{\overfullrule0pt\def\cite#1{[#1]} % Warning: This is an input file for AMSTeX, rather than for LaTeX. % This is just so that I can use the commands for arrows % mentioned on pages 140-141 of "The Joy of TeX". % It may not be "good AmSTeX", since it's the first time % I've used AmSTeX seriously! \title{Could \LaTeXbf\ do more for chemists?} From time to time I have seen queries in {\TeXhax} and elsewhere from people who wish to use \LaTeX\ for documents that involve chemical formulae, etc. Since there are now plans for \LaTeX\ 3.0 \cite{1}, I thought this might be a good time to consider what facilities one could reasonably ask the people implementing the new \LaTeX\ to provide for chemists. I'm not a chemist myself, but have been involved in helping chemists use \LaTeX\ 2.09 \cite{2} for producing theses, etc. \section{Typesetting and artwork} It may help to consider the division of labour within a traditional publishing house. A `copy-editor' \cite{3, p236} might divide the work on chemical formulae up between `the printer' and `the draughtsman'. Thus, some formulae can be typeset, but others are treated as `artwork' and are drawn. Although there have been valiant attempts \cite{4-5} to define \TeX\ macros for drawing chemical structure diagrams, I think it is inevitable that, whatever \TeX\ macros are defined, there will be chemists who come along with requirements that are beyond the abilities of the macro packages. Therefore, it seems sensible to retain the traditional division between typesetting and artwork: to typeset those formulae that can be typeset easily and to get the other formulae drawn in some way. In \TeX\ terms, artwork can be treated as `graphics' to be `pasted' into a typeset document via |\special|. For example, {\sans Chemdraw} \cite{6} can produce `encapsulated {\PS}', so a \LaTeX-ed document with {\sans Chemdraw} diagrams can be printed on a {\PS} printer. \section{Desirable facilities for chemists} \LaTeX\ is never going to be a system for producing `artwork', so it seems to me that it would be unwise to attempt to provide comprehensive facilities for `chemical artwork' in \LaTeX. However, it does seem worth providing a limited number of facilities to make it easier to produce the chemical formulae that should be treated as `typesetting'. \subsection{Environments} Chemists' problems start when they use \LaTeX's `mathematics' environments for `chemistry'. Chemistry is not mathematics, and the conventions for typesetting chemistry are different from those for typesetting mathematics. In terms of the \LaTeX\ philosophy \cite{2, p6}, `mathematics' and `chemistry' represent `logically distinct structural elements'. It would seem natural to: \bi define environments for chemists that are analogous to the environments that are available for mathematicians \bi within these `chemistry' environments, aim to keep to whatever typesetting conventions are usual in chemistry. \noindent How about defining {\tt chem}, {\tt displaychem} and {\tt chem\-equation} environments, by analogy with {\tt math}, {\tt displaymath} and {\tt equation}? If such environments were defined, the style-file writer would then have control over `mathematics' and `chemistry' separately. In particular: \bi The default would be {|\rm|} for chemistry (although a designer could change the default in a {\tt .sty} file). Individual authors would no longer have to search through ``double bend'' sections of the \TB\ themselves \cite{7, pp163 \& 179}. \bi A designer could implement a house-style in which mathematical and chemical equations are numbered in separate sequences \cite{3, p224} or a house-style in which there is only one sequence of numbers \cite{8, p32}. \bi It might be possible to arrange that subscripts will normally be at the same level \cite{7, p179} inside the environments for chemistry. \subsection{Commands} At first sight, the \LaTeX\ manual \cite{2, ch.\ 3} gives the impression that \LaTeX\ 2.09 provides the `building blocks' to give all the arrows, harpoons and annotation that a chemist could want. However, it is often difficult to get these building blocks arranged in the ways required. For example: \bi How does one obtain CH${_3}$(C=O)OCl \cite{3, p235} in `math mode'? We can't use `=' to mean ``double bond'', since \TeX\ puts space around it. \bi To represent a reversible reaction with rate constants above\slash below a pair of harpoons, I ended up with \begintt \renewcommand{\arraystretch}{0.5} A \begin\{array}{c} \scriptstyle k_1\\[1mm] \rightleftharpoons\\ \scriptstyle k\_2 \end{array} B \endtt before it looked right. Surely individual \LaTeX\ users shouldn't have to re-do the `tuning' needed to get these things right? \bi As in the above example, arrows and harpoons are often labelled to show reaction conditions. It is not clear how to get arrows\slash harpoons that expand to the width of the labels. \noindent Many of these difficulties are another consequence of trying to use, for chemistry, the structural elements that were designed for mathematics. So what commands might usefully be made available inside some future `chemistry' environments? It seems desirable \cite{3, p237} to have documented facilities for single and double bonds. Triple bonds might also be needed \cite{9}. A documented facility for representing single bonds by raised dots would also be useful \cite{10, p59}. Might commands such as |\bond|, |\doublebond| and |\triplebond| be appropriate? It seems desirable to have specific commands for typesetting arrows\slash harpoons with labels above\slash below (to indicate conditions or rate constants). An indication of the combinations of arrows\slash harpoons that have been typeset traditionally is given in \cite{11,~p371}. Thus, as well as providing simple arrows for one-way reactions, it might be worth aiming to provide commands for: equilibrium reactions (beginning at left and right); reversible reactions (beginning at left and right); reactions beginning at left\slash right and completed to left\slash right. Might it be worth defining some commands such as \centerline{\vbox{\tt\halign{\char'134#\quad&\char'134#\cr oneway\cr equilibriumR &equilibriumL\cr reversibleR &reversibleL\cr rightright &rightleft\cr leftright &leftleft\cr }}} \noindent that each accept two parameters: one to give a label above the symbol, the other to give a label below the symbol? For example, \begintt \begin{chem} 2H_2 + O_2 \oneway{catalyst}{300 K; 4 bar} 2 H_2O \end{chem} \endtt might be a natural way to specify % This is the line that contains the command that made me use AMSTeX. % I couldn't find anything else ready-made in plain TeX or LaTeX. \newdimen\bigaw \def\oneway>#1>#2>{% \setbox0\hbox{$\scriptstyle#1$} \setbox1\hbox{$\scriptstyle#2$} \bigaw\wd0\ifdim\wd1>\bigaw\bigaw\wd1\fi \mathrel{\mathop{\hbox to\bigaw{\rightarrowfill}}\limits^{#1}_{#2}}} $${\rm2H}_2 + {\rm O}_2 \oneway >{\rm catalyst} > {\rm 300 K; 4 bar} > {\rm 2H}_2{\rm O} $$ \section{Work involved} I'd guess that my suggestions about environments could be implemented by slight modications of the code that implements the corresponding environments for mathematics. Some new work would be required for commands such as |\oneway|, \dots\ , |\leftleft|. The only similar facility that I've found in an existing macro package is that for arrows in \AmSTeX\ \cite{12, p140}, but (from a chemist's point-of-view) this doesn't provide sufficient choice of symbols. It might be worth seeking advice about objectives from people who typeset chemistry professionally. % Perhaps the UK's Royal Society of Chemistry would be prepared to advise % about details. I've been in e-mail correspondence with their Publications % Manager about authors submitting mansuscripts electronically to them, % although they seem to be going the wordprocessor + chemdraw + SGML route. \section{Documentation} If the facilities that I've outlined above were provided, I would be inclined to give them less prominence in the documentation than the analogous facilities for mathematicians. This would help give the impression that, whereas mathematicians can expect \LaTeX\ to do everything they want, chemists can only expect \LaTeX\ to do a certain proportion of what they want. For example, in the successor to \cite{2}, the description of facilities for chemists might be relegated to an appendix, which could start with a paragraph explaining that the facilities are intended to support `typesetting' but not `artwork'. \section{Bibliography}{\parindent0pt\frenchspacing\raggedright \everypar{\hangafter1\hangindent20pt} \par [1] Frank Mittelbach and Rainer Sch\"opf, 1989, With \LaTeX\ into the Nineties TUGboat 10, 681--690. [2] Leslie Lamport, 1986, \LaTeX: A Document Preparation System, Addison-Wesley. [3] Judith Butcher, 1981, Copy-editing Cambridge University Press, [4] Roswitha T. Haas and Kevin C. O'Kane, 1987, Typesetting chemical structure formulae with the text formatter \TeX\slash \LaTeX, Computers and Chemistry 11, 251-271. [5] Michael Ramek, 1990, Chemical structure formulae and $x/y$ diagrams with \TeX\ {\it in:} \TeX: applications, uses, methods, Malcolm Clark (editor), Ellis Horwood. [6] {\sans Chemdraw} (A program for the Macintosh), Cambridge Scientific Computing, Cambridge, Massachusetts. % [7] Donald E. Knuth, 1986, The \TeX book, Addison-Wesley. [8] Janet S. Dodd, 1986, The ACS Style Guide American Chemical Society [9] Handbook for Chemical Society Authors, 1961, The Chemical Society [10] Hart's Rules, 1983, Oxford University Press 1983 [11] Chicago Manual of Style, 1982, Chicago University Press % \publaddr Chicago [12] Michael Spivak, 1986, The Joy of \TeX, American Mathematical Society }} % \author{David Rhead}