FIELD OF THE INVENTION

The present invention relates to a fuel additive and a liquid fuel that has a high energy content, is safe and environmentally friendly.

BACKGROUND

It has been estimated that the health of many thousands of people are adversely affected by engine exhaust within the European Union and worldwide. To decrease the toxicity of vehicle exhausts, many methods have been investigated.

Even though advances have been done within the field of engines with improved combustion, it is still difficult to avoid emissions of toxic compounds, especially during quick changes of torque or speed. New engines have also been developed, e.g., Homogenous Charge Compression Ignition (HCCI) engines which have been successfully tested. However, the disadvantages of these are that new engine types require new types of fuels, and also that the capital investment for completely new engines in many vehicles will be very high. Catalysts for after-treatment of the exhaust are mainly reducing light hydrocarbons and nitrous oxides. The most toxic compounds, such as heavy polyaromatics cannot be reduced at temperatures much below 1500 ⁰ C, temperatures which are rarely achieved in commercial catalytic systems. Another approach is to hydrotreat conventional fuels, saturating many of the molecules in the fuels with hydrogen. This treatment will reduce the most toxic compounds such as polyaromatics, however, since the basic chemical structure is not affected by hydrogenation, said polyaromatics will to some degree be regenerated during the first stage of combustion, in which de-hydrogenation takes place.

EP 1452579 (EP'579) shows a fuel based on alkylated monocyclic alkanes, non- cyclic alkanes and additives. The fuel according to EP'579 shows a flashpoint above 90 ⁰ C and an energy increase of up to 4MJ/litre compared with a pure iso- and n- paraffinic fuel. The fuel further shows low levels of toxic emissions.

EP0208541 (EP'541) presents a lubricant composition comprising a polycyclic structure wherein the cycles are non-aromatic and may be substituted with alkyl groups. The cycles are further separated with either a methylene, ethylene or trimethylene group. The composition further comprises a thio phosphate, an alkenyl succinimide or its boron derivative and a carboxylic acid of a polyalcohol.

SUMMARY OF THE INVENTION

The object of the present invention is to provide a fuel that is both high in energy content and environmentally friendly. The present invention further relates to a method of producing said fuel and additive and the use of the same.

The first aspect of the present invention is a fuel additive and a fuel additive for use in Homogeneous Charge Compression Ignition engines, fuel cell systems, compression ignition engines, or turbine engines, comprising cyclic organic compounds containing: -at least two ring structures with 4-9 atoms in each ring; -said atoms are selected from hydrogen, carbon, oxygen and nitrogen; and wherein

-each ring is separated from each other by a saturated or unsaturated organic chain of at least three atoms, e.g. propylene.

In one embodiment of the present invention the organic compounds have a molecular weight of more than 100, preferably more than 200, more preferably more than 400 and more preferably more than 600 but less than 800 Dalton.

In one embodiment of the present invention the organic compounds consist of hydrogen and carbon atoms.

In one embodiment of the present invention the ring structures are saturated with hydrogen atoms.

In one embodiment of the present invention the amount of the cyclic organic compound is more than 3%, preferably more than 10%, more preferably more than 25 %, even more preferred more than 50% and most preferred more than 80%.

In one embodiment of the present invention the additive further contains other additives.

In one embodiment of the present invention at least one ring structure comprise a side chain of a saturated or unsaturated hydrocarbon.

In one embodiment of the present invention all the ring structures comprise at least one side chain of a saturated or unsaturated hydrocarbon. In one embodiment of the present invention the additive further comprises less than 25%, preferably less than 15% and more preferred less than 10% of alkenes.

In one embodiment of the present invention the flash point is at least 100 ⁰ C. In one embodiment of the present invention the content of molecules containing ring structures which share at least one atom is less than 1 %.

In another embodiment the additive comprises a mixture of compounds having 2, 3 or 4 rings and where the linker between the rings is a propylene group, without any side chains and wherein the end rings should have alkyl chains of 5-12 carbon atoms in the para position.

Another aspect of the present invention is a liquid fuel for use in Homogeneous Charge Compression Ignition engines, fuel cell systems, compression ignition engines, or jet engines, comprising a fuel additive comprising cyclic organic compounds containing:

-at least two ring structures with 4-9 atoms in each ring; -said atoms are selected from hydrogen, carbon, oxygen and nitrogen; and wherein -each ring is separated from each other by a saturated or unsaturated organic chain of at least three atoms, e.g. propylene. In one embodiment the amount of fuel additive is more than 3%, preferably more than 10%, preferably more than 25%, preferably more than 50%, preferably more than 75% and more preferably more than 90%.

In one embodiment the fuel has a density between 760 and 850 kg/m ³ .

Yet another aspect of the present invention is a method of producing a fuel or a fuel additive comprising the steps of: -adding a mixture C2-C6 alkenes to a reactor;

-adding a catalyst; -polymerizing said alkenes; and -hydrogenating formed polymers. In one embodiment the polymerization is conducted at a temperature of more than 150 ⁰ C, preferably more than 250 ⁰ C , preferably more than 350 ⁰ C but less than 400 ⁰ C.

In one embodiment the hydrogenation is conducted at a temperature of more than 350 ⁰ C, preferably more than 450 ⁰ C, preferably more than 550 ⁰ C but less than 600 ⁰ C.

In one embodiment the polymerization is conducted at a pressure of more than 35, preferably more than 45, preferably more than 55 but less than 65 bar.

In one embodiment the hydrogenation is conducted at a pressure of more than 35, preferably more than 45, preferably more than 55 but less than 65 bar.

In one embodiment the formed polymers are separated via distillation or chromatography or any other suitable technique.

In one embodiment a lubricant is added to the hydrogenated polymers.

Another aspect of the present invention is a liquid fuel or a fuel additive according to any of the preceding claims, wherein the fuel is used in a Homogenous Charge Compression Ignition Engine, a fuel cell system, a compression ignition engine, e.g. a diesel engine, or in a turbine engine.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 shows the structure for the C/O20 peak of sample A.

DETAILED DESCRIPTION OF THE INVENTION

There is a need for new fuel additives and new fuels that are environmentally friendly, can be mixed with other commonly used fuels and are high in energy content. The present invention fulfils all of these criteria.

The present invention involves a fuel additive containing a mixture of organic compounds wherein at least one compound is a polycyclic compound. These polycyclic compounds are not aromatic in order to minimize the amount of toxic products that are emitted during use. The cyclic structures should contain 4-9 atoms, for example carbon, and each cycle is separated with a chain, a linker, of preferably at least 3 atoms, for example carbon atoms. Said linker could be a saturated or unsaturated chain but could also be branched. The organic compounds may contain oxygen or nitrogen as well, but preferably, the compounds should only contain carbon and hydrogen.

In order to obtain the right viscosity of the fuel, the molecular weight of the fuel or the fuel additive should be in the range of 100 to 800 Dalton, and preferably between 200 and 400 Dalton.

The fuel additive according to the present invention is a mixture that may contain, beside the polycyclic compounds described above, non-cyclic alkanes and alkenes. The amount of compounds in the mixture containing compounds that have aromatic structures or ring structures that share at least one atom should be minimized. These compounds may form toxic or environmentally harmful byproducts. Preferably, these compounds should only constitute a few percent of the fuel, preferably below 5% and more preferred below 1%.

According to the present invention, the fuel additive may further contain other additives such as lubricants (e.g. BASF AG, LA99C), anti-corrosion agents, freezing point depressants, conductivity additives, etc. as used in the art.

The present invention does not only provide a fuel additive or a fuel that is less toxic and less harmful to the environment, but the invention has also much higher energy content. The fuel, and the fuel additive, of the present invention may contain up to 20% more energy than standard diesel or jet engine fuel. This means that an aircraft could save an intermediate landing during longer routes. This would save, time, money and the environment. In a preferred embodiment the cyclic structures contain one or more side chains, preferably saturated or unsaturated hydrocarbon groups. These groups could be, e.g. methylene, ethylene, propylene, butylene, or pentylene groups. These side chains may function as an initiation point for combustion and bacterial digestion and makes the additive, or the fuel, at the same time more readily biodegradable and will facilitate complete combustion, thus increasing the efficiency.

One embodiment of the cyclic organic compound is represented by formula I

FORMULA I wherein each Ri is independently a saturated or unsaturated hydrocarbon chain of 3 to 10 carbon atoms, e.g. 3, 4, 5, 6, 7, 8, 9 or 10 carbon atoms; each R ² is independently H or a saturated or unsaturated hydrocarbon of 1 to 12 carbon atoms, e.g. 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 carbon atoms; and x is 0, 1, 2, 3, 4 or 5. R2 may contain an ester, keton or an ether group, but preferably the compound should only contain carbon and hydrogen atoms.

Some preferred embodiments of the cyclic organic compounds are:

FORMULA II and all corresponding isomers.

Another preferred embodiment is an additive comprising a mixture of compounds having 2, 3 or 4 rings and where the linker between the rings is a propylene group, preferably without any side chains. In a preferred embodiment of said embodiment, the end rings should have alkyl chains of 5-12 carbon atoms in the para position. The compounds could contain unsaturated bonds but preferably only saturated bonds. This embodiment would give a fuel additive or a fuel with high energy content, low melting temperature and high density. Some of the said preferred embodiments are shown below:

Hi ₃ C ₆ -/ Vc ₃ H ₆ -/ V-C ₃ H ₆ V N-C ₆ H ₁₃

FORMULA III

The present invention could be used in a Homogeneous Charge Compression Ignition engines, fuel cell systems, compression ignition engines, or jet engines. The present invention may be used either as an additive to a fuel or as a fuel itself.

The present invention provides a fuel and a fuel additive that unlike EP'579 contains cyclic organic compounds comprising more than one ring structure and thereby increasing the energy content of the fuel substantially. In comparison with EP'579 the average molecular weight of the fuel according to the present invention will be higher which means a higher energy density, since the energy-carrying atomic bonds are optimally packed within molecules. Additionally, the flash point of the present invention may be up to 1 15 ⁰ C which makes the fuel even safer to handle.

However, each ring should be separated from each other by preferably at least a propylene group or a derivative thereof. The reasons for this are: a) that the fuel will be more bio-degradable with at least a propylene linking group, since bacterial degradation is more easily achieved in molecules that are not sterically hindered. b) that the fuel will burn more easily without steric hindrance. c) that the viscosity of the fuel will be lower without steric hindrance. d) there is much less risk of formation of toxic compounds when the ring structures are separated by several atoms, which in addition does not introduce a steric structure which in itself may have toxic effects.

This linking part between the cyclic structures makes the present invention differ from EP'541 wherein the linking part is not more than three methylene groups. Additionally, EP'541 refers to a lubricating compound which is something completely different from a fuel additive or a fuel. The lubricating composition of EP'541 further comprises thio phosphates, esters of polyalcohols and succinicimides.

EXPERIMENTS

Sample preparation

Four different fuel samples were prepared in the following way:

1. A mixture of hydrogen gas and carbon monoxide in molar ratio of 1.0 to 1.7 was contacted with a iron Fischer-Tropsch-catalyst, according to patents 1 to 7 in the reference list below, at 230°C. From the products of the Fischer- Tropsch synthesis, C2-C6 olefins (alkenes) were separated using a 13 X zeolite adsorbent. It is known in the art that the main olefin product from the

Fischer-Tropsch synthesis are alpha-olefins (1 -alkenes).

2. The mixture consisting mainly of C2-C6 1 -alkenes was polymerized in a lab reactor of 1.2 meters length and 30 mm inner diameter. The alkenes are fed at lkg/h for approximately 8 hours. A group of thermocouples was arranged along the central axis of the reactor inside a 6 mm outer diameter stainless steel pipe in order to measure the temperature. 0.7 liters of a commercially available zeolite -based polymerization catalyst was used, COD 900 delivered by Sϋd-Chemie AG, Lenbachplatz 6, Munich, Germany. The polymerization was conducted at a reactor maximum temperature of 300°C and a maximum pressure of 50 bars.

3. The polymerized product from step 1 was hydrogenated in a lab reactor using a commercially available hydrogenation catalyst, TK-553 from Haldor-

Topsoe, Nymoellevej 55, Kgs. Lyngby, Denmark. The hydrogenation was conducted at a maximum temperature of 450 ⁰ C and a maximum pressure of 50 bars. 4. Fivehundred ppm by weight of a commercial lubricating additive from BASF

AG, LA99C, was added to the fuel to give good lubrication properties to the fuel.

Said produced oil was denoted sample "A". The oil was separated by standard distillation in three fractions, a sample "B" collected above 300 ⁰ C, a sample "C" collected between 250 and 300 ⁰ C, and a sample "D" collected below 250 ⁰ C.

For comparison, a sample of commercial Swedish Ultra Low Sulphur Diesel (ULSD) was bought at a OKQ8 filling station in Gothenburg, Sweden, on February 1 lth, 2009, denoted sample "E".

Sample characterization

The samples were characterized using Gas Chromatography with Mass Spectroscopy (GCMS), and pyroly sis -field ionisation mass spectroscopy (py-FIMS).

By GCMS, following essentially the method ASTM D2425-93, it was found that the amount of compounds containing 2 or carbon rings sharing at least one atom, i.e. the sum of all di- and polyaromatics and di- and polynaphthenics, was below 1.0% in all samples according to the invention (A to D). The amount of compounds containing 2 or more carbon rings sharing at least one atom was found to be approximately 24.0 wt-% in the ULSD sample E.

By py-FIMS, it was found that the amount of compounds according to the invention containing two, three or four cyclohexane rings in the fuel could be up to 54.0%, using the measurement according to Analysis method. The amount of compounds according to the invention in the ULSD sample was not detectable, i.e. certainly below 1.0%.

Table 1. Analysis results of the fuel samples using GCMS and py-FIMS. N.d. = not detected. Trace = traces seen before filterering of the data, probably in the range of 0.1-0.3 mol-%.

By measurements according to ASTM standards, it was found that the energy content per kg was almost 20% higher for fuels according to the present invention (sample B) compared to the ULSD sample E (see table 2).

A B C D E Method

Calorific Value 46.6 50.2 45.9 42.2 42.3 ASTM D240 Gross [MJ/kg]

Cetane 53 51 53 55 54 ASTM D613 number

Flash point 102 1 15 97 87 73 ASTM D93 [ ⁰ C]

Lubricity 350 290 320 370 390 ISO 12156- HFRR wear 2: 1998 scar [μm]

Density 802 830 801 790 817 ASTM D4052

Viscosity at 3.0 9.5 3.8 1.9 2.0 ISO 3104 40 ⁰ C [mm ² /s]

Cold Flow - 27 -20 <-30 <-30 < -30 ENl 16 Plugging Point (CFPP) [ ⁰ C]

Table 2. Measurements on the fuel samples

Example 1.

The environmental advantages of fuel according to the invention were demonstrated by running a Scania truck engine in a test bench using the driving cycle ECE R49. Measurements of regulated emissions

Regulated emissions were similar for the two fuels, as shown in table 2. The small differences for NOx and soot emissions are not statistically significant. Thus, it can be stated that the differences between the fuels were minor with regard to regulated emissions.

Table 3. Comparison of regulated emissions between fuel according to the invention and ULSD fuel

Measurements of unregulated emissions

Unregulated emissions were analyzed using naphthalene as an indicator substance.

The reason for this choice was that napthalene is the smallest polyaromatic substance, and is usually found in relatively large concentrations in exhausts. It has been found in earlier measurements that measured emissions of naphthalene correlate well with emissions of aromatics, aldehydes, and heavier polyaromatics. It is therefore reasonable to use naphthalene as an indicator of unregulated emissions. The emissions of naphthalene from sample E were 0.27 mg/kWh, whereas the emissions of naphthalene were 0.1 1 mg/kWh when using fuel sample A, i.e. a decrease by approximately 60%. Aromatics, aldehydes, heavier polyaromatics, and other toxic substances are therefore also expected to decrease by 60% or more. This will mean a decrease in the toxicity of the exhaust with 60% or more, thus a huge leap in terms of lower health effects from vehicle exhausts. This is a surprisingly large positive effect considering that there were only small differences between the two fuels in terms of regulated emissions.

Example 2 A "Ranger Basic" unmanned aircraft model with a wingspan of 1.9 meters was fitted with a "Jetcat P70" turbine. The "Ranger Basic" model is available from JetCom e.K., Robert-Bosch-Strasse 24, D-68766 Hockenheim, Germany. The "Jetcat P70" turbine is available from Ing.- Bύro CAT, M. Zipperer GmbH, Etzenbach 16, D- 79219 STAUFEN, Germany.

As a preparation, the speed of the model was compared with standard jet Al turbine fuel and fuel sample B according to the invention. The fuels were in all experiments here described mixed with 5% fully synthetic oil ("Marin TT" from OKQ8 of Stockholm, Sweden) according to the turbine manufacturer's instructions. In all experiments described here, the model aircraft was fitted with a special throttle that gave a constant fuel flow measured in milliliters per second.

The aircraft model was filled with 1 kg of fuel, and clocked manually at the start and end of a 500 meter test run. At approximately 75% throttle, the aircraft reached a speed of approximately 100 km/h with jet Al turbine fuel. The same speed was attained with approximately 60% throttle for fuel sample B according to the invention. At 75% throttle, a speed of 120 km/h was attained with fuel according to the invention.

Fuel consumption was evaluated in the following way: A test path of approximately 1000 meters around the perimeter of the airfield was used. After taking off with 1 kg of fuel, a reasonably constant throttle of 75% was held for jet Al fuel, and 60% for fuel sample B, with the aim of keeping as stable and similar speed as practically possible. After 10 runs around the test path, the aircraft was landed and weighed.

It was found that the fuel consumed was 600 grams for standard Jet Al fuel, and 530 grams for fuel according to the invention. Thus, it is obvious from this example that fuel sample B according to the invention can give either a higher speed at the same throttle, or a longer length of flight in the case that the speed is kept constant. Though the full potential of the fuel should theoretically be close to 20% fuel savings or performance improvements, only approximately 10- 15% savings were shown in these experiments.

It should also be noted that commercial aircraft use up the major part of the total fuel use during the take-off and landing cycles. Thus, a fuel containing approximately 20% more energy could potentially give a considerably increased range of operation for commercial aircraft, leading to very substantial savings in fuel use in the case that the aircraft can avoid an extra landing and take-off cycle.

Example 3

Sample D in a reformer and subsequently a fuel cell A gliding arc reformer according to French patent FR 2768424 and figure X, from ECP SARL of Orleans, France, was used to produce hydrogen from fuel sample D according to the following procedure:

The reformer has two High-Voltage electrodes inside a 0.18-L plasma space and a 0.57-L post plasma zone filled with ECP SARL granules according to French patent FR 2872149. The total inside volume of the reformer is 0.75 L. The reformer is devoted to work under high pressure (up to 22 bars) but our experiments were performed at a close to atmospheric pressure. Pure oxygen is injected in the tiny central tube of the injector. At the end of the injector (downstream) both flows mix and cross the GlidArc discharge. Partially reformed reactants then cross the post-plasma zone where three thermocouples are installed for the temperature sensing. Hot gas product exits the reformer, is cooled to ambient temperature, and is then flared or sent to a consumer, e.g. a fuel cell. A small gas sample is analyzed online using a gas chromatograph.

An experiment was run with the following operating conditions: - 50 W GlidArc power

10 g/min fuel sample D - 3 g/min oxygen (>99%) A 2 cm bottom layer of a commercial high temperature shift catalyst (SK-

201-2 from from Haldor-Topsoe, Nymoellevej 55, Kgs. Lyngby, Denmark) was used in the reformer

Maximum reforming temperature 1 100 ⁰ C

The gas was analyzed, and was found to contain approximately 33% CO2, 2% CH4, 61% H2, and 4% CO.

A filter paper placed in the path of the cool gas for 1 hour showed no trace of soot in the gas.

Thus, this experiment showed that fuel according to the invention can be reformed to hydrogen, and potentially used in a fuel cell. The produced gas in this small experiment can be used directly in a solid oxide fuel cell, which accepts carbon monoxide as fuel. For fuel cells that do not accept carbon monoxide, e.g. PEM fuel cells, a further low temperature shift and probably also a methanation step as used in the art would need to be introduced.

Analysis methods

Selection of samples

Three samples of fuels were chosen, all originally from Fischer-Tropsch (FT) processes using Fe catalysts. One sample was chosen from a low-temperature FT process (sample "AvI"), one from a medium- temperature FT process (sample "ECPl"), and one sample produced using polymerisation of olefins from the FT process (sample "A"). Sample A can also be manufactured from other feedstock sources using other production processes. a) Sample "AvI" is produced using precipitated iron based catalyst in a fixed bed reactor in a low temperature FT process. This type of FT technology has been widely published, and can be seen as a relatively stable, low-cost and commercially available technology. The feedstock of this sample is currently wood pellets, but using appropriate reforming/gasification technology, many other types of bio-feedstock can be used. Sample A was produced by Aviosol AB in its demonstration plant in Overkalix. The plant uses a fluidized bed gasifier to gasify 4 kg/h wood pellets at 1.5 bar absolute pressure and a temperature of 800-900 ⁰ C. The gas is cleaned in a Venturi scrubber, subsequently cooled and further cleaned in an alkali scrubber. After compression, the gas is passed to a 4 cm inner diameter fixed bed Fischer-Tropsch reactor of 2.8 meters length. The precipitated iron catalyst is produced using a proprietary variation of the basic recipe given in a number of German patents [1-7]. Reaction temperature is 220-250 ⁰ C, and pressure is 20-25 bars. The products from the reactor are separated in three steps, first a wax separator at 120 ⁰ C, then a diesel separator at 90 ⁰ C, and finally a gasoline separator at 2O ⁰ C. The sample analysed here comes from the diesel separator. b) Sample "ECPl" is produced using a proprietary iron catalyst in a fixed bed reactor in a medium temperature FT process. Little has until now been known about the characteristics of the products from this process. In principle, this type of product can be manufactured from any type of biomass, using appropriate technology. Sample B was produced by ECP SARL in La Ferte St. Aubin in France. Pipeline grade natural gas is reformed in a high-voltage electric arc reformer to syngas, whereby the syngas composition can be varied to emulate various feedstocks, e.g. wood chips. The syngas is compressed up to 150 bars and stored in gas cylinders. In a separate experiment, the syngas is passed into a fixed bed Fischer-Tropsch reactor [8] containing a proprietary iron catalyst. The reaction temperature is 310-330 ⁰ C and the pressure is 20-25 bars. One-pass conversion rates of CO reaching 90% are observed at high space velocities. The products from the reactor are separated in two steps: a wax separator and remaining liquids separator, while very limited amounts of gaseous hydrocarbons are generated. The products from the wax and liquid separators are mixed before the analysis, c) Sample A is a fuel produced according to the invention Sample preparation and analysis

All analyses were performed in single measurements. For Py-FIMS, about 0.2 mg of solid or 0.1 μl liquid samples are transferred to a quartz tube and introduced into the fore-vacuum chamber of a double-focussing Finnigan MAT 900 mass spectrometer (Finnigan, MAT, Bremen, Germany). The quartz tube with sample is inserted into the ion source and heated under high vacuum (10 ⁴ Pa) from 50 to 500 ⁰ C in steps of 10 ⁰ C per min. During about 25 min of total measurement time, 48 magnetic scans are recorded for the mass range 15 to 900 Dalton (single spectra). These are combined to obtain one thermogram of total ion intensity (TII) and a summed Py-FI mass spectrum. All Py-FIMS data were normalised with respect to mg/μl sample. For detailed descriptions of the Py-FIMS methodology and of statistical evaluations of sample weight and residue, volatilised matter and total ion intensities (TII) see [9- 10].

A raw data file containing the temperature -resolved mass traces will be the results of the Py-FIMS analysis. For each sample, there will be 47 temperature steps from 5O ⁰ C to 510 ⁰ C for each of the 885 mass traces, a total of 41 595 data points per sample.

Interpretation of measurement data Fischer-Tropsch fuels mainly contain carbon and hydrogen [H]. A small content of oxygen may also occur, mainly in crude samples and for low molecular weights. Since both carbon and oxygen have an even molecular weight, and all stable hydrocarbons also have an even molecular weight, it could be expected that all occuring compounds in Fischer-Tropsch fuels should have an even molecular weight.

Surprisingly, the py-FIMS spectra also contain substantial amounts of molecules with odd (uneven) molecular weight. There are two possible sources for such molecules with odd molecular weight: 1. Occurrence of isotopes in the samples, in this case mainly ¹³ C. The

¹³ C isotope of carbon has a typical natural occurrence of 1.08 %, and the likelihood of finding a specific number of C-atoms in a hydrocarbon molecule can therefore be calculated using the binomial distribution. The natural occurrences of other carbon, hydrogen and oxygen isotopes are so low that they can safely be neglected in comparison to ¹³ C.

2. In the py-FIMS process, protons can be transferred between molecules. Some molecules, e.g. alkenes, can accept a proton (hydrogen ion) from other compounds, e.g. alcohols or acids. Even though two compounds have the same molecular weight, it may be possible to distinguish between them by the amount of protons transferred, e.g. alkenes will easily accept protons from other compounds, whereas alkanes will not accept protons at all. For the purposes of this study, it will be practical to group oxygen-containing compounds together with non-cyclic alkanes with a similar molecular weight. Thus, acetic acid with a molecular weight of 60 Dalton will be grouped together with butane with a molecular weight of 58 Dalton. Thus, the compounds will be grouped according to the total number of carbon plus oxygen atoms in the molecule. The general formula for a non-cyclic alkane is C _n H2n+2- The molecular weight in Dalton for a non-cyclic alkane will therefore be 14-n + 2, where n is the number of carbon atoms (14 being the weight of one "segment" of the hydrocarbon chain consisting of one carbon and two hydrogen atoms). To identify individual compounds, corrections in relation to the non-cyclic alkane can be used, according to table 4.

Table 4. Group corrections in relation to the respective non-cyclic alkane. The table is used to infer probable molecular composition of hydrocarbons from molecular weights. This table is logically derived from the molecular structure of the compounds mentioned.

Analysis of raw data

The structure of one group of compounds with a total of 20 carbon and oxygen atoms for sample AvI is shown in Figure 1. The main value representing non-cyclic alkanes, C20H42, is found at 282 Dalton.

Estimating the amount of non-cvclic alkanes in the samples Referring to table 4, a reasonable approximation of the amount of alkanes in the samples should be to simply add the 282 and 283 Dalton molecule weight values, since the 283 Dalton value will consist mainly of alkanes containing a ¹³ C isotope. The likelihood of two or more ¹³ C isotopes in the same molecule is less than 2%, so as a first approximation this should work.

In fact, the measured ratio between the D283 and D282 value is 0.202, whereas the expected value from the binomial distribution is 0,218, in excellent agreement. This difference can e.g. be explained by a feedstock source that has less occurence of ¹³ C isotopes than the average of the earth's crust.

For samples ECP and EcoPar, the agreement for the ratio between the D283 and D282 values are even better, both are within +/-2% of the expected value. Referring to table 4, carboxylic acids, esters, and ethers that have donated a proton may also occur in the D283 value. However, this content is known to be small [11], probably much smaller than the amount of molecules with two or more C atoms. Estimating the amount of alkenes and cyclo-alkanes in the samples

From table 4 and the discussion concerning alkanes above, it can in analogy be deduced that the total amount of mono-alkenes and mono-cyclo-alkanes can be estimated with the sum of the values for D280 and D281.

In analogy, the content of di-cyclo-alkanes and di-alkenes can be estimated, etc.

In table 5, first approximation values are given for main groups of compounds in the samples.

Table 5. First approximation analysis results in overview, corrected for isotope pattern.

Detailed analysis of data after filtering

For a quantitative analysis of the data, some data points were excluded according to the following criteria:

1. Only peaks which had a total occurrence of more than 0.1% were included in the analysis.

2. Temperature-resolved data points where either the main value or the next higher Dalton value had an occurrence below 0.01 % were excluded. The reason that this filtering of data was necessary was that it was observed that the isotope pattern in some cases showed unrealistic values. Thus, a more detailed analysis required that some non-significant data was removed by filtering. Heuristic testing determined that the limits mentioned above gave realistic values of the isotope pattern.

It was noted that the Alkene/mono-cyclo-alkane values of sample AvI had a much higher aberation from the expected than the other two samples. The alkane/mono- cyclo-alkane peak of sample AvI is known to represent alkenes, whereas this peak is known to represent cyclo-alkanes for sample A.

Since the curve of sample ECPl is much closer to sample A than to sample AvI , the sample ECPl peak would also be expected to mainly contain mono-cyclo-alkanes. It was noted that for carbon numbers 18-20, the aberrations from the expected isotope patterns were very close for samples AvI and A. Since curve AvI is known to consist of alkenes, this seems to indicate that compounds in the alkene/mono- cyclo-alkane peak of sample A mainly consists of alkenes for lower carbon numbers than approximately 18.

Analysis results after detailed data analysis and filtering is shown in table 6. All significant peaks have been taken into account.

Table 6. Analysis results after data analysis and filtering. It has been assumed that the alkene/mono-cyclo-άikane peak of samples ECPl and A consists of alkenes below a carbon+oxygen number of 18.

References

1. Patent DE919288 (1949)

2. Patent DE881497 (1951) 3. Patent DE937047 (1954)

4. Patent DE937706 (1951)

5. Patent DE923127 ( 1950)

6. Patent DE959911 (1952)

7. Patent DE1061306 (1952) 8. M. Czernichowski, A. Czernichowski, Reacteur a plaques et son fonctionnement dans un procede catalytique exothermique, French Patent 2824755, May 15, 2001.

9. Schulten H-R. 1999. Analytical pyrolysis and computational chemistry of aquatic humic substances and dissolved organic matter. Journal of Analytical and Applied Pyrolysis 49: 385-415. 10. Sorge C, Schnitzer M, Schulten H-R. 1993. In-source pyrolysis-field ionization mass spectrometry and Curie-point gas chromatography/mass spectrometry of amino acids in humic substances and soils. Biology and Fertility of Soils 16: 100- 1 10.

1 1. CD. Frόhning, H. Kolbel, M. Ralek, W. Rottig, F. Schnur, H. Schulz, in: J. Falbe (Ed.), Chemierohstoffe aus Kohle, G. Thieme-Verlag, Stuttgart, 1977