CA2205535C - Sucralose pentaester production - Google Patents

Abstract

In copending SN 2,114,180, a process for producing substantially pure sucralose pentaester from a mixture of 6-O-acyl-4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose in a reaction medium comprising a tertiary amide is described wherein said process comprises the steps of:
(a) recovering the 6-O-acyl-4,1',6'-trichloro-4,1', 6'-trideoxygalactosucrose from said mixture;
(b) peracylating the 6-O-acyl-4,1',6'-trichloro-4,1',6'trideoxygalactosucrose product of step (a) to produce thereby 4, 1',6'-trichloro-4,1',6'-trideoxygalactosucrose pentaester; and (c) crystallizing the 4,1',6'-trichloro-4,1',6'-trideoxygalactosucrose pentaester product of step (b) to produce substantially pure 4,1',6'-trichloro-4,1',6'-trideoxygalactosucrose pentaester.
This application is for a specific sucralose pentaester product that can be obtained from the above process, namely 6-O-benzoyl-4,1',6'-trichloro-4,1',6'-trideoxygalactosucrose tetraacetate.

Description

CA 0220~3~ 1997-06-10 -SU~RALOSE PENTAESTER PRODUCTION
The invention relates to a process ~or the production of sucralose pentaester and to a specific sucralose pentaester prcduct.
Backqround of the Invention The arti~icial sweetener 4,1',6'-trichloro-4,1',6'-trideo~yqalacotosucrose ("sucralose") is derived from sucrose by replacing the hydroxyls in the 4,1', and 6' positions with chlorine. (In the process of making the sweetener, the stereo configuration at the 4 position is reversed - hence the compound is a qalactosucrose.) The direction of the c~lorine a~oms to only the desired positions is a major synthesis problem because the hydroxyls that are replaced are of differing reactivity; two are primary and one is secondary. The synthesis is further complicated by the fact that the primary hydroxyl ~n the 6 position is unsubstituted in the final product.
A number of different synthesis routes for the preparation of sucralose have been developed in which the reactive hydroxyl in the 6 position is first blocked, as by an ester group, prior to the chlorination of the hydro~yls in the 4, 1', and 6' positions, followed by hydrolysis to remove the ester suhstituent to produce sucralose. Several of such synthesis routes involve tin-mediated syntheses of sucrose-6-esters.
Illustrative are the tin-mediated routes disclosed by Navia (U.S.
Patent No. 4,950,746), Neiditch et al. (U.S. Patent No.
5,023,329), Wingard et al. (U.S. Patent Application No. 870,1g0, CA 0220~3~ 1997-06-10 ,.

filed April 13, 19g2 - published as EP-A-0 475 619 A1), and Walkup et al. (u.s. Patent No. 5,08g,608) - Walkup et al.-I).
The above-illustrated tin-mediated syntheses have in common the preparation of a sucrose-6-ester that can be chlorinated to produce a sucralose-6-ester (that is, sucralose having an ester group substituent at the 6 position). The sucrose-6-ester can be chlorinated by the process described in Walkup et al., U.S. Patent No. 4,980,463 (Walkup et al.-II). The Walkup et al.-II process produces as a product of the chlorination a mixture of 6-O-acyl-4,1',6'-trichloro-4,1',6'-trideoxy~alactosucrose ("sucralose-6-ester") in a tertiary amide solvent such as N,N-dimethylformamide. The chlorination reaction praduct mixture also contains water, salts, and chlorinated carbohydrate byproducts.
In copending SN 2,114,180, an improved method is described for producing sucralose from a mixture of 6-~acyl-4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose in a tertiary amide solvent such as N,N-dimethylformamide.
This present invention provides a novel compound, prcduct of the improved method, namely 6-O-benzoyl-4,1',6'-trichloro-4,1',6'-trideoxygalactosucrose tetraacetate.
Brief Summary of the Invention The invention relates to a process for producing substantially pure 4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose pentaester from a reaction mixture comprising 6-C-acyl-4,1',6'-trichloro-4,1',6'-CA 0220~3~ 1997-06-10 tri.deoxyqalactosucrose and a tertiary amide such as N,N-dimethyl~ormamide, wherein said process comprises the steps o~:
(a) recovering the 6-C~acyl-4,1~,6'-trichloro-4,1~,6'-trldeoxyqalactosucrose from said mixture;
(b) peracylating the 6-C~acyl-4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose product of step (a) to produce thereby 4,1~,6~-trichloro-4,1~,6~-trideoxygalactosucrose pentaester; and (c) crystallizing the 4,1',6'-trichloro-4,1',6'-trideogyqalactosucrQse pentaester product of step (b) from amixture of water and a substantially water-immiscible solvent to prc~duce substantially pure 4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose pentaester.
The invention is a product of the a~ove process, 6-G-ben.zoyl-4,l',6'-trichloro-4,1',6'-trideoxyqalactosucrose tetraacetate.
The invention also relates to a process for recoverin~
sucralose-6-ester from a feed mixture of (a) 6-O-acyl--1,1',6~-trichloro-4,1',6'-trideoxyqalactosucrose, (b) salt comprising alkali metal or alkaline earth metal chloride, (c) water, and (d)=other chlorinated sucrose by-products, in a reaction medium comp~ising a tertiary amide, wherein said process comprises removing said tertiary amide by steam distillation, to produce an aqueous solution product of (a), (b~ and (d) containing not more than 0.5 weight % of the tertlary amide.
~ he product of step (c) can then be de-acylated to produce sucralose in high yield and purity.

-2a-' ' CA 02205535 1997-06-10 - ' NSP--4 Brief Description o~ the Drawinqs Fig. 1 is a diagram of a laboratory-scale dual stream quench app~ratus designed for neutralizing the acid present in sucralose-6 ester chlorination product reaction mixtures; and Fig. 2 is a diagram of a laboratory-scale falling-film packed-coltlmn steam distillation apparatus designed for stripping the DMF
fro~n quenched sucralose-6-ester chlorination product reaction mixtures.

Detailed Descri~tion of the Invention Nomenclature and Abbreviations As used in this application, the foll~wing short names and abb:reviations have the indicated meaning:

Sucralose = 4,1',6~-trichloro-4,1~,6~-trideoxyqalactosucrose;
DMF = N,N-dimethylformamide;

S-6-A or sucrose-6-acetate1 = 6-O-acetylsucrose;

S-6-B or sucrose-6-kenzoate = 6-O-benzoylsucrose;

Sucralose-6-acetate or 4,1~,6'-trichloro-4,1',6~-trideoxyqalacto-sucrose-6-acetate=4,1',6'-trichloro-4,1',6'-trideoxyqalactosucrose-6-acetate;
Sucralose-6-benzoate or 4,1',6'-trichloro-4,1',6'-trideoxy-galactosucrose-6-benzoate = 4,1~,6~-trichloro-4,1~,6'-trideoxy-galactosucrose-6-benzoate;

~SP--4 Sucralose pentaacetate = 4,1',6'-trichloro-4,1',6'-trideoxy-qalactosucrose pentaacetate; and The process of the invention employs as its starting reaction mixture a composition comprising 6-O-acyl-4,1',6'-trichloro-4,1',6~-trideoxy~alactosucrose in a tertiary amide (pre~era~ly DMF) re~ction medLum, such as the neutralized (quenched) product o~ the chlorination reaction described by Walkup et al. II, cited above.

On the laboratory scale, the crude chlorination product may be quenched in a batch operation by the addition ~in one portion) of one molar equivalent (~asis phosgene) of ice-cold aqueous solutions or slurries of the al~ali or alkaline earth metal hydroxides following the teachings of Walkup et al.-II. Preferred alkaline agents include the hydroxides of sodium, potassium, and calcium. More dilute aqueous alka~ine solutions, such as for exam~le 4 to 8N sodium ~ droxide, are preferred.
.

In a preferred method of practice of this quench method, cold aqueous alkali is added with vigorous stirring as rapidly as possible in a quantity sufficient to raise the pH to 8-10. After stirring several minutes at this mildly elevated pH, the quenched solution is neutralized to p~ 5-7 by the addition of an acid, such as, for example, concentrated aqueous hydrochloric acid or glacial acetic acid. The brief treatment of the quenched chlorination reaction mixture at pH 8-1~ has the beneficial effect of insuring that all of the hydroxyl groups that have not been replaced by chlorine atoms are returned to their original hydroxyl group form (i.e., they are deprotected ) .
The batch method for quenching the crude chlorination product mixture su~fers ~rom scale limitations owing to inef~iciencies in heat and mass transport. An improved method, known as the ~dual-stream" or "concurrent addition" method, involves mixing streams of aqueous alkali and cooled (to about room temperature~ crude chlorination product together at carefully metered rates with vigorous agitation und~!r conditions of pH and temperature control. The ~rimary advantag-es c,f the dual-stream quench method are that it proviaes for complete control of pH, temperature, and rate of mixing throughout the course s of t~e quench. Thus, side reactions resulting in product losses are minimized. A further advantage of-the dual-stream quench method is that it may be operated continuously by using a quench vessel fitted with either a bottom drain or a pump. By operating the dual-stream quench method in a continuous mode, a relati~ely large amount of crude chlorination product can be processed using a quench ~essel of modest size. This continuous operation is a rough approximation of an in-line m; ~; ng process that might be employed for ~uenchinq in a ~om~rcial operation.
.
i,~
/~ The laboratory-scale dual-stream quench apparatus consists of a temperature-compensated pH control pump for the addition of aqueous alkali, a second pump for the constant addition of the crude chlori-nation product mixture, a quench vessel fitted with an external jacket to allow for the flow of coolant, a thermostated chiller to both cool and pump the coolant, and various pieces of auxiliary equipment such as a mechanical stirrer, thermocouples, etc. The apparatus is operated by adding the crude chlorination product mixture to the vess~l at a constant rate. The pH control pump is fit~ed with a pH
meter and a pH probe which is placed in the quench vessel. The control pump adds aqueous caustic automatically in response to programmed instructions for maint~;~;ng the pH of the mixture at a certain value. Vigorous agitation of the solution in the quench vessel is required. Experiments have indicated that inadequate mixing will result in domains o~ inadequate p~ control within the quench mixture, resulting in the loss of product-to side reactions.

Using a 1500-ml ~a~eted quench vessel, it was de~ermined that crude sucralose-6-ester product mixtures could be quenched efficiently using a chlorination mixture constant feed rate of about 10 ml per minute, a quench mixtura tPmr~rature of about 15~C (coolant tempera~

ture 5~C), a four-bladed propeller-type stirrer with a stirring rate sufficie~.t to ir.sure good m; x; n~, and a pH control setting of pH 8.5 on the pH control pump. These results were obtained with 3N or 4N
NaOH as the alkaline agent, and with a starting charge of about 100 ml of between 3:1 to 1:3 DMF-H2O in the quench vessel ~in order to have suff;cient solution volume for accurate pH measurement during the early stages of the quench). A diagram of a laboratory-scale dual-stream quench apparatus is shown in Fig. 1.

DMF REMOVAL

Following the quench, sucralose-6-ester is recovered from a mixture containing DMF, water, salts, and chlorinated carbohydrate b~o~ucts. The salts are approximately 1:1 sodium chloride:di-methylamine hydrochloride, but small amounts of sodium formate alsoappear to be present. The direct extraction of sucralose-6-ester from thelquenched product mixture is complicated by the presence of DMF and its propensity to distribute between both phases. Laboratory experimentation established that DMF can be removed from the quenched chlorination product mixture by steam distillation without any detectable decomposition of the desired chlorination product. The DMF
can subsequently be recovered from the aqueous overheads by distilla-tion, and can then be recycled.

25An example of a laboratory-scale falling-film pac~ed-column steam dist:illation apparatus designed for stripping the DMF from quenched sucralose-6-ester chlorination products is shown in Fig. 2. The str~pping column is a 5.0-cm diameter, 90-cm long vacuum-jacketed dist:illation column packed with 5-mm Raschig rings or other suitable 30packing. Alternatively, a 15-plate, jac~etedi Oldershaw column has been used. The quenched product, which is typical}y preheated, is introduced into the top of the column at a rate of about 5.0-5.5 grams per minute. Steam is introduced into the column through a sidearm located at the bottom of the column. As condensate-free steam is ' ' CA 02205535 1997-06-10 -requ:ired, the steam is past through a "preboiler" to trap any condensate carried over. In the laboratory, this preboilér is typic,ally a small multineck flask fitted with a heating mantle.
Typical steam feed rates are in the range of 38-47 gra~s per minute (calculated by adding the weights of overhead and bottom products, and then subtracting the weight of chlorination feed), which corresponds to a stea~-to-feed ratio ranging from 4:1 to 12:1, with steam to feed ratios of between 7.5:1 and 9:1 being typical for the packed column assembly. The preferred embodiment would use more plates with a lower steam:feed ratio, e.g., 15 plates with a steam/feed ratio of about 4:1.

The preheating of the que~ched chlorination feed before it is introduced onto the top of the column is conducted in order to lS increase the efficiency of the stripping operation. Preheating is typically conducted in the labo-atory by passing t~e feed through an enclosed glass coil apparatus heated with a secondary source of steam.
The feed is normally heated to about 90-95~C. The efficiency of DMF
remo~Jal can also be enhanced by employing a "reboiler" (i.e., by heat:ing the bottoms product in such a way that it refluxes up into the stripping column).

Temperatures are advantageously measured at two places on the apparatus using thermocouple devices. In addition to the quenched chlorination feed temperature described above, the temperature of the vapors passing through the distillation column head are also measured.
Head vapor temperatures are typically in the range of from about 99~C
to about 104~C.

A typical quenched chlorination product of sucrose-6-acetate contains about 1.5-5 wt % sucralose 6-ester, about 0.5-1.5 wt % of various other chlorodeoxysucrose derivatives, about 35 45 wt % DM~, about 3S-45 wt % water, and about 12-18 wt ~ salts. After passage of such product mixtures through the laboratory-scale steam-stripping apparatus, bottoms products will typically consist of about 1-3 wt %

~SP-4 sucralose-6-ester, about O.3-1.0 wt % of various other chlorodeoxy-sucrose derivatives, about 0.1-0.5 wt ~ DMF, about 80-~0 wt % water, and about 8-12 wt ~ salts (expressed as NaCl, n~C~ on sodium and chloride assays).

Under typical laboratory conditions (see Example 1), which involve a column residence time o~ 7-10 minutes, no decomposition of sucralose-6-acetate is detectable, provided the pH of the ~uenched chlorination feed is neutral to slightly acidic (pH 5.0-7.0).
SUCRALOSE-6-~S~ER EX~RACTION --Following the steam strip, sucralose-6-ester may be readily isol~ted by extraction of the DMF-depleted aqueous brine solution with a variety of organic sol~ents. ~hese solvents include methyl acetate, ethyl ac~tate, methyl ethyl ketone, methyl iso-~utvl ketone, methyl iso - amyl ketone, methylene chloride, chloroform, diethyl ether, methyl tert~butyl ether, and the like. A preferred solvent, for reasons of extr'~ction selectivity, ease of recycle, and toxicological safety, is ethyl acetate.

Sucralose-6-ester isolation is typically conducted in the laboratory by first partially evaporating the crude steam-stripped product. About half the water present may optionally be removed, producing a solution containing about 2-5 wt ~ carbohydrates and about 15-2S wt % salts. Isolation is normally conducted by carrying out three seguential extractions with ethyl acetate or other appropriate solvent. The extracts are combined, and may optionally be washed with water (to partially remove DMF and dichlorodideoxysucrose derivatives which to some extent are partitioned into the organic phase).
Evaporation of the sol~ent produces crude solid sucralose-6-ester.

These crude solid products typically contain about 70-80 wt ~
sucralose-6-ester and about 7-16 wt ~ of various chlorodeoxysucrose deri~atives (both acylated and nonacylated), with va-ying degrees of c chlo~.o-substitution. These crude solids also typically contain small residual amo~m~s of DMF, water, and ethyl acetate. It is desirable to minimize the water content of these crude solids (e.g., by conducting the brine wash of the combined extracts as described above3, because the next step of the process involves treating the material with acetic anhydride, which will be partially con~ by the water present. Typical experimental prQcedures for the extraction and isolation of sucralose-6-ester are provided in Examples 1-3.

SUC~LOSE PERACYLATE OR P~NTAESTER P~EPARATION AND PURIFICATION

Sucralose-6-ester is exhaustively acylated by treatment with an acylating agent such as acetic anhydride, and the sucralose pentaester thus produced is purified by extractive crystallization. Peracylation lS is typically conducted by heating the crude solid sucralose-6-ester such as s7~crose-6-acetate w~th a moderate excess of acetylating agent such as acetic anhydride in the presence of an acylation catalyst, such as pyridine, triethylamine, sodium acetate, or other art-known materials. There are a number of nonhydroxylic organic solvents which ~0 can be employed as cosolvents, if desired. These include ethyl acetate, methyl ethyl ketone, methylene chloride, methyl tert-butyl ether, toluene, and the like. However, in the case of acetylation, the acetic anhydride is itself a satisfactory solvent for the conversion, thus eliminating the need for a cosolvent in this case.
The amount of excess acetic anhydride employed is minimized for economic reasons. A 25-50 molar % excess (basis all free carbohydrate hydroxyl groups and water present) has been found to be sufficient for reactions conducted in adequately dry media. Reaction temperatures in the 30-50~C range are satisfactory for providing complete conversion within several hours, although temperatures of up to the ~oiling point of acetic anhydride (138~C) may be employed if a faster rate of conversion is desired.

. .

.

' ' CA 02205535 1997-06-10 After the peracylation is complete, the reaction mixture is typically diluted with an app~o~riate solvent such as toluene (from abou~ S to about lO volumes, relative to weight of isolated sucralose-6-ester; e.g., from 5-lO ml toluene/g sucralose-6-ester starting reactant in the peracylation reaction), cooled to ~elow about 20QC, and treated with water (from about 2 to about 4 volumes, basis isolated sucralose-6-ester?. The biphasic mixture is then cooled to ~elow about 5~C, sP~P~ with authentic sucralose pentaester, and agitated until crystallization is complete.

The presence of-water in the crystallization medium serves-two purposes. The water destroys the residual excess acetic anhydride present, and it also provides a second phase during the crystalli-zation which, in effect, turns the toluene crystallization into an extractive purification. In the extracti~e crystallization the sucralose pentaester is soluble in the toluene phase, while the polar materials present, such as acetic acid, DMF, and trace amounts of salt, are soluble in the aqueous phase. Since the vast bulk of the impurities present in the product mixture following peracetylation are both extremely polar and water soluble, this biphasic crystallization is effective at producing a high yield of a high qu~lity product.

Toluene may be replaced by other solvents in the above-described purification scheme. These alternate solvents include, but are not limited to, benzene, mixed xylenes, cyclohexane, methyl tert-butyl ether, methyl ethyl ketone, and the like, plus mixtures of these. The primary criteria for this solvent are that it be a suitable re-crystallization solvent for sucralose pentaester.

',ucralose pentaester products isolated from toluene-based extractive crystallizations are typically in the range of from about 85 wt % to about 95 wt ~ pure. The bulk of the remainder of the weigh1: of these products is made up of water and (primarily) toluene.

~ MSP-4 Contamination of the sucralose pentaester produced by the bipha~;ic crystallization by carbohydra~e-based impurities is nn~in~l.
It is this high level of carbohydrate purity which makes this sucralose pentaester product suitable for conversion to sucralose.
Typically, the carbohydrate-based purity of the sucralose pentaester following the toluene-water cryst~]l;~ation is greater than 98 wt %, most c~ften greater than 99 wt ~. ~

The yields for the isolated crystalline sucralose pentaester afforded at this step of the process are typically from ~bout 90% to about 95% based upon crude solid sucralose-6-ester. Yields based on sucrose-6-ester utilized in the chlorination typically range from about 45% to about 55~. Overall yields from sucrose are normally in the range of from about 30~ to about 40%. The Examples provide experi.mental details for the conversion of crude solid sucralose-6-ace'~t:e and crude solid sucralose 6-benzoat~ into crysta~line sucralose pentaester. Example 2 provides an experimental procedure for the conversion of a sample of this sucralose pentaester into sucra].ose.
OPTIONAL SUCRALOSE PENTAESTER RECRYSTALLIZATION

If sucralose pentaester of higher purity than that produced in the biphasic crystallization is desired, it may be generated by recryc;tallization. This may be carried out by dissolving the sucral.ose pentaester in about from 7 to about 10 volumes (basis sucralose pentaester mass) of toluene at from about 80~C to about 100~C, and then allowing the solution thus produced to cool slowly, with agitation, to room temperature. Recovery is on the order of from about 84% to about 89%.

Solvents other than toluene may be employe~ for this optional recrystallization. These other recrystallization solvents include benzene, mixed xylenes, methanol, ethanol, ethyl acetate, methyl ethyl ketone, methyl tert-butyl ether, and the like, plus mixtur~s of these ~xample 3 provides experimental details for the recrystallization of a samp~e of crystallized sucralose pentaester from toluene, followed ~y t:he conversion of this additionally purified material into sucralose .

The above-described process has a number of advantages over other art-known processes for the production of sucralose pentaester. Said advantages are especially relevant to the commercial-scale manufacture of this valua~le precursor of the nonnutritive sweetener sucralose.
10One of these advantages is that, for example, S-6-A and S-6-B (which are used to produce the 6-0-acyl-4,1',6'-trichloro-4,1',6'-trideoxy_ galactosucrose-6-ester starting reactant of the present invention) can be produced cont~m;n~ted by residual sucrose and undesirable sucrose mono--and diacylates, and can be chlorina~ed without a requirement for 15producing isolated and purified solid S-6-A or S-6-B. Crystallization of the sucralose pent2cster in accordance with this invention removes the c:arbohydrate contaminants.

Another of these advantages is that the DMF is readily and 20essentially completely removed from the product stream by a falling-film packed-column steam distillation operation termed "steam stripping". Efficient removal of DMF is important for two reasons.
First, since DMF is a relatively expensive organic solvent, it is important that it be recovered and recycled in a cost-effective 25manner. The steam-stripping process allows for this to be accom-plished. Second, removal of the DMF from the chlorination product stream allows the extraction of sucralose-6-ester to function smoothly and efficiently. If the DMF is not removed prior to the extraction, undesirable partitioning of the various chlorinated carbohydrate 30products between the aqueous and organic phases is observed.

Additionally, since the sucralose-6-ester can be cleanly and efficiently extracted, crude sucralose-6-ester (i.e., that afforded by the direct evaporation of the extraction solvent~ is obtained in high 35yield (extraction efficiencies are typically 90-95~. Sucralose-6-CA 02205535 l997-06-lO

ester can be crystallized from ethyl acetate at a high state-of purity (normally 85-90% ~ure exclusive of residual organic solvents and moisture, which are not detrimental to the ~oi~ s). ~his high purity of the crude sucralose-6-ester provides for a high-yield conversion to high-~uality sucralose pentaester. The sucralose pentaester generated by this method is normally of s~fficient purity that it may be directly used for sucralose production without the need to resort to additional purification (by, for example, recrystallization) prior to deacylation to sucralose (as is the case for other art-known sucralose pentaester production procPsses). If any purification is desired, one optional recrystallization from, for example, toluene is norm~l-ly sufficient to produce sucralose pentaester suitable for sucralose produrtion.

The examples below illustrate the invention utilizing sucralose-6-acetate or sucralose-6-benzo~e as the sucralo~c 6 cster in the starting reaction mixture. However, other sucralose-6-esters, such ~s sucralose-6-propionate, sucralose-6-butanoate, and the like, may be used as the sucralose-6-ester.
The Examples below illustrate the invention.

ExamP:Le 1 ISOLATION OF SOLID SUCRALOSE PENTAACETATE FROM A PHOSGENE
CHLORINATION PRODUCT MADE FROM A SUCROSE-6-ACETAT~ SYRUP
710 Grams of crude aqueous chlorinated mixture prepared by the methocl of Neiditch et al. was employed as the starting reaction mixture. This reaction mixture, a dark agueous solution, was determined by HPLC analysis to contain 4.35 wt % sucralo~e-6-~cetate (30.9 g, 70.2 mmol), 0.7 wt % of various other chlorodeoxysucrose derivatives. Additional assays showed that this solution cont~ine~
about 40 wt % D~F, about 38 wt % water, and about 15 wt % chloride salts (NaCl and dimethylamine hydrochloride), in addition to lesser amount:s of sodium formate.

~ he above-described mixture was stPam stripped of DMF using a 5.0-CDI diameter, 90-cm length, vacuum-jacketed distillation COlD
packecl with 5-mm ~C~h; g rings. A reboiler was not employed. The product mixture (heated to about 105~C) was i-lLrGduced into the top of the cclumn at a rate of about 5.0-5.5 grams per min at a steam-to-feed ratio (by wt) of about 7.5-8.5. The distillation overheads were assaye!d by gas chromatography and found to contain about 5.6 wt % DMF.
~he distillation bottoms (982 g) were assayed by various techniques and found to contain about 0.2 wt % DMF, about 85 wt % H20, and about 11 wt % salts (expressed as NaCl). Carbohydrate composition was determined by ~PLC analysis to be 3.15 wt ~ sucralose-6-acetate and 0.11 wt ~ of other chlorodeoxysucrose derivatives.
.

The steam-stripped carbohydrate solution was subjected to rotary evaporation (water-aspirator vacuum, 50~C bath) to reduce its volume to about 500 ml. The solution was transferred to a 1000-ml separatory funnel and extracted with ethyl acetate (3 x 250 ml). The combined extracts were washed with water (1 x 100 ml), saturated aqueous sodium chloride solution (1 x 50 ml), and evaporated (rotary evaporator, water-aspirator vacuum, 40~C bath) to produce a light-tan solid which was fu~ther vacuum dried (25~C/l.0 mm ~g/24 hr) to a weight of 36.7 g.
Car~ohydrate composition was determined by HPLC analysis to be 76.2 wt % sucralose-6-acetate (27.9 g, 63.5 mmol, 90.5~ recovery from the crude quenched chlorination product mixture) and 11.5 wt ~ of various other chlorodeoxysucrose derivatives. Additional assays showed the solid t:o contain about 1.4 wt ~ DMF, about 8.4 wt ~- ethyl acetate, and about ;~.4 wt ~ water.

The crude-solid was treated with 65.0 g (637 mmol) of acetic anhydride and a few drops of pyridine at 50~C with magnetic stirring under argon for 24 hr. Silica-gel TLC (Et~0) was employed to follow -the formation of sucralose pentaester (Rf 0.7). ~he reac~ion mixture w~s dil~ted with 300 ml of toluene, cooled in an ice ba~h, t~e~ted with :L00 ml of H20 in three portions over 30 min, seeded with authentic sucralose pentaester, and then stirred at 5~C overnight. The resulting cryst~ll;ne solid was collected on a coarse-frit sintered glass filter, washed with 50 ml of ice-cold toluene, and vacu~m dried (45~C/1.0 mm Hg/60 hr). The drie* product weighed 40.1 g, and was found to consist of 9S.1 wt ~ sucralose pentaester (38.1 g, 62.7 mmol, 98.7% yield of crude solid sucralose-6-acetate) by HPLC assay.

The overall yield of solid sucralose pentaester from sucrose for this ~;et of experiments was 38.0%. This solid product is suitable for conversion into high-quality sucralose.

ExamPle 2 PREPARATION OF SUCRALOSE FROM SUCRALOSE PENTAACETATE

A 52.0 g sample of 88.6% pure sucralose pentaacetate (46.1 g, 75.8 mmol), prepared according to the method of Example 1, was slurried in 500 ml of methanol in a 1000-ml! three-neck, round-bottom flask equipped with mech~n;cal stirrer and argon inlet. The slurry was treated with 20.0 g of 20 wt % sodium methoxide (~.00 g, 74.1 Z5 mmol) in methanol, and stirred at room temperature under argon. The reaction mixture was homogeneous after 10 min, and sucralose (Rf 0.5) formation was judged complete by silica-gel TLC (4:1, CH2C12-CH3OH, sprayed with 5% ethanolic H2SO4 and charred) after 120 min.

The reaction mixture was quenched with acetic acid (5.00 g, 83.3 mmol), evaporated to dryness (rotary evaporator, aspirator vacuum, 30~C water bath), and then dried at high vacuum (25~C/0.5 mm Hg/18 hr) to remove as much of the methanol, methyl acetate, and excess acetic acid as possible. The solid mixture of sodium acetate and sucralose thus produced (36.6 g) was dissolved in about 40 ml of water at 80~C, and t.he resulting solution allowed to cool to room temperature with magnetic stirring and s¢eded with authentic sucralose. After stirring overnight, the ~uct was filtered, washed with a small amount of cold water, and vacuum dried (2S~C/0.5 mm HgJ12 hr). The crystalline solid (20.4 g) was shown by HPLC assay to consist o~ 99.5 wt ~
sucralose (20.3 g, 51.0 mmol, 67.3% yield) and 0.5 wt % other chlorinated sucrose derivatiYes.

Exam~le 3 PURIFICATION OF SUCRALOSE PENTAACETATE BY OPTIONAL
TOLUENE RECRYSTALLIZATION AND CONVERSION INTO SUCRALOSE

In order to challenge the ability of the optional toluene crystallization to purify sucralose pentaester, a DMF-based sucrose-6-acetate syrup (prepared according to the method of Navia) was employed as the starting reaction mixture. This syrup was shown by HPLC
analysis to contain 40.4 wt % sucrose-6-acetate (285 g, 0.742 mol, 67.4~ yield). A com~ination of further analyses showed the syrup to ~o also contain a 7.1 wt % other acetylated sucrose derivatives, 2.1 wt %
unreacted sucrose, 0.1 wt % tin, and 0.1 wt % water, with the remainder being DMF.

The syrup was "doped" with sucrose to a final HPLC assay of 32.3 wt % sucrose-6-acetate, 4.4 wt % other sucrose acetates, and 2.7 wt ~
sucrose. This syrup was chlorinated according to the method described by Navia, and the resulting chlorination product was peracetylated essentially as described in Example 1, except that the steam-strip (DMF removal) operation was not performed (i.e., the ethyl acetate extractions were conducted with the D~F still present in the chlorina-tion product mixture).

~ 2.00 g sample of the crude sucralose pentaacetate thus produced was treated with 5 drops of 20 wt % sodium methoxide in 15 ml of methanol with stirring at room temperature for 120 min. After .

quenching with 5 ~rops of glacial acetic acid, the pro~uct solution was analyzed by HPLC and ~ound to consis~ of ~3.5% sucralose, 6.4 wt % other chlorinated sucrose derivatiYes, and 0.4% sucralo~e 6 acetate, ~asis total ~rh~hydrate contsnt.

A 15.1 g sample of the crude sucralose pentaacetate ~as dissolved in 100 ml of toluene at 80CC, and the so7ution thus produced ~lter~d, cooled, and seeded. After fil~ration and vacuum drying the puri~ied product was found to weigh 13.1 g. A small sample was deacety~ated as described above in Example 2 to pro~ide a product consistin~ of 97.8 wt % sucralose, 1.8 wt % other chlorinated sucrose derivatives, and 0.4 wt % sucralose-6-acetate.

The once-purified sucralose pentaacetate was -cryst~ ed a second time from 100 ml of toluene (10.9 g recovery). A ~m~ll sample was deacetylated as described above to give a product oonsisting of (basis total car~ohydrate content) 99.2 wt % sucralose, 0.~ wt % other chlorlnate~ sucrose deriYatives, and 0.3% sucralose-6-acetate.

Examl~le 4 Acetvlation of crude 6-0-benzoyloxy-a,1' 6~-trichloro~alactosucrose Crude 6-0-~enzoyloxy-4,1',6'-tricholoroqalactos~crose (lQ.1 g, 84.0'~ chrom. purity) was dissolved in a mixture of ethyl acetate (100 mL~ and pyridine (10 mL) in a 3-neck 250-mL round bottom flas~
equipped with a thermometer, nitrogen purge and a drying tube. The solu1:ion was stirred (~agnetic~, cooled to Q~C with an ice bath, acet:ic anhydride (10 mL) was ~ in a sinsle portion, and the solul:ion was allowed to warm to am~ient temperature. The ~-oyLess of t~le reaction was monitored by T.L.C. (toluene~ethyl acetate, 1:1) The reaction was not completa after about 2 hrs., so an addit:ional 5 mL acetic anhydride was added to the solution and the react:ion allowed to continue at ambient tP~rat~re overnight (T.L.C. single spot, Rf 0.65-0.67).

Water (50 mL~ was added to the mixture to destroy unreacted anhydride, the solution was e~tracted with two 25 mL
portions of lN HCl, two 25-mL portions of saturated aqueous bicarbonate, and once with water. The organic layer was ev~porated to a thick syrup (14.7 g) Which was soluble in toluene and methanol, and sparingly soluble in aqueous methanol. A
portion o~ this material was crystalllzed from a saturated solution in a~ueous methanol, to give 6-C~benzoyl-4,1~,6'-trlchloro-4,1',6'-trideoxyqalactosucrose tetraacetate.

Claims

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE PROPERTY
OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. 6-O-benzoyl-4, 1', 6'-trichloro-4,1', 6'-trideoxygalactosucrose tetraacetate.