CC BY
43
YflK 54.061:543.544.32:547.022
BecTHHK Cn6ry. Cep. 4. 2012. Bun. 3
A. O. Razgoniaev, A. I. Ukolov, I. G. Zenkevich
APPLICATION OF COMBINED GC AND MS DATA IN GC-MS DETERMINING THE STRUCTURES OF PRODUCTS OF PHENOL ALKYLATION BY BUTYL ALCOHOLS
Introduction. The generally accepted approach in identification of previously non characterized organic compounds assumes their isolation followed by determination of their physicochemical properties (boiling point, melting point, density, refractive index, etc.) and/or more informative spectral characteristics (first of all, NMR and mass spectra). However, in the analytical practice there are samples consisting of components which preparative isolation is impossible or unreasonable due to complexity of these objects.
An alternative approach in the identification of organic compounds is based on the application of hyphenated analytical techniques, namely gas chromatography-mass spectrometry (GC-MS) which allows us to register the mass spectra of constituents without their isolation from mixtures. Nevertheless, mass spectra alone are not enough often to identify the components of the mixtures unambiguously and the determination of structures of analytes molecules requires the use of additional data. As such data the gas chromatographic retention indices (RI) could be used. Combined application of two independent parameters increases the probability of correct identification of the analytes.
As an example, it is advisable to consider three hydrocarbons commonly encountered in mixtures one with another, namely 1-ethyl-4-methylbenzene (RI 960 ± 5), 1-ethyl-2-methylbenzene (RI 979 ± 7) and 3-methylnonane (RI 972 ± 3). The mass spectra of alkyl benzenes differ insignificantly by the ratio of intensities of peaks with the same m/z values, while the RI values of 1-ethyl-2-methylbenzene and 3-methylnonane are overlapped. Therefore it is impossible to identify unambiguously such components using both mass spectral information and chromatographic RIs separately. However, the joint using of these two analytical parameters allows us to identify them unambiguously. It should be noted this approach is used preferably in identification of compounds already characterized by mass spectra and/or gas chromatographic RIs at present. For example, the last version of NIST/EPA/NIH 2011 database includes mass spectra of 212'961 compounds and GC RIs on standard polar and non-polar phases for 70'838 compounds [1, 2]. Accepting all the advantages of this approach, its limitation is those that only small part of organic compounds is characterized by analytical parameters. At the same time, the major criterion of informativeness of contemporary algorithms of the identification of organic compounds seems to be the effective interpretation of the analytical data for analytes not represented in the reference databases. The possible mode of solving such problems is a precalculation of RIs for these compounds. At first, this is the task of the most importance for products of non-regioselective organic reactions which cannot be isolated from reaction mixtures. Naturally, to find any special mentions concerning non-regioselective reactions is highly difficult, because it is not a classification criterion. Nevertheless, just such objects are most important in analytical practice [3-7].
In our work the reaction of phenol alkylation by isomeric butyl alcohols in the presence of aluminum chloride is considered. This method of alkylation was described even in the middle of 20th century [8], but no attempts of the detailed analysis of all reaction products are known
© A. O. Razgoniaev, A. I. Ukolov, I. G. Zenkevich, 2012
up to present. These reaction mixtures contain sufficiently large number of constituents to make the task of their separation and/or isolation too time- and labor-consuming. Besides that the differences in the mass spectra of isomeric alkyl phenols may be quite insufficient. The goal of the present work is to provide the comparative analysis of different methods of RI calculation as well as to prove the possibility of structures evaluations for reaction products using RI for compounds not characterized yet.
In reaction mixtures of phenol alkylation by alcohols in the conditions chosen the following classes of compounds have been found: butyl phenyl ethers (average content ~ 39 %), butyl phenols 42 %), dibutyl phenols 4 %), butyl butylphenyl ethers 3 %), alkenyl phenols (less than 0.1 %), bis-(4-hydroxyphenyl)butanes 8 %), and propyl phenols 0.5 %) as the products of destructive alkylation. It is noteworthy that most of them are poorly characterized by both physicochemical and spectral data [1], that complicates their identification.
Experimental. Anhydrous aluminum chloride (0.03 mol, 4 g) was placed in round-bottomed 100 ml flask equipped with a reflux condenser, followed by dropping the solution of 0.02 mole of phenol (1.9 g) in 0.06 mole of the corresponding alcohol (4.5 g). The mixture was heated in an oil bath during 3 h (for isobutanol and n-butanol) or 1 h (for ieri-butanol and sec-butanol), washed with water and dilute hydrochloric acid; the products were extracted with chloroform.
Gas chromatographic analysis of extracts was performed with gas chromatograph-mass spectrometer Shimadzu QP 5000 (Research institute of hygiene, occupational pathology and human ecology, St.Petersburg) equipped with quadrupole mass analyzer (electron impact mode of ionization) and column HP-5 MS, 30 m length, internal diameter 0.25 mm, stationary phase film thickness 0.25 The temperature programming was 80-280 °C, heating rate 4 °/min, cut-off time 11 min, injector temperature 260 °C; interface temperature 280 °C. The flow rate of carrier gas (helium) was 1.2 ml/min. Micro syringe Hamilton (10 ml) was used for injection of liquid samples; volume was 1.0 ^L. Mass spectra were registered in total ion current scan mode within m/z range 39-700. To measure the retention times and peak areas the software package GC-MSsolution was used. The mixture of reference n-alkanes (Supelco) was used for RI determination. To distinguish the dibutyl phenols and butyl butylphenyl ethers the samples were dissolved in deuteromethanol (CD3OD). The replacement of active hydrogen atoms in hydroxyl groups of phenols onto deuterium leads to the increasing the intensities of the peaks [M + 1] in their mass spectra; no effects are observed for butyl ethers. The linear-logarithmic retention indices were calculated using the QBasic program given in the manual [9]. Evaluation of RI was performed using the non-incremental additive scheme [3].
Results and their discussion. Among the numerous methods of RI evaluation [10] known at present it is reasonable to use the simplest and therefore most accessible of them. At least three groups of them should be mentioned specially.
1. The algorithms based on physicochemical characteristics of target compounds. A key problem of this calculation mode is the description of RI dependence on selected physico-chemical properties using mathematical equations whose parameters are chosen empirically (see, e. g., [11]). The general mathematical form of such equations is the following:
RI = a\x\ + ... + anxn,
where xi,.. .,xn is the set of physicochemical parameters; a\,...,an are constants.
The obvious disadvantage of this approach is that the linear form of these equations often do not have any justification. Among the nonlinear relations the equation of RI dependence
vs. boiling point (Tbp) of homologues or congeners is known:
log RI = a log Tbp + bnc + c,
where Tbp is boiling point; nC is number of carbon atoms in the homologue; a, b, c are coefficients calculated by the least squares method.
There are examples of the reciprocal logic, i. e., the calculation of physicochemical parameters of compounds on the base of retention indices [12].
2. An example of the "classical" additive scheme is the algorithm developed at NIST [13]. Target RIs are calculated by summarizing atom and group increments:
N
RI
J2ARii + f,
i=l
where f is the average deviation of calculated RIs from experimental values (used for additional correction of results).
3. "Non-incremental" (or "modified") additive scheme.
RI calculation by using a modified additive scheme [3] is as follows (Scheme (2)). At first, select the structures (characterized by RIs) corresponding with structure of the target compound as closely as possible. Secondly, form the "superposition" of these structures and subtract the structure formed by duplicated (overlapping) fragments. The corresponding arithmetic operations should be performed with the retention indices. As an example, Scheme (2) shows the "assembly" of 2,4-dimethyl phenol from structures of o- and p-cresol. The result of RI evaluation "in parallel" with this "assembly" is shown too:
OH
OH
CH
OH
l^CH,
1031
CH
1059
OH
OH
CH
(2)
CH3
"2090"
957
CH
= 1133
In this case, the difference between RIs of 2,4-dimethyl phenol and o-cresol is essentially the increment of the replacement of hydrogen atom in 4th position of o-cresol by methyl group. It is principal that reference data are not the averaged increments but RI values of compounds which structures match the structure of the target analyte most precisely. Maximum accuracy of estimated RIs can be achieved only when the selected precursor structures include all the features of the target structures influencing their gas chromatographic retention parameters. It was this additive scheme used in identification of the 839 congeners of polychlorinated hydroxybiphenyls [3] and 211 structural isomers of nonyl phenols [7].
Table 1 includes the results of RI evaluation of isomeric butyl phenols based i) on the algorithm developed at NIST (RInist), ii) the relation connecting RI and Tbp of homologues or congeners (RIT), iii) using a modified additive scheme (RIadd); the experimentally determined RI values (RIexp) and RI values know from the reference (RIref) are presented, as well.
Table 1 permits us to conclude the algorithm developed at NIST principally does not allow us to distinguish ortho- and para-isomers, as well as some structural isomers, namely, sec- and isobutyl phenol.
+
+
Evaluation of retention indices for isomeric butyl phenol by different methods and their identification in the reaction mixtures
RInist RIt Rl&dd -ß-^exp Rlrei Component
1228 1317 * 1297 1296 4-ferf-butyl phenol
1248 1329 1319 1319 1318 4-sec-butyl phenol
1248 1310 1312 1312 No data** 4-isobutyl phenol
1312 1367 1360 1360 1361 4-n-butyl phenol
1228 1275 * 1274 1274 2-tert-butyl phenol
1248 No data** 1294 1281 No data** 2-sec-butyl phenol
1248 No data** 1292 1270 No data** 2-isobutyl phenol
1312 1315 1338 1321 No data** 2-n-butyl phenol
* The RI values of 4- and 2-tert-butyl phenol were selected as starting data for calculations.
** No reference data for boiling point or reference RI values.
RI Calculations using the correlation equation (1) allows us to identify some isomeric butyl phenols. However, the objective increasing the number of isomers of higher homologues of any series is accompanied by obvious decreasing the part of isomers characterized by physicochemical properties (including normal boiling point). This fact restricts the applicability of this RI evaluation method. Thus, Tbp values are known for 6 from 8 isomeric butyl phenols (assuming the boiling points of the para- and meta-isomers are close each other), as well for pentyl phenols Tbp values are known for 4 from 16 isomers only [14]. In the RI evaluating for more complex products of phenol alkylation by butyl alcohol the problem of the availability of reference data on the boiling point becomes more actual.
RI Evaluation for isomeric butyl phenol based on "modified" additive scheme was as follows: 2-tert-butylphenol and 4-tert-butylphenol were selected as the basic structures to calculate RI values for other ortho- and para-isomers, respectively, "stacked" with structure of butyl benzene (reference RI values are known for all four isomers), following by subtracting the structure of tert-butylbenzene (Scheme (3)):
OH OH
The similar way was used for evaluating the RIs of butyl phenols, phenyl butyl ethers dibutyl phenols and butyl butylphenyl ethers. Corresponding tert-butyl derivative was selected as the basic structures followed by corresponding structural transformations to obtain RI values of target analytes. It is noteworthy that the term "additive scheme" does not imply a single RI evaluation algorithm (the calculations can be realized in different ways) but the option used seems to be useful for the following reasons:
— tert-butyl derivatives can be easily identified in the mixtures using their mass spectra;
— they are better characterized by the reference mass spectra and RI data;
— the additive schemes drawn up on the basis of tert-butyl derivatives reflect all structural features of other isomers.
If the number of isomeric monosubstituted butyl phenols is eight, the presence of the second alkyl radical in the benzene ring increase the number of isomeric dibutyl phenols up to 80. The calculating RI values for all 80 isomers is solvable but time-consuming problem. This task can be simplified by taking the following fact into account: the location of two
alkyl substituents in the neighboring positions seems to be highly unlikely. This feature prevents the formation of 2,3- and 3,4-dibutyl phenols during alkylation. Keeping in mind this condition the total number of 80 possible isomers can be reduced up to 48 structures for further consideration (Scheme (4)):
OH
OH
OH
OH
OH
R
R
OH
R
OH OH
R
(4)
OH
Rl R2
The RI values of 3,5- and 2,5-di-ieri-butyl phenols were evaluated in the following ways (Scheme (5)):
OH
OH
(5)
Table 2 presents the results of RI evaluation for dibutyl phenols using additive schemes (1) and (5) in comparison with RI values of the constituents of reactions mixtures.
In relation to 2,6-disubstituted butyl phenols an appropriate RI evaluation scheme reflecting the location of two sterically hindered substituents in the ortho-positions to the hydroxy group was not found. Large deviations of evaluated RIs from experimentally measured RI values of 2,6-dibutyl phenols are illustrated by Table 3.
The alkylation of aromatic compounds by Friedel—Crafts method leads to the formation of significant amounts of by-products those can be not only structural isomers but belong
+
+
+
Results of RI evaluation for dibutyl phenols using additive schemes (4) and (5), and their identification in the reaction mixtures
Rl&dd R^exp Component
1524 1524 2-sec-butyl-4-ferf-butyl phenol
1533 1536 2-tert-butyl -4-isobutyl phenol
1544 1534 2,5-di-tert-butyl phenol
1547 1556 2,4-di-sec-butyl phenol
1571 1565 3,5- di-tert-butyl phenol
1560 1565 2-tert-butyl -5-isobutyl phenol
1574 1566 2,5-di-sec-butyl phenol
1588 1571 3-sec-butyl -5-tert-butyl phenol
1611 1592 3,5-di-sec-butyl phenol
OH
Low precision of additive RI evaluations for 2,6-dibutyl phenols*
(6a)
(6b)
Table 3
Rl&ddl RI&dd2 -ß-^exp Rltsbi Component
1564 1445 1444 2,6-di-tert-butyl phenol
1578 1459 1502 2,6-di-sec-butyl phenol
* RIaddl is RI calculated according to scheme (6a) and RIadd2 is RI calculated according to scheme (6b).
to different homologous series. This fact makes the identification of reaction products more complicated. Together with the main reaction the formation of not alkylation, but etherifi-cation products (alkyl phenyl ethers) is observed, as well.
The additive scheme of RI evaluation for phenyl ethers is the following:
OR
RCH
However, this scheme does not allow us to distinguish isobutyl and sec-butylphenol ethers because both modes of calculation imply the same alkane RCH3, namely 2-methylbutane. Nevertheless, the use of mass-spectrometric data (in the case of sec-butyl derivative the weak, but characteristic signals [M-29] are observed) permits us to establish the structures of such isomers unambiguously. The results of RI evaluations for butyl phenyl ethers in comparison with their experimentally measured RI values are presented in Table 4.
Results of RI evaluations for butyl phenyl ethers and their identification
in reaction mixtures
ß-^add Rlexp Component
* 1074 ierf-butyl phenyl ether
1126 1122 sec-butyl phenyl ether
1126 1139 isobutyl phenyl ether
1161 1191 ??-butyl phenyl ether
* Reference values for calculations.
RI Evaluations for alkyl alkylphenyl ethers was performed in the similar way as those for alkyl phenols, namely corresponding tert-butyl derivatives were selected as the basic structures, followed by the sequence of structural transformations illustrated by the Scheme (7):
X X
The results of RI evaluating for butyl butylphenyl ethers using the additive scheme (7) in comparison with experimentally determined RI values are presented in Table 5.
Table 5
Results of RI evaluations for butyl butylphenyl ethers and their identification
in reaction mixtures
Rladd RIexp Component
* 1353 2rferf-butylphenyl terf-butyl ether
* 1385 4-tert-butylphenyl tert-butyl ether
1408 1426 2-sec-butylphenyl sec-butyl ether
1433 1437 4-ierf-butylphenyl sec-butyl ether
1456 1454 4-sec-butylphenyl sec-butyl ether
1450 1463 4-tert-butylphenyl isobutyl ether
1517 1498 2-n-butylphenyl ??-butyl ether
1564 1576 4-n-butylphenyl n-butyl ether
* See footnotes to Tables 1 and 4.
Summarizing the discussion of the results, it should be noted that mass spectra of previously non characterized compounds (not included in the NIST MS database [1]) should be presented in the complete form (Table 6).
Table 6
Mass spectra of compounds not presented in the NIST MS database
Compound Mass spectrum, m/z > 50 (Irel > 2 %)
2-sec-butyl-4rferf-butyl phenol 207(5), 206(18)M, 192(13), 191(100), 178(3), 177(25), 175(2), 163(5), 162(2), 161(3), 147(4), 135(4), 133(2), 13l(2), 129(3), 128(3), 121(6), 119(2), 117(2), 115(3), 107(7), 105(3), 103(2), 91(5), 79(2), 77(4), 65(2), 57(26), 55(3), 53(2), 51(2)
Ending of Table 6
Compound Mass spectrum, m/z > 50 (Jrei > 2 %)
2-1erf-butyl -4-isobutyl phenol 206 M(6), 192(12), 191(100), 164(3), 163(14), 161(2), 149(2), 148 (3), 147(5), 135(3), 133(2), 131(3), 123(3), 121(3), 120(3), 119(2), 116(2), 107(3), 106(2), 105(4), 103(2), 95(2), 91(5), 79(2), 77(4), 73(2), 65(2), 57(12), 55(2), 53(2)
2-ierf-butyl -5-isobutyl phenol 207(9), 206 M(39), 192(12), 191(100), 164(10), 163(93), 149(7), 148(13), 135(10), 134(6), 133(18), 131(7), 121(4), 119(6), 117(5), 115(14), 107(8), 105(12), 103(4), 102(3), 93(4), 91(9), 89(3), 88(4), 84(3), 83(3), 79(7), 78(2), 77(5), 74(6), 73(6), 67(2), 64(4), 63(2), 58(3), 57(32), 55(10), 53(4), 52(3), 51(4)
3-ifec-butyl-5-ferf-butyl phenol 207(3), 206 M(8), 192(3), 191(19), 178(15), 177(100), 175(2), 163(5), 162(10), 161(2), 149(2), 147(8), 135(12), 133(3), 129(3), 128(4), 121(7), 117(2), 116(2), 115(4), 107(7), 105(3), 103(3), 91(7), 79(2), 77(4), 65(2), 57(25), 55(5), 53(2), 51(2)
3,5-di-ifec-butyl phenol 207(2), 206 M(14), 178(13), 177(100), 163(3), 161(2), 149(2), 148(7), 147(4), 132(2), 133(5), 129(2), 128(3), 121(11), 117(2), 115(4), 107(8), 105(3), 103(2), 91(6), 77(4), 65(2), 57(7), 55(3), 53(2)
4-1erf-butyl phenyl-1erf-butyl ether 206 M(1), 151(2), 150(10), 136(10), 135(100), 119(2), 107(12), 95(4), 91(6), 79(2), 77(4), 74(2), 65(2), 57(6), 56(2), 55(3), 51(2)
2-sec-butyl phenyl-sec-butyl ether 207(2), 206 M(10), 177(2), 151(3), 150(23), 135(4), 123(2), 122(12), 121(100), 108(2), 107(7), 104(2), 103(3), 94(2), 93(3), 92(2), 91(9), 79(2), 78(3), 77(7), 65(3), 57(2), 55(4)
4-ierf-butyl phenyl-sec-butyl ether 206 M(7), 151(2), 150(7), 149(4), 147(2), 136(11), 135(100), 133(2), 131(3), 119(4), 115(2), 107(11), 103(2), 95(4), 91(4), 89(2), 79(2), 78(2), 77(5), 65(2), 57(5), 55(3), 53(2), 51(2)
4-sec-butyl phenyl-sec-butyl ether 206 M(7), 177(2), 150(7), 135(3), 122(9), 121(100), 107(7), 103(3), 93(2), 91(6), 78(2), 77(5), 65(2), 57(2), 55(2)
4-ierf-butyl phenylisobutyl ether 207(2), 206 M(12), 192(2), 191(19), 150(3), 136(10), 135(100), 119(2), 115(2), 107(12), 105(2), 95(3), 91(6), 79(2), 77(4), 65(2), 57(5), 55(2)
2-n-butylphenyl-n-butyl ether 206 M(8), 163(15), 150(35), 108(14), 107(100), 105(6), 103(6), 95(5), 91(9), 89(5), 79(4), 78(7), 77(11), 65(3), 52(5)
4-n-butyl phenyl-??-butyl ether 207(6), 206 M(16), 163(7), 151(4), 150(15), 119(3), 108(12), 107(100), 94(3), 92(2), 91(2), 90(2), 78(2), 77(4), 57(3), 52(3), 51(2)
Thus, the approach considered appears to be acceptable in establishing the structures of phenol alkylation by isomeric butyl alcohols, when the interpretation of mass spectral information only does not provide unambiguous results.
References
1. NIST/EPA/NIH Mass Spectral Library [Электронный ресурс]. Software/Data Version (NIST 05, NIST 08) (CD-ROM).
2. NIST Chemistry WebBook [Электронный ресурс]. URL: http://webbook.nist.gov/ (accessed January 2012).
3. ZenkevichI. G., Moeder M., Koeller G., Schrader S. Using new structurally related additive schemes in the precalculation of gas chromatographic retention indices of polychlorinated hydrox-ybiphenyls on HP-5 stationary phase //J. Chromatogr. (A). 2004. Vol. 1025. P. 227-236.
4. ZenkevichI. G., Eliseenkov E. V., Kasatochkin A. N. Using retention indices for GC-spec-trometric identification of halogen-containing organic compounds // Mass-spectrometry. 2006. Vol. 3, N 2. P. 131-140.
5. ZenkevichI. G., Makarov A. A. Identification of alkylarene chloromethylation products using gas chromatographic retention indices // Rus. J. General Chem. 2007. Vol. 77, N 4. P. 611-620.
6. ZenkevichI. G., Eliseenkov E. V., Kasatochkin A. N. Chromatography "component" of GC-MS identification of cyclohexane chlorination products // Mass-spectrometry. 2009. Vol. 6, N 2. P. 137-148.
7. Zenkevich I. G., Makarov A. A., Schrader S., Moeder M. A new version of an additive scheme for the prediction of gas chromatographic retention indices of the 211 structural isomers of 4-non-ylphenol // J. Chromatogr. (A). 2009. Vol. 1216. P. 4097-4106.
8. Adams R. Organic Reactions. N.-Y.: John Wiley & Sons, Inc, 1946. Vol. 3. 460 p.
9. Stoliarov B. V., SavinovI. M., Vitenberg A. G. et al. Practical Gas and Liquid Chromatogra-phy. S.-Petersburg: University press, 2002. 610 p.
10. ZenkevichI. G. Compilation "100 years of chromatography". Moscow: Science, 2003. P. 311.
11. Buydens L., MassartD. L., Geerlings P. Prediction of gas chromatographic retention indexes with topological, physicochemical, and quantum chemical parameters // Anal. Chem. 1983. Vol. 55. P. 738-744.
12. Weiner P. H., Parcher J. F. Prediction of some physical properties of organic molecules by factor analysis of gas chromatographic retention indices // Anal. Chem. 1973. Vol. 45, N 2. P. 302-307.
13. Stein S. E., Babushok V. I., Brown R. L., Linstrom P. J. Estimation of Kovats Retention indices using group contributions // J. Chem. Inf. Model. 2007. Vol. 47. P. 975-980.
14. PotekhinA. A. Properties of Organic Compounds: Handbook. Leningrad: Chemistry, [W. y.]. 518 p.
Статья поступила в редакцию 28 февраля 2012 г.
