gents of this type generally are known by the names oftheir discoverers (Hiinig base,'
or by some other dis-tinctive appellation (Proton Sponge,2
Interest inthis laboratory in such compounds derives:
fromsuccess3 in generating carbenes from diheteroatom sub-stituted carbonium ions [(RX)EH+
when alternative nucleophilic displacement (at R) andaddition (at C+) processes were predominant with otherbases, and
from failure4 in attempts to extend thisreaction to cations whose predicted C-H acidity wasless than that required for attack by the tertiary amines,
MeScH]. The preceding com-munication5 introduces a very strong base, lithium
(LiTMP), whose value ina practical new cyclopropane preparation stems fromthe selective reactivity elucidated above. The failure inthis reaction of other classes of strong bases previouslyrecommended for their excellence as selective protonabstractors6 (including LiR, NaH, KO-t-Bu, NaCPh3,i-PmNMgBr, and the recently acclaimed LiN(SiMe3)z) salso recorded. In this report the relative merit for thesame purpose of structurally similar, potentially eco-nomical LiNR, reagents is examined and further ap-plications in which these bases can be used to prepara-tive advantage are ~utlined.~LiTMP is not the first LiNRz reagent to be used insynthetic chemistry as a strong proton-selective base.Since the early investigations of Levine,s the value oflithium diethyl- and diisopropylamide [LiN(Et)z andLiN(i-Pr)z] in this area has been widely rec~gnized.~More recently the
[LiN-(i-Pr)Cy] has been touted as an improved reagent of thistype for ester condensations.'O In Table
the yield ofthe cyclopropane, 7-phenylnorcarane
with Li-TMP used as the base in its production from benzylchloride and cyclohexene is compared with the samedata for several other LiNRR' species.The variationin yield is easily rationalized with simple steric argu-ments.'Z3 Note that the previously unsung lithium di-cyclohexylamide [L~N(CY)~]s a more effective basethan either LiN(i-Pr)z or LiN(i-Pr)Cy.The ability totranslate this superiority to other published syntheticapplications of the last two basess-'O would have sig-nificant economic consequences, since HN(CY)~s as
Hunig and M. Kiessel,
91, 380 (1958);J.
 5, 224 (1958);
Aldrich trademark; R. W. Alder, P.
Steele. and D. R. Winterman.
P. Marino, and
R. A. Olofson and
R. A. Olofson and C. M. Dougherty,
J. Amer. Chem.
Presented at the
National Meeting of the American Chem-ical Society, New York, N. Y., Aug
see also C. M.Dougherty, Dissertation, The Pennsylvania State University,
G. Stork and L. Maldonado,
93, 5286 (1971);
89, 2500 (1967); 92,1396, 1397 (1970;;
Reiffers, H. Wynberg, and
K. Crandall, L. C. Crawley, D. B. Banks,and
J. Org. Chem.,
G. W. Moersch and A. R.Burkett,
36, 1149 (1971);
P. Thummel and B. Rickborn,ibid.,
36, 365 (1971).(10)
M. W. Rathke and
93, 2318,4605 (1971).(11)
Note also the increased simplicity
dicyclohexyl-amine, HTMP, and analogs in industrial applications.
Lithium Amide Structure on Product Yields
39 14h (64i)
18 66 
54 80 89 [86Ik
Special abbreviations in text.
reaction conditions seetext and ref
satisfactory combustionanalyses and corroborative spectral data were obtained for all newcompounds. Yields are of isolated, pure IV prepared by refluxing,o-chloroanisole, LiNR?, and PhCSCLi
n THF over-night. m-An-NRR' and anisolel5 were major side products withfirst nine bases (0-chloroanisole
pure, vacuum distilled P-keto ester unless noted; pre-pared by addition of LiNR2
equiv) to ester
equiv) in THF at
for VI) followed after
equiv). Major products were isobutyryl- and benzoylpiperidides.
Nmr-tlc analysis only.
PhCONEto was a major product.
Low yield not accounted
by slow decomposition of
withLiN(i-Pr)z in refluxing THF. The yield was cut in half when
base was used in the reaction. Nmr assay; product notisolated pure; distillation fraction also contained i-PrCON(i-Pr)?
The by-product, HN(Cy)?, is removed by extraction withaqueous citrate, not aqueous HCl!
Yield of VI after
the yield after
Reaction period doubled, further yield increases possible.
cheap as HN(i-Pr)z and HN(Et), and much less expen-sive than HN(i-Pr)Cy.From Table
it is evident that RzN- reagents with
attached to two tertiary carbons (here named asH+arpoons12) re best for the production of phenyl car-benoids from ArCH2C1. Although the yield spread inthe four compounds listed is not large, the followingpoints are important in consideration of the broaderand relative synthetic utility of these bases. First, allfour reagents [and L~N(CY)~]re soluble in nonpolar-even hydrocarbon-solvents. The insolubility
many strong bases in such solvents seriously limits theirversatility. Second, reaction rates are very different.While benzyl chloride reacts exothermally withLiTMP, the reaction with lithium tevt-amyl-tevt-butyl-amide [LiN(t-Am)?-Bu] requires several hours in re-fluxing ether. N-(1-Ethylcyclohexy1)- 1,1,3,3-tetra-methylbutylamine is
hindered that formation of itslithium salt (LiETA) with methyllithium in ether re-quires
hr. With all other HNR, substrates inTable
this reaction is a titration. From these ob-servations it is anticipated that LiN(t-Am)t-Bu andLiETA will be especially useful reagents in reactions ofproton acids more acidic than benzyl chloride in whichit is necessary for the base to have a higher basicity tonucleophilicity ratio than that identified with LiTMP.A third factor is base cost." While LiTMP is poten-tially inexpensive,j no practical synthesis of HN(t-Bu),is yet known.I3 Both HN(t-Am)t-Bu14 and HETA
stress the pKdifference