Citation
Boroxines

Material Information

Title:
Boroxines
Series Title:
Boroxines
Creator:
Povlock, Thomas Paul
Place of Publication:
Gainesville FL
Publisher:
University of Florida
Publication Date:
Language:
English

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Thomas Paul Povlock. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000423880 ( alephbibnum )
11022345 ( oclc )

Downloads

This item has the following downloads:


Full Text













BOROXINES: THEIR REACTIONS

AND RING STABILITY












By
THOMAS PAUL POVLOCK


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY











UNIVERSITY OF FLORIDA
August, 1960















ACKNOWLEDGMENTS


The author wishes to express his sincere appreciation to

Dr. W. T. Lippincott for the encouragement, aid, and guidance which he so generously gave during the course of the investigation and the preparation of this dissertation.

The author is indebted to the members of his graduate committee for their helpful suggestions in writing this dissertation.

The author also wishes to express his gratitude to Mr. and Mrs. John Cribbs for the technical preparation of the dissertation.
















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS ii

LIST OF TABLES vi

INTRODUCTION I

A. Purpose 1

B. Historical 2

I. Borinic Acids 2

II. Boronic Acids 5

III. T rimethoxybo roxine 6

EXPERIMENTAL 12

A. Reaction With Grignard Reagents 12

I. General Procedure 12

II. Preparation of Intermediates 13

III. Preparation of Grignard Reagents 15

IV. Reaction of Grignard Reagents With
T rimethoxyboroxine 16

V. Reaction of Phenylmagnesium Bromide With
Triphenylboroxine 18

VI. Reaction of Phenylmagnesium Bromide With a
Mixture of Trimethoxyboroxine and Trimethyl
Borate 18










TABLE OF CONTENTS (continued)


Page

VII. Isolation and Purification of Reaction
Products 19

VIII. Characterization of Products 21

B. Reaction With Other Reagents 25

I. Reaction With the Sodium Salt of MAalonic Ester 25 II. Reaction With Carbonyl Compounds 26

RESULTS 27

A. The Nature of the Reaction 27

B. Summary of Yields of Borinic Acids Prepared Under
Comparable Conditions 27

C. Effect of Temperature 29

D. Effect of Varying the Ratio of Boroxine to Grignard
Reagent 31

E. Effect of Solvent 33

F. Effect of Some Electronic Factors 35

G. Effect of Some Steric Factors 35

H. Results of Reaction of Phenylmagnesium Bromide
With Other Boroxines 39

I. Tri-n-butoxyboroxine 39

II. Triphenylboroxine 40

I. A Comparison of the Reactivity of Trimethylboroxine
With Trimethyl Borate 40










TABLE OF CONTENTS (continued)


Page

J. Cyclic Borinic Acids 43

I. Introduction 43

II. Results of Studies With Di-Grignard Reagents 43

K. Results of Studies With the Sodium Salt of Malonic
Ester 44

L. Results of Studies With Carbonyl Compounds 45

DISCUSSION 46

A. The Role of the Grignard Reagent: The Case for
Stepwise Addition 46

B. The Role of the Borox-ine 49

C. Formation of the Carbon-Boron Bond: The Case
for Displacement on Magnesium 52

CONCLUSION 56

BIBLIOGRAPHY 58

BIOGRAPHICAL SKE TCH 61















LIST OF TABLES


Table Page

I. Physical Constants of Aromatic Borinic Acids and
Esters 8

II. Physical Constants of Aliphatic Borinic Acids and
Esters 10

III. Physical Constants of Unsymmetrical Borinic Esters 11 IV. Boiling Points and Molecular Weights of Borinic Acid Anhydrides 22

V. Analysis of Borinic Acid Anhydrides 24

VI, Yields of Borinic Acids in Reaction of Grignard Reagents With Trimethoxybo roxine 28

VII. Effect of Temperature in the Reaction of Trimethoxyboroxine *With Phenylmagnesium Bromide 30

VIII. Effect of Temperature in the Reaction of Trimethoxyboroxine With n-Butylmagnesium Bromide 30

IX. Effect of Changing the Mole Ratio of Trimethoxyboroxine to Phenylmagnesium Bromide at Z50C 32

X. Effect of Changing the Mole Ratio of Trimethoxyboroxine to n-Butylmagnesium Bromide at 25'C 32

XI. Effect of Solvent in the Reaction of Trimethoxyboroxine With Phenylmagnesium Bromide 34

XII. Effect of Some Electronic Factors 36

XIII. Effect of Steric Factors in the Reaction of Aryl
Grignard Reagents With Trirnethoxyboroxdne 37










LIST OF TABLES (continued)


Table Page

XIV. Effect of Steric Factors in the Reaction of Alkyl
Grignard Reagents With Trimethoxyboroxine 38

XV. A Comparison of the Reactivity of Trimethoxyboroxine
With Trimethyl Borate 41














INTRODUCTION

A. Purpose


This study was undertaken to examine the reaction of trimethoxyboroxine with organic bases and carbanion reagents, and to obtain more information about the nature of the reaction in which a carbon-boron bond is formed. The following reactions were studied:

1. Reaction with Grignard reagents

2. Reaction with sodio salts

3. Reaction with carbonyl compounds Only the first of these was studied in detail.

This study was divided into two sections. The first was reported by the author in his Master of Science thesis. In the work reported therein, it was found that aryl Grignard reagents react with t. imethoxyborodne to yield aryl borinic acids. Eight aryl borinic acids, isolated as the aminoethyl esters, were prepared in yields of 21-62 percent. The effects of temperature, ratio of Grignard reagent to trimethoxyboroxine, steric, and electrical factors were studied.

The second section of the study of the reaction of Crignard reagents with trimethoxyboroxine constitutes the major portion of this








a

dissertation. It consists of a study of the reaction of alkyl Grignard reagents with trimethoxyboroxine, From this study nine alkyl borinic acids were prepared and characterized. In an attempt to elucidate a possible mechanism for the reaction, a study of the effects of temperature, ratio of trimethoxyboroxine to Grignard reagent, solvent, and steric factors was conducted along with the study of methods of isolation and purification of the borinic acids. In addition, reactions with di-Grignard reagents, preparation of a possible intermediate and its reaction with a Grignard reagent, and a series of competition experiments involving mixture of trimethoxyboroxine and trimethyl borate were studied.

An examination of the reaction of the sodium salt of malonic ester and simple aldehydes and ketones with trimethoxyboroxine was undertaken.


B. Historical

I. Borinic Acids

Aryl borinic acids, having the general formula R BOH, have been known since 1894 when Michaelis (1) described the isolation of diphenylborinic acid from the hydrolysis of the corresponding chloride. Since the methods of isolation and characterization were doubtful, it was not known whether or not the earlier workers had actually prepared the reported compounds. In the case of the diphenylborinic acid, it was not until 1955 that Letsinger (2) isolated and characterized this acid.








3

The yield of borinic acid was unspecified in some of the earlier work and when reported ranged from Z0-40 percent.

Tables I, II, and III list the aryl, allyl, and unsymmetrical

borinic acids and esters for which the melting or boiling points have been reported. The boiling points of the acids which have been recorded are actually those of the anhydride. Letsinger (2) showed that when diphenylborinic acid wac distilled, the anhydride was recovered.

In some cases there are discrepancies in the reported physical

constants of the borinic acids as shown in Table I. Di(p-bromophenyl)borinic acid is listed with two different melting points. Di(p-biphenyl)borinic acid is listed as melting above 300�C, but the isolated compound was never proved to be the borinic acid. It was found as a by-product in the preparation of the boronic acid and was assumed to be the borinic acid.

Borinic acids have been prepared by use of one of the following reactions, followed by hydrolysis.

1, Moist air oxidation of trialkylborane (R3B) (3).

2. Reaction of trialkylborane with iodine to prepare the

dialkyliodoborane (4).

3. Reaction of trialkyl borates [(RO) 3B] or cyclic borates

. ,O-CH 2
RO-B 2 with an excess of alkyl or aryl Grignard reagents



at -60C (2, 5, 6, 7, 8).









4. Reaction of trialhoxyboroxine [(ROBO)3] with an excess of

aryl Grignard reagent at 25*C (9, 10).

5. Reaction of boronic esters [RB(OR') I with alkyl or aryl

Grignard reagent at -60*C (7, 8,11, 12, 13, 14).

Of these preparations method 3 is the most widely used procedure, although this author (9) showed that method 4 using trimethoxyboroxine gave higher yields of aryl borinic acids. Method 5 is used for the preparation of unsymmetrical borinic acids.

Two of the principal difficulties in the preparation of borinic acids are their separation from the boronic acid and the preparation of a suitable derivative for characterization. Letsinger (5) developed a method of separation of the borinic ester from the boronic ester. This method consists of treating the mixtureoof esters in ether solution with ammonia. The borinic ester complexes with ammonia and precipitates from solution. The complex is then treated with a solution of ethanolamine in water or toluene to prepare the aminoethyl borinate. This ester is an excellent derivative because it can be prepared easily, recrystallizes readily from an ethanol-water mixture, melts sharply, analyzes well, and over a period of months shows no sign of decomposing. The aminoethyl esters of aryl (5, 9) and mixed aryl alkyl borinic acids (12) were reported, but there are no reports of this ester of alkyl borinic acids.








5

Letsinger (15) prepared a cyclic borinic acid, 5-, 8-aminoethoxy10, 11 -dihydrodibenzo- (b, f)-borepin





B
I
0- CH2- CH2- NH2

Lappert (16, 17, 18) prepared a series of esters of diphenyl- and di(n-butyl)borinic acids and studied the reactions of the esters with reagents such as aqueous base, hydrogen halides, boron chlorides, and phosphorus pentabromide,


II. Boronic Acids

Boronic acids, having the general formula RB(OH)2, have received far more attention in the literature than the borinic acids, In a review article by Lappert (19), seventy alkyl and aryl beronic acids were reported along with their melting points.

The most important methods for the preparation of alkyl or aryl boronic acids depends on one of the following procedures.

1. The addition of a metal alkyl or aryl to a trialkyl borate

and the hydrolysis of the boronic ester (20).

2. The addition of a metal alkyl or aryl to a boron trihalide

to obtain the alkyl or aryl boron dihalide, which on hydrolysis

gives the boronic acid (3).







6

3. Anhydrous air oxidation of a trialkylborane to a boronic

ester and the hydrolysis of the latter (3).


III. Trimethoxyboroxine

Trimethoxyboroxine, I, is believed to be a six membered ring containing alternate boron and oxygen atoms with one methoxy group attached to each boron atom.





0 0
I I
H 3-0 -B B-OGl3
\o/



I

Ogle (21) has verified the structure by molecular weight determinations, elemental analysis, Raman spectra, and the solubility in anhydrous ether while one of its components, boric acid anhydride, is insoluble.

Ogle, Lippincott, and Quill (22, 23) studied the reaction of trimethoxyboroxine with alcohols and aromatic amines. The reaction with alcohols is illustrated in the following equation:

6 ROH + (CH3OBO)3 ), 3 (RO)2BOCH 3 + 3 H 20







7

The following equation illustrates the reaction of trirnethoxyboro:rine with aniline.


2 (CH 3OBO)3 +


QNI


OCH
B
0 0


S3 - \ I
HG3O-13 B
3J


-N

k


OCH3
j3
/B

0 0
1 I
13 B-OCH3 \0/


+ 2 CH 3OH


r










TABLE I

PHYSICAL CONSTANTS OF AROMATIC BORINIC ACIDS AND ESTERS


R BOR' Melting Point Boiling Point Reference

R R' C


Phenyl

















o- Tolyl m- Tolyl p- Tolyl p-Anisyl p- Biphenyl p- Chiorophenyl


Hydrogen ,6 -Aminoethyl Ethyl

n-Propyl n- Butyl 3-Methylpropyl 1 -Methylpropyl Phenyl

,e -Aminoethyl 3- Methylpropyl ,6 -Aminoethyl 8 -Aminoethyl

- Aminoethyl Hydrogen

/3-Aminoethyl Hydrogen

0 -Aminoethyl


210/ 1


mm.


189-190


138/10 152/10 158/10

153/10 152/10 140/ 0.


am. mm.



mm. mm.

05mm.


181-18115


127/ 5.5 mm.


180-181 186-187

173-174 107

217-218

300

223-224


1
2;5;9


15 15 15 15 15

15

9

10

9

9;10

9

24

9

25

9







9

TABLE I (continued)


R BOR Melting Point Boiling Point Reference
R R' 0C 0C


Hydrogen 75 26

&.- Naphthyl .-Aminoethyl 202-203 9

Hydrogen 105-106 5

,6-Naphthyl Hydrogena 172 24

p-Bromophenyl Hydrogena 113 24

Hydrogen 82-84 1

-Aminoethyl 236-237 2

2-Methyl- a
5- Chlorophenyl Hydrogen 81 26


alsolated as anhydride.












TABLE II

PHYSICAL CONSTANTS OF ALIPHATIC BORINIC ACIDS AND ESTERS R L!OR, Melting Point Boiling Point Reference

R R' 0C 0C


n- Butyl n- Butyl 114/20 mm, 17;27
Hydrogena 137/ 9 mm. 17


3-Methylpropyl 94/ 8 mrn 17

1-Methylpropyl 96/ 9. mm. 17

Phenyl 116/ 0.1 mm. 17

Ethylene glycol 147/ 0.48mm. 6

4-Methylbutyl Ethylene glycol 147/ 0,48rnm. 6

Allyl n-Butyl 61/ 6 mm. 27
Ethyl Hydrogena -51 to -48 35/75 mm, 28


2- Chlorovinyl Hydrogena 66/ 3 mm. 29






TABLE III

PHYSICAL CONSTANTS OF UNSYMMETRICAL BORINIC ESTERS


RRtBOR" Melting Point Boiling Point Reference
R R' R" ac 0c


Phenyl o(- -Naphthyl 6 -Aminoethyl 228 5

Phenyl Benzyl . -Aninoe thyl 141-142 12

Phenyl 1 -Methylheptyl de -Aminoethyl 93- 95 12

Phenyl n-Butyl .8 -Arminoethyl 108-108.5 12

3-kiethylpropyl 96-97 / 3mm, 13

n-Butyl n-Propyl n-Butyl 84-86 /10mm, 11

Phenyl n-Propyl 3-Methylpropyl 83-87 / 3mm, 13

Phenyl o-Tolyl n- Propyl 156-159/ 9mn. 30

o Tolyl p- Tolyl n-Propyl 173-175/10mm. 14

o-Tolyl rn-Tolyl n-Propyl 173-175/11mm,. 14

o-Tolyl p-Chlorophenyl n- Propyl 176-177/10mm. 14

Phenyl Ethyl n-Butyl 110-111/ 9mm. 14

Phenyl Methyl n-Propyl 87-89 / 9mm. 14














EXPERIMENTAL

A. Reaction With Grignard Reagents I. General Procedure


The general procedure used in carrying out the reaction of

Grignard reagents with trimethoxyboroxine was previously reported by the author (9), and is summarized as follows: Grignard reagents from both aliphatic and aromatic bromides were prepared in anhydrous ether solution. Trimethoxyboroxine, prepared from trimethyl borate and boric acid anhydride, was dissolved in anhydrous ether and added to each Grignard reagent. The reaction mixture was hydrolyzed and the ether layer removed and distilled yielding a liquid containing a mixture of boronic and borinic acids. The borinic acids derived from aliphatic Grignard reagents were distilled and converted to the corresponding anhydride. The borinic acids from aromatic Grignard reagents were converted to the aminoethyl esters.

The experimental details for each of the processes described above are given in Section IIA, IIIA, IVA, and VII.








13

II. Preparation of Intermediates A. Preparation of Trimethoxyboroxine

The boroxine was prepared following a procedure of Goubeau and Keller (31)i In a 500 ml. one-necked flat-bottomed flask were placed 69. 6 grams (1 mole) of boric acid anhydride and 103. 8 grams (1 mole) of trimethyl borate along with a glass covered stirring bar. The flask was connected to a reflux condenser and placed in a heating mantle on top of a magnetic stirrer. The reaction mixture was stirred and refluxed from two to four hours or until the boric acid anhydride dissolved. The flask was cooled to room temperature and then immersed in a Dry Ice-isopropyl alcohol bath until the contents of the flask froze. The flask was removed from the bath, permitted to warm to room temperature, and the trimethoxyboroxine stored in a glass stoppered Erlenmeyer flask.

Special precautions were taken in the preparation of the boroxine. Moisture was excluded at all times. Freshly distilled trimethyl borate gave the best results. Boric acid anhydride obtained from Baker Chemical Company contained both the powder and glass structure. Only the powder was suitable for the boroxine and was separated from the glassy structure by screening through a 40 mesh screen. B. Preparation of Triphenylboroxine (20)

In a 2-liter three-necked flask equipped with a stirrer, condenser, and addition funnel were placed 48.6 grams (2 moles) of magnesium








14

turnings. The reaction was initiated with 50 ml. of anhydrous ether and 10 ml. of bromobenzene. After the reaction had started, the remainder of 316 grams (2 moles) of bromobenzene, which had previously been diluted with 1450 ml, of anhydrous ether, was added to the flask over a period of five hours. The reaction mixture was refluxed for three hours after the addition,

In a 3-liter three-necked flask equipped with a stirrer, condenser, and addition funnel were placed 500 ml, of anhydrous ether and 208 grams (2 -moles) of freshly distilled trimethyl borate. The reaction flask was immersed in a Dry Ice-isopropyl alcohol bath and the. temperature maintained between -60 and -70GC. The phenylmagnesium bromide was added slowly over a period of eight hours. The reaction mixture was allowed to slowly warm to room temperature, and hydrolyzed using a solution of 75 ml. of concentrated sulfuric acid in 600 ml. of water. The ether layer was removed and the water layer washed with ether. The ether extracts were combined and distilled on a steam bath to remove the solvent. The resulting liquid was extracted with a dilute sodium hydrox-ide solution. The aqueous layer was acidified and the solid removed by filtration. The phenylboronic acid was converted to the anhydride (triphenylboroxine) by dissolving in benzene and distilling off the water. The triphenylboroxdne was recrystallized from benzene and dried in a vacuum desiccator. The yield of phenylboronic anhydride was 105 grams, 43 percent based on trimethyl borate. Melting point 214-215"C (lit. m. pt. Z160C (25)).








15

III. Preparation of Grignard Reagent A., General Procedure

Grignard reagents were prepared following the procedure from Organic Syntheses (32). In a I-liter three-necked flask equipped with a stirrer, addition funnel, and a parallel side arm holding a reflux condenser and a thermometer were placed 9.7 grams (0. 4 mole) of magnesium turnings, The system was swept with dry nitrogen and a nitrogen atmosphere -was maintained over the system throughout the reaction (33). Ten milliliters of anhydrous ether and

5 ml. of the bromide were added to the flask. After the reaction was initiated, the stirring was started, and the remainder of 0. 4 mole of the bromide, which had previously been diluted with 170 ml. of anhydrous ether, was added to the flask over a period of one hour. After the addition, the reaction mixture was refluxed for two hours. The contents of the flask were cooled to room temperature, a 5 ml. aliquot was removed for analysis, and the ether solution of trimethoxyboroxine was added to the remainder of the Grignard reagent as described in Section IV.

The 5 ml. aliquot was analyzed by titration (34). Twenty milliliters of distilled water was added along with an excess of standard hydrochloric aciul to the aliquot, and the excess of acid was neutralized with standard sodium hydroxide to a methyl orange end point.








16

IV. Reaction of Grignard Reagents With Trimethoxyboroxine A. General Procedure

The addition of trimethoxyboroxine to aryl and alkyl Grignard reagents was carried out in exactly the same manner for all the reactions.

To a 1-liter three-necked flask containing the Grignard reagent equipped with a stirrer, addition funnel, and a parallel side arm holding a reflux condenser and a thermometer was added 100 ml. of anhydrous ether. The nitrogen atmosphere was maintained over the system. From the Grignard analysis, the amount of trimethoxyboroxine needed was calculated. The boroxine was dissolved in 150 ml. of anhydrous ether and was added dropwise over a period of two hours. After the addition of the boroxine, the reaction mixture was stirred for five hours at constant temperature. Forty milliliters of concentrated hydrochloric acid in 300 ml. of water was used to hydrolyze the reaction mixture. The ether layer was separated and washed twice with 100 ml. portions of water. The borinic acid was isolated from the ether solutior. by the procedure given in Section V.


B. Effect of Temperature

The effect of temperature on the yield of borinic acid was studied. The procedure was the same as given in the preceding paragraph with temperatures of -60�, 6, 16o, 250, and 35* and mole ratio of Grignard reagent to boroxine of 9:1. The temperature was controlled










at -.60 by immersing the reaction flask in a Dry Ice-isopropyl alcohol bath with the temperature range between -65 and -55�C. At 60 the temperature was controlled between two degrees, At 25* the flask was immersed in a water bath with the temperature range between 24-26�C, and at 350 the reaction was heated to the reflux temperature of ethyl ether with the temperature remaining constant over the reaction period. These results are summarized in Section C of the Result Section.


C. Effect of Ratio Change

The effect of changing the mole ratio of trimethoxyboroxine to Grignard reagent at 25�C was also studied. The reaction procedure was identical with that given except the mole ratio of trimethoxyboroxine to Grignard reagent of 1:3, 1:6, 1:9, and 1:12 were used. These results are summarized in Section D of the Result Section.


D. Effect of Solvents

The effect of changing the solvent on the yield of borinic acid was studied. Three solvents, ethyl ether, tetrahydrofuran, and toluene were used in this study. The procedure was the same as given in the preceding paragraph with a 1:9 mole ratio of trimethoxyboroxine to phenylmagnesium bromide and varying the temperature. When toluene was used as the solvent, the Grignard reagent was prepared using ether as the solvent, toluene was added, and the ether







18

distilled off. The results from this studyiare summarized in Section E of the Result Section.


V. Reaction of Phenylmagnesium Bromide With Triphenylboroxine

One tenth of a mole of phenylmagnesiumtbromide was prepared by the method described previously. Fifty milliliters of anhydrous ether was added to the Grignard reagent. A solution of 7.8 grams (06 025 mole) of triphenylboroxine dissolved in 50 ml. of ethyl ether and 50 ml, of tetrahydrofuran was added to the Grignard reagent at 250C. The tetrahydrofuran was added because the triphenylboroxine was not soluble in ether. The remainder of the reaction is the same as described for the reaction of Grignard reagents with trimethoxyboroxine. A yield of 7 grams of aminoethyl diphenylborinate was isolated, 41.5 percent based on phenylboronic anhydride. VI. Reaction of Phenylmagnesium Bromide With a Mixture of Trimethoxyboroxine and Trimethyl Borate

A study was conducted in which a mixture of trimethoxyboroxine and trimethyl borate was reacted with phenylmagnesium bromide in ethyl ether as solvent at 250. In one series of experiments, solutions containing mixtures of the two boron compounds in mole ratios of trimethoxyboroxine to trimethyl borate of 1:1, 1:3, and 2:1 were added to the Grignard reagent. The mole ratio, of phenylmagnesium. bromide to boron was 3:1. In the other series of experiments, mixtures of the








19

two boron compounds in mole ratios of trimethoxyboroxine to trimethyl borate of 1:3 and 2:1 were added to phenylmagnesium bromide. The mole ratio of Grignard reagent to boron was 1:1. These results are summarized in Section I of the Result Section.


VII. Isolation and Purification of Reaction Products A. General Procedure for Aromatic Borinic Acids

The following isolation and purification method, adapted from a procedure of Letsinger (2), was used for all aromatic borinic acids. The ether solution from the reaction of trimethoxyboroxine with the Grignard reagent was heated on a steam bath to remove the solvent. One hundred milliliters of water was added to the resultant oil and the mixture was heated on a steam bath for 15 minutes to remove the boronic acid that had dissolved in the ether. The oil containing nearly pure borinic acid was separated from the water and diluted with 250 ml. of ether. A mixture of ethanolamine in an equal volume of water was added to the ether solution and stirred until the precipitate, aminoethyl borinate, ceased to form. The solid was collected by filtration, and washed with three 50 mll. portions of water. The ester was recrystallized by dissolving it in the least possible amount of hot ethyl alcohol. Water was added until the first appearance of crystals. The solution was cooled in an ice-water bath, the solid ester removed by filtration, and dried in a vacuum desiccator.









B. Procedure for Isolating Aliphatic Borinic Acids

Three general methods were used to isolate the aliphatic borinic acids.

1. Direct isolation as the anhydride. The ether solution from the reaction of trimethoxyboroxine with the Grignard reagent was distilled under reduced pressure to remove the solvent. The borinic acid was then distilled as the borinic acid anhydride under reduced pressure through a ZO cm. glass helices packed column. The anhydrides were kept under a nitrogen atmosphere throughout the isolation. The boiling points of the anhydrides are listed in Table IV.

2. Attempt to obtain stable solid esters. The borinic acid

anhydride and an excess of alcohol or phenol were placed in a flask with benzene as the solvent. The flask was equipped with a stirrer, Dean-Stark azeotrope trap, and a condenser. The solutions were refluxed and the water removed by azeotroping with the solvent. The following alcohols or phenols were used: p-nitrophenol, ethanolamine, o-aminophenol, benzoin and c--naphthol. An ester that was stable to air was never obtained.

3. Use of vacuum apparatus. The cyclic borinic acids were

very susceptable to oxygen or moisture. For this reason a vacuum manifold was built in order to distill the anhydride. The manifold consisted of a 25 mm. glass tube that was attached to a manometer, Dry Ice-isopropyl alcohol trap, and vacuum pump. Attached to the








a'

manifold were three three-way stopcock and standard tapered joints to which the distillation flask and receivers were connected. A flask containing the ether solution from the reaction was placed in the system and the ether distilled under reduced pressure. The product was distilled using the following technique. After the solvent was removed, the stopcock to the distillation flask was closed and the system evacuated. The manifold was closed to the vacuum pump and the stopcock to the distillation flask opened. The material vaporized and collected in the receiver which was immersed in the Dry Iceisopropyl alcohol bath.


VIII. Characterization of Products A. Molecular Weight

The molecular weights were determined by the freezing point depression using cyclohexane as the solvent. The following is a typical calculation of the molecular weight using di(n-butyl)borinic anhydride as the example.

Freezing point of cyclohexane 5. 700C

Freezing point of solution 4. l5*C

Freezing point depression 1. 550 C

Molality of solution 0. 0768m

Weight of cyclohexane 8. 3452g

Weight of anhydride 0. 1708g

Molecular Wieight 263

Calculated Molecular Weight 266






TABLE IV

BOILING POINTS AND MOLECULAR WEIGHTS OF BORINIC ACID ANHYDRIDES


Name Boiling Point �C Molecular Weight

Obs. Obs. Catc.

Di(n-propyl)borinic Anhydride 34 /1 mm. 215 211

Di(iso-propyl)borinic Anhydride 49 /Z2mm. 213 211
a
Di(n-butyl)borinic Anhydride 107-8/2 mm. 263 266

Di(soc-butyl)borinic Anhydride 90-2/2 mm. 268 266

Di(n-amyl)borinic Anhydride 130-3/1 mm. 309 313

Di(sec-amyl)borinic Anhydride 120-2/1 mm. 315 313

Di(cyclohexyl)borinic Anhydride 147-9/1 mm.

Cyclotetramethyleneborinic Anhydride 70 / 35mm.

Cyclohexamethyleneborinic Anhydride 80 / 33mmtur


'Literature boiling point 137/9mn. (17).










B. Boron Analysis

Boron analyses were obtained following a modified procedure

of Johnson (35). The following is a typical analysis using di(n-butyl)borinic acid anhydride. A sample, 0. 1217 grams, was placed in a 50 ml. flask along with 1 ml. of 30 percent hydrogen peroxide and 10 ml. of water. The solution was reflu.-ed for two hours. A small amount of manganese dioxide was added to act as a catalyst in the decomposition of the peroxide. The solution was cooled to room temperature, 10 grams of mannitol was added, and the solution was titrated with 0. 0974 N sodium hydroxide. The titration was followed by a Beckman Model N pH meter.

A typical calculation to, illustrate the method used for determining the boron content of the products is given below.

9. 32 ml. of NaOH used

0. 00932 1 x 0. 0974 N = 0. 000908 eq.

0. 000908 x 10. 82 =0.O00982 gof boron

0.00982 x 100 = 8.09% boron
0.1217

Theoretical 8.13% 'boron

These results are summarized in Table V.







TABLE V

ANALYSIS OF BORINIC ACID ANHYDRIDESa


Name Found Calc.
C H B C H B

Di(n-propyl)borinic Anhydride ---- ---- 9. 91 ---- 10. 03

Di(iso-propyl)borinic Anhydride ---- ---- 9.90 101 03

Di(n-butyl)borinic Anhydride 66.8 12.8 8.09 72.2 13,55 8.13

Di(sec-butyi)borinic Anhydride 73.51 14.0 8,17 72.2 13,55 8,13

Di(n-amyl)borinic Anhydride 77.8 14.3 6.69 74.6 13.66 6,71

Di(sec-amyl)borinic Anhydride 75.0 13.7 6.70 74.6 13.66 6.71

Di(cyclohexyl)borinic Anhydride ---- ---- 5.87 5.84

Cyclotetramethyleneborinic Anhydride ---- ---- 9, 95 ---- 14, 4

Cyc lohexamethyleneborinic Anhydride ---- - .--- --- -- ----


Skokie, Illinois.


Boron


aCarbon and hydrogen analysis were by Micro-Tech Laboratory, analysis obtained from a procedure of Johnson (35).







25

B. Reaction W�ith Other Reagents

I. Reaction With the Sodium Salt of Malonic Ester A. General Procedure

In a 1-liter flask equipped with a stirrer and condenser was

placed 80 grams (0. 5 mole) of redistilled diethyl malonate, 400 ml. of anhydrous ether, and 200 ml. of 1, 4-dioxane. To this solution was added 11. 5 grams (0. 5 mole) of sodium metal, The oxide coating on the sodium metal had been removed by first washing the sodium in ethyl alcohol and then toluene. The solution was refluxed until all of the sodium metal had reacted.

In a 2-liter flask equipped with a stirrer, condenser, and addition funnel was placed a solution of 30 grams (0. 2 mole) of trimethoxyboroxine in 200 ml. of anhydrous ether. The sodium salt of malonic ester was added dropwise over a period of three hours. Precipitation occurred immediately on the addition of the sodium salt. The reaction mixture was stirred for twenty-four hours.


13. Isolation of the Products

The solid from the reaction of trimethoxyboroxine with the sodium salt of malonic ester was removed by filtration and dried. The solid was hydrolyzed by two methods.

1. Basic hydrolysis, The solid was added to a concentrated sodium hydroxide solution. The mixture was refluxed until a clear solution was obtained. The basic solution was cooled to room tem-










perature and acidified with concentrated hydrochloric acid. The only boron-containing compound isolated was boric acid.

Z. Acidic hydrolysis. The solid was added to a 1 N hydrochloric acid solution and stirred until the complex had hydrolyzed. A waterinsoluble liquid and water-insoluble solid formed. The liquid was distilled at reduced pressure and was proved to be diethyl malonate. The solid was boric acid.


II. Reaction With Carbonyl Compounds A. General Procedure

In a 125 ml. Erlenmeyer flask was placed the carbonyl compound, either acetaldehyde, propionaldehyde, or acetone, and trimethoxyboroxine in various concentrations, The reaction mixture was placed in the refrigerator for three days. In some cases a solid precipitated from solution,


B. Isolation of Products

The solid that formed from the reaction was removed by filtration and was proved to be boric acid. The liquid was washed with water, dried over anhydrous magnesium sulfate, and distilled, With a mole ratio of trimethoxyboroxine to carbonyl compound of 1:3, the trimers of acetaldehyde, propionaldehyde, and acetone were formed, but no boric acid was formed. With a mole ratio of trimethoxyboroxine to propionaldehyde of 1:200 a viscous liquid of unknown composition was isolated along with boric acid.















RESULTS

A. The Nature of the Reaction


Trimethoxyboroxine reacts readily with both aliphatic and

aromatic Grignard reagents at temperatures between -60 and 35*C: Near room temperature and with stoichiometric or higher mole ratios of Grignard reagent to boroxine, the products are bcrinic acids along with smaller quantities of boronic and boric acids. The yield of boronic acid can be increased by lowering the mole ratio of Grignard reagent to boroxine or by lowering the temperature. During the reaction a precipitate forms which is destroyed when the reaction mixture is hydrolyzed. Ether, tetrahydrofuran, and toluene are satisfactory solvents. An equation to represent the reaction is: (CH3OBO)3 + 6 RMgBr -3 3 R 2 B-O-MgBr + 3 CH3 OMgEBr


13. Summary of Yields of Borinic Acids Prepared
Under Comparable Conditions

Table VI shows the yield of borinic acids isolated as the aminoethyl esters for the aromatic compounds, and as the anhydride for the aliphatic compounds. The yields range from 17 to 77 percent, and are usually higher than those reported for other methods of preparation (24, 25, 26).










TABLE VI

YIELD OF BORINIC ACIDS IN REACTION OF GRIGNARD REAGENTS WITH TRIME THOXYBOROXINEC


Compound Formula Yieldd

Diphenylborinic Acida (C6H5)zBOH 6z. 4

Di(o-tolyl)borinic Acida (o-CH 3-C 6H4)zO 59,3

Di(m-tolyl)borinic Acida (M-CH3 -C 6H4)BOH 60.3

Di(p-tolyl)borinic Acida (-CH3- C6H5)z BOH 33.0

Di(p-anisyl)borinic Acida (p-CH30- 6H4)2BOH 38.0

Di( c--naphthyl)borinic Acid ( - CoH7) BOH 62.0
Di(p-biphenyl)borinic Acida (p- 6H5- C6H4)2 BOH 41.7


Di(p-chlorophenyl)borinic Acid a (p-Ci-C6H4) BOH 21. 5;21. 3

Di(n-propyl)borinic Acidb (n-C 3H7)2 BOH 60,3

Di(iso-propyl)borinic Acidb (iso-C H ) BOH 32,0
__ 3 72
Di(n-butyl)borinic Acidb (n-C 4H9) 2BOH 75. 0;77, 0

Di(sec-butyl)borinic Acidb (sec- C4H9)2BOH 21. 0;24, 0

Di(n-amyl)borinic Acidb (n-C5HI I) BOH 50. 0;52, 0

Di(sec-amyl)borinic Acidb (sec-C5H1 1)2BOH 46.0;47.0

Di(cyclohexyl)borinic Acidb (C6H 1 )23OH 16, 0;18.0

Cyclotetramethyleneborinic Acidb (C4H8)BOH 17.0

Cyclohexamethyleneborinic Acid (C5H10) BOH 20.0
aIsolated as aminoethyl ester. bIsolated as anhydride.
CReaction conducted at 250 with a mole ratio of 9:1 of Grignard reagent
to boroxine.
dyield based on boroxine.







29

C. Effect of Temperature

A study of the effect of temperature on the reaction was undertaken in an effort to find an optimum reaction temperature and with the hope of obtaining some insight into the reaction mechanism. The results of this study are summarized in Tables VII and VIII. The reaction temperature was varied from -60 to 350 using phenylmagnesium bromide and from 4 to 35' using n-butylmagnesium bromide. In each case the ratio of Grignard reagent to boroxine was 9:1. The phenylborinic acid was isolated as its aminoethyi ester. The n-butylborinic acid was isolated as its anhydride, The yields of the borinic acid derivatives in each series increased with increasing temperature up to 250C. Above 25� the yields decreased moderately.

The small yield of the borinic acid at low temperatures may be

explained by reasoning that the activation energy is too high to form an appreciable amount of the borinic acid at these temperatures. The fact that appreciable yields of boronic acids were obtained at the lower temperatures suggests that the activation energy is higher for the addition of the second aryl group than for addition of the first aryl group. The decrease in yield at 35' may be due to the instability of the ring and formation of the trisubstituted borane.

The yields of the n-butylborinic anhydride were significantly

higher at each temperature than the yields of the aminoethyl phenylborinate. This may be due to a difference in the reactivities of the












TABE.LE VII

EFFECT OF TEMPERATURE IN THE REACTION OF
TRIMETHOXYBOROXINE WITH PHENYLMAGNESIUM BROMIDEa


Temperature 0C Percent Yield of
Aminoethyl Diphenylborinateb


-60 11.2

6 18.4

16 42.6

25 62.4

35 51.7

aOne to nine mole ratio of trimethoxyboroxine to Grignard reagent.
byield based on boroxine.



TABLE VIII

EFFECT OF TEMPERATURE IN THE REACTION OF
TRIMETHOXYBOROXINE WITH n-BUTYLMAGNESIUM BROMIDEa


Temperature 0C Percent Yield of
Di (n-butyl)borinic Anhydrideb


4 56.4;58. 0

25 75.0;77.0

35 66. 4;67. 0

aOne to nine mole ratio of trimethoxyboroxine to Grignard reagent.
bYield based on boroxine.










aromatic and the aliphatic Grignard reagents with boroxine. It could also be due to the different isolation procedures used in each case. Interestingly, less material precipitated from solution during the reaction of the aliphatic reagent than during the reaction with aromatic Grignard reagent, This suggests that the lower yield with the aromatic compound is due in part to the removal of some reagent from the reaction zone.


D. Effect of Varying the Ratio of Boroxine to Grignard Reagent Tables IX and X summarize the results obtained by varying the mole ratio of trimethoxyboroxine to either phenylmagnesium bromide (Table IX) or n-butylmagnesium bromide (Table X) at 25�C. Maximum yields of borinic acids--62 percent in the case of the aromatic Grignard reagent and 76 percent with the aliphatic Grignard reagent--were obtained at a mole ratio of one mole of boroxine to nine moles of Grignard reagent. This mole ratio corresponds to an excess of Grignard reagent. The stoichiometric ratio is one mole of boroxine to six moles of Grignard reagent. Lower yields of borinic acids were obtained at mole ratios of 1:3 and 1:12. In the former case the yield of borinic acid was only 16 percent. Since the Grignard reagent is in short supply here, this low yield probably reflects the higher reactivity of the boroxine toward initial attack of the Grignard reagent. In support of this, a good yield of boronic acid was observed. The decreased yields of borinic acids at a mole ratio of 1:12 may be caused by











TABLE IX

EFFECT OF CHANGING THE MOLE RATIO OF TRIMETHOXYBOROXINE
TO PHENYLMAGNESIUM BROMIDE AT 250C


Mole Ratio of Percent Yield of
Boroxine to Grignard Aminoethyl Diphenylborinate a

1: 2b 13.0

1:3 16. 7;15.0

1:6 56.2

1:9 62.4

1:12 43. 7;45. 0

aYield based on boroxine.
b Addition of Grignard reagent to boroxine.



TABLE X

EFFECT OF CHANGING THE MOLE RATIO OF TRIMETHOXYBOROXINE
TO n-BUTYL MAGNESIUM BROMIDE AT 25*G


Mole Ratio of Percent Yield of
Boroxine to Grignard Di(n-butyl)borinic Anhydridea


1:6 44. 0;45. 0

1:9 75.0;77.0

1:1Z 52. 0;56. 0
aYield based on boroxine.







3 3

rupture of the boroxine ring or by destruction of the Grignard reagent by reaction with itself.

Reversing the order of adding reactants, adding the Grignard reagent to the boroxine, and using a 2:1 mole ration of phenylmagnesium bromide to boroxine resulted in a six fold increased in the yield of boronic acid compared to the normal reaction. The yield of borinic acid under these conditions was 13 percent, approximately the same as that obtained in the normal procedure.


E. Effect of Solvent

The reaction was carried out successfully in three solvents, i. e. ethyl ether, tetrahydrofuran, and toluene, and at several temperatures in each solvent. The results of these experiments are given in Table XI. The yields of borinic acid were highest in ethyl ether and significantly lower in tetrahydrofuran. This is probably due to the greater complexing tendency of tetrahydrofuran. An especially low yield of borinic acid was obtained when the reaction was carried out in toluene at 1000C. This was caused by thermal decomposition of the boroxine ring into trimethyl borate and boric anhydride prior to reaction with the Grignard reagent. Since the anhydride does not react with the Grignard reagent while the trimethyl borate reacts to give a good yield of triarylborane, R3 B, the yield of borinic acid was decreased sharply.













TABLE XI

EFFECT OF SOLVENT IN THE REACTION OF TRIMETHOXYBOROXINE
WITH PHENYLMAGNESIUM BROMIDEa


Temperature �C Solvent Percent Yield of
Aminoethyl Diphenylborinateb


6 Ethyl Ether 18.4

25 Ethyl Ether 62.4

25 Tetrahydrofuran 15. Z

25 Toluene 34.5

35 Ethyl Ether 51.7

50 Tetrahydrofuran 43.0

73 Tetrahydrofuran 35.0

100 Toluene 9.2

aOne to nine mole ratio of trimethoxyboroxine with Grignard reagent.
byield based on boroxine.







35

F. Effect of Some Electronic Factors

The influence of some electron-donating and withdrawing substituents in the Grignard reagent has been examined. These results are summarized in Table XII. In general, the results indicate that the highest yields of ester are obtained with the phenyl and O--naphthyl compounds. The p-biphenyl and p-anisyl Grignard gave approximately the same yield (40 percent of theoretical), somewhat lower than that obtained from the first two members of the series. The yield in the case of p-tolyl compound was 33 percent, somewhat less than the p-anisyl and p-biphenyl and significantly less than the m-tolyl which was 60, 2 percent. Finally the p-chlorophenyl Grignard gave a poor yield (21.5 percent).

In general, it appears that both electron-donating and electronwithdrawing substitutes decrease the yield of borinic acid in this reaction.


G. Effect of Some Steric Factors

The study of steric effects in this reaction is illustrated in Tables XIII and XV. The results of reacting phenyl, o-tolyl, ck-naphthyl, and mesityl Grignard reagents with boroxine, Table XIII, gives a clear indication of the susceptibility of the reaction to steric factors. Substitution of one ortho-methyl group has virtually no effect on the yield of borinate, whereas blocking both ortho positions of the Grign.rd completely prevents the reaction. Only a













TABLE XII

EFFECT OF SOME ELECTRONIC FACTORSa


Compound


Aminoethyl Aminoethyl Aminoethyl Aminoethyl Aminoethyl Aminoethyl Aminoethyl


Percent Yiehd of
Borinate


Diphenylborinate Di( o.-naphthyl)borinate Di(p-biphenyl)borinate Di(p-anisyl)borinate Di(p-tolyl)borinate Di(p-chlorophenyl)borinate Di (m- tolyl)bo rinate


aMole ratio of one to nine of boroxine to Grignard


reagent at 256C


62.4 62.0

41.7 38.0 33.0 21.5

60.3
















TABLE XIII

EFFECT OF STERIC FACTORS IN THE REACTION OF ARYL GRIGNARD REAGENTS WITH TRIMETHOXYBOROXINEa


Compound Percent Yield of
Borinateb


Aminoethyl Diphenylborinate 62.4

Aminoethyl Di(o-tolyl)borinate 59.3

Aminoethyl Di( c--naphthyl)borinate 62.0

Aminoethyl Dimesitylborinate 0.0


nOne to nine mole ratio of boroxine to Grignard reagent at 250C.
byield based on borinate.












TABUTL XIV

EFFECT OF STERIC FACTORS IN THE REACTION OF ALKYL
GRIGNARD REAGENTS WITH TRIMETHOXYBOROXINEa


Compound Percent Yield of
Anhydrideb


Di(n-butyl)borinic Anhydride 75; 77

Di(n-propyl)borinic Anhydride 60

Di(n- amyl)borinic Anhydride 50; 52

Di(sec-amyl)borinic Anhydride 46;47

Di (iso-propyl)borinic Anhydride 32

Di(sec-butyl)borinic Anhydride 21;Z4

Di(cyclohexyl)borinic Anhydride 16; 18

Di (te rt-butyl)bo rinic Anhydride 0

Di(tert-amyl)borinic Anhydride 0

aOne to nine mole ratio of boroxine to Orignard reagent at 250C.
byield based on boroxine.







39

small yield of the boronic acid was isolated when the reaction time between mesitylmagnesium bromide and boroxine was increased from five to fifteen hours,

Table XIV summarizes the study of steric factors in the prepataion of alkyl borinic anhydride. Highest yields are obtained when primary Grignard reagents are used. The lower yield in the preparation of di(n-amyl)borinic anhydride may be e:-plained by assuming that the alkyl chain coils on itself preventing the attack of the Grignard reagent. Single branching on the 4--carbon decreases the yield as evidenced from the reaction of sec-amyl, iso-propyl, and sec-butyl Grignard reagents. Two groups attached to the o--carbon prevent reaction of the Grignard reagent as seen in the reaction of tert-butyl and tert- amyl Grignard reagents. If the two groups on the v-carbon are tied back as in the case of the cyclohexyl Grignard reagent, small yields of borinic anhydride are obtained.


H. Results of Reaction of Phenylmagnesium Bromide With Other Boroxines

I. Tri-n-butoxyboroxine

Substituting tri-n-butoxyboroxine for the methoxy compound resulted in a decrease in the yield of borinic ester from 62.4 to 35. 0 percent.










II. Triphenylboroxine

Substituting triphenylboroxine for the methoxy compound resulted in a yield of 41. 5 percent of the borinic ester. This decrease in yield from the methoxy compound may be due partly to tl tetrahydrofuran that was added to dissolve the anhydride. This reasoning is possible because the yield of borinic acid decreases when tetrahydrofuran is used as the solvent,

I A Comparison of the Reactivity of Trimethoxyboroxine With Trimethyl Borate

A study was undertaken to compare the reactivity of trimethoxyboroxine with trimethyl borate. Letsinger (2) has reported that the reaction of Grignard reagents with borates must be carried out at

-600C to prevent the formation of trisubstituted borane, R3Bj which is the major product at room temperature. Since the boroxine and borate have similar structures, this study may give some insight into whether the boroxine ring is stable during the addition of the Grignard reagent.

Table XV summarizes the results obtained from the study of the reaction of trimethoxyboroxine and trimethyl borate with phenylmagnesium bromide. When the yield of borinate is calculated on total boron product, the yield of ester is highest (56 percent) when only boroxine is present and lowest (4. 8 percent) when only the borate is present. The borate is assumed to react with Grignard reagent to








TABLE XV

A COMPARISON OF THE REACTIVITY OF TRIMETHOXYBOROXINE WITH TRIMETHYL BORATEa



Mole Ratio of Mole Ratio of Percent Yield of Aminoethyl Diphenylborinate
Boroxine to Borate Grignard to Boron Based on Based on Based on
Total Boron Borate Boroxine


1:0 3:1 56.0 ---- 56.0

2: 1 3:1 44.0 313.0 51.7

2:1 1:1 10.0 70.2 11.7

1:1 3:1 42.5 170.0 56.8

1:3 3:1 47.0 94.0 94.0

1:3 1:1 16.6 33.2 33.2

0:1 3:1 4.8 4.8

a
Reaction carried out at 25�C.










form the triaryl borane, R3 B. W]ith the mole ratio of Grignard reagent to boron of 1:1, the stoichiometry is in favor of the boronic acid, instead of the borinic acid. The lower yield of borinate at these concentrations may be rationalized by assuming that the borate, reacting with three Grignard reagents, decreases the amount of reagent available for reaction with the boroxine.

When the yield of borinate is based on the amount of borate present, the yields are above 100 percent in the cases when the Grignard reagent is in excess (Grignard to boron ratio of 3:1) except when the boroxine-borate ratio is 1:3 (94 percent). This would indicate that the borinate is produced from both the boroxine and borate.

Basing the yield on the boroxine, these results suggest that in

mixtures of borate and boroxine, the yields of borinate depend almost entirely on the boroxine present, and are not appreciably affected by the presence of the borate. For example, with a mole ratio of boroxine to borate of 2:1, using excess Grignard reagent, the yield of borinate is 51. 7 percent whereas when this mole ratio of boroxine to borate is 1: , again using excess Grignard reagent, the yield of borinate is 56. 2 percent.

The reason for carrying out the reaction with a limited amount of Grignard reagent (Grignard-boron ratio of 1:1) was to verify the difference in reactivity of the boroxine and borate. With an excess of boroxine (boroxine to borate ratio of Z: 1), the yield of borinate based











on total boron was 10.0 percent, while a ratio of 1:3 produces a yield of 16. 6 percent. This may indicate that the first aryl group reacts with equal ease with the boroxine and borate. This could account for the increased yield when more borate is present.


J. Cyclic Borinic Acids

I. Introduction

The previous work has indicated that the substitution of one R group on the boron atom facilitates the attack of a second molecule of the Grignard reagent. This suggests that reaction with di-Grignard reagents might produce cyclic borinic acids. Such acids might be of interest, but in addition this synthetic route might provide an opening for compounds of the type

CH-CH

CH CH


OH

which have theoretical interest because of the possibility of aromatic character. This aromatic character would be due to the interaction of the five electrons with the five pi-orbitals, four on carbon and one on boron.


II. Results of Studies With Di-Grignard Reagents

Di-Grignard reagents prepared from 1, 4-dibromobutane and 1, 5-dibromopentane were reacted with trimethoxyboroxine in an










attempt to prepare the cyclic borinic acids, cyclotetramethyleneborinic acid, I, and cyclohexamethyleneborinic acid, II.

CH 2 CHU2 CH

CH2 CH2 CH2

B CH2 CH2
OH B

OH
II
I II

The material isolated from the reaction was spontaneously combustable and reacted with oxygen even in an ether solution. Boron analysis on the material was low indicating that either the compound has previously oxidized or that the compound prepared was not the borinic acid.


K. Results of Studies With the Sodium Salt of Malonic Ester

When the sodium salt of malonic ester was caused to react with trimethoxyborox:ine a solid, containing boron, precipitated from solution. On hydrolysis the only products isolated were boric acid and diethyl malonate. This could be interpreted as meaning that the boron-carbon bond is susceptible to hydrolysis or that a complex between the boroxine and the sodium salt of malonic ester formed. This complex would be hydrolyzed by water giving the products, boric acid and diethyl malonate. More experiments are necessary to reach any definite conclusions.










L. Results of Studies With Carbonyl Compounds

The formation of the trimers of acetaldehyde, propionaldehyde, and acetone is not surprising because other acids like zinc chloride, sulfur dioxide, sulfuric acid catalyze the reaction. The formation of the material when a small concentration of trimethoxyboroxine was used cannot be explained and more experiments are necessary.















DISCUSSION


The results of this study have raised a number of questions

concerning the mechanism of the reaction. These questions can be conveniently discussed under three headings:

A. The role of the Grignard reagent

B. The role of the boroxine compound

C. Formation of the carbon-boron bond

Such a classification is only for convenience since the mechanism of the reaction involves a simultaneous interplay among a variety of factors including the Grignard reagent, the boroxine, and the process whereby the carbon-boron bond is formed.


A. The Role of the Grignard Reagent: The Case for Stepwise Addition

Perhaps the initial question concerning the role of the Grignard reagent in the reaction should be: Does this reagent react with the boroxine to form the borinic acid in one or two steps? Stated in another way the question becomes: Is the path of the reaction more nearly represented by the equation:
0 - R O _/
R2Mg + MeO-B B + MeOMg

0- R 0-










or by the sequence of equations


/O /O
*%Mg + MeO-B - R-B + RMgOMe

00

0- o* 2 Mg +R-B B \3 + RMg
R -0


Some insight into this question might be provided by examining the products obtained when the reaction is carried out with less than stoichiometric quantities of Grignard reagent. Two cases given in Table IX will suffice to illustrate the point. In the first case the reaction was carried out so that there was one mole of boron for each mole of Grignard reagent present. Product analysis showed that a large yield (over 50 percent) of boronic acid was formed along with a low yield (16 percent) of borinic acid. In a second experiment a slightly smaller mole ratio of Grignard reagent was employed and the Grignard reagent was added to the boroxine. Product analysis showed less borinic acid (13 percent) and more boronic acid than in the previous case. This strongly suggests a two step addition process. More support for this mechanism was obtained by analyzing the products from reactions containing higher mole ratios of Grignard reagent to boroxine. In these experiments the yield of boronic acid decreased and the yield of borinic acid increased, exactly as would be predicted for a stepwise addition process.











If the two step process occurs, the activation energies for the

individual steps should be different. This means the ratio of the two steps should have different temperature coefficients. Hence the ratio of the yield of boronic to the yield of borinic acid should change markedly with temperature. The results indicate that this is the case. At -60' this ratio is approximately 6, and it decreases systematically with increasing temperature reaching a value of less than 0. 5 at 25*C. Crude estimates of relative ratio indicates that at -600 the rate of step one might be as much as thirty times faster than step two while at 250 step one might be as much as five times that of step two.

A final experiment pertinent to the case for the two step process was conducted by preparing triphenylboroxine and causing this compound to react with phenyimagnesium bromide. The reaction proceeds readily with a good yield of the borinic acid, demonstrating that the second step in the process is a feasible one.

The experiments discussed above provide strong support for the two step mechanism. They also raise at least two more questions: (1) What species from the Grignard reagent attacks the boroxine? (2) Why does the second step in the sequence appear to be so much slower than the first step? An attempt to answer the first question will be postponed until Part C of the Discussion. Some insight into the second question can be obtained by examining the role of the boroxine in the reaction.










B. The Role of the Boroxine

Before discussing specific experiments to elucidate the role of the boroxine in the reaction, a brief discussion of the reaction of trimethyl borate with Grignard reagents will be given. Methyl borate reacts with Grignard reagents at -60*C to give fair yields of borinic acids. Above this temperature good to excellent yields of triarylor trialkylboranes are obtained. V 'orkers in this field have postulated

(2) that the reaction proceeds by a stepwise process with the first step being the slowest. It has been observed (12) that boronic esters, lIB(OR),, react much more readily than borate esters, B(OR)3, with Grignard reagents.

Since trimethoxyboroxine is very similar in structure to methyl borate, the sluggish rate of the second step in the reaction sequence with Grignard reagents should be explained. The most obvious explanation is that the boroxine ring is stable during the reaction and that while the first step in the sequence involves the displacement of a methoxy group by another organic group, the second step involves formation of an intermediate of the type R R
B

0 0

3GO-b B-OCH
\ 0 3











While this intermediate accounts for most of the observed facts such as the sluggish second step and the very low yield of triarylboranes, some workers (36) consider it unlikely since water, alcohol. and other bases easily decompose the boroxine ring. They reason that the attacking Grignard reagent may also be regarded as a base, ostensibly with sufficient strength to rupture the ring- -eospecially in the second step. However, the results of the following data show conclusively that the ring is not ruptured during the course of the reaction. Perhaps the most convincing evidence emerges from an examination of what would happen if the ring were to rupture during the process, In this event two possibilities exist. Either the open chain structure will rapidly rearrange to give methyl borate and boric acid anhydride according to the equation (MeOBO) 3 B 2 03 + (MeO) 3B

or the Grignard reagent will react with the open chain structure with the same or greater rapidity than it reacts with methyl borate. Now the difference between these two alternates is this: If boric acid anhydride forms, each mole of boroxine wrill produce one mole of methyl borate; if the open chain is involved, each mole of boroxine is equivalent to three moles of methyl borate. In either case, because of the great reactivity of the borate, the yield of borinic acid would be only a small fraction of the yield actually observed. This is strong support for ring stability.










Two other pieces of evidence add support to this notion. The first is the result of an experiment in which the boroxine ring was forced to decompose in the presence of Grignard reagent. The second is the result of a study of the reaction of Grignard reagents with mixtures of methyl borate and boroxine.

When the boroxine is added to the Grignard reagent in toluene at 1000C, the ring is ruptured thermally and the open chain form rearranges rapidly to methyl borate and a precipitate of boric acid anhydride. The methyl borate then reacts in the usual way giving a very low yield (7. 5 percent) of borinic acid. This is much lower than obtained in any reaction involving the boroxine and shows that the ring is stable during reactions at lower temperatures. In the study involving mixtures of methyl borate and the boroxine, experiments with excess Grignard reagents and with excess boron compounds were conducted (Table XV). Yields of borinic acids in reactions with excess Grignard reagent were, within experimental error, those expected from the boroxine present. Apparently the methyl borate reacted almost completely to give triphenylborane. This again illustrates that the ring has remained intact. The more critical experiments in this study were those in which less than the stoichiometric quantities of Grignard reagent were used. In these experiments the yields were all low, but the higher yields of borinic acids were obtained for those solutions richer in methyl borate than in boroxine. These experiments










indicate that the first step in the reaction of both methyl borate and the boroxine with Grignard reagents proceeds at about the same rate but that the second step is much faster for the borate.

The preceding paragraphs summarize the role of the boroxine in the reaction, but they also raise several questions. One of these is: If a methoxy group of the boroxine is displaced in the first step, what is displaced in the second step? Some insight into this and other questions can be obtained by concentrating on the information available concerning formation of the carbon-boron bond.


C. Formation of the Carbon-Boron Bond: The Case
for Displacement on Magnesium

Experimental data which provides information pertaining to bond formation in reactions usually comes from kinetic studies or from the effect of ele ctron- donating or electron- withdrawing substituents on rates or yields. In this work the experiments have established that the first step in the reaction is displacement of a methoxy group of the boroxine at a rate comparable to that of a similar displacement in methyl borate. ,The nature of the bond formation process, in this step might be described by one of three processes. These are represented by the following equations:

OMe
1.) R + B - R---- B----OMe - R-B + OMe
O 0 0 0 0 0
1 1 1 1 1 1












R Me 0Mg + 0-B
0-


Mg-OMe

R


X-Mg +
I
14


- fR-B
0 0 1 1


+/

O-


O-Me



0 0
I I


X Me 0O/p- Mg--Q--13

R 0/0
R -B

0-


X-Mg--- O-Me
I I
I I
R----3

0 0 I I


+ XMgOMe


In the first of these processes a carbanion attacks the boron atom displacing the methoxide ion. In the second process, the magnesium atom attacks an oxygen atom with the resulting displacement of the R groups as a carbanion. The third process consists of a simultaneous attack on the oxygen and boron atom, The result suggests but does not establish that the third process is responsible for bond formation. Evidence for this comes from studies of the effect of electrical factors (Table XII) and steric factors (Tables XIII and XIV) on the yields of products, From the studies of electrical factors, it has been observed that electron-donating substituents in tV Grignard










reagent consistently decrease the yield of borinic acids. For example, the yield of borinic acid is reduced by a factor of 1. 5: 2 when methoxy, methyl, or phenyl groups are substituted in the para position of phenylmagnesium bromide. Since these substituents should increase the strength of the carbon-boron bond formed in the reaction, it is unlikely that the bond is formed by direct attack of a free carbanion (process 1) on the boron atom. If such attack occurred, the presence of electron-donating substituents would be expected to increase rather than decrease the yields. However, if the initial attack involved displacement on a magnesium atom (process 2 or 3), the presence of electron-donating groups would retard the process by strengthening the magnesium-carbon bond. This bond must be broken while the magnesium-oxygen bond is being formed.

Some unusual steric factors observed in this reaction can be explained in terms of the magnesium displacement. The sensitivity of the reaction to steric factors appears to be contradictory when cornparing the results of the study of steric factors in the reaction of aryl and alkyl Grignard reagents. While no decrease in yield is observed with one ortho substituent on the phenyl ring as compared to the unsubstituted Grignard reagent, a decrease in yield of one-third is observed when the alkyl Grignard reagent has branching on the o.-carbon. This observation may be explained in terms of one of the three processes. All three processes would be sensitive to steric requirements, but










process 3, the concerted mechanism, should be most sensitive to attack of bulky reactants, for in this case four atoms must attain an exact arrangement before reaction. More hindrance to formation of this strongly oriented four-rnembered ring transition state would be provided by the alkyl groups than the aromatic groups, This is because of the great rigidity of the carbon atoms in the aromatic ring as compared with the free-to-rotate carbon atoms in the aliphatic structure.

The extreme sensitivity of the reaction to steric factors is evidenced by the comparison of the yields of borinic acids obtained, by reaction of the following Grignard reagents: n-butyl (76 percent), phenyl (62 percent), o-tolyl (59 percent), iso-propyl (32 percent), and sec-butyl (21 percent). The n-butyl Grignard reagent would show the least steric hindrance because of the unsubstituted ot.-carbon, The aryl Grignard reagents, phenyl and o-tolyl, would be intermediate in the order because of the rigid structure, and the aliphatic Grignard reagents, iso-propyl and sec-butyl, because of free rotation of the bonds would show the most steric hindrance.

The discussion of the formation of the boron-carbon bond was mainly concerned with the attack of the first reagent. Although no evidence was obtained about the second attack, the similarity of-the two displacements would indicate the same concerted mechanism.















CONCLUSION


The reaction between trimethoxyboroxine and either aryl or

alkyl Grignard reagents was found to proceed to give borinic acids. The yields ranged from 17 to 77 percent. Borinic acid anhydrides of di(n-propyl), di(iso-propyl), di(n-butyl), di(sec-butyl), di(n-amyl), di(sec-amyl), and dicyclohexylborinic acids were prepared while only the di(n-butyl)borinic anhydride had previously been reported,

Di-Grignard reagents, prepared from 1, 4-dibromobutane and 1, 5-dibromopentane were reacted with trimethoxyboroxine in an attempt to prepare cyclic borinic acids. The results were inconclusive.

The versatility of the reaction was also demonstrated in the

reaction of phenyl Grignard reagents with triphenylboroxine to yield the diphenylborinic acid.

From a study of the effects of temperature, ratio of trimethoxyboroxine to Grignard reagents, solvent, steric factors, and competition between trimethoxyboroxine and trimethyl borate, a reasonable mechanism has been postulated. This mechanism accounts for all facts observed in this study.










The initial attack is probably


R-MgX +


OCH
B

0 0 i 1


R
-4 B

o 0y0


Followed by.


R-MgX


+ I
/
0 1


X-Mg---- -OCH{
I I 3
I I
R- --- B

0 0
I I


+ MgX(OCH3)




R
/
/R ------B
I z /\
B Mg ----- 0 0
\/ I I
0 X
I


R R
B

XMg- -O 0
I I


This same series of reactions can occur on the other boron atoms of the ring. On hydrolysis the ring is ruptured yielding the borinic acids.














BIBLIOGRAPHY


1. A. Michaelis, M. Benrens, J. Robinson, and W. Geisler, Ber., 27 244 (1894); J. Chem. Soc., 66, 190 (1894).

2. 1%. L. Letsinger and I. Skoog, J. Am. Chem. Soc., 77, 2491 (1955).

3. J. R. Johnson and M. G. van Campen, J. Am. Chem. Soc., 60, 121 (1938).

4. L. H. Long and D. Dollimore, J. Chem. Soc., 1953, 3902.

5. R. L. Letsinger, I, Skoog, and M. Remes, J. Am. Chem. Soc., 76. 4047 (1954).

6. R. L. Letsinger and I. Skoog, J. Am. Chem. Soc., 76, 4174 (1954).

7. R. L. Letsinger and H. Remes, J. Am. Chem. Soc., 77, Z489 (1955).

8. D. M. Mikhailov and V. A. Vaver, Dokla' Akad. Nauk, S.S.S.R., 102. 531 (1955); C. A., 50 4813d (1956).

9. T. P. Povlock and W. T. Lippincott, J. Am. Chem. Soc., 80, 5409 (1958).

10, B. MA. Mikhailov and V. A. Vaver, Izvest. Akad. Nauk. S.3S. S. R.,
Otdel. Khim, Nauk., 1957, 989; C. A., 52, 4532f (1958).

11. B. M. Mikhailov and T. A. Shchegoleva, Izvest. Akad. Nauk.
S.S. S.R., Otdel Khim. Nauk., 1955, 1124-5; C. A., 50,
IIZ335c (1956).

12. R. L. Letsinger and 1. R. Nazy, 3. 2r Chem., 23, 914
(1958),









13. B. M. Mikhailov and P. M. Aronovich, Izvest. Akad. Nauk.
S. S. S. R., Otdel. Khim. Nauk., 1955, 859; C. A., 50, 9320f
(1956).

14. K. Torssell, Acta. Chem. Scand., 9, 239 (1955); C. A., 49,
I0214c (1955).

15. R. L. Letsinger and I. Skogg, J. Am. Chem. Soc., 77, 5176
(1955).

16. E. V. Abel, W. Gerrard, and M. F. Lappert, J. Chem. Soc.,
1957, 112.

17. W, Gerrard, M. F. Lappert, and R, Shaffermann, J. Chem.
Soc., 1957, 3828.

18. E. W. Abel, W. Gerrard, and M. F. Lappert, J, Chem. Soc.,
1957, 3833.

19. M. F. Lappert, Chem. Revs., 56, 959 (1956).

20. F. Bean and J. R. Johnson, J. Am. Chem. Soc., 54, 4415
(1932),

21. P. R. Ogle, Jr., Ph.D. Thesis, Michigan State University,
1955.

22. L. L. Quill, P. R. Ogle, Jr., and W. T. Lippincott,
Unpublished Report.

23. L. L. Quill, P. R. Ogle, Jr., L. G. Kallender, and W. T.
Lippincott, "The Reaction of Trimethoxyboroxine With
Aromatic Amines, " Abstracts 129th Meeting, American
Chemical Society, Dallas, April, 1956.

24. W. Konig and W. Scharrnbeck, J. prakt. Chem., 128, 153 (1930);
C. A., 25, 927 (1931).

25. N. N. Mel'nikov and M. S. Rokitskaya, J. Gen. Chem. (U. S. S. R.),
8, 1768 (1938); C. A., 33, 4969 (1939).

26. N. N. Mel'nikov, J. Gen. Chem. (U.S.S.R.), 6, 636 (1936);
C. A., 30, 5571 (1936).







60

27. B. M. Mikhailov and F. B. Tutorskaya, Izvest. Akad. Nauk.
S.S.S.R., Otdel. Khim. Nauk., 1959, 1127; C. A., 53, 6990g
(1959).

28. H. I. Schlesinger, L. Horvitz, and A. B. Berg, J. Am. Chem.
Soc., 58, 407 (1936).

29. A. L. Borisov, Izvest. Akad. Nauk. S.S. S.R., Otdel. Khim.
Nauk., 1951, --02; C. A., 46, 2995d (1952). 30. V. R. Nev, Der., 88, 1761 (1955). 31. J. Goubeau and Keller, Z. anorg. Chem., 267, 1 (1951). 32. C. F. H. Allen and S. Converse, "Organic Syntheses, t" Coil.
Vol. I, John Wiley and Sons, Inc., New York, 1943, p. 226. 33. L. J. Fieser, "Experiments in Organic Chemistry, " Third
Edition, D. C. Heath and Co., Boston, 1955, p. 266.

34. H. Gilman, P. D. Wilkinson, W. P. Fishel, and C. H. Meyer,
J. Am. Chem. Soc., 45, 150 (1923).

35. H. R. Snyder, I. A. Kuck, and J. R. Johnson, J. Am. Chem.
Soc., 60, 105 (1938).

36. J. M. Davidson and C. M. French, J. Chem. Soc., 1960, 191.















BIOG11APHICAL SKETCH


Thomas Paul Povlock was born in Salamanca, New York, on January 22, 1934. Upon graduating from Salamanca High School in 1951, he attended Kent State University, Kent, Ohio, and he was awarded the degree of Bachelor of Science in 1955.

He was admitted to the Graduate School of Michigan State

University, East Lansing, Michigan, and received the Master of Science degree in 1957.

In 1957 he was admitted to the Graduate School of the University of Florida as a graduate assistant. In 1959 he was appointed as a graduate teaching assistant.















This dissertation was prepared under the direction of the

chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Arts and Sciences and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 13, 1960


Dean', College of Arts and *ScAnces



Dean, Graduate School Supervisor Committee: Chairman


0 �r,




Full Text

PAGE 1

BOROXINES: THEIR REACTIONS AND RING ST ABILITY By THOMAS PAUL POVLOCK A DISSERTATION PRESENTED TO THE GRADUATE OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA August, 1960

PAGE 2

ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. W. T. Lippincott for the encouragement, aid, and guidance which he so generously gave during the course of the investigation and the preparation of this dissertation. The author is indebted to the members of his graduate committee for their helpful suggestions in writing this dissertation. The author also wishes to express his gratitude to Mr. and Mrs. John Cribbs for the technical preparation of the dissertation. ii

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES INTRODUCTION A. Purpose B. Historical I. Borinic Acids II. Boronic Acids III. T rimethoxybo roxine EXPERIMENTAL A. Reaction With Grignard Reagents Page ii vi 1 1 2 2 5 6 12 12 I; General Procedure 12 II. Preparation of Intermediates 13 III. Preparation of Grignard Reagents 15 IV. Reaction of Grignard Reagents With Trimethoxyboroxine 16 V. Reaction of Phenylmagnesium Bromide With Triphenylboroxine 18 VI. Reaction of Phenylmagnesium Bromide With a 1v1ixture of Trimethoxyboroxine and Trimethyl Borate 18 iii

PAGE 4

VII. Vlll; TABLE OF CONTENTS (continued) Isolation and Purification of Reaction Products Characterization of Products B. Reaction With Other Reagents Page 19 21 25 I. Reaction With the Sodium Salt of Malanie Ester 25 Reaction With Carbonyl Compounds RESULTS 26 27 A. The Nature of the Reaction 27 B. Summary of Yields of Borinic Acids Prepared Under Comparable Conditions 27 C. Effect of Temperature 29 D. Effect of Varying the Ratio of Boroxine to Grignard Reagent 31 E. Effect of Solvent 33 F. Effect of Some Electronic Factors 35 G. Effect of Some Steric Factors 35 H. Results of Reaction of Phenylmagnesium Bromide With Other Boroxines 39 I. II. Tri-_!!-butoxyboroxine Triphenylboroxine I. A Comparison of the Reactivity of Trimethylboroxine 39 40 With Trimethyl Borate 40 iv

PAGE 5

TABLF. OF CONTENTS (continued) J. Cyclic Borinic Acids I. Introduction I II. ' Results of Studies With Di-Grignard Reagents K. Results of Studies With the Sodium Salt of Malonic Ester L. Results of Studies With Carbonyl Compounds DISCUSSION A. The Role of the Grignard Reagent: The Case for Page 43 43 43 44 45 46 Stepwise Addition 46 B. The Role of the Boroxine C. Formation of the Carbon-Boron Bond: The Case for Displacement on Magnesium CONCLUSION BIBLIOGRAPHY BIOGRAPHICAL SKETCH V 49 52 56 58 61

PAGE 6

LIST OF TABLES Table Page I. Physical Constants of Aromatic Borinic Acids and Esters II. Physical Constants of Aliphatic Borinic Acids and Esters III. Physical Constants of Unsymmetrical Borinic Esters IV. Boiling Pointo and Molecular Weighta of Borinic Acid Anhydrides v. Analysis of Borinic Acid Anhydrides VI. Yields of Borinic Acids in Reaction of Grignard Reagents Vv'ith Trimethoxyboroxine VIL Effect of Temperature in the Reaction of Trimethoxyboroxine 'With Phenylmagnesium Bromide VIII. Effect of Temperature in the Reaction of Trirncthoxyboroxine With ~-Butylmagnesium Bromide I}~. Effect of Changing the Mole Ratio of Trimethoxyboroxine to Phenylmagnesium Bromide at 25 C x. Effect of Changing the Mole Ratio of Trimethoxyboro:,dne to ,!!-Butylmagnesium Bromide at 25C XI. Effect of Solvent in the Reaction of Trimethoxyboroxine With Phenylmagnesium Bromide XII. Effect of Some Electronic Factoro XIII. Effect of Steric Factors in the Reaction of Aryl Grignard Reagents With Trimethoxyboroxine vi 8 10 11 22 24 28 30 30 32 32 34 36 37

PAGE 7

LIST OF TABLES (continued) Table Page XIV. Effect of Steric Factors in the Reaction of Alkyl Grignard Reagents With Trimethoxyboroxine 38 XV. A Comparison of the Reactivity of Trimethoxyboroxine Vlith Trimethyl Borate 41 vii

PAGE 8

INTRODUCTION A. Purpose This study was undertaken to examine the reaction of tri methoxyboroxine with organic bases and carbanion reagents, and to obtain more information about the nature of the reaction in which a carbon-boron bond is formed. The following reactions were studied: 1. Reaction with Grignard reagents 2. Reaction with sodio salts 3. Reaction with carbonyl compounds Only the first of these was studied in detail. This study was divided into two sections. The first was re ported by the author in his Master of 'Science thesis. In the work reported therein, it was found that aryl Grignard reagents react with t: imethoxyboroxine to yield aryl borinic acids. Eight aryl borinic acids, isolated as the aminoethyl esters, were prepared in yields of 21-62 percent. The effects of temperature, ratio of Grignard reagent to trimethoxyboroxine, steric, and electrical factors were studied. The second section of the study of the reaction of Grignard re agents with trimetho:x:yboroxine constitutes the major portion of this 1

PAGE 9

2 dissertation. It consists of a study of the reaction of alkyl Grignard reagents with trimethoxyboroxine. From this study nine alkyl borinic acids were prepared and characterized~ In an attempt to elucidate a possible mechanism for the reaction, a study of the effects of tem perature, ratio of trimethoxyboroxine to Grignard reagent, solvent, and steric factors was conducted along with the study of methods of isolation and purification of the borinic acids. In addition, reactions with di-Grignard reagents, preparation of a p ' ossible intermediate and its reaction \Vi.th : a Grignard reagent, and a series of competition ex periments involving mixture of trimethoxyboroxine and trimethyl borate were studied. An examination of the reaction of the sodium salt of malonic ester and simple aldehydes and ketones with trimethoxyboroxine .was under taken. B. Historical I. Borinic Acids Aryl borinic acids, having the general formula R 2 BOH,, have been known since 1894 when Michaelis (1) described the isolation of di phenylborinic acid from the hydrolysis of the corresponding chloride. Since the methods of isolation and characterization were doubtful~ it was not known whether or not the earlier workers had actually prepared the reported compounds. In the case of the diphenylborinic acid, it was not until 1955 that Letsinger (2) isolated and characterized this acid.

PAGE 10

3 The yield of borinic acid was unspecified in some of the earlier work and when reported ranged from 20-40 percent. Tables I. II. and III list the aryl, alkyl, and unsymmetrical borinic acids and esters for which the melting or boiling points have been reported. The boiling points of the acids which have been re corded are actually those of the anhydride. Letsinger (2) Dhowed that when diphenylborinic acid wac distilled, the anhydride was re covered. In some cases there are discrepancies it: the reported physical constants of the borinic acids as shown in Table I. Di(.E_-bromophenyl) borinic acid is listed with two different melting pointe. Di(.1?.-biphenyl)borinic acid is listed as melting above 300 C; but the isolated compound was never proved to be the borinic acid. It was found as a by-product in the preparation of the boronic acid and was assumed to be the borinic acid. Borinic acids have been prepared by use of one of the following reactions 1 followed by hydrolysi s . 1. . Moist air oxidation of trialkylborane (R 3 B) (3). 2. Reaction of trialkylborane with iodine to prepare the dialkyliodoborane (4). 3. Reaction of trialkyl borates [(R0) 3 B] or cyclic borates . /O-CH 2 RO-B I with an excess of alkyl or aryl Grignard reagents "0-CH 2 at -60 C (2, 5, 6, 7, 8).

PAGE 11

4 4. Reaction of trialkoxyboroxine [(ROB0) 3 ] \-vith an excess of aryl Grignard reagent at 25 C (9, 10). 5. Reaction of boronic esters {RB(OR 1 ) 2 ] with alkyl or aryl Grignard reagent at -60C (7, 8, 11, 12., 13, 14). Of these preparations method 3 is the most widely used procedure, although this author (9) showed that method 4 using trimethoi::yboroxine gave higher yields of aryl borinic acids. Method 5 is used for the preparation of unsymmetrical borinic acids. Two of the principal difficulties in the preparation of borinic acids are their separation from the boronic acid and the preparation of a . suitable derivative for characterization. Letsinger (5) developed a method of separation of the borinic ester from the boronic ester. This method consists of treating the mixture .of esters in ether solution with ammonia. The borinic ester complexes with ammonia and precipitates :from solution. The complex io then treated with a solution of ethanol amine in water or toluene to prepare the aminoethyl borinate. Thia ester is an excellent derivative because it can be prepared easily, recrystallizes readily from an ethanol-water mixture, melts sharply, analyzes well, and over a period of months shows no sign of decom posing. The aminoethyl esters of aryl (5, 9) and mixed aryl alkyl borini c acids ( 12) we re reported, but there are no reports of this ester of alkyl borinic acids.

PAGE 12

5 Letoinger (15) prepared a cyclic borinic acid, 5/3-aminoethoxy10, 11-dihydrodibenzo-(b, f)-borepin o:C)) B I 0-CH -CH -NH 2 2 2 Lappert (16, 17, 18) prepared a series of esters of diphenyland di(n-butyl)borinic acids and studied the reactions of the esters with reagents such as aqueouo base, hydrogen halides, boron chlorides, and phosphorus pentabromidc. II. Boronic Acids Boronic acids, having the general formula RB(OH) 2 , have received far more attention in the literature than the borinic acids, In a review article by Lappert (19), seventy alkyl and aryl bcronic acids were reported along with their melting points. The most important methods for the preparation of alkyl or aryl boronic acids depends on one of the following procedures. 1. The addition of a metal alkyl 01 aryl to a trialkyl borate and the hydrolysis of the boronic ester (20). 2. The addition of a metal alkyl or aryl to a boron trihalide to obtain the alkyl or aryl boron dihalide, which on hydrolysis gives the boronic acid (3).

PAGE 13

6 3. Anhydrous air oxidation of a trialkylborane to a boronic ester and the hydrolysis of the latter (3). III. Trimethoxyboroxine Trimethoxyboroxine, I. is believed to be a six membered ring containing alternate boron and oxygen atoms with one methoxy group attached to each boron atom. 0-CH I 3 B I\ 0 0 I I H C-0 -B B-OCH 3 \ I 3 0 I Ogle (21) has verified the structure by molecular weight determinations, elemental analysis, Raman spectra. and the solubility in anhydrous ether while one of its components, boric acid anhydride. is insoluble. Ogle, Lippincott, and Quill (22, 23} studied the reaction of tri metho,,yboroxine with alcohols and aromatic amines. The reaction with alcohols is illustrated in the following equation:

PAGE 14

7 The following equation illustrates the reaction of trimethoxy boroxine with aniline.

PAGE 15

n u TABLE I PHYSICAL CONSTANTS OF AROMATIC BORINIC ACIDS AND ESTERS R2BOR 1 Melting Point Boiling Point Reference R R' oc oc Phenyl a 210/ 1 Hydrogen mm. 1 /3 -Aminoethyl 189-190 2;5;9 Ethyl 138/10 mm. 15 !_!-Propyl 152/10 mm. 15 !_!-Butyl 158/10 mm. 15 3-Methylpropyl 153/10 mm. 15 1-Methylpropyl 152/10 mm. 15 Phenyl 140/ 0. 05mm. 15 2_-Tolyl f3 -Aminoethyl 181-181.5 9 3-Methylpropyl 127/ 5.5 mm. 10 ,!!!•Tolyl 8 -Aminoethyl 180-181 9 ..e,-Tolyl ./9 -Aminoethyl 186-187 9; 10 .:e,-Anisyl /3-Aminoethyl 173-174 9 Hydrogen a 107 24 1?, Bi phenyl R,-Aminoethyl 217-218 9 a Hydrogen 300 25 _E,-Chlorophenyl /3 -Aminoethyl 223-224 9

PAGE 16

9 TABLE I (continued) R BOR' 2 Melting Point Boiling Point Reference R R' oc oc Hydrogen a 75 26 ct. -Naphthyl .,8-Aminoethyl 202-203 9 Hydrogen a 105-106 5 /3-Naphthyl Hydrogen a 172 24 p-Bromophenyl Hydrogen a 113 24 Hydrogen a 82-84 1 /3 -Aminoethyl 236-237 2 2-Methyla 5-Chlorophenyl Hydrogen 81 26 alsolated as anhydride.

PAGE 17

10 TABLE II PHYSICAL CONSTANTS OF ALIPHATIC BORINIC ACIDS AND ESTERS P :-;OR' '"""" 2-Melting Point Boiling Point Reference R R' oc oc !!-Butyl !:•Butyl 114/20 mm. 17;27 a Hydrogen 137 / 9 mm. 17 3-Methylpropyl 94/ 8 mm. 17 1-Methylpropyl 96/ 9. mm. 17 Phenyl 116/ 0.1 mm. 17. Ethylene glycol 147/ O. mm. 6 4-Methylbutyl Ethylene glycol 147/ 0, 48mm. 6 Allyl !!-Butyl 61/ 6 mm. 27 Ethyl a -51 to -48 35/75 28 Hydrogen m.m. 2-Chlorovinyl Hydrogen a 66/ 3 29 mm.

PAGE 18

TABLE III PHYSICAL CONSTANTS OF UNSYMMETRICAL BORINIC ESTERS RR'BOR" Melting Point Boiling Point Reference R R' Rll oc oc Phenyl ci_ -Naphthyl fJ -Aminoethyl 228 5 Phenyl Benzyl /3 -A1ninoethyl 141-142 12 Phenyl 1-Methylheptyl /3 -Aminoethyl 9395 12 Phanyl _!:-Butyl /3 -Aminoethyl 108-108.5 12 I-' ..... 3-1v1ethylpropyl 96-97 / 3mm, 13 ~-Butyl _!:-Propyl _!:-Butyl 84-86 /l0mm. 11 Phenyl ~-Propyl 3-Methylpropyl 83-87 I 3mm. 13 Phenyl _-Tolyl _!:-Propyl 156-159/ 9mm. 30 .2,-Tolyl ,E_-Tolyl _!:-Propyl 173-175/ 10mm 14 2_-Tolyl !!!_-Tolyl ~-Propyl 173-175/llmm •. 14 _-Tolyl 2.-Chlorophenyl _!:-Propyl 176-177 / 10mm. 14 Phenyl Ethyl ~-Butyl 110-111/ 9mm. 14 Phenyl lvfothyl ~-Propyl 87-89 / 9mm. 14

PAGE 19

EXPERIMENTAL A. Reaction With Grignard Reagents I. General Procedure The general procedure used in carrying out the reaction of Grignard reagents with trimethoxyboroxine was previously reported by the author (9), and is summarized as follows: Grignard reagents from both aliphatic and aromatic bromides were prepared in anhy drous ether solution. Trimetho,:yboroxine, prepared from trimethyl borate and boric acid anhydride, was dissolved in anhydrous ether and added to each Grignard reagent. The reaction mixture was hydrolyzed and the ether layer removed and distilled yielding a liquid containing a mixture of boronic and borinic acids. The borinic acids derived from aliphatic Grignard reagents were distilled and converted to the corresponding anhydride. The borinic acids from aromatic Grignard reagentz were converted to the aminoethyl esters. The experimental details for each of the processes described above are given in Section IIA, IIIA, IVA, and. VII. 12

PAGE 20

13 II. Preparation of Intermediates A. Preparation of Trimethoxyboroxine The boroxine was prepared following a procedure of Goubeau and Keller (31). In a 500 ml. one-necked flat-bottomed flask were placed 69. 6 grams (1 mole) of boric acid anhydride and 103. 8 grams (1 mole) of trimethyl borate along with a glass covered stirring bar. The flask was connected to a reflux condenser and placed in a heating mantle on top of a magnetic stirrer. The reaction mixture was stirred and refluxed from two to four hours or until the boric acid anhydride dissolved ~ The flask was cooled to room temperature and then immersed in a Dry Ice-isopropyl alcohol bath until the contents of the flask froze. The flask was removed from the bath, permitted to warm to room temperature, and the trimethoxyboroxine stored in a glass stoppered Erlenmeyer flask. Special precautions were taken in the preparation of the boroxine. Moisture was excluded at all times. Freshly distilled trimethyl borate gave the best results~ Boric acid anhydride obtained from Baker Chemical Company contained both the powder and glass struc• ture. Only the powder was suitable for the boroxine and was separated from the glassy structure by screening through a 40 mesh screen. B. Preparation .2_! Triphenylboroxine (20) In a 2-liter three-necked flask equipped with a stirrer, condenser, and addition funnel were placed 48. 6 grams (2 moles) of magnesium

PAGE 21

14 turnings. The reaction was initiated with 50 ml. of anhydrous ether and 10 ml. of bromobenzene. After the reaction had started, the remainder of 316 grams (2 moles) of bromobenzene, which had previously been diluted with 1~150 ml. of anhydrous ether, was added to the flask over a period of five hours. The reaction mixture was refluxed for three hours after the addition. In a 3-liter three-necked flask equipped vn.th a stirrer,. concl;enser, and addition funnel were placed 500 ml. of anhydrous ether and 208 grams (2 moles) of freshly distilled trimethyl borate. The reaction flask was immersed in a Dry Ice-isopropyl alcohol bath and the. temperature maintained between -60 and -70C, The phenylmag nesium bromide was added slowly over a period of eight hours. The reaction mixture was allowed to slowly warm to room temperature; and hydrolyzed using a solution of 75 ml. of concentrated sulfuric acid in 600 ml. of water. The ether layer was removed and the water layer washed with ether. The ether extracts were combined and distilled on a steam bath to remove the solvent. The resulting liquid was extracted with a dilute Dodium hydro::dde solution, The aqueous layer was acidified and the solid removed by filtration, The phcnyl boronic acid was converted to the anhydride (triphenylboroxine) by dissolving in benzene and distilling of the water. The triphenylboro::dne was recrystallized from benzene and dried in a vacuum desiccator. The yield of phenylboronic anhydride was 105 grams, 43 percent based on trimethyl borate. Melting point 214-215C (lit. m. pt. 216G (25)).

PAGE 22

\ 15 III. Preparation of Grignard Reagent A.. General Procedure Grignard reagents were prepared following the procedure from Organic Syntheses (32). In a 1-liter three-necked flask equipped with a stirrer, addition funnel, and a parallel side arm holding a reflux condenser and a thermometer were placed 9. 7 grams (0. 4 mole) of magnesium turnings. The system was swept with dry nitrogen and a nitrogen atmosphere was maintained over the system throughout the reaction (33). Ten milliliterfl of anhydrous ether and 5 ml. of the bromide were added to the flask. After the reaction was initiatedt the stirring was started, and the remainder of O. 4 mole of the bromide, which had previously been diluted with 170 ml. of anhydrous ether, was added to the flask over a period of one hour. After the addition, the reaction mixture was reflu,;ed for two hours.. The con tents of the flask were cooled to room temperature, a 5 ml. aliquot was removed for analysis, and the ether solution of trimethoxyboroxine was added to the remainder of the Grignard reagent as described in Section IV. The 5 ml. aliquot was analyzed by titration (34). Twenty milli liters of distilled water was added along with an excess of standard hydrochloric acid to the aliquot, and the excess of acid was neu tralized with standard sodium hydroxide to a methyl orange end point.

PAGE 23

16 IV. Reaction of Grignard Reagents With Trimethoxyboroxine A. General Procedure The addition of trimethoxyboroxine to aryl and alkyl Grignard reagents was carried out in exactly the same manner for all the reactions. To a I-liter three-necked flask containing the Grignard reagent equipped with a stirrer. addition funnel. and a parallel . side arm holding a reflux condenser and a thermometer was added 100 ml. of anhydrous ether. The nitrogen atmosphere was maintained over the system~ From the Grignard analysis, the amount of trimethoxyboroxine needed was calculated. The boroxine , was dissolved in 150 ml. of anhydrous ether and was added dropwise over a period of two hours. After the addition of the boroxine, the reaction mbcture was stirred for five hours at constant temperature. Forty milliliters of concen trated hydrochloric acid in 300 ml. . of water waa used to hydrolyze the reaction mixture. The ether layer was separated and washed twice with 100 ml. portions of water. The borinic acid was isolated from the ether solutior .. by the procedure given in Section V. B. Effect of Temperature The . effect of temperature on the yield of borinic acid was studied, The procedure was the same as given in the preceding paragraph with temperatures of -60", 6 16, 25 and 35 and mole ratio of Grignard reagent to boro:,dne of 9: 1. The temperature was controlled

PAGE 24

1 ... t at -60 by immersing the reaction flask in a Dry Ice-isopropyl alco hol bath with the temperature range between -65 and -55C. At 6 the temperature was controlled between two degrees, At 25 the flask was immersed in a water bath with the temperature range be tween 24-26C, and at 35 the reaction was heated to the reflux temperature of ethyl ether with the temperature remaining constant over the reaction period. These results are summarized in Section C of the Result Section. C. Effect..! Ratio Change The effect of changing the mole ratio of trimethoxyboro,dne to Grignard reagent at 25 C was also studied. The reaction procedure was identical with that given except the mole ratio of trimethoxy boroxine to Grignard reagent of 1:3, 1:6. 1:9, and 1:12 were used. These results are summarized in Section D of the Result Section. D. Effect of Solvents The effect of changing the solvent on the yield of borinic acid was studied. Three solvents, ethyl ether, tetrahydrofuran, and toluene were used in this study. The procedure was the same as given in the preceding paragraph with a 1: 9 mole ratio of trimethoxy boroxine to phenylmagnesium bromide and varying the temperature. When toluene was used as the solvent, the Grignard reagent was prepared using ether as the solvent, toluene was added, and the ether

PAGE 25

18 distilled off. The results from this study are summarized in Section E of the Result Section. V. Reaction of Phenylmagnesium Bromide 1Vith Triphenylboro,dne One tenth of a 1nole of phenylmagnesium,bromide was prepared by the method described previously. Fifty milliliters of anhydrouo ether was added to the Grignard reagent. A solution of 7. 8 grams (0. 025 mole) of triphenylboroxine dissolved in 50 ml. of ethyl ether and 50 ml. of tetrahydrofuran was added to the Grignard reagent at 25C. The tetrahydrofuran was added because the triphenylboroxine was not soluble in ether. The remainder of the reaction is the same as described for the reaction of Grignard reagents with trimethoxy boroxine. A yield of 7 grams of aminoethyl diphenylborinate was isolated, 41. 5 percent based on phenylboronic anhydride. VI. Reaction of Phenylmagnesium Bromide a lv1ixture of Trimetho:::ryboroxine and Trimethyl Borate A study was conducted in which a mixture of trimethoxyboroxine and trimethyl borate was reacted with phenylmagnesium bromide in ethyl ether as solvent at 25, In one series of experiments, solutions containing mbctures of the two boron compounds in mole ratios of trimethoxyboroxine to trimethyl borate of 1: 1. 1: 3, and 2:1 were added to the Grignard reagent. The mole ratio, of phenylmagnesium. bromide to boron was 3: 1. In the other series of experiments, mixtures of the

PAGE 26

19 two boron compounds in mole ratios of trimethoxyboroxine to trimethyl borate of 1: 3 and 2: 1 were added to phenylmagnesium bromide. The mole ratio of Grignard reagent to boron was 1: 1. These results are summarized in Section I of the Result Section. VIL Isolation and Purification of Reaction Products A. General Procedure for Aromatic Borinic Acids The following isolation and purification method, adapted from a procedure of Letsinger (2), was used for all aromatic borinic acids. The ether solution from the reaction of trimethoxyboroxine with the Grignard reagent was heated on a steam bath to remove the solvent. One hundred milliliters of water was added to the resultant oil and the mixture was heated on a steam bath for 15 minutes to remove the boronic acid that had dissolved in the ether. The oil containing nearly pure borinic acid was separated from the water and diluted with 250 ml. of ether. A mixture of ethanolamine in an equal volume of water was added to the ether solution and stirred until the precipitate, aminoethyl borinate, ceased to form. The solid was collected by filtration, and washed with three 50 ml . portions of water. The ester was recrystallized by dissolving it in the leaot possible amount of hot ethyl alcohol. Water was added until the first appearance of crystals. The solution was cooled in an ice-water bath, the solid ester removed by filtration, and dried in a vacuum desiccator.

PAGE 27

B. Procedure for Isolating Aliphatic Borinic Acido Three :general-methodo were used to isolate the aliphatic borinic acids. 1. Direct isolation~ the anhydride, The ether solution from the reaction of trimethoxyboroxine with the Grignard reagent was distilled under reduced pressure to remove the solvent. The borinic acid was then distilled as the borinic acid anhydride under reduced pressure through a 20 cm. glaso heliceo packed column. The anhy drides were kept under a nitrogen atmosphere throughout the isola tion. The boiling points of the anhydrides are listed in Table IV. 2. Attempt ,!2 obtain stable solid esters. The borinic acid anhydride and an excess of alcohol or phenol were placed in a flask with benzene as the solvent. The flask was equipped with a stirrer, Dean-Stark azeotrope trap, and a condenser. The solutions"were reflu::iced and the water removed by azeotroping with the solvent. The following alcohols or phenols were used: _E-nitrophenol, ethanol amine; .2,-aminophenol, benzoin and o<...-naphthol. An ester that was stable to air was never obtained. 3. ~~vacuum apparatus. The cyclic borinic acids were very susceptable to oxygen or moisture. For this reason a vacuum manifold was built in order to distill the anhydride. The manifold consisted of a 25 mm. glass tube that was attached to a manometer, Dry Ice-isopropyl alcohol trap, and vacuum pump. Attached to the

PAGE 28

21 manifold were three three-way stopcock and standard tapered joints to which the distillation flask and receivers were connected. A flask containing the ether solution from the reaction was placed in the system and the ether distilled under reduced pressure. The product was distilled using the following technique. After the solvent was removed, the stopcock to the distillation flask was closed and the system evacuated. The manifold was closed to the vacuum pump and the stopcock to the distillation flask opened. The material vaporized and collected in the receiver which was immersed in the Dry Ice isopropyl alcohol bath. VIII. Characterization of Products A, Molecular Weight The molecular weights were determined by the freezing point depression using cyclohexane as the solvent. The following fa a typical calculation of the molecular weight using di(n-butyl)borinic anhydride as the example. Freezing point of cyclohexane Freezing point of solution Freezing point depression Molality of solution Weight of cyclohexane Weight of anhydride Molecular V/eight Calculated Molecular Weight s. 10c 4.15C 1. 55 C O. 0768m 8.3452g o. 1708g 263 266

PAGE 29

TABLE IV BOILING POINTS AND MOLECULAR WEIGHTS OF BORINIC ACID ANHYDRIDES Name Di(n•propyl)borinic Anhydride Di(iso-propyl)borinic Anhydride Di(n-butyl)borinic Anhydride Di(sec-butyl)borinic Anhydride Di(n-amyl)borinic Anhydride Di(sec-amyl)borinic Anhydride Di{cyclohexyl)borinic Anhydride Cyclotetramethyleneborinic Anhydride Cyclohexamethyleneborinic Anhydride a Literature boiling point 137/9mm. (17). Boiling Point .. C Obs. 34 /1 mm. 49 /22mm. 107-8/ 2 a mm. 90-2/2 mm. 130-3/1 mm, 120-2/1 mm. 147-9/1 mm. 70 /35mm. 80 /33mm. Molecular Weight Obs~ Cale. 215 211 213 211 263 266 268 266 309 313 315 313 ;:'-J t',J

PAGE 30

23 B. Boron Analysis Boron analyses were obtained following a modified procedure of Johnson (35). The following is a typical analysis using di(n-butyl) borinic acid anhydride. A sample, 0, 1217 grams, was placed in a 50 ml. flask along with l ml. of 30 percent hydrogen peroxide and 10 ml. of water. The solution was refhrn:ed for two hours. A small amount of manganese dioxide was added to act as a catalyst in the decomposition of the peroxide. The solution was cooled to room temperature, 10 grams of mannitol was added, and the solution was titrated with 0. 0974 N sodium hydroxide. The titration was followed by a Beckman Model N pH meter. A typical calculation to illustrate the method used for deter mining the boron content of the products is given below. 9. 32 ml. of NaOH used O. 00932 l x O. 0974 N = O. 000908 eq. O. 000908 x 10. 82 = O. 00982 g of boron O. 00982 x 100 = 8. 09% boron 0.1217 Theoretical 8. 13% boron These results are summarized in Table V.

PAGE 31

TABLE V ANALYSIS OF BORINIC ACID ANHYDRIDESa Name Di(n-propyl)borinic Anhydride Di(iso-propyl)borinic Anhydride Di(n-butyl)borinic Anhydride Di(scc-butyl)borinic Anhydride Di(n-amyl)borinic Anhydride Di(sec-amyl)borinic Anhydride Di(cyclohexyl)borinic Anhydride Cyclotetramethyleneborinic Anhydride Cyclohexamethyleneborinic Anhydride C ---66.8 73.51 77.8 75.0 Found H B 9.91 9.90 12.8 8.09 14.0 8. 17 14.3 6,69 13.7 6.70 5,87 9,95 Cale. C H B 10.03 10,03 72.2 13,55 8~ 13 72.2 13,55 8,13 74.6 13,66 6,71 74.6 13.66 6.71 5.84 14,4 aCarbon and hydrogen analysis were by Micro-Tech Laboratory, Skokie. Illinois. Boron analysis obtained from a procedure of Johnson (35). N ,r~

PAGE 32

B, Reaction With Other Reagents I. Reaction With the Sodium Salt of Malonic Ester A. General Procedure In a 1-liter flask equipped with a .stirrer and condenser was. placed 80 grams (O, 5 mole) of redistilled diethyl malonate. 400 ml, of anhydrous ether, and 200 ml. of 1, 4-dioxane, To this solution was added 11. 5 grams (0. 5 mole) of sodium metal. The oxide coating on the sodium metal had been removed.by first :washing the sodium in ethyl alcohol and then toluene. The solution was refluxed until all of the sodium metal had reacted. In a 2-liter flask equipped with a stirrer. condenser, and additio.n funnel was placed a solution of 30 gramn (0. 2 mole) of trimethoxyboro,dne in 200 ml. of anhydrous ether, The sodium salt of malonic ester was added dropwise over a period of three hours. Precipitation occurred immediately on the addition of the sodium salt. The reaction mixture was stirred for twenty-four hours. B. Isolation of the Products The solid from the reaction of trimethoxyboroxine with the sodium salt of malonic ester was removed by filtration and dried. The solid was hydrolyzed by two methods. l. Basic hydrolysb:i. The solid was added to a concentrated sodium hydroxide solution. The mixture was refluxed until a clear solution was obtained. The basic solution was cooled to room tern

PAGE 33

26 perature and acidified with concentrated hydrochloric acid. The only boron-containing compound isolated wao boric acid. 2. Acidic hydrolysis. The solid was added to a 1 N hydrochloric acid solution and stirred until the complex had hydrolyzed. A waterinsoluble liquid and water-insoluble solid formed. The liquid was distilled at reduced pressure and was proved to be diethyl malonate. The solid was boric acid. I!. Reaction With Carbonyl Compounds A, General Procedure In a 125 ml, Erlenmeyer flask was placed the carbonyl compound, either acetaldehyde, propionaldehyde. or acetone, and trimethmcy boroxine in various concentrations, The reaction mixture was placed in the refrigerator for three days. In some cases a solid precipitated from solution, B. Isolation of Products The solid that formed from the reaction was removed by fil tration and was proved to be boric acid. The liquid was washed with water, dried over anhydrous magnesium sulfate, and distilled. With a mole ratio of trimethoxyboroxine to carbonyl compound of 1: 3, the trimers of acetaldehyde, propionaldehyde, and acetone were formed, but no boric acid was formed. With a mole ratio of trimethoxyboroxine to propionaldehyde of 1: 200 a viscous liquid of unknown composition was isol:.J.ted along with boric acid.

PAGE 34

RESULTS A.. The Nature of the Reaction Trimethoxyboroxine reacts readily with both aliphatic and aromatic Grignard reagents at temperatures between -60 and 35 C, Near room temperature and with stoichiometric or higher mole ratios of Grignard reagent to boroxine, the products are bcrinic acids along with smaller quantities of boronic and boric acids. The yield of boronic acid can be increased by lowering the mole ratio of Grignard reagent to boro:idne or by lowering the temperature.. During the reaction a precipitate forms which is destroyed when the reaction mixture is hydrolyzed. Ether, tetrahydrofuran, and toluene are satisfactory solvents. An equation to represent the reaction is: (CH 3 OBO) 3 + 6 RMgBr -+ 3 R 2 B-O-MgBr + 3 CH 3 OMgBr B. Summary of Yields of Borinic Acids Prepared Under Comparable Conditions Table VI shows the yield of borinic acids isolated as the amino ethyl esters for the aromatic compounds, and as the anhydride for the aliphatic compounds. The yields range from 17 to 77 percent, and are usually higher than those reported for other methods of prep aration (24, 25, 26). 27

PAGE 35

28 TABLE VI YIELD OF BORINIC ACIDS IN REACTION OF GRIGNARD REAGENTS WITH TRlMETHOXYBOROXINEc Compound Formula Yieldd Diphenylborinic Acida (C 6 H 5 ) 2 BOH 62.4 Di(o-tolyl)borinic Adda (o-CH 3 -c 6 H 4 ) 2 BOH 59.3 Di(_!!!-tolyl)borinic Acida (m-CH 3 -C 6 H 4 ) 2 BOH 60.3 Di(E-tolyl)borinic Adda (2-CH 3 C 6 H 5 ) 2 BOH 33.0 Di(E-anisyl)borinic Acida {.e-CH 3 o-c 6 H 4 ) 2 BOH 38.0 Di( c(. -na.phthyl)borinic Acid a ( c,(.-c 10 H 7 ) 2 BOH 62.0 Di(E.-biphenyl)borinic Acida (E.-C 6 H 5 -,c 6 H 4 ) 2 BOH 41. 7 Di (E.chlorophenyl)borinic Acid a (.E,-Cl-C 6 H 4 ) 2 BOH 21. 5;21. 3 Di(_!!-propyl)borinic Acid b (n-C 3 H 7 ) 2 BOH 60.3 Di{iso-propyl)borinic Acid b (iso-C 3 H 7 ) 2 BOH 32.0 Di(_!!butyl)borinic Acid b (n-C 4 H 9 ) 2 BOH 75.0;77.0 Di(sec-butyl)borinic Acidb (s_ec-C 4 H 9 ) 2 BOH 21. 0;24. 0 Di~-amyl)borinic Acidb (n-C 5 H 11 ) 2 BOH 50.0;52.0 Di(sec-amyl)borinic Acidb (sec-C 5 H 11 ) 2 BOH 46.0;47.0 Di(cyclohexyl)borinic Acid b (C 6 H 11 ) 2 BOH 16.0;18.0 Cyclotetramethyleneborinic Acidb (C 4 H 8 )BOH 17. 0 Cyclohexamethyleneborinic Acidb (C 5 H 10 )BOH 20.0 alsolated as aminoethyl ester. blsolated as anhydride. cReaction conducted at 25 with a mole ratio of 9: 1 of Grignard reagent to boroxine. dYield based on boroxine.

PAGE 36

29 C. Effect of Temperature A study of the effect of temperature on the reaction was under taken in an effort to find an optimum reaction temperature and with the hope of obtaining some insight into the reaction mechanism. The results of this study are summarized in Tables VII and VIII. The reaction temperature was varied from -60 to 35 using phenylmag nesium bromide and from 4 to 35 using .!!-butylmagnesium bromide. In ea.ch case the ratio of Grignard reagent to boroxine was 9: 1. The phenylborinic acid was isolated as its aminoeth,rl ester. The .!!-butyl borinic acid was isolated as its anhydride. The yieldo of the borinic acid derivatives in each series increased with increasing temperature up to 25G. Above 2.5 the yields decreased moderately. The small yield of the borinic acid at low temperatures may be explained by reasoning that the activation energy is too high to form an appreciable amount of the borinic acid at these temperatures. The fact that appreciable yields of boronic acids were obtained at the lower temperatures suggests that the activation energy is higher for the addition of the second aryl group than for addition of the first aryl group. The decrease in yield at 35 may be due to the instability of the ring and formation of the trioubstituted borane. The yields of the .!!-butylborinic anhydride were significantly higher at each temperature than the yields of the aminoethyl phenyl borinate. This may be due to a difference in the reactivities of the

PAGE 37

30 TABLE VII EFFECT OF TEMPERATURE IN THE REACTION OF TRIMETHOXYBOROXINE WITH PHENYLMAGNESIUM BROMIDEa Temperature C -60 6 16 25 35 Percent Yield of b Aminoethyl Diphenylborinate 11. 2 18.4 42.6 62.4 51. 7 aOne to nine mole ratio of trimethoxyboroxine to Grignard reagent. bYield based on boroxine. TABLE VIII EFFECT OF TEMPERATURE IN THE REACTION OF TRIMETHOXYBOROXINE WITH n-BUTYLMAGNESIUM BROMIDEa Temperature C 4 25 35 Percent Yield of b Di(n-butyl)borinic Anhydride 56. 4;58. 0 75.0;77.0 66.4;67.0 aOne to nine mole ratio of trimethoxyboroxine to Grignard reagent. bYield based on boroxine.

PAGE 38

31 aromatic and the aliphatic Grignard reagents with boroxine. It could also be due to the different isolation procedureo used in each case. Interestingly, less material precipitated .from solution during the re action of the aliphatic reagent than during the reaction with aromat_~_c Grignard reagent. This suggests that the lower yield with the aro matic compound is due in part to the removal of some reagent from the reaction zone. D. Effect of Varying the Ratio of Boroxine to Grignard Reagent Tables IX and X summarize the results obtained by varying the mole ratio of trimethoxyboroxine to either phenylmagnesium bromide (Table IX) or .!!-butylmagnesium bromide (Table X) at 25C. Maximum yields of borinic acids--62 percent in the case of the aromatic Grignard reagent and 76 percent with the aliphatic Grignard reagent-were obtained at a mole ratio of one mole of boro:::dne to nine moles of Grignard reagent. This mole ratio corresponds to an excess of Grignard reagent. The stoichiometric ratio is one mole of boroxine to six moles of Grignard reagent. Lower yields of borinic acids were obtained at mole ratios of 1: 3 and 1: 12. In the former case the yield of borinic acid was only 16 percent. Since the Grignard reagent is in short supply here. this low yield probably reflects the higher reactivity of the boroxine toward initial attack of the Grignard reagent. In support of this. a good yield of boronic acid was observed. The decreased yields of borinic acids at a mole ratio of 1: 12 may be caused by

PAGE 39

32 TABLE IX EFFECT OF CHANGING THE MOLE RATIO OF TRIMETHOXYBOROXINE TO PHENYLMAGNESIUM BROMIDE AT 25C Mole Ratio of Boroxine to Grignard 1: 3 1:6 1: 9 1: 12 aYield based on boro::dne. Percent Yield of Aminoethyl Diphenylborinate a 13.0 16.7;15 .. 0 56.2 62.4 43. 7;45. 0 b Addition of Grignard reagent to boroxine. TABLE X EFFECT OF CHANGING THE MOLE RATIO OF TRIMETHOXYBOROXINE TO n-BUTYL MAGNESIUM BROMIDE AT 25C Mole Ratio of Boroxine to Grignard 1: 6 1: 9 1: 12 aYield based on boroxine. Percent Yield of Di(n-butyl)borinic Anhydride a 44. 0;45. 0 75.0;77.0 52.0;56.0

PAGE 40

33 rupture of the boroxine ring or by destruction of the Grignard reagent by reaction with itself. Reversing the order of adding reactants, adding the Grignard reagent to the boroxinc, and using a 2: 1 mole ration of phcnylmag nesium bromide to boroxine resulted in a six fold increased in the yield of boronic acid compared to the normal reaction. The yield of borinic acid under these conditions was 13 percent, approximately the same as that obtained in the normal procedure. E. Effect of Solvent The reaction was carried out successfully in three solvents, i.e. ethyl ether, tetrahydrofuran, and toluene, and at several temperaturen in each solvent. The results of these experiments are given in Table XI. The yields of borinic acid were highest in ethyl ether and signifi cantly lower in tetrahydrofuran. This is probably due to the greater complexing tendency of tetrahydrofuran. An eopccially low yield of borinic acid was obtained when the reaction was carried out in toluene at 100C. Thio was caused by thermal decomposition of the boroxine ring into trimethyl borate and boric anhydride prior to reaction with the Grignard reagent; Since the anhydride does not react with the Grignard reagent while the trimethyl borate reacts to give a good yield of triarylborane, R 3 B; the yield of borinic acid was decreased sharply.

PAGE 41

34 TABLE XI EFFECT OF SOLVENT IN THE REACTION OF TRIMETHOXYBOROXINE WITH PHENY LMAGNESIUM BROMIDE a Temperature C Solvent Percent Yield of b Aminoethyl Diphenylborinate 6 Ethyl Ether 18.4 25 Ethyl Ether 62.4 25 Tetrahydrofuran 15.2 25 Toluene 34.5 35 Ethyl Ether 51.7 50 Tetrahydrofuran 43.0 73 Tetrahydrofuran 35.0 100 Toluene 9.2 aOne to nine mole ratio of trimethoxyboroxine with Grignard reagent. bYield based on boroxine.

PAGE 42

35 F. Effect of Some Electronic Factors The influence of some electron-donating and withdrawing sub stituents in the Grignard reagent has been examined, These results are summarized in Table XII. In general, the re9.~lts indicate that the highest yields of ester are obtained with the phenyl and d...-naphthyl compounds. The ,E.-biphenyl and .E_-anisyl Grignard gave approximately the same yield (40 percent of theoretical), somewhat lower than that obtained from the first two members of the series. The yield in the case of ..e,-tolyl compound was 33 percent, somewhat less than the E-anisyl and l?,• biphenyl and significantly less than the _!!!-tolyl which was 60, Z percent. Finally the _e-chlorophenyl Grignard gave a poor yield (21. 5 percent). In general, it appears that both electron-donating and electron withdrawing substitutes decrease the yield of borinic acid in this reaction. G, Effect of Some Steric Factors The study of steric effecto in this reaction is illustrated in Tables XIII and :XIV. The results of reacting phenyl, .2,-tolyl, oL-naphthyl, and mesityl Grigna.rd reagents with boroxine, Table XIII, gives a clear indication of the susceptibility of the reaction to steric factors, Substitution of one ortho-methyl group has virtually no effect on the yield of borina.te, whereas blocking both ortho positions of the Grign?sd completely prevents the reaction. Only a

PAGE 43

TABLE XII EFFECT OF SOME ELECTRONIC FACTORSa Compound Aminoethyl Diphenylborina.te Aminoethyl Di( c:,(_-naphthyl)borinate Aminoethyl Di(E-biphenyl)borinate Aminoethyl Di(.e-anisyl)borinate Aminoethyl Di(,E_-tolyl)borinate Aminoethyl Di(,E_-chlorophenyl)borinate Aminoethyl Di(!:,:-tolyl)borinate Percent Yiebd of Borinate 62.4 62.0 41. 7 38.0 33.0 21. 5 60.3 aMole ratio of one to nine of bormdne to Grignard reagent at 25 6 C

PAGE 44

37 TABLE XIII EFFECT OF STERIC FACTORS IN THE REACTION OF ARYL GRIGNARD REAGENTS WITH TRIMETHOXYBOROXINEa Compound Aminoethyl Diphenylborinate Aminoethyl Di(o-tolyl)borinate Aminoethyl Di( ct.-naphthyl)borinate Aminoethyl Dimesitylborinate Percent Yield of Borinateb 62.4 59.3 62.0 o.o aOne to nine mole ratio of boro,dne to Grignard reagent at 25 C. bYield based on borinate.

PAGE 45

38 TABLE XIV EFFECT OF STERIC FACTORS IN THE REACTION OF ALKYL GRIGNARD REAGENTS WITH TRIMETHOXYBOROXINEa Compound Di(n .. butyl)borinic Anhydride Di(n-propyl)borinic Anhydride Di(n-amyl)borinic Anhydride Di(aec-amyl)borinic Anhydride Di(iso-propyl)borinic Anhydride Di(sec-butyl)borinic Anhydride Di(cyclohexyl)borinic Anhydride Di(tertbutyl)bo rinic Anhydride Di(tcrt-amyl)borinic Anhydride Percent Yield of Anhydrideb 75;77 60 50;52 46;47 32 21;24 16;18 0 0 a One to nine mole ratio of boroxine to Grignard reagent at 25 C. bYield based on boroxine.

PAGE 46

39 small yield of the boronic acid was isolated '\Vhen the reaction time between mesitylmagnesium bromide and boroxine was increased from five to fifteen hours. Table XIV summarizes the study of steric factors in the prep ataion of alkyl borinic anhydride. Highest yields are obtained when primary Grignard reagents are used. The lower yield in the prep aration of di(n-amyl)borinic anhydride may be e~-tplained by assuming that the alkyl chain coils on itself preventing the attack of the Grignard reagent. Single branching on the o<..-ca.rbon decreaseo the yield as evidenced from the reaction of ~-amyl, ~-propyl, and sec-butyl Grignard reagents. Two groups attached to the Cl(.•carbon prevent reaction of the Grignard reagent as seen in the reaction of .!!_t• butyl and .!!_1:amyl Grignard reagentn. If the two groups on the cit-carbon are tied back as in the case of the cyclohexyl Grignard reagent, small yields of borinic anhydride arc obtained. H. Results of Reaction of Phenylmagnesium Bromide With Other Boroxines I. Tri-~-butoxyboroxine Substituting tri-nbutoxyboroxine for the rnethoxy compound re oulted in a decrease in the yield of borinic eoter from 62. 4 to 35. 0 percent.

PAGE 47

II. Triphenylboroxine Substituting triphenylboro:ldne for the metho,::y compound re sulted in a yield of 41. 5 percent of the borinic ester. This decrease in yield from the methoxy compound may be due partly to tJ:e tetra hydrofuran that was added to dfosolve the anhydride. This reasoning is possible because the yield of borinic acid decreases when tetra hydrofuran is used ao the solvent. I. A Comparison of the Reactivity of Trimethmcyboroxine With Trimethyl Borate A study was undertaken to compare the reactivity of trimethoxy .. boroxine with trimethyl borate. Letsinger (2) hao reported that the . reaction of Grignard reagents ._~.,ith borates mu0t be carried out at -6OC to prevent the formation of trisubstituted borane; R ... B. which ., is the major product at room temperature, Since the boroxine and borate have similar structures, this study may give some insight into whether the boroxinc ring is stable during the addition of the Grignard reagent. Table XV summarizes the results obtained from the study of the reaction of trimethoxyboroxine and trimethyl borate with phenylmag nesium bromide. When the yield of borinate is calculated on tota.l boron product, the yield of ester is highest (56 percent) when only boroxine is present and lowest (4. 8 percent) when only the borate is present. The borate is assumed to react with Grignard reagent to

PAGE 48

TABLE XV A CO1v1PARISON OF THE REACTIVITY OF TRIMETHOXYBOROXINE WITH TRIMETHYL BORATEa Mole Ratio of .Mole Ratio of Percent Yield of Aminoethyl Diphenylborinate Boroxine to Borate Grignard to Boron Based on Based on Based on Total Boron Borate Boroxinc 1:0 3:1 56.0 56.0 2:1 3:1 44.0 313.0 51. 7 2:1 1:1 10 •. 0 70. 2 11. 7 :>. l-' 1:1 3: 1 42.5 170.0 56.8 l:3 3: 1 47.0 94.0 94.0 1: 3 1: 1 16.6 33. 2 33.2 0:1 3:1 4.8 4.8 aRea.ction carried out at 25" C.

PAGE 49

form the triaryl borane, R 3 B. With the mole ratio of Grignard reagent to boron of 1: 1, the stoichiometry io in favor of the boronic acid, instead of the borinic acid. The lower yield of borinate at these concentrations may be rationalized by assuming that the borate, re acting with three Grignard reagents, decreases the amm.mt of reagent available for reaction with the boroxine. When the yield of borinate is based on the amount of borate present, the yields are above 100 percent in the cases when the Grignard re agent is in excess (Grignard to boron ratio of 3: 1) except when the boroxine-borate ratio is 1: 3 (94 percent), This would indicate that the borinate is produced from both the boroxine and borate, Basing the yield on the boroxine, these results suggest that in mixtures of borate and boroxine, the yields of borinatc depend 8-lmost entirely on the boroxine present, and are not appreciably affected by the presence of the borate. For example. with a mole ratio of boroxine to borate of 2: 1, using excess Grignard reagent, the yield of borinate is 51, 7 percent whereas when this mole ratio of boroxine to borate is 1: 1. again using excess Grignard reagent, the yield of borinate is 56. 2 percent. The reason for carrying out the reaction with a limited amount of Grignard reagent (Grignard-boron ratio of 1:.1) was to verify the difference in reactivity of the boroxine and borate. With an excess of boroxine (boroxine to borate ratio of 2: 1 ), the yield of borinate baaed

PAGE 50

on total boron was 10. 0 percentt while a ratio of 1: 3 produces a yield of 16. 6 percent. This may indicate that the first aryl group reacts with equal ease with the boroxine and borate. This could account for the increased yield when more borate is present. J. Cyclic Borinic Acids I. Introduction The previous work has indicated that the substitution of one R group on the boron atom facilitates the attack of a second molecule of the Grignard reagent. This suggests that reaction with di-Grignard reagents might produce cyclic borinic acids~ Such acids might be of interestt but in addition this synthetic route might provide an opening for compounds of the type CH--CH II JI CH CH ~B/ I OH which have theoretical interest because of the possibility of aromatic character. This aromatic character would be due to the interaction of the five electrons with the five pi-orbitals, four on carbon and one on boron. II. Results of Studies With Di-Grignard Reagents Di-Grignard reagents prepared from 1, 4-dibromobutane and 1, 5-dibromopentane were reacted with trimethoxyboroxine in an

PAGE 51

attempt to prepare the cyclic borinic acids~ cyclotetramethylene borinic acid, I, and cyclohexamethyleneborinic acid, II. CH 2 CH 2 I I CH 2 CH 2 ~/ B I OH I /CH 2 '\ fH2 IH2 CH 2 CH 2 ~B/ I OH II The material isolated from the reaction was spontaneously com bustable and reacted with oxygen even in an ether solution. Boron analysis on the material was low indicating that either the compound has previously oxidized or that the compound prepared was not the borinic acid. K. Results of Studies With the Sodium Salt of Malonic Ester When the sodium salt of malonic ester was caused to react with trimethoxyboro:dne a solid, containing boron, precipitated from solution. On hydrolysis the only products isolated were boric acid and diethyl malonate. This could be interpreted as meaning that the boron-carbon bond is susceptible to hydrolysis or that a complex be tween the boroxine and the sodium salt of malonic ester formed. This complex would be hydrolyzed by water giving the products, boric acid and diethyl malonate. More experiments are necessary to reach any definite conclusions.

PAGE 52

L. Results of Studies with Carbonyl Compounds The formation of the trimers of acetaldehyde, propionaldehyde, and acetone is not surprising because other acids like zinc chloride, sulfur dioxide, sulfuric acid catalyze the reaction. The formation of the material when a small concentration of trimethoxyboroxine was used cannot be explained .and more experimento are necessary.

PAGE 53

DISCUSSION The results of this study have raised a number of questions concerning the mechanism of the reaction. These questions can be conveniently discussed under three headings: A. . The role of the Grignard reagent B. The role of the boroxine compound C. F'ormation of the carbon-boron bond Such a classification is only for convenience since the mechanism of the reaction involves a simultaneous interplay among a variety of factors including the Grignard reagent, the boroxine, and the process whereby the carbon-boron bond is formed. A. The Role of the Grignard Reagent: The Case for Stepwise Addition Perhaps the initial question concerning the role of the Grignard reagent in the reaction should be: Does this reagent react with the boroxine to form the borinic acid in one or two steps? Stated in another ...,,ay the question becomes: Is the path of the reaction more nearly represented by the equation: o/ R 2 Mg + MeO-B . " oR '-.~/ B + MeOMg+ / "-R O

PAGE 54

or by the sequence of equations R Mg+ 2 / oMeO-B "'oo/ R-B "' o+ RMgOMe + + RMg Some insight into this question might be provided by examining the products obtained when the reaction is carried out with less than stoichiometric quantities of Grignard reagent. Two cases given in Table IX will suffice to illustrate the point. In the first case the reaction was carried out so that there was one mole of boron for each mole of Grignard reagent present. Product analysis showed that a large yield {over 50 percent) of boronic acid '-Vas formed along with a low yield (16 percent) of borinic acid. In a second experiment a slightly smaller mole ratio of Grignard reagent was employ-ed and the Grignard reagent was added to the boroxine. Product analysis showed less borinic acid {13 percent) and more boronic acid than in the previous case. This strongly suggesto a two step addition pro cess. More support for this mechanism was obtained by analyzing the products from reactions containing higher mole ratios of Grignard reagent to boroxine. In these e::::periments the yield of boronic acid decreased and the yield of borinic acid increased, exactly as would be predicted for a stepwise addition process.

PAGE 55

If the two step process occurs, the activation energies for the individual steps should be different. This means the ratio of the two steps should have different temperature coefficients. Hence the ratio of the yie Id of boronic to the yield of borinic acid should change markedly with temperature. The results indicate that this is the case. At -60 this ratio is approximately 6 1 and it decreaseo systematically with increasing temperature reaching a value of less than O. 5 at 25 C. Crude eatimates of relative ratio indicates that at -60 the rate of step one might be as much as thirty times faster than step two while at 25 step one might be as much as five times that of step two. A final experiment pertinent to the case for the two step process was conducted by preparing triphenylboroxine and causing this com pound to react with phenylmagnesium bromide. The reaction proceeds readily with a good yield of the borinic acid, demonstrating that the second step in the procesG is a feasible one. The experiments discussed above provide strong support for the two step mechanism. They also raise at least two more questions: (1) What species from the Grignard reagent attacks the boroxine? (2) why does the second step in the sequence appear to be so much slower than the first step? An attempt to answer the first question will be postponed until Part C of the Diocussion. Some insight into the second question can be obtained by examining the role of the boroxine in the reaction.

PAGE 56

/.0 B. The Role of the Boroxine Before discussing specific e":perimento to elucidate the role of the boroxine in the reaction, a brief discussion of the reaction of trimethyl borate with Grignard reagents will be given. Methyl borate reacts with Grignard reagents at -60 C to give fair yields of boriuic acids. Above this temperature good to excellent yields of triarylor trialkylbora.nes are obtained. Workers in this field have postulated (2) that the reaction proceeds by a stepwise process with the first step being the slowest. It has been observed (12) that boronic esters, RB(OR) 2 , react much more readily than borate esters, B(OR) 3 , with Grignard reagents. Since trimetho:,:yboroxine is very similar in structure to methyl borate, the sluggish rate of the oecond step in the reaction sequence with Grignard reagents should be explained. The most obvious ex planation is that the boroxine ring is stable during the reaction and that while the first step in the sequence involves the displacement of a methoxy group by another organic group, the second step involves formation of an intermediate of the type R R ""-/ B" d o I I H,, CO-B B-OCH 3 :, " / 0

PAGE 57

so While thio intermediate accounts for most of the observed facto such as the sluggish second step and the very low yield of triarylborancs, some workers (36) consider it unlikely since water, alcohol. and other bases easily decompose the boroxine ring. They reason that the attacking Grignard reagent may also be regarded as a base. ostensibly with sufficient strength to rupture the ring--especially in the second step. However, the results of the following data show conclusively that the ring is not ruptured during the course of the reaction. Perhaps the most convincing evidence emerges from an examination of what would happen if the ring were to rupture during the process. In this event two possibilities exist. Either the open chain structure will rapidly rearrange to give methyl borate and boric acid anhydride according to the equation (MeOB0) 3 --+ B 2 0 3 + (Me0) 3 B or the Grignard reagent will react with the open chain structure with the same or greater rapidity than it reacts with methyl borate. Now the difference between these two alternates is this: If boric acid anhydride forms. each mole of boro~dne v1ill produce one 1nole of methyl borate; if the open chain is involved, each mole of boroxine is equivalent to three moles of methyl borate. In either case, because of the great reactivity of the borate, the yield of borinic acid would be only a small fraction of the yield actually observed. This is strong support for ring stability.

PAGE 58

Two other pieces of evidence add support to this notion. The firot is the result of an experiment in which the boroxine ring was forced to decompose in the presence of Grignard reagent. The second is the result of a study of the reaction of Grignard reagents with mix tures of methyl borate and boroxine. When the boroxine is added to the Grignard reagent in toluene at 100 C, the ring is ruptured thermally and the open chain form re• arranges rapidly to methyl borate and a precipitate of boric acid anhydride. The methyl borate then reacts in the usual way giving a very low yield (7. 5 percent) of borinic acid. This is much lower than obtained in any reaction involving the boroxine and shows that the ring is stable during reactionEJ at lower temperatures. In the study in volving mixtures of methyl borate and the boroxine, experiments with excess Grignard reagents and with excess boron compounds were conducted (Table XV). Yields of borinic acids in reactions with excess Grignard reagent were, within experimental error, those expected from the boroxine present. Apparently the methyl borate reacted almost completely to give triphenylborane, This again illustrates that the ring has remained intact. The more critical experimento in this study were those in which less than the stoichiometric qua::itities of Grignard reagent were used. In these experiments the yields were all low, but the higher yields of borinic acids were obtained for those solutions richer in methyl borate than in boroxine. These experiments

PAGE 59

52 indicate that the first step in the reaction of both methyl borate and the boroxine with Grignard reagents proceeds at about the same rate but that the second step is much faster for the borate. The preceding paragraphs summarize the role of the boroxine in the reaction, but they also raise several questions, One of these is: If a methoxy group of the boroxine is displaced in the first step. what is displaced in the second step? Some insight into this and other questions can be obtained by concentrating on the information available concerning formation of the carbon-boron bond. C. Formation of the Carbon-Boron Bond: The Case for Displacement on Magnesium Experimental data which provides information pertaining to bond formation in reactions usually comes from kinetic studies or from the effect of electron-donating or electron-withdrawing substituents on rates or yields. In this work the experiments have established that the first step in the reaction is displacement of a methoxy group of the boroxine at a rate comparable to that of a similar displacement in methyl borate. The nature of the bond formation process in this step might be described by one of three processeo. These are represented by the following equations: 1.) R----B----OMe 6 "-o R-B + OMe cl'o I I I I

PAGE 60

2.) ---+3. ) 53 R t1e /0" /Mg + 0-B "'-o... ,.. .,(1. oX " Mg-OMe +/ \ 0X-Mg + I 0-Me I R B o/"o I I ,,. Me 0" I / Mg--0--B ,,/ " R 0/ R-B 0\ 0X-Mg---0-Me I I I I R---B /" 0 0 I I R-B / " + XMg01v1e 0 0 I I In the first of these processes a carbanion attacks the boron atom displacing the metho:,dde ion. In the second process, the magnesium atom attacks an oxygen atom with the resulting displacement of the R groups a~ a carbanion. The third process consists of a simultaneous attack on the oxygen and boron atom, The result suggests but does not establish that the third process is responsible for bond formation. Evidence for this comes from studies of the effect of electrical factors (Table XII) and flteric .factors (Tables XIII and XIV) on the yields of products. From the studies of electrical factors, it has been observed that electron-donating substituents in tho Grignard

PAGE 61

54 reagent consistently decrease the yield of borinic acids. For example, the yield of borinic acid is reduced by a factor of 1. 5: 2 when methoxy, methyl, or phenyl groups arc substituted in the para position of phenylmagnesium bromide. Since these substituents should increase the strength of the carbon-boron bond formed in the reaction, it is un likely that the bond is formed by direct attack of a free carbanion (process 1) on the boron atom. If such attack occurred, the presence of electron-donating substitucnts would be expected to increase rather than decrease the yields. However, if the initial attack involved displacement on a magnesium atom (process 2 or 3), the presence of electron-donating groups would retard the process by strengthening the magnesium-carbon bond. This bond must be broken while the magnesium-oxygen bond is being formed. Some unusual steric factors observed in this reaction can be explained in terms of the magnesium displacement. The sensitivity of the reaction to steric factors appears to be contradictory when comparing the results of the study of steric factors in the reaction of aryl and alkyl Grignard reagents. While no decrease in yield is observed with one substituent on the phenyl ring as compared to the unsub stituted Grignard reagent, a decrease in yield of one-third is observed ' when the alkyl Grignard reagent has branching on the c:i(,_-carbon. This observation may be explained in terms of one of the three processes. All three processes would be sensitive to steric requirements, but

PAGE 62

55 proceos 3, the concerted mechanism, should be most sensitive to attack of bulky reactants, for in this case four ato.ms must attain an e,:act arrangement before reaction. More hindrance to formation of this strongly oriented four-membered ring transition state would be provided by the alkyl groups than the aromatic groups. This is because of the great rigidity of the carbon atoms in the aromatic ring as compared with the free-to-rotate carbon atoms in the aliphatic structure. The extreme sensitivity of the reaction to steric factors is evidenced by the comparison of the yields of borinic acids obtained, by reaction of the following Grignard reagents: n-butyl (76 percent). phenyl (62 percent), o-tolyl (59 percent). iso-propyl (32 percent), and !!,-butyl {21 percent). The .,!!-butyl Grignard reagent would show the least steric hindrance because of the unsubstituted al-carbon. The aryl Grignard reagentn, phenyl and 2,-tolyl, would be intermediate in the order because of the rigid structure, and the aliphatic Grignard reagents, ~-propyl and ~-butyl, because of free rotation of the bonds would show the most stcric hindrance. The discussion of the formation of the boron-carbon bond was mainly concerned with the attack of the first reagent. Although no evidence was obtained about the second attack, the similarity of the two displacemento would irnlicate the same concerted mechanism.

PAGE 63

CONCLUSION The reaction between trimethoxyboroxine and either aryl or alkyl Grignard reagents was found to proceed to give borinic acids. The yields ranged from 17 to 77 percent. Borinic acid anhydrides of di(n-propyl), di(iso-propyl), di(n-butyl), di(sec-butyl), di(n-amyl), di(sec-amyl), and dicyclohm:.:ylborinic acids were prepared while only the di(n-butyl)borinic anhydride had previously been reported. Di-Grignard reagents, prepared from 1, 4-dibromobutane and 1, 5-dibromopentane were reacted with trimetho:>:.:ybormdne in an attempt to prepare cyclic borinic acids. The results were incon clusive. The versatility of the reaction was also demonstrated in the reaction of phenyl Grignard reagents with triphenylboroxine to yield the diphenylborinic acid. From a study of the effects of temperature, ratio of trimethoxy .. boroxine to Gr\gnard reagents, solvent, steric factors, and compe tition between trimethoxyboroxine and trimethyl borate, a reasonable mechanism has been postulated. This mechaniom accounts for all facts observed in this study. 56

PAGE 64

57 The initial attack is probably Followed by R-MgX + R I B /'-.. 0 0 I I R-MgX + R R "'\./ B /'-.. Xlv"ig--0 0 I I + R I B /" 0 0 I I X-Mg----OCH I I 3 I I R----B /"'. 0 0 I I /R /R------B / /"' Mg----0 0 / I I X This same series of reactions can occur on the other boron atoms of the ring. On hydrolysis the ring is ruptured yielding the borinic acids.

PAGE 65

BIBLIOGRAPHY 1. A. Michaelis, M. Benrens, J. Robinson, and Vi. Geisler, Ber., 27 244 (1894); J. Chem. Soc., 66, 190 (1894). 2. R. L. Letsinger and I. Skoog, .:!• ~Chem. ~-, ]1, 2491 (1955). 3. J. R. Johnson and M. G. van Campen, J. Am. Chem. Soc., 60, 121 (1938). 4. L. H. Long and D. Dollimore, J. Chem. Soc., 1953, 3902. 5. R. L. Letsinger, I, Skoog, and M. Remes, J. ~Chem. ~1 76, 4047 (1954). 6. R. L. Letsinger and I. Skoog, J. ~Chem. Soc., 76, 4174 (1954). 7. R. L. Letsinger and H. Remes, I_. ~• Chem. Soc., 77, 2489 (1955). 8. B. 1v1. Mikhailov and V. A. Vaver, Doklay ~~"-• s. s. s. R., 1oz, 531 (1955);_. fl .• 5o 4813d (1956). 9. 'I'. P. Pov lock and W. T. Lippincott, J. Am. Chem. ~•, 80, 5409 (1958). 10, B. M. Mikhailov and V. A. Vaver, Izvest. ~~• S.S. s. R., ~Khim. Nauk., 1957, 989; c~. !::.:~• 4532 (1958). 11. B. M. },,1ikhailov and T. A. Shchegoleva, Izvest. Akad. S. S.S. R., Otdel Khim. Nauk., 1955, 1124-5; C. A., 50, 112335c (1956). 12. R. L. Letsinger and I. R. Nazy, .!!. Org. Chem,, 23, 914 (1958). 58

PAGE 66

59 13. B. M. Mikhailov and P. M. Aronovich, Izvest. Akad. Nauk. -s. S, S, R., Otdel. Khim. Nauk,, 1955, 859; : A., 2., 9320 (1956). 14. K, Torssell, Acta. Chem. Scand., J_, 239 (1955); C. A,, 49, 10214c (1955):-15. R, L. Letsinger and I. Skogg, l.• .:m• Chem. Soc., 11• 5176 (1955). 16. E. W. Abel, W, Gerrard, and M, F. Lappert, J, Chem, ~., 1957, 112. 17. W. Gerrard, M. F. Lappert, and R. Shaffermann, l•. Chem. ~-, 1957, 3828. 18. E. VI. Abel, W, Gerrard, and M. F, Lappert, J. Chem,~•• 1957, 3833., 19. M. F. Lappert, Chem. Revs., i.~ 959 (1956), 20. F. Bean and J, R. Johnson, l• ~• Chem. Soc,, 54, 4415 (1932). 21, P. R. Ogle, Jr,, Ph, D, Thesis, Michigan State University, 1955. 22. L. L, Quill, P, R. Ogle, Jr,, and W. T, Lippincott, Unpublished Report. 23. L, L. Quill, P, R. Ogle, Jr., L. G, Kallender, and W, T. Lippincott, "The Reaction of Trimethoxyboro:>dne With Aromatic Amines, 11 Abstracts 129th Meeting, American Chemical Society, Dallas, April, 1956. 24. W. Konig and W. Scharrnbeck, l_. eral~t. Chem., 128, 153 (1930); . ~-.• ~. 927 (1931). 25. N. N. Mel'nikovandM. S. Rokitskaya, l_. Gen. Chem, (U.S.S,R.):1 ,!!, 1768 (1938); C. A., _g, 4969 (1939). 2.6. N. N. Mel 1 nikov, J. Gen. Che.m. (U.S. S. R. ), 6, 636 (1936); : f;.:, 30, 5571 (l 93~

PAGE 67

60 27. B. M. Mikhailov and F. B. Tutor ska ya, Izvest. Akad. Nauk. s. s. s. R. ~l. ~. ~,;:., 1959, 1127; c. A.• El• 6990g (1959). 28. H. I. Schlesinger, L. Horvitz, and A. B. Berg, !,_. l::!!: Chem. Soc., 58, 407 (1936). 29. A. L. Borisov, Izveot. Akad. Nauk. S.S. S. R., Otdel. Khim. Nauk., 1951, -02; C. ~. 46, 2995d (1952). 30. V. R. Nev, Ber., 88, 1761 (1955). 31. J. Goubeau and Keller, anorg. Chem., 267, l (1951). 32. C. Jr. H. Allen and S. Converse, "Organic Syntheses, 11 Coll. Vol. I, John Wiley and Sons, Inc., New York, 1943, p. 226. 33. L. J. Fieser,, "Experiments in Organic Chemistry, 11 Third Edition, D. C. Heath and Co., Boston, 1955, p. 266. 34. H. Gilman, P. D. "Wilkinson, W. P. Fishel, and C. H. Meyer, J. ~. Chem. Soc., 45, 150 (1923). 35. H. R. Snyder, I. A. Kuck, and J. R. Johnson, J. Am. Chem. -Soc., 60, 105 (1938). 36. J. M. Davidson and C. M. French, J. Chem. Soc., 1960, 191.

PAGE 68

BIOGRAPHICAL SKETCH Thom , as Paul Povlock was born in Salamanca, New York, on January 22, 1934. Upon graduating from Salamanca High School in 1951, he attended Kent State University, Kent, Ohio, and he was awarded the degree of Bachelo1 of Science in 1955. He was admitted to the Graduate School of Michigan State University, East Lansing, Michigan, and received the Master of Science degree in 1957. In. 1957 he was admitted to the Graduate School of the Univer sity of Florida as a graduate assistant. In 1959 he was appointed as a graduate teaching assistant. 61

PAGE 69

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was oubmitted to the Dean of the College of Arts and Sciences and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 13, 1960 Supervisor CW:-T-~~---:;;:---e;;tl