摘要:

Proceedings of the Society for Analytical Chemistry CONTENTS ?roc. SOC. Analyt. Chem. Vol. 7 No. I Pages 1-20 Reports of Meetings . . .. I ence .. .. .. . * 2 “Application of Microchemical Techniques in Pet roc he mica1 and Allied Industries” .. 4 Notice . . . . . . . . 19 Papers Accepted for The Analyst 20 Forthcoming Meetings Back cover Honorary Secretaries Confer- Summaries of Papers January 1970 Vol. 7 No. I January 1970 PROCEEDINGS OF THE SOCIETY FOR ANALYTICAL CHEMISTRY President of the Society T. S. West Hon. Secretary of the Society W. H C. Shaw Hon. Treasurer of the Society G. W. C. Milner Hon. Assistant Secretaries of the Society D. I. Coomber; D. W. Wilson Secretary Miss P. E. Hutchinson 9/10 SAVILE ROW LONDON WIX IAF Telephone 01-734 6205 Editor J. B. Attrill Telephone 01-734 3419 Proceedings is published by The Society for Analytical Chemistry and distributed to all members of the Society and to subscribers with The Analyst; subscriptions cannot be accepted for Proceedings alone.Single copies may be obtained direct from the Society’s Distribution Agents The Chemical Society Publications Sales Office Blackhorse Road Letchworth Herts. (NOT through Trade Agents) price 0 The Society for Analytical Chemistry 5s. post free. Remittances MUST accompany orders. INDEXES 1969 The Index to Volume 6 of the Proceedings and the Index to Volume 94 of The Analyst will be distributed to Members of the Society and to Subscribers in February. The Index to Volume 17 of Analytical Abstracts covering July to December 1969 is expected to be ready for distribution during April and copies will automatically be sent to Members and Subscribers entitled to receive them.

ISSN:0037-9697    

DOI:10.1039/SA97007FX001

出版商:RSC    

年代:1970

数据来源: RSC

摘要:

TS TINGS t. January 19701 REPORTS OF MEETINGS 3 -year to avoid some duplication of dates but by holding this meeting any major conflictions of both subject matter and dates can usually be avoided. I t is hoped that this is now so for 1970-7 1. Following this Session a most enjoyable lunch was held at the Westbury Hotel Bond Street a t the invitation of the President and Honorary Officers of the Society. Mr. K. Smith (Biological Methods Group) thanked the hosts on behalf of all the Honorary Secretaries present. In the afternoon a meeting which was attended by the President and Honorary Officers of the Society afforded an opportunity for the Secretaries to discuss matters of general policy and over-all organisation of the Groups and Sections. DY. J . M . Ottaway (Scottish Section) MYS. D. E. Butterworth (+crochemical Methods Group) and M Y . J . B. Aldred (North of bngland Section) The President entertains the Secretaries at luncheon

ISSN:0037-9697    

DOI:10.1039/SA9700700002

出版商:RSC    

年代:1970

数据来源: RSC

摘要:

THE SOCIETY FOR ANALYTICAL CHEMISTRY Forthcoming Meetings-couttiutwd Wednesday 25th MICROCHEMICAL METHODS GROUP London Discussion Meeting. J,ONDON Discussion on “The Direct Determination of Oxygen in Organic Compounds ” The Leicester Lounge Glasshouse Street London W. 1 ; 6.30 p.m. CHROMATOGRAPHY AND ELECTROPHORESIS GROUP. Discussion on “Pre-coated Plates and Films for Thin-layer Chromatography,” to be introduced by Professor E. J . Shellard and G. Whitechurch. The Leicester Lounge Glasshouse Street London W.l; 6.15 p.m. to be introduced by F. H. Oliver. Thursday 26th J o ~ n ON THE SOCIETY FOR ANALYTICAL CHEMISTRY Forthcoming Meetings January Thursday 29th LONDON February Thursday 5th LONDON Tuesday 10th LEEDS Tuesday 10th NEWCASTLE Wednesday 1 1 th LONDON Thursday 12th LIVERPOOL Thursday 12th LONDON Tuesday 17th ED INBURGH Thursday 19th LONDON Tuesday 24th JOINT PHARMACEUTICAL ANALYSIS GROUP Inaugural Meeting followed by a Discussion Meeting on “The RBle of the Analyst in Pharmaceutical Control.” Pharmaceutical Society of Great Britain 17 Bloomsbury Square London W.C.2; 2.30 p.m.RADIOCHEMICAL METHODS GROUP Meeting on “Radiometric Standards.” “The Measurement of Radioactivity,” by P. J. Campion. “Practical Aspects of Using Radioactive Standards,’’ by L. C. Myerscough. “Criteria of Purity for Radiopharniaceuticals,” by J . C. Charlton. Lecture Room Laboratory of the Government Chemist Cornwall House NORTH OF ENGLAND SECTION jointly with the Leeds University Union Chemical “Recent Work on Dithizone and its Selenium Analogue,” by Professor The University Leeds; 5.15 p.m.NORTH EAST SECTION jointZy with the Newcastle upon Tyne Section of the Royal Institute of Chemistry. “Future Trends in Chemical Analysis,” by J . B. Headridge. Chemistry Department The University Newcastle upon Tyne ; 6.30 p.m. SOCIETY. “Analytical Cosmochemistry (or Migrants from Outer Space Enter the Lecture Theatre 342 Department of Mechanical Engineering Imperial NORTH OF ENGLAND SECTION SPECIAL TECHNIQUES and ATOMIC SPECTRO- Liverpool. BIOLOGICAL METHODS GROUP. “Immunological Assays.” Pharmaceutical Society of Great Britain 17 Bloomsbury Square London SCOTTISH SECTION jointly with the Edinburgh University Chemical Society. “Amplification Reactions and their Applications ” by Professor R. Belcher. University of Edinburgh Edinburgh ; 4.30 p.m AUTOMATIC METHODS GROUP.Discussion on “Bridging the Gap An open discussion of the Problems of Manufacturers and Users in the Field of Automatic Instrumentation,” to be introduced by D. H. Loebl and R. J. Weir. Stamford Street London S.E. 1 ; 2.30 p.m. Society. H. M. N. H. Irving. Laboratory),” by A. A. Smales. College London S.W.7; 6.30 p.m. SCOPY GROUPS on “Fluorescence Spectroscopy.” W.C.1; 5 p.m. The Leicester Lounge Glasshouse Street London W. 1 ; 6.30 p.m. NORTH EAST SECTION jointly with the Teesside Section of the Royal Institute MIDDLESBROUGH of Chemistry. L. S. Phillips. “Yorkshire’s Pink Gold its Discovery Exploration and Evaluation,” by “The Analysis of Crude Potash Materials,” by J. M. Skinner. Constantine College of Technology Middlesbrough ; 8 p.m. NORTH OF ENGLAND and MIDLANDS SECTIONS and THERMAL ANALYSIS GROUP jointlly with the Chemical Society on “Newer Techniques of Thermal Analysis with Special Reference to Calorimetry.” Wednesday 25th KEELE Speakers to include M. M. Faktor Professor I>. Rouquerol and I. Wiley. University of Keele Staffs.; 2.30 p.m. [continued on inside back cover Printed by W Heffer & Sons Ltd Cambridge England

ISSN:0037-9697    

DOI:10.1039/SA97007BX003

出版商:RSC    

年代:1970

数据来源: RSC

摘要:

4 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [Proc. SOC. AnaZyt. Chew. The Application of Microchemical Techniques in the Petrochemical and Allied Industries The following are summaries of some of the papers presented at a Joint Meeting of the Microchemical Methods Group with the Mid-Southern Counties Section of the Royal Institute of Chemistry held on September 19th and 20th 1969 and reported in the October 1969 issue of Proceedings (p. 182). Recent Advances in the Application of Organic Elemental Analysis BY P. GOUVERNEUR (KoninkZijke/Shell-Laboratoyium Amsterdam Holland) QUANTITATIVE elemental analysis of organic products has attracted interest for more than a century. This is surprising particularly as many other methods for determining the nature of organic substances have also been developed.At least four circumstances are responsible for the present state of elemental analysis which are illustrated with examples from this laboratory. “COMMON” ELEMENT ANALYSIS- Here most often milligram amounts of sample are used to determine the elements carbon hydrogen oxygen nitrogen sulphur the halogens phosphorus etc. with a precision usually to the first decimal place. We are entering the era of “Element Analysers.” Extensive use of the Perkin-Elmer Model 240 C.H.N. Analyser has given rise to two experiences. In the nitrogen determination it was observed that with high quality oxygen nitrogen contents down to 0.05 per cent. could be determined on a 2-mg sample together with carbon and hydrogen thus making sep- arate Kjeldahl analysis superfluous. In the determination of hydrogen in the high hyd.rogen range from about 10 per cent.upwards positive discrepancies occurred of several per cent. relative. This was caused by a deviation in the hydrogen calibration graph which did not pass through zero as did those for carbon and nitrogen. The hydrogen factor is therefore not generally valid but is on a sliding scale. Now that the reason is known this no longer causes trouble. For oxygen determination the preferred method is based on the carbon reduction principle and automatic colorimetric titration of the resulting carbon dioxide with sodium methoxide. The application to metal-containing compounds with Indulin Base RM as a pyrolysis aid has proved usefu1.l The method most commonly used at present for the other elements such as sulphur chlorine bromine iodine fluorine phosphorus and mercury is the Schoniger flask method which was developed about 15 years ago and is now well established.An ultimate target would be a multi-element analyser with which the component elements could be traced simultaneously in a single operation even if only approximate values (to within 1 per cent. perhaps) were obtained. Interesting developments have been r e p ~ r t e d ~ . ~ and recently an attempt was made with a small mass spectrometer to monitor the combustion products.* ULTRA-PRECISE ANALYSIS- A better than usual precision is often required particularly for carbon and hydrogen. Examples are the “ring-carbon” content of saturated oil fractions the styrene content of styrene - butadiene rubbers and also in the chemical hydrogen uptake in the hydroprocessing of oils.A flexible method for obtaining precise carbon and hydrogen values is based on careful combustion of a large ample.^ By variation in the size of the sample (from 0.1 to 1 d) products can be determined with a precision to 1 2 or 3 decimal places as required. The product (weighed in a tube) is burned quickly and completely (20 minutes) in pure oxygen without any risk of explosion. The experiment is concluded conventionally i.e. the water January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES 5 and carbon dioxide formed are weighed. With the highest sample intake (1 ml) a precision (standard deviation) of 0.002 per cent. of hydrogen and 0.008 per cent. of carbon can be obtained. As yet we have found no instrument apart from the balance that permits such high accuracy.TRACE AKALYSIS- In this field the established Wickbold oxy-hydrogen method barely manages to satisfy present-day requirements regarding speed and sensitivity. Better possibilities are offered by the so-called Dohrmann microcoulometric titration system. It is suitable for automatic titration of extremely small (nanogram) amounts of reactive gases such as hydrogen sulphide hydrogen chloride ammonia and sulphur dioxide and appropriate argentometric acid - base or iodimetric titrations can be performed. The principle has been illustrated elsewhere,6 and in Table I established routes for the determination of traces of nitrogen chlorine and sulphur in oils and other organic products are given. TABLE I TRACE ELEMENT ANALYSIS BY MICROCOULOMETRY Conversion Nitrogen ... . Reduction Chlorine.. .. .. Reduction or Sulphur.. .. .. Oxidation Sulphur.. . . .. Reduction Element compound formed NH3 oxidation HC1 SO 13- + H2S Titration L4cid - base H+ + NH3 -+ NH,+ Argentometric Iodimetric Argentometric 2Ag+ + s2- -f Ag,S Ag- -1- C1- + AgCl SO +- H,O 3 31- + 2H+ + SO3 The p.p.m. concentrations are therefore measured as nanogram amounts of element in milligram amounts of sample something that was scarcely even dreamed of a few years ago. A further advantage with small samples is the speed; a recorder peak appears a few seconds after the sample has been injected. Unfortunately such rapid techniques are not always possible. So for example for the trace oxygen determination a much slower method must be used again based on the carbon reduction principle.The small volume of carbon dioxide involved here is no longer deter- mined by titrimetry but by manometry after selective freezing out with a pressure transducer. Products analysed include oil fractions and high po1ymers.l ELTRA-MICRO ANALYSIS- Here the determination of elements in normal concentrations with extremely small amounts of sample e.g. a fraction of one milligram are considered. Examples are carbon and nitrogen determined by manometry with pressure transducers and halogen by microcoulometry. They show that useful results can be obtained on sample masses in the 5 to 20-pg range; results are produced within 10 minutes. Volatile and hygroscopic samples however present problems. CONCLUSION It is evident that we are rapidly approaching extremes. Multi-elemental analysis ultra-precise aspects trace analysis and ultra-micro analysis are significant examples.We can say that we are able to detect elements in normal concentrations in extremely small amounts of material in fact in almost nothing (ultra-micro). We are also able starting with a slightly larger sample to detect extremely minute amounts (almost nothing) of an element. Just one more step and we shall be able to detect almost nothing in almost nothing. While we now detect nanograms in milligrams of sample (microcoulometry) the time will come when we can detect picograms or perhaps ferntograms in a single microgram. Truly almost nothing in almost nothing. 6 ~WCROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [PYOC. SOC. A.tzaZyt. Chem. REFEREKCES 1. 2. 3. 4. 5. 6. 7. Gouverneur P. and Bruijn A.C. Talanta 1969 16 827. Beuerman D. R. and Meloan C. E. Analyf. Letf. 1967 1 195. Dugan G. and Aluise V. L4. A n a l y f . Chem. 1969 41 495. Leuven H. C. E. van in preparation. Boes J. and Gouverneur P. 2. analyt. Chem. 1964 205 58. Gouverneur P. and van dc Craats F. ,4nalyst 1968 93 782. Gouverneur P. van Leuven H. C. E. Belcher R. and Macdonald A. 31. G. Analylica Chim. Acta 1965 33 360. Determination of Trace Components in Petrochemical Products BY G. B. CRUMP (Shell Research Ltd. Egham Research Laboratories Suvrejf) THE background to micro analysis in the petrochemical industry was described and the diversity of the products examined was emphasised. The inadequacy of classical definitions of micro analysis to describe the scope of the trace analysis undertaken by the petrochemical industry was indicated.In particular the determination of carbon hydrogen oxygen sulphur nitrogen halogens and a few simple organic functional groups represent only a minor part of the analytical load. The need for rapid and accurate methods has led to extensive analytical instrumentation. Many of the available techniques were described and it was shown how the choice of technique was conditioned by several important factors economics including cost of equipment and experience of staff; need for quick answers; need for accurate answers; instrument reliability; skill of staff; and versatility of technique. It was emphasised that many problems could not be uniquely solved by a single method however sophisticated. For example a high resolution mass spectrometer could sometimes provide less useful information than thin-layer chromatography or an infrared spectrometer might provide less information than a potentiometric titration.However the more techniques that could be applied the more certain would be the answers. The technique that has probably found widest application in the petrochemical industrj- is gas chromatography. Examples were given in which otherwise intractable problems had been resolved by it. The advent of specific element detectors is expanding its scope even further. Its impact on the elemental analysis field is evidenced by some of the automatic C.H.N. analysers in which gas chromatography is used to separate nitrogen carbon dioxide and water. Moreover the coupling of gas chromatographs to mass and infrared spectrometers enables analysis of single constituents of complex mixtures to be made.Another of the newer techniques that has revolutionised trace analysis is atomic absorption. The versatility of this technique is such that colorimetric determinations are now infrequently made for trace metals. Microcoulometry has dramatically simplified the determination of traces of sulphur nitrogen and halogens. It was emphasised that it was the aim of the petroleum - chemical analyst to be able to determine any trace component required in however complex a mixture and with com- ponents (whether compound or element) even at sub-p.p.m. levels. Polymeric Analysis by Micropyrolysis Gas - Liquid Chromatography BY G. BAGBY (B.P. Chemicals I-td. Research avzd Developnzent Depavtmen f Epsoni Suvrey) MICROPYROLYSIS gas - liquid chromatography basically involves the pyrolysis of a polymeric sample in the carrier-gas stream of a gas - liquid chromatographic apparatus so that the volatile products can be resolved chromatographically.There are several techniques of heating samples which can be divided into four categories vix. the furnace method; the simple filament method; the boosted filament method ; and the induction heating method. The furnace method has several disadvantages which result in poor gas - liquid chromato- graphic resolution. In the simple filament methodl a sample of less than 10-3g is placed January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL IXDUSTRIES 7 on a filament through which a current pulse is passed for not longer than 30 seconds if longer there may be poor gas - liquid chromatographic resolution.Sample sizes of 1 mg can be treated in this way by using a katharometer detector and the volatile products can be individually trapped from the effluent gas and then identified by mass spectrometry. This apparatus is adequate for characterisation and analysis of polymers but is not suitable for quantitative kinetic studies where several stringent requirements must be met .2,3 Firstly the duration and the temperature of the pyrolysis must be accurately specified for kinetic studies. Secondly the sample size must be such that the degradation reaction is not con- trolled by the rate of diffusion of volatile products through the molten polymer. These problems are solved by using a boosted filament apparatus. The first problem is overcome by depositing the sample within a limited region of a straight nickel ribbon filament and supplying an initial current boost to bring the filament to the desired degradation temperature within 1 second.The second problem can be overcome by using sufficiently small samples (less than lo-* g) and very sensitive gas - liquid chromatographic apparatus. The fourth technique is that of induction heating in which a sample is placed on a ferromagnetic wire. This wire is rapidly heated by eddy currents induced in it by a high frequency supply which feeds an induction coil surrounding the wire. The wire continues to consume power until it reaches its Curie temperature when the power consumption is reduced and the temperature of the wire is stabilised. The type of chromatographic apparatus is determined by the sample size and required sensitivity.The resolution of pyrolysis products presents the same problems as those generally encountered in gas - liquid chromatography. There are three principal methods of characterisation by comparison. The method of complete pyrolysis at a single temperature to provide a good yield of the characteristic volatile products results in a pyrolysis gas - liquid chromatographic pattern which then allows a comparison of peak heights and retention times. A more useful method is sequential degradation of the same sample throughout a range of temperatures so that a series of gas - liquid chromatographic patterns is obtained and these provide fingerprints of the samples. The simple filament apparatus is suitable for either of these characterisation methods.The third method requires the use of the boosted filament apparatus to make kinetic measurements. The most convenient approach is to determine the specific rate of degradation of a sample of known molecular weight. It is desirable to fractionate the polymeric sample by gel-permeation chromatography4 and then degrade these fractions the molecular weights of which can be calculated from the gel-permeation chromatograms. By this method it may be possible to characterise very small changes in the composition of a polymer (e.g. the presence of a fragment of a chain transfer agent in the polymer molecule). Copolymers and polymer mixtures can easily be analysed by micropyrolysis gas - liquid chromatography by any of the four pyrolysis methods outlined above. The procedure here is first to perform a preliminary pyrolysis of each component of the sample at several high temperatures (between 500" and 700" C) to choose a principal characteristic peak for each component and also to determine a suitable pyrolysis temperature at which the total yield of each characteristic product can be obtained within an acceptable degradation period.Secondly a calibration with pure components is performed. Several known weights of each pure component are completely pyrolysed at the selected temperature and calibration plots are drawn for the pure components. Thirdly the mixture or the copolymer is completely degraded at the selected temperature and its composition then calculated from the calibration graphs. Finally a check is made to ensure that the total weight of the mixture as deduced above is equal to the original total weight placed on the filament.This procedure has been applied to many systems and results are comparable with those obtained by analytical and infrared methods. The speed and facility of the method make it attractive especially for routine analytical work on small samples. REFERENCES 1. 2. 3. 4. Barlow A. Lehrle R. S. and Robb J. C. PoZymev 1961 2 27. Barlow *4. Lehrle R. S. Robb J. C. and Sunderland D. Ibid. 1967 8 523. Lehrle R. S. Lab. Pvact. 1968 17 696. Bagby G. Lehrle R. S. and Robb J. C. Polyuzev 1968 9 284. 8 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [Proc. SOC. AnaZyt. Chem. A Gas-chromatographic System to Facilitate the Massspectrornetric Analysis of Minor Deuterated Components in a Gas Mixture BY D. SHEPHERD (I.C.I.Ltd. Petrochemical and Polymer Laboratory Runcorm Cheshire) STUDIES of a gaseous-phase reaction required the resultant mixture of products to be analysed by a gas-chromatograph - mass spectrometer (directly coupled) combination. The use of deuterated starting materials and subsequent mass-spectrometric determination of the position of the deuterium in the product molecule helped in the evaluation of the reaction mechanism. Economic use of deuterated compounds and the need for maximum sensitivity of analysis led to the use of a 0-030-inch i d . stainless-steel packed column with a flow-rate equal to the maximum input to the mass spectrometer (1 ml of helium minute-l). A sample volume containing sufficient mass of products for detection was too large relative to the column plate volume and had to be reduced by removal of the nitrogen present.The established technique of removing components from a gas stream by passing it through a cooled packed tube has been used; re-generation by heating the tube presented the products to the column in a volume determined by their total molecular volumes. The apparatus described is a means of incorporating such a trap into a gas-chromatographic system to permit complete automation of the concentration and separation steps of the analysis. The basic chromatograph is a Pye Series 104 model 44 instrument fitted with a Servomex micro katharometer. The trap is placed in the carrier-gas stream before the analytical column the normal injection point having been removed. On the carrier-gas side of the trap a T-piece introduces a vent to atmosphere controlled by a solenoid valve.On the column side of the trap a T-junction makes an entry point for a pressure-controlled helium supply provided via a sample loop. This latter arrangement when the vent to atmosphere is open gives the gas flow needed to transfer the sample from the loop to the cooled trap. A loop-and-column by-pass line shares its gas supply with the sample transfer system and its detector entry point with the analytical column but a careful balance of pressures keeps gas flows in both independent and unmixed. The values and directions of the three gas flows are fixed by the choice of mass flow and pressure-regulator settings coupled with made-to-order gas-flow restrict ors (crimped stainless- steel capillary tubing) placed at the following points the vent to atmosphere as a continuation of the analytical column immediately after both detector cells and on the inlet of the mass spectrometer.Helium continuously leaves from the detector through a restrictor to atmos- phere and the mass spectrometer will take in through its restrictor 1 ml minute-l of helium supplied at atmospheric pressure. This produces a stable system when a direct coupling is made. The mass-flow controlled carrier gas transfers the concentrated products from the heated trap to the column and continues throughout the following analysis period. The switch between the two gas supplies is made by a 3-way Schrader spool valve. A specially modified 3-way Schrader spool valve forms the injection system one position stops the flow across the valve while allowing a sample loop to be filled the other places the loop in the gas stream.The trapping tube is cooled from the outside by nitrogen gas previously passed through a heat-exchange coil immersed in liquid nitrogen. A current acting on the ohmic resistance of the stainless-steel tubing is used to heat the trap. The compressed air to activate the spool valves is controlled by 3-way solenoid valves as is the nitrogen cooling gas. The microswitches of a Crouzet (England) Ltd. punched card programmed timer are used to control the solenoids the primary circuit of the heater transformer and the oven temperature programmer unit. The order of operation and the required time intervals found by experi- ment are punched out on the programme card which then controls further runs automatically. The order of events is as follows.A sample is placed in the loop and the trap is then cooled. The vent is opened and the loop switched into circuit simultaneously to start the sample-trapping procedure. The vent and loop states are reversed and the trap heated. -4 switch to the mass-flow controlled carrier gas transfers the sample to the column and starts the analysis. At the same time the temperature programmer unit is switched on. At the end of the analysis a switch to the pressure-controlled helium supply takes place just before the column is cooled ready for the next sample. January 19701 MICROCHEMICAL TECHNIQUES I N PETROCHEMICAL INDUSTRIES 9 A typical sample consists of N, 88; O, 8; C3H (10 per cent. of C3D,) 3; CO, 1 ; CH,.=CHCN (10 per cent. deuterated) 0-2 (per cent. v/v).Chromatograms showed a straight base-line with negligible disturbance during the trapping period. Peaks for CO, C3H and CH,=CHCN appeared to be quantitative and had plate heights comparable with standard values. For this analysis a 3-inch trap and 6-foot column both 0430-inch i.d. and packed with Porapak Q were used. Dewar insulation of the trap from the atmosphere allowed a trapping cycle time of 90 seconds. A 5-ml sample volume containing 100 p.p.m. of total acrylonitrile gave a mass spectrum from which the deuterated species can be identified. Modification of the injection system gives two further potential uses the concentration of impurities in air to give increased detection sensitivity in air pollution studies and the transfer of peaks from packed columns to capillary columns for further resolution.New Developments in the Chromatography of Petroleum Products BY S. G. PERRY (Esso Research Centre A bingdon Berks.) SEVERAL applications of both gas and liquid chromatography that have been used in the Esso laboratories at Ahingdon are described to give a broad picture of the different roles chromatography plays. TRACE AKALYSIS WITH SELECTIVE GAS-CHROMATOGRAPHIC DETECTORS- The introduction of highly selective detectors in gas chromatography is probably the most significant recent advance. There are at least six sensitive and selective gas-chromato- graphic detectors commercially available but attention was concentrated on the flame thermionic and the flame photometric devices. The flame thermionic detector gives a high and selective response to phosphorus compounds and its principal use has been in the deter- mination of phosphorus-containing pesticides ; however since as little as 10-l1 g of trialkyl phosphates can be detected these materials can also be used as tracers for example in following the movement of different petroleum products in distribution systems.The flame photometric detector has a sensitivity to sulphur in the p.p.b. range but this varies with the degree of chromatographic separation between the sulphur compound and for example hydrocarbons in the sample. These compounds quench the emission of the S species so that elution of a sulphur compound on the tail of a large hydrocarbon peak leads to diminished response. A particularly unfavourable analysis is for traces of thiophene in high purity benzene. Separation of these two substances is always in the sequence benzene before thiophene and the low degree of separation is such that the 99 per cent.benzene peak tails through the p.p.m. thiophene peak. The limit of detection of thiophene therefore is only 0.1 p.p.m. SIMULATED DISTILLATION BY GAS CHROMATOGRAPHY- Gas-cliromatographic distillation affords better resolution of the front end of the sample higher speed smaller sample size and it is believed but has not yet been proved better precision than conventional methods. Efforts are being made to standardise the technique and controversial features are being considered. The points to be standardised and at present debatable are stationary phase type and amount; final column temperature; detector type and sample size ; and injection conditions.Preference was expressed for methyl silicone gums operated at temperatures up to 350" C with on-column injection small (1 1.1) sample size and a flame ionisation detector. LIQUID CHROMATOGRAPHY- Resurgence of interest in liquid chromatography in columns has involved radical re- thinking of column technology and a technique has been evolved that has many points in common with gas chromatography. The essential features of a modern liquid chromatograph are long narrow-bore columns packed with fine materials (e.g. 10 to 20-pm particles) through which the mobile phase is pumped perhaps at pressures as high as 100 atmospheres. In addition a continuous recording detector is used to monitor composition of the eluent. In 10 MICROCHEMICAL TECHSIQUES IN PETROCHENICAL INDUSTRIES [PYOC.S O L . Analyt. Chew. return for this considerable sophistication in instrumentation a high speed high resolution chromatograph equivalent in performance and ease and convenience of operation to a modern gas-chromatographic unit is available. In particular one has a tool which if required is capable of providing vastly greater resolving power than thin-layer chromatography in faster time. Considerable use has been made in the Esso laboratories of the large pored ion-exchange resins now available which are stable in organic media. These resins are best regarded as selective (acidic or basic) adsorbents and they afford a convenient means of concentrating certain organic species from petroleum products. The behaviour of some representative classes of organic compounds on these macro-reticular resins was indicated.They are invaluable for analysis of the minor but often important naturally occurring non-hydro- carbon constituents of petroleum. The final example of progress in chromatograph>- related to fluorescent indicator adsorp- tion analysis of hydrocarbons by type. The standard procedure takes about 1 hour for completion and is carried out by applying a pressure of up to 5 p.s.i.g. to the column head. By using higher pressures up to about 30 p.s.i.g. and adding sample a t a more suitable point analyses can be carried out in less than 10 minutes with scarcely reduced precision. The Identification of Hydrocarbon Oils by Added Markers BY K. FIELD (Laboratovy of the Government Chemzst .Minzstvy of Technology Corizwall House Stamford Street Loizdon S.E. 1) Ix accordance with the Hydrocarbon Oil Regulations operative in this country it is an offence to use either a duty-free or a duty-rebated hydrocarbon oil as a road fuel.This laboratory is responsible for the provision of an analytical service that is necessary if adequate control is to be maintained over the distribution and use of these hydrocarbon oils. The type of problem referred to in this paper concerns the misuse of the following three groups of hydrocarbon oils when used either in place of or in admixture with the duty-paid oils petrol and Derv. (i) Oils containing neither markers nor dyes (benzole toluole xylole white spirit naphthas technical kerosenes and Avtur). (ii) Oils containing the official prescribed markers (gas oil and burning oil) (iii) Oils containing soluble dyes (premium burning oils and tractor vaporising oils) When we receive a sample of hydrocarbon oils we have little or no knowledge of its com- position and we rarely have available any reference samples that could be used for comparison purposes.Gvoup (i)-Initially the samples are examined by standard methods of hydrocarbon oil analysis (distillation flash-point gravity etc.) . From these results it is usually possible to classify the types of hydrocarbon oils present in the mixture. Assessment of the sample can then be made in terms of the volatility of the oil recorded a t a suitable reference tem- perature. This reference temperature is pre-determined by the examination of the volatility characteristics of many reference samples for example petrol kerosene and Derv that are distributed throughout the country a t the time of sampling.The use of gas - liquid chromatography has enabled us to make useful advances in this type of analysis. A study was recently carried out in which the distribution of the major peaks present in the chromatograms of petrol tractor vaporising oil burning oils and Derv were noted. Although in many cases the major peaks present in these oils coincide their relative proportions present may vary considerably. By this approach it is possible to make assessments of petrol (rich in C5 fractions) in kerosene or Derv and tractor vaporising oil (rich in C to C, fractions) also in petrol or Derv. The observed overlap between burning oils and Dervs makes assessments in these cases more difficult. This work was carried out on a Pye 104 gas chromatograph with a column packed with 5 per cent.Apiezon L on Chromo- sorb P DCMS treated. January 19701 MICROCHEMICAL TECHXIQUES IX PETROCHEMICAL INDUSTRIES I1 Gmq5 (ii)-An ideal situation would be one in which the identification and determination of each type of hydrocarbon oil could be achieved by reference to its own “marker” but experience has shown that this is not easily achieved. This is not surprising as the criteria that the “ideal” marker ought to meet are as follows. (1) It must leave no residue on evaporation of the hydrocarbon oil. (2) It must not possess any odour or colour that could be imparted to goods during (3) I t must not react with any of the constituents of the goods under manufacture or (4) It must not have any acute toxicity or be a carcinogenic agent.(5) I t must not be present as any natural impurity in any oil likely to be encountered. (6) It must be reasonably stable over lengthy periods and must not be easily removed from the oils. (7) I t must be detectable in low concentrations by a test which is easy to operate in on-the-spot checks and efficient methods must be available for the determination and positive identification of the markers. Of these conditions the first three reflect industrial requirements the fourth is generally essential and the remainder are necessary for the purposes of official control. Quinizarin and furfuraldehyde have proved to be very effective markers in the hydro- carbon oils most likely to be used as road fuels (gas oil and kerosene). It is in the search for markers for hydrocarbon oils used in the industrial field that the above mentioned criteria are difficult to satisfy.In this context the laboratory has considered many chemicals such as food additives vitamins anti-oxidants and chlorinated hydrocarbons but as yet the “ideal marker” has not been found. The analytical methods used for the determination and identification of quinizarin and furfuraldehyde are described in The A~zaZyst.l?~~~~* In this paper recent developments in the techniques that are used in this laboratory are described. 1. The gas-chromatographic method previously used for the identification of furfuralde- hyde has been replaced by a quicker and more sensitive thin-layer chromatographic techniq~e.~ 2. A supplementary method has been devised for the identification of quinizarin in hydrocarbon oils which is useful when contaminated samples or samples from unusual sources are examined (e.g.rags cloths filters and grass). An aliquot of the sample or of a solvent extract diluted with a light petroleum fraction is passed through a column of sodium carbonate wherein the quinizarin is held near the top of the column presumably in the form of a complex. The column is washed with light petroleum to remove the heavy oils and other contaminating materials and with chloroform to remove the quinizarin. The chloroform is evaporated the quinizarin re-dissolved in cyclohexane and the identification completed as described in the official method. 3. An automatic procedure has been developed for the determination and identification of quinizarin in hydrocarbon oils6 I t has been established that the precision and repro- ducibility of the automatic method is equivalent to that of the manual procedure.The new method should be operative in the near future. Grou$ (iii)-Methods have been devised for the identification and determination of the oil-soluble dyes present in premium burning oils and tractor vaporising oils. Concentration and isoiation of the dyes are achieved by column-adsorption chromatography identification by thin-layer chromatography and determination by absorption chromatography. This type of work is obviously very complex and every hydrocarbon oil sample examined has to be treated individually with examination being in accordance with any of the tests or combination of the tests as described. REFERENCES processing. render the oils to which it is added unsuitable for their intended use.1. 2. 3. 4. 5. 6. Harrison R. B. Palframan J. F. and Rose B. A. Analyst 1961 86 561. Harrison R. B. and Heaysman L. T. Ibid. 1961 86 566. Harrison R. B. Ibid. 1963 88 644. Field K. and Godley E. W. Ibid. 1966 91 287. Soutar N. M. Ibid. 1970 95 in the press. Tucker K. B. E. Sawyer R. and Stockwell P. B. Ibid. 1970 95 in the press. 12 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [Proc. SOC. AnaZyt. Chem. Some Uses of Flask Combustion in the Analysis of Lubricants BY M. LONSDALE (Castvol Lid. Bvacknell Bevks.) SOME applications of the flask combustion procedure as practised in the speaker’s laboratory were described. The first step in the analysis of lubricants and additives is decomposition where the sample is converted into the inorganic state thus liberating the element concerned in a form that can be readily determined.This decomposition is often lengthy and may require great care to avoid losses. Flask combustion which is the decomposition of the sample in an atmosphere of oxygen in a stoppered flask is a convenient way of doing this. Hempll first used this technique in 1892 for the determination of sulphur in coal. He enclosed the sample in platinum gauze suspended it in a 10-litre flask and fired the sample electrically. Little interest was shown in the method until the 1950’s when Roth used it for sulphur and Mikl and Pech2 used it on a semi-micro scale for determining chlorine and bromine. Schoniger3 applied the technique to the micro scale and showed that the halogens sulphur phosphorus and many other elements could be determined by flask combustion with an accuracy comparable with that of other established methods in a fraction of the time.Many papers have appeared since Schoniger’s publications proposing variations of absorbents and methods of finish and extending the technique to include many other elements. These were reviewed by Macdonald in 1961.4 Perusal of the literature will show how versatile flask combustion is in terms of sample size and composition. Most samples received for analysis in this laboratory are mobile liquids and hence sample container material for weighing and combustion purposes has been investigated. Gelatin capsules give rise to high and variable blank values and methyl cellulose capsules are no longer obtainable. Capsules are now made in the laboratory from cellulose wadding.This is formed into cups while wet around a thin glass rod removed from the rod and allowed to air dry. These capsules are suitable for all but the thinnest liquids. For these polythene cups prepared from thin polythene sheet are preferred. Usually between 5 and 80 mg of sample are taken and 500-ml flasks are used. The use of 1 or even 2-litre flasks allows for much greater sample weight when the element sought is at a low level. QUALITATIVE ANALYSIS- In view of the inability of the sodium fusion test to detect some elements present at relatively low levels we are using the methods described by Haslam Hamilton and Squirrel15 in some of our investigations. As these authors state it is possible to arrive a t a semi- quantitative value for some of the elements in question and this we do for sulphur chlorine and phosphorus.Nitrogen is detected as nitrite and the authors state that some nitrite was formed from all the compounds they combusted except from a stabilised diazonium compound. Our investigations with a large number of nitrogen-containing compounds of known com- position have shown this to be so. Compounds containing nitrogen - nitrogen bonds do not yield nitrite unless other forms of nitrogen bonding are also present. This conversion of combined nitrogen into nitrite varies from one type of compound to another. We intend to devote further time to these studies. Some of the uses of flask combustion in this laboratory are as follows. QUANTITATIVE ANALYSIS- The benefits of flask combustion are most clearly seen in the determinations of sulphur phosphorus and chlorine.Szd$h.ur-Some of the methods for sulphur determination are the lamp method high temperature combustion and the oxygen bomb. The determination by flask combustion is rapid and accurate. Dilute hydrogen peroxide is used as absorbent the solution is boiled to reduce the volume and remove excess of peroxide and isopropyl alcohol is added to the cooled solution in the ratio 4 to 1. Two drops of 0.2 per cent. aqueous thorin6 and 1 drop of 0.01 per cent. aqueous methylene blue are added and the solution titrated to the first pink colour with barium perchlorate solution. The use of an E.E.L. Quantitrator prevents any personal error experienced in the detection of this end-point. When metals are present the January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES 13 sample is burned under sodium carbonate and the platinum gauze is lowered into the absorbent solution when absorption is complete.The solution is boiled cooled and passed down a cation- exchange column.6 Phosphorus is removed by adding magnesium carbonate to the solution after absorption and boiling. The cooled filtered solution is passed down the cation-exchange column to sharpen the thorin end-point. Phos$horus-Flask combustion is simpler and more rapid than the volumetric quinoline molybdate method and gives good results down to below 0.1 per cent. The method in use in this laboratory is that of Gedansky Bowen and Milner.’ The absorbent is 10 per cent. sulphuric acid and the sample is burned under sodium carbonate to ensure the formation of orthophosphate.Ammonium molybdate and hydrazine sulphate are added and the resultant molybdenum blue is measured on a photometer. ChZorine-This element is determined after flask combustion by titration of the halide solution with mercury(I1) nitrate. The indicator is diphenylcarbazone and the absorbent is dilute hydrogen peroxide. A determination can be completed in 20 minutes by this method. Metals-Some metals present in additive concentrates such as calcium zinc and mag- nesium are determined after flask combustion by titration with EDTA solution. The absorbent is 10ml of 10 per cent. nitric acid. Metals present at very low levels can be determined by a sensitive photometric finish and selective colour-forming reagents. Thus calcium can be determined with calcichrome zinc with zincon and magnesium with titan yellow.The methods described are simple in operation and apply to a large variety of sample types and determinations. Obviously other workers may prefer absorbents and methods of finish other than those described but this paper is intended to demonstrate how one laboratory makes use of this elegant technique. REFERENCES 1. 2. 3. 4. 5. 6. 7. Hempl W. 2. angew. Chem. 1892 13 393. Mikl O. and Pech J. Chemicke’Listy 1952 46 383. Schoniger W. Mikrochim. Acta 1955 123. Macdonald A. M. G. Analyst 1961 86 3. Haslam J. Hamilton J. B. and Squirrell D. C. M. Ibid. 1961 86 239. Fritz J. S. and Yamamura S. S. Analyt. Chem. 1955 27 1461. Gedansky S. J. Bowen J. W. and Milner 0. I. Ibid. 1960 32 1447. The Determination of Low Concentrations of Metals in Lubricating Oils by Spectrometric Methods BY D.PICKLES AND C. C. WASHBROOK (Ministry of Defence Chemical Inspectorate Harefield Middlesex) BARIUM calcium and zinc derived from additives are encountered in lubricating oils in the range 0.01 to 0.5 per cent. and their determination is necessary for quality-control purposes. Traces of metals corroded and worn from moving surfaces in mechanical environments are found in a finely divided form in used oils in which concentrations can vary from a few parts per million up to 1000 p.p.m. Their determinations by emission and flame-absorp- tion spectrometry are described. EMISSION SPECTROMETRY Determination of trace metals and phosphorus in lubricating oils by emission spectro- metry dates from 1946 when Calkins and Whitel introduced the quenched electrodes sampling technique.This was followed by the porous cup in 194g2 and the rotating graphite disc in 1951 .3 These techniques were initially applied to the determination of additive elements in new engine oils but the rotating graphite disc was also used for determination of traces of metals in used 0ils.~9~ Earlier workers used photographic recording and line density measurement but direct- reading instruments with electronic read-outs were eventually used. 14 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [PYOC. SOC. Analyt. Chem. Inter-element interference has been controlled with magnesium and sodium buffers and Curnmins and Mason6 impregnated their graphite discs with sodium chloride solution instead of adding the buffer to the sample. We used a rotating graphite electrode and nickel internal standard for our work until emission methods for additive elements were abandoned for various reasons in favour of at omic-absorption met hods.ADDITIVE METALS IN NEW OILS- Our experience covers both photographic and direct-reading spectrographs. Hilger and Watts medium dispersion quartz prism instruments have been used throughout this work. In the determination of barium we found the nickel internal standard procedure gave the most satisfactory results on both types of instrument but for the much lower concentrations of calcium and zinc there was no obvious superiority among several variations of the rotating disc technique. We found that the photographic method gave considerably superior results to the direct-reading techniques although the former is more time consuming.This is attributed to the use of different standard lines each in close proximity to the line of the element being determined whereas in the direct-reading method reference is made to one nickel line only. IVEAK METALS I N USED OILS- Used petroleum oils from petrol and diesel engines and synthetic oils from aircraft turbine engines are under current investigation. The rotating graphite disc is applied for both types but the differences in background emission require different types of standards. Standards used for petroleum oils are solutions of metal naphthenates in light viscosity mineral oil. Those for synthetic oils are solutions of metal cyclohexane - butyrates' in a base fluid. The base fluid is a mixture of di-octyl sebacate and polyglycol ether adjusted to give about the same viscosity as the test samples.The instrument currently used is a Hilger and Watts medium dispersion prism spectro- meter with direct-reading attachment transistorised electronics and digital presentation of data by electric typewriter. An intermittent a.c. arc type source units is used with a voltage regulator. The wavelengths (nm) in present use are zinc 213.86 iron 259-94 magnesium 279.08 lead 283.31 silicon 288.16 chromium 301.52 aluminium 308-22 copper 327.40 silver 338.29 and nickel 349.29. Background at 242.74nm is used as the internal standard. A +-inch diameter rotating graphite electrode with 3-mm gap to pointed rod electrode is used. The lower electrode rotates at 4 r.p.m. The electrodes are pre-burned for 30 seconds to remove surface contamination and to render the surface more oil-retentive.The exposure is controlled by the background emission. When this reaches a pre-set value the exposure is automatically terminated. Digital read-out is related to a calibration graph concentration of metal in p.p.m. This graph is prepared daily with the appropriate metal standards. PRECISIOX- Our experience with the photographic method for additive metals indicates a coefficient of variation of 3 per cent. for barium calcium and zinc whereas with the direct-reading method we have found 15 per cent. for barium and 10 per cent. for zinc and calcium. For the direct-reading method for wear metals it is difficult to determine precision data with samples because of their low concentration and variability. Our results indicate the following This takes about 20 seconds.Limit of detection Repeatability (1 a) p.p.m. Metal p.p.m. (determined at 10 p.p.m.) Chromium .. 1 . 3 Aluminium . . .. 2 Lead . . . . .. 2 Silver . . .. .. 1 Iron . . * . .. 1 Magnesium . . .. 2 Copper . .. .. 1 1 2.5 1 1 2 2 1.5 January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES Limits of detection are derived from analysis of standards using the expression 15 d s w h e r e d is the difference between duplicates and n is the number of determinatio~is.~ ATOMIC-ABSORPTION SPECTROSCOPY We have introduced this technique for routine quality control of additive metals in lubricating oils and also for determining wear metals in used oils. The equipment used in our preliminary work is an Evans Electroselenium Model 140 with a scale expansion galvano- meter.ADDITIVE METALS IN NEW OIL- For the additive metals barium calcium and zinc occurring together we used the nitrous oxide - acetylene burner system and minimised inter-element interference by the use of a potassium buffer. Standards are prepared from chemically analysed oil additives dis- solved in white spirit and the buffer is a potassium alkyl sulphonate dissolved in white spirit to give 1 per cent. of potassium. The sample is dissolved in white spirit and buffer solution is added to give 1000 p.p.m. of potassium. A blank solution is used containing mineral oil of similar viscosity to the sample. The wavelengths (nm) used are barium 553.56 calcium 422-67 and zinc 213.86. The slit width varies from 0.05 to 0.20 mm and the lamp current from 15 to 18 mA. Comparison with other results indicates that the atomic-absorption results for barium calcium and zinc are within 3 to 5 per cent.of the chemically obtained data. WEAR METALS- Iron and copper derived from corrosive attack and mechanical wear have been deter- mined in used lubricating oils. These elements are present in samples as a combination of solution of metallic salt and finely divided particles. Sampling is therefore a critical factor and wide variations in test results can be experienced. The method is applicable to both petroleum and synthetic oils. The air - acetylene burner is used for copper and iron. For petroleum oils a white spirit solvent system is used and for synthetic oils isobutyl methyl ketone. Standards are based on solutions of metal naphthenates in di-octyl sebacate. The technique is similar to that for additive metals.The wavelengths (nm) used are iron 284.3 and copper 324.75. The slit width is 0.1 mm and the lamp current is from 10 to 24 mA. Precision data with used oils has not been derived but comparisons with chemical analy- sis and emission spectrometry indicate that the method gives acceptable results. Acknowledgement is made to the Chief Scientist (Army) Ministry of Defence for permission to publish this paper and to Mr. B. Bedford and Mr. P. S. Bromley of the Chemical Inspectorate for experimental work. REFERENCES 1. 2. 3. 4. 5. 6. 7. National Bureau of Standards U.S.S. Monograph No. 54. 8. 9. Calkins L. E. and White M. M. Natw PetroE. News 1946 38 519. Gassmann A. G. and O’Neill W. R. Analyt. Chem. 1949 21 419. Pagliassotti A. G. and Porsche F.IT. Ibid. 1951 23 198 and 1820. Gambrill C. M. Gassmann A. G. and O’Neill W. R. Ibid. 1951 23 1365. Pagliassotti J. P. Appl. Spectrosc. 1955 9 153 Cummins R. A. and Mason P. R. J . Inst. Petrol. 1962 48 237. Kingsbury G. W. J. and McClelland J. A. C. “Photographic and Spectrographic -4nalysis of High Purity Zinc and Zinc .4lloys for Die Casting,” H.M. Stationery Office London 1945. “-4.S.T.M. Suggested Practices for Use of Statistical Methods in Spectrographic A\nalysis,” -4.S.T.M. E 25*%-4 1964. 16 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [Proc. SOC. Afialyt. Chem. Micro-analytial Techniques in the Analysis of Fuels BY F. C. A. KILLER (Esso Research Centre Abingdon Berks.) RECENT developments in analytical techniques of interest to the petroleum analyst have been based on the automation of well established methods or the introduction of new tech- niques.Typical representatives of the automation of established methods are the automatic elemental analysers for carbon hydrogen and nitrogen which based on classical combustion procedures will automatically combust a few milligrams of sample separate the combustion products (carbon dioxide water and nitrogen) by selective absorption or by gas-chromato- graphic separation and obtain catharometer signals on a chart. These signals are converted by comparison with microchemical standards into percentages of carbon hydrogen and nitro- gen. An analysis cycle takes about 10 minutes. Low results are obtained in the analysis of lowboiling liquids such as gasolines or subliming solids for example naphthalene and anthracene because of volatilisation in the open boat before combustion.We have obtained good results on a Perkin-Elmer Model 240 Elemental Analyser by sealing the sample in an aluminium pan such as the one supplied together with a simple sealing device with the Perkin-Elmer DTA apparatus. With pure hydrocarbons or petroleum products containing small amounts of hetero-elements it is sufficient to determine the carbon-to-hydrogen ratio therefore the sample need not even be weighed. Further automation of such instruments is possible and may be advantageous in laboratories with a great demand for this type of analysis. Another possibility for the automation of classical procedures is given with the Technicon AutoAnalyzer which consists of modules that carry out automatically at a rate of up to 70 samples hour-l the operations required in wet-chemical analysis with a spectrophotometric flame-photometric fluorimetric or turbidimetric finish.We have used the AutoAnalyzer for determining phosphorus in gasolines. The sample is evaporated after adding a few drops of a heavier material e.g. tridecanol and the residue decomposed with sulphuric nitric and perchloric acids. The diluted digest solution is transferred to the AutoAnalyzer and the phosphorus content determined spectrophoto- metrically as molybdenum blue with hydrazine sulphate as the reducing agent. Another successful application of the AutoAnalyzer is in determining hydrogen sulphide in hydrocarbon gases. The gas is passed through an absorbing solution of zinc acetate and sodium citrate at a pH of 5.7.Aliquots of this solution are then analysed for hydrogen sulphide by the formation and spectrophotometric determination of methylene blue. The method gives satisfactory results in the range 0.1 to 1.2 pg of hydrogen sulphide per millilitre of absorbing solution With adequate sampling the method can be used for con- tinuous monitoring of hydrogen sulphide in hydrocarbon gases and possibly in air and other gases. An important new technique in the analysis of petroleum products is the Dohrmann microcoulometer. The instrument was originally designed as a selective sulphur detector for gas chromatography. We found that in this form the instrument was not suited for determining total sulphur by straight injection of the sample into the combustion tube because the capacity of the tube was too small resulting in incomplete combustion.We therefore replaced it with a larger diameter tube and adapted the sample inlet for straight injection of gases or liquids. Dohrmann now offer an improved combustion tube suitable for the determination of total sulphur in petroleum products. The amount of liquid sample required for one determination is 1 to 2 p1. Analysis time is about 1 minute and the precision is very good. Halogens and nitrogen compounds when present in the sample are likely to interfere in the titration of sulphur dioxide with iodine. The interference results from the oxidation of nitrogen to nitrogen oxides and halide to halogen each of which liberates iodine from the potassium iodide present in the electrolyte thus producing a negative signal. We have confirmed the suggestion of Bremanis Deering Meade and Keyworthl that these interferences can be eliminated by adding sodium azide to the electrolyte.Sodium azide reacts rapidly with chlorine and oxides of nitrogen to form chloride and molecular nitrogen respectively. With iodine however the azide reacts at an extremely low rate January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES 17 unless catalysed by bivalent sulphur compounds. Compounds of sexavalent sulphur exert no influence on the reaction hence the iodimetric titration of sulphur dioxide is unaffected by the addition of sodium azide. We have established that under our operating conditions about 98 per cent. of the chlorine present in the sample is converted into hydrogen chloride and only 2 per cent. leaves the combustion tube as elemental chlorine.Bromine has a much stronger effect on sulphur determination because only 40 to 50 per cent. of bromine present in the sample is converted into hydrogen bromide the rest leaves the combustion tube as elemental bromine. Sodium azide in the electrolyte has no effect as the reaction with bromine is too slow. This is important as it prevents the use of the microcoulometer for determining sulphur in leaded gasolines which contain dibromoethane as lead scavenger. The concentration range over which the microcoulometer can be successfully applied is 2 to lo00 p.p.m. The precise and accurate determination of sulphur below 2 p.p.m. in gases and low-boiling liquids remains an unresolved problem. The best approach is still the oxy-hydrogen or Wickbold combustion although the precision in the p.p.m.range obtained by this method is far from satisfactory. Other interesting developments in trace sulphur analysis include the flame photometric detector of Brody and Chaney,2 which is based on the fact that in a hydrogen-rich flame sulphur compounds emit a characteristic radiation in the ultraviolet and violet regions of the spectrum. This signal is extremely sensitive so that picogram amounts of sulphur can be detected. Disadvantages to be overcome before the detector can be brought into general use are as follows. 1. The response differs from compound to compound so a separation must be carried out if several sulphur compounds are present in the sample e.g. with petroleum products. Total sulphur could be obtained either by adding the individual sulphur compounds present (having calibrated the detector for each of them) or alternatively by converting all sulphur compounds present into one form either by oxidation or reduction.Either treatment will dilute the sample and therefore decrease the sensitivity by several orders of magnitude. 2. Hydrocarbons have a strong quenching effect on the sulphur signal so the sulphur species must not only be separated one from each other but also from all hydrocarbons present in the sample. Such a separation is feasible in natural gas or liquefied petroleum gas but would be difficult in a naphtha or a gasoline. 3. The response is not a linear function of sulphur concentration so samples must be compared with standards whose concentrations closely bracket the sample concentration.The detector has been used successfully by Gibbons and Goode3 for determining tetra- hydrothiophene and dimethyl sulphide added to natural gas at concentrations of 2 to 4 p.p.m. v/v. We used the detector for determining thiophene in benzene at concentrations ranging from 0.1 to 10 p.p.m. v/v. The detector can be made sensitive to phosphorus compounds as in the same hydrogen- rich flame phosphorus emits a characteristic radiation at 526 nm. Another development in the field of trace sulphur analysis is the Coulson* conductivity detector which consists of a combustion tube and a conductivity cell with water continuously circulating through a bed of ion exchanger. The sample is injected into the combustion tube and the ions formed when the combustion products enter the cell give a signal that is amplified and displayed on a chart.As the water in the conductivity cell is being con- tinuously purified it ensures a strong and reproducible response of the system. The data available show that the sensitivity is comparable with but not better than that obtainable with the microcoulometer. Other combustion products such as carbon dioxide halogens and nitrogen oxides interfere as the detector is non-specific and every ion contributes to the conductivity of the system. The detector can also be used for chlorine and nitrogen. The response of the microcoulometer however is a linear function of concentration and is highly specific for certain ions. The iodine cell will also give a signal for hydrogen sulphide and mercaptans. The microcoulometer can be used in both the oxidative and reduc- tive mode with direct sample injection for total sulphur or combined with a gas-chromato- graphic column for detecting individual sulphur compounds.With an in-line catalytic 18 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES [Proc. SOC. AnaZyt. Chenz. reactor or non-catalytic pyrolysis followed by gas chromatography it has been used by Drushe15 for determining sulphur compound types in heavy petroleum fractions. By replacing the iodine cell with a cell producing silver ions as the titrant the microcoulometer can be used for the determination of halogens. The same cell will also titrate hydrogen sulphide and mercaptans. With another cell which generates hydrogen ions as the titrant the coulometer can be used for the determination of nitrogen at very low levels6 The sample is subjected to hydrogenolysis over a nickel or platinum catalyst at 850” C which converts nitrogen compounds into ammonia which is then titrated in the cell.Acidic components formed such as hydrogen chloride must be removed from the combustion products before entering the cell by absorbers packed with potassium carbonate calcium oxide or magnesium oxide at 450” C i.e. above the temperature of dissociation of ammonium salts. Dohrmann have recently introduced a solid-sampling device which now extends the use of the microcoulometer to solids and non-volatile liquids. However further work is still required to investigate all the possibilities of the micro- coulome t er thoroughly . REFERENCES 1. Bremanis E. Deering J. R. &lea.de C . F. and Keyworth D. X. Bull.Anzer. SOC. Test. LWatev. 2. Brody S. S. and Chaney J. E. J. Gas Chrornat. 1966 2 42. 3. Gibbons P. A. and Goode K. A4. Gas J . 1968 336 27. 4. Coulson D. M. J. Gas Chvornat. 1965 1 134. 5. Drushel H. V. Analyt. Chem. 1969 41 569. 6. McNulty J. L4. and Hoppe L. W. “Submicro Elemental Analysis by Microcoulometry,” Dohr- 1967 7 459. mann Instruments Company Mountain View California 1969. Determination of Trace Metals in Fuels BY J. SCOTT AND F. C. A. KILLER (Esso Research Centre Abingdon Berks.) ATOMIC-ABSORPTION spectrophotometry provides a rapid and accurate method for determining low concentrations of metals in petroleum products. It compares favourably with other methods of elemental analysis its major advantages being relatively low cost of equipment speed of analysis freedom from inter-elemental effects and high sensitivity and specificity.In our laboratory the technique has been applied to the determination of trace metals in fuels and feedstocks; the paper discusses the procedures available and the problems in- volved in the determinations. The elements of interest are lead nickel copper iron zinc van- adium and sodium in products such as naphthas gasolines aviation and diesel fuels and fuel oils. Lead-Lead is determined in naphthas where it may be present at the sub-p.p.m. level and can cause catalyst poisoning or affect storage stability of finished products. The usual procedure involves treatment of the sample with bromine and extraction of lead with dilute nitric acid which is then measured colorirnetrically as the dithizonate. The method is lengthy and avoiding contamination represents a major problem.A procedure has been evolved which consists of the extraction of lead from naphtha with iodine monochloride followed by the determination of the extracted lead by atomic- absorption spectrophotometry. The method avoids time-consuming evaporation steps and the use of any other reagents except the iodine monochloride solution thus reducing contamination. A known volume of naphtha (about 500ml) is extracted once with 10ml of iodine monochloride solution which is then directly aspirated into an air - acetylene flame and the absorbance measured at 217.1 nm. At least 95 per cent. of the lead present was recovered by this single extraction procedure thus confirming the findings of Campbell and M0ss.l The repeatability of results obtained by this procedure is comparable with that of other methods but the method is much faster results can be obtained in less than 1 hour provided reagent and standard solutions are available.Co++er-Copper may be present in petroleum stocks as a contaminant introduced during refining or storage or because of its presence in the original crude oil. Its presence is un- January 19701 MICROCHEMICAL TECHNIQUES IN PETROCHEMICAL INDUSTRIES 19 desirable particularly in gasolines heating fuels and jet fuels as it accelerates oxidative deterioration of the finished products. These products differ greatly in viscosity causing similar variations in nebuliser intake and response. Several organic diluents have been tried to eliminate the viscosity effect. Acetone2 was satisfactory at a solvent-to-sample ratio of 4 1 and heptane (3 -+ 1) and p-xylene have been used.Naphtha samples are aspirated without dilution and the absorbance measured at 324.7 nm. Nickel-Nickel is found in crude oils in the form of metal porphyrins. I t accompanies the hydrocarbons through the refining stages and if present even in sub-p.p.m. amounts in feed stocks creates problems in catalytic cracking because it tends to deposit and poison the catalyst. Nickel is determined essentially in the same way as copper. Naphtha or gasoline samples are aspirated without dilution heavier products require dilution to reduce viscosity. Kerber3 tested solvents such as heptane xylenes cyclohexane alcohols and ketones. 9-Xylene was chosen as the best compromise between solvency and properties in the flame (solvent-to- sample ratio 5 1 for gas oils).Barras4 obtained a precision of 0-05 p.p.m. of nickel and a detection limit of 0.05 p.p.m. of nickel using a method of standard additions and measuring the absorbance at 232.0nm. Zinc-Zinc is sometimes determined in jet fuels in which ash-forming constituents are undesirable and in gasolines where zinc contamination may lead to oxidation instability. The method is similar to that used for copper and nickel but no diluent is required as more rapid aspiration does not give an advantage over the reduction of the signal because of dilution. The resonance line a t 213.8 nm is used. Iron-Iron was determined in gas oils by Barras4 who used a 3 + 1 dilution with heptane. The detection limit was 0.1 p.p.m. of iron and the standard deviation was 0.02 p.p.m.for diluted oils. Sodi~m-Robinson~ determined sodium in gas oils and claimed that the determination of 10 p.p.m. of sodium was free of interference from 0.5 per cent. of potassium or lithium. Flame photometry provides a simple technique for the determination of sodium but suffers from interference from other alkali metals. Vanadium-The vanadium content of fuel oil intended for burning under boilers is important because of corrosion problems and of hazards resulting from the toxic nature of the ash formed. The standard method for vanadium in fuel oils involves charring the oil with sulphuric acid burning off the carbonaceous matter in a muffle furnace and measuring a solution of the ash colorimetrically. The method is lengthy and subject to errors and interferences. Atomic-absorption spectrophotometry offers a method for the direct determination of vanadium in petroleum products with 2-methyl-4-pentanone as the solvent and a solvent- to-sample ratio of 10 1.A nitrous oxide - acetylene flame must be used to overcome chemical interferences. The limit of detection is estimated to be about 0.1 pg of vanadium per millilitre of solution. These methods for the direct determination of metals in fuels by atomic-absorption spectrophotometry are rapid and relatively free from contamination problems. Above the 1 p.p.m. level they are almost always superior to other techniques. However at the sub-p.p.m. level our experience has shown that the sensitivity and precision obtainable is often in- adequate. The problem of determination of metals at this level therefore splits into two separate problems the concentration of metals to make them amenable for determination and the determination of the metals.Trace metals can be concentrated from a gas oil by ashing chromatography or extraction. Dry ashing is not acceptable as metal losses may occur either through entrainment in the smoke or by volatilisation of some metal chelates. The sample is therefore decomposed with sulphuric acid followed by ashing of the charred mass which it is hoped prevents the above-mentioned losses. The ash is dissolved and the solution used for analysis. This method is time consuming as relatively large amounts of sample are treated (100 g or more) and are subject to contamination caused by the introduction of metals with sulphuric acid from other sources particularly of iron and sodium.Collection of trace metals by adsorption chromatography directly from a gas oil is not 20 NOTICE LYroc. SOC. Analyt. Chem. thought to be feasible as the metals may be present in a form in which they would not be quantitatively retained by the adsorbent. Ion-exchange chromatography is an attractive method of collecting traces of metals (from aqueous media). Bergman Ehrhardt Granatelli and Janik6 have proposed a method for determining nickel and vanadium at the p.p.b. level in petroleum by ion-exchange chromatography and X-ray fluorescence spectrometry. They claim a relative error of 5 to 10 per cent. over the range 0.1 to 1.0 p.p.m. Analysis time is 4 hours. This approach offers a significant saving in time compared with the standard procedure and the precision of the method makes it a good choice.Expensive X-ray equipment is required but the procedure could be adapted for atomic-absorption spectrophotometry as the finishing technique. We have tried to avoid incineration of the sample by using extraction to concentrate the metals. The results obtained on artificial blends with 2 N hydrochloric acid as the extractant indicate that at least 90 per cent. of vanadium nickel and iron can be extracted from gas oil. The aqueous phase containing the metals can be aspirated directly into the nitrous oxide - acetylene flame of an atomic-absorption spectrophotometer and the metal content calculated from a standard treated in the same way. There are no interferences from other elements the only problem being contamination from the extractant and from the glassware used.A volume ratio of gas oil to extractant of 25 1 (250 ml of gas oil extracted with 10 ml of hydrochloric acid) brings the metal concentration into the range of the atomic- absorption spectrophotometer. The precision of our method has not yet been firmly established but we believe it to be comparable with that of the ion-exchange X-ray method. Analysis time is less than 1 hour. Further work is required to establish the optimal instrumental conditions and procedures but the development so far shows that atomic-absorption is the best approach to fast and reliable analysis of trace metals in petroleum fuels. REFERENCES 1. 2. 3. 4. 5. 6. Campbell K. and Moss R. J . Inst. Petrol. 1967 53 194. Moore E. J. Milner 0. I. and Glass J. R. Microclzewz. J. 1966 10 148. Kerber J.D. A$$. Spectrosc. 1966 20 212. Barras R. C. Jarrell-Ash Newsletter 1962 13 June. Robinson J. W. Analytica Chim. Acta 1960 23 458. Bergmann J. G. Ehrhardt C. H. Granatelli N. and Janik J. L. Analyt. Chew. 1967,39 1258. Notice Automatic Methods Group MEMBERS’ DEMONSTRATION MEETING APRIL 9th 1970 WE are pleased to inform you that the response to our preliminary enquiry regarding the Members’ Demonstration Meeting at the Middlesex Hospital Medical School on April 9th 1970 was most encouraging and the meeting will therefore go ahead as planned. The usual Notices of Meeting and Registration forms for lunch etc. will be sent out during March. Further applications from private members who wish to demonstrate equipment should be sent IMMEDIATELY to the Group Secretary Mr. R. Sawyer Laboratory of the Government Chemist Cornwall House Stamford Street London S.E.1.

ISSN:0037-9697    

DOI:10.1039/SA9700700004

出版商:RSC    

年代:1970

数据来源: RSC

摘要:

20 NOTICE [ R o c . SOC. Analyt. Chem. Papers Accepted for Publication in The Analyst THE following papers have been accepted for publication in The Analyst and are expected to appear in the near future. “Microcoulometric Determination of Trace Sulphur in Light Petroleum Products,” “Laser Raman Spectroscopy,” by P. J. Hendra and C. J. Vear. “Rapid Determination of Nitrogen-13 and Fluorine-18 in Reactor Cooling Water by An Ion-exchange Method,” by S. Ohno and T. Tsutsui. “Ion-selective Membrane Electrodes,” by E. Pungor and K. Toth. by F. C. A. Killer and K. E. Underhill. THE SOCIETY FOR ANALYTICAL CHEMISTRY “Evaluation of Patton and Reeder’s Dye and Hydroxynaphthol Blue as Metallochromic Indicators in the EDTA Titration of Calcium,” by A. Itoh and K. Ueno. “The Identification of Polyethylene Glycols and Polypropylene Glycols by Thin-layer Chromatography,” by T.Salvage. “Determination of Fluoride by Substoicheiometric Isotope Dilution,” by I. A. Carmichael and J. E. Whitley. “An Atomic-absorption Method for the Determination of Gold and Silver in Plant Liquors and Electrolytes,” by T. P. Michailova and V. A. Respina. “Evaluation and Correction of Interference between Aluminium Silicon and Iron in Atomic-absorption Spectrophotometry,” by A. P. Ferris W. B. Jepson and R. C. Shapland . “The Chemical Analysis of Disodium Octaborate in Preserved Softwoods,” by A. I. Williams. “Potentiometric Titration of Some N aphthylhydrogensulphat es and Naph t hyldihydro- genphosphates,” by V. Chromy and A. Groagova. “Potentiometric Titration of Naphthylcarboxylates in Non-aqueous Solvent,” by V.Chromy and A. Groagova. “The Electrophoretic Separation of Penicillins and Penicilloic Acids,” by A. H. Thomas and R. A. Broadbridge. “The Examination of Organomercury Compounds and their Formulations by Thin- layer Chromatography,” by G. W. Johnston and C. Vickers. “An Investigation of the Determination of Germanium by Flame Photometry and Atomic-fluorescence Spectroscopy with a Separated Nitrous Oxide - Acetylene Flame,” by R. M. Dagnall G. F. Kirkbright T. S. West and R. Wood. “Automatic Titration by Step-wise Addition of Equal Volumes of Titrant. Part 11. An Automatic Titration System,” by A. Johansson and L. Pehrsson. “Field Methods for the Determination of Nitrogen Dioxide in Air,” by A. A. Christie R. G. Lidzey and D. W. F. Radford. “Precipitation of Lead Chromate from Homogeneous Solution by pH Increase,” by J. A. Baynes and P. F. S. Cartwright. “An Improved System for Automatic Amino-acids Analysis,” by S. Jacobs. “Spectrophotometric Determination of Yttrium in Chromium and Chromium Base Alloys with Arsenazo 111,” by D. F. Wood and M. R. Adams.

ISSN:0037-9697    

DOI:10.1039/SA9700700020

出版商:RSC    

年代:1970

数据来源: RSC