Sample holder and microscope optics

In Figure 3.3 the sample holder, sample cell and light source are shown in detail.

The construction of the interior is very subtle as small changes might inuence the sample behavior drastically via illumination, thermal gradients and atmosphere inside the sample chamber. The main dierence compared to all former 2D colloid setups is that the sample is now observed from below to enable access of the optical tweezers from top. This change requires various adjustments of the sample holder and the illumination: for a clear observation the bottom of the chamber has to be transparent.

At the same time, the chambers atmosphere has to be saturated with water vapor to minimize the evaporation from the sample cell. That, however, causes fogging on ordinary transparent windows based on silicon dioxide (SiO₂) or conventional polymers like Polymethylmetacrylat (PMMA). The solution of this problem is presented in the following where the sample holders geometry is explained.

Figure 3.3A gives schematic insight into the sample holders geometry. The holder is made from massive copper to provide a sucient heat sink being unsusceptible to quick temperature variations. For thermal contact of the glass sample cell heat con-ductive paste¹⁹ is used. Additionally, this seals the chamber against evaporation of water. From below the chamber is closed by a composite window that consists of a

19Arctic Cooling, Product: Arctic Silver 5.

conventional cover slip glued with an anti-fogging sheet²⁰ by transparent UV glue²¹. It is sealed to the copper block with epoxy glue. This window provides three necessary properties: (i) anti-fogging behavior, (ii) clear transparency for observation, and (iii) impermeability for water vapor. It was found that the impermeability for water vapor is a crucial point. A strong evaporation was always correlated with strong particle drift (up to0.5mm/day), at least in a binary mixture²². A data acquisition without particle drift was only possible using the composite window. Furthermore, the lifetime of the sample is limited by water volume of the syringe which is depleted after approximately half a year for a high evaporation rate.

Additionally to the composite window a water basin in a side pocket of the sample holder is necessary to saturate the atmosphere. It is lled by the water basin actuator 2mm before the water front inside the channel enters the chamber. Over the experi-mental duration (several months) water slowly condensates at the tip of the water front in this channel. The experimentalist has to take care with the actuator that the water drop is not entering the sample chamber for this severely inuences the observation optics. Very slow adjustment is needed, as a sudden change in the order of millimeters at the water front causes a change in pressure. This easily deforms the interface by several ten micrometers which cannot be compensated by the control mechanisms.

The gold platelet²³reects light back into the sample cell. With the light source No. I (below sample) this platelet is necessary for sucient illumination. Further, the light is reected by the inner walls of the copper block which makes this particle illumination sensitive to water condensation.

The glass sample cell²⁴ is shown in Figure 3.3B. The center bore contains the colloidal suspension which is held by surface tension at the sharp edges. To enhance wetting contrast, the at area outside the cylindrical bores are treated with silane²⁵ making this area water repellant. The bores are treated afterwards with 20% RBS solution²⁶ to ensure they are hydrophilic. The small bore is accessed by the nozzle of a teon hose (see Figure 3.3A) to control the amount of water in both bores. The curvature of the monolayer is thereby controlled directly with the water supply actuator.

Figure 3.3C shows the microscope optics. A gray scale 8-bit CCD camera²⁷ is

con-20PINLOCK, www.pinlock.nl (December 9, 2008), a motorcycle helmet shield of≈1mmthickness.

The working principle was undisclosed by the manufacturer. It is assumed that the sheet consists of a bulk polymer that swells when in contact with water. Condensation of small water droplets (fog) at the surface is prevented as they are soaked by the material. This corresponds to the observation that the sheet alone is permeable for water vapor. A decrease in water evaporation from the sample by an order of magnitude has been observed when the sheet alone was exchanged with the composite construction. This was measured by comparing the feed rates of the water supply control.

21Norland Products, Norland optical adhesive 61 Lot226.

22The binary mixture appears to be much more susceptible to this perturbation than a one-component sample. See also section 3.3.5.

23100 nm gold layer vaporized on a conventional coverslip.

24Helma, purpose-built glass cell.

25Amersham, PlusOne Repel-Silane ES.

26Roth, RBS35 Konzentrat.

27Allied Vision Technology, Firewire camera Marlin 145B.

anti-fogging sheet UV glue

glass cover slip epoxy sealing

copper block

glass sample cell heat conductive paste

gold vaporized cover slip

water reservoir water supply

A

C

particle observation

B

removable IR filter

CCD camera microscope tube 4x microscope objective surrounded by light guides

LED block

6 mm

Figure 3.3: A: Schematic drawing of the glass cell holder with composite bottom window to illuminate and observe the sample. B: Picture of the glass sample cell, upside down.

The rim of the center bore is where the water-air interface is attached. The bore is connected by a channel to a side bore that serves as a reservoir for the water supply nozzle. C: The microscope optics mount with wave guides is directly located below cell holder (A). The camera with microscope tube and objective is surrounded by bre optics that guide light from 24 LEDs (λ = 624nm, located in heat sink copper block) to the sample (light source No. I). The sample is illuminated with diusive light. Here, sample holder and magnetic eld coils are dismounted.

nected via a microscope tube²⁸ to a microscope objective²⁹. The sample holder and coils are dismounted. The microscope objective has a working distance of 18.5mm and is located directly underneath the sample chamber. In this arrangement there is enough space to turn an IR lter³⁰ between objective and composite window to block the laser tweezers beam (which is focused on the CCD chip of the camera as the laser focus is in the observed particle plane). The light source No. I consists of 24 LEDs³¹ placed in a copper block as heat sink. Light guides³² lead towards the sample with the ends located around the objective and point directly at the colloidal monolayer from below. The ends have to be prepared with a smooth cut³³to ensure sucient coupling of light into the bers and a uniform emerging illumination cone. To avoid thermal disturbance, the diodes are located far away from the colloidal suspension.

A microscope image of the particles is imported with a repetition rate of 10Hz via rewire connection to a computer for further processing. A typical image obtained with this optics is shown in Figure 3.4 with approximately 3000 particles³⁴. The whole sample contains about up to10⁵ particles. The eld of view has a size of1158×865µm² with a resolution of 1392×1040P ixel. The diusive illumination (light source No. I) provides a clear contrast, and particle species can easily be distinguished. In Figure 3.5 a small cutout is displayed from the raw image of Figure 3.4. There, the particle in the center is less bright than the other big particles. Particles with such an appearance are identied as aggregates of two big particles aligned parallel to the external magnetic eld. The image processing routine can hardly distinguish these particles from small particles by their size (depending on processing parameters). However, as there are typically just a few aggregates present in a suciently stabilized sample (in Figure 3.4 there are exactly ve which is≈0.2%of the particles in the eld of view) this imperfec-tion does not signicant reduce the quality of the data. The reason for this appearance of aggregates might be that a sphere attached on top of another is absorbing light. A considerable fraction of the diusive intensity is coming from top due to the reect-ing gold platelet. Both light sources No. I and No. II have dierent advantages and disadvantages: Light source No. II provides a more homogeneous illumination over the whole sample than light source No. I illuminating from below. The advantage of LED source No. I is that data can be recorded right at the edge of the sample cell (see Figure 3.14, section 3.3.5) where other illumination techniques fail, like classical Köhler illumination [40, 41, 69] or illumination with light source No. II. There, light is strongly scattered at the edge of the cell. This is of interest for investigations of the monolayer close to a hard wall.

28Stemmer Imaging, c-mount microscope tube with magnication1×.

29Olympus, UIS2 series PLN4X, 0.10 numerical aperture,18.5mmworking distance.

30Edmund Optics, TechSpec Shortpass Filter - 850NM 25mm Dia.

31LEDsλ= 624nmwith narrow beam divergenceθ= 6^◦ and brightness10000mCd.

32Goodfellow, PMMA ber wave guides, ber diameter1mm.

33Hirtz Bühler, diamond wafering blade, series 15HC Diamond No.11-4244.

34In a one-component sample, densities with up to 10000 particles in the eld of view can be prepared using this optics. In the binary case, the distinction between particle species becomes dicult for more than≈4000particles (see section 3.2.2).

Figure 3.4: Raw image of 1613 big and 1159 small particles (ξ = 42%) in the eld of view of 1158 × 865 µm² in diusive illumination (light source No. I, 8-bit gray scale CCD camera, 4× microscope objective). The particles were conned at a at and horizontal water-air interface. An external magnetic eld of B = 4.70mT is applied corresponding to an interaction strength of Γ = 660. Big and small particles can be clearly distinguished by their apparent size.

Figure 3.5: A cutout (125×125µm²) of the image displayed in Figure 3.4 is shown.

The particle in the center is apparently darker than other big particles which is typical for aggregates of two big particles sitting on top of each other.