PRAMANA c© Indian Academy of Sciences Vol. 83, No. 3— journal of September 2014

physics pp. 427–434

Thermal carbonization of nanoporous silicon: Formationof carbon nanofibres without a metal catalyst

GUNJAN AGGARWAL, PRABHASH MISHRA, BIPIN JOSHIand S S ISLAM∗Nano-Sensor Research Laboratory, Department of Applied Sciences and Humanities, F/O ofEngineering and Technology, Jamia Millia Islamia (A Central University), New Delhi 110 025,India∗Corresponding author. E-mail: safiul5996@gmail.com

MS received 9 April 2013; revised 11 August 2013; accepted 28 January 2014DOI: 10.1007/s12043-014-0773-y; ePublication: 29 June 2014

Abstract. An interesting phenomenon is observed while carrying out thermal carbonization of po-rous silicon (PS) with an aim to arrest the natural surface degradation, and it is a burning issuefor PS-based device applications. A tubular carbon structure has been observed on the PS surface.Raman, Fourier transform infrared spectroscopy (FTIR) and electron microscope studies, revealedthat the tubular structure is nothing but amorphous carbon nanofibres sprouted within the pores inthe absence of a metal catalyst, for which a suitable explanation is proposed.

Keywords. Porous silicon; electrochemical etching method; thermal carbonization; carbon nanofi-bre and stability.

PACS Nos 61.82.Rx; 73.63.Fg; 78.67.Rb

1. Introduction

Porous silicon (PS) nanostructure plays an important role in the area of material scienceand engineering and it has huge technological impact in nanoelectronic, field emission,photonic and nanosensor devices [1–5]. Over the years, researchers have reported PS for-mation by various techniques, notably chemical etching [6], electrochemical- and photo-induced electrochemical anodization [7], plasma etching [8] etc. The major drawback inall these cases is the surface degradation while being exposed to ambient atmosphere. Thishinders the possibility of PS from further use as a device. It has already been reported thatthe PS is oxidized when stored in ambient air and hence its electronic properties change[9,10].

Several treatments to stabilize the PS structure have been reported [11–17]. Thermaloxidation and thermal carbonization, in particular, are very efficient techniques reportedby many researchers for surface stability of PS [11–15]. One of the many methods to

Pramana – J. Phys., Vol. 83, No. 3, September 2014 427

Gunjan Aggarwal et al

stabilize the surface was through the use of Si–C bonds. Buriak et al first demonstrateda Lewis acid-mediated approach [17] to functionalize the PS surface, which resulted inincreasing stability. In previous published reports [15,16], when PS surface was thermallytreated above 400◦C, Si–H bonds broke with desorbed hydrogen and absorbed acetylenemolecule produced an SiC layer on the porous surface.

In this work, the authors report the formation of amorphous carbon nanofibres withinthe pores of PS during thermal carbonization process in the presence of C2H2. Simulta-neously, a stable surface is also achieved due to the carbonization process.

2. Experimental

PS samples were manufactured by electrochemical etching (figure 1) of boron-dopedp-type Si(1 0 0) wafers with a resistivity of 1–10 �·cm in an HF and ethanol mixture(1:1 by volume). The etching current density and time were fixed as 25 mA/cm2 and25 min, respectively.

The carbonization was carried out in a conventional tube furnace (figure 2). The ther-mal annealing pressure was kept at 1.5 mBar with a heating rate of 300◦C/h. During

Figure 1. Electrochemical etching cell.

Figure 2. CVD chamber setup.

428 Pramana – J. Phys., Vol. 83, No. 3, September 2014

Thermal carbonization of nanoporous silicon

the carbonization of PS, the argon gas flow rate was kept at 30 sccm. In the mean time,acetylene was introduced into the chamber with a flow rate of 60 sccm at 850◦C for15 min. Experimental protocol of the carbonization process is shown in figure 3a and theschematic diagram of carbonization process is shown in figure 3b.

Fourier transform infrared spectroscopy (FTIR) measurements were done with Brukervertex 70 V instrument. FTIR spectroscopy was characterized by diffuse reflectance mode

Figure 3. (a) Experimental protocol and (b) schematic diagram of carbonizationprocess.

Pramana – J. Phys., Vol. 83, No. 3, September 2014 429

Gunjan Aggarwal et al

to remove the influence of the substrate. For Raman measurement, we used micro-Ramanspectrophotometer (LabRAM HR800, JY) fitted with peltier-cooled CCD detector and anOlympus BX-41confocal microscope. The samples were excited with an air-cooled Ar-ion laser (Spectra Physics) tuned at 488 nm. The spot size was 1.19 μm at the samplesurface under optimal conditions. Measurements were carried out in the backscatteringgeometry using a 50X LWD microscope objective. The laser power was kept low on thesample surface to avoid excessive heating.

3. Results and discussion

3.1 SEM analysis

Field emission scanning electron microscope (FEI, Nova Nano SEM 450) was used forsurface structure analysis. The SEM image (figure 4a) indicates the presence of well-dispersed nanopore distribution of PS having diameter in the range of 40–60 nm. After thethermal carbonization process, the SEM results show the formation of a-C nanofibre-likestructure on the sample surface. HR-TEM image (figure 5) confirms the formation ofCNFs within the pores during thermal carbonization.

(a) (b)

Figure 4. SEM image of the sample (a) before thermal carbonization and (b) afterthermal carbonization.

Figure 5. HR-TEM image of thermally carbonized sample.

430 Pramana – J. Phys., Vol. 83, No. 3, September 2014

Thermal carbonization of nanoporous silicon

3.2 Raman analysis

To confirm the growth of a-C tubular structure, Raman analysis was done. Figure 6ashows the Raman spectra of freshly etched PS. The peak of PS shifts at 517 cm−1 and hasstrong asymmetry in Raman profile. Raman spectra of PS confirm (a) faster etching rate,(b) increase in porosity of PS and (c) decrease in silicon crystallite size [18]. Raman

Figure 6. Raman spectra of (a) freshly etched porous silicon and (b) thermallycarbonized sample.

Pramana – J. Phys., Vol. 83, No. 3, September 2014 431

Gunjan Aggarwal et al

results show that the smaller the mean crystallite size, the greater is the redshift andasymmetry in the lower energy of the LO phonon mode. Raman spectrum after car-bonization (figure 6b) shows the first-order Raman scattering (i.e., G-band) at 1585 cm−1,while the second-order Raman scattering (i.e., D-band) is at 1357 cm−1. The Ramanresult confirms the growth of a-CNF while carrying out thermal carbonization of PS.

The growth of carbon tubular structure has been reported earlier by Sun Wen et al[19] where alumina was deposited inside the pores to facilitate the growth of CNT as acatalyst. In our case, as such, no catalyst was deposited and we believe that acetylenemolecule decomposes at such high temperature and interacts with nanopore, which helpsas a catalyst in the formation of fibre-like structure of carbon atom.

In the Raman spectra (figure 6b), the intensity of the D-band vis-à-vis the G-band ismuch higher due to high degree of defect density with the CNFs and it is confirmed inthe HR-TEM micrograph (figure 5). So, to ascertain the distinction between CNT andCNFs, an analysis on G′-band is done and this approach was earlier implemented also byresearchers [20] where it was shown that for amorphous nanocarbon, G′-band (∼2680–2780 cm−1) should be absent in Raman spectra. In our sample, G′-band is also absent asshown in the given spectra (figure 6b).

3.3 FTIR analysis

FTIR studies have been performed in the range of 500 to 4000 cm−1 for identifying thefunctional group attached on the surface of the thermally carbonized sample. Figure 7shows FTIR spectrum of thermally carbonized sample. The peaks at 3848 and 3649 cm−1

are attributed to γ (Si–OH) stretching modes. At lower wave numbers, the C–O and C–Cabsorbencies at 1630 and 1260 cm−1 are present. Similarly, the Si–H modes at 858 cm−1

and various deformation modes overlapping the silicon crystal modes that are observed at667 cm−1 and 510 cm−1 are also present. The FTIR spectra show the C–C bonds formedby the carbonization process. The samples were periodically checked for six months and

Figure 7. FTIR spectra of thermally carbonized sample.

432 Pramana – J. Phys., Vol. 83, No. 3, September 2014

Thermal carbonization of nanoporous silicon

Figure 8. FTIR spectra shows the surface stability after six months.

it was found that all the specific modes mentioned above retain their characteristic peakpositions and intensities and this confirms the achievement of surface stability (figure 8).

4. Conclusion

A very interesting result has been seen while carrying out the thermal carbonization pro-cess of PS. SEM, Raman and FTIR studies of PS samples, which were thermally carboni-zed in acetylene gas, confirmed the growth of some tubular structure of a-C. It is con-cluded that the thermal carbonization process of PS was also advantageous for the growthof carbon nanofibre under proper growth conditions of CVD.

Acknowledgements

The authors gratefully acknowledge the financial support provided by the Department ofScience & Technology, Govt. of India, through its Grant No. SR/S2/CMP-0053/2009 tocarry out this work.

References

[1] R Martel, T Schmidt, H R Shea, T Hertel and P Avouris, Appl. Phys. Lett. 73, 2447 (1998)[2] S Chopra, K McGuire, N Gothard, A M Rao and A Pham, Appl. Phys. Lett. 83, 2280 (2003)[3] A Jorio, R Saito, T Hertel, R B Weisman, G Dresselhaus and M S Dresselhaus, MRS Bull. 29,

276 (2004)[4] S S Fan, W J Lian, H Y Dang, N Franklin, T Tombler, M Chapline and H J Dai, Physica E 8,

179 (2000)

Pramana – J. Phys., Vol. 83, No. 3, September 2014 433

Gunjan Aggarwal et al

[5] S Hofmann, C Ducati, B Kleinsorge and J Robertson, Appl. Phys. Lett. 83, 4661 (2003)[6] H Föll, M Christophersen, J Carstensen and G Hasse, Mater. Sci. Eng. R 280, 1 (2002)[7] Saakshi Dhanekar, S S Islam and Harsh, Nanotechnology 23, 235501 (2012)[8] C Z Gu, J J Li, Z L Wang, C Y Shi, P Xu, K Zhu and Y L Liu, J. Appl. Phys. 97, 093501

(2005)[9] H Koyama and P M Fauchet, Appl. Phys. Lett. 77, 2316 (2000)

[10] R J Martina-Palma, L Vazquez, P Herrero, J M Martinez-Duart, M Schell and S Schaefer, Opt.Mater. (Ámsterdam, Neth.) 17, 75 (2001)

[11] J Salonena, M Bjorkqvista, E Laine and L Niinisto, Appl. Surf. Sci. 225, 389 (2004)[12] A E Pap, K Kordás, T F George and S Leppävuori, J. Phys. Chem. B 108, 12744 (2004)[13] S T Lakshmikumar and P K Singh, J. Appl. Phys. 92, 3413 (2002)[14] J Salonen and E Laine, J. Appl. Phys. 91, 456 (2002)[15] A E Pap, K Kordás, G Tóth, J Levoska, A Uusimäki, J Vähakangas and S Leppävuori, Appl.

Phys. Lett. 86, 041501 (2005)[16] L T Canham, Properties of porous silicon (IEE INSPEC, London, 1997)[17] J M Buriak and M J Allen, J. Am. Chem. Soc. 120, 1339 (1998)[18] O Bisi, S Ossicini and L Pavesi, Surf. Sci. Rep. 38, 1 (2000)[19] S Wen, S Mho and I Yeo, J. Power Sources 163, 304 (2006)[20] R A DiLeo, B J Landi and R P Raffaelle, J. Appl. Phys. 101, 064307 (2007)

434 Pramana – J. Phys., Vol. 83, No. 3, September 2014