Metal or metal containing additives have been incorporated in fuel compositions for many years. The additives may provide a number of effects on the fuel. Certain additives are known to improve the combustion properties of the fuel, for example certain additives may increase the octane number of petroleum fuels. The additives may also provide an effect during combustion, in particular during combustion in an internal combustion engine. For example metal or metal containing additives may deposit metal or metal compounds on surface of an internal combustion engine during combustion. In particular metal or metal compounds may deposit on the valves or valve seats of an internal combustion engine. Such deposits may protect these components of the engine from wear caused during operation, for example the deposits my protect the valve seats from wear and consequential recession.

There is a considerable history of technical papers over many years that teach as to the causes and the means of prevention of VSR.

In a paper published in the Transactions for 1930 and 1931 of the Society of Automotive Engineers, Inc. of the United States of America, A T Colwell describes the problems of operating engines with cast iron exhaust valve seats under high load conditions. These were frequently encountered by gasoline engine trucks and motor coaches on long distance highways. The operating problems encountered centre on the formation of extremely hard warts or nodules on the surface of the exhaust valve face, where it contacts the valve seat. The presence of these hard nodules leads to rapid wear or abrasion of the seat, particularly at high exhaust valve temperatures, as experienced under conditions of sustained high speed cruising.

Colwell describes the entire phenomenon of VSR with accuracy and with great insight into the probable mechanism for the wear process, and its possible solution. He states "There are several remedies for this condition, (referring to VSR)..... In many cases the

use of ethyl gasoline (i. e. containing tetra ethyl lead) alone will stop the trouble. This is probably because the products of combustion of ethyl gasoline form a thin coating on the valve seat that acts as an insulator between valve and block."For many decades, the use of gasoline containing tetra ethyl lead provided almost complete protection from VSR. However, the phasing out of tetra ethyl lead from gasoline has resulted in a search for alternative fuel additives, which can provide protection from VSR in cast iron engines.

Many researchers since have highlighted the role played by metallic fuel additives in providing protection from VSR in gasoline engines. Barker proposed an explanation for protection by lead additives in his paper C291/73 presented at the I. Mech E tribology conference in London in 1973. The explanation is the simiiar to that proposed in the Colwell paper. namely the formation of a thin film between the exhaust valve and its seat. Metal salts, typically oxides are considered to form and to provide a high melting point solid lubricant preventing metal to metal contact.

In his 1973 conference paper, Barker gives indications of the effect of various metallic fuel additives in preventing VSR in gasoline engines. The metals considered include lead, zinc, iron, sodium and vanadium. Lead at a treat rate of 13.0 mgPb/l was very effective in preventing VSR. followed by zinc, vanadium, sodium and iron. All these latter metals were markedly less effective than lead despite being added at a metal treat rate of 18.5mgM/I, where M denotes the metal tested. Examinations of wear debris showed that oxides of iron were present on the valve seat. These abrasive materials were implicated in the wear process itself, suggesting that the presence of iron in the gasoline would not necessarily be conducive to protection against VSR. The relatively poor performance of iron as an additive to protect against VSR is consistent with this view.

A later paper on the subject of VSR additives by McArragher et al., presented at a Co- ordinating European Council conference in Birmingham in 1993, covered the use of a range of chemistries including phosphorus and alkali metals. Phosphorus compounds of various types were shown to offer a significant level of protection from VSR.

Phosphorus compounds are also mentioned in the paper as being beneficial in preventing spark plug fouling. In addition the paper showed that potassium provides limited but acceptable levels of valve seat protection in gasoline engines at a treat rate

of about 10mgK/kg. It is acknowledged in the McArragher paper that phosphorus provides a superior level of protection from valve seat recession compared to potassium. Nevertheiess, potassium has subsequently been used in many European countries to provide protection from VSR in commercial retailed fuel intended for vehicles previously fuelled with leaded gasoline. The performance of this metallic additive is well known to those skilled in the art. Similarly, its limitations as a VSR protection additive are well known to those in the Industry.

Ferrocene is a well-known metallic fuel additive with a significant capability to increase octane quality in unleaded gasoline. It is used as. an octane trimming additive at refineries to enhance octane quality in gasoline, to assist meeting gasoline octane specifications. The performance of this product as an additive to protect against VSR was explored by Barker as discussed above, and found to be relatively poor at a treat rate of 18. 5mgFe/litre, which equates to 25mgFelkg. For octane enhancement purposes, iron added as ferrocene is used typically at a treat rate of 9mgFe/kg. This treat rate of additive would be expected to provide very limited protection from VSR.

This can in fact be shown to be the case.

The mechanism for VSR protection from the phosphorus additive Valvemaster lies in the formation of P, 05 in the engine. Deposits are laid down between the exhaust valve and its seat, preventing the metal to metal contact which leads to erosion or recession of the valve seat. The deposition of such protective deposits was postulated as described earlier by Colwell in 1931 and by Barker in 1973. The products of combustion of PLUTOcent) are iron oxides, which are slightly abrasive materials not expected to provide VSR protection.

The present applicants have identified a composition which provides improvement of combustion properties of a fuel and/or prevention/inhibition of valve seat recession (VSR) In a first aspect there is provided use of a composition for the prevention and/or inhibition of valve seat recession of an internal combustion engine, the composition comprising (i) (a) phosphorus and/or a phosphorus compound; and/or (b) potassium and/or a potassium compound; and (ii) iron and/or an iron compound.

In a second aspect there is provided a fuel additive composition comprising (i) (a) phosphorus and/or a phosphorus compound ; and/or (b) potassium and/or a potassium compound: and (ii) iron and/or an iron compound.

The composition is present in amount to provide the required improvement of the combustion properties and prevention of valve seat recession.

VSR is an abbreviation of valve seat recession. In this context it generally means valve seat recession of an internal combustion engine, such as a petrol/gasoline internal combustion engine.

The mechanism for VSR protection from a phosphorus additive, for example VatvemasterTM is believed to lie in the formation of P205 in the engine. Deposits are laid down between the exhaust valve and its seat, preventing metal to metal contact, which leads to valve seat recession.

The combustion product of iron and/or iron compounds such as Plutocen is iron oxide, which is a slightly abrasive material not expected to provide VSR protection. However we have found that iron and/or iron compounds such as Plutocene, which is an octane enhancing additive, when added to phosphorus products (e. g. Valvemaster) provide a combined VSR and octane boosting additive formulation, and has demonstrated a clear unexpected additional VSR benefit. Without being bound by theory the probable explanation for this unexpected benefit lies in the formation of reaction products of iron and phosphorus which provide enhanced VSR protection. Typically this will be of benefit in formulations of Valvemaster and Plutocen, where Valvemaster additions as low as 10mg/kg in combination with Plutocen additions ranging from 30mg/kg to 60mg/kg would allow simultaneous octane enhancement and VSR protection of a high level. Normally, 30mg/kg of phosphorus (600mg/kg Valvemaster TM) are added to provide a high level of VSR protection regarded as satisfactory for vulnerable cast iron engines. This level is supported by significant field experience.

In addition to the considerable advantage of improved VSR protection unexpectedly obtained from the combination of iron and phosphorus, certain additional advantages can be seen.

The use of a volatile iron-containing additive may contribute to increased driveability, in both the short and long term. Initially, an octane improvement results from the use of PLUTOcenO. Iron compounds are known to reduce in-cylinder deposits through oxidation of carbonaceous material, again, a volatile species would be expected to provide such benefits throughout the cylinder as opposed to only those parts wetted by liquid fuel. Thin films of iron containing material deposited on the cylinder walls are suspected as providing fuel consumption and emissions reduction benefits, especially for CO. NOx and unburned hydrocarbons. The combination iron-phosphorus additive reduces the risk of spark plug fouling. The role of phosphorus containing compounds in reducing spark plug fouling is well known to those skilled in the art and is described in the 1993 paper by McArragher et al. The combination of reduced combustion deposits, hence reduced ORI, improved combustion, and increased octane number of the fuel results in an improved driveability.

In addition or in place of the phosphorus material, the iron and/or iron compound may be combined with potassium and/or a potassium compound and unexpected advantages observed. The VSR prevention performance of potassium at a metal treat rate of 8mgK/kg is well established as moderate. However, we have found that when combined with iron in the form of ferrocene, whose VSR prevention performance is poor. as referred to above, the combination of the two metals provides a level of protection from VSR which is surprising, and unexpected. In addition, the combination of potassium with ferrocene increases the octane quality of the blend to which the combination is added.

In addition to the considerable advantage of improved VSR protection unexpectedly obtained from the combination of iron and potassium, certain additional advantages can be seen.

The lower treat rates so available allow adequate VSR protection at metal treat rates less likely to give any problems with regard to issues such as spark-plug fouling or the general growth of in-cylinder deposits. Further, any alkali-metal induced corrosion risks are minimised.

A misplaced concern regarding the effects of VSR additives on 3-way catalysts or the lambda sensors used in their control is frequently expressed. Any vehicle equipped

with a 3-way catalyst will be designed to run on unleaded fuel, and therefore be equipped with hardened valve seats. It will not require a VSR additive in the fuel and dispenser nozzles are designed so as to prevent misfueling, which should thus only be capable of occurring where an aftermarket additive is inappropriately used. Where the additive comprises a combination of iron and phosphorus, VSR protection and various other benefits can be obtained with a further reduction even in this small risk, because of the reduced phosphorus content of the combination.

The combination additive (s) are believed to function by deposition of high temperature lubricant thin films on and around the valve face and seat. Without being bound by theory it is believed that the mechanism (s) by which the combination additives are successful is/are: * catalytic oxidation of carbonaceous material prevents deposit growth, reactions leading to the deposition of carbonaceous material are suppressed, * otherwise harmful deposits are rendered soft and friable.

PHOSPHORUS Preferably the phosphorus and/or phosphorus compound is an phosphorus compound.

Preferably the phosphorus and/or the phosphorus compound is an amine salt of a phosphorus based acid.

All references in the present specification to Valvemaster may be read to mean an amine salt of a phosphorus based acid and/or the reaction product of the following reaction C13 alcohol + P2Os < organic acid organic acid + amine amine salt of a phosphorus based acid and/or a product described in US-A-4720288.

Thus in a preferred aspect the phosphorus based acid is obtainable or obtained from the reaction of (i) the reaction of a C13 alcohol and P205 to form an organic acid and (ii) the reaction of the organic acid and an amine.

POTTASIUM Preferably the potassium and/or potassium compound is an potassium compound.

A very extensive range of compounds have been claimed to be suitable as a means to provide alkali metals, in particular potassium, in a suitable gasoline-soluble form for use as VSR additives.

Potassium salts used may be acidic. neutral or basic (that is over-based, hyperbased or superbased).

Acidic salts may be prepared with an excess of organic acid over potassium, neutral salts react essentially stoichiometric quantities of acid and base and basic salts contain an excess of cations, and are typically prepared by'blowing'a suspension of metal base in a solution of organic acid with gaseous C02.

In aspects of the present invention colloidal suspensions of inorganic salts of potassium may be used.

Suitable organic acids for use in preparing the potassium compound are extensively reviewed in W087/01126 to Johnston et al. These include sulphur acids. carboxylic acids and phosphorus acids.

Some workers have expressed fears that catalyst poisoning may limit the usefulness of the phosphorus acids.

In one aspect the potassium compound is prepared from a sulphur acid.

Sulphur acids include sulphonic, sulphamic, thiosulphonic, sulphenic, sulphinic, partial ester sulphuric, sulphurous and thiosulphuric acids. The sulphur acids may be aliphatic or aromatic, including mono-or poly-nuclear aromatic acids or cycloaliphatic compounds. Sulphonates from detergent manufacture by-products are frequently encountered.

Carboxylic acids include aliphatic, cycloaliphatic and aromatic mono-and poly-basic

carboxylic acids, naphthenic, alkyl or alkenyl cyclopentanoic and hexanoic acids and the corresponding aromatic acids. Branched chain carboxylic acids, including 2- ethylhexanoic acid and propylene tetramer substituted maleic acids may be used.

Carboxylic acid fractions featuring various, mixed hydrocarbon chains, such as tall oils and rosins are also encountered.

Salts of phenols (generally referred to as phenates) may be used. These are of the general formula: (R*) a- (Ar*)- (OH) m where R* is an aliphatic group of 4 to 400 C atoms, a is an integer of 1-4. Ar* is a polyvalent aromatic hydrocarbon nucleus of up to about 14 C atoms and m is an integer from 1-4, provided that there are at least bout 8 C atoms per acid equivalent provided by the R* groups. The R groups may be substituted provided that this does not alter the essentially hydrocarbon character of the groups.

Phosphorus acids may also be used, for example the phosphonic and thiophosphonic acids prepared by reaction of P2S5 with petroleum fractions such as bright stock or with polymeric materials prepared from C2 to C5 mono-olefins, such as poly- (butenes).

Appropriate technology for preparation of a range of phosphorus additives is referenced in WO 87/01126.

EP 207.560 and EP 555,006 describe ranges of succinic acid derivatives. substituted on at least one of the alpha carbon atoms with a C20 to C200 hydrocarbyl group. optionally connected to the other alpha-carbon atom by a hydrocarbon moiety of from 1 to 6 carbon atoms. Such derivatives may be further derivatised by reaction of one carboxyl group with an alcohol or an amine preparing, respectively, the hemi-ester or the amide.

Preferred acid salts are those of potassium with the succinic acid derivatives, as described immediately above, or of alkyl benzene sulphonic acids, especially dodecyl benzene sulphonic acid, from detergent manufacture.

Preferred neutral salts are over basic salts. Salts which are resistant to extraction into aqueous phases are preferred.

The alternative of providing a fuel-stable colloidal suspension of a metal salt having a mean particle size of 1 micron, preferably 0.5 micron or less is illustrated in US-A- 5 090.966 to Crawford et al. An emulsion of a solution of a suitable metal salt, whether potassium borate, carbonate, bicarbonate or acetate is prepared, optionally using an emulsifying agent is prepared in some carrier oil. The solvent is then removed, typically by heating whilst subjecting to rapid agitation. Preferred in-situ preparations of metal borate products, preferred carrier oils and preferred emulsifying agents are set out in the Patent. Such colloidal suspensions are also preferred sources of potassium for use according to the invention.

Mixtures of any or all of the above-mentioned acids may be employed in order to provide a fuel-soluble and stable source of potassium ions. Potassium ions may be employed as a mixture of solution and colloidal suspension sources.

LIRON Preferably the iron and/or iron compound is an iron compound.

Preferably the iron compound is a ferrocene and/or a substituted ferrocene.

In a highly preferred embodiment the composition of the present invention comprises (i) an amine salt of a phosphorus based acid ; and (ii) a ferrocene and/or a substituted ferrocene.

Preferably the iron compound is an iron complex selected from dicyclopentadienyl and substituted-dicyclopentadienyl, The iron compound may be an iron complex of dicyclopentadienyl or substituted- dicyclopentadienyl, wherein the substituents can be, for example, one or more C s alkyl groups, preferably C1 2 alkyl groups. A combination of such iron complexes may also be used.

Suitable alkyl-substituted-dicyclopentadienyl iron complexes are cyclopentadienyl- (methylcyclopentadienyl) iron, cyclopentadienyl (ethyl-cyclopentadienyl) iron, bis- (methylcyclopentadienyl) iron bis- (ethylcyclopentadienyl) iron, bis- (1, 2-dimethyl-

cyclopentadienyl) iron. iron pentacarbonyl, and bis- (1-methyl-3-ethylcyclo-pentadienyl) iron. These iron complexes can be prepared by the processes taught in US-A-2680756, US-A-2804468, GB-A-0733129 and GB-A-0763550.

Suitable iron complexes are dicyclopentadienyl iron and/or bis- (methylcyclo-pentadienyl) iron.

A highly preferred iron complex is ferrocene (i. e. dicyclopentadienyi iron).

The co-ordination chemistry relevant to the solubilisation of transition metals, including iron, in hydrocarbon solvents, e. g. diesel fuel is well known to those skilled in the art (see e. g. WO-A-87/01720 and WO-A-92/20762).

A wide range of so-called"substituted ferrocenes"are known and may be used in the present invention (see e. g. Comprehensive Organic Chemistry, Eds. Wilkinson et al., Pergamon 1982, Vol. 4: 475-494 and Vol. 8: 1014-1043). Substituted ferrocenes for use in the invention include those in which substitution may be on either or both of the cyclopentadienyl groups. Suitable substituents include, for example, one or more Ci. s alkyl groups, preferably C12 alkyl groups.

Particularly suitable alkyl-substituted-dicyclopentadienyl iron complexes (substituted ferrocenes) include cyclopentadienyi (methylcyclopentadienyl) iron. bis- (methylcyclopentadienyl) iron. bis- (ethylcyclopentadienyl) iron. bis- (1.2- dimethylcyclopentadienyl) iron and 2,2'-diethylferrocenyi-propane.

Other suitable substituents that may be present on the cyclopentadienyl rings include cycloalkyl groups such as cyclopentyl, aryl groups such as tolylphenyl, and acetyl groups, such as present in diacetyl ferrocene. A particularly useful substituent is the hydroxyisopropyl group, resulting in (-hydroxyisopropyl) ferrocene. As disclosed in WO-A-94/09091, (-hydroxyisopropyl) ferrocene is a room temperature liquid.

Other organometallic complexes of iron may also be used in the invention, to the extent that these are fuel soluble and stable. Such complexes include, for example, iron pentacarbonyl, di-iron nonacarbonyl, (1,3-butadiene)-iron tricarbonyl, (cyclopentadienyl)-iron dicarbonyl dimer and the diisobutylene complex of iron

pentacarbonyl. Salts such as di-tetralin iron tetraphenylborate (Fe (C, oH, 2) 2 (B (C6Hs) 4) 2) may also be employed.

As a result of a combination of their solubility, stability, high iron content and, above all, volatility, the substituted ferrocenes are particularly preferred iron compounds for use in the invention. Ferrocene itself is an especially preferred iron compound on this basis. Ferrocene of suitable purity is sold in a range of useful forms as PLUTOcen@ and as solutions. Satacen0. both by Octel Deutschland GmbH.

The iron compounds for use in the invention need not feature iron-carbon bonds in order to be fuel soluble and stable. Salts may be used; these may be neutral or overbased. Thus. for example, overbased soaps including iron stearate. iron oleate and iron naphthenate may be used. Methods for the preparation of metal soaps are described in The Kirk-Othmer Encyclopedia of Chemical Technology, 4th Ed, Vol.

8 : 432-445. John Wiley & Sons, 1993. Suitable stoichiometric. or neutral, iron carboxylates for use in the invention include the so-called'drier-iron'species. such as iron tris (2-ethylhexanoate) [19583-54-1].

Iron complexes not featuring metal-carbon bonds and not prepared using carbonation may also be used in the invention provided these are adequately fuel soluble and stable. Examples include complexes with-diketonates, such as tetramethylheptanedionate.

Iron complexes of the following chelating ligands are aiso suitable for use in the invention: aromatic Mannich bases such as those prepared by reaction of an amine with an aldehyde or ketone followed by nucleophilic attack on an active hydrogen containing compound, e. g. the product of the reaction of two equivalents of (tetrapropenyl) phenoi. two of formaldehyde and one of ethylenediamine, hydroxyaromatic oximes, such as (polyisobutenyl)-salicylaldoxime. These may be prepared by reaction of (polyisobutenyl) phenol, formaldehyde and hydroxylamine; Schiff bases such as those prepared by condensation reactions between aldehydes or ketones (e. g. (6tbutyl)-salicylaldehyde) and amines (e. g. dodecylamine). A tetradentate ligand may be prepared using ethylenediamine (half equivalent) in place of dodecylamine;

substituted phenols, such as 2-substituted-8-quinolinols, for example 2-dodecenyl- 8-quinolinol or 2-N-dodecenylamino-methylphenol; substituted phenols, such as those wherein the substituent is NRz or SR in which R is a long chain (e. g. 20-30 C atoms) hydrocarbyl group. In the case of both a-and 3- substituted phenols, the aromatic rings may beneficially be further substituted with hydrocarbyl groups, e. g. lower alkyl groups; carboxylic acid esters, in particular succinic acid esters such as those prepared by reaction of an anhydride (e. g. dodecenyl succinic anhydride) with a single equivalent of an alcohol (e. g. triethylene glycol) ; acylated amines. These may be prepared by a variety of methods well known to those skilled in the art. However, particularly useful chelates are those prepared by reaction of alkenyl substituted succinates, such as dodecenyl succinic anhydride, with an amine, such as N, N'-dimethyl ethylene diamine or methyl-2-methyiamino-benzoate; amino-acids, for example those prepared by reaction of an amine. such as dodecylamine, with an a, 3-unsaturated ester, such as methylmethacrylate. In cases where a primary amine is used, this may be subsequently acylated, such as with oleic acid or oleyl chloride; hydroxamic acids, such as that prepared from the reaction of hydroxylamine with oleic acid, * linked phenols. such as those prepared from condensation of alkylated phenols with formaldehyde. Where a 2 : 1 phenol: formaldehyde ratio is used the linking group is CH2.

Where a 1 : 1 ratio is employed, the linking group is CH2OCH2 ; alkylated, substituted pyridines, such as 2-carboxy-4-dodecylpyridine ; * borated acylated amines. These may be prepared by reaction of a succinic acylating agent, such as poly (isobutylene) succinic acid, with an amine. such as tetraethylenepentamine. This procedure is then followed by boronation with a boron oxide, boron halide or boronic acid, amide or ester. Similar reactions with phosphorus acids result in the formation of phosphorus-containing acylated amines, also suitable for providing an oil-soluble iron chelate for use in the invention; 'pyrrole derivatives in which an alkylated pyrrole is substituted at the 2-position by OH, NH2, . NHR, CO2H, SH or C (O) H. Particularly suitable pyrrole derivatives include 2- carboxy-t-butylpyrroles; sulphonic acids, such as those of the formula R'S03H, where R'is a Clo to about Ceo hydrocarbyl group, e. g. dodecylbenzene sulphonic acid;

organometallic complexes of iron, such as ferrocene, substituted ferrocenes, iron naphthenate, iron succinates, stoichiometric or over-based iron soaps (carboxylate or sulphonate), iron picrate, iron carboxylate and iron-diketonate complexes.

Suitable iron picrates for use in the invention include those described in US-A- 4.370,147 and US-A-4,265,639.

Other iron-containing compounds for use in the invention include those of the formula M (R) x. nL wherein: M is an iron cation; R is the residue of an organic compound RH in which 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; x is 2 or 3; n is 0 or a positive integer indicating the number of donor ligand molecules forming a dative bond with the metal cation: and L is a species capable of acting as a Lewis base.

FUEL In a third aspect there is provided a fuel composition comprising (i) (a) phosphorus and/or a phosphorus compound; and/or (b) potassium and/or a potassium compound; and (ii) iron and/or an iron compound; and (iii) a fuel.

In the context of VSR the term'fuel'covers compositions containing a major amount of gasoline base fuel suitable for use in spark-ignition engines. This includes hydrocarbon base fuels boiling in the so-called gasoline boiling range of 30 to 230°C. These base fuels may comprise mixtures of saturated, olefinic and aromatic hydrocarbons. They can be derived from straight-run gasoline, synthetically produced aromatic hydrocarbon mixtures, thermally or catalytically cracked hydrocarbon feedstocks, hydrocracked petroleum fractions or catalytically reformed hydrocarbons. Motor gasolines are defined by ASTM D-439-73, aviation gasolines typically have a narrower boiling range of 37 to 165°C. The gasoline may also contain various blending components designed to provide octane number, such as MTBE, TAME or ETBE as non-limiting examples. A proportion of the hydrocarbons may also be replaced for example by alcools, ethers (as above), esters or ketones. Generally the octane number of the gasoline will be greater than 65.

In a preferred aspect the phosphorus compound provides elemental phosphorus in an

amount of at least 10 mg per kg of fuel. More preferably the phosphorus compound provides elemental phosphorus in an amount of from 10 to 40 mg per kg of fuel, more preferably in an amount of from 10 to 20 mg per kg of fuel or in an amount of from 25 to 35 mg per kg of fuel.

In a preferred aspect the iron compound provides elemental iron in an amount of at least 5 mg per kg of fuel. More preferably the iron compound provides elemental iron in an amount of at least 30 mg per kg of fuel or in an amount of from 7 to 10 mg per kg of fuel.

Preferably the fuel is gasoline.

The fuei may further comprise performance-enhancing additives. A non-limiting list would include corrosion inhibitors, rust inhibitors, gum inhibitors, anti-oxidants. solvent oils, anti-static agents, dyes. anti-icing agents, ashless dispersants and detergents.

The fuel additives according to the invention may be added as part of a package to the fuel prior to combustion. This may be done at any stage in the fuel supply chain (for example, at the refinery or distribution terminal) or may be added via a dosing device on-board the vehicle, either to the fuel or even separately direct into the combustion chamber or inlet system. The fuel additives may be added to the fuel in the vehicle fuel tank by the user. a so-called'aftermarket'treatment.

The invention further comprises an additive solution for addition to a fuel. Such an additive might be dosed at any stage in the fuel supply chain prior to combustion of the fuel. The fuel additives of the invention may be dosed to the fuel at any stage in the fuel supply chain. Preferably, each additive is added to the fuel close to the engine or combustion systems, within the fuel storage system for the engine at the refinery, distribution terminal or at any other stage in the fuel supply chain, including aftermarket use.

How an additive solution is to be employed significantly influences the optimum formulation. For example, the additive may be added to the fuel at the refinery or at the distribution terminal. Here the iron and phosphorus and/or potassium components may be added together or separately, providing an additional valuable flexibility in use. If

added together, they will be dissolved in the minimum amount of fuel compatible solvent commensurate with the need to provide a pumpable solution and avoid crystallisation/separation of any of the components at low temperatures, e. g. about- 30°C.

Where the advantages of separate addition are desired, the iron material such as PLUTOcent) is added at the refinery as a blending component for octane trimming, to meet the required product octane specification, thus fulfilling the well known and valuable role to the refiner of an octane enhancing agent. The phosphorus or potassium component such as ValvemasterT can be added to the finished fuel at the distribution terminal. in order to produce a product known to those in the Industry as a 'lead replacement gasoline" (LRG) or'lead replacement petrol" (LRP). Addition of a level of ValvemasterT" from 100mg/kg up to 600mg/kg, equivalent to 5mg/kg up to 30mg/kg phosphorus will provide VSR protection to enable the product to be retailed as LRG or LRP. However the significant advantage provided by the combination of the present composition (for example by PLUTOcen (D and Valvemaster TM) is that the level of protection provided by the combination can be adjusted to suit market requirements independently of the octane enhancement which the combination also provides. Also the very high level of VSR protection provided by the combination allows for the minimum inclusion of a phosphorus additive (e. g. ValvemasterTM) while still providing a very high level of VSR protection.

Where, however. the additive combination is intended to be added as an aftermarket' treatment, the volume of solvent used will be such as to provide a non-viscous solution, suitable for use in a dispenser bottle or syringe pack. The concentration of iron and phosphorus and/or potassium will be such that some convenient and easily recalled treat rate (e. g. about 1 cm3 per litre of fuel) is required. In any case the solvents to be used should be readily fuel soluble and compatible, including with respect to boiling point range, and preferably will have flash points in excess of 62°C for ease of storage.

The additive solution may optionally contain additional components beyond the iron and phosphorus and/or potassium compounds. These components include corrosion inhibitors, rust inhibitors, gum inhibitors, anti-oxidants, solvent oils, anti-static agents, dyes, anti-icing agents, ashless dispersants and detergents as a non-limiting list.

Where any additional component is employed, the use of detergents, especially poly-

(butenyl) succinimide based detergents, is preferred.

Further aspects of the present invention include * use of a composition for the prevention and/or inhibition of valve seat recession of an internal combustion engine, the composition comprising (i) (a) phosphorus and/or a phosphorus compound ; and/or (b) potassium and/or a potassium compound; and (ii) iron and/or an iron compound; with the proviso that when the phosphorus compound is an amine salt of a phosphorus based acid, the iron and/or iron compound is other than a ferrocene or a substituted ferrocene. a fuel additive composition comprising (i) (a) phosphorus and/or a phosphorus compound; and/or (b) potassium and/or a potassium compound; and (ii) iron and/or an iron compound; with the proviso that when the phosphorus compound is an amine salt of a phosphorus based acid. the iron and/or iron compound is other than a ferrocene or a substituted ferrocene. a fuel composition comprising (i) (a) phosphorus and/or a phosphorus compound; and/or (b) potassium and/or a potassium compound; and (ii) iron and/or an iron compound; and (iii) a fuel; with the proviso that when the phosphorus compound is an amine salt of a phosphorus based acid, the iron and/or iron compound is other than a ferrocene or a substituted ferrocene.

'use of a composition for the prevention and/or inhibition of valve seat recession of an internal combustion engine, the composition comprising (i) potassium and/or a potassium compound; and (ii) iron and/or an iron compound.

* a fuel additive composition comprising (i) potassium and/or a potassium compound; and (ii) iron and/or an iron compound.

* a fuel composition comprising (i) potassium and/or a potassium compound: (ii) iron and/or an iron compound; and (iii) a fuel.

The invention will now be further described in further detail by way. of example only with reference to the accompanying drawing in which: Figure 1 is a diagram showing measurement of valve stand-down height

EXAMPLES Example 1-Iron and Phosphorus It will be shown that the combination of PLUTOcen and ValvemasterT has demonstrated a clear and unexpected benefit from enhanced protection from VSR, in engine tests. The tests were conducted by the Motor Industry Research Association (MIRA), under the auspices of the Federation of British Historic Vehicle Clubs (FBHVC). The tests were performed in accordance with documented method No.

FBHVC 98/01. Details of this method are given in Appendix I.

The test protocol utilised a 1.3 litre 4 cylinder engine having a cast iron cylinder head without valve seat inserts. The engine was operated for a total of 70 hours comprising 50 hours at 3,800 rev/min and 23 kW load, and 20 hours at 5,500 rev/min and 42 kW load. In practice, this condition constituted wide open throttle (WOT) operation. Prior to the start of the two test stages, the engine was operated for a"shakedown"period of approximately one hour using unleaded petrol. This process was carried out for consistency and to allow the engine to bed in after the refitting of the cylinder head. Head removal and refitting was necessary after the completion of each separate test run. Valve clearances were checked after every ten hours during the first 50 hour operational period, and every five hours during the second 20 hour period of operation.

Tests were carried out with phosphorus alone and with phosphorus combined with iron, according to the FBHVC test procedure. Results form these engine tests are given below. Test Additive active Valve seat recession over 70h, mm content Mean t Worstvalve 30mg/kg P 0. 35 0. 2 37mg/kg P 0.13 1 0.27 ! 3 8mg/kg P + 0. 21 0.52 ! 9mg/kgFe 4 18mg/kg P + 0.14 0.28 9mg/kg Fe 5 30mg/kg P + 0.05 0.06 9mg/kg Fe These data indicate that the protection against VSR provided by the additives tested, increases as the treat rate increases, which is consistent with published literature. The data also indicated that a significantly and unexpectedly higher level of protection is

provided by the combination of phosphorus and iron, than is provided by phosphorus alone.

This is clearly demonstrated initial by the exceptional protection provided by the combination of 30mg/kg phosphorus combined with 9mg/kg of iron, as shown by test 5.

Furthermore, the protection provided by 18mg/kg of phosphorus combined with 9mg/kg of iron is almost equivalent to that provided by 37mg/kg of phosphorus alone. Finally, the combination of 8 mg/kg of phosphorus combined with 9 mg/kg of iron provides better protection than that provided by 30mg/kg of phosphorus alone.

Example 2-Iron and Potassium Vehicle tests are carried out using a Rover"A"series engine with a cast iron cylinder head. Cast iron cylinder heads are noted for their susceptibility to valve seat recession. Before each test is commenced, the cylinder head is refurbished. and valve seats re-cut, to ensure that no trace of lead deposits can influence the findings. The rebuild is to standard specification. The car is then operated on a chassis dynamometer at speeds of 50-70 km/h for 1,000 km to allow the exhaust valves to bed in. Experience of previous testing is that at these speeds, little or no valve seat recession is observed (see M W Vincent and T J Russell"A Review of World-wide Approaches to the Use of Additives to Prevent Exhaust Valve Seat Recession"4th Annual Fuels and Lubes Asia Conference, January 14-16,1998). Details of the engine are as below : Capacity, cc _275 No of cylinders 4 Valve operation OHV Bore,mm 70.6 Stroke.mm 81.3 CR:1 9.75 Power, kW @ m 51 5800 Tor ue. Nm r m 104 3500 Fuelsystem Carburettor Type SU HIF 44 The test car is operated according to the cycle shown below. Valve stem to rocker-pad clearances are checked every 4 hours during the actual test. Overall test duration is 100 cycles, but tests are terminated early when significant valve recession is observed.

Overall wear and hourly wear rates for the additised fuels are compared to those from gasoline containing 0.03 to 0.15 9/l of lead as tetra-ethyl lead. Tests using non- additised unleaded gasoline are of somewhat short duration. Time. min Speed, km/h rpm Cumulative distance, km 30006.67580 375040.020100 10 120 4500 60. 0 300073.331080 1 20 100 3750 106.

Overall duration 65 minutes.

Overall average speed 98.5 km/h Whilst this cycle is realistic, it is also severe. Tests are also carried out using a modified cycle as shown below. Time. min Speed. km/h rpm Cumulative distance. km 22505.0560 10 80 3000 11.67 15 100 3750 36.67 590337544.17 375015100 69.17

Overall duration 50 minutes Overall average speed 83 km/h The results of the tests are summarised in the table below:

Fuel! Summary Test Result 1 (Cycie 1 and 2 combined) Base Gasoline ! Tests of limited duration. VSR unacceptabte Base Gasoline plus 0.03 to 0.15 g/l of lead as Full 100 hours reached. excellent VSR tetra-ethyl lead protection Base gasoline plus 9 ppm m/m Fe as Some limited VSR protection observed ferrocene Base gasoline plus 8 ppm m/m K as Fair VSR protection on less severe cycle, commercially available product modest protection over severe test I Base gasoline plus 9 ppm m/m Fe as Good VSR protection on both cycles. ferrocene and 8 ppm m/m K as commercially available product In each case, the performance of the combination of additives is superior to that which would be expected by comparing the performance of the individual components at or around the dose rates used. That is, indications of a synergistic effect are observed.

Example 3-Iron and Potassium Testing was completed with a Rover'A'Series engine as previously described. The test cycle was Time. min Speed, km/h rpm 5 60 2250 20 80 3000 10 100 3750 5 90 3375 15 100 3750

The following fuel compositions were tested Base gasoline plus 9 ppm Fe as ferrocene Base gasoline plus 8 ppm K as commercially available product Base gasoline plus 9 ppm Fe as ferrocene and 8 ppm K as commercially available product Each run was followed by replacement of valve seat inserts in the cylinder head with cast iron of constant hardness.

The following data were obtained Fue) Compositton: Recession Rate mm/1000km ! Fuel + Worst Value Mean Value 19 ppm Fe 0. 105 0.059 1 8 ppm K 0. 058 0.032 9 ppm Fe & 8 ppm K 0. 044 0.030

The mean and more significantly the critical worst value show substantially less recession with the iron and potassium combination of the present invention. These data demonstrate a synergy when iron and potassium are combined in a VSR inhibiting additive.

All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the

scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments.

Indeed, various modifications of the described modes for carrying out the invention which are obvious to those skilled in chemistry or related fields are intended to be within the scope of the following claims.

APPENDIX I FEDERATION OF BRITISH HISTORIC VEHICLE CLUBS

DOCUMENTED METHOD NO. FBHVC 98/01 MEASUREMENT OF EXHAUST VALVE SEAT RECESSION USING ROVER"A" SERIES ENGINE Originator: Fuels Committee Federation of British Historic Vehicle Clubs Approvedby: IEdmunds Issued by: M. Holt-Chasteauneuf

EXHAUST VALVE SEAT RECESSION TEST PROCEDURE.

FEDERATION OF BRITISH HISTORIC VEHICLE CLUBS 1. SCOPE This document defines a test procedure for evaluating claims made for devices, and fuel additives. to enable a spark ignition engine designed for leaded petrol to operate continuously on unleaded petrol.

2. OBJECTIVE The objective of the test procedure is to quantify and to measure exhaust valve seat recession experienced with any device or fuel additive assessed. From measurements recorded. an assessment of the engine protection provided by candidate devices or fuel additives. and their potential suitability to prevent valve seat recession with continuous use of unleaded petrol, can be made.

3. Definitions 3.1 Engine-a reciprocating spark ignition internal combustion engine 3.2 Engine system-any part of the engine assembly including fuel, induction, ignition, lubrication, cooling, exhaust and management systems.

3.3 Device-any equipment or apparatus applied to the engine system, fuel storage tank or pipework 3.4 Additive-fuel soluble medium added to unleaded petrol in the fuel storage tank or pipework 3.5 Unleaded petrol-fuel containing less than 0.013g Pub/ ! and meeting the specification of EN 228 or BS7070 3.6 Leaded petrol-fuel containing lead alkyl antiknock additives and meeting the BS 4040 specification 3.7 Shall-indicates a mandatory requirement.

4. Test engines The test engine shall have the following specification: TypeRover"A"series Capacity, cc 1275 No of 4 Valve operation OHV Bore, mm 70.6 Stroke,mm 81.3 1 Compression ratio 9. 75 : 1@ Fuel system Carburettor Type SU HIF 44

5. Engine preparation The engine shall be inspected prior to its use in testing to ensure that the cylinder head fitted, has not been modified for operation on unleaded petrol has not been fitted with valve seat inserts has not experienced valve seat recession.

N. B.: Due to the unavailability of new cylinder heads, reconditioned units may be used, with great care exercised in selection and preparation to meet the above requirements.

The engine shall be rebuilt prior to its use in testing with the following new components: inlet and exhaust valves valve stem seals cylinder head and other gaskets as needed All valves shall be ground in to ensure removal of lead deposits from valve seats, from previous operation on leaded petrol.

After reassembly, the engine shall be operated over a range of speed and load conditions to ensure normal operation. Ignition advance and exhaust CO level shall be checked and set to manufacturer's specification. As a final check, a full load power curve shall be carried out.

Valve tip location shall be measured using a jig in combination with a micrometer depth gauge. See Figure 1. The distance"a"is defined as the valve stand down height.

Valve stand down heights shall be measured as follows, and measurements recorded: after engine reassembly and before starting. This is denoted as"initial" condition after engine break in and power check, allowing 30 minutes after shut down for engine cooiing. This is denoted as"post power check"condition.

6. Engine running conditions The valve seat recession test shall be run in two stages, as follows: Stage 1 Operation for 50 hours at 3800 rev/min and 23 kW output Stage 2

Operation for 20 hours at 5500 rev/min and 42 kW output.

The following operating conditions shall be maintained during the test: Coolant outlet temperature 90 2oC Oil gallery temperature 100 2°C Exhaust back pressure 133 mbar at 5500 rev/min 7. Valve Seat Recession Measurement Details of valve tip location measurements prior to the start of test are given in Section 5. The same technique is used to measure valve seat recession at intervals during the test. The valve stand down height, after 30 minutes cooling, shall be measured and recorded at the following intervals: Stage 1: every 10 hours and at the end of 50 hours Stage 2: every 5 hours and at the end of 20 hours After each valve tip location measurement, valve clearances shall be checked and adjusted to manufacturer's specification.

8. Test Fuel The test fuel hall be taken from a batch of unleaded petrol of adequate size to enable all candidate devices or fuel additives to be tested using the same type of fuel. Where a device is to be tested, no other additive shall be added to the fuel unless the additive comprises an integral part of the device. Where a fuel additive is to be tested, it shall be added to the test fuel prior to commencing the test, using the mixing procedure defined in the Appendix A.

9. Test Protocol The test engine shall be dismantled (cylinder head removed) and prepared according to the requirements of Section 5, in preparation for the test on each candidate device or fuel additive.

Crankcase lubricating oil shall be drained and refilled as part of the preparation for each test.

The engine reciprocating parts ("bottom end") shall be inspected at least every 4th test to ensure satisfactory mechanical condition e. g. blow by, piston slap, between tests.

Replacement pistons shall be fitted, and bores honed to maintain the engine in a satisfactory operating condition.

On completion of testing of candidate devices and fuel additives, a test fuel containing 0.03g Pb/l shall be employed for a further test to assess relative valve seat recession performance. This test fuel hall be produced by adding the required amount of lead alkyl additive to the unleaded test fuel employed for the previous tests.

10. Pass-Fail Criteria (i) Ideally there should be no significant recession of the exhaust valves throughout all stages of the test.

(ii) A borderline pass is one where no individual valve shows recession in Stage 1* and no individual valve shows recession of more than 0.25mm or twice the value recorded with leaded petrol, whichever is the greater, during Stage 2.

*NB: The regrinding of valves prior to the start of the test may allow a slight change in valve stand down height, due to"bedding in", near to the start of test. For this reason, a single change in valve stand down height, of up to 0.05mm during Stage l. is permitted within the definition of a borderline pass, provided there is no further valve seat recession during Stage l.

(iii) A fail is one where any valve shows a valve recession greater than (ii).

11. Results 11.1 Cumulative valve recession (mm) Test Valve 1 Valve 2 Valve 3 Valve 4 Valve 5 Valve 6 Valve 7 Valve 8 Hours (ini) (exh) (exh) (inl) (inl) (exh) Stage 1 10 20 30 40 50 Stage 2 55 60 65 70 11.2 Seat face wear after 70 hours (mm) Seat Valve 1 Valve 2 Valve 3 Valve 4 Valve 5 Valve 6 Valve 7 Valve 8 face (exh) (ini) (inl) (exh) (exh) (inl) (inl) (exh) Valve Cylinderhead 12. Operational Summary Test Torque Power Fuel Flow Air in Exh. A Exh. B Ign. Adv. CO Hours run kg/hr°C°C°C°btdc%kW Stage1 10 20 30 30 40 i 50 Stage 2 55 60 65 70

Appendix A Mixing procedure 1. Fuel/additive mixing procedure The following procedure is employed for preparing the fuel for test: a) Using clean and dry 205 litre fuel barrel fill with 200 litres of base fuel b) Take a 1 litre sample of fuel from the drum c) Calculate the required amount of additive to achieve the correct dose. d) Measure out the required volume of additive e) Add the additive to the base fuel. If necessary add some fuel to the additive from the 1 litre sample to assist mixing. Rinse the additive container with fuel to ensure all the additive has been transferred to the drum Agitate the mixture using a pneumatic stirrer for 10 minutes g) Take a 1 litre fuel sample.

The base fuel for the test shall be unleaded petrol meeting the requirements of Sections 3.5 and 8.

2. Fuel analysis results A sample of fuel shall be routinely taken from each of the barrels used for the test. The samples can be sent for analysis if required.