Study on the synthesis of toluene-2,4-diisocyanate via amine and carbonyl fluoride

Article abstract of DOI:10.1016/j.jfluchem.2015.07.025

Abstract We presented recently a synthesis of toluene-2,4-diisocyanate (TDI) as one of the nine examples to verify the feasibility of an industrial appealing two-step method for isocyanates synthesis via amines and carbonyl fluoride (COF₂). Because more investigation was considered to be necessary for future industrial application, the two-step synthesis processes of TDI were studied in detail herein. The total yield of TDI increased 5.6% under the optimized experimental conditions. The influence of the excess COF₂ left in the reaction system was studied carefully according to a one-pot method.

Full text of DOI:10.1016/j.jfluchem.2015.07.025

Journal of Fluorine Chemistry 178 (2015) 208–213
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Study on the synthesis of toluene-2,4-diisocyanate via amine and
carbonyl fluoride
b
Ni Zhang a, Xiaomeng Zhou b, Hengdao Quan a,b, , Akira Sekiya
*
a School of Chemical Engineering & Environment, Beijing Institute of Technology, 5 South Zhongguangcun Street, Haidian District, Beijing 100081, PR China
b National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan
A R T I C L E I N F O
A B S T R A C T
Article history:
We presented recently a synthesis of toluene-2,4-diisocyanate (TDI) as one of the nine examples to verify
the feasibility of an industrial appealing two-step method for isocyanates synthesis via amines and
carbonyl fluoride (COF2). Because more investigation was considered to be necessary for future
industrial application, the two-step synthesis processes of TDI were studied in detail herein. The total
yield of TDI increased 5.6% under the optimized experimental conditions. The influence of the excess
COF2 left in the reaction system was studied carefully according to a one-pot method.
ß 2015 Elsevier B.V. All rights reserved.
Received 22 June 2015
Received in revised form 16 July 2015
Accepted 24 July 2015
Available online 29 July 2015
Keywords:
Toluene diisocyanate
Carbonyl fluoride
Amine
Synthesis
1. Introduction
and a chemical intermediate to prepare other fluorinated
compounds [13–15] and generally more expensive than phosgene.
Using COF2 instead of phosgene to react with amines for
isocyanates production will inevitably increase the raw material
costs, but the costs of equipment, facilities, security and
environmental protection, etc. will be decreased accordingly due
to the relatively lower potential risk of COF2. Besides, the novel
synthesis method of COF2 invented by our laboratory [16] and the
recycle of by-product HF and raw material COF2 could decrease
the total costs further [12]. Synthetically, there is no much
difference about the total costs between the COF2 method and
phosgene method if the total yields are basically the same.
However, the investigation for the synthesis of TDI via TDA and
COF2 was not detailed in the previous report [12], in which the
other eight different isocyanates were synthesized and the focus
was placed on the verification about the feasibility of COF2 method.
Herein, six main factors of the experimental conditions for the two-
step method are further studied and optimized, and the influence
of the excess COF2 left in the reaction system is discussed carefully
via a one-port method.
Toluene-2,4-diisocyanate (TDI) is the major raw material for
synthesis of both flexible and rigid polyurethane, which could be
used as automotive cushions, carpet underlay, furniture, insulators
in construction, pipeline, etc. [1]. Presently, the phosgenation of
the 2,4-diaminotoluene (TDA) or its salt precursors is still the main
production method in a commercial scale [1–3], but several
alternative methods have been developed to reduce or eliminate
its security and environment risks, such as the one-step
carbonylation of 2,4-dinitrotoluene (DNT) [4,5]. During the recent
two decades, many researchers paid their attentions to one such
method that involves the production of carbamate esters by
oxidative carbonylation of TDA [6–9] or reductive carbonylation of
DNT [10,11], dealcoholysis of which gives TDI. But to the best of our
knowledge, none of them are applied in industry so far due to
several drawbacks.
We showed recently a facile two-step method that reaction of
amines and COF2 to yield the corresponding isocyanates could be
facilitated under relatively mild conditions and in good yields, by
which the TDI was synthesized as an example in 87.4% yield based
on TDA [12]. It is well known that COF2 is an attractive and
important compound as a semiconductor etching gas, cleaning gas,
2. Results and discussion
2.1. Synthesis of TDI via two-step method
The method is consisted of two steps. The first step is the
reaction between COF2 and TDA at the room temperature or near.
The second step is the decomposition of the intermediate
*
Corresponding author at: National Institute of Advanced Industrial Science and
Technology (AIST), 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan.
E-mail address: hengdao-quan@aist.go.jp (H. Quan).
http://dx.doi.org/10.1016/j.jfluchem.2015.07.025
0022-1139/ß 2015 Elsevier B.V. All rights reserved.

N. Zhang et al. / Journal of Fluorine Chemistry 178 (2015) 208–213
209
CH3
CH3
CH3
Table 2
The experimental results by using different solventsa.
NCO
NH2
2COF2
NHCOF
T1
2HF
T2
+
Entry TDA
COF2
Solvent
Solvent
Yield of 2 (%)
2HF
(mmol) (mmol)
amount (g)
1
2
3
4
8
8
8
8
32
32
32
32
Toluene
15
15
59.6
90.3
96.2
0
NH2
1
NHCOF
2
NCO
3
Chlorobenzene
Dichloromethane 15
THF 15
Scheme 1. The main reactions of TDA with COF2.
a
Reaction temperature: room temperature; reaction time: 2 h; the yield of
intermediate 2 was determined by 1H NMR.
produced in the first step at a higher temperature. According to
our previous report [17], the whole process mainly took place by
way of Scheme 1.
under the experimental conditions optimized above. The results
are shown in Table 2.
From Table 2, the intermediate 2 could be produced in different
yields except Entry 4 although THF can dissolve TDA. Among the
2.1.1. The reaction between TDA and COF2
Several factors, for example reaction time, may affect the results
of the first step reaction. To check their relevance for the yield of
target intermediate 2, several experiments were conducted and
the results are shown as Table 1.
other three Entries, the yield of 2 was the highest when
dichloromethane was used as the solvent; besides, due to the
lower boiling point of dichloromethane (40 8C) than chloroben-
zene (132 8C), dichloromethane was more easy to be taken away
and could gain purer intermediate. So, dichloromethane was a kind
of preferable solvent than chlorobenzene. For toluene as the
solvent, not only was the yield of 2 unsatisfactory, but also a foam-
liked material was generated with intermediate 2 together, when
we discharged the excess COF2, the solution will be sprayed out.
Moreover, when intermediate 2 was dissolved in acetone a night
later or in DMF for 6 h, the quantity were decreased, of which the
possible causes are still on study.
Actually, the reactions could happen at other temperature, such
as 40 8C, 60 8C, etc. But the room temperature could keep the
energy consumption low and made the application of a solvent
with low boiling point possible, which is considered to be the best
reaction temperature.
From Table 1 – (1), we can know that when n (TDA): n (COF2)
was 1:1, there was no 2 produced. With the increase of COF2, the
yield of 2 was growing up until n (TDA): n (COF2) reached 1:4, at
which the highest yield was obtained. Whereas when n (TDA): n
(COF2) was 1:6, the yield of 2 was decreased, which may be
explained that the excess COF2 addition at the ‘C55O’ double bond
of 2 were happened and generated a by-product with ‘C (F)–O–
COF’ structure [15]; and we do not believe that this decrease was
caused by the reverse reaction of 2 with HF owing to the no
detected TDA in the products. Compare the experimental results in
Table 1 – (2), it is known that with the increase of TDA amount, the
yield of 2 was growing up. Because the mole ratio of TDA and COF2
was fixed, when TDA amount increased, the COF2 amount will also
increase, which means the pressure in the reactor will increase
correspondingly. Thus, the yield cannot be increased indefinitely
just by adding more substrates because higher requirement for the
pressure tolerance of the reactor and the resulting increase of
apparatus cost and dangerousness should be considered; and for
the 10 ml stainless reactor used here, the amount of TDA was
preferred to be 4 mmol. Based on the results of Table 1 – (3), the
preferable reaction time was 2 h. The possible reason about the
yield decrease of 2 after 2 h is considered to be similar to the result
of Table 1 – (1) at 1:6 mole ratio of TDA and COF2.
In sum, the optimized experimental conditions of the first step
are: amount of TDA: 4 mmol; n (TDA): n (COF2) = 1:4; reaction
temperature: room temperature; reaction time: 2 h; solvent:
dichloromethane.
2.1.2. Thermolysis of the carbamyl fluoride intermediate
The investigation of the decomposition process of intermediate
2, which was produced in the first step under the optimized
experimental conditions, was conducted herein under different
decomposition temperature.
Moreover, the solvent is indispensable for the first synthesis
step, thus the investigation about the effects of different
solvents is necessary. The intermediate 2 is soluble in acetone,
ethanol, DMF, DMSO, and THF; and it is insoluble in toluene,
dichloromethane, chlorobenzene, and m-dichlorobenzene.
According to synthesis procedure that descried in Section
4.3.1.1, we selected four different solvents for comparison
The intermediate solid was added into a PTFE tube after
weighting precisely, the tube was sealed by rubber stopper. The
gases that produced during the decompose process was taken
away by N2. The tube was put into oil bath with different
temperature, the decomposed process could be observed clearly
and the phenomena are as Table 3 shown.
Based on Table 3, the intermediate solid began to crack
obviously at the temperature of 160 8C; and about 30 min later,
there was no bubble generated in the tube, a yellowish brown
viscous liquid was gained. Meanwhile, the intermediate 2 was
analyzed by TG-DTA (shown in Section 4.3.1.3 Fig. 2), from which
the loss of weight was mainly happened between 100 8C and
Table 1
The yields of intermediate 2 under different experimental conditions.
a
(1) in different mole ratio of TDA and COF2
n (TDA):n (COF2)
Yield of 2 (%)
1:1
0
1:1.5
6.74
1:2
1:4
1:6
24.4
52.5
37.8
(2) in different TDA amountb
TDA amount (mmol)
Yield of 2 (%)
Table 3
The decomposition phenomena of intermediate solid under different temperature.
2
4
6
48.4
52.5
83.4
Temperature
Phenomenon
(3) in different reaction timec
Reaction time (hour)
1
2
4
12
120 8C
140 8C
There was no obvious visible change.
Yield of 2 (%)
70.9
94.5
70.9
52.5
The surface of the intermediate was becoming yellow after
2 min.
Reaction temperature: room temperature; Solvent: chlorobenzene, 7.5 g; the yield
160 8C
180 8C
The intermediate liquefied very quickly and some gases were
of intermediate 2 was determined by 1H NMR.
generated.
a
Amount of TDA: 4 mmol; reaction time: 12 h.
The intermediate liquefied very quickly and some gases were
generated.
b
n (TDA): n (COF2) =1:4; reaction time: 12 h.
c
Amount of TDA: 4 mmol; n (TDA): n (COF2) =1:4.

210
N. Zhang et al. / Journal of Fluorine Chemistry 178 (2015) 208–213
CH3
NHCOF
+
CH3
CH3
CH3
COF
CO NH
NCO
N
Δ
NHCOF
2
NCO
3
NCO
NHCOF
5
Scheme 3. The possible side reaction of the second step.
Scheme 2. The main reactions of the decomposition of intermediate 2.
process. Thus, 180 8C is considered to be the optimized
180 8C and it was an obvious heat absorbing process. The observed
phenomena and the TG-DTA result were consistent.
decomposition temperature.
Moreover, the decomposed products of the intermediate 2
under the temperature of 160–180 8C and lower than 160 8C were
analyzed by NMR. When the temperature was lower than 160 8C,
an intermediate 4, that only one HF was removed from the
intermediate 2 and the other carbamyl fluoride group was not
decomposed, was detected. When the temperature was 160–
180 8C, the decomposition was complete and the main product was
compound 3 (TDI). Hence, the second step reaction of Scheme 1
could be corrected to take place as Scheme 2.
The final product 3 was analyzed by FT-IR and the spectrum is
shown in Fig. 1. The IR peaks of 3 synthesized in this work are
almost the same as those of the authentic TDI sample (Wako, Osaka
Japan), except two peaks of 1779 cmÀ1 and 1815 cmÀ1. These two
peaks are close to the peaks of C55O in –NHCOF group of
intermediate 2, but it is not believed to be the undecomposed 2
based on the measurements of NMR. It may be explained that in
the process of dehydrofluoride, some side reactions occurred and
the products have the similar C55O structure of intermediate 2. Ma
and Tan [18] reported the side reactions in phosgenation process of
TDI. They believed that N–H group, in raw materials or in
intermediates, can react with TDI. Besides, TDI can polymerize by
themselves on heat. In our experiment, the similar reactions are
likely to take place, shown in Scheme 3.
The final product TDI was reported to be obtained in 93% yield
based on the corresponding intermediate 2 preciously [12], and in
this work it was raised to be 95.9% when the decomposition
temperature was set as 180 8C. This may be explained that the
higher temperature increased the decomposition speed of 2 and
thus decreased the formation possibility of impurity 5 or
structure-liked by-products. However, it is not suggested just
to use a higher temperature to raise the yield because surface
sintering possibility of the intermediate may be increased and the
higher energy consumption will reduce the industrial value of this
2.2. Synthesis of TDI via one-pot method
We believed that the excess COF2 need to be removed
completely before the decomposition process [12] based on the
study of Fawcett et al. [15], in which the mechanism of reaction for
COF2 with C55N double bounds was reported. In general, the
reaction appears to occur via reversible formation of a nucleophilic
anion by addition of a fluoride ion at the carbon atom of isocyanate.
So, some unwanted by-products are likely to be generated if the
excess COF2 of the first step is left in the reactor during the second
step. In this paper, we studied the reaction process when the excess
COF2 existed in the reaction system by synthesizing the TDI via a
one-pot method. The products were separated to be a white
powder and a colorless solution by filtering.
The colorless solution was analyzed by GC–MS firstly. Checked
against the standard sample, the peak
z = 174) was identified to be TDI 3. The peak (retention = 6.1 min
m/z = 148) maybe 4-aminotoluene-2-isocyanate 6. Peak and
peak , with almost the same fragment ion peaks, may be isomers
and the structures may be N,N-bis(fluoroformyl)-toluene isocya-
nates (7 and 8), products of adding one molecule of COF2 to one of
(retention = 6.3 min m/
C–N double bonds of TDI 3. Peak
maybe 2,4-di(N,N-bis(fluor-
oformyl))-toluene 9, a product of adding two molecule of COF2 to
one molecule of TDI 3. Except the reactions shown as Schemes 1
and 2, there were other possible reactions may take place as
Schemes 4 and 5.
Compound 6 was considered to be produced as Scheme 4 rather
than by the decomposition of an intermediate with only one
carbamyl fluoride group. Firstly, this kind of intermediate was not
detected either in the one-pot method or in the two-step method
when the COF2 was over amount to TDA. Secondly, in the one-port
method, the reactor was closed, in which amounts of by-product
HF existed from the beginning of the decomposition process, and
this may lead to the addition of HF to –NHCOF group.
Then, the colorless solution was analyzed by GC and the results
are shown in Table 4.
CH3
CH3
NCO
NCO
HF
+
+
COF2
NHCOF
NH2
6
4
Scheme 4. The possible reaction of intermediate 4 with HF.
Fig. 1. IR spectrum of authentic TDI purchased from Wako and synthesized TDI 3 in
this work.
Scheme 5. The possible reactions of TDI 3 with COF2.

N. Zhang et al. / Journal of Fluorine Chemistry 178 (2015) 208–213
211
Table 4
raised quickly by more work, nevertheless, subsequent research
should be done on reaction mechanism and industrial chemical
process.
The GC analysis results of the colorless solution.
No.
GC area%
TDA
6
3
7
8
9
4. Experimental
1a
2b
—
—
Trace
Trace
31.8
29.2
55.9
65.8
12.3
5.0
Trace
Trace
4.1. Chemicals
TDA: 8 mmol; solvent: chlorobenzene (15 g); COF2: 76 mmol.
a
80 8C for 4 h, then 150 8C for 4 h.
b
2,4-Diaminotoluene (TDA) 99.5+% was purchased from Tokyo
Chemical Industry Co., Ltd. (Tokyo, Japan). Toluene-2,4-diisocya-
nate (TDI) 95.0+%, dehydrated dichloromethane 99.0+%, chloro-
benzene 99.5+%, super dehydrated acetone 99.5+%, super
dehydrated ethanol 99.5+%, N,N-dimethyl formamide (DMF)
99.5+%, m-dichlorobenzene 98.0+%, dimethyl sulfoxide (DMSO)
99.5+%, tetrahydrofuran (THF) 99.5+%, toluene 99.8+%, dimethyl
sulfoxide-d6 (DMSO-d6) 99.9+% (0.05 v/v% tetramethylsilane),
tetramethyl silane (TMS) 99.9+% were purchased from Wako Pure
Chemical Industries, Ltd. (Osaka Japan). CCl3F 99.0+% was
purchased from SynQuest Labs, Inc. (USA). COF2 99.0+% was
prepared in our laboratory [16]. N2 (399.999%) was supplied by
Taiyo Nippon Sanso Corporation (Japan).
80 8C for 4 h, then 105 8C for 4 h.
From the Entries 1 and 2, TDA was not found in the GC curves
indicated that amine was almost conversed. The primary products
of one-pot method were 3, 7 and 8. When the dehydrofluoride
reaction, shown in Scheme 2, was conducted at the temperature
150 8C for 4 h, as Entry 1, GC peak compositions of 3, 7 and 8 were
31.8%, 55.9% and 12.3% respectively; but for Entry 2, reacted at the
temperature of 105 8C for 4 h, the according compositions were
29.2%, 65.8% and 5.0% respectively. With the dehydrofluoride
reaction temperature raised, the content of TDI increased slightly
and that of 8 increased distinctly. Like showed above, the
dehydrofluoride reaction of intermediate 2 was a heat absorbing
process, so when the reaction temperature was raised, more 2 will
converse to TDI. Meanwhile, COF2 will add to C55N double bonds to
give N,N-bis(fluoroformyl)-toluene isocyanate (7 and 8) and 2,4-
di(N,N-bis(fluoroformyl))-toluene (9). Owing to the steric effect of
bulky substituent –N(COF)2, the adding reaction prefers to
occur on the para position than the adjacent position. The rate
of 4-N,N-bis-(fluoroformyl)- toluene-2-isocyanate formation will
thus be higher than that of 2-N,N-bis-(fluoroformyl)-toluene-4-
isocyanate. When the reaction temperature was raised, the steric
effect will declined, and the content of 2-N,N-bis-(fluoroformyl)-
toluene-4-isocyanate will rising obviously. So we got a rudimen-
tal judgment that 7 was 4-N,N-bis-(fluoroformyl)- amino-tolu-
ene-2-isocyanate and 8 was 2-N,N-bis-(fluoroformyl)-toluene-4-
isocyanate.
The white powder was identified to be intermediate 2 by NMR,
and no 2 was detected in the colorless solution, which may
because of the insolubility of 2 in chlorobenzene and incomplete
dehydrofluorination at the temperature lower than 160 8C.
Besides, unlike the two-step method, there was no intermediate
4 detected in the colorless solution, this may because the
extension of the reaction time compensated the insufficient
energy for the dehydrofluorination of the second HF when the
reaction temperature was not enough. However, if we increased
the reaction temperature to promote decomposition of the solid
material, the unwanted reaction between TDI and COF2 will be
strengthened together and the reaction pressure will increase
dramatically.
4.2. Instruments and apparatus
1H NMR and 19F NMR spectra of the substrate, intermediate and
product during the synthesis were recorded on a JNM-EX300 (JEOL,
300 MHz) NMR with TMS and CCl3F as internal standards in DMSO-
d6 solvent, respectively.
GC analysis was conducted on a Shimadzu GC-2014A with a
thermal conductivity detector (TCD) and a wide-bore capillary
column (DB-5; length: 60 m; i.d.: 0.25 mm; J&W Scientific Inc.).
The operation condition of GC was as follows: column temperature
80 8C for 2 min, heated to 250 8C at the rate of 80 8C/min, and held
for 10 min; detector temperature 280 8C; injection port tempera-
ture 300 8C; carrier gas, ꢀ3.4 cm3 He/min.
MS (EI, 70 eV) spectrum was measured using the Shimadzu
GCMS-QP2010 system equipped with GC-2010. The column type
and the parameters were the same as the GC conditions mentioned
above.
TG and DTA patterns were recorded on a Rigaku Thermoplus
DSC-8230 and a Thermoplus 8120 equipped with an IBM PS/V
Model 2411 (Master 100) computer. The flow rate of N2 was 25 ml/
min. The range of temperature was from room temperature to
200 8C. The reference material was Al2O3 and the weight of sample
is 10 mg.
FT-IR spectra were obtained on a JASCO FT/IR-620 spectrometer
with a TGS detector. Spectra were measured on a KBr pellet at
4 cmÀ1 resolution after accumulation of 16 scans.
The inert glove box was used to offer a <1 ppm oxygen and
moisture inert atmosphere. The samples of TDA, the intermediate,
and the final product were prepared in this system.
The vacuum line was used to add the COF2 into the reactor
precisely. The air in the reactor will be push out by vacuum line and
the reactant and dissolve solution have no touch with air and
water.
3. Conclusions
In this paper, the synthesis processes of TDI via the TDA and
COF2 by the two-step method were studied more detailed.
Different reaction conditions, including reaction time, reaction
temperature, and mole ratio of substrates, etc., could affect the
experiment results to different extent. Six main factors were
optimized and the total yield of TDI based on TDA was raised to be
92.3%. Owing to the high reactivity of COF2 with C55N double
bounds, some by-products containing fluorine could be generated
and affect the formation of TDI. The experiment was conducted via
a one-pot method, based on which the necessity of removing COF2
completely after the first step reaction was verified more
effectively.
4.3. Reaction processes
4.3.1. Two-step method
4.3.1.1. The reaction between TDA and COF2. TDA (4 mmol) and
chlorobenzene (7.5 g) as a solvent were added into a stainless
reactor in the inert glove box. The COF2 (16 mmol) was introduced
into the reactor at a constant flow rate precisely by the vacuum line
after freezing the reactor by liquid nitrogen. The mixture reacted at
room temperature for 4 h with magnetic stirring. Then, COF2 and
To sum up, the two-step method by reacting TDA with COF2
is efficient for the synthesis of TDI and the yield could be

212
N. Zhang et al. / Journal of Fluorine Chemistry 178 (2015) 208–213
Fig. 2. TG-DTA spectrum of intermediate 2.
HF left in the reactor were taken away by nitrogen gas. After that,
the reactor was heated to 60–80 8C by oil bath and the solvent was
be vaporized and removed. A white solid product will be obtained
and analyzed by NMR. Reactions under different experimental
conditions were carried out.
After that, the residuary of COF2 in the reactor was taken away by
nitrogen. A milky white suspension was collected. After filtering, a
white powder and a colorless solution were obtained. The colorless
solution was analyzed by GC–MS and GC. The white powder was
analyzed by NMR.
For the GC–MS spectrum, besides the peaks of solvent
chlorobenzene, there were five peaks. In the order of retention
4.3.1.2. Thermolysis of the carbamyl fluoride intermediate. The
white solid product obtained in the first step was heated at the
temperature of 160–180 8C and lower than 160 8C for 30 min with
a flow of N2 (10 ml/min), respectively. The product was analyzed
by NMR and FT-IR.
time, they were marked with
,
,
,
, and respectively. The
data of GC–MS are listed as follows: , retention time: 6.1 min,
m/z: 148, (+)C7H6(NH2)N55C55O; 104, 76, 50, 38; : retention time:
6.3 min, m/z: 174, (+)C7H6(N55C55O)2, 145, 132, 118, 103, 91, 76, 64,
51, 39, 27;
:
retention time: 6.7 min, m/z: 240,
4.3.1.3. Analytical results.
(+)C7H6(N55C55O)N(COF)2, 200, 192, 173, 164, 132, 90, 76, 47;
:
retention time: 7.1 min, m/z: 240, (+)C7H6(N55C55O)N(COF)2, 211,
192, 173, 165, 132, 90, 77, 47; : retention time: 7.2 min, m/z: 306,
(+)C7H6[N(COF)2]2, 266, 258, 239, 211, 191, 173, 150, 95, 47.
(1) NMR data of TDA 1: 1H NMR (300 MHz, DMSO-d6, TMS):
(s, 3H), 3.8 (bs, 4H), 5.9–6.0 (m, 2H), 6.7 (d, 1H).
(2) NMR data of the intermediate 2: 1H NMR (300 MHz, DMSO-d6,
d 2.0
d
d
d
Acknowledgements
TMS):
d
2.2 (s, 3H),
d
7.6–7.7 (m, 3H),
d
10.1 (br, 1H),
d
10.7 (br,
1H); 19F NMR (300 MHz, DMSO-d6, CCl3F):
d
À6.7 (s, 1F),
d
The authors thank Dr. Guangcheng Yang, Dr. Xiaoqing Jia, Dr.
Masanori Tamura, and Dr. Mizukado Junji for their valuable
comments and advices.
À10.8 (s, 1F).
(3) NMR data of the intermediate 4: 1H NMR (300 MHz, DMSO-d6,
TMS):
(300 MHz, DMSO-d6, CCl3F):
(4) NMR data of the final product TDI 3: 1H NMR (300 MHz, DMSO-
d6, TMS) data: 2.2 (s, 3H); 7.6–7.7 (m, 3H).
d
2.2 (s, 3H),
d
6.9–7.2 (m, 3H),
d
10.7 (br, 1H); 19F NMR
d
À6.2 (s, 1F).
References
d
d
(5) TG-DTA spectrum of intermediate 2:
[1] R.H. Richter, R.D. Priester Jr., in: J.I. Kroschwitz, M. Howe-Grant (Eds.), 4th
edn., Kirk-Othmer Encyclopedia of Chemical Technology, vol. 14, Wiley, New
York, 1998, pp. 446–461.
[2] K. Biskup, R. Bruns, W. Lorenz, F. Pohl, F. Steffens, V. Michele, Process for the
Preparation of Isocyanates in the Gas Phase, U.S. Patent 2010/0152484 A1
(June, 2010).
[3] A. Woelfert, C. Mueller, E. Stroefer, A. Brodhagen, Method for Producing
Isocyanates, U.S. Patent 7112694 B2 (September, 2006).
[4] P.D. Hammond, J.A. Scott, Preparation of Aromtic Isocyanates in a Tubular
Reactor, U.S. Patent 3812169 (May, 1974).
4.3.2. One-pot method
TDA (8 mmol) and chlorobenzene (15 g) were added into a
25 ml stainless reactor. Then, the COF2 (76 mmol) was introduced
into the reactor precisely by the vacuum line after freezing the
reactor by liquid nitrogen. With magnetic stirring, the reactor was
heated at 80 8C for 4 h, and then beyond 100 8C for 4 h by oil bath.
[5] S. Ozaki, Chem. Rev. 72 (1972) 457–496.

N. Zhang et al. / Journal of Fluorine Chemistry 178 (2015) 208–213
[6] Y.J. Wang, X.Q. Zhao, F. Li, S.F. Wang, J.Y. Zhang, J. Chem. Technol. Biotechnol. [12] H.D. Quan, N. Zhang, X.M. Zhou, H. Qian, A. Sekiya, J. Fluorine Chem. 176
213
76 (2001) 857–861.
(2015) 26–30.
[7] K. Murayama, Y. shimokawatoko, T. Kanno, K. Takamatsu et al., Method for
Producing Toluenedicarbamate, Method for Producing Roluenediiocyanate, and
Toluenedicarbamate, U.S. Patent 2013/0109881A1 (May, 2013).
[8] M. Shinohata, N. Miyake, Process for Producing Isocyanates Using Diaryl
Carbonate, U.S. Patent 8895774 B2 (November, 2014).
[9] F.L. Serrano, B.A. Munoz, A.P. Polo, et al., One-Step Catalytic Process for the
Synthesis of Isocyanates, U.S. Patent 7423171 B2 (September, 2008).
[10] R. Ugo, R. Psaro, M. Pizzotti, et al. J. Organomet. Chem. 417 (1991)
211–233.
[13] H. Ohno, T. Ohi, H. Yokoo, Process for Producing Carbonyl Difluoride, U.S.
Patent 8664438 B2 (March 2014).
[14] Y. Mitsui, Methods for Producing Carbonyl Fluoride and Apparatus for
Production, U.S. Patent 7371898 B2 (May, 2008).
[15] F.S. Fawcett, C.W. Tullock, D.D. Coffman, J. Am. Chem. Soc. 84 (1962) 4275–4285.
[16] H.D. Quan, M. Tamura, A. Sekiya, Process for Production of Carbonyl Fluoride,
W.O. Patent 2007/037468 (April, 2007).
[17] H.D. Quan, M. Tamura, A. Sekiya, Method for Producing Isocyanate Compound,
J.P. Patent 5322183 (October, 2013).
[11] V.I. Manov-Yuvenskii, B.A. Redoshkin, B.K. Nefedov, G.P. Belyaeva, Bull. Acad.
Sci. USSR 29 (1980) 117–120, Division of chemical science.
[18] W.C. Ma, X.S. Tan, Chem. Ind. Times 18 (2004) 14–17 (in Chinese).