GSAS-EXPGUI Workshop

00:57:47
https://www.youtube.com/watch?v=rG14YjLK9xQ

Resumo

TLDRHierdie video verskaf 'n gedetailleerde uiteensetting van hoe om die GSAS en X3 sagteware suite te gebruik vir die analise van poeiervormige X-straal diffraksiepatrone. Dit beklemtoon die stappe wat betrokke is by die skep van 'n generiese lêer in GSAS en die gebruik van X3 om die data visueel te sien. Die gids fokus op ontmodelskepping, insluitend herkalibrering en demping, en bied insigte oor struktuur modifikasie, fase toevoeging en parameters wat gebruik word om die kwaliteit van die modelle te verbeter. Belangrike modelle soos die Reitveld metode en die gebruik van die etiketmetode word bespreek, tesame met hoe om met fase mengsels en dunfilms met gepreferensieerde oriëntasie om te gaan. Verskeie nutsmiddels en parameters soos LX, GW, GU word verduidelik om die breedte en simmetrie van pieke aan te pas. Die gebruik van gespesialiseerde databasisse soos die ICSD (Inorganic Crystal Structure Database) om kristaldata te bekom en die gebruik van sferiese harmoniese modelle om gepreferensieerde oriëntasie te evalueer en te kwantifiseer, word ook uitgelig.

Conclusões

  • 🧑‍💻 GSAS en X3 stel gebruikers in staat om visuele analise van poeiervormige X-straal diffraksiepatrone te doen.
  • 🔍 Die proses behels die skep van 'n generiese lêer in GSAS en gebruik X3 om data te visualiseer.
  • ⚙️ Rietveld-herkalibrering is 'n kerntegniek gebruik vir modellering van kristal struktuur.
  • 🌐 Databasisse soos ICSD is nuttig vir die verkryging van kristaldata.
  • 📊 Geskikte dempwaardes verbeter die affiniteit van modelle.
  • 📊 LX en ander parameters is noodsaaklik om pieksimmetrie aan te pas.
  • 🔄 Reitveld metode en etiketmetode help by die verwerking van verskillende fase mengsels.
  • 🧪 Gebruik van sferiese harmoniese modelle kwantifiseer gepreferensieerde oriëntasie.
  • 🗃️ Fase toevoegings in modelle moet akkuraat wees vir beter analise.
  • 💡 Die video bied 'n stap-vir-stap gids vir die gebruik van die GSAS en X3 sagteware suite.

Linha do tempo

  • 00:00:00 - 00:05:00

    Die spreker bespreek die gebruik van GSAS en X3 sagteware vir die analise van poeier X-straal diffraksiepatrone. Die proses begin met die skep van 'n generiese lêer in GSAS voordat dit in X3 gevisualiseer kan word. Die voorbeeld behels annotase titaandioksied en die opstel van 'n eksperiment om rekristallisasie te modelleer.

  • 00:05:00 - 00:10:00

    Die spreker beklemtoon die verkryging van fase-inligting uit die ICSD databasis, die seleksie van hoë kwaliteit kristaldata, en die gebruik van SIF-lêers as 'n beginpunt vir modellering. Hierdie fases word by die projek gevoeg en vorm die basis vir verdere verfyning.

  • 00:10:00 - 00:15:00

    Die gesprek verduidelik die invoer van data in die GSAS-formaat vir histogramme en die gebruik van die PowerPref-funksie vir aanpassing. Daar word gefokus op die pasmaak van modelle aan werklike data deur middel van verfyning en die belang van korrekte agtergrondinstelling.

  • 00:15:00 - 00:20:00

    Verfyningsprosesse sluit die aanpassing van piekhoogtes, posisies, en die agtergrond in om die kristalmodel akkuraat te maak. Die spreker gebruik 'n Chi-kwadraat waarde om die akkuraatheid van die model aan te dui en stel verdere aanpassings voor.

  • 00:20:00 - 00:25:00

    Die spreker ondersoek die invloed van kristallietgrootte met die LX-parameter en bespreek verdere verfyningsmetodes om die modelresultate meer akkuraat te maak, insluitend die aanpassing van piekprofiele en die agtergrond.

  • 00:25:00 - 00:30:00

    Daar word gewys op die hantering van veelvuldige fases en hoe die sagteware fasefraksies kan bereken met behulp van gewigsverhoudings. 'n Nuutlakeerde voorbeeld word gedemonstreer met annotase en rutile mengsels.

  • 00:30:00 - 00:35:00

    Dunfilmvoorbeelde word bespreek met spesiale aandag aan voorkeursorientasie-effekte, wat die tradisionele Rietveld-metodes uitdaag. Daar word gebruik gemaak van die Label-metode om voorkeursorientasie-effekte te verreken.

  • 00:35:00 - 00:40:00

    Die gebruik van die rekordmetode en die aanpassing van ander parameters soos simmetrie, spanning, en basiese lyne om akkurate modelle uit te werk word gedemonstreer. Verskeie aanpassings van agtergrond en parameters word bespreek.

  • 00:40:00 - 00:45:00

    Die spreker bespreek die gebruik van die sferiese harmoniese model om voorkeursorientasie te kwantifiseer. Hierdie metode laat 'n meer gedetailleerde analise toe en die tekstuurindeks dien as 'n maatstaf vir tekstuureffekte.

  • 00:45:00 - 00:50:00

    Die proses word herhaal vir alternatiewe konfigurasies van dunlae met verskillende deposito-tye, en vergelyk die voorkeursorientasie-indekse tussen monsters. Daar is verdere bespreking oor die gebruik van tekstuurindekse vir vergelykings.

  • 00:50:00 - 00:57:47

    Die samevatting sluit af met die belangrikheid van praktyk en die gebruik van handleidings om die volledige potensiaal van die sagteware te leer ken, en beklemtoon die relatiewe intensiteit van pieke vir kristallografiese analises.

Mostrar mais

Mapa mental

Mind Map

Vídeo de perguntas e respostas

  • Wat is GSAS en X3?

    GSAS en X3 is sagteware gereedskap wat gebruik word vir die analise van poeiervormige X-straal diffraksiepatrone.

  • Hoe werk die Reitveld model?

    Reitveld model word gebruik om die kristalstruktuur binne X-straal diffraksie data te verfyn deur die posisies en breedtes van pieke aan te pas.

  • Wat is die etiketmetode en hoe word dit toegepas?

    In X3 word die etiketmetode gebruik om gepreferensieerde oriëntasie in dunfilms te kwantifiseer en aan te pas.

  • Wat is ICSD en hoe word dit gebruik?

    ICSD staan vir Inorganic Crystal Structure Database, wat gebruik word om kristaldatale vir modellering te bekom.

  • Wat is die doel van sferiese harmoniese modelle in hierdie konteks?

    Sferiese harmoniese modelle word gebruik om die kwantifisering van gepreferensieerde oriëntasie in kristalmodelle te verbeter.

Ver mais resumos de vídeos

Obtenha acesso instantâneo a resumos gratuitos de vídeos do YouTube com tecnologia de IA!
Legendas
en
Rolagem automática:
  • 00:00:00
    today I'm going to be talking to you
  • 00:00:02
    about how to use the gsas and X3
  • 00:00:06
    software suite for
  • 00:00:10
    analyzing your X-ray diffraction
  • 00:00:12
    patterns
  • 00:00:13
    happens this is for analyzing powder
  • 00:00:16
    x-ray diffraction patterns rather than
  • 00:00:20
    looking at single Crystal data the first
  • 00:00:24
    thing that you do when you use gsas and
  • 00:00:26
    X3 is your first open gsas
  • 00:00:29
    and you just have to create a generic
  • 00:00:31
    file that then you can then
  • 00:00:34
    visualize and excrete so gsas is kind of
  • 00:00:38
    like the older version that you can do
  • 00:00:40
    stuff in DOS and you can just type lots
  • 00:00:43
    of information into that but to actually
  • 00:00:45
    visually see what's going on XP is what
  • 00:00:48
    you use but but you need to set up an
  • 00:00:51
    experiment using gsess so sorry you do
  • 00:00:53
    that by pressing setup experiment name
  • 00:00:55
    you give this photo an arbitrary name
  • 00:00:58
    and our first example we'll be looking
  • 00:01:00
    at is is annotase titanium dioxide so
  • 00:01:03
    just call it annotase
  • 00:01:06
    yes we'll create that file
  • 00:01:10
    and I think I made it on the desktop
  • 00:01:11
    here
  • 00:01:13
    so I hope it
  • 00:01:15
    I'm not sure where it was then you have
  • 00:01:18
    to press setup expedit so experiment
  • 00:01:21
    edit
  • 00:01:23
    okay and it says do you want to create
  • 00:01:25
    this file on your desktop yes
  • 00:01:27
    what will we call it annotes and now
  • 00:01:29
    it's appearing
  • 00:01:31
    do I want to do anything else so it
  • 00:01:33
    gives you a whole list of options of
  • 00:01:34
    what you can actually do within the gsas
  • 00:01:38
    software
  • 00:01:40
    but I don't like dos format I like to
  • 00:01:42
    see what's going on I like a user
  • 00:01:45
    interface so I'll just press X for now
  • 00:01:47
    and exit it so I've just basically made
  • 00:01:49
    myself a nice blank file here and I
  • 00:01:52
    can't but that's
  • 00:01:55
    nothing's been entered yet
  • 00:01:57
    now
  • 00:01:59
    um I've got some examples in X3 that
  • 00:02:03
    I've already got set up okay so here are
  • 00:02:06
    the examples
  • 00:02:09
    and first we're going to look at and
  • 00:02:11
    annotate powder and we're going to do a
  • 00:02:14
    rebuild refinement for that
  • 00:02:16
    now I know how to do recoil refinement
  • 00:02:18
    and label refinements okay so if you're
  • 00:02:22
    powder there's no thermodynamic effects
  • 00:02:25
    on
  • 00:02:26
    which
  • 00:02:28
    Crystal planes it grows so it grows into
  • 00:02:31
    what what is in the literature you know
  • 00:02:33
    the the different Peak Heights are what
  • 00:02:36
    the literature powder pattern would have
  • 00:02:37
    so basically there was no strain on
  • 00:02:40
    which in which it was growing it just
  • 00:02:42
    grow to where it's thermodynamically
  • 00:02:44
    Optimum state in different Crystal
  • 00:02:46
    planes different uh orientations
  • 00:02:49
    so the refill model is great for that
  • 00:02:51
    okay
  • 00:02:53
    but then if you want to do a label
  • 00:02:54
    refinement
  • 00:02:56
    where you just relax that structure
  • 00:02:58
    Factor that's that's it you know
  • 00:03:02
    constraining that model
  • 00:03:04
    so you can use it there will find model
  • 00:03:06
    if you have preferred orientation
  • 00:03:08
    effects
  • 00:03:09
    and I'll teach you how to deal with that
  • 00:03:11
    as well so for now we'll just model
  • 00:03:14
    a powder
  • 00:03:16
    which grows very similar to what you see
  • 00:03:19
    in the literature
  • 00:03:20
    in terms of its preferred orientation so
  • 00:03:23
    we'll just use the x-free software now
  • 00:03:26
    and it will ask you okay open up this
  • 00:03:29
    file so I've created this blank file
  • 00:03:32
    here annotates experiment I'm just going
  • 00:03:35
    to move it to the powder
  • 00:03:37
    here to delete this blank that I made
  • 00:03:39
    earlier
  • 00:03:42
    so in in export you I can go to these G6
  • 00:03:45
    examples and it taste powder wheat but I
  • 00:03:48
    can then open this experiment
  • 00:03:51
    and this is the user interface okay so
  • 00:03:54
    I'm going to teach you how to do a
  • 00:03:55
    rebuild model
  • 00:03:56
    and a label equally weighted model
  • 00:04:00
    um so the first thing you want to do is
  • 00:04:03
    enter what phase you're trying to model
  • 00:04:06
    so before you even go into this software
  • 00:04:09
    and you're thinking I'm going to model
  • 00:04:10
    my pattern you should first use the
  • 00:04:13
    correct software to identify what phases
  • 00:04:16
    are present
  • 00:04:18
    so once your identify it is and for this
  • 00:04:20
    particular example it's annotase
  • 00:04:22
    you can then add that phase here
  • 00:04:25
    so what phase where do I get this phase
  • 00:04:28
    from well you can get it online
  • 00:04:32
    from the crystal Orchestra database TSD
  • 00:04:38
    so if you're going to be modeling
  • 00:04:40
    patterns quite Lily
  • 00:04:42
    what I recommend you do
  • 00:04:44
    is you become a member of the icsd
  • 00:04:49
    do using the icsd
  • 00:04:53
    is you can just search for that
  • 00:04:55
    particular material that you know it is
  • 00:04:57
    okay so then
  • 00:05:00
    um I know chemistry
  • 00:05:03
    of this material called Anna taste
  • 00:05:05
    that's the middle of the material
  • 00:05:07
    called Anna taste
  • 00:05:12
    okay and then it will give you
  • 00:05:15
    um you know all of the standards that
  • 00:05:17
    have been uploaded to the system
  • 00:05:18
    and what I like to do is I'd like to
  • 00:05:21
    click on this filter and just take high
  • 00:05:23
    quality crystal data only
  • 00:05:25
    and then you can click on any of these
  • 00:05:26
    that are of interest to you say this one
  • 00:05:29
    and show a detailed view of it
  • 00:05:32
    and it will give you
  • 00:05:35
    the paper in which it was published
  • 00:05:38
    um
  • 00:05:39
    sorry the the title of that publication
  • 00:05:42
    where it was published I think is
  • 00:05:45
    somewhere as well I can't see oh here it
  • 00:05:47
    is
  • 00:05:48
    so then you have a reference
  • 00:05:50
    yeah
  • 00:05:52
    um and it will give you some structural
  • 00:05:54
    parameters and information
  • 00:05:55
    it will also give you
  • 00:05:58
    um a pattern of how it looks and you can
  • 00:06:00
    config configure that pattern
  • 00:06:02
    you can tell it to display the indices
  • 00:06:07
    so then you can you can know oh okay
  • 00:06:09
    that's the one zero one plane one zero
  • 00:06:11
    three zero zero four one one two Etc
  • 00:06:14
    and you can even change it so you can
  • 00:06:16
    change the minimum or maximum values
  • 00:06:18
    that you want to change this range here
  • 00:06:21
    for instance from 10 to 20 degrees two
  • 00:06:23
    feet of there's no information so you
  • 00:06:25
    might just want to go from 20 elements
  • 00:06:26
    and then you can extract that by
  • 00:06:29
    pressing export X Y data okay
  • 00:06:33
    so you also need to know when you're
  • 00:06:35
    doing your X-ray diffraction
  • 00:06:37
    you need to know what was your X-ray
  • 00:06:39
    source
  • 00:06:40
    so
  • 00:06:42
    um here to get this
  • 00:06:44
    standard you know as a sort of a
  • 00:06:47
    starting point for the model you have to
  • 00:06:49
    export a Sif file
  • 00:06:52
    so export that Sif
  • 00:06:54
    and using that Sif
  • 00:06:56
    I'm ready to put it here so it's the
  • 00:06:58
    same it's the same thing here
  • 00:07:00
    so this will be kind of like my starting
  • 00:07:02
    point of what I'm trying I'm going to
  • 00:07:04
    take the starting point and actually
  • 00:07:06
    make it reach what I've got Okay so
  • 00:07:10
    in this phase section you add the phase
  • 00:07:13
    I've got a sip file
  • 00:07:15
    [Music]
  • 00:07:17
    and I'm going to upload this zip file
  • 00:07:20
    press continue to just say okay okay
  • 00:07:22
    yeah yeah
  • 00:07:24
    put all of those things in thank you
  • 00:07:26
    and now this is the first phase that
  • 00:07:29
    we're gonna it's a starting point and
  • 00:07:31
    we're going to refine that starting
  • 00:07:32
    point to match our actual data that
  • 00:07:34
    we've got but I haven't even uploaded my
  • 00:07:36
    data yet that's the next thing I have to
  • 00:07:38
    do so under the histogram tab you add
  • 00:07:42
    so when you go to histogram you click on
  • 00:07:44
    add new histogram
  • 00:07:47
    and if you select your data file
  • 00:07:50
    but it has to be in the correct format
  • 00:07:53
    which I haven't shown you how to do
  • 00:07:56
    so
  • 00:07:58
    if you have
  • 00:07:59
    your data it's simply like this so you
  • 00:08:02
    have degrees two feet on the left and
  • 00:08:03
    counts on the right
  • 00:08:05
    that that's something that's that's
  • 00:08:07
    handy
  • 00:08:08
    a handy way of having your data then you
  • 00:08:10
    can convert that into a gsas format okay
  • 00:08:13
    so using this called convex.exe tool you
  • 00:08:18
    can do that and so the file type that
  • 00:08:20
    I've got is just a simple ASCII and I
  • 00:08:22
    want to convert that to a GSS
  • 00:08:26
    so I select the file that I want to
  • 00:08:28
    convert
  • 00:08:29
    so that is um
  • 00:08:32
    where is that it's under C drive
  • 00:08:46
    I thought it was here
  • 00:08:52
    [Music]
  • 00:08:57
    okay
  • 00:08:59
    there it is
  • 00:09:01
    looking forward to adapt file the ATF
  • 00:09:03
    but it was a DOT txt file so I didn't
  • 00:09:05
    see that but they're both equivalent I
  • 00:09:08
    think for the software so it's fine so
  • 00:09:10
    now press do that convert thing
  • 00:09:15
    and then here I get my this is that
  • 00:09:18
    ASCII file but just in a gsas format
  • 00:09:20
    that likes to see the data of your X-ray
  • 00:09:24
    pattern
  • 00:09:26
    so now within the software I can load
  • 00:09:28
    you know as I said before go to
  • 00:09:30
    histogram add new histogram I can load
  • 00:09:32
    this histogram here by looking at
  • 00:09:35
    for that gsas file
  • 00:09:38
    if you're pressing open
  • 00:09:41
    we like it it is a gsas format I know
  • 00:09:44
    it's fine
  • 00:09:45
    then you have to just select bank then
  • 00:09:47
    you have to put an instrument collab
  • 00:09:49
    instrument parameter file so this is
  • 00:09:52
    something that whoever's in charge of
  • 00:09:54
    that machine they'll have an instrument
  • 00:09:55
    parameter file for that specific xrd
  • 00:09:57
    machine and within that instrument
  • 00:09:59
    parameter file it'll have information as
  • 00:10:02
    I told you before about is it coppers my
  • 00:10:06
    source
  • 00:10:07
    um
  • 00:10:09
    of X-rays and other information that's
  • 00:10:12
    vital
  • 00:10:14
    for this software I've already got that
  • 00:10:17
    information here it's got broad Phillips
  • 00:10:18
    the ins it's a bit old but it works
  • 00:10:23
    so I'll just add that
  • 00:10:27
    and then press add and everything's done
  • 00:10:29
    now
  • 00:10:31
    now if I ever want to see what's going
  • 00:10:34
    on I press the live plot button
  • 00:10:36
    but I can't see anything yet because I
  • 00:10:39
    haven't told it any changes that I've
  • 00:10:41
    made I haven't told the software okay
  • 00:10:43
    make those changes and to do that I
  • 00:10:46
    don't know what the hell these buttons
  • 00:10:47
    mean but you have to press powerpref
  • 00:10:50
    wow
  • 00:10:51
    and then join us
  • 00:10:54
    so all of the stuff I've just done so
  • 00:10:56
    adding the face adding my so that's a
  • 00:10:59
    phase in which we're going to starting
  • 00:11:01
    point
  • 00:11:01
    and then adding my actual x-ray data now
  • 00:11:04
    I can actually see it on the live plot
  • 00:11:07
    so these X's are my data
  • 00:11:12
    the green line is just the background
  • 00:11:13
    but it's drawn for me and the red is the
  • 00:11:16
    model
  • 00:11:17
    so as you can see the model
  • 00:11:20
    is kind of in the right place but the
  • 00:11:22
    heights aren't right okay so that's
  • 00:11:24
    something that that needs to be refined
  • 00:11:26
    I'm guessing also the bits won't be
  • 00:11:28
    correct so you can see the position is
  • 00:11:30
    slightly out there are many things that
  • 00:11:31
    we need to start to refine to get this
  • 00:11:33
    get this to work
  • 00:11:36
    but I think first thing first is just to
  • 00:11:37
    check that the background's okay
  • 00:11:42
    and if I zoom in
  • 00:11:44
    I can see that my background pretty much
  • 00:11:46
    goes through my noise
  • 00:11:49
    there's no point in really editing it in
  • 00:11:51
    this case
  • 00:11:52
    um but there are cases where you have
  • 00:11:54
    like a broad background because the
  • 00:11:56
    amorphous uh glass substrate is is
  • 00:12:00
    giving you something there could be
  • 00:12:02
    other reasons
  • 00:12:04
    so then you have to draw it back on
  • 00:12:06
    through that
  • 00:12:08
    and you can do that by clicking add and
  • 00:12:11
    so forth but I'll show you how to do
  • 00:12:12
    that later for an actual scenario where
  • 00:12:16
    it's important to do it
  • 00:12:18
    now let's actually start modeling or
  • 00:12:21
    we'll soon start modeling things one
  • 00:12:23
    thing that's important is the profile
  • 00:12:24
    now a lot of you guys are dealing with
  • 00:12:27
    thin films
  • 00:12:28
    and this peak cutoff is just simply not
  • 00:12:31
    wide enough to deal with the widths of
  • 00:12:34
    the piece that you get because we're
  • 00:12:37
    dealing with quite Nano crystalline
  • 00:12:39
    materials quite often so we need to
  • 00:12:41
    change this P card off we need to add a
  • 00:12:43
    couple of zeros okay
  • 00:12:45
    and there's something called damping now
  • 00:12:49
    the higher my damping value the shorter
  • 00:12:52
    the steps the software will take in
  • 00:12:55
    finding the local minimum okay when it's
  • 00:12:57
    trying to fit a model to your data okay
  • 00:13:02
    so it's always advisable to start with a
  • 00:13:04
    high number
  • 00:13:05
    and move slowly towards that local
  • 00:13:08
    minimum rather than going huge steps
  • 00:13:10
    because you can really mess things up
  • 00:13:12
    so I just like the number seven I think
  • 00:13:14
    it's quite lucky
  • 00:13:15
    so I've done that to phase as well cell
  • 00:13:18
    damping
  • 00:13:19
    so I think we're ready to start modeling
  • 00:13:23
    things
  • 00:13:24
    so
  • 00:13:25
    you're going to do a week Board model
  • 00:13:27
    it's already stuck on the beatboard here
  • 00:13:30
    and each cycle will do seven cycles of
  • 00:13:32
    modeling
  • 00:13:33
    so I'm going to start modeling the face
  • 00:13:36
    and nothing else so I'm just going to
  • 00:13:38
    move the positions of the Peaks so one
  • 00:13:40
    thing here is called scaling so if you
  • 00:13:42
    remember the heights of my model
  • 00:13:44
    we're not close to matching the heights
  • 00:13:46
    of my
  • 00:13:48
    um actual data
  • 00:13:50
    this is ticked on so it's gonna it's
  • 00:13:52
    gonna move that the heights the peak
  • 00:13:54
    Heights to the correct place I don't
  • 00:13:55
    change the damn thing here that's not a
  • 00:13:57
    big problem
  • 00:13:58
    so now to start modding it I press
  • 00:14:00
    Powerpuff
  • 00:14:02
    and then join us
  • 00:14:05
    and that's modeling it
  • 00:14:07
    and here my reduced Chi Squared is 32 so
  • 00:14:10
    it's obviously not great there are some
  • 00:14:11
    issues
  • 00:14:13
    that we're going to have to fix
  • 00:14:16
    [Music]
  • 00:14:20
    again it's just not getting there I
  • 00:14:22
    think maybe we need to keep it going so
  • 00:14:25
    we'll press genre again
  • 00:14:31
    all right
  • 00:14:34
    and we'll just keep pressing until it's
  • 00:14:36
    replied there we can see that slowly
  • 00:14:38
    slowly my Chi Squared is going down
  • 00:14:41
    now it's 17.
  • 00:14:43
    and let's have a look at the Life part
  • 00:14:45
    see how we're doing
  • 00:14:47
    so we're getting much closer
  • 00:14:50
    it was just the scale factor was a
  • 00:14:52
    problem so I'm just going to keep going
  • 00:14:54
    so I'm going to increase this number
  • 00:15:00
    just so we can get the scale factor in
  • 00:15:02
    the correct place but it's not really
  • 00:15:03
    changing much now there are other things
  • 00:15:05
    that I need to change
  • 00:15:07
    now in terms of the profile
  • 00:15:09
    the profile
  • 00:15:12
    is things that change the width and
  • 00:15:14
    asymmetry of your Peaks it can also
  • 00:15:16
    shift your Peaks as well so the main
  • 00:15:19
    three things that I use
  • 00:15:21
    that I think are really important are
  • 00:15:22
    really handy ones are LX l y and shift
  • 00:15:26
    now LX
  • 00:15:28
    it's to do with
  • 00:15:29
    the average crystallite of your material
  • 00:15:32
    the average crystallite size
  • 00:15:34
    so that's probably one of the first
  • 00:15:36
    things that you want to model but it
  • 00:15:38
    just depends so shift
  • 00:15:41
    sometimes you need to shift your two
  • 00:15:43
    Theta data because you didn't align your
  • 00:15:46
    sample very well okay so when you do an
  • 00:15:49
    x-ray diffraction pattern you need to
  • 00:15:51
    make sure that the height of the sample
  • 00:15:55
    is in the correct place
  • 00:15:57
    in this case I'll be starting with just
  • 00:16:00
    modeling LX now whether whenever you
  • 00:16:03
    check or uncheck a box in retail mode
  • 00:16:06
    you always have to press Powerpuff first
  • 00:16:08
    and then join us again before you start
  • 00:16:10
    modeling things
  • 00:16:14
    and now my Chi Squared is down to 12. it
  • 00:16:17
    means things are getting quite good and
  • 00:16:19
    as you can see it is getting quite good
  • 00:16:23
    it's getting quite close
  • 00:16:25
    I just need to continue refining and
  • 00:16:29
    hopefully we'll get there
  • 00:16:34
    okay
  • 00:16:35
    so right now I finally have my refined
  • 00:16:37
    cell that is changing a A and C because
  • 00:16:41
    a is equivalent to B it's changing a and
  • 00:16:43
    C of my unit cell
  • 00:16:45
    the width of my unit cell and and length
  • 00:16:48
    and right now it's currently changing
  • 00:16:50
    the average crystallite size so it's
  • 00:16:52
    doing all of these things so try and
  • 00:16:54
    match my pattern
  • 00:16:58
    but as you can see
  • 00:17:00
    it's not really changing much now
  • 00:17:04
    and it's not quite reaching the width
  • 00:17:06
    that's predicted
  • 00:17:09
    I think there's perhaps there's a
  • 00:17:11
    problem with the pattern
  • 00:17:16
    well the way it's trying to model these
  • 00:17:17
    points too much down here rather than
  • 00:17:19
    the points up here
  • 00:17:21
    so maybe I have to change this manually
  • 00:17:23
    because it's not doing a very good job
  • 00:17:26
    so right now it's
  • 00:17:29
    four
  • 00:17:32
    yeah that number is four so I'm just
  • 00:17:34
    going to change it manually to say
  • 00:17:37
    two
  • 00:17:40
    and not model it and see if that helps
  • 00:17:42
    things along
  • 00:17:49
    let's make things worse
  • 00:17:56
    that's a bit weird isn't it shouldn't be
  • 00:17:58
    doing that
  • 00:18:04
    I think the issue is that this peak
  • 00:18:07
    is quite wide
  • 00:18:10
    whereas it matches quite well with the
  • 00:18:11
    other piece
  • 00:18:14
    so I should leave it as it is and let it
  • 00:18:16
    just refine that but I'll change the
  • 00:18:18
    damping a bit so it can go in bigger
  • 00:18:20
    steps
  • 00:18:24
    so it's modeling quite well
  • 00:18:28
    on average all of the pizza being
  • 00:18:29
    modeled quite well
  • 00:18:36
    I think that's the best it's going to
  • 00:18:38
    get for that Peak but for the rest of
  • 00:18:39
    these Peaks it's doing quite a good job
  • 00:18:43
    ish
  • 00:18:44
    so
  • 00:18:46
    there are other sort of fighting things
  • 00:18:47
    that you can do to try and improve that
  • 00:18:49
    so let's just try it
  • 00:18:52
    and changing
  • 00:18:54
    The Strain perhaps The Strain is causing
  • 00:18:57
    an issue I doubt it
  • 00:19:03
    it's not strong maybe a there's a slight
  • 00:19:06
    error in the measurements needs to be
  • 00:19:08
    shifted
  • 00:19:15
    again that's not the issue now we come
  • 00:19:19
    to these parameters gug and GW
  • 00:19:22
    now this just changed the changes the
  • 00:19:24
    gaussian shape
  • 00:19:26
    um of your material to try and make it
  • 00:19:28
    match
  • 00:19:29
    what I prefer to deal with LX first
  • 00:19:32
    before I deal with anything else because
  • 00:19:33
    this is actually a real
  • 00:19:35
    a number that you can then relate to
  • 00:19:37
    crystallite size and these numbers you
  • 00:19:38
    can't really relate to crystallite size
  • 00:19:40
    so it's best not to fight with these
  • 00:19:41
    first it's best to fight about these
  • 00:19:44
    things afterwards okay
  • 00:19:47
    so
  • 00:19:48
    let's start with gu and see if we can
  • 00:19:50
    improve things
  • 00:19:55
    and that's made things not better at all
  • 00:20:01
    [Music]
  • 00:20:04
    I just have a feeling that it's not
  • 00:20:06
    going to get any better
  • 00:20:08
    it's not bad
  • 00:20:12
    but it's not great either
  • 00:20:14
    after this example I hope you understand
  • 00:20:16
    how then we've got these we've got the
  • 00:20:18
    lettuce parameters now
  • 00:20:20
    and again from the profile we've got LX
  • 00:20:22
    which can be related to the average
  • 00:20:24
    crystal out size of the material using
  • 00:20:27
    this equation here
  • 00:20:32
    so normally we take it from the share
  • 00:20:34
    operation normally we take k
  • 00:20:36
    as some arbitrary number around 0.9 is
  • 00:20:39
    often what's used
  • 00:20:41
    and X here will be your LX number and
  • 00:20:44
    this will give you the average Crystal
  • 00:20:45
    Light size in amstrongs
  • 00:20:47
    okay so don't forget to convert it to
  • 00:20:49
    nanometers which is the convention
  • 00:20:51
    in our field
  • 00:20:53
    as materials
  • 00:20:55
    scientists
  • 00:20:57
    so I think that example is a bit much
  • 00:20:59
    done but one thing it can do is if you
  • 00:21:01
    press results
  • 00:21:02
    in list View
  • 00:21:06
    it lists some really important things
  • 00:21:09
    that we've just modeled for instance Chi
  • 00:21:11
    Squared so that's the measurement of how
  • 00:21:14
    good our fit is
  • 00:21:15
    but generally
  • 00:21:17
    what people like as as crystallographers
  • 00:21:19
    as the wrp value and if this wrp value
  • 00:21:23
    is low so 0.58 I don't think is that
  • 00:21:27
    great we didn't get it so the model fit
  • 00:21:29
    very well to the model
  • 00:21:33
    um and also you can then if you scroll
  • 00:21:35
    down
  • 00:21:36
    you've got your a b and c that's my unit
  • 00:21:40
    cell parameters for my check trigonal
  • 00:21:42
    annotation itself and this is the error
  • 00:21:45
    associated with each value is coded
  • 00:21:47
    below
  • 00:21:48
    so that's also interesting to have
  • 00:21:52
    okay
  • 00:21:55
    so move on to the next example
  • 00:21:59
    so the next example we have
  • 00:22:03
    is a phase mixture
  • 00:22:10
    [Music]
  • 00:22:13
    so we have an annotate root tile powder
  • 00:22:16
    phase mixture
  • 00:22:23
    [Music]
  • 00:22:24
    let's have it have to give it a blank
  • 00:22:26
    experiment file
  • 00:22:28
    I'll just call it annotated root cell
  • 00:22:33
    mixture
  • 00:22:35
    not open it here
  • 00:22:41
    so again I'll add the phases so from the
  • 00:22:45
    database I've already got my annotating
  • 00:22:47
    retail files
  • 00:22:52
    I'll add those
  • 00:22:54
    so my first phase is annotase
  • 00:22:57
    my second phase I'm adding is router
  • 00:23:01
    foreign
  • 00:23:03
    [Music]
  • 00:23:14
    which I might need to convert using
  • 00:23:16
    convex yep
  • 00:23:39
    put my instrument parameter file
  • 00:23:42
    press add
  • 00:23:45
    and then press power prep
  • 00:23:48
    and join us so it can register those
  • 00:23:50
    changes those additions I've just made
  • 00:23:53
    and see it on the live clock
  • 00:23:56
    at this time you can see we have more
  • 00:23:59
    Peaks
  • 00:24:00
    so this is the P that we saw earlier
  • 00:24:03
    this is the annotate101 peak and now
  • 00:24:06
    this is the root r110p
  • 00:24:09
    there'll be more we'll be modeling both
  • 00:24:11
    phases
  • 00:24:12
    together
  • 00:24:13
    so if we can see the backgrounds
  • 00:24:17
    it's a pretty flat background so I don't
  • 00:24:21
    need to modify that at all so it's quite
  • 00:24:23
    fine
  • 00:24:26
    so what I'll do is
  • 00:24:28
    again I'll dampen my cells
  • 00:24:40
    change my Peak cutoffs make it slightly
  • 00:24:42
    wider
  • 00:24:50
    and then I'll start to model things
  • 00:24:52
    first thing I'll do is I'll just model
  • 00:24:54
    the phases the Emma tase number one
  • 00:24:56
    phase root of number two face
  • 00:24:59
    remember in default modes you check
  • 00:25:02
    things press Powerpuff first then drawn
  • 00:25:04
    us afterwards in about in the barrel
  • 00:25:06
    mode it's different
  • 00:25:09
    now I'll just let it refine for a while
  • 00:25:12
    17 cycles and my
  • 00:25:16
    chi-squared value is not going down a
  • 00:25:19
    huge deal
  • 00:25:20
    now we're starting to get better
  • 00:25:29
    I will find again
  • 00:25:31
    it's not getting much better let's take
  • 00:25:34
    a look at how that model is fitting my
  • 00:25:37
    data it's not doing a great job okay and
  • 00:25:40
    that's because I haven't modeled the
  • 00:25:43
    face factions
  • 00:25:45
    okay so under scaling you can tell these
  • 00:25:49
    face fractions okay start start to let
  • 00:25:52
    it go to what is the true base fraction
  • 00:25:55
    now because it starts off with a number
  • 00:25:57
    one and one that that doesn't
  • 00:25:58
    necessarily mean 50 50. okay
  • 00:26:03
    um because
  • 00:26:06
    your different phases have different
  • 00:26:09
    densities and they defect and x-rays
  • 00:26:12
    slightly differently
  • 00:26:14
    so you have it the software takes us all
  • 00:26:16
    into account so one one here does not
  • 00:26:19
    necessarily mean 50 50. I'll tell you
  • 00:26:21
    how to get the real
  • 00:26:23
    face fractions out afterwards but for
  • 00:26:26
    now we will find those
  • 00:26:30
    file press
  • 00:26:32
    and jobless before second side
  • 00:26:43
    we'll do that again
  • 00:26:46
    thank you
  • 00:26:53
    it's doing okay it's just not going to
  • 00:26:55
    the right Peak Heights I think I need to
  • 00:26:57
    start modeling the width of my piece
  • 00:27:04
    so now under profile again I'll start
  • 00:27:06
    with my LX
  • 00:27:09
    which is related to my average Crystal
  • 00:27:11
    Light size
  • 00:27:15
    and you're finding that
  • 00:27:21
    it doesn't oh it's gone a bit weird
  • 00:27:29
    and it's fine now
  • 00:27:36
    some reason it's going very strange at
  • 00:27:38
    the end
  • 00:27:42
    [Music]
  • 00:27:43
    what's going on
  • 00:27:48
    I guess I have too many things checked
  • 00:27:51
    so
  • 00:27:55
    um I'll just put some arbitrary numbers
  • 00:27:57
    here and try and figure out what's going
  • 00:27:59
    on
  • 00:28:00
    it's a good thing when things go wrong
  • 00:28:01
    in the sense that you can learn how to
  • 00:28:04
    fix them as
  • 00:28:13
    okay that's working fine now what I'll
  • 00:28:16
    do is I'll just turn the scaling off
  • 00:28:19
    and
  • 00:28:22
    the letters parameters off
  • 00:28:23
    but then now you find
  • 00:28:26
    the average Crystal size because
  • 00:28:28
    sometimes in your model if you have too
  • 00:28:29
    many variables
  • 00:28:31
    it's not a good thing
  • 00:28:36
    it seems behaving itself a bit better
  • 00:28:39
    now
  • 00:28:41
    again it's not perfect
  • 00:28:42
    [Music]
  • 00:28:44
    but it's quite close
  • 00:28:47
    okay
  • 00:28:48
    so that's not bad
  • 00:28:50
    and again you can go through many
  • 00:28:51
    iterations of trying to improve the
  • 00:28:53
    average crystallite size or if there's
  • 00:28:55
    an error in your shift because you think
  • 00:28:57
    the samples misaligned
  • 00:28:59
    or if there's any strain
  • 00:29:01
    things like this you can play around
  • 00:29:03
    with go through many iterations until
  • 00:29:05
    you reach the optimum
  • 00:29:07
    the one thing that I think is pretty
  • 00:29:08
    interesting using the light bot if
  • 00:29:11
    you're not familiar with where the Peaks
  • 00:29:13
    are you can look at tick marks so if you
  • 00:29:16
    press tick marks phase one it shows you
  • 00:29:19
    where all of the Peaks are the annotate
  • 00:29:21
    space I'm just tick marks
  • 00:29:26
    phase two
  • 00:29:28
    it shows you all the Peaks where
  • 00:29:30
    the rutile phases
  • 00:29:35
    okay
  • 00:29:37
    so here here
  • 00:29:40
    Etc and one thing once you're done
  • 00:29:42
    modeling your pattern if you want to
  • 00:29:44
    have all of this information in a way
  • 00:29:46
    that you can then publish it
  • 00:29:48
    you can press file export plot as a CSV
  • 00:29:51
    file
  • 00:29:54
    and then you can open this CSV file
  • 00:30:03
    all right
  • 00:30:10
    in Excel
  • 00:30:16
    okay
  • 00:30:24
    so then you have potato and you can play
  • 00:30:27
    about it however you want
  • 00:30:37
    so now in terms of getting the phase
  • 00:30:40
    fractions
  • 00:30:41
    what we can do is we can again press
  • 00:30:43
    results
  • 00:30:45
    list View
  • 00:30:50
    and it tells you the wrp which is a
  • 00:30:52
    measure of the error
  • 00:30:58
    so it's too wind
  • 00:31:02
    your Chi Squared again is another
  • 00:31:03
    measure of the arrow
  • 00:31:05
    so used to for your histogram
  • 00:31:09
    it gives you information about
  • 00:31:13
    uh let's say okay
  • 00:31:17
    oh because we weren't modeling the
  • 00:31:19
    lattice parameters in the last cycle it
  • 00:31:22
    doesn't show us to those parameters
  • 00:31:24
    so what I'll do is I'll turn these
  • 00:31:26
    things off
  • 00:31:27
    these things back on
  • 00:31:41
    check that everything's okay
  • 00:31:45
    it's fine now I can press results
  • 00:31:48
    this View
  • 00:31:52
    and then once we scroll down
  • 00:31:56
    foreign
  • 00:32:04
    this is my a b and c my lattice
  • 00:32:06
    parameters the errors
  • 00:32:08
    phase two the last parameters and errors
  • 00:32:13
    and I think
  • 00:32:17
    okay
  • 00:32:21
    okay here we are
  • 00:32:22
    so phase element fractions for phase
  • 00:32:25
    number one weight fraction is 0.42
  • 00:32:29
    and weight fraction is 0.57 for phase
  • 00:32:31
    number two
  • 00:32:32
    so it's from these weight fractions that
  • 00:32:35
    then you get your uh
  • 00:32:39
    your composition
  • 00:32:41
    okay
  • 00:32:43
    now I'm dealing with the same phase the
  • 00:32:45
    same
  • 00:32:47
    um chemical composition is that there
  • 00:32:49
    are two different polymorphs of to2
  • 00:32:53
    so if 0.42
  • 00:32:56
    weight fraction is annotated and 0.57
  • 00:33:00
    weight fraction
  • 00:33:01
    is rutile then that is actually my
  • 00:33:03
    composition it's 43 annotates and 57
  • 00:33:07
    router because both are different
  • 00:33:09
    polymorphs of titanium dioxide
  • 00:33:11
    but if you have two different materials
  • 00:33:13
    then the weight factions won't
  • 00:33:15
    necessarily be the atomic fractions so
  • 00:33:18
    you need to account for that
  • 00:33:22
    okay and that's that example done
  • 00:33:27
    so the next example I want to look at is
  • 00:33:30
    looking at
  • 00:33:32
    um a thin film rather than a powder
  • 00:33:45
    [Music]
  • 00:33:49
    yeah
  • 00:33:51
    so again using the software X3 we open
  • 00:33:55
    that
  • 00:33:56
    go to that
  • 00:34:00
    and because it's a thin film we have
  • 00:34:02
    preferred orientation you don't you
  • 00:34:03
    don't always get preferred orientation
  • 00:34:05
    effects within films without you very
  • 00:34:07
    often do
  • 00:34:08
    so
  • 00:34:09
    same routine
  • 00:34:12
    remember whenever you have
  • 00:34:14
    an x-ray pattern you have to identify
  • 00:34:16
    which phases are there I've already done
  • 00:34:18
    this it's just annotates
  • 00:34:20
    so I'm uploading a good starting point
  • 00:34:24
    um and annotates from the literature Sif
  • 00:34:27
    bar I've shown you how to do that
  • 00:34:28
    earlier
  • 00:34:29
    and add my histogram which I've already
  • 00:34:31
    converted to gsas before
  • 00:34:40
    okie doke so I want it to register what
  • 00:34:43
    I've just done
  • 00:34:49
    and then I look at the live plots
  • 00:34:51
    [Music]
  • 00:34:52
    so this is the the sort of
  • 00:34:55
    founding background I was telling you
  • 00:34:56
    about before it's Broad and I know that
  • 00:34:59
    I know this feature because I see it
  • 00:35:00
    often this is just due to the amorphous
  • 00:35:03
    silica glass that's underneath the
  • 00:35:05
    substrate
  • 00:35:06
    so to get rid of that rather than try
  • 00:35:09
    and model it it's made me amorphous you
  • 00:35:12
    don't really want to care to model it
  • 00:35:14
    you just have to edit the background
  • 00:35:18
    so what you do is you zoom in
  • 00:35:21
    and you click here and there where you
  • 00:35:23
    think
  • 00:35:24
    that background should be
  • 00:35:28
    and the more time and period that you
  • 00:35:30
    take with it the better
  • 00:35:36
    now sometimes with certain machines you
  • 00:35:38
    get little bumps and values of two Theta
  • 00:35:41
    because there's an error of the detector
  • 00:35:43
    or something like that so you need to be
  • 00:35:45
    aware where the artifacts are in that
  • 00:35:48
    spectrometer if that's the case if
  • 00:35:50
    there's a problem that's spectrometer
  • 00:35:52
    and model the background report will
  • 00:35:54
    draw the background accordingly
  • 00:35:57
    and sometimes at the edges the
  • 00:35:59
    background goes a bit weird so you just
  • 00:36:01
    have to add a few more points but but
  • 00:36:04
    then not add any more terms and it does
  • 00:36:06
    a better job
  • 00:36:10
    now I'll press paragraph
  • 00:36:13
    join us it should register what I've
  • 00:36:16
    just done those changes
  • 00:36:19
    and now I can see that the background
  • 00:36:21
    goes much better
  • 00:36:23
    through the data
  • 00:36:27
    the one thing I can notice straight away
  • 00:36:30
    so let's just look at the relative
  • 00:36:31
    Heights of my Peaks
  • 00:36:34
    this peak here the 112 peak in annotates
  • 00:36:37
    it doesn't show the same
  • 00:36:39
    distribution
  • 00:36:41
    as it did before
  • 00:36:43
    so
  • 00:36:47
    I know those preferred orientation in
  • 00:36:48
    that respect and over here I think this
  • 00:36:51
    is the two one one or one one two peak
  • 00:36:54
    vanities there's a problem there as well
  • 00:36:56
    so I know there's preferred orientation
  • 00:36:58
    so if I try and fit things to a record
  • 00:37:00
    model it's just not going to work
  • 00:37:02
    it'll only work if I use
  • 00:37:05
    um
  • 00:37:06
    a refill plus a March doll Ace or
  • 00:37:10
    cerebral harmonic preferred orientation
  • 00:37:12
    model added on top
  • 00:37:15
    okay
  • 00:37:16
    but if you don't care to do that you
  • 00:37:18
    don't care to quantify your preferred
  • 00:37:20
    orientation in some in some respect you
  • 00:37:23
    don't care to quantify it
  • 00:37:24
    then you can just relax those preferred
  • 00:37:26
    orientation effects using the label
  • 00:37:28
    method and I'll show you how to do that
  • 00:37:30
    but first things first let's change why
  • 00:37:33
    people cut off make it wider and change
  • 00:37:36
    my dampings accordingly
  • 00:37:46
    and now I'm ready to do my label method
  • 00:37:48
    now with lateral methods you have to
  • 00:37:52
    have everything unchecked when you press
  • 00:37:55
    half breath
  • 00:37:57
    the only thing that you can have checked
  • 00:37:58
    is the scaling nothing else can be
  • 00:38:01
    checked
  • 00:38:02
    and with the with the label method you
  • 00:38:04
    have to press power prep once
  • 00:38:07
    and then join this twice
  • 00:38:11
    I don't know why interested
  • 00:38:14
    okay now I can start to check things so
  • 00:38:17
    I want to start to refine my face so
  • 00:38:24
    I'll do it a couple of times
  • 00:38:28
    now look at my material
  • 00:38:31
    if you remember before the peak heights
  • 00:38:33
    were quite different it was
  • 00:38:36
    the distribution was the middle was the
  • 00:38:38
    highest and the two on the edges were
  • 00:38:39
    the same height but you see it's
  • 00:38:41
    starting to relax that structure Factor
  • 00:38:44
    and allow things just to fit
  • 00:38:47
    in any thermodynamic Direction okay
  • 00:38:50
    so we just need to continue to press
  • 00:38:53
    dramas
  • 00:38:55
    try and reach the local minimum
  • 00:38:57
    I can now start to model the crystallite
  • 00:39:00
    with
  • 00:39:03
    doing a much better job now my Chi
  • 00:39:05
    Squared is down to 1.8
  • 00:39:08
    it's not going to get much better than
  • 00:39:09
    that
  • 00:39:12
    and we'd be
  • 00:39:14
    almost there I'd say and you just need
  • 00:39:16
    to go and do more and more iterations to
  • 00:39:18
    try and get things to match more
  • 00:39:19
    perfectly
  • 00:39:20
    but as you can see there's some slight
  • 00:39:22
    asymmetry
  • 00:39:23
    perhaps potentially there's some rutile
  • 00:39:26
    impurity there where it could just be
  • 00:39:28
    noise
  • 00:39:30
    um so you know
  • 00:39:32
    you can then model the asymmetry the uh
  • 00:39:38
    the strain
  • 00:39:40
    so once you model The Strain you can
  • 00:39:42
    then quantify The Strain using this
  • 00:39:45
    equation here
  • 00:39:49
    where rather than just using Y minus y i
  • 00:39:52
    you just put the l y value there
  • 00:39:55
    so we can model those things or perhaps
  • 00:39:57
    you know the beam wasn't perfectly
  • 00:39:59
    aligned so let's just go through each
  • 00:40:02
    one in turn
  • 00:40:05
    and try and see if things improve or if
  • 00:40:07
    they don't
  • 00:40:08
    so there wasn't much strain
  • 00:40:12
    the beam was probably well aligned
  • 00:40:19
    and there's not a huge amount of
  • 00:40:21
    asymmetry
  • 00:40:24
    you just go through those iterations and
  • 00:40:26
    it slowly slowly will get better
  • 00:40:28
    okay another thing you can do is you can
  • 00:40:31
    model
  • 00:40:32
    your Baseline okay you can set it to
  • 00:40:35
    relax a bit and move so that piece can
  • 00:40:37
    fit better
  • 00:40:38
    and that's advisable because I've drawn
  • 00:40:40
    a line by eye I don't know what exactly
  • 00:40:42
    is the contribution from the amorphous
  • 00:40:44
    silica background
  • 00:40:45
    so it's surprisable to to put a high
  • 00:40:49
    very high damping factor and to refine
  • 00:40:52
    your background very slowly
  • 00:41:01
    I'll do that whilst refining my face
  • 00:41:04
    and my Crystal Light size
  • 00:41:11
    1.802
  • 00:41:15
    1.795 is getting slowly slowly better
  • 00:41:19
    but now it's got worse that's 1.8 and
  • 00:41:22
    that's where you call it a day for
  • 00:41:23
    refining the background it's just not
  • 00:41:25
    going to get better
  • 00:41:28
    but as you can see it's moved the
  • 00:41:30
    background a little bit shifted it to
  • 00:41:32
    allow it
  • 00:41:34
    to try and improve things
  • 00:41:36
    what you can sometimes do as well is for
  • 00:41:38
    your histogram because there's no data
  • 00:41:40
    here
  • 00:41:41
    there's no point in trying to improve
  • 00:41:43
    things so you can always
  • 00:41:46
    you can set data limits in exclude
  • 00:41:48
    regions okay so you can
  • 00:42:17
    [Music]
  • 00:42:22
    let's see if that actually works I don't
  • 00:42:24
    know if it works
  • 00:42:30
    oh it has worked okay that is fine okay
  • 00:42:36
    so now we start to model things again
  • 00:42:40
    so we'll have to edit this background
  • 00:42:50
    [Music]
  • 00:42:55
    um
  • 00:42:55
    [Music]
  • 00:42:57
    Powerpuff John is twice remember we're
  • 00:43:00
    in the bail mode
  • 00:43:07
    and now we can start to just try and
  • 00:43:10
    improve things again
  • 00:43:25
    just increase the number here so
  • 00:43:33
    the live plots he has doing and then I
  • 00:43:35
    can try and find that background again
  • 00:43:36
    to make things better but I'd say that's
  • 00:43:39
    pretty darn good I mean for a thin film
  • 00:43:41
    it's not bad at all
  • 00:43:44
    so then again as before you can go to
  • 00:43:47
    list View
  • 00:43:48
    get your errors out
  • 00:43:50
    to 0.18 that's respectable
  • 00:43:54
    and
  • 00:43:55
    you can then extract
  • 00:43:57
    the loudest parameters
  • 00:43:59
    and that's pretty much all you can get
  • 00:44:00
    from this information
  • 00:44:02
    from what we've done one thing you can
  • 00:44:05
    do if you care to quantify
  • 00:44:08
    you know the texturing they call it the
  • 00:44:10
    preferred orientation
  • 00:44:11
    you can go to results reflist
  • 00:44:20
    oh we want to look at histogram one
  • 00:44:23
    and
  • 00:44:27
    press s for all of the data
  • 00:44:32
    you just press enter and enter Etc now
  • 00:44:35
    this data here
  • 00:44:38
    it will give you
  • 00:44:40
    a couple K Alpha One and copper KL for
  • 00:44:42
    two contributions so for the 101 Peak
  • 00:44:45
    from care for one care for two
  • 00:44:47
    contributions we've got compute
  • 00:44:48
    contribution of 14.5 and 6.7 for the 103
  • 00:44:52
    we've got the 1.05 and 6.687 so you can
  • 00:44:56
    take these numbers add them and that
  • 00:44:59
    will give you the relative intensities
  • 00:45:00
    of each Peak
  • 00:45:02
    okay and then you can quantify it
  • 00:45:04
    relative to each other or relative to
  • 00:45:07
    changes versus a single Crystal
  • 00:45:12
    okay and that's that example done
  • 00:45:14
    oh
  • 00:45:21
    now we're going to look at the exact
  • 00:45:22
    same film again but this time what I'm
  • 00:45:25
    doing in the barrel mode we've been
  • 00:45:26
    doing refill mode and we're going to use
  • 00:45:27
    the spherical harmonic model
  • 00:45:32
    which I I do not claim to be any kind of
  • 00:45:34
    experts they're using
  • 00:45:37
    but it's
  • 00:45:39
    it's a way of
  • 00:45:41
    quantifying or semic quantifying the
  • 00:45:43
    amount of preferred orientation you have
  • 00:45:49
    so again I'm just going through the
  • 00:45:51
    motions now
  • 00:45:54
    have to add the face
  • 00:45:56
    you guys probably getting bored of this
  • 00:45:58
    but
  • 00:46:04
    practice makes perfect I'm going to add
  • 00:46:06
    our histogram
  • 00:46:09
    and I can't stress that anymore I mean
  • 00:46:10
    to use this software
  • 00:46:13
    practice is essential otherwise you'll
  • 00:46:15
    forget how to do these little things
  • 00:46:21
    and you might find much better ways of
  • 00:46:24
    doing it than I do
  • 00:46:25
    probably will
  • 00:46:27
    so um
  • 00:46:32
    read the manual
  • 00:46:34
    and explore things
  • 00:46:36
    so the manual is available online if you
  • 00:46:38
    just type in gsas manual
  • 00:46:43
    but it's quite detailed and quite in
  • 00:46:45
    depth and you wouldn't get the kind of
  • 00:46:47
    information the basic way of doing
  • 00:46:49
    things I'm telling you but if you want
  • 00:46:51
    to look for specific things
  • 00:46:53
    you can look here
  • 00:46:54
    under the in the uh index section
  • 00:46:59
    um
  • 00:47:03
    okay so this spherical harmonics thing
  • 00:47:05
    that I'm about to tell you is talk
  • 00:47:07
    spoken about in the gsas technical
  • 00:47:09
    manual and they refer to a paper here
  • 00:47:12
    which if you're going to use that
  • 00:47:14
    spherical harmonic model you should
  • 00:47:16
    references paper and read it and try and
  • 00:47:19
    get a grasp of
  • 00:47:22
    what it truly means
  • 00:47:26
    now I'm just gonna
  • 00:47:28
    fit my background as I did before but
  • 00:47:30
    what I think I'll do first is delete the
  • 00:47:32
    background so
  • 00:47:38
    I'm just gonna actually exclude this
  • 00:47:40
    data here
  • 00:48:03
    see if it's worked it seems to work is
  • 00:48:06
    fine
  • 00:48:07
    um
  • 00:48:09
    now I'm going to start to model things
  • 00:48:12
    first and then I'm going to show you how
  • 00:48:14
    to use the spherical harmonic model now
  • 00:48:16
    there is the marsh dialysis model as
  • 00:48:18
    well but
  • 00:48:20
    um I don't like it because
  • 00:48:23
    um
  • 00:48:23
    it's only good if you know for a fact
  • 00:48:26
    which plane
  • 00:48:29
    your material was growing so sometimes
  • 00:48:31
    if you have epitaxial growth yeah I'm
  • 00:48:33
    not sure many of us deal with epitaxial
  • 00:48:35
    growth so we're doing sort of chemical
  • 00:48:38
    depositions where it gives it that
  • 00:48:40
    flexibility to grow in many different
  • 00:48:42
    directions but whereas in other
  • 00:48:45
    materials growth for instance physical
  • 00:48:47
    Vapor depositions where you're growing
  • 00:48:48
    on a lattice matched substrate
  • 00:48:51
    you often get growth in one particular
  • 00:48:53
    direction and the march to the latest
  • 00:48:55
    function is very good so if you know
  • 00:48:57
    that it's growing State just in the 101
  • 00:48:59
    Direction you can specify one zero one
  • 00:49:02
    and give it that ratio here
  • 00:49:05
    to grow or not grow in that in that
  • 00:49:07
    particular plane and the other planes
  • 00:49:09
    will obviously go down in growth
  • 00:49:12
    subsequently but what I found was films
  • 00:49:16
    the symptoms I've made is that once you
  • 00:49:18
    enter one plane it doesn't often work
  • 00:49:21
    for one plane and then you're into two
  • 00:49:23
    or three planes to make it work
  • 00:49:25
    and you're just creating
  • 00:49:26
    whatever I mean you're opening Pandora's
  • 00:49:29
    Box because there's so many different
  • 00:49:30
    possible combinations that could work
  • 00:49:33
    so rather than doing it manually it's
  • 00:49:35
    better to just
  • 00:49:36
    let the spherical harmonic model do that
  • 00:49:39
    for you in many different planes and try
  • 00:49:41
    and get a quantitative analysis in many
  • 00:49:42
    different things
  • 00:49:47
    foreign
  • 00:49:51
    [Music]
  • 00:50:02
    just modeling the lattice parameters
  • 00:50:07
    and the crystallite width
  • 00:50:10
    it doesn't seem to get once again much
  • 00:50:12
    better because of
  • 00:50:14
    the problem with your prefer orientation
  • 00:50:16
    here you can see there's a big problem
  • 00:50:18
    with preferred orientation
  • 00:50:21
    also my background sucks so I need to
  • 00:50:23
    fix that
  • 00:50:28
    so whenever I uncheck something first I
  • 00:50:31
    press Powerpuff and joiners just so it
  • 00:50:33
    registers the change
  • 00:50:34
    then now I can fix my background because
  • 00:50:37
    that was a big problem as well for the
  • 00:50:39
    model because that background really
  • 00:50:40
    doesn't fit
  • 00:50:43
    [Music]
  • 00:50:45
    that's good job
  • 00:50:49
    refining the original one
  • 00:50:53
    it's it's similar but I like to give it
  • 00:50:58
    um
  • 00:50:59
    I'd like to start off with what my eye
  • 00:51:01
    looks good
  • 00:51:03
    and then let it move because sometimes
  • 00:51:06
    it just moves into a completely long
  • 00:51:07
    face because it doesn't know what to do
  • 00:51:10
    so I like to give it a good starting
  • 00:51:11
    point
  • 00:51:13
    so it again it's personal choice I mean
  • 00:51:15
    you could you can just go ahead and
  • 00:51:17
    press refine background without manually
  • 00:51:19
    editing it and just see how that goes
  • 00:51:21
    and plays out for you but I like to
  • 00:51:24
    do it myself
  • 00:51:26
    so now I'm going to start working the
  • 00:51:28
    spherical harmonic it's really easy to
  • 00:51:31
    use you just tell it the order
  • 00:51:34
    so I found that six or eight works
  • 00:51:37
    really well but you can you can try
  • 00:51:38
    whatever works for you and the idea is
  • 00:51:41
    to get the lowest order to work for you
  • 00:51:43
    okay but whatever you do if you're
  • 00:51:46
    analyzing a series of samples
  • 00:51:49
    um be consistent
  • 00:51:51
    so I'm going to go to eight sorry eight
  • 00:51:55
    I'm just going to try and push growth
  • 00:52:00
    in these different axes
  • 00:52:09
    I'm going to give it a high damping
  • 00:52:13
    foreign
  • 00:52:29
    things must be improving
  • 00:52:32
    and they are
  • 00:52:34
    if you look at my spherical harmonic
  • 00:52:36
    numbers they've all changed and it gives
  • 00:52:39
    you something called a texture index
  • 00:52:41
    now how it gets that texture index I'm
  • 00:52:43
    not sure I haven't read the paper I need
  • 00:52:45
    to
  • 00:52:46
    but it's something that we need to
  • 00:52:47
    understand that gives us a semi
  • 00:52:50
    quantitative parameter or how textured
  • 00:52:53
    your material is
  • 00:52:58
    so again you just continue to refine the
  • 00:53:01
    preferred orientation I've unchecked it
  • 00:53:03
    for now because I want to start to
  • 00:53:05
    refine my background
  • 00:53:12
    just a tad
  • 00:53:21
    I think that's enough maybe just a
  • 00:53:23
    little bit more
  • 00:53:24
    make sure that
  • 00:53:29
    1.329
  • 00:53:30
    [Music]
  • 00:53:33
    it's improving about the steadily
  • 00:53:42
    and again you just go through many
  • 00:53:44
    iterations of this
  • 00:53:56
    until you reach
  • 00:53:58
    a true minimum
  • 00:54:00
    and that's not bad I mean that's very
  • 00:54:03
    good actually
  • 00:54:06
    in terms of the this is the blue is the
  • 00:54:08
    air is the measurement of the difference
  • 00:54:10
    between your model and the data and it's
  • 00:54:12
    very flat so that's great that's a good
  • 00:54:15
    fit
  • 00:54:16
    okay this number here I mean these
  • 00:54:20
    numbers you can note them down
  • 00:54:23
    um but this texture index is just going
  • 00:54:25
    to give us a semi-constitutive parameter
  • 00:54:27
    that we can then compare with another
  • 00:54:28
    film that I've made
  • 00:54:31
    and that'll be the last example today so
  • 00:54:33
    it's just basically going to be a
  • 00:54:35
    repetition of what I've just done
  • 00:54:38
    talk to another sample
  • 00:54:41
    and the difference between these samples
  • 00:54:43
    is so simple
  • 00:54:46
    one sample
  • 00:54:47
    are made by chemical vertical deposition
  • 00:54:50
    using a set parameters
  • 00:54:53
    and now my deposition time is two
  • 00:54:55
    minutes
  • 00:54:57
    now
  • 00:54:59
    the second material
  • 00:55:01
    I did exactly the same thing
  • 00:55:03
    but my deposition time was for four
  • 00:55:05
    minutes
  • 00:55:07
    but there's a very big difference in
  • 00:55:09
    looking forward orientation
  • 00:55:12
    and you'll see that very soon
  • 00:55:21
    see huge preferred orientation in this
  • 00:55:24
    film there's absolutely no growth in the
  • 00:55:26
    one zero one plane under these
  • 00:55:27
    conditions for some strange reason I
  • 00:55:30
    don't know why but that that is the case
  • 00:55:31
    that was the case
  • 00:55:33
    so
  • 00:55:35
    I'm just going to
  • 00:55:36
    make sure
  • 00:55:39
    everything's okay
  • 00:55:47
    [Music]
  • 00:56:00
    foreign
  • 00:56:12
    I forgot to model my background but it
  • 00:56:15
    was pretty flat anyway
  • 00:56:17
    so I'm not sure if that's a problem
  • 00:56:21
    let me just double check my background
  • 00:56:23
    is okay
  • 00:56:29
    okay
  • 00:56:53
    because there's so much preferred
  • 00:56:55
    orientation in this form
  • 00:56:57
    the model got really confused at first
  • 00:56:59
    but now I think it's it's finally found
  • 00:57:01
    its way
  • 00:57:07
    and there we are
  • 00:57:09
    it's got an almost perfect fit as well
  • 00:57:12
    don't forget that you can export this
  • 00:57:14
    data back on some file explorer CSV file
  • 00:57:19
    but there's there's massive preferred
  • 00:57:21
    growth I think this is the 211 plan
  • 00:57:24
    and if we look at the referred or
  • 00:57:26
    notation texture index now it's 470 I
  • 00:57:29
    mean there's a huge difference in that
  • 00:57:31
    texture index
  • 00:57:32
    so it had to really really flush things
  • 00:57:34
    to make it work
  • 00:57:36
    um so this texture index gives you an
  • 00:57:38
    idea
  • 00:57:39
    of the period orientation in your form
  • 00:57:44
    um
  • 00:57:45
    and that's how you do that
Etiquetas
  • GSAS
  • X3
  • X-straal diffraksie
  • poeierpatrone
  • Reitveld metode
  • struktuur modifikasie
  • fase mengsels
  • dunfilms
  • sferiese harmoniese model
  • databasis
  • gepreferensieerde oriëntasie