SOLVOLYSIS KINETICS OF
PYRIDINEBORANES
By
ERNEST RODMAN BIRNBAUM
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
January, 1961
Tic author tiches to hia:'.: his fricnic and nscociatoc for many
helpful rwu:;cotions and extreely stinu3latin; discussions,
To his faculty:. co..itoo, who have gone far beyond the call
of dut: to C:epc31t.o the writin3 of this dissertation, the author
er:roseco his sincere appreciation*
For efforts in obtaining equi;:icnit a .C dimicalo, for a firn,
but patient gtuidanco, and for an interest in the author's wellbeing,
the author is dccp;l:, grateful to )Dr. tschklcitsch, his research
director.
The author would als o like to thank the Research Corporation
for its generous financial support.
AC11CWLn^irrTIT
TABLE OF COITINTS
Page
Chapter I
Introduction .* . . . . . .
Chapter II
Experiuental2.. .. *
Chapter III
ReCul.ts * .*. * *. * *
Chapter IV
Discussion . * *. * * *
Statistical Appendix . . . . . * * *
Reforences .* * * a * * * *
iii
LIST OF TABLES
Table Page
1. Analysis of purified pyridineborane . . 5 5
2. Amines w * * * * 5
3, Pyridineborane rate constant data . . . . 31
4, 11ixedsolvent experiments, 50,00C. . . . . . 32
5. Surmary of kinetic data for 4picolineborane . 40
6, Su a ru of kinetic data for 3picolineborane . 43
7. S'u:ar. of kinetic data for 2picolineborane .. 6
8. Sumary of kinetic data for 2othylpyridineborane 4 . 19
9. Sunrmary of kinetic data for 2,41utidineborane . 52
10. Sunmary of kinetic data for 2,6lutidineoborane . .
11. Kinetic data for 2,h,6collidineborane . 5. . 7
12, Fit of rate constants to Arrhenius equation,
log k A+ B/T (sece,)1 . . . 58
13. Physicochemical data for the pyridines . . 77
!h, Test for homogeneity of several least squaes slopes . 88
LIST OF FIGURES
Figure
1.
2.
3.
4.
5.
7.
8,.
9.
10.
lla.
lib.
lc,.
12,
13.
3i.
15.
16.
17.
18.
19.
20.
es
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Fyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
Pyridineborane
* 0 . 0 9 *
* 09. 0 99900. 9 . 9
00. 090*99 9 99.
*,
0999
@9900
* 99*
43.120C.
43.120C.
4.5490C.
5.4L9C .
L7.6300.
47.630C
47.630C.
50000.
50.000C.
50.000c.
506000C.,
50.OCC.
0.000oo.
5o.oooc.
$o.Oooc.
S6
Behavior
Behavior
Behavior
Pyridineborano, 52.430C. . .
Pyridineborane, 52.30C . .
Pyridineborane, 52.330. .
Pyridineborane, 54$260C. . .
Pyridineborane, 54.260C. . .
Arrhenius Plot for pyridineborane
Pyridineborane 5~.300C. Iagnetica:
Pyridineborane h5424C. Added pyr:
Pyridineborane 54.24OC. Added pyr:
over
over
over
* 0 a *
several h
several hi
several h
* . *
alflives
alflives
alflives
ly stirred . . .
idine, 0,26M
idine, 0.421I
* . .9
* S 9 5 9
* 9
9
9.. SO 9* 9 *999*
Page
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
34
35
36
37
* 09
9..
LIST OF FIOURES(Continued)
Page
Pyridineborane 50.00C. Mixed solvent 30% water 70%
propanol by weight *. *. *. . . 38
Arrhenius plot for
Arrhenius plot for
Arrhenius plot for
Arrhenius plot for
Arrhenius plot for
Arrhenius plot for
Rate dependence on
4picolineborane . .
3picolineborane . .
2picolineborane . .
2ethylpyridineborane .
2,4lutidineborane .
2,6lutidineborane .
alcohol concentration .
29, Kirkwood plot for propanolwater mixtures at
30. Steric requirement of transition state relat
of methanesulfonic acid . . . .
31. Steric requirement of transition state relate
of a borane group .. . .
* 4 5 *
ive to
Ive to
...
that
* *t
that
...
32. Steric requirement of transition state relative to that
of boron trifluoride . . . *. . . .
Figures
21.
22.
23.
24.
25.
26.
27.
28.
CHAPTER I
INTRODUCTION
The earliest known addition compound between a tertiary amine
and a borane group, trimethylanineborane, was synthesized in 1937 by
Burg and Schlesingera. Originally prepared by the base displacement
reaction between borine carbonyl and trimethylamine, the amineborane
was also formed when trimethylamine and diborane were brought together.
Melting at 940C., trimethylamineborane is a white, crystalline solid
which can be heated for severalhours at 125C. without a detectable
change in its physical properties. Even though a number of physico
chemical investigations, including electron diffraction,2 xray
diffraction,3 Raman and infrared spectra,4,5j6 dipole moment measure
ments,7 heats of reaction,8 BU magnetic resonance spectra,9 and use
as a stereospecific reducing agent,10 have been made on tritaethylamine
borane, relatively little knowledge has been obtained about the reac
tivity of its boronnitrogen or boronhydrogen bonds.
Five years after the discovery of trimethylamineborane, Brown,
Schlesinger, and Cardonll reacted pyridine with diborane to produce
pyridineborane. Utilizing the analogous reaction in nitrobenzene
solutions of pyridine and various alkylsubstituted pyridines, Brown
and Domash12 in 1956 prepared pyridineborane and seven additional
alkylsubstituted pyridineboranes, and measured the heats of reaction
with diborane calorimetrically,
The pyridineboranes are either colorless liquids or white,
crystalline solids. In contrast to trimethylamineborane, however,
they are not stable to heat, but readily evolve hydrogen at tempera
tures below 1000C. to form red resins. The few investigations of
pyridineborane to be found in the literature concern its cryoscopic
molecular weight and differential thermal analysis,13 its dipole
moment,7 its infrared spectra,5'6 its synthesis and typical reac
tions,1 and its B11 magnetic resonance spectra.9 No further ex
perimental work has been reported for the alkylsubstituted pyridine
boranes.
Since the heats of formation were the only data common to all
of the tertiary amineboranes, and the sole data besides melting
points for the substituted pyridineboranes, it was proposed to
undertake a series of quantitative studies relating the reactivities
of these compounds to their structures. The results of the first
set of experiments in the series, which pertain to trimethylamine
borane, have already been published, Comprising the second part
of the series, the experiments to be described in this dissertation
will be devoted to a kinetic investigation of the solvolysis reaction
of the pyridineboranes in npropyl alcohol
C9Hgmrl3 + 3C3HH > 3H2 + CAsN + B(oC3H7)3
The purpose of these experiments is to determine if there is a
correlation between the mechanism and rates of solvolysis of the
pyridineboranes, and their structures and heats of formation.
CHAP'ITFR II
FYXPJ1RTEITTAL
Chemicals .A sample of Callery Chemical Company's pyridine
borane used for preliminary experiments was a clear, colorless liquid.16
Purification was easily accomplished by dissolving the amineborane
in ether, extracting with dilute sodium hydroxide, discarding the
aqueous phase, and removing the ether from the ethereal phase by
vacuum distillation, The purified material was analyzed for active
hydrogen by a slight modification of the iodate method.17 Because
of the limited miscibility of pyridineborane in water, it was
necessary to dissolve the weighed sample of amineborane in a few ml,
of npropanol, and then add the excess standard potassium iodato
immediately to this solution. No difference in iodate titer was
found when standard potassium iodato solution containing 10 ml, of
npropanol was converted to iodine with excess acid and potassium
iodide, neutralized with sodium bicarbonate, and titrated to the
starch end point with standard sodium arsenite.
A boron determination was also made on the purified pyridine
borane. To the methanol solution of the amineborane, concentrated
hydrochloric acid was added crop:wio until there was no further
evolution of hydrogen. An additional 5 ml. of concentrated acid and
10 grams of anhydrous calcium chloride were added, and the resultant
solution heated to distill methyl borate into a delivery tube immersed
in a beaker of water. After neutralization of the hydrochloric acid
3
which distilled along with the methyl borate, the determination was
completed by titration of the boric acidmannitol complex with
standard base. Table 1 contained the results of these analyses.
When it became necessary to obtain additional pyridineborane,
a pound was purchased from the Callery Chemical Company. This
material turned out to be a somewhat turbid, distinctly yellow liquid.
Application of the above purification procedure in conjunction with a
filtration through alumina or magnesia did not produce a stable, color
less product. The method of purification finally adopted consisted of
putting the pyridineborane over calcium hydride, slowly heating it to
700C. with continual pumping in a vacuum system, and vacuum distil
ling it at 70800C. onto a cold finger. Samples purified in this
manner were clear, colorless liquids which typically gave active
hydrogen analyses within 97.3983% of the theoretical values. No
difference in kinetic behavior was observed between the several
purified batches of pyridineborane.
All of the amines used in this study were purified by 812
hours of stirring over calcium hydride, and distillation from the
calcium hydride on a Todd column at a 0:1t reflux ration. Table 2
lists the pertinent data. Synthesis of the alkylsubstituted pyridine
boranes was accomplished by bubbling diborane, generated by the method
of Brown and Tierney,18 from a delivery tube immersed in mercury into
an excess of the respective free amines. 'Host of the excess amine
was removed from the solid pyridineboranes by decantation; the re
mainder was taken off by continual pumping in a vacuum system for
several hours. The liquid pyridineboranes were subjected to high
TABLE 1
ANALYSIS OF PURIFIED PYRIDINEBORANE
Weight of pyridineborane Calc. Value Exptl., Value
0.1149 gram 3.71 mr.mole H" 3.67 m.mole IH"
0.1332 gram 4.30 m.mole H" 4.26 m.mole H"
0.2882 gram 3.10 m.mole B 3.06 m.mole B
0.2636 gram 2.84 rm.mole B 2.81 m.mole B
TABLE 2
AIMIES
Amine Manufacturer Distilling range oC.
4picolino Matheson, Coleman, &. Bell 144.0144.7
3picoline Matheson, Coleman, &* Bell 143.5143.8
2picoline Matheson, Coleman, &. Bell 128.0128.0
2ethylpyridine Reilly Tar &, Chemical Corp. 148.8148.8
4ethylpyridine Reilly Tar &. Chemical Corp. 167.0167.3
2,hlutidine Reilly Tar &. Chemical Corp. 156.9157.1
2,6lutidine Reilly Tar &. Chemical Corp. 142.0143.0
2,4,6collidine Reilly Tar &. Chemical Crop. 170.0170.0
vacuum overnight, or until no further reduction in volume was observed.
4Picolineborane was recrystallized from nnonane as a white,
crystalline solid melting at 7273oC. It was analyzed by the
Schwartzkopf licroanalytical Laboratory with the following results:
% C calc*, 67.37; found, 67.20
% N calc., 13,10; found, 13.32
% B calc., 10.12; found, 10.31
% H, calc., 9.42; found, 9.66
A liquid, 3picolineborane was purified by vacuum distil
lation onto a cold finger. Active hydrogen analyses gave 98.3 and
97.1% of the theoretical values.
Repeated recrystallizations of 2picolineborane from nnonane,
1:1 benzene2,3,5trirethylhexane, or benzenenhexane mixtures failed
to yield a product melting above 4608C., (literature value 50510C.)
or analyzing higher than 95.5% of the calculated active hydrogen.
Following a recrystallization from nnonane, 2 thylpyridine
borane melted at 49.851.20C., and gave an active hydrogen content
97.4 and 97.3'" of the theoretical values.
4Ethylpyridineborane, a liquid at room temperature, was
not further purified, characterized, or used for Kiinetic studies.
2,4Lutidineborane is a white, crystalline solid. Both the
material synthesized as above and a sample from the Gallery Chemical
Company were recrystallized from nnonane. The melting points were
76.577.70C. for the fonrer, and 76.577.5C. for the latter. The
Callcry product was analyzed by the Schwarzkopf Microanalytical
Laboratory as follows:
% C calc., 69.49; found, 69.69
% H calc., 10.0 ; found, 9.98
% N calc., 11.58; found, 11.17
In addition to repeated recrystallizations, a vacuum sublima
tion was tried to purify 2,6lutidineborane. The highest purity
obtained was 96% of the theoretical active hydrogen content.
A white, crystalline solid from nnonane, 2,h,6collidine
borane melted at 99.2,100,30C. The Schwarzkopf Microanalytical
Laboratory also analyzed this material:
% C calc., 71,16; found, 71.32
% H calc., 10.45; found, 10.39
% N calc., 10.37; found, 10.12
Fisher Scientific Company's rcagent grade normal propyl
alcohol was purified by refluxing it over anhydrous barium oxide for
812 hours, and distilling it on a Todd column at a reflux ratio of
50Sl. About the first and last 10% would be discarded, while the
middle fraction, distilling between 97.0 and 90.00C, would be collected.
Purified grade pdioxane uas also obtained from the Fisher
Scientific Company. It was refluxed over calcium hydride for several
days and distilled from the calcium hydride just before use. Distil
ling at 1010C., this material gave a negative test for peroxide upon
addition of potassium iodide, dilute hydrochloric acid, and starch.
Apparatus.An E. H. Sargent &. Co. "Thermonitor" control and
heatercirculator were used with a Raytheon voltageregulating trans
former to provide the constant temperature bath required for kinetic
studies. In addition, the rlass tank supplied with the Sar;,nt unit
was fitted with an aluminum top and surrounded with a 23 inch layer
of venniculite. As measured by a Beckmann thernometor, the bath
could be maintained to within t 0.0050C at any temperature between
30 and 6500. For temperatures below 30C., the Sargent "Thermonitor"
unit was used in conjunction with a temperaturecontrolled, refrig
erated water bath. In this case the Sargent glass tank containing
the "Thermonitor" heatercirculator was placed directly into the
reservoir of the cooling bath, and the thermostat of the latter was
set a few degrees below the desired te::iperature. The "Thermonitor"
control was then adjusted to bring the temperature inside the Sargent
bath up to, and maintained at, the desired value. The constancy of
this arrangement was t 0.030C. maximum at 1000., and less than t 0.01C.
for temperatures between 18 and 300C.
Procedure.All of the kinetic experiments to be reported in
this dissertation involved determining the "reducing strength," i.e.,
the number of milliequivalents of reducible species per 5 ml. sample,
of a reaction mixture as a function of time by means of the iodate
method.17 For convenience the "reducing strength" will henceforth be
designated by the abbreviation RS. The detailed kinetic procedure, in
order of operation, consisted of the following: (1) direct weighing
of the anineborane into a 100 r.l volumetric flask used as a reac
tion vessel; (2) diluting to the mark with npropanol at room tempera
ture; (3) through mixing of the solution and discarding of about 10 ml.
so that there would be no liquid in the neck of the flask; (4) immersing
the volumetric flask in the constant temperature bath to within about
an inch of the standard taper and waiting a minimum of 30 minutes for
therntal equilibrium to be established; (5) pipetting 5$ ml of the
reaction mixture directly into an excess of standard potassium iodate;
(6) adding solid reagent grade potassium iodide to the iodate solu
tion; (7) acidifying the iodateiodide solution with an excess of 2 I
sulfuric or hydrochloric acid; (C) neutralizing the excess acid with
excess reagent grade sodium bicarbonate; and (9) titrating the lib
erated iodine to the starch end point with standard sodium arsenite.
The amount of 0,01 IT standard potassium iodate solution taken
for step (5), as measured by a 50 ml. burette, was regulated to give
an approximately constant back titration. Tb minimize experimental
error, the excess iodateand hence the arsenite titrationwas kept
as small as possible consistent with obtaining a quantitative oxida
tion of the kinetic sample. Typically the arsenite back titration
amounted to 1 ml. of 0.025 LN solution. Pipetting presented a problem
at high reaction rates due to the copious evolution of hydrogen bubbles.
It was found that shaking the volumetric flask prior to taking a
sample eliminated most of this trouble. A further precaution taken
in connection with pipetting was to rinse the pipette with methanol
and dry it on an aspirator after each sampling.
lTmperaturo differences were measured by using three Beckmann
thermometers which had been set so that the bottom range of one over
lapped with the top range of another. Once a temperatare on a O.10,
thermometer was taken as an arbitrary standard, temperature intervals ,
of 12140C, could be determined to within a few thousandths of one
dcgrce centigrade. Time intervals ranging from 15 minutes to several
10
hours were obtained to within $ seconds by means of a stopwatch.
With the exceptions of 2,6lutidineborane and 2,h,6collidine
borane, at least tLo kinetic "runs" of widely differing concentration
were made at each temperature, giving an estimate of the experimental
precision. Tnose concentrations were usually between 0.06 and
0.1 N.
CHAP TER III
RESULT
Very similar kinetic results were obtained for all of the
pyridineboranes. Those for the alkylsubstituted pyridineboranes
will be given in tabular form; while those for pyridineborane itself
will be presented graphically as a representative example. In either
treatment the time scale for many of the experiments has been shifted
so that zero time refers to the time the first sample was taken, and
not to the time the experiment was begun. This change will not affect
the interpretation of the data, because as will be considered subse
quently, the rate constant for a firstorder reaction is independent
of both initial time and concentration.
Since propyl alcohol was used as the solvent for these studies,
its concentration rained essentially constant during the course of
a kinetic experiment, making the solvolysis reaction ajpiarcntly zero
order in alcohol. Graphs of log (RS), log (milliequivalents of
reducible species/5 ml. sample), against time will be utilized to
present the results of the observed firstorder reaction in amine
borane, These results for pyridineborane are shown in figures 1 to
10, 11a, and 12 to 16. The size of the circles representing the
data is an estimate of the relative experimental error. Included on
each graph is the least squares slope, as determined from those points
which fall on the line.
12
Lea
0.
.600
60
0.560
+
CMJ
CM
rt
.520
69 309
Time (Min.)
Figure 1. Pyridineborane, '3.120C.
st Squares Slope
000218 (min.)l
I9
549
.820.
Least Squares Slope
0.000211 (min.)
.7800
.O
* .7)40
I I I I
131 251 371 491
Time (Min.)
Figure 2. Pyridineborane, h3.120C.
Least Squares Slope 30
0.000292 (min.) 300
.240
180 I
120
60
.600 .580 .560 .54o
1.222 + log (RS)
Figure 3. Pyridineborane, 45.90C.
Least Squares Slope
0.000290 (min.)'
.800
.780
+
CM
CM
S.760
.7h0 
60 120 180 240 300
Time (Min.)
Figure 4. Pyridineborane, 454.90C.
Least Squares Slo e
0.000378 (min.)"
.680
.660
0
+
c .6o0 o
.620
120 180 2h0 300
Time (Min.)
Figure 5. Pyridineborane, 47.630C.
Least Squares Slope
0.000398 (min.) 
.660
.6ho
.620
.600
0o 100 160 220
Time (Miin.)
Figure 6. Pyridineborane, 47.630C.
Least Squares Slope
0.000353 (min.)1
.760
0o
+
0H
.700
16b 320b 80
Time (Min.)
Figure 7. Pyridineborane, 47.630C.
.82C
Least Squares Slope
0.000491 (min.)1
Time (Hin.)
Pyridineborane, 50.C00C.
.300.
. 280
0
C"
+
c\J
j .260
r
.240.
Figure 8.
Least Squares Slope
0.000o93 (min.)1
180
Pyridineborane,
.980
.960
.900
Time (Min.)
50.000.
Figure 9.
Least Squares Slope
0.0oo093 (min.)
Figure 10.
60 120
Time (Min.)
Pyridineborane, 50.00C.
.640
o .620
rI
+
C\1
(14
(04
*
r6
.600
Least Squares Slope 0.000490 (min.)1
Sun of squares of deviations from Least
Squares Line 0.000018.
.48o0
0 .h60
CM
C\M
.o44
S10hho
51 100 150
Figure 11a.
Time (Min.)
Pyridineborane, 50.000C. Behavior over several halflives.
Least Squares Slope
0.000489 (min.)
.300
. 200
C%
S.100
Jo 5o0 700 o9b
Time (Min.)
Figure 11b.
Pyridineborane, 50.00C.
Behavior over several halflives.
Least Squares Slope
0.000530 (min.)
Figure 11c. Pyridineborane,
50.00C. Behavior over several
halflives.
1600
Time (Min.)
100
2000
0
.650
CM
CM
(M
CM
oJ
oJ
1200
1400
1 I _4I I I
Least Squares Slope
0.000631 (min.)"
Pyridineborane, 52.h30C.
Time (Min.)
0.
(\j
cuj
cM
Hu
Figure 12.
Least Squares Slope
0.000632 (min.)
Time (Min.)
Figure 13. Pyridineborane, 52.h30C.
.640
0
Cr
CM
C\M
, .620.
.600.
Least Squares Slope
0.000617 (min.)1
Figure 14.
60 90
Time (Min.)
Pyridineborane, 52.430C.
.94o
, .900
bf
0
r1
+
C\J
cli
CM
8 .880
.860
180
Least Squares Slope
0.000788 (min.)
Figure 15.
Time (Min.)
Pyridineborane, 5h.26C.
.560
S.520
0
.480
.h4o
Least Squares Sloe
0.C00770 (min.)
120
Figure 16. Pyridineborane, 5h.260C.
_ II I I _ _
180 2h6 30b 360
Time (iin.)
.800.
.7Q0
S.680
b0
0
O
+
C(M
CM
.620.
Pseudofirstorder rate constants for pyridineborane were
calculated from the average slope of two or more experiments with
different initial concentrations at each temperature. A swumary of
this rate constant data is given in table 3. Figure 17 is the Arrhenius
activation energy plot, i.e., log k against l/T, for pyridineborane.
From a least square analysis (see statistical appendix), the equation
of the Arrhenius plot was detcriained ast
loc k 11.05 S100/T (sec.)"
with a standard error of 1 0.10 for the intercept and t 31 for the
slope, iMultiplication of the slope by 2.303 R yielded an experi.
mental activation energy of 23.34 t 0.1l kcal./mole.
The results of a lacutica1y stirred pyridineborane "run"
at 54.300C, are given in figure 1G. Showing the continued linearity
of the log (RS) vs. tine plots, figures 11b and 11c extend the o50C,
experiment of figure 11a to almost 3" halflives. For the combined
data of these three figures, the overall rate constant was found to
be 0.00117 (min.)1. Added pyridine 0.26 and 0.h2 M, figures 19 and 20,
yielded respective rate constants of 0.00193 and 0.00188 (min.)"l at
54.20c.
Table h .ivcs log (PS) and time values together with the least
squares slopes for a number of pyridineborane experiments using
propanolwater mixtures as solvents, and for one experiment using a 50% by
volume pdioxanepropanol .mixture, all at 500.00C. As typical of the
results from these mixedsolvent experiments, figure 21 contains a
plot of the data for the 307 water70% propanol by veiht solvent
mixture,
TABLE 3
TFPYITIE:3n'?A;.E rPJT CONSTANT DATA
Temp* 0C. Avg. Slope k = 2.303 Slope log k (min.)1
43.12 0.000216 o0.oo497 3.304
45.*9 0.000291 0.000670 3.174
47.63 0.000376 o0.oo866 3.063
50.00 0.000492 0.00113 2.947
52.43 0.000627 0.00144 2.842
54.26 0.000779 0.00179 2.747
TABLE 4
IilXFESOLVFJiT EXPERIME1 S 50.0000C,
10% water 20% water 3c0 water 4 water
90% PrOH by wt. 80C PrOH by wt. 70% PrOH by wt. 60% PrOl by wt.
1 + log (RS) Tine (Hr.)
0.596 0
0.513 5.5
0.470 8
0.434 10.5
0.387 13.5
0.339 16.5
0.284 20
Least Squares Slope
0.000260 (min.)1
1 + log (RS) Time (Hr.)
0.618 0
0.603 1.5
0.524 8
0.494 10.5
0.458 13.5
0.423 16.5
0.383 20
Least Squares Slope
0.000198 (min.)"
1 + log (RS) Time (Hr.)
0.683 0
0.670 1.5
0.635 5.5
0.610 8
0.589 10.5
o.561 13.5
0.533 16.5
Least Squares Slope
0.000152 (min.)1
1 + log (RS) Time (Hr.)
0.693 0
0.665. 3
0.639 6
0.614 9
0.587 12
0.562 15
0.536 18
Least SquaresSlope
o.oool45 (nin.)"l
~~~
TABLE 4Continued
50, water $55.% water* 60Z water 50% pdioxane
50% PrOH by wt. 44.6% PrOH by wt. 40 % PrOH by ut. 50% Pr0H by volume
1 + log (PS) Time (Hir.) 1.222 + log (RS) Time (Min.)
0.633 0 0.658 4,
0.610 3 0.654 75
0.588 6 0.651 120
0,568 9 0.629 285
0.547 12 0.612 435
0.524 15 0.606 495
0.502 18 0.597 570
Least Squares Slope 0.591 630
0.000120 (min.)"1 Least Squares Slope
0.000116 (min.)"'
1 + log (RS) Time (ir.) 1.222 + log (RS) Time (Min.)
0.656 0 0,572 37
0,639 3 0.546 53
0.620 6 0.523 70
0.601 9 0.499 85
0.582 12 0.479 100
o.566 15 0.453 115
0.542 18 0.43 130
Least Squares Slope O.412 143
0,0001.o (min.' 0.596 156
0.370 172
Least Squares Slope
0.001U9 (min.)1
*Calculated for a 50% by volume solution at 2000.
2.741Q 34
Least Square
2.940
I
.r
0
3.14o
3.340
305 310
1/T x 105
is Slope 5100
Figure 17. Arrhenius Plot for Pyridineborane.
Least Squares Slope
0.00788 (min.)"
180
120
60
I I 1
.550 .520 .490 .460 .430
1 + log (RS)
Figure 18. Pyridineborane, 54.300C. Magnetically stirred.
Least Squares Slope
0.000837 (min.)
Figure 19. Pyridineborane,
300 450
Time (Min.)
54.240C. Added Pyridine 0.26M
0h
45
CM
CM
C(
20
r2
.250
Least Squares Slope
0.000818 (min.) 
Figure 20.
1u 300 450
Time (Min.)
Pyridineborane, 5$.24C. Added Pyridine 0.42M
.650o
0
CMj
CMj
CM
.2501
Least Squares Slope
0.000152 (min.)
Pyridineborane, 50.000C.
Time (Min.)
Mixed Solvent 30% Water, 70% Propanol by weight.
H..
Figure 21.
39
The kinetic results for the alkylsubstituted pyridine
boranes are given in tables 5 to 11 and figures 22 to 27* Table 12
summarizes the derived thermodynamic quantities (see discussion ) for
all of the pyridineboranes,
TABLE 5
SUI2IAY OF KINETIC TDTA FOR 4PICOLINKEDORAJE
58.2o00. 58.200C. 60.100C. 60.10C,
1 + log (iRS) Time (in.) 1 + log (RS Time (Min.) 1 + log (RS) Time (Min.) 1 + log (RS) Tine (Min.)
0.376 0 0.131 0 0.456 0 0.349 0
0.349 60 0.099 60 0.422 60 0.305 60
0.321 120 0.072 120 0.382 120 0.268 120
0.274 210 0.028 210 0.343 180 0.233 180
0.246 270 0.015 270* 0.304 240 0,189 240
0.215 330 0.039 330 0.267 300 0.151 300
0.184 390 0.071 390* 0.227 360 0.112 360
Least Squares Slope Least Squares Slope Least Squares Slope Least Squares Slope
0.000494 (rain.)"' 0.000509 (rin.)1 0.000640 (min.)"1 0.000654 (min.)~1
*Not included in least squares calculation.
TABLE 5Continued
62.00c. 62.00oc. 62.00oc. 64.90c.
1 + log (RS) Time (in.)
o.455 0
0.423 45
0.386 90
0.350 135
0.306 180
0.256 240
0.208 300
Least Squares Slope
0.000837 (min.)"
1 + log (RS) Time (MIn.)
0.277 0
0.243 45
0,205 90
0.180 120
0,151 150
0.125 180
0.100 211
Least Squares Slope
0.00oo081 (min.)1
1 + log (RS) Timne (MiLn.)
0.581 0
0.558 30
0.533 60
0.507 90
0.483 120
0.456 150
o.434 180
Least Squares Slope
0.000827 (min.)"1
1 + log (RS) Time (Uin.)
0.303 0
0.108 180
0.077 210
0.043 240
0.008 270
0.037* 30oo*
o.o37" 330*
.0.067* 330*
0.109 360
Least Squares Slope
0.00109 (min.)1
*Not included in least squares calculation.

TABLE 5Continued
64.94tC. 66.43oC. 66.643C.
1 + log (RS) Time (Hin.)
0.526 0
0.493 30
0.461 60
0.430 90
0.397 120
0.364 150
0.332 180
Least Squares Slope
0.00108 (min.)"'
1 + log (RS)Time (Min.)
0.504 o
0.471 30
0.433 60
0.393 90
0.356 120
0.277 180
0,213 225
Least Squares Slope
0.00130 (rin,)"1
1 + log (PS) Time (Min.)
0.357 0
0.320 30
0.282 60
0.240 90
0.200 120
0.158 150
0.123 180
Least Squares Slope
0.00131 (in.)1
 ~
TABLE 6
SLUP2ARY OF KINEEIC DATA FOR 3PICOLINEBORANE
52.70ca 52.70c.a 527.70oca 55.2000.
1 + log (RS) Time (Min.) 1 + log (RS) Time (lin.) 1 + log (RS) Time (Min.) 1 + log (RS) Time (Min.)
0.613 0 0.425 0 0.513b 0b 0.73 0
0.589 60 o.4oo 60 0.492 60 o0.36 75
0.564 120 0.362 150 0.460 150 0.ol5 135
0.537 180 0.338 210 0.434 210 0.374 195
Least Squares Slope 0.312 270 0.390b 270b 0.338 225
0.00022 (min.)"1 0.289 330 0.383 330 0.302 315
0.265 390 0.358 390 0.273 375
Least Squares Slope 0.338 l50 Least Squares Slope
0.00O11 (rmin.)" Least Squares Slope .0o000o 2 (min,)
o.oOo5 (min.)"
bNot included in least squares calculation.
aA weighted average was used to calculate the average slope for this temperature.
TABLE 6Continued
55.20C. 57.470,C 57.47 C. 59.97C.
1 + log (RS) Time (Min.) 1
0.376 0
0.333 75
0.306 135
0.276 195
0.240 255
0.206 315
0.178 375
Least Squares Slope
o.c0o530 (min.)"1
+ log (RS) Time ( in.)
0.532 0
0.501 45
0.465 90
0.434 135
0.405 180
0.359 240
0.332 275
Least Squares Slope
0.000723 (rain.)1
1 + log (RS) Time (MIin.)
o.449 0
o.417 45
0.385 90
0.353 135
0.314* 180*
0.281 240
0.248 275
Least Squares Slope
0.000720 (min.)"
1 + log (RS) Time (Min.)
0.517 0
0.484 30
0.443 75
0.401 120
0.358 165
0.321 210
0.278 255
Least Squares Slope
0.000928 (min.)"1
,.ot included.in least squares calculation.
TABLE 6Continued
59.97C. 62.070c. 62.07oC.
1 + log (RS) hime (Min.)
0.o38 0
O.409 30
0.368 75
0.325 120
0.283 165
0.239 210
0.200 255
Least Squares Slope
0.c00937 (rin.)1
1 + log (RS) Time (;in.)
0.560 0
0,522 30
0.685 60
0.455 90
0.419 120
0.383 150
0.323* 180*
Least Squares Slope
0.00117 (min.).1
1 + log (RS)Time (HMin.)
0.462 0
C0.27 30
0.388 60
0.359 90
0.321 120
0.288 150
0.254 180
Least Squares Slope
o.o0015 (min.)1
*Not included in least squares calculation.

TABLE 7
SU1IIAEY OF KINETIC DATA FOR 2PICOLINEBORANE
39.6700. 39.670C. b1.800C. h1,800o ,
1 + log (RS) Time (Min.) 1
0.554 o
0.517 105
o.687 180
o.459 255
0.431 330
0.4065 405
0.376 80
Least Squares Slope
0.000371 (min.)1
+ log (RS) Time (Min.)
0.452 0
0.410 105
0.385 180
0,361 255
0.333 330
0.306 4o5
0.270 495
Least Squares Slope
0.000361 (min.)'
1 + log (RS) Time (Min.)
0.572 0
0,536 75
o.516 120
c0.87 180
0.460 240
0.429 300
0.395 375
Least Squares Slope
0.000473 (rmin.)"
1 + log (RS) Time (Min.)
0.470 o
0.433 75
0,o13 120
0.382 180
0.354 240
0.319 315
0.289 375
Least Squares Slope
0.000481 (min,)"1
TABLE 7Continued
43.950. 43.950. 46.1ic. 46.h44c.
1 + log (RS) Time (iIin.)
0.577 0
0.526 45
0.499 90
o.474 135
o.446 180
0.413 225
0.388 270
Least Squares Slope
0.000624 (min.)Y1
1 + log (RS) Time (1ian.) 1
0.151 0
o0.424 15
0.391* 90*
0.376 120
0.362 150
0.344 180
0.311 225
0.286 270
0.25! 330
Least Squares Slope
0.000609 (min.)"1
+ log (R3) Time (Min.) 1
0.569 0
0.533 45
0.457 135
0.421 180
0.386 225
0.347 270
0.311 315
Least Squares Slope
0.000820 (rnin.)
+ log (RS) Time (Min.)
0.474 0
0.438 45
0.l01 90
0.363 135
0.329 180
0.293 225
0.256 270
Least Squares Slope
o.0oo806 (min.)
*Not included in least squares calculation.
TA LLE 7Continued
 ,
1 + log (RS) Time (iin.)
0,598 0
o.566 30
0.536 60
0.5o0 90
.t470 120
0.437 150
oto6 100
Least Squares Slope
0.00107 (in.)"
1 + log (RS) Time (;in.)
0.471 0
0.439 30
o.h4o 6o
0.373 90
0.337 120
0.308 150
0.274 180
Least Squares Slope
0.00110 (n.)"
4h0.8h0c.
h8.8hc.
 I'  ~ ~
TABLE 8
SU L:::A OF :I::EzIC DATA FOR 2ETYLPYrrLJ.^LO...E
30.0000. 30.0000. 32.S0. 32.0Co
32 0. 32. 0
1 + log (3S) Tioe (Min.)
0.543 0
0.525 60
0.510 120
0.479* 2o40
0.459 300
0,441 360
0.426 420
0.388 $40
Least Squares Slope
0.000284 (rin.)1
1 + log (0S) IThe (MIin.)
0.156 0
0.o45* 60*
0.424 120
0.389 240
0.372 300
0.39* 360*
0.336 420
0.320 400
0.302 $540
Least Squares Slope
0.000287 (min.)"1
1 + log (PRS) Time (Hin.)
0.497* o*
0.467 60
0.446 120
0.420 180
0.396 240
0.375 300
0.348 360
0.325 420
Least Squares Slope
0.000397 (min.)l
1 + log (i1S) Time (Uin.)
0.389 o
0.363 60
0.341 120
0.319 180
0.298 240
0.272 300
0.246 360
Least Squares Slope
0.000389 (min.)"1
*Not included in least squares calculation.
TABLE 8Continued
34.70C. 34.780C. 37.270C. 37.270C.
1 + log (RS) Time (Min.) 1
0.537 0
o0.5G 60
0.471 120
O.039 180
0.408 2140
0.374 300
0,343 360
Least Squares Slope
0.0005o 4 (~in.)1)
+ log (RS) Time (Min.) 1
O.!l1 0
0.l14 60
0.384 120
0.352 180
0.312 240
0.282 300
0.253 360
Least Squares Slope
.OOCC055 (rin.)1
+ log (RS) Time (CIin.)
o.560 0
0.529 45
0.494 90
0.6h4 135
0.).32 180
0.397 225
0.364 270
Least Squares Slope
0.000725 (nin.)"1
1 + log (PS) Time (Min.)
0.&39 0
0.403 45
0.374 90
0.337 135
0.302 180
0.273 225
0.236 270
Least Squares Slope
0.000747 (min.)"
~ ~~ ~~i~~
TAiLE 8Continued
1 + log (RS) Tiro (1iin,)
0.534 0
o.490 o4
0.U59 90
o.hoo 135
o.356 ISO
0.310 225
0.267 270
Least Squares Slope
0.0oC992 (min.)1
1 + log (RS) Tine (.in)
0..1.6 0
0.368 45
0.341 75
0.310 105
0.285 135
0.256 165
0.226 195
0.196 225
Least Squares Slope
0.000965 (min.)"1
39.770c
39.770C
~~ 
~~ ~ ~~~~
TALE 9
Su:LAR*i OF KIIETIC DATA FOR 2s,LUTIU.I:;EOJA;;E
48.49c. 48.490c. 48.h90c. 50.950c.
1 + log (RS) Time (ain.)
0.55,9" 0
0.528 60
0.500 120
0.477 10,
C.l51 240
0.431* 300*
0.397 360
0.373 U20
0.315 180
Least Squares Slope
0o.00033 (min.)1
1 + log (ES) Time (Min.)
0.610 0
0.586 60
0.553 120*
0.530 130
c.So4 24o
0.473 300
0o.43* 360
o.419 h20
Least Squares Slope
0.000458 (rain.)1
1 + log (RS) Time (Iki,.)
O.410* 0
0.389 60
0.362 120
0.334 180
0.308 24o
0.282 300
0.260* 360*
0.226 h20
Least Squares Slope
0.0ooo0051 (min.)"1
1 + log (RS) Time (Uin.)
0.531 o
0.502 60
0.h64 120
o.l31 180
o0.oo 240
0.357* 300*
0.328 360
0.294 420
Least Squares Slope
.ooo0568 (min.)1
*Not included in least squares calculation.

TABLE 9Continued
50.950. 53.350C. 53.3$c. 55.870C.
~~ J Ii 1 I
1 + log (rS) Time (Min.)
O.462 0
0.434 60
0.394 120
0.363 180
0.334 2Loe
0.293 300
0.263 360
0.227 420
Least Squares Slope
0.000563 (min.)1
1 + log (RS) Time (min.)
0.563 0
0.525 45
0,95 90
0.457 135
0.424 180
0,391 225
0.355 270
Least Squares Slope
0.000764 (min,.)1
1 + log (RS) Time (Iin.)
o.489* o*
0.463 45
0.!26 90
0.393 135
0.358 180
0.321 225
0.293 270
Least Squares Slope
O.000762 (min.)1
1 + log (is) TiMe (Min.)
0.519 0
0.468 45
0.428 90
0.382 135
0.349 165
0.323 195
0.295 225
Least Squares Slope
0.000993 (min.)"1
*Not included in least squares calculation.
I
TABLE 9Continued
55.?870. 58.37C. 8.370
... ,, = , ,
1 + log (RS) Time (llin.)
o ,!5o o
0..50 0
O.J01 45
0.356 90
0.308 135
0.281 165
0.249 195
0.220 225
Least Squares Slope
00.0102 (min.)1
1 + log (RS) Time (iin.) 1
0.581 0
0.544 30
o0.oh 60
0.l60 90
0.19 120
0.373 150
0.312 180
Least Squares Slope
0.00136 (nin.)"1
+ log (0S) Time (lin.)
0.482 0
0.432 30
0.395 60
0.35h 90
0.310 120
0.279 350
0.246* 180*
0.156 20o
Least Squarcs S1~pe
0.00o13 (min.)
*Not included in least squares calculation.
 
TA LT 10
SUI'MAE OF KIC2TIC DATA FOR 2,6LU1 Tl .0iAvE
10.000c, 10.00C. 1.oo0%. 1.oo00c
1 + log (RS) Thme (Uin.)
0.596 0
o.589 60
o.570 165
0.561* 2,5"
0.532* 435
O.4h9 615
0.394 12P5
0.367 1?.25
0.331 1605
Least Squares Slope
.00o016o (min.)"1
1 + log (RS) Time (Miin.)
0.488 0
0.481 75
0.462 165
0.W 2 255
0.425* 435*
0.392 615
0.287 1245
0.254 125
0.227 1605
Least Squares Slope
0.000165 (min,.)"1
1 + log (RS) Thie (MIin.)
0.577 0
0.558 90
0.535 165
0.510 240
0.492 315
O.466 390
o.h8 465
0.u11 570*
0.390 660
Least Squares Slope
0.000288 (min.)1"
1 + log (PS) Time (;.in.)
o.471* o*
o.450 90
o.432 165
o.410 240
0.388 315
0.368 390
0.3t44 65
0.315 570
Least Squres Slope
0.000285 (rain.)")
*Not included in least squares calculation.
~~
T.'LL? 10Contirnued
17,0800, 17.080c. 19.600c. 22,10o0.
. .,, , ,
1 + log (RS) Tine (Lin.)
0.538 o
0.520 45
o.490 105
0.470 150
0.453 195
0.428 255
O.a08* 315*
0.365 375*
Least Squares Slope
0.000436 (min.)'"
1 + log (;s) Tine (Iin.)
0.413 0
0.1422 15
0.393 105
0.372 150
0.356 195
0.329 255
0.302 315
Least Squares Slope
..cco'n4i (min,)1
1 + log (RS) Tihe (Min.) 1
0.538 0
0.524 30
O.505 60
0.o189 90
o0.65 135
o.434 1830
0,419 210
O.402 240
0.386 270
Least Squares Slope
0,000573 (min.)"1
+ log (RS) TiMe (rin.)
0.560 0
0.537 30
0.513 60
0.494 90
o0.69 120
0.37* 15*
0,420 130
0.398 210
0.370 240
Least Squares Slope
0.000785 (min,.)
L*ot included in least squares calculation.
TABLE 11
KIi:ETITC DATA FOR 2,4,6COLLIDIiEBORANE
30.00CG.
1 + log (RS)
0.51 1
0.526
0.517
0.499
0.497
0.479
0.468
0.461
Leat Sqares Slope
Least Squares Slope
Time (Min.)
0
15
30
45
60
75
90
105
120
 0.000752 (min.)"I
I" I~ ~` I ~  ""~~I
TABLE 12
FIT OF RATE CONSTANTS TO ARRHEU~IS EQUATION,
log k A + B/T (sec.)"
r 5 ; ~ A.
Amineborane A B I S e.u.
kcal../raole at 65$C.
pyridino 11.0+o0.10 5100+31 23.34+0.14 10.9
4picoline 12.13+0o.8 5576+282 25.52+1.29 5.98
3picoline 11.51+0.36 5316+117 24.33L+0.5 8.79
2picoline 11.59+0.12 5143+39 23.54+0.18 8.43
2ethlyl
pyridine 12.32+0.39 5240+119 23.98+0.54 5.07
2,4lutidine 11.43+O.42 5213+139 23.86+0.64 9.14
2,,6utidine 11.h2+,0.40 704+114 21.52+0.52 9.21
ji
2.600
H 2.700
o 2.800
2.900
Least Squares Slope 5576
259
Figure 22. Arrhenius
297/T x 10299 30
Plot for Picolineborane
Plot for hPicolineborane.
I I I i
60
2.600
2.700
S2.800
S4
0
2.900
3.00
I I
Least Squares Slope 5316
I I
306
298 300 302 30o
1/T x 105
Figure 23. Arrhenius Plot for 3Picolineborane.
I I I I
I
Least Squares Slope 5143
I I I
312 311T x 16 318
l/T x P l 
Figure 24. Arrhenius Plot for 2Picolineborane.
2.65o
2.75o
bfl
0~~
3.5o0
2700 Least Squares Slope 5240
2.800
bfl
2.900
0
3.000o
3.100
3.200
320 322 32h 326 32
1/T x 105
Figure 2$. Arrhenius Plot for 2Ethylpyridineborane.
Least Squares Slope 5213
I I I I I
310
2.650
I
S2.750
hO
0
2.850
2.950
302 304 306 308
1/T x 10i
Figure 26. Arrhenius Plot for 2,4Lutidineborane.
Least Squares Slope h70h
1/T x 105
Figure 27. Arrhenius Plot for 2,6Lutidineborane.
2.850
2.95o.
3.o0o
3.15o
bfl
3.250
3.350
CHAPTER IV
DISCUSSION
During the course of these experiments the conversion of the
excess iodate to iodine and the arsenite backtitration were carried
out at times ranging from less than a minute to several hours after
the kinetic sample had been pipetted into the iodate. Since no
difference in kinetic results was observed, it would appear that the
oxidation of a borane group by iodate is an instantaneous reaction.
This is especially interesting considering the minimal excess of
iodate used here, and the large excess required for a quantitative
oxidation of borohydride.
Failure to obtain a constant 5 ml. siple because of hydrogen
evolution in the pipette was the largest single source of experimental
error. The magnitude of the scatter caused by hydrogen bubbles may
be seen by comparing figures 15, 16, and 18, While the interval
between sampling is about the same, 2030 minutes, the experiment
which was magnetically stirred to purge the reaction mixture of hydrogen
shows considerably less scatter.
Since the iodate method gives the total amount of oxidizable
species, the RS value would be identical with the amineborane
normality only if there were no siCnificant buildup of oxidizable
intermediates. Considering the many variations in these experiments,
i.e., change of amineborane change of amineborane concentration,
65
addition of amine, change of solvent composition, change of tempera
ture, and investigation of differing extents of reaction, the excellent
fit of the data to the first order log (RS) versus time plots ar,ucs
against the existence of any oxidizable ini.cr.ediales. ',;c PS values
will therefore be taken as being equivalent to arineborane normalities.
The general rate equation for the solvolysis reaction is civcn
by
d(amineborane) = k1(iaineborane)m (alcohol)n (1)
dt
Since the concentration of alcohol remained essentially constant rela
tive to that of the amineborane, (1) may be reformulated ast
d(aminebor k( nborane) k ineborane (2)
dt
where k' = k (alcohol)n If "m" is set equal to 1, and the amine
borane concentration exi.re
a, and x, the number of roles per liter which have reacted after time
t, equation (2) becomes
dx/dt = kI (ax) (3)
A'trLr separation : the variables, (3) is readily inI.c rated to .ive
ln(ax) = k't + Consto (4)
When t O0, x = 0, so that the constant of (4) is simply In a.
Whence (4) becomes
ln(ax) = k't ln a (5)
A plot of lo:; (ax) a:.ainst ti ie should thus be linear with a
slope k'/2.303, and an intercept lo; a. Fro.i equations 35 it can be
seen that the rate of this pseudofirstorder reaction is independent
of both initial concentration and initial time, and is determined only
by !he concentration of unrcacted amineborane. The excellent linearity
of the log (PS) versus time plots attests to the solvolysis being first
order in amineborane.
When it was found that vigorous stirring would cause several
hundred milligrams of pyridineborane to dissolve in 100 ml. of
water, both quantitative and qualitative investigations were made on
the rate of reaction of pyridineborane with distilled water. In a
quantitative experiment at 5oC., only about 12% of the active hydrogen
had been lost in 19 hours, corres.ondirn; to a halflife of about 90 hr.
The halflife for the propanol solvolysis at this temperature is about
6 hours. After having remained at room temperature for a month, an
aqueous pyridineborane solution readily reduced silver nitrate and
evolved hydrogen upon addition of dilute acid, On the basis of these
results it was assumed that within experimental error, water could be
used an an inert diluent for the propanol.
Returning to the relationship that k' = k(alcohol)n ,
log k' = log k + n log (alcohol) (6)
A plot of Ic, k' against log molall concentration of alcohol) should
be linear with slope "n" and intercept log k. Figure 28 is such a
plot for pyridineborane with pure p opanol and propanolwater mixtures.
'Tr curvilinear relationship obtained in figure 28 is attributed to
strong electrostatic interactions of the amineborane with the solvent,
causing log k to vary. This result i. :it have been anticipated from
the vastly different slopes obtained with pyridineborane for 50C
propanol50C water and 50% propanol5 C. dioxane by volume (table h).
Uhile the former had a slope several times smaller than the value for
2.950
3.150
0
O0
3.550.
I I I I
.900 1.00 1.10 1.20
log molall concentration of alcohol)
Figure 28. Rate dependence on alcohol concentration
50.00C.
pure alcohol at the same temperature, the latter had a slope several
times larger than the pure alcohol slope. Further evidence of solvent
electrostatic interactions with pyridineborane comes from the work of
il:hecva and Fedneva.13 These investigators found pyridineborane to
be highly associated in benzene solutions, and only slightly associated
in the more polar nitrobenzene solutions of the same concentration.
Considering only strong electrostatic interactions, Kirkwood,19a
has derived a relationship which gives the effect of the dielectric
constant of the medium on the free ener;;: of a polar noleculo. For a
rate process this relation becomes
1 o + (7)
RT(aD+I) L 3 ri rJ
where k is the rate constant in medium of dielectric constant D; ko is
the rate constant in a condensed rcdi,. of dielectric constant unity;
the / %3 refer to the dipole moments of reactants, A and B, and of the
transition state; and the r's refer to the effective radii of these
species. While the dielectric constant of the iKirkwood equation is
actually the value for the solution, it has been found that for a dilute
solution the dielectric constant of tile pure solvent nay be used.
For various mixtures of two inert solvents, a plot of log k
against (D1)/(21+1) should be linear. In addition, if as a first
approximation the volume of the activated complex is taken to be equal
to the sum of the volumes of the reactants, a negative slope should be
obtained when the ground state has a larger dipole moment than the
transition state.
Dielectric constants for propanolwater mixtures by weight have
20
been detenained at 500C. by Akerlof.2 The dielectric constant for the
to,6% alcohol mixture was obtained from a plot of electricc constant
versus weight percent alcohol, which was found to be linear in the range
2060W alcohol. If the rate data for propanolwater mixtures were used
with AKcrlof's dielectric constants, a linear iirkiood plot could be
realized only if the solvolysis were zero order in alcohol. The Kirkwood
plot is shown in figure 29. In addition to visual linearity, this plot
was analyzed statistically (see statistical appendix) for a linear
relationship between the two variables. The correlation coefficient
proved to be 0.99, giving a highly significant linear relationship
between log k and (D1)/(21)+l). It is thus concluded that the solvolysis
reaction is zero order in alcohol and that the dipole moment of the
ground state is larger than that of the transition state. kt of equa
tion (2) must then be identical to kr,
The calculation of Arrhenius activation energies was given in
chapter 3. Tntrcipies of activation were calculated as follows: The
"thermodynanic" treatment of reaction rates9b gives
kr = (kT/h)edS*/R eH/RT (8)
where,4H" is given by
dH = EArr. RT (9)
for a reaction in solution in the liquid state. Taking the common
logaritlh of both sides of (8)
log kr = log(kT/h) + d S*/2.303 R AI1/2.303 RT' (10)
Substitution of a particular temperature, in this case 338.20i., into
3.30
S3.
o 3.50.
3.60.
.63 .h69 .h75 .481
D1
( 2D+1 )
Figure 29. Kirkwood plot for propanolwater mixtures at 500C.
the least squares equations of the Arrhenius plots gave the values of
log kr in (sec.)"1 for that temperature. If (kT/h) is taken as approxi
mately 1013, and equation (9) is substituted into (10),
log kr 338,20i. = 13 +S*/2.303 R (EArr.RT)/2.303 RT (11)
Equation (11) nay thus be solved for the entropy of activation.
Consistent with this work would be a mechanism involving a slow
dissociation of the amineborane into one or more reactive fragments
which would then undergo a fast reaction with the solvent to yield the
observed products. Two different dissociative mechanisms will be con
sidered. The first of these would involve essentially an ionization of
the amineborane to give hydride ion:
araine'BH3l ) (anine'BlH2) + H" fast
3 slow \ RH Products (12)
Equation (12) predicts a large generation of charge in the transition
state, and hence an increased rate with increasing dielectric constant
of the solvent, Exactly the opposite effect was found experimentally.
lBrg21has suggested that as the Lewis base stren:th of the donor atom
is increased, the hydridic character of the borane complexes also is
increased. If (12) were the correct mechanism the Arrhenius activa
tion energies should decrease with increasing base strengths for steri
cally si.il.r pyridineboranes. "The pKa values for the pyridineboranes
are listed in table 13 along with a number of other physicochemical
quantities. Thus both 3picolineborane and Ipicolineborane should
have Arrhenius activation e.ncr, ics less than that for pyridineborano,
while 2,tlutidineborane would be expected to have an Arrhenius acti
vation .ner,~ less than that for 2picolineborane. Again just the
opposite result was obtained experimentally. This dissociation
mechanism is hence eliminated as being inconsistent with experiment.
The second dissociative mechanism involves a slow dissociation
of amineborane into the amine and a borane group, followed by a fast
reaction of the latter with the solvent:
a~inb*3 l aiiino + BII3 fast
sI ROH > Products (13)
The transition state in (13) would have a lesser separation of charge
than the ground state because of the lesser extent of dative bonding
between boron and nitrogen.
Hence a decreased rate with increasing dielectric constant of
the solvent would result. Looked at from a different viewpoint, the
polar amineborane ground state (dipole moment of pyridineborane
5.86 D.) would be stabilized relative to the nonpolar transition state
by an increase in solvent dielectric constant, thus causing the rate to
decrease. Fxperimentally just such a decrease in rate with increasing
dielectric constant of the solvent was observed, If (13) were the
correct mechanism, the Arrhenius activation energies should increase
with increasing base strengths for sterically similar pyridinoboranes.
That such an increase was found experimentally has been mentioned
above, and is evident from the data in table 13.
Since the dissociative mechanism of equation (13) is in good
accord with experiment, it will be examined further. A steadystate
assumption for the concentration of borane yields the following rate
law t
d(products) = klk2 (,'ineborane) (alcohol)
dt k. (amine) + k2 (alcohol) (14)
Depending upon the relative nagnitudes of the kl(azinine) and k2(alcohol)
terms, a variety of rate dependencies can be predicted. If they are
both of about equal magnitude, the rate would be first order in amine
borane, and of c. .,A,x order in alcohol and arine. If the k_(amine)
toer isneglibible in comparison to the k2(alcohol) term, a shaple
first order reaction in amineborane would be observed. If, however,
the opposite were true, the rate would be found to be first order in
aaine'oorane, first order in alcohol, and inversely proportional to the
amine concentration.
It is apparent from the fonm of (lh) that either k.2 would need
to be very large or k2 very snall for the concentration of amine pro
duced by the reaction to have a detectable effect on the rate in alcohol
solvent. Tn.5 the log (iS) plots in figures llb and 11c are still
linear to almost 33 halflives. While the slope has actually in
creased in the latter plot, the overall least squares slope for figure
11 is only about le larger than the initial slope. Consideration of
the experimental error involved in determining the RS for times greater
than 2 halflives has led to the conclusion that this difference in
slope is not significant.
The further attempts to detect a retardation by adding amine
0.26M and 0.14Z figures 19 and 20 respectively, were also fruitless
as the rates in these e;:,,ri c,ts agreed with each other and with the
rate obtained in the absence of additional arinc, to within the
coti.ited experimental error. It was not possible to use a higher
concentration of amine, because this substance interfered with the
analytical procedure by completing the iodine.
If, as these results indicate, the k.l(amine) term is negligible,
then equation (lh) becomes
d(products) = kl(anineborane)
dt (15)
which is exactly the rate dependence obtained previously from a
consideration of the individual rate dependencies of amineborane and
alcohol. Hence it is concluded that the solvolysis mechanism of
equation (13) is entirely consistent with the experimental results if
the recombination rate is assumed to be insignificant with respect to
the rate of the boranealcohol reaction. The validity of this assump
tion will be considered later. In nccepAi.t: equation (13) as the most
probable mechanism consistent witi experiment, it should again be noted
that any hypothetical mechanism involving the hydridic character of
the amineborane would necessarily result in the fallacious prediction
of decreasing experimental activation ener':ics with increasing base
strength s for sterically similar pyridineboranes.
Information as to the nature of the transition state for the
solvolysis reaction can be obtained by ai.plication of an in;genious
technique developed by Brown et al.22 These investigators found that
for 1"nar Lewis acidbase reactions of the pyridines an estimate of the
relative steric requirements of the acids could be determined from
plots of an energy function for the reaction of each acid with the
pyridines against an encr,y function for some other reaction of the
pyridines. The ener:' functions may be activation energies, heats of
association, rate constants, and even pKa values.
Three plots of this ty! e, the data for which nay be found in
table 13, are sho~m in figure 3032. Following Broom's Cexaple,
energy differences between the values for the respective substituted
pyridines and that for pyridine itself are plotted, rather than Vhe
actual nr.rr:: va.l,. Figure 30 is a plot of the Arrhenius activa
tion cnr 'ic; for the solvolysis reaction :;:..rt the heats of reac
tion of the pyridines with methane sulfonic acid. "Toto that the
orthosubstituted pyridines fall below the line formed by pyrid'L;ic,
3picoline, and hpicoline.
A stxilar plot of the Arrhenius activation energies for the
solvolysis reaction .;aint the heats of reaction of the amines with
diborane is shown in fi 're 31. In this case, however, the ortho
substituted pyridines lie above the line formed by pyridine, 3picoline,
and hpicoline, Finally figure 32 is a plot of the Arrhenius activa
tion enerics for the solvolysis reaction against the heats of associa
tion of the i;rL:V'i s with boron trifluoride. Here the orthosub
stituted pyridines aain lie above the line formed by pyridine,
3picolinnr and hpiicol:nin, but are much further displaced from the
line.
It has been shown by Brown et al.22 that the steric require
ments for the acids considered in figures 3032 increase in the Cx
pected order proton boranleboron trifluoride. The generalization
may thus be made that reactions of the pyridines with acids having
a smaller steric requirement than the solvolysis transition state
will result in the orthosubstituted pyridines falling below the
pyridine, 3picoline, and hpicoline line when the inr, ,'ii data are
plotted as above. Sin.larly acids with a greater steric rcquirer:ment
TABLE
T17SICOCI!FTJCAL DATA
13
FOR 7l r IP'YP.!I!.TS
Amine pKaa Heats of reaction, kcal./molea Arrhonius b, c
Activation
CH1S3OH 1(BHl3) BF3 Energes
pyridine 5.17 17.1 17.9 32.9 23.34
4picoline 6.02 18,. 18.5 33.4 25.52
3picoline 5.68 17.8 18.2 33.2 24.33
2picoline 5.97 18.3 17.2 31.2 23.54
2ethyl
pyridino 5.92 18.2 16.9 30.6 23.98
2,4lutidine 6.79d 23.86
2,6lutidine 6.75 19.5 16.3 25.4 21.53
a reference 22
b this work
c kcal./mole
d Detennination of Organic Structures by Physical Methods, E. A.
Braude and F. C. Nachod Academic Press Inc. iic' York, U. Y.,
1955, p,,o 59h.
3Pic
O 2Etpy
O 2Pic
2,6Col
O
dHB 
Figure 30. Steric requirement
of methanesulfonic acid, for
1.6 2.1
AH kcal./mole
of transition state relative to that
the pyridine bases.
1.6
0)
4
0
C)
0.0
1.6
hPic
3Pic
O 2Etpy
O 2Pic
D 2,6Col
I I
.81 o.o0
dHB Hpy kcal./mole
Figure 31. Steric requirement of transition state relative to that
of a borane group, for the pyridine bases.
.8
1.6
0.0
1.6
'"
i
4Pic
O 2Etpy
02Pic
O 2,6Col
I
6.0'
SHB
Figure 32. Steric
to that of boron
4.o 2.0 0.0
 A Hpy kcal./mole
requirement of transition state relative
trifluoride, for the pyridine bases.
1.6
 I

~
X
I
than the solvolysis transition state will give plots with the ortho..
substituted pyridines lying above the line formed by yrridine,
3r'icoliic, and 4picoline.. INeglecting specific sterio interactions,
a statistical analysis (sec statistical appendix) of figure 31 for a
linear relationship between the Arrhcnius activation energies and the
heats of association of all the pyridines with diborane yielded a
correlation coefficient of 0.83. iTis value was statistically signi
ficant at the 0.05 level, but not significant at the 0.01 level. When
steric interactions are considered, however, it is seen by inspection
of figure 31 that the correlation is much better and 1iore closely
related to chemical behavior if the results are grouped into steri
cally similar compounds. Thus pyridine, 3picoline and 4picoline
would form one group, while 2picoline, 2ethylpyridine, and
2,6lutidine would form another.
On this basis it is concluded that the transition state for
the solvolysis reaction has a smaller steric requirement than the
association reaction of the pyridines with dioborane. Since the
equilibrium
B2116 = 2IH3
is a comraon factor in the association of the pyridines with diborane,
it is also cr ,lu.cld that the steric requirement for the solvolysis
transition state is less than that for the formation of amineborane
from the awine and a borane group. This is of course in agreement
with the proposed solvolysis mechlaniir, as a stretching of the NB
bond in a dissociation would relieve steric repulsions between BH3
and an ortho substituent.
Returning then to figure 31, if the line formed by pyridine,
3picoline, and 4picoline is taken as the relationship between these
two reactions in the absence of steric effects, then deviations from
this line may be taken as an estimate of relative steric strain. Thus
transitionstate steric strains for the solvolysis reaction were
estimated to be about 0.8 kcal./mole for 2picolineborane, and 1.1
kcal./mole for both 2ethylpyridineborane and 2,61utidineborane.
These values may be compared with the respective ground state strains
estimated by Brown and Domash,12 1.3, 1.5, and 2.7 kcal./mole.
If a value of 28.5 kcal./mole23 is taken for the gasphase
enthalpy of dissociation of diborane into 2 borane groups, and if
it is also assumed that heats of reaction in nitrobenzene do not
differ significantly from those in the gasphase, the gasphase
enthalpy change for the reaction
CoSH NBH3 C5H + BH3 (16)
may be estimated to be 32.2 kcal./mole. While it would thus appear that
the Arrhenius activation energies were far below the energetic require
ments of the proposed mechanism, it should be remembered that the 32.2
kcal./mole value was for the gasphase, where there are no solvent
effects. A strong coordination of either or both of,the products of (16)
relative to that of the ai.dneborane would supply sufficient energy to
overcome the apparent deficiencies. Ihe situation here is quite similar
to the simple reaction of dissolving an ionic solid in water, for which
reaction the energy required to overcome the ionic lattice is supplied
by coordination of the ions with water. A very strong coordination of
the boron atom of the electrondeficient borane group with the oxygen
atom of an alcohol molecule will thus be proposed not only to account
for the aparent energy differences, but also to explain the negligible
rate of the reverse of equation (16) in propanol solution.
Evidence for this tight coordination comes from the reactions
of diborane with alcohols. With methanol, for example, more than enough
coordination energy is available to dissociate the diborane even in the
gasphase.24 If a limited amount of alcohol is used, intermediates
such as (CH30BH2)x and (CH30)2BH may actually be obtained. The avail.
ability of a coordination energy of this magnitude for a borane group
in propanol solution would thus be more than sufficient to make the
proposed solvolysis mechanism energetically feasible. Concomitantly,
once a borane group were tightly coordinated, the association energy
released by the formation of amineborane would not be sufficient to
overcome the coordination energy, and hence the rate of the associa
tion reaction would be predicted to be insignificant with respect to
the dissociation reaction.
Hawthorne and Lewis25 have proposed that the hydrolysis of
pyridine diphenylborane involves the electrophilic attack of a water
proton on the electrons of the BH bond. Under the same conditions
the hydrolysis of pyridine phenylborane was too slow to measure accu
rately. In view of the considerable steric and electronic differences
between pyridine diphenylborane and the pyridineboranes, it is thus
not unreasonable for diverse solvolysis mechanisms to hold.
As final support for the solvolysis mechanism proposed in
this dissertation, Davis and Kirby26 have recently reported a
negligible isotope effect, k/kD 1.05+0.02, for the reaction of
8h
trimethylamineborane in 1.401~ hydrochloric acid. Since this reaction
has also been found to be first order in amineborane,15 a dissociation
into the amine and a borne Zroup similar to that for the propanol
solvolysis is indicated for the protolysis reaction. It is indeed
gratifying to report that the protolysis reaction for pyridineborane
has been found to be very much faster than that for trimethylamine
borane under the same conditions.27
STATISTICAL AiTE:DIX
The least squares "normal" equations are well known and have
been used extensively in the physical sciences to fit a straight line
to experimental data. How to obtain standard errors for sample esti
mates of the parameters in the least squares equation, on the other
hand, is little known and not always to be found in mathematical hand
books. The calculation of these quantities, the calculation of a cor
relation coefficient, and a statistical test for the homogeneity of
several least squares slopes will be given in this appendix.28 29
Both the mathematical expressions and the arithmetical calcula
tions will be considerably simplified if the following symbols are
defined:
S?2 _Y; (E y)2/n
x2 FX2 ( x)2/n
Sxy
where n is the number of points to be fitted by the least squares
line, and X and Y refer respectively to their Cartesian coordinates.
If the equation of the line to be determined is written as
Y' + a + hb
it can be shown that
b =fxy/'x2 and a *= Y bX
where X and Y are the corresponding averages of the X and Y variables.
When the individual deviations of the points are of no particular
interest, the sum of squares of deviations from the least squares line
85
may be obtained from
2 y2 )2/ 2.
Since a and b are estimators, computed from the sample, two degrees of
freedom have been lost, and the mean square deviation from the line
is given by
1.x d.d/(n2).
The standard errors for the slope and intercept are respec
tively determined from
sb 0 sy.X/f ) 2 and
sa = sY.x(/n + X21 )
The preceding equations have been derived from the assumptions that
the X values are measured without error, and that for each X, the Y
values will be normally and independently distributed with the same
variance. They are however, often applied even though the X's are
not exact if the error in X is negligible with respect to that of Y.
Data with a considerable amount of scatter may be tested for
a linear relationship by computing the correlation coefficient.
The correlation coefficient is a measure of the fraction of the total
variation explainable by a linear relationship between the variables.
It is defined by
r g371y/(f x2y2)t
The calculated value of the correlation coefficient is then examined
for statistical significance.
Pour kinetic experiments made with hpicolineborane at 620C.
gave least squares slopes of 0,000827, 0.000837, 0.0008~1, and
0.000926 (min.)"l. While the last slope would sceea inordinately
87
high, there was no justification for arbitrarily discarding it. Con
cequcntly a statistical test based on the hypothesis that the slopes
were identical was applied to the data from these experiments, and the
hypothesis was rejected. Elimination of the 0.000926 (min.)" value
resulted in acceptance of the hypothesis that the slopes were identi
cal when the test was repeated. Hence the fourth slope was not used
in determining the avera e slope for this temperature. The test which
resulted in rejection is illustrated in table 14.
TEST FOR :0C.. MF!ETTY.
TABEI 14
OF SEVERAL
LAST SQUSTTArr SiLPES
Experiment Degrees M22 :w2 $272 2 Degrees Mean
of ( xy) /x of Square
Freedom Freedom
1 6 25200 20.850 .017258 .000007 5
2 6 68336 57.221 .47964 .000050 5
3 6 33429 28.433 .02W206 .000022 5
S?5 39000 .36.100 .033h59 .000043 4
23 165965 142.60o .122887 .000122 19 0.00000642
Totals
.000356 22
Difference for testing Ho b, = b2 b3 = b4 .000234 3 0.0000780
3 0.0000780 12.15
19 0.00000642
For the 0.05 level, F3 is 3.13 therefore reject Ho
19
SFrT.': CS
1. A. !. 7. ur: and HI, I. ,clQcoin cr, J. An. Smhc:. soc., 29, 70o (1937)
2, S :. auer, J. A. fI:cn. Soc1.02, 9 lOO1 (1937)
3. S, Gellor, 1, E. Hushies, and J, L. Holard, Acta. Cr st. 3 0 (1.1)
4h* 4. C, Price, R. D, Be Fraser, To. 'obinson, and H, C, Longuet.
uilc;lins3. ice. 1araday Soc., 1950 ('9 1231.
5. Bi. mico, n, J, Galiano, and 9,' J. Lchm.~n, J, Ph:4s. Ch i, 61,
1222 (1957)
6. A, RIt :atrit53::., J* Chem, SOC., 2049 (1959)
7. C*. ;, *ax., A. R. .atritzcrl, and L. E.* utton, J. Chem. Soc, 1258
(1?58)
6, R. E. I'cCny and S. 1. lDauer, J. An. Chec. Soc., 70, 2061 (19?6)
9. UT 1, r'hill1ip, I. C, Irller, and '", L, t'uottrties, J. Am Chli. .
So 8c. 1, 81 96 (1959)
10. H. H. Jones, J. Am. Chl. Soc., 82, 2528 (1960)
U11 HI C. 3rown, H. I, SchlcinAcr, and 1. 2. Cordon, J. A.7 Cho. *oc.
66 325 (1942)
12. HI C, ;rc;un and L. Eco' i, J. An. Chem. Soc., 7, 53'4 (1.6)
13. V, I. "il:':ccva and E,. M. ,'ccn vr 4.ur. ::or. f.irA. 894 (1956)
4I ZI M D. 7a;lor, L, :,. "rant, and C. A. Sandas J, A., Chr o Soee,
77, 106 (1955)
15. G B. WE. aschiritsch, J. A i, Che. Soc., 02, 3290 (1960)
16, ve would like to tian.. the Callry C rheIical Copany for sawplca of
^'rldincborano and 2,L.l1utidineborane.
17 1.. Jrnsen, !. Little, and U, ltruc,:, Anal. Chin., 24, l8h3 (i.52)
18. Ii. C. irown ind P. A. Tiernf, J. Anm. Chem. Soc., 00, 1552 (156)
19a. ;inctica an ::ccnnionr by A. A* Front andi RG* O Pearson, John lilcy
aid cons, njc.3, :c:u Iori, Y. (1953) p. 130.
19b. Ibid., p. 95.
20. o. Ai:,rlo", J. An. C:5n. Soc,, 2, 125 (1932)
21. A. B, ur rc..orc of Chrcical Proreso, .35 / 159 (19h)
22. 11. C. 2rc'rn, ,. Gintis, and L, Coaash, J. An. Chcn. Soc., 78,
5387 (1956)
23, S. H. Bauer, J. Am. ChQa. Soc.. 78, 5575 (1956)
2h* A. B, DBurg and II. 1. cSchLesin.cr, J. A.i. Cho:.. Soc.* $*, 4020
(1933)
25* ':. F, i.auth2or:e anid nr, S, 1".D, J, C;c., coo., G., 4296
(193;)
26, R, F' Pavis and, C. L. Kirby, J. Ar. C;]n. Soc., 82, 5950 (196n)
27, G0 E, TrschlewIitsch, unpublished woirk
28. Statistics in Research by Bernard Ostle, Iowa State College Press,
Amcs, Iowa (195)1 p. 133138.
29. Statistical Methods by George W. Snedecor, Iowa State College Press,
Ames, Iowa (1956) p. 122159.
BIOGRAPHY
Ernest Rodman Birnbaur was born on October 4, 1933, in Newark,
New Jersey. His undergraduate work was started at U.C.L.A. and com
pleted at the University of California in June, 1955 with the attain
ment of a Bachelor of Arts Degree in Chemistry.
He began his graduate study under Dr. Anton Burg at the
University of Southern California where he was a teaching assistant
for two years. Following graduation from tlis institution with a
Master of Science degree in Chemistry he came to the University of
Florida in August of 1958 to pursue work leading toward the degree
of Doctor of Philosophy.
While at the University of Florida he has held a research
fellowship. He is a member of Kappa Nu, the American Chemical Society,
and the Chemical Society of England.
Thio dis sert.tion o~s prepared runcr tho direction of t.he
chairman of the ccndidato's .up:crisory coii,.tte.. and has been approved
by all .Tebers of that co~iittee. It was shbteitted to 'the Poan of the
CollcLc of Arts aznd Ecicnccs cnd tIo the Gracuhate Covncil, and was
ali ro.'(_ as partial fulfllT'reInt of the rrcirecen for the degree of
Doctor of Fhilosophy,
January, 1961
Dean, College of Arts and Sciences
Dean, Graduate School
Supervisory Cor mittee:
fl/Ai
L2 jZ^
Internet Distribution Consent Agreement
In reference to the following dissertation:
AUTHOR: Birnbaum, Ernest
TITLE: Solvolysis kinetics ofpyridineboranes. (record number: 424005)
PUBLICATION DATE: 1961
I, ri fv rZ a t i'rla O4 as copyright holder for the
aforementioned dissertation, hereby grant specific and limited archive and distribution rights to
the Board of Trustees of the University of Florida and its agents. I authorize the University of
Florida to digitize and distribute the dissertation described above for nonprofit, educational
purposes via the Internet or successive technologies.
This is a nonexclusive grant of permissions for specific offline and online uses for an
indefinite term. Offline uses shall be limited to those specifically allowed by "Fair Use" as
prescribed by the terms of United States copyright legislation (cf, Title 17, U.S. Code) as well as
to the maintenance and preservation of a digital archive copy. Digitization allows the University
of Florida or its scanning vendor to generate image and textbased versions as appropriate and
to provide and enhance access using search software.
This grant of permissions prohibits use of the digitized versions for commercial use or profit.
"__/ I /&z^CtLr
Signature of Copyright Holder
ERNEST R. BIRNBAUM
Printed or Typed Name of Copyright Holder/Licensee
Personal information blurred
Dae of Signature
Please print, sign and return to:
Cathleen Martyniak
UF Dissertation Project
Preservation Department
University of Florida Libraries
P.O. Box 117008
Gainesville, FL 326117008
