Nanotechnology In Action: Overbased Nano Detergents As Lubricant Oil Additives

+ MODEL

ARTICLE IN PRESS

Advances in Colloid and Interface Science xx (2006) xxx – xxx www.elsevier.com/locate/cis

Nanotechnology in action: Overbased nanodetergents as lubricant oil additives L.K. Hudson a , J. Eastoe a,⁎, P.J. Dowding b a

b

School of Chemistry, University of Bristol, Bristol, BS8 1TS, United Kingdom Infineum UK Ltd, Milton Hill Business and Technology Centre, Abingdon, Oxon, OX13 6BB, United Kingdom

Abstract The synthesis and study of oil-soluble metal carbonate colloids are of interest in the area of lubricant additives. These surfactant-stabilised nanoparticles are important components in marine and automotive engine oils. Recently introduced, environmentally driven legislation has focused on lowering of gaseous emissions by placing limits on the levels of phosphorous sulphur and ash allowed in engine oil systems. These chemical limits, coupled with improved engine performance and extended oil drainage intervals, have lead to renewed interest in the production of stable, efficient nanodetergent systems. To date, this has resulted in modification of existing surfactant structures and development of new generations of surfactants. This review covers the current state of research in the area of nanodetergents. © 2006 Elsevier B.V. All rights reserved. Keywords: Lubricant additives; Lubricants; Overbased; Detergents; Nanoparticles; Non-aqueous colloids

Contents 1. 2. 3. 4. 5. 6.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nanoparticle preparation from microemulsions . . . . . . . . . . . . . . . The detergent core . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solvent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. SANS and SAXS . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Langmuir trough measurement and molecular dynamics simulations 6.3. Acid neutralisation mechanism . . . . . . . . . . . . . . . . . . . 7. Conclusions and future outlook. . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction This review describes recent advances in a commercial application of nanoparticles, which have been routinely used as ⁎ Corresponding author. Tel.: +44 117 9289180; fax: +44 117 9250612. E-mail address: [email protected] (J. Eastoe).

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

0 0 0 0 0 0 0 0 0 0 0 0

additives in lubricant oils since the 1940's. A major challenge currently facing potential industrial applications of new nanotechnologies is the ability to generate well-defined, functionalised particles on a mega-tonne scale. Lubricant detergents represent an important case study for commercialisation of nanotechnology. Ever since the invention of the internal combustion engine lubricant oils have been an essential component. Over the last

0001-8686/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.cis.2006.05.003 CIS-00807; No of Pages 7

ARTICLE IN PRESS 2

L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

Fig. 1. Schematic picture of an overbased sulphonate detergent particle. Reproduced by permission of Elsevier.

century, an increasing range of additives have been incorporated into lubricating oils to confer chemical stability, improved performance and beneficial physiochemical properties [1]. An important class are the so-called overbased additives (detergents), which are colloidal nanoparticles of calcium carbonate (and calcium hydroxide) stabilised by a surfactant layer. These nanodetergents essentially consist of an inorganic core (15–40 mass %) stabilised by oil-soluble surfactants (20–45 mass %) incorporated into a lubricating base oil, as depicted schematically in Fig. 1 [2]. Fuel combustion yields many by-products, including inorganic and organic acids formed by oxidation of sulphurous and nitro-

genous impurities in the fuel, and oxidative degradation of the lubricant to form organo-acids. If allowed to build up, these acids would cause severe corrosion, especially under engine start-up conditions. These nanoparticles are relatively insensitive to temperature, explaining their effectiveness at high temperatures as slow-release acid neutralisers [3]. The principal functions performed by detergents in engine oil formulations are: (1) acid neutralisation, (2) high temperature detergency, (3) oxidation inhibition, and (4) rust prevention. These functions promote engine cleanliness, fuel efficiency and extended trouble-free operation. Various surfactants are used to stabilise the CaCO3 particles, typically long chain carboxylic acids, glycols, alcohols, phenates, sulphonates, salicylates or phosphonates. Recently calixarene/ stearate surfactants have been developed [4]; these novel systems address environmental concerns about existing detergents (by containing no sulphur or phosphorous). The term “overbased” refers to the fact that the quantity of base incorporated in the particle cores is greater than that needed to neutralise the acid surfactant. The neutralising strength of an overbased detergent is measured by its ‘total base number’ (TBN). Where TBN is defined as the quantity of acid, expressed in terms of the equivalent number of milligrams of potassium hydroxide that is required to neutralise all basic constituents present in 1 g of overbased detergent. 2. Nanoparticle preparation from microemulsions The commercial synthesis of colloidally stable metal carbonates is known as the “oxide/hydroxide” process; the mechanism of formation is thought to proceed through a microemulsion route [5]. The reaction system consists of an excess of metal hydroxide, appropriate surfactant and a mixture of hydrocarbon and polar solvents. The polar solvents are usually methanol and water: water, surfactant and methanol form swollen reverse micelles,

Fig. 2. XANES spectra obtained on overbased sulphonate detergent particles (in grey) and on pure calcite powder (in black). Reproduced by permission of Elsevier.

ARTICLE IN PRESS L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

3

Fig. 3. Positive ion TOF-SIMS spectrum of overbased sulphonate detergent particles showing the presence of calcium and calcium hydroxide ions. Reproduced by permission of Elsevier.

which contain dispersed Ca(OH)2. The surfactants are made ‘in situ’ by the reaction of the corresponding organic acid with the metal base. Diffusion of gaseous CO2, which is bubbled through the system, generates CaCO3 in the reverse micellar cores, which then nucleate at supersaturation [5]. The micellar cores grow through coalescence. Literature suggests that a stable system requires a certain amount of residual calcium hydroxide, which ensures that the inorganic core is amorphous, [5]. If carbonation is driven to completion, the cores transform into crystalline calcite, which agglomerates to form larger particles and these eventually sediment during the reaction [6]. The products are then filtered under high pressure and temperature to remove unreacted inorganic material and agglomerated particles. The initial waterin-oil microemulsion reaction system results in the formation of nanometre-sized particles, with a narrow particle size distribution. Although the mechanisms involved in these reactions are not yet fully understood, [5,7,8], Jacquet and co-workers [6] have studied the oxide/hydroxide reaction and have drawn some general conclusions on the behaviour of promoters and determination of factors controlling the overbasing reaction. 3. The detergent core The detergent core is typically 1–10 nm in diameter as demonstrated by small-angle neutron scattering (SANS) experiments. Many metals have been incorporated into detergents but currently, based on cost/performance, the three most commonly used are calcium, magnesium and sodium. The detergent CaCO3 core naturally occurs in many forms, in the bulk as calcite and

aragonite, and also as a variety of unusual forms produced principally by living organisms, [9]. Calcite has a rhombohedral crystal structure, whereas aragonite an orthorhombic crystal structure. Heyes and co-workers [10] have modelled CaCO3 nanoparticles, and their simulations suggest an essentially amorphous microstructure. They suggest that there is no thermodynamic benefit from forming local crystalline order in the cores due to the large surface to volume ratio. Using complimentary techniques such as, X-ray photoelectron spectroscopy (XPS), X-ray adsorption near edge structure (XANES) and time-of-flight secondary ion mass spectroscopy (ToF-SIMS) with examples shown in Figs. 2 and 3, the authors [11] have been able to prove the existence of Ca(OH)2 in the reversed micelle core and suggest that it is preferentially located as an outer shell in the micelle core. This can be explained by residual calcium hydroxide that is not fully carbonated, but still is present in the core thus creating an amorphous structure. Further, simulations [2,10] have shown that once nanoparticles of CaCO3 are formed, the stabilising surfactant molecules are essentially “locked” in place on the surface, perhaps by strong coulombic forces originating from the inorganic core material. In 1996 new inorganic colloidal dispersions of calcium, sodium and potassium thiophosphate were produced and subsequently characterised [12,13]. These systems have beneficial anti-wear properties. 4. Surfactants As mentioned above, a range of surfactants can be used to stabilise CaCO3 nanoparticles. The chain lengths of these Table 1 Comparison of detergent properties

Fig. 4. Molecular structures of surfactants commonly employed in overbased detergent systems a) alkyl suphonate, b) sulphurised alkyl phenol (SAP) and c) alkyl salicylate where R1 is a long chain alkyl tail group.

Property

Phenates

Sulphonates

Salicylates

Phosphonates

Total base number (TBN) range Hydrolytic stability Oxidation stability Thermal stability Detergency Rust inhibition Antioxidant effect

0–300

0–500

0–300

0–80

Good Very good Excellent Good Low Very good

Moderate Poor Excellent Good Good None

Good Very good Excellent Excellent Low Very good

Moderate Moderate Moderate Good Good Good

ARTICLE IN PRESS 4

L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

properties [14]. There are four major classes: sulphonates, phenates, salicylates and phosphonates shown schematically in Fig. 4 [10]. Table 1 summarises the detergent properties of the different surfactants commonly employed in stabilising CaCO3 nanodetergents, [15]. Sulphonates are the most widely used detergent additives followed by phenates, salicylates and phosphonates. Environmental concerns, primarily the problems of sulphur emissions and sulphur/phosphorous poisoning of automobile catalytic converters, have led to the demand for sulphur-free overbased detergents such as salicylates and alkyl bridged phenates (of which calixarenes are an example, generic structures of these are shown in Fig. 5) and combination of such structures (e.g. salixarenes) [16]. 5. Solvent

Fig. 5. Schematics of generic calixarene structures.

surfactants are generally between C9 and C60. On neutralising, the surfactant prevents both the formation of “varnish” on pistons, bears polishing of the cylinder and also confers antifriction

The solvent system employed in preparing these overbased detergents is very important. Small changes in the composition of the solvent mixtures result in very different final products. Jacquet and co-workers [17] have methodically explored the effects of varying the solvent level in the preparation of overbased calcium sulphonate. It was shown that: (1) a minimum amount of water is required, (2) a minimum amount of methanol is required to give a significant TBN; however, an excess of water has a detrimental effect on the rheology, and may even result in an organogel, and (3) increasing the amount of xylene increases the rate of

Fig. 6. Measured small-angle neutron scattering curves and analysed core shell form factors for surfactant-stabilised nanodetergent particles in mixtures of d8-toluene and h8-toluene. a) and b) are for 6.2 nm particles with an 1.8 nm shell, a) in 100% h8-toluene, b) in 23% d8-toluene, 77% h8-toluene, c) and d) are for 10.5 nm particles with a 2.6 nm shell, c) in 60% d8-toluene, d) in d8-toluene. Reproduced by permission of Springer.

ARTICLE IN PRESS L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

5

6.2. Langmuir trough measurement and molecular dynamics simulations

Fig. 7. Space filling (calcium ions) and stick (all other molecules) representations of overbased sulphonate detergent nanoparticles using repulsive alkyl tails. Reproduced by permission of the Royal Society of Chemistry.

Various molecular simulations have been performed on detergent systems. Heyes and co-workers [4,23] have shown, by combined Langmuir trough measurements and molecular dynamics simulations, that different geometries in the detergent particle are observed when the surfactant chemistry is varied. It was shown that sulphonate and salicylate stabilised particles were approximately spherical in shape (Fig. 7 shows an example of modelled overbased sulphonate detergent); however, phenates and calixarates generated oblates and prolates, respectively (Fig. 8 depicts a modelled overbased phenate detergent). The addition of a stearate cosurfactant gave more spherical particles by breaking the natural regular packing tendency of the surfactants. The simulations have shown that the internal structure and arrangement of the ions in the core are sensitive to the surfactant type, and produce quite specific core shape and internal ionic structures. Strong coulombic forces between the ions provide the driving mechanism for the Ca2+, CO32−, and surfactant moieties to arrange into a reverse micellar structure, with the

carbonation without changing the final TBN. It is known that the methanol also acts as a promoter at 70 °C for the preparation of CaCO3 nanoparticles [18]. 6. Characterisation Detergents have been characterised by many different methods over the years including acid neutralisation, SANS and Transmission electron microscopy [11,19–22,25,26]. 6.1. SANS and SAXS Both small-angle neutron scattering (SANS) and small-angle X-ray scattering (SAXS) are ideal methods to study nanodetergents [19–22]. Markovic and Ottewill [19,20,21] prepared calcium carbonate particles, stabilised by alkylbenzene sulphonate surfactants in toluene and dodecane. They used SANS to characterise the system and the Guinier approximation and full model fitting to analyse the data. Using these approaches, they could independently determine the metal carbonate core diameter and the thickness of the surfactant layer. The data and analysis in Fig. 6 provide a firm evidence for the core shell structure proposed by others. More detail can be found in relevant papers [20,21]. Further work by Ottewill and co-workers [22] investigated magnesium carbonate systems. During the synthesis of these systems, conversion of magnesium oxide stopped naturally when the basic salt 3MgCO3·Mg(OH)2·3H2O was formed. The stabilising surfactant used was alkylbenzene sulphonate, and the solvents were iso-octane and toluene. When using dilute systems, it was found that a single shell model could be used for iso-octane, but a double shell model had to be employed to accurately account for the scattering data with toluene solvent systems. This was attributed to a partial solvation of the steric stabilising layer by the toluene. These detailed analyses were broadly consistent with a core dimension of 1–10 nm and a steric stabilising layer of 1–5 nm.

Fig. 8. a) Front and b) side projections for space filling (calcium ions) and stick (all other molecules) representations of overbased phenate detergent nanoparticles using repulsive alkyl tails. Reproduced by permission of the Royal Society of Chemistry.

ARTICLE IN PRESS 6

L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

calcium carbonate in the core and surfactant anions forming an external stabilising shell [3]. The level of penetration of the core is greatest for phenate systems and least for sulphonates [23]. Simulations performed by Tobias and Klein [24] modelled sulphonate stabilised nanodetergents, suggesting that the structure was roughly spherical. They also hypothesized that due to large dipole moments, it was conceivable that long-range electrostatic effects could drive strong attractions between free polar acid molecules dispersed on the bulk non-polar solvent and the polar micelle core. This suggests a possible mechanism for the enhanced stabilisation in the oil medium. 6.3. Acid neutralisation mechanism The main purpose of nanodetergent oil additives is to neutralise acids generated by combustion of petrochemical fuels. Various studies have been made to attempt to elucidate the mechanism of neutralisation [7,25]. Robinson et al. [26] have studied the reaction of organic and inorganic acids with overbased detergent additives, in organic media, using Fourier transform infra-red spectroscopy (FTIR) and stopped flow UV/Vis spectroscopy; an example of a typical profile generated by a stopped flow experiment is shown in Fig. 9. The systems they studied were amorphous calcium carbonate and magnesium carbonate cores stabilised alkylbenzene sulphonate surfactant with TBN of 300. For hydrochloric acid, it was noted the acid was neutralised on a timescale of milliseconds to seconds, with the reaction being both temperature and solvent dependent. Results showed that for a ‘matched set’ of calcium and magnesium overbased sulphonate detergents magnesium detergents neutralised inorganic acid more rapidly than their calcium analogues. The molecular weight and structure of the sulphonate surfactant appeared to have a less marked influence on neutralisation rate. However, detergents containing higher levels of surfactant were seen to neutralise acid

more rapidly. The greater neutralising ability of magnesium detergents was attributed to the fact that neutralisation of inorganic acid could occur within the aqueous phase, and that magnesium carbonate had greater aqueous solubility than calcium carbonate. For organic acids (propanoic, cyclohexanoic, heptanoic, nonanoic and decanoic acids), it was found that neutralisation occurred over a longer timescale (up to minutes) than observed for inorganic acid neutralisation, and never appeared to reach 100% completion. This was explained by the possibility that neutralised metal carboxylate salt could remain attached to the particle surface rather than penetrate the inner core. Over time, formation of such a layer on the periphery of the nanodetergent core would be expected to hinder any further reaction between the basic core and acid molecules. The relative rates of neutralisation between magnesium and calcium carbonate nanodetergents were reversed for organic acids as compared to inorganic acids, and this is thought to be linked to the relative basicity of the inorganic species, i.e. CaCO3 > MgCO3 (and Ca(OH)2 > Mg(OH)2). No significant effects of structure or molecular weight of the sulphonate surfactant were reported with regard to rate of acid neutralisation: again, higher surfactant levels gave rise to faster rates of acid neutralisation. Studies of neutralisation kinetics as a function of acid molecular weight showed that the rate of neutralisation decreased with increased acid molecular weight. This was explained by the bulkier acids being more effective at screening the inorganic core, by forming a thicker surface layer, and hence hindering neutralisation. In addition, the relative solubilities of the carboxylic acids in the solvent (heptane) will also affect neutralisation rate: those with relatively low solubilities (such as propanoic acid) effectively ‘driving’ the acid towards neutralisation. Organic acid neutralisation has also been investigated by Papke [27] using FTIR and overbased calcium sulphonate nanodetergents. In addition, Robinson and co-workers [26] measured – COOH absorption at 1705 cm− 1 to follow the percentage of acid

Fig. 9. Typical stopped flow trace for the reaction between overbased detergent particles and acid containing water-in-oil microemulsion droplets using methyl range as the pH indicator. The time (t1/2) taken for the absorbance to decrease by half of the total change (ΔA / 2) is used as a measure of the neutralisation rate. Reproduced by permission of Elsevier.

ARTICLE IN PRESS L.K. Hudson et al. / Advances in Colloid and Interface Science xx (2006) xxx–xxx

neutralised as a function of time. Papke used the sulphonate band at 1065 cm− 1 as a reference, due to the fact that no sulphonate was lost in the precipitate (carboxylate salt). This enabled a direct determination of the relative amount of solubilized carbonate (865 cm− 1) by comparison to the sulphonate level. The measurements were performed with hexanoic acid; a plateau region was observed initially during neutralisation, where the amount of carbonate did not change, followed by a linear region corresponding to reaction of carbonate with acid. The author concluded that a non-carbonate base was selectively neutralised, which is residual hydroxide in the core of the nanodetergent. Furthermore, it must also be accessible and, therefore, is probably located near the surface of the inorganic core, consistent with the simulations of Heyes and co-workers [2,10] suggesting an outer hydroxide shell. 7. Conclusions and future outlook Overbased detergents have been an integral element of lubricant additive packages for over 50 years. Indeed detergents are examples of nanoparticles, which can be reproducibly produced on an industrial scale. Production of uniform nanoparticles with tailored properties on a large scale is a focus of current research interest. The surfactants used to stabilise detergents are generally aromatic in nature, often based on structures commercially available in petrochemical plants. The need for optimised nanoparticle detergents is increasing, driven by: • Improvements in engine design for better fuel economy and longer oil drain intervals • Environmental legislation aimed at providing reduced gaseous emissions (by imposing chemical limits on the fuel and lubricant) • With the use of more refined base oils (which are more aliphatic in nature), the necessity for development of stable detergents with optimised performance is increasing. This has manifested in the development of novel surfactant structures. However, a significant amount of research is still required to characterise and optimise these new systems before they become commercially viable. In addition, other issues should be addressed with regard to environmental acceptability, in particular the synthesis routes. Current synthesis methods rely on using mega-tonne quantities of solvents (xylene, methanol, etc.), here work could seek to

7

reduce the levels or replace petrochemical solvents with green alternatives. Acknowledgments LH thanks Impact Faraday (EPSRC) and Infineum UK Ltd for the provision of a Ph.D. Scholarship. References [1] Ullmann's encyclopaedia of industrial chemistry, vol 20, sixth edition, chapter 6, p 118 Jürgen Braun and Jürgen Omeis. [2] Bearchell CA, Heyes DM, Moreton DM, Taylor SE. Phys Chem Chem Phys 2001;3:4774. [3] Griffiths JA, Heyes DM. Langmuir 1996;12:2418. [4] Cunningham ID, Courtois J-P, Danks TN, Heyes DM, Moreton DJ, Taylor SE. Colloids Surf A Physicochem Eng Asp 2003;229:137. [5] Bandyopadhyaya R, Kumar R, Gandhi KS. Langmuir 2001;17:1015. [6] Roman JP, Hoornaert P, Faure D, Biver C, Jacquet F, Martin J-M. J Colloid Interface Sci 1991;144:324. [7] Galsworthy J, Hammond S, Hone D. Curr Opin Colloid Sci 2000;5:274. [8] Kandori K, Kon-No K, Kitahara K. J Colloid Interface Sci 1991;144:324. [9] Bearchell CA, Heyes DM. Mol Simul 2002;28:517. [10] Bearchell CA, Danks TM, Heyes DM, Moreton DJ, Taylor SE. Phys Chem Chem Phys 2000;2:5197. [11] Cizaire L, Martin JM, Le Mogne Th, Gresser E. Colloids Surf 2004;238: 151. [12] Delfort B, Chivé A, Barré L. J Coll Colloid Interface Sci 1997;186:300. [13] Delfort B, Normand L, Dascotte P, Barré L. J Colloid Interface Sci 1998;207: 218. [14] Mansot JL, Hallouis M, Martin JM. Colloids Surf 1993;75:25. [15] Connor SPD, Crawford J, Cane C. Lubrication Sci 1994;6:297. [16] Moreton DJ. United States patent, 6200936. [17] Gallo R, Jacquet F, Hoornaert P, Roman J-P. Rev Inst Fr Pét 1991;46:251. [18] Abou el Naga HH, Abd El-Azim WM, Bendary SA, Awad NG. Indian Eng Chem Res 1993;32:3170. [19] Markovic I, Ottewill RH, Cebula DJ, Field I, Marsh JF. Colloid Polym Sci 1984;262:648. [20] Markovic I, Ottewill RH. Colloid Polym Sci 1986;264:65. [21] Markovic I, Ottewill RH. Colloid Polym Sci 1986;264:454. [22] Markovic I, Ottewill RH, Singra E, Marsh JF, Heenan RK. Colloid Polym Sci 1992;270:602. [23] Bearchell CA, Heyes DM, Moreton DJ, Hanks TN. Phys Chem Chem Phys 2000;2:5197. [24] Tobias DJ, Klein ML. J Phys Chem 1996;100:6637. [25] Wu RC, Papadopolous KD, Campbell CB. J AIChE 1999;45:2011. [26] Hone DC, Robinson BH, Steytler DC. Langmuir 2000;16:340. [27] Papke BC. Tribol Trans 1988;31:420.