CHEMICAL ENGINEERING TRANSACTIONS

VOL. 81, 2020

A publication of

The Italian Association
of Chemical Engineering
Online at www.cetjournal.it

Guest Editors: Petar S. Varbanov, Qiuwang Wang, Min Zeng, Panos Seferlis, Ting Ma, Jiří J. Klemeš

ISBN 978-88-95608-79-2; ISSN 2283-9216

Selective Adsorption for Toluenediamine Recovery from TDA

Tar through Surface Modification of Activated Carbon

Dongliang Wang*, Wei Wang, Wenliang Meng, Guixian Li, Yong Yang

School of Petrochemical Engineering, Lanzhou University of Technology, Gansu Lanzhou 730050, China

wangdl@lut.edu.cn

Toluenediamine (TDA) tar, as the heavy fraction from the dinitrotoluene hydrogenation to produce the important

product of TDA, contains significant amounts of meta-TDA of 10 - 30 wt.% and is currently incinerated as waste

at great cost. In this paper, the adsorption separation technology was adopt for the recovery of meta-TDA from

TDA tar with some porous adsorbent materials. The activated carbon has the highest adsorption ratio for ethanol

solution of meta-TDA. After treatment by HNO3 or CH3COOH, the modified activated carbon has the slightly

decreased adsorption capacity for TDA tar of 238.8 mg·g-1 or 243.3 mg·g-1 with the meta-TDA selectivity of

96.36 % or 98.15 %. While the activated carbon modified by NaOH or NH3·H2O leads to the higher adsorption

capacity of 287.9 mg·g-1 or 261.2 mg·g-1 for TDA tar with relative lower meta-TDA selectivity of 90.10 % or 96.02

%. The FT-IR and BET results confirm that surface oxygen-containing groups can be created through the acid

or alkali modification, which result in enhancing the adsorption selectivity of meta-TDA. The proposed selective

adsorption based on modified activated carbon would be an efficient and low-cost separation technology for

TDA recovery from TDA tar.

1. Introduction

Toluenediamine (TDA) is the important intermediate for the production of toluene diisocyanate (TDI) which is

the raw materials for the preparation of polyurethane. During the hydrogenation of dinitrotoluene (DNT) to obtain

TDA, the low-boiling constituents are removed overhead in the rectification column, the bottoms comprise 2,4-

TDA, 2,6-TDA and oligomer mixture, called as TDA tar. TDA tar contains significant amounts of meta-TDA (10

- 30 wt.% of 2,4-TDA and 2,6-TDA) in order to keep it pumpable or flowable. TDA tar is classified as hazardous

waste due to its toxicity. The disposal or incineration of TDA tar leads to a waste of the desired TDA and is not

environmental friendly.

Several technologies have been developed to recover meta-TDA from TDA tar. BASF (China) Company Ltd.

Shanghai Branch proposed the thin-film evaporator process with a rotary scraper film (Yu et al., 2012). Beckhaus

et al. (1998) raised a rectification process with ortho-TDA as the entrainer to reduce meta-TDA content in TDA

tar. After the treatment, meta-TDA content is generally below 10 wt.%. Brady et al. (2002) added a stripping

column in the Beckhaus process, and at least 97 wt.% up to 99.8 wt.% of meta-TDA presenting in the original

mixture could be recovered. Monobasic alcohol such as cyclohexanol, dibasic alcohol, or ethylene glycol were

also used as entrainer (Lai et al., 2016). Oliveira et al. (2019) invented the hydrogenation process of TDA tar

with a heterogenous catalyst comprising at least one metal with Ni, Co, Ru, Pd, Pt on the support of carbon,

TiO2 or ZrO2. The selective adsorption is an alternative method for TDA recovery due to its mild conditions. The

13X molecular sieve was used as adsorption material to recycle meta-TDA from TDA tar, and the yield of meta-

TDA can achieve 70.56 wt.% (Li et al., 2009). Su et al. (2013) used resin adsorbent to separate meta-TDA from

TDA tar.The choice of these adsorption materials is derived from practical experience in industrial production,

and there is no clear theoretical basis. Activated carbon is attractive adsorbent due to its wide availability, high

thermal stability, low cost, and low sensitivity to moisture (Rashidi et al., 2013). Activated carbon was used to

remove astaxanthin from microalgae (Casella et al., 2020). In this study, the activated carbon with abundant

surface oxygen-containing groups was found to be beneficial for the recovery of meta-TDA. The acid

DOI: 10.3303/CET2081202

Paper Received: 03/05/2020; Revised: 02/07/2020; Accepted: 03/07/2020
Please cite this article as: Wang D., Wang W., Meng W., Li G., Yang Y., 2020, Selective Adsorption for Toluenediamine Recovery from TDA
Tar through Surface Modification of Activated Carbon, Chemical Engineering Transactions, 81, 1207-1212 DOI:10.3303/CET2081202

1207

modification and alkali modification were adopted to create oxygen-containing groups, and the surface of the

adsorbents were analyzed to reveal their influence on the adsorption capacity and selectivity.

2. Experiment method

The effect of five adsorbents on the absorption performance for meta-TDA were firstly investigated. Because

the adsorption performance was better than other adsorbents, the activated carbon was modified with acid or

alkali to test the adsorption for meta-TDA in TDA tar. The adsorption ratio was measured using a UV/VIS

spectrophotometer at the maximum absorption wavelength.

2.1 Selection of adsorbent

Activated carbon, diatomite, ZSM-5 molecular sieve and permutite were chosen as adsorbents to study the

adsorption performance for the meta-TDA.

• Four groups of 100 mL ethanol solution wtih 0.1 mol/L meta-TDA were placed in three-necked flask (250

mL).

• Then, 1 g adsorbent was added into each solution, keeping stirring conditions for some time.

• Samples were centrifuged at 4,000 rpm and the concentrations of the supernatants were determined by

UV/VIS spectrophotometer to analyze the adsorption performance.

2.2 Acid or alkali modification of selected adsorbent

Nitric acid, acetic acid, NaOH and ammonia were used to modify the activated carbon to create oxygen-

containing groups.

• Activated carbon was grinded and sifted to powders of 100-mesh. Four equal parts of the powders were

placed in four conical flasks.

• 100 mL solution of 0.1 mol·L-1 nitric acid, acetic acid, NaOH or ammonia water were added into each conical

flasks under the ultrasonic treatment for 12 h at 25 °C.

• The modified powders were recovered through filtration. Each filter cake was washed with distilled water

until the pH of filtrate was 7, and the filter cake was dried for 24 h at 110 °C.

2.3 Adsorption performance of modified adsorbent

Adsorption performance of the different modified activated carbons were studied.

• Four groups of TDA tar (1 g) were dissolved in 20 mL anhydrous ethanol in four three-necked flasks.

• 1 g modified activated carbon was added into each three-necked flask under stirring condition for some

time.

• Four samples were filtered and dried for 24 h at 110 °C.Unit adsorption amount was measured for each

sample.

• The dried activated carbons were desorbed in 10 mL anhydrous ethanol at 50 °C under stirring for 3 h.

• The GC internal standard method was used to determine the composition of desorption ethanol solution, in

which the internal standard substance is p-nitrotoluene.

2.4 Calculation of adsorption ratio and quantity

The solution concentration was measured using a UV/VIS spectrophotometer at the maximum absorption

wavelength (meta-TDA at 212 nm). The adsorption ratio and meta-TDA mass (mm-TDA) were calculated

according to Eq(1) and Eq(2).

Adsorption ratio %=(1 - Ct/C0)×100 % (1)

mm-TDA = Cm-TDA × V (2)

where C0 and Ct were the initial and final concentration of meta-TDA in ethanol solution. Cm-TDA was the mass

concentration of meta-TDA, mg·L-1. V was the volume, L.

Meta-TDA content of the desorption ethanol solution could be determined by the GC internal standard method.

3. Results and discussion

The Influence of surface chemistry for the four adsorbents and the enhanced adsorption of the

modified activated carbon were discussed in detail.

1208

20 40 60 80 100 120
0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

A
d
so

rp
ti

o
n
r

a
ti

time (min)

Activated carbon

Diatomite

ZSM-5

Permutite

Figure 1: Effect of different adsorbents on the adsorption of meta-TDA

0 20 40 60 80 100 120
0.00

0.02

0.04

0.06

0.08

0.10

0.12

d
V

/d
lo

g
(D

)
P

o
re

V
o
lu

m
e

(
c
m

3
/g

·A
)

Pore Diameter (nm)
0 20 40 60 80 100 120

0.00

0.02

0.04

0.06

0.08

0.10

0.12

d
V

/d
lo

g
(D

)
P

o
re

V
o

lu
m

e
(

c
m

3
/g

·A
)

Pore Diameter (nm)

0 20 40 60 80 100
0.00

0.02

0.04

0.06

0.08

0.10

0.12

d
V

/d
lo

g
(D

)
P

o
re

V
o
lu

m
e

(
c
m

3
/g

·A
)

Pore Diameter (nm)
0 20 40 60 80 100 120

0.00

0.02

0.04

0.06

0.08

0.10

0.12

d
V

/d
lo

g
(D

)
P

o
re

V
o

lu
m

e
(

c
m

3
/g

·A
)

Pore Diameter (nm)

Figure 2: Characterization of pore size distribution of (A) activated carbon; (B) diatomite; (C) ZSM-5; (D)

permutite (inset: N2 physical adsorption / desorption curve)

The N2 adsorption/desorption isotherms of adsorbents are shown in Figure 2. The pore size distribution chart

shows that there are two distinct peaks at 4 nm and 7.7 nm for activated carbon in Figure 2(A), indicating that

the pore size is mainly centered at 4 nm with mesopores of a diameter about 2 - 10 nm. The diatomite has a

strong peak at 4.19 nm in Figure 2(B) because of the collapse for pore structure after high-temperature

calcination to form amorphous silica. The pore size of ZSM-5 is generally below 2 nm, so there is no obvious

A
B

C
A

1209

3.1 Influence of adsorbent surface

The adsorption performances of meta-TDA for activated carbon, diatomite, ZSM-5 molecular sieve and
permutite are showed in Figure 1. It can be seen that activated carbon has the best adsorption effect on meta-
TDA. With the increase of adsorption time, the adsorption ratio of activated carbon also increased more
significantly than that of the other three adsorbents.

strong peak in the range of mesopores and macropores in Figure 2(C). Permutite has a strong peak at 42.92

nm with a distribution ranging from 10 nm to 60 nm in Figure 2(D). The surface area and pore structure of

different adsorbents were summarized in Table 1. The specific surface area of activated carbon is relatively

small, but the pore size of activated carbon is suitable for the adsorption of meta-TDA molecules. Appropriate

pore size is beneficial for the adsorption of meta-TDA.

Table 1: Surface area and pore structure of different adsorbents

No. Adsorbents Surface area (m2·g-1) Pore volume (cm3·g-1) Pore size (nm)

A Activated carbon 21.5 0.052 8.9

B Celite 522.4 0.044 5.4

C ZSM-5 296.8 0.043 1.9

D Permutite 2.9 0.041 45.0

The FT-IR spectrums of diatomite, ZSM-5 and permutite mainly show the vibration peak of the skeleton of the

materials themselves in Figure 3. The difference between activated carbon and other silica-alumina-based

framework adsorption materials is the surface acid group. Although these functional group peaks are weak for

activated carbon, they have an important influence on the adsorption of meta-TDA.

4,000 3,500 3,000 2,500 2,000 1,500 1,000 500

wavelength(cm
-1

)

Activated carbon

Celite

ZSM-5

Permutite

1,643

790

563

3,617

1,029

1,064

1,620

Figure 3: FT-IR spectrums of different adsorbents

3.2 Influence of activated carbon modification

After the acid or alkali modification, the adsorption effect of activated carbon can be seen in Table 2.

Table 2: Adsorption capacity of modified activated carbon

Modification method
Adsorption capacity

for TDA tar (mg·g-1)

Adsorption capacity

for meta-TDA (mg·g-1)

Adsorption selectivity

for meta-TDA (%)

Unmodified 252.9 153.4 60.66

HNO3 238.8 230.1 96.36

CH3COOH 243.3 236.6 98.15

NaOH 287.9 259.4 90.10

NH3·H2O 261.2 250.8 96.02

For unmodified activated carbon, the adsorption capacity per unit mass adsorbent for TDA tar is 252.9 mg·g-1,

for meta-TDA is 153.4 mg·g-1, and the adsorption selectivity is 60.66 %. After acid modification by HNO3 or

CH3COOH, the unit adsorption capacity of activated carbon for TDA tar decreases slightly to 238.8 mg·g-1 or

243.3 mg·g-1, while the adsorption capacity for meta-TDA has improved to 230.1 mg·g-1 or 236.6 mg·g-1

regarding the special selectivity obviously up to 96.36 % or 98.15 %. After NaOH or NH3·H2O modification, the

unit adsorption capacity increases both for both TDA tar and meta-TDA with adsorption selectivity of 90.10 %

1210

or 96.02 %. The weak acid modification is more conducive to improve the adsorption selectivity of activated

carbon for meta-TDA.

Figure 4 shows the FT-IR spectrums of different modified activated carbons that represent the characteristics

of the functional groups on the surface of the activated carbon. It can be seen that the unmodified activated

carbon has few weak peaks. The activated carbon modified with HNO3 is similar to the surface characteristic

peak of unmodified activated carbon in Figure 4(B). The stretching vibration of C-O bond at 1,026.0 cm-1 can

been seen after the CH3COOH, NaOH or NH3·H2O modification. The activated carbon modified with CH3COOH

has a strengthened stretching vibration of C-O at 1,322.8 cm-1 and the absorption peak of carbonyl group at

1,622.1 cm-1, which indicates that the carbonyl group number of oxygen-containing groups has increased

significantly. The surface oxygen-containing groups can be created through the acid or alkali modification, which

may result in enhancing the adsorption selectivity of meta-TDA.

4,000 3,500 3,000 2,500 2,000 1,500 1,000 500

wavelength(cm
-1

)

A. Unmodified

B. HNO
3

C. CH
3
COOH

D. NaOH

E. NH
3
·H

2
O

1,322.8

1,623.83,621.7

1,026.0

775.3

1,622.1

Figure 4: FT- IR spectrum of different modified activated carbons

3.3 Analysis of the desorption product

Activated carbon containing adsorbed TDA tar was desorbed in 10 mL anhydrous ethanol at 50 °C under stirring

for 3.0 h. The desorption liquid was analyzed with a Fourier infrared analyzer. The FT-IR spectrum of the

desorption product was compared with the meta-TDA of 2,4-TDA in Figure 5. It can be seen that stretching

vibrations of N-H at 3,422.0 cm-1 and 3,351.7 cm-1 represent the primary amine groups in the mixture. These

characteristic peaks indicate that the desorption product consists mainly of 2,4-TDA.

4,000 3,500 3,000 2,500 2,000 1,500 1,000 500

wavelength(cm
-1

)

3,430.8

3,353.6

2,927.5

2,852.2

2,4-TDA

Desorption product

Figure 5: FT-IR spectrum of the desorption product

1211

0 2 4 6 8 10 12 14
0

retention time (min)

re
la

ti
v
e

r
e

s
p

o
n

s
e

ethanol

p-nitrotoluene

TDA

Figure 6: The desorption product gas chromatogram using internal standard method

4. Conclusions

The proposed selective adsorption has demonstrated excellent adsorption performance for TDA recovery from

TDA tar. Due to the suitable pore size, highest pore volume and typical surface chemistry, the activated carbon

has the highest adsorption ratio. The surface oxygen-containing groups of activated carbon can be created

through the acid or alkali modification, which may result in enhancing the the adsorption selectivity of meta-

TDA. After acid modification by HNO3 or CH3COOH, the unit adsorption capacity of activated carbon for TDA

tar remains at 238.8 mg·g-1 or 243.3 mg·g-1 corresponding to the selectivity for meta-TDA of 96.36 % or 98.15

%. While NaOH or NH3·H2O modification leads to the higher adsorption capacity of 287.9 mg·g-1 or 261.2 mg·g-

1 for TDA tar with relative lower selectivity for meta-TDA of 90.10 % or 96.02 %. Adsorption separation

technology based on modified activated carbon would be an efficient and low-cost method for TDA recovery

from TDA tar. Besides meta-TDA, the identification and recovery of other components in TDA tar should be

focused in furure, and the adsorption equipment should also be investigated for the development and application

of this TDA recovery technology.

Acknowledgements

It is gratefully for the financial supports from the Major Projects for Science and Technology of Gansu Province

(No. 18ZD2GD014).

References

Beckhaus H., Witt B., Zarnack U., Greger, G.,1998, Process for the separation of high-boiling materials from

the reaction mixture generated during the production of diaminotoluene, United States Patent, US 5728880.

Brady B.L, Weymans G., Keggenhoff B.,2002, Process for separating mixtures of materials having different

boiling points, United States Patent, US 6359177B1.

Casella P., Musmarra D., Dimatteo S., Chianese S., Karatza D., Mehariya S., Molino A., 2020, Purification of

astaxanthin from microalgae by using commercial activated carbon, Chemical Engineering Transactions,

79, 295-300.

Lai J., Jiang N., Li B., Xiang Y., Feng G., Du C., Pu Z., 2014, Methods for recycling the toluene diamine from

The dinitrotoluene product tar. China Patent, CN 201410328429.3.

Li G., Li X., Zhang Y., Teng Z., Mao L., Fan L., 2009, The method for the mate-toluenediamine recovery from

dinitrotoluene hydrogenation tar. China Patent, CN 200910250473.6.

Oliveira L. A., Lottenburger I., Bechold C., Heidemann T., 2019, Process for hydrogenating toluenediamine

(TDA) tar, United States Patent, US 20190241501A1.

Rashidi N.A., Yusup S., Lam H.L., 2013, Kinetic studies on carbon dioxide capture using activated carbon,

Chemical Engineering Transactions, 35, 361-366.

Su D., Kou J., Meng F., Yang G., Duan L., Zhang S., 2013, Dinitrotoluene hydrogenated tar recycling

system,China Patent, CN 201320221412.9.

Yu Z., Li D., Jia P., David B., 2012, A method for the tar recycling toluene diamine, China Patent,

CN201210576787.7.

1212