Effects of Surface Energy on Rubrene Crystallization

By: Christopher Phenicie

External Advisor: Russell Holmes (CEMS Department)

Internal Advisor: Elias Puchner

Date Submitted: May 11, 2015

Abstract

The correlation between the activation energy of crystallizing rubrene and the energy of the surface it was grown on was determined. Rubrene was grown on silicon substrates modified with different molecules in order to change its surface energy. This rubrene was then thermally annealed on the modified surfaces. It was determined that there is no correlation between the surface energy and activation energy of crystal growth. This knowledge is useful to further predict and eventually control the growth kinematics of crystalline organic semiconductors.

1 Introduction

It was postulated in [1] that the activation energy of rubrene crystallization depends on the surface it is grown on. Namely, they postulated that the activation energy would be larger on surfaces with large surface energy, and the activation energy would be smaller on surface with small surface energy. In order to test this, rubrene was grown on four different surfaces with different surface energy. The activation energy of crystal growth was then determined on each surface.

2 Theory

2.1 Crystallization

Crystallization of rubrene can be thought of as a reaction that transforms amorphous rubrene into crystalline rubrene. Therefore, the Arrhenius equation [2] (equation 1) can be used to describe the rate of crystallization of rubrene (dr/dt) at a given temperature (T) and activation energy (Ea) with just knowledge of Boltzmann's constant (k) and initial conditions (A).

(1)

The rate of crystallization can be determined from the time derivative of the radius of the crystal. Since the crystals grow in circles, this measurement is well defined, as shown in figure 1.

Figure 1: The growth rate of the crystal is used to determine the activation energy from the Arrhenius law. The growth rate is determined from the change in the size of the crystal with respect to time.

2.2 Surface Energy

The energy of a surface is related to the contact angle of water on that surface, as shown in figure 2. A surface of higher energy will have a lower contact angle as described in [3]. Therefore, we used the contact angle as the indicator of the surface energy.

Figure 2: Example of contact angle measurement.The red line shows the bottom of the substrate. The part of the drop that appears to be below the red line is reflection off the surface of the substrate. The contact angle is a function of surface energy, and decreases with increasing surface energy.

3 Experimental Methods

3.1 Substrate Functionalization

The surfaces the rubrene was grown on were silicon substrates functionalized with different molecules. Functionalization is the process of modifying a surface to change what molecule (“functional group”) sits on top. In order to modify the surface, the specific molecules chosen were Hexamethyldisilazane (HMDS), Octenyltrichlorosilane (OTS), and Trichloro(1H,1H,2H,2Hperfluorooctyl)Silane (FTS) for their wide range of contact angles. Rubrene was also grown on bare silicon substrates (ie substrates that were not functionalized at all). The substrates were functionalized through a process called spin coating. The molecule was dissolved in a solution and then one drop of the solution was placed on the substrate. The substrate was set on the spin coater, shown in figure 3, which spins the substrate at 3000rpm to spread the drop evenly on the substrate.

One substrate with each molecule was used to determine the contact angle of each surface. Each substrate was placed in front on a camera and had one drop of deionized water placed on it. An image was then immediately captured. The contact angle was measured at the intersection of the water drop and the substrate on both the left and right side. This process was repeated five times on each surface, with the final contact angle determined to be the average of the 10 values. The contact angles for all 4 surface are shown in table 1.

Figure 3: A spin coater inside of a glove box. The substrate is placed on the disk in the center of a cavity. This disk then spins around at 3000pm, ensuring that the drops of molecules in solution sit evenly on the surface of the substrate.

3.2 Growth

An 85nm thin film of rubrene was deposited on all the substrates. The machine used for this is an Angstrom Engineering thermal evaporation chamber (henceforth known as the “growth chamber”) shown in figure 4. The chamber was pumped below 8 × 10⁻⁷ Torr in order to lower the sublimation point of solid rubrene. The rubrene was sitting in a tungsten boat with high current running through it. This heated the boat through Joule heating beyond the sublimation point of rubrene, at which point the rubrene vapor was emitted isotropically in the chamber, a process known as physical vapor deposition.

Figure 4: Thermal evaporation chamber. The rubrene in placed in a tungsten boat at the bottom of the chamber. This tungsten boat is then heated through Joule heating to sublimate the rubrene. This vapor fills the entire chamber, depositing a thin and even film on the substrates sitting on the disk at the top of the chamber.

4 Data Collection Process

Once the thin films were deposited, they were immediately thermally annealed (heated at a specific temperature) to prevent room-temperature growth of crystals. The crystal growth was imaged with a Nikon polarized optical microscope equipped with a heated stage (to anneal the substrates) and CCD camera, shown in figure 5.

Figure 5: Microscope used to image rubrene during annealing process. Each substrate is placed on the heated stage. The microscope lens is focused on a CCD camera, which is then read out to a computer for later analysis. The heated stage was controlled by an external PID controller that controls the temperature to within 0.1^oC.

5 Data Analysis

These images were analyzed by a macro written for the FIJI (Fiji Is Just ImageJ) image analysis software. This macro took the raw images and converted them to a binary image: black is crystalline rubrene and white is anything else in the image. This conversion is shown in figure 6. Then, the macro detects the distance from the center to the edge of the the crystal at a given time, with this detected edge shown as a green dot in figure 6. All these radii (each represented by one green dot) were average together, giving the radius at that time. This radius is a single point on the plot in figure 7. This plot was then fit to a line with a least squares fit. The slope of this fit is the growth rate of the crystal.

Figure 6: Left: Raw images taken with the microscope. Right: Image modified with the FIJI macro to define crystal boundaries. The edge of the four black lobes are defined to be the radius of the crystal at a given time. This radius is measured at each time to determine the growth rate. The average of this growth rate over all crystals on the image is the crystal growth rate for the substrate.

Figure 7: Example of a fit of crystal growth rate. The slope of the least squares fit of the data is the growth rate of the crystal. Notice the data cut from the fit at the very beginning and end of the growth. It takes some time for the crystal to nucleate, so any data prior to nucleation should not be included in the fit. At the end of the growth, crystals start to merge, so the edge-detection algorithm is no longer able to find the edge of the original crystal. Thus, any data after crystals merge also should not be included in the fit.

This process was repeated for five crystals on each substrate, and three substrates at each temperature. The growth rate at a given temperature was taken to be the mean of these 15 values. This process was repeated for each surface at approximately six different temperatures. This data of growth rate vs. temperature was then fit to the Arrhenius law, as shown in figure 8. The axes were chosen such that the slope of the line is the activation energy of crystal growth in eV, as indicated on each graph.

Figure 8: Plots of the fits to the Arrhenius equation. The slope of each line (indicated on the graph) is the activation energy of crystal growth on that surface.

6 Discussion

The results of the all data collected is summarized in table 1. Recall that the expectation from [1] is that as contact angle increases, surface energy should decrease. Notice that the molecules in table 1 are organized in order of increasing contact angle. However, looking at column 3, we see that the surface energy is certainly not decreasing. Indeed, there seems to be no pattern whatsoever, despite having a wide range of contact angles. This suggests that the surface energy is a relatively unimportant factor in determining the contact angle.

In hindsight, the argument presented in [1] should not have been vindicated by the data in the first place. The assumption was that the presence of an electric field on the surface of the substrate should inhibit the motion of the rubrene molecules, making it harder for them to crystallize. This makes the assumption that rubrene is a polar molecule, that is, it has a non-negligible dipole moment. However, rubrene is a hydrophobic molecule [1]. Then, this means that it has a negligible dipole moment. Therefore, the presence of an electric field should not inhibit the crystallization of rubrene molecules.

Table 1: Summary of all data collected during the experiment. The measurement of the activation energy, E_a, on each of the surfaces was well represented by the data: uncertainties are on the order of a few percent and the reduced χ2 of the fits in figure 8 are near 1. Notice that the hypothesis was rejected: There seems to be no correlation between contact angle and surface energy.

7 Conclusion

Upon comparison, it was determined that there is no correlation between the surface energy and activation energy of crystal growth. This result is consistent with other known properties of rubrene, especially its non-polar nature [1]. Future studies could investigate what properties of these surfaces (other than their surface energy) is controlling the activation energy of crystal growth.

8 Acknowledgments

I would like to thank Elias Puchner for his support and guidance as a faculty mentor, Kurt Wick for his support and financial assistance, Russell Holmes for use of his lab equipment and time, and Tom Fielitz, for his suggestion of this project and unending help.

References

[1] S. W. Park, J. M. Choi, K. H. Lee, H. W. Yeom, S. Im, and Y. K. Lee, "Amorphous to crystalline phase transformation of thin fi.lm rubrene," Journal of Physical Chemistry B, vol. 114, pp. 5661-5665, 2010.

[2] S. Arrhenius, "Uber die Dissociationswarme und den Ein us.der Temperatur auf den Dissociationsgrad der Elektrolyte," Zeitschrift fur Physikalische Chemie, vol. 4, pp. 96-116, 1889.

[3] R. Tadmor, "Line energy and the relation between advancing, receding, and Young contact angles," Langmuir, vol. 20, no. eq 1, pp. 7659-7664, 2004.

[4] D. Chandler, "Interfaces and the driving force of hydrophobic assembly," Nature, vol. 437, pp. 640-647, 2005.

-- Main.pheni004 - 14 May 2015