Chemical resistivity of self-assembled monolayercovalently attached to silicon substrate to hydrofluoric

acid and ammonium fluoride

N. Saito a,*, S. Youda a, K. Hayashi a, H. Sugimura a, O. Takai b

a Department of Materials Processing Engineering, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa,

Nagoya 464-8603, Japanb Center for Integrated Research in Science and Engineering, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8603, Japan

Abstract

Self-assembled monolayers (SAMs) were prepared on hydrogen-terminated silicon substrates through chemical vapor

deposition using 1-hexadecene (HD) as a precursor. The HD-SAMs prepared in an atmosphere under a reduced

pressure (� 50 Pa) showed better chemical resistivities to hydrofluoric acid and ammonium fluoride (NH4F) solutions

than that of an organosilane SAM formed on oxide-covered silicon substrates. The surface covered with the HD-SAM

was micro-patterned by vacuum ultraviolet photolithography and consequently divided into two areas terminated with

HD-SAM or silicon dioxide. This micro-patterned sample was immersed in a 40 vol.% NH4F aqueous solution. Surface

images obtained by an optical microscopy clearly show that the micro-patterns of HD-SAM/silicon dioxide were

successfully transferred into the silicon substrate.

� 2003 Elsevier Science B.V. All rights reserved.

Keywords: Self-assembly; Chemical vapor deposition; Silicon; Halides

1. Introduction

Self-assembled monolayers (SAMs) on silicon

substrates are a promising material for applica-

tions to electronic devices, organic templates, resistfilms and so fourth [1–4]. In particular, organosi-

lane SAMs have been most widely studied and

used in such applications. These SAMs are formed

on the basis of the dehydration between hydroxyl

groups in a precursor and surface silanol groups

introduced onto native silicon dioxide on silicon

substrates. The siloxane network, which is formed

as the result of dehydration, binds between organic

layer and oxide-covered silicon. However the

monolayer does not have good chemical resistivi-ties particularly to hydrofluoric acid (HF) and

ammonium fluoride (NH4F) solutions due to the

presence of the siloxane network which is readily

damaged with such solutions. The organic mono-

layer covalently attached to silicon substrate

through Si–C bonds without inserting native oxide

layer is expected to have better chemical resistivi-

ties. In addition, the Si–C interface may providefavorable electronic properties to molecular de-

vices on silicon substrate. Many researchers have

*Corresponding author. Fax: +81-52-789-2796.

E-mail address: nagahiro@plasma.numse.nagoya-u.ac.jp

(N. Saito).

0039-6028/03/$ - see front matter � 2003 Elsevier Science B.V. All rights reserved.

doi:10.1016/S0039-6028(03)00158-4

Surface Science 532–535 (2003) 970–975

www.elsevier.com/locate/susc

reported on structural configurations and chemical

bonding states of such SAMs due to these benefits

[5–12]. However, chemical resistivities of the

SAMs to HF and NH4F solutions have not been

studied in detail.

In this study, we have prepared 1-alkene SAMon hydrogen-terminated Si(1 1 1) substrates

through chemical vapor deposition in an atmo-

sphere under a reduced pressure or in air. Chem-

ical resistivities of the SAM to HF and NH4F

solutions are elucidated.

2. Experiments

Fig. 1 shows a schematic diagram of the exper-

iments in this study; (a) SAM preparation and

chemical resistivity tests, (b) application of the

SAM to micro-patterning of silicon substrates. In

process (a1), silicon (1 1 1) substrates (p-type) were

cleaned ultrasonically in acetone, methanol and

deionized water in that order. The substrates werefurther etched in aqueous HF solution (5 vol.%) at

70 �C in order to remove surface oxide and to

terminate with hydrogen. In process (a2), 1-hexa-

decene (HD, Tokyo Kasei Co Ltd.) was used as a

precursor. The hydrogen-terminated Si(1 1 1) (Si–

H) substrates were alkylated as follows. The Si–H

substrate and 1-hexadecene were sealed into an

autoclave of which volume was 100 cm3. Thevolume of 1-hexadecene put into the autoclave

was determined to be 4 l l in order to avoid

being condensed in the autoclave. After the auto-

clave modified for the use under reduced pres-

sures had been evacuated down to at a pressure

of 50 Pa, it was heated to 150 �C. Hexadeceneliquid vaporized and reacted with Si–H groups

on the substrate, resulting in formation of analkyl monolayer (BSiH+CH2@ CHC14H29!BSiC16H33). In addition, the alkylation was con-

ducted in air a well in order to reveal the influence

of oxygen in the alkylation. After the alkylation,

the silicon substrates were ultrasonically cleaned in

toluene, methanol and deionized water in that

order. In process (a3), the resistivities of the or-

ganic films to aqueous HF and NH4F solutionswere examined. The concentrations of HF and

NH4F were 5 and 40 vol.%, respectively. Water

Fig. 1. Schematic diagram of the experimental procedures: (a)

preparation and characterization of 1-hexadecene SAM and (b)

application of 1-hexadecene SAM as a resist film to aqueous

NH4F solution.

N. Saito et al. / Surface Science 532–535 (2003) 970–975 971

contact angles of the etched samples were mea-

sured.

In process (b), HD-SAM was applied to a resist

film for micro-fabrication of silicon. In process

(b1), the SAM was micro-patterned by vacuum

ultraviolet (VUV) light. The source of light used inthis study was an excimer lamp with k ¼ 172 nm

and 10 mW/cm2 (Ushio Electric, UER20-172V).

The surface covered with HD-SAM was irradiated

under a reduced pressure of 10 Pa with VUV light

through a photomask for 20 min. This process

decomposed alkylchains of the SAM, oxidized the

underlying silicon substrate with active oxygen

species generated from atmospheric oxygen speciesby VUV excitation. The substrate surface was

divided into micro-patterned two regions, i.e. un-

dercomposed SAM and silicon oxide. In process

(c2), the substrate was immersed in 40% NH4F

solution at 25 �C. Microstructures were fabricatedon the silicon surface because Si etching was pro-

tected in the region covered with the undecom-

posed SAM.The water repellency of SAM surfaces was

evaluated by measuring their water-contact angles

with a sessile drop of deionized water in air using

an automatic contact anglemeter (CA-X150,

Kyowa Interface Science).

3. Results and discussion

Fig. 2 shows Si2p spectra before and after the

HF etching acquired by X-ray photoelectron

spectroscopy (XPS). Two peaks corresponding to

bulk Si (� 100 eV) and SiO2 (� 104 eV) were

confirmed in the spectrum before the HF etching.

On the other hand, only a peak of bulk Si was

detectable in the spectrum after the HF etching.The water contact angle of the substrate etched for

3 min was about 80�, which agrees with that of Si–H [13]. A root mean square roughness of this Si–H

was approximately 0.25 nm. These results show

that native oxide was completely removed from

the substrate and it was terminated with hydrogen.

Fig. 3 shows a typical relationship between

water contact angle and alkylation time. The watercontact angle of HD-SAM prepared at the reduced

pressure and in air reached saturated ones of 104�

Inte

nsity

[Arb

itrar

y U

nits

]106 104 102 100 98 96

Binding Energy [eV]

Si 2p

Hydrogen-terminatedsilicon

Silicon substratebefore HF treatment

SiO2

Si

Fig. 2. Si2p spectra of the silicon substrates obtained by XPS:

(a) before and (b) after HF treatment.

110

105

100

95

90

85

80

4003002001000

Hydrogen-terminated silicon

Preparation time / min

Wat

er c

onta

ct a

ngle

/ de

gree

104o

101o

CH3-termination

CH2-termination

in air at a reduced pressure

Fig. 3. Relationship between water contact angles and prepa-

ration time: (a) CVD in a reduced pressure (� 50 Pa) and (b)

CVD in air.

972 N. Saito et al. / Surface Science 532–535 (2003) 970–975

and 101� after the alkylation for 100 min, respec-tively. In our HD-SAM, molecular alkylchains are

assumed to be more highly oriented perpendicular

to the substrate and/or more densely packed than

those in similar SAMs reported previously, be-

cause the water contact angle of 104� is closer tothe ideal value of the CH3-terminated surface [14].

Fig. 4 shows Si2p and C1s XPS spectra before and

after the alkylation. In the case of the reduced

pressure, the Si2p spectra before and after the al-

kylation were not changed at all, indicating that

the alkylation proceeded without oxidizing the Si–

H substrate. Locations of the C1s peaks before and

after the alkylation are identical, while the inten-

sity increases due to the alkylation. The thicknessof the monolayer was approximately 1.5 nm,

which is approximately equal to that of a 1-hexa-

decene molecule. These results show that the al-

kylation of the Si–H substrate is accomplished

without forming the oxide. In the case of the al-

kylation in air, the spectra of Si2p after the alky-

lation have peaks of silicon (100 eV) and silicon

dioxide (104 eV). The silicon dioxide layer wasformed at the interface between the silicon sub-

strate and organic layer during the alkylation. The

spectra of C1s have a peak of carboxyl group

(COOH) other than hydrocarbons. The peak of

COOH group show that the reaction between

precursors and Si–H did not completely precede

and the precursors and hydrogen-terminated sili-

con oxidized partly. Thus the preparation in airprovides the oxidized organic layers with no Si–C

bonds. Here we refer to the former organic layers

as HD-SAM and to the later one as partially ox-

idized organic layer (POOL). The film thickness of

HD-SAM, which was roughly estimated by the

attenuation of Si2p spectra, was approximately 1.9

nm [15,16]. A semi-empirical molecular orbital

calculation with AM1 Hamiltonian shows the filmthickness is 2.1 nm when the tilt angle of the mo-

lecular chains is 0�. Thus, the film is a monolayer

with slight tilt angle.

Fig. 5(a) and (b) show the changes in water

contact angles when the HD-SAM and POOL

samples were immersed in aqueous 5 vol.% HF

and 40 vol.% NH4F solutions at 25 �C, respec-tively. For control experiments, an organosilaneSAM prepared from octadecyltrimethoxysiloxane

(ODS-SAM) on oxide-covered silicon substrates

was also examined in the same solutions. As

shown in Fig. 5(a), the water contact angle of

ODS-SAM drastically decreases down to less than

85� within 10 min. Within the first few minutes, thewater contact angles of ODS-SAM and POOL

decrease drastically, while that of HD-SAM de-creases a little and remains at around 98� evenafter immersing for 30 min in the HF solution.

Inte

nsity

[A

rbitr

ary

Uni

ts]

106 104 102 100 98 96

Binding Energy [eV]

Si 2p

SiO2

Si

Inte

nsi

ty [

Arb

itra

ry U

nit

s]

295 290 285 280

Binding Energy [eV]

C 1s -CO-

Organic layerprepared in air

Organic layerprepared at reduced pressure

Hydrogen-terminatedsilicon

-CH2-

Organic layerprepared in air

Organic layerprepared at reduced pressure

Hydrogen-terminatedsilicon

Fig. 4. (a) Si2p and (b) C1s XPS spectra of the silicon substrates

covered with organic layers prepared in air and under reduced

pressure, and with terminated with hydrogen.

N. Saito et al. / Surface Science 532–535 (2003) 970–975 973

Through this HF etching, ODS-SAM was con-

cluded to be removed almost completely. In case

of POOL, a part of the monolayer where was not

covalently attached to silicon was assumed to be

removed by the etching. As shown in Fig. 5(b), the

water contact angles of ODS-SAM and POOL

immersed in aqueous 40 vol.% NH4F solution

drastically decreases down to less than 85� as

similarly to the results in aqueous 5 vol.% HF

solution. However, the water contact angle of HD-SAM decreases slower than these monolayers. Its

water contact angle remains around 95� even after10 min. These chemical resistivities of HD-SAM

originate in the covalent Si–C bond.

The silicon substrates with the micro-patterns of

HD-SAM and silicon dioxide (Fig. 1(b)) were

immersed in aqueous 40 vol.% NH4F solution at

25 �C for 7 min. The areas irradiated by VUV lightare intermediately etched. While the areas covered

with HD-SAM protect the etching of the silicon

substrate. The micro-patterns are thus transferred

into the silicon substrate due to the difference in

the etching rates. Fig. 6 shows a surface image

obtained by optical microscopy. The microstruc-

tures of silicon were clearly seen in this image. The

image indicates that HD-SAM can be applied to a

Fig. 6. Optical micro-graph of the micro-fabricated silicon

surface.

110

105

100

95

90

85

80

14121086420

HD-SAM POOL ODS-SAM

in 40%NH4 F solution

Etching time / min

Wat

er c

onta

ct a

ngle

/ de

gree

Hydrogen-terminated silicon

110

105

100

95

90

85

80

35302520151050

Wat

er c

onta

ct a

ngle

/ de

gree

Etching time / min

HD-SAM POOL ODS-SAM

in 5%HF solution

Hydrogen-terminated silicon

(a)

(b)

Fig. 5. Water contact angles of HD-SAM, POOL and ODS-

SAM etched in (a) aqueous 5 vol.% HF and (b) aqueous 40

vol.% NH4F solutions.

974 N. Saito et al. / Surface Science 532–535 (2003) 970–975

resist film for silicon etching with aqueous NH4F

solution.

4. Conclusions

We have prepared the organic films of 1-hexa-

decene, HD, through chemical vapor deposition

at a reduced pressure and in air. Only the film

prepared at a reduced pressure became SAM im-

mobilized through S-C bonds as evidenced by the

XPS spectra and the film thickness. The HD-SAM

shows better chemical resistivities for aqueous HF

and NH4F solutions than the organosilane SAM.The chemical resistivities originated in the Si–C

bonds at the interface between the silicon substrate

and organic layer. Finally, we demonstrated that

HD-SAM was applicable to a resist film for micro-

fabrication of silicon substrates.

Acknowledgements

This work has been supported by the Research

Project ‘‘Biomimetic Materials Processing’’ (no.

JSPS-RFTF 99R13101), Research for the Future

(RFTF) Program, Japan Society for the Promo-

tion of Science.

References

[1] A. Ulman, An Introduction to Ultrathin Organic Films,

Academic Press, Boston, MA, 1991.

[2] A. Ulamn, Chem. Rev. 96 (1996) 1533.

[3] V. Chechik, R.M. Crooks, C.J.M. Stirling, Adv. Mater. 12

(2000) 1161.

[4] A.B. Sieval, R. Linke, G. Heji, G. Meijer, H. Zuilhof,

E.J.R. Sudholter, Langmuir 17 (2001) 7554.

[5] M.R. Linford, C.E.D. Chidsey, J. Am. Chem. Soc. 115

(1993) 12631.

[6] M.R. Linford, P. Fenter, P.M. Eisenberger, C.E.D. Chid-

sey, J. Am. Chem. Soc. 117 (1995) 3145.

[7] J. Terry, M.R. Linford, C. Wigren, R.Y. Cao, P. Pianetta,

C.E.D. Chidsey, Appl. Phys. Lett. 71 (1997) 1056.

[8] J. Terry, M.R. Linford, C. Wigren, R.Y. Cao, P. Pianetta,

C.E.D. Chidsey, J. Appl. Phys. 85 (1999) 213.

[9] R.L. Cicero, C.E.D. Chidsey, G.P. Lopinski, D.D.M.

Wayner, R.A. Wolkow, Langmuir 18 (2002) 305.

[10] E.E. Quayum, T. kondo, S. Nihonyanagi, D. Miyamoto,

K. Uosaki, Chem. Lett. 2 (2002) 208.

[11] D.D.M. Wayner, R.A. Wolkow, J. Chem. Soc., Perkin

Trans. Part 2 1 (2002) 23.

[12] J.M. Buriak, Chem. Rev. 102 (2002) 1271.

[13] E. Kondoh, M.R. Baklanov, F. Jonckx, K. Maex, Mat.

Sci. Semicon. Proc. 1 (1998) 107.

[14] A.B. Sieval, V.V.H. Zuihof, E.J.R. Sudholter, Langmuir 15

(1999) 828.

[15] Z. Li, M. Lieberman, Langmuir 17 (2001) 4887.

[16] S. Tanuma, C.J. Powell, D.R. Penn, Surf. Interface Anal.

21 (1993) 165.

N. Saito et al. / Surface Science 532–535 (2003) 970–975 975