SYNTHESES AND PROPERTIES OF HIGHLY STERICALLY
HINDERED ALIPHATIC COMPOUNDS
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
TADAMIGHI FUKUNAGA, M. S,
***************
The Ohio State University 1959
Approved by
Adviser Department of Chemistry ACKNOWLEDGMENTS
The author expresses his sincere appreciation
to Professor Melvin S. Newman for proposing this problem and for the many suggestions he has offered
throughout the course of this work.
The author also wishes to acknowledge the
receipt of the Fellowship sponsored by Wright Air-
Development Center for the periods October 1955 to
March 1956 and April 1957 to June 1959, and by
Research Corporation for the period April 1956 to
March 1957.
- 1 1 - TABLE OF CONTENTS
ACKNOWLEDGMENT ...... i l
INTRODUCTION ...... 1
I. SYNTHESIS OF HIGHLY STERICALLY HINDERED ALIPHATIC NITRIDES ...... 3
A. Historical ...... 3
B. Results and Discussion ...... 12
II. REACTIONS OF HIGHLY STERICALLY HINDERED ALIPHATIC NITRIDES ...... 21
A. Hydrolysis ...... 21
B. Lithium Aluminum Hydride Reduction .... 25
I I I. SYNTHESIS OF HIGHLY STERICALLY HINDERED ALIPHATIC ACIDS ...... 28
A. Historical ...... 28
B. Results and Discussion ...... 33
IV. REACTIONS OF HIGHLY STERICALLY HINDERED ALIPHATIC ACIDS AND THEIR DERIVATIVES ...... 41
A. Ionization Constants of Acids ...... 41
B. Reaction of Acids with Ethoxy- acetylene ...... 52
C. Reaction of Acids with T rifluoro- acetic Anhydride ...... 56
D. Reaction of Acids with Thionyl Chloride ...... 60
E. Estérification of Acid Chlorides ...... 63
F. Reaction of Acid Chlorides with Ammonia and Sodium Amide (Ketene Formation and Reactions) ...... 63
G. Reaction of Esters and Amides with Lithium Aluminum Hydride ...... 75
- i i i - TABLE OF CONTENTS (continued) Zâ£e
V. SUGGESTION FOR FUTUlffi WORK AND GENERAL PROPERTIES OF HIGHLY STERICALLY HINDERED COMPOUNDS ...... 92
VI. EXPERIMENTAL ...... 98
Conventions Used in the Discussion of Experimental Work ...... 98
A. Synthesis of Highly Sterically Hindered Aliphatic N itrile ...... 99
B 1. D iisopropylacetonitrile ...... 99 2. t-B utylacetonitrile ...... 101 3 . Alkylation of diisopropylaceto- n itr ile ...... 102 4. Alkylation of t-hutylacetonitrile 104
B. Reactions of Highly Sterically Hindered Aliphatic Nitrile ...... 106
1. Hydrolysis ...... 106 2. Lithium aluminum hydride reduction 107
C. Synthesis of Highly Sterically Hindered Aliphatic Acids ...... 110
1. Triisopropylacetic acid ...... 110 2. Di-t-butylacetic acid ...... 114 3 . Isopropyl-t-butylacetic acid ...... 122 4. Attempted alkylation of triethyl- methyl dl-t-butylacetate ...... 124 5. Attempted alkylation of triethyl- methyl isopropyl-t-butylacetate .. 126
D. Reactions of Highly Sterically Hindered Aliphatic Acids ...... 127
1. Ionization constants of acids .... 127 2. Reaction of acids with ethoxy- acetylene ...... 137 3. Reaction of acids with triflu o ro - acetic anhydride ...... 139 4. Reaction of acids with thionyl chloride ...... 139 5. Estérification of acid chlorides . I 40
- i v - TABLE OF CONTENTS (continued) Zs-gs
6. Reaction of acid chlorides with ammonia and sodium amide ...... I 4 I 7. Reaction of di-t-hutylketene ...... 143 8. Attempted preparation of diiso- propylketene ...... 145 9. Reaction of esters and amides with lithium aluminum hydride ...... I 46
AUTOBIOGRAPHY ...... 155
-V- LIST OF TABLES
Table Page
I. Alkylation of Nitrile-1 ...... 14
II. Alkylation of Hitrile-II ...... 15
III. Alkylation of Ilitrile-III ...... 16
IV. Conversion of Witriles to Acids ...... 23
V. Physical Constants of Acids ...... 42
VI. Ionization Constants of Acids-I, ...... 45
VII. Ionization Constants of Acids-II...... 47
VIII. Ionization Constants and Structure of Acids ...... 51
IX, Lithium Aluminum Hydride Reduction ..... 79
X, Determination of Ionization Constants of Acids ...... 135
—vi “* LIST OF FIGURES
■Page
1. Direct alkylation of nitriles ...... 4.
2. C- vs. N-ilkylation of hindered n itr ile s ...... 10
3 . Synthesis of di-t-butylacetic acid ...... 35
4.. Lithium aluminum hydride reduction of unsubstituted amides ...... 85
- v i i - INTRODUCTION
The synthesis of highly sterically hindered ali phatic compounds has received little previous attention by synthetic organic chemists and, even though a number of tria lk y la c e tic acids and th eir derivatives have been prepared, they bear n-alkyl groups except in a few cases. In this laboratory a systematic approach to the preparation of hindered aliphatic nitriles^ has been
(1) M. S. Newman and T. Miwa, unpublished resu lts, i n i t i a t e d .
The hindering steric effect on the rate of esteri- 2 fication of hindered aliphatic acids and on the
(2 ) (a) K. L. Loening, k, B. G arrett and M. S.
Newman, J . Am. Ghem. 8 0 c . . 74-. 3929 (1952) j (b) See
M, S. Newman, "Steric Effects in Organic Chemistry,"
John Wiley and Sons, In c., New York, N. Y., 1956,
p. 205 f£.
hydrolysis of nitriles^ has been studied and it has
(3 ) L. Tsai, T. Miwa and M. S. Newman, J. Am.
Ghem. S o c ., 23., 2530 (1957).
- 1 - —2 — 2b,4 been found that the higher the six number, the more
(4 ) M. S. Newman, J. Am. Ghem. Soc., 72, 4783
(1950). slowly the acid is esterified and the nitrile is hydro lyzed. Ionization constants of hindered aliphatic 5 acids have also been shown to be subject to steric
(5 ) G. S. Hammond and D. H. Hogle, J. Am. Ghem.
Sqæ., XL, 3384 (1955). e ffe c ts .
The purpose of th is work was to estab lish general methods for preparing highly sterically hindered aliphatic acids and their derivatives and to study the chemical and physico-chemical properties of these compounds.
The highly hindered compounds of interest in this work have the following general structure: R^CCOOH,
R^CCOGl, R^CCOOR', R^GGONHg and R^CGN where R stands for
ethyl, isopropyl and t-b u ty l.
In order to emphasize the alkyl groups on the a-
carbon atom of these compounds, they w ill be named,
throughout this dissertation, as trialkylsubstituted
acetic acids and their derivatives and not as dialkyl-
substituted acids of RGH2GOOI, although the latter
system is more commonly used. I. SYNTHESIS OF HIGHLY STERICALLY HINDERED
ALIPHATIC NITRILES
I-A. Historical
The most widely used method for preparing tri- alkylacetonitriles is direct alkylation^ of substituted
(1) See A. 0. Cope, H. L. Holmes and H. 0. House,
"Organic Reactions," Vol. IX, John Wiley and Sons, Inc.,
New York, N. Y ., 1957, p. 107 ff.
acetonitriles, as shown in Figure 1. This method was 2 first used for preparing triethylacetonitrile from
(2) M. Bockmühl and J. Ehrhard, German Patent,
473,329 (Chem. Zentr.. 1929 II, 218).
butyronitrile. The general procedure, which consists of
conversion of nitrile to its conjugate base by means of
amide ion followed by heating with alkylating reagents,
was developed by Ziegler.^ It has been satisfactorily
(3) K. Ziegler and H. Ohlinger, Ann.. 495. 84
(1 9 3 2 )} K. Ziegler, French Patent, 728,241 (Cham. Zentr..
1933 I, 1 1 9 7 ).
applied to the synthesis of a number of trialkylaceto-«. nitriles in 50-80% yield.^ Liquid ammonia^*^ has also
-3- - 4"
E l ? 1 \ base , . GH-GW + ------^ E g - G - G N (l) / 3) * I ^2 R3
0*^ggQ R^-CHg-CN + 2R^X »------^ R 3 -C-GN (2 )
*3
?3 OHj-OK + 3K3X K3-O-CK (3)
^3
Fig, 1. - Direct alkylation of nitriles. - 5 -
(4.) L. &. Walter and S, M. McElvain, J. Am. Ghei
Soc. y 5 6 , 1 614 (1 9 3 4 ); G. Newbery and W. Webster, i*.
Chem. S oc.. 738 (1947); C. Schuerch, J r. and E. H.
Huntress, J. Am. Chem. Soc., 70, 2824 (1948); N. Sperber,
D. Papa and E. Schwenk, ib id .. 70., 3091 (1948).
(5 ) F. W. Bergstrom and R. Agostinho, J. Am. Chem.
Soc. . b l, 2152 (1945) .
(6) Liquid ammonia-alkali amide system has been
used for alkylation or benzylation of arylsubstituted
acetonitriles. For example, see D. E. Whyte and A. C.
Cope, ib id . . 65. 1999 (1943); G. R. Hauser and W. R.
Brasen, ibid. , iâ.» 82 (1956).
been used as a solvent for alkylation of acetonitrile
with ethyl and butyl bromides to give the corresponding
mono- and dialkylated acetonitriles.
The lithium , sodium, and bromomagnesium salts of 3 7 secondary amines have also been u s e d , * but to a lim ited
(7 ) J. Cason and G. Sumrell and R. S. Mitchell, J.
Ore. Ghem.. H , 850 (1950).
extent.
Other methods which have been used for the preparation g of tria lk y la c e to n itrile s include dehydration of amides —6“
(8) For example, see R. F. Raffauf, J. Am. Chem.
S o c.. Ik , 4460 (1952).
and hydrogen cyanide addition to o l e f i n s . 9 The f i r s t
(9 ) G. R. Harris and W. W. Deatley, U. S. Patent,
2,455,995 4^, 3439 (1949)].
method has never been applied to the synthesis of homo
logs higher than triethylacetonitrile, probably because
of the scarcity of trialkylacetamides, The last method
may be illustrated by the formation of dimethylneo-
pentylacetonitrile^ from diisobutylene and of dimethyl-
isopropylacetonitrile^ from tetramethylethylene. They
were prepared by passing a mixture of hydrogen cyanide
and the olefin over alumina at 380-450°, 3 Except for diethylisopropylacetonitrile and ethyl- 3 diisopropylacetonitrile , no nitrile having a six number
of more than 12 has been synthesized by any of these
methods,
A systematic study of the preparation of trialkyl-
acetonitriles was initiated in this laboratory^^ by direct
(1 0 ) M. S. Newman and T. Miwa, unpublished re s u lts.
alkylation of less substituted nitriles as shown in - 7 -
Figure 1. Alkylation was carried out by adding a mixture
of nitrile and alkylating reagent in proper ratio to a
known amount of sodium amide suspension in liquid ammonia
followed by refluxing for various times,
Methyldiisopropyl-, ethyldiisopropyl-, dimethyl-t-
butyl- and diethyl-t-butylacetonitrile were thus prepared
in 65-81% yield as summarized in Table I (see page 14 )•
Alkylation of t-b u ty lac e to n itrile with two moles
of isopropyl iodide, however, gave no tr ia lk y la c e to n itrile ,
Instead W-isopropyl isopropyl-t-butylacetamide was ob
tained after treating the reaction mixture with aqueous
acid. When the product was distilled without the prior
acid treatment, W-isopropyl isopropyl-t-butylketenimine
was obtained. Formation of the ketenimine^^ was con-
(ll) (a) Ketenimines were first prepared H. Staud-
inger and J. Meyer, Bar... 53. 72 (l920)j H. Staudinger and
E. Hauser, Helv. Ghim. Acta. 887 (l92l) by treating
phosphinimine with ketenes or phosphine-methylene with
isocyanates and characterized by the hydrolysis to amides;
(b) E. Bergmann and H. A. Wolff, Ber. , 63, 1176 (I 9 3 0 )
reported isolation of a ketenimine-type compound, but
la te r work R. Fosse, Bull. soc_._ chim., [4 ] A2* 171 (1931)
showed that the structure assigned was incorrect;
(c) Recently, C. L. Stevens and J. C. French, J. Am. Chem. - 8-
Soc.. 23, 657 (1953), 2â, 4398 (1954), prepared several ketenimines by dechlorination of a-chloroiminochlorides by means of sodium iodide or by dehydrochlorination of iminochlorides having a single a-hydrogen atom by means of triethylamine. They also reported that ketenimines had a characteristic infrared absorption band at 4*9-
5.0 u and ultraviolet absorption maxima at 268,5 and
375 muj (d) See also G. L. Stevens et a l ., ib id . , 79.
6057 (1957); 80, 4 0 6 5 , 4069 (1958); M. Talat-Erben and
8, Bywater, ib id .. 77, 3710 (1955).
firmed by the presence of a strong infrared absorption 11c band at 5.0 u and by the formation of N-isopropyl
isopropyl-t-butylacetamide on hydrolysis. Alkylation of
isopropyl-t-butylacetonitrile with ethyl and isobutyl
bromides, however, yielded mixtures of the corresponding
ketenimine and trialkylacetonitrile. Alkylation of
acetonitrile, mono- and diisopropylacetonitrile with iso-
propyl iodide also afforded a mixture of triisopropyl-
acetonitrile and N-isopropyl diisopropylketenimine, ,
These resu lts are summarized in Table II (see page 15).
Simultaneous formation of trialkylacetonitrile and
ketenimine may be explained by C- and N-alkylation of the 12 ambident conjugate base of the starting nitrile as
(1 2 ) W. Kornblum, H. A. Smiley, R. K. Blackwood and
D. 0. Ifflund, J. am. Ghem. Soc. . 77, 6269 (1955). - 9- shown in Figure 2, The result represents the first ex ample in which the course of alkylation of disubstituted acetonitrile is controlled by steric effects, although 13 N-alkylation has been predicted from the fact that the
(13) M. Prober, J. km. Chem. Soc. . 2Û., 2274 (1956), conjugate bases of acetonitrile and phenylacetonitrile react with trimethylchlorosilane to give in part silicon analogs of ketenimine as shown below.
CH3CN + 2(0^^)38101 -Sâ-^ L'(CH3)3Si] gCHCN +
(0H3)3SiCH=C=NSi(CH3)3 (,)
G/Hc NaNHo D 5 \
C^H^-GH=C=N-Si(OH3)3 (5)
13 These results have been attributed to the stronger nucleophilicity of the nitrogen as compared with the a-carbon atom of the nitrile anions as well as to a large steric requirement of trimethylsilyl group, which renders difficult its approach to the a-carbon atom.
Some disubstituted acetonitriles bearing strongly electronegative groups such as methylsulfonyl, phenyl- sulfonyl, methylsulfoxy or carbethoxy groups have been known^^ to react with diazomethane to give the corres-
(1 4 ) R. Kijkstra and H. J. Backer, Rec. trav. chim.,
.73, 575 (1 9 5 4 ). -10-
Rl CH - C = N R.
\ / R R.
Rgl
^1 I Ri R, — c — c G = C = N - R . t R, Ro
Fig. 2. - C- N-Alkylation of nitriles, —11— ponding ketenimines as shown in equation 6,
R R ^CH-C=N + GHoNo ^ ^C=G=N-CHo (6) R- 2 2 3 where R stands for CH^SO^-, GH^SO- or G2H^0G0—,
In this case, however, the reaction appears to he con
trolled by polar effects. It is noteworthy that the crude
alkylation product of ethyl t-butylcyanoacetate with iso-
propyl iodide in the presence of sodium ethoxido had^^
an infrared absorption band a t 5.0 u ch aracteristic of
ketenimines^^® although the product was not characterized.
The Ritter reaction^^ and the alkylation of nitriles
(1 5 ) J. J. R itte r ejb al ^, J. Am. Ghem. S oc.. 7A,
763 (1 9 5 2 ) and references contained therein,
with active alkylating reagents in the presence of inor
ganic silv er salts^^ also yield amides as fin a l products
(1 6 ) F. W. Upson, R. T, Maxwell and H. M. Parmelee,
J , Am, Chem. Soc.. 12, 1971 (1930); J. Cast and T. S,
Stevens, J. Ghem. Soc., 4180 (1953)j S. J. Gristol and
J. E. Leffler, J. âm. Ghem. Soc.. 76, 4468 (1954).
as illustrated by equation 7 and 8, respectively. These
reactions, however, probably proceed through imidol-
type intermediates (l). - 1 2 - CHo CH3 I ^ TT m I RlC=N + R i-G .= N-C-Rg R1CONHC-R2 (?)
^ OSO3H R3 R3
AgNOo , . H 2O CHjCN+tCaHgigO'Cl — CH3-C=N-C >
ONO2 GH^GONH'CfC^Hgjg (8)
I
The initial aim of this work was to study the forma
tion of ketenimine during alkylation of nitriles. This
plan was, however, eventually set aside because of the
difficulty of separating ketenimines from reaction mix
tures as well as the fact that new and interesting reac
tions of highly hindered acid derivatives were discovered
in the course of th is work.
I-B. Results and Discussion
Results. Diisopropyl- and t-butylacetonitrile which
were needed for alkylation studies were prepared from
ethyl cyanoacetate in 70 and 43% overall yields, respec
tively, according to the previously described methods^^"^^
(17) E. C. B. Marshall, J. Ghem. Soc.. 2754 (1930).
(1 8 ) S. Wideqvist, Acta Ghem. Scand. , 3. 303 (1949).
(1 9 ) A. J. Birch, J. Ghem. Soc.. 2721 (1949). - 13- with a few modifications.
The alkylation of nitriles was carried out by allowing the nitrile to react with alkali amide in liquid
ammonia for one hour to form the conjugate base of the
nitrile which was alkylated by the alkyl halide over a o period of 12-30 hours at the boiling point (-33 ) of
liquid ammonia. The re su lts and reaction conditions are
summarized in Table II I,
Formation of ketenimines (N-isopropyl diisopropyl
ketenimine and N-isopropyl isopropyl-t-butylketenimine)
was confirmed by the characteristic strong absorption band 20 at 5.0 u and by hydrolysis to the corresponding N-
(20) G. L, Stevens and J, G, French, J. Am. Ghem.
Soc... 16, 4398 (1954).
substituted amides. Separation of the ketenimines by
fractional distillation was not possible because the
boiling points were toe close to those of other components,
such as the isomeric or the starting nitriles, in the re
action products.
The ketenimines in the mixture were quite stable and
could be stored at room temperature for at least three 21 months without any change in the infrared spectra,
(21) It is reported in reference 20 that N-n-butyl
ethyl-n-butylketenimine was unusually stable but others. TABLE I
Alkylation of N itriles, R^RgCHCN, with
Mole Ratio Reaction Yield % Run R^R^GHCN R_ X Time 3 R3X /n itrile hr. (crude)
Methyldiisopropylacetonitrile
1 CH3 H i-G_Hy I 2 .1 27 59 2 i —G^Hy 1—G^Hy CH^ Hr 4*4 30 81 I H I Ethyldiisopropylacetonitrile 3 GgH^ H i — Br 2 .1 24 58 4 GgHc H i-G^H? I 2 .2 24 77 5 i-GjHy i-Gj&y GgHg Br 1.5 30 78
Dimethyl-t-butylacetonitrile 6 t-G Eg H CH3 Br 4*4 35 78
Diethyl-t-butylacetonitrile 7 t-G Eg H GgEg Br 2.2 40 65
Sodium amide was employed as reagent. TABLE II
Carbon- and Nitrogen-alkylation of N itriles
Mole Ratio Run R^RgCHCN Mole Ratio Reaction Yield (ÿ C RjZb Time d C-alkyl Rl Eg R^X/nitrile h r , C-alkyl N-alkyl N-alkyl
8 H H i-C^H? 3,3 25 18 14 1.3
9 i-CgH? H 2.2 30 34 12 2,8 I H 10 1—G i —C^Hy i-C^H? 2,1 25 37 23 1 .6 vn I 11 I-G3H7 CzHg* 1,2 24 56 small^ large 12 H i-C^Hy 2,1 23 0 38 0
13 t —G^Hg H i —C^Hi^ 3.6 30 0 28^ 0
14 t-C^Hg i —C^H^ l-CjHgS 1,1 24 38 12^ 3.2
8 Sodium amide was employed as reagent. Iodides were used unless otherwise stated, ® Crude yield. Isolated as N-substituted amide unless otherwise stated, ® Bromides were used. ^ Isolated as ketenimines. TABLE III
Carbon- and Nitrogen-alkylation of Nitriles®
R^X^ Mole Ratio Reaction Mole Ratio Run Yield % G-alkyl E3 K3X /n ltrile ^1 %2 G-alkyl N-alkyl° N-alkyl
1 5 " i —G^Hy i —G^Hy 1.3 30 39° 2 6 ^ 1,5 1 6 ® i-GgHy i-G_Hy 1.3 18 50^ 2 5 ^ 2 ,0 17 i.—G^Hy i — i-C^Hy 1.3 30 45^ 19^ 2,4 3 3 k 18 l — C ^Hy i —C 1 —G ^Hy J 1.4 12 2 5f 1.3 I 19 ^Hy i —G^Hy i —G^Egj 1.1 15 8 4 ^ 1 ® large I 2 08 t —G^Hg H i —G^Ey 2,1 30 0% 56° 0
® Potassium amide was employed as reagent unless otherwise stated. ^ Alkyl iodides were used as alkylating reagent unless otherwise stated. ° Estimated by isolation as N-substituted amides. d Lithium amide was used. ® b.p. 98-100° at 10 mm. M. S. Newman and T. Miwa, unpublished, reported b.p. 220-221° for triiso- propylacetonitrile, f.m.p. 167-168°. M. S. Newman and T. Miwa, unpublished, re ported m.p. 168,0-168.8 for N-isopropyl diisopropylacetamide. * Sodium amide was used. ^ b.p, 107-108° at 15 mm. i b.p. 98-100° at 10 mm. j Bromide was used. b.p. 80-83° at 4 Dim. ^ b.p. 79-80 at 3 mm., n^^ 1.44^0. Anal. Galcd. for G1 2 H2 3 N: G, 79.5; H, 12.8; N, 7.7. Found: G, 79.6, 79.7; H, 13.0, 13.0; N, 7.7, 7.7. m.p. 1180 (uncorrected), ^ A 43^ yield of isopropyl-t-butylacetonitrile, b.p, 7 2 , 0- 7 2 . 5 ° at 15 mm. [L. Tsai, T. Miwa and M. S. Newman, J. Am. Chem. Soc.. 79, 2 5 3 0 (l957)J, was obtained. ° m.p. 149—150°. M. S. Newman and T. Miwa, unpublished, reported m.p. 150,0-150.6° for N-isopropyl isopropyl-t-butylacetamide. - 1 7 - such as N-p-tolyl dimethylketenimine and W-methyl di- phenylketenimine, were unstable at room temperature and turned to viscous oils during several days.
When the ketenimines were distilled at atmospheric pressure a small amount of gas (probably olefin) which decolorized permanganate solution was very slowly evolved during distillation.
The ketenimines were very stable to cold water and not hydrolyzed when washed with water. However, they were 22 quite sensitive to dilute aqueous acid and hydrolyzed
(2 2 ) It is reported in reference 20 that ketenimines were only slowly hydrolyzed in aqueous acetone at room temperature but the addition of hydrochloric acid greatly accelerated the reaction, exothermically to give the corresponding amide.
Discussion, The ratio of C- to N-alkylation obviously decreases with increasing steric requirement of the sub stituent groups of nitrile to be alkylated (compare run
1 , 4 , 16 and 2 0 ) as well as that of alkylating reagent
(compare run 2, 5, 18 and 19), As long as the expected trialkylacetonitrile has a six number of less than 1 5 ,
N-alkylation is negligible. In reactions leading to nitriles having a six number of 18 C~ and N-alkylation - 1 8 - become equally important (run 8 - 1 1 and 14-18), and in those leading to nitriles having a six number of 21, only
N-alkylation is observed (run 12, 13 and 20),
The change of the course of reaction is thus in good accordance with the Rule of Six^^ originally proposed for
(2 3 ) See M. S. Newman, "Steric Effects in Organic
Chemistry," John Wiley and Sons, Inc., New York, N. Y.,
1 9 5 6 , p. 206j L. Tsai, T, Miwa and M. S. Newman, J . Am.
Chem. S oc.. 2 1 , 2530 (1957). carbonyl addition reactions and may be attributed to the steric strain (F strain) involved in the transition state leading to trialkylacetonitrile. The transition state for
N-alkylation, on the other hand, is less subject to hinder ing steric effects, consequently ordinarily insignificant
N-alkylation becomes important as steric effects about carbon increase.
The fact that the yield of product was increased by allowing one hour for formation of the conjugate base of the parent nitrile indicates that steric factors also in fluence the acidity of the a-hydrogen in hindered nitriles
(compare run 10 and 16; 12 and 20),
Potassium and sodium amides were equally effectiv e, 2 / but lithium amide appeared to be a poor reagent (compare - 1 9 -
(24) C. R. Hauser and W. J. Chambers, J. Am. Ghem.
8 o c ., 78, 3837 (1 9 5 6 ), reported a similar trend for alkylation of triethylmethyl esters of dialkylacetic acid s,
1 5 , 16 and 1 7 ). Isopropyl iodide seems to be a better 2 5 alkylating reagent than the bromide (compare run 3
(2 5 ) M. S. Newman and T. Miwa, unpublished re s u lts,
reported that diethyl sulfate was a better alkylating re
agent than halides for éthylation of n-butyronitrile.
Ethyl methanesulfonate and benzenesulfonate did not give
triethylacetonitrile but diethylacetonltrlle in ca. 30#
yield.
and 4; 17 and 18).
It may be noteworthy that reaction temperature plays
an important role at least when isopropyl iodide is used
as alkylating reagent. Diisopropylacetonitrile was re- 25 covered after treating the anion with isopropyl iodide
in tetrahydrofuran at room temperature. On the other
hand, butyronitrile anion reacted with isopropyl iodide
in liquid ammonia at -33° for 2 4 hours to give ethyl-
diisopropylacetonitrile in 77# yield whereas at -70° for
40 hours the yield was only 4 6 #. This is probably due
to the larger temperature coefficient of the rate of - 2 0 - base-catalyzed dehydrohalogenation of isopropyl iodide as compared with that of alkylation of the nitrile anion.
Although it is anticipated that the yield of tri- alkylacetonitrile containing at least two different alkyl groups will not be affected much by the order of intro duction, i t seems b e tte r, judging from the few available data, to introduce the smaller group or primary alkyl group la s t (compare run 1 and 2; 3 and 5). II. REACTIONS OF HIGHLY STERICALLY HINDERED
ALIPHATIC NITRIDES
II-A. Hydrolysis
Historical. Conversion of moderately hindered nitriles to the corresponding amides or acids has received consider able attention and a number of different reagents has been used. These reagents^ include alkali metal hydroxides in
(1) il, Sperber, D. Papa and E. Schwenk, J. Am. Chem.
Soc.f 2S1, 3091 (194 #) and references contained therein. various solvents, concentrated hydrochloric acid, alkaline hydrogen peroxide, phosphoric acid, acetic and sulfuric acids mixture and sulfuric acid of various concentrations.
These reagents, nowever, failed to hydrolyze tr i- n - butylacetonitrile to the acid. Conversion of the nitrile
to the corresponding amide was successful only by heating % 1 on the steam bath with 60^ sulfuric acid for 12 hours.
By taking advantage of this reaction, complete hydrolysis
of moderately hindered nitriles to the corresponding acids
has been accomplished by a two-stage process, hydrolysis
to amide by concentrated sulfuric acid followed by treat
ment of the amide with sulfuric and nitrous acids.
(2 ) C. Schuerch and E. Huntress, J. Am. Chem. Soc., 2 0 , 2824 (1 9 4 8 ). ______
— 21 — —22—
(3) L. Tsai, T. Miwa and M. S. Newman, i b i d . . 79
2530 (1957).
(4) This method first applied to hindered aromatic nitriles by L. Bouveault, Bull, soc. chlm.. [3]368
(1893); J. J. Sudborough, J . Chem. Soc. ^ 67, 601 (1895).
Hindering steric effects on the hydrolytic step are 5 well correlated by the Rule of Six originally proposed
(5) M. S. Newman, J. Am. Chem. Soc.. 72, 4783 (1950). for carbonyl addition r e a c t i o n s . ^ Thus, nitriles having a six number of less than 12 were hydrolyzed to amides by treating with 75^ sulfuric acid at 140° for one-half hour.
The amides were then converted to acids by addition of ex cess sodium nitrite at about 50° in more than 63% overall y ie ld s. ^
Results and discussion. By this method, t-b u ty l-, diisopropyl-, methyldiisopropyl- and ethyldiisopropylaceto- nitrile, ethyl-t-butyl- and dimethyl-t-butylacetamide were converted to the corresponding acids. The results and reaction conditions are summarized in Table IV.
The two-step sulfuric acid-nitrous acid treatment is, however, of little value for highly hindered nitriles having a six number of more than 1 5 . For example, ethyl- diisopropylacetonitrile gave only a 30^ yield^ of the acid TABLE IV Conversion of Nitriles to Acids via Amides
NaN02 N itriles, 3G.GN S2SO. Time Temp. mole/mole Yield g./g. n itr ile $ n itr ile hr. °G. or amide Acid $ «3 i — i —G^Hfjf E 75 1.9 0.25 150 1.5 85 i-GjHy i-GjHy CE3 75 2.4 1 140 1.5 63 wI i-GjEy i-GjEy C2E3 85 5.9 20 98 3.0 30® w I E E 75 1 .8 0.25 140^ 1 .6 92
B° 75 7.1 - - 1 .8 76 t —G GE3 CE ° 75 9.3 -- 2 .6 70 3
® The amide was isolated in 33$ yield under the described conditions and con verted with sodium nitrite in 75$ HgSO^ to the acid in 92$ yield. ^ When the nitrile was heated in sulfuric acid at 130° violent reaction occurred so that external heat ing was discontinued. After '-he exothermic reaction had subsided it was heated at 140° for 15 min, ° The amides were employed instead of n itr ile and treated with a specified amount of sodium n itr ite at approximately 50°. —24-“
(6 ) A less than 20% yield is reported in reference 3. under more drastic conditions than those required for the others. Triisopropylacetonitrile (six number 18) was re covered almost quantitatively after heating with 97% sulfuric acid for a week.^ Attempted hydrolysis of this nitrile under the following conditions was not successful; concentrated hydrochloric acid in a sealed tube at 2 0 0 ° for 3 days; potassium hydroxide in ethylene glycol under
(7) M. S. Newman and T. Miwa, unpublished re s u lts.
7 reflux for 7 days; and boron trifluoride-acetic acid complex at 125° for one-half hour. ^ Polyphosphoric acid
(8 ) C, R. Hauser and D. S. Hoffenberg, J. Ore.
Chem.. 2 0 , 144-8 (1955).
(80 %) was not effective? even for ethyldiisopropylaceto- nitrile (at 85° for 23 hours). In these reactions,the starting nitriles were recovered in high yield.
The N-substituted amides obtained by acidic hydroly sis of the ketenimines were extremely resistant to further hydrolysis. For example, N-isopropyl diisopropylacet- amide was not attacked at all by refluxing with 35% potassium hydroxide solution for 24 hours, with 40% -2 5 - 7 hydrobromic acid in acetic acid for 45 hours or with
concentrated hydrochloric acid in dioxane for 24 hours,
or by treating with sodium nitrate in 75^ sulfuric acid
at 50°. The inertness of this amide may be comparable
to or more pronounced than that of triisopropylaceto-
n itr ile since both compounds have 18 atoms in the six
position if one considers both acyl and amino portions
of the amide.
II-B, Lithium Aluminum Hydride Reduction of Nitriles
Introduction. Since triisopropylacetonitrile was
not convertible to the corresponding acid by any hy
drolytic method (see previous section), the reduction
with lithium aluminum hydride was undertaken to obtain
the imine or amine which might have provided a possible
precursor of triisopropylacetic acid. It has been found^
(9) M. S. Newman and T. Miwa, unpublished re su lts.
that triisopropylacetonitrile was quantitatively recovered
after treating with lithium aluminum hydride for 4 hours
at room temperature, but was reduced by refluxing with
ona mole of lithium aluminum hydride in te trahydrofuran
for 24 hours to give a mixture consisting of 2 ,2 , 2- t r i -
isopropylethylamine ( 4?%), triisopropylacetaldimine ( 46#)
and the starting nitrile (7#).^^*^^ —2 6—
(10) When one-quarter mole of lithium aluraunum hy dride was used a mixture consisting of the amine (12%), the imine (67%) and the n itr ile (21%) was obtained.
(11) The proportion of the components in the mixture was determined, without separation, by preparing picrates of the amine and imine, and the 2,4—dinitrophenylhydra-
zone, m.p. 157.5-158.7°, of the aldehyde derived from the imine.
Results. Reduction of triisopropylacetonitrile with
lithium aluminum hydride was repeated and an e ffo rt was
made to isolate the reduction products. When the nitrile
was treated with 1.3 mole of lithium aluminum hydride in
refluxing ether for 4 days, 2,2,2-triisopropylethylamine
was obtained in 97% yield.
Similar runs with less than one-half mole of lithium
aluminum hydride, however, afforded always a mixture of
the amine, aldimine and the starting nitrile, and close
similarity of their boiling points and adsorption pro
perties on alumina prevented successful separation of
the aldimine. Hydrolysis of the mixture with aqueous acid
afforded triisopropylacetaldehyde which formed a 2 ,4 -d i-
nitrophenylhydrazone and had a characteristic aldehyde 12 absorption band at 5.8 u. On keeping at room temperature, (12) L. J. Bellamy, "The Infrared Spectra of Complex
Molecules," John Wiley and Sons, Inc. New York, 1958, p. 155. -2 ? - the mixture slowly evolved ammonia to yield the sûüehyde.
Although pure triisopropylacetaldimine could not be separated, it is noteworthy that, even though the reaction mixture was decomposed with small amounts of water and aqueous base, triisopropylacetaldimine was not attacked, Aldimines are in general extremely unstable and hydrolyzed instantaneously to aldehydes on treatment with water. They behave like aldehydes in inert solvents.
1 3 ,1 4 Triisopropylacetaldimine might, thus, be the most
(1 3 ) Triethylacetonitrile was reduced with insuffi cient amount of lithium aluminum hydride to give the
corresponding amine as well as the Schiff base, see reference 9.
(1 4 ) W. Busch, B ar., 29, 2143 (I 8 9 6 ), isolated benzaldimine hydrochloride; E. P. Kohler and N. L. Drake,
J . Am. Chem.-. Soc,, 45. 1281 (1923), isolated diphenyl-
acetaldimine; V. Grignard and E. Escourrou, Gomnt. rend.,
180. I 883 (1 925 ), isolated phenylacetaldimine.
stable aliphatic aldimine presently known although some
aromatic aldimines, which decomposed in a ir with loss
of ammonia, have been isolated, Ill. SYNTHESIS OF HIGHLY STERIGALLY HINDERED
ALIPHATIC ACIDS
III-A. Historical
The most widely used and well developed method for the synthesis of trialkylacetic acids is the hydrolysis of the corresponding nitrile by two step sulfuric acid- nitrous acid treatment. This method has been discussed elsewhere^ in more detail.
(l) See Chapter II-A, p. 21 ff.
The second most widely used method is probably the 2 3 Haller-Bauer reaction ' involving the cleavage of aryl
(2) For an early review, see A. Haller, Bull, soc. c-him., France, 141 21, 1073 (1922).
(3) C. L. Carter and S. N. Slater, J. Chem. Soc.,
130 (1946 ); H. P. Buu-Hoi and P. Cagniant, Rec. trav. chlm., 65, 246 (1946 ). t-a lk y l ketones by means of sodium amide to tria lk y l- acetamides and aryl hydrocarbon. Hydrolysis of the amides yields the corresponding acid. Both the cleavage and hydrolysis steps, however, become extremely difficult when the molecular weight and the complexity of the parent ketones increase. Thus, only a very poor yield
— 20 — -2 9 - of dlethylhexylacetic acid was obtained from diethyl- hexylacetophenone which cleaved mainly to form benzamide
instead of the desired trialkylacetamide,^
Direct alkylation of esters of dialkylacetic acids^
(4-) See k. C. Cope, H. L, Holmes and H. 0. House,
"Organic Reactions," Vol. IX, John Wiley and Sons, Inc.,
New York, W. Y., 1957, Chapter U, p. 107 ff,
has been carried out s a tis fa c to rily by means of sodium
triphenylmethide^ in ether, but the hydrolysis of the
(5) B. E. Hudson and C. R. Hauser, J. Am. Chem. Soc.,
62, 2457 (1940).
resulting esters, generally ethyl or methyl esters, is
difficult or impossible^ if the acids are highly sterically
(6) M. S. Newman, "Steric Effects in Organic Chem
is try ," John Wiley and Sons, Inc., New York, N. Y., 1956,
p. 205 f f .
hindered.
A general method for the synthesis of trialkylacetic
acids, involving alkylation of triethylmethyl dialkyl-
acetates by means of alkali amide followed by the acidic 7 hydrolysis of the resulting esters, has been developed. —30—
(7) C, R. Hauser and W. J. Chambers, J. Am. Ghem.
S oc.. _Za, 3837 (1956).
The success of this method depends on hindering the usual
attack of amide ion at the carbonyl carbon of the dialkyl
acetic acid esters by using the bulky triethylmethyl group
as the alkyl portion of the ester and also on the fact
that the alkylated esters may be acid-cleaved. In this
way the abstraction of the a-hydrogen of the ester as re
quired for the alkylation is the only reaction course
permitted. The alkylated esters were readily converted
to trialkylacetic acids in over 90^ yield. The ease of
this hydrolysis is due to the facile alkyl-oxygen fission
(8) See G. K. Ingold, "Structure and Mechanism in
Organic Chemistry," Cornell University Press, Ithaca,
N. Y., 1953, p. 779.
of the t i r t i a r y e ste r. Thus, triethylm ethyl esters of
diethylacetic and ethyl-n-butylacetic acids were alkylated
with propyl, butyl and octyl halides, and, after the
hydrolysis, the corresponding trialkylacetic acids were
obtained in 70-84% yield.
Other methods for preparing trialkylacetic acids
which have been used to a limited extent involve the
carbonation of appropriate Grignard r e a g e n t s , the -3 1 -
( 9 ) F. C. Whitmore and D. E. Badertcher, J. Am.
Chem. Soc... 1559 (1933).
(10) G. Schuerch, J r. and E. H. Huntress, J. Am.
Chem. Soc.. 70, 2824 (1948). rearrangement of a-bromoketones (Faworaki rearrangement)?'^
(11) A. A. Sacks and J. G. Aston, J. Am. Chem.
S oc., 73, 3902 (1951) and references contained therein. and the oxidation of olefins.The f i r s t method
(12) F. C, Whitmore and K, G. Laughlin, J. Am. Chem.
Soc., 56, 1128 (1934)Î F . G. Whitmore and C. D. Wilson, ibid. . M., 1397 (1934).
(1 3 ) A. Byers and W. J. Hickinbottom, J. Ghem. Soc.,
1331 (1948). involves the difficulty of preparing Grignard reagents from highly branched tertiary alkyl halides. For example, triethylacetic acid is obtained in only 7% yield^^ from triethylmethylmagnesium chloride.
The Faworski method may be illustrated by the syn
thesis of neopentyldimethylacetic acid^^ from 3,5,5-tri- methylhexan-2-one. The rearrangement was carried out in ether to give methyl dimethylneopentylacotate in 78% yield. —32— GH3 CH3 0 ÇH3 CH^
CHj-Ç-OHj-M - Ô-CH3 GH3-Ç-OE2-Ç-OOOOH3
GH3 CH, GH^
GH_ CH„ I ^ I 3
HI CH-3—G-CHo“ C-COOH J f * I GH^ GH^
Hydrolysis of the ester with constant boiling hydroiodic acid afforded the acid in 70% yield. This method is limited by the availability of the parent a-bromoketones.
It may be noted that diisopropylacetic acid and isopropyl- t-butylacetic acid which are of in te re st in this work have 1 1 been prepared in 60 and 42^ yields from the a-bromo- derivatives of isopropyl isobutyl ketone and isopropyl neopentyl ketone, respectively.
The third method, the oxidation of olefins, has been used in only a few cases. For example, chromic oxide
oxidation of triisobutylene 12 afforded dineopentylacetic
and methyl-t-butylneopentylacetic acids, and diisobutyi-
ene^^ gave dim ethyl-t-butylacetic acid.^^
(1 4 ) F . G. Whitmore, R. E. Marker and L. Plambeck,
Jr., J . Am. Chem. Soc. . 63, 1626 (l94l), prepared this
acid by permanganate oxidation of 2,2,3,3-tetramethyl-
butanol which was obtained from 2,2,3,3-tetramethylbutyl-
magnesium chloride by air oxidation. -33~ D i-t-butylacatic acid,^^ which is the most highly hindered
(1 5 ) à. Arkell, Ph.D. Dissertation, The Ohio State
University, 1958,
(six number 18) dialkylacetic acid presently known, has been prepared from di-t-butyl ketone in 21% yield. The details will be discussed later.
III-B. Results and Discussions
Results. In relation to further work t-butyl-, ehtyl-t-butyl-, dimethyl-t-butyl-, diisopropyl-, methyl diisopropyl- and ethyldiisopropylacetic acids were prepared from the corresponding n itr ile or amide by sulfuric acid-sodiura nitrite treatment as discussed in the previous chapter,
(1 6 ) See Chapter II-A, p, 21 ff.
Since triisopropylacetonitrile could not be converted to the corresponding acid by any hydrolytic method,the
Hauser alkylation procedure was applied to the synthesis of this acid. Diisopropylacetic acid was converted to the acid chloride (89-93#) by means of thionyl chloride. The acid chloride was reacted with sodium triethylm ethoxide in ether to give the triethylmethyl ester (80-88#), which was alkylated with isopropyl iodide in the presence of -3 4 - potassium amide in liquid ammonia containing a little ether to yield triethylmethyl triisopropylacetate ( 46^).
The ester was then hydrolysed to give triisopropylacetic acid (95%) by refluxing with concentrated hydrochloric acid in dioxane.
Since this alkylation was successful, the synthesis
of more hindered acids was attempted. Di-t-butylacetic
acid was prepared by a modification of the method^^ de
veloped in this laboratory as shown in Figure 3. Diiso
propyl ketone was methylated with methyl iodide and sodium
amide to give isopropyl t-butyl ketone (83%) which was
further methylated with dimethyl sulfate to give di-t-
*1 rj butyl ketone (60%). The ketone was then reacted with methyl
(1 7 ) A. Arkell, loc. cit.. reported a 41% overall
yield by using dimethyl sulfate for the both steps. See
also F. C. Whitmore and E. E. Stahly, J. Am. Ghem. S o c.,
55, 4153 (1933 ); A. Haller and E. Bauer, Gomnt. rend.,
150. 582 ( 1910 ); I. N. Nazarov and I. L. Kotlyarevskii,
J . G8n._Ghem._ U.S.S,r , ^ 20, 1449 (1950), [C.A.. Ai,
1965 (1951)] .
magnesium iodide to give 1,1-di-t-butylethanol (71%).^^
(1 8 ) A. Arkell, loc. cit.. reported 75% yield by
using methyl lithium and 46% by méthylmagnésium iodide.
See also J. B. Gonant and A. H. B latt, J. Am. Ghem. Soc., 0 0 n 1. NaNH2 II (CH.)pCH-C-GH(CH,)------> (GH„)-C-C-CH(CHo)P 1. NaHH2 J ^ 3 / 2. GH3 I ^ ^ ^ 2. (GHjjgSO^
OH GHjMgI I SOGl, (GHjjjG-G-GfGHg)^ (GH3 ) 3G—G—G ( GH3 )3 pyridine GHo
(GH ),G^ NaBH/ (03^ ) 3%, C=GHp ----- ^ ^CH-GHg B (GH^)^(/ AIGI 3 (GH^JjG OH- I \jj VJ1 I
GrO^^ (GH^)^G GH-GHpOH \ y'GH.COOH
Fig. 3, - Synthesis of di_t-butylacetic acid. -3 6 -
51. 1227 (1929)î F. C. Whitmore and K. C. Laughlin, i b i d . . 55. 3732 (1933); N. Nazarov, 62, 18 (1936).
The alcohol was dehydrated with thionyl chloride and
pyridine to give 1,1-di-t-butylethylene (78%). The ole
fin was then reacted with diborane^^ generated from sodium’
(1 9 ) H. C. Brown and B. G. Subba Bab, J. Am. Chem.
Soc.. 28, 5694 (1956).
borohydride and aluminum chloride in diglyme to give pre
sumably tri-(2 ,2 -d i-t-b u ty leth y l-)b o ran e.^ 0 The crude
(2 0 ) Although this compound was not isolated,
ordinary olefins formed the corresponding trialkyl borane,
see reference 19.
borane was oxidized with 30% hydrogen peroxide in alooholic
sodium hydroxide solution to 2,2-di-t-butylethanol ( 63-
70%). The pure sample of the alcohol, m.p. 54*0-55.0°,
obtained by recrystallization followed by sublimation,
was convertible to the p-nitrobenzoate by refluxing with 21 p-nitrobenzoyl chloride in benzene for one hour.
(2 1 ) The crude alcohol, m.p. 51-54°, gave oil by
treating with this reagent.
The failure of the previous worker^^ to make the crystal- -3 7 - line derivatives was probably due to impurities in the alcohol.
The alcohol was readily oxidized with two equivalents
of chromic oxide in acetic and sulfuric acids to give di-
t-butylace tic acid (77-02%), The product was pure
enough for further reactions, although it melted eight
degrees lower than the pure sample, m.p. 80,5-81.5°.
Di-t-butylacetic acid was then treated with thionyl 2 2 chloride to give the acid chloride (97%), which was
(22) See Chapter IV-D, p. 60 f f .
converted to the trie thylme thyl ester (30%>)^^ by reacting
(2 3 ) The low yield and the high conversion (9%.%)
suggest the formation of di-t-butylketene as a by-product.
This experiment was done before the reaction of the acid
chloride with sodium amide, see Chapter IV-F, p. 64-ff.
with sodium triethylmethoxide in ether.
This ester was subjected to alkylation with methyl
and ethyl iodides by means of potassium amide in liquid
ammonia and ether, but the starting ester was recovered
quantitatively and no alkylated ester was obtained.
Triethylraethyl dialkylacetate anions are generally
green^4 under the reaction conditions. However, the — 38—
(24) The anions formed from esters of diisopropyl acetic and t-butylacetic acids are green. See also reference 7, system of the di-t-butylacetate and potassium amide never became green even when the ratio of the two solvents, 2 5 liquid ammonia and ether, was changed, While the
(2 5 ) The ra tio of solvents was changed by evaporating liquid ammonia slowly from the reaction mixture at room temperature for several hours,
amount of liquid ammonia exceeds considerably that of ether, however, the reaction mixture consists of two
immiscible layers, and as the amount of ammonia decreases
black solid; probably potassium amide, deposits on the wall of the flask,
Triethylraethyl isopropyl-t-butylacetate was treated with potassium amide in liquid aramonia-ether. The mix ture became green but 83^ of the ester was recovered after treating with isopropyl iodide for 2 hours. When ethyl iodide was used, a small amount of higher boiling fraction, probably a mixture of the di- and trialkylacetate, was
obtained. The infrared spectrum of the residue, 1,5 g,
(ca. 30%), exhibited a strong band at 5,93 u although O ^ the starting ester had a band at 5,88 u. -3 9 -
(26) The infrared spectrum of the residue differed from that of the starting ester also in the 7.6-8,8 u region.
This alkylation method was satisfactorily applied to monoalkylation of triethylmethyl t-butylacetate with isopropyl iodide to give triethylmethyl isopropyl-t- butyladetate (70%), which was then acid-hydrolyzed to the corresponding acid (90%).
Discussion. The unsuccessful attempt to alkylate the triethylmethyl esters of di-t-butylacetic and iso- propyl-t-butylacetic acids must be due to: (l) the insolubility of the esters in liquid ammonia and potassium amide in ether; (2) steric effects in the esters which inhibit the approach of amide ion to the a-hydrogen of the esters to abstract the a-hydrogen; or (3) the failure of the ester anions to be alkylated. The third factor may be the result of the following two effects: (a) the inertness of the ester anions to alkylation; (b) the un favorable ratio of the rate of dehydrohalogenation of alkylating reagent to that of alkylation of the ester anions. Large steric requirements of substituent groups around the a—carbon of the ester anions probably make it quite difficult for alkylating reagents to approach the a-oarbon. On the other hand, the inherent ability of alkyl halides, especially secondary and tertiary ones, to —4.0— undergo base-catalyzed dehydrohalogenation would not be subject to hindering steric effects of base as profoundly as the alkylation of the ester anion might be.
The failure of the attempted alkylation of triethyl methyl dl-t-butylacetate Is most likely attributed to the failure of the ester to form a carbanlon since no color change was observed throughout the reaction In spite of the vigorous stirring to Insure good mixing. If this Is the case, this reaction represents the f i r s t example In which an acid-base reaction Involving amide Ion Is s te ric ally Inhibited,
On the other hand, In the alkylation of the ester of isopropyl-t-butylacetlc acid the third factor must be mainly responsible. Although two effects discussed before are not distinguishable from the data, it must be that the resistance of the ester anion toward alkylation Is causing the reaction to fall since the base-catalyzed elimination reaction of ethyl Iodide Is not significant 27 In most circumstances,
(27) C. K. Ingold, "Structure and Mechanism In
Organic Chemistry," Cornell University Press, Ithaca,
N, Ï., 1953, p. 4-3ÔJ J. Hlne, "Physical Organic Chemistry,"
McGraw-Hill Book Company, Inc., New York, N. Y., 1956,
p. 180. VI. REACTIONS OF HIGHLY STERIGALLY HINDERED
ALIPHATIC ACIDS AND THEIR DERIVATIVES
Throughout this chapter, discussion will be developed on triisopropylacetic, di-t-butylacetic acids and their derivatives as the representatives of the most highly sterically hindered tri- and dialkylacetic acids known to date. Reactions of less hindered acids and their derivatives will be discussed only as needed,
VI-A. Ionization Constants of Acids
Results. Ionization constants of twelve aliphatic acids were determined by potentiometric titration by using a glass electrode, calomel reference electrode and Beckman pH meter, model G, in 50 volume per cent raethanol-water at 4-0°.^
(1) See G. S. Hammond and D. H. Hogle, J. Am. Chem. Soc^, 77, 338 (1955).
The acids titrated are listed in Table V together with their physical constants.
The ionization constants were calculated by the 2 Henderson equation (l) using 1/4, 1/2 and 3/4
pH = PK + - 0 ^ (1)
(2) See S. Glas stone, "Text Book of Physical Chemis try," D. Van Nostrand Co., Inc., New York, N.Y., 1940, p .982,
-4 1 “ TABLE V
Physical Constants of Acids
Neutral Eouivalent
R- m .p. °C. Galcd. Found
HH ee B.p. 118° 6 0 .1 60.7
38.8-39.5° 144.2 1 4 5 .0 G2B5 G2B5 t f- Ed B.p. 1 03. 0- 103.5 ÎO I-G3H7 I-C3H7 I (6 mm.)® 144.2 146.2 i-C^E? i-G Hy CB3 50.2- 50 .8 ^ 156.2 156.8 i-G_Hy i—GgHy 55.5-56.2® G2H5 172.3 172.5 1—C 2^7 i*“G Y I-G^Hy 146.5-149.3 166.3 165.9 t—G^Hg B Eb B.p. 7 9 .0 (9 mm.)^ 116.2 116.9 t-GjHg CgEg E 78.0-78.5j 144.2 144.0 t— i-GjHy^ E B.p. 100.0-100.2 (4 mm.)1 156.2 156.9 t-GjHg t-G^Eg E 60 .5-61 .5* 172.3 170.8 t— G GE3 GE3 199.6-200.2* 144.2 144.6 neo-G^H^j t-G^Hg GEg° 130. 0- 1 3 0 . 5P 200.3 196.2 -43- TABLE V (continued)
^ Neutral equivalents were calculated from the end point of the potentiometric titration, ^ Acetic acid was purified by chromic oxide treatment followed by d is ti l l a t i o n . ° L, Tsai, T, Miwa and M, S, Newman, J . Am.
Chem. Soc., 22; 2530 (1957), reported m.p. 38.9-39.5°.
1 .4269 . ® J. G. Traynham and M. A. B attiste, jLx—Qrg. Chem., 22, 1551 (1957), reported b.p. 108°
(12 mm.), n^^ I . 42 6 4 . ^ m.p. 49.5-50.2° reported in r e ference c. ® m.p. 5 6 .4 - 5 7. 0° reported in reference c.
^ I . 4 IO8 , ^ M. Hommelen, Bull. Soc. Ghim. B ele., 42.
243 (1933 ), reported b.p. 183,1-183.8° (741 mm.), n^*^ 1 ,4 1 0 5 . ^ G. S. Hammond and D. H. Hogle, J. Am. Chem.
Go£j., J77, 338 (1955 ), reported m.p. 79.5-80.5°.
^ 1 .4 3 5 0 . ^ A. A. Sacks and J. G. Astoh, J. Am.
-Chgi!}, Soc.,, J41, 3902 (1951 ), reported b.p. 122.5-123.8° 2 0 (22 mm.), nj) 1.4343. ^ A. Arkell, Ph. D. Dissertation,
The Ohio State University, 1958, reported m.p. 71.0-72.8°.
^ m.p. 1 9 8 . 5- 19 9 . 0° reported in reference c. ^ The author wishes to thank Dr. W. J, Hickinbottom, Queen Mary
College, University of London, England. ^ F. C. Whitmore and K. G. Laughlin, J. Am. Chem. Soc.. 56. 1128 (1934 ), reported m.p. 130-130.5°. - 44- neutralization points. The values thus calculated were accurate to ±0.03 pK unit and are summarized in Table VI,
Discussion. The solvent, 50 volume per cent methanol- water, used in the measurements was chosen in order to get the best available compromise based upon the slight solubility of the acids in water and the non-ideal be havior of electrolytes in non-aqueous solvents. The use
of an electrode system standardized by aqueous buffers for the mixed solvent, however, requires an unknown
correction due to the difference in the junction potential.
It is anticipated, however, that the correction is small
and constant throughout the series.
The second factor which should be taken into considera
tion is that due to the activity coefficient. The common equilibrium expression involves the activity coefficient term as shown in equation 2, but the Henderson equation (l)
does not. The ionization constants calculated in this work are thus classical or apparent. The activ ity
coefficient term, however, may be neglected if one takes
account of the relative values of ionization constants
(^2/^ 1 » or pKi-pKg); because the ra tio of two activ ity
coefficient terms of two acids which dissociate to a
similar extent is probably very close to unity. TABLE VI
Ionization Constants of Acids, in 50 (vol.) % Methanol- •Water at 40°
KGH3GOOH pK K X 10? %1 R2 ^3 Six No. Glass Kacid
H H H 0 5.69 2040 1
t —C^Hg HH 9 3 6 .2 4 575 3.5 I i —G^Hy i —GgHy H 12 5 6.48 331 6.2 VJl I t —C^Hg H 12 5 6 .5 0 316 6.45
CzHg 224 °2^5 C2H5 9 5 6.6 5 9.1 t-C^Hg i-C^Hy H 15 6 6.76 174 12
t-C^Hg GH3 GH3 9 5 6.95 112 18
i —G i —C^Hy GH3 12 6 6.97 107 19
t —G^Hg H 18 7 7 .0 4 91.2 22
i-GjHy i-G,Ey 15 7 58.8 C2%5 7 .2 3 35 ns o—G t-G^Hg 0H3 12 6 7.31 49.0 42
i —G^Hy i-G^Hy 18 8 7 .3 6 43.7 47 /O hÎ - ^ î = 1 (3) 3 HA 2 / ? HA 2
The values in Table VI, therefore, represent a reasonable approximation to the re la tiv e thermodynamic dissociation constants.
For comparison, data previously reported for the ionization constants of various acids 1 3 in a similar
(3) (a) W, L. Wright and H. T. Briscoe, J . Phys.
Ghem. . 37, 787 (1933)j see also (b) G. S. Hammond, "Ster- ic E ffects in Organic Chemistry," John Wiley and Sons,
Inc., New York, N. Y., 1956, p. 425 f f . medium are summarized in Table VII, While the values for
some acids do not check well with the present ones the
difference may be due to a medium effect as the earlier workers did not specify the way in which their solvents were purified and sample solutions were prepared.
It is however evident that the ionization constants
of highly hindered acids are significantly smaller than
those of unhindered acids, and that the observed variation
among unhindered acids^ in Table VII is much smaller than
(4) See also, J. F. J. Dippy, Ghem. Rev,, 25. 151
(1939); J. Ghem. Soc.. 1222 (1938), -hfl-'
TABLE VII
Ionization Constants of Acids, R^R^R^C'COOH
in 50 (vol.) % Methanol-Water
^2 Six Glass pKa® “ 3 Number
Benzoic -- 5.62%
Benzoic —- 5.15°
HHH 0 - 5 .5 6 n-alkyl H H 0 or 3 1 5.84-5.91° i-C^Hy HH 6 2 5.89° neo-C^H^^ GH3 H 3 2 6.05
CH3 GHg H 0 - 6.23°
GH_ H 9 4 6 .2 5 H G2H5 12 5 6 .3 1 1—G i-G_Hy H 12 5 6 .4 0
C2 H5 G2H5 C2%5 9 5 6.44 ne 0—G3H22 GH3 GH3 3 3 6 .5 0 ne 0—G^H^^ neo-G^H^^ H 6 3 6 .5 6 t —G^H^ GH3 GH3 9 5 6.72 neo-G^H^^ t-G^Hg GH3 12 6 6 .9 6
® G. s . Hammond and D. H. Hogle, J. Am. Ghem. Soc.. jujuy \-^ y J ^ ) • A’l t J a o u x t ; m e 11 Ü w a o u a i r j . e u oux# unless otherwise stated. ^ at 2 3 0 . <= Wright and H. T. Briscoe, J. Phys. Ghem.. 3?. 7Ô7 (1933). Measure ment was carried out at 25°. — that among the hindered acids. The anomalous change in the ionization constants of the highly hindered acids thus must be attributed to steric effects.
The ionization constant of an acid is directly
related to the free energy change (AF°) accurring in the
equilibrium ,^ Although the free energy change from acid
(5) See L. P. Hammett, "Physical Organic Chemistry,"
McGraw-Hill Book Company, Inc., New York, N. Y ., 1940,
pp. 69-37.
A F° = -RT In K (4)
molecule to carboxylate ion has been explained in differ
ent ways,& it seems most likely to attribute the present
(6) For example, see J. F. J. Dippy, D. P. Evans,
J. J. Gordon, R. H. Lewis and H. B. Watson, J. Chem. Soc.,
1421 (l937)j H. C. Brown and A. Cahn, J. Am. Chem. Soc. .
1 2 , 2939 (1950).
7 results to steric hindrance to solvation of the
(7 ) H. L. Goering, T. Rubin and M. S. Newman, J .
Am. Chem. Soc.. 76, 787 (1954)J 0. S, Hammond, "Steric
E ffects in Organic Chemistry," John Wiley and Sons, Inc.,
New York, N. Y., 1956, p. 426 ff. —49— carboxylate ions,®
(8) The solvation is more pronounced in the carboxy late ion than in the acid molecule.
Steric effects may be operative to the solvation in the following two ways,^ First, the alkyl groups lower the effective dielectric constant in the vicinity of the negative carboxylate group, thence the classical electro static free energy of the carboxylate ions is increased and less solvation results. Secondly, the alkyl groups in the carboxylate ion exclude the solvent molecules g from the vicinity of the negative carboxylate group ,
(9 ) Molecular model of the hindered acids used here shows, however, that the oxygen atoms of the anion, the center of the negative charge, are not completely shielded from the solvent.
thence stabilization of the carboxylate ion by special
interactions, such as hydrogen bonding, with solvent mole
cules is diminished.
These two factors are not distinguishable by use of
the present data. However, there might be a relationship
between the ionization constants of these hindered acids
and the number of solvent molecules solvated to the
carboxylate anions. -50-
The Rule of is useful^^ for estimating steric
(10) M. S. Newman, J . Am. Ghem. Soc. . 72, 4-7^3
(1950).
(11) See M. S. Newman, "Steric Effects in Organic
Chemistry," John Wiley and Sons, Inc., New York, N. Y.,
1956 , p. 206 ff.
(1 2 ) K. L. Loaning, à. B. G arrett and M. S. Newman,
J . Am. Chem. Bpc. . 74.. 3929 (1952); L. Tsai, T. Miwa and
M. S. Newman, i b i d . , 13, 2530 (1957). hindrance to carbonyl addition type reactions. It was not originally proposed to explain variations in the
ionization constants of acids, however a fair relationship
is obtained if the six number as well as the branching on
the a-carbon atom of acids are considered together. The
ionization constant decreases in the order shown in
Table VIII. For example, class 2 includes monoalkylacetic
acids having a six number of 6 and dialkylacetic acids
having a six number of 3 , and class 3 includes mono
alkylacetic acids having a six number of 9 , dialkylacetic
acids having a six number of 6 and trialkylacetic acids
having a six number of 3 .
The classification, as shown in Table VIII and also
in the fifth column in Table VI and VII, is well correlated
to the ionization constants of the acids except those -5 1 -
TABLE V III
Ionization Constants® and Structure of Acids
Branching on a-Carbon Atom Primary Sec ondary Tert iary Class Six Six Six Wo. pKa No. pKa No. pKa
1 3 5.84-5.91%
2 6 5.89% 3 6.05®
3 9 6 .2 4 6 - 3 -
4 9 6.25° 6 -
5 12 6.48,6.50 9 6 . 65 , 6.95
6 15 6.76 12 6.97
7 18 7 .0 4 15 7 .2 3 8 18 7.36
® Values were determined by the present work unless otherwise stated. See Table VI. b w, l . Wright and H. T. Briscoe, J. Phys. Ghem,. 17, 787 (1933). Measure ments were carried out at 25° in 50% methanol-water. ^ 0. S. Hammond and D. H. Hogle, J. Am. Chem. Soc., 77, 3384 (1955 ). Measurements were carried out at 40° in 50% methanol-water. — 52— bearing a neopentyl group on the a-carbon. Thus, the acids of class 5, 6 and 7 are approximately 7.5, l6 and
30 fold weaker than acetic acid, respectively, Triiso- propylacetic acid, class 8, is the weakest aliphatic acid ever prepared and 47 fold weaker than acetic acid.
The classification is merely a tentative tool to estimate the relative a cid ity of hindered a lip h a tic acids,
Although no theoretical explanation could be made for this classification, it may imply that the solvation fac tor is influenced both by the number of atoms in six position and branching on the a-carbon atom of acids,
IV-B Reaction of Acids with Ethoxyacetylene
Introduction. A study of ethoxyacetylene^-^* has
(1 3 ) J. F. Arens and P. Modderman, Proc. Noninkl.
Ned. Akad.. Wetenschap, 5J2., 1163 (1950) [ O.A. , 45. 6152
(1951)] .
(1 4 ) G. Eglinton, E. R. H. Jones, B. L. Shaw and
M. C. Whiting, J. Chem.Soc.. i860 (1954).
revealed that the reaction with carboxylic acids in
ether yields anhydrides corresponding to the starting
acids and ethyl acetate.
Results. By refluxing a mixture of diisopropyl-
acetic acid and ethoxyacetylene in ether for one hour
diisopropylacetic anhydride was obtained in 39^ yield. -53- On running a similar reaction with triisopropylacetic acid only a trace of the corresponding anhydride was obtained after 18 hours, but a 61^ yield after 5 days.
Discus sion. Although triisopropylacetic acid was much less reactive than diisopropylacetic acid, it is still noteworthy that, in spite of the fact that inter- molecular carbonyl addition type reactions of highly
hindered acids are Inhibited,the triisopropylacetic
(15) See M. S. Newman, "Steric Effects in Organic
Chemistry," John Wiley and Sons, In c., New York, N. Y.,
(1956), p. 203 ff.
anhydride formation still occurred. The mechaniam^^
proposed for the anhydride formation reaction, however, / provides a reasonable explanation for this successful
r e s u l t .
The proposed mechanism, as shown below, involves
addition of carboxylate and hydrogen ions across the
triple bond of ethoxyacetylene in the first step. It is
expected that this type of reaction would proceed even
with hindered a c i d s . T h e addition of the second
(16) E. Feith, Ber., 25. 503 (1892), reported that
silver salt of 2,4,6-trimethylbenaolc acid reacts with
methyl iodide to give the methyl ester of the acid. - 54- (17) M. S. Newman and W. S. Fones, J . Am. Chem.
Soc.. 62, 1046 (1947 ), reported good yields of butyl esters by the reaction of sodium s a lt of 2 ,4 ,6 -trialk y l- benzoic acids and tri-n-butylacetic acid with n-butyl chlorosulfite.
0 OCgHg ti R.COOH + HCSC-OGgHj (5 ) GHc
0 /O-O2H5 /OGgHg R G + RCOOH 0^ ( 6 ) C - o' CHo R^ II
I? s R - G + GHjCOOGgH^ (7) /
0=0
acid molecule to the monoadduct (I) proceeds in a sim ilar
manner.
The final step, equation 7, is the decomposition of
the diaduct through a cyclic transition state analogous 18 to the decomposition of ethylidene diacetate into
(IÔ) See J. Hine, "Physical Organic Chemistry," Mc
Graw-Hill Book Company, Inc., New York, N. Y., 1956, p. 458, -55“ acetaldehyde and acetic anhydride. This step obviously involves the highly hindered carbonyl group of triiso propylacetic acid in the reaction center. It has been shown, however, that hindered carbonyl groups which are very slow to undergo intermolecular carbonyl addition type reactions may undergo reactions proceeding through intramolecular cyclic transition states^^*^^
(1 9 ) See R. H. Dewolfe and W. G. Young, Chem. R ev..
875 ( 1956 ). Highly hindered ketones only enolise
with the usual Grignard reagents but react with allyl
Grignard reagent to give normal products, carbinols.
Thus the following cyclic transition state was proposed.
C ^ ' M g-B r
GHz) ^ GHg
R, R' - isopropyl or t-butyl ; H = mesityl, R' = mesityl,
methyl or isobutyl,
(3D) F. Meyer, W„ Sch&fer and J. Rosenbach, &rch.
PhaXïïjL, 2,67, 571 (1929 ) [ C ...A., 2 A. 838 (1930) J, reported
that 2 ,/4--dime thyl-6-hydroxymethylbenzonitrile afforded 5,7-
dimethylphthalide on the hydrolysis with sodium hydroxide.
Since 2,4,6-trimethylbenzonitrile is known to be inactive
toward sodium hydroxide, i t is probable th at the reaction
proceeds through a cyclic transition state as shown below. — 56—
CE CHCH CE
CH CH
Thus, the success of this reaction with triiso propylacetic acid is reasonably explained, and also furnishes a good example showing the a b ility of highly hindered carbonyl compounds to undergo reactions proceed ing through cyclic transition states.
XV-.G. Reaction of Acids with T rif luoroacetic Anhydride
Introduction. It has been reported^^ that acetic
(2 1 ) E. J. Bourne, M. Stacey, J. C. Tatlow and R.
Worrall, J. Chem. Soc., 2006 (1954). anhydride is obtained by reaction of acetic acid with trifluoroacetic anhydride under certain conditions.
Results. When triisopropylacetic acid was treated with trifluoroacetic anhydride slow evolution of gas occurred even at 0°. The mixture separated into two layers. The gas was probably carbon monoxide since no precipitate was formed when passed through a barium hy droxide solution. It was also confirmed that the top layer, which was almost colcniess, contained olefin as it decolorized permanganate and bromine solutions. The -57- bottom layer became dark brown as the reaction proceeded»
Discussion. This result is in accordance with many ? 2 previous findings." Trisubstituted acetic acids, such
(2 2 ) See L. P. Hammett, "Physical Organic Chemistry,"
McGraw-Hill Book Company, Inc., New York, H. Y. , 194-0 p. 283; M. S. Newman, "Steric Effects in Organic
Chemistry," John Wiley and Sons, Inc., New York, N, Y.,
1 9 5 6 , p. 217. as triphenylacetic and trimethylacetic acids, are known to be decarbonylated in strongly acidic media according to the following equations. The tertiary carbonium ion
RjCGOOH + H+ ----^ R^CCOOHJ >R^C'CO* + HgO (8)
R_COO+ -----) RgC* + CO (9)
formed is then stabilized either by attacking other species
in the system or by yielding olefin, with or without re
arrangement, by loss of a proton. The olefin, or olefins,
formed may further polymerize.
It is also known^^ that when trisubstituted acetic
(2 3 ) K. C, Laughlin and F. G. Whitmore, J1 Am. Chem.
Soc. . 54. 4462 (1932 ); F. C. Whitmore and H. M, Crooks,
i b i d .. 60, 2078 (1938).
acids are heated at 155-160° with phosphorous pentoxide, -5 8 - decarbonylation takes place. The yield of carbon monoxide increases with increasing branching of the tertiary acid.
Thus, trimethylacetic and methyl-t-butylneopentylacetic acids yield 55 and 90^ of the theoretical amount of car
bon monoxide, respectively.
It is furthermore quite general that tertiary alkyl
oxocarbonium ions readily decarbonylate whenever they
. 24 are formed.
(2 4 ) See H. J. Boëseken, Rec. trav. chim.. 29, 100
(19 0 0 ) J E. Rothstein, et a l . , J. Ghem. Soc., 581 (1958)
and the preceding papers.
Acid anhydride formation from carboxylic acids by
means of trifluoroacetic anhydride has been reported to 21 proceed according to the following reaction scheme.
R.COOH + (GF^CO)qO -— > R.QO.O.GOOF^ + GF^GOOH (lO)
R.GO'O'GOFq, + RGOOH ^—> (RGCOgO + GF^GOOH (ll)
The intermediate mixed anhydride dissociates to a slight
extent into oxocarbonium and trifluoroacetate ions, the 2 c former being the principal acylating spedies to
(2 5 ) E. J. Bourne, J. E. B. Randles, M. Stacey,
J. G. Tatlow and J. M. Tedder, J. Am. Ghem.Soc., 76,
3206 (1954). -59- R'GO'O'GOCF^ ^ ^ R CO ■’■ + CF3COO (12) form the final product.
When triisopropylacetic acid is reacted with tri fluoroacetic anhydride, the intermediate mixed anhydride
(III) probably dissociates into triisopropylmethyl
BGH^^gCHl^G'GO'O'GO'CF.
I l l oxocarbonium and trifluoroacetate ions to a great extent because of the steric facilitation due to the three isopropyl groups. Decarbonylation of the former ion is similarly aided by a large release of steric compression
(B s tr a in ) . 26 On the other hand, formation of an acid
(26) See W. G. Dauben and K, S. P itzer, "Steric
Effects in Organic Chemistry," John Wiley and Sons, Inc.,
New York, N. Y., 1956, p. 3 f f ; E. L. E lie l, ib id . , p.
62 ff. anhydride by combining triisopropylm ethyl oxocarbonium ion and the acid accompanies an increase of coordination number of the oxocarbonium ion carbon atom and encounters a large F strain^^ in the transition state, consequently the rate of anhydride formation must be incomparably smaller than that of decarbonylation. This results in olefin formation and polymerization of the olefin in the strongly acidic medium. -6 0 -
IV-D. Reaction of Acids with Thionyl Chloride
Results. Triisopropylacetic acid reacts normally with thionyl chloride to give the corresponding acid chloride in over 90% yield, whereas di-t-butylacetic a cid yields a mixture consisting of the corresponding acid chloride and a small amount of impurity. Theimpurity, although it was not isolated from the mixture, was probably 27 di-t-butylketene since the infrared spectrum of the
(27) See Chapter IV-F, p.64-ff.
2 7 mixture exhibited an absorption band a t 4*8 u.
Discussion. Although the mechanism of acid chloride formation from acid with thionyl chloride is complicated 2 8 and has not been established yet, a mechanism involving
(28) For example, see ¥. Gerrard and A. M, Thrush,
J. Cham. Soc.. 2117 (1953).
2 9 chlorosulfite as an intermediate, analogous to that in
(2 9 ) G. K. Ingold, "Structure and Mechanism in
Organic Chemistry," Cornell University Press, Ithaca,
N. Y., 1953, p. 392.
30 the reaction of alcohols with thionyl chloride seems to - 6 l -
(3 0 ) See J. Hine, "Physical Organic Chemistry,"
McGraw-Hill Book Company, Inc., New York, N. I . , 1956 p. 115. explain best the present results. The intermediate chlorosulfite (IV) yields the corresponding acid chloride
8 Ç1 R'COOH + SOClg ----^ R - = 0 + HCl (13)
IV
9 G1 R - C^t=^ 8 = 0 > R - GOGl + SO2 (1 4 ) 0 ^ 31 through a cyclic transition state. The successful
(3 1 ) See Chapter IV-B, p . 55 f f . formation of acid chlorides, thus, is reasonably explained even in the case of highly hindered acids such as triiso propylacetic acid.
Carboxylic acids with an a-hydrogen atom, however, furnish another possible reaction path throuth another cyclic transition state to yield ketenes^^ as shown in
(3 2 ) Although thermal decomposition of acid chloride is another possible route to ketenes, it is not likely in the case of di-t-butylacetyl chloride which can be distilled at 90 ° under reduced pressure. —62— .9 ^S. / R J ( 0 > ^0=0=0 + HCl + 8Op (15)
Cl
In equation 15.
Exclusive formation of acid chlorides from unhindered
acids with an a-hydrogen atom may be attributed to the
inherent characteristic of the reaction intermediate (IV)
to follow the reaction 14. I t is, however, apparent from
inspection of the overall structural change in the two
competitive cyclic reactions that in acid chloride forma
tion, equation 14, there is no change in bond angle or
coordination number at the carbonyl carbon, whereas in
ketene formation, equation 15, the coordination numbers
of the carbonyl carbon and of the a-carbon are decreased
and the bond angles increased. Consequently the latter
reaction may be aided by relief of strain. The contri
bution to the overall reaction thus may become important
as steric effects in the acids increase and a noticeable
amount of ketene may be formed.
The above consideration may further suggest that acid
chloride formation from any acid bearing an a-hydrogen
atom with thionyl chloride proceeds through a ketene
followed by addition of hydrogen chloride to the ketene.
Most ketenes probably react so rapidly with hydrogen
chloride that isolation is not possible. This is —63— probably the reason that the ketene intermediate mechan ism has never been considered. On the other hand, di-t- butylketene may react slowly enough with hydrogen chloride
that isolation is possible. Since these two possible
mechanisms are not distinguishable by any available data
at present, more work must be done to establish mechanism.
It may be noteworthy that in the reaction of diiso
propylacetic acid no indication of the existence of the
corresponding ketene in the reaction product was obtained,
this is probably due to the fact that the steric factors
are much less in the derivatives of diisopropylacetic
acid than in those of di-t-butylacetic acid.
IV-E. E sté rifica tio n of kcid Chlorides
Triisopropylacetyl chloride reacts exothermically
with methanol to give methyl triisopropylacetate in
90% yield. This result as well as those of the estérifi
cation of di-t-butylacetyl and diisopropylacetyl chlorides 3 3 with sodium triethylmethoxide will be discussed in the
(33) See Chapter III-B , p. 37.
next section.
IV-F. Reaction of Acid Chlorides with Ammonia and Sodium
Amide (Di-t-butylketene and Its Reactions)
Results. Triisopropylacetyl chloride reacted with
liquid ammonia in the presence or the absence of sodium •“6 4 - amide to give triisopropylacetamide in good yield. How ever, reaction in ether saturated with ammonia and in concentrated ammonium hydroxide failed to produce the amide. The last reagent caused a vigorous reaction but the product was the corresponding acid. .
The reaction of triisopropylacetyl chloride with liquid ammonia was quite slow even in the presence of an equivalent amount of sodium amidej 47 and 93 ^ yields were obtained after 35 and 20 hours reflux, respectively.
Liquid ammonia alone reacted more slowly and after a
20 hour reaction period a considerable amount of the corresponding acid was recovered after working up with water, although a 68 % yield of the amide was obtained after 30 hours.
Di-t-butylacetyl chloride behaves somewhat different ly toward sodium amide in liquid ammonia. When i t was reacted for 24 hours with ammonia containing 1 .5 mole equivalent amount of sodium amide, the corresponding amide and di-t-butylketene were obtained in 29 and 50% yields, respectively.
Di-t-butylketene is yellow oil and can be kept at room temperature in a flask with a ground glass stopper for at least a week without any sign of decomposition.
It was identified by infrared and elemental analyses and by preparing its derivatives. A strong infrared ab sorption band at 4,80 u^^ and the absence of absorption —6$—
(3 4 ) W. R. Harp, J r. and R. S. Rasmussen, J. Chem.
Phys. , 15, 778 (1947), reported a strong absorption band at 4 .6 4 u and the absence of absorption in the 6 u region
for ketene.
in the region of 6 u suggest that the ketene exists as
(3 5 ) D. H. Whiffen and H. W. Thompson, J. Ghem.
Soc.. 1005 (1946 ), reported strong absorption bands in
5.6-6,0 u region but no absorption in the 4 . 5- 4 .8 u. region
for ketene dimer.
monomer^^ (b.p. 73.0° at 45 mm.). The ketene did not
(3 6 ) M. S. Newman and A. Arkell, J . Ovp.. Chem. . 24 ,
385 (1959 ), reported that di-t-butylketene was present in
less than 3% yield by thermal decomposition of a diazo-
ketone prepared from dipivaloyl followed by gas chromato
graphic separation of the reaction products. The reported
infrared absorption spectrum in which strong bands
appear at 4.83 u and 5.80 u, however, differs from th at
of di-t-butylketene obtained here particularly at 5.80 u
and in the region of over 9 u. I t is believed th at the
previous sample was impure.
37 react at all with liquid ammonia during 24 hours. -66-
(37) Addition of ammonium chloride did not affect
the reaction.
If sodium amide was present, however, amide formation
occurred but so slowly that even after 12 hours a small
amount of the ketene was still unreacted. The ketene did
not react with aniline in refluxing benzene during 2 hours
but did react to give di-t-butylacetanilide in the
presence of a few drops of concentrated sulfuric acid.
Discussion. .A study of steric factors in reactions O rt of acid chlorides has revealed that two or possibly
(38) See M. S. Newman, "Steric Effects in Organic
Chemistry," John Wiley and Sons, Inc., New York, W. Y.,
1956 , p. 225 ff. and references contained therein.
39 three different routes are operative in most reactions
(39 ) See R. W. Taft. Jr., ibid.. p. 630 f f .
of acid chlorides. These may be illu stra te d by methanoly-
sis as shown in equations 16, 17 and 18,
OH I R'GOGl + CH3OH > R-G-Gl ^ RGOOGH3 + HCl (16 ) OCH3 R.OOGl --- RGO+ + Cl- ^£Ël9L) RGOOGH3 + HCl (l7) -67-
? ? R*G 0C1 + CH3OH -----> CH_0 '-- G-.- 01 -----) R'COOCHj + HCl (l8 ) 0
The first mechanism involves carbonyl addition and is subject to hindering steric effects. Conceivable correlation between acids and acid chlorides indicates that th is type of reaction does not occur^^ with the highly hindered acid chlorides used in this work.
The second mechanism (17) involves ionization to give oxocarbonium ion followed by nucleophilic attack of methanol and known to be facilitated not only by steric effects- but a Iso by the ion-solvating power of the medium,These facts indicate that the first step is
(4 0 ) B, L, Archer and R. F, Hudson, J , Chem, Soc.,
3259 (l950)j D. k. Brown and R. F, Hudson, ib id . , 883
(1953). the rate controlling step. The higher reactivity of water and methanol than liquid ammonia toward triisopropylacetyl
chloride is probably the result of the higher solvating / n y p power ’ " of the former solvents as compared with the
(4 1 ) The following dielectric constants were reported;
8 1 ,8 for water at 18°, 3 5 .4 for methanol at 13°, and 22 for
ammonia at -33°. See T. Moeller, "Inorganic Chemistry,"
John Wiley and Sons, Inc., New York, N. Y ,, 1954, p. 341. (4-2) The solvation effect is thought to be largely centered on the chloride ion and very little on the oxocarbonium ion since there would be great hindering steric effect to the solvation of the latter, latter. The intermediate triisopropylmethyl oxocarbonium ion is thought to react with nucleophilic reagents, such as solvent molecule in these cases, before it is decar- 42 bonylated,
(4 2 ) See Chapter IV-C, p. 56,
The th ird mechanism (18) is an Sn 2 type reaction
in which chloride ion is removed from the molecule by the
attack of nucleophiles at the carbonyl carbon. This mechanism is probably not affected by hindering ste ric
effects as seriously as the carbonyl addition mechanism
(16), and may be facilitated if stronger nucleophilic
reagents are used. The observed acceleration of the amide
formation from triisopropylacetyl chloride with sodium
amide is proabably due to the contribution of this mechan
ism, and due to the far greater nucleophilic character A3 of amide ion^"^ as compared with the ammonia molecule,
(4 3 ) In th is mechanism the relativ e steric req u ire
ment of the group replaced and the one added is probably
also important. In this case, however, the both groups
probably have a similar size, thence this factor may be sm all, •“é>9“
The result of the reaction of di-t-butylacetyl chloride with sodium amide in liquid ammonia brings up a new possible mechanism for reactions of acid chlorides having an a-hydrogen atom with bases. The observed formation of di-t-butylketene is undoubtedly due to the abstraction of the a-hydrogen by means of the amide ion ■44
(44) The failure of the amide ion to abstract the
a-hydrogen of triethylmethyl di-t-butylacetate (see
Chapter III-B, p. 40) thus must be partly attributed to
the considerably larger steric requirement of triethyl-
methoxy group as compared with chlorine.
followed by ^-elimination of the chloride ion. The ketene
then reacts with amide ion (not with ammonia molecule in
this particular case) to yield the corresponding amide
probably via the enol form.
H I ,0 R - C - C + NHp ----^ 0 = 0 = 0 + NHq + 01 (19) I ^01 R'' ^ R
R. R. 0=0=0+ MHg 0 - o! R' NR: or
R\ OH.0ONE, ( 2 )
+NE2 o r NH3 -70-
The ketene mechanism is probably a more favorable course for highly hindered disubstituted acetyl chlorides and may be aided by much larger release of hindrance^^ in the
(4.5 ) See Chapter IV-D, p. 62, molecule than the other mechanisms involving carbonyl
addition, ionization or substitution step.
The results of estérification^^ of di-t-butylacetyl
(4 6 ) See Chapter III-B, p. 33 ff.
and diisopropylacetyl chlorides with triethylmethoxide
also suggest a ketene intermediate mechanism. When d i- t-
butylacetyl chloride was reacted with an equivalent
amount of sodium triethylmethoxide in dry ether for 2 hours
the expected ester and the corresponding acid were ob
tained. in 30 and 62^ yields after treating with aqueous
base,47 respectively. This might imply that d i-t-b u ty l-
(4 7 ) Treatment of the reaction product with water
was not accompanied by heat evolution but aqueous base
caused a violent reaction with evolution of heat.
ketene had been present in the reaction product and that
the reaction had proceeded at least partly through the
intermediate ketene, although isolation was not attempted -71- in this case,
A similar reaction with diisopropylacetyl chloride afforded under certain conditions a mixture consisting of the triethylmethyl ester and a noticeable amount of impurity. The infrared spectrum of the mixture exhibited an extra band at 5.5 u^^ besides those observed in the
(4 8 ) Although the startin g acid chloride has an absorption band at 5.5 u, the reaction product does not have an absorption band at 13.2 u characteristic of carbon-chlorine bond, pure ester. Although this substance was not isolated, the extra absorption band at 5.5 u^^ might be due to d iiso - propylketene dimer.
Alternative methods for preparing diisopropylketene^^
(4 9 ) H. Staudinger, Helv. Chim. Acta, $, 306 (1925),
reported the isolation of this ketene but gave no
property of i t .
or its dimer were attempted but without success : Diiso-
propylacetyl chloride was treated with triethylamine or
M-me thyl piperidine in benzene at room temperature for cer
tain periods. In both cases, the corresponding amine
hydrochloride was formed but distillation afforded only
the recovered acid chloride in such yields as expected -72-
from the amounts of amine hydrochloride isolated. The
rest of the product was undistillable tar, probably poly
mer of the ketene. The same reaction in the presence of
triethylmethanol, however, afforded during 2^. hours 91%
of the theoretical amount of amine hydrochloride and a
60% yield of the corresponding crude ester. The crude
ester again showed a weak absorption band at 5,5 u.
All these results suggest the intervention of ketene
intermediates in reactions of hindered acid chlorides
having an a-hydrogen with bases. The fact that this ke
tene intermediate mechanism has not been considered in
reactions of acid chlorides is probably because no datum
has been available to support th is mechanism since the
ketene formed with most acid chlorides is so reactive and
it immediately reacts further to give final stable product.
In the case of highly hindered acid chloride such as di-t-
butylacetyl chloride-, however, the ketene is isolated from
the reaction mixture because it is unusually unreactive as
described above.
The inertness of di-t-butylketene towards nucleo-
philic reagents, such as aniline and ammonia, must be
attributed to considerable increase of steric hindrance
occurring in the reaction, since in any reaction of ketenes
the coordination numbers of the carbonyl carbon and the a- carbon are increased. “73“ 50-58 ^ Although several s table monomeric ketenes have
(50) See W. E. Hanford and J. C. Sauer, "Organic
Reaction," Vol. I l l , John Wiley and Sons, Inc., i'iew York,
N. Y., I 94 &, p. 108 ff; see also C. M. Hill, et al..
J. Am. Chem. Soc.. 71, I 66O (l95l); Z5, IO 84 (1953).
(5 1 ) Aldoketenes are isolated only as dimer. For example, R. C. Fuson, L. J. Armstrong and W, J. Shenk, J r ., ib id . f 66. 964 (1944), reported that mesitylketene was isolated only as the dimeric form: See also J . G. Sauer, i b i d . . 6 5 , 2445 (1 947 ).
(5 2) II. Staudinger, H. Schneider, P. Schotz and
P. M. Strong, Helv. Chim. Acta, 6, 291 (1923), measured degree of polymerization, which was determined by cryo- scopic method, of various ketoketenes a fte r standing for various periods of time.
Ketene Tijng merization.
Dimethyl 6 hr. 70 Methyle thyl 24 hr. 52 Diethyl 20 days 28 Di-n-propyl 28 days 9 Methylallyl 24 hr. 69 D iallyl 5 days 75
Di-n-propyIketene, yellow oil boiling at 30° at 11 mm., dimerized after heating at 100° for 10 days to give the white crystalline dimer, m.p. 61-62°. Aromatic ketoketenes dimerized less readily than aliphatic ketoketenes. -74-
(53) J. G. Sauer, U. S. Patent, 2,2^8,169 |'c.4...
36, 2737 (1942 )], reported dlheptylketene, b.p, 133-5° at 5 mm., and ethyldodecylketene, b.p. 143-5° at 7 mm.
They were probably monomeric although no other description was given.
(5 4 ) H. Staudinger, Helv. Chim. Acta. 306 (1925), prepared diisopropylketene from diisopropylmalonic an hydride inca.50^ yield, but no other description was given.
(55) R. G. Fuson, L. J. Armstrong, J. ¥. Kneisley and W. J. Schlenk, J r ., J. Am. Ghem. S o c.. 66, I 464
(1944 ), reported that phenylraesitylketene showed some tendency to polymerize at room temperature but the mono mer could be regenerated upon distillation.
(56) H. Staudinger, Ber. . 39, 3062 (1906), isolated golden yellow crystalline, monomeric biphenyleneketene, m.p. 90 - 9 0 . 5°.
(5 7 ) E. Shilow and S. Burmistrow, ibid., 68, 582
(1935 ), isolated yellow crystalline monomeric dibiphenyl- ketene, m.p. 197°.
(58 ) R. G. Fuson, L. J. Armstrong, D. H. Ghadwick,
J. W. Kneisley, S. P. Rowland, W. J . Shlenk, J r ., and Q.
F. Soper, J._Am. Ghem. Soc., 62, 386 (1945), isolated yellow crystalline monomeric dimesitylketene, m.p. 126-
127°. -7 5 - "been isolated they seem to be more reactive to nucleo- philic reagents than di-t-bntyIketene. For example, most of these ketenes were reported to react very readily 52 55—57 with aniline to give the corresponding anilide, * Dimesitylket e n e , although the reaction with aniline was not carried out, reacted with ethanol after refluxing for
8 hours to give ethyl dimesitylacetate.
Thus, di-t-butylketene probably represents the most
hindered and the least reactive monomeric ketene ever prepared, although dimesitylketene may be comparably
unreactive.
The resu lts in this work, however, showed that the
inertness of di-t-butylketene could be overcome by using
either very strong nucleophiles such as amide ion or a
protonic acid catalyst such as concentrated sulfuric acid.
IV-G. Reaction of Esters and Amides with Lithium
Aluminum Hydride
Results. Esters. Methyl triisopropylacetate, on
treatment with lithium aluminum hydride (LiAlflA) in
refluxing ether, yielded 2,2,2-triisopropylethanol in 6/^%
yield.
Amides. Treatment of triisopropylaoetamide with
LiAlH^, in refluxing ether or tetrahydrofuran (THF) under
certain conditions yielded triisopropylacetonitrile in
stead of the expected amine. When one mole of the -7 6 - 59 amide was treated with one-half mole of LiAlH^ in
(5 9 ) Reduction of R-unsuhstituted amides to amines requires one mole of LiAlH^ per mole of the amides as discussed later, see Figure 4, P* 85. refluxing THF for 12 hours, 25% of the starting amide was recovered and triisopropylacetonitrile was isolated in
55% yield. A similar reaction was observed in refluxing ether, and a 66^ yield of the nitrile was obtained to gether with a very small amount of the expected amino.
Since i t was obvious that LiAlH^ behaved as a basic dehydrating reagent under these conditions, the following preliminary studies were made in order to obtain further information about LiAlH^ reduction of unsubstituted amides. Benzamide and 2 ,3 ,5 ,6-tetramethylbenzamide were chosen as the model compounds of unhindered and hindered amides, respectively. Standardized LiAlH^ solution was used as a reagent so that complications^^ due to the
(6 0 ) Since solid LiAlH^ might be contaminated with salts derived by the action of the water formed by dehy dration of amide, the concentration of active LiAlH^ in solution might not have been the same in each case, presence of solid LiAlH^ could be avoided. The progress of reaction was followed in general by taking infrared -7 7 “ spectra of samples withdrawn, after known periods of time.
Both benzamide and the tetramethylbenzaraide evolved two moles of hydrogen gas if more than one-half mole of
LiAlH^ were employed. When the tetramothylbenzamide was
treated with one-half mole of LiAlH^ in refluxing THF,
the characteristic nitrile absorption band at 4-. 55
(61) L. Tsai, T. Miwa and M. S. Newman, J. Am.Chem.
13., 2530 (1957).
became more intense as the reaction proceeded. After a
40 hour reaction period 2,3^5,6-tetramethylbenzonitrile
was actually isolated in 20% yield together with the
starting amide (62%) and a very small amount of the corres
ponding amine. When more than one mole of LlAlH^ was
used, only traces of the nitrile were detectable and the
main product was 2,3,5,6-tetramethylbenzylamine.
The results were similar with benzamide. When less
than one-half mole of LIAIH^, was used the characteristic
nitrile band at 4.5 u^^ became intense as the reaction
(6 2 ) L. J. Bellamy, "The Infrared Spectra of ^omplex
Molecules," John Wiley and Sons, Inc., New York, N. Y.,
1958, p. 210,
proceeded. In one experiment of 4 hours duration benzo-
nitrile was isolated in 45% yield together with the —V8— / *3 amide (.35%) and benzaldimine (A-%) • When more than one
(6 3 ) Isolated as benzoic acid after oxidation with aqueous permanganate solution. mole of LiAlH^ was used, however, only traces of n itr ile were found in the reaction product.
On the other hand, when benzonitrile was mixed with one mole of LIAIH^ in THF at 0-5° the ch aracteristic nitrile absorption band at 4,5 u instantaneously and com p le te ly disappeared and the ch aracteristic imine and amine absorption bands appeared in the 6 u region,
(6 4 ) See reference 6 2 , p, 268,
The resu lts of the above reactions are summarized in Table IX together with reductions of triisopropylaceto n itr ile and methyl triiso p ro p y lacetate. Thus, regardless of steric effects amides produce nitriles if insufficient amounts of LiAlH^ are used and if su ffic ie n t amounts of
LiAlH^ are used two moles of hydrogen a re evolved and traces of nitriles are found.
Discussion, Esters and N-disubstituted amides of carboxylic acids are known to be readily reduced with
consumption of one—half mole of LiAlH^ to give primary alcohols and tertiary amines, respectively. The reduc
tion reaction has been pictured^$ gg a carbonyl addition TABLE IX _ Lithium Aluminum Hydride Reduction
Molar Sol Rexn.c Rexn. Yield % (crude) Compound Ratio^ (mole) vent Temp. Time Nitrile Aldimine Amine à mi de hr - d Triisopropyl- 0.67 T ref 1. 48 55 25 acetamide 0.52 I d E ref 1. 120 66 — trace 26
Tetramethyl- 0.50 _ d T re fl. 40 20 — trace 6 2 ® benzamide 1.5 2.0 E room 70 trace 50 20®
0.38 1.2 T re fl. Benzamide 4 45 35 -Î3 1.4 2.0 T-E room 2 trace main trace traced ^
Benzonitrile^ 1.0 0 T 1 major^ minor 2.1 0 T 2 trace mainJ
Triisopropyl acetonitrile 1.3 0^ T re fl. 96 mm. —. 97 Methyl T riiso propylacetate 1.5 0^ E-T r e fl. 12 64^ (2,2,2-Triis opropylathanol) ® Molar ratios of the hydride^to substrate. ^ T=tetrahydrofuran and E=ether. ° refl.» reflux temperature; T,65°j E,35 and room=room temperature, Volume of hydrogen was not measured. ® Remaining reaction mixtures, a fte r several aliquots had been taken out, were worked up and the yields of components were calculated, f Isolated as benzoifi acid after oxidation with permanganate. ë No pure component was isolable. ^ No quantitative separa tion was attempted. ^ Isolated as the 2,4-âinitrophenylhydrazone of benzaldehyde. J Isolated as the N-p-nitrobenzoyl derivative of benzylamine. ^ Pure yield. -8 0 -
(65) See N, G. Gaylord, "Reaction with Complex Metal
Hydride," Xnterscience Publisher, Inc., New York, N. Y.,
1956, p. 5U ff. type reaction which may be illustrated with disubstituted amides as shown in equation 21.
0 OM g R-G-NR:+MH ----) R-C-NR^ R-C-NR^+MgO (21) H H
The fact that the derivatives of highly hindered acids, such as methyl triisopropylacetate and N,N-dimethyl
4-benzyloxy-2,6~dimethylbenzamide (VI),are reduced to
(66) D. Lednicer and G. R. Hauser, J. Ore. Ghem.,
2j, 2008 (1958).
0^ NCoajjj c
OCHgCaHg
VI give the corresponding alcohol and amino in good yield
indicates that moderate steric effects do not prevent —8 l — the reaction and that the reduction proceeds by hydride ion attack on the carbonyl group,
K-Disubstituted amides bearing bulky groups on nitrogen, however, have been reported to produce aldehydes
and alcohols.T hese reactions have been explained by
(67) F. Weygand and D, Tietjen, Ghem. B e r,. 8 4 ,
625 (1 9 5 1 ); S. Chiavarelll, F. F. Rogers and G. B. Marini-
Bettolo, Gazz. chim. i t a l . . 83. 347 (1953).
(68) V. M. Micovic and M. L. Mihailovic, J. Ore.
Chem.. IS, 1190 (1953). the steric effects leading to the carbon-nitrogen bond cleavage of the intermediate V by the attack of the second hydride ion on the carbonyl carbon instead of the carbon- oxygen bond cleavage which normally occurs.
On the other hand, mono- and unsubstituted acid 6 5 amides require an additional one-quarter mole of Li.AlH^ for each hydrogen atom present to yield the corresponding amines and the reaction proceeds very slowly as compared with disubstituted amides and esters. For example, 69 pivalamide is recovered after treating with LiAlH^ in
(6 9 ) D. Y. Curtin and S. M. Gerber, J. Am. Ghem.
Sjogjt, lA , 4052 (1952 ). Detailed reaction conditions were not given. —82 — ether, and N-a~phenylethyl pivalamide (VIl) is reduced
(70) D. J. Gram and F. A. Abd Elhafez, ,J. Anij. Ghem«-
Sqc.., 24, 5851 (1952).
GH3 (GH2)2G.G0NH.CH'G6Hg
VII to the correç)onding amine by refluxing with excess LiAlH^
in ether for 4 hours. The slowness of the reduction of 65 mono- and unsubstituted amides has been attrib u ted to
the action of LiAlH^ on the hydrogen atom attached to
nitrogen to form a slow-reacting complex.
Sluggishness has also been reported on the reduction 71 of acylhydrazines containing the group -GONH- and
(71) R. L. Hinman, J. Am.Ghem.. .Sqc.^ , Z&, 2463 (1956)
72 explained by intervention of an enolization step as
(7 2 ) Reference 71 reported that replacement of
hydrogen was demonstrated by the evolution of hydrogen gas,
but no quantitative data were given.
shown in equation 22. The intervention of this step has
O H OM It I I -C - N- + MH ----) -C = N- + Hg (2 2 ) VIII —83— been assumed from the fa c t th a t the rapid and usually
quantitative'^^ replacement of an active hydrogen by
(7 3 ) Compounds exhibiting keto-enol tautoraerism are
only p a rtia lly enolized with Li/ilH^. For example, ethyl
acetoacetate and diethylmalonate are 50-70/O enolized
according to their reaction with LiAlH^ L^ee J. A. Krynit-
sky, J. E. Johnson and H, W. Carhart, J. Am. Ghem. Soc.,
70, 486 (1948 ); F. A. Hochstein, ibid.. %1, 305 (1949);
E. HBfling, H. Lieb and W. Schbniger, Monatsh.. 83. 60
(1952 ) J[, although they are 100# enolized according to
the Grignard reaction [T. Zerewitinoff, Ber.. 41. 2233
(1908 ); W. Fuchs, li. H. Ishler and A. G. Sandhoff, Ind.
Eng. Chem.. Anal. E d., 12, 507 (l940)J.
reaction with LiAlH^ is accompanied by the evolution of
hydrogen gas and the formation of a complex between the
substrate and some form of the reducing agent. Thus the
slowness of the reaction was attributed to repulsion of
hydride ion by the enolate (VIIl) to be reduced and to
the lower re activ ity of the carbon-nitrogen double bond
as compared with the carbon-oxygen double bond,^^
( 74 ) Garboxylate ions are reduced faster than
enolates of monosubstituted amides. See references 65
and 71. —84.“ On the basis of the results of the present work and the above discussion we propose a mechanism, as shown in
Figure 4 , for reduction of unsubstituted amides with
L I A I H 4 .
The first and second steps represent enolization of unsubstituted amides w ith consumption of one-quarter mole 75 of LiAlH^ and evolution of one mole of hydrogen gas in
(7 5 ) Any differences in the reactivity of the four active hydrogen atoms in LiAlH^ molecule is disregarded. each step to give doubly charged (dimetalated) enolates
(XI) of amides. The intervention of these steps accounts for the evolution of two moles of hydrogen gas^^ and the
(76) Compare with the footnote T3. The complete enolization implies that amide carbonyl is less reactive to LiAlH^ than ketonic carbonyl group. consumption of one mole of LiAlH^ per mole of amide in the overall reaction.
The fact that the doubly charged enolate (Xl) may regenerate the starting amide on hydrolysis is probably responsible for the reported poor yields^^ in the re duction of unsubstituted amides with in su fficien t amounts of LiAlH^.
Special attention should be drawn to the fact that - 8 5 - 0 OM " ,H ' R - G - N + MH ------> R-G=N-H + Hm (23) \ h
IX X
OM OM I I R - G = N - H + M H ---- ) R - G = N - M + Hg (24)
X XI
OM OM ’ ’ -Mo 0 MH , R-G=N-M + MH :— » R-C-N — R-G=N-M ---^ RGH,NM„ (25) ^ I ^ H H XI XII XIII XIV
OM I R — G = N — M - R — G SN + MgO (26)
XI XV
R_GzN + MH ------> RGH=NM RGHgNMg (27)
XV XIII XIV
M = LiAl/4
Fig, 4. - Lithium aluminum hydride reduction of unsubstituted amides. - 8 6 - the reduction of unsubstituted amides regardless of ste ric effects proceeds through the intermediate (XI), since both benzamide and 2 ,3 ,5 ,6-tetramethylbenzamide evolved two moles of hydrogen gas in the presence of sufficient amounts of LiàlH^, In other words th is fact indicates that the reduction does not proceed by carbonyl addition of hydride (MH) which occurs in the case of esters and disubstituted amides.
The doubly charged enolate (Xl) may be reduced to yield amines via aminoalcohol (XII) and imine (XII-l) as shown in equation 25. Since this step involves an attack of hydride ion to the doubly charged enolate (Xl) the slowness 77 of the reaction must be much more pro-
(77) Although the e le c tro sta tic repulsion is thought to be the main operating factor, the lower reactivity of carbon-nitrogen double bond as compared with carbon-oxygen double bond as proposed in reference 71 may be another factor. If this is true a species to give the addition porduct XII must not be the doubly charged enolate XI but the tautomeric keto form XVI, Since the contribution of
0 /M MH 9" yM R-G = N-M R-G-r R-G-N tlH H M XI XVI XII the keto form XVI to the equilibrium mixture is probably small, the rate of reaction leading to the intermediate —87—
XVI must be much smaller than that of reaction of the enolate XI, the concentration of which in the mixture is much larger than that of the former anion (XVI), nounced than the reduction of monosubstituted amides such as acylhydrazines discussed before.
Another possible route yielding amine from the intermediate XI is the reaction involving nitrile as the next intermediate. The doubly charged enolate (XI) may eliminate MgO to produce nitrile (XV), as shown in equation 26, This type of elimination reaction, dehydr- 7A—0 0 ation of unsubstituted amides, has been reported
(78) L, Tsai, T, Miwa and M, S. Newman, loc. c i t , . reported that 2,4,6-trimethyl-, 2,4»6-triisopropyl- and
2,3,5,6-tetramethylbenzamide yielded the corresponding nitriles in $6- 62% yield by treatment with sodium hydroxide
in ethylene glycol at 190-210°.
• (7 9 ) E, Wenkert and B, G. Jackson, J. àm. Ghem.
Soc., 8 0 . 211 (1958 ), reported that desoxypodocarpic acid
chloride (XVII, Y=COCl) yielded the corresponding amide
(XVII, I=C0NH2) and n itr ile (XVII, Y=GN) when treated with
sodium amide in liquid ammonia. •—88
(80) J. Leroide, Ann, chim.. [9] l 6 , 377 (l92l), reported that treatment of methyldl-n-propylaoetophenone with sodium amide in benzene yielded methyldi-n-propyl acetonitrile and a small amount of the corresponding amide. Since the Haller-Bauer reaction yields normally
trialkylacetamide, the nitrile formation is believed to
be due to the dehydration of the acetamide by means of
amide ion.
in the case of hindered amides and pictured as a base-
catalyzed elimination reaction of the enol form of the
amides as shown in equation 28, This transformation
B"------N t G - R -----> BH + N = 0 - R + OH" (28)
78 79 has been attributed * to a large release of steric
hindrance in the elimination step and also to the slow
ness of competitive carbonyl addition resulting in the
hydrolysis of amides. Thus, formation of nitrile from
the doubly charged enolate (Xl) is thought to be also 8 1 subject to steric assistance.
(8 l) Reaction temperature may also be important.
The nitrile thus formed may further be reduced to
imine and amine as shown in equation 27. This trans
formation is not affected much by steric effects as - 8 9 - illustrated by esters and N-disubstituted amides of hin- 82 83 dered carboxylic acids as well as by hindered nitriles. ’
(82) Trimethylacetonitrile^^ is reduced to give neopentylamine in 86% yield.
(83 ) 2,2-Diisopropyl-4—phenoxybutyronitrile is
reduced to give a cyclic compound as shown below, see re
ference 6 5 , p .733.
GH(CH-)„ ' ^ C&2— . C,H_*0*GH* CH_«G - GN --- > ‘ ‘ ^ ^ 2 2 , CHo GHp GH(GH^)p ^ H
(8 4 ) Triphenylacetonitrile is reported not to be
reduced at 25° in ether, see R. F. Wystrom and W. G.
Brown, J. Am. Ghem. Soc.. 70, 3738 (1948).
Thus, the doubly charged enolate (Xl) of both the
unhindered and hindered amides may be converted to
imines or amines by the addition mechanism (equation 25)
and/or the n itr ile intermediate mechanism (equations 26
and 2 7 ). The contribution of the la tte r mechanism is
supported by the fact that nitriles are isolable when
in su ffic ie n t amounts of LiAlH^ are used, that traces of
n itr ile s are found even when su ffic ie n t amounts of LiAlH^
are used and th at n itr ile s (for example benzonitrile) are
reduced much faster than the corresponding amides. —90—
In the case of hindered amides one might expect that the elimination of MgO from the intermediate XI would be sterically assisted and hence the preference of the n itr ile intermediate mechanism is expected. Should
steric assistance play a role one would expect that unhindered amides would be transformed into n itr ile s more
slowly than hindered ones. The slowness of most N-unsub-
stituted amides to be reduced is thus attributed to lack
of steric assistance for yielding nitriles and also
to the large electrostatic repulsion encountered in the
addition mechanism. However, the qualitativ e experiments
performed with hindered and unhindered amides have shown
a similar result, hence no conclusion can be made at
this point as to which mechanism re a lly operates.
The reduction of unsubstituted amides may be sum
marized as follows.
1. They react very rapidly with LiAlH^ to give
doubly charged enolates of the amides,
2. In the absence of sufficient Li&lH^ these
enolates eliminate MgO (M=LiAl/4) to form nitriles.
3 . In the presence of sufficient LiAlH^ the doubly
charged enolates are reduced to give imines and amines by
the n itr ile intermediate mechanism and/or the addition
mechanism. —91—
Judging from the above conclusion, the reduction of monosubstituted amides, which consume three-quarter mole of LiAlH^ per mole for the complete reaction, pro ceeds probably by formation of singly charged enolate
(XVIII) of the amides with evolution of one mole of hydrogen gas. The enolates (XVIII) may be reduced to amines through aminoalcohols (XIX) and the Schiff bases (XX),
OM DM • ' yR R-C=N-R R-G-N R-GH=N-R I ^M XVIII H XX
XIX V. GENERAL PROPERTIES OF HIGHLY HINDERED COMPOUNDS
AND SUGGESTIONS FOR FUTURE WORK
The highly sterically hindered compounds discussed in this work exhibit considerable resistance toward the reactions in which the coordination numbers of the un saturated carbon and/or the a-carbon atoms are increased.
This is best represented by;
1. The inertness of triisopropylacetonitrile and
N-isopropyl diisopropylacetamide to hydrolytic reactions, OH. I RjO'OzN + HgO - S — ^ RjG'CsNH (l) 0 OH n I EgGH'G-NHR + HgO — RgGH-G-NHR (2)
OH
K = (CH ) GH- 3 ^
2 . The inertness of di-t-butylketene to ammonia and aniline in the absence of catalyst,
R* R». NHR” ^G=G=0 + R«NH!2 — 3t—> ;G=G (3 ) R* R* ^OH
R* = (GH^)^G- , R"=H or C^H^
3. The failure of formation of diisopropyl-t- butylacetonitrile from the conjugate base of isopropyl- t-butylacetouitrile with isopropyl iodide,
-9 2 - -9 3 - R
R ^ - ' , , C-G2N + R-I — R-C-C5N (4) R / ’ R«
R = (GHjjgCH- , R« = (CH^)^C-
On the other hand, the reactions in which the coordination number of the carbon atom are decreased are greatly facilitated, Ag a consequence general re actions which occur with ordinary unhindered compounds are often dismissed and other types of reaction leading to less s te ric a lly demanding species take place. Those reactions are represented by;
4. Ketenimine formation by alkylation of conjugate base of the h itr ile s .
R^- R\ ^ R G—G=N ) G=G=N + R ’X -—^ G=G=N—Ri (5) R^ R^ R"^
5. Decarbonylation of triiso p ro p y lacetic acid by means of trifluoroacetic anhydride.
R 3C.COOH + (CF^G0)20 -- ^ R^C-CO-OCO-CF^ +
GF3GOOH (6 )
RjG-CO-OGOGFj ^ ^ R3G*G0+ + GF^COQ- (?)
R q G ' G O * -- ) R o G + + GO. (g) R = (GH^^gOH-
6 , Ketene formation from di-t-butylacetlc acid with thionyl chloride. - 94- 0 o' Il
R» > ^/ G _ = G = 0 + 8 On + HGl (9)
(I) R* = (GHj^jG-
7. Ketene formation from disubstituted acid chlorides by means of bases. H ' /O R\ R - G - G ^ + B -- 4 ^ G=0=0 + BHG l (lO) « ^G1 R^ R
R = (GH2)2GH- or (GB^)^G- j B => NHg or alkoxide.
8 . Nitrile formation from amides by means of lithium aluminum hydride. 0 CM II I R-G-ÏÏH^ + 2MB -- ^ R - G = B M + ZHg (ll) CM 1 r » G = N - M — > R - G 5 N + M gO (12)
M = L iAl/4
The reactions which involve the carbonyl carbon
atom as a reaction center and do not alter the coordina
tion number of either the carbonyl carbon or the a-carbon
may still occur if the reactions proceed through cyclic
transition states. For example;
9. Formation of triisopropylacetic anhydride from
the corresponding acid with ethoxyacetylene. - 9 5 - 0
/O Q It- R-G.GOOa + GpH.O-G=GH ---) " ^ C (13) \ - o / \: h .
? ^ R ^ C - C n - O , ACiH,,- Ix OCjHs:
F«C^
Suggestion for future work. Some of the above
results bring up new aspects of reaction mechanisms of
ordinary unhindered compounds,
10, Example 6 indicates that ordinary acids bearing
an a-hydrogen atom may react with thionyl chloride to
give ketenes, as shown in equation 9, which immediately
react with hydrochloric acid to give acid chlorides.
Since ordinary ketenes are highly reactive, their isola
tion would not be possible because of rapid reaction
with hydrogen chloride. The intervention of a ketene
intermediate would be shown by using a-deuteroacids,
This result may furnish information about the relative
importance of four- and six-merabered cyclic transition s t a t e s .
11, Example 7 suggests again a ketene intermediate
mechanism for estérification or amidation of ordinary -9 6 - acid chlorides bearing an a-hydrogen atom. Again deuterium experiments may provide information about the mechanism,
12, The nitrile formation from triisopropylacetamide with lithium aluminum hydride has lead to the proposition of the n itr ile intermediate mechanism for lithium alumi num hydride reduction of unsubstituted amides. It has been shown that this mechanism appears to be more lik e ly to take place than the alternative path which involves direct reduction of doubly charged enolate II, and that
OM OM I I M R-C=N-M R-C-N^- > R-CH=NM RGHgNMg (15)
H II the relative importance of these two paths may vary with steric effects in the enolates. In order to clarify the relative importance further studies must be undertaken.
However, in the case of both the hindered and unhindered amides, the nitrile formation is the main reaction when in su ffic ie n t amounts (optimum amount being one-half mole) of lithium aluminum hydride are employed,
13, Another point of theoretical interest lies in the ionization constants of highly hindered acids. The ionization constants measured in this work decrease with increasing steric hindrance. It will be of interest to study a correlation of the ionization constants with the number of molecules solvating to the carboxylate anions. - 97 -
14-. In re la tio n to the above problem, i t w ill be of interest to see how weak the more hindered acids, such as di-t-butylalkylacetic acids, will be, or whether the lowest limit of the ionization constant of aliphatic hindered acid has already been reached.
15. Turning to the practical point of view, the failure of alkylation of triethylmethyl di-t-butylacetate may be overcome by using the esters having less sterically demanding alkoxy groups. The fa ilu re of the abstraction of the a-hydrogen has been attributed to hindering steric effects of both acyl and alkoxy alkyl groups which inhibit the attack of amide ion on the a-hydrogen. This can be checked by means of isotopic (deuterium exchange) study.
Decrease of the steric requirements in the alkoxy portion of the ester probably will diminish the hindering steric effects around the a-hydrogen atom and enable the desired anion to be formed. Since the alkylated esters must be readily convertible to acids, the t-butyl and benzyl esters may be the alkyl groups of choice. The former may be hydrolytically removed and the la tte r by hydrogenolysis,
16. The failure of the separation of ketenimines and triisopropylacetaldimine may be overcome by gas chromatography. VI. EXPERIMENTAL
Conventions Used in the Discussion of Experimental Work
1. Fractionating Column;
Column Length Diameter Packing Heating Jacket (cm.) (cm.)
a A 90 1.8 glass helices e le c tric a l
B 30 1.6 glass helices® e le c tric a l
C 16 1.6 glass helices® e le c tric a l
E 90 2.9 plates^ e le c tric a l
E 30 0.6 unpacked water® no description Modified Claisen head
1/ 8 " single-turn. 31 perforated plates, bubble cap type. Hot water or steam,
2. Melting points of pure compounds are corrected.
3. Analyses were performed by Galbraith Microanalyti-
cal Laboratories, Knoxville, Tenn.
4 . The wavelength in microns is shown after the symbol
IR for all infrared absorption spectra. The absorption
in te n sity is given by the symbols (s)-strongj (m)-medium;
(w)-weakj and (sh)-shoulder.
5. Dry ether used was dried with calcium hydride
followed by the distillation over ethyl Grignard reagent.
6. The phrase "treated in the usual manner" means that
the organic layer, to which were added the ether-benzene
(ca. 3:2 by volume) extracts of the aqueous layer, was washed with water and saturated sodium chloride solution —98— -9 9 - and dried over or by passing through anhydrous magnesium su lfate. The bulk of the solvent was then removed by d is tilla tio n . For the separation of acidic components
of a reaction mixture 10^ sodium bicarbonate or 5%
sodium hydroxide solution was employed, and for basic components 2N hydrochloric acid was used,
7, The following trade-named materials were used:
“Skellysolve B" - a mixture of saturated hydro
carbons, b,p. 65- 69 °.
“Skellysolve G" - a high boiling hydrocarbon fraction,
b.p, 90 - 97 °,
“Skellysolve F” - a low boiling hydrocarbon fraction,
b,p. 35-55°.
"Diglyme” - diethylene glycol dimethylether,
VI-A. Synthesis of Highly S terically Hindered
Aliphatic Hitriles
1. Diisopropylacetonitrile. Ethyl diisopropyl-
cyanoacetate, b,p. 98-103° (5 mm,), was prepared from
ethyl cyanoacetate as described^ in 84-% yield,^
(1) F. C. B. Marshall, J. Chem. Soc.. 2754 (1930),
An approximately 60% yield is reported,
(2 ) For the introduction of the second isopropyl
group, the reaction mixture was heated at reflux for 2
hour and at about 80° overnight with continuous stirring. -10 0 -
A mixture of 334.8 g. (1.7 mole) of the ester, 100 ml, of ethanol and 1200 g. of 35/6 potassium hydroxide solution was refluxed for 26 hours until the mixture be came homogeneous.^ The alcohol was then removed under
(3) The time required for the hydrolysis appeared to be cut down by adding more alcohol to the reaction mixture. reduced pressure and, after diluting with water, the mixture was strongly acidified with concentrated hydro
chloric acid and worked up in the usual manner to give
282.5 g. i9B%) of solid residue, crude diisopropyl-
cyanoacetic acid.^
(4 ) F. C. B. Marshall, loc. cit.. reported a
V quantitative yield for this reaction in a similar manner and m.p. 97.5° for th is acid.
The crude acid was then heated at 190-200° with 2 g.
of copper powder (Copper Metal, Precip. Powder, J. T.
Baker Chemical Co., Phillipsburg, N. J.). Vigorous
evolution of carbon dioxide sometimes ensued and moderate
cooling was necessary. After the evolution of gas had
subsided, the mixture was distilled. Redistillation of
the crude product through column A afforded 181.3 g. - 101-
(85%) of diisopropylacetonitrile, b.p, I 68 -I 69
(reported^ b.p. 170°).
2. t-Butylacetonitrile. Ethyl is opropylidene- cyanoacetate, b.p. 81-86° (3 mm.), was prepared from ethyl cyanoacetate with acetone as described^ in 78%
(5 ) S. Wideqvist, Acta Chem. Scand.., _2., 303 (1949), reported a 75% yield and b.p. 107° at 10 mm. for this e s te r . yield,^ The isopropylidenecyanoacetate (3.9 mole) was
(6) Column B was used together with an azeotropic water separator for the reaction. A 20 hour reaction period was needed for completion of the reaction. converted to the t-butylcyanoacetate , b.p. 73-79° (3 mm. )
(reported^ b.p. 93.0-93.5° at 8 mm.), by means of methyl-
(7 ) S. Wideqvist, Arkiv. Kemi, 321 (1950), reported a 75% yield.
n magnesium iodide as described in 75% yield. The t-butyl- cyanoacetate was then saponified and decarboxylated^ to
(8) A. J. Birch, J. Chem. Soc.. 2721 (1949), reported an 80% yield. —102 — give t-butylacetonltrile, b.p. 135-137° (reported^ b.p,
137°) in 12% yield, after redistillation of the crude product through column C,
3. Alkylation of diisopropylacetonitrile. To a stirred solution (or suspension) of 0.17 mole of alkali amide^ in 300 ml. of liquid ammonia was added 20 g.
(9 ) Sodium and lithium amides were prepared accord ing to the method described by C. R. Hauser, F. W. Swamer and J . T. Adams, "Organic Reactions," Vol. VIII, John
Wiley and Sons, Inc., Hew York, N. Y,, 1954, p. 122;
Potassium amide was prepared according to R. S. Yost and
G. R. Hauser, J. âm. Chem. Soc., 69, 2325 (1947).
(0.16 mole) of d iiso p ro p y laceto n itrile. Within a few minutes the reaction mixture became orange. After stirring for one hour, 0.2 mole of alkyl halide was added over a 20 minute period and the resulting mixture was stirred for a certain period under reflux. The liquid ammonia was then evaporated on the steam bath and a suffi cient amount of water (ca. 50 ml.) was added to dissolve salts in the residue.
Hydrolytic separation. The organic layer was then washed with 2N hydrochloric acid and 10^ sodium th io su l- fate solution and worked up in the usual manner. The organic residue was cooled in a low temperature bath (-78°) - 1 0 3 - anà crystalline N-substituted amide was filtered and washed with cold Skellysolve F. The mother liquor and washings were combined and, after removal of the solvent, the residue was d is tille d through column C to give alkyl
ated n itr ile together with a small amount of low boiling
substance.
The resu lts are summarized in Table I I I ,
A11emoted senaration. o_f..N-i3 om;oj3ya....dli^.Q.P.r.Q.pyl=
ketenimine. The organic layer obtained from the reaction
of diisopropylacetonitrile and isopropyl iodide by the
procedure described above was worked up in the usual
manner. The roughly distilled organic liquid was sub
jected to fractional distillation through column C.
Approximately one-third of the total distillate, b.p.
46- 83 ° (6-7 mm.), was divided into five fractions, all
of which had nitrile and ketenimine characteristic absorp
tion bands at 4.5^® and respectively. The n itr ile
(10) L. J. Bellamy, "The Infrared Spectra of
Complex Molecules," John Wiley and Sons, Inc., New York,
N. Y,, 1958, p, 263.
(11) G. L, Stevens and J, C. French, J. Am. Chem.
àssu, 2Â, 4398 (1954).
in these fractions were probably mainly diisopropyl
acetonitrile, A fraction boiling at 83-84° (6 mm.) weighed
approximately two-thirds of the total distillate and was - 1 0 4 - practically pure triisopropylacetonitrlle. All fractions 12 were colorless liqu id s.
(12) C. L. Stevens and J. G. French, loc. c it.. reported that the aliphatic ketenimine, N-n-butyl ethyl- n-butylketenimine, was colorless, but aryl-alkyl- or arylketenimines were yellow.
4. Alkylation of t-bûtvlacetonitrile. To a stirred suspension of 1.0 mole of sodium amide in 600 ml. of liquid ammonia was added 58 g, (0.5 mole) of t-b u ty l- a c e to n itrile in 50 ml. of dry ether over a 30 minute period. The mixture became green. After stirring for one hour 190 g. ( l . l mole) of isopropyl iodide was added during 40 minutes. The resulting black colored mixture was stirred for 30 hours under reflux and worked up as described for the attempted separation of N-isopropyl diisopropylketeniraine,
The residual organic liquid was distilled through column G to give 82.6 g. of a mixture, b.p. 97-101°
(45 mm.), IR: 4.5^^(w) and 5.0^^(s), of isopropyl-t- 13 butylacetonitrile and N-isopropyl isopropyl-t-butyl- ketenim ineRepeated distillation under reduced
(13 ) L. Tsai, T, Miwa and M. S. Newman, J . Am.
Chem. Soc.j, 79y 2530 (1957), reported b.p. 178-181° - 1 0 5 - for the nitrile.
(14) M. S. Newman and T. Miwa, unpublished, reported b.p, 74° (14 mm.) for the ketenimine. pressure through column B did not give separation.
On attempted distillation of a part of the mixture at atmospheric pressure, a small amount of gas, probably olefin was continuously but very slowly evolved during the course of d is tilla tio n and decolorized permanganate solution. The distillate, b.p. 181-184°, was identical in the infrared spectrum with the starting mixture.
The distillate, 38.2 g., obtained by vacuum distilla tio n was dissolved in 20 ml. of Skellysolve F and treated with 20 ml. of water and 10 ml. of concentrated hydro
chloric acid with shaking in an ice-water bath. The addition of concentrated hydrochloric acid caused a highly
exothermic reaction and precipitated a large amount of white c ry stallin e substance which was filte re d and washed with a small amount of cold Skellysolve F to yield 26,0 g.
of N-isopropyl isopropyl-t-but#ketenimine, m.p. 149-150°
(reported ^ 150.0-150.6°), The mother liquor and washings
(15) M, S. Newman and T. Miwa, unpublished data,
were treated in the usual manner and the residual organic
liquid was distilled through column C to give 14.0 g. of —106" isopropyl-t-butylacetonitrile, b.p. 72,0-72,5° (15 mm,), ^ the infrared spectrum of which was identical with that of 13 the authentic specimen in all respects. Thus the yields of the ketenimine or the amide and the dialkylacetonitrile were calculated to be 56 and 45%, respectively, 12 The ketenimine-nitrile mixture, which was colorless, could be kept at room temperature in a reagent bottle with a ground glass stopper without any change in the infrared spectrum for at least 3 months,
VI-B. Reactions of Highly__^erically Hindered A liphatic N itriles
1. Hydrolysis. A mixture of nitrile and sulfuric acid of known concentration was heated at a specific temperature. After a certain period of heating the re sulting clear, slightly colored solution was cooled down to approximately 50° and a known amount of solid sodium n itr ite was added l i t t l e by l i t t l e so that formation of nitric oxide was minimized. The mixture was gently warmed on the steam bath until evolution of gas ceased. A large flask should be used since the evolution of gas is some times violent. After cooling and diluting with water, the mixture was worked up in the usual manner and the acids were purified either by distillation through column C or by recrystallization from Skellysolve F at -78° followed by sublimation in vacuum. -1 0 7 -
When amide was used as the starting substance in stead of nitrile, the same procedure was followed without the prior hydrolytic treatment.
The reaction conditions and re su lts are summarized in Table IV, and the physical constants of the acids are listed in Table V together with those of other hindered a cid s,
2. Lithium aluminum hydride reduction. T riiso- propylacetonitrile. To a suspension of 2,5 g. (0.066 mole) of lithium aluminum hydride in 300 ml. of purified
tetrahydrofuran^^ was added 8.4 g. (0.050 mole) of
(16) Commercial tetrahydrofuran kept over potassium
hydroxide pellets for months was filtered and distilled
from lithium aluminum hydride.
triiso p ro p y lac e to n itrile in 10 ml. of purified te tr a -
hydrofuran. The mixture was refluxed for 4 days under
nitrogen and decomposed, after cooling in an ice-water
bath, by successive additions of 3 ml. of water, 4 ml, of 17 10^ sodium hydroxide solution and 8 ml. of water. The
(17) V. M. Micovic and M. L. Mihalovic, J. Ore.
G hem. , 18, 1190 (1953), recommend n ml. of water, n ml.
of 15^ sodium hydroxide solution and 3 n ml. of water for
reduction of amides to amines by means of n g. of lithium - 1 0 8 - aluminum hydride. precipitate was filtered and washed with tetrahydrofuran several times, and the filtrate was dried over anhydrous magnesium sulfate. A fter removal of solvent the residue was distilled to give 8.4 g. (97#) of crude 2,2,2-triiso- propylethylamine, b.p. 87-90° (8 mm.), IR: 6.1 (m)^^,
(18) L. J. Bellamy, loc. c i t . . p. 255, which so lid ified to form a wax-like solid, m.p. 40- 45°.
The amine was too soluble in any organic solvent to be recrystallized and did not crystallize from Skellysolve
F at -78°.
The crude amine (2,4 g.) was dissolved in 50 ml, of dry benzene and refluxed with 5 g. of p-nitrobenzoyl chloride for 20 minutes. After working up in the usual manner (base and acid washing), the residue was recrystal lized from alcohol to give N—(2,2,2-triisopropylethyl)—p— nitrobenzamide, m.p. 124,5-125.5°,
Anal^ Calcd. for G, 67.5; H, 8,8; N, 8,7. Found; C, 67.3,67.5; H, 8.7, 8.9; W, 8,8, 8.8,
The crude amine was treated with 6ll hydrochloric acid and an exothermic reaction occurred to precipitate white crystalline hydrochloride. Recrystallization from hot water gave pure 2,2,2-triisopropylethylamine hydro chloride, m.p, 355° in a sealed tube. - 1 0 9 -
Anal. Calcd. for Cn^g^NCl: C, 6 3 . 6 j H, 12.6; N,
6.7; 01, 17.1. Found; 0, 63. 6 , 63.7; H, 12.9, 12.7;
N, 6.7, 6.7; 01, 17.0, 17.3.
The amine hydrochloride did not absorb moisture in
the air and was stable enough to be handled in the a ir.
When triisopropylacetonitrile was treated with
one-half or one-quarter mole of lithium aluminum hydride
in refluxing tetrahydrofuran^^ for 4O-6O hours, the
product obtained after working up in the usual manner
contained nitrile, imine and amine, IR: 4.5, 6.0-6.2.
Attempted separation by means of d is tilla tio n or chromato
graphy over alumina failed. On keeping at room tempera
ture, the mixture decomposed with evolution of ammonia.
By washing the mixture with 2N hydrochloric acid and
working up in the usual manner, a neutral portion, IR:
4.5 (w), 5.Ô (s), was separated. Treatment of this sub
stance with 2,4-dinitrophenylhydrazine gave the triiso-
propylacetaldehyde 2,4-dinitrophenylhydrazone, m.p. 152.0- 152.8°.
M ai. Calcd. for CiyHgaN^O^: 0, 58.3; H, 7,5;
If, 16.0. Found: 0, 58.4, 58.4; H, 7.7, 7.6; N, 15.9, 1 6 . 0 .
No quantitative estimation of the relative amount
of the product was possible because the basic portion
obtained by the above extraction method s t i l l evolved
ammonia at room temperature. This indicates that the -1 1 0 - imine is quite re s ista n t to hydrolysis by aqueous hydrochloric acid,
VI-0. Synthesis of Highly Sterically Hindered
Aliphatic Acids
1, Triiospornylacetlc acid. TriethyImethanol. To a stirred solution of ethylmagnesium brom ide,prepared
(19 ) The use of iodine as the in itia to r should be
avoided since the product, triethylcarbinol, is readily
dehydrated in the presence of iodine,
from 14-0 g, ( 1,3 mole) of ethyl bromide, 31 g. (l .3 g.
atom) of magnesium turnings in 500 ml. of dry ether, was
added as rapidly as possible 86 g, (l.O mole) of diethyl-
ketone in 130 ml. of dry ether. After s tirrin g for 2 hours
and standing overnight, the reaction mixture was cooled
and decomposed with 5ÏÏ sulfuric acid. The ethereal
solution was then separated as quickly as possible. After
working up in the usual manner, the organic liquid was
d is tille d through column C to give 102 g. (88%) of t r i -
ethyImethanol, b.p. 65-67° (45 mm,),^^ 1.4296
(20) Distillation at atmospheric pressure resulted
in a large quantity of olefin,
(reported^^ b.p. 53-54° at 20 mm., 1.4276). -1 1 1 -
(21) C. R. Hauser and J. Chambers, J. Am. Chem. Soc..
2E, 3837 (1956).
Diisopropvlacetyl chloride. A mixture of 52 g,
(0,36 mole) of dilsopropylacetic acid and 90 g, of thionyl chloride was allowed to stand for 2 hours and refluxed for one-half hour. After evaporating the excess thionyl chloride, the residue was distilled through column C to give 53 g. ( 90 #) of diisopropylacetyl chloride, b.p,
87-88° (48 mm.), np° 1.4388 (reported^^ b.p. 75-78° at
(2 2 ) A. A. Sacks and J. G. Aston, J . Am. Chem. Soc.,
72, 3902 (1951).
28 mm.).
TrlethV1 me_thv 1 diisopropvlacetate. To a stirred suspension of 0.52 mole of sodium amide in 500 ml. of liquid ammonia was added 64 g. (0.55 mole) of tr ie th y l- methanol in 50 ml. of dry ether. The mixture became dark gray. The liquid ammonia was evaporated on the steam bath while 4OO ml. of dry ether was added. The remaining
ammonia was removed by refluxing the ethereal solution for
3 hours, sufficient heat being applied during the last
half hour to distill approximately 100 ml. of ether
through the reflux condenser.
To the stirre d suspension of sodium triethylm ethoxide -1 1 2 - was added at room temperature over one-half hour period
81,4 g. ( 0,50 mole) of diiaopropylacetyl chloride in
50 ml, of dry ether. After refluxing for one hour the mixture was cooled and 200 ml, of water was added with stirring. After working up in the usual manner (base washing), the residual organic liquid was distilled through column C to give 106,6 g, {B0%) of triethylmethyl diisopropylacetate, b.p, 99-102° ( 3 mm.), IR; 5,5 (sh),
5.75 (s). Redistillation of the product through the same column furnished the analytical sample, b.p, 101.0°
(3 ram,), n^° 1,4442, IR: 5.75 (s).
Anal, Calcd, for C^^H^gOg: C, 74.3; H, 12,5.
Found: 0, 74.2, 74.3; H, 12,4, 12,2,
On running the same reaction where diisopropyl- acetyl chloride was added into an ice-cold suspension of sodium triethylmethoxide, a sudden violent reaction
occurred after adding approximately one-half of the acid
chloride and a large amount of solid was formed. After dissolving the solid by adding more ether, the rest of
the acid chloride solution was added and the resulting
cloudy mixture was refluxed for one hour. After working up in the same manner as in the previous run, 20,2 g,
(28^) of the free acid, b.p, 89-97° (4 mm,), was recovered
and 44.5 g. of a frac tio n , b.p, 109-116° (4 mm.), IR; 5,5
(s), 5.75 (s); was obtained. The high boiling fraction was probably a mixture of the expected ester and - 1 1 3 - diisopropylketene dimer, although the latter compound was not isolated.
TriethyImethyl trlisooropylac_etate_. To a stirred solution of potassium amide prepared from 16,5 g. (0.42 g. atom) of potassium and 900 ml. of liquid ammonia was added 97.0 g. ( 0,40 mole) of triethylm ethyl diisopropyl acetate in 200 ml. of dry ether. The resulting mixture became green after one-half hour and stirring was con tinued for another hour. To the stirred mixture was
then added 70 g. (O. 4I mole) of isopropyl iodide in 100 ml. of dry ether. The resulting mixture became bright
yellow and then turned to light gray within 5 minutes.
After s tirrin g for 2 hours, 200 ml. of wet ether was
carefully added and the liquid ammonia was allowed to
evaporate on the steam bath until the ether started to
boil. Water was then added to the cold mixture with
stirring. After working up the reaction mixture in the
usual manner, the residual organic liquid was distilled
through column G to give 52.7 g. ( 46#) of triethylmethyl
triisopropylacetate, b.p. 141- 146° (4 mm.), n^*^ 1.4673-
1.4689 and 41.0 g. of a mixture of the di- and t r i i s o
propylacetate, b.p. 107- 141° (4 mm.), n§^ 1,446-1.459.
The low boiling fraction was usually subjected to the
following hydrolysis together with the high boiling
fractio n .
Triisopropylacetic acid. A mixture of 21 g. —11 4 ”*
(0.074 mole) of triethylm ethyl triiso p ro p y lacetate, 15 ml. of concentrated hydrochloric acid and 25 ml, of dioxane was refluxed for 4 hours. The resulting mixture was dis tilled until the temperature rose to 100°. Crystals deposited on cooling were filtered to give 13.3 g. {97%) of white crystalline triisopropylacetic acid, m.p. 148-
149°. The purest sample was obtained by sublimation of the above sample in vacuum, m.p. 148.5-149.3°.
Anal. Calcd. for ^11^22^2' 70.9; H, 11.9. Found; C, 70.8, 70.9; H, 11.8, 11.9.
The use of the crude ester required a recrystalliza tion of the product from dioxane-water to get the pure sample, the yield of which was 70-90%.
2. Di-t-butvlacetic acid. Isopr.opvl t-butyl ketone.
To a sodium amide suspension prepared from 55.4 g. (2.4 g. atom) of sodium and 2 1. of liquid ammonia was added
241 g. ( 2,1 mole) of diisopropyl ketone (b.p. 112. 5-
123. 5°, 1.4008 )^^ in one 1. of dry ether. The liquid
(2 3 ) F . C. Whitmore and E. E. StaM y, J. &m. Chem.
S oc., 15, 4155 (1933), report b.p. 120-125°, n§° 1.4001
for diisopropyl ketone.
ammonia was replaced by another I 4OO ml. of dry ether.
To the well stirred ethereal suspension was added 36O g.
(0.25 mole) of methyl iodide over a 2 hour period at - 1 1 5 - rooni temperature. The resulting mixture was stirred at reflux for 2 hours and stirring was continued overnight
without external heating. The reaction mixture was
cooled and a sufficient amount (ca. 700 ml.) of water was
added to dissolve salts. The ethereal layer was separated
and treated in the usual manner. The residual organic
liquid was distilled through column B to give 233.6 g.
(85%) of isopropyl t-butyl ketone, b.p. 73-74° (97 mm.),
1 .4060- 1.4064 (reported^^ b.p. 132-134.5°, n§° 1.4073, 1.406o24).
(24) S. S. Nametkin and K. S, Zabrodina, Dokladv
Akad. Mauk. U.S.S.R.. 75, 395 (1950), [C.A., 45 , 6998
(1951)].
The average yield from four runs was 83% and 763 g.
(6.0 mole) of isopropyl t-butyl ketone was prepared.
D i-t-butvl ketone. To a sodium amide suspension
prepared from 65 g. (2.8 g; atom) of sodium and 2 1. of
liquid ammonia was added 332.5 g . (2. 6 mole) of isopropyl
t-butyl ketone. The liquid ammonia was then replaced by
3 1. of toluene.The mixture was heated on the heating
(25) Commercial reagent grade toluene was d is tille d
over sodium wire through column D and the fraction boil
ing at 109 . 0- 109 . 5° was employed. -1 1 6 - mantle to enforce a brisk reflux with stirring until evolu tion of ammonia ceased. After cooling the mixture, 370 g.
(2,9 mole) of dimethyl sulfate was added at a rate s u ffi cient to sustain a moderate reflux. Vigorous stirring and slow addition were required for control of the reac tion, After refluxing for 2 hours, the reaction mixture was cooled and 200 ml, of ammonium hydroxide and 450 ml, of water was added over a 1,5 hour period. The resu ltin g mixture was allowed to stand with occasional agitation for 12 hours and the toluene layer was separated and treated in the usual manner.
The above procedure was repeated on a 1 and 2,3 mole scale and the combined residual liquid from the three runs was distilled through column A to give 504,0 g, (60#) of di-t-butyl ketone, b,p, 145- 151°, ng^ 1,4215 (report-
8d23 b,p, 144-150° at 740 mm., n^° 1,4191-1,4197),
Fractions, b,p, 139-145°, n^° 1,4180, and b, p,
151-157°, n^^ 1 ,4222, which weighed 42,0 and 47,4 g,, respectively, were also subjected to the following Grig nard reaction,
1,1-Di-t-b u ty leth an o l. To a méthylmagnésium iodide reagent prepared from 73 g, ( 3,0 g, atom) of magnesium turnings and 440 g (3.1 mole) of methyl iodide in 800 ml, of dry ether was added 284,5 g, ( 2,0 mole) of d i-t-b u ty l ketone in 700 ml, of dry ether. After refluxing for approximately one hour a crystal was found on the wall -1 1 7 - of the flask . The external heat was then removed^^ and
(26) Continuous heating resulted in a sudden violent reflux of the mixture probably due to the heat of crystal lis a tio n , the surface of the flask was rubbed with a piece of ice.
Immediately a large amount of crystals was formed with evolution of heat. The resulting mixture was refluxed for an additional 5 hours and poured on crushed ice, A su fficien t amount of 6N sulfuric acid was added to dissolve salts and the ethereal solution was separated and worked up in the usual manner.
The same procedure was repeated several times on a to ta l of 575,6 g , (4.1 mole) of d i-t-b u ty l ketone. The combined residual organic liquid was distilled through column C to give 454,9 g, (71%) of 1,1-di-t-butylethanol, b,p, 78-02° (17 mm.), m.p, 42,0-42,7° (sublimed sample),
(reported^^ b,p, 122,5-123° at 100 mm., m.p, 42°).
(27) F, C, Whitmore and K, G. Laughlin, J . Am.
SMAU-âaSui., Ü , 3732 (1933),
1,1-Di-t-butylethylene, To a well stirred, ice-cold mixture of 95.0 g, (0.6 mole) of 1,1-d i-t-b u ty leth an o l and 300 ml, of reagent grade pyridine was added 106 g,
(0,9 mole) of thionyl chloride over a 1,5 hour period. -1 1 8 -
After stirring for an additional hour at 0-10°, approxi mately 100 ml. of water was added and the resulting mixture was worked up in the usual manner (acid washing).
The above procedure was repeated with 2,25 mole of
the alcohol and the combined organic residue was distilled
over potassium hydroxide pellets in order to remove a
bad-smelling sulfur compound. Redistillation of the
slightly cloudy distillate through column B gave 310,3 g,
(78%) of 1,1-di-t-butylethylene as a colorless oil, b,p.
146- 150°, 1,4358-1,4361 (reported^? b.p, 149.5°,
1 .4364).
2,2-Di-t-butylethanol. To a solution of 9.5 g.
(0.25 mole) of sodium borohydride in 250 ml, of purified 2 8 diglyme were added 70 g, (0,50 mole) of d i-t-b u ty l-
(28 ) See H, C, Brown, E, J. Mead and B, C, Subba Rao,
J. Am. Chem. Soc,. 77.. 6209 (1955).
ethylene and then a solution of 11,2 g, (O.O 84 mole) of
aluminum chloride in 50 ml, of diglyme. After s tirrin g
for one hour at room temperature and for 7 hours on the
steam bath, the diglyme (230 ml,) was removed under re
duced pressure. To the residue was added 200 ml, of 2W
hydrochloric acid with stirring and the organic layer
was worked up in the usual manner to give 73,8 g , of
residue. - 1 1 9 -
The residue was dissolved in an alcoholic sodium hydroxide solution, prepared from 8,0 g, of sodium hydroxide and 100 ml. of grain alcohol. To the mixture was added 68 g. ( 0.6 mole) of 30^ hydrogen peroxide at a rate sufficient to sustain gentle reflux. After stirring for one hour, 350 ml, of water was added. The organic layer was separated, washed with 2N hydrochloric acid and lOi? sodium th io su lfate solution and worked up in the usual manner. The residual organic liquid was distilled through column 0 to give 51.5 g, (65%) of 2 , 2- d i- t- butylethanol, b.p. 105- 110° (29 mm.), which cry stallized in long needles, m.p. 52-54°.
The analytical sample, m.p. 54.0-55.0°, was
(29 ) A. Arkell, Ph.D. Dissertation, The Ohio State
University, reported m.p. 51.8-53.8°.
obtained by recrystallization of the crude alcohol from
Skellysolve F followed by sublimation.
Anal. Calcd. for CioHgjp: C, 75.9; H, I 4 .O.
Found: C, 76.0; 76.1; H, 1 4.O, I 4 .O.
2,2-Di-t-butylethyl p-nitrobenzoate. A mixture of
0.2 g. of the pure alcohol, 0.5 g. of p-nitrobenzoyl
chloride and 4 ml. of benzene was refluxed for one hour.
The hot mixture was poured onto 10 ml. of water .with s t i r
ring and the organic layer was worked up in the usual -1 2 0 - manner to give a residual oil, which was dissolved in a small amount of alcohol. On adding water fine needles precipitated, re c ry sta lliz a tio n of which from alcohol gave 2,2-di-t-hutylethyl p-nitrobenzoate as large prisms, m.p. 78.8-79.4°.
Anal. Calcd. for C, 66.4; H, 8,2;
N, 4*6. Found: C, 66.3, 66.5; H, 8,1, 8,3; H, 4.6, 4.7. Di-t-butvlacetic acid. To a well stirred solution of 86.5 g. (0,54 mole) of 2,2-di-t-butylethanol and
270 ml, of sulfuric acid-acetic acid mixture^^ was added
(30) The mixture was prepared by mixing 50 ml, of concentrated sulfuric acid and 100 ml. of water followed by the dilution with acetic acid to 500 ml,
240 ml, (0,60 mole) of a chromic oxide solution^! over a
(31) This solution was prepared by dissolving 125 g.
(1,25 mole) of chromic oxide in 125 ml. of water followed by the dilution with acetic acid to 500 ml,
1,5 hour period. After standing overnight and heating
on the steam bath for one hour, 300 ml. of water was added
and the organic substance was taken up into ether-benzene
mixture. The ethereal solution was worked up in the
usual manner to give 35.5 g, of crystalline acid and -1 2 1 - a neutral fraction.
The reoxidation of the neutral fraction in the same manner yielded a further 41.1 g. of acid. The total yield of crude di-t-butylacetic acid, m.p. 72-74°, was thus 76.6 g. (82^).
This acid was pure enough for further reactions and had an infrared spectrum identical in all respects to that of the analytical sample, m.p. 80.5-81.5°,^^ which
(32) A. Arkell, loc. cit.. reported m.p. 71.0-72.8° for this acid. was obtained by re crystallization of the crude acid from
Skellysolve F at -78° followed by sublimation in vacuum.
Anal. Calcd. for CioH20°2* 69.7; H, 11.7.
Found: C, 69.6, 69.5; H, 11.6, 11.6. 33 A nuclear magnetic resonance analysis of the pure
(33) The author is indebted to Dr. George Tiers of
the Minnesota Mining and Manufacturing Company for the
nuclear magnetic resonance analysis.
sample agrees perfectly with the structure of di-t-butyl-
acetio acid: one proton of carboxyl group at -2.03 , one
proton on a-carbon at 7.83 , and much more than ten
protons due to t-butyl groups at 8.84 , -1 2 2 -
3. Isopropyl-t-butylacetio acid. Triethylmethyl t-butylacetate. To a sodium triethylmethoxide suspension prepared from 0,52 mole of sodium amide and 60.4 g *
(0,52 mole) of triethylm ethanol in approximately 400 ml, of dry ether was added 62,2 g, ( 0,46 mole) of t-b u ty l- acetyl chloride (b,p, 125-128°, n^^ 1,4213)^^ prepared
(34) J . 0. Traynham and M, A, B attiste , J. Ore.
Chem.. 22. 1551 (1957), report b,p, 68-71° (lOO mm,), 20 ng 1,4229. from 70 g, (0,60 mole) of the corresponding acid, in 50 ml, of dry ether. After refluxing for one hour, the re
sulting reaction mixture was worked up in the same manner
as described for triethylmethyl diisopropylacetate to give
85.3 g. (86%) of triethylmethyl t-butylacetate, b.p,
94.5-96,5° (9,5 mm,), n^° 1,4292,
Anal. Ogled, for 0, 72,8; H, 12,2,
Found; G, 72,9, 72,7; H, 12,2, 12,4,
Triethylmethyl isopropyl-t-butylaestate, To a
potassium amide solution prepared from 16 g, (0,41 g, atom)
of potassium and one 1, of liquid ammonia was added 75 g.
(0,35 mole) of triethylm ethyl t-b u ty la cetate in 100 ml,
of dry ether. The pale green reaction mixture was stirred
for 45 minutes, and 72 g, (0.42 mole) of isopropyl iodide
in 100 ml, of dry ether was added. The resulting colorless - 1 2 3 - reaction mixture was stirred for one hour and worked up in the same manner as described for triethylmethyl triiso- propylacetate.
R e d istilla tio n of the crude alkylated ester gave
62.4 g. (70%) of triethylmethyl isopropyl-t-butylacetate, b.p. 94 - 96 ° (1.5 mm.), n^° 1 .4482 .
Anal. Calcd. for C16H32O2 : C, 74.9; H, 12.6.
Found: C, 75.1, 75.0; H, 12.7, 12.4.
From the reaction mixture 18.0 g. (27%) of the starting ester was recovered.
Isopropvl-t-butvlacetic acid. A mixture of 12.3 g.
(0.48 mole) of the triethylm ethyl e ste r, 8 ml, of con
centrated hydrochloric acid and 5 ml. of dioxane was
refluxed for 2 hours. The dioxane was then distilled and
the residue, consisting of two layers, was worked up in
the usual manner to give 6.8 g. (90%) of isopropyl-t-
butylacetic acid, b.p. 92-95° (3 mm.), redistillation of
which through column 0 furnished the analytical sample,
b.p. 100.0-100.2° (4 mm.), n^^ 1.4350 (reported^^ b.p.
(35 ) A, A. Sacks and J. G. Aston, J. Am. Chem.
Soc^, 71, 3902 (1951).
122. 5- 123.8 ° at 22 mm., n^° 1.4343). - 124—
4. Attempted Alkylation of Triethylmethyl dl-t-butyl- acetate. Dl-t-butvlacetyl chloride. A mixture of 8.6 g.
(0.05 mole) of d i-t-b u ty lacetic acid and 25 ml. of thionyl
chloride was refluxed for one hour after vigorous spon
taneous gas evolution had subsided. The excess thionyl
chloride was evaporated and the residue was distilled
to give 9.2 g. (96.5/^) of the crude acid chloride as a
colorless o il, b.p. 83-86° (12 mm.), IR; 4.6 (w)^^.
(36) Characteristic of ketene, see Chapter IV-F,
p. 64 ff.
5.57 (s).37
(37) Characteristic of acid chloride carbonyl, see
L. J. Bellamy, loc. cit.. p. 125.
The crude acid chloride was used for further re
actions without purification.
Triethylmethyl di-t-butvlacetate. To a suspension
of 0.22 mole of sodium triethylmethoxide in 300 ml. of
dry ether, prepared as described for the ester ofdiiso-
propylacetic acid, was added 36 g. (0.19 mole) of d i- t-
butylacetyl chloride in 20 ml. of dry ether. The re
action mixture was refluxed for 2 hours and decomposed,
after cooling, with 50 ml. of water but no violent
reaction was observed. The ethereal layer was separated -1 2 5 - and washed with 10^ sodium hydroxide solution several times. The first two washings caused considerable evolution of heat. After working up the ethereal solu tion in the usual manner, the residual organic liquid was cooled in the low temperature bath (-78°) to separate
8.8 g. of recovered d i-t-b u ty la ce tic acid. The mother liquor was then distilled to give crude ester, redistilla tion of which through column C gave 16.3 g. (30^) of
triethylmethyl di-t-butylacetate, b.p. 105- 111° (2 mm.).
The center cut, b.p. 110° (2 mm.), n^*^ 1.4-570, of the distillate was analyzed.
Anal. Calcd. for G^yH^^Og: C, 75.5; H, 12.7.
Found; 0, 75.8, 75.6; H, 12.5, 12.4.
From the alkaline extracts 12.7 g. of d l-t-b u ty l-
acetic acid was recovered. Thus, the total recovery of
the acid was 21.5 g. ( 63%) and the conversion of the
acid chloride to the ester was 94 %»
M_tempt.ed..,alkvlation. With etJivl l.odide. To a
potassium amide solution prepared from 1.7 g. (O.O 44 g.
atom) of potassium and 200 ml. of liquid ammonia was
added 10.8 g. (O.O 4O mole) of triethylmethyl di-t-butyl-
acetate in 60 ml. of dry ether. The resulting mixture
was stirred for 3 hours but no color change was observed.
To the mixture was then added 7 g. (0.045 mole) of ethyl
iodide in 15 ml. of dry ether, which was stirred for 2
hours. The almost colorless mixture consisting of two - 1 2 6 - layers was decomposed by adding 60 ml. of wet ether and liquid ammonia was evaporated on the steam bath. The re action mixture was treated in the usual manner and the organic residue was distilled through column C to give
9.8 g. ( 91 ^) of the startin g ester, b.p. 104-107°
(15 mm,).
With methyl iodide. To a suspension of O.O 6 mole of potassium amide in 200 ml. of liquid ammonia was added
10.8 g. ( 0.04 mole) of the ester in 10 ml. of dry ether.
The mixture was stirred for 4 hours without color change.
The liquid ammonia was then slowly evaporated at room temperature over a 7 hour period while 200 ml. of dry ether was added. The color of the mixture never changed but when the amount of ether increased black solids pre cipitated. To the ethereal solution was added 10 g.
(0,07 mole) of methyl iodide in 15 ml. of dry ether.
After stirring overnight, the reaction mixture was worked up in the usual manner to give 10.0 g. ( 96 %) of the starting ester, b.p. 100-110° (1.5 mm.),
5. Attempted alkylation of triethylmethyl isonropvl- t-b u ty la c e ta te . With isopropvl iodide. To a solution of
0,064 mole of potassium amide in 200 ml. of liquid am monia was added 15.4 g. (O.O 6O mole) of triethylm ethyl isopropyl-t-butylacetate in 15 ml. of dry ether. Within
10 minutes, the mixture became deep yellow but consisted of two layers; the top layer was yellow and the bottom -1 2 7 - colorless. Addition of another 50 ml. of dry ether in creased the volume of the bottom layer. After stirring for 2 hours, 11.0 g. (0.065 mole) of isopropyl iodide in
15 ml. of dry ether was added. The mixture became color less within 10 minutes. After refluxing for 2 hours, liquid ammonia was evaporated on the steam bath and the reaction mixture was worked up in the usual manner. The residual organic liquid was distilled through column C to give 12.7 g. (83%) of the starting ester. The dis tillation residue (2.2 g.) showed a slightly different absorption in the 8-9 u region.
With ethyl iodide. On running a similar experiment using ethyl iodide in place of isopropyl iodide on an
0.02 mole scale, the starting acid was recovered in 40% yield (2.0 g.), b.p. 93-120° (2 mm.), IR: 5.88 (s),
8.3 (s), 8.65 (s), 8.9 (s). On addition to this, 1.5 g.
of a high boiling fraction, b.p. 120- 125° (2 mm.), and
1.5 g. of residue, IR: 5.93 (s), no strong band in the
8.2-9.0 u region, were obtained,
VI-D. Reactions of Highly Sterically Hindered Aliphatic
Acids and Their Derivatives
1. Ionization constants of acids. General Procedure 38
(38) A modification of procedure used by G. S.
Hammond and D. H. Hogle, J. Am. Chem. Soc. , 77. 338 (1955).
Ionization constants of acids were determined by -1 2 8 - potentiometric titration. The titration was made at
4-OiO,l° using a Beckman pH meter, model G, a g la s s
electrode, Beckman 1190-72, and a calomel reference
electrode, Beckman 1170, The electrodes were in s e r te d
directly into sample solutions. Temperature was con
trolled by using a thermostatic water bath and an elec
tronic relay controlled by a mercury to mercury thermo
regulator, Precision Scientific Co,, catalog numoer 66532,
The system was standardized before and after each
titration with three buffers recommended^^ for calibration
(39) H. H, W illard, 'L. L, Merritt and J, A. Dean,
"Instrumental Methods of Analysis," D, Van Nostrand
Company, Inc,, Princeton, W. J,, 1958, p, 4.47-69,
of equipment: 0.05 M phthalate buffer for pH 4,03,
0,05 M phosphate buffer for pH 6,84 and 0,01 M Borax
buffer for pH 9,07,
Sample solutions were prepared in the following way:
an accurately weighed amount of acid ranging from 0.352
to 0,773 mmole depending upon the solubility was placed
in a titrating beaker of 180 ml, capacity. The acid was
dissolved by adding 50 ml, of methanol pipetted from a
methanol reservoir kept in the water bath at 40°. A fter
the acid had completely dissolved, 50 ml. of water was
added in the same way as for methanol. -1 2 9 - The sample solutions were titrated with 0.0967 i 0.0001 N carbonate-free methanolic sodium hydroxide solution in a 25 ml. needle valve burette with a side arm, reading to 0.05 ml., held at room temperature.
During the titration the solution was stirred by a steady
stream of purified nitrogen, both to keep the solution
free from carbon dioxide and to avoid the pH meter fluctu
ation encountered when a magnetic stirrer was used for
agitation. The temperature (40i0.1°)of the solution was
checked by a thermometer, reading to 0.1^, inserted into
the solution.
The acids titrated are shown in Table V together
with their physical constants. The solid acids were re
crystallized and sublimed to insure that the highest
melting point had been reached. The liquid acids were
purified by distillation through column G or D,
At least two titrations were made on each acid ex
cept methyl-t-butylneopentylacetic and t-butylacetic
acids.
From the data of the differential titration pH was
plotted against volume of the standard base and a smooth
curve was drawn. The end-point was determined by both
the Gran^^ and the Fenwick'^^ methods.
(40) G. Gran, Acta Chim. Scand.. 4. 559 (1950). é (41 ) F. Fenwick, Ind. Eng. Ghem.. Anal. E d., 4, I 44 (1932 ). - 1 3 0 -
Ths Gran method for determining the end-point of a weak acid-strong base system consists of plotting
V*10^“^^ and (Vo+V)•10^^ against V before and after the end-point, respectively. The V and Vo stand for the volume (ml.) of standard base added and that of original solution, respectively. The k is an arbitrary constant and most conveniently the nearest integer to the pH value at the approximate end-point. The plot gives two straight lines and the intercept from their point of intersection on the abscissa gives the end-point (ml.). The initial volume of the sample solution was calculated to be 101.8 ml. from the data of density and cubic expansion coeffici- end of water and methanol.
(42) N. A. Lange, "Handbook of Chemistry," Hand book Publishers, Inc., Sandusky, Ohio, 1949, p. 1275.
Fenwick proposed that the end-point of d iffe re n tia l titration could be calculated by assuming that a titration curve may be represented by a cubic equation (l) in the vicinity of the end-point,
aV^ + bV^ + cV + d = pH (l) where V stands for the volume of standard base. A^ the end-point the equation (2) must be satisfied.
d^pH - 6aV + 2b = 0 (2 ) —131“
Thus, by taking four values of V, equidistant (O.l ml.) from one another in the region of the end-point, the following equation ( 3) can be derived as an end-point expression,
2 (pH2-pHi)“ (pH^-pH^) Vend. = V + 0.1 - Yo "* (3) (pH3“pH i)-3 (pH2“pH3_) where pHn stands for the pH value at (V+n/lO) ml.
The end-points obtained by these two methods agreed to within 10.01 ml. The neutral equivalent calculated from the end-point thus obtained are shown in Table V.
The ionization constants were calculated by the
Henderson equation^^ (4) using l/4, 1/2 and 3/4
(43 ) S. Glasstone, "Text-Book of Physical Chemistry,”
D. Van Nostrand Company, Inc., New York, N. Y., 1940, p. 982 .
pKg = pH - log ( 4 )
neutralization points. The values thus calculated were accurate to 10.03 pH unit and the re su lts are summarized in Table VI.
The general procedure may be illustrated by triiso- propylacetic acid. A sample solution was prepared as described above from O.O 668 g. (O .358 mmole) of the - 132-
aciâ and titrated with the standard base (0.0967f0.0001N)
The following pH values were read against volume (V ml.)
of the standard base.
V(ml. ) pH V(ml.) pH V(ml.) pH
0 5.28 3.10 8.10 4.00 10.34 0.25 6.19 3.20 8.20 4.05 10.41 0.60 6.61 3.30 8.31 4.10 10.45 0.80 6.79 3.35 8.36 4.20 10.56 1.00 6.91 3.40 8.46 4.30 10.64
1.20 7.03 3.45 8.51 4.40 10.71 1.40 7.13 3.50 8.62 4.50 10.78 1.60 7.23 3.55 8.71 4.60 10.82 1.80 7.32 3.60 8.90 4.70 10.87 2.00 7.43 3.65 9.13 4.80 10.90
2.20 7.52 3.70 9.36 5.00 10.98 2.40 7.62 3.75 9.70 5.25 11.05 2.60 7.73 3.80 9.85 5.50 11.11 2.80 7.87 3.85 10.04 5.75 11.17 2.90 7.93 3.90 10.13 6.00 11.21 3.00 8.01 3.95 10.24 6.25 11.24
Buffer pH 4# 03 6.84 9. 07 Before titration A. 04 6.84 9. 06 After titration 4.02 6.83 9. 05
The pH values were plotted against V and from a smooth curve thus obtained the following : fifteen points were read.
V(ml .) pH Gran's Value
3.0 8.01 293 3.1 8.10 246 3.2 8.20 2 02 3.3 8.30c 164 3.4 8.42 129 3.5 8.62 84 3.6 (V) 8.90 (pHo) 45 3.7 9.36 (pHl) 16 3.8 9.86 (pH2) 96 3.9 10.235 (PH3 ) 182 -133- V(ml.) pH Gran's Value
4.0 10.32 221 4.1 10.45 299 ■ 4.2 10.56 385 4.3 10.64 463 4.4 10.71 546 4.5 10.775 633
By takihg four values (V=3.6, 3.7, 3.8 and 3.9 ml.) and substituting them into equation (3), the end-point was
Ven.. . 3.6.0.1-
calculated to be 3.72^ ml.
On the other hand the values for the Gran equations were calculated by using 10 as the arbitrary constant k and 101.8 ml. as Vo, and assuming that the end-point lies between 3.7 and 3.8 ml. The values are shown in the third column in the last table. The, are then plotted against V to give two straight lines. The end-point was determined from the reading of V at the intercept of these two lines to be 3.72 ml.
Thus, the neutral equivalent was found to be 185.8
(calcd., 186. 3 )
0.0668 X 103 = 185.8 3.72 X 0.0967
The pKa was then calculated according to the Hender son equation ( 4 ) by using the pH readings at the I/ 4 ,
1/2 and 3/4 neutralization points. —13 4 —
Neutralization point 1/4 1/2 3/4
V, (ml,) 0,93 1,86 2.79
pH 6,87 7,36 7,86 A- 0.48 0 - 0,48 KA pKa 7,35 7,36 7,38
End-points of the titration and pKa values at the
three neutralization points of the acids (R 2R2R2GCOOH)
titrated are listed in the following table together
with amounts of the acids used.
Reagents used. Watep. Boiled distilled water was used.
Methanol. Commercial absolute methanol was refluxed for a few hours with sodium metal and dimethyl phthalate and distilled through column B with exclusion of moist air. A fraction boiling at 64, 5° was employed.
Carbonate-free methanolic sodium hydroxide solution,
Approximately 6 g, of sodium hydroxide was dissolved in one 1. of warm methanol. After standing in a tightly stoppered flask for 24. hours, the top clear solution was decanted into a reagent bottle and standardized against potassium acid phthalate by potentiometrie titration.
The standard base solution (0,0967l0,0001N) was protected from atmospheric carbon dioxide by an Ascarite (sodium hydrate asbestos absorbent) absorption tube.
Nitrogen. Commercial nitrogen was purified by TâBLE X
Determination of Ionization Constants of Acids
Amount of End-point R1R2R3GGOOH pKa Acid Used ml. %1 ^2 K3 mg. mmole of base 1/4 1/2 3/4
H HH 60.4 1.006 10.39 5.67 5.68 5.70 56.9 0.948 9.70 5.68 5.69 5.72 t —C^Hg H H 83.7 0.720 7.42 6.22 6.24 6.25 i-CjHy i-GgHy H 85.7 0.594 6.05 6.45 6.47 6.49 I 94.7 . 0.657 6.70 H 6.47 6.49 6.51 Vo VR H I C2E5 109.3 0.758 7.92 6.47 6.49 6.52 106.9 0.741 7.68 6.49 6.50 6.50 G2H5 G2H5 G2H5 119.1 0.827 8.39 6.64 6.66 6.67 109.6 0.759 7.82 6.64 6.65 6.67 t —G^Hg i —G gHy H 110.3 0.697 7.18 6.72 6.76 6.77 116.7 0.738 7.58 6.74 6.77 6.78 t —G^Hg GH3 GH3 108.1 0.750 7.73 6.93 6.95 6.97 107.5 0.746 7.75 6.92 6.94 6.96 i —GgHy 1“G G H3 120.4 0.762 7.84 6.98 6.98 7.00 118.1 0.747 7.58 6.94 6.96 6.96
H 60 .6 0.352 3.68 7.04 7.05 7.07 65.9 0.383 3.99 7.04 7.04 7.05 TABLE X (continued)
R^RgR^CGOOH Amount of End-point pKa Acid Used ml. R3 mg. mmole of base 1/4 1/2 3/4 i —C i —G qH/7 GpHc 133.2 0.773 7.98 7.21 7.22 7 .24 J) ( J ( ^ 5 126.6 0.734 7.63 7.22 7.24 7.25 ne 0—C t — C g CH3 71.7 0.358 3.73 7.28 7.31 7.32 i —C oHy i —G oHY 1—G3HY 66.8 0.358 3.72 7.35 7.36 7.38 I 67.4 0.362 3.74 7.36 7.36 7.37 H* V.O o\ I -1 3 7 - passing through two wash bottles of the Fieser solution,
(44 ) L. F, Fieser, "Experiments in Organic Chemis try," 3rd ed., D. G. Heath and Company, Boston, Mass.,
1955, p. 299 . a bottle of saturated lead acetate solution and a calcium
chloride drying tower.
Buffer solutions. Phthalate buffer, pH 4.03 at 40°, was prepared by dissolving 10,21 g. of potassium acid phthalate (dried for 2 hours at 110°) in boiled distilled water to make one 1. solution.
Phosphate buffer, pH 6.84 at 40°, was prepared by
dissolving 3.44 g. of potassium dihydrogen phosphate and
3.55 g. of disodium hydrogen phosphate (each dried for
2 hours at 110°) in boiled d is tille d water to make one 1.
solution.
Borax buffer, pH 9.07 at 40°, was prepared by dis
solving 3.81 g. of borax in boiled distilled water to make
one 1. solution.
2. Reaction of acids with ethoxvacetvlene. Diiso-
pj_opylacetic anhydride. â mixture of I .4 g. (O.Ol mole)
of diisopropylacetic acid and 0.5 g. (0.007 mole) of
ethoxyacetylonej b.p. 48-51°, in 3 ml. of dry ether was
kept at room temperature for 2 hours and refluxed for
one hour. The solvent and low boiling substance —1 3 8 — smelling of ethyl acetate were evaporated and the residue was distilled to give 0.52 g. (39%) of diisopropylacetic 2 Q anhydride, b.p. 105° (2 mm.), n^ 1.4^20, IR: 5.5c (s),
5 . ? 2 (m). 45
(45) L, J. Bellamy, loc. cit.. p. 127.
fi-nal. Calcd. for C, 71.1; H, 11.2.
Found; C, 70.8, 71.0; H, 10.9, 11.1
From the reaction mixture 0.59 g. (42#) of the
starting acid was recovered as a low boiling fraction
of the distillation. Tri is opr opyla ce.t.ic anhydride . k mixture of 1.9 g.
(O.Ol mole) of triisopropylacetic acid and 0.5 g. (0.007
mole) of ethoxyacetylene in 25 ml. of dry ether was re
fluxed for 5 days. Evaporation of low boiling substance
smelling of ethyl acetate gave a solid residue which was
recrystallized from Skellysolve 0 to give 1.1 g (60.8#)
of the crude anhydride, m.p. 88-90°, and 0.74 g. (39#)
of the crude starting acid. The fractional recrystal
lization of the crude anhydride from the same solvent
gave large hexagonal crystalline triisopropylacetic an
hydride, m.p. 92-93°, IR: 5.60 (s), 5.79 (m).^^ Anal.. Calcd. for G, 74.5; H, 11.9. Found: C, 74.8, 74.8; H, 11.7, 11.8. On running the same reaction for 18 hours, a very
small amount of triisopropylacetic anhydride was formed but could not be isolated. -1 3 9 -
3. Reactions of acids with trifluoroacetio anhydride..
TriisQ-prooylacetic acid . Into 2,2 g. (O.Ol mole) of t r i - fluoroacetic anhydride was added 1,9 g. (0,01 mole) of triisopropylacetic acid, Within 5 minutes the acid completely dissolved and slow evolution of gas took place. The gas was probably carbon monoxide since no precipitate was formed by passing through a barium hydrox ide solution. The gas evolution became much slower on cooling the mixture in an ice-water bath. While standing for approximately 1,5 hours the mixture became brown and then separated into two layers. The top layer was color less and, after washing with water, decolorized perman ganate solution or bromine in carbon tetrachloride. The bottom layer was dark brown.
Djisooropvlacetic acid. This acid did not evolve any gas when treated with the anhydride in a similar way.
4. Reaction of acids with thionvl chloride. Triiso- propylacetvl chloride. A mixture of triisopropylacetic acid and a large excess thionyl chloride was heated to reflux while the acid slowly dissolved with evolution of gas. After refluxing for 1-2 hours the excess thionyl chloride was evaporated. The residue was distilled under reduced pressure to give a 90-95# yield of triiso p ro p y l- acetyl chloride, b.p, 100-101° (6 mm,), which solidified into a wax-like solid, m.p. 49-52°, The distilled sub stance was employed for further reactions. —14.0—
Recrystallization of the crude product from Skelly solve B at -78° gave pure triisopropylacetyl chloride, m.p. 54.2-55.2° in a sealed tube, IR: 5.60 (s), as a wax-like solid.
The acid chloride was very sensitive to the moisture of the air. Since the surface of the solid became powdery when transferred into a sample tube, the elemental analy sis was not attempted.
Di-t-butvlacetvl chloride. See Chapter VI-C, p . 1 2 4 .
Diisopropylacetyl chloride. See Chapter VI-C, p. 111.
5. E stérificatio n of acid chloride. Methyl t r i i s o - j?ropylacetate. Approximately 20 ml. of anhydrous methanol was added to 8.1 g. (0.04 mole) of triisopropylacetyl chloride at 0°, The addition caused an instantaneous reaction and the mixture separated into two phases. After standing one hour at room temperature, the excess methanol was evaporated and the residue was distilled through column C to give 7,2 g. ( 90 %) of methyl triisopropyl- acetate as a colorless oil, b.p. 91.5-92.5° (6.5 mm,),
1.4518, IR; 5.75 (s).
Anal. Calcd. for OigHg/Og: C, 72.0; H, 12.1.
Found: C, 72.0, 71.7; H, 12.0, 12.1
Triethylmethyl di-t-butvlacetate. See Chapter VI-C, p . 1 2 4 . - 1 4 1 - Triethvlmethvl diisopropvlacetate. 'See Chapter VI-G p .I l l and VI-D, p. I 45,
6. Reaction of acid chlorides with ammonia and sodiup amide. Triis opropylacetamide. Into a suspension of 0.065 mole of sodium amide in 150 ml. of dry liquid ammonia^G was added 9.9 g. (O .048 mole) of triiso p ro p y l-
(4 6 ) Commercial anhydrous liquid ammonia was dried with sodium and d is tille d into a reaction flask equipped with a Dry Ice-acetone condenser.
acetyl chloride in 30 ml. of dry ether. After stirring for 2 0 hours, 4 g. of ammonium chloride was added and the liquid ammonia was evaporated on the steam bath, while
100 ml. of dry ether was added. Insoluble inorganic salts were filtered and washed with dry ether several times.
Evaporation of the ether from the filtrate gave solid residue, which was recrystallized from benzene-Skelly- solve B to give 8.3 g, (93%) of triisopropylacetaraide, m.p. 141.8 - 142.8 °, IR; 6.1 (s).47
(47) See L. J. Bellamy, lo c . c i t . . p. 216.
Anal. Calcd. for C^Hg^ON: C, 71.3; H, 12.5;
N, 7.6. Found; C, 71.4, 71.6; H, 12.6, 12.4; U, 7.5, 7.6,
The same reaction with 13.0 g. (O.O 64 mole) of tr i i s o -
propylacetylchloride gave 5.5 g. (47%) of the amide after -1 4 2 - a 2 0 h o u r reaction period.
On running a similar reaction in the absence of sodium amide for 30 hours a 68!» yield of the amide was obtained. A portion of the reaction mixture withdrawn after a 20 hour reaction period had, a fte r treating with water, infrared absorption bands of a similar intensity at 5.9 and 6.1 characteristic of acid and amide, respec tiv ely .
Dj-t-butylacetamide and di-t-butvlketene. To a
sodium amide suspension prepared from 1.0 g. ( 0.0435 g. atom) of sodium in 50 ml. of dry liquid ammonia was added
5.0 g . (0.026 mole) of d i-t-b u ty lacety l chloride in
approximately 20 ml. of dry ether during 15 minutes, while considerable frothing was observed.
After stirring for 24 hours 2.5 g. of ammonium chlor
ide was added l i t t l e by l i t t l e , and the liquid ammonia was
evaporated while 100 ml. of dry ether was added. Inorganic
salts were filtered and washed several times with dry
ether. After removing the solvent from the filtrate (and
washings) 0.3 g. of cry stallin e plates formed was filte re d
and the liquid was distilled to give 2.3 g. (57$) of di-
t-butylketene as a yellow oil, b.p. 74-76° (4? mm.).
Redistillation through column E gave 2,0 g. (50$) of the
pure sample, b.p. 73° (45 mm.), n^° 1.4370, IR; 4.80 (a),
no band in the 5-6 u region, - 1 4 3 -
(48) See Chapter IV-F, p. 64 f f .
Anal. Calcd. for SQ_QH]_gO; C, 77.9; H, 11.8.
Found; C, 77.9, 77.9; H, 12.0, 12.1.
The distillation left 1.9 g. of nonvolatile solid residue. All the solid products were combined and re- crystallized from aqueous ethanol and then from benzene-
Skellysolve B to give 1.3 g. (29%) of di-t-butylacetamide, m.p. 109 . 5- 110.2°, IR; 6.0 (s).4?
Anal. Calcd. for C^qH^^ON: C, 70.1; H, 12.4; N,
8.2. Found; C, 70.1, 70.2; H, 12.4, 12.7; N, 8.2, 8.3
7. Reactions of d i-t-b u tv lk eten e. With sodium amide,
To a sodium amide suspension prepared from 0.5 g. (0.02 g. atom) of sodium and 100 ml. of dry liquid ammonia was added O.40 g. (0.0026 mole) of di-t-butylketene in 20 ml. of dry ether. The progress of the reaction was followed by taking infrared spectra of samples withdrawn after known periods of time. A portion of the bulk was siphoned out and a drop of water was added to decompose the unre acted sodium amide. The mixture was immediately dried with anhydrous magnesium sulfate and filtered. After the
ammonia and ether were evaporated, a few drops of chloro form were added and the solution was again concentrated
to furnish samples for spectral analysis. After a 3 hour reaction period both ketene, 4,8 u, and amide, 6.0 u. - 1 4 4 - characteristic bands showed comparable intensity, and after 8 hours the relative intensity of the ketene to amide band was considerably decreased. However, even after 12 hours the sample exhibited a weak absorption characteristic of the ketene. The solid product isolated after 12 hours was identical with di-t-butylacetamide prepared from the acid chloride and sodium amide.
With liquid ammonia. k homogeneous mixture of 0.4 g. of d i-t-b u ty lk eten e, 120 ml. of dry liquid ammonia and
50 ml. of dry ether was stirred at the boiling point of liquid ammonia. The progress of the reaction was follow ed as in the previous run. The infrared spectra of samples after 3, 12 and 24 hours were identical and had a strong absorption band at 4.8 u characteristic of ketene but no absorption in the 6 u region, showing that no re action had occurred.
To the reaction mixture was then added a small amount of solid ammonium chloride and the reaction progress was again followed by taking infrared spectra.
No change in the spectrum was, however, observed after
3, 9 and 19 hour reaction periods.
With a n ilin e . A mixture of 0.5 g. of d i-t-b u ty l- ketene and 1 g. of aniline in 5 ml. of anhydrous benzene was refluxed for 2 hours but, after evaporating the benzene, the infrared spectrum of the residue showed that no re action had taken place. —14-5“
To the residue was then added 5 ml. of anhydrous
benzene and two drops of concentrated sulfuric acid.
Immediately a small amount of white precipitate was formed
but the mixture was refluxed for one hour. After removal
of volatile substance the residue was recrystallized from
benzene-Skellysolve B to give di-t-butylacetanilide, m.p,
1 4 2 .5-1 4 3 .5 °.
Anal. Calcd. for C^^Hg^NO: C, 77.7; H, 10.2; N, 5.7.
Found: C, 77.6, 77.7; H, 10.3, 10.2; N, 5.7, 5.8.
8. Attempted preparation of diisopropvlketene.
Diisopropvlacetvl chloride with tertiary base. A solution
of 16.3 g. ( 0.10 mole) of diisopropylacety l chloride and
11.0 g. ( 0.11 mole) of N-methylpiperidine in 250 ml. of
benzene was kept at room temperature for 2 weeks. The
mixture gave 9.8 g. (72^) of the amine hydrochloride and
distillation of the liquid gave 4.5 {27%) of the starting
acid chloride and 10 g. of undistillable tar.
A similar result was obtained when triethylamine was
used in place of N-methylpiperidine.
Diisopropvlacetvl chloride with tertiary base in the
presence of alcohol. A solution of 16.3 g, (O.IO mole) of diisopropylacetyl chloride, 11.0 g. (O.ll mole) of
N-methylpiperidine and 15 g. (0,13 mole) of trie th y l-
methanol was kept at room temperature for 24 hours and
heated on the steam bath for 2 hours. After filtering
12.3 g. ( 91 %) of amine hydrochloride, the liquid was —14 6 — distilled to give 16.0 g. of product, the infrared spectrum of which exhibited a weak absorption band at
5.5 u besides those characteristic of triethylmethyl diisopropylacetate. Redistillation of the product, after washing with acid and base, gave 11.5 g. of the pure ester.
9. Reactions of esters and amides with lithium
aluminum hydride. Methyl triisopronvlacetat e . Into a
suspension of 1.7 g. (0.045 mole) of powdered lithium
aluminum hydride in 50 ml. of dry ether was added 6.0 g.
(0.03 mole) of methyl triisopropylacetate in 35 ml. of
dry ether. After refluxing 12 hours, the reaction mix
ture was decomposed with 10 ml. of water and approximately
30 ml. of 6K sulfuric acid and worked up in the usual
manner. D istilla tio n of the organic residue gave 5.1 g.
of a mixture, b.p. 75-82° (2 mm.), m.p. 36-41 , of the
desired alcohol and a small amount of the starting ester.
Recrystallization from Skellysolve F at -78° followed
by sublimation gave 3.3 g. ( 64%) of 2, 2, 2-triisopropyl-
ethanol, m.p. 43.0-44.0°.
Anal. Calcd. for G, 76.7; H, 14.O.
Found; 0, 76.6, 76.5; H, 14.O, 14.2.
A mixture of 1.0 g. of the alcohol and 0.8 g. of
phenyl isocyanate was warmed on the steam bath for 5
minutes. On cooling the mixture completely solidified.
The solid was heated with Skellysolve C and insoluble -1 4 7 - diphenyl urea was removed by filtration of the hot mixture
From the filtrate was isolated 2,2,2-triisopropylethyl
N-phenylcarbamate, m.p. 126-127°.
Anal. Calcd. for GigH29^02: G, 74.2; H, 10.0;
N, 4.8. Found: G, 74.3, 74.0; H, 10.0, 9.8; W, 4.8, 4.8.
Triisopropylaoetamide. In a three-necked, round- bottom flask containing a magnetic stirrer and equipped with a dropping funnel and two condensers, were placed
0.6 g. ( 0.016 mole) of finely powdered LiAlH^ and 50 ml.
of purified tetrahydrofuran (THF)^^, and the mixture was refluxed for 12 hours under purified nitrogen.
(49 ) See Chapter VI-D, p. I 34,
To the cooled, resulting suspension was added 4.4 g.
(0.024 mole) of triisopropylacetamide in 30 ml. of THF
under nitrogen. During the addition spontaneous hydrogen
evolution was observed although the volume was not
measured. After refluxing for 48 hours the reaction mix
ture was cooled in an ice-bath and decomposed by the
successive addition of 1 ml. of water, 1 ml. of 10%
sodium hydroxide solution and 3 ml. of water as recommend
ed^^ for amide reductions. The mixture was filtered and
(50) V. M. Micovic and M. L. Mihailovic, J. Ore.
Ghem., 18, 1190 (1953). —14-8— the precipitate was washed with THF and ether several times. The combined solution was dried and, after removal of the solvent, the residue was distilled to give 2,2 g.
(55^) of triisopropylacetonitrile, b.p. 75-73° (4 mm.).
From the distillation residue was isolated 1.1 g. (25%) of the startin g amide.
On running a similar reaction with 4.7 g. (0.025 mole) of triisopropylacetam ide and 0.5 g. (0.013 mole) of lithium aluminum hydride in refluxing ether for 5 days there were isolated 2.8 g. (66%) of triisopropylaceto nitrile and 1.2 g. (26%) of the starting amide together with a trace of amine as the hydrochloride.
The amide recovered in these runs was not pure and the infrared spectra showed a slight contamination,
2,3,5,6-TetramethyIbenzamide. One-half molar lithium aluminum hydride-ether solution (LiAlH^-ether) and 0.25 molar lithium aluminum hydride-purified tetra- hydrofuran solution (LiâlH^-THF) were prepared, standard ized and employed for the following experiments.
A three-necked, round-bottom flask containing a magnetic stirrer was equipped with a dropping funnel, a ground glass stopper and a reflux condenser connected through a calcium chloride tube to a gas collecting cy linder. Reactions were performed either by normal or reverse addition. The volume of gas evolved was calibrated by a blank test to determine the net gas evolution due -1 4 .9 “ to known reactions, After known periods of time approxi mately 10 ml. of reaction mixture was pipetted out and several drops of water was added to decompose the unreacted
LiAlH^, The mixture was dried with anhydrous magnesium sulfate, filtered and concentrated to furnish samples for the infrared spectral analysis.
The reaction mixture was fin a lly decomposed and worked up in the same manner as described for the reaction with triisopropylacetamide.
(i) To 100 ml. ( 0.025 mole) of LiAlH.-THF solution 4 was added 8.9 g. (0.05 mole) of 2 ,3 ,5 ,6-tetramethylbenz- amide in 100 ml. of THF, A sample withdrawn after 20 hours duration showed a medium absorption band at 4.55 u characteristic of 2,3,5,6-tetramethylbensonitrile.
(51) L. Tsai, T. Miwa and M. S. Mewman, J . Am.
Ghem. S o c , 22, 2530 (1957).
After refluxing 40 hours the reaction mixture was decomposed and worked up by the general method. Recrystallization of the crude product from benzene gave 5.5 g. ( 62%) of the startin g amide. D istilla tio n of the mother liquid gave 1.5 g. ( 20%) of low melting substance, b.p. 100-110°
(2 mm.), which was recrystallized from hot water to give
1.0 g. of 2,3,5,.6-tatramethylbenzonitrile, m.p. 73-74°
(reported m.p. 73-74 ), IR: 4.55 (s). From the - 1 5 0 - distillation residue a trace of amine was isolated as the p-nitrobenzoate derivative.
(ii) To 150 ml. (0.075 mole) of LiAlH^-ether solution was added 8.9 g. (O.O 5 mole) of 2 ,3 ,5 ,6 -te tra - methylbenzamide in 150 ml. of THF over a period of 20 minutes while a little more than 0.1 mole of hydrogen gas was evolved. The mixture was stirred at room temperature
and subjected to infrared spectral measurement at 3, 10,
23 and 70 hour reaction periods. In no sample was a
nitrile absorption band observed. Although shifts of
absorption bands in the v icin ity of 6 u were noticed,
no analysis was possible. After stirring for 70 hours
the reaction mixture was decomposed and worked up by the
general method. The crude product was taken up in ether-
benzene and, when it was washed with 2N hydrochloric acid,
a voluminous white precipitate was formed. After adding
more concentrated hydrochloric acid, the precipitate was
filtered and recrystallized from alcohol to give 4.5 g.
(50^) of 2,3,5,6-tetramethylbenzylamine hydrochloride,
m.p, 300° with decomposition in a sealed tube. Anal. Calcd. for CnHj^gNCl; 0, 66.2; H, 9.1;
N, 7.0; 01, 17.8. Found: G, 66.0, 66.2; H, 9.0, 9.1;
N, 7.0, 7.2; 01, 17.6, 17.8.
The amine hydrochloride was dissolved in benzene and pyridine and refluxed with p-nitrobenzoyl chloride for several hours. After filtering pyridine hydrochloride. - 1 5 1 - the solution was worked up in the usual manner. Recrystal- libation of the crude product from alcohol gave N-(p-nitro- benzoyl-)-2,3,5,6-tetramethylbenaylamine, m.p. 218.0-
218.6°.
Anal. Calcd. for G2gH20^203: G, 69.2; H, 6.5; N,
9.0. Found: C, 69.1, 69.3; H, 6.4, 6.7; N, 8.9, 8.9.
From the neutral fraction of the reaction product
1.5 g. (20/0 of the starting amide was recovered and the
remaining liquid showed a medium infrared absorption band
at 4.55 u characteristic of nitrile.
Benzamide. (i) To a solution of 6.0 g. (0.05 mole)
of benzamide in 100 ml. of THF was added 75 ml. (0.019
mole) of LiAlH^-THF solution over a 1.5 hour period at
30-35° while 0.06 mole of hydrogen gas was collected.
The first sample withdrawn as soon as possible after the
addition had a weak nitrile characteristic absorption
band at 4*5 u and that after refluxing for 4 hours ex- 52 hibited strong infrared absorption bands at 4.5 and
(52) See L. J, Bellamy, loc. cit ,, p. 210.
53 and 6.05 u characteristic of nitrile and imine, respec'
(53) See ib id . . p . 268.
tively, and the amide characteristic band at 5.95 became
weak. The reaction mixture was decomposed at this period - 1 5 2 - and worked up as described for triisopropylacetam ide. The crude product was mixed with Skellysolve F and cooled at
-78° to separate 2.1 g. (35%) of the startin g amide. The mother liquor was washed with 6W hydrochloric acid and distilled to give 3.2 g. of a mixture, b.p. 74.-76° (20 mm.),
IR; 4.5 (s),^^ 5.85 (s).^^ The distillate was then
(54) Characteristic of aldehyde, see L. J. Bellamy,
loc. ■Cit^., p. 155.
oxidized with aqueous alkaline permanganate according to
a conventional method^^ to give 0.5 g. ( 4^) of benzoic
(55) R. L. Shriner and R. G. Fuson, "The Systematic
Identification of Organic Compounds," John Uiley and Sons,
Inc., New York, N. Y., 1948, p. 170.
acid and 2.3 g. (45/â) of benzonitrile, b.p. 75-76°
(20 mm.).
(ii) k solution of 3.0 g, (0.025 mole) of benzamide
in 50 ml. of THF was added into 70 ml. (0.035 mole) of
LiAlH^-ether solution over a 100 minute period, while
0.05 mole of hydrogen gas was evolved. The reaction mix
ture was stirred at room temperature. The infrared
spectra of samples withdrawn after 0, 10, 20 and 120
minutes were taken but no nitrile band was observed. - 1 5 3 “
àfter stirring for 18 hours at room temperature the reac tion mixture was decomposed and worked up as described above. Although no pure component was isolated, the neu tral fraction, IR: 4.5 (w), 5.85 (a), contained aldehyde and a small amount of n itr ile .
Benzonitrile.- (i) A solution of 2.6 g. (0.025 mole) of benzonitrile in 100 ml. of THF was added into 100 ml.
(0.025 mole) of LiAlH^-THF solution during 5 minutes at 0-5° and the mixture was stirred at room temperature. Ho gas
evolution was observed. Even the first sample withdrawn as
soon as possible after the start lacked the nitrile char
acteristic band at 4.5 but contained a strong band at 6.05 5 3 ch aracteristic of imine. There was no significant change
in the infrared spectra of samples withdrawn at 0, 10, 40
and 60 minutes of reaction time. After working up the re
action mixture in the usual manner benzaldehyde was obtained
as the major product and characterized as the 2,4-dinitro-
phenylhydrazone, m.p. 238-239° (reported^^ m.p. 237°),
(56) See reference 55, p. 229.
From the basic fraction a small amount of benzylamine
was obtained,
(ii) A sim ilar reaction with 1.3 g . (0.012 mole)
of benzonitrile in 50 ml. of THF and 100 ml, (0.025 mole)
of LiAlH^-THF solution by normal addition at 0-5° showed
again that even at the start the nitrile had completely - 1 5 4 “ disappeared and the reac tio n mixture consisted of amine 57 and a small amount of imine, IK: 6.05 (m), 6.2 (m).
(57) See L. J. Bellamy, loc. c i t . . p. 250 and 255. ks time proceeded, the imine hand disappeared and the reaction seemed to be complete during 100 minutes.
After working up the mixture, the p-nitrobenzoyl deriva tive of benzylamine, m.p. 142-143° (reported^^ m.p. I 4I-
(58) L. Rligheimer, Her. , 49. 596 ( 19 I 6) .
143°), was obtained by treating the basic fraction with p-nitrobenzoyl chloride. AUTOBIOGRAPHY
I, Tadaraichi Fukunaga, was born at Fukuoka, Japan,
March 21, 1931, and received my secondary school education in the public schools of Osaka, Japan, and my undergradu ate training at Osaka University, which granted me the
Bachelor of Science in 1953. From the same University I received the Master of Science degree in 1955. In the autumn of 1955 I entered the Graduate School of The Ohio
State University. While completing the requirements for the degree Doctor of Philosophy, I held in turn the following positions and scholarships; Research Assistant,
1955-1956, sponsored by Wright Air-Development Center in
The Ohio State University Research Foundation; Research
Fellow, I 956 -I 957 , sponsored by Research Corporation in
Department of Chemistry, The Ohio State University; and
Research Fellow, 1957-1959, sponsored by Wright Air-
Development Center in The Ohio State University Research
Foundation.
- 1 5 5 -