Gas Chromatographic - Mass Spectrometric Identification of AI kylcyclohexanes and Cyclohexenes

J. Albaig~s / X. G u a r d i n o * Instituto de Qu[mica Bio-Org~nica (CSIC), Jorge Girona Salgado, Barcelona-34, Spain

The purpoose of this paper was to study the GC behaviour of a series alkylcyclohexane hydrocarbons, structurally related to p-menthane (Fig. 1, II-V), on five stationary phases of varying polarity (squalane, SE-30, Apiezon L, OV-225 and PEG 20M). All saturated, monoolefinic and aromatic isomers were considered, as well as their structural precursors from cyclohexane (see Table I). In order to exploit this information, an olefin-rich synthetic hydrocarbon mixture (pyrolysis naphtha) has been analyzed by GC-MS, resulting in the positive identification of 17 (10 for the first time) of the 34 hydrocarbons studied.

Key Words Gas chromatography Kov~its retention indices Alkylcyclohexanes Mass spectrometry Summary A series of 34 alkylcyclohexane hydrocarbons structurally related to p-menthane, including all saturated, monoolefinic and aromatic structural precursors from cyclohexane, have been characterized by GC and MS. Kov~ts retention indices have been determined on five stationary phases (squalane, SE-30, Apiezon L, OV-225, PEG 20M) and the corresponding polarity, temperature and structural increments have been calculated. From the information acquired I 0 alicyclic hydrocarbons have been identified for the first time in a pyrolysis naphtha.

Experimental

Introduction Cyclohexane derived hydrocarbons are widely distributed among synthetic and natural products, such as gasolines, cracked naphthas and essential oils, their identification being of special interest from chemical and geochemical standpoints. Although GC-MS is the most suitable technique for the analysis of such mixtures, the identifications are not always unequivocal, especially when there is a lack of reference compounds and when several positional and conformational isomers, which can display very similar mass spectra, can be present. In this respect, Kovats retention indices, including polarity (AI), temperature (8I/~5T) and structural increments (~iI) [1] can be used to complement the spectral identification. Retention data on mono- and disubstituted C6-C~s alkylcyclohexanes, cyclohexenes and benzenes have already been described [ 2 - 5 ] , however, no systematic study has been undertaken on the series of monoterpene-like hydrocarbons. * Present address: Instituto de Higiene y Seguridad del Trabajo (SSHISET), Dulcet, Barcelona34, Spain Chromatographia Vol. 13 No. 12, December 1980 0009-5893/80/12 0755-08 $ 02.00/0

The hydrocarbons used in the present study were obtained, with a purity over 95%, from Fluka A.G. (Nos. 1-12, 17, 18 and 2 4 - 2 6 in Table I), and by synthesis in the laboratory. Thus, cyclohexanone and 4-methylcyclohexanone afforded, throughout a reaction with suitable Wittig and Grignard reagents, compounds 15, 16, 20, 22, 29 and 32 (Table I). 4-Isopropyleyclohexanone [6] gave, by similar reactions and by reduction to the alcohol and subsequent dehydration, compounds 21, 30 and 31, respectively. The isopropenyl derivatives (compounds 23, 33 and 34) were obtained by coupling 2-brompropene to the corresponding cyclohexylbromides [7]. Alternatively, compounds 23, 33 and 34 were synthesized by a Wittig reaction on the corresponding methylcyclohexylketones, previously prepared by hydrogenation of the Diels-Alder adduct

?o00 ? I

II

ill

IV

v

Fig. 1 9 Series of alkylcyclohexane hydrocarbons studied in this paper.

Originals

9 1980 Friedr. V ieweg & Sohn Verlagsgeseltschaft mbH

755

~o

~

o

~

~

~

~

~

=o

~~

~~

~

~~

~

~

~

o

~

~

o d d d d o o o d d d d d d d d d d d d d d d d d d o o o o o o o d

~g

~

0

~

~

~

~

0

0

0

0

0

0

0

~

0

0

0

0

d d d d d o d d d d d o o o o o o o o o o o o o o o o o d o o d d o O4

_g

d d d d d d o d d d d d d d d d d d d d d d d d d d d d d d d d d d

co

coco

c0 co o~ o~

m o~ o~ o~

o

o

o~ on o

~

o

~

Co

d d d d d o d d d d d d d d d d d d d d d d d d d d d d d d d d d d

%0

8 N N g $ ~ 8 ~ S g ~ h S N s 1 6 3

d d d o d o d o d d d d d d d d d d d d d d d d d o d d d d d d d o

0o m~

G

E

>~.__

0

~

0

~

0

~

~

0

0

0

0

~

0

~

m

m

~

m

0

0

0

0

~

m 0

m 0

~

m

0

~

o

~

m

~

0

~

0

0

0

0

0

0

m 0

r%

0

o >. e-

._o "5 >. 0 >-

~

cyo "O

0

0

0

0

~

0

0

0

0

0

0

~

~

0

0

0

~

0

0

0

0

0

0

~

0

0

~

0

0

~

0

0

0

0

~

0

0

0

0

0

0

~6 ._

~

~

~

0

~

~

~

~

~

O.

o

O.

.u_

C CX X ~ ~~ 0 o >

>. C~ -0

oo

2z

g

go

ooo

-~ -0 -0

>--~

E

-O

2 = 2o ~o ~o ~

E

E o

756

x x o ~ ~ 1 7 6

oo==~oo~

x

c~

.m

o>

~ ~ - -

UJ

,F.

n"

C C ~ ~ X X ~ ~ x ~ ~ C c ~ O ~

o~

o "D

0

C ~X x ~ oO

0 -o >. I

d Z

~ ; 2 _ _ zzz~.~_~

.a~

~

~

~ ~ ~

0u O ~ m e~ ~~ 1 ~7 -6 1~ 7' 6~ 1 7O 6

~ o~~ g o ~ o~ o o o o~ ~ ~ ~T

I

r ~

0

~

~

~

0

~

O (.O v

Chromatographia V01. 13 No. 12, December 1980

Originals

between methylvinylketone and 1,3-butadiene or isoprene. Finally, compounds 13, 14, 19, 27 and 28 were obtained from the 2-alkylcyclohexanones by selective elimination of their tosylhydrazone derivatives with methyllithium [ 8 ] . Reduction of p-xylene, isopropylbenzene and p-cymene by lithium in ethylamine [9] also gave mixtures of these olefinic isomers. Conformational assignments were established by hydrogenation of the olefins to the corresponding saturated hydrocarbons.

The general fragmentation pathways that have been accepted for the isomeric menthanes (CloH2o) and menthenes (CIoHts) are indicated in the following schemes: -CsH7

Retention indices were calculated by means of an HP 9830 A desk computer after optimization of the operating conditions as previously reported [10]. The standard deviations from five measurements were 0.1-0.6 i.u., the higher values corresponding to the aromatic hydrocarbons in the shorter stainless steel columns. Mass spectra were recorded at 70 eV on an AEI MS 902S instrument interfaced with the gas chromatograph by a Watson-Biemann separator. A make-up gas flow was used at the end of the column.

Results and Discussion Mass Spectrometric Identification

The previously described mass spectra of methyl- and methylisopropylcyclohexane andcyclohexene hydrocarbons [ 1 1 - 1 4 ] have shown the difficulty of establishing the unequivocal identification of certain isomers on the basis of mass spectral data alone.

Table II. Columns used during the study Stationary phase SquNane Column material

SE-30

Apiezon L OV-225 PEG-20M

s.s.

Glass

s.s.

s.s. Glass

m

100

64

30

30

40

inner diameter mm

0.25

0.30

0.25

0.25

0.30

Nr. plates (n-decane) 2 8 0 0 0 0

150000

Temperature (~

60-80-100

Length,

/ --C3H6 ~ Call 7 (55)

C1oH2o (140)-

Pyrolysis naphtha was obtained from an ethylene cracker installation (Enpetrol Spain) and analyzed as received). Chromatographic analysis was performed on a PerkinElmer 990 instrument equipped with a flame ionization detector and a digitizer unit Autolab 6300 for the retention time measurements. The carrier gas was helium. Stainless steel and glass capillary columns were coated by the dynamic and static methods, respectively, exhibiting the characteristics indicated in TabIe II.

-C2H4 ' CTH,a (97) _~ ' CsH9 (69)

-CsHs

-CH 3

-CHs

' C7H,2 (96)

' CgH,s (123)_]

, C6H 9 (81)

-C3H 6

' CoH9 (81)

/

/ - C 4 H 8 ' CsH7 (67) CtoH18 (I38)

-C3H7 ~

CTH~ (95)

-C2H 4

' CsH7 (67)

-CsHlOCsH8 (68) As shown, the elimination of methyl and isopropyl groups and retro-Diels-Alder fragmentations occur most readily. This affords relatively simple and common spectral patterns for different isomers. Furthermore, fragments from by isomeric cycloalkenes may represent about the same fraction of the total ionization, due to the possibility of allylic rearrangements before the fragmentation. In consequence, individual spectra are not always sufficiently characteristic for the structural elucidation of unknown alicyclic hydrocarbons. This is cleary exemplified in Figs. 2 and 3, where the mass spectra of the less investigated 1,4-dimethyl and isopropyl cyclohexene hydrocarbon series are presented. As can be seen, only slight differences are evident among the spectra of compounds 10, 11; 14, 15, 16 and 20, 22. Some of these differences are due to the ion fragments formed by retroDiels-Alder reactions, which are apparently more abundant in the methyl-substituted olefins, or by the loss of the methyl or isopropyl groups (M-15 and M-43), which is in turn slightly more favoured in the trans as compared with the cis isomers. Therefore it seems justified that in the analytical approach for the identification of isomeric hydrocarbons, a study of their chromatographic behaviour should be included. Gas Chromatographic Characterization

60-80

50000 60-80-100

45000

66000

60-430

60-80

Chromatographia Vol. 13 No. 12, December 1980

The GC behaviour of monoalkylcyclohexanes and cyclohexenes has been extensively studied by Rang et al. [ 2 ] , and a number of regularities described. Similary, Engewald and Wennrich [5] and Svob and Deu-Siftar [3] have studied different series of dialkylbenzenes. The Kov~ts retention indices and some physical properties of the hydrocarbons examined here are listed in Table I. The elution order follows those of boiling points, densities

Originals

757

!

I b5

155

9

97

97

@

9 112

112

69

69 83

!

!

Lll

I

II "

[ L ..... L~ 1 68

I

81

@

68

95

J

!

95

81 5b II0

I

55

.l_l

I

9 I

I

IL

68

i. i

110

I

i

d Jl.

J

jJ ii

i 68 I

95

i i

T

~5

r

S

~

F

55

110

I10 81 ~b

]l

I[|

I ,LI,

II

I!

!

F

I[ ,LI !

I

I.li,

,,I

1

Fig. 2 9 Mass spectra of dimethylcyclohexane and cyclohexenes. Encircled numbers are referred to compounds listed in Table I.

and refractive indices, because retention is determined by the vapor pressure and the molecular polarity of the solutes. Thus, on non-polar stationary phases, the correlation between retention indices (IK) and boiling points (Tv) displays linear relationships, the higher deviations being obviously exhibited by the most polar solutes, namely by the aromatic hydrocarbons. Coefficients a and b from the equation IK = a + b 9Tv are indicated in Table III. As shown, the term b increases with the column polarity and the correlations are improved when the three hydrocarbon families (alkanes, alkenes and aromatics) are considered separately. The most accurate correlation was obtained on squalane (Sq) which has enabled us to estimate the unreported boiling points of some hydrocarbons, as indicated in Table I.

758

On the other hand, the values obtained for the coefficient Kn in the second rule of Kovfits (6 I~: = K n 6 T v ) were 3.66 (Sq), 3.57 (SE-30) and 3.58 (ApL) and the mean absolute error in the prediction of retention indices on squalane, according to this relationship, was 6.5 i.u. The variation of I K with column temperature (6I/6T)increases from 0.20-0.30 (Sc0 to 0.31-0.52 (PEG 20M). Within the olefin series the higher increments correspond to exocyclic isomers (compounds Nos. 9, 16, 31) and the lower to the trisubstituted ones (compounds Nos. 8, 15, 20, 29, 30). On non-polar phases there is a slight increase of 8I/8T with increasing alkyl substitution, whilst on polar stationary phases the situation is reversed. The structural increments (ZXI,6H and 6I) for the present series of compounds are given in Tables I and V and Fig. 4.

Chromatographia Vol. 13 No. 12, December 1980

Originals

1

81

5,:

I

t

8J

Fss

|

| 67

67

,L, J 81 L._..,

|

|

67

109. 109

67

112t,

81

81

@

@

67 i09

95

~12,~

Fig. 3

~lass spectra of isopropylcyclohexane and cyclohexenes. Encircled numbers are referred to compounds listed in Table I.

As stated before, retention indices increase with the polarity of the stationary phase. The A1 (PEG-Sq) values exhibited by the cyclohexanes are about the same as those displayed by acyclic olefins with the same number of carbon atoms (50-70 i.u.) [ 2 ] . Notably, cycloalkenes and aromatics show AI values which are respectively close to two and four unsaturations (100-140 and 260-300, respectively). Schomburg and Dielmann [15]have previously reported that cyclopropanation of alkenes did not modify appreciably the polarity of the molecule. The AI values also depend on the number, type and position of the substituents. Thus, they decrease within each family of compounds with increasing relative molecular mass, more markedly in the aromatic series (314-261 i.u.) and for the disubstituted derivatives, because of the decrease of the relative importance of the polar moieties of the molecule. Chromatographia Vot. 13 No. 12, December 1980

Sterically hindered double bonds in cycloalkenes (for example isopropyl substituted ones like compounds 20 and 29) also display lower AI values, whereas the higher ones are exhibited by 1,1- and 4-substituted cycloalkenes (compounds 7, 9, 16, 21, 23, 31 and 33), whose double bonds are more accessible to the stationary phase. Hence, these are the most strongly retained in argentation chromatography [16] where the retention mechanism depends on double bond complexation. The resolution of the cis/trans isomers increases with the polarity of the stationary phase (see Table IV), the trans isomers being the first to elute, as in other isomeric series [17]. The trans isomers also show lower AI values compared with their corresponding cis counterparts, in agreement with the lower polarity of the former.

Originals

759

Table II I. Correlation of retention indices (I K) with boiling (T v ). TeeI = 80 ~

Confidence limits at (~ = 0.05

IK=a-Tv+b

Squatane SE-30 Apiezon L

,

51

a

b

cc

cc(sat)

e

3.73 + 0.09 3.67 + 0.11 3 67 + 0.09

351.4 + 13.2 361.8 + 15.5 383.2 + 13.4

0.998 0.997 0.997

0.999 0.998 0.998

0.7 0.8 0.8

Sq

1,4-Dimethylcyclohexane

20.0

3,6-Dimethylcyclohexene

0V-225

PEG 20M

24.5

36.2

37.0

SE-30

5.7

11.5

12.9

13.3

1-Isopropyl4-methylcyclohexane

10.8

10.5

2.1

10.0

3-1sopropyl6-methylcyclohexene

1.9

6.9

8.9

7.5

15.1

19.3

27.1

47.2

1-I sopropenyl4-methylcyclohexane

250

3.66 + 0.10 0.8 3.57 -+ 0.15 1.0 3.58+-0.10 . 0.8

The presence of alkyl substituents in the ~ position seems to have no specific influence on retention, because the 6Hsq remains practically constant within the corresponding series (7.6-7.8 for compounds 3, 7 and 21;34.5-36.0 for compounds 8, 15 and 30). The correlation is more diffi. cult in the case of a-substituted olefins, however, 6H increases regulary with the alkyl substitution in positions: 3<1, 1<4<1. The calculation of ~Hsq for the dialkyl cyclohexenes and aromatics studied according the following procedure:

~,liPr

'N 200

e

As shown before, the alkyl substitution on the double bonds has a remarkable influence on retention. Thus, triand tetrasubstituted cycloolefins (compounds 8, 15, 22, 30 and 32) exhibit the highest retention increments (SHsq: 29.2-36.0), with the exception of the isopropyl derivatives (-6.7, -7.8). On the other hand, the values displayed by disubstituted olefins (compounds 3, 6, 7, 9, 14, 16, 19, 21, 23, 28, 31, 33 and 34) are somewhat lower (-5.6 to 9.4) whilst the cis-3,6-dialkyl isomers exhibit the lowest ones (-16.4, -16.5).

,,

.

0.6 0.7 0.7

Kn

The contribution of double bonds and alkyl substituents to the retention index can be better defined by the structural increments ~ H (cyclene_cyetane) and ~ I (alkylcyclane.cyclane), which are indicated in Table V and Fig. 4, respectively. These values may be used in analytical practice for structural elucidation of related compounds, according to the additivity of the individual increments corresponding to the various adhering zones on the molecule [ 18].

Table IV. Retention differences between cis and trans isomers Alcis.tran s

-e(sat)

Kn-ST V

\

H p-eymene

,/'

= 6 Htoluene + ~ Hcumene - ~ nbenzene

.

afforded a mean absolute error of 2.8 units. 100

,/

(./~ AI Me

%

50

-,~,

\

t

J \

i

\

ooooo

99999@

I

F 2R22 RF

Fig. 4 9 Structural increments for methyl and isopropyl groups. + Squalane, 9 Carbowax 20M. 760

Finally, the structural increments 81 shown in Fig. 4, illustrate a close similarity between the three series of compounds examined. The Alsq values for methyl and is0propyl substituted aromatics fall between 103.1-107.1 and 253.4-257.4 i.u., respectively. The methyl contribution in 1-methylcyclohexenes is 86.0-90.5 i.u., and lower (50-70 i.u.) when it is located elsewhere in the ring. In cyclohexane derivatives, axial methyl substituents display 51 values (68.8-80.0) about 1 0 - 2 0 i.u. higher than the equatorial ones (57.8-60.0). The 6Isq values exhibited by a methylene group are 67.2-69.4 i.u., whereas the isopropylydene group contribution is 286.1-287.2 i.u.. On the other hand, the isopropenyl substitution represents a similar contribution to that of the isopropyl group (248.4-252.4).

Chromatographia Vol. 13 No. 12, December 1980

Originals

Q

%

1 I

++.

IL +

i

~

!

+J

t-i.:

~t +__8__-----: ! t

i "

:

i ~ . . . . . . . . . .

!~ t

'

-

I

+

|

t+++,

+

9

__2 .

.

~_++_--+++_~,

.

.

~

.

.

1

Fig. 5 Gas chromatogram of a pyrolysis naphtha. Column: Squalane (100 m x 0.25 mm), temperature programmed from 20 to 90 ~ at 0.75 ~ Numbered peaks correspond to those quoted in the Gallegos paper [20]. Chromatographia Vol. 13 No. 12, December 1980

Originals

-

761

Analysis of a Pyrolysis Naphtha

Table V. 6H(sat.) (80 ~ Sq. 2 3 5 6 7 8 9 12 13 14 15 16 18 19 20 21 22 23 26 27 28 29 30 31 32 33 34

Benzene Cyclohexene Toluene 3-Methylcyclohexene 4-Methylcyclohexene 1-Methylcyclohexene Methylenecyclohexane p-Xylene

PEG 20M

--25.2 7.8 18.1 3.8 7.6 34.5 5.6 62.3 cis-3,6-Dimethylcyclohexene --16.4 trans-3,6-Dimethylcyclohexene --2.1 1,4-Dimethylcyclohexene 34.8 4-Methyl methylenecyclohexane 8.8 Cumene -25.8 3-1sopropylcyclohexene O.0 1-Isopropylcyclohexene --6.7 4-1sopropylcyctohexene 7.8 Isopropylidenecyclohexane 29.2 Isopropenylcyctohexane --5.6 p-Cymene 19.5 cis-3.1sopropyl-6-methylcyclohexene --16.5 trans-3-1sopropyl-6-methylcyclohexene --7.6 1-Isopropyl-4-methylcyclohexene --7.8 4-1sopropyl-1 -methylcyclohexene 36,0 1-Isopropyl-4-methylenecyclohexane 9,4 1-Isopropylidene-4-methylcyctohexane 34,1 cis-1-Isopropenyl-4-methylcyclohexane 0.7 trans-l-lsopropenyt-4-methytcyclohexane -3.6

216.3 80.7 225.2 77.9 88.1 113.9 83.5 307.8 56,1 79.8 116.0 95.8 179.6 55.5 45.0 72.4 80.0 63.2 215.1 19.0 21.3 26.0 95.2 66.3 93.1 78.3 40.9

The availability of chromatographic and spectrometric data on the above mentioned hydrocarbons prompted us to investigate them in some industrial products. In this regard it is well known that pyrolysis naphtha contains large quantities of olefins and aromatics formed during the pyrolysis of the feed naphtha to produce ethylene. Using HRGC-MS and selective ion monitoring, Gallegos et al [19] were able to identify most o f the 152 individual components into which the pyronaphtha was resolved. However, several compounds remained unknown, whose mass fragment0grams indicated their molecular type to be olefins or cycl0olefins (2; = 0 or -2). From the mass spectral analysis and reference retention times we have obtained positive identification of some of them. The gas chromatogram of the analyzed naphtha is shown in Fig. 5, where peaks are numbered in the same order as that given by Gallegos. As was expected the major components are aromatic hydrocarbons. Among the large number of minor peaks, some have been identified as alkylcyclohexane and cyclohexene derivatives which are listed in Table VI. In fact, five of these hydrocarbons have already been assigned by Gallegos et al. to the series Z = 0 or Z = - 2

R eferences

Table Vl. Hydrocarbons identified in pyrolysis naphtha (Fig. 5). (a) Z-2 from Gallegos, (b) Z-0 from Gallegos and (c) No identified by Gallegos No (Fig. 5) 2 1 3 4 6 7 5 8

Benzene Cyclohexane Cyclohexene Methylcyclohexane 3-Methylcyclohexene 4-Methylcyclohexene Toluene 1-Methylcyclohexene

11 trans-l,4-Dimethylcyclohexane 10 cis-l,4-Dimethylcyclohexane 15 1,4-Dimethylcyclohexene 12 p-Xylene 18 Cumene 20 1-1sopropylcyclohexene 17 Isopropylcyclohexane 21 4-1sopropylcyclohexene 22 Isopropylidenecyclohexane

42 49 51 71 73 74 78 80 90 97 100 108 122 123 123 125 132

bis (c) (a)

(a) bis (c) bis (b)

(a)

bis (c) tris (c)

(a) bis (c)

[1] A. Wherli, E. Kovdts, Helv. Chirn. Acta 42, 2709 (1969). [2] S. Rang, A. Orav, K. Kuningas, O. Eissen, Chromatographia 10, 55 and 115 (1977) and references therein. [3] V. Svob, D. Deu-Siftar, J. Chromatogr. 91,677 (1974). [4] L. So]ak, Z A. Ri/ks, J. Chromatogr. 119, 505 (1976). [5] W. Engewald, L. Wennrich, Chromatographia 9, 540 (1976). [6] N.H. Inhoffen, D. Kampe, W. MolkowskL Liebig Ann. Chem. 674, 28 (1964). [7] O.P. Vig, A. L. Khurorna, K. L. Matta, J. Indian Chem. Soc. 45,615 (1968). [8] A. K. Bose, N. G. Steinberg, Synthesis, 595 (1970). [9] R . A . Benkeser, E. M. Kaiser, J. Am. Chem. Soc. 85, 2558 (1963) and references therein. [i0] X. Guardino, J. Albaig~s, G. Firpo, M. Gassiot, J, Rodriguez. Vihals, J. Chromatogr. 118, 13 (1976). [11] A . F . Thomas, B. Willharn,, Helv. Chim. Acta 47,475 (1964). [12] D.S. Weinberg, C. D]erassi, J. Org. Chem. 31,115 (1966). [ 13 ] G. Von Biinau, G. Schade, K. Gollnik, Z. Anal. Chem. 244, 7 (1969). [14] K.K. Mayer, C. D]erassi, Org. Mass Spectr. 5,817 (1971). [15] G. Schornburg, G. Dielman, Anal. Chem. 45, 1647 (1973). [ 16 ] aT.Herling, s ShabtaL E. GilA v, J. Chromatogr. 8,349 (1962). [ 17] J. C. Pascal M. Heintz, A. Druihle, D. Lefort, Chromatographia 7, 236 (1974). [18] G. Schornburg, J. Chromatogr. 23, 1 (1966). [19] E. s Gallegos, L M. Whitmore, R. F. Klaser, Anal. Chem. 46, 157 (1974). Received: June 24, 1980 Accepted: July 17, 1980 C

762

Chromatographia Vol. 13 No. 12, December 1980

Originals