Home

HBA User Manual - Institut für Anorganische Chemie

1. June 4 2015 Chapter 2 Reference 2 1 15 Discard Initial Guess The option to discard the initial guess is available from the menu via Fit Discard initial guess The iterative refinement p 27 requires an initial guess as starting values This starting position may determine whether the iteration falls into a local minimum or finds the global minimum 5o far three different routes are available to obtain a starting guess 1 the default action if there is no user specified starting guess is to perform a grid search a grid search can also be initiated by the display of rms errors p 25 2 astarting point can be provided by the user via the data reader p 26 3 a starting point can be provided by the user via the interactive simulation p 30 In the two last cases a flag is set to indicate that initial values are provided by the user and no grid search is performed then Also the menu option to discard this starting guess is enabled 2 1 16 Iterative Refinement The iterative determination of the principal components is activated by e the menu item Fit Iterate the button bar the keyboard accelerator key I The Iteration Procedure The iterative refinement as implemented here requires an initial guess This is common to all iteration algorithms Only the way this starting guess or guesses in the case of a SIMPLEX algorithm is obtained will vary from one problem to the other If there is no user
2. PDF reader should work abandoned the 16bit line HBA32 is now hba exe If someone is desperate to use the old file names everything should work if the hba exe hba ini and hba pdf are renamed to hba32 or any name New in Version 1 5 8 07 07 2005 bug fix entering data in the Enter intensities Read Peaklist Read Integrals dialog boxes did not work caused by the transition from 16 to 32 bit Windows New in Version 1 5 1 26 01 2005 most importantly first 32 bit Windows release dealt with enabling disabling of toolbar buttons in a better way added tooltips introduced command line parameters p 33 for reading of sessions peak picking lists or integral lists etc New in Version 1 4 4 29 05 2001 started to work on calculation of fit parameter errors but this is not ready yet more important incorporated new recalculated Herzfeld Berger tables calculations took roughly a week on a Pentium 400 MHz under Windows NT 4 each spectrum required ca 1200 seconds added writing and reading of session data as ini type file Read session p 21 Write session p 21 source of intensity data is now documented in the output p 28 if information is available increased the sensitivity of the scroller bar in interactive simulation Simulate p 30 New in Version 1 3 1 15 06 1999 Upon startup HBA is automatically in contour display mode p 24 using data from the previous session if available Include
3. 3 2 5 Exceeding the Limits The experimental data can be inappropriate for use with the Herzfeld Berger tables For example sample spinning might be so slow that the parameter y exceeds the limits of the precalculated tables To protect yourself yourself from the above pitfalls This situation might be encountered if spinning sidebands of order higher than 15 are visible e Quite likely the parameter y will bounce against the borders of the Herzfeld Berger tables i e will have a value of 30 Generate stick spectra to visually inspect the agreement If your experimental data exceed the limits of the precalculated Herzfeld Berger ta bles the method of moments p 50 might be used or a direct simulation using other software 3 2 6 Axially Symmetric or Not In the literature there is a certain fascination with proving that a certain chemical shift tensor is ax ially symmetric In my opinion such proof is impossible Principal components might be degenerate or nearly so by chance it could be just a matter of going to a magnetic field strength sufficiently high to lift the degeneracy that will never be exact unless required by the symmetry about the nucleus of interest As stated by Clayden 16 part of the problem arises from the fact that the contour lines have similar curvatures for values of p close to 1 see Fig 3 7a hence run almost parallel to each other and a unique convergence is hard to establish at least w
4. 99 96 99 96 Table 3 7 The three equivalent solutions Eqs 3 16 3 18 for the principal components of the chemical shift tensor from the spectrum shown in Fig 3 10 and use the angle to solve one of the three equations for a _ 1 2 3 05 4 3 16 2 2 5 cos 120 3 17 5 3 2 5 cos dl 240 3 18 With one of the values for we solve eq 3 9 for y 15 M5 Although this procedure generally results in three different sets of values for 6 and y they end up in equivalent solutions see Sec 3 1 4 for the three principal elements of the chemical shift tensor Tab 37 54 June 4 2015 Bibliography 1 Sideband Intensities in NMR Spectra of Samples Spinning at the Magic Angle Herzfeld J Berger A E J Chem Phys 1980 73 6021 6030 DOI 10 1063 1 440136 2 Improvements in the quantitation of NMR spectra by the use of statistical methods Bain A D Fahie B J Kozluk T Leigh W J Can J Chem 1992 69 1189 1192 DOI 10 1139 v91 177 3 Press W H Teukolsky S A Vetterling W T Flannery B P Numerical Recipes in C Cambridge University Press Cambridge 1992 4 Rigorous Statistical Analysis of Errors in Chemical Shift Tensor Components Obtained from Spinning Sidebands in Solid State NMR Olivieri A C J Magn Reson Ser A 1996 123 207 210 DOI 10 1006 jmra 1996 0237 Recommendations for the Presentation of NMR Data for Publica
5. 9 distorted data 46 exit 23 export 22 features 3 fictitious data 39 fit 27 guess 26 27 Haeberlen 37 HBA 1 Herzfeld Berger 36 inappropriate data 41 INI file 32 integrals 13 intensities 12 introduction 1 iteration 27 level selector 26 license 3 limits 47 Marat 13 16 Marquardt Levenberg 27 menu items 11 meta file 22 method of moments 50 moments method 50 overview 4 peak list 13 problems 33 program parameters 33 purpose 1 radial sidebands 46 reference 11 refinement 27 report 28 revisions 5 rms 25 session 21 setup 23 32 simulation 30 spectrum settings 21 spin system 22 spinning sidebands 12 13 SpinWorks 13 16 starters 1 TopSpin 13 15 trademarks 9 troubleshooting 3 39 warranty 9 WinNMR 13 18 WSolids1 22 XML 22 XWinNMR 13 15 16 June 4 2015
6. Fig 4b in Ref 16 will yield a spectrum that contains more than three peaks and more importantly the first order spinning sidebands have intensities much lower than that of the isotropic peak you can reproduce this spectrum using HBA by using in File Enter intensities p 12 the experimental settings stated above then Fit Simulate p 30 adjust y to 3 16 and p to 0 0 In contrast to the single solution of the moment method any of the intersections of the I lg and I 41 1o contour lines represent solutions where the center peak and first order spinning sidebands will have equal intensities three example spectra corresponding to the indicated intersections in Fig 3 1 are shown in Fig 3 2 In each case the center peak and the first order spinning sidebands have equal intensities However comparison of experimental and calculated spectra reveals that the calculated spectra are increasingly poor representations of the experimental data because they contain more spinning sidebands than required You can reproduce the three cases by entering under File Enter intensities p 12 the experimental settings mentioned above and setting the intensities of the 1 CB and 1 peaks to 1 0 while the intensities of all other peaks are set to 1 this neglects these spnning sidebands during the iteration setting the initial guess at or close to the desired intersection using the Contours Data Reader p 26 and performing an i
7. Neels J Z Anorg Allg Chem 1989 576 117 130 DOI 10 1002 zaac 19895760114 31 Retrieval of solid state 9 P nuclear magnetic resonance chemical shielding parameters proposal of an approach concerning variable temperature P NMR MASS spectra of urea phosphate and comparison of different methods Lagier C M Olivieri A C Solid State Nucl Magn Reson 1994 3 163 170 DOI 10 1016 0926 2040 94 90011 6 32 Novel Method to Determine the Principal Values of a Chemical Shift Tensor from MAS NMR Powder Spectra Fenzke D Mael B Pfeifer H J Magn Reson 1990 88 172 176 DOI 10 1016 0022 2364 90 90122 P 33 Radial field sidebands in MAS Tekely P Goldman M J Magn Reson 2001 148 135 141 DOI 10 1006 jmre 2000 2215 34 The reliability of the determination of tensor parameters by solid state nuclear magnetic resonance Hodgkinson P Emsley L J Chem Phys 1997 107 4808 4816 DOI 10 1063 1 474844 35 NMR in Rotating Solids Maricq M M Waugh J S J Chem Phys 1979 70 3300 3316 DOI 10 1063 1 437915 57 June 4 2015 Index absolute magnetic shielding 36 acknowledgements 8 axially symmetric 47 background 35 beginners 1 Bruker 13 16 18 bugs 33 49 button bar 31 caveats 39 41 46 47 49 chemical shift 36 command line 33 comparison 29 contours 24 26 conventions 36 copyright 9 credits 8 Dalhousie University 1 data reader 26 disclaimer
8. 1 Getting Started This chapter covers the following topics Introduction An introduction to HBA Overviews An overview of some of the common tasks and actions Revision History Summary of changes versus previous versions Acknowledgements Credits for important contributions etc 1 1 Introduction 1 1 1 Purpose of the program HBA derives its name from the first letters of the term Herzfeld Berger Analysis 1 And it simply does just that it is aimed at assisting the NMR experimentalist in analyzing spinning sideband inten sities in magic angle spinning nuclear magnetic resonance MAS NMR spectra using the Herzfeld Berger approach to obtain the principal components of nuclear spin interaction tensors There are already several programs available doing this also so why another one The primary reason why I wrote this program is to gain some experience in iterative data fitting The Herzfeld Berger analysis could be considered as one of the simpler problems in iterative fitting because it requires only two parameters to be fitted However the nature of the data requires to use non linear least squares techniques with the additional problem of boundaries imposed on the u parameter by the size of the data tables and on the parameter p by the very definition of the skew of a tensor Also the iterative procedure is subject to the problem of local and global minima The second reason has to do with something that the other programs don t
9. 1 and of equal intensities Spectra corre sponding to the three intersections representing possible solutions are shown in Fig 3 2 40 June 4 2015 Chapter 3 Background Information zal Figure 3 2 Spectra corresponding to the three intersections of contour lines shown in Fig 3 1 For SF 80 96 MHz vr 1 0 kHz the spectra correspond to principal components of O O44 35 0 699 0 0 633 35 0 ppm O 11 52 5 699 0 0 633 52 5 ppm OW 114 3 699 74 7 633 189 0 ppm To protect yourself yourself from the above pitfalls Don t trust the reported errors blindly Look at and compare the experimental spec trum to the calculated spectrum If there is some systematic deviation such as signif icant spinning sidebands that have not been observed Fig 3 2 try to figure out the reason 3 2 2 Inappropriate Data The Herzfeld Berger method is based on the assumption that the chemical shift interaction is the only interaction affecting the intensitites of the spinning sidebands Problems may arise if the dipolar or quadrupolar interaction contribute to the line shape Actually some papers state that magic angle spinning removes the dipolar interaction and therefore the spinning sidebands can be used to extract information on the chemical shift tensor However this is a misconception MAS usually removes the effect of dipolar interactions on the line shape of the isotropic peak unless couplin
10. 46 144 6673 109480 13 19 2692 8 13000 25 160 4502 56732 7 20 2755 5 14277 38 176 2127 24902 3 Figure 2 8 A WinNMR peak list June 4 2015 Chapter 2 Reference 2 1 3 Read Session The option to read data from a previous session is available from the menu via Filel Read session This action retrieves parameters intensities fitting results etc from a previous session with HBA Note that HBA can read session files via the command line p 33 2 1 4 Write Session The option to write session data to a file is available from the menu via Filel Write session This action writes the current session parameters intensities fitting results etc to a file 2 1 5 Define Spectrum Settings Define Spectrum Settings y xi Chemical shift Intensity Xx Cancel AHE These values are only guesses it is important that you specify the proper values Reference frequency 81 02357 Spinning frequency 1280 13525 Isotropic chemical shift 2 67500 Figure 2 9 Dialog Read Integrals After using File Read integrals p 13 to retrieve the spinning sideband intensities from a file it is necessary to define the spectrum settings reference frequency the frequency of the chemical shift reference spinning rate isotropic chemical shift HBA attempts to guess these values Usually the guesses for the reference frequency and the spin ning rate could be acceptable however th
11. 85030 9 11 Absolute Shielding Scale for P from Gas Phase NMR Studies Jameson C J De Dios A Jameson A K Chem Phys Lett 1990 167 575 582 DOI 10 1016 0009 2614 90 85472 O 12 An Absolute Chemical Shielding Scale for Tin from NMR Relaxation Studies and Molecular Dynamics Simulations Laaksonen A Wasylishen R E J Am Chem Soc 1995 117 392 400 DOI 10 1021 ja00106a044 13 Conventions for the Reporting of Nuclear Magnetic Shielding or Shift Tensors suggested by Partici pants in the NATO ARW on NMR Shielding Constants at the University of Maryland College Park July 1992 Mason J Solid State Nucl Magn Reson 1993 2 285 288 DOI 10 1016 0926 2040 93 90010 K al Os Ml oo June 4 2015 Bibliography 14 Haeberlen U High Resolution NMR in Solids Selective Averaging In Waugh J S ed Adv Magn Reson Suppl Academic Press New York volume 1 pp 1 468 15 Mehring M Principles of High Resolution NMR in Solids Springer Verlag Berlin 2nd revised edition 1983 16 Chemical Shift Tensor Analyses and Simulations of Slow Spinning MAS NMR Spectra Clayden N J Dobson C M Lian L Y Smith D J J Magn Reson 1986 69 476 487 DOI 10 1016 0022 2364 86 90159 9 17 Slow Magic Angle Rotation C NMR Studies of Solid Phosphonium lodides The Interplay of Dipolar Shielding and Indirect Coupling Tensors Harris R K Packer
12. Display contour lines using the Herzfeld Berger nomograms Show rms p 25 Display rms errors as contour levels Show fit p 26 Toggle display of iteration result on off Show initial guess p 26 Toggle display of initial guess on off Show level selector p 26 Toggle display of level selector tool on off Data reader p 26 Use the mouse to read data off the Herzfeld Berger nomograms or to set an initial guess Fit Discard initial guess p Discards the user defined starting guess for the 27 iterative refinement 11 June 4 2015 Chapter 2 Reference Iterate p 27 Starts the iterative refinement of the Herzfeld Berger parameters Report p 28 Show the fitting results by calling Notepad Compare p 29 Compare the experimental and fitted sideband intensities using stick spectra Simulate p 30 Simulate the experimental sideband intensities interactively 2 1 1 Enter Intensities The dialog box Enter Intensities is available from the menu via File Enter intensities by using the button bar by using the keyboard accelerator key E Intensities and Spectrometer Settings F x Tip Use TAB or SHIFT TAB to move between edit fields Reference frequency 15 OLD 10 0 00000 5 6 44000 e1 02357 MHz 14 9 0 00000 a 12 00000 13 8 o 00000 3 10 76000 Spinning frequency 12 0 00000 7 0 68000 2 6 36000 _ 1277 00000 Hz 11 6 2 10000 1 8
13. The letter p represents high frequency positive sidebands the letter m stands for the low frequency negative minus sidebands SETTINGS This section contains entries characterizing the experimental spectrum SF spectrometer frequency actually the frequency of the chemical shift reference Rot spinning frequency Shift isotropic chemical shift Mu Rho the Herzfeld Berger parameters of the last fit rms0 rms4 rms error contour levels the colors used correspond to those of spinning sideband orders 1 to 5 COLORS All color entries specify the values for red green and blue RGB triple constituting the color Such values can be determined for example by using Windows color selection box 51 515 This section contains color entries for the spinning sidebands of different order S1 to 15 contours for sidebands of negative order have the same color as the sideband of positive order but the lines are broken EXP The color of the experimental stick spectrum CALC The color of the calculated stick spectrum CANVAS The color of the client window background color CONTBG The color of the background rectangle for displaying contours AXES The color of the axes June 4 2015 Chapter 2 Reference 2 4 Command Line Parameters Using command line parameters HBA can read on start up session files peak lists or integration files followed by iterative fitting output to a report file and quit Such opti
14. be determined The isotropic peak is determined by an analysis of the first moment It is crucial that these values are set properly You can check or modify the result using Enter Inten sities p 12 TopSpin Peak Picking Files The native file name is peaklist xml see also XWinNMR Peak Picking Files p 16 but HBA does not care If your spectrum contains several sites I recommend the following procedure perform peak picking for site 1 and save the result copy peaklist xml to peaklist1 xml for site 1 peak pick site 2 copy peaklist xml to peaklist2 xml etc This way you can make the data available to TopSpin again by copying peaklist2 xml back to peaklist xml Note that the file peaklist xml does not provide sufficient information to determine the refer ence frequency default value 1 MHz If you keep the peak lists in their native subdirectory pdata lt procno gt HBA will try to retrieve the reference frequency from the file procs in the same directory parameter SF TopSpin peak picking files are recognized by these features e an XML tag at the beginning lt xml the following root element in the second line lt PeakList gt e data are contained in attributes such as lt Peak1D F1 intensity An example of a file is shown in Figure 2 2 Note that you can highlight and copy the content of the figure box paste it into the Windows Editor save the file and read it into HBA The reference frequency wil
15. data Fig 3 5 bottom However experiments and simulations on ADP in 22 demonstrate that homonuclear dipolar interactions to neighbouring phosphate groups play an important role in the P NMR spectra and cause this apparent deviation from an axially symmetric case The crystal symmetry at the phosphorus site dictates an axially symmetric chemical shift tensor and this has been confirmed by a NIS single crystal NMR study of ADP 22 Olivieri and coworkers have subsequently shown that the inclusion of the effects of dipolar 42 June 4 2015 Chapter 3 Background Information 1 0 0 5 0 0 1 0 1 0 0 5 0 0 1 0 0 5 10 15 20 25 30 Figure 3 4 Contour lines top and error levels bottom for the spectrum and fit for the P MAS NMR spectrum of 8 VOPO shown in Fig 3 3 The best fit is obtained with y 5 393 and p 0 024 corresponding to 64 9 4 0 8 d22 10 1 0 4 033 30 5 0 7 ppm 43 June 4 2015 Chapter 3 Background Information 100 0 r F T H h I H T T T T Figure 3 5 Experimental 3 P MAS NMR spectrum of NH4H2PO obtained at 81 033 MHz and a spin ning speed of 383 Hz dis 0 81 ppm 22 Intensities obtained from peak picking are indicated Bottom comparison of experimental intensities and best fit corresponding to the fit indicated in Fig 3 6 interactions in the spinning sideband analysis allows to extract chemical shift tensor information
16. defined starting guess HBA performs a grid search automatically For this purpose it scans through the Herzfeld Berger tables using fixed increments and calculates the sg error at each grid point N Calc x 2 ve F R 2 Lal ve he p Q3 I intensity of spinning sideband or isotropic peak N number of experimental points p number of refined parameters This procedure should help HBA to find the global minimum rather than getting stuck in a local minimum During this stage HBA needs to calculate several hundred spectra Therefore this is the most time consuming step However on a Pentium processor the price paid for this safety measure seems to be inconsequential Alternatively the user can specify a starting guess p 27 interactively thereby greatly reducing the time required for the iteration Having obtained a starting guess HBA calls a Marquardt Levenberg routine to iteratively refine the required parameters using gradient methods and least squares fitting techniques 3 After finishing 7 June 4 2015 Chapter 2 Reference successfully the results p 36 are written to a scratch file Using the Report p 28 option the results can be inspected and sent to a printer or they can be displayed on the screen either on the contour map p 24 or as a comparison of stick spectra p 29 Once a fitting result has been obtained HBA explores the topography about the fit parameters in order to obtain er
17. in Fig 3 4 ssb order components up to 22 ppm compared to those of other carbons up to 4 ppm 18 Proper analysis of such spectra requires requires that the dipolar and N quadrupolar interaction are included in the simulation 19 20 Another example is the 31P MAS NMR Spectrum of B VOPO 21 shown in Fig 3 3 and the cor responding HBA analysis shown in Fig 3 4 The MAS spectrum looks normal but the systematic spread of intersections of the contour lines Fig 3 4 and the systematic deviations of calculated spinning sideband intensities from experimental data Fig 3 3 bottom indicate that there could be a problem The results in 21 demonstrate that the anisotropy in the P spectra is mainly the result of heteronuclear V P dipolar interactions The strongest indicator is the observation that the apparent principal components of the chemical shift tensor show a pronounced field dependence 21 A more heineous case is the P MAS NMR spectrum of ammonium dihydrogen phosphate ADP 22 Fig 3 5 Evidence has been brought forward that the 31P chemical shift tensor is not axially symmetric 23 24 The HBA analysis shown in Fig 3 6 confirms that the shift tensor clearly appears nonaxially symmetric the intersection of contour lines is well defined and the error surface indicates that this non zero value of p is unambiguous Accordingly the simulation of spinning sidebands shows excellent agreement with the experimental
18. of the first moment determine the principal components of the chemical shift tensor via analysis of second and third moments Starting from the high frequency end of the spectrum our first task will be to note the frequency and intensity of each significant peak as shown in Table 3 5 In the following we shall use the intensities I obtained from peak picking but the integrals could be used as well In the table we normalize the intensities of the peaks such that their sum gives 1 obtaining I norm 51 June 4 2015 Chapter 3 Background Information peak order frequency I integral I norm integral norm M Hz 1 5 21679 678 0 0007 0 1132 0 000228938 0 000218942 20000 52329 2 4 17675 780 0 0060 1 0002 0 001962323 0 001934502 16000 15396 3 3 13671 875 0 0413 6 9778 0 013507326 0 013495866 12000 52329 4 2 9667 969 0 2127 35 9616 0 069564364 0 069553862 8000 523286 5 1 5712 891 0 9870 166 9209 0 322802198 0 322844178 4000 523286 6 0 1708 984 0 8710 147 3136 0 284863946 0 284921409 0 523286238 7 1 2294 922 0 4650 78 6330 0 152080063 0 152085246 3999 476714 8 2 6298 828 0 3538 59 8367 0 115711669 0 115731045 7999 476714 9 3 10302 735 0 0934 15 7900 0 030546834 0 030539672 11999 47671 10 4 14306 641 0 0226 3 8147 0 007391418 0 007378068 15999 47671 11 5 18310 545 0 0036 0 5952 0 001177394 0 001151185 19999 47671 12 6 22314 445 0 0005 0 0755 0 000163527 0 000146026 23999 476
19. tensor In the case of an axially symmetric tensor a dyy xx will be zero and hence y 0 37 June 4 2015 Chapter 3 Background Information The Haeberlen Mehring convention is related to the Standard convention via for gt 0 i e zz 911 for lt 0 i e 7 933 611 diso 033 diso 629 dia 6 1 4 2 822 iso 8 1 41 2 633 digo 1 4 2 611 daa 0 1 79 2 June 4 2015 Chapter 3 Background Information 3 2 Possible Caveats Some of the pathological situations and their potential cure are listed below additions or comments welcome 3 2 1 Fictitious Data The major problem is that the Herzfeld Berger method and the moment method sect 3 3 will always give a result no matter how bad or fictitious the input parameters sideband intensities are an issue discussed by Clayden et al 16 Consider a hypothetical MAS spectrum obtained at 80 96 MHz at a spinning rate of 1 KHz and con sisting of the center peak at dig 0 ppm and first order spinning sidebands all three peaks having equal intensities cf the red spectra in Fig 3 2 N B there will be no chemical shift tensor that under any experimental conditions will produce exactly such a spectrum As stated by Clayden 16 a moment analysis Sect 3 3 will yield principal components of 61 19 5 22 0 0 633 19 5 ppm However a simulation using these parameters cf
20. the IUPAC rules 13 They follow the high frequency positive order Thus 61 corresponds to the direction of least shielding 011 with the highest fre quency while 633 corresponds to the direction of highest shielding 733 with the lowest frequency The isotropic values digo Or Ciso are the average values of the principal components and correspond to the center of gravity of the line shape 3 1 3 Herzfeld Berger Convention In the Herzfeld Berger notation 1 a tensor is described by three parameters which are combina tions of the principal components in the standard notation The isotropic value i e the center of gravity is the average value of the principal components The span describes the maximum width of the powder pattern June 4 2015 Chapter 3 Background Information 8 Isotropic Value isa 611 822 533 3 180 Span Q 64 633 020 Skew k 3 822 digg O 1 lt k lt 1 The skew of the tensor is a measure of the amount and orientation of the asymmetry of the tensor As indicated x is given by 3a Q Depending on the position of 622 with respect to so the sign is either positive or negative If 622 equals iso a and the skew are zero In the case of an axially symmetric tensor equals either 9 1 or 633 and a Q 3 Hence the skew is 1 The parameter u used with the Herzfeld Berger tables is related to the span of a tensor by u Q SF spinning rate The parameter p
21. 0 98 00035 6 the extended tables used by HBA since 2001 are tables that I calculated using a home made program based on this knowledge helped me to figure out how to do an estimation of fitting errors Olivieri A C J Magn Reson Ser A 1996 123 207 210 DOI 10 1006 jmra 1996 0237 Dr Michael H Frey Jeol for suggesting the use of command line parameters and for beta testing Dr William P Power performed the calculations to extend the original Herzfeld Berger tables to 15 spinning sidebands and for y to range up to 30 These extended tables were used by HBA up to 2001 Drs Judith Herzfeld and Alan E Berger kindly provided to Dr William P Power the code to generate the extended tables The original reference J Herzfeld A E Berger J Chem Phys 1980 73 6021 6030 Dr A Kentgens kindly provided the program HBLSQFIT which we used prior to HBA and which also employed Bill Power s extended tables The contouring algorithm is relatively simple but works sufficiently well It follows the BASIC subroutine CONREC written by Paul D Bourke BYTE 1987 6 143 150 The algorithms for bicubic spline interpolation and Marquardt Levenberg non linear least squares procedures are adopted and adapted from W H Press S A Teukolsky W T Vetterling B P Flannery Numerical Recipes in C Cambridge University Press Cambridge 1992 Microsoft for providing Visual C 2008 Express Edition for free Jordan Russell for maki
22. 0000 Isotropic chemical shift 2 67500 ppm t J E eo N n CB 812 34 mu 4 6 3 06000 11 0 39000 2 7 3 16000 12 0 0000 3 8 2 72000 13 o 00000 S th fi 64000 Uk jo ooo00 X 2 Hep 5 1 40000 10 0 06000 15 0 00000 Figure 2 1 Dialog Enter Intensities Purpose This dialog box serves to enter the intensities of the spinning sidebands and of the center peak manually It also contains the parameters defining the spectrum settings ie reference fre quency spinning rate and isotropic chemical shift One can also use this dialog box to check the result from parsing integration or peak picking p 13 files Units for the intensity data are irrelevant because all data will be normalized with respect to the intensity of the center peak The intensities might have been obtained by using a ruler and measuring the peak heights using a peak picking listing with intensities or by integrating the experimental spectrum June 4 2015 Chapter 2 Reference Convention The convention used in the HBA program for numbering the spinning sidebands is indicated by the picture It follows the direction of the frequency axis i e spinning sidebands to high frequency of the center peak have positive ordinal numbers spinning sidebands to the low frequency side of the center peak have negative ordinal numbers note th
23. 2 Show Fit Show Initial Guess While displaying contour lines one can toggle the display of initial guess and fit by the menu item Contours Show initial guess menu item Contours Show fit While viewing the contour lines one can toggle the display of the position corresponding to the data of the initial guess or the fit on and off The colors of the points can be customized using the File Setup p 23 dialog box 2 1 13 Show Level Selector Tool While displaying contour lines using this tool one can customize the amount of information displayed The Hide All button is useful to quickly erase the list lt lt X X I 1 X4 3 4 5 6 F 8 AIII XI 2 9 2 1 14 Data Reader Bec The mouse data reader can be activated by using the US menu item Contours Data reader button bar While displaying the spinning sideband intensities as contour lines one can select the mouse cursor as a data reader For any point within the range of the Herzfeld Berger tables the status bar will display the values of 1 0 for the current mouse cursor position Also using the parameters describing the spectrum settings p 21 the values of the principal components of the chemical shift tensor will be calculated One can use this data reader to set an initial guess p 27 for the iterative refinement p 27 by pressing the left mouse button at the desired position over the contour plot 26
24. 25 In my opinion you should be able to recognize that there is something not right in the case of the analysis of the P MAS spectrum of B VOPO However the only indication that something is wrong in the case of the P MAS spectrum of ADP is the discrepancy between expectations based on the X ray structure and the non axial symmetry determined in the analysis Note that the absolute errors in the principal components may not be good indicators that something is wrong they are less than 1 ppm in both cases In the case of B VOPO the error ratio Ryarc conp Eq 2 5 determined for y and p 3 59 might be a better indication the ratios are 1 37 for ADP Here I need more experience in assessing these parameters The conclusions of this section are quoted from Olivieri s paper 25 It should be noted that all anisotropic interactions affecting the nucleus of interest will exert an effect on the intensities of the side band manifold The most ubiquitous is the dipolar coupling both direct and anisotropic indirect In principle therefore the study of the side band intensities will give useful results when all anisotropic interactions other than that arising from the chemical shift tensor have been removed or are negligible in comparison with the latter To protect yourself yourself from the above pitfalls Check the contours they may give an indication of how bad or inappropriate the input data are Generate stick spectra
25. 5 3956 984 003 39 767 17 256 15 5858 2843 004 12 753 4 356 33 2712 6069 005 9 759 28 669 54 0414 9858 006 35 872 53 882 63 9669 11668 007 57 484 79 095 69 1362 12611 008 81 797 105 209 81 0864 14791 009 108 811 131 323 78 8533 14384 010 133 124 154 735 74 2664 13547 011 156 536 178 147 63 9885 11672 012 183 550 203 361 48 4859 8844 013 206 062 226 773 41 7548 7617 014 231 275 251 986 42 5950 7770 015 258 290 277 200 46 8464 8545 016 282 602 300 612 41 5667 7582 O17 306 015 326 726 32 3039 5893 018 331 228 350 138 37 6363 6865 019 356 441 374 451 36 5269 6663 020 382 555 398 764 27 6157 5037 021 405 067 423 977 29 5102 5383 022 429 380 450 091 32 2221 5878 023 455 494 473 503 28 3826 5177 024 480 707 498 716 22 2018 4050 025 505 920 522 129 21 5916 3939 026 530 233 547 342 26 3894 4814 027 554 546 572 555 28 4303 5186 028 579 759 596 868 24 2943 4432 029 604 972 622 081 16 9160 3086 030 631 987 644 593 28 9432 1631 031 656 299 668 906 4 2139 769 032 678 811 690 518 1 0642 194 Figure 2 6 A SpinWorks integrals file peaks txt place the mouse cursor on the desired peak and hit the right mouse button a peak label will appear on the screen e select Peak Pick List to confirm that your peaks ended up in the file peaks txt Note that in the case of a peak picking list the file peaks txt contains sufficient information to determine the reference frequency Ther
26. 6 0 007391418 1 182626897 189 2203035 30275 24856 5 0 0036 0 001177394 0 235478807 47 09576138 9419 152276 6 0 0005 0 000163527 0 039246468 9419152276 2260 596546 Sum 3 0576 1 000000 0 005232862 2885 086342 74231 29252 Table 3 6 Contribution of each spinning sideband and the center band in the spectrum shown in Fig 3 10 to the first second and third moments Herzfeld Berger approach more robust because only some of the spinning sidebands are required rather than the complete lineshape as in the case of the moment analysis The last row of the table lists the second and third moments for the total lineshape the values of which are used in combina tion with the isotropic chemical shift to obtain the three principal components of the chemical shift tensor Solving the Cubic Equation Given the second and third moments we need to solve for the reduced anisotropy the reduced cubic equations created by the combination of eqs 3 9 amp 3 10 under elimination of the asymmetry of the chemical shift tensor a b 0 3 12 with 1 M 3 13 b M3 3 14 Using a trigonometric solution we compute the value of the angle in the expression b cos 2 3 15 June 4 2015 Chapter 3 Background Information solution 1 solution2 solution 3 83 145 116 63 33 489 y 1 806 0 4260 5 9655 Siso 16 67 16 67 16 67 11 99 82 99 83 99 82 635 50 16 50 14 50 16 633 99 96
27. 71 y 3 0576 517 0324 1 000000000 1 000000000 Table 3 5 Frequencies and intensities or integrals of spinning sidebands and centerband in the spec trum shown in Fig 3 10 The M is calculated according to Eq 3 11 with respect to the peak in that row as center band Determination of Isotropic Peak Then we start the trial by assigning one of the peaks as center band and calculating the first moment about that peak according to the following equation N V ixwxl M Mi Hz 3 11 where the sideband order runs from the spinning sideband at lowest frequency of order M to the spinning sideband at highest frequency of order N Sum up all the products of order times intensity I of that peak to obtain M about that trial center peak Now take the next peak at higher or lower frequency to this trial peak and calculate the first moment for that new trial assignment If the absolute value of that new first moment is smaller than the old value then you should keep going into that direction until the magnitude of the first moment starts to increase again Now you have located the position of the isotropic peak The solution for the example shown in Fig 3 10 is already indicated in Tab 3 5 for each peak the last column gives the value of M calculated for the spectrum about that peak If peak 6 is taken to be the center band the first moment has the smallest magnitude It is not exactly zero because we have neglected some smaller peak
28. 77 0005 3163 533 19 5306 0 7031 1865232 0006 7162 284 44 2176 0 8495 2253664 0007 11163 056 68 9171 0 8651 2295016 0008 15163 274 93 6131 0 9931 2634531 0009 719163 879 118 3116 1 0311 2735493 0010 23164 447 143 0098 0 9557 2535396 0011 27159 128 167 6717 0 8196 2174433 0012 31163 564 192 3938 0 6308 1673497 0013 35169 052 217 1223 0 5281 1401075 0014 39162 846 241 7787 0 5306 1407680 0015 43161 960 266 4680 0 6137 1628051 0016 47166 051 291 1879 0 5308 1408305 0017 51163 796 315 8687 0 3954 1049017 0018 55163 414 340 5611 0 4976 1320201 0019 59166 330 365 2738 0 4795 1272095 0020 63163 477 389 9509 0 3529 936114 0021 67162 918 414 6421 0 4038 1071376 0022 71168 260 439 3698 0 4337 1150522 0023 75163 613 464 0358 0 3727 988660 0024 79165 947 488 7449 0 2965 786614 0025 83166 840 513 4451 0 2902 769786 0026 87165 281 538 1302 0 3489 925547 0027 91167 412 562 8381 0 3997 1060454 0028 95170 999 587 5549 0 3405 903275 0029 99167 878 612 2304 0 2239 594128 0030 103167 578 636 9232 0 1384 367095 0031 107172 653 661 6493 0 0690 183144 0032 111165 095 686 2973 0 0267 70949 E wsolids TopSpin cpniphms 1 Figure 2 5 A SpinWorks peak picking file peaks txt Y June 4 2015 Chapter 2 Reference SpinWorks integral list for E wsolids TopSpin cpniphms 1 peak start ppm end ppm intensity intensity Absolute 001 88 393 71 284 1 0000 182 002 62 279 46 071
29. 7v Hz L ik x L 3 3 i M N M Hz k v Hz y Kx l 3 4 i 2 M k N June 4 2015 Chapter 3 Background Information us VIS lt gt s65 0 lt 65L0 0 lt ETTO lt TOO T lt LL6 9 lt gt y 8z LT lt l 88 991 T T t T T T T T T T H T i T T T PPM 200 160 120 80 40 0 40 80 120 160 200 240 Figure 3 10 Theoretical MAS NMR spectrum calculated for a chemical shift tensor of 911 100 622 50 633 100 ppm at an observation frequency of 100 MHz and a spinning frequency of 4000 Hz where v is the spinning frequency in Hz vo the frequency of the chemical shift standard and the sideband order i runs from the spinning sideband at lowest frequency of order M to the spinning sideband at highest frequency of order N and its intensity I is normalized such that pes 3 6 For the first few moments Maricq and Waugh have shown 35 Mo 1 total intensity normalized to unity 3 7 M 0 the first moment about the isotropic shift is zero 3 8 Mp 67 15 3 47 3 9 M3 26 35 1 n 3 10 where is the reduced anisotropy and y the asymmetry parameter defined according to the Hae berlen convention see 3 1 4 3 3 2 Application In Fig 3 10 we show a calculated MAS spectrum that shall be used to illustrate the application of the method of moments to determine the isotropic peak via analysis
30. 831 44 66 7372 5407 29 26 96 1 639402E 006 46 3981 3759 34 27 0634 2192 77 49 30 2 997328E 006 6 7242 544 82 9 8484 797 9b 22 55 1 370712E 006 32 4474 2629 01 47 0113 3809 02 6 71 4 078642E 005 69 1081 5599 39 86 9362 7043 88 22 75 1 382878E 006 107 7776 8732 52 123 5969 10014 26 10 13 6 160914E 005 146 4470 11865 66 158 7510 12862 57 4 91 2 983119E 005 184 8654 14978 45 196 1649 15893 98 1 04 6 309909E 004 Figure 2 7 A WinNMR integration file Note that in both cases peak picking list or integrals the file contains sufficient information to deter mine the reference frequency Therefore you can keep the file in any location you like separate from the rest of the spectrum WinNMR peak picking or integrals files are recognized by these features e at the beginning Peak List or Integral List An example of a peak picking file is shown in Figure 2 8 and a integrals file in Figure 2 7 If you don t have WinNMR but want to use peak lists it is relatively straightforward to create such an ASCII file because HBA does not care about the prosa All it wants for each spinning sideband is a row containing two integers specifying the peak number and data point followed by four float ing point numbers specifying the frequency in Hertz chemical shift in ppm intensity and relative intensity in per cent HBA actually uses frequency ppm and intensity If you want to use integrals HBA requires Integral List as the fi
31. File Open WSolids Document 3 If the spectrum consists of several sites read additional spin system files and tell WSolids1 to add to the spin system press the No button on the corresponding dialog box 4 After all spin systems have been read read an experimental spectrum if available File Open Spectrum or create a spectrum window File New window 5 So far HBA does not provide information on the observed nucleus fix this before performing a calculation Simulation Default parameters 6 Perform a calculation If necessary change the convolution parameters 22 June 4 2015 Chapter 2 Reference 2 1 8 Setup x Contour Color Select Item Color Select 1 BEN 7 01 exp spectrum S z NENNEN _ calc spectrum MA NENNEN canvas EA T MEN contour backgr p NEN axes NENNEN W p NENNEN fit point 7 _ guess point A g B 9 ma RMS contour levels at UU rms1 0 008 rms4 0 030 1 MEN 12 BEEN rms2 0 010 rms5 0 050 13 mm rms3 0 020 1 BENE 3 15 mz 8 P Save settings as default MESES Figure 2 10 Dialog Setup The option to customize colors and set a few parameters is available from the menu via File Setup The dialog box allows to customize the colors of the contour lines and several other data items used by HBA Also the levels for the dis
32. HBA Herzfeld Berger Analysis USER MANUAL Klaus Eichele June 4 2015 Contents 1 Getting Started 1 1 1 2 1 3 1 4 Gone 1 cr 111 Purpose of the program 2o one kd don 64 be yo n dent gi ted VI VOV PTT labo A case ates koe eX euo x ue ems Se ES onu eg E eds 114 Troubleshoc pg ccoo nd e m n eee ERA Eee n Ep udt go cl Av DT REVISION PISEOPV ka du x ee UR kai A S dee eq x ROO Se hee d Essi wciccpundl rr Uu s so pP 142 Trademarks 2 cocco et Ao deine E ef dee etes i Sat ade dao Copyright dorado oi u mu ua k ae au 144 Disclaimer or Wananby coce a sa RR RIPE Was oP EES k 2 Reference 21 T2 23 24 2 5 31 3 2 Helpon Menu Eme is i ea a e std e das ALI Enter Intensities 2220 59st Rom kk Re a e e wey u tks Re dodo LLLI Read Me DOSE Lagi um esi 5 8 RU E Rane wae Bee nr mm eg 21 3 Regd Sess cea v6 ee koe aa RH X x ee 3303 kk us 214 WIKI io rar Se Pb E eA Eher e Eo OY u ak SUR OE pdat 21 5 Define Spectrum Settings ca ske a Grd He R REOR sva d Rer ie Cea 216 We Meta File 212 2 9239 79493 b 5 b EWA DNO udi Zl JVuebpmn5ysleWb 22s 4 d la a RUE A es CARA o A rrr AI 2110 Show Contours s sso a ke mS RE E Y 9 o doe Rs 2111 Show RMS ndis eR edo doe deb a d na eL HG Se ee we UY S 2 112 Show Fit Show Initial Guess c eee 2 1 13 Show Level Selector Tool 0 2 2 0 0 ec u k k s k eee 2144 Data Reader 4 2229 a a Ra Oe oy e eco 2113 Disc
33. K J Thayer A M J Magn Reson 1985 62 284 297 DOI 10 1016 0022 2364 85 90061 7 18 BC High Resolution Nuclear Magnetic Resonance Studies in Polycrystalline Tetracyanoquin odimethane Nunes T Vainrub A Ribet M Rachdi E Bernier P Almeida M J Chem Phys 1992 96 8021 8025 19 13C Chemical Shielding Anisotropy in the Solid Phases of CH3 CN Sepa J Gorte R J Suits B H White D Chem Phys Lett 1996 252 281 286 DOI 10 1016 0009 2614 96 00156 X 20 The Effects of Interplay between Quadrupolar Dipolar and Shielding Tensors on Magic Angle Spinning NMR Spectra Shapes of Spinning Sidebands Davies N A Harris R K Olivieri A C Mol Phys 1996 87 669 677 DOI 10 1080 00268979600100461 21 Phosphorus 31 and vanadium 51 solid state nuclear magnetic resonance spectroscopy of B vanadyl phos phate Effects of homo and heteronuclear spin spin electrostatic and paramagnetic interactions Eichele K Grimmer A R Can J Chem 2011 89 870 884 DOI 10 1139 V11 025 31P NMR Study of Powder and Single Crystal Samples of Ammonium Dihydrogen Phosphate Effect of Homonuclear Dipolar Coupling Eichele K Wasylishen R E J Phys Chem 1994 98 3108 3113 DOI 10 1021 j100063a012 Structure and Cation Effects on Phosphorus 31 NMR Chemical Shifts and Chemical Shift Anisotropies of Orthophosphates Turner G L Smith K A Kirkpatrick R J Oldfield E J Mag
34. Windows Meta File format can be selected from the file menu using the Meta file p 22 menu item The simulation results can also be written to a file for later use as spin system file p 22 in WSolids The session data can also be written to a data file p 21 for later retrieval continuation 1 3 Revision History This page summarizes the changes made compared to previous versions of HBA Experienced users may use this information as a quick update on new program features New in Version 1 7 5 17 04 2015 New feature the location of the INI file p 32 has changed in the pograms directory for the portable version or in the user s documents directory New feature added support for PDF XChange Editor by Tracker Software Products Ltd http www tracker software com as PDF viewer to display context sensitive help Bug fix the relative scaling of experimental and calculated stick spectra did not always work properly New in Version 1 7 3 23 08 2012 New feature when selecting Report p 28 the result file is now written in the users documents path New feature reading of peak picking and integrals files got combined into one function Read Integrals Peak Lists 13 HBA itself figures out the file format Also TopSpin native peak lists and SpinWorks peaks and integrals can be retrieved New feature included Olivieri s error treatment in the iteration p 27 the normalized inten sities of spinning side
35. also available in the literature in Fig 3 of Clayden s paper 16 and in Fig 6 of the Herzfeld Berger paper 1 the N 1 contour does not pass through the intersection of the other contour lines the presence of a saddle point could also be part of the problem Fortunately Tekely 33 was also observant and actually had the math and physics to unveil the cause of this phenomenon This intensity aberration is caused by radial radio frequency components at the edge of the coil and may especially affect the 1 spinning sideband Interestingly the intensity dis tortion is independent of the spinning frequency and this provides a test as to the origin Also TOSS total suppression of spinning sidebands is not able to suppress this effect and its intensity aberra tion survives this sequence Note that Tekely and the two examples quoted above use a shielding convention to label the spinning sidebands i e their 1 spinning sideband is to the high frequency side of the isotropic peak To protect yourself yourself from the above pitfalls Check the contours is the 1 contour line off the intersection of the other contours Note that you can exclude the bad sideband intensity from fitting by setting its inten sity to 1 The radial sidebands are caused by sample molecules at the edge of the coil restrict ing the sample to the center of the coil should remove radial sidebands June 4 2015 Chapter 3 Background Information
36. ard frai GUESS os ceana aaaea Re ee See ded we 0 2 1 16 Iterative Refinement a ES E eena 4939 ede EU aile dock enu A 2115 Compas RARER 2 ite vu eee dE Te A up SERS SEA aW e EX as 2 19 Simulate Interactvel 222699 SUS A S ARE eS 6S LT Helpon Baton Bar Heime E E RR wk 4 ER ee ue um SS eod o colour rd ed RE RR Re Ro oa T A ee ae ae ars Command Line Parameters uu u u k k k k k k k k k c Known Problems 224222229 stu si nuk Background Information LOWER ca 6 we ROR RAS ER Ow ERASER we OR e da ld Sis l eE a T ee AAA a 31 2 Standard Conventio ous l4 ua kno RUE ER EY X o DA E a us 313 PIeutekbBergerionvenlon 42 nn v b V da Gr ee E E Rs 3 1 4 Haeberlen Convention lt ees s e s oe o Y ad ti Possible Caveats 22s nux bbe een cee x RO dea boa GO S Xo e eee cd 32 1 PicH ousDabhb cras 6 c ac s da baa dkk iii June 4 2015 NO 6 D GO GO U E LUA UA UA A LE Contents S2 Inmappip ste Date cos s deum ee Rege o Sew ee end 41 S23 Distorted Data 222 222 eaa doe a Rem E mm RERO ro e m Gta 46 S24 Radial Sidebands 22 RR xc d a ar ed coe a Ee ea wa oxo3 46 peo Exes tie Lite 664 0 k oe A s b E SUR xU 47 325 Axially Symmetricor Not 6056 cos ee or E n 47 227 MalceousSolw tf 2222 REGE B SE RUE GE ek a 49 33 Method of Moments 2 2242659Rx Gc Roc RR Ek o E X Eos omo Ek 50 CHER T DP 50 SOL APPO LITT 51 Bibliography 55 Index 58 June 4 2015
37. at the Herzfeld Berger paper uses the opposite convention because they use a magnetic shielding convention instead of a chemical shift convention The absolute value of the ordinal number increases with the separation of the spinning sideband from the center peak Clear Button By default HBA loads the intensities used in the last session from the HBA INI file Use the Clear button to reset the contents of each field This action will set the intensities of all spinning sidebands to 1 and the intensity of the center peak to 1 Neglecting Spinning Sidebands Use a negative intensity to indicate that a spinning sideband should be neglected in the iterative calculation This option is not allowed for the center peak Neglecting sidebands may cause HBA to fall into a false local minimum One might be better off guessing the intensity of a sideband than in neglecting it 2 1 2 Read Integrals Peak Lists The option to read an integration file is available from the menu via Filel Read integrals peak lists by using the keyboard accelerator key P the command line p 33 The purpose of this action is to retrieve peak position intensity data pairs from a specified file This file may contain integrals or a peak picking list produced by any of the programs detailed below In older versions of HBA pre 1 7 the user had to preselect the kind of file integrals or peak picking then specify the file name Current versions of HBA will try
38. ative refinement To disable this initial guess select Discard initial guess p 27 June 4 2015 Chapter 2 Reference 2 2 Help on Button Bar Items Button Menu Item Action Alius Enter Intensities p 12 Show Contours p 24 Data Reader p 26 Iterate p 27 Show rms p 25 Compare p 29 Report p 28 Simulate p 30 Help p iii Enter intensities of spinning sidebands and center peak all following actions are based on these data Display the contour lines in the Herzfeld Berger tables using the normalized intensities Use the mouse cursor to read data off the contour diagrams or to set an initial guess for the iteration Starts the iterative determination of the chemical shift tensor parameters Show the rms error as contour levels Based on the results from the iterative determination experimental and calculated stick spectra are compared Report the results of the iterative determination of chemical shift tensor parameters Simulate the experimental data interactively Opens the HBA help file with the Contents page 3 June 4 2015 Chapter 2 Reference 2 3 INI File Settings Several of the items of the INI file can be modified via the Setup p 23 item of the File menu The HBA INI file contains the following sections INTENSITIES This section contains the spinning sideband and centerband intensities used in the last session with HBA
39. bands and center band are now fitted instead of fitting the ratios New feature added support for other PDF viewers to display context sensitive help added handling of PDF XChange Viewer by Tracker Software Products Ltd http www tracker software com currently my prefered PDF viewer and Sumatra PDF Viewer http blog kowalczyk info software sumatrapdf also deal with the new Adobe Reader X Modified the Background Information p 35 section in the documentation got expanded caveats alternatives New in Version 1 6 14 27 05 2010 The result of the iteration can now be written as spin system file p 22 for later retrieval in WSolids1 New in Version 1 6 12 02 01 2010 Added handling of TopSpin integrals p 13 and peak lists p 13 Added command line parameters p 33 to supply information on spectrometer reference fre quency spinning frequency and isotropic chemical shift New in Version 1 6 9 06 01 2009 switched to a different development platform Microsoft Visual C 2008 Express Edition the user interface has thus changed a little introduced a level selection tool p 26 for the contour display this allows to select deselect the levels to be displayed the documentation help is now in Adobe Acrobat PDF format for context sensitive help from HBA Acrobat Reader or Acrobat Professional is required Just for viewing the PDF file any June 4 2015 Chapter 1 Getting Started
40. d a Setup dialog p 23 where the user can customize color settings directly instead of editing the INI file p 32 In contour display mode p 24 display of initial guess p 27 and iteration result p 27 can be toggled individually New in Version 1 2 12 10 10 1998 In reading peak picking p 13 or integral p 13 files the estimation of the isotropic chemical shift is done via analysis of the first moment Fixed a few bugs New in Version 1 2 10 09 04 1998 The result of the iteration can now be written as spin system file p 22 for later retrieval in WSolids 2 0 The rms errors p 25 can be displayed as contour plots This required that the grid calculation uses a finer step size INI p 32 file contains some additional parameters The menu got rearranged a little New in Version 1 2 9 10 03 1998 The default extension for peak picking p 13 and integral p 13 files is now set to TXT as produced by WinNMR The format of the output p 28 file changed The LOG entry in the first line forcing Notepad to append date and time got removed Date and time are now given in the file header Derived tensor parameters using different conventions p 36 are also listed The initial guess p 27 is now shown in the contour plot as a small circle The About dialog box shows now detailed version and system information The dialog box for input of intensities got changed in layout The help file documents INI fil
41. do Herzfeld and Berger 1 suggested that their method could be used as a graphical approach to the problem of determin ing the tensor parameters During the first few years of my dabbling in solid state NMR I used to analyze hundreds of MAS spectra by drawing and tracing contour lines on transparencies and de termining the intersections Certainly tedious but also quite instructive With the aid of a computer this can be done in a much pleasanter way HBA is a simulation package developed during my stay at the Department of Chemistry Dalhousie University Halifax Canada The initial version has been written in C using Borland C 4 52 and has been developed to run on IBM compatible personal computers under the MS Windows en vironment in the 16 bit subsystem For the 32 bit version Borland C 5 0 has been used initially The current version has been produced using Microsoft Visual C 2008 Express Edition HBA is actually a spin off of a larger package WSolids aimed at simulating solid state NMR spectra for a variety of nuclear spin interactions and experimental conditions zs June 4 2015 Chapter 1 Getting Started Figure 1 1 Dalhousie University chemistry department E v d s Mis YN IF Z mah i ut Figure 1 2 Dalhousie University administration building and faculty club left S eo June 4 2015 Chapter 1 Getting Started 1 1 2 Features Extended Herzfeld Berger tables allowing spinning sidebands
42. e p 32 settings June 4 2015 Chapter 1 Getting Started New in Version 1 2 5 06 08 1997 Added a menu item File Meta file p 22 to create a meta file containing the graphics shown on screen New in Version 1 2 4 05 08 1997 Added entries to ini file which allow customization of background and control colors colors for contours were available previously New in Version 1 2 Added reading of WinNMR 5 0 5 1 integration p 13 or peak picking p 13 files ASCII Added a data reader p 26 to the contour display one can set an initial guess for the iteration using this option Added interactive simulation of the experimental spectrum one can also set an initial guess for the iteration using this option New in Version 1 1 Only some minor improvements in the display 7 June 4 2015 Chapter 1 Getting Started 1 4 Acknowledgements This section covers the following topics Credits Credits for important contributions b Trademarks Trademark acknowledgements cu Copyright The obligatory copyright statement Disclaimer The obligatory disclaimer message 1 4 1 Credits Dr Alejandro C Olivieri kindly provided BASIC programs that taught me how to calculate spinning sidebands in the frequency domain Olivieri A C Solid State Nucl Magn Reson 1997 10 19 24 DOI 10 1016 50926 2040 97 00023 4 Olivieri A C Solid State Nucl Magn Reson 1998 11 181 187 DOI 10 1016 50926 204
43. e write report to a file and quit hba exe s e nmr projects phosphate ini f r f my results phosphate_fit txt q 2 5 Known Problems Some programs are not happy with the metafiles produced by HBA June 4 2015 3 Background Information Background information might be available some day on the following topics depending on whether someone is actually going to read this Chemical Shift Anisotropy CSA What is CSA Conventions p 36 What is sample rotation What are spinning sidebands Herzfeld Berger Analysis What are the Herzfeld Berger parameters How does the Herzfeld Berger method work What are possible caveats in working with the Herzfeld Berger Method p 39 What are the alternatives p 50 June 4 2015 Chapter 3 Background Information 3 1 Conventions Unfortunately there are many different conventions for labeling the principal components of chem ical shift tensors around in the literature Most of the conventions have advantages for certain sit uations but drawbacks in others Often it is not obvious which convention has been chosen The collection given here attempts to summarize some of the most frequently used conventions 3 1 1 Shifts and Shieldings It is recommended that the IUPAC conventions 5 6 are obeyed The absolute magnetic shielding c in ppm is the difference in shielding between the fre quency of the bare nucleus 1 1 and the frequency of the same n
44. e isotropic chemical shift might be off because it is deter mined by an analysis of the first moment This analysis relies on the assumption that all spinning June 4 2015 Chapter 2 Reference sidebands of significant intensity are available Enter the correct value or double click with the left mouse button in the list box on the peak representing the center peak 2 1 6 Write Meta File The option to create a graphics file in Windows meta file format is available from the menu via Filel Write meta file Selection of this menu item allows to write the graphics content of the client area to a file using the Windows meta file format WMF This format can be read by a variety of graphics programs and further processed there The user will be asked to specify a file name It is recommended to use the WMF extension Possibly the labels of the p axis are missing and the font size for the y axis labels will be too small 2 1 7 Write Spin System The option to write chemical shift tensor information as WSolids spin system file is available from the menu via File Write spin system After an iteration the chemical shift tensor information can be written in XML format to a WSolids compatible spin system file for later display of simulated spectra In order to use this file in WSolids1 follow this procedure 1 In HBA save the fitting results as WSolids spin system file 2 In WSolids1 read the created spin system file
45. e the experimental data by using the simulation interactively feature The Herzfeld Berger parameters are simply modified via two scroll bars and the effect is shown immediately as comparison of stick spectra Simulate p 30 This feature sets also an initial guess p 27 for the iterative refinement p 27 Enter The iteration procedure requires an initial guess p 27 on the parameters to initial be refined There are three ways to obtain this guess uess E the default action a grid search is performed automatically if there are no user defined starting values selection of rms error p 25 display also performs a grid search and updates the initial guess if no iteration of the data has been performed one can use the data reader p 26 one can use the interactive simulation p 30 Iterative Starting from an initial guess a Marquardt Levenberg algorithm is employed refinement to find the best fit p 27 to the experimental data The result can be examined by using the contour display p 24 the comparison p 29 of experimental and calculated intensities the display of rms errors p 25 as contours the text report p 28 file generated by the fitting procedure 4 June 4 2015 Chapter 1 Getting Started Output Output of the fitting results is achieved by the report p 28 action which fitting opens the text file generated by the fitting procedure result Creation of a graphics meta file
46. efore you can keep the file in any location you like separate from the rest of the spectrum In the case of integrals the file peaks txt does not provide sufficient information to determine the reference frequency default value 1 MHz If your spectrum is a TopSpin data set and you keep the integrals file in the location natural to SpinWorks the subdirectory where the fid is located HBA will try to retrieve the reference frequency parameter SF from the file procs in the pdata 1 subdirectory SpinWorks peak picking or integrals files are recognized by these features e atthe beginning SpinWorks integral or SpinWorks peak An example of a peak picking file is shown in Figure 2 5 and an integrals file in Figure 2 6 WinNMR Peak Picking or Integrals Files To generate a peak picking or integrals file with WinNMR perform peak picking or integration for all spinning sidebands belonging to one spinning sideband manifold Display the results using the Report button and save the result as an ASCII file using a name and location at your convenience 18 June 4 2015 Chapter 2 Reference Integral List Date 25 04 2007 Time 18 31 File Name D nmr aa KOOPH2BU 002001 1R Integration regions saved in File D NMR AA KOOPH2BU 002001 TXT Start ppm Start Hz End ppm End Hz Integral norm Integral abs 154 6224 12528 06 144 3273 11693 91 1 00 6 079783E 004 118 9661 9639 06 104 9045 8499 74 5 23 3 181884E 005 84 3143 6
47. ependency with other refined parameters Note that realistic errors could be greater than the reported standard deviations My estimation is that likely errors are on the order of 1 2 ppm or up to 5 of the span whichever number is greater 2 1 17 Report The output of the fitting results can be inspected by using the menu item Fit Report the button bar the keyboard accelerator key R At the end of the iteration p 27 procedure the results p 36 are written to a scratch file From version 1 7 3 on the file is named hba_iterate txt and is located in the users My Documents path under subdirectory WSolids Right now the Report option just calls Notepad with the output file From Notepad one can modify the output file to include a descriptive title or save the results under a different file name or print the results The experimental sideband intensities are also saved in the HBA INI file to serve as defaults for the June 4 2015 Chapter 2 Reference next session If one is interested in keeping the data one could copy the ini file to a different file name 2 1 18 Compare Results The display of stick spectra representing experimental and calculated intensities is activated by the menu item Fit Compare the button bar the keyboard accelerator key C Herzfeld Berger Analysis 15 10 0 5 Aj 15 Figure 2 13 Display of experimental and theoretical s
48. g to a quadrupolar nucleus is involved but it does not remove their effect on the intensities of the spinning sidebands If on top of that also indirect spin spin coupling is involved that separates the subspectra associated with different spin states of the coupled nucleus into discrete spinning sideband manifolds one can actually analyze the subspectra to obtain information on the chemical shift tensor and the dipolar interaction as well as their relative orientations The theory involved has been outlined by Harris and coworkers 17 In contrast if the presence of a dipolar interaction is not taken into account properly the resulting principal components of the chemical shift tensor may have large errors One example is the carbon of a cyano group here the spinning sideband intensities in the IC MAS spectrum are also affected by the dipolar interaction with the directly attached 4 N nucleus It has been noted that a Herzfeld Berger analysis of the spinning sideband intensities results in relatively large errors in the principal June 4 2015 Chapter 3 Background Information 100 0 PPM 50 40 30 20 10 0 10 20 30 40 50 60 70 80 al FE Figure 3 3 Experimental P MAS NMR spectrum of B VOPO obtained at 161 974 MHz and a spin ning speed of 1200 Hz dis 10 4 ppm 21 Intensities obtained from peak picking are indicated Bottom comparison of experimental intensities and best fit corresponding to the fit indicated
49. ger analysis 1 js corresponds approximately to the number of visible spinning side bands and is therefore spinning rate and field dependent while p corresponds to the skew axiality of the tensor and is field and spinning rate independent 24 June 4 2015 Chapter 2 Reference 2 1 11 Show RMS The contour lines of the rms errors can be displayed by using the menu item Contours Show rms button bar keyboard accelerator key M Herzfeld Berger Analysis ini xi Ele Contours Fi Help Figure 2 12 Display of RMS error contour lines The display of rms errors as contours provides a great opportunity to visually assess the accuracy and uniqueness of a fit This requires that the rms error at each grid point of the Herzfeld Berger tables is calculated on slow computers a quite lengthy process If an iteration p 27 after a grid search has been performed already the rms errors are available for display immediately Otherwise with no iteration done yet a grid search is initiated which updates the initial guess p 27 for an iteration also The rms errors used for the display are actually sg errors see iteration p 27 for details The levels for the contours are stored in the ini file p 32 and the colors correspond to those of spinning sideband orders 1 to 5 They can be manipulated via the Setup p 23 item of the File menu June 4 2015 Chapter 2 Reference 2 1 1
50. gure 2 3 A TopSpin XWinNMR integration file integrals txt frequency to 1277 Hz XWinNMR Peak Picking Files The native file name is peak txt but HBA does not care Note that TopSpin can export its peak picking into XWinNMR format however from this file you cannot go back to TopSpin format If your spectrum contains several sites I recommend the following procedure only works in TopSpin perform peak picking for site 1 and save the result copy peak txt to peak1 txt for site 1 peak pick site 2 copy peak txt to peak2 txt etc With XWinNMR it is not possible to perform peak picking manually hence with several sites you should use integrals Note that the file peak txt contains sufficient information to determine the reference frequency Therefore you can keep the file in any location you like separate from the rest of the spectrum XWinNMR peak picking files are recognized by these features atthe beginning DU and USER An example of a file is shown in Figure 2 4 Note that you can highlight and copy the content of the figure box paste it into the Windows Editor save the file and read it into HBA Change the spinning frequency to 1277 Hz SpinWorks Peak Picking or Integrals Files The native file name is peaks txt but HBA does not care The file name is the same no matter whether it contains integrals or a peak picking list If your spectrum contains several sites I recom mend the following procedure perform pea
51. in the range from 15 to 15 and values of y in the range 0 lt y lt 30 the original tables have 5 spinning sidebands and 0 u 15 Intensities can be entered manually or parsed from peak picking or integration files The intensities can be displayed as contour diagrams giving a very instructive visualization of the quality of the data The initial guess required for the iterative refinement may result from a systematic grid search or can be entered interactively This provides different starting positions for the fitting proce dure The iterative refinement of the starting guess is achieved via the Marquardt Levenberg nonlin ear least squares fitting procedure The best fit can be graphically displayed as symbol in the contour plot or by comparison of stick spectra using experimental and calculated intensities So far HBA doesn t read and fit experimental spectra 1 1 3 License This program package can be used by any individual or organization without any fee as detailed in the Copyright statement If you find this program useful and publish results obtained by using HBA we would appreciate if you cite this program as HBA 1 74 K Eichele Universit t T bingen 2015 1 1 4 Troubleshooting Although HBA has been tested and used both in house and by others it is always possible that errors still exist Some errors may become apparent after detailed use on the wide variety of chemical systems It is the
52. ith experimental data that are subject to noise and other sources of error Therefore it is not surprising that often y is better defined than p see the error contours in Fig 3 7a on the right hand side they are elongated in the p direction This problem is not limited to axially symmetric chemical shift tensors see Fig 3 7c In this exam ple the situation is comparable to Fig 3 7a To deal with such situations one should think of the spinning speed as some sort of resolution power i e if the spinning speed is much higher than the difference between two principal components in Hz then HBA will have problems to discriminate between them On the other hand if the spinning speed is sufficiently slow so that not only verti cal contours are available but also horizontal lines then the situation improves cf Figs 3 7a and 3 7b and 3 7c and 3 7d In the slow spinning cases Figs 3 7b and 3 7d the intersection is defined much better These examples with y gt 20 also illustrate why it has been worth while to extend the original Herzfeld Berger tables to values of u up to 30 and to correct the tables in version 1 4 4 p 6 Hodgkinson et al 34 stated that for values of the asymmetry parameter smaller than 0 3 the reliability of the determination is a monotonically decaying function of the spinning speed and that static spectra always provide a more reliable determination of the asymmetry parameter than spinning spectra L Ju
53. k picking or integration for site 1 and save the result copy peaks txt to peaks1 txt for site 1 peak pick or integrate site 2 copy peak txt to peak2 txt etc This way you can make the data available to SpinWorks again by copying peaks2 txt back to peaks txt Manual peak picking is easy in SpinWorks 6 June 4 2015 Chapter 2 Reference DU D u USER klaus NAME KOOPH2BU EXPNO 5 PROCNO 1 Fi 515 705ppm F2 512 803ppm MI 0 20cm MAXI 10000 00cm PC 1 000 ADDRESS FREQUENCY INTENSITY Hz PPM 1 1623 1 8752 227 108 0208 0 68 2 1686 3 7466 867 92 1567 2 10 3 1749 6 6177 484 76 2431 6 44 4 1812 4 4901 389 60 4934 12 00 5 1874 8 3630 764 44 8112 10 76 6 1937 5 2355 498 29 0718 6 36 7 2000 3 1078 195 13 3072 1 80 8 2063 9 216 748 2 6751 3 60 9 2126 1 1481 216 18 2813 5 13 10 2189 0 2760 907 34 0754 3 35 11 2252 8 4058 491 50 0903 2 42 12 2314 8 5320 644 65 6679 3 68 13 2378 0 6605 740 81 5286 1 40 14 2440 9 7887 141 97 3438 3 06 15 2504 2 9174 000 113 2263 3 16 16 2567 1 10453 771 129 0214 2 72 17 2630 0 11733 727 144 8187 1 64 18 2692 9 13012 852 160 6058 0 86 19 2755 7 14290 209 176 3710 0 39 Figure 2 4 A XWinNMR TopSpin peak list peak txt SpinWorks peak list for peak freq Hz freq ppm intensity intensity Absolute 0001 12835 858 79 2444 0 0261 69151 0002 8838 911 54 5686 0 0863 228852 0003 4835 805 29 8547 0 2278 604416 0004 831 592 5 1340 0 4555 12083
54. l default to 1 MHz enter 161 978043 and move the cursor away This should update the spinning frequency from 24 70341 in ppm to 4001 41089 in Hz Change the spinning frequency to 4000 Hz June 4 2015 Chapter 2 Reference lt xml version 1 0 encoding UTF 8 gt lt PeakList gt lt PeakList1D gt lt PeakListiDHeader expNo 1 name cpniphms procNo 1 source D u gt lt PeakPickDetails gt F1 146 992ppm F2 707 941ppm MI 0 37cm MAXI 16 00cm PC 4 000 lt PeakPickDetails gt lt PeakList 1DHeader gt lt Peak1D F1 54 2179 intensity 1 31 type 0 gt lt Peak1D F1 29 4939 intensity 3 35 type 0 gt lt Peak1D F1 4 7639 intensity 6 69 type 0 gt lt Peak1D F1 19 8985 intensity 10 54 type 0 gt lt Peak1D F1 44 5997 intensity 12 34 type 0 gt lt Peak1D F1 69 3021 intensity 12 72 type 0 gt lt PeakiD F1 94 0019 intensity 14 8 type 0 gt lt Peak1D F1 118 7134 intensity 15 0 type 0 gt lt Peak1D F1 143 4123 intensity 13 95 type 0 gt lt PeakiD F1 168 0808 intensity 12 13 type 0 gt lt Peak1D F1 192 8133 intensity 9 09 type 0 gt lt Peak1D F1 217 5482 intensity 7 62 type 0 gt lt Peak1D F1 242 2048 intensity 7 88 type 0 gt lt PeakiD F1 266 908 intensity 8 88 type 0 gt lt Peak1D F1 291 6273 intensity 7 68 type 0 gt lt Peak1D F1 316 3145 intensity 5 84 type 0 gt lt Peak1D F1 341 0205 inten
55. l problem with the graphical approach to Herzfeld Berger fitting and the approach used by HBA prior to version 1 7 it assumes that the intensity of the center peak is error and distortion free This is certainly not justified Using intensity ratios Iy Ip for the spinning sidebands will contaminate all data if the intensity of the isotropic peak is in error e g due to the presence of impurities Grimmer et al suggested to use intensity ratios not involving the center band in such cases 30 Otherwise one advantage of the Herzfeld Berger method is that distorted peaks can be excluded from the analysis 31 in contrast to methods that use all sideband intensities e g the method proposed by Fenzke et al 32 or the method of moments p 50 Another source of distortions are radial sidebands outlined in the following section To protect yourself yourself from the above pitfalls Check the contours is the scatter of intersections systematic or statistical Generate stick spectra are the differences in spinning sideband intensities systematic or statistical 3 2 4 Radial Sidebands While analyzing MAS spectra by hand using the graphical method of Herzfeld Berger I frequently noticed that the contour line of one of the first order spinning sidebands seemed to behave strangely and not in line with the other sidebands i e there was a clear intersection of most contour lines ex cept for the 1 spinning sideband Examples are
56. n Reson 1986 70 408 415 DOI 10 1016 0022 2364 86 90129 0 24 Magic Angle Spinning P NMR Spectra of Solid Dihydrogen Phosphates Lagier C M Olivieri A C Apperley D C Harris R K Solid State Nucl Magn Reson 1992 1 205 210 DOI 10 1016 50926 2040 10 80005 0 25 Retrieving P Chemical Shift Tensor Information for Dihydrogen Phosphates in the Presence of Homonuclear 3 p P Dipolar Coupling Lagier C M Olivieri A C J Magn Reson 1997 126 138 141 DOI 10 1006 jmre 1997 1122 26 Spinning Sideband Pattern Deviations in Cross Polarization MAS NMR Spectra Jeschke G Grossmann G J Magn Reson Ser A 1993 103 323 328 DOI 10 1006 jmra 1993 1173 27 Variable Amplitude Cross Polarization MAS NMR Peersen O B Wu X L Kustanovich I Smith S O J Magn Reson Ser A 1993 104 334 339 DOF 10 1006 jmra 1993 1231 28 Towards quantitative CP MAS NMR Metz G Ziliox M Smith S O Solid State Nucl Magn Reson 1996 7 155 160 DOI 10 1016 50926 2040 96 01257 X 29 Nuclear magnetic shielding tensors for the carbon nitrogen and selenium nuclei of selenocyanates a combined experimental and theoretical approach Bernard G M Eichele K Wu G Kirby C W Wasylishen R E Can J Chem 2000 78 22 23 June 4 2015 Bibliography 614 625 DOI 10 1139 v00 046 30 Tensor der P chemischen Verschiebung in KaNa POsF Grimmer A R
57. ne 4 2015 Chapter 3 Background Information 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Figure 3 7 Contour lines left and error levels right for spectra obtained at 100 MHz and at spinning rates of 2000 Hz a and c or 1000 Hz b and d corresponding to an axially symmetric chemical shift tensor a b 911 622 100 633 100 ppm or non axially symmetric chemical shift tensor c d 11 110 522 90 633 100 ppm p 0 81 48 June 4 2015 Chapter 3 Background Information 3 2 7 Malicious Software Another pitfall could be malicious software The program might not work properly Although tested and found working fine there might be some extreme situation where the program fails Never trust anything All warranty is disclaimed One problem that we became aware of has been associated with the extended Herzfeld Berger tables themselves for relatively large values of y gt 10 some of the outer spinning sidebands did not have the proper intensity This gets particularly bad for y 30 R Herzfeld Berger Analysis Figure 3 8 Problem with the extended Herzfeld Berger tables aliasing in the calculations causes an increase in intensity for spinning sideband orders greater than 12 the spinning sidebands on the left hand side are too high This issue was fixed in version 1 4 4 p 6 Some programs by other people may also show this problem To protect yourself yourself from the above pi
58. ng Inno Setup available http www jrsoftware org Jochen Kalmbach for demonstrating how to statically link against the Microsoft CRT and thus get rid of VCREDIST X86 EXE http blog kalmbach software de chicks for demonstrating in his pdfp PDF tools how to establish Dynamic Data Exchange DDE with Adobe Acrobat Reader http www esnips com web PDFTools This manual has been produced using the MiKTEX http www miktex org distribution of KIR in combination with the TeXnicCenter editor http www ToolsCenter org IrfanView http www irfanview com has been used to process bitmapped images and June 4 2015 Chapter 1 Getting Started Inkscape www inkscape org for dealing with vector graphics 1 4 2 Trademarks Microsoft MS is a registered trademark and MS DOS and MS Windows are trademarks of Mi crosoft Corporation WIN NMR and TopSpin are products of Bruker Franzen Analytik GmbH and Bruker Biospin Other brand and or product names are used for identification purposes only and are trademarks registered trademarks or copyrights of their respective owners 1 4 3 Copyright Information Copyright O 1995 2015 Klaus Eichele All rights reserved This program executable help file and related files may be distributed freely and may be used without fee by any individual for non commercial use and by any govern ment organization Although the copyright holder retains all rights to this document and
59. ntegrals file in their native subdirectory pdata lt procno gt HBA will try to retrieve the reference frequency from the file procs in the same directory parameter SF TopSpin XWinNMR integrals files are recognized by these features atthe beginning Current data set e in the second line NAME An example of a file is shown in Figure 2 3 Note that you can highlight and copy the content of the figure box paste it into the Windows Editor save the file and read it into HBA The reference frequency will default to 1 MHz enter 81 023559363091 and move the cursor away This should update the spinning frequency from 15 78658 in ppm to 1279 08447 in Hz Change the spinning 15 June 4 2015 Chapter 2 Reference Current data set NAME KOOPH2BU EXPNO 5 PROCNO 1 USER klaus DU D u Number Integrated Region Integral 1 129 324 119 220 0 99161 2 113 495 104 402 3 25990 3 99 014 86 889 14 97713 4 82 848 70 387 44 46394 5 69 377 54 558 81 38064 6 53 885 37 719 68 14137 7 34 352 23 238 40 19295 8 18 860 6 736 14 68601 9 4 041 8 083 25 05822 10 11 451 24 585 31 95773 11 29 637 38 056 20 61244 12 44 118 54 895 16 28154 13 60 957 69 714 23 33287 14 76 449 88 237 11 32771 15 91 941 103 055 20 89910 16 107 096 118 884 21 40571 17 123 262 134 712 17 52124 18 139 090 149 531 10 06131 19 155 929 164 686 4 95554 20 172 432 180 851 1 79446 21 186 913 196 006 0 81569 Fi
60. o loss of data or data being rendered inaccurate or losses sustained by you or third parties even if such holder or other party has been advised of the possibility of such damages In any case liability will be limited to the amount of money that the copyright holder received from you for the use of this program June 4 2015 2 Reference This section covers the following topics Help on Menu Items Help on Button Bar Items INI File Settings Command Line Parameters Known Problems 2 1 Help on Menu Items Explains the meaning of specific menu items Explains the actions of specific buttons Documents the INI file parameters Documents the parameters that can be passed to HBA on the command line Summarizes known problems Category Menu Item Action File Enter intensities p 12 Enter spinning sideband intensities manually Read integrals Read spinning sideband intensities from an inte peak lists p 13 gration file or peak picking file Read session p 21 Retrieve parameters intensities fitting results etc from a previous session with HBA Write session p 21 Save parameters intensities fitting results etc to a file for later retrieval Write meta file p 22 Creates a graphics file in Windows meta file for mat Write spin system p 22 Writes chemical shift tensor information into a WSolids spin system file Setup p 23 Customize color settings Exit p 23 Exit HBA Contours Show contours p 24
61. ons can be useful for batch processing or for integrating HBA in other applications Usage hba switch parameter switch parameter where switch s for session files a file name as parameter is mandatory i for integration files a file name as parameter is mandatory p for peak lists a file name as parameter is mandatory sf for reference frequency followed by the value in MHz as parameter nur for spinning frequency followed by the value in Hz as parameter diso for isotropic chemical shift followed by the value in ppm as parameter f do iterative fitting no further parameter required r create a report file if no file name is given output will be written to ITERATE TXT q quit HBA parameter filename a fully qualified file name full path including drive specifying the session integration or peak list file If the file name or path contain blanks filename should be put between double quotes value a floating point value Note Parameters will be processed in the order given on the command line In newer versions of HBA you can use the switch i or p followed by the file name independent of the actual content of the file peak picking or integration because HBA will check the content anyway Examples Start HBA with a peak picking file d program files winsolids hba exe p c documents and settings klaus my files hba data triphenylphosphine pp Start HBA with a session file iterat
62. play of RMS errors can be set If the option Save settings as default is enabled the selected settings are stored in the ini file of HBA Otherwise they only affect the current session 2 1 0 Exit To exit HBA simply use the menu item Filel Exit the keyboard keys ALT F4 This action exits HBA June 4 2015 Chapter 2 Reference 2 1 10 Show Contours Ve The contour lines can be displayed as an overlay of M Herzfeld Berger nomograms by using the J e menu item Contours Show contours e button bar keyboard accelerator key S Herzfeld Berger Analysis Mu 127 Rho 0 15 D11 91 9 D22 25 9 D33 64 3 Figure 2 11 Display of contour lines Ideally the contour lines for all spinning sidebands should intersect in one unique point This unique point would correspond to the correct y p parameters for this particular chemical shift tensor In less ideal cases the intersection of all contours might not be as clear The spread of contours and their intersections gives a good visual indication of the quality of the data It may also indicate if this particular case is more likely to have problems in determining the skew of the tensor or the span Spinning sidebands of positive and negative ordinal number are represented by solid and dashed lines respectively The horizontal axis corresponds to the u parameter and the vertical axis to the p parameter of the Herzfeld Ber
63. responsibility of the user to determine the correctness of the results As stated in the Disclaimer we disclaim all warranty If errors are noticed please notify us of your problems and the prescribed or suggested corrections so that others may benefit from the improved code Also suggestions for improvements are always welcome Dr Klaus Eichele Institut f r Anorganische Chemie Universit t T bingen Auf der Morgenstelle 18 72076 T bingen Germany klaus eichele uni tuebingen de June 4 2015 Chapter 1 Getting Started 1 2 Overview The ordering in this list corresponds to the order of actions taken in a typical session New users may want to follow each step Action How to achieve this action Enter Before starting the Herzfeld Berger analysis input of the spinning sideband intensities intensities is required Basically there are three different ways of entering the spinning sideband intensities Enter the intensities manually p 12 Read an integration or peak picking file p 13 Read a previous session p 21 Display After the intensities of the spinning sidebands have been entered one can contours display the corresponding contour lines Show contours p 24 One can use the Data reader p 26 to query the principal components of the chemical shift tensor corresponding to a specific 1 0 p 36 pair or to set an initial guess p 27 for the iteration p 27 Simulate One can attempt to simulat
64. ror estimates The procedure follows the rigorous statistical analysis of errors doc umented by Olivieri 4 However while his approach uses the chemical shift anisotropy Ac and asymmetry parameter 7 to describe the chemical shift tensor and that are used as fit parameters HBA employs the Herzfeld Berger parameter y instead of Ac and x instead of y see conventions p 36 The results should be comparable The matrix B dl Ol calc Bu 2 2 j D Opi Op is evaluated numerically The summation involves all sidebands observed experimentally p are the parameters y and x The partial derivatives dI calc dp are estimated as quotients I calc Sp where Jl calc is the change in calculated intensity when p is changed by a small amount from its refined value Currently u is changed by 0 2 and x by 0 05 The diagonal elements of B are used to calculate the conditional standard deviation for pj SconD Pi sm x Bi 2 3 while the diagonal elements of the inverse of the matrix B give the marginal standard deviation of the refined p values SMARG Pi sm x B 55 24 Also given in the report is the ratio Ri siarc pi Sconp Pi 2 5 the value of which is close to 1 if the parameters are reasonably well defined by the experimental data Note that syarc p takes into account the presence of random errors contained in Spit e the sensitivity of sg toward small changes in pj the interd
65. rst line otherwise it wants for each spinning sideband a row containing five floating point numbers specifying the beginning in ppm and Hertz and ending in ppm and Hertz of the integration range followed by the actual value of the integral In short hand notation ppm1 nul ppm2 nu2 Int 19 June 4 2015 Chapter 2 Reference Peak List File Name nmr kooph2bu 005001 1R Peak Results saved in File F NMR KOOPH2BU 005001 PP Nucleus E cuiz SF 81 023559 MHz OFFSET 515 7048 ppm SW_p 83333 33 Hz SI 4096 Peak Picking Parameter Peak constant PC 1 00 Noise 2236 Sens level 8944 Peak Picking region Start ppm Hz End ppm Hz MICA MAXI 515 70 41784 2 512 80 41549 1 05916 100 00 Peak Picking results Peak Nr Data Point Frequency PPM Intensity Int 1 1560 5 10034 81 123 8505 10618 1 2 1623 1 8762 29 108 1450 45345 5 3 1686 3 7477 14 92 2835 142188 17 4 1749 6 6188 38 76 3775 438379 53 5 1812 3 4912 09 60 6255 813997 100 6 1874 8 3641 48 44 9435 717164 88 7 1937 5 2365 97 29 2011 436170 53 8 2000 3 1088 10 13 4295 122305 15 9 2063 9 205 77 2 5397 243424 29 10 2126 1 1470 45 18 1485 340400 41 11 2189 0 2750 08 33 9418 223965 27 12 2252 7 4046 86 49 9467 164764 20 13 2314 7 5309 27 65 5275 247248 30 14 2377 9 6594 69 81 3923 96832 11 15 2440 9 7875 35 97 1983 208677 25 16 2504 1 9161 80 113 0757 214262 26 y 2567 0 10441 46 128 8694 183193 22 18 2629 9 11721
66. s vide infra The value of M given in Tab 3 5 may indicate by how many spinning sideband orders the current test peak is off from being the center peak This result might be fortuitous nor have I seen anyone mentioning this in the literature nor do I have any mathematical proof Ishould do some more tests Determination of Higher Moments In the next step we calculate the second and third moments about the isotropic peak The contribu tion of each spinning sideband to the second and third moments is listed in Tab 3 6 It is obvious that high order spinning sidebands still make large contributions to the second and third moments albeit their low normalized intensity We have noted above the effect of missing some high order spinning sidebands to cause a non zero value of the first moment This is an issue that generally makes the June 4 2015 Chapter 3 Background Information N I I norm M34 ppm M ppm M3 ppm 5 0 0007 0 000228938 0 045787546 9 157509158 1831 501832 4 0 0060 0 001962323 0 313971743 50 23547881 8037 676609 3 0 0413 0 013507326 1 620879121 194 5054945 23340 65934 2 0 2127 0 069564364 5 565149137 445 2119309 35616 95447 1 0 9870 0 322802198 12 91208791 516 4835165 20659 34066 0 0 8710 0 284863946 0 0 0 1 0 4650 0 152080063 6 083202512 243 3281005 9733 124019 2 0 3538 0 115711669 9 256933543 740 5546834 59244 37467 3 0 0934 0 030546834 3 665620094 439 8744113 52784 92936 4 0 022
67. sity 7 15 type 0 gt lt Peak1D F1 365 7328 intensity 6 92 type 0 gt lt Peak1D F1 390 4183 intensity 5 28 type 0 gt lt Peak1D F1 415 1213 intensity 5 86 type 0 gt lt PeakiD F1 439 8502 intensity 6 3 type 0 gt lt Peak1D F1 464 5187 intensity 5 59 type 0 gt lt Peak1D F1 489 2395 intensity 4 31 type 0 gt lt Peak1D F1 513 937 intensity 4 21 type 0 gt lt Peak1D F1 538 6281 intensity 5 25 type 0 gt lt Peak1D F1 563 3579 intensity 5 9 type 0 gt lt Peak1D F1 588 0827 intensity 4 96 type 0 gt lt Peak1D F1 612 7285 intensity 3 47 type 0 gt lt Peak1D F1 637 4625 intensity 2 06 type 0 gt lt Peak1D F1 662 1812 intensity 1 02 type 0 gt lt PeakList1D gt lt PeakList gt Figure 2 2 A TopSpin peak picking file peaklist xm1 TopSpin XWinNMR Integrals Files The native file name is integrals txt but HBA does not care If your spectrum contains sev eral sites I recommend the following procedure perform integration for site 1 and save the result copy integrals txt to integrals1 txt for site 1 integrate site 2 copy integrals txt to inte grals2 txt etc This way you can make the data available to TopSpin again by copying inte grals2 txt back to integrals txt Note that the file integrals txt does not provide sufficient information to determine the refer ence frequency default value 1 MHz If you keep the i
68. teration Have a look at the errors reported by HBA Fit Report p 28 Because only three peaks have been considered in the iteration the agreement between experimental and calculated spectra is perfect and the errors zero in all three examples spinning sidebands present in the calculation but not in the experiment have been neglected because their intensities have been set to 1 To indicate such potential problems the display of stick spectra under Fit Compare p 29 and the intensities listed in the Report p 28 utilize normalized intensities of the total spectra hence the visual agreement gets increasingly worse for the examples shown in Fig 3 2 Note that the simulations shown by Clayden et al in Fig 4c d of their paper are crude the three peaks of sideband orders 1 0 1 should have equal intensities also the intensities of sidebands should be symmetric about the isotropic peak Actually the proper procedure to obtain a fit for this hypothetical example is to set the intensities of all spinning sidebands not observed experimentally to zero explicitly In this case the fit will be the best compromise 011 29 5 2 0 022 0 0 1 1 633 29 5 1 7 to a non existant solution June 4 2015 Chapter 3 Background Information 0 5 10 15 20 25 30 Figure 3 1 Contour lines top and error levels bottom for a hypothetical three line MAS spectrum consisting of spinning sidebands of orders 1 0
69. tfalls Test the software that you are using Try to reproduce literature data June 4 2015 Chapter 3 Background Information Figure 3 9 Definition of the moment of a line shape illustrated for the line shape of a static chemi cal shift powder pattern The mean angular frequency w corresponds to the isotropic chemical shift the value of f w corresponds to the intensity of the line shape at w 3 3 Method of Moments 3 3 1 Theory Although the theory necessary to calculate MAS spectra of nuclei under the influence of chemical shift anisotropy is well known Maricq and Waugh stated that a calculation of this type requires approximately 15 minutes on a state of the art computer in 1979 35 Because only two parameters the span and skew of the chemical shift tensor should be extracted from line shape one has the task of constructing a theory of two independent moments of this line shape The jth moment Mj for a line shape f w is defined in angular frequency units as Fig 3 9 T w w flw dau M 3 1 w gt 3 2 Under MAS the theoretical second and third moments of the spectrum remain unaffected by sample spinning and one needs to only consider the intensities of the spinning sidebands and neglect the intensity of the baseline Because of the discreteness of the spinning sidebands the k th moment of a magic angle spinning spectrum in different units is given as N My radks 2
70. the software package you are allowed to copy and distribute verbatim copies of them as you received them in any medium provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty keep intact all the notices that refer to this license and to the absence of any warranty and distribute a copy of this license along with it This package may not be distributed as a part of any commercial package You are expressly not allowed to sell or license this package 1 4 4 Disclaimer of Warranty Because this software package is licensed free of charge there is no warranty to the extent per mitted by applicable law This package is provided as is without warranty of any kind either expressed or implied including but not limited to the implied warranties of merchantability and fitness for a particular purpose the entire risk as to the quality and performance of the contents of this package is with you should this package prove defective you assume the cost of all nec essary servicing repair or correction in no event unless required by applicable law or agreed to in writing will any copyright holder or any other party who may modify and or redistribute this package as permitted in the license be liable to you for damages including any general special incidental or consequential damages arising out of the use or inability to use the package including but not limited t
71. tick spectra The direct visual comparison of calculated and experimental spinning sideband intensities allows for a better assessment of the quality of the fit In earlier versions of HBA both spectra were nor malized to have the same center peak intensity Now the total sum of sidebands and centerband are normalized and experimental and calculated spectra are plotted with a slight horizontal offset to avoid overlap 29 June 4 2015 Chapter 2 Reference 2 1 19 Simulate Interactively The display of stick spectra representing experimental and Ri calculated intensities is activated by the menu item Fit Simulate the button bar the keyboard accelerator key M lola File Ci p ive Tensor Analysis 4 n 30 Rho0500 1 F Lp oin D11 1006 D22 51 0 D33 98 1 15 10 5 0 5 4i DL Figure 2 14 Simulate the experimental stick spectrum manually Using this feature one can easily vary the Herzfeld Berger parameters y and p via scroll bars and directly observe the effect on the calculated spectrum In earlier versions of HBA both spectra were normalized to have the same center peak intensity Now the total sum of sidebands and centerband are normalized and experimental and calculated spectra are plotted with a slight horizontal offset to avoid overlap This action automatically sets the current values of y and p as initial guess for the iter
72. tion in Chemical Journals Jones R N Cole A R H Miller F A Elyash vich M A Forster T Hadni A Morino Y Sheppard N Fluck E Lippincott E R Lord R C Nagakura S Pliv J Sir Thomp son H Turner D W Herzberg G Urbanski T Pure Appl Chem 1972 29 625 628 DOI 10 1351 pac197229040625 Presentation of NMR Data for Publication in Chemical Journals B Conventions Relating Spectra from Nuclei other than Protons Recommendations 1975 Sheppard N Elyash vich M A Miller F A Becker E D Beynon J H Fluck E Hadni A Zerbi G Herzberg G Jezowska Trzebiatowska B Morino Y Nagakura S Rao C N R Sir Thompson H Turner D W Pure Appl Chem 1976 45 217 219 DOI 10 1351 pac197645030217 Gas Phase C Chemical Shifts in the Zero Pressure Limit Refinements to the Absolute Shielding Scale for PC Jameson A K Jameson C J Chem Phys Lett 1987 134 461 466 DOI 10 1016 0009 2614 87 87173 7 An Improved Carbon 13 Nuclear Shielding Scale Raynes W T McVay R Wright S J J Chem Soc Faraday Trans 2 1989 85 759 763 DOI 10 1039 F29898500759 9 A revised experimental absolute magnetic shielding scale for oxygen Wasylishen R E Bryce D L J Chem Phys 2002 117 10061 10066 DOI 10 1063 1 1518683 10 Absolute Shielding Scale for Si Jameson C J Jameson A K Chem Phys Lett 1988 149 300 305 DOI 10 1016 0009 2614 88
73. to figure out the kind of file and the software producing the file Currently the following file formats are recognized more details will be provided in the following subsections Bruker TopSpin 2 1 peak picking p 14 and integrals p 15 other TopSpin versions may work as well but have not been checked Bruker XWinNMR 3 5 peak picking and integrals p 16 Bruker WinNMR 5 0 5 1 or 6 peak picking or integrals files p 18 in ASCII format Kirk Marat s University of Manitoba SpinWorks 3 1 peak picking or integrals p 16 ftp davinci chem umanitoba ca pub marat SpinWorks If your NMR files do not correspond to any of the Bruker formats specified above or originate from a different vendor consider using SpinWorks to process your data A few other points are worth considering If your spectrum contains several sites that need to be analyzed separately be aware that most of the above processing software packages write their data to a file of constant fixed name Hence analyzing a second site will overwrite the data of the first site The following subsec tions on file formats will contain some hints on that Should you use peak picking or integrals Integrals are more robust with respect to small vari 13 June 4 2015 Chapter 2 Reference ations in spinning speed the spinning sidebands will show a systematic increase in line width with increasing order of the spinning sideband but require e
74. to visually inspect the agreement Are there systematic devia tions The best test if dipolar interactions play a role the results should show a systematic field dependence 44 June 4 2015 Chapter 3 Background Information 1 0 0 5 0 0 1 0 0 5 10 15 20 25 30 Figure 3 6 Contour lines top and error levels bottom for the spectrum and fit for the 31P MAS NMR spectrum of NH H gt PO4 shown in Fig 3 5 The best fit is obtained with u 8 456 and p 0 346 corresponding to 911 18 5 0 2 927 5 4 0 1 035 21 5 0 2 ppm June 4 2015 Chapter 3 Background Information 3 2 3 Distorted Data For a variety of reasons the experimental intensities can be inaccurate or distorted for example A low signal to noise ratio will render the determination of intensities difficult it could be preferable to use peak heights from peak picking rather than integrals 2 the errors however should be statistical If MAS spectra have been obtained after cross polarization CP the sideband intensities may suffer CP efficiency artifacts because the Hartmann Hahn matching condition is violated for large frequency offsets from the true Hartmann Hahn match a simple empirical correction function has been suggested 26 alternatively application of variable amplitude CP 27 28 will also reduce such problems 29 The intensity distortions should show systematic devia tions There is a fundamenta
75. ucleus in the species under investigation vs c ppm 1e06 Vnucl Vs Vnucl The chemical shift is the difference in shielding between the nucleus in the species under investigation cs and the shielding of the same nucleus in a reference compound ref ppm aret Os 1 Orer Shifts commonly used in solution and solid state NMR studies are thus positive to high frequency Absolute shieldings are positive to low frequency and are only accessible via theoretical calculations The establishment of a correspondence between a chemical shift scale and a magnetic shielding scale is not a trivial task and requires both careful theoretical calculations and experimental measurements Some examples of established shielding scales are available for carbon 7 8 oxygen 9 silicon 10 phosphorus 11 and tin 12 Comments The symbol c should only be used for absolute shieldings Often however authors use a pseudo shielding scale where the shielding is obtained by simply reversing the sign of the chemical shift In our opinion this adds only to the confusion without providing any additional insight 3 1 2 Standard Convention Principal Components 611 gt 022 gt 633 011 022 033 Isotropic Value iso 611 022 633 3 Viso 011 022 033 3 In what we shall call the standard convention the principal components of the chemical shift tensor 611 622 and 633 are labeled according to
76. used with the Herzfeld Berger tables corresponds to the skew of a tensor described here For historical reasons we used p throughout this manual but generally we prefer x 13 The Herzfeld Berger convention is related to the Standard convention via 099 dian k 2 3 633 3 diso 22 A 2 011 3 iso 622 033 3 1 4 Haeberlen Convention Principal Components l zz diso gt xx o 2 l yy m diso Os 3 Isotropic Value iso 011 022 633 3 Reduced Anisotropy 627 biso Anisotropy Ag bzz xx dyy 2 Ao 35 2 Asymmetry y yy xx 0 lt y 1 The Haeberlen Mehring 14 15 convention uses different combinations of the principal components to describe the line shape This convention requires that the principal components are ordered ac cording to their separation from the isotropic value The center of gravity of the line shape is described by the isotropic value which is the average value of the principal components The anisotropy and reduced anisotropy describe the largest separation from the center of gravity The term reduced anisotropy is not used in the literature but we introduce it here in order to be able to distinguish between and Av The sign of the anisotropy indicates on which side of the isotropic value one can find the largest separation The asymmetry parameter indicates by how much the line shape deviates from that of an axially symmetric
77. xcellent signal to noise ratios and well defined baselines If spinning is stable peak picking is not worse than integration is often done quicker especially for spectra containing only one spinning sideband family and with low signal to noise ratios may even give better results than integrals 2 Once File Read Integrals Peak Lists has been selected from the menu HBA opens a dialog box to retrieve the name of the file containing the integrals or peak picking results After parsing the file HBA prompts for input to define the reference frequency frequency of the chemical shift standard spinning rate and isotropic chemical shift using the dialog box Define Spectrum Settings p 21 Using the information contained in the files HBA attempts to guess these values Thereference frequency is determined from the ratio of the frequency ppm values given for the peak position or integrals if available The TopSpin peak picking format has the disadvantage that it does not contain sufficient information to work out the reference frequency Hence its default value is set to 1 MHz HBA will recalculate the spinning frequency if you enter the correct reference frequency and move away the cursor e g hit the Tab key on the keyboard The average distance between peaks or integrals serves as a default for the spinning rate in Hz if a proper value for the reference frequency has been retrieved or in ppm if the reference frequency could not

HBA User Manual - Institut für Anorganische Chemie

Contents

Download Pdf Manuals

Related Search

Related Contents

HBA User Manual - Institut f&uuml;r Anorganische Chemie

Contents

Download Pdf Manuals

Related Search

Related Contents

HBA User Manual - Institut für Anorganische Chemie