Removal of Microcystis aeruginosa through the Combined Effect of Plasma Discharge and Hydrodynamic Cavitation

wa e A icle Remo al o Mic ocys is ae uginosa h ough he Combined E ec o Plasma Discha ge and Hyd odynamic Ca i a ion Blahosla Ma šálek 1,*, Eliška Ma šálko á1, Klá a Odehnalo á1, F an išek Pochylý2, Pa el Rudol 2, Pa el S ahel 3, Joze Rahel 3, Jan ˇ Cech 3, Simona Fialo á2and Š ˇepán Zezulka 1 1Ins i u e o Bo any, Czech Academy o Sciences, Lidická25/27, CZ-602 00 B no, Czech Republic; [email p o ec ed] (E.M.); [email p o ec ed] (K.O.); [email p o ec ed] (Š.Z.) 2V. Kaplan Depa men , Facul y o Mechanical Enginee ing, B no Uni e si y o Technology, 601 90 B no, Czech Republic; [email p o ec ed].cz (F.P.); [email p o ec ed].cz (P.R.); [email p o ec ed].cz (S.F.) 3Depa men o Physical Elec onics, Facul y o Science, Masa yk Uni e si y, Ko lᡠská2, CZ-611 37 B no, Czech Republic; [email p o ec ed] (P.S.); [email p o ec ed] (J.R.); [email p o ec ed] (J. ˇ C.) *Co espondence: blahosla [email p o ec ed]; Tel.: +42-0603-8729-55 Recei ed: 19 No embe 2019; Accep ed: 13 Decembe 2019; Published: 18 Decembe 2019   Abs ac : Cyanobac e ial wa e blooms ep esen oxicological, ecological and echnological p oblems a ound he globe. When p esen in aw wa e used o d inking wa e p oduc ion, one o he bes s a egies is o emo e he cyanobac e ial biomass gen ly be o e ea men , a oiding cell des uc ion and cyano oxins elease. This pape p esen s a new me hod o he emo al o cyanobac e ial biomass du ing d inking wa e p e- ea men ha combines hyd odynamic ca i a ion wi h cold plasma discha ge. Ca i a ion p oduces p ess s ess ha causes Mic ocys is gas esicles o collapse. The cyanobac e ia hen sink, allowing o emo al by sedimen a ion. The cyanobac e ia showed no signs o e i alisa ion, e en a e se en days unde op imal condi ions wi h nu ien en ichmen , as pho osyn he ic ac i i y is nega i ely a ec ed by hyd ogen pe oxide p oduced by plasma bu n in he ca i a ion cloud. Using his me hod, cyanobac e ia can be emo ed in a single ea men , wi h no inc ease in mic ocys in concen a ion. This no el echnology appea s o be highly p omising o con inual ea men o aw wa e in low in d inking wa e ea men plan s and will also be o in e es o hose wishing o ea su ace wa e s wi hou he use o algaecides. Keywo ds: cyanobac e ial bloom; wa e ea men ; d inking wa e ; su ace wa e ; cold plasma 1. In oduc ion Inc easing eu ophica ion o su ace wa e s esul s in he de elopmen o cyanobac e ial blooms, esul ing in a de e io a ion o wa e quali y and dis up ions in he use o wa e bodies o ec ea ional, echnological and ishe ies pu poses o d inking wa e p oduc ion [ 1 ]. Fu he mo e, he e is clea e idence indica ing ha cyanobac e ial blooms a e inc easing in equency, magni ude and du a ion globally and ha hey a e becoming a majo echnological, oxicological and hygienic p oblem [1]. Va ious me hods a e a ailable o educing o emo ing cyanobac e ial blooms. A b oad spec um o chemical compounds can be used as p e en i e o cu a i e measu es [ 2 ]; howe e , hese can cause addi ional eco oxicological isks o aqua ic ecosys ems [ 3 ]. Fu he s a egies o p e en ing he de elopmen o cyanobac e ial blooms a e nu ien limi a ion [ 4 ], semi-na u al me hods such as he use o plan ex ac s [ 5 , 6 ] o ba ley s aw, and biological me hods such as he use o p obio ic mic oo ganisms and bio ic in e ac ions [ 7 ]. O he many chemical and biological me hods a ailable, physical me hods a e being inc easingly in es iga ed, as hey no only emo e cyanobac e ial biomass Wa e 2020,12, 8; doi:10.3390/w12010008 www.mdpi.com/jou nal/wa e Wa e 2020,12, 8 2 o 14 bu also cyano oxins, which is especially impo an o d inking wa e ea men . While elec ochemical me hods, such as elec opo a ion, elec ocoagula ion o elec o lo a ion [ 8 ], ha e been he subjec o pa icula in e es , plasma discha ge in combina ion wi h ad anced oxida ion echnologies o i ied by ul asonic o hyd odynamic ca i a ion a e p obably he mos s udied me hods a his ime [9–11]. In a s udy Kim e al. [ 12 ] ha e alua ed he educ ion o algal biomass by applying non- he mal o cold-plasma p ocesses, i was ound ha such me hods des oyed he cell walls o mic oalgae. Based on hei esea ch, cold plasma seems o be a use ul op ion o e ec i ely ea ing pollu ion caused by algal blooms in su ace wa e . Plasma and ca i a ion echnologies ha e been applied o pu i ying wa e con amina ed by mic oo ganisms [ 10 , 13 ]; howe e , p ac ical applica ions ha e been se e ely es ic ed by excessi e ea men imes and ene gy demand (e.g., 24 h o cold plasma ea men o se e al hund ed ca i a ion cycles). As such, he e is a clea need o echnologies ha can emo e cyanobac e ial biomass wi hou eleasing cyano oxins, and ha can be applied in con inuous mode o d inking wa e p e- ea men . In his s udy, we ocus on physical me hods ha combine hyd odynamic ca i a ion (HC) and plasma discha ge (HC +plasma). In p inciple, he me hod elies on he o ma ion o apou ca i ies in a liquid, wi h plasma being p oduced in he bubble cloud ha is o med downs eam o he eac o nozzle h ough ca i a ion. Plasma is p oduced h ough he applica ion o high ol age o an elec ode pai moun ed in he eac o [ 14 ]. We modi ied his me hod o gen le ea men o cyanobac e ial biomass wi h he aim o educing cyanobac e ial cell pho osyn he ic ac i i y wi hou cell lysis o elease o cyano oxins. By op imising he ime, ene ge ic and ea men aspec s o he me hod, we aim o de elop a p ac ically applicable me hod o use in d inking wa e ea men plan s. 2. Ma e ials and Me hods 2.1. Hyd odynamic Ca i a ion De ice We cons uc ed an expe imen al ci cui consis ing o a wa e ank, a cen i ugal pump (Calpeda, Vicenza, I aly, a ed powe ou pu 0.1 kW) con olled by a equency con e o , a se o p essu e ansduce s (p 1 and p 2 ; bo h BD Senso , 0–6 ba ) and a lowme e (Flomag DN 15) (Figu e 1). A con e ging–di e ging (CD) nozzle wi h an inle diame e o 10 mm and a minimum diame e o 4 mm a he h oa sec ion was manu ac u ed om plexiglass. We designed a nozzle ha p oduced a ca i a ion egion illed wi h sa u a ed apou ollowing a p essu e d op om c oss-sec ional a ea educ ion (i.e., es ablishmen o HC acco ding o he Be noulli p inciple). As apou p essu e is sensi i e o wa e empe a u e and he amoun o dissol ed gas, a he mome e (Sensi , ype PT100) and p obe measu ing dissol ed oxygen con en (Oxymax) we e inco po a ed in o he expe imen al ci cui . Subsequen moni o ing con i med ha dissol ed oxygen emained a ca. 8.3 mg/L h oughou he expe imen , ha s a ic p essu e a he nozzle h oa eached sa u a ed apou p essu e le els a 3500 Pa and ha liquid empe a u e emained ela i ely s able o he i s i e cycles o HC +plasma ea men , a ying only sligh ly be ween 23 and 25 ◦ C. All es s we e ca ied ou a a discha ge a e o 0.45 L/s. Wa e 2019, 11, x FOR PEER REVIEW 2 o 17 cyanobac e ial biomass bu also cyano oxins, which is especially impo an o d inking wa e ea men . While elec ochemical me hods, such as elec opo a ion, elec ocoagula ion o elec o lo a ion [8], ha e been he subjec o pa icula in e es , plasma discha ge in combina ion wi h ad anced oxida ion echnologies o i ied by ul asonic o hyd odynamic ca i a ion a e p obably he mos s udied me hods a his ime [9–11]. In a s udy Kim e al. [12] ha e alua ed he educ ion o algal biomass by applying non- he mal o cold-plasma p ocesses, i was ound ha such me hods des oyed he cell walls o mic oalgae. Based on hei esea ch, cold plasma seems o be a use ul op ion o e ec i ely ea ing pollu ion caused by algal blooms in su ace wa e . Plasma and ca i a ion echnologies ha e been applied o pu i ying wa e con amina ed by mic oo ganisms [10,13]; howe e , p ac ical applica ions ha e been se e ely es ic ed by excessi e ea men imes and ene gy demand (e.g., 24 h o cold plasma ea men o se e al hund ed ca i a ion cycles). As such, he e is a clea need o echnologies ha can emo e cyanobac e ial biomass wi hou eleasing cyano oxins, and ha can be applied in con inuous mode o d inking wa e p e- ea men . In his s udy, we ocus on physical me hods ha combine hyd odynamic ca i a ion (HC) and plasma discha ge (HC + plasma). In p inciple, he me hod elies on he o ma ion o apou ca i ies in a liquid, wi h plasma being p oduced in he bubble cloud ha is o med downs eam o he eac o nozzle h ough ca i a ion. Plasma is p oduced h ough he applica ion o high ol age o an elec ode pai moun ed in he eac o [14]. We modi ied his me hod o gen le ea men o cyanobac e ial biomass wi h he aim o educing cyanobac e ial cell pho osyn he ic ac i i y wi hou cell lysis o elease o cyano oxins. By op imising he ime, ene ge ic and ea men aspec s o he me hod, we aim o de elop a p ac ically applicable me hod o use in d inking wa e ea men plan s. 2. Ma e ials and Me hods 2.1. Hyd odynamic Ca i a ion De ice We cons uc ed an expe imen al ci cui consis ing o a wa e ank, a cen i ugal pump (Calpeda, Vicenza, I aly, a ed powe ou pu 0.1 kW) con olled by a equency con e o , a se o p essu e ansduce s (p 1 and p 2 ; bo h BD Senso , 0–6 ba ) and a lowme e (Flomag DN 15) (Figu e 1). A con e ging–di e ging (CD) nozzle wi h an inle diame e o 10 mm and a minimum diame e o 4 mm a he h oa sec ion was manu ac u ed om plexiglass. We designed a nozzle ha p oduced a ca i a ion egion illed wi h sa u a ed apou ollowing a p essu e d op om c oss-sec ional a ea educ ion (i.e., es ablishmen o HC acco ding o he Be noulli p inciple). As apou p essu e is sensi i e o wa e empe a u e and he amoun o dissol ed gas, a he mome e (Sensi , ype PT100) and p obe measu ing dissol ed oxygen con en (Oxymax) we e inco po a ed in o he expe imen al ci cui . Subsequen moni o ing con i med ha dissol ed oxygen emained a ca. 8.3 mg/L h oughou he expe imen , ha s a ic p essu e a he nozzle h oa eached sa u a ed apou p essu e le els a 3500 Pa and ha liquid empe a u e emained ela i ely s able o he i s i e cycles o HC + plasma ea men , a ying only sligh ly be ween 23 and 25 °C. All es s we e ca ied ou a a discha ge a e o 0.45 L/s. Figu e 1. Schema ic illus a ion o he hyd odynamic ca i a ion (HC) de ice [15]. CD: con e ging– di e ging. Figu e 1. Schema ic illus a ion o hehyd odynamicca i a ion (HC)de ice[ 15 ]. CD: con e ging–di e ging. Wa e 2020,12, 8 3 o 14 2.2. Plasma Je Gene a ion in Liquid Gas Ca i ies A plasma je was p oduced in he ca i a ion bubble cloud gene a ed in wa e using an expe imen al se up consis ing o wo pa s: (a) a PMMA (poly me hyl me hac yla e, i.e., plexiglass) pipe ca i a ion gene a o (nozzle) connec ed o a wa e pumping and eci cula ion uni (black ci cle in Figu e 2), and (b) a plasma gene a o wi h a discha ge elec ode sys em and a high- ol age (HV) powe gene a o (g een ci cle in Figu e 2). Spec al analysis o he discha ge was ob ained jus abo e he HV elec ode using A aSpec ULS3648TEC-USB2 spec ome e s (A an es BV, The Ne he lands). Spec ome e s we e con igu ed o h ee spec al egions co e ing he ange om nea -ul a iole (UV) o he nea -in a ed (NIR) egion. In he con igu a ion o su ey spec a (UV-VIS-NIR), he spec ome e was equipped wi h he g a ing UA (200–1100 nm), a sli o 25 um, Deep UV de ec o coa ing and o de -so ing coa ing wi h 350 and 600 nm long-pass il e . Fo highe esolu ion measu emen s in UV egion, esp. NIR egion, he spec ome e s we e equipped wi h he g a ing UE (290–395 nm), sli 10 um and DUV coa ing, esp. g a ing NC (740–924 nm), sli 25 um and o de -so ing il e a 600 nm. Wa e 2019, 11, x FOR PEER REVIEW 3 o 17 2.2. Plasma Je Gene a ion in Liquid Gas Ca i ies A plasma je was p oduced in he ca i a ion bubble cloud gene a ed in wa e using an expe imen al se up consis ing o wo pa s: (a) a PMMA (poly me hyl me hac yla e, i.e., plexiglass) pipe ca i a ion gene a o (nozzle) connec ed o a wa e pumping and eci cula ion uni (black ci cle in Figu e 2), and (b) a plasma gene a o wi h a discha ge elec ode sys em and a high- ol age (HV) powe gene a o (g een ci cle in Figu e 2). Spec al analysis o he discha ge was ob ained jus abo e he HV elec ode using A aSpec ULS3648TEC-USB2 spec ome e s (A an es BV, The Ne he lands). Spec ome e s we e con igu ed o h ee spec al egions co e ing he ange om nea -ul a iole (UV) o he nea -in a ed (NIR) egion. In he con igu a ion o su ey spec a (UV-VIS-NIR), he spec ome e was equipped wi h he g a ing UA (200–1100 nm), a sli o 25 um, Deep UV de ec o coa ing and o de -so ing coa ing wi h 350 and 600 nm long-pass il e . Fo highe esolu ion measu emen s in UV egion, esp. NIR egion, he spec ome e s we e equipped wi h he g a ing UE (290–395 nm), sli 10 um and DUV coa ing, esp. g a ing NC (740–924 nm), sli 25 um and o de - so ing il e a 600 nm. Figu e 2. Expe imen al se -up. We gene a ed a plasma je in he ca i a ion egion using a pai o PVC-insula ed coppe wi e elec odes placed inside he ca i a ion nozzle (Figu e 3). The i s HV elec ode was placed inside he nozzle a he poin o minimum diame e , while he second, g ounded, elec ode was posi ioned downs eam o he ca i a ion bubble cloud. The insula ion was s ipped om he ips o he wi es o c ea e a di ec conduc ing con ac wi h he wa e /bubble en i onmen (see Figu e 3). The discha ge was gene a ed using an HV gene a o wi h sinusoidal HV ou pu , wi h ope a ion equency se o 50 kHz and HV powe a 400 W. The plasma je was gene a ed be ween he ips o he HV elec ode and he g ounded elec ode, whe e i o med a plasma a ea ca. 20 mm in leng h (Figu e 4). To al wa e olume inside he sys em was 6 L, and con ac - ea men ime was 1.48 s. Figu e 3. De ails o he elec ode sys em and i s placemen wi h espec o he ca i a ion cloud. Figu e 2. Expe imen al se -up. We gene a ed a plasma je in he ca i a ion egion using a pai o PVC-insula ed coppe wi e elec odes placed inside he ca i a ion nozzle (Figu e 3). The i s HV elec ode was placed inside he nozzle a he poin o minimum diame e , while he second, g ounded, elec ode was posi ioned downs eam o he ca i a ion bubble cloud. The insula ion was s ipped om he ips o he wi es o c ea e a di ec conduc ing con ac wi h he wa e /bubble en i onmen (see Figu e 3). The discha ge was gene a ed using an HV gene a o wi h sinusoidal HV ou pu , wi h ope a ion equency se o 50 kHz and HV powe a 400 W. The plasma je was gene a ed be ween he ips o he HV elec ode and he g ounded elec ode, whe e i o med a plasma a ea ca. 20 mm in leng h (Figu e 4). To al wa e olume inside he sys em was 6 L, and con ac - ea men ime was 1.48 s. Wa e 2019, 11, x FOR PEER REVIEW 3 o 17 2.2. Plasma Je Gene a ion in Liquid Gas Ca i ies A plasma je was p oduced in he ca i a ion bubble cloud gene a ed in wa e using an expe imen al se up consis ing o wo pa s: (a) a PMMA (poly me hyl me hac yla e, i.e., plexiglass) pipe ca i a ion gene a o (nozzle) connec ed o a wa e pumping and eci cula ion uni (black ci cle in Figu e 2), and (b) a plasma gene a o wi h a discha ge elec ode sys em and a high- ol age (HV) powe gene a o (g een ci cle in Figu e 2). Spec al analysis o he discha ge was ob ained jus abo e he HV elec ode using A aSpec ULS3648TEC-USB2 spec ome e s (A an es BV, The Ne he lands). Spec ome e s we e con igu ed o h ee spec al egions co e ing he ange om nea -ul a iole (UV) o he nea -in a ed (NIR) egion. In he con igu a ion o su ey spec a (UV-VIS-NIR), he spec ome e was equipped wi h he g a ing UA (200–1100 nm), a sli o 25 um, Deep UV de ec o coa ing and o de -so ing coa ing wi h 350 and 600 nm long-pass il e . Fo highe esolu ion measu emen s in UV egion, esp. NIR egion, he spec ome e s we e equipped wi h he g a ing UE (290–395 nm), sli 10 um and DUV coa ing, esp. g a ing NC (740–924 nm), sli 25 um and o de - so ing il e a 600 nm. Figu e 2. Expe imen al se -up. We gene a ed a plasma je in he ca i a ion egion using a pai o PVC-insula ed coppe wi e elec odes placed inside he ca i a ion nozzle (Figu e 3). The i s HV elec ode was placed inside he nozzle a he poin o minimum diame e , while he second, g ounded, elec ode was posi ioned downs eam o he ca i a ion bubble cloud. The insula ion was s ipped om he ips o he wi es o c ea e a di ec conduc ing con ac wi h he wa e /bubble en i onmen (see Figu e 3). The discha ge was gene a ed using an HV gene a o wi h sinusoidal HV ou pu , wi h ope a ion equency se o 50 kHz and HV powe a 400 W. The plasma je was gene a ed be ween he ips o he HV elec ode and he g ounded elec ode, whe e i o med a plasma a ea ca. 20 mm in leng h (Figu e 4). To al wa e olume inside he sys em was 6 L, and con ac - ea men ime was 1.48 s. Figu e 3. De ails o he elec ode sys em and i s placemen wi h espec o he ca i a ion cloud. Figu e 3. De ails o he elec ode sys em and i s placemen wi h espec o he ca i a ion cloud. 2.3. Cyanobac e ial S ain and Cul u e Condi ions Fo ou expe imen s, we used Mic ocys is ae uginosa PCC 7806 s ain (Pas eu Cul u e Collec ion o Cyanobac e ia; Pa is) cul u ed in BG11 medium o cyanobac e ia ( o medium composi ion, see Supplemen a y Ma e ials Table S1). The cul u e has now been placed in ou labo a o y cul u e collec ion and is g own unde s anda d condi ions o cyanobac e ia. Wa e 2020,12, 8 4 o 14 Wa e 2019, 11, x FOR PEER REVIEW 4 o 17 Figu e 4. Image o he plasma je gene a ed in he liquid/ca i a ion bubble cloud en i onmen . 2.3. Cyanobac e ial S ain and Cul u e Condi ions Fo ou expe imen s, we used Mic ocys is ae uginosa PCC 7806 s ain (Pas eu Cul u e Collec ion o Cyanobac e ia; Pa is) cul u ed in BG11 medium o cyanobac e ia ( o medium composi ion, see Supplemen a y Ma e ials Table S1). The cul u e has now been placed in ou labo a o y cul u e collec ion and is g own unde s anda d condi ions o cyanobac e ia. 2.4. De e mina ion o Hyd ogen Pe oxide We de e mined hyd ogen pe oxide p oduc ion using a colo ime ic me hod based on he eac ion o pe oxide wi h i anyl ions ha p oduces a yellow-colou ed complex o pe i anic acid: Ti4+ + H2O2 + H2O → TiO2.H2O2 + 4H+ [16] 2.5. G ow h Inhibi ion Tes Immedia ely a e each expe imen , a cul u e o he Mic ocys is biomass was assessed using a modi ied e sion o s anda d me hod ISO 8692 [17,18]. Samples om he expe imen al ea men (HC o HC + Plasma) we e mixed and homogenised, ollowing which 1 o 2 mL o sample was mixed wi h 50% g ow h medium a a a io o 1:1 in o de o p o ide op imal g ow h condi ions. The samples we e hen pipe ed in o a 96-well pla e and chlo ophyll luo escence was measu ed each 24 h using a Tecan SPARK mul i unc ional mic opla e eade (Aus ia). 2.6. Chlo ophyll Quan i ica ion Chlo ophyll concen a ion was assessed using a bbe Moldaenke Fluo oP obe (Ge many), which bo h de e mines and quan i ies he algae classed unde g een algae, cyanobac e ia, dia oms and c yp ophy es. Samples we e ob ained om 2 cm below he essel su ace. 2.7. Measu emen o Pho osyn he ic Ac i i y Induced chlo ophyll luo escence (ChlFl) cha ac e is ics we e assessed using he AquaPen AP 100-C luo ome e (PSI, D áso , Czech Republic) equipped wi h cu e e holde and ambe - ed (620 nm) LEDs as measu ing, ac inic and sa u a ion ligh sou ces. A se o ChlFl pa ame e s was achie ed a e 10 min o da k p e-adap a ion a oom empe a u e (22 ± 1 °C). Immedia ely be o e he measu emen , samples in he cu e e we e ho oughly shaken o a oid sedimen a ion and/o su ace loa ing o cyanobac e ial cells, hen a sa u a ion pulse (3000 µmol m−2 s−1, 500 ms) was applied and he maximal quan um yield (FV/FM) was measu ed. A e ano he 30 s o elaxa ion in he da k, as Commen ed [M1]: 补补补补补补 Figu e 4. Image o he plasma je gene a ed in he liquid/ca i a ion bubble cloud en i onmen . 2.4. De e mina ion o Hyd ogen Pe oxide We de e mined hyd ogen pe oxide p oduc ion using a colo ime ic me hod based on he eac ion o pe oxide wi h i anyl ions ha p oduces a yellow-colou ed complex o pe i anic acid [16]: Ti4++H2O2+H2O→TiO2.H2O2+4H+ 2.5. G ow h Inhibi ion Tes Immedia ely a e each expe imen , a cul u e o he Mic ocys is biomass was assessed using a modi ied e sion o s anda d me hod ISO 8692 [ 17 , 18 ]. Samples om he expe imen al ea men (HC o HC +Plasma) we e mixed and homogenised, ollowing which 1 o 2 mL o sample was mixed wi h 50% g ow h medium a a a io o 1:1 in o de o p o ide op imal g ow h condi ions. The samples we e hen pipe ed in o a 96-well pla e and chlo ophyll luo escence was measu ed each 24 h using a Tecan SPARK mul i unc ional mic opla e eade (Aus ia). 2.6. Chlo ophyll Quan i ica ion Chlo ophyll concen a ion was assessed using a bbe Moldaenke Fluo oP obe (Ge many), which bo h de e mines and quan i ies he algae classed unde g een algae, cyanobac e ia, dia oms and c yp ophy es. Samples we e ob ained om 2 cm below he essel su ace. 2.7. Measu emen o Pho osyn he ic Ac i i y Induced chlo ophyll luo escence (ChlFl) cha ac e is ics we e assessed using he AquaPen AP 100-C luo ome e (PSI, D á so , Czech Republic) equipped wi h cu e e holde and ambe - ed (620 nm) LEDs as measu ing, ac inic and sa u a ion ligh sou ces. A se o ChlFl pa ame e s was achie ed a e 10 min o da k p e-adap a ion a oom empe a u e (22 ± 1 ◦ C). Immedia ely be o e he measu emen , samples in he cu e e we e ho oughly shaken o a oid sedimen a ion and/o su ace loa ing o cyanobac e ial cells, hen a sa u a ion pulse (3000 µ mol m −2 s −1 , 500 ms) was applied and he maximal quan um yield (F V /F M ) was measu ed. A e ano he 30 s o elaxa ion in he da k, as luo escence ansien (OJIP cu e) unde ac inic ligh (300 µ mol m −2 s −1 , 2 s) was eco ded. Th ee independen samples om each ea men we e analysed and ChlFl pa ame e s including basal and maximal luo escence le el (F 0 , F M o F P ) o ‘Fix A ea’ alue we e eco ded. In o de o compa e he i ness o cyanobac e ial cells ela i ely and independen ly om he F V /F M alue, a “Fi ness ac o ” was in oduced as a a io o he Fix A ea alue (de ined ma hema ically as he a ea below he OJIP cu e limi ed by F 0 ime o es ima ion (40 µ s) and ime 1 s, a e he ex ac ion o backg ound signal le el) o F0acco ding o he ollowing equa ion: Fi ness ac o =(Fix A ea/F0)−1000 (uni less) Wa e 2020,12, 8 5 o 14 This alue in eg a es bo h he o e all luo escence signal based on cell suspension p ope ies (densi y, chlo ophyll con en ) and changes in he shape o he OJIP cu e. 2.8. Mic ocys in Analysis Quan iPla e Ki o Mic ocys ins (En i oLogix USA) was used o ex acellula mic ocys ins de e mina ion. A e expe imen al ea men s, samples we e il e ed by 0.2 µ m Sa o ius il e s. Resul s a e in e p e ed as he sum o mic ocys ins. 2.9. S a is ical Analysis To ind ou i he da ase s o he expe imen al esul s we e s a is ically signi ican , analysis o a iance (ANOVA) was used. A e e i ica ion o no mal dis ibu ion and homogenei y o a iance, he da a we e p ocessed by analysis o a iance (ANOVA), wi h di e ences compa ed he Tukey Hones ly signi ican di e ence (HSD) ange es and signi icance se a p<0.05. 3. Resul s and Discussion 3.1. Plasma Gene a ion While ac i i y o eac i e oxygen species (ROS) has been p oposed as he d i e o cyanobac e ial cell damage in p e ious s udies epo ing on he use o plasma in wa e ea men , he e is no empi ical da a o suppo his. In his s udy, we speci ically se ou o measu e he p esence o adicals gene a ed. In doing so, we no only con i med he p esence o oxygen adicals [OH] bu also an app eciable le el o ni ogen adicals (see Figu e 5). Spec al analysis o he discha ge ( aken jus abo e he HV elec ode; spec al lines and bands iden i ied using [ 19 – 21 ]; see Figu e 5, o de ails see Supplemen a y Ma e ials) unde bo h nea UV (Figu e 6) and NIR (Figu e 7) clea ly showed ha he combina ion o HC +plasma p oduced bo h oxygen adicals (single and/o iple spec a) and hyd oxyl o ni oxyl adicals. While he wa e was only exposed o plasma o ca. 7 ms pe cycle, he eac i e oxygen/ni ogen species (RONS) gene a ed (e.g., pe oxides) ha e a longe li e ime. The spec a also included a omic coppe lines a ising om he me al elec odes hemsel es, wi h he s onges lines iden i ied a 324.8 and 327.4 nm in he UV egion; 510.6, 515.3 and 521.8 nm in he isible egion and 793.3 and 809.3 nm in he NIR egion (see Figu es 6and 7). Wa e 2019, 11, x FOR PEER REVIEW 6 o 17 Figu e 5. Su ey spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . The s onges emission in he UV spec al egion came om OH adicals (Figu e 6), wi h he s onges emission o he OH 2Σ→2Π g ound s a e sys em occu ing a 306 nm (bands 0-0 and 1-1) and a weake emission a 281 nm (band 1-0). Unde isible and NIR, he spec um also showed s ong a omic hyd ogen lines (Balme se ies; Hα a 656 nm, Hβ a 486 nm and Hγ a 434 nm) and a omic oxygen lines (3s-3p iple s a 777.4 nm, 5S*-5P and 844.6 nm, 3S*-3P). The p esence o a wa e laye be ween he discha ge and he PMMA ube con ibu ed o abso p ion o he discha ge emission [22]. Fo example, ni ogen adical molecules p esen in he discha ge a mosphe e we e dilu ed, as shown by a compa ison be ween he emission o he second posi i e N2 sys em (SPS; C 3Π→B 3Π) wi h he i s nega i e N2+ ion sys em (FNS; 2Σ→2Σ, g ound s a e). The SPS comp ised nume ous ib a ion bands (e.g., band 0-0 a 337.1 nm), while he FNS spec um only showed a weak band a 391.4 nm (band 0-0). Based on he second posi i e ni ogen sys em spec um, he o a ional empe a u e o he N2 molecules was es ima ed a 3600 ± 300 K. The emission spec a, showing a omic oxygen, a omic hyd ogen and exci ed ni ogen molecules/ions in he discha ge, clea ly indica e ha wa e apou molecules we e dissocia ed o OH adicals. The p esence o exci ed ni ogen molecules and a omic oxygen also sugges s he p esence o ni ogen oxide species [23], commonly obse ed in discha ge spec a o discha ges used o p oduce RONS in wa e [24,25]. While o he spec oscopic me hods (e.g., Fou ie ans o m in a ed abso p ion, lase abso p ion, elec on pa amagne ic esonance), along wi h luo escence/colo ime ic measu emen s [22,26,27], may be be e placed o con i m and quan i y RONS, ou own spec oscopic e alua ion clea ly showed p oduc ion o RONS by HC + plasma. The spec um also consis s o a omic coppe lines (see Figu es 6 and 7) a ising om he ma e ial o me al elec odes. The s onges iden i ied lines we e a 324.8 and 327.4 nm in he UV egion, 510.6, 515.3 and 521.8 nm in he isible egion and 793.3 and 809.3 nm in he NIR egion. Fo he iden i ica ion o he spec al lines and bands he ollowing li e a u e was used: [19–21] Figu e 5. Su ey spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . The s onges emission in he UV spec al egion came om OH adicals (Figu e 6), wi h he s onges emission o he OH 2Σ→2Π g ound s a e sys em occu ing a 306 nm (bands 0-0 and 1-1) and a weake emission a 281 nm (band 1-0). Unde isible and NIR, he spec um also showed s ong a omic hyd ogen lines (Balme se ies; H α a 656 nm, H β a 486 nm and H γ a 434 nm) and a omic Wa e 2020,12, 8 6 o 14 oxygen lines (3s-3p iple s a 777.4 nm, 5S * -5P and 844.6 nm, 3S * -3P). The p esence o a wa e laye be ween he discha ge and he PMMA ube con ibu ed o abso p ion o he discha ge emission [ 22 ]. Fo example, ni ogen adical molecules p esen in he discha ge a mosphe e we e dilu ed, as shown by a compa ison be ween he emission o he second posi i e N 2 sys em (SPS; C 3Π→ B 3Π ) wi h he i s nega i e N 2+ ion sys em (FNS; 2Σ→2Σ , g ound s a e). The SPS comp ised nume ous ib a ion bands (e.g., band 0-0 a 337.1 nm), while he FNS spec um only showed a weak band a 391.4 nm (band 0-0). Based on he second posi i e ni ogen sys em spec um, he o a ional empe a u e o he N2molecules was es ima ed a 3600 ±300 K. Wa e 2019, 11, x FOR PEER REVIEW 7 o 17 Figu e 6. Nea UV/VIS spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . Figu e 7. VIS/NIR spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . 3.2. Mic ocys is G ow h Inhibi ion unde Op imal Condi ions wi h Nu ien En ichmen We unde ook an ini ial 8-day expe imen al ial, wi h samples aken a e one, h ee and i e ea men cycles, in o de o assess he numbe o cycles needed o (a) p omo e Mic ocys is damage, and (b) assess whe he cyanobac e ia a e able o eco e g ow h and me abolic ac i i y ollowing HC + plasma ea men unde op imal condi ions. This i s es con i med h ee cycles as being su icien o inhibi cyanobac e ial g ow h unde bo h HC and HC + plasma ea men s, wi h he esul s o HC + plasma e en mo e p onounced han o HC alone (Figu es 8 and 9). While g ow h o HC- ea ed Mic ocys is began o eco e one week a e a single cycle o ea men (Figu e 8), i emained s ongly inhibi ed ollowing ea men by HC + plasma, wi h li le me abolic ac i i y e iden a e eigh days unde op imal cul i a ion condi ions (Figu e 9). Figu e 6. Nea UV/VIS spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . Wa e 2019, 11, x FOR PEER REVIEW 7 o 17 Figu e 6. Nea UV/VIS spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . Figu e 7. VIS/NIR spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . 3.2. Mic ocys is G ow h Inhibi ion unde Op imal Condi ions wi h Nu ien En ichmen We unde ook an ini ial 8-day expe imen al ial, wi h samples aken a e one, h ee and i e ea men cycles, in o de o assess he numbe o cycles needed o (a) p omo e Mic ocys is damage, and (b) assess whe he cyanobac e ia a e able o eco e g ow h and me abolic ac i i y ollowing HC + plasma ea men unde op imal condi ions. This i s es con i med h ee cycles as being su icien o inhibi cyanobac e ial g ow h unde bo h HC and HC + plasma ea men s, wi h he esul s o HC + plasma e en mo e p onounced han o HC alone (Figu es 8 and 9). While g ow h o HC- ea ed Mic ocys is began o eco e one week a e a single cycle o ea men (Figu e 8), i emained s ongly inhibi ed ollowing ea men by HC + plasma, wi h li le me abolic ac i i y e iden a e eigh days unde op imal cul i a ion condi ions (Figu e 9). Figu e 7. VIS/NIR spec um o plasma in he liquid/ca i a ion bubble cloud en i onmen . The emission spec a, showing a omic oxygen, a omic hyd ogen and exci ed ni ogen molecules/ions in he discha ge, clea ly indica e ha wa e apou molecules we e dissocia ed o OH adicals. The p esence o exci ed ni ogen molecules and a omic oxygen also sugges s he p esence o ni ogen oxide species [ 23 ], commonly obse ed in discha ge spec a o discha ges used o p oduce RONS in wa e [ 24 , 25 ]. While o he spec oscopic me hods (e.g., Fou ie ans o m in a ed abso p ion, lase abso p ion, elec on pa amagne ic esonance), along wi h luo escence/colo ime ic measu emen s [ 22 , 26 , 27 ], may be be e placed o con i m and quan i y RONS, ou own spec oscopic e alua ion clea ly showed p oduc ion o RONS by HC +plasma. The spec um also consis s o a omic coppe lines (see Figu es 6and 7) a ising om he ma e ial o me al elec odes. The s onges iden i ied lines we e a 324.8 and 327.4 nm in he UV egion, 510.6, 515.3 and 521.8 nm in he isible egion and 793.3 and 809.3 nm in he NIR egion. Fo he iden i ica ion o he spec al lines and bands he ollowing li e a u e was used: [19–21] Wa e 2020,12, 8 7 o 14 3.2. Mic ocys is G ow h Inhibi ion unde Op imal Condi ions wi h Nu ien En ichmen We unde ook an ini ial 8-day expe imen al ial, wi h samples aken a e one, h ee and i e ea men cycles, in o de o assess he numbe o cycles needed o (a) p omo e Mic ocys is damage, and (b) assess whe he cyanobac e ia a e able o eco e g ow h and me abolic ac i i y ollowing HC + plasma ea men unde op imal condi ions. This i s es con i med h ee cycles as being su icien o inhibi cyanobac e ial g ow h unde bo h HC and HC +plasma ea men s, wi h he esul s o HC +plasma e en mo e p onounced han o HC alone (Figu es 8and 9). While g ow h o HC- ea ed Mic ocys is began o eco e one week a e a single cycle o ea men (Figu e 8), i emained s ongly inhibi ed ollowing ea men by HC +plasma, wi h li le me abolic ac i i y e iden a e eigh days unde op imal cul i a ion condi ions (Figu e 9). Wa e 2019, 11, x FOR PEER REVIEW 8 o 17 Figu e 8. G ow h o a Mic ocys is cul u e ollowing HC ea men . RFU –Rela i e luo escence uni s. Figu e 9. G ow h o a Mic ocys is cul u e ollowing HC + plasma ea men . 3.3. Mechanisms o Mic ocys is Biomass Remo al Based on he esul s o he i s expe imen al ial, we decided on h ee cycles as he op imal ea men egime. In he second expe imen al ial, he e o e, cyanobac e ial ea men samples we e aken a e one, wo and h ee cycles. We included he measu emen o chlo ophyll-a o ou expe imen al design in o de o con i m quan i a i e emo al o cyanobac e ial biomass om he wa e column. The esul s ended o mi o hose om he g ow h inhibi ion es , wi h a mo e apid dec ease in chlo ophyll concen a ion ollowing wo o mo e cycles o HC + plasma ea men compa ed wi h HC ea men alone (Figu es 10 and 11). Cyanobac e ial biomass only showed a slow decline ollowing ea men by HC alone, wi h a no iceable e ec only de ec able a e 3–6 days (Figu e 10). In compa ison, ea men by HC + plasma Figu e 8. G ow h o a Mic ocys is cul u e ollowing HC ea men . RFU –Rela i e luo escence uni s. Wa e 2019, 11, x FOR PEER REVIEW 8 o 17 Figu e 8. G ow h o a Mic ocys is cul u e ollowing HC ea men . RFU –Rela i e luo escence uni s. Figu e 9. G ow h o a Mic ocys is cul u e ollowing HC + plasma ea men . 3.3. Mechanisms o Mic ocys is Biomass Remo al Based on he esul s o he i s expe imen al ial, we decided on h ee cycles as he op imal ea men egime. In he second expe imen al ial, he e o e, cyanobac e ial ea men samples we e aken a e one, wo and h ee cycles. We included he measu emen o chlo ophyll-a o ou expe imen al design in o de o con i m quan i a i e emo al o cyanobac e ial biomass om he wa e column. The esul s ended o mi o hose om he g ow h inhibi ion es , wi h a mo e apid dec ease in chlo ophyll concen a ion ollowing wo o mo e cycles o HC + plasma ea men compa ed wi h HC ea men alone (Figu es 10 and 11). Cyanobac e ial biomass only showed a slow decline ollowing ea men by HC alone, wi h a no iceable e ec only de ec able a e 3–6 days (Figu e 10). In compa ison, ea men by HC + plasma Figu e 9. G ow h o a Mic ocys is cul u e ollowing HC +plasma ea men . 3.3. Mechanisms o Mic ocys is Biomass Remo al Based on he esul s o he i s expe imen al ial, we decided on h ee cycles as he op imal ea men egime. In he second expe imen al ial, he e o e, cyanobac e ial ea men samples we e aken a e one, wo and h ee cycles. We included he measu emen o chlo ophyll-a o ou expe imen al design in o de o con i m quan i a i e emo al o cyanobac e ial biomass om he wa e column. The esul s ended o mi o hose om he g ow h inhibi ion es , wi h a mo e apid dec ease in chlo ophyll concen a ion ollowing wo o mo e cycles o HC +plasma ea men compa ed wi h HC ea men alone (Figu es 10 and 11). Wa e 2020,12, 8 8 o 14 Wa e 2019, 11, x FOR PEER REVIEW 9 o 17 esul ed in a 95% d op in cyanobac e ial biomass wi hin a single day (Figu e 11). Mic oscopic examina ion con i med ha HC ea men a ec ed cell s uc u e, he cellula su ace and, especially, he gas esicles and subcellula s uc u e ha allow Mic ocys is o loa . This was con i med by mic oscopic obse a ion, which showed abundan gas esicles (black do s inside cells) in con ol cells (Figu e 12A) bu a la ge numbe o emp y cells a e jus one ca i a ion cycle (Figu e 12B). Hyd odynamic ca i a ion can be used in wa e ea men echnologies o di e en pu poses ( o e iew, see [28]) and can be combined wi h o he echnologies like ad anced oxygena ion echnologies, e c. He e, we combined HC wi h plasma and, in compa ison o HC use only, all gas esicles had collapsed in e e y cell ollowing he combina ion HC + plasma ea men (Figu e 12C). Howe e , di e en mechanisms we e employed when we combined HC wi h plasma. As a esul , Mic ocys is biomass was almos comple ely emo ed om he wa e column a e one day o ea men , and we hypo hesize ha he cells had insu icien ene gy o main ain me abolic ac i i y. This hypo hesis can be p o en by measu emen o pho osyn he ic ac i i y. Figu e 10. Remo al o cyanobac e ial biomass ollowing HC ea men . Figu e 10. Remo al o cyanobac e ial biomass ollowing HC ea men . Wa e 2019, 11, x FOR PEER REVIEW 10 o 17 Figu e 11. Remo al o cyanobac e ial biomass ollowing HC + plasma ea men . (A) Figu e 11. Remo al o cyanobac e ial biomass ollowing HC +plasma ea men . Cyanobac e ial biomass only showed a slow decline ollowing ea men by HC alone, wi h a no iceable e ec only de ec able a e 3–6 days (Figu e 10). In compa ison, ea men by HC +plasma esul ed in a 95% d op in cyanobac e ial biomass wi hin a single day (Figu e 11). Mic oscopic examina ion con i med ha HC ea men a ec ed cell s uc u e, he cellula su ace and, especially, he gas esicles and subcellula s uc u e ha allow Mic ocys is o loa . This was con i med by mic oscopic obse a ion, which showed abundan gas esicles (black do s inside cells) in con ol cells (Figu e 12A) bu ala genumbe o emp ycellsa e jus one ca i a ion cycle(Figu e12B). Hyd odynamic ca i a ion can be used in wa e ea men echnologies o di e en pu poses ( o e iew, see [ 28 ]) and can be combined wi h o he echnologies like ad anced oxygena ion echnologies, e c. He e, we combined HC wi h plasma and, in compa ison o HC use only, all gas esicles had collapsed in e e y cell ollowing he combina ion HC +plasma ea men (Figu e 12C). Howe e , di e en mechanisms we e employed when we combined HC wi h plasma. As a esul , Mic ocys is biomass was almos comple ely emo ed om he wa e column a e one day o ea men , and we hypo hesize ha he cells had insu icien ene gy o main ain me abolic ac i i y. This hypo hesis can be p o en by measu emen o pho osyn he ic ac i i y. Wa e 2020,12, 8 9 o 14 Wa e 2019, 11, x FOR PEER REVIEW 10 o 17 Figu e 11. Remo al o cyanobac e ial biomass ollowing HC + plasma ea men . (A) Wa e 2019, 11, x FOR PEER REVIEW 11 o 17 (B) (C) Figu e 12. (A) Abundan gas esicles (black do s inside cells) in con ol Mic ocys is cells; (B) Mic ocys is cells a e he i s HC ca i a ion cycle. Red a ows show he cells wi h collapsed gas esicles; (C) Mic ocys is cells a e HC + plasma ea men , indica ing in ac cells wi h collapsed gas esicles. Cyanobac e ial pho osyn he ic ac i i y, measu ed using dynamic o chlo ophyll luo escence (ChlFl) cha ac e is ics, exhibi ed signi ican changes al eady a e 24 h a e HC ea men , wi h he e ec s pa icula ly no iceable ollowing ea men wi h HC + plasma (Figu e 13). Compa ed o he un ea ed con ol, basal ChlFl (F0) alues inc eased in samples exposed o h ee cycles o HC and HC + plasma (Figu e 13A). A conside able d op in F0 alues was obse ed 48 and 72 h a e HC + plasma Figu e 12. ( A ) Abundan gas esicles (black do s inside cells) in con ol Mic ocys is cells; ( B )Mic ocys is cells a e he i s HC ca i a ion cycle. Red a ows show he cells wi h collapsed gas esicles; (C)Mic ocys is cells a e HC +plasma ea men , indica ing in ac cells wi h collapsed gas esicles.