Are the initial TiO2 nanorods totally amorphous?

Maybe the phase transformation from amorphous to crystalline anatase occurs at the whole volume of the nanorods and not due to the formation um small crystals, in way that the nanorods are not a mixture from amorphous and crystalline parts but they get entirely more crystalline with the annealing procedure.

Recently I have published a paper that can help you

Annealing temperature effects on sol-gel nanostructured mesoporous TiO 2/SiO 2 and its photocatalytic activity

Strauss, M., Maroneze, C.M., De Souza E Silva, J.M., Sigoli, F.A., Gushikem, Y., Mazali, I.O.

Materials Chemistry and Physics 126 (1-2) , pp. 188-194

doi: http://dx.doi.org/10.1016/j.matchemphys.2010.11.041

Does "see" mean that you already know (by TEM or by another characterization method) you have amorphous phase along with crystalline phase and you want to observe them in an image instead of a diffraction pattern?

Zimimg is right, the SAED patterns for amorphous and "poly" (or quasi -- whatever!!) crystalline (given the size and process, that would be the case here) TiO2 is very similar. Plus, given the size of the nanorods I am assuming the thickness (or the sample length that the beam is going through) is pretty large, messing up any good SAED pattern / image if there is any. Try to locate thinner part of the rod and heat the sample in-situ, inside the TEM. As Mathias points out, if the initial nanorods are not totally amorphous you have a harder task at hand to discern the change. Given right conditions, you can probably polish the nanorod down using FIB and then HRTEM to look for change.

Perhaps this will help you and answer your quest. Http://pubs.acs.org/doi/full/10.1021/nl302017w

Linear scan on STEM may be helpful to confirm amorphous phase containing Ti.

Yes, it can be done. People have done it for Al/Al2O3 transition. Here is the link - http://iopscience.iop.org/0953-8984/14/8/316/pdf/c20816.pdf

Thank you all for the responses, quite helpful actually.

Mathias- good paper. How did you get the transition to anatase at such high temperature? I thought it would convert to rutile above 650C.

Dilek- the film is deposited PVD and is known to be amorphous (XRD) and converts to anatase (with oxygen supplied) at 550C and rutile at >650C. I would like to observe the location of crystal growth at various annealing stages and such need to be able to distinguish between the phases.

Zimimg and Palash- Since the conversion partially depends on oxygen availability since it is O2 deficient during deposition, I would assume the crystal growth would begin on the surface since the oxygen must diffuse inwards. Why would you get a diffraction ring for SAED for a completely amorphous material? Perhaps I need to anneal for less time to catch the very beginning of growth.

Dmytro- Thanks for the paper reference. Yes, they did it, but I guess I am looking for a way to image a whole rod and pull out the amorphous sections versus get a lattice image for every grain.

I see. The size of the rod is too large compared to size of features that can identify amorphous vs crystal sections. You probably could try image the every end of rods. Due to flux O2 through cross section and side wall (at the edge) the crystal growth should start there first. Just thinking loudly.

Hi Michael,

we have not observed any phase transition from anatase to rutile as it is expected (at 650 oC or less). But in our case the TiO2 nanoparticles are dispersed in a SiO2 host and stabilized by Si-O-Ti bondings.

The phase transition from anatase to rutile occurs from the surface to the volume of the particle so that in presence of this superficial bondings the phase transition is not possible or happens at higher temperatures.

Recently I made some tests with TiO2 NPs dispersed in another mesoporous SiO2 matrix and even at 1200oC I have not observed the phase transition!!!!

One additional thing. Are you supposing by XRD analysis that your initial nanorods are amorphous? If each nanorod is formed by very tiny (

Effectively dark field can be useful but you have to used conic integration. You select some diffracted dots or circle (if polycristallized) and proceed by angle or conical rotation. Or you can also select 3 different dark field position at 0, 45 and 90° for one diffracted point. Best option because I think that amorphous layer around or between grain boundaries are very thin is to used HR mode. You can easily illustrated some amorphous layer. The most critical point is the sample preparation which supposed ultramicrotomy. If you some problems feel free to send me one sample and I can try it for free. If success please be kind by assoicating me to the publication. Sincerely yours

Philippe Le Coustumer

You will need to use a high resolution TEM, (one with a field emmission or LaB6 filament) and accelerating voltage of at least 200 kV. Then you will be able to image the lattice of the crystalline regions; amorphous regions will not have dicernable lattice fringes but should still be visable if present. Also imaging in dark field mode may help you to pick out crystalline regions with greater contrast.

Yes you can go for SAXS scan with synchronous source for the purpose, which is an important tool for particle size, shape and surface more precisely in comparison with TEM.

Yes ofcourse TEM be used to image amorphous sections and grain boundaries

I wish I could respond to these within each message, instead of this way. Anyways:

Mathias- thanks yes I guess I can see how a matrix material well-mixed, could interfere with phase transitions.

Sajjad- I will look into the capabilities of Gatan to see if there are any image processing capabilities that would help.

Philippe- I will have to look into conic integration to see better what you mean. If you are referring to tilting the diffracted beam and then imaging it in dark field, then that is my current approach however I suspect you are referring to another strategy I will have to investigate. Thanks for the offer, it is appealing but I will try these suggestions first. I'll give it a couple more tries first.

Dillip- We have access to SAXS and that may actually be a viable tool to use. No synchrotron source thought, I'd have to go down to Brookhaven for that. Without an SLS, what do you estimate the smallest feature size I can discern would be?

Zulkernain- Well, it's nice to know I can do it, but perhaps you could elaborate a little? See the attached images (top of question) for some of what I've already done.

How about HRTEM? At different regions you would be able to see the crystallinity or otherwise.

Yes.. TEM can be used to distinguish amorphus and crystalline material... From SAED pattern we can see the difference.

Yes you can go for SAXS facility, where the size can be resolved even less then 10 nm.

If any more help is required I am ready to support you.

I could be. Please see the following Journal Article: Nanocompositesbased on titaniumdioxide and polythiophene: Structure and properties

Quoc-Trung Vua, Martin Pavlikb, Niels Hebestreita, Ursula Rammelta, Waldfried Plietha, Jiri Pflegerb, a Institute of Physical Chemistry, Electrochemistry, Dresden University of Technology, Mommsenstrasse, Germany b Institute of Macromolecular Chemistry, Academy of Science of the Czech Republic, Heyrovsky Sq, CZ-16206 Prague, Czech Republic

ABSTRACT: A composite of polythiophene (PTh) and nanoscopic titaniumdioxide (TiO2), possessing core-shell structure, was prepared via oxidative polymerization of thiophene by iron (III) chloride in the presence of TiO2 particles. The morphology of the obtained composite particles was studied by transmission e... [more]

Reactive and Functional Polymers

In my view it is not possible to resolve the amorphous domain as TEM has a good resolution up to minimum of 100 nm i.e. the maximum resolution of the crystalline materials for crystal domains.

With HRTEM one can see the lattice spacings.

Yes. You can use TEM to detect amorphous phases in cross sectional nanorods. Be careful about distinguishing between annealing induced and e-beam induced amorphous layers under 200 KeV irradiation, in particular on the thinnest portions of the nanorods (e.g. external surface). To minimize the problem the irradiation time as well as the image and SAD acquisition should be as fast as possible.

I think in image mode you can easily differentiate between an amorphous and crystalline phase, if you are not able to differentiable in diffraction mode by performing some intelligent measurement. Like before going to Image more, You orient the gain of you Guessed crystalline phase along some proper-zone direction, then in image mode it will show a strong contrast as compared to the other gains and only by small tilting, the gain will be out of proper zone direction and its contrast will decrease very fats with a small tilt and with have a contrast as that of other gains. while for amorphous phase the is no zone axis, so will not show a big difference in contrast of image by tilting the sample.

Giancarlo- Yes, it is possible I'm getting some rearranging due to the beam, although I can't see it happen. I will try to minimize beam time in one area. If that is happening then it is likely to induce conversion to anatase phase from amorphous, versus the other way around.

Syed- I will try that. I think I understand what you mean by 'gain' (tilt/shift) but not sure. Also, the problem is not see crystals, I can, but only can distinguish one crystal orientation at a time (at 200kx). Since it is polycrystalline, if I have it set up to image one crystal plane, than all the other crystals not oriented correctly are not visible, and are indistinguishable from amorphous parts.

Thanks again for the input.

TEM is the perfect tool to discriminate crystallized zone from amorphous one. You presented in your ppt data obtained by Dark Field mode which the most accurate one to images the amorphous zone from crystallized one. Please refer to the wotk of Agnes Oberlin on carbon. It is possible using conical Dark Field to observe amorphous and crystal. While crystallized domains will appear lighted for two azimutal position, amorphous zones will look always slightly bright-grey. Then when you ll defocalize the objective lens, bright domains of the crystal will appear unfocus while amorphous will look focused. You can get also data allowing images of the different domains using Lattice Fringes modes with an HRTEM. Your micrography look good to test HRTEM LF. Some recommandations: start at 120 keV accelerating voltage with an emission gun at 2 or 3 (on a scale of 6) with a small objective diaphragm for DF, after used a larger diaphragm objective for LF micrography. TiO2 are quite stable under electron beam and you can work without too much artefact. If you tell me which TEM you used I can gives you more details if necessary