Platinum nanoflowers on scratched silicon by galvanic displacement for an effective SALDI substrate

Article abstract of DOI:10.1002/chem.201001038

We report a new and facile method for synthesizing 3D platinum nanoflowers (Pt Nfs) on a scratched silicon substrate by electroless galvanic displacement and discuss the applications of the Pt Nfs in surface-assisted laser desorption/ionization-mass spectrometry (SALDI-MS). Surface scratching of n-type silicon is essential to induce Pt Nf growth on a silicon substrate (to obtain a Pt Nf silicon hybrid plate) by the galvanic displacement reaction. The Pt Nf silicon hybrid plate showed excellent SALDI activity in terms of the efficient generation of protonated molecular ions in the absence of a citrate buffer. We propose that the acidity of the Si-OH moieties on silicon increases because of the electron-withdrawing nature of the Pt Nfs; hence, proton transfer from the Si-OH groups to the analyte molecules is enhanced, and finally, thermal desorption of the analyte ions from the surface occurs. Signal enhancement was observed for protonated molecular ions produced from a titania nanotube array (TNA) substrate on which Pt nanoparticles had been photochemically deposited. Moreover, surface modification of the Pt Nf silicon hybrid plate by perfluorodecyltrichlorosilane (FDTS) (to obtain an FDTS-Pt Nf silicon hybrid plate) was found to facilitate soft SALDI of labile compounds. More interestingly, the FDTS-Pt Nf silicon hybrid plate acts 1) as a high-affinity substrate for phosphopeptides and 2) as a SALDI substrate. The feasibility of using the FDTS-Pt Nf silicon hybrid plate for SALDI-MS has been demonstrated by using a β-casein digest and various analytes, including small molecules, peptides, phosphopeptides, phospholipids, carbohydrates, and synthetic polymers. The hybridization of Pt Nfs with a scratched silicon substrate has been found to be important for achieving excellent SALDI activity.

Full text of DOI:10.1002/chem.201001038

DOI: 10.1002/chem.201001038
Platinum Nanoflowers on Scratched Silicon by Galvanic Displacement for an
Effective SALDI Substrate**
Hideya Kawasaki,*[a] Teruyuki Yao,[a] Takashi Suganuma,[a] Kouji Okumura,[a]
Yuichi Iwaki,[a] Tetsu Yonezawa,[b] Tatsuya Kikuchi,[b] and Ryuichi Arakawa*[a]
Abstract: We report a new and facile
method for synthesizing 3D platinum
nanoflowers (Pt Nfs) on a scratched sil-
icon substrate by electroless galvanic
displacement and discuss the applica-
tions of the Pt Nfs in surface-assisted
laser desorption/ionization-mass spec-
on silicon increases because of the elec-
tron-withdrawing nature of the Pt Nfs;
by
perfluorodecyltrichlorosilane
(FDTS) (to obtain an FDTS-Pt Nf sili-
con hybrid plate) was found to facili-
tate soft SALDI of labile compounds.
More interestingly, the FDTS-Pt Nf sil-
icon hybrid plate acts 1) as a high-affin-
ity substrate for phosphopeptides and
2) as a SALDI substrate. The feasibility
of using the FDTS-Pt Nf silicon hybrid
plate for SALDI-MS has been demon-
strated by using a b-casein digest and
various analytes, including small mole-
cules, peptides, phosphopeptides, phos-
pholipids, carbohydrates, and synthetic
polymers. The hybridization of Pt Nfs
with a scratched silicon substrate has
been found to be important for achiev-
ing excellent SALDI activity.
ꢀ
hence, proton transfer from the Si OH
groups to the analyte molecules is en-
hanced, and finally, thermal desorption
of the analyte ions from the surface
occurs. Signal enhancement was ob-
served for protonated molecular ions
trometry
(SALDI-MS).
Surface
scratching of n-type silicon is essential
to induce Pt Nf growth on a silicon
substrate (to obtain a Pt Nf silicon
hybrid plate) by the galvanic displace-
ment reaction. The Pt Nf silicon hybrid
plate showed excellent SALDI activity
in terms of the efficient generation of
protonated molecular ions in the ab-
sence of a citrate buffer. We propose
produced from a titania nanotube
array (TNA) substrate on which Pt
nanoparticles had been photochemical-
ly deposited. Moreover, surface modifi-
cation of the Pt Nf silicon hybrid plate
Keywords: galvanic displacement ·
hybrid materials · mass spectrome-
try · platinum · surface chemistry
ꢀ
that the acidity of the Si OH moieties
Introduction
lyzed.[3,4] However, matrix ion interference and detector sat-
uration are inevitable in the MALDI-MS analysis of low-
mass molecules (m/z<500), and this makes the characteriza-
tion of small molecules difficult despite the significance of
such characterization. Various approaches involving the use
of organic-matrix-free LDI-MS have been investigated for
analyzing small molecules in a MALDI instrument.[5] Siuz-
Matrix-assisted laser desorption/ionization mass spectrome-
try (MALDI-MS) using organic matrices is a soft ionization
technique that causes very little fragmentation of the ana-
lytes.[1,2] By using UV-absorbing organic acids, such as 2,5-
dihydroxybenzoic acid (DHBA) and a-cyano-4-hydroxycin-
namic acid (CHCA) in MALDI-MS, various compounds
such as polymers, peptides, and lipids, as well as complex
mixtures in high-salt matrices and buffers, can be ana-
dak and co-workers developed
a matrix-free LDI-MS
method for generating intact molecular ions by using porous
silicon, which has high UV absorbance and a high surface
[a] Dr. H. Kawasaki, T. Yao, T. Suganuma, K. Okumura, Dr. Y. Iwaki,
Prof. R. Arakawa
[b] Prof. T. Yonezawa, T. Kikuchi
Division of Materials Science and Engineering
Faculty of Engineering, Hokkaido University
Kita 13, Nishi 8, Kita-ku, Sapporo
Hokkaido 060-8628 (Japan)
Department of Chemistry and Materials Engineering
Faculty of Chemistry, Materials and Bioengineering
Kansai University, 3-3-35 Yamate-cho, Suita-shi
Osaka, 564-8680 (Japan)
Fax : (+81)6-6339-4026
E-mail: hkawa@kansai-u.ac.jp
[**] SALDI=surface-assisted laser desorption/ionization
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001038.
arak@kansai-u.ac.jp
10832
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

FULL PAPER
area.[6] This method is known as desorption ionization from
porous silicon (DIOS). In contrast to MALDI, DIOS does
not use organic matrices, and hence, one of its most impor-
tant features is the absence of matrix interferences in the
low-mass region. Therefore, in DIOS, the detectable mass
range of small biomolecules, pharmaceutical compounds,
amino acids, and oligopeptides can be extended to below
m/z 500.[7–9]
nanoparticles, for the analysis of peptides, and we observed
that the SALDI performance was superior when Pt nano-
particles were used.[63] In addition, we developed a SALDI-
MS method by using Pt nanoparticles with thin projections
on the surface (termed Pt Nfs); in this case, the performance
of the SALDI-MS for biomolecules, that is, the sensitivity to
peptides and the efficiency of detecting proteins in the pres-
ence of a citrate buffer,[42] was further improved. However,
these Pt nanomaterials and other nanomaterials have limit-
ed applications in SALDI-MS for the following reasons:
1) the efficiency of generating protonated molecular ions is
low, and 2) the SALDI process, being “harder” than
MALDI, causes fragmentation of labile compounds. Pt
nanomaterials show sufficiently high UV absorbance, but
are inefficient for producing protonated species in the ab-
sence of citrate buffer additives. The use of a citrate buffer
as an effective proton source helps to obtain proton adducts,
but molecular ion peaks from the citrate often appear in the
low-mass region (below m/z 500) in the spectrum; hence,
this organic-matrix-free approach becomes less advanta-
geous for SALDI-MS. Such an inefficiency in the generation
of protonated species has also been reported in other
SALDI-MS methods involving the use of Au,[29,43,46] TiO2,[18]
and CNTs.[64] Another disadvantage of SALDI is that it is
relatively “hard” compared with MALDI. When using Pt
nanomaterials, a very small peak corresponding to the intact
molecular ion generated from 1,2-dimyristoyl-sn-glycero-3-
phosphocholine (DMPC) is observed in the SALDI mass
spectrum; this is because the spectrum includes significant
peaks corresponding to the fragments (m/z 184 and 86)
from the polar head group.[42] In DIOS-MS as well, fragment
ion peaks corresponding to DMPC are predominant in the
spectrum, and the molecular ion peak is indistinct.[65] This
hard ionization tendency has also been reported for SALDI-
MS employing Au nanoparticles[63] or carbon-based materi-
als.[32] Thus, achievement of a soft SALDI process with a
low degree of fragmentation and the efficient generation of
protonated molecular ions without the use of a citrate
buffer additive remain unresolved issues in SALDI-MS.
To increase the efficiency of generating protonated molec-
ular ions and to achieve a “soft” SALDI process, we pro-
pose a new and facile method that employs Pt nanoflowers
(Pt Nfs). We first synthesized 3D Pt Nfs on a silicon sub-
strate by electroless galvanic displacement and investigated
the application of these Pt Nfs in SALDI-MS. Here, we
define Pt nanoflowers as Pt nanoparticles with many nano-
meter-scale surface projections. The plate was simply pre-
pared by immersing an n-type silicon substrate into an aque-
ous solution of HF containing a Pt salt, whereupon direct
growth of Pt Nfs on the silicon surface proceeded spontane-
ously through an electroless galvanic reaction between the
aqueous solution of Pt salt and the silicon substrate. Surface
scratching of the n-type silicon proved to be essential for in-
ducing Pt Nf growth on a silicon wafer (to obtain a Pt Nf sil-
icon hybrid plate) by galvanic displacement. This method
differs from the previously reported electroless galvanic dis-
placement methods, which afforded spherical Pt particles on
More recently, several other types of nanostructured sub-
strates for organic-matrix-free LDI-MS have been reported.
In particular, silicon-based nanomaterials, including nano-
structured silicon films,[10] silicon nanowires,[11,12] silicon
nanocavity arrays,[13,14] silicon microcolumn arrays,[15] and
amorphous silicon,[16] have been extensively studied, and a
new technique known as nanostructure-initiator mass spec-
trometry (NIMS) has recently been proposed.[17] Metal-
oxide-based semiconductors with good UV absorbance are
also promising candidates for matrix-free LDI-MS applica-
tions, such as titania sol–gel films,[18] titania nanotube
arrays,[19] zinc oxide nanowires,[20] mesoporous tungsten tita-
nium oxides,[21] and germanium nanodots.[22] In addition,
double- or multilayer-coated hybrid substrates, such as
metal-coated porous alumina (platinum/alumina),[23,24] two-
layered amorphous silicon,[25] titania-printed aluminum foils
(titania/aluminum),[26] silver-particle-deposited porous sili-
con (silver/silicon),[27] gold nanorods on porous alumina
(gold/alumina),[28] layer-by-layer (LBL) self-assembled films
(polymer/gold),[29,30] DVDs coated with diamond-like
carbon,[31] cationic-polymer-coated graphite sheets (polymer/
graphite),[32] and cobalt-coated silicon substrates (cobalt/sili-
con),[33] are considered as promising materials for matrix-
free LDI-MS because the layer properties of these substan-
ces can be varied independently; further, these hybridization
effects can improve the efficiency of matrix-free LDI-MS.
Another approach for the LDI-MS analysis of small mole-
cules involves the use of nanoparticles as matrices, such as
graphite,[34] carbon nanotubes (CNTs),[35] ZnO,[36] TiO2,[37–39]
Fe2O3,[40] Fe2O3/TiO2,[41] Pt,[42] Au,[43–47] Ag,[48,49] FePtCu,[50]
ZnS,[51] CsSe,[52] and EuF3 nanoparticles. The use of nano-
[53]
particles with a high surface area is extremely advantageous
for sample pretreatment, especially for the separation and
enrichment of target molecules from mixtures in LDI-MS
analysis. In this paper, “surface-assisted laser desorption/ion-
ization” (SALDI) refers to both nanoparticle matrices and
nanostructured substrates, although the meaning of the tech-
nical terms used in this field is somewhat unclear. There is
no consensus on the mechanism underlying SALDI, but
some factors that promote desorption/ionization efficiency
have been reported: 1) laser-induced rapid temperature in-
crease,[54] 2) substrates with a high surface area (porous,
groove-like, nanowires, and nanodots),[54,55] 3) solvent mole-
cules,[14,56] 4) surface functionalities such as terminal OH
groups or a hydrophobic surface,[12,32,46,57–59] 5) electrically
conductive surfaces,[23] and 6) laser-induced surface melting/
restructuring.[60–62]
In our previous studies, we focused on SALDI-MS using
various metal nanoparticles, such as Pt, Au, Ag, and Cu
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10833

H. Kawasaki, R. Arakawa et al.
the silicon surface.[66,67] The Pt Nf silicon hybrid plate
showed excellent SALDI activity in terms of the efficient
generation of protonated molecular ions in the absence of a
citrate buffer. To achieve “soft” SALDI, the Pt Nf silicon
hybrid plate was further modified with perfluorodecyltri-
chlorosilane (FDTS) via the surface OH groups on the sili-
con substrate (to obtain an FDTS-Pt Nf silicon hybrid
plate). More interestingly, we found that, similar to TiO2
substrates,[26,41] the FDTS-Pt Nf silicon hybrid plate acts as
an efficient solid-extract plate and selectively absorbs phos-
phorylated peptides from a complex analytical mixture. The
FDTS-Pt Nf silicon hybrid plate can act as an affinity sub-
strate for phosphopeptides as well as a SALDI substrate.
Results and Discussion
Figure 1. SEM images of Pt nanoflowers deposited on manually scratched
silicon (#1000) by galvanic displacement. Bar: 100 nm. Galvanic deposi-
tion of Pt on the silicon substrates from an aqueous solution of 1 mm
H2PtCl6 containing 5m HF at different immersion times of a) 3 min,
b) 10 min, c) 30 min, d) 60 min. Scale bars=300 nm.
Fabrication of Pt Nfs on scratched silicon surfaces by gal-
vanic displacement and their characterization: Galvanic dep-
osition of Pt on silicon was carried out according to a local
cell mechanism, which involves local cathode deposition of
Pt particles by the injection of holes into the valence band
of n-type silicon and local anode dissolution of silicon by
the injected holes and HF.[66,67] Thus, the reduction of Pt
ions and the chemical etching of silicon occur simultaneous-
ly at the silicon surface in the absence of an externally ap-
plied potential. Previously, galvanic deposition of Pt parti-
cles on a silicon wafer has been reported to afford spherical
Pt particles;[68–70] these studies were focused on controlling
the size and density of the Pt particles. By using a scratched
silicon substrate in the present study, we induced the direct
growth of 3D Pt Nfs with anisotropic shapes on a silicon
wafer (to obtain a Pt Nf silicon hybrid plate) by galvanic dis-
placement. The scratched silicon wafer was prepared by
manual grinding of an n-type silicon wafer with sandpaper
(#1000). Galvanic deposition of Pt on these scratched silicon
substrates was performed by using an aqueous solution of
1 mm H2PtCl6 for various immersion times. After immersion
times in excess of 10 min, SEM images of the Pt Nf silicon
hybrid plate (Figure 1) revealed the formation of non-spher-
ical structures with numerous thin irregular projections on
their surfaces, similar to petals; these projections give rise to
flower-like structures (Pt Nfs) on the scratched silicon plate.
The size of the individual Pt Nfs was in the range 100–
300 nm; these Pt Nfs further assembled to form large 3D ag-
gregates that were a few microns in size. The Pt Nf silicon
hybrid plate was obtained without the use of any capping
molecule/surfactant. When the immersion time was 10 min,
the Pt Nfs did not cover the silicon substrate completely, but
tiny Pt Nfs formed around large Pt Nfs. When the immer-
sion time was increased to 60 min (Figure 1c and d), the tiny
Pt Nfs grew in size and assembled to form large 3D Pt Nfs.
Interestingly, the projections on the surfaces of the Pt Nfs
grew further, as shown in Figure 1d). In contrast, when we
used an n-type silicon wafer without scratching, as in our
previous paper,[68] spherical Pt particles were formed on the
silicon surface. This difference clearly shows that the use of
a scratched silicon wafer is essential for the formation of Pt
Nfs. Although the detailed mechanism of the Pt Nf growth
is not clear at present, it is apparent that surface scratching
of the silicon wafer produces defect sites such as step edges,
pits, and dangling bonds on its surface. This results in the
preferential deposition of Pt nuclei at the defect sites and
the further growth of these nuclei into Pt Nfs. Departure
from local equilibrium at such defect sites may cause mor-
phological instability and subsequent patterns, especially in
the event of rapid solidification, to produce nanoflower
growth.[69,70]
The morphology of the Pt Nfs is expected to depend on
the growth conditions. Thus, we examined the effect of the
following factors on the morphology of the Pt Nfs: 1) sur-
face roughness of the scratched silicon and 2) concentration
of the Pt ions. The surface roughness was varied by mechan-
ical grinding of the silicon wafer with different scratching
plates (#400, #1000, and #4000). The arithmetic mean devia-
tions (Ra) of the profiles of the scratched silicon wafers were
estimated to be 329 nm, 169 nm, and 61 nm, respectively, by
confocal scanning laser microscopy (CSLM). From the SEM
images of the Pt Nfs on scratched silicon surfaces with vari-
ous degrees of roughness, it is apparent that the morphology
of the Pt Nfs is almost independent of the surface roughness.
This, in turn, suggests that surface roughness (i.e., degree of
scratching) is not an important factor affecting the morphol-
ogy of the Pt Nfs (see the Supporting Information, S1).
In contrast, the morphology of the Pt Nfs does depend on
the concentration of H2PtCl6. Figure 2 compares the mor-
phologies of the Pt particles formed on a scratched silicon
wafer (#1000) by galvanic deposition from aqueous H2PtCl6
solutions of different concentrations (0.2 mm and 10 mm);
the deposition time in this case was 10 min. The Pt particle
density clearly increased with the H2PtCl6 concentration.
More importantly, the morphology of the Pt particles
changed from one of nanoflowers (1 mm, Figure 1a; 0.2 mm,
10834
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

Platinum Nanoflowers as SALDI Substrates
FULL PAPER
Figure 2. SEM images of Pt nanoflowers formed on manually scratched
silicon (#1000) by galvanic displacement (immersion time: 10 min). Gal-
vanic deposition of Pt on the silicon substrates from an aqueous solution
containing 5m HF and different concentrations of H2PtCl6: a) and
b) 0.2 mm; c) and d) 10 mm. Scale bars: a) and c) 20 mm; b) and
d) 300nm.
Figure 3. a) SALDI mass spectrum of angiotensin II (5 pmol) obtained
using Pt nanoflowers on manually scratched silicon (#1000) in positive-
ion mode. Galvanic deposition of Pt on the silicon substrates from an
aqueous solution of 1 mm H2PtCl6 and 5m HF (immersion time: 10 min).
b) SALDI mass spectrum of angiotensin II (5 pmol) obtained using the
+
Figure 2b) to disc-like particles (10 mm, Figure 2d) as the
concentration of H2PtCl6 was increased. At 10 mm, parallel
rows of Pt particles were observed in the scratching direc-
tion (Figure 2c), implying scratching-induced Pt deposition
on the silicon surface. Thus, it is likely that the growth of Pt
Nfs on the scratched silicon substrate occurs when the
H2PtCl6 concentration is at least in the range 0.1–1 mm.
Signal enhancement was achieved for protonated molecu-
lar ions in the SALDI mass spectra of peptides when using
the Pt Nf silicon hybrid plate. In the above subsection, we
presented the galvanic deposition of 3D Pt Nfs on scratched
silicon (to obtain a Pt Nf silicon hybrid plate). The galvanic
deposition time was set to 10 min because in our previous
study we found that longer deposition times caused signifi-
cant background noise in the SALDI mass spectra using
spherical Pt particles on unscratched silicon.[70] Additionally,
the dramatic decrease in silicon surface area by the surface
modification is unfavorable for SALDI-MS. In this study,
angiotensin II was chosen as a model peptide, and the
SALDI performance was evaluated when using the Pt Nf
silicon hybrid plate (Figure 1b); angiotensin II is typically
used for preliminary screening when investigating the
SALDI sensitivity for peptides. For comparison, the SALDI
performance when using a scratched silicon plate without Pt
deposition was also examined. As shown in Figure 3a),
when using the Pt Nf silicon hybrid plate and 0.1% tri-
fluoroacetic acid (TFA; proton source), the obtained mass
spectra showed only peaks attributable to the protonated
molecular ion of the peptide. The SALDI mass spectrum
showed low chemical background noise in the low-mass
region (m/z<500). In contrast, the mass spectrum obtained
when using the scratched silicon plate without Pt deposition
showed peaks corresponding to sodium and potassium
adduct ions of the peptides and unknown ions, in addition
~
*
manually scratched silicon in positive-ion mode. =[M+H] , =[M+
+
+
&
Na] , =[M+K] .
to peaks corresponding to the protonated ions, irrespective
of the presence of the TFA proton source (Figure 3b). High
chemical background noise was also observed in the low-
mass region (m/z <300). These results indicate that the hy-
bridization of Pt Nfs with the silicon substrate enhances the
efficiency of protonation of the peptide molecular ions in
SALDI. Our previous study indicated that the deposition of
Pt Nfs from a suitable suspension onto a MALDI plate (i.e.,
stainless steel plate) was inefficient for producing protonat-
ed species in the absence of a citrate buffer additive.[42,63]
Thus, Pt Nfs alone cannot enhance the protonation of pep-
tide molecular ions in SALDI. Combining Pt Nfs with the
silicon substrate is important for generating protonated mo-
lecular ions.
The chemical state of the Pt–silicon interface formed
during the deposition of Pt on the silicon surface may be im-
portant for charge separation of the electron/hole pair be-
tween silicon and Pt. To examine the chemical state of the
Pt–silicon interface, we performed X-ray photoelectron
spectroscopy (XPS) measurements on the Pt Nf silicon
hybrid plate. Figure 4 shows the X-ray photoelectron spectra
of the Pt 4f region for the Pt Nfs silicon plate, which was
prepared by immersing a silicon substrate scratched with
sandpaper (#1000) in an aqueous solution containing 1 mm
H2PtCl6 and 5m HF for deposition times of a) 10 s and
b) 10 min. Pt 4f7/2 and 4f5/2 doublet peaks are seen in the
spectrum. When the Pt deposition time was short (10 s), the
Pt 4f7/2 peak in the obtained spectrum was further split into
two peaks: one at 71.1 eV and the other at 72.8 eV. The
former peak indicates the presence of metallic Pt, whereas
the latter is indicative of platinum silicide (PtSi). These re-
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10835

H. Kawasaki, R. Arakawa et al.
12.7 (deposition time: 10 min), confirming the growth of Pt
on silicon by galvanic deposition.
On the basis of the above-mentioned hypothesis, we state
that Pt nanoparticle deposition helps to enhance the proto-
nation of molecular ions not only when employing silicon
substrates but also when employing other UV-absorbing
semiconductor materials. Similar to silicon, TiO2 also acts as
a semiconductor and is a promising SALDI-MS substrate
because of its high UV absorbance and high chemical stabili-
ty. However, TiO2 nanomaterials have been reported to be
inefficient for producing protonated species for SALDI-MS
in the absence of a citrate buffer additive.[19,37–39] The deposi-
tion of Pt nanoparticles on a TiO2 surface is known to sup-
press electron hole/pair recombination.[71,72] Thus far, studies
on enhanced charge separation on Pt-decorated TiO2 surfa-
ces have focused extensively on photocatalysis. Effective
UV-irradiation-mediated charge separation on Pt-decorated
TiO2 surfaces would help in the production of protonated
species in SALDI-MS, as in the case of using the Pt-decorat-
ed silicon substrates. To confirm this possibility, we exam-
ined the effectiveness of Pt nanoparticle deposition on a
TiO2 nanotube array (TNA). SEM images revealed the
porous morphologies of the TNA and the Pt-decorated
TNA substrate (see the Supporting Information S2). The
SALDI mass spectra of a peptide mixture (Des-ArgI-Brady-
kinin, angiotensin I, and neurotensin) were obtained using
the TNA and Pt-decorated TNA substrate. Peaks due to the
protonated molecular ions of the peptides (5 pmol) could be
detected in the spectrum when using the Pt-decorated TNA
plate along with 0.1% TFA; on the other hand, when using
the TNA plate, peaks due to these ions could not be detect-
ed in the absence of citrate buffer additives (Supporting In-
formation S2). These results confirm that Pt nanoparticle
deposition is effective for enhancing the protonation of mo-
lecular ions in TiO2-based and silicon-based SALDI-MS.
Figure 4. a) and b) XPS Pt 4f spectra obtained for Pt nanoflowers on
manually scratched silicon (#1000). Galvanic deposition of Pt on the sili-
con substrates was from a 1 mm aqueous solution of H2PtCl6 containing
5m HF for immersion times of 10 s (a) and 10 min (b). c) and d) XPS Si
2p spectra measured for Pt nanoflowers on manually scratched silicon
(#1000). Galvanic deposition of Pt on the silicon substrates from a 1 mm
aqueous solution of H2PtCl6 containing 5m HF for immersion times of
10 s (c) and 10 min (d).
sults suggest that the Pt is well adhered to the silicon sub-
strate because of the formation of PtSi. Recently, similar ob-
servations on PtSi formation have been reported for Pt gal-
vanic deposition on unscratched silicon.[69] When the immer-
sion time was short, the size of the deposited Pt nanoparti-
cles was small. Thus, the XPS provides information about
the interface between the deposited Pt metal and the silicon
substrate. When the deposition time was long (10 min), only
peaks due to the Pt metal present near the Pt Nf surface
were observed in the XPS; the Pt 4f7/2 peak appeared at
71.0 eV. The close contact between the n-type silicon sub-
strate and Pt interface owing to PtSi formation may be im-
portant for transferring the UV-irradiation-induced elec-
trons from the silicon to the deposited Pt; this would result
in near-surface hole-trapping on the silicon surface and a
Laser threshold for SALDI-MS when using the Pt Nf silicon
hybrid plate: The minimum laser energy required to gener-
ate stable ion signals (i.e., the laser threshold) is an impor-
tant benchmark for characterizing the effectiveness of
MALDI,[73] SALDI,[23,61] and DIOS.[58,74] In this study, we ex-
amined the laser threshold for angiotensin II and GGYR in
SALDI-MS when using a Pt Nf silicon hybrid plate and a
disc-like Pt Nf silicon hybrid plate (shown in Figure 2d). The
laser threshold was found to be smaller when using the Pt
Nf silicon hybrid plate (20 mJcmꢀ2 for angiotensin II and
25 mJcmꢀ2 for GGYR) than when using the disc-like Pt sili-
con hybrid plate (94 mJcmꢀ2 for angiotensin II and
67 mJcmꢀ2 for GGYR). The smaller laser threshold in the
former case suggests the importance of the Pt Nf morpholo-
gy in enhancing the efficiency of peptide desorption/ioniza-
tion. In a previous study, we suggested that Pt Nfs can effi-
ciently absorb energy during desorption and that the mor-
phological features of the Pt Nfs, such as the roughness and
porosity generated by the formation of petals and the tip
sharpness of the petals on the surface, are very important in
improving the SALDI performance.[42] Among these factors,
ꢀ
consequent increase in the acidity of the Si OH groups.
The presence of SiOx was confirmed by XPS. The spectra
of the Si 2p region for the Pt Nf silicon hybrid plate were re-
corded for deposition times of 30 s and 10 min, as shown in
Figure 4c) and d), respectively. In the Si 2p region, two main
peaks are visible in each spectrum: one attributable to sili-
con at 99.7 eV and the other attributable to SiOx at
102.8 eV, from SiOx that was probably formed by the oxida-
tion of silicon, as the Pt Nf silicon hybrid plate was exposed
to air for a short time before being placed in the XPS cham-
ber. Similar oxidation of the Pt Nf silicon hybrid plate may
have occurred prior to the SALDI-MS experiments. The Pt/
Si XPS ratio increased from 0.3 (deposition time: 30 s) to
10836
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

Platinum Nanoflowers as SALDI Substrates
FULL PAPER
we consider the tip sharpness of the petals on the surface to
be effective for increasing the laser-induced temperature;
this is because of the low thermal conductivity and thermal
conferment of the Pt Nfs. As a result, a low laser fluence is
sufficient to ensure that the Pt Nfs attain the same tempera-
ture as the disc-like Pt Nfs. A similar low laser threshold has
been reported for SALDI-MS employing silicon nanowires
grown on silicon surfaces.[74]
As mentioned above, we prepared Pt Nf silicon hybrid
plates with the desired surface roughness by mechanical
grinding of the silicon wafer (SEM images shown in the
Supporting Information S1). For all of the Pt Nf silicon
hybrid plates, the mass spectra showed peaks due to the pro-
tonated molecular ion of angiotensin II when 0.1% TFA
was used. A small laser threshold was required for Pt Nf sili-
con hybrid plates with higher surface roughness (see the
Supporting Information S3). This is consistent with previous
reports, according to which a small laser threshold is re-
quired to produce ions on a silicon substrate with high po-
rosity or a large total surface area.[14] However, in contrast
to the mass spectra obtained when using the silicon hybrid
plates prepared by manual grinding (Figure 3a), those ob-
tained when using the Pt Nf silicon hybrid plates prepared
by mechanical grinding showed a number of unknown sig-
nals in the low-mass region. Since the spectra obtained for
the mechanically ground hybrid plate included chemical
background noise, we used the Pt Nf silicon hybrid plate
produced by manual grinding of the silicon wafer with sand-
paper (#1000) for the following SALDI-MS studies.
Figure 5. a) SALDI mass spectrum of BP-Cl using Pt nanoflowers (Pt
Nfs) on manually scratched silicon (#1000). b) Comparison of molecular
ion survival yields of BP-Cl in SALDI-MS using Pt Nfs, FDTS-modified
Pt Nfs, FDTS-Pt Nfs (with citrate buffer), DIOS, and MALDI with
CHCA matrix.
Survival yield measurements of thermometer ions on the Pt
Nf silicon hybrid plate: In the SALDI process, excess
energy may be transferred to individual molecules, and this
may lead to undesired fragmentation, as opposed to the soft
desorption in MALDI. To examine the degree of internal
energy transfer in the desorption process, we measured the
survival yield (SY) of a model substance, benzylpyridine
(BP), which affords a series of benzylpyridinium ions, the
so-called “thermometer ions” (TM). Details of the SY
method using TM ions have been previously reported in the
literature.[57,59,61,64,76] We performed SY measurements on
BP-Cl by SALDI-MS using the Pt Nf silicon hybrid plate.
To further improve the SALDI efficiency in terms of sensi-
tivity[58] and soft desorption,[61] we modified the Pt Nf silicon
hybrid plate by derivatizing the surface OH groups of the
substrate with FDTS. The reaction time for FDTS silylation
was set at 5 h to avoid morphological changes in the Pt Nfs.
When the reaction time was increased to 24 h, a dramatic
morphological change in the Pt Nfs was observed (Support-
ing Information S4). The SALDI mass spectrum of BP-Cl
exhibits two primary peaks: a molecular ion M+ peak (m/z
204) and the corresponding fragment F+ peak (benzyl
cation, m/z 125), as shown in Figure 5a). The SY value is de-
termined according to Equation (1):
in which IM and IF are the experimentally measured intensi-
ties of the M+ and F+ peaks, respectively. The internal
energy is inversely proportional to the SY. The change in
the SY of BP-Cl with laser fluence was also examined. For
comparison, SY measurements were also carried out for
DIOS with CHCA in MALDI.
As shown in Figure 5b), in the case of the Pt Nf silicon
hybrid plate, the threshold for ion production was reached
at a low laser fluence (14 mJcmꢀ2), which is comparable
with that in MALDI (13 mJcmꢀ2). Although the SY values
of the Pt Nf silicon hybrid plate (ꢁ0.65) are lower than
those observed in the MALDI process (SYꢁ1), the FDTS-
modified Pt Nf silicon hybrid plate (FDTS-Pt Nf silicon) has
a major impact on the increase in the SY values (ꢁ0.92),
that is, the decrease in the degree of fragmentation. The ad-
dition of citrate buffer further increased the SY (ꢁ0.97) of
the FDTS-Pt Nf silicon plate. This suggests that surface
modification by fluorinated chains helps to decrease the in-
ternal energy during analyte desorption from the surface.
One possible reason for this is as follows: the weak interac-
tion of the surface-fluorinated chains with BP-Cl may lead
to a decrease in the energy barrier for analyte desorption
IM
ðIM þ IFÞ
SY ¼
ð1Þ
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10837

H. Kawasaki, R. Arakawa et al.
from the surface (i.e., an increase in the SY value). Another
possible reason for the high SY of the FDTS-Pt Nf silicon
hybrid plate is non-thermal desorption. The laser-induced
explosive evaporation of the surface-fluorinated chains may
help to desorb the BP-Cl and ensures that the critical
energy for surface desorption is lower than that for simple
thermal desorption. Note that the SY values (i.e., fragmen-
tation behavior) change significantly with the laser fluence
in the case of the FDTS-Pt Nf silicon hybrid plate. This im-
plies the possibility of secondary molecular collisions in the
gas phase, during which energy exchange occurs.[14] Co-de-
sorption of the fluorinated chains and BP-Cl ions at a high
laser fluence allows molecular collisions in the gas phase be-
tween these species; hence, the SY values decrease at a high
laser fluence.[14] In contrast, for the Pt Nf silicon hybrid
plate, the SY values remain practically unchanged in the ex-
amined fluence range because there is no co-desorption of
the surface-modified molecules and BP-Cl ions.
Surface reconstruction/destruction or melting of the
SALDI substrate upon laser irradiation can contribute to
non-thermal desorption and thus trigger the desorption pro-
cess.[62–64] Since the melting temperature of nanosized metals
is lower than that of bulk metals, deformation or ablation of
Pt nanostructures may occur during irradiation by a pulsed
laser. Therefore, we observed the structure of Pt Nfs after
laser irradiation at 50 mJcmꢀ2, which is sufficient for detect-
ing peptides by SALDI-MS. The SEM images demonstrated
that there was very little change in the morphology of the Pt
Nfs after laser irradiation (see the Supporting Information
S5). Therefore, surface reconstruction/destruction of Pt Nfs
upon laser irradiation may contribute to a small extent to
non-thermal desorption for triggering the SALDI process.
This suggests that in the case of
hybrid plate upon exposure to ambient air was mainly indi-
cated by the appearance of unknown peaks in the mass
spectra at m/z <100 and peaks due to sodium and potassi-
um ions, which were probably formed by air contamination.
Applicability of FDTS-Pt Nf silicon hybrid plate to various
analytes
Phospholipids: When using the FDTS-Pt Nf silicon hybrid
plate, a high SY was observed for molecular ions at a low
laser fluence. For this reason, we consider that the FDTS-Pt
Nf silicon hybrid plate should have potential application in
the SALDI-MS analysis of labile compounds. We selected a
phospholipid as a model labile compound. In the detection
of phospholipids by a SALDI process, such as DIOS, peaks
due to the fragment ions obtained from the polar head
group are predominantly observed in the mass spectrum,
while the molecular ion peak is indistinct. However, such
fragment ion peaks are not predominant in the spectrum
when phospholipids are analyzed by MALDI-MS using a
typical chemical matrix such as DHBA. This suggests that
the ionization conditions in SALDI-MS are harsher than
those in MALDI-MS. In this study, we found that the
FDTS-Pt Nf silicon hybrid plate showed a relatively lower
fragmentation tendency. Figure 6 shows the mass spectra of
DMPC obtained when using the Pt Nf silicon hybrid plate
and the FDTS-Pt Nf silicon hybrid plate, in addition to the
MALDI spectra obtained with DHBA. The spectrum ob-
tained using the Pt Nf silicon hybrid plate shows peaks cor-
responding to the lipid molecular ions [M+Na]+and
[M+H]+, but peaks due to the fragments (m/z 184 and 86)
from the polar head group are still predominant (Figure 6a).
the Pt Nf silicon hybrid plate
and the FDTS-Pt Nf silicon
hybrid plate, SALDI activity
can be observed at
a laser
energy lower than the laser
threshold for surface recon-
struction/destruction or melting.
Durability of FDTS-Pt Nf sili-
con hybrid plates: The SALDI
activity of silicon has been re-
ported to decay after a few
days of exposure to ambient
air.[16] However, we did not ob-
serve any such degradation for
the Pt Nf silicon hybrid plate
even when it was exposed to
air. Good quality signals were
obtained from angiotensin II
when using
a one-month-old
plate, and the laser threshold
was 23 mJcmꢀ2 (see the Sup-
porting Information S6). Degra-
dation of the Pt Nf silicon
Figure 6. SALDI mass spectra of DMPC (60 pmol) in the presence of 0.5 mm NaTFA using a) Pt nanoflowers
(Pt Nfs) on manually scratched silicon (#1000), b) FDTS-modified Pt Nfs, and c) MALDI with DHBA matrix.
+
+
~
&
*
=[M+H] , =[M+Na] , =fragment ion.
10838
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

Platinum Nanoflowers as SALDI Substrates
FULL PAPER
In contrast, the mass spectrum obtained using the FDTS-Pt
Nf silicon hybrid plate clearly shows peaks due to the lipid
molecular ion [M+Na]+; from the spectrum, it is apparent
that the degree of fragmentation becomes low (Figure 6b),
comparable to that observed in the MALDI process with
DHBA (Figure 6c).
sor ions of PEG 6000 have also been detected in ZnO-
SALDI, in which the degree of fragmentation is low.[36] In
this study, we performed a SALDI-MS analysis of PEG
6000 by using the FDTS-Pt Nf silicon hybrid plate. The
SALDI mass spectrum of PEG 6000 showed peaks due to
the precursor ions with a molecular weight distribution
around m/z 6000 with little fragmentation (Figure 7b), simi-
lar to the MALDI mass spectrum obtained with CHCA
(Figure 7a). The FDTS-Pt Nf silicon hybrid plate was also
effective for the analysis of hydrophobic polymers, such as
poly(methyl methacrylate) (PMMA), as shown in Fig-
ure 7d); this was because high-molecular weight PMMA
ions could be detected at a laser power lower than that re-
quired for MALDI (Figure 7c). Thus, the performance ach-
ieved with the FDTS-Pt Nf silicon hybrid plate was compa-
rable to that achieved in MALDI in terms of the molecular
weight distribution of the polymer ions and limited fragmen-
tation.
Synthetic polymers: Polyethylene glycol (PEG) is used in a
number of pharmaceutical, industrial, and food additive ap-
plications, for example laxatives, skin creams, personal lubri-
cants, toothpastes, and thickening agents; binding agents for
ceramic molds, casting, and powder metallurgy; oligomers
for polyurethane manufacture; and non-ionic detergents.
Recently, we reported that fragment ions with m/z 500–3000
are predominantly generated in the TiO2-SALDI of PEG
6000; however, no precursor ions (m/zꢁ6000) are formed in
this case.[39] Further, in SALDI-MS using Au nanoparticles,
PEG fragment ions with molecular weights of more than
3000 are predominantly formed, and precursor ions are also
observed.[63] It has been reported that EuF3 microdiscs with
hollow interiors can be successfully used for the detection of
PEG precursor ions with molecular weights up to 30000, al-
though the fragmentation pattern is still not clear.[53] Precur-
Peptides, carbohydrates, and small molecules: The use of the
FDTS-Pt Nf silicon hybrid plate helps to improve the detec-
tion sensitivity for peptides, carbohydrates, and small mole-
cules (see the Supporting Information S7). When using the
Figure 7. Mass spectra of PEG 6000 (1 mgmLꢀ1) in the presence of 0.5 mm NaTFA using a) MALDI with CHCA matrix and b) FDTS-modified Pt Nfs.
LDI mass spectra of PMMA (1 mgmLꢀ1) in the presence of 0.5 mm NaTFA using c) MALDI with CHCA matrix and d) FDTS-Pt Nfs.
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10839

H. Kawasaki, R. Arakawa et al.
FDTS-Pt Nf silicon hybrid plate, the minimum amount of
angiotensin II and GGYR samples required for obtaining
spectra with reasonable signal-to-noise ratios was about
5 fmol in the absence of citrate buffer. The detection limits
for b-cyclodextrin and caffeine (500 fmol and 10 pmol, re-
spectively) are relatively lower than those for angiotensin II
and GGYR. For amine-containing drugs, such as verpamil
hydrochloride, the detection limit is 5 fmol. However, the
detection of small acidic drug molecules, such as deprotonat-
ed aspirin and vitamin C proved impossible by SALDI-MS
when using the FDTS-Pt Nf silicon hybrid plate, even at
high concentrations (100 pmol) of the analytes (see the Sup-
porting Information S8). The detection of deprotonated ions
in negative-ion mode is very difficult when using the FDTS-
Pt Nf silicon hybrid plate, probably because of the surface
chemical properties of the plate. The silicon oxide surface is
peptides used, only one showed a weak peak (m/z 1369.7) in
the mass spectrum, which corresponded to the protonated
molecular ion (Figure 8a). This indicated that the signals
due to the phosphopeptides were completely suppressed in
the SALDI mass spectrum. Peptides with high molecular
weights, such as renin substrates, enolase T35, enolase T37,
and melittin were not detected. When the FDTS-Pt Nf sili-
con hybrid plate was subjected to washing procedures, peaks
due to all of the phosphopeptides could be seen in the
SALDI mass spectrum (Figure 8b). This result indicates that
the FDTS-Pt Nf silicon hybrid plate has some affinity for
phosphopeptides, and hence, this plate can be used for the
enrichment of phosphopeptides from mixtures. However,
the reason for this affinity is not clear at present. The inter-
action of phospholipid (Lewis base) with Pt Nfs (Lewis
acid) might contribute to the on-plate phosphopeptide en-
richment.
To further confirm the phosphopeptide-trapping capacity
of the FDTS-Pt Nf silicon hybrid plate, we used the tryptic
digest product of b-casein as a sample; this is because b-
casein is a commonly used phosphoprotein. Peaks corre-
sponding to phosphopeptides from the tryptic digest product
of b-casein were detected at m/z 2062, 2556, and 3122 only
when using the washed PFD-Pt Nf silicon hybrid plate (Fig-
ure 8c, d). The phosphopeptides identified are summarized
in Table S2 in the Supporting Information. The results indi-
cated that the PFD-modified Pt Nf silicon hybrid plate
acted as an affinity substrate for phosphopeptides and as a
SALDI substrate, which is similar to the affinity work on
TiO2 substrates. By optimizing the extraction and washing
steps, the procedure will be further improved.
intrinsically acidic because of the chemical properties of the
[77]
ꢀ
terminal Si OH groups, as in the case of DIOS chips. The
ꢀ
acidity of the Si OH moieties may be further increased be-
cause of the electron-withdrawing nature of the surface fluo-
rocarbon chains and Pt particles on the silicon plate. There-
fore, the analytes deposited on the surface are in an acidic
environment; this results in differences in the degree of de-
protonation of acidic compounds. On the other hand, in-
creased surface basicity may facilitate the formation of neg-
ative ions from small acidic molecules. In fact, we reported
in a previous paper that a polyethyleneimine-modified
graphite substrate is suitable for the effective detection of
the deprotonated ions of perfluorooctanoic acid in negative-
ion mode.[32]
On-plate phosphopeptide enrichment using FDTS-Pt Nf sili-
con hybrid plates: The selective detection of phosphopepti-
des from proteolytic digests is of very high relevance in
many proteomics applications. In proteolytic digests, phos-
phopeptides are often present in small amounts. The coexis-
tence of non-phosphopeptides decreases the peak ion inten-
sity of phosphopeptides in the LDI mass spectra because of
the efficient LDI of the non-phosphopeptides (the so-called
ion-suppression effect). Therefore, selective isolation and/or
enrichment of phosphopeptides from a mixture is essential
before identification by LDI-MS. Thus, to date, TiO2 materi-
als have been employed for the enrichment of phosphopep-
tides from complex mixtures because of the specific affinity
of TiO2 for phosphopeptides.[26,41] Here, we found that the
FDTS-Pt Nf silicon hybrid plate could be used for the en-
richment of phosphopeptides by direct SALDI-MS analysis.
The following peptide mixture containing ten kinds of non-
phosphopeptides (allanton, angiotensin fragment 1–7,
RASG-1, angiotensin II, bradykinin, angiotensin I, renin
substrate, enolase T35, enolase T37, and melittin) and four
kinds of phosphopeptides (T181-1P, T191-1P, T43-1P, and
T43-2P) was used in this study. The molecular weights of
these peptides are summarized in Table S1 in the Supporting
Information. Figure 8a) presents the SALDI mass spectrum
of the peptide mixture containing phosphopeptides before
the washing treatment. Among the four spiking phospho-
Conclusion
We have reported for the first time the fabrication of 3D Pt
nanoflowers (Pt Nfs) on a scratched silicon substrate by
electroless galvanic displacement and the application of
these Pt Nfs in SALDI-MS. Surface scratching is essential to
induce Pt Nf growth on the silicon substrate (to obtain a Pt
Nf silicon hybrid plate) by galvanic displacement. Surface
scratching of the silicon wafer produces defect sites, such as
step edges, pits, and dangling bonds on the silicon surface,
and hence preferential deposition of the Pt nuclei occurs at
the defect sites. These Pt nuclei can then be grown into Pt
Nfs. When there is supersaturation at such defect sites, mor-
phological instability can occur, and nanoflower-like growth
forms may appear.
Using this Pt Nf silicon hybrid plate, we achieved excel-
lent SALDI activity in terms of the efficient generation of
protonated molecular ions in the absence of a citrate buffer.
The acidity of the Si OH moieties on the silicon increases
because of the electron-withdrawing nature of the Pt Nfs.
ꢀ
ꢀ
As a result, proton transfer from the Si OH groups to the
analyte molecules is enhanced, and there is thermal desorp-
tion of the analyte ions from the surface. Signal enhance-
ment of protonated molecular ions has also been observed
10840
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

Platinum Nanoflowers as SALDI Substrates
FULL PAPER
Pt Nf silicon hybrid plate acts
as an affinity substrate for
phosphopeptides as well as a
SALDI substrate.
The Pt Nf silicon hybrid plate
may be readily extended to
other transition metals, such as
palladium and rhodium, and ef-
forts towards achieving this are
in progress in our laboratory.
The study is of significance in
shape-controlled synthesis of
metal nanoparticles, and is of
importance not only with
regard to SALDI applications,
but
also
surface-enhanced
Raman scattering (SERS), elec-
trocatalysis, sensors, and fuel
cells.[76–78]
Experimental Section
Preparation of Pt Nfs on scratched sil-
icon by galvanic displacement: A sili-
con wafer was cut into 1.5 cmꢁ1.5 cm
squares and sonicated in acetone and
methanol for 10 min each. Scratched
silicon wafers were produced by two
methods: i) manual grinding of the sili-
con wafer by using sandpaper (#1000)
and ii) mechanical grinding of the sili-
con wafer with
a grinding machine
Figure 8. SALDI mass spectra of the peptide mixture containing phosphopeptides a) before and b) after wash-
(Marumoto Struers S5629) and differ-
ent wet SiC papers of #400, #1000, and
#4000 (Marumoto Struers) for 60 s.
During mechanical grinding, the rota-
tion speed of the papers was adjusted
to 50 rpm. The scratched silicon wafer
was sonicated in acetone and metha-
nol for 10 min and then etched with
*
ing treatment. A peptide mixture containing ten kinds of non-phosphopeptides ( ; allanton, angiotensin frag-
ments 1–7, RASG-1, angiotensin II, bradykinin, angiotensin I, renin substrate, enolase T35, enolase T37, and
melittin) and four kinds of phosphopeptides ( ; T181-1P, T191-1P, T43-1P, and T43-2P) was used. SALDI
mass spectra of the tryptic digest product of b-casein c) before and d) after washing treatment.
~
46% HF for 20 min. Typically, electroless galvanic deposition of Pt Nfs
on the silicon surface was carried out by immersing the scratched silicon
substrate in a 1 mm aqueous solution of H2PtCl6 containing 5m HF for
10 min. After immersion, the silicon substrate was washed with water
and preserved in EtOH until further use in the SALDI-MS experiments.
for a TiO2 nanotube array substrate upon photochemical
deposition of Pt nanoparticles.
Surface modification of the Pt Nf silicon hybrid plate by
FDTS (an FDTS-Pt Nf silicon hybrid plate is obtained as a
result) helps to improve the detection sensitivity for pep-
tides, carbohydrates, and small molecules. Another advant-
age of using the FDTS-Pt Nf silicon hybrid plate is the soft
SALDI process, which has been demonstrated by the surviv-
al yields of a thermometer ion (BP-Cl). As a result, by using
the FDTS-Pt Nf silicon hybrid plate, labile compounds, such
as phospholipids and PEG 6000, can be detected with a very
low degree of fragmentation in SALDI-MS. The feasibility
of using FDTS-Pt Nf silicon hybrid plates (or Pt Nf silicon
hybrid plates) for matrix-free SALDI-MS analysis has been
demonstrated by using a b-casein digest and various ana-
lytes, including small molecules, peptides, phosphopeptides,
phospholipids, carbohydrates, and synthetic polymers. More
importantly, we have found for the first time that the FDTS-
Caution! HF is a hazardous acid that can cause serious tissue damage if
burns are not appropriately treated. Thus, HF should be handled in a
well-ventilated fume hood while wearing appropriate chemical safety
items: a face shield and double-layered nitrile gloves.
Surface silylation of Pt Nf silicon hybrid plate: The Pt Nf silicon hybrid
plate was modified with FDTS or 3-aminopropyltrimethoxysilane
(APTMS) via the surface OH groups on the substrate. The surface silyla-
tion reaction was induced by immersing the Pt Nf silicon hybrid plate in
hexane containing 1 mm perfluorodecyltrichlorosilane (PFDS) for 5 h at
258C. Subsequently, the substrate was washed with excess hexane and
then with EtOH.
SALDI-MS: The substrates prepared for SALDI-MS were fixed to a
stainless steel sample plate using double-sided conductive carbon tape.
The sample solution containing 0.5 mL of a cationization agent (0.1%
TFA or 1 mm NaI) was spotted onto this substrate and dried under re-
duced pressure. SALDI mass spectra were obtained in linear mode by
using an AXIMA CFR TOF mass spectrometer (Shimadzu, Kyoto,
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10841

H. Kawasaki, R. Arakawa et al.
Japan) fitted with a pulsed nitrogen laser (337 nm). One hundred laser
shots were used to acquire the mass spectra. The analyte ions were accel-
erated at 20 kV under delayed extraction conditions.
[19] C. Y. Loa, J. Y. Lina, W. Y. Chena, C. T. Chena, Y. C. Chen, J. Am.
Soc. Mass Spectrom. 2008, 19, 1014.
[20] M. J. Kang, J. C. Pyun, Y. J. Choi, J. H. Park, J. G. Park, J. G. Lee,
H. J. Choi, Rapid Commun. Mass Spectrom. 2005, 19, 3166.
[21] M. Yuan, Z. Shan, B. Tian, B. Tu, P. Yang, D. Zhao, Microporous
Mesoporous Mater. 2005, 78, 37.
[22] T. Seino, H. Sato, A. Yamamoto, A. Nemoto, M. Torimura, H. Tao,
Anal. Chem. 2007, 79, 4827.
[23] S. Okuno, R. Arakawa, K. Okamoto, Y. Matsui, S. Seki, T. Kozawa,
S. Tagawa, Y. Wada, Anal. Chem. 2005, 77, 5364.
[24] R. Nayak, D. R. Knapp, Anal. Chem. 2007, 79, 4950.
[25] V. Jokinen, S. Aura, L. Luosujꢂrvi, L. Sainiemi, T. Kotiaho, S. Frans-
sila, M. Baumann, J. Am. Soc. Mass Spectrom. 2009, 20, 1723.
[26] H. Bi, L. Qiao, J. M. Busnel, V. Devaud, B. Liu, H. H. Girault, Anal.
Chem. 2009, 81, 1177.
Survival yield (SY) measurements of thermometer ion: Benzylpyridinium
chloride (BP-Cl) was synthesized by the reaction of pyridine (anhydrous,
purity >99.8%, Sigma Aldrich) and the corresponding substituted benzyl
chloride (95–99% pure, Sigma Aldrich).[64] Benzyl chloride was mixed
with anhydrous pyridine (3 mL; pyridine/benzyl chloride molar ratio:
20:1), and this mixture was heated in a water bath for 5 h at 608C.
Excess pyridine was removed by vacuum evaporation. The identities of
the compounds with BP-Cl were confirmed by SALDI-MS analysis and
used without further purification. A stock solution of BP-Cl (0.166 mm)
was prepared in methanol.
[27] H. Yan, N. Xu, W. Y. Huang, H. M. Han, S. J. Xiao, Int. J. Mass
Spectrom. 2009, 281, 1.
[28] L. C. Chen, J. Yonehama, T. Ueda, H. Hori, K. Hiraoka, J. Mass
Spectrom. 2007, 42, 346.
[29] a) H. Kawasaki, T. Sugitani, T. Watanabe, T. Yonezawa, H. Moriwa-
ki, R. Arakawa, Anal. Chem. 2008, 80, 7524; b) A. Tarui, H. Kawa-
saki, T. Taiko, T. Watanabe, T. Yonezawa, R. Arakawa, J. Nanosci.
Nanotechnol. 2009, 9, 159.
[30] M. Hong, F. Qiu, L. L. Huang, J. Zhu, J. Phys. Chem. C 2008, 112,
11078.
[31] M. Najam-ul-Haq, M. Rainer, C. W. Huc, P. Hausberger, H. Kraus-
haar, G. K. Bonn, Anal. Chem. 2008, 80, 7467.
[32] H. Kawasaki, N. Takahashi, H. Fujimori, K. Okumura, T. Watanabe,
C. Matsumura, S. Takemine, T. Nakano, R. Arakawa, Rapid
Commun. Mass Spectrom. 2009, 23, 3323.
Acknowledgements
This study was partially supported by a Grant-in-Aid for Scientific Re-
search (B) (Nos. 22350040 (to R.A.), 21310072 (to T.Y.)) from the Japan
Society for the Promotion of Science (JSPS) and the Ministry of Educa-
tion, Culture, Sports, Science and Technology (MEXT), Japan. T.Y. is
also grateful for partial financial support from Grant-in-Aid for Scientific
Research in Priority Area, “Molecular Science for Supra Functional Sys-
tems (477)” (No. 22018001) from MEXT. This research was partly sup-
ported by the Core-to-Core Program promoted by the Japan Society for
the Promotion of Science (Project No. 18004). This study was supported
by the “Strategic Project to Support the Formation of Research Bases at
Private Universities”: Matching Fund Subsidy from MEXT.
[33] T. Yalcin, L. Li, Appl. Surface Sci. 2009, 256, 1309.
[34] J. Sunner, E. Dratz, Y. C. Chen, Anal. Chem. 1995, 67, 4335.
[35] M. V. Ugarov, T. Egan, D. V. Khabashesku, J. A. Schultz, H. Peng,
V. N. Khabashesku, H. Furutani, K. S. Prather, H. W. J. Wang, S. N.
Jackson, A. S. Woods, Anal. Chem. 2004, 76, 6734.
[36] T. Watanabe, H. Kawasaki, T. Yonezawa, R. Arakawa, J. Mass Spec-
trom. 2008, 43, 1063.
[37] C. T. Chen, Y. C. Chen, Anal. Chem. 2004, 76, 1453.
[38] K. H. Lee, C. K. Chiang, Z. H. Lin, H. T. Chang, Rapid Commun.
Mass Spectrom. 2007, 21, 2023.
[39] T. Watanabe, K. Okumura, H. Kawasaki, R. Arakawa, J. Mass Spec-
trom. 2009, 44, 1443.
[40] W. Y. Chen, Y. C. Chen, Anal. Bioanal. Chem. 2006, 386, 699.
[41] C. T. Chen, Y. C. Chen, Anal. Chem. 2005, 77, 5912.
[42] H. Kawasaki, T. Yonezawa, T. Watanabe, R. Arakawa, J. Phys.
Chem. C 2007, 111, 16278.
[43] J. A. McLean, A. A. Stumpo, D. H. Russell, J. Am. Chem. Soc. 2005,
127, 5304.
[44] Y. F. Huang, H. T. Chang, Anal. Chem. 2006, 78, 1485.
[45] H. P. Wu, C. L. Su, H. C. Chang, W. L. Tseng, Anal. Chem. 2007, 79,
6215.
[46] E. T. Castellana, D. H. Russell, Nano Lett. 2007, 7, 3023.
[47] K. Shibamoto, K. Sakata, K. Nagoshi, T. Korenaga, J. Phys. Chem.
C 2009, 113, 17774.
[48] S. D. Sherrod, A. J. Diaz, W. K. Russel, P. S. Cremer, D. H. Russell,
Anal. Chem. 2008, 80, 6796.
[1] K. Tanaka, H. Waki, Y. Ido, S. Akita, Y. Yoshida, T. Yoshida, Rapid
Commun. Mass Spectrom. 1988, 2, 151.
[2] M. Karas, D. Bachmann, F. Hillenkamp, Anal. Chem. 1985, 57, 2935.
[3] a) M. W. F. Nielen, Mass Spectrom. Rev. 1999, 18, 309; b) S. D.
Hanton, Chem. Rev. 2001, 101, 527; c) K. Deisewerd, Chem. Rev.
2003, 103, 395; d) M. Karas, R. Kruger, Chem. Rev. 2003, 103, 427;
e) R. Knochenmuss, R. Zenobi, Chem. Rev. 2003, 103, 441.
´
[4] F. Hillenkamp, J. Peter-Katalinic, MALDI-MS: A Practical Guide to
Instrumentation, Methods and Applications, Small-Molecule Desorp-
tion/Ionization Mass Analysis, Wiley-VCH, Weinheim, Chapter 9,
2007.
[5] D. S. Peterson, Mass Spectrom. Rev. 2007, 26, 19.
[6] J. Wei, J. M. Buriak, G. Siuzdak, Nature 1999, 399, 243.
[7] R. A. Kruse, S. S. Rubakhin, E. V. Romanova, P. W. Bohn, J. V.
Sweedler, J. Mass Spectrom. 2001, 36, 1317.
[8] Z. Shen, J. J. Thomas, C. Averbuj, K. M. Broo, M. Engelhard, J. E.
Crowell, M. G. Finn, G. Siuzdak, Anal. Chem. 2001, 73, 612.
[9] E. P. Go, J. E. Prenni, J. Wei, A. Jones, S. C. Hall, H. E. Witkowska,
Z. Shen, G. Siuzdak, Anal. Chem. 2003, 75, 2504.
[10] J. D. Cuiffi, D. J. Hayes, S. J. Fonash, K. N. Brown, A. D. Jones,
Anal. Chem. 2001, 73, 1292.
[11] E. P. Go, J. V. Apon, G. Luo, A. Saghatelian, R. H. Daniels, V. Sahi,
R. Dubrow, B. F. Cravatt, A. Vertes, G. Siuzdak, Anal. Chem. 2005,
77, 1641.
[12] G. Piret, H. Drobecq, Y. Coffinier, O. Melnyk, R. Boukherroub,
Langmuir 2010, 26, 1354.
[49] T. C. Chiu, L. C. Chang, C. K. Chiang, H. T. Chang, J. Am. Soc.
Mass Spectrom. 2009, 20, 2008.
[50] H. Kawasaki, T. Akira, T. Watanabe, K. Nozaki, T. Yonezawa, R.
Arakawa, Anal. Bioanal. Chem. 2009, 395, 1423.
[13] N. H. Finkel, B. G. Pervo, O. D. Velev, L. He, Anal. Chem. 2005, 77,
1088.
[14] Y. Xiao, S. T. Retterer, D. K. Thomas, J.-Y. Tao, L. He, J. Phys.
Chem. C 2009, 113, 3076.
[51] S. K. Kailasa, K. Kiran, H. F. Wu, Anal. Chem. 2008, 80, 9681.
[52] K. Shrivas, S. K. Kailasa, H. F. Wu, Proteomics 2009, 9, 2656.
[53] Z. Chen, Z. Geng, D. Shao, Y. Mei, Z. Wang, Anal. Chem. 2009, 81,
7625.
[54] S. Alimpiev, S. Nikiforv, V. Karavanskii, T. Minton, J. Sunner, J.
Chem. Phys. 2001, 115, 1891.
[15] Y. Chen, A. Vertes, Anal. Chem. 2006, 78, 5835.
[16] S. Alimpiev, S. Nikiforv, V. Karavanskii, T. Minton, J. Sunner, J.
Chem. Phys. 2008, 128, 014711.
[17] T. R. Northen, O. Yanes, M. T. Northen, D. Marrinucci, W. Urit-
boonthai, J. Apon, S. L. Golledge, A. Nordestrom, G. Siuzdak,
Nature 2007, 449, 1033.
[55] R. A. Kruse, X. Li, P. W. Bohn, J. V. Sweedler, Anal. Chem. 2002,
74, 3639.
[18] C. T. Chen, Y. C. Chen, Rapid Commun. Mass Spectrom. 2004, 18,
1956.
10842
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10832 – 10843

Platinum Nanoflowers as SALDI Substrates
FULL PAPER
[56] Y. Chen, H. Chen, A. Aleksandrov, T. M. Orlando, J. Phys. Chem. C
2008, 112, 6953.
[67] M. Cerruti, G. Doerk, G. Hernandez, C. Carraro, R. Maboudian,
Langmuir 2010, 26, 432.
[57] S. A. Trauger, E. P. Go, Z. Shen, J. V. Apon, B. J. Compton, E. S. P.
Bouvier, M. G. Finn, G. Siuzdak, Anal. Chem. 2004, 76, 4484.
[58] G. Luo, Y. Chen, G. Siuzdak, A. Vertes, J. Phys. Chem. B 2005, 109,
24450.
[68] T. Yao, H. Kawasaki, T. Watanabe, R. Arakawa, Int. J. Mass Spec-
trom. 2010, 291, 145.
[69] R. F. Sekerka, Cryst. Res. Technol. 2005, 40, 291.
[70] K.-S. Choi, Dalton Trans. 2008, 40, 5389.
[59] S. Dagan, Y. Hua, D. J. Boday, A. Somogyi, R. J. Wysocki, V. H.
Wyoski, Int. J. Mass. Spectrom. 2009, 283, 200.
[71] M. Sadeghi, W. Liu, T. G. Zhang, P. Stavropoulos, B. Levy, J. Phys.
Chem. 1996, 100, 19466.
[60] T. R. Northen, H. K. Woo, M. T. Northen, A. Nordstrçm, W. Urit-
boonthai, K. L. Turner, G. Siuzdak, J. Am. Soc. Mass Spectrom.
2007, 18, 1945.
[61] Y. Wada, T. Yanagishita, H. Masuda, Anal. Chem. 2007, 79, 9122.
[62] H. W. Tang, K. M. Ng, W. Lu, C. M. Che, Anal. Chem. 2009, 81,
4720.
[72] S. Sakthivel, M. V. Shankar, M. Palanichamy, B. Arabindoo, D. W.
Bahnemann, V. Murugesan, Water Res. 2004, 38, 3001.
[73] G. Luo, I. Marginean, A. Vertes, Anal. Chem. 2002, 74, 6185.
[74] G. Luo, Y. Chen, H. Daniels, R. Dubrow, A. Vertes, J. Phys. Chem.
B 2006, 110, 13381.
[75] L. P. Law, Int. J. Mass Spectrom. 2010, 290, 72.
[76] N. Tian, Z. Y. Zhou, Z. G. Sun, L. Cui, B. Ren, Z. Q. Tian, Chem.
Commun. 2006, 4090.
[63] T. Yonezawa, H. Kawasaki, A. Tarui, T. Watanabe, R. Arakawa, T.
Shimada, F. Mafune, Anal. Sci. 2009, 25, 339.
[64] W. Y. Chen, L. S. Wang, H. T. Chiu, Y. C. Chen, J. Am. Soc. Mass
Spectrom. 2004, 15, 1629.
[77] M. Subhramannia, V. K. Pillai, J. Mater. Chem. 2008, 18, 5858.
[78] J. N. Tiwari, F. M. Pan, K. L. Linb, New J. Chem. 2009, 33, 1482.
[65] Q. Liu, Y. Xiao, C. Pagan-Miranda, Y. M. Chiu, L. He, J. Am. Soc.
Mass Spectrom. 2009, 20, 80.
[66] I. Lombardi, S. Marchinna, G. Zangaru, S. Pizzini, Langmuir 2007,
23, 12413.
Received: April 21, 2010
Published online: August 2, 2010
Chem. Eur. J. 2010, 16, 10832 – 10843
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10843