Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
2nn7~26
The invention relates to a process of separating acids from
substrates contain1ng salts and carbohydrates, in which the substrates
are made to pass over res1ns adsorptively removing the acids from the
substrates and the aclds are subsequently separated from the resins
by e1ut10n.
Organ1c and inorganic ac1ds are used in many branches of industry:
1n appl1cat10ns 1n the fields of pharmaceut1cals and foodstuffs, but
also when used 1n pur1fying chemicals, these must be of part1cularly
h1gh qual1ty.
~ p to now, proposed pur1f1cation methods for ac~ds, ln addltion
to prec1p1tat10n as heavy metal salt or alkaline earth metal salt,
have been 11qu1d/11quid extract10n and the format10n of volat11e
der1vat1ves (for 1nstance methyl esters). All these methods are also
used on an 1ndustr1al scale for the process1ng of techn~cally 1mportant
u1ds such as lact1c ac1d, mal1c ac1d, succ1n1c ac1d, tartar1c ac1d,
c1tr1c ac1d, glucon1c ac1d and phosphoric ac1d. However, the processes
ment10ned have ser10us drawbacks. The pur1ficat10n methods provid1ng
or the ac1d to be 1solated v1a an 1nsoluble salt consume large amounts ;;
of aux111ary chem1cals. ln most cases, a spar1ng1y soluble calc1um
sa1t ~s formed and subsequent1y reacted with sulfur1c ac1d to form
the free ac1d and an equ1molar amount of gypsum. ln addit10n to the
j expense for 11me and su1fur1c ac1d, the gypsum formed must be
d1scarded. Pur~f1cat10n methods based on the d1stribut10n between
two mutua11y 1nsoluble phases and processes 1n wh1ch the acid 1s
reacted to a more volat11e derivat1ve operate w1th large amounts
of organ1c solvents const~tuting a safety and environmenta1 hazard.
The extraction of lactic acid is disclosed in EP O 159 585; DE-OS
':;'':,'
'~,
- 2 - :
.~
~ ,
~ æ{~7 ~L~2
32 22 837 presents a process in which lactic acid is purified by the
formation of methyl lactate.
Methods of the pur1f1cat10n of citrlc acld by means of liquid/liquid
extract10n are d1sclosed in US-PS 4 251 671 and 4 334 095 as well
as 1n AT-PS 382 364.
A process of 1solat1ng citric acid from fermentat10n broths by
the format10n of spar1ngly soluble salts is disclosed in US-PS 2
810 755.
Further processes of extract1ng tartar1c actd,c1tric acid, mal1c
ac1d and lact1c ac1d were proposed 1n DE-OS 1 443 538 and EP O 049
429.
Methods of pur1fytng phosphor1c acid are disclosed 1n DE-OS 23
20 ~77 and 17 67 442 and 1n DE-AS 23 21 751.
for c1tr1c ac1d, EP-A O 151 470 further d1scloses a process where1n
f1rst, h1gher molecular compounds are separated from the contam1nated
c1tr1c ac1d solut10n by means of ultraf11trat10n; impur1t1es st111
present are subsequently removed by adsorpt10n on non-ionogen1c restns
of large spec1f1c surface, the substant1a11y more polar c1tr1c acid
rema1n1ng unbound.
As a further pur1ficat10n measure, US-PS 4 720 579 d1scloses the
adsorpt10n of citr1c ac1d on neutral, non-10nogen1c, macroret1cular,
water-1nsoluble res1ns. The eluant used 1s water or a mixture of water
and acetone. The process 1s capable of separating salts and
carbohydrates from the citric acid.
The present 1nvention is based on the state of the art represented
by US-PS 4 720 579. On principle, the ob~ect is to pass substrates
conta1n1ng acids, salts and carbohydrates over resins selectively
-- 3 --
, ~ .
.i .
. .
:i;
.j .
2~C~7~l26
acting as adsorbers for the acid;, while the other components of
the substrates are passed on. The acids whose purification or
recovery is intended are subsequently obtained by elution of the charged
res1ns.
It was now surprisingly found that basic ion exchange resins are
su1table for the chromatographic isolation and/or purification of
ac1ds 1n general, 1.e. as well as of organic as of inorganic acids,
wlth fru1t ac1ds to be particularly mentioned among the carboxylic
ac1ds. Of the typ1cal ac1dic organic edible acids , tartaric acid,
c1tr1c acld, mal1c ac1d, fumaric acid and succinic acid are usually
des1gnated as fruit ac1ds 1n 11terature.
The ac1ds are adsorbed selectively and reversibly and the ;~
pur1f1cat10n effects obta1ned and the charging capac1ty of the resins
are substant1ally h19her than those of the non-ionogen1c resins.
The process accord1ng to the invent10n of chromatographically
pur1fy1ng or separat1ng acids is thus ma1nly characterized 1n that
bas1c 10n exchange res1ns are used for the separation.
The use of bas1c 10n exchange resins for the purificat10n of
fru1t ac1ds 1s known per se 1n processes 1n wh1ch the fru1t ac1ds
themse1ves pass through the 10n exchanger bed while impurities in the
substrate such as stronger ac1ds or salts are retained on the resin
by 10n exchange. The 10n exchange resins are regenerated in this ;-,
known process pr10r to the break-through of the impur1ties , the charge
be1ng removed. ; DD-PS 203 533 discloses an ion exchange process for the recovery
of carboxylic and oxycarboxylic acids from their salt solutions
conta1ning foreign salts in which the salts are first converted to ~-~
~.,
4 - ;~
~)07~26
ac1ds via cation exchangers in H~-form. The result is a m~xture of
carboxylic acids and foreign acids. A fraction of this mixture of
low concentrat10n is then used for first charg~ng an anion exchanger
present 1n the base or hydroxyl form , thus substantially converting
1t to the carboxylic acid form. A higher concentrated fraction of
the ac1d m1xture obtained in the decationization step is then passed
over the 10n exchanger charged with acid, with the acid to be recovered
pass1ng the res1n bed unhampered and merely the stronger foreign anions
(ch10r1de be1ng ment10ned) be1ng bound to the resin by 10n exchange.
DE-OS 29 31 759 relates to the separat10n of citric ac1d from a
m1xture w1th oxallc acid and n~tric acid ln wh1ch this mixture is passed
over an anion exchanger wh1ch was optionally previously equillbrated
w1th c1tr1c ac1d. The 10n exchanger withdraws nitr1c acid and oxal1c
ac1d from th1s m1xture by ion echange wh11e the c1tr1c acid passes
through the column.
DE-OS 25 43 332 relates to a process of separat1ng maleic acid
1mpur1t1es from synthet1c tartar1c acid wh1ch 1s passed over bas1c
an10n exchangers present 1n hydroxyl or tartrate form. The 10n
exchanger aga1n w1thdraws the stronger maleic ac1d from the mixture,
wh11e the tartarlc acid to be recovered passes through the column.
In the process accord1ng to the invention, the acids to be recovered
or pur1f1ed are not recovered by 10n exchange, but instead adsorptively
bound to the res1ns and recovered from there by elution.
The bas1c 10n exchange resins are conveniently used for the
1solat10n and /or pur1fication of ac1ds of a PKS value within a range
of from 1.2 to 4.8, preferably of from 2.1 to 3.9.
The resins are preferably employed in acid-charged form. According
. ':
- 5 -
, . i
'~ .
2~ 7 ~L
tQ ~he ~nventiQn~ the resin is conveniently charged by means of an
anlon to which the resin has an equal, or higher affinity than, to
thè acid to be purif1ed. The charging of the resin does not have to
be carr1ed out as an expliclt operat10n, in many cases, it automatically
occurs 1n the course of extended use. Particularly used are bas1c
10n exchange res1ns charged wlth hydrochloric acid, nitric acid,
phosphoric acid, citric acid or sulfuric ac1d, preferably w1th sulfuric
ac1d~
The an10n exchange res1ns according to the invention can be strongly
or weakly basic and of gel or macroreticular structure. The ion
exchangers may among others contaln a styrene/d1vinyl benzene basis
or a cross-linked acrylic resin bas1s.
The elut10n can be carried out with water in a known manner; 1t
1s preferred, however, to use a diluted acid, in particular sulfuric
ac1d 1n a concentrat10n range of from 0 to 1 percent. Low pH values
known to be beneficial for other processes are also convenient for
the process according to the 1nvention. The elution temperature can
be selected between room temperature and the stabil~zation threshold
temperature of the res1n, the elution is preferably carrled out at
a temperature preventing the microbial contamination of the resin
The process 1s preferably carried out in a quasi-continuous apparatus
(s1mulated mov1ng bed), although fully continuous or batch operation
1s also poss1ble.
The 1nvent10n 1s expla1ned 1n the follow1ng by means of examples
The res1ns used are conventional, commerc1ally available products
accord1ng to the following table:
~,~
- 6 - `
~,'
:::
.~ .
'',
. :
~007~26
Des~gnatlon Polymer Basis Active Group Type Manufacturer
AMBERLITE XAD-4 polystyrene - O Rohm & Haas
AMBERLITE IRA904 styrene/DVB amine --- Rohm & Haas
AMBERILTE IRA900 styrene/DVB amine --- Rohm & Haas
AMBERLITE IRA958 acrylic/DVB amine --- Rohm & Haas
AMBERLITE IRA93 styrene/DVB amlne - Rohm & Haas
AMBERLITE IRA67 acrylic amine - Rohm & Haas
AMBERLITE IRA35 acrylic/DVB amine - Rohm & Haas
LEWATIT MP-64 polystyrene amine - Bayer
LEWAtIT MPSOOA polystyrene amine -- Bayer
DOWE~ ~GR2 epoxy amine amine - Dow Chemical
RELITE MG1 acrylic/DVB polyamine - Tanatex
O ..... non-ionogenic
- ..... weakly bas~c
-- ..... limited strongly basic
--- .......... strongly basic
EXAMPLE 1
Following complete regeneration by means of 4 percent NaOH, the
an10n exchange resins were converted to the sulfate form by means of
10 percent sulfuric ac~d and conditioned by means of 0.1 percent
sulfuric acid. The resins thus prepared were charged into columns
of 25 mm diameter and 800 mm length. Following the inject~on of 5 ml
of a solution consisting of 40 percent citric acid and 10 percent
maltose ln water, elution was carried out with 15 ml/min 0.1 percent
sulfuric acid. The column temperature was ad~usted to 80C by means
of a c1rcul~t~on thermost~t. The eluate concentrations were determiend
',
~ 7~L~6
by measuring the refractive index. Under these conditions, maltose
and citric acid were completely separated on nearly all the res~ns.
The table shows the retention volumes of maltose and citr~c acid.
Resln Elution Volume Elution Volume
Maltose Citric Ac~d
(ml) (ml) ;;
AMBERLITE IRA 904 285 505
AMBERLITE IRA 900 270 911 ~;AMBERLITE IRA 958 270 656
AMBERLITE lRA 93 SP 338 806
AMBERLITE IRA 67 210 566
AMBERLITE IRA 35 296 544 ;~LEWATIT MP 64 341 731
LEWATIT MP 500A 251 690
WWEX WGR 2 no separation possible
RELlTE MG 1 no separat1On possible
EXAMPLE 2
As descr1bed 1n Examp1e 1, a column was packed with AMBERLITE ;~
IRA 67 and equllbrated w1th a flow of 30 ml/m1n of 0.1 percent sulfur1c
ac1d. The temperature of the column was ad~usted to 75C. The exctusion
volume of the resin was determined by in~ection of 1 ml of a 5 ~'^i
percent Dextran T10 solution . It can be assumed that a high molecular,
uncharged substance w111 not be retained under these condlt1Ons. ;,
The exclusion volume was calculated as follows ;;
,.',
,~ ,
I - 8 - ~
',,
3 :
.. .
' :~
.; ...... , .. ,., .. ,,,., .. ,.- ,.. . .. , ., ,'~
7~L26
T.
e - ~--
V
Fl
Ttr ........... retention time of Dextran in seconds
V ............. column volume in ml
Fl ............ f10w ~n ml . s 1
F'rec~sely 1 ml of var~ous substrates was now ~ntroduced ~nto the
medlum flow by means of a "drag" in~ector and detected at the column
exit v~a the refractive index. The time up to the occurrence of the
maxlmum concentration in the eluate and the width of the peak at half
he1ght were measured. These data were used for calculat~ng the
separation stage number and the distribution coefficient:
Tm~u--T--T
T~
n-16~ In ~ 2)
d2
u1~ Tm~
K- --------_-_______
(1-e) ~ z
T ...... retent~on time of substance in seconds
n ...... separatlon stage number of column
d ...... w1dth of substance peak at half height in seconds
u ...... area-related flow ~n mlts . crn2
z ...... column length in cm
K ...... distribution coefficient of substance
7~6
:,
The follow1ng table shows the results obtained:
Substance Concentrat~on Separat1On Stage D~str1but1On
( % by we~ght) Number Coeff1cient
n K .:
Sod1um sulfate 10 9.09 0.219 ..
10.11 0.262 ~.:
Maltose 10 2.25 0.149
2.17 0.146 `
2.25 0.149 ~
2.25 0.149 ;;
Glucose 10 6.24 0.362
6.02 0.356
5.81 0.349
Glucon1c Ac1d 10 6.46 0.541 ~
6.22 0.514 ;.
6.22 0.514 ~
Lact1c Ac1d 10 15.11 0.673 ~;.
14.07 0.663
C1tr1c Acid 10 17.50 1.864
18.06 1.815
16.48 1.748
17.25 1.775 ;.~
,'.,"
These values clearly show that a purif~cat~on of c1tric acid under ~.
the selected conditions is poss1ble. The data also show a separat1On
of c1tric ac1d - glucon1c ac~d and between c1tric ac~d - lact~c "
ac1d. .
. ~'
~ - 10 -
'".',
:i ;
!~
~ ~007~26
EXAMPLE 3
Example 2 was repeated at a temperature of 60C.
Substance Concentrat1OnSeparat1On Stage D1str1but1On
(% by weight) Number Coeff1cient
n K
Sodlum Sulfate 10 7.70 0.192
8.71 0.242
Maltose 10 1.62 0.110
1.70 0.112
Glucose 10 4.93 0.322
4.57 0.312
G1ucon1c Ac1d 10 5.33 0.489
4.85 0.483
C1tr1c Ac1d 10 13.83 1.743
12.69 1.617
EXAMPLE 4
Example 2 was repeated at a temperature of 90C.
~ 2 6
Substance Concentration Separation Stage Distribution
(% by we~ght) Number Coeffic~ent
n K
Sod1um Sulfate 10 11.09 0.227
12.39 0.280
Ma1tose 10 2.97 0.183
3.06 0.186
Glucose 10 8.76 0.404
8.22 0.392
Glucon1c Acid 10 10.96 0.628
9.80 0.594
Lact1c Ac1d 10 18.99 0.776
17.77 0.752
C1tr1c Ac1d 10 30.00 2.514
25.10 2.147
A compar1son of the results of Examples 3 and 4 shows the favorable
1nfluence of the temperature on the separation.
EXAMPLE 5
.
A polypropylene column with a d1ameter of 27 mm and a length of
146 mm was packed with AMBERLITE IRA-67 in the sulfate form. At a
temperature of 28C, 4.97 ml/min 0.1 percent sulfuric acid were pumped
over the column. Then, 0.5 ml hydrofluoric ac~d or phosphoric acid
were ln~ected 1nto the med1um flow. The discharge of the substances
from the column was detected by means of a conduct1vity detector.
12 -
~,
,, ,
.~
x~n7l26
Substance ConcentrationSeparation Stage Distribution
(X by weight)Number Coefficient
n K
Hydrofluor1c Ac1d 10 15.45 1.06
Phosphor1c Ac1d 40 15.01 3.14
Th1s example demonstrates the possibility to purify inorganic acids
as we11.
EXAMPLE 6
Example 2 was repeated w1th 0.1 percent sulfurlc acid as flow med~um
and IRA900 as 1On exchanger at a temperature of 75C.
Substance ConcentrationSeparat1On Stage Distribut1On
(% by we19ht) Number Coefflcient
n K
Sod1um Sulfate 10 6.80 0.319
8.01 0.358
Maltose 10 2.77 0.316
2.77 0.309
61ucose 10 6.24 0.485
6.24 0.485
Glucon1c Ac1d 10 7.02 0.726
6.43 0.717
Lact1c Ac1d 10 16.73 1.036
16.41 1.043
Tartar1c Acid 10 19.47 2.909
Mal1c Ac1d 10 16.86 2.212
Itacon1c Ac1d 10 15.96 3.095
r` ~ 13
~ .
.~ ,
. . .
7~26
Example 7
Example 2 was repeated with water as the flow medium and IRA900
as lon exchanger at a temperature of 75C.
Substance Concentrat~on Separat~on Stage D~str~but~on
(X by we19ht) Number Coeff~cient
n K
Sod1um Sulfate 10 7.70 0.343
7.90 0.370
Maltose 10 3.09 0.340
2.96 0.340
Glucose 10 7.62 0.545
7.03 0.53
Lact1c Ac1d 40 35.47 2.549
EXAMPLE 8: ~
.:
A sem1cont~nuous chromatography plant cons1sting of ten columns
~d1ameter 25 mm, length 1600 mm) was f111ed w~th AMBERLITE IRA900 ln
the sulfate form. The plant ~s layed out ~n such a manner that the
columns are connected by means of p1pe condu~ts to form a rlng. Th~s
r1ng 11ne is connected between the columns to four supply 11nes
v1a valves. These supply 11nes are crude product ~nlet, eluant inlet,
product outlet and raff1nate outlet. The flows ~n the plant are kept
constant by means of adjustable rec~procating pumps. In th1s process,
there are always exactly four valves open and div~de the ring formed
of columns into four zones. These zones are advanced ~n flow d~rect~on
by one column after predeterm~ned per~ods by an ~n-process computer.
~",
.', .
,, ,
.. ,
., .
2~)7~26
,,
These sw1tch1ng operations simulate a resin flow counter to the flow
direct10n, thls is why it is designated as "simulated moving bed"
process 1n 11terature.
.
~ :::
The dtv1sion 1nto zone was effected according to the fo110w1ng
scheme:
Zone 1 2 columns
Zone 2 3 columns
Zone 3 4 columns
I Zone 4 1 column
¦ 8 ml of crude product and 32 ml of water per m1nute were 1ntroduced
and 16 ml of product or 24 ml of raffinate were withdrawn during th1s
t1me. The temperature of the plant was kept constant at 70C. The
1nternal c1rculat10n volume in the column r1ng was ad~usted to 4
ml/min. The valves were sw1tched every 1550 seconds. A lactic ac1d
fermentat10n mash concentrated to approximately 40 percent was
_ 1 5
." .,
. ~
:
2007126
1ntroduced 1nto this plant. A purified lactic acid with a yield of
more than 95 percent was recovered. The raffinate contained, ~n addition
to lact1c ac1d traces, colorants, carbohydrates and salts, above all
polymers which can cons~derably 1nterfere with lactic ac1d production.
The purlfy~ng effect 1s also shown by the h~gh chemlcal oxygen
consumpt10n of the raffinate of 13.5 9/l. The follow1ng table shows
a compar1son of lact1c ac1d before and after chromatograph1c
separat10n:
before after
,, - .
Lact1c Ac1d (%) 40.52 19.53
Color (405 nm, 1 cm) 1.94 0.43
Sulfate (X) 1.78 0.03
Chlor1de (mg/kg) 330 23
Sod1um (mg/kg) 7820 330
Magnes1um (mg/kg) 7 0
Iron (mg/kg) 27 5
EXAMPLE 9:
"
A sem1-1ndustr1al chromatography plant cons1st1ng of twelve columns
(d1ameter 0.5 m, length 2.0 m) was f111ed w1th AMBERLITE IRA-67. The
res1n was converted to the sulfate form by flush~ng with 0.1 percent
sulfuric ac1d. The columns were connected as shown 1n the followlng
~ d1agram:
,1 - 16 -
f
~ :
,1 '
~, .
Z007~Z6
contaminated
solution -- _ eluant ~ _
purified ~ contamina~s ~
' ~
At a temperature of 75C, 215 l/h of a citric acid fermentation
broth concentrated to approx1mately 45 perent by weight were 1ntroduced
and 825 l/h of 0.1 percent sulfuric acid were added as eluant. At
the same t1me, 500 l/h of pur1f1ed citr k ac1d solution (product)
and 533 l/h of a solut1On containing the impurities (raff1nate) were
w1thdrawn. 192 1/h were c1rculated in the column r1ng, the 1nlet
and outlet s1tes were sw1tched by one column in flow direction every
1800 seconds.
An analys1s of the so1utions added and withdrawn yielded the
follow1ng values:
17
'
.~ '
,. .:
.~. ~,.
20~7~ 6
Fermentation Broth Product Rafffinate
Cltr~c Acid 43.45 ~ 22.36 X 0.02 %
Glucose 0.25 % 0.00 % 0.16 X
Fructose 0.10 X 0.00 % 0.11 %
Ma1tose ~
Isomaltose 1.40 % 0.00 % 0.80 %
Total Sugar 4.60 X 0.03 % 2.35 %
Sulfate 0.98 % 0.27 % 0.43 %
Chlor~de 110 ppm 16 ppm 42 ppm
Iron 8 ppm 2 ppm 2 ppm
Potassium 370 ppm 0 ppm 190 ppm
Sodlum 55 ppm 0 ppm 29 ppm
Ca1c1um 20 ppm 1 ppm 8 ppm
Magneslum 120 ppm 0 ppm 60 ppm
Density 1.27 g/ml 1.07 g/ml1.00 g/ml
The productivity of the plant amounts to 2.87 tons (metr~c) of citric
acld anhydrate per day, the yield of 99.9 percent is substantially
better than those of the known purificat~on processes for citric
ac1d. It is evident that virtually all impurities, whether cationic,
I anionlc or uncharged, are separated.
' EXAMPLE 10:
!: Example 9 was repeated with crude phosphoric acid. A prefiltered
~ wet phosphoric acid (green acid) of a concentration of 37 percent¦ of P205 was introduced. The columns were switched in flow direction
':
- 18 -
d
''~
.
2007~26
every 1500 seconds. An analysis of the samples of product and raffinate
in the equilibrated state showed the following values:
Product Raff~nate
t1trable ac1d as P205 15.60 1.90 X
color (405 nm, 1 cm) 0.047 1.486
The y1eld of phosphor1c ac1d amounted to 89.9 percent, a comparison
of cat1On charge before and after chromatographic separat1On shows
the h1gh pur~f1cat1On performance of the system (all concentrat~ons
based on 100 percent P205).
beforeafter
Magnes1um 20800 42 mg/kg
Calc1um 358 9 mg/kg
Alum1num 3390 410 mg/kg
Iron 4620 1200 mg/kg ;.
Z1nc 855 15 mg/kg . :
Chrom1um 614 102 mg/kg .
Copper 29 0 mg/kg
N1ckel 56 0 mg/kg
Manganese 21 0 mg/kg
Cadm1um 23 0 mg/kg
~ .,,
~ - 19 -
,~ I ...
.'~ ~, .
