2 Physical Methods and Techniques Part (iv) Chromatography (a) High-pressure Liquid Chromatography By J. P. LEPPARD and E. REID Wolfson Rioanalytical Centre University of Surrey Guildford GU2 5XH 1 Introduction Terminology is in a state of turmoil. ‘HSLC’ (high-speed liquid chromatography) seems a misnomer where as practised by the pioneer group at Oak Ridge National Laboratory runs may occupy nearly 24 hours in the interests ofgood ion-exchange separation (e.g. refs 1 and 2). Whilst the term ‘HPLC’ is widely accepted some practitioners regard the ‘P’ as an abbreviation not for ‘pressure’ -which can indeed be low with some systems even 40p.s.i. -but for ‘performance’. This as a purist might object pre-supposes optimization of the conditions for a particular separation (e.g.refs 1 and 2). Whilst the term ‘HPLC’ is widely accepted some manufacturers’ literature and even in research papers but is ambiguous :it could wrongly imply demise of the old-established practice of running columns under gravity with a coarse packing where this gives adequate speed and resolution. Indeed with packings such as dextran gels HPLC is precluded because of blockage under pressure. During 1973 and 1974 this being the period covered by the present survey general introductions to HPLC have appeared in several books3 and including reviews that cater for particular interests such as petroleum,’ pharma- ceuticals,8 and nucleic acids.g Reviews of ion-exchange methods allude to HPLC,’O and recent issues of two journals” are rich in HPLC reviews and ’ C.D. Scott D. D. Chilcote S. Katz and W. W. Pitt jun. J. Chromatog. Sci.,1973 11 96. J. E. Mrochek S. Katz W. H. Christie andS. R. Dinsmore Clin. Chem. 1974,20,1086. P. R. Brown ‘High Pressure Liquid Chromatography Biochemical and Biomedical Applications’ Academic Press New York 1973; L. R. Snyder and J. J. Kirkland ‘Introduction to Modern Liquid chromatography’ Wiley New York 1974. J. H. Knox Ann. Rea. Phys. Chem. 1973 24 29. G. E. Zweig and J. Sherma Analyt. Chem. 1974 46 73R (see 78R). J. H. Knox Chem. and Ind. 1975,29. ’ D. L. Camin and A. J. Raymond J. Chromatog. Sci.,1973 11 625. A. F. Michaelis D. W. Cornish and R. Vivilecchia J. Pharm. Sci. 1973 62 1399. C. Horvath Methods of Biochemical Analysis 1973 21 79.lo P. Jandera and J. ChurhEek J. Chromatog. 1974 98 1 55; H. F. Walton Analyt. Chem. 1974 46 398R. 44 Physical Methods-Part (iva) High-pressure Liquid Chromatography 45 original papers. Reviews of apparatus12 may be helpful to uninitiated readers as may a surveyi3 of theory and capabilities in relation to improved packing materials. Developments in packing^:.'^ including chemically bonded stationary phases,I5 have in fact been a notable feature of recent HPLC technology. 2 Column Packings The following categories as used below (with abbreyiations as naw given parenthetically) in a tabulation of illustrative separations are somewhat arbitrary but will serve as a reminder of alternative separatory principles. (a) Steri’c exclusion (‘Steric’) also termed ‘Molecular size exclusion’ or inappropriately in the case of porous glass beads,16 ‘Gel permeation’.(b) Adsorption (‘Adsorb’) also termed ‘Liquid-solid,. (c) Partition (‘Partit’) also termed ‘Liquid-liquid’. The further term ‘Reversed phase‘ (‘RP’)is applied to this and the following category if necessarily or option- ally it is the mobile phase rather than the stationary phase that is aqueous (polar). (d) Bonded phase (‘Bonded’) where a compound such as an alkyl silicone is chemically bonded on to the surface groups of the particles ;this category is really a sub-category of (c) or possibly (b),unless in (e). (e) Anion-exchange (‘AX’) or Cation-exchange (‘CX’) usually made by bond- ing. As is considered below packings within most of these categories are now available in a ‘pellicular’ (also termed ‘superficially porous’ or ‘porous layer’) form as an alternative to the (totally) porous form.A range of particle sizes down to ‘micro-particulate’ may also be available as may a choice between spherical and irregular particles the irregularity being without adverse effect on separ- ations.13 Unfortunately for readers whose memory for catalogue pages is poor journal editors often allow authors to use trade names for packings with no reminder of their nature. A useful list is to be found in an informative survey of column packings. l4 Pellicular packings have a thin ‘skin’ of the chromatographically active material coated on the outside of hard inert spheres of diameter only ca.40 pm. Comparable efficiencies are obtainable with totally porous packings (usually non-spherical)-only if their apparent diameter is much smaller commonly < 10pm. For both types of packing a narrow particle size distribution is essential for the highest efficiency Pellicular packings allow of faster speeds for given pressures.* Hansen’s prediction’ that the optimal pellicle thickness is 5-10 ”/ I J. Chromatog. Sci. 1974 12,425 (to end of Volume); J. Chromatog. 1973,83 I et seq.; ibid. 1974 99 13 et seq. l2 C. D. Chandler and H. M. McNair J. Chromatog. Sci. 1973 11,468; H. M. McNair ibid. 1974 12 425; J. H. Knox Lab. Practice. 1974 23 505. l3 G. R. Laird J. Jurand and J. H. Knox Proc. SOC.Analyt. Chem. 1974 11 310. l4 R. E. Leitch and J.J. DeStefano J. Chromnrog. Sci. 1973 11 105; cf. R. E. Majors Proc. Analyt. Diu. Chem. Soc. 1975 12 25. A. Pryde J. Chromatog. Sci. 1974 12 486. l6 ’’ M. J. Telepchak J. Chromatog. 1973 83 125. L. C. Hansen and T. W. Gilbert J. Chromarog. Sci. 1974. 12 464. 46 J. P. Leppard and E. Reid of the particle diameter has been borne out by tests on polystyrene-divinylbenzene beads having varying thicknesses of ion-exchange coating. Whilst on theoretical grounds the column efficiency should increase with diminution of particle diameter investigation has suggested that factors such as detector dead volume and recorder response time will limit the usefulness of diameters below 5pm.I’ This study was performed with specially prepared totally porous silica spheres of closely controlled diameter a form of packing known to give high column efficiencies.20 Much work with porous silica gel makes use of irregularly shaped micro-particles and the effects of various activation and deactivation steps upon these have been examined.21 Silica gel packings from different commercial sources have been compared.2 Impregnation of silica gel with cadmium ions has given improved selectivity in the separation of various chloroaniline and toluidine isomer^.^ Moreover pellicular silica gel has beqn impregnated with dimeric rhodium(i1) acetate for the separation of alkenes through complex formation.24 Whilst adsorption chromatography is conventionally performed with materials such as silica gel alumina,13 or polyamide various other materials have been investigated.The porous acrylic resin XAD-7 has been used to separate phenols.25 Compounds such as phenols and organic bases have likewise been separated with essentially aqueous eluents by use of the porous polystyrene-divinylbenzene resin XAD-2 crushed and sieved to give particle sizes down to 45 pm ;the HETP values were 3 mm or poorer and the separation mode was apparently adsorp- tive.26 Such material may however behave like a reversed-phase partition packing. A packing coated with Bentone-34 an organo-clay derivative has enabled various phenols and polycyclic aromatic hydrocarbons to be separated.28 Engelhardt29 has described how to apply a heavy coat of liquid phase to small- particle adsorbents in situ as may also occur inadvertently through adsorption of water from chromatographic solvents used for separation of non-polar compounds on silica gel.Rigid and flexible foams based on polymers such as phenol-formaldehyde polyurethane and polyethylene have been described for use in ion-exchange ’’ L. C. Hansen and T. W. Gilbert J. Chromatog. Sci. 1974 12,458. R. Endele I. HalBsz and K. Unger J. Chromatog. 1974 99 377. *O J. J. Kirkland J. Chromatog. 1973,83 149; K. Unger J. Schick-Kalb and K.-F. Krebs ibid. p. 5. I. I. Frolov R. G. Vorobyoeva I. V. Mironova A. Z. Chernov and Ya. I. Yashin J. Chromatog. 1973 80 167. 22 R. P. W. Scott and P. Kucera J. Chromatog. Sci. 1974 12 473. 23 D. Kunzru and R. W. Frei J. Chromatog. Sci. 1974 12 191. 24 F. Mikes V.Schurig and E Gil-Av J. Chromatog. 1973 83 91. 25 J. S. Fritz and R. B. Willis J. Chromatog. 1973 79 107. 26 C. H. Chu and D. J. Pietrzyk Analyt. Chem. 1974 46 330. 27 H. Takahagi and S. Seno J. Chromatog. Sci. 1974 12 507. 28 D. W. Grant R. B. Meiris and M. G. Hollis J. Chromatog. 1974 99 721. 29 H. Engelhardt J. Asshauer U. Neue and N. Weigand Analyt. Chem. 1974 46 336. Physical Methods-Par t (iva) High-pressure Liquid Chroma tograp hy 47 ~hromatography.~’ Polyurethane has also been used in the form of agglomerated particles in column form the polymerization being done in sit^.^' Although the agglomerates consist of spherical particles of only 1-10 pm diameter rapid flow rates are obtained readily the resistance being slight and sample capacity is high.Separations by steric exclusion have been successfully performed with silica gel of diameter as little as 3 pm with high controlled porosity.32 A re-evaluation of a porous silica for steric exclusion has shown that contrary to the findings in previous work pore structure was not homogeneous but that the material was nevertheless suitable for the separation of macr~molecules.~ Experiments with columns made of glass rods sintered together have shown that it is possible to separate molecules of different molecular weights by flow alone i.e. without the operation of steric exclusion principles.34 An evaluation of various steric exclusion packings has been reported.35 Unfortunately as in the case of a com-pendium of separations with porous glass beads,16 the enthusiasm of manu- facturers concerned with HPLC apparatus or materials in encouraging their staff to publish is not always matched by the depth and objectivity of the study.Besides ion-exchange packings mentioned above and below triethylaminoethyl cellulose (TEAE-cellulose) warrants mention because of its successful use for HPLC notwithstanding the susceptibility of such materials to compre~sion.~~ When packed by dry tamping it enabled prostaglandins and organic acids to be separated without blockage. Chemically Bonded Packins.-The development and application of ‘bonded phase’ packings has been and will continue to be a notable growth area in HPLC as well discussed by Locke3 and by Pryde.’ Such packings are almost invariably for use in an ion-exchange or predominantly partition mode.In the latter mode they may obviate a need that can arise with non-bonded packings i.e.to circum- vent stationary-phase ‘bleeding’. Yet as mentioned below they are not neces- sarily free from deterioration problems such as may be encountered (e.g. ref. 38) with non-bonded packings. Gratifyingly however the phenomenon of bleeding of traces of U.V.-absorbing material (with detriment to U.V. detection) was not encountered with a novel type of cation-exchanger having sulphobenzyl groups bonded to ~ilica.~’.~’ ” T. Braun 0. Bekeffy I. Haklits K. Kadar and G. Majoros Analyt. Chim. Acta 1973 64 45. 3’ L. C. Hansen and R. E. Sievers J. Chromatog. 1974 99 123. 3z K. K. Unger R. Kern M. C. Ninou and K.-F.Krebs J. Chromatog. 1974 99 435. 33 A. R. Cooper and E. M. Barrall J. Appl. Polymer Sci. 1973 17 1253. 34 S. Mori R. S. Porter and J. F. Johnson Analyt. Chem. 1974 46 1599. 3J W. A. Dark and R. J. Limpert J. Chromatog. Sci. 1973 11 114. 36 W. Morozowich J. Chromatog. Sci. 1974 12,453; J. Pharm. Sci. 1974,63 800. 37 D. C. Locke J. Chromatog. Sci. 1974 12,433. 38 G. G. Vaughan B. B. Wheals and M. J. Whitehouse J. Chromatog. 1973 78 203. ’9 R. A. Barford L. T. Olszewski D. H. Saunders P. Magidman and H. L. Rothbart J. Chromatog. Sci. 1974 12 555. 40 D. H. Saunders R. A. Barford P. Magidman L. T. Olszewski and H. L. Rothbart Analyt. Chem. 1974 46,834. 48 J. P. Leppard and E. Reid The preparation and characterization of a number of bonded phase packings with hydroxy cyano and ester functional groups have been described:’ and some commercially available packings evaluated.42 A novel polar bonded phase l-trimethoxysilyl-2-chloromethylphenylethane, has been described and tested.43 Gilpin et have described a technique for forming the bonded phase on the support after it has been packed into the column.Many of the earlier bonded phases employed Si-0-C or Si-0-Si bonds to link the functional group to the support ;but these are comparatively easily broken by hydrolysis. The C-Si bond should be much more stable and is a feature of some recently developed packings such as the above-mentioned sulphobenzylsilica cation exchangers4’ and those described by Sebestian and Halasz4’ The latter group have also described stationary phases involving the Si-N linkage.46 Packing Techniques.-Columns can be dry-packed or slurry-packed.Almost every author has his own variation on these two basic approaches as compared by Laird et a1.I3who feel that slurry packing is the only successful method for particles below 10 pm. In most hands dry packing gives efficient columns only with particles of 20 pm diameter or more. The simplest technique is to tamp small aliquots of packing with a rod of appropriate diameter (cj ref. 36) but a generally more satisfactory method is to use a variant of the ‘rotate-bounce-tap’ method,13 in which adsorbent is slowly added to the column while it is being rotated and bounced and tapped at the level attained by the packing.Hazelton4’ has designed a machine to do this automatically. A widely used method of slurry packing is that of Majors4* in which the adsorbent is slurried in a mixture of solvents whose density is adjusted to that of the adsorbent; the slurry is then pumped at high speed into the column. Cassidy et aL4’ have studied a number of packing techniques for small-particle silica gels and recommend an incremental slurry-packing method. Asshauer and HalaszS0 favour a high-viscosity slurry-packing method. A centrifugal system has been described” for packing and running microparticulate silica gel columns. 3 Applications and Approaches Choice of Packing and Separation Conditions.-Model separations that may give guidance on if not a blueprint for a particular separation problem have been collated into a useful compendi~m,~~ which however pre-supposes that the 41 M.Novotny S. L. Bektesh and K. Grohmann J. Chromatog. 1973,83 25. 42 J. H. Knox and G. Vasvari J. Chromatog. 1973,83 181. 43 E. Grushka and E. J. Kikta jun. Analyt. Chem. 1974 46 1370. 44 R. K. Gilpin J. A. Korpi and C. A. Janicki Analyr. Chem. 1974 46 1314. 4s I. Sebestian and I. Halasz Chromatographia 1974 7 371. 46 0. E. Brust I. Sebestian and I. Halasz J. Chromatog. 1973 83 15. 4i H. R. Hazelton Lab. Practice. 1974 23 178. 48 R. E. Majors J. Chromatog. Sci. 1973 11 88. 49 R. M. Cassidy D. S. LeCay and R. W. Frei Analyt. Chem. 1974 46,340. 50 J. Asshauer and I. Halasz J. Chromatog. Sci. 1974 12 139. 51 C. R.Wynne-Roberts and D. L. Davis J. Chromarog. Sci. 1973 11 406. 52 J. N. Done J. H. Knox and J. Loheac ‘Applications of High Speed Liquid Chroma- tography’ Wiley London 1974. Physica1 Methods-Part (iua) High-pressure Liquid Chromatography 49 sample is already in a fit state for loading. The latter aspect as further considered later is touched on in Table 1 which collates some recent literature with a classification based primarily on the type of packing employed (see beginning of Section 2 above). With the steric exclusion approach the molecular weight distribution of polymers can readily be determined particularly with on-line data handling;53 but the tendency of some polymers to adhere to surfaces can be detrimental.54 Whilst Table 1 gives only one example of the separation of small molecules by steric exclusion another is to be found in a report73 of the influence of hydrogen bonding the behaviour of chloroform methanol and benzoic acid on a polystyrene packing was studied with carbon tetrachloride as the eluent.Steric exclusion would be the approach of choice for small molecules only if other approaches were ineffective. Whilst porous adsorptive or partitioning packings are commonly used (Table l) for some purposes it is advantageous to use pellicular packings particularly those of bonded reversed-phase or ion- exchange type. The choice of packing and of running conditions hinges on empiricism and experience although a few of the papers cited do report compari- sons. Where conventional approaches are ineffective an ion-pair technique66 may warrant trial with attempted optimization of hydrophobic interactions of the solutes with the stationary (aqueous) phase.Initial derivatization is sometimes practised either to facilitate chromato- graphic separation as with the triacetyl derivatives of adrenaline and noradren- ali~~e,~~ or more commonly to aid detection through the formation of e.g. a coloured metal derivative7' or a fluorescent product (Table l).69The fluorogenic approach may not only improve sensitivity but also make u.v.-absorbing 53 W. W. MacLean J. Chromatog. 1974 99 425. 54 P. M. James A. C. Ouano E. Gipstein and A. R. Gregges J. Appl. Polymer Sci. 1972 16 2425. 55 G. P. Belue J. Chromatog. 1974 100 233. 56 A. F.Machin Proc. Soc. Analyt. Chem. 1973 10 92; J. Pesticide Sci. 1973 4 425. 57 C. D. Chandler G. R. Gibson and W. T. Bolleter J. Chromatog. 1974 100 185. A. J. Falk J. Pharm. Sci. 1974 63 274. 59 M. Koreeda G. Weiss and K. Nakanishi J. Amer. Chem. SOC.,1973 95 239. 6o E. H. Pfadenhauer J. Chromatog. 1973 81 85. 61 J. N. Seiber and D. P. H. Hsieh J. Assoc. Ofic. Analyt. Chemists 1973 56 827. 62 D. W. Smith T. H. Beasley R. L. Charles and H. W. Ziegler J. Pharm. Sci. 1973,62 1691. 63 W. C. Landgraf and E. C. Jennings J. Pharm. Sci.,1973,62,278. 64 A. Stolyhwo and 0. S. Privett J. Chromatog. Sci. 1973 11 20; c/ J G. Lawrence J. Chromatog. 1973 84 299. 65 W. Mechlinski and C. P. SchaRner J. Chromatog. 1974 99 619. 66 B.-A. Persson and €3. L. Karger J.Chromatog.Sci. 1974 12 521; B. L. Karger S. C. Su S. Marchese and B.-A. Persson ibid. p. 678. 67 C. V. Manion D. W. Shoeman and D. L. Azamoff J. Chromatog. 1974 101 169. 68 M. C. Olson J. Pharm. Sci.,1973,62 2001. 69 R. W. Frei J. F. Lawrence J. Hope,and R. M. Cassidy J. Chromatog. Sci.,1974 12 40. 'O W. Dunges G. Naundorf and N. Seiler J. Chromatog. Sci.,1974 12 655. 71 T. F. Gabriel and J. E. Michalewsky J. Chromatog. 1973 80 263. 72 J. H. Knox and J. Jurand J. Chromatog. 1973,87,95. 73 D. H. Freeman and R. M. Angeles J. Chromatog. Sci. 1974 12 730. 74 J. Merzhauser E. Roder and Ch. Hesse Klin. Wschr. 1973 51 883. 75 J. S. Fritz and J. N. Story Analyt. Chem. 1974 46 825. VI 0 Table 1 Illustrative separations of organic compounds in recent literature Starting materials used Chromatographic Notes on conditions Remarks especially on results besides pure solutions; category -usually e.g.mobile phase and conclusions (sensitivities soh = solvent pellicular packing if (grad = gradient) refer to amount chromatographed; Compound(s) separated extract (ion) not stated ‘porous’ detector ifnot U.V. typically x 10‘ to get amount [e.g. Juor(imetric) per ml of sample) refiactometric)] Sulphur-containing polymers Steric (gel) Aqueous; refr Troubles due to surface adherence Polymers ;blood proteins ;drugs ; Blood plasma etc. Steric (glass beads Various arbitrarily stated Theme of paper (from a flavours; etc. porous) conditions ;refr and U.V. cowpany laboratory) is merit of the packing Polyhydric alcohols e.g.Adsorb porous Partly aqueous ;refr Relevant to degradation studies isoerythritol ethylene glycol on polysaccharides Hydroxylated metabolites of Urine solv Adsorb porous Iso-octane (mainly) Hard to get good resolution on diazinon (a phosphorothionate) a preparative scale Nitrate esters e.g. Waste water (effluent Adsorb porous Dichloromethane/n-hexane; nitroglycerin monitoring) grad (optional) o-Iodohippurate Na salt Pharmaceutical solution Adsorb porous G.1.c. (derivatization difficulty) solv two-stage or t.1.c. poorer than HPLC ( +)-Abscisic acid precursor Adsorb porous n-Hexane ;9-ft column Diastereoisomers separated on ‘MTP esters’-cz and cis-diols several recycles preparative scale Xanthine and hypoxanthine Blood plasma Adsorb porous Solvent mixture Sensitivity 7 ng (endogenous) ultrafiltrate Aflatoxins Mould cultures solv Adsorb Chloroform-iso-octane T.1.c.gave better resolution Thebaine (an alkaloid) Poppies -roots etc. Adsorb Solvent mixture ;pressure XAD-2 (resin) extract only 40 p.s.i. Steroids e.g. fluocinonide Lipids (polar and non-polar) Steroids; dansyl amino-acids Polyenes (anti-fungal) Biogenic amines sulpha drugs thyroid hormones etc. Theoph ylline Polycyclic hydrocarbons Cortisone and other steroids N-Methyl carbamates (insecticides) Barbiturates Acetaminophen (4hydroxyacet- anilide) and metabolites Oligodeoxyribonucleotides Morphine alkaloids etc. Creams ointments soh (elaborate) Red blood cells soh Blood plasma solv ; urine XAD-2 extract Used engine oil Creams ointments solv Water and soil solv then dansylate Blood plasma soh then dansylate Blood serum ultrafiltrate; urine Adsorb and Partit (modified surface) Partit (pre-treat with NH solution) Partit porous Partit RP Partit porous ion-pair Bonded (nitrile) porous Bonded RP Bonded R.P.Bonded R.P. Bonded RP AX porous AX (weak type) CX (AX for one) Ternary mixture novel Ammoniacal solvent mixture grad; FID Coating applied in situ Methanol-THF-water (30,- or N(butyl),+ in stationary phase. Solvent mixtures. Propan-2-ol/n-hexane ; grad (optional) Aqueous methanol; fluor and U.V. Water with 1% methanol n-Hexane-ethanol ;fluor Aqueous methanol ;grad fluor 70 "C,grad 21 h ;fluor after Ce"' and U.V.Sulphate grad at 50°C Cations + acetonitrile b Good discussion of the methods 63 $ & Good discussion of the methods 64 ts 2 High load can be tolerated 29 8-65 I Ion-pair separation mechanism 66 discussed s e. E Caffeine etc. in plasma 67 -jeopardize g.1.c. but not HPLC 3 5 Forensic application envisaged 38 6 5 Degradation products examined 68 2 ;E Sensitivity 1 ng;initial clean-up 69 Q obviated by the dansylation 9 3 Sensitivity 1 ng 70 2 Glucuronisidates sulphates 2 $ etc.,detected incidentally 71 39 'rJ Good discussion of the methods 72 ,$ 52 J. P.Leppard and E.Reid contaminants less trouble~ome.~~ Yet good sensitivity may be attainable with U.V. detection if a suitable concentrate can be prepared for loading and if the compound (without derivatization) has a sufficiently high U.V. absorbance at the detection wavelength (commonly 254 nm). Differences in absorbance preclude generalization about sensitivity from published values such as are cited in Table 1. It has to be taken into account that with a crude sample as distinct from the authentic compound in pure solution sensitivity may be impaired due to inter- fering substances and that the loading volume to which sensitivity values com- monly refer is typically only 101.11. Thus a ‘sensitivity’ of 1ng referring to the amount loaded may imply that a concentration below (say) 1OOpgl-l in the starting sample would hardly be detectable.Many trace constituents of practical interest occur at lower concentrations than this or absorb so poorly that this sensitivity is not attainable with a U.V. detector. In Table 1 and in additional examples of separations now to be mentioned the emphasis is on ‘realistic’ applications often with a crude starting material which can hardly be successfully chromatographed without initial semi-purifi- cation by a procedure such as solvent extraction. Pilot HPLC runs with authentic material in pure solution are of course an essential step in method development ; but literature on so-called applications often from manufacturers’ laboratories is rather dominated by model separations which have not been tested on the ‘real’ starting material even to ensure that there will not be detriment to the column.A disposable ‘pre-column’ may give some protection besides being of use (e.g.refs. 6,66) to ensure that the mobile phase is saturated with the stationary phase in partition systems. Analysis of Fluids especially for Trace Constituents.-Table 1 cites a study5’ related to water pollution (cf.refs. 28 and 90) and several studies on blood or urine in relation to an ingested foreign substance such as a pesticidd6 or drug whose metabolism or pharmacokinetics can be elucidated with the aid of HPLC methodology. Other examples (see also ref. 8) are the assay of a diphenylhydan- toin metabolite in hydrolysed urine by a partition method after solvent extrac- ti~n,~~ and assays (on raw samples) for p-aminobenzoic acid and metabolites in serum or urine7? and for cefoxitin cephalothin and metabolites in urine,78 by ion-exchange methods.Beverages and soluble foods can be examined for various additives including saccharin by an anian-exchange method without pre- treatment of the sample.79 Other fluids that have been analysed by HPLC include culture media6’ and for forensic reasons engine {Table 1). Appli-cations in the petroleum industry have been re~iewed.~ Analysis of Pharmaceuticals and Materials.-The HPLC approach is being widely adopted in the assay of pharmaceutical preparations for active ingredients as evidenced by a review8 and diverse papers in the Journal of Pharmceuticai 76 T.Inaba and J. F. Brien J. Chromatog. 1973 80 161. 77 N. D. Brown R. T. Lofberg and T. P. Gibson J. Chromarog. 1974 99 635. 78 R.P.Buhs T. E.Maxim N. Allen T. A. Jacob and F. J. Wolf J. Chromatog. 1974 99 609. 79 J. J. Nelson J. Chromatog. Sci. 1973 11 28. Physical Methods-Part (iva) High-pressure Liquid Chromatography 53 Sciences during 1973 and 1974. Illustrative examples are given in Table 1 including a steroid notable for its exposition of the methodology. High sensitivity is needed only exceptionally as may be the case when stored pre- parations have to be examined for degradation products.6g Such products in multi-vitamin preparations the HPLC assay of which has been can vitiate the official fluorimetric assay for riboflaving A further instance of a field of applied chemistry where HPLC has proved useful is the analysis of food wrappings :the examination of polyolefins and PVC for phenolic antioxidants and organophosphite or organotin stabilizers has been reported.g2 HPLC as an Aid in Pure Chemistry and Biochemistry.-HPLC proved useful in an investigation of the mechanism of oxidation of xylenol by oxygen in the presence of copper and ~yridine.~~ In the reaction kinetics field HPLC has also been used to facilitate study of the rate and the mechanism of azo-group migration from a cyclic position.84 HPLC has enabled diastereoisomers to be separated sometimess9 (Table 1) but not necessarilyg5 with painstaking methodology.HPLC in the partition mode was helpful in distinguishing different 1,2-dicarba- closo-dodecaborane derivatives and discerning impurities.g6 It can also be a means of determining activity coefficients." Anion-exchangers can handle oligodeoxyribonucleotides7 (Table 1) and other nucleic-acid component^.^ 4 Apparatus Developments Detectors.-For organic compounds which absorb light poorly no highly sensitive detector is on the market.Different photometric detectors have been compared.gg One advance is the design of a reagent-addition facility :this enables Ce"' formed by Ce" oxidation of the separated solutes to be measured fluori- metri~ally~~~~~ (cf. ref. 69). Electrometric detectors offer promise but are not without limitation^.^^ The approach may be electrochemi~al,~' coul~metric,~~ polarographic,6 or condu~timetric.~~ Descriptions have appeared94 of novel systems which make use of flame-ionization or electron-capture detectors as in g.1.c.A novel detection principle is involved in the 'spray-impact detector' in R. C. Williams D. R. Baker and J. A. Schmidt J. Chromatog. Sci. 1973 11 618. " D. Wittmer and W. G. Haney,jun. J. Pharm. Sci. 1974 63 588. J. Protivovh J. PospiSil and L. Zikmund J. Polymer Sci. Polymer Symposia 1973 No. 40,233 (Chem. Ah. 1974,80 15 575). 83 A. Revillon Analyt. Chem. 1974 46 1589. 84 J. N. Done J. H. Knox R. McEwan and J. T. Sharp J.C.S. Chem. Comm. 1974 532. 85 G. Helmchen and W. Strubert Chromatographia 1974 7 713. J. H. Kindsvater P. H. Weiner and T. J. Klingen Analyt. Chem. 1974 46 982. 87 P. Alessi and I.Kikic Chromatographia 1974 7 299. 88 D. R. Baker R. C. Williams and J. C. Steichen J. Chromarog. Sci. 1974 12 499. R9 S. Katz W. W. Pitt jun. and G. Jones Clin. Chem. 1973 19 817; S. Katz W. W. Pitt jun. J. E. Mrochek and S. Dinsmore J. Chromatog. 1974 101 193. 90 A. W. Wolkoff and R. H. Larose. J. Chromatog. 1974 99 731. 9' P. T. Kissinger L. J. Felice R. M. Riggin L. A. Pachla and D. C. Wenke Clin. Chem. 1974 20 992; F. A. Schultz and D. E. Mathis Analyt. Chem. 1974 46,2253. 92 L. R,Taylor and D. C. Johnson Analyt. Chem. 1974 46 262. 9J K. Tesarik and P. Kalab J. Chromatog. 1973 78 357. 54 J. P.Leppard and E. Reid which the effluent is sprayed past a glassy carbon or gold electrode generating a high potential on the latter which varies with the type and amount of the com-ponent.Mass Spectrometry (MS) in Conjunction with HPLC.-In efforts to couple an MS detector to HPLC a major difficulty is the rather long time needed to volatilize each portion of the effluent for MS examination of components. Each such cycle took 5-7 minutes in a system designed around electron-impact MS.96 Accordingly the effluent flow sometimes had to be interrupted,as was also the case with a chemical-ionization MS systemg7 (effective if the HPLC peak contained a few micrograms of solute) in which a small proportion of the effluent was continuously fed into the MS apparatus via a capillary. Whilst atmospheric pressure ionization like chemical ionization is unsuitable for poorly volatile solutes it has given promising results with test compounds such as progesterone in an ethanolic medium the whole effluent is vaporized through a corona discharge and the resulting ions enter the MS chamber via a narrow apert~re.~' High-resolution MS seems rather slow for HPLC,98 and field desorption MS seems unsuitable for on-line ~peration.'~ A particularly interesting approachg9 (tried with electron-impact MS)is the use of a moving wire to convey a proportion of the effluent into an evaporation chamber and thence carrying the dried solute into the MS chamber.This approach is not handicapped by a long cycle time and allows of latitude in the nature of the solvent since it is pre-evaporated. Other Apparatus Developments.-Carr loo has described a sequential separator which with a moving-bed arrangement enables samples to be purified at a throughput rate of up to 100 g h- A more conventional approach to preparative HPLC is the use of wide-bore columns up to 8 cm in diameter."' Pressure can be dispensed with in a liquid chromatography approach where electro-osmotic solvent flow is created by applying a high potential across the ends of the Another way to achieve solvent flow is to use a ~entrifuge.~'The centrifugal multi-column system reported from Oak Ridge'03 is an adaptation 94 F.W. Willmott and R.J. Dolphin J. Chromatog. Sci. 1974 12 695; J. J. Szakasits and R. E. Robinson Anafyt. Chem. 1974 46 1648. 95 R. A. Mowery and R. S. Juvet jun. ibid. p. 687. 96 R. E. Lovins S. R. Ellis G. D. Tolbert and C. R. McKinney Adu.Mass Spectrometry 1974 5 457. 97 M. A. Baldwin and F. W. McLafferty Org. Muss Spectrometry 1973 7 1111; P. Arpino M. A. Baldwin and F. W. McLafferty Biomed. Mass spectrometry 1974 1 80; P. J. Arpino B. G. Dawkins and F. W. McLafferty J. Chromatog. Sci. 1974 12 574. 98 E. C. Horning D. 1. Carroll I. Dzidic K. D. Haegele M.G. Horning and R. N. Stilwell J. Chromatog. 1974 99 13. 99 R. P. W. Scott C. G. Scott M. Munroe and J. Hess jun. J. Chromatog. 1974,99 395. 100 P. J. Carr Proc. Analyt. Div. Chem. SOC. 1975 12 28. 101 D. R. Baker R. A. Henry R. C. Williams D. R. Hudson and N. A. Parris J. Chromatug 1973 83 233; E. Godbille and P. Devaux J. Chromatog. Sci. 1974 12 564. 102 V. Pretorius B. J. Hopkins and J. D. Schieke J. Chromarog. 1974 99 23. 103 C.D. Scott W. W. Pitt jun. and W. F. Johnson J. Chromatog. 1974 99 35. Phy sica1 Methods-Part (iua)High-pressure Liquid Chroma tograp hy of the centrifugal fast analyser' with in situ monitoring and with computerized data handling Chil~ote''~ has advocated a scheme to facilitate computer identi- fication of chromatographic peaks. As an aid to the automation of HPLC a high-pressure sample injection valve has been designed. 'O5 5 Theoretical Aspects Basic theory which was cardinal to the development of HPLC has been well recapitulated by Knox's gro~p.~,~,' Differences between theory and practice have been discussed for both g.1.c. and HPLC.'06 A unifying concept has been proposed for the mechanism of action of chromatographic gels."' Snyder'** has critically reviewed three theories of adsorption chromatography and con- cluded that his own is the most generally useful.Whether or not the theory of Scott and Kucera'" is valid it does enable a 'rational series of solvents' to be developed for gradient elution. Retention volume bandwidth and resolution have been theoretically related in an exposition of gradient-elution performance.' The causes and importance of solute band spreading in HPLC have been discussed."' For partition chromatography retention volumes' ' and partition coefficients' can be predicted as can suitable conditions for 'ion pair' operation.66 For resolving overlapping HPLC peaks a mathematical approach has been proposed in conjunction with an array of different detectors.114 With a given detector and a given weight load of solute great improvement in sensitivity should be obtain- able by increasing the sample injection volume say from 5 to 150pl ;there may be little impairment of chromatographic performance.' 15,' Other ways to optimize experimental conditions on the basis of theory have been well surveyed by Kirkland.' In connection with the advent of packings of ultra-small diameter,I3 Martin et a[.' have argued that there need be no great increase in pressure to achieve a given separation since the column length can be reduced as the particle size decreases. The benefit to laboratory budgets is obvious. D. D. Chilcote J. Chromatog. 1974 99 243. '05 J. Fleischer Chromatographia 1974 7 80. M. J. E. Golay Chromatographia 1973,6 242.lo' D. H. Freeman J. Chromatog. Sci. 1973 11 175. lo* L. R. Snyder Analyt. Chem. 1974,46 1384. log R. P. W. Scott and P. Kucera Analyt. Chem. 1973,45 749. P. Jandera and J. ChurBCek J. Chromatog. 1974, 91 207 223. 'I' E. Grushka Analyr. Chem. 1974,46 510A. 'l2 C. Eon B. Novosel and G. Guiochon J. Chromatog. 1973 83 77. 'I3 P. Menheere C. Devillez C.Eon,and G. Guiochon Analyr. Chem. 1974 46 1375. 'I4 N. Ostojic Analyr. Chem. 1974 46 1653. '" B. L. Karger M. Martin and G. Guiochon Analyt. Chem. 1974 46 1640. '" J. J. Kirkland Analyst 1974 99 859. M. Martin C. Eon and G. Guiochon J. Chromatog. 1974 99 357; M. Martin G. Blu C. Eon and G. Guiochon J. Chromatog. Sci. 1974 12 438.