Data Mining with DDView+ and the PDF-4 Databases

Data Mining with DDView+ and the PDF-4 Databases

FeO Non-stoichiometric Oxides

Some slides of this tutorial have sequentially-layered information that is best viewed in ‘Slide Show’ mode

This is one of three example-based tutorials for using the data mining capabilities of DDView+ with the PDF-4+ database and it covers the following topic:

• FeO Non-stoichiometric Oxides– sorting out temperature and stoichiometric effects

on cell parameters

Two other similar tutorials for data mining exist and cover the following topics:

• CIGS Photovoltaics– solid solution / cell parameter relationship

• Carbamazepine Polymorphs– a PDF-4/Organics application – investigating polymorphic forms of an active

pharmaceutical ingredient (API)

Stoichiometric Factors Affecting the Diffraction Pattern of FeO

• FeO is frequently non-stoichiometric with Fe-site vacancies.

• These defects have crystallographic effects and can cause shifts in the observed powder diffraction peaks.

• Summaries of this effect can be “mined” from the PDF-4+ database and displayed for further study.

Ideal Crystal Structure of FeO*

Cubic SystemSpace Group: Fm-3mNaCl type structure

*Structure taken from PDF entry 04-004-7638 calculated from the LPF database.

Data Mining for FeO Entries: Step 1 Use the Preferences Window to establish what will be displayed in the Search Results table . . . 1. Click the ‘Preferences’ icon1. Click the ‘Preferences’ icon

2. Click the ‘Search’ tab in the ‘Preferences’ window

Selecting Fields for the Results Table

Selected Fields:Use these buttons to move a selected item up or down in the listed order for the results table.

Available Fields:Use these buttons to move selected items between the ‘Available Fields’ list of 60 items and the ‘Selected Fields’ list of items that will be displayed in the results table.

Selecting Fields for Results Table (FeO)

Setting up the fields as shown here will serve the purposes of this example.

Chemistry Criterion for Search: Fe and O only entered on ‘Periodic Table’ tab of ‘Search’ window

First click elements Fe and O . . .Then click ‘Only’ . . .

Finally click ‘Add’ to include this criterion for the search.

Once entered, the chemistry criterion can be verified here.

Structure Criterion for Search: FCC Space Group #225 – ‘Fm-3m’entered on ‘Structures’ tab of ‘Search’ window

Perform search using specified chemistry and space group criteria.

Results of DDView+ Search for FeO

A total of 63 entries for Fe1-xO compounds

Analysis of the Resulting Database EntriesMost fields in the results table can be graphically illustrated in either X-Y plot or histogram form.

For the current application, we will use an X-Y plot to illustrate the a-axis cell parameter as a function of atomic % Fe.

To do this, ‘Graph Fields…’ is first selected from the ‘Results’ drop down menu of the ‘Results’ window.

Choose X-axis field from drop down menu: ‘Atomic %’

Choose element for Atomic % values from drop down menu: ‘Fe’

Choose Y-axis field from drop down menu: ‘XtlCell-a’

Click ‘OK’ to draw Graph

FeO a-axis Cell Parameter vs. Atomic % Fe

Ambient

High Temperature

• This graph shows little apparent correlation between atomic % Fe and a-axis. • Note that entries reported as being stoichiometric (1:1) are in a vertical line at

right (50 at% Fe). All others report less than stoichiometric amounts of Fe.• Further examination of individual entries separates ambient and high

temperature determinations of the a-axis.

Stoichiometric – 1:1

• Individual entries from this chart may be examined by left-clicking on individual spots. The circled spot corresponds to the entry on the next slide.

PDF Card for FeO (01-073-2144)

Fe.942O

The ‘PDF’ tab of this window displays the actual formula and stoichiometry.


The ‘Experimental’ tab of this window shows (among other things), the source of the information. Authors, in this case, are B.T.M. Willis & H.P. Rooksby.

This window may be closed to return to the previous graphical plot of ‘a’ vs. atomic % Fe.

Ambient

High Temperature


Ambient

High Temperature


FeO a0 Cell Parameter vs. Atomic % Fe

Choosing another entry from the graph (circled in blue), one can bring up the information shown on the next slide.


Fe.9536O

Reported stoichiometry for this entry is shown on the ‘PDF’ tab.


The reference shown on the ‘Experimental’ tab, for this entry, cites a study by E.R. Jette & F. Foote.

Close window to return to search results graph.

FeO a0 Cell Parameter vs. Atomic % Fe

Formula Reference1. Fe0.942O Willis &

Rooksby2. Fe0.9536O Jette &

Foote3. Fe0.9570O Jette &



Foote 6. Fe0.9712O Jette &


Foote8. Fe0.974O Willis &

Rooksby9. FeO Jette &

Foote

Ambient

High Temperature


1 2 3 4 5 6 7 8 9

A number of points on this graph appear to be linearly aligned with the two examined thus far.

One could check the references of these points to determine if they are from the same sources.

Combining the FeO Data from Just the Willis & Rooksby and Jette & Foote Sources

• To analyze the apparent relationship within the roughly linear data identified in the previous slide, the data must be limited to just the sources identified.

• A particular author can be added to the search criteria to restrict results to just that author.

• Results of multiple searches can be combined, such as results found from two different sources.

• This process is outlined here and illustrated on the next few slides:

– Add author ‘Willis’ criterion to previous FeO search and perform search

– Change author ‘Willis’ to author ‘Jette’ and perform search again

– Combine results from these two searches via the ‘History’ tool

Search FeO Structures for Author ‘Willis’

2 Results

With the ‘Just (Fe And O)’ and ‘Space Group #225’ still in place on their respective tabs on the ‘Search’ window, the author ‘Willis’ can be added to the ‘References’ tab and the search performed.

The results of this, and any other ‘Search’, are automatically saved in the ‘History’ file and can be accessed for the duration of the DDView session.

Search FeO Structures for Reference ‘Jette’

7 Results

Similarly, a second search is performed substituting author ‘Jette’ for ‘Willis’ of the previous search.

Use of the ‘History’ Tool to Combine Searches

The ‘History’ feature is accessed via the ‘Tools’ menu.

Combine the Results of These Two Searches

The top area shows a history of searches performed during this session. These searches may be individually selected and copied to the ‘Combined Searches’ area via the down arrow button. Alternatively, they can both be selected using ‘Ctrl-click’ and copied together via the down-arrow button.

The selected searches may be combined with ‘And’ to include only those entries that appear in both searches or with ‘Or’ to include all entries that appear in either search. For this exercise, we are interested in all entries, so ‘Or’ is chosen, then the ‘Combined Results’ button performs the operation.

Combined Author Results for FeO EntriesThe results table includes all 9 entries, which may be graphed similarly to the way the original 63 hits were graphed.

‘Atomic % Fe’ vs. ‘XtlCell-a’

Click ‘OK’ to create graph

Graph of FeO Cell Parameters derived by Willis & Rooksby and Jette & Foote

These data may be exported to a ‘.csv’ file for more involved mathematical or statistical analysis using a spreadsheet program. For example, one could explore whether these data are linear (Vegard’s Law) or if there is a systematic (and significant) deviation from linearity.

FeO a-axis Cell Parameter vs. Atomic % Fe - High Temperature Studies

Ambient

High Temperature


Several entries from this high temperature region reference a single study, “Point Defect Clusters in Wuestite” by Radler, Cohen, & Faber, J. Phys. Chem. Solids, 51, 217 (1990). To quickly examine which points come from this study, the search can be reperformed using one of these authors.

Change Author for Search

On the ‘References’ tab, ‘Radler’ (or ‘Cohen’ or ‘Faber’) can be entered in the ‘Author’ field to obtain the desired PDF entries. The ‘Structures’ and ‘Periodic Table’ tabs contain the previously entered information regarding FeO and space group 225. This search will yield the 13 entries found on the next slide.

PDF Entries from Radler, Cohen, & Faber Study (1990)

The a-axis cell parameter for just these entries can now be plotted vs. atomic % Fe and the resulting graph is shown on the following slide.

Graph of Radler, Cohen, and Faber FeO Entries

1

2

1. Data for varying T from 1123 to 1373 C with slightly increasing O content

2. Data for varying stoichiometry at 1323 C from Fe0.88O to Fe0.95O

Return to the original list of 63 FeO entries . . .

Use the ‘History’ icon to retrieve any searches already performed during this session of DDView+.

Choose the original search:

{Only (Fe And O)} And {International Space Group Number Exactly ‘225’}

and click the ‘Results’ button.

Another Look at FeO results - Density

Another Look at FeO results - DensityAnother way to analyze this data is to look at the density as a function of cell parameter. This is performed by using the ‘Results’ drop down menu to access ‘Graph Fields…’

Entries with 1:1 atomic ratio reported for FeODensity is inversely proportional to cubic cell parameter

“An -ray Study of the Wuestite (FeO) Solid Solutions” - Jette & Foote & “Change of Structure of Ferrous Oxide at Low Temperature” - Willis & Rooksby. Density increases very slightly as cell parameter goes up with increasing Fe content.

“Point Defect Clusters in Wuestite” - Radler, Cohen, & Faber Density decreases as a axis increases (inverse relationship) as T is varied from 1123 to 1373 C with slightly increasing O content

Results: density vs. a axis cell parameter for reported FeO structures

Both density and a axis increase dramatically as stoichiometry at 1323 C is increased from Fe0.88O to Fe0.95O.

Summary for Non-stoichiometric Cubic FeO

• Multiple explanations exist for unit cell parameter variations in non-stoichiometric FeO in the PDF

• Systematic studies regarding stoichiometry and/or temperature can be “mined” from the database

• No single relationship describes all the data, thus different “defect” arrangements must exist for these materials

• Ability to access PDF entries directly from graphs’ facilitates obtaining other data and references

International Centre for Diffraction Data

12 Campus Boulevard

Newtown Square, PA 19073

Phone: 610.325.9814

Fax: 610.325.9823

Thank you for viewing our tutorial. Additional tutorials are available at the ICDD web site (

www.icdd.com).

http://www.icdd.com/

Documents

Data Mining with DDView+ and the PDF-4 Databases