G Model
APCATA-15012; No. of Pages7
ARTICLE IN PRESS
2
F. Diehl et al. / Applied Catalysis A: General xxx (2014) xxx–xxx
◦
in the severe conditions of the engine combustion chamber. In this
study, hydrocarbon-air mixtures were injected on the catalyst after
vaporization of the hydrocarbon in a saturator. This investigation
was limited to HC having a sufficient volatility to get 1500 ppm C
in the saturator (alkanes up to C20 and polycyclic hydrocarbons
up to C13). For heavier hydrocarbons, this technique could not be
applied. Oxidation tests were then carried out using the classical
method employed for diesel soot oxidation: the solid hydrocarbon
is intimately mixed with the platinum catalyst and light-off profiles
are recorded with oxygen only in the gas [30–32].
550 C, NO exp, can be compared to NO2th, which gives the percent-
2
age of theoretical O2 consumed in TPO:
N
O2 exp
%O2 cons = 100 ×
(2)
N
O2th
Due to the fact that the hydrocarbon starts to vaporize before
it can be oxidized, NO2 exp is always smaller than NO2th. In this
approach, HC conversion is calculated on the basis of O2 con-
sumption profile NO2 vs. T. For instance, T50 is the temperature
at which half of NO2 exp has been consumed. Several tests were
repeated to determine the interval of confidence of the light-off
temperatures. It was concluded that the temperatures at 50% con-
2
. Experimental
◦
version are given at ± 5 C. Oxidation of solid hydrocarbons may
2.1. Catalysts
also give rise to aldehyde/ketones, etc. or to lighter hydrocarbons
(by cracking/dealkylation, etc.). These products were not system-
The condition of the test made that a relatively high sample
atically analyzed but their presence (generally less than a few
percents) was evidenced by GC/MS coupling.
weight was used for each light-off test. Therefore, it was decided
to prepare a less loaded catalyst for the present study. Experi-
ments were carried out on a 0.55 wt%Pt/Al O catalyst prepared by
3. Results and discussion
2
3
wet impregnation of a ␥-alumina support with an aqueous solu-
tion of hexachloroplatinic acid. The support (grain size: 2–4 mm;
3.1. Preliminary study: influence of Pt loading
2
−1
3
−1
1
0
05 m g ; pore volume: 1.18 cm g ) was crushed and sieved to
.125–0.250 mm, calcined at 500 C and impregnated with chlorhy-
◦
Oxidation of n-tetradecane and naphthalene was studied under
constant HC and O2 concentration at the reactor inlet (1500 ppm C
in air, see details in [28]) over four Pt catalysts of variable load-
ing between 0.1 and 1 wt%Pt. All these catalysts were prepared
dric acid (2%) before Pt impregnation. This chlorination ensures
a homogeneous distribution of Pt in the alumina balls. The chlo-
◦
rinated support was dried at 150 C. After impregnation with
2
−1
hexachloroplatinic acid the catalyst was slowly dried in vacuum
on the same alumina (105 m g ). H2-O2 titration and electron
microscopy showed that metal dispersion was close to 100% for
the three less loaded catalysts (0.1 to 0.55%Pt) while it is slightly
lower (80%) for the 1 wt% catalyst. The results of the oxidation tests
are reported in Table 1. Specific activity and turn-over frequency
are virtually constant for the 0.1–0.55 wt% catalysts while they are
a little bit greater for the 1 wt% catalyst. This may be ascribed to
the moderate increase of particle size in this latter catalyst. Other
results (not shown here) have confirmed that HC oxidation was a
structure sensitive reaction over Pt, TOF increasing with the particle
size. Substituting the 1 wt%Pt catalyst for the 0.55 wt% one increases
◦
◦
at 100 C (10 h) and then at 150 C (10 h). The solid was dechlo-
rinated by washing with ammonia (0.1 M) and rinsed with pure
◦
water up to pH 7. It was dried again at 100–150 C and calcined
◦
at 500 C (2 h). No change of textural properties was noticed after
Pt impregnation, drying and calcination. Chemical analyses led to
the following composition (wt% or ppm): 0.55% Pt; 110 ppm K;
4
0 ppm Cl; 590 ppm Ca; 260 ppm Fe; 500 ppm Na and 300 ppm Si.
The catalyst was finally crushed and sieved to less than 50 m.
In a preliminary study, the impact of the metal loading (between
0.1 and 1 wt%) was investigated. The low-loaded catalysts were
◦
prepared by the same procedure.
the light-off temperature by 14 C when HC oxidation is performed
The catalysts were characterized by H -O titration in a pulse
in the saturator reactor (constant concentration at the reactor inlet
during the light-off test).
2
2
chromatic apparatus [33]. A dispersion close to 100% was calculated
assuming a stoichiometry H/PtS and O/PtS close to unity. Trans-
mission Electron Microscopy showed very small Pt particles below
3.2. Hydrocarbon oxidation: comparison between the two
reaction systems
1
.2 nm.
2
.2. Light-off tests
Catalytic oxidation of some hydrocarbons could be carried out
both in the saturator reactor (results in ref. 26, 1 wt%Pt) and in the
soot oxidation reactor (present study, 0.55 wt%Pt). Fig. 1 compares
the light-off curves recorded in the two reaction systems (N being
the name of the hydrocarbon, N corresponds to light-off in the
saturator reactor while N-S corresponds to light-off in the soot oxi-
dation reactor). Except for naphthalene, oxidation occurs at higher
temperatures in the soot oxidation reactor. The vaporization is an
essential step before oxidation. Naphthalene being relatively easy
to vaporize, oxidation is not delayed and can occur at low tem-
perature: in the soot oxidation reactor, naphthalene concentration
may become very high, which favors a “light-off’‘-type combustion.
Fluorene and hexamethylbenzene having a lower volatility (see
Table 2), their oxidation could be delayed up to reach a sufficient
concentration in the stream.
The preliminary study on the impact of Pt loading was
performed in the saturator reactor described in ref. [28]. All
the other experiments were carried out in a soot oxidation
reactor. Two grams of 0.55 wt%Pt/Al O3 catalyst (50 m) were
physically mixed with 0.133 mmol of hydrocarbon. This ratio
corresponds, for instance, to 40 mg coronene in 2 g of catalyst.
Temperature-programmed oxidation of 100 mg of this HC/catalyst
mixture (6.65 mol HC) was carried out from ambient tempera-
ture to 550 C at 5 C min in a 1%O /He mixture (gas flowrate:
2
by catharometry (after trapping of CO , H O and not reacted HC
on Zeolite 13X and KOH), which allows to calculate the integrated
2
◦
◦
−1
2
3
−1
0 cm min ). Oxygen concentration was continuously recorded
2
2
◦
◦
amount of O2 consumed between 20 C and T C and the HC con-
version at temperature T. It is assumed that the oxidation reaction
produces only CO2 and water according to eq. (1):
3.3. Hydrocarbon oxidation in the soot oxidation reactor
CxHy + (x + y/4)O → xCO + y/2 H O
The amount of HC introduced in the reactor (6.65 mol) gives
the theoretical amount of O2 (NO2th) required for total oxidation.
(1)
2
2
2
Twenty-two hydrocarbons having a number of carbon atoms
HC are reported in Fig. 2 (HC of less than 20 atoms C) and Fig. 3 (HC
having 20 atoms C or more). For all the tested hydrocarbons, the
The experimental amount of O actually consumed between 20 and
2