Highly selective catalysts for conversion of ammonia to nitrogen in gasified biomass

Article abstract of DOI:10.1039/a904088i

Almost zero emissions of NO(x) can be achieved in the catalytic combustion of simulated biomass mixtures containing substantial amounts of ammonia by use of a heteropolyacid catalyst.

Full text of DOI:10.1039/a904088i

Highly selective catalysts for conversion of ammonia to nitrogen in gasified
biomass
Robert Burch* and Barry W. L. Southward
Catalysis Research Centre, Department of Chemistry, University of Reading, Whiteknights, Reading, UK RG6 6AD.
E-mail: r.burch@reading.ac.uk
Received (in Cambridge, UK) 21st May 1999, Accepted 29th June 1999
Almost zero emissions of NOx can be achieved in the
catalytic combustion of simulated biomass mixtures contain-
ing substantial amounts of ammonia by use of a hetero-
polyacid catalyst.
Renewable energy sources, such as biomass, will be of
increasing importance in the future as part of a strategy to lower
the total emissions of CO2. Recently, the combustion of
biomass-derived gas (biogas) for combined heat and power
generation has been studied.1–3 However, conventional flame
combustion processes create problems because biogas contains
significant quantities of NH3 (600–4000 ppm) in addition to fuel
components (CO, H2, CH4) and on combustion the NH3 is
largely converted into NOx. Catalytic combustion may over-
come this problem but until now the selectivity for the
conversion of NH3 to N2 is unsatisfactory, typically < 70%.2,4
Here, we describe a newly discovered process for removing
NH3 from biogas with almost zero production of NOx.
Fig. 1 NH3 TPD results for HPW. Key: (8) NO (m/z) 30, (:) N2 (m/z 28),
The novel solution to the selective oxidation of NH3 in a
biogas fuel which we have discovered is to differentiate the feed
components on the basis of their chemical properties. The
crucial discovery is that ammonia, being a basic molecule, can
be differentiated from carbon monoxide and hydrogen by using
a catalyst which contains acidic sites, to preferentially adsorb
the ammonia. Combination with redox sites allows the selective
oxidation of ammonia to nitrogen.
The catalyst selected was 12-tungstophosphoric acid,
H3PW12O40 (ex Acros, hereafter denoted HPW). It is a
heteropoly acid material based upon the Keggin unit structure.5
Such materials are well known for their strong and uniform acid
sites arising from the charge balancing protons associated with
the Keggin unit anion. In addition they also possess strong
redox properties arising from surface and bulk electron transfer
processes.5
Catalyst testing was performed in a standard quartz flow
microreactor described previously6 at a gas hourly space
velocity (volume of reactants per volume of catalyst per hour) of
250 000 h21. The reaction mixture comprised 6.0% CO, 4.0%
H2, 0.5% O2, 1050 ppm NH3, and balance He. Product analysis
was by mass spectrometry (Hiden DSMS) with NOx emissions
and residual NH3 levels being confirmed using an external NH3
oxidation reactor (with independent oxygen supply) coupled
with a NOx chemiluminescence detector (Signal series 4000).
The partial salts of HPW were prepared by reflux of the parent
acid with varying stoichiometries of KNO3 (Analar ex Aldrich)
to produce K2.06H0.94PW12O40 (hereafter KHPW) and
K2.66H0.34PW12O40 (hereafter KPW) using a standard exchange
process.5 Temperature-programmed desorption of NH3 was
performed by saturation of the sample at 150 °C in 1% NH3–He
followed by purging in He. The sample was then ramped at
12 °C min21 and the evolved species monitored by mass
spectrometry.
(!) H2O (m/z 18), (/) NH3 (m/z 17, corrected for H2O contribution), (µ)
NH3 (m/z 16).
acid sites of HPW and then converted into N2 and H2O by an
internal reaction with labile oxygen from the Keggin anion. This
reaction is presumed to proceed via the condensation of NO
with NH3, a proposal which is supported by temperature
programmed reaction of a NO–CO–H2 mix (1050 ppm
6.0%+4.0%) over NH3 pretreated HPW. This yielded N2 as the
only product with peak N2 production occurring at ca. 600 °C,
the dissociation temperature of the NH4–Keggin unit complex.7
The amount of N2 produced was consistent with a catalytic
reaction between the NO and NH3.
Fig. 2 shows the reaction of HPW pre-treated with NH3 and
then heated in the full reaction mixture. The results show that
this material displays little or no catalytic function at tem-
+
peratures below the NH4 dissociation temperature. However,
+
simultaneous with the onset of NH4 dissociation there is a
Fig. 1 illustrates the NH3 TPD profile of HPW. The strong
and uniform acidity of the material is demonstrated by the
single, sharp combination of desorption peaks at ca. 600 °C.
These comprise H2O, NH3, N2 and NO, the latter at two orders
of magnitude lower concentration. This was a significant result
since it demonstrated that NH3 may be fixed on the Brønsted
Fig. 2 Conversion profiles for simulated biomass stream over HPW. (6.0%
CO, 4.0% H2, 0.5% O2, 1050 ppm NH3 balance He). Key: (8) NH3
conversion by MS (Ω) H2 conversion by MS, (2) CO conversion by MS,
(µ) % N2 yield by NOx chemiluminescence.
Chem. Commun., 1999, 1475–1476
1475

sharp evolution of NOx, as reflected in the negative N2
production.
After this initial burst of NOx, the catalyst then becomes
active and highly selective for N2 production, giving ca. 85%
conversion of NH3 and 100% selectivity to N2. There is no
measurable production of NO, NO2, N2O or HCN, and the
nitrogen mass balance from both the mass spectrometry and
NOx analysis is 100% within experimental error. At the same
time as we observe this very high conversion of ammonia, we
see that the conversion of CO is < 1% and there is < 20%
conversion of H2 in the temperature range 600–750 °C.
These data for ammonia oxidation are consistent with an
‘internal selective catalytic reduction’ mechanism which can be
summarised as follows. The Keggin unit oxidises the adsorbed
+
NH3 (trapped as NH4 ) to NOx using labile oxygen from the
anion. The NOx formed is retained briefly as part of the anion,
and reacts with an NH3 molecule to give N2 and H2O. The
Keggin unit then re-oxidises by reaction with gas phase oxygen.
This last step further accounts for the high chemical specificity
of the process as it limits the concentration of active oxygen to
react with CO and/or H2. The rapid turnover between the proton
and NH4+ states also prevents over reduction and collapse of the
Keggin unit through dehydration,8 with the result that the HPW
is stable over the course of 6 h.
The requirement for protonic sites for the reaction was
confirmed by the synthesis and testing of full and partial salts.
The results in Fig. 3 show that, in comparison with the parent
acid, KHPW displays a smaller NOx formation (negative) peak
and a lower activity, which increases gradually with tem-
perature. KPW shows no negative NOx peak and then an
increasing level of N2 production with temperature. For both the
full and partial salts (KPW and KHPW) the activity for
ammonia conversion to N2 is much lower than for the HPA over
most of the temperature range from 600–800 °C. The lower
activity at ca. 600 °C is consistent with the requirement for acid
sites to adsorb NH3.
In conclusion, we have developed a strategy for the selective
conversion of NH3 from biomass-derived gases into N2 by
selective oxidation in competition with a large excess of CO and
H2. The key feature is the use of a catalyst which contains acid
sites, to differentiate ammonia from CO and H2, and redox
properties, to oxidise the adsorbed species. Heteropoly acid
catalysts have been used to demonstrate that the concept is
viable and provides a novel way to overcome the NOx formation
associated with the direct combustion of biogas.
Fig. 3 N2 production profiles for HPW and its salts. (6.0% CO, 4.0% H2,
0.5% O2, 1050 ppm NH3 balance He). Key: (µ) HPW (H3PW12O40), (5)
KHPW (K2.06H0.94PW12O40), (:) KPW (K2.66H0.34PW12O40).
We are pleased to acknowledge the financial support of
Alstom, DTI, and EPSRC through the FORESIGHT Challenge
initiative. Helpful discussions with colleagues at Reading (Mr
Matthieu Amblard), Cranfield University (Mr J. J. Witton,
Professor B. Moss, Mr J. M. Przybylski, and Dr E. Noordally)
and at Alstom (Mr M. Cannon) are gratefully acknowledged.
References
1 M. F. M. Zwinkels, G. M. Eloise Heginuz, B. H. Gregertsen, K. Sjöström
and S. G. Järås, Appl. Catal. A, 1997, 148, 325.
2 L. Lietti, C. Groppi, and C. Ramella, Catal. Lett., 1998, 53, 91.
3 Development of Improved Stable Catalysts and Trace Element Capture
for Hot Gas Cleaning, DTI/ETSU/Clean Coal Power Generation Group,
Project Profile 178, 1996.
4 E. M. Johansson, S. G. Jaras, Catal. Today, 1996, 47, 359.
5 T. Okuhara, N. Mizuno and M. Misono, Adv. Catal., 1996, 41, 113.
6 M. Amblard, R. Burch and B. W. L. Southward, Appl. Catal. B, in
press.
7 M. Amblard, R. Burch and B. W. L. Southward, Clean Catalytic
Combustion of Nitrogen-Bearing Biomass, 4th International Workshop
on Catalytic Combustion, San Diego, 1999.
8 B. W. L. Southward, J. Vaughan and C. T. O’Connor, J. Catal., 1995,
153, 293.
Communication 9/04088I
1476
Chem. Commun., 1999, 1475–1476