Media.rsc.org

Unilever
1. Mass spectrometry
Mass spectrometry in some form is familiar to many people whose chemistryeducation has gone beyond compulsory education.
The theory
Diagrams of simple mass spectrometers, such as Fig. 1, are common. The principlesof such spectrometers are quite straightforward: a sample is introduced into the spectrometer and vaporised; ions of charge z (where z is a multiple of the charge on an electron) areproduced by bombarding the sample with electrons in the ionisation chamber; the ions are accelerated by an electric field so that they have similar kineticenergy; the ions of mass m are deviated by a magnetic field such that ions of low orhigh m/z value strike the sides of the spectrometer, but ions of one particularm/z value continue along the spectrometer body; by varying the magnetic field strength all ions are sequentially focused into thedetector; and the ions are detected and the mass spectrum plotted.
Unilever
Because the ions have to travel distances from 25 cm up to maybe 3 m inside themass spectrometer it is important that they are not scattered or otherwise interferedwith by other species – ie by reaction with neutral molecules. This is most likely tooccur when air is present. This problem is overcome by subjecting the inside of thespectrometer to a high vacuum, typically 1.33 x 10-5 Nm-2 (10-7 mmHg) pressure.
(1 Nm-2 ≡ 1 Pa; 1 torr ≡ 1 mmHg; 1 torr ≡ 1.33 x 10-2 Nm-2). There are two waysof determining the m/z values of the ions produced: the accelerating electric field canbe kept constant and the magnetic field scanned (in practice an electromagnet isused); or the magnetic field can be kept constant and the electric field scanned. It isusually the magnetic field that is scanned.
If masses are required to one atomic mass unit (one dalton) a single focusing instrument such as the one described will suffice. However, an accuracy of 1 in 106– ie an accuracy that allows determination of precise atomic composition rather thannominal mass – can be achieved by using a double-focusing spectrometer. Theprinciples for this type of spectrometer are identical to those for the single-focusinginstrument, but include an electric sector before separation by a magnetic field(Fig. 2). Ions with a similar kinetic energy (ke) enter the electrostatic analyser and arefocused into a narrower energy range. The magnetic field can then separate the ions,which have a much smaller energy range, giving a much greater resolving power.
Practical considerations
The sample
Mass spectrometry is about 1000 times more sensitive than infrared (IR) or nuclear
magnetic resonance (NMR) spectroscopy. Only a microgram or less is required to
record a mass spectrum (many modern spectrometers require nanogram samples).
Different methods for introducing the sample in the vapour phase are used.
Gaseous samples can be allowed to diffuse into the spectrometer – a 10-5 dm3 Unilever
(10 microlitre) sample would be ample for this. Volatile liquids can be injected in,but using rather smaller volumes, because the sample will vaporise under the lowpressures present in the spectrometer.
Involatile liquids and solids are vaporised by placing the sample on a ceramic tip or in a glass capillary made of disposable soda-glass or reusable quartz (which has tobe heated to very high temperatures to clean – ie to remove any residue) or in a metalcrucible (typically 5 mm deep and 2 mm in diameter). These are inserted into thespectrometer (right up to, but not inside, the ion production area itself) and aresubjected to temperatures of up to 300 °C (Fig. 3). The vapour is then allowed todiffuse into the ion production area. Solids dissolved in a solvent can be used byputting the solution into a crucible and allowing the solvent to evaporate.
The ionisation chamber of the mass spectrometer is heated to 150–250 °C to ensure that the vaporised sample remains in the gas phase. Thermally stablenon-polar organic molecules – eg perfluorokerosene (fully fluorinated kerosene) –with masses up to 1000 daltons can generally be vaporised at temperatures below300 °C. If polar groups are present (eg OH, COOH and NH groups) the volatility decreases, and molecules with moderate polarities will only vaporise readily if theirmasses are below 500 daltons.
The rate of production of vapour is important – too slow and insufficient ions will be produced to obtain an appreciable signal; too rapid and the dominant ions willsaturate the detector and information regarding relative ion intensities will be lost. Ifthere are too many ions in the ionisation chamber of the spectrometer ion moleculeinteractions might occur, forming ions with a mass greater than the mass of thesample molecules.
Ion production
The method of ionisation most commonly discussed is electron impact (EI). Electrons
are ‘boiled off’ from a heated filament, which is made the cathode with respect to an
anode set typically at +10 to +70V (Fig. 4). As the electrons accelerate towards the
anode they can collide with the vaporised sample. Their energies are therefore up to
70 eV (1 eV ≡ 96.5 kJ mol-1; 70 eV ≡ 6750 kJ mol-1).
Collision between the high energy electron and the sample ‘knocks out’ an electron from an electron orbital, generally from the highest energy level: Unilever
(Multiple charged ions can be formed. Doubly charged ions are detected at half theirmass value on the final spectrum, which has mass/charge as its horizontal axis.) Some negative ions are formed (less than 0.1 per cent of the positive ions formed), and those that are formed are attracted away from the electric sector and magnet.
However, it is possible to set the spectrometer to monitor negative ions by reversingthe polarity of the repeller.
Ionisation of an element requires from 381 kJ mol-1 (francium) to 2370 kJ mol-1 (helium) (ie 4–25 eV), and organic molecules need about 600–1000 kJ mol-1 (7–10eV). With organic molecules the energy of the ionising electrons is so great comparedwith the bond energies in the molecules that fragmentation is possible – nb thestrongest common single bond (bond energy 485 kJ mol-1, 5 eV) is C–F. Some of theresidual energy (possibly up to 600 kJ mol-1) of the fast moving electron might also betransferred to the ion as internal energy.
When an organic molecule forms a positive ion, it becomes a radical-cation because one electron has been removed from a pair in a filled orbital.
By convention, the radical electron is omitted from the representation of these If the radical-cation fragments, the molecular ion can lose either a radical or a neutral molecule eg the butyl ethanoate molecular ion (radical-cation) fragments asfollows: CH COOCH CH CH CH + → CH C+=O + C H O CH COOCH CH CH CH + → C H + + CH COOH Unilever
The remaining ions can fragment further to give the spectrum shown in Fig. 5. Themolecular ion peak is not significant because the ion is unstable.
The ionising electrons for this spectrum had energy 70 eV. These cause more fragmentation according to equation 1. The peak at m/z = 43 is due to CH C+=O, and the one at m/z = 56 due to C H +. If ionising electrons of energy 15 eV had been used, the peak at m/z = 56 would have been the most abundant – ie fragmentationwould predominantly have been according to equation 2. However, in each case themolecular ion would not be significant.
So, although it is possible to obtain a unique spectrum which will act as a ‘fingerprint’ for an organic compound, all the operating conditions (including theenergy of the ionising electrons) must be quoted. This is especially true if thespectrum obtained is to be compared with library spectra for identification.
Ionisation using lower potentials often gives more detailed information about the molecule because less energy is transferred to the sample and the molecular ion (M+)can become more significant. A disadvantage of this method is that fewer ions areproduced. One way of increasing the amount of molecular weight informationobtained on a compound is to use a ‘soft’ ionisation technique such as chemicalionisation (CI).
The molecular ion and the arrangement and sizes of the peaks formed by the breaking apart (the fragmentation pattern) of an organic compound can often be usedto identify it. However, there are cases when electron impact ionisation isinappropriate because it is impossible to ionise the molecule without it breaking up.
The molecular ion can usually be observed by using other ionisation techniques,such as CI and FAB (see below). Other, more elaborate methods exist, such as laserdesorption, secondary ion mass spectrometry (SIMS), electrospray and californiumplasma desorption, which are compatible with large, polar molecules. (Descriptionsof these techniques can be found in the bibliography.) Deflection of the ions
The ions are accelerated through a series of plates set at increasingly negative
potentials and emerge with a broadly similar energy, before passing into the electric
Unilever
sector where they are deviated so that the emergent ions all have a more sharplydefined range of energies. Deflection in the electric field is independent of mass, butfor a given charge it is proportional to the energy of the ions.
In the radial electrostatic field within the analyser the ions follow a circular path of radius r, broadly given by the equation: V = electrostatic potentialE = energy of ion The ions continue in a straight path outside the influence of the electric field, until they enter the magnetic field where they are deflected so that only the ions of aspecific mass to charge ratio will continue on towards the detector, (Box).
Dependence of ion deflection on accelerating
voltage and magnetic flux density
The kinetic energy of the ions after being accelerated is given by: m = mass of the ion z = charge on the ion v = velocity of the ion The force on the ions in the magnetic field is described by: The force on a body as it accelerates towards its centre of curvature is: r = radius of circular path.
Combining equations 2 and 3 we get Bzv = mv2, so v = Bzr Substituting equation 4 into equation 1, 1⁄2 m B2z2r2 = zV thus m = B2r2 So, for a given m/z value, the radius of curvature of the deflected ion isdependent on the magnetic field strength and on the accelerating voltage.
Unilever
The mass spectrum can be obtained by varying the accelerating voltage, V, or by varying the magnetic field (magnetic scanning). Voltage scanning can be done at highspeed but this usually gives a distorted spectrum – the relative abundances of thefragments decreases as their mass increases. Magnetic scanning is the norm althoughit is a little slower, because it is restricted by the response time of the analysermagnet.
The double-focusing mass spectrometer is so precise that the relative mass of a compound can be determined to such an accuracy that its molecular formula can beassigned. The molecular ion peak of the unknown in Fig. 8 is 122.036776, which canbe assigned the molecular formula C H O if it is assumed that only carbon, hydrogen, nitrogen and oxygen are present (Table 1). However without furtherinformation the precise isomer cannot be identified.
Some formulae corresponding to nominal m/z = 122 Formulae
Actual mass
These are based on the following relative atomic masses: Ion detection
Many modern mass spectrometers use electron multipliers (Fig. 6). The ions strike a
plate (the conversion dynode) made of a material (eg a copper/beryllium alloy) that
emits electrons when struck by energetic particles . This is set at ca –1.4 kV.
Secondary electrons are emitted from it, and they are accelerated and focused onto
the second and subsequent dynodes, which are set at potentials progressively closer
to earth. At each dynode there is an increase in the number of electrons emitted, such
that at the end of the multiplier a gain of approximately 106 has been achieved. (A
multiplier might contain 10 –14 dynodes.) Since ions are sharply focused onto one
position the conversion dynode degrades after a period of time and must then be
replaced.This occurs after perhaps two years on an instrument that is in daily use.
Unilever
Schematic diagram of an electron multiplier The amount of information obtained from a mass spectrometer is so vast that it is only realistic to capture the data using a microprocessor. The signal from the electronmultiplier is fed to an analogue to digital converter, then stored in a computer. Thecomputer stores the mass value at the centre of each ion distribution, and amaximum value which corresponds to the number of ions of that mass (Fig. 7). Thecomputer can then plot mass spectra. If plotted graphically the computer will set m/zvalues on the x-axis and relative abundance on the y-axis, setting the most abundantion as 100 per cent abundance (Fig. 8). The intensities of other ions are then plottedrelative to this peak.
Recording of data from the mass spectrometer Unilever
Benzaldehyde, 2-hydroxy-
Mass spectra of an unknown and five isomers of the samemolecular formula (continued overleaf) The vacuum system
It is vital that the path of the ions in the spectrometer is not affected by residual gas
molecules or by neutral species formed during the fragmentation or deflection of
particles. The average distance travelled by ions between successive collisions is
known as the mean free path, and is of the order of several hundred metres when the
internal pressure of the spectrometer is 10-5 Nm-2. In modern spectrometers two types
of pumps are used to achieve this low pressure. A rotary, or backing, pump reduces
the pressure to approximately 10-2 Nm-2, in tandem with a diffusion pump (which
traps gas molecules in the droplets as high speed jets of oil vapour condenses), or a
turbomolecular pump (which has a powerful fan that turns very quickly).
Unilever
Benzaldehyde, 3-hydroxy-
Benzoic acid
Ions that do not reach the detector can pick up an electron from the walls of the spectrometer, and if of low relative mass they become gaseous and can be removedby the vacuum line, along with any particles that were not ionised in the first place. Ifthe species are involatile, they condense on the inside walls of the spectrometer andare periodically removed by surrounding the instrument with electric band heatersand vaporising them off. Heavy or stubborn deposits have to be removed bymechanically stripping the instrument down and cleaning it. Fortunately, this doesnot happen very often.
The ionisation chamber of the spectrometer is particularly susceptible to contamination by condensed samples, therefore it is likely to need to be strippeddown and mechanically cleaned every three months.
Unilever
2,4,6,-Cycloheptatrien-1-one, 2-hydroxy-
2,5-Cyclohexadiene-1,4-dione, 2-methyl-
Calibration
If low resolution data is sufficient, perfluorokerosene, for example, (containing fully
fluorinated hydrocarbon molecules – ie no hydrogen atoms remain) can be used as a
reference to calibrate the instrument. Many peaks are observed, separated by 12 or
50 mass units (Fig. 9). Once calibrated, the spectrometer can function without
recalibration for about one month.
More detailed information (higher resolution data) can be obtained by putting a reference material in with the sample so that the instrument can be calibrated at thesame time as analytical data are obtained. The reference will depend on the mass ofthe compound, but is likely to be perfluorokerosene with a mass range appropriate tothe approximate mass of the unknown. (Mass ranges of perfluorokerosene are Unilever
available in the same way as boiling point ranges of petroleum ether.) The peaks dueto the reference material are used for calibration, and are then subtracted from thefinal spectrum by a computer (the data system). It is unlikely that the reference peakswill coincide with the sample peaks because fluoro compounds tend to have masseslower than the nominal mass, whereas most other compounds have higher masses.
Mass spectrum of a perfluorokerosene, used for calibrating amass spectrometer Other ionisation techniques
Electron impact frequently fails to yield the information required from substances. Avariety of alternative methods for ionising substances exist.
Chemical ionisation (CI)
In this technique a reagent gas such as methane, methylpropane or ammonia is
ionised by electron bombardment and is then allowed to react with a neutral
molecule to produce a molecular ion, eg
Unilever
The reactant ion, CH + in this case, then reacts with the sample gas and In this way positive ions are produced, but these are one mass unit higher than the parent molecule. The ions produced almost always have less internal energy thanions formed by electron impact, and thus fragment less as a consequence.
In practice the reagent gas is allowed into the ion chamber at a partial pressure of 102 Nm-2 (Pa). The amount of sample in the ion source can be very small – on thenanogram scale. The bombarding electrons come from a hot filament, and theirenergy is in the range 9650–48250 kJ mol-1 (100–500 eV) – higher than for electronimpact. Statistically the reagent gas is more likely to be ionised than the sample.
The choice of reagent gas depends on the ease of fragmenting the sample. The internal energy of the species MH+ decreases in the order CH + > C H + > NH +.
This results from the relative strength of the bonding between the reagent gas and theproton it transfers to the sample: ammonia bonds most strongly and methane least.
Consequently, if the ammonium ion is the protonating agent it transfers very littleenergy to the sample. Conversely, the CH + ion readily protonates samples and the energy liberated is transferred as the internal energy of the sample – it is then likely tofragment more extensively. However, one disadvantage of using methane ormethylpropane as the reagent gas is that the spectra obtained are complicated abovethe (M+H)+ molecular ion because adducts such as (M+CH )+ and (M+C H )+ are The difference between the spectra obtained by electron impact and chemical ionisation is immediately apparent in Figs. 11 a and b, where the spectra of Tenormin(Fig. 10) a drug produced by ICI, used for treating heart conditions are shown.
Figure 10
Under CI conditions equal abundances of positive and negative ions are usually produced. The processes involved in negative ion formation are complicated, but oneway that they can be formed is: The molecular ion peaks obtained by CI are predominantly (M+H)+ and (M–H)–.
One major advantage of negative ion chemical ionisation is that some compoundscan be detected at pico- (10-12) or femto- (10-15) gram levels.
Clearly the potentials inside the spectrometer have to be altered to enable negative ion spectra to be measured, but good, reproducible data can be recorded.
Unilever
(M+H)+ – sometimes seen in N containing compounds that fragment extensively m/z = 222 also arises from loss of C rearrangement involvingloss of this carbon Figure 11a
Mass spectrum of Tenormin using electron impact.
Unilever
Figure 11b Mass spectrum of Tenormin using chemical ionisation
Unilever
(Protonated 3-nitrophenyl methanol – H O)+ Figure 11c
Mass spectrum of Tenormin using fast atom bombardment Fast atom bombardment (FAB)
If the sample is susceptible to decomposition when heated, FAB is a useful technique
because it involves no heating at all. A beam of atoms with high kinetic energy,
usually xenon atoms with several keV energy, is used to strike a solution of the
sample in a ‘matrix’ compound such as glycerol (propan-1,2,3-triol) or 3-nitrophenyl
methanol (m-nitrobenzyl alcohol), on the end of a probe (Fig. 12).
Unilever
Figure 12
The fast atoms are generated by accelerating xenon ions to 6–9 keV, then neutralising them as they pass xenon atoms at low pressure. Neutralisation takesplace by electron transfer: When the sample is struck by the fast atoms, it is desorbed from the surface of the probe by the transfer of momentum, usually as an ion. In common with chemicalionisation the sample molecule in FAB is usually detected as (M+H)+ or as (M–H)–.
The ions produced are analysed in the mass spectrometer in the same way as ionsproduced by other methods. The FAB mass spectrum of Tenormin is shown inFig. 11c.
Salts of the type AX can be examined by FAB, and the A+ ions are detected (usually as AXA+) in the positive ion mode, and the X– ions in the negative ion mode.
Interpretation of the mass spectrum of a compound
With the aid of a computer it is possible to obtain information from the spectrumquickly – eg the mass of the molecular ion peak can be measured with sufficientprecision to be able to assign the molecular formula directly. Libraries of massspectra are commercially available on computer software, and the spectra of all thelibrary compounds with the same molecular formula can be called up forcomparison. The first spectrum in Fig. 8 is of an unknown compound with amolecular ion peak corresponding to the formula C H O , and with it are the library spectra of five isomers of this formula. Although the computer can be used tocompare one spectrum with another it is not necessary in this case. The peak at m/z= 105 (corresponding to the C H CO+ ion) is sufficient to be able to eliminate all the isomers except benzoic acid. However, the spectrum of the unknown does notalways correspond exactly to the library spectrum. This is entirely normal becausethe operating conditions under which the sample and the library compound wererecorded might not have been exactly the same. Make, and to a greater extent, typeof instrument, can influence the final spectrum. Unless there is a significantdifference in the two spectra the major peaks should be sufficient to identify thecompound.
If the spectrum is of a new or commercially sensitive compound then it will not Unilever
appear in a commercial library, although the elemental composition can bedetermined. In these cases it is usual to confirm a suspected structure rather than tryto determine the structure from first principles.
In confirming a suspected structure, or trying to elucidate a completely unknown structure, a rather different approach is used. The ratio of the two isotopes carbon-12to carbon-13 (12C:13C) is 100:1.1, which is significantly greater than the isotopicratios for hydrogen, nitrogen and oxygen. Therefore the isotopic peak at (M+1)+ willbe 1.1 per cent of the height of the molecular ion peak for every carbon atom presentin the molecule. In the spectrum of benzoic acid (Fig. 14) the (M+1)+ peak at m/z =123 is approximately 8 per cent (7.7 per cent) of the height of the molecular ionpeak. This is consistent with there being seven carbon atoms in the sample.
The presence of two other elements, chlorine and bromine, is usually easy to determine. The two isotopes of chlorine have mass numbers 35 and 37, and occurnaturally in the ratio 76:24. Thus the appearance of two molecular ion peaks differingby 2 mass units, and in the approximate ratio 3:1 suggests the presence of a chlorineatom in the compound. Bromine has isotopes of relative masses 79 and 81 inapproximately 1:1 ratio (50.5:49.5), so two molecular ion peaks of similar heightdiffering by 2 mass units indicates the presence of a bromine atom. Compounds withmore than one chlorine and/or bromine atom give the molecular ion and ionfragment peaks with the patterns shown in Fig. 13. All the peaks differ by 2 massunits.
Figure 13
Peak patterns due to the presence of chlorine and bromine atoms in molecular ions and ion fragments Unilever
In the absence of chlorine and bromine atoms, if the relative mass of the molecular ion is odd, then probably an odd number of nitrogen atoms is present; andif the relative mass is even then either nitrogen is not present, or there is an evennumber of nitrogen atoms.
Once some idea of the molecular composition has been calculated an attempt can be made to determine the structure of the compound. This is not always easybecause without some prior knowledge of the structure it is difficult to work outexactly what has been lost from the molecular ion. However, there are some groupsthat are commonly lost from the molecular ion and some of these are shown inTable 2.
For compounds containing a benzene ring, monosubstituted compounds might be expected to give a peak at m/z = 77, corresponding to C H +. While such a peak is often observed, bond fission occurs more frequently one bond away from thebenzene ring.
Figures 14 and 15 show the mass spectra of benzoic acid and methyl benzoate with the assignment of the major peaks.
Figure 14
Unilever
Groups commonly
associated
with the mass lost
Possible inference
Unilever
Figure 15
Applications of mass spectrometry
So far discussion of the application of mass spectrometry has been restricted to
determining relative atomic/isotopic mass and identification of organic compounds.
However, there are other applications.The mass spectrometer is sufficiently sensitive
to measure the isotopic composition of substances accurately. Although the isotopic
composition of elements is usually regarded as constant, there are some variations.
Atmospheric oxygen, for instance, is richer in 18O than fresh surface water.
Hydrogen, carbon, oxygen and sulphur are just four elements whose isotopic
compositions in different locations have been found to vary, and these variations
have been studied using the mass spectrometer.
Mass spectrometer applications are also found in medicine and biochemistry.
Many of the substances found in living systems are complex mixtures and aresensitive to fragmentation by using EI. Separation of individual compounds is done byusing gas chromatography (GC) or high performance liquid chromatography (HPLC)and the mixture from the column (the eluate) is fed into a mass spectrometer (seechapter 5). Once inside the spectrometer soft ionisation methods are frequently usedto ionise the sample. For example, FAB is often used to determine the residuesequence of peptides. Similarly, radioactively labelled compounds and compoundslabelled with stable isotopes such as 13C are used to determine the fragmentation Unilever
mechanisms of organic molecules and extended studies have led to the identificationof metabolic pathways. Labelled drugs allow pathways of drug metabolism to bestudied on a picomole scale.
Examples of GC-MS studies (gas chromatography interfaced to a mass drug abuse by athletes; oestrogens in the urine of pregnant females; determination of prostaglandins in human semen; the presence of ergotamine (an alkaloid) and other hallucinogens in blood; the separation and identification of the compounds responsible for givingfoodstuffs their aromas and tastes; the identification of bile acids in the gall bladder of an Egyptian mummy; the separation and detection of urinary acids thought to be connected withsudden infant death syndrome (SIDS) or cot death; and the highly specific and sensitive quantitative procedure for the determinationof residues or contaminants in foodstuffs thereby facilitiating increased safety.
Mass spectrometry is a powerful technique but it does have some disadvantages; for example: optical isomers cannot be distinguished from their spectra (although cis/trans isomers sometimes can); and although only small samples are required foranalysis the technique is destructive, and the sample cannot be recovered.
Applications also exist in environmental analysis. Although concentration methods are sometimes required to supply the spectrometer with sufficient sample toanalyse, GC-MS is sensitive enough to detect pollutants such as polychlorinatedbiphenyls (PCBs), which are found in surface water in concentrations on thenanogram per dm3 (ng l-1 or 10-9 g l-1) scale. It is important that they are detectedbecause they do not degrade in nature and build up in the food chain. Some seals,for example, have been found to have several per cent PCBs in their total fat content.
Monitoring for dioxin can also be done by using mass spectrometry. Dioxin can be found in the effluent stacks of incinerators, and is a most undesirable substance inthe environment because of its toxicity and stability.
Mass spectrometers have been adapted for use on board spaceships, where less attention has to be paid to the inclusion of vacuum pumps because the atmosphere isalready at low pressure. Indeed, in some cases the atmosphere has been sampledusing an open source. The exhaust gases of the space rocket must be avoided, ofcourse, as must the emissions from manned modules. The Viking space craft thatlanded on Mars was equipped with GC-MS, and both the Martian atmosphere andsoil were analysed (it was found that the partial pressures of water, nitrogen andcarbon dioxide are very low, and that the soil contained a lot of chlorine andsulphur).
Biomarkers in the petrochemical industry
Biomarkers are organic compounds whose carbon skeletons provide an unambiguous
link with a known natural product, and these are particularly useful in the
petrochemical industry.
When the marine organisms that formed crude oil died, their bodies were subjected to extreme heat and pressure, resulting in the formation of the oil. In thechemically reducing environment present many of the natural products – eg steroids Unilever
and terpenoids – were saturated. These saturated products are now present in crudeoil, and serve as biomarkers. It is thus possible to identify the organisms from whichthe oil was formed, enabling the terrain at the time of formation to be deduced. Oncethe biomarkers from a particular oilfield have been identified it is also possible todetermine where an oil sample originated. Thus, if an oil spillage occurs at sea andnobody admits responsibility, the biomarkers could be used to identify its source. Itmay then be a simple matter to find out which tankers were carrying oil from thatfield at the time of the incident, and so identify the culprit.
Identifying the biomarkers is done by separating the components of the crude oil using gas chromatography (see page 123) and feeding them directly into the chamberof a mass spectrometer. Once the components have been identified, by theirretention times on the chromatography column, and their fragmentation patterns, thebiomarkers can be detected on subsequent occasions by selective ion monitoring, inwhich only the characteristic peaks are looked for.
Geological dating
Naturally occurring potassium contains 0.01167 per cent of the radioactive isotope
40K. One of its daughter products is 40Ar. If a rock sample contains potassium, and
the argon formed by the decay of 40K is unable to diffuse out of the crystal, it is
possible to determine the age of the rock sample.
The argon content of a sample is determined by using a mass spectrometer to analyse the gas released when the sample is heated to melting in a vacuum. It isassumed that none of the argon produced in situ diffuses out of the crystal lattice ofthe rock, and that there is no loss or gain of potassium – ie it is assumed that thesample has remained a closed system since its formation.
A tracer of 38Ar is used to determine the 40Ar in a method known as isotope dilution, in which the unknown has a reference sample added against which it canbe measured. The total remaining potassium can be measured by using severalmethods, including flame photometry, atomic absorption spectrometry and isotopedilution.
It is possible that some atmospheric argon is trapped in geological specimens, and this will register on the mass spectrum once it has been released from the crystallattice. Atmospheric argon consists predominantly of three isotopes, 36Ar, 38Ar and40Ar. The latter isotope will clearly interfere with the results expected from the rocksample, but this can be compensated for from a knowledge of the naturalabundances of the isotopes – the ratio of 40Ar:36Ar is 295.5:1. Thus the peak at m/z =40, due to the argon in the rock, can be calculated.
By using the half life of 40K and the radiogenic 40Ar value, the amount of potassium that has decayed can be determined. From a knowledge of the ratio ofnaturally occurring isotopes it is then possible to determine how much 40K wasoriginally present. From this the proportion that has decayed can be calculated, andhence the age of the sample can be found. This is known as potassium argon dating(Fig. 16).
Recent trends
Improved spectrometer technology over the past decade has made detection ofsmaller and smaller concentrations of materials possible. Electron multipliers in thedetectors of spectrometers can now achieve a gain of 107 electrons per ion, andinstruments are available that can routinely measure elemental compositions down to10-12g g-1 or 10-12g cm-3 (pg g-1 or pg cm-3) of sample. By interfacing other techniqueswith mass spectrometry it is possible to analyse solid, liquid and gaseous samples –often automatically.
Mass spectrometers employing a magnetic field are now being complemented, and in some cases superseded, by quadrupole spectrometers. The ions produced Unilever
Figure 16
Schematic diagram of the potassium-argon dating method travel down the gap between four rods through which a radiofrequency and a dccurrent are passed. The ion path in such spectrometers is helical. These have theadvantage of being cheaper than magnetic sector instruments, but are not capable ofsuch a high resolution. In theory, magnetic sector instruments will differentiatebetween ions of m/z values of 60 000 and 60 001, although in practice are likely toresolve 600.00 from 600.01 or 60.000 from 60.001. However, quadrupolespectrometers can only produce unit resolution (ie 60 and 61).
The size and cost of quadrupole instruments has decreased significantly in recent years, and benchtop models are now available. The lower potentials necessary inthese instruments mean that they are much less prone to arcing inside thespectrometer. They are frequently interfaced to other techniques, such as gaschromatographs, where the presence of ions at particular nominal mass values issought.
In the future this technique will probably be more accessible to non-specialists who need specific data quickly, perhaps without the detailed interpretation a skilledspectrometrist could give.
Unilever
Exercises
Below are the mass spectra of some organic compounds. Try to determine theirstructures. Interpretations are given.
Exercise 1
Figure 17
The peaks of similar abundance, at m/z =94 and 96, are likely to be isotope peaks. Their relative proportions and their mass difference of two would suggest Br.
This is supported by the peaks at m/z = 79 and 81. Subtracting 79 from 94 or 81 from96 leaves 15, which corresponds to CH . The peaks at m/z = 12-14 correspond to the loss of successive protons from this group, and the peaks at m/z = 91, 93 and 95correspond to the loss of protons from the molecular ion. The peak at m/z = 94 islarger than the peak at 96 because it comprises CH81Br+ and CH 79Br+, as well as the 79Br isotope being very slightly more abundant. Therefore, we can conclude that thecompound is bromomethane, CH Br.
The peak at m/z = 97 is due to the 13CH 81Br+ ion.
Unilever
Exercise 2
Figure 18
No isotopic cluster due to a halogen is present, and the peak at m/z = 73, ie (M+1)+, suggests that 2.7 per cent ÷ 1.1 per cent = 2.5 carbon atoms are present. Ifthree carbon atoms are present, 36 of the 72 mass units remain and either twooxygens, two nitrogens or one of each is likely to be present. One nitrogen is unlikelybecause the relative formula mass is even. Formulae which fit are C H N and C H O . The absence of a peak at M-16 eliminates a primary amine (NH group).
Cyclic and secondary amine structures are also unlikely because the peaks in thespectrum are not abundant enough to suggest either CH –NH masses or losses. This leaves us with two oxygen atoms to try to account for. The peak at m/z = 55 is M-17,which would correspond to the loss of an OH group. The peak at m/z = 27corresponds to C H , CH =CH, and its abundance suggests its relative stability. This leaves us with C H , C, OH, and O to piece together. The possibility of a carboxylic acid group is confirmed by the peak at m/z = 45 (COOH+). This suggests that thecompound is propenoic acid (acrylic acid), CH =CHCOOH.
Unilever
Exercise 3
Figure 19
The abundance of the peaks at m/z = 96 and 98 lead us to consider whether the peak at 100 is truly the molecular ion peak, rather than an isotopic peak. A glance atm/z = 35 and m/z = 37 shows the presence of peaks in the ratio 3:1, stronglysuggesting chlorine. The abundance of the peaks at m/z = 96, 98 and 100 in theapproximate ratio 9:6:1 suggests that two chlorine atoms are present. The peaks atm/z =61 and m/z = 63 (formed by the loss of one chlorine atom) have heights in theratio of 3:1 – typical of the pattern expected for a single chlorine atom remaining inthe molecule. This accounts for all but 26 mass units. The group C H fits this, so the compound must be dichloroethene, C H Cl , but without further information it is impossible to decide which isomer it is (actually it is trans-1,2-dichloroethene).
Unilever
Exercise 4
The two compounds are known to be 1-aminobutane, CH CH CH CH NH , (A) and N,N-dimethyl aminoethane, CH CH N(CH ) ,(B). Try to decide which spectrum corresponds to which isomer, and explain why the spectra appear as they are.
Figure 20
Figure 21
Although it might appear difficult to assign the spectra, it is not. What must be considered is the relative stabilities of the ions that might be formed. Basically, apositive charge on a highly substituted ion will be more stable than an ion with a Unilever
straight chain because the effect of the positive charge can be reduced by theinductive effect of the side chain groups. Figure 20 shows a major peak at (M-15) – ieat m/z = 58. This corresponds to loss of a CH group because loss of an NH group is highly unlikely in the case of (A) and impossible in the case of (B). The remainingions would be: [CH CH CH NH ]+ and [CH N(CH ) ]+ or [CH CH NCH ]+ Ion (1) would be less stable than ions (2) and (3), although it is not possible to say which of these two ions gives the peak on this information alone. A tentative guess atthis stage would be that Fig. 20 is the mass spectrum of the tertiary amine.
Figure 21 shows only one significant peak, at m/z = 30. The only sensible ions that can give this peak are [CH NH ]+, and [CH NH]+. In neither case can (B) give these ions because it has no hydrogen atoms bonded to the nitrogen atom. Thus, thepeak at m/z = 30 must be due to the [CH NH ]+ ion because there is no methyl group bonded to the nitrogen in (A) so Fig. 21 is the mass spectrum of (A).
An experienced mass spectrometrist would be able to interpret the spectra directly from the knowledge that amines are susceptible to fragmentation one bond awayfrom the C–N bond (the bond to the nitrogen) – ie (B) would fragment as follows: The peak at m/z = 58 in Fig. 20 would then correspond to ion (2) above.
Similarly (A) would fragment here:

Source: http://media.rsc.org/Modern%20chemical%20techniques/MCT1%20Mass%20spec.pdf