Resea ch A icle
Enhanced Ammonia Adso p ion on Di ec ly Deposi ed
Nano ib ous Ca bon Films
Alexande G. Banno ,
1,2
Ondřej Jašek,
3
Jan P ášek ,
4,5
Jiří Bu šík ,
6
and Lenka Zajíčko á
1,3
1
Cen al Eu opean Ins i u e o Technology, Masa yk Uni e si y, Kamenice 5, CZ-62500 B no, Czech Republic
2
Depa men o Chemis y and Chemical Technology, K. Ma x 20, No osibi sk S a e Technical Uni e si y,
630073 No osibi sk, Russia
3
Depa men o Physical Elec onics, Facul y o Science, Masa yk Uni e si y, Ko lářská 2, CZ-61137 B no, Czech Republic
4
Cen al Eu opean Ins i u e o Technology, B no Uni e si y o Technology, Technická 3058/10, CZ-61600 B no, Czech Republic
5
Cen e o Senso s, In o ma ion and Communica ion Sys ems, Facul y o Elec ical Enginee ing and Communica ion,
Technická 3058/10, CZ-61600 B no, Czech Republic
6
Ins i u e o Physics o Ma e ials, Academy o Sciences o he Czech Republic, Žižko a 22, CZ-61662 B no, Czech Republic
Co espondence should be add essed o Alexande G. Banno ; [email p o ec ed]
Recei ed 12 Ap il 2018; Re ised 9 July 2018; Accep ed 29 July 2018; Published 17 Sep embe 2018
Academic Edi o : Michele Penza
Copy igh © 2018 Alexande G. Banno e al. This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion
License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
The ammonia adso p ion on he nanos uc u ed ca bon hin film was significan ly influenced by he choice o deposi ion
empe a u e and deposi ion ime o hin film syn hesis. The hin films we e p epa ed on Si/SiO
2
subs a es by chemical apou
deposi ion in A /C
2
H
2
gas mix u e using i on ca aly ic nanopa icles. The analysis o he g own laye by he scanning and
ansmission elec on mic oscopy showed he ansi ion om long mul iwalled nano ubes (MWCNTs) o bamboo-like hollow
ca bon nanofibe s uc u e wi h he dec ease o he deposi ion empe a u e om 700 o 600
°
C. Fu he , he ma e ial was
analyzed by ene gy-dispe si e X- ay spec oscopy and Raman spec oscopy confi med he ansi ion om g aphi ic sp
2
s uc u e o highly de ec i e s uc u e a lowe deposi ion empe a u e. The esis ance o he p epa ed laye s ongly depends on
deposi ion empe a u e (Td) and deposi ion ime ( d). High esis ance laye , 38.6 kΩ, was o med a Td600
°
C and d10 min,
while a Td700
°
C and d60 min, he esis ance dec eased o 860 ohms. Such beha iou is consis en wi h MWCNTs being
esponsible o he o ma ion o he conduc i e ne wo k. Such sys em was s udied using chemi esis o ammonia gas senso
configu a ion. The senso esis ance inc eased when exposed o ammonia in all he cases, bu hei esponse a ied conside ably.
A dec ease in deposi ion ime, om 60 o 10 min, and he deposi ion empe a u e, om 700 o 600
°
C, led o he 10- old inc ease
in he senso esponse. The measu emen s ca ied ou a oom empe a u e showed he highe senso esponse han he
measu emen s ca ied ou a 200
°
C. This beha iou can be explained by he change in adso p ion-deso p ion equilib ium a
diffe en empe a u es. Analysis o dependence o he senso esponse on he ammonia concen a ion p o ed ha he
unde lying esis ance change mechanism is chemiso p ion o ammonia molecules on he ca bon ne wo k co esponding o he
Langmui iso he m.
1. In oduc ion
A sa e y and sus ainable de elopmen equi e c ea ion o new
highly sensi i e de ices o he de ec ion o oxic and flamma-
ble gases along wi h he con ol o en i onmen al pollu ion.
Such de ices equi e new ma e ials which ex end he numbe
o de ec able gases and lowe hei de ec ion limi s. One o he
mos dange ous gases ha ha e a nega i e influence on human
heal h and en i onmen is ammonia (NH
3
). Acco ding o US
OHSA (Occupa ional Sa e y and Heal h Adminis a ion) and
CDC (Cen e s o Disease Con ol and P e en ion) egula-
ions, he long- e m exposu e limi s a e 25–35 ppm o
Hindawi
Jou nal o Senso s
Volume 2018, A icle ID 7497619, 14 pages
h ps://doi.o g/10.1155/2018/7497619
wo ke s. The smell h eshold o ammonia anges om 5 o
15 ppm. The concen a ions abo e 50 ppm induce i i a ion
o he mou h, nose, wheezing, e c. The concen a ions
om 300 o 500 ppm a e dange ous o li e. In indus y,
ammonia can be de ec ed a highe concen a ions ( om
1000 ppm o 40000 ppm, depending on ooms, acili ies, e c.).
The e o e, s udy o ammonia adso p ion on nanos uc-
u ed su aces and de elopmen o new ma e ials o ammo-
nia gas senso s a ac ed a lo o a en ion o scien ific
communi y. Single-wall ca bon nano ubes (SWCNTs) and
mul iwall ca bon nano ubes (MWCNTs) belong o one o
he mos p omising candida es in he gas sensing field
[1–9] because o hei abili y o change he elec ical p ope -
ies (e.g., esis i i y) unde gas adso p ion. This ac is com-
plemen ed by he enhanced adso p ion o gases by ca bon
nanoma e ials [8]. The e a e a wide ange o gases on which
he ca bon nano ube- (CNT-) based senso s can be used,
such as NH
3
[4, 10–16], NO
2
[17, 18], CH
4
[19, 20], H
2
[21], H
2
S[22–25], CO
2
[26], e hanol [25], me hanol [25],
hyd oca bons [7], and o he gases.
Comme cially a ailable senso s equi e a high empe a-
u e o hei ope a ion. Fo wide use o such de ices, he
oom empe a u e ope a ion o ammonia gas senso s is an
impo an equi emen and was ecen ly s udied by se e al
au ho s. In [27], he au ho s c ea ed HCl-doped MWCNT/
polyaniline composi e senso s wi h good sensing esponse
and high ep oducibili y. In [28], he au ho s c ea ed
SWCNT- (40% me allic and 60% semiconduc ing) based
ammonia senso s using inkje -p in ed elec odes which pos-
sessed he maximal esponse o 27.3% o 500 ppm a oom
empe a u e. Cui e al. [29] de eloped he oom empe a u e
ammonia senso based on Ag nanoc ys al- unc ionalized
MWCNTs ha exhibi ed enhanced esponse o 9% and as
esponse a oom empe a u e wi h he ull eco e y wi hin
se e al minu es in ai . In [12], he au ho s c ea ed he oom
empe a u e senso based on SWCNT o ammonia sensing
wi h he ex emely low de ec ion limi (3 ppb). The enhance-
men o he esponse was ealized by wo echniques: d op
cas ing and sonica ion.
In his pape , a p omising po en ial o managing he NH
3
gas senso pe o mance by he con ol o nanofib ous ca bon
(NFC) chemical apou deposi ion (CVD) syn hesis condi-
ions is in es iga ed in de ail. Up o now, he CNT-based
NH
3
gas senso s we e closely in es iga ed om he pe o -
mance poin o iew bu he influence o he syn hesis condi-
ions o hese ma e ials on he sensing p ope ies has no ye
been s udied. P e ious s udies we e based on he c ea ion o
he senso s by he di ec deposi ion o CNTs using only ce -
ain condi ions [30–33] wi hou de ailed in es iga ion o
hei syn hesis ole in he o ma ion o senso esponse.
The influence o NFC deposi ion pa ame e s (g ow h em-
pe a u e and ime) on he ma e ials’p ope ies and NH
3
sensing cha ac e is ics is de e mined. The in e connec ion
be ween he senso esis ance and esponse has been ound.
2. Ma e ials and Me hods
2.1. Nanofib ous Ca bon Film P epa a ion. NFC laye s we e
g own by a mosphe ic p essu e chemical apou deposi ion
using an i on ca alys . Polished single-c ys al Si (c-Si) pieces
(8 mm ×8 mm) coa ed by a he mal SiO
2
film, 92 nm in
hickness, we e used as subs a es. Nanopa icles (NPs) o
i on ca alys we e deposi ed by a mic owa e (MW) plasma
o ch om he i on pen aca bonyl, Fe(CO)
5
, apou s mixed
wi h a gon. The expe imen al se -up is desc ibed by Synek
e al. [34, 35] in de ail. The c-Si/SiO
2
subs a es we e pu in
a special holde o 4 samples. The flow a e o a gon h ough
he cen al pa o he nozzle was 700 sccm. The ou e con-
cen ic pa o he nozzle was used o deli e Fe(CO)
5
apou s (0.1 sccm) ca ied by A flow o 28 sccm. The MW
o ch was igni ed wi h he powe o 210 W. The deposi ion
ime o he nanopa icles was 15 s. Di ec deposi ion o nano-
pa icles was used as p e e ed o m o he ca alys because i
enabled us o o m a spa se ne wo k o fib ous ca bon. In
case o he hin film ca alys , dense s uc u e o he nanofi-
b ous ca bon ne wo k wi h low esis ance and negligible
esponse was o med.
NFC g ow h was ca ied ou in a qua z ubula u nace.
The p ocesses s a ed by hea ing he u nace om he oom
empe a u e o deposi ion empe a u e (Td): 600
°
C, 650
°
C,
and 700
°
C, unde he a gon flow o 1400 sccm. The hea ing
speed was 25
°
C/min. Then, he ca aly ic nanopa icles
we e educed in A /H
2
flow (1400 sccm o A and 500 sccm
o H
2
) o 10 min. A e educ ion, he hyd ogen flow was
swi ched offand NFC g ow h was ca ied ou in A /C
2
H
2
mix u e wi h he flow a es o 1400 and 25 sccm, espec i ely.
The g ow h las ed o 10 min, 40 min, and 60 min as summa-
ized in Table 1.
The eac o was cooled down o he oom empe a-
u e unde a gon flow. The samples we e aken ou
and placed in a acuum e apo a o o he deposi ion o
6.65 mm ×2.33 mm gold con ac ing pads o e lapping by
1 mm wi h he field o NFC (Figu e 1). Thickness o he Au
laye was 350 nm, and a 15 nm hick Ni/C laye was used
o imp o e Au adhesion o he subs a e. Such p epa ed sub-
s a es we e used as senso s o measu emen o esis ance
change unde a ious gas a mosphe es.
2.2. In es iga ion Me hods o NFC Films. The as-p epa ed
ca aly ic NPs and NFCs on c-Si/SiO
2
subs a es we e in es i-
ga ed by scanning elec on mic oscopy (SEM) wi h MIRA II
Table 1: Summa y o condi ions used o he g ow h o NFC
by CVD.
Sample
Reduc ion and
deposi ion empe a u e
(Td),
°
C
Reduc ion
ime, min
Deposi ion
ime ( d), min
CNF600-10 600 10 10
CNF600-40 600 10 40
CNF600-60 600 10 60
CNF650-10 650 10 10
CNF650-40 650 10 40
CNF650-60 650 10 60
CNF700-10 700 10 10
CNF700-40 700 10 40
CNF700-60 700 10 60
2 Jou nal o Senso s
(TESCAN, B no, Czech Republic) equipped wi h he EDX
de ec o (Ox o d Ins umen s, UK). Raman spec a o NFCs
we e ob ained using he Renishaw inVia (Renishaw, Glouces-
e shi e, UK) spec ome e in he ange 100–3200 cm
−1
(λ= 514 nm). Raman spec a we e ea ed using Lo en zian
fi ing. S uc u e and mo phology o ca bon nanoma e ials
and ca aly ic nanopa icles we e addi ionally in es iga ed
using he CM12 STEM ansmission elec on mic oscope
(Philips, Eindho en, Ne he lands).
Gas sensing cha ac e is ics o he senso s we e de e -
mined by measu ing he changes o senso esis ance du ing
ammonia exposu e in a cus om-buil sys em equipped wi h
wo gas channels and a measu emen chambe (Figu e 1).
Syn he ic ai (80% N
2
, 20% O
2
, Linde, B no, Czech Republic)
was used as a gas ca ie in one gas channel. Along wi h i , he
second gas line was flowed by ammonia dilu ed in ni ogen
(calib a ion gas 5000 ppm o NH
3
in N
2
, Linde, B no, Czech
Republic). To al olume flow a e o gases was se cons an a
500 sccm o all s abiliza ions and measu emen s, bu he
concen a ion o NH
3
in he mix u e wi h syn he ic ai and
N
2
was changed. The addi ional channel o oxygen was used
in he se up, and oxygen was admixed when inc easing
ammonia concen a ion in o de o keep he concen a ion
o ai and ni ogen mix u e he same as in syn he ic ai .
Gas dis ibu ion sys em was made o s ainless s eel
(connec ions, ubes). The olume o he chambe whe e he
senso s we e examined was 160 cm
3
(leng h: 10 cm, wid h:
8 cm, heigh : 2 cm).
The senso was placed on a hea e inside he measu e-
men chambe , and he measu emen s we e ca ied ou a
wo empe a u es: oom empe a u e (25 ±2
°
C) and 200
°
C
(±2
°
C). The hea ing empe a u e was con olled by he
DC powe supply Agilen U3606A (Agilen , San a Cla a,
Cali o nia, USA) using empe a u e calib a ion cu es. The
esis ance was measu ed by wo elec odes ha we e placed
o Au pads. Elec odes we e plugged o a high ol age sou ce
me e Kei hley 2410 (Kei hley, Cle eland, OH, USA) using
1 V bias ol age. In he p esen wo k, no effo s we e made
o dope o ea NFCs o esponse enhancemen . Solely,
he co ela ions be ween he NFC deposi ion condi ions,
influencing he s uc u e and composi ion o he deposi ,
and he gas senso p ope ies we e in es iga ed. Rela i e
humidi y (RH) in he chambe was con olled by he
SHT25 senso (Sensi ion, S ae a, Swi ze land). Also, he
measu emen s o humidi y influence on sensing p ope ies
we e ca ied ou . The le el o RH du ing measu ing he
esponse in d y gases om cylinde s was 2.5–3%. The mea-
su emen s in diffe en RH condi ions we e ca ied ou by
eeding he we ai o he chambe addi ionally wi h d y ai
and analy e.
Be o e each esponse measu emen , he senso was con-
di ioned and eco e ed o 30 min a 200
°
C in he syn he ic
ai flow o 1000sccm. The baseline o he senso esis ance
was s abilized a desi ed empe a u e in 500 sccm syn he ic
ai flow o 60 min be o e he fi s measu emen o each
sample. A e he baseline s abiliza ion, he senso esponse
o NH
3
was measu ed using al e na ing cycles, 10 min in
syn he ic ai and 10 min in he mix u e o he NH
3
wi h
syn he ic ai . This cycle was epea ed o se e al es ed
concen a ions, 100 ppm, 250 ppm, and 500 ppm o NH
3
.
Ai Ac i e ma e ial
NiC / Au con ac s
Val e
Flow con olle
Measu ing cell
P essu e con olle
Exhaus
Si/SiO2 subs a e
NH3
Figu e 1: The senso layou and expe imen al se -up o measu emen o senso esponse o ammonia.
3Jou nal o Senso s
The baseline esis ance, R0, was linea ly ex apola ed along
he measu emen ime, and he senso esponse was de e -
mined by sub ac ing he baseline esis ance signal om he
sample esis ance unde ammonia exposu e, R. The measu e-
men o baseline and ex apola ion has been done o each
measu emen o all samples. The senso esponse was defined
as ollows:
ΔR
R0
=R−R0
R0
⋅100% 1
The ins umen al de ec ion limi o he senso esponse
was 10 ppm, he alue was limi ed by he flow con olle
cha ac e is ics (low accu acy o flow a e con ol a he alue
up o 1 sccm) o he senso measu emen . To es ima e he
senso selec i i y, hey we e also es ed o de ec ion o H
2
and iC
4
H
10
. These gases we e also ed dilu ed in N
2
(5000 ppm o NH
3
). The esolu ion o scanning o senso
esis ance was 0.5 s (one expe imen al poin pe 0.5 s).
3. Resul s and Discussion
3.1. Cha ac e iza ion o Ca aly ic Nanopa icles and Ca bon
Nanos uc u es. The ypical SEM and TEM mic og aphs o
ca aly ic nanopa icles a e shown in Figu e 2. The ca alys
o med agg ega es consis ing o nanopa icles wi h he size
o 5–25 nm. I was difficul o de e mine he phase compo-
si ion o he ca alys by X- ay diff ac ion o Raman spec-
oscopy because o he low amoun o nanopa icles on
he subs a e. The e o e, he phase composi ion was s udied
by elec on diff ac ion in TEM. The analyses e ealed ha
he ca alys consis ed o i on oxides, p edominan ly maghe-
mi e (γ-Fe
2
O
3
). Maghemi e (γ-Fe
2
O
3
) phase was de ec ed
(Figu e S1 Supplemen a y Ma e ials) by selec ed a ea
elec on diff ac ion (SAED) by TEM. I is in ag eemen
wi h he da a ob ained by Synek e al. [34].
The SEM and TEM mic og aphs o he NFCs g own o
10 min a 600
°
C (Figu e 3) show bo h he ypical hollow
ca bon s uc u es obse ed also in o he samples, long
MWCNTs (2–5μm) wi h walls consis ing o g aphene laye s
pa allel o he ube axis, and s ongly cu ed sho bamboo-
like hollow s uc u es. Besides, he samples con ained also
ca bon-encapsula ed Fe and Fe
3
C nanopa icles.
The diame e o MWCNTs, 10–30 nm, sugges s ha he
g ow h was ini ia ed om small, 5–25 nm, ca aly ic NPs.
The bamboo-like s uc u es, 50–120 nm in diame e , a e
expec ed o g ow om la ge NPs (>45 nm in diame e ).
The diffe ence in hei g ow h mechanism, as compa ed o
MWCNTs, induces a s ong cu a u e wi h chain-like mo -
phology ha has some simila i ies o CNFs [36]. Ini ially, a
highe amoun o small NPs, a lowe ac i i y o la ge NPs,
and limi ed g ow h ime we e he easons o much highe
amoun o MWCNTs han bamboo-like CNFs on he sub-
s a es (Figu e 3(a)) [37].
SEM mic og aphs o he NFC samples p epa ed a diffe -
en empe a u es and imes a e p esen ed in Figu e S2
(Supplemen a y Ma e ials). A 600
°
C, he inc eased
deposi ion imes, 40 and 60 min, led o he g ow h o longe
nano ubes and a o ma ion o mo e dense nano ube
agg ega es compa ed o 10 min. Acco ding o EDX analysis
o he CNF600-10 sample, he C/Fe weigh a io was 26.5
and i inc eased o 85 and 107 wi h inc easing deposi ion
ime o CNF600-40 and CNF600-60 samples, espec i ely.
A highe empe a u es (650, 700
°
C), he yield o CNT
inc eased and a dense MWCNT ne wo k was o med.
Raman spec oscopy o all he NFC samples was ca ied
ou o assessing he o e all s uc u e o he deposi s. Raman
spec a o he CNF600-10, CNF650-10, and CNF700-10
samples a e shown in Figu e 4(a). The spec a we e domina ed
by wo peaks co esponding o diso de ed Dand g aphi ic G
bands [38]. The posi ions o he Dand Gpeaks anged om
1349 o 1352 cm
−1
and om 1581 o 1594 cm
−1
, espec i ely
(Table 2).
The second-o de peaks we e p esen ed abo e 2500 cm
−1
.
The peaks o silicon a 520 cm
−1
and 960 cm
−1
we e obse ed
in case o spa sely coa ed subs a es (low g ow h o em-
pe a u e and ime). The a io o Dand Gpeak in ensi ies,
ID/I G , p o ided in o ma ion abou ma e ial’s diso de
(a) (b)
Figu e 2: SEM (a) and TEM (b) mic og aphs o ca aly ic nanopa icles syn hesized by he MW o ch.
4 Jou nal o Senso s
deg ee [38]. The dependencies o he I D /I G on he
NFC g ow h ime a e shown in Figu e 4(b). The diso de
deg ee inc eases wi h inc easing g ow h ime. I can be
a ibu ed o he loss o nanopa icle ca aly ic ac i i y. Less
de ec i e samples, syn hesized o only 10 min a 600, 650,
and 700
°
C, had I D /I G = 0.88, 0.7, and 0.44, espec-
i ely. I e eals ha highe empe a u e had a posi i e
effec on he NFC g aphi iza ion deg ee, bu he ole o
he empe a u e in he supp ession o de ec s is lowe o
a longe g ow h ime.
3.2. Ammonia Gas Response Measu emen s o NFC Films.
The senso esis ances a ied in a wide ange om 0.87 kΩ
o 38.60 kΩ(Table 3). The highes esis ance o each deposi-
ion ime was ob ained a 600
°
C. The esis ance d opped
om 38.60 o 2.72 kΩ( o 10 min deposi ion ime) when
he deposi ion empe a u e inc eased om 600 o 700
°
C.
The esis ance a ia ions can be explained wi h he help
o high esolu ion SEM images (Figu e 5) ha di ided he
senso s in o h ee g oups. The fi s g oup is ep esen ed by
only one sample CNF600-10 which possesses he highes
esis ance. The sensing ma e ial is composed o sho ca bon
nanofibe s connec ed wi h each o he by a ely dispe sed
MWCNTs. The second g oup consis s o CNF650-10,
CNF600-40, and CNF600-60 samples. The leng h o he
CNFs and CNTs inc eases, and i c ea es addi ional connec-
ions and o ms he ne wo k wi h a lowe senso esis ance
(R≈6–9kΩ). The hi d g oup is ep esen ed by CNF650-60,
CNF700-10, CNF700-40, and CNF700-60 samples. Inc eas-
ing he amoun o MWCNTs enhances he o ma ion o he
conduc i e ne wo k be ween nano ubes whe eas he ole
o ca bon nanofibe s becomes negligible (senso esis ance
R<3 kΩ). The conduc i e ne wo k o med du ing longe
deposi ion ime and wi h he inc ease o he deposi ion
empe a u e, and i has a ce ain simila i y wi h he o -
ma ion o pe cola ing ne wo ks. The sample in he fi s
g oup can be unde s ood as a quasi-insula ing s a e o
he ne wo k (in e ms o pe cola ion heo y); he second
g oup o samples is si ua ed in he ansi ion egion, and
he hi d g oup ep esen s a conduc i e ne wo k in which
esis ance is weakly influenced by he u he inc ease in
deposi ion ime and empe a u e.
The changes o he senso esis ance could be pa ially
explained also by a changed de ec i eness o ca bon s uc-
u es. Indeed, a highe nano ube g aphi iza ion deg ee was
p o ed by Raman spec oscopy, i.e., lowe ID/I G a io,
(a) (b)
(c) (d)
Figu e 3: SEM (a) and TEM (b, c, d) mic og aphs o he CNT600-10 sample.
5Jou nal o Senso s

when he empe a u e du ing he g ow h was inc eased om
600 o 650 and 700
°
C (Figu e 4(b)). Howe e , he senso
esis ances did no inc ease wi h inc easing g ow h ime
al hough ID/I G was highe . The inc eased ime esul ed
in he deposi ion o highe amoun o in e connec ing long
MWCNTs, and i had much s onge influence on he senso
esis ance ha d opped significan ly o 40 min o he
g ow h, especially o 600
°
C.
I is wo h no ing ha acco ding o chemical equip-
men and plasma equipmen in indus y, i is possible o
ca y ou g owing o CNTs on la ge wa e s. Mo eo e , a
CVD p ocess is a e y simple way o CNT g ow h, and
i is app op ia e o p ac ical applica ion. The p ocesses
wi h a floa ing ca alys o he deposi ion o i on nanopa -
icles a e also successi ely used o plasma-enhanced CVD
o nanopa icles o subsequen g ow h o CNTs. The
possibili y o con ol he g ow h ime and deposi ion empe -
a u e is an ad an age o he p ocess used in his pape ,
because i is possible o ob ain he defined esis ance o he
sensing laye ha makes he p ocess flexible and a o able
o indus y.
3.3. Ammonia Adso p ion on Ca bon Nanos uc u ed Films.
All he senso s exhibi ed inc eased esis ance upon ammonia
exposu e. The inc ease o he NFC senso esis ance du ing
an ammonia adso p ion has been explained on he basis o
he hole deple ion in p- ype MWCNTs [39, 40].
Adso bed ammonia molecules dona e elec ons o CNTs
inducing a dec ease o cha ge ca ie s (holes). The same
effec was obse ed o p- ype SWCNTs [40]. The esis ance
o CNF600-10, 38.60 kΩ, was qui e high o he MWCNT-
based senso s [40, 41], bu his sample exhibi ed he highes
esponse, 2.8% o 500 ppm a oom empe a u e (Figu e 6).
In addi ion, he sample CNF600-40 exhibi ed he esponse
compa able o CNF600-10, e.g., 2.5% a 500 ppm ( oom
empe a u e). Compa ing esul s o Raman spec oscopy
0
5000
10000
15000
CNT600-10
CNT650-10
In ensi y, a b. un.
CNT700-10
500 1000 1500 2000 2500 3000
Raman shi , cm−1
(a)
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
I(D)/I(G)
600 ºC
650 ºC
700 ºC
10 20 30 40 50 60
Time, min
(b)
Figu e 4: (a) Raman spec a o CNF600-10, CNF650-10, and CNF700-10 samples. (b) In ensi y a io I D /I G s. NFC g ow h ime.
Table 2: Raman spec oscopy da a o he samples.
Sample Dpeak posi ion, cm
−1
Dpeak FWHM, cm
−1
Gpeak posi ion, cm
−1
Gpeak FWHM, cm
−1
I D /I G
CNF600-10 1352 102 1590 70 0.88
CNF600-40 1350 150 1593 73 0.92
CNF600-60 1350 125 1594 71 1.03
CNF650-10 1350 80 1586 55 0.70
CNF650-40 1349 84 1587 64 0.78
CNF650-60 1351 158 1592 72 0.92
CNF700-10 1349 71 1581 44 0.44
CNF700-40 1352 79 1589 62 0.76
CNF700-60 1350 82 1586 66 0.84
6 Jou nal o Senso s
(Figu e 4(b)), SEM images (Figu e 5), and senso esis ance
(Table 3), i can be concluded ha he senso esponse is
mos ly linked o he mic os uc u e and high senso
esis ance. The high esis ance senso s possessed low cha ge
ca ie concen a ion, and hus, hey we e mo e sensi i e o
any change o NH
3
adso p ion. Mish a e al. [32] supposed
Table 3: Summa y o senso p ope ies. Senso esis ances a oom empe a u e and 200
°
C a e deno ed RRT and R200°C, espec i ely. The
senso esponses, S
100 ppm
,S
250 ppm
, and S
500 ppm
o 100, 250, and 500 ppm o ammonia, espec i ely, a oom empe a u e (RT) and 200
°
C
we e calcula ed as ΔR/R0.
Sample RRT,ΩR200°C,ΩS
100 ppm
,% S
250 ppm
,% S
500 ppm
,%
RT 200
°
C RT 200
°
C RT 200
°
C
CNF600-10 38600 24100 1.8 1.1 2.4 1.4 2.8 2.1
CNF600-40 8940 2670 1.8 0.7 2.3 1.4 2.5 1.9
CNF600-60 8610 2580 0.6 1.1 0.8 n/a
1
1.1 n/a
1
CNF650-10 6860 4380 1.2 0.5 1.7 0.7 2.0 1.1
CNF650-40 2660 800 0.4 0.1 0.7 0.2 1.0 0.2
CNF650-60 1680 670 0.6 0.1 0.8 0.2 0.9 0.3
CNF700-10 2720 1340 0.4 0.4 0.6 0.7 0.7 1.1
CNF700-40 1420 530 0.3 0.2 0.4 0.4 0.5 0.5
CNF700-60 870 440 0.1 n/a
1
0.2 n/a
1
0.2 n/a
1
1
The senso esponse was compa able wi h noise.
(a) (b)
(c) (d)
Figu e 5: SEM images o NFCs: (a) CNF600-10 (sample wi h he highes esis ance), (b) CNF600-40 and (c) CNF650-10 (samples om he
g oup wi h medium esis ance), (d) CNF700-10 (sample om he g oup wi h he lowes esis ance).
7Jou nal o Senso s
ha he inc ease o he diso de deg ee imp o ed he senso
esponse because a highe de ec densi y inc eased he
numbe o ac i e si es in which ammonia can be adso bed.
The ID/I G o he fi e senso s om he hi d g oup
( he esis ance below 3 kΩ) was qui e diffe en , 0.44–0.92,
bu hei esponse did no p o e o be dependen on i .
F om s uc u al s udies and ammonia adso p ion cha ac-
e iza ion, one can conclude ha no only p ope ies o
indi idual componen s o hin film (CNFs and CNTs) bu
also he amoun and s uc u al a angemen o nanos uc-
u es a e impo an o hin film esponse owa ds ammonia.
The abo e discussed esul s sugges ha he esponse o
he senso s can be inc eased by he dec ease o he deposi-
ion empe a u e below 600
°
C. Howe e , he p elimina y
expe imen s ca ied ou o he NFC ac i e laye s syn he-
sized a 500
°
C and 550
°
C, showed ha hese senso s
possessed ex emely high esis ance and we e almos wi h-
ou esponse o NH
3
. So much educed empe a u es had a
nega i e effec on he educ ion o ca aly ic NPs p io o
he CNT and CNF g ow h. I esul ed in a e y low yield
o NFCs, and i on oxide and Fe
3
C NPs we e co e ed by
ca bon. I is wo h no ing ha no i on was de ec ed on he
su ace o ou samples by XPS analysis e en o he deposi-
ion empe a u e o 600
°
C. In his connec ion, he i on
nanopa icles ha e no di ec con ac wi h ammonia and
ha e no influence on he esponse om he poin o adso p-
ion. The ca aly ic nanopa icles can change he conduc i i y
o ca bon nanoma e ials bu only o ca bon nanoma e ials
syn hesized using low deposi ion ime, e.g., 10 min, o low
empe a u e (600
°
C). Thei p esence in he ac i e ma e ial
inc eases he esis ance o he ac i e laye bu only o
limi ed samples, and we supposed ha i has a li le effec
on he senso esponse, because conduc i i y in such ma e-
ials is mainly de e mined by he numbe o con ac s o med
be ween he nano ube and nanofibe ne wo k ( his effec can
be ea ed as pe cola ion).
The esponse o he senso s a 100 ppm, 250 ppm, and
500 ppm NH
3
was in he anges o 0.1–1.8%, 0.2–2.4%,
and 0.2–2.8%, espec i ely. In compa ison wi h p e iously
published esul s, such as Cui e al. [29] ob ained he
esponse o 2.8% o ba e MWCNTs unde exposu e o
1% o NH
3
(i.e., 10000 ppm). Hoa e al. [42] achie ed
app oxima ely 8% esponse o 6% o NH
3
(60000 ppm)
using CNTs on anodized alumina empla e; ou NFC hin
films exhibi ed he 2.8% esponse unde much lowe expo-
su e o 0.05% (500 ppm) o NH
3
. This is compa able o a
mo e expensi e SWCNT-/cellulose-based senso [16] o
chemically modified CNTs epo ed by Randeniya e al.
[4], which showed 1–2% esis ance g ow h o an acid-
ea ed CNT ya n senso unde 550 ppm o ammonia. The
senso s discussed in his wo k we e no su ace ea ed,
and i is expec ed ha some su ace ea men will imp o e
u he hei sensing p ope ies. We ecen ly epo ed ha
plasma ea men o PECVD-g own CNTs could enhance
hei esponse [43].
The adso p ion on ammonia on he su ace can be
di ided in o wo ca ego ies: physiso p ion and chemiso p-
ion. Banno e al. [43] discussed ha he esponse cu e is
a combina ion o bo h ca ego ies; he fi s s eep pa is ela ed
o physiso p ion o ammonia on he s uc u e su ace and
la e , an asymp o ic pa o chemiso p ion. In case o he
eco e y p ocess, he esponse can quickly eco e by he
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Room empe a u e
500 ppm
250 ppm
ΔR/R0, %
Time, min
100 ppm
200 ºC
0 102030405060
(a)
600 620 640 660 680 700
1
2
3
Deposi ion empe a u e, ºC
100 ppm
( oom empe a u e) 100 ppm (200 ºC)
250 ppm
( oom empe a u e) 250 ppm (200 ºC)
500 ppm
( oom empe a u e) 500 ppm (200 ºC)
ΔR/R0, %
(b)
Figu e 6: (a) Response cu es o he CNF600-10 sample o ammonia a oom empe a u e and 200
°
C. Ammonia exposu e pe iod and i s
concen a ion a e depic ed in he g aph. (b) CNF600-10 senso esponse a he oom empe a u e (solid symbols) and 200
°
C (open
symbols) as unc ion o he NFC deposi ion empe a u e o he deposi ion ime o 10 min.
8 Jou nal o Senso s
deso p ion o physiso ped molecules bu he eco e y o che-
miso bed molecules is slow o i eco e able. Fo ou films,
we p opose ha he main esponse mechanism is physiso p-
ion and can be modelled by Langmui iso he m. I is wo h
no ing ha i is no possible o sepa a e chemiso p ion and
physiso p ion. Ne e heless, we canno exclude he pa ial
chemiso p ion, since he e is an incomple e eco e y o sen-
so esponse wi hou using hea ing. The use o he mal
eco e y makes i possible o ully ca y ou he deso p ion
o ammonia molecules.
The Langmui iso he m is usually used o he desc ip-
ion o NH
3
adso p ion on ca bon ma e ials aking in o
accoun ew assump ions: each si e holds each molecule, all
si es a e equi alen , and he e is no in e ac ion be ween
NH
3
molecules on si es [44]. The adso p ion equilib ium
cons an can be es ima ed by he ollowing equa ion:
θ=θ∞
K⋅p
1+ K⋅p, 2
whe e θis he ac ion o ac i e si es co e ed by NH
3
(wi h
he assump ion ha he alue is p opo ional o he change
o esis ance), θ∞is he o al numbe o ac i e si es, Kis
he adso p ion equilib ium cons an , and pis he pa ial
p essu e o NH
3
. Resul s o he sample fi ing by adso p ion
iso he m a e shown in Figu e 7(b). The fi ing was p esen ed
aking in o accoun ha he concen a ion o ac i e si es is
p opo ional o he senso esis ance. Acco ding o he
fi ing, he adso p ion cons an o K=005 Pa
−1
was diffe en
compa ed wi h he epo ed one o g aphene-like ma e ials
[45] (K=016 Pa
−1
) and he g aphene-based NH
3
senso
epo ed in [46].
The senso s s udied can be used as indus ial senso s o
he de e mina ion o ammonia concen a ion in chemical
enginee ing (appa a uses), oil efining, e c. The da a on fi -
ing he expe imen al da a using Langmui iso he m makes
i possible o find he esponse o low NH
3
concen a ions
ha makes i possible o p edic he senso esponse o
14 ppm (acco ding o OSHA egula ions) and below ha is
app op ia e o en i onmen al con ol senso s. I is wo h
no ing ha he e is a necessi y o con ol also he highe con-
cen a ions (1000–20000 ppm) o u ning on he eme gency
en ila ion in comp esso ooms a he acili ies used anhy-
d ous ammonia and he senso s s udied can be also used
o his pu pose.
3.4. Influence o he Ope a ing Tempe a u e on NFC Film
Response and I s Reco e y. NFC-based senso s possessed
good esponse a oom empe a u e ha is an ad an age as
compa ed wi h semiconduc o -based con en ional senso s
[47]. The senso s wo king a oom empe a u e we e also p e-
pa ed using g aphene-based ma e ials. Fo example, Ka ko
e al. [45] c ea ed a fluo ine- unc ionalized g aphene senso
and eached 10.2% esponse o 10000 ppm o ammonia. The
esponse 3–4% o 0.1% (1000 ppm) NH
3
in A was ob ained
on educed g aphene oxide-sil e nanowi es [48].
The senso esponses a oom empe a u e and 200
°
C
we e qui e diffe en . The dependence o ΔR/R0on he con-
cen a ion had almos linea beha iou a 200
°
C, whe eas i
became nonlinea a oom empe a u e (Figu e 6(b)). The
maximum esponse a 200
°
C was 2.5% in compa ison wi h
2.8% a oom empe a u e. The enhanced senso esponse
a he oom empe a u e can be explained by he he mody-
namics o adso p ion. The adso p ion is an exo he mic p o-
cess, and he e o e, an inc ease o empe a u e enhances he
deso p ion and highe empe a u es (compa ed o oom
empe a u e) a e a ou able o his p ocess. Addi ionally, i
can be caused by weake deso p ion a oom empe a u e ha
lead o accumula ion o small amoun o ammonia on NFC
su ace. Fo example, he eco e y pe cen age (measu ed a e
10 min eco e y in syn he ic ai ) o he senso CNF600-10 o
100 ppm a oom empe a u e is only 23% in compa ison wi h
62% a 200
°
C. The ull adso p ion-deso p ion cycle o wide
ange o concen a ions o NH
3
(50–500 ppm) can be seen
in Figu e 7(a). We can see ha al hough he eco e y a
200
°
C was significan ly be e han a oom empe a u e, we
can obse e a highe concen a ion sa u a ion o he NFC
hin film esponse an inc ease o backg ound esis ance.
0
2
4
6
500 ppm
375 ppm
300 ppm
250 ppm
175 ppm
100 ppm
ΔR/R0, %
Time, min
50 ppm
0 20 40 60 80 100 120 140 160
(a)
0 1020304050
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
Pa ial p essu e, Pa
ΔR/R0 = 3.54⁎0.05⁎p/(1+0.05⁎p)
ΔR/R0, %
(b)
Figu e 7: CNF600-10 esponse cu e o NH
3
in a wide concen a ion ange a 200
°
C and da a fi ing by Langmui adso p ion iso he m
(R2=0984).
9Jou nal o Senso s