Formation of N,O-acetalsin the production of x-ray contrast agents

Article abstract of DOI:10.1021/op500177w

Acetylation of 1 with acetic anhydride, an important step in the preparation of iohexol, gave the corresponding acetanilide and minor amounts of a few byproducts, whose structures have been elucidated. Among the byproducts were some 1,3-oxazolidines which survived when the reaction was quenched by adding methanol and water under strong acidic conditions.

Full text of DOI:10.1021/op500177w

Article
pubs.acs.org/OPRD
Formation of N,O‑Acetals in the Production of X‑ray Contrast Agents
Torfinn Haland†,‡ and Leiv K. Sydnes*,†
̊
†Department of Chemistry, University of Bergen, Alleg
́
aten 41, NO-5007 Bergen, Norway
‡GE Healthcare, Lindesnes Site, NO-4521 Lindesnes, Norway
S
* Supporting Information
ABSTRACT: Acetylation of 1 with acetic anhydride, an important step in the preparation of iohexol, gave the corresponding
acetanilide and minor amounts of a few byproducts, whose structures have been elucidated. Among the byproducts were some
1,3-oxazolidines which survived when the reaction was quenched by adding methanol and water under strong acidic conditions.
containing these compounds was analyzed by LC-MS.
Somewhat to our surprise the compounds denoted a−f in
Figure 1 gave rise to two different mass spectra only; the
spectra of compounds a−d were in essence identical and so
were those of compounds e and f. This observation clearly
indicates that the six compounds constitute two groups of
isomers, and we decided to focus our attention on the isolation
and structure elucidation of these impurities.
1. INTRODUCTION
Process development has always been in focus in the chemical
industry, but with increasing production costs in recent years
due to the price of chemicals, the improvement of chemical
processes has become more important than ever. Many aspects
are important to consider in such an endeavor, but two issues
are particularly important to address, chemical yield and
byproduct formation. Overall the best way to improve a
chemical process is by reducing the latter in every step. Lower
levels of impurities may not only pave the way for more
efficient purification; in the best cases it may even be possible
to skip one such step and significantly reduce production costs
without compromising product quality.
To reduce the level of impurities formed in a synthesis, it is
important to know the structures of the byproducts and
understand how they are formed and react. This approach has
successfully been applied to improve many industrial syntheses.
One such example is the acetylation of the aniline derivative 5-
amino-N,N′-bis(2,3-dihydroxypropyl)-2,4,6-triiodoisophthala-
mide (1) to make the corresponding acetamide, 5-acetamido-
N,N′-bis(2,3-dihydroxypropyl)-2,4,6-triiodoisophthalamide (2)
(Scheme 1), which is a key intermediate in the production of
iohexol, the active ingredient in the X-ray contrast agent
Omnipaque. Over the years thorough studies of byproduct
formation and reactivity have made it possible to manufacture
pure 2 both simpler and more efficiently.
Although process development increased the yield and purity
of 2, there were still unknown impurities present that would be
beneficial to study with the aim, ultimately, to prevent their
formation completely. The presence of such impurities is
illustrated in Figure 1, which shows a typical HPLC
chromatogram of an unpurified sample of 2. The structures
of all compounds with a retention time shorter than 12 min
have been known for some time, whereas several of those giving
rise to the peaks with retention times between 15 and 22 min
were unknown. In order to improve the production of 2
further, it was therefore decided to study the latter group of
impurities more thoroughly, and our findings are reported here.
2.1. Byproducts e and f. In order to obtain impurities e
and f in reasonable amounts for structure elucidation, a reaction
mixture from the manufacturing process was used to isolate the
two impurities. Unfortunately, neither compound could be
obtained pure in reasonable quantities even after multiple
preparative HPLC separations, but by investigating three
mixtures of the two compounds (approximate ratios of 1:9,
1:1, and 9:1) by two-dimensional NMR techniques, it could be
concluded that e and f are stereoisomers of 3-acetamido-N-
(2,3-dihydroxypropyl)-5-(5-hydroxymethyl-2,2-dimethyloxazo-
lidine-3-carbonyl)-2,4,6-triiodobenz-amide (3) (Figure 2).
The formation of 3 can be explained by a two-step
transformation from 1, i.e., acetylation of the amino group
and N,O-ketalization of one of the benzamide moieties. The
latter reaction requires the presence of acetone to occur, which
is a significant observation because acetone was not added to
the reaction mixture. Therefore, acetone must have been
generated in situ or been present as an impurity in one or
several of the reagents (vide infra). It is also noteworthy that an
oxazolidine and not a 1,3-dioxolane is obtained, because the
latter group of compounds should be formed much more easily
than the former, by reaction with the vicinal diol in the side
chain.1
2.2. Byproducts a−d. To elucidate the structure(s) of
byproducts a−d, another reaction mixture from the manu-
facturing process was used for the isolation, which was
performed in three steps by preparative HPLC. It was necessary
to use an acidified eluent in the isolation to achieve stable
retention times for compounds a−d, and this indicates that the
byproducts contain a carboxylic acid moiety. Compounds a and
b were obtained with purities of 93 % and 97 %, respectively.
Unfortunately, compounds b and c could not be obtained pure
2. RESULTS AND DISCUSSION
In order to get an indication of the structural differences
Received: June 2, 2014
between the impurities under investigation, here a mixture
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Acetylation step in the manufacturing of iohexol
Figure 3. 1,3-Oxazolidine 4, which appeared to be formed as a
stereoisomeric mixture when 1 was acetylated under standard
conditions.
because the latter group of compounds is in general formed
much more easily than the former.1
2.3. Impurities in Reagents. The formation of impurities
3 and 4 in the synthesis of 2 in the production of iohexol is
intriguing since acetone and acetoacetic acid are not used in the
process at all. One explanation for the formation of these
impurities may be that one or several of the reagents applied
contain a low level of acetone and acetoacetic acid. All of the
reagents were therefore analysed thoroughly with respect to
acetone, but not even traces were detected. Another possibility
is that the two ketones are formed during the reaction. If this is
the case, it is reasonable to believe that chemicals recovered in
the manufacturing process should contain at least acetone.
Samples collected from fractions of chemicals recovered during
the manufacturing of 2 were therefore analysed by GC-MS, and
significant amounts of acetone were indeed detected in
recovered acetic anhydride. Thus, the source of this ketone is
at least one acetic-anhydride impurity which is able to form
acetone under the reaction conditions prevailing during the
acetylation of 1.
The potential acetone precursors are of course limited by the
processes applied to manufacture acetic anhydride on an
industrial scale. Today this anhydride is produced predom-
inantly by two processes. The most recent method involves
carbonylation of methyl acetate in the presence of a rhodium-
iodide catalyst, and in this process ethylidene diacetate, acetone,
carbon dioxide and methane are known byproducts.4 The other
process involves addition of acetic acid to ketene, and acetic
anhydride manufactured in this way is known to contain
ketene-related byproducts including diketene.5,6 This is
significant because diketene reacts with nucleophiles, and
when exposed to water under acidic conditions acetoacetic acid
is indeed formed.6 This acid can undergo a number of
reactions, including oxazolidine formation to give 4 and
decarboxylation to give acetone which is required to form
impurity 3.
Figure 1. A typical impurity profile prior to purification of 2. In this
study the focus has been on impurities a−f.
Figure 2. 1,3-Oxazolidine 3, which appeared to be formed as a
stereoisomeric mixture when 1 was acetylated under standard
conditions.
due to very similar retention under a variety of conditions, but
two mixtures with a b:c ratio of approximately 3:2 and 1:9 were
obtained.
Notably, most of the IR and NMR spectra of a, d, and the
two mixtures of b and c were so similar to parts of the
corresponding spectra of the two isomers of oxazolidine 3 that
the structures of the four impurities were believed to be almost
identical to that of 3. The only important differences were the
presence of a broad IR absorption in the 3300−2800 cm−1
region, indicative of a COOH group,2 and an additional
carbonyl signal in the 170−171 ppm region in the 13C NMR
spectrum, which also indicates the presence of a COOH
moiety.3 This, combined with the absence of one methyl signal,
means it could be concluded that a−d are stereoisomers of 2-
(3-(3-acetamido-5-(2,3-dihydroxypropylcarbamoyl)-2,4,6-triio-
dobenzoyl)-5-hydroxymethyl-2-methyloxazolidin-2-yl)acetic
acid (4) (Figure 3).
Just like for compound 3 the formation of 4 can be explained
by a two-step transformation from 1, i.e., acetylation of the
amino group and N,O-ketalization at one of the benzamide
moieties to give a five-membered ring. The latter step requires
the presence of acetoacetic acid or a corresponding ester of this
acid, and this is a significant observation since this acid was not
added to the reaction mixture (vide infra). Again it is
noteworthy that 4 is an oxazolidine and not a 1,3-dioxolane,
2.4. N,O-Acetal versus O,O-Acetal Formation. For-
mation of 1,3-dioxolanes from vicinal diols and aldehydes and
ketones is a well-known transformation in organic chemistry,7
but conversion of 2-aminoethanols to 1,3-oxazolidines by
reaction with aldehydes and ketones has also been
reported.8−11 In order to get a good yield of both heterocycles,
an acid has to be used as catalyst, and the water generated in
both reactions has to be removed, for instance by using
B
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
molecular sieves, performing azeotropic distillation with
benzene or toluene, or by adding a chemical that traps water
by a chemical reaction.
then acidified with sulfuric acid until the pH was approximately
0, and subsequent analysis showed that only the main product 2
and oxazolidine 3 were present; not even traces of 5 and 6
could be detected (Figure 4b).
Although syntheses of oxazolidines are reported in the
literature, it was surprising to find that oxazolidines are formed
in considerable amounts directly from amides under acidic
conditions since amides are significantly less nucleophilic than
amines. Therefore, it was suspected that acetic anhydride plays
the role as a water-removing reagent and changes the
equilibrium position in the oxazolidine direction. In order to
check if this might be the case a model compound was mixed
with acetone and PTSA and heated to reflux (ca. 60 °C) (vide
infra), and indeed, in the absence of acetic anhydride
oxazolidines were not formed.
On this basis, it is interesting to observe that 1,3-dioxolane
impurities were not detected at all in the reaction mixture from
the industrial production of 2. One reason could of course be
that such compounds are not formed at all, but a more likely
explanation is that 1,3-dioxolanes are indeed generated but
decompose during the reaction or during the workup. In order
to explore the latter possibility, the starting material (1) was
added to a mixture of acetone, acetic anhydride, and acetic acid
in the presence of PTSA and stirred under reflux. A sample was
collected after 15 min and after modified workup (hydrolysis by
aqueous base treatment followed by adjustment of pH to 4−5),
it was analysed by HPLC. The analysis revealed as expected the
presence of acetanilide 2 and oxazolidine 3 (as a stereoisomeric
mixture), but in addition that two other major products,
denoted 5 and 6, had been formed (Figure 4a). The sample was
To determine the structures of 5 and 6, the reaction above
was repeated with a short reaction time (15 min) and
subsequently worked up by applying the modified procedure
described above (hydrolysis by addition of NaOH and
neutralization to pH 4−5). Due to the significant difference
in the retention of the two compounds, the separation by
preparative HPLC was straightforward, and samples of 5 and 6
were obtained in adequate quantity and purity for spectroscopic
and spectrometric investigations (see Figure 5). The data
obtained proved that the two products were 3-(3-acetamido-5-
(2,2-dimethyl-1,3-dioxolan-4-yl)methylcarbamoyl-2,4,6-
triiodobenzamido)propane-1,2-diol (5) and 5-acetamido-N,N′-
bis(2,2-dimethyl-1,3-dioxolan-4-yl)methyl-2,4,6-triiodoiso-
phthalamide (6).
With the structures of 5 and 6 elucidation of the impurity
profile of 2, prepared by acetylation of 1 in the manufacturing
process, can be understood. All compounds 3−6 are indeed
formed initially, but 1,3-dioxolanes 5 and 6, which are formed
the fastest and in the highest amounts, are much more easily
cleaved than oxazolidines 3 and 4 and revert to the
corresponding tetraol (2) under strongly acidic aqueous
conditions.
2.5. Studies with Model Compounds. The fact that
oxazolidines 3 and 4 are formed as mixtures of stereoisomers is
not really surprising since the starting material contains two
chiral centres and the three iodine atoms are so sterically
demanding that rotation around the C−C and N−C bonds to
the benzene ring is prevented.12 The steric hindrance caused by
the iodine atoms is illustrated by the ChemDraw-generated
drawing in Figure 6, but to substantiate the effect
experimentally we decided to study the model compound
N,N′-bis(2,3-dihydroxypropyl)-5-nitro-isophthalamide (7), a
common, iodine-free starting material used to produce X-ray
contrast agents.
When 7 was exposed to a mixture of acetic anhydride and
acetone under strongly acidic conditions, HPLC analysis of the
crude product showed that several new compounds had been
formed (Figure 7). Four compounds were by means of
preparative HPLC obtained in adequate quantity and purity for
subsequent spectroscopic and spectrometric investigations, and
the data obtained proved that the four products were 5-nitro-N-
(2,3-dihydroxypropyl)-N′-(2,2-dimethyl-1,3-dioxolan-4-yl)-
methylisophthalamide (8), 5-nitro-N,N′-bis(2,2-dimethyl-1,3-
dioxolan-4-yl)-methylisophthalamide (9), N-(2,3-dihydroxy-
propyl)-3-(5-hydroxymethyl-2,2-dimethyloxazoli-dine-3-car-
bonyl)-5-nitrobenzamide (10), and N-(2,2-dimethyl-1,3-dioxo-
lan-4-yl)methyl-3-(5-hydroxymethyl-2,2-dimethyloxazolidine-3-
carbonyl)-5-nitrobenzamide (11) (Scheme 2). The structures
of the compounds giving rise to the peaks with retention times
21.5 and 25.1 min are still unknown, but their mass spectra
prove that they are not isomers of 10 and 8, respectively. Thus,
unlike oxazolidines 3 and 4, each of the acetals 8−11 gives rise
to one chromatographic peak only, and this clearly indicates
that the stereoisomerism exhibited by 3 is due to hindered
rotation caused by the iodine atoms, whereas that of 4 is a
combination of hindered rotation and cis/trans isomerism in
the oxazolidine moiety.
Figure 4. Chromatograms from HPLC analyses of a sample collected
from the reaction between aniline 1, acetone, acetic anhydride, acetic
acid, and PTSA. (a) Sample collected after reflux for 15 min; worked
up by hydrolysis with addition of NaOH and neutralization to pH 4−
5; (b) sample in a after subsequent treatment with concentrated
sulfuric acid until pH approximately 0.
Another relevant point to consider is the potential formation
of tetrahydro-1,3-oxazines during the acetylation of 1. Such
C
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 5. 1,3-Dioxolane derivatives 5 and 6 formed during the acetylation of aniline 1. The compounds were isolated when the reaction time was
short, and the workup was carried out under slightly acidic conditions.
Figure 6. A ChemDrawn-generated drawing of oxazolidine 3. The following colour code has been applied: Hydrogen atoms are white, carbon grey,
nitrogen blue, oxygen red, and iodine violet.
product isolation and full product characterization, and the
results were conclusive: N,O-acetal formation did only take
place with the bis(2-hydroxyethyl)amides (12a and 12b), not
with bis(3-hydroxypropyl)-amides (12c and 12d). On the basis
of these observations the absence of 1,3-oxazine impurities in 2
is in accordance with expectations.
3. EXPERIMENTAL SECTION
General. All commodity chemicals were purchased from
commercial suppliers and used without further purification.
Also, unless otherwise stated, all reagents were commercially
available and used as received. All melting points reported are
uncorrected. IR absorptions are given in wave numbers (cm−1),
and intensities are characterized as (s) for strong, (m) for
1
medium, (w) for weak, and (br) for broad. H NMR spectra
Figure 7. A HPLC chromatogram of the crude product after
were recorded at ambient temperature at 400 or 600 MHz with
DMSO-d6 (δH = 2.50 ppm) as solvent and internal reference.
Chemical shifts are reported in ppm. Multiplicity is given as (s)
for singlet, (d) for doublet, (t) for triplet, (q) for quartet, (qui)
for quintet, (dd) for doublet of doublet, (m) for multiplet, and
(bs) for broad singlet. 13C NMR spectra were recorded at
ambient temperature at 100 or 150 MHz with the central peak
of the DMSO-d6 septet (δC = 39.50 ppm) as internal reference.
Mass spectra were obtained on a Waters qTOF-micro or a
JEOL AccuTOF T100LC, which both were operated in the ESI
mode.
Analysis and Isolation by HPLC. Analytical HPLC was
performed on an Agilent 1100 or a TSP (Thermo Separation
products) instrument with UV detection (254 nm). Four
different RP-18 columns were used, two from Supelco (75 ×
4.6 mm, 3 μ; 250 × 4.6 mm, 5 μ), one from LUNA (250 × 4.6
acetylation of 7 and workup using a modified procedure that makes
it possible for 1,3-dioxolanes to survive.
compounds are known in the literature,13 but no such product
was obtained from 1. Several reasons for this outcome are
conceivable, but the most likely one is that tetrahydro-1,3-
oxazines are formed more reluctantly than 1,3-oxazolidines and
1,3-dioxolanes under the reaction conditions prevailing under
the acetylation. In order to see if this is the case N,N′-
disubstituted isophthalamides 12a−12d were acetylated by
acetic anhydride in acetone in the presence of PTSA. The
workup was carried out under mild conditions to make sure
that any N,O-acetals formed did not decompose (Scheme 3).
Thorough LC-MS analyses of the crude product mixtures were
performed before they were subjected to preparative HPLC for
D
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 2. N,O- and O,O-Acetals from treatment of model compound 7 with acetone in acetic anhydride containing p-
toluenesulfonic acid (PTSA)
Scheme 3. N,O-Acetals from treatment of model compounds 12a−12d with acetone in acetic anhydride containing p-
toluenesulfonic acid (PTSA)
mm, 10 μ), and one from Brownlee (250 × 4.6 mm, 5 μ). Five
different gradient programs with mixtures of water (A) and
acetonitrile (B) as eluent were applied. All gradients are linear.
Program 1: 0.4−2.2% B (0−10 min), 2.2−14.3% B (10−20
min), 14.3% B (20−24 min).
performed by means of preparative HPLC. The injection
solution was prepared by diluting 100 mL of the reaction
mixture with H2O until 1 L and then adjusting the pH to 7−9
by adding HCOOH. Fifteen runs were performed on a LUNA
260 × 101 mm column, and the two compounds were collected
in the same fraction and then concentrated under vacuum on a
rotary evaporator. The resulting concentrate was purified
further on a LUNA 250 × 50 mm column in two steps. In
the first the two compounds were again collected in the same
fraction, which was concentrated on a rotary evaporator and
freeze-dried to give a white solid (530 mg). The solid was
dissolved in water for the second purification step. Six runs
were performed, and three fractions were collected, concen-
trated on a rotary evaporator, and freeze-dried. The weights of
the obtained white solids were 83 mg (fraction 1), 130 mg
(fraction 2), and 65 mg (fraction 3). The HPLC analyses
(Program 1, Supelco 75 × 4.6 mm) of the solids gave the
following results (stereoisomers of 3 are denoted α and β, in
Figure 1 they are given as e and f): Solid 1: Purity 97%, 3α: 3β
(90:10). Solid 2: Purity 96%, 3α: 3β (54:46). Solid 3: Purity
89%, 3α: 3β (12:88). Data for 3α: Mp 265 °C (dec); IR (ATR)
3251 (br), 2983 (w), 2937 (w), 2885 (w), 1635 (s), 1540 (m),
Program 2: 2.0−5.0% B (0−10 min), 5.0−25.0% B (10−20
min), 25.0% B (20−25 min).
Program 3: 2.5−5.0% B (0−10 min), 5.0−30.0% B (10−30
min), 30.0% B (30−50 min).
Program 4: 5.0−10.0% B (0−10 min), 10.0−50.0% B (10−
30 min), 50.0% B (30−35 min).
Program 5: 2.5−5.0% B (0−10 min), 5.0−30.0% B (10−30
min), 30.0% B (30−60 min).
The flow rate was 1.0 mL/min (Program 1) or 1.5 mL/min
(Programs 2−5). No acid was added into the eluents unless
otherwise stated. The amount of each component is given in %
on the basis of its relative area.
Preparative HPLC was carried out by using a Shimadzu LC-
8A or a Varian 800 mL SDS instrument, both with UV
detection (254 nm). A LUNA, RP-18, 250 × 50 mm, 10 μ
column or a LUNA, RP-18, 260 × 101 mm, 10 μ column was
used. The flow rate was 75 mL/min (Shimadzu) or 250 mL/
min (Varian). No acid was added into the eluents (water and
acetonitrile) unless otherwise stated
Isolation of 3-Acetamido-N-(2,3-dihydroxypropyl)-5-(5-
hydroxymethyl-2,2-dimethyl-oxazolidine-3-carbonyl)-2,4,6-
triiodobenzamide (3). Aniline 1 was acetylated on a
production scale by acetic anhydride in the presence of a
strong acid catalyst. When the reaction was complete, the
mixture was concentrated, and remaining acetic anhydride was
quenched by addition of methanol and water under strong
acidic conditions. Then deacetylation was performed by
addition of sodium hydroxide, and a small portion of the
reaction mixture obtained was used for the isolation, which was
1
1362 (m), 1248 (m), 1044 (m) cm−1; H NMR (600 MHz,
DMSO-d6) δ 9.86 (m, 1H, Ar−NH), 8.60−8.35 (m, 1H, HN−
CH2), 4.80 (m, 1H, N−CH2−CH−CH2−OH), 4.55 (m, 1H,
HN−CH2−CH−OH), 4.37 (bs, 1H, HN−CH2−CH−CH2−
OH), 4.17 (m, 1H, N−CH2−CH−CH2−OH), 3.71 (m, 1H,
HN−CH2−CH), 3.53 and 3.48 (2m in a 1:1 ratio, 2H, N−
CH2−CH−CH2−OH), 3.48 and 3.42 (2m in a 1:1 ratio, 2H,
HN−CH2−CH−CH2−OH), 3.31 and 3.17 (2m in a 1:1 ratio,
2H, HN−CH2), 3.19 and 3.01 (2m in a 1:1 ratio, 2H, N−
CH2), 2.03 (s, 3H, HN−C(O)−CH3), 1.69 (s, 3H, C−CH3),
1.63 (s, 3H, C−CH3); 13C NMR (150 MHz, DMSO-d6) δ
169.2 (Ar-C(O)−NH), 167.5 (Ar−NH−C(O)), 165.1 (Ar−
E
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
C(O)−N), 150.5 (ArC−C(O)−NH), 148.4 (ArC−C(O)−N),
143.7 (ArC−NH), 99.1 (ArC−I), 97.8 (ArC−I), 94.5 (CH3−
C−CH3), 89.2 (ArC−I), 75.1 (N−CH2−CH), 69.8 (HN−
CH2−CH), 63.8 (HN−CH2−CH−CH2−OH), 61.3 (N−
CH2−CH−CH2−OH), 49.0 (N−CH2), 42.5 (HN−CH2),
25.0 (C−CH3), 23.3 (C−CH3), 22.8 (C(O)−CH3); HRMS
Calcd for C19H24I3N3O7Na+ [M + Na+] 809.8646, found
809.8645. Data for 3β: Mp 265 °C (dec); IR (ATR) 3231 (br),
2988 (w), 2936 (w), 2880 (w), 1631 (s), 1523 (m), 1361 (m),
4.6 mm, 0.1% HCOOH) of the solids gave the following
results: Solid 1:93%, 4β:4γ (59:41). Solid 2:99%, 4β:4γ
(12:88). Data for 4α: Mp 230 °C (dec); IR (ATR) 3231
(br), 2930 (w), 1712 (m), 1626 (s), 1525 (m), 1356 (m), 1237
(m), 1108 (m), 1032 (m) cm−1; 1H NMR (400 MHz, DMSO-
d6) δ 10.00 (m, 1H, Ar−NH), 8.75−8.50 (m, 1H, HN−CH2),
5.30 (bs, 3H, OH), 4.23 (m, 1H, N−CH2−CH), 3.68 (m, 1H,
HN−CH2−CH), 3.54 and 3.50 (2m in a 1:1 ratio, 2H, N−
CH2−CH−CH2−OH), 3.47 and 3.39 (2m in a 1:1 ratio, 2H,
HN−CH2−CH−CH2−OH), 3.30 and 3.16 (2m in a 1:1 ratio,
2H, N−CH2), 3.30 and 3.15 (2m in a 1:1 ratio, 2H, HN−
CH2), 3.24 and 2.85 (2m in a 1:1 ratio, 2H, CH2−COOH),
2.03 (s, 3H, HN−C(O)−CH3), 1.83 (s, 3H, C−CH3),
carboxylic acid proton not observed; 13C NMR (100 MHz,
DMSO-d6) δ 170.3 (COOH), 169.5 (Ar−C(O)−NH), 167.9
(Ar−NH−C(O)), 165.6 (Ar−C(O)−N), 150.7 (ArC−C(O)−
NH), 148.1 (ArC−C(O)−N), 143.8 (ArC−NH), 100.1 (ArC−
I), 98.5 (ArC−I), 94.3 (CH3−C−CH3), 89.5 (ArC−I), 75.9
(N−CH2−CH), 70.0 (HN−CH2−CH), 64.0 (HN−CH2−
CH−CH2−OH), 61.3 (N−CH2−CH−CH2−OH), 49.0 (N−
CH2), 42.6 (HN−CH2), 41.0 (CH2−COOH), 23.3 (C−CH3),
23.0 (C(O)−CH3); HRMS Calcd for C20H24I3N3O9Na+ [M +
Na+] 853.8544, found 853.8547. Data for the mixtures of 4β
and 4γ: IR (ATR) 3247 (br), 2935 (w), 1714 (m), 1631 (s),
1522 (m), 1357 (m), 1242 (m), 1112 (m), 1042 (m) cm−1;
HRMS Calcd for C20H24N3I3O9Na+ [M + Na+] 853.8544,
found 853.8547. Data for 4β: 1H NMR (400 MHz, DMSO-d6)
δ 10.00 (m, 1H, Ar−NH), 8.75−8.50 (m, 1H, HN−CH2), 4.21
(bs, 3H, OH), 4.21 (m, 1H, N−CH2−CH), 3.68 (m, 1H, HN−
CH2−CH), 3.54 and 3.50 (2m in a 1:1 ratio, 2H, N−CH2−
CH−CH2−OH), 3.47 and 3.39 (2m, 2H, HN−CH2−CH−
CH2−OH), 3.37 and 2.75 (2m in a 1:1 ratio, 2H, CH2−
COOH), 3.30 and 3.16 (2m in a 1:1 ratio, 2H, N−CH2), 3.30
and 3.15 (2m in a 1:1 ratio, 2H, HN−CH2), 2.03 (s, 3H, HN−
C(O)−CH3), 1.79 (s, 3H, C−CH3), carboxylic acid proton not
observed; 13C NMR (100 MHz, DMSO-d6) δ 170.4 (COOH),
169.5 (Ar−C(O)−NH), 167.9 (Ar−NH−C(O)), 165.6 (Ar−
C(O)−N), 150.7 (ArC−C(O)−NH), 148.1 (ArC−C(O)−N),
143.8 (ArC−NH), 100.0 (ArC−I), 98.3 (ArC−I), 94.3 (CH3−
C−CH3), 89.5 (ArC−I), 76.0 (N−CH2−CH), 70.0 (HN−
CH2−CH), 64.0 (HN−CH2−CH−CH2−OH), 61.1 (N−
CH2−CH−CH2−OH), 48.9 (N−CH2), 42.6 (HN−CH2),
42.6 (CH2−COOH), 23.1 (C(O)−CH3), 21.4 (C−CH3).
1
1249 (m), 1108 (m), 1043 (m) cm−1; H NMR (600 MHz,
DMSO-d6) δ 9.88 (m, 1H, Ar−NH), 8.60−8.35 (m, 1H, HN−
CH2), 4.80 (bs, 1H, N−CH2−CH−CH2−OH), 4.55 (bs, 1H,
HN−CH2−CH−OH), 4.36 (bs, 1H, HN−CH2−CH−CH2−
OH), 4.17 (m, 1H, N−CH2−CH), 3.72 (m, 1H, HN−CH2−
CH), 3.55 and 3.49 (2m in a 1:1 ratio, 2H, N−CH2−CH−
CH2−OH), 3.49 and 3.42 (2m in a 1:1 ratio, 2H, HN−CH2−
CH−CH2−OH), 3.32 and 3.17 (2m in a 1:1 ratio, 2H, HN−
CH2), 3.16 and 3.09 (2m in a 1:1 ratio, 2H, N−CH2), 2.04 (s,
3H, HN−C(O)-CH3), 1.70 (s, 3H, C−CH3), 1.63 (s, 3H, C−
CH3); 13C NMR (150 MHz, DMSO-d6) δ 169.2 (Ar−C(O)−
NH), 167.5 (Ar−NH−C(O)), 165.1 (Ar−C(O)−N), 150.5
(ArC−C(O)−NH), 148.4 (ArC−C(O)−N), 143.6 (ArC−
NH), 99.2 (ArC−I), 97.9 (ArC−I), 94.5 (CH3−C−CH3),
88.9 (ArC−I), 75.1 (N−CH2−CH), 69.8 (HN−CH2−CH),
63.8 (HN−CH2−CH-CH2−OH), 61.4 (N−CH2−CH−CH2−
OH), 49.1 (N−CH2), 42.5 (HN−CH2), 25.0 (C−CH3), 23.2
(C−CH3), 22.8 (C(O)−CH3); HRMS Calcd for
C19H24I3N3O7Na+ [M + Na+] 809.8646, found 809.8646.
Isolation of 2-(3-(3-Acetamido-5-(2,3-dihydroxypropylcar-
bamoyl)-2,4,6-triiodobenzo-yl)-5-hydroxymethyl-2-methyl-
oxazolidin-2-yl)acetic Acid (4). The reaction mixture used for
the isolation by preparative HPLC was obtained as described
for 3. Thirteen runs were performed on a LUNA 260 × 101
mm column, and HCOOH (0.2%) was added to the eluents.
The injection solution was prepared by diluting 50 mL of the
solution from the experiment described above with water until
1 L. Then the pH was adjusted to approximately 4 by adding
HCOOH. One fraction was collected and then concentrated on
a rotary evaporator. The concentrate was purified further on a
LUNA 250 × 50 mm column in two steps. For the first step, six
runs were performed. Three fractions were collected, and they
were then concentrated on a rotary evaporator and freeze-dried.
The weight of the white solids obtained from the fractions was
0.6 g, 1.9 g, and 0.7 g, respectively. The solid from fraction 3
(0.7 g) was dissolved in water (60 mL) for the second
purification step on the LUNA 250 × 50 mm column. Three
runs were performed, and one fraction was collected followed
by concentration on a rotary evaporator and freeze-drying.
(Stereoisomers of 4 are denoted 4α−4δ; in Figure 1 they are
given as a−d, respectively). The purity (HPLC, LUNA 250 ×
4.6 mm, Program 2, 0.1% HCOOH) was 97% (4δ). The solid
from fraction 1 (0.6 g) was dissolved in water (40 mL) for the
second purification step on the LUNA 250 × 50 mm column.
Two runs were performed, and one fraction was collected
followed by concentration on a rotary evaporator and freeze-
drying. The purity (HPLC, LUNA 250 × 4.6 mm, Program 2,
0.1% HCOOH) was 93% (4α). The solid from fraction 2 (1.9
g) was dissolved in water (60 mL) for the second purification
step on the LUNA 250 × 50 mm column. Three runs were
performed, and two fractions were collected, concentrated on a
rotary evaporator and then purified further in additional
purification steps. Finally, after freeze-drying, two white solids
were obtained. The HPLC analyses (Method 2, LUNA 250 ×
1
Data for 4γ: H NMR (400 MHz, DMSO-d6) δ 9.99 (m, 1H,
Ar−NH), 8.75−8.50 (m, 1H, HN−CH2), 4.99 (m, 1H, N−
CH2−CH−CH2−OH), 4.75 (bs, 1H, HN−CH2−CH−OH),
4.53 (bs, 1H, HN−CH2−CH−CH2−OH), 4.21 (m, 1H, N−
CH2−CH), 3.68 (m, 1H, HN−CH2−CH), 3.54 and 3.50 (2m
in a 1:1 ratio, 2H, N−CH2−CH−CH2−OH), 3.47 and 3.39
(2m in a 1:1 ratio, 2H, HN−CH2−CH−CH2−OH), 3.41 and
3.22 (2m in a 1:1 ratio, 2H, N−CH2), 3.30 and 3.15 (2m in a
1:1 ratio, 2H, HN−CH2), 3.24 and 2.85 (2m in a 1:1 ratio, 2H,
CH2−COOH), 2.03 (s, 3H, HN−C(O)−CH3), 1.83 (s, 3H,
C−CH3), carboxylic acid proton not observed; 13C NMR (100
MHz, DMSO-d6) δ 170.5 (COOH), 169.4 (Ar−C(O)−NH),
167.8 (Ar−NH−C(O)), 165.6 (Ar−C(O)−N), 150.7 (ArC−
C(O)−NH), 148.1 (ArC−C(O)−N), 143.9 (ArC−NH), 100.0
(ArC−I), 98.3 (ArC−I), 94.3 (CH3−C−CH3), 90.0 (ArC−I),
75.9 (N−CH2−CH), 70.0 (HN−CH2−CH), 63.9 (HN−
CH2−CH−CH2−OH), 61.0 (N−CH2−CH−CH2−OH), 48.8
(N−CH2), 42.5 (HN−CH2), 41.0 (CH2−COOH), 23.1 (C−
CH3), 23.0 (C(O)−CH3). Data for 4δ: Mp 270 °C (dec); IR
(ATR) 3230 (br), 2929 (w), 1718 (m), 1627 (s), 1526 (m),
F
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
1
1359 (m), 1205 (m), 1107 (m) cm−1; H NMR (400 MHz,
(2m in a 1:1 ratio, 2H, HN−CH2−CH−CH2−O), 3.69 (m,
1H, HN−CH2−CH−OH), 3.47 and 3.40 (2m in a 1:1 ratio,
2H, HN−CH2−CH−CH2−OH), 3.38 and 3.20 (2m in a 1:1
ratio, 2H, NH−CH2−CH−O), 3.30 and 3.15 (2m in a 1:1
ratio, 2H, NH−CH2−CH−OH), 2.02 (s, 3H, HN−C(O)−
DMSO-d6) δ 10.00 (m, 1H, Ar−NH), 8.75−8.52 (m, 1H,
HN−CH2), 4.95 (bs, 1H, N−CH2−CH−CH2−OH), 4.73 (m,
1H, HN−CH2−CH−OH), 4.51 (bs, 1H, HN−CH2−CH−
CH2−OH), 4.19 (m, 1H, N−CH2−CH), 3.68 (m, 1H, HN−
CH2−CH), 3.53 and 3.50 (2m in a 1:1 ratio, 2H, N−CH2−
CH−CH2−OH), 3.47 and 3.39 (2m in a 1:1 ratio, 2H, HN−
CH2−CH−CH2−OH), 3.37 and 2.76 (2m in a 1:1 ratio, 2H,
CH2−COOH), 3.30 and 3.15 (2m in a 1:1 ratio, 2H, HN−
CH2), 3.17 and 3.11 (2m in a 1:1 ratio, 2H, N−CH2), 2.03 (s,
3H, HN−C(O)−CH3), 1.79 (s, 3H, C−CH3), carboxylic acid
proton not observed; 13C NMR (100 MHz, DMSO-d6) δ 170.4
(COOH), 169.4 (Ar−C(O)−NH), 167.9 (Ar−NH−C(O)),
165.6 (Ar−C(O)−N), 150.6 (ArC−C(O)−NH), 148.1 (ArC−
C(O)−N), 143.8 (ArC−NH), 100.1 (ArC−I), 98.5 (ArC−I),
94.4 (CH3−C−CH3), 89.4 (ArC−I), 75.9 (N−CH2−CH),
69.9 (HN−CH2−CH), 64.0 (HN−CH2−CH−CH2−OH),
61.3 (N−CH2−CH−CH2−OH), 49.3 (N−CH2), 42.6 (HN−
CH2), 42.6 (CH2−COOH), 23.0 (C(O)−CH3), 21.3 (C−
CH3); HRMS Calcd for C20H24N3I3O9Na+ [M + Na+]
853.8544, found 853.8547.
Synthesis of 3-(3-Acetamido-5-((2,2-dimethyl-1,3-dioxo-
lan-4-yl)methylcarbamoyl)-2,4,6-triiodobenzamido)-
propane-1,2-diyl Diacetate (5) and 5-Acetamido-N,N′-bis-
(2,2-dimethyl-1,3-dioxolan-4-yl)methyl-2,4,6-triiodoisoph-
thaldiamide (6). Acetone (14 mL, 11 g, 0.19 mol), acetic
anhydride (7.6 mL, 8.2 g, 0.081 mol), and acetic acid (22 mL)
were mixed at rt, and PTSA (0.4 g) and 1 (7.0 g, 0.0099 mol)
were then added. The reaction mixture was heated to reflux
(90−95 °C), and after 15 min the solution was cooled to rt.
Precipitation started during the cooling, and diethyl ether (40
mL) was added at rt to precipitate more solid. The precipitate
was filtered, washed with diethyl ether, and finally dried in a
vacuum oven at 50 °C. A HPLC sample was treated with a drop
of NaOH (50%, aq), and the pH was approximately 14 (pH
paper). After neutralizing with a little bit of acetic acid (pH 4−
5), the HPLC sample was analyzed (Supelco 250 × 4.6 mm;
Program 3). The main compounds were 2 (38%), 5 (39%), and
6 (17%).
CH3), 1.35 (s, 3H, O−C−CH3), 1.27 (s, 3H, O−C−CH3); 13
C
NMR (100 MHz, DMSO-d6) δ 169.7 (Ar−C(O), 2C), 167.7
(Ar−NH−C(O)), 150.1 (ArC−C(O)), 149.7 (ArC−C(O)),
143.4 (ArC−NH), 108.5 (CH3−C−CH3), 99.4 (ArC−I, 2C),
90.2 (ArC−I), 73.4 (HN−CH2−CH−O), 70.0 (HN−CH2−
CH−OH), 67.7 (HN−CH2−CH−CH2−O), 64.0 (HN−CH2−
CH−CH2−OH), 42.6 (NH−CH2−CH−OH), 41.9 (HN−
CH2−CH−O), 27.0 (O−C−CH3), 25.4 (O−C−CH3), 23.0
+
(HN−C(O)−CH3); LC-HRMS Calcd for C19H25I3N3O7 [M
+ H+] 787.8821, found 787.8851. Data for 6: Mp 306 °C (dec);
IR (ATR) 3246 (m), 2984 (w), 2935 (w), 2876 (w), 1633 (s),
1553 (m), 1509 (m), 1369 (m), 1267 (m), 1212 (m) cm−1; 1H
NMR (400 MHz, DMSO-d6) δ 9.96 (s, 1H, Ar−NH), 8.86−
8.43 (m, 2H, HN−CH2), 4.25 (m, 2H, HN−CH2−CH), 4.06
and 3.81 (2m in a 1:1 ratio, 4H, HN−CH2−CH−CH2), 3.39
and 3.19 (2m in a 1:1 ratio, 4H, HN−CH2), 2.02 (s, 3H, HN−
C(O)−CH3), 1.36 (s, 6H, (O)C(O)−CH3), 1.27 (s, 6H,
(O)C(O)−CH3); 13C NMR (100 MHz, DMSO-d6) δ 169.6
(Ar−C(O), 2C), 167.7 (Ar−NH−C(O)), 149.8 (ArC−C(O),
2C), 143.4 (ArC−NH), 108.5 (CH3−(O)C(O)−CH3, 2C),
99.5 (ArC−I, 2C), 90.2 (ArC−I), 73.3 (HN−CH2−CH, 2C),
67.7 (HN−CH2−CH−CH2, 2C), 41.9 (HN−CH2, 2C), 27.0
((O)C(O)−CH3, 2C), 25.4 ((O)C(O)−CH3, 2C), 23.0
+
(HN−C(O)−CH3); LC-HRMS Calcd for C22H29I3N3O7 [M
+ H+] 827.9134, found 827.9086.
Decomposition of 5 and 6 under Aqueous Con-
ditions. Acetone (7.0 mL, 5.5 g, 0.095 mol), acetic anhydride
(3.8 mL, 4.1 g, 0.040 mol), and acetic acid (11 mL) were mixed
at room temperature, and PTSA (0.2 g) and 1 (3.5 g, 0.0050
mol) were then added. The resulting mixture was heated to
reflux, and after 15 min a sample was collected for HPLC
analysis (Supelco 250 × 4.6 mm; Program 3). Prior to analysis
the sample was diluted with water, treated with a drop of
aqueous NaOH (50%), and neutralized with acetic acid (pH
4−5). The chromatogram revealed that the product mixture
contained 2 (19%), 3 (5%), 5 (32%), and 6 (17%) (see Figure
4a). The sample was further treated with concentrated H2SO4
until the pH was approximately 0 (pH paper), and after storage
for 5−10 min the sample was analyzed again under the
conditions mentioned above. The chromatogram revealed that
the product mixture now contained 2 (70%) and 3 (10%), but
not even traces of 5 and 6 (see Figure 4b).
Synthesis of N,N′-Bis(2,3-dihydroxypropyl)-5-nitroisoph-
thaldiamide (7). Dimethyl 5-nitroisophthalate (40.0 g, 0.167
mol) and 3-amino-1,2-propandiol (36.4 g, 0.400 mol) were
mixed in methanol (140 mL). The mixture was heated to reflux
(70−75 °C), stirred for 70 h, and then cooled to rt. During the
cooling crystallization occurred. The slurry was filtered, the
solid was washed with methanol, and the crystals were finally
dried under vacuum at 50 °C to give 7 (54.4 g, 91%) as a white
solid. Mp 131−134 °C (literature 128−132 °C).14 The purity
(HPLC; Brownlee 250 × 4.6 mm; Program 5) was 99%. NMR
data was in accordance to the literature.15
Preparative HPLC was used for isolation which was carried
out in one step. Three runs were performed on the LUNA 250
× 50 mm column. For each run, solid (200−1100 mg) from the
experiment was diluted with methanol (5 mL) and then water
(45 mL). Some drops of an aqueous KOH solution (10 M)
were added to hydrolyse the ester groups. When the hydrolysis
was complete, the pH was adjusted to 4−5 by adding some
drops of acetic acid. After the pH adjustment, the solution was
injected into the column, and two fractions were collected. The
fractions were then concentrated on a rotary evaporator in the
presence of small amounts of ammonia to keep a weak alkaline
solution during the concentration. Precipitation occurred in the
second fraction, and the precipitate was filtered and air-dried.
The concentrate from the first fraction was freeze-dried, and
the weight of the solid obtained was 525 mg. For the second
fraction the weight was 160 mg (white solid). The purity of 5
(HPLC; LUNA 250 × 4.6 mm; Program 4) was >99% and of 6
98%. Data for 5: Mp 285 °C (dec); IR (ATR) 3244 (br), 2984
(w), 2935 (w), 2880 (w), 1633 (s), 1553 (m), 1510 (m), 1370
(m), 1268 (m), 1216 (m), 1047 (m) cm−1; H NMR (400
N-(2,3-Dihydroxypropyl)-N′-(2,2-dimethyl-1,3-dioxolan-4-
yl)methyl-5-nitroisophthaldi-amide (8) and N,N′-bis((2,2-
dimethyl-1,3-dioxolan-4-yl)methyl-5-nitroisophthaldiamide
(9). Acetone (50 mL, 40 g, 0.68 mol) and acetic anhydride (8.0
mL, 8.6 g, 0.085 mol) were mixed at rt, and PTSA (0.4 g) and 7
1
MHz, DMSO-d6) δ 9.95 (s, 1H, Ar−NH), 8.84−8.66 and
8.19−8.09 (m, 1H, HN−CH2−CH−O), 8.62−8.44 (m, 1H,
HN−CH2−CH−OH), 4.71 (m, 1H, CH−OH), 4.50 (m, 1H,
CH2−OH), 4.24 (m, 1H, HN−CH2−CH−O), 4.06 and 3.82
G
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(3.5 g, 0.0098 mol) were then added. After stirring the mixture
for 3 h at rt, the solution was diluted with CH2Cl2 (150 mL),
and the organic phase was washed with 2 M aqueous NaOH (2
× 50 mL), H2O (50 mL), and brine (30 mL). The organic
phase was then dried over MgSO4, filtered, and concentrated
on a rotary evaporator. A HPLC sample of the residue was
treated with NaOH and then neutralized with acetic acid before
being analysed by HPLC (Supelco 250 × 4.6 mm; Program 3)
which revealed the presence of 7 (2%), 8 (20%), and 9 (56%).
Preparative HPLC was used for isolation which was carried out
in two steps. For the first step, two runs were performed on the
LUNA 250 × 50 mm column. For each run, the solution
injected was made from some of the residue from the
experiment (10−15 mL) which was diluted with methanol
(10−20 mL) followed by water (60 mL). One fraction only was
collected during the two runs. The fraction was concentrated
under vacuum on a rotary evaporator in the presence of
ammonia to keep it alkaline. The resulting concentrate was
then treated with KOH and stirred for 5 min followed by acetic
acid until pH was approximately 5 (pH strip). The second
purification step was carried out on the same column as the first
step, and two fractions were collected, followed by freeze-
drying. The weight of the pale yellow, highly viscous liquid
obtained from the first fraction was 270 mg. The purity of 8
(HPLC; LUNA 250 × 4.6 mm; Program 4) was >99%. The
weight of the off-white glazed solid from the second fraction
was 1030 mg. The purity of 9 (HPLC; LUNA 250 × 4.6 mm;
Program 4) was >99%. Data for 8: IR (ATR) 3305 (br), 3085
(w), 2986 (w), 2936 (w), 2882 (w), 1643 (s), 1529 (s), 1211
Acetone (25 mL, 20 g, 0.34 mol) and acetic anhydride (4.0 mL,
4.3 g, 0.042 mol) were mixed at room temperature, and PTSA
(0.2 g) and 7 (1.8 g, 5.0 mmol) were then added. The reaction
mixture was stirred at rt for 72 h. The mixture was then diluted
with ethyl acetate (200 mL), and the organic phase was washed
with 2 M aqueous NaOH (3 × 50 mL), water (50 mL), and
brine (30 mL). Then the organic phase was dried over MgSO4,
filtered, and concentrated on a rotary evaporator. A HPLC
sample of the residue was treated with NaOH and then
neutralized with acetic acid before being analysed by HPLC
(Supelco 250 × 4.6 mm; Program 3) which revealed the
presence of 7 (9%), 8 (23%), 9 (19%), 10 (8%), and 11 (14%).
Preparative HPLC was used for isolation which was carried
out in two steps. For the first step, two runs were performed on
the LUNA 250 × 50 mm column. For each run, residue (10−
15 mL) from the experiment was diluted with methanol (30−
40 mL) and then water (60 mL). One fraction was collected
followed by concentration on a rotary evaporator in the
presence of small amounts of ammonia to keep a weak alkaline
solution during the concentration. The concentrate was then
freeze-dried. Prior to the second purification step the solid (250
mg) was dissolved in methanol (20 mL) and then diluted with
water (30 mL). Aqueous 10 M KOH was then added, and after
stirring for 5 min at rt acetic acid was added until pH was
approximately 5 (pH strip). The second purification step was
carried out on the same column as the first step. Several
fractions were collected during the two runs. Each fraction was
analysed by HPLC, and based on the HPLC-results some of
them were combined to give two fractions. The weight of the
off-white solid (freeze-dried) obtained from the first fraction
was 19 mg. The purity of 10 (HPLC, Program 4, LUNA 250 ×
4.6 mm) was 88%. The weight of the off-white solid (freeze-
dried) from the second fraction was 99 mg. The purity of 11
(HPLC) was 89%. Data for 10: Mp 240 °C (dec); IR (ATR)
3307 (m), 3080 (w), 2983 (w), 2934 (w), 2875 (w), 2728 (w),
1
(m), 1045 (m) cm−1; H NMR (400 MHz, DMSO-d6) δ 9.11
(t, 1H, HN−CH2−CH−O), 8.92 (t, 1H, HN−CH2−CH−
OH), 8.82 (t, 1H, ArH), 8.80 (t, 1H, ArH), 8.76 (t, 1H, ArH),
4.88 (d, 1H, CH−OH), 4.61 (t, 1H, CH2−OH), 4.24 (qui, 1H,
HN−CH2−CH), 4.01 and 3.71 (2dd in a 1:1 ratio, 2H, HN−
CH2−CH−CH2), 3.69 (m, 1H, HN−CH2−CH−OH), 3.45
and 3.23 (2m in a 1:1 ratio, 2H, NH−CH2−CH−OH), 3.43
(m, 2H, NH−CH2−CH−O), 3.37 (m, 2H, HN−CH2−CH−
CH2−OH), 1.35 (s, 3H, CH3), 1.26 (s, 3H, CH3); 13C NMR
(100 MHz, DMSO-d6) δ 164.1 (Ar−C(O)), 164.0 (Ar−
C(O)), 147.8 (ArC−NO2), 136.3 (ArC−C(O)), 135.8 (ArC−
C(O)), 132.5 (ArC), 124.4 (ArC), 124.2 (ArC), 108.5 (CH3−
C−CH3), 74.1 (HN−CH2−CH−O), 70.2 (HN−CH2−CH−
OH), 66.8 (HN−CH2−CH−CH2−O), 64.0 (HN−CH2−CH−
CH2−OH), 43.3 (HN−CH2−CH−OH), 42.4 (HN−CH2−
CH−O), 26.9 (CH3), 25.4 (CH3); LC-HRMS Calcd for
C17H24N3O8+ [M + H+] 398.1558, found 398.1554. Data for 9:
Mp 83−86 °C; IR (ATR) 3310 (w), 3082 (w), 2985 (w), 2936
(w), 2882 (w), 1644 (s), 1529 (s), 1349 (m), 1210 (m), 1154
(m), 1080 (m), 1051 (m), 834 (m) cm−1; 1H NMR (400 MHz,
DMSO-d6) δ 9.11 (t, 2H, NH), 8.81 (s, 2H, ArH), 8.76 (s, 1H,
ArH), 4.24 (qui, 2H, HN−CH2−CH), 4.01 and 3.71 (2dd in a
1:1 ratio, 4H, HN−CH2−CH−CH2), 3.43 (m, 4H, HN−CH2),
1.35 (s, 6H, CH3), 1.26 (s, 6H, CH3); 13C NMR (100 MHz,
DMSO-d6) δ 164.0 (Ar−C(O), 2C), 147.8 (ArC−NO2), 135.9
(ArC−C(O), 2C), 132.5 (ArC), 124.4 (ArC, 2C), 108.5
(CH3−C−CH3, 2C), 74.1 (HN−CH2−CH, 2C), 66.8 (HN−
CH2−CH−CH2, 2C), 42.4 (HN−CH2, 2C), 26.9 (CH3, 2C),
25.4 (CH3, 2C); LC-HRMS Calcd for C20H28N3O8+ [M + H+]
438.1871, found 438.1829.
1
1628 (s), 1533 (s), 1348 (m), 1040 (s) cm−1. H NMR (400
MHz, DMSO-d6) δ 8.95 (t, 1H, NH), 8.75 (t, 1H, ArH), 8.44
(t, 1H, ArH), 8.39 (t, 1H, ArH), 4.90 (m, 1H, CH−OH), 4.90
(m, 1H, N−CH2−CH−CH2−OH), 4.62 (t, 1H, HN−CH2−
CH−CH2−OH), 4.12 (m, 1H, N−CH2−CH), 3.67 (m, 1H,
HN−CH2−CH), 3.49 (t, 2H, N−CH2−CH−CH2), 3.47 and
3.40 (2m in a 1:1 ratio, 2H, N−CH2), 3.45 and 3.21 (2m in a
1:1 ratio, 2H, HN−CH2), 3.36 (m, 2H, HN−CH2−CH−CH2),
1.67 (s, 3H, CH3), 1.63 (s, 3H, CH3); 13C NMR (100 MHz,
DMSO-d6) δ 163.9 (Ar−C(O)), 163.7 (Ar−C(O)), 147.8
(ArC−NO2), 139.3 (ArC−C(O)−N), 136.2 (ArC−C(O)−
NH), 131.5 (ArC), 123.6 (ArC), 123.0 (ArC), 94.8 (CH3−C−
CH3), 75.4 (N−CH2−CH), 70.1 (HN−CH2−CH), 64.0
(HN−CH2−CH−CH2−OH), 61.2 (N−CH2−CH−CH2−
OH), 50.2 (N−CH2), 43.4 (HN−CH2), 25.6 (CH3), 23.7
(CH3); LC-HRMS Calcd for C17H24N3O8 [M + H+]
+
398.1558, found 398.1555. Data for 11: Mp 115−120 °C; IR
(ATR) 3326 (br), 3082 (w), 2984 (w), 2936 (w), 2882 (w),
1630 (s), 1534 (s), 1348 (m), 1247 (m), 1209 (m), 1048 (s),
1
841 (m), 716 (m) cm−1; H NMR (400 MHz, DMSO-d6) δ
9.13 (t, 1H, NH), 8.75 (s, 1H, ArH), 8.46 (s, 1H, ArH), 8.39 (s,
1H, ArH), 4.90 (t, 1H, OH), 4.24 (qui, 1H, HN−CH2−CH),
4.13 (m, 1H, N−CH2−CH), 4.01 (dd, 1H, HN−CH2−CH−
CH2), 3.71 (m, 1H, HN−CH2−CH−CH2), 3.49 (m, 2H, N−
CH2−CH−CH2−OH), 3.48 and 3.41 (2m in a 1:1 ratio, 2H,
N−CH2), 3.43 (m, 2H, NH−CH2), 1.68 (s, 3H, N−C−CH3),
1.63 (s, 3H, N−C−CH3), 1.35 (s, 3H, O−C−CH3), 1.26 (s,
3H, O−C−CH3); 13C NMR (100 MHz, DMSO-d6) δ 163.8
N-(2,3-Dihydroxypropyl)-3-(5-hydroxymethyl-2,2-dime-
thyloxazolidine-3-carbonyl)-5-nitrobenzamide (10) and N-
(2,2-Dimethyl-1,3-dioxolan-4-yl)methyl-3-(5-hydroxymethyl-
2,2-dimethyloxazolidine-3-carbonyl)-5-nitrobenzamide (11).
H
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(Ar−C(O)), 163.7 (Ar−C(O)), 147.8 (ArC−NO2), 139.4
(ArC−C(O)−N), 135.8 (ArC−C(O)−NH), 131.1 (ArC),
123.8 (ArC), 123.0 (ArC), 108.5 (CH3−(O)C(O)−CH3),
94.8 (CH3−(O)C(N)−CH3), 75.4 (N−CH2−CH), 74.1
(HN−CH2−CH), 66.8 (HN−CH2−CH−CH2), 61.2 (N−
CH2−CH−CH2), 50.2 (N−CH2), 42.4 (HN−CH2), 26.9
((O)C(O)−CH3), 25.6 ((N)C(O)−CH3), 25.4 ((O)C(O)−
CH3), 23.7 ((N)C(O)-CH3); LC-HRMS Calcd for
The solution was carefully cooled to 5 °C, and during the
cooling diethyl ether (650 mL) was added. Next day the slurry
was filtered and washed with a mixture of methanol and diethyl
ether. The recrystallized white solid was finally dried in a
vacuum oven to obtain 12c (106 g, 74%). The purity (HPLC)
was 98%. Data for 12c: Mp 128−129 °C; IR (ATR) 3283 (br),
3094 (w), 2952 (w), 2868 (w), 1628 (s), 1567 (m), 1546 (s),
1336 (m), 1314 (m), 1281 (m), 1052 (s), 1000 (m), 675 (s)
+
1
C20H28N3O8 [M + H+] 438.1871, found 438.1848.
cm−1; H NMR (400 MHz, DMSO-d6) δ 8.57 (t, 2H, NH),
Synthesis of Model Compounds 12. N,N′-Bis(2-
hydroxyethyl)isophthalamide (12a). Dimethyl isophthalate
(70.0 g, 0.360 mol) was added to 2-methoxyethanol (140 mL),
and then ethanol amine (56.0 g, 0.918 mol) was added to the
mixture. The solution was immediately heated to reflux (125−
130 °C), and the mixture was refluxed in 20 h. Complete
conversion was confirmed by HPLC. The solution was cooled
to room temperature, and then diethyl ether (250 mL) was
added for crystallization. The slurry was filtered, and the solid
was washed with a mixture of diethyl ether and 2-
methoxyethanol. The white solid was finally dried in a vacuum
oven to obtain 12a (51.0 g, 56%). The purity (HPLC, Program
5, Brownlee 250 × 4.6 mm) was >99%. Data for 12a: Mp 145−
146 °C; IR (ATR) 3384 (m), 3254 (br), 3075 (w), 2970 (w),
2925 (w), 2870 (w), 1635 (s), 1542 (s), 1301 (m), 1056 (s),
875 (m), 687 (s) cm−1; 1H NMR (400 MHz, DMSO-d6) δ 8.54
(t, 2H, NH), 8.33 (s, 1H, ArH), 7.97 (d, 2H, ArH), 7.54 (t, 1H,
ArH), 4.77 (t, 2H, OH), 3.53 (m, 4H, HN−CH2−CH2), 3.36
(m, 4H, HN−CH2); 13C NMR (100 MHz, DMSO-d6) δ 166.0
(Ar−C(O), 2C), 134.7 (ArC−C(O), 2C), 129.7 (ArC, 2C),
128.3 (ArC), 126.2 (ArC), 59.8 (HN−CH2−CH2, 2C), 42.3
(HN−CH2, 2C); HRMS Calcd for C12H16N2O4Na+ [M + Na+]
275.1008, found 275.1007.
N,N′-Bis(2-hydroxyethyl)-5-nitroisophthaldiamide (12b).
Dimethyl 5-nitroisophthalate (40.0 g, 0.167 mol) was added
to methanol (140 mL), and then ethanol amine (24.9 g, 0.408
mol) was added to the mixture. The solution was immediately
heated to reflux (70−75 °C), and after 70 h it was cooled to
room temperature. During the cooling crystallization occurred.
The slurry was filtered, and the solid was washed with
methanol. The white solid was finally dried in a vacuum oven to
give 12b (46.8 g, 94%). The purity (HPLC; Brownlee 250 ×
4.6 mm; Program 5) was >99%. Data for 12b: Mp 150−151
°C; IR (ATR) 3281 (br), 3076 (w), 2921 (w), 1647 (s), 1549
(m), 1529 (s), 1346 (m), 1327 (m), 1300 (m), 1083 (m), 1049
(m), 918 (m), 674 (s) cm−1; 1H NMR (400 MHz, DMSO-d6)
δ 8.95 (t, 2H, NH), 8.80 (s, 2H, ArH), 8.76 (s, 1H, ArH), 4.81
(t, 2H, OH), 3.55 (m, 4H, HN−CH2−CH2), 3.38 (m, 4H,
HN−CH2); 13C NMR (100 MHz, DMSO-d6) δ 163.9 (Ar−
C(O), 2C), 147.8 (ArC−NO2), 136.2 (ArC−C(O), 2C), 132.3
(ArC), 126.2 (ArC, 2C), 59.5 (HN−CH2−CH2, 2C), 42.5
(HN−CH2, 2C); HRMS Calcd for C12H15N3O6Na+ [M + Na+]
320.0859, found 320.0858.
8.28 (s, 1H, ArH), 7.95 (d, 2H, ArH), 7.54 (t, 1H, ArH), 4.51
(bs, 2H, OH), 3.47 (t, 4H, HN−CH2−CH2−CH2), 3.33 (q,
4H, HN−CH2), 1.69 (qui, 4H, HN−CH2−CH2); 13C NMR
(100 MHz, DMSO-d6) δ 165.9 (Ar−C(O), 2C), 134.8 (ArC−
C(O), 2C), 129.6 (ArC, 2C), 128.3 (ArC), 126.1 (ArC), 58.6
(HN−CH2−CH2−CH2, 2C), 36.7 (HN−H2, 2C), 32.4 (HN−
CH2−CH2, 2C); HRMS Calcd for C14H20N2O4Na+ [M + Na+]
303.1321, found 303.1320.
N,N′-Bis(3-hydroxypropyl)-5-nitroisophthalamide (12d).
Dimethyl 5-nitroisophthalate (40.0 g, 0.167 mol) was added
to methanol (140 mL), and then 3-amino-1-propanol (30.2 g,
0.402 mol) was added to the mixture. The solution was
immediately heated to reflux (70−75 °C). After 70 h the
solution was cooled to room temperature, and during the
cooling crystallization occurred. The slurry was filtered, and the
white solid was washed with methanol before drying in a
vacuum oven to obtain 12d (50.0 g, 92%). The purity (HPLC;
Brownlee 250 × 4.6 mm; Program 5) was >99%. Data for 12d:
Mp 129−130 °C; IR (ATR) 3279 (br), 3084 (w), 2943 (w),
2878 (w), 1638 (s), 1555 (m), 1534 (s), 1330 (s), 1303 (m),
1
1050 (s), 1033 (s), 1000 (s), 696 (s) cm−1; H NMR (400
MHz, DMSO-d6) δ 8.94 (t, 2H, NH), 8.77 (s, 2H, ArH), 8.73
(s, 1H, ArH), 4.51 (t, 2H, OH), 3.48 (q, 4H, HN−CH2−CH2−
CH2), 3.36 (m, 4H, HN−CH2), 1.71 (qui, 4H, HN−CH2−
CH2); 13C NMR (100 MHz, DMSO-d6) δ 163.7 (Ar−C(O),
2C), 147.9 (ArC−NO2), 136.3 (ArC−C(O), 2C), 132.2 (ArC),
124.1 (ArC, 2C), 58.6 (HN−CH2−CH2−CH2, 2C), 37.0
(HN−CH2, 2C), 32.2 (HN−CH2−CH2, 2C); HRMS Calcd for
C14H19N3O6Na+ [M + Na+] 348.1172, found 348.1171.
3-(2,2-Dimethyloxazolidine-3-carbonyl)-N-(2-
hydroxyethyl)benzamide (13a). Acetone (75 mL, 59 g, 1.0
mol) and acetic anhydride (6.0 mL, 6.5 g, 0.064 mol) were
mixed at room temperature. PTSA (0.6 g) and 12a (3.8 g,
0.015 mol) were then added. After 3 h the solution was diluted
in dichloromethane (200 mL), and the organic phase was then
washed with aqueous NaOH (2 M, 2 × 50 mL), water (50 mL)
and brine (30 mL). The organic phase was finally dried over
MgSO4, and solvents were removed on a rotary evaporator.
The white solid obtained was dried further under vacuum at 50
°C and weighed 2.7 g. A HPLC sample of the solid was treated
with NaOH, neutralized with acetic acid, and analysed by
HPLC (Supelco 250 × 4.6 mm; Program 3). The amount of
13a was 5%. Preparative HPLC was used for isolation. Two
runs were performed on the LUNA 250 × 50 mm column.
Solid (1150−1300 mg) was dissolved in methanol (20 mL) and
then diluted with water (80 mL). Some aqueous KOH (10 M)
was added, and when the pH was stable at 13−14 (pH paper),
it was adjusted to approximately 5 by adding a little acetic acid.
One fraction was collected and subsequently concentrated on a
rotary evaporator in the presence of small amounts of ammonia
to keep the solution slightly alkaline. The concentrate was
finally freeze-dried, and the weight of the pale yellow, highly
viscous liquid obtained was 41 mg. The purity of 13a (HPLC;
LUNA 250 × 4.6 mm; Program 4) was 83%. IR (ATR) 3310
N,N′-Bis(3-hydroxypropyl)isophthalamide (12c). Dimethyl
isophthalate (100 g, 0.515 mol) was added to 3-amino-1-
propanol (117 g, 1.56 mol). The solution was immediately
heated to 120−140 °C, and after 2.5 h methanol (330 mL) was
added for crystallization. Then the solution was carefully cooled
to 5 °C. During the cooling crystallization occurred. The slurry
was stirred overnight at 5 °C, and then the solid was filtered
and washed with methanol. The white solid was finally dried in
a vacuum oven to obtain 12c (130 g, 90%). The purity (HPLC;
Brownlee 250 × 4.6 mm; Program 5) was 96%. The crude was
dissolved in methanol (400 mL) at 60 °C for recrystallization.
I
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Notes
(br), 3077 (w), 2983 (w), 2937 (w), 2883 (m), 1619 (s), 1535
1
(s), 1433 (m), 1240 (s), 1145 (m), 1054 (s) cm−1; H NMR
The authors declare the following competing financial
interest(s): T. Hland is employed by GE Healthcare which is
a co-funder of the scholarship that paid for his PhD studies.
(400 MHz, DMSO-d6) δ 8.58 (t, 1H, NH), 7.97 (s, 1H, ArH),
7.94 (d, 1H, ArH), 7.64 (d, 1H, ArH), 7.52 (t, 1H, ArH), 4.76
(bs, 1H, OH), 3.90 (t, 2H, N−CH2−CH2), 3.50 (m, 2H, N−
CH2), 3.50 (m, 2H, HN−CH2−CH2), 3.34 (m, 2H, HN−
CH2), 1.61 (s, 6H, CH3); 13C NMR (100 MHz, DMSO-d6) δ
166.1 (Ar−C(O)−N), 165.5 (Ar−C(O)−NH), 137.8 (ArC−
C(O)−NH), 134.5 (ArC−C(O)−N), 128.9 (ArC), 128.5
(ArC), 128.4 (ArC), 125.1 (ArC), 94.9 (CH3−C−CH3), 63.1
(N−CH2−CH2), 59.7 (HN−CH2−CH2), 48.2 (N−CH2), 42.2
(HN−CH2), 24.2 (CH3, 2C); LC-HRMS Calcd for
ACKNOWLEDGMENTS
■
Financial support from the Research Council of Norway, the
University of Bergen and GE Healthcare is gratefully
acknowledged. Skillful recording of mass spectra by Dr. Bjarte
Holmelid and Dr. Erlend Hvattum is highly appreciated.
Thanks are also due to Dr. Nils Å. Frøystein, for valuable
discussions about NMR problems, and Willy Skjøld, for
valuable assistance with preparative HPLC isolation.
C15H21N2O4 [M + H+] 293.1496, found 293.1460.
+
3-(2,2-Dimethyloxazolidine-3-carbonyl)-N-(2-hydroxyeth-
yl)-5-nitrobenzamide (13b). Acetone (100 mL, 79.1 g, 1.36
mol) and acetic anhydride (8.0 mL, 8.7 g, 0.085 mol) were
mixed at rt. PTSA (0.8 g) and 12b (5.8 g, 0.015 mol) were then
added. After 5 h the solution was diluted with ethyl acetate
(300 mL), and the organic phase was washed with NaOH (2
M, 2 × 50 mL), water (50 mL) and brine (30 mL). The organic
phase was finally dried over MgSO4, and solvents were removed
on a rotary evaporator. A HPLC sample of the residue was
treated with NaOH, neutralized with acetic acid, and analysed
by HPLC (Supelco 250 × 4.6 mm; Program 3). The amount of
13b was 8%. Preparative HPLC was used for isolation. Three
runs were performed on the LUNA 250 × 50 mm column. The
residue (5−10 mL) was diluted in methanol (10 mL) and then
diluted further with water (80 mL). Some aqueous KOH (10
M) was added, and when the pH was stable at 13−14 (pH
paper), it was adjusted to approximately 5 by adding a little
acetic acid. One fraction was collected and subsequently
concentrated on a rotary evaporator in the presence of small
amounts of ammonia to keep the solution slightly alkaline. The
concentrate was finally freeze-dried, and the weight of the white
solid obtained was 205 mg. The purity of 13b (HPLC; LUNA
250 × 4.6 mm; Program 4) was 97%. Mp 173−174 °C; IR
(ATR) 3424 (br), 3292 (m), 3088 (w), 2991 (w), 2935 (w),
2880 (w), 1622 (s), 1535 (m), 1328 (m), 1251 (m), 1057 (s)
REFERENCES
■
(1) Wuts, P. G. M.; Greene, T. W. Greene’s Protective Groups in
Organic Synthesis; Wiley & Sons: Hoboken, 2007.
(2) Bellamy, L. J. The Infra-red Spectra of Complex Molecules;
Chapman and Hall: London, 1975; ch. 10.
(3) Stothers, J. B. Carbon-13 NMR Spectroscopy; Academic Press:
New York, 1972; ch. 8.
(4) Zoeller, J. R.; Agreda, V. H.; Cook, S. L.; Lafferty, N. L.;
Polichnowski, S. W.; Pond, D. M. Catal. Today 1992, 13, 73.
(5) Painter, E. S.; Scott, H. A. Manufacture of lower aliphatic acid
anhydrides. US Patent US2743296 A, 1956.
(6) Boese, A. B. Ind. Eng. Chem. 1940, 32, 16.
(7) Carey, F. A. Organic Chemistry; McGraw Hill: New York, 2003;
ch. 17.
(8) Taylor, W. G.; Hall, T. W.; Schreck, C. E. Can. J. Chem. 1992, 70,
165.
(9) Taylor, W. G. J. Agric. Food Chem. 1982, 30, 409.
(10) Yamada, S.; Inoue, M. Org. Lett. 2007, 9, 1477.
(11) Tanaka, M.; Oishi, S.; Ohno, H.; Fujii, N. Int. J. Pept. Res. Ther.
2007, 13, 271.
(12) Fossheim, R.; Gulbrandsen, T.; Priebe, H.; Aasen, A. J. Acta
Chem. Scand. 1995, 49, 589.
(13) Bates, R. W.; Lu, Y.; Cai, M. P. Tetrahedron 2009, 65, 7852.
(14) Haavaldsen, J.; Nordal, V.; Kelly, M. Acta Pharm. Suecica 1983,
20, 219.
(15) Gonzal
́
ez Tavares, L.; Carretero, J. M.; Harto Martinez, J. R.;
Martin Jimen
́
es, J. L.; Martinez Sanz, A.; Riefke, B.; Gries, H. New
brominated compounds as contrast media for X-ray mammography.
1
cm−1; H NMR (400 MHz, DMSO-d6) δ 8.99 (t, 1H, NH),
EP1186305 A1, 2002.
8.74 (t, 1H, ArH), 8.45 (t, 1H, ArH), 8.41 (t, 1H, ArH), 4.80
(bs, 1H, OH), 3.92 (t, 2H, N−CH2−CH2), 3.54 (t, 4H, N−
CH2 and HN−CH2−CH2), 3.37 (q, 2H, HN−CH2), 1.63 (s,
6H, CH3); 13C NMR (100 MHz, DMSO-d6) δ 163.8 (Ar−
C(O)−N), 163.5 (Ar−C(O)−NH), 147.8 (ArC−NO2), 139.3
(ArC−C(O)−N), 136.2 (ArC−C(O)−NH), 131.1 (ArC),
123.6 (ArC), 123.1 (ArC), 94.3 (CH3−C−CH3), 63.3 (N−
CH2−CH2), 59.5 (HN−CH2−CH2), 48.1 (N−CH2), 42.6
(HN−CH2), 24.1 (CH3, 2C); LC-HRMS Calcd for
C15H20N3O6 [M + H+] 338.1347, found 338.1320.
+
ASSOCIATED CONTENT
■
S
* Supporting Information
1H NMR and 13C NMR spectra for all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +4755583450. E-mail: leiv.sydnes@kj.uib.no.
J
dx.doi.org/10.1021/op500177w | Org. Process Res. Dev. XXXX, XXX, XXX−XXX