Round Granules of Dimethyl Fumarate by Three-in-One Intensified Process of Reaction, Crystallization, and Spherical Agglomeration in a Common Stirred Tank

Article abstract of DOI:10.1021/acs.oprd.7b00183

Pure, isomorphic, round, and free-flowing dimethyl fumarate granules in a size range of 250-2000 μm were successfully produced directly from esterification through the three-in-one intensified process of three distinctive steps of reaction, crystallization, and spherical agglomeration (SA) in a 0.5 L sized jacketed glass stirred tank. Dimethyl fumarate was prepared by sulfuric acid catalyzed esterification of fumaric acid with methanol. The reaction temperature was below the maximum allowable limit of 65 °C as determined by reaction kinetics to avoid the runaway situation. The dissolution rate of primary crystals of dimethyl fumarate was inversely proportional to the particle size which was strongly affected by the antisolvent addition and temperature cooling schemes during crystallization. However, the dissolution rate of the round granules was mainly dependent on the exterior dimension of the granules and not so much on the primary crystal size inside the granules. The mechanical properties such as density, porosity Carr's index, friability, and fracture force of round dimethyl fumarate granules generated in (1) three-in-one processes with the final temperatures at either 5 or 25 °C (Three-in-one I and II) and (2) SA of dimethyl fumarate, which was done separately and disconnected from the train of reaction and crystallization process at either 5 or 25 °C (SA I and II), were thoroughly studied and compared. The concept of scale-up for Three-in-one I and II was also verified in a 10 L sized jacketed glass stirred tank. Powder manufacturability such as flowability, blend uniformity, and compressibility had been substantially enhanced by spherical agglomeration. The added values of free-flowing and easy-to-pack properties to dimethyl fumarate in addition to its original intrinsic slip planes in the crystal lattice would make direct compaction into tablets feasible.

Full text of DOI:10.1021/acs.oprd.7b00183

Subscriber access provided by AUBURN UNIV AUBURN
Full Paper
Round Granules of Dimethyl Fumarate by Three-in-
One Intensified Process of Reaction, Crystallization and
Spherical Agglomeration in a Common Stirred Tank
Chih Wei Chen, and Tu Lee
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.7b00183 • Publication Date (Web): 31 Jul 2017
Downloaded from http://pubs.acs.org on July 31, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Organic Process Research & Development is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Round Granules of Dimethyl Fumarate by ThreeꢀinꢀOne
Intensified Process of Reaction, Crystallization and Spherical
Agglomeration in a Common Stirred Tank
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*
Chih Wei Chen and Tu Lee
Department of Chemical and Materials Engineering, National Central University,
3
00 Jhongda Road, Jhongli District, Taoyuan City 32001, Taiwan R.O.C.
ABSTRACT
Pure, isomorphic, round and freeꢀflowing dimethyl fumarate granules in a size
range of 250 to 2000 ꢀm were successfully produced directly from esterification
through the threeꢀinꢀone intensified process of three distinctive steps of reaction,
crystallization and spherical agglomeration (SA) in a 0.5 Lꢀsized jacketed glass stirred
tank. Dimethyl fumarate was prepared by sulfuric acidꢀcatalyzed esterification of
fumaric acid with methanol. The reaction temperature was below the maximum
allowable limit of 65°C as determined by reaction kinetics to avoid runaway situation.
The dissolution rate of primary crystals of dimethyl fumarate was inversely
proportional to the particle size which was strongly affected by the antisolvent
*
Corresponding Author
Phone: +886ꢀ3ꢀ422ꢀ7151 ext. 34204. Fax: + 886ꢀ3ꢀ425ꢀ2296.
Eꢀmail: tulee@cc.ncu.edu.tw.
1
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
addition and temperature cooling schemes during crystallization. However, the
dissolution rate of the round granules was mainly dependent on the exterior dimension
of the granules and not so much on the primary crystal size inside the granules. The
mechanical properties such as density, porosity Carr’s index, friability and fracture
force of round dimethyl fumarate granules generated in: (1) threeꢀinꢀone processes
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
o
with the final temperatures at either 5 C or 25 C (Threeꢀinꢀone I and II), and (2) SA of
dimethyl fumarate was done separately and disconnected from the train of reaction
o
o
and crystallization process at either 5 C or 25 C (SA I and II), were thoroughly
studied and compared. The concept of scaleꢀup for Threeꢀinꢀone I and II was also
verified in 10 Lꢀsized jacketed glass stirred tank. Powder manufacturability such as
flowability, blend uniformity and compressibility had been substantially enhanced by
spherical agglomeration. The added values of free flowing and easyꢀtoꢀpack
properties to dimethyl fumarate in addition to its original intrinsic slip planes in the
crystal lattice would make direct compaction into tablets feasible.
KEYWORDS
Process Intensification, Dimethyl Fumarate, Reaction, Crystallization, Spherical
Agglomeration
2
ACS Paragon Plus Environment

Page 3 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
INTRODUCTION
Pharmaceutical companies have long competed mainly through the innovation
of new drugs for medical needs. However, new biotechnologies and paradigms,
such as proteomics, genomics and the use of biomarkers, have made the situation
more competitive. Many customized small volume drugs for patients’ specific
genetic makeup have been generated and translated into manufacturing of many more
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
futuristic products. In addition, the increases in competition and proportion of
generic utilization, opening of new markets, and the socioeconomic pressures for
price controls, have driven the industry to look for other ways to improve quality
2
while reducing manufacturing costs which account for 36% of the industry’s cost.
3
One possible solution is continuous processing of solid dosage products from
the synthesis of active pharmaceutical ingredient (API) to the final packaging of
tablets or capsules all year round because of less scaleꢀup difficulties, smaller in
equipment size, shorter in residence time, better process control, enhanced margins of
4,5
safety, increase in production rate (> 50 tons/day), and improved quality and yields.
However, the industry as a whole is reluctant to embrace continuous manufacturing
because: (1) batch processing equipment in many ways is the most flexible and
convenient, and unit operations such as API synthesis, milling, blending, roller
compaction, extrusion and compression, are not continuous, (2) the additional capital
3
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
investment for the current vast disposal of batch process equipment and staff training
required is enormous, and (3) the particle size distribution (PSD) of the APIs are
6
highly process dependent.
A tight control of PSD is critical because it has
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
significant effects not only on final drug product performance such as dissolution and
bioavailability, but also on powder manufacturability such as flowability, blend
uniformity and compressibility. For example, API particles may suffer from
troublesome problems in powder processing such as wall adhesion, ratꢀholing,
bridging and segregation due to its small size and inhomogeneity. In addition,
pharmaceutical particles, which usually range in size from 1 to 120 ꢀm, are required
to mix with other excipients to granulate before tablet compaction or capsule filling,
7
and granule size used in solid dosage products often ranges from 0.2 to 4.0 mm.
The size range of 0.2 to 4.0 mm is hard to be achieved by seeding and milling.
8
, 9
9
Therefore, granulation methods such as roller compaction,
spray drying,
1
0
11 , 12
9
supercritical fluid,
high shear mixing,
fluid bed granulation, and
13
extrusion/spheronization
have been successful in manufacturing of various
pharmaceutical solid dosage forms. However, those methods are complicated,
timeꢀconsuming, requiring a large space and many expensive equipment. On the
1
4 ꢀ 17
other hand, spherical agglomeration (SA)
is an appealing technique for
intensifying some of the processes especially when round API granules composed of
4
ACS Paragon Plus Environment

Page 5 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
small and highꢀaspectꢀratio API crystals can be prepared by SA right after
crystallization in the same stirred tank. The relatively large round API agglomerates
can improve filtration and dryer performance in downstream. Their excellent
flowability will allow them to be blended directly and uniformly with the excipients in
the formulation step as well.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Therefore, our aims in this project are twofold: to achieve satisfactory powder
flowability of API on one hand, and a reasonable compromise between batchꢀcentered
mentality and process intensification on the other. These can be realized by
immediate preparation of API granules right after chemical reaction and
crystallization with a high purity profile in the same common stirred tank by SA,
soꢀcalled “threeꢀinꢀone intensified process.” In this type of threeꢀinꢀone intensified
process, filtration and drying steps are needed only once after SA, whereas in the
14ꢀ17
previous studies on SA,
without having the chemical reaction of solid materials,
SA had been achieved mainly through reꢀcrystallization. That is, the filtration and
drying steps were carried out twice, one after the chemical reaction and crystallization,
and the second one after SA. Hopefully, some of the downstream activities in
formulation such as granulation used for enhancing particle property will soon to be
reduced by implementing the threeꢀinꢀone intensified process of reaction,
5, 18
crystallization and spherical agglomeration.
Its robustness may be readily
5
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
adopted for contract development and manufacturing organizations (CDMO) and
manufacturing companies in APIs, generic drugs, nutraceuticals and other fine
chemicals whose raw materials are supplied from various sources, and main concerns
are flexibility, capacity optimization and cost reduction for staying competitive and
increasing the market share.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Taking the API as an example, basic SA involves suspending the API crystals
in a poor solvent and agglomerating them by the addition of bridging liquid. The
bridging liquid is immiscible with the poor solvent, and wets the suspended API
1
9
crystals preferentially.
When SA is coupled with reꢀcrystallization, the API crystals
2
0
are firstly dissolved in a good solvent, crystallized out by adding a poor solvent, and
2
1
then coagulated by introducing the bridging liquid with polymers.
When SA is
coꢀcrystallization/reꢀcrystallization and
agglomeration are controlled separately to tailor the crystal size internally and the
2
2
coupled with coꢀcrystallization,
16
agglomerate size externally, and sometimes, with the assistance of excipient
23
19
16
surface.
So far, SA has only been implemented to batch, twoꢀstage continuous,
17
24
continuous, and microfluidic modes,
and conducted indirectly on crystals
25,26
27
produced from neutralization
and solvothermal process in separate vessels.
However, SA has not yet been coupled with the train of reaction and crystallization,
such as esterificationꢀcrystallization system. Since our research group has had
6
ACS Paragon Plus Environment

Page 7 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
9
experiences with SA of dimethyl fumarate suspension in a common stirred tank, we
would like to continue with the same system, but this time, promoting the telescopic
concept of integrating chemical reaction, crystallization and SA seamlessly in the
same stirred tank.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
®
Dimethyl fumarate is chosen as a model API (Tecfidera ) because of its
multiple applications in antifungal, antiseptic and antimoth in food industry, and the
28
treatment of psoriasis, and multiple sclerosis.
High purity dimethyl fumarate can
also be synthesized easily by esterification of fumaric acid with methanol in the
2
9
presence of homogeneous acid catalyst such as sulfuric acid.
Esterification is an
important but common reaction for making organic solvents, fragrances, flavor
chemicals, agrochemicals and pharmaceuticals. In addition, due to the inherent slip
planes in the crystal lattice of dimethyl fumarate, direct compaction should have been
3
0
technically feasible if not because of its undesirable plateꢀlike habit which has
prevented the dimethyl fumarate crystals from freeꢀflowing in the tableting machine.
To enhance its manufacturability, dimethyl fumarate crystals would need to be
agglomerated first into round granules. As a first step, small primary crystals will
have to be generated from a high degree of supersaturation to serve as ideal building
blocks. This could be achieved by antisolvent addition such as cold water addition,
and temperature cooling at the end of esterification reaction, to create a large
7
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
temperature gradient. Therefore, esterification will preferably be carried out at a
high temperature but below the maximum allowable reaction temperature as
determined by the reaction kinetics (Supporting Information) to avoid the runaway
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
situation due to the low boiling point of the methanolꢀbased reaction medium.
Since a wide range of particle size distribution (PSD) of API can be achieved
in the crystallization step, and the PSD of API can in turn influence the growth
mechanism of the round granules in SA, round granules of different sizes will be
3
2
produced which are easily filtered and dried. Strategically, the two separate events
16
of crystallization and SA can be orchestrated to tune the majority of round granules
to fall mainly in the range of 250 to 2000 ꢁm at will which is difficult to achieve by
conventional methods, such as seeding, within a reasonable batch time. The absence
of filtration and drying steps between the nonꢀstop action of crystallization and SA
steps during the threeꢀinꢀone intensified process also allows the favorable production
of small sized crystals in the crystallization step without worrying that they would
create any high wet cake resistance in filtration or powder lumping in drying, which
33
usually take place after the crystallization step.
In the threeꢀinꢀone intensified
process, reaction conversion, crystal yield, agglomerate yield, the purity levels in the
solution and solid phase, granular flowability, friability, strength, tabletability and
dissolution rate will be fully investigated. Finally, to pave the way for scaling up the
8
ACS Paragon Plus Environment

Page 9 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
threeꢀinꢀone intensified process for dimethyl fumarate in the future, a 10 Lꢀsized
vessel will be used to prove the feasibility as well.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
. MATERIALS AND METHODS
.1. Chemicals and Solvents
2
o
Dimethyl fumarate (C
6
H
8
O
4
, 99% purity, MW = 144.13, mp = 102ꢀ105 C, Lot:
o
1
1
0177580) and fumaric acid (C
4
H
4
O
4
, 99% purity, MW = 116.07, mp = 299 C, Lot:
0175189) were purchased from Alfa Aesar (Heysham, England). Monomethyl
o
fumarate (C
5
H
6
O
4
, 97% purity, MW = 130.1, mp = 144ꢀ145 C, Lot: MKBW0455V)
14 5
was purchased from Aldrich (St. Louis, MO, USA). Phenol red (C19H O S, MW =
o
354.38, mp > 300 C, Lot: N07A006) was purchased from Alfa Aesar (Ward Hill, MA,
USA). Sodium hydroxide (NaOH, 97% assay, MW = 40.00, mp = 318°C, Lot:
o
KXꢀ2842), dimethyl sulfate (C
2
H
6
O
4
S, 99% purity, MW = 126.13, bp = 188 C, Lot:
o
KWꢀ3436C) and sulfuric acid (H
2
SO
4
, 97% assay, MW = 98.08, bp = 337 C, ρ = 1.84
3
g/cm , Lot: 4A267) were received from Showa Chemical Co. Ltd. (Tokyo, Japan).
o
Methanol (CH
3
OH, 99.9% assay, MW = 32.04, bp = 64.7 C, Lot: 14050368) was
6 5 3
purchased from Tedia Company Inc. (Fairfield, OH, USA). Toluene (C H CH , ACS
o
grade 99.5% purity, MW = 92.14, bp = 111 C, Lot: ETA140403) and ethanol (C
2
H
6
O,
o
9
9.5% purity, MW = 46.7, bp = 78 C, Lot: 262611) were obtained from Echo
9
ACS Paragon Plus Environment

Organic Process Research & Development
Page 10 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6 6 3 2
Chemical (Miaoli, Taiwan). Dimethyl sulfoxideꢀd (DMSOꢀd , (CD ) SO, 99 atom
%
D, MW = 84.17, Lot: A0372803) was purchased from Acros Organics (New Jersey,
o
3
NJ, USA). Acetonitrile (CH CN, HPLC grade, MW = 41.05, bp = 82 C, Lot:
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0
000142848), and monopotassium phosphate (KH
2
PO
52.6 C, Lot: 0000154227) were purchased from Avantor Performance Materials, Inc.
Phosphoric acid (H PO , 85% assay, MW = 97.994, bp
158 C, Lot: K3422573 504) was purchased from Merck KGaA (Darmstadt,
4
, MW = 136.086, mp =
o
2
(Center Valley, PA, USA).
3
4
o
=
Geremany). Reversible osmosis (RO) water was clarified by a water purification
system (model MilliꢀRO Plus) bought from Millipore (Billerica, MA, USA).
2.2. Three-in-One Intensified Process of Reaction, Crystallization and
Spherical Agglomeration of Dimethyl Fumarate
Threeꢀinꢀone intensified process was carried out in a 0.5 Lꢀsized jacketed glass
stirred tank with an inner diameter of 7.5 cm, equipped with four equally spaced
o
vertical baffles with a width of 9 mm, a fourꢀbladed 45 turbine impeller with a
19
diameter of 3.5 cm and the clearance of the impeller from the vessel bottom was
lowered to 1.5 cm due to the small solution volume, a water jacket and a total reflux
condenser to prevent product loss.
For esterification, 8.357 g (0.072 mol) of fumaric acid were introduced to a 0.5
10
ACS Paragon Plus Environment

Page 11 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Lꢀsized jacketed glass stirred tank. The tank was then heated to the maximum
allowable reaction temperature of 65°C (Supporting Information). 100 mL of
preheated methanol at the corresponding reaction temperature were then rapidly
added. The agitator was set to at 600 rpm. Once fumaric acid was dissolved, to
initiate the reaction, 0.96 mL of sulfuric acid (molar ratio of fumaric acid to sulfuric
acid = 4:1) was fed into the stirred tank by a glass pipette in one portion below the
liquid surface and near the tip of the impeller for good microꢀmixing. The total
reaction time was 3 h. The percent reaction conversion is defined as: the number of
moles of reacted fumaric acid that divided by the initial 0.072 mol of fumaric acid fed
to the stirred tank multiplied by 100%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For crystallization, the stirring speed was kept at 600 rpm. The clearance of
the impeller from the vessel bottom was adjusted from 1.5 to 3 cm because of (1) the
1
9
increase in solution volume, and (2) the consistency with our last study.
300 mL of
cold water as an antiꢀsolvent at either 5 C or 25 C were rapidly fed into the same
stirred tank near the tip of the agitator through a funnel within about 1.5 min. The
o
o
o
o
o
stirred tank was cooled to either 5 C or 25 C with the cooling rate of 0.86 C/min for 3
h. Percent crystal yield is defined as: the actual yield of crystals, in grams, divided
by the theoretical yield of crystals, in grams, based on the reaction conversion
multiplied by 100%, assuming that the solubility of dimethyl fumarate is negligible in
11
ACS Paragon Plus Environment

Organic Process Research & Development
Page 12 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the 1:3 (v/v) methanolꢀwater mother liquor from 5° to 25°C.
1
9
For spherical agglomeration, 6 mL of toluene as a bridging liquid in a glass
pipette were added rapidly to the dimethyl fumarate suspension below the liquid
surface, near the tip of the impeller, which was stirred for 2 h. After the 2ꢀh
maturation time, the spherical agglomerates were filtered out, and oven dried at 40°C
until no change in weight. The percent agglomerate yield is calculated by dividing
the actual yield of agglomerates, in grams, by the theoretical yield of agglomerates, in
grams, based on the reaction conversion multiplied by 100%, assuming that the
solubility of dimethyl fumarate is negligible in the 1:3 (v/v) methanolꢀwater mother
liquor from 5° to 25°C.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.3. Disconnection of Spherical Agglomeration of Dimethyl Fumarate
SA of dimethyl fumarate was done separately and disconnected from the train
of reaction and crystallization process. However, the operation conditions for the
remaining two steps of reaction and crystallization of dimethyl fumarate were
unchanged. Filtration would now be carried out after 3 h of crystallization process,
o
and oven drying of wet cakes was performed at 40 C until no change in weight. For
the disconnected SA, about 7 to 8 g of dried dimethyl fumarate crystals were first
introduced to a 0.5 Lꢀsized jacketed glass stirred tank, and 400 mL of water at the
12
ACS Paragon Plus Environment

Page 13 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
corresponding crystallization temperature of 5° or 25°C were then added. The
agitator was set to 600 rpm to make the slurry suspend. 6 mL of toluene as a
bridging liquid in a glass pipette were added rapidly to the dimethyl fumarate
suspension below the liquid surface near the tip of the impeller for good micromixing,
which was stirred for 2 h. After the 2ꢀh maturation time, the round granules were
filtered out and oven dried at 40°C until no change in weight.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
.4. Large Scale for Three-in-One Intensified Process of Dimethyl Fumarate
Threeꢀinꢀone intensified process was carried out in a 10 Lꢀsized jacketed glass
stirred tank with an inner diameter of 20.4 cm, equipped with four equally spaced
o
vertical baffles with a width of 2.8 cm, a fourꢀbladed 45 turbine impeller with a
diameter of 7 cm having a clearance from the vessel bottom of 3 cm, and a total reflux
condenser to prevent product loss. For esterification, 167.14 g (1.44 mol) of fumaric
acid were introduced to a 10 Lꢀsized tank. 2 L of preheated methanol at 65°C were
then rapidly added. The agitator was set to at 350 rpm based on the Zwietering
19
equation.
Once fumaric acid was dissolved, to initiate the reaction, 19.2 mL of
sulfuric acid (molar ratio of fumaric acid to sulfuric acid = 4:1) was fed into the
stirred tank by a dropping funnel in one portion below the liquid surface and near the
tip of the impeller for good microꢀmixing. The total reaction time was 3 h. For
13
ACS Paragon Plus Environment

Organic Process Research & Development
Page 14 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
crystallization, the stirring speed was kept at 350 rpm. The clearance of the impeller
from the vessel bottom was adjusted from 3 to 7 cm because of the increase in
1
9
o
o
solution volume.
6 L of cold water as an antiꢀsolvent at either 5 C or 25 C were
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
rapidly fed into the same stirred tank near the tip of the agitator through a dropping
funnel within about 5 min. For spherical agglomeration, 120 mL of toluene in a
dropping funnel were added rapidly to the dimethyl fumarate suspension below the
liquid surface, near the tip of the impeller, which was then stirred for 2 h. After the
2
4
ꢀh maturation time, the spherical agglomerates were filtered out, and oven dried at
0°C until no change in weight.
2.5. Analytical Methods for Reaction Kinetics Study
The kinetics of esterification was studied by the method of initial rates for a set
of initial fumaric acid concentrations of 0.36, 0.54 and 0.72 M in 100 mL solution at a
o
o
o
constant temperature of 45 , 55 and 65 C (Supporting Information). The maximum
allowable reaction temperature was determined by the reaction kinetics (Supporting
Information) to avoid the runaway situation.
2
.5.1. Gravimetric Analysis
ꢀmL samples were withdrawn periodically through a glass pipette and titrated
1
14
ACS Paragon Plus Environment

Page 15 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
against 1 M of methanolic NaOH solution with the addition of a drop of phenol red
3
4
pH indicator solution to determine the reactant concentration of fumaric acid.
An
excess amount of NaOH would be added to ensure all unreacted fumaric acid was
converted to white sodium fumarate precipitate as indicated by the color change of
phenol red from yellow to bright pink in the mother liquor. Each sample was
centrifuged at 6000 rpm for 10 min and then decanted. To avoid the undesired
dimethyl fumarate crystals mixed with sodium fumarate in the wet centrifuged cake,
the cake was rinsed with copious amount of methanol, centrifuged again with a speed
of 6000 rpm for 10 min and decanted. All sodium fumarate wet cakes were vacuum
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
o
dried at 40 C until no change in weight. The weight of each wet cake at different
time point was used to estimate the conversion of fumaric acid as a function of time.
2
.5.2. Nuclear Magnetic Resonance (NMR) Analysis
To verify the validity of the simple and onꢀsite gravimetric analysis, 1ꢀmL
samples were also withdrawn periodically from the 0.5 Lꢀsized jacketed glass stirred
tank to determine the reactant concentration of fumaric acid in one of the trials with
an initial fumaric acid concentration of 0.72 M at 45°C. All withdrawn samples
were immediately dropped into an ice bath to extinguish esterification. All samples
were vacuum dried at 0°C for 1 h to evaporate off methanol, water, monomethyl
15
ACS Paragon Plus Environment

Organic Process Research & Development
Page 16 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
sulfate (MMS), dimethyl sulfate (DMS), and sulfuric acid. Dimethyl fumarate solids
mixed with unreacted fumaric acid were obtained and their sublimation was
minimized at such a low temperature. About 50 mg of lyophilized reaction mixture
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
were dissolved in an adequate amount of deuterated dimethyl sulfoxide (DMSOꢀd ).
6
About 0.8 mL of the DMSOꢀd sample solution was added into an NMR tube having a
1
5
ꢀmm O.D. and 178 mm in length for solution H NMR quantitative analysis. The
concentration of each sample was calculated by comparing the integral peak area for
3
the –COOꢀCH ester proton chemical shift at 3.74 ppm with the one from a 0.25 M
dimethyl fumarate standard solution prepared by dissolving the purchased dimethyl
fumarate with a purity of 99% in DMSOꢀd
Bruker (Billerica, Massachusetts, USA).
6
through the Top Spin 3.5 software from
2
.5.3. Solubility Test
About 10 mg of solid samples of dimethyl fumarate were weighed in a 20 mL
scintillation vial. Solvent or solution drops were titrated carefully by a micropipet
into the vial with intermittent shaking until all solids were just dissolved. The
solubility value of solids in the solvent or solution at a given temperature was
calculated as the weight of solids divided by the total volume of solvent or solution
added to the vial. Solubility values of solids in a solvent or solution at four different
16
ACS Paragon Plus Environment

Page 17 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
o
o
o
o
temperatures of 15 , 25 , 40 and 60 C were determined in a water bath. Although
3
5
this particular gravimetric method for solubility determination appeared to have an
inherent inaccuracy of about ±10%, its advantages were its robustness, simplicity,
without the need of performing any calibration and the concern of solvate formation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
.6. Sieve Analysis Method
To determine the size distribution of dimethyl fumarate crystals/agglomerates,
a series of metal sieve plates from the largest to the finest aperture were utilized in the
order of 2000, 1410, 1000, 710, 500, 355, 250, 180, 125 and 90 ꢁm (Kuang Yang,
Taiwan). About 1 to 5 g of dried samples were fed onto the 2000 ꢁm sieve plate
directly and vibration was then generated by holding the 2000 ꢁmꢀsized sieve plate
with one hand and tapping the sieve plate sideways with a spatula by another until no
more powders (or agglomerates) on the 2000 ꢁmꢀsized sieve plate passed through by
eye. The same procedure was carried out repeatedly on the following smallerꢀsize
opening sieve plates until the finest powder passed through the 90ꢀꢁm sized sieve
plate and collected by the bottom receiver. The weight of the remaining particles on
each sieve plate was recorded and then divided by the total powder weight to obtain
the corresponding mass percent on sieve plotted as a histogram and cumulative mass
percent larger. With this method, there was no need to worry about factors which
17
ACS Paragon Plus Environment

Organic Process Research & Development
Page 18 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
may affect the end results like loading and time.
2
.7. Mechanical Properties of Spherical Agglomerates
.7.1. Density of Spherical Agglomerates
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
0 to 80 round granules of dimethyl fumarate in the size range of 710 to
1
1
000 ꢁm were selected and weighted. The diameter of round granules was 710 and
000 ꢁm on the average assuming each agglomerate was a sphere. The density was
calculated by dividing the mass of all agglomerates with the total spherical volumes of
agglomerates.
2.7.2. Porosity of Spherical Agglomerates
Porosity is a measurement of the void spaces in a material and its value is
between 0 and 1. If the void spaces are filled with air, the porosity, ϕ, can be defined
as:
ρbulk
φ
= 1ꢀ
ρparticle
(1)
3
where the bulk density (g/cm ), ρbulk = mass of agglomerates divided by the volume of
3
agglomerates, the particle density (g/cm ), ρpartcle = true density of solids (ρdimethyl
3
36
fummarate = 1.12 g/cm ).
18
ACS Paragon Plus Environment

Page 19 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
2
.7.3. Carr’s Index
A portion of 710 to 1000 ꢁm round granules was poured into a 5 mL graduated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
cylinder. The flowability of agglomerates were quantified by the Carr’s index, CI,
19
defined as:
ρt
−
ρ p
CI =
×100%
ρt
(2)
3
where the poured density (g/cm ), ρ
p
= mass of agglomerates divided by the
undisturbed volume in a 5 mL graduated cylinder after filling, and the tapped density
3
(g/cm ), ρ
t
= mass of agglomerates divided by the disturbed volume after tapping until
no change in the volume was seen. The number of times for tapping was about 100.
2
.7.4. Friability
.5 g of the 710 to 1000 ꢁm round granules was placed in a 50 mL scintillation
0
vial with a 35ꢀmm O.D. and 58 mm in length which was subjected to a rotating speed
of 50 rpm on a ballꢀmilling machine (MUBMꢀ236ꢀRTD, Shin Kwan Machinery,
37
Taipei) for 30 min.
The weight of 710ꢀ to 1000ꢀꢁm sized round granules before
and after this friability test was measured. Friability of round granules was
3
8
determined by using the formula:
19
ACS Paragon Plus Environment

Organic Process Research & Development
Page 20 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
W ꢀW
0
Friability(%) =
×100%
W0
(3)
where W
weight of the round granules did not pass through the 710ꢀꢀm sized sieve after
rotation.
0
= initial weight of the round granules placed on the 710ꢀꢀm sized sieve, W
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
2
.7.5. Particle Strength
The fracture forces of 10 to 30 granules from the same size cut of 710 to 1000
1
9
ꢁ
m were determined together by compression under a punch with a water reservoir.
Water titration would continue until all round granules were fractured due to the
increasing weight of the water reservoir. The force applied was the total weight of
water and the water reservoir, and the average fracture force exerted on each round
granule was the total force applied divided by the number of granules.
2
.7.6. Tabletability
Tabletability is the ability to make a tablet of adequate mechanical strength by
powder compaction. 200 mg of 710ꢀ to 1000ꢀꢁm sized round dimethyl fumarate
granules were charged into the 13 mm die set of the FTꢀIR pellet press (Specac Inc.,
Woodstock, GA, USA) with about 8 tons of compaction force to produce a tablet.
20
ACS Paragon Plus Environment

Page 21 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
2
.8. Dissolution Test
A dissolution test station (SR6, Hanson Research Corporation, Chatsworth,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
California) Type II (paddle method) at a rotational speed of 100 rpm was used for in
vitro testing of the dissolution behavior of dimethyl fumarate crystals. Dissolution
was carried out on 120 mg of the size range of 710ꢀ1000 ꢀm dimethyl fumarate
agglomerates. RO water of pH 6.8 with potassium phosphate buffer was used as the
dissolution medium. The volume and temperature of the dissolution medium were
o
5
00 mL and 37.0 ± 0.2 C, respectively. 1ꢀmL samples were withdrawn at the time
points of 0, 1, 2.5, 5, 6.5, 8, 10, 12.5, 15, 20, 30, 40, 60, 90 and 120 min by a plastic
syringe near the stirring paddle. Each sample was filtered by a 0.22 ꢁm PVDF
membrane (PALL Corporation, PN), diluted 40 times in a 20 mL scintillation vial
with RO water, and then assayed for the absorbance of dimethyl fumarate by UVꢀvis
spectrophotometer at λ = 210 nm. The concentration of dimethyl fumarate at each
time point was converted from the absorbance value to the corresponding
concentration by a linear calibration: Absorbance (a.u.) = 116.4 × Concentration of
dimethyl fumarate (mg/mL), according to Beer’s law established from five standard
dimethyl fumarate solutions, each with a known concentration of dimethyl fumarate.
21
ACS Paragon Plus Environment

Organic Process Research & Development
Page 22 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2.9. Analytical Instrumentations
.9.1. Thermocouple
The solution temperature in the 0.5 L baffled glass reactor was recorded by a
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Kꢀtype thermocouple and a temperature logger (TMꢀ947SD, Lutron Electronic, Taipei,
Taiwan)
2
.9.2. Nuclear Magnetic Resonance Spectroscopy (NMR)
1
H NMR was used for carbonꢀhydrogen framework, concentration and
1
impurity determination. All of the H NMR spectra were performed at 298 K using
Bruker Avance III HD spectrometer at 600 MHz proton frequency with 5 mm BBFO
probe in Zꢀgradient.
2
.9.3. Gas Chromatography–Mass Spectroscopy (GC-MS)
The GC–MS analyses for samples were performed on a Thermo Scientific
Trace 1300 Gas Chromatograph instrument coupled with a Thermo Scientific ISQ
Single Quadrupole massꢀselective detector and a Thermo Scientific autosampler
TriPlus RSH injector. A Rxiꢀ5MS capillary column with a dimension of 30 m × 0.25
mm I.D. × 0.25 ꢁm film thickness was used for the chromatographic separation. The
o
o
o
initial oven temperature was 40 C, and then increased to 160 C at a rate of 10 C/min.
22
ACS Paragon Plus Environment

Page 23 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The injection size was 1 ꢁL, which was prepared by dissolving about 2 mg of sample
in 1 mL of methanol. Helium was used as the carrier gas with a constant flow rate of
o
o
1
mL/min. The injector and MS source temperatures are set at 180 and 250 C,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
respectively. The electron impact (EI) mode at 70 eV was utilized for sample
ionization. The GCꢀMS spectra for DMS and internal standards were obtained
through the injection of the stock solutions and scanning in the range of m/z 29–300.
The GCꢀMS chromatograms are recorded with selected ion monitoring (SIM) mode.
2
.9.4. High Performance Liquid Chromatography (HPLC)
HPLC measurement of samples was performed using a Shimadzu LCꢀ20AT
HPLC with SILꢀ20A autosampler and SPDꢀ20A UV/vis spectrophotometric detector.
Sample solution was prepared by about 10 mg of sample dissolving in 20 mL of
methanol. 5 ꢁL of sample drop was injected onto a YMCꢀPack reverse phase
ODSꢀAQ 250 mm × 4.6 mm I.D. × 5ꢁm particle size × 12 nm pore size column with a
3 4
mobile phase of 0.01M H PO /ACN (80/20 v/v) pumped at 0.8 mL/min. Sample
was detected by its native UV absorbance at λ = 220 nm.
2
.9.5. Gas Chromatography (GC)
Gas chromatographic analysis was carried out by using a China
23
ACS Paragon Plus Environment

Organic Process Research & Development
Page 24 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Chromatography 2000 fitted with a flame ionization detector. Nitrogen was used as
the carrier gas and was set at 15 mL/min. The separation was performed with a 10%
Carbowax 20M phase, matrix 80/100 Chromosorb WꢀHP column, 3 m × 1/8 in O.D. ×
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.1 mm I.D. (SUPELCO metal packed GC column). The oven temperature and the
o
o
injector temperature were set at 190 C and 200 C, respectively. The detector
o
temperature was held at 220 C. About 15 mg of sample was dissolved in 1 mL of
ethanol. The injection volume for a sample was 1 ꢁL.
2
.8.6. Fourier Transform Infrared (FT-IR) Spectroscopy
Infrared spectroscopy provided information about the kinds of functional
groups in the molecule. FTꢀIR spectroscopy was conducted on Perkin Elmer
Spectrum One (Norwalk, CT, USA). 1 mg of each sample was ground gently with
9
9 mg of KBr powders, then a manual press was used to form the pellet. The pellet
ꢀ1
ꢀ1
was scanned 8 times with a resolution of 2 cm in the region of 4000 to 400 cm .
2
.8.7. Differential Scanning Calorimetry (DSC)
DSC analysis was mainly used to identify the enthalpy of fusion, solidꢀliquid
(melting) temperature, and solidꢀsolid transition temperature. Thermal analytical
data of samples in perforated aluminum sample pans were collected on a Perkin Elmer
24
ACS Paragon Plus Environment

Page 25 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
DSCꢀ7 calorimeter (Perkin Elmer Instruments LLC, Shelton, CT, USA) with a
o
o
o
temperature scanning rate of 10 C/min from 30 to 150 C under a constant nitrogen of
9
9.990% purge.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
.8.8. Powder X-Ray Diffraction (PXRD)
The PXRD diffractogram detected by Bruker D8 Advanced (Germany) could
provide information about crystal structure identification and crystallinity. The
source of PXRD was Cu Kα (λ = 1.542 Å) and the diffractometer was operated at 40
o
o
o
kV and 40 mA. The scanning rate was set at 0.05 2θ/sec ranging from 5 to 35 ,
and the experiment was operated under room temperature.
2
.8.9. Single Crystal X-ray Diffraction (SXD)
SXD provided detailed information about the internal lattice of crystalline
substances, including unit cell dimensions, bondꢀlengths, bondꢀangles, and details of
siteꢀordering. SXD data of dimethyl fumarate samples were obtained from the
30
literature.
The view of crystal structure from SXD was drawn by Diamond 3.1
computer software (Crystal Impact GbR, Brandenberg Germany).
2
.8.10. Optical Microscopy (OM)
25
ACS Paragon Plus Environment

Organic Process Research & Development
Page 26 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To observe the actual size of dimethyl fumarate primary crystals, the dried
dimethyl fumarate powders harvested from antisolvent and cooling crystallization step
were dispersed in mineral oil, birefringence and crystal habits were observed by
Olympus BXꢀ51 (Tokyo, Japan) equipped with a digital camera (Hong Kong, China)
Moticam 2000. These images were transformed by Motic Images Plus (version 2.0)
into digital pictures and analyzed by Measure Tool (version 4.10, copyright by
Jieꢀchuang international Co., Ltd.) The other optical microscope (Olympus SZII,
Tokyo, Japan) with a charge couple device (CCD) camera (SONY, model:
SSCDC50A, Tokyo, Japan) was used to characterize the morphology of dimethyl
fumarate granules.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.8.11. Ultraviolet Visible (UV/Vis) Spectrophotometer
UV/Vis spectrophotometer was used to identify organic compounds and
measure concentrations of solutions. The clear solution was assayed for dimethyl
fumarate based on the characteristic UV absorbance peak at 210 nm by an UV/vis
spectrophotometer (Lambda 25, PerkinꢀElmer, Norwalk, CT, USA.).
3
.
RESULTS AND DISCUSSION
3
.1. Reaction
26
ACS Paragon Plus Environment

Page 27 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Dimethyl fumarate was prepared by sulfuric acidꢀcatalyzed esterification of
fumaric acid with methanol (Scheme 1). This was a classical homogeneous
31
liquidꢀphase reaction.
Guzowski et al. had reported of using the different molar
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ratios of fumaric acid to sulfuric acid of 4:1, 20:1 and 100:1 and achieved the same
29
conversion to dimethyl fumarate when reacted for 3, 8 and 18 h, respectively.
In
this study, the molar ratio of fumaric acid to sulfuric acid of 4:1 was chosen for
achieving a reasonable reaction time length. The initial concentration of sulfuric
acid catalyst was fixed at 0.73 mol% (2.18 wt%) in the methanolic reaction medium.
However, the larger amount of sulfuric acid used in esterification could raise the level
29
of side products such as monomethyl sulfate MMS and DMS.
Due to the
genotoxic nature of DMS, its maximum daily intake level should not exceed 1.5 ꢀg
3
9
per person per day, and the risk of creating MMS and DMS must be carefully
evaluated.
Scheme 1. Esterification of fumaric acid and methanol.
1
The H NMR spectra in Figures S1(a) to S1(p) revealed the kinetics of
esterification of fumaric acid with methanol (Scheme 1) in different time samples.
The chemical shifts for alkene protons of fumaric acid and dimethyl fumarate slightly
27
ACS Paragon Plus Environment

Organic Process Research & Development
Page 28 of 63
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
moved from 6.61 ppm (Peak (I)) to 6.77 ppm (Peak (1)) during the progress of
3
esterification. The integral peak area of the resonance for the –COOꢀCH ester
protons at 3.71 ppm (Peak (2)) did grow over time. The total reaction time of 3 h
was sufficient to reach a ~ 85% conversion (Figure 1). Because of the low
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
concentration of sulfuric acid used, esterification of sulfuric acid was suppressed
39
(Scheme 2).
0
.9
.8
100% Conversion (0.72 M)
0
0.7
0.6
(
(
a)
b)
0.5
0.4
0.3
0.2
0
.1
.0
End point:
(a) 90.1% Conversion (0.65 M)
(
b) 84.5% Conversion (0.61 M)
0
0
20
40
60
80 100 120 140 160 180
Reaction Time (min)
Figure 1. Kinetic studies of esterification of fumaric acid with an initial concentration
o
of 0.72 M based on 100 mL of methanol at 45 C by (a) gravimetric analysis, and (b)
1
H NMR analysis (based on Figures S1(a) to S2(p)).
28
ACS Paragon Plus Environment

Page 29 of 63
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Esterification of sulfuric acid and methanol.
Noticeably, the chemical shift of hydroxyl –OH protons for fumaric acid
showed a broad singlet with a large downfield shift at 13.06 ppm (Peak (II)) because
of the high initial concentration of –OH protons and increased hydrogen bonding in
6
DMSOꢀd (Figure S1). As the esterification proceeded, fumaric acid was being
consumed and its concentration was lowered. The –OH proton signal then moved to
a higher chemical shift to around 4.78 ppm (Peak (7)) due to the diminishing solvent
effect. In general, the chemical shift for –OH proton is strongly dependent on
40
concentration, temperature, and solvent.
The peak of reaction intermediate,
3
monomethyl fumarate (MMF) (Scheme 3), at chemical shift of methyl –CH proton
around 3.72 ppm (Peak (8)) was detected in the beginning of esterification.
However, it was gradually decreased over time, and finally disappeared towards the
end of reaction (Figure S1 Insets). For the other chemical shifts at 2.50 ppm (Peak
6
(3)), 3.16 ppm (5)) and 3.38 ppm (Peak (4)), they were from the protons of DMSOꢀd ,
6
methanol, and water in “neat” DMSOꢀd , respectively.
29
ACS Paragon Plus Environment