Applied Catalysis, 48 (1989) 265-277 265

Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

The Bioethanol-to-Ethylene (B.E.T.E.) Processa

RAYMOND LE VAN MAO*, THANH MY NGUYEN and GERALD P. MCLAUGHLIN

Catalysis Research Laboratory, Department of Chemist? and Laboratories for Inorganic
Materials, Concordia University, 1455 De Masionneuve Blvd. West, Montreal, Quebec H3G
1M8 (Canada)
(Received 6 June 1988, revised manuscript received 11 October 1988)

ABSTRACT

Ethanol present at low concentration in a fermentation broth which also contains some leftover
glucose as well as other fermentation residues. is totally converted into ethylene over a solid su-
peracidic catalyst. Trifluoromethanesulfonic acid bound to the silica rich surface of the acidic
ZSM-5 zeolite is thermally stable up to 300 ‘C and can catalyse the reaction with nearly complete
conversion at a temperature as low as 17O’C. Below 2OO”C, the catalyst is extremely stable even
in the presence of the high amount of water found in the feed. A rapid economical evaluation
shows that ethylene produced by the B.E.T.E. (bioethanol-to-ethylene) process has a production
cost 10% to 40% higher than petroleum ethylene. However, a more promising figure can be envis-
aged with a better integrated process working with cheaper raw materials.

INTRODUCTION

Ethanol is currently produced by hydration of ethylene [ 1]which is derived

mainly by steam-cracking of petroleum or natural gas feedstocks. Ethanol can
also be obtained by the anaerobic fermentation of sugars (mostly glucose)
which in turn are the products of (acid or enzymatic) hydrolysis of cellulosic
compounds [ 2-61. These compounds can be “extracted” from biomass, partic-
ularly wood materials. In the latter case, a sequence of four main steps is nec-
essary. Fig. 1 reports the entire process which starts from biomass or lignocel-
lulosic raw materials and leads to ethanol. Due to today’s low cost of petroleum
ethanol and the relatively high production cost of fermentation ethanol, the
competitiveness of the latter is regularly the subject of discussion [ $61.
On the other hand, ethanol is rather easily dehydrated to ethylene over acidic
alumina. However, the reaction must be carried out with at least a 95 vol.-%

“Presented to the Third Chemical Congress of North America - American Chemical Society,
Toronto, June 6-2,1988

0166.9834/89/$03.50 0 1989 Elsevier Science Publishers B.V.

 PRODUCTION COST(%)

Fig. 1. Comparison of costs for production of ethanol and ethylene from biomass. A:cost of raw
material and pretreatment; B: cost of acid or enzymatic hydrolysis, C: cost of alcohol fermentation,
D: cost of conversion of ethanol to ethylene by B.E.T.E. process, E: cost of distillation and dehy-
dration of ethanol from the fermentation broth. F: cost of conversion of pure or 95% ethanol to
ethylene by conventional catalytic process 17 j.

ethanol and at rather high temperatures, 300-400°C. The practical ethylene

yield is approximately 94% of the theoretical value [ 71. The presence of water
in the ethanol feed is definitely detrimental to the dehydration reaction.
Gasoline can be produced from ethanol by catalytic conversion over ZSM-5
type zeolite [&lo]. In the latter process, a concentration of ethanol of at least
60 vol.-% is necessary [lo]. Thus, according to the today’s state of art, two
steps are needed to produce ethylene or gasoline (the first step being the ethanol
distillation from the fermentation broth).
In our laboratory, zeolite catalysts which allow the production of practically
pure ethylene from aqueous ethanol (concentration of ethanol in the aqueous
solution ranging from 2 to 19 vol.-% ) have been developed [ 11-15). This also
means that ethylene can be produced directly from ethanol fermentation broth
in a one-step process (called B.E.T.E. or bioethanol-to-ethylene, see also Fig.
1) [ 151. The latest version of these catalysts, a superacidic ZSM-5 zeolite, can
catalyse the reaction at temperatures as low as 170’ C and under atmospheric
pressure [ 151. As a consequence, its total energy requirement appears ex-
tremely low, when compared to the conventional ethanol distillation followed
by a catalytic dehydration to ethylene over acidic alumina.
A fully integrated and continuous process which uses in the final step a spe-
cially designed reactor has an excellent chance of producing ethylene at a com-
petitive cost [ 151. Much work must still be accomplished, primarily in the
hydrolytic step, in order to further decrease the production cost of ethylene as
end-product.
However, “Should this route become of industrial significance it could lead
to the paradox that in some parts of the world ethanol would be dehydrated to
ethylene while elsewhere ethylene would be hydrated to produce ethanol” [ 71.
 267

EXPERIMENTAL AND RESULTS

Preparation and characterization of the catalysts

The hydrothermal synthesis of ZSM-5 zeolites and their conversion into the
acidic catalysts were reported elsewhere [ 16,171. The preparation of t.he steam
treated zeolite (first generation of B.E.T.E. cata1yst.s) and that of the super-
acidic ZSM-5 zeolite (latest catalyst generation) were described in some recent
patent applications [ 13-151.
(i) In particular, the st.eamed H-ZSM- 5 (Si/Al atomic ratio = 21) in powder
form, was obtained by stream treating at 500°C for 3 h prior to its extrusion
with bentonite [ 131.*‘Al magic angle sample spinning NMR (“Al-MASS
NMR) of such a treated powder did not show any significant presence of the
octahedral aluminium; thus, the existence of a new aluminium species assumed
to be the pentacoordinated aluminium was suggested [ 18,191. Moreover, the
hydrothermal treatment of the ZSM-5 zeolite seems to significantly affect the
overall acidity, as shown by our latest ammonia TPD studies [ 201.
(ii) One way of preparing superacidic zeolites is to coat the silica rich surface
of ZSM-5 zeolite (acid form) with triflic acid (trifluoromethanesulfonic acid
or TFA). The incorporation of triflic acid onto the H-ZSM-5 (powder) is done
by slowly adding a solution of TFA dissolved in a minimal amount of solvent
(acetone, for instance) with gentle stirring. The mixture is allowed to air dry
at room temperature for several hours with periodic mixing. The solid obtained
is then heated at 120’ C for (at least ) 12 h [ 141. The final catalyst was prepared
by extrusion in the presence of bentonite as binder [ 171.
The resulting material (H-ZSM$/TFA, powder) is stable up to 300°C
whereas pure triflic acid starts boiling at 162°C. In fact, thermal gravimetric
analysis shows a loss of weight starting at 300’ C which corresponds exactly to
the amount of TFA incorporated into the ZSM-5 zeolite (see Fig. 2). Moreover,
the volume of n-hexane adsorbed into the TFA treated zeolite decreased with
the increase of TFA loading, suggesting a certain pore size narrowing due to
the presence of TFA on the zeolite pore walls (Table 1).
The adsorption of n-hexane and water was done following the procedure
described by Le Van Mao [ 211 and Le Van Mao et al. [ 221.
Solid state *‘Si-MASS NMR of this material (spectra recorded on a Varian
XR-300 spectrometer at 59.59 MHz) shows some deshielding effects of the
bonding between the TFA and the surface silicon atoms of the ZSM-5 zeolite.
“Pure” H-ZSM-5 zeolite shows two bands (chemical shifts 6= - 112 ppm and
6= - 107 ppm, with respect to tetramethylsilane) which are attributed to the
Si(-0-Si), and Si(-O-Al) (-0-Si), bonding, respectively. Spectra of H-ZSM-
5/TFA samples have an additional band (shoulder) at 6= - 116 ppm; their
- 107 ppm band have a significantly lower intensity.
Although the - 116 ppm band is not strictly peculiar for the H-ZSM-5/TFA
268

200 300 400 500

temperat”re [“c)

Fig. 2. Thermal gravimetric analysis of pure (A j ZSM-5 and (B) ZSM-5 with TFA coated on
surface (TFA loading: 4 wt.-%).

TABLE 1

Variations in the adsorption properties of the H-ZSM-5 zeolite (silicon-to-aluminium 10) with
the TFA loading.

TFA loading n-Hexane adsorbed Water adsorbed

(wt.-%) (vol.-%) iv01 %)

0 9.3 10.5
1.1 6.8 9.0
3.3 5.9 10.8
5.3 6.0 9.9
1.2 4.3 12.4

zeolite (titanium-silicalite TS-1 [ 231, for instance, shows the same band), these
NMR spectra can provide useful indications for further investigation on the
modified zeolite surface.

Catalytic testing and analysis of the reaction products

The fermentation broth from which the yeast has been removed, was sent
directly to a special vaporiser. The latter was set at 120-150°C to vaporize the
volatile compounds without decomposing the leftover glucose and the yeast
nutrient - which can then be recycled to the fermentor or periodically removed
as valuable by-products such as fodder. The gaseous mixture which consisted
of ethanol, steam and some other volatile substances coming from the previous
steps, flowed through a preheating zone to the catalyst bed, as shown in Fig. 3.
 269

catalyst bed

=l.(..Io180t
preheating zone i
J

outlet for non-volatile substances

I’

Fig. 3. Laboratory reaction system for the B.E.T.E. process

In most of our laboratory experiments, glucose (2-4 wt.-%) was added to a 10

wt.-% ethanol in water, to roughly simulate the fermentation broth. In some
runs, formic and acetic acids (1 wt.-%) - representing the carboxylic acids
which may be present in the fermentation broth - along with some salts (1
wt.-% ) - which simulate yeast nutrients - were also added to the feed.
The reaction products were then condensed and the gaseous phase, sepa-
rated from the liquid one. The gaseous phase was analysed by gas chromato-
graphy (GC), the liquids by GC mass spectrometry as already reported [ 121.
Table 2 reports the catalytic data of two reference H-ZSM-5 zeolites (H
indicates the acid form), the steam treated zeolite and some non-zeolitic cat-
alysts, some of which have been treated with triflic acid.
The yield in product ethylene is defined as the weight percent yield of prod-
uct ethylene to the maximum theoretical (weight) yield of ethylene, according
to the following equation:

ethanol -ethylene + water

100 wt.-% 60.9 wt.-% 39.1 wt.-%

Table 3 reports the catalytic data of the H-ZSM-5 zeolites treated with 2 wt.-
% of triflic acid. The influence of the reaction temperature, the W.H.S.V., and
the silicon-to-aluminium atomic ratio of the zeolite catalyst, on the ethylene
yield can be particularly emphasized.
TABLE 2

Catalytic data of the comparative catalyst

Catalyst chemical properties Reaction conditions Conversion Product selectivity (% )

(wt.-%)
Catalyst Si/Al TFA Temp WHSV Feed Ethylene Other light Other hydrocarbons
ratio loading (“Cl (h-l) ethanol olefins
(wt.-%) (vol.-%)
--
H-ZSM-5 21 - 400 3.2 15 96 49 28 23
H-ZSM-5 21 - 400 3.2 15 98 99 1 0
(steamed)
H -ZSM-,5 10 _ 200 1.0 10 14.0 99.8 0.2 0.0
Alumina, Acidic - - 200 1.0 10 0.0 - _-
(Stem Chem)
;I-Alumina/TFA - 2 200 1.0 10 0.0 - -
(92% Al 0)
(Stem Chem)
Silicalite > 100 2 200 1.0 10 13.7 99.6 0.4 0.0
(Union Garb.)
H-Y (Linde)/TFA 2.0 2 200 1.0 10 43.4 93.8 5.3 0.9
Nafion NR-50 - - 200 1.0 10 12.0 100.0 0.0 0.0
(Aldrich, Beads)
 271

,
O, ethylene
0 yield

0 I 2 3
triflic
4 acid5 6 ’ 0
‘0
Fig. 4. Effect of TFA loading on the ethylene yield

Fig. 4 reports the effects of the triflic acid loading on the ethylene yield at
two reaction temperatures (185 “C and 170aC ) .

DISCUSSION

A high ethylene production can be obtained with catalysts of the first gen-
eration (hydrothermally treated H-ZSM-5 zeolites). However, the required
reaction temperatures are so high (at least 300’ C ) that these catalysts undergo
a slow but definite activity decay under the joint effects of heat and steam
(water from the feed). This decay results in a decrease in acid density which
stems from both the annealing effects on hydroxyl groups [24,25] and the
removal of framework aluminium atoms, the latter occurring mainly under
more severe hydrothermal conditions ]26].
Although to our knowledge no strong experimental evidence has been found
for the way in which the triflic acid is bound to the zeolite surface, results from
thermal analysis and Si-MASS-NMR seem to suggest that triflic acid is not
merely dissociatively adsorbed onto the zeolite surface (Fig. 5, Cl ) but that it
actually reacts with the zeolite hydroxyl group to yield the surface species as
tentatively described in Fig. 5, C2. The probability of having such a species is
supported by previous works on reaction of TFA over silicon and tin com-
pounds [ 271. Furthermore, Tanabe and Hattori (281 have hypothesized a sim-
ilar situation (Fig. 5, A) where SbF, is allowed to react on silica-alumina
support.
Therefore, there is a clear cooperative effect between the zeolite surface acid
sites and the superacidic centres created by the triflic acid, and this leads to a
tremendous enhancement in the yield of ethylene. The hypothesized super-
acidic centers (Fig. 5, C2) also reflect the “positive” influence of the zeolite.
Aluminium atoms content on the cataiytic activity (the lower the silicon-to-
aluminium atomic ratio of the ZSM-5 zeolite, the higher the ethylene yield of
272

(+,I 5
i 0
O\ A/4 /O lo\ /O PISi p
/- /,.F\ /*‘\*/ \
/“‘a ‘bYi\ - 0

Fig. 5. Probable TFA/zeolite surface bonding; (A ) superacidic centres on silica-alumina modified

by SbF, [28], (B) “common” acid centres on the H-ZSM-5 zeolite, (C,) acid and base centres
originating from dissociative adsorption of TFA on the H-ZSM-5 zeolite, (C,) new Lewis acid
centres created by the reaction of TFA with the zeolite silanol group.

TABLE 3

Catalytic data for the H-ZSM-5 zeolites treated with TFA (2 wt.-%)

Feed is 10 wt.-‘% ethanol in water

Catalyst Si/ Temp. WHSV Conversion Product selectivity (%)

Al (“Cl (h-‘) to hydroc.
Ethylene Propylene Butenes Others

H-ZSM-5/TFA 10 205 0.9 99.2 95.6 1.3 2.6 0.5

205 2.4 98.2 99.0 0.3 0.7 _
205 4.5 90.0 99.7 0.1 0.2 _
200 1.1 97.3 99.9 0.1 _ -
190 0.9 94.8 99.1 0.3 0.5 -
185 1.0 77.3 99.8 0.1 0.1 -
180 1.1 73.9 99.2 0.3 0.4 -
170 1.0 41.9 99.1 0.6 0.3 _

H-ZSM-5/TFA 21 200 1.0 95.8 97.7 1.1 0.9 0.3

285 1.0 73.7 99.3 0.3 0.3 0.1

H-ZSM-5/TFA 54 200 1.0 72.0 98.1 0.4 1.5 -

the catalyst, see Table 3)) as well as a high resistance to the leaching action of
steam at temperatures well above the boiling point of TFA.
Other “indirect” experimental evidence of such a cooperative action is found
 273

from the following observations:

_ A much lower ethylene yield was observed with the silicalite treated with
triflic acid (Table 2). Silicalite (S-115 type from Union Carbide, silicon-to-
aluminium atomic ratio> 100) which is a crystalline silicate, has the same
channel framework as the ZSM-5 zeolite, but does not show any significant
surface acidity.
- The shape selectivity of the ZSM-5 zeolite (pore size = 0.55 nm) plays a ma-
jor role in preventing catalytic decay by blocking the formation of bulky oli-
gomeric materials. This is probably not the case for the H-Y zeolite (pore
size=0.74 nm) treated with TFA, which undergoes a rapid deactivation, al-
though the initial ethylene yield is rather interesting.
- H-Nafion resin (a copolymer of perfluorinated epoxide and vinylsulfonic
acid) which exhibits “free” surface sulfonic acid groups, can catalyse several
reactions at relatively low temperatures. For example, the gas-phase alkylation
of benzene with ethylene and propylene proceeds at 110°C over H-Nafion resin
[ 291. However, its activity in the B.E.T.E. process is very low even at reaction
temperatures as high as 200” C (Table 2).
The catalyst efficiency reaches a maximum at a triflic acid loading of circa
3.5-4 wt.-% (Fig. 4). This is apparently due to the pore size narrowing upon
the TFA incorporation. In fact, a constant decrease of n-hexane adsorption is
observed when the TFA loading increases, while the amount of water adsorbed
remains practically unchanged (see Table 1). Such a difference in sorptive
properties can be explained by the large difference in kinetic diameter between
n-hexane (0.43 nm) and water (0.265 nm) [30]. Thus, n-hexane being much
bulkier than water is more sensitive to any variations of the zeolite pore size.
Coating the zeolite surface with triflic acid has as immediate consequence, a
decrease of the n-hexane adsorption as reported in Table 1.
On the other hand, if the TFA loading increases but still remains lower than
3.5 wt.-%, there is a sharp increase in reactivity due to an enhancement in
acidic character of the catalyst (see Fig. 4). However, when the pore size re-
duction becomes severe, the reactivity drops significantly. The explanation
can be found in a lower ethanol uptake when the zeolite pore size is reduced
(by the TFA loading) below 0.39 nm which is that of the kinetic diameter of
ethanol (and ethylene).
In terms of catalyst stability, if the reaction is carried out at a temperature
lower than 2OO”C, the TFA bearing zeolite does not show any significant cat-
alytic decay over periods of hundreds of hours. This indicates that the en-
hanced shape-selectivity of the H-ZSM-5/TFA (3.5 wt.-%), the massive pres-
ence of water in the feed and the low reaction temperature (at these
temperatures, carboxylic acids which may act as catalyst fouling agents through
their condensation products, do not undergo any conversion [ 311) are key
parameters for such a catalyst stability. However, some TFA leaching is ob-
served at reaction temperatures higher than 250’ C: this is understandable be-
274

cause the system is approaching the 300°C the unstability temperature for
bound TFA.
The equipment required for the B.E.T.E. conversion is very simple (atmos-
pheric pressure, temperature: 170-185 ‘C ). No corrosion problems are envis-
aged at present. The vaporiser design allows in principle the recycling of excess
glucose and yeast nutrients, and thus, this will provide some flexibility to the
previous alcoholic fermentation step.
The net energy requirement (calculated in basis of the thermodynamic and
kinetic data from ref. 18) is only 20% of that usually calculated for the con-
ventional two-stage distillation of ethanol (see Table 4 and ref. 10).
Table 5 shows a very approximate estimate of the ethylene production cost
with the B.E.T.E. process, as well as some values for the production cost of
petroleum ethylene.
The value of ethanol is U.S.$O.25 per liter for production from fermentation.
This assumption reflects and corrects (in consideration of the today’s tech-
nology [ 61) the value proposed by the New Zealand Forest Research Institute.
The latter agency in 1979 gave a figure of less than U.S.$ 0.20 per liter of
ethanol, according to the data they had obtained from a pilot plant (a two-
stage Madison percolation process for the hydrolytic step, starting from wood
materials) [ 32,331.
As a result of our estimates, the production cost of 1 lb of bio-ethylene is

TABLE 4

Net process energy requirement for some processes of ethanol recovery or conversion

Process Net process requirement

Btu/ga” lo6 J/l”

Recovery of ethanol from fermentation broth [lo]:

(i) Conventional two-stage distillation 27 400 7.6
(ii) Conventional distillation (95% )
followed by dehydration 19 200 5.4
(iii) Conventional distillation (95% )
followed by low temperature blending
with gasoline 21000 5.9

Distillation (to 60% ) followed by reactions

to gasoline (based on 1 gal of gasoline processed
which produces 0.678 gal of gasoline [lo] 1800 0.5

Direct conversion of ethanol (from the

fermentation broth) to ethylene or B.E.T.E.
process (our data) 5 500 1.5

” Of ethanol recovered or processed.

 275

TABLE 5

Production cost estimates (April 1988, in U.S. dollars)

Rio-ethylene (ethylene from the B.E.T.E. process)

. Raw material: wood
. Production cost (assumed of fermentation ethanol:
$0.25/l (or Can. $0.31/l)
. Production cost of unpurified fermentation broth which
contains 11 of recoverable ethanol (R.E. ) :
$0.21 (85% of $0.25)
. B.E.T.E. process cost: 5% or $0.01/l of R.E.
. 11 of R.E. provides 1.05 (0.48 kg) of ethylene
. Credit for by-products (fodder, etc.) and yeast nutrients recycling: 20%
. Final production cost of ethylene: $O.l7/lb or $0.375/kg

Petro-ethylene (ethylene from petroleum or natural gas derivatives feedstocks)

. Current selling prices (April 1988):
highest: $0.24/lb or $0.53/kg
lowest: $O.l7/lb or $0.375/kg
. Estimated production cost (with by-product credit):
highest: $O.l6/lb or $0.35/kg
lowest: $ O.ll/lb or 8 0.24/kg

circa $0.17 ($0.375 per kg) whereas that of 1 lb of petro-ethylene ranges from
$0.11 to 0.16 (from $0.24 to $0.35 per kg) (such a large range is due to several
factors: location of processing plants, cost and nature of petroleum feedstocks,
etc.). The selling price of ethylene ranges from $0.17 and 0.24 per lb (from $
0.375 to $0.53/kg) [34].
Thus the difference ranges from 10% to 40%. However, this is not the least
discouraging because of the following factors:
- Ethylene demand and price continue to increase.
- Ethylene production cost in some countries is increasing as the steam-crack-
ing process (actual major source of ethylene) is shifting to heavier petroleum
feedstocks.
- Substantial progress is expected in the hydrolytic step within the entire pro-
cess from biomass or cellulosic materials. The B.E.T.E. reaction can accom-
modate any ethanol content in the fermentation broth ranging from 2 to 19
vol.-%, thus, providing a high flexibility to the fully integrated process “bio-
mass-to-ethylene”. New superacidic catalysts are in the development stage for
converting by-products (carboxylic acids, furfural derivatives, etc.) [31] into
more commercially valuable chemicals and will contribute to increasing the
by-product credit of the process.
- Renewable biomass is available in some countries at very low cost.
- Several countries with no petroleum or natural gas resources, are ready to
276

pay a high “political” price for a technology to produce such an important

commodity chemical as ethylene.

CONCLUSION

Triflic acid which boils at 162°C can react with the hydroxyl groups of the
ZSM-5 zeolite. The resulting surface species is thermally stable up to 300°C.
Such a surface modification provides a higher acidity and an enhanced shape-
selectivity for the selective conversion of aqueous ethanol into practically pure
ethylene. The reaction is carried out at 200°C or less, and there is practically
no need for prepurification or predistillation of the fermentation broth. Fur-
ther, ethanol content as low as 2 vol.-% can be used in the B.E.T.E. process,
this would be advantageous for any process of biomass conversion using dilute
acid in the hydrolysis step.

ACKNOWLEDGEMENTS

The authors thank NSERC-Canada and Actions Structurantes Program of

Quebec for financial support. Precious technical assistance from Dr. J.J. Hech-
ler (NRC-IMI, Boucherville) and Mr. B. Sjiarel is gratefully acknowledged.

REFERENCES

1 P. Leprince, A., Chauvel, J.P. Catry and I. Castrex, Pro&d& de Petrochimie, Technip, Paris,
1971, p. 305.
2 N.P. Cheremisinoff, P.N. Cheremisinoff and F. Ellerbursch, Biomass: Applications, Tech-
nology, and Production, Marcel Dekker, New York, 1980, p. 168.
3 D.L. Klass and G.H. Emert (Eds.), Fuels from Biomass and Wastes, Ann Arbor Sci. Publ.
(Butterworth group), Ann Arbor, 1981, p. 343.
4 S.S. Sofer and O.R. Zaborsky (Eds.), Biomass Conversion Processes for Energy and Fuels,
Plenum Press, New York, 1981, p. 196.
5 J.S. Robinson (Ed.), Fuels from Biomass, Noyes Data Corp., New Jersey, 1980.
6 D.L. Klass, Energy from Biomass and Wastes: 1985 Update, Institute of Gas Technology,
Chicago, IL, 1987, p. 13, and references therein.
7 L. Kniel, 0. Winter and K. Stork, Ethylene: Keystone to the Petrochemical Industry, Marcel
Dekker, New York, 1980, p. 74.
8 E. Costa, A. Uguina, J. Aguado and P.J. Hernandez, Ind. Eng. Chem,. Process Des. Dev., 24
(1985) 239.
9 D.R. Whitcraft, X.E. Veryklos and R. Mutharasan, Ind. Eng. Chem. Process Des. Dev., 22
(1983) 452.
10 G.A. Aldridge, X.E. Veryklos and R. Mutharasan, Ind. Eng. Chem. Process Des. Dev., 23
(1984) 733.
11 R. Le Van Mao and L.H. Dao, U.S. Patent 4 698 452 (Oct. 6,1987).
12 R. Le Van Mao, P. Levesque, G.P. Mclaughlin and L.H. Dao, Appl. Catal., 34 (1987) 163.
13 R. Le Van Mao and L.H. Dao, US. Patent Appl. (Sept. 29,1987).
14 R. Le Van Mao and T.M. Nguyen, Patent Appl. (Apr. 5,1988).
15 R. Le Van Mao, U.S. Patent Appl. (Apr. 25,1988).
16 R.J. Argauer and G.R. Landolt, U.S. Patent 3 702 8886 (Nov. 12,1972).
17 R. Le Van Mao, P. Levesque, B. Sjiariel and P.H. Bird, Can. J. Chem., 163 (1985) 3464.
18 T.M. Nguyen, Conversion of Aqueous Ethanol to Ethylene over ZSM-5 Zeolite and Chryso-
zeolite Catalyst, M. SC. Thesis, Concordia University, Montreal, (July, 1988).
19 Reference cited for the pentacoordinated aluminium species in zeolites: J.P. Gilson, G.C.
Edwards, A.W. Peters, K. Rajagopalan, R.F. Wormsbecher, T.G. Roberie and M.P. Shatlock,
J. Chem. Sot., Chem. Commun., (1987) 91.
20 R. Le Van Mao and J. Yao, Can. J. Chem. Eng., submitted for publication.
21 R. Le Van Mao, React. Kinet. Catal. Lett., 12 (1) (1979) 69.
22 R. Le Van Mao, 0. Pilati, A. Marzi, G. Leofanti, A. Villa and V. Ragaini, React. Kinet. Catal.
Lett., 15 (2) (1980) 293.
23 B. Kraushaar and J.H.C. van Hooff, Catal. Lett., 1 (1988) 81.
24 R.M.Dessau,K.D. Schmitt,G.T.Kerr, G.L. WooleryandL.B.Alemany, J. Catal., 104 (1987)
484.
25 R.M. Dessau, K.D. Schmitt, G.T. Kerr, G.L. Woolery and L.B. Alemany, J. Catal., 109 ( 1988)
472.
26 D. Barthomeuf, J. Phys. Chem., 83 (1979) 249.
27 R.D. Howells and J.D. McCown, Chem. Rev., 77 (1) (1977) 69 and references therein.
28 K. Tanabe and H. Hattori, Chem. Lett., (1976) 625.
29 G.A. Olah, G.K. Surya Prakash and J. Sommer, Superacids, Wiley, New York, 1985, p. 278,
and references therein.
30 D.W. Breck, Zeolite Molecular Sieves, Wiley, New York, 1974, pp. 636 and 637.
31 R. Le Van Mao, G.P. McLaughlin, B. Sjiariel and M.C. Le Van Mao, Concordia University
(Montreal) unpublished results (Sep. 1988).
32 Analysis of ethanol production potential from cellulosic feedstocks, J. Stone and Associates
Ltd., Report to Energy, Mines and Resources Canada, (March 1982).
33 New Zealand Energy Research and Development Committee, The potential of energy farm-
ing for transport fuels in New Zealand, Report No. 46 (1979).
34 Chemical Marketing Reporter, Apr. 8,1986.