**Light Scattering and Absorption Complementarities to Neutron Scattering: In Situ FTIR and DLS Techniques at the High-Intensity and Extended Q-Range SANS Diffractometer KWS-2**

**Livia Balacescu 1,2, Georg Brandl 2, Fumitoshi Kaneko 3, Tobias Erich Schrader <sup>2</sup> and Aurel Radulescu 2,\***


**Abstract:** Understanding soft and biological materials requires global knowledge of their microstructural features from elementary units at the nm scale up to larger complex aggregates in the micrometer range. Such a wide range of scale can be explored using the KWS-2 small-angle neutron (SANS) diffractometer. Additional information obtained by in situ complementary techniques sometimes supports the SANS analysis of systems undergoing structural modifications under external stimuli or which are stable only for short times. Observations at the local molecular level structure and conformation assists with an unambiguous interpretation of the SANS data using appropriate structural models, while monitoring of the sample condition during the SANS investigation ensures the sample stability and desired composition and chemical conditions. Thus, we equipped the KWS-2 with complementary light absorption and scattering capabilities: Fourier transform infrared (FTIR) spectroscopy can now be performed simultaneously with standard and time-resolved SANS, while in situ dynamic light scattering (DLS) became available for routine experiments, which enables the observation of either changes in the sample composition, due to sedimentation effects, or in size of morphologies, due to aggregation processes. The performance of each setup is demonstrated here using systems representative of those typically investigated on this beamline and benchmarked to studies performed offline.

**Keywords:** SANS; FTIR; DLS; semi-crystalline polymers; proteins in buffer solution

#### **1. Introduction**

Small-angle neutron scattering (SANS) is a powerful technique that yields important low-resolution structural information on macromolecular systems. The small-angle neutron diffractometer, KWS-2 [1], operated by the Jülich Centre for Neutron Science (JCNS) at the Heinz Maier-Leibnitz Zentrum (MLZ), Garching, Germany, is dedicated to the investigation of mesoscopic multi-scale structures and structural changes in soft condensed matter and biophysical systems. Following demands from the user community, it was recently considerably upgraded to boost its performance with respect to the intensity on the sample, instrumental resolution, counting rate capabilities, and maximum Q-range (Q = (4π/λ)sinθ is the momentum transfer, where λ is the neutron wavelength and 2θ is the scattering angle). The instrument, with high-intensity, extended Q-range and tunable resolution, is optimized for the study of mesoscopic structures and structural changes due to rapid kinetics.

The ability to observe chemical and physical processes simultaneously with smallangle scattering (SAS) is highly desirable for a deeper understanding of biological and soft

**Citation:** Balacescu, L.; Brandl, G.; Kaneko, F.; Schrader, T.E.; Radulescu, A. Light Scattering and Absorption Complementarities to Neutron Scattering: In Situ FTIR and DLS Techniques at the High-Intensity and Extended Q-Range SANS Diffractometer KWS-2. *Appl. Sci.* **2021**, *11*, 5135. https://doi.org/ 10.3390/app11115135

Academic Editor: Antonino Pietropaolo

Received: 30 April 2021 Accepted: 29 May 2021 Published: 31 May 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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 (https:// creativecommons.org/licenses/by/ 4.0/).

matter systems. This is particularly crucial for dynamic systems, to ensure sample stability, purity, chemical conditions, to reduce ambiguities in data reduction, and for when it is not possible to perform further measurements or characterization on intermediate states ex situ. Given the high brilliance of synchrotron radiation, simultaneous analyses by X-ray and differential scanning calorimetry (DSC) or Raman spectroscopy were developed quite early at the Synchrotron Radiation Source in Daresbury, U.K. [2,3], and used successfully since then [4,5]. More recently, in situ size-exclusion chromatography (SEC) with SAXS [6] was developed for the first time at the Advanced Photon Source of Argonne National Laboratory (ANL, Lemont, IL, USA). This brought advantages in the field of protein scattering, in which samples very often typically show poor stability or aggregation tendency.

As unique probes, neutrons interact differently with the 1H and 2H (deuterium, D) hydrogen isotopes: the large difference in the coherent scattering length density between H and D represents the basis of the contrast variation and contrast matching methods. As most of the soft matter and biological samples consist of hydrocarbon systems, the H/D substitution offers the possibility to vary the coherent scattering length density of a compound over a broad range. With this technique, selected constituents in a complex multi-component system can be labeled by isotope exchange. Depending on the created contrast—the squared difference between the scattering length density of one component and that of the other components or the environment medium—selected components or regions within a complex soft matter or biophysical morphology can be made visible or invisible in the scattering experiment without the chemical alteration of the system.

Owing to technological developments in high-intensity specialized SANS instruments such as detectors and neutron guides [7,8], exciting developments in the nature and number of characterization techniques that can be performed in situ on large-scale SANS instruments started to emerge. For example, the first implementation of an in situ SEC-SANS system was reported with a D22 instrument at the Institute Laue-Langevin (ILL, Grenoble, France) [9], while first results obtained with a simultaneous SANS and FTIR measuring system were collected with the KWS-2 instrument [10]. In situ DLS setups have also been developed to complement SANS at several neutron beamlines so far, for example a SANS-I instrument at the SINQ spallation source (Paul-Scherrer Institute, PSI, Villigen, Switzerland) [11], D11 instrument at the Institute Laue-Langevin (ILL, Grenoble, France) [12,13], and LOQ and ZOOM instruments at the ISIS Pulsed Neutron and Muon Source (Rutherford Appleton Laboratory, Didcot, UK) [14].

Therefore, the simultaneous combination of light scattering or absorption methods with SANS (or WANS) represents very powerful experimental approaches to be used in the microstructural characterization of soft and biological systems.

In this paper we communicate recent upgrades to the sample environments available using the KWS-2, which is now equipped to perform in situ FTIR and dynamic light scattering (DLS). All of which can be performed simultaneously with standard or real-time SANS and enable in-beam monitoring of the sample quality or provide complementary information to neutrons.

#### **2. Materials and Methods**

Small-angle neutron scattering (SANS) experiments were performed using the KWS-2 instrument at the JCNS at MLZ in Garching, Germany [1]. The measurements were typically performed using sample-to-detector distances of 1.5, 8, and 20 m, and neutron λ of 5 and 10 Å (Δλ/<sup>λ</sup> = 10%), to achieve a maximum dynamic *<sup>q</sup>*-range of 1 × <sup>10</sup><sup>−</sup>3–0.7 Å−1. Scattering data was collected in static and real-time modes. Each raw scattering data set was corrected for detector sensitivity, electronics background, and empty cell contribution and converted to scattering cross-section data using instrument software QtiKWS [15]. Data was then converted to the absolute scale (cm<sup>−</sup>1) through reference to the scattering from a secondary standard sample (Plexiglass).

Hydrogenated polyethylene glycol dimethyl ether (hPEGDME, MW = 20,000 g mol−<sup>1</sup> with a polydispersity of D = 1.01) that was synthesized in house [16] was dissolved in

deuterated tetrahydrofuran (dTHF, Sigma-Aldrich Chemie GmbH, Munich, Germany) by weighing the components in appropriate amounts to ensure a polymer volume fraction (vol/vol%) φpol = 2% in the final solution. The solution was first heated to 50 ◦C under stirring, to achieve a homogeneous distribution of polymer chains in solution and then transferred into a sample container with ZnSe windows (Laser components GmbH, Olching, Germany), which is appropriate for simultaneous FTIR and SANS measurements. The sealed sample container with a beam path length of 1 mm was installed in a Peltiercontrolled holder of own production that can enable the variation of temperature on the sample between 5 ◦C and 130 ◦C in a controlled way. To determine the crystallization temperature of PEGDME from the dTHF solution in decreasing temperature, the system was investigated by differential scanning calorimetry (DSC) using a DSC131 EVO Calorimeter (SETARAM Instrumentation, KEP Technologies EMEA, Kirchheim unter Teck, Germany) prior to the simultaneous FTIR and SANS analysis.

Uniaxially oriented amorphous deuterated syndiotactic polystyrene (d-sPS) films, about 50 μm thick, were prepared by quenching a melt of d-sPS in an ice water bath, drawing the melt-quenched film four times at 373 K, and clipping well-oriented portions from the drawn films [17]. d-sPS co-crystal films containing protonated toluene as guest (d-sPS/hTol) were obtained by exposing the oriented amorphous films to a vapor of hTol (Sigma-Aldrich Chemie GmbH, Munich, Germany). The d-sPS/hTol co-crystalline films were transferred into the sample container with ZnSe windows for simultaneous FTIR and SANS analysis at room temperature.

In situ FTIR investigations were carried out during the SANS measurements with a JASCO VIR-200 spectrometer (JASCO Deutschland GmbH, Pfungstadt, Germany) installed at the sample position of the KWS-2 in an appropriate geometry that enabled a simultaneous irradiation of the sample using IR and neutron coaxial beams. Polarized IR beams were obtained by using a precision automated PIKE polarizer KRS-5 (PIKE Technologies, Madison, WI, USA).

For the in situ DLS measurements, a 15 mg/mL solution of apomyoglobin (apoMb) molten globule was prepared. ApoMb was obtained from horse heart myoglobin (Sigma-Aldrich Chemie GmbH, Munich, Germany) by the butanone method (as performed in [18]) and then refolded by dialysis in 20 mM NaH2PO4/Na2HPO4 (Sigma-Aldrich Chemie GmbH, Munich, Germany) pH 7 buffer. Before storage in the freezer at −20 ◦C, the solution was lyophilized. To replace the exchangeable protons by deuterium, the freezedried apoMb powder was dissolved in heavy water (D2O, 99.9% D, Sigma-Aldrich Chemie GmbH, Munich, Germany), incubated for 1 day, and lyophilized again. To obtain the molten globule state of apoMb the powder was dissolved in D2O and centrifuged to remove the large aggregates. In the supernate solution of concentration 2 mg/mL and pD 6, DCl 0.1 M (Sigma-Aldrich Chemie GmbH, Munich, Germany) was added until the pH value was 3.6 (monitored by pH meter Methrom), corresponding to a pD value of 4. The buffer exchanged protein solution was centrifuged (Heraeus Instruments) to the final concentration of 10 mg/mL using Vivaspin 3000 MWCO concentration units (Sartorius, Göttingen, Germany). The protein concentration was determined in a 0.1 mm thick cell (Hellma, Germany), using the molar extinction coefficient (280 nm = 13,980 M−<sup>1</sup> cm−1) which was calculated from the amino acid sequence [19].

All samples were prepared in deuterium solvents or using deuterated polymer films to provide low incoherent neutron background and enable appropriate scattering contrasts for neutrons.

#### **3. Results**

In the following sections we will present results and discussion of the proof-ofprinciple tests on the in situ FTIR and DLS setups performed using the KWS-2 simultaneously with the SANS characterization of samples.

#### *3.1. In Situ FTIR*

FTIR allows for the determination of the chemical and physical structure of material via interpretation of the spectrum of characteristic bands that is produced by the excitation vibrations of molecular bonds following absorption of IR photons in the material [20]. FTIR is particularly powerful in the analysis of the polymeric systems under different environmental conditions. The chemical identity of the chains is achieved by detecting the chemical groups present in the polymer, the monomer sequences, and the stereoregularity. Additionally, one can observe the local conformation of individual chains as well as conformational change when macroscopic fields such as temperature, pressure, and relative humidity are applied to the sample. Due to these advantages, a combination of FTIR and SANS represents an outstanding investigation approach for understanding the detailed mechanism of structure formation and structural changes in multiphase polymeric systems such as semi-crystalline polymers, copolymers, or polymer composites: the conformational change of the molecules that make up such a complex system can be monitored by FTIR, while the hierarchy of morphologies occurring in different phases can be qualitatively and quantitatively characterized by SANS with contrast variation. Polymeric materials are very sensitive to slight variations of external conditions; therefore, it is more useful to apply the combination of these techniques in a simultaneous approach, when in situ or time-resolved analyses of rapid structural changes or irreversible processes can be performed.

We previously conducted pioneering work in developing a simultaneous SANS and FTIR spectroscopy measuring system using the KWS-2 diffractometer in 2014. Using this approach we studied the structure of a d-sPS co-crystalline film with various guest molecules [10]: higher-order crystalline structures were characterized by SANS while the local molecular configuration and molecular packing were characterized by FTIR. We demonstrated that the FTIR data were useful for interpreting and analyzing SANS profiles when larger molecules are included in the sPS films [21]. To extend the measurable Q-range for wide-angle neutron scattering (WANS), we used this approach on the time-of-flight SANS instrument TAIKAN installed at the Material and Life Science Experimental Facility (MLF), Japan Proton Accelerator Research Complex (J-PARC, Tokai, Japan) [22,23]. The old measurement system tested with the KWS-2 consisted of a portable FTIR spectrometer Perkin-Elmer Spectrum Two (PerkinElmer, Rodgau, Germany) and an optical system made up of six mirrors. For an easier optical path adjustment, higher light utilization efficiency, and proper use of the available space on the sample stage of the KWS-2, and for the installation of a device for the measurement of polarization-dependent infrared absorption and a more complex sample environment, the optical system including the FTIR spectrometer was redesigned, using the experience accumulated in [22]. A new FTIR spectrometer (JASCO VIR-200, JASCO Deutschland GmbH, Pfungstadt, Germany) is currently in regular use for conducting FTIR measurements in transmission geometry, with the IR beam going through the sample coaxially with the neutron beam (Figure 1): the two optical ports in the same lateral face, the outlet port for emitting parallel light and the inlet port for the incoming light, enable a great flexibility in designing the optical system and accommodating special sample environments (Figure 1b,c). At the same time, this geometry is paving the way for the installation of an in situ ATR-FTIR option that can be used simultaneously with SANS in the near future for an optimal investigation of liquid samples. In the current geometry, only the two Al-coated quartz plates (LASER COMPONENTS GmbH, Olching, Germany) are left (elements 3 and 4 in the scheme of Figure 1a).

**Figure 1.** Schematic presentation (**a**) and detailed (**b**) and general (**c**) views of the new operational setup for in situ FTIR analysis simultaneous with the SANS investigation using the KWS-2. The numbers indicate the main components, which can be identified in the photos on the (**b**,**c**) panels: the outlet and inlet beam ports (1,2), the windows (3,4), and the new ancillary equipment (5) for the control of the temperature on the sample (in-house designed and produced Peltier-based thermostated cuvette holder), while P indicates the IR polarizer.

Figure 1b,c shows photos of the new setup for simultaneous FTIR and SANS analysis using the KWS-2: the components indicated by numbers in the scheme in Figure 1a are identified in the photographs, which show the complete experimental arrangement at the sample position of the KWS-2 between the "collimation nose" that holds the sample aperture on the left side of photos, and the entrance window of the evacuated SANS detector tank, which is visible on the right side of the images. The new FTIR setup is entirely built on a breadboard (Thorlabs GmbH, Bergkirchen, Germany) which is equipped with "zero-point" fixation systems (AMF GmbH, Fellbach, Germany) for pneumatic clamping (activated with pressurized air) for an easy and precise installation on the sample stage of the KWS-2.

The control of the FTIR spectrometer, IR polarizer position (in beam and out of beam) and orientation were implemented in the NICOS instrument control software of the KWS-2 [24]. This is mainly meant for time-resolved investigations when FTIR measurements with the selected time acquisition can be carried out simultaneously with SANS measurements on samples in different external fields or composition conditions. The start of the FTIR acquisition is triggered by the detector start of the SANS instrument and runs until the SANS acquisition time is consumed. All relevant parameters for SANS: wavelength λ, collimation length LC, detection length LD, beam-size on sample, etc. [15], sample field parameters (temperature), and the FTIR setup parameters (polarizer position and orientation, etc.) can be combined in a single complex experimental routine. For conventional SANS measurements, standard FTIR acquisitions can be carried out in parallel using the operation menu of the FTIR spectrometer. A precise synchronization of the results obtained with the two methods is possible by using the SANS data acquisition in list mode [8], which enables the assignment of the SANS pattern collected at a specific time to the corresponding FTIR spectrum.

Proof-of-principle measurements were carried out by simultaneous use of FTIR in transmission geometry and SANS on soft matter systems that are relevant for the research topics developed using the KWS-2: (i) a hPEGDME solution in dTHF (φpol = 2%) at 30 ◦C and 12 ◦C and (ii) a d-sPS film in co-crystalline δ-form with hTol as guest molecules in the cavities between the sPS helices.

#### 3.1.1. FTIR Sample Holder and Neutron Background

Figure 2 shows a technical drawing of the sample cells designed for the liquid and film samples for the simultaneous FTIR and SANS setup using the KWS-2. The cell is equipped with two ZnSe windows, each with a thickness of 1 mm, which enable a beam path of either 1 mm or 2 mm. A shorter beam path can be obtained by placing an additional ZnSe window with a thickness of 0.5 mm and an appropriate diameter inside of the holder. The chosen geometry allows for an easy filling and a tight closing of cells using the series of lateral screws, appropriate O-rings, and Teflon screws for the top openings used for filling in the samples. Tests of the holders have shown that they remain tight up to a temperature of at least 120 ◦C.

**Figure 2.** Technical drawing of the new sample cuvette equipped with ZnSe windows, which is used for the simultaneous FTIR and SANS investigation of samples using the KWS-2. The liquid samples can be filled in by using the top openings, while the black O-rings, the series of screws (violet), and Teflon screws for the top openings provide the cuvette tightness. Cuvettes with 1 mm and 2 mm sample thickness are provided; the sample thickness can be further decreased to 0.5 mm or 0.2 mm by using additional ZnSe windows of appropriate thickness and diameter that can be placed inside of the cuvette, between the yellow windows shown here.

The background yielded by the set of Al-coated mirrors and ZnSe windows in SANS is low (Figure 3) and, despite the weak Bragg reflexes from the ZnSe material, which can be observed in the high-Q scattering conditions (the small spots indicated by arrows in Figure 3c), very good background subtraction for SANS data of high quality can be achieved.

**Figure 3.** The two-dimensional SANS patterns corrected for detector sensitivity collected at LD = 1.5 m from: empty beam (**a**), when the two Al-coated quartz windows (items 3 and 4 in Figure 1a) are added (**b**), and when the two ZnSe windows are added (**c**); data from the same setup as in (**c**) but collected at LD = 20 m are shown in (**d**). The yellow arrows indicate the Bragg reflections from the ZnSe material.

#### 3.1.2. Simultaneous FTIR and SANS on PEGDME in Solution

Unlike crystallization from bulk or melt, a variety of processes, such as the formation of polymer and solvent compounds, a physical gelation, or a liquid–liquid phase separation, may take place when polymers crystallize from solution, depending on the solvent type and polymer concentration and molecular weight [25].

PEGDME is a semi-crystalline polymer, soluble in water and compatible with most organic solvents. When PEGDME crystallizes from solution, crystallites grow as aggregates of small layers of lamellae with branching and splitting. The morphology of this hierarchical organization of structures over a wide length scale, between nm and μm, can be resolved by SANS. In the crystalline state, PEGDME takes a regular helical conformation, which can be regarded approximately as a uniform (7/2) helix [26,27]. When taking a long helical structure, PEGDME chains exhibit some sharp conformational regularity IR bands, such as the 1345 cm−<sup>1</sup> A2 band [28,29]. Thus, the combination between SANS and in situ FTIR is a powerful method for the study of the kinetics and morphology of polymer crystallization from solution.

In Figure 4 we report the results of the DSC scan taken under nitrogen atmosphere between 50 ◦C and 10 ◦C from the solution of the PEGDME in dTHF with a cooling rate of 5 ◦C/min. The polymer crystallization occurs at about TC = 26–27 ◦C, with two peaks observed in the heat flow curve, an effect similar to that observed for the bulk crystallization of PEG [30]. Therefore, in order to investigate the change in conformation of PEGDME chains and the formation of polymer crystalline morphology at mesoscopic scale when passing through the TC, we performed the simultaneous FTIR and SANS analysis of the hPEGDME in dTHF solution at 30 ◦C, above the TC, followed by a rapid cooling to 12 ◦C, below the TC.

**Figure 4.** The thermal behavior of hPEGDME (MW = 20,000 g mol<sup>−</sup>1) in solution of dTHF (φpol = 2%) during the crystallization process at a rate of 5 ◦C/min.

Figure 5 shows a selected range of the infrared spectra from the hPEGDME in dTHF at 30 ◦C (blue line) and at 12 ◦C (with the cyan and red lines indicating the initial and final stages, and the other lines in between depicting the time evolution of the spectrum at this temperature) in parallel with the spectrum of dTHF taken at 30 ◦C. Upon reaching the temperature of 12 ◦C, infrared spectra from the sample were collected every 30 s within a time interval of 35 min, to follow the kinetics for isothermal crystallization of polymer in solution. The time-resolved analysis clearly shows the evolution and transformation of the band at 1352 cm<sup>−</sup>1, which is characteristic of amorphous polymer structure (single coil in solution) and well visible in the initial stage at 12 ◦C, into the bands at 1345 cm−<sup>1</sup> and 1364 cm−1, which are characteristic of the helical conformation of polymer chain in the crystalline state and well visible in the final stage at 12 ◦C. All other bands, corresponding to either the amorphous conformation in the initial stage or the helical conformation in the final stage, correspond to frequencies reported for the PEG system [31,32].

**Figure 5.** FTIR spectral change observed during the crystallization of hPEGDME (MW = 20,000 g mol<sup>−</sup>1) from solution in dTHF in decreasing temperature from 30 ◦C to 12 ◦C. The arrow indicates the evolution of the FTIR spectra collected between the initial and final states at 12 ◦C with a time step of 30 s.

The polymer morphology in solution at 30 ◦C and 12 ◦C in the initial and final stages of crystallization was investigated by SANS in parallel with the FTIR analysis. Figure 6 reports the corrected and calibrated SANS results collected simultaneously with the acquisition of the infrared spectra. The power law exponents (Q−*p*) specific for different polymer morphologies are given. Hence, swollen single coils (*p* = 5/3), one-dimensional morphologies or Kuhn segments (*p* = 1), lamellar (*p* = 2), and large 3D morphologies with sharp interfaces (*p* = 4) can be directly identified and characterized [33]. The model curves of the semi-flexible single coil and the crystalline lamellar stacks morphologies described later are also depicted.

At 30 ◦C all polymers are dissolved as semi-flexible coils in solution, as shown by the single-chain form factor features identified in the scattering profiles: the plateau towards low Q yields information about the volume fraction and molecular weight while the bending down of the intensity is characteristic of the Guinier regime, from which information about the radius of gyration of the polymer coil is obtained. The Q−5/3 power law regime towards high Q is indicative of excluded volume interactions between the chain segments (with −5/3—Flory's exponent). A transition to Q−<sup>1</sup> behavior is observed at higher Q, which indicates that short stiff (linear) polymer segments with a persistence length lP = *b*/2, with *b* the Kuhn segment, exist at a local length scale [34]. The small-angle scattering pattern at 30 ◦C was fitted using the unified Beaucage model [35] for multiple structural levels in hierarchical morphologies. The approach described the experimental data well and delivered the sizes involved in each structural level: the Rg = 37 Å of the polymer coil and the persistence length lP = 7.5 Å. Thus, both FTIR and SANS results are in good agreement regarding the polymer conformation at 30 ◦C as semi-flexible coil in solution.

**Figure 6.** SANS cross sections from solution of hPEGDME (MW = 20,000 g mol<sup>−</sup>1) in dTHF (φpol = 2%) measured at different temperatures, during the FTIR analysis (Figure 7). Data collected at different LD and temperatures are differentiated by symbols and colors as shown in the legend. The solid lines indicate the power law behavior in different Q ranges, whereas the red curves represent the model description of the experimental data, as discussed in the text. The dotted curve represents the scattering pattern from the lamellar stacks, as resulted from the global fit of the experimental data collected in the final state at 12 ◦C (green symbols).

**Figure 7.** *Cont.*

**Figure 7.** FTIR spectra measured with polarized IR beam parallel (blue line) and perpendicular (red line) to the uniaxially deformation direction of the d-sPS co-crystalline film with hTol: whole spectra (**top**) and detail (**bottom**) are shown.

At 12 ◦C, due to methodical reasons, we used the typical procedure for carrying out experiments over an extended Q-range by combining different LD's rather than the time-resolved measurements on a selected narrow Q-range. Thus, the raw data were collected for one minute at two detection distances, LD = 8 m and 20 m (always in this order), with the detector movement from one position to another taking 10 min. Two sets of measurements were performed at 12 ◦C, starting at the time when the set temperature was attained by the Peltier-based variable temperature cuvette holder, which corresponds to the acquisition start of the cyan IR spectrum in Figure 5.

A decrease in the temperature resulted in the formation of polymer crystalline morphology that led to an increase of the scattering level towards low Q compared with the single chain scattering feature, as observed in the SANS results corresponding to the initial state at 12 ◦C (blue symbols in Figure 6). As observed in the FTIR spectra, a coexistence between an amorphous conformation of PEGDME, on one hand, and a slow increase in number density of crystallites by nucleation and growth, on the other hand, is observed, which explains the SANS pattern.

The scattering data collected in the end state at 12 ◦C indicate the formation at mesoscale of a morphology made of crystalline lamellae alternating with amorphous interlamellar regions, as denoted by the peak-like feature which indicates an interlamellar domain spacing of D\* = 2π/Q\*: where Q\* represents the peak position, and the steep increase of intensity towards low Q, due to association of the lamellar stacks in bundles and branches, which led to the formation of a large scale spherulite morphology.

The scattering features characteristic of such morphology appear at much lower Q values than the range covered in pinhole mode with the KWS-2 [1] and in the current experiment only the high-Q asymptotic behavior of the scattering from the polymer spherulites could be observed (green symbols, Q−<sup>4</sup> power law behavior). A fit with the model combining the form factor of individual lamellar slabs with the paracrystalline structure factor describing the stacking effects over distances larger than the lamellar thickness [33] successfully described the experimental SANS data and delivered the interlamellar distance D\* = 110 Å and its large smearing σD\* = 30 Å. A power law term with the exponent *p* = 4, as discussed above, was considered for a proper description of the data at low Q, which far exceeds the scattering from the polymer lamellae (depicted as a red dotted curve in Figure 6, based on the fitted parameters). A value for the lateral extension of lamellae

(assumed as discs) much larger than their thickness was assumed as fixed parameter (RL = 500 Å), as long as this size level cannot be observed towards low Q, while a fixed lamellar thickness *d* = 45 Å [17] was considered in the fitting procedure, since no scattering data were collected at high Q at 12 ◦C. The lateral extension of the lamellae does not affect the fitting of the D\* and σD\* parameters, as long as it is supposed to be much larger than the lateral size of the lamellar stacks. A more detailed model interpretation of the SANS data in terms of "forward scattering" and scattering length density parameters for different regions of the complex morphology [33] is beyond the goal of this report. The SANS results indicate that crystalline lamellar stacks, which form larger objects (spherulites), are the dominating morphology in the sample and are in very good agreement with the FTIR observations, which indicate that PEGDME chains are mainly in helical conformation, thus crystalline state, in these conditions.

#### 3.1.3. Polarized FTIR Measurements

In Figure 7 we report polarized FTIR spectra measured on a d-sPS/dTol co-crystalline uniaxially deformed film with the polarized IR radiation parallel (blue line) and perpendicular (red line) to the drawing direction of the film. The FTIR analysis was done in this case independently to the SANS investigation and was meant to prove the performance of the IR polarizer using well-known samples [21]. SANS measurements on similar systems and guest exchange process were reported elsewhere [36]. The d-sPS/hTol-oriented film exhibits clear polarization in the 3100–2800 cm−<sup>1</sup> region, which suggests that toluene guest molecules are kept oriented in the crystalline region of the d-sPS, as they are located between the oriented helices of the d-sPS host system. The hTol provides IR bands due to benzene ring C–H stretch modes in 3100–3000 cm−1, while its methyl group shows two bands due to antisymmetric stretch modes at about 2945 and 2920 cm−<sup>1</sup> and a band due to symmetric stretch mode at about 2860 cm<sup>−</sup>1.

#### *3.2. In Situ Dynamic Light Scattering*

DLS, also known as photon-correlation spectroscopy (PCS) or quasi-elastic light scattering spectroscopy (QELSS), is based on the temporal analysis of the intensity fluctuations of the scattered light caused by the Brownian motion of particles (protein molecules, aggregates, polymer particles, etc.) in solution. The velocity of the Brownian motion is defined by a property known as the translational diffusion coefficient (*D*). *D* is related to the average hydrodynamic radius of the particles (*RH*) through the Stokes–Einstein equation [37]:

$$D = \frac{k\_B T}{6\pi\eta R\_H}$$

where *kB* is the Boltzmann constant, *T* is the temperature, and *η* is the viscosity of the medium. The radius obtained by this technique is the radius of a sphere that has the same diffusion coefficient as the particle. If several particles are present in the solution, they can be easily distinguished if they are at least one order of magnitude different in size. This is particularly useful for samples, which show a tendency to aggregate over time, such as biological samples, which are sensitive to slight changes of temperature, pH of the solvent, radiation damage, etc. Since SANS measurements offering structural information on the nanometer scale often require integration times of hours, monitoring the sample composition on the minute timescale with DLS provides an additional insight into the aggregation state of the sample. This is why in situ DLS has been developed and used at many SANS beamlines so far [11–14].

A DLS measurement requires a highly monochromatic light beam, for example, produced by a laser illuminating an area of particles within a solution. The refractive index jump between those particles and the surrounding solution causes the light to be scattered. The particle size (and magnitude of Brownian motion) will affect the rate at which the intensity of this scattered light fluctuates. A single photon counting detector converts the scattered photons into standardized electric pulses. Those are fed into a digital autocorrelator which sorts the photons into time bins with respect to the time (*t*) when a first photon arrived at the detector and time (*t* + δ*t*), when the next photon arrived on the detector. Thereby it constructs a correlation function (gII(*t*)). For larger particles, the scattered light changes slowly and the correlation persists for a long time. In contrast, if the particles are small and move rapidly then the correlation function decays more quickly.

#### 3.2.1. DLS Setup

Similar to the setup developed by Heigl et al. [12], the DLS setup using the KWS-2 is built modularly, consisting of five major components: light source, optics defining the scattering geometry, detector, correlator, and computer. The optics defining the scattering geometry are placed at the sample position on the instrument, whereas the other components are mobile connected with optical fibers (Figure 8). The light from a He–Ne laser (20 mW, 632.8 nm) is coupled to a fiber (Schäfter + Kirchhoff GmbH, Hamburg, Germany) and brought to the neutron instrument. A collimator (Schäfter + Kirchhoff GmbH, Hamburg, Germany) and a mirror focus the light from the fiber into the center of the quartz cell, which was previously aligned in the center of the neutron beam. The scattered light under the angle ϑ (in the present case ϑ = 120◦) is then guided by a mirror and collimated into a second fiber. The optics are mounted using standard optic holders (Thorlabs) and placed on a platform anchored on the vacuum tank of the neutron detector. This concept allows the use of a variety of sample changers with the existing optics. For alignment purposes, one can mark a spot on the quartz cuvette in the sample holder, where the laser hits the cell surface. After disconnecting the laser fiber and reconnecting it to the collimator for the scattered light, one can see and align the light path for the scattered light to the same sample spot marked in the quartz cell. This alignment is often good enough to obtain a first correlation curve which can then be optimized to a better intercept using nanoparticle solutions as highly scattering test samples. This alignment concept relies on the fact that both the laser light and the scattered light are guided through a single mode optical fiber, a specialty of this realization of in situ DLS.

**Figure 8. Left**: Schematic view of the setup for simultaneous DLS and SANS analysis of samples using the KWS-2. **Right**: view of the setup on the sample stage of the neutron SANS diffractometer, between the sample aperture of the collimation system and the entrance window of the evacuated detector tank. Light beam trajectories to and from the sample are indicated in red. The sample cuvettes are placed in the locations of a Peltier-based multi-position variable temperature holder (Quantum Northwest Inc., Liberty Lake, WA, USA) that can move each sample in neutron/light beams in a controlled way.

After passing the mirror, the collimator, and the single mode fiber, the scattered photons arrive at an APD detector (Excelitas Technologies Corp. SPCM-AQRH-FC, Waltham, MA, USA). The recorded electrical pulses are autocorrelated with an ALV-7004 Multiple Tau Digital Correlator (ALV-GmbH, Langen, Germany). A computer displays the measured intensities and the autocorrelation functions. The correlation function and the recorded intensities are saved for further analysis. To protect the detection components, a filter wheel with neutral density filters is placed before coupling the laser light into the fiber. Based on the photon counts detected for each sample, the wheel adjusts to the optimum filter for each measurement.

A DLS measurement begins shortly after the start of the neutron measurement, which has been triggered by the instrument control software NICOS. Light scattering data are recorded in parallel and saved on a separate computer with a unique identifier containing the run number of the neutron measurement. The correlator determines the normalized intensity–intensity (second-order) auto-correlation function gII(*τ*). Data processing is performed post-experiment, by obtaining the electrical field first-order correlation function gI (*τ*) from gII(*τ*), according to the Siegert relation. gI (*τ*) is interpreted using a tripleexponential fit, corresponding to a triple-modal distribution [38]. For samples with higher polydispersity, a CONTIN fit routine can be used [39].

#### 3.2.2. Proof of Concept: Investigation of Silica Nanospheres

Silica nanospheres (20 nm, NanoXactTM, nanoComposix, San Diego, CA, USA) were investigated as a proof of concept. The water-based solution with a reported concentration of 5 mg mL−<sup>1</sup> was used as received. According to the producer, these nanospheres are sensitive to lower temperatures, leading to aggregation. Their diameter is 21.4 ± 2.7 nm, as determined by JEOL 1010 transmission electron microscope gravimetric analysis. In terms of absorption characteristics, their optical density at the 632.8 nm wavelength of the DLS-laser is lower than 0.005, meaning the light absorption by the silica nanosphere at this wavelength is negligible. The SANS data shown in Figure 9 confirms the silica nanosphere radius (Rns = 11.9 nm with polydispersity σ = 12%) and is not altered by the formation of any aggregates. On the other hand, the DLS investigations performed in parallel reveal that there are also larger structures in the solution of the silica nanospheres. The intensity of the first-order correlation function gI (*τ*) can be attributed as follows: 18% monomer (with a signal at 0.48 ms, corresponding to an average hydrodynamic radius of 13.3 nm) and 63% to aggregates (average RH of 185 nm). Less than 15% of the intensity can be attributed to larger-sized aggregates (RH ~2 μm) or can be assumed to occur because of noise. Given that the scattering intensity strongly depends on the particle's volume, the percentage corresponds to 98% monomer and 2% aggregates. The reason for the presence of aggregates can be assumed as a thermal mistreatment of the sample due to temperature variation.

**Figure 9. Left**: the electrical field first-order correlation function gI (*τ*) obtained from the DLS measurement of the solution of silica nanospheres (with nominal diameter of 20 nm). The red line is a triple-exponential fit, identifying two decay times: 0.48 ms and 6.64 ms with considerable amplitudes and a third one with a comparably small amplitude which will be attributed to noise here. **Right**: the SANS cross section of the same solution (open symbols). The red line is the data fit based on the spherical form factor model (Rns = 11.9 nm) with polydispersity in size (σ = 12%) and instrument resolution considered [15]. Although the SANS curve does not signal the presence of aggregates in the solutions, the DLS does.

3.2.3. A Typical Sample: Apomyoglobin

Apomyoglobin in the molten globule state was investigated by SANS and in situ DLS. Because the protein is not in its native-alike state and the sample concentration is significantly higher than in the living organism, this system is prone to aggregation. Even slight changes in temperature and pH of the environment can lead to sudden aggregation. To ensure that the sample does not change during the SANS experiment, in situ DLS was employed. Every 5th minute during the hours-long SANS experiment, a DLS measurement was performed.

The first and the last measurement are shown in Figure 10 and one can see that the sample did not change. The first-order correlation function was evaluated using the CONTIN algorithm [39] and the particle-size distribution according to the scattering intensity can be seen in Figure 11. This confirms that the molten globule apomyoglobin solution did not form more aggregates during the SANS experiment and the amount of aggregates present already at the start of the SANS experiment could be determined. Beyond the RH~2 nm monomers, larger structures of RH ~100 nm are present in the solution. Signals of particles of RH > 10 μm which are attributed to noise can also be identified. Nevertheless, based on the fact that the scattering intensity is strongly dependent on the particle's volume, the largest number of particles in the solution is represented by monomers and the second largest by the RH ~100 nm aggregates.

**Figure 10.** The electrical field first-order correlation functions of the first and the last DLS measurements performed on the apomyoglobin sample during the neutron beam time. Although larger structures were permanently present in the sample, the two curves nearly coincide. This means that the aggregation state of the sample did not change during the long integration time needed for the SANS measurement. The red lines mark the fit to the data using the CONTIN algorithm.

**Figure 11. Left**: The derived size distribution obtained from the measurements presented in Figure 12. **Right**: close-up of the smaller particles in the solution; besides the monomers with RH ~2 nm, larger structures of RH ~110 nm are present in the solution. The black line corresponds to the distributions obtained from the first DLS measurement, whereas the blue line indicates the results of the last measurement.

**Figure 12.** SANS cross sections of the 15 mg/mL apomyoglobin molten globule solution. The blue line marks the simulation for the monomer scattering intensity based on the polymer with excluded volume model (Rg = 2.5 nm, ν = 0.46). The green line is the simulated scattering curve for a larger structure. The red line is obtained by adding the green and blue lines and describes the data well.

Figure 12 shows the corresponding SANS curve to the in situ DLS data discussed above. The scattering profile of the monomeric protein apomyoglobin is represented by the blue line and was determined based on the polymer with excluded volume model [40] and previous knowledge about the protein: excluded volume parameter ν = 0.46 and Rg = 2.5 nm [41].

The green line shows the scattering contribution from the aggregates with RH = 150 nm (which can be roughly approximated to Rg = 116 nm, according to the ratio Rg/RH = (3/5)0.5 [42]). Since for such large morphologies the Guinier regime occurs at lower Q values than the Q-range of the current SANS investigation, this contribution mostly resembles the power law regime of the Beaucage model with the Porod exponent *p* = 5/3 [35]. The red line is obtained by adding the blue and green lines, with an appropriate power law scaling factor at low Q, to describe the data well. Using this additional information from the in situ DLS measurement, one can now interpret the SANS curve better.

#### **4. Conclusions**

We reported here the concept and proof-of-principle tests of the in situ Fourier transform infrared (FTIR) spectroscopy and dynamic light scattering (DLS) analysis using the high intensity and extended Q-range SANS diffractometer KWS-2. Simultaneous approaches using light absorption and scattering with standard or real time SANS are now possible with the KWS-2, which enable in-beam control of the sample quality or provide complementary structural and composition information to neutrons. The performance of each setup was demonstrated using systems representative of those typically investigated on this beamline and benchmarked to studies performed offline.

These developments have been greatly assisted by the short exposure times provided by the high neutron flux of the FRM II reactor, the increased maneuverability allowed at the sample position of the instrument, and the flexibility of the instrument control software, NICOS, which is developed in-house at Garching.

**Author Contributions:** Conceptualization, L.B., T.E.S., A.R.; methodology, F.K., T.E.S., A.R.; software, G.B.; validation, L.B., F.K., T.E.S., A.R.; investigation, L.B., T.E.S., A.R.; data curation, L.B., T.E.S., A.R.; writing—original draft preparation, L.B., F.K., T.E.S., A.R.; writing—review and editing, L.B., T.E.S., A.R. All authors have read and agreed to the published version of the manuscript.

**Funding:** Not applicable.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** We are thankful to Simon Staringer, Vladimir Ossovyi, Thomas Kohnke, and Andreas Nebel (JCNS) for their technical support during installation and commissioning of the in situ FTIR and DLS setups with the KWS-2 SANS diffractometer. Design and realization of the Peltierbased variable temperature cuvette holder by Vladimir Ossovyi (JCNS) is specially acknowledged. The help from Maria-Maddalena Schiavone with the DSC measurements is gratefully acknowledged. A. R. is thankful to Hiroki Iwase (CROSS Neutron Science and Technology Center, Tokai, Japan) for help with the concept drawings of the sample cuvettes for the FTIR and SANS setup. This project was partially financially supported by the BMBF projects 05K16PA1 and 05K19PA3. T. E. S. wishes to thank Stefan Rustler, Simon Lechlmayr, and Raimund Heigl for their pioneering work on the in situ DLS setup.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Simultaneous SAXS/SANS Method at D22 of ILL: Instrument Upgrade**

**Ezzeldin Metwalli 1, Klaus Götz 1,2,3, Tobias Zech 1,2,3, Christian Bär 1, Isabel Schuldes 1, Anne Martel 4, Lionel Porcar <sup>4</sup> and Tobias Unruh 1,2,3,\***


**\*** Correspondence: tobias.unruh@fau.de; Tel.: +49-9131-85-25189

**Abstract:** A customized portable SAXS instrument has recently been constructed, installed, and tested at the D22 SANS instrument at ILL. Technical characteristics of this newly established plugand-play SAXS system have recently been reported (*J. Appl. Cryst.* 2020, 53, 722). An optimized lead shielding arrangement on the SAXS system and a double energy threshold X-ray detector have been further implemented to substantially suppress the unavoidable high-energy gamma radiation background on the X-ray detector. The performance of the upgraded SAXS instrument has been examined systematically by determining background suppression factors (SFs) at various experimental conditions, including different neutron beam collimation lengths and X-ray sample-todetector distances (SDDX-ray). Improved signal-to-noise ratio SAXS data enables combined SAXS and SANS measurements for all possible experimental conditions at the D22 instrument. Both SAXS and SANS data from the same sample volume can be fitted simultaneously using a common structural model, allowing unambiguous interpretation of the scattering data. Importantly, advanced in situ/real time investigations are possible, where both the SAXS and the SANS data can reveal time-resolved complementary nanoscale structural information.

**Keywords:** small angle X-ray scattering; small angle neutron scattering; SAS; SANS; SAXS; nanomaterials

#### **1. Introduction**

Small angle scattering (SAS) of X-rays (SAXS) and neutrons (SANS) are non-destructive powerful techniques which have been utilized successfully for investigating nanostructured materials such as metal nanoparticles [1,2], emulsions [3], micelles [4,5], electrolytes [6,7], liquid crystals [8,9] and organic nanoparticles [10–13]. The SAS technique provides nanoscale information including shape, size, and size distribution as well as spatial distribution of dispersed materials in solution [14,15]. In particular, they provide the diameter of gyration and the mass of isotropically distributed shaped nanoscale particles [16], as well as the linear mass density and cross-sectional size of anisotropically shaped particles [14,17]. Moreover, SAXS and SANS methods are important tools for examining materials for various applications, e.g., solar cells, lithium-ion batteries and sensors [6,7,18–21]. Combining both SAXS and SANS methods to investigate the same sample volume is remarkably advantageous, where two radiations offer two different contrast situations (X-ray and neutron scattering cross section). Using such multiple modalities is very beneficial, especially for cases when a single modality provides incomplete or ambiguous results. For instance, SANS provides the form factor of the shell in a representative hollow-shell-like particle [2]. In contrast, SAXS measurements of the same core–shell sample are sensitive

**Citation:** Metwalli, E.; Götz, K.; Zech, T.; Bär, C.; Schuldes, I.; Martel, A.; Porcar, L.; Unruh, T. Simultaneous SAXS/SANS Method at D22 of ILL: Instrument Upgrade. *Appl. Sci.* **2021**, *11*, 5925. https:// doi.org/10.3390/app11135925

Academic Editor: Sebastian Jaksch

Received: 1 June 2021 Accepted: 23 June 2021 Published: 25 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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 (https:// creativecommons.org/licenses/by/ 4.0/).

porcar@ill.fr (L.P.)

to the high-electron-density core. The structures of both shell and core can be uniquely obtained from the same sample volume via complementary SAXS and SANS methods. Thus, simultaneous SAXS/SANS measurements will allow the exactness of the probed samples and the potential for using a common analysis model for both SAXS and SANS data, and consequently, a straightforward and unambiguous interpretation of the scattering data. For instance, structural modifications of supramolecular gels upon switching between H2O and D2O have been reported [22]. The deuteration effect on the structure of some nanoscale structured samples cannot be ignored. Thus, the simultaneous SAXS/SANS method is an essential tool for eliminating any ambiguity regarding sample exactness.

A key contribution of the simultaneous SAXS/SANS method is its ability to perform in situ and operando studies on advanced materials that undergo temporal modifications under external stimuli [23], allowing access to their structural rearrangements from the molecular scale to nanoscale assemblies. This means not only the synergy that results when data sets are obtained simultaneously from the same sample volume, but also access to the cross-correlation of phases/components when two different simultaneous contrast situations are probed for multi-phase/component samples. For instance, the cross-correlation between evolved gold nanorods size/shape (SAXS) and the corresponding engaged structure of organic moieties (SANS) during surfactant-assisted seed-mediated synthesis of gold nanorods will reveal new information related to the exact growth mechanisms involved [23]. Moreover, the SAXS/SANS system will enable contrast-dependent successive length-scale characterization of novel nanomaterials, addressing unsolved scientific questions.

A few instruments at neutron research reactors are currently utilized for both neutron and X-ray radiation experiment, obtaining complementary results of investigated samples. These dedicated instruments are mainly reflectometry and imaging techniques. A neutron imaging setup at PSI is outfitted with an additional X-ray tube to perform computerized tomography (CT) investigations using X-ray and neutron radiation [24,25]. Similarly, an instrument (NeXT) at ILL hosts two complementary imaging setups; neutron and Xray tomography [26,27]. Neutron reflectometers such as the NREX instrument at MLZ allows combined X-ray and neutron specular reflectivity measurements using both Xray and neutron beams placed in an orthogonal geometry [28]. Unfortunately, in most cases for the latter instruments, high energy background signal on the X-ray detector is a significant challenge and hampers a simultaneous mode of combined X-ray and neutron methods. The X-ray detection system is often not optimized for simultaneous acquisition, thus both X-ray and neutron data sets are mainly acquired successively and not in a synchronous mode [29,30].

An advanced SAXS system mounted on a standalone metal rack that makes it easily movable for use on the D22 SANS instrument has successfully been installed and tested at ILL [23]. For the first time, SAXS and SANS measurements in a simultaneous mode have been demonstrated [23]. One challenge is cutting down background signal that originates mainly from high-energy gamma radiation [31] in the experimental zone of the D22 instrument. A preliminary lead shielding design around our SAXS instrument at D22 of ILL has shown the capability of acquiring background discriminated SAXS data on a single energy threshold X-ray detector, enabling simultaneous SAXS/SANS experiments [23]. To enable a combined SAXS/SANS method in a simultaneous mode for all possible experimental conditions at the D22 instrument, an upgrade of the SAXS system was an essential requirement. In the current work, an optimized lead shielding has been implemented and the background suppression has been systematically examined. Additionally, a double energy threshold X-ray detector has been employed, replacing the single energy threshold detector. Overall reduced background noise and acceptable signal-to-noise ratios of SAXS data have been demonstrated upon combining both modifications, for all possible experimental configurations at the D22 instrument. Combined SAXS/SANS measurements were acquired for two nanoparticle (NP) dispersions of different sizes, to demonstrate the performance of the SAXS/SANS measurements under various experimental conditions.

#### **2. Materials and Methods**

An advanced SAXS system based on a Copper/Molybdenum (Kα; 8.0/17.4 keV) switchable microfocus rotating anode X-ray generator (Rigaku MM007 HF DW) and a Dectris EIGER2 X 1M detector with a continuously variable sample-to-detector distance (SDDX-ray) from 0.55 m up to 1.63 m in a vacuum tube is dimensionally suitable for installation on the D22 SANS instrument at ILL (Figure 1). For Cu Kα radiation of 8 keV (λ ≈ 0.1540 nm wavelength), the absolute value of the scattering vector *q* can be selected to range between 0.005 and 0.5 Å−<sup>1</sup> by the SDDX-ray. Due to the unsuitability of copper radiation for iron containing samples (such as magnetite, maghemite, and hematite) and for reaching higher penetration depths, the X-ray source can easily be switched to Mo Kα radiation of 17.4 keV. The short wavelength (0.071 nm) of the Mo radiation also allows the *q* range of the SAXS instrument to be extended [23]. With a Rigaku VariMax™ beam divergence control (BDC) system mounted on the X-ray source, the dual wavelength (Cu and Mo) optics can be operated. A control software can be used to automate the VariMax DWTM optics when changing from one X-ray source to the other, so that the operation of the whole system can proceed easily. A collimation length of approximately 56 cm is composed of three different types of slits. The first tungsten carbide slit suitable for both Cu and Mo radiation is placed directly after the mirror on the source, however, two scatterless slits (JJ X-ray) are positioned at the end of the collimation. The last two scatterless slits, either silicon (used for Cu radiation) or GaAs (used for Mo radiation), define the divergence of the primary beam. The whole assembly (X-ray source, optics and collimation system) can be moved along three axes (x, y, z), enabling a fine tune of the X-ray beam position, so that both the X-ray and neutron beam can be superimposed at the same sample position. The beam size (FWHM) of 0.5 × 0.5 mm at the sample position provides a flux of approximately 1.5 × <sup>10</sup><sup>7</sup> photons/s (Cu source).

The components of the system are mounted on a standalone metal rack that makes it easily movable for use on the D22 instrument. The floor of the experimental zone is equipped with an adjustable metal support, allowing fast and precise positioning of SAXS system. Via a plug-and-play operation, the instrument can be used within half an hour following the installation process (1 h). NICOS control software (MLZ), combined with a TANGO environment, is used to control the whole instrument, including the motors and the SAXS data detection system. A goniometer (Huber) with six degrees of freedom is employed to achieve movements of the sample. The sample is typically positioned at 45◦ relative to both of the orthogonal X-ray and neutron beams. Following the sample position, an evacuated detector tube hosts a double energy threshold EIGER2 X 1M detector (pixel size = 75 × <sup>75</sup> <sup>μ</sup>m2). No beam stop is required to attenuate the direct beam, thus enabling investigations with direct beam intensity at the detector (no beam-stop shadow). We refer readers to our previous work [23] for more details on the instrument specifications.

Background intensity measurements at the X-ray detector were measured at different experimental conditions, including neutron collimation lengths and SDDX-ray. Upon installing an optimized lead shielding arrangement on the metal chassis of SAXS setup, as well as employing a double energy threshold detector, the overall background reduction factor was evaluated compared with both the unshielded SAXS instrument and a single energy threshold detector. An extensive preliminary test prior to the latest instrumental upgrade was performed. The current results on background suppression demonstrate the achievements of increasing the signal-to-noise ratio on the X-ray detector. This was essential in order to enable a combined SAXS and SANS measurement in a synchronous mode. SANS data were collected using the D22 instrument at a neutron wavelength of 6Å(Δλ/λ = 10%), an aperture of 10 mm and SDDNeutron of 1.4, 5.6, 8, 17.6 m to cover <sup>a</sup> *<sup>q</sup>*-interval ranging from 3 × <sup>10</sup>−<sup>3</sup> to 0.7 Å−1, corresponding to sizes 2R ∼ <sup>π</sup>/*<sup>q</sup>* varying between 5 and 1050 Å. Aiming at examining the feasibility of acquiring an acceptably high signal-to-noise ratio for SAXS data at certain experimental conditions, two different sized silica NPs (100 nm and 26 nm) were tested, as these samples are of typical sizes (covering *q*-ranges) measured at various experimental conditions (collimation lengths 17.6, 8.0 and 5.6 m). The first sample was a powder of monodisperse silica NPs with a nominal size of 100 nm. The second sample was an aqueous dispersion of commercial silica NPs (Ludox TM50, 50 wt %) with an average diameter of approximately 26 nm, purchased from Sigma–Aldrich, Germany. A 5 mL solution (2 wt %) was prepared from each silica sample in a deuterated solvent (D2O). Both SAXS and SANS data were simultaneously collected at different SDDX-ray, SDDneutron and neutron collimation lengths. Both SAXS and SANS data sets were processed using Grasp software [32]. In the case of generating large numbers of SAXS and SANS data (for instance during in situ investigations), easy online data reduction and qualitative analysis of scattering data using this common software assists with the interactive optimization of experimental analysis.

#### **3. Results and Discussion**

A challenging requirement in order to perform experiments using combined X-ray and neutron radiation at large scale neutron research facilities is to shield the sensitive X-ray detector against the high-energy X-ray, neutron, and gamma radiation in the experimental zone. For high signal-to-noise ratio SAXS data, it is important to minimize the background radiation level, both from neutrons scattered by the sample and from gamma rays produced by neutron absorption within the sample. For this purpose, lead shielding and a double energy threshold detector were further employed.

#### *3.1. Reduction of the Gamma Background by Lead Shielding*

10 cm thick lead metal sheets were attached to the concrete walls towards the neighboring IN15 and WASP instruments to reduce the gamma radiation at the SAXS detector position in the D22 experimental zone. Additionally, several lead shielding walls were constructed and installed on the chassis of the SAXS instrument (Figure 1). The shielding walls include (i) a cone-shaped wall (nose) of lead-based material in front of the detector vacuum chamber, (ii) a lead wall along the final neutron collimation line, and (iii) a lead wall on the side of SAXS vacuum chamber. The current design has been pre-tested so that the X-ray detector inside the vacuum chamber is appropriately shielded from radiation for all possible positions of the X-ray detector in the detector tube. In the photograph of Figure 1, the installed lead-based shielding walls attached to the SAXS chassis are visible.

A double-energy threshold EIGER2 X 1M detector replacing a single-energy threshold EIGER R 1M detector is currently in operation and has been placed inside the vacuum chamber of the SAXS system. It allows the use of an upper and a lower adjustable energy discriminator (for instance, at 4 keV and 10 keV) to select a sensitive energy range close to the characteristic energy of the employed X-ray source (Cu Kα 8 keV).

Here, background measurements were recorded for two extreme configurations of neutron collimation lengths, referred to in the following as 'short' (2.8 m) and 'long' (17.6 m). The SAXS background measurements were performed inside the D22 zone using the singleenergy (4 keV) and the double-energy threshold (4 and 10 keV) modes for both the lead shielded and the unshielded SAXS instrument. The unshielded system represents a configuration where the lead shielding walls i, ii, and iii (cf. Figure 1) are not installed. From the data visualized in Figure 2, it can be concluded that the detected background signal can be substantially reduced by the lead shielding for both neutron collimation lengths (2.8 and 17.6 m). Further reduction of the background can be achieved by switching the X-ray detector from the single-energy threshold mode (th1 mode: 4 keV, dashed area) to the difference mode (diff mode: 4–10 keV, solid area) (cf. Figure 2). Measurements were performed for both the short (SD1 = 0.55 m red bars) and the long (SD2 = 1.64 m blue bars) SSDX-ray of the SAXS system. The achieved total background SFs are summarized in Table 1.

**Figure 1.** (**a**) Three lead shielding walls: (i) a cone of lead-based material, (ii) lead wall along the final part of the neutron collimation line, and (iii) a lead wall (2 tons) along the vacuum chamber of the X-ray detector. The shielding walls can be easily craned and then firmly attached to the metal chassis of the SAXS setup. Red and yellow arrows indicate the directions of the X-ray and neutron beams, respectively. The unshielded SAXS instrument (detector) represents a configuration where the lead shielding walls i, ii, and iii are not implemented during SAXS operation. A double-energy threshold X-ray detector was placed inside the vacuum tube, replacing the previously employed single-energy threshold detector. (**b**) A zoomed-in detail of the sample environment of the SAXS/SANS instrument, displaying a sample at an angle of 45◦ relative to both of the orthogonal neutron and X-ray beams.

**Figure 2.** Background intensity collected using EIGER2 X 1M detector while neutron shutter is opened for two different neutron collimation lengths (2.8 m and 17.6 m). Two measurements for the singleenergy threshold mode (dashed area, th1 mode = 4 keV) and two for the double-energy mode (solid area, diff = 4–10 keV) were performed for both the shielded and unshielded SAXS instrument. All background measurements were repeated for short (red, SD1 = 0.55 m) and long (blue, SD2 = 1.63 m) SDDX-ray. Lab environment data are based on background intensity measurements while the reactor power was off. The intensity (cts/s) data are shown on a logarithmic scale.


**Table 1.** Background suppression factors (SFs) for all investigated experimental configurations in an optimized lead shielding setup with a double-energy threshold energy detector (total SF ≈ lead shielding SF × detector SF).

> The gamma background is originated from the facility environment and nuclear reactions of the neutrons with the sample material. The short collimation length of the SANS instrument (2.8 m) leads to a higher neutron flux compared to longer collimation lengths. The increase of neutron flux and the gamma radiation background level is roughly proportional to the inverse of the squared collimation length. To shield the X-ray detector (at different SDDX-ray) against this high-energy gamma radiation, several preliminary tests were performed to choose the best lead wall dimensions and geometric arrangements for our SAXS system. The detector SF was calculated as the ratio of background intensity for single-energy and double-energy threshold modes. The total suppression factor is theoretically equal to the product of both shielding and detector SFs. An uncertainty of the total SF of ≤ 1% was observed under identical experimental conditions. Finally, large SFs were achieved, especially at short SDDX-ray (cf. Table 1). For SDDX-ray = 0.55 m, the background intensity could be reduced by a factor of 505 and 275 for short (2.8 m) and long (1.7.6 m) neutron collimation lengths, respectively. For both the unshielded and shielded instrument, the background SF for the double-energy threshold detection was almost equal. The gamma background which originates from the neutron interaction with the sample or from the environment and transmitted through the sample towards the detector cannot be shielded. But the double-energy threshold detector is able to reduce this background component. The tremendous reduction in the background signal of the upgraded instrument for the benefit of simultaneous SAXS/SANS measurements is demonstrated in Section 3.3 by SAXS/SANS data from NP samples.

#### *3.2. Shielding the X-ray Source against External Magnetic Fields*

It has occasionally been observed that the X-ray beam intensity periodically varies during SAXS measurements. Such beam intensity fluctuations had not been observed before operating it at the D22 SANS instrument. Using a slit size of 0.5 × 0.5 mm2 in front of the sample, it was observed that the X-ray beam moves partially outside the slit's openings, leading to an intensity reduction of up to 60%. It was a challenge to identify a reason behind such occasional instability behavior of X-ray intensity during operation inside the experimental zone. Finally, the variation of the residual weak magnetic field (~2–5 Gauss) from the IN15 spin-echo spectrometer was found to cause these X-ray beam intensity fluctuations. A plate of soft ferromagnetic alloy was installed to shield the static and low-frequency magnetic field at the position of the X-ray source, which leads to a stable beam intensity (Figure 3) for all extreme magnetic fields employed at the IN15 instrument.

**Figure 3.** X-ray beam intensity before (no magnetic shielding, green line) and after (magnetic field shielding, black symbols) installing a 1 mm thick plate of a soft ferromagnetic alloy at the X-ray source. A magnetic field strength of approximately 2–5 Gauss at the X-ray source causes an X-ray beam drift.

#### *3.3. Overall SAXS Performance*

For simultaneous SAXS/SANS experiments, it is important to achieve sufficient data quality in a given measuring time for both methods. For that purpose, it was important to reduce the background signal originating from gamma rays on the 2D X-ray detector of the SAXS instrument for all experimental conditions represented by different SANS collimation lengths and primary neutron beam fluxes and sizes, as well as different SDDX-ray. As demonstrated above, this could be achieved by a targeted enhancement of the neutron and magnetic field shielding, as well as using a double-energy threshold X-ray detector. To demonstrate the success of these upgrades for the performance of the SAXS instrument, two NP samples of nominal sizes of 100 nm and 26 nm were studied by simultaneous SAXS/SANS experiments. The scattering data was collected while the nuclear reactor was in operation (neutron shutter opened, reactor power = 43 MW). SAXS data of the same samples were also collected when the neutron reactor was in the shutdown time (neutron shutter closed, reactor power < 1.5 MW). The latter experiments allow a comparison of the SAXS data of the investigated samples under simultaneous SAXS/SANS data recording conditions with standard laboratory standalone SAXS experiments, where only natural cosmic background contributes to the gamma background of the SAXS data.

SAXS data of silica NP samples were collected for 15 min for both short (0.55 m) and long (1.63 m) SDDX-ray (Figure 4). For long SDDX-ray, SANS measurements of both silica NPs were simultaneously acquired at a long collimation length of 17.6 m. For short SDDX-ray, SANS data were simultaneously measured at two different collimation lengths of 8 m and 5.6 m for 100 nm and 26 nm silica NPs, respectively. The SAXS 2D detector patterns were corrected and normalized, respectively, for measuring time, sample thickness, transmission, flat field, solid angle covered by each detector pixel, and incident intensity. The data were calibrated to absolute intensities, and the *q*-scale was azimuthally averaged. Background (dark current) correction was performed by subtracting the background intensity measured on the detector at relevant operating conditions. For each neutron collimation length used, the background intensity on the X-ray detector was collected for several minutes while the neutron shutter was opened but the X-ray shutter was closed. Besides this, each SAXS data set was corrected for a constant background determined from a measurement of the corresponding sample with a closed neutron shutter (and reactor power < 1.5 MW).

**Figure 4.** 1D SAXS profiles at various neutron collimation lengths and SDDX-ray of (**a**,**b**) 100 nm and (**c**,**d**) 26 nm silica NPs collected with the neutron beam switched on (orange curves) or off (blue) (simultaneous and asynchronous modes). The asynchronous mode represents a standard laboratory SAXS experimental condition where neutron shutter was closed and the nuclear reactor was almost switched off (below < 1.5 MW). SAXS/SANS simultaneous mode represents an experimental condition where the neutron shutter of D22 was open (reactor power = 43 MW). At long SDDX-ray, the SAXS profiles in simultaneous mode are well matching those of the asynchronous mode, which leads to a good signal-to-noise ratio for the SAXS data. For short collimations, noticeable background which was approximately 4 and 8 times larger than that for standard lab SAXS measurements was found for both collimations 8 m (**b**) and 5.6 m (**d**), respectively.

At a long collimation length (17.6 m), the SAXS 1D profiles are well matching those collected in absence of the high-energy gamma radiation environment (Figure 4a,c). This reveals that the SAXS data quality of a simultaneous SAXS/SANS measurement at those conditions (collimation length = 17.6 m, SDDX-ray = 1.63 m) is not compromised with respect to the background level of a state-of-the-art lab-based SAXS measurement. At a short distance SDDX-ray (0.55 m), SANS images were collected in a synchronous mode at collimation lengths of 8 m (for 100 nm NPs) and 5.6 m (for 26 nm NPs). In the latter configurations, the SAXS profiles exhibited a background intensity approximately 4 and 8 times higher than the standard lab-based SAXS measurements for both 8 m and 5.6 m collimation lengths, respectively (Figure 4b,d). This gamma background originated predominantly from the gamma rays produced by neutron capture within the sample. Nevertheless, the overall background level even for the short neutron collimation lengths is still low and allows high quality simultaneous SAXS/SANS experiments. Based on the *q*-range of characteristic interference features, scattering power of investigated samples, and experimental conditions (such as collimation length, SDDX-ray, X-ray slit size, and measuring time), one needs to carefully optimize the experimental conditions to achieve the best quality simultaneous SAXS/SANS experiments.

Luckily, the D22 instrument is currently equipped with a second fixed detector at 1.4 m from the sample position in addition to a movable detector, enabling extended SANS *q* ranges (from 0.003 up to 0.7 Å−<sup>1</sup> in a single configuration). Hence, typical SANS experiments are performed with collimation lengths > 5 m. Thus, the high signal-tobackground level of the SAXS data detected at different collimation lengths (17.6, 8 and 5.6 m) allows state-of-the-art time-resolved in situ simultaneous SAXS/SANS measurements. This has been demonstrated by studies on the growth of gold nanorods using surfactantassisted seed-mediated synthesis using simultaneous SAXS/SANS at a neutron collimation length of8m[23].

For SAXS/SANS data evaluation of the studied NP dispersions (Figure 5), a simple sphere model with a log-normally distributed radius coupled with a distribution width parameter implemented in SASfit software [33] was used. In SASfit, simultaneous fits of the model to the SAXS and SANS experimental data sets were performed with a convergence χ << 1. The scattering length density contrast ΔSLD = SLDNPs − SLDsolvent was fixed to the known values for silica to be 13.3 and 2.2 Å−<sup>2</sup> for SAXS and SANS, respectively. The data can be nicely fitted with the simple model and reveals a particle diameter (with a distribution width) of 110 ± 7 nm (nominal 100 nm) and 26 ± 3 nm (nominal 26 nm), respectively. This simple example demonstrates that a simultaneous fit of a structural model to the SAXS and SANS data set can be easily performed on an absolute intensity scale and reveal the structural parameters, circumventing ambiguities which might be left if only one contrast (X-ray or neutron) is used. This is a particular advantage for in situ or operando studies of complex nanoscale systems which slowly change their structure, e.g., upon a reaction, crystallization, aggregation, gelation, particle formation, growth or aging process.

**Figure 5.** Solvent subtracted SAXS/SANS intensity profiles of 100 nm (**a**) and 26 nm (**b**) silica NPs acquired in a simultaneous mode. A common structural model is used for each sample to simultaneously fit both SAXS and SANS data collected from the same sample volume.

#### **4. Conclusions**

A customized SAXS system was successfully installed and operated in combination with the D22 SANS instrument of ILL, resulting in simultaneously collected SAXS and SANS data of the same part of a sample. Due to high-energy gamma radiation inside the D22 instrumental zone, the SAXS instrument was upgraded with a double-energy threshold detector and optimized lead shielding. Background suppression factors were evaluated for different collimation lengths and SDDX-ray. Extreme background reduction by factors up to 505 was achieved for short collimation lengths and SDDX-ray. Acceptable signal-tobackground ratios of the SAXS patterns comparable to SAXS data collected in the absence of gamma background radiation could be achieved. For long neutron collimation lengths of up to 17.6 m and long SDDX-ray, the SAXS profiles did not exhibit an increase in background at all when compared to a typical lab experiment. An increase of the corresponding low background by factors of 4 and 8 were observed for neutron collimation lengths of 8 m and 5.6 m, respectively. In the latter experimental configurations, SAXS data of selected test samples prove the suitability of the SAXS instrument for combined SAXS/SANS experiments. The upgraded SAXS system is now compatible with all collimation settings

of the D22 SANS instrument, enabling in situ and operando simultaneous SAXS/SANS measurements for a variety of complex nanoscale sample systems.

**Author Contributions:** Conceptualization, T.U., C.B., E.M., K.G., T.Z., L.P.; methodology, E.M., K.G., T.Z., L.P., I.S., C.B., A.M.; data curation, E.M., I.S.; formal analysis, E.M., I.S.; writing—original draft preparation, E.M.; writing—review and editing T.U., L.P., A.M., I.S.; technical details, C.B., E.M., T.U., L.P.; funding acquisition, T.U. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research is primarily funded by the Federal Ministry of Education and Research of Germany (BMBF) in the framework of project No. 05K16WE1. As well, this research received funding from the Deutsche Forschungsgemeinschaft (DFG) through the 'Cluster of Excellence Engineering of Advanced Materials (EAM)', the research training group GRK 1896 'In-situ Microscopy with Electrons, X-rays and Scanning Probes' and the research unit FOR 1878 'Functional Molecular Structures on Complex Oxide Surfaces'.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data presented in this study can be obtained from the authors upon request.

**Acknowledgments:** We gratefully thank Herbert Lang (ICSP), Jürgen Grasser (ISCP), Mark Jacques (ILL) for their technical support. We acknowledge ILL for the beamtime DOI:10.5291/ILL-DATA.LTP-9-9.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**

The following abbreviations are used in this manuscript:


#### **References**

