Tetracyanoethylene π-complex chemistry. Indirect spectrophotometric determination of Diels-Alder-active 1,3-dienes

Wanta,

Vol. 20, pp. lOES-1096. Pergamon Press, 1973. Printed in Great Britain

TETRACYANOETHYLENE

n-COMPLEX

CHEMISTRY

INDIRECT SPECTROPHOTOMETRIC DETERMINATION DIELS-ALDER-ACTIVE 1,3-DIENES

OF

DALE A. WILLIAMS*and GEORGEH. SCHENK Department of Chemistry, Wayne State University,Detroit, Michigan48202,U.S.A. (Received 30 November 1972. Revised 25 April 1973. Accepted 1 May

1973)

Summary-An indirect spectrophotometric method, based on the rapid Diels-Alder reaction betweencisoid 1,3-dienesand tetracyanoethylene(TCNE) and the destruction of an aromaticTCNE x-complex, was developed to determine eleven 1,3-dienesin the O-05-1*00x 10e3M range. These dienes were: cyclopentadiene; 1,Ecyclohexadiene; trans-1,3-pentad; 2,4dimethyl-1.3~pentadiene;trnns-2-methyl-1,3-pentadiene;2-methyl-1,3-butadiene;!%methylanthracerie; 9,10-dimethylante; 1,6-diphenyl-1,3,5-hexatriene; 2,3-dimethyl-1,3-butadiene; and l&diphenyl-1,3-butadiene.Three 1,3-dienesweredeterminedin theO.O5-1 x 10-4Mrange: cyclopentadiene, trans-2-methyl-1,3-pentadiene, and anthracene. The limit of detection for cyclopentadiene in carbon tetrachloride solutions is O-11pg/ml. Fourteen 1,3-dienes were found to form stable ?r-complexes and could not be determined by the proposed method. For these 1,3_dienes, the spectra of some of the complexes are reported; in addition, relative equilibrium constants for the s-complexes of 2,5-dimethyl-2,4diene, cis-1,3-pentadiene, 4-methyl-1,3pentadiene, and 1,3-cycle-octadiene were estimated. An explanation of the transient colour in the I,3-diene-TCNE Diels-Alder reaction is suggested.

The Diels-Alder reaction of a 1,3-diene and a dienophile to form a cyclohexene derivative is a major synthetic organic reaction. However, the literature does not contain any reports of the use of this reaction for the trace analysis of a series of 1,3-dienes, though isolated examples have been found.lA With the synthesis of tetracyanoethylene (TCNE) in 1957,’ a new extremely electrophilic dienophile (and n-acid) became available for use in Diels-Alder reactions. The great reactivity of this dienophile with 1,fdienes was first cited by Middleton er ~1.~ and later by Stewart.’ In 1961, Schenk and Ozolins* demonstrated the utility of this rapid Diels-Alder reaction by the indirect determination of anthracene in trace quantities. Small amounts of certain 1,3-dienes have also been determined by a diazo reaction, * by formation of coloured fulvenes,’ by spectrophotometry,” gas chromatography,” mass spectrometry,‘* and coupled spectrometry,“’ i3 coupled gas chromatography/mass coulometry/spectrophotometry.ls In this paper a simple method with the utilization of relatively inexpensive instrumentation is described. The basis of the method is the formation of a coloured aromatic_TCNE n-complex and measurement of the decrease in absorbance of the x-complex in the presence of a Diels-Alder-reactive 1,3-diene. The decrease is usually proportional to the concentration of 1,3-diene present although the reaction is not exactly stoichiometric. The formation * On leave 1972-73 from Thiel College, Department of Chemistry, Greenville, Pennsylvania 16125, U.S.A. 1085

1086

DALEA. WILLS

of the yellow mesitylene-TCNE of these complexes

and GEORW H. SCHENK

x-complex from colourlcss reactants illustrates the utility

The equilibrium constant, K,, for this reaction is defined by K, _ [mesitylene-TCNE complex] [mesitylene] [TCNE]

(1)

Unfortunately, the equilibrium constants reported in the literature for TCNE (see the KN values in Table 1) are defined by means of a mole-fraction term (footnote, Table 1). The K, constant is defined entirely in terms of molarities so that the values of K, are somewhat smaller than those of KN . In methylene chloride, K, = K&6. To design an optimum method, the optimum aromatic hydrocarbon (x-base) must be selected to form a n-complex with TCNE. Some of the parameters involved are listed in Table 1 for various aromatic hydrocarbons. The choice of the optimum aromatic hydrocarbon is determined by its availability in quantity, its purity, its K, value and the molar absorptivity of its TCNE complex, E. After substitution of absorbance for concentration of the x-complex in equation (1) and adjustment for the mole fraction of the n-base, the rearranged Benesi-Hildebrand equation’ 6 giving the relationship between the initial molar&y of the x-acid, COPeid,and the absorbance A, E, and K, is: (2)

CO,cid = A K E ‘Co + i [ X. - bare I

where Cobye is the initial molarity of the x-base. After selection of a given aromatic hydrocarbon for evaluation, equation (2) can be used to calculate the initial’molarity, C’pcid, of TCNE needed. (This is the molarity of TCNE present before it reacts with the 1,fdiene.) In choosing an arbitrary value for A in equation (2), it is important to select a large value. This is because the absorbance will decrease after TCNE reacts with the 1,3-diene to be determined. To achieve a large A, a large excess of aromatic hydrocarbon is needed because the complexes of TCNE are rather weak. A value of 14M4 for CObarewas arbitrarily chosen to achieve a value of 0.50 for A. Table 1. Constants for some TCNE nccomplexes in CHzCll solvenP’ x-Donor Naphthalene Mesitylene Durene o-Xylene Pyridine Hexamethylbenzene Benzene

KN*

E,

11.7 17.3 54.2 6-97 12.0 263 2.00

* Kn = ‘Ar~,!$~~(~~~~‘exl fraction of the aromatic hydrocarbon

1.m01e-~. cm-’ 1.24 3.12 z-08 3.86 1.05 4.39 3.57

x x x x x x x

lo3 10” 10” lo3 104 10” 10’

Amu,nm 550

El 430 422 545 384

, where (Aromatic) is the mole

Determination of Die&Alder-active 1.3-dienes

1087

Sample calculations for both the pyridine-TCNE and mesitylene-TCNE n-complexes will now be discussed. For pyridine-TCNE, the KN value in Table 1 corresponds to a value of 075 for K, . For this complex then, in methylene chloride solvent, CO.,.cNE= 050[1*27 x 1O-4 + 0.95 x 1O-4] = 1.1 x 10-4M This implies that the upper limit of 1,3-diene that can be measured in methylene chloride solvent is about 1 x 10V4M; the lower limit should be about 1% of this or 1 x lo- 6M. For mesitylene-TCNE, the KN value in Table 1 corresponds to a value of 1.1 for K,. For this complex in the same solvent, COTCNE= 050[2.9 x lO-4 + 3.2 x lO-4] = 3.0 x 10-4M This implies that the upper limit of 1,3-diene that can be measured in methylene chloride solvent is almost 3 x 10-4M; the lower limit should be about 1% of this or 3 x lo- 6M. Pyridine is thus a better choice than mesitylene for forming a ?r-complex to measure the lowest concentration of 1,3-dienes. Initial work involved the use of this n-complex; however, it was found that precipitates formed when 1,Zdienes were added. Hence most of the experimental work was done with mesitylene (and naphthalene). To compensate for the lower stability of the mesitylene-TCNE complex chloroform and/or carbon tetrachloride were substituted for the methylene chloride solvent. In these solvents K, is appreciably larger than it is in methylene chloride. The effect of this substitution is to lower the calibration curve range for mesitylene-TCNE to a range close to that of pyridine-TCNE in methylene chloride. EXPERIMENTAL Reagents Standard 0.05OOM TCNE. Dissolve 06406 g of doubly sublimed TCNE in 100 ml of reagent grade methylene chloride and store in the dark. (This solution appears to be stable for about two months.) Standard I-5OM naphthalene. Weigh 192.3 g of reagent-grade naphthalene into a 1-1. volumetric flask and dissolve it in about 750 ml of reagent-grade carbon tetrachloride or chloroform. (The dissolution of naphthalene is endothermic and slow.) Make the final dilution with the same solvent after the solution has attained room temperature (about one day). Standard 2.5 x IO-‘M 1,3-diene solutions. Weigh exactly 0.62 mmole of the 1,3-diene (reagent-grade or doubly distilled and checked for purity) into a flask, cover the flask with ahnninium foil, and add 250 ml of reagent-grade carbon tetrachloride or chloroform to dissolve the diene. Store in the dark since photodecomposition may occur with some solutions. Because cyclopentadiene polymerizes very quickly at room temperature, store the pure cyclopentadiene at “ dry ice”-acetone temperatures and use prepared solutions within a couple of hours. Prepare other standard 1,3-diene solutions by making the appropriate dilution. Apparatus

Spectrophotometric measurements were made in round and square 1.1 cm cells on the Spectronic 20, and square 1.00 cm cells on the Beckman DB-G and Cary 14 spectrophotometers. For purity checks, a Carle gas chromatograph and Abbe refractometer with temperature bath were used. Weighing of tng quantities of volatile 1,3-dienes. Place about 4 ml of solvent in a 5-6 ml capacity ampoule. Over the top of the solvent-containing ampoule place and squeeze a rubber pipette-bulb which will fit a Pasteur disposable pipette. Weigh the ampoule. Introduce small amounts of 1.3-diene into the ampoule by means of a 2-ml syringe equipped with a small hypodermic needle and filled with the pure diene. Weigh the ampoule with the diene added. Take off the rubber bulb and immediately transfer the contents of the ampoule to an appropriate volumetric flask. Rinse out the ampoule three times with solvent. Dilute the sample of diene as quickly as possible and store in the dark if necessary. Calibration plot for 1,3-dienes in the 0.5-1.0 X 1Ob3M range By pipette transfer 1 ml of O*OSMTCNE into a 50-ml volumetric flask, and add l-20 ml of 2.5 x 10’)M

1,3-diene solution, washing down the sides of the flask with l-2 ml of solvent. Allow to react in the dark for the time specified in Table 2. After the reaction is essentially complete, add 20 ml of 1*5OMnaphthalene

DALEA. WILUAMSand GEORGE H. Scnnrnc

lo&8

Table 2. Reaction times for 1,3-diene-TCNE reactions Reaction times 0.05-l x lo-SM 0.05-l x 10-*&f

IJ-Diene Cyci~tadiene

smin

zragr-2-Methyl-1,3-pentadiene 2&DiiC1,3-pentadiene 2,3-Dimethyl-1,3-butadlene 1.3~Cyclohexadiexle 2-Methyl-1,3-butadiene 1,4-Diphenyl-1,3-butadiene 1,~Dip~nyl~~~e Anthmcene 9-Methylanthracene 9,10-Dimethylanthracene rrans-1,3+entadiene

0.5 hr 1 hr zz

12 hr 24 hr -

2days 24 hr 2 days O-33hr 10 min 10 min 24 hr

24 hr -

solution by pipette and dilute to the mark. Prepare a reagent blank by the addition of the same quantities of TCNE and naphthakne to a 5&m1 volumetric flask, and dilute to the mark with the samesolvent as above. Read the absorbance or transmittance of each solution at 550 nm. Plot absorbance us. concentration to prepare the indirect calibration plot. To utilii the more accurate Reilley-Hi&brand’s method for indirect spectrophotometric measurements, set the transmittance to 100% at 550 nm for the dienc solution and read the transmittance of the reagent blank. Repeat this procedure for each solution and setting transmittance to 100% for each diene solution. (For a double-beam spectrophotometer, place the diene soiution in the reference beam and the reagent blank in the sample beam.) A plot of absorbance us. concentration of 1,3-diene yields a typical calibration plot. Table 3. Stability of complexes of TCNE with I,J-dienes

1,3-Diene

&1,3Pentadiene tmns-1,3Pentadiene 4-Methyl-1,3pentadiene tranr-2-Methyl-1,3-pentadienc 2,4-Dimethyl-1,3-pentadiene 2,5-Dimethyl-2&hexadiine cis, tranr-lJ-Hexadiene ~3-~~yI*l,3-~~~e 2-Methyl-1,3-but 1.3~Cycle-octadiene Anthracene 9.Methylanthracene 9-Anthraldehyde 9,10-Dimethyku&racene l~s,~~~t~e ~~~~~ l&Diphenyl-1,3-butadiene 1.3-Cyclohexadiene 1,6-Diphenyl~triene Cycfopentadiene 1,~3,~Te~~yl-l,~fo~~ FUran ZMethylfuran 2,5-Dimethylfuran

Initial cofour (&., , mm) Yellow-orange (480) Yellow-orange (490) Violet (535) Reddish violet Bluish violet Blue (645) Grange (4e51 Orange Yellow-orange (455) Reddish orange (490) Green Yellow Green Zlr(7) Light violet Yellow Bluish green Violet Green to yellow Not observable Yellow ~~(~l~~-~) Blue (575)

Tie for disappearance of colour, min Q, (stable) >5 cc (stable) 0 (fast reaction) 0 (fast reaction) 00 (stable) >5 0 (fast reaction) >5 cc (stable) >5

0 (fast reaction) cc (stable1 0 (fast re&tion~ co (stable) 60 00 (stable) 15-20 >5 15-20 cc (stable) a~ (stable) co (stable) Q, (stable)

Estimated & o-74 1.5 9.7

054 -

0.35

Determination Calibrationplot

of1,34ienes

in the 0.5~1.0

Transfer 2 ml of 25 x lO_jM TCNE with l-2 ml of solvent. Allow reaction mesitylene and make the final dilution. 5 ml of mesitylene to a 50-ml volumetric the transmittance or absorbance of each at 460 nm.

of Diels-Alder-active x

1,3-dienes

1089

10m4M range

into a 50-ml volumetric flask and wash down the sides of the flask to occur for the time specified in Table 2. Then add 5 ml of pure Prepare a reagent blank by the additon of 5 mmole of TCNE and flask and dilution to the mark with the same solvent as above&ad solution at 460 nm (Table 2), or use the Reilley-Hildebrand method

Estimation of KN and E values for some 1,3-dienes Weigh a IO-ml volumetric flask and stopper. With a syringe place about 1 mmole of 1,3-diene in the flask and weigh again. Add (by pipette) IO ml of 0.504025M TCNE to the flask and weigh. As soon as possible, measure the absorbance of the solution at h ,,,.%for the 1,3-diene n-complex with TCNE (Table 3), against reagent-grade methylene chloride as reference. RESULTS Eflect

AND

DISCUSSION

of solution variables

Concentration of 1,Zdienes. With both the indirect and Reilley-Hildebrand methods, it was found that the spectrophotometric measurement of reactions of cyclopentadiene, 1,3cyclohexadiene, 9-methylanthracene, 9,10_dimethylanthracene, 1,6-diphenyl-1,3,5_hexatriene, 1,4-diphenyl-l,3-butadiene, trans-2-methyl-l,3-pentadiene, 2,3-dimethyl-l,3-butadiene, and Z+dimethyl-1,3-pentadiene generates essentially straight line plots in the 0~05-140 x lo-‘M range while cyclopentadiene, anthracene, and rrans-Zmethyl-1,3-pentadiene give straight line plots in the O-05-140 x 10e4M range. Although these Diels-Alder reactions are not exactly stoichiometric, differences in the calibration plots are slight, but a calibration plot should be prepared for each 1,Ediene to be determined. The plots for isoprene and piperylene (trans-1,fpentadiene) were found to deviate from linearity as shown in Figs. 1 and 2. The usefulness of these methods in the vicinity of 140 x 10m3M concentrations of 1,3-diene is limited by the curvature of the line. It may be that these two 1,3-dienes polymerize faster than the others and that the rate of polymerization is close to the rate of Diels-Alder reaction with TCNE. On the other hand, at higher 1,3-diene concentrations, the strength of the n-complex between TCNE and the transoid conformation of the 1,3-diene may hinder the formation of the cisoid conformation (this conformation is necessary for a Diels-Alder reaction) of the 1,3-diene. This point is discussed

Concn,

Fig. 1. Deviations

10%

in the indirect spectrophotometric plots of 1,3-dienes (Spectronic 20). (1) Isoprene. (2) trans-1,3pentadiene.

DALE A. WILLIAMSand GEORGE H. SCHENK

lo!40

0

2

6

4 Concn.

a

IO

12

104M

Fig. 2. Deviations in the Reilley-Hildebrand plots for 1,3-dienes (Spectronic 20). (1) Isoprene. (2) tram-1,3-Pentadiene.

in the section on the nature of l,Idiene-TCNE n-complexes. Probably each of these effects contributes to the abnormally high deviations at higher concentrations of 1,3dienes. The following compounds were not amenable to this method of analysis: calciferol; 1,3-cycle-octadiene ; 9-anthraldehyde ; 4-methyl-l ,bpentadiene ; 2,5-dimethyL2,4_hexadiene; cis- 1,3-pentadiene; benz[u]anthracene; 1,2,5,6dibenzanthracene; 1,2,3,4-tetraphenyEl,3cyclopentadiene; furan; 2-methylfuran; and 2,ldimethylfuran. Most of these compounds form stable complexes with TCNE (Table 3); however, benx[u]anthracene and calciferol do undergo very slow reactions with TCNE. Probably tr~s- 1,Zhexadiene could be determined in the O-05-1.00 x 10W3M range by the method since a preliminary test on the cir, tr4ns mixture of 1,3-hexadiene indicated that its reaction time is slightly faster than that of the piperylene. No separation for the mixture was found; therefore, no calibration plot was prepared. In addition, the determination of 9-methylanthracene and 9,10_dimethylanthracene appears to be feasible in the 0.05-1.00 x 10V4M range, according to the reaction times; however, no results were collected for these two aromatic dienes. Solvent e$kt. As is indicated in Table 4, change of solvent drastically a&&s the absorbance; therefore, the calibration plots should be prepared with the samples and standards in comparable solvent systems. Stability of solutions. The stability of the naphthalene-TCNE solutions is excellent, the absorbance varying by no more than 0.006 units over some hours. Except for slight fading, the mesitylene-TCNE solutions are also stable for long periods of time. A precision study of ten solutions for the 5.0 x 10m5M trans-Zmethyl-1,3-pentadiene system gave absorbance values of 0.164 f OW2 (indirect) and 0.141 + 0.002 (Reilley-Hildebrand). The Reilley Hildebrand absorbance measurements were taken one day after the indirect measurements were taken. This indicates any fading is consistent throughout the set of solutions. Table 4. Effect of solvent on diene reagent blank

Solvent

Absorbance, 550 nm (Spectronic 20)

CCL CHCIJ CH&lz

1.28 0.85 0.50

Absorbance, 550 nm Gary 14) 1.078 O-730 0.432

Determination of Die&Alder-active 1,3-dienes

1091

Table 5. Effect of concentrations of naphthalene on absorbance of the naphthalene-TCNE complex Concentration of naphthalene, M 0.75 0.69 0.60 0.51 0.45

Absorbance of blank*

Absorbance of sample?

1.109 I.120 1so75 1W9 0950

0642 0590 0567 0.522 0.520

Difference in absorbance, AA

0467 0530 0508 0.487 0.430

* The blank was prepared by the addition of 1 ml of BOSOM TCNE and the appropriate volume of 150M naphthalene to a W-ml flask and diluting to the mark with Ccl,. All readings were taken on the Gary 14 spectrophotometer. t The sample was prepared in the same manner as the blank except for the addition of 10 ml of 2.5 x 10m3M cyclopentadiene before dilution. Concentration of naphthalene and mesitylene. Schenk and Ozolins4 showed that the naphthalene-TCNE reagent was suitable for the indirect spectrophotometric determination of anthracene. Hence essentially the same reagent was used for the 1,3-diene determinations in the 0~05-190 x 10T3M 1,3-diene concentration range. Table 5 indicates that the 0*60M naphthalene-TCNE reagent gives nearly the highest sensitivity. The 06OM naphthalene concentration was chosen because (a) at higher concentrations of naphthalene solvent evaporation is great enough to leave a white powdery film of crystalline naphthalene on the glassware, and (b) the absorbance of the reagent blank is approximately in the centre of the minimum error range for a double-beam spectrometer.’ ’ Doubly distilled mesitylene gave no better reagent blank values than did Baker White Label mesitylene. Therefore, Baker or other commercially available mesitylene was used for measurements in the 0.05-1.00 x 10s4M range. A study of various concentrations of mesitylene for the reagent blank indicated that 5 ml of mesitylene per 50 ml in carbon tetrachloride gave nearly the optimum absorbance measurement. Interferences. Schenk and Ozolins delineated the major interference effects of aromatic hydrocarbons in this type of analysis. Experiments have shown that mixtures of 1,3-dienes which react at comparable rates can be analysed by an indirect differential kinetic spectrophotometric method, and that very slow-reacting 1,3-dienes do not significantly interfere with the determination of the faster-reacting 1,3-dienes. This subject will be presented in a subsequent publication. Order of addition of reagents. To minimize the reaction time, it is important to adopt a specific order of addition of reagents. If the 1,3-diene and TCNE solutions are mixed before the aromatic donor is added, the reaction is found to take only half as long as when the aromatic donor and TCNE are mixed before addition of the 1,3-diene sample. Instrumental efict. Since the absorption bands of aromatic-TCNE n-complexes are unusually wide, the methods were found to be feasible on all four spectrophotometers utilized. Nevertheless, the band-pass of the instrument is reflected in the absorbance measurements. This is indicated for the Spectronic 20 and the Cary 14 spectrophotometers by the solvent effect study on the naphthalene-TCNE solutions cited previously, and also for the Beckman DB-G and Spectronic 20 spectrophotometers. The following trend was observed:

1092

DALEA. WIL~JAMS and GEORGE H. SCHENK

the larger the bandwidth of the spectrophotometer, the higher the absorbance readings. This study simply indicates that the method is feasible on both narrow and wide bandwidth spectrophotometers. Precision and reproahibility. With carbon tetrachloride as the solvent, the naphthaleneTCNE reagent blank was found to yield absorbances of 1.03 & 0.02 and I.090 f OGM for ten independent solutions measured on the Beckman DB and Cary 14 spectrophotometers, respectively. Precision studies for ten independent solutions of 750 x 10m4M 9-methylanthracene and 7.20 x 10S4iU trans-2-methyl-1,3pentadiene solutions in carbon tetrachloride gave absorbances of 0.726 f OGO4 and 0664 + 0.008, respectively (Beckman DB spectrophotometer). Precision measurements on other 1,3-diene systems gave similar results in both ranges for both methods. With the values of precision found for a number of systems, the reproducibility of the calibration plots should be good. This was observed qualitatively since the shapes of the calibration plots did not vary even though the slopes did, especially for slow-reacting dienes when the reaction times were changed. Therefore for optimum results, the reaction times in Table 2 should be used to within about f 10 %. Sensitivity and detection limit. The only meaningful sensitivity for these systems is the quantity M/AC over the straight-line portion of the calibration plot. This quantity is equivalent to the effective molar absorptivity, E,~, for a particular aromatic-TCNE complex. For the naphthalene-TCNE solutions with carbon tetrachloride as the solvent, the AA/AC value for most 1,3-diene systems is 1050 &50 1 * mole-’ - cm-’ whereas with chloroform as the solvent the sensitivity is significantly lowered to 670 f 50. With the mesitylene-TCNE system, see is 3.20 sf O-50 x lo3 for the three 1,Zdiene systems studied with carbon tetrachloride as the solvent. Practically, the lowest possible detection limit is 25 x 10”M for the naphthalene-TCNE system and 2.5 x 10’ 6M for the mesitylene-TCNE system, although measurements were not made at these levels. These concentrations correspond to I -1 and O-11 ppm, respectively, for cyclopentadiene in carbon tetrachloride. Nature of 1,3-diene-TCNE complexes Absorption spectra. Transient colours are observed in the Diels-Alder reactions of 1,3-dienes with TCNE; however, some of the 1,3-dienes do not readily undergo the DielsAlder reaction but do form relatively stable complexes with TCNE (see Table 3). The spectra of some of these complexes (Fig. 3) show an increased bathochromic shift of about

Waveleoqth.

nm

Fig. 3. Spectra of Some 1.3~diene-tetracyanoethylenecomplexes. (1) cis-1.3-Pentadiene. 10-4M. (2) 4-Methyl-1,3-pentadiene, 10-4hf. (3) 2,5-Dimethyl-2.4-hexadiene, 10-4hf.

Determination of Diels-Alder-active 1,3-dienes

1093

60 nm with each additional methyl group or the equivalent, compared to the parent 1,3-diene. Estimations of Kmand E values for aliphatic I,3-dienes. The spectra of the 1,3-diene-TCNE molecular complexes are very similar to those found by Merrified and Phillips’ ‘I for x-complexes of aromatic compounds with TCNE. In order to determine the magnitude of the n-complex formation constant, K, , a plot of &,,,,/A vs. 1/CoTmE was made. The slope of this plot is equal to l/(K, * E) and the intercept is equal to the reciprocal of the molar absorptivity of the n-complex. Usually the Benesi-Hildebrand equation would predict a linear plot of C”rcNE/A vs. 1/Ndonor3 where honor is the mole fraction of the donor. However, when the donor is an aliphatic 1,3-diene, small concentrations of TCNE and an excess of the donor will force an addition reaction of some type to occur, and the result is a marked fading of the Ir-complex colour. If the TCNE-mesitylene n-complex values for K, and E from Table 1 are used as standards, meaningful plots for the TCNE-diene complexes and the mesitylene-TCNE x-complex can be made as shown in Fig. 4. Similar plots were obtained for cis-pentadiene&

I/C TCNE

Fig. 4. Comparison of 1,3-diene and mesitylene-TCNE complexes. (1) Mesitylene. (2) 2,5-DimethyL2,4iene.

methyl-1,3-pentadiene and cycle;octadiene. The mole-fraction term becomes meaningless in these plots for the system described since the solvent concentration is so large that the only real variable is the concentration of TCNE. Hence it was decided to use just l/C”,cm . If the assumption is made that s1 ,3diene N E,,,ityienc, then the slope of the line is equivalent to the reciprocal of K, . The molar absorptivity assumption is valid for these crude measurements since E for the series of benzene and methylated benzene derivatives is about 3.20 x 10’ 1. mole-’ * cm-‘, and the slopes for the aliphatic 1,3-dienes studied coincide closely with the mesitylene slope in Fig. 4. If the molar absorptivity of the aliphatic 1,fdienes is smaller than E for mesitylene, then the K, values reported here would be increased, not decreased. The K, values for the aliphatic 1,3-dienes listed in Table 3 are not absolute; they indicate only the relationship of one 1,3-diene to another and the similarity of the l,fdiene-TCNE rr-complexes to their aromatic analogues. The K, values reflect an increase in complex formation with an increase of methyl substituents in the diene. This is similar to the behaviour of aromatic-TCNE systems. The most salient feature of this study is to show that there exist finite K, values for the I,3 diene-TCNE systems.

1094

DALEA. WILLMU and GEORGE H.

SCHENK

The transient colour in the DiebAlder reaction. Since the existence of the 1,3-dieneTCNE z-complexes has now been established, the transient colour in the 1,3-diene-TCNE Diels-Alder reaction may be understood by considering the cisoid and transoid conformations of the aliphatic 1,3-dienes:

Cisoid

Transotd

A cisoid conformation of the 1,3-diene is necessary for the formation of a Diels-Alder reaction adduct;2’ however, if a transoid conformation of the aliphatic 1,3-diene is formed, no Diels-Alder adduct is observed. For aliphatic 1,3-dienes with free rotation around the middle single bond, the transoid conformation exists in much larger concentrations than the cisoid conformation.22 This leads to the conclusion that if the 1,3-diene is in the cisoid conformation, as for 2-methyl- 1,3-pentadiene, the observed transient n-complex assists the rate of reaction. However, if the 1,3-diene is in the transoid conformation, the Diels-Alder reaction is inhibited since the n-complex between the 1,3-diene and TCNE has to dissociate and rotation of the aliphatic 1,Idiene to the cisoid conformation has to take place before reaction can occur. The proposed explanation can be pictured as:

NC

CN

n-Complexes offuran and its derivatives. Since furan is known to undergo a quantitative Diels-Alder reaction with maleic anhydride in about half an hour,23 and since TCNE is known to react many times more rapidly with 1,3-dienes than does maleic anhydride, it was thought that furan and its derivatives could be determined by the proposed methods. As noted by Ozolins, 24 furan does not react to form the Diels-Alder adduct as rapidly as expected. In fact, when TCNE is mixed with furan, 2-methylfuran, or 2$dimethylfuran, stable coloured x-complexes are formed as noted in Table 3. These coloured complexes have still been in existence months after they were prepared; with slow evaporation, crystals of the complexes are formed. No characterization was made of these crystals; however, an addition product or z-complex appears to be formed. Since cyclopentadiene reacts so rapidly with TCNE and maleic anhydride reacts fairly quickly with furan, it should be reasonable to assume that furan should react rapidly with TCNE. However, the opposite is true. Butler and his associates23 found that maleic anhydride forms a 1 : 1 n-complex with furan, and the equilibrium constant is about 046 as determined by ultraviolet-visible spectrophotometric and nuclear magnetic resonance

Determination

of Diels-Alder-active

IJ-dienes

1095

techniques. Yoshida and Kabayashl ‘20 found the equilibrium constant for the furan-TCNE complex to be 0.35 with an E of 2.11 x lo3 1. mole-’ * cm-’ at 445 nm. A priori, the TCNE must position itself on the furan and its derivatives in a manner that would not lead immediately to a Diels-Alder adduct. If maximum overlap produces the most stable situation, it appears that the TCNE would position itself over the furan systems in this manner: CN

Q

CN

NC

CN

rather than the Diels-Alder-active

position : NC

CN

R

NC

0

CN

This proposed position would allow the electron-deficient TCNE to share the electron-rich oxygen atom, which might be explained on the basis of resonance and a co-ordinate covalent bond. Acknowledgemenf-One for part of this work.

of the authors (D.A.W.) wishes to thank Thiel College for use of their instruments

REFERENCES 1. 2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

S. F. Birch and W. D. Scott, Znd. Eng. Chem., 1932,24,49. A. Mora and E. Blasco, Inst. Nacl. Tee. Aeronaut. Madrid Cornman., 1944, No. 4, 25. S. Jifuku and H. Suzamara, Coal Tar (Japan), 1956, 8, 316. G. H. Schenk and M. Ozolins, Talanta, 1961,8, 109. T. L. Cairns, R. A. Carboni, D. D. Coffman, V. A. Engelhardt, R. E. Heckert, E. L. Little,E. G. McGeer, W. J. Middleton, R. M. Scribner, W. C. Theobald and H. E. Winberg, J. Am. Chem. Sot., 1958, 80, 2775. W. J. Middleton, R. E. Heckert, E. L. Little and C. G. Krespan, ibid., 1958, 80, 2783. C. A. Stewart, Jr., J. Org. Chem., 1963,28, 3320. A. P. Terent’ev, Org. Chem. Znd. (USSR), 1937,4, 532. K. Uhrig, E. Lynch and H. Becker, Znd. Eng. Chem., Anal. Ed., 1946, 18, 550. J. S. Fritz and G. E. Wood, Anal. Chem., 1968, 40, 134. H. S. Knight, ibid., 1958, 30, 9. E. B. Clairborne, H. M. Davis and C. A. Rivet, ibid., 1956, 28, 1104. A. A. Polyakova, R. A. Khmel’nitskii, F. A. Medevedev, and K. I. Zinimer, Khim. i Tekhtzol. Topliv i Masel, 1962, 7, 59. D. Joly, 2. Anal. Chem., 1968,236, 259. J. M. Miller and D. D. DeFord. Anal. Chem.. 1957.29.475. H. A. Benesi and J. H. Hildebrand, J. Am. &em. ioc.; 1949,71,2703. R. E. Merrified and W. D. Phillips, ibid., 1958, 88, 2778. C. N. Reilley and G. P. Hildebrand, Anal. Chem., 1959, 31, 1763. H. H. Willard, L. L. Merritt, Jr. and J. A. Dean, Instrumental Methods of Analysis, 4th Ed., p. 91. Van Nostrand, Princeton, 1965. Z. Yoshida and T. Kabayaski, Tetrahedron, 1970,X, 267. A. Wasserman, Diels-Alder Reactions. Elsevier, London, 1965. G. S. Szasz and N. Sheppard, Trans. Faraday Sot., 1958,49, 358. G. B. Butler, J. T. Badgett and M. Scharabash, J. Macromol. Sci., 1970, 4, 51. M. Ozolins, M. S. The&, Wayne State University, 1961.

1096

DALEA. WIUUM~ and GEORGE H. SCHENK Zrmamman~Es wurde eine indirekte spektrophotometrische Methode zur Bestimmung van elf 1,3-Dienen im Ko~trationsbemich 0,05-I,00 - 10e3M entwickelt. Se beruht auf der raschen Dieis-Alder-Reaktion zwischen cisoiden 1,3-Dienen und Tetracyanlthylen (TCNE) und der Zerst&ung eines n-Komplexes zwischen einem Aromaten und TCNE. Die Diene waren: Cyclopentadien; 1,3-Cyclohexadien; trans-lf-Rentadien; Z,CDimethyl-1,3-pentadien; trans2-Methyl-1,3-pentadien; 2-Methyl-1.3-butadienn; 9-Methylanthracen; 9,10-Dimethylanthracen; 1.6Diphenyl-1,3,5_hexatrien; 2,EDimethyl-1,3-butadien; 1.4Diphenyl-1.3-butadien. Drei 1,3Diene wurden im Bereich 0.05-l * lo-•M bestimmt: Cyclopentadien, tram-2-Methyl-1,3-pentadien und Anthracen. Die Nachweisgrenze fur Cyclopentadien in Tetrachlorkohlenstofflosung betrtigt 0,ll &ml. Vierzehn 1,3-D&e bilden stabile Ir-Komplexe und lie&n sich nach der vorgeschlagenen Methode nicht bestimmen. Fur diese 1,3-Diene werden die Spektren van einigen der rr-Komplexe angegeben; d&u wurden die relativen Gleichgewichtskonstanten der +Komplexe von 2,5-Dimethyl-2,4-hexadien, cis-1,3-Pentadien, 4-Methyl-1.3-pentadien und 1,3_Cyclooctadien abgesch&zt. Es wird eine Erkliirung fur die vortibergehende Farbe bei der Diels-Alder-Reaktion zwischen 1,3-Dienen und TCNE vorgeschlagen. R&sum&Gn a developpe une m&hode spectrophotometrique indirecte, bas& sur la reaction de Diels-Alder rapide entre des 1,3-d&nes cisoldes et le t&racyanCthylene (TCNE) et la destruction d’un complexe 7r aromatique-TCNE pour doser onze 1,3-dienes dans le domaine O,OS-1,OO x lo-) M. Cea di&nes sent: cyclopentadi≠ 1,3_cyclohexadi≠ rrans-1,3-pentadiene; 2,4din&thy1 1,3-pentadiene; rrmu-2-methyl 1,3-pentadiene; 2-methyl 1,3-butadiene; Pmethylanthracene; 9,lOdim&hylanthra&eractne;l$diph&nyi 1.3.5.hexatriene; 2,3-dim&hyl 1,3-butadi≠ et 1,4diphCnyl 1,3-butadiene. Trois 1,3-dienes ant 6tC dCtermi& dans le domaine 0,054 x 10~*M: cyclopentadiene, tronr-2-methyl lJ-pentadiene, et anthracene. La limits de detection pour le cyclopentad&e dans des solutions en tttrachlorure de carbone eat 0,ll &ml. On a trouve quatorze 1,3-dibcs formant des complexes w stables et qui ne peuvent pas etre determines par la m6thode propos&. Pour cm 1,3_dienes, on rapporte les spectres de quelques-uns de-scomplexes P; de plus, on a e&m& les con&antes d%quilibre relatives pour les complexes 7r des 2.5-dim6thyl 2&hexadiene, cis-1,3-pentadiene, emethyl 1,fpentadiene et 1,3-cycleoctadiene. On sugg& une explication de la couleur fugace dam la reaction de Diels-Alder IJdiene-TCNE.