This item is only available as the following downloads:
5':;:T :F:.: OF H.TE: .'CYCLIC COrMPOUNDS CONTAIMNiI B--N COCRINA[E
B9C.....- AS MODELS
FOPR TI: I-.L:.Y STABLE POLYMERS
CHARLES LEWI MCCOR!ICK III
A DISSERTATION PRESENTED TO Th-E G -))ADUATF 7U0J;CEL 0, TE JNIVE-T: *F
FLOPIDA IN PASTA FULLILL T OF REQI 'L.i'' "'LT
OCR THE DEGRE OrF DOCTORR G FHIL.SuPH
t' r~; -'''
This dissertation is dedicated in memory of
Mr. Claude Stuart
who dedicated his life to teaching chemistry
in the Greenville, Mississippi, Public School
I wish to express my gratitude and appreciation to Dr. G. B. Butler
for his patience, guidance, and understanding during the course of this
research. My appreciation is also extended to Dr. T. E. Hogen Esch, Dr.
Paul Tarrant, Dr. Martin Vala, and Dr. Henry C. Brown for serving on my
Grateful thanks are extended to my colleagues in the laboratory for
their frienship and good humor, which made work quite enjoyable.
The financial assistance received from the National Science Founda-
tion is gratefully acknowledged (grant number GH32766 and GH17926i)
I also wish to thank my parents for their support and encoragement
during my graduate studies.
A special expression of gratitude goes to my wife, Pat, for her love,
encouragement, and assistance in completion of this project.
TABLE OF CONTENTS
Acknowledgements..... ... .
List of Tables .
List of Figures. .
I. IIITF.CDUCTI .
A. Polymers Exhibiting Principles of Thermal Stability
B. Boron-Nitrogen polymers .
C. Boron-Oxygen Polymers .
D. Boron-Carbon Polymers. .. ..
E. Amine-Boranes as Hydroborating Agents .
F, Boron-Nitrogen Coordinate Bonds in Cyclic and
Bicyclic Compounds .... .
G. Statement of Problem .
II. SYNTHESES OF UNSATURATED TERTIARY ANILINES .
A. Syntneses of Para-substituted-N.N-Diallylanilines
B. Synthesis of N,N,N',N'-Tetraallyl-p-
C. Synthesis of N,N-Di-3-butenylaniline .
III. 'FFiPAF.'iION OF BORON INTERMEDIATES .
Synthesis of Grignard Reagents
Preparation of Substituted Pbenylboronic Acids
Syntheses of Borate Esters of Substituted
PhenylDooronic Acids . .
Cyclotriboroxenes . .
Preparation of Triethyl- .- fr.-!;, Lboranes
Preparation of Pyridine-Phenylborane .
IV. SYNTHESES AND REACTIONS.OOF BORON HETEROCYCLES 31
A. Reaction of Triethylamine-Phenylborane with
N,N-Diallylaniline .. 31
B. Mechanism of Formation of Azaborolidines and
Azaborabicyclo(3.3.O.)octanes .. .35
C. Deuterium Labeling Studies 41
D. Preparation of 3-Deuteropropene by an
Alternate Method .. .. 50
E. Preparation and Reactions of p-Substituted 1,5-
Diphenyl-l-aza-5-borabicyclo(3.3.0. )cctanes 50
F. Preparation of Bic-L,L4-[L5-( 4-m-ethylphenyl.)-1-
Bis-1.,4-[5-( 4-chlorophenyl )-l-aza-5 -borabicycio-
G. Reaction of N,N,N',N'-Tetraallyl-p-phc.ylene-
diamine with p-Phenylenediborane .. 62
H. llBoron Nuclear Magnetic Resonunc Studies .
I. Temperature Studies by Differential Scanning
J. Synthesis of 1,6-Diphenyl-"-a.zt-6-borabic,'..-
(4 4.0.)decans .
K. Conclusions .. .
V. EXPERIMENTAL .
A. Equip,:tert and Trx:-atment: of a,t: 7-
B. Syntheses and Characterizaton 7'
2i li.ography .
Biograph ica Ske. tch
LIST OF TABLES
1 Skeletal Bond Energies 4
2 Substituted 1,5-Diphenyl-l-aza-5-borabicyclo(3.3.0)octanes
and i,2-Diphenyl-1 ,2-azaborolidines 54
3 Physical Data of Substituted 1,5-Diphenyl-l-aza-5-bora-
4 Spectral Data of Substituted 1,5-Diphenyl-l-aza-5-bora-
LIST OF FIGURES
1 Nmr Spectrum of 1,5-Diphenyl-1-aza--5-borabicyclo(3.3.0)-
2 Infrared Spectrum of 3-Deuteropropene 45
3 Nmr Spectrum of 3-Deuteropropene 46
4 Infrared Spectrum of 3,7-Dideutero-l,5-diphenyl-1-aza-
5 Nmr Spectr'un of 3,7-Dideutero-l,5-diphenyl-i-aza-5-bora-
6 Nmr Spectra of 1,5-Bis-(4-cblocropheiiyl)-i-aza-5-bo'a-
bicyclo(3.3.0)octane and 1,2-.ic-(4-chlorophenyl)-1,2-
Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
SYNTHESES OF HETEROCYCLIC COMIPCUOIOD CONTAINING B-N COORDINATE
BONDS AS MODELS
FOR THERMALLY STABLE POLYMERS
CHARLES LEWIS MCCORMICK III
Chairman: Dr. G. B. Butler
Major Department: Chemistry
The major goals of this research were to exaAine the mechanisnI of
tha reaction of triethylaimine-phernyloranrT. with NN-d- alylani.L:' i a"-
to extend this reaction to the preparation of compounds which could be
utilized for the synthesis of high molecular weight, thermally stable
1,5-Diphenyl-l-aza-5-borabicyclo (3.3.0) octane, 1 ,2-d.phernyl-, 2-
azaborolidine, and propene were isolated as the major products of the
reaction of triethylamine-pheny Iborane with N,N-diallylaniline. Thes-
compounds were characterized by nuclear magnetic resonance, infrared,
mass spectroscopy, and elemental analyses.
Two mechanisms were proposed for the forration of propene arnd
Triethylami ne-dideuterophanriyboran-;? ,a prepsard anri s'bsequelntly
reacted with N,N--diallylanilin'e to give 3,7-dideutero-i,5-diphenyi-l--aza
5-borabicyclo( 3.0) octane, 3-deutero- 2--diphieny1- 2- xa;borolidiTie,
and 3-deuteropropene. These products were consistent with one of the
proposed mechanisms, a concerted, facile elimination of propene. This
elimination mechanism was supported by model studies of the transition
Triethylamine-phenylborane was reacted with N,N-di-3-butenylaniline
to give 1,2-diphenyl-l-(3-butenyl),2-hydro-azaboracyclohexane and 1,6-
diphenyl-l-aza-6-borabicyclo(4.4.0)decane. No butene gas was eliminated,
giving further support for the proposed mechanism.
Several substituted derivatives of 1,5.-diphenyl-l-aza-5-borabicyclo-
(3.3.0)octane were prepared. The B-N coordinate bonds in these compounds
were studied by nuclear magnetic resonance, 1B nuclear magnetic resonance,
infrared spectroscopy, and differential scanning calorimetly. The B
chemical shifts in these ccrapounds varied from 2.2 to 6.0 p.p.m. relative
to trimWethylborate. The B-N coordinate bonds e:xhibited absorb'inc-, near
1275 cm. Temperatures for the dissociation of the boron-nitrogen coor-
dinate bond, determined by differential scanning calorimetry appeared to
be near 5200 K. The heats of dissociation of the B-N bond for bis-l,4-
[5- 4-methylphenyl)-l-aza-5-borabicyclo(3.3.0)octyl]benzeue, and bis-1,4-
[5- (4-o-'~.o ophenyl)-l-aza-5-borabicy.lo(3.3.0)octyl]benzene were calciu-
lated to 28.5 krai/m-ole and 30.3 kcal/zioli,, respectively.
Sev-eral substituted derivatives of i,5-diphenyi-i-aza-5-borabicyclo-
(.3.3.0)octane were prepared which are excellent monomer precursoTs for
A. Polymers Exhibiting Principles of Thermal Stability
The demands of the space program, and more recently the garment
industry, have led to extensive research in the area of thermally stable
polymers and flame retardant polymers.
Ladder polymers, 1, or double-stranded polymers have been made
which show good thermal stability. The increased thermal stability of
these polymers was attributed to the added stability cf the ring systems.
Cleavage must occur twice within a given ring of the polymer for chain
degradation. i,2 ,4,5
Recently, ladder polyamerrs cabls in air above '09' C and based
on ,inyJ-acetylenes, have been p -paro.
Hete.roatom pclyners show a higher resistance to oxidative degra-
d.cion at elevated telera:ture's. keletl bonds, s'nch as Al-0, Si-O,
Si-N, P-N, B-0, and 3-N, might: b cxpected to show resistance to ther-
mal oxidation, and polymers wicih r.hei:e units should be --ore inert ra
high temperature. Bon c.A is:i:, ..c:rgies arc given in Table
Polyphenyleneoxadiazoles, polyphecQylenetriazoles, polyphenylaee-
thiazoles, and polybenzimidazoles have been shown to be capable of
forming thermally stable films aid fibers.8,9,10,11 Another series of
thermally stable heterocyclic polymers, 2, was prepared by the conden-
sation of 3,3'-diaminobenzidine with tetrabutyl esters of diboronic
NH--/\ -\N \
I NH ./ "NH
butler and Stackman demonstrated the high resistanc" of
cyclic polydia.lylriphenyl silanes, 3, to thermal degr-a-atici. Po..-
atrs ;containiag bodith eteroatonm stability aid "ladder' ralili.y 'i-
been prepared by the hydrolysis of phenyltrichlorosilane in ai o.rg c ic
"' -- L .-' *- I- --
i recent yvet:s, a number of hete;o:or3 a riug and chain syyste.;s';,
sucI; as; carbazeno, 4, orazenes, 5, phosphazenes, .6, bcr~oeaes, .,
si.lo-':na"e-, n3, ad silazanes, 9, have sho.:n remarkable therial st .biity.
S4 5 I 6
0 0 0 0 NN N
I I I I
7 8 9
This stability has been interpreted in terms of resonance stabili-
zation inherent in these systems. An intensive investigation into the
possibilities of incorporating these compounds into polymers for flame
retardance is currently of top priority in the fibers industry.
.-:trcatom polyiners containing C-N, C-0, B-N, F-C, P-N, Si--N, and Si-0
bonds have been reiewLed.
Tberaoplasti.: ccati-gs and fibers have been prepared from cyclic
polys loxanes, Methylphenylcyclotetrasiloxanes have been polym..ri:ed to
give polymers with thermal stability dependent upon phenyl substitution.
He~eroato:m pclymers containing boron have been known for Tiany years.
Boric oxide consi..-s of a three-dimensicnal network of distorted 30,
retrahedra. HexagRorna boron nitcide, _10, has a hexagonal layer struc-
ture and is som-atl n ca.ll.d ,h.ite graphite. These materials have ex-
cellent th er:i. ability 2'"h0'0C) and other useful. properties whii
bave coamnerci.ai utility. Th iey have stimulated research efforts toward
.- .. ', -,
TABLE 1. SKELETAL BOND ENERGIES (kcal/m.ole)
high molecular weight polymers based on boron. Certain phases of boron
chemistry also find their counterparts in the technologically important
area of silicon chemistry. Since the bond energies of boron-oxygen
(~130 kcal) and boron-nitrogen ('100 kcal) are high, there should be
high temperature applications.
Most polymeric boron compounds have not been characterized ade-
quately. Little information is available on reactivity, molecular
weights, and solubility in many cases.2
B. Boron-Nitrogen Polymers
Since the discovery of borazene and its stability at 500C,22
mai-y publications have appeared attempting to incorporate this com-
pound into thermally stable polymers. Reviews covering all phases
of borazene chemistry may be found in the literature. An impor-
tant aspect of borazene chemistry is the strong driving force for
formation of cyclotriborazenes from dehydrogenation of substituted
amine-borane adducts (Scheme I).
PNH2 + H2BR' ---- R2 N-->BH2R' --- rN-.-' '
(1) Monoborazans (2) Monoborazeac
H 2 H
R' R' -3H (4)
I I /
H A --.-3H. ()/
Cyclotriborazane R- B-R'
Cy clotri -raze
Cy clotril 2r"zene.:
The monoborazine would be electronically unstable and should form
polymers. However, the stabilizing influence of suitable bond angles
and n-bonding in the cyclic trimer favors this homologue.
Three main avenues have been investigated for the formation of
boron-nitrogen polymers. The first involves the opening of various
borazene rings, and the second, reactions of the unstable mcnoborazines.
The final method has been an attempt to link borazene rings in an
ordered fashion, avoiding cross-linking.
As a general rule, cyclotriborazenes do not undergo ring opening
to liear polymers, However, the polymerization of N- and/or B-methyl--,
ary]-, or triazinylcyclotriborazenes in a closed system at 3500 to
600'C, oc in the presence of catalysts, led to hard, glassy, irsolub~le
resins.' No molecular weight was given. An open-chain, low
,o.leccular weight polyiser of the structure, (PhB-NC4H9)n, where n=20 to 40
was prepared by rhhe reaction of pheny1borondichloride with isobucyi.amias."
Ctross-.lin2ked polymer was formed by the pyrolysis of B-trichloro-N--tri-'
arlcyclotrlborazeues, (C6-BAr) 3, from room temperature up to 500"C."
r'y,:.oboraene rings, havpb been linked to give polymers, ,1 of high
mol.cuiar weight. Benzene solivle polymer i7as formed fhan ;yycloborazene
vings ware linked by :oe process:2
Me Me !
n i + 2n KC5.
.-.N -. N.-Hs- hen B..h :.
Cvc3tr'iboraznc- : rics ,hav,, a.l.so bee :inked v react etg dio!p ithi
HBP B H -
2 I i + HO(CH2)200H -2B. 20
PhN BNPh PhN 2BNPh
H 6(CH2 )2
Other examples have been reported in which B,N'-dichlorocycletri-
borazenes and diamines reacted to give polymers stable up to 250-
300CC.29 Copolymers have been prepared by the free-radical poly-
merization of B-trivinyi--N-triphenylcyclotriborazene wich styrene
and methyl methacrylate.30 The reaction of aniline with an equi-
molar concentration of trimethox'-yclotriborexene was reported to
give a poly :er with the following structure:
.B ----- N
13 h -n
High boiling diamines have been known to polymerize upon
heati'-ng with various alkylborates. For example, ethylenediamiine
reacts with tri'ethyLbcrate to givw" a polymer with the following
N-- --B -
The reaction of p-aminophenol and tripropoxyborane resulted
in the formation of a resin. The structure assigned to the poly-
B NH 0- -- --NH -- --- B
C. Boron-Oxygen Polymere
Oligomeric compounds containing boron-oxygen links have been
made by heating al ky.-- or arylboronic acids. Most of the resulting
compounds are easily hydrolyzed, although they exhibic remarkable
thermal stability. As in the borazene series, the trimers, cyclo-
triboroxenes, sec. To be the most favored thermndyaa;micaly.y Cycle.
triboroxenes have bce'n reported to have 10-20% aroima-.ic chadrcrx:e.
boronic acids under !vauu',- le first ainhydrlde of phenyiboroni.
acid was prepared in 1836 Te structure of triphenycyclotri-
phenylboronic acid yielded th cyclic triTer:
Cyclotriboroxenes disproportionate to boric oxide and trialkyl or
triaryl boranes at elevated temperatures. This fact seems to
rule out the possibility of ring-opening polymerizations of these
Benzenediboronic acids have been prepared as starting materials
for polymerization. These acids were formed by the extension of the
synthesis of phenylboronic acid, 19.,41
-gBr (1) (CH30)3B -B(OH).
Diboronic acids were prepared by Nielsen4 and Musgrave43 by
reaction of a bis-Grignard reagent with trimethylborate in tetra-
hydrofuran. Subsequent hydrolysis yielded the diboronic acid, 20.
Polyesters from diboronic acids have been reported, but no mo3!ecular
weights were given.4 Mixtures of boric acids with glycols gave high
molecular weight polymers. No structure det.erminations were made,
alti;'..",h cross-l-inking is inevitable in these systres. Polymeric
glycol borates and esters of borcnic acids have hben studied. A
careful steady of the condensation reactions of orthoboric aci-3l with
diols has also been made. RecenLly, Svarcs a.t al. reported th-
preparation of a copolymer, 21, of pentaerythritol and boric a.cid.
D. Boron-Carbon Polymars
Polymers containing only boron and carbon are rare. Tri-n-
hexylboron decomposed on heating to produce two polymeric materials.4
A dark solid was assigned the structure [-B(C6H13)BH-]nand a vis-
cous liquid [-B(C6H13)-(C6H12)-B(CHI13)-In Pyrolysis of Me2BCH2CHIa Me2
yielded trimethylboron and polymers with the following structures:0
B-CH-CH-B -- -CH2--CHi B / -
1, \_2 -
A patent has been awarded for the synthesis of a polymer, 24, pre-
pared by reacting a boron halide or ester with the di-Grignard of
Several other polymeric boron-carbon compounds have been re-
The first organoboranes were synthesized in 1862 by FranklanAd
by tle reaction of diaikyl zinc compounds with Lriethylborate. He
He observed the ability of these boranes to coordinate with ammonia.
Since that time countless papers on other boron-nitrogen coordi-
nation compounds have appeared.
Koster first utilized triethylaiine-borane in 1957 in the hyiro-
boration of cyclododeca-l,5,9-triene. Hawthorne57 observed that
pyridine borane would react with olefins in diglyne at 100C to pro-
duce trialkylboranes. Nielsen58 observed that phenylborane readily
disprcportionated to triphenylborane and diborane. Hawthorne59 pre-
pared pyridine and triethylamine adducts of phenylborane and aryl-
substituted phenylboranes which were white crystalline materials stable
in air and had melting points from 50 to 80*C. Pyridine- and tri-
ethylamine-alkylboranes0 and -arylboranes59,61 were prepared by the
presence of the amine bases.
Hawthorne reported the reduction of substituted cyclotriboroxenes
by lithium aluminum hydride and the presence of pyridine, trimerhyl-
amine, or triethylamine to give the corresponding amine-bcranes.
Ashby64 reacted several trialkynamine-boranes with olefins to
form trialkylboranes in 78--95% yields.
The mechanism for hydrobor.ltion by amine-boariic is believed ;:o
involve the rate determining dissociation of the smine-borane. A
sufficiently high temperature is required for the dissociation of the
B-N bond.65 B-H addition can then proceed via the boron p-orbital
interaction with che T-elect-rcns of the olefin (Schem.e II).66
:.' -R RN + RMH,
tBH/2 + ..=.. ... = -- --- -- C--- C -
H ---- -H.....uuR
Cyclic boranes were prepared by Koster6768 by the reactions of
trialkylamine-boranes with diolefins. Hawthorne9,70 reacted trimethyl-
amaine-t-butylborane with a number of olefins and diolefins. Reaction
with divinyl ether and divinyl silane gave corresponding heterocyclic
The coordination compounds formed by substituted alkyl- and aryl-
borons with amines have been studied relating steric factors to B-N
bond stability. Heats of dissociation of the complexes of trimethylboron
with various amines are all about -7.26 0.21 kcal/mole.7
Aaine-boranes may undergo dehydrogenation to give monoborazenes
or aemnoboranes (Scheme I). When B-trimethylborazane, 25, was heated
to 2800C under 20 atm.. pressure, B-dimethylborazena, 26, was forqpd.
1his compound was further dealkylated to yield B-trimethylcyclotri.-
borazene, (HN-BMe)3, Both of the above steps required cbe eliim ati-.on
H3N-BMe ------- N-BMe CH
3 3 Me. 2 4 t
Further examples of elimination were the formation of B-a-llyl-
vonoborazenes by the following reactions:
-:B(C 3 ------.- EtN-B(C + C h
H3N+B(C3)3 -------- f ,2N-B(C3H5)2 + C3
Pyrol.-:. of amine--phenylborane adducts leads to aminoboranes. The
heating of diethylamine-phenylborane at 80--110C under vacuum resulted
in 88% yield of diethylaminoborane, 29.'
Et2NH-BH2Ph -- H2 + Et2N-BHPh
Several papers were presented dealing with amine-boranes and
aninoboranes at an international symposium on boron-nitrogen chemistry.7
Heteroaromatic boron-nitrogen compounds have also been reviewed in
F. Cyclic and Bicyclic Compounds with Boron-Nitrogen Coordinate Bonds
The bicyclic ester, 30, formed by the azeotropic dehydration of an
equimolar mixture of l,l,l-trimethy.letLhane and boric acid was re-
ported in a patecrn in 1959.76 This compound upon heating yielded a
C-- ------- Polymer
Similar bicyclic compounds bad been prepared by 3rown in 1951
by the reaction of trit.anolam~ne vi.th boric acid'. 'he prodlu;:t was
a white crystalli:ne solid, m.p. 236.5-237.5C. Two possible struc-
tures were proposed:
\ C\ \ \,i
)--B--0 C) -V--yB~-~G
In the cage structure, 31, boron maintains trigonal coplanar hybridi-
zation. In the second structure, 32, a transannular boron-nitrogen
interaction gives boron sp hybridization. Triethanolamine borate
was concluded to have structure 32 on the basis of unreactivity with
methyl iodide as compared to quinuclidine, 33, and triethanolamine.
Further evidence for the B-N transannular bond was found in the 11B
nuclear magnetic resonance chemical shift, relative to boron tri-
fluoride etherate, of triethanolamine borate, which is found at
higher fields (6-10.7) in aqueo'us solution than most aikyl- and aryl-
borates (6-14 to 6-19). The shift was interpreted to be due ;-o
increased sl.iel'i.g around the boron atom as a result of B-N 'r : 'ir,g
mad, therefore, increased tetrahedral character and number of bonding
electrons around the boron atom.
Later studies showed that an equilibrium could exist bet w ern
structures 31 and 32. In water and butanol the equiliibrium mixture
contained 20% of the tetrahedral species, 32, and in aprctic solvents
95% of the tetrah2dral for. Iquil.ibrium s tudics of the estr-
ification of 1:oric acid -ith triethanol.amine indicate a shift in
eq.ii.ibriu- tr ard structure 3. w0 th an -icrhesze in tefperatu-re.
Zi.merman reported the resultS off -ydril;vysis exi-erilments on bcrox-
azoldinCs." H -elated T.he -T bond stc.nc tn to rat s of hydroyivis
for several c- .'n.-i ic.itudii. aze following, 34-37:
B -- C)0
Ph --<-N---CHK-CH-- --?--Ph
/ \ 2 2
A vir.ety of boronates, b.or.inat?, >*
triisopropylamine borates 99 have been postulated to ha;-r tra;ns-
annular boron-nitrogen bonds. Syntheses directed towa-;d boron-
containing amino acids for cancer chemotherapy have led to prepa-
ration of a series of substituted trietlanolamine borates. -' i -
isopropy1-91,92 and tri-n-propylamine borates have also been s uiJ.ed.
Substituted aryl boronic acids and 2-aminoethanol were heat-6 in
toluene to give the corresponding diethanolamiine arylborate,:, 4;0.",
Ar -- B<-NH
Recently polymers with improved antistatic properLies and flame
resistance were prepared by addition c f B-N ircllsioan .irpcunds. Fcr
example, the following compound (3% by weight) was mixed with poly--
/ MZie -CH-2 OPh
PhOCH 2- 0 CH2Ph
Adams and Poholsky96 prepared the first reported 1,2-aza-boro-
cycloalkane, 41, by reacting N,N'-dimethyl-allylamine with triethyl-
amine-borane in toluene. The following structure was assigned:
CH --- B N H
In 1963 White reported the synthesis of 1-methyl-2-phenyl-
1,2-azoboro.id:lne, 42_, and 2--2phenyl-!,9--azboracyc che-an.', -3-, fr.:
the reactions of triethy lamine-pher'ylboraae with N-!'.ethylailyl-n.-i
and 3-butenylaiine, respectively.
Ph___ 'B H--N 3___., f P...--- F h
P- i B
In i95o3 Statton and Butler o reported t he foni.-ation of .& ef.'j.rs
of diallyletiryJ amire with crimethylam ne-phenyhborane yielded com--
pounds with the following structures:
N"--./ k ....-- N '.
The B-N coordinate bond in the bicyclic compound, 45, was believed
responsible for a strong infrared absorbance at 1266 cm. The
azaborolidine, 44, showed strong peaks at 1512 cm.- which were at-
tributed to the =N bond. 11B resonance signal in the nmr spec-
trum of 44 was recorded at -23 p.p.m. and of 45 at 8.9 p.p.m. rela-
tive to trimethylborate.
In 1965 Butler and Statton99 reported the preparation of four
substituted l-aza-5-borabicyclo(3.3.0)octanes and four previously
unreported 1,2-azaborolidines. These compounds were prepared by the
reaction of triethylamine-phenylborane with tertiary diallylamines.
A mechanism for the reaction was proposed based on qualitative data
which involved the cleavage of an allylic carbon-nitrogen bond. Two
of the 1,2-azaborolidines were also prepared by the reaction of tri-
ethylaziine-phanyiborane with secondary allyla.ines. The l-aza-5-
borabicyclo(3.3.0)octanes exhibited strong infrard absorption bands
at 1250-1270 cm.-, which were assigned to the B-N coordinate bond.
A new compound, 1-phenyl-1-bora-5-azoniaspiro(4.5)decane, 46, was
prepared as well as a telomer, 4Y.
Ph 3 6
A new bicyclic .icoipound containing a B-P coordinate bond was
prepared having the following structure:
i _.j)"' -- ,/
G. Statement of Problem
The major goals of this research were to examine the mechanism
of the reaction of triethylamine-phenylborane with N,N'-diallylaniline
and to extend this reaction to the preparation of compounds which could
be utilized in the synthesis of high molecular weight, thermally stable
The first step was to prepare 1,5-diphenyl-l-aza-5-borabicyclo-
(3.3.0)octane, 49, and 1,2-diphenyl-l,2-azaborolidine, 50, (R,R'=H) to
serve as models.
k -R-~^' RB=N --(=--= N---R'
The next step wuuJd be to e.a.in; i-. ,chaaistica:!ly -h-:.- o r ncot
formation of 49 and 50 occurs in a competitive -manner, 01hvioirsly,
this reaction requires loss of three carbon units for formation of
50. Since no gaseous products or other products consistent with
this three-carbon chain had been previously detected, one goal
w.is ro isolate and identify :;his product.
After i.ni:ial investigation begau, it became obvious that deu-
terium labeling st.udics would be necessary. Therefore, an idea wa.r
conceived which would differentiate between two possible mechanisms.
Ano::ber goal .a.3 pr-inarat-.on of several selected derivatives of
f9, both for -synthetic purposes and property studies. Substituents
should have ielectrcnic and stacic .ffec-:s on the re.laive yields of
49 and 50. The boron.- it.o.en coordinat-. bond could be detected
by infrared, 1B nuclear magnetic resonance, and proton magnetic
resonance. The B-N bond dissociations for the various compounds
could be followed by differential scanning calorimetry. These
studies should have important meaning when extended to polymeric
Another important objective was the study of the stability
of these compounds to various conditions necessary for introducing
functionality for monomer formation.
The ultimate objective of the boron research in these labora-
tories is the incorporation of the bicyclic structure into several
thermally stable polymeric systems, 51.
(CH,) r(CH )
(CH 2x H2
These polymers should have some i:tult tst'irn properties. Thermal
dissociation of the dati-re bonds along the polymer backbone could
oczur without cleavage of the polyTer chains, alliing the polym'er
chain to "breathe" or function liK bellows. This cleaavage should
result in an iLncrease in the dei-r.;- of freedom along the longUI udin-'
axis of the polymer chain, and should impart additional elasticity
to the polymer.
SYNTHESES OF UI Sj'ATED TERTIARY ANILINES
A. Syntheses of Para-substituted-NN-Diallylanilines
p-Substituted-N,N-diallylanilines, 52, were prepared by the
method of Butler and Bunch.0 3-Bromopropene was added dropwise
to a slurry of sodium carbonate and the para-substituted-aniline
in water. The reactions were allowed to proceed at 100 C for 24
to 48 hours. The resulting organic layer was separated, dried, and
vacuum dis killed,
--H,( -:- 2 3-bromoprcpene -- X- )
The lol. owing compounds were prepared by the above method:
N.N--diallyla.nline, 37.6%; p-bromo-N,N-diallylariline, 74.5%;
p-.lc ro--N :::-I -.. -I l.lanil ine, 72.5% ; p-me thyl-N ,N-diallylaniline,
70 .1a; and p-methoxy-N.,.-diallylaniline, 88%
The ccrr spending se-ondar' :,onoallyian.ilnes were also ob-
t:Aine-d: fror, early .ractitns in the di.sillation.
B. Sy z h.iofN N' -T etralyl -p- henLvenediamine
p-Pnen)ilnediaminfe was i; ic.ed ,'t:ith 4 moles of allylbromide
in a slurry of sodium carbonate in water. The resulting viscous,
syrupy siatarial. wes diluted with water and the organic layer ex-
tracted in ether. This solution was filtered, dried over sodium
sulfate, and then distilled under vacuum yielding a pale yellow
liquid (46.5%, b.p. 138-140* C @ 0.55 mm.).
C. Synthesis of N,N-Di-3-butenylaniline
N,N-Di-3-butenylaniline was prepared using a similar proce-
dure as in preparation of the other diallylanilines. 4-Bromo-
butene was added dropwise to a slurry of aniline in sodium car-
bonate in water. The organic layer was extracted, filtered, and
vacuum distilled yielding a colorless liquid (37%, b.p. 74-75 C
@ 0.10 mm.),
PREPARATION OF BORON INTERMEDIATES
A. Syntheses of Grignard Reagents '
The boron intermediates were prepared by modifications of
procedures reported in the literature (Scheme III).
The mono-Grignard reagents of bromobenzene, p-dibromobenzene,
p-chlorobromobenzene, p-methylbromobenzene, p-methoxybromobenzene,
and p-ethoxybromobenzene were prepared by reacting the substituted
bromcbenzenes with an equimolair quantity of Ta!nesi.m in diethyl
ether, dried over zcdium. The substituted bromobenzene was added
dropwise at a rate which maintained a jow reflux.. In each reaction,
a small crystal of iodine or i d:op of ethylenabromicde was necessary
to initiate the reaction. T1.:e induction periods varied froTm 10 to
40 minute .
An attempt was made to prepare the Grignard reiaent of p-bromo-
phenylbenzyl ethei. T1'r. ether wvs prepared by reeccing p-bromophenol
with benzylch..rid.e in acetone, .'te reutin-lg p7brtiphenylbenzyl
ether was recrysta,'lized twice from hot netlhno!. It wzas then diS-
s,-Ived in diethyl ethai, distilled from lithium a.lum'-num hydride.
Initial addition of the p-braphenylbenyl, et'-he to the -.>c.;n:liun
did not result ir iny -caction.r Cystals of iodii.r : :-d ethylene
bromide were employed to initiate the reaction. However, the addition
of further p-bromophenylbenzyv ether appea're to stop tche reaction.
Entrai~nenit and heat were eiloyed,I yet tt: re-.c:iOer' c;Old not be
\ / H
t /I) LiAH4
forced. Only a small amount of the magnesium reacted. Perhaps
another solvent, such as tetrahydrofuran, might yield better results.
The bis-Grignard reagent was prepared from p-dibromobenzene,
Tetrahydrofuran was employed as a solvent. It became necessary to
control the heat evolved in the reaction with an ice bath. The high
viscosity of the resulting Grignard reagent presented problems in
filtering and addition through the small opening in the dropping
B. Preparation of Substituted Phenylboronic Acids
The phenylboronic, or phenylboric, acids, 53, were prepared by
the reaction of substituted phenylmagnesium bromides with trimethyl
borated, followed by hydrolysis.04
(CH 0)3B H 30
X-,- MgBr --- -- X- -B(OII) + 2CH^OO
The trimethyl borate was added to dry diethyl ether and cooled under
nitrogen to -70 C in a dry ice-isopropanol bath., The appropriate
substituted phenylmagnesiumbromide was then added s3cwl' to the stir-
ring mixture. Too fast an addition resulted in formation of sub-
stituted biphenyls from coupling reactions, The resulting white solid-
slush was hydrolyzed slowly oy dropwise addition to the reaction mix-
ture, cooled in an ice bath. Addition without the ice bath resulted
in formation of a large yield of the substittutje t-rihenylCyclo-
triboroxine, 54. The cyclotriboroxenes were probably fo-nred by the
acid catalyzed dehydration of the phenylboro-ic ;cids. 53.
3 X ( B(H H30 /Bo
XPh N 0- PhX
The crude phenylboronic acids were recrystallized from hot
water. The following boronic acids were prepared by the above
method: phenylboronic acid (m.p. 214-216 C, 88.8%), p-chloro-
phenylboronic acid (m.p. 280-281" C, 83.5%), p-bromophenylboronic
acid (m.p. 271-2720 C, 93.5%), p-methylphenylboronic acid (m.p. 255-
2570 C, 31.0%), p-methoxyphenylboronic acid (m.p. 200-202* C, 69.0%),
and p-ethoxyphenylboronic acid (m.p. 120-122 C, 68.5%).
p-Phenylenediboronic acid 0was prepared by addition of a
tetrahydrofuran solution of the bis-Grignard reagent of p-dobromo-
benzene i-o a solution of trimethyl borate in tetrahydrofuran i.lder
nitrogen and cooled to -700 C. Addition presented problems due Lo
the thick, milky texture of the Grignard reagent, The resulting
solution was hydrolyzed with 15% sulfuric acid to give p-phpnyl-
diboronic acid (m.p. >3150 C, 74.5%).
C. Syntheses of Borate Esters of Substituted Phenylboronic Acdid
The substituted phenylboronic acids were converted to the-
corresponding dierhyl esters, 5_5, by reaction with ethanol, followed
by azeotropic distillation.10
x 5-- B(5(OH)2 + EtOH + H, 0
After all water had been distilled, the remaining solvent was re-
moved from the crude esters. The esters were then distilled under
vacuum. The following esters were prepared: diethyl phenylboronate
(b.p. 68-69 C @ 0.35 mm., 90%), diethyl p-bromophenylboronate (b.p.
76-77 C @ 0.25 mm., 41%), diethyl p-chlorophenylboronate (b.p. 92-
930 C @ 0.75 mm., 65.3%), diethyl p-methylphenylboronate (b.p. 58-
590 C @ 0.15 mm., 93.2%), diethyl p-ethoxyphenylboronate (b.p. 130-
1350 C @ 1.5 mm., 40%), diethyl p-methoxyphenylboronate (b.p. 108-
1100 C @ 2.0 mm., 60%), and tetraethyl p-phenyldiboronate (b.p. 102-
104 C @ 0.35 mm., 41%). Several of these esters have not been re-
ported previously in the literature.
After distillation of the substituted diethyl phenylborcaate, a
whitee residue remained in the distillation flask. These residues ve:e
recrystallized and examined by nmr, ir, and analyses. The nmr speztra
showed intense aromatic resonance signals in the 6 7.0 to 6 7.3 region.
The infrared spectra showed intense absorption in the 1350-1450 cMu.
regicT. assigned to the B-0 asymmetrical stretching modes. The melting
points were extremely high. The compounds were identical to those
formed by the dehydration of the substituted phenylboronic acids under
heat and vacuum. On this basis and elemental analysis, the structures
were assigned to be the corresponding substituted triphenylcyclocri-
boroxenes, 56. The following reaction would explain product formation.
0' 0 56
3 X-( -B(OEt) --- > + 3 EtOH
'- PhX o FhX
Several cyclotriboroxenes were recovered as by-products from
the distillation of their parent esters: triphenylcycloboroxene,
boroxene, tri(p-methylphenyl)cyclotriboroxene, tri(p-ethoxyphenyl)-
cyclotriboroxene, and tri(p-methoxyphenyl)cyclotriboroxene.
The distillation of tetraethyl p-phenyldiboronate yielded a
tan-colored solid which did not melt and was insoluble in all
common solvents tried. In acetone nte compound swelled and even-
tually became gel-like in structure. In analogy with the formation
of cyclotriboroxenes, 56, the compound was assumed to be a poly-
meric, 3-dimensional network, 57.
r i I.
B--- Ph OP Ph
IL -' /0
E. Preparation of Triethllamine-Phenylbor-anes
Several new triethylamine-phenylboranes, 58, were prepared by
modification of the method of Hawthorne. The p-substituted diethyl
phenylboronates were reduced in. the presence of triethyilamine at low
temperature under nitrogen. The reactions were carried out with slow
addition of the diethy phenyiboronate to a mixture of lithium a],in-
inum hydride in diethyl ether and triethylamine. A"ter complete di-
tion, the mixture was allowed to warm to room temperature, Filrration
followed to remove the excess lithiiru alu-ninum hydride ::d other in-
soluble salts. The triethylamine-phenyiboranes were isolated by
cooling the filtrate to -70 C and quick vacuum filtration under
X Q B(OEt)2 -, X-- -BH-2NEt3 + salts
Several of the triethylamine-phenylboranes exhibited peculiar
properties on attempted filtration. For example, the triethylamine-
(p-bromo)phenyl-borane formed a solid (crystals) at -72* C in a dry
ice-isopropanol bath. On removing the ether solvent, these crystals
seemed to melt away. At room temperature only a white residue re-
mained. This reaction could possibly be a disproportionation to
p-bromotriphenylboron, iriethylamine, and diborane. Evidence
against disproportionaticn was that the residue was insoluble in
every solvent tried, and the infrared gas spectrum showed no diborane
evolved. The second possibility would be the elimination of ethane
across the boron-nitrogen coordinate bond to give the monoborazene.
This process, though, usually requires heac and vacuum conditions.
Also, no ethane gas was evolved,
The following compounds were prepared, which are relatively
stable at room temperature: triethylamine-phenylborane (m.p. 65-6b6 C,
73%), triethylamine-(p-chloro)phenylborane (m,p. 62-63 C, 75%),
triethylamine-(p-methiyl)phenylborane (m.p. S0-830 C, 869,.), triethyl-
ainine-(p-ethoxy)phenylborane (m.p. 63-64' C, 84%), and triethylamine-
(p-toethoxy)phenylborane (m.p. 63-64' C, 82%).
The triethylamine-(p-ethoxy)phenylborane and triethylamine-
(p-methoxy)phenylborane did, however, decompose after one week,
even under nitrogen. These compounds and triethylamine-(p-bromo)-
phenylborane could be kept in toluene or benzene solvents without
F. Preparation of Pyridine-Phenylboranes
An attempt was made to prepare the di(triethylamine)-p-
phenylenediborane, 59, by the same general method as described
above. Tetraethyl-p-phenyldiboronate was added dropwise to a cold
mixture of lithium aluminum hydride in diethyl ether and triethyl-
amine. After complete addition, the resulting mixture was allowed
to warm to room temperature and filtered. The filtrate was cooled
to -72 C in a dry ice-isopropenol bath. No crystals were foimed.
The solution was concentrated to one-half its volume and cooled once
more. Fine, white crystals formed. These crystals were found to be
unstable in nitrogen and were, therefore, kept in the solvent. The
nmr spectrum showed only one mole of triethylamine completed to the
p-phenylenediborane. The ir spectrum showed the B-H stretching ab-
sorbance at 2200 to 2420 cm.
NtN HT- i --
It was thought that, perhaps, a different base, such as pyridine
with less face-strain, would complex more strongly with the p-phenylene-
diborane. The same general procedure was followed using pyridine as the
base. A yellow oil was obtained from the initial react:tjn.. This ol :as
crystallized from cold diethyl ether. A small amount of yellow crystals
(m.p. 60-68 C) was obtained. The crystals were unstable in air. The nmr
spectrum showed a broad multiple at 5 8.8, which was consistent with the
expected pyridinium resonance, and an aromatic resonance at 67.3. How-
ever, only one pyridine molecule was completed to the p-phenylenediborane.
A possible reason for the complexation of only a single base mole-
cule to the diborane could be a resonance or electronic effect. A signi-
B.-B B..-/ -B
ficant contribution from 60 would allow only single coordination with the
nucleophilic bases. This problem could possibly be eliminated by pre-
paring bis-boranes, 62, with insulation from electronic effects.
EtI- BI0 -32 Bp----^^
SYNTHESES AND REACTIONS OF BORON HETEROCYCLES
A. Reaction of Triethylamine-Phenylborane with N,N-Diallylaniline
Triethylamine-phenylborane was reacted with an equimolar quantity
of N,N-diallylaniline in refluxing toluene. After 12 hours, the
solvent was removed on a rotary evaporator. The residual, phospho-
rescent-green liquid was distilled under vacuum to give three fractions.
The first fraction proved to be unreacted N,N-diallylaniline. The
second fraction was identified as 1,2-diphenyl-l,2-azaborolidine; the
third was 1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane. A gumy
residue in the distilling flask defied all attempts at purification
dile to its insolubility. These products were consistent with those
reported by Butler and Statton.9
Examination of the two major, isolated products of the above
reaction indicated that a unit of 3 carbon atoms was needed to complete
tne balanced chemical equation (Scheme IV). Statton100 had suggested
that the loss of an allyl group would give the 1,2-dipher'yl-1,2-azaboro-
lidine, 61-. However, he was unable to trap or detect any alkcne during
the course of the reactions studied.
Since ona goal included improving the yield of the bicyclic
compound, 63, over that of the monocyclic compound, 64, while studying
the mechanism of the reaction, ic became important to isolate the miss-
ing 3-carbon unit. The reaction outlined in Scheme IV was repeated.
This time, however, the apparatus was modified to include a gas outlet
1-,2- Diphe:yl- ,2 ozaboro!idine
1,5 Dipheny -- zaa--boroaicyclo (3,3,0) octane
to an infrared gas cell. The triethylamine-phenylborane and N,N-
diallylaniline were dissolved in toluene in a flask equipped with
a magnetic stirrer and heater. The solution was slowly heated. Gas
samples, as well as solution samples, were taken at regular intervals
over a temperature range from 260 C to 1100 C. The boron-hydrogen
bond absorbance at 2340 cm.- was monitored, and it was found that
the initial hydroboration occurred at approximately 500 C. Little
change occurred in the intensity up to 970 C. At this point, the B-H
absorbance decreased with time. The infrared gas samples gave the
best information. Only triethylamine and toluene vapors were observed
below 92 C. However, at 95 C the spectra began to show traces of
propene gas. The broad, spiked peak at 910 cm. I was the most easily
followed absorbance. At 980 C large amounts of propene were being
pumped into the ir gas cell. The spectra were identical with the one
for propene given in the Sadtler Midget Edition, No. 6403.
After 12 hours, the green reaction solution was cooled, filtered,
and roto.;aced to remove the toluene. The viscous, green liquid was
transferred under nitrogen to a small, round-bottomed flask for vacuum
distillation. Again, the 1,2-diphenyl-1,2-azaborolidine, 64 (b.p. 8&-
860 C @ 0.30 imm.), was isolated. The nmr spectrum showed resoInance
signals at 67.2 (i, 10); 53.8 (t, 2); and 61.3 (n, 4), :corespc:ring
to the aromatic, carbon-5, and carbons-6 and-7, respectively (Figure 1),
The infrared spectrum exhibited an absorbance at 1389 cm~-, which was
assigned to the boron-nitrogen double bond.
1,5-lipheny-l-l-aza-5-borabicyclo(3.3.0)oc-ane, 63, distilled as a
light ye.liow oil (b.p. 125-130' C @ 0.30 mm.). This compound was
rccrystallized from hot ethanol, yielding white crystals (m.y, 80~-81i' C)
Figure 1. Nmr spectrum of 1,5diphenii 1aza5orabicyclo(3.3.0)octane
r i' i
4.0 3.o 2.0 1.0 0
ppm (8) t.)
Figure 1. Nmr spectrum of 1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane
The nmr spectrum is shown in Figure 1. The aromatic protons give
resonance signals centered at 66.85 (m, 10). The a protons give
a multiple (area 4) at 63.35. The b protons exhibit a distorted
quintet (area 4) at 62.1. The c protons are assigned to the multiple
(area 4) centered at 61.1. All chemical shifts are relative to tetra-
methylsilane in deuterochloroform. The B1 nmr spectrum showed a
chemical shift at 3.6 p.p.m. relative to trimethylborate. This chem-
ical shift value is strong evidence for the boron-nitrogen coordinate
bond since increased shielding around the boron atom would result in
an upfield shift. The chemical shift value for diethylamine-borane
is 3.1 p.p.m. and for the 3.3.0. bicycloborate esters of triethanolamine
The infrared spectrum showed a strong absorbance at 1285 cm. ,
which was assigned to the boron-nitrogen coordinate bond. The mass
spectrum showed a parent peak at 263 tl. The elemental analysis for
1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane agreed with that cal-
culated for the model compound.
Anal. Calcd. for C8 H22 N: C, 82.13; H, 8.42; 3, 4.11; and N,
-- 10 z22
5.32. Found: C, 82.15; H, 8.51; B, 4.21; and N, 5.41.
B. Mechanism of Formation of Azaborolidines and Azabo1rabicYclo-
The proposed mechanism for the reaction cf triethylaaine-
phenylborane with N,N-diallylaniline is outlined in Schemes V and VI.
The triethylamine-phenylborane, 65, dissociates to give the free
phenylborane, 66, and triethylamine, One of the allyl groups of the
N,N diallylaniline, 76, undergoes a single hydroboration to give the
I,---.. ~ N -5
I, s ---F;-
+ prope ne
uncoordinated intermediate, 69. This involves a four-membered
transition state, 68. In solution there is probably an equilibrium
between the uncoordinated, 69, and the coordinated, 70, intermediates.
Above 90 C in toluene, competitive pathways (Scheme VI) would lead
to formation of the final products 71 and 72 (X=H). The 1,5-diphenyl-
l-aza-5-borabicyclo(3.3.0)octane, 72, could be formed by a second
hydroboration with boron adding to the terminal end of the remaining
double bond. This might occur by either pathway 1 or pathway 2.
Brown66 proposed a four-centered transition state for the general
process of hydroboration. Following this proposal, the transition
state for the coordinated fonr, 73, would have much more strain than
that of the uncoordinated form, 74. Therefore, it seems likely that
the second hydroboration proceeds along pathway 2,
Tle foi.r-tion of the 3,2-jiphenyl-1,2-azaborolidine, 71, is
accompanied by the elimination of propene gas. Amnine-boranes have
been reported to dealkylate at very high temperatures under vacuum,
as discussed in the introduction. Obviously, this system is a special
one which allows for very facile propene elimination at relatively
low temperatures, atmospheric pressure, ancr mild conditions.
The coordinated intermediate, 70, is perfectly si'ited for a
concerted elimination of propene as shown below:
-X--X X-: piopene
A six-membered transition state would allow for a relatively
strain-free addition of hydrogen to the terminal end of the double
bond of the allyl group and the ensuing electron shifts.
A second mechanism may be proposed which would lead to the
same products (Scheme VII). The triethylamine in the reaction
mixture could act as a base in abstracting a proton in the 1,5-
diphenyl-l-aza-5-borabicyclo(3.3.0)octane, 72, at either of the
two beta-carbons to the quaternary ammonium group. This is the
well-known Hofmann elimination.0 The resulting allylborane, 75,
then could conceivably undergo a thermal cleavage to give the aza-
borolidine, 71, and propene. Alkylboranes have been reported to
undergo this type of thermal cleavage.64
The most important aspect of the concerted mechanism (Scheme
VI) involves the cquilibri~u between structures 69 and 70. This
equilibrium must likely exist for the competetive formation of
the mcnocyclic, 71, and bicyclic, 72, products. A stronger
coordination, or shift toward intermediate 70 of the equilibrium
should favor the monocyclic product. A shift toward the un-
coordinated form should yield more of the bicyclic product. The
most obvious studies, therefore, would be directed toward deter-
mining tihe electronic effects of substituents demonstrated beicw -
/ \\ ),-,
Electron-withdrawing subscituents in the para-position should
increase the electron deficiency on boron aTd would shift the equi-
librium toward the coordinated form of the intermediate, 76.
Electron-donating substituents would be expected to have the
opposite effect. Electron-donating substituents on the aniline
ring would increase the basicity of the amine, resulting in
stronger coordination. Again, the opposite effect would be
expected for withdrawing substituents.
The Hofmann elimination mechanism would also be expected
to show similar effects of substitution. In this case, the
elimination would be possible only if the quaternized form,
72, were present.
C. Deuterium Labeling Studies
After several syntheses of substituted l,5-diphenyl-l-aza-
'-borabicyclo (3,3.0)octanes (discussed later in t!h-is 7-r ),
it was evident that substituent effects would not give a clear-
cut distinction between the concerted rrechanism and the Hofmann
elimination. Therefore, deuterium labeling studies were proposed.
If triethylamine-dideutero-phenylborane could be prepared, it
should then be possible to distinguish quite clearly the two pro-
The proposed concerted mechanism is shown in Scheme VIII.
The triethylamine-dideuterophenylborane, 78,would dissociate to
give the free dideuterophenylborane, 79, which would undergo an
initial deuteroboration to give intermediate 80 and the coordinated
form of the intermediate, 81, by an equilibration of the two forms.
The deuterium ends up in the 3-position on the ring. A second
deuteroboration through pathway A would lead to the dideuterated
x-< )H-B--,.B3 --- <- ( xy .... -.,NEis
f_; g-- D:))-/ D
-+ ,-x .--
DC cr~CH --CH 2
bicyclic product, 84. A concerted elimination, with deuterium
adding to the terminal end of the double bond in intermediate 81
and the ensuing electron shifts, wguld yield the mono-deuterated
azaborolidine, 82, and 3-deuteropropene, 83.
The expected Hofmann elimination mechanism is shown in Scheme
IX. The triethylamine in solution would attack the dideuterated
1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane, 84, at a hydrogen
atom or deuterium atom on the beta-carbon to the quaternary nitro-
gen. Abstraction of a proton should be slightly favored over
deuterium abstraction, due to an isotope effect. Intermediate 85
could then undergo thermal cleavage to give the mono-deuterated
azaborolidine, 82, and 2-deuteropropene. If a deuterium were
abstracted rather than a proton, propene gas would be generated.
Triethylamine-dideuterophanylborane was prepared by reducing
dicthyl phenylboronate with litbiuni aluminu~n deuteride i-: ether
with triethylamine at low temperature. The triethylamine-dideutero-
borane crystals (m.p. 64-655 C) were characterized by mass spectral,
nmr, ir, and elemental analyses. The infrared spectrum showed a
broad boron-deuterium absorbance at 1680 to 1765 cm. .
N,N-Diallylaniline was added to an equimolar quantity of the
triethylamine-dide uterophenylborane in p-xylene. The reaction
vessel was provided with a gas outlet, which was attached to an
infrared gas cell and, in turn, to a dry ice--isoprcpanol gas trap.
Several gas samples were taken at: various temperatures. Above
950 C deuteropropene gas was evolved. The infrared spectrum is
given in Figure 2. The absorbance at 2175 em. was consistent:
with that expected for the alylic carbon-deuterfrm stretching
WAVE'LENGTH IN MICRONS
Figure 2. Infrared spectrum of 3-deuteropropene.
r .'c Hl II
~ I /
Ij i ii
--lr. J~~~~~ Ic.c.: ,!.. .~~... rr-Llr-* ---~-U-r j1^I -~
E 55.; .3G -.
Figure 3. Nmr spectrum of 3-deuteropropene.
frequency. The mass spectrum gave a parent peak at 43, consistent
with the molecular weight.
The nmr spectrum (Figure 3) of the gas was obtained by adding
deuterochloroform to the trapped liquid propene. Resonances were
assigned as follow: the allylic protons at 51.7 (finely split
signal, area 2); protons a and b at 65.0 (finely split triplet, 2);
and proton c at 65.8 multiplee, 1). The above spectral evidence
surely indicates that the trapped gas was 3-deuteropropene rather
than 2-deuteropropene. 2-Deuteropropene would have shown no signal
in the 65.8 region and, of course, would not have shown the 2:2:1
At this point, a model gas, 1-butene, was obtained to be sure
that the chemical shift assignments for 3-deuteropropene were correct.
The only expected difference in the two nnmr spectra would be the
added splitting due to the methyl group being substituted for
deuterium. The allylic proton signal moved slightly to 62.0 and was
a slightly distorted, finely divided quintet. The signals at 65.0
and 65.8 were almost identical to those of 3-deuteropropene.
The structure of the 3-deutero-1,2-diphenyl-l,2-azaborolidine,
82, was confirmed by mass, nmr, and ir spectra
The structure of the 3,7-dideutero-l,5-dipbeinyl-l-aza-5-hora-
bicyclo(3.3.0)octane, 84, was also confirmed by spectral data and
analysis of the high-boiling fraction (b.p. 64-680 C @ 0.15 mm.).
This material was recrystallized to give white, flaky crystals
(m.p. 78-79 C). The infrared spectrum exhibited absorbances
characteristic of the dipheny3-aza-borabiryclc(..3.0)octanes. In
addition, the carbcu-deuteriumn absorbance appeared a'it 2150 cm.
WAVELENGTH IN MICRONS
Figure 4. Infrared spectrum of 3,7-dideutero-l,5-diphenyl-l-aza-
.... ..-.L _. .. .*- I I I 1 I I I I *
4.0 3.0 2.0 1.0 o
Figure 5. Nmr spectrum of 3,7-dideutero-1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane.
(Figure 4). The nmr spectrum of 72 is given in Figure 5. Resonances
occur at 61.1 (m, 2); 62.15 (broad m, 1); 63.45 (m, 2); and the aromatic
resonance at 67.0 (m, 5). The mass spectrum gave a parent peak at
265 -1, corresponding to the correct molecular weight. The analysis
agreed with structural assignment.
Anal. Calcd. for C8H20D 2BN: C, 81.52; H, 7.60; D, 1.52; B, 4.08;
and N, 5.28. Found: C, 81.74; B, 4.20; and N, 5.18.
D. Preparation of 3-Deuteropropene by Alternate Method
To cast aside any lingering doubts as to the identity of the
isolated gas, 3-deuteropropene was prepared by an alternate method.
The infrared and nmr spectra were compared.
Allylmagnesium bromide was added to deuterium oxide in a small
flask equipped with a gas outlet. Gas samples were trapped it' an
infrared gas cell. Samples for the nmr were obtained by liquefying
the gas in a trap, cooled by dry ice-isopropanol, and mixing with
The infrared and nmr spectra were identical to those of the
3-deuteropropene, 83, obtained from the reaction in Schere VIII.
E. Preparation and Reactions of -Substituted 1, .5-Diphenv1-1-aza -
The next goal was the synthesis of a series of substituted 1,5-
diphenyl-l-aza-5-borabicyclo(3.3.0) octanes which would have functional
groups in the para-positions of the phenyl rings capable of conversion
The most obvious approach might appear to be that of electro-
philic aromatic substitution. However, boron is, in theory, a
better leaving group than hydrogen in these substitutions. Tri-
phenylboroxene was reacted with aluminum chloride and, also,
phosphorus pentachloride to yield benzene as the only identifiable
product.110 The boron atom on triphenylboroxene, or triphenyl-
cyclotriboroxene, however, might be a better leaving group than
the coordinated structure of the 1,5-diphenyl-l-aza-5-borabicyclo-
(3.3.0)octane, 63. Li 11 reacted 63 with acetyl chloride and
aluminum chloride to give acetophenone and other unidentified
.63 ACI4 4
A second approach would involve nucleophilic aromatic substi-
tut.Icn. .,o major problems present themselves in this method. Both
are due to the stre:.gt'n of the bases employed in these substitutions.
If an equilibrium exists between the coordinated, 63, and un-
coordinated, 63A, forms of the bicyclic system, different chemical re-
activity would be expected for the two forms. Strong bases would be
expected to attack the electronically-deficient boron atom in struc-
BN- Q '
Structure 63 would be susceptible to Hofmann elimination by
strong bases as shown below:
Some experimental data support the possibility of cleavage
of the bicyclic unit by the above mechanism. The 1,5-diphenyl-
l--aza-5-borabicyclo(3.3.0)octane system was destroyed by .'
reaction with butyllithium. The starting material was not
recovered, but tne products were not identified. Trinapthyl-
borane was reacted with methanol to give trimethoxybcrane.112
The only feasible approach in preparing 1,5-diphenyl--azn-
5-borabicyclo(3.3.0)octane with phany! substituents was to start
with the substituted bromobenzenes and to build the desired pro-
ducts. The synthesis of trietbylaitine-phenylboranes (Scheme III)
involves Grignards, acids, lithium aluminum hydride, alcohols,
heat, and Lewis bases. The preparation of substituted NN-diailyl-
anilines involves the heating of a sodium carbonate-water solution
to reflux. The desired substituents must also be readily converted
to monomer without destroying the bicyclic structure.
Preliminary studies of the 1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)-
octane showed the stability to boiling water, dilute mineral acids, tri-
ethylamine, and dilute sodium hydroxide. The compound was not stable
to peroxides in sodium hydroxide or strongly acidic conditions. Oxi-
dizing conditions with bromine in acetic acid destroyed the bicyclic
Carbon-boron bond cleavage has been reported to occur with n-butyl-
lithium, bromine, peroxides, and acetylchloride in pyridine.1
Several previously unreported, substituted 1,5-diphenyl-l-aza-5-
borabicyclo(3.3.0)octanes were prepared along with their monocyclic
by-products. These are listed in Table 2. Each synthesis involved
essentially the same procedures as were outlined for the model com-
pounds. Some modifications were, however, necessary in nearly every
case due to the new functional groups. The average time required for
synthesis and isolation of each pair of bicyclic and monocyclic cos--
pounds was approximately 4 to 5 weeks. Physical properties, nmr, ir,
mass spectra, and elemental analyses are given in Tables 3 and 4 for
the bicyclic compounds. The details of each synthesis are given in
the experimental section and will not be repeated in this chapter.
Instead, the preparation of 1,5-bis-(4-chloropbenyl)-l-aza-5-borabi-
cyclo(3.3.0)occane. 93, and 1,2-bis-(4-chlorophenyl)-1,2-azaoor-lidine,
94, will be presented to give the reader an idea of the approach to
synthesis, separation, work-up, and analysis of these compounds.
Triethylamine-(p-chloro)phenylborane was dissolved in a large
volume of toluene in a round-bottopied flask under nitrogen. The
contents were heated to 550 C, and then the p-chloro-N,N-diallyl-
aniline was added dropwise to the stirring mixture. As the reactio-.
(3.3.0)OCTANES AND 1,2-DIHENYL-1,2-AZABOROLIDINES
j )- Y
-\_. ^ _T
Cpd. No. X
63 H H
87 H Er
89 H Cl
91 Ci HC
U. 3 Cl Cl
95 Br H
97 Dr Br
'OC i, OCH3
1 "C3 ,CHC CCHI3
Cpd. No. X Y
64 H H
88 H Br
90 H Cl
92 Cl H
94 Cl Cl
96 Br H
98 Br Br
100 CH3 CIH
104 OC23 OCH3
102 OCH3 OCH
104 OCH2CH3 OCH3
TABLE 3. PHYSICAL DATA OF SUBSTITUTED
aza-5-borabicyclo( 3.3.0. )octane
91 l-Phenyl,5-(4-c l--. p.eiyl)-l-
aza-5 -borabicyclo (3.3.0.)octane
C H 21ENBr
C H2i ::C1
Analysis (Calculated, Found)
C,82.13; H,8.42; N,5.32; B,4.11
C,82.15; H,8,51; N,5.41; B,4.21
116-118 C: C,63.19;
H,7.60; D,1.52; N,5.28;
H,6.19; N,4.09; B,3.16;
H,6.29; N,3.96; B,2.90;
H,7.11; N,4.70; B,3.64;
H,7.06; N,4.59; B,3.73;
H,6.07; N,4.22; B,3.26;
H,5.99; N,4.15; 01,21.13
TABLE 3 (continued)
Analysis (Calculated, Found)
C20H26BN 107-108 C:
H,4.78; N,3.33; B,2.57;
C,82.46; H,9.00; N,4.81; B,3.72
C,82.21; H,9.05; N,4.96; B,3.71
5-borabicyclo (3.3. )octane
C20 H2 T)j'2
C2 H 2BNO2
112-114 C: C,74.32;
CoH 36B 2N2Cl2 198-200
3.0 36 i'
H,8.11; N,4.33; B,3.35;
H,8.01; N,4.16; B,3.36
C,80.69; H,8.89; N,5.88; B,4.54
C,80.39; H,8.71; N,5.70; B,4.72
H,7.02; N,5.42; B,4.18;
H,6.86; N,5.45; B,4.27;
TABLE 4. SPECTRAL DATA OF SUBSTITUTED
* Chemical shift relative to trimiethyl borate
** Chemical shift relative to tetramethylsilane
TABLE 4 (continued)
1273 and 1286
1260 and 1273
proceeded and the temperature raised, a brilliant green color appeared.
Infrared gas samples indicated the formation of propene gas. The reac-
tion was allowed to proceed at 920 C for 24 hours. The contents were
cooled, filtered, and then the solvent was removed on a rotary evapo-
rator. The residual liquid was transferred to a small flask for dis-
tillation on a spinning band column. Several fractions were collected.
The first fraction proved to be unreacted p-chloro-N,N-diallylaniline.
The second fraction was a clear, viscous liquid (b.p. 130-135 C @ 0.20
mm.). This compound turned brown on exposure to air. The third frac-
tion was the same compound. The infrared spectrum was consistent with
the azaborolidine structure, 94, with the B=N absorbance assigned at
1400 cm. -. The mass spectrum gave a parent peak at 289 1, consistent
with the molecular weight. The nmr spectrum (Figure 6) shows resonances
at 57,05 (m, 8); 63.75 (t, 2); 61.7 (m, 4) for the aromatic, a, and b.c
Anal. Calcd. for C ,H 1BNC1 : C, 62.09; H, 4.87; N, 4.83; B, 3.73;
la 14 2
and Cl, 24.46. Found: C, 61.78; H, 5,01; and Cl, 24.45.
The fourth fraction (b.p. 170-175' C) was very viscous and solidi-
fied in the side arm. This material was recrystallized from ethanol to
give a crystalline solid (m.p. 93-950 C). The ir spectrum gave a strong
absorbance at 1275 cm.-1 which was assigned to the B-4 coordinate bond.
The mass spectrum gave a peak at 332 *1, corresponding to the correct
molecular weight. The nmr spectrum (Figure 6) showed resonances at 67.2
(m, 8); 63.4 (m, 4); 62.15 (q, 4); and 61.15 (in, 4) for the aromatic, a,
b, and c protons respectively.
Anal. Calcd. for C H 20BNCi,: C, 65.08; H, 6.07; N, 4.22; B, 3.26;
and Cl, 21.36. Found: C, 64.97; H, 5.99; N, 4.15; and Cl, 21.13.
-~~~ ~~ t' .r-,-. -~---- r~u --
Figure 6. Nmr spectra of 1,5-bis-(4-chlorophenyl)-l-aza-5-borabicyclo-
(3.3.0)octane and 1,2-bis-(4-chlorophenyl)- ,2-azaborolidine.
-i-r ~l-r.---+- --i ui--C----. -~- -)- -J--..rL--c--~---Lu~~- -1' C- ~ 1
LiiL/lt /*-Wy^' -.!'
A comparison of the spectra in Figure 6 reveals some interesting
effects. A look at models gives some insight into these effects. In
the monocyclic structure, 94, the phenyl rings are far apart and the
ring must be nearly planar. As a result, the b protons are moved
upfield and the c protons are deshielded, resulting in an overlap of
the protons on carbons b and c. The a protons are also slightly
deshielded, probably due to the planar positioning of the phenyl groups
relative to the 5-membered ring. In the bicyclic structure, 93, the
a, b, and c protons give three distinct resonance absorptions. The
model of this compound shows the phenyl groups tied back, much like
butterfly wings, relative to the B-N bond axis. The two propylene
bridges are also forced back in the opposite direction to the phenyl
groups. The b_ protons are in the region expected for 5-membered,
bicyclic rings. The a and c protons are s-ielded relative to the
azaborolidine, resulting in upfield shifts of 4 co 5 p.p.m.
The B1 nmr spectrum of 93 exhibited a chemical shift of 2.2
p.p.m., relative to trimethyl borate. This compares with 3.6 p.p.m.
for the unsubstituted 1,5-diphenyl-l-aza-5-borabicyclc(3.3.0) otane.
F. Preparation of Bis- 14-15--(4-methj Lphenvl)-l-aza-5-borabicyclo-
(3.3.0)octyl]benzene and Bis-1,4- [5- (4-hioEronvl )-l-aza-5-bora-
Bis-1, 4- [5- (4-me thy lphenyl)-l-aza-5-bora-bicyclo ( 3..0) oetyl] -
benzene, 104, and bis-l,4-[5-(4-chlorophenyl)-l-aza--5-bcrabicyclo-
(3.3.0)octyljbenzene, 105, were prepared by reacting N,N,N',N''-
tetraallyll--pphcnylen di a-ine with tried thy la:mine-(4-;:e thy ) phenyl-
borane and triethylamine(4-chicro)phenyiborane. These compounds,
of course, were not distillable and were purified using silica ge.i
chromatography and recrystallization. Physical and spectral data are
given in Tables 3 and 4.
CH Q ----CH3 Cl- -Cl
The high molecular weights of these compounds and their molecular
structures make them excellent models for the study of polymers with
similar structures. The boron-nitrcgen coordinate bonds in these com-
pounds exhibited double absorbances in the 7.8 to 8.0 micron region
compared to single absorbances in the 1,5-diphenyl-l-aza-5-borabicyclo-
(3.3.0)octaees. These compounds were the best models for differentiAl
scanning calorimetry studies, described later in this chapter.
G. Reaction of N,N,N',N'-Tetraallyl-p-phenylenediaamine with
The relatively high yields of 104 and 105 encouraged attempts
to prepare oligomeric or polymieric products. It seemed that the
above synthesis should be extended by reacting a bis-borane com-
pound, 107, with 1,N,N',N'-tetraallyl--pphenylneediamiue, 106, to
yield a polymer, 1.08. Attempts to prepare the pure di(pyridine)-
p-phenylenediborane and di(triethylamine)-p-phenylenediborane are
described in Chapter III. The resulting crude compound, 107, was
reacted with the tetraallyl compound, 106. A viscous, brown com-
pound was obtained. This compound was insoluble in every solvent
CH2=CHCH CHCH=CH rR-
2 2,N & /z 2c R.oN
2 2 2 107
tried, but swelled to a gel-like material in acetone after several
weeks. The material appeared to be cross-linked, as might be
expected by examination of the above equation.
H. !Boron Nuclear Magnetic Resonance Studies
11 nmr spectra were obtained for the substituted bicyclic
compounds listed in Table 4. The chemical shifts in chlorof rm
were measured in p.p.m. relative to trirethy borate by a Varian
XL-100 instrument. The chemical shift values varied from 1.9 to
5.2 p.p.m. for the compounds studied. These shifts are very close
to those given in the literature for dietbylamine-borane (3.2 p.p.m.),
pyridine-borane (4.7 p.p.m.), and for the (3.3.3)bicycloborate
ester of the triethanolainines (3.2 p.p.m.).8 The chemical shift
of these bicyclob.,rate caters to somewhat higher fields than other
alkyl borates was attributed to increased shielding around the
boron atom duep to the boron-nitrogen coordinate bond, thus increasing
the tetrahedral character and the number of bonding electrons around
the boron atoms.
Some of the iB cheraical shifts in Table 4 might be interpreted
in light of the above discussion. The electronic effects of the
phenyl substituents on the bicyclic structure 109 should be reflected
in the boron-nitrogen bond strength and in the tetrahedral character
of this bond. The bicyclic structure, for which both _x and y are
x 0 --y
chlorine atoms, gives a chemical shift of 2.2 p.p.m. When x is
chlorine and y is hydrogen, the chemical shift is 5.2 p.p.m. This
may be interpreted as a result of the withdrawing effect of chlorine
increasing the electron deficiency of boron and, therefore, strength-
ening the B-N coordinate bond. If x is hydrogen and y is bromine,
the chemical shift drops to 1.9 p.p.m. This effect is consistent
with a decrease in the basicity of the amine, thus decreasing the
B-N bond strength. Less reduction is noticed when x is hydrogen and
y is chlorine; this compound, 89, gives a chemical shift of 2.8 p.p.m.
In the dibromo-bicyclic compound, 97., electronic effects of the sub-.
stituents might be expected to cancel each other. The chemical shift,
3.1 p.p.m., is close to that of the unsubstituted compound, 63, at 3.6
p.p.m. The symmetrical bicyclic compounds 104 and 105 show chLem.cal
shifts of 4.4 and 4.7 p.p.m., respectively.
I. TeTrperature Studies by Differential Scanning Calorimetry
A Perkin-Elmer DSC-1B Differential Scanning Calorimeter was used
for studies of the B-N coordinate bond dissociation for all substituted
1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octanes which were prepared.
Samples were weighed and then placed in special sample holders to
reduce volitalization. Temperature scanning increments were set at 20
per minute. Melting points were easily observable and served ao
excellent calibration marks. At higher temperatures, sublimation of
the samples occurred. No decomposition peaks were observed prior to
sublimation, showing the remarkable stability of these relatively
heavy compounds. A broad endotherm appeared in all spectra beginning
at 520-540* K. This temperature seemed to be consistent with that
expected from B-N coordinate bond dissociation. However, the peak
area could not be measured due to volitalization of the sample,
beginning at a slightly higher temperature.
The bis-l,4-[5-. (4-methylphenyl)-I-aza-5-borabicyclo(3.3.0)octyl]-
benzene, 104, and the bis-l,4 -[5-(4-chlorophenyl)-l-aza-5-borabicyclo-
(3.3.0)octyl]benzene, 105, were much better models for the temperature
studies. These compounds sublimed at a much higher temperature and
allowed the observance of endotherai.s which seemed consistent with
those expected for the boron-nitrogen bond dissociation. Broad,
double peaks were observed beginning at 5320 K and ending at 6170 K
for compound 104. Similar peaks were observed for compound 105
beginning at 5410 K and ending at 6270 K. The heat of this transition
for compound 10_4 was calculated to be 60 cal./g. or 28.5 kcal./mole.
Compound 105 gave a value of 58.2 cal./g. or 30.3 kcal./mole. These
values are very close to those predicted by Butler.1
J. Synthesis of 1,6-Dipbeny-l--aza-6-borabicyclo(4.4.0)decane
The final synthetic effort in this research was an extension
of the hydroboration by triethylamine-phenylborane of N,N-di-
3-butenylaniline, 10. This reaction would be expected to yield
1,6-diphenyl-l-aza-6-borabicyclo(4.4.0)decane, 114. 1,2-Diphenyl-
<0 4. Cy
azaboracyclohexane, 113, should not be formed by a concerted
mechanism as was the case for the azaborolidines in our previous
studies (Scheme X). This compound could, however, conceivably be
formed by the pyrolysis of intermediate 111 at high temperature
under vacuum. The elimination could be tested, as before, by
monitoring gases produced from the reaction in p-xylene. If the
competing pathway 1 were followed at the reaction temperature,
butene gas would be detected. If only pathway 2 were followed,
the bicyclic compound should be formed in relatively higher yield.
This higher yield, obviously, would be more attractive for monomer
preparation than the yields obtained for the 1,5-dipheny-l--aza-
N,N-Di-3-butenylaniline, triechylamine-phenyiborane, and p-
xylene were slowly heated under nitrogen. Gas samples were taken
at various temperatures and time intervals. A bright green color
developed during the course of the reaction, but no butene gas was
evolved during heating. The resulting solution was filtered, roto-
vaced, and divided into two portions. The first portion was then
dissolved in benzene and passed through a neutral silica gel column.
The eluted solution was rotovaced. A white, gummy material was
obtained. This was recrystallized from ethanol to give white
crystals (m.p. 143-145 C). The compound was consistent with the
structure assigned to 1,6-diphenyl-l--aza-6-borabicyclo(4.4.0)decane.
The mass spectrum gave a parent peak at 291 tl. The infrared spectrum
gave a broad absorbance at 1280 cm. consistent with that expected
for the B-N coordinate bond. The nmr spectrum exhibited resonance
signals at 60.95 (m, 4); 61.6 (m, 8); 63.5 (m, 4); and 67.3 (m, 10).
Anal. Calcd. for C20H26BN: C, 82.48; H, 9.00; B, 3.71; and
N, 4.81. Found: C, 81.98, H, 8.61; B, 3.64; and N, 4.76.
The second portion of the reaction mixture was distilled under
vacuum on a spinning band column. Several fractions were obtained.
The first fraction proved to be unreacted N,N-di-3-butenylaniline.
The second fraction (b.p. 75-78 C @ 0.50 mm.) exhibited absorbances
in the infrared spectrum and resonances in the nmr spectrum con-
sistent with the structure of 1,2-diphenyl-azacyclohexane, 113.
The third fraction (b.p. 90-95 C @ 0.50 mm.) was a pale yellow
liquid. The spectra were consistent with the assignment of 1,2-
diphenyl-l-(3-butenyl) ,2-hydro-azaboracyclohexane, 112. The nmr
spectrum showed the presence of an allyl group with resonances at
65.8 (m, 0.5); and 65.1 (m, 1). Other resonances for compound 112
occurred at 60.8 (m, 2); 61.65 (m, 6); 63.35 (m, 4); and 67.4 (m, 12).
The ir spectrum showed the presence of a C=C double bond at 1645 cm.
as well as the presence of a B-H bond at 2250 cm. The distillation
flask contained a gummy, brown residue, which was recrystallized from
ethanol to give a white, crystalline material. This compound was
identical to that obtained from the silica gel column chromatography.
The above results gave strong support to the mechanism proposed
for the facile, concerted elimination of propene gas in the competitive
formation of l,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octanes, 72,
and 1,2-diphenyl-1,2-azaborolidine, 71 (Scheme VI).
This research has attained many of its projected goals. The
model compound, 1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)octane was
prepared and its properties investigated. All of the products of
the reaction of triethylamine-phenylborane with N,N-diallylaniline
were isolated, purified, and carefully characterized by nmr, ir,
mass spectral, and elemental analyses. Propene gas was isolated
and identified as a by-product.
Mechanistic studies with deuterium labeling gave strong
support to the proposed mechanism of the facile, concerted elimi-
nation of propene. These studies ruled out a Hofmann-type elimi-
The stability of the model compou-:nd, 49, to various synthetic
conditions, such as thy;:.e necessary for iu!Icleophilic or' electro-
philic aromatic substitutions, waq deternmied. Its reactivity in
various solvents, acids, and bases was also studied.
Several substituted derivatives of 49 have been prepared which
should be excellent monomer precursors. P3relimTinary studies .sh
evidence for the formatiecn of the bis-Grignard reagt from ~ 5-
bis- ( '--bromophcnyl-l--..n. .-' -5-bor~-abi .yclc, (3.3.0 )oc.t-.::. e corra-
sponding bis.-(4- -methcx:yphenyl) compound has be-n converted to snall.
amount& of the 1,5-bis- (4-hydro;eyphenyl' --l--aa-borabi.cyclo (3.3.0)-
octarn by the reaction of lithiiml iodide in collidine. Thess two
reactions have excellent potential fir the formation of dicarboxylil
and diphenolic derivatives of -9. These compounds 1.ould be interest-
ing co-monomers for several] possible condensation pol7ym-erization
reactions. Several other compounds listed in Table IV have function-
ality which could be utilized in monomer preparation. In all, over
40 new compounds have been synthesized and characterized in this
The boron-nitrogen coordinate bond has been investigated
extensively by nmr, ir, 1B nmur, and by differential scanning
calorimetry. Chemical shifts, boron-nitrogen stretching frequencies,
and heat of transition studies have given us valuable information
which has opened many avenues for future research on these hetero-
cyclic compounds and other related systems.
A. Equipment and Data
All temperatures are reported uncorrected in degrees centigrade.
Nuclear magnetic resonance (nmr) spectra were obtained with a Varian
A-60A Analytical NMR Spectrometer. The chemical shifts were measured
in deuterochloroform, unless otherwise specified, relative to tetra-
Infrared spectra were obtained with a Beckman IR-8 Infrared
1IBorcn nuclear magnetic resonance spectra were obtained,
courtesy of Dr. Wallace S. Brey, with a Varian X-L 100 High Resolution
Mass spectra were obtained with a Hitachi Perkin-Elmer RilU
Thermal analyses were carried out on a Perkin--Elmer Differential
Scanning Calorimeter DSC-1B.
Elemental analyses were performed by Galbraith Laboratories,
Inc., Knoxville, Tennessee.
B. Syntheses and Characterization
Preparation of N,N-Diallylaniline
Aniline (186.0 g., 2.0 mol.), sodium carbonate (420.0 g., 4.0 mol.),
and 300 ml. of water were placed in a 3-liter, 3-necked, round-bottomed
flask equipped with an overhead stirrer, dropping funnel, and water-
cooled condenser. The mixture was heated to low reflux using a
heating mantle and variac. Allyl bromide (484.0 g., 4.0 mol.) was
added dropwise over a period of 4 to 5 hours with constant stirring.
The brown, syrupy liquid which resulted was stirred for 4 more hours
and then filtered into a 2-liter separatory funnel. The two resulting
layers were separated and the aqueous layer washed with 3 100 ml. por-
tions of ether. The ether was removed on a rotary evaporator. The
resulting residue was combined with the crude organic layer. This oily,
brown liquid was then added carefully to the 30% solution of sodium
hydroxide in 300 ml. of water. This solution was then transferred to
a 3-necked, 3-liter, round-bottomed flask equipped with mechanical
stirrer and addition funnel. Benzenesulfonyl chloride (100 mi.) was
added dropwise with stirring. T:.r resulting slush was cooled in a:-
ice bath and neutralized with 10% hydrochloric acid. The organic
layer was separated and dried with sodium sulfate. The crude N,,- di-
allylaniline was filtered and placed in a 1--liter, round-bottomed
flask for distillation. Using a vacuum distillation apparatus, consist-
ing of a Vifereaux column, fraction cutter, and receiving flasks, an
97.6% yield' of a clear liquid (293.0 g., b.p. 104-i06 C ( 3.5 mm.) was
Synthesis of p-methyl-N,N-diallylaniline
p-Toluidine (214.0 g., 2.0 mol.), sodium bicarbonate (420.0 g.),
and 1000 ml. of water were placed in a 3-liter, 3-necked, round-
bottomed flask equipped with thermometer, overhead stirrer, addition
funnel, and condenser. The mixture was heated on a heating mantle
until the temperature of the liquid reached 700 C. Allyl bromide
(581.0 g., 4.8 mol.), was added dropwise with stirring. The resulting
mixture was heated overnight. The brown liquid separated into 2 phases
on sitting. The layers were separated using a 1-liter separatory
funnel. The oily layer was washed with 3 100 ml. portions of water
and then dried over sodium sulfate. The crude p-methyl-N,N-diallyl-
aniline was filtered and then vacuum distilled yielding 262.3 g. (70.1%)
of a pale yellow liquid (b.p. 80-850 C. @ 0.350 mm.).
The infrared spectrum (ir) gave absorbances at 1520 (m), 1240 (s,
split), 800 (m), 1620 (s, split), 920 (s, split) 1340 (m), 1360 (s),
1390 (s), 1410 (m), 1425 (m), 1450 (w), 1180 (s), 990 (s), 2990 (s),
2950 (s), 3000 (s), 3020 (s), and 3110 (m) cm. .
The nmr gave resonance signals at 68.8 (m, 4); from 64.9 to 66.2
(m, 6); 63.8 (m, 4); and 62.2 (s, 3).
Synthesis of p-bromo-N ,N-diall.ylani ine
p-Bromoaniline (100 g., 0.582 mol.), sodium carbonate (126.0 g.,
1.2 mol.), and 230 ml. of water were placed in a 2-liter, 3-necked,
round-bottomed flask equipped with a mechanical stirrer, drcpi.-:n
funnel, and water-cooled condenser. The mixture was heated on a heat-
ing mantle until low reflux was obtained. Allyl bromide (146.0 g.,
1.2 mol.) was added over a 3 to 4 hour period through an addition
funnel. Stirring was continued for 12 hours, after which a syrupy,
brown liquid was obtained, along with an aqueous layer. The layers
were separated in a 1-liter separatory funnel. The aqueous layer
was washed with 2 350 ml. portions of ether. The ether was then re-
moved on a rotary evaporator and the resulting layer combined with
the crude p-bromo-N,N-diallylaniline. This crude liquid was added
dropwise to a 30% solution (100 ml.) of sodium hydroxide in water
in a 2-liter, round-bottomed flask equipped with a mechanical stirrer
and addition funnel. Benzenesulfonyl chloride (34.0 ml.) was added
dropwise. Since the reaction was exothermic, an ice bath was used
to control the temperature. The resulting cooled solution was neu-
tralized with 10% hydrochloric acid solution. The organic layer was
separated and dried with sodium sulfate. The crude product was fil-
tered and placed in a 200 ml. round-bottomed flask for vacuum distil-
lation. The distillation yielded 69.5 g. (a7.5%) of a colorless liquid
(b.p. 93-940 C @ 0.250 nm.).
Preparation of p--bromo-N,N-diallylaniline
D-Bromoaniline (200.0 g., 1.15 mol.), sodium bicarbonate (242.0 g.,
2.9 mol.), and 400 ml. of water were placed in a 2-liter, 3-necked,
round-bottomed flask equipped with stirrer, condenser, addition funnel,
and heating mari-e. The mixture was heated to low reflux. Allyl
broride (278.3 g., 2.3 mol.) was added dropwise over a two-hour period
with constant stirring. The lower, oily layer was dried with sodium
sulfate. The crude product was vacuum distilled yielding 215.3 g. (74.5%)
of a clear liquid (b.p. 113-1140 C @ 0.100 rm.).
Synthesis of p-methoxy-N,N-diallylaniline
p-Methoxyaniline (187.0 g., 1.0 mol.), sodium carbonate (210.0 g.,
2.0 mol.), and 500 ml. of water were placed in a 2-liter, 3-necked,
round-bottomed flask equipped with stirrer, addition funnel, and con-
denser. The mixture was heated to reflux on a Glas-col heating mantle.
Allyl bromide (266.0 g., 2.2 mol.) was added dropwise to the mixture.
The resulting mixture was allowed to stir at reflux overnight and was
then transferred to a 2-liter separatory funnel. The water layer was
discarded; the organic layer was dried with magnesium sulfate. The
liquid was filtered and then vacuum distilled yielding 183.4 g. (88.0%)
of a clear liquid (b.p. 87-880 C @ 0.350 mm.).
The ir spectrum showed only a small residual N-H absorbance at
3420 cm.-. Other absorbances were present at 3100 (m), 2840 to 3020
(s, broad, detailed), 1840 (w, broad), 1645 (m), 1622 (w), 1580 (w),
1515 (s, broad), 1463 (m), 1442 (m), 1422 (m), 1403 (w), 1385 (w),
1360 (w), 1332 (w), 1300 (w), 1278 (w), 1240 (s, broad), 1180 (s),
1130 (m), 1040 (s), 918 (s), 815 (s), 760 (w), 712 (m) cm.-1
The nmr spectrum showed resonance signals at 66.65 (m, 4); 65.85
(m, 2); 65.1 (m, 4); and 67.1 (m, 7).
Synthesis of p-chloro-N,N-diallylaniline
p-Chloroaniline (225.0 g., 2.0 mol.), sodium carbonate (420.0 g.),
and 500 ml. of water were placed in a 2-liter, 3-necked, round-bottomed
flask equipped with stirrer, addition funnel, and condenser. The
mixture was heated to reflux on a heating mantle. Allyl bromide (484.0
g., 4.0 mol.) was added dropwise with stirring. Heating and stirring
were continued overnight. The resulting liquid was poured through a
filter into a separatory funnel in which two layers formed. The lower
layer was discarded and the brown, oily layer was dried with sodium
sulfate. The crude p-chloro-N,N-diallylaniline was distilled under
vacuum yielding 300.4 g. (72.5%) of a colorless liquid (b.p. 85-860
C @ 0.350 mm.). Nmr and ir spectra agreed with the assigned structure.
Preparation of N,N-di-3-butenvlaniline
Aniline (14.0 g., 0.15 mol.), sodium carbonate (32.0 g.), and
50 ml. of water were placed in a 250 ml., 3-necked, round-bottomed
flask equipped with a mechanical stirrer, addition funnel, and cold-
water condenser. The mixture was heated to a low reflux. 4-Bromobutene
(40.5 g., 0.30 mol.) was added dropwise with constant stirring. The
reaction was allowed to proceed for 24 hours. The contents were
cooled and filtered. The amine layer was separated, dried over sodium
sulfate, and distilled under vacuum. A clear, colorless liquid (11.2
g., 37%) was obtained (b.p. 74-750 C @ 0.100 mm.).
The ir spectrum exhibited absorbances at 695 (s), 748 (s), 915
(s), 992 (s), 1040 (w), w) ), 1180 to 1220 (w, detailed), 1283 (m),
1360 (m), 1395 (w), 1420 (w), 1430 to 1460 (w, detailed), 1503 (s),
1600 (s), 1640 (m), and 2950 to 3050 (s, detailed, broad) cm. .
The nmr spectrum gave resonance signals at 52.25 (quartet, 4);
63.3 (m, 4); 65.1 (m, 4); 65.8 (m, 2); and 66.8 (m, 5).
Synthesis of N,N,N',N'-tetraallyl-p-phenylenediamine
p-Phenylenediamine (125.0 g., 1.15 mol.) and sodium carbonate
(530.0 g.) were added to a 3-liter, 3-necked, round-bottomed flask
fitted with a Claisen adapter, thermometer, addition funnel, and
reflux condenser. Water (1500 ml.) was added; the mixture was heated
to reflux to dissolve the solid. The solution was then cooled to 700
C and 3-bromopropene (559.7 g., 4.63 mol.) added dropwise with stirring.
The solution was refluxed overnight before extracting the oily layer
with ether. The ether layer was then dried over sodium sulfate and
filtered. The brown, oily compound was then distilled yielding 124.6 g.
(46.5%) of a pale yellow liquid (b.p. 138-1400 C @ 0.550 mm.).
Synthesis of phenylboronic acid
A mixture of ethyl ether and magnesium (36.0 g., 1.5 g. atoms)
was placed in a 1-liter, 3-necked, round-bottomed flask equipped with
a mechanical stirrer, addition funnel, and cold-water condenser. Bromo-
benzene (157.0 g., 1.0 mol.) in 200 ml. of ether was added dropwise
with stirring. After the reaction was started, the round-bottomed
flask was cooled in ice water to control the temperature. After all
the bromobenzene had been added, the contents of the flask were allowed
to warm to room temperature, Trimethylborate (103.9 g., 1.0 mol.) and
800 ml. of diethyl ether were placed in a 3-liter, 3-necked, round-
bottomed flask equipped with a mechanical stirrer, low-temperature
thermometer, and Claisen adapter with nitrogen inlet tube and addition
funnel. The Grignard reagent was filtered through glass wool into the
addition funnel. The solution in the 3-liter flask was cooled to -700
C in a dry ice-jsopropanol bath. The Grignard reagent was added drop-
wise and the temperature was controlled between -70 and -550 C. The
mixture, after cooling overnight, was hydrolyzed with 10% sulfuric
acid. The ether layer was separated and placed in a 3-liter, 3-necked,
round-bottomed flask fitted with a Claisen distilling head, mechanical
stirrer, addition funnel, and heating mantle. The ether solution was
concentrated while adding 1000 ml. of water dropwise. At this point
the temperature at the distilling head had reached 980 C. The solu-
tion was cooled. Fluffy, white crystals were observed. These were
filtered and washed with 3 10 ml. portions of hexane yielding 108.1 g.
(88.8%) of phenyl boronic acid (m.p. 214-2160 C).
Preparation of diethyl phenylboronate
Phenylboronic acid (50.0 g., 0.41 mol.), benzene (320.0 g.), 138
g. of absolute ethanol, and 2 drops of sulfuric acid were placed in a
1-necked, 1-liter, round-bottomed flask. An azeotropic distillation
apparatus consisting of a packed column, cold-water condenser, ther-
mometer, Claisen head, drying tube, and Dean-Stark trap was assembled.
The flask was heated on a Glas-col heating mantle. A ternary azeotrope
of water, ethanol, and benzene distilled over at 640 C. fi;. layers
formed in the trap. The lower (water) layer was removed as the reaction
progressed. After 2 days, the temperature at the distilling head had
reached 680 C, indicating the reaction was complete. The binary azec-
trope of benzene and ethanol distills at this temperature. The remain-
ing solvent was distilled and the resulting liquid placed in a 100 ml.
round-bottomed flask for vacuum distillation. The distillation yielded
65.5 g. (90.1%) of a clear liquid (b.p. 68-690 C @ 0.350 mm.).
Preparation of triethylamine-phenylborane
Lithium aluminum hydride (12.0 g., 0.32 m'ol.) was added to 400 ml.
of dried diethyl ether in a 3-necked, 1-liter, round-bottomed flask
equipped with condenser and drying tube, addition funnel, low-tempera-
ture thermometer, nitrogen inlet tube, addition funnel connector,
and a mechanical stirrer. The mixture was refluxed for thirty
minutes and then cooled to -720 C in a dry ice-isopropanol bath.
Triethylamine (37.4 g., 0.37 mol.) was added with stirring. Then
a solution of diethyl phenylboronate (65.5 g., 0.37 mol.) in ether
was added dropwise with stirring, keeping the temperature below
-650 C. The solution was stirred at -720 C for another hour and
then allowed to warm to room temperature. The mixture was filtered
through a sintered-glass funnel. The filtrate was concentrated and
then cooled in a dry ice-isopropanol bath resulting in the formation
of 51.9 g. (73.0%) of white, needle-like crystals (m.p. 65-660 C).
Preparation of 1,2-diphenyl-l,2-azaborolidine and 1,5-diphenyl-l-aza-
A solution of 1.25 liters of toluene, triethylamine-phenylborane
(18.0 g., 0.094 mol.), and N,N-diallylaniline (16.3 g., 0.094 mol.)
was distilled at atmospheric pressure in a 2-liter, 1-necked, round-
bottomed flask until the temperature reached 1200 C at the distilling
head. The remaining toluene was removed on a rotary evaporator,
resulting in a phosphorescent-green solution. This solution was dis-
tilled on a spinning band distillation column yielding two major pro-
ducts. The lower boiling component, 4.10 g., was a clear liquid (b.p.
84-860 C @ 0.30 mm.). The spectral data agreed with the assignment
of 1,2-diphenyl-1,2-azaborolidine. The ir spectrum gave absorba.ces
at 3430 (w), 3250 (w), 3060 (s), 2950 (s), 2880 (s), 1600 (s), 1490 (s),
1400 to 1440 (s, broad), 1300 (s), 1190 (m), 1150 (w), 1080 (w), 1050
(m), 1000 (m), 750 (s), 700 (s) cm.-1
The nmr spectrum gave resonances at 67.2 (m, 10); 63.8 (t, 2);
and a broad multiple at 61.8 (n, 4). The mass spectrum showed a
peak at 220 1.
Anal. Calcd. for C15H16BN: C, 81.47; H, 7.29; N, 6.34; and
B, 4.89. Found: C, 81.42; H, 7.31; N, 6.21; and B, 5.01.
The higher boiling fraction (b.p. 125-1300 @ 0.30 mm.) was
dissolved in acetone and cooled in a dry ice-isopropanol bath after
which white crystals formed. These crystals were filtered and dried
yielding 4.47 g. (m.p. 80-810 C). The spectral data and analysis
agreed with the assignment of 1,5-diphenyl-l-aza-5-borabicyclo(3.3.0)-
octane.9 The ir spectrum showed absorbances at 3040 (m), 2900 (s),
2820 (m), 1.925 (s), 1850 (w), 1580 (s), 1420 to 1480 (s, broad), 1260
(s), 1215 (m), 1175 (s), 1150 (m), 1125 (s), 1075 (m), 1040 (m), 1020
(s), 980 (s), 950 (s), 930 (m), 880 (s), 830 (w), 810 (s), /75 (s),
745 (w), 725 (s), 670 (s), and 610 (!w) cm. .
The nmr spectrum gave resonance signals at 66.85 (m, 10); 63.3
(m, 4); 62.1 (q, 4); and 61.1 (m, 4).
The mass spectrum gave a parent peak at 262 1.
Anal. Calcd. for CSH22 BN: C, 82.13; H, 8.42; N, 5.32; and
B, L.11. Found: C, 82.15; H, 8.51; N~ 5.41; and F, 4.21.
Synthes-is of p-bLroophenylboronic acid
A mixture of magnesium (27.0 g., 1.1 g. atoms) and 100 ml. of dry
diethyl ether was placed in a 1-liter, 4-necked, round-bottomed flask
equipped with a stirrer, addition funnel, and reflux condenser. p--Di-
bromobenzena (237.0 g., 1.0 moi.) in dLethyl ether was added dropwise
with stirring after the initial reaction was started. An ice bath was
used to control the reaction rate. The reaction mixture was stirred
for an additional 30 minutes at room temperature. Trimethylborate
(103.9 g., 1.0 mol.) and 800 ml. of diethyl ether were placed in a
3-liter, 3-necked, round-bottomed flask equipped with a mechanical
stirrer, low-temperature thermometer, and Claisen adapter for an
addition funnel and nitrogen inlet tube. The Grignard reagent was
filtered through glass wool into the addition funnel. The solution
in the 3-liter flask was cooled to -700 C in a dry ice-isopropanol
bath. The system was flushed with a slow stream of nitrogen. The
Grignard reagent was added dropwise with stirring and the temperature
was kept below -650 C. The mixture was allowed to warm to room tem-
perature with continuous stirring and was hydrolyzed with 10% sulfuric
acid. The ether layer was separated and placed in a 3-necked, 3-liter,
round-bottomed flask equipped with mechanical stirrer, Claisen head,
and addition funnel. Using a heating mantle and variac, the ether was
distilled away while adding 1000 ml. of water dropwise. When the tem-
perature at the distilling head reached 980 C, the solution was trans-
ferred to a l-.liter Erlenmeyer flask and allowed to cool. White, fluffy
crystals were obtained and washed with 3 10 ml. portions of hexane.
The total yield was 187.4 g. (93.5%) of p-bromophenylboronic .aci (a:.p.
The ir spectrum showed absorbances at 3300 (broad, s), 2400 (w),
600 (s), 1500 (m), 1435 (s), 1350 (broad, s), 1280 (w), 1185 (s), 1090
(s), 1070 (w), 1010 (broad, s), 920 (w). 760 (m), 690 (s), and 630 (s)
The nmr spectrum exhibited resonance signals at 67.5 (mi, 4) and
62,75 (s, 2).
Preparation of diethyl p-bromophenylboronate
p-Bromophenylboronic acid (100.8 g., 0.5 mol.), benzene (320.0 g.),
and absolute ethanol (138.0 g.) were placed in a 2-liter, 1-necked,
round-bottomed flask. An azeotropic distilling apparatus consisting
of a packed column, Claisen head, thermometer, condenser with drying
tube, and a Dean-Stark trap was assembled. The round-tottomed flask
was heated on a Glas-col mantle until a ternary azeotrope of benzene,
ethanol, and water distilled at 640 C. The water layer that resulted
was continuously removed through the Dean-Stark trap. After all of
the water had been collected, the temperature had reached 680 C the
temperature at which the binary azeotrope of benzene and ethanol dis-
till. The remaining benzene and ethanol were distilled away and the
remaining brownish liquid placed in a 100 ml. round-bottomed flask
and vacuum distilled. The distilltion yielded 47.2 g. (41.0%) of a
clear liquid (b.p, 76-770 C @ 0.250 mm.).
The ir spectrum showed absorbances at 2995 (s), 2940 (s), 2920 (s),
1920 (w), 1650 (w), 1585 (s), 1560 (w), l1187 (s), 1470 (w), 1415 (s),
1375 (s), 1270 to 1350 (broad, s), 1255 (s), 1175 (4), 1125 (s), 1375
(s), 1100 (s), 1070 (s), 1040 (s), 1005 (s), 900 (s), 815 (s), 720 (s),
650 (m), and 620 (m) cm. .
The nmr spectrum showed resonances at ~7.L5 (s, 4), 64.05 (quartet,
4), and 61.25 (t, 6).
Magnesium (27.0 g., 1.1 g. atoms) and 500 ml. of dry diethyl ether
were placed in a flamed, 1-liter, 3-necked, round-bottomed flask
equipped with mechanical stirrer, addition funnel, and cold-water
condenser. The mixture was refluxed for 30 minutes. p-Dibromobenzene
(235.9 g., 1.0 mol.) was dissolved in dry diethyl ether and placed in
a 500 ml. addition funnel. Then the mixture was activated with a small
crystal of iodine; a few ml. of the solution were added. An ice bath
was necessary to control the reflux rate while the remainder of the di-
bromobenzene was added dropwise to the stirring mixture. The resulting
solution was filtered to remove the unreacted magnesium.
p-Bromophenylboronic acid (80.0 g., 0.4 mol.), benzene (320.0 g.),
and ethanol (138.0 g.), along with a small amount of sulfuric acid,
were placed in a 2-liter, 1-necked, round-bottomed flask. An azeotropic
distillation apparatus was assembled, consisting of a packed column.
Claisen head, thermometer, condenser with drying tube, and Dean-Stark
trap. A ternary azeotrope of benzene, water, and ethanol distilled at
640 C, forming two layers in the trap. Ths lower (water) layer was
continuously removed as the reaction proceeded. The reaction appeared
complete as the temperature at the distilling head reached 670 C-the
temperature at which the binary azeotrope of benzene and ethanol dis-
tills. Most of the remaining benzene and ethan:,l were distilled away
and the remaining liquid placed in a 200 ml. round-bottomed flask.. The
brown liquid was then distilled under vacuum yielding 77.4 g. (60%) of
a clear liquid (b.p. 94-950 C @ 0.50 mm.).
The infrared spectrum showed absorbances at 2995 (s), 2940 (s),
2920 (s), 1920 (w), 1650 (w), 1585 (s), 1560 (w), 1487 (s), 1470 (w),
1415 (s), 1375 (s), 1270 to 1350 (broad, s), 1255 (s), 1175 (w), 1125 (s),
1100 (s), 1070 (s', 1040 (s), 1o005s 0 s), 00 (s), S.S (s), 720 (s), 550
(m), and 920 (m) cm.-
The nmr spectrum exhibited resonances at 67.45 (s, 4); 64.05
(quartet, 4); and 61.25 (t, 6).
Preparation of triethylamine-(p-bromo)phenylborane
Lithium aluminum hydride (5.5 g., 0.15 mol.) was added to 400 ml.
of diethyl ether in a 4-necked, 1-liter, round-bottomed flask equipped
with a mechanical stirrer, condenser with drying tube, addition funnel
with nitrogen inlet tube, and low temperature thermometer. The mix-
ture was refluxed for 30 minutes on a heating mantle and then cooled
to -720 C in a dry ice-isopropanol bath. Triethylamine (15.8 g., 0.192
mol.) was added with stirring. A solution of diethyl p-bromophenyl-
boronate (49.3 g., 0.192 mol.) was added dropwise with stirring, keep-
ing the temperature below -650 C. The reaction mixture was allowed to
warm to room temperature and then filtered to remove the unreacted
lithium aluminum hydride. The filtrate was cooled in a dry ice-iso-
propanol bath to -700 C. White, needle-like crystals were formed.
These were unstable and appeared to melt during filtration, even under
nitrogen. The remaining material was immediately dissolved in toluene.
Preparation of 1-(4-bromophenyl)-2-phenyl-1,2-azabcrol idine and
-( 4-bromophenyl) -5--phenyl-l-aza-5-borabicyclo. 3.3.0 )octane
Triethylamine-phenylborane (19,3 g., 0.1 mol), p-bromo-N,N--diallyl-
aniline (25.2 g., 0.1 mol.), and 1.25 liters of toluene were placed in
a 2-liter, round-bottomed flask. The toluene was slowly distilled from
the flask through a fractionating column and condenser. On initial
heating the contents of the flask turned a bright yellow-green color.
When the temperature at the distilling head reached 1200 C, the residual
green liquid was transferred to a 50 ml. round-bottomed flask and
distilled on a spinning band column. Four fractions were obtained.
The two lowest boiling fractions (7.8 g.) were clear, pale yellow,
viscous liquids (b.p. 120-1230 C @ 0.15 mm.) Nmr and ir data were
consistent with the assignment of l-(4-bromophenyl)-2-phenyl-l,2-
azaborolidine. The ir spectrum gave absorbances at 3080 (s), 3060 (s),
2950 (s), 2870 (s), 1960 (w), 1890 (w), 1820 (w), 1765 (w), 1700 (w),
1630 (w), 1590 (s), 1570 (m), 1480 (broad, s), 1425 (broad, s), 1385
(s), 1300 (broad, s), 1240 (s), 1185 (m), 1150 (m), 1140 (w), 1100 (m),
1067 (s), 1050 (s), 1030 (m), 1000 (s), 960 (w), 910 (w), 880 (m), 820
(s),770 (w), 740 (s), 695 (s), and 640 (m) cm. The nmr gave
resonance signals at 67.2 (m, 9); 63.75 (t, 2); and 61.7 (broad m, 4).
Anal. Calcd. for C, H BNBr: C, 60.03; H, 5.04; N, 4.67; B,
_- 5 15
3.61; and Br, 26.65. Found: C, 60.14; H, 5.12; N, 4.59; B, 3,64;
and Br, 26.52.
The two remaining fractions (b.p. 130-1320 C @ 0.15 mm.) were
dissolved in small amounts of acetone and cooled in a dry ice-iso-
propanol bath, resulting in the formation of white crystals. These
were filtered and dried yielding 6.7 g. (m:.p. 116-1180 C).
The ir, r.mr, and mass spectra were consistent with the.assignment
of structure as (4-bromophenyl)--5-phenyl-!-aza- 5-bopabicyclo(3.3.0)--
octane. The ir spectrum showed absorbances at 3050 (m), 3010 (m),
3000 (m), 2920 (s), 2825 (s), 1950 (w), 1880 (w), 1820 (w), 1700 (w),
1475 (broad, s), 1425 (m), 1400 (m), 1350 (w), 1310 (w), 1265 (s),
1227 (s), 1190 (w), 11.70 (w), 1.155 (m), 1130 (broad, -), 1080 (m), 1050
(m), 1035 (m), 1020 (m), 1000 (s), 980 (m), 950 (m), 930 (n), 890 (m),
810 (s), 740 (s), 700 (broad, s), 675 (s) cm." The nmr showed
resonance signals at 66.9 (m, 9); 63.3 (m, 4); 62.05 (quintet, 4);
and 61.1 (m, broad, 4). The mass spectrum gave a parent peak of
Anal. Calcd. for C18H21BNBr: C, 63.19; H, 6.19; N, 4.09;
B, 3.16; and Br, 23.24. Found: C, 63.11; H, 6.29; N, 3.96; B, 2.90;
and Br, 23.36.
Preparation of l-(4-chlorophenyl)-2-phenyl-1,2-aza-borolidine and
Triethylamine-phcnylborane (32.7 g., 0.17 mol.), p-chloro-N,N-
diallylaniline (35.3 g., 0.17 mol.), and 1.5 1. of toluene were
placed in a 2-liter, round-bottomed flask. The contents were heated
on a heating mantle. Toluene (200 ml.) was slowly distilled through
a fractionating column and condenser. The mixture was slowly re-
fluxed; then the remainder of the toluene was distilled until the
temperature at the distilling head reached 1200 C. The remaining
toluene was removed on the rotary evaporator under reduced pressure.
The gir n, phosphorescent liquid was transferred to a 50 ml. flask
and distilled on a spinning band column. Three fractions were
collected. The two lowest boiling fractions (15.6 g.) were viscous,
colorless liquids (b.p. 130-1320 C @ 0.20 mm.) which turned brown
in air. A portion of the viscous liquid was sealed under vacuum for
analysis. The ir, nmr, and mass spectra were consistent with the
structure assignment of l-(4-chlorophenyl)-2-phenyl-l,2-azaborolidine.
The infrared spectrum exhibited absorbances at 3450 (w), 30'C (m),
3060 (m), 2950 (s), 2880 (s), 1960 (w), 1390 (w), 1820 (w), 1765 (w),
1640 (w), 1595 (s), .575 (w), 1490 (broad, s), 1430 (broad, s) 1393
(s), 1315 (s), 1290 (s), 1265 (m), 1245 (m), 1185 (m), 1153 (m), 1140
(w), 1093 (s), 915 (m), 925 (s), 740 (s), 720 (m), and 595 (s) cm.1
The nmr spectrum gave resonance signals at 67.1 (m, 9); 63.75 (t, 2);
62.25 (m, L). The mass spectrum gave a molecular weight of 255 1l.
Anal. Calcd. for C H5 BNC1: C, 70.48; H, 5.92; B, 4.23; and
01,13.88. Found: C, 70.57; H, 6.06; N, 5.27; B, 4.40; and Cl, 13.88.
The highest boiling fraction (b.p. 144-146 C @ 0.20 mm.) was
dissolved in a small amount of acetone and cooled in a dry ice-iso-
propanol bath, resulting in the formation of 4.6 g. of flaky, white
crystals (m.p. 95-960 C.). The ir, nmr, mass spectrum and analysis
were consistent with the structure of 1-(4-chlorophenyi)-5-phenyl-l-
aza-5-borabicyclo(3.3.0)octane. The "r spectrum showed absorbances
at 3080 (m). 3060 (m), 3040 (m), 3020 (m), 2920 (broad, s), 2837 (s),
2600 (w), 1958 (w), 1893 (w), 1823 (w), 1765 (w), 1640 (w), 595 (),
1485 (broad, s), 1450 (m), 1430 (s), 1390 (m), 1345 (n), 1330 (s), 1275
(broad, s), 1255 (w), 1230 (s), 1190 (s), 1175 (s), 1160 (s), 1132 (s),
1120 (m), 1096 (s), 1053 (s), 1038 (m), 1022 (m), 1008 (s). 980 (s),
957 (m), 933 (m), 895 (s), 860 (w), 820 (s), c10 (m), 745 (s), 700 (s),
and 648 (s) cm. The nmr spectrum exhibited resonances at 67.0 (broad
m, 9); 63.3 (m, 4); 62.05 (quintet, 4); 61.1 (broad m, 4). Th mass
spectrum gave a parent peak at 297 -3
Anal. Calcd. for C18H2,BNCI: C, 72.63; H, 7.11; N, 4.70; B, 3.64;
and Cl, 11.91. Found: C. 72.66: H, 7.06; N, 4.59; B, 3.73; and Cl,
Synthesis of p-chloro-phenylboronic acid
Magnesium turnings (36.4 g., 1.5 g. atoms) were placed in a 2-liter,
3-necked, round-bottomed flask equipped with mechanical stirrer, con-
denser with drying tube, and a 500 ml. addition funnel. Dried diethyl
ether (50 ml.) was added and the mixture refluxed for 15 minutes. p-Bromo-
chlorobenzene (287.1 g., 1.5 mol.), dissolved in I liter of dried diethyl
ether, was added dropwise after the initial reaction had started. After
complete addition, the mixture was refluxed for an additional 30 minutes.
The resulting Grignard reagent was filtered through glass wool into a 500
ml. addition funnel. Trimethylborate (155.8 g., 1.5 mol.) and 1 liter
of dry diethyl ether were placed in a 3-liter, 3-necked. round-bottomed
flask equipped with mechanical stirrer, low-temperature thermometer, and
Claisen adapter for an addition funnel and nitrogen inlet tube. The
solution was cooled to -700 C in a dry ice-isoprcpanol bath and the
Crignard reagent added dropwise with stirring. Nitrogen flow was main-
tained throughout the system during the course of the addition. The
temperature was kept below -650 C. The solution was allowed to warm to
room temperature overnight with continuous stirring. The resulting
mixture was hydrolyzed with a 15% sulfuric acid solution. The temperature
was controlled by cooling the flask in an ice bath. The two resulting
layers were separated in a 2-liter sparatory funnel and the aqueous
layer discarded. The ether layer was then placed in a 3-liter, 3-necked,
round-bottomed flask equipped with mechanical stir.rer, addition funnel,
Claisen distilling head, condenser, and heating mantle. The ether layer
was slowly distilled while adding water dropwise. when the temTiperature
at the distilling head reached 980 C, the hot solution was transferred
to a 2-liter Erlenmeyer flask and allowed to cool. :.pFroximrately 1500
ml. of water had been added as the ether was replaced. After cooling,
filtering, and washing with 3 20 ml. portions of hexane, white, fluffy
crystals, 195.2 g. (83.5%), were obtained (m.p. 280-281 C).
The ir spectrum showed absorbances at 3200 (broad, s), 1920 (w),
1880 (w), 1800 (w), 1730 (w), 1650 (w), 1590 (s), 1560 (m), 1350 (broad,
s), 1250 (w), 1170 (m), 1080 (s), 1010 (s), 820 (s), 720 (s), 670 (s),
and 640 (m) cm.-
The nmr spectrum exhibited resonance signals at 57.6; and 63.1
Synthesis of p-chlcrophenylboronate
p-Chl:c.'chen lboronic acid (100.0 g., 0.64 mol.) was placed in a
2-liter, round-bottomed flask. To this, ethanol (230.0 g., 5.0 mol.)
and benzene (546.0 g.) were added. An azeotropic distillation
apparatus was constructed consisting of a fractionating column,
Claisen head, thermometer, Dean-Stark trap, and condenser with dry-
ing tube. A small amount of sulfuric acid was added. The mixture
was heated on a Glas-col mantle. A ternary azeotrope of benzene,
ethanol, and water distilled ar 64 C. Two layers formed in the
trap and the lower lave" was con tinuously removed. Approximately
23 to 24 ml. of water were collected in the trap. After the temp-
erature had stabilized at 58 C, the temperature at which the binary
azeotrope of benzene and ethanol boils, the remaining solvent was
removed on the rotary evaporator. The residual crude eCter was
placed in a 200 -m., rout:d-bottomed flask and distilled under vacuum.
A clear liquid, 60.2 g. ('44.3%), was obtained (b.p. 92-930 C @ 0.75 mmn.).
The ir spectrum showed absorbances at 3100 (w), 3050 (w), 2990
(s), 2945 (s), 2920 (s), 1920 (w), 1660 (w), 1593 (s), 1563 (m),
1487 (s), 1430 (s), 1415 (s), 1375 (s), 1325 (broad, s), 1280 (s),
1258 (s), 1175 (w), 1127 (s), 1100 (s), 1087 (s), 1040 (s), 1018 (s),
904 (s), 820 (s), 725 (s), 650 (s), and 625 (s) cm.-1
The nmr spectrum showed resonances at 67.4 (quartet, 4); 64.0
(quartet, 4); and 61.24 (t, 6).
Synthesis of triethylamine-(p-chlorophenyl)borane
Lithium aluminum hydride (5.32 g., 0.14 mol.) was added to 600
ml. of dry diethyl ether in a l-liter, 4-necked, round-bottomed flask
equipped with mechanical stirrer, condenser with drying tube, addition
funnel with nitrogen inlet tube, and low-temperature thermometer. The
mixture was refluxed for 30 minutes on a heating mantle and then
cooled to -70 C in a dry ice-isopropanol bath. Triethylamine (28.6
g., 0.28 mol.) was added with stirring. The system was flushed with
nitrogen; diethyl p-chlorophenylboronate (58.9 g., 0.28 mol.) was added
dropwise. The temperature was kept below -55 C. The mixture was
then filtered through a sintered-glass funnel. A watch-glass was
placed over the funnel mouth to prevent evaporation of the ether. The
filtrate was concentrated and then cooled in a dry ice-isopropanol-bath,
resulting in the precipitation of a white cry3talline substance. The
material, 47.7 g. (75%), was filtered and dried yielding whites, needle-
like crystals (m.p. 63-64 C).
The ir spectrum exhibited absorbances at 3080 (w), 3000 (s), 2950
(s), 2980 (m), 2360 (broad, s), 1910 (w), 1798 (w), 1660 (w), 1578 (s),
1475 (s), 1450 (m), 1420 (w), 1380 (s), 1345 (m), 1300 (s), 1288 (w),
1193 (s), 1170 (s), 1150 (s), 1085 (s), 1065 (m), 1040 (m), 1013 (s),
900 (m), 860 (m), 820 (s), 775 (m), 750 (s), 720 (w), and 630 (s) cm. .
The nmr spectrum showed resonance signals at 67.3 (m, 4); 62.7
(quartet, 6); and 61.2 (t, 9).
Preparation of l-phenyl-2-(4-chlorophenyl)-1,2-azaborolidine and
Triethylamine-(p-chlorophenyl)borane (22.7 g., 0.10 mol.), N,N-
diallylaniline (17.3 g., 1.0 mol.), and 1.25 1. of dry toluene were
placed in a 2-liter, round-bottomed flask. The contents were heated
on a Glas-ccl mantle and distilled through a packed column, Claisen
head and condenser. Initial heating caused a light green coloration
of the solution. After the temperature reached 1200 C at the distill-
ing head, the remainder of the solvent was removed on a rotary evap-
orator. During the distillat.ion, several infrared samples of the
solution, distillate, and gas above the solution were obtained. Propene
gas was trapped. Confirmation of this was obtained by comparison of
the gas sample spectrum with that in the Sadtler Spectra Index. The
crude, green-brown liquid was transferred to a 50 ml. round-bottomed
flask and placed on a spinning band distillation column. Four fractions
were obtained. The lowest boiling fractions were pale yellow liquids
(b.p. 120-1210 C @ 0.10 man.). This material, 15-16 g., was unstable
in air; several samples were sealed in ampoules. The nmr, ir, mass
spectrum, and analysis were consistent with the structure assignment
of l-phenyl-2-(4-chlorophenyl)-l,2-azaborolidine. The ir spectrum ex-
hibited absorbances at 3090 (m), 3060 (m), 2960 (s), 2890 (s), 1945 (w),
1920 (w), 1875 (w), 1800 (w), 1740 (w), 1595 (s), 1.560 (m), 1500 (s),