charges and radicals

In the default font size in MarvinSketch, it is very difficult to distinguish a minus sign (e.g., C-) from a lone electron of a radical (e.g., C·) with the naked eye. One is two pixels wide, the other is one pixel. IMHO, it would be much better if Marvin used a bullet (•, a big, fat dot) for the radical and an en dash (–, longer than a hyphen) for the minus sign.





Also, it is customary, when the letter C is used to represent a C atom, to also show all the H atoms attached to that C. You always show the C when the C is charged or has a radical, but you show the H atoms attached to that C only when the implicitH tag is set to do so. IMHO, the implicitH tag should be ignored for C atoms that are charged or have unshared electrons, and the implicit H atoms should always be shown on those C atoms. Otherwise, it looks very confusing.





Finally, Marvin does a great job in rectifying H atoms as charges are added to atoms and underlining atoms with incorrect valences. But it fails occasionally. For example, if I draw a tetravalent, uncharged N, the N atom is not underlined, though it should be. Also, if I draw a pentavalent, negatively charged C atom, the C atom is not underlined, even though it has far exceeded its correct valence of 4. I understand why the pentavalent negative C is mathematically OK, but chemically it is not OK.

P.S. Instead of showing the implicit H atoms of C• or C(-), you could omit the letter C and just show the (-) or • next to the vertex representing the C atom. But I think it's better to show the C and the implicit H atoms.

Quote:

In the default font size in MarvinSketch, it is very difficult to distinguish a minus sign (e.g., C-) from a lone electron of a radical (e.g., C·) with the naked eye. One is two pixels wide, the other is one pixel. IMHO, it would be much better if Marvin used a bullet (•, a big, fat dot) for the radical and an en dash (–, longer than a hyphen) for the minus sign.

The bullet issue has been solved. Please see the latest test version:


http://www.chemaxon.com/test/marvin/doc/dev/example-sketch1.1.html width="90%" cellspacing="0" cellpadding="3" border="0" align="center"> Quote: Also, it is customary, when the letter C is used to represent a C atom, to also show all the H atoms attached to that C. You always show the C when the C is charged or has a radical, but you show the H atoms attached to that C only when the implicitH tag is set to do so. IMHO, the implicitH tag should be ignored for C atoms that are charged or have unshared electrons, and the implicit H atoms should always be shown on those C atoms. Otherwise, it looks very confusing.


Sounds reasonable. We will solve that.

Quote:

Finally, Marvin does a great job in rectifying H atoms as charges are added to atoms and underlining atoms with incorrect valences. But it fails occasionally. For example, if I draw a tetravalent, uncharged N, the N atom is not underlined, though it should be. Also, if I draw a pentavalent, negatively charged C atom, the C atom is not underlined, even though it has far exceeded its correct valence of 4. I understand why the pentavalent negative C is mathematically OK, but chemically it is not OK.

We will fix this, too.


Thanks,


Ferenc

Ferenc wrote:

The bullet issue has been solved. Please see the latest test version: http://www.chemaxon.com/test/marvin/doc/dev/example-sketch1.1.html

Doesn't look any different to me....





Oh, another bug. If you draw a bond starting at an element that has a radical or charge, the radical or charge is erased. It shouldn't be.

Quote:

Doesn't look any different to me....

Please restart the browser. When you click on "?" the version number is displayed. It should be MarvinSketch 3.4pre5 or later

Quote:

Oh, another bug. If you draw a bond starting at an element that has a radical or charge, the radical or charge is erased. It shouldn't be.

It will be fixed.

Ferenc wrote:

The bullet issue has been solved. Please see the latest test version: http://www.chemaxon.com/test/marvin/doc/dev/example-sketch1.1.html

bobgr wrote:

Doesn't look any different to me....

Ferenc wrote:

Please restart the browser. When you click on "?" the version number is displayed. It should be MarvinSketch 3.4pre5 or later

Yes, it is version 3.4pre5, but the radical is still just one pixel in size and hard to distinguish from a minus sign. The font is SansSerif size 12 (the default), and I'm using Safari 1.2.2 on MacOS 10.3.4.





Remember, I also suggest that you use an en dash (–) instead of a hyphen (-) for the minus sign. The en dash is longer and easier to distinguish from a dot.

bobgr wrote:

Oh, another bug. If you draw a bond starting at an element that has a radical or charge, the radical or charge is erased. It shouldn't be.

Ferenc wrote:

It will be fixed.

I need to clarify this bug report. In 3.4pre5, if you draw a bond by clicking on a charged atom (or radical) with the bond tool, the charge is not erased. However, if you draw a bond by clicking and holding, then dragging, with either the bond tool or an atom button, then the charge is erased.

You should also know that I tested MarvinSketch/AWT 3.4pre5 with Netscape 4.77 on MacOS 9.2, and when I made a radical, I got a question mark instead of either a dot or a bullet!





Sounds like we have font issues.

bobgr wrote:

In the default font size in MarvinSketch, it is very difficult to distinguish a minus sign (e.g., C-) from a lone electron of a radical (e.g., C·) with the naked eye. One is two pixels wide, the other is one pixel. IMHO, it would be much better if Marvin used a bullet (•, a big, fat dot) for the radical and an en dash (–, longer than a hyphen) for the minus sign.

Ok, it seems we found the proper unicode characters for bullet ('\u2022') and en dash ('\u2013')


http://www.fileformat.info/info/unicode/char/search.htm


We plan to use these characters for lone electron and negative charge. They work under Linux, Windows, Mac OS 9, and Mac OS X


Thanks a lot for the help!


Ferenc

Ferenc wrote:

Ok, it seems we found the proper unicode characters for bullet ('\u2022') and en dash ('\u2013') http://www.fileformat.info/info/unicode/char/search.htm We plan to use these characters for lone electron and negative charge. They work under Linux, Windows, Mac OS 9, and Mac OS X Thanks a lot for the help! Ferenc

I just checked out Marvin3.4pre6 on Netscape 4.77 and Safari 1.2.2, and the bullets and en dashes are very visible. It's much improved. Very nice.





Don't forget to show implicit Hs on charged and radicalized C atoms, and don't forget to fix the bug where the charge or radical is erased upon drawing a new bond.

bobgr wrote:

Don't forget to show implicit Hs on charged and radicalized C atoms, and don't forget to fix the bug where the charge or radical is erased upon drawing a new bond.

Ok, but these will probably be fixed in a later release, not in the coming one.


Best regards,


Ferenc

bobgr wrote:

Don't forget to show implicit Hs on charged and radicalized C atoms, and don't forget to fix the bug where the charge or radical is erased upon drawing a new bond.

These issues have been solved. The next release coming in a few days will contain the fix.

I'm not a chemist, but as a student, its confusing to me when I see a CH+ (H is implicit H displayed along with a C+). Should it be C+H or CH+ ?

Although, the formal charge belongs to the carbon, it is never drawn between the carbon atom symbol and its implicit hydrogens. The position of charge and radical symbols can vary in different molecule representation models, but they often appear right to the implicit hydrogens.


Please see this IUPAC page: http://www.iupac.org/goldbook/C00812.pdf





Anyway, this is not just an organic chemistry tradition, the inorganic hydroxyl anion OH- and ammonium cation NH4+ formats might be familiar for you.

Ferenc wrote:

Quote: Also, if I draw a pentavalent, negatively charged C atom, the C atom is not underlined, even though it has far exceeded its correct valence of 4. I understand why the pentavalent negative C is mathematically OK, but chemically it is not OK. We will fix this, too. Thanks, Ferenc

Hey guys,





In Marvin 3.5.0, a pentavalent, negatively charged C is not underlined in red if one or more of its bonds is a pi bond, only if all five bonds are sigma bonds.





-- Bob

bobgr wrote:

In Marvin 3.5.0, a pentavalent, negatively charged C is not underlined in red if one or more of its bonds is a pi bond, only if all five bonds are sigma bonds.

... and that, too

Ferenc wrote:

bobgr wrote: In Marvin 3.5.0, a pentavalent, negatively charged C is not underlined in red if one or more of its bonds is a pi bond, only if all five bonds are sigma bonds. ... and that, too

Still not fixed in Marvin 3.5.1....

bobgr wrote:

Still not fixed in Marvin 3.5.1....

It will come in Marvin 4. It was fixed just today (in the main branch and not in the release branch, so Marvin 3.5.x will not have the fix).