This is the second Q&A blog from the series of X-ray diffraction webinars. During this webinar, Dr. Daniel Lee discussed how powder XRD research can be applied to many material types such as battery powders, geological samples, powder metallurgy for 3D printing, catalyst deterioration, respirable silica or pharmaceutical APIs, you can leverage on XRD. To watch the first webinar click here

General XRD

Why is the angle marked as 2 theta and not as theta?

Because the geometrical reason for X-ray diffraction (XRD) during the measurement, the diffracted beam must be considered from the incident X-ray beam. Please refer to the theory of XRD and Bragg’s law.

What do you mean by boxes. One box is of one element with crystal size and orientation.

The slides were prepared to give you more understanding about XRD theories. So, the boxes in slide 12,13 were unit cell, and the others were grain/crystallite of the sample in slide 13-18.

Can there be a detailed discussion on use of XRD in understanding clay minerals? ­

We could plan one if there are many requests.

How to determine Fe2+ in any rock sample? ­

Firstly, we need to do a Search/Match to judge what phases (minerals) are present. Then we could start a Rietveld refinement to quantify them, and as a result, divalent iron can be calculated based on chemical composition and wt.%.

­LOD can be determined using calibration curve­

Yes, it is a spiking method. If there are no standard/reference samples, we can do the math according to the ICH guideline. That uses signal and noise.

­Is it possible tell about amount of doping % with XRD spectra with and without doping??

You can refer to “Vegard’s law” which describes that peak position shifts as a function of how much dopants were substituted by seeing a peak shifting from the non-doped sample. There is some linear correlation between doping ratio and shifting.

In case of metal powders, how do the different particle sizes of particles affect peak intensities? ­

A peak broadening is observed when particle size is close to nano range. And because X-ray penetration depth of the sample can be limited against large metallic particles having a greater X-ray absorption coefficient.

Does the u-strain % distinguish between tensile and compressive stress­?

I do not think so. Theoretically, the micro-strain starts from zero. But, if you measure your sample using residual stress method, you can derive + and – value of stress by measuring one (hkl) plane with a various tilt in omega or chi to monitor peak shifting.

­In CHC plus, there is stated humidity meanwhile TTK 600 was not stated a humidity, what is the different? ­

CHC and MHC+ chambers are designed for controlling humidity and temperature for organics. But, TTK600 does not have any humidity sensor and humidifier in the system, but you can still insert humid gases by connecting hoses and flask, then flow inert gas into the water of flask. It is kind of indirect way.

How to differentiate between phases if they happen to have almost same peak positions in PXRD pattern? ­

It sounds like your sample is consisted of two similar structures. Even though those phases are chemically different from each other, they could be identical in crystallographic properties including crystal system, space group and unit cell parameters. This is called isomorphs which is the most difficult job for XRD. Example: Steel304 vs Steel403, LiNiMnO2 vs LiMnCoO2, Fe vs Fe: Co…

Can I measure each phase at different layers without destructing sample? I mean I need to know the phase at layer 1 and to know the phase layer 2 and something like that. ­

In Bragg-Brentano geometry, you cannot because the X-ray will penetrate whole layers of the sample if the thickness is below several tenth of micrometres.

So, you need to use a Parallel beam geometry instead. You can fix the incident beam angle (omega) during the measurement around 1~2 degrees, then you can only move detector in 2 theta. The appropriate incident angle to capture a very thin skin layer of your sample is depending on a) incident angle, b) density of sample, c) mass absorption coefficient and d) layer thickness in accordance with Lamber-Beer equation. You can vary the omega angle (fixed value) to get a depth resolved XRD patterns. It is called GI-XRD (Glazing Incidence X-Ray Diffraction).

Amorphous

How can we calculate the amorphous area and crystalline area­?

By finding peaks and fitting a profile. HighScore software will make your job easier. Feel free to contact us.

Will you cover quantification of amorphous phase content in this presentation?

It could be covered in a separate session in the future. It indeed takes half a day.

There is a broad peak for amorphous material. is there any characteristic structural information related to this peak position and width for the relative amorphous material? ­

It is hard to derive structural information from the amorphous material through XRD instrument. By the way, the peak position of the hump is related to the original crystalline phase of that material. So, organic compounds show their peak position at around 5 to 20 degrees while inorganic compounds have 20 to 50. Metallic is far away from them, 35 to >90. If you really want to extract more structural information of your amorphous sample, you can refer to PDF (Pair-Distribution Function) Analysis that can also be done by our XRD equipped with Ag-tube and some specially required parts. Once you measure amorphous or disordered material by PDF technique, you finally get the boding lengths between atoms from the short-range orderings even they are not crystalline.

How do we differentiate between two amorphous phases in the same sample?

If two amorphous phases have a similar chemical composition, then we have no chance to distinguish them due to the humps that appear at the same 2θ position. However, if they are quite different in chemistry, we may see two separated humps at different positions.

How to quantify the amorphous pattern, using HSP software, for example silica?

There are several ways to do that in HighScore software. If you can add a certain amount of known crystalline powder into the sample, we could use an “internal standard method” by using Rietveld algorithm. If adding is not allowed, then we could use an “external standard methods” by measuring 100% crystalline materials with very similar chemical composition as a standard. By doing so, we will get some constant to calibrate it and calculate amorphous contents of the unknown material.

Battery Applications

How can I distinguish mixed battery material, LCO, NCM, LMO, LNO? I found that most of the peaks of the materials are overlapped­

They are all rhombohedral structure in crystal system, and the space group also same. So, you can distinguish them by seeing peak intensity ratio of (003) and (104) for example. The best way is measuring XRF. Please get more info from us.

XRD Software

which software did you tell us to convert the XRD data in image form to XY data form? ­

HighScore Plus from Malvern Panalytical

How to extract data points from additional graphics pane (FWHM micro-strain data points)?

You can use any commercial XRD software like our HighScore software that helps you to extract FWHMs easily.

Sample Preparation

The outcome of quantitative analysis depends on sample preparation such as consistency of particle size of the calibration set and samples, right?

Yes, right. But those factors can be handled and corrected by software afterwards. More serious issue would be inhomogeneity of the powder sample. So, keeping a homogeneity is key by mixing them well or measuring many samples to get an average.

We have some cases in which we need to do refinement of atomic co-ordinates to obtain good profile fitting and quantification of multiphase mixtures. What are the precautions that we must take in this case?

In case of mixture sample, you need to be careful about the refinement of any atomic coordinates when the peaks are overlapped and/or other factors are influencing the intensity as well. So, do the sample prep and measurement as perfect as possible to get a randomly oriented PXRD pattern. The best way is to use K-alpha1 system and fill your powder into the glass capillary for the Debye-Scherrer geometry in order to ignore preferred orientation effect.

If some doping causes the crystal to shrink by imparting the strain… what will happen to peaks?

If the dopant is smaller than the host ion, the peak will be shifted to the higher angle side, vice versa. Moreover, if the dopant deteriorates the host lattice, the peak widths will be broadened with respect to the influences of the dopant.

Suppose I have 2 samples with A-5%B and A-10%B. I did the XRD. Will there be any differences in the peak and their positions? ­

Yes or No. It depends on the case. Can you tell us a bit more information?

Data Analysis and Results

Currently my interest is on metals and alloys dislocation density, stress-strain analysis by XRD data analysis as a result of various processing of metals and alloys

We need to separate your request into two: firstly, dislocation density in the metals/alloys could be handled by powder guys, but residual stress could be covered by metallurgy team who are specialized in stress/texture analysis.

Can we get the dimension of crystal cell from PXRD data only?

No. You can also use SC-XRD (Single Crystal XRD) or TEM. But it is no doubt that the easiest way is use of PXRD. The advantages of PXRD is giving you short measurement time, minimum sample preparation and representative result in statistics compared to the others.

How can we differentiate grains and crystallites from XRD data?

Those are the same.

What is the minimum peak height to consider it a significant peak?

It strongly depends on the sample quality and measurement conditions. But, in mathematics, the peak intensity must be higher than 10 times of standard deviation of the background. Then it can be recognized as a peak and quantifiable.

In case polymer composites is there any detection limit in terms of volume fraction of the filler? ­

We need to say in weight fraction. So LoD (limit of detection) would be 1 wt.% as a best.

When split peaks seen. Is it suggested to smooth it as single peak and analyse?

Depends. But the smoothing is always not recommended for analysing a PXRD pattern due to the mistreatment.

Will you get the same size from all the peaks in the pattern?

If you are talking about a single phase, it would be relatively similar. But every single reflection has its own directions. So, same size of crystallites will not be obtained. That is why we use an average value to ignore that effect.

­Does composition affect peak height? ­

Yes. Because the peak intensity is a function of number of electrons at the certain lattice plane.

Can you explain why the high angle peaks become more broadened in comparison to low angle when defects are there? ­

Because a peak at higher angle represents a small distance between atomic planes. If some defects existed in your lattice, overall structure would still be okay. But local part is not.

While indexing the peaks, what is the minimum number of peaks that need to be matched with a known pattern­

At very least, 3 peaks are needed. But, recommend getting more than 5.

Does the peak height in X-Ray diffraction is depends on the volume fraction of material? ­

Usually, we need to say that it depends on a weight fraction since scattering power of X-ray is related to the number of electrons.

How much shift in peak should be consider as peak shift due to doping??­

In general, below 0.02 deg would be considerable. But make sure that what is the angular accuracy or reproducibility of your instrument. We have 0.01 deg in error.

Why there is background at lower 2 thetas? ­

Due to the presence of air, and some reflections from the sample and holder surface.

In analysis most of the characteristic peaks in alloys are at lower angles? So, what significant results can be concluded from high angle peaks? ­

At higher angle, you can see more small changes in the lattice, for example, local distortion/inhomogeneity in atomic position/defects (dislocation, point or screw defects, stacking faults, anti-domain, and so on)

This webinar is ready to watch on-demand. Please click the link below to watch the video. Stay tuned for more XRD webinars!