Indiall Journal of Chemical Technology VoL 6, March 1999, pp. 107-111

Synthesis of cheaper resolving agent, N-alkyl glu.camines Some aspects of process development

Sudip Mukhopadhyay, Ganesh Kumar Gandi & S B Chandalia University Department of Chemical Technology, University of Mumbai, Matunga (East), Mumbai 400019. Indi a

Received I July 1998;-.accepled 30 November 1998

Reducti ve aminati on of D-glucose was stud ied from the view point of process research and development to synthes ize N-alkyl g lucami nes which are recently used as the reso lvi ng agent for racem ic modifications of products contai ning carboxylic ac ids. Different amines like methy l amine, ethyl amine and butyl amines were taken as the aminat ing agent and Raney nickel was used as the catalyst. At a temperature of 85°C and a hydrogen pressure of 95 atm, 97% se lect ivity \Vas achieved at a glucose conversion level of 92%.

Among the various routes for the production of amines from' carbonyl compounds, reductive amination is a process of considerable industrial significance (Table I ) and various fine-chemicals and pharmaceuticals are prepared by this method . This involves the reaction 1·5 of the carbonyl compound with ammonIa or amme and subsequent hydrogenation of the Imme formed . Several investigators claimed the use of Raney nickel catalyst on the reductive amination of various aliphatic and aromatic ketones and aldehydes6

•16

.

Usually, costly resolving agents like cinchonidine are used to separate racemic modifications of products containing carboxylic acid groups. The use of cheaper reso lving agents like N-alkyl glucamine is worth considering. N-a lkyl glucamines generally prepared b d · .. 17 18 fl · d y re uctlve ammatlon · 0 g ucose, IS use as a reso lving agent l 9

-21 for racemic modifications of

products contammg carboxy lic acids . But the informati on rega rding cata lyst loading and life of catalyst for the synthesis of glucamine from D­glucose is inadequate and needs substanti ation . In the present study, attem pts have been made to find out the suitable process parameters from the view point of process research and deve lopment for the reductive aminati on o f gl ucose (Scheme I) and to draw a si mple kinetic interpretati on.

Exper imental Procedure

\la leri a l

D-G lucose, tec hni cal grad(:; methylamine, L.R; ethy!a ll1 ine, L.l{: butylanline_ L. R: Raney nicke l "FI '

type, technical grade were lI seo in this study.

CI-O I

H -?-Q--l

1-0- ?-H

H-?-OH

H-C-OH I C~OH

0(+) - glLCOse

R- ttl ~

[H] / Rarey Ni

aq. rreciiLlll

R = tv1e, Et, rrbLiyt

~ ~C~- r-.l1

I I-rG-OH

I rD-y-H

H-y-OH

I-rC-Q--l I C~OH

N-al~ glu::anire

Scheme I - Process-scheme for reducti ve ami nation of glucose

Experimental set-up

The experiments were carried out in a 100 mL capacity hastalloy B2-Parr autoc lave eq uipped with four bladed magnetically driven impeller. The autoc lave was heated externally by a heating e lement . The temperature of the reaction was maintained within ± I °c of the desired value .

Procedure

Predetermined quantity of D-glucose, N-alkyl­amine and Raney Nickel catalyst were charged in the au toc lave . Before hydrogen was passed , the react ion mixture was allowed to keep under stirrin g fo r I h with a view to complete th e formatio n of imine by the reaction of ammon ia and aldehyde . Subsequently. the hea ting was started to get the desired temperature . Once the desired tcmpera turc was attained . the autoclave was pressur:zed \\ ith hydrogen to the

108 INDIAN J CHEM. TECHNOL., MARCH 1999

Table I-Industrial relevance of~ductive amination

Starting material Product Use

Phenylacetone Amphetamine Stimulators for central-+ammonia nervous system Novoloketone Novoldiamine Intermediate for anti-+ammonia malarial drug Cyclohexanone Cyclohexylamine Manufacture of +ammonia pesticide, organic

intermediate Acetophenone a-phenylethylamine Organic +ammonia synthesis

Materials

D-Glucose (taken) N-Glucamine Secondary amine GJucose unreacted Unaccounted Total

Table 2 - Material balance

Gmol

0.0555 0.04953 0.0008325 0.0039 0.001256 0.0555

%Glucose acccounted for

100 89.24 1.5

7 2.263

100

Reaction conditions Glucose concentration, 20% w/v; temperature, 85°C; mole ratio of glucose : amine, I : 5; reaction time. 3 h; hydrogeJ:) pressure, 90 atm, catalyst loading, 1% w/v; speed of agitation, 1200 rpm, reaction volume, 50 mL.

desired pressure level. A constant pressure was maintained throughout the reaction period.

Analysis The unreacted glucose was estimated by the usual

oximation reaction using hydroxyl amine hydrochloride at 45 °C. The liberated hydrochloric acid was estimated by titrating potentiometrically against standard sodium hydroxide solution.

To ensure any by-product formation in the reaction mixture, samples. were taken for H.P.L.C. The conditions are - coUumn used, C-18 (Octadecyl column); solvent f system, acetonitrile-water (65:35); flow rate, I mLimin; wave length, 254 nm.

Results and Discussion

Material balance The material balance of D-glucose was estimated

for a typical run (Table 2). It would be possible to account 97.7% D-glucose. Unaccounted material balance may be due to loss in handling or error in analytical methods.

Effect of speed of agitation Speed of agitation was varied from 500 to 1200

rpm (Table 3). There was a significant change in the

Table3 -- Effect of speed of agitation on conversion of glucose

Speed of agitation, % Conversion of % Selectivity to rpm glucose product

0 0 500 62 96 800 80 97 1100 92 97 1200 93 97

Reaction conditions : Glucose concentration, 20% w/v; temperature, 850 C; mole ratio of glucose: amine, I : 5; reaction time, 3 h; hydrogen pressure, 90 atm, catalyst loading, I %w/v; reaction volume, 50 mL.

Table 4 - Effect of initial concentration of glucose and period of reaction on overall conversibn

Time, % Conversion for % Conversion for % Conversion for mIn 5% reactant 15% reactant 20% reactant

concentration concentration concentration

0 0 0 0 30 30 32 36 60 51 55 58 120 76 79 81 180 88 89 92 210 92 91 93

Reaction conditions: Temperature, 850 C; mole ratio of glucose : ami ne, I : 5; re~ction time, 3 h; hydrogen pressure, 90 atm, catalyst loading, I %w/v; reaction volume, 50 mL.

overall conversion when the agitator speed was varied from 500 to 1100 rpm indicating the presence of the mass transfer limitation for the diffusion of hydrogen from the gas liquid interface to the bu lk liqu id. When the speed of agitation was further increased to 1200 rpm, there was no significant change in the overall conversion indicating that mass transfer effects are mostly eliminated above 1100 rpm. So all the subsequent reactions were carried out at 1200 rpm.

Effed of partial pressure of hydrogen The partial pressure of hydrogen was varied from

50 to 95 atm (Fig 1). When the pressure was increased from 90 to 95 atm, it was observed, that the overall conversion remained almost same in the above range. Thus, the reaction is independent of partial pressure of hydrogen in the range of 90-95 atm.

Effect of concentration of D-glucose and period of reaction There was a slight change in the overall conversion

of D-glucose, when the concentration of D-glucose was varied from 5-20%w/v (Table 4).

With an increase in the reaction period from 50 min ··to 200 min, the overall conversion of D-glucose increased from 35 to 92% for a 20% initial reactant concentration (Table 4).

~

MUKHOPADHY A Y el at. : SYNTHESIS OF N-ALKYL GLUCAMINES 109

c: 0 e II > c: 0 0

*'

100

90

eo

70

eo

50

40

30

20

10

Time, min Reaction Conditions: Temperature, 8SoC; catalyst loading, 1%wlv; partial pressure of hydrogen, 90 alm;

glucose, 20%w/v; speed of agitation, 1200 rpm; reaction volume, 50 mL.

Fig. I - Effect of hydrogen pressure on overall conversion

c ~ .....

3.5 r------------~----...,

!.O.3"'~1 3 ~ .0. 7 '!1o WIV I

!A l '!IowlV I

! X 12'!1. wlV ~ 2.5

~ 15

0.5

o 50 100 150 200

Time, min

Fig. 2 - - In( I -X,J VS I at different catalyst loading

6 r--------

c >< ..:. 3

2

. 65'C

:. 7S'C

,A 65 'C

Xl 05 'C

250

o~~=====--__ --.J o 50 100 150 200

Time, min

Fig. 3 --In(l-XA) vs I at different temperature

-3r-------------------,

-3.5

~ .E -4S

-5

-5.5

• ~ ~--------------------~

2.6 265 2.7 2 75 2.8 285 2 9 295

11T X 103, K-1

Fig. 4 - Arrhenius plot

Effect of catalyst loading The catalyst loading expressed as weight percent of

catalyst based on the total reaction volume was varied from 0.3% to 1.2% w/v (Table 5). The overall conversion increased significantly with an increase in catalyst loading .

Effect of temperature The temperature was varied in the range of 65 to

105 °C (Table 6). The conversion of D-glucose increased from 38 to 98%. The selectivity with respect to glucami-ne was maximum at 85 °C. At 105 °C, the selectivity was as low as 69% due to residue forl'hation .

Effect of mole ratio The mole ratio of aldehyde to ammonia was varied

from I: I to I : lOusing 1% w/v of catalyst at 85 °c (Table 7). As expected, with increase in the mole ratio, the selectivity with respect to the glucamine increased. From a pragmatic point of view, one may use a mole ratio of 1 :5, since the selectivity obtained was as high as 97%.

Effect of different type of catalyst and reusability of the catalyst

Different type of nickel catalysts obtained fro m different sources were examined (Table 8) for this reaction and it was found that Raney nickel " F-Type" was the most effective. Attempts were made to see whether the catalyst can be used repeatedly. However, results obtained indicating that with fresh catalyst 92% conversion was achieved in 3 h, in the very first reuse the conversion of D-glucose was 81 % and in the

110 INDIAN 1 CHEM. TECHNOL., MARCH 1999

Table 5 - Effect of catalyst loading on conversion and selectivity

Catalyst loading, % w/v % Conversion % Selectivity

0 0 0.3 42 97 0.7 69 96 1.0 92 97 1.2 93 97

Reaction conditions : Glucose concentration, 20% w/v; temperature, 850 C; mole ratio of glucose: amine, I : 5; reaction time, 3 h; hydrogen pressure, 90 atm; reaction vo lume, 50 mL.

Table 6 - Effect of temperature

Temp,oC % Overall % Conversion % Selecti vi ty to conversion to product the product

65 38 36 97 75 60 57 96 85 81 79 97 105 96 66 69

Reaction conditions : Glucose concentration, 20% w/v; mole I16tio of glucose: am ine, I : 5; reaction time, 2 h; hydrogen pressure, 90 atm, catalyst loading, 1% w/v; speed of agi tation, 1200 rpm, reaction volume, 50 mL.

Mole ratio of %Overall % Conversion % Selectivi ty glucose: con version to product with respect to

N-methylamine the product

I : 1 79 71.1 90 1 : 2 87 82.65 95 I : 5 93 90.2 1 97 I : 7 94 9 1.1 8 97

I : 10 94 92 .12 98

React ion condit ions : Glucose concentration, 20% w/v; reaction temperature, 850 C; reaction time, 3 h; hydrogen pressure, 90 atm. catalyst load ing, 1% w/v: speed of agitation, 1200 rpm. reaction vol ume, 50 mL.

second reuse the conversion was only 70% (Table 9). In a view to use the catalyst in success ive runs without loss of much activity 15% wlw fresh cata lyst was added to the 2nd reused catalyst -and it was observed that the convers ion leve led off with that of fresh one. Therefore, catalyst could be reused repeated ly if 15% fresh cata lyst was mixed with the spent one.

Effect of differe nt type of a mines

Different type of N-a ikyl am ines like methy l amine, ethyl am ine and buty l amine were used . It was observed (Table 10) that a lmost salTle conversion leve l was obta ined w ith th ese arn ines for a spec ified reaction time.

Kinetic interpretation T he reactions are carried out at a high speed of

Table 8 - Effect of different type of catalyst on overall conversion and selectivity

Catalyst

Raney Nickel Supported Nickel (HLL) Supported Nickel (Engelhard)

% Conversion

81 77 60

% Selectivity

97 96 95

Reaction cond itions : Glucose concentration, 20% w/v; reaction temperature, 850 C; reaction time, 2 h; hydrogen pressure, 90 atm, catalyst loading, 1% w/v; speed of agitat ion, 1200 rpm, reaction vo lume, 50 mL.

Table 9 - Reusability property of Raney Nickel catalyst

Catalyst used

Fresh First reuse Second reuse Second reuse + 15"'10 w/w of fresh catalyst

% Conversion

92 81 70 9 1

% Selectivity

97 96 96 97

Reaction conditi ons: Glucose concentration, 20% w/v; reaction temperature, 85°C; react ion time, 3 h; hydrogen pressure, 90 atm, catalys t loading, 1% w/v; speed of agi tat ion, 1200 rpm, reaction volume, 50 mL.

Table 10 - Effect of different amines

Substrate % Conversion % Selectivity

N-methyl amine 93 97 N-ethyl ami ne 9 1 95 N-butyl amine ' 89 96

Reaction condi tions : Glucose concentration, 20% w/v; temperature, 850 C; mole rat io of glucose : amine, ) : 5; reaction time, 3 h; hydrogen pressure, 90 atm, catalyst loading, ) % w/v; speed of agi tation , 1200 rpm, reac ti on volume. 50 mL.

ag itation to eliminate the mass transfer effects. There was absolute ly no increase in the conversion leve l when hydrogen pressure is increased from 90 atm to 95 atm. It indicates that at or above 90 atm pressure the reaction is zeroth order w ith respect to hydrogen.

. It is seen from Table I I, that for a g iven initial concentration of the reactant the reaction is first order w ith respect to the g lucose at a hydrogen pressure of 90 a tm, speed of ag itati on o f 1200 rpm and a catalyst loading of 1 % ,!"/v .

From the s lopes of Fig 2, the rate constants are fo und out and given in Tab le 12. It is observed that the initia l rate in creases li nearly with the increase in catalyst load ing. Since fo r the s ize of the catalyst used and conditions empl oyed, the ro le of diffusion of hydrogen from the bulk liqu id to the so lid is estimated to be unimportant, the data appears to represent the true kinetics of the process at a cata lyst loading of 1.0% w/v.

..ot

MUKHOPADHY A Yet af.. SYNTHESIS OF N-ALKYL GLUCAMINES III

Table 11 - Effect of initial concentration of glucose on initial rate constant

Initial concentration of glucose, %w/v 5 15 25

k, sol atm-' gear'

4.2 x 10-6

4.4 X 10-6

4.9." 10-6

Table 12 - Effect of catalyst loading on initial rate constant

Catalyst loading, %w/v

0.3 0.7 1.0 1.2

k, sol atm-' gear'

3.3 x 10-6

3.7 X 10-6

4.9 X 10-6

5.0 X 10-6

At different temperature, -In (l -XA) vs t is plotted (Fig 3) and from the slopes, k at different temperatures are found to be I .3x 10-6

, 2.8x 10-6 ,

4.9x I0·6 and 9.8x I0-6 S- I atm- I g-caC I at 65,75,85

and 105 °c under the reaction conditions. Energy of activation is found to be 12.3 kcal/mol from the slope of the Arrhenius plot (Fig 4).

Conclusion The reductive amination of D-glucose to N-methyl

glucamine gave as high as 97% selectivity towards N­methyl glucamine using Raney nickel as catalyst.

The reusability problem of Raney nickel catalyst was solved and it will definitely cut short the process cost.

A high speed of agitation of 1200 rpm was used to eliminate the mass transfer effects to study the true kinetics of the reaction.

The reductive amination of D-glucose was found to be zeroth order with respect to hydrogen concentration above a hydrogen partial pressure of 90 atm.

It would be necessary to allow the mixture of aldehyde and ammonia to react for I h before hydrogenation because the reduction to alcohol could be a competing side reaction .

Under the best suitable conditions (D-glucose 20%w/v; mole ratio of glucose: methyl amine, 1 :5;

temperatur~, 85°C; solvent, water;· hydrogen pressure 95 atm.; catalyst loading 1 % w/v; reaction period, 3h; speed of agitation, 1200 rpm), D-glucamine could be obtained with 97% selectivity when the overall conversion of D-glucose was 92%.

Acknowledgement Ganesh kumar Gandi and Sudip Mukhopadhyay

are grateful to the University Grants Commissions, New Delhi, for the a~ard of a Junior and Senior research fellowships, respectively.

References 1 Radisson J (SANOFI); Eur Pat Appl EP 321,349, ( 1990), cf,

Chern Abstr, 112, 35671. 2 Friefelder M. Practical catalytiC hydrogenation techniques

and application. (Wiley Interscience, New York), (1971 ). 3 Adam S R, OrganiC Reactions. Vol. 4. 1st. Edn., (John Wiley

and Sons, New York), ( 1947). 4 Schwaegler L & Adkins H, J Am Chern. Soc, 6 1 ( 1939) 3499. 5 Winans C F,;Adkins H, J Am. Chern. Soc, 55 ( 1933) 205 1. 6 Brimmer M R, US Pat 2,518,659, ( 1950) . 7 Omelo Y, Masuda K & Nakamora J, Koggyo Kagaku Zashi.

70 (1979) 1343 . 8 Tamaki A, Yamamoto K, Samajima M, Jpn Kokai Tokf...yo

Koho JP 74 132,534, ( 1979); cr. , Chern Abstr, 92, 146409. 9 Woyori G, Jpn Kokai Tokkyo Koho JP 79 103 ,804,( 1979). cf..

Chern Abstr, 92. 110646. 10 Taira S, Jpn. Kokai Tokkyo Koho Jf 60 248, 647. ( 1985); cr. ,

Chern Abstr, 104, 129609. II Milton G, (Polaroid Corp.): US Pat 3, 187,047, (1 965). Chem

Abstr, 63 , 9873 . 12 Ometo Y, Masuda K. Nakamora J, Koggyo Kagaku Zashi 70

(1967) 1335. 13 Shah S A, Reductive amination of ketones. M. Chem. Eng.

Thesis, Uni versi ty of Bombay, ( 1986). 14 Trehan B S A, Production of pharmaceutical intermediates.

M. Chern . Engg. Thesis, University of Bombay, 1986. 15 Satagopan V, Reduction and reductive amination of organic

compounds. Ph.D. Thesis, Uni versity of Bombay, 1990. 16 Kar S, Studies in process development Synthesis of

pharmaceuticals and agrochemicals. PhD. Thesis. Uni versity of Bombay, 1996.

17 Holton P G, (Syntex Corp.), US Pat. 4,515 ,811 , ( 1985 ). 18 Arnold R A & Mathews G J, Ger Pat. 2,805,488 , ( 1978). 19 Masahito S, Chern Abstr, I 14, 20726. 20 Masaru S&Takehijof, ChemAbstr, ( 1990), 11 3, 131843. 21 Harrington P J & Lodcwijk E, Org Proc Res Dev. I ( 1977)

72.