Full text of DOI:10.1070/MC2002v012n01ABEH001522

Mendeleev Commun., 2002, 12(1), 9–11
Thermal-lens spectrometry for studying molecular layers covalently bonded to a
flat glass surface
Mikhail Yu. Kononets, Svetlana N. Bendrysheva, Mikhail A. Proskurnin,* Elena V. Proskurnina,
Evgenii M. Min’kovskii, Aleksei R. Tarasov and Georgii V. Lisichkin
Department of Chemistry, M. V. Lomonosov Moscow State University, 119992 Moscow, Russian Federation.
Fax: +7 095 939 4675; e-mail: michael@analyt.chem.msu.ru
10.1070/MC2002v012n01ABEH001522
The molecular layers of the Reactivrot B5A dye covalently bonded to flat quartz glass surfaces was studied by thermal lensing in
combination with electron-probe microanalysis.
Table 1 Parameters of the dual-beam thermal lens spectrometer.
Studies of the structure of chemically bonded molecular layers
are of importance in the chemistry of surface compounds.1 A
small contribution from the layer mass to the total sample
mass does not allow methods traditionally used for solving the
problems of three-dimensional chemistry to be applied. There-
fore, the composition and structure of bonded molecular layers
on porous substrates with high specific surface areas are usually
studied. In the case of developed surfaces, the mass fraction of
a bonded layer can reach 10%, which is sufficient for elemental
microanalysis, IR, UV, and EPR spectroscopy etc.2 However,
in the case of a chemical modification of the surface of flat
substrates, the study of the bonded layer becomes a compli-
cated problem. We applied thermal lens spectrometry (TLS) to
study chemically modified flat surfaces of quartz glasses.
Thermal lensing is a thermooptical method commonly used
in analytical practice.3 This method is based on detecting heat
effects due to the non-radiative relaxation of molecules that ab-
sorbed electromagnetic radiation.3,4 In thermal lensing, an equi-
librium redistribution of temperature in the irradiated sample
results in a Gaussian profile of refractive index determined by
the spatially Gaussian form of the excitation laser beam. This
profile can be treated as an optical element similar to a lens
with the focal distance being a function of the absorbance of the
test sample.3
Thermal lensing has such features as (i) high instrumental
sensitivity, which makes it possible to detect absorbances in
liquid samples down to 10–7–10–8, to determine concentrations
down to 10–12 mol dm–3, to analyse volumes down to 10–15 dm3
and to detect hundreds of molecules in such volumes; (ii) the
possibility to use all the variety of methods and approaches
of conventional spectrophotometry; and (iii) non-destructivity,
which makes it possible to apply TLS to a wide range of test
substances.3–5 Usually, thermal lensing is applied to determine
substances like dyes or metal chelates absorbing the radiation
of an excitation laser in liquids.3–5
Excitation laser
Wavelength le (nm)
514.5
300
19.5
180
Focussing lens focal length (nm)
Confocal distance (mm)
Laser power at cell (mW)
Spot size at the waist (µm)
60 1
Probe laser
Wavelength lp (nm)
632.8
185
3.1
Focussing lens focal length (nm)
Confocal distance (mm)
Laser power at cell (mW)
Spot size at the waist (µm)
4
25 1
Other parameters
Cell length (mm)
2
Sample-to-detector distance (cm)
m in equation (2)
V in equation (2)
200
2.0 0.1
3.1 0.1
cathode. Due to the very low conductivity of the samples,
measurements were made at low acceleration voltages at the
equilibrium point (2 kV or lower). At this potential, the total
secondary current is equal to the primary electron current. For
each sample, the equilibrium points were selected to eliminate
charging effects on the images. At this potential, photographs
were made every 2 min during 20 min (the images did not
change). Focussing was performed at low magnifications (×25)
to prevent charging. Secondary-electron images were obtained
with a magnification of 200 to 1500.
The thermal lens signal qj for a single j-th measurement of a
sample was calculated as the average of i = 50–100 signals of
on-off thermal-lens excitation cycles resulting in a series of the
relative change in intensity in the central part of the probe beam
for a steady-state thermal lens [qj1, qj2, … qji]. The values of qji
were calculated from the equation:3,4
Ioff – Ion
Ion
qji =
,
(1)
where Ioff is the time-averaged probe beam intensity without
a thermal lens (excitation beam is off) and Ion is the time-
averaged probe beam intensity for a steady-state thermal lens
(excitation beam is on) for a single cycle. Recalculation of the
signal q to absorbance A was made using the equation:3,5
2
However, the sensitivity of the method is sufficient to meas-
ure the concentration of a dye on a nonporous glass surface.
Shimosaka et al.6 studied the Acridine Orange dye adsorbed on
glass using total-internal-reflection thermal lensing. The aim of
this study was to examine the uniformity and surface concen-
tration of a bonded layer.
J
2
2mV
1 + 2mV + V2
qj = 1 – arctan
– 1
(2)
A dual-laser parallel-beam (collinear) thermal lens spectrometer
was used.7 The thermal lens was excited with an INNOVA 90-6
argon-ion laser (Coherent, USA) with le = 514.5 nm (TEM00
mode); the maximum power at the sample was 180 mW. The
probe was a SP-106-1 helium–neon laser (Spectra Physics, USA)
with lp = 632.8 nm (TEM00 mode). The signal (the intensity at
the centre of the probe beam) was measured with an FD-7K
photodiode. Next, the signal was amplified, and it entered an
ADC–DAC board of an IBM PC/AT computer. The measure-
ment was synchronised using the special software.7 The optical
parameters of the spectrometer are summarised in Table 1.8 The
beam waist location and waist spot sizes of laser beams were
measured using an 818-SL digital power meter (Newport Corp.,
Fountain Valley, CA, USA) according to Snook and Lowe.5
Electron-microscopic images were taken on a CamScan 44
scanning electron microscope with a thermal-emission tungsten
where J = 2.303E0PeA, Pe is the power of the excitation laser at
the sample, m is the mode-mismatching factor and V is the dis-
tance from the probe waist to the sample normalised to the cor-
responding confocal length.5 The parameter E0 = (dn/dT)(lpk)–1
is the reduced enhancement factor for thermal-lens measurements,
dn/dT is the temperature gradient of the refractive index, lp is
the wavelength of the probe laser, and k is the thermal con-
ductivity. In thermal-lens measurements, the following parameters
were measured: an averaged signal of the sample for j sub-
sequent cycles qj and the measurement precision (deviation of
the signal from cycle to cycle). For each test sample with the
concentration of the test component c, the average signal qn(c)
was calculated for n replicate measurements. The repeatability
of measurements was characterised by the relative standard
deviation.
– 9 –

Mendeleev Commun., 2002, 12(1), 9–11
Si(OEt)2(CH2)3NH2
SiOH + (EtO)3Si(CH2)3NH2
SiO
SiO
– EtOH
Si(OEt)(CH2)3NH2
O
+ H2O
SiOH + (EtO)3Si(CH2)3NH2
Scheme 1 Reactivrot B5A.
– EtOH
[Si(OEt)(CH2)3NH2]n
The value of J is proportional to absorbance A = al, where
a (cm–1) is the absorptivity of the layer of the test compound
and l (cm) is the sample depth. The absorptivity can be esti-
mated as a = escs, where es (cm mol–1) is the surface absorption
coefficient and cs (mol cm–2) is the surface concentration of the
absorbing layer. Taking into account the sufficient precision of
measurements of beam radii at the sample (see Table 1), this
makes it possible to calculate the surface concentration of the dye
from the thermal-lens signal. In our experiments (le = 514.5 nm),
es = 9×106 cm mol–1, E0 = 6.3 W–1, and Pe = 0.180 W.
The selection of the dye was dictated by its absorption-band
parameters, by the absence of radiation losses under light absorp-
tion, and by the possibility of its chemical immobilization on
the surface. We selected the Reactivrot B5A dye (Scheme 1).
Modification of glasses with a sublayer of (3-aminopropyl)-
trimethoxysilane was made from a toluene solution in the pres-
ence of trace moisture according to Lisichkin et al.9 The dye
was bonded to a (3-aminopropyl)trimethoxysilane sublayer under
the conditions providing an incomplete surface coverage to reveal
the possibility of detection of ‘holes’ in the bonded layer. A dye
solution in dry methanol was used, and the reaction was carried
out in a sealed ampoule at 70 °C for 16 h. Next, the ampoule
was opened, and the sample was washed with dry methanol.
The conditions of bonding the dye to aminated quartz glasses
were selected from preliminary experiments with a porous silica
substrate, Silochrom S-120. The dye was bonded by amino
groups of (3-aminopropyl)trimethoxysilane as a modifier, which
react with a SiO2 surface (see Scheme 2).9
Thermal-lens studies of modified samples with the bonded
dye showed that certain areas of a single glass plate are char-
acterised by two significantly different signals. One of them
is the response of the modified surfaces without the dye (a
‘background’ line in Table 2), which is mainly determined by
background photothermal defocusing and light scattering. Another
signal corresponds to the bonded dye and shows a curve with a
plateau, and the signal of the plateau also corresponds to the
thermal-lens signal of the modified surface before dye bonding
(Figure 1). Subsequent measurements at the same sample position
give the same signal.
The data obtained evidenced that the selected modification
conditions result in ‘holes’ or sites without the dye. This is con-
firmed by the measurements of the same samples using scanning
electron microscopy (see Figure 2). Thus, thermal lensing can
be used for estimating the uniformity of the bonded layer in the
scale corresponding the linear dimensions of the thermal lens
(about 60 µm in our case).
In electron-probe experiments, the stability of so called ‘chemi-
cally labile’ substances, i.e., substances with low thermo- and
electroconductivity, which may decompose under vacuum and
electron-probe effects, is a very complicated problem that requires
Scheme 2 The bonding of (3-aminopropyl)trimethoxysilane to a glass sur-
face under dry conditions and in the presence of water traces.
special studies. Previously, we developed several approaches to
the analysis of chemically labile samples, including non-con-
ducting organic samples.10,11
Especially important is the time dependence of the thermal-
lens signal for areas with a bonded dye. As the temperature
gradient of the refraction index for the used glass is positive, the
signal J increases with the concentration of the test substance
on the surface (photothermal focussing). Thus, the term in paren-
theses in equation (2) and the signal q increase with a decrease
in the absorption of the bonded layer. Therefore, a reproducible
increase in the signal for glass surfaces with a bonded dye is
likely due to the decomposition of the dye under laser irradiation.
This was confirmed by comparing electron-probe images of the
plates before and after thermal lens measurements: the images
show new ‘holes’ in dye covers with a round symmetry and a
rough size of 60 µm. Under the same conditions, Reactivrot B5A
in an aqueous solution is resistant to laser radiation. Probably,
the rigid fixation of the dye on a solid substrate results in a
decrease in its photostability. However, the residence time of
the dye at the surface was much longer than the characteristic
time of the thermal lens effect (~100 ms). Thus, it provided
reliable thermal-lens determination of the dye by photodestruc-
tion kinetic curves (Figure 1).
The photodestruction curves showed that the process is com-
plex. We suppose that TLS would provide data on the thermal
stability of bonded molecular layers.
The surface concentrations of bonded dyes were found exper-
imentally (Table 2). It is noteworthy that these data are not
averaged through the whole surface and relevant only to the
surface areas with the dye. Previously, it was shown12 that the
bonding of trifunctional silanes on porous substrates may result
in the formation of polylayers. It is also known1,2 that the maxi-
mum bonding of monofunctional silanes to silica is limited by the
area of the anchor group of a modifier, and it is 2.5 group nm–2.
Taking into account steric hindrances, the limiting density of a
Reactivrot B5A monolayer is about 1.1 group nm–2 (calculated
using the PM3 method, WinMOPAC, Chem3D, CambridgeSoft).
These calculations are in good agreement with the experimental
data for Sample 1 (very close to the limiting density of a mono-
0.035
0.030
0.025
0.020
0.015
0.010
Table 2 The bonding density of the dye to samples. Sample 1: (3-amino-
propyl)trimethoxysilane was applied directly without a solvent. Sample 2:
(3-aminopropyl)trimethoxysilane was applied from dry toluene. Sample 3:
(3-aminopropyl)trimethoxysilane was applied from wet toluene. The bonding
density of (3-aminopropyl)trimethoxysilane to Silochrom S-120 is about
1.6 group nm–2. Blank sample is a glass treated with (3-aminopropyl)-
trimethoxysilane only (n = 15).
0.005
0.000
Bonding density of the dye
Average
signal qn(c)
Number
of layers
0
50
100
150
200
250
300
Sample
mol cm–2
group nm–2
t/s
Background
Sample 1
Sample 2
Sample 3
Silochrom S-120
0.025 0.002
—
—
—
Figure 1 Thermal-lens measurements. Photodestruction of the Reactivrot
B5A dye bonded to a glass surface. The curve corresponds to the photo-
destruction of the bonded dye, dashed line is the average signal of the sur-
face modified with (3-aminopropyl)trimethoxysilane, solid line is the average
signal of the initial unmodified glass surface. le = 514.5 nm, Pe = 180 mW.
0.022 0.002 (1.6 1.0)×10–10 1.0 0.6
0.018 0.003 (3.2 1.4)×10–10 1.9 0.8
–0.01 0.01 (1.6 0.5)×109 9 3
0.87
1.74
8.7
—
8.3×10–11
0.5
0.45
– 10 –

Mendeleev Commun., 2002, 12(1), 9–11
References
1
G. V. Lisichkin and A. Yu. Fadeev, Zh. Ross. Khim. O-va im. D.I.
Mendeleeva, 1996, 40, 65 (in Russian).
2
A. Yu. Fadeev and G. V. Lisichkin, in Adsorption on New and Modified
Inorganic Sorbents. Series ‘Studies in Surface Science and Catalysis’,
Elsevier, Amsterdam, 1996, vol. 99, p. 191.
3
S. E. Bialkowski, Photothermal Spectroscopy Methods for Chemical
Analysis, Wiley, New York, 1996, and references therein.
J. Georges, Talanta, 1999, 48, 501.
R. D. Snook and R. D. Lowe, Analyst, 1995, 120, 2051.
T. Shimosaka, T. Sugii, T. Hobo, J. B. Alexander Ross and K. Uchiyama,
Anal. Chem., 2000, 72, 3532.
4
5
6
7
8
9
M. A. Proskurnin, A. G. Abroskin and D. Yu. Radushkevich, Zh. Anal.
Khim., 1999, 54, 101 [J. Anal. Chem. (Engl. Transl.), 1999, 54, 91].
M. A. Proskurnin and V. V. Kuznetsova, Anal. Chim. Acta, 2000, 418,
101.
Modifitsirovannye kremnezemy v sorbtsii, katalize i khromatografii
(Modified Silicas in Adsorption, Catalysis and Chromatography), ed.
G. V. Lisichkin, Khimiya, Moscow, 1986 (in Russian).
Figure 2 Secondary-electron image of a glass surface with the bonded
Reactivrot B5A dye (acceleration voltage of 2 kV, magnification of 1500.
Darker regions correspond to the bonded dye).
layer of the dye). If the bonding density exceeds this value, this
unambiguously shows the formation of a polylayer, which is
the case of Sample 3 in Table 2. However, if the experimental
bonding density is lower than the theoretical value, it hardly
gives an unambiguous answer. Nevertheless, the homogeneity of
a monolayer can be estimated indirectly, e.g., from the repeat-
ability of TLS measurements at different sites of a sample.
The data (Table 2) suggest that the bonding density for the
dye on a glass surface unambiguously correlates with the amount
of water at the stage of surface modification with a (3-amino-
propyl)trimethoxysilane sublayer. High amounts of water result
in a developed structure of the modifier sublayer (see Scheme 2),
which is in good agreement with the data of Fadeev and
Lisichkin.2 This leads to high concentrations of the dye bonded
to the surface (Sample 3). The oligomerization of this trifunc-
tional modifier was completely suppressed only in a pure modifier
without a solvent (Sample 1).
10 F. A. Guimelfarb, M. N. Filippov and E. V. Kletskina, Fresenius’ Z.
Anal. Chem., 1997, 357, 796.
11 E. V. Kletskina, M. N. Filippov, A. G. Borzenko and F. A. Guimelfarb,
Microscopy and Analysis, 1997, 49, 5.
12 A. Yu. Fadeev and T. J. McCarthy, Langmuir, 2000, 16, 7268.
Thus, thermal lensing combined with scanning electron micro-
scopy is a promising approach to study chemically modified sur-
faces of non-conducting substrates.
This study was supported by the Russian Foundation for
Basic Research (grant no. 01-03-33149a).
Received: 25th September 2001; Com. 01/1848
– 11 –