Determination of Olefins by Means of Iodine Complexes Ultraviolet Absorption Method DONALD R. LONG and RICHARD W. NEUZIL Universal O i l Products Co., Des Plaines, 111. This paper describes a procedure for distinguishing types of alkyl-substituted olefins by means of the ultraviolet spectra of reversible complexes formed with iodine. Yery intense bands are observed, as follows: RCH=CHa, 27.5 mp; R?C=CH,, 290 to 295 mp; RCH= CHR, 295 to 300 mp; R,C=CHR, 317 mp; and RaC= CR,, 337 mp. Cyclopentene and cyclohexene behave like the cis forni of the corresponding open-chain olefin RCH=CHR. The method is of particular interest for the determination of the tri- and tetrasubstituted olefins because these types are difficult to determine by infrared or Raman spectroscopy. Olefin (0.03 to 0.1 mole per liter) and iodine (0.002 mole per liter) are dissolved in iso-octane, and the ultraviolet spectrum of the complex is measured in a cell of 1-cm. path length. For these concentrations it is sholrn that a simple extension of the Beer’s law equation applies in which the measured absorbance is proportional to the product of the olefin and iodine concentrations. .ipparent molecular “absorptivities” can thus be calculated and are found to he reasonably constant for a given olefin type.
ultraviolet spectrometers. The method is of particular interest because the tri- and tetrasubstituted olefins can be distinguished. PROCEDURE
Summary of Method. The experimental procedure is coniparatively simple. Olefin (0.05 mole per liter) and iodine (0.002 mole per liter) are dissolved in purified iso-octane. The spectrum is then recorded from 400 to 250 mp with a Cary spectrometer (Model 11) and an absorption cell of 1-cm. path length.
30
A
S U J I B E R of workers ( 2 , 3 , 5 , 6) have shown that iodine forms reversible complexes with a variety of organic molecules such as ethers, alcohols, ketones, sulfides, olefins, diolefins, aromatics, and even paraffins. I n particular, the iodine-olefin complexes have a very intense band in the 275- to 350-mp region of the ultraviolet spectrum. It is shown here that mono-! di-, tri-, and tetrasubstituted ethylenes can be determined because the band for each type occurs a t a characteristic wave length (Figure 1). The most widely used method for olefin-type analysis depends on infrared absorption bands due to ethylenic hydrogen vibrations, which occur in the 800- t o 1000-em. region ( 1 , 9 ) . This method is good for t,he simpler olefin types, RCH=CH,, RZC=CH*, and trans-RCH=CHR. For these olefins, the characteristic band for each type falls within a very narrow absorption region, regardless of the nature of the R group. However, the trisubstituted olefins, R2C==CHR, have an absorption maximum which may occur over the whole 800- to 850-crn.-l region. Since the exact position of the absorption band for this olefin type depends strongly on the nature of the R groups, the analysis is unsatisfactory. Finally, the fully substituted olefin, R&=CR,, has no ethylenic hydrogens and therefore no infrared bands of this type are even available for analysis. Raman spectra can also be used to distinguish the various olefin types by means of the C=C stretching frequencies in 1650 to 1680-cm.-l region ( 4 ) . .4lthough all of the olefin types have characteristic Raman bands, the Irans-RCH=CHR, R?C=CHR, and RzC=CR2 types cannot he distinguished because their bands substantially overlap. .\nalysis for olefin types by ultraviolet spectroscopy is possible in the “vacuum region’’ beloil- 2000 A. (8). The simpler olefins have bands in this region, and increasing alkyl substitution shifts the absorption maxima toward longer wave lengths. Unfortunately, the experimental technique is difficult, and few laboratories are equipped with the necessary vacuum spectrometers. The iodine-olefin complex procedure described here is useful for all of the olefin types, and the work can be done with ordinary
20
IO
0 250
Figure 1.
300
350
400
Olefin-iodine complexes for various olefin types
Reagents. Iso-octane of 99% purity (Phillips pure grade) is used as the solvent. The iso-octane is purified by shakingit thoroughly with concentrated sulfuric acid. A liter of the acidwashed iso-octane is then percolated through a %foot column of 1-inch diameter filled with 200-mesh silica gel (Davison Chemical, Code 950). This treatment removes traces of impurities such as aromatics and dienes. The solvent should have an absorbance less than 0.05 down to 230 mp, when compared to distilled water in a cell of 1-cm. path length. The iodine reagent is a solution of iodine crystals in the purified iso-octane, at, a concentration of 1 gram per liter. Approsimately 0.1 gram of iodine, weighed to the nearest 0.1 mg., is dissolved and diluted to 100 ml. in the iso-octane. Calibration. API (American Petroleum Institute) standard samples of olefins are recommended for calibration. A dilution of the olefin is prepared at a known concentration of approximately 0.1 mole per liter in the puiified iso-octane. Five milliliters of the prepared olefin solution are then added to 5 ml. of the iodine reagent. The spectrum of the resulting solution is meas1110
V O L U M E 2 7 , NO. 7, J U L Y 1 9 5 5
1111
ured from 400 to 250 mb*,using a silica absorption cell of 1-cni. path length. The solution is measured immediately after mixing to avoid the catalytic effect of light on the formation of undesired diiodo compounds. To keep the absorbance of the comples betryeen 0.3 and 1.0, it may be necessary to vary the olefin concentration. If so, a second dilution of the olefin in the iso-octane is prepared. and then the iodine reagent is added. Mere dilution of the final iodine-olefin solution with solvent is not satisfactory, since the iodine concentration should not be varied. A correction for background due to uncomplesed iodine is obt,ained by adding 5 nil. of the iodine reagent to 5 ml. of iso-octane and then measuring the solution as described above. S o background correction for the uncomplexed olefin is made in the calibration procedure 1)eciLuse the pure olefins are transparent in this region. Measurements reported in this parier are made at 25” 5 1’ C. As the Car>- spectrometei. used does not warm the solutions appreciably and the AH of the iodine-olefin complexes is small [ca. -500 cal. per mole (GI!! no significant temperature effects are to he espected.
centrations and the method of espressing ahsorptivities, it is necessary to examine briefly the underlying theory of iodine complex formation. It is known that iodine and hydrocarbons undergo a reversible reaction to form 3.11 addition complex of an acid-base type ( 3 , 7 ) . For olefins the equation is: 11
The particular prolileni in the analytical application of iodine complexes is that the equilibrium constant, K , the true niolar absorptivity, E , of the complex, and the concentration of the complex cannot readily be determined with precision. Benesi and Hildebrand ( 3 )have described a mathematical procedure for deriving these quantities from absorbance measurements on a series of solutions. However, this is esacting and tedious work, and will be shown to be unnecessary for analytical applications. For analytical purposes it is essential that the absorbance measured be directly proportional to the concentration of olefin in the final Jolution. In order to choose concentrations of iodine and olefin to meet this requirement, an approsimat’e value for the equilibrium constant must be known. I t has been shown that the true molar absorptivity, e . of iodine complexes falls within the rather narrow range from 10,000 to 25,000 liter8 per mole-vm. for such diverse molecules as aromatic, cycloparaffin, and paraffin hydrocarbons, and even for an alkyl halide (.7. 7 ) . Therefore, if
Amax.. nl,u
275 290-29.5 295-300 317 337
RCH=CHR RsC=CHR RzC=CR*
The measured slmorl)ances are first corrected for the iodine background absorption. “hIolecular absorptivities” are then calculated as follows: where E‘
A,
= molecular absorptivity = absorbance a t A m < .
~
corrected for 11background (CI) = concentration of 1 2 in the final solution, it1 grams per liter [CO] = concentration of olefin in the final solution, i n moles per liter hlolecular absorptivities at for a numthe appropriate A,, ber of olefins are given in Table I. For a multicomponent analysis, overlap absorptivities are obtained in a similar manner a t each of the wave lengths used. For ordinary quantitative analysis, as distinguished from type analysis, it is preferable to calculaJe a “weight absorptivity.” a , defined as follows:
a‘
A,
= weight absorptivity = absorbance a t ,A ,
,
corrected for I t background (CI) = concentration of I? in final solution, grams liter ( C O ) = concentration of olefin in the final solution, grams/liter
Basic Theory. I n order to understand the choice of con-
_
_
_
~
~ -
-
~
-
~ _ _.
~
____
~~
Table 1. Molecidar Absorptivities for Iodine-Olefin complexes e’,
RCH=CRz
Compound 2,3-3lez-2-butene 2,3-Mez-Z-pentene 2 3-Mez-2-hexene Z-hle-Z-butene 3-~le-t,ans-2-pentenr 3-hle-cis-2-penterie 2-hIe-2-pentene 2-Me-2-hexene 3-hle-c%s-3-hexene 3-Et-2-pentene
I, Concn., Grami Liter 0 , 4405 0 5105
Olefin Concn.. hlole/Litt.r 0.063t;
0.5105
0 .O R l t i
0,5040 0,5040 0.4405 0.5470 0.4405
0 0773 0 0795 0 0647 0.0761 0 0703 0 0465 0.0477 0.0389 0.0817 0,0364 0.0615 0.0664 o 0795 0.0999 0 0689 0 0700 0.0855 0.0832 0.0923 0 0898 0 0704 0.0713 0 0767 0 0847 0 0893 0 07liO 0,0731; 0 .om0 0,04.56 0.0733 0.0668 0.0972 0.0713 0.0693
n 5280 0.6840
0.6840
RHC-CHR
R HC= C H R cyclic RCH=CHz
cis-2-Pentene trans-2-Pentene cis-2-Hexene trans-2-Hexene cis-3-Hexene trans-3-Hexene 4-Me-cis-Z-pentenP 4-Me-t rans-%pen t enr 4, GMez-cis-2-pentene 4,4-MeArans-2-pentt.ne trans-2-Heptene trans-3-Heptene trans-4-Octene 2.2-~~e?-cis-3-hexenp 2,2-hle~-traiis-3-he~ene Cyclopenten~ Cyclohexene I-Hexene 3,3-Mez-l-butene 3-Me-1-pentene 4-Me-1-pentene 1-Heptene 4-Me-1-hexene S-hl e-1 -hexene 4.4-Ales-1-pentene 1-Octene 1-NonrnP . ~....
DI SCU SSlON
cal. per mole (.7)!
The iodine functions as an electron-acceptor (Lewis acid), and the hydrocarbon functions as the electron-donor (Lewis base ,. At equilibrium, the concentrations of the comples and the equilibrium constant, K , are defined hy the niass law equation:
T o obtain citlibration absorptivities, measurements are made a t the wave lengths of maximum absorption (Amx.) characteristic of each olefin type, as follow (see Figure 1): Olefin Type RCH=CH? RzC=CHs
+ olefin e I? X olefin [ A H ea. -500
1-Deoene I-Dodecene I-Tetradecene I-Pentadecene
0.5470 0.6840 0,5280 0.5470 0,6840 0.5105 0.6765 0 6765 0.5000 0 5000 0.5470 0 5103
nm o
0 .m40 0,507.5 0 5280 0 5470 0 5280 0,5280 0.5095 0,5095 0,5040 0.5470 0.5280 0.5470 0.5365 0 5280 0,5280 0.5365 0.6840 0,5365 0.5105 0.5075 0.5365 0.6840
0.0453
0.090ti 0 06.53
0 0773 O.Ofi23
0 0671 0,0626 0 0577 0 0645 0 0622
1 llole/Liter
,
Amas.
lI$ 337 337 337 313 317 317 317 317 317 317 293 293 290 293 295 293 19.5 294 29.5 295 293 29s ‘97 29.5 29.5 L”)7 800
?,nn
300 300 300 293 27.1 27.5 275 27.5 27i 27.5 ‘27.5 27.5 ‘27.; 27; 273 27.5 275 273
1-Crn. Cell 1,081; 0.711 0 937
1 Grani/I.iter of Olefin,
of I 2 38.7 30.8 29.8
1.090 1.089 0.823 1 180 0.804 0,648 0.790
28.0 27.1 28.9 28.4 25.9 26.4 24.2
0.634 1.074 0 862 1.090 0.918 1 134 0 692 0 915 0 018 0 884 0 ,567 0 . !j82 0.311 0 788 0 370 0 ,524 0 ,i:3 i 0 481 O.ri84 0 . 2Rli
23 24 32 25 25 20 13
0.712
20.8
0,6tj7
0.449 0.428 0,505 0,393 0,480 0.489
0 558 0 3R8 0.553 0 460 0.424
o.ain
0 .4 i i i O.Ii33
8 1
5
3 I
9
2
18.r
13.1 20.7 13.7 16.2 11 2 lli.4 7 li 13 i 11.9 10.R
17.1 6.8
29.2 12.: 11.i
9.9
10 1
12.9 10 2 12 4 8 9 13.0 12.8 13.3 13.4 13 4 14 9
ANALYTICAL CHEMISTRY
1112
an average absorptivity of 15,000 liters per mole-cm. is assumed for olefin-iodine complexes, the concentration of the complex can be calculated directly from experimental absorbance data (Table I). An illustrative calculation from these data is given below for 3-methyl-cis-3-hexene : [ISX olefin] = A / EX path length = 0.648/15,000 X 1 cm. = 4.3 X 10-5 mole per liter The equilibrium constant, K , can now be calculated from the mass law equation, using the calculated concentration of the complex and the known iodine and olefin concentrations. The K value for 3-methyl-cis-3-hexene is found to be 0.45 liter per mole. Similar calculations for the other olefin types show that K is of the order of magnitude of 1 liter per mole. This approximate K value of 1 liter per mole can be used to select suitable concentrations of iodine and olefins. Calculation shows that it is not possible to use an excess of iodine so as to make the absorbance independent of iodine concentration. The feasible approach is to use concentrations such that the absorbance is proportional to the product of the iodine and olefin concentrations. This will be nearly true for concentrations for which 5 % or less of both components are in the complexed state. Such an approach has an important analytical advantage-the added concentrations of olefin and iodine can be used to calculate the absorptivity of the complex, since little has been used to form the complex. The concentrations of 0.002 mole per liter of iodine and 0.03 to 0.1 mole per liter of olefin used for analysis meet the conditions discussed above. These concentrations permit the use of standard 1-cm. absorption cells. Furthermore the background absorbance due to uncomplexed iodine does not rise above 0.2 in the 276- to 400-mp region. The justification for the use of an “absorptivity,” E ‘ , (see calibration section), is that the absorbance is proportional to the product of the iodine and olefin concentrations within the selected range of concentrations. Therefore, the E’ can be defined as the absorbance divided by the product of iodine and olefin concentrations. This is analogous to the conventional molar absorptivity, e, defined as the absorbance divided by the concentration of a single component. The apparent molecular absorptivity,
~
E‘, as defined in this paper is related to the true molecular absorptivity, E , of the complex as follows: E’ = K E / M ( K is the equilibrium constant of formation of the complex in liters per mole, and M is the molecular weight of iodine in grams per mole). Finally, the iodine concentration is expressed in grams per liter rather than in moles per liter in order to obtain an absorptivity unit of convenient numerical size. Experiments vere carried out to show that, for the conditions chosen, the absorbance measured is proportional to the product of iodine and olefin concentration. In view of the 30- to 100-fold mole ratio of olefin to iodine which is used, a very slight reaction of iodine to form diiodo compounds would apparently destroy the desired relationship. For this experiment a propylene polymer was used as the test olefin. Table I1 shows that the E’ values a t 295, 317, and 337 mp are independent of concentration even though the mole ratio of olefin to iodine was varied from 1.4:l to 2 i : l .
Table 11. Constancy of Absorptivity of Iodine-Olefin Complexes for Varying Concentrations of Propylene Polymer and Iodine Concentrations, SIoles/Liter Olefin Iodine 0.053 0,00198 0.0127 0,00909
2.8
2.8
-
0
0
E 2.0 -
0
E
-
W
6‘
13.7 13.7
0.366 0.402
337 XI* Net A e’ 0.325 12.2 0.354 12.0
2.4
2.0
Q)
16 .
1.6 0
0
e
e
2
317 XI@ Xet -4
Photochemical Effect. I t is important for the present method that no appreciable fraction of the iodine added be consumed by reaction with the olefins to form diiodo compounds. The authors have found that ordinary room light (from fluorescent bulbs and north daylight) catalyzes the undesired reaction. Figures 2 and 3 show that the spectra of iodine-olefin complexes are stable if the solutions are kept in the dark, but that a change occurs i f the solutions are exposed to light. The authors find that the simpler olefins react appreciably in a few minutes if exposed to light, whereas the more fully substituted olefins react much more slovly. Another interesting proof that iodine adds to olefins in the 3.2
0
295 ll@ Xei A e’ 9.8 0.262 9.4 0.275
~
3.2
2.4
-
2
1.2
1.2
U
Q .8
.e
.4
.4
.o
.O 250
350
300
400
250
300
350
400
Am9 Figure 2.
Effect of light on spectra of l-octeneiodine complex
Figure 3.
Effect of light on spectra of trans-2-hexene iodine complex
V O L U M E 2 7 , NO. 7, J U L Y 1 9 5 5
1113 vacuum distillation of such a sample, no change in the net absorbance due to the iodine complex is observed. Correlation of Olefin-Iodine Complexes with Molecular Structure. Iodine complexes for a variety of structural types of olefins have been measured (Table I). The following conclusions from these data appear to be justified: 1. The olefins can be arranged by degree of alkyl substitution, and each type then has an approximately constant A, and e’, as follows: Type RCH=CHz RC=CHz cis-RCH=CHR trans-RCH=CHR R,C=CHR RzC=CRa
0
E -
1.2
0
a u K O
a L
0 ul
n
a
Butene
250
Figure 4 .
300
350
+
.4405g./1.
4 00
.inal>-sis of two-component olefin blend by means of iodine complexes
presence of light is observable by means of the iodine absorption spectrum in the visible region. Iodine in iso-octane solution has only one band in this region, a broad one n i t h, , ,A = 520 mp. The authors have calculated that ea. 27, of the iodine added is used to form the olefin complex under the chosen concentration conditions. This \vas verified by an immediate decrease of 2 % in the 520-mp band intensity when 0.0’7 mole per liter of trans-2hexene n-as added to a solution of iodine in iso-octane (0.002 mole per liter). After this solution was exposed to room light for 2 hours, the 520-mp band had decreased another 14yoin intensity, indicating that addition of iodine to the double bond had occurred. For satisfactory analytical work with olefin-iodine complexes it is therefore concluded that the spectra should be measured within a short time (ea. 1 minute) after mixing the components. .%lternatively, the solutions can be preserved unchanged for a t least an liour if kept in the dark until measured. I t should be noted that no catalytic effects are observed from the lon--intensity, monochromatic radiation incident on the sample during the ahsorption measurements. Interferences. So far the method has been applied to mixtures containing only olefins, so there are not many data available on interfering substances. However, it is known that saturated hydrocarbons would not affect the analysis ( 7 ) . Aromatic hydrocarbons Tvill interfere because their iodine complexes overlap those of the olefins a n d a r e of comparable apparent intensity ( 7 ) . I t may be feasible to correct for the interference of particular aromatics if their concentrations are determined independently. Diolefins add iodine rapidly a t room temperature ( 5 ) ,and would therefore interfere. I t has been found that the olefin analysis is not appreciably affected by the traces of color bodies and peroxide present in a propylene polymer sample that is a few days old. Although the spectral background is markedly lowered by a bulb-to-bulb
Mp 275 290-295 295-300 295-300 317 337
e’
Xmax.,
12 25 19 11 27 23
More limited subclasses such as linear olefins of the RCH=CH2 type, or 01- and 8-methyl substituted olefins of this class, have a In considerably more constant molecular absorptivity, E ’ . general the absorptivities, e’, increase markedly with increasing alkyl substitution of the ethylenic hydrogens. 2. The trans forms of the RCH=CHR type have considerably lower values of E’ than the corresponding cis forms. 3. With few exceptions, branching at a carbon atom 01- or pto the ethylenic carbon atom considerably lowers the E’. This effect is evident for such olefins as 4,4-dimethyl-trans-2-pentene and 4,4-dimethyl-l-pentene. This marked effect of branched R groups is the reason that no statement of average deviations is given for the E‘ values in the above table. No olefins of the R2C=CHR and R2C=CR2 types with branched-chain R groups are available from .%PIProject 44. and it would be misleading to give an average deviation figure based only on olefins n-ith unbranched R groups. 4. Cyclopentene and cyclohexene form olefin complexes with A,, and e’ values typical of cis open-chain olefins of the RCH=CHR type. However, cyclohexene has the highest E’ observed for any cis olefin. APPLICATIOYS TO ANALYSIS
Ordinary Quantitative Analysis. The iodine complex procedure can be used to make a quantitative determination of particular olefins which are qualitatively known to be present in a mixture. Figure 4 illustrates such an analysis for a two-component test blend containing 72.1 weight %of 2-methyl-2-hexene and 27.9 weight % of 2,3-dimethyl-2-hexene. The characteristic wave lengths of 317 and 337 mg are used for analysis, since these are tri- and tetrasubstituted olefins. The only unusual step is in calculating absorptivities, a’, as the quotient of the absorbance and the product of olefin and iodine concentrations (see calibration). Otherwise the calculation follon s the conventional spectroscopic procedure involving the usual linear simultaneous equations. The results are shown in Table 111. d more difficult analysis of the conventional type (as distinguished from a type analysis) is shown in Table IV.
Table 111. Analysis of Two-Component Olefin Blend Olefin Type Rz=CHR RzC=CRz
Known Value, Compound 2-Methyl-2-hexene 2,3-Dirnethyl-Z-hexene
%
72.1 27.9
Analysis by Iodine Complex, % 71.5 28.5
~-
Table IV.
-
Analysis of a Five-Component Olefin Blend
Compound
Known Value,
Analysis by Iodine Complex,
%
%
ANALYTICAL CHEMISTRY
1114 Examination of a Propylene Polymer. Limited success hap been achieved in extending the iodine-olefin procedure to an olefintype analysis of a complex olefin mixture. A propylene polymer was selected because it is thought to consist largely of tri- and tetrasubstituted olefins which are detectable by the iodine-olefin procedure I t n a s ’thought best to analyze such a complex olefin mixture by a combination infrared and iodine-olefin method. This takeq advantage of the certainty with which the simpler olefin types can he determined bv infrared spectroscopy. The infrared method used is essentially equivalent to that published by Saier and other% (9), except that a 0 1-mm. cell and n-heptane w r e used. The matrix used for the iodine-olefin procedure, including the overlaps for the olefins determined by infrared analysis, is give? in Table V.
The propylene polvmer used was prepared by polymerization of propylene over UOP sohd phosphoric acid catalyst. Its average molecular weight was approximately Clo as determined by boiling range. The analysis obtained is given in Table VI. The low total obtained in Table VI is believed to be due to the lack of tri- and tetrasubstituted olefins with branched-chain R groups for calibration. It has been pointed out that blanching of the R groups of the simpler olefin types generally decreaws the absorptivity, e‘. It is possible that this effect iq even more pronounced for the tri- and tetrasubstituted olefin^. T h r propylene polymer is thought to contain principally theqe high11 branched structures This type analysis is the first spectroscopic determination that the RzC=CRZ type of olefin is present in propylene polymei Further, the RzC=CHR type is shown to be a major constituent This confirms infrared evidence based on a broad. geneial abSOI’ptiOn in the 800- to 850- cm.-l region.
Table I-. Matrix for Combined Infrared and Iodine Complex Methods for Olefin-Type 4nalysis cis-
RHC= CHR
295 317 337
E 11 6 3 1
Molecular Absorptivity e ’ transRzC= R?C= RHC= CHR CR, CHR 17 1 86 113 86 22 3 270 2 6 3 x 192
RCH= CH? 5 5
09 03
RICE CH2
248
127 31
Table YI. Analysis of Propylene Polymer Olefin Type RCH=CHz
R?C=CHga trans-RCH=CHRa
Percentage 2.4
7.0
cis-RCH=CHR
RL!=CHR RzC=CR% Total a Determined by infrared analysis . .-
~
~
LITERATURE CITED
(1) Anderson. J. .I.,Jr., and Seyfried, K. D.. .%N.I., J . A m . CAenz. Soc., 74, 458 (1952). (3) Benesi, H.A , , and Hildebrand. J. H., I h i d . , 71, 2703 (1949) (4) Fenske, M. R., Braun. W. G.. Wiegand, R. V., Quiggle, D.. McCormick, R. H., and Rank, D. H.. AN.
