Nume ical in es iga ion on he combus ion cha ac e is ics o p emixed
NH
3
-ai lames using gliding a c plasma
Ziyu Wang , B. A a ind
*
, Syed Mash uk , Agus in Vale a-Medina
College o Physical Sciences and Enginee ing, Ca di Uni e si y, Wales, CF24 3AA, UK
ARTICLE INFO
Handling Edi o : D . Paul Williams
Keywo ds:
Ammonia
Non-equilib ium plasma
Ca bon- ee combus ion
Nume ical model
Emission
ABSTRACT
In his s udy, a nume ical model is de eloped o p edic he combus ion cha ac e is ics o gliding a c plasma
(GAP) assis ed ammonia (NH
3
)-ai mix u e, in eg a ing ZDPlasKin and Chemkin. To he bes o he au ho s’
knowledge, his is he i s alida ed model capable o accu a ely p edic ing NO emissions om GAP-assis ed
NH
3
-ai combus ion. Ini ially, h ee well-known plasma mechanisms a e e alua ed agains non- eac ing GAP
expe imen s o assess hei e ec i eness in modelling NH
3
-ai plasma chemis y. The mos accu a e mechanism is
hen coupled wi h an op imized combus ion mechanism o imp o e NO p edic ion accu acy. The esul s indica e
ha NH
2
adical o ma ion is enhanced by app oxima ely 7 % a a educed elec ic ield o 30 Td, playing a
c ucial ole in NO educ ion. Addi ionally, NH
2
is p ima ily gene a ed h ough wo key eac ions: O(
1
D) +NH
3
→ OH +NH
2
and N
2
(A) +NH
3
→ NH
2
+N
2
+H, occu ing be o e combus ion. Fu he mo e, inc easing plasma
powe signi ican ly accele a es NO consump ion by p omo ing he o ma ion o exci ed NH
3
s a es (NH
3
(e
1
),
NH
3
(e
2
)), which enhance NH
2
and NH adical p oduc ion. Sensi i i y analysis e eals ha NH
2
exhibi s a 52.1 %
sensi i i y o he eac ion N(
2
D) +NH
3
→ NH
2
+H +N
2
a 90 Td, highligh ing i s dominan ole in NO
educ ion.
1. In oduc ion
Ammonia (NH
3
) is gaining signi ican a en ion as a p omising g een
uel due o i s ca bon- ee combus ion, high ene gy densi y compa able
o ossil uels (22.5 MJ/kg) [1], and cos -e ec i e s o age, posi ioning i
as a iable solu ion o he cu en ene gy c isis [2]. Despi e hese ad-
an ages, NH
3
has high NO
x
emissions due o i s inhe en ly low lame
empe a u e, low lamina combus ion a e, and low lammabili y [3]. To
ci cum en hese issues, co i ing o NH
3
wi h highly eac i e uels [4],
such as me hane [5] and hyd ogen [6], s aged combus ion [7], humid-
i ica ion [8], plasma assis ed combus ion [9] e c. a e commonly adop ed
echniques.
Among hese echniques, plasma assis ed combus ion [10], is ela-
i ely new s a egy and, wi h g ea po en ial o enhance he lame s a-
bili y and o he combus ion cha ac e is ics [9,11]. Al hough i s impac
on NH
3
combus ion is compa a i ely less explo ed, ecen s udies
[12–14] ha e shown ha plasma can signi ican ly add ess he d aw-
backs o NH
3
as a u u e uel [15]. Howe e , plasma assis ed combus ion
o NH
3
does no al e he p ima y NO gene a ion pa hway, whe e HNO is
in ol ed in 70 % o he eac ion [16]. This sugges s ha a ca e ul
analysis o he eac ion pa hways in ol ed in plasma-assis ed combus-
ion o NH
3
is essen ial o unde s and NO
x
consump ion h ough he
p oduc ion o NH adicals [17]. The dielec ic ba ie discha ge (DBD)
assis ed NH
3
combus ion on he p emixed NH
3
-H
2
-ai , e ealed ha NO
x
emissions inc eases wi h he plasma in lamina / u bulen lames [18].
While Ju e al. [14] ound ha NO has been educed signi ican ly using
gliding a c plasmas (GAP), by p omo ing he NO consump ion eac ions.
In plasma assis ed combus ion modelling, many esea che s ha e
made a conside able con ibu ion in DBD and nanosecond epe i i ely
pulsed discha ges (NRP) [16–22]. Faingold e al. [19] conduc ed a
de ailed pa ame ic s udies on he e ec o pulse epe i ion equency,
numbe o pluses o NRP on he igni ion delay ime (IDT) cha ac e is ics
o NH
3
-O
2
-He mix u es. Shahsa a i e al. [20] in es iga ed he impac o
NRP on lame cha ac e is ics. They ound ha he pulse ene gy densi y
inc easing in he ange o 0–20 mJ/cm
3
, a a gi en educed elec ic ield,
dec eases he IDT. In 2019, Mao e al. p oposed a nume ical model o
p edic he Igni ion enhancemen o CH
4
-O
2
-He mix u es unde NRP and
DC discha ges [21]. Recen ly, Zhong e al. de eloped a obus kine ic
model o plasma combus ion o NH
3
-O
2
-N
2
ha a e alida ed wi h he
expe imen s [22]. The same plasma mechanism was used by Mao e al.
* Co esponding au ho .
E-mail add ess: [email p o ec ed] (B. A a ind).
Con en s lis s a ailable a ScienceDi ec
Jou nal o he Ene gy Ins i u e
jou nal homepage: www.else ie .com/loca e/joei
h ps://doi.o g/10.1016/j.joei.2025.102314
Recei ed 6 July 2025; Recei ed in e ised o m 16 Sep embe 2025; Accep ed 29 Sep embe 2025
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
A ailable online 30 Sep embe 2025
1743-9671/© 2025 The Au ho s. Published by Else ie L d on behal o The Ene gy Ins i u e. This is an open access a icle unde he CC BY license
( h p://c ea i ecommons.o g/licenses/by/4.0/ ).
[23] o DBD-p omo ed igni ion delay o NH
3
-ai and in he analysis o
NO o ma ion mechanisms. The IDT and NOx emission mechanisms o
NH
3
-ai by NRP we e simila ly in es iga ed by Taneja e al. [24].
Shahsa a i e al. [25] cons uc ed a speci ic mechanism o he ac ion o
NRP and hyd ogen on he combus ion o ammonia, e i ying he change
in empe a u e unde speci ic condi ions.
Al hough nume ous expe imen al and nume ical s udies [19,20,
22–27] ha e in es iga ed NRP and DBD o NH
3
igni ion and NO o -
ma ion, esea ch on GAP o ammonia combus ion has la gely emained
expe imen al [12,14,18,28]. Ne e heless, Compa ed wi h DBD and
NRP, he o a ing gliding a c plasma p o ides b oad olume ic
co e age and can be di ec ly inco po a ed in o he combus ion zone,
making i highly p omising o enhancing combus ion. Meanwhile, GAP
also ea u es a combina ion o high ene gy densi y and he mal e ec s,
enabling con inuous and s able discha ge ha e ec i ely p omo es
chemical eac ions in combus ion. Recen s udies ha e shown ha GAP
plasma combus ion signi ican ly educes NO
x
emissions, d awing
inc easing a en ion om he esea ch communi y [14,28,29]. Howe e ,
he unde lying mechanisms o NO educ ion in GAP emain unclea , and
exis ing expe imen al esul s lack comp ehensi e in e p e a ion and
analysis.
In his wo k, o he i s ime, nume ical model is de eloped o s udy
he GAP-assis ed combus ion cha ac e is ics o NH
3
-ai mix u e.
Ini ially, di e en plasma mechanisms a e compa ed, he mos sui able
one o GAP is selec ed based on expe imen al alida ion. This is ol-
lowed by he de elopmen o a combus ion mechanism in eg a ed wi h
he plasma mechanism. Finally, he mechanism behind he NO emission
educ ion unde he in luence o GAP is analysed in de ail.
2. Nume ical me hods and model alida ion
In o de o alida e he nume ical me hod, GAP expe imen s a e
conduc ed o comp ehend he e ec o GAP on he NH
3
con e sion a e
o non- eac ing cases. The de ails o he expe imen se up a e shown in
he Supplemen a y Ma e ial Fig. S1. Di e en a ios o NH
3
-ai a e
injec ed om he ube, and he gas mix u e is ionized by an a c o med
by he high- ol age elec ode. The “ esh gas” is aken abo e he elec-
ode and he olume ac ion o NH
3
is measu ed using Agilen 990
mic oGC. The chemiluminescence spec um o he discha ge a he
bu ne ’s exi and 10 cm om i s cen al axis was eco ded using a
lexible A aSpec-ULS spec ome e , whe e he s ay ligh (0.19–1.0 %)
ep esen ed he main sou ce o e o . This me hod o de e mining
elec on empe a u e based on he H-
α
/H-β spec al line in ensi y a io
has ela i e e o s o only ±0.017 % [30]. The e o e, he po en ial
o e all unce ain y is es ima ed o below 1.017 %. The ypical empo al
e olu ion o he ol age and cu en wa e o m o GAP in he p esen
s udy is shown in he Supplemen a y Ma e ial Fig. S2. As e iden om
Fig. S2 ha ol age shows a pe iodic jagged shape, and cu en spikes
a e obse ed a each b eakdown a he smalles gap. This b eakdown
ype GAP is obse ed o many plasma assis ed combus ion s udies [28].
In GAP, bo h spa k- ype discha ge and glow- ype discha ge a e obse ed
[12], In he p esen s udy, cu en eaches abo e 1 A and below 0.5 A a e
e med as spa k and glow discha ges.
2.1. Nume ical me hod
This s udy adop s a nume ical me hodology simila o ha p oposed
by C ispim e al. o modelling he GAP [31]. The simula ion is di ided
in o wo domains, as shown in Fig. 1. The i s domain, 0–
1
, ep esen s
he discha ge leng h o esidence ime. Du ing his pe iod, he a c ini-
ia es a he minimum gap, peaks in cu en , p opaga es helically,
elonga es, and e en ually b eaks as he cu en d ops o ze o. The
minimum clea ance be ween he conical elec ode and he ube wall is
2.5 mm, wi h a cha ging leng h o 14 mm om his loca ion o he apex.
This discha ge ime is calcula ed om he ol age-cu en cu e. The
second domain,
1
–
2
, is e med as he a e glow discha ge. The
discha ge powe is main aining be ween 130 and 150 W, and he e-
quency o plasma is 11 kHz. The discha ge ime is be ween 4 and 7 ms,
depends on he gas low a e. The plasma chemis y in he a e glow
egion is signi ican ly di e en om ha o he plasma glow [23].
Howe e , he a e glow e ains mos o he plasma p ope ies.
To simpli y he calcula ions, he educed elec ic ield and elec on
densi y a e es ima ed as pe iod a e ages using he ol age-cu en
cha ac e is ic cu es. This s udy was ca ied ou using a dual-swi l
plasma-assis ed bu ne a CEAT, Ca di Uni e si y. Fu he de ails o
his bu ne a e a ailable in Wang e al. [29] and A a ind e al. [32]. In
he ioniza ion egion, he ini ial gas mix u e consis s o NH
3
/N
2
/O
2
. The
elec on densi y is calcula ed using Eq. (1) [33]:
Ne =J
a
/e
μ
e
E (1)
In Eq. (1), Ne is he cu en densi y, J
a
is he a io o cu en o c oss-
sec ional a ea, A/cm
2
. The elec on mobili y (
μ
e
) is calcula ed om
BOLSIG+, de i ed om collision in e ace da a wi h ini ial E/N in-
ensi y, di ided by ini ial gas densi y, cm
2
/Vs. E is elec ic ield, V/cm.
In his s udy, he cu en cu e p o ides he da a be ween 0.1 and
0.11 A, he a e age elec on conduc ion cu en densi y was calcula ed
o 1–1.16 A/m
−2
, he elec ic ield (E) wi hin he discha ge egion
measu ed 2.1 kV/cm, he elec on mobili y (
μ
e) is calcula ed a 500
cm
2
/V. The elec on densi y a e aged ac oss he conical elec ode was
es ima ed o be 5.7 ×10
12
- 6.4 ×10
12
cm
−3
.
The E/N is de e mined om he expe imen al ol age, elec ode gap
leng h and gas densi y [28]. The ini ial gas densi y is es ima ed by ini ial
p essu e and empe a u e, 2.4 ×10
19
cm
−3
. The ol age is es ima ed a
1.8–4.5 kV. In his s udy, he E/N is 30–90 Td. This ange o educed
elec ic ield s eng h is simila o ha epo ed by Dong e al. [34].
The ini ial elec on empe a u es we e p elimina y de e mined based
on spec al measu emen s. Spec al measu emen s a di e en mix u e
a ios a e shown in Fig. 2. The elec on empe a u e o he plasma is
es ima ed using Eq. (2) based on he spec al cha ac e is ics obse ed in
expe imen s [30]. The elec on empe a u e is used as an ini ial
pa ame e in o de o assume Maxwellian ene gy dis ibu ion o elec-
ons; he alue o educed ield is calcula ed in his case o sa is y
elec on ene gy balance. This es ima ion can enhance he nume ical
me hod o di e en NH
3
-ai a ios.
In (I
32
/I
42
) =In (
υ
32
/
υ
42
) +1.08+h(
υ
42
-
υ
32
)/ (kT
e
) (2)
Whe e T
e
is he elec on empe a u e in Kel in. I
32
and I
42
a e he in-
ensi y o H-
α
(486.1 nm), H-β(656.3 nm).
υ
32
and
υ
42
a e equencies o
adia ion ini ia ed in nm. h is he Planck cons an in J⋅s. k is Bol zmann
cons an , J/K. The sys em is assumed o be in local he mal equilib ium
when he elec on empe a u e o he plasma anges be ween 4000 and
64,000 K. The elec on empe a u e is used as an ini ial pa ame e in
Fig. 1. Theo y o gliding a c plasma.
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
2
o de o assume Maxwellian ene gy dis ibu ion o elec ons; he alue
o educed ield is calcula ed in his case o sa is y elec on ene gy
balance.
As a ansien p ope y, he mix u e densi y (
ρ
mix) dynamically ad-
jus s o eal- ime composi ion a ia ions, compu ed h ough he
componen -based o mula ion in Eq. (3).
ρ
mix =1
NA∑
i
Mi⋅ni(3)
Whe e i ep esen s di e en species, [NH
3
, N
2
, O
2
…]. M
i
deno es he
mola mass o species i. The numbe densi y o each componen is ep-
esen ed by n
i
. N
A
is A ogad o’s cons an , 6.022 ×10
23
mol
−1
.
The GAP eac o exhibi s subs an ially highe gas empe a u es,
obse ed by Dong e al. [34]. Compa ed o con en ional DBD eac o s,
ep esen ing a undamen al ope a ional di e ence be ween hese
plasma sys ems. Consequen ly, he GAP gene a ed he mal e ec s on gas
mix u es should be explici ly conside ed in he analysis. When i aken
in o accoun , he gas empe a u e (T
gas
) can be de e mined h ough he
hea ans e equa ion (4) unde he physical condi ion o adiaba ic
equidis an app oxima ion [35].
Ngas
γ−1
dTgas
d =∑
imax
i=1
±δ
ε
i⋅Ri+Pelas ⋅[Ne] + Pex (4)
λmix =1
ρ
mix ∑
i
(λi⋅Ci)(5)
cp,mix =∑
i
Ni⋅γi
γi−1(6)
Pex =Nu⋅λ⋅(Tgas −Twall )
Rou 2(7)
∑imax
i=1±δ
ε
i⋅Ri ep esen s he hea o chemical eac ions; Pelas co e-
sponds o Joule hea ing induced by elec on cu en , associa ed wi h
elas ic elec on-neu al collisions ( his e m is compu ed using he
BOLSIG +sol e . Pex is a use -de ined hea sou ce ha can be a bi a ily
speci ied in equa ion (7). The symbol λ is used o he gas-phase he mal
conduc i i y. c
p
deno es he speci ic hea capaci y a cons an p essu e.
N
i
and γ
i
a e he densi y and speci ic hea a io o componen i,
espec i ely. C
ᵢ
ep esen s he concen a ion o he i- h gas species
(whe e i is de ined consis en wi h Equa ion (3)). The Nussel numbe is
ixed a Nu =8 o his analysis. The e ec i e he mal conduc i i y (λ) o
he gas mix u e is calcula ed as a weigh ed a e age o he cons i uen
species (NH
3
, O
2
, and N
2
…) based on hei espec i e concen a ions.
The eac o wall empe a u e (T
wall
) is held cons an a he ini ial eac-
ion empe a u e h oughou he simula ion. The sys em geome y is
cha ac e ized by eac o adius R
ou
.
2.2. Model alida ion
The me hod o coupling ZDPlasKin [35] wi h Chemkin [36] has been
gene ally alida ed by nume ous s udies, demons a ing i s e ec i eness
in accu a ely simula ing plasma-assis ed combus ion p ocesses,
including he e olu ion o chemical species and changes in gas em-
pe a u e [22].
The plasma mechanisms o NH
3
-ai mix u es unde plasma condi-
ions ha e also been he subjec o se e al s udies [23–25]. Taneja e al.
[24] assembled a eac ion mechanism comp ising 53 species and 383
eac ions o NRP. Shahsa a i e al. [25] assembled a plasma eac ion
mechanism comp ising 61 species and 790 eac ions. A plasma kine ic
mechanism was cons uc ed by in eg a ing he NH
3
-O
2
-He eac ion
mechanism p oposed by Faingold e al. [19]. Mao e al. [23] p o ided a
plasma mechanism which includes 77 species and 894 eac ions. A
plasma mechanism in ol ing NH
3
-O
2
-N
2
was u ilized in DBD.
Al hough se e al mechanisms ha e been de eloped o NRP [24,25],
no mechanism is cu en ly a ailable o GAP-assis ed combus ion o
NH
3
. As an ini ial s ep owa d iden i ying he mos sui able mechanism
o GAP, he pe o mance o models p oposed by Mao e al. [23], Taneja
e al. [24], and Shahsa a i e al. [25]is compa ed wi h he p esen
expe imen al esul s on GAP-assis ed NH
3
con e sion in non- eac ing
cases. The NH
3
con e sion a e was calcula ed om he ini ial NH
3
olume ac ion and he sampled ammonia olume ac ion [37].
I is e iden om Fig. 3 ha mechanism p oposed by Mao e al. [23]
demons a es good ag eemen wi h he p esen expe imen s and is
he e o e selec ed o he subsequen simula ions. Following he selec-
ion o a sui able mechanism o GAP, he plasma mechanism o Mao
e al. [23] is combined hen wi h he NH
3
combus ion mechanism p o-
posed by Alnasi e al. [38,39].
Du ing he coupling p ocess, Alnasi ’s mechanism se es as he hos ,
while Mao e al.’s mechanism unc ions as he dono . All plasma- ela ed
eac ions om Mao’s mechanism a e in eg a ed in o Alnasi ’s mecha-
nism, wi h o e lapping eac ions and species ca e ully emo ed. The
equi alence a io is de ined as ollows:
Φ=(F/O) eal
(F/O)s oichiome ic
(8)
Fig. 2. Spec a o plasma a di e en NH
3
a ios.
Fig. 3. Valida ion o NH
3
plasma mechanism sui able o GAP.
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
3
He e, (F/O) eal is he ac ual mola o uel o oxidize , (F/O)s oichiome ic is
he mola o uel o oxidize unde s oichiome ic condi ions
Then is he pe o mance o he coupled mechanism on NO
x
p edic-
ion is compa ed wi h he a ious GAP assis ed NH
3
combus ion
expe imen al esul s o Ju e al. [14] and Tang e al. [28], as shown in
Fig. 4. No ably, he expe imen al da a a e all based on NO
x
s udies o
p emixed NH
3
–ai lames in a plasma-assis ed swi l bu ne . Fo
plasma-assis ed ammonia/ai lames, he NO peak appea s a ound ϕ =
0.8 and g adually dec eases as he equi alence a io inc eases. Unde
uel- ich condi ions, he ob ained NO is app oxima ely ze o. The nu-
me ical simula ion esul s co espond well wi h his end. Expe imen al
esul s a e well cap u ed by he Mao–Alnasi mechanism. The de ia ion
o he mechanism om he expe imen is less han 5 % When ϕ is
0.7–0.82 and 0.95–1.0.
3. Resul and discussion
3.1. E ec o GAP on densi y o main species
This sec ion discusses he a ia ion in gas densi y and empe a u e o
he main componen s o p emixed NH
3
-ai gas unde GAP a E/N =30
Td. This E/N alue is selec ed o analysis as i alls wi hin he es ima ed
ange o he GAP expe imen s.
The ans o ma ion o gas species du ing he gliding a c plasma
ac ua ion can be di ided in o wo dis inc phases as shown in Fig. 5(a)
and (b). The i s phase in ol es gas ioniza ion du ing he discha ge
phase which is up o 6.25 ms, whe e a high elec on densi y acili a es
collisions be ween high-ene gy elec ons and gas molecules. These in-
e ac ions ans e ene gy o he gas molecules, exci ing hem om hei
g ound o low-exci ed s a es o ib a ionally exci ed s a es [37]. This
esponds o he ac ha he densi ies o NH
3
, ni ogen (N
2
), and oxygen
(O
2
) all dec ease signi ican ly du ing he esidence ime in Fig. 1. As
shown in Fig. 5 (a), a e plasma ioniza ion, he inal gas densi y o N
2
and O
2
dec eased by 26 % and 44 % espec i ely. The inal obse ed
change in densi y is 26 % o N
2
and 44 % o O
2
. Fu he mo e, oxygen
a oms p esen he highes pe cen age o H, N and O. I is p esumed ha
his is since oxygen has a lowe ioniza ion (12.07 eV) and dissocia ion
ene gy (5.12 eV) and is mo e easily ionized and dissocia ed [40]. Oxy-
gen a oms, being chemically ac i e, can be gene a ed ia a ious eac-
ion pa hways. Du ing he discha ge phase, he a c aises he gas
empe a u e, which is u he go e ned by he speci ic hea , leading o a
g adual empe a u e inc ease, as shown in Fig. 5 (a). I is also obse ed
ha a e he cu en disappea s, he empe a u e dec eases be o e
e en ually ising by app oxima ely 100 K. Some in e media e compo-
nen s a e gene a ed du ing he plasma p ocess as shown in Fig. 5 (b).
The inal gas composi ion o NH
2
and OH has inc eased by 6.75 % and
4.74 %, espec i ely.
3.2. Beha iou o NH
3
in GAP
In his sec ion, he beha iou o NH
3
ioniza ion a di e en E/N is
analysed and he pa hways o NH
3
gene a ion and consump ion wi h
GAP is discussed.
The empo al e olu ion o beha iou al change o NH
3
du ing he
plasma ac ua ion a e shown in Figs. 6 and 7. Du ing he ini ial b eak-
down momen , NH
3
ib a es om he g ound s a e o di e en exci ed
s a es. A e wa ds, he di e en exci ed s a es o NH
3
eac wi h he
ini ial gas componen s o o m new NH
3
molecules. I is in e es ing o
no e ha he exci ed s a es gene a ed by he main ib a ions o NH
3
a
di e en ene gy injec ions ( o di e en E/N) a e all NH
3
( 2). This is
also consis en wi h he conclusions o Zheng e al. [37]. The a es o
NH
3
( 4) and NH
3
( 13) p oduc ion dec ease sligh ly wi h inc easing
E/N. The gene a ion a e is always smalle han he consump ion e i-
ciency, and he o e all componen s show a dec easing end. I means
ha in Fig. 7, he black line indica es ha he eac ion a e o NH
3
con e sion o di e en ib a ionally exci ed s a es o NH
3
is always
mo e han he eac ion a e o NH
3
gene a ion indica ed by he ed line.
The pa hway o NH
3
p oduc ion and consump ion is shown in Fig. 8
(a) and (b). I is obse ed ha in e con e sion eac ions be ween NH
3
and di e en exci ed s a es o NH
3
domina e he ioniza ion phase as
depic ed in Fig. 8 (a). In addi ion o he in e con e sion o NH
3
wi h
di e en exci ed s a es, he e a e u he modes o consump ion o NH
3
as shown in Fig. 8 (b). Reac ing wi h di e en elec onically exci ed
Fig. 4. Valida ion o NO emissions. (line: nume ical, symbol: expe imen ).
Fig. 5. Tempo al e olu ion o species densi y and gas empe a u e o di e en
main species. (ϕ =1, T =300 K, P =1 a m, E/N =30 Td).
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
4
s a es o N
2
, ni ogen and oxygen a oms, dehyd ogena e o o m NH
2
.
This is one o he easons o he highe p oduc ion o NH
2
men ioned in
3.1. In pa allel, he a e o eac ion o p oduce is mo e signi ican in he
discha ge phase han in he no-discha ge phase.
3.3. Gene a ion o main g oups a ec ing NO p oduc ion
In his sec ion, majo adicals such as OH, NH, and NH
2
o NO
gene a ion and consump ion wi h di e en E/N a e discussed. This was
aken as he eason o analysing he educ ion in NO emissions when
GAP p omo es NH
3
-ai combus ion.
The sensi i i y analyses o he g oups (OH, NH, NH
2
) ha play a
p ima y ole in NO p oduc ion/consump ion wi h plasma a e shown in
Fig. 9. Acco ding o Fig. 9 (a) and (c), NH
2
and NH, as he main pa hways
o NO consump ion, show he s onges sensi i i ies o he eac ions
om NH
3
o NH
3
(e1) and NH
3
(e2), espec i ely. The elec on collision
eac ions o NH
3
as a ec ed by he s eng h o he elec ic ield showed
ha he a e o he eac ion o p oduce NH
3
in he di e en exci ed s a es
inc eased apidly wi h an inc eased E/N as illus a ed in Fig. 6. This
indica es ha high ield s eng h helps o p omo e he consump ion pa h
o NO. Fo he NH
2
gene a ion eac ion, he O(
1
D) +NH
3
→ OH +NH
2
and N
2
(A) +NH
3
→ NH
2
+N
2
+H eac ion domina es (30 Td). As E/N
inc eases, N(
2
D) +NH
3
→NH
2
+H +N
2
g adually akes he lead (52.1 %
in 90 Td). Unde low- empe a u e condi ions, he he mal mo ion o
molecules is educed, and chemical eac ions p ima ily ely on non-
he mal p ocesses (such as elec on impac o he in ol emen o
exci ed species). N(
2
D) and O(
1
D) loses ene gy h ough collisions wi h
su ounding molecules o ene gy ans e , leading o he gene a ion o
in e media e adicals (e.g., OH and NH
2
) [23]. Ne e heless, he in-
c ease o OH is o en accompanied by he gene a ion o NO, h ough
HNO eac ions. In Fig. 9 (b), he main pa hway o OH o ma ion is he
eac ion be ween H a oms and oxygen in di e en ib a ional exci ed
s a es.
3.4. E ec o GAP on combus ion cha ac e is ics
In his sec ion, sensi i i y analyses and eac ion pa hways o NO in
NH
3
-ai combus ion unde GAP we e in es iga ed. P e ious expe imen s
ha e shown ha he maximum ϕ o NO change wi h plasma is be ween
0.8 and 0.9 [14,32], so his ϕ was selec ed o he p esen analysis.
The sensi i i y analysis unde plasma-assis ed combus ion e sus
pu e combus ion wi h pa h lux o NO is shown in Figs. 10 and 11. The
sensi i i y o plasma assis ed combus ion a di e en ϕ is app oxima ely
he same o he di e en eac ions, he maximum sensi i i y co-
e icien s all occu in he eac ion o HNO wi h H as shown in Fig. 10.
The wo eac ions ha consume NO show a signi ican magni ude o a e
due o he gene a ion o NH
2
and NH. The eac ion N +NO =N
2
+O a
ϕ =0.8 and 0.9 wi h sensi i i y coe icien s o 0.91 and 0.98. The a e
inc eases due o he inc ease in N a oms p oduced by ioniza ion, which
con ibu es mo e o he consump ion o NO.
Di e si ica ion o NO consump ion pa hs a di e en ϕ ( eac ions NH
+H =N +H
2
, HNO +OH =NO +H
2
O) a ibu ed o plasma-gene a ed
in e media e g oups. Simila ly, in Fig. 11, combus ion wi h o wi hou
Fig. 6. Va ia ion o elec on eac ions o NH
3
a di e en E/N (30 Td-180 Td).
Fig. 7. Va ia ion o a e coe icien s o NH
3
and ib a ionally exci ed s a e NH
3
du ing he GA discha ge ime a 30 Td.
Fig. 8. (a) P oduc ion and (b) losses o NH
3
unde GAP.
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
5
plasma, HNO is he main pa hway o NO p oduc ion, he same as ha o
li e a u e [16,41]. Almos 50 % o NO is p oduced h ough he pa hway
o HNO wi h and wi hou plasma. The pe cen age a e o p oduc ion o
NO consumed by NH
2
in GAP (12.9 %) is abou wice as high as wi hou
GAP (5.9 %). Meanwhile, OH and N a oms a e p oduced du ing he
ioniza ion phase, leading o he eac ion OH +N=H +NO exace ba ed
by NO p oduc ion wi h GAP (14.4 %, compa ed wi h 8.7 % wi hou
GAP).
4. Conclusion
This s udy p esen s a nume ical model o p edic ing NO
x
emissions
du ing gliding a c plasma (GAP)-assis ed combus ion o an NH
3
–ai
mix u e. The NO p oduc ion and consump ion mechanisms unde GAP
ac ua ion in NH
3
-ai mix u es a e also in es iga ed in de ail. The main
conclusions a e summa ized below:
(1) Th ee plasma eac ion mechanisms o NH
3
-ai combus ion we e
compa ed wi h he p esen GAP expe imen . The mechanism o
Mao e al. [23] was ound o be op imal and was subsequen ly
coupled wi h he Alnasi e al. [38] op imized combus ion
mechanism o de elop a plasma-assis ed combus ion model.
(2) The esul s e ealed ha NO p oduc ion p ima ily occu s ia he
HNO pa hway, ega dless o GAP ac ua ion. The GAP Plasma
educes NO p ima ily by b eaking down NH
3
in o NH
2
(6.75 % in
30 Td), which enhances NO consump ion, mainly h ough N
2
(A)
+NH
3
→ NH
2
+N
2
+H and O(
1
D) +NH
3
→ OH +NH
2
.
(3) I is obse ed ha highe plasma powe u he acili a es NO
consump ion. S ong elec ic ields p omo e he gene a ion o
exci ed ammonia s a es, which exhibi a highe endency o o m
NH
2
and NH, enhancing NO educ ion. NH
2
showed high sensi-
i i y (52.1 %) o he eac ion N(
2
D) +NH
3
gNH
2
+H +N
2
, N
2
(A)
+NH
3
→ NH
2
+N
2
+H. NH showed high sensi i i y (68.9 %) o
he eac ion N(
2
D) +NH
3
→NH
2
+NH (90 Td).
Fu u e s udies should ake in o accoun he in luence o he swi ling
lame s uc u e, he p opaga ion o he a c wi hin he GAP pa icula ly
he impac o a c o a ion on spa ial a ia ions as well as o he ac o s
ha may a ec lame beha iou .
Fig. 9. Sensi i i y analysis o (a) NH
2
(b) OH and (c) NH on he NO p oduc ion
a 30( ed), 60(blue), 90(o ange)Td in plasma. (Numbe s indica e he o al eac-
ion pe cen age). (Fo in e p e a ion o he e e ences o colou in his igu e
legend, he eade is e e ed o he Web e sion o his a icle.)
Fig. 10. Sensi i i y analysis o NO p oduc ion and consump ion a ϕ =0.8
and 0.9.
Fig. 11. Pa h lux o NO emissions a ϕ =0.8(wi h plasma and no plasma).
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
6
CRediT au ho ship con ibu ion s a emen
Ziyu Wang: W i ing – o iginal d a , Me hodology, In es iga ion,
Fo mal analysis, Da a cu a ion, Concep ualiza ion. B. A a ind: W i ing
– e iew & edi ing, Supe ision, Me hodology, In es iga ion, Concep-
ualiza ion. Syed Mash uk: W i ing – e iew & edi ing, Supe ision.
Agus in Vale a-Medina: W i ing – e iew & edi ing, Supe ision,
Funding acquisi ion.
Decla a ion o compe ing in e es
The au ho s decla e ha hey ha e no known compe ing inancial
in e es s o pe sonal ela ionships ha could ha e appea ed o in luence
he wo k epo ed in his pape .
Acknowledgemen s
This wo k was suppo ed by he Eu opean Union P ojec CAIPIRI-
NH3A, unde he GA Numbe 101191768. Views and opinions exp essed
a e howe e hose o he au ho (s) only and do no necessa ily e lec
hose o he Eu opean Union o CINEA. Nei he he Eu opean Union no
he g an ing au ho i y can be held esponsible o hem. The esea ch
was unde aken a Ca di Uni e si y’s The mo luids Lab (W/0.17) wi h
in aluable echnical suppo om M . Jona han Ma in. Fo he pu pose
o open access, he au ho has applied a CC BY copy igh licence o any
Au ho Accep ed Manusc ip e sion a ising. In o ma ion on he da a
unde pinning his publica ion, including access de ails, can be ound in
he Ca di Uni e si y Resea ch Da a Reposi o y a h ps://doi.o g/10
.17035/ca di .30104149. 1.
Appendix A. Supplemen a y da a
Supplemen a y da a o his a icle can be ound online a h ps://doi.
o g/10.1016/j.joei.2025.102314.
Re e ences
[1] C. Zam i escu, I. Dince , Using ammonia as a sus ainable uel, J. Powe Sou ces 185
(1) (2008) 459–465.
[2] A. Vale a-Medina, H. Xiao, M. Owen-Jones, W.I.F. Da id, P.J. Bowen, Ammonia o
powe , P og. Ene gy Combus . Sci. 69 (2018) 63–102.
[3] B. Mei, X. Zhang, S. Ma, M. Cui, H. Guo, Z. Cao, Y. Li, Expe imen al and kine ic
modeling in es iga ion on he lamina lame p opaga ion o ammonia unde
oxygen en ichmen and ele a ed p essu e condi ions, Combus . Flame 210 (2019)
236–246.
[4] S. Mash uk, M.O. Vigue as-Zuniga, M.E. Tejeda-del-Cue o, H. Xiao, C. Yu, U. Maas,
A. Vale a-Medina, Combus ion ea u es o CH4/NH3/H2 e na y blends, In . J.
Hyd ogen Ene gy 47 (70) (2022) 30315–30327.
[5] A. Vale a-Medina, R. Ma sh, J. Runyon, D. Pugh, P. Beasley, T. Hughes, P. Bowen,
Ammonia–me hane combus ion in angen ial swi l bu ne s o gas u bine powe
gene a ion, Appl. Ene gy 185 (2017) 1362–1371.
[6] R. Meloni, A. Vale a-Medina, G. Babazzi, E. Pucci, S. Cas ellani, A. And eini,
Nume ical in es iga ion o he NOX emissions o a pe ec ly p emixed NH3-H2
lame a mode a e p essu e le els, Fuel 385 (2025) 134128.
[7] H. Yamashi a, A. Hayakawa, E.C. Oka o , S. Colson, K.D.K.A. Soma a hne,
T. Tsujimu a, S. I o, M. Uchida, T. Kudo, H. Kobayashi, Op imum p ima y
equi alence a io o ich-lean wo-s age combus ion o non-p emixed ammonia/
me hane/ai and ammonia/hyd ogen/ai lames in a swi ling low, Fuel 368
(2024) 131598.
[8] J. Da ies, S. Mash uk, D. Sa o, L. Mazzo a, D. Pugh, A. Vale a-Medina, Emissions
analyses o humidi ied c acked ammonia swi ling lames, Combus . Flame 274
(2025) 113984.
[9] Y. Ju, W. Sun, Plasma assis ed combus ion: dynamics and chemis y, P og. Ene gy
Combus . Sci. 48 (2015) 21–83.
[10] Y. Ju, J.K. Le kowi z, C.B. Reu e , S.H. Won, X. Yang, S. Yang, W. Sun, Z. Jiang,
Q. Chen, Plasma assis ed low empe a u e combus ion, Plasma Chem. Plasma
P ocess. 36 (1) (2016) 85–105.
[11] J. Tian, P. Liu, Z. Liu, J. Tang, W. Yin, Y. Cheng, In es iga ion on he e ec s o
nanosecond pulsed discha ge on combus ion cha ac e is ics o C3H8/ai mix u es
using a apid comp ession machine, J. Ene gy Ins . 119 (2025) 101983.
[12] T. Yu, H. Zhang, Z. Zhao, C. Kong, R. Zhang, J. Zhu, B. Zhou, Cha ac e is ics o an
AC o a ing gliding a c discha ge in NH3 and ai a mosphe es, Phys. Plasmas 31 (2)
(2024).
[13] H. Zhong, X. Mao, A.C. Rousso, C.L. Pa ick, C. Yan, W. Xu, Q. Chen, G. Wysocki,
Y. Ju, Kine ic s udy o plasma-assis ed n-dodecane/O2/N2 py olysis and oxida ion
in a nanosecond-pulsed discha ge, P oc. Combus . Ins . 38 (4) (2021) 6521–6531.
[14] R. Ju, J. Wang, M. Zhang, H. Mu, G. Zhang, J. Yu, Z. Huang, S abili y and emission
cha ac e is ics o ammonia/ai p emixed swi ling lames wi h o a ing gliding a c
discha ge plasma, Ene gy 277 (2023) 127649.
[15] J. Choe, W. Sun, T. Omb ello, C. Ca e , Plasma assis ed ammonia combus ion:
simul aneous NOx educ ion and lame enhancemen , Combus . Flame 228 (2021)
430–432.
[16] A.M. Elbaz, S. Wang, T.F. Guibe i, W.L. Robe s, Re iew on he ecen ad ances on
ammonia combus ion om he undamen als o he applica ions, Fuel Commun. 10
(2022) 100053.
[17] A.M. Radwan, M.C. Paul, Plasma assis ed NH3 combus ion and NOx educ ion
echnologies: p inciples, challenges and p ospec i e, In . J. Hyd ogen Ene gy 52
(2024) 819–833.
[18] R. Ju, J. Wang, M. Zhang, H. Mu, Y. Wu, G. Zhang, Z. Huang, Expe imen al s udy
on bu ning eloci y, s uc u e, and NOx emission o p emixed lamina and swi l
NH3/H2/ai lames assis ed by non- he mal plasma, Appl. Ene gy Combus . Sci. 14
(2023) 100149.
[19] G. Faingold, J.K. Le kowi z, A nume ical in es iga ion o NH3/O2/He igni ion
limi s in a non- he mal plasma, P oc. Combus . Ins . 38 (4) (2021) 6661–6669.
[20] M. Shahsa a i, A.A. Konno , A. Vale a-Medina, M. Jangi, On nanosecond plasma-
assis ed ammonia combus ion: e ec s o pulse and mix u e p ope ies, Combus .
Flame 245 (2022) 112368.
[21] X. Mao, A. Rousso, Q. Chen, Y. Ju, Nume ical modeling o igni ion enhancemen o
CH4/O2/He mix u es using a hyb id epe i i e nanosecond and DC discha ge,
P oc. Combus . Ins . 37 (4) (2019) 5545–5552.
[22] H. Zhong, X. Mao, N. Liu, Z. Wang, T. Omb ello, Y. Ju, Unde s anding non-
equilib ium N2O/NOx chemis y in plasma-assis ed low- empe a u e NH3
oxida ion, Combus . Flame 256 (2023) 112948.
[23] X. Mao, H. Zhong, N. Liu, Z. Wang, Y. Ju, Igni ion enhancemen and NOx o ma ion
o NH3/ai mix u es by non-equilib ium plasma discha ge, Combus . Flame 259
(2024) 113140.
[24] T.S. Taneja, P.N. Johnson, S. Yang, Nanosecond pulsed plasma assis ed combus ion
o ammonia-ai mix u es: e ec s on igni ion delays and NOx emission, Combus .
Flame 245 (2022) 112327.
[25] M. Shahsa a i, A.A. Konno , X.-S. Bai, A. Vale a-Medina, T. Li, M. Jangi,
Syne gis ic e ec s o nanosecond plasma discha ge and hyd ogen on ammonia
combus ion, Fuel 348 (2023) 128475.
[26] Y. Qiu, Y. Zhu, Y. Wu, N. Zhao, Z. Li, M. Hao, B. Zhang, D. Pan, Nume ical
in es iga ion o he hyb id pulse–DC plasma assis ed igni ion and NOx emission o
NH3/N2/O2 mix u e, Combus . Flame 258 (2023) 113078.
[27] Z. Xin, X. Li, Z. Zheng, Y. Hu, R. Cao, A. Sun, F. Zhao, W. Yu, S udy on he kine ics
and ene gy ans e du ing igni ion o me hane exci ed by NRP-SDBD non-
equilib ium plasma, J. Ene gy Ins . 118 (2025) 101929.
[28] Y. Tang, D. Xie, B. Shi, N. Wang, S. Li, Flammabili y enhancemen o swi ling
ammonia/ai combus ion using AC powe ed gliding a c discha ges, Fuel 313
(2022) 122674.
[29] Z. Wang, B. A a ind, S. Mash uk, A. Vale a-Medina, Expe imen al in es iga ion on
he e ec s o gliding a c plasma on he combus ion cha ac e is ics o ai s a i ied
NH3 lames, Ene gy (2025) 138413.
[30] J. Cui, Z. Xu, J. Zhang, Q. Nie, G. Xu, L. Ren, Online diagnosis o elec on exci a ion
empe a u e in CH4+H2 discha ge plasma a a mosphe ic p essu e by op ical
emission spec a. Science in China se ies G: physics, Mech. As on. 51 (12) (2008)
1892–1896.
[31] L.W.S. C ispim, P.H. Hallak, M.S. Benilo , M.Y. Balles e , Modelling spa k-plug
discha ge in d y ai , Combus . Flame 198 (2018) 81–88.
[32] B. A a ind, Z. Wang, Syed Mash uk, Deanna A. Lacos e, A. Vale a-Medina, No el
S a egy o Combus ion Enhancemen o NH3-ai Mix u e Using Gliding A c
Plasma, ol. 41, P oceedings o he combus ion ins i u e, 2025.
[33] D.P. Subedi, R.P. Gu again, U.M. Joshi, Su ace modi ica ion o polyme s by 50 Hz
dielec ic ba ie discha ge (DBD) plasma p oduced in ai a 40 To , Fund. Plasma
Phys. 10 (2024) 100058.
[34] G. Dong, Y. Zhou, P. Ming, Z. Wu, H. Chen, Y. Huang, L. Li, Kine ics-Based analysis
o he gliding a c plasma assis ed ammonia decomposi ion p ocess owa ds ehicle
on-boa d applica ions, Chem. Eng. J. 505 (2025) 159443.
[35] S. Pancheshnyi, B. Eismann, G. Hagelaa , L. Pi ch o d, Compu e Code Zdplaskin,
Uni e si y o Toulouse, LAPLACE. Tech. Rep, CNRS-UPS-INP, Toulouse, F ance,
2008.
[36] R.J. Kee, F.M. Rupley, E. Meeks, J.A. Mille , CHEMKIN-III: a FORTRAN Chemical
Kine ics Package o he Analysis o Gas-phase Chemical and Plasma Kine ics,
1996.
[37] Z. Zheng, C. Wang, Z. Xin, Y. Hu, Q. Zhu, W. Yang, F. Zhao, W. Yu, Kine ic
modelling o non-equilib ium plasma enhanced ca aly ic ammonia decomposi ion,
J. Ene gy Ins . 116 (2024) 101715.
[38] A. Alnasi , S. Mash uk, M. Hayashi, J. J´
ojka, H. Shi, A. Hayakawa, A. Vale a-
Medina, Pe o mance in es iga ion o cu en ly a ailable eac ion mechanisms in
he es ima ion o NO measu emen s: a compa a i e s udy, Ene gies 16 (9) (2023)
3847.
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
7
[39] A. Alnasi , J. J´
ojka, M. Papp, A.G. Szan ho e , M. Ko ale a, T. Tu anyi, A. Vale a
Medina, T. Nagy, Compac Kine ic Model o he Combus ion o NH3/H2 Mix u es,
J Ammonia Ene gy, 2025.
[40] R. Pa ajuli, S. Ma , A. S ama o ic, T.D. M¨
a k, P. Scheie , Unimolecula
dissocia ion o non-s oichiome ic oxygen clus e ions On+* (n=5, 7, 9, 11): a
swi ch om O3 o O2 loss abo e clus e size n=5, In . J. Mass Spec om. 220 (2)
(2002) 221–230.
[41] S. Mash uk, H. Shi, L. Mazzo a, C.E. Us un, B. A a ind, R. Meloni, A. Alnasi ,
E. Boule , R. Jankowski, C. Yu, M. Alnajideen, A. Paykani, U. Maas, R. Sle a ski,
D. Bo ello, A. Vale a-Medina, Pe spec i es on NOX emissions and impac s om
ammonia combus ion p ocesses, Ene gy Fuel. 38 (20) (2024) 19253–19292.
Z. Wang e al.
Jou nal o he Ene gy Ins i u e 123 (2025) 102314
8
