Heard D.E. (editor) Analytical Techniques for Atmospheric Measurement

Подождите немного. Документ загружается.

242 Analytical Techniques for Atmospheric Measurement

lenses. In a magnetic field-free linear separation system, an ion with mass m and total

charge q = ze, has a kinetic energy KE defined by Equation (5.13).

KE =

= zeV (5.13)

Since t = v/d, we can rearrange this equation to give



2Ve



(5.14)

Equation 5.14 shows that ions with lower masses will therefore reach the detector first.

By accurately measuring the time t and keeping V and d constant m/z can be derived.

An advantage of this technique is that although the time differences are extremely small,

time can be very accurately measured. The principle is depicted in Figure 5.6.

Time of flight mass filters have very high transmission efficiencies, which leads to very

high sensitivities. All the ions formed are directed to the detector and so, in principle,

all ions are analysed. This is in contrast to the quadrupole mass filter, described in

Section 3.2.1, which measures masses sequentially. An important drawback of time of

flight instruments is difficultly in achieving good resolution. This stems from the physical

limits of creating ions, of the same energy, instantaneously, in the same place. These are

the following: the length of time in the ion formation pulse is finite, creating a time

distribution; there is a minimum size of the ion-generating volume creating a spatial

distribution; and the ions produced do in reality have a kinetic energy distribution. All

these factors contribute to uncertainty in the measurement of the flight time and hence

to lower resolution. To counter this, TOF-MS practitioners lengthen the flight tube and

hence the flight time, thus increasing the resolution. A long flight tube would be typically

1–2 m in length. This is a considerable restriction when operating this instrument on small

jet aircraft where space is a precious commodity. One major advantage of current TOF

systems is that they may be operated at high measurement frequencies. Commercially

available systems may be operated at 500 Hz (in contrast to typical quadrupole data acqui-

sition frequency of 10 Hz), making them ideally suited to comprehensive chromatography

applications, see Chapter 8.

5.3.2.3 Magnetic sector

The mass separating effect of a magnetic field comes from the fact that ions of equal

energy but different m/z describe different radial trajectories when passing through a

Figure 5.6 A representation of the principle of time of ﬂight mass spectrometry (Jochum, 1988 with

permission from Expert Verlag GmbH).

Mass Spectrometric Methods for Atmospheric Trace Gases 243

Figure 5.7 A schematic of the Nier-Johnson type magnetic sector mass spectrometer design.

magnetic field. A schematic diagram of a magnetic sector mass filter is given in Figure 5.7.

In a homogeneous magnetic field of strength B, ions with a velocity v experience a

force F , in a direction perpendicular to the field lines and to the direction of motion

of the ions. In order to traverse the circular path, the magnetic or Lorentz force FL

and the centripetal force Fc must be equal. The radius of the resulting trajectory can be

calculated by equating the Lorentz force FL =evB with the centripetal force Fc =mv

/r.

Substituting the relationship eU =mv

/2 for the ion energy (where U is the acceleration

potential), we obtain:

r =

2mU

(5.15)

Rearranging we can obtain an expression for m/z.

(5.16)

From Equation (5.16) it can be seen that mass spectra can be obtained by varying one

of three variables (B V ,orr) while holding the other two constant. Classical magnets

were not well suited to the fast scanning necessary for coupling with GC due to magnetic

heating by Foucauld currents induced by rapidly changing the magnetic field. Most

modern mass spectrometers contain an electromagnet in which ions are sorted by holding

V and r constant while varying the current in the magnet and hence the field strength B.

In practice, most magnetic sector instruments are built to deflect ions through circular

paths of 60, 90 or 180



. When a collection of ions with the same m/z but with small

diverging directional distribution is acted upon by the magnetic field, a converging

directional distribution is produced as the ions leave the field (directional focussing).

For this reason the simplest magnetic sector mass filters, are called single focussing

spectrometers. It is important for mass spectrometry with magnetic sectors that the

incident ions have the same energy as far as possible. A large span of energies leads to a

low resolution. Therefore to obtain the high resolution required for isotope measurements

an electric and a magnetic field are combined so that the energy dispersing effect of the

magnet is compensated by the electric field. By this so-called double focussing method,

directional and energy focussing is achieved. Figure 5.7 shows a schematic diagram of

the so-called ‘Nier-Johnson geometry’ design.

244 Analytical Techniques for Atmospheric Measurement

A wide variety of double focussing mass spectrometers are available commercially.

Resolutions of up to 10

can be obtained and this is necessary when analysing isotopic

ratios in ambient air.

5.3.2.4 Ion trap

An ion trap is a device in which gaseous ions can be formed and confined for extended

periods by electric and/or magnetic fields. In a sense, the ion trap may be thought of

as a ‘three dimensional’ quadrupole (see Section 3.2.1.). The electric fields in a standard

quadrupole assembly constrain ions into stable two-dimensional motion between the

quadrupole rods. By constructing an arrangement of two end-cap electrodes, one on

either side of a third annular electrode and all of hyperbolic cross section, ions injected

into the volume between the three electrodes can be ‘trapped’ into a complex circular

motion by applying suitable DC and RF electric fields. Ions with an appropriate m/z

circulate in a stable orbit within the cavity. When the radio frequency is increased, the

orbits of the heavier ions become stabilised, while those of the lighter ions become

destabilised and eventually collide with the electrode. The ions repel one another in the

trap and as a result their trajectories expand as a function of time. To avoid ion losses

through this expansion, a small pressure of helium gas (typically 0.13 Pa) is maintained

in the trap. This removes excess energy from the ions by collision.

Like a quadrupole, the ion trap is usually a low resolution mass spectrometer, being

usually able to resolve unit mass (1 amu). A considerable advantage of the ion trap

mass filter is that it can be used to perform MS/MS experiments. This involves trapping

ions of a characteristic m/z as described above and then collisionally fragmenting them

further before detection. In this way two different species having the same m/z can be

differentiated provided they produce distinct fragments. Usually the precursor ions are

given more kinetic energy through the use of an extra RF voltage applied to the end caps

so that total collision energy can be increased and the selected ions induced to fragment.

This is of great help in atmospheric studies where many different species may lie on the

same mass (e.g. propanal, acetone, glyoxal all have mass 58 at unit mass resolution, see

Table 5.2) but when fragmented can be expected to be different.

5.3.3 Detectors

Following the successful application of the mass filters described above we obtain a

current of ions with a well-defined m/z. These ion currents lie typically between 10

−8

and 10

−16

A. To measure these currents a number of systems are available. The three

most commonly in use in atmospheric science are described below.

5.3.3.1 The Faraday cup

The Faraday cup collector is the simplest and most inexpensive of detectors available.

The cup is aligned so that ions exiting the mass filter strike the collector electrode. The

charge carried by the ions that hit the cup is earthed via a high ohmic resistance and the

potential difference across this resistance is measured with a high impedance amplifier.

Mass Spectrometric Methods for Atmospheric Trace Gases 245

The electrode is surrounded by a cage that prevents the escape of reflected ions and

ejected secondary electrons. The collector electrode is inclined with respect to the entering

ions so that particles striking or leaving the electrode are reflected away from the entrance

to the cup. The response of this detector is independent of the mass, the energy and the

chemical nature of the ion. The main disadvantage of the Faraday cup is the amplifier

required, since this limits the speed at which a spectrum can be obtained. Moreover,

as the cup detector uses no internal amplification (see Sections 5.3.3.2 and 5.3.3.3) it is

less sensitive. The Faraday cup is widely used in systems requiring high precision, as the

charge of the cup is not dependent on the mass and energy of the detected ion. It is

necessary to use such detectors for isotopic analyses, see Section 5.4.3.

5.3.3.2 The Secondary electron multiplier (SEM)

Where extremely high sensitivity is required to detect ions, a secondary electron amplifier

is used. The secondary electron multiplier comprises a conversion dynode, a series of

further dynodes and a collector. Every mass-filtered ion that hits the conversion dynode

releases a burst of electrons, which are accelerated to the next dynode. In turn, each

electron hitting the second dynode release a similar burst of secondary electrons. With

each successive dynode the electron current is amplified with the result that the current

measured at the final cup is several orders of magnitude greater than the incident ion

current. A 20-stage SEM can produce a current gain of ca 10

. A combined schematic

diagram and picture of an SEM is shown in Figure 5.8. The SEM pictured is one used

in the proton transfer mass spectrometer described in Section 5.4.2. To better judge the

size of this detector a 1 Euro coin is included in the photograph.

Figure 5.8 Schematic diagram and photograph of a secondary electron multiplier.

246 Analytical Techniques for Atmospheric Measurement

(a) (b) (c)

–2

–4

–150

Ions Ions

–

Conversion

dynode

Conversion

dynode

Channel electron

multiplier

(Channeltron)

Signal Signal

Figure 5.9 A schematic diagram of a channeletron when used for (a) negative and (b) positive ion

detection and (c) a cross section picture of a commercial version (Picture courtesy of Sjuts Optotechnik

GmbH).

In general, SEM detectors are robust and reliable, providing high current gains in

nanosecond response times. The detector response is, however, dependent on the mass,

the charge, the speed and the molecular structure. This necessitates empirical assessment

of the ion calibration factors for quantification work. For magnetic sector instruments,

the detector may be set directly behind the mass filter exit slit as ions reaching the detector

usually have enough energy to eject electrons from the first dynode. When used with

quadrupoles the ions must be accelerated prior to the first impact. The lifetime of these

detectors is 1–2 years because of surface contamination of the Cu/Be surface from the

incident ions. They are not as precise as Faraday cups but the higher sensitivity allows

for rapid scanning.

5.3.3.3 The Channeletron

The channeletron (or channel electron multiplier) is a variant of the secondary electron

multiplier described above, but it has a single continuous dynode rather than several in

series. It consists of a glass or ceramic tube with semiconducting inner surfaces. This

detector is used in the API-CIMS instrument described in Section 5.4.1. When used for

positive ions, see Figure 5.9a, a conversion dynode at high negative potential (3–4 kV)

is used to release electrons when the ion impacts. The resulting electrons are led into

the channeletron where they cause an electron avalanche, sufficient to be detected by a

preamplifier. In the case of negative ions, Figure 5.9b, the ions themselves are led into the

channeletron and detected. Current gains of 10

are typical for such detectors. A cross

section of a commercially available channeletron is shown in Figure 5.9c.

5.4 Mass spectrometry used in atmospheric chemistry

5.4.1 Atmospheric pressure ionisation-chemical ionisation

mass spectrometer (API-CIMS)

As the name suggests, in API-CIMS primary ions are produced in an ion source operating

under approximately atmospheric pressure. This instrument has also been termed as

Mass Spectrometric Methods for Atmospheric Trace Gases 247

ion molecule reaction mass spectrometer (IMRMS). In such an instrument, positive or

negative reagent ions interact with the trace gas molecules of interest and produce new

ions by chemical ionisation, as described in Section 5.3.2. For these reactions to be fast

and effective, the ionisation has to occur at high pressures, and for atmospheric scientists

it is practical to choose approximately atmospheric pressure.

Figure 5.10 shows a schematic diagram of a mass spectrometer that has been used to

measure SO

(when in negative ion mode) and acetone (when in positive ion mode) on

various field campaigns on the ground and from on-board aircraft. Ions are produced by

corona discharge between a needle and the aperture lens housed within the ion source,

which is kept at constant temperature 80



C and pressure (280 hPa). The polarity of the

needle is chosen to have the same polarity as the ions to be detected since this improves

the transmission of ions through the system. Further details of such chemical ionisation

systems can be found in Huang et al. (1990). In this case the pressure in the ion source is

chosen to be slightly lower than the lowest pressure encountered in flight (at maximum

altitude of the aircraft, in this case 8.5 km–350 hPa).

The ion source is separated from the decluster chamber by a small 0.25 mm orifice.

As the decluster chamber is at lower pressure 0.5–1.0 hPa, a small flow (200 ml) of air is

drawn through the source region into the decluster chamber. When ionisation of ambient

air is done at these pressures, significant clustering of water molecules can occur on the

ions of interest. Since the number of water clusters can vary greatly, the ion signal can

be spread over many masses (e.g. X

H

O

, complicating the mass spectra in similar

fashion to fragmentation. In the decluster chamber the mean free path of the gas particles

is typically 100m, which is enough for them to be accelerated in the electric field present

in the decluster chamber. The energy thus imparted is sufficient to induce declustering

Figure 5.10 A schematic diagram of an atmospheric pressure chemical ion mass spectrometry. (Jost,

2002)

248 Analytical Techniques for Atmospheric Measurement

of the ions upon collision with another gas molecule. The desired ion-molecule reactions

can occur from when the ions are first produced until the ions are drawn into the high

vacuum of the spectrometer, where the pressure is too low for ion-molecule reactions to

be effective. A number of ion lenses are used to improve the transmission of the ions

from the decluster chamber to the channel electron detector. The ion lenses always have

the same polarity as the ions to be detected, and the lens potential must be optimised.

If it is too low the transmission will be poor, while too high causes the ions to enter the

quadrupole too fast and poorly resolved mass spectra result.

Let us now examine the underlying principle of operation. Having produced a positive

or negative reagent ion A

, we wish to react it with a neutral molecule of interest, X.

This can occur in a time dt, between ion source and high vacuum.

The rate of reaction can therefore be described,

dA



=−kA

X (5.17)

where k is the rate coefficient and the terms in square brackets are concentrations.

Assuming that only a small fraction of the trace gas X reacts then it follows that [X] is

approximately constant. Therefore we may integrate to derive the expression,

X =

A



t=0

A



t=T

(5.18)

By taking into account that the number of ions does not change in-ion molecule reactions

we may further transform Equation (5.18) to include B

, the product ions, as shown

below

X =

A



t=T

+B



t=T

A



t=T

1 +B



t=T



A



t=T

(5.19)

If the reaction time T is sufficiently short, the concentration of reactant ions A

remains

much higher than the concentration of the product ions B

, and the approximation

ln1 +x ≈x when x<<1 can be applied, simplifying this equation to

X =

B



t=T

A



t=T

(5.20)

The concentration ratio of the reagent ion A

 to the product ion B

 can be determined

from the relative count rates of the mass spectrometer. Therefore the concentration of

gas X may be determined if we know the reaction time t, the rate coefficient k and

the count rate ratio of B

and A

. The same principle is used in the proton transfer

mass spectrometer described in Section 5.4.2. The reaction time T can be calculated or

measured although the measurement is complicated. The rate of an ion–molecule reaction

is very fast and the rate coefficient k typically has a value of 10

−9

molecule

−1

In the case of the atmospheric pressure ionisation mass spectrometer described here,

multiple reagent ions can be formed (e.g. in positive mode H

H

O

. The distribution

of these ions can change markedly with humidity. Each may react with X with different

Mass Spectrometric Methods for Atmospheric Trace Gases 249

rates, which are known only with ca 30% accuracy. Additionally, back reactions and

continuing reactions of the product ions can also be sources of error. Clearly, careful

calibration is essential to determine the proportionality coefficient between the X, the

molecule of interest and the product/reagent ion ratio under a variety of temperature and

humidity conditions. The measurement principle of API-CIMS is in many ways similar to

the PTR-MS technique. The accuracy, precision and errors of atmospheric measurements

made with these techniques are discussed in Section 5.4.2.

The API-CIMS instrument described here has been successfully deployed on several

ground and airborne campaigns. Over the west coast of Africa it has been used to track the

emission and subsequent photochemical production of acetone (using H

as a reagent

ion) and sulphur dioxide (using the CO

−

ion) from the large-scale biomass burning in

the dry season (Jost et al., 2003). Furthermore, the instrument was employed in the

laboratory to investigate the direct emission of compounds (e.g. acetonitrile, CH

CN)

when selected African biomass materials were burned under controlled conditions, see

Figure 5.11.

If the global distribution of biomass is known, then laboratory emission rates deter-

mined in this way may be used to derive global emission strengths. Furthermore the

emissions rates may be included in global models in order to estimate the atmospheric

effect of these global emissions on species such as ozone.

By adopting different reagent ions, other gases of atmospheric interest may be selectively

detected. The SF

−

reagent ion has been used in laboratory or field studies to detect

molecules of atmospheric interest such as HNO

,NO

, HOCl, CF

O and SO

(Huey

et al., 1995) and in 2004 hydrochloric acid (HCl) was measured from aircraft in the

upper troposphere/lower stratosphere with a 5 pptv detection limit and an accuracy of

25% (Marcy et al., 2004).

5.4.2 The proton transfer reaction mass spectrometer

The proton transfer reaction mass spectrometer (PTR-MS) is a specific example of

the general chemical ionisation scheme described above, being developed by Lindinger

and co-workers at the University of Innsbruck (Lindinger et al., 1998). The underlying

principle is the same as described in Section 5.4.1 but in this case, the reagent ion is H

Complications resulting from water clusters present in API-CIMS are reduced by judicious

Mass 42 (CH

CN)

14:25 14:30 14:35 14:40 14:45

Time

0.0

0.5

1.0

1.5

2.0

CO and CO

Figure 5.11 Example of a laboratory application of API-CIMS – Biomass burning analysis.

250 Analytical Techniques for Atmospheric Measurement

Table 5.3 Proton afﬁnities

Compound Formula Proton afﬁnity kJ mol

−1



Nitrogen N

464.6

Oxygen, Ozone O

396.3, 595.9

Noble gases Ar, Ne, He 346.3, 174.4, 148.5

Carbon mon/di oxide CO, CO

562.8, 515.8

Methane CH

544

Water H

O 660.0

Methanol CH

OH 725.5

Acetonitrile CH

CN 779.2

Acetone CH

COCH

782.1

Dimethylsulphide CH

SCH

801.2

Isoprene CH

CCH

CHCH

797.6

Source: Data from Hunter and Lias (1998).

choice of pressure and energy parameters, as detailed below. The key step is the transfer

of a proton from the reagent ion H

 to the molecule of interest (X). Through proton

transfer reactions it is possible to measure all compounds with a proton affinity larger

than that of H

O. Fortunately for atmospheric scientists, it is ‘blind’ to the major air

constituents N

, Ar, and CO

, as can be seen from Table 5.3. However, it can detect

many important atmospheric species such as acetone and isoprene. The ionisation occurs

at relatively low energies so that protonation usually does not cause the molecule to

fragment, or if it does, by ejecting only one H

O molecule. This considerably simplifies

the mass spectra. The proton affinity or gas-phase basicity is defined as the negative

molar Gibbs energy, −G, of the hypothetical protonation reaction,

+X → XH

(5.21)

A selection of these proton affinities is given here in Table 5.3. A comprehensive list can

be found at the NIST website – http://webbook.nist.gov/chemistry/. All the species in

Table 5.3 with proton affinities greater than water have been measured by PTR-MS.

A schematic diagram of the PTR-MS apparatus is shown in Figure 5.12. The ion source

consists of a hollow cathode and an earthed anode, through which water vapour flows at

about 8 cm

min

−1

. When held at a pressure of about 2 mbar (200 Pa), a cathode voltage

of 600 V is sufficient to cause an electric discharge between cathode and anode. The

result is an intense source of H

ions: count rates of typically 1×10

counts s

−1

being

detected by the mass spectrometer. An alternative source of H

ions has also been

developed using alpha particles emitted from a strip of Am

241

to ionise water vapour

(Hanson et al., 2003). Connected to the ion source is the reaction chamber or ‘drift tube’,

where the proton transfer reactions occur between H

and the compounds of interest

in ambient air. The reactor consists of a series of stainless steel rings separated by thin

isolating teflon rings. The steel rings are connected by resistors so that a voltage of up to

600 V can be applied over the entire set of rings to obtain a homogenous electric field.

Ambient air is pumped through the drift tube which is maintained at 2 mbar (200 Pa)

pressure. The residence time of air in the drift tube is about 1 second. During this time

Mass Spectrometric Methods for Atmospheric Trace Gases 251

Pump

Hollow

cathode

O vapor

inlet

Ambient

air inlet

Drift tube

600 V, 2 mbar

Secondary electron

multiplier

Quadrupole

mass

filter

Figure 5.12 A schematic diagram of the PTR-MS.

ions are accelerated through the drift tube containing ambient air and ionising

reactions may occur. As we have seen in Section 5.4.1, H

ions can form clusters with

water of the type H

H

O

. The distribution of the clusters is highly dependent on

the sample humidity. More humid conditions promote larger clusters. Their presence

complicates the mass spectra and for some species it can lead to a strong humidity

dependence in the calibration (see Section 5.4.1). This is because some trace gas species

react with H

but not with H

H

O

ions, an effect that has been shown to be

important for non-polar compounds such as benzene (Warneke et al., 2001). The drift

tube electric field is applied to increase the average kinetic energy of the ions to prevent

clustering so that in PTR-MS, H

is the predominant reagent ion. The disadvantage

of the electric field is that it effectively limits the reaction time and pressure in the

reaction chamber, making the PTR-MS technique less sensitive than API-CIMS. From

the drift tube the ions are directed to a quadrupole mass filter held at 10

−5

bar by a turbo

molecular pump.

Quantification is based on the difference in counts between ambient air and air that has

been scrubbed of organics by passing it through a heated trap containing a platinised wool

catalyst. The interpolated background signal before and after an ambient measurement

is then subtracted from the ambient signal to derive a mass signal corrected for any

background contamination originating from the mass spectrometer. As the effectiveness

of the mass filter and detector is mass dependent the ion counts must be corrected for this

transmission difference. The transmission as a function of mass can readily be determined

by sequentially injecting high concentrations of species covering the entire mass range.

At sufficiently high concentrations all the reagent ions (e.g. H

) are converted to

product ions (e.g. CH

COCH

. The ratio of the initial reagent ion counts per second

(typically 1 ×10

 to the acetone counts indicates the relative transmission of the two