The main bottleneck to maintain the long-term stability of the partial nitritation-anammox processes, especially those operated at low temperatures and nitrogen concentrations, is the undesirable development of nitrite oxidizing bacteria (NOB). When this occurs, the punctual addition of compounds with the capacity to specifically inhibit NOB without affecting the process efficiency might be of interest. Sodium azide (NaN3) is an already known NOB inhibitor which at low concentrations does not significantly affect the ammonia oxidizing bacteria (AOB) activity. However, studies about its influence on anammox bacteria are unavailable. For this reason, the objective of the present study was to evaluate the effect of NaN3 on the anammox activity. Three different types of anammox biomass were used: granular biomass comprising AOB and anammox bacteria (G1), anammox enriched granules (G2) and previous anammox granules disaggregated (F1). No inhibitory effect of NaN3 was measured on G1 sludge. However, the anammox activity decreased in the case of G2 and F1. Granular biomass activity was less affected (IC50 90 mg/L, G2) than flocculent one (IC50 5 mg/L, F1). Summing up, not only does the granular structure protect the anammox bacteria from the NaN3 inhibitory effect, but also the AOB act as a barrier decreasing the inhibition.

The anammox process consists of the oxidation of ammonium to nitrogen gas, in anaerobic conditions, using nitrite as electron acceptor and producing small amounts of nitrate (Strous et al. 1998). It was discovered in the 1990s and immediately identified as a promising process allowing the establishment of the completely autotrophic nitrogen removal, with N2 as main product. The required nitrite for this process is produced from the oxidation of half of the incoming ammonium to nitrite by ammonia oxidizing bacteria (AOB) in the so-called partial nitritation (PN) process.

Nowadays, the technologies based on PN-anammox processes have been successfully implemented in more than 100 full scale plants, operated at mesophilic conditions and treating high nitrogen concentration effluents (Lackner et al. 2014). A relevant number of these plants face the problem of nitrate build up, revealing the presence of nitrite oxidizing bacteria (NOB) as the weak point of the PN-anammox system stability (Wang et al. 2015). From previous studies, performed in a stable single stage PN-anammox granular biomass system, the appearance of NOB has been associated with the development of flocs or small granules (Winkler et al. 2011; Morales et al. 2016). However, this observation is not of general application, and NOB wash-out strategies have not been found yet. For this reason, it is crucial to identify the operational conditions for the NOB suppression, as it is also the main bottleneck to extend the application of the anammox based processes to mainstream conditions (nitrogen concentration <70 mg N/L and temperature <25 °C) and to some industrial effluents.

To date, different strategies to induce the NOB depletion have been proposed in PN-anammox systems. One studied action in a one-stage PN-anammox system was the re-inoculation with anammox biomass to improve the nitrite depletion and avoid its use by the NOB. However, this method is expensive, due to the large amounts of biomass needed in a full scale plant, and not permanent, as nitrate concentrations appear again after some time (Wang et al. 2015). Another option relies on the control of the solids retention time (SRT), as anammox bacteria need higher SRT than the aerobic ones; however, it is only feasible if the anammox bacteria are grown as biofilm and the NOB as suspended biomass (Han et al. 2016). Another strategy is based on the control of the dissolved oxygen (DO) concentration at low values, profiting from the fact that at high temperature the oxygen affinity of AOB is lower than that of NOB. But for long-time operation at low temperatures this is not applicable, as this behaviour is inverted (higher oxygen affinity for AOB than NOB) (Ma et al. 2016). This may be explained by the dominant NOB species, since the abundance of Nitrospira-like NOB (k-strategists) increases under DO limited conditions and lower temperature against Nitrobacter-like NOB (r-strategists) (Ma et al. 2016). Recently, new strategies have arisen, such as the use of free ammonia or free nitrous acid NOB inhibitory concentrations (more toxic to NOB than to AOB) or real time control of parameters like pH, ammonia or nitrate concentrations (Ge et al. 2015; Wang et al. 2015). From these studies it is inferred that the stability problem due to NOB development is still far from being solved. Therefore, new strategies such as the use of specific inhibitory compounds can be considered as an alternative in some cases (Wang et al. 2015).

It is known that in general NOB are more sensitive than AOB to the presence of substances such as organic matter, sulphide, hydroxylamine, salts, chlorates, hydrazine or azide (Ge et al. 2015). However, substances like organic matter, salt or sulphides are also known to be inhibitors of the anammox bacteria (Dapena-Mora et al. 2007). Although some of them exert worse effects on the anammox bacteria than on the NOB, others affect preferentially the NOB. For example, hydroxylamine, an intermediate in both nitrification and anammox process, is toxic only for NOB. Wang et al. (2015) proposed the use of this compound (20 mg/L) combined with the SRT control (reduced to 40 days) to restore the stable operation of the PN-anammox and suppress the NOB activity. Another compound, sodium azide (NaN3), has been identified as a specific NOB inhibitor with no detrimental effect on AOB activity at low concentrations (IC50 values of 40 and 0.025 mg NaN3/L for AOB and NOB, respectively) (López-Fiuza et al. 2002). However, the evaluation of its potential effect on the anammox biomass has not been evaluated yet in either suspended or aggregated biomass. Having in mind that anammox bacteria are operated in many cases in the form of granules together with AOB, this aggregation stage can be exploited as a beneficial parameter to resist the presence of the potential NOB inhibitors. It has been observed that the granule matrix acts as a mass transfer barrier that produces lower internal local concentrations of toxic compounds than those in the bulk liquid (Adav et al. 2008).

When the suppression of NOB activity is accomplished via addition of inhibitors, certain amounts of nitrite might be accumulated in the system with the consequent production of N2O gas (undesired due to its global warming effect). It is a byproduct of the nitrification process under aerobic conditions or of an incomplete denitrification carried out by nitrifiers or heterotrophic denitrifiers under anoxic conditions (Campos et al. 2016).

The aim of this study was to determine the NaN3 inhibitory effects on the specific anammox activity (SAA) of flocculent anammox biomass and granular biomass, performing the anammox and/or PN-anammox processes, respectively. The activity restoration capacity was also assessed for the anammox granules. The N2O production in the tests was evaluated.

Origin of biomass

Three types of biomass were used at concentrations of approximately 3 g volatile suspended solids (VSS)/L in the experiments. Granular biomass (G1) was collected from a pilot plant performing the PN-anammox in a single unit. The pilot plant was operated with the ELAN® process treating the reject water from a municipal wastewater treatment plant containing 540–1,045 mg NH4+-N/L at 30 °C (Morales et al. 2015). Granular anammox enriched biomass (G2) was taken from a laboratory reactor, inoculated with biomass from the ELAN® pilot plant, treating a synthetic medium with 60 mg NH4+-N/L and 60 mg NO2-N/L at 30 °C. Finally, previous granular anammox enriched biomass was mechanically disaggregated to obtain flocculent biomass (F1). Previously performed fluorescence in situ hybridization (FISH) analysis revealed that the anammox dominant species in all the samples was Brocadia fulgida (Morales et al. 2015).

Batch activity tests

The SAA was determined in batch tests carried out according to Dapena-Mora et al. (2007). The biomass samples were washed with phosphate buffer. The headspace of the vials, hermetically closed, was flushed with helium gas. The vials were incubated at 30 °C and 150 rpm. After addition of substrates (70 mg N/L of ammonium and nitrite, respectively), and pressure equalization to atmospheric pressure, the overpressure evolution during time was recorded using a differential pressure transducer (0–5 psi, linearity 0.5% of full scale) manufactured by Centerpoint Electronics.

To check the inhibitory effect of sodium azide on the SAA, experiments with eight different concentrations of this compound (in the range from 0 to 100 mg/L) were carried out. The desired amount of sodium azide solution was added previous to closing the vials and flushing them. SAA tests to evaluate the reversibility of the inhibitory effect or the NaN3 were carried out only with G2. In this case, each vial was washed with phosphate buffer by consecutively filling it to the top and draining the supernatant 10 times. Then SAA tests were performed.

Heterotrophic denitrification activity tests were carried out to determine if some of the produced nitrogen in the vials of the SAA tests could originate from the presence of denitrifying activity. A procedure similar to that used in the SAA tests was performed, but only nitrite (70 mg N/L) was added as substrate.

Calculations

The SAA was estimated from the measurement, throughout the time, of the overpressure inside the headspace of the vials (Dapena-Mora et al. 2007). The measured overpressure corresponded to N2 production (composition higher than 99%) when the anammox was the only activity inside the vial (Dapena-Mora et al. 2007). However, in this case gas samples at the end of the SAA test were analysed to determine the N2 production when N2O was simultaneously produced and determine the actual SAA in g N/(g VSS·d).

The inhibitory effect of sodium azide on the anammox activity was expressed as percentage of activity maintained and calculated according to Equation (1):
formula
1
where SAA and SAA0 are the SAA measured in the presence of NaN3 concentrations and corresponding to the control test (maximum SAA), respectively.

The IC50 was determined as the concentration of the NaN3 which led to an activity percentage (%SAA) of 50%.

Analytical methods

At the end of each activity test, the biomass concentration (g VSS/L) was determined according to Standard Methods (APHA-AWWA-WPCF 2005) and gas samples collected from the headspace of the vials were analysed by gas chromatography (GC). The GC system, a Hewlett Packard 5890 Series II instrument, was equipped with a flame ionization detector (thermal conductivity detector) and 80/100 Porapak Q column (2 × 1/8″, Supelco). The mobile phase consisted of helium gas with a flow rate of 16 mL/min, and oven, detector and injector temperatures were 35, 110 and 110 °C, respectively. The average diameter of the granules was determined using a stereomicroscope (Stemi 2000-C, Zeiss) for image acquisition and the software Image ProPlus® for image analysis.

Short-term anammox inhibition by sodium azide

Results from the batch experiments performed with the three types of biomass showed that the inhibitory effect exerted by NaN3 on the maximum SAA is highly dependent on the biomass aggregation state (Figure 1). The PN-anammox granular biomass (G1) showed no inhibitory effect on sodium azide concentrations from 0 to 100 mg/L and its SAA value remained around 0.330 ± 0.051 g N/(g VSS·d) (Table 1). The granular structure (Figure 2(a)) and the presence of AOB in the outer layer of the granules (Morales et al. 2015) presumably helped to mitigate the toxic effect of NaN3 on the anammox bacteria, located in the core of the granule.
Table 1

Characteristics of sludge samples and sodium azide effects

SampleType of biomassDiameter (mm)SAA0 (g N/g VSS·d)NaN3 inhibitionIC50 (mg/L)
G1 Granular PN-anammox 2.0 0.330 ± 0.051 No – 
G2 Granular anammox 2.4 0.366 ± 0.007 Yes 90 
F1 Disaggregated anammox – 0.366 ± 0.007 Yes 
SampleType of biomassDiameter (mm)SAA0 (g N/g VSS·d)NaN3 inhibitionIC50 (mg/L)
G1 Granular PN-anammox 2.0 0.330 ± 0.051 No – 
G2 Granular anammox 2.4 0.366 ± 0.007 Yes 90 
F1 Disaggregated anammox – 0.366 ± 0.007 Yes 
Figure 1

Effect of sodium azide at concentrations between 0 and 100 mg/L over the maximum SAA of: PN-anammox granules (G1) (●), enriched anammox granules (G2) (●) and anammox flocs (F1) (○).

Figure 1

Effect of sodium azide at concentrations between 0 and 100 mg/L over the maximum SAA of: PN-anammox granules (G1) (●), enriched anammox granules (G2) (●) and anammox flocs (F1) (○).

Figure 2

Images of (a) PN-anammox granules (G1), (b) enriched anammox granules (G2) and (c) anammox flocs (F1). The size bar represents 2 mm.

Figure 2

Images of (a) PN-anammox granules (G1), (b) enriched anammox granules (G2) and (c) anammox flocs (F1). The size bar represents 2 mm.

However, when the enriched anammox biomass (G2 and F1) was evaluated, the inhibitory effect of the NaN3 was detected even at its lowest concentration. Mechanical disintegration did not affect the SAA measured in the absence of the inhibitor, which was of 0.366 ± 0.007 g N/(g VSS·d) (Table 1). When both types of anammox biomass were exposed to increasing concentrations of NaN3, a similar inhibition pattern was observed. The NaN3 inhibitory effect increased faster at low inhibitor concentrations (up to 10 mg/L), showing a sharp decrease of anammox activity. The SAA diminution was less relevant for higher sodium azide concentrations. These assays, with entire (Figure 2(b)) and disintegrated (Figure 2(c)) granules, showed the importance of the aggregation state of the biomass, as the percentage of remaining activity is more than twice as small for F1. The estimated value of IC50 with NaN3 was around 90 mg/L for the anammox granular biomass (G2) and 5 mg/L for the disintegrated granules (F1). The granular structure presumably originates a gradient of the sodium azide (due to its diffusivity) throughout the granule that led to a much lower inhibition (Crank 1975). This effect can also be observed in unsteady state.

The PN-anammox granules (G1) presented a smaller average diameter than the anammox enriched granules (G2) with values of 2.0 mm and 2.4 mm, respectively (Figure 3). There were some bigger particles in G1 (up to 7.5 mm), but a wider size distribution was observed and huge amounts of small particles were present. More than 60% of the total number corresponded to particles with a size below 0.5 mm, while in the case of G2 about 50% were granules with sizes below 1.5 mm (Figure 3). Although the mass transfer resistance is higher in the bigger granules, the present results show higher inhibitory effect of NaN3 in G2. Therefore, presumably not only the granular structure but also the presence of AOB located in the external layers of the granules could act as a physical barrier to avoid the direct contact of the anammox bacteria with this toxic compound. Moreover, the presence of extracellular polymeric substances has been thought to play an important role in the resilience of the granular biomass against toxic compounds (Chen et al. 2017).
Figure 3

Size distributions of (a) PN-anammox granules (G1) and (b) enriched anammox granules (G2) in percentage of volume with respect to the total volume. Grey line indicates the accumulative frequency of each class of particles.

Figure 3

Size distributions of (a) PN-anammox granules (G1) and (b) enriched anammox granules (G2) in percentage of volume with respect to the total volume. Grey line indicates the accumulative frequency of each class of particles.

López-Fiuza et al. (2002) has previously reported the IC50 value of 16.3 μg N3/L for NOB, equivalent to 24.7 μg NaN3/L, which is a value much lower than the one obtained for the anammox biomass in this study. Moreover, the sodium azide concentration reported to totally inhibit the NOB (in batch tests) is 1.5 mg/L according to Guisasola et al. (2005). This concentration caused, in the present study, the loss of around 5% of the SAA of anammox granules (G2) and 20% of flocculent anammox sludge (F1), respectively.

Production of N2O

The composition of the gas phase in the SAA experiments was characterized. Although N2 was the most abundant gas, with a percentage between 98 and 99%, small amounts of CO2 and N2O were also detected. In all cases, the higher the sodium azide concentration applied, the higher the N2O concentration measured (Figure 4), while the composition in terms of CO2 was maintained practically constant (around 0.4% for G1 and G2 and 0.8% in the case of F1). Since the anammox bacteria metabolism does not produce N2O (Kartal et al. 2007), further research about the reasons for these emissions is needed. Possible causes of the nitric oxide origin could be either the presence of heterotrophic denitrification, which used the products from the decayed biomass as organic carbon, or the occurrence of the nitrifying denitrification process (Ali et al. 2013). Moreover, N2O might also be chemically produced by the reaction of sodium azide and nitrite in acid media to produce nitrogen and nitric oxide (Stedman 1959). However, a vial was incubated with the substrates (ammonia and nitrite) and phosphate buffer to test the gas production from the chemical reaction and no N2O was detected.
Figure 4

Effect of sodium azide at concentrations between 0 and 100 mg/L over the % of produced N2O in the gas phase of: PN-anammox granules (G1) (●), enriched anammox granules (G2) (●) and anammox flocs (F1) (○).

Figure 4

Effect of sodium azide at concentrations between 0 and 100 mg/L over the % of produced N2O in the gas phase of: PN-anammox granules (G1) (●), enriched anammox granules (G2) (●) and anammox flocs (F1) (○).

Despite the unknown origin of this gas, again the aggregation state of the biomass has a strong influence. Comparing the results from the experiments with G2 and F1, the percentage of N2O measured in the vials with the former (granules) was half of the amount produced with the latter (flocs). As the flocculent biomass (F1) was also more inhibited than the granular one (G2), the N2O production might be related to inhibition of the anammox bacteria activity, associated with their death, together with the production of another biological activity yet to be identified. Okabe et al. (2011) also found greater N2O production in anammox granules when their activity is inhibited (in that case due to the presence of HNO2 due to the pH decrease). These authors concluded that the heterotrophic denitrification was the main responsible process for N2O production due to the lack of enough organic matter to complete the process (Okabe et al. 2011). This might also be the cause in the present study, since the only organic matter source present in the experiments comes from the decayed biomass.

In the case of the biomass G1, the SAA remained constant at all tested NaN3 concentrations (Figure 1), but the N2O percentage in the gas phase was similar to that from the flocculent biomass. The presence of the AOB that might carry out the nitrifying denitrification process might be responsible for this observation in G1. However, in F1 the AOB population is not expected to be present in significant amounts.

In order to elucidate the possible heterotrophic denitrification role in the N2O production, a denitrification test adding only nitrite as substrate was carried out. The test was performed with the biomass G2 in the most unfavourable case (100 mg NaN3/L). The denitrification activity measured was around 0.020 ± 0.008 g N/(g VSS·d) and the gas composition was considerably different from that obtained in the SAA test (Table 2). The N2O percentage obtained in the denitrification test was almost six times higher than the one measured in the anammox test, showing that presumably the heterotrophic denitrification activity may be the main process responsible for N2O emissions. It should be pointed out that the gas phase is not greatly enriched in N2, since it only accounts for 96%.

Table 2

Gas composition of the headspace in the SAA test and denitrification test with granular (G2) and flocculent (F1) anammox biomass in the presence of 100 mg sodium azide/L

BiomassTestN2 (%)CO2 (%)N2O (%)
F1 SAA 98.21 ± 0.17 0.72 ± 0.05 1.07 ± 0.22 
G2 SAA 99.08 ± 0.03 0.40 ± 0.02 0.51 ± 0.01 
G2 SAAa 99.64 ± 0.01 0.19 ± 0.01 0.17 ± 0.01 
G2 Denitrification 96.02 ± 0.39 1.04 ± 0.18 2.94 ± 0.22 
BiomassTestN2 (%)CO2 (%)N2O (%)
F1 SAA 98.21 ± 0.17 0.72 ± 0.05 1.07 ± 0.22 
G2 SAA 99.08 ± 0.03 0.40 ± 0.02 0.51 ± 0.01 
G2 SAAa 99.64 ± 0.01 0.19 ± 0.01 0.17 ± 0.01 
G2 Denitrification 96.02 ± 0.39 1.04 ± 0.18 2.94 ± 0.22 

aRecovery SAA test, without NaN3 after the exposure of the biomass to 100 mg/L.

Based on the results obtained in this study, it is not possible to conclude a clear explanation to identify the N2O emissions source. But, it may be hypothesized that for G1 the main cause is the nitrifying denitrification and for G2 and F1 the heterotrophic denitrification. Presumably, the higher the inhibition of the anammox activity, the higher the biomass decay, and the more important the heterotrophic denitrification becomes for the gas production and therefore the N2O emissions. More biomass decay may take place in the F1, with the consequent higher heterotrophic denitrification, as can be seen by the CO2 production (almost double) and higher N2O emissions (Table 2).

Restoration of anammox activity

The possible reversibility of the inhibitory effect of NaN3 on the anammox activity was evaluated in batch activity tests with anammox granules (G2) previously exposed to the toxic substance. Results indicated that the activity of the anammox biomass was almost completely restored after exposure to sodium azide concentrations lower than 20 mg/L for 7 hours and after a washing step. However, in the case of exposure to higher concentrations, the activity was not completely recovered after biomass washing, and about 30% of inhibition was detected in the biomass exposed to 100 mg/L of NaN3. The produced gas composition was also determined in these activity recovery experiments, and the N2O concentration was negligible in experiments performed with biomass previously exposed to concentrations up to 20 mg/L of NaN3. In all cases, it was lower than in the experiments with sodium azide. For example, the percentage of N2O in the test with G2 and 100 mg/L was 0.5%, and after washing the biomass the obtained percentage was 0.2% (Table 2). However, after washing it was not possible to guarantee that sodium azide was fully removed from the biomass, especially in the experiments with high concentrations, and further research is required to evaluate the effects of the long-term exposure.

In case these future long-term inhibition experiments provide similar results to those obtained in the present study, sodium azide could be an option to be used to suppress the NOB activity not only in the one-stage PN-anammox process (no inhibition found) but also in the two-stage configuration using granular biomass in the anammox reactor. G2 at the concentration needed to achieve the complete inhibition of NOB (1.5 mg/L) (Guisasola et al. 2005) will lose around 5% of its SAA, and this inhibition would be reversible, meaning that to operate the reactor under these conditions is feasible.

To sum up, the biomass grown as granules together with the presence of AOB in the external layers can act as a physical barrier for the anammox bacteria to mitigate the toxic effects of chemical compounds such as azides. The sodium azide did not affect the PN-anammox granular biomass. The enriched anammox biomass was inhibited by NaN3 at all the concentrations tested, but the effect on anammox granules was considerably lower than on anammox flocculent biomass, resulting in an IC50 of 90 mg/L for granular biomass and IC50 5 mg/L for the flocculent one. For enriched anammox biomass (G2 and F1), N2O emissions are related to the extension of SAA inhibition, being lower when the inhibition is lower. In the case of PN-anammox biomass, the nitrifying denitrification is the most probable source of N2O, and it increased at higher sodium azide concentrations.

The authors want to thank the Pioneer_STP (ID 199) project funded by the WaterWorks2014 Cofunded Call (Water JPI/Horizon 2020). This work was also funded by the Spanish Government through FISHPOL (CTQ2014-55021-R) and GRANDSEA (CTM2014-55397-JIN) projects co-funded by FEDER. The authors from the USC belong to CRETUS (AGRUP2015/02) and the Galician Competitive Research Group (GRC 2013-032), programs co-funded by FEDER. The authors want to thank FCC Aqualia for the ELAN® biomass samples.

Adav
S. S.
Lee
D.-J.
Show
K.-Y.
Tay
J.-H.
2008
Aerobic granular sludge: recent advances
.
Biotechnology Advances
26
(
5
),
411
423
.
APHA-AWWA-WPCF
2005
Standard Methods for the Examination of Water and Wastewater
.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
Campos
J. L.
Valenzuela-Heredia
D.
Pedrouso
A.
Val del Río
A.
Belmonte
M.
Mosquera-Corral
A.
2016
Greenhouse gases emissions from wastewater treatment plants: minimization, treatment, and prevention
.
Journal of Chemistry
2016
,
12
.
doi: 10.1155/2016/3796352
.
Chen
Q.-Q.
Sun
F.-Q.
Guo
Q.
Shen
Y.-Y.
Zhu
W.-Q.
Jin
R.-C.
2017
Process stability in an anammox UASB reactor with individual and combined thiocyanate and hydraulic shocks
.
Separation and Purification Technology
173
,
165
173
.
Crank
J.
1975
The Mathematics of Diffusion
.
Clarendon Press
,
Oxford
,
UK
.
Dapena-Mora
A.
Fernandez
I.
Campos
J. L.
Mosquera-Corral
A.
Mendez
R.
Jetten
M. S. M.
2007
Evaluation of activity and inhibition effects on Anammox process by batch tests based on the nitrogen gas production
.
Enzyme and Microbial Technology
40
(
4
),
859
865
.
Guisasola
A.
Jubany
I.
Baeza
J. A.
Carrera
J.
Lafuente
J.
2005
Respirometric estimation of the oxygen affinity constants for biological ammonium and nitrite oxidation
.
Journal of Chemical Technology & Biotechnology
80
(
4
),
388
396
.
Kartal
B.
Kuypers
M. M.
Lavik
G.
Schalk
J.
Op den Camp
H. J.
Jetten
M. S.
Strous
M.
2007
Anammox bacteria disguised as denitrifiers: nitrate reduction to dinitrogen gas via nitrite and ammonium
.
Environmental Microbiology
9
(
3
),
635
642
.
Lackner
S.
Gilbert
E. M.
Vlaeminck
S. E.
Joss
A.
Horn
H.
van Loosdrecht
M. C. M.
2014
Full-scale partial nitritation/anammox experiences – an application survey
.
Water Research
55
,
292
303
.
López-Fiuza
J.
Buys
B.
Mosquera-Corral
A.
Omil
F.
Méndez
R.
2002
Toxic effects exerted on methanogenic, nitrifying and denitrifying bacteria by chemicals used in a milk analysis laboratory
.
Enzyme and Microbial Technology
31
(
7
),
976
985
.
Ma
B.
Wang
S.
Cao
S.
Miao
Y.
Jia
F.
Du
R.
Peng
Y.
2016
Biological nitrogen removal from sewage via anammox: recent advances
.
Bioresource Technology
200
,
981
990
.
Morales
N.
Val del Rio
A.
Vazquez-Padin
J. R.
Gutierrez
R.
Fernandez-Gonzalez
R.
Icaran
P.
Rogalla
F.
Campos
J. L.
Mendez
R.
Mosquera-Corral
A.
2015
Influence of dissolved oxygen concentration on the start-up of the anammox-based process: ELAN(R)
.
Water Science and Technology
72
(
4
),
520
527
.
Morales
N.
Val del Río
Á.
Vázquez-Padín
J. R.
Méndez
R.
Campos
J. L.
Mosquera-Corral
A.
2016
The granular biomass properties and the acclimation period affect the partial nitritation/anammox process stability at a low temperature and ammonium concentration
.
Process Biochemistry
51
(
12
),
2134
2142
.
Stedman
G.
1959
591. Mechanism of the azide-nitrite reaction. Part II
.
Journal of the Chemical Society (Resumed)
,
2949
2954
.
DOI: 10.1039/JR9590002949
Strous
M.
Heijnen
J. J.
Kuenen
G. J.
Jetten
M. M. S.
1998
The sequencing batch reactor as a powerful tool for the study of slowly growing anaerobic ammonium-oxidizing microorganisms
.
Applied Microbiology and Biotechnology
50
(
5
),
589
596
.
Winkler
M. K. H.
Kleerebezem
R.
Kuenen
J. G.
Yang
J.
van Loosdrecht
M. C. M.
2011
Segregation of biomass in cyclic anaerobic/aerobic granular sludge allows the enrichment of anaerobic ammonium oxidizing bacteria at low temperatures
.
Environmental Science & Technology
45
(
17
),
7330
7337
.