#189 depict donor-acceptor energies

open
5
2011-05-01
2011-04-11
No

The energy of stabilization of a donor-acceptor interaction is given as the E(2) value in this NBO report:

SECOND ORDER PERTURBATION THEORY ANALYSIS OF FOCK MATRIX IN NBO BASIS

Threshold for printing: 0.01 kcal/mol
E(2) E(j)-E(i) F(i,j)
Donor NBO (i) Acceptor NBO (j) kcal/mol a.u. a.u.
===============================================================================

within unit 1
1. BD ( 1) C 1- C 2 204. RY*( 1) O 3 1.86 1.86 0.053
1. BD ( 1) C 1- C 2 1308. BD*( 1) N 7- C 8 5.31 1.46 0.079
152. LP ( 1) O20 1237. BD*( 1) N69- H74 8.25 1.68 0.105
140. CR ( 1) C80 1110. RY*( 3) C83 1.32 13.77 0.120

It would be really nice if the E(2) value could be depicted graphically. A point in a color spectrum corresponding to the value could appear in the overlap of the donor PNBO and the acceptor PNBO. PNBOs are better for showing overlap than the NBOs. Donor acceptor interactions are directional, so a particular interaction is specified by an ordered pair of NBO IDs, but there is no conflict with an interaction in the reverse direction since in how these things work there won't be one. If the mouse pointer was placed on one of the overlaps, the numeric E(2) value associated with the color would be shown. The ability to show the overlaps from a particular donor to all its acceptors and show overlaps to a particular acceptor from all its donors is highly desirable. A threshold of E(2) value would further subset which overlaps were shown.

Also, it is desirable that the pairwise steric exchange energies (another NBO report) receive a similar treatment. Both directions exist in this case, but the value should be the same in both cases so the same approach (showing a color in the PNBO overlap) would work.

As it stands these values have to be looked up from a long text report and a graphical depiction would be so much more convenient.

Discussion

  • Bob Hanson

    Bob Hanson - 2011-05-01

    First question I would ask is this: Is this a totally novel idea, or do you know of examples where programs do this?

    Then: It seems to me we would need to depict two orbitals, and then indicate by color (or a vector?) the strength and the direction of the donor-->acceptor E(2) value.

    "In the overlap" is a bit of a problem, I think. How would you define that?

    Bob

     
  • Bob Hanson

    Bob Hanson - 2011-05-01
    • assigned_to: nicove --> hansonr
     
  • Granger Hermitage

    I am unaware of this being done elsewhere but it would be quite surprising if it were not since interaction between pairs of orbitals is such a central feature of some quantum analyses. Donor-acceptor interactions (charge transfer) and steric pressure are examples of pairwise orbital interactions, the first being unidirectional and the second being bidirectional.

    A colored arrow representing a vector would be a fine way of depicting donor-acceptor interactions. It might be best to allow the orbitals themselves not to be shown just the arrows as orbitals might obscure the arrows. Allowing drawing of just the arrows would means that sense could be made in a volume of numerous donor-acceptor interactions. The arrow would be from and to nominal centers of orbitals. The arrow of the vector could be of user-definable line width. When the mouse pointer was placed on the arrow, the identities of both orbitals (NBO IDs and bond type details) would be shown in text as would the E(2) and F(I,J) of the interaction.

    For depiction of bidirectionally symmetric interactions, a line segment rather than an arrow might be used.

    The overlap would be the intersection of the visual representations of PNBOs. However, colored arrows would have advantages over that.

     
  • Bob Hanson

    Bob Hanson - 2011-05-01

    version=12.1.45_dev

    # new feature: NBO second-order data saved in modata:
    # Var x = getProperty("auxiliaryInfo.models[2].moData.secondOrderData")
    # show x
    # >> x = [[69.0,115.0,1.2800001],[69.0,134.0,0.67],...[mo1, mo2, E(2)]
    # where mo1 and mo2 are indexes into the MO list (Note that they are floats, not integers.)
    # and E(2) is the E(2) value calculated by genNBO.

    Now, what we DO with those is another question....

     
  • Bob Hanson

    Bob Hanson - 2011-05-02

    correction:

    version=12.1.45_dev

    # new feature: NBO second-order data saved in modata:
    # Var x = getProperty("auxiliaryInfo.models[2].moData.secondOrderData")
    # show x
    # >> x = [[69.0,115.0,1.2800001,0.044],[69.0,134.0,0.67,0.026],...[mo1, mo2, E(2), f(i,j),...]
    # where mo1 and mo2 are indexes into the MO list (Note that they are floats, not integers.)
    # and E(2) and f(i,j) are the E(2) value calculated by genNBO.

     
  • Bob Hanson

    Bob Hanson - 2011-05-02

    All for now.... Question: Can you find some depictions of these in journal articles? That might point us to the relevant software.

    Also for now it will be up to you to create some arrows using DRAW and the data provided. Then we can see how to incorporate this into Jmol more generally.

     
  • Granger Hermitage

    If an arrow were pointed at by mouse the interaction and the id and type of both orbitals is shown textually, and perhaps if the arrow were clicked on both orbitals could be drawn. The NBOs rather than PNBOs could be drawn since the arrow depicts the interaction and PNBO overlap is not needed to show the interaction. In this particular context, drawing NBOs might be better than drawing PNBOs since NBOs are be less cluttered due to less overlap.

    If clicking on the arrow causes both orbitals (NBOs) to be drawn, left clicking on an orbital might draw all arrows from that orbital and right clicking on an orbital might draw all arrows to that orbital. The arrows so drawn could in turn be clicked on to follow chains of orbital interactions.

    A threshold E(2) value might be set for what arrows would be drawn. Bond type for donor orbitals and bond type for acceptor orbitals might be set to further filter what arrows are drawn.

     
  • Granger Hermitage

    Perhaps the two orbitals would be transiently drawn if the mouse pointed at an arrow, that is if the pointer were moved off the arrow the orbitals would disappear. If the arrow were clicked on, the drawing of the two orbitals would be persistent and the pointer could be moved to an orbital to left or right click to get all arrows from or to that orbital.

     
  • Bob Hanson

    Bob Hanson - 2011-05-02

    I think you have enough to work with. Give it a try with a web page and let's see how it works. Probably the best way is to construct a drop-down list of the interactions on the page, and let the user select what that menu does -- show the donor, show the acceptor, or show the arrow. Or something of that sort. From JavaScript you can access the second-order values using:

    var Data = jmolGetPropertyAsArray('auxiliaryInfo.models[0].modata.secondorderdata')

    see http://chemapps.stolaf.edu/jmol/docs/examples-12/Jmol-12.zip for files.

     
  • Granger Hermitage

    It would be best if the E(2) value of the arrow were shown by its color rather than its length, since varying the length to show E(2) would disturb the 3D perspective of distance from point of view.

    It would be nice if the width of the arrow varied along the arrow in accordance with distance perspective.

    A fairly minimal arrowhead might be best - the arrow being a line or column, perhaps just sharpened to a point at one end to denote direction.

    NBOView is oriented to showing contours of nominated orbitals. One looks at overlaps of the contours of different orbitals. The graphical depiction of scalar results of orbital interactions may be novel. I have not seen it the literature, though it would be very convenient relative to looking through a long text report.

    The list of programming projects more central to my thesis is long, and I simply can't add to the list.

     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks