Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
8~
Large numbers of azeotropes are known; for
example, the azeotrope of water and perchloroethylene
and the azeotrope of water and HBr are known. The
ternary azeotropic mixture of water, HBr and perchloro-
~ ethylene, however, is not known. Moreover, the tradi-
tional method of removing HBr from a reaction mixture
is neutralization with a base and removal of the salt
formed by washing. Products containing undesirably
large quantities of inorganic salts are obtained by this
technique.
Removal of minor amounts of HCl from halo-
genated organic compounds utilizing epoxy compounds as
scavengers is also known. Thus, in United States Patent
3,303,107 it is disclosed that epoxy compounds con-
taining at least one oxirane group such as styrene oxide,
ethylene oxide, epichlorohydrin, and the like, may be
used as an HCl scavenger to purify vinyl chloride using
a distillation step to then remove the epoxy compound.
In the preparation of brominated pentaerythritols,
wherein HBr and pentaerythritol are reacted in a hydro-
carbon or halogenated organic solvent, using an acid
catalyst produces a brominated pentaerythritol product
useful as a fire retardant in polyesters, polyurethane
foam and the like. However, any trace amounts of HBr
left in the product will cause undesirable color in the
final polymer product and undue corrosion in the "cooking"
of the polymer product when the fire retardant is incor-
porated, for example, in a polyester resin.
16,591-F ~ -1-
,, ~
This undesirable effect is substantially avoided
by the present invention, said invention comprising a
method for removing HBr from a reaction mixture formed by
the reaction of pentaerythritol and HBr in a perchloro-
ethylene solvent, said reaction mixture including brominated
pentaerythritols, HBr, water and perchloroethylene, said
method comprising (a) separating the bulk of the HBr, the
water and the perchloroethylene from the brominated penta-
erythritols by a~eotropic distillation and (b) reacting
the residual HBr remaining with the brominated penta-
erythritols with an epoxy compound having an oxirane or
oxetane group, thereby producing brominated pentaeryth-
ritols free of HBr, wherein said epoxy compounds having
oxetane groups are defined as including trimethylene oxide
and any oxetane having 3,3-bis(haloalkyl) substitution or
3,3-bis(alkyl) subs~itution wherein the halogen is chlorine
or bromine and the alkyl moiety contains from 1 to 4 carbon
atoms. The novel use of either class of compounds in the
present invention effectively removes the color and acid
from the product. By "epoxy compound having an oxirane
group" is meant a compound containing at least one 1,2-
-epoxy or oxirane group. Suitable oxides
16,591-F -la-
.. ~,
3'Z~;~
for use in the instant invention include, for example, ethylene oxide, propy-
lene oxide and other alkylene oxides; 1,2-epoxy butane and the like; epi-
chlorohydrin, epibromohydrin, lJ2-epoxy-5-bromohexane, and the like; and
other compounds containing the oxirane group such as triphenyl ethylene
epoxide, epoxidi~ed soybean oil and the like. Preferred epoxides include
epichlorohydrin, epibromohydrin, styrene oxide and alkylene oxides.
When the above epoxides are used to scavenge the residual HBr in
the reaction product in the practice of the present process, a neutralization
reaction takes place whereby the epoxide and the HBr react to form a brominat-
ed product. Thus~ where epichlorohydrin is used to scavenge the HBr, the HBr
opens the oxirane ring to form the 1,3-halogenated propanol-2. This product
and similar products formed when other epoxides are used as HBr scavengers in
the instant process need not be removed from the final brominated pentaery-
thritol product in most applications of that product. Brominated pentaery-
thritols have their greatest utility as fire retardant intermediates and may
be used directly for such purposes and also as reaction components in polymer
systems. In some of these latter systems, as will be delineated below, the
use of these epoxides as HBr scavengers is less desirable because the reaction
product of the epoxide and the HBr is volatile and likely to dehydrohalogenate
when the fire retardant brominated pentaerythritols are cooked into poly-
esters and may also cause foam scorching in polyurethane foams. In such
applications, it is more desirable to use as the HBr scavenger, the oxetanes
of the present invention.
"Epoxy compounds having an oxetane group," as used herein, are
defined to mean trimethylene oxide, and any oxetane having 3,3-bis~haloalkyl)
substitution or 3,3-bis(alkyl) substitution, wherein the halo is chlorine or
bromine and the alkyl moiety contains from 1 to 4 carbons. Said substituted
oxetanes include 3,3-bis~bromomethyl)-oxetane, 3,3-bis(chloromethyl)oxetane,
3-methyl-3-ethyl oxetane, 3-chloromethyl-3-bromomethyl oxetane, 3,3-bis
(propyl)oxetane and the like. When such oxetanes react with IIBr, a haloneo-
-- 2 --
~5~3fA~i~3
pentyl alcohol is formed which is stable and high boiling. Consequently, as
will be seen below, discoloration, corrosion and foam scorching is eliminated
when the brominated pentaerythritol product is used as a fire retardant in
polymers. Preferred among these oxetanes are the halo-substituted ones as
they add fire retardancy due to the presence of the additional halo group or
groups. Especially preferred is 3,3-bis(bromomethyl)oxetane ~BBM0) because
it forms tribromoneopentyl alcohol when it reacts with HBr which, of course,
is one of the desired end products of the overall process. This reaction is
shown by the following equation:
CH2Br CH2
\ / \ CH2Br
/ \ / 0 + HBr -~ CH2Br-C-CH20H
CH2Br CH2 CH2Br
In this method, the reaction mixture containing HBr, water and
perchloroethylene is distilled in such a manner that the HBr, water and per-
chloroethylene form an azeotropic mixture which is found in the distillate.
The azeotropic mixture of the invention consists of three components,
HBr, water and perchloroethylene in the proportions of approximately 34:27:39
parts by weight, respectively. The azeotropic mixture has a boiling point
of 63-67C. at 175 mm. of Hg. The distillate received forms two layers -
one of which is essentially concentrated hydrobromic acid and the other of
which is perchloroethylene. This separation facilitates recycle of the per-
chloroethylene, if desired.
It is to be understood that the distillation can take place under
a wide range of pressures - under vacuum up to above atmospheric. Depending
on the economics of heat costs versus sophisticated vacuum equipment, the
pressure can range from 3 mm. to 15 p.s.i.g. The preferred range is from 25
mm. to atmospheric while especially preferred in this system is a range of
from about 50 mm. to about 300 mm. Hg.
The azeotropic removal of HBr in the described process is conven-
1~5~ 3
iently carried out to remove practically all of the HBr in the reaction mix-
ture. The removal of all the HBr by this technique, however, is virtually
impossible for there is a very small residual quantity of HBr that is most
difficult to remove by these techniques. Thus J the second step of the in-
vention is utilized for complete removal of the HBr. After the stripping
operation by the azeotropic distillation, the epoxide or oxetane is introduced
into the reaction product in small amounts sufficient to neutralize the free
}IBr remaining.
A special advantage of this invention is that the bulk of the
excess HBr in the reaction that is azeotropically distilled can be convenient-
ly recycled to the reaction as aqueous HBr. The minute J but harmful J amount
of HBr remaining in the reaction product is then convertedJ thus avoiding the
processing problems referred to above.
PentaerythritolJ as used hereinJ is defined to include pentaery-
thritol that has been partially halogen-substituted. ThusJ the process of
the invention is applicable to the preparation of brominated pentaerythritols
where one to three of the hydroxyls have been replaced by bromine.
Examples
Example 1 - Removal of HBr After the Bromination of Pentaerythritol with HBr
A three-liter reactor equipped with a condenser and pressure
regulator was charged with 350 ml. of perchloroethyleneJ 54 g. ~0.9 mole) of
acetic acid and 544 g. (4.0 moles) of pentaerythritol. Over a period of two
hours and 40 minutes at a temperature of 110 to 118C. and a pressure of 5
p.s.i.g., 842 g. ~10.4 moles) of HBr was added. After the addition, the
reactor was maintained at 113C. for an additional 30 minutes. An azeotropic
recycle head was placed on the reactor and the HBr, water and perchloroethylene
were distilled from the reaction mixture. The distillation was conducted at
175 mm. of Hg until the last 15 minutesJ when the final vacuum was brought
down to 50 mm. The pot temperature ranged from 67 to 102C. while a head
temperature ranged from 63 to 67C. The distillation was conducted over a
Z~;~
a time of 3 hours and 10 minutes. One hour and 55 minutes from the beginning
of the distillatio~, the aqueous layer containing water and HBr was isolated.
The aqueous layer had a volume of 158 ml., a density of 1.456 and was analyzed
to contain 45.4% by weight HBr. At the end of the distillation, the remaining
aqueous layer was isolated. This cut was found to have a volume of 57 ml., a
density of 1.634 and contained 59.2 weight percent of HBr. The product re-
covered weighed lO91 g. and had the following analysis by weight percent as
determined from gas-liquid chromatography: 10.8% monobromopentaerythritol,
74.1% dibromoneopentyl glycol, 9.74% tribromoneopentyl alcohol and 0.96% HBr.
The product as a melt had a light orange color and as a solid was cream color-
ed.
Example 2 - Removal of HBr by Azeotropic Distillation and Treatment with an
Expoxide
In the same manner as shown in Example 1, the reactor was charged
with 350 ml. of perchloroethylene, lO g. of acetic acid and 544 g. (4.0 moles)
of pentaerythritol. To this mixture was added 907 g. (11.2 moles) of HBr over
a period of 5 hours and 20 minutes at a temperature of 104 to 119C, and a
pressure of 5 p.s.i.g. The reaction was heated for an additional 30 minutes
at 114C. The HBr, water and perchloroethylene were removed by azeotropic
distilla~ion at 86 to 100C. and 50 mm. of Hg. To a melt of the orange
colored product maintained at 90C., 25 ml. of epichlorohydrin was added with
stirring and the product became a light amber color. Upon solidification, the
liquid was decanted and a white solid weighing 1046 g. was obtained having an
analysis by weight percent of: 14.0% monobromopentaerythritol, 74.1% dibromo-
neopentyl glycol, 5.3% tribromoneopentyl alcohol and 0.42% H2O. No HBr was
detected in the product. Thus, treatment with the epichlorohydrin not only
eliminates the HBr but also substantially improves the color of the final
product.
To compare the efficacy of the oxetanes and epoxides as scavengers,
similar brominated pentaerythritol products prepared by the process of Example
1, above, containing trace amounts of HBr, were analyzed for acidity and color
as shown in Examples 3-7 below.
Examples 3-7 - Removal oi Trace ~IBr by Treatment with Epoxides
and BBMO
Five test tubes were prepared of the brominated
pentaerythritol products from which the bulk o-f the HBr,
water and perchloroethylene had been azeotropically distilled
as in Example 1, above, by mel-ting 20 grams in each test tube
at 90C. and maintaining this temperature while the iree
acid was neutralized utilizing 0.2 cc. of the various
epoxides and BBMO. One scavenger was added to each o~ these
-test tubes with stirring and where the color did not disappear,
an additional 0.1 cc. was added to insure neutralization of
the free HBr present. Acidities were determined by titrating
to give a phenolphthalein end point with N/10 NaOH and
calculated as HBr. Light transmittance data were ob-tained
on a B. & L. Spectronic 20 (Registered Trademark~ and a
Gardner color calculated irom these data. The data -irom
these determinations are given in Table I, below.
~,591-~' -6-
1~58'~'~3
O a~ I ~ _ ~ O~ oo
Z ~ L~
~, h ¢
h O
h ~ h ¦ ~ _
O ~
H C~l 111 ~0
h
a~ ~ ~
~ a> ~ Ln ~ o ~ oo oo o o~ co O a) ~ ~ oo o
¢ ~ ~ ~t r~ co ~ 1~ oo ~ t~ 0~ t` 00 ~ ~ ~ 00
a~
~ a~,." ." .r,
.,1 h ~ ~
e O ~ ~ ~ O u~ ~ O ." ~ O u~ O u~ N cn 00 0
h 4~ ~ N ~ 1~ N ~D 1~ N ~ I~ 11~ 1` a) N It~ ~
o\
~ I
~4o o n o o u~ o o In o o n o o Ln
~ IJ'~ N IJ~ N Ln 11~ N n U'l N Lt) Ll~ N
H 3 ~ ~~ 11~ ~0 ~ Ir~ ~ ~ n ~ ~ In \1~ d' Lt) ~D
a~
a>
E~ h h ~ 1 --I o
tra~ ~ ~ ,, _ _ - o
o\~ ~ N Z O
~d
~10
C~ o\o
~¢h h ~ O N \
~1 :S N , _ _ o o
~D z.rl O O
'h ~ ~::
~> ~ .,1 c~ h
:~ X
'X ~
~,1 h ~ O O
~ h
¢ .S
U
.,1 p, h t~ O
h
X ~ d~
U~
-- 7 --
~QS~'~Z3
As shown above, oxiranes and oxetanes are equally effective in
removing the residual HBr from the reaction product. However, in some end
uses of the reaction products the oxetanes are preferred; for example, where
the brominated pentaerythritol is to be used as a reactive fire retardant in
slab stock polyurethane foam. Slab stock polyurethane foam is made in large
buns which evolve considerable heat in their formation. In such uses, the
oxetane is preferred as the reaction product of the oxetane in the HBr is
stable and high boiling. By contrast the oxirane-HBr reaction product is less
stable, more volatile and dehydro-halogenates more readily. However, it is
to be understood that the oxirane neutralized brominated pentaerythritol may
be used where smaller polyurethane foam parts are molded. In such cases,
the temperatures are less and the scorching effect of the oxirane-HBr reaction
product does not occur.
Similarly, both the oxirane and oxetane neutralized brominated
pentaerythritol may be used as fire retardants for polyesters. However, when
the polyester is cooked in a stainless steel reactor it is found that the
oxirane-HBr reaction product will corrode the reactor mOTe readily. In such
cases the oxetane is the preferred neutralizer. Where glass systems are
used to cook the polyesters, use of the oxirane is not a problem.
The above advantages of the oxetane in particular circumstances are
shown in the series of e,xamples, below.
Examples 8-15 - Comparison of Oxirane and Oxetane Neutralized Tribromoneopentyl
Alcohol in Polyurethane Foam
A brominated pentaerythritol was prepared and the bulk of the HBr
was removed along with the water and the perchloroethylene utilizing the first
step of the process of this invention, i.e., ternary azeotropic distillation.
This product contained about 97.8% tribromoneopentyl alcohol and about 2.2%
dibromoneopentyl glycol. Its acidity was in the range of 0.02 to 0.05%.
This material was divided into four equal samples. Sample A was not neutral-
ized, Sample B was neutralized with epichlorohydrin (20 meq./kg.), Sample C
was neutralized with 3,3-bis(chloromethyl)oxetane (20 meq./kg~), and Sample D
was neutralized with BBMO (20 meq./kg.). Each sample was then separately
dissolved in a polyol at a level of about 10% and then blended in the convent-
ional manner with catalysts, blowing agents and surfactants. The four samples
were then mixed with an isocyanate and foams A, B, C and D were prepared
corresponding to the above samples and cured at 150C. for 45 minutes.
After cure, visual inspection showed Sample A to be definitely
yellow in color while Samples B, C and D were near white.
The bottom one inch of each foam bun was removed and 1 1/2 inch
thick semi-circular slices from each bun were wrapped in aluminum foil and
placed in a 150C. circulating air oven for 16 hours as a heat aging test.
Visual inspection showed that Sample C was slightly lighter colored than
Sample D which in turn was much lighter and less colored than Samples B and
A. Sample B which was the foam prepared utilizing the brominated pentaery-
thritol neutralized with epichlorohydrin gave a foam which was almost as bad-
ly discolored as the nonneutralized Sample A.
A second series of three foams (Samples E, F and G) were prepared
as in the above formulation using the same concentration of the same brominat-
ed pentaerythritol product. Sample E was neutralized with epichlorohydrin,
Sample F was neutralized with BBMO and Sample G had all of the HBr removed
through distillation and recrystallization of the alcohol. The three samples
were again heat age tested as above and the order of discoloration of the
foams after 16 hours at 150C. was as follows:
E > F > G
Examples 15-21 - Comparison of Corrosive Effects of an Oxirane System with
an Oxetane System
A series of polyester cooks were carried out in a stainless steel
reactor in order to compare the effect of the neutralizing agent upon the
corrosion of the reactor. ~he relative amounts of corrosion were determined
by analyzing the polyester alkyd for iron - the higher the iron content of
the resin, the higher the corrosion rate. Seven runs were made Imder iden-
tical conditions, quantities and materials except that, as indicated in Table
II, below, batches A and C of the brominated pentaerythritols
had an initial iron content of 2 p.p.m. while batch B of the
brominated pentaerythritols had an initial iron content of
5 p.p.m. ~urther, batches A and B had their residual HBr
removed by reaction with epichlorohydrin (ECH) while batch C
utilized BBM0. The brominated pentaerythritol product com-
prised approximately 82% dibromoneopentyl glycol, 7% mono-
bromoneopentyl triol and 11% tribromoneopentyl alcohol.
In each of the seven runs, the reactor was charged
with 281 g. tl.9 m.) of o-phthalic anhydride and 186 g.
(1.9 m.) of maleic anhydride. The mixed anhydrides were
melted and brough-t to about 115C. under a nitrogen blanket.
In each run, equal amounts (1048 g.) of solid brominated
pentaerythritol product was added to the reactor in two
portions, two-thirds of the material being added in one
portion and 15 minutes later the last one-third was added.
The reaction temperature was raised to 185C. during the
space of about l/2 hour and maintained at this temperature
until the acid number indicated in Table II was reached.
This took about 5 1/2 hours. The nitrogen flow through
the reaction was held at about 450 cc./min. throughout the
cooking cycle. Approximately 55 ml. of water was collected
in the Dean-Stark tube.
The alkyd was cooled to about 150C. and 0.76 g.
of hydroquinone inhibitor added. After mixing, the mixture
was poured into a Teflon-lined tray to cool. ("Teflon" is
a Registered Trademark) The samples of the alkyd were then
analyzed for iron.
15,591-F ~ -lO-
lq~S~
Runs 15 and 21 are comparable in that both were
carried out in a clean reactor. In this connection it
should be noted that the brominated pentaerythritol product
neutralized with the oxetane (BBM0) showed less corrosion
in spite of the iact that it was cooked to a much lower
acid number than was the brominated pentaerythritol
neutraliæed with epichlorohydrin. Runs 16 through 20 are
comparable in that they were run sequentially without cleaning
the reactor between runs. Run 16 was run in the reactor after
Run 15 without cleaning the reactor, etc.; however, the
residual polymer remaining in the reactor was small -
less than 2% oi the charge in each case. Nevertheless,
16,591-~' -10~-
Z'~3
some iron carry-over occurred. The mean iron content of those resins cooked
using epichlorohydrin was about 38 p.p.m. while those using the oxetane was
about 25 p.p.m. Correlating these results and acid numbers, Table II shows
that the corrosion rate when epichlorohydrin was used was about 150% of the
corrosion rate that occurred when the brominated pentaerythritols were neu-
trali~ed using BBM0. Comparable results would be expected with any oxetane
having a 3,3-bis~haloalkyl) or a 3,3-bis(alkyl) substitution.
O h
~ O
h h
~ a~
z ¦ L" In n O ~P N ~
C~ N 1~ `.0 ~ ~ ~ 1`
~ N N N t~ N H
P1,~
h
U~ ~ I 00 N 00 ~
O ~ N ~ N t-~ ~t N _1
H~
Q,
E-~bl\
S~ ~
~d
H ~ h
h O ~ ~C X ~ ~
b~ ~ m ~ ~) C4 ~ E3
h
,~
O p~
h
N
O h
~,1 ~ . .
h ~ E3
3 ~ ~ ~ ~ ~ ~ ~ ~
o ~ ~
h ~ 3
o C4
h
~ P
ID 1 N
n u~ ~ ~ coo~ o _I o 1l
::~ ,~ l N a~ ¢
Vl 1~
- 12 -
` ~ ~)5~Z~3
In the same manner as shown for epichlorohydrin, other epoxides,
such as ethylene oxide, styrene oxide, propylene oxide and butylene oxide are
added to the reaction mixture after the azeotropic distillation to remove
the last traces of the HBr from the reaction mixtures. Also, in the same
manner as shown for 3,3-bis(bromomethyl)oxetane and 3,3-bis(chloromethyl)-
oxetane, other oxetanes, such as trimethylene oxide, 3-methyl-3-ethyl oxetane,
3-chloromethy].-3-bromomethyl oxetane, and the like are added to the reaction
mixture after the azeotropic distillation to remove the last traces of the
` HBr from the reaction mixtures. These oxetane-neutralized reaction products
are effectively used in the manufacture of polyurethane foams without causing
scorching and in the manufacture of alkyd resins in stainless steel reactors
without causing undue corrosion.
