Next Article in Journal
Advanced Structural Analysis of n-Derivations and n-Automorphisms in Nest Algebras via Exponential Mappings
Previous Article in Journal
Symmetry-Driven Gaussian Representation and Adaptive Assignment for Oriented Object Detection
Previous Article in Special Issue
Some New Bounds for Bateman’s G-Function in Terms of the Digamma Function
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Some New Inequalities for the Gamma and Polygamma Functions

Department of Mathematics and Statistics, College of Sciences and Humanities, Fahad Bin Sultan University, P.O. Box 15700, Tabuk 71454, Saudi Arabia
*
Authors to whom correspondence should be addressed.
Symmetry 2025, 17(4), 595; https://doi.org/10.3390/sym17040595
Submission received: 11 March 2025 / Revised: 9 April 2025 / Accepted: 12 April 2025 / Published: 14 April 2025

Abstract

:
In this paper, we present some new symmetric bounds for the gamma and polygamma functions. For this goal, we present two functions involving gamma and polygamma functions and we investigate their complete monotonicity. Also, we investigate their completely monotonic degrees. This concept gives more accuracy in measuring the complete monotonicity property. These new bounds are better than some of the recently published results.

1. Introduction

The gamma function was presented by Euler [1] as:
Γ ( l ) = lim n n l l r = 1 n r l + r , l R +
and this leads to
Γ ( n + l ) Γ ( l ) = r = 0 n 1 ( r + l ) , l R + .
ln Γ ( l ) has the asymptotic expansion [2]:
ln Γ ( l ) ( l 1 / 2 ) ln l + 1 2 ln ( 2 π ) l + r = 1 B 2 r ( 2 r ) ( 2 r 1 ) l 2 r 1 , l
where B 2 r denotes the Bernoulli numbers. The Psi and polygamma functions are presented by [2]:
ψ ( l ) = γ l 1 l + r = 1 l r ( r + l ) = 0 ( t e t ) 1 e ( l + 1 ) t 1 + e t d t , l R +
where the Euler–Mascheroni’s constant γ 0.5772156649 , and
ψ ( s ) ( l ) = r = 0 ( 1 ) s + 1 s ! ( r + l ) s + 1 = 0 ( 1 ) s + 1 t s e ( 1 l ) t 1 + e t d t , l R + , s N
and they have the relation:
ψ ( s ) ( 1 + l ) = ( 1 ) s s ! l 1 + s + ψ ( s ) ( l ) , s = 0 , 1 , 2 , .
Also, they have the asymptotic expansions [2]:
ψ ( l ) 1 2 l + ln l r = 1 B 2 r ( 2 r ) l 2 r , l
and
ψ ( s ) ( l ) ( 1 ) s 1 s ! 2 l s + 1 + ( 1 ) s 1 ( s 1 ) ! l s + ( 1 ) s 1 r = 1 ( s + 2 r 1 ) ! B 2 r ( 2 r ) ! l 2 r + s , s N , l .
For extra information about the gamma and polygamma functions, see [3,4,5,6]. A function S defined on an interval J is completely monotonic (CM) if it satisfies that
( 1 ) s S ( s ) ( l ) 0 l J ; s = 0 , 1 , 2 , . . . .
The necessary and sufficient condition for the function S ( l ) to be CM for l R + is that [7]:
S ( l ) = 0 e l t d ϑ ( t ) ,
where ϑ ( t ) is a non-negative measure such that the integral converges for l R + . Let S ( l ) be a CM function for l R + and suppose the notation S ( ) = lim l S ( l ) . If l ρ [ S ( l ) S ( ) ] is a CM function for l R + if and only if ρ 0 , σ , then the number σ R + is called the CM degree of S ( l ) for l R + and denoted by deg C M l [ S ( l ) ] = σ . For extra information about this topic, see [8,9,10].
In 2007, Alzer and Batir [11] introduced some CM functions involving the gamma the digamma functions and the following symmetric inequalities are deduced:
exp l 1 2 ψ ( l + 1 / 3 ) < Γ ( l ) 2 π l l < exp l 1 2 ψ ( l ) , l R +
1 2 ψ ( l + 1 / 3 ) < ln l ψ ( l ) < 1 2 ψ ( l ) , l R +
and for s = 2 , 3 , ,
( 1 ) s + 1 2 ψ ( s ) ( l + 1 / 3 ) < ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1 < ( 1 ) s + 1 2 ψ ( s ) ( l ) , l R + .
Batir [12] improved (8) by:
exp l 1 2 ψ ( l ) 1 6 ( l 1 4 ) < Γ ( l ) 2 π l l < exp l 1 2 ψ ( l ) 1 6 l , l R + .
Next, Şevli and Batir [13] investigated some CM functions involving the Γ and ψ functions and the following symmetric inequalities are deduced:
1 l exp l + 1 12 ψ ( l + 1 / 2 ) < Γ ( l ) 2 π l l < 1 l exp l + 1 12 ψ ( l ) , l R +
1 2 l 1 12 ψ ( l + 1 / 2 ) < ln l ψ ( l ) < 1 2 l 1 12 ψ ( l ) , l R +
and for s = 2 , 3 , and l R + ,
( s 1 ) ! 2 l s + ( 1 ) s 12 ψ ( s + 1 ) ( l + 1 / 2 ) < ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1 < ( 1 ) s 12 ψ ( s + 1 ) ( l ) + ( s 1 ) ! 2 l s .
In 2024, Moustafa [14] refined the upper bounds of (8) and (11) for all l R + and l > 1 8 consecutively by:
exp l 1 2 ψ ( l ) 1 6 ψ ( l ) < Γ ( l ) 2 π l l < exp l 1 2 ψ ( l ) 1 6 ψ ( l + 1 / 4 ) , l R +
and he improved the upper bounds of (9) and (13) for all l R + by:
1 2 ψ ( l ) + 1 6 ψ ( l ) < ln l ψ ( l ) < 1 2 ψ ( l ) + 1 6 ψ ( l + 1 / 4 ) , l R +
and for l R + and s = 2 , 3 , , he improved the upper bounds of (10) and (14) by:
( 1 ) s + 1 2 ψ ( s ) ( l ) + ( 1 ) s + 1 6 ψ ( s + 1 ) ( l ) < ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1
< ( 1 ) s + 1 2 ψ ( s ) ( l ) + ( 1 ) s + 1 6 ψ ( s + 1 ) ( l + 1 / 4 )
And finally, he improved the lower bound of (8) for all l 2.2 by:
exp l 1 2 ψ ( l ) 1 6 ψ ( l ) + 1 24 l 2 < Γ ( l ) 2 π l l < exp l 1 2 ψ ( l ) 1 6 ψ ( l ) + 1 24 l 3 10 2 .
Most of the last results were generalized to the generalized gamma and polygamma functions in [15,16,17].
The researchers studying mathematics and applied sciences benefit from the new inequalities involving the gamma and polygamma functions, which gives better approximations and a better understanding of the behavior of special functions. Our paper aims to present new inequalities involving Γ and ψ ( s ) ( s N { 0 } ). These new results are better than the previous inequalities presented by Alzer and Batir, Batir, Şevli and Batir and Moustafa. For this goal, we will study some CM functions containing Γ , ψ and ψ and we will present some new symmetric bounds for Γ and ψ ( s ) ( s N { 0 } ), which improve the results (8)–(18).

2. Auxiliary Results

In [10], (Lemma 1), the next corollary was introduced:
Corollary 1.
Suppose that the real-valued function T ( l ) is defined on ( l 0 , ) and lim l T ( l ) = 0 . Then, for υ R + , T ( l ) > 0 , if  T ( l + υ ) < T ( l ) for every l ( l 0 , ) and T ( l ) < 0 , if T ( l + υ ) > T ( l ) for every l ( l 0 , ) .
In the following Lemma, we present some inequalities involving the digamma and polygamma functions, which will be useful in Section 4, in proving that the new bounds for Γ and ψ s are better than the old ones.
Lemma 1. 
1. 
For l [ 0.5 , ) , we have
ψ ( l ) > 1 l 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) ,
ψ ( l ) < 1 l 2 1 6 ψ ( l ) 1 12 ψ ( 4 ) ( l + 3 / 10 ) .
ln l ψ ( l ) > 1 6 ψ ( l ) + 1 12 ψ l + 3 / 10 + 1 3 l ,
and
ln l ψ ( l ) > 1 6 ψ ( l ) + 1 12 ψ ( l + 3 / 10 ) .
2. 
For all l 5.1 , we have
ln l 1 6 ψ ( l ) 1 12 ψ ( l ) < ψ ( l + 1 / 3 ) ,
1 l 1 6 ψ ( l ) 1 12 ψ ( l ) > ψ ( l + 1 / 3 )
and
1 l 2 1 6 ψ ( l ) 1 12 ψ ( 4 ) ( l ) < ψ ( l + 1 / 3 ) .
3. 
For all l 13 10 , we have
ln l ψ ( l ) > 1 6 ψ ( l ) + 1 12 ψ ( l + 3 / 10 ) + 1 3 ψ ( l + 1 / 4 ) ,
1 l 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) < ψ ( l ) + 1 3 ψ ( l + 1 / 4 ) ,
ψ ( l ) < 1 l 2 1 6 ψ ( l ) 1 12 ψ ( 4 ) ( l + 3 / 10 ) 1 3 ψ ( l + 1 / 4 ) ,
and
ln l ψ ( l ) < 1 2 ψ ( l ) + 1 12 ψ ( l ) + 1 20 l 3 1 12 l 2 .
Proof. 
First, let  Ω 1 ( l ) = 1 l ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) . Then Ω 1 ( l ) = 1 l 2 ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( 4 ) ( l + 3 / 10 ) and by using (5), we obtain
Ω 1 ( l ) + Ω 1 ( 1 + l ) = h ( l ) ( 1 + l ) 2 l 4 ( 3 + 10 l ) 5 < 0 , l 1 2
where
h ( l ) = 11741 + 225636 l 1 2 + 1206636 l 1 2 2 + 3013600 l 1 2 3
+ 4054000 l 1 2 4 + 3030000 l 1 2 5 + 1200000 l 1 2 6 + 200000 l 1 2 7 .
Using the asymptotic Formula (7), we obtain lim l Ω 1 ( l ) = 0 and by Corollary 1, we obtain Ω 1 ( l ) is positive for l 0.5 and hence we obtain (20). Then, Ω 1 ( l ) is increasing on [ 0.5 , ) with lim l Ω 1 ( l ) = 0 , and consequently, Ω 1 ( l ) is negative for l 0.5 and this proves (19). Next, we assume that Ω 2 ( l ) = ln l ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) 1 3 l and then Ω 2 ( l ) = 1 l ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) + 1 3 l 2 and by using (5), we have
Ω 2 ( l ) + Ω 2 ( 1 + l ) = 1 4 3 l 3 ( 3 + 10 l ) 4 ( 1 + l ) 2 [ 491 + 41830 l 1 2 + 191400 l 1 2 2
+ 322000 l 1 2 3 + 230000 l 1 2 4 + 60000 l 1 2 5 ] > 0 , l 1 2 .
Similarly, as before, we obtain Ω 2 ( l ) < 0 for all l 1 2 and then Ω 2 ( l ) > 0 for l 0.5 and this proves (21), and consequently, we have (22). Now, we assume that Ω 3 ( l ) = ψ ( l + 1 / 3 ) ln l + 1 6 ψ ( l ) + 1 12 ψ ( l ) . Then Ω 3 ( l ) = ψ ( l + 1 / 3 ) 1 l + 1 6 ψ ( l ) + 1 12 ψ ( l ) and Ω 3 ( l ) = ψ ( l + 1 / 3 ) + 1 l 2 + 1 6 ψ ( x ) + 1 12 ψ ( 4 ) ( x ) and hence
Ω 3 ( l ) + Ω 3 ( l + 1 ) = t ( l ) ( 1 + l ) 2 l 5 ( 1 + 3 l ) 3 < 0 , l 5.1
where
25000 t ( l ) = 24458152 + 432369375 l 51 10 + 306040500 l 51 10 2 + 82610000 l 51 10 3
+ 9950000 l 51 10 4 + 450000 l 51 10 5 .
By the same way, we obtain Ω 3 ( l ) > 0 for all l 5.1 and then we have (25). Then, Ω 3 ( l ) is increasing on [ 5.1 , ) with lim l Ω 3 ( l ) = 0 and then Ω 3 ( l ) < 0 for all l 5.1 and this proves (24). Also, Ω 3 ( l ) is decreasing on [ 5.1 , ) with lim l Ω 3 ( l ) = 0 and then we obtain (23). After that, we assume that Ω 4 ( l ) = ln l ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) 1 3 ψ ( l + 1 / 4 ) . Then Ω 4 ( l ) = 1 l ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( l + 3 / 10 ) 1 3 ψ ( l + 1 / 4 ) and Ω 4 ( l ) = 1 l 2 ψ ( l ) 1 6 ψ ( l ) 1 12 ψ ( 4 ) ( l + 3 / 10 ) 1 3 ψ ( l + 1 / 4 ) and similarly, we have
Ω 4 ( l ) + Ω 4 ( l + 1 ) = η ( l ) ( 1 + 4 l ) 4 l 4 ( 1 + l ) 2 ( 3 + 10 l ) 5 < 0 , l 1.3
where
3125 η ( l ) = 128702632481 + 2752894096900 l 13 10 + 12291217933500 l 13 10 2
+ 26291877140000 l 13 10 3 + 32922564950000 l 13 10 4 + 25998075950000 l 13 10 5
+ 1318515 ( 10 7 ) l 13 10 6 + 4183225 ( 10 6 ) l 13 10 7 + 758 ( 10 9 ) l 13 10 8 + 6 ( 10 10 ) l 13 10 9 .
Then, Ω 4 ( l ) > 0 for all l 1.3 and then we have (28). Also, Ω 4 ( l ) is increasing on [ 1.3 , ) with lim l Ω 4 ( l ) = 0 and then we obtain (27). Then, Ω 4 ( l ) is decreasing on [ 1.3 , ) with lim l Ω 4 ( l ) = 0 and then we obtain (26). Finally, we set Ω 5 ( l ) = ln l ψ ( l ) 1 2 ψ ( l ) 1 12 ψ ( l ) 1 20 l 3 + 1 12 l 2 and then
Ω 5 ( l ) = 1 l ψ ( l ) 1 2 ψ ( l ) 1 12 ψ ( l ) + 3 20 l 4 1 6 l 3
and we obtain, for l 13 10 ,
Ω 5 ( l + 1 ) Ω 5 ( l ) = 1 125 4694 + 23115 l 13 10 + 19300 l 13 10 2 + 4500 l 13 10 3 60 l 4 ( 1 + l ) 4 < 0 .
By the same way, we obtain Ω 5 ( l ) > 0 for all l 13 10 and then Ω 5 ( l ) is increasing on [ 1.3 , ) with lim l Ω 5 ( l ) = 0 and then we obtain (29). □
The following lemma is deduced from (3) and (4):
Lemma 2.
The following limits are valid for l R + :
lim l 0 l s + 1 ψ ( s ) ( l ) = ( 1 ) s + 1 s ! , s N { 0 } ,
lim l 0 l s ψ ( s ) ( l + τ ) = 0 , s N , τ > 0
and
lim l 0 l s + 3 ψ ( s ) ( l ) = 0 , s = 0 , 1 , 2 , .

3. CM Degrees of Some Functions Containing Γ , ψ and ψ

In this section, we present two completely monotonic functions involving the gamma and polygamma functions, which will be useful in Section 4 in deducing some new bounds for the gamma and polygamma functions. Also, we investigate their completely monotonic degrees.
Theorem 1.
Suppose that l R + . Then, the function
T ϱ ( l ) = ( l 1 / 2 ) ln l + ln 2 π l ln Γ ( l ) + 1 12 ψ ( l ) + 1 24 ψ ( l + ϱ ) , ϱ 0
is CM on R + if and only if ϱ 3 10 with 1 deg C M l T 3 10 ( l ) < 2 . Also, the function T ϱ ( l ) is CM on R + if and only if ϱ = 0 with 2 deg C M l T 0 ( l ) < 3 .
Proof. 
We use the Binet first Formula [18]:
ln Γ ( l ) = ( l 1 / 2 ) ln l + ln 2 π l + 0 1 / 2 + 1 e t 1 1 t e l t t d t , l R +
we obtain
T ϱ ( l ) = 0 1 / 2 1 e t 1 + 1 t e l t t d t + 1 12 ψ ( l ) + 1 24 ψ ( l + ϱ )
and by using (4), we have T ϱ ( l ) = 0 e l t t 2 ( e t 1 ) χ ϱ ( t ) d t , where
χ ϱ ( t ) = 1 + e t t ( e t + 1 ) 2 + t 3 e t 12 t 4 e ( 1 ϱ ) t 24 .
Let ϱ 3 10 , then we obtain
χ ϱ ( t ) 1 + e t t ( e t + 1 ) 2 + t 3 e t 12 t 4 24 e 7 10 t = 13 t 6 14400 + p = 3 a p 24 ( p + 4 ) ! t p + 4 > 0 ,
where
10 p ( p + 1 ) a p ( p + 2 ) = 2 ( 6 + p ) 10 p 7 p ( 4 + p ) ( 3 + p ) = ( 12 + 2 p ) s = 3 p   s p 7 p s 3 s + p ( p 1 ) ( 9 p + 47 ) 7 p 2 > 0 , p 3
and consequently, T ϱ ( l ) is CM on R + for ϱ 3 10 . On the other side, if T ϱ ( l ) is CM on R + , then by using the asymptotic Formulas (2) and (7), we obtain for l > 0 :
l 3 T ϱ ( l ) = l 3 1 24 l 2 1 24 ( l + ϱ ) 2 + 1 60 l 3 1 24 ( l + ϱ ) 3 + O 1 l 4 > 0 .
Then, lim l l 3 T ϱ ( l ) = ϱ 12 1 40 0 and then ϱ 3 10 . Next, letting ϱ = 0 , we obtain
T 0 ( l ) = 0 p = 1 ( p + 5 ) ( p + 2 ) ( p + 1 ) p t p + 4 24 ( p + 4 ) ! e l t t 2 ( 1 e t ) d t
and then the function T 0 ( l ) is CM on R + . On the other side, assuming that T ϱ ( l ) is CM on R + with ϱ > 0 , and then
l 2 T ϱ ( l ) < 0 , l > 0 , ϱ > 0 .
Using (1), (30), and (31), we have lim l 0 l 2 T ϱ ( l ) = 1 12 > 0 , which is a contradiction with (34). Then, ϱ = 0 . By using the asymptotic Formulas (2) and (7), we obtain T 3 10 ( ) = lim l T 3 10 ( l ) = 0 .
  • Next,
l T 3 10 ( l ) = 0 Δ 1 ( t ) 1 + e t 5 + e 2 t 5 e 3 t 10 e t 10 2 240 t 3 ( e t 10 1 ) 2 ( e t 10 + 1 ) 2 1 + e t 5 + e 2 t 5 + e t 10 + e 3 t 10 2 e l t d t , l R +
where
Δ 1 ( t ) = 480 1 + 2 e t e 2 t + 120 t e 2 t 1 + 240 t 2 e t + 20 t 3 e 2 t e t 20 t 4 e 7 t 10 + e t + e 17 t 10 + t 5 7 e 7 t 10 + 3 e 17 t 10 = 13 t 7 20 + 551 t 8 700 + 896094 t 9 5 ( 9 ! ) + 96987924 t 10 125 ( 10 ! ) + 3575122551 t 11 1250 ( 11 ! ) + 29569195002 t 12 3125 ( 12 ! ) + 7239762907611 t 13 250000 ( 13 ! ) + 10445121705591 t 14 125000 ( 14 ! ) + 5755638359336499 t 15 25000000 ( 15 ! ) + p = 11 b p ( p + 5 ) ! t p + 5 > 0 , p 11
with
b p = 960 1 + ( p 3 ) 2 p + 1 + 20 ( p + 5 ) ( p + 4 ) 12 + ( p + 3 ) 2 p + 2 1 + ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) 7 ( p + 3 ) 7 10 p + ( 3 p 31 ) 17 10 p 20 = 960 1 + ( p 3 ) 2 p + 1 + 20 ( p + 5 ) ( p + 4 ) 12 + ( p + 3 ) 2 p + 2 1 + ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) ( 7 ( p + 3 ) 7 10 p + 3 ( p 11 ) + 2 s = 3 p   s p 7 10 s + 1 200 10260 + 23728 ( p 11 ) + 3605 ( p 11 ) 2 + 147 ( p 11 ) 3 ) > 0 , p 11 .
Hence, 1 deg C M l T 3 10 ( l ) . But,
l 2 T 3 10 ( l ) = Δ 2 ( t ) 2400 t 4 ( e t 10 1 ) 3 ( e t 10 + 1 ) 3 1 e t 10 + e t 5 e 3 t 10 + e 2 t 5 3 1 + e t 10 + e t 5 + e 3 t 10 + e 2 t 5 3 e l t d t , l R +
where
Δ 2 ( t ) = 2400 e 3 t 6 + t 2400 6 + t e 27 t 10 200 120 t + 9 t 2 t 4 2 e 17 t 10 t 4 200 80 t + 71 t 2 + 200 e 2 t 216 + 12 t 24 t 2 12 t 3 2 t 4 + t 5 + 200 e t 216 + 12 t + 24 t 2 12 t 3 + 2 t 4 + t 5 e 7 t 10 t 4 200 + 280 t + 49 t 2
with Δ 2 ( 11.1 ) = 1.59388 ( 10 17 ) and Δ 2 ( 11.2 ) = 1.57813 ( 10 18 ) . Hence, l 2 T 3 10 ( l ) is not CM on R + and then deg C M l T 3 10 ( l ) < 2 . Now, we have T 0 ( ) = lim l T 0 ( l ) = 0 and
l 2 T 0 ( l ) = 0 Δ 3 ( t ) 24 t 4 ( e t 1 ) 3 e l t d t , l R +
where
Δ 3 ( t ) = 144 1 3 e t + 3 e 2 t e 3 t + 24 t 1 e t e 2 t + e 3 t + 48 t 2 e 2 t e t + 24 t 3 e t + e 2 t + 2 t 4 e t + e 3 t + 2 t 5 e t 3 e 2 t + t 6 e t + e 2 t = 3 t 7 5 + 7 t 8 5 + 54 t 9 35 + 4033440 t 10 10 ! + 23923152 t 11 11 ! + 124711488 t 12 12 ! + 593986224 t 13 13 ! + 2649720528 t 14 14 ! + 11256805296 t 15 15 ! + p = 10 c p ( p + 6 ) ! t p + 6 > 0
with
c p = 144 3 + 3 ( 2 p + 6 ) 3 p + 6 + 24 ( p + 6 ) 1 2 p + 5 + 3 p + 5 + 48 ( p + 6 ) ( p + 5 ) 2 p + 4 1 + 24 ( p + 6 ) ( p + 5 ) ( p + 4 ) 1 + 2 p + 3 + 2 ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) 3 p + 2 1 + 2 ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) 1 3 ( 2 p + 1 ) + ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) ( p + 1 ) 1 + 2 p = ( 320640 + 487464 ( p 10 ) + 171322 ( p 10 ) 2 + 25467 ( p 10 ) 3 + 1885 ( p 10 ) 4 + 69 ( p 10 ) 5 + ( p 10 ) 6 ) 2 p + 2304 + 4392 p + 2858 p 2 + 1033 p 3 + 213 p 4 + 23 p 5 + p 6 + 18 108 + 1974 ( p 3 ) + 335 ( p 3 ) 2 + 30 ( p 3 ) 3 + ( p 3 ) 4 3 p > 0 , p 10 .
Hence, 2 deg C M l T 0 ( l ) . But,
l 3 T 0 ( l ) = 0 Δ 4 ( t ) 24 t 5 ( e t 1 ) 4 e l t d t , l > 0
where
Δ 4 ( t ) = 72 e 4 t ( 8 + t ) + 72 ( 8 + t ) 4 e 2 t 864 72 t 2 + 24 t 4 3 t 5 2 t 6 + t 7 e 3 t 2304 144 t + 144 t 2 + 72 t 3 + 24 t 4 + 12 t 5 8 t 6 + t 7 e t 2304 + 144 t + 144 t 2 72 t 3 + 24 t 4 + 4 t 6 + t 7
with Δ 4 ( 6 ) 1.91564 ( 10 11 ) and Δ 4 ( 6.1 ) 6.76109 ( 10 10 ) . Hence, l 3 T 0 ( l ) is not CM on R + and then deg C M l T 0 ( l ) < 3 .
Theorem 2.
For l R + and δ 0 , the function
M δ ( l ) = ( l + δ 1 / 2 ) ln ( l + δ ) + ln 2 π + 1 12 ψ ( l + δ ) + 1 24 ψ ( l + δ ) + 1 40 l 3 ln Γ ( l + δ ) ( l + δ )
is CM on R + if and only if δ 5 18 with 2 deg C M l M 5 18 ( l ) < 3 and the function M δ ( l ) is CM on R + if and only if δ = 0 with 2 deg C M l M 0 ( l ) < 3 .
Proof. 
We use the identity 1 l p = 1 ( p 1 ) ! 0 t p 1 e l t d t for l R + , (see [2]) and (4), (33), we obtain M δ ( l ) = 0 e ( l + δ ) t t 2 ( e t 1 ) Υ δ ( t ) d t , where
Υ δ ( t ) = 1 + e t t ( e t + 1 ) 2 + t 3 e t 12 t 4 e t 24 + t 4 80 ( e t 1 ) e δ t .
Letting δ 5 18 , we obtain Υ δ ( t ) 1 + e t t ( e t + 1 ) 2 + t 3 e t 12 t 4 e t 24 + t 4 80 e 23 t 18 e 5 t 18 . Then,
Υ δ ( t ) 121 t 7 362880 + 1319 t 8 4898880 + p = 5 d p 80 ( p + 4 ) ! t p + 4 > 0 ,
where
d p ( p + 2 ) ( p + 1 ) = 10 p ( p + 5 ) 3 + ( p + 4 ) ( p + 3 ) 23 18 p 5 18 p = ( p 2 ) 18 32 + 53 ( p 5 ) + 5 ( p 5 ) 2 + ( p + 4 ) ( p + 3 ) s = 2 p 1   s p 5 18 s > 0 , p 5 .
Hence M δ ( l ) is CM on R + for δ 5 18 . On the other hand, assuming that M δ ( l ) is CM on R + , then we obtain M δ ( l ) > 0 for l > 0 and we use (2) and (7), we obtain
lim l l 4 M δ ( l ) = lim l l 4 1 40 l 3 1 40 ( l + δ ) 3 1 48 ( l + β ) 4 + O 1 l 5 = 3 δ 40 1 48 0
and then δ 5 18 . Next letting δ = 0 , we obtain
M 0 ( l ) = 0 p = 2 ( 1 + p ) ( 1 + p ) ( 2 + p ) ( 36 + 7 p ) t p + 4 240 ( p + 4 ) ! e l t t 2 ( 1 e t ) d t
and then M 0 ( l ) is CM on R + . On the other side, assuming that M δ ( l ) is CM on R + with δ R + , and then
l 3 M δ ( l ) < 0 , l R + , δ R + .
We use (4), we have lim l 0 l 3 M δ ( l ) = 1 40 > 0 which has a contradiction with (35) and then δ = 0 . Using the asymptotic Formulas (2) and (7), we obtain
M 5 18 ( ) = lim l M 5 18 ( l ) = 0 .
Next,
l 2 M 5 18 ( l ) = 0 Δ 5 ( t ) 38880 t 4 e 5 t 18 ( e t 18 1 ) 3 ( e t 18 + 1 ) 3 1 e t 18 + e t 9 3 1 + e t 18 + e t 9 3 1 e t 6 + e t 3 3 1 + e t 6 + e t 3 3 e l t d t , l R +
where
Δ 5 ( t ) = 233280 e 3 t 3 e 2 t + 3 e t 1 + 4320 t 19 + 39 e t 21 e 2 t + e 3 t 120 t 2 115 813 e t + 633 e 2 t + 65 e 3 t 60 t 3 25 + 263 e t + 983 e 2 t + 25 e 3 t + 36 t 4 27 e 5 t 18 + 40 e t + 81 e 23 t 18 + 100 e 2 t 81 e 41 t 18 140 e 3 t + 27 e 59 t 18 + 10 t 5 299 e t + 742 e 2 t + 205 e 3 t 5 t 6 169 e t + 454 e 2 t + 25 e 3 t = u ( t ) + p = 20 A p ( p + 6 ) ! t p + 6 > p = 20 A p ( p + 6 ) ! t p + 6 ,
where u ( t ) is a polynomial of 25-th degree of t with positive coefficients (by using Wolfram Mathematica 10), and
A p = 233280 3 p + 6 3 ( 2 p + 6 ) + 3 + 4320 ( p + 6 ) 39 21 ( 2 p + 5 ) + 3 p + 5 120 ( p + 6 ) ( p + 5 ) 813 + 633 ( 2 p + 4 ) + 65 ( 3 p + 4 ) 60 ( p + 6 ) ( p + 5 ) ( p + 4 ) 263 + 983 ( 2 p + 3 ) + 25 ( 3 p + 3 ) + 36 ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) 40 + 81 23 18 p + 2 + 100 ( 2 p + 2 ) + 7 59 18 p + 2 + B p + 10 ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) 299 + 742 ( 2 p + 1 ) + 205 ( 3 p + 1 ) 5 ( p + 6 ) ( p + 5 ) ( p + 4 ) ( p + 3 ) ( p + 2 ) ( p + 1 ) 169 + 454 ( 2 p ) + 25 ( 3 p ) ,
where
B p 3 p + 2 = 27 5 54 p + 2 81 41 54 p + 2 140 + 20 59 54 p + 2 > 0.128139 > 0 , p 20
as B p 3 p + 2 is the sum of some increasing sequences, we have B p 3 p + 2 > 0.128139 > 0 for p 20 . By using the binomial theorem, we have
A p > 233280 s = 1 p 5   s p + 6 2 s + s = p 4 p + 6   s p + 6 2 s 3 ( 2 p + 6 ) + 4 + 4320 ( p + 6 ) s = 1 p 5   s p + 5 2 s + s = p 4 p + 5   s p + 5 2 s 21 ( 2 p + 5 ) + 40 120 i = 5 6 ( p + i ) 65 s = 1 p 5   s p + 4 2 s + 65 s = p 4 p + 4   s p + 4 2 s + 633 ( 2 p + 4 ) 748 60 i = 4 6 ( p + i ) 25 s = 1 p 5   s p + 3 2 s + 25 s = p 4 p + 3   s p + 3 2 s + 983 ( 2 p + 3 ) + 288 + 36 i = 3 6 ( p + i ) ( 7 s = 1 p 5   s p + 2 2 s 23 18 p + 2 s + 7 s = p 4 p + 2   s p + 2 2 s 23 18 p + 2 s + 100 ( 2 p + 2 ) + 88 23 18 p + 2 + 40 ) + 10 i = 2 6 ( p + i ) 205 s = 1 p 5   s p + 1 2 s + 205 s = p 4 p + 1   s p + 1 2 s + 742 ( 2 p + 1 ) 94 5 i = 1 6 ( p + i ) 25 s = 1 p 5   s p 2 s + 25 s = p 4 p   s p 2 s + 454 ( 2 p ) + 194 . = 2 p 9 295245 ( 504 ) ( 1280188765694361600 + 2098647666504997920 p + 1003140178986264408 p 2 + 302033387875264572 p 3 + 57665513357826134 p 4 + 9907268194068903 p 5 + 1659133062425685 p 6 + 214596730342998 p 7 + 21045208371756 p 8 + 1583093175207 p 9 + 61823537617 p 10 ) + C p + 3168 i = 3 6 ( p + i ) n = 7 p + 2   n p + 2 5 18 n + D p > 0 , p 20 .
where
C p = 1 17006112 ( 183149035104407040 + 320624709301629504 ( p 19 ) + 93286455487888896 ( p 19 ) 2 + 12866690543164964 ( p 19 ) 3 + 1053248892332396 ( p 19 ) 4 + 56077574544255 ( p 19 ) 5 + 2020495004535 ( p 19 ) 6 + 49561734750 ( p 19 ) 7 + 802353750 ( p 19 ) 8 + 7789375 ( p 19 ) 9 + 34375 ( p 19 ) 10 ) > 0 , p 20
and
D p = 233280 s = 1 p 5   s p + 6 2 s + 4320 ( p + 6 ) s = 1 p 5   s p + 5 2 s 7800 ( p + 6 ) ( p + 5 ) s = 1 p 5   s p + 4 2 s 1500 ( p + 6 ) ( p + 5 ) ( p + 4 ) s = 1 p 5   s p + 3 2 s + 252 i = 3 6 p + i s = 1 p 5   s p + 2 2 s 23 18 p + 2 s + 2050 i = 2 6 p + i s = 1 p 5   s p + 1 2 s 125 i = 1 6 p + i s = 1 p 5   s p 2 s = i = 1 6 p + i s = 1 p 5   s p 2 s m s
with
i = 1 6 p s + i m s = 5 ( 246240 345444 ( p s ) 191140 ( p s ) 2 44875 ( p s ) 3 3825 ( p s ) 4 + 115 ( p s ) 5 + 25 ( p s ) 6 ) + 252 i = 3 6 p s + i 23 18 p s + 2 = 1 136048896 ( 11034829531434240 + 5577195933147744 ( p s 5 ) + 1027054854920088 ( p s 5 ) 2 + 75913157989196 ( p s 5 ) 3 + 1090067398942 ( p s 5 ) 4 + 23389059495 ( p s 5 ) 5 + 24455541975 ( p s 5 ) 6 + 2126055750 ( p s 5 ) 7 + 78960000 ( p s 5 ) 8 + 1894375 ( p s 5 ) 9 + 21875 ( p s 5 ) 10 ) + 252 i = 3 6 p s + i n = 7 p s + 2   n p s + 2 5 18 n > 0 , 5 p s p 1 .
Then, 2 deg C M l M 5 18 ( l ) . But,
l 3 M 5 18 ( l ) = 0 Δ 6 ( t ) 139968 t 5 e 5 t 18 ( e t 18 1 ) 4 ( e t 18 + 1 ) 4 1 e t 18 + e t 9 4 1 + e t 18 + e t 9 4 1 e t 6 + e t 3 4 1 + e t 6 + e t 3 4 e l t d t , l R +
where
Δ 6 ( t ) = 3359232 1 + e t 4 279936 4 + e t 1 + e t 3 t + 5184 1 + e t 2 35 + 187 e t + 10 e 2 t t 2 + 48 1 + e t 400 + 3363 e t + 13458 e 2 t + 275 e 3 t t 3 + 12 125 + 4894 e t + 41256 e 2 t + 23834 e 3 t + 125 e 4 t t 4 + 108 e t 1 + e t 65 + 598 e t + 115 e 2 t t 5 2 e t 6929 + 5613 e t + 34833 e 2 t + 1475 e 3 t t 6 + e t 2197 + 21003 e t + 11667 e 2 t + 125 e 3 t t 7
with Δ 6 ( 5.8 ) = 2.51761 ( 10 15 ) and Δ 6 ( 5.9 ) = 1.40575 ( 10 16 ) . Hence, l 3 M 5 18 ( l ) is not CM on R + and then deg C M l M 3 10 ( l ) < 3 . We also have M 0 ( ) = lim l M 0 ( l ) = 0 and
l 2 M 0 ( l ) = 0 Δ 7 ( t ) 120 t 4 ( e t 1 ) 3 e l t d t , l R +
where
Δ 7 ( t ) = 720 e 3 t 3 e 2 t + 3 e t 1 + 120 t e 3 t e 2 t e t + 1 + 240 t 2 e 2 t e t + 120 t 3 e 2 t + e t + t 4 3 19 e t + 9 e 2 t + 7 e 3 t 10 t 5 3 e 2 t e t + 5 t 6 e 2 t + e t = 5 t 8 2 + 111 t 9 28 + 1667 t 10 504 + 1937 t 11 1008 + 195 t 12 224 + 326213 t 13 997920 + p = 8 θ p ( p + 6 ) ! t p + 6
with
θ p = 8280 + 18882 p + 13219 p 2 + 5003 p 3 + 1056 p 4 + 115 p 5 + 5 p 6 + 9360 + 20138 ( p 4 ) + 3017 ( p 4 ) 2 + 238 ( p 4 ) 3 + 7 ( p 4 ) 4 3 p + 2 + ( 141264 + 542940 ( p 8 ) + 326060 ( p 8 ) 2 + 66735 ( p 8 ) 3 + 6311 ( p 8 ) 4 + 285 ( p 8 ) 5 + 5 ( p 8 ) 6 ) 2 p > 0 , p 8 .
Hence, 2 deg C M l M 0 ( l ) . But,
l 3 M 0 ( l ) = 0 Δ 8 ( t ) 24 t 5 ( e t 1 ) 4 e l t d t , l R +
where
Δ 8 ( t ) = 72 e 4 t ( 8 + t ) + 72 ( 8 + t ) 4 e 2 t 864 72 t 2 + 24 t 4 3 t 5 2 t 6 + t 7 e 3 t 2304 144 t + 144 t 2 + 72 t 3 + 24 t 4 + 12 t 5 8 t 6 + t 7 e t 2304 + 144 t + 144 t 2 72 t 3 + 24 t 4 + 4 t 6 + t 7
with Δ 8 ( 6 ) 1.91564 ( 10 11 ) and Δ 8 ( 6.1 ) 6.76109 ( 10 10 ) . Hence, l 3 M 0 ( l ) is not CM on R + and then deg C M l M 0 ( l ) < 3 .

4. Some New Symmetric Bounds for Γ and ψ s

In this section, we present some new bounds for Γ and ψ s , which are better than some of the recently published results. The following results are deduced from Theorem 1 and 2.
Corollary 2.
Let 0 ω , τ < and l R + . Then,
1 l exp l + ψ ( l ) 12 + ψ ( l + τ ) 24 < Γ ( l ) 2 π l l < 1 l exp l + ψ ( l ) 12 + ψ ( l + ω ) 24 ,
where ω = 3 10 and τ = 0 being the best.
Proof. 
The right-hand side (R.H.S) of (36) is deduced from l 3 T ω ( l ) > 0 which gives ω 3 10 as we mentioned in the proof of Theorem 1. Since ψ ( l ) is increasing on R + , we obtain ψ ( l + 3 10 ) ψ ( l + ω ) for ω 3 10 . Hence, ω = 3 10 is the best in (36). The other side of (36) for τ = 0 is deduced from Theorem 1. We assume that (36) is true for τ > 0 and l R + , then we obtain
lim l 0 l 2 ln Γ ( l ) > lim l 0 l 2 l 1 2 ln l l 3 + l 2 ln 2 π + l 2 12 ψ ( l ) + l 2 24 ψ ( l + τ )
and we use (30) and (31), we obtain lim l 0 l 2 ln Γ ( l ) > 1 12 and this contradicts that lim l 0 l 2 ln Γ ( l ) = 0 . Then, τ = 0 is the best in (36). □
Remark 1.
  • Since ψ ( l ) is CM on R + , we have that the upper bound of (36) improves its counterpart of (12) for all l R + .
  • By using the relations (22) and (23), we obtain that the upper and lower bounds of (36) at ω = 3 10 and τ = 0 , respectively, improve their counterparts of (8) for all l 0.5 and l 5.1 , respectively.
  • Using (21), we have that the upper bound of (36) at ω = 3 10 , improves its counterpart of (11) for all l 0.5 .
  • Using (26), we obtain that the upper bound of (36) at ω = 3 10 , improves its counterpart of (15) for all l 1.3 .
Corollary 3.
Let l R + and 0 ω , τ < . Then,
1 2 l ψ ( l ) 12 ψ ( l + τ ) 24 < ln l ψ ( l ) < 1 2 l ψ ( l ) 12 ψ ( l + ω ) 24 ,
where ω = 3 10 and τ = 0 are the best.
Proof. 
The R.H.S (37) is deduced from T ω ( l ) < 0 . Using (6) and (7), we obtain
lim l l 4 T ω ( l ) = l 4 1 12 ( l + ω ) 3 1 12 l 3 + 1 8 ( l + ω ) 4 1 20 l 4 + O 1 l 5 = 3 40 ω 4 0
and then ω 3 10 . Since ψ ( l ) is increasing on R + , we deduce that ω = 3 10 is the best in (37). The other side of (37) for τ = 0 is concluded from Theorem 1. We assume that (37) is true for τ > 0 and l R + , and then we obtain lim l 0 l 3 ψ ( l ) + ln l > lim l 0 l 3 1 2 l ψ ( l ) 12 ψ ( l + τ ) 24 and this leads to
lim l 0 l 3 ψ ( l ) > 1 12 lim l 0 l 3 ψ ( l ) 1 24 lim l 0 l 3 ψ ( l + τ ) .
We use (30), (31), and (32), we obtain lim l 0 l 3 ψ ( l ) = 0 and lim l 0 l 3 ψ ( l ) 12 lim l 0 l 3 ψ ( τ + l ) 24 = 1 / 6 , which have a contradiction with (38). Then τ = 0 is the best. □
Remark 2.
  • Since ψ ( l ) is completely monotonic on R + , we obtain that the upper bound of (37) improves its counterpart of (13) for all l R + .
  • By using (19) and (24), we have that the upper and lower bounds of (37) at ω = 3 10 and τ = 0 , respectively, improving their counterparts of (9) for all l 0.5 and l 5.1 , respectively.
  • Using (27), we obtain that the upper bound of (37) at ω = 3 10 improves its counterpart of (16) for all l 1.3 .
Corollary 4.
Let 0 ω , τ < , l R + and s = 2 , 3 , . Then,
( s 1 ) ! 2 l s + ( 1 ) s ψ ( s + 1 ) ( l ) 12 + ( 1 ) s ψ ( s + 2 ) ( l + τ ) 24 < ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1
< ( s 1 ) ! 2 l s + ( 1 ) s ψ ( s + 1 ) ( l ) 12 + ( 1 ) s ψ ( s + 2 ) ( l + ω ) 24 ,
where ω = 3 10 and τ = 0 are the best.
Proof. 
The R.H.S. of (39) is deduced from ( 1 ) s T ω ( s ) ( l ) > 0 for s = 2 , 3 , , and by using (7), we obtain
lim l l s + 3 ( 1 ) s T ω ( s ) ( l ) = lim l l s + 3 [ ( s + 1 ) ! 24 1 l s + 2 1 ( l + ω ) s + 2 + ( s + 2 ) ! 120 l s + 3 ( s + 2 ) ! 48 ( l + ω ) s + 3 + O 1 l s + 4 ] = ( s + 2 ) ! ω 24 ( s + 2 ) ! 80 0
and then ω 3 10 . Since ψ ( l ) is CM on R + , we obtain ( 1 ) s ψ ( s + 3 ) ( l ) > 0 on R + , and then ( 1 ) s ψ ( s + 2 ) ( l ) is increasing on R + , and then ω = 3 10 being the best in (39). The other side of (39) for τ = 0 is deduced from Theorem 1. We assume that (39) is true for τ R + and l ( 0 , ) , then we obtain
lim l 0 l s + 2 ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1 > lim l 0 l s + 2 ( s 1 ) ! 2 l s + ( 1 ) s ψ ( s + 1 ) ( l ) 12 + ( 1 ) s ψ ( s + 2 ) ( l + τ ) 24
and then
( 1 ) s lim l 0 l s + 2 ψ ( s 1 ) ( l ) > ( 1 ) s 12 lim l 0 l s + 2 ψ ( s + 1 ) ( l ) + ( 1 ) s 24 lim l 0 l s + 2 ψ ( s + 2 ) ( l + τ ) .
We use (30), (31), and (32), we obtain lim l 0 l s + 2 ψ ( s 1 ) ( l ) = 0 and
( 1 ) s 12 lim l 0 l s + 2 ψ ( s + 1 ) ( l ) + ( 1 ) s 24 lim l 0 l s + 2 ψ ( s + 2 ) ( l + τ ) = ( s + 1 ) ! 12
and this in contradiction with (40). Then, τ = 0 is the best. □
Remark 3.
  • Since ψ ( l ) is CM on R + , we have ( 1 ) s ψ ( s + 2 ) ( l ) < 0 and hence the upper bound of (39) improves its counterpart of (14) for all s = 2 , 3 , , and l > 0 R + .
  • We use (20) and (25), we obtain that the upper and lower bounds of (39) at ω = 3 10 , τ = 0 , and s = 2 improve their counterparts of (10) at s = 2 for all l 0.5 and l 5.1 , respectively.
  • Using (28), we obtain that the upper bound of (39) at ω = 3 10 and s = 2 , refines its counterpart of (17) at s = 2 , for all l 1.3 .
Corollary 5.
For 0 ξ < and l R + , we have
1 l exp l + ψ ( l ) 12 + ψ ( l ) 24 + 1 40 l 3 < Γ ( l ) 2 π l l
< 1 l exp l + ψ ( l ) 12 + ψ ( l ) 24 + 1 40 ( l ξ ) 3 ,
where the upper bound is valid for l > ξ and the other bound is valid for l R + with ξ = 5 18 is the best.
Proof. 
The R.H.S. of (41) is deduced from M ξ ( l ξ ) > 0 for l > ξ and we use (2) and (7), we obtain
lim l l 4 M ξ ( l ξ ) = lim l l 4 1 40 ( l ξ ) 3 1 40 l 3 1 48 l 4 + O 1 l 5 = 3 ξ 40 1 48 0
and then ξ 5 18 . As the function 1 y 3 is decreasing on R + , we deduce that ξ = 5 18 is the best in (41). The other side (41) is deduced from M 0 ( l ) < 0 in Theorem 2. □
Remark 4.
  • The lower bound of (41) improves its counterpart of (36) at τ = 0 , for all l R + .
  • By using (29), we deduce that the lower bound of (41) improves its counterpart of (18) for all l 1.3 .
Corollary 6.
For 0 ξ < and l R + , we have
1 2 l ψ ( l ) 12 ψ ( l ) 24 + 3 40 l 4 < ln l ψ ( l ) < 1 2 l ψ ( l ) 12 ψ ( l ) 24 + 3 40 ( l ξ ) 4 ,
where the upper bound is valid for l > ξ and the other bound is valid for l R + with ξ = 5 18 is the best.
Proof. 
The R.H.S of (42) is deduced from M ξ ( l ξ ) < 0 for l > ξ and then
l 5 M ξ ( l ξ ) = l 5 ln l ψ ( l ) 1 2 l + 1 12 ψ ( l ) + 1 24 ψ ( l ) 3 40 ( l ξ ) 4 < 0 .
Using (6) and (7), we have
lim l l 5 M ξ ( l ξ ) = lim l l 5 3 40 l 4 3 40 ( l ξ ) 4 + 1 12 l 5 + O 1 l 6 = 1 12 3 ξ 10 .
From (43), we conclude that ξ 5 18 . Similarly, as before, we obtain ξ = 5 18 , which is the best in (42). The other side of (42) is deduced from M 0 ( l ) > 0 in Theorem 2. □
Remark 5.
The lower bound of (42) improves its counterpart of (37) at τ = 0 , for all l R + .
Corollary 7.
For 0 ξ < , l R + and s = 2 , 3 , , we have
( s 1 ) ! 2 l s + ( 1 ) s ψ ( s + 1 ) ( l ) 12 + ( 1 ) s ψ ( s + 2 ) ( l ) 24 + ( s + 2 ) ! 80 l s + 3 < ( 1 ) s ψ ( s 1 ) ( l ) ( s 2 ) ! l s 1
< ( s 1 ) ! 2 l s + ( 1 ) s ψ ( s + 1 ) ( l ) 12 + ( 1 ) s ψ ( s + 2 ) ( l ) 24 + ( s + 2 ) ! 80 ( l ξ ) s + 3
where the upper bound is valid for l > ξ and the other bound is valid for l R + with ξ = 5 18 is the best.
Proof. 
The R.H.S of (44) is deduced from
( 1 ) s M ξ ( s ) ( l ξ ) = ( s 2 ) ! l s 1 + ( 1 ) s + 1 ψ ( s 1 ) ( l ) + ( s 1 ) ! 2 l s + ( 1 ) s 12 ψ ( s + 1 ) ( l )
+ ( 1 ) s 24 ψ ( s + 2 ) ( l ) + ( s + 2 ) ! 80 ( l ξ ) s + 3 > 0 , s = 2 , 3 , , l > ξ .
Using (7), we have
lim l l s + 4 ( 1 ) s M ξ ( s ) ( l ξ ) = lim l l s + 4 ( s + 2 ) ! 80 ( l ξ ) s + 3 ( s + 2 ) ! 80 l s + 3 ( s + 3 ) ! 288 l s + 4 + O 1 l s + 5 = ( s + 3 ) ! ξ 5 18 80 > 0
and then ξ 5 18 . By the same way, we have that ξ = 5 18 is the best in (44). The other side of (44) is deduced from ( 1 ) s M 0 ( s ) ( l ) < 0 in Theorem 2. □
Remark 6.
The lower bound of (44) improves its counterpart of (39) at τ = 0 , for every s = 2 , 3 , and l > 0 .

5. Conclusions

The main conclusions of this paper are stated in Theorems 1 and 2. Concretely speaking, we studied the monotonicity of two functions involving the gamma and polygamma functions to deduce some symmetric bounds for the gamma, psi, and polygamma functions in terms of some exponents of the polygamma functions, which are mentioned in Corollaries 2–7. These results give more accurate bounds for the gamma, psi, and polygamma functions, which are better than the results presented by Alzer and Batir, Batir, Şevli and Batir, and Moustafa, which are mentioned in (8)–(18).

Author Contributions

Writing the original draft, W.A.S. and H.M. All authors contributed equally to the writing of this paper. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Sandor, J. Selected Chapters of Gometry, Analysis and Number Theory. In RGMIA Monographs; Victoria University: Melbourne, Australia, 2005. [Google Scholar]
  2. Abramowitz, M.; Stegun, I.A. (Eds.) Handbook of Mathematical Functions with Formulas; Graphs and Mathematical Tables: Dover, NY, USA, 1965. [Google Scholar]
  3. Batir, N. Sharp bounds for the psi function and harmonic numbers. Math. Inequal. Appl. 2011, 14, 917–925. [Google Scholar] [CrossRef]
  4. Olver, F.W.J.; Lozier, D.W.; Boisvert, R.F.; Clark, C.W. (Eds.) NIST Handbook of Mathematical Functions; Cambridge University Press: New York, NY, USA, 2010. [Google Scholar]
  5. Qi, F. Limit formulas for ratios between derivatives of the gamma and digamma functions at their singularities. Filomat 2013, 27, 601–604. [Google Scholar] [CrossRef]
  6. Yang, Z.-H.; Tian, J.-F. Monotonicity rules for the ratio of two Laplace transforms with applications. J. Math. Anal. Appl. 2019, 470, 821–845. [Google Scholar] [CrossRef]
  7. Widder, D.V. The Laplace Transform; Princeton University Press: Princeton, NJ, USA, 1946. [Google Scholar] [CrossRef]
  8. Mahmoud, M.; Qi, F. Bounds for completely monotonic degrees of remainders in asymptotic expansions of the digamma function. Math. Inequal. Appl. 2022, 25, 291–306. [Google Scholar] [CrossRef]
  9. Qi, F. Completely monotonic degree of a function involving trigamma and tetragamma functions. AIMS Math. 2020, 5, 3391–3407. [Google Scholar] [CrossRef]
  10. Qi, F.; Guo, S.; Guo, B.-N. Complete monotonicity of some functions involving polygamma functions. J. Comput. Appl. Math. 2010, 233, 2149–2160. [Google Scholar] [CrossRef]
  11. Alzer, H.; Batir, N. Monotonicity properties of the gamma function. Appl. Math. Lett. 2007, 20, 778–781. [Google Scholar] [CrossRef]
  12. Batir, N. On some properties of the gamma function. Expo. Math. 2008, 26, 187–196. [Google Scholar] [CrossRef]
  13. Şevli, H.; Batir, N. Complete monotonicity results for some functions involving the gamma and polygamma functions. Math. Comput. Model. 2011, 53, 1771–1775. [Google Scholar] [CrossRef]
  14. Moustafa, H. Inequalities for completely monotonic degrees of functions involving gamma and polygamma functions. Turk. J. Inequal. 2024, 8, 68–85. [Google Scholar]
  15. Mahmoud, M.; Almuashi, H.; Moustafa, H. An asymptotic expansion for the generalized gamma function. Symmetry 2022, 14, 1412. [Google Scholar] [CrossRef]
  16. Moustafa, H.; Almuashi, H.; Mahmoud, M. On some complete monotonicity of functions related to generalized k-gamma function. J. Math. 2021, 2021, 9941377. [Google Scholar] [CrossRef]
  17. Zhang, J.M.; Yin, L.; You, H.L. Complete monotonicity and inequalities related to generalized k-gamma and k-polygamma functions. J. Inequal. Appl. 2020, 2020, 21. [Google Scholar] [CrossRef]
  18. Andrews, G.E.; Askey, R.; Roy, R. Special Functions, Encyclopedia of Mathematics and its Applications 71; Cambridge University Press: Cambridge, UK, 1999. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Al Sayed, W.; Moustafa, H. Some New Inequalities for the Gamma and Polygamma Functions. Symmetry 2025, 17, 595. https://doi.org/10.3390/sym17040595

AMA Style

Al Sayed W, Moustafa H. Some New Inequalities for the Gamma and Polygamma Functions. Symmetry. 2025; 17(4):595. https://doi.org/10.3390/sym17040595

Chicago/Turabian Style

Al Sayed, Waad, and Hesham Moustafa. 2025. "Some New Inequalities for the Gamma and Polygamma Functions" Symmetry 17, no. 4: 595. https://doi.org/10.3390/sym17040595

APA Style

Al Sayed, W., & Moustafa, H. (2025). Some New Inequalities for the Gamma and Polygamma Functions. Symmetry, 17(4), 595. https://doi.org/10.3390/sym17040595

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop