Pretty simple: by HRTEM, you 'see' the structure of your object which, in your case, is at least partly crystalline. In fact, what you really see, as Erico rightly pointed out, is the result of interference from your electron beam and crystal structure (any good textbook on HRTEM will explain that). So you're visualizing the crystal in real space (the object space), i.e. you 'see' the electron density distribution which is stronger at the atomic position. From this, you can directly measure lattice spacings etc. if you are in the right orientation with respect to the electron beam.

If you take the FFT^2 (i.e. you loose the phase but keep the modulus of the FFT) of your HRTEM image, what you basically get is the frequency distribution of the intensity of your object, i.e. a pattern looking like the XRD one (in reciprocal/Fourier space). That is because XRD is (by theory) the square of the 3D Fourier transform of your electron density (see any introductory textbook of XRD).

So remember, the FFT(HRTEM image)^2 = XRD image.

Now, SAED means selected area electron diffraction. What you do there, is that you image the electron diffraction of your object (by imaging the Fourier plane) and then select a specific diffraction peak which is imaged for each scan point. In this way, only the scan position of your object which is in the correct orientation will give you some intensity.

So all of those provide you complementary information, you can't really say that one makes more sense than the other. The only 'detail' you should remember, is that in HRTEM, you're only visualizing a (very, very, very) small portion of your sample (in the 100 nm range), while the beam size in lab source XRD is huge (in the mm range). If your sample is homogeneous, both would give the same information (assuming you have a fully crystalline sample). If your sample is heterogeneous, then you are getting a statistical average by XRD. So for practical reasons, you mayt want to use one or the other.

Hi Bidhan, a HRTEM image is built up from the wave function coming out of the sample. The fringes periodicity observed in a HRTEM image depends on the diffraction contrast mechanism in the image formation process. So, when you apply the Fourier Transform function to the HRTEM image you will have the pattern observed in the FFT, and it will show you the symmetry of the "lattice" fringes on your HRTEM image. You might know that. Regarding your question, the electron diffraction (SAD, CBED, or NDB) patterns give you direct information on the crystal lattice, but not the FFT. You can rely on the FFT once you make sure your imaged crystalline particle is well oriented along a zone axis (keep in mind a TEM image is a 2D projection of your 3D material). In this case, the symmetry on the FFT will correspond to that of the SAD pattern. Then you can use the FFT to index the observed spots.

Regards,

Pretty simple: by HRTEM, you 'see' the structure of your object which, in your case, is at least partly crystalline. In fact, what you really see, as Erico rightly pointed out, is the result of interference from your electron beam and crystal structure (any good textbook on HRTEM will explain that). So you're visualizing the crystal in real space (the object space), i.e. you 'see' the electron density distribution which is stronger at the atomic position. From this, you can directly measure lattice spacings etc. if you are in the right orientation with respect to the electron beam.

If you take the FFT^2 (i.e. you loose the phase but keep the modulus of the FFT) of your HRTEM image, what you basically get is the frequency distribution of the intensity of your object, i.e. a pattern looking like the XRD one (in reciprocal/Fourier space). That is because XRD is (by theory) the square of the 3D Fourier transform of your electron density (see any introductory textbook of XRD).

So remember, the FFT(HRTEM image)^2 = XRD image.

Now, SAED means selected area electron diffraction. What you do there, is that you image the electron diffraction of your object (by imaging the Fourier plane) and then select a specific diffraction peak which is imaged for each scan point. In this way, only the scan position of your object which is in the correct orientation will give you some intensity.

So all of those provide you complementary information, you can't really say that one makes more sense than the other. The only 'detail' you should remember, is that in HRTEM, you're only visualizing a (very, very, very) small portion of your sample (in the 100 nm range), while the beam size in lab source XRD is huge (in the mm range). If your sample is homogeneous, both would give the same information (assuming you have a fully crystalline sample). If your sample is heterogeneous, then you are getting a statistical average by XRD. So for practical reasons, you mayt want to use one or the other.

thnx Erico T. F. Freitas for your answer,

how the SAED pattern originate compare to FTT then?

and from FTT spots, what can be calculated related to crystal information?

regards,

hello Aurelien Gourrier sir,

thank u for your reply. now the origin of FTT is clear to me.

but, from FTT spots, miller indices can be calculated. I've seen it in some of the research papers. How? can u plz help me?

regards,

Question is quiet authentic for the every materials science researchers, I have also the same problem, but what I think is the FFT is an important for geting the preliminary  idea regarding the planes of the materials which can be then focused for the SAED pattern as per the ratio method.

thank you Liu Ping,

but, from FTT spots, miller indices can be calculated. I've seen it in some of the research papers. How? can u plz help me?

Miller indices can be indexed from both the FTT spots and SAED results, HOWEVER, it's a big project, you may find some tutorial from the website. For the reason of language difference, my resource may not suitable for you. Sorry for that i can't help you as you expected.

Sorry for the misdirection in my first answer, the BUT the SAED is more accurate benefiting from a further correction of the angle of copper grid is a wrong description. Actually, a correction was conducted in the HRTEM for a high quality of  lattice fringes BUT not in SAED. We recorded a HRTEM in 550k magnification and SAED in 30k, so the main difference may originate from the difference in magnification. MORE CLEARLY, a SAED is an observed result in a larger area, the FTT is a translation from the HRTEM results, you can do FTT on your PC if you have the HRTEM results. Benefiting from the DigitalMicrograph software, you can obtain spots with high quality in a rather small area with some treatments.

Very thanx Mr. Ping for your answer

Hi Bidhan,

As you have already been told, yes you may index a FFT image taken from a HRTEM image and attibute the Miller indeces to the observed spots once you make sure your crystalline particles are in the right orienation as Aurelien have mentioned. Also, as he told you the FFT gives you information of a very small area of your sample. You may generalize providing your sample is homogeneous.

As Aurelien has mentioned, use XRD if you need statistical average crystallographic information. If you need very localized crystallographic information, then use electron diffraction performed in a TEM.

Always prefer to get crystallographic information from an actual electron diffraction pattern (SAED, CBED, NBD, LACDIF/LACBED, microdiffraction...) or from and XRD.

Going back to the core of your original question, the HRTEM image is essencially a "diffractogram", because it is built up from interference of the electron waves diffracted by the crystal lattice of your sample. However, the phase information of the diffracted electron waves are *not* transfered into the HRTEM image, but the their amplitude. The FFT gives you the frequency distribution of the pixels intensities in the HRTEM image, so any periodicity observed in the HRTEM image will give you spots in the FFT. So the FFT may interpreted as a "diffraction pattern" providing the crystalline particles are in the right orientation.

If your TEM does not give you enough capability to perform electron diffraction in a very small region of your sample, you will still able to do so by getting crystallographic information from the HRTEM image, through its FFT though. However, take in mind the uncertainty in any measurement from FFT is larger than from a electron diffraction pattern.

Cheers,

Erico

Thank you so much Erico T. F. Freitas sir.