The uniqueness of the integration factor associated with the exchanged heat in thermodynamics Yu-Han Ma1,2, Hui Dong2, Hai-Tao Quan3 and Chang-Pu Sun1,2* Address: Beijing Computational Science Research Center, Beijing 100193, China Address: Graduate School of China Academy of Engineering Physics, No. 10 Xibeiwang East Road, Haidian District, Beijing, 100193, China Address: School of Physics, Peking University, Beijing, 100871, China * Correspondence: [email protected] Abstract State functions play important roles in thermodynamics. Different from the process function, such as the exchanged heat 훿푄 and the applied work 훿푊, the change of the state function can be expressed as an exact differential. We prove here that, for a generic thermodynamic system, only the inverse of the temperature, namely 1/푇, can serve as the integration factor for the exchanged heat 훿푄. The uniqueness of the integration factor invalidates any attempt to define other state functions associated with the exchanged heat, and in turn, reveals the incorrectness 2 of defining the entransy 퐸푣ℎ = 퐶푉푇 /2 as a state function by treating 푇 as an integration factor. We further show the errors in the derivation of entransy by treating the heat capacity 퐶푉 as a temperature-independent constant. I. Introduction State functions, e.g., the internal energy and the entropy, in thermodynamics characterize important features of the system in a thermal equilibrium state [1]. Physically, some quantities are process-dependent and thus cannot be treated as state functions, e.g., the exchanged heat 훿푄 and the applied work 훿푊. An integration factor 푓 can be utilized to convert the process function, e.g., the exchanged heat, into an exact differential (the change of a state function). Mathematically, the requirement of the state function can be expressed as follows: the change of the state function remains unchanged with any topological variation of the integration path on the parameter space [2]. In addition, the number of the integration factors is usually limited for any system with more than two thermodynamic variables. Such number of the integration factor is further reduced in order to define a universal state function without the dependence on the system characteristics [2]. It is a common sense that the inverse of the temperature, i.e., 1/푇, serves as the integration factor for the exchanged heat 훿푄, and thus a state function, the entropy can be defined. A relevant question arises here: is there any other universal integration factor associated with the exchanged heat 훿푄 for an arbitrary system? Several attempts on this issue have been made for both specific systems [3] and generic systems [4]. However, the uniqueness of the integration factor of interest for a generic thermodynamic system remains unexplored. In our current paper, from the first principle, we prove that only one factor, namely 1/푇, can serve as the integration factor to convert the exchanged heat into the state function. Such uniqueness invalidates any attempt to find new state functions associated with the exchanged heat 훿푄. It is worth mentioning that the introduction of the so-called “state function”, the entransy defined via 훿퐸푣ℎ = 푇훿푄 [5], in the realm of the heat transfer is an example of such attempt. The entransy is claimed to be a “state function” of a system characterizing its potential of heat transfer [5]. Since its appearance, it has triggered a lot of debates [6-10] over its validity as state function [6] as well as its usefulness [7-10] in the practical application in the field of heat transfer. In this paper, from the fundamental principles of thermodynamics [11-13] and with the help of statistical mechanics, we reveal the improperness and incorrectness of such a concept. The rest of the paper is organized as follow. In Sec. II, we prove the uniqueness of the integration factor 1/푇 and the resultant entropy as the state function. In Sec. III, we further discuss the errors in the definition of the entransy. The conclusions are given in Sec. IV II. Proof of the uniqueness of the integration factor ퟏ/푻 In this section, we will prove the uniqueness of the integration factor associated with the exchanged heat 훿푄. As an illustration, we first demonstrate the result in the ideal gas with the internal energy 푈, the temperature 푇, the volume 푉, and the pressure 푃. According to the first law of thermodynamics, the changed heat of the gas reads 훿푄 = 푑푈 − 훿푊, which can be further written as (with the work done on the ideal gas 훿푊 = −푃푑푉) 훿푄 = 푑푈 + 푃푑푉. (1) An infinitesimal change of a generic thermodynamic function Λ can be defined as 훿Λ = 푓(푇)훿푄, namely, 훿Λ = 푓(푇)(푑푈 + 푃푑푉). (2) Figure 1. The evolution path of the ideal gas in the parameter space 푇 − 푉 As illustrated in Fig. 1, the state (푇0, 푉0) at point A can be connected by two paths (푙1 푎푛푑 푙2) to the state (푇푓, 푉푓) at point B in the 푇 − 푉 space. If Λ is a state function, it is required that 푙1 푙2 ∫A→B 훿Λ = ∫A→B 훿Λ [2]. Accordingly, the loop integral of 훿Λ in the 푇 − 푉 space is strictly zero, namely, 푙1 푙2 푙1 푙2 ∮ 훿Λ = ∫ 훿Λ + ∫ 훿Λ = ∫ 훿Λ − ∫ 훿Λ = 0. (3) A→B B→A A→B A→B With Eq. (2), the above equation can be specifically written as 푛푅푇 ΔΛ = ∮ 훿Λ = ∮ 푓(푇) (퐶 푑푇 + 푑푉) = 0, (4) 푉 푉 where we have used 푑푈 = 퐶푉푑푇 and the equation of state of the ideal gas 푃푉 = 푛푅푇. Here, 퐶푉 is the heat capacity at a constant volume, 푛 is the number of moles of the gas, and 푅 is the gas constant. With Green’s theorem, the loop integral in Eq. (3) can be rewritten in the form of the surface integration as 휕 휕 푛푅푇 ΔΛ = ∬ { [푓(푇)퐶 ] − [푓(푇) ]} 푑푉푑푇 = 0. (5) 휕푉 푉 휕푇 푉 This equation is valid irrespective of the integral intervals in the 푇 − 푉 space. Hence, we have 휕퐶 휕 푛푅푇 푓(푇) 푉 − [푓(푇) ] = 0. (6) 휕푉 휕푇 푉 The fact 휕퐶푉/휕푉 = 0 for the ideal gas further simplifies the above equations to 휕 푛푅푇 [푓(푇) ] = 0, (7) 휕푇 푉 whose solution can be uniquely determined as 훼 푓(푇) = , (8) 푇 with 훼 being an arbitrary constant independent of the temperature 푇. Without losing generality, we choose 훼 = 1, i.e., 훿Λ = 훿푄/푇. Thus, 훿Λ is nothing but the change of the thermodynamic entropy 푑푆. In summary, we prove that for the classical ideal gas only one state function associated with the exchanged heat, namely the entropy, can be defined. A similar proof for such an ideal gas system was proposed by Weiss [3]. Having shown the uniqueness of the integration factor associated with the exchanged heat for the classical ideal gas, in the following, we will prove a theorem on the uniqueness of the integration factor associated with 훿푄 for a generic thermodynamic system. Theorem: For a generic thermodynamic system, the universal thermodynamic state function Λ can be defined as 푑Λ = 푓(푇, 휆)훿푄 , if and only if 푓(푇, 휆) = α/푇 with α, a system- independent constant. Proof: For a generic thermodynamic system with the internal energy 푈, the first law of thermodynamic law reads 훿푄 = 푑푈 − 훿푊 = 푑푈 − 푌푑휆, (9) where 푌 and 휆 are the generalized force and the generalized displacement, respectively. Similar to the discussions for the ideal gas system, the condition for 훿Λ = 푓(푇, 휆)훿푄 to be an exact differential (Λ to be a state function) is: For an arbitrary loop in the 푇 − 푉 space, we always have ΔΛ = ∮ 푓(푇, 휆)훿푄 = ∮ 푓(푇, 휆) (푑푈 − 푌푑휆) = 0. (10) Noticing the relation for the internal energy 휕푈 휕푈 푑푈 = 푑푇 + 푑휆, (11) 휕푇 휕휆 Eq. (10) can be rewritten in the form of the surface integration with Green’s theorem as 휕 휕푈 휕 휕푈 ΔΛ = ∬ { [푓(푇, 휆) ] − [푓(푇, 휆) ( − 푌)]} 푑휆푑푇 = 0. (12) 휕휆 휕푇 휕푇 휕휆 The above equation is valid irrespective of the integration intervals. Hence, we have 휕 휕푈 휕 휕푈 [푓(푇, 휆) ] − [푓(푇, 휆) ( − 푌)] = 0, (13) 휕휆 휕푇 휕푇 휕휆 which can be further simplified as 휕푈 휕푓(푇, 휆) 휕푌 휕푈 휕푓(푇, 휆) + 푓(푇, 휆) + (푌 − ) = 0. (14) 휕푇 휕휆 휕푇 휕휆 휕푇 In statistical mechanics, the internal energy and the generalized force can be written as [1] 푗=푁 푈 = ∑ 푝푗퐸푗 , (15) 푗=1 and 푗=푁 푑퐸푗 푌 = (∑ 푝 ) , (16) 푗 푑휆 푗=1 where 퐸푗 = 퐸푗(휆) (푗 = 1,2 ⋯ 푁) is the 푗-th energy level of the system [For simplicity, we consider a system with discrete energy levels. But it is straightforward to extend our discussions to systems with a continuous energy spectrum.], and it is 휆-dependent. 푒−훽퐸푗(휆) 푒−훽퐸푗 푝푗 = 푗=푁 ≡ (17) −훽퐸푗(휆) 푍 ∑푗=1 푒 is the corresponding thermal equilibrium distribution of the system on the 푗-th energy level with 훽 = 1/(푘퐵푇) as the inverse temperature and 푘퐵 as the Boltzmann constant. Combining Eqs. (15-17), we find 푗=푁 −훽퐸푗 2 2 휕푈 휕 푒 〈퐸푗 〉 − 푈 = (∑ 퐸푗) = 2 ≡ 퐶휆, (18) 휕푇 휕푇 푍 푘퐵푇 푗=1 푗=푁 푗=푁 −훽퐸푗 휕푌 휕 푒 푑퐸푗 훽 푑퐸푗 = (∑ ) = (∑ 푝 퐸 − 푈푌) , (19) 휕푇 휕푇 푍 푑휆 푇 푗 푗 푑휆 푗=1 푗=1 and 푗=푁 푗=푁 −훽퐸푗 휕푈 휕 푒 푑퐸푗 = (∑ 퐸 ) = 푌 + 훽푈푌 − 훽 ∑ 푝 퐸 . (20) 휕휆 휕휆 푍 푗 푗 푗 푑휆 푗=1 푗=1 Substituting the above three relations into Eq.