Ana Maria Freitas Da Silva - RUN: Página Principal
Ana Maria Freitas da SilvaLicenciatura em Ciências da Engenharia Química e BioquímicaA SIMPLE CLOSED-LOOP MEMBRANEPROCESS FOR THE PURIFICATION OFACTIVE PHARMACEUTICAL INGREDIENTSDissertação para obtenção do Grau de Mestre emEngenharia Química e BioquímicaOrientadores:Co-orientador:Professor Andrew G. Livingston (IC)Professor João G. Crespo (FCT-UNL)Jeong F. Kim (IC)IMPERIAL COLLEGE LONDONFaculty of EngineeringDepartment of Chemical Engineering and Chemical TechnologyUNIVERSIDADE NOVA DE LISBOAFaculdade de Ciências e tecnologiaDepartamento de QuímicaNovembro 2012
Copyright Ana Maria Freitas da Silva, FCT-UNL, UNLA Faculdade de Ciências e Tecnologia e a Universidade Nova de Lisboa tem o direito, perpétuoe sem limites geográficos, de arquivar e publicar esta dissertação através de exemplaresimpressos reproduzidos em papel ou de forma digital, ou por qualquer outro meio conhecido ouque venha a ser inventado, e de a divulgar através de repositórios científicos e de admitir a suacópia e distribuição com objectivos educacionais ou de investigação, não comerciais, desdeque seja dado crédito ao autor e editor.
IV
ACKNOWLEDGMENTFirst of all, I would like to thank Professor Andrew Livingston for receiving me in hisresearch group.I am deeply grateful to Jeong Kim, my daily supervisor, who helped me through the wholeproject, including writing this thesis.I would like to thank all the other members of the Separation Engineering and Technologyresearch group for their support during my stay at Imperial College.I would like also to thank Professor João Crespo and Professor Isabel Coelhoso for beingalways available to help with any questions and showing interest in my work.Last but not least, I would like to thank my family and friends for their emotional help,encouragement and understanding.V
VI
ABSTRACTHere we present a simple closed-loop process for the purification of active pharmaceuticalingredients (API) that combines two Organic Solvent Nanofiltration (OSN) membranes, one forpurification and another for solvent recovery. Its success depends on the membrane used forsolvent recovery that should only let the solvent pass through it.-1A mixture of Solvent Yellow 7 (MW 198.2g.mol ) (SY7) and Brilliant Blue R-1(MW 826.0g.mol ) (BBR) in N,N-dimethylformamide (DMF) – model mixture A - and a mixture-1of Martius Yellow (MW 274.16 g.mol ) (MY) and BBR in methanol (MeOH) – model mixture C-, were purified using the system proposed. Although the process operated as predicted, thesemixtures were challenging in terms of separation so it was difficult to find a membrane tighterenough for the solvent recovery purpose. Thus, to achieve yields 90% it was necessary todisconnect the two membrane units at some point and continue the diafiltration with only thefirst membrane, using fresh solvent.Experiments with model mixture A showed that the tighter the membrane used for solventrecovery, the small the volume of solvent required to achieve the target purity and yield.Experiments with model mixture C showed a maximum reduction of 59% in the MeOH usage,comparing to the amount of solvent required by a single membrane process to achieve thesame yield and purity.The effect of increasing the number of membranes for purification was assessed throughsimulations. It was found that the product yield can be increased from 1% to 97% by justincreasing the number of membranes units from 1 to 3. However, the purity can drop from 99%to 54% due to the exponential increase of the overall rejections with the number of membraneunits in the membrane cascade.Keywords: Organic Solvent Nanofiltration (OSN), Active Pharmaceutical Ingredients (API),Solvent Recovery, Membrane CascadesVII
VIII
RESUMOApresenta-se aqui um processo simples em circuito fechado para a purificação deingredientes farmacêuticos activos, que combina duas membranas de nanofiltração comsolventes orgânicos, uma para purificação e outra para recuperação de solvente. O sucessodeste processo depende da membrana utilizada na recuperação de solvente, que deverá serapenas permeável ao solvente.-1Uma mistura de Solvent Yellow 7 (MW 198.2g.mol ) (SY7) e Brilliant Blue R-1(MW 826.0g.mol ) (BBR) em N,N-dimetilformamida (DMF) – mistura modelo A – e uma-1mistura de Martius Yellow (MW 274.16 g.mol ) (MY) e BBR em metanol (MeOH) – misturamodelo C-, foram purificadas utilizando o sistema proposto. Apesar do processo ter funcionadocomo previsto, estas misturas revelaram-se difíceis de separar, pelo que foi difícil encontraruma membrana suficientemente densa para a recuperação de solvente. Assim, para alcançarum rendimento de produto 90%, foi necessário desconectar as duas membranas a dada alturae continuar a diafiltração com apenas a primeira membrana, utilizando solvente novo.Experiências com a mistura modelo A demonstraram que quanto mais densa é amembrana utilizada na recuperação de solvente, menor o volume de solvente necessário paraatingir o rendimento e a pureza desejados. Experiências com a mistura modelo Cdemonstraram uma redução máxima do consumo de MeOH em 59%, comparando com aquantidade de solvente exigida por um processo com uma única membrana para alcançar omesmo rendimento e a mesma pureza.O efeito de aumentar o número de membranas para purificação foi avaliado através desimulações. Verificou-se que o rendimento pode aumentar de 1% para 97% aumentando onúmero de membranas de 1 para 3. No entanto, a pureza pode diminuir de 99% para 54%devido ao aumento exponencial das rejeições globais com o número de membranas nacascata.Termos chave: Nanofiltração com Solventes Orgânicos, Ingredientes Farmacêuticos Activos,Recuperação de Solventes, Cascatas de MembranasIX
X
TABLE OF CONTENTSLIST OF FIGURES . XVLIST OF TABLES. XXIABREVIATIONS. XXVNOMENCLATURE . XXVIGREEK SYMBOLS . XXVIISUBSCRIPTS. XXVIICHAPTER 1 LITERATURE REVIEW11.1.1Membrane technology1.1.1.The Membrane11.1.2.Membrane types21.1.3.Membrane preparation techniques31.1.4.Membrane processes51.1.5.Membrane characterization71.1.6.Transport in membranes91.1.7.Strengths and limitations of membrane processes1.2.Solvent use in the pharmaceutical industry11161.2.1.Solvent utilization161.2.2.Waste minimization and solvent recovery171.2.3.Use of membrane technology for solvent recovery181.3.Membrane cascades211.3.1.Module configurations and mode of operation211.3.2.Membrane cascades configurations and modes of operation.221.3.3.Membrane cascades applications241.4.Implication of the Literature Review and Research Motivation27CHAPTER 2 MATERIALS & METHODS282.1.Model mixtures282.2.Membranes30XI
2.2.1.Integrally Skinned Membranes302.2.2.Thin Film Composite Membranes322.2.3.Membrane performance332.3.Membrane Filtration342.3.1.Cross-flow Filtration342.3.2.Experimental set-ups352.4.Analytical Methods382.4.1.UV/Vis Spectroscopy382.4.2.High Pressure Liquid Chromatography (HPLC)382.5.Process Modeling39CHAPTER 3 CLOSED-LOOP MEMBRANE PROCESS403.1.41Model Mixture A3.1.1.Experimental413.1.2.Results & Discussion423.1.3.Conclusions643.2.Model mixture B663.2.1.Experimental663.2.2.Results & Discussion663.2.3.Conclusions68Model Mixture C703.3.3.3.1.Experimental703.3.2.Results & Discussion723.3.3.Conclusions96CHAPTER 4 MEMBRANE CASCADE974.1.99Single membrane experiments4.1.1.Experimental4.1.2.Results & Discussion991004.2.Membrane cascade modelling1064.3.Conclusions107CHAPTER 5 CONCLUSION REMARKS AND FUTURE WORKXII109
REFERENCES110APPENDIX A. HPLC CALIBRATION CURVES118APPENDIX B. EFFECT OF TEMPERATURE ON THE VISIBLE SPECTRAOF MY AND BBR COMPOUNDS119APPENDIX C. PROCESS MODELING120C.1 Single membrane system120C.2 Closed-loop membrane process121C.3 Membrane cascade121XIII
XIV
LIST OF FIGURESFigure 1.1Schematic representation of a membrane separation: component 1 selectivelypasses through the membrane, driven by a chemical or electrical potential gradient. Adaptedfrom (Mulder, 1996). . 1Figure 1.2Principal types of membranes. Adapted from (Baker, 2004, Mulder, 1996). . 2Figure 1.3Typical rejection curves for membranes with a A) sharp cut-off and a B) diffusecut-off. Adapted from (Mulder, 1996). . 8Figure 1.4Schematic representation of the most important transport mechanisms throughmembranes: (a) the pore-flow model and (b) the solution-diffusion model (Baker, 2004). . 10Figure 1.5Concentration polarization: concentration profile under steady-state conditions. 12Figure 1.6Flux as a function of the applied pressure both for pure water and for a solution:the flux of pure water increases lineary with applied pressure, considering that no compactionoccurs; however when solutes are added to water the flux starts platteauing after a certainapplied pressure (Mulder, 1996). . 13Figure 1.7Typical pharmaceutical batch operation. Adapted from (Dunn, 2010). . 16Figure 1.8The two basic module operations: (a) Dead-end and (b) Cross-flow (Mulder,1996). . 21Figure 1.9The arrangement of membrane units and stages in a cascade. Adapted from(Benedict, Pigford & Levi, 1981). 22Figure 1.10Modes of cascade operation: (a) simple cascade and (b) countercurrent recyclecascade. Adapted from (Villani & Becker, 1979). . 23Figure 2.1Process diagram of the cross-flow cell rig used for membrane screening tests (F:Flow meter; P: Pressure gauge; T: Temperature thermocouple). . 34Figure 2.2Schematics of the (a) front and (b) top views of the custom-made membrane cellsused for the experimental set-ups of this work. Adapted from (Lin & Livingston, 2007). . 35XV
Figure 2.3Process diagram of the single membrane system (HPLC-P: HPLC pump; BPR:Back pressure regulator; PRV: Pressure relief valve; PI: Pressure gauge; TC: Temperaturecontroller). 36Figure 2.4Process diagram of the closed loop system. (HPLC-P: HPLC pump; BPR: Backpressure regulator; PRV: Pressure relief valve; PI: Pressure gauge; TC: Temperaturecontroller). 37Figure 3.1Schematic diagram of the closed-loop membrane process. . 40Figure 3.2Photograph of the feed solution at the beginning (F0, 40 times diluted) and after 8diafiltration volumes (F8, 8 times diluted) and of the permeate samples at each diafiltrationvolume (P1-P8, without dilution). . 45Figure 3.3Calculated and experimental mass profiles of product (SY7) and waste (BBR) in(a) feed tank and (b) membrane cell for the purification of model mixture A using singlemembrane process with a PI2411 membrane disc at 30 bar and 22 C. Curve fitting was doneby assigning different values to/Figure 3.4: Model 1-; Model 3-/, Model 2-/. . 46Yield and purity profiles for the purification of model mixture A using singlemembrane process with a PI2411 membrane disc at 30 bar and 22 C. Curve fitting was doneby assigning different values to/Figure 3.5: Model 1-; Model 3-//, Model 2. . 47Calculated and experimental mass profiles of the product (SY7) and waste (BBR)in (a) feed tank, (b) membrane cell 1 and (c) membrane cell 2 for the purification of modelmixture A using the closed-loop membrane process with a PI2411 and a DM150 membranedisc. Curve fitting was done by assigning different values to, Model 2-and: Model 1, Model 3-andand. . 50Figure 3.6Yield and purity profiles for the purification of model mixture A using closed-loopmembrane process with a PI2411 and a DM150 membrane discs. Curve fitting was done byassigning different values to: Model 12-/and/and/and/,Model/,Model/. . 523-XVI
Figure 3.7Calculated and experimental mass profiles of the product (SY7) and waste (BBR)in (a) feed tank and (b) membrane cell 1 after disconnecting the second cell from the systemPI24411/DM150 (Figure 3.8). . 54Yield and purity profiles for the purification of model mixture A after disconnectingthe second cell from the system PI24411/DM150 (Figure 3.9). . 55Calculated and experimental mass profiles of the product (SY7) and waste (BBR)in (a) feed tank, (b) membrane cell 1 and (c) membrane cell 2 for the purification of modelmixture A using closed-loop membrane process with a PI2411 and a TFC-MPD discs. Curvefitting was done by assigning different values toand/,: Model 1-Model2-/3-and/,/. . 57Figure 3.10Model//andYield and purity profiles for the purification of model mixture A using the closed-loop membrane process with a PI2411 and a TFC-MPD disc Curve fitting was done byassigningdifferentFigure /. . 59Calculated and experimental mass profiles of the product (SY7) and waste(BBR) in (a) feed tank and (b) membrane cell 1 after disconnecting the second cell from thesystem PI24411/TFC-MPD. Curve fitting was done by assigning different values toModel 1-/Figure 3.12, Model 2-/:. . 61Yield and purity profiles for the purification of model mixture A afterdisconnecting the second cell from the system PI24411/TFC-MPD. Curve fitting was done byassigning different values toandFigure 3.13: Model 1-and, Model 2-. . 62Effect of product rejection in the solvent recovery stage in a closed-loopmembrane process to the product yield. Simulations were performed assuming thatandFigure 3.14. . 64Calculated yield and purity profiles for the purification of model mixture B usingthe closed-loop membrane process with PI2211 membrane (purification and TFC-MPD membrane (/XVII/) for) for solvent recovery. . 68
Figure 3.15Flux profiles over time at 10 bar and 27 C for the membranes screened usingmodel mixture C. The flux at time 0 is the flux of pure MeOH. . 73Figure 3.16Calculated and experimental mass profiles of product (MY) and waste (BBR) in(a) feed tank and (b) membrane cell for the purification of model mixture C using singlemembrane process with a PI2211 disc at 30 bar and 22 C. . 78Figure 3.17Calculated and experimental concentration profiles in the permeate for thepurification of model mixture C using the single membrane process with a PI2211 disc at 30 barand 22 C. 79Figure 3.18Photograph of the initial and final feed and of retentate and permeate samples ateach diafiltration volume. Feed and retentate samples were 40 times diluted. . 79Figure 3.19Yield and purity profiles for the purification of model mixture C using singlemembrane process with a PI2211 disc at 30 bar and 22 C. . 80Figure 3.20Calculated and experimental mass profiles of the product (MY) and waste (BBR)in (a) feed tank, (b) membrane cell 1 and (c) membrane cell 2 for the purification of modelmixture C using the closed-loop membrane process with a PI2211 and a TFRO-SG disc at 25 C. Curve fitting was done by assigning different values to the rejections: Model 1,/,/andFigure 3.21and/,Model2-. . 83Yield and purity profiles for the purification of model mixture C using closed-loopmembrane process with a PI2211 and a TFRO-SG disc at 25 C. Curve fitting was done byassigning different values to the rejections: Model 1/,Model2-,//, andand. . 84Figure 3.22Yield and purity profiles for the purification of model mixture C afterdisconnecting the second cell from the system PI2211/TFRO-SG. . 86Figure 3.23Experimental and calculated concentration profiles for the product (MY) andwaste (BBR) in the permeate. . 86Figure 3.24Calculated and experimental mass profiles of product (MY) and waste (BBR) in(a) feed tank and (b) membrane cell for the purification of model mixture C using singlemembrane process with a PBI 24xDBX disc at 21 bar and 23 C. Curve fitting was done byXVIII
assigning different values to/and: Model 1-; Model 3-Figure 3.25//; Model 2-. . 89Yield and purity profiles for the purification of model mixture C using singlemembrane process with a PBI 24xDBX disc at 21 bar and 23 C. Curve fitting was done byassigning different values to/and: Model 1-; Model 3-Figure 3.26//; Model 2-. . 90Experimental and calculated concentration profiles for the product (MY) andwaste (BBR) in the permeate. Curve fitting was done by assigning different values to: Model 1-//; Model 2-/and; Model 3-. . 91Figure 3.27Calculated and experimental mass profiles of the product (MY) and waste (BBR)in (a) feed tank, (b) membrane cell 1 and (c) membrane cell 2 for the purification of modelmixture C using closed-loop membrane process with a PBI 24xDBX and a PBI 26xDBB disc.Curve fitting was done by assigning different values to,andandFigure 3.28and: Model 1-; Model 2-and,,. . 93Yield and purity profiles for the purification of model mixture C using closed-loopmembrane process with a PI2211 and a TFRO-SG disc. Curve fitting was done by assigningdifferent values toand; Model 2-Figure 3.29: Model 1-,and,,andand. . 94Yield and purity profiles for the purification of model mixture C afterdisconnecting the second cell from the system PBI 24xDBX/PBI 26xDBB. Curve fitting wasdone by assigning different values toModel 2-Figure 4.1/and: Model 1-/;; . 95Schematic representation of the closed-loop membrane process integrating acascade of n membrane units for purification. . 97Figure 4.2Flux profiles over time at 10 bar and 25 C for the membranes screened usingmodel mixture D1. The flux at time 0 is the flux of pure MeCN. . 100Figure 4.3Rejection curves in for the membranes screened in MeCN at 10 bar and 25 Cafter a) 5 hours and b) 24 hours of continuous operation. . 101XIX
Figure 4.4Flux profiles over time at 10 bar and 25 C for the membranes screened usingmodel mixture D2. The flux at time 0 is the flux of pure MeCN. . 102Figure 4.5Experimental and calculated mass profiles for the polymer (product) and monomer(waste) using the single membrane process with a PI2341 membrane at 10-15 bar and 21 C.Calculated curves were obtained considering the an average of the calculated rejections,and. (Model 1) and corrected values,and. (Model 2). . 105Figure 4.6Yield and purity profiles as a function of the number of membrane units used forpurification. Simulations performed by assuming thatand. 106XX
LIST OF TABLESTable 1.1Classification of membrane processes according to their driving forces (Mulder,1996). . 5Table 1.2Comparison of various pressure driven membrane processes. Adapted from(Mulder, 1996). . 6Table 1.3Comparison of solvent use at GSK based on overall manufacturing operations(1995-2000) and more recent pilot plant operations (2005) (Constable, Jimenez-Gonzalez &Henderson, 2007). . 17Table 2.1Properties of the dyes used as model compounds. . 29Table 2.2Summary of the membranes not prepared but used in this work . 30Table 2.3Summary of the integrally skinned PI membranes prepared. 31Table 2.4Summary of the integrally skinned PBI membranes prepared. . 32Table 2.5Summary of the TFC membranes prepared. . 33Table 3.1Summary of the screening results for 14 cm membrane discs using model mixture2A (Mix A). . 43Table 3.2Summary of the data recorded during the diafiltration of model mixture A using thesingle membrane process with a PI2411 membrane disc. . 44Table 3.3Summary of the data recorded along the diafiltration of model mixture A using theclosed-loop membrane process with a PI2411 membrane disc for purification and a DM150membrane disc for solvent recovery. . 49Table 3.4Summary of the data recorded along the diafiltration of model mixture A using thesingle membrane process with a PI2411 membrane disc after disconnected the cells. . 53Table 3.5Summary of the data recorded along the diafiltration of model mixture A using theclosed-loop process with a PI2411 membrane disc for purification and a TFC-MPD membranedisc for solvent recovery. . 56XXI
Table 3.6Summary of the data recorded along the diafiltration of model mixture A using thesingle membrane process with a PI2411 membrane disc after disconnected the cells. . 60Table 3.7Comparison between the three processes used to purify model mixture A. . 63Table 3.8Summary of the screening data. . 67Table 3.9Physical parameters of the solvents. . 74Table 3.10Rejections after 5h and 24h of operation at 10 bar and 27 C for the membranesscreened using model mixture C. . 74Table 3.11Summary of the data recorded along the diafiltration of model mixture C using thesingle membrane process with a PI2211 membrane. . 77Table 3.12Summary of the data recorded along the purification of model mixture C using theclosed-loop membrane process with a PI2211 membrane disc for purification and a TFRO-SGmembrane disc for solvent recovery. . 81Table 3.13Summary of the data recorded along the purification of model mixture C using thesingle membrane process with a PI2211 membrane disc after disconnecting the second cellfrom the system. . 85Table 3.14Comparison between the performance of the single membrane process withPI2211 membrane and the closed-loop membrane process with PI2211 and TFRO-SGmembranes. 87Table 3.15Summary of the data recorded along the purification of model mixture C using thesingle membrane process with a PBI 24xDBX membrane. . 88Table 3.16Summary of the data recorded along the purification of model mixture C using theclosed-loop membrane process with a PBI 24xDBX membrane disc for purification and a PBI26xDBB membrane disc for solvent recovery. . 92Table 3.17Summary of the data recorded along the diafiltration of model mixture C using thesingle membrane process with a PBI 24xDBX membrane disc after disconnected the cells. . 95Table 3.18Comparison between the performance of the single membrane process with PBI24xDBX membrane and the closed-loop membrane process with PBI 24xDBX and PBI 26xDBBmembranes. 96XXII
Table 4.1Rejections after 2h of operation at 10 bar and 25 C for the different membranestested against model mixture D2. . 103Table 4.2Performance data for the purification of the product crude using PI2341 membraneat 21 C. 104Table 4.3Overall rejections, Rcascade,i, of the separation stage as a function of the number ofmembrane units used for purification. . 107XXIII
XXIV
ABREVIATIONSAPI(s)Active pharmaceutical Ingredient(s)BBRBrilliant Blue RBPRBack pressure etylacetamide (DMAc)DMFN,N-dimethylformamideGHGsGreen house gasesHDA1,6-HexanediamineHPLCHigh pressure liquid thanolMFMicrofiltrationMPDm-phenylene diamineMWMolecular weight (g.mol )MWCO-1Molecular weight cut-offMYMartius YellowNFNanofiltrationOSNOrganic solvent oly(ether ether ketone)PEGPolyethylene glycolPIPolyimidePPPolypropylenePRVPressure relief valvePSPolystyreneROReverse OsmosisSRNFSolvent resistant nanofiltrationSY7Solven Yellow 7TFCThin-film avioletVOCsVolatile organic compoundsXXV
NOMENCLATUREFraction in the retentate (dimensionless)Fraction in the permeate (dimensionless)Rejection of solute (%)-1Concentration (g.L )Pressure (bar)Temperature ( )-2-1Solute flux (g.m .h )-2-1Solvent flux (L.m .h )-2-1-1Hydrodynamic permeability (L.m .h .bar )̅-1Logarithmic average of solute concentration across the membrane (g.L )Radius (m)Solute hindrance factor for convection (dimensionless)-1Solvent velocity (m.s )-1Faraday constant (96487 C.mol )-1-1Gas constant (8.341 J.mol .K )2-1Corrected diffusive coefficient according to (Bowen & Welfoot, 2002) (m .s )Valence (dimensionless)Fick’s aw diffusiv c2ffici-1t (m .s )Sorption coefficient (dimensionless)Membrane thickness (m)-3Concentration of the solute in the feed side (mol.m )-3Concentration of solute in the permeate side (mol.m )2Membrane area (m )-2-1-1Solvent permeability ((L.m .h .bar )Volume (L)Time (h)Number of diafiltration volumes (dimensionless)Mass (g)Yield of product (%)Purity of product (%)Number of membrane units in the cascade (dimensionless)-1Volumetric flow rate (L.h )XXVI
GREEK SYMBOLSSeparation factor (dimensionless)Reflection coefficient (dimensionless)Osmotic pressure (bar)-2-1-1Osmotic permeability (g.m .h .bar )Surface porosity (dimensionless)Pore tortuosity (dimensionless)-
1.2.3. Use of membrane technology for solvent recovery 18 1.3. Membrane cascades 21 1.3.1. Module configurations and mode of operation 21 1.3.2. Membrane cascades configurations and modes of operation. 22 1.3.3. Membrane cascades applications 24 1.4. Implication of the Literature Review and Research Motivation 27 CHAPTER 2 MATERIALS & METHODS .
Nagila Freitas Souza, Kílvia de Freitas Alves, Morsyleide de Freitas Rosa, Maria Cléa Figueiredo de Brito . 222 Environmental impact analysis of different
2 La hoomaka iho Hei keia KaIlUa Haoulu 0 ka Oihana 1Ia hiai 0 Amelika ma Honolulu a ma Hila HO hoi, e hooulu i ke kalo malalo 0 na ana hou 0 keia au naauao i ikeia ai ke ano ku pono loa a ka lehulchu e hoohana ai i maopopo ke ana 0 ka mahi ana, ke kanu ana, ke kanu kahuli ana i ke kalo i kahi 00 ai, a i mea okoa aIm i kekahi kau, a me ke ana 0 na mea hoomomona .
SANTA ANA CITY COUNCIL Vicente Sarmiento Mayor vsarmiento@santa-ana.org David Penaloza Mayor Pro Tem, Ward 2 dpenaloza@santa-ana.org Thai Viet Phan Ward 1 tphan@santa-ana.org Jessie Lopez Ward 3 jessielopez@santa-ana.org Phil Bacerra Ward 4 pbacerra@santa-ana.org Johnathan Ryan Hernandez Ward 5 jryanhernandez@santa-ana.org Nelida Mendoza
su modelo. Leen poemas ante su tumba los poetas Guillermo Valencia ("Leyendo a Sil-va") y Víctor María Londoño ("A José Asun-ción Silva"). Roberto Suárez publica en Re-pertorio Colombiano"Paréntesis", una antología de poemas de Silva. En 1908 aparecen publicados en España las Poesías de Silva editadas por
serradilla garzon, maria fraile conde, maria de las mercede del campo alvarez, ana maria gomez hernandez, monica santos bautista, fatima morchÓn gil, irene gÓmez isabel, luisa maria terron marcos, sara isabel calderon guzman, raquel uriarte palomino, maria del mar alonso gonzalez, maria
Doña Ana County Community Report Figure 3: Alcohol-involved Crashes by Hour in Doña Ana County, 2019 Figure 2: Crashes by Hour in Doña Ana County, 2019 * In 2019, Doña Ana County had 0 alcohol-involved crashes for which hour data were missing. * In 2019, Doña Ana County had 4 crashes for which hour data were missing. 67 50 47 47 33 35 78 .
Ave maria Caccini Ave maria Schubert Avé Maria Beyoncé Ave maria Gounod Avé Maria Michal Lorenc Avé Maria Mascagni Avé Maria Camille Saint-Saëns Avé Verum Corpus Mozart Agnus Dei Lloyd Webber Agnus Dei Michael Smith Agnus Dei Samuel Barber Agn
ANA CAROLINA SILVA ANA CAROLINA SILVA ANA CAROLLINA ROCHA QUINTAO CAMPUS UNIVERSITÁRIO, S/N MORRO DO CRUZEIRO CEP 354000-000 OURO PRETO M.G BRASIL Home page: www.prace.ufop.br - Telefone: 0xx(31) 3559-1271 - E-mail faleconosco@prace.ufop.br Lista gerada pelo Sistema Integrado de Moradia (SIM) em, 14 de Fevereiro de 2019
