Remove extra space generated by a LaTeX command

Question

My LaTeX document contains the following lines:

\documentclass{article}

\usepackage{etoolbox}

\newcommand{\defineSpecies}[2]{\csdef{spec@#1}{#2}}
\newrobustcmd{\spec}[1]{%
  \ifcsname spec@#1\endcsname
    \csuse{spec@#1}
  \else
    \GenericError{}{Undefined species `#1'}{}{}
  \fi
}
\defineSpecies{h216o}{H$_2^{~16}$O}

\begin{document}
\spec{h216o}. \spec{h216o} molecule
\end{document}

When compiling this document, an extra space (denoted with a red box) is put before the dot in the PDF file:

Could you please help me to redefine \spec{} that (a) it does not provide the red space before the dot, but (b) it includes the blue space before the world molecule.

Please focus on repairing this command instead of suggesting packages, like mchem.

Answer 1

Add % at the end of the lines.

\documentclass{article}

\usepackage{etoolbox}

\newcommand{\defineSpecies}[2]{\csdef{spec@#1}{#2}}
\newrobustcmd{\spec}[1]{%
    \ifcsname spec@#1\endcsname
    \csuse{spec@#1}% <<<<
    \else
    \GenericError{}{Undefined species `#1'}{}{}% <<<<
    \fi
}
\defineSpecies{h216o}{H$_2^{~16}$O}

\begin{document}
    \spec{h216o}. \spec{h216o} molecule
\end{document}

Answer 2

Apart from the missing % to mask endlines, you might exploit the chemformula package:

\documentclass{article}

\usepackage{etoolbox}
\usepackage{chemformula}

\newcommand{\defineSpecies}[2]{\csdef{spec@#1}{\ch{#2}}}
\newrobustcmd{\spec}[1]{%
  \ifcsname spec@#1\endcsname
    \csuse{spec@#1}%
  \else
    \GenericError{}{Undefined species `#1'}{}{}%
  \fi
}
\defineSpecies{h216o}{H2\,{^{16}}O}
\defineSpecies{caco3}{CaCO3}

\begin{document}

\spec{h216o}. \spec{caco3} molecule

\end{document}

A possible improvement is to use a “centralized” species definition.

\documentclass{article}

\usepackage{chemformula}

\ExplSyntaxOn

\prop_new:N \g_tobir_species_prop
\cs_generate_variant:Nn \chemformula_ch:nn { ne }

\NewDocumentCommand{\defineSpecies}{m}
 {
  \prop_gset_from_keyval:Nn \g_tobir_species_prop { #1 }
 }
\NewDocumentCommand{\spec}{m}
 {
  \prop_if_in:NnTF \g_tobir_species_prop { #1 }
   {
    \chemformula_ch:ne {} { \prop_item:Nn \g_tobir_species_prop { #1 } }
   }
   {
    \GenericError{}{Undefined~species~`#1'}{}{}
   }
 }

\ExplSyntaxOff

\defineSpecies{
  h216o = H2\,{^{16}}O,
  caco3 = CaCO3,
}

\begin{document}

\spec{h216o}. \spec{caco3} molecule

\end{document}

Answer 3

There might be good reasons for you to use etoolbox, but to write chemical equations, you do not need it. Like highly recommended for the body of questions/answers/comments on sibling chemistry.se, you may simply use mhchem (entry CTAN) to yield, e.g.

with an easier to read MWE like

\documentclass{article}
% \usepackage{etoolbox}  % neither requirement for, no inhibitor of a compilation
\usepackage[version=4]{mhchem}

\begin{document}
Our water is labelled as $\ce{H2{}^{16}O}$. $\ce{CaCO3}$ then is added.
\end{document}

The explicit call for version 1, 2, or 3 yield output with perhaps slightly different kerning; for the problem presented in your question, this however is irrelevant. To use {} past H2 prevents 2 to be misunderstood as a mass number of oxygen instead of an atom multiplier of hydrogen; the manual explicitly mentions this assistance by the user to the usepackage.

In addition to the project's own documentations, chemistry.se showcases some peculiarities in a dedicated FAQ, too.