2 Physical Methods and Techniques Part (ii) Gas Chromatography ~~~~ ~~~~~ ~ ~~~ By D. A. COWAN Pharmacy Department Chelsea College Manresa Road London S W3 6LX 1 Introduction The intention of this Report is to highlight some of the newest techniques and developments in gas chromatography. A comprehensive review is published in the annual applied review issue of Analytical Chemistry,’ and a bibliography appears four times a ear^-^ with an annual subject index6 in the Journal of Chromatography. Reviews of publications in specific areas of chromatography are published as a part of the Journal of Chromatography and a list covering the period 1959 to 1976 has recently a~peared.~ Among books and reviews of a general nature published recently are a text by Miller8 on ‘Separation Methods in Chemical Analysis’ a compilation of 11 papers’ from a meeting of the American Chemical Society in 1974 entitled ‘New Developments in Separation Methods’ a review by Ettre,” on the development of gas chromatography and a dictionaryof chromatography by Denney,” almost half of which is devoted to gas chromatography.A book covering organic functional group analysis by gas chromatography by Ma and Ladas12 is referenced here but other specific texts will be referred to in later sections. 2 Columns Various methods may be used for assessing the efficiency of a chromatographic column and Ettre13 has discussed the advantages of each. The ideal column will have a sufficiently high separation efficiency and be as short as possible to ensure rapid analysis with low column temperatures thus minimizing the risk of decomposition.The physicochemical processes which occur between the sample and the column will determine whether a sufficient differential rate of migration will occur with a S. P. Cram and R.S. Juvet jun. Analyt. Chem. 1976,48 411R. 2 J. Chromatog. 1976,118 B1. 3 J. Chromatog. 1976 121 B71. J. Chromatog. 1976,124 B145. J. Chromatog. 1976 128 B219. 6 J. Chromatog. 1976 128 B341. J. Chromatog. 1976 127 259. * J. M. Miller ‘Separation Methods in Chemical Analysis’ Wiley New York 1975. E. Grushka ‘New Developments in Separation Methods’ Marcel Dekker. New York 1976. lo L. S. Ettre J. Chromafog.,1975 112 1. l1 R.C. Denney ‘A Dictionary of Chromatography’ Macmillan London 1976.l2 T. S.Ma and A. S. Ladas ‘Organic Functional Group Analysis by Gas Chromatography’ Academic Press London 1976. 13 L. S. Ettre Chromatogruphiu 1975.8. 291. 11 12 D. A. Cowan minimum of solute band br~adening.’~ The factors which affect these processes are therefore continually under investigation and include column diameter and length quantity and type of stationary phase and its particle size column temperature and carrier gas and its pressure and flow rate. Many papers on retention indices continue to appear largely as a result of the research programme initiated by Janik and co-workers in 1964. Examples of subjects discussed are the precision of the determination of the retention indices of alkylbenzenes,” considering the method of time measurement the non-ideality of the carrier gas tile column temperature and the ageing of the column; the use of retention illdiceh in the identification of some organochlorine pesticides;16 the pre-calculanon of retention indices of alkanes on the basis of their molecular structures;l7 the application of Rohrschneider’s and McReynold’s concepts in a discussion of the contribution to the polarity of stationary phases expressed by retention indices;18 the use of topology-information correlation^;'^ and the predic- tion of resolutions from Kovats retention indices as an aid to column selection.” Instrumental contributions to band broadening have been evaluated experimentally by Cram and Glenn,21 and a new equation has been derived from the kinetic viewpoint to describe the effect of the column length on the height equivalent to a theoretical plate (HETP).22 A systematic approach for the optimization of gas chromatographic separations using the simplex method has been described by Morgan and Demi~~g,~~ and more recently by Holderith T6th and Varadi.24 Solid Supports.-Diatomaceous earth or kieselguhr is still the most popular basis for solid supports.Vermiculite from a deposit in the Soviet Union (Kovdor) is to make a support which will improve the selectivity towards aromatic hydrocarbons compared with commonly used bentones. The use of carbon in various forms has attracted quite a lot of attention during the past few A method for coating glass capillary columns with graphite using and its use in packed columns3’ have been described.Carbon supports have also been modified by treatment with various quantities of organic vapour or liquid Di Corcia and co-workers have investigated the analysis of various 14 S. P. Cram and T. H. Glenn jun. J. Chromatog. 1975 112 329. 15 L. Sojlk and J. A. Rijks J. Chromatog. 1976,119 505. 16 F. I. Onuska and M. E. Comba J. Chromatog. 1976,119 385. 17 T. S. Lombosi E. R. Lombosi I. Bernat Zs. Sz. Bernat E. C. Taklcs and J. M. Takacs J. Chromatog. 1976,119,307. 18 G. Tarjln A. Kiss G.,Kocsis S. Meszaros and J. M. Taklcs J. Chromatog. 1976 119 327. 19 J. R. ChrCtien and J.-E. Dubois J. Chromatog. 1976 126 171. 20 S. D. West and R. C. Hall J. Chromatog. Sci. 1976 14 339.21 S. P. Cram and T. H. Glenn jun. J. Chromatog. 1976 119,55. 22 K. Ohzeki T. Karnbara and K. Kodama J. Chromatog. 1976 121 199. 23 S. L. Morgan and S. N. Deming J. Chromatog. 1975,112 267. 24 J. Holderith T. T6th and A. Varadi J. Chromatog. 1976,119 215. 25 V. G. Berezkin and V. S. Gavrichev J. Chromatog. 1976 116 9. 26 V. Patzelovl 0.Kadlec and P. Seidl J. Chromatog. 1974,91 313. 27 T. V. Barmakova A. V. Kiselev and N. V. Kovaleva KofloidZhur. 1974 36 133. 28 T. V. Barmakova A V. Kiselev and N. V. Kovaleva Kolloid Zhur 1974 36 934. 29 F. Bruner G. Bertoni and P. Cicciolo J. Chromatog. 1976 120 307. 30 G. Nota G. C. Goretti M. Armenante and G. Marino J. Chromatog. 1974,95 229. 31 F. Bruner P. Ciccioli and G. Bertoni J. Chromatog. 1974,90 239. 32 A.Di Corcia A Liberti and R. Samperi Analyt. Chem. 1973 45 1228. 33 A. Di Corcia and R. Sarnperi J. Chromatog. 1973 77,277. 34 A. V. Kiselev and K. D. Shcherbakova J. Chromatog. Sci. 1974 12 788. Physical Methods -Part (ii) Gas Chromatography 13 aliphatic and aromatic hydrocarbons using a 2,4,5,7-tetranitrofluorenone-modified graphitized carbon black following their earlier work which used a 2,4,7-trinitrofluorenone-modified stationary phase.36 In addition they have used a graphitized carbon black column partially coated with poly(ethy1ene glycol) 1500for the separation of C and C5 hydrocarbons at 50 0C.37 Other supports include aluminium glass beads,39 and organic materials such as Rohm and Haas XAD-type resins.,' Some applications of molecular sieves in the gas chromatographic analysis of hydrocarbons and alcohols have been described.41 Liquid Phases.-There are many hundreds of liquid phases but the search for a few standard phases continues.Hawkes and co-workers4* have selected five preferred phases 24 phases of secondary preference and 13 phases for special applications. Moff at43 has reviewed the use of SE-30 for the gas chromatographic analysis of some 480 drugs. A recent book4 summarizes the basic terms used in gas chromatography in connection with retention data; it deals with the selectivity of gas chromato- graphic packings and considers the chemical characterization of the main types of phase. A nematic liquid crystal NN'-bis(p-methoxybenzy1idene)-act'-bis-p-toluidine has been used as the stationary phase €or the gas chromatographic separation of naphthalene homologue~,~~ aza-heterocyclic and steroid epimers.,' Di-n-butyltetrachlorophthalate has been used by R~ba,~ for the identification of 78 hydrocarbons.He points out that this stationary phase displays certain features not found in other stationary phases especially in that alkylbenzenes can be disting- uished from alkylcyclohexanes. Polyols on Chromosorb P have been used to separate volatile compounds from excess amounts of ethanol in the analysis of The properties of column packings containing small amounts of stationary phase have been in~estigated,~' and it was shown that uni- and bi-molecular layers of the stationary phase were formed on the support surface.A new set of test compounds has been recommended5' for the determination of the selectivity of liquid phases at temperatures between 250 and 400 "C. The upper temperature limit of various poly(ethy1ene glycols) has been raised to 230°C by 35 A. Di Corcia A. Liberti and R. Samperi J. Chromatog. 1976 122 459. 36 A. Di Corcia R. Samperi and G. Capponi .I. Chromatog. 1976 121 370. 37 A.Di Corcia and R. Samperi J. Chromatog.. 1976 117 199. 38 E. E. Kugucheva Khim. Teckhnol. Topl. Masel 1976 57. 39 L.Zoccolillo and F. Salomoni J. Chromatog. 1975 106 103. 4O D. F.Fritz and R. C. Chang Analyt. Chem. 1974,46 938. 41 M. Long V. Raverdino G. Di Tuilio and L. Tomarchio J. Chromatog. 1976,117 305. 42 S.Hawkes D. Grossman A. Hartkopf T. Isenhour J. Leary and J. Parcher J.Chromatog Sci. 1975 13,115. 43 A. C. Moffat J. Chromatog. 1975 113 69. 44 G.E.Baiulescu and V. A. Hie 'Stationary Phases in Gas Chromatography' Pergamon Press Oxford 1975. 45 S. Wasik and S. Chesler J. Chromatog. 1976 122 451. a 46 M. Pailer and V. Hloiek J. Chromatog. 1976. 128 163. 47 W.L. Zielinski jun. K. Johnston and G. M. Muschik Analyt. Chem. 1976,48,907. 48 M. Ryba J. Chromatog. 1976 123 327. 49 H.Verachtert D. van Oeveen and J. Bevers J. Chromatog. 1976 117 295. 50 A Waksmundzki and J. Rayss J. Chromatog. 1976,119 557. 51 R. D.Schwartz and R. G. Mathews J. Chromatog. 1976,126,113. 14 D.A.Cowan bonding them to silaceous supports by the formation of oxirons in the preparation A hydrocarbon of empirical formula C87H176 has been proposed as a possible standard non-polar stationary phase.53 This paper describes the disadvantages of using either squalane or methylsilicanes as standard phases and explains the design of a hydrocarbon to fulfil as far as possible the following requirements (i) the hydrocarbon should be a pure substance with no chiral centres; (ii) its melting point should be as low as possible (iii) it should be of low viscocity; (iv) the upper temperature limit of its use should be determined by its pyrolytic stability and not by volatility; and (v) the price of the substance should not be prohibitive for its use as a stationary phase.Consideration of these criteria and preliminary investigations led to the synthesis of the hydrocarbon 24,24-diethyl-19,29-dioctadecylheptatetra-contane (C87HI87) which may be used between 30 "Cand at least 250 "C.Grant54 has reviewed the highlights of gas chromatographic column development from a kinetic rather than from a thermodynamic standpoint. He compares the theoretical and practical performances of columns in relation to the distribution and load of the liquid phase and considers the feasibility of using packed and open tubular columns in the optimization of column parameters to achieve the minimum analysis time. Open Tubular Columns.-Discussion still continues on the use of capillary columns for routine gas chromatographic analysis despite the fact that high-efficiency columns were first used successfully more than 15 years ago. In a critical but short review55 the progress in the production and connection of glass capillary columns is surveyed the authors pointing out that the commercial gas chromatographs now available have been developed mainly for packed columns and hence the various components of the instruments must be re-optimized if high-performance analyses are to be undertaken.The typical size of sample per component for a capillary column is about 200 times less than that for a packed column and is in the order of nanograms. Flow splitters are normally used with capillary columns to allow between one-thousandth and one- twentieth part of the sample onto the column. Schomburg and co-worker~~~ have discussed the implications of the restricted sample load and present in their review article the design of an improved splitter.An injector using a pre-column and a pneumatically controlled split point is described in a paper by Hartigan and Ettre56 with the interesting title of 'Questions related to gas chromatographic systems with glass open-tubular columns'. It is generally accepted that capillary columns are invaluable for the analysis of complex mixtures which cannot be separated by the use of packed columns. Capillary columns however are more easily damaged than packed columns and some skill is needed in their preparation. The novice to capillary columns is advised to purchase a commercially prepared column of guaranteed quality before he attempts to prepare his own. Nevertheless many new and simpler techniques have been developed to 52 S.Ho Chang K. M. Gooding and F. E. Regnier J. Chromatog. 1976 120 321. 53 F. Riedo D. Fritz G.Tarjhn and E. Sz. KovLs J. Chromatog. 1976 126 63. s4 D. W. Grant I. Chromatog. 1976 122 107. 55 G. Schomburg R. Dielmann H. Husmann and F. Weeke J. Chromatog. 1976,122 55. 56 M. J. Hartigan and L. S. Ettre J. Chromatog. 1976 119 187. Chysical Methods-Part (ii) Gas Chromatography 15 enable the preparation of laboratory-made columns. Grob and GrobS7 have discus- sed some of the factors which influence the successful preparation of glass capillary columns. Indeed they recommend that individuals prepare their own columns on psychological grounds claiming that the analyst who can rely on his own ability to replace a ruined column has far better pre-conditions for acquiring a correct technique in their use.The liquid phase must be spread homogeneously and permanently over the walls of the capillary column and this spreading will depend on two independent factors. First the liquid film may be stabilized by intermolecular forces that attract the liquid to the solid as occurs when untreated glass is coated with apolar polysiloxane phases. The second stabilizing factor is known as micro-roughness in which the surface of the column is roughened to a degree which is of similar magnitude to the thickness of the liquid film. Usually both factors will operate to some extent at the same time although the predominance of the former is likely to produce higher separation efhcienciesss because of the more homogeneous liquid film produced.Unfortu- nately more polar phases require a higher degree of roughening since the surface tension of these phases tends to decrease their ability to wet the glass surface. The glass surface may be modified by etching or by the growth of crystals or by both. Gaseous hydrogen chloride may be used to etch soft glass when a layer of sodium chloride crystals is simultaneously pr~duced.~~,~* Repeating the etching process has enabled even the most polar phases to be successfully coated.61 However columns made from soda glass tend to be alkaline and the crystal growth depends on the glass structure which makes the production of a standard degree of micro-roughness ~ncertain.~’ In addition the solubility of the crystals in polar liquids may be a limitation.Gaseous hydrogen fluoride has also been used6* for the modification of the surface of both soft soda-lime glass and sodium borosilicate (Pyrex)glass. Sodium chloride crystals may be deposited on the glass surface by coating with a solution of sodium chloride and the size and number of crystals deposited when solutions of different strengths are used has been in~estigated.~~ Barium carbonate is a useful interlayer as it is easily used and is insoluble in water. Although the micro-crystals of barium carbonate produced on borosilicate glass differ from those on soda glass the variations usually observed within a selected type of glass can be neglected. Dehydration of the glass during the drawing of the tubes is claimed to produce an oxide-type glass surface which is completely wettable by non-polar and by most polar liquid phases without further treatment.This procedure results in high efficiency columns of 3000 to 10 000 theoretical plates per metre. The stationary phase may be bonded to the glass as has been demon~trated~~ using octadecyltrich- lorosilane to form an octadecylsilane bonded to the capillary wall or by using a base-catalysed reaction of a siloxane polymeric mixture with the previously etched 57 K. Grob and G. Grob J. Chromatog. 1976 125 471. 58 L.Blomberg J. Chromatog. 1975 115 365. s9 J. J. Franken G. A. F. M. Rutten and J. A. Riiks J. Chromatog. 1976 126 117. 6o J. KrupEik M. Kristin M. ValachoviCova and S. Janiga J. Chromatog. 1976,126 147. 61 J.L.Marshal and D. A. Parker J. Chromatog. 1976,122,425. F.I. Onuska and M. E. Comba J. Chromatog. 1976,126 133. 63 C.Watanabe and H. Tomita J. Chromatog. 1976 121 1. 64 J. Simon and L. Szepesy J. Chromatog. 1976,119,495. 6s R. G.Einig and J. L. MacDonald Analyt. Chem. 1976,48 2281. 16 D.A. Cowan glass surface.66 Various other means of stabilization of the stationary phase include the addition of a polymer which has an improved ability to wet glass,67 the addition of a surfactant,68 first reported in 1962 or the coating of the stationary phase together with a finely powdered Metal capillaries may also be used and the importance of the initial cleaning process of stainless-steeI capillaries on their adsorption properties has been st~died.~’ Tapeworm columns formed by flattening a 1mm internal diameter tube in a ‘micro-mangle’ are claimed71 to combine high separating power with high capacity.3 Detectors A book by SevEik7’ covers all the main types of detector in use today. His systematic treatment of the subject describes not only the function but also the optimal working conditions of the devices including gas flow effect of additives in the gases and the temperature. The use of selective detectors in gas chromatography has been reviewed with a bias towards their application to forensic science,73 and ~~~ A u has reviewed detectors for use in the gas chromatographic analysis of pesticides. Flame Ionization.-The fundamental treatment by Blades75 on the mechanism of ion formation has been added to by Sevc‘ik and co-worker~.~~*~’ The normal mixing of the eluate and hydrogen after the gas chromatograph column gives a constant but temperature-dependent response related to the methylene groups of organic com- pounds.The ionization efficiency of flame ionization detection has been shown to increase rapidly with the number of CH functional groups in the organic sample and the proportion of methyl radicals is correlated with the resulting detector current. A semi-specific flame ionization detector (FID) called an ‘energized flame ionization detector’ (EFID) has been designed77 and is based on a pre-reaction zone containing a heated hydrogen path leading to a FID. The output of the EFID and the signal from a normal-mode FID are measured simultaneously both detectors being connected in parallel to the outlet of the gas chromatograph column.The EFID signal depends on sample cracking leading to radical formation and in the presence of electronegative groups or atoms to the higher probability of recombination. Branched alcohols were found to produce a greater signal than straight-chain alcohols owing to the greater number of CH groups per molecule. The technique shows promise for making qualitative identifications. 66 C. Madani E. M. Charnbaz M. Rigaud J. Durand. and P. Chebroux J. Chromatog. 1976,126 161. 67 A. L. Gordon P. J. Taylor and F. W. Harris J. Chromafog.Sci. 1976 14 428. 68 F. Farre-Rius J. Henniker and G. Guichon Nature 1962 196 63. 69 R. S. Deelder J. J. M. Ramaekers J. H.M. Van Den Berf and M. L. Wetzels J. Chromatog. 1976,119 99. 70 M. Ryba J. Chromatog. 1976 123 317. 71 H._D. Papendick and J. Baudisch J. Chromatog. 1976,122 443. 72 J. SevEik ‘Detectors in Gas Chromatography’ Journal of Chromatography Library Vol. 4 Elsevier New York 1976. 73 J. F. Taylor Roc. Analyt. Div. Chem. SOC.,1976 13 168. 74 W. A. Aue J. Chromatog. Sci. 1975,13 329. 75 A.,T. Blades J. Chromafog.Sci. 1973 11 251. 76 J. SevEik and M. Klima Chromatogruphia 1976 9 69. 77 J. SevEik R. E. Kaiser and R. Rieder J. Chromatog. 1976,126 263. S Physical Methods-Part (ii) Gas Chromatography 17 A flame ionization detector using a negatively polarized collector electrode mounted high above the flame burning in a mixture of hydrogen and air to which a small amount of silane is added has been to respond well to organometallic compounds of aluminium chromium iron lead and tin.As little as 6 X g of one test compound aluminium hexafluoroacetylacetonate could be detected and the minimum detectable amount of a typical hydrocarbon standard tetradecane was only 4 X g. It appears that this detector will be suitable for determining certain trace metal organics contained in complex mixtures. The use of ammonia instead of hydrogen for flame formation in a FID increases the sensitivity of detecting chlorinated hydrocarbon^.^^ In addition the low viscosity of ammonia as carrier gas especially in long packed columns makes it possible to obtain high velocities using a decreased pressure gradient. A new thermionic detector has been described" by Verga and Poy which uses a potassium chloride pellet as the alkali source and which has high sensitivity and variable selectivity towards nitrogen and phosphorus.The alkali source may be positioned over the flame to produce the maximum ionization current or lowered beneath the flame to allow normal flame ionization detection. The polarization electrode may also be moved to alter the electric field; movement towards the flame increases the standing current and gives an increased phosphorus/nitrogen signal ratio. Conversely movement of the electrode away from the flame reduces the standing current and improves the selectivity towards nitrogen. The detector is claimed to be 50-100 times more sensitive to nitrogen compounds and 500-1000 times more sensitive to phosphorus compounds than the conventional FID.The Optimization of a three-electrode thermionic detector by adjustment of the hydrogen flow rate may allow increased selectivity between nitrogen and Bleeding 2,6-dinitrotoluene onto the column from a 7 cm 100 pI syringe needle has been used8* to simplify the optimization of a thermionic detector towards nitrogen. Brazhnikov and Shmide183 have proposed a mechanism for the processes that occur in the thermionic ionization detector. Alkali-metal atoms are formed in the volume of the flame not on the salt surface by reaction of the vapours of alkali-metal salts with hydrogen. The metal atoms are excited in the flame and collide with other atoms to produce metal ions.Aerosols of alkali-metal particles are formed near the collector electrode because of its lower temperature and the background current is probably due to thermoemission of the aerosols since they absorb radiant energy from the flame. When organic compounds containing phosphorus or nitrogen enter the flame a useful thermionic detector signal appears which may be explained by consideration of the following facts (i) alkali-metal salts are active inhibitors of combustion; (ii) heavy ions of low mobility are formed by combination of alkali-metal ions with those produced by the combustion of phosphorus- or nitrogren- containing organic compounds; (iii) a reducing pyrolysis of phosphorus-and nitrogen-containing organic compounds resulting in the formation of hydrocarbon radicals may occur.The flame temperature is substantially reduced by the presence 78 H. H. Hill jun. and W. A. Aue J. Chromatog. 1976 122,515. 79 V. G. Berezkin and L. A. Shkolina J. Chromatog. 1976 119,33. G. R. Verga and F. Poy J. Chromatog. 1976 116,17. 81 N. Mellor J. Chromatog. 1976 123,396. ** A. T. Chamberlain J. Chromatog. 1976 116,180. 83 V.V. Brazhnikov and E. B. Shmidel J. Chromafog.,1976 122,527. 18 D.A. Cowan of alkali-metal ions and the efficiency of ionization described above depends on the temperature that is established. When phosphorus- or nitrogen-containing organic compounds enter the flame the temperature of the flame increases because the concentration of alkali-metal salt in the flame is reduced as a result of heavy-ion formation.In addition combustion of hydrocarbon radicals formed will raise the flame temperature and this will lead to an increase in the efficiency of the ionization of the alkali-metal atoms and to an increase in the thermoemission of aerosol particles and hence to the appearance of the useful detector signal. The design of a new dete~tor,'~ the sensitivity of which is not so dependent on the flow-rates of the gases as other designs was also discussed by Brazhnikov and Smidele3 and a comparison made with other thermionic detectors. Flame Photometry.-The flame photometric detector has been extensively used for the selective detection of volatile sulphur- and phosphorus-containing compounds in the atmosphere pesticides urinary volatile sulphur-containing compounds and sulphur gas in soil atmospheres (see paper by Bl~mberg'~ for list of references).The problems encountered in the use of the flame photometric detector for gas chromatographic analysis include (i) extinguishing of the flame when the solvent peak elutes; (ii) recorder baseline drift due to thermal instability of the photomultip- liers; (iii) damping of luminescence by the presence of other compounds which may lead to a loss of specificity; (iv) small dynamic range for sulphur-containing com- pounds; (v) the optimum flow-rates of feed gases are different for detecting phosphorus- and sulphur-containing compounds; (vi) sulphur produces interference signals on the phosphorus channel. Several designs of detector have been publi~hed~~-*~ which do not suffer the problem of the flame extinguishing when using liquid samples of as much as 50 p1by passing the column effluent to the detector by means of the hydrogen line instead of the oxidant line.The design of Hasinski86 uses a thermal insulation plate between the luminescence and burner parts enabling changes to be made to the geometric arrangement of the system to optimize the signal-to-noise ratio the background noise with the flame on the detector sensitivity to carbon-containing compounds and the independent thermal control of the detector system to maintain the dark current of the photomultipliers at a constant level. The detector of Joonson and LOO^^^ overcomes the problem of the dependence of the detector sensitivity on changes in the gas flow.In addition the optimum flow-rates for phosphorus- and sulphur-containing compounds are equal and this enables the reduction of the interference to the phosphorus signals from the sulphur signals by electrical compen- sation. Blomberg" uses ice-cooled acetone to keep the photomultiplier at 4"Cto reduce the detector noise and he has successfully used his apparatus with capillary columns to analyse the gas phase of fresh tobacco smoke. The response of the Pye flame photometric detector has been optimized88 and the detection limits for some phosphorus- and sulphur-containing insecticides have been determined. 84 E. B. Shmidel L. I. Kalabina E. N. Vorona and K. I. Sakodynskii Bull. Isobret. 1975,No.30 123. 85 L. Blomberg J.Chromatog. 1976 125 389. 86 S.Hasinski J. Chromatog. 1976 119 207. 87 V. A. Joonson and E. P. Loog J. Chromatog. 1976,120 285. 88 R.Greenhalgh and M. A. Wilson J. Chromatog. 1976 128 157. Physical Methods-Part (ii) Gas Chromatography 19 Electron Capture.-The effect of the nitrogen carrier gas pressure on the response of a commercial direct current nickel-63 electron capture detector has been measuredSg using 2,4,5,6-tetrachloronitrobenzenein the pressure range 1-5 atm by making the detector leak-tight and by the connection of a pressure gauge and an outlet valve. The response increased linearly with the cell pressure in the range 2-5 atm and as little as 20 fg of the test compound was easily detected at the higher pressures. The response under the normal operating conditions of 1atm was found to be less dependent on pressure.The advantages of iron-55 as an Auger electron emitter over conventional sources have been rep~rted.~' Signal fluctuations which determine the smallest useable signal in electron capture detectors are assumed to be as a consequence of the statistical nature of radioactive decay. Thus the p -emitting nuclides tritium and nickel-63 are generally used as the radiation source as a-emitters are considered too 'noisy' and y -emitting sources would cause radiological exposure problems at the strength required to give suitable ion currents. The current flowing at the detector anode is made up of a series of random pulses which varies as a function of the square root of the number of ionizing particles (N)emitted in a chosen time interval and the number of ion pairs (A) generated by each ionizing particle.Low noise for a given current is favoured by making N as large as possible and A as small as possible. A source giving a low ion current is required for the pulse feedback method of electron capture detection to achieve a linear response characteristic. The orbital electron which may be emitted in extra-nuclear re-adjustments which follow radioactivity decay by orbital electron capture is known as an Auger electron. The Auger yield of iron-55 the number of ions generated by each particle the half-life and the specific activity are all satisfactory for its use as the source in an electron capture detector. A source using iron-55 has been devised which is stable to moisture and to heat up to 4OO0C which is above the temperature to which these detectors are normally subjected.A 5mCi iron-55 source gave a better signal-to-noise ratio than that obtained for other more conventional sources. An atmospheric pressure ionizer using a nickel-63 source for mass spectrometry has been usedg1 to study the formation of ions as well as electrons produced in what is essentially an electron capture detector. This technique has demonstrated that in the electron capture detector the positive and the negative charge densities are numerically equal in the active volume and that the charged particle density is relatively insensitive to whether the negative charge consists of electrons or negative ions.It has also been shown that the differential and ultimate sensitivity and the dynamic range of the electron capture detector is strongly affected by the nature of the positive ions present. It is suggested that atmospheric pressure ionization coupled to the mass spectrometer allows the systematic determination of the operating conditions for specific sample types. MassSpectrometer.-The mass spectrometer is perhaps the best gas chromatograph detector at present available for selectivity and sensitivity. The mass spectrometer analyser and generally its ion source must operate at very low pressures often less 89 S. Kapila and W. A. Aue J. Chromatog. 1976 118 233. 90 D. J. Dwight E. A. Lorch and J. E. Lovelock,J. Chromatog.. 1976 116 257.91 M. W.Siege1 and M. C. McKeown J. Chromatog. 1976,122 397. 20 D. A. Cowan than 1x Torr and since the gas chromatograph column outlet is usually at atmospheric pressure a coupling system which will produce the necessary pressure reduction without too much loss of eluted compound must be used. Various interfacing devices are available but with the use of faster pumping systems on modern mass spectrometers and the small volume flow-rate of capillary columns direct coupling to the ion sources becomes a possibility. A recent designg2 uses such a direct coupling together with a device to enable the removal of the large volume of the eluted solvent peak to prevent pressure build-up in the ion source. Ryhageg3 has described the design of a mass spectrometer ion source which may be switched rapidly between the electron impact and chemical ionization modes enabling a mass spectrum to be recorded in each mode for an eluted gas chromatog- raph peak over the mass range of m/e 5 to m/e 500 in less than nine seconds.A simple modification is described94 to allow the dual operation of a packed and a capillary column with switching of either column outlet to the mass spectrometer. This modification is intended to reduce the need to change columns when packed columns are required and removes the limitations of sample injection volume associated with capillary columns and stream splitting. Palmer and Holmstedtgs have described briefly the principle of selective ion monitoring (mass fragmentography) and provided a list of 319 references on the technique and its use as a selective and sensitive detection system for quantitative gas chromatographic analysis.The use of glass capillary columns with the mass spec- trometer as the selective ion monitor for the quantitative analysis of trace compo- nents in extracts of biological fluids has been de~cribed.’~ The review by Burlingame Kimble and Derrick” covers all aspects of mass spectrometry as well as aspects relating to gas chromatograph linked systems. A minicomputer-based method has been described98 which will produce mass spectra free from neighbouring compo- nent contributions. The systematic errors which may occur when isotope ratios are measured using gas chromatography coupled mass spectrometry have been discus- ~ed.~~ It was concluded that unidirectional scanning is superior to bidirectional scanning and that a bias of less than one per cent may be obtained by its use for ten or more cycles on any given gas chromatographic peak.Radio-isotope.-A very sensitive flow-through proportional counting tube is now available for gas chromatography. Some of the theoretical and practical aspects of the use of this equipment have been discussed,’oo in particular the calculation of specific activities of gas chromatograph peaks the optimization of the tube voltage using external radiation sources and the effect of the composition of the counting gas. Under otherwise identical conditions the width of the gas chromatographic peak has no effect on the count number and thus the activities of narrow and broad peaks can be evaluated equally well.92 F. A. Thome and G. W. Young Analyt. Chem. 1976,48,1423. 93 R. Ryhage Analyt. Chem. 1976,48 1829. 94 L.Kazyak Analyt. Chem. 1976,48 1826. 95 L.Palmtr and B. Holmstedt Sci. Tools 1975 22 25. 96 J. Eyem Sci Tools 1976,23,43. 97 A.L.Burlingame B. J. Kimble and P. J. Derrick Analyt. Chem. 1976,48,368R. 98 R. G. Dromey M. J. Stefik T. C. Rindfleisch and A. M. Duffield Anafyr. Chem. 1976,48 1368. 99 D.E.Matthews and J. M. Hayes Analyt. Chem. 1976,48 1375. 100 I. Kiricsi K. Varga and P. Fejes J. Chromatog. 1976 123 279. Physical Methods-Part (ii ) Gas Chromatography 21 Electrochemical.-The Hall"' microelectrolytic conductivity detector surpasses the earlier Coulson'02 detector in sensitivity.A valve has been to vent the solvents away from a Hall conductivity detector thus significantly reducing solvent tailing peak masking and detector contamination. The influence of furnace temper- ature flow-rate of water through the conductivity cell and mode of operation (pyrolytic or reductive) has been investigated."* The use of the detector in the pyrolysis mode for N-nitrosamines gave a specificity lo7times greater than that for hexane and a detection limit of 50 pg with a linear response up to at least 200 ng. Miscellaneous.-A piezo-electric with an operation range of 25-100 "C will it is suggested increase the scope of analytical capabilities over the normal device despite a reduction in response with an increase in the temperature.Some further comments have been made by Novhk Guha and JanBklo6 following their earlier work'" on the effect of columnn temperatures on the sensitivity of the katharometer detector. Although the effect of katharometer temperature on peak area was found to be insignificant the column temperature does have an effect on the response which has not been explained. A gas chromatograph using glass capillary columns has been coupled'o8 to a multiple detector arrangement consisting of a flame ionization detector a flame photometric detector in the sulphur mode a thermionic ionization detector for nitrogen and a sniffing port for the sensoric evaluation of column effluent. This exotic arrangement has been used for the analysis of flavours including that obtained from roasted meat.4 Pyrolysis Gas Chromatography The pyrolysis of samples and subsequent gas chromatographic separation of the products continue to find new applications. A regulated temperature pyrolyser has been describedlog which uses a filament whose resistance is a linear function of the temperature and which reaches the required temperature in less than 5 ms with very good reproducibility. This pyrolyser has been used to study epoxy polymers. Other applications include the determination of alkylene oxides in their co-polymers,"o the characterization of bitumens,"' and the investigation of some of van Meegeren's faked Vermeers and Pieter de Hooghs."' 5 Derivatization Without the preparation of suitable derivatives the gas chromatographic analysis of many compounds would not be possible because of their low volatility or thermola- 101 R.C. Hall J. Chromatog. Sci. 1974 12 152. lo2 D. M. Coulson J. Gas Chromatog. 1965 3 124. 103 J. MacDonald and J. W. King J. Chromatog. 1976,124 364. '04 E. von Rappard G. Eisenbrand and R. Preussrnann,J. Chromatog. 1976,124 247. lo5 F. W. Karasek P. Guy H. H. Hill jun. and J. M. Tiernay J. Chromatog. 1976,124 179. 106 J. Novak 0.K. Guha and J. Janik J. Chromatog. 1976 123 497. 10' J. Novik 0.K. Guha and J. Janlk J; Chromatog. 1975 112 365. 108 M. HiivniE W. Frischknecht and L. Cechovi Analyt. Chem. 1976,48 937. Io9 C. Waysrnan D. Matelin and C. L. Duc J. Chromatog. 1976 118 115. ll0 I. Zernan L. Novik L. Mitter J. Stckla and 0.Holendova J.Chromatogr. 1976 119 581. 111 Z. Ramljak D. Deur-Siftar and A. Solc J. Chromatog. 1976,119 445. R. Breek and W. Froentjes Sad. Consem. 1975 20 183. D.A. Cowan bility. The general principles involved in the preparation of derivatives and the application of derivatives in individual groups of compounds have been reviewed by Dr0~d.l'~ The consultation of this excellent paper with its 604 references is highly recommended before any derivatization work is undertaken. The preparation of derivatives should also be considered for improving the separation of closely related compounds or to increase the detector response; for example chlorinated and fluorinated derivatives usually yield very large responses from the electron capture dete~tor."~ However despite its usefulness derivatization is not a panacea.For instance low volatility caused as the result of intermolecular cohesion from disper- sion forces because the substance has a large molecule obviously cannot be increased by derivatization. The use of derivatization in the gas chromatographic analysis of pharmaceutical compounds has been reviewed.' '5*116 Recent examples of derivatization methods and their applications include the formation of N-isobutyloxycarbonylaminoacid methyl esters of protein amino-acids (except arginine) by simple treatment with isobutylchloroformatein aqueous medium117 and the separation in 35 min of the 20 protein amino-acids as their N-trifluoroacetyl n-butyl derivatives.' l8 J. Drozd J. Chromatog.1975,113 303. 114 J. E. Lovelock Nature 1961,189 729. llS S. Ahuja J. Phamt. Sci. 1976,65 163. J. Nicholson Proc. Analyr. Diu. Chem. Soc. 1976 13 16. 117 M. Makita S. Yamamoto and M. K6no J. Chromatog. 1976 120 129. L.-A. Appelqvist and B. M. Nair J. Chromatog. 1976 124 239.