- Email: [email protected]
Substituted methyl groups
Tcrrokdron.
Vol 27. pp.945
10 951
Per:moa
Press 1971.
Pnnlcd
SUBSTITUTED
In Gru1
METHYL
SUBSTITUENT W. A.
Bntam
GROUPS
EFFECTS’
SHEPPARD
Contribution No. 1666 from the Central Research Department, Experimental Station, E. 1. du Pont de Ncmours and Company, Wilmington, Delaware 19898 (Received in USA 29 September
1970; Received in the UK/or
publication 6 October 1970)
Abatract-The substitucnt parameters for the series of mono-, di- and trisubstituted Me groups,CHJ, CHX,. and CX, where X is F, Cl, Br. OMc, SCF,. and CN. have been determined from F” NMR measurements on the corresponding meto- and para-tluorotoluenes. An additive linear effect of substitution is found only for the cyano substituent. A saturation effect noted for the groups where X is F. Cl, Br and SCF, (particularly chloro and bromo) after disubstitution is attributed to through-space interactions of unshared p-electrons of these substituents with the x system of the ring
THE electronic effects of halogenated substituents are of considerable interest, but the origin and mechanism of transmission of the observed effects continue to be controversial.’ In particular, the trifluoromethyl group has been found to have a resonance withdrawal effect3** which is not explained by conventional resonance pictures as normally used for nitro or cyano groups. This +R effect was initially explained by a fluoride ion hyperconjugation or a x-inductive effect. As a result of new data, an alternative explanation of p-x interaction was proposed.4 Like the accepted picture for resonance interaction of fluorine with an aromatic ring, a through-space return of electron density from the fluorine of the trifluoromethyl group was suggested to in part cancel the inductive electron withdrawal from the meta position of the ring. Thus the pura position appeared to have an enhanced deactivation. Through-space interaction involving fluorine has also been concluded from ESR studies.’ This p-n interaction mechanism has been criticized6 and was not apparent from analysis of pKa data on bicyclooctane acids.’ This paper reports and summarizes results on a series of substituted Me’s that are pertinent to the question of interaction between the unshared p-electrons of substituents on a Me group and the n system of an aromatic nucleus to which the Me group is bonded.’ DISCUSSION The trisubstituted Me groups CX3, trifluoromethyl, trichloromethyl, and tribromomethyl. are all strongly electron-withdrawing inductively since the chemical shift for the meta fluorine is downfield from that of fluorobenzene. The inductive effect of the CX3 group decreases from fluorine to bromine in relative proportion to the electronegativity of the halogen. The larger downfield shift for the para fluorine (positive uR) would seem to indicate a resonance-enhanced electron widthdrawal for all trihalol
A - R effect is reported for CF,(u,
053. (rD048) in reaction of the benxoic acids with diazomethanc. 945
para
meI0 pore mefa pm0 wsela pm0 meta pma meta PIItd meta pma Ma pm0 meta
CHClx
CHBr,
CBr,
1.6196 1.6195 14447 14450 1.4471 14468 1,4597 14601 1.4761 1.4750 1.4718 1.4709 1.4706 I 4693 79-80 (hexane) 75-75.5 (hexaaej 1.4990 1.4905
85 (07) 82 (07) 67.5 (76) 61(52) 71 (12) 74(12) 62 (D5) 59 (05) 65 (16) 60 (2-Q 84(16) 84 (16) 64(1.2) 64 (08)
65 (035) 70 (050)
-
I.5849
1.5255
-
80 (2.4)
_ 61” (60 mm) 61”(58 mm) 83 (6)
(m)
ni’, or mp. “C and solvent recrystallization
(146.1) C,H,CI,F (1791)) C,H,Br,F (2680) C,H.Br,F (346-9) C,H,aF, (1806) C&F,% (310.3) CWHIFWS, (4103) CsH,FO (140-2) C&LFOz (172.2) G,HI,FOS (2002) GHsFN, (1602) CIOH,NIF (174.2)
CY57) cz(68) BW) BW W4) W91) W9Y DW) 437) E(59) F(M) F(47) G(76) G(47) H(w) H(43) U78) U87)
Cl(l7)
C&F,
A(70) e(U)
Formula mol wt
AW
(% yield)
Preparation’
Of
Method
67.5 _-
-
29.3 635 -
24.2 466 34.8
31.4
57.5 _. -
Calc
63.1 63.1 6Q3 604 67.7 674
24.9 245 46.4 46.3 3440 34.8 29.3 29.5
31.5
57.5 581 -
Found
Carbon .
3.15 -
098 6.52 6.54
2.24 I .62
1.16
1.88
345
CA
_. 5.97 6.34 6.71 6.57 2.84 2.84 -
I .39 I44 2.39 2.55 1.70 I.74 1.30 I.17
211
3.73 4Q7 _-
Found
__ Hydrogen
-
-
46.3 __ 13.6 Il.2 9.49 Il.9 109
46+I 46G 13.8 13.6 Il.9 110 9.49 944 11.9 12.1 IQ9 109
546 548 31.9
I@4
_.
548 316
1@6
-
Fluorine _. .___ Calc Found
Analysis
N
N
-
S _.
Cl S
Br
Br
16.1
17.5
-
-
23.4 -
19.6 20.7
69.1
59,7
396
-
Calc
cl
Other
_-
3 4
B_ q
206 23.4 23.5 17.8 17-8 16.3 160
f
2 g 2. 19.8 198
69.1 69.2
596
394
Found
’ Letter (and numerical) designation refers to procedure listed in Experimental -ion. b A purple impurity, apparent also from F I9 NMR spectra. was nmoved by treatment with concentrated sulfuric acid. ’ Reported used for solvolysis studies by A. M. Avedikian, J. Chaput, S. C. N’Ketsia, J. Dausquc, A. Kcrgomard. J.-M. Rondier and H. Tatou. Bull. Sot. Chim. Fr., 95 (1966) but no characterization given.
C&H,) (CN), 2 paa
CH(CN)x
WCH,),
CH(OCH,)x
CH,OCHs
C(SCF,),
CH(SCF,),
CCIFl
met4 pma pora
CHFx
X of FC6H.X
bp. “C
W. A.
948
SHEPPARD
TABLE 2. KNOWN ~-SUBSTITUTEDFLUOROTOLUEMS: PHYSICAL PROPWW
x of
bp, “C
FC,H,X
(mm)
CHs CHsF
CFs CH,Cl CHCl, CCI, CHsBr CHBr, CH,SCF, CH,CN C(CN)s
ANTI SOURCE
Source or Method of Preparation and Reference
n’,’
mera para
115 116-117
14671
a
1.4663
a
meca
35 (20)
14633
b
para meta para meta pora mefa
5ft(lO) 100 103
1.4673
b, c AgF + FC,H,CH,Br
1.3980 1.3984
a a
64 (6) 67 (6)
15109 15109
d
F&H.CH,
+ Cl2 -&-
77 (6)
1.5238
e,/
FCsH,CH,
+ 2C1, $&-
88 (7.3) 76 (4.3)
1.5325 1.5321
FC6H,CH,
+ excess Cl, ,--&
81 (7.5) 64.5 (6)
15443 15450
g B h
FC6H,CH,
+ Br, --$-
77 (2.4)
1.5847
e
FC,H,CHs FC,H,CH,Br
meta wt met0 wra meta meta pora mera prrra mera wra
d
HgF, + FC,H,CH,Br
1.4582 1.4573
i i
1.4988 1.4980
a a
92 (3.5) 91-92 (3.3)
1.4954 1.4958
i
tctr~~~l;d~~‘c”c
.
.
.
h
64.5 (12) 64(12) 124-126(10) 117-118(18)
J
uy~~~c;
+ 2Br, &
FC6H,CHsCN
4
CCI,F
+ Hg(SCF,), -
s
i
’ Pierce Chemical Company.
b J. Bernstein. S. J. Roth and W. T. Miller, J. Amer. C/tern. Sot.. 70, 2310 (1948); see also C. B&+uo. Bull. Sot. Chim. Fr.. 4214 (1967) and rcf in footnote d. ’ The literature procedure using mercuric difluoride gave only a very low yield of impure pfluorobenzyl fluoride. The silver fluoride method gave pure product in respectable yield. ’ T. Yokoyama, G. R. Wiley and S. I. Miller, J. Org. Cheat. 34, 1859 (1969) ’ P. S. Varma. K. S. Venkat Raman and P. hi. Nilkantiah. J. Indian Chem. Sot. 21. 112 (1944) ; Chem. Abstr., 39. 1395 (1945) / G. Scbiemaott, 2. Phys. Chem. (Leipzig), 156.397 (1931). s H. Frciscr. M. E. Hobbs and P. M. Gross, J. Amer. Chem. Sot.. 71.111 (1949). * G. A. Olah, A. E. Pavloath, J. A. Olah and F. Herr, J. Org. Chem., 22,879 (1957). ‘ V. V. Grda, L. M. Yagupol’skii, V. F. Bystrov and A. U. Stepaoyaots, 1. Gen. Cheat. U.S.S.R. (English Traosl.), 3, 1631 (1965). ’ J. K. Williams, E. L. Martin and W. A. Sheppard, J. Org. Chem., 31,919 (1966).
normal resonance donation of the Me group to zero. However, the halogen substituents quickly cause an apparent positive resonance withdrawal. A possible explanation for this effect is through-space interaction between the unshared p electrons of the halogens and the n system of the ring. For the larger halogens, chlorine and bromine, the trisubstituted groups must have considerable steric interaction with possible restricted rotation. With disubstitutioo, the hydrogen can be located in a position such that interaction with the orrho hydrogen of the rings could restrict rotation of the group so that the halogens above and below
Substituted TABLE 3. NMR
F”
CHEMICAL
SHIF&
PARAMETERS
Substitucnt
X
ff
OF z-SuBFllNTED FOR
949
methyl groups l’LUOROTOLUEMES.
SUEsTrrUnD
METHYL
K”
Kx
FC6H.X
AND
SUBSTINENT
ououP3
01
61’
+ 4.34 -I-4.25
-007 -002
-015
PX *X
+ I~lO(l~18)
+ 5.44 (540)
CH,CH,
( + 0.75)*
CH(CH,), C(CH,),
( + Q45Y
(+5a)b (4.75)’ (+s.ss)b
+ 5.10
+M2
+Ql7
CF, CCIF,
-028 - 1.48 - 2.20 -209
+Q25 -3.14 -5Q9 - 439
+Q53 -166 - 289 - 2.30
+Q12 +Q29 +Q39 + Q38
-002 +OQ6 +QlO +0-08
CH,CI CHCI, CCI,
-Q53(-045) - I.56 ( - 1.53) - 1.51
+ Q50 ( + 045) -2.16(-2.25) -2.31
+1Q3 -Q72 - Q80
+Q16 +Q30 +Q30
-004 +OQ2 +QO3
CH,Br CHBr, CBr,
- 0.50 - I.51 - I.26
+ 0.22 -2.21 -2.21
+Q72 -Q70 - 095
+Q16 +Q30 + 026
-002 +OQ2 +OQ3
CH,CN
-1.28(-1.1) - 3.2Y - 3.88 - 646
+lQ7(+1*20) - 2.3Q -2.16 -6.71
+ 2.35 +Q99 + 1.72 -025
+Q26 +Q55 +Q63 + Q98
-008 -03 -0% +OQl
+053 + Q50 +Q4l
+2.41 + 1.38 + 062
+ 1.88 + 0.88 +Q21
-0Ql -002 -003
-0% -003 -Ml
-Q92 - 2.53 - 2.90
+ 073 - 2.43 - 3.99
+ I.65 +Q10 -IQ9
+Q2l +Q44 + Q49
-006 0 +OQ4
CH,
CH,F CHF,
CH(CN), C(CH,) (CN), C(CN), CH,OCH, CH(OCH,), C@CH,), CH,SCF, CH(gCP,), CGCF,),
-Q14
’ In parts per million. Probable error is 004 ppm. Unless indicated otherwise all data are in CFCI, solvent at infmitc dilution. Values in parenthesis are literature values in Ccl, or a hydrocarbon solvent from ref. 15. unless indicated otherwise. b Private communication from Professor R. W. Taft. c Benzene solvent (insoluble in CCI,F).
can interact to the fullest extent with the n system of the ring. These steric differences are clearly seen on space-filling molecular models. Because cyano is a rod-like substituent, the cyan0 group has only a small steric effect and does not appear to interact directly with the II system. However, the idea that the saturation effect results from steric interactions can also be argued for the hyperconjugation mechanism. The chlorodifluoromethyl group is intermediate between CF3 and Ccl, as expected. The methyldicyanomethyl group is very close to the dicyanomethyl, and the small differences could be due to the different solvent required because of solubility problems. The slightly smaller inductive withdrawal for the dicyanomethyl could result from hyperconjugative effects derived from the strongly acidic benzylic hydrogen. Solvent effect studies on the cyano series only confirm previous observations that field effects are extremely important in transmission of electronic effects of cyanocarbon groups.’ 4
950
W. A. SHEPPARD
‘The effect of SCF3 group substitution on the electronic properties of the methyl group is very similar to those observed for halogens including the saturation effect from two to three substituents. Again, p-lr interaction from the p-electrons of both sulfur and fluorine could be significant, although the consequences of d-orbital participation, which is enhanced by the strong electron-withdrawing effect of the trifluoromethyl groups,‘* is not readily assessed. The substitution in the Me group by OMe surprisingly results in very little apparent electronic change, almost the same as for substitution by Me. Possibly the Me groups are primarily oriented to interact with the n system, so that the unshared p-electrons of the oxygen do not. A similar result is the lack of any apparent interaction or backand suggests that bonding effects” in the system (CH,)2N(CH,),C,H,C02CH, p-lr interaction may be repulsive or antibonding in nature and occurs only when forced by the configuration of the molecule. In conclusion, this comparison of substituent effects strongly suggests that unshared p-electrons on substituents X of CX, group interact with the n system of the adjacent aromatic ring and may be of some importance in transmission of electronic effects. Although a through space p-1~ donation4 is an attractive way to interpret this interaction, a a repulsion2’ or some other distortion of the 1c system is also very reasonable. However the data can also be explained by hydridization changes in bonding induced by concentration of highly electronegative halogens and can be represented as hyperconjugation. The saturation effects noted for the chloro- and bromo-substituted methyl groups should be compared to NMR chemical shift2’ and coupling constant22 data where halogen substitution produces anomalous effects that are not readily explained theoretically. EXPERIMENTAL All new compounds prepared for this study are reported in Table 1 with physical properties and analytical data. The IR, UV and proton (and F”) NMR spectra were obtained on all new compounds and used to verify structures. Typical experimental procedures are given below and used for reference in Table 1. All known compounds are summarized in Table 2 with literature reference and method of preparation. 1. Synthesis A. Fluorobenzul difluoride. Following the literature procedure, ’ 124 g (@lo mole) m-fluorobenzaldehydc and 20 g SF, were heated at lso” for 6 hr in a Hastcloy autoclave. The product was distilled at r&cod pressure. B. a-Chlorotoluenes. A stoichiomctric amount of Cl, (condensed in a calibrated trap) was passed into a soln of lluorotoluene in Ccl. irradiated by a low-pressure mercury lamp. C. a-Bromololuenes. A stoicbiometric amouot of Br, was added slowly to a solo of fluorotoluenc in Ccl, irradiated with a sunlamp. D. a~Trijluoromethylthio)roluenes. FIuorobenzotribromide 8.7 g (0025 mole) was mixed with 15.2 g and heated with stirring to loo” for almost 30mio. The (@I5 mole) bis(trifluoromethylthio)mercury’” mixture was cooled, diluted with trichlorofluoromethane, and saturated with dry HCI gas to destroy any unreacted (trifluoromcthylthio)ercury. The trichlorofluoromcthanc soln was filtered and distilled. For mono and di-a-bromotoluenes, molar quantities of Hg salt were used. E. Flwrobmzyl methyl ether. To a soln of 9.25 g lluorobenxyl bromide in 50 ml ether was added 271 g NaOMe under dry N2. The solo was rcfluxed for 1 hr. filtered and distilled. F. Fluorobenzclldehydedfmethylocetal. Following the literature procedure” a soln of 1g6 g lluorobenmldehyde in 200 ml McOH containing 0.3 ml cone HCl waa stirred 40 hr under N1 and then made basic with sat NaOMe in MeOH, filtered, and distilled.
Substituted methyl groups
951
G. Methyl orthofluorobenzwte. According to a literature procedure,” 1@6g tluorobenzotrichloridc was added to a cold soln of 8.42 g NaOMc in 50 ml MeOH. The soln was stirred at room temp overnight, then rctluxed for 4 hr, cooled and filtered to remove NaCl. The product was distilled. H. Fluorophenybnalononitrife. Adopting a literature procedure,” a soln of I.7 g fluorobenzotricyanidc in 30 ml acetonitrilc was cooled to -40” and 065 g dry KCN added. After being stirred at -30 to -40 for 3 hr. the soln was filtered into 500 ml ether. A gummy ppt was added to water at 0’ and acidified with 10% H,SO,. The fluorophenylmalononitrile was recrystallized from hcxane. I. (Fluorophenyl)methylmolononitrile. The potassium fluorophenylmalononitrile was prepared as above from fluorobenzotricyanide and KCN in acetonitrile. Mel, 3 g, was added and the soln warmed to room temp and stirred for 3 hr. The soln was filtered and the product distilled. 2. NMR calibrations. FL’ NMR chemical shifts were measured at 25” with a Varian A56-60 spectrometer using techniques previously described. Trichlorotluoromethane was used as solvent and internal calibrant ; measurements were made at 20,10,5’~ concentration and the chemical shift was obtained by extrapolation to infinite dilution. Where the compound was not su8iciently soluble in trichlorolluoromethane, the solvent was employed with 5% 1,1,2,2-tetrachloro-3.3,4,4-tetrafluorocyclobutane as internal calibrant as described previously.‘* The NMR chemical shift data are given in Table 3. 3. Substituent prometers. The inductive and resonance substituent parameters, o, and ai, were calculated for the series of substituent Me groups by the procedure described by Taft et al.” as used in earlier studies of fluorinated substituents4.‘6 and are given in Table III. Acknowledgements-We wish to acknowledge with thanks the special assistance received from Mr. Louis Walther for NMR calibrations and Mr. Ed Wonchoba on synthesis. REFERENCES
1 This paper was presented in part at the 5th International Symposium of Fluorine Chemistry, Moscow, USSR, July. 1969 r a For a review and tabulation of fluorinated substituent effects, see W. A. Sheppard and C. M. Sharts, Orgunic Fluorine Chemistry. pp 33-40 and 348355. W. A. Benjamin, New York, N.Y. (1969); b D. HoIt& A Critical Evaluation of the Concept oj Fluorine Hyperconjugation in frog. Phys. Org. Chem. in press 3 J. D. Roberts, R. L. Webb and E. A. McElhill, J. Am. Chem. Sot. 71408 (1950). l W. A. Sheppard, Ibid. 87.2140 (1%5) ’ For a compilation of these references to ESR studies, set reference 20. page 38. See also K. Kosman and L. M. Stock, J. Am. Gem. Sot. 9z 409 (1970) 6 M. J. S. Dewar and A. P. Marchand, Ibid. 88,354 (1966) ’ F. W. Baker, R. C. Parish and L. M. Stock, Ibid. 89,5677 (1967) e Similar studies have been done by L. M. Yagupolskii. University of Kiev, USSR, reported at the 5th Intemurionul Symposium ojFluorine Chemiary, Moscow. July (1969); and by E. T. McBee, I. Serfaty and T. Hodgins, Purdue University. We thank Dr. Hodgins for a preprint of manuscript prepared for publication. 9 W. R. Hasek, Org. Syn. 41, 104 (1961) ” E. H. Man, D. D. Coffman and E. L. Muetterties, J. Am. Chem. Sot. al,3575 (1959) ” R. N. Icke, C. E. Redemann 9. 9. Wisegarver and G. A. Alles. Org. Syn. Coil. Vol. 3.644 (1955) ” S. M. McElvain and J. T. Venerable, J. Am. C/tern. Sot. 71 1661 (1950) ” 0. W. Webster, W. Mabler and R. E. Benson, Ibid. 84.3678 (1962) ” W. A. Sheppard and R. M. Henderson, Ibid. 89.4446 (1967) ” R. W. Taft, E. Price, J. R. Fox_ 1. C. Lewis, K. K. Andersen, and G. T. Davis, Ibid. 85.709.3l46 (1963) I6 F. S. Fawcett and W. A. Sheppard, Ibid. 87.4341 (1965) ” ’ M. J. S. Dewar, Hyperconjugation. p 159. Ronald Press, New York, N.Y. (1962); b 0. Exner, Tetrahedron Letters 815 (1963) ‘a W. A. Sheppard, J. Am. Chem. Sot. 85, 1314 (1963) I9 J. H. Smith and F. M. Menger, J. Org. Chem. 34. 77 (1969) r” J. N. Murrell, The Theory of ElectronicSpectra ojOrganic Molecules. Wiley, New York, N.Y. (1963) r’ Y. Nomura and Y. Takuchi. Tetrahedron Letters 639 (1969) ” ’ N. Muller and D. T. Carr. J. Phys. Chem. 67. 112(1963); b W. M. Litchman and D. M. Grant, J. Am. Chem. Sot. 90, 1400 (1968)
Vol 27. pp.945
10 951
Per:moa
Press 1971.
Pnnlcd
SUBSTITUTED
In Gru1
METHYL
SUBSTITUENT W. A.
Bntam
GROUPS
EFFECTS’
SHEPPARD
Contribution No. 1666 from the Central Research Department, Experimental Station, E. 1. du Pont de Ncmours and Company, Wilmington, Delaware 19898 (Received in USA 29 September
1970; Received in the UK/or
publication 6 October 1970)
Abatract-The substitucnt parameters for the series of mono-, di- and trisubstituted Me groups,CHJ, CHX,. and CX, where X is F, Cl, Br. OMc, SCF,. and CN. have been determined from F” NMR measurements on the corresponding meto- and para-tluorotoluenes. An additive linear effect of substitution is found only for the cyano substituent. A saturation effect noted for the groups where X is F. Cl, Br and SCF, (particularly chloro and bromo) after disubstitution is attributed to through-space interactions of unshared p-electrons of these substituents with the x system of the ring
THE electronic effects of halogenated substituents are of considerable interest, but the origin and mechanism of transmission of the observed effects continue to be controversial.’ In particular, the trifluoromethyl group has been found to have a resonance withdrawal effect3** which is not explained by conventional resonance pictures as normally used for nitro or cyano groups. This +R effect was initially explained by a fluoride ion hyperconjugation or a x-inductive effect. As a result of new data, an alternative explanation of p-x interaction was proposed.4 Like the accepted picture for resonance interaction of fluorine with an aromatic ring, a through-space return of electron density from the fluorine of the trifluoromethyl group was suggested to in part cancel the inductive electron withdrawal from the meta position of the ring. Thus the pura position appeared to have an enhanced deactivation. Through-space interaction involving fluorine has also been concluded from ESR studies.’ This p-n interaction mechanism has been criticized6 and was not apparent from analysis of pKa data on bicyclooctane acids.’ This paper reports and summarizes results on a series of substituted Me’s that are pertinent to the question of interaction between the unshared p-electrons of substituents on a Me group and the n system of an aromatic nucleus to which the Me group is bonded.’ DISCUSSION The trisubstituted Me groups CX3, trifluoromethyl, trichloromethyl, and tribromomethyl. are all strongly electron-withdrawing inductively since the chemical shift for the meta fluorine is downfield from that of fluorobenzene. The inductive effect of the CX3 group decreases from fluorine to bromine in relative proportion to the electronegativity of the halogen. The larger downfield shift for the para fluorine (positive uR) would seem to indicate a resonance-enhanced electron widthdrawal for all trihalol
A - R effect is reported for CF,(u,
053. (rD048) in reaction of the benxoic acids with diazomethanc. 945
para
meI0 pore mefa pm0 wsela pm0 meta pma meta PIItd meta pma Ma pm0 meta
CHClx
CHBr,
CBr,
1.6196 1.6195 14447 14450 1.4471 14468 1,4597 14601 1.4761 1.4750 1.4718 1.4709 1.4706 I 4693 79-80 (hexane) 75-75.5 (hexaaej 1.4990 1.4905
85 (07) 82 (07) 67.5 (76) 61(52) 71 (12) 74(12) 62 (D5) 59 (05) 65 (16) 60 (2-Q 84(16) 84 (16) 64(1.2) 64 (08)
65 (035) 70 (050)
-
I.5849
1.5255
-
80 (2.4)
_ 61” (60 mm) 61”(58 mm) 83 (6)
(m)
ni’, or mp. “C and solvent recrystallization
(146.1) C,H,CI,F (1791)) C,H,Br,F (2680) C,H.Br,F (346-9) C,H,aF, (1806) C&F,% (310.3) CWHIFWS, (4103) CsH,FO (140-2) C&LFOz (172.2) G,HI,FOS (2002) GHsFN, (1602) CIOH,NIF (174.2)
CY57) cz(68) BW) BW W4) W91) W9Y DW) 437) E(59) F(M) F(47) G(76) G(47) H(w) H(43) U78) U87)
Cl(l7)
C&F,
A(70) e(U)
Formula mol wt
AW
(% yield)
Preparation’
Of
Method
67.5 _-
-
29.3 635 -
24.2 466 34.8
31.4
57.5 _. -
Calc
63.1 63.1 6Q3 604 67.7 674
24.9 245 46.4 46.3 3440 34.8 29.3 29.5
31.5
57.5 581 -
Found
Carbon .
3.15 -
098 6.52 6.54
2.24 I .62
1.16
1.88
345
CA
_. 5.97 6.34 6.71 6.57 2.84 2.84 -
I .39 I44 2.39 2.55 1.70 I.74 1.30 I.17
211
3.73 4Q7 _-
Found
__ Hydrogen
-
-
46.3 __ 13.6 Il.2 9.49 Il.9 109
46+I 46G 13.8 13.6 Il.9 110 9.49 944 11.9 12.1 IQ9 109
546 548 31.9
I@4
_.
548 316
1@6
-
Fluorine _. .___ Calc Found
Analysis
N
N
-
S _.
Cl S
Br
Br
16.1
17.5
-
-
23.4 -
19.6 20.7
69.1
59,7
396
-
Calc
cl
Other
_-
3 4
B_ q
206 23.4 23.5 17.8 17-8 16.3 160
f
2 g 2. 19.8 198
69.1 69.2
596
394
Found
’ Letter (and numerical) designation refers to procedure listed in Experimental -ion. b A purple impurity, apparent also from F I9 NMR spectra. was nmoved by treatment with concentrated sulfuric acid. ’ Reported used for solvolysis studies by A. M. Avedikian, J. Chaput, S. C. N’Ketsia, J. Dausquc, A. Kcrgomard. J.-M. Rondier and H. Tatou. Bull. Sot. Chim. Fr., 95 (1966) but no characterization given.
C&H,) (CN), 2 paa
CH(CN)x
WCH,),
CH(OCH,)x
CH,OCHs
C(SCF,),
CH(SCF,),
CCIFl
met4 pma pora
CHFx
X of FC6H.X
bp. “C
W. A.
948
SHEPPARD
TABLE 2. KNOWN ~-SUBSTITUTEDFLUOROTOLUEMS: PHYSICAL PROPWW
x of
bp, “C
FC,H,X
(mm)
CHs CHsF
CFs CH,Cl CHCl, CCI, CHsBr CHBr, CH,SCF, CH,CN C(CN)s
ANTI SOURCE
Source or Method of Preparation and Reference
n’,’
mera para
115 116-117
14671
a
1.4663
a
meca
35 (20)
14633
b
para meta para meta pora mefa
5ft(lO) 100 103
1.4673
b, c AgF + FC,H,CH,Br
1.3980 1.3984
a a
64 (6) 67 (6)
15109 15109
d
F&H.CH,
+ Cl2 -&-
77 (6)
1.5238
e,/
FCsH,CH,
+ 2C1, $&-
88 (7.3) 76 (4.3)
1.5325 1.5321
FC6H,CH,
+ excess Cl, ,--&
81 (7.5) 64.5 (6)
15443 15450
g B h
FC6H,CH,
+ Br, --$-
77 (2.4)
1.5847
e
FC,H,CHs FC,H,CH,Br
meta wt met0 wra meta meta pora mera prrra mera wra
d
HgF, + FC,H,CH,Br
1.4582 1.4573
i i
1.4988 1.4980
a a
92 (3.5) 91-92 (3.3)
1.4954 1.4958
i
tctr~~~l;d~~‘c”c
.
.
.
h
64.5 (12) 64(12) 124-126(10) 117-118(18)
J
uy~~~c;
+ 2Br, &
FC6H,CHsCN
4
CCI,F
+ Hg(SCF,), -
s
i
’ Pierce Chemical Company.
b J. Bernstein. S. J. Roth and W. T. Miller, J. Amer. C/tern. Sot.. 70, 2310 (1948); see also C. B&+uo. Bull. Sot. Chim. Fr.. 4214 (1967) and rcf in footnote d. ’ The literature procedure using mercuric difluoride gave only a very low yield of impure pfluorobenzyl fluoride. The silver fluoride method gave pure product in respectable yield. ’ T. Yokoyama, G. R. Wiley and S. I. Miller, J. Org. Cheat. 34, 1859 (1969) ’ P. S. Varma. K. S. Venkat Raman and P. hi. Nilkantiah. J. Indian Chem. Sot. 21. 112 (1944) ; Chem. Abstr., 39. 1395 (1945) / G. Scbiemaott, 2. Phys. Chem. (Leipzig), 156.397 (1931). s H. Frciscr. M. E. Hobbs and P. M. Gross, J. Amer. Chem. Sot.. 71.111 (1949). * G. A. Olah, A. E. Pavloath, J. A. Olah and F. Herr, J. Org. Chem., 22,879 (1957). ‘ V. V. Grda, L. M. Yagupol’skii, V. F. Bystrov and A. U. Stepaoyaots, 1. Gen. Cheat. U.S.S.R. (English Traosl.), 3, 1631 (1965). ’ J. K. Williams, E. L. Martin and W. A. Sheppard, J. Org. Chem., 31,919 (1966).
normal resonance donation of the Me group to zero. However, the halogen substituents quickly cause an apparent positive resonance withdrawal. A possible explanation for this effect is through-space interaction between the unshared p electrons of the halogens and the n system of the ring. For the larger halogens, chlorine and bromine, the trisubstituted groups must have considerable steric interaction with possible restricted rotation. With disubstitutioo, the hydrogen can be located in a position such that interaction with the orrho hydrogen of the rings could restrict rotation of the group so that the halogens above and below
Substituted TABLE 3. NMR
F”
CHEMICAL
SHIF&
PARAMETERS
Substitucnt
X
ff
OF z-SuBFllNTED FOR
949
methyl groups l’LUOROTOLUEMES.
SUEsTrrUnD
METHYL
K”
Kx
FC6H.X
AND
SUBSTINENT
ououP3
01
61’
+ 4.34 -I-4.25
-007 -002
-015
PX *X
+ I~lO(l~18)
+ 5.44 (540)
CH,CH,
( + 0.75)*
CH(CH,), C(CH,),
( + Q45Y
(+5a)b (4.75)’ (+s.ss)b
+ 5.10
+M2
+Ql7
CF, CCIF,
-028 - 1.48 - 2.20 -209
+Q25 -3.14 -5Q9 - 439
+Q53 -166 - 289 - 2.30
+Q12 +Q29 +Q39 + Q38
-002 +OQ6 +QlO +0-08
CH,CI CHCI, CCI,
-Q53(-045) - I.56 ( - 1.53) - 1.51
+ Q50 ( + 045) -2.16(-2.25) -2.31
+1Q3 -Q72 - Q80
+Q16 +Q30 +Q30
-004 +OQ2 +QO3
CH,Br CHBr, CBr,
- 0.50 - I.51 - I.26
+ 0.22 -2.21 -2.21
+Q72 -Q70 - 095
+Q16 +Q30 + 026
-002 +OQ2 +OQ3
CH,CN
-1.28(-1.1) - 3.2Y - 3.88 - 646
+lQ7(+1*20) - 2.3Q -2.16 -6.71
+ 2.35 +Q99 + 1.72 -025
+Q26 +Q55 +Q63 + Q98
-008 -03 -0% +OQl
+053 + Q50 +Q4l
+2.41 + 1.38 + 062
+ 1.88 + 0.88 +Q21
-0Ql -002 -003
-0% -003 -Ml
-Q92 - 2.53 - 2.90
+ 073 - 2.43 - 3.99
+ I.65 +Q10 -IQ9
+Q2l +Q44 + Q49
-006 0 +OQ4
CH,
CH,F CHF,
CH(CN), C(CH,) (CN), C(CN), CH,OCH, CH(OCH,), C@CH,), CH,SCF, CH(gCP,), CGCF,),
-Q14
’ In parts per million. Probable error is 004 ppm. Unless indicated otherwise all data are in CFCI, solvent at infmitc dilution. Values in parenthesis are literature values in Ccl, or a hydrocarbon solvent from ref. 15. unless indicated otherwise. b Private communication from Professor R. W. Taft. c Benzene solvent (insoluble in CCI,F).
can interact to the fullest extent with the n system of the ring. These steric differences are clearly seen on space-filling molecular models. Because cyano is a rod-like substituent, the cyan0 group has only a small steric effect and does not appear to interact directly with the II system. However, the idea that the saturation effect results from steric interactions can also be argued for the hyperconjugation mechanism. The chlorodifluoromethyl group is intermediate between CF3 and Ccl, as expected. The methyldicyanomethyl group is very close to the dicyanomethyl, and the small differences could be due to the different solvent required because of solubility problems. The slightly smaller inductive withdrawal for the dicyanomethyl could result from hyperconjugative effects derived from the strongly acidic benzylic hydrogen. Solvent effect studies on the cyano series only confirm previous observations that field effects are extremely important in transmission of electronic effects of cyanocarbon groups.’ 4
950
W. A. SHEPPARD
‘The effect of SCF3 group substitution on the electronic properties of the methyl group is very similar to those observed for halogens including the saturation effect from two to three substituents. Again, p-lr interaction from the p-electrons of both sulfur and fluorine could be significant, although the consequences of d-orbital participation, which is enhanced by the strong electron-withdrawing effect of the trifluoromethyl groups,‘* is not readily assessed. The substitution in the Me group by OMe surprisingly results in very little apparent electronic change, almost the same as for substitution by Me. Possibly the Me groups are primarily oriented to interact with the n system, so that the unshared p-electrons of the oxygen do not. A similar result is the lack of any apparent interaction or backand suggests that bonding effects” in the system (CH,)2N(CH,),C,H,C02CH, p-lr interaction may be repulsive or antibonding in nature and occurs only when forced by the configuration of the molecule. In conclusion, this comparison of substituent effects strongly suggests that unshared p-electrons on substituents X of CX, group interact with the n system of the adjacent aromatic ring and may be of some importance in transmission of electronic effects. Although a through space p-1~ donation4 is an attractive way to interpret this interaction, a a repulsion2’ or some other distortion of the 1c system is also very reasonable. However the data can also be explained by hydridization changes in bonding induced by concentration of highly electronegative halogens and can be represented as hyperconjugation. The saturation effects noted for the chloro- and bromo-substituted methyl groups should be compared to NMR chemical shift2’ and coupling constant22 data where halogen substitution produces anomalous effects that are not readily explained theoretically. EXPERIMENTAL All new compounds prepared for this study are reported in Table 1 with physical properties and analytical data. The IR, UV and proton (and F”) NMR spectra were obtained on all new compounds and used to verify structures. Typical experimental procedures are given below and used for reference in Table 1. All known compounds are summarized in Table 2 with literature reference and method of preparation. 1. Synthesis A. Fluorobenzul difluoride. Following the literature procedure, ’ 124 g (@lo mole) m-fluorobenzaldehydc and 20 g SF, were heated at lso” for 6 hr in a Hastcloy autoclave. The product was distilled at r&cod pressure. B. a-Chlorotoluenes. A stoichiomctric amount of Cl, (condensed in a calibrated trap) was passed into a soln of lluorotoluene in Ccl. irradiated by a low-pressure mercury lamp. C. a-Bromololuenes. A stoicbiometric amouot of Br, was added slowly to a solo of fluorotoluenc in Ccl, irradiated with a sunlamp. D. a~Trijluoromethylthio)roluenes. FIuorobenzotribromide 8.7 g (0025 mole) was mixed with 15.2 g and heated with stirring to loo” for almost 30mio. The (@I5 mole) bis(trifluoromethylthio)mercury’” mixture was cooled, diluted with trichlorofluoromethane, and saturated with dry HCI gas to destroy any unreacted (trifluoromcthylthio)ercury. The trichlorofluoromcthanc soln was filtered and distilled. For mono and di-a-bromotoluenes, molar quantities of Hg salt were used. E. Flwrobmzyl methyl ether. To a soln of 9.25 g lluorobenxyl bromide in 50 ml ether was added 271 g NaOMe under dry N2. The solo was rcfluxed for 1 hr. filtered and distilled. F. Fluorobenzclldehydedfmethylocetal. Following the literature procedure” a soln of 1g6 g lluorobenmldehyde in 200 ml McOH containing 0.3 ml cone HCl waa stirred 40 hr under N1 and then made basic with sat NaOMe in MeOH, filtered, and distilled.
Substituted methyl groups
951
G. Methyl orthofluorobenzwte. According to a literature procedure,” 1@6g tluorobenzotrichloridc was added to a cold soln of 8.42 g NaOMc in 50 ml MeOH. The soln was stirred at room temp overnight, then rctluxed for 4 hr, cooled and filtered to remove NaCl. The product was distilled. H. Fluorophenybnalononitrife. Adopting a literature procedure,” a soln of I.7 g fluorobenzotricyanidc in 30 ml acetonitrilc was cooled to -40” and 065 g dry KCN added. After being stirred at -30 to -40 for 3 hr. the soln was filtered into 500 ml ether. A gummy ppt was added to water at 0’ and acidified with 10% H,SO,. The fluorophenylmalononitrile was recrystallized from hcxane. I. (Fluorophenyl)methylmolononitrile. The potassium fluorophenylmalononitrile was prepared as above from fluorobenzotricyanide and KCN in acetonitrile. Mel, 3 g, was added and the soln warmed to room temp and stirred for 3 hr. The soln was filtered and the product distilled. 2. NMR calibrations. FL’ NMR chemical shifts were measured at 25” with a Varian A56-60 spectrometer using techniques previously described. Trichlorotluoromethane was used as solvent and internal calibrant ; measurements were made at 20,10,5’~ concentration and the chemical shift was obtained by extrapolation to infinite dilution. Where the compound was not su8iciently soluble in trichlorolluoromethane, the solvent was employed with 5% 1,1,2,2-tetrachloro-3.3,4,4-tetrafluorocyclobutane as internal calibrant as described previously.‘* The NMR chemical shift data are given in Table 3. 3. Substituent prometers. The inductive and resonance substituent parameters, o, and ai, were calculated for the series of substituent Me groups by the procedure described by Taft et al.” as used in earlier studies of fluorinated substituents4.‘6 and are given in Table III. Acknowledgements-We wish to acknowledge with thanks the special assistance received from Mr. Louis Walther for NMR calibrations and Mr. Ed Wonchoba on synthesis. REFERENCES
1 This paper was presented in part at the 5th International Symposium of Fluorine Chemistry, Moscow, USSR, July. 1969 r a For a review and tabulation of fluorinated substituent effects, see W. A. Sheppard and C. M. Sharts, Orgunic Fluorine Chemistry. pp 33-40 and 348355. W. A. Benjamin, New York, N.Y. (1969); b D. HoIt& A Critical Evaluation of the Concept oj Fluorine Hyperconjugation in frog. Phys. Org. Chem. in press 3 J. D. Roberts, R. L. Webb and E. A. McElhill, J. Am. Chem. Sot. 71408 (1950). l W. A. Sheppard, Ibid. 87.2140 (1%5) ’ For a compilation of these references to ESR studies, set reference 20. page 38. See also K. Kosman and L. M. Stock, J. Am. Gem. Sot. 9z 409 (1970) 6 M. J. S. Dewar and A. P. Marchand, Ibid. 88,354 (1966) ’ F. W. Baker, R. C. Parish and L. M. Stock, Ibid. 89,5677 (1967) e Similar studies have been done by L. M. Yagupolskii. University of Kiev, USSR, reported at the 5th Intemurionul Symposium ojFluorine Chemiary, Moscow. July (1969); and by E. T. McBee, I. Serfaty and T. Hodgins, Purdue University. We thank Dr. Hodgins for a preprint of manuscript prepared for publication. 9 W. R. Hasek, Org. Syn. 41, 104 (1961) ” E. H. Man, D. D. Coffman and E. L. Muetterties, J. Am. Chem. Sot. al,3575 (1959) ” R. N. Icke, C. E. Redemann 9. 9. Wisegarver and G. A. Alles. Org. Syn. Coil. Vol. 3.644 (1955) ” S. M. McElvain and J. T. Venerable, J. Am. C/tern. Sot. 71 1661 (1950) ” 0. W. Webster, W. Mabler and R. E. Benson, Ibid. 84.3678 (1962) ” W. A. Sheppard and R. M. Henderson, Ibid. 89.4446 (1967) ” R. W. Taft, E. Price, J. R. Fox_ 1. C. Lewis, K. K. Andersen, and G. T. Davis, Ibid. 85.709.3l46 (1963) I6 F. S. Fawcett and W. A. Sheppard, Ibid. 87.4341 (1965) ” ’ M. J. S. Dewar, Hyperconjugation. p 159. Ronald Press, New York, N.Y. (1962); b 0. Exner, Tetrahedron Letters 815 (1963) ‘a W. A. Sheppard, J. Am. Chem. Sot. 85, 1314 (1963) I9 J. H. Smith and F. M. Menger, J. Org. Chem. 34. 77 (1969) r” J. N. Murrell, The Theory of ElectronicSpectra ojOrganic Molecules. Wiley, New York, N.Y. (1963) r’ Y. Nomura and Y. Takuchi. Tetrahedron Letters 639 (1969) ” ’ N. Muller and D. T. Carr. J. Phys. Chem. 67. 112(1963); b W. M. Litchman and D. M. Grant, J. Am. Chem. Sot. 90, 1400 (1968)