Grazing Incident Small Angle X-ray Scattering (GISAXS)

GISAXS Application in Polymer Physics - Master Thesis Content

4.2.2 Grazing incidence small angle X-ray scattering (GISAXS)

The grazing incidence small angle X-ray scattering is also a SAS technique, which is similar to SAXS but with grazing incidence mode. This means the X-rays impinges on the sample surface at a very shallow angle below 1 degree. Thus it is considered as very powerful technique for the investigation of thin film samples. To enable the X-ray penetrates the film, the grazing incidence angle should be larger than the critical angle of the materials, otherwise only X-ray reflection exists.

Figure 4-7: Geometry of grazing incidence small angle X-ray scattering technique.

In a GISAXS experiment, besides the same components used in the SAXS experiment, a special sample stage with two adjustable angles controlled by rotational motors is used to align the film sample. In addition, one beam-stop needs to stop the direct beam and another one needs to stop the specular beam which is quite strong due to the mirror reflection. Normally the in-house GISAXS experiment can take more than 12 hours to get enough statistic scattering event information on the detector because of the limited X-ray energy for in-house instrument. Therefore, the GISAXS experiments in this thesis are taken at the large scale synchrotron facility where sufficient X-rays fluxes are provided and the time for one high-quality pattern could be less than 0.1 second. This advantage brings the possibility to carry on in-situ GISAXS experiments like sputtering metal nanoparticles on the film surface and use GISAXS to probe the structure changes.

• Static GISAXS experiment at Elettra

The static GISAXS experiments were performed in the Australian SAXS beamline at Elettra Sincrotrone Trieste in Italy. Elettra is the third generation synchrotron with 2 and 2.4 GeV storage ring that has been in operation since 1993. The Australian GISAXS beamline is equipped with two detectors: a 1M Pilatus detector (984 x 1043 pixels) for GISAXS and a 100k Pilatus detector (487 x195 pixels) for GIWAXS with pixel size 172µm x 172 µm for both. And the beam-stops are positioned inside the flying tube, which is sealed with Kapton windows on both sides. The sample is positioned on a rotational stage between the beam shutter and the flying tube. The shutter is operated by the control panel in the controlling room, which is isolated by the safety gate made by concreate and lead. When the sample is positioned ready, the operator should get the safety key and close the gate within 30 seconds. For the static GISAXS experiments, the sample to detector distance is 1967.02 mm and the synchrotron radiation wavelength λ is 0.154 nm with 8keV photon energy. The beam size is narrowed as 0.5 mm x 0.2 mm. The incident angle is set as 0.43o and the exposure time is 300s.

Figure 4-8: The Australian SAXS beamline at Elettra Sincrotrone Trieste.

• Sputtering GISAXS experiment at DESY

The sputtering GISAXS experiments were performed in P03 Micro- and Nanofocus X-ray Scattering (MINAXS) beamline of the PETRA III storage ring at Deutsches Elektronen-Synchrotron in Hamburg. It is equipped with a Pilatus 300k detector (487 x619 pixels, pixel size 172µm x 172 µm) and an ultra-high vacuum sputter chamber which is mounted on a goniometer (Huber). The synchrotron radiation wavelength is 0.954 Å and the photon energy is 13 keV. The sample to detector distance equals to 3988 mm and the incident angle 0.4o is used to clearly separate the Yoneda peak of the involved materials (Au, Si, PS-b-PNIPAM) from the specular peak. The beam is narrowed to 38µm x 19 µm by beryllium lenses. The sputter chamber deposits gold nanoparticles (99.999%, MaTeck GmbH) onto film surface with a sputtering rate at 0.007 nm/s. And it is operated at 150 w power with pre-evacuated vacuum at 0.02 mbar.

Figure 4-9: P03 MINAXS beamline of the PETRA III storage ring at DESY in Hamburg with a) axial view of flying tube and b) side view of sputter chamber.

5.4 Structural Characterizations of PS-b-PNIPAM DBC / Iron Oxide Hybrid Thin Films: Static and Sputtering GISAXS Study

5.4.1 Static GISAXS study of PS-b-PNIPAM DBC/iron oxide hybrid thin film

The thin films of PS-b-PNIPAM / iron oxide nanocomposite with increasing content of PS tailored Fe2O3 nanoparticles are prepared according to the method in chapter 3, and ex-situ GISAXS experiments were performed at SAXS beamline of Elettra Sincrotrone Trieste, Italy. The results are presented in figure 5-19 (a) and (b).

The bare diblock copolymer structure is relatively ill-defined and the Fe2O3 nanoparticles give a very wide peak which does not change with increasing its contents. The ill-defined structure of the DBC thin film compared with bulk samples is mainly due to surface physical/chemical effect on the evolved DBC structures. Possibly further optimization of the surface chemistry and film thickness is essential for further examinations. As we learned from the SAXS profile fitting results that the structure of bare PS-b-PNIPAM DBC bulk samples is lamellar with inter-lamellar distance equals to 26 nm. And the PS-coated Fe2O3 nanoparticles used here is around 10 nm, thus there is possible large confinement effect to incorporate these large size NPs. Possible thin film optimization of this PNIPAM-based DBCs and smaller sized metal-oxide NPs may allow formation of nanostructured hybrid responsive systems in thin film format. This part has not been further investigated in this thesis. However, similar in-situ scattering investigation on PNIPAM based nanostructured DBCs in thin film format to those performed for free-standing bulky films (in this thesis) is an interesting topic to further investigate.

5.4.2 Sputtering GISAXS study of bare PS-b-PNIPAM DBC thin film

The sputtering GISAXS experiments were performed at P03 beamline in DESY, Hamburg. The gold nanoparticles are sputtered on the surface of bare PS-b-PNIPAM thin film with a rate of 0.007 nm/s and sputter power of 150 W, and the GISAXS measurements are operated at the same time to probe the structure change of the deposited gold layer and the growth kinetics of the gold nanoparticles.

From figure 5-20, it is clearly to see the growth process of the gold NPs. Around 3 min after the start of sputtering, the first gold metal characteristic peak appears as a cloud and a quasi-uniform gold layer around 15 min is formed, during this time, also the second and third ordered gold characteristic peaks come from high q values toward a low q values.

The Yoneda peak of the PS-b-PNIPAM is a material-specific characteristic, which reveals the strongest scattering position depending on the critical angle of the material.

Formula 5-2

Where Y is the Yoneda peak position (in pixel) of the material, and y is the pixel position of the direct beam. d is the distance between sample and detector, and p is the size of the pixel. is the incident angle of the x-ray, and is the critical angle of the investigated material.

Thus the Yoneda peak position can be calculated for PS and PNIPAM, it is qz= 0.567 nm-1 (245.5 in pixel) and qz= 0.576 nm-1 (248.19 in pixel) respectively. And the 2D time-resolved mapping (figure 5-20) shows intensity evolution of the horizontal cuts at the Yoneda peak position of PS-b-PNIPAM.

Figure 5-20: Time evolution of GISAXS 2D intensity patterns from a) 0 min to i) 40 min for bare PS-b-PNIPAM thin film during Au nanoparticles sputtering process.

Figure 5-21: a) 2D time-mapping showing the temporal intensity evolution of PS-b-PNIPAM bare thin film GISAXS horizontal cuts at qz= 0.567 ~ 0.576 nm-1. The grey area is covered due to an accidently beam dump, and it is the same reason for all the following graphs. b) Intensity evolution of PS-b-PNIPAM DBC GISAXS horizontal cuts at qz= 0.567 ~ 0.576 nm-1 with stacking on intensity scale for comparison.

Figure 5-22: a) 2D time-resolved map showing the temporal intensity evolution of Au GISAXS horizontal cuts at qz= 0.851 nm-1. b) Intensity evolution of Au GISAXS horizontal cuts at qz= 0.861 nm-1 with stacking on intensity scale for comparison.

The same calculation can be applied for gold NPs, the Yoneda peak position of Au is given as qz= 0.851 nm-1 (345 in pixel). The time evolution of the gold scattering peak can be seen from figure 5-22 (b), it undergoes exponentially increasing in the early stage and finally stopped after sputtering for 30 min. The gold layer formation can be interpreted from nucleation, diffusion, adsorption to grain growth. In the very beginning, the gold nanoparticles exist as small particles and they are the nucleation center. The gold particles grow soon into gold cluster in the diffusion phase. Then these gold clusters adsorb and merge together until percolation threshold. Afterwards a dominated cluster grows by a movement of grain boundaries [46]. The gold layer thickness, however, will still increase and it forms multi-layer structure as shown in figure 5-23 (a) and (b).

Figure 5-23: a) 2D time-resolved mapping showing the temporal intensity evolution of GISAXS vertical cuts at qy= 0 nm-1. The Yoneda peak positons of Au, PS-b-PNIPAM and Si are marked on the left. b) Intensity evolution of GISAXS vertical cuts at qy= 0 nm-1 with stacking on intensity scale for comparison.

From the vertical cuts of GISAXS profiles the layer information can be acquired and in figure 5-23, the 2D time-resolved mapping shows the intensity evolution of the vertical cuts at qy= 0 nm-1. The Yoneda peak positions of Au, PS-b-PNIPAM and Si are also indicated on the map. Since the sputtering chamber gives the constantly sputtering rate of 0.007 nm/s for gold nanoparticles, thus the thickness of the gold layer can be calculated theoretically. After 40 min sputtering of gold nanoparticles, the nominal thickness of gold layer is 16.8 nm.

Figure 5-24: Evolution of Yoneda intensities versus sputtering time for PS (blue), PNIPAM(red), silicon (black) and gold (yellow).

The figure 5-24 presents the evolution of PS, PNIPAM, Si and Au Yoneda peak intensities versus Au sputtering time. Due to ill-defined structure of the prepared DBC thin films (as indicated by absence of any initial polymer characteristic peak), it is not obvious any selectivity phenomenon of gold to either PS or PNIPAM block of the DBC thin film. The selective wetting of gold to one or other block for block copolymer film systems during metal sputtering has been previously investigated by E. Metwalli et al [46-48]. And it has been observed that a strong selectivity to PS domains of the nanostructure PS-based DBC. The growth of the gold NPs on the polymer film show a typical exponential behavior as those reported for homo and block copolymers indicating a metal growth mechanism including four main steps; nucleation, diffusion, aggregation/coalescence processes and boundary movement between NPs [46-47].