- Email: [email protected]
Gas chromatographic analysis of aliphatic alcohols, diols and ethers using aromatic polymers
JOURNAL
CI-iROM.
GAS AND J.
3579
CHROMATOGRAPHIC ETHERS
13. LlNDSAY
Department (lieccivecl
I45
OF CHROMATOGRAPHY
USING
SMITH
of Chemistry, April
5th,
AND
ANALYSIS AROMATIC
D.
University
OF ALIPHATIC
ALCOHOLS,
DIOLS
POLYMERS
J, WADDINGTON
of York
(Great 23vitai.n)
IgGS)
SURmlARY
The gas-solid separations of alcohols, dials and ethers on porous aromatic polymers are essentially the same as those using gas-liquid separation on CelitcApiezon L columns, although the retention times on the latter are much shorter. 2. For liquids with similar boiling points the order of elution is diol, alcohol, ether, which reflects their increasing solubility in the polymer or in Apiezon L. 3, Within a class of compounds there is a linear relationship between log tr and boiling point of the eluent. Small divergencies from linearity arise from changes in the structure of the allcyl sidechain. 4. Modifying PAR-I with polyethylene glycol has little effect on separation of alcohols and ethers, and the behaviour of the resulting column contrasts strongly with the Celite-PEG 1500 column. 5, Diols are separated on aromatic polymers (e.g. PAR-I) without decomposition. I.
INTRODUCTION
’
Saturated aliphatic alcohols are often separated by gas chromatography using a polar liquid phase such as a polyglycol, an ester or a nitrile derivative. However, HOLLIS et nl.l-3 have shown that porous polyaromatic beads can be used. For example, on an ethylvinylbenzene-divinylbenzene (EVB-DVB) column, a mixture of 16 C,-C, saturated alcohols gave 14 peaks, only mixtures of 2-pentanol and 3-pentanol and of z-methyl-I-butanol and isopentyl alcohol not separating. HOLLIS et al. suggested that, from elution data, the solubility of the compound in the polymer is the most important factor in determining the order of elution, and that boiling point (or volatility) is of less consequence. As an example, he showed that, on an EYB-DVB polymer, qz-butyl alcohol eluted before terE.-pentyl alcohol although the boiling point of the tertiary alcohol is lower. In order to investigate the interaction between alcohols and aromatic polymers during gas chromatographic elution of the alcohols, results obtained in this work are compared with results from similar experiments using Celite coated with non-polar and polar stationary phases. Moreover, some analyses of diols and ethers are described ,., in order to compare their behaviour with that of the alcohols. J. Chvomatog.,
36 (IgGS)
145-151
146
J, R. LINDSAY
SMITH,
D. J. WADDINGTON
ZXPERIMENT.4L
Ofierating conditions of the gas chzromntogma$h All results were determined isothermally. The retention tin1es for amines were determined on a Pye 104 gas chromatograph equipped with a flame ionisation detector coupled to an RE 511 (Goerz Servoscribe) recorder. In order to find the corrected retention time (&) of each compound, the time taken for the carrier gas (nitrogen) to pass through the column was measured using a thermal conductivity cell. Glass columns (4 mm I.D.) were used, the length being 1.6 m except for the column of Porapak Q which was 0.7 m long. The flow rate for nitrogen was 25 ml * nlin-I. In some experiments, the gas chromatograph was connected to an A.E.1, MS 12 mass spectrometer. Ma terinls Solids used for chromatography were PAR-I (So--120 mesh) (Hewlett-Packard), Porapak Q (100-120 mesh) (Waters Associates Inc.) and Celite (acid washed (A/W) 100-120 mesh) (W. G. Pye and Co. Ltd.). The liquid phases used were PEG 1500 (British Drug Houses Ltd.) and Apiezon L (Apiezon Products Ltd.). Alcohols, diols and ethers used were all commercially available.
A weighed sample of the coating material was dissolved in an excess of solvent, and a known weight of support material was added to the solution. The mixture was evaporated, on a rotary evaporator, under reduced pressure, and the column packing was subsequently dried in an oven at 100~ for several hours. When preparing columns using alkali-washed (Alk/W) (5 o/Ow/w potassium hydroxide) Celite, a weighed sample of potassium hydroxide was dissolved in methanol, and Celite (A/W) was added. The mixture was evaporated under reduced pressure. The following packing nlatcrials were used during this study: (1) PAR-I ; (11) Porapak Q ; (III) PAR-I + PEG 1500 (a) 5% w/w, (b) 10% w/w; (IV Celite (Alk/W) + Apiezon L (a) IO % w/w, (b) 20% w/w, (c) 30% w/w; (V) Mite (A/IV) + Apiezon L 20 yO w/w; (VI) Celite (A/IV) + PEG 1500 10% w/w. RESULTSAND
DISCUSSION
Se$aration of nli$hntic aZcohoLs and ethers by PAR-I, Porafiak Q and Celite-A’;biexom L (CoLztmns I, II a92d IV (a-c)) The adjusted retention times, Er, for aliphatic alcohols and etl1ers on PAR-I, Porapak Q and Celite-Apiezon L are given in Table I. The retention times on CeliteApiezon L are shorter than those on the polymers but in most respects the chromatographic properties of the polymers resemble those of conventional Celite-Apiezon L. Thus the order of elution is generally similar for the two polymers and Celite-Apiezon L, although there are instances where the retention times of a pair of alcohols or J, Cirromatog., 3G (1968)
145-151
__ -.
..-
. . -. ...-_.-.- . _.
. ..
12.5
. . .
4-o 8.4 5.3
. .. . . .
47-3’ 54.1 2-3 3-9 4.0 5-o 7-o 20.3
22.2
11.3 17-3
36-5 30-9 57-2 66.1
22.9
‘3.7
29-j
3.2 5 .2 4.5 8.4 9-4 ‘3.2 11.7 ‘7.” 19.4 18.3
1.6
0.6
_
::; 11.6 _.__ .
_
4.6 13.8
5.1
I.5
2.4 2.4 1-9 I-9
‘7.4 20.3 35.2 44-z
12.0
i:f 7.9 9.5 9.6 8.9
3.6 4.1 5-I 4.7 6.7 6.S
-.-.-.
J-7 2.2 3.6 4.3 j-4 6.1
...
3.0 3.5 5.3 5.3 5.1
3-7 22.2
. . ..
42-7
-7-d
2.3 3.0
. . . 142-g ‘2.3
35.3
27.5
1q.s
13.3
9.8 10.1
IO.9
Z.
10.3 64-3
.. .
37-o 144.9 1G6.3 1.9 3-5 4-j
1j.S 30.6
11.6 15.1
6-4 8.S 9.3 12.4 11.5
I.2
1.0 2.0
0.9
0.5
40.6 73-I 112.3 156.2 237.0 308.0 0.s I.2 1.0 I.9 2.3 s-9
j0.S
7-6 7.! 6.1 16.0 13.6 23.8 13-5 35.3 26.0 27.6 26.7 24.9 51.2
j-8 0-j
1.d
0.5
60 1.6
1.G
Cditc
60
Celite
60
Cditc
o-7 I-3 1.6
5-o j-0 7.5 8.5 16.4 21.0 67.2 82.2
3.4 3-O 3-6 3.7 6.0
2.2
I-5 I.8
I.4
o-3 0.4 o-5 0.7
1.2 1.s 2.0 2.2
I.6
60
cclitc
1.6
I50
PAR-I
1.1
17.0 31.4 39.3
14.1
Z:, j-9 6-7 7.8 7-6 7.8 9-5
3.6 4-o 4-o 5 .2
o-7 I-3 1-5 1.9 2.6
1.6
150
o-7
111 (a)
IjO
PAR-I
11
Il.2
S:: S.6 9-4 9.6 9.6
3.4 4-s 5.3 5.8 72
I-9 2.9
I-5
o-9
1.6
Column length (m)
Ethyl alcohol Isopropyl alcohol ted-Butyl alcohol ?z-Propyl alcohol sec.-Butyl alcohol lsobutyl alcohol tcrt.-Pentyl alcohol s-Butyl alcohol 2,2-~~ethvl-I-propanol 2-Pentaiol 3-Pentanol 2-Methyl-2-pentanol Isopentyl alcohol z-Methyl-I-butanol 4-Methyl-2-pentanol n-Pentyl alcohol n-Ethyl-I-butanol gz-Hexyl alcohol n-Octanol 2-Ethyl-I-hexanol Ethyl ether te&Butyl methyl ether Isopropyl ether ~J3Utyl methyl ether n-Propyl ether wButy1 ether
1
Porapak c!
(min) FOR ALIPHATIc
PAR-I
mm
Colunm No.
m_mmD.RmwrloN
TABLE
J. R. LINDSAY
14s
‘sMITH,
D. J. WADDINGTON
1.5
1.0
d-
5
0.5
50
50
Temperature
100
150
‘C
Fig. I. Plot of log &boiling point for aliphatic alcohols and ethers. (a) PAR-1 at 150~. (b) CelitcPrimary alcohols; I = methyl alcohol; 2 = ethyl alcohol; 3 = Apiezon L 20% w/w at GOO. (a) alcohol. ( 0) secondary 5 = qz-pentyl alcohol; 6 = diexyl +z-propyl alcohol ; 4 = n-butyl alcohol; (4)) Tertiary alcol~ols ; alcohol; 2 = sec.-butyl alcohol ; 3 = 2-pentanol. alcol~ols; I = isopropyl I = ted-butyl alcohol ; 2 = tevC.-pentyl alcohol; 3 = ~-methyl-a-pentar701. (3) Ethers; I = ethyl ether:2 = methyl led.-butyl ether; 3 = isopropyl ether; 4 = methyl wbutyl ether; 5 = wpropyl ether; 6 = wbutyl etlwr.
are reversed (for example, qz-butyl alcohol and tert.-pentyl alcohol; isopcntyl and z-methyl-2-pentanol; and Wt.-butyl methyl ether and isopropyl ether). The similarity between the columns is further shown by an examination of the linear plots for log tr against the boiling point of the eluent” (Fig. I, a and b). In general, the order of elution of alcohols from these columns parallels the order of their boiling points. The effect of structure on retention time is also illustrated in Fig. I, where separate lines can be drawn for primary [Me* (CH,)n’ OH], secondary [Me. (CH,), l CHMe*OHJ and tertiary [Me. (CM2),~ CMe, OH] alcohols. Furthermore, primary alcohols of varying structure [Me (CH,) 7L- OH and Me * (CH,) n - CHEt - CH 4 OH] lie on individual lines (calculated from Table I). This is also seen amongst secondary alcohols [Me* (CH,), CHMe - OH and {Mea (CH,)n},CH *OH]. Thus, as seen ,in Table I, a primary alcohol may be eluted before a tertiary alcohol although the former has a higher boiling point. This effect is even more apparent on Porapak Q where the lines are wider apart than on PAR-1 and on Celite-Apiezon L columns (calculated from Table I). rt-Alkyl aliphatic etliers also have a linear relationship between log tr and their boiling point (Fig. I). However, the line for ethers is considerably displaced from those of the ‘alcohols indicating that replacement of the alcoholic hydrogen by an alkyl group has a much more drastic effect than similar changes in the alkyl side-chain. The longer retention times of ethers relative to alcohols can be attributed to the difference ethers
alcohol
l
l
l
l
J. Clwomato&?., 36 (196s)
145-1.51
l
GAS CHROMATOGRAPWY
OF ALIPWATIC
ALCOHOLS,
DIOLS AND ETHERS
149
in solubility of each in the non-polar polymer or Apiezon L, for an ether will be more soluble in a non-polar solvent than an alcohol with a comparable boiling point. Increasing the concentration of Apiezon L increases the retention times of the alcol~ols and ethers, but does not affect the isodromos point relative to the boiling points, for normal alcohols and ethers (calculated from Table I), The effect of coding PAR-r with PEG rgoo (Cohnn I and. III (a> anti (b)), ad a com$arison of PAR-r-PEG 1300 am-l Celite-PEG r5oo (Column VI> Retention times for alcohols on PAR-r are reduced on adding 5 oh w/w PEG 1500 and are then increased on adding a further 5 o/ow/w of the liquid (Table I). A similar effect is noted for retention times of amines on polymers coated with Apiezon L and with Carbowax 20 M 6. Such a finding has been attributed to competition between the solid and liquid for the eluent. As the solid becomes progressively more covered with liquid phase, the role played by the solid becomes less important. Since log &-boiling point plots for PAR-I-PEG 1500 columns bear a closk resemblance to those obtained with PAR-r, it is clear that up to a loading of 10% w/w polyethylene glycol, separation is still due principally to the polymer, and not to the liquid phase (Figs. Ia and 2a). The PEG r500 coating modifies PAR-I sufficiently to move the plot for ethers closer to that for normal alcohols and as a result the isodromos point, relative to the boiling point, is lowered. Modification of the polymers with a liquid phase may prove useful in the separation of alcohols having similar boiling points but with widely different structures. The chromatographic properties of alcohols and ethers on PAR-I-PEG 150o contrast markedly with those on Celite-PEG 1500 (Table I and Fig. 2), for the polar liquid phase on the latter ensures that the alcohols are held back with respect to the 1.. 5
1 *I
-
o-
3
F 0.:
i i -/ Pl
0.0,
C
SO
Temperature
100
‘C
150
,
Fig. 2. Plot of log &-boiling point for aliphatic alcohols and ethers. (a) PAR-I-PEG at 150’. (b) Celite-PEG 1500 10% w/w at 60~. ,
rgoo
J. Clrromatog., 36 (1968)
10%
w/w
145-151
, J. R. LINDSAY
150
SMITEI, D., J. WADDINGTON
1.:
I’ , -, 100
110
100
1.1 I
110 Temperature
110
100
110
'C
Fig. ,3, Plot of +boiling point alcohols. l?AR:r at ~50~. 1500 ~0% w/w at 60~. 20% w/w 60~. (c) rgoo 10% at 150~. (cl)Ccl$x-PEG (0) Primary alcohols; (0) secondary alcohols: (a) tertiary alcohols. I = n-Propyl alcohol: 2 = = a,a-methyl-I-propanol. sec.-bukyl alcohol; 3 = terC.-perky1 alcohol; 4 = isobutyl alcohol; ,,
ethers. Vurther among the alcohols the primary compounds are held back most and the tertiary, .least. A particularly striking .difference between the Celite-PEG 1500 column, on the one hand, and the coated and uncoated polymers and Celite-Apiezon columns on the other, is seen on comparing the retention times, and hence the log Clboiling point plots, for the alcohols [Me 9CR, OH and CR2H CH,* OH] (Fig. 3). l
l
Se$aration of dials on PAR-r (Cohmn I), Pora#ak Q (Cohmn II), Celite (ALk/ W)A+zon L 20 % w/w (Column IV b) , CeLite (A/ W)-A$ iezon L 20 o/o w/w (Column V) and Celite (A/ W)-PEG 1500 IO y0 w/w (Cohmn VI) The peaks obtained for diols on the Celite columns are grossly distorted whether the liquid phase is polar or non-polar or the support is acid- or alkali-washed. In extreme instances, glycols give double peaks. That this peak asymmetry arises in part from decomposition of the diol was confirmed by mass spectrometric analysis (for example x,+butanediol decomposes to tetrahydrofuran) , Retention times for the diols were irreproducible on the Celite columns. Previous investigations into the gas chromatographic separation of diols has concentrated on eliminating this peak tailing by reducing the interaction between the diols and the solid support. This has been attained in two ways, either by using a highly polar liquid phase such as tetrahydroxyethylethylenediamine7-0, water lo ox a combination of .both 7~10, or by deactivat,ing the support by silanisationlO. In contrast to our results with conventional gas-liquid columns, there appears to ,be little or no decomposition of diols on Porapak Q, although there is still some tailing. Similar results were obtained for ethylene glycol and 8-methylpentane-2,41. ChW?2atO&,
36 (1968) 145-151
GAS CHROMATOGRAPHY TABLE
OF .AI+IPHATIC ALCOHOLS, PIOLS AND ETHERS
x51
II
ADJUSTED RETENTION TIMES (I&l) FOR DIFPERENT TEMPERATURES
ALIPHATIC
DIOLS
ON
PAR-I
AND
ON PORAPAIC
Poragak
PAR-r 15O0
Ethylene glycol Propane-r,a-diol Butane-a.3-diol Propane-x,3-diol Butane-x ,3-diol Butane-r .4-diol
8.1
I8*o
",
65::
x8.6
8.9
21.0
9.9
2.9
33:; 5.6 6.2
X5.4
9.8
Q
:
185’
2000
4.7
... 11.6
36.8
Q AT
6.0 9.0 r4.0 x4.3 O-2.0 32*5
separation came from diol on Porapak S 11, We found that the most satisfactory PAR-I, the peaks being symmetrical and there being no evidence of decomposition. Moreover, butane-2,3-diol is separated more satisfactorily from propane-x,3-diol. Retention times of a series of diols on PAR-r and on Porapak Q are given in Table II. Linear plots for log &-boiling point are obtained for diols containing two primary alcohol groups, and for diols with one primary and.one secondary group. The separation between these lines is wider than found for aliphatic alcohols, showing that changes of structure in the alkyl side chain have a greater effect than for alcohols. Thus ethylene glycol is eluted before I:&propanediol and x,3-propanediol before x,3-butanediol, although the boiling points of these pairs of diols are ,reversed. Similar-’ ly, the presence of two secondary groups, as in z,3butanediol, kau.ses an eimi greater ap$arent disparity between boiling point and retention, time. ”
ACKNOWLEDGEMENTS
We thank Dr. C. B. THOMAS for the mass spectrometric for most helpful technical assistance.
analyses and Miss M. A.
WARRISS
REFERENCES I 0. L. HOLLIS, Ana2. Ckem., 38 (2966) 3bg. L, HOLLIS AND W. V. HAYES, J. Gas CAromaCog., 4 (1966) 23.5. 3 O.L. HOLI.ISANDW.V.HAYRS, inA.B. LITTLEWOOD (Editor),Gas Chromatograpiry
2 0.
4 5 6 7 8
2966, The Institute of Petroleum, London, 1967, p. 57. Dictionary of Organic Compounds, Eyre and Spottiswoode, and E. and F. N. Spon, Bonbon, xg6jj. J, R. LINDSAY SMITH AND D.J.WADDINGTON, AnaLChem., 40 (x968) 52% J, R. LINDSAY SMITH AND D. J.~ADDINGTON, University of.York, unpublished results,rgb?. D. M. OTTBNSTEIN, Seventh Anackem. Confevence, Dehoit, rgsg. 1-I. G. NADEAU AND D. M. OAKS, An&Z. Clrem;,32 (1960) 1760. B. A. SWINEHART, Anal. Ckem., 40 (1968) 427.
g 'IoL. H.PIIIBERAND H.IC.PLUMMER,A~~~.C~~~~., II J. F.~?AL.BRAMANAND E.&WALKER, AlzaZyst,gl
38 (1966)x652.
(x967)535. ’
J. Ch-vomatog., 36 (1968) x45-15x
CI-iROM.
GAS AND J.
3579
CHROMATOGRAPHIC ETHERS
13. LlNDSAY
Department (lieccivecl
I45
OF CHROMATOGRAPHY
USING
SMITH
of Chemistry, April
5th,
AND
ANALYSIS AROMATIC
D.
University
OF ALIPHATIC
ALCOHOLS,
DIOLS
POLYMERS
J, WADDINGTON
of York
(Great 23vitai.n)
IgGS)
SURmlARY
The gas-solid separations of alcohols, dials and ethers on porous aromatic polymers are essentially the same as those using gas-liquid separation on CelitcApiezon L columns, although the retention times on the latter are much shorter. 2. For liquids with similar boiling points the order of elution is diol, alcohol, ether, which reflects their increasing solubility in the polymer or in Apiezon L. 3, Within a class of compounds there is a linear relationship between log tr and boiling point of the eluent. Small divergencies from linearity arise from changes in the structure of the allcyl sidechain. 4. Modifying PAR-I with polyethylene glycol has little effect on separation of alcohols and ethers, and the behaviour of the resulting column contrasts strongly with the Celite-PEG 1500 column. 5, Diols are separated on aromatic polymers (e.g. PAR-I) without decomposition. I.
INTRODUCTION
’
Saturated aliphatic alcohols are often separated by gas chromatography using a polar liquid phase such as a polyglycol, an ester or a nitrile derivative. However, HOLLIS et nl.l-3 have shown that porous polyaromatic beads can be used. For example, on an ethylvinylbenzene-divinylbenzene (EVB-DVB) column, a mixture of 16 C,-C, saturated alcohols gave 14 peaks, only mixtures of 2-pentanol and 3-pentanol and of z-methyl-I-butanol and isopentyl alcohol not separating. HOLLIS et al. suggested that, from elution data, the solubility of the compound in the polymer is the most important factor in determining the order of elution, and that boiling point (or volatility) is of less consequence. As an example, he showed that, on an EYB-DVB polymer, qz-butyl alcohol eluted before terE.-pentyl alcohol although the boiling point of the tertiary alcohol is lower. In order to investigate the interaction between alcohols and aromatic polymers during gas chromatographic elution of the alcohols, results obtained in this work are compared with results from similar experiments using Celite coated with non-polar and polar stationary phases. Moreover, some analyses of diols and ethers are described ,., in order to compare their behaviour with that of the alcohols. J. Chvomatog.,
36 (IgGS)
145-151
146
J, R. LINDSAY
SMITH,
D. J. WADDINGTON
ZXPERIMENT.4L
Ofierating conditions of the gas chzromntogma$h All results were determined isothermally. The retention tin1es for amines were determined on a Pye 104 gas chromatograph equipped with a flame ionisation detector coupled to an RE 511 (Goerz Servoscribe) recorder. In order to find the corrected retention time (&) of each compound, the time taken for the carrier gas (nitrogen) to pass through the column was measured using a thermal conductivity cell. Glass columns (4 mm I.D.) were used, the length being 1.6 m except for the column of Porapak Q which was 0.7 m long. The flow rate for nitrogen was 25 ml * nlin-I. In some experiments, the gas chromatograph was connected to an A.E.1, MS 12 mass spectrometer. Ma terinls Solids used for chromatography were PAR-I (So--120 mesh) (Hewlett-Packard), Porapak Q (100-120 mesh) (Waters Associates Inc.) and Celite (acid washed (A/W) 100-120 mesh) (W. G. Pye and Co. Ltd.). The liquid phases used were PEG 1500 (British Drug Houses Ltd.) and Apiezon L (Apiezon Products Ltd.). Alcohols, diols and ethers used were all commercially available.
A weighed sample of the coating material was dissolved in an excess of solvent, and a known weight of support material was added to the solution. The mixture was evaporated, on a rotary evaporator, under reduced pressure, and the column packing was subsequently dried in an oven at 100~ for several hours. When preparing columns using alkali-washed (Alk/W) (5 o/Ow/w potassium hydroxide) Celite, a weighed sample of potassium hydroxide was dissolved in methanol, and Celite (A/W) was added. The mixture was evaporated under reduced pressure. The following packing nlatcrials were used during this study: (1) PAR-I ; (11) Porapak Q ; (III) PAR-I + PEG 1500 (a) 5% w/w, (b) 10% w/w; (IV Celite (Alk/W) + Apiezon L (a) IO % w/w, (b) 20% w/w, (c) 30% w/w; (V) Mite (A/IV) + Apiezon L 20 yO w/w; (VI) Celite (A/IV) + PEG 1500 10% w/w. RESULTSAND
DISCUSSION
Se$aration of nli$hntic aZcohoLs and ethers by PAR-I, Porafiak Q and Celite-A’;biexom L (CoLztmns I, II a92d IV (a-c)) The adjusted retention times, Er, for aliphatic alcohols and etl1ers on PAR-I, Porapak Q and Celite-Apiezon L are given in Table I. The retention times on CeliteApiezon L are shorter than those on the polymers but in most respects the chromatographic properties of the polymers resemble those of conventional Celite-Apiezon L. Thus the order of elution is generally similar for the two polymers and Celite-Apiezon L, although there are instances where the retention times of a pair of alcohols or J, Cirromatog., 3G (1968)
145-151
__ -.
..-
. . -. ...-_.-.- . _.
. ..
12.5
. . .
4-o 8.4 5.3
. .. . . .
47-3’ 54.1 2-3 3-9 4.0 5-o 7-o 20.3
22.2
11.3 17-3
36-5 30-9 57-2 66.1
22.9
‘3.7
29-j
3.2 5 .2 4.5 8.4 9-4 ‘3.2 11.7 ‘7.” 19.4 18.3
1.6
0.6
_
::; 11.6 _.__ .
_
4.6 13.8
5.1
I.5
2.4 2.4 1-9 I-9
‘7.4 20.3 35.2 44-z
12.0
i:f 7.9 9.5 9.6 8.9
3.6 4.1 5-I 4.7 6.7 6.S
-.-.-.
J-7 2.2 3.6 4.3 j-4 6.1
...
3.0 3.5 5.3 5.3 5.1
3-7 22.2
. . ..
42-7
-7-d
2.3 3.0
. . . 142-g ‘2.3
35.3
27.5
1q.s
13.3
9.8 10.1
IO.9
Z.
10.3 64-3
.. .
37-o 144.9 1G6.3 1.9 3-5 4-j
1j.S 30.6
11.6 15.1
6-4 8.S 9.3 12.4 11.5
I.2
1.0 2.0
0.9
0.5
40.6 73-I 112.3 156.2 237.0 308.0 0.s I.2 1.0 I.9 2.3 s-9
j0.S
7-6 7.! 6.1 16.0 13.6 23.8 13-5 35.3 26.0 27.6 26.7 24.9 51.2
j-8 0-j
1.d
0.5
60 1.6
1.G
Cditc
60
Celite
60
Cditc
o-7 I-3 1.6
5-o j-0 7.5 8.5 16.4 21.0 67.2 82.2
3.4 3-O 3-6 3.7 6.0
2.2
I-5 I.8
I.4
o-3 0.4 o-5 0.7
1.2 1.s 2.0 2.2
I.6
60
cclitc
1.6
I50
PAR-I
1.1
17.0 31.4 39.3
14.1
Z:, j-9 6-7 7.8 7-6 7.8 9-5
3.6 4-o 4-o 5 .2
o-7 I-3 1-5 1.9 2.6
1.6
150
o-7
111 (a)
IjO
PAR-I
11
Il.2
S:: S.6 9-4 9.6 9.6
3.4 4-s 5.3 5.8 72
I-9 2.9
I-5
o-9
1.6
Column length (m)
Ethyl alcohol Isopropyl alcohol ted-Butyl alcohol ?z-Propyl alcohol sec.-Butyl alcohol lsobutyl alcohol tcrt.-Pentyl alcohol s-Butyl alcohol 2,2-~~ethvl-I-propanol 2-Pentaiol 3-Pentanol 2-Methyl-2-pentanol Isopentyl alcohol z-Methyl-I-butanol 4-Methyl-2-pentanol n-Pentyl alcohol n-Ethyl-I-butanol gz-Hexyl alcohol n-Octanol 2-Ethyl-I-hexanol Ethyl ether te&Butyl methyl ether Isopropyl ether ~J3Utyl methyl ether n-Propyl ether wButy1 ether
1
Porapak c!
(min) FOR ALIPHATIc
PAR-I
mm
Colunm No.
m_mmD.RmwrloN
TABLE
J. R. LINDSAY
14s
‘sMITH,
D. J. WADDINGTON
1.5
1.0
d-
5
0.5
50
50
Temperature
100
150
‘C
Fig. I. Plot of log &boiling point for aliphatic alcohols and ethers. (a) PAR-1 at 150~. (b) CelitcPrimary alcohols; I = methyl alcohol; 2 = ethyl alcohol; 3 = Apiezon L 20% w/w at GOO. (a) alcohol. ( 0) secondary 5 = qz-pentyl alcohol; 6 = diexyl +z-propyl alcohol ; 4 = n-butyl alcohol; (4)) Tertiary alcol~ols ; alcohol; 2 = sec.-butyl alcohol ; 3 = 2-pentanol. alcol~ols; I = isopropyl I = ted-butyl alcohol ; 2 = tevC.-pentyl alcohol; 3 = ~-methyl-a-pentar701. (3) Ethers; I = ethyl ether:2 = methyl led.-butyl ether; 3 = isopropyl ether; 4 = methyl wbutyl ether; 5 = wpropyl ether; 6 = wbutyl etlwr.
are reversed (for example, qz-butyl alcohol and tert.-pentyl alcohol; isopcntyl and z-methyl-2-pentanol; and Wt.-butyl methyl ether and isopropyl ether). The similarity between the columns is further shown by an examination of the linear plots for log tr against the boiling point of the eluent” (Fig. I, a and b). In general, the order of elution of alcohols from these columns parallels the order of their boiling points. The effect of structure on retention time is also illustrated in Fig. I, where separate lines can be drawn for primary [Me* (CH,)n’ OH], secondary [Me. (CH,), l CHMe*OHJ and tertiary [Me. (CM2),~ CMe, OH] alcohols. Furthermore, primary alcohols of varying structure [Me (CH,) 7L- OH and Me * (CH,) n - CHEt - CH 4 OH] lie on individual lines (calculated from Table I). This is also seen amongst secondary alcohols [Me* (CH,), CHMe - OH and {Mea (CH,)n},CH *OH]. Thus, as seen ,in Table I, a primary alcohol may be eluted before a tertiary alcohol although the former has a higher boiling point. This effect is even more apparent on Porapak Q where the lines are wider apart than on PAR-1 and on Celite-Apiezon L columns (calculated from Table I). rt-Alkyl aliphatic etliers also have a linear relationship between log tr and their boiling point (Fig. I). However, the line for ethers is considerably displaced from those of the ‘alcohols indicating that replacement of the alcoholic hydrogen by an alkyl group has a much more drastic effect than similar changes in the alkyl side-chain. The longer retention times of ethers relative to alcohols can be attributed to the difference ethers
alcohol
l
l
l
l
J. Clwomato&?., 36 (196s)
145-1.51
l
GAS CHROMATOGRAPWY
OF ALIPWATIC
ALCOHOLS,
DIOLS AND ETHERS
149
in solubility of each in the non-polar polymer or Apiezon L, for an ether will be more soluble in a non-polar solvent than an alcohol with a comparable boiling point. Increasing the concentration of Apiezon L increases the retention times of the alcol~ols and ethers, but does not affect the isodromos point relative to the boiling points, for normal alcohols and ethers (calculated from Table I), The effect of coding PAR-r with PEG rgoo (Cohnn I and. III (a> anti (b)), ad a com$arison of PAR-r-PEG 1300 am-l Celite-PEG r5oo (Column VI> Retention times for alcohols on PAR-r are reduced on adding 5 oh w/w PEG 1500 and are then increased on adding a further 5 o/ow/w of the liquid (Table I). A similar effect is noted for retention times of amines on polymers coated with Apiezon L and with Carbowax 20 M 6. Such a finding has been attributed to competition between the solid and liquid for the eluent. As the solid becomes progressively more covered with liquid phase, the role played by the solid becomes less important. Since log &-boiling point plots for PAR-I-PEG 1500 columns bear a closk resemblance to those obtained with PAR-r, it is clear that up to a loading of 10% w/w polyethylene glycol, separation is still due principally to the polymer, and not to the liquid phase (Figs. Ia and 2a). The PEG r500 coating modifies PAR-I sufficiently to move the plot for ethers closer to that for normal alcohols and as a result the isodromos point, relative to the boiling point, is lowered. Modification of the polymers with a liquid phase may prove useful in the separation of alcohols having similar boiling points but with widely different structures. The chromatographic properties of alcohols and ethers on PAR-I-PEG 150o contrast markedly with those on Celite-PEG 1500 (Table I and Fig. 2), for the polar liquid phase on the latter ensures that the alcohols are held back with respect to the 1.. 5
1 *I
-
o-
3
F 0.:
i i -/ Pl
0.0,
C
SO
Temperature
100
‘C
150
,
Fig. 2. Plot of log &-boiling point for aliphatic alcohols and ethers. (a) PAR-I-PEG at 150’. (b) Celite-PEG 1500 10% w/w at 60~. ,
rgoo
J. Clrromatog., 36 (1968)
10%
w/w
145-151
, J. R. LINDSAY
150
SMITEI, D., J. WADDINGTON
1.:
I’ , -, 100
110
100
1.1 I
110 Temperature
110
100
110
'C
Fig. ,3, Plot of +boiling point alcohols. l?AR:r at ~50~. 1500 ~0% w/w at 60~. 20% w/w 60~. (c) rgoo 10% at 150~. (cl)Ccl$x-PEG (0) Primary alcohols; (0) secondary alcohols: (a) tertiary alcohols. I = n-Propyl alcohol: 2 = = a,a-methyl-I-propanol. sec.-bukyl alcohol; 3 = terC.-perky1 alcohol; 4 = isobutyl alcohol; ,,
ethers. Vurther among the alcohols the primary compounds are held back most and the tertiary, .least. A particularly striking .difference between the Celite-PEG 1500 column, on the one hand, and the coated and uncoated polymers and Celite-Apiezon columns on the other, is seen on comparing the retention times, and hence the log Clboiling point plots, for the alcohols [Me 9CR, OH and CR2H CH,* OH] (Fig. 3). l
l
Se$aration of dials on PAR-r (Cohmn I), Pora#ak Q (Cohmn II), Celite (ALk/ W)A+zon L 20 % w/w (Column IV b) , CeLite (A/ W)-A$ iezon L 20 o/o w/w (Column V) and Celite (A/ W)-PEG 1500 IO y0 w/w (Cohmn VI) The peaks obtained for diols on the Celite columns are grossly distorted whether the liquid phase is polar or non-polar or the support is acid- or alkali-washed. In extreme instances, glycols give double peaks. That this peak asymmetry arises in part from decomposition of the diol was confirmed by mass spectrometric analysis (for example x,+butanediol decomposes to tetrahydrofuran) , Retention times for the diols were irreproducible on the Celite columns. Previous investigations into the gas chromatographic separation of diols has concentrated on eliminating this peak tailing by reducing the interaction between the diols and the solid support. This has been attained in two ways, either by using a highly polar liquid phase such as tetrahydroxyethylethylenediamine7-0, water lo ox a combination of .both 7~10, or by deactivat,ing the support by silanisationlO. In contrast to our results with conventional gas-liquid columns, there appears to ,be little or no decomposition of diols on Porapak Q, although there is still some tailing. Similar results were obtained for ethylene glycol and 8-methylpentane-2,41. ChW?2atO&,
36 (1968) 145-151
GAS CHROMATOGRAPHY TABLE
OF .AI+IPHATIC ALCOHOLS, PIOLS AND ETHERS
x51
II
ADJUSTED RETENTION TIMES (I&l) FOR DIFPERENT TEMPERATURES
ALIPHATIC
DIOLS
ON
PAR-I
AND
ON PORAPAIC
Poragak
PAR-r 15O0
Ethylene glycol Propane-r,a-diol Butane-a.3-diol Propane-x,3-diol Butane-x ,3-diol Butane-r .4-diol
8.1
I8*o
",
65::
x8.6
8.9
21.0
9.9
2.9
33:; 5.6 6.2
X5.4
9.8
Q
:
185’
2000
4.7
... 11.6
36.8
Q AT
6.0 9.0 r4.0 x4.3 O-2.0 32*5
separation came from diol on Porapak S 11, We found that the most satisfactory PAR-I, the peaks being symmetrical and there being no evidence of decomposition. Moreover, butane-2,3-diol is separated more satisfactorily from propane-x,3-diol. Retention times of a series of diols on PAR-r and on Porapak Q are given in Table II. Linear plots for log &-boiling point are obtained for diols containing two primary alcohol groups, and for diols with one primary and.one secondary group. The separation between these lines is wider than found for aliphatic alcohols, showing that changes of structure in the alkyl side chain have a greater effect than for alcohols. Thus ethylene glycol is eluted before I:&propanediol and x,3-propanediol before x,3-butanediol, although the boiling points of these pairs of diols are ,reversed. Similar-’ ly, the presence of two secondary groups, as in z,3butanediol, kau.ses an eimi greater ap$arent disparity between boiling point and retention, time. ”
ACKNOWLEDGEMENTS
We thank Dr. C. B. THOMAS for the mass spectrometric for most helpful technical assistance.
analyses and Miss M. A.
WARRISS
REFERENCES I 0. L. HOLLIS, Ana2. Ckem., 38 (2966) 3bg. L, HOLLIS AND W. V. HAYES, J. Gas CAromaCog., 4 (1966) 23.5. 3 O.L. HOLI.ISANDW.V.HAYRS, inA.B. LITTLEWOOD (Editor),Gas Chromatograpiry
2 0.
4 5 6 7 8
2966, The Institute of Petroleum, London, 1967, p. 57. Dictionary of Organic Compounds, Eyre and Spottiswoode, and E. and F. N. Spon, Bonbon, xg6jj. J, R. LINDSAY SMITH AND D.J.WADDINGTON, AnaLChem., 40 (x968) 52% J, R. LINDSAY SMITH AND D. J.~ADDINGTON, University of.York, unpublished results,rgb?. D. M. OTTBNSTEIN, Seventh Anackem. Confevence, Dehoit, rgsg. 1-I. G. NADEAU AND D. M. OAKS, An&Z. Clrem;,32 (1960) 1760. B. A. SWINEHART, Anal. Ckem., 40 (1968) 427.
g 'IoL. H.PIIIBERAND H.IC.PLUMMER,A~~~.C~~~~., II J. F.~?AL.BRAMANAND E.&WALKER, AlzaZyst,gl
38 (1966)x652.
(x967)535. ’
J. Ch-vomatog., 36 (1968) x45-15x