18

Gas chromatography

S Dawling, S Jickells and A Negrusz

Introduction ...... 469 Quantitative determinations ...... 505 Gas chromatography Optimising operation conditions to columns ...... 470 customise applications ...... 506 Inlet systems ...... 483 Specific applications...... 508 Detector systems ...... 493 References ...... 510 Specimen preparation ...... 498 Further reading ...... 511

Introduction When a mixture of substances is injected at the inlet, each component partitions between the Gas chromatography (GC) is applicable to a wide stationary phase and the gas phase as it is swept range of compounds of interest to toxicologists, towards the detector. Molecules that have greater pharmaceutical and industrial chemists, envir- affinity for the stationary phase spend more time onmentalists and clinicians. Adsorption GC was in that phase and consequently take longer to developed by a German scientist, Fritz Prior, in reach the detector. The detector produces a the late 1940s. In the early 1950s Archer J. P. signal proportional to the amount of substance Martin and Richard L. M. Synge, two scientists that passes through it, and this signal is from the UK, invented partition chromatog- processed and fed to an integrator or some other raphy, for which they received the Nobel Prize in recording device. Each substance that elutes chemistry in 1952. This marks a true beginning from the column has a characteristic retention of GC as a broadly used analytical technique. If time, defined as the time interval from injection a compound has sufficient volatility for its mo- to peak detector response. Figure 18.1 shows a lecules to be in the gas or vapour phase at or schematic of a GC system. below 400ЊC, and does not decompose at these Identification of components was traditionally temperatures, then the compound can probably based primarily on peak retention time, but it is be analysed by GC. becoming increasingly more reliant on the The separation is performed in a column nature of the response obtained from the (containing either a solid or liquid stationary detector. The analyst has two main goals: firstly, phase) that has a continuous flow of mobile to make each different compound appear in a phase passing through it – usually an inert discrete band or peak with no overlap (or co- carrier gas, but more recently supercritical fluids elution) with other components in the mixture, (SCFs) have been used for some applications – and, secondly, to make these bands uniform in maintained in a temperature-regulated oven. shape and as narrow as possible. This is achieved 470 Clarke’s Analytical Forensic Toxicology

Gas supply

Injector CarrierMake up Detector

Regulators and dryers

Sample tray

Detector Oven Instrument controller

Data Data Column processor monitor

Storage

Figure 18.1 A modern GC system. partly by judicious choice of the column provided much higher resolution efficiencies stationary phase and its loading, and partly by and smaller peak volumes, which provoked optimising the operating conditions of the further modifications in detector design. The column. In addition, the method of introducing latest column developments mean that peaks of the sample into the chromatograph, the choice only a few milliseconds are a real possibility in of detector, and chemical modification to both GC and liquid chromatography (LC). Since improve the volatility of the compounds also this matches the response time of the current contribute to converting a mediocre analysis sensor electronics, the ingenuity of the detector into a first-class one. design engineers is again being tested. There has been a continuous synergism between enhanced detector performance and column performance, each advance being recip- Gas chromatography columns rocally dependent on the other. Initially, high- sensitivity detectors permitted the development of a precise column theory that, in turn, enabled Packed columns made of glass or steel contain a the design of columns that had much higher stationary phase (see below), either loaded efficiencies. The improved efficiency of columns directly into the column if it is a solid at its oper- resulted in substances eluting from the separa- ating temperature, or coated onto the surface of tion columns as very low volumes compared to a solid support if it is liquid at its operating the volume of the detectors and dispersion of the temperature. Thus, the operating principle of GC substance into that relatively large volume can be distinguished as either gas–solid chro- resulted in reduced efficiency. Thus the effi- matography (mainly an adsorptive process by ciency of separations became limited by the the stationary phase) or gas–liquid chromatog- geometry of the detector, and not by its intrinsic raphy (GLC; mainly partition of the analytes sensitivity. Detector design was modified to between the mobile and stationary phases), incorporate smaller tube dimensions, and the based on the physical characteristics of the volume of the sensing cell was thereby reduced stationary phase. Capillary columns, introduced greatly. The introduction of capillary columns in the early 1980s, have now replaced packed suitable for routine use in GC in the mid-1980s columns for most applications. The original glass Gas chromatography 471 columns were fragile and have been superseded allows for more flexibility and greater durability by fused-silica capillary columns. Fused silica is and reduces the possibility of degradation of high-purity synthetic quartz, with a protective analytes which can be catalysed by the presence coating of polyimide applied to the outer of metal surfaces. surface. Since these columns retain their flexi- Both bonding and cross-linking impart bility only as long as the coating remains enhanced thermal and solvent stability to the undamaged, their operating temperature must stationary phase (often designated ‘DB’), and be maintained below 360ЊC for standard should be used if they are available. For con- columns (400ЊC for high-temperature polyimide ventional packed column chromatography, coatings). The first capillary columns were 0.2, however, the stationary phase must first be 0.25 or 0.32 mm internal diameter (i.d.) and coated onto the surface of a solid support. between 10 and 50 m long. Subsequently ‘mega- bore’ (0.53 mm i.d.), ‘minibore’ (0.18 mm i.d.) and ‘microbore’ (0.1 or 0.05 mm i.d.) columns Stationary phases and support materials have been developed. When coated with a heat- resistant polymer, these have the advantages of Solid stationary phases flexibility and strength, and can be threaded with ease through intricate pipework. A single In gas–solid chromatography, the stationary column can be fitted into almost any manufac- phase is an active solid at its operating tempera- turer’s GC apparatus. Capillary columns provide ture. A conventional packed column is filled improved resolution, sensitivity and durability, completely by stationary phase particles, but in a less bleed with increasing temperatures, and ease capillary column a fine layer (usually less than of maintenance and repair; they yield reliable 10 lm) of particles is adhered by proprietary and highly reproducible separations, typically processes to the inner surface of the tubing, to over many hundreds or thousands of injections. create a PLOT column. These solid phases may This has reduced the number of columns be inorganic materials (e.g. aluminium oxides, required to achieve a satisfactory separation of molecular sieves, silica gel, or graphitised quite complex mixtures. carbon), or they may be organic polymers The internal surface of the silica is deactivated such as styrene. Both packed and capillary by a variety of processes that can react silanol columns use similar solid-phase materials. groups (Si1OH) on the silica surface with a Sample compounds undergo a dynamic gas– silane reagent (usually a methyl or phenyl- solid adsorption–desorption process with the methyl surface is created). For gas–solid capillary stationary phase, and since the particles are chromatography, a fine layer (usually less than porous, size exclusion and shape selectivity 10 lm) of stationary phase particles is adhered to processes also occur. The carrier gas (mobile the tubing (porous layer open tubular columns, phase) merely serves to sweep towards the or PLOT columns). For gas–liquid capillary chro- detector those solute molecules that are not matography, the stationary phase may be coated currently adsorbed. The resultant columns are or covalently bonded directly onto the walls of highly retentive, and separations impossible the column (wall-coated open tubular columns, with liquid phases can be accomplished easily on or WCOT columns) or onto a support (e.g. PLOT columns above ambient temperature. microcrystals of sodium or barium chloride) These columns are generally reserved for the bonded to the column wall (support-coated separation of low-molecular-weight materials, open tubular columns, or SCOT columns). such as hydrocarbon and sulfur gases, noble and Stainless-steel capillaries are usually reserved for permanent gases, and low boiling solvents. Since applications that require extremes of tempera- PLOT columns occasionally shed particles, their ture, or where the possibility of column breakage use is not advised with detectors that are affected cannot be tolerated. Nowadays, the internal adversely by the intrusion of particulate matter surface of metal columns is specially deactivated (the mass spectrometer is particularly vulnerable, chemically or by lining with fused silica, which as the column interface operates under vacuum). 472 Clarke’s Analytical Forensic Toxicology

Graphitised carbon black methanol to pentanol. Tenax-TA is a porous Carbopaks are graphitised carbon black, having polymer of 2,6-diphenyl-p-phenylene oxide, adsorptive surfaces of up to 100 m2/g. They are used both as a chromatographic phase and as a usually modified with a light coating of a polar trap for volatile substances prior to analysis. It is stationary phase. Difficult separations of the C1 also used for high-boiling alcohols, polyethylene to C10 hydrocarbons can be achieved rapidly. glycols (PEGs), phenols, aldehydes, ketones, Carbopak C with 0.2% Carbowax 20M has been ethanolamines and chlorinated aromatics. used to resolve substances abused by ‘glue- sniffers’. Carbopak C modified with 0.2% Liquid stationary phases Carbowax 1500 and Carbopak C with 0.8% tetrahydroxyethylenediamine (THEED) are useful In GLC, the stationary phase is a liquid or gum for the analysis of ethanol and ethylene glycol, at the normal operating temperature. Compo- respectively, in blood. Resolution is superior to nents injected into the column are partitioned that obtained with the Porapak and Chromosorb between the moving (mobile) gas phase and the polymers, although the elution order is similar. stationary phase. Molecules that have greater affinity for the stationary phase spend more time Molecular sieves immobilised in the column and consequently Activated alumina is unique for its extremely take longer to reach the detector. The process of wide pore-diameter range, and is very useful for immobilisation and subsequent release back into separating most C1 to C4 molecules, separating the mobile phase occurs thousands of times light hydrocarbon saturates from unsaturates in during the course of the analysis. The separation the C1 to C5 range, and separating benzene, of components is dependent, to a large extent, toluene and xylenes. Deactivation of alumina on the chemical nature of the stationary phase. with potassium hydroxide reverses the elution of Stationary phases are essentially two types of some molecules (acetylene and n-butane). high-boiling polymers, siloxanes (often incor- Carbosieves are granular carbon molecular sieves rectly called silicones) and PEGs. Chiral that give good separation of C1 to C3 hydrocar- stationary phases based on cyclodextrins (cyclic bons. Carboxen 1006 is useful in resolving glucose chains) have specific applications in the formaldehyde, water and methanol, and impuri- separation of enantiomers and are discussed ties in ethylene. Zeolites (5A, 13X) give a good separately. general separation of inorganic gases. Carbon dioxide is irreversibly adsorbed below 160ЊC. Polysiloxanes Oxygen, nitrogen, carbon monoxide and Standard polysiloxanes (PSXs) are characterised methane are well separated. These columns by their repeating siloxane backbone in which have a tendency to adsorb water and carbon each silicone atom has the potential to attach dioxide, which results in changes in retention two functional groups, the type and amount of over time. which distinguish the stationary phase and its properties (Fig. 18.2). The basic PSX is 100% Polymers methyl-substituted. When other groups are Chromosorb 101–108 and Porapak are divinyl- incorporated into the stationary phase, the benzene cross-linked polystyrene copolymers. amount is indicated as a percentage of the total Incorporation of other functional groups, such substituent groups. For example, if 5% of silicon as acrylonitrile and acrylic esters, into the atoms contain two phenyl groups and the polymer matrix provides moderately polar to remaining 95% of silicon atoms are methyl- polar surfaces with different pore sizes and substituted, this may be written as (5% surface areas (polarity increases with ascending diphenyl/95% dimethyl)-PSX, (5% phenyl/95% number or letter). HayeSep phases are polymers methyl)-PSX or simply (5% phenyl)-PSX. In of divinylbenzene and ethylene glycol dimethyl- some instances, two different groups are present acrylate. Separations range from free fatty acids on the same silicon atom, so a (10% cyanopropyl- to free amines, and small alcohols from phenyl/dimethyl)-PSX contains a total of 5% Gas chromatography 473 cyanopropyl, 5% phenyl and 90% methyl cyanopropylphenyl groups (80:20 or 90:10). residues (see Fig. 18.2). 100% biscyanopropyl substitution gives the While PSXs are generally less polar in nature highest polarity of the PSXs. This phase can be than PEGs, the substitution of polar residues for operated at both high and low temperatures. To a proportion of the methyl groups confers added polarity to the column. Polar phases retain polar compounds more effectively than do nonpolar compounds, and vice versa. The 100% methyl- CH substituted-PSX is often considered the ‘stand- 3 HO CH CH O CH CH O ard’ nonpolar phase, and has been used 2 2 2 extensively in compilations of retention indices. nm PEG backbone. poly(ethylene glycol) This column is an ideal choice for starting a new application. However, substitution by n-octyl Ј ϭ R or R methyl CH3 groups (up to 50%) renders the column R cyanopropyl CH2CH2CH2CN trifluoropropyl extremely nonpolar, and similar to squalene. O Si CH2CH2CF3 phenyl Ј Substitution with up to 5% phenyl groups still R n furnishes an essentially nonpolar column with Polysiloxane backbone, PSX improved thermal stability. This phase has also been used for retention index (RI) work and is another good column with which to start a new CH3 CH2CH2CH2CN CH3 application. Increasing the phenyl substitution O Si O Si O Si O Si to 20%, 35% or 50% yields columns classed as CH3 CH3 intermediate in polarity, which predictably retain aromatic compounds relative to aliphatic 5% 95% 14% 86% solutes. All these are available as bonded phases PSX-5 PSX-1701 that can be solvent rinsed, are not damaged by Poly(5%-diphenyl-95%- Poly(14%-cyanopropylphenyl-86%- dimethyl)siloxane dimethyl)siloxane organic acids or bases, and can tolerate small injections of water if sufficiently highly loaded, R RЈ but are sensitive to strong inorganic acids and O Si Si O Si bases. R RЈ Substitution of cyanopropylphenyl groups n (typically 6% or 14%) creates an intermediate Arylene backbone, polyphenylene siloxane polarity mixed phase with unique elution rela- R tive to simple phenyl substitution, but renders R R Si the column more susceptible to damage from R R Si O oxygen, moisture and mineral acids. 50% substi- O R O tution is specifically designed to separate cis- and O O HO trans-fatty acid methyl esters (FAMEs). However, OH O O HO HO R HO R even low-level bleeding of the stationary OH Si phase produces a high background signal with R O O O O R Si HO OH certain types of detector (nitrogen–phosphorus R R OH OHO O HO detector; NPD). Bleeding is the term given to the OH O O phenomenon wherein stationary phase degrades O R during analysis and elutes from the column. O O Si Si R R Depending on the nature of the stationary phase R R and the detector used, the degraded material R may be detected and can contribute to the Substituted b-cyclodextrin detector signal, potentially causing problems with the analysis. More polar columns are Figure 18.2 Structures of some common capillary GC produced by substitution of bis-cyanopropyl and stationary phases. 474 Clarke’s Analytical Forensic Toxicology date, these are nonbonded columns, and should can tolerate repeated injections of water, alco- not therefore be rinsed with solvent. Trifluoro- hols, aldehydes and acids without the need for propyl/methyl-PSX is a mid-to-high polarity acidification treatment. phase especially suited for otherwise difficult-to- separate positional isomers. Its unique interac- Chiral phases tions with nitro, halogen, carbonyl and other Second-generation chiral phases are based on electronegative groups give it application in the cyclodextrins (toroidal shaped structures formed analysis of herbicides and pesticides. by a1Ϫ4 linkages of multiple glucose units). The Increasing polarity in general is associated enzyme cyclodextrin glucosyl transferase is used with some negative effects. Polar phases tend to to cleave partially digested starch, and link the have a narrower operating temperature range glucose units into three forms, referred to as a, b, (higher minimum, lower maximum), are more and c, that have six, seven and eight glucose prone to bleed at higher temperatures, are more units, respectively. The mouth of the cyclodex- sensitive to moisture and oxygen, and conse- trin molecule has a larger circumference than the quently have a shorter life expectancy than base and is linked to secondary hydroxyl groups nonpolar phases. More recently, low-bleed of the C2 and C3 atoms of each glucose unit. The arylene stationary phases (sometimes designated primary hydroxyl groups are located at the base mass spectrometry or ‘ms’) have been introduced of the torus, on the C6 atoms. The number of that have phenyl groups incorporated into the glucose units thus determines the cavity size and siloxane backbone (see Fig. 18.2). The incorpor- electrophilic orientation (see Fig. 18.2), and ated phenyl groups confer additional strength to affects the order of the enantiomeric forms. The the backbone, which prevents the formation of hydroxyl groups can be functionalised select- cyclic fragments and associated ‘bleed’ at higher ively to provide various physical properties and temperatures. inclusion selectivities. Six different cyclodextrin derivatives are manufactured: permethylated Polyethylene glycols hydroxypropyl (PH), dialkylated (DA), trifluoro- PEGs are widely used polar stationary phases and acetylated (TA), propionylated (PN), butyryl- their general structure is shown in Fig. 18.2. ated (BP) and permethylated (PM). Changes in They are less stable, less robust (especially to elution order can be seen between the different oxidative damage) and have lower temperature derivatives, and also between cyclodextrin cavity limits and a shorter life expectancy than polar sizes. Unlike the cyclodextrins used in high- PSX phases, but they have unique resolving performance liquid chromatography (HPLC), qualities. Acid-modified PEG (also referred to as these phases separate both aromatic and nonaro- free fatty acid phase, FFAP) substituted with matic enantiomers of a wide range of chemicals, terephthalic acid is especially useful for separ- including saturated alcohols, amines, carboxylic ating acidic polar compounds, such as acids, acids, epoxides, diols, lactones, amino alcohols, alcohols, acrylates, ketones and nitriles. amino acids, esters, pyrans and furans. Derivat- Nitroterephthalic acid-substituted PEG (e.g. ised cyclodextrins are thermally stable, highly Nukol) is designed for volatile fatty acids and crystalline and virtually insoluble in most phenols. Both are highly resistant to damage organic solvents. Chiral phases are fragile, from water-based samples. Base-modified PEG however, and unless chemically bonded or cross- (CAM) is suited to analysis of strongly basic linked, they cannot be washed with solvent or compounds, such as primary amines, that do not taken to temperatures outside the 0 to 225ЊC chromatograph well on polar PSXs. Since this range. phase is usually cross-linked rather than cova- A subsequent development has been the lently bonded, it cannot be used with water or embedding of PM cyclodextrins (usually 10% or alcohol, but it can be solvent rinsed. New 20% by volume) into columns that contain bonded and cross-linked PEG phases are now standard liquid stationary phases of interme- available to separate free fatty acids and other diate polarity, such as 35% phenyl-PSX. Silyl- organic acids. These show superior inertness and substituted cyclodextrins, such as 2,3-di-O- Gas chromatography 475 methyl-6-O-t-butyldimethylsilyl are also avail- strong, and the hydrogen-bonding interactions able embedded (usually 25–35% by weight) in as weak or moderate, if the analyte has func- 20% phenyl-PSX, another intermediate polarity tional groups that can undergo hydrogen stationary phase. These columns are useful for bonding with the particular stationary phase separating positional isomers (phenols, xylenes, employed, the hydrogen-bonding interaction is etc.), as well as enantiomers. likely to be stronger than interaction by disper- sion, unless the analyte also has a high propor- tion of groups in its molecular structure that can Solute–stationary phase interactions participate in dispersion interactions, as would For liquid stationary phases, three major types of be the case for example, for long-chain fatty interaction between the stationary phase and the acids. solute determine chromatographic elution: Dipole interactions of PEG phases and the dispersion, dipole and hydrogen bonding. The cyanopropyl- and trifluoropropyl-substituted volatility of a substance is the most important PSXs enable these phases to separate solute factor defining its separation in GC. The more molecules that have different dipole moments. volatile the compound (the lower its boiling Such solutes are those with positional isomers of point), the more likely it is to be in the mobile electronegative groups, such as pesticides, halo- phase and so the faster it elutes from the carbons and many pharmaceuticals. column. Although this holds true for groups of Functional groups of solutes that exhibit compounds with similar functional groups or strong hydrogen bonding include alcohols, within homologous series, it cannot be applied carboxylic acids and amines. Aldehydes, esters universally. In general, a difference of 30ЊC in and ketones generally have less effect; hydro- boiling point is sufficient to predict and main- carbons, halocarbons and ethers produce negli- tain elution order, but differences of less than gible hydrogen bonding. Moderate hydrogen 10ЊC can be overturned by the influence of other bonding is exhibited by PEGs and interactions. cyanopropyl-substituted PSXs, with less marked Table 18.1 shows the contribution of each of effects shown by phenyl- and trifluropropyl- these interactions for the common types of substituted PSXs. liquid stationary phases. It should be remem- Although the amount of separation obtained bered that hydrogen-bonding interactions are through dipole interactions or through considerably stronger than dipole–dipole inter- hydrogen bonding can be difficult to predict, actions, which are themselves stronger than resolution of compounds with smaller differ- dispersion interactions. Thus, although the ences in dipoles or in hydrogen bonding dispersion interaction between the various strengths requires larger percentages of siloxane stationary phases is listed as strong or very substitution in order that solute–stationary

Table 18.1 Contribution of different types of interactions to solute separation on GC stationary phases

Functional group Type of interaction

Dispersion Dipole Hydrogen bonding

Methyl-PSX Strong None None Phenyl-PSX Very strong None Weak Cyanopropyl-PSX Strong Very strong Moderate Trifluoropropyl-PSX Strong Moderate Weak PEG Strong Strong Moderate 476 Clarke’s Analytical Forensic Toxicology phase interaction via dispersion forces can be • butan-1-ol for alcohols, nitriles, carboxylic exploited to bring about separation. acids and diols (electron-attracting effect) • pentan-2-one for ketones, ethers, aldehydes, esters, epoxides and dimethylamino deriva- McReynolds constants tives (dipole–dipole effect) • nitropropane for nitro and nitrile derivatives The retention behaviour of probe compounds (electron-donating effect) has been used traditionally to classify stationary • pyridine for bases (nonbonding electron phases in terms of their polarity. Lutz attraction and hydrogen-bonding effects). Rohrschneider in 1966 pioneered this type of classification using five probe compounds, Moffat et al. (1974a) devised a system to assess followed closely by McReynolds (1970). the effectiveness of liquid stationary phases by McReynolds increased the number of probe calculating the discriminating power, and exam- compounds to ten but the following five ined a number of phases commonly used in toxi- compounds are considered as the most import- cology (Moffat et al. 1974b). Contrary to popular ant: benzene, butanol, pentan-2-one, nitro- belief, it was shown that one column could be propane and pyridine. The retention indices of used to elute all the drugs studied, and that for each of these five reference compounds are screening purposes a single column, either SE-30 measured on the stationary phase being tested, or OV-17 (100% dimethyl-PSX or 5% phenyl- and then compared to those obtained under the PSX capillary equivalents), was sufficient for the same conditions on squalene (a standard reliable identification of drugs. nonpolar stationary phase). The differences in the retention indices between the two phases (DI) for the five probe compounds are added Solid supports for packed columns together to give a constant, known as the As noted above, in packed column GC where the McReynolds constant, which is used to compare stationary phase is a liquid, it is typically coated the ability of stationary phases to separate onto the surface of a solid support. The raw different classes of compounds. Phases that material for the most commonly used supports is provide McReynolds values of Ϯ4 can be substi- diatomaceous earth, calcined, usually with a tuted freely for each other as they should give flux, then crushed and graded into a number of similar retention and specificity; those differing particle sizes (60–80, 80–100 and 100–120 by Ϯ10 units generally yield similar separations mesh). Diatomaceous earths are available with a but may show some greater differences in terms variety of properties and include the Chromo- of retention and elution order. Phases with sorb series of supports. Chromosorb G has a McReynolds values below 100 are considered surface area of about 0.5 m2/g, and is suitable for nonpolar, those above 400 indicate a highly low-loaded columns – the amount of stationary polar phase and values between 100 and 400 phase should not exceed 5% (w/w). Chromosorb indicate intermediate polarity. Table 18.2 shows W has a larger surface area (1 m2/g) and accepts the McReynolds constants, operating tempera- higher loadings, but is more fragile. Chromosorb ture range, the relationship between capillary P, obtained from crushed firebrick, accepts load- and packed column nomenclature and example ings up to 35% (w/w) for some phases. Supelco- applications for the most popular stationary port is the most inert of the diatomite supports, phases. DI values for individual probes indicate and can accept 20% (w/w) loadings. Carbopack, the deviation from boiling point order and Porapak, HayeSep and Tenax-TA (see above) can consequently represent the contribution of also be used as solid supports. Table 18.3 shows forces other than dispersion to elution for that the maximum suggested coating percentages for probe. The probes are chosen to represent the most common solid supports. different functional groups as follows: Deactivated support materials are nearly • benzene for aromatics and olefins (p-type always preferred. Deactivation procedures interactions) include acid or base washing to remove Gas chromatography 477 impurities and fines, and treatment with a adjusted to give a flow of approximately 20 silanising agent that reacts with surface hydroxyl mL/min of carrier gas for a packed column or 1 groups to reduce adsorptive effects. Support to 2 mL/min for a capillary, the column may be materials that have the liquid phase chemically heated and a test mixture injected. Commercial bonded to them are available. These offer columns are invariably supplied with a chro- decreased bleed rates of stationary phase, an matogram obtained from a test mixture, and it advantage when operating a temperature should be possible to obtain a performance equal programme or when using a mass spectrometer or close to the supplied chromatogram. Various as the detector. test mixtures are used, including a mixture of dimethylphenol and dimethylaniline with straight-chain paraffins. Any acidity or alkalinity of the column is apparent by loss of the peak Installing, conditioning and maintaining shape of the amine or phenol. The efficiency columns obtained is a function of the entire chromato- graphic system. Poor efficiency or peak shape Column installation often results from a non-swept volume some- A GC column is attached at one end to the where in the system. injector and at the other end to the detector. It may be necessary to add an additional gas Attachment is typically via a nut and ferrule, the supply to the column outlet to ensure that the nut attaching to a screwthread on the injector detector is purged effectively, because most and detector. As the nut is tightened, the ferrule detectors are designed to operate with packed is compressed and helps produce a gastight columns and a flow rate of about 30 mL/min, as fitting. opposed to the 1 or 2 mL/min delivered by a Glass columns must have straight smooth capillary column (see Fig. 18.1, p. 470). Make-up ends to allow them to fit into unions at the gas is integral in the design of most modern GC injector and detector using either graphite or instruments. Vespel (polyimide) ferrules inside the nuts. Fused-silica capillaries should also have their Column conditioning ends freshly cut after insertion through ferrules, to eliminate blockages. The injector end of the A new column requires conditioning before use column should be fitted first, adjusting the to remove volatile impurities that remain after height of the protrusion above the ferrule into deactivation of the support, and/or the coating the injector according to the type of inlet being and packing processes. The column should be used (this information is supplied by the GC installed in the injector port only, with the manufacturer for the particular system in use), detector end disconnected. With the column at then tightening the fittings just enough to room temperature, a low carrier gas pressure prevent leakage when tested with a proprietary (14–35 kPa) should be maintained for half an leak-testing fluid (not soap solution which leaves hour to purge oxygen from the system. The a residue). The detector end of a packed column temperature may then be raised by about can be attached to a bubble-flow meter to ensure 1ЊC/min, to 10ЊC above the anticipated working adequate flow through the column before temperature or the maximum operating temper- connecting the detector end. The detector end of ature, whichever is less, and the temperature the capillary column may be immersed into a is maintained (2 h for a manufactured capillary, small tube of methanol to ensure adequate flow 12 h for a packed column or self-packed capillary). and the capillary end re-cut and then attached to Care must be taken not to exceed the maximum the detector and checked for leaks. The detector operating temperature. After conditioning, the is activated and the column tested at room column is connected to the detector, and a temperature with an injection of 1 or 2 lL of period of further conditioning is carried out if methane, when a needle-sharp peak should be the background signal is excessive. Some phases obtained. When the carrier-gas pressure has been (e.g. OV-17) are particularly oxygen-sensitive 478 Clarke’s Analytical Forensic Toxicology FAMEs, stereoisomers FAMEs, phenols, sulfur compounds, flavours, fragrances compounds, aromatic compounds organic acids trimethylsilyl (TMS) sugars organophosphates nitriles ethers, glycols, solvents, primary amines polychlorinated biphenyls (PCBs) cis–trans ) Applications

I D ( R Ј

s Ј

u b Ј

z Ј

y Ј

x C) Њ ( range (min/max) SP2100 Packed equivalent Temperature McReynolds values a Polarity (McReynolds values) of some common stationary phases, and example applications *-5*-1301*-35*-1701 SE-52, OV-73 SE-54, CP-624*-50, *-17*-210 –60/320 OV-11 OV-1701 –20/280*-225 SP-2250 OV-17, *-23 OV-25 30/310 19 69 10/280 XE-60, OV-210 113 0/300 74 OV-225 111 64 125 171 40/230 175 –45/250 67 93 128 101 183 170 62 592 146 268 153 40/250 146 312 Aroclors, alcohols, phenols, volatile 151 178 220 228 238 halogenated Alkaloids, drugs, FAMEs, 204 219 971 171 358 202 208 Drugs, glycols, pesticides, steroids 228 789 468 305 728 369 Aroclors, herbicides, pesticides, 310 280 Aroclors, amines, pesticides, drugs 338 1520 1175 alditol acetates, neutral sterols 492 FAMEs, Aldehydes, ketones, organochlorines, 386 1813 *-wax, *-20M*-FFAP Carbowax-20M TPA 35/280 Carbowax-1500 PEG 305 50/250 551 360 340 562 484 580 2262 397 Alcohols, free acids, essential oils, 602 627 2546 Acids, alcohols, aldehydes, acrylates, Table 18.2 Table Capillary phase SPB-Octyl*-1 Squalene, Apiezon L –60/300 OV-101, SE30, OV-1, –60/320 3 14 4 11 58 12 43 11 56 51 38 Separates by boiling point, 199 Amines, hydrocarbons, pesticides, PCBs, Gas chromatography 479 SP supplied ϭ FAMEs, positional isomers FAMEs, positional isomers, FAMEs, positional isomers FAMEs, RT supplied by Restek; * RT ϭ ethers, glycols, solvents cis–trans cis–trans alditol acetates cis–trans CPSil supplied by Chrompack; * ϭ DB supplied by J&W; * DB supplied by J&W; ϭ pyridine. Јϭ

s HP supplied by Hewlett Packard/Agilent; * ϭ 1-nitropropane; Јϭ

u pentan-2-one; Јϭ

z butan-1-ol; Јϭ OV supplied by Ohio Valley. This list is not intended to be exhaustive. OV supplied by Ohio Valley. y ϭ benzene; Јϭ -Cyclodextrin in-Cyclodextrin in 30/240 30/240 102 243 119 142 264 221 154 170 134 878 187 Enantiomers and isomers 858 Enantiomers and isomers * is the proprietary prefix for the phase, example; * x *-2330 SP-2330 10/250 382 610 506 710 591 2799 Nukol SP-1000, OV-351 60/200TCEPa 35% phenyl-PSX 314b 56935% phenyl-PSX TCEP 372a 578b 504 2337 Alcohols, free acids, essential oils, 10/145 594 857 759 1031 917 4158 Flavours, fragrances, essential oils *-2380*-2340–– – SP-2340, OV-275 25/250 DEGS EGS 10/275 419 654 541 20/200 402 758 100/200 629 637 520 3009 496 744 537 746 623 787 590 2918 643 837 903 835 889 3504 3759 Acids, esters, phenols, terpenoids TMS or methyl sugars, acidic drugs by Supelco; * 480 Clarke’s Analytical Forensic Toxicology

of a large injection volume (Ͼ2 lL) or when Table 18.3 Maximum recommended loadings for the there are solvent–stationary phase polarity most commonly used packed column mismatches. Greatest improvement is seen in supports early-eluting peaks, or for solutes with similar polarity to that of the solvent. Phase Maximum coating (%w/w)

Carbopak B 1–6 non-silicone phase Maximum operating temperatures Carbopak C 0.1–1.0 non-silicone phase Carbopak F 0.1–1.0 non-silicone phase Maximum operating temperatures for stationary Chromosorb G 20 (15 for gums) phases are usually quoted assuming isothermal Chromosorb P 30 (25 for gums) operation with a flame ionisation detector (see Chromosorb T 15 (7 for gums) Table 18.2). Other detectors may impose Chromosorb W 20 different limits, the mass spectrometer being Gas Chrom Q 15 much more susceptible to bleeding of the HayeSep Polymers 15 (5 for gums) stationary phase than the thermal conductivity Porapak 15 (5 for gums) detector. All phases bleed very slightly at high Supelcoport 20 temperatures because of the loss of smaller-sized Tenax-TA 15 (5 for gums) (and hence lower-boiling) polymer chains, although normally this is not noticeable. Oper- and can be ruined by careless conditioning. A ating temperature has a profound effect on constrictor fitted to the detector end of the column life, particularly for capillary columns. column helps prevent back diffusion of oxygen Loss of stationary phase, or breakdown of the if air or oxygen is supplied to the detector. thin film into pools to expose part of the tubing surface, results in serious loss of performance. Additionally, in columns that contain PSX Guard columns and retention gaps phases with two different functional groups, one A guard column and a retention gap are essen- group (usually that which confers additional tially the same thing, but are installed to serve polarity) is preferentially lost. This results in a different purposes. These 1–10 m lengths of change of relative separation as well as a loss of fused silica tubing are attached to the front of resolution. The temperature limit of a column the chromatography column via a press-snap may be determined by the deactivation proce- connector or zero dead-volume union, and then dure used in production, rather than by the installed into the injector port. The surface of stationary phase itself. Newer silica columns the silica is deactivated to minimise solute inter- have a very low metal oxide content, thought to actions, but no stationary phase is added. The act as a catalyst for the degradation of both tubing diameter should be the same as that of sample and stationary phase, and thus enable the column, but if different it should, ideally, be phases to be run at higher temperatures. Fused- of a wider bore. silica capillary columns have a protective The function of a guard column is to trap external coating of polyimide that is slowly deposits of nonvolatile residues, thereby degraded at elevated temperatures (maximum preventing their contamination of the analytical temperature 360ЊC, but now to 400ЊC), column. Solutes are not retained by the guard which can also limit column life. However, column (since there is no stationary phase) and separations are usually achieved at much pass directly onto the column. Portions can be cut lower temperatures. periodically from the top of the guard column as deterioration in chromatography requires, Temperature programming without any appreciable loss of resolution from the analytical column. For complex mixtures with components of A retention gap is used to improve peak shape widely varying retention characteristics, it is either when poor chromatography is the result often impractical to choose a single column Gas chromatography 481 temperature that allows all the components to Retention time of a non-retained compound or be resolved. Increasing the column temperature hold-up time (tM) throughout the analysis dramatically reduces the The retention time tM is the time taken for a non- time taken for higher-boiling compounds to retained solute to travel along the column; it elute, and simultaneously improves the sensi- represents the transit time for the mobile phase tivity of the assay, as the peaks are remarkably (carrier gas) in the column and is a column- sharper. If the early-eluting compounds are specific parameter, applicable only under the resolved inadequately, a lower starting tempera- prevailing conditions of gas flow and oven ture or slower initial ramp should be used, taking temperature. No other peak can be expected to care to observe the temperature requirements of elute earlier than this time. tM is obtained by the type of injector used. All instruments injecting a non-retained compound suitable for currently manufactured are available with a the detector system being used (butane or temperature program option, and a multi-ramp methane for flame ionisation detection (FID) or programmer is particularly useful for capillary thermal conductivity detection (TCD); acetoni- chromatography. The first ramp can be used trile for nitrogen–phosphorus detection (NPD); during splitless injection (see p. 485) to bring the methylene chloride for electron-capture detec- column rapidly up to the initial chromatography tion (ECD); vinyl chloride for photoionisation temperature, followed by a slower analytical detection (PID) or electrolytic conductivity ramp to perform the separation. One problem detection (ELCD)). with temperature programming is that the back- pressure increases with temperature and reduces Average linear velocity the carrier gas flow if a mass-flow controller _ is not used. For polar stationary phases, the The average linear velocity (l ) represents the polarity increases with temperature. Column average speed of carrier gas through the column, bleed also increases, which results in an usually expressed in cm/s, and is considered increasing baseline. For this reason the column more meaningful than measuring the flow should be well conditioned before use. (usually expressed in mL/min) at the column effluent, since flow is dependent on column diameter. This term directly influences solute retention times and column efficiency. Velocity Evaluating column performance is controlled by altering the column head pres- sure, and is calculated from the equation: Column performance, whether of capillary or packed columns, is made on the basis of effi- (18.1) ciency (the narrowness of a peak), the peak shape (whether it tails or fronts) and ability to resolve compounds. This section deals with separ- where L is the column length (cm), and tM is the ation theory, and the reader may find it useful to retention time (s) of a non-retained solute. refer at intervals to Fig. 18.3.

Retention factor Retention time Retention factor (k) is the ratio of the amount of Retention time (t ) is the time taken for a given time a solute spends in the stationary and R mobile phases and is calculated from t and t solute to travel through the column, and is the R M time assigned to the corresponding peak on the using the equation: chromatograph. It is a measure of the amount of time the solute spends in the column, and is therefore the sum of time spent in both the (18.2) stationary and the mobile phases. 482 Clarke’s Analytical Forensic Toxicology

A BCDE

Wh Injection

XYZ

Wb

0 tR(A) tR(B) Retention time

DtR

FGH 5.59 5.77 5.59 5.83 Retention time (min)

5.59 5.77

Width W1/2 0.126 min 0.071 min 0.071 min Resolution Rs 0.84 1.50 1.99 % Resolution % 50 100 100

Figure 18.3 A and B are symmetrical peaks that show the measurement of significant parameters. Wb, width at the base of the peak; Wh, width at half peak height; tR(A) and tR(B), retention times of peaks A and B, respectively; DtR, the difference in retention time between A and B; XYZ, a line drawn at 10% of peak height. Peak C is symptomatic of column overload with solute. Peak D is symptomatic of degradation of a thermally unstable solute. Peak E is symptomatic of adher- ence to active sites in the injection port or on the column. Resolution of two compounds: In F and G the peaks of the two compounds have identical retention times (5.59 and

5.77 min, respectively), but in G the peaks are narrower (Wh 0.071 versus 0.126 min, respectively), and are fully resolved (Rs 1.50 versus 0.84, respectively). In H the peak widths of the two compounds are the same as in G, but the retention time of peak B is later (5.83 versus 5.77 min, respectively). Again, the peaks are fully resolved and Rs is larger than in G (Rs ϭ 1.99 versus 1.50).

where tM is the retention time of a non-retained Separation factor solute, t is the retention time of the solute and R a tЈ is the adjusted retention time of the solute. The separation factor ( ) is a measure of the time R a Since all compounds spend an identical time interval between two peaks. If equals 1, then in the mobile phase, k is a measure of retention the peaks have the same retention time and co- by the stationary phase. A compound with a elute. Separation factor is calculated using the retention factor of 4 spends twice as much time equation: in the stationary phase (but not twice as much (18.3) time on the column) as a compound with a retention factor of 2. Thus, k provides relative rather than absolute information, and is to a where k1 is the retention factor of the first large degree independent of the operating peak, and k2 the retention factor for the second conditions. peak. Gas chromatography 483

The value of a, however, does not indicate Peak shape is usually expressed by the peak whether the peaks are resolved completely from asymmetry (As). In Fig. 18.3, the peak asymmetry one another. Two peaks may have only 0.01 min factor for substance B is given by: between them on one column but still be resolved completely, while on another column (18.5) they may have 0.1 min between them but not be resolved adequately (refer to Fig. 18.3). where a vertical line is drawn through the peak Number of theoretical plates or column maximum and XYZ is drawn at 10% of the peak ϭ efficiency height. A symmetrical peak has As 1. The theoretical plate is an indirect measure of peak width at a specific retention time. Higher plate numbers indicate greater column efficiency Inlet systems and narrower peaks. The number of plates per metre of column (N) is calculated from either The inlet system provides the means of intro- form of equation (18.4): ducing the specimen into the GC. Obtaining a ϭ 2 narrow sample band at the start of the chro- N 16(tR/Wb) (18.4a) matographic process is critical to achieve good N ϭ 5.54(t /W )2 (18.4b) R h resolution, since broad sample bands usually produce broad peaks, especially for analytes that where tR is the time from injection to peak elute early. The choice of injector depends on: maximum for the solute, Wb is the peak width at base in units of time and Wh is the peak width at • the characteristics of the specimen or residue half height in units of time. • the quantity and characteristics of the Efficiency is a function of the column dimen- analytes to be separated sions (diameter, length, film thickness or • the temperature and nature of the stationary loading), the type of carrier gas and its flow, as phase and the column. well as the chemical nature of the solute and the stationary phase. Solid samples are normally chromatographed by dissolving in a suitable solvent and injecting with a micro-syringe. Liquids can be injected Peak shape or asymmetry using a micro-syringe, but with sensitive detec- A well-designed system should give symmetrical tion systems the sample should be dissolved in a peaks, as tailing or fronting adversely affects suitable solvent to reduce the sample size and resolution. Tailing may result from non-swept avoid overloading the detector. Gases and volume in the system or from component– vapours may be introduced by injection through stationary phase or component–support interac- the inlet port septum using a gas-tight syringe. tions. Tailing of polar compounds can often be The four main types of GC injectors are mega- remedied by the use of a more polar stationary bore direct (or packed column), split, splitless phase. Fronting (shark’s fin peaks) is usually and cold on-column. In reality, split and splitless caused by overloading, particularly with capil- injection are carried out using the same lary columns, and can be resolved by making a hardware; the so-called split/splitless injector. smaller injection, by diluting the sample, or by Conventional glass syringes of 1–10 lL volume using a column with a higher stationary phase with stainless-steel needles can be used on the ratio. Column capacity is the maximum amount packed column, or vaporisation capillary inject- of a solute that can be chromatographed success- ors, and the injection is made by piercing a sili- fully without loss of peak shape. Peak fronting cone rubber septum. Care must be taken to select caused by thermal decomposition can be septa that have low bleed characteristics at the reduced by either lowering the injection temper- operating temperature, and those with Teflon ature or using a cold on-column injector system. backs are most reliable in this respect. Unstable 484 Clarke’s Analytical Forensic Toxicology solutes can be decomposed by the high tem- column. A deactivated glass-wool plug at the top perature of the injection system, particularly of the column serves as a filter to retain if the system is constructed of metal. For nonvolatile co-injected material, and must be labile substances, cold on-column injection is replaced regularly to prevent turbulence or preferred, but clean extracts must be used to blockage of the carrier gas flow. Packed column minimise column contamination. The injection inlets (6.3 mm diameter) can usually be modi- system typically utilises a removable liner to fied by insertion of a reducing fitting and glass minimise contact with metal surfaces. These liner to take megabore capillary (0.53 or 0.45 liners are usually made of glass. They come in mm i.d.) columns, but the high flow require- various configurations (see Fig. 18.4) and the ments to sweep the compounds onto the type used depends on the nature of the sample column in a narrow band (4 mL/min minimum) being analysed, the injector system used and the preclude the use of narrower capillaries. One injection volume. With repeated injection of type of injection liner usually has a restriction at samples, liners inevitably become contaminated the top to prevent backflush and septum contam- with nonvolatile material. This can cause ination (see Fig. 18.4E), to which the top of the unwanted interactions with subsequent injec- column is abutted to reduce interaction of the tion of samples. Hence liners can be removed, sample with the steel surface of the injection cleaned by solvent washing and replaced in the port. The injection is made directly onto a capil- injector. lary column (sometimes referred to as hot on- column injection), and it is essential to maintain clean specimens to avoid poor chromatographic Megabore direct (or packed column) performance. The installation of a retention gap injection (see p. 480) may be useful in this situation. This type of injection is ideally suited to high-boiling The packed column is usually inserted such that compounds, and minimises degradation of ther- the top butts directly onto the septum or septum mally labile compounds. An alternative type of plate, and is housed in a heated port. The sample liner (flash vaporisation) avoids direct injection is injected directly on to the top of the column, onto the capillary by having a second restriction and the carrier gas (typically 10–40 mL/min) about half way down (Figure 18.4F), to which sweeps the compounds directly along the the top of the column is abutted. This creates a vaporisation chamber above the column into AB C DE F GH which the injection is made. Chromatography is generally more efficient, since the second taper acts as a concentrating zone for the solutes, and produces a narrow solvent front, which enables analysis of highly volatile compounds. Glass wool should not be inserted into these liners. The injector temperature is usually held about 50ЊC above the boiling point of the solvent, with the initial column temperature (if a tempera- ture programme is available) some 10ЊC below the boiling point. Usually, experimentation is needed to balance the injector temperature with the column temperature and carrier gas flow to obtain the most efficient chromatography. This is the simplest capillary injector available and is compatible with most samples. The highest Figure 18.4 Examples of injector liners. Note the concentration is usually limited by the column different shapes that are available and the fact that glass capacity, and the smallest amount by the wool is used in some liners. sensitivity of the detector. Gas chromatography 485

Split and splitless injectors sample down the column and may result in column overload. High split ratios waste large As noted above, modern GC instruments have amounts of carrier gas and insufficient analyte the split/splitless injector combined in a single may reach the column. The analyst therefore injection system (see Figure 18.5). needs to find the optimum split ratio for a partic- The split mode of injection is used for more ular analysis. concentrated samples, since only a fraction of In splitless injection, all the carrier gas passes the sample actually enters the column. An inlet to the column. This is useful for very volatile splitter allows a high flow of carrier gas through compounds, for low sample concentrations or the injector while maintaining a low flow (1–4 for trace analysis. The flow rate in the injector is mL/min) through the column: the excess gas the same as that in the column (1–4 mL/min), and associated sample components are vented to and the only path for the injection to take is into the atmosphere through the split line (also the column, since the split vent is closed. At a referred to as the split outlet/split vent). The fixed time after injection (usually 15–60 s), the ratio of these two flows (the split ratio) controls injector is purged by opening the split vent to the proportion of the injected sample that introduce a much larger flow of carrier gas reaches the column. The total flow through through the injector (typically 20–60 mL/min) the injector may be from 10 to 100 mL/min, and any remaining sample in the injector is which gives split ratios of 10:1 to 100:1. The discarded through the split vent. Injection of a injection of a sample in split mode is shown sample in splitless mode is shown in Figure 18.7 in Figure 18.6. A good splitter should be linear, Since the rate of sample transfer onto the that is it should split high- and low-boiling point column is so slow in splitless mode (because of compounds equally. The function of the splitter the low gas flow), peaks are usually somewhat is not primarily to reduce sample volume, but broader than for split injections. Temperature rather to ensure that the sample enters the conditions can be adjusted to narrow or focus column as a compact plug. Split injections, the sample band at the top of the column. Split- therefore, produce some of the most efficient less injections should therefore be made with the chromatographic separations, and allow the use initial column temperature at least 10ЊC below of very narrow capillary columns. A lower split the boiling point of the solvent, and the initial ratio channels a larger fraction of the injected temperature should be held at least until after the purge activation time. Solvent condenses on the front of the column and traps the solute molecules, which focuses the sample into a Septum narrow band (known as the solvent effect) (see Septum purge outlet Fig. 18.8). Individual solutes with boiling point Carrier gas Њ inlet 150 C above the initial column temperature condense and focus at the top of the column in

Split outlet a process known as cold trapping. Either the Heated metal solvent effect or cold trapping must occur before block efficient chromatography can be obtained. Vaporisation Some newer chromatographs have the option chamber Liner of a pulsed splitless injection. In this mode, the column head pressure is increased immediately Column upon injection (typically to 174 kPa) and held there for 30–60 s, before returning to the normal operating pressure. This facilitates band sharp- ening and, while the process is not guaranteed to increase the fraction of the injection delivered Figure 18.5 Cross section through a split/splitless onto the column, sensitivity is often improved injector. because of improved chromatography. 486 Clarke’s Analytical Forensic Toxicology

Split injector at point of injection of sample

Carrier flow through split injector before injection of sample

Septum Septum Valve Septum Septum Valve nut nut

Mobile Septum Mobile Septum phase purge phase purge O-ring O-ring

Split flow Split flow

Liner Liner

Column Column

Split-splitless injector a few seconds Sample some seconds after injection after injection of sample in split mode

Septum Septum Valve Septum Septum Valve nut nut

Mobile Septum Mobile Septum phase purge phase purge O-ring O-ring

Split flow Split flow

Liner Liner

Column Column

Figure 18.6 Injection of sample in split mode.

Glass liners for split and splitless injectors deactivated glass wool in the liner helps prevent come in a variety of shapes and volumes (see the deposition of nonvolatile or particulate Fig. 18.4) and it is prudent to start with a material on the column, but may cause some straight liner, and to investigate some of those peak discrimination, and for the best results that cause turbulence (e.g. the inverted cup needs to be placed at a consistent position in style) later if this is unsatisfactory. A plug of the liner. Packing splitless injection liners with Gas chromatography 487

Split/splitless injector at point of injection of sample in splitless mode

Split/splitless injector just before injection in splitless mode

Septum Septum Valve Septum Septum Valve nut nut

Mobile Septum Mobile Septum phase purge phase purge O-ring O-ring

Split flow

Liner Liner

Column Column

Split/splitless injector a few seconds Split/splitless injector after split valve opened after injection in splitless mode ca 15–60 s after injection in splitless mode

Septum Septum Valve Septum Septum Valve nut nut

Mobile Septum Mobile Septum phase purge phase purge O-ring O-ring

Split flow

Liner Liner

Column Column

Figure 18.7 Injection of sample in splitless mode.

deactivated glass wool may decrease the chro- Large-volume injectors matographic performance, but this must be weighed against the potential for damage to the The analysis of trace amounts of components or stationary phase from the repeated injection of contaminants in complex matrices such as nonvolatile or particulate material. foods, beverages, faeces and environmental 488 Clarke’s Analytical Forensic Toxicology

capillary column, typically by low-temperature Mobile evaporation through the split vent. As the phase Just after sample is concentrated towards the bottom of sample the injector, the injector is heated, the split vent injected is closed and the analytes are introduced onto the GC column in splitless mode. Those injectors Allow that can be heated selectively and cooled allow solvent to vaporise the precise introduction of selected components only from the sample, and thus reduce the quan- Analytes tity of nonvolatile components (e.g. sugars) that focused but still might overload or destroy the analytical column. condensed The TCRC has a small mobile oven (2–8 mm width) that can be scanned along the length of Figure 18.8 Solvent focusing and cold trapping. After the column to produce band compression. Prior the sample is injected, solutes and solvent condense on the to the next injection, the injector columns are head of the column, which is at a lower temperature than usually baked to vent high-boiling compounds the boiling point of the solvent. The solvent evaporates to waste. Sensitivity can often be improved 50- owing to the flow of mobile phase (carrier gas). The uni- to 100-fold and time is saved in sample prepara- directional flow of the carrier gas causes the solvent and associated solutes to focus as a narrow band on the tion, since extensive clean-up or extraction column. As the temperature program is initiated, the procedures are no longer required. solutes themselves vaporise and move along the GC column. Cold on-column injection samples is difficult. Adequate sensitivity to Cold on-column injection is most suited to detect trace components is provided by specific compounds that are thermally labile. The injec- detectors such as the NPD or ECD, but regulatory tion needle must be fine enough to enter the standards require positive identification of these column bore, usually fused silica or stainless compounds by mass spectrometry (MS). To over- steel with a fused silica insert (see Fig. 18.9). The come the inferior sensitivity of MS compared to top of the column is held at a temperature low these sensitive and specific detectors, large- enough for the solvent that contains the sample volume injectors have been developed. Examples to condense, usually by an air- or carrier gas- include the Apex pre-column separating inlet cooled sleeve. The solvent temporarily swamps (PSI), the temperature-programmed sample inlet the stationary phase and ensures that the sample (PTV) from Gerstel, and time-coupled time- components concentrate in a narrow band. Any resolved chromatography (TCRC). The inlet typic- solvent or sample that remains in the injector is ally consists of a length (10–50 cm) of standard backflushed with carrier gas, often by automatic (2 mm i.d.) glass chromatography column that valves. The proximal end of the column is then can be deactivated or packed with traditional brought rapidly to the operating temperature, materials. The first two injectors are mounted when the solvent vaporises and chromatography directly in the GC injector port; the latter is a begins. The potential for rapid column contam- free-standing column coupled by a four-way ination or deterioration means that cold on- valve into the GC inlet. Injection volumes range column injection is usually restricted to those from 125 lL for the PSI, 1 mL for the PTV and up applications where its use is essential. Cold on- to 20 mL for the TCRC. Injection of larger column injection is usually used in conjunction volumes (up to 60 mL) is possible for some appli- with a retention gap (see p. 480). This is typically cations, but results in discrimination in favour of a short (1–5 m) piece of fused silica capillary high-boiling components and loss of volatiles. tubing with no stationary phase which is Large-volume injectors remove the solvent attached between the injector and the analytical from the sample prior to its introduction onto the column. This piece of tubing becomes contami- Gas chromatography 489 nated with nonvolatile material after many in the fields of industrial air monitoring, analysis samples have been injected, but it can then of residues in food, soil and water, petrochemical simply be removed and replaced with a new analysis, and environmental monitoring. The piece of tubing, thereby protecting the main methods of preparing samples for analysis are analytical column and prolonging its use. described in the section on specimen prepara- tion. These samples require special interfaces with GCs to ensure good chromatography. In Volatiles interface some instances the sample preparation device and injector are manufactured as stand-alone The volatiles interface allows automated analysis pieces of equipment that require very little of gaseous samples. The interface is a low- modification of conventional injectors, while volume highly inert switching block, and is others must be dedicated pieces of equipment. ideally suited to trace-level detection. A portion Once collected, the concentrated sample must be of the carrier gas supply is diverted through the desorbed into the chromatograph using the specimen sampler and released under controlled heated injector port. The major problem here is conditions onto the column. The remainder of the possible introduction of water into the chro- the carrier gas goes to a flow sensor, which matograph from moisture adsorbed during prevents fluctuations in column gas flow that collection from high-humidity samples. Release would otherwise occur when the switching of solutes from the adsorbent should be as rapid valves are opened and closed. The interface and complete as possible to allow for rapid and can be run in split, splitless or cold-on column sensitive analysis and for a narrow sample band modes as described in the sections above. to be introduced into the chromatograph. This is Samples may be introduced from external achieved either by cooling the column oven devices, such as air samplers or purge and trap cryogenically to refocus the sample in the devices (see below), or from headspace analysis, injector prior to injection or by using a dry purge which permits analysis of volatile substances in a system coupled to the GC via a volatile interface liquid sample while minimising contamination (see above) designed to operate above ambient of the column. This technique is used in the temperature. Here, the specimen is thermally assay of ethanol and other solvents in blood and desorbed from the collection tube onto a for complex household preparations, such as narrower (1 mm i.d.) tube of the same adsorbent polishes, which contain volatile substances (see material. The concentrated solute is then p. 311). released into the chromatograph, ensuring rapid and complete sample introduction. Adsorbents must be thermally stable to reduce interference Thermal desorption and purge-and-trap from background contaminants. injection With solid-phase microextraction (SPME) the adsorbed sample is introduced into the heated The analysis of samples that have been pre- injector port via a special sleeved needle (see concentrated onto solid adsorbents is common p. 501). This technique requires the injector liner to be narrow (usually 0.75 mm as opposed to 2 or 4 mm) to increase the linear velocity of carrier Oven wall or Syringe oven top gas through the liner and ensure a narrow band of sample is introduced onto the column.

Wide-bore Solid injection Silicone capillary septum column Carrier gas When solvent interference is serious the sample may be injected as a solid. The ‘moving needle’ Figure 18.9 On-column injector. injector has found application in steroid analysis 490 Clarke’s Analytical Forensic Toxicology and for the determination of anticonvulsant drugs. A solution of the material to be injected is Table 18.4 Boiling points and expansion volumes for placed on the tip of the glass needle with a commonly used injection solvents syringe. A small flow of carrier gas sweeps the Solvent Boiling Expansion volume (lL) solvent out of the top of the device to waste. The point (ЊC) per lL of solventa dry residue is then introduced by moving the needle into the heated injection zone of Methylene chloride 40 330 the chromatograph with a magnet. This form of Carbon disulfide 46 355 injection can only be used with compounds that Acetone 56 290 do not volatilise with the solvent. Methanol 65 525 n-Hexane 69 165 Ethyl acetate 77 215 Backflush Acetonitrile 85 405 iso-Octane 99 130 Upon vaporisation, the injected sample under- Water 100 1180 goes considerable expansion, sometimes up to Toluene 111 200

100 to 1000 times its original volume, which a Values are given at 250ЊC and 105 kPa head pressure. creates a pulse of pressure that often exceeds the column carrier-gas pressure. If the volume of the liner is smaller than the expanded solvent volume (see Table 18.4), some of the sample is Injector discrimination propelled out of the injector in a process known as backflush. This can appear as a broad tailing Injector discrimination occurs because not all solvent front, since it now takes longer to flush the compounds in the sample vaporise at the the expanded solvent out of the injector and same rate. Since the sample remains in the liner carrier-gas line. Backflush can also cause injector for a limited time, this usually results in some contamination, since the analytes condense in loss of higher-boiling solutes, potentially the cooler carrier-gas line, from where they may resulting in lower sensitivity for these bleed continuously into the injector and cause compounds (see Fig. 18.10). It is particularly high background or spurious peaks. Carryover or notable with split injection or splitless injection peak ghosting can occur when the next injection with a low purge time. Discrimination can be backflushes and carries previously condensed alleviated by increasing the residence time of the compounds back into the vapour phase and sample within the injector, or by using a higher onto the column. Backflush can usually be injector temperature (Fig. 18.10) or smaller injec- solved by using a smaller injection volume, a less tion volume. However, there is usually a expansive solvent, a lower injector temperature, compensatory loss in lower-boiling compounds. a liner with an upper restrictor or a faster carrier Discriminating behaviour can usually be gas flow. The use of an adjustable septum purge managed by making reproducible injections. gas (0.5–1 mL/min usually) also decreases the potential for backflush, as components that would normally condense on the cooler septum Gas pressure and flow control and travel into the carrier-gas lines are swept away by the septum purge. Too high a purge flow For accurate and reproducible gas chromatog- results in loss of highly volatile components. raphy, either a constant carrier-gas flow or a Boiling points and expansion volumes for constant carrier-gas pressure must be main- commonly used injection solvents are presented tained. Under isothermal conditions, simple in Table 18.4 pressure control is adequate for packed or capil- Gas chromatography 491

120°C

400 000

C8 300 000

200 000

Detector response 100 000 C22

4.00 8.00 12.00

200°C

400 000 C8

C 300 000 22

200 000

Detector response 100 000

4.00 8.00 12.00 Time

(minutes)

Figure 18.10 Analysis of alkanes, showing discrimination during injection. Injection in splitless mode. Note the lower signal-to-noise ratio for the alkanes with the injector at 120ЊC. When the injector temperature is raised to 200ЊC there is improved transfer of alkanes, most notably the higher boiling point alkanes. lary columns and back pressure can be moni- throughout the analysis. Figure 18.11 shows the tored by a pressure gauge between the flow effects of increasing the column temperature on controller and the injector. A decrease indicates the carrier gas flow and velocity if the head pres- a leaking septum and an increase suggests con- sure is held constant during the run. As flow and tamination of the injector liner or the top of the velocity do not respond identically to increasing column. Flow control is highly desirable, if not temperature (see Fig. 18.11D), late-eluting essential, during temperature programming with analytes are recovered more quickly using packed columns and can be used to advantage constant flow than under constant pressure with capillary columns. The added convenience conditions. Furthermore, since column effi- of a digital (electronic) flow controller may be ciency is a function of the carrier gas velocity worthwhile. (Fig. 18.12), resolution at the end of the chro- Since the carrier gas becomes more viscous in matogram is improved under constant flow condi- the column as the temperature rises, the gas tions. Switching between conditions of either pressure must be increased as the run progresses constant flow or constant pressure can sometimes to maintain constant velocity (or constant flow) resolve otherwise co-eluting compounds. 492 Clarke’s Analytical Forensic Toxicology

A 5 Hydrogen B 100 Nitrogen 4 Helium 80

3 60 2 Flow (mL/min) Velocity (cm/s) Velocity 40 1

0 20 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Column temperature (°C) Column temperature (°C)

C 50 D 175

45 140

40 105 35 70

Velocity (cm/s) Velocity 30 Head pressure (kPa) 35 25

20 0 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Column temperature (°C) Column temperature (°C)

Figure 18.11 Effect of temperature on carrier-gas flow and velocity. (A) and (B) are under conditions of constant carrier- gas head pressure (140 kPa). (A) shows the change in column flow (mL/min) with change in temperature from 50 to 300°C. (B) shows the change in velocity (cm/s) with change in temperature from 50 to 300°C. (C) and (D) are under conditions of constant carrier-gas flow (1 mL/min). (C) shows the change in carrier-gas velocity (cm/s) with change in temperature from 50 to 300°C. (D) shows the change in column head pressure (kPa) with change in temperature from 50 to 300°C. All calculations are for a 25 m column of 0.25 mm i.d. operating at atmospheric pressure.

The way in which carrier-gas velocity affects theoretical plate; HETP) but it can be seen that column efficiency is best demonstrated by refer- the chromatography is much less tolerant to ence to the van Deemter curves in Figure 18.12. changes in nitrogen velocity than to helium. These demonstrate that the optimum column Helium is favoured by most users, as analysis efficiency (minimum plate height, H) occurs at times are half that with nitrogen, with only a intermediate velocity, and that column effi- slight loss in efficiency. Although hydrogen gives ciency is compromised at both low and very the best dynamic range and shortest analysis high velocity. A small loss in efficiency for a times, there are safety issues relating to its use. shorter analysis time is usually tolerated. Curves While the gas used for the carrier gas should are shown for the three most common carrier always be of the highest purity available, a lower- gases (helium, nitrogen and hydrogen). Of the quality gas can sometimes be used for the three gases, nitrogen produces the highest effi- makeup or detector, since these do not ciency (smallest value for height equivalent of a contribute to column deterioration by oxidation. Gas chromatography 493

sensitivity and selectivity. The use of detectors 1.2 N2 such as the ECD to identify amenable 1.0 compounds, and the NPD to detect compounds that contain phosphorus and nitrogen, removes 0.8 many of the extraneous peaks frequently He 0.6 observed when using nonselective detectors, such as the FID. However, these selective detec- HETP (mm) 0.4 H2 tors have also led to the detection of substances 0.2 such as plasticisers from blood-collection tubes or transfusion lines, which interfere in many 0.0 toxicological analyses. Detectors that detect the 0 10 20 30 40 50 60 70 80 90 Carrier gas velocity (cm/s) presence of a solute and also give information about its structure are increasingly popular and Figure 18.12 Van Deemter plots for a 25 m ϫ 0.25 mm MS, Fourier transform infrared spectroscopy and i.d. WCOT OV-101 column (HETP, height equivalent of a atomic emission spectrometry have been theoretical plate). invoked to achieve this goal. Detector sensitivity is measured as signal-to-noise ratio, in which the signal corresponds to the height of the peak, and Regardless of quality, it is advisable always to use the noise to the height of the baseline variability. a scrubber (to remove oxygen and hydrocarbons) A signal-to-noise ratio of 8 to 10 is considered followed by a dryer (to remove water vapour) sufficient to confirm the presence of a peak. Each between the supply and the instrument. Metal type of detector has a linear operating range in trap bodies are recommended, as plastics are which the response obtained is directly propor- permeable to impurities in laboratory air, espe- tional to the amount of solute that passes cially when large amounts of organic solvents through, although this can be modified slightly are used. Most traps have an indicator to show by the nature of the solute and the chromato- when they are saturated, and they can be graphic conditions (mobile phase type and flow, changed without interruption to the gas flow. detector temperature). The linear operating Stainless steel or copper tubing is recommended range is considered to be exceeded when the for plumbing of all gases, as plastics are perme- incremental response obtained from the detector able to moisture and oxygen, and Teflon, nylon, varies by more than 5% from that expected. polyethylene, polypropylene and PVC contain Most detectors (except MS) rely on gas other contaminants that degrade gas purity. than the mobile phase (combustion, reagent or purge gas) for their operation. Usually, a total flow of at least 30 mL/min is necessary to sweep the solute molecules physically Detector systems through the body of the detector at sufficient speed to prevent flow disturbances and The choice of chromatography detector for an produce narrow peaks. Thus, the addition of a application depends on factors such as cost, ease ‘makeup’ gas is invariably required with capil- of operation, consumables’ supply, sensitivity, lary columns. Recommended gases and their selectivity and the linear working range (see later flows for each detector are included in the in this section). manufacturer’s instruction manuals, and it is Some detectors respond to almost all solutes, important to follow these guidelines (and while others (selective detectors) respond only to those on maintenance) to achieve the stated solutes with specific functional groups, atoms or performance. structural configurations. Additional functional Here, only the most widely used detectors are groups can often be added to solutes, generally considered in detail. Several other types of detec- after extraction (see p. 503), to achieve a response tors are available; for a more detailed discussion, from a selective detector and gain additional the reader is referred to the text by Scott (1996). 494 Clarke’s Analytical Forensic Toxicology

Flame ionisation detector (FID) which must be periodically removed. The insen- sitivity of the detector to water is a useful feature The FID is the most widely used of all detectors, that allows aqueous solutions to be used. since it responds to nearly all classes of compound. The effluent from the column is mixed with hydrogen and the mixture burnt at a small jet in Nitrogen–phosphorus detector (NPD) or a flow of air. A polarising current is applied alkali flame ionisation detector between the jet and an electrode situated above it (Fig. 18.13). The introduction of alkali metal vapours (usually When a component elutes from the column, it supplied by an electrically heated bead of burns in the flame to create ions that carry a rubidium or caesium chloride) into the flame or current between the electrodes and provide the ‘plasma’ of an FID confers an enhanced response signal. The background current and noise are to compounds containing phosphorus and both low. Any of the usual carrier gases can be nitrogen. By adjustment of the plasma gases, the used and minor changes in gas flow are without detector can be made virtually specific for phos- effect. Sensitivity is moderate (0.1–10 ng), with phorus compounds (e.g. a phosphorus:carbon linearity extending sometimes as high as six response ratio of 50 000:1 and a phosphorus: orders of magnitude. The response of the FID is nitrogen response ratio of 100:1). Even when dependent on the number of carbon atoms in optimised for nitrogen compounds, it retains its the molecule, but the response is lowered if response to phosphorus (e.g. a nitrogen:carbon oxygen or nitrogen is also present in the mol- response ratio of 5000:1 and a nitrogen:phos- ecule. It responds to all organic compounds that phorus response ratio of 10:1). This detector is contain carbon–hydrogen bonds with the excep- particularly useful for drug analysis, since most tion of formic acid. Both the sensor design and drugs contain nitrogen, while the solvent and electronics are simple, and manufacturing cost is the bulk of the co-extracted material from a therefore low. The FID is easy to clean, and when biological sample do not. The NPD is ideal to operating with capillary columns it is virtu- detect pesticides that contain phosphorus, and ally maintenance free. With packed columns, therefore has wide application in environmental however, there is a tendency for a build up of and regulatory analysis (air, soil, water and stationary phase bleeding from the column, residues in food) and in clinical and toxicolog- ical analysis where pesticide poisoning is suspected. The extreme sensitivity to compounds that contain phosphorus can be further exploited by the preparation of derivatives that contain this element. Sensitivity is excellent

Collector electrode (1–10 pg), with a good linear range of up to four Flame or six orders of magnitude. A disadvantage is the ignition coil need for the supply of three gases and, unlike ϩ300 V Polarising voltage with the FID, their control is absolutely critical to selectivity. The detecting element (bead) lasts between 1 and 3 months depending on usage. Stationary-phase bleeding from packed columns coats the bead and collector assembly and can be rinsed off using methanol or dilute (0.1 M) sulfuric acid. Most of the early problems that Air Hydrogen arose from poor reproducibility in bead coating

Column have now been resolved, and the most stable detectors nowadays have a geometry that Figure 18.13 Cross-section though a flame ionisation enables the bead to be located and fixed in its detector. optimal position with relative ease. Gas chromatography 495

Electron-capture detector (ECD) The ECD is a selective detector with a very high sensitivity to compounds that have a high The early form of this detector consists of a small affinity for electrons; for many compounds, the chamber with a pair of electrodes and a radio- sensitivity of the ECD often exceeds that of MS, active source, usually 63Ni, placed close to and sometimes even that of the NPD. the cathode to ionise the carrier gas. Potential Compounds that contain a halogen, nitro group applied to the electrodes produces a steady back- or carbonyl group are detected at concentrations ground current. There is an interaction with the of 0.1–10 pg, 1–100 pg and 0.1–1 ng, respect- electrons emitted from the 63Ni source with the ively. This makes it very useful for compounds carrier gas to produce ‘low-energy’ electrons. such as the benzodiazepines or halogenated These electrons then interact with electronega- pesticides and herbicides. Alternatively, the great tive solutes (Fig. 18.14). The response of the sensitivity of the detector may be utilised by detector is therefore a loss of signal rather than preparing derivatives with halogenated reagents, an increase, as is given by most other detectors. such as trifluoroacetic, heptafluorobutyric Although the ECD can be polarised from a suit- or pentafluoropropionic (PFP) anhydrides. able low-voltage direct-current supply, it is more Linearity (at best only two or three orders of sensitive when a pulsed power supply is used, magnitude) is a limiting factor for quantitative and in modern detectors the polarising pulses analysis. In older models, the addition of a small are modulated to maintain a constant current. A amount of quench gas, such as methane, voltage dependent on the modulation frequency improves stability and linearity, and is essential is generated as the output signal. Additional if argon or helium carrier gas is used. Newer carrier gas is necessary, even with packed models can be operated successfully with helium columns, to obtain a flow of at least 60 mL/min as both carrier and detector gas. The ECD, to purge the detector adequately and avoid peak because of its high sensitivity, can be contam- broadening and distortion. Sensitivity can also inated easily: an impure cylinder of gas can be improved dramatically by raising the oper- damage a detector beyond repair in a matter of ating temperature of the detector, and decreasing only a few hours. The use of electrophilic the makeup gas flow. solvents as dissolution agents for solutes should

Anode ϩve

Cathode Ϫve Ϫ e Ϫ Ϫ Ϫ b e Ϫ Ϫ Ϫ Ϫ b Ϫ b e b Ϫ Ϫ Ϫ b b Ϫ e b b e Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ e e b b e e Ϫ Ϫ b b

Ϫ Ϫ b particles emitted from b particles interact Thermal electrons Electronegative analytes radioactive source with CH4 in carrier gas are attracted to elute from column, to produce secondary anode – standing ‘capture’ thermal energy thermal electrons current set up between electrons – standing anode and cathode current reduced

Figure 18.14 Schematic of operation of electron capture detector (ECD), simplified to note the most important principles of how the detector operates. 496 Clarke’s Analytical Forensic Toxicology be avoided. Cleaning is difficult, although some ecule. Computer libraries allow for easy and material can be removed by heating the rigorous comparison of spectra and sample iden- detector to its maximum operating temperature tification. FTIR can be fully quantitative, but it is overnight, and the injection of water in 100 lL relatively insensitive (10 ng range). Its advan- aliquots through an empty glass column can also tages are that it is nondestructive and it can help. However, if contamination is avoided, it is distinguish between isomers (MS cannot). virtually maintenance-free. The radioactive source requires special handling procedures that may be subject to regulation. More recently, it Atomic emission detector (AED) has been shown that this detector can work with greater sensitivity and operate over an increased With the AED, carrier gas that elutes from the linear range using a helium plasma in place of column delivers solutes into a high-temperature the radioactive source. helium plasma, where heat energy is absorbed by the constituent elements. In returning to their ground state, they emit energy as light, the Fourier transform infrared detector (FTIRD) wavelength of which is characteristic for each element. Emitted light is focused by a quartz lens In the FTIRD, the column effluent is conducted and spherical mirror onto a diffraction grating, through a light pipe and swept by a scavenging and the dispersed light is focused onto a diode gas into the path of an infrared light beam that array that is continuously scanned (wave- has been processed by an interferometer. The length usually 170–800 nm). Typically, some 15 interferometer directs the entire source light to elements can be monitored simultaneously, and a beam splitter, which sends the light in two each is plotted against time. The composite chro- directions at right angles. One beam takes a matogram allows the percentage elemental fixed pathlength to a stationary mirror, while composition of each peak to be determined. the other takes a variable pathlength to a Sensitivity is very good and, although these computerised moving mirror. The two beams detectors are complex and expensive to operate, are recombined, and the difference in path they find use in environmental, nutritional and lengths creates constructive and deconstructive clinical and toxicological analysis. interference, or an interferogram (see the associ- ated discussion in Chapter 16 on infrared, Fig. 16.6). Mass spectrometry (MS) detector The recombined beam is then passed through the sample. Analyte molecules absorb light A gas chromatograph is an almost ideal inlet energy of specific wavelengths from the inter- device for quadrupole MS. The detector is main- ferogram, and the sensor reports variation in tained under vacuum, and in the most common energy versus time for all wavelengths simulta- technique of electron impact (EI) ionisation the neously. For molecules to be infrared active they column effluent is bombarded with electrons. must be able to undergo a change in dipole Compounds absorb energy, which causes them moment with the transition to their excited to ionise and fragment in a characteristic and state. As a result, many compounds that are reproducible fashion. The resultant ions are symmetrical do not respond. focused and accelerated into a mass filter that Fourier transform refers to the mathematical allows fragments of sequentially increasing mass computation that converts the data from an to enter the detector stepwise. The mass filter intensity versus time plot into an intensity (% scans through the designated range of masses transmission) versus frequency spectrum. Each (usually up to about 700 a.m.u.) several times dip in the spectrum corresponds to light per second. The abundance of each mass at a absorbed, and can be interpreted as character- given scan time produces the mass spectrum, istic of specific functional groups in the mol- which can be summed and plotted versus time to Gas chromatography 497 obtain a total ion chromatogram (TIC). The MS Dual detector systems detector can be operated either in full scan mode (collecting all the ions within a given mass The simultaneous use of a combination of a range) or selected ion monitoring (SIM) mode, universal detector (FID) with a specific detector which collects only pre-selected masses charac- to monitor the effluent of a column can provide teristic for the compound under study. Sensi- useful information about the properties of func- tivity for the two modes of operation is quite tional groups and substituents in a molecule. different, 1–10 ng for full scan, increasing to The FID response is roughly dependent on the 1–10 pg in SIM because of the dramatic decrease number of carbon atoms in a molecule and is in background noise. The linear range is excel- quite predictable. However, the ECD response lent and often spans five or six orders of magni- varies widely for different compounds, is depen- tude. Recent advances in computer technology, dent on the electron-deficient part of the coupled with improved detector design, have compound, and is difficult to predict. The NPD revolutionised the use of the MS detector from a response of a compound depends to some extent research tool to one of routine application. on the number of phosphorus or nitrogen atoms This technique is described in more detail in in a molecule, but it also depends on their envir- Chapter 21. onment. Thus, by using the FID as a reference, and measuring the ECD or NPD response relative to it, another characteristic for identification is obtained in addition to retention behaviour (see Ion-trap mass spectrometer Fig. 18.15). In ion-trap mass spectrometers the production of Dual detector systems can be used in several ions in EI or chemical ionisation (CI) mode is ways. The column can be split at the detector achieved in pulses rather than continuously. The end and the effluent passed into two different fundamental difference is that all the solute ions detectors that operate in parallel. This approach generated over the entire pulse period are allows the most flexibility, since the choice of trapped in the detector and are then sequentially detectors is wide, and the effluent can be split in ejected in increasing mass number from the trap proportion to the sensitivity required from each into the electron multiplier. The addition of detector. For capillary columns this is accom- helium into the trap (133 mPa) contracts the ion plished easily with zero dead volume press-fit tee trajectory to the centre of the trap, where it is connectors, but it is a more complicated opera- further focused by the ring electrode to form tion for packed columns. Additional makeup gas dense ion packets that are expelled more effi- may be required to ensure a good flow through ciently than diffuse clouds, and thus greatly the detectors, and care should be taken to use improves resolution. The spectral patterns can be tubing of area smaller than or equal to the total quite different from those produced by mass area of the analytical column to avoid loss of filter spectrometers, and are often characteristic peak shape through refluxing at the detector. of the conditions under which the instrument is Alternatively, the GC oven houses two completely run, which makes comparison difficult between separate but identically matched columns, each instruments. However, because the ion collec- connected to a single detector. This is not an ideal tion period is longer, the sensitivity of the ion approach, as matching of columns is difficult and trap in full scan mode is similar to that obtained has to be checked at frequent intervals. Another in SIM on the average mass spectrometer. approach is to stack the detectors in series, and Furthermore, an improved mass range (some- some manufacturers deliberately provide detec- times up to several thousand a.m.u.) gives this tors in identical modules for this purpose. There type of detector many applications, particularly are limitations to the choice of possible detector for quantitative trace analysis, and for higher combinations, as the first detector must always be mass components. This technique is described in a nondestructive detector, such as the ECD, AED more detail in Chapter 21. or FTIRD. 498 Clarke’s Analytical Forensic Toxicology

TIC: 2201031.D 900000 21.75 C 800000 16.29

700000

600000

500000 12.58

Abundance 400000 14.16 300000

200000 14.40 A B 17.38 20.16 100000

4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.00 29.00

2201031.D\NPD1A 15.86 300000 C

250000

200000

150000 Abundance A 100000 B 11.23 50000 3.52 12.36

4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.00 29.00 Time

Figure 18.15 Analysis of basic drugs in blood using dual GC-MS and GC-NPD detection. A, caffeine; B, propy- phenazone; C, proadifen (used as internal standard for the analysis). Upper trace: GC-MS TIC chromatogram. Lower trace: GC-NPD chromatogram. Several nitrogen-containing drugs barely detected by GC-MS in scan mode are clearly detected by the NPD detector, indicating the high sensitivity of NPD. The lower chromatographic trace is much simpler, indicating the high specificity offered by NPD detectors. Note: there is a slight difference in retention times between the two columns because columns of different length were used for the analysis. (Reprinted from A. Beat and Werner, Gas chromatography with dual mass spectrometric and nitrogen-phosphorus specific detection: a new and powerful tool for forensic analyses, Forensic Science International, 1999, 102, 11, with permission from Elsevier.)

Specimen preparation Hydrolysis

Prior to chromatography, it is usually necessary Recovery of conjugated drug metabolites from to isolate the compound(s) of interest from biological fluids can be increased by hydrolytic either a biological matrix (plasma, urine, cleavage of the conjugate bond prior to extrac- stomach contents, hair and tissue) or some other tion. This offers a vast improvement in sensitivity matrix, such as soil, food and drink, tablets and for qualitative analysis, particularly from urine, other drug formulations, air or water. Removal of and is essential to identify drugs (e.g. extraneous material and concentration of the laxatives) that are excreted almost exclusively compounds of interest usually take place simul- as conjugated metabolites. However, reliable taneously. The high water solubility of some quantitative analysis of conjugated metabolites drug metabolites (e.g. glucuronide conjugates) requires that the unconjugated metabolite must requires chemical conversion to a less polar first be removed or quantified, and then the total entity to permit isolation from water-based (conjugated plus unconjugated) metabolite be samples, and a hydrolysis procedure is often measured after hydrolysis in a subsequent sepa- used for this purpose. rate procedure. For quantitative work, appro- Gas chromatography 499 priate standards that contain conjugated metabo- Isolation and concentration lites must be carried through the procedure to monitor the efficiency of the hydrolysis step. Protein precipitation If the analyte is present in blood in high concen- Enzymatic hydrolysis tration, a simple protein precipitation step often provides a suitable extract, although the possi- The use of a specific enzyme to cleave chemical bility of losing significant amounts of analyte bonds is the more specific of the two approaches, with the precipitate must be considered. Mixing but incurs additional cost and time. It also with a solution of mercuric chloride or barium provides cleaner extracts, and therefore prolongs sulfate readily precipitates plasma proteins, and the life of the chromatography column. There centrifugation provides a supernatant for direct are a number of commercial preparations of injection onto the chromatography column. Use purified glucuronidases and sulfatases harvested of perchloric or trichloroacetic acids (10%) is from different species. It is important to pay not advised, unless the resultant solution is attention to the pH and temperature optima of neutralised prior to injection. Dimethylform- the specific enzyme preparation. The best results amide is a good organic precipitation reagent are achieved by overnight hydrolysis at 37ЊC that is well tolerated by most GC stationary (C. Luckie et al., unpublished data); however, phases. Other organic precipitating agents are temperature-tolerant preparations allow heating methanol, acetone and acetonitrile, all of which up to 60ЊC, which permits relatively short (2 h) should be added in the proportion of two incubation times. volumes to each volume of blood. While the extract is still water-based, most columns with a Chemical hydrolysis high stationary-phase loading (5 lm film thick- ness) can tolerate the injection of 1 lL of water. This quicker and less expensive approach can If the column is not water tolerant, it is possible provide suitable extracts for chromatography for to evaporate small volumes of the supernatant to some analytes, although they are generally more dryness for reconstitution in a more suitable demanding in terms of clean-up procedures. solvent. Caution must be exercised for some Typically, strong mineral acids or alkalis are solutes when evaporating aqueous solutions used, often with boiling or treatment in a because volatile components may be lost, e.g. microwave or pressure cooker. Extracts must be amfetamine and methamfetamine in base form neutralised, otherwise the chromatography are relatively volatile. column deteriorates quickly. Care should be taken to ensure the stability of the analytes to the hydrolysis conditions. Vigorous hydrolysis Liquid–liquid extraction conditions often yield undesirable by-products or, if several compounds can be hydrolysed to a Liquid–liquid extraction is the most frequently single entity, preclude accurate identification of used method to isolate and concentrate solutes the original compound present. For example, for GC. The pH of the specimen is adjusted to both the acid and the enzymatic hydrolysis of ensure that the compounds to be extracted are benzodiazepines remove glucuronide conju- not ionised (basic for bases, acid for acidic gates, but acid hydrolysis also converts two or compounds). Bearing in mind that some portion three drugs to the same benzophenone of the aqueous acid or base will dissolve in the compound. Diazepam, temazepam and ket- solvent, the use of strong mineral acids or alkalis azolam are all converted into 2-methylamino-5- is not advised as this adversely affects column chlorobenzophenone; while this compound has performance. Best results are obtained with good chromatography characteristics, the acidic buffers (phosphate or acetate) and with approach is unsuitable for those applications ammonium hydroxide or basic buffers (borate), (such as forensic analysis) that require absolute using a 5:1 ratio of solvent to specimen. The identification of the drug ingested. solvent chosen should be sufficiently polar to 500 Clarke’s Analytical Forensic Toxicology partition the compound of interest without co- packing at the base of the reservoir. On intro- extracting excessive amounts of polar contam- duction of the sample matrix, the compounds of inants. For more water-soluble drugs, such as interest are withheld by the packing. Impurities b-blockers, the addition of 2–10% of a polar are then rinsed selectively from the column, and solvent (e.g. propan-2-ol or butanol) is helpful, the final elution releases the compound of or solid sodium chloride can be added to ‘salt interest (Fig. 18.16). out’ the analyte. If a derivatisation step is to be Evaporation followed by reconstitution in a carried out subsequently, the use of a solvent suitable solvent provides a clean, concentrated compatible with the derivatisation eliminates sample ready for analysis by GC. Bonded-phase the need for an evaporation step. Use of solvents packings that have been modified by the addi- with a higher density than the sample (e.g. tion of various functional groups are available. dichloromethane) can lead to difficulty in isola- The mechanisms of interaction for the matrix, tion of the organic phase. Purification of extracts analytes and packings are similar to those in by back extraction (re-extraction of the analytes liquid chromatography (see Chapter 19). Polar from the organic solvent at the opposite pH stationary phases preferentially retain polar followed by re-extraction into solvent at the analytes (normal phase) and are eluted with original pH) may be helpful for trace analysis. organic solvents, while nonpolar stationary The use of a small volume of solvent for the final phases preferentially retain nonpolar analytes extraction serves as a concentration step without (reversed phase) and are eluted with aqueous the need for separation and evaporation of the solvents. Ion-pair extraction uses a nonpolar organic phase. stationary phase and polar analyte, with a counterion added to the sample solution, and allows retention of the (now neutral) analyte by Solid–liquid or solid-phase extraction a reversed-phase mechanism. In ion-exchange Solid-phase extraction (SPE) uses a polypropyl- extraction, the adsorbent surface is modified ene cartridge with a small amount (200 mg to with ionisable functionalities. Analytes with 3 g) of high-capacity (1–20 mL) silica-based ionic charges opposite to those on the packing

Step 1 Step 2 Step 3 Step 4 Precondition the Load sample into Rinse with weak Rinse with medium- cartridge by washing it cartridge solvent to elute strength solvent with organic solvent weakly bound to remove product (mainly methanol) and water contaminants of interest

Figure 18.16 Steps involved in SPE. Gas chromatography 501 are retained. Solvents that contain counterions inserted through a septum into a vial that of greater strength are used to elute the analytes contains the sample, and the fibre is exposed by of interest from the tube. depressing the plunger either into the liquid or the headspace for 20–30 min. The retracted fibre is inserted into the injection port of the GC, Solid-phase microextraction and is desorbed when the plunger is depressed Solid-phase microextraction (SPME) requires no (Fig. 18.17). solvents or complicated apparatus and can The unit may be reconditioned and used 50 to concentrate volatile and nonvolatile compounds 100 times. For field analysis, adsorbed samples in both liquid and gas samples. The unit consists can be stored and transported in the needle of a fused-silica fibre attached to a stainless-steel sealed in a special container for subsequent plunger coated with a stationary phase (mixed analysis by GC (or LC). Pesticides recovered from with solid adsorbents as required). The plunger is water samples have been shown to be more

Plunger

Adjustable needle guide GC injector port Septum- piercing needle SPME fibre Vial containing SPME assembly Fibre inserted in sample placed over GC injector. injector. Desorption Plunger depressed and allowed to take place injector septum pierced. with transfer of desorbed analytes to GC column. SPME assembly Plunger depressed Fibre retracted placed over vial. to immerse SPME into needle guide Plunger depressed and fibre in sample. and removed vial septum pierced. Fibre held in from vial. sample to allow analytes to partition into the phase, coating the fibre.

Figure 18.17 Extraction using solid-phase microextraction (SPME). The SPME fibre is coated with a phase similar to the types of stationary phases used in GC. Analytes partition into the phase and hence are extracted from solution. It is also possible to carry out headspace analysis using SPME in which case the fibre is held in the headspace above the sample. 502 Clarke’s Analytical Forensic Toxicology stable when stored in this way than in water. The volumes of co-solvent to the SCF, to extract special small-volume injection liner fits any highly polar solutes with excellent efficiency. In model of chromatograph, and produces sharper contrast to the conventional extracting solvents, peaks because of the higher linear gas velocity, the fluid most often used in supercritical fluid with little or no backflush. Suitable stationary extraction (SFE) – supercritical carbon dioxide – phases are: is nonpolluting, nontoxic and relatively inex- pensive. Additionally, extractions are carried out • 100 lm dimethyl-PSX film for low-molecular- quickly at temperatures that avoid degradation weight compounds or volatiles, or a thinner of temperature-sensitive analytes and provide film (7 lm) for higher-weight semivolatile clean extracts with extremely high efficiency. compounds Several dedicated SFE analysers are available; • 85 lm polyacrylate film for polar compounds each consists of a gas supply, pump and • 65 lm film of dimethyl-PSX–divinyl benzene controller used to pressurise the gas, temperature- for volatile alcohols and amines controlled oven, extraction vessel, internal • for surfactants, 50 lm Carbowax-templated diameter regulator and collection device. The resin supply carbon dioxide is compressed to a • for trace level volatiles, a 75 lm selected pressure (e.g. 28 000 kPa) and its Carbowax–carboxen phase is suitable. temperature is adjusted (e.g. 50ЊC). As the super-

An alternative approach uses a small magnetic critical CO2 passes through the sample material, stir bar encapsulated in glass and coated with a the solutes are extracted to an equilibrium solu- layer of dimethyl-PSX. The bar is left to stir in bility level, typically about 10% (w/w). The the sample for 30–120 min and then removed gaseous solution that leaves the extractor is and placed in a thermal desorption tube. From passed through the pressure reduction valve, there, it is introduced onto the GC as described where the pressure (and thus the dissolving in the section on thermal desorption injectors power) of the CO2 is reduced. The solutes precipi- (p. 489). Both approaches give similar perform- tate in the separator, and the CO2 is recycled ance for higher-boiling compounds (Ͼ350ЊC), through the system several times until the but SPME is inferior for lower-boiling com- extraction is completed, when it is vented to pounds such as naphthalene and fluorene waste. (b.p. 218 and 298ЊC, respectively). Headspace analysis Supercritical fluid extraction This method of isolation is used for analytes A supercritical fluid (SCF) is a substance that is with volatility higher than that of the common maintained above its critical temperature and extraction solvents. A detailed description of the pressure, where it exhibits physicochemical technique is given on p. 311. properties intermediate between those of a liquid and a gas. Properties of gas-like diffusivity, Purge and trap gas-like viscosity and liquid-like density combined with a pressure-dependent solvating Purge and trap is a powerful procedure for power provided the impetus to apply SCFs to extracting and concentrating volatile organic analytical separation. The initial applications compounds from soil, sediment, water, food, most often involved isolation of flavours and beverages, etc. It is especially useful for poorly contaminant residues from food and soil. These water-soluble compounds and those with boiling have now been extended to the isolation of points above 200ЊC. The procedure involves drugs from blood and other aqueous-based bubbling an inert gas (nitrogen or helium) media by using adsorbents added in-line (such as through an aqueous sample or suspension at molecular sieves, diatomaceous earth, silica gel, ambient temperature, which causes volatile and so on) to filter proteinaceous material and organic compounds to be transferred into the adsorb water. It is possible, by adding small vapour phase. Alternatively, an inert gas is Gas chromatography 503 passed across the headspace of the sample able adsorbent tubes, the tube is fitted into a and onto the trap. The removal of volatile specially designed thermal desorption unit. The components from the headspace causes the adsorbents must have high capacity to remain sample–headspace equilibrium to be perturbed active during the entire sampling period, and such that further vaporisation of the volatile show an acceptable pressure drop during components occurs, resulting in concentration sampling. Ideally, a minimal amount of of sample in the trap. During the purge step, unwanted analytes should be absorbed, as these purge gas sweeps the vapour through a trap will contribute to the background noise. containing adsorbent materials that retain the volatilised compounds. Water vapour may be Tissues and hair removed by dry purging. The trap is rapidly heated to 5–10ЊC below the desorption tempera- Tissues and hair require treatment prior to drug ture. The valve is then switched to join the trap extraction to break down the biological matrix flow to the carrier-gas flow, and the trap heated and enable a good recovery of the drug. For solid to its desorption temperature for a fixed time. tissues, good results are obtained by incubation Adsorbent tubes are usually packed with of a portion of the tissue with a mixture of a multiple beds of sorbent materials, each one collagenase, a protease and a lipase in a buffer of more active than the preceding one, which suitable pH. For small amounts of tissues (100 allows compounds with a wide range of boiling mg), overnight treatment at room temperature points and polarities to be analysed simultane- suffices, although gentle agitation or occasional ously. During purge, the smaller and more mixing speeds up the process. Larger amounts of nonpolar solutes are readily carried down the tissue benefit from mechanical homogenisation beds, and since the carrier gas passes in the prior to incubation. For the analysis of hair, an opposite direction during the desorption phase, initial washing to remove residues from cosmetic the larger and more polar compounds do not products or environmental contaminants is come into contact with the innermost active recommended, followed by incubation with beds, from which their release may be difficult to either caustic alkali (for basic drugs) or mineral effect. acid (for acidic drugs). After adjustment of the pH, drug recovery can proceed by the usual procedures established for the specific Thermal desorption compounds under investigation. For additional This technique is used extensively for air moni- information see p. Chapters 6, 7 and 8. toring in industrial hygiene, environmental air, indoor air or source-emission monitoring and for examination of accelerant residues in arson Derivative formation investigations. Devices may be portable or fixed and of varying size. In some devices the trap is The main reasons derivatisation is performed integral whereas in others the trapping device is are: a removable tube packed with the adsorbent. • to permit analysis of compounds not directly This tube is then attached to a pump or a syringe amenable to analysis owing to inadequate for sampling. Air is pumped continuously volatility or stability through the device or adsorbent tube at a fixed • to improve the analysis by improving rate or, in the case of a syringe attached to the chromatographic behaviour or detectability. adsorbent tube, drawn up by hand. Components are concentrated onto the adsorbent beds. The To some extent the availability of stable polar arrangement of the beds may be the same as stationary phases in capillary columns and the described above for the purge and trap or it may use of temperature programming has negated be a single adsorbent. The direction of the flow the requirement for derivatisation, although it is is simply reversed during desorption. Analysis still widely used. Choice of reagent is based on requires a special interface to the GC. For remov- the functional group that requires derivatisation, 504 Clarke’s Analytical Forensic Toxicology the presence of other functional groups in the Alkylation. Typical derivatives are formed by molecule and the reason for performing the reac- replacement of an active hydrogen (carboxylic tion. Although the retention characteristics are acids and phenols) by an aliphatic or aliphatic- changed, the order of elution of a series of deriv- aromatic (i.e. benzyl) group. The reagents atives will be the same as that for the parent include 3 M HCl in butanol or N,N-dimethylform- compounds. The preparation of derivatives amide dimethyl acetal. modifies the functionality of the solute molecule Acylation is the conversion of compounds to increase (or sometimes decrease) volatility, that contain active hydrogen (1OH, 1SH, and thereby shortens or lengthens the retention and 1NH) into esters, thioesters and amides, time of a substance, or to speed up the analysis. respectively. Typical derivatising reagents The major derivatisation reactions for GC are include heptafluoropropionic anhydride (HFPA), silylation, alkylation and acylation. pentafluoropropionic anhydride (PFPA), and Silylation is usually used as an abbreviation for others. An example of a derivatising reaction is trimethylsilylation. Examples of derivatising presented in Fig 18.18. agents include: N,O-bis(trimethylsilyl)trifluoro- Derivatisation can improve resolution and acetamide (BSTFA); a mixture of BSTFA reduce tailing of polar compounds (hydroxyl, and trimethylchlorosilane (TMCS); N-methyl- carboxylic acids, hydrazines, primary amines trimethylsilyltrifluoroacetamide (MSTFA); N- and sulfhydryl groups). For instance, hydroxyl- methyl-N-(t-butyldimethylsilyl)trifluoro ated compounds often have long retention acetamide (MTBSTFA); and others. Typically times and column adsorption causes tailing, derivatives are formed by replacement of active which results in low sensitivity. However, they hydrogens from acids, alcohols, thiols, amines, readily form silyl ethers and these derivatives amides, ketones and aldehydes with the show excellent chromatography; sensitivity can trimethylsilyl group. often be improved by a factor of 10 or more.

CH 3 O

10 N 2 9 1 F F FO O F F F 6 3 5 4 7 8 ϩ 11 C N F C C C C O C C C C F H2N F F F F F F F

HFBA 7-Aminoflunitrazepam

CH 3 O

10 N 2 9 1 F F FOH 6 3 5 4 7 8 11 N F C C C CN C F F F F

7-Aminoflunitrazepam HFB derivative

Figure 18.18 Derivatisation of 7-aminoflunitrazepam with heptafluorobutyric anhydride (HFBA). Gas chromatography 505

Derivatisation can also help to remove the with enantiomeric amines. Excess reagent is substance peak away from interfering material. washed off with 6 M HCl and the organic phase Derivatives may also be used to make the is dried over magnesium sulfate. For chiral molecule amenable to detection by selective alcohols, (1R,2S,5R)-(Ϫ)-menthylchloroformate detectors (e.g. introducing halogens to the mol- (MCF) reacts well if pyridine is used as a catalyst. ecule can increase detectability by an electron capture detector, see Fig 18.18), or can be used to improve the fragmentation pattern of the Quantitative determinations compound in the mass spectrometer. Derivatisation reactions may be carried out during extraction (e.g. extractive alkylation), on Quantitative work usually requires some form of the dry residue after solvent extraction (e.g. silyl- sample preparation to isolate the drug from the ation), or during injection (e.g. methylation). In bulk of the sample and some degree of concen- choosing a suitable reagent, certain criteria tration or, more rarely, dilution. These processes must be used. A good reagent produces stable inevitably introduce a degree of analytical error. derivatives without harmful by-products that A further problem is caused by the difficulty of interact with the analytical column, in a reaction reproducibly transferring the same mass of all that is almost 100% complete. Poor reagents sample components to the GC column with cause rearrangements or structural alterations each injection. To compensate for these errors, it during formation, and contribute to loss of is usual to compare the response of the unknown sample during reaction. Most manufacturers of with the response of an added internal standard. derivatising reagents provide information on the The internal standard should be added as early as potential uses of each product, along with stand- possible in the assay process and should have ard operating instructions. Entire texts, such as chromatographic properties matching those of that by Blau and Halket (1993), are devoted to the solute as closely as possible, preferably with this topic. a longer retention time. It is often possible to obtain unmarketed analogues of drugs, or compounds specially synthesised for use as Chiral separations internal standards (e.g. a methyl addition or a Chiral compounds can be derivatised to improve halogen substitution). However, the internal their chromatographic characteristics, and the standard usually does not behave exactly as the enantiomers separated on a chiral stationary drug and careful control of variables, such as pH, phase. Both enantiomers behave similarly, is necessary. If a derivative is to be prepared, the provided that steric hindrance does not preclude internal standard should also be amenable to a reaction with one enantiomer. An alternative derivatisation. Use of an inappropriate internal approach is to use a chiral derivatising reagent standard can seriously affect precision (Dudley which, when reacted with enantiomers, pro- 1980). If a mass spectrometer is being used as the duces diastereoisomers that can then be separ- detector, then the ideal internal standard is a 2H- ated on a conventional stationary phase. As with substituted (deuterated) analogue of the drug, a enantiomers, diastereoisomers still produce number of which are readily available at reason- similar mass spectra, but they are resolved in able cost. Some examples of deuterated internal time by the chromatography column. This standards are shown in Figure 18.19. Potential approach is less expensive and also less restrict- disadvantages of deuterated internal standards ive, since a dedicated column is not required. include a possibility of chemical exchange and Care should be taken to ensure the enantiomeric loss of a label. In addition, the presence of too purity of the derivatising reagent, and to many 2H atoms may alter the chromatographic guard against racemisation during the reaction. properties of a labelled compound. The alterna- n-Trifluoroacetyl-1-propyl chloride (TPC) in tive solution is to use compounds labelled triethylamine and chloroform (or ethyl acetate) with stable isotopes such as 13C. Calibration is a commonly used chiral reagent that couples should include points of higher and lower 506 Clarke’s Analytical Forensic Toxicology concentrations than the sample, and quality and using the detector at a higher sensitivity assurance samples should be included at level. appropriate concentrations in frequently run When attempting a new analysis, it is advis- assays. able first to review published literature for a Peak measurement may be by peak height or method that can be copied or for a method that by the peak area obtained by integration. If the involves a similar type of compound and can be peaks show even a modest degree of tailing, use adapted. Column manufacturers’ catalogues or of peak area usually provides a more accurate websites are a useful source of information and quantitative result. A plot of the ratio of peak invariably show examples of separations height (or area) of the drug to internal standard performed with their columns, and manufac- versus concentration is a straight line with most turers typically have technical departments to detectors. Care should be taken in the prepara- offer specific advice. Data on boiling points and tion of standards to match the matrix to that of RI are also useful indicators, as is consideration the specimens, and to allow for any associated of the chemical structure of the substances salt or water of crystallisation in the calculation present in the sample. If the review is not of the concentration. The best results are helpful, a start can be made with a standard obtained when the amount of internal standard column, selecting either a 100% methyl-PSX used produces a peak response ratio of unity at capillary column (25 m with a 0.25 lm film or the mid point of the calibration range. an OV1-packed column (1.7 m with a 3% loading on 100–120 mesh support) and using standard flow conditions (1–2 mL/min helium for a capillary or 30–60 mL/min for a packed Optimising operation conditions to column). The oven temperature should be taken customise applications from 80 to 300ЊC at 10ЊC/min (or started at 200 or 250ЊC if only an isothermal oven is available). If sensitivity is an issue, it can be increased by Lower temperatures may be needed for increasing sample size, using a concentration substances which are relatively volatile. A solu- step, derivatisation, injecting a larger sample tion of the compounds of interest in ethanol or volume, selecting a different stationary phase methanol should be injected with the injector

Cl Cl

D D

N CH3 D N CH3 N N O D D O

Diazepam D5-Diazepam

HO HO HO D D D O O O N N N

CH3 CD3 CD3 HO HO HO

Morphine D6-Morphine D3-Morphine

Figure 18.19 Diazepam, morphine and their deuterated analogues. Gas chromatography 507 temperature set at 250ЊC. If a peak tails, derivati- columns and helium for capillary columns. sation or use of a more polar stationary phase Certain detectors impose restrictions on the should be considered. Fine-tuning is carried out choice of carrier gas, but an additional supply of once some peaks have been obtained. Having gas can be added to the column effluent to purge established the chromatography, the extraction the detector. Experimenting with higher flow and concentration steps can be determined. and a lower operating temperature (or vice versa) Manufacturers’ catalogues are again a useful can give rewarding results for the separation of source for both derivatisation and solid-phase compounds that elute closely. This effect is extraction procedures. particularly noticeable for two compounds that Good preventative maintenance is essential. have different polarities, as the retention of the The injector (or liner) should be cleaned period- more polar compound is influenced to a greater ically, and any glass wool changed regularly extent the longer it resides in the column (approximately every 100 to 1000 injections, (nonpolar compounds elute in boiling point depending on the quality of the extracts). For sequence). Conditions of constant flow improve capillary columns, the performance is improved the efficiency of late-eluting peaks and produce by periodically removing the first 5–10 cm of faster chromatography than do constant pres- capillary tubing (a retention gap could be sure conditions. considered for dirty samples), and for packed For a particular separation, the lowest temper- columns by replacing the glass wool and first few ature compatible with a reasonable analysis time centimetres of packing. It is advisable to monitor should be used. In general, retention times performance by selecting certain performance double with each 20ЊC decrease in temperature. criteria (e.g. a certain response size, running a If the time is excessive, it is generally better to standard test mix or amount of acceptable separ- reduce the stationary phase loading or use a ation between two closely eluting components) shorter column than to increase the column to indicate when maintenance is required. The operating temperature. There is a maximum manufacturer’s instructions for cleaning detec- temperature at which a column can be operated tors should be followed. and there is also a minimum temperature below The presence of traces of contaminants in the which efficiency drops sharply. Manufacturers carrier gas supply shortens the column life dras- give the temperature operating ranges for each of tically, and also causes detector deterioration. In- their stationary phases (see Table 18.2). For GLC, line filters (to remove oxygen, hydrocarbons, the stationary phase must be a liquid at the etc.) and molecular sieves (to remove water temperature of operation, and if a column is run vapour) are strongly recommended, and the use at too low a temperature to obtain longer reten- of stainless-steel gas tubing minimises further tion times the stationary phase may still be contamination. Carrier-gas flow should be opti- in the solid or semi-solid form. When using mised for a particular column and a particular temperature programming, experimentation carrier gas. This is most important for capillary with a faster initial ramp followed by a slower columns. Fig. 18.12 shows the relationship subsequent ramp or an isothermal period can between efficiency expressed as the HETP versus help resolve problematic separations. carrier-gas velocity (van Deemter plot) for a 25 m Efficiency can also be improved by decreasing by 0.25 mm i.d. WCOT OV-101 column. Modi- the column diameter or increasing the column fying the nature of the mobile phase in GC has length. The resultant increase in analysis time very little effect compared with that observed (particularly if the flow must be reduced to with HPLC or thin-layer chromatography (TLC) accommodate the increased pressure demand and, in general, affects efficiency rather than imposed by a narrower column), can usually be selectivity. Nitrogen gives higher efficiency, but offset by using a slightly higher operating at the expense of longer analysis time, while the temperature (temperature increases affect reten- less dense, but more hazardous, hydrogen gives tion time much more than do increases in gas slightly lower efficiency, but faster analysis. In flow). Reducing the diameter of a capillary practice, nitrogen is usually used for packed column markedly increases efficiency, but the 508 Clarke’s Analytical Forensic Toxicology retention time remains constant only as long as emerge before the main component of a the same phase ratio is maintained. Therefore, mixture, while they may be lost completely in unless there is a simultaneous reduction in film the tail if they elute just after the large peak. thickness, retention increases in direct propor- Early peaks are also sharper and thus, for the tion to the phase ratio. same peak area, higher – an effect that can The solvent used for the sample can some- contribute enormously to the successful times produce unexpected derivatives that give detection of trace substances. different retention times (traces of acetic anhyd- ride that remain in butyl acetate avidly derivat- ise primary amines at room temperature). An Specific applications inert nonpolar solvent should be used if possible to minimise the co-extraction of unwanted con- taminants. Acetone, other ketones, ethyl acetate Amfetamines and other stimulants and carbon disulfide readily form derivatives with primary amines and should be avoided. Amfetamines are basic drugs that require strongly The choice of injector type and injection alkaline conditions to be extracted from aqueous solvent also plays an important part in the chro- solution. These conditions are too basic to extract matography. A solvent volume should be chosen the phenolic metabolites, but these can be recov- that does not expand to exceed the capacity of ered at pH 8 or 9 and the extracts combined prior the injector (see Table 18.4), otherwise backflush to chromatography. For high sensitivity, back and irreproducible results are obtained. Split extraction into dilute sulfuric acid (0.05 M) is a injection significantly reduces the amount of useful clean-up procedure. When using packed solvent and associated contaminants that enter columns, derivatives are almost always required the column and, although the analyte response for the primary and secondary amines, since the is reduced, the improvement in the signal-to- peaks tail badly. Suitable derivatives are acetyl, noise ratio often results in enhanced sensitivity. trifluoroacetyl, pentafluoropropionate or TMS The use of a selective detector, such as an ECD (see p. 503, Derivative Formation). With capillary (with the preparation of a strongly responsive columns, derivatives are used most often to derivative if appropriate), can improve sensi- improve mass spectral patterns or to modify the tivity typically up to 100-fold. Similarly, separation of compounds that elute closely. For switching from full scan to selected-ion moni- hydroxylated metabolites, derivatisation is toring in MS improves the sensitivity, usually by invariably required to achieve acceptable chro- a factor of 10. However, selective detectors matography. Care must be taken to avoid drug should not be used as a substitute for cleaning up loss during solvent evaporation, which can be of sample extracts, as loading contaminants obviated by adding a small amount of concen- onto the column affects the chromatography trated aqueous acid (20 lL of 6 M HCl) to the adversely, even if the selective detector does organic solvent. Unless otherwise stated, GC not respond to the compounds. Increasing the retention data and mass spectral data are identical detector temperature may also improve for both d- and l-(ϩ and Ϫ) enantiomers. To sensitivity. differentiate enantiomers (such as d- and l- Fronting or splitting of peaks indicates column methamfetamine or amfetamine), a chiral overload. If the detector sensitivity permits, the column or chiral derivatising reagent is required best option here is to inject a smaller sample (Cody and Schwarzhoff 1993). At present, all volume (or a more dilute sample), rather than amfetamine- or methamfetamine-producing to increase the column loading or diameter, drugs (aminorex, amfetaminil, clobenorex, ethyl- otherwise efficiency is also affected. amfetamine, fencamine, fenethylline, fenpro- If trace levels of solutes are sought in the pres- porex, mefenorex, prenylamine, benzfetamine, ence of a preponderant component, a number of dimethylamfetamine, famprofazone, furfenorex) stationary phases of differing polarities should are racemates (with the exception of l-selegiline, be tried. Trace impurities are seen easily if they l-methamfetamine and dexamfetamine). Stereo- Gas chromatography 509 inversion does not occur in humans (Nagai et al. extraction solvent should be moderately polar 1991). Drugs that are metabolised to amfet- (ethyl acetate is appropriate), and TMS deriva- amines, but are not themselves classified as such, tives form easily in 20–30 min at 60ЊC using 50% are also listed BSTFA with 1% TCMS in acetonitrile. These derivatives markedly improve peak shape and sensitivity. All compounds except 7-aminonitr- Antidepressants azepam show electron-capture responses with high sensitivity. However, quantification with Antidepressants (tricyclics, selective serotonin ECD is problematic as it has a narrow linear reuptake inhibitors (SSRIs), monoamine oxidase range, and a multiple-point calibration is essen- inhibitors (MAOIs)) can be extracted readily tial. Alternatively, for most compounds a nitro- under mildly basic conditions (pH 10) into many gen detector (NPD) gives adequate sensitivity solvents, such as ethyl acetate, hexane, diethyl with a much-improved linear range, although ether. Less polar solvents, such as hexane, limit it is not advisable to make TMS-derivatives if the extraction of hydroxylated metabolites. An using this detector. MS detection is required to acidified (0.05 M H2SO4) back extraction is a confirm the identity useful clean-up procedure where sensitivity is important. Chromatography of primary and secondary amines is poor on packed columns, Hydrolysis of benzodiazepines but is adequate on well-maintained capillary (preparation of benzophenones) columns, particularly those of low-medium polarity such as PSX-5. Some authors prefer to An aqueous solution (or urine) should be boiled chromatograph the secondary amines and with concentrated hydrochloric acid (1 part to hydroxylated metabolites as acetylated deriva- 10 parts urine or solution) for 30–60 min, tives, prepared by heating the dried residue with cooled, and neutralised with solid KHCO3 or the acetic anhydride and pyridine (3:2, v/v; Maurer pH adjusted to 8–9 with 10 M KOH. It is then and Bickeboeller-Friedrich 2000). Others employ mixed with an equal volume of petroleum ether an enzymatic hydrolysis procedure to improve for 10 min and then centrifuged, and the upper recovery of both parent drug and metabolites, organic phase is evaporated to dryness at 60ЊC. although the additional sensitivity gained is The reconstituted extract can be used for GC or often negated by the increased analytical time in other analytical procedures such as TLC. Not all the emergency setting. Acid hydrolysis is benzodiazepines make benzophenones when quicker, but some relevant compounds are hydrolysed with acid, and a number of other destroyed under these conditions. degradation products are furnished. The a-OH- metabolites of alprazolam, brotizolam and tri- azolam are partly altered by the elimination of Benzodiazepines formaldehyde. Hydrolysis products of bis- desethylflurazepam and di-OH-tetrazepam are The analysis of benzodiazepines in biological dehydrated; OH-bromazepam, lorazepam and specimens is hampered by their high potency oxazepam form artefacts by rearrangement; and resultant low-plasma concentrations, and by the nor-metabolites of clobazam are cleaved their interconnected metabolic pathways. and rearranged to benzimidazole derivatives; Several benzodiazepines appear in urine almost tetrazepam, and its two hydroxylated metabol- exclusively as glucuronide-conjugated metabol- ites, are transformed into a pair of cis- and trans- ites, and these can be hydrolysed with isomeric hexahydroacridone derivatives. glucuronidase (1000 U glucurase/mL of urine at Since the metabolism of benzodiazepines is 60ЊC for 1–2 h), although some can degrade with complex, assays that convert drugs and metabol- prolonged heating. Extraction can be performed ites into hydrolysis products are not ideal, since at any pH between 3 and 12, but basic extracts they do not permit unequivocal identification of (pH 9 to 11) give cleaner chromatograms. The the parent compound. After acid hydrolysis, care 510 Clarke’s Analytical Forensic Toxicology must be taken to ensure that the acid is required, and most laboratories use TMS as the neutralised prior to extraction or before injecting derivatising reagent. the solvent onto the chromatograph, otherwise For analysis of cannabis metabolites, hydroly- the column deteriorates rapidly. sis of conjugates with 10 M potassium hydroxide is usually performed on urine prior to weakly acidic extraction (pH 6.5); TMS is the Narcotic analgesics, opiates and opioids derivative of choice. Phencylidine (PCP) analysis is complicated by the low concentration present, Many laboratories perform specific assays for although extraction is straightforward and opiates for federal or legal purposes; these are derivatisation is only required for metabolite generally limited to codeine, morphine and measurement (Nakahara et al. 1997). Chromato- more recently 6-monoacetyl morphine (MAM; graphic confirmation of lysergide (LSD) is Paul et al. 1999). However, for clinical purposes a hampered by the low concentrations and acidic wider range of analytes is desirable and can nature of the metabolites, which necessitates include codeine, dihydrocodeine, hydrocodone, both derivatisation (TMS) and tandem MS hydromorphone, oxycodone and oxymorphone. (Nelson and Foltz 1992). All assays involve a hydrolysis step (acidic or enzymatic; see p. 498 for an evaluation of these) to cleave the glucuronide conjugates, followed by a basic extraction (often using solid phase or References acidic back extraction for cleanliness). Derivati- sation is possible with a number of reagents (PFP, TMS, TFA or AC derivatives are the most K. Blau and J. Halket (eds), Handbook of Derivatives for Chromatography, 2nd edn, New York, Wiley, 1993. common; Maurer and Pfleger 1984; Chen et al. L. A. Broussard et al., Simultaneous identification and 1990; Grinstead 1991). The derivatising reagent quantitation of codeine, morphine, hydrocodone, is selected on the basis of personal preference for and hydomorphone in urine as trimethylsilyl and a desired separation or the formation of unique oxime derivatives by gas chromatography–mass ions on MS fragmentation. Analysis of hydro- spectrometry, Clin. Chem., 1997, 43, 1029–1032. morphone, oxycodone and oxymorphone is B. H. Chen et al., Comparison of derivatives for deter- complicated by the possibility that several mination of codeine and morphine by gas chro- structurally different derivatives will form in matography/mass spectrometry, J. Anal. Chem., non-reproducible proportions from the 1990, 14, 12–17. tautomerisation of the enol and keto forms. J. T. Cody and R. Schwarzhoff, Interpretation of methamfetamine and amfetamine enantiomer However, these compounds can be stabilised in data, J. Anal. Toxicol., 1993, 17, 321–326. their keto forms by incubating with hydroxyl- K. H. Dudley, Trace organic sample handling, in amine or methoxyamine–pyridine, and then Methodological Surveys Sub-series (A), E. Reid (ed.), yield only a single derivatised oxime product Chichester, Ellis Horwood, 1980, p. 336. (Broussard et al. 1997; Meatherall 1999). G. F. Grinstead, A closer look at acetyl and pentafluoro- propionyl derivatives for quantitative analysis of morphine and codeine by gas chromatography/ Non-amfetamine stimulants and mass spectrometry, J Anal. Toxicol., 1991, 15, hallucinogens 293–298. H. H. Maurer and J. Bickeboeller-Friedrich, Screening Non-amfetamine stimulants and hallucinogens procedure for detection of antidepressants of the selective serotonin reuptake inhibitor type and have a variety of clinical and toxic actions. their metabolites in urine as part of a modified Extraction of cocaine is straightforward under systematic toxicological analysis procedure using basic conditions, and most metabolites, except gas chromatography–mass spectrometry, J. Anal. benzoylecgonine, can be detected in the clinical Toxicol., 2000, 24, 340–347. setting without derivatisation. For regulated H. H. Maurer and K. Pfleger, Screening procedure for testing, quantification of benzoylecgonine is the detection of opioids, other potent analgesics Gas chromatography 511

and their metabolites in urine using a computerized taneous detection of codeine, morphine, and 6- gas chromatographic–mass spectrometric tech- acetylmorphine in urine, Clin. Chem., 1999, 45, nique, Fresenius Z. Anal. Chem., 1984, 317, 42–52. 510–519. W. O. McReynolds, Characterization of some liquid R. P. W. Scott, Chromatographic Detectors – Design, Func- phases, J. Chromatogr. Sci., 1970, 8, 685–691. tion, and Operation, Chromatographic Science R. Meatherall, GC-MS confirmation of codeine, Series, Vol. 73, J. Cazes (ed.), New York, Marcel morphine, 6-acetylmorphine, hydrocodone, hydro- Dekker, 1996. morphone, oxycodone, and oxymorphone in urine, J. Anal. Toxicol., 1999, 23, 177–186. A. C. Moffat et al., Optimum use of paper, thin-layer and gas–liquid chromatography for the identifica- Further reading tion of basic drugs. I. Determination of effective- ness for a series of chromatographic systems, K. Blau and J. Halket (eds), Handbook of Derivatives for J. Chromatogr., 1974a, 90, 1–7. Chromatography, 2nd edn, New York, Wiley, 1993. A. C. Moffat et al., Optimum use of paper, thin-layer K. Grob, Split and Splitless Injection in Capillary Gas and gas–liquid chromatography for the identifica- Chromatography, Heidelberg, Hüthig, 1993. tion of basic drugs. III. Gas–liquid chromatography, K. Grob, On-Column Injection in Capillary Gas J. Chromatogr., 1974b, 90, 19–33. Chromatography, Heidelberg, Hüthig, 1993. T. Nagai and S. Kamiyama, Simultaneous HPLC H. H. Hill and D. G. McMinn (eds), Detectors for analysis of optical isomers of methamfetamine and Capillary Chromatography, New York, Wiley, 1992. its metabolites, and stereoselective metabolism of W. Jennings et al., Analytical Gas Chromatography, 2nd racemic methamfetamine in rat urine, J. Anal. edn, London, Academic Press, 1997. Toxicol., 1991, 15, 299–304. K. Jinno, Chromatographic Separations Based on Y. Nakahara et al., Hair analysis for drugs of abuse. Molecular Recognition, New York, Wiley, 1997. XVII. Simultaneous detection of PCP, PCHP, and D. Rood, A Practical Guide to the Care, Maintenance, and PCPdiol in human hair for confirmation of PCP use, Troubleshooting of Capillary Gas Chromatographic J. Anal. Toxicol., 1997, 21, 356–362. Systems, 2nd edn, Heidelberg, Hüthig, 1995. C. Nelson and R. L. Foltz, Determination of lysergic R. P. W. Scott, Chromatographic Detectors – Design, Func- acid diethylamine (LSD), iso-LSD, and N-demethyl- tion, and Operation, Chromatographic Science LSD in body fluids by gas chromatography/tandem Series, Vol. 73, J. Cazes (ed.), New York, Marcel mass spectrometry, Anal. Chem., 1992, 64, Dekker, 1996. 1578–1585. D. Stevenson and I. D. Wilson (eds), Sample Preparation B. D. Paul et al., A practical approach to determining for Biomedical and Environmental Analysis, New York, cutoff concentrations for opiate testing with simul- Plenum Press, 1994.