1. Introduction
All graphs in the paper are finite and have no loops or multiple edges. Let
be a graph with
vertices. An r-matching in a graph
is a set of
edges, no two of which have a vertex in common. The number of r-matching in
will be denoted by
. We set
and define the matching polynomial of
by
(1)
The matching polynomial has very important applications in mathematics, statistical physics, and chemistry. In statistical physics, it serves as a mathematical model for describing a certain physical system. Physicists Heilmann and Lieb first introduced the matching polynomial of a graph in reference [1] to study this physical system. In theoretical chemistry, the sum of the absolute values of the roots of the matching polynomial is called the matching energy level of the graph, which is related to the activity of the aromatic hydrocarbon represented by this graph (see [2]). The sum of the absolute values of all its coefficients (that is, the total number of all matchings) is the Hosoya index of the hydrocarbon represented by this graph (see [3]).
For any graph
, the roots of
are all real numbers. Assume that
, the largest root
is referred to as the largest matching root of
.
Throughout the paper, we denote by
and
the path and the cycle on
vertices, respectively.
denotes the tree with a vertex
of degree 3 such that
, and
denotes the tree obtained from the path with versices
(in order) by attching a pendant edge at each of the verices
and
.
is obtained by appending a cycle
to a pendant vertex of a path
. Two graphs are matching equivalency if they share the same matching polynomial. A graph
is said to be determined by its matching polynomial if for any graph
,
implies that
is isomorphic to
. For the notations and terms not explained in this article, please refer to [4]. Since Farrell E. J. studied the problem of graphs characterization by its matching polynomial, many matching characterization graphs have been found (see [5]-[8]). However, as most of the methods used are to compare the coefficients of the matching polynomials of two graphs, this work is quite difficult and progresses slowly. There are still a large number of problems that have not been solved. the paper makes full use of the information about the roots of the matching polynomial
, we prove
are determined by its matching polynomial iff
is even or 3 and
.
2. Basic Results
Lemma 2.1 [4]. The matching polynomial
satisfies the following identities:
a)
.
b)
if
is an edge of
.
Lemma 2.2 [4]. Let
be a connected graph, and let
be a proper subgraph
. Then
.
Lemma 2.3 [4]. Let
be a vertex in the graph
. Then the roots of
interlace those of
. If
is connected then the largest root of
is simple ,and is strictly greater than the largest root of
.
Lemma 2.4 [5] [6]. Let
be a connected graph, then
a)
iff
.
b)
iff
.
Lemma 2.5 [5] [6]. The connected graphs
with the largest matching root
in the interval
iff it is precisely the graphs of the following types:
a)
,
,
,
.
b)
,
,
.
c)
, for
or
where
and
(2)
Lemma 2.6 [5]. Let
be a tree and let
be obtained from
by subdividing the edge
of
, then
a)
if
not lies on an internal path of
.
b)
if
lies on an internal path of
, and if
is not isomorphic to
.
Lemma 2.7 [7] [8].
,
.
Lemma 2.8.
.
Proof. If
, clearly
is a proper subgraph of
, By Lemma 2.2,
. If
,
is a proper subgraph of
, then
.
By Lemma 2.6,
. So
and
is a proper subgraph of
, by Lemma 2.2, thus the lemma holds.
Lemma 2.9. If
, then
and
may only be 5, 6, 7, 11.
Proof. By Lemma 2.1,
.
So, we have
. Direct calculate the largest matching root of
and
(using Matlab 8.0), we immediately have the following:
(3)
(4)
(5)
(6)
By Lemma 2.8, If
,
. If
,
This completes the proof.
3. Main Results
Theorem 3.1. Let
. Then
is uniquely determined by its matching polynomial iff
is even or 3 and
.
Proof. The necessary condition follows immediately from Lemma 2.1.
We have
(7)
(8)
(9)
(10)
Now suppose that
is even or 3 and
,
is a graph being matching equivalency with
, hance
. We proceed to prove that
must be isomorphic to
. By Lemma 2.5, 2.7,
. Set
,
,
, then by Lemma 2.8, 2.9, we get
(where
denotes
re-elements form
,
). By Lemma 2.3,
. We distinguish the following cases:
Case 1. If
. We further consider several cases:
Subcase 1.
Then
. By Lemma 2.7,
.
Direct computation shows
, hence we have
matching equivalency with
.
Which is a contradiction.
Subcase 2.
Then
. By Lemma 2.1, we get
, which is a contradiction.
Subcase 3.
. The same argument as subcase 2 can be used to get a contradiction.
Subcase 4.
.
Then
. By Lemma
2.7, and
, we have
, thus
be isomorphic to
.
Case 2. If
. By lemma 2.9, we get
or 7. First suppose that
. Note that
and
. Thus we have
or
.
Which contradicts to Lemma 2.9. Secondly assume
. By Lemma 2.1, direct Direct computation shows
and
, hence we get
or
.
which is a contradiction.
Combing cases1, 2,
is isomorphic to
. The proof is complete.
For a graph, its matching polynomial determine the matching polynomial of its complement (see [9]), so the complement of
is determined by its matching polynomial iff
is even or 3 and
.