單晶衍射儀Apex+II使用系列教程-1

1、Single Crystal Diffractometry using the Apex2 SoftwareBruce M. FoxmanBrandeis UniversityA tutorialVersion 1.151byBruce M. Foxman (Brandeis University)Use of this tutorial is governed by the terms of the Creative Commons Attribution 3.0 Unported License, February 2010.2This tutorial is dedicated to t

2、he pioneers of Single Crystal Diffractometry. The author was influenced by reading their papers, as well as by meeting all of the wonderful people (except for Siem Poot) listed or shown below, while he was a student in the late Professor F. Albert Cottons research group at MIT, 1964-1968, and during

3、 later years during his appointment as Sometime Professor of Crystallography at Brandeis University, 1972-present. There will be more about these important scientists later (see Chapter 4).Leroy E. AlexanderRobert A. SparksPhotos needed:Thomas C. Furnas; Siem PootWilliam R. BusingHenri A. LevyJohn S

4、. Rollett3Here is the front cover of Tom Furnas excellent treatise on single-crystal diffractometrythe SCOIM, as we called itit can still be found if you have the time and interest!4.and heres the Title Page: it was a product of the General Electric X-ray Department !5CONTENTSThe following links pro

5、vide navigation once you have started to use the tutorial, but please begin by selecting Unit 1 below (later, you can return here or leave the tutorial by pushing the ESC key one or more times): (black ink topics are under development):Apex2: A tutorial using the Ylid Standard CrystalApex2 Quick Tri

6、cks, Sidebars and HOWTOsTwinned Crystals; Strange and Nearly Awful SamplesA Brief Pictorial History of Single-Crystal DiffractometryOrientation Matrix Operations; Phase Transitions & TopotaxyCredits: Exact web address of all links included in Version 1.xx6Chapter 1Apex2: A tutorial using the Ylid St

7、andard Crystal7Before placing this tutorial on the web, I let a few of my friends see an earlier version, and asked that they provide both comments and corrections. Unlike previous passes, where I obtained comments on my Space Group and High School Crystallography tutorials, I learned quickly that t

8、here were comments that corrected errors, and others that represented “l(fā)ocal approaches”. The latter ranged from some very cool things which I incorporated, to preferences that I did not wish to include, rather taking a different approach in my own laboratory. Our techniques usually result in values

9、 of 0.009-0.030 for the conventional R-factor.Sooo.when you go through this and later versions, youll find our approach to data collection and analysis. If you are an experienced user, no doubt youll see things that you do differently and do not wish to change: thats fine. You may change the raw pow

10、er point to suit your preferences, but I ask, then, that you do not distribute it without noting that you have altered slides on certain pages, as noted. Add a slide at the beginning to indicate “who says what”!All that is then in concert with the Creative Commons Attribution 3.0 Unported License.Im

11、portant : How To Use and Distribute This Tutorial8Before your instrument left the factory, the folks there collected data using the Bruker Standard Crystal (an ylid first chosen in the early days of diffractometry by Arild Christensen of Syntex Analytical Instruments). You can read more about this i

12、n a nice study carried out by Ilia Guzei of the University of Wisconsin.1 When your instrument was installed, the data set was recollected under the same conditions for comparison before the installation was certified. The Ylid data set can usually be found under the name “cust” in your structure di

13、rectory. This is a small but useful data set, with four standard “runs” (more about that later), which is collected in a relatively short time.In this Chapter we will collect data on an Ylid standard crystal, but well carry out a 24-hour data collection. It will be of some interest to briefly compar

14、e the two data collections. The present Chapter outlines the setup and execution of a “l(fā)ong” Ylid run, and, at the end, makes some observations about the differences in statistical quality of the “cust” and the “l(fā)ong” data sets. There will be “how-tos”, as well as “whys” hopefully you will find the

15、mix of pedagogy and instruction helpful. Not everything will be explained immediately, since we need to learn something of the operations first, but as we go through later examples, we will consider some of the deeper issues in diffractometry and X-ray crystallography.1Guzei, I. A.; Bikzhanova, G. A

16、.; Spencer, L. C.; Timofeeva, T. V.; Kinnibrugh, T. l.; Campana, C. F. Cryst. Growth Des. 2008, 8, 24112418.9“The Ylid” : C11H10O2SThe Man: Arild ChristensenThe Molecule: Arilds YlidThe crystal, ground to a spherical shapeThe paper (DOI link)10Well now assume that weve mounted our Ylid Standard Crys

17、tal on the instrument, and centered it in the X-ray beam.We punch the Apex2 logo on our computer near the diffractometer, and were away:11First, we select “New” on the “Sample” tab. Then this screen pops up. We pick a sensible name (a la compound, notebook number, user initials & compound, blah, bla

18、h, blah.!).A logical name for future reference, he hopes.Then we punch OK and were off.12Here, “Name” is the compound and weve already chosen, and cannot now be changed.What we put in the Compound area is optional, but if we use it, its a good place to record some short but pertinent information.Nex

19、t, its important at this point to get the formula as close to the actual cell contents as possible this can be very helpful at several key points in the structure analysispart of our upcoming Apex2 tour. Well remind you & explain why it was important later. The color should certainly be entered, alo

20、ng with any useful observations under the other three tabs. We leave the crystal dimensions blank, as well actually measure these later. Finally, the crystal shape: our standard xtal has been ground to a sphere. We often use prism, acicular (= needle-shaped), or plate, possibly accompanied by the wo

21、rd fragment if we cut the crystal, or it is obviously a broken piece.Moving down the steps on the left side, we next choose Evaluate.13The main screen contains, at the far right, two ways to proceed with the initial sample analysis. One can push the Run button, and the analysis will proceed through

22、the five steps at the far right, starting with Collect Data. Weve never done that, as we like to look carefully at each step (watch my act !) and see whether things look OK at each stage.So, lets punch Collect Data, and step through the process.14No doubt your diffractometer has some default values

23、herewhich may or may not be the same. The detector distance most folks use is either 50 mm for a platform system or 60 mm for a Kappa system. There are many other possibilities, which, applied with intelligence, are perfectly appropriate. This has been a subject of discussion on the Bruker Users Gro

24、up email list, and you are invited to join (you should!) and peruse it further. At this stage, we dont have to worry much about this, and your lab default should be just fine. In one of the advanced topic areas, as this tutorial evolves, we shall certainly give the distance idea a workout.Ten second

25、s is a good initial choice for the average xtal with an Apex2 detector. Whats going to happen is that well sweep 6 degrees in omega, 0.5 degree at a time, for a total of 12 frames. Then (watch your goniometer!) omega will reset back six degrees, and phi will rotate another 120 degrees. After another

26、 set of 12 frames, this happens one more time: at the end there will be a total of 36 frames.Now well hit the collect button at the bottom right, and were off and running!15The generator is ramping up to the preset values of 50 kV, 30 ma. Weve set a 35-second “wait time” to allow this to happen. Som

27、ething in the 20-35 second range should be fine. You can set your preferred value in BCP!When you see underlined green text, thats a link to an interesting sidebar, that you should explore sometimedoesnt have to be now!16Wowheres our first frame.really run-set 1, frame 1.See the matrix_01_0001.srfm

28、? (01 is set one, 0001 is frame 1).Its saving the files to the hard drive.Spot shape looks qualitatively very nice.17This is run-set 1, frame 518Now weve reached frame 12 on run-set 1.19Omega resets, phi rotates 120 degrees, and we press on with run-set 2. In a short while we are up to frame number

29、7. Note that the mouse cursor position gives us the value of two-theta at a particular place on each frame, here 57.4020We get twelve frames for run-set 2, then again omega resets, phi rotates 120 degrees, and we press on with run-set 3. This is frame number 2.21Run-set 3, frame 922We get the twelft

30、h frame from run-set 3, and after an exciting six minutes were ready to “harvest” reflections.Lets tug this slider over to the 5-10 rangebetter to have some of the weaker reflections in when we are attempting to determine the unit cell.Hit the Harvest button.and.23We now have 202 reflections! This i

31、s a good number. Usually, we hope to have at least 100-200.The ylid crystal has excellent resolution. Note this table, and it will help us plan our analysis later.Now, we are ready to get the unit cell punch the Index button!24Well leave the slider where it isand hit Index25Two different methods of

32、determination are used by default. The first has been selected in this case. The unit cell has three angles near 90; after refinement well consider the significance of that observation!The “HKL histogram” looks great most of the reflections index with integral indices, just as they should for a good

33、 quality single crystal.Well accept this result, and press on with refinement26Heres our starting cell from the indexing runWe click the refine button and watch the refinement proceed in the unit cell window.27Well click on the refine button a few times to refine the unit cell.The cell looks like a

34、plausible orthorhombic unit cellWell now click on the Accept button, and the select “Bravais”.28Generally, we want the highest symmetry shown with a high FOM (Figure of Merit). In a later lesson well discuss cases where this might not be appropriate.Sowe click on “Accept”29Now that we have our cryst

35、al system, we refine againclick!30Well click on the refine button a few times to refine the unit cell. As we do that, the values of a, b and c only change. The values of the angles are constrained to be 90. After that , we hit “Accept”.31Wowwhat a nice sample (ohits the Ylid, I knew that!). At 20s,

36、the resolution is 0.61 . Very nice. If we are able, with Mo radiation, we always collect data to a resolution of 0.71 , or 2 = ca. 60.Now well look at the lattice metrics using a pretty option called RLATT. Click!32The RLATT screen appears and we select the “Orientation” option under the RLATT pulld

37、own. 33Well click on the first three tabs, in turn, to view the reciprocal lattice along a*, b* and c*, respectively. Those views are shown on this and the next two slides. The richness of the pattern at this point depends upon how many reflections we picked up in our 36 matrix frames. This looks so

38、 pretty that Im even gonna include the weakest reflections, by sliding the filter over to the right.Note the axis system labels; and, everything is where it oughta be; its regular and very pretty!34Heres the view down b*; again, no spots “in the wrong place”.35Finally, here we are looking down c*. I

39、n later lessons, well look at some of these that are not so prettyWere done. Click on the X box to close RLATT.36Were back after closing the RLATT screen; now we click on Collect, and select the Knight: Data Collection Strategy37Look: in our lab, weve developed a set of runs that does a good job wit

40、h completeness and redundancy.I dont think we need the next stuffwe can skip it! Ill just import my favorite run set.Well. That can be done, but you ought at least to examine the predicted redundancy and completeness of your run set. Its quite straightforward to do thatand you will learn something.

41、Click here for an illustration.Bill BusingHugh Canduit38Generally, theres a bit of editing to do here.first, we set our desired detector distance; in our lab this is usually 60 mm.Provided the crystal is good enough, Ill set resolution to 0.71 , or, equivalently, 2 to 60 .While the screen on slide 2

42、4 suggested 5-10 sec would be adequate for a scan, Ill use 15 here, as I want even stronger signal-to-noise. First I enter 15, then click “same” to use the same time at every resolution. I really want to compare my data to the “Bruker Calibration Run” done at installation.Finally, Ill select a Custo

43、m run of 24 h, with the greatest priority on completeness, then next on time, and last on redundancy (the number of equivalent measure-ments for a given reflection).39These priorities can be set manually, but we find that the default values in the Custom mode work pretty well.Next we select Refine,

44、and start the strategy selection. Well be optimizing completeness, time and redundancy based on the parameters we just selected.In newer versions of Apex2, theres a box here that lets us even out redundancy through-out the angular rangeswe usually choose that option!40Oops! Almost forgot. The Ylid i

45、s a chiral crystal (might as well take advantage of what I know), so Ill collect the Friedel pairs by unchecking this box.Now Im ready to go!41Were off and running (refinement is 1% complete) ! At the outset, completeness is not 100%; redundancy is relatively low, and the time has not settled down.W

46、atch these two screens as we let the refinement proceed: the upper screen displays completeness and redundancy vs. angle rangethe bottom screen the two parameters vs. time.On the next several screen, watch the changes in redundancy and completeness as the refinement proceeds!423%439%4418%Note that t

47、here was little change between 9 and 18%, so Ill stop it there.45Now Ill sort those runs for completenessthat is, Id like them to be reordered such that Ill have a complete data set as early as possibleWeve converged on 100% completeness, a redundancy of 11.0, and our 24 h requested timespan.Andwe h

48、ave 14 runs46Sorting in progresswhen it completes, watch the runs jump at the bottom left!100% completeness is scheduled in an 8-10h time frame.47After sorting the runs, completeness is reached in only 6 h!So. All set, and time to set up the data collection. Click on the Experiment icon.The redundan

49、cy is really quite high. Having an orthorhombic sample makes it easy to get high redundancy. In our present example, theres a hefty taste of overkill. In real life, here in the everyday lab, wed like to shoot for a minimum redundancy of 4. That gives us good scaling and absorption corrections.48We n

50、eed to import our runs.simply click on the Append Strategy button49Wow! There they are.now Ill add an operation to turn down the generator when Im done50After the last frame is collected, Ill turn the generator down to the lowest values: 20 kV, 5 maNow Ill click Validate to make certain that all run

51、s are OKthere will be no collision conditions, etc.51All is well. Ready to go. Click on Execute!52Excellent! Heres my first frame.Run 1 frame 1.53Run 1 frame 7.Note the instrument status panel: we can check sample and CCD temperature, what the instrument is doing (writing frames just now), etc. The

52、third line contains the expected completion time.54Run 5 frame 60.Later, the data collection finishes. Then we need to get the intensities of the reflections by Integrating the frames. Click, please!55Time to integrate the framesthats the only option here. Click!56Well bring in the 14 runs now, by c

53、licking the Import button.57Andhere they are!Cool.Next, select Integration Options58ThenMore Options, please59This item will appear as 0.00. It is used in “mask” generation. E. g., the result masks out the beamstop image and other obscuration features. Ideally, you should determine it empirically (t

54、his will appear in a later version of the tutorial). With Mo radiation, we find that values of 0.45- 0.50 work well.This was in older versions of the software and should be unchecked. If you are up-to-date (AHEM!) you wont see it!Ready to rollclick OK !60Click “Start Integration” and off we go.61Ori

55、entation parameters are being refined during the “passes” that you will see initially; there will be 14 “scans”, each corresponding to one of the 14 runs. The number of “passes” depends upon how quickly the initial refinement converges.We like to look also at the “Spot position overlay” while the in

56、tegration is running. Click!This is no doubt the busiest screen youll see: many things to think about and absorb: well introduce some very basic things now, and in later lessons, expand our knowledgelooking at some of the details.62For quality crystals, the correlation between predicted and observed

57、 spot position should be as high as possible. Here were averaging ca. 0.95 !The passes have completed and integration of Run 1 has begun; were just looking at frame 66 from Run 1.Observed and predicted positions look very nice!63Integration continuesRun 1, frame 121Spot shape is very pretty ( I can

58、examine it in 9 detector regions by clicking on the 9 tabs below the image.64Integration continuesRun 2, frame 9165Integration has completed, and the last frame of Run 14 is displayed.When the integration finishes, I click in this window, and scrool up to look at the “unconstrained” cell, the comple

59、teness, and the Rsym.uhhhclick!66Here are the unconstrained cell constants. For the angles, we expect values as close as possible to 90 degrees. These look good. Ill note the cell constants, as I will want to compare their values again in a few slides.67Ive scrolled up a bit more, and am looking at

60、a useful Table. Completeness is 100%. Redundancy is nearly 18 (normally Id like 4-6 at least, but we were doing this to compare to the Factory Ylid run). Rsym is 2.7%: this is fine! And, in my last “shell” (”to 0.710”), about 74% of the data has I/(I) 2. Very nice data set.!Next, Ill go back to the