вход по аккаунту


Chemical IonizationЧA Mass-Spectrometric Analytical Procedure of Rapidly Increasing Importance.

код для вставкиСкачать
Chemical Ionization-A Mass-Spectrometric Analytical Procedure of
Rapidly Increasing Importance
By Wilhelm J. Richter and Helmut Schwarz[*l
The mode of ionization of a molecule has a strong influence on its behavior in the mass
spectrometer and thus on the information that can be obtained from its mass spectrum.
In chemical ionization a reagent gas, e.g. methane, is first ionized by electron impact. The
ions formed in ion-molecule reactions, in particular [CH,]’, [CZH5]+, and [C3H5]+, then
react “chemically” with the substrate M in fast acid/base type reactions to form ions of the
type [MH] +,[M(C2H5)]+,etc., which subsequently fragment to various extents. Alternatively,
chemical ionization can be effected by charge exchange, in that ions of a reagent gas, e.g.
[He]“, react with the substrate M to form molecular ions [MI+’. Chemical ionization can
thus be conducted in a more or less mild fashion and the extent of the fragmentation can be
controlled over a very wide range.
1. Introduction
Given the use of a suitable instrument, the versatility of
mass spectrometryY*]permits the investigation of a wide variety of problems ranging from investigations of primarily scientific character to investigations of practical and highly topical
significance.Studies of the kinetic, thermodynamic, and chemical behavior of ions in the gas phase, including studies of
extremely fast to relatively slow reactions on the picosecond
to millisecond time scale, exploration of surface properties
of solids, structure elucidation of unknown substances, as
well as detection and quantitative determination of known
substances are but a few examples to which mass spectrometry
can be successfully applied“]. Mass spectrometry qualifies
especially for analytical applications because of its very high
sensitivity (present detection limits about 10- l 4 g) and because
of the wealth of information afforded, quite unattainable by
any other routine method of instrumental analysis for minute
sample size. As judged by the number of published papers
and instruments in use, analytic chemistry represents the principal field of application of the method. Mass spectrometry
is almost indispensable for trace analysis, especially in the
environmental, biological, and medical sciences. In depth, its
potential and scope are demonstrated by geochemical, paleontological, archeological, and cosmochemical studies-e. g. in
the Apollo and Viking projects.
In analytical applications the reliable determination of molecular weights (when using high-resolution instruments also
the determination of molecular formulas), along with correlation of fragmentation with structural features are often the
main objectives.In conventional ionization by electron impact,
Dr. W. J. Richter
Zentrale Funktion Forschung, CIBA-GEIGY AG
CH-4002 Basel (Switzerland)
Prof. Dr. H. Schwarz
lnstitut fur Organische Chemie der Technischen Universitat
Strasse des 17. Juni 135, D-1000Berlin 12
[**I Abbreviations: amu, atomic mass units; CD, charge exchange; CI,
chemical ionization; EI, electron-impact ionization; FD, field desorption;
FI, fieid ionization; GC, gas chromatography; ICR, ion cyclotron resonance;
MS, mass spectrometry; PA, proton affinity; RE, recombination energy;
SIM, selected ion monitoring; TIC, total ion current.
limitations or difficulties may arise because the sample has to
be ionized in the gas phase, yielding molecular ions [MI+ often
insufficiently stable to permit detection. Further limitations
may arise because insufficiently volatile or thermally labile
compounds can pyrolyze or isomerize before ionization takes
place and because isomerization can also occur after ionization.
In recent years it has become possible to overcome these
disadvantages to a considerable degree by an extension of
the methodology. So far as the correlation of fragments with
structural features is concerned, collisional activation (CA)
mass spectrometry‘’’ and other techniques[31now make it
possible to verify structures of ions. Thus conclusions can
be drawn from fragments on a much better-founded basis
regarding the structure of molecular ions and, hence, also
about the structure of intact molecules. In the determination
of molecular weights of intact substrates there have even
been several breakthroughs, represented by three new techniques: 1)Field desorption mass spectrometry (FD), developed
by Beckey and his associates[481,especially in its variant of
c a t i ~ n i z a t i o n [ ~ ~2)- ~the
~ ]still
; very recent development of
ionization by radioisotopes (californium-252source)[51 ;and 3)
the method of chemical ionization (CI)Pldeveloped by Munson
and Field6],which constitutes the subject of the present article.
The first two of these procedures allow very mild ionization
of the substrate, i. e. ionization with little excess energy transfer;
owing to the suppression of fragmentation, well-defined molecular ions are produced, or their adducts with protons or
alkali-metal ions. The two processes truly extend the area
of application of mass spectrometry in that, unlike electronimpact ionization (EI), they permit the study of intact compounds that do not volatilize without decomposition. Chemical ionization (CI) shares only the advantage of mild ionization; yet, in contrast to the FD techniques, the amount of
fragmentation can be controlled within a wide range, whereby
the proportions of information about molecular weight on
the one hand and structure on the other can be adapted,
The term “chemical ionization”(C1) is not synonymous with chemi-ionization. Chemi-ionization achieves ionization of a substrate by interaction with
electronically excited neutral particles [8].
Angew. Chem. Int. Ed. Engl. 17, 4 2 4 4 3 9 ( I 978)
almost at will, to requirements by varying the experimental
conditions. Chemical ionization is therefore often an alternative to electron-impact ionization, whereas the two other modes
ofionization represent more or less complementary techniques.
In accordance with its special merits and promise“] for
modern mass-spectrometric analysis, chemical ionization has
been discussed in several earlier reviews[lb,71 sometimes placing emphasis on special aspects of the topic; in the present
article we can thus concentrate on some of the more recent
viewpoints and results after first describing the principles of
the phenomenon.
2. Physicochemical Aspects
2.1. Production of the Ionizing Plasma and Ionization of the
When electron-impact ionization of a compound, e. g. methbar in contrast to the
ane, is carried out at ca. 1 x
usual high-vacuum conditions (ca. lo-’ bar), the molecular
ions and fragments produced [eq. (a)] can react further in
rapid ion/molecule reactions1’] with neutral methane molecules as a result of the diminished mean free path. Concentrations of “secondary” and “tertiary” ions are thus built up
[eq. (b)-(h)], in dependence on the pressure, temperature,
and instrumental parameters that determine the residence
time of the charged particles in the collision chamber (ion
source). At this operating pressure of about 1 x 1 O - j bar 95 %
of the total ion current is carried by the following species:
[CH,]’ (48 %), [C2H5]+ (41 %), and [C3Hs]+ (6 %).
usual CI working pressure) thus acts as a “reagent gas”, the
compound M as a CI substrate.
M + [MH]+ + CH4
[C2H5]+ + M + [MH]+ + C2H4
[CZH~]’+ M + [M(CzHs)]’
Besides this acidbase type of substrate ionization, which is
often used in practice and which invariably involves “closed
shell” ions, in chemical ionization in the widest sense“],
another type of ion/molecule reactions is of importance, in
which the substrate is ionized by charge exchange (CE).
Such CE processes correspond to the redox-reaction type
and, like electron-impact ionization, lead to “true” molecular
ions, i. e. to species with “open shell” character [eq. (I)]; they
occur particularly with reagent gases that contain no H atoms
(noble gases, N2, NO, 02,CO).
+ M-t [MI” + He
Like electron-impact ionization, acidjbase and also redoxtype ionization find a counterpart in the formation of negative
ions. Whereas “negative CE” (NCE) occurs by resonance capture of thermal electrons, such as are present in a nitrogen
plasma (electron attachment), in “negative CI” of the acidbase
type (NCI), proton transfer occurs in analogy to the positive
mode (but in the opposite direction) together with formation
of addition complexes. Reagent anions such as [CH30](e.g. from CH30NO as the reagent gas [(eq. (rn)][’la1)or
[Cl] - (from the substrate itself by dissociative electron addias Br$nsted or
tion [eq. (n) and ( o ) ] ~ ’ ’ ~) ,now
~ ’ ~function
Lewis bases, respectively.
[CH,O]- + R-H[R]e- + R-CI -t [Cl] + R’
[Cll- + R-CI-t [RC12]-
2.2. Energetics of Ionization and Fragment Formation
If such a “methane plasma” contains as an additive a second
compound M in a concentration smaller by orders of magniit will be converted into ions not
tude (generally 1: <
directly by primary electron impact (a physical process), but
by ion/molecule reactions with the above species ( i e . by
chemical processes). Not only can the [ C H s ] + ~ * ]and the
[C2HS]+ ions function as Bransted acids with respect to
M in proton-transfer reactions [eqs. (i) and (j)], but also
the [C2Hs]+ ion can function as a Lewis acid [eq. (k)] in
the formation of collision-stabilized complexes. In the formation of the ions [MH]’ and [M(C2Hs)]+ (“quasimolecular
ions” [QM]+)p**], methane at a pressure of ca. 1 mbar (the
[*I The growing importance of chemical ionization can be recognized.
by the fact that as few as about 30 papers appeared on this topic in the
period from 1966 to 1971 [7i] but more than a thousand up to 1976.
We are grateful to Dr. H.-M. Schiebel, Technical University of Braunschweig,
for his assistance with the literature search.
p’] For the structure of the ion [CH,]’ and some homologous carbonium
ions see [lo].
[***I Although this description is, strictly speaking, incorrect [l b, 741, it is
retained here for convenience.
Angew. Chem. I n t . Ed. Engl. 17,424-439 ( 1 9 7 8 )
2.2.1. Thermodynamic Factors
The contributions of processes(i)-(k) to the total ionization
of a substrate are in the first place determined by the structure
of the substrate and the nature of the reagent gas selected.
In general, only exothermic ion/molecule reactions proceed
at high ratesfg1.In the acidbase case this fundamental condition for chemical ionization is always fulfilled for proton
transfer [eqs. (i) and (j)] if the proton affinity (PA)[“] of M
exceeds that of the conjugate base (e.g. CH4 or C2H4)
of the relevant reagent ion ([CHs]’ or [C2Hs]+). The PA
values of some common reagent gases are listed in Table 1.
[*I Some authors restrict the term “chemical ionization” to this acid/hase
type of reaction [eq. (i)-(k)] and regard charge-exchange processes [eq.
(I)] as a separate type of ionization. Common to both types of ionization
is the occurrence of ionJmolecule reactions.
[**I Proton affinity (PA) is defined as the standard enthalpy AH” of the
reaction [MH]+-. M [HI’. For a comprehensive collection of PA values
and critical comments on the methods of measurement see 17 h] and [i2b12d].
Table 1 . Protonating reagent gases, principal reagent ions, and proton affinity
(PA) of corresponding conjugate bases (chemical ionization (CI)) [7, 12 b,
Reagent gas
Brmnsted acid
PA (kcal/mol)
193 (bC4Hg)
A first consequence of this energy requirement is that some
degree of selectivity as to effective ionization of the substrate
can be achieved by the choice of the reagent gas. Thus, chemical
ionization with NH3 as the reagent gas [CI(NH3)], unlike
universally protonating chemical ionization with CH4
[CI(CH4)], causes no protonation to [MH]' ions neither in
saturated nor non-conjugated unsaturated ketones (PA values
<200 kcal.mol-'[7h~121);
yet it causes formation of their precursors, i. e. of collision-stabilized complexes [M(NH4)] +,in
which (endothermic) H+-transfer is not completed. However,
owing to the increase in proton affinity as a result of conjugation u,P-unsaturated ketones afford high yields of [MH]'
ions under these conditions['3]. A second, no less important
consequence of the energetics of the ionization process is
that the exothermicity (difference between the PA values of
the substrate and of the conjugate base of the reagent ion)
controls to a considerable degree the stability of the [MH]'
ions and thus the extent of decomposition into ions of smaller
mass with loss of neutral particles (CI fragmentation). In
this way, energy-rich reagent ions such as [H3]' or [CH,]'
afford e.g. highly excited [ROH2]+ ions by protonation of
alcohols ROH, which, in general, have a much higher protonaffinity than the corresponding reagent-gases (H2 or CH4);
consequently, these excited ions often decompose completely
with loss of water molecules into [R] fragments. If, however,
[H,O]+ ions are used for the ionization, ( H 2 0 as reagent
gas) relatively stable and thus detectable [ROHz]+ ions are
obtained owing to the smaller PA difference between HzO
and ROH. As mentioned above, this possibility of controlling
the fragmentation by appropriate choice of the reagent gas,
together with adjustable selectivity, endows the method with
a degree of freedom that is of great value in many analytical
A similar situation is encountered in the ionization of substrates by charge exchange (CE). Here too, the reactivity of
the initially formed [MI" ions can be controlled by a suitable
choice of the ionizing agent. In this case the energy content
of the molecular ions [MI +'
by the difference
between the recombination energy (RE) of the reagent ions
(in eq. (1) of [He]+') and the ionization potential (IP) of M,
the latter being in the range 7-12eV for most organic compound~['~
] values for some typical CE reagent gases are
listed in Table 2). Correspondingly, when helium is used
complete fragmentation often results, while heavier noble
gases tend to give molecular ions that fragment to a much
smaller extent.
How such energy effects can be used to advantage, and
how these two types of chemical ionization can be effectively
combined is illustrated in Figure 1. It shows the combined
CE/CI spectrum of 4,5a-dihydrocorticosterone (I) obtained
with a two-component reagent gas system['51 (Ar/H20; cf.
Section 4.2.1)[""]. On the one hand, the mild protonation
by the weak Br$nsted acid [H30]+ (m/e= 19)leads to [MH]+
ions of remarkable stability for an ROH-type compound (cn.
50% relative intensity in spite of the repeated subsequent
loss of HzO) as reliable indicator of molecular weight. On
, ,
Table 2. Redox type reagent, recombination energy (RE) (ionization by charge
exchange (CE)) [12a].
Reagent gas
RE [eV]
[zl ; ,
295 299
180 2 4
Fig. 1. CE/CI(Ar/HzO) spectrum of 4,5a-dihydrocorticosterone ( I j
the other hand, [Ar]" ions (m/e=40), act mainly as an aggressive agent of the redox type, simultaneously producing highly
unstable [MI" molecular ions (barely 2 % are recorded as
such).These latter yield valuable information about the structure because of their marked fragmentation, whose course
resembles that observed on electron impact.
2.2.2. Kinetic Factors
For a given substrate, not only the exothermicity of the
ionization process but also the reaction temperature (here
in chemical ionization the temperature of the reagent gas)
is decisive for the extent of fragmentation, since this temperature provides contributions to the internal energy ofthe ionized
particle. In electron-impact ionization the internal energy of
the ions formed in the high vacuum cannot be represented
by a Boltzmann distribution, which would be the prerequisite
Angew. Chem. Int. Ed. Engl. 1 7 , 4 2 4 4 3 9 (1978)
for an exact mathematical description and formal kinetic treatment of the temperature-dependence of the spectrap]. In sharp
contrast to this, in chemical ionization the large number of
collisions of ions with excess reagent gas (the number of
collisions lies between 1 x 10' and 2 x 10317jl,depending on
pressure, temperature, and residence times of ions in the source)
usually permits equilibrium to be sufficiently established in
order to assume a Boltzmann distribution of the internal
energies of the ions. According to Field7J3'I, the kinetic
parameters, i. e. activation energy and pre-exponential factor
(frequency factor) v, can consequently be obtained by studies
of the temperature-dependence of the fragmentation.
2.3. Special Investigations
The possibility of determining the kinetic parameters of
a fragmentation process experimentally is, of course, very
valuable for the investigation of fundamental aspects of reaction mechanisms. Thus, it has been shown[' 81 that the elimination of RCOOH from the [MH]' ions of methoxymethyl
esters (2) (Scheme 1 ) occurs by the same mechanism as their
acid-catalyzed solvolysis[''I. From the relatively low values
of the frequency factors [v=4.5 x lo9 and 6.5 x 108s-' for
( 2 a ) and ( 2 b ) , respectively] one can conclude that the ratedetermining step is not a simple bond fission [(3)+ (5)]
but a rearrangement by way of a relatively highly ordered
transition state [ ( 4 ) -+ ( 5 ) ] . For the methylthiomethyl ester
analogous to (2) no such strict correspondence of the kinetics
of ester fission in the gas and in the condensed phase was
R-C -0-C H2-0-C
(2a), R = H
( 2 b ) . R = C H3
stricted as a result of its partial double bond character (stabilization of the incipient carbenium center by rr-participation);
the reduced number of rotational states of the activated
complex is reflected in a somewhat lower frequency factor v.
The same processes are, however, also of interest because
of their temperature eflects. When benzyl acetate is protonated
with a weak Brgnsted acid, e.g. [C4H9]' (isobutane as the
reagent gas, isobutene as the conjugate base of the reagent
ion), then at T= 37°C one observes almost exclusively [MH] '
ions. At T=196"C, however, the concentration of [MH]'
is considerably reduced and the fragment [C,H7]' now provides the base peak of the spectrum. When [CH,]' is used
as the protonating agent, only a minimal temperature effect
is, however, observed ([C7H7]' is the base peak in the entire
temperature range). This contrasting behavior is ascribed to
the fact that decomposition of the protonated benzyl ester
is a weakly endothermic reaction and can thus occur sufficiently rapidly only on additional thermal activation; in protonation by [CH,]', a supply of additional thermal energy
is unnecessary because of the strong exothermicity of the
process (chemica/ activation).
An informative study of the influence of the reagent gas
on the nature and the importance of competing fragmentation
processes was carried out by Harrison et ~ 1 . ' These
~ ~ ~ authors
were able to show, again for protonated esters, that-in loose
analogy to the quasi-equilibrium theory (QET) governing EIa process with a relatively high activation energy is kinetically
favored if a) the internal energy of the precursor ion considerably exceeds the activation energy and b) the process has a
high frequency factor v (simple bond fission). The authors
showed that on ionization of (6) (Scheme 2) with [H3]'
(H2 as the reagent gas) the ion [R]' (10) is formed as the
main fragment in an energetically rather demanding but entropically favored process (except when R = CH3).When a weaker
Brdnsted acid is used the ion (10) will be suppressed since
condition a) is no longer fulfilled, and the formation of (8)
and ( 9 ) will be favored.
H2C :O-C H,
Scheme 1. Fragmentation of methoxymethyl esters ( 2 ) under CI conditions.
The CI-induced decomposition of benzyl and tert-amyl
acetate with loss of acetic acid from [MH]' ions has also
been investigated in kinetic detail[", ''I.
[C7H7]' and
[C5Hl'3' ions are formed with almost identical activation
energies (12.3 and 12.4kcal x mol-', respectively), whereas v
for the formation of [C7H7]' is smaller by an order of magnitude than that for the formation of [C5Hll]'. This difference
can be accounted for by assuming that, in contrast to (2),
the tert-pentyl ester decomposes by simple bond fission [analogous to (3)-+(5)], whereas in the transition state of the
benzyl ester, rotation around the Ph-CH'
bond will be re-
[*] The fact that also under E l conditions temperature can considerably
influence the extent of fragmentation was emphasized by Spireller [16].
Angew. Ckem. I n t . Ed. Engl. 17, 424-439 (1978)
Scheme 2. CI fragmentation of alkyl formates (6), R=CH3, C2H5,n-C3H7,
n-C4H9, i-C4H9, sec-C4H9.
Finally, it should be mentioned that in chemical ionization
structural features of the transition state of SN2 reactions[231
can be revealed. Dougherty et al. and Brauman demonstrated
which inherent properties[241(structure of the transition state
[Y ...R...X]-, nucleophilicity of [Yl-, degree of substitution
of R) contribute to the energy and entropy factors of the
reaction investigated [eq. (p)]. On comparison of the results
with data for the reaction carried out in the condensed phase
the authors indirectly established solvent effects.
3. The Chemistry of Chemical Ionization
3.1. Classification of Principal Reaction Typesr]
If we neglect the special cases of charge exchange (CE)
and "negative" chemical ionization (NCI, NCE), then, owing
to the absence of radical reaction centers, most of the behavior
of chemically ionized substrates can be ascribed to but a
few general modes of reaction such as are familiar from condensed phase ion chemistry. To a first approximation, in which
only changes in the relative positions of atoms are considered,
we can distinguish the following types of processes: 1)addition
reactions, 2) heterolysis, 3) rearrangement, 4) elimination reactions, and 5 ) replacement (substitution).
3.1.1. Addition Reactions
Irrespective of the nature of the newly formed bond, we
may consider, for example,the formation of collision-stabilized
complexes (described in Section 2.1) of the substrate M with
reagent ions (in general with an electrophile [El') as an
intermolecularaddition [eq. (k)]. Besides the "classical" electrophiles [C2H5]' and [C3H5]+ of the methane plasma,
[i-C4H9]' ions from isobutane, [NO]+['4b,251
, [SiMe3]+r261,
and recently [Li] '[271 are of importance, Relatively long-lived
adducts are obtained through formation of regular two-electron bonds from substrates with nucleophilic centers (n-electrons ofmultiple bonds,functional groups X with lone electron
pairs). If these centers are basic heteroatoms (e.g. in X), this
is also true for the adducts [M(NH4)]+ mentioned above,
in which partial bonds are responsible for complex formation
(hydrogen bonds M...H...NH3 and perhaps further H-bonds
in chelate-like structures when M contains additional basic
centers), and for the [M,E]' cluster ions[281that arise under
certain conditions (relatively long ion extraction times, high
partial pressure of M). Common to all these stable adducts
is their analytical significance for reliable molecular weight
determination of unknown compounds (cf. Sections 5.1.1 and
As opposed to these long-lived products of addition reactions, short-lived collision complexes can be recognized only
indirectly, due to their rapid decomposition into fragments
of often unexpected composition. Formation of complexes
of this type with X-substituted arenes by [CH,]' and [H3]+,
i. e. strong Bransted acids, was demonstrated only recently[2g1.
In both cases HX was eliminated from these complexes. In
the net result, this corresponds to the unexpected substitution
of X by CH3 or H [eqs. (9) and (r)].
(which are higher than those of analogous monofunctional
compounds) and also by strongly negative protonation entropies. Both ab initio calculations[321and the correlation of
thermochemical data with the ring sizes of ( 1 I ) suggest linear
geometry of the (X...H...X) bonds.
In conformationally mobile systems macrocyclic structures
of the type ( 1 1 ) can be formed with up to 49 ring
as is to be expected, with short-chain diols the entropically
favored ring formation does not take place if triple bonds
or trans-double bonds are present[341.Consequences of the
formation of such internal adducts for fragmentation are discussed in Section 3.1.5.
3.1.2. Heterolysis
Whereas addition reactions can occur both intermolecularly
or intramolecularly, heterolysis must, by definition, be an
intramolecular reaction. Effecting the simple cleavage of a
bond, heterolysis is the counterpart of an addition reaction.
The most simple, though not directly observable case of
heterolysis would be the reversal of eq. (k); such a reaction is,
in principle, possible on thermodynamic grounds, since dissociation and recombination energies are equal ; in practice,
the reverse reaction is largely prevented by stabilizingcollisions
with neutral molecules of the reagent gas. Heterolytic decomposition of [MH]' ions (M=RX) is, however, very often
directly observable; examples in Section 2.3 are the CI(H2)induced decomposition of the 0-protonated alkyl formates
(7) to form the alkyl ions ( 1 0 ) (Scheme 2) or of protonated
tert-amyl acetate in the temperature study discussed. In
general, the extent of heterolytic fission of protonated RX
compounds [eq. (s)] under standard conditions (same reagent
gas, same temperature, same residue R) is, to a first approximation, inversely proportional to the ionization potential of X
and to the basicity of HX['J,351.The [R]+/[RXH]+ ratios
thus increase within a series of compounds RX in the order
X =NH2< SCH3,OCH3, COOH <CN < SH < OH < I, Br, CI.
Quantitative predictions from the CI spectra are, however,
often thwarted by the fact that heterolysis of [MH]' ions
overlaps with heterolysis of [ME]' adducts. For example,
[C6Hll]+ is formed from protonated cyclohexanol in chemical ionization with i-C4HI0,but also in part by the decomposition of [M(C4H9)]+ ions by loss of H 2 0 or C4H90H,
Hydrogen bonding, for example, in [MH]' ions of specific
bifunctional compounds can be regarded as an intramolecular
addition reaction, i. e. as counterpart to the formation of bimolecular adducts [ME]' and [M2E]'. Corresponding formation of cyclic structures ( 1 I ) from c~,w-diamines[~~l
and a,wdimetho~yalkanes[~is, e. g., reflected in the proton affinities
[*] An attempt at a detailed classification is given in [7j].
+ H-X
Heterolysis of C-X bonds in onium ions can be considered
along with the subsequent fission of further C-C bonds,
as can occur under special conditions in (formal) carbenium
ions. Examples of this special case of heterolysis (usually
referred to as P-cleavage)are provided by the CI(CH4)analysis
of the photodimers ( 1 2 ) and (13) of cyclohexenone (Scheme
3), as reported by Field et ~ l . [ ~An
~ 'abundant
[MH]' ion
is observed for both isomers, but only ( 1 2 ) gives an intense
[M/2+H]' fragment ( 1 4 ) . It is assumed that ( 1 2 ) decomAngew. Chem. I n t . Ed. Engl. 1 7 , 4 2 4 4 3 9 (1978)
poses promptly to (14) by double 0-cleavage via a p-oxocarbenium ion (12a) as an intermediate; with (13) an analogous
genesis is likely to be quenched at this intermediate stage,
since the corresponding a-oxocarbenium ion ( 1 3 a ) represents
an energetically unfavorable intermediate. This case illustrates
how chemical ionization permits constitutional isomers to
be readily differentiated. As in this very example, this is often
not possible on electron-impact ionization.
effecting the usual proton transfer by Brqhsted acids [eq. (i)
and (i)]. Abstraction of a hydride ion by Lewis acids [eq. (t)]
can be regarded, in a way, as its counterpart. This fatter
process is often considered as a primary ionization event
along with protonation and adduct formation, although,
according to their masses, the resulting [M - H] ions already
represent fragments; moreover, it is not immediately possible
to differentiate hydride abstraction from decomposition of
short-lived [MH]+ ions (cf. Section 3.3.1) through loss of
Hz, i. e. from a true fragmentation process.
It was shown by an ion cyclotron resonance (ICR) experiment PI[39a-39clthat both hydride abstraction and H2-loss
from [MH]' ions take place in the chemical ionization of
alkanes with CH4 as reagent gas. The study revealed that
[C2H5]+ reacted exclusively as a hydride-acceptor as in eq. (t),
and not as a proton donor according to eq.
would be possible on thermodynamic grounds). In contrast,
[CH,]' was, however, found to furnish extremely unstable
carbonium ions (17) and (18) by protonation of the alkane
(16) (Scheme 5); these ions decompose quantitatively with
[ M H l + (13)
Scheme 3. Different CI fragmentations of constitutionally isomeric photoproducts ( 1 2 ) and ( 1 3 ) from 2-cyclohexen-1-one. Reagent gas: CH4.
When several functional groups are available for protonation a certain multiplicity of fragments may result from the
parallel occurrence of various possible heterolytic processes;
frequently,this multiplicity reflects the gross structural features
of the substrate more evenly and completely than is the case
with electron-impact ionization, in spite of the latter's
numerous competitive and consecutive fragmentation paths.
An illustrative example is provided by a simplified consideration of the competing primary heterolyses of ephedrine
(Scheme 4) (these are actually likely to proceed in a more
complex manner, presumably because of neighboring group
participation; cf. Section 3.1 .5)[381.
[ M H - HzO]'
7 0%
[ M H - CHsNHz]+
Scheme 4. C1 fragmentation of ephedrine: diverse reactions as a consequence
of protonation at different sites. Reagent gas: CHI.
Scheme 5. C1 fragmentation of n-alkanes with partial structure (16).
loss of H2 or smaller alkane molecules ("dissociative ionizati~n''['~])to yield [M-H]+ (19) and various [CnH2,,+l]+
ions (20). At the same time, the alkyl ions (20) also arise
from (19) by (formal) loss of alkene molecules, presumably
as a result of jkleavage (cf. Section 3.1.2); extensive C/H
rearrangements precede this terminal step[301.
The fact that in the protonation of C-C and C-H bonds
[(16)+ (17), (18)], and also in hydride abstraction
[(16)-+(19)] the attack by [CH,]' or [C2H5]+ occurs largely statistically at every bond (random-attack concept)['],
causes saturated hydrocarbons to undergo a multiplicity of
competing degradation reactions, which is unusual for chemical ionization. This more or less El-like variety of fragmentation processes is in distinct contrast to the behavior of functionalized molecules (e.g. RX in Section 3.1.2); in these the
primary ionization process is governed by ion/dipole interactions, causing a strictly localized attack of the reagent ion
and thus correspondingly localized subsequent decomposition.
3.1.3. Rearrangements
This type of reaction, more complex than the preceding
ones, arises formally from the combination of bond cleavage
with bond formation. An intermolecular variety of H-rearrangement can again be encountered in substrate ionization,
Angew. Chem. Int. Ed. Engl. 17, 4 2 4 4 3 9 ( 1 978)
['I In the case of competing reaction channels, it is possible to show conclusively which products of ion/molecule reactions arise from which ionic precurs o r ~by performing a double resonance experiment. As an integrating method
of measurement mass spectrometry with conventioiial chemical ionization
cannot disclose individual processes; for a review cf. [39d, 39eJ
The chemistry of carbenium ions in solution leads one
to suspect that in chemical ionization, too, intramolecular
rearrangements will neither be rare nor be limited to Htransfer. With unlabeled compounds, skeletal rearrangement
is usually hard to spot unless it involves rearrangement of
groups containing hetero-atoms whose dislocation is readily
apparent from fragment mass or competition. This applies,
for example, to the CI(CH4)-induced decomposition of 2,2,6,6tetramethylcyclohexanone. The [MH]+ ion (21 ) of this compound, formally corresponding to a carbenium ion, undergoes
Scheme 6. CI(CH4)-induced skeletal rearrangements of the Wagner-Meerwein
type in cyclohexanone derivatives with a high degree of a-alkylation.
a dominant fragmentation process in which acetone is ejected
as a neutral particle; this can be plausibly explained by ring
contraction to (21 a) with subsequent migration of the hydroxyl group [(21 a)+ (21 b ) ] (Scheme 6). In cis- and trans-2,6dimethylcyclohexanone the analogous process can likewise
be observed, though with ejection of acetaldehyde being much
less p r o n o ~ n c e d l ~ ~ ] .
3.1.4. Elimination Reactions
In chemical ionization, elimination reactions have in common with heterolyses that saturated or unsaturated neutral
particles are lost from ionized substrate. However, elimination
is more complex in that bond rupture is formally combined
with intramolecular rearrangement. The rearranged entity is
often an H atom. In heterolysis [eq. (s)] the neutral molecule
to be lost is already present as a leaving group: in elimination
this neutral molecule must first be formed in a reactive intermediate or transition state.
We have encountered examples of elimination reactions
in the preceding sections. The methoxymethyl esters ( 2 a)
and ( 2 b ) decompose after protonation following elimination
routes, by way of ( 4 ) , and not by heterolysis via ( 3 ) . In
the decomposition of [MH]’ ions (7) the heterolysis to (10)
is only the “high-temperature’’ path; under milder conditions
elimination becomes prevalent in this case as well. In the
decomposition to (8) and ( 9 ) , elimination processes even
o c c u r k two ways (1,l- and 1,2-elimination).In general, eliminations (like heterolyses) are of great diagnostic value for
structural analysis; in useful extension of EI fragmentation,
they frequently disclose several elements of the overall structure simultaneously rather than spotlighting one particular
detail. Since protonated hetero-atoms (onium centers) often
provide similarly favored reaction sites, substituents at these
centers are mostly better recognized in CI than in EI spectra;
in the latter, one single cleavage process may dominate fragmentation entirely. This is illustrated by the CI(i-C4HI,,)induced decomposition of the macrolide antibiotic oleandomycin (Scheme 7)17g1.By elimination processes the protonated
molecule loses both sugar constituents in succession
[ ( 2 3 ) + ( 2 4 ) - ( 2 5 ) l ; thereby not only these moieties are
disclosed but also the intact macrolide ring as the structural
core. Heterolyses of the protonated glycosidic bonds yield
the oxonium ions ( 2 6 ) and ( 2 7 ) , thus providing valuable
complementary findings. A diagnostically especially important
fragment is ( 2 9 ) : it is formed by elimination (opening of
the lactone ring in a manner similar to a McLafferty rearrangement) and, after protonation of the carbonyl group in position
9, by heterolytic P-cleavage (similar to a retroaldol fission).
This process also occurs with other macrolides containing
a 14-membered ring, e. g. with erythromycin B17g1,
from which
an ion with m/e = 99 results due to the presence of a 13-ethyl
substituent. This fragment is thus suitable for an approximate
localization of substituents within the ring section encompassed (C-I 1 to C-23). Decomposition of these compounds
on electron impact is characterized by complete disintegration
of the basic structural units; only residual structures of the
amino-sugar components are recognizable (due to the strong
fragmentation-inducing effect of the Me2N substituent).
Analogously, a direct insight is often obtained into the way
subunits are linked in oligomeric molecules,as is often revealed
[ M H ] + ( 2 3 ) m / e = 688, 100%
mle = 158, 30%
- MeOH
mJe( 2=7 )1 4 5 , 25%
Scheme 7. CI(i-C,H,,)-induced heterolyses and elimination reactions in oleandomycin
Angew. Chem. lnt. Ed. Engl. 17, 4 2 4 4 3 9 ( 1 9 7 8 )
only by extensive chemical degradation. The CI-fragmentation
of peptides can serve as an example representative of many
other potential applications to the analysis of natural products
(e.g . nucle~sides[~'~
or olig~saccharides[~~I).
The principal
advantage of chemical ionization over electron-impact ionization for peptide-sequence analysis-a classical[43]and yet also
recent[441objective of mass spectroscopy-lies in the fact
that both N-terminal ions and C-terminal ions are produced
simultaneously [eq. (u)][~']. Unless very special derivatization
is employed ['I, electron-impact ionization affords only ions
of the former kind.
and [H3N-R]+
In CI, the C-terminal ammonium ions are thereby formed
by elimination (ejection of a neutral aminoketene molecule);
for the genesis of the N-terminal acylium ions the question
of involvement of elimination or heterolysis (cf. Scheme 1)
remains unresolved. Since, in principle, protonation can occur
at every single peptide bond, two series of fragments are
produced which, owing to the redundancy of their information,
should permit the deduction of the amino-acid sequence even
when isobaric fragment ions interfere. Thus, there is little
need to resort to high-resolution mass spectrometry:
3.1.5. Substitution Reactions
According to the formalism used, substitution reactions
in chemical ionization consist of a combination of two simple
steps: the displacement of a leaving group in the substrate
by a nucleophile amounts to a heterolysis associated with
the addition of the nucleophilic reagent [eq. (v)]. As outlined
in Section 3.1.2, in a substrate of type RX the formation
ofpotential leaving groups is a direct consequence of ionization
at X. Depending on the external or internal origin of the
nucleophile Y: (reagent gas or functional groups within the
substrate), again an intermolecular and an intramolecular
mode (SN2or SNitype) are conceivable.
+ [R-X-H]++
+ H-X
(31), n = 1, 2
More direct indications of interactions of the SNitype can,
of course, be expected for stereochemical effects[52a1.Marked
reactivity differences may thus be exhibited by polyfunctional
alicyclic systems, depending on the spatial arrangement of
the substituents. An elegant study of 1,3-, 2,3-, and 3Pdifunctionalized steroidal amino alcohols was reported by Longev i d e et a1.[s2b1.
These authors found that in all compounds
in which intramolecular H bridges between the substituents
were sterically possible [e.g. in (32) and (33)], there was
no ejection of H 2 0from [MH]+; such ejection was prominent
when the leaving group was not hydrogen-bonded and SNi
attack was configurationally favored [e.g. in the 3,4-diaxial
or 1,3-equatorial/axial trans-compounds ( 3 4 ) and (35)].
Furthermore, proportions of specific conformations of the
protonated forms were estimated from the [MH]+/
[MH - H2O]+ ratio. The results showed surprisingly good
agreement with IR-spectroscopic measurements.
It is uncertain whether direct intermolecular substitution,
i.e. a process corresponding exactly to the SN2 type, plays
a role in chemical ionization [eq. (p) and Section 2.31. In
their net results, occasionally observable hydrolyses like those
of trimethylsilyl and perfluoroacyl derivatives on use of H 2 0
or other insufficiently dried reagent gases, fall into this category, as do the a m m o n o l y ~ e s [and
~ ~ ]transesterification react i o n ~ [ ~ofesters
on use of NH3 or CH30H. However, detailed
investigation of such reactions of the ionized substrates with
the reagent gas itself, e.g. the hydrolysis of aliphatic dimethyl
u,o-dicarboxylates in dependence on the number of carbon
atoms in the chain (3-12)'471, compel one to assume more
complex pathways, namely elimination of HX from [RXHY]'adducts with neighboring group participation by the second
[*] Reduction of the polypeptide chain to polyamino alcohols by Biemann's
method [43, 441 with LiAIH, intentionally generates optimum conditions
for a uniform formation of C- and N-terminal fragments.
Angew. Chem. l n t . Ed. Engl. 1 7 , 4 2 4 4 3 9 (1978)
ester group instead of direct HX/Y-substitution in [RXH]'
In contrast to this intermolecular mode, intramolecular
substitution in chemical ionization is relatively well documented. SNireactions are reflected, for example, in protonated
open-chain bifunctional molecules in heterolytic processes that
are absent or unimportant in monofunctional analogs under
comparable conditions. [MH]' ions of simple amjno acids,
for instance, undergo no appreciable ejection of NH3, in sharp
contrast to those containing additional functionalities; thus,
S-methylcysteine is readily converted into a cyclic sulfonium
ion (30) by an Sxi reaction[481.Analogous reactivity was
observed with a series of polyamines under CI conditions*491,
which is paralleled by their EI behavior'501.The formation
of bicyclic oxonium ions of structure (31) must be interpreted
along similar
The fact that H-bond formation in bifunctional molecules
(cf. Section 3.1.1) can also block sterically unhindered SNi
reactions follows from investigations of open-chain compounds. In the CI(CH4) spectrum of n-decanol, [MH]' ions
are completely missing, mainly due to the loss of H20[61;
in the CI(CH4) spectrum of 1,lO-decanediol ( I l ) , X=OH,
n=10, the [MH]'
ions retain as much as 60% relative
As shown by investigations of alicyclic diols and t r i o l ~ [ ~ ~ I , care of in commercial instruments) to allow the ion source
alkyl effects play an additional role. In general, in compounds
to be operated at a working pressure of around 1 mbar (in
of the type (36) the loss of water from LMH]’ is quite
mbar) through fairly gas-tight conEI instruments ca.
struction“]; suitably high pumping capacities have to maintain the necessary minimum vacuum
outside the ion source for operation of a heated cathode[**]
as well as in the analyzer to avoid collisions of ions with
residual gas. Pump performances-of 600 to 12001~s-’ at the
source housing and differential pumping of the analyzer are
(37), R = O H , R‘ = H (11P-OH)
thus necessary. In magnetic field instruments, in which the
( 3 8 ) . R = H, R’ = O H ( l l a - O H )
ion source is held at high potentials (acceleration voltage
in the kV range), precautious must also be taken for electrical
sensitive to the configuration of the group R. For the
insulation of the reagent gas, which will become conducting
[MH]+/[MH-H20]+ ratio, the case R = O H provides the
the prevailing
591. The technical simplicity
upper and R=CH3 the lower extreme value. Alkyl effects
out by the fact that
may prevent the formation of linear H-bonds for example
are used for analysis
in the diterpenediols (37) and (38); this effect may be so
of single substances as in electron-impact ionization. As in
marked that the spatial arrangement of the hydroxyl groups
EI, in CI mass spectrometry the combination with gas chromais no longer of importance for the extent of loss of
tography (GC) leads to a major broadening of the scope
Apparently, the space-filling by the lop- and 13P-methyl
of the technique, since analyses of complex mixtures become
groups of (37) precludes the formation of such a bridge;
possible (cf. Section 5.1).
in (38) it is ruled out in any event for configurational reasons.
Technically, combination of gas chromatography with CI
Indirectly, cooperative effects of this type may also be of
spectrometry is simpler than that with its EI counterimportance for the behavior of open-chain compounds.
least as far as the level of performance of monogas
the ejection of
As a result of a geminal dialkyl
6 3 *641 is concerned. The reagent gas, usually CH4,
methanol from the [MH]+ ions of, e.g., a-(benzoylamino)
through the gas chromatograph and thus serve
acid methyl esters is strongly dependent on additional a-substiat
as a carrier gas. Because of the “high”
tution: in ( 3 9 ) the amount of loss of CH30H is considerable
the ion source, molecular separators or
(25 % relative intensity, [MH]’ =base peak), in (40) it is
unnecessary, i.e. can be replaced in
not even detectable1561(Scheme 8).
favor of direct connections which operate largely without
loss of sample and with minimum failure and maintenance
requirements[58.631. Gas chromatograms are obtained by
recording total ion current (TIC) representative of the sample[***].Independent optimization of separation and reagent
gas properties is made possible in the dual gas concept[**’*]
(39), R CH,
which different carrier and reagent gases are employed.
(40), R = H
to avoid interference with the separation process,
Scheme 8. Loss of C H 3 0 H from protonated u-benzoylamino acid methyl
the reagent-gas is admixed with the carrier gas flow only
esters in dependence on u-alkylation (alkyl effect).
shortly before entering the ion source[65-681. On introduction
of capillary-GC/CI mass spectrometry with typically low car4. Instrumental and Methodological Aspects
rier gas flows (ca. Iml/min) relative to that of reagent gas
(20 to 30ml/min) this concept could be realized in its “purest”
4.1. Instrumentation for Chemical Ionization
form, i. e. without compromising the versatility of CI operation
by major carrier gas contributions (Fig. 2)1653 6 7 - 701.
As far as instrumental aspects are concerned, the only essenFor packed columns with considerably higher carrier gas
tial difference between chemical ionization (CI)p] and electronflow (30-50ml/min) either the carrier gas (noble gas, N2)
impact ionization (EI) lies in the manner in which ions are
produced. Mass separation and detection are based on the
same principles in both techniques. Conventional EI mass
[*] The number and the dimensions of the source apertures into the housing
spectrometers can thus be modified for CI or combined EI/CI
are much smaller than in EI sources. The electron entrance and ion exit
operationC“1 with little effort. No loss of resolution was
apertures have together an area of only 1-2 mm’.
observed for high-resolution instruments adapted to CI operap*] For primary ionization of the reagent gas other principles than the
customary emission of thermal electrons are available (corona discharge,
tionc5’, 711.Modifications are required (they are already taken
use of radioisotopes such as 63Ni).Plasmas of aggressive reagent gases (e.g.
0 2 ,NO) are produced by gas discharge [60-621.
[‘I Since the working pressure is significantly higher in chemical ionization
than in electron-impact ionization, the term “high-pressure” ionization is
occasionally used.
In principle, every CI source becomes an EI source as soon as the
flow of reagent gas is interrupted. Unless special modifications are made 1571
the sensitivity in the EI mode considerably decreases, however, because
of the smaller aperture dimensions. The simultaneous or alternative operation
ofa Combination ofan independent CI and EI source has been advocated [58],
and at least one such dual source is commercially available.
[***I The total ion current (TIC) signal cannot he sampled before mass separation, as is possible in GC/EIMS under certain conditions (helium as the
carrier gas, 20eV ionization energy), since it is predominantly due to unused
reagent ions. In dynamic spectrometers, total ion current representative of
the sample can be conveniently obtained after mass separation by integrating
the single ion currents over preselected mass ranges, i.e. with exclusion
of reagent ions, in rapid scan cycles (0.1 s). In magnetic field instruments
this .is less easily achieved.
First described for (non-chromatographic) direct analysis [63].
Angew. Chem. lnt. Ed. Engl. 17, 424-439 (1978)
Fig. 2. Block diagram of a GCjCIMS system in typical dual gas operation
(cf. [71]). Principal components: a) gas chromatograph with Grob injector,
b) capillary GC column, c) split systems for dual FID/TIC recording, d)
flame ionization detector (FID), e) coaxial interface, 1) reagent gas inlet,
g) mass spectrometer, h) dedicated computer.
is modified[’] in the desired manner by addition of appropriate
amounts of a reagent gas (e.g. i-C4HI0 or NO)[661,or most
of it is removed in molecular separators or flow dividers
before the reagent gas is added. In the former case the full
sensitivity of the system is retained at the expense of some
flexibility in the choice of the CI conditionsP*]; in the latter
case complete flexibility of the experiment is retained as in
capillary-GC/CI mass spectrometry, yet at the expense of
sensitivity owing to unavoidable losses of material on pressure
4.2. Methods of Chemical Ionization
4.2.1. Choice of the Reagent Gas
The choice of the reagent gas is of decisive importance
for the achievement of the analytical purpose. This choice
determines not only the type of ionization (redox or acidbase
type)[”*], but also the exothermicity of the ionization process.
As noted before, these two factors largely shape the information
to be derived from the CI spectrum. According to the magnitude of the difference (A) between the recombination energy
or the proton affinity of the reagent gas and the substrate,
the result will range between extremes of high ionization
selectivity, reduced fragmentation, and thus low substrate specificity (low A value) on the one hand, and high substrate
specificity, extensive fragmentation, and low selectivity as a
result of universal ionization (large A value) on the other.
Control of selectivity and fragmentation is important not
only for the analysis of multicomponent systems, in which
it allows discrimination of disturbing components, but also
for structural analyses of isolated, single compounds that contain several functional groups distinguished by different proton
[*] Addition of approximately 10% reagent gas to the carrier gas [15d]
usually sufficesto moderate the latter’s “hard” charge-exchange (CE) character
(addition of NO), or to modify it to a CI reagent (addition of i-C4HLO).
[**I In this case, the carrier gas serves only as an inert collision gas [63],
which is essential to maintain CI conditions.
[***I Depending on the substrate (differences between recombination energy
and proton afinity), one reagent gas can act both ways simultaneously:
in the typical acidic CH4 plasma [C,H,]”
and [C2H5]+ ions are potential
charge-exchange (CE) reagents [7j].
Angew. Chem. I n t . Ed. Engl. 1 7 , 4 2 4 4 3 9 (1978)
In this connection it is worth mentioning experiments with
binary reagent gas systems, including partially modified syssuch as
tems (addition of modifying gas <
Ar/H20[’5a1, He/H20[’5b1,and many others“ 5d1; mild protonation is generally used to provide stable [MH]’ ions,
while strongly exothermic charge exchange[’4b1 produces
highly reactive [N]” ions that decompose very much in
an EI-like fashion. In this way one attempts to combine the
advantages of chemical ionization with those of electron
impact (cf. also Fig. 1). In single-compound analysis there
may be real advantages to such a “hybrid” concept; in the
analysis of complex mixtures by GC/CIMS, however, a sequential approach by independent, yet ~orrelatable[~’~
single experiments with only one reagent gas at a time (a “puristic” concept)
may frequently be
4.2.2. Selected Ion Monitoring
This recording technique, which was initially developed
and proposed for EI mass spectrometry under the name of
“mass fragment~graphy”[~~][*I,
has today acquired similar
importance for CI mass spectrometry, especially in GC/CIMS
analysis. In this technique the usual recording of complete
spectra(i. e. repetitive scanning of wide-largely empty-mass
ranges)during a separation process (e.g. fractional evaporation
or GC/MS) is abandoned; instead, during the same period
of time only a few selected masses are monitored continuously
by rapid switching rather than sweeping electric or magnetic
fields. Thus, for these few ions deadtimes are reduced by
several orders of magnitude, with the result that detection
limits are lowered dramatically from the nanogram down
to the lower picogram levellsome 10- g). Since this detection
principle is independent of the mode of the ion production,
it lends itself immediately to GC/CIMS analysis.
5. Analytical Applications
5.1. Analysis of Organic Compounds
This Section intends to point out applications of chemical
ionization that are characteristic of both the analytical objective (determination of structure, identification,detection, quantitative determination) as well as of the fields in which it
is chiefly applied (biomedical field, environmental sciences).
For more complete presentations the literature should be
Of the advantages CI may offer Over EI mass spectrometry
for analytical purposes reliable molecular weight information
and the possibility of controlling fragmentation are of prime
importance. This is true not only for structural analysis but
also for the identificationof known, yet unexpected compounds
present as minor components in mixtures. The detection and
quantitative determination of expected, known compounds
in the presence of unknown constituents or ballast material
may likewise profit from control of fragmentation behavior.
Pronounced [MH] ions and deliberately suppressed frag+
[*I After considerable terminological uncertainty [74] (the synonyms “multiple ion monitoring”, “multiple ion detection”, “selective ion recording”,
“specific ion plotting” and many more were in use; cf. also [73a]) the acronym
SIM (“selected ion monitoring”) seems to have found wide acceptance [73].
mentation are, of course, important non-instrumental prerequisites for high detection sensitivity, as is essential in trace
analysis since they reduce the risk of spectral interference
by foreign substances (which will also fragment to lesser
degrees, and hence be more readily detected than in EI mass
spectrometry). The ease of suppressing interferences, the
greater reliability in qualitative applications, and the higher
accuracy in quantitative work by measurement in the less
disturbed region of highest possible mass (thus excluding systematic errors) are direct advantages. The control of ionization
also results in great flexibility as to the selectivity desired;
its optimization frequently permits the complete exclusion
of disturbing accompanying materials from ionization, and
thus a reduction of background in favor of lower detection
limits. In these respects, electron-impact ionization is rather
Finally, in structural analysis and identification, the great
versatility of chemical ionization allows for the use of a series
of reagent gases, in order to probe, step by step, for specific
structural evidence[70! This extension to multiple analyses
of one and the same sample is certainlyjustified in the analysis
of complex mixtures by GCjCI mass spectrometry; the expenditure of work is then economized when the structural characterization of more than only a few compounds is required,
and when difficult problems of correlation can be avoided
by keeping the GC parameters constand65*67’70.711
throughout the series of GC/CIMS runs. It also has to be kept
in mind that other ways are currently unavailable if one
attempts to compensate for the lack of structural information
otherwise provided by other spectroscopic means.
(426) [MH]’
( 4 4 ) (3%)
Scheme 9. CI(CH4)-induced decomposition of botryodiplodin ( 4 2 ) .
to chemical
(incidentally,this was the first application of CI to structure analysis). Although the intensity
of the [MH]’ ions was relatively low due to excessive decomposition of ( 4 2 a ) into (43)-(as usual at that time, methane
was used as the reagent gas) the molecular weight and empirical
formula could be determined. Indication of a P-oxygenated
carbonyl system with no substituent at C-5 was afforded by
the loss of CH2=0 from the [MH]’ ion [ ( 4 2 b ) + ( 4 4 ) ] .
This type of fragmentation is similar to the retroaldol decomposition of oleandomycin [(23)+ (29), Scheme 71; since
double j3-cleavage occurs, it can also be considered as a Grob
in the gas phasd”. ’’I. From the above and
the results of ‘H-NMR analysis, structure (42) was finally
arrived at in which only the stereochemistry at C-3 and C-4
remained unestablished.
5.1.1. Elucidation of the Structure of Unknown Compounds
Chemical ionization, like electron-impact ionization, will
frequently allow solution of only part of a complex structural
problem. In any case, its most likely minimum contribution
will be the determination of the molecular weight and, when
using high-resolution instruments, the determination of the
elemental composition. This information alone is especially
valuable because of its high reliability[**].However, the greater
simplicity and better predictability of CI fragmentation, which
is often complementary in character to that of EI mass spectrometry, and in particular its better correlation with structural
elements will make its extensive evaluation highly attractive
in structural work. Little effort has as yet been made in exploiting this feature in practice.
The situation is illustrated by the determination of the
constitution of the antibiotic botryodiplodin ( 4 2 ) (Scheme
9). In view of the poor reproducibility of the EI behavior
of the compound (tautomerizable five-membered ring lactol
structure with an anomeric center at C-2), recourse was made
[*] The attempt to achieve selective ionization in EI by reducing electron
energy to below the ionization threshold of individual classes of compounds
is known to cause a prohibitive loss in sensitivity.
[**I As in the EI method, limitations arise from the necessity to vaporize
the sample without thermal decomposition. However, pyrolysis is easier
to recognize in CI than in E l spectra, particularly when more than one
product, i.e. several quasimolecular ions (cf. footnote [*I on p. 436), as is
the case with certain quaternary nitrogen compounds [75]. If in this latter
case only one pyrolysis product is formed, the presence of a covalent compound
can be simulated (e.g. in the phospholipid dioleylphosphatidylcholine
ROP(O)(O-)OCH2CH2NMe: by intramolecular methylation to thecovalent
Scheme 10. CI fragmentation and biotransformation of phencyclidine ( 4 5 ) .
(45) with a perdeuterated phenyl group affords an ion with m/e=249 instead
of ( 4 6 ) , and an ion with m/e=164 instead of ( 4 8 ) . ( 4 7 ) remains unshifted.
As a result of their small absolute amount and low degree
of enrichment, compounds “isolated” from physiologicalfluids
often cannot be subjected to parallel spectroscopic investigations like those used for ( 4 2 ) . Structural characterization
of drug metabolites is clearly such ;I L‘J\C. 11 ork of this kind
is admittedly easier than that on natural products because
the structure of the starting material is known and the biotransformation is almost invariably confined to the molecular peripherypl. Determination of the structure of two metabolites
[*] In principle, skeletal changes are not entirely excluded. Cases of ring
contraction (7-ring --t 6-ring) are known, e. g. from the biotransformation
of SH-dibenz[bf]azepine-5-carboxamide (carbamazepine) [SO].
Angew. Chem. Int. Ed. Engl. 1 7 , 4 2 4 4 3 9 ( 1 9 7 8 )
of the local anesthetic phencyclidine (45) can serve as an
example (Scheme 10f811.GC/CIMS analysis (packed column,
monogas system with CH4 as universal gas) of a silylated
crude extract from human urine showed, besides unchanged
parent drug (45), two GC peaks whose spectra corresponded
to trimethylsilylderivatives of isomeric monohydroxy metabolites (49) and (50) ([MH]+: m/e=332). Marked losses of
trimethylsilanol in both cases ruled out substitution of aryl
positions. Comparison of these spectra with the fragmentation
of the parent compound [larger structural elements are
revealed in a manner typical of chemical ionization:
(45) -+ [MH]' -+ (47) by elimination of phenylcyclohexene;
( 4 5 ) - t [MH]+ --t (48) by heterolysis with loss of piperidine]
rapidly permitted the substitution region to be narrowed down
to one of the saturated rings in both cases. Silylated (49)
decomposed to silyloxy-(47) (rn/e= 174) and unchanged (48),
while silylated ( 5 0 ) decomposed to silyloxy-(48) (m/e= 247)
and unchanged (47). The complementary nature of EI and
CI is borne out by the fact that in the EI mass spectra
[MI" ions were much less intense than [MH]' ions in
CI, but that the EI fragmentation allowed an exact placement
of the hydroxyl group in one of the two metabolites (50).
The structural proposals (49) and ( 5 0 ) were subsequently
confirmed by synthesis.
A similar situation may arise in structural investigations
of hard-to-separate reaction mixtures from preparative organic
chemistry; admittedly sample sizes are mostly larger and
sample compositions less complex than when analyzing
biological materials. Although the starting materials are
known, the variety and number of the products may be large
enough to thwart their isolation and application of conventional methods for structure elucidation or confirmation.
As the following example shows, multiple GC/CIMS analysis
with variation of the reagent gas[65,67,7'1can be of help in
such cases. The sample studied originated from a co-oligomerization experiment with butadiene and the Schiff base (51)
(Scheme 1 1). Under Ni(0)-ligand catalysis, among other
products, 2: 1 adducts mainly of the types ( 5 2 ) and (53)[821
were formed.
E tzCH-C H=N-C HzC HEtz (51)
/N-C H& HE tz
(54) E tzC H-C?
c 8%
.N=C H-C HE tz
E t zC H-CP
,N=C H-C HE tz
lIH1 7
Scheme 11. Co-oligomerization of 1,3-butadiene with the SchX base ( 5 1 )
under nickel@)-ligandcatalysis with subsequent hydrogenation.
Figure 3IS3lshows the gas chromatogram of a distillation
fraction after (incomplete) hydrogenation to ( 5 4 ) and (55).
The difficult separation of the components 3/4 and 5/6 required
the high resolution of glass capillary columns in the GC/CIMS
experiments. Gross structural characterization of the imporAngew. Chem. lnt. Ed. Engl. 1 7 , 4 2 4 4 3 9 ( 1 9 7 8 )
I t
Fig. 3. CI(CH4)-totalion current-gas chromatogram (TIC-GC) of a partially
hydrogenated reaction mixture obtained by co-oligomerization of butadiene with the Schiff base ( 5 1 ) under Ni(0)-ligand catalysis (glass capillary
column, 20m x 0.3 mm, SE-54). Components of the mixture: 1)
H2NCH(C8H,7)CHEt2,2) and 3) Et 2CH-CH=N-CH(C8H
,,)CHEt2 (55 ),
Et2CHCH2NHCH(CsH, 7)CHEtz f 5 4 ) ,
5) EtZCH-CHzN,)CHEt2, 7)
CH(CgH13)CHEt2 (Srl), 6 ) EtZCH-CH=N-CH(CaHI
Et2CHCH2NHCH(CsH15)CHEt2(see Scheme 1 I).
besides confirming the [MH] findings, furnished distinct
structural information based on fragmentation similar to that
of electron-impact ionization: pronounced [M - H I + , [M C5H11]+,and [M-C8Hl7]+ ions indicate which substituents
occupya-positions with respect to the N atom in the secondary
amines and Schiff bases. CH30D, like D20IE4],
is an efficient
reagent gas for the determination of the number of exchangeable H atoms (two in component 1, one in 4 and 7, none
in 2, 3, 5, and 6); in the present case it serves as a probe
for the recognition and distinction of primary and secondary
amines in the presence of Schiff basest69%85!
5.1.2. Identification and Detection of Known Substances
, N?/PPhp
,N-C HzC HEtz
(52) E t z C H - C P
tant components sufficed for an evaluation of the oligomerization experiments, and was achieved by employing the reagent
gas sequence CH4/i-C4Hto/CH30D. Isobutane invariably
gives [MH]+ ions as the base peaks and thus reveals molecular
weights, which are hard to obtain by electron-impact ionization in the case of secondary aliphatic amines (54). Methane,
A certain minimum of structural characterization is normally required when there is no specific clue as to the identity
of the compound in question, i. e. there is only an indication
of the class to which the compound belongs. If additional
complications arise from hard-to-separate mixtures, from a
low degree of enrichment, or from small sample quantities,
then, for reasons given above, chemical ionization will be
the method of choice.
The widespread abuse of drugs provided a challenge to
the exploratory application of the new technique for the identification of these materials in clinical analysis. Direct sample
introduction, applied to crude extracts of gastric contents,
urine, or blood was performed at the National Institutes of
Health by a group centered around Fales and Milne some
years ago, as a means of simple rapid screeningC'1[88].The
[*] The potential of GC/MS for this field was clearly recognized 1861, but
instruments then available operated relatively slowly and were generally
reserved for non-routine problems [87].
procedure consisted in taking a series of CI(i-C4H1o) spectra
during evaporation of the sample and searching them for
[MH]+ ions“]; visual comparison ([MH]+ as search parameter) against the 48 reference spectra then available made
it possible to identify rapidly at least the most often abused
drugs. The spectrum of a gastric extract (Fig. 4) illustrates
an overdose case in which five constituents [(56)-(60)]
80 -
3 00
Fig. 4. CI(i-C4Hlo) spectrum of an extract of gastric contents after drug
abuse (overdose of Percodan). ( 5 6 ) , acetylsalicylic acid (aspirin); (57), phenacetin; (M),caffeine; ( 5 9 ) , amytal; (60), oxycodon; (611, dioctyl phthalate.
of the preparation Percodan, even in spite of pronounced
fragmentation of one component [aspirin (5611, could be
identified[88! While contaminants, e. g. the ubiquitous dioctyl
phthalate (61 ), barely interfere, the formation of cluster ions
of the type [MzH]+ (cf. Section 3.1) by phenacetin (57)
is to some extent a disadvantage; cluster formation of this
kind is a consequence of high momentary sample pressure
which is not always easy to avoid, yet fortunately readily
In forensic medicine common “street drugs” and narcotics
such as heroin can be similarly identified by direct analysis
of the confiscated material as
Diluents and current
“narcotic substitutes” are simultaneously identified, and this
often enables the investigating authorities to draw conclusions
as to the origin of the material[**].
In the past years the increasing spread of technically more
sophisticated GC/MS/computer systems, and the resulting
[*] The presence of several [MH]+ ions (e.g. when combined preparations
were ingested) causes irregular changes of the intensity profiles within a
series of subsequently recorded spectra, in contrast to the presence of CI
[**I For detailed analysis, the GCjCIMS version is, of course, likewise
recommended; in this context, the use of HejNO (lO/l) as a moderated
CE reagent gas has been reported [66].
automation of spectra comparisons, have replaced much of
the initial skillful improvisation in this field. Importance still
attaches to direct analyses in cases in which thermal lability
or low volatility excludes the use of GC/MS analysis in spite
of derivatization. Examples are the identification of LSDrgol,
or of explosives for which chemical ionization with HzO
as reagent gas has proven ~ a l u a b l e ~ The
~ ’ ~ .general trend
toward automated identification is reflected in the inclusion
of CI spectra in larger general collections of mass spectra,
and also in the growth of specific collections of CI reference
spectra for drug identification to currently 450 compounds
(including the main metabolites and customary associated
s u b s t a n ~ e s ) ~An
~ ~approach
developed at the Batelle Institute
in Columbus, Ohio, is based on a GCjCIMS analysis of
crude extracts; a monogas system operating with packed GC
columns and methane as the carrier/reagent gas is used[g0.921.
For compounds of equal molecular weight (which serves as
the main search parameter) fragmentation provides sufficient
distinguishing criteria. Because of the dependence of CI fragmentation on experimental parameters, strict standardization
of the conditions is, of course, necessary.
Detection sensitivities very much greater than in the preceding examples are required for analyses of trace components,
while at the same time, high substance specificity must be
maintained. This applies particularly to the detection of trace
components in very complex mixtures. A breakthrough in
this direction (detection limits in the lower picogram range
and less) was achieved by the technique of selective ion monitoring (SIM), described in Section 4.2.2[72J.A result of general
interest, relying on this technique, is the detection of traces
of nicotine in the urine of non-sm~kers[~~].
This analysis,
which required an extremely high detection sensitivity and
very narrow selectivity of ionization, was performed by Horning et al. These authors developed a special v e r ~ i o n I ~ ~ . ~ ~ l
of chemical ionization using reagent gases at atmospheric
pressure (API) in an “open” ion source; the detection limits
of such a system are in the upper femtogram range (Ifg=
10- g). In a typical experiment a benzene extract of the urine
is injected directly into the source flushed with Nzat atmospheric pressure, and the solvent vapor also acts as a highly
selective CE reagent gas ([C6H6]+’ ions). The presence of
nicotine in urine was evidenced by the SIM profile of m/e= 162
([M]”). Surprisingly, the alkaloid was also detected in the
urine of all non-smokers tested: the nicotine levels of these
apparently “passive smokers” were ca. 5 % of the average
content for smokers examined at the same time.
“Negative chemical ionization” (NCI and NCE ; cf. Section
2.1) has opened up some interesting and highly topical possibilities for the detection of traces of special compounds. This
applies particularly to 2,3,7,8-tetrachlorodibenzo[l,4]dioxin
(TCDD), which recently made the headlines. The molecule
of (62) is remarkably stable in every respect but, after the
formation of a radical-ionic adduct (63) with Oz (reagent
gas in a Townsend ion source), it decomposes readily into
an ionic fragment ( 6 4 ) (Scheme 12).
By means of this fragment, (62) can be detected in extremely
low concentration (a 2 x
g sample gave a signal/noise
ratio of over 50: 1 on the SIM trace[961)and, owing to the
specificity of the process and the relatively low mass of the
fragment, can be easily differentiated from other wide-spread
Angew. Chern. Int. Ed. Engl. 17. 4 2 4 4 3 9 (1978)
(64) mfe = 1 7 6 , %1
Scheme 12. Decomposition of 2,3,7,8-tetrachlorodibenzo[l,4]dioxin ( 6 2 ) after
negative chemical ionization (NCE) with O2 as the reagent gas.
polyhalogenated compounds which are hard to separate by
present techniques.
5.1.3. Quantitative Analysisr]
Here too, direct introduction of the sample with fractionating evaporation may be of certain interest as the simplest
mode of analysis when, for example, only one or a few main
components of a simple mixture are to be determined quantitatively. A typical case in which derivatization and GC/CIMS
analysis proved hardly worthwhile has been reported by FaIes
et ~1.['~].
These authors were concerned with the quantitative
determination of phenylthiohydantoin (PTH) derivatives of
amino acids which they obtained in decreasing purity in the
course of a sequential Edman degradation of sperm whale
myoglobin, i.e. in admixture with PTH derivatives of the
amino acids next in the sequence. After identification of all
the PTH-amino acids present by means of a preliminary CI
run, quantitative determination was successfully achieved by
the addition of the corresponding C6D5derivatives in known
amounts as internal standards (isotope dilution analysis). This
permitted the identification of the chief component in the
mixtures obtained at every single step (in contrast to visual
inspection of spectra beyond step 20), and thus the deduction of the correct sequence of the protein chain. The
concentrations of the PTH derivatives were obtained from
the relative intensities of the [MH]' ions from the sample
and standard in the very simple CI(i-C4Hlo)spectra.
For quantitative determination of very low concentrations
in multicomponent mixtures, GC/MS with the use of selective
ion monitoring (SIM) is, of course, much more effective. The
GC/CIMS combination is a very promising alternative to
GC/EIMSrg8]because of the advantages mentioned above,
especially the lower susceptibility to interference by material
co-eluted with the sample. As in the direct-introduction mode,
an internal standard is added in a known amount (depending
upon the expected concentrations) to the original medium,
e. g. blood plasma, before isolation and clean-up (generally
by extraction) of suitably enriched fractions. As with the above
phenylthiohydantoins, the current practice is to use standards
multiply labeled with stable isotopes of high isotopic purity[95b1.Multiply deuterated or 13C-labeled standards with
at least 3 amu mass difference from the unlabeled substrate
are particularly suitable, because they preclude overlap in
the simultaneous recording of selected [Do]/[Dn] mass pairs
even in the presence of the [MH]+/[M]+'/[M - H I t peak
groups in CI(CH4) spectra (should CH4 be unavoidable as
reagent gas).
A typical case of pharmacokinetic investigations is the quantitative determination of phencyclidine ( 4 5 ) (Scheme 10)
(already mentioned in Section 5.1.1) as unchanged parent
drug in blood and uriner811.The derivative with a perdeuterated phenyl group is the internal standard and CH4 is the
carrier and reagent gas (monogas system).As mentioned above,
only the two main fragments ( 4 7 ) (m/e=86) and ( 4 8 )
(m/e=159) are prominent apart from [MI" and [MH]'
(m/e=243 and 244; relative intensity 100 and 50%, respective1y)yl. Unlike ( 4 8 ) , fragment ( 4 7 ) contains no heavy isotope
in the case of the [D,]-derivative and, thus, cannot be used
as a reference mass. The mass pairs m/e= 243/248 and 159/164
are, therefore, used for selective ion monitoring. The recording
of two ions characteristic of the compound, together with
practically identical retention of the substrate and the added
standard provides sufficient compound specificity and is nearly
equivalent to a duplicate analysis by independent routes. Interference at one of the two selected masses (e.g. by metabolites)
would result in a discrepancy between the two values found.
By establishing two calibration curves for the two selected
mass pairs with known amounts ([Do]/[D5] intensity ratios
are plotted against increasing amounts of the [Do] compound),
the necessary linearity of response is ascertained for the
required concentration range; unknown concentrations can
then be interpolated directly. The versatility of chemical ionization and its special advantages in the field of pharmacological
chemistry are amply demonstrated by a multitude of further
reports on the quantitative determination of drugs such as
methadone (and its metabolite^)['^', phenformin" Ool, morphinerga1,quinidine and lidocaine['"], tolbutamide (and its
several barbiturates and hydantoinst' 03], and
also of endogenous substances such as purine and pyrimidine
bases and nucleosides (as permethylated derivatives)" 04].
5.2. Analysis of Inorganic and Organometallic Compounds
The advantages of chemical ionization outlined in Section
5.1, and the complementary nature of chemical ionization
and electron-impact ionization naturally apply also to inorganic and organometallic substrates; however, little information is so far available on this topic['05sl 06]. Borohydrides,
as an example, have been investigated in some detail; in
CI(CH4)mass spectrometry they always yield intense [MH]+
ions when terminal BH groups are present; in contrast [M H] fragments result from compounds containing terminal
BH2 groups['071.Among the organometallic compounds there
is apparently a more or less pronounced tendency to undergo
redox reactions in accordance with eq. (I), depending on the
nature of the metal, the ligand, and the reagent gas['o81.Such
charge-exchange reactivity has been utilized to detect, for
example, molecular ions from thermally very labile arsoIane~['~~].
p] Aspects of quantitative analysis by
means of mass spectrometry have
recently been reviewed by W D. Lehmcmn and H.-R. Schulten, Angew. Chern.
90, 233 (1978);Angew. Chem. Int. Ed. Engl. 1 7 , 221 (1978).
Angew. Chem. Int. Ed. Engl. 17, 424-439 ( 1 978)
[*] [MI" ions often occur with high intensity in CI(CH4)spectra of aryl-substituted compounds or arenes (low values of ionization potential or recombination energy) as a result of charge-exchange processes.
This article and the investigations of chemical ionization carried out at the Technical University of Berlin were kindly supported by the following Institutions: Technische Universitat
Berlin (Research Area “Massenspektrometrie” FPS 511 ), Fonds
der Chemischen Industrie, and Gesellschaft von Freunden der
Technischen Universitat Berlin. We are also indebted to CIBAGEIGY AG, Base1,for similar support.
Received: June 24, 1977 [A 214 IE]
German version: Angew. Chem. 90,449 (1978)
Translated by Express Translation Service, London
a) J . Seibl, Experientia 25, 1009 (1969); b) A . L . Burlingame, B. J .
Kimble, P. J . Derrick, Anal. Chem. 48, 368R (1976).
K. Levsen, H . Schwarr, Angew. Chem. 88, 589 (1976); Angew. Chem.
Int. Ed. Engl. 15, 509 (1976); references cited therein.
Extensive discussion of the advantages and disadvantages of individual
methods for structure determination of ions in the gas phase, see
J . H . Beynon, R. G. Cooks, Adv. Mass Spectrom. 6, 835 (1974).
a) H . D. Beckey, H. R. Schulten, Angew. Chem. 87, 425 (1975); Angew.
Chem. Int. Ed. Engl. 14, 403 (1975); b) F. W Rollgen, H. R. Schulten,
Z. Naturforsch. 30a, 1685 (1975); c) U . Giessmann, F. W Rdllgen,
Org. Mass Spectrom. 11, 1094 (1976); d) H . J . Veith, Angew. Chem.
88, 762 (1976); Angew. Chem. Int. Ed. Engl. 15, 696 (1976).
D. F. Torgerson, R. P. Skowronski, R. D. Macfarlane, Biochem. Biophys.
Res. Commun. 60, 615 (1974); b) R. D. Macfarlane, D . F. Torgerson,
Science 1 91, 920 (1 976).
M . S. B. Munson, F . H. Field, J. Am. Chem. SOC.88, 2621 (1966).
a) F. H. Field, Acc. Chem. Res. 1, 42 (1968); b) F. H. Field, Adv.
Mass Spectrom. 4, 645 (2968); c) 7: Axenrod, Chim. Ind. (Milano)
52, 550 (1970); d) H. D. Beckey, F. J . Comes in A . L. Burlingame:
Topics in Mass Spectrometry. Wiley-Interscience, New York 1970,
p. 74; c) 8. Munson. Anal. Chem. 43, 28A (1971); f) H. M. Fales
in G. W A. Milne: Mass Spectrometry: Techniques and Applications.
Wiley-Interscience, New York 1971, p. 190; g) R. L . Foltr, Lloydia
35, 344 (1972); h) E. J . Bonelli, M . S . Story, Finnigan Spectra 2,
1 (1972); i) G. P. Arsenault in G. R . Waller: Biochemical Applications
of Mass Spectrometry. Wiley-Interscience, New York 1972, p. 817;
1) F. H. Field in A. Maccoll: MTP International Review of Sciences.
Butterworth, London 1972, Vol. 5, p. 133; k) G. W A. Milne, M .
J . Lacey, Crit. Rev. Anal. Chem. 4, 45 (1974); I) N. F. Volkov, G.
I/: Karachevtsev, A. Z. Marutkin, A. I. Kholodoti, Instrum. Exp. Techn.
(USSR) 17, 631 (1974); m) A. L . Burlingame, R. E. Cox, P. J. Derrick,
Anal. Chem. 46, 2 4 8 1 (1974); n) J . M . Wilson in R . A. W Johnstone:
Mass Spectrometry. Specialist Periodical Reports. Chemical Society,
London 1975, Vol. 3, p. 86; 0) D. F. Hunt, Finnigan Spectra 6, 1
(1976); p) W D. Lehmann, H. R. Schulten, Chem. unserer Zeit 10,
163 (1976); q) B. Munson, Anal. Chem. 49, 772A (1977).
H . F . Calcote, Combust. Flame I , 385 (1957); b) V: Cermrik, J. Chem.
Phys. 44, 3781 (1966).
Review of this subject, see P. Ausloos: Interaction between Ions and
Molecules. Plenum Press, New York 1975.
a) G. A. Olah: Carbocations and Electrophilic Reactions. Wiley-Interscience, New York 1974; h) M . D. Sefcik, J . M . S. Henis, P. P. Gaspor, J.
Cbem. Phys. 61,4321 (1974); c) K. Hiraoka, P. Kebarle, J . Am. Chem.
SOC. 98, 6119 (1976).
a) D . F . Hunt, G. C. Stafford, Jr., F. W Crow, 7: M . Harvey, J .
W Russell, Anal. Chem. 48, 2098 (1976); b) R. C . Dougherty, J . Dalton,
F . J . Biros, Org. Mass Spectrom. 6, 1171 (1972); c) R . C. Dougherty,
J . D. Roberts, F. J . Biros, Anal. Chem. 47, 54 (1975).
a) J . L . Franklin, J . G. Dillard, H. M . Rosenstock, J . 7: Herron, K .
Draxl, Nat. Stand. Ref. Data Ser., Nat. Bur. Stand. 1969; b) R . Yamdagni,
P. Kebarle, J . Am. Chem. SOC.98, 1320 (1976); c) A. G. Harrison,
P.-H. Lin, C. W Tsang, Int. J . Mass Spectrom. Ion Phys. 19, 23 (1976);
d) P. Kebarle, R. Yamdagni, K . Hiraoka, 7: B. McMahon, ibid. 19,
71 (1976).
I. Dzidic, J . A . McCloskey, Org. Mass Spectrom. 6, 939 (1972).
a) F. H . Field, P. Hamlet, W F. Libby, J . Am. Chem. SOC.89, 6035
(1967); b) N . Einolf, B. Munson, Int. J. Mass Spectrom. Ion Phys.
9, 141 (1972).
a) D. F. Hunt, J . F. Ryan I I I , Anal. Chem. 44, 1306 (1972); b) G.
P. Arsenault, J. Am. Chem. SOC.94, 8241 (1972); c) F. Hunt, Adv.
Mass Spectrom. 6, 517 (1974); d) B. L. Jelus, B. Munson, C. Fenselau,
Biomed. Mass Spectrom. I , 96 (1974); e) J . L. Franklin, J. G. Dillard,
H. M . Rosenstock, J . 7: Herron, K . Draxl, Nat. Stand. Ref. Data
Ser., Nat. Bur. Stand. 1969.
a) G. Remberg, E. Remberg, M . Spiteller-Friedmunn, G. Spiteller, Org.
Mass Spectrom. I , 87 (1968); b) M . Spiteller-Friedmann, G. Spiteller,
Justus Liebigs Ann. Chem. 712,179 (1968); c) R . C. Dougherty, Fortschr.
Chem. Forsch. 45, 93 (1974).
[17] F. H . Field, J . Am. Chem. SOC.91, 2827 (1969).
[18] D. P. Weeks, F. H. Field, J . Am. Chem. SOC.92, 1600 (1970).
[19] D. P. Weeks, G . W Zuorick, J . Am. Chem. SOC.91, 477 (1969).
[20] F . H . Field, D. P. Weeks, J . Am. Chem. Soc. 92, 6521 (1970).
1213 a) F. H . Field, J. Am. Chem. SOC.91, 6334 (1969); b) S. Vredenberg,
L. Wojcik, J . H . Futrell, J . Phys. Chem. 75, 590 (1971); c) W A.
Laurie, F. H. Field, ibid. 76, 3917 (1972); d) W A. Laurie, F . H .
Field, J. Am. Chem. SOC.94, 2913, 3359 (1972).
[22] A. G. Harrison, C . W Tsang, Can. J. Chem. 54, 2029 (1976).
[23] a) R. C. Dougherty, J . Dalton, J. D. Roberts, Org. Mass Spectrom.
8, 77 (1974); b) R. C. Dougherty, J . D. Roberts, ibid. 8 , 81 (1974);
c) R. C. Dougherty, ibid. 8, 85 (1974); d) W N . Olmstead, J . I . Bruumann, J. Am. Chem. SOC.99,4219 (1977).
1241 Further discussion of this question, cf. E. M . Arnett, Acc. Chem. Res.
6, 404 (1973).
D. F. Hunt, J . F . Ryan, J . Chem. SOC.Chem. Commun. 1972, 620.
a) ‘IJ . Odiorne, D. J . Harvey, P. Vouros, J . Phys. Chem. 76, 3217
(1972); b) 7: J . Odiorne, D . J . Hartiey, P. Vouros, J . Org. Chem. 38,
4274 (1973); c) J . R. Krause, P. Potringer, Int. J . Mass Spectrom.
Ion Phys. 18, 303 (1975).
R. V: Hodges, J . L. Beauchamp, Anal. Chem. 48,825 (1976).
See, e.g., a) D . P. Beggs, F . H. Field, J . Am. Chem. SOC.93, 1567,
1576 (1971); b) L. W Sieck, R. Gordon, Jr., Int. J . Mass Spectrom.
Ion Phys. 19, 269 (1 976).
a) A. G. Harrison, P.-H. Lin, Can. J . Chem. 53, 1314 (1975); b) H. W
Leung, A . G. Harrison, Can. J . Chem. 54, 3439 (1976); c) M . Speranza,
M . D. Sefcik, J . M . S. Henis, P. 7: Gaspar, J. Am. Chem. Soc. 99,
5583 (1977).
a) D. H. Aue, H. M . Webb, M . 7: Bowers, J. Am. Chem. SOC. 95,
2699 (1973); b) R. Yamdagni, P. Kebarle, ibid. 95, 3504 (1973).
7: H. Morton, J . L. Beauchamp, J . Am. Chem. Soc. 94, 3671 (1972).
P. A. Kollmann, L . C . Allen, J. Am. Chem. SOC.92, 6101 (1970).
I. Dzidic, J. A. McCloskey, J . Am. Chem. SOC.93, 4955 (1971).
H. Schwarz, U . Rapp, B. Windel, Org. Mass Spectrom. 11, 219 (1976).
a) I. Jardine, C . Fenselau, J. Am. Chem. SOC.98, 5086 (1976); b) A . G.
Harrison, F. I . Onuska, Org. Mass Spectrom. 13, 35 (1978); c) J .
Respondek, H. Schwarz, F. Van Gaever, C. C. Van de Snnde, ibid., in press.
B. L. Jelus, R. K . Murray, Jr., B. Munson, J. Am. Chem. SOC.97,
2362 (1975).
H . Ziffer, H . M . Fales, G. W A . Milne, F . H. Field, J . Am. Chem.
Soc. 92, 1597 (1970).
H. M . Fales, H . A . Lloyd, G. W A. Milne, J . Am. Chem. SOC.92,
1590 (1970).
a) R. P. Clow, J . H. Futrell, J . Am. Chem. SOC.94, 3748 (1972);
b) R. Houriet, G. Parisod, T Gaumann, ibid. 99, 3599 (1977); c) D.
F. Hunt, C. N . McEweo, Org. Mass Spectrom. 7, 441 (1973); d) ‘I
A. Lehmann, M . M . Bursey: Ion Cyclotron Resonance Spectrometry.
Wiley-Interscience, New York 1976; e) J . L. Beauchamp, Annu. Rev.
Phys. Chem. 22, 527 (1971).
W J . Richter, W Blum, unpublished work.
a) J. A. McCloskey, J . H. Futrell, ‘I A. Elwood, K . H . Schramm,
R . P. Panzica, L. B. Townsend, J . Am. Chem. SOC.95, 5762 (1973);
b) M . S. Wilson, J. A . McCloskey, ibid. 97, 3436 (1975).
0. S. Chizhov, !l I . Kadentsev, A. A . Solov’you, P. F . Levonowich,
R. C. Dougherty, J . Org. Chem. 41,3425 (1976); references cited therein.
a) K . Biemann: Mass Spectrometry. McGraw Hill, New York 1962;
b) K. Biemann, W Vetter, Biochem. Biophys. Res. Commun. 2, 93
a) H. J . Forster, J. A . Kelley, H . Nau, K . Biemann in J . Meienhofer:
Chemistry and Biology of Peptides. Ann Arbor Science Puhl., Ann
Arbor 1972, p. 679; b) J . A. Kelley, H. NRU,H . J . Forster, K . Biemann,
Biomed. Mass Spectrom. 2, 313 (1975); c) H . Nau, H. J . Forster,
J . A . Kelley, K. Biemann, Biomed. Mass Spectrom. 2, 326 (1975);
d) H. Nau, Angew. Chem. 88, 74 (1976); Angew. Chem. Int. Ed. Engl.
15, 75 (1976).
A. A. Kiryushkin, H. M . Fales, T Axenrod, E. J . Gilbert, G . W A.
Milne, Org. Mass Spectrom. 5, 19 (1971).
K. D. Haegele, D. M . Desiderio, Jr., J . Org. Chem. 39, 1078 (1974).
R. J. Weinkam, J . Gal, Org. Mass Spectrom. 11, 197 (1976).
G. W A. Milne, T Axenrod, H . M . Fales, J . Am. Chem. SOC. 92,
5170 (1970).
T A. Whitney, L. P. Kleman, F . H. Field, Anal. Chem. 43, 1048
H. Bosshardt, M . Hesse, Angew. Chem. 86, 256 (1974); Angew. Chem.
Int. Ed. Engl. 13, 252 (1974).
C . C . Van de Sande, M . Vanhooren, F. Van Gaever, Org. Mass Spectrom.
11, 1206 (1976).
a) C. C. Van de Sande, F. Van Gaever, 3. Motistrey, Z. Naturforsch.
B32, 573 (1977); b) P. Longeuialle, G. W A. Milne, H. M . Fales, J . Am.
Chem. SOC.95,6666 (1973).
J. Winkler, F . W McLafferty, Tetrahedron 30, 2971 (1974).
H . Schwarz, F. Bohlmann, U. Rapp, B. Windel, Org. Mass Spectrom.
11, 652 (1976).
Angew. Chem. Int. Ed. Engl. 17,4244439 (1978)
[5S] E. L. Elielr Stereochemistry of Carbon Compounds. McGraw-Hill,
New York 1962.
[56] G. Hope, G. Hohne, J . Respondek, H . Schwarz, Org. Mass Spectrom.
12, 447 (1977).
[57] R. Ryhage, Anal. Chem. 48, 1829 (1976).
[58] G. P. Arsenault, J . J . Dolhuhn, K . Biemann, Anal. Chem. 43, 1720
(1 971).
[59] J . H. Futrell, L. H. Wojcik, Rev. Sci. Instrum. 42, 244 (1971).
[60] D. F. Hunt, C . N . McEwen, 7: M . Harvey, Anal. Chem. 47, 1730
[61] a) E. C . Horning, D. I . Carroll. I. Dzidic, K . D. Haegele, M . G.
Horning, R. N . Stillwell, J . Chromatogr. 99, 13 (1974); b) 12, 725
(1974); c) D. I. Carroll, 1. Dzidic, R. N . Stillwell, K. D. Haegele, E.
C. Homing, Anal. Chem. 47, 2369 (1975); d) I. Dzidic, D. I . Carroll,
R. N . Stillwell, E. C. Homing, ibid. 48, 1763 (1976).
[62] B. Hoegger, P. Bommer, Int. J . Mass Spectrom. Ion Phys. 13, 35
[63] a) G. P. Arsenault, J . J . Dolhuhn, K . Biemann, Chem. Commun. 1970,
1542: h) G. P. Arsenault, J . J . Dolhuhn, 18th Annual Conference on
Mass Spectrometry and Allied Topics. San Francisco 1970, p. B 423.
[64] D. M . Schoengold, B. Munson, Anal. Chem. 42, 1811 (1970).
[65] W Blum, W J . Richter, Helv. Chim. Acta 57, 1744 (1974).
[66] I. Jardine, C. Fenselau, 22nd Annual Conference on Mass Spectrometry
and Allied Topics, Philadelphia 1974.
1671 W Blum, W J . Richter, Finnigan Spectra 4, 1 (1974).
[68] W Blum, W J . Richter, Finnigan Spectra 5, 1 (1975).
[69] a) W J . Richter, W Blum. J . G. Liehr, 23rd Annual Conference on
Mass Spectrometry and Allied Topics. Houston 1975, p. 591 ; b) ci!
Blum, E. SchlumpJ J . G. Liehr, W J . Richter, Tetrahedron Lett. 1976,
[70] W Bluni, W J . Richter, Tetrahedron Lett. 1973, 835.
[71] W Hum, W J . Richter, J . Chromatogr. 132, 249 (1977).
[72] a) C . G. Hammar, B. Holnistedt, R. Ryhage, Anal. Biochem. 25, 532
(1968); b) C. C. Sweelej. W H. Elliot, I . Fries, R. Ryhage, Anal.
Chem. 38, 1549 (1966).
[73] a) J . 7: Watson, F . C . Falkner, B. J . Sweetman, Biomed. Mass Spectrom.
I , 156 (1974); b) W L. Budde, ibid. I , 427 (1974); c) C. Fenselau,
B. Millard, ibid. 2, 226 (1975).
[74] IUPAC Information Bulletin No. 51, September 1967.
[75) J . Shabanowitz, P. Brynes, A. Maelicke, D. K Bowen, F. H. Field,
Biomed. Mass Spectrom. 2, 164 (1975).
[76] G. P. Arsenault, J . R. Althaus, P. V Divekar, Chem. Commun. 1969,
[77] C . A. Grob in: Theoretical Organic Chemistry. Papers Presented to
the Kekule Symposium, Butterworth, London 1959, p. 114.
[78] W M. Bryant 111, A. L. Burlingame, H. 0. House, C. G. Pitt, B.
Tefertiller, J. Org. Chem. 31, 3120 (1966).
[79] W J . Richter, W Vetter, Org. Mass Spectrom. 2, 781 (1969).
[80] J . W Faigle, S. Brechbiihler, K . F. Feldmann, W J . Richter in W
Birkmayer :Epileptic Seizures-Behaviour-Pain.
Hans Huber Verlag,
Bern 1976, p. 127.
[81] D. C. K . Lin, A. F. Fenriman, Jr., R. L. Foltz, R. D. Forney, 1. Sunshine,
Biomed. Mass Spectrom. 2, 206 (1975).
Angew. Chem. Int. Ed. Engl. 17, 4 2 4 4 3 9 (1978)
[82] P. Heimbach, B. Hugelin, E. F. Nabbefeld, D. Reinehr, A. Roloff, E.
Troxler, Angew. Chem. 89, 261 (1977): Angew. Chem. Int. Ed. Engl.
16, 253 (1977).
[83] W Blum, E. Schlumpf, W J. Richter, unpublished work.
[84] D. F . Hunt, C . N . McEwen, R. E. Upham, Anal. Chem. 44, 1292
[85] D. F. Hunt, C. N. McEwen, R. E. Upham, Tetrahedron Lett. 1971,
[86] J . R. Althaus, K. Biemann, J . Biller, P. F. Donaghue, D. A. Evans,
H. J . Forster, H. S. Herrz, C. E. Hignire, R. C. Murphy, G. Preti,
l! Reinhold, Experientia 26, 714 (1970).
[87] N . C . Law, K Aandahl, H. M . Fales, G. W. A. Milne, Clin. Chim.
Acta 32, 221 (1971).
[88] G . W A . M i l n e , H . M. Fales, T Axenrod, Anal. Chem. 43, 1815 (1971).
[89] J.-M. Chao, R. Saferstein, J . Manura, Anal. Chem. 46, 296 (1974).
[90] R. L. Foltz: Final Report on the Rapid Identification of Drugs from
Mass Spectra. BATELLE Columbus Laboratories, Columbus, Ohio
1974, p. 10.
[91] J . Enon, Biomed. Mass Spectrom. I , 393 (1974).
[92] B. S . Finkle, R. L. Foltz, D. M . Taylor, J . Chromatogr. Sci. 12, 304
( 1974).
[93] E. C . Homing, M . G. Horning, D. I . Carroll, R. N . Stillwell, I . Dzidic,
Life Sci. 13, 1331 (1973).
[94] D. I. Carroll, I . Dzidic, R. N . Stillwell, E. C . Horning, Anal. Chem.
46, 706 (1974); 47, 1956 (1975).
[95] a) E. C. Horning, M. G. Homing, D.I. Carroll, 1. Dzidic, R. N . Stillwell,
Anal. Chem. 45, 936 (1973); b) Adv. Biochem. Psychopharmacol. 7,
15 (1973).
[96] D. F. Hunt, 7: M . Harvey, J . W Russell, J . Chem. SOC.Chem. Commun.
1975, 151.
[97] H. M. Fales, l! Nagai, G. W A. Milne, H. B. Brewer, Jr., 7: J .
Bronzert, J. J . Pisano, Anal. Biochem. 43, 288 (1970).
[98] P. A. Clarke, R. L. Foltz, Clin. Chem. 20, 456 (1974).
[99] D.l!Bowen, A. L. Smit, M. J . Kreel, 7th International Mass Spectrometry Conference. Florence 1976, p. 280.
[too] S. 8. Martin, J . H . Karam, P. H . Forsham, J . B. Knight, Biomed.
Mass Spectrom. I, 320 (1974).
[101] W G. Garland, ci! F. Traeger, S. D. Nelson, Biomed. Mass Spectrom.
1, 124 (1974).
[lo21 S. B. Martin, J . B. Knight, Biomed. Mass Spectrom. I , 323 (1974).
[I031 M . G. Homing, W G. Stillwell, J . Nowfin, K. Lertratanangkoon. D.
Carroll, 1. Dzidic, R. N . Stillwell, E. C . Horning, J . Chromatogr. 99,
413 (1974).
[I041 C . Pantarotto, A. Martini, G. Belvedere, A. Bossi, M . G. Donelfi, A.
Frigerio, J . Chromatogr. 99, 519 (1974).
[lo51 7: R. Spalding in R. A. W Johnstone: Mass Spectrometry. Specialist
Periodical Reports. Chemical Society, London 1975, Vol. 3, p. 143.
[106] J . Miiller, Angew. Chem. 84, 725 (1972): Angew. Chem. Int. Ed. Engl.
11, 653 (1972).
[lo71 a) J . J. Solomon, R. F. Porter, J. Am. Chem. SOC. 94, 1443 (1972);
b) R. C . Pierce, R. F. Porter, ibid. 95, 3849 (1973).
[108] D. F . Hunt, J . W Russell, R. L. Torian, J . Organomet. Chem. 43,
175 ( 1 972).
[lo91 F . Kober, 2. Chem. 16, 159 (1976).
Без категории
Размер файла
1 745 Кб
analytical, mass, procedur, chemical, rapidly, importance, increasing, ionization, spectrometry
Пожаловаться на содержимое документа