Development of FTIR Spectroscopy Methodology for Characterization of Boron Species in FCC Catalysts

catalysts

Article Development of FTIR Spectroscopy Methodology for Characterization of Boron Species in FCC Catalysts

Claire Chunjuan Zhang , Xingtao Gao and Bilge Yilmaz * BASF Corporation, 25 Middlesex-Essex Tpk., Iselin, NJ 08830, USA; [email protected] (C.C.Z.); [email protected] (X.G.) * Correspondence: [email protected]

Received: 23 October 2020; Accepted: 13 November 2020; Published: 15 November 2020

Abstract: Fluid Catalytic Cracking (FCC) has maintained its crucial role in refining decades after its initial introduction owing to the flexibility it has as a process as well as the developments in its key enabler, the FCC catalyst. Boron-based technology (BBT) for passivation of contaminant metals in FCC catalysts represents one such development. In this contribution we describe Fourier Transform Infrared Spectroscopy (FTIR) characterization of boron-containing catalysts to identify the phase and structural information of boron. We demonstrate that FTIR can serve as a sensitive method to differentiate boron trioxide and borate structures with a detection limit at the 1000 ppm level. The FTIR analysis validates that the boron in the FCC catalysts studied are in the form of small borate units and confirms that the final FCC catalyst product contains no detectable isolated boron trioxide phase. Since boron trioxide is regulated in some parts of the world, this novel FTIR methodology can be highly beneficial for further FCC catalyst development and its industrial application at refineries around the world. This new method can also be applied on systems beyond catalysts, since the characterization of boron-containing materials is needed for a wide range of other applications in the fields of glass, ceramics, semiconductors, agriculture, and pharmaceuticals.

Keywords: ﬂuid catalytic cracking; resid-FCC; boron; FTIR; spectroscopy; Environmental Health and Safety (EHS)

1. Introduction As one of the cornerstones of petroleum refining, Fluid Catalytic Cracking (FCC) is a key conversion processes that provides majority of gasoline consumed throughout the world in addition to other important transportation fuels and petrochemical feedstocks. In terms of the type of crude oil processed, the global trend has been towards heavier and more contaminated feedstocks in recent years; significantly increasing the importance of mitigating the negative effects of contaminants that come with these heavier crudes. Today, the significance of nickel as one of the most challenging contaminant metals on FCC catalysts has been thoroughly acknowledged and there has been considerable effort dedicated to developing technologies to be able to process feeds with elevated nickel levels. Boron Based Technology (BBT), represents one such development [1]; and since its industrial introduction, it has been used in many refineries around the world. This FCC catalyst technology utilizes boron compounds as precursors for introducing the novel nickel passivation feature to enhance in-unit performance such as lower levels of undesirable hydrogen and coke yields and increased yields of valuable hydrocarbon products [1]. Boron compounds are already widely used in different fields such as glass and ceramics, polymers, semiconductors, bleaching products, insecticides, agricultural fertilizers, etc. Among the variety of boron compounds, boron trioxide (B2O3) is commonly used as a raw material additive in glass, ceramics, and semiconductors to improve the electrical, thermal, and/or mechanical properties of

Catalysts 2020, 10, 1327; doi:10.3390/catal10111327 www.mdpi.com/journal/catalysts Catalysts 2020, 10, 1327 2 of 11 these materials. Boron trioxide is composed of six-membered boroxol rings with 3-coordinate boron and 2-coordinate oxygen. Boron trioxide can react with oxide materials and turn to borate form at high temperatures. Direct contact, inhalation, or ingestion of boron trioxide may irritate eye, skin and respiratory system. Toxicity testing studies on animals indicate that extended exposure to high doses of boron trioxide may cause issues with reproduction, although statistical studies on workers at boron trioxide manufacturing sites did not support the results in this claim [2–5]. On the other hand, borate species are usually the final forms of boron-species in products after high temperature reactions between boron trioxide and other oxide materials. These borates are seen as benign with regards to reproductive health and are exempted from the regulated/controlled chemicals list of the European Chemicals Agency (ECHA) [6–8]. The aim of this study is to provide a spectroscopic characterization methodology to identify the boron species in FCC catalysts, to be able to categorize them as bulk boron trioxide or as borate species. This analysis is important from a regulatory perspective especially for the use of these products in regions where there are regulations in place for use of boron trioxide, such as in European Union with the regulatory requirements by ECHA [2,3]. Boron-containing materials are known to be challenging to characterize using scanning electron microscopy (SEM) and electron dispersive X-ray spectroscopy (EDS), owing to boron’s low atomic number, difficulty for ionization, and low fluorescence yield upon excitation. [9] It is also challenging to use X-ray photoelectron spectroscopy (XPS) to analyze boron directly using B 1s due to its low photoelectron cross-section and overlapping with P 2s and Si 2s plasmon peaks [10]. Nuclear Magnetic Resonance (NMR) spectroscopy can be a sensitive technique for detecting some boron species, especially for BO4 structure/phase; however, it is challenging to differentiate the BO3 in borate with BO3 in boron trioxide phase [11]. Infrared spectroscopy is an important tool to characterize the structural properties of amorphous materials, especially for boron-containing materials. When there is a change in the dipole moment of a bond vibration, infrared (IR) signals can be generated and detected due to the match of excitation light energy with the bond vibration energy level. Valid IR signals can come from surface-adsorbed gas molecules, or from the structural vibrations in the solid phases. In this contribution, IR-spectroscopic characterization of boron trioxide and several borate-containing materials is provided when the boron content is low in the materials, which provides a quick way to achieve the structural information about the nature of the boron species. This method can serve as an important tool for product development and commercial utilization for boron-containing catalysts as well as other materials, and help align these products with the regulatory requirements.

2. Results and Discussion

2.1. Boron Trioxide Characterization via FTIR Boron trioxide is a well-known glass former that can form glass on its own under appropriate conditions. The vitreous glass structure of boron trioxide can be formed by simply heating boric acid to temperatures higher than 176 ◦C. The primary structure in vitreous boron trioxide glass consists of 6-member boroxol rings that maintain the short-range order, and bridged oxygen atoms to inter-connect the small units. The red plot in Figure1a shows the FTIR spectrum for vitreous boron trioxide glass that was formed after heating ground boron trioxide powder at 200 ◦C for an hour. In literature, the vitreous 1 1 boron trioxide glass shows two IR peaks at 1260–1280 cm− and 720–760 cm− , which are assigned to transverse-optical-longitudinal-optical (TO-LO) splitting due to the IR light interaction propagated 1 in longitudinal and transverse directions [12]. Some others also assigned the peak at ~740 cm− to the bending mode of B-O-B linkage structure in BO3 units [13]. In the higher wavenumber range 1 (>1600 cm− ), the IR light is almost completely absorbed by the vitreous glass structure as shown in 1 the cliﬀ-like absorbance in the FTIR spectrum. The negative feature at 1286 cm− is likely due to light reﬂection at this wavenumber (energy/frequency) that enhances the signal in the detector. Catalysts 2020, 10, x FOR PEER REVIEW 3 of 12

dehydration. The sharp peak at 995 cm−1 is assigned to the BO4 structure, while the 665 cm−1 peak is due to PO4 vibrations [14,15]. The peak sharpness is related to the short-range ordering in the glass Catalystsnetwork2020 as, 10 compared, 1327 to broader absorbance in amorphous structures. The total absorbance cutoff3 is of 11 at ~ 1400 cm−1 for glassy PBO4.

2.0 2.0 a B-O-B b BO4

1.5 1.5 665

995

1.0 1.0

757

0.5 0.5

Absorbance (a.u.) 0.0

1286 Boron trioxide Phospor borate PBO4 -0.5 -0.5 2000 1800 1600 1400 1200 1000 800 600 2000 1800 1600 1400 1200 1000 800 600 Wavenumber (cm-1) Wavenumber (cm-1)

FigureFigure 1. 1.FTIR FTIR spectra spectra forfor ((aa)) glassyglassy boron trioxide trioxide and and (b (b) )glassy glassy phosphor phosphor borate borate (PBO (PBO4). 4).

Similarly,When boron phosphor trioxide borate is mixed (PBO with4) canother also oxide form materials, a glass such by itself. as silica, The and blue thermally plot in Figuretreated1 b showsat high the temperatures, FTIR spectrum different collected types on of glassy borate PBO structures4 that wascan be formed formed after [16] being. The polymeric treated at network 400 ◦C for 1 1 dehydration.of boroxol rings The in sharp pure peak boron at trioxide 995 cm −is interrupted.is assigned toMuch the BOsmaller4 structure, structural while BO3 the and 665 BO cm4 units− peak are is formed as borate species. As the boron content gets lower, it becomes less likely that the boroxol ring due to PO4 vibrations [14,15]. The peak sharpness is related to the short-range ordering in the glass networkstructures as compared remain intact, to broader and more absorbance likely that in small amorphous borate structures.structures are The formed. total absorbance This phenomenon cutoff is at is simply1 due to higher mobility of small units of borate species at high temperatures, which tend to ~1400 cm− for glassy PBO4. disperseWhen uniformly boron trioxide in the is solid mixed solution. withother oxide materials, such as silica, and thermally treated at high temperatures, different types of borate structures can be formed [16]. The polymeric network of 2.2. Borate structure characterization via FTIR boroxol rings in pure boron trioxide is interrupted. Much smaller structural BO3 and BO4 units are formedFTIR as borate characterization species. As of the boron boron-containing content getsmaterials lower, is it widely becomes used less in likely the field that of the glasses boroxol and ring structuresceramics. remain The borate intact, vibration and more modes likely are that detected small in borate three structuresmain infrared are formed.regions. The This IR phenomenon absorbance is −1 simplyin the due1200 to-1500 higher cm mobility range is assigned of small unitsto asymmetric of borate stretching species at of high B-O temperatures,bond in trigonal which BO3 units. tend to −1 disperseThe B-O uniformly stretching in in the tetrahedral solid solution. BO4 units is in the range of 800-1200 cm . The bending vibrations of bridging oxygen (B-O-B) between BO3 and BO4 units are active in the range of 600-800 cm−1 [17-19]. 2.2. BorateFor alumina Structure and Characterization boron trioxide via mixtures, FTIR glassy structures, or polymeric network structures of boron trioxide cannot be formed even after high temperature treatments. Depending on the preparationFTIR characterization temperature and of boron-containing the stoichiometric materials ratio between is widely aluminum used in and the boron, field of aluminum glasses and ceramics.borate may The form borate differen vibrationt cluster modes structu areres detected [20,21] in. Figure three main2 shows infrared the FTIR regions. spectrum The collected IR absorbance for 1 inAl the18B 1200–15004O33 purchased cm− fromrange Alfa is Aesar assigned, which to asymmetric shows dominating stretching BO3 ofstructures B-O bond centered in trigonal at 1320, BO 12603 units., 1 Theand B-O 700 stretching cm−1, and in less tetrahedral amount of BO BO4 4units structures is in the at 1075 range and of 800–1200985 cm−1. cm− . The bending vibrations of 1 bridging oxygen (B-O-B) between BO3 and BO4 units are active in the range of 600–800 cm− [17–19]. For alumina and boron trioxide mixtures, glassy structures, or polymeric network structures of boron trioxide cannot be formed even after high temperature treatments. Depending on the preparation temperature and the stoichiometric ratio between aluminum and boron, aluminum borate may form different cluster structures [20,21]. Figure2 shows the FTIR spectrum collected for Al 18B4O33 purchased 1 from Alfa Aesar, which shows dominating BO3 structures centered at 1320, 1260, and 700 cm− , and less 1 amount of BO4 structures at 1075 and 985 cm− . Some of the novel FCC catalysts, such as those from Boron-based Technology (BBT), use boron species as precursor in the preparation processes [1]. The function of boron in this technology is to help passivate nickel in the fluid catalytic cracking unit to achieve higher yields of valuable hydrocarbons such as liquefied petroleum gas (LPG) and gasoline, reduction of unwanted hydrogen, and reduction of coke. As demonstrated in Figure3 and discussed in detail below, the spectroscopic study confirms that during preparation/manufacturing of the catalysts, after the final calcination of the product, boron stays in borate structures (BO3 and BO4) in borate species rather than in boroxol rings in B2O3. Catalysts 2020, 10, x FOR PEER REVIEW 4 of 12

2.0

1.8

1.6

1320

1260 1.4

Catalysts 2020, 10, 1327 700 4 of 11 1.2 Catalysts 2020, 10, x FOR PEER REVIEW 4 of 12

985 BO3 1075 1.0 2.0 0.8 1.8 BO4 BO3 0.6 Absorbance (a.u.) Absorbance 1.6

1320 0.4 1260 1.4

0.2 700 1.2 Aluminum Borate

985 0.0 BO3 1075 1.0 1800 1600 1400 1200 1000 800 600 Wavenumber (cm-1) 0.8 BO4 BO3 Figure 2. FTIR0.6 spectrum for aluminum borate (Al18B4O33), which was obtained by processing the

Absorbance (a.u.) Absorbance single beam spectrum using a KBr reference spectrum. 0.4

Some of the novel0.2 FCC catalysts, such as those from Boron-based Technology (BBT), use boron species as precursor in the preparation Aluminum processesBorate [1]. The function of boron in this technology is to help passivate nickel0.0 in the fluid catalytic cracking unit to achieve higher yields of valuable 1800 1600 1400 1200 1000 800 600 hydrocarbons such as liquefied petroleumWavenumber gas (LPG) and gasoline, (cm-1 reduction) of unwanted hydrogen, and reduction of coke. As demonstrated in Figure 3 and discussed in detail below, the spectroscopic study confirms that during preparation/manufacturing of the catalysts, after the final calcination of Figure 2. FTIR spectrum forfor aluminum aluminum borate borate (Al (Al1818B4OO3333),), which which was was obtained obtained by by processing the the product, boron stays in borate structures (BO3 and BO4) in borate species rather than in boroxol single beam spectrum using a KBr reference spectrum. rings in B2O3. Some of the novel FCC catalysts, such as those from Boron-based Technology (BBT), use boron species as precursor in the preparation processes [1]. The function of boron in this technology is to help passivate nickel in the fluid catalytic cracking unit to achieve higher yields of valuable BO hydrocarbons such as liquefied (a.u.) petroleum gas (LPG) and gasoline, 4reduction of unwanted hydrogen, and reduction of coke. As demonstrated in Figure 3 and discussed in detail below, the spectroscopic BO BO3 study confirms that during preparation/manufacturingSi-O-Si 3 of the catalysts, after the final calcination of the product, boron stays in borate structures (BO3 and BO4) in borate species rather than in boroxol rings in B2O3.

1102

835

1433

1879

Absorbance

2005

BO (a.u.) 2000 1500 10004 0.05 BO WavenumberBO (cm-1) 3 Si-O-Si 3

FigureFigure 3. 3.FTIR FTIR spectrumspectrum ofof aa FluidFluid CatalyticCatalytic CrackingCracking (FCC)(FCC) catalystcatalyst (Sample(Sample #4)#4) showingshowing BOBO33, BO4

1102 borate structures. borate structures. 835

1433

1879

Absorbance SinceSince the the boron boron content content is is usually usually2005 at at relatively relatively low low levels levels in in FCC FCC catalysts, catalysts, a KBr KBr reference reference backgroundbackground cannotcannot bebe usedused toto process thethe absorbance spectrumspectrum here.here. Using a KBr background would simplysimply havehave thethe borateborate signalssignals overwhelmedoverwhelmed2000 byby1500 muchmuch strongerstronger1000 absorbanceabsorbance signals from otherother majormajor components in the catalysts.0.05 Instead, a background spectrum was collected on a specifically generated components in the catalysts. Instead,Wavenumber a background spectrum (cm-1) was collected on a specifically referencegenerated samplereference that sample has the that identical has the compositionidentical composition and structure and structure but without but without the boron the species. boron Thus, the borate signals stand out by eliminating/canceling other IR signals during data processing. Figure 3. FTIR spectrum of a Fluid Catalytic Cracking (FCC) catalyst (Sample #4) showing BO3, BO4 The proper selection of background spectra is the critical step in this methodology as it significantly borate structures. lowers the detection limit to sub-percentage levels. As shown in Figure3, BO (at 1433 and 835 cm 1) and BO structures (at 1102 cm 1) are identified Since the boron content 3 is usually at relatively− low levels4 in FCC catalysts, − a KBr reference in IR spectra. The incorporation of borate in the support (containing silica, alumina, and zeolite) background cannot be used to process the absorbance spectrum here. Using a KBr background would caused slight decrease in the Si-O-Si overtone signals at 2005 and 1879 cm 1, which indicates some simply have the borate signals overwhelmed by much stronger absorbance −signals from other major components in the catalysts. Instead, a background spectrum was collected on a specifically generated reference sample that has the identical composition and structure but without the boron

Catalysts 2020, 10, x FOR PEER REVIEW 5 of 12

species. Thus, the borate signals stand out by eliminating/canceling other IR signals during data processing. The proper selection of background spectra is the critical step in this methodology as it significantly lowers the detection limit to sub-percentage levels. As shown in Figure 3, BO3 (at 1433 and 835 cm−1) and BO4 structures (at 1102 cm−1) are identified in IR spectra. The incorporation of borate in the support (containing silica, alumina, and zeolite) Catalysts 2020, 10, 1327 5 of 11 caused slight decrease in the Si-O-Si overtone signals at 2005 and 1879 cm−1, which indicates some silicon being replaced by boron in the support. This result is expected since after being subjected to siliconhigh temperatures being replaced as part by boron of the inFCC the manufacturing support. This resultprocess, is expectedborate species since are after known being to subjected form upon to highinteraction temperatures of boron as-containing part of the FCCprecurs manufacturingor species with process, the oxide borate materials species are(such known as silicoaluminate to form upon interactionphases in typical of boron-containing FCC catalysts and precursor additives) species that with serve the as oxidethe support materials structure. (such as silicoaluminate phasesIn inorder typical to illustrate FCC catalysts the interaction and additives) and thattransformation serve as the of support boron structure.trioxide with other materials usedIn in order FCC, to two illustrate sets of the experimental interaction andsamples transformation containing of boron boron were trioxide prepared with other specifically materials for used the inFTIR FCC, characterization two sets of experimental study. In order samples to achieve containing a cleaner boron comparison were prepared for this specifically part of the for study, the FTIR an characterizationadditive setup was study. chosen In order as the to model achieve system. a cleaner ZSM comparison-5 additives for are this typi partcally of used the study,in FCC an operations additive setupto boost was propylene chosen as theselectivity model system.as well ZSM-5as to incre additivesase gasoline are typically octane used value inFCC [22,23] operations. They crack to boost the propylenegasoline range selectivity olefins as into well smaller as to increase hydrocarbons, gasoline and octane to serve value that [22, specific23]. They purpose crack the they gasoline are typically range olefinssimpler into in design smaller compa hydrocarbons,red to base and FCC to catalysts serve that since specific cracking purpose matrices they and are metals typically traps simpler that are in designpart of compared the base resid to base-FCC FCC catalysts catalysts are since not crackingincluded matrices in the additive and metals construct. traps that To arebe partable ofto theprovide base resid-FCCinformation catalysts on the are transformation not included in of the boron additive trioxide construct. in an To FCC be microsphere able to provide pro informationduction and on use, the transformationsamples prepared of boron with trioxidean identical in an additive FCC microsphere recipe were production calcined andat typical use, samples FCC catalyst/additive prepared with anmanufacturing identical additive conditions recipe were as a calcined final step at typical of the FCCprocedure. catalyst These/additive comparisons manufacturing were conditions repeated as at avarious final step boron of the loadings. procedure. These comparisons were repeated at various boron loadings. TheThe firstfirst setset of of samples samples presented presented below below in in Figures Figure4 and4 and5 contain Figure the5 contain physical the mixture physical of mixture boron trioxideof boron and trioxide ZSM-5 and with ZSM alumina-5 with binder. alumina The binder. second The set of second samples set contains of samples the contains same ingredients the same (i.e.,ingredients chemical (i.e., composition chemical of composition boron trioxide of boron and ZSM-5 trioxide with and alumina ZSM-5 binder) with alumina but finished binder) with bu at finalfinished calcination with a final step calcination at 649 ◦C bystep following at 649 °C a by protocol following typical a protocol to FCC typical catalyst to /FCCadditive catalyst/additive production. Theproduction. comparison The ofcomparison the two sets of ofthe samples two sets is of shown sampl ines Figures is shown4 and in 5Figures. 4 and 5.

0.1% B O / ZSM-5 with binder a Physical mixtures 2 3 0.25% B2O3 / ZSM-5 with binder

0.5% B2O3 / ZSM-5 with binder

1.0% B2O3 / ZSM-5 with binder

3704

3689

Absorbance (a.u.)

3743

3725 0.02 B-OH 3680 b 649 oC - calcined samples

3680

3725

3746

Absorbance (a.u.)

1.0% B2O3 / ZSM-5 with binder

1.5% B2O3 / ZSM-5 with binder

0.02 2.0% B2O3 / ZSM-5 with binder

3800 3750 3700 3650

Wavenumber (cm-1)

Figure 4. FTIR spectra in the OH region for (a) physical mixtures (Samples #5–8), and (b) 649 ◦C—calcined boron trioxide with ZSM-5 (Samples #9–11). Catalysts 2020, 10, x FOR PEER REVIEW 6 of 12

CatalystsFigure2020, 104., 1327FTIR spectra in the OH region for (a) physical mixtures (Samples #5-8), and (b) 649 °C 6– of 11 calcined boron trioxide with ZSM-5 (Samples #9-11).

a Physical mixtures

646

716

1199

1244

1404

1490 BO3 BO3 BO3

Absorbance (a.u.) 0.1% B O / ZSM-5 with binder 2 3 Si-O-Si 0.25% B2O3 / ZSM-5 with binder

0.5% B2O3 / ZSM-5 with binder 0.1 1094 1.0% B2O3 / ZSM-5 with binder

o 1.0% B O / ZSM-5 with binder 649 C - calcined samples 2 3 1.5% B O / ZSM-5 with binder b 2 3 BO4 2.0% B2O3 / ZSM-5 with binder

BO3 BO4 BO3

1237

707

788

1122

936

Absorbance (a.u.)

1505

0.1

1600 1400 1200 1000 800 600 Wavenumber (cm-1)

FigureFigure 5.5.FTIR FTIR spectraspectra inin thethe borateborate regionregion forfor (a(a)) physical physical mixtures mixtures (Samples (Samples #5–8), #5-8), andand ((bb)) 649649 ◦°C–C – calcinedcalcined boron boron trioxide trioxide with with ZSM-5 ZSM-5 (Samples (Samples #9–11). #9-11).

TheThe physicalphysical mixturesmixtures ofof boronboron trioxidetrioxide andand ZSM-5ZSM-5 withwith aluminaalumina binderbinder showshow aa characteristic characteristic 1 OHOH onon boronboron trioxide trioxide at at 3704 3704 cm cm−−1 (Figure(Figure4 4a)a) and and planar planar BO BO33 structures atat 1404,1404, 1244,1244, 1199,1199, 807,807, 716,716, 1 andand 646646 cmcm−−1 asas shown shown in in Figure Figure 5a.5a. Here Here the the boron boron trioxide trioxide powder powderss were were prepared prepared by by grinding grinding the theglassy glassy boron boron trioxide trioxide large large particles particles into into fine fine powders. powders. Some Some nonuniformity nonuniformity may may exist exist during during the thegrinding grinding and and mixing mixing processes processes that that may may affect affect the the borate borate peak peak intensity intensity slightly, forfor exampleexample thethe 1 sharpsharp peakpeak at 646 cm cm−1− forfor th thee physic physicalal mixture mixture sample sample with with 0.25 0.25 wt.% wt.% B2O3 B may2O3 mayindicate indicate longer longer range rangeordering ordering structures structures for larger for particles. larger particles. A noticeable A noticeable boron-silicon boron-silicon exchange exchange interaction interaction can be clearly can beseen clearly in the seen light in blue the lightspectrum blue for spectrum the physical for the mixture physical sample mixture with sample 1.0 wt. with% B2O 1.03 with wt.% dramatically B2O3 with 1 dramaticallydecreased Si decreased-O-Si peak Si-O-Si intensity peak at intensity 1094 cm at−1. 1094The cmFTIR− . spectra The FTIR on spectra the physically on the physically mixed samples mixed samples(B2O3/ZSM (B2-O5 3 with/ZSM-5 alumina with alumina binder) binder)provide provide IR features IR features associated associated with fine with powder fine powder B2O3 with B2O 3 a 1 withcharacteristic a characteristic B-OH B-OH(3704 cm (3704−1) and cm− planar) and planarBO3 structures. BO3 structures. TheThe samplessamples wherewhere thethe preparationpreparation waswas completedcompleted withwith aa 649649◦ °CC calcinationcalcination stepstep areare presentedpresented inin FiguresFigures4 b4b and and5b. 5b. Experiments Experiments were were conducted conducted at at 1 1 wt.% wt.% B B2OO33 concentrationconcentration in in order order to to serve serve a acomparison comparison basis basis against against the the physical physical mixture mixture sampl samples;es; and and also also at 2 atwt.% 2 wt.% B2O3 Blevel2O3 tolevel see tothe see nature the natureof transformation of transformation at higher at higher boron boron loadings. loadings. In both In both cases, cases, after after high high temperature temperature calcination, calcination, the 1 thecharacteristic characteristic B-OH B-OH peak peak disappears. disappears. The The BO BO3 3featurefeature at at 1404 1404 cm−1− shiftsshifts to higher wavenumbers wavenumbers 1 1 (~1505(~1505 cm cm− −1),), and and the the bending bending B-O-B B-O for-B BO for3 features BO3 features (807, 716, (807, and 716 646, and cm − 646) are cm significantly−1) are significantly reduced. 1 Thereduced. BO4 feature The BO at4 936feature cm− atappears. 936 cm−1 These appears. changes These are changes associated are withassociated the breakdown with the ofbreakdown planar BO of3 structuresplanar BO and3 structures formation and of formation borate BO of3 and borate BO 4BOunits.3 and BO4 units. AsAs shownshownin in thethe resultsresults presentedpresentedin inFigures Figures4 4and and5, 5, with with the the choice choice of of right right reference reference samples samples andand controlcontrol experiments,experiments, FTIRFTIR cancan serveserve asas aa sensitivesensitive techniquetechnique forfor thethe characterizationcharacterization ofof boronboron trioxide and borate species. The isolated B2O3 phase can be detected by FTIR at low concentrations

Catalysts 2020, 10, 1327 7 of 11

(0.1%Catalysts by weight 2020, 10 as, x inFOR provided PEER REVIEW samples). The small borate BO3 and BO4 units can be detected7 of at12 0.5% by weight even in actual industrial catalyst samples as presented in Figure3. trioxide and borate species. The isolated B2O3 phase can be detected by FTIR at low concentrations 2.3. Characterization(0.1% by weight as of Boronin provided via Other samples). Spectroscopic The small Techniques borate BO3 and BO4 units can be detected at 0.5% by weight even in actual industrial catalyst samples as presented in Figure 3. For the second phase of the investigation other spectroscopic analyses, namely Raman, XPS and NMR,2.3. were Characterization carried out of to boron explore via other their spectroscopic potential techniques utilization for characterization of boron species in FCC catalysts and additives for Environmental Health and Safety (EHS) purposes as described above For the second phase of the investigation other spectroscopic analyses, namely Raman, XPS and in SectionNMR,1 .were Again, carried the ZSM-5out to explore additive their system potential was utilization chosen as for the characterization model case for of this boron part species of the in study. Pilot-scaleFCC catalysts ﬂuidizable and additives FCC microsphere for Environmental samples Health were preparedand Safety with (EHS 1) purposes wt.% B2O as3 describedloading as above well as a referencein section sample 1. Again, without the ZSM boron.-5 additive For validating system was the chosen complete as the transformationmodel case for this and part disappearance of the study. of boronPilot trioxide-scale fluidizable under typical FCC FCC microsphere catalyst /samplesadditive were processing prepared conditions, with 1 wt.% a B sample2O3 loading with as higher well as B2O3 contenta reference was also sample prepared without at boron. 2 wt.% For B2 validatingO3 loading. the complete In order transformation to ensure that and there disappearance is no interference of withboron other trioxide species, under this sample typical FCC contained catalyst/additive only Boron processing on the silicoaluminate conditions, a sample matrix with support higher B and2O3 was preparedcontent without was also any prepared zeolite at or 2 binder.wt.% B2O3 loading. In order to ensure that there is no interference with other species, this sample contained only Boron on the silicoaluminate matrix support and was Raman Spectra in Figure6 shows the comparison of calcined FCC additive Samples #12 and prepared without any zeolite or binder. #13 (collected using a 488-nm laser) and a silicoaluminate support loaded with B O before and after Raman Spectra in Figure 6 shows the comparison of calcined FCC additive Samples2 3 #12 and #13 calcination using a 785-nm excitation laser, where the spectra collected under diﬀerent lasers were (collected using a 488-nm laser) and a silicoaluminate support loaded with B2O3 before and after 1 normalizedcalcination using using peak a 785 intensity-nm excitation at 635 cmlaser,− . where In this the experiment, spectra collected a reference under sampledifferent of lasers standard were pure boronnormalized trioxide was using also peak included intensity and at 635 showed cm−1. In Raman this experiment, shifts at 258, a reference 271, 391, sample 408, 425, of standard 486, 648, pure 854, 877, 1 1 1012,boron and 1274 trioxide cm− was, where also included peaks at and 258, showed 271, 391, Raman 408, 425,shifts 648, at 258, 854, 271, 877 391, cm− 408,are 425, assigned 486, 648, to 854, BO3 and −1 1 −1 peaks877, at 486,1012,1012, and 1274 1274 cm cm, −whereare peaks assigned at 258, to 271, BO4 391,[24 –408,28]. 425, The 648, silicoaluminate 854, 877 cm are support assigned loaded to BO3 with and peaks at 486, 1012, 1274 cm−1 are assigned to BO4 [24-28]. The silicoaluminate support loaded B2O3 (before calcination) shows these B2O3 Raman peaks, while the calcined samples do not have any with B2O3 (before calcination) shows these B2O3 Raman peaks, while the calcined samples do not have of these peaks. This result demonstrates that after calcination, the precursor B2O3 phase disappears, any of these peaks. This result demonstrates that after calcination, the precursor B2O3 phase which is likely due to the formation of borate species. disappears, which is likely due to the formation of borate species.

After calcination, Sample #13

After calcination, Sample #12

After calcination, Sample #14

Before calcination, Sample #14

Intensity (a.u.)

877

854

1012

805

486

Boron trioxide, Sample #2

408

391

5000 425

400 600 800 1000 Raman Shift (cm-1)

FigureFigure 6. Raman 6. Raman spectra spectra collected collected for for FCC FCC additive additive samples samples (#12(#12 and #13) using using a a488 488-nm-nm laser laser and and an FCC catalystan FCC catalyst sample sa #14mple using #14 ausing 785-nm a 785 laser-nm laser before before (the (the 4th trace4th trace from from top) top) and and after after (the (the 1st, 1st, 2nd2nd and 3rd tracesand 3rd from traces top) from calcination top) calcination with comparisonwith comparison to pure to pure boron boron trioxide trioxide (bottom (bottom trace) trace) reference referenceshow that boronshow that trioxide boron phase trioxide disappears phase disappears after calcination. after calcination. The top The three top spectrathree spectra were normalizedwere normalized based on based on the peak intensity1 at 635 cm−1 to the elemental content of TiO2 in the two samples. the peak intensity at 635 cm− to the elemental content of TiO2 in the two samples.

Catalysts 2020, 10, 1327 8 of 11

Raman spectra in Figure7 show a close look of the top two traces of spectra in Figure6, which demonstrate the presence of BO4 borate unit in the ZSM-5 sample with boron. Compared to the reference sample #12 without any boron oxide precursor, the sample #13 with 1 wt.% B2O3 as precursor 1 shows an additional signal at 463 cm− , which is assigned to BO4 unit vibration based on the peak assignments using a commercial standard phosphor borate material. Other peaks at 390, 514, 637, 1 806 cm− exist both in the control sample (no boron) and the 1% boron addition sample, which are possibly due to some TiO2 present in the samples. Catalysts 2020, 10, x FOR PEER REVIEW 8 of 12

BO4

637

390

806

514

463

733

Intensity (a.u.)

0% B O / ZSM-5 with binder 200 2 3 1% B2O3 / ZSM-5 with binder

300 400 500 600 700 800 900 1000 Raman Shift (cm-1)

FigureFigure 7. Raman7. Raman spectra spectra collectedcollected using using a 488 a 488-nm-nm laser laser for Samples for Samples #12 and #12 #13. and This #13.is a higher This- is a higher-resolutionresolution view view of the of top the two top tra twoces traces of spectra of spectra in Figure in 6. Figure 6.

In additionRaman tospectra FTIR andin Figure Raman 7 show spectroscopy a close look methods, of the top additional two traces characterization of spectra in Figure techniques 6, which such as XPSdemonstrate and NMR thespectroscopy presence of experiments BO4 borate unit were in also the ZSM conducted-5 sample for with boron boron. material Compared characterization to the as shownreference in the sample Supplementary #12 without Materials. any boron Since oxide the precursor, FCC catalysts the / sampleadditives #13 contains with 1 wt. silicon% B2 oxideO3 as and precursor shows an additional signal at 463 cm−1, which is assigned to BO4 unit vibration based on some phosphor components/impurities, XPS analysis shows B 1s signal overwhelmed by P 2s and Si 2s the peak assignments using a commercial standard phosphor borate material. Other peaks at 390, plasma peaks, which agrees with results in literature [10]. As expected [11], the 11B NMR analysis 514, 637, 806 cm−1 exist both in the control sample (no boron) and the 1% boron addition sample, shows very similar BO chemical shifts for standard materials aluminum borate and boron trioxide, which are possibly due3 to some TiO2 present in the samples. which makesIn addition it challenging to FTIR toand precisely Raman spectroscopy identify boron methods, phases additional in practice. characterization techniques Insuch summary, as XPS the and Boron NMR characterization spectroscopy experiments using infrared were spectroscopy also conducted is clearly for boron a better material approach comparedcharacterization to the Raman as shown spectroscopy, in the Supplementary XPS and NMR. Materials The. borateSince the signal FCC cancatalysts/additives be significantly contains improved in FT-IRsilicon analysis oxide with and all some other phosphor signals in components/impurities, the supporting materials XPS subtracted analysis shows when processedB 1s signal using a properlyoverwhelmed selected by backgroundP 2s and Si 2s plasma spectrum. peaks, However, which agrees the with Raman results Spectrum in literature reads [10]. As counts expected directly, 11 where[11] the, the borate B NMR signals analysis might shows be overwhelmed very similar BO easily3 chemical by other shifts signals for standard from the materials supporting aluminum materials. borate and boron trioxide, which makes it challenging to precisely identify boron phases in practice. 3. MaterialsIn summary, and Methods the Boron characterization using infrared spectroscopy is clearly a better approach compared to the Raman spectroscopy, XPS and NMR. The borate signal can be significantly improved 3.1. Samplein FT-IR Information analysis with all other signals in the supporting materials subtracted when processed using a properly selected background spectrum. However, the Raman Spectrum reads counts directly, Thewhere detailed the borate sample signals information might be is overwhelmed listed in Table easily1. The by boron other oxide signals content from wasthe supporting analyzed and confirmedmaterials. by X-ray fluorescence (XRF) spectroscopy analysis. Samples #1–11 were characterized in the boron trioxide and borate analysis. There are reference samples such as boron trioxide, phosphor borate3. and Materials aluminum and Methods borate. Sample #4 is the industrial FCC catalyst sample utilizing the Boron Based Technology. Samples #5–8 are physical mixtures of boron trioxide with ZSM-5 additive microspheres 3.1. Sample information The detailed sample information is listed in Table 1. The boron oxide content was analyzed and confirmed by X-ray fluorescence (XRF) spectroscopy analysis. Samples #1-11 were characterized in the boron trioxide and borate analysis. There are reference samples such as boron trioxide, phosphor borate and aluminum borate. Sample #4 is the industrial FCC catalyst sample utilizing the Boron Based Technology. Samples #5-8 are physical mixtures of boron trioxide with ZSM-5 additive

Catalysts 2020, 10, 1327 9 of 11

that consist of ZSM-5 with alumina binder, P2O5 and clay to serve as control experiments against Samples #9–11 are identical to Samples #5–8 in composition but are subjected a calcination step typical to FCC catalyst/additive manufacturing as a final step. Samples #5–11 are representing the main ingredients of an FCC additive preparation recipe with various amounts of boron loaded on ZSM-5 additive microspheres as supporting materials; but only for Samples #9–11 the preparation is finished by a calcination step to demonstrate the impact of that step on the nature of boron species in a typical FCC catalyst/additive manufacturing procedure. A typical ZSM-5 additive recipe consisting of ZSM-5 with alumina binder, P2O5 and clay was chosen as the model system to study via controlled experiments with tailor-made reference samples in order to achieve a cleaner comparison. This is mainly because ZSM-5 additives are typically simpler in design compared to base FCC catalysts since cracking matrices and metals traps that are part of the resid-FCC catalyst design are typically not included in the additive construct [29,30]. Samples #12–13 are pilot-scale samples generated following a FCC additive preparation procedure with 0% and 2% B2O3 in a mixture of ZSM-5, alumina binder, P2O5, and clay, followed by a calcination step typical to FCC catalyst/additive manufacturing. The resulting materials are fluidizable ZSM-5 additive microspheres with physical properties characteristic of commercial FCC particles.

Table 1. Powder sample IDs and detail description.

Sample ID Description Sample Contents and Treatment

1 Phosphor borate, PBO4, Alfa Aesar commercial material 2 Boron trioxide commercial material 3 aluminum borate, Al18B4O33, BOC Sciences commercial material 4 0.5% B2O3/USY zeolite, (silico)aluminate matrices Cyclic Propylene Steaming (CPS) deactivated 5 0.1% B2O3/ZSM-5 with alumina binder physical mixture 6 0.25% B2O3/ZSM-5 with alumina binder physical mixture 7 0.5% B2O3/ZSM-5 with alumina binder physical mixture 8 1% B2O3/ZSM-5 with alumina binder physical mixture 9 1% B2O3/ZSM-5 with alumina binder 649 ◦C-calcination 10 1.5% B2O3/ZSM-5 with alumina binder 649 ◦C-calcination 11 2% B2O3/ZSM-5 with alumina binder 649 ◦C-calcination 12 0% B2O3 ZSM-5 with alumina binder Fluidizable microspheres 13 1% B2O3/ZSM-5 with alumina binder Fluidizable microspheres 14 2% B2O3/Silicoaluminate matrix Fluidizable microspheres

3.2. Raman Spectroscopy Raman spectra were collected on a Renishaw InVia Raman microscope (Renishaw plc., Gloucestershire, United Kingdom) with a 785-nm laser as the excitation source. A Leica N PLAN microscope/50 objective lens (Leica Microsystems, Wetzlar, Germany) was used to focus the 785-nm × laser beam on the sample surface at 1% in power usage.

3.3. Infrared Spectroscopy Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) experiments were performed on a Varian 7000 FTIR spectrometer (Agilent Technologies, Santa Clara, CA, United States) equipped with an MCT detector and a DiffusIR high-temperature environmental chamber with a KBr window (Pike Technologies, Madison, WI, USA). Spectrum collection was performed under diffuse reflection mode. The samples were ground into fine powders with a mortar and pestle, and then packed into the sample cup. The samples were dehydrated in Ar at 200 or 400 ◦C for 1 h with a ramp rate of 20 ◦C/min. IR spectra were collected after dehydration, once the chamber had been cooled to 30 ◦C. 4. Conclusions Environmental Health and Safety (EHS) classification of industrial catalysts requires suitable methods to be able to analyze and identify—as well as confirm the absence of—regulated species. In this Catalysts 2020, 10, 1327 10 of 11 contribution an FTIR Spectroscopy methodology that provides phase and structural characterization of boron-containing FCC catalysts is presented. It shows that FCC catalysts with Boron Based Technology (BBT) do not contain any detectable amount of boron trioxide. It also shows that the typical FCC catalyst synthesis procedures would transform boron trioxide if it were used, since the FCC microspheres go through the high-temperature calcination process that breaks down boron trioxide planar BO3 network structures to form small borate BO3 and BO4 units. FTIR proves to be a sensitive method to differentiate boron trioxide and borate structures with a detection limit at 0.1% by weight. Other spectroscopic techniques such as Raman, NMR, and XPS were also utilized as part of this investigation. Even though Raman spectroscopy provides insights on the complete transformation of any boron trioxide (if it were to be used as a precursor) in the FCC catalysts/additive after calcination, and confirms the absence of boron trioxide in the final product, it did not provide clear signals of the new borate species at the same sensitivity level as FTIR. Thus, Raman spectroscopy together with XPS and NMR techniques were not seen as suitable standalone methods for phase identification. These observations highlight the importance of the FTIR based method for utilization in EHS studies. Since boron trioxide is regulated in parts of the world, this new FTIR method is beneficial for the application of boron-containing FCC catalysts as well as other wide range applications of boron-containing materials in the fields of glass, ceramics, semiconductors, agriculture, and pharmaceuticals.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4344/10/11/1327/s1, Figure S1: XPS P 2s and B 1s spectra for FCC ZSM-5 additive samples without boron (top) vs. with 1 wt.% B2O3 (bottom). B 1s region is overlapping with P 2s region, Figure S2: 11B NMR spectrum of ZSM-5 additive with 11 1 wt.% B2O3, Figure S3: B NMR spectra for (a) Al-borate and (b) B-phosphate (c) Boron trioxide. Author Contributions: Conceptualization, C.C.Z., X.G. and B.Y.; methodology, data curation, visualization, investigation, C.C.Z., X.G.; writing—original draft preparation, writing—review and editing, C.C.Z., X.G. and B.Y. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: The authors thank Jasmine Luo, Gary M. Smith, Sage Hartlaub, and Ivan Petrovic for their support for this work. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Pan, S.; Shackleford, A.; McGuire, R., Jr.; Smith, G.M.; Yilmaz, B. Creative catalysis. Hydrocarb. Eng. 2015, 20, 46–52. [CrossRef] 2. European Chemicals Agency. Available online: https://echa.europa.eu/registration-dossier/-/registered- dossier/6449/1 (accessed on 15 March 2019). 3. European Chemicals Agency. Available online: https://echa.europa.eu/documents/10162/6b35228c-45c4- 4ed8-88b0-823fafb0d795 (accessed on 15 March 2019). 4. Duydu, Y.; Ba¸saran,N.; Ustundag, A.; Aydın, S.; Ündeger, Ü.; Ataman, O.Y.; Aydos, K.; Düker, Y.; Ickstadt, K.; Waltrup, B.S.; et al. Assessment of DNA integrity (COMET assay) in sperm cells of boron-exposed workers. Arch. Toxicol. 2012, 86, 27–35. [CrossRef] 5. Duydu, Y.; Ba¸saran,N.; Ustundag, A.; Aydın, S.; Ündeger, Ü.; Ataman, O.Y.; Aydos, K.; Düker, Y.; Ickstadt, K.; Waltrup, B.S.; et al. Reproductive toxicity parameters and biological monitoring in occupationally and environmentally boron-exposed persons in Bandırma, Turkey. Arch. Toxicol. 2011, 85, 589–600. [CrossRef] 6. Glass Alliance Europe. Available online: https://www.glassallianceeurope.eu/images/cont/gae-reach- position-paper-on-boron-trioxide_1_ﬁle.pdf (accessed on 15 March 2019). 7. IPC—Association Connecting Electronics Industries. Available online: http://www.ipc.org/4.0_Knowledge/4. 1_Standards/Position-Statement-3-12d-TG_REACH-SVHC_E-Glass-B2O3.pdf (accessed on 15 March 2019). 8. Intel. Available online: https://www.intel.com/content/dam/www/programmable/us/en/pdfs/support/ reliability/environmental/reach/intel-dcl-product-content-reach-oct2017.pdf (accessed on 15 March 2019). 9. Berlin, J. Analysis of boron with energy dispersive X-ray spectrometry. Imaging Microsc. 2011, 13, 19–21. Catalysts 2020, 10, 1327 11 of 11

10. Ong, C.W.; Huang, H.; Zheng, B.; Kwok, R.W.M.; Hui, Y.Y.; Lau, W.M. X-ray photoemission spectroscopy of nonmetallic materials: Electronic structures of boron and BxOy. J. Appl. Phys. 2004, 95, 3527. [CrossRef] 11. Sen, S.; Xu, Z.; Stebbins, J.F. Temperature dependent structural changes in borate, borosilicate and boroaluminate liquids: High-resolution 11B, 29Si and 27Al NMR studies. J. Non-Cryst. Solids 1998, 226, 29–40. [CrossRef]

12. Galeener, F.L.; Lucovsky, G.; Mikkelsen, J.C., Jr. Vibrational spectra and the structure of pure vitreous B2O3. Phys. Rev. B 1980, 22, 3983–3990. [CrossRef] 13. Gautam, C.; Yadav, A.K.; Singh, A.K. A review on infrared spectroscopy of borate glasses with effects of different additives. ISRN Ceram. 2012, 2012, 17. [CrossRef] 14. Kmeel, P.; Bukovee, P. Boron phosphate: Its synthesis, gradual crystallisation and characterisation of bulk properties. Acta Chim. Slov. 1999, 46, 161–171. 15. Baykal, A.; Kizilyalli, M.; Toprak, M.; Kniep, R. Hydrothermal and microwave synthesis of boron phosphate, BPO4. Turk. J. Chem. 2001, 25, 425–432. 16. Kamitsos, E.; Patsis, A.; Karakassides, M.; Chryssikos, G. Infrared reflectance spectra of lithium borate glasses. J. Non-Cryst. Solids 1990, 126, 52–67. [CrossRef] 17. Stoch, L.; Sroda,´ M. Infrared spectroscopy in the investigation of oxide glasses structure. J. Mol. Struct. 1999, 511–512, 77–84. [CrossRef]

18. Ram, S. Infrared study of the dynamics of boroxol rings in the crystallization of BaFe12O19 microcrystals in borate glasses. Phys. Rev. B 1995, 51, 6280. [CrossRef][PubMed] 19. Pisarski, W.A.; Pisarska, J.; Ryba-Romanowski, W. Structural role of rare earth ions in lead borate glasses

evidenced by infrared spectroscopy: BO3 <-> BO4 conversion. J. Mol. Struct. 2005, 744–747, 515–520. [CrossRef] 20. Gielisse, P.J.M.; Foster, W.R. The system Al2O3–B2O3. Nature 1962, 195, 69. [CrossRef] 21. Gutsev, G.L.; Weatherford, C.A.; Johnson, L.E.; Jena, P. Structure and properties of the aluminum borates Al(BO )n and Al(BO )n , (n = 1–4). J. Comput. Chem. 2011, 33, 416–424. [CrossRef] 2 2 − 22. Blay, V.; Louis, B.; Miravalles, R.; Yokoi, T.; Peccatiello, K.A.; Clough, M.; Yilmaz, B. Engineering zeolites for catalytic cracking to light oleﬁns. ACS Catal. 2017, 7, 6542–6566. [CrossRef] 23. Clough, M.; Pope, J.C.; Lin, L.T.X.; Komvokis, V.; Pan, S.S.; Yilmaz, B. Nanoporous materials forge a path forward to enable sustainable growth: Technology advancements in ﬂuid catalytic cracking. Microporous Mesoporous Mater. 2017, 254, 45–58. [CrossRef]

24. Rulmont, A.; Almou, M. Vibrational spectra of metaborates with infinite chain structure: LiBO2, CaB2O4, SrB2O4. Spectrochim. Acta Part A Mol. Spectrosc. 1989, 45, 603–610. [CrossRef] 25. Umari, P.; Pasquarello, A. Fraction of boroxol rings in vitreous boron oxide from a first-principles analysis of raman and NMR spectra. Phys. Rev. Lett. 2005, 95, 137401. [CrossRef] 26. Simon, G.; Hehlen, B.; Vacher, R.; Courtens, E. Courtens, Hyper-Raman scattering analysis of the vibrations in vitreous boron oxide. Phys. Rev. B 2007, 76, 054210. [CrossRef] 27. Voronko, Y.K.; Sobol, A.A.; Shukshin, V.E. Structure of boron-oxygen groups in crystalline, molten, and glassy alkali-metal and alkaline-earth metaborates. Inorg. Mater. 2012, 48, 732–737. [CrossRef] 28. Barrio, R.A.; Castillo-Alvarado, F.L.; Galeener, F.L. Structural and vibrational model for vitreous boron oxide. Phys. Rev. B 1991, 44, 7313. [CrossRef][PubMed] 29. Pan, S.S.; Lin, L.T.X.; Komvokis, V.; Spann, A.; Clough, M.; Yilmaz, B. Nanomaterials fueling the world. In Nanomaterials for Sustainable Energy; ACS Symposium Series; American Chemical Society: Washington, DC, USA, 2015. 30. Komvokis, V.; Tan, L.X.; Clough, M.; Yilmaz, B. Zeolites in fluid catalytic cracking (FCC). In Zeolites in Sustainable Chemistry; Springer: Berlin/Heidelberg, Germany, 2016.

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

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).