Nanoscale Advances View Article Online PAPER View Journal | View Issue Using evolved gas analysis – mass spectrometry to This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Cite this: Nanoscale Adv., 2019, 1, 2740 characterize adsorption on a nanoparticle surface† Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. Jordi Martı́nez-Esaı́n, ab Teresa Puig, b Xavier Obradors, b Josep Ros, a Jordi Farjas, c Pere Roura-Grabulosa, c Jordi Faraudo, b Ramón Yáñeza and Susagna Ricart *b The surface chemistry of nanoparticles is the key factor to control and predict their interactions with molecules, ions, other particles, other materials, or substrates, determining key properties such as nanoparticle stability or biocompatibility. In consequence, the development of new techniques or modification of classical techniques to characterize nanoparticle surfaces is of utmost importance. Here, a classical analysis technique, thermally evolved gas analysis – mass spectrometry (EGA-MS), is employed to obtain an image of the nanoparticle–solvent interface, unraveling the molecules present on the surface. As the use of complementary techniques is urged, the validity of EGA-MS characterization is corroborated by comparison with a previously reported surface characterization method. Previous studies were based on several experimental techniques and MD simulations using YF3 nano/ supraparticles and LaF3 nanoparticles as model systems. We demonstrate the applicability of this technique in two differently sized systems and two systems composed of the same ions on their surface but with a different inorganic core (e.g. LaF3 and YF3 nanoparticles). The results described in this paper agree well with our previous results combining experimental techniques and MD simulations. EGA-MS Received 18th February 2019 Accepted 3rd June 2019 not only revealed the ions attached to the nanoparticle surface but also shed light on their coordination (e.g. citrate attached to one or two carboxylate moieties). Thus, we show that EGA-MS is a useful and DOI: 10.1039/c9na00098d efficient technique to characterize the surface chemistry of nanoparticles and to control and predict rsc.li/nanoscale-advances their final properties. methodologies. Currently, nuclear magnetic resonance (NMR) Introduction is a crucial technique to characterize the surface chemistry of Nanoscale systems are formed basically from a core and as-synthesized NPs or those aer carrying out ligand a protective shell that stabilizes the system kinetically prevent- exchange.5–8 In addition, 2D NMR techniques (i.e. two dimen- ing its coagulation.1 Nowadays, the complete characterization of sional nuclear Overhauser effect (NOESY) experiment)1,4 and the nanoparticle (NP) surface is performed with the combina- saturation-transfer difference (STD)-NMR9 are useful tion of several techniques to understand the complete system approaches for the complete characterization of NP surface image.2 Surface chemistry has arisen as a new multidisciplinary chemistry. However, these common techniques have drawbacks approach to unravel and study the surface-to-ligand interac- such as the fast H-D exchange and re-dispensability problems in tions in nanoscale systems.3,4 The complete knowledge of the NMR, or the overlapping of different ligands containing the NP surface composition is currently a complex discipline in same functional group in infrared spectroscopy (IR), among nanoscience, due to the intricacy of developed synthetic others. Given these limitations and difficulties in characterizing nanoscale interfaces, it is of great interest to add additional techniques to the nanoscale characterization portfolio. Our a Departament de Quı́mica, Universitat Autònoma de Barcelona, 08193, Bellaterra, objective here is precisely to add a technique to this list, Spain showing how evolved gas analysis – mass spectrometry (EGA- b Institut de Ciència de Materials de Barcelona (ICMAB-CSIC), 08193, Bellaterra, MS) can be used to characterize nanoscale surfaces. Spain. E-mail: ricart@icmab.es c EGA-MS is a key technique for the identication of species University of Girona, Campus Montilivi, Edif. PII, E17071 Girona, Spain when a substance is subjected to a controlled temperature † Electronic supplementary information (ESI) available: Graphical summary of previously characterized systems as a reference to our results, fragmentation program. Among the vast number of applications (periodically patterns of the ionic species identied in EGA analysis, example of the reviewed by Materazzi et al.10 since 2006), pyrolysis and thermal representative mass : charge spectrum for typical EGA-MS and schematic stability studies on organic materials,11,12 and metallorganic13 representation of tetramethylammonium and ammonium on the NP surface. and inorganic salts14,15 are the most usual. EGA is also used to See DOI: 10.1039/c9na00098d 2740 | Nanoscale Adv., 2019, 1, 2740–2747 This journal is © The Royal Society of Chemistry 2019 View Article Online Paper Nanoscale Advances identify adsorbed gases on polymers,16 the surface structure of Nanoparticle synthesis catalysts,17 the functionalization of nanoparticles18,19 and In a 50 mL round-bottom ask equipped with a condenser and carbon nanotubes to prepare graphene nanoribbons.20 In a magnetic stirrer, citric acid (2.25 mmol) in 16 mL of MilliQ a vacuum, EGA-MS offers very high sensitivity and reduces the water was neutralized with tetramethylammonium hydroxide chance of secondary reactions between volatiles. Consequently, (6.75 mmol), followed by the addition of Ln(CH3COO)3$H2O it is very suitable to analyze decomposition processes where (1.5 mmol). The initial solution was heated until 100  C or several radicals evolve from the surface of NPs. EGA-MS allows cooled to 5  C, and then NH4F (4.5 mmol) in 4 mL of MilliQ us to monitor the evolution of the fragmentation pattern of the water was injected dropwise. Aer 2 h of reaction, the nal detected volatiles as a function of the sample temperature. The This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. mixture was cooled to room temperature. LnF3 particles were use of the fragmentation pattern of the molecules attached onto separated from the reaction medium by the addition of 10 mL of the NP surface ensures not only the detection of the presence of Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. ethanol (supraparticles) or acetone (nanoparticles), followed by capping agents (as common in reported studies), but also centrifugation at 10 000 rpm for 20 minutes. The separated extraction of crucial information about their behavior on the nanoparticles were re-dispersed in 20 mL of MilliQ water surface (e.g. coordination to the NP surface). forming a stable dispersion. EGA-MS of NPs has been used to determine the organic amount on the surface or how strongly the stabilizer is adsor- Characterization bed. Celik et al.18 reported the use of this technique coupled to a thermobalance (TG) to simultaneously measure the evolution Dynamic light scattering (DLS) analyses have been carried out at of the NP mass. Their main objective was to detect how the the Characterization of So-Materials Services at ICMAB using stabilizer is adsorbed onto the NP surface. Finally, they postu- a Zetasizer Nano Zs with a measurement range of 0.3 nm to 10.0 lated three different layers in the shell of the NP. On the other mm and a sensitivity of 0.1 mg mL1. Transmission electron hand, the work reported by Slostowski et al.19 is based on the microscopy (TEM) micrographs were obtained on a 120 kV JEOL surface characterization of CeO2 NPs by EGA-MS and IR to 1210 TEM, which has a resolution point of 3.2 Å. Nanoparticles unravel the presence of the alcohol used in the synthesis on the were washed ve times with a non-solvent (ethanol or acetone), NP surface. In this case, IR was used as the key technique to re-dispersed in water and dried with a N2 current. The EGA-MS detect the formation of carboxylates that bind to the NP surface setup consists of a quartz tube that is kept at 106 mbar. To with different coordinative mechanisms. In view of all this measure the sample temperature, samples are placed directly evidence, we aimed to analyze deeply different nanoscale on a platinum sheet that is in tight thermal contact with a K systems using EGA-MS in a vacuum as the main technique to thermocouple. Gases evolving from the sample are analyzed by study the surface chemistry. means of a quadrupole mass spectrometer, MS, (model Micro- In this work, we present a simple and effective way to char- vision Plus from MKS). The energy of ionizing electrons is 70 eV. acterize the surface chemistry of NPs containing several ions The MS is placed in a chamber next to the quartz tube so that adsorbed on the surface via the EGA-MS technique. We have the distance from the sample to the gas analyzer is about 40 cm. applied the method to three different systems reported previ- The quartz tube is placed inside a tube furnace. Experiments are ously: (i) YF3 nanoparticles of 5 nm, (ii) YF3 supraparticles of performed at a constant temperature rise of 5 K min1 or 10 80 nm and (iii) LaF3 nanoplatelets of 7 nm size.21,22 The use K min1. This setup allows us to monitor the evolution of the of (i) and (ii) allows us to compare compounds of different sizes volatiles and fragments as a function of the sample while maintaining the composition. Group (iii) allows us to temperature. observe the effect of different metals in the LnF3 compounds. The results agree with the previous set of experimental and Results and discussion computational characterization performed to obtain the real surface image of these systems.21,22 We show that EGA-MS Nanoparticles allows us to identify all the ions attached to the NP surface Particles were obtained following a previous study,23 based on and to know the coordination of the multidentate ligand (e.g. a modied co-precipitation method in water at 100  C to obtain citrate), disclosing if it is adsorbed by one or two carboxylate YF3 supraparticles and LaF3 NPs. To obtain YF3 NPs with a small moieties. size (5 nm), we used kinetic control during the synthesis at a temperature of 5  C.21 As shown in Fig. 1, these NPs are monodispersed in water and show a cubic crystalline structure Experimental section (YF3 NPs and supraparticles)21 and a hexagonal crystalline structure for LaF3 NPs.22 During the synthesis, several ionic Materials species were introduced: acetate (from the metallic precursor), Yttrium(III) acetate hydrate 99.9%, lanthanum(III) acetate tetramethylammonium citrate (stabilizer) and ammonium hydrate 99.9%, citric acid 99%, tetramethylammonium (counterion of the uorinating source). Previous studies21 have hydroxide 25% v/v in water and ammonium uoride > 99.99% shown the use of different cations in these systems to have an were purchased from Sigma-Aldrich. Ethanol 96% was inuence, not only in the NP size but, in general trends, on the purchased from Panreac and acetone 99.5% from Scharlau. All behavior and charge density.24,25 The nal composition of the reagents were used as received without further purication. system in this study contains two carboxylate species and two This journal is © The Royal Society of Chemistry 2019 Nanoscale Adv., 2019, 1, 2740–2747 | 2741 View Article Online Nanoscale Advances Paper This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. Fig. 1 DLS and TEM brightness images of the selected particles for the EGA-MS study. (a) YF3 nanoparticles synthesized at 5  C (DLS in black), (b) YF3 supraparticles obtained at 100  C (DLS in green) and (c) LaF3 nanoparticles obtained at 100  C (DLS in blue), all of them via a co-precipitation method in water. nitrogen-based compounds, impeding direct identication of strongly reduced because under the working conditions, the all adsorbed species by classical experimental techniques. mean free path is much longer than the distance from the In previous studies,21,22 we used a set of experimental tech- sample to the detector (see Fig. S2† as an example of EGA-MS at niques in combination with MD simulations to identify all a determined temperature showing all the volatile fragments species present on the surface and to create the complete image identied in this work). of the system. Several techniques were necessary due to the complexity of studying the surface chemistry of these particles. Monovalent species by EGA-MS From MD simulations we found that the surface of YF3 particles First, we compare YF3 NPs (Fig. 1a) to YF3 supraparticles and LaF3 particles is spontaneously covered by acetate, citrate (Fig. 1b). EGA-MS allows detection of the ligands attached to the and ammonium ions, while tetramethylammonium plays NP surface. Also, the evolution of the fragments as the a counterion role neutralizing the global negative charge of the temperature increases is related to the kinetics of the decom- particles (see Fig. S1 in the ESI†).21,22 position and provides information about the stability of the Given the drawbacks of commonly used experimental tech- different ligands; we have detected the presence of NH4+ on the niques, the implementation of a general protocol to charac- surface of the NPs as well as strongly adsorbed water, playing terize the surface of nanoparticles is required. Besides, EGA-MS a stabilizer role. The mass-to-charge ratio (m/z) fragments of 16, is a suitable technique to detect the volatiles released from the 17 and 18 (Fig. 2) can be interpreted as arising from H2O and NP surface during thermal treatment. The high sensitivity and NH3. Their relative intensities should be compared to the reliability of MS make it an advantageous technique to be fragmentation pattern of H2O (I16/I17/I18 ¼ —/0.22/1) and NH3 applied in nanoscale systems containing inorganic NPs. (I16/I17/I18 ¼ 0.8/1/—) as shown in Scheme S1.† The peak at Compared to the traditional measurement based on a TG 100  C corresponds mainly to water desorption with a non- coupled to EGA, the use of EGA in a vacuum signicantly negligible contribution of NH3 as revealed by the m/z ¼ 16 increases the accuracy and prevents the occurrence of signal. As the temperature increases above 200  C, two addi- secondary reactions or reactions with residual gases that may tional processes at around 220 and 300  C are detected in which affect the characterization of the groups evolved from the the contribution of NH3 dominates (higher intensity of 16 and surface. Also, volatile fragments on their way to the detector are 17 fragments). Finally, the sharp peak between 375 and 400  C 2742 | Nanoscale Adv., 2019, 1, 2740–2747 This journal is © The Royal Society of Chemistry 2019 View Article Online Paper Nanoscale Advances This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. Fig. 2 EGA-MS analysis of YF3 particles synthesized at 100  C and those synthesized at 5  C and LaF3 NPs synthesized at 100  C. All spectra are shown with detected m/z main fragments of water and ammonia, corresponding to water and ammonium cations adsorbed Fig. 3 EGA-MS analysis of YF3 particles synthesized at 100  C and 5  C on the NP surface. and LaF3 NPs synthesized at 100  C. All spectra are shown with detected m/z main fragments of tetramethylammonium release as trimethylamine. can be attributed to a combustion process. It does not provide any relevant information as most measured m/z fragments are identication (coming from the tetramethylammonium cation) embedded in this peak. In addition, LaF3 NPs show a similar are m/z ¼ 58, 59, 30 and 42 with expected ratios of (I58/I59/I30/ plot with a high adsorbed water contribution of the m/z frag- I42 ¼ 1/0.70/0.35/0.27). The exceeding contribution of fragment ment 18. EGA-MS has evidenced that water and ammonium are m/z ¼ 58 is related to the simultaneous release of acetate to adsorbed onto the NP/supraparticle surface (at the chamber form acetone.26,27 Acetone (coming from acetate) can be detec- pressure, the boiling point of free water is around 70  C). ted from the presence of m/z ¼ 58 and 43 with an expected ratio These two molecules are hard to detect with the experimental of I43/I58 ¼ 4 as shown in Scheme S3.† Also, the contribution of techniques used in previous studies, and XPS and IR were fragments m/z ¼ 30 and 42 is larger than expected; the higher needed to observe the presence of ammonium (see Fig. S1†) amount of small fragments is due to an enhancement of because a suitable technique depends on the particular NPs. spontaneous dissociation with temperature.28–30 The presence of tetramethylammonium in our system has The presence of acetate on the NP and supraparticle surface also been stated as trimethylamine for all NP and supraparticle (Fig. 4) is conrmed although the relative intensities of the m/z systems, thanks to the detection of fragments m/z ¼ 58 and 59 peaks cannot be determined due to the overlapping with tet- (see Fig. 3) with a ratio of I59/I58 ¼ 0.4 (see the fragmentation ramethylammonium fragments. Concerning EGA-MS results of path in Scheme S2†). The main peaks for trimethylamine YF3 particles vs. LaF3 NPs, we observe the same volatiles, but This journal is © The Royal Society of Chemistry 2019 Nanoscale Adv., 2019, 1, 2740–2747 | 2743 View Article Online Nanoscale Advances Paper behavior. In our case, we observed a lower temperature for LaF3 than for YF3 (Fig. 3). Tetramethylammonium plays the role of a counterion stabilizing the negative charge coming from citrates. It is released at a different temperature from the NP surface due to the different charge densities of the metals (La3+ vs. Y3+) and their citrate–metal interactions, affecting the binding energy between citrate and tetramethylammonium, as explained below for the case of citrate and summarized in Fig. S3.† This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. Unravelling the coordination of citrate To detect citrate attached to the NP surface, rst we need to know the decomposition path of this molecule. Two decom- position paths for citric acid have been described in the litera- ture:31 one at 175  C with the formation of citraconic anhydride and the other at 250  C with the formation of 1,3-acetonedi- carboxylic acid. Under the conditions used in this study, we expect the formation of citraconic anhydride because it takes place at a lower temperature (Scheme S4†). In consequence, the detection of citraconic anhydride (that has been identied through fragments m/z ¼ 39, 40 and 68 as shown Fig. 5a) is crucial. The intensity ratios of fragments 39 and 40 agree with the fragmentation pattern of citraconic anhydride (I39/I40/I68 ¼ 1/0.7/0.7). Again, the lower intensity of the fragment m/z ¼ 68 can be attributed to thermally induced dissociation that favors the formation of smaller fragments.28–30 From the EGA-MS curves, we can identify the presence of citraconic anhydride in YF3 particles thanks to the peaks of the above fragments at 200 and 285  C (Fig. 5a). The peak at 200  C is more intense in particles obtained at 5  C while the peak at 285  C is predominant for those synthesized at 100  C. This difference reveals the dependence of citrate coordination onto the NP surface on the synthesis temperature. At 5  C, the low thermal energy of the system explains the preferential coordi- nation with one carboxylic group. This interaction could be broken by a mechanism that starts from citrate to form cit- raconic anhydride, releasing water and carbon dioxide (Scheme 1a). Finally, citraconic anhydride is released from the surface of NPs at a temperature (T1) of approximately 200  C. On the other Fig. 4 EGA-MS analysis of YF3 particles synthesized at 100  C and hand, for particles synthesized at 100  C, the coordination with 5  C, as well as LaF3 NPs synthesized at 100  C. All spectra are shown two different carboxylic moieties accounts for the most intense with detected m/z main fragments of acetate release as acetone. peak at a higher temperature (T2 ¼ 285  C), as can be observed in Scheme 1b. These two citrate coordinations in YF3 NPs and supra- with an interesting difference: molecules/ions attached to particles were already postulated in our previous work via 1H uorine atoms (ammonium) on the NP surface are released at NMR analysis and MD simulations.21 From the present EGA-MS the same temperature while the ones attached to lanthanide results, we can establish a surface chemistry containing atoms (acetate and water) evolve at a lower temperature. This ammonium, acetate, citrate and tetramethylammonium in YF3 effect is related to the different behavior of the metals (e.g. their NPs/supraparticles, i.e. the same composition that was identi- cationic radii). ed using several experimental techniques and MD simula- Tetramethylammonium decomposition plots (Fig. 3) show tions. In addition, given the two different volatilization signicant differences in its release temperature, meaning that temperatures for the same molecule, we postulate a surface with it is not directly attached to uorine surface positions. This two different kinds of coordination for citrate whose relevance phenomenon could be explained considering the role of coun- depends on the preparation temperature. terions in this system. If tetramethylammonium is attached to Concerning the citrate bridge between YF3 NPs postulated as uoride positions, all NPs (LaF3 and YF3) should have the same the main reason for their self-assembly into YF3 2744 | Nanoscale Adv., 2019, 1, 2740–2747 This journal is © The Royal Society of Chemistry 2019 View Article Online Paper Nanoscale Advances This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. Fig. 5 Unravelling the coordination of citrate on the NP surface. (a) EGA-MS analysis of YF3 particles synthesized at 100  C (upper) or 5  C (bottom) with detected m/z peaks for citraconic anhydride. Schematic indicating the coordination of different ions on the NP surface for YF3 particles synthesized at (b) 5  C and (c) 100  C. (d) EGA-MS analysis of LaF3 particles with corresponding m/z peaks of citraconic anhydride. (e) Radial correlation function g(r) between oxygen atoms of citrate and metal atoms of the NP surface obtained for YF3 and LaF3 in MD simulations at 100  C.21,22 supraparticles,21 we consider that the citrate fragments The comparison between the results obtained in the case of contributing to this bridge evolve at a temperature similar to YF3 and LaF3 also provides additional physical insight. We that for mono- and bi-coordinated citrates and, consequently, observe that water and acetate are released from the NP surface their EGA-MS signals are overlapped. at lower temperatures in LaF3 than in YF3. This indicates In the case of LaF3 particles, EGA-MS gives a single peak a stronger interaction of these molecules with Y3+ which could related to the three main fragments of citraconic anhydride at be due to the smaller radius of yttrium. These differences in the 210  C (Fig. 5d). This means that we have one single kind of interaction of organic molecules with metals of different sizes citrate coordination onto the NP surface. However, it cannot be are common in coordination complexes in which the charge directly assigned to a given coordination from only the data of density of the metal cations is highly inuenced by their size. An EGA-MS of YF3 particles, due to the difference in charge density analogous effect should be expected with citrate. In fact, this is between Y3+ and La3+ cations. Our previous MD simulations22 what we obtained in our atomic MD simulations performed for predict that each adsorbed citrate moiety is coordinated to the LaF3 and YF3 in our previous publications.21,22 In this case, we LaF3 NP surface by two carboxylic groups. Therefore, the pres- used the results of the previous MD simulations to reinforce our ence of a single coordination mode in MD simulations is affirmation about the different coordination and to clarify this consistent with the presence of a single peak in EGA-MS. point with the use of this complementary technique. Scheme 1 Mechanism involved in the decomposition and release of citrate from the NP surface. (a) Citrate is attached to one carboxylic acid and is released at temperature T1 and (b) citrate is adsorbed by two different carboxylic moieties and is released at T2 as citraconic anhydride. This journal is © The Royal Society of Chemistry 2019 Nanoscale Adv., 2019, 1, 2740–2747 | 2745 View Article Online Nanoscale Advances Paper The results for the citrate–NP interaction are compared for moieties of citrate to the metal cation in the NP core. In LaF3 and YF3 in Fig. 5e. In this gure, we show the radial particular, YF3 NPs obtained at 5  C present a preferential correlation function g(r) between the oxygen atoms of citrate coordination with one carboxylate moiety. On the other hand, and NP metal surface atoms obtained in simulations (for YF3 supraparticles and LaF3 NPs synthesized at 100  C show detailed discussion on calculation and meaning of the radial a preferential coordination with two carboxylate moieties. The correlation function, see ref. 32). The location of the rst peak metal present in the NP (Y3+ or La3+) has a direct inuence on in the g(r) function indicates the distance between adsorbed the release temperature of the molecules or ions attached to citrate oxygen atoms and surface metal ions, and the relative these atoms (water, acetate and citrate) due to a different intensity of the peak indicates the strength of the interaction strength in the metal–molecule or metal–ion interaction This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. (technically, the potential of mean force). According to Fig. 5e, (charge density effect). In addition, the counterion role played the oxygen–metal distance is larger for lanthanum (2.35 Å by tetramethylammonium is also stated thanks to the difference Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. compared with 1.85 Å for yttrium). Furthermore, the rst between its decomposition patterns in LaF3 and YF3 NPs. These coordination peak is smaller for lanthanum by a factor of z0.87 differences mean that tetramethylammonium is interacting than for yttrium indicating a weaker metal–citrate oxygen with citrate adsorbed on the metal cation position, the metal– interaction for La3+. Therefore, MD simulation results imply ligand charge density being the main reason for the difference that citraconic anhydride will be released from the LaF3 NP in thermal stability. We found that the ions attached to surface at lower temperatures than from YF3, due to its weaker lanthanum were released at a lower temperature than those interaction (this shi to a lower temperature has also been adsorbed on yttrium atoms. This behavior is as expected observed for water and acetate release). This prediction from because of their cationic radii (La3+ is bigger than Y3+); the MD simulations is consistent with the EGA-MS peak tempera- bigger the metal radius is the lower the binding energy is. This tures of 285  C and 210  C for citraconic anhydride evolution is also consistent with the NP–ion interaction observed in our from YF3 and LaF3, respectively (Fig. 5a and d). Comparing the previous MD simulations.21,22 thermal energies  KBT (T, absolute temperature and KB, The present study is of interest not only for metal uoride Boltzmann constant) associated with these peak temperatures, NPs but, in general, for all kinds of nanoscale systems in which we obtain a ratio of 4.01 kJ mol1 (corresponding to 285  C)/ their surface composition is the key parameter to control and 4.64 kJ mol1 (corresponding to 210  C) z 0.86. This value design stable nanoscale suspensions. The use of EGA-MS as one coincides with the ratio between the rst peak of the g(r) func- of the main techniques to fully characterize nanoscale systems tion (Fig. 5e) measuring the relative strength of oxygen–citrate– avoids the necessity of performing several experimental char- metal interactions for LaF3 and YF3 NPs discussed above. In acterization studies to unravel the actual surface picture. summary, we can say that the citrate-related peaks at 285  C and Although in some cases the use of different techniques is 210  C for YF3 and LaF3, respectively (Fig. 5a and d), correspond necessary to gain greater insights into uncovering specic to the same decomposition mechanism, i.e. that identied at details of the obtained systems (e.g. spectroscopic techniques), the decomposition temperature T2 in Scheme 1. The tempera- here we demonstrate the efficiency of EGA-MS in NP surface tures are different for each metal because of the different characterization. strengths of the citrate–metal interaction. Note that EGA-MS and MD simulations are complementary techniques to Conflicts of interest unravel the dynamics and complex surface chemistry of NPs. There are no conicts to declare. Conclusions Acknowledgements We report an easy, fast and complete surface characterization method applied to nanoscale systems which allows the identi- This work was supported by the Spanish Ministry of Economy cation of all species attached to the NP surface. Using EGA-MS, and Competitiveness through the “Severo Ochoa” Programme we unraveled the surface composition of a set of three nano- for Centres of Excellence in R&D (SEV-2015-0496), the CON- scale systems (YF3 NPs, YF3 supraparticles and LaF3 NPs). The SOLIDER Excellence Network (MAT2015-68994-REDC), and EGA-MS results are in good agreement with a previous full SuMaTe RTI2018-095853-B-C21 and RTI2018-095853-B-C22 characterization study that involved several techniques and projects (both co-nanced by the European Regional Develop- works as a complementary technique to our previous MD ment Fund). We also acknowledge the support from the Euro- simulations. In other words, EGA-MS is able to provide pean Union for the EUROTAPES Project (FP7-NMP-Large-2011- complete characterization of surface chemistry avoiding the 280432) and from the Catalan Government with 2017-SGR- massive use of experimental techniques, but needs ancillary 1519. The authors acknowledge the technical support of Ser- techniques such as MD simulations to assess the complete vei de Microscòpia at the UAB. image. The surface chemistry of the NPs is composed of a combi- References nation of acetate, citrate, ammonium and tetramethylammo- nium. The same composition is obtained in all cases, but EGA- 1 J. De Roo, K. De Keukeleere, Z. Hens and I. Van Driessche, MS reveals differences in the coordination of the carboxylate Dalton Trans., 2016, 45, 13277–13283. 2746 | Nanoscale Adv., 2019, 1, 2740–2747 This journal is © The Royal Society of Chemistry 2019 View Article Online Paper Nanoscale Advances 2 H. Al-Johani, E. Abou-Hamad, A. Jedidi, C. M. Widdield, 18 D. Celik, M. Krueger, C. Veit, H. F. Schleiermacher, J. Viger-Gravel, S. S. Sangaru, D. Gajan, D. H. Anjum, B. Zimmermann, S. Allard, I. Dumsch, U. Scherf, S. Ould-Chikh, M. N. Hedhili, A. Gurinov, M. J. Kelly, M. El F. Rauscher and P. Niyamakom, Sol. Energy Mater. Sol. Eter, L. Cavallo, L. Emsley and J.-M. Basset, Nat. Chem., Cells, 2012, 98, 433–440. 2017, 9, 890–895. 19 C. Slostowski, S. Marre, O. Babot, T. Toupance and 3 M. A. Boles, D. Ling, T. Hyeon and D. V. Talapin, Nat. Mater., C. Aymonier, Langmuir, 2012, 28, 16656–16663. 2016, 15, 141–153. 20 B. Genorio, W. Lu, A. M. Dimiev, Y. Zhu, A. R. O. Raji, 4 J. De Roo, Y. Justo, K. De Keukeleere, F. Van den Broeck, B. Novosel, L. B. Alemany and J. M. Tour, ACS Nano, 2012, J. C. Martins, I. Van Driessche and Z. Hens, Angew. Chem., 6, 4231–4240. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Int. Ed., 2015, 54, 6488–6491. 21 J. Martinez-Esain, J. Faraudo, T. Puig, X. Obradors, J. Ros, 5 T. K. Egner, P. Naik, N. C. Nelson, I. I. Slowing and S. Ricart and R. Yáñez, J. Am. Chem. Soc., 2018, 140, 2127– Open Access Article. Published on 04 June 2019. Downloaded on 7/23/2021 9:14:48 AM. V. Venditti, Angew. Chem., Int. Ed., 2017, 56, 9802–9806. 2134. 6 R. R. Knauf, J. C. Lennox and J. L. Dempsey, Chem. Mater., 22 J. Martı́nez-Esaı́n, T. Puig, X. Obradors, J. Ros, R. Yáñez, 2016, 28, 4762–4770. J. Faraudo and S. Ricart, Angew. Chem., Int. Ed., 2018, 5, 7 L. E. Marbella and J. E. Millstone, Chem. Mater., 2015, 27, 14747–14751. 2721–2739. 23 J. Martı́nez-Esaı́n, J. Ros, J. Faraudo, S. Ricart and R. Yáñez, 8 Z. Hens and J. C. Martins, Chem. Mater., 2013, 25, 1211–1221. Langmuir, 2018, 34, 6443–6453. 9 S. Using, S. Di, Y. Zhang and L. B. Casabianca, J. Phys. Chem. 24 C. He, T. D. Nguyen, K. Edme, M. Olvera de Cruz and Lett., 2018, 9, 6921–6925. E. A. Weiss, J. Am. Chem. Soc., 2017, 139, 10126–10132. 10 S. Materazzi and R. Risoluti, Appl. Spectrosc. Rev., 2014, 49, 25 C. He, D. J. Weinberg, A. B. Nepomnyashchii, S. Lian and 635–665. E. A. Weiss, J. Am. Chem. Soc., 2016, 138, 8847–8854. 11 L. Sanchez-Silva, D. López-González, J. Villaseñor, P. Sánchez 26 R. A. Hites and K. Biemann, J. Am. Chem. Soc., 1972, 94, and J. L. Valverde, Bioresour. Technol., 2012, 109, 163–172. 5772–5777. 12 M. Brebu, T. Tamminen and I. Spiridon, J. Anal. Appl. 27 J. Farjas, J. Camps, P. Roura, S. Ricart, T. Puig and Pyrolysis, 2013, 104, 531–539. X. Obradors, Thermochim. Acta, 2011, 521, 84–89. 13 J. Farjas, J. Camps, P. Roura, S. Ricart, T. Puig and 28 L. Nattermann, O. Maßmeyer, E. Sterzer, V. Derpmann, X. Obradors, Thermochim. Acta, 2013, 556, 58–62. H. Y. Chung, W. Stolz and K. Volz, Sci. Rep., 2018, 8, 1–7. 14 C. Maton, N. De Vos and C. V. Stevens, Chem. Soc. Rev., 2013, 29 R. M. Reese, V. H. Dibeler and F. L. Mohler, J. Res. Natl. Bur. 42, 5963. Stand., 1951, 46, 79–84. 15 A. Llordés, K. Zalamova, S. Ricart, A. Palau, A. Pomar, 30 S. Rasi, S. Ricart, X. Obradors, T. Puig, P. Roura and J. Farjas, T. Puig, A. Hardy, M. K. Van Bael and X. Obradors, Chem. J. Anal. Appl. Pyrolysis, 2018, 133, 225–233. Mater., 2010, 22, 1686–1694. 31 I. A. Farbun, I. V. Romanova and S. A. Kirillov, J. Sol-Gel Sci. 16 T. Shimoaka, C. Wakai, T. Sakabe, S. Yamazaki and Technol., 2013, 68, 411–422. T. Hasegawa, Phys. Chem. Chem. Phys., 2015, 17, 8843–8849. 32 D. Frenkel and B. Smit, Understanding Molecular Simulation: 17 I. Murillo Leo, M. López Granados, J. L. G. Fierro and From Algorithms to Applications, Academic Press, San Diego, R. Mariscal, Appl. Catal., B, 2016, 185, 141–149. 2002. This journal is © The Royal Society of Chemistry 2019 Nanoscale Adv., 2019, 1, 2740–2747 | 2747