Article

Cite This: J. Chem. Educ. 2019, 96, 2467−2475 pubs.acs.org/jchemeduc

Representation of Three-Center−Two-Electron Bonds in Covalent

Molecules with Bridging Hydrogen Atoms
Gerard Parkin*
Department of Chemistry, Columbia University, New York, New York 10027, United States
*
S Supporting Information

ABSTRACT: Many compounds can be represented well in

terms of the two-center−two-electron (2c−2e) bond model.
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

However, it is well-known that this approach has limitations;

for example, certain compounds require the use of three-
center−two-electron (3c−2e) bonds to provide an adequate
Downloaded via UNIV DE ALICANTE on October 14, 2022 at 17:53:06 (UTC).

description of the bonding. Although a classic example of a

compound that features a 3c−2e bond is provided by
diborane, B2H6, 3c−2e interactions also feature prominently in transition metal chemistry, as exemplified by bridging hydride
compounds, agostic compounds, dihydrogen complexes, and hydrocarbon and silane σ-complexes. In addition to being able to
identify the different types of bonds (2c−2e and 3c−2e) present in a molecule, it is essential to be able to utilize these models to
evaluate the chemical reasonableness of a molecule by applying the octet and 18-electron rules; however, to do so requires
determination of the electron counts of atoms in molecules. Although this is easily achieved for molecules that possess only 2c−
2e bonds, the situation is more complex for those that possess 3c−2e bonds. Therefore, this article describes a convenient
approach for representing 3c−2e interactions in a manner that facilitates the electron counting procedure for such compounds.
In particular, specific attention is devoted to the use of the half-arrow formalism to represent 3c−2e interactions in compounds
with bridging hydrogen atoms.
KEYWORDS: Second-Year Undergraduate, Graduate Education/Research, Inorganic Chemistry, Analogies/Transfer,
Problem Solving/Decision Making, Covalent Bonding, Lewis Structures, Main-Group Elements, Organometallics, Transition Elements

■ INTRODUCTION
The octet1,2 and 18-electron2a,3 rules, together with the
deficient”5,7 to indicate that there is an insufficient number of
electrons for all bonds to be 2c−2e.
construction of so-called “Lewis structures”,4 are of pivotal The structures of so-called “electron deficient” compounds
importance because they provide a rudimentary means to may, however, be rationalized by the use of multicenter
evaluate the chemical reasonableness of a molecule. Central to bonding descriptions, of which the three-center−two-electron
the application of these rules is the necessity of determining (3c−2e) model is the simplest.8−10 In view of this explanation
the electron count (or electron number, EN) of the atoms in
the molecule of interest. For many molecules (e.g., H2 and
CH4), the bonding can be described well in terms of a two-
center−two-electron (2c−2e) model, and the electron counts
of atoms in such molecules are typically easily evaluated from
the Lewis structure. In contrast, the bonding in some
molecules cannot be described by using only 2c−2e bonds,
as exemplified by those with bridging hydrogen atoms, of
which diborane, B2H6 (Figure 1),5,6 is an excellent example.
For example, whereas ethane, C2H6, has a total of 14 valence
Figure 1. Comparison of the structures of C2H6 and B2H6. C2H6 has
electrons derived from two carbon atoms and six hydrogen
14 valence electrons, which is sufficient to form seven 2c−2e bonds,
atoms (i.e., (2 × 4) + (6 × 1)) and can hence possess seven whereas B2H6 has only 12 valence electrons and can only form a
2c−2e bonds to give rise to the familiar structure shown in maximum of six 2c−2e bonds; therefore, B2H6 cannot adopt the same
Figure 1, B2H6 has a total of only 12 valence electrons (i.e., (2 structure as C2H6, and multicenter bonding is required to rationalize
× 3) + (6 × 1)) and can therefore only support a maximum of its structure. The dotted lines connecting the bridging hydrogen and
six 2c−2e bonds; as such, the number of electrons is boron atoms are intended to indicate that these are not 2c−2e bonds.
insufficient for B2H6 to adopt a structure akin to that of
C2H6. Therefore, B2H6 adopts a different type of structure in Received: August 10, 2019
which two of the hydrogen atoms bridge the two boron centers. Revised: September 7, 2019
Compounds of this type are often referred to as “electron Published: October 9, 2019
© 2019 American Chemical Society and
Division of Chemical Education, Inc. 2467 DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 2. Molecular orbital diagram for H2. The bonding orbital (σ)
is occupied by two electrons, so each hydrogen is assigned a duet
configuration. The + and − symbols on the orbitals refer to their
relative phases, as do their colors.
of the bonding, Rundle once quipped, “There are no such
things as electron deficient compounds, only theory deficient
chemists.”11,12
However, although the bonding in “electron deficient”
compounds such as B2H6 may be understood by the
application of theoretical methods, this approach lacks the
convenience of using simple electron counting procedures to
evaluate the chemical reasonableness of a covalent molecule.
Unfortunately, the electron counts for atoms in molecules with
3c−2e interactions are not as easily determined as those with
2c−2e interactions because (i) the structural representations
employed often lack clarity, and (ii) there is also the need to
evaluate the electron count of more than one atom. The
purpose of this article is, therefore, to describe how to
represent the bonding in molecules with 3c−2e bonds in such
a manner that it is simple to evaluate the electron count of
each atom that is involved in the interaction, and thereby
utilize effectively the octet and 18-electron rules for classes of
molecules such as bridging hydride, dihydrogen, and agostic
complexes, which are often encountered in undergraduate
inorganic chemistry courses.13
Figure 3. Covalent bond classification (CBC) of L, X, and Z ligands. Each partner provides one electron for a normal covalent bond (b), whereas a
single partner provides both electrons for a dative covalent bond (a, c).
2468
Article
Figure 4. Two alternative representations for a dative bond, as
illustrated for H3NBH3. Despite the different appearances, the
representations are not resonance structures because the electrons
are in the same location (recall that resonance structures differ in the
locations of the electrons).
■ REVIEW OF THE COVALENT BOND
CLASSIFICATION AND THE REPRESENTATION OF
2C−2E BONDS
Prior to discussing 3c−2e interactions, it is first pertinent to
review the representation of 2c−2e bonds. The molecular
orbital description of a 2c−2e bond is straightforward: the
overlap of two atomic orbitals on adjacent atoms results in the
formation of bonding and antibonding orbitals, of which the
former is occupied by two electrons, as illustrated by H2
(Figure 2). However, although this is a common feature of all
2c−2e bonds, there are two fundamentally different types of
covalent bond that are differentiated by the number of
electrons that each partner contributes to the interaction
(Figure 3). Thus, a normal covalent bond is one in which each
neutral partner contributes one electron to the bonding
orbital,14 whereas a dative covalent (or coordinate) bond is
one in which a single atom provides both electrons.15
Common examples of species that provide one electron to a
bond are H, Cl and CH3 (i.e., radicals), whereas species that
provide two electrons are illustrated by H2O, NH3, PR3, and
CO (i.e., neutral closed shell molecules with lone pairs).
Normal covalent and dative covalent bonds are represented
by different symbols. Thus, a normal covalent bond is depicted
as a line between two atoms, whereas a dative covalent bond is
represented as an arrow with the head pointing to the acceptor
atom (Figure 4). The latter can also be denoted by an
equivalent form that uses a line but also depicts formal charges
on the atoms, as illustrated for H3NBH3 (Figure 4).16,17 It
must be emphasized that the two depictions of the dative
covalent bond are not resonance structures because the
electrons are in the same location (i.e., between N and B;
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 4); as such, the two depictions are just different
representations of the same bonding description, and the form
that is used is often just a matter of convenience.
The covalent bond classification18 provides a means to
categorize covalent molecules and, for this approach, ligands
are classified according to the number of electrons that they
contribute to the bonding orbital. Thus, a neutral two-electron
donor is classified as L, a one-electron donor is X, and a zero-
electron donor is Z (Figure 3).
Many ligands possess more than one L, X or Z function, or
combination thereof, and in such cases, the ligand is classified
as [LlXxZz], where l, x, and z are the respective numbers of L,
X, and Z functionalities. For example, an η6-benzene ligand is
classified as L3 on the basis that each of the “double bonds”
may be conceptually viewed as an L donor akin to that of C2H4
(Figure 5). Correspondingly, an η5-cyclopentadienyl ligand is
Figure 5. [LlXxZz] classifications of two common multidentate
ligands. A cyclopentadienyl ligand is classified as L2X because the
Lewis structure contains two double bonds (L) and an alkyl moiety
(X), whereas a benzene ligand is classified as L3.
classified as L2X on the basis that the cyclopentadienyl radical
may be conceptually considered to be composed of two
“double bonds” (each of which is an L donor) and an alkyl
radical (which is an X donor). In terms of overall electron
count, cyclopentadienyl and benzene are respectively consid-
ered to be five-electron and six-electron donors (Figure 5).
The electron count about an atom (M) in a molecule of the
type [MLlXxZz]Q±, where l, x, and z are the respective numbers
of L, X, and Z functions, and Q± is the charge on the molecule,
is, therefore, given by the expression EN = m + 2l + x − Q±,
where m is the number of valence electrons on a neutral M
atom. For the transition metals in groups 3−11,19 m
corresponds to the Periodic Table group number, (e.g., Sc,
3; Ti, 4; V, 5; etc.), whereas for the main group elements in
groups 12−18, the value of m is the last digit.
As an illustration of the determination of the electron count,
the carbon atom in CH4 is classified as MX4 and therefore
possesses an octet configuration (i.e., EN = 4 + (4 × 1) = 8,
Figure 6). A simple example of a molecule that possesses a
Figure 6. Procedure for determining the electron counts (EN) for
atoms in some simple compounds (EN = m + 2l + x − Q±).
2469
Article
dative covalent bond is provided by the ammonia−borane
adduct, H3N→BH3, in which the boron is classified as MLX3
and the nitrogen is classified as MX3Z. Thus, both the boron
(i.e., EN = 3 + 2 + (3 × 1) = 8) and nitrogen (i.e., EN = 5 + (3
× 1) = 8) possess octet configurations. Alternatively, a
procedure that is often more convenient for compounds with
multidentate ligands is to determine the electron count by the
expression EN = m + Σ(ligand electrons) − Q±, where
Σ(ligand electrons) is the sum of the contributions of
individual ligands. For example, the electron count for
Cp2TaH3 is EN = 5 + (2 × 5) + (3 × 1) = 18; likewise,
Cp2W(CO) also has an 18-electron configuration (Figure 7).
Figure 7. Alternative procedure for determining the electron counts
(EN) for atoms in some compounds with multidentate ligands (EN =
m + Σ(ligand electrons) − Q±).
■
MOLECULAR ORBITAL DESCRIPTION OF 3C−2E
INTERACTIONS AND STRUCTURE−BONDING
REPRESENTATIONS
The simplest molecule that possesses a 3c−2e interaction is
[H3]+, which has been spectroscopically observed but not
isolated. Both linear and triangular structures may be
considered, and the molecular orbitals for the two forms,
which are related by an orbital correlation diagram, are
illustrated in Figure 8.20 Since the system contains a single pair
of electrons, the preferred structure is determined only by the
occupancy of the fully bonding orbital (i.e., all orbitals in
phase), which is lowest in energy for the triangular form
because it maximizes the overlap of the three hydrogen 1s
Figure 8. Molecular orbitals for linear and triangular [H3]+. Each
hydrogen possesses a duet configuration. The + and − symbols on the
orbitals refer to their relative phases, as do their colors.
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 9. (a) Connectivity and (b) three equivalent structure−
bonding representations of [H3]+.
Figure 10. Half-arrow representation for a single 3c−2e interaction
for a bridging hydride compound (top) and comparison with the
representation for a bridging chloride ligand that has two 2c−2e
bonds because of the availability of lone pairs on chlorine (bottom).
orbitals. Since the pair of electrons in the bonding orbital of
[H3]+ is associated with all of the hydrogen atoms, each
hydrogen atom possesses a duet configuration; this situation is
analogous to both hydrogen atoms in H2 being assigned a duet
configuration because they share a pair of electrons (Figure 2).
Thus, with respect to electron counting, an important feature
of a 3c−2e interaction is that each atom is assigned a shared
pair of electrons. It should be noted that the same concept
would apply if the linear form of [H3]+ (or a bent form with an
H−H−H angle not equal to 60°) were to be more stable (i.e.,
each atom would still be assigned a duet configuration).
Rather than having to construct an MO diagram to
determine the electron counts for atoms involved in a 3c−2e
interaction, however, it is more convenient to use structure−
bonding figures that allow one to evaluate the electron counts.
Consider, for example, [H3]+ being derived from two hydrogen
atoms (HA and HB) and a proton (HC+), as illustrated in the
center structure of Figure 9. In valence bond terms, the [H3]+
molecule can be viewed as an adduct of H2 and H+, which
differs from a conventional adduct in the sense that the donor
pair of electrons is derived from an H−H bond rather than a
lone pair on a single atom. The bonding can, therefore, be
Figure 11. Representations of the 3c−2e B−H−B interaction in diborane. Coordination of BH3 by a B−H bond of another molecule allows boron
to achieve an octet configuration. The 3c−2e interaction can be represented by either an arrow from the center of the B−H bond or an equivalent
form in which a half-arrow is drawn from the hydrogen atom.
2470
Article
represented by the structures shown in Figure 9. Evaluation of
each hydrogen atom in turn illustrates how the duet
configuration may be derived. Thus, HA and HB have a duet
configuration because they share a pair of electrons, whereas
HC has a duet configuration because HC+ (with no electrons)
shares the pair of electrons in the HA−HB bond.
This approach for representing a 3c−2e interaction can also
be used if the three atoms are not part of an equilateral triangle
such that the interactions between all pairs of atoms are not
equal. However, an alternative way to represent this bonding
situation is to employ the so-called “half-arrow” notation that
was introduced in 1980 by Green for bridging hydride
compounds (Figure 10).21,22 With this notation, a half-arrow
is drawn from a central bridging hydrogen atom to an outer
atom. This representation is very similar to the well-known use
of a full-arrow for bridging halide compounds (e.g., Al2Cl6) to
indicate the donation of a lone pair of electrons (Figure 10);
the purpose of using a half-arrow (rather than a full-arrow) for
the bridging hydrogen atom is to emphasize that the arrow
does not correspond to donation of a lone pair but rather
donation of a bond pair.22 Otherwise, there is no difference in
the impact of a half-arrow or full-arrow on the electron count
of the outer atom.
It is important to emphasize that the two representations for
the 3c−2e interaction (i.e., a full-arrow from the center of the
bond and a half-arrow from the central atom) are intended to
convey the same information with respect to electron counting
purposes, and the form that is used is simply a matter of
convenience, similar to the manner in which a dative covalent
bond can represented as either A→B or A+−B−. For example,
the full-arrow from the center of the bond is most commonly
employed if the angle at the central atom is distinctly nonlinear
and the two “outer” atoms are considered to be within a
bonding distance (i.e., a “closed” system23), whereas the half-
arrow representation is used if the angle at the central atom is
significantly obtuse and there is little interaction between the
outer atoms.24
■ EXAMPLES OF 3C−2E INTERACTIONS
Molecules that feature 3c−2e interactions are rather common,
and the application of the above structure−bonding
representations to these molecules is illustrated below by
consideration of three classes of compounds, namely, (i)
bridging hydride compounds, (ii) agostic compounds, and (iii)
dihydrogen compounds.13
Bridging Hydride Compounds
Bridging hydride compounds constitute the largest class of
molecules with 3c−2e interactions, as exemplified by B2H6, as
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 12. Often encountered incorrect representation of B2H6 that
appears to depict two 2c−2e bonds to hydrogen.
noted above. The bonding in B2H6 may be considered in terms
of its formation by dimerization of BH3. Thus, whereas the
boron atom in BH3 has only a sextet configuration,
coordination by the pair of electrons in a B−H bond of a
second molecule results in an octet configuration (Figure 11).
The half-arrow representation provides a very convenient
means to evaluate the electron count of each boron center. It
also provides a clear depiction of the 3c−2e nature of the
interaction, in contrast to representations that simply draw two
lines from each hydrogen atom,25 which can thereby give the
misleading impression that hydrogen is forming two normal
covalent bonds (Figure 12).
Transition metal compounds also feature bridging hydride
ligands, and some examples are presented in Figures 13 and
14.26 For example, [(CO)5W(μ-H)W(CO)5]− can be depicted
by the two structure−bonding representations shown in Figure
13, in which both tungsten centers possess 18-electron
configurations.
Figure 13. Half-arrow bonding representations for [(CO)5W(μ-
H)W(CO)5]−. The 3c−2e interaction allows each tungsten center to
possess an 18-electron configuration without requiring an M−M
bond.
Figure 14. Half-arrow bonding representations for [(CO)4Re]2(μ-
H)2 and [CpReH2]2(μ-H)2. Whereas the two 3c−2e interactions in
[(CO)4Re]2(μ-H)2 allow the rhenium centers to achieve an 18-
electron configuration, [CpReH2]2(μ-H)2 achieves only a 17-electron
configuration; therefore, a Re−Re bond is required to obtain an 18-
electron configuration for the latter.
One aspect of evaluating the electron count of atoms
involved in an A−H−A 3c−2e interaction is that it provides a
2471
Article
means to determine whether any additional direct bonding
exists between the outer A atoms. For example, if A was a
transition metal, an apparent electron count of EN would
indicate that each A atom would require an additional (18 −
EN) A−A bonds in order to achieve an 18-electron
configuration, because each A−A bond would increase the
electron count by one. As an illustration, the two rhenium
atoms of [CpReH2]2(μ-H)2 would possess 17-electron
configurations, and so a direct Re−Re bond is required to
achieve an 18-electron configuration (Figure 14). In contrast,
[(CO)4Re]2(μ-H)2 possesses an 18-electron configuration for
each rhenium atom, and so no additional Re−Re bond is
required (Figure 14).
A good illustration of the use of the half-arrow to represent
the structure of bridging hydride compounds is provided by
Os3(CO)10(μ-H)2, which has been depicted in the literature by
a variety of different structures that have ambiguous bonding
descriptions, with Os−Os bond orders that vary from zero to
two (Figure 15). The ambiguity is also highlighted by the fact
that some representations simply utilize ill-defined dotted lines
to represent the bonding. The half-arrow representation,
however, clearly indicates that each osmium center associated
with the bridging hydrogens adopts an 18-electron config-
uration without requiring an additional Os−Os bond.
Significantly, this description is in accord with theoretical
calculations, which indicate that there is no OsOs double
bond and that the interaction is best represented as two 3c−2e
interactions.27
With respect to the origin of the different bonding
representations of Os3(CO)10(μ-H)2 in Figure 15, it is
pertinent to note that the derivation of an OsOs double
bond is a consequence of essentially apportioning only “half of
an electron” of a bridging hydrogen atom to each metal
center.28 The problem with this approach can be readily
ascertained by recognizing that this method would also predict
a BB double bond in B2H6 in order for boron to achieve an
octet configuration, which is clearly not feasible. More
generally, this approach predicts the number of metal−metal
bonds (m#) for transition metals by using the formula m# =
(18n − N)/2, where n is the number of M atoms, and N is the
total valence electron count for the molecule. However, this
simple formula does not take into account the 3c−2e nature of
bridging hydride interactions, and so the predicted M−M bond
order is greater than that determined by computational
analysis; in fact, the bond order predicted by the formula
does not actually depend on whether the hydrogen atoms are
terminal or bridged, which is clearly not a satisfactory situation.
In contrast, the half-arrow representation provides a bonding
description that is much more in accord with that predicted
theoretically.
In addition to the above compounds that feature hydrogen
bridging between two boron centers and between two
transition metals centers, hybrid compounds are known in
which hydrogen bridges boron and a metal. For example, there
is a large variety of transition metal borohydride compounds in
which the metal and boron are bridged by one (κ1), two (κ2),
or three (κ3) hydrogen atoms (Figure 16).29 The bonding in
these compounds may likewise be described in terms of the
half-arrow representations, which provides a convenient means
to evaluate the electron count of a metal center, as illustrated
for Cp2Nb(κ2-H2BH2) in Figure 17.
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 15. Structural representations for Os3(CO)10(μ2-H)2. Only E and F provide good representations of the bonding.
Figure 16. Coordination modes of metal borohydride compounds.
Figure 17. Example of a κ2-borohydride complex.
Agostic Compounds
An interesting class of compounds is one in which there is an
interaction between a C−H bond and a metal center. Such
compounds, and their interactions, are referred to by the term
“agostic”.13,30 Agostic interactions are recognized to be an
important feature of organometallic chemistry, and they can
play important roles in modifying the structure and stability of
a molecule in which the metal center would otherwise be
electron deficient. Classic examples of compounds that possess
agostic interactions are provided by the bis-
(dimethylphosphino)ethane titanium alkyl compounds
(dmpe)TiMeCl3 and (dmpe)TiEtCl3, in which the titanium
centers interact, respectively, with α- and β-C−H bonds
(Figure 18).30 The significance of the agostic C−H−M
Figure 18. Examples of compounds that feature α- and β-agostic
interactions. Coordination of the C−H bond to Ti increases the
electron count by two.
2472
Article
interaction is provided by the observation that the titanium
ethyl compound, (dmpe)TiEtCl3, exhibits a very unusual acute
Ti−C−C bond angle of 85.9° because of the interaction of the
titanium with the β-C−H bond. This interaction is a
consequence of the titanium center of (dmpe)TiEtCl3 being
highly electron deficient (with a formal electron count of 12 in
the absence of an interaction), and the electronic unsaturation
is partially relieved by interaction of the titanium center with
the electron density associated with a C−H bond.
Dihydrogen Compounds and Other σ-Complexes
Dihydrogen compounds compose an extremely interesting
class of molecules, first discovered by Kubas,31 in which the
H−H bond of H2 remains intact upon coordination to a metal
center, [M](η2-H2), as illustrated in Figure 19. Prior to Kubas’s
discovery, the only known interaction between H2 and a metal
center was oxidative addition resulting in a dihydride
compound in which the H−H bond has been cleaved,
[M]H2. The 3c−2e interaction can be viewed as donation of
the pair of electrons associated with the H−H bond into an
empty orbital on the metal center (cf. [H3]+); for this reason,
dihydrogen compounds are typically represented by drawing
an arrow from the center of the H−H bond to the metal center
(Figure 20).
The ability to isolate a dihydrogen compound, rather than a
dihydride derivative, requires that there be little backbonding
from the metal center into the H−H σ* orbital, because such
transfer of electron density promotes cleavage of the H−H
bond. For this reason, dihydrogen compounds are typically
associated with metal centers that have ligands that remove
electron density via backbonding (e.g., CO).32
Figure 19. Dihydrogen and dihydride tautomers. The former
possesses one 3c−2e interaction, whereas the latter possesses two
2c−2e interactions.
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
Figure 20. Examples of transition metal dihydrogen complexes. The
dihydrogen molecule donates its bonding electrons to the metal to
result in 18-electron configurations.
Dihydrogen complexes are a subset of a class of molecules
referred to as σ-complexes, so named because the compounds
are considered to be derived by donation of the electron
density of an X−H σ-bond to a metal. Specific examples are
provided by hydrocarbon33 and silane34 adducts, in which
C−H and Si−H bonds interact with a metal center. However,
such compounds are much less commonly encountered than
dihydrogen complexes. Alkane σ-complexes are particularly
unstable but are often invoked as intermediates in reactions
involving the oxidative-addition and reductive-elimination of
C−H bonds. By comparison, silane σ-complexes are more
stable than alkane σ-complexes, as illustrated by
Mo(dppe)2(η2-HSiH2Ph), which has been isolated in the
solid state. Since the 3c−2e interactions are asymmetric, with
longer M−C and M−Si bonds than M−H bonds, these
compounds are frequently represented by using either the half-
arrow or full-arrow representations, although both are intended
to convey the same information with respect to electron
counting purposes (Figure 21).
Figure 21. Two alternative representations of silane σ-complexes,
both of which convey the same information with respect to the
purpose of electron counting.
Other Examples of 3c−2e Interactions
Many examples of compounds with 3c−2e interactions feature
a bridging hydrogen because the nondirectional spherical
nature of the 1s orbital facilitates an interaction with more than
one atom. Despite this observation, a bridging hydrogen atom
is not a requirement for the existence of a 3c−2e interaction, as
illustrated by Al2Me6, which possesses bridging methyl groups
(Figure 22). In addition, although less common, a variety of
other ligands, such as PR3 and MeCN, are also known to
participate in 3c−2e interactions.8
■
SUMMARY
In summary, the half-arrow notation provides a means to
represent the structures of molecules with 3c−2e interactions
in such a manner that the electron count of each atom can be
conveniently evaluated. Therefore, in conjunction with the
octet and 18-electron rules, the approach affords a simple
process to evaluate the chemical reasonableness of a molecule.
This notation has broad applicability to a variety of classes of
2473
Article
Figure 22. Example of a bridging methyl compound with 3c−2e
interactions.
compounds, which include bridging hydride and agostic
complexes, and furthermore provides a means to evaluate
metal−metal bond orders in bridging hydride compounds.
■
ASSOCIATED CONTENT
*
S Supporting Information
The Supporting Information is available on the ACS
Publications website at DOI: 10.1021/acs.jchemed.9b00750.
Worksheets to help students practice the concepts
(PDF)
Worksheet answers and comments for instructors (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: parkin@columbia.edu.
ORCID
Gerard Parkin: 0000-0003-1925-0547
Notes
The author declares no competing financial interest.
■ ACKNOWLEDGMENTS
The National Science Foundation (CHE-1465095) is thanked
for support.
■ REFERENCES
(1) (a) Lewis, G. N. Valence and The Structure of Atoms and
Molecules; The Chemical Catalog Company, 1923. (b) Lewis, G. N.
The atom and the molecule. J. Chem. Phys. 1933, 1, 17−28.
(2) (a) Langmuir, I. Types of valence. Science 1921, 54, 59−67.
(b) Langmuir, I. The structure of atoms and the octet theory of
valence. Proc. Natl. Acad. Sci. U. S. A. 1919, 5, 252−259.
(3) (a) Jensen, W. B. The origin of the 18-electron rule. J. Chem.
Educ. 2005, 82, 28−28. (b) Pyykkö, P. Understanding the eighteen-
electron rule. J. Organomet. Chem. 2006, 691, 4336−4340.
(c) Rasmussen, S. C. The 18-electron rule and electron counting in
transition metal compounds: Theory and application. ChemTexts
2015, 1, 10. (d) Mingos, D. M. P. Complementary spherical electron
density model and its implications for the 18 electron rule. J.
Organomet. Chem. 2004, 689, 4420−4436. (e) Mingos, D. M. P.
Development of bonding models based on the Periodic Table. Chimia
2019, 73, 152−164. (f) Landis, C. R.; Weinhold, F. 18-Electron rule
and the 3c/4e hyperbonding saturation limit. J. Comput. Chem. 2016,
37, 237−241.
(4) Gillespie, R. J.; Robinson, E. A. Gilbert N. Lewis and the
chemical bond: The electron pair and the octet rule from 1916 to the
present day. J. Comput. Chem. 2007, 28, 87−97.
(5) (a) Wade, K. Electron Deficient Compounds; Springer, 1971.
(b) Wade, K. Boranes: Rule-breakers become pattern-makers. New
Scientist, June 6, 1974, pp 615−617 (c) Stone, F. G. Electron-deficient
compounds. Endeavour 1961, 20, 61−67.
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education
(6) Laszlo, P. A diborane story. Angew. Chem., Int. Ed. 2000, 39,
2071−2072.
(7) Rundle, R. E. Electron deficient compounds. J. Phys. Chem. 1957,
61, 45−50.
(8) (a) Green, J. C.; Green, M. L. H.; Parkin, G. The occurrence and
representation of three-centre two-electron bonds in covalent
inorganic compounds. Chem. Commun. 2012, 48, 11481−11503.
(b) Green, M. L. H.; Parkin, G. The Covalent Bond Classification
method and its application to compounds that feature three-center
two-electron bonds. Struct. Bonding 2016, 171, 79−140.
(9) (a) Bridgeman, A. J.; Empson, C. J. Detecting delocalization.
New J. Chem. 2008, 32, 1359−1367. (b) De Kock, R. L.; Bosma, W. B.
The three-center, two-electron chemical bond. J. Chem. Educ. 1988,
65, 194−197.
(10) Three-center−four-electron interactions are also known. See,
for example, Green, M. L. H; Parkin, G. The classification and
representation of main group element compounds that feature three-
center four-electron interactions. Dalton Trans. 2016, 45, 18784−
18795.
(11) The source of this quote attributed to Rundle is the opening
comment in Wade’s 1971 book on Electron Def icient Compounds (ref
5a). Interestingly, a very similar comment was attributed to Mulliken
by Burdett in his 1997 book (Burdett, J. K. Chemical Bonds: A Dialog;
Wiley, 1997, p 103): “Robert Mulliken said at one time, referring to
the term ‘electron deficient’ molecules, that there was nothing
deficient about the molecules, just the theory.” Neither Wade nor
Burdett give the sources for their quotes, and so it is difficult to
ascertain the originator. However, in private correspondence with
G.P. (January 2011), Wade commented that he believes he first heard
the comment from Rundle at a lecture in England during the period
1954−1959.
(12) Another problem with the term “electron deficient” is that it is
also used to refer to molecules in which one of the atoms is
hypovalent (i.e., it has less than an octet configuration for a main
group element or less than an 18-electron configuration for a
transition metal). For example, monomeric BH3 is considered to be
electron deficient (hypovalent) because the boron has only a sextet
configuration.
(13) (a) Miessler, G. L.; Fischer, P. J.; Tarr, D. A. Inorganic
Chemistry, 5th ed.; Pearson: Boston, MA, 2014. (b) Housecroft, C. E.;
Sharpe, A. G. Inorganic Chemistry, 5th ed.; Pearson: Boston, MA,
2018.
(14) This statement assumes that the molecule is neutral. Obviously,
if the molecule were to be charged, then one of the partners that
provides one of the electrons must necessarily also be charged.
(15) Haaland, A. Covalent versus dative bonds to main group
metals, a useful distinction. Angew. Chem., Int. Ed. Engl. 1989, 28,
992−1007.
(16) Parkin, G. Valence, oxidation number and formal charge: Three
related but fundamentally different concepts. J. Chem. Educ. 2006, 83,
791−799.
(17) It should be noted that dative bonds are often drawn without
using an arrow or including the formal charges. Although such
representations indicate connectivity and minimize clutter in complex
structures, they do not provide a good indication of the electronic
nature of the molecule.
(18) (a) Green, M. L. H.; Parkin, G. Application of the Covalent
Bond Classification method for the teaching of inorganic chemistry. J.
Chem. Educ. 2014, 91, 807−816. (b) Parkin, G. Classification of
organotransition metal compounds. In Comprehensive Organometallic
Chemistry III; Crabtree, R. H., Mingos, D. M. P., Eds.; Elsevier:
Oxford, 2007; Vol. 1, pp 1−57. (c) Green, M. L. H. A new approach
to the formal classification of covalent compounds. J. Organomet.
Chem. 1995, 500, 127−148. (d) Covalent Bond Classification. http://
www.covalentbondclass.org/ (accessed Sept 19, 2019).
(19) (a) Jensen, W. B. The place of zinc, cadmium, and mercury in
the periodic table. J. Chem. Educ. 2003, 80, 952−961. (b) Constable,
E. C. Evolution and understanding of the d-block elements in the
periodic table. Dalton Trans. 2019, 48, 9408−9421.
2474
Article
(20) Albright, T. A.; Burdett, J. K.; Whangbo, M.-H. Orbital
Interactions in Chemistry, 2nd ed.; John Wiley & Sons, Inc.: New York,
NY, 2013.
(21) Berry, M.; Cooper, N. J.; Green, M. L. H.; Simpson, S. J.
Synthesis and reactions of binuclear molybdenocene and tungsteno-
cene derivatives. J. Chem. Soc., Dalton Trans. 1980, 29−41.
(22) Arrows are used in a multitude of ways in chemistry. For
example, in addition to a dative bond, a straight arrow can be used to
indicate a reaction, evolution of a gas, precipitation of a solid, or an
electron in a molecular orbital diagram. Likewise, a straight, half-
barbed arrow may be used as a component of an equilibrium symbol
or an electron in a molecular orbital diagram. In addition to straight
arrows, curved arrows are also used. For example, a curved arrow is
used to indicate the motion of a pair of electrons in a reaction
mechanism, whereas a curved, half-barbed (fishhook) arrow is used to
indicate the motion of a single electron in radical reactions. In each
case, the meaning of an arrow must be interpreted in the context of its
use, but it must be emphasized that the straight half-arrow used to
represent a 3c−2e interaction has a distinctly different meaning from
that of a curved half-arrow. See (a) Alvarez, S. Chemistry: A panoply
of arrows. Angew. Chem., Int. Ed. 2012, 51, 590−600. (b) Lakshminar-
ayanan, A. Arrows in chemistry. Resonance 2010, 15, 51−62.
(23) Bau, R.; Teller, R. G.; Kirtley, S. W.; Koetzle, T. F. Structures of
transition-metal hydride complexes. Acc. Chem. Res. 1979, 12, 176−
183.
(24) In addition, even if the interaction with the outer atom is
significant, the half-arrow representation is sometimes preferable
because it minimizes clutter in the drawing.
(25) Wulfsberg, G. Inorganic Chemistry; University Science Books:
Sausalito, CA, 2000.
(26) (a) Parkin, G. Metal−metal bonding in bridging hydride and
alkyl Compounds. Struct. Bonding 2010, 136, 113−145. (b) Baik, M.-
H.; Friesner, R. A.; Parkin, G. Theoretical investigation of the metal-
metal interaction in dimolybdenum complexes with bridging hydride
and methyl ligands. Polyhedron 2004, 23, 2879−2900.
(27) Sherwood, D. E.; Hall, M. B. Electronic structure of metal
clusters. 2. Photoelectron spectrum and molecular orbital calculations
of decacarbonyldihydridotriosmium. Inorg. Chem. 1982, 21, 3458−
3464.
(28) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987. (b) Owen, S. M.
Electron counting in clusters: A view of the concepts. Polyhedron
1988, 7, 253−283. (c) Housecroft, C. E.; Sharpe, A. G. Inorganic
Chemistry; Prentice Hall: New York, NY, 2001. (d) Bennett, M. J.;
Graham, W. A. G.; Hoyano, J. K.; Hutcheon, W. L. Synthesis and X-
ray structure of di-μ-hydrido-octacarbonyldirhenium. J. Am. Chem.
Soc. 1972, 94, 6232−6233.
(29) Besora, M.; Lledós, A. Coordination modes and hydride
exchange dynamics in transition metal tetrahydroborate complexes.
Struct. Bonding 2008, 130, 149−202.
(30) (a) Brookhart, M.; Green, M. L. H.; Parkin, G. Agostic
interactions in transition metal compounds. Proc. Natl. Acad. Sci. U. S.
A. 2007, 104, 6908−6914. (b) Brookhart, M.; Green, M. L. H.; Wong,
L. L. Carbon hydrogen transition-metal bonds. Prog. Inorg. Chem.
2007, 36, 1−124. (c) Brookhart, M.; Green, M. L. H. Carbon-
hydrogen-transition metal bonds. J. Organomet. Chem. 1983, 250,
395−408.
(31) (a) Kubas, G. J. Metal Dihydrogen and σ-Bond Complexes:
Structure, Theory, and Reactivity; Kluwer Academic/Plenum Publish-
ers: New York, NY, 2001. (b) Kubas, G. Chem. Rev. 2007, 107, 4152−
4205.
(32) Upmacis, R. K.; Gadd, G. E.; Poliakoff, M.; Simpson, M. B.;
Turner, J. J.; Whyman, R.; Simpson, A. F. The photochemical
synthesis of [Cr(CO)5(H2)] in solution: I.r. evidence for coordinated
molecular dihydrogen. J. Chem. Soc., Chem. Commun. 1985, 27−30.
(33) (a) Weller, A. S.; Chadwick, F. M.; McKay, A. I. Transition
metal alkane-sigma complexes: Synthesis, characterization, and
reactivity. Adv. Organomet. Chem. 2016, 66, 223−276. (b) Hall, C.;
DOI: 10.1021/acs.jchemed.9b00750
J. Chem. Educ. 2019, 96, 2467−2475
Journal of Chemical Education Article

Perutz, R. N. Transition metal alkane complexes. Chem. Rev. 1996, 96,

3125−3146. (c) Crabtree, R. H. Aspects of methane chemistry. Chem.
Rev. 1995, 95, 987−1007.
(34) Corey, J. Y. Reactions of hydrosilanes with transition metal
complexes and characterization of the products. Chem. Rev. 2011, 111,
863−1071.

2475 DOI: 10.1021/acs.jchemed.9b00750

J. Chem. Educ. 2019, 96, 2467−2475