EP0904337A1

EP0904337A1 - Fuel additives

Info

Publication number: EP0904337A1
Application number: EP97918251A
Authority: EP
Inventors: Donald Barr; Stephen Leonard Cook; Paul Joseph Richards; Maurice William Rush
Original assignee: Associated Octel Co Ltd
Current assignee: Innospec Ltd
Priority date: 1996-04-24
Filing date: 1997-04-24
Publication date: 1999-03-31
Also published as: WO1997040122A1; GB9622026D0

Abstract

A process for improving the combustion of fuel and/or improving the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel, the process comprising adding to the fuel before the combustion thereof a composition comprising a mixture of two or more organo-metallic complexes of Group I metals, and fuel additives comprising a mixture of two or more organo-metallic complexes of Group I metals, together with a fuel-soluble carrier liquid miscible in all proportions with the fuel.

Description

Fuel Additives

The present invention relates to a process for improving the combustion of fuel and/or improving the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel. The invention further relates to fuel additives suitable for use in such a process.

Products from the combustion or pyrolysis of hydrocarbon fuels include carbon monoxide, nitrous oxides (N0_X) , unburnt hydrocarbons and particulates. These particulates include not only those particulates which are visible as smoke emission, but also unburned and partially oxidised hydrocarbons from fuel and the lubricants used in engines. The particulate and soot emission are known to be harmful and themselves contain harmful pollutants. In this regard, there is a growing recognition of the health risks associated with particulates emissions. In particular, unburned or partially oxidised hydrocarbons emitted to the atmosphere are irritant astringent materials. Further, in a problem recently highlighted for diesel fuel, emissions of particulate matter of less than 10 micrometers of principle dimension ("PM10 matter") is claimed to cause 10,000 deaths in England and Wales and 60,000 deaths in the USA annually, as published in the New Scientist, March 1994, pl2. It is suspected that these smaller particles penetrate deeper into the lung and adhere.

Diesel fuels and diesel engines, and fuel combustors for heating units, are particularly prone to the emission of small size soot particulate material in the exhaust gas. Diesel engines especially are prone to emission of high levels of particulate matter when the engine is overloaded, worn or badly maintained. Particulate matter is also emitted from diesel engines exhausts when engines are operated at partial load and these emissions are normally invisible to the naked eye. Combustors fuelled by liquid hydrocarbon fuels are also prone to emission of unburned and partially burned substances especially when operated on a frequent start- stop programme or when the burner parts are inadequately maintained. As energy regulations become more stringent the control and stop start operation of combustors must be improved.

Legislation now exists in many countries of the world that is designed to control pollution from diesel engines. More demanding legislation is planned. A number of ways are being examined to enable diesel engines to run and comply with the developing legislation. Engine designs to give effective combustion within the cylinder are being developed. The engine designs developed to achieve low levels of emission are well known to those familiar with the art and examples of such designs are given in S.A.E.

International Congress (February 1995) S.A.E. Special Publication SP - 1092. The drawbacks to the various engine management solutions include cost, complexity and the poor capability for retrofitting. Many modern engine designs use a technology known as Exhaust Gas Recirculation (E.G.R.) . In this regard, exhaust gas recycled in a controlled way to the intake of a diesel exhaust can contribute to the reduction of certain emissions species, mainly oxides of nitrogen. However, there is a drawback in using E.G.R. in that soot particles in the exhaust gas also become recirculated within the engine. Thus, engines running with E.G.R. for prolonged periods of time can become choked with carbon particulate in areas such as the exhaust gas recycle lines and control valves, inlet ports and valves, and the piston top ring glands. Even the piston rings themselves can become choked in the ring grooves. Also, the carbon and other particles become deposited m the engine lubricant so causing premature deterioration of the lubricant.

Particulate traps having the capability to oxidise collected material are also proposed in the light of forthcoming legislation. Such devices are well known to those familiar with the art and some examples are discussed in "Advanced techniques for thermal and catalytic diesel particulate trap regeneration", SAE International Congress (February 1985) SAE Special

Publication - 42. 343-59 (1992) and S.A.E. International Congress (February 1995) S.A.E. Special Publication SP - 1073 (1995) . However, the trap oxidation solutions also suffer from the problems of expense, complexity and poor capability for retrofit. An additional problem is that of trap blockage which causes an increase in exhaust back pressure and a loss of engine efficiency and/or "chimney fires" resulting from sudden and intense burn off of soot from highly loaded traps. Catalytic devices can assist the control of emissions from diesel engines. However, these devices require low sulphur fuel (< 500 ppm) to enable benefits to exhaust emission to be achieved.

Also, low speed engine operation can cause carbonaceous deposits to form on the active parts of the diesel engine oxidation catalyst and so inhibit the effectiveness of the catalyst until a sufficiently high enough gas temperature is available to regenerate the catalyst active surface. Exhaust catalyst devices fitted to diesel and gasoline fuelled engines become effective after engine start up when the exhaust gas passing the catalyst substrate exceeds about 250 C. Experimental work is proceeding to develop catalyst systems effective from temperatures below this level. Details are given in the proceedings of the S.A.E. International Congress February 1995; S.A.E. publications 950404 to 950412, inclusive. Cold engine operation such as stop start driving in gasoline vehicles, or prolonged engine idling for diesel engines, can cause a layer of soot and other carbonaceous material to form over the active catalyst surfaces. The emissions control of catalysts with active surfaces covered with soot and other carbonaceous material is poor, and additional vehicle driving distance or engine operation is necessary to heat the catalyst surfaces to regeneration conditions. Similarly, the performance of a lambda oxygen sensor in the exhaust gas of a gasoline fuelled engine can become degraded by cold engine stop-start driving and the formation of carbonaceous deposits on the exhaust gas sensing surface. Carbonaceous deposits can even form on the combustion surfaces of engines. Particularly affected are gasoline engines where the deposits and residues from the combustion or pyrolysis of fuel and lubricant cause spark knock or can increase the emissions from the engine. Details on these aspects are given in the proceedings of the S.A.E. International Congress February 1995; S.A.E. publication 950680.

Two stroke engines are also prone to the formation of deposits in the combustion chamber, such as on the piston crown and around the piston rings and ring grooves. Deposits also form in the exhaust ports of two stroke engines causing a loss of engine performance efficiency and emission control.

Additives have been used in an attempt to provide solutions to many of these problems.

WO-A-94/11467 to Platinum Plus discloses the use of platinum compounds in conjunction with a trap to lower the unburned hydrocarbon and carbon monoxide concentration of diesel exhaust gases. Lithium and sodium compounds are also claimed to be useful in lowering the regeneration temperature of the trap.

DE-A-40 41 127 to Daimler-Benz describes the use of various fuel soluble, stable lithium and sodium salts in reducing the ignition temperature of the material retained within a diesel particulate filter. Frequent partial unblocking of the filter is observed at sodium levels of around 32 ppm m/m, 28 ppm m/m with lithium.

EP-A-207 560 to Shell concerns the use of succinic acid derivatives and their alkali or alkaline earth metal (especially potassium) salts as additives for increasing the flame speed within spark ignition internal combustion engines.

EP-A-555 006 to Slovnaft AS discloses the use of alkali or alkaline earth metal salts of derivatised alkenyl succinates as additives for reducing the extent of valve seat recession in gasoline engines designed for leaded fuel but used with non-leaded.

GB-A-2 248 068 to Exxon teaches the use of additives containing an alkali, an alkaline earth and a transition metal to reduce smoke and particulate emissions during the combustion of diesel fuel. EP-A-0 476 196 to Ethyl Petroleum Additives teaches the use of a three part composition including a soluble and stable manganese salt, a fuel soluble and stable alkali or alkaline earth metal and a neutral or basic detergent salt to reduce soot levels, particulates, and the acidity of carbonaceous combustion products.

EP-A-0 423 744 teaches the use of a hydrocarbon soluble alkali or alkaline earth metal containing composition in the prevention of valve seat recession in gasoline engines designed for leaded but run on unleaded fuel.

As there is still a need to control the formation of particulates and/or to prevent or to remove carbonaceous deposits, so there is still a need to prepare improved additives which will be of benefit in reducing the rate of deposition or in cleaning up existing deposits.

It has now surprisingly been found that a mixture of two or more organo-metallic complexes of Group I metals acts synergistically to improve the combustion of fuel and/or the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel. According to a first aspect, the present invention thus provides a process of improving the combustion of fuel and/or improving the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel (such as with the use of a particulate trap used with diesel engines) , the process comprising adding to the fuel before the combustion thereof a composition comprising a mixture of two or more organo-metallic complexes of Group I metals. Preferably, the composition comprises a mixture of two identical or different organometallic complexes of different Group I metals .

According to a second aspect, the invention provides the use of a combination of organo-metallic complexes as hereinbefore defined for improving combustion of fuel and/or improving the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel (such as with the use of a particulate trap for use with diesel engines) , wherein the complexes are added to the fuel before the combustion thereof, preferably wherein the total concentration of the metals of the Group I organo-metallic complexes in the fuel before combustion is 100 ppm or less, preferably 50 ppm or less, more preferably 30ppm or less.

Viewed from a yet further aspect, the invention provides a fuel additive comprising a mixture of two or more organo-metallic complexes of Group I metals, together with a fuel-soluble carrier liquid miscible in all proportions with the fuel. Concentrations of the organo-metallic complexes in the additive composition may range from 10 to 90% by weight. As high a concentration as may practically be achieved is preferred. Practical considerations include solubility of the complex and in particular the viscosity of the resulting concentrate. Compositions containing from 40 to 60% by weight of organo-metallic complexes are often preferred as typically offering a good compromise between concentration and viscosity.

Suitable organic carriers for the formulations include aromatic hydrocarbon solvent fractions such as Shellsol AB™, Shellsol R™ and Solvesso 150™. De- aromatised solvent fractions such as Shellsol D70 are also suitable. Other suitable carrier liquids miscible with diesel and other similar hydrocarbon fuels will be apparent to those skilled in the art.

Many types of particulate traps are known to those skilled in the art including as non-limiting examples 'cracked wall' and 'deep bed' ceramic types and sintered metal types. The invention is suitable for use with all particulate traps; the optimum dose rate is a function of the trap type. For use with a particulate filter trap of the 'cracked wall' type, such as the Corning

TM EX80 , a preferred total concentration of the metals of the Group I organometallic complexes in the fuel is 100 ppm or less, preferably 20ppm or less. For use with a particulate filter trap of the 'deep bed' type, such as

TM one constructed from 3M Nextel fibre, a preferred total concentration of the metals of the Group I organo¬ metallic complexes in the fuel is 50 ppm or less, preferably 20ppm or less, e.g. 5ppm or less.

The key advantages of the present invention are that it provides additives for diesel and other hydrocarbon fuels that give an overall emissions benefit to the environment on combustion by any one or more of: improving the combustion process,- controlling the formation of soot and carbonaceous deposits in engines and combustors; and improving the oxidation of particulates within trap systems, engines or exhaust systems. The composition of the present invention promotes and sustains combustion in the trap. Another key advantage is that the composition of the present invention may be used in low dosage amounts. Whilst any combination of two or more alkali metals may be used, combinations of potassium and sodium are particularly preferred. This is on the grounds of their low toxicity, low cost and the ready availability of sources of these metals compared to the other alkali metals lithium, rubidium and cesium. Preferred organometallic complexes of potassium and sodium are those which may readily be prepared from an inexpensive salt of the metal, such as the hydroxide or oxide.

Preferably, each organo-metallic complex is fuel soluble.

Preferably, each organometallic complex is soluble in a fuel-compatible solvent such that each organometallic complex is soluble to the extent of 10 wt%, preferably 25 wt% and most preferably 50 wt% or more in the solvent. Conveniently, the fuel-compatible solvent may comprise a poly(butene) .

Where two Group I organometallic complexes are present, the ratio of these is preferably in the range of from 20:1 to 1:20, e.g. from 10:1 to 1:10. More preferably, the ratio of these is in the range of from

20:1 to 1:1, preferably from 10:1 to 1:1. Where the two complexes are of sodium and potassium, these are advantageously added such that the ratio by weight of sodium to potassium is in th range of from 20:1 to 2:1, more preferably from 10:1 t :1.

Preferably, each of th rgano-metallic complexes is of the formula M(R)_m.nL wiι_re M is the cation of an alkali metal of valency m; R is the residue of an organic compound RH, where R is an organic group containing an active hydrogen atom H replaceable by the metal M and attached to an O, S, P, N or C atom in the group R; n is a positive number indicating the number of donor ligand molecules forming a bond with the metal cation, but which can be zero; and L is a species or functional group capable of acting as a Lewis base.

As used herein, the term "species capable of acting as a Lewis base" includes any atom or molecule that has one or more available electron pairs in accordance with the Lewis acid-base theory.

Preferably, R and L for at least one of the complexes, preferably for each of the complexes, are in the same molecule.

Each organometallic complex may be dosed to the fuel at any stage in the fuel supply chain.

Preferably each complex is added to the fuel close to the engine or combustion systems, withm the fuel storage system for the engine or combustor, at the refinery, distribution terminal or at any other stage in the fuel supply chain.

The term "fuel" includes any hydrocarbon that can be used to generate power or heat. The term also covers fuel containing other additives such as dyes, cetane improvers, rust inhibitors, antistatic agents, antioxidants, reodorants, gum inhibitors, metal deactivators, de-emulsiflers, upper cylinder lubricants, and anti-icing agents. Preferably, the term covers diesel fuel.

The term "diesel fuel" means a distillate hydrocarbon fuel or for compression ignition internal combustion engines meeting the standards set by BS 2869 Parts 1 and 2 as well as fuels in which hydrocarbons constitute a manor component and alternative fuels such as rape seed oil and rape oil methyl ester.

The combustion of the fuel can occur in, for example, an engine such as a diesel engine, or any other suitable combustion system. Examples of other suitable combustion systems include recirculation engine systems, domestic burners and industrial burners The present invention therefore relates to additives for liquid hydrocarbon fuel, and fuel compositions containing them.

The composition of the present invention can have many uses, some of which are now described.

In engine management approaches, there is a well- known trade-off between N0_X and particulates emissions. Diesel engines emissions tests now include specified levels for many pollutants. In some instances, the composition of the present invention achieves a useful level of particulates suppression and to such an extent that it decouples this trade off, thereby giving the engineer more freedom to achieve power output or fuel economy within a given emission standard. In trap approaches, the composition of the present invention may be effective in reducing engine out emissions or as a combustion catalyst aiding the oxidation of trapped particles. Either way, the composition of the present invention provides for simpler, safer and less costly traps by enabling less frequent, less intense or less energetic regeneration, whether the heat required for the regeneration is provided by the exhaust gas or through some external mechanism. In some instances, the combustion of fuel containing the composition of the present invention enables engines to be run at a full load and at a fractional load with a suitable trap arrangement and in doing so a self regenerating mechanism is initiated. In some instances, when an engine and associated particulate trap are run burning a fuel containing the composition of the present invention there are provided two broad modes of trap function. First, a soot and particulate trapping stage associated with a minor clogging function can be observed. This is then followed by an automatic burn off or self-regeneration function. Trap conditions which favour self regeneration are influenced by particulate size and formation, the composition of unburned hydrocarbons, the back pressure and composition of the exhaust gas in the exhaust system. These discrete functions of trapping then burn off are particularly recordable at light to medium engine duty.

Up until now, many diesel trap devices have required complicated devices to initiate and control the exotherm of trap regeneration. In some instances, the composition of the present invention can significantly reduce or eliminate the need for regeneration initiation and control devices. The need for energy input to initiate the regeneration can also be substantially reduced or eliminated for many engine designs. At conditions of medium to full engine load the trapping and regeneration mechanisms operate simultaneously giving excellent control of the particulate emissions from diesel exhaust.

Preferably, the composition of the present invention is designed to remain compatible with hydrocarbon fuels and remain stable up to the point of entry to the combustion zone. The composition of the present invention when burned with the fuel can reduce the soot and carbonaceous material entrained in the exhaust gas recycle system of certain engines. Thus, the levels of soot and carbonaceous material that are subsequently trapped in the engine becomes reduced.

Burning of a fuel comprising the composition of the present invention gives particulate matter remaining in the exhaust gas which is in a form readily collectable on a trap. Further, when the fuel is burned with the additive of the present invention the trapped material exhibits a reduced ignition temperature and oxidation of the trapped material is enhanced, when compared to that of fuel burned without the composition of the present invention. The burning of soot and other hydrocarbons from the surfaces of a trap therefore provides a way to regenerate the filter and so prevent the unacceptable clogging of particulate traps.

As used herein, the term "regeneration" or "regenerating" means cleaning a particulate trap so that it contains minimal or no particulates. The usual regeneration process includes burning off the trapped particulates in and on the particulate trap. Regeneration of the trap is accompanied by a reduction in pressure drop across the trap. The synergistic combination of two alkali metals offers a number of advantages. Firstly, the regeneration of the trap may be enhanced such that there is a lower average back pressure across the trap. Secondly, regeneration of the trap may be caused to occur with greater reliability and frequency, such that there is a less extreme variation in back pressure across the trap. This can be detected e.g. by logging trap back pressure at regular discrete intervals whilst running the engine under steady conditions and determining the standard deviation as well as the mean of the back pressure readings.

Preferably, the composition of the present invention is designed so that very low levels of combustion or pyrolysis ash are formed. In this way clogging of the trap from additive residue is kept to a minimum.

When a fuel comprising the composition of the present invention is burned the fuel may cause the carbonaceous deposits that form during stop start driving on the active surfaces of catalytic converters to be cleared away even from low driving duty thereby enabling a fast light off or early regeneration to full conversion efficiency.

When a fuel comprising the composition of the present invention is burned the fuel provides a significant reduction in levels of soot and carbonaceous deposits that form on the combustion surfaces of engines in the piston rings and piston ring bands, and also in the exhaust ports, thereby contributing to a maintenance of engine performance emissions and longevity. Preferably, the composition of the present invention is designed such that the soot and hydrocarbons burned become emitted as water vapour, carbon monoxide and carbon dioxide.

In a preferred embodiment, the composition of the present invention provides ultimate products that are readily water soluble, or soluble in solvents non- corrosive towards exhaust system components thereby simplifying any recycle of the system.

Preferably the composition of the present invention is fuel-soluble or fuel miscible. This serves to reduce the complexity and cost of any on-board dosing device. A further advantage of a highly preferred composition of the invention is that it can be supplied in concentrated form in a suitable solvent that is fully compatible with diesel and other hydrocarbon fuels, such that blending of fuel and additive may be more easily and readily carried out.

A further advantage of a highly preferred composition of the present invention is that it is at least resistant and preferably totally inert towards water leaching, thus providing a fuel additive that is compatible with the fuel handling, storage and delivery systems in common use. In particular, diesel fuel often encounters water, especially during delivery to the point of sale and so the composition of the present invention is not affected by the presence of that water. In one aspect of the present invention, the alkali metal complexes have the general formula

M(R)_m.nL

where M is the cation of an alkali metal of valence m, R is the residue of an organic compound of formula RH where H represents an active hydrogen atom reactive with the metal M and attached either to a hetero atom selected from O, S and N in the organic group R, or to a carbon atom, that hetero or carbon being situated in the organic group R close to an electron withdrawing group, e.g. a hetero atom or group consisting of or containing 0, S or N, or aromatic ring, e.g. phenyl, n is a number indicating the number of organic electron donor molecules (Lewis bases) forming dative bonds with the metal cation in the complex, usually up to five in number, more usually an integer from 1 to 4, and L is one or more organic electron donor ligands (Lewis base) . R and L may be combined in the one molecule, in which case n can be and often is zero and L is a functional group capable of acting as a Lewis base.

In a more detailed aspect, the Lewis base metallo- organic co-ordination complexes used in accordance with the present invention contain the residue of an organic molecule RH which contains an active hydrogen atom H which is replaceable with a metal cation. In the organic compound RH the active hydrogen atom will be attached to a hetero atom (O, S, or N) or to a carbon atom close to an electron withdrawing group. The electron withdrawing group may be a hetero atom or group consisting of or containing O, S, or N, e.g. a carbonyl (>C=0) , thione (>C=S) or imide (>C=NH) group, or an aromatic group, e.g. phenyl. When the electron withdrawing group is a hetero atom or group, the hetero atom or group may be situated in either an aliphatic or alicyclic group, which, when the active hydrogen group is an >NH group, may or may not, but usually will contain that group as part of a heterocyclic ring.

Suitable complexes are derived from a β-diketone of the formula R¹C(0)CH₂C(0)R²

where R¹ or R² is C,-C₅, alkyl or substituted alkyl, e.g. halo-, amino-, alkoxy- or hydroxyalkyl-, C₃-C₆ cycloalkyl, benzyl, phenyl or Cj-C₅ alkylphenyl, e.g. tolyl, xylyl, etc., and where R¹ may be the same as or may be different to R². Suitable β-diketones include: hexafluoroacetylacetone: CF₃C(0)CH₂C(0)CF₃ (HFA) ; 2,2,6,6-tetramethylheptane-3,5-dione: (CH₃)₃CC(0)CH₂C(0)C(CH₃)₃.

If the active hydrogen atom is attached to oxygen in the organic compound RH, then suitable compounds include phenolic compounds containing from 6-30 carbon atoms, preferably substituted phenols containing from 1- 3 substituents selected from alkyl, alkylaminoalkyl, and alkoxy groups of 1-8 carbon atoms, e.g. cresols, guiacols, di-t-butylcresols, dimethylaminomethylene- cresol. The substituted phenols are particularly preferred.

Especially preferred compounds wherein the hydrogen atom is attached to oxygen in the organic compound RH are those derived from reaction of a metal hydroxide or other alkali metal source with an alkyl or alkenyl substituted succinic anhydride or the hydrolysis product. Typically such anhydrides are those prepared by reaction of oligomerised isobutenes or other simple olefins with maleic anhydride. A wide variety of such alkyl or alkenyl substituted succinic anhydrides and a range of techniques for their preparation are known to those skilled in the art. In general, a high molecular weight poly(isobutene) substituent provides the resulting complex with good hydrocarbon solubility at the cost of lower metal content. We have found the alkenyl substituted succinic anhydride derived from the

TM thermal reaction of BP Napvis X-10 with maleic anhydride to give a good compromise between hydrocarbon solubility and metal content. Whilst not wishing to be bound by theoretical considerations, it is believed that in such compounds one carboxylic acid group is deprotonated and bound in salt-like fashion to metal ion and the second carboxylic acid group to be protonated and to bind as a Lewis base.

If the active hydrogen is attached to a nitrogen atom in the organic compound RH, then suitable compounds are heterocyclic compounds of up to 20 carbon atoms containing a -C(Y)-NH- group as part of the heterocycle, Y being either O, S or =NH. Suitable compounds are succinimide, 2-mercaptobenzoxazole, 2-mercapto- pyrimidine, 2-mercaptothiazoline, 2- mercaptobenzimidazole, 2-oxobenzoxazole.

In more detail, L can be any suitable organic electron donor molecule (Lewis base) , the preferred ones being hexamethylphosphoramide (HMPA) , tetramethylethylenediamine (TMEDA) , pentamethyldiethylenetriamine, dimethylpropyleneurea (DMPU) , dimethylimidazolidinone (DMI) , dimethylcarbonate (DMC) , dimethylsulphoxide (DMSO) , dimethylformamide (DMF) . Other possible ligands are diethylether (Et₂0) , 1,2-dimethoxyethane (monoglyme) , bis(2- methoxyethyl) ether (diglyme) , dioxane, tetrahydrofuran. Where R comprises L, L is a functional group capable of acting as a Lewis base donor, preferred ones being dimethylaminomethyl (-CH₂N(CH₃) ₂) , ethyleneoxy(-OCH₂CH₂0-) , ethyleneamine (-N(R)CH₂CH₂N(R) -) , carboxy( -C0₂H) and ester (-C0₂CH₂R) . It is to be understood that these listings are by no means exhaustive and other suitable organic donor ligands or functional groups (Lewis bases) may be used. The alkali metal complex will usually contain 1-4 ligand molecules to ensure oil solubility, i.e. the value of n will usually be 1, 2, 3, or 4. Where R comprises L, n can be and often is zero.

Whilst the organometallic compounds described may be added directly to the fuel, either external to the vehicle or by using an on board dosing system, they will preferably first be formulated as a fuel additive composition or concentrate containing the substance, or mixtures thereof possibly along with other additives, such as detergents, anti foams, dyes, cetane improvers, corrosion inhibitors, gum inhibitors, metal deactivators, de-emulsiflers, upper cylinder lubricants, anti-icing agents, etc., in an organic carrier miscible

It is desirable to have a high concentration of organo-metallic complex present in the fuel additive composition. There is a trade-off between the viscosity of the composition, the concentration of organo-metallic complex and the temperature at which the composition is held. The most suitable concentration is thus a function of the type of mixing equipment, if any, to be employed, the ambient temperatures expected to be encountered and whether heated containers, pumps, lines etc. are to be used. Conveniently, the concentration of the organo-metallic complex may range from 10 to 90% by weight. However, compositions containing from 40 to 60% by weight of organo-metallic complex are preferred.

The composition of the present invention reduces the ignition temperature and/or promotes oxidation of particulate matter. Without wishing to be bound by theory, it is believed that there are four basic mechanisms to explain soot formation and decay. These are: mass growth, coagulation, pyrolysis and oxidation. Earlier workers have suggested that metallic additives appear to work by enhancing oxidation rather than reducing soot formation. Alkali metals, particularly metal oxides thereof, have been shown to be effective in rich pre-mixed flame studies. Suggested mechanisms for alkali metals include a charge transfer process which limits coagulation, especially in the combustion space of a diesel engine cylinder, thus promoting soot burn out and limiting the formation of larger more stable soot particles. In this context "larger" refers to particle sizes in the ranges of 300 to 700 nanometres principle dimension. Thus it is believed this contributes to the surprising synergistic combustion influence of the combination of the alkali metal complexes of the composition of the present invention. In addition, the seemingly random low temperature oxidation of soot and the auto regeneration in the range of 185°C to 220°C for the preferred composition of the present invention may be due to the formation of short lived species, during the combustion or pyrolysis event, such as a superoxide or peroxide radical. Dependent upon exhaust gas temperature and partial pressure of oxygen in the exhaust gas it is possible that metal peroxides may form from metal carbonates (see McKee and Chatterji, Carbon JL2.:381-90) . It is expected that the metals will be emitted from the cylinder as carbonate salts or soon form such salts under conditions present in the exhaust manifold.

A particular advantage of the complexes of this invention is their low nuclearity, many being monomeric in character, although some are dimeric and trimeric, tetrameric or higher. This low nuclearity means that, in contrast to overbased metal soaps (i.e. the traditional method of providing oil-soluble metal compounds) the complexes used in accordance with the present invention provide a uniform distribution of metal atoms throughout the fuel, each metal atom theoretically being available to enhance combustion of particulates both within the engine and exhaust system and in traps. In contrast, the overbased metal soaps essentially consist of individual micelles containing a number of metal (e.g. alkali or alkaline earth metal) cations and inorganic anions, typically carbonate, surrounded by a shell of dispersant type molecules on the surface of the particle. Whilst some overbased soaps are stably dispersed, the metal will not be uniformly dispersed throughout the fuel as individual atoms, but in clusters, or micelles. Further, only a limited number of metal atoms are available on the surface of the micelle for action, so the effectiveness of those soaps is low. Also, since the soaps are non¬ volatile there is a significant risk of increased deposit formation in the engine itself and in the fuel injectors, including the fuel injectors of oil fire boilers etc.

The effectiveness of the composition of the present invention is also attributable to its volatility as the combustion process is a vapour phase reaction, essentially requiring the particulate suppressant to be volatile in order to have an effect.

The present invention will now be described by way of the following non-limiting examples and with reference to the accompanying figures in which:

Figure 1 indicates the pressure drop across a 'cracked wall' trap with time on combustion of a fuel in accordance with the invention.

Figure 2 indicates the effect of Na:K ratio on mean exhaust gas pressure.

Figure 3 indicates the effect of Na:K ratio on exhaust gas pressure (mean + 2 standard deviations) .

Example 1: Preparation of 1,3-dimethylimidazolidinone adduct of sodium 2,2, 6,6-tetramethylheptane-3,5-dionate: [Na(TMHD) .DMI]

A round bottom flask was charged under nitrogen with sodium hydride (NaH, 4.8 g, 200 mmol) , dry toluene (100 cm³) and dimethylimidazolidinone (23.8 cm³, 22.8 g, 200 mmol) . 2,2, 6, 6-tetramethylheptane-3, 5-dione (HTMHD, 43 cm³, 37.97 g, 206 mmol) was then added dropwise by syringe against nitrogen flush. After the addition of a few drops an effervescence was noted. The solution was stirred and gently warmed (oil bath, 60°C) during one hour before filtration. A large batch of crystals grew on refrigeration. Melting point 70-72°C C/H/N found versus (calculated) wt%, C 60.09(60.00) , H 9.14 (9.06) and N 8.67(8.85) , ^lH nmr in C₆D₆ shifts rel. to TMS 5.873 ppm (s, H, COCHCO) , 2.609 (s, 6H, NCH₃) , 2.570 (s, 4H, CH₂CH₂) and 1.396 (s, 18H, C(CH₃)₃) •

Example 2: Preparation of sodium salt of poly(isobutenyl) succinic acid, approx. 1,000 molecular weight [Na(PIBSA₁₀₀₀) ]

A suspension of powdered solid sodium hydroxide (8.04g, 200 mmol) in a solution of poly(isobutenyl) succinic anhydride (PIBSA, 198.8 g, 200 mmol) in dry toluene (995 cm³) was allowed to stir at ambient temperature during several days. The solids dissolved to yield a cloudy suspension and ultimately a clear solution.

Example 3 : Preparation of dimethylcarbonate adduct of the sodium salt of 2, 6-ditertiarybutyl-4-methyl phenol: (NaBHT)₂.3DMC]

A solution of 2, 6-ditertiarybutyl-4-methyl phenol (butylated hydroxy toluene, BHT, 21.8 g, 100 mmol) in dry toluene 100 cm³) is added to a suspension of sodium hydride (2.4 g, 100 mmol) in dry toluene (100 cm³) and dimethyl carbonate (12.64 cm³, 13.51 g, 1.5 equiv) under inert atmosphere. Precipitation of white material accompanied the evolution of hydrogen gas and heat. Alter completion of the addition the reaction mixture was stirred at ambient temperature during some 60 minutes. The solids were isolated by filtration and dried under vacuum.

C/H/N found versus (calculated) wt%, C 62.40(62.07) and H 8.28 (8.49) . Example 4: Preparation of the 1,3- dimethylimidazolidinone (DMI) adduct of potassium 2,2,6,6-tetramethyl-3,5-heptanedionate: [{K(TMHD) }₂.DMI]

Potassium hydride (KH, 0.90 g, 22.5 mmol) was washed of mineral oil, dried and placed in a Schlenk tube. Hexane was then added followed by DMI (7 ml, 64.22 mmol) . Some effervescence occurred, implying reaction or dissolution, and a green coloration was apparent. TMHD (4.4 ml, 21.05 mmol) was then added slowly, as a very vigorous reaction takes place. After about fifteen minutes the reaction subsided and an oil settled out of solution. The two-phase liquid was cooled in an ice-box (to -10°C) and a solid crystalline mass formed from the oil part over about half an hour.

The crystalline solids were washed with hexane, isolated and determined to be a dimethylimidizolidinone adduct of potassium 2,2,6, 6-tetramethylheptane-3, 5-dione: [{K(TMHD) }₂.DMI] . Yield 1.7 g, 16% first batch based on a 1/2 ligand:donor ratio

Formula:

K[ (CH₃)₃C(-0)CH₂C(=0)C(CH₃)₃] .0=CN(CH₃) CH₂CH₂N(CH₃) , Mw 450.678, m.p. 64-68°C

Example 5: Dimethylimidazolidinone adduct of potassium 2, 6-ditertiarybutyl-4-methyl phenol, [K(BHT) .2DMI] .

The method of example 3 is used, with the appropriate change in Lewis base and Lewis base:metal ratio. The adduct is sufficiently soluble in toluene to permit recrystallisation. The microcrystals obtained have melting point 92-96°C. Example 6: Dimethylimidazolidinone adduct of sodium 2,6' ditertiarybutyl-4-methyl phenol, [Na(BHT) .3DMI] .

The method of example 3 is used, with the appropriate change in Lewis base and Lewis base:metal ratio. The adduct is sufficiently soluble in toluene to permit recrystallisation. The crystals obtained have melting point 96-98°C.

Example 7: Dimethylimidazolidinone adduct of sodium 2- methoxy phenol, [Na(TMP) .DMI] .

The method of example 4 is used, with the appropriate change in Lewis base:metal ratio. The adduct is sufficiently soluble in toluene to permit recrystallisation. The crystals obtained have melting point 87-89°C.

Example 8: Preparation of the sodium salt of molecular weight 420 poly(isobutenyl)succinic anhydride.

TM A thermostatted 'Soverel' reactor was charged with BP

TM Hyvis XD-35 poly(isobutene) (665.79 g, no. av. mol. wt

320,2.08 mol) and maleic anhydride (411.79 g, 4.2 mol, 2.02 equivalents). The contents were heated to 200°C with oil circulated through the jacket by an external oil bath and strongly stirred during 8 hours. A viscous, dark brown solution formed. The unreacted maleic anhydride was removed under vacuum, along with some of the unreacted poly(isobutene) . A material analysing at 11.2 wt% poly(isobutene) was recovered.

A sample of the material prepared above (535.78 g, theoretical 1.125 moles PIBSA₄₂₀) was charged to a flat- bottomed glass vessel fitted with turbine agitator, thermocouple well and charging port. The vessel was further charged with Solveso 150 (502.26 g) . The contents were warmed to 82°C via an external oil bath and stirred until homogenous. Beaded sodium hydroxide (46.03 g, 1.15 moles) was then charged. The resulting suspension of white 1 mm beads in brown solution was stirred overnight at 78°C. By visual inspection, no unreacted NaOH remained by the following morning. FTIR inspection of the 1863 cm^"1 peak suggested that about 7% of the PIBSA remained unreacted. The solution was allowed to stand for 2 hours without stirring, but with heating. It was then carefully decanted. Some 1066.19 g of material was collected.

The reactor was carefully washed and the washings analysed for sodium. By mass balance the NaPIBSA₄₂₀ solution was expected to contain 2.37 wt% Na, it was found to contain 2.13% by analysis.

Example 9: Preparation of No. Average Molecular Weight 420 Poly(isobutylene) Succinic Anhydride - PIBSA₄₂₀.

TfΛ

A reactor was charged with BP-Hyvis XD-35 poly(isobutylene) (12.906 kg, 40.33 mol) and heated to 100°C with stirring before adding maleic anhydride (5.966 kg, 60.88 mol) . The temperature of the oil bath supplying the reactor jacket was set to 220°C, the internal reactor temperature reached 185°C after three hours. This was taken as the start of the reaction time. The oil bath temperature was lowered to 212°C and the reaction mix stirred during some 30 hours. At the end of this period a vacuum was applied and the excess maleic anhydride distilled out. After 15 hours under vacuum, residual maleic anhydride content was 0.0194 wt% and residual PIB 19.9 wt% . Some 13.888 kg of brown, viscous material was recovered. Example 10: Preparation of Potassium Salt of PIBSA₄₂₀.

An oil-jacketed reactor was charged with material prepared in example 13 (440.78 g, 0.85 mol PIBSA₄₂₀) , and

TNI Solveso 150 (462.53 g) . The contents were warmed to 50°C and stirred until homogenous. KOH flake (47.88 g, 0.77 mol if 10% H₂0) was then added with stirring and the resulting suspension left to stir overnight. The solids dissolved and FTIR analysis showed an absence of the 1863 cm^"1 absorption due to the PIBSA. The solution was analysed by AAS as containing 3.33 wt% K.

Example 11: Preparation of No. Average Molecular Weight 360 poly(isobutylene) succinic anhydride (PIBSA₃₆₀) .

A number average molecular weight 260 poly(isobutylene) (PIB₂₆₀, BP-Napvis XlO , 586.2 g, 2.257 moles) was charged to a one litre oil-jacketed reaction vessel. The vessel was further charged with maleic anhydride (442.71 g, 4.52 moles). The mixture was heated to 200°C and stirred during 24 hours. At the end of this period, the maleic anhydride was removed by vacuum distillation.

A dark brown viscous oil was recovered. This analysed as 8.1% m/m PIB₂₆₀.

Example 12: Preparation of sodium salt of No. Average Molecular Weight 360 Poly(isobutylene) Succinic Acid - Na(PIBSA_3S0) .

A reactor was charged with a sample of poly(isobutylene)succinic anhydride prepared as above (412.91 g, 392.26 g PIBSA₃₆₀, 1.096 moles, 20.65 g PIB₂₆₀) .

TM

The vessel was further charged with Solveso 150 (526.19 g) and the liquids heated and stirred to form a homogenous deep brown solution. Sodium hydroxide as dry pellets (43.84 g, 1.096 mol) was then added. The resulting suspension was stirred overnight at 70°C. FTIR indicated complete consumption of the PIBSA and formation of carboxylic acid and carboxylic acid salt. The stirring was switched off and a small amount of NaOH (<0.5 g) allowed to settle out whilst the solution was kept warm. The solution was carefully decanted and analysed as containing 2.35 wt% Na by AAS versus 2.56% calculated.

Example 13: Preparation of Sodium salt of 2-ethyl, 2- phenyl butyric acid, 1,3-Dimethyl-3, 4, 5, 6-tetrahydro- 2 (IH) -pyrimidinone (DMPU) Adduct.

PhEt₂CC0₂H (3.45 g, 20 mmol) was added to a slurry of NaH (0.48 g, 20 mmol) and DMPU (2.42 cm³, 10 mmol) in dry toluene (30 cm³) in a dry Schlenk tube under nitrogen atmosphere. The solids dissolved in a few minutes accompanied by a mild exotherm and some gas evolution to give a clear pale straw coloured solution. The solution was filtered and toluene removed under vacuum until solids began to crystallise out. The solids were warmed to dissolve, then recrystallised on slow cooling, first to ambient, then -20°C.

STA and C/H/N analysis indicated the adduct to contain non-integer amounts of DMPU.

Addition of NaOH granules (5.36 g, 0.134 mol) to a solution of PhEt₂CC0₂H (25.02 g, 0.13 mol) and DMPU (35.35 g, 0.276 mol) in Shellsol AB™ (154.20g) . A homogeneous solution formed. FTIR analysis showed a single peak for the DMPU carbonyl, indicating all of the material to be complexed to sodium, or none.

The experiment was performed with PhEt_?CC0₂H (50.17 g,

0.261 mol) and NaOH (10.70 g, 0.268 mol) in Shellsol AE (243.71 g) in the absence of DMPU. A gel-like, highly viscous, cloudy solution resulted.

Example 14: Preparation of Sodium Salt of 2-Ethyl- Hexanoic Acid, DMPϋ Adduct.

A dry Schlenk tube was charged with NaH (0.48 g, 20 mmol), toluene (30 cm³) and DMPU (2.42 cm³, 20 mmol) . 2-Ethyl-hexanoic acid (3.8 cm³, 20 mmol) was then added dropwise with caution. A vigorous effervescence accompanied the dissolution of the solids to a clear, yellow solution. The solution was filtered, then reduced in volume to a pale yellow oil. Even on prolonged evacuation and cooling to -20°C oil did not provide crystals.

C/H/N analysis gave (% m/m) C 56.85, H 9.74 and N 9.34 versus 57.14, 9.18 and 9.52, respectively, for Na(2- ethylhexanoate) .DMPU.

A similar experiment, in the absence of DMPU, gave a viscous gel which could not be filtered.

A further similar experiment, using NaOH in place of NaH also yielded an oil. C/H/N and nmr analysis gave no indication of the presence of water in the adduct.

Trap Regeneration Tests Using Cracked Wall Trap

Additised fuel was prepared by dissolving the required amounts of additive or additives (depending on the test) in one litre of base diesel fuel, then diluting in the base fuel until the fuel finally contained an additional total 17 ppm m/m of total metal or metals, as appropriate, above background level. Base fuel used was BPD28, as specified below. Additives used were sodium and potassium salts of PIBSAs prepared from BP Napvis X-

TM

10 . Sodium and potassium concentrations in each additive were determined by mass balance,

DIESEL ANALYSIS

DESCRIPTION OF SAMPLE BPD28

SAMPLE NO. 954953

DENSITY @ 15 C 0.8359

VISCOSITY @ 20 C

VISCOSITY @ 40 C 2.827

CLOUD POINT C -5

CFPP C -20

POUR POINT C -24

FLASH POINT C 66

SULPHUR % WT. 508

FIA ANALYSIS % vol saturates 74.3

% vol olefins 1.5

% vol aromatics 24.2

Initial boiling point @ C 166.0

5% VOL. @ C 190.5

10% VOL. @ C 204.5

20% VOL. @ C 228.0

30% VOL.® C 248.0

40% VOL. @, C 264.5

50% VOL. @ C 277.0

65% VOL. @ C 293.5

70% VOL. @ C 300.5

85% VOL. @ C 322.5

90% VOL. @ C 332.5

95% VOL. @ C 347.0

FBP @ C 375.5

% VOL. RECOVERY 98.3

% VOL. RESIDUE 1.7

% VOL. LOSS 0.0

CETANE IMPROVER - % NIL

CETANE NUMBER 53.1 The test vehicle was a Peugeot 309 diesel, specified as bbeellooww,, aanndd ffiitttteedd with a 'cracked wall' trap prepared

T TTΛ from Corning EX80

Model 309D Body 4 seat saloon

Arrangement Front wheel drive Kerb Weight kg 990 Engine type Diesel indirect injection Swept volume 1 1.905, normally aspirated

Compression ratio 23.5:1 Bore, stroke mm 83,88 Fuel pump Rotary type Rotodiesel

Transmission 5 speed manual

The car was driven on a single rolls endurance dynamometer at steady speed and level road drag power for the unladen vehicle such that at the inlet to the trap the exhaust gas temperature was about 195°C for a clean trap. The dynamometer used was outdoors and so the exact speed was a function of ambient temperature and weather conditions, typically the speed was 60 km/hr. Test duration was about 24 hrs.

The pressure-drop across the filter was measured using a pressure transducer and the data logged at programmed intervals. A plot of pressure drop versus time yielded a characteristic 'saw-tooth' curve consisting of a series of gradual increases in pressure drop interspersed at intervals with sharp decreases associated with ignition of the trapped material and hence regeneration of the filter. A typical such trace is shown as Figure 1 hereto. At the conclusion of each test the engine speed and load were steadily increased until regeneration occurred. This meant that succeeding additives started from a minimally loaded trap. The readings for pressure drop were taken and, using conventional techniques, the mean value of the pressure drop and the standard deviation of the pressure drop readings were determined. The following table shows total ppm m/m metal above background value for each of sodium and potassium as well as the mean back pressure and the sum of mean back pressure plus twice standard deviation therein resulting from use of the listed doses in the test described above.

Na K Pressure drop aero

(ppm) (ppm) Mean Mean + 2σ

17 0 166 362

17 0 195 486

17 0 160 200

17 0 147 231

15 2 142 220

13 4 131 203

13 4 145 241

8.5 8.5 240 532

4.25 12.75 254 558

0 17 210 548

This data indicates that 1:1 and lower ratios of Na:K perform less well than Na alone. The 15:2 and 13:4 Na:K mixtures perform as well as the best performance seen from fuel additised with sodium additive alone. Additionally, the action of the sodium additive is seen to be somewhat prone to fluctuation, even amongst such prolonged runs. By contrast, the action of the synergistic combination of additives is evidently much more consistent and reliable. Such consistency is a highly desirable feature.

The effectiveness of the synergistic combination at the favoured Na:K ratios for reducing pressure drop and pressure drop plus 2σ is better illustrated by the charts shown in Figures 2 and 3, respectively. Further, the average values of mean back pressures and mean back pressures plus twice standard deviation may be compared.

Na (% total Averages of metal, Mean . back pressure Mean plus 2σ residual is K) (mbar) (mbar)

100 167 (4 runs) 320 (4 runs)

88.24(15:2) 142 220

76.47(13:4) 138 (2 runs) 222 (2 runs)

50(1:1) 240 532

25(1:3) 254 558

0 210 548

Using these comparisons both the increased reliability (increased regeneration frequency) and improved completion of regeneration by the synergistic combination can be seen. The advantages of the chosen range of synergistic combinations are thus that they provide a much more reliable filter regeneration operation and give a more complete and/or more frequent filter trap regeneration.

Additional Trap Regeneration Tests using Cracked Wall

Trap

In an attempt to reduce the spread of results observed in the earlier tests through reduction of variability in conditions, the tests were repeated in immediate succession. A total additive concentration of 20 ppm m/m metal was prepared, as set out in the previous example. Additives were again salts of PIBSAs prepared

TM from BP Napvis X-10 . The sodium:potassium ratios shown below were used. Blend Number Sodium Potassium

Concentration Concentration

Ppm m/m Ppm m/m

1 20 0 2 17.5 2.5 3 15 5 4 13.3 6.7 5 10 10 6 0 20

The base fuel use was BPD 29, specified as below:

Density @ 15 C 0.8355 Viscosity ® 40 C 2.931 Cloud point C -6 CFPP C -20

Pour point C -27 Sulphur ppm m/m 457

FIA Analysis, % vol saturates 75.0 % vol olefins 2.0 % vol aromatics 23.0

Distillation, IBP @ C 175.0

5% vol @ C 206.0

10% vol @ C 219.5

30% vol @ C 239.0

40% vol ® C 255.0 50% vol @ C 267.0

65% vol @ C 277.5 70% vol @ C 298.5 85% vol @ C 320.0 90% vol @ C 330.0 95% vol @ C 344.5

FBP @ C 352.5 % vol recovery 97.8 % vol residue 2.0

% vol loss 0.2

Cetane Number 54.4 CCI (IP380) 54.9

The test vehicle, trap, dynamometer, driving and monitoring methods were as detailed for the previous example. The readings for pressure drop were logged. The data log was taken and values at periods of engine downtime or where pressure was not changing by more than 1 mBar per minute were removed. The mean and standard deviations of the values for the remaining measurements were determined by standard methods.

The results set out in the table below were obtained:

Run No. Na K Mean Back Mean Back ppm ppm Pressure mBar Pressure plus two mm//mm mm//mm standard deviations mBar

1 20 0 197 426

2 17.5 2.5 188 372 3 3 1 155 5 5 2 20000 404

4 13.3 6.7 213 432

5 10 10 244 468

6 0 20 261 515

A benefit from simultaneous use of a major proportion of sodium and a minor proportion of potassium over either metal alone can again clearly be seen.

Claims

1. A process for improving the combustion of fuel and/or improving the oxidation of carbonaceous products derived from the combustion or pyrolysis of fuel, the process comprising adding to the fuel before the combustion thereof a composition comprising a mixture of two or more organo-metallic complexes of Group I metals.

2. A process as claimed in claim 1 wherein the composition comprises a mixture of two identical or different organo-metallic complexes of different Group I metals.

3. A process as claimed in claim 1 or claim 2 wherein the total concentration of the metals of the Group I organo-metallic complexes in the fuel before combustion is 100 ppm or less.

4. A process as claimed in claim 3 wherein the total concentration of the metals of the Group I organo¬ metallic complexes in the fuel before combustion is 50 ppm or less.

5. A process as claimed in claim 3 wherein the total concentration of the metals of the Group I organo¬ metallic complexes in the fuel before combustion is 30 ppm or less.

6. A fuel additive comprising a mixture of two or more organo-metallic complexes of Group I metals, together with a fuel-soluble carrier liquid miscible in all proportions with the fuel.

7. A fuel additive as claimed in claim 6 wherein said organo-metallic complexes are present at a total concentration of from 10 to 90% by weight.

8. A fuel additive as claimed in claim 6 wherein said organo-metallic complexes are present at a total concentration of from 40 to 60% by weight.

9. A fuel additive as claimed in any one of claims 6 to 8 for use with a particulate filter trap of the 'cracked wall' type, wherein the total concentration of the metals of the Group I organo-metallic complexes is 100 ppm or less.

10. A fuel additive as claimed in any one of claims 6 to 8 for use with a particulate filter trap of the 'deep bed' type, wherein the total concentration of the metals of the Group I organo-metallic complexes is 50 ppm or less .

11. A fuel additive as claimed in any one of claims 6 to 10, wherein the Group I organo-metallic complexes are complexes of potassium and sodium.

12. A fuel additive as claimed in any one of claims 6 to 11 wherein each organo-metallic complex is fuel soluble.

13. A fuel additive as claimed in any one of claims 6 to 12 comprising two organo-metallic complexes of Group I metals in a ratio in the range of from 20:1 to 1:20.

14. A fuel additive as claimed in claim 13 wherein said ratio is m the range of from 10:1 to 1:10.

15. A fuel additive as claimed in claim 11 wherein the ratio by weight of sodium to potassium is in the range of from 20 : 1 to 2:1.

16. A fuel additive as claimed in claim 11 wherein the ratio by weight of sodium to potassium is in the range of from 10:1 to 3:1.

17. A fuel additive as claimed in any one of claims 6 to 16 wherein each of the organo-metallic complexes is of the formula M(R)_m.nL where M is the cation of an alkali metal of valency m; R is the residue of an organic compound RH, where R is an organic group containing an active hydrogen atom H replaceable by the metal M and attached to an 0, S, P, N or C atom in the group R; n is a positive number indicating the number of donor ligand molecules forming a bond with the metal cation, but which can be zero; and L is a species or functional group capable of acting as a Lewis base.

18. A fuel additive as claimed in claim 17 wherein R and L, for at least one of the complexes, are in the same molecule.

19. A fuel additive as claimed in claim 17 or claim 18 wherein the organo-metallic complex comprises an alkali metal salt of an alkyl or alkenyl-substituted succinic acid.