Precipitation of potassium as hazenite from washing water of spent alkaline batteries

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY-NC-ND 4.0 h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ P ecipi a ion o po assium as hazeni e om washing wa e o spen alkaline ba e ies © 2022 The Au ho s. Published by Else ie B.V. Published e sion Lapinkangas, Su i; Rau io, Lasse; Kauppinen, Toni; Hu, Tao; Pesonen, Janne; Lassi, Ulla Lapinkangas, S., Rau io, L., Kauppinen, T., Hu, T., Pesonen, J., & Lassi, U. (2022). P ecipi a ion o po assium as hazeni e om washing wa e o spen alkaline ba e ies. Chemical Enginee ing Jou nal Ad ances, 12, A icle 100426. h ps://doi.o g/10.1016/j.ceja.2022.100426 2022 Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 A ailable online 15 No embe 2022 2666-8211/© 2022 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by- nc-nd/4.0/). P ecipi a ion o po assium as hazeni e om washing wa e o spen alkaline ba e ies Su i Lapinkangas a , Lasse Rau io a , b , Toni Kauppinen a , c , Tao Hu a , Janne Pesonen a , * , Ulla Lassi a , c a Uni e si y o Oulu, Resea ch Uni o Sus ainable Chemis y, P.O. Box 3000, FI-90014, Finland b T aceg ow Oy, Teollisuus ie 21, K¨ a s¨ am¨ aki, Finland c Kokkola Uni e si y Conso ium Chydenius, Uni e si y o Jy ¨ askyl¨ a, Talonpojanka u 2 B, FI-67100 Kokkola, Finland ARTICLE INFO Keywo ds: S u i e Hazeni e P ecipi a ion Alkaline ba e y Black mass Fe ilize ABSTRACT Hazeni e (KNaMg 2 (PO 4 ) 2 ×14 H 2 O), a new ype o s u i e mine al, was p ecipi a ed om he po assium- ich washing wa e o spen alkaline ba e y black mass. Hazeni e can be used as a e ilize , which would be an addi ional bene i de i ed om he sus ainable eco e y o ba e y ma e ials. P ecipi a ion expe imen s we e pe o med using di e en pH alues (9.5–12), Mg:K:PO 4 a ios ((1.0–1.5):1:(1.0–1.5)) and empe a u es (10–40 ◦C). Based on he esul s, hazeni e p ecipi a ed in a wide pH ange unde alkaline condi ions. The p e- cipi a ion kine ics we e as , and he pu i y o he hazeni e was high. O e all, hazeni e can be p ecipi a ed a oom empe a u e wi hou he addi ion o excess chemicals, which minimizes he consump ion o chemicals and ene gy. 1. In oduc ion Phospho us and po assium a e essen ial nu ien s o c ops, animals and humans. Mos o he phospho us and po assium used a e de i ed om mine al s ones, making hem non- enewable esou ces. Thus, o he sou ces o phospho us and po assium a e needed o ensu e he s able and sus ainable de elopmen o ag icul u e [1]. Po assium in was ewa e can be emo ed simul aneously wi h phospho us and magnesium in alkaline condi ions (pH a ound 11) o p oduce K-s u i e (KMgPO 4 ×6 H 2 O), which is a po assium analog o s u i e (NH 4 MgPO 4 ×6 H 2 O) [2]. Hazeni e (KNaMg 2 (PO4) 2 ⋅14H2O) is a newly disco e ed s u i e g oup mine al ha con ains wo s uc u ally dis inc mono alen ca ions (Na + and K + ) [3]. The i s obse a ion o hazeni e was made a he highly alkaline (pH ≈10) and saline (84–92 g/l) Lake Mono in Cali o nia, whe e i was syn hesized by biodeg adable cyanobac e ia. Like s u i e and K-s u i e, hazeni e could also be sui able o e ilize use [3,4]. In a ecen s udy, hazeni e eleased phospho us slowe han con en ional comme cial wa e -soluble e il- ize s bu as e han s u i e [5]. The e o e, hazeni e could be a po en ial phospho us sou ce o long- e m elease. Indeed, hazeni e inc eased he biomass o suga cane by 21% compa ed o comme cial supe phospha e in a 180-day g ow h expe imen [6]. In his esea ch, hazeni e p ecipi a ion was s udied using he washing wa e o spen alkaline ba e ies as a sou ce o po assium, which c ea es an addi ional bene i o he sus ainable eco e y o ba - e y ma e ials. Spen alkaline ba e ies a e haza dous was e, and he ea men o hem is needed. As ba e y black mass is washed, a po assium- ich was ewa e is p oduced. He e, he po assium is p ecipi- a ed om he washing wa e as hazeni e, which is a po en ial po assium e ilize . This educes he need o use p ima y aw ma e ials o po- assium e ilize s, since po assium is mos ly eco e ed om mine al s ones, making i a non- enewable esou ce. Fu he , he elease o po- assium in o wa e cou ses om was ewa e s can be p e en ed using his p ecipi a ion p ocedu e. Alkaline ba e ies use po assium hyd oxide as an elec oly e, whe eas zinc-ca bon ba e ies use ammonium chlo ide. When alkaline ba e ies a e collec ed and c ushed, a small ac ion o zinc-ca bon ba e ies o en eside among he alkaline ones. The e o e, when po as- sium is sepa a ed om he c ushed ba e y mass by washing wi h wa e , some ammonium also ends up in he washing wa e [7,8]. Cu en ly, hese po assium- ich washing wa e s end up in municipal was ewa e ea men plan s o pu i ica ion. I is also possible o hyd o- me allu gically ea he ba e y mass wi hou emo ing he po assium i he p esence o po assium in he end solu ion is app o ed. The e o e, he * Co esponding au ho . E-mail add ess: [email p o ec ed] (J. Pesonen). Con en s lis s a ailable a ScienceDi ec Chemical Enginee ing Jou nal Ad ances jou nal homepage: www.sciencedi ec .com/jou nal/chemical-enginee ing-jou nal-ad ances h ps://doi.o g/10.1016/j.ceja.2022.100426 Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 2 washing phase is no always pe o med. Howe e , he ele a ed po as- sium con en in he zinc and manganese con aining sul a e solu ions could suppo o ma ion o Tu on’s sal s [9,10]. The e o e, o p oduce mo e concen a ed ZnSO 4 and MnSO 4 solu ions, emo al o he po as- sium om he sys em is needed. P ecipi a ion expe imen s we e pe - o med using di e en pH alues (9.5–12), Mg:K:PO 4 a ios ((1.0–1.5):1: (1.0–1.5)) and empe a u es (10–40 ◦C). The e is only a e y limi ed numbe o publica ions ela ed o hazeni e, and his is he i s in which he p ecipi a ion condi ions o hazeni e ha e been s udied. 2. Expe imen al 2.1. Po assium- ich washing wa e The washing wa e o spen alkaline ba e ies was p o ided by T aceg ow Oy (K¨ a s¨ am¨ aki, Finland). T aceg ow is a company ha p o- duces zinc and manganese con aining ace elemen e ilize s om spen alkaline ba e ies. In his expe imen , c ushed and sie ed alkaline black mass was dosed o a 220 l con aine , and wa e was pou ed among he black mass (0.8 kg wa e /1 kg black mass). The mix u e was hen mixed wi h a ba el mixe o abou 1 h. A e his, he mixing was s opped, and he black mass was le o sedimen a he bo om o he ba el o one day. A e his, abou hal o he wa e was anspo ed om he su ace laye wi h a pump o a sepa a e con aine . The solu ion con ained ine g aphi e, which had o be il e ed away wi h a il e p ess. The esul was g een-shaded po assium- ich alkaline washing wa e (Table 1), which was hen deli e ed o he Uni e si y o Oulu. Po as- sium, sodium and magnesium concen a ions o he washing wa e we e measu ed using a omic abso p ion spec oscopy AAS (Va ian AA240FS, Aus alia) and ammonium concen a ion was measu ed using an HACH HQ40d- NH 4 -selec i e elec ode (Model ISENH418101, Lo eland, CO, USA). 2.2. P ecipi a ion expe imen s Po assium- ich washing wa e was dilu ed in a a io o 1:5 be o e he p ecipi a ion expe imen s due o i s abundan po assium con en . 1.6 l o dilu ed washing wa e was used in each p ecipi a ion expe imen . Du ing he p ecipi a ion expe imen s, he e ec s o se e al pa ame e s (pH, mola a ios and empe a u e) we e s udied (Table 2). The o al p ecipi a ion ime was 4 h, and samples we e wi hd awn a e 0, 15, 30, 60, 90, 120, 180 and 240 min. All p ecipi a ion expe imen s we e con- duc ed in duplica e o iplica e. A e he expe imen , he p ecipi a e was sepa a ed by il a ion, and he solid p ecipi a e was d ied a oom empe a u e in a desicca o o i e days. In each expe imen , phospha e sal (NaH 2 PO 4 x 2 H 2 O; AnalaR NORMAPUR, 99.0–102.0%, VWR Chemicals, Ge many) was added o he washing wa e in he p ecipi- a ion eac o and mixed (500 pm) by an o e head s i e (IKA EUROSTAR 40 digi al o e head s i e ; IKA®-We ke GmbH & CO.KG, Ge many) wi h a 2-blade impelle and s a o s un il he sal was comple ely dissol ed. The pH was adjus ed o he desi ed alue wi h 10 M NaOH (AnalaR NORMAPUR, 99.2%, VWR Chemicals, Sweden). P e- cipi an (solid magnesium sal , MgSO 4 x7H 2 O, pH. Eu ., 99.6%, VWR Chemicals, Belgium) was hen added while s i ing he solu ion a a cons an speed (1 min 500 pm, hen 250 pm o 50 pm). The pH o he solu ion was kep cons an by using 10 M NaOH and concen a ed HCl (37%, EMSURE®, Me ck KGaA, Ge many). The e ec o he ammonium ions was s udied by ca ying ou a wo-s ep p ecipi a ion. In he i s s ep, he pH was se o 9 and ammonium was a ge ed by ca ying ou s u i e p ecipi a ion; he pH was hen aised o 11 and he po assium was p ecipi a ed as hazeni e. 2.3. Cha ac e iza ion o samples Liquid samples we e il a ed h ough 0.45 um il e pape . Magne- sium and po assium concen a ions we e measu ed using a omic ab- so p ion spec oscopy AAS (Va ian AA240FS, Aus alia) using s anda d e e ence solu ions. Phospha e concen a ions we e measu ed using ion ch oma og aphy IC (Me h ohm 761 Compac IC; Swi ze land). Solid p ecipi a es we e analyzed using X- ay di ac ion (XRD). The XRD pa e ns we e eco ded by a PANaly ical X´Pe P o-X- ay di ac- ome e (Mal e n Panaly ical, Almelo, The Ne he lands) using mono- ch oma ic CuK α 1 adia ion (λ =1.5406 Å) a 45 kV and 40 mA. Di ac og ams we e collec ed in he 2θ ange o 10–85◦a 0.017◦in- e als. The c ys alline phases and s uc u es we e analyzed using HighSco e Plus so wa e (Ve sion 4.0, PANaly ical B.V., Almelo, he Ne he lands). The elemen al analysis was measu ed using a PANaly ical Axios mAX X- ay luo escence spec ome e . The sample was ed as loose powde h ough Myla ilm in He-a mosphe e. A geos anda d was used and he con ibu ion o pu e Myla ilm was sub ac ed om he measu emen esul s. 3. Resul s and discussion 3.1. E ec o ammonium on hazeni e p ecipi a ion The washing wa e con ained ammonium ions, as desc ibed in Sec- ion 2.1. This migh ha e a ec ed he p ecipi a ion and pu i y o he p ecipi a e. The e o e, he e ec o ammonium ions was accoun ed o by ca ying ou s u i e p ecipi a ion ollowed by hazeni e p ecipi a ion ( wo-s ep p ecipi a ion). In he s u i e p ecipi a ion expe imen s, he mola a io o Mg:N:PO 4 was 1:1:1, he pH was 9 and he mixing a e was 50 pm. The ob ained il a e was u he p ecipi a ed wi h an Mg:K:PO 4 o 1:1:1, a pH o 11 and a mixing a e o 50 pm. In addi ion, he hazeni e was di ec ly p ecipi a ed om he washing wa e (one-s ep p ecipi a- ion) wi hou he p eceding p ecipi a ion o ammonium. In hese ex- pe imen s, he condi ions we e equal (a Mg:K:PO 4 o 1:1:1, a pH o 11 and a mixing a e o 50 pm). The emo al o po assium, magnesium and phospha e in he one- and wo-s ep p ecipi a ion p ocedu es is p esen ed in Fig. 1. Simila p e- cipi a ion e iciencies o po assium we e obse ed a 54% and 56% o Table 1 Composi ion o he washing wa e . K mg/l NH 4 mg/l Mg (mg/l) Na (mg/l) pH 34,500 1700 31 1220 10.8 Table 2 P ecipi a ion condi ions. Va iable O he pa ame e s E ec o pH 9.5 10.0 10.5 11.0 11.5 12.0 Mola a io Mg:K:PO 4 1:1:1 T =20 ◦C =4 h E ec o mola a io Mg:K:PO 4 1:1:1 1.2:1:1 1.5:1:1 1:1:1.2 1:1:1.5 1.2:1:1.2 1.5:1:1.5 T =20 ◦C pH 11 =4 h E ec o empe a u e 10 ◦C 20 ◦C 30 ◦C 40 ◦C Mola a io Mg:K:PO 4 1:1:1 pH 11 =4 h Two s age p ecipi a ion Mola a io Mg:K:PO 4 1:1:1 T =20 ◦C =4 h s age pH 9 s age pH 11 S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 3 he one-s ep and wo-s ep p ecipi a ions, espec i ely; ammonium’s p esence in he washing wa e did no a ec he p ecipi a ion o he po assium. This is p obably because he ammonium concen a ion was e y small (340 ppm) compa ed o he concen a ions o po assium (6900 ppm) and sodium. Based on p e ious s udies, a high po assium con en and low ammonium con en a o K-s u i e o ma ion ins ead o s u i e o ma ion [11,12]. Simila obse a ions we e also made in he case o phospha e ( he emo al e iciencies we e 92% and 99% o he wo-s ep and one-s ep p ecipi a ions) and magnesium ( he emo al e iciencies we e o e 98.8% o he wo-s ep and one-s ep p e- cipi a ions). Based on XRD analysis (no p esen ed), he p ecipi a es ob ained om s u i e p ecipi a ion con ained mos ly K-s u i e (76–99.9%) and a li le s u i e (0–23.2%). Since ammonium did no in luence po assium emo al, di ec hazeni e p ecipi a ion was pe - o med in he la e s ages o he esea ch. 3.2. E ec o pH The e ec o pH on he p ecipi a ion o hazeni e was s udied in he pH ange o 9.5–12 using Mg:K:PO 4 1:1:1 and a mixing a e o 250 pm. The emo al e iciencies o po assium, magnesium and phospha e a e p esen ed in Fig. 2. As can be seen in Fig. 2, he pH had no e ec on he p ecipi a ion e iciency. Fo po assium, he emo al e iciency was 60%–65%. In addi ion, in he case o magnesium and phospha e, simila obse a ions we e made. The o al emo al e iciency was 98.5%–100% o magnesium and 96%–99% o phospha e. Kine ically, he p ecipi- a ion was e y as a all pH alues, as he eac ion occu ed du ing he i s 15 min. The e a e no s udies in he li e a u e on he e ec o pH on hazeni e p ecipi a ion. Since hazeni e is na u ally ound in alkaline condi ions (a pH 10), i will also p ecipi a e a high pH alues, like o he s u i e g oup compounds [3,4]. Gene ally, po assium emo al e iciency in- c eases signi ican ly a high pH alues due o he dissocia ion o H 2 PO 4 − o HPO 4 2− and PO 4 3- [1,13,14]. The inc eased PO 4 3− concen a ion in- c eases he p ecipi a ion o po assium as phospha e compounds, such as K-s u i e and hazeni e. As he pH ises abo e 11, he o ma ion o magnesium hyd oxide [15] could weaken he e iciency o po assium p ecipi a ion, bu his was no obse ed in his s udy. I he pH d ops below 11, he o ma ion o K-s u i e slows down and s u i e may begin o o m, wi h an op imum p ecipi a ion pH o 8–9 [2]. Based on he XRD analysis (Fig. 3), all main peaks we e associa ed wi h hazeni e ega dless o he pH alue, con o ming he esul s o emo al e iciencies (Fig. 2). This indica es ha hazeni e p ecipi a es in a wide pH ange unde alkaline condi ions. 3.3. E ec o empe a u e The e ec o empe a u e on p ecipi a ion e iciency was s udied a empe a u es o 10–40 ◦C (Fig. 4). Based on he esul s, empe a u e does no seem o ha e any impac on p ecipi a ion e iciency. Unde he gi en condi ions (a pH o 11, 250 pm and a Mg:K:PO 4 o 1:1:1), he emo al e iciencies we e 60%, 99.7%–99.9% and 97.3%–99% o po- assium, magnesium and phospha e, espec i ely, a e ou hou s o p ecipi a ion. The p ecipi a es ob ained a di e en empe a u es we e pu e hazeni e (Fig. 5). Fo s u i e p ecipi a ion, empe a u e has a less signi ican e ec han pH o mola a io [16]. Howe e , he solubili y o s u i e inc eases wi h inc easing empe a u e un il a empe a u e o a ound 35 ◦C [16–18]. A highe empe a u es, he solubili y once again dec eases. Ta ag´ o e al. (2017) epo ed ha he op imum empe a u e o K-s u i e o ma ion would be 38 ◦C [19]. Ou esul s indica e ha hazeni e p ecipi a ion is no dependen on empe a u es unde he 10–40 ◦C empe a u e ange. Gene ally, lowe p ecipi a ion empe a- u es should be p e e ed o a oid he high ene gy consump ion caused by hea ing. 3.4. E ec o Mg:K:PO 4 a io The e ec o mola a ios on p ecipi a ion e iciency was s udied wi h mola a ios Mg:K:PO 4 (1.0–1.5):1:(1.0–1.5) (Fig. 6). The Fig. 1. Remo al e iciency o po assium (a), magnesium (b), and phospha e (c) in one-s ep (hazeni e) and wo-s ep (s u i e +hazeni e) p ecipi a ions (Mg:K:PO 4 1:1:1, pH 11 and 50 pm). S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 4 Fig. 2. Remo al e iciency o po assium (a), magnesium (b), and phospha e (c) as a unc ion o pH (Mg:K:PO 4 1:1:1 and 250 pm). Ini ial concen a ions o K: 6900 ppm, Mg: 4750 ppm and PO 4 3− : 17,000 ppm. Fig. 3. XRD di ac og ams o he samples p ecipi a ed a di e en pH alues. S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 5 Fig. 4. The e ec o empe a u e on he emo al e iciency o po assium (a), magnesium (b) and phospha e (c) (Mg:K:PO 4 1:1:1, pH 11 and 250 pm). Ini ial concen a ions o K: 7100 ppm, Mg: 4750 ppm and PO 4 3− : 17,000 ppm. Fig. 5. XRD di ac og ams o he samples p ecipi a ed a di e en empe a u es. S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 6 po assium p ecipi a ion e iciency inc eased om 63.4% o 78% when he mola a io o Mg:K:PO 4 chanced om 1:1:1 o 1.5:1:1.5. Thus, he mo e magnesium and phospha e sal s we e added o he washing wa e , he mo e po assium was emo ed. Howe e , i he e was an excess o only one componen (magnesium o phospha e), i did no imp o e he emo al o po assium om he washing wa e . The excess o phospha e o magnesium and phospha e had li le e - ec on he magnesium emo al a e, as i was o e 99% (Fig. 6a). When excess magnesium was p esen , 95.7% (Mg:K:PO 4 =1.2:1:1) and 96.9% (Mg:K:PO 4 =1.5:1:1) o he magnesium we e emo ed. Phospha e emo al e iciencies (see Fig. 6c) we e simila when he p ecipi a ions we e pe o med wi h a 1:1:1 mola a io o Mg:K:PO 4 and a simul a- neous excess o magnesium o phospha e and magnesium. In all hese p ecipi a es, mo e han 98.5% o he phospha e was emo ed. When magnesium was p esen in excess, he pe cen age o phospha e emo al was sligh ly (abou 1%) highe han wi h Mg:K:PO 4 mola a ios o 1:1:1, 1.2:1:1.2 o 1.5:1:1.5. Thus, excess magnesium sligh ly inc eased he emo al o phospha e om he washing wa e . A mola a ios o Mg:K: PO 4 o 1:1:1.2 and 1:1:1.5, abou 85% and 69% o he phospha e we e emo ed, espec i ely. The amoun o phospha e added in excess emained in he washing wa e and did no p ecipi a e. Theo e ically, po assium should be emo ed comple ely om washing wa e when he Mg:K:PO4 mola a io is 1:1:1. Xu e al. (2011) s udied he emo al o po assium and phospha e as K-s u i e om syn he ic u ine and ound ha he concomi an excess o magnesium and phospha e inc eased he emo al o po assium [13]. Howe e , an excess o magnesium and phospha e ha is oo high can cause he p e- cipi a ion o Mg 3 (PO 4 ) 2 . Excess phospha e (wi hou excess Mg) leads o esidual phospha e in he il a e. Gao e al. (2018) also s udied syn he ic u ine and concluded ha excess magnesium inc eased phospha e emo al [14]. Howe e , hese esul s con adic he esul s o Xu e al. (2011), in which excess mag- nesium educed po assium emo al, inc eased he amoun o p ecipi a e and dec eased i s pu i y [13]. They also ound ha excess phospha e dec eased phospha e emo al e iciency bu sligh ly inc eased po as- sium emo al e iciency. Howe e , excess phospha e did no play a majo ole in he amoun o pu i y o he K-s u i e o med. The simul aneous excess o phospha e and magnesium inc eased he phos- pha e’s and po assium’s emo al e iciencies. The highes po assium emo al e iciency (78.1%) was achie ed wi h a Mg:K:PO 4 mola a io o 1.8:1:1.8 when he pH o he solu ion was 11. Howe e , in hei s udy he simul aneous excess o phospha e and magnesium inc eased he p ecipi a e o med bu dec eased i s pu i y due o he o ma ion o i- magnesium phospha e. The XRD pa e ns o p ecipi a es a di e en mola a ios a e shown in Fig. 7. All he main spikes we e associa ed wi h hazeni e. The p e- cipi a es o med wi h di e en mola a ios Mg:K:PO 4 we e also cha - ac e ized using XRF (Table 3). The p ecipi a es we e nea ly pu e hazeni e, excluding he expe imen wi h excess Mg (mola a io 1.5:1:1), whe e un eac ed MgSO 4 was also p esen . This explains he highe - han- expec ed Mg emo al obse ed using his mola a io. 4. Conclusions C ushed and sie ed alkaline ba e y black mass was washed wi h wa e , and his il e ed washing wa e con aining po assium and ammonium was ea ed by emo ing po assium using phospha e and magnesium sal s. The p ecipi a ion e iciencies o po assium we e Fig. 6. Remo al e iciency o po assium (a), magnesium (b) and phospha e (c) wi h di e en mola a ios Mg:K:PO 4 (250 pm and pH 11). S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 7 ypically 60–65%, and hazeni e was o med. The simul aneous excess o magnesium and phospha e inc eased he po assium emo al e iciency o 78%. When p ecipi a ion was pe o med a di e en pH alues, he e was no di e ence in po assium emo al e iciencies. Fu he , an excess o magnesium o phospha e o di e en empe a u es did no a ec he po assium emo al a e. Thus, i can be concluded ha hazeni e p e- cipi a es o e a wide pH ange, bu i he pH is alkaline, i does no equi e excess magnesium o phospha e o hea ing du ing p ecipi a ion. In addi ion, he p ecipi a ion is kine ically as , and sho eac ion imes a e enough. The e was no need o emo e ammonium ions om he washing wa e p io o hazeni e p ecipi a ion, as ammonium did no inhibi hazeni e o ma ion. E en hough, he excess amoun o magne- sium and phospha e imp o ed he emo al e iciency o po assium, he op imal condi ions o he p ecipi a ion o hazeni e om washing wa e would be oom empe a u e a pH 9.5, wi hou he addi ion o excess magnesium o phospha e. This would mean ha he consump ion o chemicals and ene gy would be minimized. Hazeni e can be used as a e ilize , bu i is a ela i ely new compound ha has no been s udied in dep h. The e o e, mo e esea ch is needed on hazeni e p ecipi a ion and i s use as a e ilize . Funding in o ma ion This esea ch did no ecei e any speci ic g an om unding agencies in he public, comme cial, o no - o -p o i sec o s. Decla a ion o Compe ing In e es The au ho s decla e ha hey ha e no known compe ing inancial in e es s o pe sonal ela ionships ha could ha e appea ed o in luence he wo k epo ed in his pape . Da a a ailabili y Da a will be made a ailable on eques . Acknowledgmen s Sa i Tuikkanen is acknowledged o IC analyses. Re e ences [1] H. Huang, D. Zhang, J. Li, G. Guo, S. Tang, Phospha e eco e y om swine was ewa e using plan ash in chemical c ys alliza ion, J. Clean. P od. 168 (2017) 338–345, h ps://doi.o g/10.1016/j.jclep o.2017.09.042. [2] K. Shih, H. Yan, Chap e 26 - he c ys alliza ion o s u i e and i s analog (K- s u i e) om was e s eams o nu ien ecycling, in: M.N.V. P asad, K. Shih (Eds.), En i onmen al Ma e ials and Was e, 1s ed., Academic P ess, 2016, pp. 665–686, h ps://doi.o g/10.1016/B978-0-12-803837-6.00026-3. [3] H. Yang, H.J. Sun, R.T. Downs, Hazeni e, KNaMg2(PO4)2⋅14H2O, a new biologically ela ed phospha e mine al, om Mono Lake, Cali o nia, U.S.A, Ame . Mine alogis . 96 (2011) 675–681, h ps://doi.o g/10.2138/am.2011.3668. [4] C. Wa son, J. Clemens, F. Wiche n, Hazeni e: a new seconda y phospho us, po assium and magnesium e ilise , Plan Soil En i on. 66 (2020) 1–6, h ps://doi. o g/10.17221/492/2019-PSE. [5] H.R. Rani o, A.P. Be oni Teles, C. Adam, P.S. Pa ina o, Phospho us solubili y and dynamics in a opical soil unde sou ces de i ed om was ewa e and sewage sludge, J. En i on. Manag. 302 (2022), 113984, h ps://doi.o g/10.1016/j. jen man.2021.113984. [6] H.R. Rani o, T.D.M. Soa es, C. Adam, P.S. Pa ina o, Was e-de i ed e ilize s can inc ease phospho us up ake by suga cane and a ailabili y in a opical soil#, J. Plan Nu . Soil Sci. 185 (2022) 391–402, h ps://doi.o g/10.1002/ jpln.202100410. [7] F. Fe ella, I.D. Michelis, F. Pagnanelli, F. Beolchini, G. Fu lani, M. Na a a, F. Vegli` o, L. To o, Reco e y o zinc and manganese om spen ba e ies by di e en leaching sys ems, Ac a Me all. Slo aca 12 (2006) 95–104. [8] A. Sobianowska - Tu ek, W. Szczepaniak, M. Zabłocka-Malicka, Elec ochemical e alua ion o manganese educe s – Reco e y o Mn om Zn–Mn and Zn–C ba e y was e, J. Powe Sou ces 270 (2014) 668–674, h ps://doi.o g/10.1016/j. jpowsou .2014.07.136. Fig. 7. XRD di ac og ams o he samples p ecipi a ed a di e en mola a ios o Mg:K: PO 4 . Table 3 Composi ion o he p ecipi a es a di e en mola a ios measu ed wi h XRF. Mola a io Mg:K: PO 4 Hazeni e (%) Na 2 O (%) MgO (%) P 2 O 5 (%) SO 3 (%) Cl (%) CaO (%) 1:1:1 98.483 0.00 0.00 0.40 0.73 0.14 0.19 1.2:1:1 98.391 0.00 0.02 0.29 0.88 0.11 0.20 1.5:1:1 89.528 0.56 4.32 0.38 4.49 0.44 0.19 1:1:1.2 98.606 0.00 0.00 0.64 0.40 0.08 0.19 1:1:1.5 98.635 0.00 0.00 0.51 0.50 0.09 0.18 1.2:1:1.2 98.174 0.00 0.00 0.80 0.65 0.09 0.21 1.5:1:1.5 96.069 0.54 0.08 2.17 0.77 0.10 0.21 S. Lapinkangas e al. Chemical Enginee ing Jou nal Ad ances 12 (2022) 100426 8 [9] G.F. Lipscomb, G.A. Hule , A S udy o Double Sal s in S anda d Cells, J. Phys. Chem. 20 (1916) 75–82, h ps://doi.o g/10.1021/j150163a005. [10] G. Wollmann, W. Voig , Solid–liquid phase equilib ia in he sys em K2SO4–MnSO4–H2O a 298 K and 313 K, Fluid Phase Equilib. 291 (2010) 76–80, h ps://doi.o g/10.1016/j. luid.2009.12.008. [11] R.C. de S. Mei a, S.P.A. da Paz, J.A.M. Co ˆ ea, XRD-Rie eld analysis as a ool o moni o ing s u i e analog p ecipi a ion om was ewa e : P, Mg, N and K eco e y o e ilize p oduc ion, J. Ma e . Res. Technol. 9 (2020) 15202–15213, h ps:// doi.o g/10.1016/j.jm .2020.10.082. [12] E. Company, M. Fa ´ es, J. Colp im, A. Mag í, Explo ing he eco e y o po assium- ich s u i e a e a ni i ica ion-deni i ica ion p ocess in pig slu y ea men , Sci. To al En i on. 847 (2022), 157574, h ps://doi.o g/10.1016/j. sci o en .2022.157574. [13] K. Xu, C. Wang, H. Liu, Y. Qian, Simul aneous emo al o phospho us and po assium om syn he ic u ine h ough he p ecipi a ion o magnesium po assium phospha e hexahyd a e, Chemosphe e 84 (2011) 207–212, h ps://doi.o g/ 10.1016/j.chemosphe e.2011.04.057. [14] Y. Gao, B. Liang, H. Chen, P. Yin, An expe imen al s udy on he eco e y o po assium (K) and phospho ous (P) om syn he ic u ine by c ys alliza ion o magnesium po assium phospha e, Chem. Eng. J. 337 (2018) 19–29, h ps://doi. o g/10.1016/j.cej.2017.12.077. [15] E.-.T. Tolonen, J. R¨ am¨ o, U. Lassi, The e ec o magnesium on pa ial sulpha e emo al om mine wa e as gypsum, J. En i on. Manag. 159 (2015) 143–146, h ps://doi.o g/10.1016/j.jen man.2015.05.009. [16] C. Gonz´ alez-Mo ales, B. Fe n´ andez, F.J. Molina, D. Na anjo-Fe n´ andez, A. Ma amo os-Veloza, M.A. Cama go-Vale o, In luence o pH and empe a u e on s u i e pu i y and eco e y om anae obic diges a e, Sus ainabili y 13 (2021) 10730, h ps://doi.o g/10.3390/su131910730. [17] M. Hanhoun, L. Mon as uc, C. Azza o-Pan el, B. Biscans, M. F ` eche, L. Pibouleau, Tempe a u e impac assessmen on s u i e solubili y p oduc : a he modynamic modeling app oach, Chem. Eng. J. 167 (2011) 50–58, h ps://doi.o g/10.1016/j. cej.2010.12.001. [18] S. Ben Moussa, M.M. Tlili, N. Ba is, M.B. Amo , In luence o empe ea u e on S u i e p ecipi a ion by CO2-deagassing me hod, C ys . Res. Technol. 46 (2011) 255–260, h ps://doi.o g/10.1002/c a .201000571. [19] E. Ta ag´ o, M. Ruscalleda, J. Colp im, M.D. Balague , S. Puig, Towa ds a me hodology o eco e ing K-s u i e om manu e, J. Chem. Technol. Bio echnol. 93 (2018) 1558–1562, h ps://doi.o g/10.1002/jc b.5518. S. Lapinkangas e al.