Dec.
, 1950 HYDROLYSIS
OF ETHYL
SILICATE
Summary
1. The kinetics of the alkaline hydrolysis of
potassium ethyl malonate has been studied over
a temperature range 5 to 35’ in two isocomposi-
tion media, water and 27.57 wt. % dioxane in
water; in three isodielectric dioxane-water mix-
tures corresponding to dielectric constants of 50,
55 and 60; and a t 25’ in three isodielectric t-
butyl alcohol-water mixtures at the above
dielectric values.
2. A positive primary salt effect has been
found, which is in accord with the mechanism
postulated for the reaction.
3. The distance of closest approach of the
reacting ions is shown to be 4.8 A. in water and in
the dioxane-water mixtures. It is shown to be
4.3 A. in the t-butyl alcohol-water mixtures.
4. A comparison is made between the use of
volume concentration and mole fraction units for
the expression of rate constants.
[CONTRIBUTION
FROM THE INSTITUTE
OF POLYMER
RESEARCH,
POLYTECHNIC
Hydrolysis of Ethyl Silicate*
BY R.AELION,A. LOEBELAND F. EIRICH
The hydrolysis of ethyl silicate is used in in-
dustrial applications as a convenient process
for obtaining an adhesive type of silica. Hy-
drolysis will occur readily in the presence of acidic
or basic catalysts, but due to the immiscibility of
water and ethyl silicate, a mutual solvent is often
used to obtain homogeneous reaction. A great
deal of attention has been focussed on the de-
velopment of the most convenient methods of
obtaining the right degree of conversion of ethyl
silicate to silica. For example, diagrams of
solubilities of ternary mixtures of water, ethyl
silicate and mutual solvent have been determined,l
and special commercial processes for hydrolysis
have been developed.2 Quantitative informa-
tion on the rate of hydrolysis of ethyl. silicate in
acidic and basic media and on the stability of
hydrolysis products to gel formation is, however,
still very scant.a The purpose of the work de-
scribed in this paper was the collection of quantita-
tive information regarding the rate of ethyl
silicate hydrolysis, and the formulation of a
possible reaction mechanism.
The over-all hydrolysis of ethyl silicate and
the subsequent dehydration of the product
formed may be formulated as
* Taken in part from a thesis in preparation by M. Arnold h b e i
for presentation to the Graduate School of the Polytechnic Institute
of Brooklyn for the degree of Doctor of Philosophy in Chemistry.
(1) H. D. Cogan and C. A. Setterstrom, Chem. Eng. Nems, 44,
2499 (1946); I d . Eng. Chem., 89, 1364 (1947).
(2) C. Shaw and J. E. Hackford, I r d . Chcm., 01, 130 (1945);
as, 61 (1946); aa, 1230 (lace).
(3) G.King, J . Oil and Colour Chem. Assoc., 18, 116 (1930).
5705
5. The energy of activation at zero ionic
strength is greater in isodielectric media than in iso-
composition media which is in accord with theory
for a reaction between ions of like sign. Com-
parisons are made with theoretical calculations.
6: The effect of ionic strength upon the energy
of activation is in excellent agreement with theory.
7. The influence of dielectric constant upon
the rate constants is in good agreement with the
Scatchard-Christiansen theory in dioxane-water
mixtures.
8. Various thermodynamic activation values
have been calculated. It is shown that the free
energy of activation, energy of activation, and the
entropy of activation, if calculated for D = a,
are independent of solution composition.
9. It is shown that the simple collision theory
and the thermodynamic treatment of La Mer
are not in agreement for this reaction.
COLLEGE PARK, MARYLAND RECEIVED MAY22, 1950
INSTITUTE
OF BROOKLYN]
rzSi(OCzH& + 4nH20 +
nSi(0H)r + 4nGHsOH (la)
nSi(0H)c +(Si0t)n + 2nH2O Ob)
Depending on conditions, one or both of these
consecutive reactions may only go to partial
completion, leaving ethoxy or hydroxyl groups
unreacted. The changes with time in the follow-
ing quantities were experimentally determined :
(1) The total loss of water, LT,equal to the
difference between the water taken up by the
hydrolysis and the water regenerated by de-
hydration. (2) The production of alcohol, LH,
equivalent to the amount of water used in the
hydrolysis reaction. (3) The degree of conversion
of monomer.
Expressing LT and LH in moles per mole of
ethyl silicate, a simple equation relates these
two quantities at any stage of the reaction
LT = LE -aL~/2 (2)
where a is the degree of completion of the de-
hydration reaction.
By determination of LH and LT the rate and
extent of both reactions and therefore the number
of unreacted ethoxy and hydroxyl groups can be
calculated. The average composition of the poly-
ethoxysiloxanes formed is then obtained by evalu-
ating the degree of conversion of the monomer.
Experimental
Titration with Karl Fischer reagent‘ allows the instan-
taneous determinations of water content. This method
(4) J. Mitchell and D. M. Smith, “Chemical Analysis,” Vol. 5.
Aquametry, Interscience Publishers, New York, N. Y.,1948,Chap. 4.
5706 I<. AEIJON, A . LOEBELAND F. EIRICH Vol. 72

requires the use of a intitiial solvent to obtain a homo- were found to be greatly influenced by the pres-
geneous reaction medium. If 2cfh and -?!$ are the number ence and concentration of strong acids or bases.
of moles of water per liter a t time lo and time t , respectively,
then LT is given by Temperature and the type of solvent produce
LT = (Mta - -%fd/S, secondary effects. To obtain results which lend
(3 1
where SC is the initial concentration of silicate in moles
themselves to easy interpretation, either the
per liter. Although the absolute accuracy of this method catalyst concentration or the concentrations of
is similar to the accuracy of the usual volumetric titra- water and ethyl silicate were kept practically
tions (lye), the relative accuracy is much better and constant. The data listed in Table I were ob-
allows assessment of very small variations of the n ater tained from runs made at decreasing concentra-
content (0.2%).
The amount of alcohol formed cannot be titrated by tions of hydrochloric acid, employing dioxane
direct chemical methods rvhich rely on a hydroxyl group as solvent.
determination (e. g., Zeisel’s method), since the products TABLEI
of the reaction have the same functioual groups. The
alcohol has, therefore, t o be separated from the reacting HYDRO[
\SI5 Ob E r H Y L S I L I C A l E WITH STROYG HYnRo-
medium, e. g., by distillation. A complete separation CHLORIC ACID SOLUTIONS
can be obtained by using a mutual solvent of higher boil- Unre-
acted
ing point than that of the ethyl alcohol. A rapid partial mono-
distillation under vacuum of an aliquot sample removed h a LT LEI mer
during reaction yields a distillate containing all the alcohol ethyl m./ m./ m./
si& Mt mole mole Si-
and the unreacted water. The concentration of alcohol HCI. cate, water Si- Si- (1, f 0 C ~Time,
in the mixture is determined by density, refractive index Expt m./l m /I. m./l.’ (0CnHa)a (OCnH6)a % Ha), min.
measurements, ete., which have been previously stand- I n063 0 . 1 9 ~ 1.432 o 0 1 0
ardized. If Ct is the concentration of ethyl alcohol found, 1.044 2.017 1
expressed i n weight per cent., LE is obtained by 0.975 2.37 12

La == [TA - (AftVB x 18.02)/100~ L: 1000 0.983 2.34 4.08 SB 0.0 120

~ - 1 I

11 0281 ,2043 1.592 o n 1 0

VB X 4b.07 1.556 0.176 2
(4 1.203 1.904 i

where VB is the volume of the sample before distillation, 1.181 2.00 ’2

arid Rh the weight of the distillate. The accuracy of the 1.156 2.13 3.89 91 (1 0 ao
method varies from 1to 4% depeiiding mainly on the com- 1.156 2 . 1 3 120
pleteness and the speed of separation of the aIcohol and on 111 ,0053 ,2132 1 . 4 7 0 ti I 0
the determination of Ct. 1.09 1.78 170
The distillation carried out for the determination of al- 1.05 2.05 4.0 97 o 400
cohol ( L E )is continued until all the unreacted monomer is
removed; the monomer which distils over can be trans- The rates of these three reactions (Fig. 1)
formed into pure silica by reacting with an excess of very were too fast to enable determinations of in-
concentrated hydrochloric acid solution arid its amount stantaneous values of LH,and consequently only
he determined hy drying and weighing the silica. If m is
the weight of silica, the numher of moles of unreacted the final values are tabulated. The reactions a t
ethyl silicate per mole of ethyl silicate initially used is lower acidity were carried out with dilute hydro-
equivalent to: 1000m/G0.06 VgStp. chloric acid or distilled water. The results are
All of the experiments reported as acid-catalyzed hy- summarized in Table I1 and Fig. 1.
drolyses were carried out with hydrochloric acid but tests
showed that with other strong acids similar rewlts were TABLE I1
obtained. Of weak acids much larger quantities are needed
to obtain comparable speeds of hydrolyqis. The result5 HYDROLYSIS
OF ETHYLSILICATE WITH DILUTE HYDRO-
will be published in a later paper CHLORIC ACID SOLUTIONS
The reaction between ethyl silicate .tnd concentrated St,.
hydrochloric acid (12 N ) was noted to be very exothermic. ethyl
The acid solution introduced drop by drop dissolves readily $111- Mt
in pure ethyl silicate and gel formation occurs immediately. HCI, catr. water, a, Time,
Expt. m./l m /l. In /I LI LE C hours
In the range of moderate and low acidity the reaction iz
I V O.DOi6 0.2095 1.207 0 0 0 0
slower, samples can he removed during the reaction and
1.197 0 . 0 4 7 0:13
instantaneous values, of thr alcohol produced can t x 1.61 103 l:oo
obtained.
0.880 1 . 0 7 1:46
For the hydroly5is in h a w media, sodium hydroxide,
0.935 1.29 2.65 102 2:40
ammonium hydroxide and pyridine have been used at ‘1
0.852 1.69 3.50 102 5:20
iiumher of concentrationq. Dioxane, methyl and ethyl
0.851 1.69 3.B3 170:OO
:tlcohol served in all Cases ‘is mutual solvents. v .000022 ,2182 1.249 0 0
As pointed out above, though the hydrolysis and de-
1.225 0.18 24 :00
hydration reactions occur side by side, the kinetics of the
1.220 .20 0.4 170:00
hydrolysis can be studied separately by determining L A
and the variation of the total water content, so that the VI Distil. Hz0 ,220 1.355 0 0
1.34 ,068 18:OO
contribution due to dehydration during the reaction can 1.82 ,159 44:oo
he deducted. The order of the reaction can then be de- 1.25 ,310 196:OO
Lermined by comparing the ratios ~ i ’ / 7 1 of the respective
times 7 required to complete dflerent experiments t o the The results of both tables indicate that in acid
same degree ‘‘1.” These experiments are carried out at
constant HC1 concentrations and with different over-all catalyzed reactions the hydrolysis step is rate
concentrations of equivalent amounts of water and ethyl determining, except a t very high HC1 concentra-
bilirate tions where dehydration lags somewhat behind.
Experimental Results I n experiments (VII) and (VIII), Table 111,
The rate and extent of the hydrolysis reaction the concentration of hydrochloric acid was
nec., 1950 HYDROLYSIS SILICATE
(IF ETHYL 5707

the number of moles of water per liter consumed

in hydrolysis, 2 x , is given by

I
< 1.5
4 Taking this into account in cases of incomplete
1.0
dehydration, the plots of log - versus
0.5
ast, -x
time for all experiments performed give straight
0 lines (Fig. 2). From the slopes of the lines the
0 100 200 300 400 rate constants are calculated. As regards the
Time in minutes. variations of reaction rate with catalyst, Table
Fig. 1.-Acid-catalyzed hydrolysis of ethyl silicate in IV shows the effect of acid concentration.
dioxane. LT is the total loss of water in moles per mole TABLEIV
of ethyl silicate for several acid 3)"( concentrations.
VARIATION OF THE RATE OF THE REACTION
WITH THE
chosen within the range where complete dehydra- ACID CONCENTRATION
tion occurs (0.0037 mole acid per liter). The HCl concn., Rate K,
mole/l. 1. mole-' sec. - 1
ratio of the initial concentrations of the reactants 0.063 5.72 x 10-3
in the two experiments was 2.02: 1. The average ,0281 2.86 x 10-3
value of ~ j ' / ~ was
i found to be 2.04, indicating ,0053 1.80 x 10-4
that the reaction is second order. ,0016 1.04 x 10-4
TABLEI11 .000022 5.16 x 10-6
OF ETHYL
HYDROLYSIS SILICATE-ORDER
OF THE REACTION
Distill. H20 4.12 x 10-6
Ethyl silicate By plotting log K versus log [HCl] the following
1Stqs Mtm
HCI, Sb, equiv. water, Order of reaction linear relationship is obtained (Fig. 3).
Expt. mJ1. m./l. p. m./L (relative times)
log K = 1.0 log [HCl] f log 0.051 (9)
VI1 0.00368 1.248 2.496 2.34 Ti
VI11 .0037 0.622 1.245 1.16 ~ i =
' 2.04~i The rate is thus directly proportional to the
IX ,0034 0.624 1.248 2.32 r0.b" = 0.258~0.1 [acid] and the rate constant k expressed in liter
= 0.126 70.6 mole-' sec.-l [acid] -l is 0.051 a t 20' in dioxane
as a mutual solvent.
Experiment (IX), carried out with an excess
of water shows an increase of rate, appropriate
to a second order reaction with respect to the -1 -
i

water and ethyl silicate, and follows the usual

expression
- dx/dt = K(Mb - ~ ) ( 2 S t -, X ) (6)
which by integration gives

If the extent a of dehydration is less than unity,

-7 I , I

-6 -5 -4 -3 -2 -1
log [ H C l ] .
1.5
h Fig. 3.-Determination of absolute rate constant (k) of
x
I acid-catalyzed hydrolysis of ethyl silicate.

29 1.0 Concerning the activation energy, a value of

\
c 6.S kcal./mole has been obtained by comparing
c1
v the rates of reactions carried out in dioxane a t
20, 45.3 and 54' (Fig. 4). It was found that
-2 0.5 the increase of temperature from 20 to 45.5'
1 increases the rate of hydrolysis ten-fold while
01
the extent rises from 89 to 930/,; however,
a t higher temperatures (T2 50') the dehydra-
I 1

0 3 6 9 13 tion reaction is no longer complete.

Time in hours. No marked influence of the amount of silicate
Fig. 2.-Rate of reaction of acid-catalyzed hydrolysis on the nature of the hydrolysis reaction was ob-
of ethyl silicate in dioxane for several acid concentrations. served in several experiments in which ethyl and
 R. \ITLION, A. LOEBELAND F. EXRICH Vol. 72

same time a stronger dependence on the nature of

the solvent is observed in Table VII.
TABLE VI1
HYDROLYSIS
OF ETHYL IN BASICMEDIA
SILICATE
St,,
Concn. ethyl .Ut,
of NaOH, silicate, water, Time,
Solvent m,/l, m.A. m.h LT LE min.
Methyl alcohol 0.0347 0.235 1.06 0 0 11
0.976 0.11 .3
,768 0 . 9 9 3
. .i97 1 . 7 2 2;<
, .i.i7 1.89 4.0 5.3
.;,20 2.04 4.0 600
0 1 3 3 4 ,018: 1.2: z.44 0 0 0
2 "9 0.12 4
Time in hours. 2.12 ,266 21
Fig. 4.-Activation energy of acid-catalyzed hydrolysis: 1.51 ,505 135
curve 1, T = 20 0 " ; curve 2, T = 45.3'. 1 io .59 220
1.13 tY7 10%
methyl alcohol were used. In ethyl alcohol the Hrs.
extent of hydrolysis (la) decreases a t acid concen- .I)l6R 0.2305 0 . 9 8 5 0 0
trations smaller than 0.003 mole/liter. Table .89 0.41 2:25
,835 .65 23: 25
V shows that in experiment XI the extent has 7-
. , I .93 47:OO
fallen off to =LT/~). ,782 1.14 2.2 119:oo
,035 .236 2.0 0 0 0
TABLB \ 0.75 1.2 3:oo
HYDROLYSIS OH ETHYL SILICATE IN ETHYL ALCOHOL ,0145 ,2144 1.62 0 0 0
Ethyl 0.35 0 . 5 2 24o:oo
HCI. silicate, Water, Time, ,0083 ,236 1.92 0 0 0
Expt. m./l. m./l, ni./l. LT hr.
0.3 0.5 300:oo
S 0.0029 0.168 1.594 0 0 ,00391 ,236 1.67 0 0 0
1.38 1.7 la .a 0.5 3oo:oo
,000205 ,236 1.421 0 0 0
1.23 2 05 44
0 0 3oo:oo
XI ,0022 --
, i:,s
1.23
2.15
2 05
0
160
0
1 56
(NHrOH)
,220 0 8.40
,855
0
.01
0
0.10
1.08 I 81 8 ,698 .75 24.00
144 227 1 028 0 0
1 08 1 83 500 (NHIOH) 1 002 .I14 22: 30
0 9G ,365 147 :00
Reactions carried out in methyl alcohol (Table
VI) give an average value of the rate constant of The hydrolysis as well as the dehydration of low
the hydrolysis almost identical with the value concentrations of ethyl silicate in methyl alcohol
found for runs in dioxane, and agree well with with large amounts of hydroxyl ion catalyst is
runs in ethyl alcohol. This would indicate that fast and complete. But as soon as the concen-
re-esterification docs not play an important tration of the monomer is increased, the polymer
part. For the silicates of higher alcohols the does not stay in solution any longer but precipi-
mass action of the alcohol used as solvent is a tates, giving a white turbid appearance to the
little more marked, but still of minor importance. reaction mixture. This same phenomenon occurs
The value there of the constant depends mainly in ethyl alcohol, where the reaction is slower,
on the nature of the alkoxy groups attached to with smaller concentrations of ethyl silicate.
the silicon atom." Reactions carried out in dioxane, which exhibit
TABLE bI the slowest observed speed, a t very high concen-
trations of sodium hydroxide result in the forma-
HYDROLYSIS OF ETHYLSILICATE IN METHYLALCOHOL
tion of two liquid phases. The lower layer con-
k,
rate sists mainly of an aqueous solution of sodium
constant hydroxide and of partially hydrolyzed polymer,
Hydro- St,, liter,
chlotic ethyl Mb, mole - 1 the upper one contains the dioxane and the re-
acid, silicate, water, Rate K, liter sec - 1
Expt. m./l m/I. mp. rnole-lsec.-l HCl-1 maining monomer. Thus the growing polymer,
XI1 0 0034 0.624 2.32 0.20 X 0.059 having a silicate structure, collects most of the
XI11 ,0016 ,210 1.207 . l o 4 X ,065 water of hydration and coalesces into a liquid pre-
SIV .0037 .624 1.16 .12 X .042 cipitate. At lower concentrations of base than
about m. no reaction occurs. Ammonium
As in acid media, the hydrolysis of ethyl silicate hydroxide and pyridine produce a measurable
in basic solution is a function of the concentration speed of the reaction only if they are present in
of the catalyst, in this case of the base. At the large concentrations. In the case of ammonium
Dec., 1950 OF ETHYL
HYDROLYSIS SILICATE 5709

hydroxide-catalyzed reactions] flocculation of the

polymer formed occurs almost immediately.
The rate constant in dilute solutions of ethyi
silicate can be represented by the first order
relation
h. -0.5
x
ko
2.303 log (2Sh - X) = -Kt +C (10) I I \\
K being the rate of the reaction. For very dilute
solutions, the values of K vary proportionally with
the concentration of the base (Fig. 5 ) . In the
particular example given in Table VIII, the data -0.9 1 i1
TABLEVI11 I \
i
sto,
Rate
K X IO’* Rate constant, k,
-1.1 L.-t ( i - - I - - L 2

Ethyl silicate, liter mole-1 liter mole-’ 0 10 20 30 40

NaOH, mJ1. m./l. sec. -1 sec. -1 [OH] -1
Time in minutes.
0.0347 0.235 1.67 0.047
Fig. 5.-Rate of reaction of alkaline-catalyzed hydrolysis
.0180 .161 0.70 .039
of ethyl silicate in methyl alcohol: curve 1, ethyl silicate
are from hydrolysis in methyl alcohol at a tem- 0.161 mlliter; NaOH 0.018 mlliter; curve 2, ethyl sili-
perature of 20.0’ yielding values for k which are cate 0.235 mlliter; NaOH 0.0347 mlliter.
very close to those found for acid-catalyzed but becomes proportional to “i.” In experiment
reactions. XX the reaction medium clouds immediately
The order of the reaction was again deter- and the reaction is very slow. The lack of con-
mined by comparing the times required to com- stancy of ri’/ri may be taken as a sign that the
plete definite degrees, “2’ of the hydrolysis of reaction does not proceed any longer in true
very dilute solutions of ethyl silicate in methyl solution but changes into a heterogeneous re-
alcohol. The data are found in Table IX. A action. Similarly, the part played by the base
large increase of the Concentration of water does becomes more complex for higher ethyl silicate
not affect the rate of the reaction. The experi- concentrations (Table XI). By changing only
TABLE
IX TABLESI
HYDROLYSISOF ETHYL SILICATE IN BASIC MEDIA. Sb. -bft!
KINETICSOF THE REACTION NaOH, ethyl silicate, water,
mJ1. m./l. m./l. Time, n
s,. 0.00927 0,612 1.19 71
ethyl Mto,
NaOH, silicate, water, Order of .0189 .612 1.24 71’ = 3.957i
Expt. m.A. m./l. m./l. reaction
.0370 ,612 1.224 TiU = 11.0 Ti
xv 0.0180 0.161 0.386 Ti

XVI ,0185 ,312 0.644 ri’ = 1.06 ~i the catalyst concentration, the times, 73, required
XVII .01805 ,312 1.26 Ti“ = 0.98~L to complete definite degrees of reaction, “z‘,” are
found to form a well defined ratio, independent
ments, therefore, show the reaction to be first of “i.” This ratio is not proportional to the
order with respect to the silicate concentration. catalyst concentration but follows the empirical
In the case of more concentrated solutions of relation (15) deduced from the data of Table X I .
ethyl silicate] the rate of the reaction decreases
very rapidly (Table X), and the ratios Ti’/Ti -K= - - 3.1[Base] 2.1
K‘ [Base]’
are not constant but vary with “i,” the degree
Concerning the reaction products, those from
TABLE X acid or alkaline solution are widely different and
HYDROLYSIS O F ETHYLSILICATE I N BASICMEDIA in each case again the concentration of catalyst
st0s or of silicate creates conditions for rather different
ethyl Mq, condensation products.6
“OH, silicate, water,
I3xpt. m /1. m./L m./l. Time, si In strongly acid solutions, where the hydrolysis
XVIII 0.0185 0.312 0.644 Ti reaction (la) is fast and complete, and the extent
XIX ,0189 0.936 1.85 i <0.3,71’= Ti of dehydration increases from 86 to 100% as the
i > 0.3, Ti’ = 6 . 8 ~ i acid concentration decreases] no unreacted mono-
XX .0185 1.25 2.45 Ti” =a 12.871 mer remains, while the polymer in solution shows
no tendency to gel formation. The polymer
of completion. This indicates that the reactions molecules do not contain unreacted ethoxy groups
do not any longer follow a simple order but but some hydroxy groups remain. The average
become complicated by secondary reactions. I n formula of the polymer under the conditions
experiment XIX, e. g., the reaction follows a of experiment (I) shows four remaining OH
first order rate up to 3/10 of its completion, at (6) R. Aelion, A. Loebel and F. Eirich, Rec. Tuau. Chim., T B B , 61
which time the reaction medium becomes turbid (19M)), also appears in “Proc. Internat. Coll. Macromolecules,”
and the ratio q’/q does not remain constant Amsterdam, 1949.
 groups per seven Si atonis. Under the condition5 TABLE XI1
of experiment (11)only two undehydrated hydrox! x-KA\ I)AlA PROM A FILM OF POLYETHOXYSILOXANE
groups remain for seven Si atoms. The lack of SPREAD 0 % A MERCURY SURFACE
reaction of these remaining groups may be due to Distances Line J,inr
an enveloping action by the rapid growth of the in J, breadth intensity
polymer chains and to the sinal1 probability of :I Vrlu I tnri,il iwasuremrrits
reactive collision between OH groups situated 111 RrOdtl Strong
the interior of large three dimensional iiioleculei. Hroad Mediurii
I n reaction (111) the hydrolysis has beconre sui- Sharp Faint
ficiently slow t u permit a practically complete BrOdd Faint
il ehydration. Sharp Weak
.It all lower acid concentrations dehydr,ttlon Sharp Faint
remains complete but hydrolysis become\ in- Sharp l'ery faint
creasingly slow and incomplete, dropping to an Sharp L.ery faint
extent of 159, for neutral solutions. These
extents are obviously influenced by equilibria
which vary with the concentration of the acid. Wrolid Strong
-it the same time the hydroxy groups formed are lirontl Mediurii
competitively dehydrated without dependence Sharl) Medium
upon the HCI concentration. The dry polymer, Broad Weak
even for incomplete reactions, is always an in Sharp Medium
fusible and insoluble white powder similar t o Sharp Strong
silica. A t very low acid concentrations the Sharp Very faint
miount of unreacted inonomer I S greatly iw Sharp Very faint
creased arid the reaction mixture resembles in goes Simultaneously with the formation of the
molecular weight distribution that of <i vinyl precipitate the reaction velocity falls off but does
polymerization a t a small degree of conversion, not go to zero. This proves that the surface
The polymer is again similar to silica, but now retains the ability to grow for some time provided
rather unstable in solution and gels rapidly. it stays i n contact with the solution. Together
In generd, the rate of gelation depends on pH, with the fact mentioned earlier of the phase
concentration of unreacted ~noiionier anti con- separation in dioxane, the complications found
centration of polyiner. The higher the concen- during alkaline hydrolysis may be attributed to
tration of polymer, the faster gelation occurs. the formation of a polysilicate which becomes
X-Ray analysisaa of samples removed during rapidly insoluble with increasing molecular weight.
hydrolysis after various times show the same Thus, the formation of silicate may be considered
hind of amorphous patterns, 2haracterized by as competing with the dehydration, and as inter-
two broad rings at 11.6 and 4.1 A The polymer fering further by using up catalyst.
ceases to be soluble and gel formation occurs
rapidly whenever the hydrogen ion concentration Discussion
i n the reaction mediuin drops below a value oi
0.002 mole, liter. -kt higher concentration of particular from our kineticIn trying to arrive from our results, and in
HCI, gelation wa5 not observed. The degree of data, a t a possible
(*on\ersiori of the iiioiimier x dries widely with
reaction mechanism, a number of difficulties are
the conditioris oi the reaction and its value
encountered. For the acid catalysis the over-
plays a very iniportant part in the stability of intoreaction all is third order. This can be fitted
the solutions, since gelation can be prevented by order reactions,schemes the kinetic of two subsequent second
namely, either the formation of
;Idding unreacted ethyl silicate. This solubilizing
action of the monomer can b c explained by the the conjugate acid from the addition of a proton
to the ester oxygen and the subsequent hydrolysis
siimlanty of structure (as witli it. inolccule of water, or the formation of an
diffraction) o f irozen ethyl si oxonium ion followed by its reaction with the
structure of oriented films obtained by evapor'i- ester. The difficulty with both mechanisms lies
tion of a polymeriied solution on I inerciirv sur- in the fact that they are based on a hydrogen ion
t a w (Table XIl:. catalysis, which is by no iiieaiis assured. General
111contrast to tlie 3 c d reactioii, 11ifili illoiloilier :icid catalysis may be tndicilted since, despite the
COll~eIltrdtl~ll~ 111 d1!-.cLllllC' 5OlUtlOll I I l O i SOlU practical dbseilCt- of curiductivity, tlie wine rates
h i k e the polynirr but aid precipitation, vithri are observed in dioxane as iii alcohols with their
by non-solvent dction, or by increasing the 10 times higher dielectric constants. The dif-
polymer concentration. This shows t€iatthe poly ficulty with general acid catalysis again rests
mer from alkdiiir 111drolysis itlust Iiitvt: rather with the observatioii that replaceiiierit of hydro-
tliiferetit structure, least as fat- '1s it5 surf,tc.e chloric acid by equivalent amounts of weaker
a' R e niih t o than!. M r I R l i l l w for h i acids cuts the rate parallel with the strength of the
x i c l ; furtlicr (to be pnhlisherl later) hydrolvsis in
Dec., 1950 HYDROLYSIS
O F ETHYL
SILICATE 5711

glacial aceticacid as solvent is not particularly fast. oxonium salt is likely in analogy, e. g., to its
This leads one to the conclusion that an acid formation in sulfur dioxide as shown by Bagster
catalysis must be involved (I, 11) of which the and Cooling.’ If this salt does in fact exist and
essential feature consists in the participation of is regenerated by a fast reaction, we would expect
the acid itself in the transition state, whereby. a reaction I to outstrip I1 in view of its bimolecular
proton is donated toward the formation of the nature and of the spreading of the activation
alcohol. The apparent contradiction between entropy over two reactions.
the required ease of acid dissociation and the In alkaline solution too we must not expect
practical absence of such dissociation in the ordinary hydrolysis of carboxyesters (which pro-
solvent would thereby be removed. ceeds by attack of the hydroxyl ion on the
As to the action of water, the acid has either carbonyl oxygen), nor can the pattern of carbon
to be present in the form of its oxonium salt, or orthoesters be followed which shows no catalysis
attachment of a molecule of water would have by bases.* The observations are rather com-
to take place as a separate event rendering the patible with nucleophilic displacement whereby
reaction truly termolecular. There are some the direct participation of water is excluded by
points in favor of the former scheme, such as the kinetics. In contrast to the acid catalysis
the difference in the kinetics of acid and base we find further a large drop in reaction velocity
catalysis and the rate dependence on the size of with decreasing dielectric constant of the medium,
the alcohol, but they are not conclusive. We i. e., the participation of ions is indicated. We
feel, however, that our findings are sufficiently suggest therefore the simple displacement mech-
akin to those of C. G. Swain, et aL16in the case anism
of the hydrolysis of triphenylsilyl fluoride, to \ ( - ) I (-)
apply their reaction scheme with minor modifica- (111) OH + -SiOR +H0:SiOR +HOLi + 6 R
tions. / /\ A
Swain postulates that rather than proceeding OR + HOH +HOR + OH
Via the siliconium ion, a molecule of water is addedWe have here retained the expanded volume shell
to form the quinquevalent coordination com- for the silicon, in accordance with Swains and
pound with subsequent charge separation and with Sidgwicklg to indicate the absence of sili-
ionization into the conjugated silo1 acid and fluo- conium ion in the transition state. In view of the
rine ion. We see from the extremely slow and in- ionic character of the displacement and of the
complete reaction of our silicates a t PH 7 and lesser complexity of the transition state, we think
from the influence of the molecular weight, that the decomposition of the latter more likely to be
the uncatalyzed complex formation must be the rate determining step than in the case of the
very minor; further, the resulting ethylate ion acid catalysis.
is much less stable than the halogen ion. Bearing Although the hydrolysis rates could be found
the practical absence of ionization in mind, we independently by our method, the rate of dehydra-
propose the following modified schemes of Swain’s tion had to remain undetermined because of their
dimolecular, or termolecular, reactions, respec- large speed. We wish only to remark that, in
tively, (B = base) view of the reluctance of silicon to form double
(I) - SiOR + HaOB +
\ bonds, the reaction has to occur intermolecularly,
/ and that the speed, as pointed out by Swain6
HOHHB and by Sauer,’O may also be considered a con-
1 Si OR + \ - SiOH + HOR + HE sequence of the power of silicon to add a fifth
/ / covalent bond. Further aspects of the con-
+
\
(11) HzO 7SiOR +
HE -+- densation reaction will be found in our earlier
paper,6 where we also pointed out that all re-
HE action constants, being functions of the degree
1 .. + \
Hz0:SiOR -- SiOH +
HOR + HE of polymerization of the silicates as well as of the
/\ / molecular weights of the alkoxy groups, must be
The acid is here written undissociated and in the considered as average values. Their remarkable
most general way, including conjugate acids constancy is due to the fact that the individual
formed with the solvent. All charge separations reaction times from monomer to polymer are
are assumed to occur in the transition state, after insignificant compared with the shortest time
the molecules of water and acid have been at- intervals which can be measured during the course
tached by means of the unshared pairs of electrons. of the total reaction.
The rate determining step may be the regrouping Acknowledgment.- We gratefully acknowledge
of bonds or the attachment to the hydrophobic the financial help and assistance with materials
silicate molecule. The effective rate in either (7) L. S. Bagster and G. Cooling, J . Cham. Soc., 117, 693 (1920).
case would depend on the product [acid] X ( 8 ) L. P. Hammett, “Physical Organic Chemistry,” McGraw-Hill
Co., lac., New York, N. Y.,1940, p. 219.
[HzO]. However, the existence of a preformed Book (9) N. V. Sidgwick, “Electron Theory of Valency,” Oxford
(6) C. G . Swain, R. M. Esteve and R. H. Jones, THIS
JOURIAL, University Press, Oxford, 1927, p. 159.
17, 965 (1948). (10) R. 0. Sauer, THISJOURNAL, 66, 1707 (1944).
5712 F. RILEY
RICHARD
from the Union Carbide and Carbon Corp., and
we wish to thank Dr. H. Mark for his encourage-
ment of this work.
Summary
A kinetic study of acid- and alkaline-catalyzed
hydrolysis of ethyl silicate reveals that the acid
catalyzed hydrolysis is second order with respect
to water and ethyl silicate, and that the rate is
further directly proportional to the acid con-
centration used. The basic hydrolysis is first
order with respect to ethyl silicate, proportional
to the concentration of base and indenendent of
the amount of water present. It is, moreover,
[CONTRIBUTION FROM THE n I v I S I O N OF PHARMACOLOGY AND TOXICOLOGY,UNIVERSITY OF ROCHESTER, SCHOOL OF
MEDICINEAND DENTISTRY]
The Isolation and Characterization of Two Metabolic Products of Myanesin
(3-(o-TOlO~y)-l,Z-propanediol)l p 2
13%’RICHARD
F.
Graves, Elliott and Bradley4 have reported the
isolation of P-(o-to1oxy)-lactic acid, m. p. 146-147’,
from the urine of man and rabbits which had
received myanesin. Working independently, we
isolated the same compound as well as a second
metabolic acid, m. p. 16S0, which is responsible
for the positive Ehrlich diazo reaction6 given by
metabolic urines of man and experimental animals
which have received the drug. The chromo-
genic substance, m. p. 16S0, has now been identi-
fied as P- (2-methyl-4-hydroxyphenoxy)-lactic acid
on the basis of its chemical and physical properties
and by degradation to toluhydroquinone. A
preliminary communication regarding this work
has appeared.6
It is the purpose of the present paper to present
(a) a method of isolation of these compounds,
(b) pertinent data which have now led to a de-
termination of the structure of the chromogen
and (c) some information concerning the extent
of metabolic conversion and excretion of the two
substances.
Initially, we were successful in obtaining small
quantities of the two compounds from human
urine by a lengthy process involving adsorption
on Darco G-60, elution with acetone and con-
centration of the carbon eluate followed by
solvent and chromatographic fractionation.
(1) Presented in part before the Division of Biological Chemistry
of the American Chemical Society at the San Francisco, California,
Meeting, 1949.
(2) Aided by a grant from The National Foundation for Infantile
Paralysis, Inc.
(3) University of California Atomic Energy Project, Post Office
Box 81, Beverly Hills, California.
(4) Graves, Elliott and Bradley, N-rc. 169, 267 (1948).
( 5 ) Hawk and Bergeim, “Practice1 Physiological Chemistry.”
11th ed., P Blakiston’s Son aud Co , Inc , Philadelphia, P a , 1937,
p. 669.
‘6) Riley arid Berger, A ; c A B~orhrnt, 20, 15P (1040)
Vol. 72
complicated by silicate formation a t higher con-
centrations,
The value of the activation energy of acid
hydrolysis has been determined. The nature of
the polymer formed and its stability in solution
has been investigated. For the acid hydrolysis
a mechanism intermediate between general acid
and oxonium ion catalysis has to be postulated.
A dimolecular as well as a termolecular reaction
path appear possible. The base catalysis in-
volves a nucleophilic displacement of the alkoxy
group by the hydroxyl ion.
BROOKLYN,KEW YORE RECEIVED
DECEMBER
21, 1949
RILEY3
More recently, the counter-current distribution
technique has been employed as a convenient
means of isolation of both compounds, and has
provided sufiicient amounts of the chromogen
to simplify degradation studies.
It was readily demonstrated, on the basis of
carbon and hydrogen analyses, the neutral equiv-
alent and ultraviolet absorption spectra, that
the first metabolic product was &(o-to1oxy)-
lactic acid.6 That the chromogen was meta-
bolically derived from fi-(o-toloxy)-lactic acid
was indicated since urine of rats which had re-
ceived this compound showed the characteristic
chromogenicity of urine of animals which had
received myanesin. Combustion analyses on
several different samples indicated that the
empirical formula of the chromogen differed
from that of the first metabolite by the addition
of one oxygen atom. The second compound was,
like the first, optically inactive. Electronietric
titration showed the second compound to be a
relatively strong acid, pK 3.5.
Front these observations we concluded that
the chromogen must be one of the four isomeric
8- (hydroxy-2-methyl phenoxy)-lactic acids.
Earlier observations that o-cresol, phenetol and
diphenyl ether are metabolized to toluhydro-
quinone,’ the glucuronide of 9-ethoxyphenol
and p-hydroxydiphenyl ether,g respectively, led
us to suspect that the chromogen was, in fact,
p- (2-methyl-4-hydroxy phenoxy)-lactic acid,
formed from 0-(0-to1oxy)-lactic acid by an
analogous oxidation of the benzene ring para
to the phenyl ether linkage. In further support
(7) Preusse, 2. phydol. Chsm., 6, 67 (1881).
(8) Kossel, ibid , 4, 296 (1880); ‘7,292 (1883); Lehmann, ibid., 13,
181 (1889)
‘9) Strcmil, J Endocirnol , 2, 53 ( 1 9 i O )