1. Introduction and Preliminaries
The binomial coefficient
is defined for
and
by
where
denotes the classical gamma function. In particular, when
, the binomial coefficient reduces to the familiar form
with the convention
. For a historical overview of the binomial coefficients and a detailed treatment of related fundamental identities, such as
see Jordan [
1] (§§ 22–23).
Using a known identity along with mathematical induction, Sun [
2] established the following formula:
Throughout this work, we use
,
, and
to denote the sets of complex numbers, integers, and positive integers, respectively. Also, for a given
, we use
(or
and
(or
)) to denote the sets of integers greater than or equal to (or strictly greater than)
ℓ, and less than or equal to (or strictly less than)
ℓ, respectively. As illustrated by (
3), finite or infinite sums involving reciprocals of binomial coefficients have attracted considerable interest due to their rich structure and surprising properties (see, for example, [
3] (Equations (2.1)–(2.26), (4.1)–(4.30) and (5.1) and (5.2))).
Gould [
3] (Equation (2.12)) presented an infinite analogue of identity (
3):
where
denotes the classical Riemann zeta function, defined by
and
is the second-order harmonic number. The generalized harmonic number of order
is defined as
By convention,
, and when
,
denotes the
nth harmonic number.
Combining (
3) with the identity given in [
3] (Equation (5.1)) yields the following result for
:
where identities (
4) and (
6) are applied in the second and third equalities, respectively.
Batır and Sofo [
4] presented several finite sum identities involving the reciprocals of binomial and central binomial coefficients, along with harmonic numbers, Fibonacci numbers, and Lucas numbers.
Sprugnoli [
5] investigated combinatorial sums involving reciprocals of central binomial coefficients, deriving several infinite and finite sum results—some involving the golden ratio—using generating functions and the method of coefficients.
Batır and Chen [
6] established a general combinatorial formula involving reciprocals of binomial coefficients and partial sums of arbitrary sequences, from which they derived numerous identities—some previously known and others new—featuring reciprocals of binomial and central binomial coefficients and harmonic numbers.
Apéry [
7] used the rapidly converging alternating series
involving reciprocals of central binomial coefficients to prove the irrationality of
.
Guillera [
8] derived a remarkably fast-converging series for the Catalan constant
G, involving reciprocals of central binomial coefficients:
a value classically defined by the alternating series
For other similar fast-converging series, see [
9].
Sofo and Batır [
10] presented an alternative expression for the left-hand side of identity (
3). For
,
It follows from (
3) and (
6) that
Batır and Choi [
11] investigated the following sums:
where
,
, and
.
In this paper, we derive an explicit formula for a broad class of generalized sums of the form
where
,
, and
. Here,
, with
denoting an arbitrary sequence of complex numbers. Notably, this general expression reduces to known identities under specific parameter choices; for instance, the first sum in (
7) becomes (
3) when
,
,
, and
for all
.
In
Section 2, we present the main identity (Theorem 1) and its derivation.
Section 3 is devoted to examining the three particular cases
, as detailed in Corollaries 1–3. In particular, Corollary 1 offers an alternative explicit formula for the case
, accompanied by a direct proof.
Section 4 investigates further concrete examples of the identity by selecting specific sequences for
, thereby illustrating the flexibility of our general framework.
In
Section 5, we differentiate the results from Corollaries 1–3 with respect to the parameter
r, leading to new identities involving sums of reciprocals of binomial coefficients. These are formalized in Propositions 1–3, and are complemented by several interesting special cases.
Finally,
Section 6 concludes the paper by highlighting the depth and adaptability of the central identity presented in Theorem 1.
For completeness, we also introduce several auxiliary functions, including extensions of the classical harmonic numbers
and generalized harmonic numbers
, as discussed in [
12,
13] (Section 1.3). The psi (or digamma) function, denoted by
, is defined as the logarithmic derivative of the gamma function
:
The harmonic number
is expressed in term of the digamma function as follows:
where
is the Euler–Mascheroni constant (see, for example, [
13] (Section 1.2) and [
14], and the references therein). Since the psi function
is defined on
, the harmonic number
admits the following extension to the complex domain:
For non-integer values of the order
n, such as a complex number
, the generalized harmonic number
can be defined in terms of the polygamma function
where
is the gamma function and
is the digamma function. This definition gives
for
and
, where
denotes the Riemann zeta function. In the special case when
, (
10) reduces to
, where
is the extended harmonic number of order
in (
9).
This method of extending harmonic and generalized harmonic numbers, pioneered by Sofo and Srivastava [
12], has proven to be remarkably fruitful. It not only deepens our understanding of these classical objects but also opens the door to a wealth of elegant identities, particularly those related to classical Euler sums.
By applying the logarithmic derivative and invoking identity (
8), we obtain the following formula:
where (
9) is employed for the second equality.
We remark that Pascal’s Triangle exhibits symmetry about its central vertical axis. This property arises from the fundamental combinatorial identity
which asserts that, in the
nth row, the
kth entry from the left equals the
th entry from the right. As a result, each row forms a palindromic sequence, reading identically in both directions.
That is, in row
n, the
kth entry from the left is equal to the
th entry from the right. Consequently, each row of Pascal’s Triangle is palindromic, reading identically from left to right and right to left. The identities and results developed in this paper are deeply connected to this underlying combinatorial symmetry. Many of the recurrence relations and binomial identities we derive inherently reflect this structural balance. Notably, Equation (
11) itself exhibits a form of symmetry: differentiating its left-hand side with respect to the parameter
a yields a result that mirrors the differentiation on the right-hand side, reinforcing the symmetrical behavior of the underlying functions.
This alignment between structural symmetry in binomial coefficients and the analytic properties of the identities we explore highlights this paper’s thematic relevance to the broader concept of symmetry in mathematical analysis and combinatorics.
6. Concluding Remarks
In this work, we have explored a variety of special cases stemming from our main result, illustrating the richness and flexibility of the underlying identities. Thanks to the generality of Theorem 1, many further instances can be generated. In particular, whenever a closed-form expression for the partial sums of a sequence is available, our framework enables the derivation of explicit identities, as evidenced by the corollaries presented.
Moreover, the identities in these corollaries admit further refinement, leading to even more concrete formulas. As an example, Formula (
40) reproduces the classical identity
Similarly, by applying Equation (
52), one can derive additional identities that extend the special cases discussed in Corollaries 10–12.
We hope that this line of investigation not only contributes to a deeper understanding of summation identities involving reciprocals of binomial coefficients but also lays a foundation for discovering new analytic results with potential applications in number theory, combinatorics, and mathematical analysis.