- Email: [email protected]
Charge-transfer complexes in organic chemistry—XII
Ttrmkdm
Vol. 31. pp. 13 to 141. Pagumn Pm5 1575. tinted in Great kitam
CHARGE-TRANSFER COMPLEXES IN ORGANIC CHEMISTRY-XII SOLVENT
EFFECTS
ON
CHARGE-TRANSFER
SPECTRA’
S. DIJPIRE,J. M. MULINDABYUMA, J. B.NAGY and 0. B.NAGY* Universite
Catholique
de Louvain, Laboratoire de Chimie C&&ale Place Louis Pasteur, B-1348 Louvain-la-Neuve,
(Receivedin
the UK, 20 June 1974; Acceptedforpuhlication
et Organique; Belgique 12 August
Institut
de Chimie,
1974)
Abatraet-The charge-transfer transition energies of 2.6dimethoxynaphthalene and 9-methylanthracene with tetrachlorophthalic anhydride, and acenaphthene and anthracene with 3,Sdinitrophthalic anhydride were measured in sixty aprotic solvents. The observed effects can be interpreted in terms of various solvent parameters if the solvents are divided into the following classes: halogen-containing, aromatic and n-donor solvents.
In the present work we extend our solvent effect studies on charge transfer spectra in order to confirm and generalize our previous results.* Hildebrand et al. have pointed out that aromatic solvents give a different straight line than the halogen containing solvents, when the charge-transfer absorption maxima of the complex solvent-iodine is plotted against the ionization potential (IO) of the rolvents.’ Later it was recognized that n-donor molecules give another relation for the same type of plot.’ The more complete work of Voigt showed finally the following classes: alkanes, alkyl halides, aromatics, oxygen compounds, sulphur compounds and nitrogen compounds.‘3.Y It seems that similar differentiation occurs when one CT complex is examined in different solvents.* Voigt’ found that hydrocarbons and perfluorohydrocarbons form two different classes, when the charge-transfer maxima of several complexes aromatics-tetracyanoethylene (TCNE) are plotted vs the polarizability parameter (n’ - l)/(2n* + I), n being the refractive index. Foster and Emslie had to distinguish between two type of solvents, protic (alcohols) and aprotic ones (chlorine-containing, hydrocarbon).6 Ilmet and Berger’ have also emphasized that aromatic solvents did not follow the relationship h,.,/n for the complexes of aromatic donors with 3-nitro-l,&naphthalic dianhydride and 1,4,5,&naphthalene tetracarboxylic dianhydride. Special effects due to aromatic solvents were invoked. (Let us notice that the theoretically correct relationship should be imar vs (n2 - l)/(2n2 + I), where V,, is the wave number in cm-‘; nevertheless the conclusion of the authors remains true). We have found in the case of the complex
acenaphthene (ACE)-tetrachlorophthalic anhydride (TCPA) that different classes of solvents must be distinguished in order to understand the solvent effect on the absorption spectra: chlorine-containing (class a), aromatic (class b) solvents which influence the chargetransfer absorption band mainly by general polaritypolarizability effect, while the ndonor solvents (class d in Ref 2) are acting mainly by their free electron pair. A theoretical explanation was proposed for the latter class. The aromatic solvents containing “one functional” group cquld not be classified clearly (class c in Ref 2). In order to check the validity of the interpretation and the validity of the proposed classification we have extended the study to the complexes 2,6_dimethoxynaphthalene (2,6DMON)-TCPA, 9methylanthracene (9MANT)-TCPA, ACE-3,5dinitrophthalic anhydride (3JDNPA) and anthracene (ANT)-3JDNPA. We also include a review of the limited literature data. We have found that the interpretation previously given2 is fully satisfied, but the classification of the solvents into the above mentioned classes must be modified. This conclusion emphasizes further that specific solvent effects do play a great r6le at least in the case of complexes where the excited state, is more polar than the ground state. RJSUL’TSAND DISCUSSION The
spectroscopic data of the four complexes are reported in Table I. The stoichiometry of the complex ACE-3, SDNPA was established as I : I,8 and that of the other complexes was assumed to be I : I on the basis of the results obtained with similar donors.‘” i&, did not follow the theoretical (eo- l)/(2ra + I) parameter” (e. is the dielectric constant) of polarity, neither did the polarisability parameter (n’- 1)/(2n’ + I).” At present only empirical parameters, such as the Z ‘,“.” E&“*” of S,.,” values are convenient to character-
tZ value is the energy of the intermolecular charge-transfer in kcal mol.’ of lethyl4-carbomethoxypyridinium iodide at 25°C. SEr value is the energy, in kcal mol-‘, of the intramolecular charge-transfer band of pentaphenyl pyridinium-N-phenol betaine at 25°C. I35
2.653
3.343 -
4.556 4681 -
3.255 2978 2.716 2.740 2.380 2.845 3.845 394Y 4.369 4.ooo
20.2 21.05 20.5 -
-
4643 3.869 2.892 2.477 2.732 2973 I .778 2987
IO II 12 I3 I4 I5 I6 Class I7 I8 I9 20 21 22 23 24 25 26 Class 27 28 29 30 31 32 33 34 35 36 37
<6 39 -
-
2800 31.50 3150 2700 2650 -
22.25 22.3 21.9 22.0 -
2200 2300 2300 -
19.9 20.05 19.95 -
19.0 18.75 19.8 20.05 -
18.6 18.4 18.4 18.4 -
18.5 18.6 18.6 18.6 -
19.05 18.5 18.4 19.05
2&O 2150 2150 2100 2350 2200 2200 2250 -
18.25 -
18.1 18.2
18.45 18.25 18.6 18.6 19.0
-
a%.. fkK)
2&o 2850 2600 -
2500 2350 2Hw) 2650 -
2150 2150 2300 2400 -
2100 2100 2050 2050 -
2200 2300 2tXm 1750 2250
2450 2ooo
2&l 2300 1750 2303 25fm
(cm-7
WC)
22; 22.4 22.4
22,6 22.6 22.35 22.4
21.75 -
21.9 22.05 22.11 22.05 22.2 21.9 21.8 21.65 21.55
2i -
21.55 22.25 22.1 21.8
21.4 21.6
2G 21.5 21.7 21.8 22.0
k..
2400 2600 -
2250 -
2200 26J30 2300
2100 2250 2wl 2400 zoo0 2300 -
2200 -
2500 2750 If!&3
2250 2200
2100 2100 1700 2200 -
km-‘)
ANT-3,SDNPA 9MANT-TCPA -_ y,n..I,,, v",..VW
2350
2450 2250
2300 2400 23tm 2100 2400 2450
(cm-‘)
21.1 -
20.5 20.35 20.4 20.25 -
4
53 21 85 83 75
-2 -I -II
-
20.4 20.5 20.6 -
0 2
20.4 20.65 21.1
20.2 20.35
-
4940 3.547
8 9
21.10 20.35 20.2 20.75 20.6 20.85
-
- I2 - I7 -21 2 -
(a) nChlorobutane Dichloromethane Chloroform Carbon tetrachloride 1.2~Dichloro ethane I,l-Dichloro ethane l.I,2,2-Tetrachlorwthane Pentachlorotthane I ,2,3-Trichloropropane nChlorohexane Chlorocyclohexane Dichloro ethylene trans I,l-Dichloro ethylene Trichloroethylene Tettachlomethylene t-Butylbromide fb) Benzene Toluene m-Xylene p-Xylene Mesitylene Ethylbenzene Chlorobenzene Bromobenzene Benzylchloride o-Dichlorobenzene (c) Benzaldehyde Nitrobenzene Propionaldehyde Butyrafdehyde Crotonaldehyde Valeraldehyde Caproic aldehyde Methylacetate Ethylacetate lsobutylacetate n-Butylacetate
3.114 4447 4.1 I4 2.146 4.584) 3.914
V,., (kK)
Class I 2 3 4 5 6 7
(cm-‘)
v",..Y -1/*
ACE-3. SDNPA
sts,”
Al&"
I. Spectroscopic data of the complexes in different solvents at 20°C
N" Solvent
Tahle
23.8 23.8 23.7 23.8 23.6 23.7 23.75 23.8 23.8
23.1 -
23<5 23.0 23.0
23.25 23.3 23.4 23.5 23.3
23-1 -
23.1 23.4 23.3 23.2 23.35
22.95 23.1
24.7 23.2 23.1 23.3 23.2 23.4
V,.. fkK)
2500 2400 2500 2400 2400
2300 2u)o 2400 2400
2450 23al 23lm 2500
2200 -
2450 2500 2400 2450 -
2500 2400
2450 2500 2400 2450 2400
-
(cm-‘)
2,6DMON-TCPA - it12 Y,"..
2900 2850 2800
21.4 21.65 -
21.3 21.05 20.9 20.3
-
0.845 I.301 2.398
-
2s
21.0
64 56 66 107 II3 I41 38
4.176 4.623 4.204 5.778 4.6d 5.204 4.591 2350 1800 zoo0 2050
3&o 2650 -
6
71 90 22.6 21.45 -
2300 2250 -
93 77 49
21.3 21.2 -
78
-
-
-
-
-
3.568 4.670 4.079 5.040 -
2.079 16&l 3.435 3.462 -
19.1 18.8 18.4
2150 1850 2150
-
2500
19.1 -
2600
19.65 -
2450 3200 -
2550 2350 3Mm 2750 2550 2600
2800 -
20.35 19.65 19.2 20.2 19.95 19.4 19.7 20.35 19.2 20.15 -
2250 -
-
19.25 -
-
2550 2600 2350
22?5 22.9 2&) 1600
21.2 20.8
-
2soo -
2150 -
2300 2400
2250 -
2G -
22.3 -
22< 22.6 -
22.55 -
“Refs 14, 25 and 26. bRef 30. ‘G. Berrebi, These de doctorat, Louvain, 1970. ‘Estimated value from the E,/S, correlation (as a proof, see Ref 1, where 41 was estimated as 3.5).
38 Isopropylacetate 39 Diethylether 40 Dibutylether 41 1.2~Dimethoxyethane 42 Tetrahydrofuran 43 2-Methyltetrahydrofuran 44 Tetmhydropyran 45 I/t-Dioxan 46 Acetonitrile 47 Capronitrile 48 Nitromethane 49 2-Nitropropane 50 Acetone 5I Acetophenone 52 Cyclohexanone 53 Dimethylformamide 54 NN’Dimethylacetamide 55 Dimethylsulfoxide 56 Benzonitrile Miscellaneous 57 n-Pentane 58 Cyclohexane 59 Cyclohexene 60 Carbon disulphide 2900 ww)
24; 23.25
23.2 -
-
2700 -
24.0
-
2900 -
2500 -
2600 2600 -
-
23.8 -
23.5 -
23.5 23.7 -
23.7 23.8 23.85 -
23.8 -
S. DUPIREet al.
138
ize the variation of I& of a complex. These parameters include both the polarity and the polarizability factor. SMis defined as log kz, where k2is the second order rate constant of the reaction of tri(n-propyl)amine with methyliodide.‘.” This parameter is used because of the great number of solvents available. Let us note that all these three solvent parameters vary in parallel with each other. Fig 1 shows the variation of &.I with SM for the complex .ACE-3JDNPA in the halogen (chlorine)containing solvents. There is a red shift with increasing polarity. It can be seen that two classes must be distinguished in these solvents: the per- or polyhalogenated solvents and those which are monohalogenated solvents or with few halogen substituents. In the former case, cm. varies less with SM,there is a sort of “saturation effect”, while in the latter case, the variation of i;, is more pronounced.* Similar variations can be found for the three other complexes. Thus the phenomenon must be rather general. Voigt’s results of the complex pdimethoxybenzeneTCNE show similar splitting into two classes of the halogencontaining solvents (only Iln values are reported, because of the very large absorption maxima; a nondependence of V,., - filn with solvent is assumed).16 We have reported in Fig 2 the variation of & vs SMin the aromatic solvents for the same complex. We can see the bathochromic shift with increasing polarity, as may be *In the previouslyreported results’ carbon tetrachloride and chloroform were exceptions in the relation LX vs S, for the complex acenaphthene-TCPA. The small number of solvents studied in that case did not permit us to interpret this particular behaviour. tin the plot of the absorption maxima of iodine vs 1” (the donor ionization potential) cyclohexane, n-heptane were also correlated with the aromatic solvents.’ SFor abbreviations see Table 2.
I6
2
expected when the ground state is less polar than the excited state.’ The same relation holds for the other three complexes. Chlorobenzene,s bromobenzen&” benzyl chloride.X and o-dichlorobenzene26 follow the aromatic solvents as well as cyclohexane” and cyc1ohexene.t” The sensitivity of fi,, vs SMis given by the slope of the straight lines (Table 2). They were calculated with 1,, and Su expressed in kcal mol-’ units and finally standardized against the Z values (in kcal mol-‘), using the known slopes of the relationships ET/!&,(1.62) and Z/ET (1*37).‘s We have chosen the Z values as reference, because they represent the “maximum” solvent effect at present. This is a particular case where the ground state is very polar and the excited state is much less polar. The dipole moment of the complex not only diminishes when going from the ground to the excited state, but it also changes direction (approximately 90”).These two effects make the solvent effect very large.” (However the slope of I,, in kcal mol-’ of the charge-transfer maxima of tropylium iodide vs Z is 1.05, which casts doubt on the effect of the change in direction of the dipole in the reference pyridinium ion.“) The slopes &,.JZ show that the solvent effect is the greatest in the halogenated solvents (slope - 0.4), smaller in the aromatic solvents (slope - 0.3) and the smallest in the per- or polyhalogenated solvents (slope -0.2). The variation of I,,,., in the case of complexes where the ground state is less polar than the excited state, is much less than in the opposite case (Z value). While for the iodide-pyridium complex”“’ the polar and polarisability effects overshadow the small specific effects, in our case these latter effects do separate the solvents into different classes. The available literature data confirms this statement. However it must be emphasized that a small number of solvents studied can lead to an erroneous conclusion. The slope L, vs Z is positive for the complex HEB.“=’ which would indicate that the complex is more
4
3
3
5+5, Figure I. Dependency
of k,,,
of the complex ACE-3JDNPA on S, halogenated(---) solvents.
in per- or polyhalogenated
(-)
and
Charge-transfer
complexes
in organic chemistry-XII
139
5+s, Figure 2. Variation
of I,..
of the complex
ACEJJDNPA
polar in the ground state. This is contrary to other aromatic-& complexes studied in detaiLn It is probable that the general polarity-polarizability parameters (Su, Z, Er, . . . .) are not adequately used in this case. If V,,,.,is plotted against A&*, a positive slope (-20) is obtained showing that the n-donor property of the halogenated solvents play the more important role. The slope of I,, in eV vs IO(eV) is -0.2, showing again the specific character of the solvation by these solvents. Let us emphasize, that the solvent effect on the absorption of iodine in the same solvents is very similar, where the n-donor character of these solvents are mainly responsible for the shift.’ While direct relationship I,, vs Z gives a slope near unity for the complexes DMA-TNB (084) and DMA-TCNB (099):’ the slope obtained initially with & (or E+) and recalculated vs Z gives 0.2 and 0.3 respectively. Those very large slopes cannot be explained easily for these complexes where the ground state is less polar. The main reason for this discrepancy is the inclusion of “anomalous” Z values (/) of cyclohexane, heptane, hexane and 2,2,4-trimethylpentane) in the correlation. Without these solvents, the correlation for the halogen containing solvents alone gives slopes equal to -0.3 and -0.2 respectively. Voigt’s results16 present also similar anomalies: while plotted directly vs Z, the slope for the complex oDMOB-TCNE is 0.7, it is only 0.35 with Sh(or Er values. These latter parameters have the advantage to give a greater number of solvents. These remarks show the precaution to take when only one solvent parameter is used with very limited number of solvents. Since specific effects play a great role in the case of complexes where the ground state is less polar than the excited state, well chosen solvents are needed (at least IO of them) in order to establish a correlation. This must be emphasized, because literature data abound in partial *A& = i, - YR where is is the stretching vibration frequency (O-D) of deuterated methanol in solvent S and Vs is the same vibration in the reference solvent benzene. AI, is thus the n-donor parameter.M
with S, in aromatic
(0) and miscellaneous
(0) solvents.
results with partial conclusions which are often erroneous. Let us note finally that the slopes of I,, vs Z for the n-n* and n -n* transitions of a/I conjugated 9decalenone are respectively -0.25 and +O*3L” The solvent effect on simple molecules or on complexes is rather very similar, as it has been noted by Symons and Davis.29 Because the Franck-Condon state is directly reached in the excited state, the magnitude of the solvent effect vs Z does not give the amount of charge-transfer in the excited state as it was proposed for the reaction rates,% where the activation complex is supposed to be in thermal equilibrium with environment. In Fig 3, I,.. is plotted against Aio for the complex 2,6 DMON-TCPA. The correlation is very good and this holds also for the complexes ACE3JDNPA and 9MANT-TCPA. In the case of n-donor solvents, their n-donor ability is the more important factor as it was shown previously: these solvents cause a blue shift proportional to their n-donor ability. A theoretical explanation was already given in Ref 2. For the complex ANT-3, SDNPA, the points are rather scattered in the region above A&=75 cm-‘. Aromatic solvents with n-donor groups (Table I: 27, 28, 51 and 56) do not follow the straight line obtained for n-donor solvents. They seem to follow a similar correlation with A& displaced somewhat to lower wave numbers from the “purely” n-donor solvents, because of the polarisable aromatic ring. Let us emphasize that these same solvents do not follow the I, vs S, relationship. For three complexes the sensitivity of &., to the n-donor parameter is about the same (slope - 5), and for the other two, the sensitivity is about twice as great (slope -IO). The reason for this phenomenon is not clear at present. The slope of aromatic n-donor solvents is similar to that of n-donor solvents, showing that a special property (n-donor ability) is added to the general effect (polarity-polarisability of the aromatic ring). Let us emphasize that literature data also give slopes equal to -5 or -10.
S. DUPLRE et al.
140
Figure 3. Relation between i,..
of the complex 2,6DMON-TCPA solvents.
and the n-donor solvent parameter AV, in n-donor
Tahle 2. The values of the slopes 5,. (kcal mall’) vs Z &al mol-‘) in the case of per- or polyhalogenated (class al), halogenated (class aI1) and aromatic (class b) solvents. The values of the slopes G,. (cm-‘) vs Al, (cm-‘) for the n-donor solvents (class c) Slope Slope &Jz
~,../A~,
Ref
Class (a) Complex ACE-3, SDNPA ANTJ, SDNPA 9-MANT-TCPA 2,6DMON-TCPA ACE-TCPA p-DMOB’-TCNE o-DMOBd-TCNE HMB’-TCPA HEB’-I, DMA”-TNB’ DMA-TCNB’ ACE-CA’ PYR’-CA HMB-TR” ANT-TR PYR-TR ANT-A” LiBr-TCNE HMBCA HMB-NTDAP ANT-NTDA HMB-NTDA PER’-NTDA 3.4BPYR’-NTDA PYR-NNA’ ANT-NNA 3,CBPYR-NNA
I -0.16 -0.16 -0.14 - 0.07 -0.3 - 0.35 + 0.2 -0.2 -0.3 -0.2 -0.2 -
- 0.05 - 0.05 -
II -
Class (b)
0.35 0.45 0.35 0.45 0.45 0.45 -
-0.2 -0.2 -0.25 - 0.25 -0.45 -0.45 -0.25 -0.55’ -
Class (c) - IO - 5(S)” II.5 5.1 4.8 - 20(-0.2) -
-
5 - IO - IO
-
59 -10 -
-0.5 -0.25 -0.25 -0.3 -0.3 -0.25 -0.4
-
0
0
2 16 I6 I8 19.20 21,32 21 21 21 27 27 27 27 28 29 29.31 7 7 7 7 7 7 7
Note: the slopes calculated from literature values are very approximate because of the paucity of data available. “This work; ‘the value in parenthesis is for aromatic n-donor solvents; cp dimethoxybenzene; “odimethoxybenzene; hexamethylbenzene; ‘in both emission and absorption; ‘hexaethylbenzene; ‘Ndimethylaniline; ‘sym trinitrobenzene; ‘I, 2, 4, 5tetracyanobenzene; *chloranil; ‘pyrene; mtropylium tetrafluoroborate; “N-methylacridinium perchlorate; “slope obtained with V, (eV) vs IO, the ionization energies of the solvents(‘); ’ I, 4, 5,8naphthalene tettacarboxylic dianhydride; ‘perylene; ‘3,4benzopyrene; ‘3-nitrol,8-naphthoic anhydride. l
Charge-transfercomplexesin organic chemistry-Xl1 We can conclude from the variation of p, in different solvents, that the following classification is suggested: class (a) halogenated solvents, group I: per- or polyhalogenated solvents; group II. monohalogenated solvents, class (b) aromatic solvents, class (b) n-donor soloents, group I: “pure” n-donor solvents, group II, aromatic n-donor solvents. This classification seems to be rather general.’ We have also examined the variation of the half-band
width (I,, - ij,,J of the CT band with the solvent (p,n is the wavenumber at the half intensity of the band at its bathochromic side). The parameter Shl shows a much better correlation, but differ from complex to complex. The aromatic solvents show an increasing tendency of I,., - p1,2with Ski in the case of the complexes ACE3JDNPA and ANT-3, SDNPA. In the case of ACE-TCPA an opposite behaviour was observed previously.’ The complex 2,6DMONTCPA does not vary with Sk(, while the complex 9MANT-TCPA shows a complete dispersion of points. The halogen-containing solvents do not give any correlation with SH for 9MANT-TCPA and ANT3,5DNPA, there is no variation in the case of 2,6DMONTCPA, while the half-band width increases with Ski for ACE3,SDNPA. No correlation seems to exist between r&- i& and Ai% of n-donor solvents except for the complex 2,6-DMON-TCPA (positive slope). For these solvents the molar volume turns out to be a much better parameter in all cases (negative slope.). It seems that the smaller the solvent molecules the better they solvate the complex.’ This causes a loosening of the complex, hence the greater value of the half-band width. At present, the behaviour of a&,,- pin as a function of various parameters cannot be satisfactorily explained.’ EXPERlMENTAL Materials. Tetrachlorophthalic anhydride (Fluka) was purified by several sublimations, mp 255°C. 3,Sdinitrophthalic anhydride was obtained from the corresponding di-acid by dehydration with acetic anhydride (mp 163.3”c). It is-handled under nitrogen. The d&acid (mp 226°C) was synthesized from the 3,Sdinitrotoluic acid by oxidation with nitric acid (d = 1.15) in sealed tubes at 143°C. Acenaphthene(Fluka) was recrystalliz.ed,five times from aqueous ethanol. rnD95-%“C. Anthracene(U.C.B.) was molten with solid KOH and’ fractionally crystallized from petroleum ether, mp 216°C. 9Methylanthracene was recrystallized thrice from methanol, mp 80-8l”C. 26Dimethoxynaphthalene (Aldrich) was washed with aqueous sodium hydroxide and recrystallized from aqueous ethanol (mp l52-153°C). The solvents were purified by normal methods and were dried on molecular sieves. Measurements. The spectrophotometric measurements were carried out with a Unicam SP 800 Spectrophotometer at 2O*O.l”C. The wave number scale was standardized against a holmium filter.
141
BIBLIOGRAPHY ‘Part X. J. B. Nagy, 0. B. Nagy and A. Bruylants, Bull. Sot. Chim. Beig. 83, 163 (1974) ‘Part VI. 0. B. Nagy, J. B. Nagy and A. Bruylants, I.C.S. Perkin II 968 (1972) ‘See R. S. Mulliken and W. B. Person, Molecular Cumplexes. A Lecture and Reprint Volume, Wiley, New York. I%9 ‘J. Walkey, D. N. Glew and J. H. Hildebrand, J. Chenr. Phys. 33. 621 (1960) ‘E. hf. Voigt, 1. Phys. Chem. 70, 598 (1966) ‘P. H. Emslie and R. Foster, Rec. Trau. Chim. 84. 255 (1965) ‘I. Ilmet and S. A. Berger, 1. Phys. Chem. 71, 1534 (l%7) ‘Part XIII. S. Dupire, J: B. Nagy. 0. B. Nagy and A. Bruylants. J.C.S. Perkin II 478 (1974); S. Dupire. These de Licence. Louvain. 1971 ‘Part VIII. J. B. Nagy, 0. B. Nagy and A. Bruylants, Bull. Sot. Chim. Be/n. 82. 337 (1973): J. B. Nagy, Ph.D. Thesis. U.C.L., Louvain, i970 -_ “‘See for example W. Liptay in Oprische Anregung orgonischer Systeme, Verlag Chemie, Weinheim/Bergstr., 1966 “E. M. Kosower, 1. Am. Chem. Sot. 80, 3253, 3260 (1958) “K. Dimroth, Chr. Reichardt, Th. Siepman and F. Bohlman, Ann. 661, I (1963) “K. Dimroth, Chr. Reichardt and A. Schweig, Ann. 669, I (l%3) “J. C. Jungers, L. Sajus, I. de Aguirre and D. Decroocq, L’Anulyse cinbque de la transformation chimique, Technip, Paris, 1968, Vol. II “E. M. Kosower, An Introduction to Physical Organic Chemistry, Wiley, New York, 1968 “E. M. Voigt, J. Am. Chem. Sot. 86, 3611, 3930 (1964) “E. M. Kosower, J. Org. Chem. 29,956 (1964) ‘J. Czekalla, K. 0. Meyer, 2. physik. Chem. 27, 185 (l%l) “H. L. Browning, Dissertation, Louisiana State University, 1960 =G. Briegleb in Oprische Anregung organischer Sysfeme. Verlag Chemie, Weim/Bergstr., 1966 “P. H. Emslie and R. Foster, Rec. Trau. Chim. 84, 255 (1%5) “G. Briegleb, Elekfronen-Donator-Acceptor-Komplexe, Springer Verlag, Berlin, I%1 “E. M. Kosower, J. Chim. Phys. 230 (1964) UM. Mohammad and E. M. Kosower, 1. Am. Chem. Sot. 93.2709, 2713 (1971) ?. Lassau and J. C. Jungers, Bull. Sot. Chim. 2678 (1%8) ‘“Y. Drougard and D. Decroocq, Bull. Sot. Chim. 2972 (1969) nT. G. Beaumont and K. M. C. Davis, J. Chem. Sot. B 1010(1968) ?I. Briegleb, W. Liptay and R. Fick, 2. Physik. Chem. 33, I81 (1%2); Z Elektrochem. Ber. Bunsenges. Physik. Chem. 66.85 I. 859 (1962) “K. M. C. Davis and M. C. R. Symons, 1. Chem. Sot. 2079(1965) ?. Kaaiva. Y. Sumida and T. Inoue. Bull. Chem. Sot. Japan 41, 767 (1968) “R. Foster, 1. Chem. Sot. 1075 (1960) “R. Foster and D. L. Hammick, J. Chem. Sot. 2685 (1954) “E. M. Voigt, I. Chem. Phys. 49, 3300 (1968) ‘J. H. Hildebrand, J. M. Prausnitz and R. L. Scott, Rqulur and related solutions, The solubility of Gases, Liquids, and Solids. Van Nostrand, New York, 1970
Vol. 31. pp. 13 to 141. Pagumn Pm5 1575. tinted in Great kitam
CHARGE-TRANSFER COMPLEXES IN ORGANIC CHEMISTRY-XII SOLVENT
EFFECTS
ON
CHARGE-TRANSFER
SPECTRA’
S. DIJPIRE,J. M. MULINDABYUMA, J. B.NAGY and 0. B.NAGY* Universite
Catholique
de Louvain, Laboratoire de Chimie C&&ale Place Louis Pasteur, B-1348 Louvain-la-Neuve,
(Receivedin
the UK, 20 June 1974; Acceptedforpuhlication
et Organique; Belgique 12 August
Institut
de Chimie,
1974)
Abatraet-The charge-transfer transition energies of 2.6dimethoxynaphthalene and 9-methylanthracene with tetrachlorophthalic anhydride, and acenaphthene and anthracene with 3,Sdinitrophthalic anhydride were measured in sixty aprotic solvents. The observed effects can be interpreted in terms of various solvent parameters if the solvents are divided into the following classes: halogen-containing, aromatic and n-donor solvents.
In the present work we extend our solvent effect studies on charge transfer spectra in order to confirm and generalize our previous results.* Hildebrand et al. have pointed out that aromatic solvents give a different straight line than the halogen containing solvents, when the charge-transfer absorption maxima of the complex solvent-iodine is plotted against the ionization potential (IO) of the rolvents.’ Later it was recognized that n-donor molecules give another relation for the same type of plot.’ The more complete work of Voigt showed finally the following classes: alkanes, alkyl halides, aromatics, oxygen compounds, sulphur compounds and nitrogen compounds.‘3.Y It seems that similar differentiation occurs when one CT complex is examined in different solvents.* Voigt’ found that hydrocarbons and perfluorohydrocarbons form two different classes, when the charge-transfer maxima of several complexes aromatics-tetracyanoethylene (TCNE) are plotted vs the polarizability parameter (n’ - l)/(2n* + I), n being the refractive index. Foster and Emslie had to distinguish between two type of solvents, protic (alcohols) and aprotic ones (chlorine-containing, hydrocarbon).6 Ilmet and Berger’ have also emphasized that aromatic solvents did not follow the relationship h,.,/n for the complexes of aromatic donors with 3-nitro-l,&naphthalic dianhydride and 1,4,5,&naphthalene tetracarboxylic dianhydride. Special effects due to aromatic solvents were invoked. (Let us notice that the theoretically correct relationship should be imar vs (n2 - l)/(2n2 + I), where V,, is the wave number in cm-‘; nevertheless the conclusion of the authors remains true). We have found in the case of the complex
acenaphthene (ACE)-tetrachlorophthalic anhydride (TCPA) that different classes of solvents must be distinguished in order to understand the solvent effect on the absorption spectra: chlorine-containing (class a), aromatic (class b) solvents which influence the chargetransfer absorption band mainly by general polaritypolarizability effect, while the ndonor solvents (class d in Ref 2) are acting mainly by their free electron pair. A theoretical explanation was proposed for the latter class. The aromatic solvents containing “one functional” group cquld not be classified clearly (class c in Ref 2). In order to check the validity of the interpretation and the validity of the proposed classification we have extended the study to the complexes 2,6_dimethoxynaphthalene (2,6DMON)-TCPA, 9methylanthracene (9MANT)-TCPA, ACE-3,5dinitrophthalic anhydride (3JDNPA) and anthracene (ANT)-3JDNPA. We also include a review of the limited literature data. We have found that the interpretation previously given2 is fully satisfied, but the classification of the solvents into the above mentioned classes must be modified. This conclusion emphasizes further that specific solvent effects do play a great r6le at least in the case of complexes where the excited state, is more polar than the ground state. RJSUL’TSAND DISCUSSION The
spectroscopic data of the four complexes are reported in Table I. The stoichiometry of the complex ACE-3, SDNPA was established as I : I,8 and that of the other complexes was assumed to be I : I on the basis of the results obtained with similar donors.‘” i&, did not follow the theoretical (eo- l)/(2ra + I) parameter” (e. is the dielectric constant) of polarity, neither did the polarisability parameter (n’- 1)/(2n’ + I).” At present only empirical parameters, such as the Z ‘,“.” E&“*” of S,.,” values are convenient to character-
tZ value is the energy of the intermolecular charge-transfer in kcal mol.’ of lethyl4-carbomethoxypyridinium iodide at 25°C. SEr value is the energy, in kcal mol-‘, of the intramolecular charge-transfer band of pentaphenyl pyridinium-N-phenol betaine at 25°C. I35
2.653
3.343 -
4.556 4681 -
3.255 2978 2.716 2.740 2.380 2.845 3.845 394Y 4.369 4.ooo
20.2 21.05 20.5 -
-
4643 3.869 2.892 2.477 2.732 2973 I .778 2987
IO II 12 I3 I4 I5 I6 Class I7 I8 I9 20 21 22 23 24 25 26 Class 27 28 29 30 31 32 33 34 35 36 37
<6 39 -
-
2800 31.50 3150 2700 2650 -
22.25 22.3 21.9 22.0 -
2200 2300 2300 -
19.9 20.05 19.95 -
19.0 18.75 19.8 20.05 -
18.6 18.4 18.4 18.4 -
18.5 18.6 18.6 18.6 -
19.05 18.5 18.4 19.05
2&O 2150 2150 2100 2350 2200 2200 2250 -
18.25 -
18.1 18.2
18.45 18.25 18.6 18.6 19.0
-
a%.. fkK)
2&o 2850 2600 -
2500 2350 2Hw) 2650 -
2150 2150 2300 2400 -
2100 2100 2050 2050 -
2200 2300 2tXm 1750 2250
2450 2ooo
2&l 2300 1750 2303 25fm
(cm-7
WC)
22; 22.4 22.4
22,6 22.6 22.35 22.4
21.75 -
21.9 22.05 22.11 22.05 22.2 21.9 21.8 21.65 21.55
2i -
21.55 22.25 22.1 21.8
21.4 21.6
2G 21.5 21.7 21.8 22.0
k..
2400 2600 -
2250 -
2200 26J30 2300
2100 2250 2wl 2400 zoo0 2300 -
2200 -
2500 2750 If!&3
2250 2200
2100 2100 1700 2200 -
km-‘)
ANT-3,SDNPA 9MANT-TCPA -_ y,n..I,,, v",..VW
2350
2450 2250
2300 2400 23tm 2100 2400 2450
(cm-‘)
21.1 -
20.5 20.35 20.4 20.25 -
4
53 21 85 83 75
-2 -I -II
-
20.4 20.5 20.6 -
0 2
20.4 20.65 21.1
20.2 20.35
-
4940 3.547
8 9
21.10 20.35 20.2 20.75 20.6 20.85
-
- I2 - I7 -21 2 -
(a) nChlorobutane Dichloromethane Chloroform Carbon tetrachloride 1.2~Dichloro ethane I,l-Dichloro ethane l.I,2,2-Tetrachlorwthane Pentachlorotthane I ,2,3-Trichloropropane nChlorohexane Chlorocyclohexane Dichloro ethylene trans I,l-Dichloro ethylene Trichloroethylene Tettachlomethylene t-Butylbromide fb) Benzene Toluene m-Xylene p-Xylene Mesitylene Ethylbenzene Chlorobenzene Bromobenzene Benzylchloride o-Dichlorobenzene (c) Benzaldehyde Nitrobenzene Propionaldehyde Butyrafdehyde Crotonaldehyde Valeraldehyde Caproic aldehyde Methylacetate Ethylacetate lsobutylacetate n-Butylacetate
3.114 4447 4.1 I4 2.146 4.584) 3.914
V,., (kK)
Class I 2 3 4 5 6 7
(cm-‘)
v",..Y -1/*
ACE-3. SDNPA
sts,”
Al&"
I. Spectroscopic data of the complexes in different solvents at 20°C
N" Solvent
Tahle
23.8 23.8 23.7 23.8 23.6 23.7 23.75 23.8 23.8
23.1 -
23<5 23.0 23.0
23.25 23.3 23.4 23.5 23.3
23-1 -
23.1 23.4 23.3 23.2 23.35
22.95 23.1
24.7 23.2 23.1 23.3 23.2 23.4
V,.. fkK)
2500 2400 2500 2400 2400
2300 2u)o 2400 2400
2450 23al 23lm 2500
2200 -
2450 2500 2400 2450 -
2500 2400
2450 2500 2400 2450 2400
-
(cm-‘)
2,6DMON-TCPA - it12 Y,"..
2900 2850 2800
21.4 21.65 -
21.3 21.05 20.9 20.3
-
0.845 I.301 2.398
-
2s
21.0
64 56 66 107 II3 I41 38
4.176 4.623 4.204 5.778 4.6d 5.204 4.591 2350 1800 zoo0 2050
3&o 2650 -
6
71 90 22.6 21.45 -
2300 2250 -
93 77 49
21.3 21.2 -
78
-
-
-
-
-
3.568 4.670 4.079 5.040 -
2.079 16&l 3.435 3.462 -
19.1 18.8 18.4
2150 1850 2150
-
2500
19.1 -
2600
19.65 -
2450 3200 -
2550 2350 3Mm 2750 2550 2600
2800 -
20.35 19.65 19.2 20.2 19.95 19.4 19.7 20.35 19.2 20.15 -
2250 -
-
19.25 -
-
2550 2600 2350
22?5 22.9 2&) 1600
21.2 20.8
-
2soo -
2150 -
2300 2400
2250 -
2G -
22.3 -
22< 22.6 -
22.55 -
“Refs 14, 25 and 26. bRef 30. ‘G. Berrebi, These de doctorat, Louvain, 1970. ‘Estimated value from the E,/S, correlation (as a proof, see Ref 1, where 41 was estimated as 3.5).
38 Isopropylacetate 39 Diethylether 40 Dibutylether 41 1.2~Dimethoxyethane 42 Tetrahydrofuran 43 2-Methyltetrahydrofuran 44 Tetmhydropyran 45 I/t-Dioxan 46 Acetonitrile 47 Capronitrile 48 Nitromethane 49 2-Nitropropane 50 Acetone 5I Acetophenone 52 Cyclohexanone 53 Dimethylformamide 54 NN’Dimethylacetamide 55 Dimethylsulfoxide 56 Benzonitrile Miscellaneous 57 n-Pentane 58 Cyclohexane 59 Cyclohexene 60 Carbon disulphide 2900 ww)
24; 23.25
23.2 -
-
2700 -
24.0
-
2900 -
2500 -
2600 2600 -
-
23.8 -
23.5 -
23.5 23.7 -
23.7 23.8 23.85 -
23.8 -
S. DUPIREet al.
138
ize the variation of I& of a complex. These parameters include both the polarity and the polarizability factor. SMis defined as log kz, where k2is the second order rate constant of the reaction of tri(n-propyl)amine with methyliodide.‘.” This parameter is used because of the great number of solvents available. Let us note that all these three solvent parameters vary in parallel with each other. Fig 1 shows the variation of &.I with SM for the complex .ACE-3JDNPA in the halogen (chlorine)containing solvents. There is a red shift with increasing polarity. It can be seen that two classes must be distinguished in these solvents: the per- or polyhalogenated solvents and those which are monohalogenated solvents or with few halogen substituents. In the former case, cm. varies less with SM,there is a sort of “saturation effect”, while in the latter case, the variation of i;, is more pronounced.* Similar variations can be found for the three other complexes. Thus the phenomenon must be rather general. Voigt’s results of the complex pdimethoxybenzeneTCNE show similar splitting into two classes of the halogencontaining solvents (only Iln values are reported, because of the very large absorption maxima; a nondependence of V,., - filn with solvent is assumed).16 We have reported in Fig 2 the variation of & vs SMin the aromatic solvents for the same complex. We can see the bathochromic shift with increasing polarity, as may be *In the previouslyreported results’ carbon tetrachloride and chloroform were exceptions in the relation LX vs S, for the complex acenaphthene-TCPA. The small number of solvents studied in that case did not permit us to interpret this particular behaviour. tin the plot of the absorption maxima of iodine vs 1” (the donor ionization potential) cyclohexane, n-heptane were also correlated with the aromatic solvents.’ SFor abbreviations see Table 2.
I6
2
expected when the ground state is less polar than the excited state.’ The same relation holds for the other three complexes. Chlorobenzene,s bromobenzen&” benzyl chloride.X and o-dichlorobenzene26 follow the aromatic solvents as well as cyclohexane” and cyc1ohexene.t” The sensitivity of fi,, vs SMis given by the slope of the straight lines (Table 2). They were calculated with 1,, and Su expressed in kcal mol-’ units and finally standardized against the Z values (in kcal mol-‘), using the known slopes of the relationships ET/!&,(1.62) and Z/ET (1*37).‘s We have chosen the Z values as reference, because they represent the “maximum” solvent effect at present. This is a particular case where the ground state is very polar and the excited state is much less polar. The dipole moment of the complex not only diminishes when going from the ground to the excited state, but it also changes direction (approximately 90”).These two effects make the solvent effect very large.” (However the slope of I,, in kcal mol-’ of the charge-transfer maxima of tropylium iodide vs Z is 1.05, which casts doubt on the effect of the change in direction of the dipole in the reference pyridinium ion.“) The slopes &,.JZ show that the solvent effect is the greatest in the halogenated solvents (slope - 0.4), smaller in the aromatic solvents (slope - 0.3) and the smallest in the per- or polyhalogenated solvents (slope -0.2). The variation of I,,,., in the case of complexes where the ground state is less polar than the excited state, is much less than in the opposite case (Z value). While for the iodide-pyridium complex”“’ the polar and polarisability effects overshadow the small specific effects, in our case these latter effects do separate the solvents into different classes. The available literature data confirms this statement. However it must be emphasized that a small number of solvents studied can lead to an erroneous conclusion. The slope L, vs Z is positive for the complex HEB.“=’ which would indicate that the complex is more
4
3
3
5+5, Figure I. Dependency
of k,,,
of the complex ACE-3JDNPA on S, halogenated(---) solvents.
in per- or polyhalogenated
(-)
and
Charge-transfer
complexes
in organic chemistry-XII
139
5+s, Figure 2. Variation
of I,..
of the complex
ACEJJDNPA
polar in the ground state. This is contrary to other aromatic-& complexes studied in detaiLn It is probable that the general polarity-polarizability parameters (Su, Z, Er, . . . .) are not adequately used in this case. If V,,,.,is plotted against A&*, a positive slope (-20) is obtained showing that the n-donor property of the halogenated solvents play the more important role. The slope of I,, in eV vs IO(eV) is -0.2, showing again the specific character of the solvation by these solvents. Let us emphasize, that the solvent effect on the absorption of iodine in the same solvents is very similar, where the n-donor character of these solvents are mainly responsible for the shift.’ While direct relationship I,, vs Z gives a slope near unity for the complexes DMA-TNB (084) and DMA-TCNB (099):’ the slope obtained initially with & (or E+) and recalculated vs Z gives 0.2 and 0.3 respectively. Those very large slopes cannot be explained easily for these complexes where the ground state is less polar. The main reason for this discrepancy is the inclusion of “anomalous” Z values (/) of cyclohexane, heptane, hexane and 2,2,4-trimethylpentane) in the correlation. Without these solvents, the correlation for the halogen containing solvents alone gives slopes equal to -0.3 and -0.2 respectively. Voigt’s results16 present also similar anomalies: while plotted directly vs Z, the slope for the complex oDMOB-TCNE is 0.7, it is only 0.35 with Sh(or Er values. These latter parameters have the advantage to give a greater number of solvents. These remarks show the precaution to take when only one solvent parameter is used with very limited number of solvents. Since specific effects play a great role in the case of complexes where the ground state is less polar than the excited state, well chosen solvents are needed (at least IO of them) in order to establish a correlation. This must be emphasized, because literature data abound in partial *A& = i, - YR where is is the stretching vibration frequency (O-D) of deuterated methanol in solvent S and Vs is the same vibration in the reference solvent benzene. AI, is thus the n-donor parameter.M
with S, in aromatic
(0) and miscellaneous
(0) solvents.
results with partial conclusions which are often erroneous. Let us note finally that the slopes of I,, vs Z for the n-n* and n -n* transitions of a/I conjugated 9decalenone are respectively -0.25 and +O*3L” The solvent effect on simple molecules or on complexes is rather very similar, as it has been noted by Symons and Davis.29 Because the Franck-Condon state is directly reached in the excited state, the magnitude of the solvent effect vs Z does not give the amount of charge-transfer in the excited state as it was proposed for the reaction rates,% where the activation complex is supposed to be in thermal equilibrium with environment. In Fig 3, I,.. is plotted against Aio for the complex 2,6 DMON-TCPA. The correlation is very good and this holds also for the complexes ACE3JDNPA and 9MANT-TCPA. In the case of n-donor solvents, their n-donor ability is the more important factor as it was shown previously: these solvents cause a blue shift proportional to their n-donor ability. A theoretical explanation was already given in Ref 2. For the complex ANT-3, SDNPA, the points are rather scattered in the region above A&=75 cm-‘. Aromatic solvents with n-donor groups (Table I: 27, 28, 51 and 56) do not follow the straight line obtained for n-donor solvents. They seem to follow a similar correlation with A& displaced somewhat to lower wave numbers from the “purely” n-donor solvents, because of the polarisable aromatic ring. Let us emphasize that these same solvents do not follow the I, vs S, relationship. For three complexes the sensitivity of &., to the n-donor parameter is about the same (slope - 5), and for the other two, the sensitivity is about twice as great (slope -IO). The reason for this phenomenon is not clear at present. The slope of aromatic n-donor solvents is similar to that of n-donor solvents, showing that a special property (n-donor ability) is added to the general effect (polarity-polarisability of the aromatic ring). Let us emphasize that literature data also give slopes equal to -5 or -10.
S. DUPLRE et al.
140
Figure 3. Relation between i,..
of the complex 2,6DMON-TCPA solvents.
and the n-donor solvent parameter AV, in n-donor
Tahle 2. The values of the slopes 5,. (kcal mall’) vs Z &al mol-‘) in the case of per- or polyhalogenated (class al), halogenated (class aI1) and aromatic (class b) solvents. The values of the slopes G,. (cm-‘) vs Al, (cm-‘) for the n-donor solvents (class c) Slope Slope &Jz
~,../A~,
Ref
Class (a) Complex ACE-3, SDNPA ANTJ, SDNPA 9-MANT-TCPA 2,6DMON-TCPA ACE-TCPA p-DMOB’-TCNE o-DMOBd-TCNE HMB’-TCPA HEB’-I, DMA”-TNB’ DMA-TCNB’ ACE-CA’ PYR’-CA HMB-TR” ANT-TR PYR-TR ANT-A” LiBr-TCNE HMBCA HMB-NTDAP ANT-NTDA HMB-NTDA PER’-NTDA 3.4BPYR’-NTDA PYR-NNA’ ANT-NNA 3,CBPYR-NNA
I -0.16 -0.16 -0.14 - 0.07 -0.3 - 0.35 + 0.2 -0.2 -0.3 -0.2 -0.2 -
- 0.05 - 0.05 -
II -
Class (b)
0.35 0.45 0.35 0.45 0.45 0.45 -
-0.2 -0.2 -0.25 - 0.25 -0.45 -0.45 -0.25 -0.55’ -
Class (c) - IO - 5(S)” II.5 5.1 4.8 - 20(-0.2) -
-
5 - IO - IO
-
59 -10 -
-0.5 -0.25 -0.25 -0.3 -0.3 -0.25 -0.4
-
0
0
2 16 I6 I8 19.20 21,32 21 21 21 27 27 27 27 28 29 29.31 7 7 7 7 7 7 7
Note: the slopes calculated from literature values are very approximate because of the paucity of data available. “This work; ‘the value in parenthesis is for aromatic n-donor solvents; cp dimethoxybenzene; “odimethoxybenzene; hexamethylbenzene; ‘in both emission and absorption; ‘hexaethylbenzene; ‘Ndimethylaniline; ‘sym trinitrobenzene; ‘I, 2, 4, 5tetracyanobenzene; *chloranil; ‘pyrene; mtropylium tetrafluoroborate; “N-methylacridinium perchlorate; “slope obtained with V, (eV) vs IO, the ionization energies of the solvents(‘); ’ I, 4, 5,8naphthalene tettacarboxylic dianhydride; ‘perylene; ‘3,4benzopyrene; ‘3-nitrol,8-naphthoic anhydride. l
Charge-transfercomplexesin organic chemistry-Xl1 We can conclude from the variation of p, in different solvents, that the following classification is suggested: class (a) halogenated solvents, group I: per- or polyhalogenated solvents; group II. monohalogenated solvents, class (b) aromatic solvents, class (b) n-donor soloents, group I: “pure” n-donor solvents, group II, aromatic n-donor solvents. This classification seems to be rather general.’ We have also examined the variation of the half-band
width (I,, - ij,,J of the CT band with the solvent (p,n is the wavenumber at the half intensity of the band at its bathochromic side). The parameter Shl shows a much better correlation, but differ from complex to complex. The aromatic solvents show an increasing tendency of I,., - p1,2with Ski in the case of the complexes ACE3JDNPA and ANT-3, SDNPA. In the case of ACE-TCPA an opposite behaviour was observed previously.’ The complex 2,6DMONTCPA does not vary with Sk(, while the complex 9MANT-TCPA shows a complete dispersion of points. The halogen-containing solvents do not give any correlation with SH for 9MANT-TCPA and ANT3,5DNPA, there is no variation in the case of 2,6DMONTCPA, while the half-band width increases with Ski for ACE3,SDNPA. No correlation seems to exist between r&- i& and Ai% of n-donor solvents except for the complex 2,6-DMON-TCPA (positive slope). For these solvents the molar volume turns out to be a much better parameter in all cases (negative slope.). It seems that the smaller the solvent molecules the better they solvate the complex.’ This causes a loosening of the complex, hence the greater value of the half-band width. At present, the behaviour of a&,,- pin as a function of various parameters cannot be satisfactorily explained.’ EXPERlMENTAL Materials. Tetrachlorophthalic anhydride (Fluka) was purified by several sublimations, mp 255°C. 3,Sdinitrophthalic anhydride was obtained from the corresponding di-acid by dehydration with acetic anhydride (mp 163.3”c). It is-handled under nitrogen. The d&acid (mp 226°C) was synthesized from the 3,Sdinitrotoluic acid by oxidation with nitric acid (d = 1.15) in sealed tubes at 143°C. Acenaphthene(Fluka) was recrystalliz.ed,five times from aqueous ethanol. rnD95-%“C. Anthracene(U.C.B.) was molten with solid KOH and’ fractionally crystallized from petroleum ether, mp 216°C. 9Methylanthracene was recrystallized thrice from methanol, mp 80-8l”C. 26Dimethoxynaphthalene (Aldrich) was washed with aqueous sodium hydroxide and recrystallized from aqueous ethanol (mp l52-153°C). The solvents were purified by normal methods and were dried on molecular sieves. Measurements. The spectrophotometric measurements were carried out with a Unicam SP 800 Spectrophotometer at 2O*O.l”C. The wave number scale was standardized against a holmium filter.
141
BIBLIOGRAPHY ‘Part X. J. B. Nagy, 0. B. Nagy and A. Bruylants, Bull. Sot. Chim. Beig. 83, 163 (1974) ‘Part VI. 0. B. Nagy, J. B. Nagy and A. Bruylants, I.C.S. Perkin II 968 (1972) ‘See R. S. Mulliken and W. B. Person, Molecular Cumplexes. A Lecture and Reprint Volume, Wiley, New York. I%9 ‘J. Walkey, D. N. Glew and J. H. Hildebrand, J. Chenr. Phys. 33. 621 (1960) ‘E. hf. Voigt, 1. Phys. Chem. 70, 598 (1966) ‘P. H. Emslie and R. Foster, Rec. Trau. Chim. 84. 255 (1965) ‘I. Ilmet and S. A. Berger, 1. Phys. Chem. 71, 1534 (l%7) ‘Part XIII. S. Dupire, J: B. Nagy. 0. B. Nagy and A. Bruylants. J.C.S. Perkin II 478 (1974); S. Dupire. These de Licence. Louvain. 1971 ‘Part VIII. J. B. Nagy, 0. B. Nagy and A. Bruylants, Bull. Sot. Chim. Be/n. 82. 337 (1973): J. B. Nagy, Ph.D. Thesis. U.C.L., Louvain, i970 -_ “‘See for example W. Liptay in Oprische Anregung orgonischer Systeme, Verlag Chemie, Weinheim/Bergstr., 1966 “E. M. Kosower, 1. Am. Chem. Sot. 80, 3253, 3260 (1958) “K. Dimroth, Chr. Reichardt, Th. Siepman and F. Bohlman, Ann. 661, I (1963) “K. Dimroth, Chr. Reichardt and A. Schweig, Ann. 669, I (l%3) “J. C. Jungers, L. Sajus, I. de Aguirre and D. Decroocq, L’Anulyse cinbque de la transformation chimique, Technip, Paris, 1968, Vol. II “E. M. Kosower, An Introduction to Physical Organic Chemistry, Wiley, New York, 1968 “E. M. Voigt, J. Am. Chem. Sot. 86, 3611, 3930 (1964) “E. M. Kosower, J. Org. Chem. 29,956 (1964) ‘J. Czekalla, K. 0. Meyer, 2. physik. Chem. 27, 185 (l%l) “H. L. Browning, Dissertation, Louisiana State University, 1960 =G. Briegleb in Oprische Anregung organischer Sysfeme. Verlag Chemie, Weim/Bergstr., 1966 “P. H. Emslie and R. Foster, Rec. Trau. Chim. 84, 255 (1%5) “G. Briegleb, Elekfronen-Donator-Acceptor-Komplexe, Springer Verlag, Berlin, I%1 “E. M. Kosower, J. Chim. Phys. 230 (1964) UM. Mohammad and E. M. Kosower, 1. Am. Chem. Sot. 93.2709, 2713 (1971) ?. Lassau and J. C. Jungers, Bull. Sot. Chim. 2678 (1%8) ‘“Y. Drougard and D. Decroocq, Bull. Sot. Chim. 2972 (1969) nT. G. Beaumont and K. M. C. Davis, J. Chem. Sot. B 1010(1968) ?I. Briegleb, W. Liptay and R. Fick, 2. Physik. Chem. 33, I81 (1%2); Z Elektrochem. Ber. Bunsenges. Physik. Chem. 66.85 I. 859 (1962) “K. M. C. Davis and M. C. R. Symons, 1. Chem. Sot. 2079(1965) ?. Kaaiva. Y. Sumida and T. Inoue. Bull. Chem. Sot. Japan 41, 767 (1968) “R. Foster, 1. Chem. Sot. 1075 (1960) “R. Foster and D. L. Hammick, J. Chem. Sot. 2685 (1954) “E. M. Voigt, I. Chem. Phys. 49, 3300 (1968) ‘J. H. Hildebrand, J. M. Prausnitz and R. L. Scott, Rqulur and related solutions, The solubility of Gases, Liquids, and Solids. Van Nostrand, New York, 1970