Contrasting Effects of Bi and Si Substitution at the Ni Site on Magnetostructural Transitions and Magnetocaloric Properties in Ni–Mn–In Heusler Alloys

Abstract

We investigated the structural, magnetic, magnetocaloric, and magnetotransport properties of Ni₅₀Mn₃₅In₁₅ Heusler alloys via partial substitution of Ni with 3 at.% Bi (Ni₄₇Bi₃Mn₃₅In₁₅) and 3 at.% Si (Ni₄₇Si₃Mn₃₅In₁₅) synthesized by arc melting. X-ray diffraction confirms a predominantly L2₁ cubic structure (space group Fm-3m), while SEM/EDX analysis verifies compositional homogeneity. Temperature-dependent magnetization measurements reveal that the Bi-substituted alloy exhibits a first-order magnetostructural transition associated with the martensitic transformation, followed by a second-order magnetic phase transition from ferromagnetic to paramagnetic behavior near the Curie temperature. In contrast, the Si-substituted alloy shows a single second-order transition with negligible thermal hysteresis, indicating suppression of the martensitic phase. The Curie temperature decreases from 324 K for the parent alloy to 313 K and 286 K for the Bi- and Si-substituted alloys, respectively. A maximum magnetic entropy change of 6.0 Jkg⁻¹K⁻¹ and 4.5 Jkg⁻¹K⁻¹ is observed for the Bi- and Si-substituted alloys, respectively, under an applied magnetic field change of 50 kOe, with corresponding relative cooling power values of 303 Jkg⁻¹ and 345 Jkg⁻¹. These results demonstrate that lattice expansion (Bi) and contraction (Si) distinctly modify Mn–Mn exchange interactions, enabling tunable magnetocaloric performance in Ni–Mn–In Heusler alloys.

1. Introduction

Magnetic refrigeration is an emerging sustainable cooling technology with strong potential to lower greenhouse gas emissions and help address global warming. This approach improves energy efficiency relative to conventional vapor-compression refrigeration systems [1,2,3]. The functionality of magnetic refrigeration is facilitated by the magnetocaloric effect (MCE), which describes the reversible temperature change of a magnetic material when an external magnetic field is applied or removed, causing heating during magnetization and cooling during demagnetization [4,5]. Researchers have identified several materials that are nearly suitable for magnetic cooling systems. While Gd-based materials exhibit sufficient magnetocaloric properties, their high cost and limited availability hinder large-scale practical applications, motivating the search for alternative materials [6,7,8,9]. In this regard, Ni-Mn-In-based Heusler alloys are the emerging viable candidates due to their tunable magnetostructural transitions and significant magnetocaloric response. These intermetallic compounds are composed of combinations of transition elements and one or more main group elements. They can be separated into different classes based on their general formula such as X₂YZ, XYZ, Y₂XZ, where X and Y commonly represent 3d transition metals and Z corresponds to a main group element such as Al, In, Ga, Si, or Sb, termed as full, semi, inverse, and quaternary Heusler alloys, respectively [10]. Off-stoichiometric compositions of these systems have attracted considerable interest due to their remarkable magneto-responsive properties, including magnetic shape memory effects, large magnetoresistance, anomalous Hall effects, and both normal and inverse magnetocaloric effects [11,12,13,14,15,16].

Previous studies have shown that off-stoichiometric Ni-Mn-In-based alloys exhibit structural phase changes from a high-temperature austenite phase (AP) to a low-temperature martensite phase (MP) near room temperature. Additionally, they undergo magnetic phase transitions from an antiferromagnetic (AFM) to a ferromagnetic (FM) phase near the martensitic temperature (T_M), and from FM to paramagnetic (PM) at the Curie temperature (T_C) [17,18,19]. The crystal structure of the alloy is temperature-dependent and can transition between an austenite phase, a low-temperature martensite phase, and mixed phases near the temperature-induced structural (martensitic) transition (T = T_M or T_A), where T_M and T_A are the forward and reverse transition temperatures, respectively [20,21]. In their austenitic phase, full Heusler alloys crystallize in an L2₁ cubic structure, while half Heusler compounds crystallize in a C1_b structure. Substitution at the Ni site in Ni₅₀Mn₃₅In₁₅-based alloys can significantly alter their physical properties. Such a substitution modifies the conduction electron concentration (e/a ratio), interatomic separation of Mn–Mn, and 3d–3d hybridization, thereby tuning the electronic, magnetic, and structural behavior of the system [14,22,23].

A.K. Pathak et al. [24] investigated the magnetocaloric effect (MCE) in Ni₅₀Mn₃₄In₁₆ and reported a magnetic entropy change (|ΔS_M|) of ≈ 6.8 Jkg⁻¹K⁻¹ near the Curie temperature (T_C = 325 K) and 5.3 Jkg⁻¹K⁻¹ around T_M ≈ 280 K under an applied field change of ΔH = 50 kOe. In another study, L. Chen et al. [25] examined Ni₄₅Co_4.75Fe_0.25Mn_36.6In_13.4, which exhibited a significantly higher |ΔS_M| of 25 Jkg⁻¹K⁻¹ at 281 K under ΔH = 5T. Stadler et al. [26] reported an exceptionally large magnetic entropy change of 65 Jkg⁻¹K⁻¹ and a relative cooling power (RCP) of 72 Jkg⁻¹ in Ni₂Mn_0.75Cu_0.25Ga near T_C ≈ 310 K. Furthermore, Bourgault et al. [27] observed a |ΔS_M| of 20 J kg⁻¹ K⁻¹ with a higher RCP of 220 J kg⁻¹ in Ni₄₅Co₅Mn_37.5In_12.5 around T_C ≈ 368 K.

The Ni₅₀Mn_50-xIn_x (x = 15, 15.05, 15.2, 16) alloys are reported to exhibit a gradual increase in |∆S_M| and a significant enhancement in RCP near T_C with increasing concentration. In the case of Ni₅₀Mn₃₅In_15-xBi_x (x = 0, 0.25, 0.5, 1.5), the saturation magnetization and magnetic entropy change (ΔS_M) decreases with increasing x, while the RCP increases [19]. Additionally, substitution of Bi at the Ni site in Ni₄₉BiMn₃₅In₁₅ [17] has been reported to yield enhanced |ΔS_M| and RCP compared with the Ni₅₀Mn₃₅In_15-xBi_x (x ≥ 0.5) alloy [19]. These studies demonstrate that compositional tuning significantly influences magnetocaloric performance; however, a direct comparison of contrasting substitutions (such as Bi and Si) at the Ni site remains limited.

Previous studies have shown that introducing Bi into such alloys decreases the martensitic transition temperature (T_M) while enhancing magnetization, magnetic entropy change (ΔS_M), and relative cooling power (RCP) [17,19]. These effects are highly desirable for magnetic refrigeration applications. Despite these advances, the impact of higher substitution levels on the Ni site remains insufficiently explored. In particular, the Ni₄₇Bi₃Mn₃₅In₁₅ (Ni-Bi) composition is of interest because it involves a relatively higher level of Bi substitution (3 at.%) at the Ni site, representing a compositional regime that remains largely unexplored in the literature. Likewise, the Ni₄₇Si₃Mn₃₅In₁₅ (Ni-Si) alloy enables a systematic comparison, allowing a direct evaluation of the contrasting effects of lattice expansion in the Bi-substituted alloy and lattice contraction in the Si-substituted alloy relative to each other on the structural, magnetic, magnetocaloric, and magnetotransport properties. A portion of this work was previously presented as an abstract at the APS March Meeting [18].

2. Experimental Methods

The pure elements Ni, Bi, Si, In, and Mn with a purity of no less than 99.99% were used to synthesize Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅ alloys with a total mass of about 5g. Excess Mn of about 5% was added to compensate for possible mass loss during arc melting. The elements were weighed accurately to match the desired chemical composition and arc-melted under a flow of ultra-high purity argon gas. The resulting ingots were re-melted three or four times to ensure homogeneity. The ingots were then cut into small pieces for magnetization and resistivity measurements and ground into powders for structural analysis. The sample purities and crystal structures were characterized using powder X-ray diffraction (XRD) with Cu- kα radiation (λ = 1.5406 Å) at room temperature, with a step size of 0.012°/step. The compositional uniformity of the alloys was examined using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX) with a Quanta 450 FEG scanning electron microscope Thermo Fisher Scientific (FEI, Hillsboro, OR, USA) fitted with an Oxford Instruments X-Max 50 EDX detector (Oxford Instruments, Abingdon, Oxfordshire OX13 5QX, UK).

Magnetization and resistivity measurements were conducted using a superconducting quantum interference device (SQUID MPMS, Quantum Design, San Diego, CA, USA). Separate pieces of the sample were used for magnetization and resistivity measurements. The temperature-dependent magnetization was measured over the range of 10–380 K during both heating and cooling cycles under applied magnetic fields of 100 Oe and 50 kOe. Field-dependent isothermal magnetization curves were used to calculate the magnetic entropy change (∆S_M) by applying a Maxwell relation: $|{∆\mathrm{S}}_{\mathrm{M}}|={\int }_{\mathrm{H}1}^{\mathrm{H}2}{\left({\displaystyle \frac{\mathrm{d}\mathrm{M}}{\mathrm{d}\mathrm{T}}}\right)}_{\mathrm{H}}\mathrm{d}\mathrm{H}$, where H₁ and H₂ are initial and final magnetic fields.

The relative cooling power (RCP) was calculated using RCP = |ΔS_M| x ẟT_FWHM, where ẟT_FWHM is the full-width-at-half-maximum of the |ΔS_M | peak. The refrigeration capacity (RC) was determined using $\mathrm{R}\mathrm{C}={\displaystyle {\int }_{{\mathrm{T}}_1}^{{\mathrm{T}}_2}}|{\mathsf{\Delta }\mathrm{S}}_{\mathrm{M}}(\mathrm{T})|\mathrm{d}\mathrm{T}$, where T₁ and T₂ are temperature corresponding to the half-maximum points of the |ΔS_M| peak [12].

3. Results and Discussion

Figure 1a shows the X-ray diffraction (XRD) patterns recorded at room temperature for Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅, confirming that both compounds crystallize in a face-centered cubic (fcc) structure with space group Fm-3m (L2₁ ordered phase) [17,28]. The lattice parameters of both alloys were calculated using the strong well-defined (220) fundamental diffraction peak. A decrease in lattice parameter is observed from 6.015 Å for the Ni-Bi alloy to 5.997 Å for Ni-Si alloy, likely due to the smaller atomic size of Si compared with Bi. Previously, the parent alloy Ni₅₀Mn₃₅In₁₅ has been reported to show a lattice parameter a = 5.884 Å with a mix of cubic and tetragonal phase at room temperature [19], while Ni₄₉BiMn₃₅In₁₅ exhibited a = 5.926 Å [17]. The relative lattice change calculated with respect to the cubic parent phase is ~0.71%(Bi₁), ~2.23%(Bi₃) and ~1.92% (Si₃), indicating significant lattice expansion upon substitution. The observed variation in lattice parameter directly influences the Mn–Mn interatomic distance, which plays a critical role in governing the magnetic exchange interactions in these alloys [29]. The inset of Figure 1a presents the variation of lattice parameter for Ni_50xBi_xMn₃₅In₁₅ (x = 0, 1, 3). For x = 0 and 1, previously reported results from [17,19] are used, demonstrating a linear increase with increasing Bi concentration. This behavior is consistent with lattice expansion due to the larger atomic size of Bi compared with Ni. Additional diffraction peaks are observed in the Bi-containing samples, indicating the presence of secondary phases, possibly due to Bi segregation. The observed Bi segregation suggests that the 3% Bi is close to the maximum amount that can be incorporated at the Ni sites in Ni-Mn-In compounds. This is further supported by the Rietveld refinement of Ni₄₇Bi₃Mn₃₅In₁₅, performed using FullProf software (Version 5.10). A single-phase model based on the Ni-Mn-In crystallographic information file (CIF) was employed; however, it does not account for all observed diffraction peaks, suggesting that the unindexed peaks may be associated with Bi-rich secondary phases. Furthermore, slight shifts and broadening of the main peaks suggest structural instability associated with a martensitic transformation near room temperature, which is further supported by the thermal hysteresis and abrupt change in magnetization observed in the M(T) curves. However, temperature-dependent XRD measurements would provide direct structural confirmation of this behavior [28].

In Figure 2, panel (a) and (b) present the SEM micrographs and corresponding EDX analyses of Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅, respectively, to examine their surface morphologies and elemental compositions. The SEM images reveal relatively uniform and homogeneous microstructures without noticeable phase segregation, while the corresponding EDX spectra confirm the presence of all constituent elements.

Table 1. Elemental compositions of Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅ alloys obtained from EDX analysis presented in both weight (wt %) and atomic% (at %). The atomic percentages are normalized after excluding carbon and oxygen contributions.

Ni47Bi3Mn35In15			Ni47Si3Mn35In15
Elements	Wt.%	At.%	Elements	Wt.%	At.%
Ni	38.9	43.2	Ni	40.9	50.1
Mn	28.2	39.7	Mn	30.7	34.2
In	23.6	14.4	In	27.2	13.2
Bi	9.3	2.7	Si	1.2	2.5

illustrates the experimentally obtained compositions from EDX analysis, reported in both weight percentage (wt.%) and atomic percentage (at.%). The results show reasonable agreement with the nominal compositions within typical experimental uncertainty (~ ± 5%), which is consistent with the typical accuracy of EDX measurements. The reported compositions represent the average values obtained from measurements at multiple regions (4–5 points) across the sample surface, confirming the reliability of the data. Minor deviations are observed, which can be attributed to the surface-sensitive nature of EDX analysis, surface enrichment, and the intentional addition of excess Mn to compensate for evaporation losses during arc melting [30]. Overall, the good match between the measured and nominal compositions, along with the consistency across multiple measurement points, suggests the successful formation of the target alloys and indicates good compositional homogeneity.

Figure 3a,b show the temperature-dependent magnetization M(T) for Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅ measured under an applied field of H = 100 Oe using field-cooled-cooling (FCC) and field-cooled-warming (FCW) protocols. For Ni₄₇Bi₃Mn₃₅In₁₅, a sharp change in magnetization occurs near the martensitic transition temperature (T_M), accompanied by pronounced thermal hysteresis, confirming a first-order phase transition (FOPT) from the ferromagnetic MP to the ferromagnetic AP [24,31]. Furthermore, a second-order transition from the ferromagnetic austenite to the paramagnetic phase is observed near the Curie temperature (T_C). The inset of Figure 3a demonstrates that, as the Bi concentration increases in Ni_50-xBi_xMn₃₅In₁₅ (x = 0, 1, 3), the T_C decreases and shifts closer to room temperature. This behavior is possibly due to the substitution of non-magnetic Bi at the Ni site, which modifies the Mn–Mn interatomic distance and weakens the ferromagnetic exchange interactions, thereby reducing the overall magnetic coupling [32].

In contrast, Ni₄₇Si₃Mn₃₅In₁₅ exhibits a single magnetic transition with nearly overlapping FCC and FCW curves, showing negligible thermal hysteresis (~1.8 K) and a predominantly second-order phase transition, as further confirmed by the corresponding Arrott plots (Figure 4b). This suggests that Si substitution suppresses (or significantly weakens) the martensitic transformation and stabilizes the ferromagnetic austenitic phase. Owing to its smaller atomic size, Si induces relative lattice contraction compared with Ni₄₇Bi₃Mn₃₅In₁₅, which can reduce the Mn–Mn interatomic distance and modifies the exchange interactions. This change can enhance antiferromagnetic contributions or weaken ferromagnetic coupling, thereby influencing the magnetic properties [23,29]. Previous studies have shown that Si substitution reduces the valence electron concentration (e/a ratio) and modifies the balance between ferromagnetic and antiferromagnetic interactions through changes in Mn–Mn spacing, leading to a reduction in both magnetization and Curie temperature [21]. In the present study the calculated e/a values for the Bi- and Si- substituted alloys are approximately 7.75 and 7.72, respectively, which are lower than that of the parent alloy (7.90). Although the difference in e/a between the Bi and Si substituted alloys is relatively small, their contrasting magnetic behavior suggests that the observed changes likely arise from the coupled influence of lattice modification and electronic structure variation rather than from a single parameter alone.

The magnetic behaviors of these Heusler alloys are primarily governed by the Mn atoms, whose large magnetic moments dominate the overall magnetization [23,33]. Due to the relatively large Mn–Mn separation, direct 3d–3d exchange is weak, and the magnetic interactions are mainly mediated by indirect Ruderman–Kittel–Kasuya–Yosida (RKKY)-type exchanges [34]. Figure 3c exhibits the M-T curves measured under an applied field of H = 50 kOe, which indicates that the first-order phase transition is significantly suppressed, as evidenced by the reduced hysteresis and the smoother nature of the transition. This weakening of the FOPT at a higher field suggests the field-induced stabilization of the ferromagnetic austenitic phase, consistent with previous findings [19,35]. The applied magnetic field favors the high-magnetization austenitic phase, thereby suppressing the low-magnetization martensitic phase due to the large difference in magnetization between the two phases [17]. The inset of Figure 3c shows the M(H) curves at 10 K, where a saturation magnetization (M_S) of ~67 emu/g is observed for the Bi-substituted alloy, which is higher compared with the parent alloy [19]. This enhancement suggests that the Bi-containing sample exhibits an improved magnetic response, likely through modifications in Mn–Mn exchange interactions associated with lattice expansion, while the possible contribution of minor secondary phases cannot be excluded.

The temperature dependence of the derivative of magnetization (dM/dT), shown in Figure 3d, provides clear evidence for the coexistence of first-order and second-order magnetic phase transitions in Ni₄₇Bi₃Mn₃₅In₁₅. The dM/dT curves exhibit two anomalies T_M1 and T_M2, with clear thermal hysteresis between cooling and heating, confirming the first-order nature of the martensitic transition [19]. This behavior reflects abrupt changes in magnetization associated with a simultaneous structural transformation in the same temperature range. In contrast, near T_C = 313 K, the dM/dT curves exhibit a sharp minimum without any thermal hysteresis, characteristic of a second-order magnetic phase transition. This is further supported by the Arrott plot shown in Figure 4a. By contrast, the inset of Figure 3d shows that the dM/dT curves for Ni₄₇Si₃Mn₃₅In₁₅ nearly overlap during cooling and heating over the full temperature range, indicating negligible thermal hysteresis and confirming a predominantly second-order transition. Unlike the Ni-Bi alloy, no low-temperature discontinuity is observed, suggesting the absence or strong suppression of the first-order transition in the Ni-Si alloy. In this study we observed that both Bi and Si substitutions lead to an increase in the lattice parameter, accompanied by a reduction in the valence electron concentration (e/a) compared with the parent alloy.

Figure 4a,b show the Arrott plots for Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅, constructed using the mean-field approximation with critical exponents β = 0.5, and γ = 1. The presence of positive slopes in all isotherms confirms that the magnetic phase transition near T_C in both alloys is predominantly of second order, according to Banerjee’s criterion. The isotherm that passes closest to the origin corresponds to the Curie temperature (T_C), providing an independent estimation of the transition temperature [36]. To further examine the nature of the FM-PM transition, a Landau fitting of the magnetic isotherms was performed near T_C using the relation ${\displaystyle \frac{\mathrm{H}}{\mathrm{M}}}=\mathrm{a}+\mathrm{b}{\mathrm{M}}^2+\mathrm{c}{\mathrm{M}}^4$. The positive values of the Landau coefficient b, shown in Figure 4c, further support the predominantly second-order magnetic behavior in both alloys [37]. Negative b values observed at temperatures well above Tc are attributed to fitting uncertainty in the paramagnetic region, where the magnetization becomes very small. From the combined analysis of the dM/dT vs. T curves, and the Arrott plots, the Curie temperatures of Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In₁₅ are determined to be 313 K and 286 K, respectively, in good agreement with the magnetization data.

Figure 5a,b show the isothermal magnetization [M(H)] curves for Ni₄₇Bi₃Mn₃₅In₁₅ and Ni₄₇Si₃Mn₃₅In_15, respectively, measured over selected temperature ranges with intervals of ΔT = 10 K and 20 K under an applied magnetic field of up to 50 kOe. Figure 5c,d present the corresponding magnetic entropy changes (|ΔS_M|) as a function of temperature for both alloys, estimated from the M(H) data using Maxwell’s thermodynamic relation under a magnetic field change of ΔH = 5 T [3]. It is noted that a larger temperature interval may lead to slight underestimation and reduced resolution of the |ΔS_M| peak; however, the overall trend and peak position remain unaffected, ensuring a reliable comparision between the samples. Magnetocaloric properties are primarily evaluated near the FM to PM transition owing to its reversible behavior and negligible hysteresis. Previously, Ni₄₉BiMn₃₅In₁₅ has been reported to exhibit a maximum magnetic entropy change of |ΔS_M| = 5.5 Jkg⁻¹K⁻¹ and a relative cooling power of 312 Jkg⁻¹ at T_C = 325 K. In the present study, increasing the Bi content leads to an enhanced magnetocaloric response, with |ΔS_M| = 6 Jkg⁻¹K⁻¹ and RCP and refrigerant capacity (RC) values of 303 Jkg⁻¹ and 257 Jkg⁻¹, respectively. In contrast, substitution with 3 at.% Si reduces the magnetic entropy change to |ΔS_M| = 4.2 Jkg⁻¹K⁻¹, while the RCP increases to 345 Jkg⁻¹ (RC = 300 Jkg⁻¹) at T_C = 286 K. The enhanced magnetocaloric response in the Ni–Bi alloy is attributed to lattice expansion, which increases the Mn–Mn interatomic distance and enhances the magnetization change across the transition [19]. In contrast, Si substitution induces lattice contraction compared with Ni₄₇Bi₃Mn₃₅In₁₅, reducing the Mn–Mn distance and modifying the balance between ferromagnetic (FM) and antiferromagnetic (AFM) interactions, thereby lowering |ΔS_M|. However, the relatively higher RCP (and RC) values observed for the Si-substituted alloy arise from the broader |ΔS_M|-T peak, which extends the effective temperature span of the magnetocaloric response [32]. These parameters represent key figures of merit for magnetic refrigeration, reflecting the heat transfer capability between the hot and cold reservoirs during a magnetic refrigeration cycle, and are widely used to evaluate material performance [21]. Furthermore, previous studies have shown that doping Ni-Mn-In alloys with magnetic elements such as Cu, Co, Fe, and Mn enhances |ΔS_M| due to the strengthening of magnetic exchange interactions, particularly Mn-Mn and Ni-Mn coupling [31].

Figure 6a exhibits the comparison of magnetic entropy changes (|ΔS_M|) for both alloys under different magnetic field changes (ΔH = 1–5 T). Although, the magnetic entropy change (|∆S_M|) is lower than that reported for some Ni–Mn-based alloys, the present system exhibits a broader magnetic transition, resulting in a competitive relative cooling power (RCP), thereby enhancing their suitability for practical magnetic refrigeration [3]. It is also noteworthy that several compositions listed in Table 2 exhibit lower |∆S_M| values than the present systems, thereby highlighting the competitive overall magnetocaloric performance observed in this study.

Figure 6b shows the temperature-dependent resistivity (ρ) of Ni₄₇Bi₃Mn₃₅In₁₅ at a zero field, along with the magnetic susceptibility (χ) measured under an applied field of 100 Oe during cooling. The resistivity was calculated from four-probe resistance measurements using $\mathsf{\rho }=\left(\mathrm{R}\mathrm{*}\mathrm{A}\right)/\mathrm{L}$, where A is the cross-sectional area and L is the distance between voltage contacts. Two distinct transitions are observed: (i) a sharp drop in resistivity near T_M, coinciding with the temperature of the first-order martensitic transformation, and (ii) a gradual change in slope near T_C, corresponding to a second-order ferromagnetic-paramagnetic transition, associated with spin disorder scattering [17]. These features are consistent with the corresponding χ(T) data and are in good agreement with previous reports on Ni-Mn-based Heusler alloys [38,39]. From the M-T measurements, the Si-substituted alloy was found to undergo only a second-order magnetic transition without a martensitic transformation. Consequently, the expected variation in ρ(T) is relatively smooth; therefore, resistivity measurements were not included in the present study.

Table 2. Comparison of magnetic entropy changes |∆S_M| and relative cooling powers (RCPs) for representative Heusler alloys near the second-order phase transition (SOPT) with changing magnetic field of ΔH = 5 T. ‘-’ indicates the data not reported.

S.N.	Systems	|∆SM| [JKg−1K−1]	RCP [JKg−1]	TC [K]	References
1.	Ni47Bi3Mn35In15	6	303	313	This work
2.	Ni47Si3Mn35In15	4.2	345	286	This work
3.	(Ni37.5Co12.5Mn35Ti15)99.9B0.4	12.2	183	305	[39]
4.	Ni50Mn34.80In15.2	7	224	328	[24]
5.	Ni49Mn35In15Bi	5.5	312	325	[17]
6.	Ni50Mn35In14.50Bi0.5	5	165	320	[19]
7.	Ni45Co5Mn37.5In12.5	20	-	368	[31]
8.	Ni2Mn0.75Cu0.25Ga	65	-	308	[26]
9.	Ni52Mn26Ga22	11.4	-	348	[40]
10.	Ni50Mn35In15	5.4	-	324	[19]
11.	Ni44Mn36In14Co6	23.5	250	368	[41]
12.	Ni47Cu2Mn38Sn13	3.54	-	316	[42]
13.	Ni2Mn1.34In0.66	4.5	201	305	[34]
14.	Ni42Mn46CoSn11	0.97	129	350	[43]
15.	Ni50Mn35In10Fe5	1.4	74	298	[44]
16.	Ni50Mn32Sn10Fe8	10.04	184	318	[45]
17.	MnNi0.85Fe0.15Ge	4.88	36.33	274	[32]
18.	MnCoGeSi0.04	26.7	-	283	[21]
19.	Ni50Mn34In15.5Al0.5	3.86	9.9	321	[22]
20.	Ni36Co14Mn35Ti15	~20	~285	305	[46]

Table 2. Comparison of magnetic entropy changes |∆S_M| and relative cooling powers (RCPs) for representative Heusler alloys near the second-order phase transition (SOPT) with changing magnetic field of ΔH = 5 T. ‘-’ indicates the data not reported.

S.N.	Systems	|∆SM| [JKg−1K−1]	RCP [JKg−1]	TC [K]	References
1.	Ni47Bi3Mn35In15	6	303	313	This work
2.	Ni47Si3Mn35In15	4.2	345	286	This work
3.	(Ni37.5Co12.5Mn35Ti15)99.9B0.4	12.2	183	305	[39]
4.	Ni50Mn34.80In15.2	7	224	328	[24]
5.	Ni49Mn35In15Bi	5.5	312	325	[17]
6.	Ni50Mn35In14.50Bi0.5	5	165	320	[19]
7.	Ni45Co5Mn37.5In12.5	20	-	368	[31]
8.	Ni2Mn0.75Cu0.25Ga	65	-	308	[26]
9.	Ni52Mn26Ga22	11.4	-	348	[40]
10.	Ni50Mn35In15	5.4	-	324	[19]
11.	Ni44Mn36In14Co6	23.5	250	368	[41]
12.	Ni47Cu2Mn38Sn13	3.54	-	316	[42]
13.	Ni2Mn1.34In0.66	4.5	201	305	[34]
14.	Ni42Mn46CoSn11	0.97	129	350	[43]
15.	Ni50Mn35In10Fe5	1.4	74	298	[44]
16.	Ni50Mn32Sn10Fe8	10.04	184	318	[45]
17.	MnNi0.85Fe0.15Ge	4.88	36.33	274	[32]
18.	MnCoGeSi0.04	26.7	-	283	[21]
19.	Ni50Mn34In15.5Al0.5	3.86	9.9	321	[22]
20.	Ni36Co14Mn35Ti15	~20	~285	305	[46]

4. Conclusions

In summary, this work investigates the structural, magnetic, magnetocaloric, and magnetotransport properties of Ni–Mn–In-based alloys with the partial substitution of Ni by Bi and Si (each at 3 at.%). A clear contrasting behavior was observed between Bi and Si substitutions. Bi substitution enhances magnetocaloric performance, with the maximum magnetic entropy change (|ΔS_M|) increasing from 5.5 (the parent) to 6 J kg⁻¹ K⁻¹ as the Bi content increases from 1 to 3 at.%. In contrast, Si substitution led to a reduction in |ΔS_M|, while the broadening of the ΔS_M peak resulted in an enhanced RCP. Furthermore, the Curie temperatures of both alloys (286 K and 313 K) are close to room temperature (~300 K), making them suitable for practical magnetocaloric applications. These results demonstrate that lattice expansion (Bi) and contraction (Si), relative to each other, distinctly modify the Mn–Mn exchange interactions and affect the magnetocaloric responses. These results highlight the contrasting roles of Bi and Si substitution in tuning the magnetocaloric response of Ni–Mn–In-based Heusler alloys.