How can I determine the size of a tropolone molecule? It's a seven-membered aromatic ring with =O on the first carbon and -OH on the second carbon.

Is there any software available?


  • 4
    $\begingroup$ You may need to clarify what you mean by size (e.g. length across, volume, collision cross section, some other measure of steric bulk) $\endgroup$ – Tyberius Oct 21 at 14:34
  • $\begingroup$ Yes, length across. $\endgroup$ – user65414 Oct 21 at 15:08
  • 10
    $\begingroup$ It's about 3 cm or so on my screen... $\endgroup$ – Todd Minehardt Oct 22 at 14:19
  • 1
    $\begingroup$ I invite you to read comments and answer but it depends on what you are really looking for. Even drawing the molecule in scale while using typical tabulated bond lengths won't drive you much out of reality. Hard to judge how accurate you must be by reading just the question. $\endgroup$ – Alchimista Oct 22 at 18:02

You wanted to measure the distance "across tropolone" but you didn't specify from where to where. The distance from a remote hydrogen to each of the oxygens and to the hydroxyl hydrogen are shown below. You can do the measurements you want yourself here: http://ursula.chem.yale.edu/~chem220/chem220js/STUDYAIDS/ESPotential/ESPmisc2.html

| improve this answer | |

According to ChemDoodle's ruler tool, this molecule is probably about 5.35 Angstroms across.enter image description here

I'm a little annoyed that the font doesn't seem to have a proper Angstrom symbol. The symbol is part of the symbol selection widget, but doesn't get used by the ruler tool. Oh well, still cheaper than ChemDraw.

EDIT: The angstrom symbol issue is font related. If I use Arial, it fails to display. But if I switch to Times New Roman, it shows up fine.

| improve this answer | |
  • 2
    $\begingroup$ This is a cool use of ChemDoodle's ruler tool. The ruler tool in 2D uses the Angstroms/bond length and default bond length styles from the style sheet to calculate distances. It is typically used with imported 3D structures where coordinates are already calculated. You can also use ChemDoodle 3D measurement tools to measure distances after optimizing structures in 3D with force fields. $\endgroup$ – iclkevin Oct 21 at 20:23
  • $\begingroup$ If you use «Arial Unicode MS» instead of «normal Arial», then you get the Angstrom comme il faut. And it is not the only font (if working in Windows) supporting this sign (examples in chat.stackexchange.com/rooms/3229/the-periodic-table). $\endgroup$ – Buttonwood Oct 22 at 20:33
  • 1
    $\begingroup$ This seems to neglect a few hydrogen though. $\endgroup$ – Martin - マーチン Oct 22 at 20:59
  • $\begingroup$ @Martin: True. Even worse, though, chemically speaking, is that the drawing of the tropolone may be scaled, while drawing the molecule according e.g., policies of a journal / style sheet the program uses implicitly by default, or explicitly to sooth wacky bonds (if one doesn't use a template for a 7-membered ring). But because the molecule's structure is one graphical object on the canvas of the sketcher, and the ruler an other, the two may be scaled independently from each other, too. For me, this is like the «poof, cinq» by Hugh Laurie and Helen Mirren, back in 2006 at the Emmy awards. $\endgroup$ – Buttonwood Oct 22 at 21:17
  • 1
    $\begingroup$ @Buttonwood (I don't know this skid, but I can imagine it. Will google.) As a rough guesstimate it may still work and it's pretty much a computerised version of drawing it by hand using standard bond lengths. It's rough which you can see, as the other values come in around 580 - 600 pm. I would suggest a more robust technique for uses other than graphical design. $\endgroup$ – Martin - マーチン Oct 22 at 21:23

A quick way to do this would be to enter the molecule in name in the search form on ChemSpider (maintained by the UK Royal Society of Chemistry). Select the 3D representation by hitting the 3D button. In the JSMol window, you can select an atom by double clicking it. To get the largest atom-to-atom distance, I've selected the hydroxyl hydrogen. By moving the mouse to the other atoms, you can see their distances. The largest distance appears to be 0.588 nm.

ChemSpider page for tropolone

The atoms themselves have size owing to their electron clouds, so a better measure of the size of the molecule could include the van der Waals radius of the atoms, which you can find a value for on the Wikipedia page. Hydrogen has a radius of 0.11 nm, so the total distance would be 0.558 nm plus the van der Waals radius of the two hydrogen atoms, which would be 0.778 nm.

This is only one way to characterize the size. More complex ways of characterizing the size would be better done in a dedicated program. You can save structure as a .mol file from ChemSpider by clicking the disk icon below the 3D image. Conformers can also be downloaded in different formats from PubChem (maintained by the US National Institutes of Health), which offers many of the same features as ChemSpider.

Commonly used standalone programs for visualizing molecular structures are the programs VMD, PyMol, and Avogadro. They are open source and completely free (Avogadro is GPL'd) or free for educational use (VMD, PyMol). There are many tutorials for using these programs. PyMol will open the .mol file natively, but VMD will not, so you can use OpenBabel to convert the .mol file to a .mol2 file:

obabel tropolone.mol -O tropolone.mol2

Then I open it in VMD: vmd tropolone.mol2

You can measure distances between pairs of atoms by hitting the "2" key and clicking atoms. Note that these distances are in angstroms (Å).

Measuring distances in VMD

VMD will give you the atom-to-atom size along the x, y, and z axis using the following commands:

mol new tropolone.mol2
set s [atomselect top all]
set m [measure minmax $s]
puts [vecsub [lindex $m 1] [lindex $m 0]]

The results are 5.42 Å along the x-axis, 5.20 Å along the y-axis, and 0.07 Å along the z-axis.

| improve this answer | |

Since we're comparing: molden has a zmat builder and a force field (amber) implemented. You can start with the obvious choices of getting a structure from a database, or generating a starting structure with Open Babel, or simply click it together. The result will come to something like this (I admit this is even less prettier than VMD, but you can use some custom settings in gmolden to make that look nicer, too):

tropolone molden image

Obviously, better results (i.e. more accurate estimates) can be achieved with better calculation techniques. A quick and open source option is xtb, which calculates the molecule on the semi-empirical level of theory.

If you want to go full quantum calculation, free options include orca and gamess. I'd say both are reasonably simple to set up and use. There are plenty of more options. For general recommendations I suggest looking through our Mater Modeling partner site's Q&A.

| improve this answer | |

Another nice free (but limited in function) tool for this job is NIH PubChem's viewer:

enter image description here

The computed distance (not clearly visible in the figure) is $\pu{6.10 Å}$

The structure is also from PubChem:


You can alternately process the coordinates with a tool such as Matlab/Octave:

data=[-2.1849   -1.2310   -0.2189 16   0  0  0  0  0  0  0  0  0  0  0  0
   -1.8861    1.4126    0.3330 16   0  0  0  0  0  0  0  0  0  0  0  0
   -0.9119   -0.6926   -0.0609 12   0  0  0  0  0  0  0  0  0  0  0  0
   -0.8564    0.7909    0.0658 12   0  0  0  0  0  0  0  0  0  0  0  0
    0.1188   -1.5452   -0.0243 12   0  0  0  0  0  0  0  0  0  0  0  0
    0.3789    1.5596   -0.2139 12   0  0  0  0  0  0  0  0  0  0  0  0
    1.5156   -1.2996    0.2099 12   0  0  0  0  0  0  0  0  0  0  0  0
    1.6492    1.1479   -0.2156 12   0  0  0  0  0  0  0  0  0  0  0  0
    2.1767   -0.1426    0.1249 12   0  0  0  0  0  0  0  0  0  0  0  0
   -0.0939   -2.6096   -0.1302 1   0  0  0  0  0  0  0  0  0  0  0  0
    0.1974    2.6114   -0.4310 1   0  0  0  0  0  0  0  0  0  0  0  0
    2.0899   -2.1870    0.4717 1   0  0  0  0  0  0  0  0  0  0  0  0
    2.4120    1.8907   -0.4446 1   0  0  0  0  0  0  0  0  0  0  0  0
    3.2469   -0.1561    0.3232 1   0  0  0  0  0  0  0  0  0  0  0  0
   -2.1177   -2.1985   -0.2883 1   0  0  0  0  0  0  0  0  0  0  0  0];
xyz = data(:,1:3);

dist2=sum((permute(mat,[2 1 3])-mat).^2,3); % compute distances between all atoms 
[d2max,imax]=max(dist2); % find max value (column by column) 
[dd2max,jmax]=max(d2max); % find absolute max

disp(['max distance:' sprintf('%10.4f',sqrt(dd2max))])
| improve this answer | |
  • $\begingroup$ @andselisk PubChem's viewer also allows visualization of an enclosing box but I haven't figured out what units are used. They are not consistent with the otherwise reasonable estimate of the longest interatomic distance. Unfortunately documentation is difficult to find. $\endgroup$ – Buck Thorn Nov 15 at 11:25

The crystal structure of tropolone has been determined [1] and is freely available via CCDC (CSD Entry: TROPOL10). One can use JSmol on the web page to find the linear parameters (RMB → Measurements → Click for distance measurement).

I would, however, suggest to use Olex2. Choosing two atoms of interest, say, H1 and H4, followed a line command, yields in linear parameter of 5.992(2) Å:

TROPOL10 - line

The issue with this approach is that atoms are treated as point particles (other answers use the same approach), which is not entirely correct. When it comes to the size of a molecule, involving a space-filling model based on van der Waals radii would be more appropriate. Conveniently, Olex2 has a tool for that too — a wbox command which

Draws a rectangular Wrapping BOX around the chemical moieties on the screen. The vdW radii employed in the computation are printed.

Like such:

TROPOL10 - wbox

Using the following element radii:
(Default radii source: http://www.ccdc.cam.ac.uk/products/csd/radii)
C   1.7
H   1.09
O   1.52
Wrapping box dimension: 3.453 x 7.348 x 8.156 A
Wrapping box volume: 206.955 A^3

meaning the length across is actually about 8.2 Å.


  1. Shimanouchi, H.; Sasada, Y. The Crystal and Molecular Structure of Tropolone. Acta Cryst B Struct Crystallogr Cryst Chem 1973, 29 (1), 81–90. DOI: 10.1107/S0567740873002013.
| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.