1
88
S. Hanspal et al. / Journal of Catalysis 352 (2017) 182–190
Table 2
Product distribution during ethanol coupling at 633 K.
Catalyst
Ethanol
conv. (%)
Rate of ethanol conversion
Rate of AcH production
Rate of BuOH production
Selectivity (C%)
ꢁ
2
ꢁ1
ꢁ2 ꢁ1
ꢁ2 ꢁ1
(mol m
s
)
(mol m
s
)
(mol m
s
)
Ethane Ethene AcH DEE BuOH
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
ꢁ
9
ꢁ9
ꢁ10
MgO
4.4
5.0
4.2
2.7
4.9
4.2
2.4 ꢄ 10
6.5 ꢄ 10
2.7 ꢄ 10
2.3 ꢄ 10
8.7 ꢄ 10
1.6 ꢄ 10
1.4 ꢄ 10
7.6 ꢄ 10
1.6 ꢄ 10
1.4 ꢄ 10
2.4 ꢄ 10
1.5 ꢄ 10
4.0 ꢄ 10
0
0
0
2
0
0
0
8
36
0
0
0
59
12
59
62
28
91
0
52
4
0
0
33
0
35
38
72
9
9
8
8
8
7
ꢁ10
ꢁ8
ꢁ8
ꢁ8
ꢁ7
Mg
3
(PO
4
)
2
ꢁ9
ꢁ9
ꢁ8
ꢁ9
b-TCP
FAP
HAP
4.7 ꢄ 10
4.3 ꢄ 10
3.1 ꢄ 10
7.2 ꢄ 10
Sr
3
(PO
4
)
2
0
0
AcH – acetaldehyde; DEE – diethyl ether; BuOH – butanol.
using density functional theory. They found that molecular adsorp-
tion of water occurs on the (001) surface of HAP with a heat of
strength and provide a low accessibility to the surface Lewis acid
sites associated with coordinatively unsaturated cations (i.e.
ꢁ
1
2+
adsorption of approximately 80 kJ mol
whereas dissociative
Mg ) [9]. These strong base sites on MgO, as measured by CO
2
adsorption of water occurs on the (010) HAP surface. The interac-
tion of water was so strong on the (010) surface that the research-
ers speculate the surface was unlikely to be present on HAP that
had been synthesized in an aqueous environment. This observation
has important implications in reactions where water is formed
in situ as a product such as Guerbet coupling and aldol condensa-
tion and may account for the significantly higher catalytic activity
of HAP for these reactions compared to MgO [19,38].
adsorption microcalorimetry (Fig. 5), also facilitate undesired etha-
nol dehydration to ethene via the E1cB pathway [39]. The relative
strength of the acid and base sites that are involved in the active
site pair must be carefully balanced to achieve high activity and
selectivity to butanol while minimizing unwanted by-product
formation.
Increasing the acid site density on the MgO surface or balancing
the relative strengths of the acid and base sites on the surface has
been beneficial for the Guerbet coupling reaction [8,40]. This opti-
mization of acid-base strength as a strategy for improving Guerbet
coupling catalysts is consistent with the results presented in
Table 2. The total rate of ethanol conversion over the b-TCP and
FAP catalysts was an order of magnitude higher than that mea-
sured over MgO. Compared to MgO, these materials expose base
sites of much weaker strength (Fig. 5b) and acid sites in signifi-
cantly higher surface density (Fig. 6a). An increased number of
modest-strength acid-base site pairs on the surface likely allows
b-TCP and FAP to catalyze the ethanol conversion reaction at a sig-
nificantly higher rate than on highly-basic MgO.
3.2.3. Performance of phosphate catalysts
The observed inhibitory influence of water on butanol forma-
tion over stoichiometric HAP during the steady-state conversion
of ethanol (Fig. 9) and prior observations that water adsorption
3
ꢁ
on HAP occurs on PO
importance of the PO
4
[22,34] motivated us to investigate the
group for C-C bond formation during the
3
ꢁ
4
ethanol coupling reaction. In addition, the potential roles of cations
2
+
2+
2+
ꢁ
ꢁ
(
Ca , Mg , Sr ) and anions (OH , F ) on the reactivity of
phosphate-based catalysts were explored. A variety of catalysts
such as b-TCP, Sr (PO , Mg (PO , and FAP were tested in Guer-
bet coupling and compared to HAP The product distribution and
3
4
)
2
3
4 2
)
The critical requirement of balanced-strength acid and base
sites for the Guerbet coupling reaction is clearly visible when com-
paring the results obtained during ethanol conversion over MgO
.
steady-state rate observed during ethanol conversion at 633 K for
the catalysts are listed in Table 2. All catalytic reactions were con-
ducted at low ethanol conversion (ꢀ5%) to ensure differential reac-
tor conditions.
The HAP catalyst exhibited the highest catalytic activity for
butanol production and the highest selectivity to butanol (72%)
among all of the catalysts tested. The b-TCP and FAP catalysts were
3 4 3 4 2
and Mg (PO )2. The phosphate group of Mg (PO ) is considerably
2ꢁ
less basic than the O anions present on the MgO surface so that
the catalysis shifts from ethanol dehydrogenation over MgO
toward acid-catalyzed ethanol dehydration over Mg
3 4 2
(PO ) . The
Mg (PO catalyst exhibited characteristic acid-like behavior lead-
3
4 2
)
ing to 52% and 36% selectivity toward diethyl ether and ethene,
respectively. Additionally, this material was catalytically inactive
for C-C coupling to butanol (Table 2).
ꢀ
3 times less active on an area basis and ꢀ50% less selective
toward butanol, compared to HAP, which implies that the hydroxyl
group of HAP likely plays a beneficial role in the catalysis during
the coupling of ethanol to butanol. Although the catalytic activity
and selectivity to butanol over b-TCP and FAP were not as impres-
sive as those over HAP, the Guerbet coupling reaction was never-
3 4 2
To further characterize the acid sites present on the Mg (PO )
surface, DRIFTS of adsorbed pyridine was performed over a fresh
catalyst diluted in KBr. Infrared spectroscopy of pyridine is com-
monly used to characterize acid sites because the IR signature of
pyridine coordinated to a Lewis acid site on the surface is very dif-
ferent from that of the pyridinium ion, i.e. when it is associated
with a Brønsted acid, which permits differentiation between acid
types on solid acid surfaces [41]. It should be noted that the acid
properties of the catalyst could be altered by exposure to reaction
conditions (which produce water), so the conclusions from DRIFTS
are relevant to the fresh catalyst.
3ꢁ
theless still observed. This observation suggests that the PO
4
group is critical for the production of butanol over b-TCP and FAP
and is likely involved in the active acid-base site pair for butanol
formation during ethanol coupling over HAP. These results are con-
sistent with the water co-feeding experiments over HAP (Fig. 9)
where the inhibiting effect of water on butanol formation was
likely the result of water interactions with adjacent Ca2+-PO
3ꢁ
4
site
pairs on the surface [34].
The DRIFTS spectrum of adsorbed pyridine on Mg
3
(PO
4 2
) at
The MgO catalyst was the least active among all of the catalysts
tested and catalyzed primarily the dehydrogenation of ethanol to
acetaldehyde (59% selectivity). The undesired ethene, which was
formed by dehydration of ethanol, was also observed (8%
selectivity).
3
73 K is presented in Fig. 11. Pyridine adsorbed on Mg
3
(PO )
4 2
ꢁ1
resulted in bands at 1608, 1576, 1491, and 1446 cm which are
characteristic of Lewis-type coordination with the surface [42].
The absence of a band at 1540 cm , characteristic of the pyri-
dinium ion, indicates a lack of Brønsted acidity associated with
ꢁ1
As discussed earlier, the Guerbet coupling of ethanol to butanol
likely proceeds over an acid-base site pair, and the poor catalytic
performance of MgO in alcohol coupling reactions has been attrib-
the Mg
on Mg (PO
dration observed over this material was Lewis-acid catalyzed. The
3
(PO
4
)
2
surface. Therefore, the surface acid sites exposed
3
4 2
) are of the Lewis type and confirm that ethanol dehy-
uted to strongly basic O2 anions that are disproportionate in
ꢁ