nanomaterials Article Modification of the Raman Spectra in Graphene-Based Nanofluids and Its Correlation with Thermal Properties María del Rocío Rodríguez-Laguna 1,2, * , Pedro Gómez-Romero 1, *, Clivia M. Sotomayor Torres 1,3 and Emigdio Chavez-Angel 1, * 1 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and The Barcelona Institute of Science and Technology (BIST), Campus UAB, Bellaterra, 08193 Barcelona, Spain; clivia.sotomayor@icn2.cat 2 Department of Chemistry, Universitat Autònoma de Barcelona, Campus UAB, Bellaterra, 08193 Barcelona, Spain 3 ICREA- Institució Catalana de Recerca i Estudis Avançats, Pg. Lluís Companys 23, 08010 Barcelona, Spain * Correspondence: rodriguez3laguna@gmail.com (M.d.R.R.-L.); pedro.gomez@icn2.cat (P.G.-R.); emigdio.chavez@icn2.cat (E.C.-A.); Tel.: +34-93-7373616 (M.d.R.R.-L.); +34-93-7373608 (P.G.-R.); +34-93-7371617 (E.C.-A.)  Received: 13 May 2019; Accepted: 24 May 2019; Published: 26 May 2019  Abstract: It is well known that by dispersing nanoparticles in a fluid, the thermal conductivity of the resulting nanofluid tends to increase with the concentration of nanoparticles. However, it is not clear what the mechanism behind this phenomenon is. Raman spectroscopy is a characterization technique connecting the molecular and macroscopic world, and therefore, it can unravel the puzzling effect exerted by the nanomaterial on the fluid. In this work, we report on a comparative study on the thermal conductivity, vibrational spectra and viscosity of graphene nanofluids based on three different amides: N, N-dimethylacetamide (DMAc); N, N-dimethylformamide (DMF); and N-methyl-2-pyrrolidinone (NMP). A set of concentrations of highly stable surfactant-free graphene nanofluids developed in-house was prepared and characterized. A correlation between the modification of the vibrational spectra of the fluids and an increase in their thermal conductivity in the presence of graphene was confirmed. Furthermore, an explanation of the non-modification of the thermal conductivity in graphene-NMP nanofluids is given based on its structure and a peculiar arrangement of the fluid. Keywords: raman of nanofluids; enhancement of thermal conductivity; nanofluids; graphene 1. Introduction The continuous trend towards miniaturization in the electronics industry has led to smaller devices with more processing power. However, as dimension decreases, the effective thermal conductivity (k) of each component also decreases and, consequently, energy losses increase [1]. In the case of the emerging charge-based nanoelectronic switches, the density of these devices is limited by a maximum allowable power dissipation of ~100 W/cm2 , and not by their size [2]. Therefore, it is a matter of high priority to design effective and novel cooling strategies, at a reasonable price to meet the cooling needs of the electronics industry. To guarantee proper performance and reliability of electronic components, excess heat must be carried away from critical components and hot spots, and then dissipated to the environment. Conduction and convection are the primary modes of heat transfer in cooling strategies of electronics, where radiation expels only a small part of the total thermal load. Conventional cooling strategies can be divided into active and passive. Passive cooling does not consume energy and is mainly based on natural convection, radiation, and heat conduction by utilizing heat spreader or a heat sink with very high thermal conductivity. Nanomaterials 2019, 9, 804; doi:10.3390/nano9050804 www.mdpi.com/journal/nanomaterials Nanomaterials 2019, 9, 804 2 of 12 Metal materials, such as aluminum, copper, zinc, are the most used as heat sinks [3]. Passive cooling technology relies on maximizing heat transfer by conduction, radiation, and convection using heat sinks, heat spreaders, heat pipes, or thermal interface materials to maintain optimal operating temperatures. These components can be found in all electronic devices and only require an architectural design of heat sinks and choice of the right material [4]. Active cooling, on the other hand, requires an external device (e.g., a fan) to force the cooling of the system. However, a major drawback of this technology is related to the consumption of extra energy to keep the device in operation although the rate of heat removal is much larger than that provided by passive cooling methods (except strategies using latent heat) [5]. Thus, active cooling can deal with higher power densities, and current cooling strategies include: Forced-air and forced-liquid cooling, direct liquid immersion, thermoelectric coolers, and forced convection [5–7]. The vast majority of active cooling systems include a heat sink or heat spreader in direct contact with the heat source. Excess heat is transported away from the electronic component by (1) conduction towards the cold edges of the solid piece or to a second piece (heat sink, pipe or cold plate) and then, by (2) forced convection in air or liquid. It is clear that the material, together with the structural and architectural design of heat sink, is crucial to increase efficiency levels of active cooling systems. Great efforts have been made to improve the performance of current heat sinks, such as the modification of the architectural designs [8–11], development of composite materials with enhanced thermal properties (metal-SiC, metal-graphene/diamond) [3,12–14], and study and control of the structure to enhance their thermal conductivity (e.g., porosity) [14]. Therefore, to improve the efficiency of active cooling systems, there must be an optimization of solid materials acting as heat sinks/spreaders/pipes and active components, such as fluids and those related to mechanical forces (fan, pump, etc.). Among these active methods, the air-based (heat-sink-and-fan) can be found in many electronic devices from the cheapest laptop to the most expensive laboratory equipment. However, this is an old technology and not always efficient. Furthermore, it is not cost-effective for the dissipation of large amounts of heat. Murshed and Nieto de Castro [5] demonstrated that forced convection of air has very low heat removal rates compared with forced convection of liquids and that natural convection has the lowest heat removal rate among all cooling modes. In this context, liquid-cooling technology has proven to be an efficient and low-cost method to cool high-power components [5,15,16], despite the potential risk of having liquid close to an electronic device. This becomes evident in the designs of high-density boards, where air cooling can become so challenging that the only viable option is the use of liquid cold plate technologies. [17] In any case, it is important to remark that the implementation of an effective cooling strategy in both high-power and low-power components may require the use of a combination of liquid- and air-cooling technologies, as well as a proper design of the passive components (heat sinks and heat spreaders). Within liquid-based active cooling, the most common is the use of heat transfer fluids (or coolants) confined in basins. The coolant flows from tower pumps through pipes along the hot side of the device, thereby reducing the temperature. Conventional coolants are mainly classified into dielectric and non-dielectric fluids. Dielectric coolants are organic-based fluids with quite low thermal conductivity but widely used in direct liquid immersion systems, while, the non-dielectric coolants are preferred for heat pipes and microfluidics due to their superior thermal properties compared to their dielectric counterparts. They are normally aqueous solutions and thus exhibit quite a high heat capacity and low viscosity. Water, ethylene glycol, and mixtures of these two are the most popular and widely used as coolants for many electronic devices [5]. However, a major problem of conventional coolants is the low heat exchange rate and thermal conductivity, which is too small to meet the coming needs and challenges in the field. One way to overcome this barrier is by using solid particles dispersed in fluids to improve their thermal properties [18–20]. Nanomaterials 2019, 9, 804 3 of 12 Nanofluids (NFs) are colloidal dispersions of nanoparticles (1–100 nm) suspended in a base liquid [21]. The presence of well-dispersed nanoparticles in a fluid can provide remarkable improvements in thermal properties, such as thermal conductivity [18,22–27] and in some cases, specific heat capacity [27–34]. The heat transfer performance of a number of nanofluids has been tested compared to that of their base fluids in commercial electronic cooling units for CPUs and chips, as well as in heat pipes (passive cooling). The results are promising, showing an effective decrease in the temperature of the electronic components (CPUs and chips) [35–39] as well as an enhancement in the heat transfer performance of heat pipes [40–44]. Several studies have been reported on the enhancement of thermophysical properties in nanofluids, however, the mechanism behind this phenomenon is not clear [45–49]. In the literature, at least six different mechanisms have been suggested for explaining this enhancement: (i) Increase of the thermal transfer due to Brownian motion of nanoparticles, (ii) localized convection created in the fluid due to Brownian motion of nanoparticles, (iii) agglomeration of nanoparticles, (iv) enhanced thermal energy transfer due to increased interatomic interactions arising from the modification of interatomic potential, (v) ordered layering of liquid around the solid, and (vi) ballistic phonon transport of heat through solid nanoparticles [45–49]. In a previous report [27], we showed that for these kinds of fluids (amides) the mechanisms: (i), (ii) and (iii) cannot explain our experimental observations. In addition, the displacement of some Raman modes of DMF-graphene NFs led us to conclude that there is a strong modification of the interatomic potential as a function of graphene concentration, as the model (iv) suggests. Taking into account all of the above, in this paper, we report an extension of our previous work related to the enhancement of thermophysical properties of graphene-based nanofluids [27]. Here, we present a thorough study of the vibrational properties of two previously reported NFs based on: N, N-dimethylacetamide (DMAc) and N, N-dimethylformamide (DMF) and we extend the work to a NF based on N-methyl-2-pyrrolidinone (NMP). The thermophysical properties of each NF are discussed based on the Raman observations. 2. Materials and Methods Graphene flakes were prepared from graphite (Sigma-Aldrich, purity > 99% and size < 20 um, St. Louis, MO, USA) by mechanical exfoliation. An extended description of the sample preparation can be found in the section on experimental methods in the supporting information of Reference [27]. The produced graphene size was around 150–450 nm and the number or layers varied from 2–10. N, N-dimethylacetamide (ACROS Organics, 99+%, Fair Lawn, NJ, USA), N, N-dimethylformamide (Scharlau, HPLC grade, Barcelona, Spain) and N-methyl-2-pyrrolidinone (Sigma Aldrich, 99.5%, St. Louis, MO, USA) were used as the base fluids. The preparation of the nanofluids consisted of direct mixing of the base fluid with graphene nanosheets. Then, graphene was dispersed in the fluid using a planetary ball mill (All-direction planetary ball mill 0.4 L, model CIT-XBM4X-V 0.4 L, Columbia International, Irmo, SC, USA) and soft ultrasonic vibration (Ovan, model ATM40-6LCD, Barcelona, Spain) for 1 hour. These techniques were used to suppress the formation of particle clusters in order to obtain stable dispersions [50]. The samples were later centrifuged for one hour at 6000 rpm to ensure the stability of the NFs. The stability of the dispersions was evaluated by periodic analyses using dynamic light scattering (DLS) as is described in [27]. The measurements of thermal conductivity of the nanofluids were carried out by using the well-known three-omega (3ω) method [51,52] in the bidirectional configuration [53,54]. An extended description of the setup can be found in the section on experimental methods in the supporting information of References [27,55]. The effective viscosity of the nanofluids was measured using a Haake RheoStress RS600 rheometer from Thermo Electron Corp. at T = 20–21 ◦ C. The shear rate used was 2880 s−1 with a measurement time of 30 seconds. The Raman spectra were recorded by T64000 Raman spectrometer manufactured by HORIBA Jobin Yvon (Chilly-Mazarin, France). It was used in single grating mode (2400 lines) with a spectral resolution better than 0.4 cm−1 . The NF was placed Nanomaterials 2019, 9, 804 4 of 12 in a transparent quartz cuvette. Then a green diode laser (λ = 532 nm) was focused on the cuvette by using a 50× long working distance microscope objective. The power of the laser was kept as low as possible (<2 mW) to avoid any possible effect from self-heating. Raman analyses of the samples were repeated under similar conditions as a function of time (day 1, after ~one month and more than one year later) to verify the reproducibility of the Raman spectra (See Figures S1–S3 in the supporting information for more details). 3. Results and Discussions 3.1. Thermal Conductivity and Viscosity Measurements The Table 1 shows the thermal conductivity for different graphene concentrations (C) in nanofluids based on N, N-dimethylformamide (DMF), N, N-dimethylacetamide (DMAc) and N-methyl-2-pyrrolidinone (NMP). Progressive enhancement of k was observed in DMAc and DMF-based nanofluids (NFs) as C increased, with a maximum enhancement of ~48% for C = 1.50 mg/mL in DMAc-based NF. Concentrations between 0.1–0.5 mg/mL in DMF and DMAc produced relatively large enhancements of ~6–25% and 3–17%, respectively. While in the same set of concentrations no significant enhancement of k was detected in NMP-based NFs. In contrast, in the case of viscosity, a maximum enhancement of ~44% was observed for 0.50 mg/mL graphene in NMP. Table 1. Thermal conductivity and viscosity of graphene based nanofluids. Samples Concentration DMAc * DMF * NMP (This Work) k (W m−1 Viscosity K (W m−1 Viscosity k (W m−1 Viscosity mg/mL wt% K−1 ) (mPa·s) K−1 ) (mPa·s) K−1 ) (mPa·s) 0.00 0 0.175 1.19 0.183 0.94 0.235 2.07 0.05 0.005 - - - - 0.234 2.19 0.10 0.01 0.180 1.17 0.194 0.99 0.236 2.21 0.25 0.03 0.196 1.18 0.203 1.01 - - 0.50 0.05 0.206 1.26 0.228 1.08 0.213 2.92 1.13 0.12 - - - 1.26 - - 1.50 0.18 0.259 1.68 - - - - * Taken from Rodríguez-Laguna et al. [27]. An increase in viscosity as the nanoparticle concentrations rise has been frequently observed in graphene-based [23,26,56–59] and other NFs [23,26]. This enhancement has been commonly associated with the tendency of the nanoparticles to agglomerate under pressure. The agglomerates raise the internal shear stress and the resistance to flow in the NF, which leads to an increase in viscosity during measurement [58]. 3.2. Raman Measurements Firstly, we will discuss the Raman characterization of DMAc and DMF as a function of graphene concentration ranging from 0 < C < 1.5 and 0 < C < 1.13 mg/mL, respectively. Due to the very high light absorption of the most concentrated dispersions (C = 1.5 and 1.13 mg/mL for DMAc and DMF, respectively), we were only able to detect one peak in the entire Raman spectrum. The vibrational assignment for DMAc and DMF was done based on the work of Chalapathi and Ramiah [60]. Secondly, we will show and discuss the Raman spectra of NMP-NFs for three different graphene concentrations of C = 0, 0.05 and 0.10 mg/mL. The vibrational assignment was carried out based on the work of Peek and McDermott [61] and Xu et al. [62]. The peak positions of all the NFs were estimated using a Lorentzian fit considering a constant background. Raman analyses of the samples were repeated under similar conditions as a function of time: Day 1, after ~one month and one year later. We found excellent reproducibility of the Raman spectra of DMAc, DMF, and NMP-nanofluids as can be seen in Nanomaterials 2019, 9, 804 5 of 12 Figures S1–S3 that are shown in the supporting information. These results confirm the high stability of these dispersions over time and are in agreement with previously performed DLS analyses [27]. Nanomaterials 2019, 9, x FOR PEER REVIEW 5 of 12 3.2.1. Raman of DMAc- and DMF-Based Nanofluids Graphene-DMAc Nanofluids Graphene-DMAc Nanofluids The Raman spectra of DMAc for different graphene concentrations are displayed in Figure 1a,c. A clear blue-shift The Raman andofaDMAc spectra broadening of thegraphene for different Raman concentrations bands can be are observed displayedas in the graphene Figure 1a,c. Aconcentration increases. clear blue-shift The largest and a broadening of displacement wascan the Raman bands found to be ~4 cm be observed forgraphene as–1the a band around ~470cm– concentration 1 (see Figure 1b). This band corresponds to a scissoring−1 −13–N–CH increases. The largest displacement was found to be ~4 cm vibration from for a band the bond around (CH ~470 cm 3) of1b). (see Figure the DMAc This bandmolecule [60].toThe corresponds displacement a scissoring of from vibration this and the other the bond bands 3to (CH3 –N–CH ) ofhigher frequencies the DMAc moleculein[60]. the presence The of graphene displacement of this is anthe and indicator of a to other bands strong highermodification frequencies inof the interatomic presence of forces within graphene the is an nanofluid. indicator of a strong modification of interatomic forces within the nanofluid. 350 450 550 650 0.0 0.1 0.3 0.5 0.7 589.4 δ (O=C-N) scissoring Peak position [cm ] -1 588.7 [mg/mL] 0.00 588.0 Intensity [a.u.] 474.6 δ (CH3-N-CH3) scissoring 0.10 472.5 0.25 470.4 0.50 420.0 r (C-CH3) rocking 419.2 0.70 418.4 350 450 550 650 0.0 0.1 0.3 0.5 0.7 -1 Raman shift [cm ] (a) Graphene concentration [mg/ml] (b) 720 735 750 960 990 1020 0.0 0.2 0.4 0.6 1.48 1.50 Fit 1012.8 r (C-CH3) rocking [mg/mL] 1011.6 Peak position [cm ] [mg/mL] -1 0.00 0.00 1010.4 Intensity [a.u.] 0.10 ν (C-CH3) assymmetric streching 0.25 0.10 957.0 0.50 0.25 956.7 0.70 0.50 736.8 νs (N-CH3) symmetric streching 1.50 0.70 736.2 735.6 720 735 750 960 990 1020 0.0 0.2 0.4 0.6 1.48 1.50 -1 Raman shift [cm ] (c) Graphene concentration [mg/ml] (d) Figure 1. Raman scattering of DMAc-based NFs and peak positions of several Raman bands: (a,c) Raman Figureof spectra 1. DMAc-based Raman scattering of DMAc-based NFs for NFs and different graphene peak positions concentrations, of several (b,d) Raman bands: peak positions (a) and of six Raman (c) Raman spectra of DMAc-based NFs for different graphene bands shown in (a,c) as a function of graphene concentration. concentrations, (b) and (d) peak positions of six Raman bands shown in (a) and (c) as a function of graphene concentration. Graphene-DMF Nanofluids A similar effect can also be observed in DMF-based nanofluids as displayed in Figure 2b,d Here, Graphene-DMF Nanofluids the largest displacement ~4 cm−1 occurred for a rocking vibration of the bond (CH3 )–N of the DMF A similar molecule. effect canstudy, In a previous also be weobserved in DMF-based related such nanofluids observation as displayed to an apparent hardeningin Figure of the2b,d Here, bonds of the largest DMF displacement molecules produced by ~4 the cm–1presence occurredof for theagraphene rocking vibration flakes. [27]ofInthe bond this (CH study, 3)–N of the simulations DMF based onmolecule. In a previous density functional study, theory we and (DFT) related such observation molecular dynamics to an apparent (MD) suggestedhardening of the bonds that the presence of theof DMF molecules graphene produced flake acts by the presence as an external force fieldofthat the induces graphene flakes. [27] a parallel In this study, orientation of thesimulations based DMF molecules inonthe density functional proximity of the theory graphene(DFT) and molecular dynamics (MD) suggested that the presence of the flakes. graphene flake acts as an external force field that induces a parallel orientation of the DMF molecules in the proximity of the graphene flakes. Nanomaterials 2019, 9, x FOR PEER REVIEW 6 of 12 Nanomaterials 2019, 9, 804 6 of 12 1070 1080 1090 1100 1110 1350 1400 1450 1500 0.0 0.2 0.4 1.1 1.2 [mg/ml] (a) 1439.8 ν (C-N) stretching (b) 1.13 1439.4 Peak position [cm ] -1 1439.0 Intensity [a.u.] 0.50 [mg/ml] 1406.6 δz(CH3)N scissoring (c) 0.50 0.25 1406.1 0.25 1405.6 0.10 0.10 1094 r(CH3)N rocking 0.00 1092 0.00 (d) 1090 1070 1080 1090 1100 1110 1350 1400 1450 1500 0.0 0.2 0.4 1.1 1.2 -1 Raman shift [cm ] Concentration [mg/ml] Figure 2. Raman scattering of DMF-based NFs and peak positions of several Raman bands: (a) Raman Figure 2. spectra ofRaman scattering DMF-based of DMF-based NFs for NFs and different graphene peak positions(b–d) concentrations, of several Raman bands: peak positions (a) Raman of three Raman spectrashown bands of DMF-based NFs for of in (a) as a function different graphene graphene concentrations, (b)–(d) peak positions of three concentration. Raman bands shown in (a) as a function of graphene concentration. Taking into account that both graphene and DMF molecules contain a p-orbital with unshared Taking electron pairsinto account that (delocalized both graphene electrons) and DMF molecules along the out-of-plane direction contain [63,64]. aThe p-orbital parallelwith unshared orientation of electron the DMF pairs (delocalized molecules electrons) with respect to thealong the out-of-plane graphene direction flakes can favor [63,64]. the overlap ofThe twoparallel adjacentorientation p-orbitals. ofcan It thefavor π-π bond between DMFamolecules with respect to the graphene the solvent moleculesflakes and thecan favor the graphene overlap surface. of interaction This two adjacent p- thus orbitals. Itthe increases can favor of rigidity π-π internal a the bond betweenbonds of the thesolvent molecules molecules closestand the graphene to graphene surface. This and, consequently, interaction modifies thusRaman their increases the rigidity spectra. of the internal It is important bonds that to mention of theour molecules closest simulations to graphene suggested that and, this consequently, interaction modifies is local theirglobal, rather than Raman i.e.,spectra. It is important just the molecules aroundto mention flakes of graphene that ourfeel simulations this change. suggested It thathow is not clear thisthe interaction rest of theis local fluid rather is alsothan global, affected by i.e., this just the molecules interaction. around Returning of graphene to DMAc NFs, aflakes feelblue-shift similar this change. It is not in several Raman clearmodes how the canrest of the fluidIf is be observed. wealso affected consider thatbyDMF this and interaction. DMAc Returning are molecules to DMAc amides NFs, with a similar blue-shift quite similar in several structures andRaman electron modes can be observed. delocalization, then itIfiswe consider reasonable that to DMF and suggest thatDMAc molecules a similar geometric are amides with quite configuration similar occurs structures and in DMAc-based electron NFs. As adelocalization, consequence, then it molecules DMAc is reasonable may toalso suggest tendthat to laya similar parallelgeometric configuration to the graphene π-π stacking between occurs ina DMAc-based flakes, favoring NFs. As a consequence, the graphene and DMAc moleculesthereby the molecules may also tend to laytoparallel contributing to the graphene the blue-shift in the Ramanflakes, favoring a π-π spectra. stacking between the graphene and the molecules thereby contributing to the blue-shift in the Raman 3.2.2. Raman in NMP-Based Nanofluids spectra. The case of NMP-based fluids is quite different, Figure 3 shows the full Raman spectra of 3.2.2. Raman NFs NMP-based in NMP-Based Nanofluids for three different graphene concentration of C = 0, 0.05 and 0.10 mg/mL. At first glance, Thea very case ofsmall displacement NMP-based fluidsofiscertain Raman modes quite different, Figurecan be detected 3 shows the full(see RamanFigure 4). However, spectra of NMP- the displacement is both within the range of the spectral reproducibility of based NFs for three different graphene concentration of C = 0, 0.05 and 0.10 mg/mL. At first glance, the equipment (~0.25 cm−1a) and verythe errordisplacement small determinationofof certaineach peak position. Raman modes In conclusion, can be detected there is(see not aFigure significant displacement 4). However, the of the Raman bands displacement is bothaswithin the graphene the range concentration of the spectral increases. reproducibility of the equipment (~0.25 cm–1) and The the Raman spectroscopy results error determination of eachfrom graphene-NMP peak position. InNFs are not surprising conclusion, there is ifnot we aconsider the significant structure of theoforganic displacement the Raman bands asN-methyl-2-pyrrolidone compound. the graphene concentration is a five-membered increases. cyclic amide (with a ring structure), The Ramanand therefore, the spectroscopy bonds results fromtend to be more rigid graphene-NMP NFsthan those are not of amides surprising having if we a linear consider the structure. Consequently, it is conceivable that small amounts of graphene structure of the organic compound. N-methyl-2-pyrrolidone is a five-membered cyclic amide (with a do not affect or modify the intramolecular ring structure), vibrations and therefore, of inner bondstend the bonds in theto solvent. be more Furthermore, rigid than those Adams et al. having of amides [65] suggested a linear that NMP molecules tend to spontaneously form dimers in the liquid state. structure. Consequently, it is conceivable that small amounts of graphene do not affect or modify the NMP molecules strongly interact in pairs.vibrations intramolecular Therefore, of the innerimpact bondsofinthe thepresence solvent. of graphene flakes Furthermore, Adams on et theal.vibration of the [65] suggested bonds that NMPshould be negligible. molecules tend to The findings of form spontaneously Adam et al. in dimers reinforce the liquidthestate. idea that NMP the solvent creates molecules stronglya natural barrier that prevents graphene from modifying the intramolecular interactions interact in pairs. Therefore, the impact of the presence of graphene flakes on the vibration of the bonds of the solvent. Moreover, Basma et al. should be negligible. [66] The reported findings of that Adam pure NMP et al. has a well-developed reinforce the idea that theintrinsic solvent order createscompared a natural to other that solvents. This graphene inner order, together with the 2 hybridization of N-C-O (which allows a barrier prevents from modifying the spintramolecular interactions of the solvent. possible Moreover, π-πBasma interaction et al. between [66] reportedNMPthatand pure graphene), NMP has can a explain the long-term well-developed dispersion intrinsic orderstability compared of graphene in NMP. However, to other solvents. This innerthe internal order, structure together with of the NMP spseems to be an obstacle 2 hybridization of N-C-O to the improvement (which allows a of possible π-π interaction the thermal propertiesbetween of the fluidNMP induced, in other cases, and graphene), by graphene. can explain the long-term dispersion stability of graphene in NMP. However, the internal structure of NMP seems to be an obstacle to the improvement of the thermal properties of the fluid induced, in other cases, by graphene. Nanomaterials 2019, 9, x FOR PEER REVIEW 7 of 12 Nanomaterials 2019,9,9,804 Nanomaterials2019, x FOR PEER REVIEW 7 of of 12 12 744 1426 1672 925 1435 616 744 1426 1672 [a.u.][a.u.] 925 1435 1457 1224 616 848 1297 1400 1023 Intensity 558 653 982 1457 893 [mg/ml] 1224 848 1297 1400 0.00 1023 Intensity 558 653 982 0.05 893 [mg/ml] 0.00 0.10 0.05 500 0.10 600 700 800 900 1000 1250 1350 1450 1620 1670 1720 -1 500 600 700 800 900 1000 Raman shift [cm ] 1250 1350 1450 1620 1670 1720 -1 Figure 3. Raman spectra of NMP-based nanofluids for three different graphene concentrations. Raman shift [cm ] Figure 3. Raman spectra of NMP-based nanofluids for three different graphene concentrations. Figure 3. Raman spectra of NMP-based nanofluids for three different graphene concentrations. 848.6 (a) 1298.2 ω (H-C-H) wagging ω (H-C-H) wagging 848.2 1297.8 [cm ][cm ] [cm ][cm ] (a) 1298.2 -1 -1 848.6 847.8 1297.4 ω (H-C-H) wagging (d) ω (H-C-H) wagging 848.2 ω (H-C-H) wagging 1297.8 744.7 τ (H-C-H) twisting (e) -1 -1 Peak position 1224.8 Peak position 847.8 744.3 1297.4 (d) 1224.4 744.7 ω (H-C-H) wagging τ (H-C-H) twisting 743.9 (b) (e) Peak position 1224.8 Peak position 1224.0 744.3 1224.4 616.0 δ (N-C-C) scissoring (c) 924.9 νs (C-C-C) symmetric stretching 743.9 (b) 1224.0 615.9 924.8 616.0 δ (N-C-C) scissoring 615.8 (c) 924.9 924.7 νs (C-C-C) symmetric stretching (f) 615.9 924.8 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.02 0.04 0.06 0.08 0.10 615.8 924.7 (f) Concentration [mg/ml] Concentration [mg/ml] 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.02 0.04 0.06 0.08 0.10 Concentration Figure 4. (a–f) Peak positions [mg/ml] of six Raman bands of NMP shown inConcentration Figure 3 as a[mg/ml] function of Figure 4. (a)–(f) Peak positions of six Raman bands of NMP shown in Figure 3 as a function of graphene concentration. graphene concentration. Figure 4. (a)–(f) Peak positions of six Raman bands of NMP shown in Figure 3 as a function of At this stage, one could consider as a possibility that higher concentrations of graphene may be Atgraphene concentration. this stage, one could consider as a possibility that higher concentrations of graphene may be preferable in order to observe a modification of the Raman spectra of NMP. However, although we preferable in order to observe a modification of the Raman spectra of NMP. However, although we preparedAt thishigher stage, concentrations one could consider of this as ananofluid, possibility it was not possible to measure their Raman prepared higher concentrations of this nanofluid, it wasthatnothigher possible concentrations to measure their of graphene may be Raman spectra. spectra. preferableIn addition in order to to the high observe light a absorption modification of in thethe most Raman concentrated spectra of samples, NMP. However,we also observed although we In addition to the high light absorption in the most concentrated samples, we also observed photoluminescence, which prepared higher concentrations masks the Raman peaks. The origin of this photoluminescence has been photoluminescence, which masksofthe thisRaman nanofluid, peaks.it was Thenot possible origin of this to photoluminescence measure their Raman hasspectra. been attributed In addition to the to sonochemical the high light degradation absorption of in the the NMP most suffered concentratedduringsamples, the preparation we alsoof observed the NFs. attributed to the sonochemical degradation of the NMP suffered during the preparation of the NFs. Previous investigations photoluminescence, [67,68] which have demonstrated masks the Raman that sonication peaks. The origin ofofpure thisNMP produces contaminating photoluminescence has been Previous investigations [67,68] have demonstrated that sonication of pure NMP produces organic nanoparticles. attributed to the Ogilvie sonochemical et al. [68] degradation also reported strong photoluminescence and enhancement contaminating organic nanoparticles. Ogilvieof etthe NMP al. [68] suffered also reported during strongthephotoluminescence preparation of the and NFs. of the optical Previous absorbance of[67,68] investigations NMP-based have NFs after a sonication demonstrated that treatment. In sonication of Figure pure 5a,b, NMP weproduces can see enhancement of the optical absorbance of NMP-based NFs after a sonication treatment. In Figure the transmissionorganic contaminating electron microscopy (TEM) nanoparticles. Ogilvie images of two et al. [68] alsographene reported flakes strong sonicated in DMAc and photoluminescence and 5a,b, we can see the transmission electron microscopy (TEM) images of two graphene flakes sonicated NMP, respectively. It is interesting to observe in Figure 5b small particles on the graphene flakes. inenhancement DMAc and of NMP,the optical absorbance respectively. of NMP-based It is interesting NFs after to observe a sonication in Figure 5b small treatment. particlesInon Figure the The 5a,b,sample we can ofsee thethe image corresponds transmission to a nanofluid electron microscopy of graphene-NMP, (TEM) images which of two was sonicated graphene flakes during sonicatedthe graphene flakes. The sample of the image corresponds to a nanofluid of graphene-NMP, which was preparation in DMAcduring stage. and NMP,Similar particles were respectively. also It isSimilar found bytoYau interesting et al. [67] after the 5bsonication of dispersion sonicated the preparation stage. particlesobserve were also in found Figure by small Yau particles et al. [67] afteronthethe of NMP graphene and single-walled flakes. The sample carbon of nanotubes. the image Their corresponds experimental to a nanofluiddata suggested of graphene-NMP,that the sonication which was sonication of dispersion of NMP and single-walled carbon nanotubes. Their experimental data leads to theduring sonicated polymerization of the stage. fluid, producing contamination by smallby organic etnanoparticulate suggested that thethe preparation sonication leads to theSimilar particles polymerization were of thealso found fluid, producingYau contamination al. [67] after by the products sonication (2–15 nm). Taking ofnanoparticulate dispersion of NMP into account and(2–15 that the single-walled thermal carbonconductivity nanotubes. of graphene-NMP Their experimental NFs did data small organic products nm). Taking into account that the thermal conductivity not change that suggested withtheincreasing sonication graphene leads toconcentration, the polymerization it seems of thatfluid, the the degradation producing of NMP and the contamination by of graphene-NMP NFs did not change with increasing graphene concentration, it seems that the resulting small organic organic contamination nanoparticulate can affect products (2–15 thenm). interaction Taking between into account thethat additive the nanomaterial thermal (i.e., conductivity degradation of NMP and the resulting organic contamination can affect the interaction between the graphene) of graphene-NMPand the solvent, NFs did asnot well as the with change thermal performance increasing graphene of the NF. This causes concentration, a detrimental it seems thatNF.the additive nanomaterial (i.e., graphene) and the solvent, as well as the thermal performance of the impact on the thermal conductivity of the nanofluid, in a manner that is still unknown. This causes a detrimental impact on the thermal conductivity of the nanofluid, in a manner that the degradation of NMP and the resulting organic contamination can affect the interaction between is additive still unknown.nanomaterial (i.e., graphene) and the solvent, as well as the thermal performance of the NF. This causes a detrimental impact on the thermal conductivity of the nanofluid, in a manner that is still unknown. Nanomaterials2019, Nanomaterials 2019,9,9, 804x FOR PEER REVIEW 88ofof1212 Figure 5. Representative transmission electron microscopy (TEM) images of graphene from: (a) a DMAc Figure 5.and dispersion Representative (b) an NMP transmission dispersion.electron microscopy Both images show(TEM) images of nature the multilayer graphene of from: (a) a the flakes DMAc dispersion and (b) an NMP dispersion. Both images show the multilayer nature of the folded graphene sheets. Highlighted by green circles, some nanoparticulates can be observed in theflakes folded graphene NMP-based sheets. Highlighted by green circles, some nanoparticulates can be observed in nanofluid. the NMP-based nanofluid. 4. Conclusions In this work, a correlation between the modification of the vibrational spectra of the fluids, 4. Conclusions and the increase of their thermal conductivity and viscosity in the presence of graphene is presented, whereIn thethis work, base fluidsa correlation were DMAc, between DMF, and the NMP. modification Our resultsof the vibrational suggest that the spectra of the presence of fluids, graphene and the increase of their thermal conductivity and viscosity in produces important changes at the macroscopic level in DMF and DMAc-based fluids. This is the presence of graphene is presented, where the reflected in base fluids were DMAc, the displacement of some DMF, and NMP. Raman bandsOurwhen results suggest that the the concentration of presence grapheneofincreases. graphene produces The blue-shift important of severalchanges Raman bands at theinmacroscopic DMF and DMAc level reveals in DMF and DMAc-based a hardening of the bondsfluids. This is associated reflected with thosein the displacement vibrational of some Ramaninteractions), modes (intramolecular bands when the which concentration requires higher of graphene energy increases. to make The vibrate. them blue-shiftIn of several a way, it canRaman bands inasDMF be explained and DMAc if graphene has the reveals abilityatohardening modify, asofanthe bonds external associated force, with those vibrational the solvent–solvent interaction, modes (intramolecular and fluid molecules, in interactions), turn, influence which the requires higher bonds intramolecular energy oftothe make them vibrate. neighbor molecules. In a This way,interpretation it can be explained as ifthat suggests graphene there ishas the ability a strong to modify, modification inas the an external force, intermolecular the solvent–solvent interaction between fluidinteraction, molecules inand fluid molecules, the presence of graphene in flakes, turn, which influencein turnthe intramolecular seems to be directlybonds of the neighbor connected molecules. This with the enhancement of interpretation the thermal conductivitysuggests that there of the is aIn fluid. strong the modification case in the intermolecular of NMP-graphene-based nanofluid interaction between no significant fluid molecules modification in theinthermal the presence of graphene conductivity was flakes,Similarly, found. which in we turn did not seems find an toimportant be directly connectedin their modification with Raman the enhancement spectra either.ofThe thestructure thermal conductivity of the fluid. In the case of NMP-graphene-based nanofluid of the NMP molecule and the peculiar arrangement of the bare fluid (dimerization and inner order), no significant modification in the can explainthermal whyconductivity the presencewas found. Similarly, of graphene does notwe didNMP affect not find an important in vibrational modification terms. NMP is a in their cyclic Raman amide and spectra either. tendency has a strong The structure to formof the NMPinmolecule dimers the liquidand state,thetherefore, peculiar arrangement the modification of the bare of the fluid (dimerization intramolecular forces andin theinner cycleorder), wouldcan explain require greatwhy the presence efforts. We can deduce of graphene does not from Raman affect NMP spectroscopy in vibrational and terms. NMP thermal conductivity is a cyclic results, amideconcentrations that those and has a strong tendencyare of graphene to insufficient form dimerstoinimpactthe liquid on the fluid’s thermal properties. We tested a higher concentration of graphene in NMP (C = 0.5 mg/mL), state, therefore, the modification of the intramolecular forces in the cycle would require great efforts. Wethe but canthermal deduceconductivity from Ramanwas spectroscopy not enhanced, and and thermal conductivity the Raman spectrum results, showedthat photoluminescence, those concentrations of graphene which masks the areRaman insufficient peaks.toPhotoluminescence impact on the fluid’s can be thermal attributed properties. We tested of to the degradation a NMP higher concentration of graphene in NMP (C = 0.5 mg/mL), but the thermal during sonication, which is a crucial step in the preparation of stable dispersions. The sonochemical conductivity was not enhanced, and the Raman degradation of NMP,spectrumtogether with showed photoluminescence, the formation of the small organic which contaminating masks the Raman peaks. particulates, Photoluminescence affects can be attributed the interaction between the grapheneto theflakes, degradation of NMP which could alsoduring sonication, contribute which is aimpact to a detrimental crucial onstep thein the preparation thermal conductivity. of stable dispersions. The sonochemical degradation of NMP, together with the formation of the small organic contaminating particulates, affects the interaction between the Supplementary graphene flakes, Materials: The following which could are available also contribute online at http://www.mdpi.com/2079-4991/9/5/804/s1, to a detrimental impact on the thermal conductivity. Table S1: Summary of the Raman measurements taken on different days over a period of time greater than one year, Figure S1: Raman spectra of graphene DMAc-NFs recorded on different dates as a function of graphene Supplementary Materials: The following are available online at http://www.mdpi.com/2226-4310/4/2/18/s1, Table S1: Summary of the Raman measurements taken on different days over a period of time greater than one Nanomaterials 2019, 9, 804 9 of 12 concentration: (a) 0.1 mg/mL, (b) 0.2 mg/mL and (c) 0.5 mg/mL, Figure S2: Raman spectra of NMP-NFs recorded on different dates as a function of graphene concentration: (a) 0.05 mg/mL and (b) 0.1 mg/mL, Figure S3: Raman spectra of DMF-based nanofluids recorded on different dates as a function of graphene concentration: (a) 0.10 mg/mL, (b) 0.25 mg/mL and (b) 0.50 mg/mL. Author Contributions: M.d.R.R.-L. and E.C.-A. wrote the article. M.d.R.R.-L. and P.G.-R. fabricated the samples. M.d.R.R.-L. and E.C.-A. measured the thermal conductivity, viscosity and Raman spectra. P.G.-R. and C.M.S.T. supervised the work. All authors discussed the results and commented on the manuscript. Funding: ICN2 acknowledges support from Severo Ochoa Program (MINECO, Grant SEV-2017-0706) and funding from the CERCA Programme/ Generalitat de Catalunya. Funding from the Spanish Ministry (MINECO/FEDER: MAT2015-68394-R NaCarFLOW and FIS2015-70862-P PHENTOM) is also acknowledged. We would also like to thank Andrew Hudson for the assistance provided with the use of English in this manuscript. Conflicts of Interest: The authors declare no conflict of interest. References 1. Shi, L.; Dames, C.; Lukes, J.R.; Reddy, P.; Duda, J.; Cahill, D.G.; Lee, J.; Marconnet, A.; Goodson, K.E.; Bahk, J.-H.; et al. Evaluating Broader Impacts of Nanoscale Thermal Transport Research. Nanoscale Microscale Thermophys. Eng. 2015, 19, 127–165. [CrossRef] 2. International Roadmap for Devices and Systems (IRDS): Beyond CMOS. Available online: https://irds.ieee. org/images/files/pdf/2017/2017IRDS_BC.pdf (accessed on 26 May 2019). 3. Ekpu, M.; Bhatti, R.; Ekere, N.; Mallik, S. Advanced thermal management materials for heat sinks used in microelectronics. In Proceedings of the 18th European Microelectronics & Packaging Conference, Brighton, UK, 12–15 September 2011; pp. 1–8. 4. Ye, H.; Zhang, G. A review of passive thermal management of LED module. J. Semicond. 2011, 32, 14008. [CrossRef] 5. Sohel Murshed, S.M.; Nieto de Castro, C.A. A critical review of traditional and emerging techniques and fluids for electronics cooling. Renew. Sustain. Energy Rev. 2017, 78, 821–833. [CrossRef] 6. Bar-Cohen, A.; Arik, M.; Ohadi, M. Direct Liquid Cooling of High Flux Micro and Nano Electronic Components. Proc. IEEE 2006, 94, 1549–1570. [CrossRef] 7. Siddique, A.R.M.; Muresan, H.; Majid, S.H.; Mahmud, S. An adjustable closed-loop liquid-based thermoelectric electronic cooling system for variable load thermal management. Therm. Sci. Eng. Prog. 2019, 10, 245–252. [CrossRef] 8. Tasaka, M.; Shinohara, K.; Hayashi, C.; Kashima, S.; Koyama, K. Cooling performance of heat sinks with corrugated-fins. In Proceedings of the ITherm’98. Sixth Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (Cat. No.98CH36208), Seattle, WA, USA, 27–30 May 1998; pp. 104–111. 9. Budelman, G.A. US-Patent: Heatsink with integrated blower for improved heat transfer. No. 6244331, 2001. 10. Mahajan, R.; Chia-pin Chiu; Chrysler, G. Cooling a Microprocessor Chip. Proc. IEEE 2006, 94, 1476–1486. [CrossRef] 11. Leng, C.; Wang, X.-D.; Wang, T.-H. An improved design of double-layered microchannel heat sink with truncated top channels. Appl. Therm. Eng. 2015, 79, 54–62. [CrossRef] 12. Molina, J.; Prieto, R.; Narciso, J.; Louis, E. The effect of porosity on the thermal conductivity of Al–12wt.% Si/SiC composites. Scr. Mater. 2009, 60, 582–585. [CrossRef] 13. Caccia, M.; Rodríguez, A.; Narciso, J. Diamond Surface Modification to Enhance Interfacial Thermal Conductivity in Al/Diamond Composites. JOM 2014, 66, 920–925. [CrossRef] 14. Molina, J.-M.; Rodríguez-Guerrero, A.; Louis, E.; Rodríguez-Reinoso, F.; Narciso, J. Porosity Effect on Thermal Properties of Al-12 wt % Si/Graphite Composites. Materials (Basel) 2017, 10, 177. [CrossRef] 15. Kheirabadi, A.C.; Groulx, D. Cooling of server electronics: A design review of existing technology. Appl. Therm. Eng. 2016, 105, 622–638. [CrossRef] 16. Lv, L.C.; Li, J. Micro Flat Heat Pipes for Microelectronics Cooling: Review. Recent Patents Mech. Eng. 2013, 6, 169–184. [CrossRef] 17. Garimella, S.V.; Persoons, T.; Weibel, J.A.; Gektin, V. Electronics Thermal Management in Information and Communications Technologies: Challenges and Future Directions. IEEE Trans. Compon. Packag. Manuf. Technol. 2017, 7, 1191–1205. [CrossRef] Nanomaterials 2019, 9, 804 10 of 12 18. Patil, M.; Kim, S.; Seo, J.; Lee, M. Review of the Thermo-Physical Properties and Performance Characteristics of a Refrigeration System Using Refrigerant-Based Nanofluids. Energies 2015, 9, 22. [CrossRef] 19. Mahian, O.; Kianifar, A.; Kalogirou, S.a.; Pop, I.; Wongwises, S. A review of the applications of nanofluids in solar energy. Int. J. Heat Mass Transf. 2013, 57, 582–594. [CrossRef] 20. Safaei, M.; Ahmadi, G.; Goodarzi, M.; Safdari Shadloo, M.; Goshayeshi, H.; Dahari, M. Heat Transfer and Pressure Drop in Fully Developed Turbulent Flows of Graphene Nanoplatelets–Silver/Water Nanofluids. Fluids 2016, 1, 20. [CrossRef] 21. Choi, S.U.S.; Eastman, J.A. Enhancing thermal conductivity of fluids with nanoparticles. ASME Int. Mech. Eng. Congr. Expo. 1995, 66, 99–105. 22. Branson, B.T.; Beauchamp, P.S.; Beam, J.C.; Lukehart, C.M.; Davidson, J.L. Nanodiamond Nanofluids for Enhanced Thermal Conductivity. ACS Nano 2013, 7, 3183–3189. [CrossRef] 23. Patil, M.; Seo, J.-H.; Kang, S.-J.; Lee, M.-Y. Review on Synthesis, Thermo-Physical Property, and Heat Transfer Mechanism of Nanofluids. Energies 2016, 9, 840. [CrossRef] 24. Azmi, W.H.; Sharma, K.V.; Mamat, R.; Najafi, G.; Mohamad, M.S. The enhancement of effective thermal conductivity and effective dynamic viscosity of nanofluids – A review. Renew. Sustain. Energy Rev. 2016, 53, 1046–1058. [CrossRef] 25. Bhanushali, S.; Jason, N.N.; Ghosh, P.; Ganesh, A.; Simon, G.P.; Cheng, W. Enhanced Thermal Conductivity of Copper Nanofluids: The Effect of Filler Geometry. ACS Appl. Mater. Interfaces 2017, 9, 18925–18935. [CrossRef] [PubMed] 26. Sekrani, G.; Poncet, S. Ethylene- and Propylene-Glycol Based Nanofluids: A Litterature Review on Their Thermophysical Properties and Thermal Performances. Appl. Sci. 2018, 8, 2311. [CrossRef] 27. Rodríguez-Laguna, M.R.; Castro-Alvarez, A.; Sledzinska, M.; Maire, J.; Costanzo, F.; Ensing, B.; Pruneda, M.; Ordejón, P.; Sotomayor Torres, C.M.; Gómez-Romero, P.; et al. Mechanisms behind the enhancement of thermal properties of graphene nanofluids. Nanoscale 2018, 10, 15402–15409. [CrossRef] 28. Wei, C.; Nan, Z.; Wang, X.; Tan, Z. Investigation on Thermodynamic Properties of a Water-Based Hematite Nanofluid. J. Chem. Eng. Data 2010, 55, 2524–2528. [CrossRef] 29. Shin, D.; Banerjee, D. Enhanced Specific Heat of Silica Nanofluid. J. Heat Transf. 2011, 133, 24501. [CrossRef] 30. Tiznobaik, H.; Shin, D. Enhanced specific heat capacity of high-temperature molten salt-based nanofluids. Int. J. Heat Mass Transf. 2013, 57, 542–548. [CrossRef] 31. Ho, M.X.; Pan, C. Optimal concentration of alumina nanoparticles in molten Hitec salt to maximize its specific heat capacity. Int. J. Heat Mass Transf. 2014, 70, 174–184. [CrossRef] 32. Shahrul, I.M.; Mahbubul, I.M.; Khaleduzzaman, S.S.; Saidur, R.; Sabri, M.F.M. A comparative review on the specific heat of nanofluids for energy perspective. Renew. Sustain. Energy Rev. 2014, 38, 88–98. [CrossRef] 33. Riazi, H.; Murphy, T.; Webber, G.B.; Atkin, R.; Tehrani, S.S.M.; Taylor, R.A. Specific heat control of nanofluids: A critical review. Int. J. Therm. Sci. 2016, 107, 25–38. [CrossRef] 34. Hu, Y.; He, Y.; Zhang, Z.; Wen, D. Enhanced heat capacity of binary nitrate eutectic salt-silica nanofluid for solar energy storage. Sol. Energy Mater. Sol. Cells 2019, 192, 94–102. [CrossRef] 35. Roberts, N.A.; Walker, D.G. Convective Performance of Nanofluids in Commercial Electronics Cooling Systems. Appl. Therm. Eng. 2010, 30, 2499–2504. [CrossRef] 36. Escher, W.; Brunschwiler, T.; Shalkevich, N.; Shalkevich, A.; Burgi, T.; Michel, B.; Poulikakos, D. On the Cooling of Electronics With Nanofluids. J. Heat Transf. 2011, 133, 51401. [CrossRef] 37. Rafati, M.; Hamidi, A.A.; Shariati Niaser, M. Application of nanofluids in computer cooling systems (heat transfer performance of nanofluids). Appl. Therm. Eng. 2012, 45–46, 9–14. [CrossRef] 38. Ijam, A.; Saidur, R. Nanofluid as a coolant for electronic devices (cooling of electronic devices). Appl. Therm. Eng. 2012, 32, 76–82. [CrossRef] 39. Nazari, M.; Karami, M.; Ashouri, M. Comparing the thermal performance of water, Ethylene Glycol, Alumina and CNT nanofluids in CPU cooling: Experimental study. Exp. Therm. Fluid Sci. 2014, 57, 371–377. [CrossRef] 40. Liu, Z.-H.; Li, Y.-Y. A new frontier of nanofluid research – Application of nanofluids in heat pipes. Int. J. Heat Mass Transf. 2012, 55, 6786–6797. [CrossRef] 41. Sureshkumar, R.; Mohideen, S.T.; Nethaji, N. Heat transfer characteristics of nanofluids in heat pipes: A review. Renew. Sustain. Energy Rev. 2013, 20, 397–410. [CrossRef] Nanomaterials 2019, 9, 804 11 of 12 42. Alawi, O.A.; Sidik, N.A.C.; Mohammed, H.A.; Syahrullail, S. Fluid flow and heat transfer characteristics of nanofluids in heat pipes: A review. Int. Commun. Heat Mass Transf. 2014, 56, 50–62. [CrossRef] 43. Das, S.; Giri, A.; Samanta, S.; Kanagaraj, S. Role of graphene nanofluids on heat transfer enhancement in thermosyphon. J. Sci. Adv. Mater. Devices 2019, 4, 163–169. [CrossRef] 44. Poplaski, L.M.; Benn, S.P.; Faghri, A. Thermal performance of heat pipes using nanofluids. Int. J. Heat Mass Transf. 2017, 107, 358–371. [CrossRef] 45. Sarkar, S.; Selvam, R.P. Molecular dynamics simulation of effective thermal conductivity and study of enhanced thermal transport mechanism in nanofluids. J. Appl. Phys. 2007, 102, 074302. [CrossRef] 46. Prasher, R.; Bhattacharya, P.; Phelan, P.E. Brownian-Motion-Based Convective-Conductive Model for the Effective Thermal Conductivity of Nanofluids. J. Heat Transf. 2006, 128, 588. [CrossRef] 47. Prasher, R.; Phelan, P.E.; Bhattacharya, P. Effect of Aggregation Kinetics on the Thermal Conductivity of Nanoscale Colloidal Solutions (Nanofluid). Nano Lett. 2006, 6, 1529–1534. [CrossRef] 48. Keblinski, P.; Thomin, J. Hydrodynamic field around a Brownian particle. Phys. Rev. E 2006, 73, 10502. [CrossRef] 49. Keblinski, P.; Phillpot, S.R.; Choi, S.U.S.; Eastman, J.A. Mechanisms of heat flow in suspensions of nano-sized particles (nanofluids). Int. J. Heat Mass Transf. 2002, 45, 855–863. [CrossRef] 50. Trisaksri, V.; Wongwises, S. Critical review of heat transfer characteristics of nanofluids. Renew. Sustain. Energy Rev. 2007, 11, 512–523. [CrossRef] 51. Cahill, D.G. Thermal conductivity measurement from 30 to 750 K: the 3ω method. Rev. Sci. Instrum. 1990, 61, 802. [CrossRef] 52. Cahill, D.G. Erratum: “Thermal conductivity measurement from 30 to 750 K: The 3ω method” [Rev. Sci. Instrum. 61, 802 (1990)]. Rev. Sci. Instrum. 2002, 73, 3701. [CrossRef] 53. Oh, D.W.; Jain, A.; Eaton, J.K.; Goodson, K.E.; Lee, J.S. Thermal conductivity measurement and sedimentation detection of aluminum oxide nanofluids by using the 3ω method. Int. J. Heat Fluid Flow 2008, 29, 1456–1461. [CrossRef] 54. Lubner, S.D.; Choi, J.; Wehmeyer, G.; Waag, B.; Mishra, V.; Natesan, H.; Bischof, J.C.; Dames, C. Reusable bi-directional 3 ω sensor to measure thermal conductivity of 100-µ m thick biological tissues. Rev. Sci. Instrum. 2015, 86, 014905. [CrossRef] 55. Chavez-Angel, E.; Reuter, N.; Komar, P.; Heinz, S.; Kolb, U.; Kleebe, H.-J.; Jakob, G. Subamorphous Thermal Conductivity of Crystalline Half-Heusler Superlattices. Nanoscale Microscale Thermophys. Eng. 2019, 23, 1–9. [CrossRef] 56. Kole, M.; Dey, T.K. Investigation of thermal conductivity, viscosity, and electrical conductivity of graphene based nanofluids. J. Appl. Phys. 2013, 113, 84307. [CrossRef] 57. Mehrali, M.; Sadeghinezhad, E.; Tahan Latibari, S.; Mehrali, M.; Togun, H.; Zubir, M.N.M.; Kazi, S.N.; Metselaar, H.S.C. Preparation, characterization, viscosity, and thermal conductivity of nitrogen-doped graphene aqueous nanofluids. J. Mater. Sci. 2014, 49, 7156–7171. [CrossRef] 58. Vallejo, J.P.; Żyła, G.; Fernández-Seara, J.; Lugo, L. Rheological behaviour of functionalized graphene nanoplatelet nanofluids based on water and propylene glycol:water mixtures. Int. Commun. Heat Mass Transf. 2018, 99, 43–53. [CrossRef] 59. Vallejo, J.; Żyła, G.; Fernández-Seara, J.; Lugo, L. Influence of Six Carbon-Based Nanomaterials on the Rheological Properties of Nanofluids. Nanomaterials 2019, 9, 146. [CrossRef] [PubMed] 60. Chalapathi, V.V.; Ramiah, K.V. Normal vibrations of N-Dimethylformamide and N, N-Dimethylacetamide. Proc. Indian Acad. Sci. 1968, 91, 109–122. [CrossRef] 61. Peek, P.S.; McDermott, D.P. Vibrational modes and frequencies of 2-pyrrolidinones and their deutero-isotopomers. Spectrochim. Acta Part A Mol. Spectrosc. 1988, 44, 371–377. [CrossRef] 62. Xu, W.; Wang, H.; Tao, Y.; Zheng, X. The structural organization of N-methyl-2-pyrrolidinone in binary mixtures probed by Raman spectroscopy: Experimental and quantum chemical results. J. Raman Spectrosc. 2018, 49, 362–371. [CrossRef] 63. Rozpłoch, F.; Patyk, J.; Stankowski, J. Graphenes Bonding Forces in Graphite. Acta Phys. Pol. A 2007, 112, 557–562. [CrossRef] 64. Kemnitz, C.R.; Loewen, M.J. “Amide Resonance” Correlates with a Breadth of C−N Rotation Barriers. J. Am. Chem. Soc. 2007, 129, 2521–2528. [CrossRef] [PubMed] Nanomaterials 2019, 9, 804 12 of 12 65. Adams, W.A.; Kruus, P.; Patraboy, T.J. The system sulfur dioxide–N-methyl-2-pyrrolidinone. Can. J. Chem. 1983, 61, 37–44. [CrossRef] 66. Basma, N.S.; Headen, T.F.; Shaffer, M.S.P.; Skipper, N.T.; Howard, C.A. Local Structure and Polar Order in Liquid N -Methyl-2-pyrrolidone (NMP). J. Phys. Chem. B 2018, 122, 8963–8971. [CrossRef] [PubMed] 67. Yau, H.C.; Bayazit, M.K.; Steinke, J.H.G.; Shaffer, M.S.P. Sonochemical degradation of N-methylpyrrolidone and its influence on single walled carbon nanotube dispersion. Chem. Commun. 2015, 51, 16621–16624. [CrossRef] [PubMed] 68. Ogilvie, S.P.; Large, M.J.; Fratta, G.; Meloni, M.; Canton-Vitoria, R.; Tagmatarchis, N.; Massuyeau, F.; Ewels, C.P.; King, A.A.K.; Dalton, A.B. Considerations for spectroscopy of liquid-exfoliated 2D materials: Emerging photoluminescence of N-methyl-2-pyrrolidone. Sci. Rep. 2017, 7, 16706. [CrossRef] [PubMed] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).