Why a Bose-Einstein condensate cannot exist in a system of interacting bosons

at ultrahigh temperatures
Maksim D. Tomchenko∗
Bogolyubov Institute for Theoretical Physics, 14b, Metrolohichna Str., Kyiv 03143, Ukraine
(Dated: January 7, 2025)
It is well known that a Bose-Einstein (BE) condensate of atoms exists in a system of interacting
(i) (i)
Bose atoms at T < ∼ Tc , where Tc is the BE condensation temperature of an ideal gas. It is
(i)
also generally accepted that BE condensation is impossible at “ultrahigh” temperatures T ≫ Tc .
While the latter property has been theoretically proven for an ideal gas, no such proof exists for an
arXiv:2501.03029v1 [cond-mat.other] 6 Jan 2025

interacting system, to our knowledge. In this paper, we propose an approximate mathematical proof
for a finite, nonrelativistic, periodic system of N spinless interacting bosons. The key point is that,
(i) P
at T ≫ Tc , the main contribution to the occupation number N0 = Z1 ℘ e−E℘ /kB T hΨ℘ |â+ 0 â0 |Ψ℘ i,
corresponding to atoms with zero momentum, originates from the states containing N elementary
quasiparticles. These states do not contain the BE condensate of zero-momentum atoms, implying
that an ultrahigh temperature should “blur” such a condensate.

(i)
Keywords: Bose-Einstein condensate, ultrahigh tem- which are typically below Tc ). For an infinite three-
peratures. dimensional ideal gas, it has been theoretically found that
(i) (i)
N0 (T )/N = 1 − (T /Tc )3/2 at T ≤ Tc and N0 = 0 at
(i)
T > Tc , where N0 is the number of zero-momentum
I. INTRODUCTION atoms [12, 13]. It is widely accepted, as supported by
(i)
experimental evidence, that at T ≫ Tc there is no con-
It is known that in a system of interacting bosons, densate also in an interacting Bose gas. However, to
Bose-Einstein (BE) condensation occurs at temperatures our knowledge, this rather obvious property has not been
below or of the order of the BE condensation temperature rigorously proven theoretically. In what follows, we pro-
(i) pose a sketchy mathematical proof, partly grounded in
Tc of an ideal gas [1, 2]. The properties of a BE con- physical considerations, that BE condensation of zero-
densate of atoms have been well studied, first for liquid momentum atoms is impossible for a system of interact-
4
He and later for trapped Bose gases. Moreover, a con- ing spinless bosons at ultrahigh temperatures.
densate of interacting mesons was theoretically predicted
long ago (see reviews [3, 4]), but, as far as we know, there
is no reliable experimental confirmation of its existence. II. MAIN PART
In addition, BE condensation has been predicted and ex-
perimentally detected for photons [10] and quasiparticles
For an ideal Bose gas, the condensate is typically eval-
such as magnons [5, 6], excitons [7, 8], and polaritons
uated by the method proposed as early as by A. Einstein
[9] (see also review [11] and books [1, 2]). Furthermore,
[14]: the system is treated as a collection of particles,
while BE condensation in liquid 4 He and trapped gases
and the occupation of the single-particle states is anal-
occurs at low (∼ 1 K) and ultralow ( < −5
∼ 10 K) temper- ysed. Such a method, however, is not applicable to a
atures, respectively, BE condensation of quasiparticles
system of N interacting bosons, since each state of such
and mesons occurs at high temperatures, ∼ 100 K and
a system is described by an N -particle wave function
∼ 1012 K respectively.
(WF) Ψ(r1 , . . . , rN ), and single-particle states are not
Despite the significant variation in temperature, in all well-defined. In this case, the condensate can be calcu-
these four cases, the condensate corresponds to tempera- lated using a more general and universal approach based
tures below or of the order of the BE condensation tem- on the diagonal expansion of the single-particle density
(i)
perature of an infinite ideal gas, Tc ≈ 3.3~2 n2/3 /m matrix F1 (r, r′ ) in the complete orthonormalised basis
[12, 13] (this is the formula for the nonrelativistic case; {φj (r)} [12]:
m and n = N/V denote the boson mass and density).
∞
Since the physical measure of temperature is precisely X
(i) F1 (r, r′ ) = λj φ∗j (r′ )φj (r). (1)
the value of Tc , all the BE condensates mentioned above
j=1
are low-temperature ones. However, there are also tem-
(i)
peratures T ≫ Tc , which we will call ultrahigh temper- Here, λj are the occupation numbers of the single-particle
atures (to distinguish them from “high” temperatures, states φj (r), and λj /N is the probability that a parti-
cle occupies the state φj (x) (we use the normalisation of
F1 (r, r′ ), for which λ1 + . . . + λ∞ = N ). The state φ1 (r)
corresponds to the condensate if λ1 ∼ N . This is the def-
∗ Electronic address: [email protected] inition of the condensate of Bose particles for a system
 2

of many interacting bosons. This definition is also valid where each term in the numerator and denominator is
(℘)
for free bosons, then each λj can take natural values or non-negative: e−E℘ /kB T > 0, N0 ≥ 0. For a small
be zero. number of quasiparticles, one has
Consider a nonrelativistic, equilibrium, uniform sys- X
tem of interacting spinless bosons with a fixed number E℘ = E0 + npj ǫ(pj ), (8)
N of particles at temperature T . For simplicity, we use pj
periodic boundary conditions. In the formalism of second
quantisation, we have where E0 is the ground-state energy and ǫ(pj ) is the
energy of the quasiparticle with momentum pj . When
F1 (r, r′ ) = hψ̂ + (r′ )ψ̂(r)i, (2) the number of quasiparticles is large (∼ N ), interac-
tions between them must be accounted for, leading to
a modification of the dispersion relation in (8): ǫ(pj ) →
1 X 1 X + −ikr ǫ(pj ; np1 , np2 , . . .). Each state of the system of N in-
ψ̂(r) = √ âk eikr , ψ̂ + (r) = √ âk e . (3) teracting bosons is an N -particle state. Accordingly, (i)
V k V k
one should describe such a system in the language of col-
lective excitations (elementary quasiparticles), even at
Here and below, hi stands for the statistical average over
ultrahigh temperatures. Experiment shows that at ul-
the canonical ensemble [13]:
trahigh temperatures the matter is in a gaseous, non-
1
Z X superfluid state. Therefore, we assume that our sys-
hÂiT >0 = dr1 . . . drN e−E℘ /kB T Ψ∗℘ ÂΨ℘ tem is such apgas, and that the Bogoliubov formula
Z ℘ [15] ǫB (k) = K 2 (k) + 2n0 ν(k)K(k), where K(k) =
2 2
1 X −E℘ /kB T ~ k /2m and n0 = N0 /V , approximately describes the
≡ e hΨ℘ |Â|Ψ℘ i, (4) quasiparticle dispersion law. It is well known that (ii)
Z ℘
such a gas can be treated as a collection of N inter-
acting atoms with the Maxwell-Boltzmann momentum
where Z = ℘ e−E℘ /kB T , {Ψ℘ } represents a complete
P
distribution (note that known to us methods of deriv-
orthonormalised set of WFs of the system with a fixed ing this distribution imply continuity and differentiabil-
number N of particles, and ℘ numerates the physically ity of the distribution function, and thus the absence of
different states. The WFs Ψ℘ and Ψ℘ i describe the same condensate). We are also aware that (iii) the free en-
stationary state in the coordinate representation and the ergy of the system, F = −kB T ln Z, and other thermo-
second quantised representation, respectively. From Eqs. dynamic quantities can be derived from the statistical
(2)–(4) we obtain sum [13]. Properties (i), (ii) and (iii) jointly imply that
at ultrahigh temperatures, the main contribution to the
1 X + ′
statistical sum Z = ℘ e−E℘ /kB T should come from the
P
F1 (r, r′ ) = hâq âk ieikr−iqr
V states, containing N quasiparticles with the dispersion
k,q
1 X + ′ X law ǫ(p) ≈ ~2 p2 /2m, which follows from the Bogoliubov
= hâk âk ieik(r−r ) ≡ Nk φ∗k (r′ )φk (r), (5) formula ǫB (p) at |p| → ∞. This establishes a correspon-
V
k k dence between the quasiparticle and particle descriptions
of the system. We henceforth use the language of quasi-
because hΨ℘ |â+ |Ψ℘ i = 0 if q 6= k; here Nk = hâ+
q âk√ k âk i particles as more rigorous. It is worth noting that al-
ikr
and φk (r) = e / V . though only single-quasiparticle states [16, 17] are usu-
Thus, for the periodic system, the expansion of ally described quantum mechanically, it is also possible
F1 (r, r′ ) directly yields a diagonal form, and the num- to accurately describe two- [18] and multi-quasiparticle
ber of zero-momentum atoms is given by the formula [18, 19] states.
The formula (7), being exact, directly implies that a
1 X −E℘ /kB T BE condensate should be absent at ultrahigh tempera-
N0 = hâ+
0 â0 i = e hΨ℘ |â+
0 â0 |Ψ℘ i. (6) tures. Indeed, at T = 0, only the ground state (i.e., the
Z ℘
state without quasiparticles) contributes to the formula
(0)
Here, each number ℘ corresponds to a set of ele- (7), yielding N0 = N0 . The structure of the ground-
mentary quasiparticles with momenta p1 , p2 , . . .: ℘ ≡ state WF of a Bose gas and a Bose liquid ensures the
{np1 , np2 , . . .}, where npj is the number of quasiparti- existence of a BE condensate of atoms with zero momen-
(0)
cles with momentum pj . Let us denote hΨ℘ |â+ tum: N0 = N0 ∼ N [13, 20–24]. At high temperatures,
0 â0 |Ψ℘ i = (i)
(℘)
N0 , then T >∼ Tc , a huge number of non-negative terms con-
tribute to the sums in the numerator and denominator of
(℘) the right-hand side of Eq. (7). Therefore, at high T : (i)
N0 e−E℘ /kB T
P
℘
the change in N0 with increasing T must be very smooth,
N0 = P −E /k T , (7) (℘)
e ℘ B and (ii) for N0 ∼ N to hold, it is necessary that N0 ∼ N
℘ be valid for the majority of the significant terms in (7).
 3

(i)
At T ≫ Tc , the main contribution to the statistical Consider a dilute Bose gas. In this case [26],
sum comes from states, containing N quasiparticles with
X c2 (k)
large momenta |p| (see above; this is apparently due to Ψ0 = A0 eS , S≈ ρk ρ−k , (13)
the fact that such states constitute the majority among 2!
k6=0
all possible states; interestingly, the maximum possible
number of elementary quasiparticles is also N [18, 19]).
(i)
Consequently, in order to get N0 ∼ N at T ≫ Tc , it is c2 (k) ≈ 1/2 − mǫB (k)/(~2 k 2 ). (14)
(℘)
necessary and sufficient that N0 ∼ N holds for states It follows from the treatment in works [23, 27] that for
with N quasiparticles with large momenta |p|. (The suf- liquid He II for configurations (r1 , . . . , rN ) correspond-
ficiency of this condition is evident; and its necessity fol- ing to a uniform atomic distribution (|ρk | ∼ 1), one has
(℘)
lows from the fact that if N0 ∼ N holds only for states S ≡ Sl ∼ −N . Consequently, for a gaseous 4 He it should
containing < N quasiparticles, the condensate would be be S ≡ Sg ∼ −|a|N , where |a| is a small number that
suppressed by the statistical sum. This is because the depends on the densities of the gas and liquid. Since
main contribution to the sum is provided by states with |Sg | ≪ |Sl |, almost all atoms of the gas are in the con-
N quasiparticles, which greatly outnumber states with densate: N0 ≈ N , N − N0 ≪ N [15]. Therefore, when
fewer quasiparticles.) The following qualitative reason- calculating the condensate in the gas at T = 0, it suffices
ing shows that this is impossible. to assume Sg = 0 in a zero-order approximation, yielding
The WF of the ground state of a Bose gas or a Bose
liquid has the form [18, 25] Ψ0 (r1 , . . . , rN ) ≈ 1. (15)

Ψ0 (r1 , . . . , rN ) = A0 eS(r1 ,...,rN ) , (9) (In (15) and below, we omit normalisation constants;
note that the approximation (15) is unsuitable for the
derivation of the dispersion law, as it results in ǫ(k) ∼ k 2
X c2 (q1 ) for all k.) A state with a single Bogoliubov’s quasiparti-
S = ρq1 ρ−q1 cle with momentum p is approximately described by the
2!
q1 6=0 WF [16–18, 28]
q1 +q
X2 6=0 c3 (q1 , q2 )
+ ρq1 ρq2 ρ−q1 −q2 + . . . Ψp (r1 , . . . , rN ) ≈ ρ−p Ψ0 . (16)
q1 ,q2 6=0
3!N 1/2
Given (15), we may write
q1 +...+qN −1 6=0
X cN (q1 , . . . , qN −1 )
+ × Ψp (r1 , . . . , rN ) ≈ ρ−p . (17)
q1 ,...,qN −1 6=0
N !N (N −2)/2
× ρq1 . . . ρqN −1 ρ−q1 −...−qN −1 , (10) Similarly, the state containing a quasiparticle with mo-
mentum p1 and another quasiparticle with momentum
PN p2 , in the zero-order approximation can be described by
where ρk = √1N j=1 e−ikrj . Any excited state (with
the WF [18, 28]
total momentum ~p and any number ≤ N of quasiparti-
cles) is described by the WF [16, 18] Ψp1 p2 (r1 , . . . , rN ) ≈ ρ−p1 ρ−p2 , (18)

Ψp (r1 , . . . , rN ) = Ap ψp Ψ0 , (11) and a state with N quasiparticles with momenta

p1 , p2 , . . . , pN is described in the same approximation
by the WF
+p6=0
q1X
b2 (q1 ; p)
ψp = b1 (p)ρ−p + ρq1 ρ−q1 −p Ψp1 p2 ...pN (r1 , . . . , rN ) ≈ ρ−p1 ρ−p2 · · · ρ−pN . (19)
q1 6=0
2!N 1/2
q1 +q
There are two ways of showing that the maximum pos-
2 +p6=0
X b3 (q1 , q2 ; p) sible number of elementary quasiparticles is N (see ap-
+ ρq1 ρq2 ρ−q1 −q2 −p + . . .
3!N pendix 1 in [18] and section 7 in [29]). For a Bose gas, this
q1 ,q2 6=0
can be seen without using formulae: quasiparticles with
q1 +...+qN −1 +p6=0
X bN (q1 , . . . , qN −1 ; p) large momentum |p| have the energy ǫ(p) ≈ ~2 p2 /2m
+ × [15] and are similar to free atoms; the number of the lat-
N !N (N −1)/2
q1 ,...,qN −1 6=0 ter is equal to N , hence the maximum possible number
× ρq1 . . . ρqN −1 ρ−q1 −...−qN −1 −p . (12) of elementary quasiparticles with large |p| is also equal
to N . The treatment in [18, 29] shows that this conclu-
The formulae (9)–(12) are exact: (10) and (12) are simply sion holds for elementary quasiparticles with arbitrary
expansions of the functions ln Ψ0 and Ψp in the complete momenta. Interestingly, the number of rotons in 4 He at
set of Bose-symmetric functions 1, ρ−q1 , ρ−q1 ρ−q2 , . . ., T = Tλ is of the order of N : The density of free rotons,
ρ−q1 ρ−q2 · · · ρ−qN , each associated with a specific total nr , can be determined in two ways: (1) from the rela-
momentum. tions ρn = Q2r nr /(3kB T ) [30] and ρn = ρ, and (2) from
 4

the formula nr = 2Q2r (µr kB T )1/2 e−∆/kb T ((2π)3/2 ~3 )−1 mentum ~p can be written as
[30]. This gives nr ≈ n/7 and nr ≈ n/6, respectively
(here n = N/V ; the parameters ρ(Tλ ) = 0.1462 g/cm3, Ψp (r1 , . . . , rN ) = d1 (p)ρ−p
∆(Tλ ) ≈ 0.56 meV , µr (Tλ ) ≈ 0.135 m4, and Qr (Tλ ) ≈ +p6=0
q1X
d2 (q1 ; p)
−1
~ · 1.93 Å are taken from [31, 32] and correspond to the + ρq1 ρ−q1 −p + . . .
q1 6=0
2!N 1/2
saturated vapour pressure). Accounting for roton inter-
actions, we obtain nr ≃ (0.1–1)n. q1 +...+qN −1 +p6=0
X dN (q1 , . . . , qN −1 ; p)
So, for a dilute Bose gas, we have the following picture + ×
q1 ,...,qN −1 6=0
N !N (N −1)/2
in the zero-order approximation. The ground state (15)
does not contain quasiparticles; in this case, all atoms ×ρq1 . . . ρqN −1 ρ−q1 −...−qN −1 −p . (21)
remain in the BE condensate of zero-momentum atoms.
The state (16) corresponds to a single Bogoliubov’s quasi- For the ground state, p = 0. If the ground state of a
particle; in this case, N − 1 atoms have zero momentum Bose system is a liquid or a gas, then both the formulae
and one atom has momentum p. The state (17) corre- (9)–(12) and (21) are applicable, but the formulae (9)–
sponds to two quasiparticles with momenta p1 and p2 ; in (12) are preferable for several reasons. However, which
that case, N − 2 atoms have zero momentum, one atom WFs can adequately describe the gaseous or liquid phase
has momentum p1 , and another has momentum p2 . And when the ground state of the system is a crystal? We have
so on. Finally, the state (19) represents N quasiparticles not found an answer in the literature. The WF (21) is
with momenta p1 , . . . , pN ; in this case, N atoms have suitable for this case, while the solutions (11), (12) yield
the same momenta p1 , . . . , pN . Thus, with an increase a non-isotropic dispersion law, because of the anisotropy
in the number of elementary quasiparticles, the BE con- of the crystalline ground-state WF Ψ0 .
densate of zero-momentum atoms gradually fades away. The WFs (9)–(12) and (21) can be represented in
And when the number of quasiparticles becomes close to the P
second quantised form Ψ℘ i, by replacing ρk6=0 with
√1 +
N , this condensate disappears completely. N q âq−k âq [33] in (9)–(12), (21) and adding the fac-

Note that the average number of quasiparticles is given tor [â+ N

0 ] |0bare i at the end of the right-hand side of for-
by the formula mulae (9) and (21) [19].
Note that in addition to the condensate N0 of zero-
1
Z X momentum atoms, the condensate Np6=0 of atoms with
N̄Q (T ) = dr1 . . . drN e−E℘ /kB T Ψ∗℘ N̂Qp Ψ℘ non-zero momentum p can also exist. According to the
Z ℘ treatment in [29], such a condensate is only possible if
1 X −E℘ /kB T (℘) there is a corresponding condensate of elementary quasi-
≡ e NQp , (20)
Z ℘ particles. It is clear that ultrahigh temperature should
also suppress such a condensate of atoms. However, it
appears feasible to create it by means of an external field
(℘)
where NQp is the number of quasiparticles for a state [29], possibly even at ultrahigh temperatures. But we
Ψ℘ . Since states containing N quasiparticles correspond cannot imagine how the condensate N0 could be pro-
(℘) duced at ultrahigh temperatures.
to the relation NQp = N and form the majority of all
(i) It is also interesting to note that although a BE con-
possible states, at T ≫ Tc we should have N̄Q (T ) ≈ N .
densate is typically associated with a large number of
The formulae (15)–(19) do not take into account the particles (N0 , N ≫ 1), it is possible — and reasonable
interatomic interaction. The latter leads to a complica- — to define a condensate even for systems with a small
tion in the form of the WFs in accordance with relations number of particles (N ≥ 2) [34]. This is because the
(9)–(12) and to an additional “blurring” of the conden- Gross-Pitaevskii equation describes few-particle systems
sate (since terms with a larger number of ρ−pi have to more precisely than many-particle systems [34–36]. A
be considered in the expansion of Ψp1 ...pj (r1 , . . . , rN ) in few-particle condensate can be introduced based on the
ρ−pi ). It implies that our conclusions above are valid density matrix (1) using the criterion λ1 ≫ λ2 + . . . + λN
not only for dilute gas but also for dense gas and liq- [34]. However, our analysis above, as well as the very
uid. This reasoning shows that gas states containing N concept of temperature, assumes that N ≫ 1.
elementary quasiparticles are characterized by the prop-
(℘)
erty N0 ∼ 1 ≪ N . This means that, according to the
(i) III. CONCLUSION
formula (7) and the ensuing treatment, at T ≫ Tc
there is no BE condensate of zero-momentum atoms:
N0 ∼ 1 ≪ N . Recapitulating, we have proposed a mathematical rea-
We also remark that WFs (9)–(12) can be written in a soning showing that a BE condensate of zero-momentum
different form. Since for a system of N Bose particles the atoms should be absent in a nonrelativistic system of
functions 1, ρ−q1 , ρ−q1 ρ−q2 , . . ., ρ−q1 ρ−q2 · · · ρ−qN form many spinless interacting bosons at ultrahigh tempera-
(i)
a complete (though non-orthogonal) set of basis functions tures, T ≫ Tc . This reasoning is not rigorous. Al-
[25], any system state (ground or excited) with total mo- though we have considered the canonical ensemble, the
 5

formula (7) can be generalised to the case of a grand tional Academy of Sciences of Ukraine (Project
canonical ensemble, and the conclusions should remain No. 0121U109612). The author thanks the Simons Foun-
the same. dation for additional financial support.
This research was supported in part by the Na-

[1] Novel Superfluids, K. H. Bennemann and J. B. Ketterson [19] M. Tomchenko, arXiv:2501.00893 [cond-mat.quant-gas].
(eds.), Oxford University Press, Oxford (2013), Vol. 1. [20] L. Reatto and G. V. Chester, Phys. Rev. 155, 88 (1967).
https://doi.org/10.1093/acprof:oso/9780199585915.001.0001 https://doi.org/10.1103/PhysRev.155.88
[2] Novel Superfluids, K. H. Bennemann and J. B. Ketterson [21] W. E. Parry and C. R. Rathbone, Proc. Phys.
(eds.), Oxford University Press, Oxford (2014), Vol. 2. Soc. 91, 273 (1967). https://doi.org/10.1088/0370-
https://doi.org/10.1093/acprof:oso/9780198719267.001.0001 1328/91/2/302
[3] A. B. Migdal, E. E. Saperstein, M. A. Troitsky, and [22] L. Reatto and G. L. Masserini, Phys. Rev. B 38, 4516
D. N. Voskresensky, Phys. Rep. 192, 179 (1990). (1988). https://doi.org/10.1103/PhysRevB.38.4516
https://doi.org/10.1016/0370-1573(90)90132-L [23] M. Tomchenko, Fiz. Nizk. Temp. 32, 53
[4] M. Mannarelli, Particles 2, 411 (2019). (2006) [Low Temp. Phys. 32, 38 (2006)].
https://doi.org/10.3390/particles2030025 https://doi.org/10.1063/1.2160507
[5] S. O. Demokritov, V. E. Demidov, O. Dzyapko, [24] A. A. Rovenchak and I. O. Vakarchuk, J. Phys. Stud. 11,
G. A. Melkov, A. A. Serga, B. Hillebrands, 404 (2007). https://doi.org/10.30970/jps.11.404
and A. N. Slavin, Nature 443, 430 (2006). [25] I. A. Vakarchuk and I. R. Yukhnovskii,
https://doi.org/10.1038/nature05117 Theor. Math. Phys. 40, 626 (1979).
[6] L. H. Bennett and E. Della Torre, J. Mod. Phys. 5, 693 https://doi.org/10.1007/BF01019246
(2014). https://doi.org/10.4236/jmp.2014.58081 [26] N. N. Bogoliubov and D. N. Zubarev, Sov. Phys. JETP
[7] A. Kogar, M. S. Rak, S. Vig, et al., Science 358, 1314 1, 83 (1956).
(2017). https://doi.org/10.1126/science.aam6432 [27] M. D. Tomchenko, JETP 102, 137 (2006).
[8] D. Wang, N. Luo, W. Duan, and X. Zou, https://doi.org/10.1134/S106377610601016X
J. Phys. Chem. Lett. 12, 5479 (2021). [28] R. P. Feynman, Phys. Rev. 94, 262 (1954).
https://doi.org/10.1021/acs.jpclett.1c01370 https://doi.org/10.1103/PhysRev.94.262
[9] T. Byrnes, N. Y. Kim, and Y. Yamamoto, Nature Physics [29] M. Tomchenko, J. Phys. A: Math. Theor. 57, 495205
10, 803 (2014). https://doi.org/10.1038/nphys3143 (2024) https://doi.org/10.1088/1751-8121/ad9188
[10] J. Klaers, J. Schmitt, F. Vewinger, and M. Weitz, Nature [30] I. M. Khalatnikov, An Introduction to the Theory of Su-
468, 545 (2010). https://doi.org/10.1038/nature09567 perfluidity, Perseus, New York (2000).
[11] J. B. Ketterson and K. H. Bennemann, Survey [31] B. N. Esel’son, V. N. Grigor’ev, V. G. Ivantsov,
of Some Novel Superfluids, in [1], pp. 74–155. E. Ya. Rudavskii, D. N. Sanikadze, and I. A. Serbin,
https://doi.org/10.1093/acprof:oso/9780199585915.003.0002 Properties of Liquid and Solid Helium, Izd. Standartov,
[12] C. J. Pethick and H. Smith, Bose-Einstein Condensation Moscow (1978) [in Russian].
in Dilute Gases, Cambridge University Press, New York [32] M. R. Gibbs, K. H. Andersen, W. G. Stirling, and
(2008), Chaps. 2, 13, 15. H. Schober, J. Phys.: Condens. Matter 11, 603 (1999).
[13] K. Huang, Statistical Mechanics, Wiley, New York https://doi.org/10.1088/0953-8984/11/3/003
(1963). [33] D. Pines and P. Noziéres, The Theory of Quantum Liq-
[14] A. Einstein, Sitzungsber. Preuss. Akad. Wiss., Phys.- uids: Normal Fermi Liquids, CRC Press, Boca Raton
Math. Kl., p. 3 (1925). (1989). https://doi.org/10.4324/9780429492662
[15] N. N. Bogoliubov, J. Phys. USSR 11, 23 (1947). [34] M. Tomchenko, J. Phys. A: Math. Theor. 57, 495202
[16] I. A. Vakarchuk and I. R. Yukhnovskii, (2024). https://doi.org/10.1088/1751-8121/ad9187
Theor. Math. Phys. 42, 73 (1980). [35] D. Blume and C. H. Greene, Phys. Rev. A 63, 063601
https://doi.org/10.1007/BF01019263 (2001). https://doi.org/10.1103/PhysRevA.63.063601
[17] R. P. Feynman, Statistical Mechanics: A Set of Lectures, [36] D. Blume, Rep. Prog. Phys. 75, 046401 (2012).
Benjamin, Massachusetts (1972). https://doi.org/10.1088/0034-4885/75/4/046401
[18] M. Tomchenko, J. Low Temp. Phys. 201, 463 (2020).
https://doi.org/10.1007/s10909-020-02498-z