Features of the chromatography-mass spectrometric identification of condensation products of the carbonyl compounds

Article abstract of DOI:10.1134/S1070363211090143

By the example of the condensation products of acetone with the simplest aromatic carbonyl compounds it was shown that the joint interpretation of mass spectrometric and chromatographic data in conjunction with a priori assumptions about the nature of the chemical structure allows the detection of the components of reaction mixtures that were not previously characterized by standard mass spectra or by retention indices on standard stationary phases. A necessary condition for solving such problems is creating a hypothesis about the sample composition that is possible for the expected products of the known organic reactions.

Full text of DOI:10.1134/S1070363211090143

ISSN 1070-3632, Russian Journal of General Chemistry, 2011, Vol. 81, No. 9, pp. 1818–1828. © Pleiades Publishing, Ltd., 2011.
Original Russian Text © I.G. Zenkevich, A.I. Ukolov, 2011, published in Zhurnal Obshchei Khimii, 2011, Vol. 81, No. 9, pp. 1479–1489.
Features of the Chromatography-Mass Spectrometric
Identification of Condensation Products
of the Carbonyl Compounds
I. G. Zenkevich and A. I. Ukolov
St. Petersburg State University, Universitetskii pr. 26, St. Petersburg, 198594 Russia
e-mail: izenkevich@mail15.com
Received April 18, 2011
Abstract—By the example of the condensation products of acetone with the simplest aromatic carbonyl
compounds it was shown that the joint interpretation of mass spectrometric and chromatographic data in
conjunction with a priori assumptions about the nature of the chemical structure allows the detection of the
components of reaction mixtures that were not previously characterized by standard mass spectra or by
retention indices on standard stationary phases. A necessary condition for solving such problems is creating a
hypothesis about the sample composition that is possible for the expected products of the known organic
reactions.
DOI: 10.1134/S1070363211090143
The carbonyl fragment is among the most common
functional groups in the composition of organic
compounds. The C=O group reactivity is sufficiently
high, especially with respect to nucleophilic reagents.
[1]. For this reason, for example, the derivatization of
carbonyl compounds for gas chromatographic analysis
[1] consists not in the transformation of analytes in
more volatile derivatives [2]. but in their conversion
into less reactive products that are more stable in
mixtures with other substances. Both in alkaline and in
acidic media carbonyl compounds readily enter in the
intermolecular condensation reactions [2]. For ex-
ample, the base-catalyzed aldol condensation followed
by dehydration of the intermediate products leads to
α,β-unsaturated carbonyl compounds:
formation (especially in the case of aliphatic aldehydes
and ketones) of a large number of the products of more
“profound”
(oligomeric)
condensation,
whose
molecules, moreover, may be capable of cyclization,
migration of C=C double bonds, and (Z,E)-isomerism.
For compounds with different R and R1≠R' the number
of potential products increases even more due to the
lack of regioselectivity of the reaction. The main
difficulty consists in the lack of the reference data for
the most of the condensation products necessary for
the interpretation of the results of gas chromatography-
mass spectrometry analysis, namely, the usual electron
impact mass spectra and gas chromatography retention
indices (RI) on the standard (especially non-polar)
stationary phases.
R2
R2
R
R1
B
To illustrate the current state of information
concerning the condensation products of even the
simplest carbonyl compounds, it is interesting to
compare the data for the acetone [reaction (1)] and 2-
butanone [reaction (2)] oligomers, corresponding to
the following sequence of molecular formula and mass
numbers (in parentheses):
R
O
O +
O
R3CH2
R1 OH R3
R2
R
O
. . . . .
_H2O
R1
R3
Despite the fact that such products may be present
and found in many reaction mixtures, the possibility of
their gas chromatography-mass spectrometric identi-
fication still is very limited. This is due to the
C3H6O (58) → C6H10O (98) → C9H14O (138)
→ C12H18O (178) → …,
(1)
C4H8O (72) → C8H14O (126) → C12H20O (180) → … . (2)
1818

FEATURES OF THE CHROMATOGRAPHY-MASS SPECTROMETRIC IDENTIFICATION
1819
One of the most detailed database of mass
spectrometric and chromatographic data NIST/EPA/
NIH[3] [3] includes both the mass spectrum and no
less than 12 references to original sources of the RI
values for the 4-methyl-3-penten-2-one (mesityl
oxide), the primary condensation product of acetone.
Inasmuch as the database [3] is in the process of
forming, the number of sources of retention indices for
each compound is continuously growing. As the main
products formed in the next stage of condensation can
be assumed to be 2,6-dimethyl-2,5-heptadien-4-one
(phorone, I), 4,6-dimethyl-3,5-heptadien-2-one (II),
and 3,5,5-trimethyl-2-cyclohexen-1-one (α-isophorone,
III). Of these isomers, the mass spectra are known
only for compounds I and III, and the RI, according to
the publications [4–16], only for α-isophorone III (the
values vary in the range 1074–1099 with an average
value of 1089±7). The reason for such strong
discrepancy in the number of data for different isomers
is that α-isophorone is one of the components forming
the odor of many foods and, therefore, was described
in detail. For the possible compounds of the same
composition, C12H18O (M = 178) (and also for more
complex ones) the currently available information is
limited to the mass spectrum of a single isomer,
1-(3,5,5-trimethyl-2-cyclohexenylidene)acetone (IV)
[3]. It is noteworthy that for all compounds in the
database [3] the RI estimates using the classical
incremental additive scheme are available [17] whose
accuracy is unfortunately low.
purpose the information from independent sources fits
best, but when such information is inavailable (as
commonly occurs), it is necessary to use estimates
based on the data for the simpler and therefore more
reliably characterized structural analogs.
As another illustrative example of the mentioned
complexity of the representation of chromatographic
data in modern databases (by the example of [3]), we
may consider a relatively simple compound such as 3-
penten-2-one. Only one value of RI = 712 is known,
which is uniquely attributed to (E)-isomer of the
ketone [18]. For the corresponding (Z)-isomer are
known two RI values not coinciding with each other
(711 [19] and 652 [20]). In addition, the RI values for
this ketone falling to the range 697–755 [without
reffering to (E)-and (Z)-isomers] are given in other
15 publications. Such data are clearly insufficient, not
only for an unambiguous conclusion about the value of
RI of (Z)-3-penten-2-one, but even about the order of
the isomers elution.
Thus, the essential features of chromatographic-
mass spectrometric identification of the condensation
products of carbonyl compounds are as follows:
(1) Informativity of the mass spectra of this class of
compounds is limited, and for many isomers the
spectra are virtually identical, but for most products
they are unknown;
(2) For the identification at the limited infor-
mativity of mass spectrometry data it is necessary to
use chromatographic retention indices, but for most
compounds of this group they are also unknown;
Even more spectacular is the poor state of
information for the condensation products of 2-butanone
[reaction (2)]. The expected primary products of its
condensation (C8H14O, M = 126) are 5-methyl-4-
heptyl-3-one (V) and 3,4-dimethyl-3-hexen-2-one
(VI); each one exists in the form of (Z)- and (E)-
isomers that has been characterized by the mass
spectra, which, naturally, for each pair of isomers
should be virtually identical. The database [3] includes
an alternative mass spectrum of compound V which
does not coincide with the first one (in one of them the
maximum peak is at m/z 55, while in another its
intensity is negligible). For the same compound V the
value of RI (1007) is know [10] attributed to the
isomer with an unknown geometry, but even in this
case this value appears to be erroneous (too high).
Thus, when solving a real problem of identification it
is necessary to consider also the problem of the
possible unreliability of reference data and the
necessity of their checking and refinement. For this
(3) In the absence of reference values of the indices
their estimates by various methods should be used,
including simple additive schemes;
(4) The necessity to obtain such estimates involves
the formation of a priori assumptions about the
structure of the condensation products.
The present communication discusses the features
of chromatography-mass spectrometric identification
of previously characterized condensation products of
acetone (as one of the simplest representatives of
aliphatic carbonyl compounds and one of the most
common organic solvents) and the simplest aromatic
carbonyl compounds, based on joint interpretation of
their mass spectra and chromatographic retention
parameters on standard nonpolar phases.
Among the methods of identification used in
modern practice of gas chromatography-mass spectro-
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

1820
ZENKEVICH, UKOLOV
metry the most popular is the comparison of mass
spectra of the analytes with reference data. The basis
for the effectiveness of this approach is the availability
of sufficiently detailed database of mass spectra. So,
the latest version of one of the most well-known
database of the National Institute of Standards and
Technology (NIST, USA, 2008) contains 220460
standard mass spectra of 192108 compounds [3]. Such
a library search does not exclude the methods of
interpretation based on known patterns of fragmenta-
tion of organic compounds, but they are significantly
less used. Since 2005, the base [3] was supplemented
by chromatographic retention indices (RI) on standard
nonpolar and polar phases. The version of 2008
included 293247 RI values for 44008 compounds,
whose use in conjunction with the mass spectra should
enhance the reliability of the identification results.
estimates of RI, since it is enough to predict the
relative order of chromatographic elution. To solve
these problems, several algorithms are known, based
for instance on an evaluation of the intramolecular
vibrational and rotational energies by the method of
the molecular dynamics [22]. Another method involves
obtaining estimates of normal boiling points (Tb) using
the ACD software, since the order of elution of
isomers in nonpolar phases unambiguously cor-
responds to an increase in Tb [32]. In a general case the
estimation of RI should be carried out using, for
instance, an additive scheme.
The most convenient for the evaluation of
chromatographic RI values turned to be not the
traditional methods of calculation involving the use of
a set of the retention indices increments (ΔRI) or their
modified versions. The operations of computing the
ΔRI values and their subsequent application can be
combined, which is equivalent to assembling the target
structures from simpler molecules by their addition
and subtraction. This method assumes selection of the
maximally close structural analog of target compounds
and combining them with the molecules containing the
missing structural fragments and subtracting the
superposition of overlapping elements.
However, despite the existence of such information,
the solution for a sufficiently large number of real
analytical problems concerns the compounds that have
not been characterized by mass spectra or chromato-
graphic retention parameters. When a direct com-
parison of experimental and reference data is
impossible, such problem is more in line with the
structure elucidation and can be very complicated. An
important group is the characterization of the products
of known chemical reactions when the solution is
simplified due to the information not only on the
chemical nature of the initial substrates and reagents,
but also on the mechanism of the process. When
preparative isolation of individual products from
complex reaction mixture is impossible, then the main
way of identifying them is just the use of the
chromatography-mass spectral method.
This algorithm corresponds to the following general
scheme. If for the formation of the target structure
ABCD are got the predecessors ABC and BCD, so that
ABC +BCD −BC ABC + BCD – BC → ABCD, then
RI ≈ (ΑΒΧΔ) RI(ABC) + RI(BCD) – RI(BC).
In other words, similar fragments of the molecular
structures of selected precursors should be cut off to
provide the desired stoichiometry of the operation. The
estimates of standard deviations of calculation results
(sRI) based on the standard deviations of the source
data can be obtained using the well-known relation:
The interpretation of combined mass spectrometric
and chromatographic data for the identification of
previously unknown compounds can be carried out in
various ways, but below we discuss only those proved
to be the most effective for these compounds. The
products of chemical reactions may be rigidly defined
by the nature of the initial substrates and mechanisms
of the processes. In such cases, it is almost possible to
predict their structures. This feature is widely used in
the practice of derivatization of target analytes, where
the preparative isolation of the reaction products is
impossible and the reactions themselves are considered
as a way to prove the structure of the products formed
[1, 21].
sRI(ABCD) ≈ [s2RI(ABC) + s2RI(BCD) + s2RI(BC)]1/2.
(3)
The main advantages of this method of the RI
estimation, besides eliminating the necessity to pre-
calculate the increments, is its structural clarity and the
possibility of varying the chosen analogs, ways of the
assembly of the target structures, and the direct use of
the reference RI values. This approach was first
proposed and used to evaluate the RI values of 839
congeners of polychlorinated hydroxybiphenyls [42],
expanded over 211 structural isomers of nonylphenol
[52] and the products of free radical chlorination of
cyclohexane [62], which in the case of nCl ≥ 2 included
The identification of a small number of products
(2–3) in the simplest cases may not require precise
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

FEATURES OF THE CHROMATOGRAPHY-MASS SPECTROMETRIC IDENTIFICATION
1821
not only the structural isomers, but also diastereomers.
At the optimal choice the scheme of the formation of
the target structure the relatively high accuracy of the
estimates is confirmed in all cases. The maximum
accuracy of the RI estimates is achieved when the
precursor molecules comprise all the features of the
structures of the target compounds that affect their gas
chromatography retention parameters.
Example 1. Establishing the structure of two
products of acetone and acetophenone condensation
(1:1), with the molecular formula C3H6O + C8H8O –
H2O = C11H12O, and the mass 58 + 120 – 18 = 160, for
which the obtained RI values are 1362 and 1396. Other
possible condensation products (acetone + acetone and
acetophenone + acetophenone) can be excluded from
consideration by their molecular weights that are not
consistent with the obtained number, and by the
chromatographic parameters. For example, the RI of
1,3-diphenyl-2-butene-1-one (C16H14O, dipnone) is
much higher (1970±6), a component with this RI value
is not found. The number of theoretically expected
products of different structure in this case is equal to
two: (Z,E)-isomers of 4-phenyl-3-penten-2-one VII
and 3-methyl-1-phenyl-2-butene-1-one VIII, that is,
corresponds to the number of detected components.
Table 1 shows the experimental spectra of components
of the reaction mixture, and also indicates the presence
or absence of the mass spectra of the expected
products in the database [3]. The probable cause of the
discrepancy of mass spectrum of compound VIII [3]
with the spectra of both components could be that the
former was recorded for the mixture of the products
isolated preparatively from the reaction mixture. Thus,
the standard use of reference mass spectra cannot only
lead to the failure at the identification of the analytes,
but also can mislead the researcher. Attempts to
interpret the mass spectra on the basis of general
patterns of fragmentation [272] for conjugate
structures are complicated by the possibility of skeletal
rearrangements of the molecular ions. For example, the
presence in the spectrum of the signal (m/z)100 = 105
can be explained both by the presence in the molecule
of benzoyl fragment VIII, and the possibility of 1,3-
migration of phenyl group VII, and therefore cannot
be interpreted unambiguously. Most informative signal
in the spectrum of compound VII is the peak at m/z 83
[M – C6H5] corresponding to the acyl fragment
(CH3)2C=CHCO, but for a choice of one structure
among two alternative structures it is not sufficient to
consider a signal of only one mass spectrum.
The principal stages of the interpretation of the
results of gas chromatography-mass spectrometry
analysis of previously uncharacterized products of
known reactions can be formalized as follows:
(1)) The estimation of the potential number and
nature of the possible products of the considered
reactions or mixtures of natural components (the
formulation of the hypothesis about the composition of
the analyzed samples);
(2) Checking in the databases of mass-spectro-
metric and chromatographic reference data for the
presence of compounds with the intended chemical
nature. In the event of insufficient information go to
step 4;
(3) The identification of the components of the
samples using known algorithms of the mass spec-
trometric library search. In the case of incertainty of
the answers go to step 5;
(4) The identification of the unidentified com-
ponents and isomers with indistinguishable mass
spectra;
(5) Depending on the complexity of the samples,
the calculation of retention indices of the assumed
products or estimating the order of elution (for
isomers);
(6) Joint interpretation of mass spectrometric and
chromatographic data for compounds at the lack of the
reference values of considered analytical parameters.
The sequence of these stages can be varied, but the
final step anyway is comprising the information in a
single logically consistent answer. It is not surprising
finally that the most difficult and time-consuming step
in the chromatography-mass spectrometry analysis are
not instrumental operations, but the interpretation of
the results.
Formally, for the identification of only two known
products of a known reaction it is enough to predict the
order of their chromatographic elution. However, in
this example, the use of molecular dynamics methods
[22] is undesirable, as they insatisfactorily describe
molecules with isomeric conjugation systems. In such
cases it is more correct to obtain estimates of normal
boiling points (Tb) using the ACD software [23]. For
To detail these general statesments it is reasonable
to illustrate specific examples of identification of the
condensation products of acetone and a few simple
aromatic carbonyl compounds, arranged by the in-
creasing complexity of the resulting reaction mixture.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

1822
ZENKEVICH, UKOLOV
Table 1. Mass spectra and retention indices of the condensation products of acetone and acetophenone (to Example 1)
Existence
in the
Mass spectrum [m/z ≥ 50 (Irel ≥ 3%)]
RIexp
RIcalc
database
[3]
3-Methyl-1-phenyl-2-butene-1-one (VII, 71%)a
161(7), 160(64) [М]+, 159(100) [М – Н], 146(9), 145(97), 144(17), 142(10), 141(21), 132(5), 131(21), 1362
130(5), 128(9), 127(17), 120(5), 118(4), 117(29), 116(7), 115(29), 106(4), 105(46), 103(4), 91(11),
89(5), 84(5), 83(63) [М – С6Н5], 82(5), 79(4), 78(3), 77(77), 76(7), 75(6), 74(6), 70(3), 67(4), 65(9),
63(8), 61(3), 57(5), 55(63), 54(9), 53(21), 52(6), 51(58), 50(22)
–
1352±9
4-Phenyl-3-penten-2-one (VIII, 29%)
b
160(4) [М]+, 159(4), 131(3), 120(20), 106(9), 105(100) [C6H5CO], 91(5), 78(9), 77(37), 65(3), 63(3), 1396
+
1378±11(Е)
1393±14(Z)
59(12), 58(4), 56(5), 53(3), 51(15), 50(7)
a
Here and hereinafter (in Tables 2 and 3) after the number of compound is shown in parentheses the estimation of its relative content in
the reaction mixture. b The mass spectrum of 4-phenyl-3-penten-2-one (VIII) in the database [3] [160 (79) [M]+ 159 (100) 145 (36), 117
(95), 116 (20), 115 (71), 91 (25), 55 (15), 43 (41), 39 (12)] differs significantly from that given here and, therefore, cannot be used for
identification (see comments in the text).
isomers VII and VIII we obtain 263.9±10 and
243.2±13°C, respectively, therefore, 3-methyl-1-
phenyl-2-butene-1-one VIII should be eluted first with
the RI = 1362. The ACD software cannot differentiate
(Z)- and (E)-isomers, so the geometry of the second
component, 4-phenyl-3-penten-2-one VII, remains
uncertain in this method of data interpretation.
If the objective is to obtain the absolute rather than
relative retention characteristics, it is necessary to
estimate the RI value using additive schemes. For 3-
methyl-1-phenyl-2-butene-1-one VIII at least two
ways of assembling the target structure of the mole-
cules can be offered characterized by the RI values of
the simpler structural analogues:
O
O
O
O
_
_
+
+
C6H5
C6H5
_
+
_
+
_
+
_
+
(783 11)
(472 12)
(1352 19)
O
(1041 9)
O
=
O
O
_
_
+
+
C6H5
(1351 18)
C6H5
_
+
_
_
+
_
+
(576 14)
+
(783 11)
(1144 5)
=
Averaging the results gives an estimate 1352±9,
the second potential product 4-phenyl-3-penten-2-one
which agrees with the RI of the first isomer (1362). For
VII the similar pattern of assembly is as follows:
O
O
_
+
C6H5
C6H5
_
_
+
_
_
+
_
+
+
(576 14)
(166 7)
(1376 17)
(966 6)
+
=
_
+
C6H5
(1381 15) (E)
C6H5
_
_
_
_
_
+
_
+
+
+
(290 3)
(576 14)
(1094 5) (E)
+
+
=
=
_
_
_
+
+
+
(576 14)
(290 3)
(1393 14) (Z)
1116 (Z)
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

FEATURES OF THE CHROMATOGRAPHY-MASS SPECTROMETRIC IDENTIFICATION
1823
The average value of the estimates RI (1383±9)
corresponds to the experimental value of the second
component (1396). However, based on these additive
estimates it is hardly possible to say that for the second
product the structure of (Z)-isomer is more likely and,
therefore, the question of its stereochemistry in this
case is better to leave unspecified.
The following example relates to a similar reaction,
but the number of isomeric products increases to four.
Example 2. Establishing structures of four isomeric
acetone and propiophenone condensation products
(1:1) with the molecular formula C3H6O + C9H10O –
H2O = C12H14O, and the mass 58 + 134 – 18 = 174 , RI
are 1383, 1414, 1435, and 1452 (Table 2). Two of
them, with (m/z)100 = 173 [M – H] (nos. 2 and 3 in the
order of elution) are the homologs of two natural α,β-
unsaturated carbonyl compounds considered in the
preceeding example, namely, 3,4-dimethyl-1-phenyl-2-
buten-1-one (XII) and (Z,E)-isomers of 4-phenyl-3-
hexen-2-one (XIII). The other two (nos. 1 and 4) are
characterized by the similar abnormal mass spectra,
(m/z)100 = 43 [CH3CO] that differ significantly from
those of the first pair of isomers. The lower intensity
signals of their molecular ions points to shorter
conjugation systems. It may be assumed that these two
anomalous products result from the alternative way of
water cleavage from the ketoalcohol XI formed
intermediately, and they are (E)- and (Z)-isomers of 4-
phenyl-4-hexen-2-one (XIV, XV). The presence in
their molecules of unconjugated CH3CO fragments
results in the predominance in the mass spectra of the
signals (m/z)100 = 43.
The summary of results is presented in Table 1.
From the consideration of even this single example a
simple rule follows regarding the order of chromato-
graphic elution of isomeric condensation products of
alkylaromatic (ArCOR) and aliphatic (RCOR')
carbonyl compounds. The minimum retention param-
eters of the two structural isomers have the products
obtained at the interaction of the carbonyl group
RCOR' and methylene fragment in the ArCOR
molecule. If this pattern is correct, then it should be
observed for the condensation products and other
carbonyl compounds that can be used for its verifica-
tion. For example, acetaldehyde and acetophenone are
precursors of 1-phenyl-2-butene-1-one (IX) and 3-
phenyl-2-butenal (X), with the RI 1192 (experimental
value) and ~1265 (estimate from database [3] by
scheme [17]), respectively, which is consistent with
the proposed rule and allows us to simplify the
interpretation of results in all such cases.
C2H5
C6H5
OH
+
C6H5
O
O
C6H5
O
ХV
(Z)-isomer
ХI
ХIV
(E)-isomer
Currently there are no methods of the theoretical
prediction of the order of chromatographic elution of
π-diastereomers, but the considered additive scheme
allows the estimation of RI of such structures. However,
the seemingly natural simplest version of assembling
the target structure starting from (1-methyl-1-propenyl)-
benzene (E)- and (Z)-isomers and acetone is unaccept-
able because it leads to highly overestimated RI values:
O
CH2COCH3
_
CH4
+
C6H5
C6H5
_
_
_
_
+
+
+
(472 12)
(100)
(1466 13) (E)
(1094 5) (E)
+
=
The elucidation of the reasons of this discrepancy
shows that the presence in the molecules XIV and XV
of the bulky acetonyl group at the double C=C bond is
important because it leads to a violation of its
conjugation with the phenyl fragment. The optimiza-
tion of the geometry of the isomers XIV and XV by
the MM+ method gives an estimate of the dihedral
angles θ[C5–C4–C(Ar)1–C(Ar)2] ~60° (E-isomer) and
71° (Z-isomer), which corresponds to a decrease in the
degree of conjugation of the fragments Ph and C=C
(proportional to cos2 θ) from the maximal value of
unity to 0.25 and 0.11, respectively. Consequently,
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

1824
ZENKEVICH, UKOLOV
Table 2. Mass spectra, retention indices and evaluation of normal boiling points of the condensation products of acetone and
propiophenone (to Example 2)
Existence in
Mass spectrum [m/z ≥ 40 (I ≥ 3%)]
(E)-4-phenyl-4-hexen-2-one (XIV, 17–23%)
RIexp
RIcalc or Тb
database [3]
175(1), 174(12) [M]+, 159(13), 132(4), 131(17), 129(5), 128(4), 117(7), 116(5), 115(8), 91(18),
77(4), 65(3), 53(3), 51(4), 43(100) [CH3CO]
1383
–
1395±13
2,3-Dimethyl-1-phenyl-2-butene-1-one (XII, 8–9%)
175(8), 174(69) [М]+, 173(100), 160(7), 159(57), 158(4), 155(2), 153(3), 146(4), 145(29), 144 1414
(11), 141(10), 132(3), 131(22), 130(4), 129(20), 128(11), 127(5), 117(10), 116(18), 115(24),
105(3), 103(7), 102(6), 92(5), 91(59), 77(14), 76(4), 75(3), 65(8), 64(6), 63(8), 56(3), 53(12),
51(17), 50(6), 43(84), 41(5)
–
–
263.7±13
280.7±10
(Z, E)-4-phenyl-3-hexen-2-one (XIII, 25–29%)
175(7), 174(60) [М]+, 173(100), 160(9), 159(57), 158(8), 150(7), 148(5), 146(9), 145(26), 144 1435
(13), 141(7), 135(5), 134(5), 132(6), 131(35), 129(15), 128(10), 127(7), 121(4), 119(4), 117
(10), 116(19), 115(27), 106(5), 105(26) [C6H5CO], 103(6), 102(11), 100(4), 97(7), 92(6), 91
(61), 89(5), 86(3), 83(5), 79(12), 78(5), 77(31), 63(11), 62(4), 56(5), 55(3), 53(12), 51(23), 50
(12), 43(92), 41(18)
(Z)-4-phenyl-4-hexen-2-one (XV, 39–49%)
175(2), 174(13) [M]+, 159(12), 132(4), 131(15), 129(5), 128(3), 117(7), 116(4), 115(7), 91(16),
1452
–
1417±14
77(20), 53(4), 51(5), 43(100) [CH3CO]
when assessing the RI for the structure XII we have to
choose as a precursor an alkenylarene with the
conjugated Ar and C=C fragments, while the double
bond C4=C5 should be included in the target structure
as a non-conjugated. For example, the following
slightly more complex versions of the additive
evaluation of RI can be implemented to satisfy these
requirements:
O
CH2COCH3
C6H5
_
+
C6H5
_
_
+
_
_
+
+
(773 10)
(400)
(1376 12)
(1003 7)
+
=
CH2COCH3
O
_
_
_
_
+
+
C4H10
(400)
+
C6H5
C6H5
(1003 7)
C3H7
+
_
_
_
_
+
_
+
_
+
+
(768 9)
(474 2)
(1414 14) (E)
(517 7)
+
=
The average RI value obtained is 1395, which is
acceptably consistent with the experimental value
1383. To estimate the RI of the (Z)-XV the difference
in the RI of (Z)- and (E)-(1-methyl-1-propenyl)
benzenes [1116 – (1094±5)] = (22±5 ) can be taken
into account which gives approximately (1395±13) +
(22±5) ≈ (1417±14). The value obtained is less
consistent with the experimental RI value (1452), but
enough to predict the order of chromatographic elution
of (Z) and (E)-isomers, (E) < (Z).
As noted above, two other products of this reaction
naturally correspond to the theoretically expected 3,4-
dimethyl-1-phenyl-2-butene-1-one (XII) and (Z,E)-
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

FEATURES OF THE CHROMATOGRAPHY-MASS SPECTROMETRIC IDENTIFICATION
1825
isomers of 4-phenyl-3-hexen-2-one (XIII). Given the
results of Example 1, for these isomers instead of
interpretation of similar mass spectra and obtaining
additive estimates of RI, one can either use the above
stated rule or compare estimates of their normal
boiling points obtained using the ACD software:
263.7±13°C (XII) and 280.7±10°C (XIII). Thus, the
order of elution is XII < XIII.
three are isomeric condensation products (1:2), C3H6O +
2C7H6O – 2H2O = C17H14O with molecular masses of
58 + 2·106 – 2·18 = 234 (XVIII–XX). It should be
noted that among them, the prevailing isomer is XIX,
whereas the relative amounts of XVIII and XX do not
exceed 1%. In addition, this reaction mixture contains
a significant amount of the condensation product of
three molecules of benzaldehyde with two molecules
of acetone, 3C6H5O + 2CH3COCH3 – 3H2O =
C27H24O2 (XXI, blurred peak with the retention time
49.1 min), presumably with the structure of
1,7-diphenyl-5-(2-phenylethenyl)-1,6-heptadien-3-on-
5-ol. Table 3 contains the mass spectra of all the
products.
Table 2 lists the mass spectral data of all four
components of the reaction mixture, neither of which
is presented in the database [3].
This example shows that the discussed version of
the additive schemes is almost the only possible way to
obtain estimates for the RI of (E)- and (Z)-isomers of
organic compounds π-diastereomers which should be
regarded as one of its major advantages. The following
examples further illustrate this possibility by con-
sidering the condensation products of acetone and
benzaldehyde, although, of course, it requires a
separate special discussion. Earlier in the case of
products of chlorination of cyclohexane it was shown
that this version of the additive scheme is applicable to
estimation of the RI of σ-diastereomers [26, 28].
The first of these products can be unequivocally
identified as 4-phenyl-3-butene-2-one (XVI) [most
likely (E)-isomer] by comparison with the known mass
spectrum [3] so that the estimation of its RI is not
formally required. The poor compliance of the
experimental (1374) and published (1346 [29]) values
of RI does not prevent this conclusion. The poor
compliance is caused by the use of a column with
polydimethylsiloxane stationary phase HP-5 MS with
5% of phenyl groups and the strong dependence of the
RI of conjugated compounds such as XVI even on a
minor variations in the polarity of stationary phase. A
component with RI 1404 is the primary condensation
product of acetone and benzaldehyde, 4-hydroxy-4-
phenyl-2-butanone (XVII). The database [3] does not
include its mass spectrum, but it can be attributed
unambiguously. In addition, its RI can be estimated in
accordance with the following additive scheme:
Example 3. Identify the five products of acetone
and benzaldehyde condensation.. The first one (XVI)
corresponds to the expected product with the stoi-
chiometry (1:1) with the molecular formula C3H6O +
C7H6O – H2O = C10H10O and the molecular mass 58 +
106 – 18 = 146. The molecular mass of the second one
is 58 + 106 = 164, which corresponds to the formula
C3H6O + C7H6O = C10H12O2 (XVII), and the following
OH
OH
O
O
_
_
+
+
C6H5
C6H5
(854 9)
C6H5
C6H5
_
_
+
_
+
_
+
+
(1048 11)
(1411 19); RIexp 1404
(1217 13)
=
The three products XVIII–XX with identical mass
spectra represent three possible (Z,E)-isomer of 1,5-
diphenyl-1,4-pentadien-3-one {the mass spectra of
(Z,Z) and (E,E )-isomers [3] are known to be almost
identical}. but their unambiguous assignment in the
absence of reference samples is a rather difficult task.
The problem consists primarily in the fact that for
many classes of compounds the RI values of (Z)- and
(E)-isomers are not very reliable, and often this is
caused just by the ambiguity of correlating the
structures of isomers with the chromatographic data.
To apply the considered additive scheme one must
choose among the accessible reference RI values those
that characterize precisely the structural fragments (Z)-
and (E)-C6H5–CH=CH–CO–. The most detailed and
reliable are the experimental data for the esters of
isomeric cinnamic acids, e.g., methyl-(Z)- and -(E)-
cynnamates: RI are equal to 1310±11 and 1372±11,
respectively [3] Using these data we can suggest the
version of assembling the structures of compounds
XVIII–XX shown below. As another predecessor, we
use 4-phenyl-3-butene-2-one identified as a product of
the same reaction to which has been attributed the
structure of the (E)-isomer XVI:
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

1826
ZENKEVICH, UKOLOV
O
O
O
O
_
+
C6H5
OMe
C6H5
C6H5
OMe
C6H5
_
_
_
_
_
+
+
+
(1374) (E)
(1374) (E)
(510 5)
(2236 12) (E,E), RIexp 2244
(1372 11) (E)
+
+
=
=
_
+
_
+
_
+
(510 5)
(2174 12) (E,Z), RIexp 2185
(1310 11) (Z)
Since the difference in RI of (E)- and (Z)-isomers is
62 index units, the estimate of RI of (1Z,4Z)-1,5-
diphenyl-1,4-pentadiene-3-one (XVIII) is 2112±16,
which agrees well with the value of RIexp 2122. As a
result, it is possible not only to predict the sequence of
elution of isomers, (Z,Z) < (Z,E) < (E,E), but also to
achieve a satisfactory agreement between experimental
and theoretically predicted RI values.
Of course, a detailed analysis of the problems and
features of gas chromatography-mass spectrometric
identification of previously characterized products of
Table 3. Mass spectra and retention indices of the condensation products of acetone and benzaldehyde (to Example 3)
Existence in
Mass spectrum [m/z ≥ 40 (I ≥ 3%)]
(E)-4-Phenyl-3-butene-2-one (XVI, 27%)
RIexp
RIcalc
database [3]
а
147(6), 146(57) [М]+, 145(66), 132(9), 131(89), 115(4), 104(8), 103(100), 102(15), 78(5), 77(45), 1374
+
–
76(8), 75(5), 74(6), 65(5), 63(9), 62(4), 58(5), 52(8), 51(43), 50(14), 43(37)
4-Hydroxy-4-phenyl-2-butanone (XVII, 4%)
165(3), 164(19) [М]+, 149(5), 147(4), 146(30), 145(13), 131(18), 108(4), 107(51) [С6Н5СНОН], 1404
106(40), 105(40), 104(11), 103(21), 102(3), 91(5), 80(5), 79(68), 78(20), 77(68), 58(29), 53(4),
52(9), 51(34), 50(16), 44(3), 43(100) [СН3СО], 41(4)
–
+
–
1418±3
2150±5
2210±5
(1Z,4Z)-1,5-diphenyl-1,4-pentadiene-3-one (XVIII, ~0.1%)b
235(6), 234(32) [М]+, 233(49), 146(24), 133(14), 131(19), 107(10), 106(13), 105(7), 104(100), 2122
103(49), 92(12), 91(24), 89(10), 85(8), 79(11), 78(18), 77(33), 76(9), 75(2), 69(9), 63(9), 53(7),
51(30), 50(5), 45(9), 43(7), 41(6), 40(3)
( (1E,4Z)-1,5-diphenyl-1,4-pentadiene-3-one (XIX, 40%)
235(18), 234(97) [М]+, 233(100), 217(2), 215(5), 206(6), 205(15), 204(5), 203(7), 202(5), 192(2), 2185
191(13), 190(6), 189(4), 179(3), 178(4), 165(4), 157(5), 156(13), 132(5), 131(46), 129(9), 128(27),
116(5), 115(9), 107(2), 105(2), 104(12), 103(72), 102(24), 101(8), 95(3), 92(3), 91(36), 89(9), 78(8),
77(71), 76(11), 75(6), 74(4), 63(7), 62(2), 53(3), 52(6), 51(35), 50(11)
(1E,4E)-1,5-diphenyl-1,4-pentadiene-3-one (XX, ~0.6%)b
235(20), 234(36) [М]+, 217(4), 187(5), 170(3), 165(3), 161(3), 160(5), 159(4), 157(3), 156(3), 2244
148(2), 147(5), 146(29), 145(14), 133(4), 132(7), 131(100), 128(3), 118(2), 117(8), 116(4), 115(5),
105(3), 104(14), 103(44), 102(6), 91(12), 89(3), 79(2), 78(7), 77(29), 76(5), 75(2), 74(2), 51(10),
50(5), 43(72)
+
–
2270±5
3С6H5O + 2CH3COCH3 – 3H2O = C27H24O2 (XXI, 28%)
c
380(4) [M]+, 265(5), 275(7), 257(4), 249(6), 248(15), 247(13), 235(7), 234(6) [M
–
–
–
C6H5CH=CH–C(OH)=CH2], 233(5), 157(15), 146(4) [M – 234], 144(5), 132(10), 131(100)
[C6H5CH=CHCO], 129(15), 129(5), 117(6), 115(10), 104(10), 103(45), 91(16), 78(6), 77(24), 51(6)
a
Component is uniquely identified by mass spectrum, estimation of RI is not required. b According to [3], in the mass spectra of isomeric
1,5-diphenyl-1,4-pentadiene-3-ones the maximum peaks belong to the molecular ions, while I(233) < I(M). The observed discrepancies
are caused by low content (<1%) of (Z,Z)- and (E, E)-isomers. c Blurred peak with retention time 49.41 min.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

FEATURES OF THE CHROMATOGRAPHY-MASS SPECTROMETRIC IDENTIFICATION
1827
condensation of carbonyl compounds is difficult to
carry out in one journal publication. However, the
consideration of even a limited number of examples of
the condensation of acetone and the simplest aromatic
carbonyl compounds illustrates the effectiveness of
joint interpretation of mass spectrometric and chro-
matographic data. To solve such problems the hypo-
theses should be created about the possible com-
position of the analyzed samples, including the
products of the known organic reactions. One of the
most effective ways to interpret the chromatographic
parameters is the use of additive schemes for
evaluation of retention indices, which allows revealing
the structures of the regio- and stereoisomers when the
interpretation of the mass spectral information only
does not lead to an unambiguous answer.
the database of NIST/EPA/NIH) were treated statis-
tically, presenting them as RI±sRI. Standard deviations
of calculated RI values (S) were estimated with the
relation S = (Σsi2)0.5, where si is standard RI deviation
of all the selected structural analogs.
The relative amounts of the condensation products
was estimated from the peak areas using internal
normalization method without considering the
differences in detector sensitivity to different isomers.
The eestimates of the normal boiling points of isomers
were obtained using the ACD software. The
optimization of molecular geometry was performed
using MM+ program within the HyperChem (version
6.0) software.
ACKNOWLEDGMENTS
EXPERIMENTAL
This work was supported by the Russian Founda-
tion for basic research, the project no. 11-03-00155.
For the synthesis of condensation products, in 2 ml
glass ampules was placed 400 µl of an individual
carbonyl compound or a binary mixture in the ratio 1:1
(acetone, benzaldehyde, acetophenone, propiophenone)
and 10 mg of dry NaOH was added. The ampules were
sealed and heated to 70ºC for 2 h. Then to the reaction
mixtures 1 ml of diethyl ether was added and 1 µl of
solution was taken for analysis.
REFERENCES
1. Zenkevich, I.G., Carbonyls: Derivatization for GC
Analysis. In: Encyclopedia of Chromatography, Cazes, J.,
Ed., New York: Taylor & Francis, 3 ed., vol. 1, p. 310.
2. McMurry, J.E., Organic Chemistry, New York: Brooks
Cole, ch. 23, 6 ed., 2003.
3. NIST/EPA/NIH Mass Spectral Library. Software/Data
Version (NIST08); NIST Standard Reference Database,
Number 69, August 2008. National Institute of
Standards and Technology, Gaithersburg, MD 20899;
http://webbook.nist.gov (January 2011).
The chromatography-mass spectrometry analysis
was performed on a Shimadzu QP5000 instrument
with a quadrupole mass analyzer and a column HP-5
MS of 25 m long and 0.20 mm internal diameter (film
thickness of stationary phase 0.33 µm). The mode of
temperature programming was heating from 40 to 280ºC
at a rate 10 deg min–1, the flow rate of carrier gas
(helium) 1.2 ml min–1. The flow separation at the
injection 1:50. Evaporator temperature 250ºC, inter-
face temperature 280ºC. To determine the gas-
chromatographic retention indices a mixture of n-
alkanes C6–C20 was used (Supelco, cat. no. 500631),
the standard deviation of RI was about 1 index unit.
The mass spectra were recorded with the standard
electron impact ionization energy 70 eV in the scan-
ning mode of the total ion current (2 scans s–1) in the
range m/z 45–700. The data processing was performed
using the GCMS Solution 2.60 and AMDIS 2.66 soft-
wares. To estimate the factors of matching the mass
spectra with the results of the library search was used
the NIST Mass-Spectral Search Program v.2.0
software. All the reference values of retention indices
determined on WCOT columns with standard nonpolar
phases under linear temperature programming (from
4. Rowland, C.Y., Blackman, A.J., D’Arcy, B.R., and
Rintoul, G.B., J. Agr. Food Chem., 1995, vol. 43,
p. 753.
5. Aaslyng, M.D., Elmore, J.S., and Mottram, D.S., J. Agr.
Food Chem., 1998, vol. 46, no. 12, p. 5225.
6. Tan, S.T., Wilkins, A.L., Holland, P.T., and McGhie, T.K.,
J. Agr. Food Chem., 1989, vol. 37, p. 1217.
7. Ibragim, E.A. and Suffet, I.H., J. Chromatogr., 1988,
vol. 454, p. 217.
8. Pino, J.A., Almora, K., and Marbot, R., Flavor Fragr. J.,
2003, vol. 18, p. 492.
9. Huang, T.-C., Bruechert, L.J., Hartman, T.C., Rosen, R.T.,
and Ho, C.-T., J. Agr. Food Chem., 1987, vol. 35, no. 6,
p. 985.
10. Habu, T., Flath, R.A., Mon, T.R., and Morton, J.F.,
J. Agr. Food Chem., 1985, vol. 33, no. 2, p. 249.
11. Yang, R. and Sugisawa, H., Nippon Shokuhin Kogyo
Gakkaishi., 1990, vol. 37, no. 12, p. 946.
12. Sefidkon, F., Jalili, A., and Mirhaji, T., Flavor Fragr. J.,
2002, vol. 17, p. 150.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011

1828
ZENKEVICH, UKOLOV
22. Zenkevich, I.G., in 100 Let khromatografii (100th
Anniverary of Chromatigraphy), Moscow: Nauka, 2003,
p. 311.
13. Pino, J.A. and Marbot, R., J. Agr. Food Chem., 2001,
vol. 49, p. 5880.
14. Heydanek, M.G. and McGorrin, R.J., J. Agr. Food
Chem., 1981, vol. 29, p. 950.
23. Zenkevich, I.G., Zh. Fiz. Khim., 2003, vol. 77, no. 1,
15. Guyot, C., Scheirman, V., and Collin, S., Food Chem.,
1999, vol. 64, p. 3.
16. Guyot, C., Bouseta, A., Sheirman, V., and Collin, S.,
J. Agr. Food Chem., 1998, vol. 46, p. 625.
p. 92.
24. Zenkevich, I.G., Moeder, M., Koeller, G., and Schrader, S.,
J. Chromatogr., A, 2004, vol. 1025, p. 227.
25. Zenkevich, I.G., Makarov, A.A., Schrader, S., and
17. Stein, S.E., Babushok, V.I., and Brown, R.L., J. Chem.
Inf. Model., 2007, vol. 47, no. 3, p. 975.
18. Chen, J. and Ho, C.-T., J. Agr. Food Chem., 1999,
Moeder, M., J. Chromatogr., A, 2009, vol. 1216, p. 4097.
26. Zenkevich, I.G., Eliseenkov, E.V., and Kasatochkin, A.N.,
Chromatographia, 2009, vol. 70, no. 5/6, p. 839.
vol. 47, p. 643.
27. Hamming, M.C. and Foster, N.C., Interpretation of
Mass Spectra of Organic Compounds, New York: Acad.
Press, 1979.
19. Buttery, R.G., Stern, D.J., and Ling, L.C., J. Agr. Food
Chem., 1994, vol. 42, no. 3, p. 791.
20. Rapior, S., Breheret, S., Talou, T., and Bessiere, J.-M.,
J. Agr. Food Chem., 1997, vol. 45, p. 820.
21. Zenkevich, I.G., Zh. Analit. Khim., 1998, vol. 53, no. 8,
28. Zenkevich, I.G. and Ukolov, A.I., Zh. Strukt. Khim.,
2010, vol. 51, no. 4, p. 671.
29. Tudor, E., J. Chromatogr., A, 1997, vol. 779, p. 287.
p. 828.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 81 No. 9 2011