Journal of Molecular Modeling (2018) 24:242

https://doi.org/10.1007/s00894-018-3782-9

ORIGINAL PAPER

Density-functional study of hydrogen cyanide adsorption on silicene

nanoribbons
Gurleen Kaur Walia 1 & Deep Kamal Kaur Randhawa 1

Received: 27 May 2018 / Accepted: 1 August 2018

# Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Adsorption of toxic hydrogen cyanide gas (HCN) on armchair silicene nanoribbons (ASiNRs) is investigated by the first
principles method using density functional theory (DFT) to compute geometric, electronic, and transport properties. Two variants
of ASiNRs are considered: pristine ASiNR (P-ASiNR) and defective ASiNR (D-ASiNR), which is created by introducing a
vacancy in P-ASiNR by removal of a Si atom. Total energy optimizations are used to find the most stable structures. The
calculated results reveal that although HCN is physisorbed in both variants, sensitivity in the case of D-ASiNR is sufficiently
enhanced owing to more adsorption energy and charge transfer between the ASiNR-gas complex. Also, the inspection of current-
voltage characteristics demonstrates that the introduction of defect has considerably increased the conductivity of ASiNR. Hence,
D-ASiNR may be used as a promising sensor for HCN gas.

Keywords Silicene nanoribbons . DFT . Hydrogen cyanide adsorption . Gas sensor

Introduction explored in the literature [21–31]. In this work, we report a

Today, the world has great industrial infrastructure in almost
all types of industries like petroleum, chemicals, fertilizers,
metals, food, etc., resulting in pesticides, fuels, detergents,
industrial effluents, especially poisonous gases, as a conse-
quence. The much needed development in effect has led to
pollution of various kinds. To deal with this environmental
concern, there is need for detection of pollutants down to the
molecule level. To date, many materials have been utilized as
gas sensors [1–9]. Silicene, a hexagonally arranged 2D allo-
trope of silicon, analogue to graphene, is the focus of research
owing to its tremendous properties and applications in the
nanoelectronics field [10–13]. Silicene is a zero band gap
material, but owing to its distinguishing buckling structure,
its band gap can be easily tuned up by a number of methods
[14–20]. The capability of silicene as a sensor has been
* Gurleen Kaur Walia
gurleenwalia.13@gmail.com
Deep Kamal Kaur Randhawa
deepkamal.ecejal@gndu.ac.in
1
Department of Electronics and Communication Engineering, Guru
Nanak Dev University, Regional Campus, Jalandhar 144007, India
first-principles simulation of the interaction of poisonous gas,
hydrogen cyanide (HCN), and armchair silicene nanoribbons
(ASiNRs). Till date, ASiNRs have not been explored as a
HCN sensor to the best of our knowledge.
HCN is a toxic and flammable liquid which boils slightly
above room temperature, i.e., 25.6 °C. HCN gas is colorless
and extremely poisonous. It has a characteristic bitter-almond
like odor. It is a lethal agent which may lead to death by
blocking the ability of the body to use oxygen [32, 33].
HCN is found in the smoke of various tobacco products and
released by the combustion of nitrogen-containing organic
materials. It has a linear molecular geometry with the carbon
atom triply bonded to the nitrogen atom, i.e., H−C≡N:. The
bond length between the H atom and C atom is 1.064 Å,
whereas between the C atom and N atom, it is 1.156 Å.
Hence, sensing of this extremely toxic gas is mandatory for
atmospheric and environmental safety. So there is need for
sensors, whose electrical properties are tuned by adsorption
of toxic gas and produce an alarm signal. HCN’s interaction
with graphene, graphene oxides, graphyne, CNT, BNNT, and
AlN has been explored in the literature [34–39]. For the effec-
tive detection of cyanide, many analytical methods have been
developed, such as chromatographic [40, 41], electrochemical
[42, 43], fluorimetric [44–46], and luminescent analyses
[47–49]. However, these methods were very expensive and
242 Page 2 of 5
Fig. 1 The most stable configurations of HCN adsorption on a P-ASiNR b D-ASiNR
time consuming. Also, optical and piezoelectric-device bases
mass sensors have been devised for sensing HCN [50–53].
Silicene nanoribbons play a great role in detecting this toxic
gas. In this study, geometric, electronic, and transport proper-
ties of pristine armchair silicene nanoribbons (P-ASiNRs), as
well as defective armchair silicene nanoribbons (D-ASiNRs),
have been calculated on HCN adsorption using density func-
tional theory (DFT).
Computational details
All calculations are performed using Virtual NanoLab
Atomistix ToolKit (VNL-ATK) package provided by
Quantumwise [54]. VNL-ATK is a DFT [55] based ab initio
tool, using nonequilibrium Green’s function (NEGF) [56].
Electron temperature has been maintained at 300 K (27 °C).
Density mesh cut-off has been selected as 150 Rydberg. The
local density approximation with Perdew-Zunger (LDA.PZ)
is employed as the exchange-correlation functional along with
double-zeta polarized (DZP) basis sets [57, 58]. All simula-
tions are carried out self-consistently. The prime objective of
this study is to analyze the sensing capabilities of pristine and
defective armchair silicene nanoribbons in the presence of
HCN gas in terms of geometric, electronic, and transport prop-
erties. All the structures are allowed to relax till the values of
Hellman-Feynman force between the atoms and stress
Table 1 The calculated adsorption energy without (Ead) and with
counterpoise correction (Ead C.P.), charge transfer (Δq), electronic band
gap (Eg), and binding distance (D), which is the shortest atom to atom
distance between molecule and ASiNR
Device Ead (eV) Ead C.P. (eV) Δq (e) Eg (eV) D (Å)
HCN on P-ASiNR −0.65 −0.41 −0.016 0.4273 3.5
HCN on D-ASiNR −1.15 −0.98 0.056 0.1699 2.63
The negative value of charge indicates a charge transfer from the gas
molecule to the ASiNR sheet and vice-versa
J Mol Model (2018) 24:242
tolerance are less than 0.05 eV/Å and 0.0006 eV/Å3, respec-
tively. K-point samplings of 1×1×12 and 1×1×100 using
Monkhorst–Pack grid are used for geometry optimization
and electronic and transport calculations, respectively [29].
The nanoribbons considered under study are nine atoms wide
with buckling height of 0.44 Å. As the edges are more reactive
than the surface [59], hence, the dangling bonds at the edges
of ASiNRs are passivated by hydrogen atoms.
Results and discussion
In order to verify the sensitivity of our selected sensor mate-
rial, P-ASiNR is considered at first. A single gas molecule of
HCN is placed at a distance of about 3 Å from the sensor. This
structure is then allowed to fully relax. The most stable con-
figuration is obtained by total energy optimization. There is
minimal change in the structure of pristine silicene
nanoribbons after HCN adsorption as shown in Fig. 1a.
HCN now sits away at a distance of 3.5 Å from the surface.
The bond angle of HCN changes to 174.4°, whereas the bond
lengths between H−C and C−N become 1.09 Å and 1.162 Å,
respectively. Also the Si−Si bond length value is as low as
2.25 Å, with maximum value of lattice constant amounting to
3.93 Å, standard values being 2.29 Å and 3.88 Å, respectively.
Further, a defect is created by removing a silicon atom from
the P-ASiNR and HCN is allowed to adsorb on this D-ASiNR.
This leads to structural deformation of the ASiNR, with Si−Si
bond length as low as 2.23 Å and lattice constant as high as
4.28 Å. The HCN molecule places itself at a distance of
2.63 Å from D-ASiNR. The bond angle for HCN is 178.6°,
with bond lengths H−C and C−N amounting to 1.105 Å and
1.162 Å, respectively. Basis set superposition error (BSSE) is
considered to calculate the adsorption energy.
The electronic (adsorption energy, charge transfer, band
structure, and density of states) and transport (current-
voltage characteristics and transmission eigenstates)
J Mol Model (2018) 24:242
Fig. 2 Electronic band structures of a P-ASiNR b D-ASiNR after HCN adsorption
properties of the ASiNR-HCN complex are studied using
DFT and NEGF formalism.
Electronic properties
Adsorption energy and charge transfer
The more negative value of adsorption energy (Ead) is indic-
ative of stronger complexation and vice-versa. It is given as
follows:
E ad ¼ E ðASiNRþHCN Þ −EðASiNRÞ −E ðHCN Þ
where Ead is the adsorption energy of HCN on silicene
nanoribbons, E (ASiNR + HCN), E (ASiNR), and E (HCN) are the
total energies of fully optimized ASiNR-HCN complex,
ASiNR, and HCN, respectively. Table 1 shows the values of
calculated adsorption energies and charge transfers between
the gas molecule and ASiNRs. The adsorption energy of HCN
Fig. 3 Comparison of Density of states between P-ASiNR and D-ASiNR
after HCN adsorption
Page 3 of 5 242
on P-ASiNR indicates that the gas molecule is weakly sensed.
Also, the charge transfer is found using Mulliken population,
which also contributes to the weak sensing of this gas mole-
cule by P-ASiNR. On the other hand, Ead and charge transfer
values of HCN on D-ASiNR are considerably high, suggest-
ing increased sensitivity due to the introduction of defect.
Undoubtedly, there is no bond formation between the gas
and D-ASiNR, still the molecule comes closer to the sensor
after optimization having considerable charge transfer, provid-
ing enough sensitivity for the molecule to be sensed as com-
pared with its adsorption on P-ASiNR. Conclusively, HCN is
ð1Þ physisorbed on both ASiNR variants, with enhanced sensitiv-
ity in the case of D-ASiNR.
Electronic band structures and density of states
Figure 2 presents the energy band gap of ASiNRs on adsorp-
tion of HCN. Fermi level is shown as horizontal dotted line
and set to zero energy. Pristine silicene is a semimetal, having
Fig. 4 Current-voltage curve
242 Page 4 of 5
Fig. 5 Transmission eigenstates
of a P-ASiNR b D-ASiNR after
HCN adsorption
zero band gap. But, without a band gap, there are fundamental
obstructions for logic applications. Therefore, band gap needs
to be tuned somehow. It can be induced by introduction of
defect and adsorption of gases, due to which it starts behaving
as a semiconductor and finds practical applications in elec-
tronics. Band gap increases to 0.4273 eV when HCN is
adsorbed on P-ASiNR, while it becomes 0.1699 eV when
adsorbed on D-ASiNR, thereby increasing conductivity.
Further, the total electronic density of states are calculated
to verify the effects on electronic properties of ASiNRs due to
HCN adsorption as shown in Fig. 3. A high value of DOS
implies that there are more available states to be occupied. It
simply means that when the band gap for a particular energy
range is zero, the corresponding DOS value is also zero. This
is in compliance with Figs. 2 and 3. Also, DOS can be calcu-
lated mathematically as [60]:
DðE Þ ¼ dN ðEÞ=d ðE Þ ð2Þ
where N designates the number of electrons with an energy E
within the energy range dE, and the number of electrons de-
pends on the energy. Equation (2) symbolizes that DOS is
inversely proportional to the slope of energy.
Transport properties
For the calculation of transport properties, the two-probe mod-
el has been formulated, consisting of silicene nanoribbons
contacted by two silicene electrodes.
I-V characteristics and transmission eigenstates
To carry out the quantitative analysis of transport properties of
ASiNR, the current-voltage analysis is done. The voltage bias
is swept from 0 to 2 V in steps of 0.2 V. Figure 4 picturizes the
nonlinear behavior of ASiNRs. The conductivity of HCN on
P-ASiNR is quite low as compared with that on D-ASiNR,
which is directly in compliance with the adsorption energy
and charge transfer values. The I-V characteristics have been
calculated using Landauer formula [61]:


I D ¼ 2e2 =h ∫T ðE; V D Þ ½ f 1 ðE Þ−f r ðE Þ dE ð3Þ
J Mol Model (2018) 24:242
where e is the electron charge, h is the Planck’s constant, E is
the electron energy, VD is the bias voltage, and fl(E) and fr(E)
denote the Fermi distribution functions at the left and right
electrodes, respectively. T (E, VD) is the transmission coeffi-
cient as a function of E and VD.
The transmission eigenstates, which are the eigenfunctions
of the transmission matrix in real space, are also considered
for further investigations. Whole calculations are performed at
a bias voltage of 1.6 V with an isovalue of 0.2. Isosurface plots
have been considered, which represent the absolute value of
the wave function. Figure 5a and b shows the transmission
eigenstates of P-ASiNR and D-ASiNR respectively with HCN
adsorbed on them. It can be noticed that for D-ASiNR, the
transmission eigenstates are delocalized, symbolizing the
electron flow from left electrode to right electrode with a
high-transmission probability. Contrariwise, for P-ASiNR,
the eigenstates are strongly localized, signifying low transmis-
sion probability of an electron. Ergo, the probability of an
electron to pass through P-ASiNR is comparatively low.
Conclusions
In this work, the study of interaction between two ASiNR
variants and the HCN gas molecule has been investigated.
The geometric structures, electronic, and transport properties
of ASiNRs are explored in the presence of the HCN gas mol-
ecule using DFT calculations. The calculated results indicate
that HCN interacts with P-ASiNR via van der Waals forces,
indicating P-ASiNR is not very sensitive in detecting the gas
molecule. In order to enhance the sensitivity, we further inves-
tigated D-ASiNR by creating a vacancy defect in P-ASiNR. It
is found that there is considerable increase in the sensitivity of
ASiNR, owing to higher adsorption energy and charge trans-
fer. HCN adsorption on D-ASiNR is energetically more favor-
able than P-ASiNR. Even the band structure diagrams imply
that the band gap of the P-ASiNR and HCN complex is
0.4723 eV, which reduces to 0.1699 eVon introducing a defect
into P-ASiNR, thereby changing the device into a narrow
band semiconductor with increased conductivity. The nega-
tive value of charge transfer shows that the HCN gas molecule
behaves as an electron donor when adsorbed on P-ASiNR and
J Mol Model (2018) 24:242 Page 5 of 5 242

vice-versa in the case of D-ASiNR. The current-voltage char- 26. Drummond ND, Zolyomi V, Fal’ko VI (2012) Phys Rev B Condens
Matter 85:075423
acteristics also exhibit enhanced conductivity in the case of D-
27. Amorim RG, Scheicher RH (2015) Nanotechnol 26:154002
ASiNR. Hence, it can be concluded that D-ASiNR behaves as 28. Walia GK, Randhawa DKK (2018a) Struct Chem 29:257. https://
a suitable sensor for the toxic HCN gas molecule. doi.org/10.1007/s11224-017-1025-9
29. Walia GK, Randhawa DKK (2018b) Surf Sci 670:33. https://doi.
Acknowledgments The authors would like to thank Quantumwise for org/10.1016/j.susc.2017.12.013
their valuable support. The corresponding author wants to acknowledge 30. Walia GK, Randhawa DKK (2018c) J Mol Model 24:94. https://
the University Grants Commission, New Delhi, India, for Senior doi.org/10.1007/s00894-018-3631-x
Research Fellowship. This study was funded by the Department of 31. Walia GK, Randhawa DKK (2018d) Struct Chem. https://doi.org/
Science & Technology (DST) of India - Promotion of University 10.1007/s11224-018-1170-9
Research and Scientific Excellence (PURSE) scheme. 32. Ansell M, Lewis FA (1970) J Forensic Med 17:148
33. Aitken D, West D, Smith F, Poznanski W, Cowan J, Hurtig J,
Peterson E, Benoit B (1977) Can Anaesth Soc J 24:651
34. Rastegar SF, Peyghan AA, Hadipour NL (2013) Appl Surf Sci 265:
References 412
35. Zhao M, Yang F, Xue Y, Xiao D, Guo Y (2014) J Mol Model 20:
1. He Q, Zeng Z, Yin Z, Li H, Wu S, Huang X, Zhang H (2012) Small 2214
8:2994 36. Beheshtian J, Peyghan AA, Bagheri Z, Tabar MB (2014) Struct
2. Late DJ, Huang Y-K, Liu B, Acharya J, Shirodkar SN, Luo J, Yan Chem 25:1
A, Charles D, Waghmare UV, Dravid VP (2013) ACS Nano 7:4879 37. Wu RQ, Yang M, Lu YH, Feng YP (2008) J Phys Chem C 112:
3. Huo N, Yang S, Wei Z, Li S-S, Xia J-B, Li J (2014) Sci Rep 4:5209 15985
4. O’Brien M, Lee K, Morrish R, Berner NC, McEvoy N, Wolden 38. Wang R, Zhang D, Liu Y, Liu C (2009) Nanotechnology 20:505704
CA, Duesberg GS (2014) Chem Phys Lett 615:6 39. Beheshtian J, Peyghan AA, Bagheri Z (2013) J Mol Model 19:2197
5. Kou L, Frauenheim T, Chen C (2014) J Phys Chem Lett 5:2675 40. Aguilar M, Farran A, Marti V (1997) J Chromatogr A 778:397
6. Abbas AN, Liu B, Chen L, Ma Y, Cong S, Aroonyadet N, Köpf M, 41. Christison TT, Rohrer JS (2007) J Chromatogr 1155:31
Nilges T, Zhou C (2015) ACS Nano 9:5618 42. Shan D, Mousty C, Cosnier S (2004) Anal Chem 76:178
7. Peyghan AA, Noei M, Yourdkhani S (2013) Superlattice Microst 43. Lindsay AE, O’Hare D (2006) Anal Chim Acta 558:158
59:115 44. Lee K-S, Kim H-J, Kim G-H, Shin I, Hong J-I (2008) Org Lett 10:
8. Behmagham F, Vessally E, Massoumi B, Hosseinian A, Edjlali L 49
(2016) Superlattice Microst 100:350
45. Chung YM, Raman B, Kim D-S, Ahn KH (2006) Chem Commun
9. Nagarajan V, Chandiramouli R (2017) Superlattice Microst 101:
2:186
160
46. Ekmekci Z, Yilmaz MD, Akkaya EU (2008) Org Lett 10:461
10. Ni Z, Liu Q, Tang K, Zheng J, Zhou J, Qin R, Gao Z, Yu D, Lu J
47. Jin WJ, Fernandez-Arguelles MT, Costa-Fernandez JM, Pereiro R,
(2012) Nano Lett 12:113
Sanz-Medel A (2005) Chem Commun 7:883
11. Tao L, Cinquanta E, Chiappe D, Grazianetti C, Fanciulli M, Dubey
48. Jin WJ, Costa-Fernandez JM, Pereiro R, Sanz-Medel A (2004)
M, Molle A, Akinwande D (2015) Nat Nanotechnol 10:227
Anal Chim Acta 522:1
12. Tsai WF, Huang C-Y, Chang T-R, Lin H, Jeng H-T, Bansil A (2013)
49. Touceda-Varela A, Stevenson EI, Galve-Gasion JA, Dryden DTF,
Nat Commun 4:1500
Mareque-Rivas JC (2008) Chem Commun 17:1998
13. Sadeghi H, Bailey S, Lambert CJ (2014) Appl Phys Lett 104:
103104 50. Jawad SM, Alder JF (1991) Anal Chim Acta 246:259
14. Gao N, Zheng WT, Jiang Q (2012) Phys Chem Chem Phys 14:257 51. Yang M, He J, Hu X, Yan C, Cheng Z (2011) Environ Sci Technol
15. Lopez-Bezanilla A (2014) J Phys Chem C 118:18788 45:6088
16. Gao N, Li JC, Jiang Q (2014) Chem Phys Lett 592:222 52. Yang M, He J, Hu X, Yan C, Cheng Z, Zhao Y, Zuo G (2011)
17. Pan F, Wang Y, Jiang K, Ni Z, Ma J, Zheng J, Quhe R, Shi J, Yang J, Sensors Actuators B 155:692
Chen C, Lu J (2015) Sci Rep 5:9075 53. Yang M, He J, Hu X, Yan C, Cheng Z (2013) Analyst 138:1758
18. Aghaei SM, Monshi MM, Torres I, Calizo I (2016) RSC Adv 6: 54. QuantumWise (2015) Atomistix Toolkit version 2015.0.
17046 QuantumWise, Copenhagen. Available from http://www.
19. Aghaei SM, Calizo I (2015a) J Appl Phys 118:104304 quantumwise.com
20. Aghaei SM, Calizo I (2015b) In: Proceeding of IEEE SoutheastCon 55. Brandbyge M, Mozos J-L, Ordejón P, Taylor J, Stokbro K (2002)
(SECon-2015), Fort Lauderdale, 9–12 April 2015, pp 1–6 Phys Rev B 65:165401
21. Kara A, Enriquez H, Seitsonen AP, Voon LC, Vizzini S, Aufray B, 56. Taylor J, Guo H, Wang J (2001) Phys Rev B 63:245407
Oughaddou H (2012) Sci Rep 67:1 57. Zhao Q, Nardelli MB, Lu W, Bernholc J (2005) Nano Lett 5:847
22. Vogt P, De Padova DP, Quaresima C, Avila J, Frantzeskakis E, 58. Abadir GB, Walus K, Pulfrey DL (2009) J Comput Electron 8:1
Asensio MC, Resta A, Ealet B, Lay GL (2012) Phys Rev Lett 59. Dávila ME, Marele A, De Padova P, Montero I, Hennies F, Pietzsch
108:155501 A, Shariati MN, Gómez-Rodríguez JM, Lay GL (2012)
23. Zhao J, Liu H, Yu Z, Quhe R, Zhou S, Wang Y, Liu CC, Zhong H, Nanotechnology 23:385703
Han N, Lu J, Yao Y, Wu K (2016) Prog Mater Sci 83:24 60. Poole CP Jr, Owens FJ (2003) Introduction to nanotechnology.
24. Ding Y, Nia J (2009) Appl Phys Lett 95:083115 Wiley, New York
25. Spencer MJS, Morishita T (2016) Silicene. In: Springer series in 61. Büttiker M, Imey Y, Landauer R, Pinhas S (1985) Phys Rev B 31:
materials science, vol 235. Springer, Cham 6207