Magnetic Supraparticles Capable of Recording High‐Temperature Events

Superparamagnetic iron oxide nanoparticles (SPIONs) are prone to oxidation at elevated temperatures (>300 °C) and lose their magnetizability upon transition from magnetite/maghemite (γ‐Fe2O3 / Fe3O4) to hematite (α‐Fe2O3). Silica (SiO2) shells can effectively prevent this undesired effect up to ≈1000 °C. Herein, the study shows how to utilize SPIONs with varying SiO2 shell thickness and thus, different oxidation susceptibility, and how to combine them in micrometers sized assemblies – so‐called supraparticles (SPs), to create a structurally emerging magnetic temperature recording functionality. The desired oxidation of non or weakly‐protected SPIONs within SPs upon temperature events reduces dipole–dipole interactions of well‐protected SPIONs in the confined SP entity. The resulting change of magnetic interactions therefore contains information on the thermal history of the SP, which can be spectrally read out via magnetic particle spectroscopy within seconds. Their working range can be tuned from 400 to 1000 °C on two independent structural hierarchy levels, namely the SiO2 shell thickness and the freely selectable ratios of different building blocks in the SP. The application of such SPs as particulate additives for magnetic recording of high‐temperature events, especially relevant in metal, alloy, and ceramic processing, representing a yet unexplored and optically‐independent option for bulk temperature recording is proposed.


Introduction
The unique magnetic properties of superparamagnetic iron oxide nanoparticles (SPIONs) are exploited in fields such as hypothermia therapy, [1] diagnostic biosensing, [2] magnetic imaging techniques, [3] drug delivery, [4] or water purification purposes. [5,6]The favorable magnetic properties are however lost at elevated temperatures, as above 300 °C, a phase transition of (superparamagnetically behaving) ferrimagnetic maghemite/magnetite (-Fe 2 O 3 /Fe 3 O 4 ) to weakly ferromagnetic hematite (-Fe 2 O 3 ) occurs. [7,8]0][11][12][13][14] Herein, we demonstrate a way how to exploit the vulnerability toward oxidation and the consequent loss of (strong) magnetic properties on the one hand and the protectability of the magnetic properties using silica on the other hand in a combined system in order to achieve a functional magnetic system, namely a magnetic high-temperature recorder (See Supporting Information for the definition of "recorder" in this context).To do so, SPIONs with and without silica protection, respectively, are combined as two distinct building blocks in a so-called supraparticle (SP) via forced assembly using spray-drying.[21][22] Recently, we demonstrated that the sum of magnetic interactions in a static SP system can yield a unique fingerprint (ID), [18,23] while an SP that can undergo a structural change might yield an altered signal and thereby offering an irreversible recording or sensing functionality. [21]Based on exploiting the softening and thus structural changes in polymer systems, we showed that temperatures can be recorded up to 170 °C within an SP built from such polymer building blocks combined with SPIONs. [21]y the nature of its design, this previously reported SP is however not suited for recording higher temperatures (>400 °C).Being capable of recording elevated temperatures up to 1000 °C would, however, be very attractive for research and industrial settings of metal, [24] glass, [25] and glass-ceramic [26] processing, in which precise control of temperature events is crucial for material properties.To date, the exact thermal history (time-temperature integral), that a bulk material experiences, is difficult or impossible to record and differs from the duration and preset temperature of the furnace process.Common methods to measure and/ or record high-temperature events like thermocouples, [27,28] radiation-based techniques like infrared pyrometers, [29] temperature-sensitive resistors, [30] optical fibers [31] or process temperature control rings (PCTRs) [28,32] can only provide information about the thermal history in a furnace volume or on the surface of an object.Ideally, a material component would record and store information about its thermal history (including the inside of the bulk material) by itself and could provide this information upon readout in a fast and simple manner.
In this work, we present an SP-based approach that envisions utilizing magnetic nanoparticle interaction changes readable via MPS to indicate elapsed temperature events between 400 and 1000 °C with a customizable temperature working range and sensitivity.SPIONs with differently thick silica shells are prepared to finetune their onset temperature for oxidation to -Fe 2 O 3 .After the assembly of tailored ratios of differently oxidation-protected SPIONs to SPs, it is the SP structure that creates the recording functionality.Due to the temperature-triggered purposeful oxidation of non or weakly protected SPIONs at elevated temperatures, the changes in magnetic interactions are exploited.While the oxidation itself diminishes their magnetizability and could thus not be distinguished from background noise, the magnetization loss additionally reduces the magnetic interaction of neighboring nonoxidized SPIONs and can be reliably detected by MPS.By correlation of MPS data with temperature-induced irreversible structural changes of SPs, elapsed temperature events are magnetically resolved in a quantifiable manner.After elucidating the working principle, we provide a detailed look into the magnetic nanoparticle toolbox, that allows finetuning oxidation kinetics in the range up to 1000 °C and 8 h and thereby determines the working range of the used SPs.The design and readout of such particle architecture (changes) with the complementary magnetic information pathway readable via MPS, opens up an unexplored research field toward improved process quality control in the processing of metals, alloys, and ceramics.In contrast to optical properties, which are limited regarding their readout distance from the inside of an object, magnetic signals can travel through the bulk material regardless of its optical properties and thus could represent an alternative and complementary readout path. [18]

Design and Working Principle of Magnetic Supraparticles Capable of Recording High-Temperature Events
The design of the herein presented magnetic high-temperature event recording SP (HT-SP) allows to combine various ratios of differently well-protected SPIONs and thereby to freely adjust the recording sensitivity at different temperature ranges between 400 and 1000 °C.
For the purpose of schematically displaying the working principle of HT-SPs (Figure 1), three differently well-protected SPI-ONs are shown: nonprotected SPIONs, SPIONs with a thin SiO 2 shell, and SPIONs with a thick SiO 2 shell.Adjusting the SiO 2 shell thickness of SPIONs via the well-known Stöber process aims to yield differently oxidation-susceptible SPIONs, which are used as nanoparticle toolboxes in this work (Figure 1a).Initially, at room temperature (RT), all SPIONs are present in the form of -Fe 2 O 3 and/ or Fe 3 O 4 .Note that for the herein presented working principle, the exact share of Fe 3 O 4 and -Fe 2 O 3 is not of importance, as they both show superparamagnetic behavior (Magnetization curves and X-ray diffractograms can be found in Figure S1, Supporting Information).After experiencing an elevated temperature T 1 (e.g., 500 °C), unprotected SPIONs are oxidized to -Fe 2 O 3 , which shows extremely weak antiferromagnetism [33] and cannot be detected in MPS, while the superparamagnetic properties should be preserved for the SPIONs with thin and thick SiO 2 shells.Increasing the temperature even further to T 2 (e.g., 1000 °C), also the thin-shelled SPIONs are desired to oxidize to -Fe 2 O 3 and only the thickshelled SPIONs remain superparamagnetic and detectable via MPS.Thus, by suitable SiO 2 shell thickness variation, the temperature susceptibility of identical SPION cores should be readily adjustable.
On a nanoparticle level (Figure 1a), this feature can only distinguish between a meaningful signal response (-Fe 2 O 3 / Fe 3 O 4 ) and no signal response (oxidized to -Fe 2 O 3 ), as the latter does not differ from pure background noise and cannot be detected via MPS (Figure 1a).When combining nonprotected and differently well-protected SPIONs in an SP via spray-drying, the high temperature (HT) event recording functionality is expected to emerge, based on the structurally defined magnetic interaction of SPIONs within SPs (Figure 1b).At RT the dense packing of SPIONs in the SP causes pronounced magnetic dipole-dipole interactions, that lead to lower intensities of higher harmonics in the MPS spectrum.Exceeding 500 °C, all nonprotected SPIONs that are oxidized to -Fe 2 O 3 are not sufficiently magnetizable anymore to substantially polarize neighboring SPIONs and now should act as a sort of spacer between the protected, intact SPIONs.Hence, the total of all dipole-dipole interactions within the SP is decreased, thereby decreasing energy barriers to follow the external magnetic field, which leads to a rise in harmonic amplitude ratios.It was recently shown that MPS spectra of such SPs are highly sensitive to changes in dipole-dipole interactions. [21]The phenomenon that nonprotected SPIONs lose their superparamagnetic properties at high temperatures does not represent an undesired effect in this concept, but instead becomes a central desired part of the presented working principle.At even higher temperatures (e.g., 1000 °C) also thin-shelled SPIONs should oxidize and only thick-shelled SPIONs remain, leading to a further decrease in dipole-dipole interactions and thus, further increase of harmonic amplitude ratios (purple diamond symbols in Figure 1b).

Proof of Concept for Magnetic Supraparticles Capable of Recording High-Temperature Events
An HT-SP concept is envisioned, where the emergence of the HT recording functionality is purely based on the structural design of the HT-SPs, and no further temperature-sensitive building block is needed additional to the differently oxidation-susceptible SPIONs.Despite being directly dependent on the temperature exposure-dependent magnetic properties of these SPIONs, the HT-SP is supposed to show differing magnetic properties after temperature exposure, although being a mere combination of these building blocks.The superparamagnetic behavior of the herein-used iron oxide nanoparticles originates from their small diameter (≈10 nm).After co-precipitation of iron salts in an alkaline milieu, they are subsequently modified with citric acid (CA) on the surface and redispersed in water to obtain a ferrofluid of well-dispersed SPIONs, which is then well-suited for further modification with a SiO 2 -shell via Stöber (further details can be found in the Experimental Section). [19,34]In short, superparamagnetism means that these particles show no remanence and no coercivity (no hysteresis in the plot of magnetization over the external field) above their blocking temperature. [35]Thus, they exhibit no magnetization, if there is no externally applied field, which simplifies their processing and handling.A proof of the desired working principle of the hereinpresented HT-SP, which is based on the changing magnetic inter-actions of SPIONs within an SP, is displayed in detail in Figure 2. The schematic cross-section of an HT-SP containing mainly nonshelled and a small share of thick-shelled SPIONs illustrates the significantly reduced interactions within such SPs after 500 °C compared to their state before temperature treatment (Figure 2a).In this example, the nonshelled SPIONs have oxidized to -Fe 2 O 3 after being exposed to 500 °C for 1 h and do not provide any signal response detectable via MPS anymore.Rather, they now affect the spacing between the remaining thick-shelled SPIONs, whose mean average distance to each other is increased and consequently, their interaction decreases.Generally, less interaction of SPIONs contributes to greater magnetic amplitude intensities of higher harmonics in the MPS spectrum, which also holds true for the herein-presented HT-SP (Figure 2b). [17,19,21]Using the amplitude ratio A5/A3 is a common way to simplify the magnetic properties displayed by MPS spectra to a single value and will be used in this work from now on to display the existing structureproperty relation. [17,21,23]uch amplitude ratios as a function of temperature are shown for differently protected NPs (Figure 2c) and SPs (Figure 2d).The amplitude ratio A5/A3 of thick-shelled SPIONs (SPION-6S) does not change significantly in the temperature range of 20-1000 °C (Figure 2c).Also, the nonshelled SPIONs show only small changes of A5/A3 up to 400 °C.After exposure to higher temperatures, no meaningful signal can be detected for nonshelled SPIONs, as they have fully transitioned to -Fe 2 O 3 (see

Figure S2
, Supporting Information for X-ray diffractograms and magnetization curves).Progressively decreasing A5/A3 values for HT-SP that contain 85 wt.% nonshelled and 15 wt.% thickshelled (SPION-6S) SPIONs are observed for temperatures up to 350 °C, which can be attributed to surface effects, including the pyrolysis of the citric acid (CA) ligand (Figure 2d).The removal of the CA allows the SPIONs to be even more densely packed than before, leading to more interaction and lower A5/A3 (see Figure S3, Supporting Information) for more details on this, including thermogravimetric analysis and Fourier-transform infrared spectroscopy).However, when the temperature exceeds 400 °C, the nonshelled SPIONs are oxidized, causing A5/A3 to rise (Figure 2a).

Magnetic Nanoparticle Toolbox with Adjustable Temperature Response
The deliberate customization of the HT-SP working range and the corresponding sensitivity is realized on two different structural design levels, that are independent from one another.One level is the freely adjustable ratios of various nanoparticle building blocks that can be combined within an SP via spray-drying, which will be discussed in the subsequent Section 2.3.The other level is the susceptibility of these magnetic nanoparticle building blocks (in this case SPIONs) toward oxidation, to fine-tune the temperature at which their superparamagnetic behavior is turned off.The latter one will be discussed in this section and concerns the oxidation protective nature of the aforementioned SiO 2 shell on SPIONs.
The SiO 2 shells are grown onto the SPIONs via a previously reported particle-tailored Stöber protocol. [14,34,36]In short, SPIONs that have been modified with citric acid to form a stable aqueous ferrofluid, are further diluted with water and ethanol in basic conditions, and tetraethoxysilane (TEOS) is added stepwise.In such basic conditions, the TEOS undergoes hydrolysis preferentially on the SPION surface.If sufficient TEOS is added, a SiO 2 shell on the SPIONs is formed, which grows thicker upon further addition of TEOS.The number of TEOS additions and the quantity of TEOS per addition can easily be varied to adjust the resulting SiO 2 shell thickness.In this work, five modifications were used: SPIONs without any SiO 2 shell (SPION), SPI-ONs after one addition of TEOS (SPION-1S), two additions of TEOS (SPION-2S), four additions of TEOS (SPION-4S) and six additions of TEOS (SPION-6S) (see Experimental Section for details).These five SPION modifications are referred to as SPION candidates of a toolbox for the HT-SP design, as they are representatives of a larger number of possible SiO 2 shell modifications.
The resulting SiO 2 shell on the herein-used SPION candidates can be derived from transmission electron microscopy (TEM) images shown in SPION-1S (Figure 3-a2) and SPION-2S (Figure 3-a3), for SPION-4S (Figure 3-a4) and especially for SPION-6S (Figure 3-a5) the shell can be easily identified.Based on the analysis of TEM images, the SPION cores exhibit diameters of 10 ± 4 nm (see Figure S7, Supporting Information for hydrodynamic diameters derived from dynamic light scattering (DLS); SPIONs D 50, number = 10 nm), while the SiO 2 shells, with regard to their thickness, are in the range of 2 ± 1 nm for SPION-1S, 4 ± 2 nm for SPION-2S, 7 ± 3 nm for SPION-4S and 10 ± 5 nm for SPION-6S (shell thickness on one side of SPIONs, which adds to the radius).Fourier-transform infrared spectroscopy (FT-IR) confirms the presence of SiO 2 also for the thin shells (SPION-1S and SPION-2S) (Figure 3b).The FT-IR spectra for SPIONs exhibit a broad iron oxide band at 560 cm −1 [19,34,37] and two bands ≈1390 and 1580 cm −1 that are characteristic of citric acid. [19,34,38]The increasing share of SiO 2 from SPION-1S to SPION-6S leads to gradually lower intensities for these three bands, whereas the occurring band at ≈1050 cm −1 , which stems from the asymmetric Si─O─Si stretching, grows in intensity. [34,39]In accordance with this, the relative iron content of the SPION candidates is expected to decrease with increasing SiO 2 shell thickness.This was confirmed by inductively coupled plasma atomic energy spectroscopy (ICP-AES) measurements and correlated with the saturation magnetization, extracted from magnetization curves measured in a superconducting quantum unit interference device (SQUID; VSM measurement mode).It was found that the saturation magnetization decrease correlates well with decreasing iron content as the number of SiO 2 shell steps is increased (Figure 3c).This is in good agreement with expectation as the mass magnetization of the SPION core of the composite nanoparticles should not undergo significant changes from SPION-1S to SPION-6S, but the diamagnetic SiO 2 , which has no significant contribution to the overall particles' mass magnetization, increases in relative share across the series.The SiO 2 shell determines to what extent the oxidation of SPI-ONs is delayed or prevented at elevated temperatures (in this case 400-1000 °C).The temperature at which the phase transition of a SPION candidate from -Fe 2 O 3 / Fe 3 O 4 to -Fe 2 O 3 occurs relates to the temperature at which an HT-SP containing this SPION candidate will change its MPS signal.Hence, knowledge about the crystallographic and corresponding magnetic properties of the individual SPION candidates in dependence on temperature is crucial for the optimal selection of the suited SPION candidates to fine-tune the HT-SP recording sensitivity in the desired temperature range.Crystallographic and magnetic data confirm the adjustable oxidation susceptibility of the SPION candidates at elevated temperatures in dependence of SiO 2 shell thickness (Figure 4).Nonshelled SPIONs have entirely transitioned from -Fe 2 O 3 / Fe 3 O 4 to -Fe 2 O 3 after 1 h at 500 °C, whereas the SPION-6S almost fully remain in their superparamagnetic form as -Fe 2 O 3 / Fe 3 O 4 even after 1 h at 1000 °C (Figure 4a).The relative share of -Fe 2 O 3 / Fe 3 O 4 (in wt.%) out of all crystalline phases present in the SPION candidate samples determined via Rietveld refinement can be seen in Figure 4b (X-ray diffractograms of all SPION candidates after temperature exposure can be found in the supporting information Figure S4, Supporting Information).The increasing SiO 2 shell thickness shifts the oxidation temperature to higher values.Also, a small amount of the thick-shelled SPIONs (SPION-4S and SPION-6S) undergoes a phase transition at 1000 °C to the metastable polymorph -Fe 2 O 3 .Due to the large coercivity of this ferromagnetic phase, -Fe 2 O 3 nanoparticles are not excited in the MPS field. [40]Thus, they are considered to be magnetically "turned off" similar to -Fe 2 O 3 in this work, although the exact nature of magnetic interactions between -Fe 2 O 3 nanoparticles and SPIONs in the MPS field remains to be investigated.
To understand the origin of the -Fe 2 O 3 phase and even more importantly for this work, to elucidate how SiO 2 shells delay or prevent the oxidation of SPIONs, an in-depth investigation of crystallite sizes of these crystallites is performed (Figure 4c,d).Generally, the phase transitions of iron oxide nanoparticles are size-dependent. [8,10,41,42]With no protective shell, SPIONs will sinter at elevated temperatures to form larger -Fe 2 O 3 particles (Figure 4c,d).The SiO 2 shells act as sintering inhibitors that conserve the SPIONs in the form of ≈10 nm sized -Fe 2 O 3 /Fe 3 O 4 nanoparticles and thereby sustain their superparamagnetic properties. [8,41]The data as well as previous publications indicate, that oxidation of SPIONs cannot take place without sintering.Thin SiO 2 shells (SPION-1S and SPION-2S) still allow sintering and oxidation, but the phase transition occurs at higher temperatures.The resulting -Fe 2 O 3 nanoparticles form at higher temperatures and initially possess smaller crystallite sizes compared to the -Fe 2 O 3 nanoparticles, which stem from nonshelled SPIONs (Figure 4d).Consequently, the thin SiO 2 shells of SPION-1S and SPION-2S (2 and 4 nm, respectively) are either not present as a uniform coating around the SPIONs covering the whole surface, which allows partial contact between SPI-ONs and the formation of sintering necks, and/ or the mechanical feature as sintering inhibitor is weakened due to a softening of the SiO 2 shell at higher temperatures.The fact that thin SiO 2 shells do not fully prevent the sintering and oxidation of SPIONs, but instead only delay this process directly depending on the easily tunable thickness of the SiO 2 shell, represents an essential and desired part of the working principle of HT-SPs.
The formation of -Fe 2 O 3 takes place when a SiO 2 matrix offers a confined space of suitable size in which in this case only a few SPIONs can sinter. [8,43,44]][43] Due to the low surface energy of -Fe 2 O 3 , the transition to -Fe 2 O 3 is favored already at low temperatures whenever the crystallites grow too large. [41,42]The formation and conservation of -Fe 2 O 3 has been reported to work when iron salts or -Fe 2 O 3 /Fe 3 O 4 nanoparticles are present in confined vacancies in SiO 2 matrices and undergo temperature treatment.Evidently, a small share of the SPION-4S and SPION-6S candidates form agglomerates that provide these conditions required for -Fe 2 O 3 formation at 1000 °C (Figure 4a,d).The TEM images (Figure 3a) and the hydrodynamic diameter (Figure S7, Supporting Information) indicate, that SiO 2 shell formation also led to small agglomerates of multiple SPIONs.These agglomerates consist of a few SPIONs and while the agglomerate as a whole has a thick SiO 2 shell, the SPIONs within the shell are partially not well-protected regarding potential sintering.Hence, these agglomerate structures are in the required size range for the formation of -Fe 2 O 3 , [8,10] as can be observed for SPION-4S and SPION-6S.However, more importantly, the majority of SPIONs are sufficiently protected up to at least 1000 °C in the case of SPION-4S and SPION-6S and thus can be used as essential high-temperature stable HT-SP building blocks.(Relative shares of crystallographic species and crystallite sizes thereof can be found in Tables S1-S5, Supporting Information for all SPION candidates after temperature exposure).
The saturation magnetization of the SPION candidates (Figure 4e) illustrates that the magnetic properties of the SPION candidates are in direct correspondence to the crystallographic phases (Figure 4b).A smaller share of -Fe 2 O 3 / Fe 3 O 4 yields lower saturation magnetization values.This is evident when comparing the SPION candidates at RT (Figure 3c), but also when setting the saturation magnetization in relation to the oxidation of -Fe 2 O 3 / Fe 3 O 4 at elevated temperatures (Figure 4b) (Magnetization curves of all SPION candidates after temperature exposure can be found in the Figure S5, Supporting Information).The amplitude intensities in MPS also represent magnetic moments that can be plotted normalized on the mass (Figure 4f).In contrast to vibrating sample magnetometry (VSM), MPS works at low fields (30 mT) and high frequencies (20 kHz) to investigate relaxation and interaction phenomena, which is why the saturation magnetization of the sample is usually not reached. [17,22]However, the amplitude intensity A1 in MPS is directly proportional to the samples' magnetization and therefore shows the same trend, revealing the temperature-dependent oxidation of the SPION candidates.
Crystallographic and magnetic data of the SPION candidates validate the desired adjustability of oxidation susceptibility realized through SiO 2 shell thickness variation, whereby the hereinused shell thicknesses are just exemplary modifications, and the oxidation susceptibility can be tuned as a continuous (not stepwise) property from nonprotected to fully protected, allowing optimal fine-tuning of temperature-dependent magnetic properties.The information about magnetizability retrieved by the two complementary analytical methods VSM and MPS about the magnetic properties of SPION candidates are in very good agreement to each other (Figure 4e,f).In contrast to VSM, MPS comes with the advantage of being a fast (order of seconds) measurement at room temperature with customizable hand-held sensor geometries, that could be applied with little effort in a large variety of use cases, including quality control scenarios during manufacturing.Considering the amplitude intensity A1 of the SPION candidates in dependence of temperature exposure (Figure 4f), an HT recording functionality readable via MPS could be realized by integrating carefully chosen amounts of these nanoparticles into the objects of interest.However, the amplitude intensity A1 in MPS of SPION candidates shown in Figure 4f is normalized on the measured powder mass and only therefore can be quantitatively compared.In most use cases, it is challenging, or even impossible when comparing different material geometries, to guarantee, that the mass of SPION candidates detected by the MPS sensor remains constant.Furthermore, the presence of diaand paramagnetic materials (e.g., the surrounding matrix) can have a contribution to the value of A1.Thus, in a more realistic application scenario, comparing the absolute amplitude intensity A1 between multiple samples or even the same sample in two different situations is prone to errors and is of very limited use.This issue can be overcome by utilizing the interaction of these different SPION candidates in a defined structure, like the subsequently presented HT-SPs.

Tuning the HT-SP Recording Sensitivity for Various Working Ranges
The emerging recording functionality of HT-SPs stems from the defined interactions of the SPION candidates within the SP structure.While the SiO 2 shell thickness of SPION candidates represents the first structural design level to finetune temperature exposure-dependent interactions and thereby customize the HT-SPs working range, the second level is the HT-SPs composition.The spray-drying forced-assembly process allows to freely combine various previously selected SPION candidates in theoretically limitless different ratios (Figure 5a), thus providing two degrees of freedom within that second structural design level.
In the following, the properties of four different HT-SPs with exemplary ratios and combinations of SPION candidates are discussed.These four HT-SPs are designed to have different temperature working ranges and recording sensitivities.To guarantee MPS readability up to 1000 °C, one high-temperature stable building block is always required.SPION-6S fulfills this purpose and is present with 15 wt.% in all four HT-SPs (Figure 5a).To address a specific temperature working range with an HT-SP, SPION candidates that oxidize at the temperature of interest, should be incorporated at a significant share in the HT-SP, as their oxidation decreases the magnetic interactions, which creates signal alteration in MPS.By selecting the other 85 wt.% to be nonprotected SPIONs (HT-SP-1), SPION-1S (HT-SP-2), and SPION-2S (HT-SP-3), the working range should be shifted from lower temperatures (HT-SP-1) to higher temperatures (HT-SP-3).More specifically, the oxidation of SPION, SPION-1S, and SPION-2S occurs ≈400, 750, and 900 °C, respectively, meaning HT-SPs containing a large share of these SPION candidates should possess a high recording sensitivity in these temperature ranges.In contrast to this, HT-SP-4 is designed to show a broader working range (600-1000 °C), realized by combining multiple SPION candidates (28 wt.% SPION, 28 wt.%SPION-1S, 29 wt.%SPION-2S, 15 wt.%SPION-6S) (Figure 5a).
The decreasing share of -Fe 2 O 3 / Fe 3 O 4 in HT-SP-1, HT-SP-2, and HT-SP-3 with increasing temperature (Figure 5b), is in good accordance with the corresponding data for SPION, SPION-1S, and SPION-2S (Figure 4b).Indeed, with increasing silica shell protection of the used SPIONs, the phase conversion is shifted from 400 °C (HT-SP-1) to 750 °C (HT-SP-2) and 900 °C (HT-SP-3).As desired, the design of HT-SP-4 leads to a more gradual decrease of -Fe 2 O 3 / Fe 3 O 4 over a broader temperature range (600-1000 °C).Similar to the direct relation of crystallographic data and magnetic properties displayed for the SPION candidates (Figure 4), also the HT-SPs indicate decreasing saturation magnetization (Figure 5c) and amplitude intensity A1 (Figure 5d) in direct correspondence to the presence of -Fe 2 O 3 / Fe 3 O 4 .However, a closer look reveals that the crystallographic and magnetic properties of the HT-SPs in dependence of temperature exposure are not just a linear combination of the properties of the individual SPION candidates that are contained.In other words, adding up the shares of -Fe 2 O 3 / Fe 3 O 4 (or the saturation magnetization, or the amplitude intensity A1), that SPION candidates show at a specific temperature, does not necessarily lead to the respective values of the HT-SP incorporating these SPION candidates.For instance, while HT-SP-1, containing 85 wt.% of nonprotected SPIONs, which oxidize ≈400 °C, exhibits the corresponding decrease in -Fe 2 O 3 / Fe 3 O 4 ≈400 °C, HT-SP-4, also containing 28 wt.% of nonprotected SPIONs, shows no decrease in -Fe 2 O 3 / Fe 3 O 4 up to 500 °C (Figure 5b-d).3]42] Since the SiO 2 -shelled SPION candidates exhibit a similar or only slightly larger hydrodynamic radius than the nonshelled SPIONs, spray-drying leads to a homogeneous distribution of the different types of SPION candidates within HT-SP-4 (Dynamic light scattering (DLS) and zetapotential measurement of the SPION candidates, as well as scanning electron microscopy (SEM) images of the HT-SPs can be found in the Figures S6 and S7, Supporting Information).Consequently, in the case of HT-SP-4, the nonprotected SPIONs are primarily surrounded by silica-shelled NPs, which provide a sort of SiO 2 matrix.Therefore, the oxidation of such arranged nonprotected SPIONs is shifted to higher temperatures, when they are incorporated into HT-SPs with a significant share of SiO 2 -shelled SPION candidates.Nonetheless, the 28 wt.% of nonprotected SPIONs in HT-SP-4 leads to a broadening of the working range toward lower temperatures compared to HT-SP-2, indicated by decreasing saturation magnetization and amplitude intensity A1 in the range of 500-700 °C (Figure 5c,d).The fact that SP structures and compositions might shift oxidation temperatures of individual SPION candidates, does not impact the toolbox-like approach of HT-SPs and their deliberate working range customization but needs to be considered while optimizing the system for specific temperature ranges.
Focusing on an application scenario in which a component has experienced a high-temperature treatment during manufacturing and nondestructive quality control is required, the amplitude intensity A1 is not suited for providing information, as knowledge of the exact mass of HT-SPs detected by the MPS sensor would be required (Figure 5d).However, HT-SPs can overcome this issue by exploiting the amplitude ratio A5/A3, which is independent of total sample mass and thus, can be self-referenced (Figure 5e,f). [18,21,23]This ratio is affected by changing magnetic interactions within their structure.To summarize this, MPS allows us to quantitatively retrieve this magnetic interaction -in this case a temperature-dependent structure-property relationand provides more information than whether magnetic material is present or oxidized.
The rise of A5/A3 of HT-SP-2, HT-SP-3, and HT-SP-4 after exposure to temperature (Figure 5e,f) occurs at similar temperatures as the decrease in -Fe 2 O 3 / Fe 3 O 4 ratio (Figure 5b), saturation magnetization (Figure 5c) and amplitude intensity A1 (Figure 5d).As previously illustrated in Section 2.2, displaying HT-SP-1 as an example (Figure 2), the oxidation of non and/ or thin-shelled SPION candidates reduces the dipole-dipole interactions of the remaining intact SPIONs, which causes A5/A3 to rise.At lower temperatures, surface effects cause A5/A3 to rise in the case of HT-SP-2 and HT-SP-4.In contrast to HT-SP-1, where the pyrolysis of citric acid, which is coordinated on the SPION surface, leads to more pronounced interactions of the nonprotected SPIONs (85 wt.% in HT-SP-1), the SiO 2 shells of the SPION candidates in HT-SP-2 and HT-SP-4 prevent this kind of increase in interaction.The fundamental reasons causing this magnetic phenomenon will require future studies as they would exceed the scope of this study but are possibly related to physicochemical alterations of the citric acid at the SPION surface.To avoid possible miss-interpretations of higher values of A5/A3 due to exposure at lower temperatures (<400 °C), HT-SPs could undergo a temperature pretreatment up to, i.e., 400 °C as a sort of formation process, before being integrated into objects of interest as HT recorder additives.
When optimizing the HT-SP in terms of sensitivity for a temperature range of interest, the nonlinear relation between the share of intact SPION and A5/A3 has to be considered, meaning that oxidation of, e.g., 30% of SPION does not cause A5/A3 to increase 30% or in another direct proportional way.This can be observed for instance in the case of HT-SP-3 after exposure to 900 °C, when 28.3 % of -Fe 2 O 3 / Fe 3 O 4 have transitioned to -Fe 2 O 3 (Figure 5b; Table S8, Supporting Information).While the saturation magnetization (Figure 5c) and the amplitude intensity A1 (Figure 5d) drop accordingly, A5/A3 does not change compared to exposure to 800 °C, after which all SPION are still intact.A similar observation can be made for HT-SP-4 at 600 °C.(X-ray diffractograms and crystallite sizes derived thereof, as well as magnetization curves for all HT-SPs after temperature exposure can be found in the supporting information Figures S8-S10 and Tables S6-S9 (Supporting Information).The crystallographic and magnetic data for all HT-SPs indicate that the oxidation of a substantial (>50 %) share of SPION candidates is required to reduce the interactions of the remaining intact SPION candidates in such a way, that A5/A3 shows a significant rise.However, at lower shares of the remaining SPION candidates, the sensitivity is enhanced.For instance, HT-SP-4 exhibits its greatest sensitivity in the temperature range 700-900 °C in which the -Fe 2 O 3 / Fe 3 O 4 content drops from 43.4 % to 8.1 % (Figure 5b; Table S9, Supporting Information).
The two degrees of freedom regarding the HT-SP composition, first, the freely selectable combinations of different SPION candidates and second, the adjustable ratios of SPION candidates, allow finetuning the recording sensitivity at any desired temperature from 400 °C up to 1000 °C.The deliberate selection of SPION candidates determines the temperature at which the HT-SP is supposed to record elapsed temperature events.The ratios of the SPION candidates can then be used to adjust the recording sensitivity in the temperature range of interest, which can be understood as the slope of A5/A3 in the plot over temperature (as in Figure 5e,f) or over time.Considering the application of HT-SPs being incorporated in two identical material components that undergo temperature treatment and being placed at two different spots in a furnace, a higher value of A5/A3 for one of the components after the furnace process would indicate higher temperatures and/ or longer durations at elevated temperatures experienced (a greater integral of temperature over time), as more SPION candidates have oxidized.
For the sake of illustrating the working principle of HT-SPs and their toolbox-like customization for various temperature ranges, the duration of temperature exposure was kept constant at 1 h with respect to the data shown in Figures 4 and 5.However, in HTs the time-temperature integral is crucial for the resulting material properties.The herein presented HT-SPs show different kinetics and varying response times at elevated temperatures, exemplary shown as a proof of concept for HT-SP-2 and HT-SP-4 at 750 °C for oven times up to 8 h in the supporting information (Figures S11 and S12, Supporting Information).This is because the oxidation of SPIONs is both, temperature and timedependent.The toolbox-like approach to designing HT-SPs offers multiple dimensions for optimizing HT-SPs toward specific time-temperature oxidation kinetics to address various application scenarios.

Conclusion
We have presented the design, working principle, and tunability of a magnetic high-temperature event recorder additive (HT-SP) in the form of micrometer-sized supraparticles (SPs), which record and memorize the thermal history of a material from 400 °C up to 1000 °C and can be read out via magnetic particle spectroscopy (MPS) within (milli)seconds.
The HT recording functionality emerges from the SP architecture, which combines different oxidation-susceptible SPIONs that are co-localized in the confined space of an SP entity.By selective oxidation of SPIONs within the HT-SP, the interaction of the remaining intact SPIONs is reduced, which provides signal contrast in MPS.The deliberate customization of the HT-SP working range is realized on two different structural design levels, that are independent from one another.The first level is the oxidation susceptibility of SPION building blocks adjusted by SiO 2 shell thickness and the second level is the freely adjustable ratios and combinations of various nanoparticle building blocks that can be combined within an SP via spray-drying.This toolboxlike approach guarantees a high level of flexibility and tunability with regard to application-specific optimization and calibration of HT-SPs.While this study focused on oxidizing conditions in air, investigating the applicability of HT-SPs for other oven atmospheres, e.g., reducing conditions, could be part of future research as altered phase transitions and thus, different magnetic phenomena can occur.
The demonstrated micrometer-sized HT recorder additive can be readily integrated into materials and stands out with all the advantages originating from its magnetic signal readout via MPS, being fast and independent of optical absorption properties enabling information readout from within dark solid objects.The utilization of high-temperature stable magnetic nanoparticle building blocks to design magnetic interaction-based recorder additives is a pioneering proof of concept toward a large variety of possible scenarios, in which recording thermal temperatures from the inside of objects is of interest.Synthesis of SPIONs: Iron oxide nanoparticles were synthesized via a well-established co-precipitation method under basic conditions, which had been reported in previous works. [6,18,19,34]FeCl 3 •6H 2 O (10.80 g, 40 mmol) and FeCl 2 •4H 2 O (3.98 g, 20 mmol) were dissolved in deionized water (125 mL) and stirred at room temperature (RT) in a beaker.In a second beaker, 180 mL of a 5 wt.% solution of ammonia was prepared.Both solutions were pumped via a peristaltic pump (Ismatec MCP, flow rate 500 mL min −1 ) and combined in a static mixer (plastic spiral bell mixer 7700924, Nordson Deutschland GmbH).Both parallel feeds had a flow rate of 500 mL min −1 each.The tube from the ammonia solution to the mixer was shorter so that the ammonia reached the mixer 1 s before the iron salt solution.The resulting black precipitate, which is SPIONs, was stirred for another 2 min and subsequently magnetically separated and washed with deionized water (250 mL) three times.For magnetic separation, a permanent magnet (Neodym, N40) was placed under the beaker.After complete sedimentation, the clear supernatant was decanted and water was added.During the first washing step, sedimentation took 30 s, 1 min for the second, and 2 min for the third.For surface capping, SPION was again magnetically separated and 500 mL of 0.05 m citric acid (CA) was added.The pH of this CA solution was adjusted with ammonia beforehand to pH 4.0.After 5 min stirring, the SPION were magnetically separated and washed with ethanol (200 mL) four times.Here, sedimentation took up to 10 min.Then, SPIONs were redispersed in water and the solid content was adjusted to 1 wt.% by adding water.The solid content was determined gravimetrically by extracting three representative samples of 2 mL each and evaporating water (Vacuum oven Memmert VO29; 110 °C, 50 mbar, 24 h).Finally, the ferrofluid was sonicated for 5 min (Branson Ultrasonic Sonifier, output: 20, duty cycle 100%).

Experimental Section
Modification of SPION with SiO 2 Shells: The Stöber-like SiO 2 shell formation was performed according to a previously reported protocol. [14,34]n aqueous dispersion of citrate functionalized SPION containing 0.7 g SPION was further diluted with water until the total solution weight was 80 g.Subsequently, ethanol (320 mL) and an aqueous ammonia solution (25 wt.%, 4.0 mL) were added and the mixture was stirred for 5 min.With continuous stirring six portions of TEOS (300, 450, 600, 750, 900, and 1050 μL) were added in 8 h intervals to obtain SPION-6S (The six represents the number of TEOS additions).Thus, for SPION-1S only 300 μL were added, 300 and 450 μL for SPION-2S, and 300, 450, 600, and 750 μL were added to obtain SPION-4S.
Afterward, ethanol was removed via rotary evaporation, and the remaining solution was purified via dialysis against water.
Fabrication of HT-SPs via Spray-Drying: The HT-SPs were produced via a spray-drying process using a lab-scale spray-dryer (Büchi Labortechnik AG, B290 mini connected to a dehumidifier B-296).For the assembly of HT-SPs containing multiple SPION candidates, the nanocomposite dispersions containing these SPION candidates were added together in the desired ratios and ultrasonicated for 2 min (Branson Ultrasonic Sonifier, output: 20, duty cycle 100%).The particles' concentration was adjusted with water to 1 wt.% and spray-dried with a 2-fluid nozzle using constant instrument parameters for all samples (inlet temperature: 130 °C, outlet temperature: ≈85 °C, pump rate: ≈0.2 L h −1 , aspirator power: 80%, N2 gas flow: 400 L h −1 ).
Heat Treatment: To simulate high-temperature events, heat treatment of the HT-SPs was performed in a muffle furnace (Nabertherm) using a rapid temperature annealing (RTA) method in the temperature range of 200-1000 °C.
Characterization: Magnetic Particle Spectroscopy (MPS) was performed with a MPS unit (PureDevices GmbH) in a sinusoidal alternating magnetic field of 20 kHz in a range of ± 23.87 kA m-1 (30 mT) at 37 °C.In the case of superparamagnetic or ferro/ ferrimagnetic samples, the magnetization response of the sample is nonlinear and the voltage signal induced from the magnetization of the sample into the measurement coil the induced signal includes amplitude intensities at higher harmonics of the excitation frequency.For more details on the physical measurement principle, please consult the supporting information (Figure S14, Supporting Information) or previous publications. [6,19]The spectra were background corrected with a reference measurement without a magnetic sample.For each measurement, 5 mg of sample (SPION candidates or HT-SPs in the form of powder) were filled into a cylindrical glass vial that fits into the MPS unit.Three different vials were prepared and each sample was measured ten times.Displayed data of absolute amplitude intensity A1 and amplitude ratio A5/A3 are the mean average values of these three samples.
For X-ray diffractometry (XRD), two different methods were applied to prepare the powders, dependent on the amount of material available for each specific sample.Powders were prepared via the front loading method into a silicon single crystal cavity sample holder if the amount was sufficient to completely fill the cavity.For samples with a more limited amount of powder, a strewn preparation using a flat silicon singlecrystal sample holder was applied.For this purpose, the sample holder was covered with a thin film of grease to enable powder adhesion.The powder was then applied through a metal sieve using a brush.If necessary, the powders were crushed with an agate mortar beforehand to render them sufficiently fine for XRD preparation.All XRD measurements were performed at a D8 diffractometer (Bruker AXS, Karlsruhe, Germany) equipped with a nine-fold sample changer.The following measurement parameters were applied: Range 15°-80°2; step size 0.011°2, integration time 1 s; radiation: Cu K  ; generator settings: 40 mA, 40 kV; divergence slit: 0.3°; sample rotation with 30 rounds per minute.Quantitative evalua-

Figure 1 .
Figure 1.Schematic working principle of high-temperature event recording magnetic SPs (HT-SPs).a) Magnetic nanoparticle toolbox containing SPIONs with silica (SiO 2 ) shells made via Stöber synthesis, that protect SPIONs from oxidation at higher temperatures (400-1000 °C) in relation to the shell thickness.Oxidized species are not detectable via MPS anymore.b) Structurally emerging functionality of SPs to record elapsed temperature events based on the purposeful oxidation of nonshelled and thin-shelled SPIONs to decrease dipole-dipole interactions.The resulting structure-property relation can be resolved in MPS in the form of changing ratios of higher harmonic magnetization amplitudes.

Figure 2 .
Figure 2. Proof of concept for magnetic high-temperature event recording SPs (HT-SPs).a) Schematic 2D cross-section of a SP containing 85 wt.% nonshelled SPIONs and 15 wt.% thick-shelled SPIONs before and after 500 °C.b) MPS spectrum of the SP depicted in a) exhibiting a higher amplitude ratio A5/A3 after heat treatment.c) Amplitude ratio A5/A3 of nonprotected SPIONs (SPION) and SPIONs with a thick SiO 2 shell (SPION-6S).No significant changes are observable for SPION-6S up to 1000 °C, whereas nonshelled SPIONs cannot be detected via MPS anymore after 450 °C.d) Amplitude ratio A5/A3 of the SP depicted in a).Emerging functionality to record the temperature exposure with a signal rise ≈450 °C.The signal drop at lower temperatures is due to surface effects including the pyrolysis of ligands.Note the weight ratio of 85 wt.% to 15 wt.% of SPION to SPION-6S is just an exemplary ratio aiming to demonstrate the concept.It is chosen because of the high share of nonprotected SPION.

Figure 3a .Figure 3 .
Figure 3. Characterization of SPIONs with tunable silica (SiO 2 ) shell thickness as exemplary candidates of a toolbox for the HT-SP design.a) Transmission electron microscopy (TEM) images of nonprotected SPIONs without SiO 2 shell a1) and with SiO 2 shells, whereby the number of SiO 2 precursor additions during the shell growth was adjusted to tune the shell thickness: one addition for SPION-1S a2), two additions for SPION-2S a3), four additions for SPION-4S a4) and six additions for SPION-6S a5) (See Experimental Section for details).b) Fourier-transform infrared spectroscopy (FT-IR) spectra of SPION toolbox candidates.c) Saturation magnetization and corresponding iron content of SPION toolbox candidates measured via vibrating sample magnetometry (VSM) and inductively coupled plasma atomic energy spectroscopy (ICP-AES).

Figure 4 .
Figure 4. Crystallographic and magnetic characterization of SPIONs with tunable silica (SiO 2 ) shell thickness as exemplary candidates of a toolbox for the HT-SP design.a) X-ray diffractograms (XRD) of SPIONs at room temperature (RT) (-Fe 2 O 3 / Fe 3 O 4 ) and after 1 h at 500 °C (-Fe 2 O 3 ), as well as SPION-6S after 1 h at 1000 °C (still -Fe 2 O 3 / Fe 3 O 4 ).b) -Fe 2 O 3 / Fe 3 O 4 content of SPION candidates determined via Rietveld refinement.c) Schematic illustration of SPION oxidation to -Fe 2 O 3 and corresponding sintering of nanoparticles, which can be delayed or prevented by SiO 2 shells.d) crystallite sizes of the crystallographic species showing that oxidation to -Fe 2 O 3 goes along with sintering and crystallite size growth.e) Saturation magnetization of SPION candidates determined via vibrating sample magnetometry (VSM) and f) absolute intensities of the first harmonic A1 in the MPS spectra, both given as magnetic moment normalized on the powder mass that was measured.

Figure 5 .
Figure 5. a) The SPION toolbox contains the different oxidation-susceptible SPION candidates that can be combined to HT-SPs via spray-drying with freely adjustable ratios and combinations.Four exemplary HT-SP configurations are presented.b) -Fe 2 O 3 / Fe 3 O 4 content of HT-SPs determined via Rietveld refinement.c) Saturation magnetization of HT-SPs determined via vibrating sample magnetometry (VSM) and d) absolute intensities of the first harmonic A1 in the MPS spectra, both given as magnetic moment normalized on the powder mass that was measured.e) Amplitude ratio A5/A3 of HT-SP-2 and HT-SP-3.f) Amplitude ratio A5/A3 of HT-SP-4.Surface phenomena (e.g., pyrolysis of ligands) affect A5/A3 at temperatures up to 550 °C.