D. Kowal et al.
Materials Research Bulletin 142 (2021) 111434
was confirmed that some shift in the band position strongly depended on
the ionic liquid structure and dye concentration [43].
highlight the difference between this value and the power emitted from
the sealed container. The excess of power measured at the “container”
zone is associated with liquid presence inside it. Although we could not
see any other explanation for the observed fluorescence spectra, it can
only be considered as indirect evidence. A natural way to examine the
interior of the reservoirs is to cut through it with focused ion beam (FIB).
However, the ion beam, due to its high energy, is not neutral to the
structure and may cause solidification of its liquid part. Thus our efforts
to identify liquid and solid part in the cross-section of the reservoir were
not conclusive.
2
ED equal to 40 mC/cm was chosen to fabricate the sidewalls of the
micro reservoirs. This value was chosen in order to provide enough
spatial resolution of the fabricated structures and in the same time to
limit the fluorescence intensity of the solid part of the structure. More-
over, the dose of 40 mC/cm2 was well above the minimum dose for
solidification of the [Allmim][Cl] liquid. In this case the thickness of the
solidified structures is practically independent on the ED variations and
it reflects the thickness of the liquid film formed by spin coating [12].
That is because the applied accelerating voltage of 30 kV was more than
enough for electrons to penetrate through the entire film and induce its
solidification from the surface to the bottom. On the other hand, this was
not the case when 5 kV accelerating voltage was applied. Measured
thickness of the “covers” fabricated by 5 kV accelerating voltage is
shown in Fig. 3g as a function of the ED used. It was preferable to
fabricate the cover part of the containers with possibly small EDs as to
keep them possibly thin. However, as we found out, the cover layers
In Fig. 6a-b the fluorescence microscope image of the disc–like
microcontainers is presented, together with quantitative data of the
fluorescent maximum intensities. The discs were sealed with cover
layers fabricated with various EDs. Fluorescent intensity increases with
2
the ED and reaches maximum for ED = 6 mC/cm . For greater EDs the
intensity decreases. This data imply that the amount of liquid inside the
fabricated containers depends on the beam parameters used for the
fabrication of the cover layers. Using small EDs for the fabrication of the
covers results in thinner layer and this maximizes the volume of the
containers. On the other hand, as it was already mentioned, thin cover
layers often collapse, somehow allowing the liquid to be washed away
during the development in alcohol. It seems the thinnest layers are in
fact leaky which point to porous character of the solidified material,
however we found no evidence of porosity during SEM examination.
Superimposing these two features, volume decrease and tendency to
2
fabricated with small EDs (1-2 mC/cm ) have the tendency to collapse
inside the cells during sample development in isopropyl alcohol. To
demonstrate this several container structures are shown in Fig. 4a.
Square and circular containers fabricated with ED of the cover layer
2
equal to 1-2 mC/cm look very much drained and collapsed. For com-
parison it seems that the containers fabricated with ED of the cover layer
2
9
-12 mC/cm are filled with RTIL to greater extent. This difference is
2
most noticeable in case of circular containers. One should note that even
the containers fabricated with ED of the cover layer equal to 12 mC/cm2
show some kind of recess. Then it seems that some amount of the RTIL
closed in the container during patterning is always lost in the develop-
ment process. Furthermore in Fig. 4b similar containers are shown
fabricated in the two studied RTILs. For the same choice of ED used for
substance solidification, physical nature of the resulting solids is
different. E-beam modified regions in [Allmim][Cl] appear much
smoother than the ones in [TPEBuImAll][Cl] which exhibit more surface
roughness. Despite this fact we managed to successfully manufacture the
micro reservoirs in both RTILs.
collapse, makes the ED = 6 mC/cm the most preferable for making
containers.
Temperature, likewise the ED, is a factor the might induce the so-
lidification of the RTIL and influence its fluorescent characteristics. As
the liquid interior of the micro reservoirs is not altered by the electron
beam it can potentially be used for monitoring the heating history of the
sample. Also, the RTILs can be applied in the future as photo-
polymerisable resists with pre- and post bake processing steps - thermal
treatment is commonly used in lithography in order to remove solvents
or increase the stability of the resist. Thus we believe it is important to
know the impact of elevated temperature on the fluorescent character-
istics of the RTIL. For this purpose some additional tests have been
conducted (Fig. 6c). Samples were annealed for 10 min in temperatures
As was mentioned in Section 2.2, to confirm the presence of the
liquid phase RTIL inside the reservoir some special patterning design has
been introduced and spectral measurements have been performed. The
designs comprised of three zones as was schematically shown in Fig. 2b:
the “cover”, the “sidewall” and the “container” (“container” zone was in
fact an oversection of the first two zones). The emission spectra have
been collected for the three zones separately. The assumption was that
the “container” spectrum not only constituted of the power emitted by
the “cover” and “sidewall” but also another factor associated with the
liquid sealed in the container cells. In Fig. 5 the measured spectra are
shown for the two examined RTILs. Additionally, summarized power of
the “sidewall” and “cover” emitted light is presented on the plots as to
◦
between 50 – 200 C subsequently. The intensity of fluorescence
increased after first annealing process which might have been connected
to drying of the substance at elevated temperature. With temperature
◦
increasing above 100 C the fluorescence intensity gradually decreased
and it fell below 25% of the initial value once the temperature reached
◦
◦
200 C. It is not surprising that the fall of intensity started after exceeding
100 C taking into account the baking of the RTILs prior to film depo-
sition. The observed change was irreversible. One might have expected
that the liquid phase RTIL inside the containers would partially harden
at elevated temperature causing an additional drop of the fluorescence
intensity measured at the covered sections of the sample. However, no
such effect was observed as the fluorescence intensity of micro reservoirs
remained distinctly higher than for sidewalls and a cover summed up,
◦
even after annealing up to 200 C, as shown in Fig. 7 (compare with
Fig. 5). This suggests that the temperature required for hardening of the
RTIL has not been reached and, consequently, the thermal degradation
threshold of the dyes is the main reason for the fluorescence drop.
◦
◦
Decomposition of pure fluorescein takes place at 290 C (554 F), as can
be found in online database [44]. Other literature data also show fluo-
◦
rescein to be only little affected by temperatures below 210 C [45].
Decomposition of TPEOH which is responsible for fluorescence of
◦
[
TPEBuIm][Cl] is expected at 215 C [46]. On the other hand the thermal
behavior of fluorophores in RTIL environment might be quite different.
The decrease of fluorescence could be also associated with two other
effects, namely increased concentration of fluorophore due to material
shrinking and change of material transparency. It is unlikely though that
these effects play significant role, as in the conducted tests we found no
Fig. 4. a) Microcontainers fabricated in fluorescein-containing [Allmim][Cl]
with square and round shapes, covered with layers solidified with various EDs
2
(
1, 2, 9, 12 mC/cm ) as indicated by numbers. b) Microcontainers fabricated in
2
◦
two fluorescent RTILs with the cover layer solidified with ED set to 12 mC/cm .
confirmation of significant material shrinking in 200 C and we observed
5