ASYMMETRIC INDUCTION IN ADDITIONS
T0 SOME ASYMMETRIC ALDEHYDES
AND KETONES

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
RICHARD NELSON NIPE
1970

 

  

L I312 A R Y
Michigan Sum
University

 
     

frfi-H‘S

This is to certify that the
thesis entitled

ASYMMETRIC INDUCTION IN ADDITIONS TO SOME
ASYMMETRIC ALDEHYDES AND KETONES

presented by
Richard Nelson Nipe
has been accepted towards fulfillment

of the requirements for

Ph .1). degree in ChemiStry

(f/Cae/Méq/

Major professor

 

Date 5-15-70

 

0-169

     
    

‘ ummua

.' nuns
mm mm gag .

. ‘ LIBRARY BIND

  
 

a. 3m "

  

’- I“ mere-LINN:

 

ABSTRACT

ASYMMETRIC INDUCTION IN ADDITIONS TO SOME
ASYMMETRIC ALDEHYDES AND KETONES

BY

Richard Nelson Nipe

The model for asymmetric induction proposed by
Karabatsos (1) predicts not only the major diastereomer,
but also the diastereomeric ratio as well. The predictions
from the model are based on the interactions of the carbonyl
oxygen with the large and medium groups on the asymmetric
center. .To test the validity of this model under a variety
of conditions (nucleophile, temperature, and solvent), the
reactions of 2-methylbutanal-2-g; 2,5,S-trimethylbutanal-Z—g;
5-methyl-2-pentanone-1,1,1,5-QA; and 5,4,4-trimethyl-2—
pentanone-1,1,1-g3 were carried out with lithium aluminum
deuteride and lithium aluminum hydride in ether, sodium
borodeuteride in tetrahydrofuran and isoPropyl alcohol,
methyllithium in ether, methylmagnesium iodide and bromide
in ether, and methylmagnesium.chloride in tetrahydrofuran.
The ratios of the isotopically related diastereomers were
determined by integration of their nmr spectra. The ratios

of the 5,4,4-trimethy1-2-pentanols (do and Z-gg) were

Richard Nelson Nipe

determined by separation on a six foot Carbowax column and
integration of the peak areas.
In contrast to earlier data, the values of AAG:B calcu-

lated according to the Curtin-Hammett principle (2) were not

AAG*

AB = —RT1n(A/B)

constant with temperature. Accordingly, values for AAH:B
were calculated from plots of 1nAéB-versus 1/T and were
found to agree with the values calculated by the Karabatsos
model. The AASiB was shown to control the stereOSpecificity
of the reactions.

The conclusions drawn from this study are: 1) the
stereospecificity in reactions of 1,2-asymmetric inductions
can be controlled by AAHiB or AAS:B7 2) the AAH:B predicted
by the model and those found experimentally agree well;

3) stereospecificity depends on the attacking Species,
although its exact dependence is difficult to predict;

4) the increase in effective steric bulk of the attacking

reagent increases stereOSpecificity.

REFERENCES

(1) G. H. Karabatsos, J.4Amer. Chem. Soc.,, 9, 1367 (1967).

(2) D. Y. Curtin, Record Chem. grogr., ;§, 111 (1954).

ASYMMETRIC INDUCTION IN ADDITIONS TO SOME

ASYMMETRIC ALDEHYDES AND KETONES

BY

Richard Nelson Nipe

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1970

To Sue

ii

ACKNOWLEDGMENTS

The author wishes to thank Professor G. J. Karabatsos
for his guidance and patience during this investigation.

-Financial support, given by the Dow Chemical Company
for the Dow Summer Fellowship (1967) and the Phillips
Petroleum Company for the Phillips Summer Fellowship (1969),
is gratefully acknowledged. Financial assistance provided
by the National Institutes of Health was also greatly

appreciated.

In addition, the author would like to thank his col-
leagues and faculty members for their assistance and sugges-
tions. He is especially indebted to Dr. J. W. Patterson and

Mr. G. P. T. Nilles for their help with synthetic methods.

 

iii

TABLE OF CONTENTS

 

Page
LIST OF TABLES. . . . . . . . . . . . . . . . . . . . vi
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . Vii
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . 1
RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . 11
A. Errors Involved in Calculations of Thermo—
dynamic Parameters. . . . . . . . . . . . . . 11
B. The Relation Between Activation Parameters
and Stereospecificity . . . . . . . . . . . . 21
C. The Contributions of Nucleophile and Solvent. 25
D. Conclusion. . . . . . . . . . . . . . . . . . 25
EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . 26
~A. General Procedures. . . . . . . . . . . . . . 26
B. Preparation of Z-Methylbutanal-Z-g, . . . . . 27
C. Nmr Studies of Z-Methyl-l-butanol-Z,O-dg. . . 28
D. Additions to Z—Methyl-i-butanol-l,27g;. . . . 28
E. Nmr Solvent Studies of 2,3-Dimethyl-2-pent-
anol. . . . . . . . . . . . . . . . . . . . . 29
F. Preparation of SeMethy1-27pentanone-
1,1,1,53g5. . . . . . . . . . . . . . . . . . 30
G. Additions to SeMethyl-Z-pentanone-i,1,1,5-QA. 51
H. Preparation of 2,5,5-Trimethylbutanoic Acid . 52

iv

TABLE OF CONTENTS--continued

J.

K.

L.

M.

N.
O.
P.
Q.

R.

Preparation of 2,3,5-Trimethylbutyry1
Chloride. . . . .~. . . . . . . . . . . . .

Preparation of S,4,4-Trimethylk2-pentanone.

Preparation of 5,4,4-Trimethyl—2-pentanone-
1'1'1-Q3. o o o o o o o o o o o o o o o o 0

Additions to 3,4,4-Trimethyl-2-pentanone. .

Additions to 5,4,4-Trimethyl-2-pentanone-
1,1,1jg3. . . . . . . . . . . . . . . . . .

Preparation of.2,3,S-Trimethyl-l-butanol. .
Preparation of 2,3,5-Trimethylbutanal . . .
Preparation of 2,3,5-Trimethylbutanal-27g_.

Additions to 2,5,5-Trimethylbutanal . . . .

~Additions to 2,5,5-Trimethylbutanal-2—d_. .

BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . .

Page

55
54

55
55

56
36
38
4O
41
41
44

LIST OF TABLES

Table Page

I. Additions to Z-Methylbutanal and ZeMethyl-
butanal-Z—g, . . . . . . . . . . . . . . . . . . 12

II. -Additions to S-Methyl-Z-pentanone and S-Methyl-
2-pentanone-1,1,1,5—§A . . . . . . . . . . . . . 13

III. -Additions to 5,4,4-Trimethy1pentanone and 5,4,4—
Trimethylpentanone-i,1,1-g3. . . . . . . . . . . 14

IV. ‘Additions to 2,3,5-Trimethylbutanal and 2,5,5-
Trimethylbutanal-Z-g_. . . . . . . . . . . . . . .15

V. The Activation Parameters of the Reactions in
Tables I-IV. . . . . . . . . . . . . . . . . . . .16

VI. List of Solvents Used to Resolve the Diastereo-
topic Methyl Groups bf 2,3-Dimethyl-2—pentanol . 50

vi

Figure

LIST OF FIGURES

The nmr Spectrum of 50% solutions of 2,5,5-
trimethyl-l-butanol in acidic methylene
chloride. . . . . . . . . . . . . . . . . . .

The neat nmr spectrum of 2,5,3-trimethyl—
butanal . . . . . . . . . . . . . . . . . . .

The nmr spectrum of 2,3,S-trimethyl—l-butanol-
1,2-g2 obtained in the reaction of 2,3,5-tri-
metBbeutanal-Z-d_with sodium borodeuteride at
-70 in tetrahydrofuran . . . . . . . . . . .

vii

Page

57

59

42

INTRODUCTION

The recognition that the arrangement of atoms in space
is three dimensional rather than two dimensional represented
a major change in the theory of molecular structure. J. H.
van't Hoff (1), who advanced this concept in 1875, explained
the differences between isomeric compounds which are identi-
cal by elemental analysis and react in a similar manner, but
are different in their interactions with plane polarized
light. This configurational theory proposed that a molecule
can be defined by stating not only the number of covalent
bonds by which the atoms are attached, but also stating the
geometrical arrangement of the atoms. .This theory can be
partially justified by the fact that all observations of
science to date are consistent with it.

A second important concept used in the development of
the theory of asymmetric induction was steric bulk. Both
HOfmann (2), who reported that no reaction occurred between
methyl iodide and dimethylmesitylamine and Meyer (5), who
reported that steric hindrance retarded the esterification
and hydrolyse of ggthgfdisubstituted benzoic acid showed
how the outcome of a reaction can be determined by the

spacial arrangements of the reactants. A further development

involved showing that restricted rotation in biphenyls could
result in optical resolution. Both Kenner g£_§l, (4) and
Turner gt‘al, (5) showed that 6,6'-dinitrobiphenic acid

could be resolved into enantiomers. These concepts of the
conformational effects and steric interactions form the

basis of the arguments used to interpret the observed results
of asymmetric induction.

Asymmetric induction represents any reaction in which
unequal amounts of two possible enantiomers or diastereomers
are formed due to asymmetry present in reactant, reagent,
catalysts, or solvent molecules. In 1952, Cram and Elhafez

(6) introduced a model for asymmetric induction shoWn below.

 

 

 

 

"cram's model" for asymmetric induction was derived empiric-
ally, based on the role of steric interactions in determin-
ing the major diastereOmeric product. The argument used

was that the carbonyl oxygen when complexed with a metallic
atom of the attacking reagent, became the largest group in

the molecule. .Therefore, it would be in the most stable

conformation when placed between the small and medium groups.
The attack of the reagent to give the major isomer A would

be past the small group, s, and the attack to give the minor
product, past the medium group, M. This model is thus based
on the assumption that the magnitude of the steric inter-
actions between the incoming group and the 3 group and those
between the incoming group and the M group determine the
product ratio. ,This rule is applied to kinetically controlled
reactions in which the groups attached to the asymmetric
center do not complex with the attacking reagent. -The
kinetic versus thermodynamic control of these reactions has
led to the coining of two terms (7): "steric approach
control" and "product development control". These terms have
been used to describe additions to cyclohexanone systems.
They represent the two possible extremes on the reaction
coordinate: first, a transition state that resembles the
starting materials with little change in bonding orders, and
second, a transition state in which extensive bond breaking
has occurred.

A simple empirical model, proposed by Karabatsos (8),
predicts not only the major diastereomer in 1,2-asymmetric
induction, but also the semiquantitative ratio of the dia-
stereomers (A/B). In this model, which correctly focuses
attention to transition states rather than ground states,
several assumptions are made: a) little bond making or break-

ing has occurred in the diastereomeric transition states,

b) the two low energy transition states that determine
product stereospecificity have the smallest group adjacent
to the incoming, reagent, and c) the diastereomeric ratios
can.be evaluated from the relative magnitudes of the calcu-
lated interactions between the carbonyl group and the group
that eclipses it at the asymmetric carbon atom.

Considering the first assumption, one concludes that
the transition states in the additions to aldehydes and.
ketones would resemble the ground state conformations found
by mmr. The conformation having an alkyl or aryl group
eclipsing the carbonyl oxygen was found to be more stable
than that having a hydrogen eclipsing the carbonyl group (9).
Three possible transition states, I-III, can be drawn lead-

ing to one diastereomeric product, A:

 

 

 

O

I

I

I

s 'M

I

R R
I II III

Referring to the second assumption, the approach of the
reagent would be less hindered while moving past the small

group than past either the medium or large group, i,§,,

structure I would be favored over either II or III as the
low energy transition state. Other interactions which would
contribute to the greater stability of I versus II or III
are'R' "->S and R "->S interactions in I versus R' ”M and
R ”M in II and the much larger R' ”Land R 9L inter-
actions in III. .The non-bonded interactions between carbonyl
oxygen and the eclipsing group, as mentioned in assumption c,
would also favor I over II and III. For example, when s is
hydrogen, M is methyl, and L is ethyl, I is favored over II
by -100 cal/mole and over III by -800 cal/mole.

These same arguments can be applied to predict the low
energy transition state leading to the other diastereomeric

product, B. 0f the three structures, IV-VI, structure V

 

9 <2 .0
I M I L I S
l I I
I
R‘-- L-- R'—-- ~-
L s s M M L
R R R
IV V VI

would be the low energy transition state leading to the
other diastereomer.
The next step was to determine which of these two low

energy diastereomeric transition states led to the major

and which to the minor product.

 

I V

The interactions of R‘ “s, R 9 s, and R' #R would cancel
each other in both I and V; to a first approximation, R'*+'L
and R"‘9'M in I might counterbalance the corresponding
interactions RQM and R' 4+ L in V, thus leaving 0 9M
and O *a'L as the interactions from which one might calculate
the relative stabilities of I and V. The evaluation of these
interactions is available from nmr studies on carbonyl com-
pounds (9) and therefore allows a semiquantitative predic-
tion of the magnitude of the diastereomeric product ratio.
The difference in the free energy of the diastereomeric
transition states is related to the diastereomeric ratio by

the following equation:

* =_
AAGAB RTlnA/B

This equation represents an application of the Curtin-Hammett

principle (10) to these systems. The results found in the

literature correlated well with the calculated results from
this model (8).

.The model as just explained seems to offer the best
starting point for further investigations of asymmetric
induction. There are several factors which have not been
evaluated and which can affect the success of the model:

1) the differences in free energy of the two diastereotopic
transition states due to entropy effects of the two transi—
tion states, 2) the degree of bond making and breaking which
marks the transition states, 5) the differences in the
gauche nonbonded interactions in the transition states, and
4) the differences in solvation energies of the diastereomeric
transition states.

In planning an experiment on asymmetric induction, the
effects of these four unknown factors should be minimized
or hopefully be amenable to evaluation. (In any system, the
differences in entropy of a given reaction can be found from
the changes in the free energy of the diastereomeric transi—
tion states with temperature. Secondly, the degree of bond
making and breaking has been suggested previously to explain
the product distribution in acyclic (11), monocyclic (7),
and bicyclic (12) ketones. To determine the extent of bond
making and breaking in the relevant diastereotopic transition
states, experiments should be chosen in which the conforma-
tional energy differences of the reactants and the energy

differences of the diastereomeric products are known (13).

Since the differences in energies of most diastereomers are
not known, a system must be chosen in which these differ-
ences are known.§_priori. The use of isotopically related
groups would reduce the differences in energy of diastereo-
meric products to a minimum because these would be due only
to secondary isotope effects. Since the energies of the
two products would be equal, the diastereomeric product
ratio expected with extensive bond breaking and making would
approach unity. ‘At the other extreme, if very little bond
making or breaking has occurred at the transition states,
the diastereomeric ratio would reflect the differences in
conformational energy of the carbonyl substrates.

Thirdly, differences in conformational energy in the
transition states are partially due to nonbonded interactions
between groups attached to the carbonyl carbon, the incoming
reagent, and the carbonyl oxygen. Since the gauche inter-
actions of the diastereomeric transition states cannot be
estimated exactly, the differences should be minimized in
the cases where the isotopically related diastereomers are

formed.

 

 

In I, the gauche interactions are R' 44‘s, R 9 s,.R' ”M,
and RQL, and in V, R' 4+3, R (+3, RQM, and R' 6L.
Since R and R' are isot0pically related, the interactions
(save that R' is bound to a metal atom) are similar and
could lead to a very small energy difference. The main
energy difference between I and V would thus arise from.the
nonbonded interactions of M'*9'Q_versust'*€>O.

The final factor, i,§,, the contribution of solvation
to the energies of the two diastereomeric transition states,
is expected to be quite complex. It is uncertain what types
of interactions the solvent has with all species present.
.However, judging from cases of asymmetric induction in
optically active solvents, one may state that the contribu-
tion of solvent to the differences in the free energies of
the two diastereomeric transition states is probably small
and of the order of 50-100 cal/mole (14).

In several examples in the literature the effect of
reagents on the stereospecificity of reactions has been
reported (15,16). In these experiments, the stereospecificity
varies with the halogen atom of the Grignard reagent and with
the alkyllithium reagents. This effect has been studied in
a systematic way by Karabatsos and Althuis (15). Some of
their results are given in the discussion section in Table V.

In continuing the work on asymmetric induction, the pur—
pose of this study was to consider the effects of tempera-

ture, reagent, and solvent on the stereOSpecificity of

10

nucleophilic additions to 2-methylbutan01-2-g;, 5~methyl-
2-pentanone-1,1,1,5-§A, 2,3,S-trimethylbutanal-Z-g;, and
5,4,4—trimethyl-2-pentanone-1,1,1-g3. Since the diastereo-
topic groups are magnetically nonequivalent, the analysis of

the diastereomeric alcohols was carried out by using nmr (16).

RESULTS AND DISCUSSION

The results of the reactions of lithium aluminum deu-
teride, sodium borodeuteride, methyllithium, methylmagnesium
iodide, bromide, and chloride with 2-methylbutanal-2-g;,
Semethyl-Z-pentanone-l,1,1,S-QA, 2,5,S-trimethylbutanal-Z-g;,
and 3,4,4-trimethyl-2-pentanone-1,1,1-g3 are summarized in
Tables I-IV respectively. Also included in these tables
are literature results from reduction or addition reactions
to similar compounds. Table V includes results from the
calculations of the AAGAB’ AAHiB,
hydride and methyl Grignard additions to the aldehydes and

and AAS:B for the metal

ketones listed above.

~A. Errors Involved in Calculatigns
of Thermodynamic Parameters

The errors in the thermodynamic parameters stem from
the errors inherent in the calculation of the percentages of
isomers A and B. Since the errors in the A/B ratios were
taken as i1%, a ratio A/B of 60/40 would not exceed the range
of 61 to 59 for A and 39 to 41 for B. These errors were in
all cases larger than the precision of the measurements made

by nmr integration or by cutting and weighing either nmr or

11

12

.md mocwnmmmu ca pcsom on cue must mmmcao

md
o
deq can
.mcoflusaom Hocooam-H>monmomw &om ca e om.m um m xmmm can 9 «m.m um d xmmm mo coflumummucfl
ecu wn Umcflaumumo mums «firm.HIHocmusnIalawcumEIm uwuoaooumummflc on» mo m\¢ moflumu exam

.ma0£\amo ca cm>Hm mum mosam>

 

 

Hmu
.omm- o.oe\o.0m I- Hmnum oo UHmm2mz omumwmmommo
m H .oma m.HH\H.mm 0m cos-
OH H .OOHI m.mw\H.Hm om Hosooam oo
mH H .ommn N.Ns\m.sm mm HaaoumomH omm Hnmmz.
OH H .OOHI m.mH\H.mm mm ooh-
ma H .OHNI m.oe\s.mm mm amuse oo
om H .oam- m.mm\H.Hm om -ouessmHHms 0mm «nmmz
e H .om, H.Hm\m.me ms ooh- emu
m H .oe: H.mw\m.Hm om oo .
OH H .omu «.mw\m.mm mm Heaps own «nach omooommonmo
newe<< mm\¢ oHHmm eHme.e Hcm>Hom .esma mHHnmomHosz mHmHHmnsm

 

 

amrmuamcmusnasnumzam ecm Hmcmusnamnumzrm 0» mcoHHHeeeru.H manna

15

.md mocmuommn ca venom on can must mmmcau

.0H0£\Hmo cw cm>wm mum mmsHm> mMo<< one

Q
.mcowu

Isaom mcflcflummwfiom ca p om.m um m xmmm cam e am.m um é xmmm mo coflumummpcw on» >9 cwcfle
Inoumc mHmB tmlm.d.a.aImHocmucmmImlamcuoEHcIm.N Ufiumeomumummac on» mo m\< mowumu mcam

 

 

.mHI m.me\m.am II can mm
.mm m.Hm\m.mH I- 00 w
.om o.Hm\o.mH II Hmnum own usmHqu mmowmommommo
ma H .omH- o.mH\o.mm mm cos- 0
mH H .omHI m.me\m.sm om oo .
mH H .OFHI H.mH\m.mm ms HmsHm 0mm Haw:
a H .om- m.sfl\m.mm ms oosu
OH H .om- m.mw\e.em Op 00
s H .omu m.mw\s.am mm amusmouessmuuma 0mm Homzms.
m H .om- o.sw\o.mm om 00»:
OH H .OHHI H.mw\m.em mm 00
OH H .OOHI m.m¢\m.em om Heaps 0mm Hmmzez
a H .omu H.>w\m.mm om cos: emu
OH H .omu H.m«\m.mm om oo _
oH H .OHHI H.me\m.em mm Hmeum 0mm Hmzmz moowoommommo
o
n mmoqq mmxc oHHmm eHmHm.& Hcm>aomr .msma mHHnmomHosz mumuumnsm

 

 

.{WIm.H.d.almcocmucmmlmlawcpmzrm cam mcocmucmmINIamcuszm Du mcofiufivc<II.HH manna

14

ecu mo m\< moflumn one
mafia Goducmumu m Um: d xmwm

...wao£\amo CH cw>fim mum mmsam>

m4
*QQQ 0389

.e Nm.m um m HoEomfl 0cm Amsoum Hmcumalmv v 05.0 pm a HmEOmH Momma mcu
mo cowumnmmucw >9 cwcfleuwumc who? nmId.d.anHocmuchINIawcumfimuumule.w.m.m UHHmEoouoummflp

.mwuscwa m.m mo mEHu coflucmumu m m xmmm Ucm mmuscwe o.¢ mo

.cecaou xmBonumU Doom xflm m co mxmmm 00> mg» no cowumumwucfl
an pmcflaumumc ouo3 cINIHocmucmmINlamcumeHuId.¢.m owuweoonmummflc ecu mo m\¢ mowumu mcam

 

 

oea H .OOOH- >.me\m.mm om 00

on H .one- H.mm\m.mm mm Hmer 0mm meme:

oh H .0mm- N.@H\m.mm os ooh-

mm H .omm- m.em\>.ms ms 00

mm H .ome m.mm\>.ms mm Hmrum 0mm Hams.

OH H .OAmI o.mm\o.es Op 00

me H .ommu o.>m\H.ms om Hmnum 0mm HmZmz

om H .omm: m.mm\>.m> om 00

mm H .onmu H.mm\m.m> mm Henum own Hmmzez
0

me H .Ommu H.mm\m.es mm 00 mm

OH H .omm- s.mm\n.oh mm :mHsHoHesanHma 0mm Humzwz «commoonxmmov

OH H .om- m.mm\H.sH mm Honouam oo». o

om H .oom m.ae\¢.mm mm HmmoumomH com ddents

OH H .os- m.Hm\H.mH 0H 00

OH H .om H.me\m.am om :mHsHOHesgmHHma 0mm «cmmz.

me H .OmHI m.Hm\m.mm mm 005: are

OH H .oe ~.m§\m.Hm mm 00 .

OH H .OOH m.mH\H.Hm om Hmnum own «QHHHA mmommoooxnmov
o

A mMWQQ m\¢ Caumm pamflm.& uco>aom .mEmB ; mawnmomausz. wumuumnsm

 

.nmrfi.d.HImcocmucmmahcumEfiuele.e.m ccm mcocmucomawcuofiwualw.w.m ou mc0auH6c<II.HHH magma

.oao£\amo ca co>Hm mum mosam>

mueHH was

.mmuscHE m.m mo wEHu coHucmumH m on: m xmmm HocHE ms» Ucm.mmuccHE o.¢ mo mEHu cowu

Icwumu m cm: d xmmm momma one

.cEsHoo xmBOQHmo Doom wa m co pmcflsuno mxmmm om> mo coHu
Imumouca ecu an pmcflfiumumc muw3 mHocmucmmINIawcumfifiqud.¢.m UHHoEomeumch ecu mo moHumn

Q

 

 

one .e mm.o um m Mama HocHE ecu 0cm H m>.m um.¢ xmmm Momma mnu Qua? HE: may no coflumum
Imucw may >2 UocHEumumc mum3 maocmusnlaIahcumEHqum.m.N Uflumeomnmummflc may no mowuwu mcam

om H .omsu o.«fi\o.mm om ooh:

om H .o~>I e.mfi\m.om 0m 00

omH H .ommHI m.0H\H.mm mm Hmeum 0mm Hmzez
oom H..o>mHI H.¢\m.mm mm 00»:
omH H..O>mHI e.>\m.mm mm 00
0mm H .ommHu m.m\¢.mm mm Hmnum 0mm HmmZez
omH H .OOHHI H.m\m.mm mm ooh: are

”m oom H .037 >.m\m.mm om oo _

omH H .ommHI m.mm\H.ss om cmHsHouemanHma mm Homzmz omomooxmmov
oe H .omeu m.wm\m.m> mm : ooh:

oe H .omen m.mm\aqas om Hozooam oo

om H .ommu m.mh\m.0m mm HaeonOHH 0mm «nmmz

me H .omsu o.mm\o.>> om 00s:

ma H .ommu o.oe\o.0m om oo

OH H .OHH- H.mw\m.mm ms :mH9H0HesanHme owe Hommz

mm H .ommn o.em\o.m> ms 00 _

om H .ONmI >.mN\m.n> Om Hmsum 0mm eaadwq omoaoomammov
QmMUQQ mm\< owpmm Gamay R ucm>aom .mEmB maflcmomaosz mumuumnsm

 

 

amrmuamcmuanmnuwsHHeum.m.m camAHnsmuanssumaHHaum.m.m oH mcoHHHneerI.>H magma

16

 

UoDCHucoo
m.mu .mHI 00s: nmo
H.m H m.m- «.mu owe H .OHmu .mm 00 Aumrumc w
H.mn .wm 0mm «mH¢Hn aroma «moose
o.mI 3 H .03.. co?
H.N H m.H- m.HI OHH H .owmu mH H .omHs oo Aumsumo
m.HI ma H .OHHI 0mm Haws
o.mu s H .omu cow:
H.m H m.HI H.Hn OHH H .OQHI OH H .omu oo Lasso
m.Hn a H .omn 0mm House:
m.Hu m H .omu ooh-
H.m H H.HI N.HI ose H .omH- OH H .OHHI oo AHmnHmv
H.HI OH H .OOHI 0mm Hmmzmz
H.HI s H .omu 00s: are
H.m H m.HI m.HI OHH H .ooeu OH H .omu oo Aumnumv
o.HI oH H .OHHI 0mm Hmzmz anerWNmommo
H.HI m H .om- o»- o
m.m H H.Hu N.HI omm H_.omHI OH H .OOHI oo AmoumIHo
m.¢- mH H .omm- omm «nmmz
m.HI OH H .OOHI 00s:
H.m H 5.0: mflou 00H H .Ommn mH H .OHNI oo Lasso
H.o- om H .OHm- 0mm «nmmz
“VoHI ¢ H oON OOFI , ”EU
H.N H o.HI m.on OHH H,.0smu m H .oeu oo Lumeuuv _
m.ou OH H .Omu 0mm «nH¢Hq omonowmommo
md Aucm>aomv
mmmmmmm mMm<< QMMW<< mmMWQQ .meB mHHcmowHosz mumuumnsm

 

 

.>HIH moanms cw mcofipommm mcu mo mHonEmumm coHum>Huo< onBII.> manna

17

Umscfiucou

 

H.m H H.Hu H.mu own H .omoHI 0H H..omu oh: Amoumqu
H.HI om H..oom oom Hommz
m.w H m.m- s.mu 0mm H .omOHu OH H .oen oo Amuse
n.mu OH H .om mm «cmmz
m.mu mH H .omHu own use
H.m H m.HI H.HI OHH H .ommHI OH H .OH 00 Aumsuma _
H.HI OH H .OOH 0mm «QHHHH mmommooxmmoo
m.mu om H_.onsn ooh:
H.m H H.HI >.HI owe H .oemHn om H .OHHI oo Aumsumo
H.o omH H .ommHI 0mm Hmzoz.
m.o oom H..0>NHI can
H.m H ~.H o.H One H .OOHHI omH H..0smHI oo Aumnumv
>.H omm H .ommHu 0mm Hmmzmz
o.H omH H .OOHHI can as
H.m H m.H m.H ooe H .oomu oom H .onHH- oo Amway H
o.m omH H .ommHu owe Homzmz omom Oaxmmov
p.mu 0H H..omHI can
m.m H o.mu m.mu omm H .OHNHI oe H..omen oo AmOHmIHV
m.mu om H .ommu 0mm «ommz.
m.mu mH H .omeu op:
H.m H m.HI o.mu 00H H .ommHu ma H .ommn oo Ammav
m.eu OH H .OHHI 0mm Hdares
«.mu mOH H .oesu ooh: «mo
H.m H m.mu m.mn OHH H ommHI mm H .ommu oo Aumnumv _
m.mu on H .ommu 0mm «caqu omoooomxmmoo
m<
m<< _ Aucm>aomo
mmmmm>< mMmQQ nmmw<< mdmdd .mema maflcmomausz mumnpmnsm

 

 

A.e.uaouvuu.> manna

18

 

 

.HWIHJ“.HumaocmucmmlNlamcumEmuumuIdfififi no“ mMmdeHo mmsHMS may Sufi: momma Hoc
mwOU QMEQQ HON UTDMHDUHMO ODHM> 05”.. mm uUwHHOUCH was 0U mfiwwm #03 HO mODHM> O3”... 05“ MO OGOU
.HmHOE wmummc\amug .D.m cH cm>Hm mum mmDHm> MMm<< mnao
.wH0£\Hmo CH cm>Hm mum mmsHm> mMW<< usan
.mHOEVHmo CH cm>Hm mum mmcam> mMU<< mafia
omm H .ooomu oea H .OOOHI oo Auosumv
e on H .03: 0mm mamas.
m.mI Os H .ommn ooh:
H.m H «.m- m.mu owe H .OHNH: mm H .omm: oo AumnHmv
o.mI mm H .onmn 0mm Ham:
m.m H >.mI o.¢I 0mm H .owmdl OH H..o>mI o Aumcumv
H.mu me H .omm- mmm Haze:
m.m H m.mn >.mu omm H .omeI om H .omou oo Aumnumv
m.mu mm H .omwu 0mm Hmmsmz
m
m.m H H.N- m.m- oom H .ommH- me H .omm- oo Anzac mu
mHn OH H .03: com Homzez «newswofnmoo
mug/H m4 m2 me 35303
ommuo>¢ Hm<< Q H344 m Heed .mEmB mHHcmomHosz oumnumnzm

 

 

A.©.uGOUvII.> magma

19

Vpc jpeaks- The error in calculating the differences in
enthalpies of activation was determined (17) by using expression

1. .The symbols in the expression are defined as follows:

T'T

5 = zinuT

d (1)

6 is the error in AAHI in cal/mole, R is the universal gas

.AB
constant, T' and T are-the temperatures at the extreme of
the temperature range used in the calculations for AAHI ,
and a represents the maximum fractional error in the ratios.
The maximum fractional error for all data was assumed to be
0.1. .This estimate includes an error in the AAGinue to
:20 in the temperature.

The formula for the error in the differences in the
entropies of activation for the diastereomeric transition
states is given below:

o=é[%+§.?$] <2)

I

 

In this expression, 6 is defined as the error in the entropy
of activation; 6, the error in the enthalpy of activation
as calculated in expression (1); and T and T', the temperature
extremes of the measurement of the diastereomeric ratios.
The results of these calculations are given in Table V.

The errors generally observed in the literature in
similar cases would indicate that the errors in vpc and ir
analysis are of the same magnitude as the values found for

nmr. .Nmr and VpC analysis for the diastereomeric alcohols

20

from the reduction of 3,4,4-trimethyl-2-pentanone with

lithium aluminum deuteride‘at 550 agreed within experimental
error. The value of AAGAB found by nmr integrations was
+95 1 9 cal/mole whereas the value of AAeiB

+100.i 10 cal/mole. Similar agreement between vpc and nmr

found by vpc was

analysis was also found in another system by Karabatsos and
Zioudrou (18). Because many investigators ignore the temper—
ature dependence of the ratios, it is difficult to compare
the diastereomeric ratios found in the literature with ours.
In most cases reported in the literature the reaction
temperatures were estimated from the description of the ex-
perimental procedure.

The absolute configuration of the products formed in
the reduction and addition reactions studied in this work has
not been proven. Previous results in the literature for
which the absolute configuration (5,19) of the diastereomers
has been proven, find the major product to be consistent with
that predicted by the Karabatsos model. In the case of the
reduction of 2,5,3-trimethylbutanal with lithium aluminum
deuteride, the nmr of the resultant diastereomeric alcohols
proves that the major diastereomer is indeed the one pre-
dicted by the Karabatsos model. Figures A and B, on the
following page, represent the most stable conformations of,
respectively, the major and minor diastereomers predicted by
the model. The proton labeled H on the major diastereomer

A
A of 2,5,S-trimethyl—l-butanol-l,2-g2 would give a different

21

 

OH ' OH
Me D2 M D2
. ) I.
3A ¢ D1 D1 HB
A
t-Bu t-Bu
A B

nmr Splitting pattern than the proton, H on the minor

Bl
diastereomer B. In the major isomer A, the coupling con-
and J

stants, would be expected to be approxi-

JHun 1 H-Dz '
mately equal and of the order of 1.5 Hz. This would give a
quintet in the nmr. In the minor diastereomer, the dihedral
angle (0 B) between H and D2 is approximately 60°. On the
basis of the Karplus relationship (20) between the coupling
constant and the dihedral angle, one would predict a much
smaller coupling constant for JH-D2 than 1.5 Hz. -The absorp—
tion signal for the minor stereoisomer (HE) in the nmr is
quite broad whereas that for the major diastereomer (HA)
appears as a well resolved quintet. Thus, at least in this
case, the major diastereomer has the absolute configuration
predicted by the model.

B. The Relation Between‘Activation Parameters
and Stereospecificity

The Karabatsos model applies only to cases where AAG:B

is controlled solely by the differences in the enthalpies

22

¢

of activation (AAHA In cases where AASt contributes

B)’ AB

significantly to AAGAB’ the model should not be valid.

In the twenty reactions studied by Karabatsos and Althuis
(15,21), only one system showed an appreciable temperature
dependence of the AAG:B
This led to the conclusion.that AAS:B contributions in

that was outside experimental error.

those systems were negligible.

The values of AAHAB were calculated for the two systems
studied by Karabatsos and Althuis. In the first system with
phenyl, methyl, and hydrogen on the asymmetric center, the
average AAH:B value was -1400 i 500 cal/mole and in the
second system with phenyl, iSOpropyl, and hydrogen on the
asymmetric center, the average AAHAB was -1550 i 500 cal/mole.
.The value of -1400 i 500 cal/mole for AAHiB, as well as the

experimental AAGI value, in the first system is actually

AB
larger than the AAHAB value of -600 cal/mole calculated by
the model. The same is true for the AAH:B values in the
second system where the average AAH:B is —1550 i 500 cal/mole
compared to the calculated value of —200 cal/mole. It should
be emphasized, however, that many of these AAHiB values were
calculated from results obtained only at two temperatures.
There appears to be better agreement between AABAB cal-
culated for the model and the actual AAHAB as found in the
two systems studied in this work. In the first system, with
ethyl, methyl, and hydrogen groups on the asymmetric center,

the average value of AAH: calculated from a plot of ln(éé§)

B

25

versus 1/T was -400 r 400 cal/mole. The value of AAH:B can
be said to agree in Sign and order of magnitude with the
-1OO cal/mole value of AAH:B eXpected from the model. In
the second system, with t_- butyl, methyl, and hydrogen
groups on the asymmetric center, the average AAH:B found
from the same plot was -1250 i 400 cal/mole. This value of
AAHAB agrees well with the value of -1,100 cal/mole calculated
from the model.
Enthalpy control of the reactions at the given tempera-

ture range means that AAG: and AAHAB values agree in Sign.

B
Referring to Table V, it is obvious that this is not always
so. -A small difference of a few entropy units could change
the values of AAG:B from a negative value (AAH:B control)

to a positive value (AA8: control). This positive value

B

for AAG: would represent the predominant formation of the

B
diastereomer opposite to that predicted by the model. This
is evident in several entries in the tables. The product
stereOSpecificity is thus controlled by the differences
either in the enthalpy or in the entropy of activation.

-C. The Contgibutions of Nucleoghile
and Solvent

From the study of the reagent as a combination nucleophile
and solvent system, the following general observations can be
made. The bulkier the nucleophile, the higher the stereo-

specificity. Dauben (7) explained the differences in product

ratios obtained from sodium borohydride and lithium aluminum

24

hydride reductions of cyclohexanones by concluding that in
more polar solvents sodium borohydride is effectively a
larger attacking species. One would have anticipated, there-
fore, that sodium borohydride in isopropyl alcohol would be
more stereospecific than lithium aluminum hydride in ether.
.Furthermore, since tetrahydrofuran is a more polar solvent
than ether, the stereosPecificity should have been lower in
the latter solvent. However, such is not the case. It is
true in one case, but not in two other cases (Tables I, III,
and IV). In view of the findings that AAS:B may control the
stereospecificity of many of these reactions, it is not
surprising that such generalizations prove to be inadequate.
The results with methyl Grignard reagents and methyl-
lithium do not follow the trends observed in previous work
(15,21). For example, it was found that methylmagnesium
chloride was more stereospecific than methylmagnesium bromide,
which in turn was more stereospecific than methylmagnesium
iodide. In this work, methylmagnesium bromide was found to
be more stereospecific than methylmagnesium iodide, which in
turn was found to be more stereospecific than methylmagnesium
chloride. It should be pointed Cut that in.the previous werk
(15,21) where the aforementioned stereospecificity was found,
one of the groups attached to the asymmetric carbon atom was

phenyl.

25
D. Conclusion

This study on 1,2-asymmetric induction in reactions of
metal hydrides and methyl Grignard reagents with 2~methyl-
butanal, 2-methyl-2-pentanone, 2,5,5-trimethylbutanal, and
5,4,4-trimethyl-2-pentanone has shown that product stereo-
specificity may be controlled by either AASAB

The role of nucleophile, solvent, and reagent concentration

r
or AAHAB.

is still not well understood. .The halogen of the Grignard

reagent does affect the stereospecificity of some reactions,
but it does so in an unpredictable manner. The correlation
between the AAHAB predicted by the Karabatsos model and the

Mn*

AB found from plots of ln(A%§) versus 1/T was good.

EXPERIMENTAL
A. General Procedures

The general apparatus used in the reduction, Grignard,
and methyllithium reactions consisted of a 50 ml three-
necked flask fitted with mechanical stirrer, condenser,
thermometer, and addition funnel. The reaction temperatures
were maintained at -700 with dry ice and isopropyl alcohol
bath, 00 with ice and water, or 550 with an oil bath.

Prepared solutions of methylmagnesium chloride, methyl-
magnesium bromide, and methyllithium were obtained from
Alpha Inorganics, Inc. The solutions of methylmagnesium
iodide were prepared from methyl iodide obtained from
Aldrich Chemical Company and from "Mallinckrodt" magnesium.
When a solvent other than that supplied in the prepared re—
agents was used, the Grignard reagent was evaporated to dry-
ness under vacuum and the desired solvent was added.

The nmr spectra were recorded on either a Varian A-60
or HA-100 nmr spectrometer and the ratios of the diastereo-
topic protons were determined by cutting and weighing,
measuring with a planimeter, multiplying the measured peak
height times the width at half-peak height, or by integrating

the area with the nmr. The average values obtained were

26

27

rounded off to the nearest percent. The concentrations of
the solutions used in the nmr spectra are expressed as volume

percents. All boiling and melting points are uncorrected.

B. Preparation of 2-Methylbutanal-25g

 

Forty-six grams (0.55 mole) of freshly distilled 2-methyl-
butanal was placed in a 250 ml flask fitted with condenser,
drying tube and a magnetic stirrer. ‘Anhydrous potassium
carbonate was added to 100 ml of 90%Ideuterium oxide until
the pH 11 was reached (using pH paper as an indicator). The
deuterium oxide and aldehyde were then heated at reflux for
eighteen hours. The pH of the solution was checked periodic-
ally and more potassium carbonate was added to maintain the
pH 11. After cooling the reaction mixture to room tempera—
ture, the organic layer was separated from the aqueous layer
and dried with molecular sieves. The organic layer was then
treated successively with 20 ml of 99.9% deuterium oxide and
two portions of 10 ml of 99.9% deuterium oxide. The organic
layer was fractionated and yielded 21 g (45%) of 2~methyl-
butanal-24g boiling at 87-90, n35 = 1.5876. The multiplet
nmr absorption of the undeuterated species centered at 7.7 T
disappeared and the aldehydic proton was reduced to a broad
singlet (T = 0.5). In addition to the broad singlet, the
nmr consisted of a singlet at 8.9 T (5 H), a complex trip-
let centered at 9.06 T (5 H), and a multiplet at 8.4 T

(2 H).

28

C. Nmr Studies of 2-Methyl-1-butanol-2,0-gg

 

ZeMethyl-l-butanol-Z,O-gg was prepared by the reduction
with lithium aluminum hydride (0.4 g, 0.011 mole) in 100 m1
ether of 2-methyl-butanal-2-g (5.0 g, 0.055 mole) over a ten
minute period. The reaction mixture was then stirred for
one hour and after the solution was cooled, 1 ml of deuterium
oxide was added. The resulting precipitate was filtered from
the ether solution of the alcohol which was dried with sodium
sulfate and distilled, yielding 0.7 g (25%).

The spectrum of the alcohol showed an AB system with the
center peaks separated by 0.5 Hz and the two outer bands
separated by 57. Hz. The difference in chemical shifts for
the A and B protons of the AB system was found from the
equation: AéAB = ((1-4)(2-5))§I By using the above values,
AéAB was calculated to be 4.5 Hz. The value measured from
the neat spectrum of the reaction of 2emethyl-butanal-24Q
and lithium aluminum deuteride was 4.4 Hz. The nmr spectrum
of 2dmethyl-1-butanol consisted of the AB system centered at

6.65 T (2 H), a singlet at 9.1 T (5 H), a complex triplet

at 9.0 T (5 H), and a multiplet centered at 8.7 T.

D. ,Additions to 2-Methyl-1-butanol-1,2-g2

 

A solution of 1.0 g (11 mmoles) of 2~methylbutanal in
5 ml ether was added to a stirred solution of 0.15 g (5
mmoles) of lithium aluminum deuteride and 20 ml ether. The

temperature was maintained at 55, O, or -700. The reduction

29

was repeated with sodium borodeuteride in tetrahydrofuran

at 66, O, and -700 and in isopropyl alcohol at 82, 0, and
-700. In each reaction, after stirring for three hours, the
excess deuteride was destroyed by the dropwise addition of
0.5 ml of water and 0.5 ml of 5% sodium hydroxide. The
ether was filtered to remove the granular salts, then passed
over a six foot Carbowax column and collected in a nmr tube
(75 to 95% yields). By using the 20% iSOpropyl alcohol
solutions, the ratios of the diastereomeric products were
obtained on the Varian HA-100. The peaks from the nmr signal
centered at 5.92 T were cut and weighed. These results are
reported in Table I along with other values from the litera-

ture.
.E. 'Nmr Solvent Studies of,2,5-Dimethyl:2-pentanol

The 2,5-dimethyl-2-pentanol was prepared from the addi-
tion of 10 g (0.1 mole) of freshly distilled 5-methyl-2-
pentanone (City Chemical.Corporation No. M1558) to methyl-
magnesium iodide prepared from 5.7 g magnesium in 150 ml
ether and 21.5 g (0.15 mole) of methyl iodide. ~After the
initial reaction had ceased, 100 ml of 10%Iammonium chloride
solution was added. The ether layer was separated, extracted
with two portions of 50 ml water and 50 ml saturated sodium
chloride solution, and dried over anhydrous sodium sulfate.
,The ether was removed by distillation and the residue was

fractionated, yielding 10.1 g (87%) of the product, bp

50

.155-50. .The nmr of this alcohol was determined in the neat

and in sixteen solvents.

Table VI.--List of Solvents Used to Resolve the Diastereotopic
Methyl Groups of 2,5-Dimethyl-2-pentanol.

 

 

Solvent Concentrations

(% by volume)

 

Neat 100
Pyridine 20,10
Dimethylsulfoxide 20,10
Benzonitrile 20,10
Methylene bromide 20,10
Chloroform .20,10
Carbontetrachloride 20,10
Methylene chloride 20
Methylene iodide 20
Benzene 20,10
Toluene 20
Nitrobenzene 20,10
Chlorobenzene 20,10
Indene 20
Thiophene 20
Acetonitrile 20
N-methylpyrrole 20

 

F.97Prepgration o§5fMethvl=2-
pentanone-1,1,1,5-gA

 

A mixture of 5~methyl-2-pentanone (20 g, 0.2 mole) and
25 ml of pH 12 deuterium oxide (95%) was heated at reflux
for fourteen hours. The reaction mixture was cooled and the
ketone was separated and dried with calcium chloride. The
ketone was treated successively with two portions of 25 ml
of 95% deuterium oxide, 15 ml of 99% deuterium oxide, and

with two portions of 5 ml deuterium oxide. The crude

51

5-methyl-2-pentanone-1,1,1,5-g5 was dried over calcium
chloride and distilled, bp 116-79 yielding 12.2 g (59%) of
pure ketone. The nmr consisted of a doublet at 8.97 T (5 H).
a triplet centered at 9.15 T (5 H) and a multiplet centered
at 8.48 T (2 H).

G. .AdditionsYto 5eMethvl-2-
pentanone-1,1,1,5-g4

 

A solution of 1.0 g (8.6 mmoles) of 5—methyl—2-pentanone-
1,1,1,3-g, in 5 ml solvent was added over a ten to fifteen
minute period to a stirred solution of 15 mmoles 0f organo-
metallic reagent diluted with 15 ml ether. The Alpha Inorgan-
ics prepared solutions of organo-metallics used in these
additions were: methylmagnesium iodide in ether, methylmag-
nesium bromide in ether, methylmagnesium chloride in tetra-
hydrofuran, and methyllithium in ether. The reaction mixture
was maintained at the desired temperature with an appropriate
bath. It was allowed to stand for a period of three to eight
hours, then 2 ml of 10% ammonium chloride was added. The
organic layer was filtered from the precipitated inorganic
salts with a sintered glass funnel, and dried with anhydrous
sodium sulfate. .After the solvent was distilled, the residue
was collected from a six foot Carbowax column. The yields
of alcohols estimated from the VpC traces varied from 70% to
95%. The resulting diastereomeric alcohols were analyzed in
20% pyridine solutions on the Varian HA-100. The ratios were

calculated by multiplying the peak height times the width of

52

the peak at half height. The diastereotOpic ratios and

corresponding AAG:B values are summarized in Table II.

H. Preparation of 2,5,5-Trimethvlbutanoic Acid

In this preparation, 100 g (1.02 moles) of 2,5,5-tri-
methyl-l-butene and sodium borohydride (19.4 g, 0.46 mole)
in 500 ml of tetrahydrofuran were placed in a.5 1. three-
necked flask fitted with condenser, mechanical stirrer, and
addition funnel. .The solution was cooled in an ice bath.
Then, boron trifluoride etherate (97 g, 0.68 mole, 86 ml)
was added over two hours. «After the addition was complete,
and the solution warmed to room temperature, stirring was
continued for an additional three hours. .The excess sodium
borohydride was decomposed with 100 ml of water and the
solvent was removed by rotary evaporation. The organoborane
was dissolved in 500 ml of diethyl ether. The oxidation of
the organoborane with a chromic acid solution followed.

The chromic acid solution consisted of 904 g (5.04 moles)
of sodium dichromate-dehydrate and 676 ml of concentrated
sulfuric acid diluted to 2,500 ml with water. ~During the
addition of the chromic acid solution, the reaction mixture
was held between 100 and 200 with a water ice bath. ‘After
the addition was complete, the reaction was stirred for two
hours. The ether layer was separated and the aqueous solu-
tion was successively extracted with three portions of 700

ml of ether. ,The combined ether extracts (about 2 l.) were

55

washed with 500 ml of water and 500 ml of saturated sodium
chloride. The acid was then extracted from the ether layer
with saturated sodium carbonate solution until the aqueous
layer was at pH 11 (using pH paper). The basic layer was
diluted with water until the acid salt dissolved. The ether
layer was separated and the aqueous layer was extracted
twice with 500 ml of diethyl ether. The aqueous solution of
sodium 2,5,5-trimethylbutanate was then acidified with con—
centrated sulfuric acid until the pH of the aqueous solution
reached three. -During this addition the solution was cooled
with ice. The aqueous solution was extracted with 1.5 l. of
diethyl ether in three portions. .The combined ether extracts
were washed with saturated sodium chloride and dried over
anhydrous magnesium sulfate. After removing the diethyl
ether by rotary evaporation, the residue was fractionally
distilled, yielding 64.8 g (66%) of 2,5,5-trimethylbutanoic
acid, bp1208-100, mp 42-5°.
I. Preparation of 2,5,5-Trimethyl-
butyryl Chloride

In a three—necked 250 ml round-bottomed flask fitted
with stirrer, condenser, and addition funnel,.57.7 g (0.44
mole) of 2,5,5-trimethylbutanoic acid was mixed with 78.6 g
(0.66 mole) of thionyl chloride. .The reaction mixture was
stirred for one hour and heated at reflux for an additional

hour. .Then, the excess thionyl chloride was distilled off

54

and the residue fractionally distilled, yielding 55.7 g

(85%), bp 150-50.
J. Preparationggf 5,4,4-Trimethylr2-pentanone

A three molar solution of Alpha Inorganic methylmagnes-
ium bromide (190 ml, 0.57 mole) was placed in a 500 ml three-
necked round—bottomed flask fitted with stirrer, condenser,
and a glass stopper. To this ice-cold solution was added
55.4 g (0.29 mole) of anhydrous cadmium chloride. The solu-
tion was stirred an additional hour after the ice bath was
removed. After distilling off the ether, 260 ml of anhydrous
thiophene-free benzene was added and an additional 50 ml of
distillate was collected. The benzene solution of dimethyl-
cadmium was cooled in an ice bath and 55.7 g (0.58 mole) of
2,5,5-trimethylbutyryl chloride was added at a rate to main-
tain reflux. .Ten minutes after addition was complete, the
solution was heated at reflux for one hour. The benzene solu-
tion was poured into 500 ml of ice and water. Sulfuric acid
(10%) was added until the solids dissolved and two clear
layers formed. The benzene layer was separated and the
aqueous layer was extracted with 200 ml of benzene. .The ben-
zene extracts were combined and dried over anhydrous magnesium
sulfate. The benzene was removed by distillation and 5,4,4-
trimethyl-Z-pentanone was fractionated, yielding 29.9 g
(62%), bp 145-60. The nmr Spectrum consisted of a singlet
at 7.95 T (5 H), a singlet at 9.1 T (9 H), a doublet centered

at 9.0 T (5 H), and a quartet centered at 7.52 T (1 H).

55

K. Preparation of 5,4,4-Trimethyl-2-
. pentanone-1,1,1-g3

 

In a 100 ml flask equipped with condenser, drying tube,
and a Teflon-covered magnetic stirring bar was placed 10 g
(0.078 mole) of 5,4,4-trimethyl-2—pentanone and 50 ml of pH
11 deuterium oxide (99.7%). ,The mixture was heated at 1000
for twelve hours, then cooled and separated. The organic
layer was dried over molecular sieves. This process was
repeated twice with 50 ml of 99.7% deuterium oxide and twice
with 20 ml of 99.7% deuterium oxide. .The resulting ketone
was distilled, bp 144-60, yielding 6.5 g (62%). The nmr ,
Spectrum consisted of a quartet centered at 7.1 T (1 H), a

singlet at 9.04 T (9 H), and a doublet centered at 9.0 T (5 H).
L. ~Additions to 5.4.4-Trimethvl-2-pentanone

In a typical addition of metal deuterides to 5,4,4-
trimethyl-2-pentanone, 1.0 g (7.6 mmoles) of ketone in 2 ml
of solvent (diethyl ether, tetrahydrofuran or isopropyl
alcohol) was added to a solution of 0.089 g (2.1 mmoles) of
metal deuteride in 16 ml of solvent. The lithium aluminum
deuteride reductions were carried out in diethyl ether, while
tetrahydrofuran or isopropyl alcohol were used as solvents
in the sodium borodeuteride reductions. The additions were
kept at constant temperature during and following the reac-
tion (six to twelve hours). They were quenched by adding
0.18 ml of water and 0.18 ml of 5% sodium hydroxide solution.

.The inorganic salts were removed by filtration and the

56

solvent by distillation. The 5,4,4-trimethyl-2-pentanol-2fid
was collected from a six foot Carbowax column. The diastereo-
meric ratios taken from the nmr of these alcohols are sum-
marized in Table III.

M. -Additions to 5,4,4-Trimethyl-2-
pentanone-1,1,1-g3

 

In a typical experiment, 1.0 g (7.6 mmoles) of 5,4,4-
trimethyl-Z—penanone-1,1,1-g3 and 2 ml of solvent (diethyl
ether or tetrahydrofuran) was added to 10.0 mmole of Alpha
Inorganics prepared organometallic reagent in 10 ml solvent.
The corresponding reagents and solvents used were methyl-
magnesium bromide and iodide in ether, methyllithium in ether,
and methylmagnesium chloride inxtetrahydrofuran. .The reac-
tion mixtures were held at constant temperatures during the
additions and for Six to twelve hours afterwards. .The reac—
tions were quenched with 0.4 ml of 10%iammonium chloride
solution. The inorganic salts were filtered and the solvent
removed by distillation. .The 2,5,5,4-tetramethyl-2-pentanol-
1,1,1eg3 was purified by vpc (six foot Carbowax column).

The ratios of the diastereotopic methyl groups at 8.78 T and

8.92 T are summarized in Table III.

N. Preparatign 9f 2,5,§-Trimethyl-1-butanol

In a 5 l. three-necked flask fitted with stirrer, con-
denser, and addition funnel, 100 g (1.02 moles) of 2,5,5-

trimethyl-l-butene and 15.7 g (0.415 mole) of sodium

57

.cuoHEmonm um ooa um can 35.30on on» «0
conon c. 0;. cu o.m 0:» MH 4 uHmucH 05H. .ooHHoHso mcoaanuofi UHoHom
SH HocmusnIHIthumEHqumfi.m uo mcoHHSHom Ron «0 guuoomm HE: manila ousmfim

p
0.0H O.m 0.0 0;. O.m O.m
_ . _ H h . _ p r . c - .

 

 

 

 

 

 

 

58

borohydride were mixed in 500 ml anhydrous diglyme (distilled
from lithium aluminum hydride). The reaction mixture was
cooled during the addition of boron trifluoride etherate

(70 ml, 79 g, 0.55 mole). The excess sodium borohydride was
destroyed with 80 ml of water. The oxidation of the organo-
borane was accomplished by using 500 ml of 5-N sodium hy-
droxide and 120 ml (176 g, 5.2 moles) of 50% hydrogen peroxide.
After the addition was complete, 1 l. of pentane was added,
and the aqueous layer was separated. The pentane solution
was extracted six times with water until no diglyme was
present in the vpc of the pentane layer. This pentane solu-
tion of the alcohol was distilled and yielded 85.5 g (75%)
of 2,5,5-trimethyl-1-butanol, bp 159-610, literature bp
161-20 (740 mm).

0. Preparation 0:72,5,5-Trimethylbutanal

The 2,5,5-trimethylbutanal was prepared by oxidation of
2,5,5-trimethylbutanol with bis (pyridinium) chromic (VI)
oxide. .The bis (pyridinium) chromic (VI) oxide was prepared
in methylene chloride and used immediately upon formation.
In a 6 l. Erlenmeyer flask, to a solution of 190 ml (2.4
moles) of pyridine (dried over potassium hydroxide) and 5 l.
of anhydrous methylene chloride was added 120 g (1.2 moles)
of Fisher chromic trioxide, N.F.X. grade. .The chromic tri-
oxide was finely powdered with a ball mill. The chromic

trioxide was added in small portions with vigorous stirring.

59

.HH on no fieHfimgu Hm Hume 083%: as» «H.
a uuouca one .Hmcmusnawzuoawqum.n.m mo Esuuummm Mac ummc oEBTI.N musmwm

P
0.0d 0.0 0.0 o.¢ 0.N 0.0
— r L p — b b n — p — L

 

 

 

 

 

 

40

The solution was yellow at the beginning of the addition,

but turned gradually to a dark red by the end of the addition.
.The reagent was allowed to stand for thirty minutes. Then,
the 2,5,5-trimethylbutanol was added in one portion (25.2 g,
0.12 mole). vAs the solution was stirred, it turned black.

The reaction mixture was stirred several times during the
following hour, it was filtered through glass wool and finally
divided into two portions. .Each portion was cooled with ice
and washed with 500 ml of 10% hydrochloric acid. .The organic
layer was dried with molecular sieves and filtered through a
sintered glass filter with a mat of 100-200 mesh Florisil.

The clear solution was then distilled and the residue was
fractionated on a Hester-Faust Spinning band column, yielding
.26.4 g (58%), bp 126-90, Vpc pure, D.N.P. mp 125-70, litera-

ture 126-70.

P. Preparation of 2,5,5-Trimethylbutanal-2—gr

 

,Ten grams (0.088 mole) of 2,5,5-trimethylbutanal was
treated with one 25 ml portion of pH 11 deuterium oxide (80%),
and three 25 ml portions of pH 11 deuterium oxide (99.9%).
.Between each exchange the mixtures were heated at reflux for
six hours. The aldehyde and aqueous layers were separated
with a separatory funnel after the aqueous layer was saturated
with sodium chloride. ”The aldehyde was dried after each step
of the exchange with molecular sieves. ‘The residue was

fractionated, yielding 6.5 g (62%).

41

Q. Additions to 2,5.5-Tgimethylbutanal

In a typical experiment, 0.5 g (4.4 mmoles) of 2,5,5—
trimethylbutanal and 2 ml of solvent was added to 6.0 mmoles
of Alpha Inorganics prepared Grignard or methyllithium re-
agents in 6 ml of solvent. The corresponding reagents and
solvents used were methyllithium, methyhmagnesium bromide or
iodide in ether, and methylmagnesium chloride in tetrahydro-
furan. The reaction mixtures were held at constant tempera-
ture in a bath during and after the addition (six to twelve
hours). .The reactions were quenched with 0.25 ml of 10%
~ammonium chloride solution. ‘After filtering the inorganic
salts with a sintered glass funnel, the solvent was removed
by distillation. The diastereomeric 5,4,4-trimethylw2-
pentanols were purified by vpc (six foot Carbowax column)
and their diastereomeric ratios obtained from the integration

of the vpc traces. These ratios are summarized in Table III.

R. ~Additions to 2,5,5-Trimethylbutanal-25Q

 

A solution of 0.1 g (2.5 mmoles) of metal deuteride was
dissolved in 16 ml of solvent and the mixture was stirred at
the boiling point of the solvent, cooled at 00 in an ice
bath, or held at -700 with a dry ice isopropyl alcohol bath.
The solvent used for the lithium aluminum deuteride reduc-
tions was ether; and that for the sodium borodeuteride reduc-
tions,tetrahydrofuran or isopropyl alcohol. .After the solu-

tions had been stirred for an additional six to twelve hours,

.cuoHsmoo3m um 00a am can Esuuoomm on» no conoH H 0.5
on 0.0 on» mH e uuomcw one ,.ooHHoHco ocoaanuoa UHoHom cw Honooam on»
mo coHuSHom Row M ca coxmu mm3 asuuoomm one .cmusmouomzmuuou cH 05:
um ooHHouSooouon ESHoon £HH3.WINIHoccusnawcuoEHHuIn.m.N Ho coHuwmou
ecu 2H emchHno mmrm.HsaocmusnaHuasnuesHuuum.m.m Ho ssuuueem He: egauu.m mHsmHm
P
0.0H o.m o.m . o.>

- r — n —

0.0 O.m
H p h — L I—

4, flI» ,/\, III,

’IIF'I II I ' NI, I"
I

{I I 4 1 Ill 1‘

 

42

 

 

 

45

the reactions were quenched with 0.2 ml of water and 0.2 ml
of 5% sodium hydroxide. The inorganic salts were filtered
and the solvent removed by distillation. .The 2,5,5-tri-
methylbutanol-1,2-gg was purified by Vpc. The ratios of the
diastereomers were determined by peak integration of the

nmr (6.22 T and 6.75 T). The results are given in Table IV.

BIBLIOGRAPHY

(1)

(2)
(5)
(4)

(5)
(6)

(7)

(8)
(9)
(10)

(11)

(12)

BIBLIOGRAPHY

J. H. van't hoff, Bull. soc. chim., 88, 295 (1875).
This article was translated from French and published
in "Classics in the Theory of Chemical Combinations",
edited by O. T. Benfey, Dover Publications, Inc.,

.New York, 1965, p. 151.

A. W. Hofmann, Chem. Ber., 8, 704 (1872); 8, 61 (1875).
V.Meyer, ibid., 541, 510 (1894).

J. Kenner and W. V. Stubbings, J. Chem. Soc., 1921, 595;
J. Kenner and G. H. Christie, ibid., 1922, 614.

E. E.-Turner and C. V. Ferriss, ibid., 1920, 1140.

D. J. Cram and F. A. Abd Elhafez, J.-Amer. Chem. Soc.,
14, 5828 (1952); D. J. Cram, F. A. Abd Elhafez and H.
Weingartner, ibid., 18, 2295 (1955); D. J. Cram, F. A.
Abd Elhafez and H. L. Nyquist, ibid., 18, 22 (1954):

D. J. Cram and J. E. McCarty, ibid., 18, 5740 (1954);

D. J. Cram and K. R. Kopecky, ibid., 81, 2748 (1959): and
D. J. Cram and D. R. Wilson, ibid., 88, 1245 (1965).

W. G. Dauben, G. J.-Fonken, and D. S. Noyce, ibid., 18,
2579 (1956).

G. J. Karabatsos, ibid., _8, 1567 (1967).
G. J. Karabatsos and N. Hsi, ibid., 81, 2864 (1965).

D. Y. Curtin, Record Chem. Progr.,.18,.111 (1954).
See also E. L. Eliel, "Stereochemistry of Carbon

,Compounds",.McGraw-Hill Book Co., Inc., New York, 1962,

pp. 151-156.

H. C. Brown and K. Ichikawa, J. Amer. Chem. Soc., 84,
575 (1962); H. C. Brown, 0. H. Wheeler, and K.-Ichikawa,
Tetrahedron, 1, 214 (1957).

H. C. Brown and J. Muzzio, J. Amer. Chem. Soc., 88,

-2811 (1966).

44

,III I INN-Ill!!! II I

(15)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(25)

(24)

(25)

(26)

(27)

45
G. J. Karabatsos and T. H. Althuis, Tetrahedron Letters,
4911 (1967).

N. Allentoff and G. F. Wright, J. Org, Chem., 88, 1
(1957); D. Seeback, H. Dorr, B. Bastani, and-V. Ehrig,

.Angew. Chem. Integnat. Edit., 8, 982 (1968).

J. H. Stocker, P. Sidisunthorn, B. M. Benjamin, and C.
J. Collins, J. Amer. Chem. Soc., 88, 5915 (1960).

M. van Gorkom and G. E. Hall, anrt. Rev., 88, 14
(1968).

K. B. Wiberg, "Physical Organic Chemistry," John Wiley
and Sons, Inc., New YOrk, 1964, pp. 576-579.

G. J. Karabatsos and C. Zioudrou, 1968, unpublished
data.

M. Cherest, H. Felkin, and N. Prudent, Tetrahedron
Letters, 2199 (1965); Y. Gault and H. Felkin, Bull. Soc.
Chim. France, 1542 (1960).

F. A. Bovey, "Nuclear Magnetic Resonance Spectroscopy."

‘Academic Press, New York, 1969, p. 155.

T. H. Althuis, Ph.D. Thesis, Michigan State University,
East Lansing, Michigan, 1968. I

G. Zweifel and H. C. Brown in "Organic Reactions",

-Vol. XIII, A. E. Cope, Ed., J. Wiley and Sons, Inc.,

New York, 1965, p. 55.

B. Helferich and W. Schaefer in "Organic Synthesis",
Coll. Vol. I, H. Gilman,.Ed., J. Wiley and Sons, Inc.,
New York, 1952, p. 147.

G. Zweifel and H. C. Brown in "Organic Reaction", V01.
XIII,-A. E. Cope, Ed., J. Wiley and Sons, Inc.,
New YOrk, 1965, p. 29.

G. Zweifel, N. R. Ayyangar, T. Munekata, H. C. Brown,
J. Amer. Chem. Soc., 88,.1076 (1964).

J. C. Collins, W. W. Hess, and F. J. Frank, Tetrahedron
Letters, 5565 (1968).

-W. H. Urry and N..Nicolaides, J. Amer.-Chem: Soc., 14,

5165 (1952).

   

M'TIIIIIIIIIIIIIIII[IIIIIIIIIIEII'IIIIIJIIIIIT