Catalysis Communications 43 (2014) 207–212

Contents lists available at ScienceDirect

Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom

Short Communication

Creation of a continuous process for bio-ethanol to butadiene conversion

via the use of a process initiator
G.O. Ezinkwo b,⁎, V.F. Tretjakov a,b, R.M. Talyshinky a, A.M. Ilolov a, T.A. Mutombo b
a
A.V. Topchieva Institute of Petrochemical Synthesis, Moscow, Russia
b
Moscow State University of Fine Chemical Technology, 86 Vernadsky Avenue, Moscow, Russia

a r t i c l e i n f o a b s t r a c t

Article history: One of the major constraints to the efficiency of Lebedev's process for butadiene production is the deposition of
Received 19 September 2013 soot on the active sites of the catalyst which results to the reduction of catalyst activity and the periodic
Received in revised form 10 October 2013 regeneration of catalyst thereby increasing time and energy spent on the process and subsequently, cost. The
Accepted 11 October 2013
use of a process initiator has been found to be a possible solution to the above mentioned challenges. A synergetic
Available online 21 October 2013
effect was observed during the initiation of the catalytic process. Furthermore, the interdependence between the
Keywords:
efficiency of the process initiator and the morphology of Al2O3, linear velocity of the feed stream and other
Butadiene (BD) process parameters was observed.
Catalyst © 2013 Elsevier B.V. All rights reserved.
Ethanol
Aluminium oxide
Initiator

1. Introduction C2 H5 OH þ CH3 CHO→CH2 ¼ CH–CH ¼ CH2 þ 2H2 O:

Butadiene, which is an important building block for the chemical

The mechanism of the ethanol to BD transformation is extremely
industry is produced by (1) pyrolysis of butane–butylene fraction of
complicated and is still a subject of debate. However, one generally
petroleum [1], (2) as a byproduct of ethylene production from steam
accepts the involvement of the following principal steps [17–30]:
crackers [2] and (3) combined process of ethanol dehydrogenation
and dehydration (Fig. 1) [1]. Step 1 Production of acetaldehyde from ethanol under the influence of
Butadiene in the 1930s was produced by a one-step catalytic process the dehydrogenating centers of the catalyst.
of both dehydrogenation and dehydration in the Soviet Union [3–14].
During the Second World War, about 60% of butadiene produced in
C2 H5 OH→CH3 CHO þ H2
the United States was via a two-step process commonly known as
Ostromislensky process [15], in which ethanol is oxidized to acetaldehyde
which then reacts with additional ethanol over a tantala-promoted silica
catalyst to yield butadiene [16] Step 2 Condensation of acetaldehyde to form aldol under the influence
of the dehydrating sites of the catalyst.
One-step process:

T ¼ 430–450 ∘C; ZnO=Al2 O3

2C2 H5 OH→CH2 CH–CH ¼ CH2 þ H2 þ 2H2 O:

Two-step process:
Step 3 Dehydration of aldol to form crotonaldehyde
C2 H5 OH→CH3 CHO þ H2

⁎ Corresponding author at: Tel.:+7 9036283618.

E-mail address: [email protected] (G.O. Ezinkwo).

1566-7367/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.catcom.2013.10.015
208 G.O. Ezinkwo et al. / Catalysis Communications 43 (2014) 207–212

Fig. 1. A brief overview of butadiene production.

Step 4 Meerwein–Ponndorf–Verley reaction between crotonaldehyde i.e. gamma and alpha. Sample A was prepared by mechanical mixing
and ethanol to obtain Crotyl alcohol and acetaldehyde. of zinc oxide with γ-Al2O3, B by impregnation of γ-Al2O3 with mixed
aqueous solution of aluminium and zinc nitrate and C by mechanical
mixing zinc oxide with α-Al2O3.
Impregnation of γ-Al2O3 was carried out using mixed aqueous
solution of aluminium and zinc nitrates at 80 °C for 2 h. The precursor
Step 5 Formation of butadiene from Crotyl alcohol on the dehydrating samples were later dried for 6 h. Further thermal treatment was carried
sites of the catalyst. out in a muffle furnace for 3 h at 300 °C with subsequent expulsion of
nitrogen oxides for an additional 3 h.
The precursor samples were loaded into a glass reactor and calcined
at 450°C for 3h. At this same temperature, activation of catalyst samples
was done in a stream of hydrogen. Further activation was done at 400°C
using bio-ethanol for 6 h at a flow rate of 1 h−1. The drying, calcination
and activation process for all samples A, B and C were the same [31,32].
Soot formation is one of the numerous side reactions that take place
in the process of butadiene formation. It forms on the surface of the
catalyst there by blocking the access to active sites. This process of 2.2. Catalyst testing and initiator introduction
soot formation results to the periodic regeneration of the catalyst in a
stream of hot air as shown below: The catalytic conversion of bio-ethanol to BD was carried out in an
integral quartz reactor with a catalyst loading of 10 cm3, in the presence
C þ O2 →CO2 þ Q and absence of the initiator (hydrogen per oxide) with a flow rate of
1.2–4.4 h−1 at 390–420 °C. Kinetic studies were also carried out using a
2C þ O2 →2CO þ Q micro reactor with a catalyst loading of 1 cm3. Products were analyzed
by gas chromatograph (GC, Crystal-2000M made in Russia by Chromatec).
The experiments were tested for reproducibility several times and the
2CO þ O2 →2CO2 þ Q: data was averaged and results were verified with independent gas-mass
spectrometry equipment, model — Finnigan MAT 95 XL. Besides
butadiene some amounts of methane, propylene, butylenes, butane,
The aim of this study was to achieve a continuous process void of
acetaldehyde, diethyl ether and other oxygenated compounds were
catalyst regeneration and to show the correlation between the
found as by-products. The conversion of ethanol and the theoretical
efficiency of the process initiator and the morphology of Al2O3, linear
yield of butadiene were found to be 42% and 88% respectively with
velocity of the feed stream, temperature and selectivity of the process.
the use of the initiator. Samples prepared using γ-Al2O3 were found to
be more active (concentration of BD in the gaseous phase was between
2. Experimental section 65 and 70%), and stable; not loosing activity for over 120h of testing in a
glass reactor. The duration of work on this sample without the use of an
2.1. Catalyst preparation initiator (hydrogen per oxide) was 48 h with subsequent decrease in
activity. However, testing in the presence of the initiator resulted to
Catalysts were prepared mainly by mechanical mixing and impreg- unabated activity of catalyst sample for 120 h which gives the basis to
nation based on the different morphologies of aluminium oxide believe the possibility of creating a continuous process needless of
 G.O. Ezinkwo et al. / Catalysis Communications 43 (2014) 207–212 209

Fig. 2. The laboratory apparatus for the experiments. 1 — metering pump, 2 — furnace, 3 — reactor, 4 — condenser, 5 — receiver for liquid products, 6 — chromatograph for analysis.

catalyst regeneration [33]. The laboratory scheme for the process is diffuse peaks belong to the γ-Al2O3. As a result of the strong line overlap,
shown in Fig. 2 below. D for gamma-Al2O3 could not be calculated.

2.3. Catalyst characterization

3. Kinetic study of butadiene production process
The phase composition of samples was studied using Advanced D8
powder XRD manufactured by the firm ‘Bruker’ with CuKα radiation, For the purpose of convenience, the reaction routes presented
λ = 0.15406 nm. Measurements were carried out between 2θ values below were used for the calculation of material balance and molar
of 10° and 100° with internal IP PSD resolution of 0.1° and a scan rate concentration.
of 20 s. The result of the phase composition analysis was obtained by
comparison with the values computed in the data base of STOE
WinXPOW. Average length of the coherent scattering was obtained 1. 2C2H5OH = C4H6 + 2H2O + H2 (X1)
using XRD based on the harmonic analysis of diffraction peak profiles. 2. 2C2H5OH = C4H8 + 2H2O (X2)
The average length of coherent scattering was calculated by Selyakov– 3. C2H5OH = C2H4 + H2O (X3)
Scherrer's formula given below: 4. C2H5OH = CH3CHO + H2 (X4)
5. 2C2H5OH = C4H8O + H2O + H2 (X5)
Kλ
D¼ :
β cos θ
Table 1
Phasic scheme of the process of conversion of ethanol catalyst on ZnO/γAl2O3.
From the XRD pattern in Fig. 3. It could be seen that our catalyst
sample, is a mixture of two phases: the main ordered phase of ZnO Steps Stoichiometric number
of steps by routes
(with D = 42 nm) and the related disordered phase of Al2O3. All blurred
I II III IV V VI VII

1* C2H5OH + Z → C2H4Z + H2O 1 0 0 0 0 0 0

Intensity 2 C2H4Z ↔ C2H4 + Z 1 0 0 0 0 −1 0
1200 3* C2H5OH + ZH ↔ CH3CHOZH + H2 0 1 0 0 0 0 0
4 CH3CHOZH ↔ CH3CHO + ZH 0 1 −1 0 −1 0 0
-ZnO 5* CH3CHOZH + C2H4 → C4H6ZH + H2O 0 0 1 0 0 0 0
1000
-γ-Al2O3 6 C4H6ZH ↔ C4H6 + ZH 0 0 1 0 1 0 0
7 C2H4 + ZO → C2H4ZO 0 0 0 1 0 0 0
800 8* CH3CHO + C2H4ZO → C4H8OZO 0 0 0 1 0 0 0
9 C4H8OZO ↔ C4H8O + ZO 0 0 0 1 0 0 0
600 10 CH3CHOZH + CH3CHO → 0 0 0 0 1 0 0
CH3CHOHCH2CHOZH
400 11 CH3CHOHCH2CHOZH → CH3CH = 0 0 0 0 1 0 0
CHCHOZH + H2O
200 2θ 12* CH3CH = CHCHOZH + H2 ↔ 0 0 0 0 1 0 0
(degree.) C4H6 ZH + H2O
13* C2H4Z + C2H4 → C4H8Z 0 0 0 0 0 1 0
0
14 C4H8Z ↔ C4H8 + Z 0 0 0 0 0 1 −1
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
15* C4H8Z ↔ C4H6Z + H2 0 0 0 0 0 0 1
16 C4H6Z ↔ C4H6 + Z 0 0 0 0 0 0 1
Fig. 3. XRD data for the sample ZnO/γ-Al2O3 synthesized in the laboratory; after the
reaction. The asterisk represent various limiting steps.
210 G.O. Ezinkwo et al. / Catalysis Communications 43 (2014) 207–212

where X1, X2, X3, X4, and X5 are the molar percents of the main products Table 2
of the above reactions. Kinetic parameters of reactions.

The main stoichiometric routes of the reaction process considered, in Routes Steps Kinetic parameter by the route с−1
order to study the selectivity of the process are as follows: −2106002100
I 1 k1 ¼ ð4:86  1:2Þ  1016 e RT
−19050190
II 3 k3 ¼ ð2:30  0:7Þ  103 e RT
−13650136

I. C2H5OH → C2H4 + H2O III 5 k5 ¼ ð1:26  0:3Þ  102 e RT

−13070130
IV 8 k8 ¼ ð2:29  0:7Þ  102 e RT
II. C2H5OH → CH3CHO + H2 −689070
V 12 k12 ¼ ð3:76  1:0Þ  103 e RT
III. CH3CHO + C2H4 → C4H6 + H2O VI 13
−10650120
k13 ¼ ð3:06  1:3Þ  102 e RT
IV. CH3CHO + C2H4 → C4H8O VII 15 k15 ¼ ð2:90  1:0Þ  1038 e
−4975004900
RT

V. 2CH3CHO + H2 → C4H6 + 2H2O

VI. 2C2H4 → C4H8
VII. C4H8 → C4H6 + H2. selectivity of the reaction process. During the initiation process ZO
(the oxidized center) gets modified by the influence of hydrogen
peroxide as shown below:
An in depth study of the reaction mechanism which was carried out
•
is presented in Table 1 and from the analysis of the EMF spectra found in H2 O2 →2OH
the Supplementary material, it was deduced that the catalysts have
three active centers as shown below: •
ZO þ OH→Z þ HO2 :

As a result of the above, more Z centers are released with subsequent

increase in the rate of formation of main products and reduction of by-
product formation due to the destruction of ZO centers responsible for
their formation. Furthermore, stage 7 gets blocked as shown in Table 1
and subsequently stage 8 which is responsible for the rate of formation
of butanal by route 4 (Fig. 4).
The above graph shows the different centers and reaction routes that
take place on them. It also shows the influence of hydrogen peroxide on
the ZO center.

4. Results and discussions

It was observed that among the catalysts prepared by the above

Z and ZH are responsible for the formation of main products while described methods, those prepared by impregnation with zinc nitrate
ZO is responsible for the formation of by-products such as aldehydes, showed the highest activity. The yield (based on reacted ethanol) and
ethers and traces of di-ethyl-ethers and butanal. The rates of formation selectivity at optimal conditions for maximum activity of industrial
of main products are expressed thus: samples of A, B and C are presented in the table below.
As shown in the table, it could be seen that the most effective is B.
w = −r1 – r2 Ethanol The addition of 1–1.5 % of the initiator in relation to the quantity of
w1 = r3 + r5 + r7 Butadiene ethanol has a complex modifying, initiating and regenerating effect.
w2 = r6 – r7 Butylene The modifying effect can be considered to be as a result of hydroxylation
w3 = r1 –r3 –r4 – 2r6 Ethylene
of the catalyst surface by hydroxyl radicals formed by the thermal
w4 = r2 – r3 – r4 – 2r5 Acetaldehyde
w5 = r4 Butanal
decomposition of hydrogen per oxide.
As seen from the results of the study (see Table 3), the initiation
effect is achieved with an optimal flow rate of at least 2.5h−1. This effect
where wi is the rate of formation for each route, ri is the rate of formation is more significant for samples A and B as shown in Fig. 5 with optimal
of the main product. flow rate of 3.0–4.4 h−1. The effect on the industrial sample of 1985
The limiting steps are 1, 3, 5, 8, 12, 13, and 15 due to the breaking of (sample used at the Russian Efremov plant suspected to be Lebedev's
C\H bond. The rates were finally calculated by the following formulas: catalyst) and lab sample C made of α aluminium oxide was practically
insignificant.
k1 X k13 X 3 2 k15 X 2 A continuous process was achieved with the use of the initiator on A
r1 ¼ ; r6 ¼ ; r7 ¼ ; For Z centre b1 ¼ 500
bX 4 bX 4 bX 4 and B samples on which a high linear velocity was used, which excluded
the recombination of OH• and HO•2 radicals before they get to the
k3 X k5 X 3 k12 X 4 catalyst. This generally explains the synergetic effect between the
r2 ¼ ; r3 ¼ ; r5 ¼ ; For ZH centre b2 ¼ 100
bX 4 b b catalyst and the initiator. The parameters obtained on C are similar to
those obtained on the industrial sample which suggests that alpha
k8 X 3 form of Al2O3 dominates in the industrial sample.
r4 ¼ ; For ZO centre b3 ¼ 550 Thus, it was found that the effectiveness of the initiator is seen only
b
for the catalyst composed of γ-Al2O3. This efficiency can be tied to the
where X, X1, X2, X3, X4, and X5 are the molar percents of ethanol, suppression of coke formation process; initiation of the process by the
butadiene, butylene, ethylene, acetaldehyde and butanal respectively; pair of radicals in a homogenous ambience, and by the modification of
K is the constant of the stages as shown in Table 2; and b is the the catalyst surface via hydroxylation.
adsorption coefficients. Our kinetic study showed that the effect has to do with the
From our kinetic model, it was observed that the competition manipulation of the linear velocity of the stream containing hydrogen
between the adsorption of acetaldehyde and butanal affects the peroxide. With linear velocities above 2.5 h−1, the initiation effect
 G.O. Ezinkwo et al. / Catalysis Communications 43 (2014) 207–212 211

Fig. 4. Graphic representation of the initiation process and catalysts centers.

becomes significant which ultimately increases the productivity and the feed ethanol, an effect of selectivity increase was observed. In the
selectivity of butadiene production process. Due to the fact that the α- case of 90% ethanol solution, the effect disappeared. The initiating effect
Alumina can only be used with a flow rate of less than 1.5 h−1 (which was however absent when 95% ethanol solution was mixed with 15%
is a major constraint), the industrial sample used both in metal and hydrogen peroxide solution. However, in all the above described
quartz reactor were not affected by the introduction of the hydrogen cases, the need for catalyst regeneration did not arise. It was, con-
per oxide i.e. no increased activity of catalyst was observed. This can sequently deduced that water inhibits the formation of OH• and HO•2
be explained by the fact that the pair of radicals is converted into radicals responsible for initiating the process.
water before they get to the catalyst bed. It was discovered that the use of 60% hydrogen peroxide solution
The integration of hydrogen per oxide into the process increases the reproduced a more stable result with the achievement of 55% butadiene
selectivity to butadiene, product yield per cycle and the efficiency of the selectivity. From these data, it was concluded that the process of
catalytic process overall. However, it was observed that the effective- initiation depends on the residence time of hydrogen per oxide in the
ness of the process of initiation on a catalyst made of γ-Al2O3 is vaporization zone of the reactor and its contact time with the catalyst.
dependent on the concentration of hydrogen peroxide in the feed This fact indicates the homogeneous–heterogeneous characteristics of
stream. When 30% hydrogen per oxide solution was mixed with 95% the initiated catalytic process. It was also noted that when the reaction
ethanol solution based on a calculation of 1% hydrogen per oxide in was conducted at 410 °C in an empty reactor and in a reactor filled
with quartz nozzle there was practically no yield of butadiene and
conversion of ethanol did not exceed 5%. However, location of the quartz
Table 3 nozzle after the catalyst bed in the quench zone of the reactor led to the
Parameters at optimal conditions for maximum activity of samples and initiator effect. reduction of selectivity. But also, location of the quartz nozzle above the
s/n. Catalyst Т, ○С Flow rate, h−1 Butadiene, % catalyst bed in the vaporization zone, led to the reduction of initiation
effect of Hydrogen per oxide. These facts depict the homogeneous–
Yield Selectivity
heterogeneous nature of the reaction mechanism which calls for more
1 Industrial sample of (1985) 420 1.2 18.1 44.2 detailed study of the observed effect.
2 А 410 3 20.0 48.0
3 В 400 4.2 22.0 53.1
Noteworthy also is that the study of the linear velocity at a given
4 С 425 1.1 17.2 42.3 contact time by the simultaneous increase of flow rate and catalyst
5 Industrial (+ initiator) 415 1.3 18.2 44.5 loading leads to the discovery of the dependence of initiation action of
6* А (+ initiator) 400 3.2 22.3 54 H2O2 on the process parameters. With the increase in the ratio of
7* В (+ initiator) 395 4.4 24.5 55
catalyst bed diameter relative to its height, it was observed that the
8 С (+ initiator) 420 1.2 17.5 43.2
effect became stronger. It was also noticed that replacing γ-Al2O3 with
The asterisk represent various limiting steps. α-Al2O3 gave rise to the need to work with a flow rate two times
lower than the former. According to studies carried out under the latest
conditions of α-Al2O3 and flow rate two times lower, the initiation effect
60 Sample B of H2O2 in relation to butadiene selectivity was not observed.
Industrial Sample A
50 Sample SampleC
selectivity %

40 References
30 [1] M.D. Jones, C.G. Keir, C.D. Iulio, R.A.M. Robertson, C.V. Williams, D.C. Apperley, Catal.
Sci. Technol. 1 (2011) 267–272.
20
[2] E.V. Makshina, W. Janssens, B.F. Sels, P.A. Jacobs, Catal. today 198 (2012) 338–344.
10 [3] S.V. Lebedev, Zh. Org. Khim. 111 (1931) 698–702.
[4] S.V. Lebedev, British Patent 331,402, 1929.
0 [5] S.V. Lebedev, J. Gen. Chem. 3 (1933) 698–708.
1 2 3 4 [6] S.V. Lebedev, Chem. Ztg. 60 (1936) 313–316.
[7] S.K. Bhattacharyya, N.D. Ganguly, J. Appl. Chem. 12 (1962) 97–104.
Fig. 5. Selectivity of the various samples with introduction of initiator. [8] S.K. Bhattacharyya, N.D. Ganguly, J. Appl. Chem. 12 (1962) 105–110.
212 G.O. Ezinkwo et al. / Catalysis Communications 43 (2014) 207–212

[9] S.K. Bhattacharyya, B.N. Avasthi, Ind. Eng. Chem. Process. Des. Dev. 2 (1963) 45–51. [21] J.-B.D. Delacaillerie, V. Gruver, J.J. Fripiat, J. Catal. 151 (1995) 420–430.
[10] B.B. Corson, E.E. Stahly, H.E. Jones, H.D. Bishop, Ind. Eng. Chem. 41 (1949) [22] G. Natta, R. Rigamonti, Chim. Ind. 29 (1947) 195–201.
1012–1017. [23] M. Leon, E. Diaz, S. Ordonez, Catal. Today 164 (2011) 436–442.
[11] B.B. Corson, H.E. Jones, C.E. Welling, J.A. Hinckley, E.E. Stahly, Ind. Eng. Chem. 42 [24] J.M. Berak, R. Guczalski, J. Wojcik, Acta Chim. Acad. Scientarium Hungaricae 50
(1950) 359–373. (1966) 163–167.
[12] Y. Kitayama, K. Shimizu, T. Kodama, S. Murai, T. Mizusima, M. Hayakawa, M. [25] W. Szukiewicz, US 2,357,855 (1944).
Muraoka, Stud. Surf. Sci. Catal. 142 (2002) 675–680. [26] G. Natta, R. Rigamonti, Chim. Ind. 29 (1947) 239–243.
[13] R. Ohnishi, T. Akimoto, K. Tanabe, Chem. Commun. (1985) 1613–1614. [27] Y. Kitayama, M. Satoh, T. Kodama, Catal. Lett. 36 (1996) 95–97.
[14] S.V. Lebedev, Jizni & Truda ONTI KHIM, TEORET Leningrad (1938) 878–882. [28] S. Teir, S. Eloneva, R. Zevenhoven, Energy Convers. Manage. 46 (2005) 2954–2979.
[15] N.A. Plate, E.V. Slivinskii, Fundamentals of Chemistry and Technology of Monomers, [29] S.K. Bhattacharyya, S.K. Sanyal, J. Catal. 7 (1967) 152–158.
Nauka/Interperiodika Publishing, 2002. 111. [30] Wm. Claude White, Chem.-Biol. Interact. 166 (2007) 10–14.
[16] J. Burla, R. Fehnel, P. Louie, P. Terpeluk, http://repository.upenn.edu/cbe_sdr/42Feb [31] R.М. Talyshinsky, R.G. Rizaev, V.S. Gadji-Kasumov et al. Method for Divinyl
28 2013. production. US patent 4198586 (1978).
[17] S.S. Kvisle, A. Aguero, R.P.A. Sneeden, Appl. Catal. 43 (1988) 117–131. [32] R.М. Talyshinsky, R.G. Rizaev, V.S. Gadji-Kasumov, et al., Method for Divinyl
[18] K. Arata, H. Sawamura, Bull. Chem. Soc. Jpn. 48/11 (1975) 3377–3378. Production. FR 2444019, BI47, 1980.
[19] V. Gruver, A. Sun, J.J. Fripiat, Catal. Lett. 34 (1995) 359–364. [33] V.F. Tretjakov, S.N. Hadjiev, R.M. Talyshinski, A.L. Maximov, A.M. Ilolov, Method for
[20] Y. Kitayama, A. Michishita, J. Chem. Soc. Chem. Commun. (1981) 401–402. Divinyl Production (Variant). RU 2 459788, 24, 2012.