Changeset 5039
- Timestamp:
- Sep 28, 2021 10:55:35 AM (8 months ago)
- File:
-
- 1 edited
-
Tutorials/SeqRefine/SequentialTutorial.htm (modified) (45 diffs)
Legend:
- Unmodified
- Added
- Removed
-
Tutorials/SeqRefine/SequentialTutorial.htm
r4457 r5039 1394 1394 </BL></P><P></P> 1395 1395 1396 <p class=MsoNormal>Sequential refinement is a way to fit a series of datasets 1397 with a set of closely related models that evolve over the course of the 1398 measurements.<span style='mso-spacerun:yes'> </span>Sequential refinements can 1396 <p class=MsoNormal>Sequential refinement is a way to fit a series of 1397 datasets collected under similar conditions with a set of closely 1398 related models. Each dataset receives an independent 1399 fit. Thus it is possible to track how the models 1400 evolve over the course of the measurements. One or more phases may 1401 be used to fit all all the datasets or some phases may appear in 1402 just some of the measurements, should there be phase transitions. 1403 <P></P> 1404 Sequential refinements can 1399 1405 be performed in three different ways: </p> 1400 1401 <p class=MsoListParagraphCxSpFirst style='margin-left:.75in;mso-add-space:auto; 1402 text-indent:-.25in;mso-list:l4 level1 lfo2'><![if !supportLists]><span 1403 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 1404 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>with 1405 differing starting parameters for each histogram</p> 1406 1407 <p class=MsoListParagraphCxSpMiddle style='margin-left:.75in;mso-add-space: 1408 auto;text-indent:-.25in;mso-list:l4 level1 lfo2'><![if !supportLists]><span 1409 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 1410 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>with 1411 the same starting parameters for each histogram</p> 1412 1413 <p class=MsoListParagraphCxSpLast style='margin-left:.75in;mso-add-space:auto; 1414 text-indent:-.25in;mso-list:l4 level1 lfo2'><![if !supportLists]><span 1415 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 1416 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>where 1417 the resulting parameters after refinement with a histogram are used as the 1418 starting point for the next histogram.</p> 1406 <OL> 1407 <LI>where the resulting parameters after a fit for a 1408 histogram are used as the starting point for the next histogram.</LI> 1409 <LI>with differing starting parameters for each histogram</LI> 1410 <LI>with the same starting parameters for each histogram</LI> 1411 </OL><P></P> 1412 Most commonly the first mode is used initially, where the parameters 1413 for each fir are copied as the starting point for the fit of next 1414 dataset. After this initial refinement of all datasets is complete, 1415 the refinement is continued where each dataset has its own 1416 parameters. 1419 1417 1420 1418 <p class=MsoNormal><o:p> </o:p></p> 1421 1419 1422 <p class=MsoNormal> 1420 <p class=MsoNormal> 1421 In this example, we will fit a temperature series where both the unit 1422 cell and atomic parameters evolve with temperature. Note that 1423 it is also possible to fit a temperature series as a combined refinement and, 1424 as will be seen, one can still deal with thermal expansion (using the 1425 Hydrostatic/Elastic strain terms), but a combined refinement would 1426 allow only a single set of (averaged) atomic parameters for 1427 all temperatures. A sequential refinement allows each 1428 data set to determine the atomic parameters and allows their variation 1429 with temperature to be determined. 1430 <P></P> 1423 1431 Data for this example will be downloaded with the tutorial, if you used 1424 1432 that option to see this page. Otherwise, download file 1425 1433 <A href="https://subversion.xray.aps.anl.gov/pyGSAS/Tutorials/SeqRefine/data/"> 1426 1434 SeqTut.zip from here</A>. 1427 In this example, we will fit a temperature series. Note that 1428 it is also possible to fit a temperature series as a combined refinement and, 1429 as will be seen, one can still deal with thermal expansion (using the 1430 Hydrostatic/Elastic strain terms), but a combined refinement would average the 1431 atomic parameters across all temperatures. A sequential refinement allows each 1432 data set to determine the atomic parameters; their variation with temperature 1433 might be of interest.</p> 1434 1435 <p class=MsoNormal>This example assumes that one is already familiar with use 1435 This example assumes that one is already familiar with use 1436 1436 of GSAS-II for simple refinements. <span style='mso-bidi-font-size:12.0pt; 1437 1437 mso-fareast-font-family:"Times New Roman";mso-fareast-language:EN-US'>Note that 1438 menu entries are listed in bold face below as </span><b style='mso-bidi-font-weight:1438 menu entries are listed in bold face below, such as </span><b style='mso-bidi-font-weight: 1439 1439 normal'><span style='mso-bidi-font-size:12.0pt;font-family:"Calibri",sans-serif; 1440 1440 mso-fareast-font-family:"Times New Roman";mso-fareast-language:EN-US'>Help/About … … 1449 1449 12.0pt;font-family:"Calibri",sans-serif;mso-fareast-font-family:"Times New Roman"; 1450 1450 mso-fareast-language:EN-US'>About GSAS-II</span></b><span style='mso-bidi-font-size: 1451 12.0pt;mso-fareast-font-family:"Times New Roman";mso-fareast-language:EN-US'>).</span>If 1452 you have not done so already, start GSAS-II. </p> 1451 12.0pt;mso-fareast-font-family:"Times New 1452 Roman";mso-fareast-language:EN-US'>).</span> 1453 </p> 1454 1455 <h1>Step I: Fit a single temperature. </h1> 1456 <p></p> 1457 1458 If you have not done so already, start GSAS-II. While it is possible 1459 to start a sequential refinement reading in all phases and datasets 1460 from the beginning, usually it is best to know something about how the 1461 fit will be performed and this usually means performing one or more 1462 fits to representative individual datasets as the conditions or phases 1463 present change. One way to do this could be to read in all datasets but select 1464 only a single individual dataset for that fit until we are ready to 1465 fit the entire sequence. Here we will start by reading in only the 1466 first dataset and fit that. Then we will add more datasets and perform 1467 a sequential fit. 1453 1468 1454 1469 <p class=MsoNormal><o:p> </o:p></p> 1455 1470 1456 <h1>Step 1: Fit a single temperature. </h1> 1457 1458 <p class=MsoNormal><o:p> </o:p></p> 1459 1460 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1461 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 1462 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Use 1471 <H3>1. Read phase files <B>CuCr2O4.cif</B> and <B>CuO.cif</B> from file <b>SeqTut.zip</B>.</H3> 1472 1473 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 1474 Use 1463 1475 the <b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1464 mso-hansi-theme-font:major-latin'>Import/Phase/from CIF file menu item</span></b> 1465 to read the phase information </p> 1466 1467 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1468 1469 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1470 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 1471 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>In 1476 mso-hansi-theme-font:major-latin'> 1477 Import/Phase/from CIF file menu item</span></b> to read the phase information.</p> 1478 <P></P> 1479 In 1472 1480 the file browser dialog window that is created, <span style='mso-bidi-font-weight: 1473 1481 bold'>change the file type from<b> </b></span><b><span style='font-family: … … 1496 1504 bold'>input files. Press<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1497 1505 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>OK</span></b>. 1498 </p> 1499 1500 <p class=MsoListParagraphCxSpMiddle>(Note that if files CuCr2O4.cif or the 1501 CuO.cif already exist, you will be asked to OK overwriting them.)</p> 1502 1503 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1504 1505 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1506 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 1507 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>You 1506 <P></p> 1507 (If this exercise is being repeated, files CuCr2O4.cif or the 1508 CuO.cif will already exist. If so, you will be asked to OK overwriting them.) 1509 1510 <P></p> 1511 <LI>You 1508 1512 will be asked sequentially for the name for each phase. The defaults (<b 1509 1513 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; … … 1516 1520 style='mso-spacerun:yes'> </span></p> 1517 1521 1522 <P></P> 1523 <LI>Note that for CuCr<sub>2</sub>O<sub>4</sub>, the space group (<I>Fddd</I>) can 1524 be set with two origin choices. Select <B>Keep current coordinates</B> 1525 since the coordinates are already set properly. 1526 <P></P> 1527 <I>Discussion: 1528 When there is a choice, GSAS-II requires the use of Origin 2, which 1529 places the origin at a center of symmetry, and this CIF has been set 1530 that way 1531 (the presence of a -x,-y,-z symmetry operation confirms that), but GSAS-II 1532 notes the possible choice and offers to apply the shift to the 1533 coordinates that would shift from Origin 1 to Origin 2. 1534 As shown in the window where this 1535 choice would be made, this shift were applied, 1536 the unit cell contents would be 1537 Cu<sub>16</sub>Cr<sub>8</sub>O<sub>32</sub> rather than the correct 1538 CuCr<sub>2</sub>O<sub>4</sub> stoichiometry, so this also confirms the 1539 origin setting is correct and no coordinate shift should be applied. 1540 <B>Choice of origin is a common problem in crystallographic 1541 computations and many CIFs do not define this setting clearly, so use care with this.</B> 1542 </I> 1543 1518 1544 <p class=MsoNormal><o:p> </o:p></p> 1519 1520 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1521 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>4.<span 1522 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 1523 style='mso-spacerun:yes'> </span>Read in the ~7 K dataset (<b style='mso-bidi-font-weight: 1545 </OL> 1546 1547 <H3>2. 1548 Read in file <B>OH_00.fxye</B> from file <b>SeqTut.zip</B> and <B>map 1549 both phases</B> to this dataset. </H3> 1550 1551 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 1552 Read in the ~7 K dataset (<b style='mso-bidi-font-weight: 1524 1553 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1525 1554 major-latin;mso-hansi-theme-font:major-latin'>OH_00.fxye</span></b>) using <b><span … … 1566 1595 src="SequentialTutorial_files/image001.png" v:shapes="Picture_x0020_54"><![endif]></span></p> 1567 1596 1568 <p class=MsoListParagraphCxSpMiddle>You will be shown the first few lines from 1597 <P></P> 1598 <LI>You will be shown the first few lines from 1569 1599 the file. Confirm that the selected file is correct by <span style='mso-bidi-font-weight: 1570 1600 bold'>pressing<b> </b></span><b><span style='font-family:"Calibri",sans-serif; … … 1579 1609 </v:shape><![endif]--><![if !vml]><img width=366 height=233 1580 1610 src="SequentialTutorial_files/image002.png" v:shapes="Picture_x0020_55"><![endif]></span></p> 1581 1611 <P></P> 1582 1612 <p class=MsoListParagraphCxSpMiddle>Note that the instrument parameter file 1583 1613 with the matching name (<b style='mso-bidi-font-weight:normal'><span … … 1587 1617 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1588 1618 1589 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1590 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>5.<span 1591 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Map 1619 <A name="I.2.C"> 1620 <LI>Map 1592 1621 the phases to the data: After reading in the dataset, a window used to select 1593 1622 phases is shown.</p> … … 1601 1630 src="SequentialTutorial_files/image003.png" v:shapes="Picture_x0020_56"><![endif]></span></p> 1602 1631 1632 <P></P> 1603 1633 <p class=MsoListParagraphCxSpMiddle><span style='mso-bidi-font-weight:bold'>Select 1604 1634 both phases and press<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1605 1635 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>OK</span>.</b><span 1606 1636 style='mso-spacerun:yes'> </span></p> 1637 </OL> 1607 1638 1608 1639 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1609 1640 1610 < p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>1611 1612 < p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span1613 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>6.<span 1614 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>We1641 <A name="I.3"></A> 1642 <H3>3. Prepare for fit: define <B>phase fraction constraint</B></H3> 1643 <P></P> 1644 <p class=MsoListParagraphCxSpMiddle> 1645 We 1615 1646 want to fit the lattice parameters, background, and a phase fraction 1616 initially. <span style='mso-spacerun:yes'> </span>To illustrate how constraints1647 initially. To illustrate how constraints 1617 1648 are used, we will refine both phase fractions, but constrain their sum to be 1618 one. To set up the constraint, <span style='mso-bidi-font-weight:bold'>click on 1649 one. <I>Discussion: a single powder diffraction pattern provides 1650 enough information to determine an overall scale factor and the 1651 relative amounts of each phase (with two phases: two observables), but 1652 with the scale factor and two phase fractions being varied 1653 independently, we have three arbitrary 1654 parameters, which would result in a very unstable refinement where the 1655 GSAS-II minimizer would likely remove a degree of freedom. By defining 1656 a constraint that sets the sum of the two phase fractions, we reduce 1657 this to an appropriate two parameter problem. The error of refining the scale 1658 factor and all phase fractions is made often enough that GSAS-II will 1659 warn about this and will offer to define this constraint.</I> 1660 1661 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 1662 To set up the phase fraction constraint, <span style='mso-bidi-font-weight:bold'>click on 1619 1663 the<b> </b></span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1620 1664 major-latin;mso-hansi-theme-font:major-latin'>Constraints</span> </b><span … … 1629 1673 style='mso-bidi-font-weight:bold'> menu item<b>.</b></span></p> 1630 1674 1631 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span></p> 1632 1633 <p class=MsoListParagraphCxSpMiddle>In the next window use a <b 1675 1676 <LI>In the next window use a <b 1634 1677 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1635 1678 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Filter</span></b> … … 1653 1696 style='mso-spacerun:yes'> </span></p> 1654 1697 1655 <p class=MsoListParagraphCxSpMiddle>In the next window, either <span 1698 <P></P> 1699 1700 <LI>After pressing "OK" a new window 1701 appears; in that next window either 1702 <span 1656 1703 style='mso-bidi-font-weight:bold'>select the other phase fraction<b> (</b></span><b><span 1657 1704 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; … … 1662 1709 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1663 1710 mso-hansi-theme-font:major-latin'>all:0:Scale</span></b><span style='font-family: 1664 "Cambria",serif;mso-ascii-theme-font:minor-latin;mso-hansi-theme-font:minor-latin'>)<o:p></o:p></span></p> 1665 1711 "Cambria",serif;mso-ascii-theme-font:minor-latin;mso-hansi-theme-font:minor-latin'>). The same result is obtained since there is only one additional 1712 phase. 1713 1714 <P></P> 1666 1715 <p class=MsoListParagraphCxSpMiddle><span style='font-family:"Cambria",serif; 1667 1716 mso-fareast-language:EN-US;mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_58" … … 1673 1722 style='font-family:"Cambria",serif'><o:p></o:p></span></p> 1674 1723 1675 <p class=MsoListParagraphCxSpMiddle>This creates the desired constraint, which 1676 forces the sum to be 1. (If the desired sum was different or different 1677 multipliers were desired, press the <b style='mso-bidi-font-weight:normal'><span 1724 <P></P> 1725 <p class=MsoListParagraphCxSpMiddle> 1726 <LI>Press OK and this creates the desired constraint, which 1727 forces the sum of phase fractions to be 1. (This is not needed here, 1728 but to change the sum and/or multipliers, press the constraint's <b style='mso-bidi-font-weight:normal'><span 1678 1729 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1679 mso-hansi-theme-font:major-latin'>Edit</span></b> button for the constraint.) </p>1730 mso-hansi-theme-font:major-latin'>Edit</span></b> button.) </p> 1680 1731 1681 1732 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 1689 1740 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1690 1741 1691 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1692 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>7.<span 1693 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Now 1694 we select the variables to refine. </p> 1695 1696 <p class=MsoNormal><o:p> </o:p></p> 1697 1698 <p class=MsoListParagraphCxSpFirst><span style='mso-bidi-font-weight:bold'>Select 1742 </OL> 1743 1744 <H3>4. Initial fit: optimize <B>lattice</B> parameters & <B>phase 1745 fractions</B>; 1746 <B>background</B> (with <B>6 terms</B>); <B>scale factor</B>. 1747 </H3> 1748 <P></P> 1749 <p class=MsoListParagraphCxSpMiddle> 1750 Now we select the variables to refine for each phase and then the 1751 histogram parameters. </p> 1752 1753 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 1754 Select 1699 1755 the first phase</span> (<b style='mso-bidi-font-weight:normal'><span 1700 1756 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; … … 1705 1761 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1706 1762 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>General</span></b> 1707 tab). ( On a Mac you will likelyneed to click on the triangle to the left of1763 tab). (You might need to click on the triangle to the left of 1708 1764 the <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1709 1765 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Phase</span></b> … … 1726 1782 src="SequentialTutorial_files/image064.png" v:shapes="Picture_x0020_60"><![endif]></span></p> 1727 1783 1728 <p class=MsoListParagraphCxSpMiddle>Repeat this with the second phase (<b 1784 <p></P> 1785 <LI>Repeat this with the second phase (<b 1729 1786 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1730 1787 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>CuO</span></b>), … … 1743 1800 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Phase 1744 1801 fraction</span> </b><span style='mso-bidi-font-weight:bold'>box</span> to cause 1745 that to be refined.<span style='mso-spacerun:yes'> </span>Also <span 1746 style='mso-bidi-font-weight:bold'>click on the single<b> </b></span><b><span 1747 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1748 mso-hansi-theme-font:major-latin'>microstrain</span></b> (isotropic) term for 1749 this second phase.</p> 1750 1751 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1752 1753 <p class=MsoListParagraphCxSpMiddle>Also <span style='mso-bidi-font-weight: 1754 bold'>select<b> </b>the<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1802 that to be refined. 1803 1804 <P></P> 1805 <LI>To refine the background,<span style='mso-bidi-font-weight: 1806 bold'> 1807 select the </span><b><span style='font-family:"Calibri",sans-serif; 1755 1808 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Background</span> 1756 </b><span style='mso-bidi-font-weight:bold'>entry</span> under <b 1809 </b><span style='mso-bidi-font-weight:bold' entry</span> 1810 entry under the <b 1757 1811 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1758 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>PWDR 1759 OH_00.fxye Bank 1</span></b> in the GSAS-II data tree (for a Mac, you will 1760 likely need to click on the triangle to the left of the entry to open the 1761 tree item.) Confirm that <b style='mso-bidi-font-weight:normal'><span 1812 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'> 1813 PWDR 1814 OH_00.fxye Bank 1</span></b> histogram sub-entries in the GSAS-II data tree. Confirm that <b style='mso-bidi-font-weight:normal'><span 1762 1815 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1763 1816 mso-hansi-theme-font:major-latin'>Background function</span></b> should already 1764 1817 be set as <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1765 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>chebyshev </span></b>1818 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>chebyshev-1</span></b> 1766 1819 and the <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1767 1820 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Refine</span></b> … … 1773 1826 to introduce more background terms. </p> 1774 1827 1775 < p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>1776 1777 <p class=MsoListParagraphCxSpMiddle>The histogram scale factor,on the <b1828 <P></P> 1829 <LI>The overall histogram scale factor should already be set to be 1830 refined by default, but you may want to confirm this by clicking on the <b 1778 1831 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1779 1832 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Sample 1780 Parameters</span></b> entry should already be set as refined by default, but you1781 may want to confirm this.</p>1833 Parameters</span></b> data tree entry. 1834 </OL> 1782 1835 1783 1836 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1784 1837 1785 <p class=MsoListParagraphCxSpMiddle>Finally, note that the first broad peak in 1786 these data is spurious. It is known to arise from the kapton window in the 1838 <H3>5.<span 1839 style='font:7.0pt "Times New Roman"'> 1840 </span></span></span><![endif]> 1841 Reduce <B>Data Range</B> to 4.5 degrees</H3> 1842 1843 <P></P> 1844 1845 <p class=MsoListParagraphCxSpMiddle>Finally, note that the first broad 1846 peak at ~1.5 degrees in these data is spurious. 1847 This artifact is known to arise from scattering from the kapton window in the 1787 1848 cryostat used for sample cooling; this was confirmed by collecting data without 1788 a sample. We could model this peak, but it is easier to simply dropthe data1849 a sample. We could model this peak, but it is easier to simply cut off the data 1789 1850 below the first peak at ~5 degrees. To do this<b style='mso-bidi-font-weight: 1790 1851 normal'>, </b><span style='mso-bidi-font-weight:bold'>select the<b> </b></span><b><span … … 1802 1863 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1803 1864 1804 < p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span1805 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>8.<span 1806 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 1807 style='mso-bidi-font-weight:bold'>Use the </span><b><span style='font-family:1865 <H3>6. Perform initial fit</H3> 1866 1867 <p class=MsoListParagraphCxSpMiddle> 1868 Use the<b><span style='font-family: 1808 1869 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 1809 major-latin'>Calculate/Refine</span></b><span style='mso-bidi-font-weight:bold'> 1870 major-latin'> 1871 Calculate/Refine</span></b><span style='mso-bidi-font-weight:bold'> 1810 1872 main GSAS-II menu item</span> to initiate the refinement. Since the project has 1811 1873 not been saved, you will be asked to select a name (and optionally a directory) … … 1815 1877 is assumed here. The refinement progresses quickly to a Rw of ~17%. </p> 1816 1878 1817 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1818 1819 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1820 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>9.<span 1821 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>There 1822 are several additional variables that should now be added. One should always 1823 refine a sample displacement parameter (or rarely one refines a zero correction.) In this 1824 case, the best choice is to <span style='mso-bidi-font-weight:bold'>add the<b> </b></span><b><span 1825 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1826 mso-hansi-theme-font:major-latin'>Sample X displ.</span> </b><span 1827 style='mso-bidi-font-weight:bold'>parameter</span> on the <b style='mso-bidi-font-weight: 1828 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1829 major-latin;mso-hansi-theme-font:major-latin'>Sample Parameters</span></b> section 1830 of histogram <b style='mso-bidi-font-weight:normal'><span style='font-family: 1831 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 1832 major-latin'>PWDR OH_00.fxye Bank 1</span></b>. (NB: a data range >>90 1833 degrees is needed before the Sample Y displ. term can be included). Set the <b 1834 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1835 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font: 1836 major-latin'>Goniometer radius</span></b> to <b style='mso-bidi-font-weight: 1837 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1838 major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>1000</span></b>. 1839 Note that since this instrument is very well characterized, there is no need to 1840 refine any of the instrumental terms. </p> 1841 1842 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1843 1844 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 1845 mso-no-proof:yes'>Selecting</span><b> </b><span style='mso-bidi-font-weight: 1846 bold'>the<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1847 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>PWDR 1848 OH_00.fxye Bank 1</span> </b><span style='mso-bidi-font-weight:bold'>entry</span> 1849 shows the Rietveld plot and zooming in on different regions of the pattern 1850 provides insight on how the fit could be improved. You can alter the view to 1851 show the square root of the intensity (use the <b style='mso-bidi-font-weight: 1852 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1853 major-latin;mso-hansi-theme-font:major-latin'>S</span></b> key to toggle 1854 this). The background is not well fit in the low angle range and examination 1855 shows a hump of low-angle scattering from ~4.5 to 6 degrees. This is more 1856 easily fit by adding a background peak than extra terms to the background 1857 function. <span style='mso-bidi-font-weight:bold'>Select the<b> </b></span><b><span 1858 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1859 mso-hansi-theme-font:major-latin'>Background </span></b><span style='mso-bidi-font-weight: 1860 bold'>item</span> in histogram <b style='mso-bidi-font-weight:normal'><span 1861 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1862 mso-hansi-theme-font:major-latin'>PWDR OH_00.fxye Bank 1</span></b> and <span 1863 style='mso-bidi-font-weight:bold'>change </span><b><span style='font-family: 1864 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 1865 major-latin'>Peaks in Background</span> </b><span style='mso-bidi-font-weight: 1866 bold'>to<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1867 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>1</span>.</b><span 1868 style='mso-spacerun:yes'> </span><span style='mso-bidi-font-weight:bold'>Change 1869 the<b> </b>peak position</span> (<b style='mso-bidi-font-weight:normal'><span 1870 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1871 mso-hansi-theme-font:major-latin'>pos</span></b>) to the approximate peak 1872 center (<b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1873 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>5.25</span></b> 1874 degrees) and <span style='mso-bidi-font-weight:bold'>set the intensity (</span><b><span 1875 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1876 mso-hansi-theme-font:major-latin'>int</span></b><span style='mso-bidi-font-weight: 1877 bold'>) and Gaussian width<b> </b>(</span><b><span style='font-family:"Calibri",sans-serif; 1878 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>sig</span></b><span 1879 style='mso-bidi-font-weight:bold'>) to be refined</span>.<span 1880 style='mso-spacerun:yes'> </span><b style='mso-bidi-font-weight:normal'><i 1881 style='mso-bidi-font-style:normal'><u>Important</u></i></b>: note that one must 1882 click on another box to ensure that input is recorded.</p> 1883 1879 <h1>Step II: Improve the single-temperature fit. </h1> 1880 1881 <P></P> 1882 There are several additional variables that should now be added, now 1883 that we have reasonable starting unit cell parameters, etc. This 1884 includes a sample displacement parameter, a better background treatment 1885 and some sample broadening terms for the profile. 1886 1887 <H3>1. Improve sample-dependent fitting.</H3> 1888 1889 <OL style='PADDING-LEFT:0.75in' type="A"> 1890 <LI>Refine the <B>Sample X displacement</B> term</LI> 1891 1892 <P></P><p class=MsoNormal> 1893 Click on the <B>Sample 1894 Parameters</B> data tree subitem for <B>PWDR OH_00.fxye Bank 1</B>. 1895 Select the refine checkbox that is 1896 labeled <B>Sample X disp. perp. to peak</B>. The value for this is the 1897 displacement in microns (10<sup>-6</sup> m), provided the 1898 diffractometer radius is entered correctly. Set the Goniometer radius to 1000. 1899 </P> 1900 1901 <P></P> 1902 <p class=MsoListParagraphCxSpMiddle><I>Discussion: Diffractometers are never 1903 perfectly aligned, which means one should always an alignment 1904 parameter. While in older programs, the alignment parameter was a zero 1905 correction (except for Bragg-Brentano instruments), this does not 1906 properly correct for mispositioning of the sample from the center of 1907 the goniometer arc. GSAS-II offers a geometrical correction for that 1908 being offset perpendicular to the 1909 direction of the beam as the Sample X displacement (and for where data is 1910 collected to angles above 120 degrees 2theta, it is also possible to fit a 1911 second term, Sample Y displacement for displacement along the beam 1912 direction.) 1913 When using GSAS-II, refine the Sample X displacement parameter rather than 1914 the 2theta-zero correction. (Refinement of both is not normally 1915 possible).</I> 1916 1917 1918 <P></P><p class=MsoNormal> 1919 Note: With a well characterized 1920 instrument, as is the case here for 11-BM at the APS, none of the 1921 Instrument Parameters or other Sample Parameters (beyond 1922 "Histogram scale factor" and "Sample X displ.") need to be fit. 1923 1924 <P></P> 1925 <LI>Add a background peak.</LI> 1926 1927 <P></P><p class=MsoNormal> 1928 Add a background peak at <B>pos=5.25</B> degrees with <B>int=5.0</B> 1929 and a <B>sig=2000</B>. 1930 </P> 1931 1932 <P></P><p class=MsoNormal> 1933 The background is not very well fit even with a six term polynomial 1934 due to a kapton scattering peak in the range of 4-6 degrees 1935 2theta. This can be seen by selecting the 1936 <b>PWDR OH_00.fxye Bank 1</b> tree entry and viewing the Rietveld 1937 plot. 1938 Then to see more detail, either press "s" to see the plot in the 1939 square-root of intensities, or zoom in so the vertical intensity scale 1940 is 0-500 counts and the horizontal region covers the first ~5 degrees 1941 in the pattern. 1942 While the fit for this could be improved by adding many more 1943 background terms, a simpler way to improve this fit is to add a broad 1944 peak in the pattern via the background fitting. This only needs 2 or 3 1945 additional parameters. 1946 1947 <P></P><p class=MsoNormal> 1948 After selecting the <b>Background</b> sub-item in the 1949 <B>PWDR OH_00.fxye Bank 1</b> histogram, change the <b>Peaks in 1950 Background</b> setting from "0" to "1". Then change 1951 the peak position (<b>pos</b>) value to the approximate peak 1952 center (<b>5.25</b> degrees) but that value should not be varied until 1953 the background is fairly well fit. Since the background generated by the 1954 Chebyschev polynomial and the background peak are highly correlated, 1955 we need to get the width and intensity into the right vicinity. Note 1956 that as the values in the peaks list are changed, the background curve 1957 (in red) is recomputed. In that way the effect of the change in values 1958 can be seen. A <B>sig</B> value of 2000 gives a Gaussian FWHM for this peak of 1959 ~2 degrees. Changing the <B>int</B> value from 1 to 5 makes the peak 1960 much easier to see and also helps start the initial fit. Set both 1961 <B>sig</B> and <B>int</B> to be <B>refined</B>. 1962 1963 <P></P><p class=MsoNormal> 1964 <i><u>Important</u></i></b>: note that after clicking on a checkbox, one must 1965 click elsewhere in the window so that the input is recorded.</p> 1966 1967 <P></P> 1884 1968 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 1885 1969 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_62" o:spid="_x0000_i1052" … … 1890 1974 src="SequentialTutorial_files/image065.png" v:shapes="Picture_x0020_62"><![endif]></span></p> 1891 1975 1892 <p class=MsoListParagraphCxSpMiddle>Once the fit has been improved we can then 1893 add the position of this peak to the fit. </p> 1894 1895 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1896 1897 <p class=MsoListParagraphCxSpMiddle>While in other projects I might postpone 1898 this until later, we can also start on improving the peak fit. We will do this 1899 by varying the sample broadening for the major phase (since the instrumental 1900 terms are very good.) <span style='mso-bidi-font-weight:bold'>Select the first 1901 phase (</span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1902 major-latin;mso-hansi-theme-font:major-latin'>CuCr2O4</span></b><span 1903 style='mso-bidi-font-weight:bold'>)</span> and its <b><span style='font-family: 1904 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 1905 major-latin'>Data</span> </b><span style='mso-bidi-font-weight:bold'>tab</span>, 1906 <span style='mso-bidi-font-weight:bold'>click on the </span><b><span 1907 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1908 mso-hansi-theme-font:major-latin'>Cryst. Size</span></b><span style='mso-bidi-font-weight: 1909 bold'> and </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1910 major-latin;mso-hansi-theme-font:major-latin'>microstrain</span></b><span 1911 style='mso-bidi-font-weight:bold'> parameters</span> to add them to the 1912 fit.<span style='mso-spacerun:yes'> </span></p> 1913 1914 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1915 1916 <p class=MsoListParagraphCxSpMiddle>There are now 20 refined parameters and a 1917 repeat of the fitting (Use the Calculate/Refine menu item) causes the fit to 1918 improve to a wR of <16%.</p> 1919 1920 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1921 1922 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l6 level1 lfo4'><![if !supportLists]><span 1923 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>10.<span 1924 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>We 1925 can now <span style='mso-bidi-font-weight:bold'>include the position flag (</span><b><span 1926 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1927 mso-hansi-theme-font:major-latin'>pos/refine</span></b><span style='mso-bidi-font-weight: 1928 bold'>) for the background peak</span> in the refinement. Do this at the <b 1929 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1930 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Background</span></b> 1931 item, as before. Be sure to click on some other box after clicking on the 1932 refine flag, so that the setting is recorded. </p> 1933 1934 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1935 1936 <p class=MsoListParagraphCxSpMiddle>Close examination of the pattern shows that 1976 <P></P><p class=MsoNormal> 1977 Once the background fit has been 1978 improved we can then add the position of this peak to the fit.</p> 1979 <P></P> 1980 1981 <LI>Treat sample broadening.</LI> 1982 1983 <P></P><p class=MsoNormal> 1984 Refine <B>Cryst. Size</B> and isotropic <B>microstrain</B> parameters for the 1985 CuCr2O4 phase and the isotropic <B>microstrain</B> parameter for the 1986 CuO impurity. 1987 1988 <P></P><p class=MsoNormal> 1989 It will be easier to start the sequential refinement with a good fit 1990 for the sample broadening. For the majority phase, we should consider 1991 both crystallite and microstrain broadening. For the CuO impurity, 1992 which only has a few visible peaks, refinement of more than one 1993 broadening parameter is not likely to succeed and use of either 1994 parameter should be fine. 1995 Select the <B>CuCr2O4 phase</B> and the <B>Data</B> tab on the top of 1996 the window and then select to refine the <B>Cryst. Size</B> and 1997 <B>microstrain</B> parameters. 1998 Then select the <B>CuO phase</B> (note that the <B>Data</B> tab is 1999 still displayed) and select to refine the <B>microstrain</B> parameter. 2000 2001 </OL> 2002 2003 <H3>2. Perform additional fit</H3> 2004 2005 <P></P><p class=MsoListParagraphCxSpMiddle> 2006 Starting an additional refinement (using the Calculate/Refine menu item) 2007 with the addition of the six new variables to the refinement improves 2008 the fit significantly visually, though the Rw value drops only to ~15.5%. 2009 2010 <H3>3. Still more sample parameters.</H3> 2011 2012 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2013 Vary background peak position</LI> 2014 2015 <P></P><p class=MsoNormal> 2016 Selecting the <b>Background</b> sub-item in the 2017 <B>PWDR OH_00.fxye Bank 1</b> histogram and click on the checkbox to 2018 refine the position (<B>pos</B>) value for the single background 2019 peak. 2020 2021 <P></P> 2022 <LI>Use Uniaxial (010 axis) microstrain broadening for CuCr2O4</LI> 2023 2024 <P></P><p class=MsoNormal>Close examination of the pattern shows that 1937 2025 the peak widths are still not as well fit as desired. Since strain broadening 1938 2026 seems more significant that sample broadening, we could use an anisotropic … … 1940 2028 can introduce only two terms with the uniaxial model.<span 1941 2029 style='mso-spacerun:yes'> </span>Trial and error will show that the 010 1942 direction works best. Click on the first phase (<b style='mso-bidi-font-weight: 1943 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1944 major-latin;mso-hansi-theme-font:major-latin'>CuCr2O4</span></b>) and <b 1945 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1946 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Data</span></b> 1947 tab and <span style='mso-bidi-font-weight:bold'>select<b> </b></span><b><span 1948 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1949 mso-hansi-theme-font:major-latin'>uniaxial</span></b> for the <b 1950 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1951 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Mustrain 1952 model</span></b> and <span style='mso-bidi-font-weight:bold'>use<b> </b></span><b><span 1953 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1954 mso-hansi-theme-font:major-latin'>0 1 0</span> </b><span style='mso-bidi-font-weight: 1955 bold'>for the<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 1956 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Unique axis</span></b><b 1957 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 1958 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>, HKL</span></b> 1959 . <span style='mso-bidi-font-weight:bold'>Select refinement for both </span><b><span 1960 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1961 mso-hansi-theme-font:major-latin'>Equatorial</span></b><span style='mso-bidi-font-weight: 1962 bold'> and </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 1963 major-latin;mso-hansi-theme-font:major-latin'>Axial mustrain</span>.</b> </p> 1964 1965 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 1966 1967 <p class=MsoListParagraphCxSpMiddle>Repeating the refinement causes the wR to 1968 drop to ~14%. Clicking on the <b style='mso-bidi-font-weight:normal'><span 1969 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 1970 mso-hansi-theme-font:major-latin'>Covariance</span></b> data tree item shows 1971 that the goodness of fit (GOF) is 1.08 (corresponding to a reduced <b 1972 style='mso-bidi-font-weight:normal'><span style='font-family:"Lucida Grande"; 1973 color:black'>χ</span></b><sup>2</sup> of 1.2) with 23 refined parameters. 1974 While atomic parameters could now be refined, as shown in the plot, this is 2030 direction works best. Click on the first phase (<b>CuCr2O4</b>) and 2031 the <b>Data</b> tab and select <b>uniaxial</b> for the <b>Mustrain 2032 model</b> and change the <B>Unique axis, HKL</B> from "0 0 1" to <b>0 2033 1 0</b>. 2034 Select refinement for both <b>Equatorial</b> and <b>Axial 2035 mustrain</b>. 2036 </OL> 2037 2038 <H3>4. Perform additional fit</H3> 2039 2040 <P></P><p class=MsoListParagraphCxSpMiddle> 2041 Starting an additional refinement (using the Calculate/Refine menu item) 2042 with the addition of two more optimized variables to the refinement improves 2043 the fit significantly visually and the Rw value drops to <14%, but the 2044 additon of the new background parameter causes issues with correlation 2045 between the background terms. Considering that we have not refined any 2046 atomic parameters, this is an very good fit despite the seemingly high 2047 Rw value, as shown in the plot below. Clicking on the 2048 <b>Covariance</b> data tree item shows the 2049 GOF=1.08 and the Reduced χ<sup>2</sup>=1.16. This is close to 2050 ideal. 2051 While atomic parameters could now be refined, this is 1975 2052 clearly a good enough place to start refinement with all the datasets. </p> 1976 2053 1977 <p class=MsoListParagraphCxSpLast><span style='mso-fareast-language:EN-US; 1978 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_63" o:spid="_x0000_i1051" 1979 type="#_x0000_t75" style='width:6in;height:370.5pt;visibility:visible; 1980 mso-wrap-style:square'> 1981 <v:imagedata src="SequentialTutorial_files/image015.png" o:title=""/> 1982 </v:shape><![endif]--><![if !vml]><img width=576 height=494 1983 src="SequentialTutorial_files/image066.png" v:shapes="Picture_x0020_63"><![endif]></span></p> 1984 1985 <h1>Step 2: Prepare for sequential fit. </h1> 1986 1987 <p class=MsoListParagraph><o:p> </o:p></p> 1988 1989 <p class=MsoNormal>There is one significant difference between fitting a 1990 sequence of datasets versus fitting a single data set. Rather than fitting 2054 <h1>Step III: Prepare for sequential fit. </h1> 2055 2056 <P></P> 2057 <p class=MsoNormal> 2058 For the initial fitting of datasets over a range of conditions, we 2059 will want to start from the fitted values we have fit, but only refine 2060 a minimal number of parameters until the scale factor, background and 2061 lattice parameters have been optimized for each dataset and more 2062 variables can be added back. 2063 <P></P> 2064 <p class=MsoNormal>There is one significant way that sequential 2065 fitting differs from the usual refinement process. Rather than fitting 1991 2066 lattice constants (for which there will be only one set for each phase in the 1992 2067 entire refinement), we will use the Hydrostatic/Elastic strain terms 1993 (parameters <p>:<h>:Dxx) to treat the change in lattice parameters. 1994 These items describe the change in lattice parameters and appear in the Phase 1995 Data tab, where there is a set for each histogram assigned to the phase. We 1996 will make a switch to use this for the current dataset, in preparation for the 1997 sequential refinement with all datasets. </p> 1998 1999 <p class=MsoNormal><o:p> </o:p></p> 2000 2001 <p class=MsoNormal>Also at this point we will temporarily remove variables that 2002 could make the fit diverge as the lattice constants shift. As an example, when 2003 initially fitting a model with peaks that overlap with the observed peaks, but 2004 not very well, the fit might be improved by a change in the phase fractions 2005 (that might eliminate our impurity CuO phase) or make peaks quite a bit broader 2006 than appropriate, these changes are improbable and could result in a model that 2007 is less correct than the starting point but still be a local LS minimum. In 2008 such cases the best-fit model may never be found. Accordingly, we will turn off 2009 refinement of all parameters except those that determine peak positions (the 2010 Dxx terms) and the overall scale factor. We might include the background terms 2011 since they are unlikely to affect convergence. </p> 2012 2013 <p class=MsoNormal><span style='mso-spacerun:yes'> </span></p> 2014 2015 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l7 level1 lfo6'><![if !supportLists]><span 2016 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2017 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2018 style='mso-bidi-font-weight:bold'>Select the first phase<b> (</b></span><b><span 2019 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2020 mso-hansi-theme-font:major-latin'>CuCr2O4</span>) </b><span style='mso-bidi-font-weight: 2021 bold'>and on the<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 2022 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>General</span> 2023 </b><span style='mso-bidi-font-weight:bold'>tab</span>, click on the <b 2024 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2025 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Refine unit 2026 cell</span></b> option to <span style='mso-bidi-font-weight:bold'>turn off 2027 refinement</span>. </p> 2068 (parameters <tt><p>:<h>:D<I>ij</i></tt>) to treat the change in 2069 lattice parameters from histgram to histogram. 2070 2071 <P></P> 2072 <p class=MsoNormal><I> 2073 Note that GSAS-II shows the user unit cell parameters in the 2074 traditional a, b, c, α, β & γ (distances in Å, angles in 2075 degrees), but computations are performed internally only 2076 after converting this cell to 2077 a version of the reciprocal lattice tensor we call the "A tensor" or 2078 the <tt>A<i>j</i></tt> terms (<a 2079 href="https://gsas-ii.readthedocs.io/en/latest/GSASIIutil.html#gsasiilattice-unit-cells">see 2080 the GSASIIlattice documentation<a/>). 2081 The actual values refined in GSAS-II are the <tt>A<i>j</i></tt> terms, as 2082 allowed by unit cell symmetry. There is one set of <tt>A<i>j</i></tt> values 2083 for each phase, but to treat the situation where the use of a phase in 2084 each histogram may have somewhat different lattice parameters (<a 2085 href="https://subversion.xray.aps.anl.gov/pyGSAS/Tutorials/CWCombined/Combined%20refinement.htm">see 2086 the PbSO<sub>4</sub> combined fitting tutorial</a>, where the x-ray and 2087 neutron data were collected at different temperatures), GSAS-II 2088 defines <tt>D<I>ij</i></tt> terms, labeled as "Hydrostatic/elastic 2089 strain coefficients", for each phase in each histogram and only the unique 2090 terms are shown on the Phase "Data" tab. These values 2091 are simply added to the <tt>A<i>j</i></tt> terms. When any <tt>D<I>ij</i></tt> 2092 terms are non-zero, the changed unit cell parameters are also shown on 2093 the "Data" tab and will differ from the unit cell values shown on the 2094 "General" tab. 2095 </I> 2096 2097 2098 <H3>1. Remove refinement of sensitive parameters (lattice, phase 2099 fractions, size/μstrain, background peak and sample displacement) 2100 and add refinement of the <tt>D<I>ij</i></tt> terms.</H3> 2101 2102 <OL style='PADDING-LEFT:0.75in' type="A"> 2103 <LI> 2104 Select the first phase (<B>CuCr2O4</B>) and on the <B>General</B> tab, 2105 click on the <B>Refine unit cell</B> option to turn off refinement. </P> 2028 2106 2029 2107 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span><span … … 2035 2113 src="SequentialTutorial_files/image067.png" v:shapes="Picture_x0020_64"><![endif]></span></p> 2036 2114 2037 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2038 2039 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l7 level1 lfo6'><![if !supportLists]><span 2040 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 2041 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>On 2042 the <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2043 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Data</span></b> 2044 tab, select <span style='mso-bidi-font-weight:bold'>the<b> </b></span><b><span 2045 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2046 mso-hansi-theme-font:major-latin'>D11</span></b><span style='mso-bidi-font-weight: 2047 bold'>, </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2048 major-latin;mso-hansi-theme-font:major-latin'>D22</span></b><span 2049 style='mso-bidi-font-weight:bold'> and </span><b><span style='font-family:"Calibri",sans-serif; 2050 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>D33</span></b><span 2051 style='mso-bidi-font-weight:bold'> terms</span> to refine the lattice strain 2052 terms. While here, to simplify convergence for the sequential refinement, where 2053 lattice constants change significantly, <B>turn 2054 off</B> the refinement of the <b>three size and mustrain terms and the phase fraction</b>. 2055 The <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2056 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Data</span></b> 2057 window should look something like</p> 2115 <P></P> 2116 <LI>On the Data tab (for <B>CuCr2O4</B>), turn <U>off</U> refinement of the <B>Phase 2117 fraction</B>, the <b>size</b> and both the <b>mustrain</b> 2118 parameters. Turn <U>on</U> refinement of the <B>D11, D22</B> and 2119 <B>D33</B> terms to allow the combined lattice parameters to vary. 2120 The Data window should look something like 2121 </p> 2058 2122 2059 2123 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span><span … … 2065 2129 src="SequentialTutorial_files/image068.png" v:shapes="Picture_x0020_65"><![endif]></span></p> 2066 2130 2067 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2068 2069 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l7 level1 lfo6'><![if !supportLists]><span 2070 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 2071 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2072 style='mso-bidi-font-weight:bold'>Click on the second phase</span> (<b 2073 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2074 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>CuO</span></b>) 2075 and on the <b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2076 major-latin;mso-hansi-theme-font:major-latin'>General</span></b><span 2077 style='mso-bidi-font-weight:bold'> tab</span>, click on the <b><span 2078 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2079 mso-hansi-theme-font:major-latin'>Refine unit cell</span></b><span 2080 style='mso-bidi-font-weight:bold'> option</span> to turn off refinement. </p> 2081 2082 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2083 2084 <p class=MsoListParagraphCxSpMiddle>On the <b><span style='font-family:"Calibri",sans-serif; 2085 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Data</span></b><span 2086 style='mso-bidi-font-weight:bold'> tab</span> for CuO, <span style='mso-bidi-font-weight: 2087 bold'>click on the </span><b><span style='font-family:"Calibri",sans-serif; 2088 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>D11</span></b><span 2089 style='mso-bidi-font-weight:bold'>, </span><b><span style='font-family:"Calibri",sans-serif; 2090 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>D22</span></b><span 2091 style='mso-bidi-font-weight:bold'>, and </span><b><span style='font-family: 2092 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 2093 major-latin'>D33</span></b><span style='mso-bidi-font-weight:bold'> terms</span> 2094 to refine the lattice strain terms and <span style='mso-bidi-font-weight:bold'>turn 2095 off refinement of the phase fraction</span>.<span style='mso-spacerun:yes'> 2096 </span>We could refine the <b style='mso-bidi-font-weight:normal'><span 2097 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2098 mso-hansi-theme-font:major-latin'>D13</span></b> term (which corresponds to the 2099 CuO β* angle), but we do not expect this to change much and expect the 2100 refinement to be relatively insensitive to this, due to the small number of CuO 2101 peaks that are in the pattern.)</p> 2102 2103 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2131 <P></P> 2132 <LI> 2133 Select the second phase (<B>CuO</B>) and on the 2134 Data tab (since that is shown first), turn <U>off</U> refinement of the <B>Phase 2135 fraction</B> and the <b>microstrain</b> 2136 parameters. Turn <U>on</U> refinement of the <B>D11, D22</B> and 2137 <B>D33</B> terms to allow the combined lattice parameters to 2138 vary. We could refine the <B>D13</B> term [which is related to 2139 cos(β*) angle for CuO] but the changes to this are small and 2140 since this is a minor phase and few CuO peaks are seen, it is best 2141 to not vary this, at least initially. The Data window should look 2142 something like </P> 2104 2143 2105 2144 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2111 2150 src="SequentialTutorial_files/image069.png" v:shapes="Picture_x0020_66"><![endif]></span></p> 2112 2151 2113 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>2114 2115 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l7 level1 lfo6'><![if !supportLists]><span 2116 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>4.<span 2117 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Also 2118 at this stage, <span style='mso-bidi-font-weight:bold'>turn off the all the 2119 background refinement terms</span> by clicking on <b style='mso-bidi-font-weight: 2120 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2121 major-latin;mso-hansi-theme-font:major-latin'>Background</span></b> on the 2122 histogram tree item and removing checks from all items (be sure to click 2123 elsewhere in the window before going on to the next step.) </p> 2124 2125 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>2152 <P></P> 2153 <LI> On the <B>General</B> tab for CuO, 2154 click on the <B>Refine unit cell</B> option to turn off refinement of 2155 that. 2156 2157 <P></P> 2158 <LI> Also 2159 at this stage, turn off 2160 refinement of the background peak refinement position and width (sig) 2161 terms by clicking on the <b>Background</b> histogram tree item and removing 2162 checks from the first and third items in the background peaks table 2163 (which has only one row). Be sure to click 2164 elsewhere in the window before going on to the next step. </p> 2126 2165 2127 2166 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2133 2172 src="SequentialTutorial_files/image070.png" v:shapes="Picture_x0020_67"><![endif]></span></p> 2134 2173 2135 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2136 2137 <p class=MsoListParagraphCxSpMiddle>Likewise, for the histogram Sample 2138 Parameters, <span style='mso-bidi-font-weight:bold'>turn off the<b> </b></span><b><span 2139 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2140 mso-hansi-theme-font:major-latin'>Sample X displ.</span> </b><span 2141 style='mso-bidi-font-weight:bold'>parameter</span> by clicking on the <b 2142 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2143 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Sample 2144 Parameters</span></b> item in the histogram entries.</p> 2174 <P></P> 2175 <LI>Likewise, select the histogram <B>Sample Parameters</B> tree entry 2176 and on that window, turn off the refinement flag for <b>Sample X displ.</b> 2177 </p> 2145 2178 2146 2179 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2152 2185 src="SequentialTutorial_files/image025.png" v:shapes="Picture_x0020_68"><![endif]></span></p> 2153 2186 2154 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l7 level1 lfo6'><![if !supportLists]><span 2155 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>5.<span 2156 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>One 2157 can confirm that the number of parameters is now 7 (3 Dxx terms for phase 1, 3 2158 Dxx terms for phase 2 and a scale factor.) If the refinement is repeated, the 2159 fit quality does not change, since these parameters are completely equivalent 2160 to the previous. There may result in very small non-zero values for the Dxx 2161 parameters, but these are much smaller than their uncertainties and are thus 2162 within noise of zero. </p> 2163 2164 <h1>Step 3: Import the remaining datasets.</h1> 2187 <P></P> 2188 <LI>At this point the refinement could be repeated to show that the 2189 fit does not change significantly. If this is done, one will see 2190 small non-zero values for the <tt>D<I>ij</i></tt> terms, but the unit 2191 cell parameters displayed on the "Data" tab are not significantly 2192 different from those computed without the <tt>D<I>ij</i></tt> values 2193 shown on the "General" tab, but due to the larger uncertainties on the 2194 CuO unit cell parameters, the <tt>D<I>ij</i></tt> terms are much 2195 larger than for CuCr2O4. 2196 2197 </OL> 2198 2199 <h1>Step IV: Import the remaining datasets.</h1> 2165 2200 2166 2201 <p class=MsoNormal><o:p> </o:p></p> 2167 2202 2168 <p class=MsoNormal>In this step we bring in the remaining data, duplicate the 2169 histogram parameters as well as the phase/histogram parameters and set up to 2170 perform a sequential fit. </p> 2171 2172 <p class=MsoNormal><o:p> </o:p></p> 2173 2174 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l2 level1 lfo8'><![if !supportLists]><span 2175 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2176 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>The 2177 remaining datasets are read <span style='mso-bidi-font-weight:bold'>using<b> </b></span><b><span 2178 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2179 mso-hansi-theme-font:major-latin'>Import/Powder Data/from GSAS powder data file</span></b> 2180 to open a file browser. Again <span style='mso-bidi-font-weight:bold'>select<b> 2181 </b></span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2182 major-latin;mso-hansi-theme-font:major-latin'>zip archive</span> </b><span 2183 style='mso-bidi-font-weight:bold'>and the downloaded<b> </b></span><b><span 2184 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2185 mso-hansi-theme-font:major-latin'>SeqTut.zip</span></b> file. <span 2186 style='mso-bidi-font-weight:bold'>Select all the .</span><b><span 2187 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2188 mso-hansi-theme-font:major-latin'>fxye</span></b><span style='mso-bidi-font-weight: 2189 bold'> files</span> except <b style='mso-bidi-font-weight:normal'><span 2190 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2191 mso-hansi-theme-font:major-latin'>OH_00.fxye</span></b> (use of the filter can 2192 again be helpful to show just the data file; press <b style='mso-bidi-font-weight: 2193 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2194 major-latin;mso-hansi-theme-font:major-latin'>Set All</span></b> and then 2195 unselect the one unneeded file). </p> 2203 <p class=MsoNormal>In this step we bring in the remaining data. 2204 Note that the values for the parameters for the newly-read histograms 2205 (on the Sample Parameters data item) and the phase/histogram 2206 parameters on each phase's Data tab are set to default values, so we 2207 copy the values from the first histogram to the new sets. 2208 We can then perform an initial sequential fit. </p> 2209 2210 <H3>1. Read the remaining datasets.</H3> 2211 2212 <P></P> 2213 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2214 Read the 2215 remaining datasets using the <B>Import/Powder Data/from GSAS powder data file</span></b> 2216 to open a file browser. Again select <b>zip archive</b> 2217 and the downloaded <b>SeqTut.zip</b> file. Select all the <b>.fxye</b> 2218 files except <b>OH_00.fxye</b> (use of the filter can 2219 again be helpful to show just the data files; press <b>Set All</b> and then 2220 unselect the one unneeded file or equivalently one can select only the 2221 <b>OH_00.fxye</b> file and press <b>Toggle All</b>). The press OK to 2222 start reading the datasets.</p> 2196 2223 2197 2224 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2203 2230 src="SequentialTutorial_files/image026.png" v:shapes="Picture_x0020_69"><![endif]></span></p> 2204 2231 2205 <p class=MsoListParagraphCxSpMiddle>This time one must <span style='mso-bidi-font-weight: 2206 bold'>select the instrument parameter file</span>. </p> 2232 <P></P> 2233 <LI>This time one must select the 2234 instrument parameter file (because the name does not match the dataset 2235 file name). </p> 2207 2236 2208 2237 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2214 2243 src="SequentialTutorial_files/image071.png" v:shapes="Picture_x0020_70"><![endif]></span></p> 2215 2244 2216 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2217 2218 <p class=MsoListParagraphCxSpMiddle><span style='mso-bidi-font-weight:bold'>Set 2219 all data sets to use the two phases</span>, as before in step 1 #6. </p> 2220 2221 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2245 <P></P> 2246 <LI>Set 2247 all data sets to use the two phases</span>, as before in <A href="#I.2.C"> 2248 step I.2.C</A> (above).</LI> 2222 2249 2223 2250 <p class=MsoListParagraphCxSpMiddle>Note that the files will be … … 2229 2256 correct order in the file browser. If files are not read in the 2230 2257 intended order, it will be necessary to 2231 manually change the positions of the tree items (by dragging them with 2258 manually change the positions of the tree items (by dragging them in 2259 the tree with 2232 2260 the right mouse button pressed). 2233 This is not an issue in this example, since the files are list ingin a2261 This is not an issue in this example, since the files are listed in a 2234 2262 zip file in the order they need to be read, but may be of concern 2235 2263 for other cases. 2236 2264 </p> 2237 2265 2238 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2239 2240 <p class=MsoListParagraphCxSpMiddle>Optional: Note that the default graphics 2266 <P></P><LI>Optional: Note that the default graphics 2241 2267 mode is to display the last dataset read in.<span style='mso-spacerun:yes'> 2242 2268 </span>If you want to see the entire ensemble of data sets, click on the main … … 2272 2298 src="SequentialTutorial_files/image072.png" v:shapes="Picture_x0020_71"><![endif]></span></p> 2273 2299 2274 < p class=MsoListParagraphCxSpMiddle>You can then select the <b2300 <P></P><LI>You can then select the <b 2275 2301 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2276 2302 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>C</span></b> … … 2290 2316 src="SequentialTutorial_files/image073.png" v:shapes="Picture_x0020_72"><![endif]></span></p> 2291 2317 2292 <p class=MsoListParagraphCxSpLast> Ive use<b style='mso-bidi-font-weight:2318 <p class=MsoListParagraphCxSpLast>Above, Ive used <b style='mso-bidi-font-weight: 2293 2319 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2294 2320 major-latin;mso-hansi-theme-font:major-latin'>D</span></b> to increase the 2295 2321 contrast & zoomed in to a location that shows the effect of temperature on 2296 2322 the peak positions.</p> 2297 2298 <h1>Step 4: Duplicate information across histograms.</h1> 2299 2300 <p class=MsoNormal><o:p> </o:p></p> 2301 2302 <p class=MsoNormal>In this step we bring in the remaining data, duplicate the histogram 2303 parameters as well as the phase/histogram parameters and set up to perform a 2304 sequential fit. </p> 2305 2306 <p class=MsoListParagraphCxSpFirst><o:p> </o:p></p> 2307 2308 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l5 level1 lfo10'><![if !supportLists]><span 2309 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2310 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Copy 2311 the histogram parameters from the first histogram to the new ones: <span 2312 style='mso-bidi-font-weight:bold'>Select the original </span><b><span 2313 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2314 mso-hansi-theme-font:major-latin'>PWDR OH_00.fxye Bank 1</span></b><span 2315 style='mso-bidi-font-weight:bold'> entry</span>. Use menu command <b><span 2316 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2317 mso-hansi-theme-font:major-latin'>Commands/Copy params</span></b>. </p> 2323 </OL> 2324 <h3>2. Duplicate information across histograms.</h3> 2325 2326 <P></P> 2327 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2328 To copy 2329 the histogram parameters from the first histogram to the new ones: 2330 select the original <B>PWDR OH_00.fxye Bank 1</b> tree entry and then 2331 use menu command <b>Commands/Copy params</b>. 2332 We need to duplicate the values in the <b>Limits</b>, 2333 <b>Background</b> and <b>Sample Parameters</b> sections between all 2334 histograms. The <b>Instrument Parameters</b> have been read from the 2335 instrument parameter file, and have not been changed in this 2336 exercise, so they are already identical. 2318 2337 2319 2338 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2320 2339 2321 <p class=MsoListParagraphCxSpMiddle>This causes a dialog box where the sections 2322 to be copied are selected. </p> 2340 <p class=MsoNormal>This opens a dialog box where the sections 2341 to be copied are selected. Limits, Background and Sample Parameters 2342 must be selected; there will be no difference if Instrument 2343 Parameters is selected, since the parameters start as identical, so 2344 there also is no harm in copying them. This, the default entries, 2345 selecting all sections, need not be changed. </p> 2323 2346 2324 2347 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2330 2353 src="SequentialTutorial_files/image033.png" v:shapes="Picture_x0020_73"><![endif]></span></p> 2331 2354 2332 <p class=MsoListParagraphCxSpMiddle>We need to duplicate the <b 2333 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2334 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Limits</span></b>, 2335 <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2336 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Background</span></b> 2337 and <b style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2338 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Sample 2339 Parameters</span></b> between all histograms. The <b style='mso-bidi-font-weight: 2340 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2341 major-latin;mso-hansi-theme-font:major-latin'>Instrument Parameters</span></b> 2342 have not been changed in this exercise and are all read from the same 2343 instrument parameter file, but there is no harm in copying them, so the default 2344 entries do not need to be changed. <span style='mso-bidi-font-weight:bold'>Press<b> 2345 </b></span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2346 major-latin;mso-hansi-theme-font:major-latin'>OK</span></b>. Note that all 2347 parameters are copied except measurement information such as setting angles, 2348 temperature, etc. </p> 2349 2350 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2351 2352 <p class=MsoListParagraphCxSpMiddle>An additional dialog window is shown which 2355 2356 <P></P><LI> 2357 Press <b>OK</B>. 2358 Note that all Sample Parameters are copied, except measurement 2359 information such as setting angles, temperature, etc. </LI> 2360 2361 2362 <P></P><LI> 2363 An additional dialog window is shown which 2353 2364 asks which histograms should receive the parameters. <span style='mso-bidi-font-weight: 2354 bold'>Press<b> </b></span><b><span style='font-family:"Calibri",sans-serif; 2355 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Set All</span></b><span 2356 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2357 mso-hansi-theme-font:major-latin'> </span>to select all remaining histograms 2358 and then <span style='mso-bidi-font-weight:bold'>press<b> </b></span><b><span 2359 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2360 mso-hansi-theme-font:major-latin'>OK</span>.</b> </p> 2365 bold'>Press <b>Set All</b> to select all remaining histograms 2366 and then press<b>OK</b>. </LI> 2361 2367 2362 2368 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> … … 2370 2376 src="SequentialTutorial_files/image034.png" v:shapes="Picture_x0020_74"><![endif]></span></p> 2371 2377 2372 <p class=MsoListParagraphCxSpLast>That the copy has been performed can be 2373 demonstrated by examining any of the other histogram plots (e.g. to see the 2374 changed lower limit), etc. Next we need to copy the Phase/Histogram parameters 2375 within each phase. </p> 2376 2377 <p class=MsoNormal><o:p> </o:p></p> 2378 2379 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l5 level1 lfo10'><![if !supportLists]><span 2380 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 2381 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2382 style='mso-bidi-font-weight:bold'>Select the first phase</span> (<b 2383 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2384 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>CuCr2O4</span></b>) 2385 and its <b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2386 major-latin;mso-hansi-theme-font:major-latin'>Data</span></b><span 2387 style='mso-bidi-font-weight:bold'> tab</span>. Then do <b style='mso-bidi-font-weight: 2388 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2389 major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>Edit 2390 Phase/</span></b><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2391 major-latin;mso-hansi-theme-font:major-latin'>Copy data</span></b></p> 2378 <P></P><p class=MsoListParagraphCxSpLast>To verify that the copy has 2379 been performed, examining any of the other histogram plots (e.g. to see the 2380 changed lower limit), etc. or view the contents of the various updated sections. 2381 2382 2383 <P></P><LI> 2384 Next we need to copy the Phase/Histogram parameters 2385 within each phase. To do this, select the first phase (<b>CuCr2O4</b>) 2386 in the data tree 2387 and then the <b>Data</b> tab in the window. Then use the <b>Edit Phase/Copy data</b> menu 2388 command, which raises the selction window below. 2389 </LI> 2392 2390 2393 2391 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span><span … … 2399 2397 src="SequentialTutorial_files/image035.png" v:shapes="Picture_x0020_75"><![endif]></span></p> 2400 2398 2401 <p class=MsoListParagraphCxSpMiddle>a dialog window appears to select where to 2402 copy the parameters. Press the <b><span style='font-family:"Calibri",sans-serif; 2403 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Set All</span></b> 2404 button and then <span style='mso-bidi-font-weight:bold'>press<b> </b></span><b><span 2405 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2406 mso-hansi-theme-font:major-latin'>OK</span></b>.</p> 2407 2408 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2409 2410 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l5 level1 lfo10'><![if !supportLists]><span 2411 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 2412 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2413 style='mso-bidi-font-weight:bold'>Copy the Phase/Histogram parameters within 2414 the second phase</span> by clicking on it (<b style='mso-bidi-font-weight:normal'><span 2415 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2416 mso-hansi-theme-font:major-latin'>CuO</span></b>) in the data tree (and 2417 possibly on the <b style='mso-bidi-font-weight:normal'><span style='font-family: 2418 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 2419 major-latin'>Data</span></b> tab.) Then do <b style='mso-bidi-font-weight:normal'><span 2420 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2421 mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>Edit 2422 Phase/Copy data</span></b> from the menu<b> </b>and <span style='mso-bidi-font-weight: 2423 bold'>select the histograms</span> as done in the previous step</p> 2424 2425 <h1>Step 5: Setup and start the sequential refinement. </h1> 2399 <P></P> 2400 <p class=MsoListParagraphCxSpMiddle> 2401 Press the <b>Set All</b> button and then press the <b>OK</b> button.</p> 2402 2403 <P></P><LI> 2404 Repeat the previous step for the CuO phase. 2405 Select <B>CuO</B> in the data tree and then copy the Phase/Histogram 2406 parameters in its <b>Data</b> tab by again using the <b>Edit 2407 Phase/Copy data</b> menu command. Again, select all histograms and 2408 then press the <b>OK</b> button. 2409 </LI> 2410 </OL> 2411 2412 2413 <h1>Step V: Setup and start the sequential refinement. </h1> 2426 2414 2427 2415 <p class=MsoNormal><o:p> </o:p></p> … … 2431 2419 will be used for the subsequent refinements.</p> 2432 2420 2433 < p class=MsoNormal><o:p> </o:p></p>2434 2435 < p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l0 level1 lfo12'><![if !supportLists]><span2436 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2437 s tyle='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>First,2438 <span style='mso-bidi-font-weight:bold'>select the<b> </b></span><b><span 2439 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2440 mso-hansi-theme-font:major-latin'>Controls</span> </b><span style='mso-bidi-font-weight: 2441 bold'>item</span> in the data tree and press the <b><span style='font-family: 2442 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 2443 major-latin'>Select data</span> </b><span style='mso-bidi-font-weight:bold'>button</span>. 2444 </p> 2421 <H3>1. Set overall Controls for a sequential fit and launch it</H3> 2422 2423 <P></P> 2424 <p class=MsoListParagraphCxSpMiddle>Here we enable sequential mode by 2425 setting the histograms to be included, set another control setting 2426 (max cycles) and launch the sequential fit. 2427 2428 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2429 Select the <B>Controls</b> data tree item, in the window (as shown 2430 below) and press the "Sequential Refinement 2431 <b>Select data</b>" button. 2432 2445 2433 2446 2434 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2452 2440 src="SequentialTutorial_files/image074.png" v:shapes="Picture_x0020_76"><![endif]></span></p> 2453 2441 2454 <p class=MsoListParagraphCxSpMiddle>This opens a window where all histograms can 2455 be selected (it would be equally valid to skip the first histogram, since that 2456 has already been fully fit). <span style='mso-bidi-font-weight:bold'>Press<b> </b></span><b><span 2457 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2458 mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>Set All</span> 2459 </b><span style='mso-bidi-font-weight:bold'>and then<b> </b></span><b><span 2460 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2461 mso-hansi-theme-font:major-latin'>OK</span></b> and note that the <b 2462 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2463 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Controls</span></b> 2464 window now lists the number of datasets in the sequential refinement and two 2465 new controls. </p> 2442 <P></P> 2443 <p class=Msonormal>This opens the window where 2444 all histograms can be selected. Press <b>Set All</b> and then press 2445 <B>OK</B>. Note that the Controls window now lists the number of 2446 datasets in the sequential refinement and three new controls below 2447 that. One is the new "Reverse order?" checkbox, which causes 2448 the sequential refinement to start with the last histogram, not the first. 2449 Sometimes this is useful, but not here. The "Clear previous 2450 seq. results" will remove previous sequential results from a .gpx 2451 file, reducing its size. The "Copy results to next histogram?" is 2452 described in the next section. 2466 2453 2467 2454 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2473 2460 src="SequentialTutorial_files/image075.png" v:shapes="Picture_x0020_77"><![endif]></span></p> 2474 2461 2475 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2476 2477 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l0 level1 lfo12'><![if !supportLists]><span 2478 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 2479 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2480 style='mso-bidi-font-weight:bold'>Check the<b> </b></span><b><span 2481 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2482 mso-hansi-theme-font:major-latin'>Copy results to next histogram</span> </b><span 2483 style='mso-bidi-font-weight:bold'>checkbox</span>. This is essential. If this 2484 is not selected, then every refinement will start with the current parameters 2485 (which are all identical, since they were set with the Copy operations.) The 2486 first few refinements will go well, but after the temperature has changed 2487 enough, the starting lattice parameters (determined from the cell and Dxx 2488 values) will be too far off and the refinements will fail. </p> 2489 2490 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span></p> 2491 2492 <p class=MsoListParagraphCxSpMiddle>Note that the Reverse order checkbox causes 2493 the sequential refinement to start with the last histogram, not the first. 2494 Sometimes this is useful, but not here. </p> 2495 2496 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2497 2498 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l0 level1 lfo12'><![if !supportLists]><span 2499 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 2500 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2501 style='mso-bidi-font-weight:bold'>Change </span><b><span style='font-family: 2502 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 2503 major-latin'>Max cycles</span></b><span style='mso-bidi-font-weight:bold'> to<b> 2504 </b></span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2505 major-latin;mso-hansi-theme-font:major-latin'>10</span>.</b> This is important, 2506 so that refinements have enough cycles to fully adjust to the changes in 2507 lattice parameters before being used as the starting point for the next 2508 refinement. </p> 2509 2510 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span></p> 2511 2512 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l0 level1 lfo12'><![if !supportLists]><span 2513 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>4.<span 2514 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Start 2515 the sequential refinement with <b><span style='font-family:"Calibri",sans-serif; 2516 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Calculate/Sequential 2517 refine</span></b><span style='mso-bidi-font-weight:bold'>; this will take some 2518 time as the refinement works its way through all the histograms.</span></p> 2519 2520 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2521 2522 <p class=MsoListParagraphCxSpMiddle>Note that if the copy operations in <b 2523 style='mso-bidi-font-weight:normal'>Step 4</b> were not performed correctly, 2524 the parameters in each refinement will not agree and the refinement will stop 2525 with an error such as this: </p> 2526 2527 <p class=MsoListParagraphCxSpLast><!--[if gte vml 1]><o:wrapblock><v:shapetype 2528 id="_x0000_t202" coordsize="21600,21600" o:spt="202" path="m,l,21600r21600,l21600,xe"> 2529 <v:stroke joinstyle="miter"/> 2530 <v:path gradientshapeok="t" o:connecttype="rect"/> 2531 </v:shapetype><v:shape id="Text_x0020_Box_x0020_3" o:spid="_x0000_s1026" 2532 type="#_x0000_t202" style='position:absolute;left:0;text-align:left; 2533 margin-left:36.55pt;margin-top:-85.9pt;width:424.45pt;height:76.65pt; 2534 z-index:251672576;visibility:visible;mso-wrap-style:square; 2535 mso-width-percent:0;mso-height-percent:0;mso-wrap-distance-left:9pt; 2536 mso-wrap-distance-top:0;mso-wrap-distance-right:9pt; 2537 mso-wrap-distance-bottom:0;mso-position-horizontal:absolute; 2538 mso-position-horizontal-relative:text;mso-position-vertical:absolute; 2539 mso-position-vertical-relative:text;mso-width-percent:0;mso-height-percent:0; 2540 mso-width-relative:margin;mso-height-relative:margin;v-text-anchor:top' 2541 o:gfxdata="UEsDBBQABgAIAAAAIQC2gziS/gAAAOEBAAATAAAAW0NvbnRlbnRfVHlwZXNdLnhtbJSRQU7DMBBF 2542 90jcwfIWJU67QAgl6YK0S0CoHGBkTxKLZGx5TGhvj5O2G0SRWNoz/78nu9wcxkFMGNg6quQqL6RA 2543 0s5Y6ir5vt9lD1JwBDIwOMJKHpHlpr69KfdHjyxSmriSfYz+USnWPY7AufNIadK6MEJMx9ApD/oD 2544 OlTrorhX2lFEilmcO2RdNtjC5xDF9pCuTyYBB5bi6bQ4syoJ3g9WQ0ymaiLzg5KdCXlKLjvcW893 2545 SUOqXwnz5DrgnHtJTxOsQfEKIT7DmDSUCaxw7Rqn8787ZsmRM9e2VmPeBN4uqYvTtW7jvijg9N/y 2546 JsXecLq0q+WD6m8AAAD//wMAUEsDBBQABgAIAAAAIQA4/SH/1gAAAJQBAAALAAAAX3JlbHMvLnJl 2547 bHOkkMFqwzAMhu+DvYPRfXGawxijTi+j0GvpHsDYimMaW0Yy2fr2M4PBMnrbUb/Q94l/f/hMi1qR 2548 JVI2sOt6UJgd+ZiDgffL8ekFlFSbvV0oo4EbChzGx4f9GRdb25HMsYhqlCwG5lrLq9biZkxWOiqY 2549 22YiTra2kYMu1l1tQD30/bPm3wwYN0x18gb45AdQl1tp5j/sFB2T0FQ7R0nTNEV3j6o9feQzro1i 2550 OWA14Fm+Q8a1a8+Bvu/d/dMb2JY5uiPbhG/ktn4cqGU/er3pcvwCAAD//wMAUEsDBBQABgAIAAAA 2551 IQDFYMPKyQIAAOYFAAAOAAAAZHJzL2Uyb0RvYy54bWysVFtv2yAUfp+0/4B4T20ndtNYdSo3UaZJ 2552 UVutnfpMMCRWbWBAYmfT/vsO2E6zbi+d9gIHznful+ubtq7QgWlTSpHh6CLEiAkqi1JsM/z1aTW6 2553 wshYIgpSScEyfGQG38w/frhuVMrGciergmkESoRJG5XhnbUqDQJDd6wm5kIqJoDJpa6JhafeBoUm 2554 DWivq2AchpdBI3WhtKTMGPhddkw89/o5Z9Tec26YRVWGwTfrT+3PjTuD+TVJt5qoXUl7N8g/eFGT 2555 UoDRk6olsQTtdfmHqrqkWhrJ7QWVdSA5LynzMUA0UfgmmscdUczHAskx6pQm8//U0rvDg0ZlkeHJ 2556 FCNBaqjRE2stupUtmrj0NMqkgHpUgLMtfEOZfahGrSV9MQAJzjCdgAG0S0fLde1uCBSBIFTgeMq6 2557 s0LhM5nMwiRKMKLAm00ncZI4u8GrtNLGfmKyRo7IsIaqeg/IYW1sBx0gzpiQq7Kq4J+klXCnkVVZ 2558 uD//0NvNotLoQKAl4tVVdLvszZ3BwHgnynwPdWZICi4D6bQ45319fyyS6TifJrPRZZ5EozgKr0Z5 2559 Ho5Hy1Ue5mG8Wszi25+9hUHeJ6zLkUudsceKde5+YRyq4VPlnXVzwE7uEkqZsFGvrRKAdigOob1H 2560 sMf7OHx87xHuMjJYlsKehOtSSN21xu9uFy+Dy7zD9y3Tx+1SYNtNCxVz5EYWR+g1LbthNYquSqj7 2561 mhj7QDRMJ3QRbBx7DwevZJNh2VMY7aT+/rd/h4ehAS5GDUx7hs23PdEMo+qzgHGaRXHs1oN/xFBR 2562 eOhzzuacI/b1QkL3RLDbFPWkw9tqILmW9TMsptxZBRYRFGxn2A7kwnY7CBYbZXnuQbAQFLFr8ajo 2563 MGKuq5/aZ6JV3/oWOuhODnuBpG8moMO6ugqZ763kpR+P16z2iYdl4gesX3xuW52/Pep1Pc9/AQAA 2564 //8DAFBLAwQUAAYACAAAACEAmMvc0uIAAAALAQAADwAAAGRycy9kb3ducmV2LnhtbEyPwU7DMAyG 2565 70i8Q2QkLmhL2wErpemEEIgJxIHBgWPWeG21xilJtnVvj3eCo+1Pv7+/XIy2F3v0oXOkIJ0mIJBq 2566 ZzpqFHx9Pk9yECFqMrp3hAqOGGBRnZ+VujDuQB+4X8VGcAiFQitoYxwKKUPdotVh6gYkvm2ctzry 2567 6BtpvD5wuO1lliS30uqO+EOrB3xssd6udlbBmL1/26vXJ7nN34YX3x2vZz+bpVKXF+PDPYiIY/yD 2568 4aTP6lCx09rtyATRK5jPUiYVTNJ5yh2YuMsybrc+rfIbkFUp/3eofgEAAP//AwBQSwECLQAUAAYA 2569 CAAAACEAtoM4kv4AAADhAQAAEwAAAAAAAAAAAAAAAAAAAAAAW0NvbnRlbnRfVHlwZXNdLnhtbFBL 2570 AQItABQABgAIAAAAIQA4/SH/1gAAAJQBAAALAAAAAAAAAAAAAAAAAC8BAABfcmVscy8ucmVsc1BL 2571 AQItABQABgAIAAAAIQDFYMPKyQIAAOYFAAAOAAAAAAAAAAAAAAAAAC4CAABkcnMvZTJvRG9jLnht 2572 bFBLAQItABQABgAIAAAAIQCYy9zS4gAAAAsBAAAPAAAAAAAAAAAAAAAAACMFAABkcnMvZG93bnJl 2573 di54bWxQSwUGAAAAAAQABADzAAAAMgYAAAAA 2574 " filled="f" strokecolor="#4f81bd"> 2575 <v:path arrowok="t"/> 2576 <v:textbox> 2577 <![if !mso]> 2578 <table cellpadding=0 cellspacing=0 width="100%"> 2579 <tr> 2580 <td><![endif]> 2581 <div> 2582 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'>**** 2583 ERROR - variable list for this histogram does not match previous<o:p></o:p></span></p> 2584 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'><o:p> </o:p></span></p> 2585 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'>current 2586 histogram PWDR OH_04.fxye Bank 1 has 6 variables<o:p></o:p></span></p> 2587 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'>Variables 2588 in previous but not in current: 1::D22, 1::D33, 1::D11, <o:p></o:p></span></p> 2589 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'><o:p> </o:p></span></p> 2590 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'>Previous 2591 refinement has 7 variables<o:p></o:p></span></p> 2592 <p class=MsoNormal><span style='font-size:10.0pt;font-family:Courier'>Variables 2593 in current but not in previous: 1::Scale, 1::Mustrain;i, <o:p></o:p></span></p> 2594 <p class=MsoNormal><span style='font-family:"Cambria",serif'><o:p> </o:p></span></p> 2595 </div> 2596 <![if !mso]></td> 2597 </tr> 2598 </table> 2599 <![endif]></v:textbox> 2600 <w:wrap type="topAndBottom"/> 2601 </v:shape><![endif]--><![if !vml]><span style='mso-ignore:vglayout'> 2602 <table cellpadding=0 cellspacing=0> 2603 <tr> 2604 <td width=48 height=0></td> 2605 </tr> 2606 <tr> 2607 <td></td> 2608 <td><img width=572 height=108 src="SequentialTutorial_files/image076.png" 2609 alt="Text Box: **** ERROR - variable list for this histogram does not match previous current histogram PWDR OH_04.fxye Bank 1 has 6 variables Variables in previous but not in current: 1::D22, 1::D33, 1::D11, Previous refinement has 7 variables Variables in current but not in previous: 1::Scale, 1::Mustrain;i, " 2610 v:shapes="Text_x0020_Box_x0020_3"></td> 2611 </tr> 2612 </table> 2613 </span><![endif]><!--[if gte vml 1]></o:wrapblock><![endif]--><br 2614 style='mso-ignore:vglayout' clear=ALL> 2615 (This error results if the parameters for the second phase were not copied in 2616 part 3 so that the phase fraction and strain broadening are retained for all 2617 but the first histogram and the 1::D<i style='mso-bidi-font-style:normal'>ii</i> 2618 terms are only included the first.)</p> 2619 2620 <h1>Step 6: Examine the sequential refinement results.</h1> 2621 2622 <p class=MsoListParagraphCxSpFirst><o:p> </o:p></p> 2623 2624 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l3 level1 lfo14'><![if !supportLists]><span 2625 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2626 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>When 2627 the refinement completes, a window will be displayed where you are asked if you 2628 want to load the results:</p> 2462 <P></P><LI> 2463 Check the "<b>Copy results to next histogram</b>?" checkbox. When this 2464 is selected, the parameters obtained after fitting each histogram 2465 will be used as the starting point for the next fit, which is 2466 essential here since the lattice parameters change considerably. If 2467 this were not selected, then refinement of each histogram would 2468 start with the current parameters -- that were obtained from fitting 2469 the first (~7 K) dataset. 2470 The first few refinements will go well, but when reaching datasets 2471 where the temperature 2472 was sufficiently different from 7 K, the starting lattice parameters may be 2473 too far off and the refinements will fail. Even if that were not 2474 true, since the temperatures increase in the series, the previous 2475 histogram will have the closest set of lattice parameters and will 2476 thus refine most quickly. </p> 2477 2478 <P></P><LI>Also on the Controls window, change the <B>Max cycles</B> 2479 setting to <B>10</B>. This makes sure that the refinements have enough 2480 cycles to fully adjust to the changes in lattice parameters before 2481 being used as the starting point for the next refinement. Note that 2482 if the refinement converges, no additional cycles are used, so 10 is a 2483 maximum that is not likely to be reached. 2484 </LI> 2485 2486 </OL> 2487 2488 <H3>2. Start the sequential refinement</H3> 2489 2490 2491 <p class=MsoListParagraphCxSpMiddle>Start the sequential refinement 2492 with the <b>Calculate/Sequential refine</b> menu command (note that 2493 once the histograms were selected for a sequential fit in the previous 2494 step, the menu command was relabeled from Calculate/Sequential 2495 refine.) The refinement may take a few minutes as each of the 17 2496 datasets are fit in turn.</p> 2497 2498 2499 <h3>3. Accept the sequential refinement results.</h3> 2500 2501 <p class=MsoListParagraphCxSpMiddle>When the refinement completes, a window will be displayed where you 2502 are asked if you want to load the results, as below. 2629 2503 2630 2504 <p class=MsoNormal style='text-indent:.5in'><span style='mso-fareast-language: … … 2636 2510 src="SequentialTutorial_files/image041.png" v:shapes="Picture_x0020_78"><![endif]></span></p> 2637 2511 2638 <p class=MsoListParagraphCxSpFirst><span style='mso-bidi-font-weight:bold'>After 2639 doing </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2640 major-latin;mso-hansi-theme-font:major-latin'>OK</span></b>, the <b 2641 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2642 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Sequential 2643 Results</span></b> item in the data tree is automatically selected, which 2644 displays a summary of the refinement results as a large table. </p> 2645 2646 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2512 <P></P> 2513 <p class=MsoListParagraphCxSpFirst>After clicking on <B>OK</B> 2514 the Sequential Results item is created in the data tree and is is 2515 selected, which displays a summary of the refinement results as a 2516 large table, as below.</p> 2517 2647 2518 2648 2519 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2662 2533 been fit). The Δχ<sup>2</sup> value in the third column shows the 2663 2534 percent improvement in χ<sup>2</sup> seen in the last cycle of refinement. 2664 (If this changes from 4.1 to 4.0, then the value will be 100*(4.1-4.0)/4.0, or 2665 2.5). Since the convergence criterion in the controls is 0.0001 (or 0.01%) we 2666 can see that only one refinement failed to converge. However, these shifts are 2667 still quite small, so that there is no reason to provide further cycles of 2668 refinement without adding new variable parameters. </p> 2535 (Meaning that if in that cycle the Δχ<sup>2</sup> value 2536 decreases from 4.1 to 4.0, then the value will be 100*(4.1-4.0)/4.0, or 2537 2.5%; small numbers indicate that the fit converged.). Note that the 2538 fits had pretty well stopped improving, so there is no reason to 2539 provide further cycles of refinement without adding new variable 2540 parameters. </p> 2669 2541 2670 2542 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> … … 2672 2544 <p class=MsoListParagraphCxSpMiddle>Subsequent columns will show any of the 2673 2545 sample parameters that vary across the datasets, here only temperature and then 2674 the direct cell lattice constants for each phase, generated from the A<i 2675 style='mso-bidi-font-style:normal'>x</i> and D<i style='mso-bidi-font-style: 2676 normal'>ij</i> terms. Scrolling further shows the values of each refined 2677 parameter. </p> 2678 2679 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2546 the direct cell lattice constants for each phase, generated from the A<i>x</i> and D<i>ij</i> terms. Scrolling further shows columns the values of most 2547 parameters that differ by histogram. Note that some parameters (such 2548 as the actual D<i>ij</i> terms) are not shown. The choices for which 2549 items that are included is determined by the "Hide columns..." command in 2550 the"Columns/Rows" menu. 2551 </p> 2552 2553 <h3>4. Examine the sequential refinement results.</h3> 2680 2554 2681 2555 <p class=MsoListParagraphCxSpMiddle>Before going further, it is useful to … … 2687 2561 this plot:</p> 2688 2562 2689 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>2690 2691 2563 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 2692 2564 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_80" o:spid="_x0000_i1034" … … 2697 2569 src="SequentialTutorial_files/image078.png" v:shapes="Picture_x0020_80"><![endif]></span></p> 2698 2570 2699 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2700 2701 <p class=MsoListParagraphCxSpMiddle>Pressing the <b style='mso-bidi-font-weight: 2702 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2703 major-latin;mso-hansi-theme-font:major-latin'>K</span></b> icon or <span 2704 style='mso-bidi-font-weight:bold'>pressing the </span><b><span 2705 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2706 mso-hansi-theme-font:major-latin'>s</span></b><span style='mso-bidi-font-weight: 2707 bold'> key</span> brings up a menu where the x-axis can be selected as 2571 <P></P> 2572 <p class=MsoListParagraphCxSpMiddle>Pressing the <b>K</b> icon or pressing the <b>s</b> key brings up a menu where the x-axis can be selected as 2708 2573 Temperature. If the plot fails to update, click on another column and then 2709 2574 return to the original column. </p> 2710 2711 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p>2712 2575 2713 2576 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2721 2584 <p class=MsoListParagraphCxSpLast><o:p> </o:p></p> 2722 2585 2723 <h1>Step 7: Continue the sequential refinement. </h1>2586 <h1>Step VI: Continue the sequential refinement. </h1> 2724 2587 2725 2588 <p class=MsoNormal><o:p> </o:p></p> 2726 2589 2727 <p class=MsoNormal>At this point we choose to continue the refinements from 2728 their current values (note that the <b style='mso-bidi-font-weight:normal'><span 2729 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2730 mso-hansi-theme-font:major-latin'>Copy results to next histogram</span></b> 2731 checkbox is automatically unchecked after a successful sequential refinement) 2732 including some additional parameters in the refinement. We will do this is a 2733 few steps. </p> 2734 2735 <p class=MsoListParagraph style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2736 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>1.<span 2737 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]><span 2738 style='mso-bidi-font-weight:bold'>Restore the refinement of background 2739 parameters</span> as before for any one histogram and then <span 2740 style='mso-bidi-font-weight:bold'>use the<b> </b></span><b><span 2741 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2742 mso-hansi-theme-font:major-latin'>File/Copy flags</span></b> menu item to copy 2743 these refinement flags (only) to all other histograms. NB: <b style='mso-bidi-font-weight: 2744 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2745 major-latin;mso-hansi-theme-font:major-latin'>Copy</span></b> would copy the 2746 parameter values as well; probably dont want to do that here. </p> 2747 2748 <p class=MsoNormal><o:p> </o:p></p> 2749 2750 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2751 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>2.<span 2752 style='font:7.0pt "Times New Roman"'> 2753 </span></span></span><![endif]>Likewise, in the Sample Parameters tree item 2754 <span style='mso-bidi-font-weight:bold'>include the<b> </b></span><b><span 2755 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2756 mso-hansi-theme-font:major-latin'>Sample X displ</span></b> parameter and again 2757 use the <b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2758 major-latin;mso-hansi-theme-font:major-latin'>Command/Copy flags</span> </b><span 2759 style='mso-bidi-font-weight:bold'>menu item<b>.</b></span> </p> 2760 2761 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2762 2763 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2764 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>3.<span 2765 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>For 2766 Phase parameters <span style='mso-bidi-font-weight:bold'> in the Data 2767 section, add the three 2768 broadening terms</span> for the CuCr<sub>2</sub>O<sub>4</sub> phase and do <b 2769 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2770 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font: 2771 major-latin'>Edit Phase/Copy flags</span></b>. (Important: not Copy!) For the <span 2772 style='mso-bidi-font-weight:bold'>CuO phase, add refinement of the single sample broadening 2773 term and the remaining lattice parameter</span> via the <b style='mso-bidi-font-weight: 2774 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2775 major-latin;mso-hansi-theme-font:major-latin'>D13</span></b> term.<span 2776 style='mso-spacerun:yes'> </span>Again, do <b style='mso-bidi-font-weight: 2777 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2778 major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>Edit 2779 Phase/Copy flags</span></b>. Note that an additional step is needed in order to 2780 vary Phase fractions, as will be discussed below. </p> 2781 2782 <p class=MsoNormal><o:p> </o:p></p> 2783 2784 <p class=MsoListParagraphCxSpFirst style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2785 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>4.<span 2786 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Start 2787 the sequential refinement again with the <b><span style='font-family:"Calibri",sans-serif; 2788 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Calculate/Sequential 2789 refine</span> </b><span style='mso-bidi-font-weight:bold'>menu item</span>. The 2790 fits are significantly improved at higher temperatures, with R<sub>wp</sub> 2791 values between 13.8 and 17.</p> 2792 2793 <h1>Step 8: Further sequential refinements. </h1> 2794 2795 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2796 2797 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2798 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>5.<span 2799 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>We 2800 can now add the phase fractions into the refinement, by clicking on the <b 2801 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2802 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Phase 2803 fraction</span></b> checkbox in the Phase/Data tab. Do this for each 2804 Phase followed by use of <b style='mso-bidi-font-weight:normal'><span 2805 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2806 mso-hansi-theme-font:major-latin;mso-bidi-theme-font:major-latin'>Edit Phase/</span></b><b 2807 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2808 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Copy 2809 Flags</span></b>. However, a constraint is still needed as before, but 2810 to be applied to each histogram. Constraints entered with a specific histogram 2811 number, as was done before, will be ignored. For sequential 2812 refinements, only constraints where the histogram number is specified 2813 as "*" are used. 2814 2815 <p class=MsoListParagraphCxSpFirst><o:p> </o:p></p> 2816 2817 <p class=MsoListParagraphCxSpMiddle>To set up the new constraint, <span 2818 style='mso-bidi-font-weight:bold'>click on the </span><b><span 2819 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2820 mso-hansi-theme-font:major-latin'>Constraints</span></b><span style='mso-bidi-font-weight: 2821 bold'> item</span> in the data tree and the <b><span style='font-family:"Calibri",sans-serif; 2822 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Select tab/Histogram/Phase</span></b><span 2823 style='mso-bidi-font-weight:bold'> menu item</span>, and then <span 2824 style='mso-bidi-font-weight:bold'>click on the </span><b><span 2825 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2826 mso-hansi-theme-font:major-latin'>Edit Constr./Add constraint equation</span></b> 2827 menu item. From the menu, we either <b><span style='font-family:"Calibri",sans-serif; 2828 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>0:*:Scale</span></b><span 2829 style='mso-bidi-font-weight:bold'> or </span><b><span style='font-family:"Calibri",sans-serif; 2830 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin;mso-bidi-theme-font: 2831 major-latin'>1:*:Scale</span></b><span style='mso-bidi-font-weight:bold'> wild 2832 card phase fraction</span>; they are at the bottom of each set </p> 2833 2834 <p class=MsoListParagraphCxSpMiddle><span style='mso-spacerun:yes'> </span><span 2835 style='mso-fareast-language:EN-US;mso-no-proof:yes'><!--[if gte vml 1]><v:shape 2836 id="Picture_x0020_83" o:spid="_x0000_i1031" type="#_x0000_t75" style='width:6in; 2837 height:276pt;visibility:visible;mso-wrap-style:square'> 2838 <v:imagedata src="SequentialTutorial_files/image050.png" o:title=""/> 2839 </v:shape><![endif]--><![if !vml]><img width=576 height=368 2840 src="SequentialTutorial_files/image081.png" v:shapes="Picture_x0020_83"><![endif]></span></p> 2841 2842 <p class=MsoListParagraphCxSpMiddle>There will only be one possible matching variable 2843 in the next window, so <span style='mso-bidi-font-weight:bold'>select that and 2844 press </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2845 major-latin;mso-hansi-theme-font:major-latin'>OK</span></b>.</p> 2846 2847 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 2848 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_84" o:spid="_x0000_i1030" 2849 type="#_x0000_t75" style='width:6in;height:161.25pt;visibility:visible; 2850 mso-wrap-style:square'> 2851 <v:imagedata src="SequentialTutorial_files/image052.png" o:title=""/> 2852 </v:shape><![endif]--><![if !vml]><img width=576 height=215 2853 src="SequentialTutorial_files/image052.png" v:shapes="Picture_x0020_84"><![endif]></span></p> 2854 2855 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2856 2857 <p class=MsoListParagraphCxSpMiddle>The initial and the most recently defined 2858 constraints will then be shown. </p> 2859 2860 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 2861 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_85" o:spid="_x0000_i1029" 2862 type="#_x0000_t75" style='width:6in;height:137.25pt;visibility:visible; 2863 mso-wrap-style:square'> 2864 <v:imagedata src="SequentialTutorial_files/image053.png" o:title=""/> 2865 </v:shape><![endif]--><![if !vml]><img width=576 height=183 2866 src="SequentialTutorial_files/image082.png" v:shapes="Picture_x0020_85"><![endif]></span></p> 2867 2868 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2869 2870 <p class=MsoListParagraphCxSpMiddle>These two constraints do not 2871 interfere, since the first will be used during a non-sequential 2872 refinement and the second is used only for sequential fits, but if 2873 desired the first constraint can be deleted by pressing the </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2874 major-latin;mso-hansi-theme-font:major-latin'>Delete</span></b><span 2875 style='mso-bidi-font-weight:bold'> button for the first</span> one to remove 2876 it, but this is optional.</p> 2877 2878 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2879 2880 <p class=MsoListParagraphCxSpMiddle style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2881 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>6.<span 2882 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Start 2883 the sequential refinement again with the <b><span style='font-family:"Calibri",sans-serif; 2884 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Calculate/Sequential 2885 refine</span> </b><span style='mso-bidi-font-weight:bold'>menu item<b>.</b></span> 2886 Five new columns appear at the end of the table for <b style='mso-bidi-font-weight: 2887 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2888 major-latin;mso-hansi-theme-font:major-latin'>::constr0</span></b> (generated 2889 by the constraint) and the selected variables, <b style='mso-bidi-font-weight: 2890 normal'><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2891 major-latin;mso-hansi-theme-font:major-latin'>0:*:Scale</span></b> and <b 2892 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2893 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>1:*:Scale</span></b>. Also 2894 included are the weight fractions for each phase. Note that the phase 2895 fractions all add to one. 2896 Changes to the R<sub>wp</sub> values are quite small, even though the value for 2897 the minor phase changes significantly. This indicates that the refinement is 2898 relatively insensitive to these values. 2590 <p class=MsoNormal>At this point we wish to continue the refinements 2591 using the the current as-refined parameter values as the starting 2592 point. This is the default action for the program as the <b>Copy 2593 results to next histogram</b> checkbox is automatically unchecked 2594 after a successful sequential refinement. Note that 2595 in subsequent actions we will use the "Copy flags" menu commands, 2596 which copy only refinement flags not parameter values (as opposed to 2597 the "Copy" commands used previously, which copy both values and flags.) 2899 2598 </p> 2900 2599 2901 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 2902 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_88" o:spid="_x0000_i1027" 2903 type="#_x0000_t75" style='width:6in;height:198pt;visibility:visible; 2904 mso-wrap-style:square'> 2905 <v:imagedata src="SequentialTutorial_files/image057.png" o:title=""/> 2906 </v:shape><![endif]--><![if !vml]><img width=576 height=264 2907 src="SequentialTutorial_files/image084.png" v:shapes="Picture_x0020_88"><![endif]></span></p> 2908 2909 <h1>Step 9: Complete the sequential refinement. </h1> 2910 2911 <p class=MsoListParagraphCxSpFirst><o:p> </o:p></p> 2912 2913 <p class=MsoListParagraphCxSpMiddle>The last thing we will add to the 2600 2601 <h3>1. Restore the parameters that were previously fit.</h3> 2602 2603 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2604 Restore the refinement of background parameters as 2605 in the single-histogram fit. Select <B>Background</B> for any 2606 histogram and add the "Peaks in Background" width (<B>sig</B>) 2607 refine flag. 2608 Note that the peak position could be refined as well, but this is 2609 not well-determined by the data since very little of the low-angle 2610 side of the peak is present, so inclusion of this adds to the 2611 refinement instability without any significant improvement to the 2612 fit. 2613 </LI> 2614 <P></P><LI> 2615 After setting the refinement flag, use the 2616 <B>Background/Copy flags</b> menu item to copy these refinement 2617 flags to all other histograms.</LI> 2618 2619 <P></P><LI> 2620 Likewise, in the <B>Sample Parameters</B> tree item, include the 2621 <B>Sample X displ</b> parameter and again 2622 use the <b>Copy flags</b> menu command (here in the <B>Command</B> menu). 2623 </LI> 2624 2625 <P></P><LI> 2626 For the <B>CuCr<sub>2</sub>O<sub>4</sub></B> phase, return to the 2627 <B>Data</B> tab and enable refinement for the <B>phase fraction</B> 2628 and the three <B>broadening</B> terms (isotropic size and equatorial 2629 and axial mustrain). Again use the <B>Edit Phase/Copy flags</b> menu 2630 command to set these flags for all histograms. </LI> 2631 2632 <P></P><LI> 2633 For the <B>CuO</B> phase, also on the <B>Data</B> tab, 2634 enable refinement for the <B>phase fraction</B> 2635 and the single isotropic <B>microstrain broadening</B> term. Also 2636 add refinement of the <B>D13</B> term. Again use the <B>Edit 2637 Phase/Copy flags</b> menu command to set this for all histograms. 2638 </IL> 2639 2640 <P></P><LI> 2641 Select the <B>Constraints</B> data tree item and note that the 2642 second line of controls that appears for sequential fits. 2643 Confirm that the <B>"Wildcard use"</B> setting is <B>"Set hist # to *"</B>. 2644 2645 <P></P> 2646 <p class=MsoListParagraphCxSpMiddle> 2647 <I>Note: 2648 Refinement of phase 2649 fractions requires a constraint as was added in <A href="#I.3">Step 2650 I.3</A>, but for all histograms. 2651 As of version 5038 of GSAS-II three modes were added for constraints 2652 in sequential refinements. 2653 <UL style='PADDING-LEFT:0.75in'> 2654 <LI>In </I>"Ignore unless hist=*"<I> mode (which was 2655 the only mode prior to this version), any constraints with a 2656 specific histogram number specified (such as </I><tt>":0:0:Scale"</tt><I> and 2657 </I><tt>":1:0:Scale"</tt><I> as used previously) would be ignored and a new 2658 constraint defining </I><tt>":0:*:Scale + :1:*:Scale = 1"</tt><I> would be 2659 required. 2660 <LI>There is now a mode </I>"Set hist # to *"<I> where any 2661 constraints are interpreted to apply to all histograms, despite the 2662 supplied number. <LI>The third mode, </I>"Use as supplied", uses both types 2663 of constraints, so differing constraints can be supplied for 2664 different histograms, but if both </I><tt>":0:*:Scale + :1:*:Scale = 1"</tt><I> and 2665 </I><tt>":0:0:Scale + :1:0:Scale = 1"</tt><I> were supplied, an error would occur 2666 because both constraints would be used for histogram 0. 2667 </UL> 2668 </I> 2669 </OL> 2670 2671 <h3>2. Perform the refinement.</h3> 2672 2673 <p class=MsoListParagraphCxSpMiddle> 2674 Start an additional refinement (using the Calculate/Refine menu item). 2675 Now that the full suite of variables is available for fitting 2676 background, peak widths, etc., the fits improve significantly for 2677 the higher temperature datasets. Before, for the highest temperature 2678 datasets, the Rwp values were in the range of 25-30%. After this 2679 fit, all the Rwp values range from 13-17%. Click on a row header to 2680 see a plot of the fit and double-click on a row header to see the 2681 convariance matrix plotted. 2682 Note that additional columns appear now in the Sequential results 2683 table for the newly-added variables. Also 2684 included are the weight fractions. Note that the sum of the phase 2685 fractions for each histogram are indeed 1.0. 2686 </P> 2687 2688 <h1>Step VII: Complete the sequential refinement. </h1> 2689 2690 <P></P> 2691 <p class=MsoListParagraphCxSpFirst> 2692 The last thing we will add to the 2914 2693 refinement will be atomic parameters. Note that these are handled a bit 2915 2694 differently than the previous things we have refined. There is only one set of 2916 2695 coordinates for each phase and these will be refined for each histogram, but 2917 are not saved individually, except in the summary table.<span 2918 style='mso-spacerun:yes'> </span>If we restart the refinement, the coordinates 2919 from the last histogram will be used to fit the first. The Reverse order 2920 checkbox on the Controls causes the sequential refinement to start with the 2921 last histogram, which would avoid this problem. </p> 2922 2923 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2924 2925 <p class=MsoListParagraphCxSpMiddle><span style='mso-bidi-font-weight:bold'>Select 2926 the </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2927 major-latin;mso-hansi-theme-font:major-latin'>CuCr2O4</span></b><span 2928 style='mso-bidi-font-weight:bold'> phase</span> to and then the <b 2929 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2930 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Atoms</span></b> 2931 tab. <span style='mso-bidi-font-weight:bold'>Double-click on the </span><b><span 2932 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2933 mso-hansi-theme-font:major-latin'>refine</span></b><span style='mso-bidi-font-weight: 2934 bold'> label</span> at the top of the atoms table, a refinement controls dialog 2935 opens, <span style='mso-bidi-font-weight:bold'>select </span><b><span 2936 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2937 mso-hansi-theme-font:major-latin'>X</span></b><span style='mso-bidi-font-weight: 2938 bold'> to refine coordinates (only O has coordinates not fixed by symmetry) and 2939 </span><b><span style='font-family:"Calibri",sans-serif;mso-ascii-theme-font: 2940 major-latin;mso-hansi-theme-font:major-latin'>U</span></b> to refine the atomic 2941 displacement (Uiso) values.<span style='mso-spacerun:yes'> </span><span 2942 style='mso-bidi-font-weight:bold'>Press </span><b><span style='font-family: 2943 "Calibri",sans-serif;mso-ascii-theme-font:major-latin;mso-hansi-theme-font: 2944 major-latin'>OK</span></b><span style='mso-bidi-font-weight:bold'>.</span> </p> 2945 2696 are saved individually only in the Sequential results table. To update 2697 the coordinates shown in the Phase data tree entry, use the "Update 2698 Phase from selected" command in the"Columns/Rows" menu. If a row is 2699 selected in the table, that will be used. If no rows are selected, you 2700 will be prompted to select a histogram. 2701 2702 </p> 2703 <h3>1. Vary CuCr<sub>2</sub>O<sub>4</sub> coordinates and 2704 U<sub>iso</sub> Parameters.</h3> 2705 2706 <P></P> 2707 <OL style='PADDING-LEFT:0.75in' type="A"><LI> 2708 Select the <b>CuCr2O4</b> phase</span> and then the <b>Atoms</b> 2709 tab. Double-click on the <b>refine</b>column label and a refinement 2710 controls dialog opens as below, select <b>X</b> to refine 2711 coordinates and <b>U</b> to refine the atomic displacement (Uiso) 2712 values. 2713 (<I>Note that only the O atom has coordinates that can be refined, 2714 all coordinates for Cu and Cr are fixed by symmetry. 2715 Use of the X flag for Cu and Cr will not cause those coordinates to 2716 change.</I>) 2717 </LI> 2718 2946 2719 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; 2947 2720 mso-no-proof:yes'><!--[if gte vml 1]><v:shape id="Picture_x0020_89" o:spid="_x0000_i1026" … … 2952 2725 src="SequentialTutorial_files/image059.png" v:shapes="Picture_x0020_89"><![endif]></span></p> 2953 2726 2954 <p class=MsoListParagraphCxSpMiddle>Note the change in the atom table.</p> 2727 <P></P> 2728 2729 <LI>Note the change that appears in the refine column of the atoms 2730 table, as below. Since these parameters are shared in all 2731 histograms, no Copy or Copy flags menu item is needed or even is 2732 available for atomic parameters.</LI> 2955 2733 2956 2734 <p class=MsoListParagraphCxSpMiddle><span style='mso-fareast-language:EN-US; … … 2962 2740 src="SequentialTutorial_files/image085.png" v:shapes="Picture_x0020_90"><![endif]></span></p> 2963 2741 2964 <p class=MsoListParagraphCxSpMiddle>Note that since these parameters are shared 2965 directly in all histograms, no Copy or Copy flags menu item is needed or even 2966 available. </p> 2967 2968 <p class=MsoListParagraphCxSpMiddle><o:p> </o:p></p> 2969 2970 <p class=MsoListParagraphCxSpLast style='text-indent:-.25in;mso-list:l1 level1 lfo16'><![if !supportLists]><span 2971 style='mso-fareast-font-family:"Times New Roman"'><span style='mso-list:Ignore'>7.<span 2972 style='font:7.0pt "Times New Roman"'> </span></span></span><![endif]>Start 2973 the sequential refinement again with the <b><span style='font-family:"Calibri",sans-serif; 2974 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Calculate/Sequential 2975 refine</span> </b><span style='mso-bidi-font-weight:bold'>menu item</span>. The 2976 fits are significantly improved at higher temperatures, with R<sub>wp</sub> 2977 values between 13.8 and 16.4. Be sure to save the project file by <span 2978 style='mso-bidi-font-weight:bold'>selecting the </span><b><span 2979 style='font-family:"Calibri",sans-serif;mso-ascii-theme-font:major-latin; 2980 mso-hansi-theme-font:major-latin'>File/Save project</span></b><span 2981 style='mso-bidi-font-weight:bold'> menu item</span>, since this will used in a 2982 later tutorial.</p> 2983 2984 <p>This completes the sequential refinement tutorial, although many more things 2985 could be attempted to further improve the refinement quality. In the <b 2986 style='mso-bidi-font-weight:normal'><span style='font-family:"Calibri",sans-serif; 2987 mso-ascii-theme-font:major-latin;mso-hansi-theme-font:major-latin'>Parametric Fitting 2988 and Pseudo Variables for Sequential Fits</span></b> tutorial, functions of 2989 fitted parameters are computed and plotted and results are fitted to equations 2990 of state.</p> 2742 </OL> 2743 <P></P> 2744 <h3>2. Perform final sequential fit, including coordinates displacement parameters.</h3> 2745 <P></P> 2746 <p class=MsoListParagraphCxSpMiddle> Start 2747 the sequential refinement again with the <b>Calculate/Sequential 2748 refine</b> menu item. Small but significant improvements are seen for 2749 fits with the higher temperature datasets, with R<sub>wp</sub> 2750 values between 13.8 and 16.5. </P> 2751 2752 2753 <p>This completes the sequential refinement tutorial, although a few more things 2754 could be attempted to try to further improve the refinement 2755 quality. Be sure to keep the final project file, 2756 since this will used in the <b>Parametric Fitting and Pseudo Variables for 2757 Sequential Fits</b> tutorial, where functions of fitted parameters are 2758 computed and plotted and these results are fitted to equations of state. 2759 </p> 2991 2760 2992 2761 </div>
Note: See TracChangeset
for help on using the changeset viewer.
