxidative attack of the peroxide TBHP at the sulfide moiety (Figure 7).

The formation of the disulfide 4c was initially considered quite unusual as a result of a homolytic cleavage followed by a dimerization of the formation of this degradation product. The confidence in the structure elucidation by high resolution MS would be increased considerably by the investigation of the fragmentation behavior. Due to the fact that the mass difference of the two isobaric ions was too low, they could not be fragmented separately by CID-MS². Hence, the formation of 4b as well as of 4c was finally confirmed by two additional MS analyses. One proof for the oxidative formation of the disulfide 4c was performed by reacting 4c with the diethyl disulfide in the presence of an alkaline aluminium oxide catalyst, as only 4c yields a disulfide moiety exchange reaction (“mixed disulfide”), but not 4. This mixed disulfide was determined in positive ion ([C₂₁H₃₄O₃S₂+NH₄]⁺ at m/z 416.2279; [C₂₁H₃₄O₃S₂+Na]⁺ at m/z 421.1840) and negative ion ([C₂₁H₃₄O₃S₂-H]⁻ at m/z 397.1883) ESI-MS mode (exemplary data are shown in the Supplementary Figure S1). The second proof was done by an on-line UHPLC-MS², i.e. LE-CID, approach. As Figure 9 shows, UHPLC-MS enabled the separation of the isobaric sulfone 4b and the disulfide 4c, in samples of the model oil L, thus enabling the separate fragmentation (LE-CID (MS²)) of the two oxidation products. Additionally, UHPLC-MS showed that the formation of the sulfone 4b is faster, since it was already observed after 3 h, whereas the disulfide 4c was detectable only after 15 h reaction time.

The other three degradation products of the phenolic antioxidant 4 were formed by cleavage reactions. Hydrolysis of one of the ester moieties of the phenol 4, as well as of the other five degradation products would lead to the formation of 4d – a phenol with a carboxyl moiety – which can only be detected in negative ion mode ([4d-H]⁻ at m/z 277.1807) due to easy deprotonation in ESI. Only the hydrolysis of the sulfoxide [4a+Na]⁺ leads to the formation of a corresponding degradation product 4e, which could be determined by ESI-MS. 4e was identified as a sulfoxide containing alcohol (positive ion mode [4e+Na]⁺ at m/z 421.2009; negative ion mode [4e-H]⁻ at m/z 397.2050). The final degradation product 4f, a sulfoxide containing alkene (positive ion mode [4f+Na]⁺ at m/z 403.1907; negative ion mode [4f-H]⁻ at m/z 379.1953) was formed by 4e due to the loss of an additional H₂O molecule.

The influence of the synergism on the thermo-oxidative degradation of antioxidants was investigated on five model oils (C, F, I, K and M) each containing a combination of the aminic antioxidant 1 and one of the three phenolic antioxidants 2–4. For none of those aged model oils, the formation of condensed oxidation products of the aminic and the phenolic antioxidants was detectable by means of the “soft ionization” technique ESI. The reason might be that under the ESI conditions applied, no desorption and ionization of high molecular mass condensation products were achievable and ESI solvent solubility of such condensation products might be challenging. MALDI (matrix-assisted laser desorption ionization) with appropriate MALDI matrix, LDI (laser desorption ionization) at different laser wavelengths, LAESI (Laser ablation ESI) or DESI (desorption ESI) might be useful alternative desorption/ionization techniques related to such structures and maybe helpful of applying the correct solvents.

All oxidation products of the three phenolic antioxidants 2–4 detectable by ESI-MS featured at least one phenolic group. These results demonstrate clearly that the specific oxidation products of antioxidants can still contain the functional groups whose IR absorption can be used for quantification of these antioxidants. Hence, the specific oxidation products of antioxidants can interfere with the quantitation of these antioxidants by FTIR.

Conclusions

A novel accelerated artificial ageing method (minimal duration 24 h) in combination with ESI MS has been developed and was compared to a standard air-flow-ageing method (minimal duration 96 h). The application of the peroxide TBHP as oxidant and n-hexadecane as model base oil provided a fast and reliable ageing process of model oils. Therefore, the novel peroxide-ageing method was considered sufficient (no effect on sampling on the rest of the aging samples; the sampling volume can be minimized; the method enables artificial ageing experiments of even volatile antioxidants; degradation of very stable antioxidants is feasible due to the increased oxidative power of the applied peroxide) for the reproducible monitoring of thermo-oxidative degradation processes. The monitoring was carried out by visual inspection, by FTIR and mainly by ESI-LIT-Orbitrap MS. On the example of one aminic and three phenolic antioxidants, it was shown that the high resolution accurate mass measurements provided by the Orbitrap MS analyzer enabled the unambiguous elucidation of the chemical structures of the specific oxidation products of the antioxidants dissolved in the bulk lubricant. The accuracy and resolving power of the applied Orbitrap analyzer were even high enough to facilitate a distinctive differentiation between isobaric ions (ions with the same nominal mass values, but different elemental compositions). For the aminic antioxidants 1, the formation of one specific oxidation product in the model oil was observed. For the phenolic antioxidants 2, 3 and 4, the formation of one, two and six degradation products, respectively, was determined unambiguously. It was also verified that the formation of certain degradation products – still containing functional groups which absorb on the IR bands used for the determination of the content of antioxidants – interferes with the quantitation by FTIR. In the case of the phenolic antioxidant 4, even an increase of the determined amount of phenol was observed since the hydrolysis of esters moieties increased the content of hydroxyl groups.

For evaluating the novel peroxide-ageing method, model oils containing the amine 1 and/ or the phenol 3 were prepared in the base oil PAO-8 (group IV oil) as well as in SN150 (group I oil) and were also aged by the mentioned conventional air-flow based ageing method. Although three different base oils and two different ageing methods were used, the same specific degradation products were determined by ESI-MS for the antioxidants in each of the corresponding samples. The influence of the quality of the base oil on the degradation was proven in the comparison of three different base oils. The use of high quality synthetic oils, consisting entirely of saturated hydrocarbons, like the model base oil n-hexadecane or the group IV oil PAO-8 suppressed the formation of an excess of hydrocarbon oxidation products (oxidation products of the oil). These degradation products would have accelerated the further depletion of the specific oxidation products of the antioxidants by degradation and highly possible polymerization. This would lead to a significant increase of the formation of sludge and varnish which cannot by analyzed directly by the means of ESI-MS. On the other hand, all model oils prepared with the low-quality group I oil SN150 darkened faster and showed a considerably increased formation of sludge and varnish. This turned out to be the case especially for the formation of the secondary oxidation product of the phenol 3. In the high-quality oils, only traces of this degradation product were detectable by MS, whereas it provided an abundant signal in aged samples of model oils of the phenol 3 prepared in the low quality SN150 base oil. The courses of the graphs of the content of antioxidants and the oxidation value determined by FTIR were similar for the corresponding air-flow-ageing reactions. The main difference was that the group I oil SN150 decreased the lifetime of each antioxidant. The differences of the results of the FTIR monitoring between them and the corresponding model oils aged by the peroxide-ageing method were caused by the significant increased oxidative power of the peroxide in comparison to air.

This work also demonstrated on two examples that the novel peroxide-ageing method also can be used for the investigation of the degradation behavior of antioxidants which cannot be aged by the air-flow method within a reasonable reaction time. On the example of the volatile phenolic antioxidant 2, it was shown that the use of glass ampoules as enclosed reactors prevented the loss of volatile molecules like small molecule additives or their low-molecular mass degradation products, thus enabling the investigation of the thermo-oxidative degradation mechanisms involved. The phenolic antioxidant 2 exhibited a high antioxidant potential. Therefore, it was not possible to deplete this antioxidant with an artificial air-flow-ageing method within a reasonable time frame. The oxidative power of the peroxide TBHP on the other hand was proven to be high enough for thermo-oxidative degradation of this antioxidant in a reasonable amount of time.

Finally, the influence of the synergism between aminic and phenolic antioxidants on the formation of specific oxidation products was investigated. Model oils containing mixtures of the amine 1 and one of the three phenolic antioxidants (2, 3 and 4) were prepared and aged. In all cases, signs of the synergism like regeneration of the amine in combination with a prolongation of its lifetime or an increased depletion of the respective phenol were observed. However, there was no indication for the formation of condensed molecules between both types of antioxidants in the bulk lubricant.

In summary, the novel peroxide-ageing method in combination with high resolution and accuracy ESI-MS has been demonstrated to be a versatile tool for the investigation of thermo-oxidative degradation mechanisms of antioxidants in a significantly shorter reaction time compared to a standard air-flow-ageing method with a very high reliability. Further developments like the addition of the possibility of sampling sludge and varnish as well as the possibility to analyze them by different advanced desorption/ionization MS techniques as LDI, MALDI, DESI or LAESI would even enhance the tribochemical value of this method.

Supplemental Material

Supplemental material for Development of an accelerated artificial ageing method for the characterization of degradation products of antioxidants in lubricants by mass spectrometry

Supplemental material for Development of an accelerated artificial ageing method for the characterization of degradation products of antioxidants in lubricants by mass spectrometry by Alexander Kassler, Ernst Pittenauer, Nicole Doerr and Guenter Allmaier in European Journal of Mass Spectrometry

Footnotes

Dedication

This paper is dedicated to Professor Jean-Claude Tabet on his 75th birthday.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was funded by the “Austrian COMET-Program” in the frame of K2 XTribology, cofinanced with “EFRE”-funding within the project “OnLab” and has been carried out within the “Excellence Centre of Tribology” – AC²T Research GmbH.

References

Dong

Migdal

Antioxidants. In: Rudnick

(ed). Lubricant additives: Chemistry and application, 2nd ed. Boca Raton, FL, USA: CRC Press, 2009, pp. 3–50.

Gatto

Moehle

Cobb

et al.

Oxidation fundamentals and its application to turbine oil testing. J ASTM Int 2006; 3: 1–20.

Bartels

Bock

Braun

et al.

Lubricants and lubrication. Ullmann's encyclopedia of industrial chemistry, Online ed. Hoboken, NJ, USA: Wiley-VCH, 2003.

Kassler

Pittenauer

Doerr

et al.

Electrospray ionization and atmospheric pressure matrix-assisted laser desorption/ionization mass spectrometry of antioxidants applied in lubricants. Rapid Commun Mass Spectrom 2009; 23: 3917–3927.

Kassler

Pittenauer

Doerr

et al.

CID of singly charged antioxidants applied in lubricants by means of a 3D ion trap and a linear ion trap-Orbitrap mass spectrometer. J Mass Spectrom 2011; 46: 517–528.

de Hoffmann

Stroobant

. Mass spectrometry: Principles and applications, 3rd ed. Hoboken, NJ, USA: John Wiley & Sons, 2007.

Brenton

Godfrey

. Accurate mass measurement: terminology and treatment of data. J Am Soc Mass Spectrom 2010; 22: 1821–1835.

Gross

. Accurate masses for structure confirmation. J Am Soc Mass Spectrom 1994; 5: 57.

Makarov

Denisov

Kholomeev

et al.

Performance evaluation of a hybrid linear ion trap/Orbitrap mass spectrometer. Anal Chem 2006; 78: 2113–2120.

10.

Besser

Schneidhofer

Dörr

et al.

Investigation of long-term engine oil performance using lab-based artificial ageing illustrated by the impact of ethanolas fuel component. Tribol Int 2012; 46: 174–182.

11.

CEC L-48-A00: Oxidation stability of lubricating oils used in automotive transmissions by artificial ageing. CEC Working Group, Brussels, Belgium, 2007.

12.

DIN 51352: Testing of lubricants; determination of ageing characteristics of lubricating oils; increase of the Conradson carbon residue after ageing by passing air through the lubricating oil. Deutsches Institut für Normung, Berlin, Germany, 1985.

13.

DIN 51554: Testing of mineraloils; test of susceptibility to ageing according to Baader; part 1: purpose, sampling, ageing; part 2: testing at 110℃; part 3: testing at 95℃. Deutsches Institut für Normung, Berlin, Deutschland, 1978.

14.

ASTM D.4742-08e1: Standard test method for oxidation stability of gasoline automotive engine oils by thin-film oxygen uptake (TFOUT). ASTM International, West Conshohocken, PA, USA, 2008.

15.

Kassler

Pittenauer

Doerr

et al.

Reversed phase UHPLC-ESI-LIT-Orbitrap mass spectrometry of antioxidants (amines and phenols) applied in lubricants. Rapid Commun Mass Spectrom 2014; 28: 63–76.

16.

Besser C. Untersuchung der thermooxidativen Stabilität von Industrieschmierölen. Master Thesis, University of Vienna, Vienna, Austria, 2007.

17.

DIN 51453: Testing of lubricants; determination of oxidation and nitration of used motoroils; infrared spectrometric method. Deutsches Institut für Normung, Berlin, Germany, 2004.

18.

Oelcheck GmbH. Why does infrared spectroscopy sometimes come up short?www.oelcheck.de/en/knowledge-from-a-z/test-methods/oxidation-of-gas-engine-oils-oelchecker-spring-2013.html (accessed 27 January2014).

19.

Vulic

Vitarelli

Zenner

. Structure-property relationships: phenolic antioxidants with high efficiency and low colour contribution. Polym Degrad Stab 2002; 78: 27–34.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB