New insights into the chemistry of laser-induced plasmas. Cutting edge strategies for sensing and characterization of molecular solids

AUTOR: Jorge Serrano Rojas http://orcid.org/0000-0002-3778-7885 EDITA: Publicaciones y Divulgación Científica. Universidad de Málaga Esta obra está bajo una licencia de Creative Commons Reconocimiento-NoComercialSinObraDerivada 4.0 Internacional: http://creativecommons.org/licenses/by-nc-nd/4.0/legalcode Cualquier parte de esta obra se puede reproducir sin autorización pero con el reconocimiento y atribución de los autores. No se puede hacer uso comercial de la obra y no se puede alterar, transformar o hacer obras derivadas. Esta Tesis Doctoral está depositada en el Repositorio Institucional de la Universidad de Málaga (RIUMA): riuma.uma.es Doctoral Thesis NEW INSIGHTS INTO THE CHEMISTRY OF LASER-INDUCED PLASMAS. CUTTING EDGE STRATEGIES FOR SENSING AND CHARACTERIZATION OF MOLECULAR SOLIDS by JORGE SERRANO ROJAS THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS TO APPLY FOR THE DEGREE OF DOCTOR Departamento de Química Analítica Facultad de Ciencias Universidad de Málaga Málaga, 2016 NEW INSIGHTS INTO THE CHEMISTRY OF LASER-INDUCED PLASMAS. CUTTING EDGE STRATEGIES FOR SENSING AND CHARACTER¡ZATION OF MOLECULAR SOLIDS por JORGE SERRANO ROJAS José Javier Laserna Vázquez Catedrático de Universidad Departamento de Química Analítica Universidad de Málaga Javier Moros Portolés I nvestigador postdoctoral Departamento de Química Analítica Universidad de Málaga Memoria de Tesis presentada para optar al grado de Doctor Jorge Serrano Rojas Málaga, Junio de 2016 JOSÉ JAVIER LASERNA VÁZQUEZ, Catedrático de Química Analítica de la Universidad de Málaga, y JAVIER MOROS PORTOLÉS, Investigador Postdoctoral del Departamento de Química Analítica de la Universidad de Málaga CERTIFICAN Que JORGE SERRANO ROJAS ha realizado bajo su dirección la presente Tesis Doctoral titulada “NEW INSIGHTS INTO THE CHEMISTRY OF LASER-INDUCED PLASMAS. CUTTING EDGE STRATEGIES FOR SENSING AND CHARACTERIZATION OF MOLECULAR SOLIDS” en el Laboratorio Láser del Departamento de Química Analítica de la Universidad de Málaga, y que el conjunto de publicaciones aportadas para avalar el trabajo científico no han sido utilizadas en Tesis anteriores, reuniendo a nuestro juicio los requisitos necesarios y autorizando, por ello, su presentación para optar al grado de Doctor. Y para que así conste a los efectos oportunos firman la presente, en Málaga, a 14 de Junio de 2016. Prof. José Javier Laserna Vázquez Dr. Javier Moros Portolés V TESIS DOCTORAL POR COMPENDIO DE PUBLICACIONES En cumplimento con los requisitos especificados en el Reglamento de Doctorado de la Universidad de Málaga, la presente Tesis Doctoral ha sido autorizada por los Directores de Tesis y el Órgano Responsable del Programa de Doctorado para ser presentada en el formato de "compendio de publicaciones". Las referencias de los artículos en los que el doctorando figura como primer o segundo autor y que avalan la presente Tesis Doctoral se detallan a continuación de acuerdo a su orden cronológico de publicación: J. Moros, J. Serrano, C. Sánchez, J. Macías, J. J. Laserna, New chemometrics in laserinduced breakdown spectroscopy for recognizing explosive residues, J. Anal. At. Spectrom. 27 (2012) 2111–2122. J. Moros, J. Serrano, F.J. Gallego, J. Macías, J. J. Laserna, Recognition of explosives fingerprints on objects for courier services using machine learning methods and laserinduced breakdown spectroscopy, Talanta 110 (2013) 108–117. J. Serrano, J. Moros, C. Sánchez, J. Macías, J. J. Laserna, Advanced recognition of explosives in traces on polymer surfaces using LIBS and supervised learning classifiers, Anal. Chim. Acta 806 (2014) 107–116. J. Serrano, L.M. Cabalín, J. Moros, J.J. Laserna, Potential of laser-induced breakdown spectroscopy for discrimination of nano-sized carbon materials. Insights on the optical characterization of graphene, Spectrochim. Acta Part B 97 (2014) 105–112. VII 1 RESUMEN Dado que la presente memoria de Tesis Doctoral está redactada en inglés, en cumplimiento del Reglamento de Doctorado de la Universidad de Málaga, se incluye un resumen de la misma en español. Desde su origen, la espectroscopia de plasmas inducidos por láser, más conocida por su acrónimo inglés LIBS (Laser-induced breakdown spectroscopy) ha sido empleada principalmente para la detección y cuantificación multielemental en el análisis de materiales inorgánicos. Sin embargo, su aplicación al análisis y caracterización de sólidos moleculares ha ido creciendo en popularidad debido a las atractivas características que dicha técnica ofrece. Entre ellas, se pueden destacar su potencial para realizar análisis in-situ de todo tipo de muestras (sólidos, líquidos y gases) sin necesidad de llevar a cabo una preparación previa de las mismas, su capacidad de detección simultánea y multielemental, su alta velocidad de análisis, y la posibilidad de su implementación en equipos portátiles y de detección a distancias remotas. En lo concerniente al análisis de compuestos orgánicos mediante LIBS, las investigaciones desarrolladas han sido abordadas desde dos vertientes. De un lado, se han realizado exploraciones fundamentales destinadas a la comprensión de los fenómenos de ablación en este tipo de Resumen 2 compuestos, de los mecanismos de ruptura de dichas moléculas y de la química de los plasmas inducidos por láser, principalmente, en la formación de las distintas especies responsables de las respuestas de emisión óptica. De forma paralela, y desde un punto de vista analítico, la técnica LIBS ha sido empleada para la caracterización e identificación de compuestos orgánicos en distintos campos de aplicación. En este contexto, el primer bloque de la presente Tesis Doctoral se centra en las investigaciones desarrolladas para la obtención de un mayor conocimiento de los procesos físico-químicos que tienen lugar desde la generación del plasma hasta su extinción. También se evalúa la influencia de las distintas propiedades del pulso láser (energía, longitud de onda y duración) sobre estos procesos, y en consecuencia, en los patrones de emisión resultantes. Además, se han llevado a cabo exploraciones para intentar descifrar algún tipo de relación entre los patrones de emisión óptica de sólidos moleculares y su estructura química. El segundo bloque de la presente Tesis Doctoral se ha encaminado más a la implementación de la técnica LIBS en distintas aplicaciones que demandan la evaluación de sólidos moleculares. Así, se valoró la capacidad de LIBS como herramienta para la caracterización de materiales nanométricos de carbono en un estudio dirigido a la caracterización de grafeno, así como su potencial en el reconocimiento de compuestos explosivos. A continuación se resumen los resultados más relevantes de la investigación desarrollada, diferenciándose los dos bloques de contenidos anteriormente mencionados. Resumen 3 1. Química de los plasmas inducidos por láser en sólidos moleculares Durante el proceso de ablación láser, las moléculas pueden experimentar una atomización total de los enlaces químicos presentes o bien sufrir una fragmentación parcial. Aunque generalmente se produce una coexistencia de ambos procesos, la estructura molecular de los compuestos orgánicos y las condiciones operacionales empleadas en la ablación láser van a determinar la predominancia de cada uno de ellos. De esto se deduce que los plasmas de compuestos orgánicos pueden estar poblados, además de por electrones, por cantidades distintas y variables de átomos, iones y diferentes fragmentos moleculares. Además, las numerosas reacciones químicas que pueden acontecer en el seno del plasma entre las especies originales liberadas –y también con especies procedentes de la atmósfera que rodea al plasma– conducen a la generación de nuevas especies moleculares. Por lo tanto, cualquier relación entre las emisiones ópticas del plasma de un sólido molecular y su estructura química puede verse notablemente desvirtuada, poniendo así de manifiesto la complejidad existente para la exacta interpretación de su espectroscopia. Típicamente, los espectros LIBS de sólidos moleculares se caracterizan por su alta similitud ya que el número de señales espectrales manifestadas está limitado a líneas atómicas e iónicas de emisión (típicamente, C, H, O y N) y a algunas bandas moleculares de emisión asociadas a pequeños radicales excitados (CN, C2, CH, OH y NH). Además, como se adelantaba, cuando los plasmas son generados en atmósfera de aire se produce una participación activa de sus constituyentes mayoritarios (N2, O2, y vapor de H2O) en la química de la pluma. De este modo, el proceso conocido como ionización secundaria del aire puede conducir a emisiones de H, O y N, y distorsionar la relación de estas emisiones con respecto a la composición química del compuesto analizado. A su vez estas especies atmosféricas pueden reaccionar con especies nativas y dar lugar a la generación de pequeñas moléculas. Resumen 4 Teniendo en cuenta estas dificultades a la hora de interpretar la espectroscopia de plasmas inducidos por láser de compuestos orgánicos, la presente memoria de Tesis Doctoral pretende proporcionar un mayor conocimiento y compresión de la química estos plasmas, así como justificar sus patrones de emisión LIBS y tratar de esclarecer cualquier vínculo con sus estructuras moleculares.  Efectos de la estructura química de sólidos moleculares en los patrones de emisión LIBS Distintos autores han demostrado la capacidad de LIBS para la caracterización de sólidos moleculares mediante el uso de las intensidades de emisión asociadas a distintas especies atómicas o radicales moleculares. Sin embargo, existe una carencia de estudios sobre la influencia que la estructura molecular de los compuestos orgánicos puede tener en sus respuestas ópticas de emisión. Una de las publicaciones incluidas en la presente memoria de Tesis Doctoral se centra en evaluar que efectos originan las distintas estructuras moleculares de los materiales orgánicos en las respuestas de emisión a través del análisis de un amplio conjunto de compuestos estratégicamente seleccionados por sus particularidades estructurales. Así, las variaciones observadas en los patrones ópticos de emisión debidas a la variación en la naturaleza, el número, y la posición de grupos funcionales han sido investigadas. Entre los compuestos analizados en este estudio se incluyen tres hidrocarburos policíclicos aromáticos, dos polímeros, una serie homóloga de bencenos para-disustituidos y dos conjuntos de isómeros de dinitrobenceno (DNB) y dinitrotolueno (DNT). Estos materiales fueron analizados en forma de pastillas, a excepción de los polímeros que se analizaron en su formato original (láminas planas). Los experimentos se desarrollaron en atmósfera de aire y se utilizó como fuente de excitación para la formación de plasmas un láser de Nd:YAG trabajando en su segundo harmónico (longitud de onda de 532 nm). Para evaluar el efecto de la longitud de onda de irradiación también se utilizaron el modo fundamental (λ = 1064 nm) y el cuarto harmónico (λ = 266 nm) de dicho láser. Para el desarrollo de los análisis se ajustaron el diámetro del Resumen 5 haz láser y su energía sobre la muestra para emplear una irradiancia constante de 3,3 GW·cm-2. Los parámetros temporales de adquisición de las señales ópticas se establecieron en un tiempo de retraso tras la incidencia del pulso láser de 1,28 µs, y un tiempo de integración de 1,1 ms, permitiendo así integrar la mayor parte del tiempo de vida del plasma. A continuación se detallan los resultados más relevantes de este estudio. Influencia de la estructura aromática de un compuesto orgánico. El análisis LIBS de tres compuestos aromáticos policíclicos –naftaleno (C10H8), antraceno (C14H10) y pireno (C16H10)– permitió evaluar la influencia de la aromaticidad sobre los patrones de ablación y las emisiones ópticas resultantes. Bajo condiciones operacionales bien definidas, los espectros ópticos de emisión manifestaron señales de C, H, N, O, CN y C2. Las emisiones de N y O, elementos no presentes en las muestras, indicaron la ocurrencia del proceso de ionización secundaria del aire. Mediante este proceso las moléculas de N2 y O2 del aire son disociadas y posteriormente excitadas debido a los mecanismos de interacción láser-plasma y plasma-aire. A su vez, las emisiones CN ponen de manifiesto la participación del nitrógeno ambiental (a nivel atómico o molecular) en la química del plasma. Tal y como se detalla a continuación, son varias las reacciones que pueden conducir a la generación de estos radicales diatómicos. A partir de carbono atómico: C + N → CN ó C + N2 → CN + N A partir de dímeros de carbono: C2 + N →CN + C ó C2 + N2 → 2CN Por otra parte, las emisiones de C2 tienen su origen en los dímeros de carbono liberados directamente de la molécula o bien en procesos de recombinación de átomos de C. Si bien la ocurrencia de ambas rutas va ligada a las condiciones operacionales –principalmente, a la intensidad de la excitación–, la presencia de enlaces carbono-carbono en los compuestos orgánicos se relaciona con mayores emisiones de C2. Teniendo esto en cuenta, la relación de intensidades de Resumen 6 emisión CN a C2 (en adelante expresada como CN/C2) se estableció como un evaluador estadístico de la contribución de los procesos de atomización y fragmentación parcial de las moléculas ablacionadas. Así, valores elevados de CN/C2 indican una mayor atomización del compuesto, mientras que valores bajos de este cociente revelan una mayor fragmentación de las moléculas. Además, se observó una conexión entre este estadístico y la estabilidad de la estructura cíclica de los compuestos conferida por la disposición de los electrones π con los orbitales p de los anillos. Se consideraron tres índices de medida de la aromaticidad: la escala efectiva de deslocalización electrónica (ESED), el índice total del modelo del oscilador armónico de aromaticidad (HOMA) y la energía de resonancia de Dewar (DRE). Los resultados revelaron un incremento de CN/C2 de acuerdo con el carácter aromático y la energía de resonancia de las moléculas según la serie C10H8 > C14H10 > C16H10. Esta circunstancia sugiere que un incremento en la deslocalización electrónica da lugar a una menor generación de fragmentos C2. Al mismo tiempo, estas deducciones esbozan que los dímeros de carbono proceden, en mayor medida, de la fragmentación parcial de las moléculas, mientras que los radicales CN son generados por recombinación atómica. Fuerzas de los enlaces de carbono: efectos inductivos. Para evaluar la influencia de los efectos inductivos en las emisiones ópticas se analizaron dos polímeros: polietileno y politetrafluoroetileno, con fórmulas químicas (CH4)n y (CF4)n, respectivamente. Ambos polímeros presentan una similar longitud y topología de los enlaces C=C. Sin embargo, la elevada electronegatividad de los átomos de F genera una fuerte polarización que debilita los enlaces σ. De este modo, la energía del enlace C=C en el fluoroetileno es aproximadamente la mitad de la que posee el enlace C-F, estando favorecida la ruptura de dicho doble enlace. Contrariamente, en el etileno el enlace C=C es más fuerte que el C-H. En consecuencia, ambos monómeros presentan distintos patrones de ablación, produciéndose la ruptura del fluoroetileno por el enlace C=C y en el etileno por el enlace C-H. Estas diferencias se reflejan en la liberación de fragmentos de C2 y, por Resumen 7 tanto, en el evaluador estadístico CN/C2, siendo mucho menor para el polietileno (2,2) en comparación con el valor para el politetrafluoroetileno (12,2). En definitiva, el efecto inductivo causado por el heteroátomo causa una mayor atomización de las moléculas de politetrafluoroetileno. Efectos inducidos por los grupos funcionales. La influencia de distintos grupos funcionales en las respuestas de emisión óptica se evaluó a partir de una serie de nitrobencenos para-sustituidos (R−C6H4−NO2, R = −CH3, −NH2, −OH, −NO2). Los efectos resonantes generados en las moléculas como consecuencia de la presencia de estos sustituyentes conllevan la transmisión de densidad electrónica a través del sistema π del anillo bencénico. Esta transferencia de carga depende del efecto polar del grupo sustituyente, es decir, de su tendencia a retirar electrones del anillo debido a su electronegatividad. Para este conjunto de compuestos el evaluador estadístico CN/C2 se comparó con los valores de la electronegatividad relativa inherente de Pauling de cada uno de los grupos funcionales, siguiendo ambos una tendencia creciente de acuerdo con la secuencia −CH3 < −NH2 < −OH < −NO2. De este modo, la menor electronegatividad del grupo metilo (grupo electrón-donante) en el p-nitrotolueno se ve manifestada por un menor valor para CN/C2. En dicha molécula la transferencia de electrones estabiliza una configuración en la que se alternan enlaces simples y dobles; una circunstancia que favorece la liberación de fragmentos C2 frente a la atomización. En los casos de la p-nitroanilina y el p-nitrofenol los pares de electrones desapareados de los grupos −NH y −OH se deslocalizan en el anillo, favoreciendo que se mantenga el sistema electrónico π y, en consecuencia, la atomización de estos compuestos se ve beneficiada frente a su fragmentación. Por último, en el dinitrobenceno el fuerte poder electrón-atrayente de los 2 grupos nitro en posición paradebilita su estructura, potenciándose mucho más el proceso de atomización y dando lugar a una mayor generación de radicales CN. Resumen 8 Por otra parte, se analizaron los regioisómeros orto-, metay paradel dinitrobenceno. Aquí se observó como las emisiones ópticas mostraban cierta sensibilidad a la variación en la posición de los grupos funcionales. Cabe destacar que el orto-dinitrobenceno manifestó un valor para CN/C2 significativamente menor que el de los otros dos isómeros. En dicho compuesto, las mayores repulsiones estéricas entre los grupos funcionales rigen una pérdida de conjugación del sistema π, que repercute en una mayor fragmentación parcial de la molécula, en detrimento de su atomización. Efectos de la longitud de onda de irradiación. Para evaluar el efecto de la longitud de onda de irradiación sobre las emisiones ópticas se analizaron los isómeros 2,3-, 2,4-, y 2,6dinitrotolueno empleando radiaciones de excitación de 266 nm (ultravioleta), 532 nm (visible) y 1064 nm (infrarroja). Aunque con el empleo de radiación visible las diferencias en los valores de CN/C2 para los distintos isómeros no fueron excesivas, se obtuvo un valor ligeramente superior en el caso del isómero 2,4-DNT. Esta circunstancia se atribuye a los menores efectos estéricos para esta disposición de los grupos funcionales y, por tanto, a la menor modificación en la conjugación del sistema π con respecto a los otros dos isómeros. Por el contrario, el empleo de radiación láser de 266 nm dejó al descubierto mayores diferencias entre las respuestas ópticas debido a que las longitudes máximas de absorción de luz de estos compuestos se sitúan en esa región del espectro electromagnético. Un valor de CN/C2 notablemente mayor se manifestó para el isómero 2,4-DNT dado que su máximo de absorción óptica se sitúa más próximo a 266 nm y, en consecuencia, la molécula sufre una mayor atomización. La distinta distribución electrónica en las moléculas también condiciona sus propiedades termodinámicas, como por ejemplo las correspondientes energías relativas de estabilización asociadas a sus estructuras optimizadas. Con el empleo de pulsos láser de 1064 nm se observó un decrecimiento para CN/C2 de acuerdo con la serie 2,4-DNT < 2,6-DNT < 2,3-DNT que, a su vez, conviene con un incremento de la energía relativa de estabilización. Valores bajos de esta energía, Resumen 9 que indican mayor movilidad de los electrones π, van ligados a una mayor atomización de la molécula. En definitiva, aunque los procesos dominantes derivados de la interacción láser–muestra puedan depender de la longitud de onda de irradiación, estos resultados ponen de manifiesto que la distribución electrónica en la molécula juega un rol crucial en los patrones de emisión resultantes.  Rutas de formación de radicales hidrogenados en plasmas de sólidos moleculares deuterados inducidos con pulsos láser de femtosegundos Desde que la espectroscopia de plasmas inducidos por láser comenzara a aplicarse al análisis de compuestos orgánicos, una gran parte de las investigaciones se ha centrado en el estudio de las rutas de formación de los radicales CN y C2. Sin embargo, las vías de generación de otras especies moleculares diatómicas también presentes en los plasmas de compuestos orgánicos apenas han sido exploradas. Uno de los capítulos incluidos en la presente memoria de Tesis Doctoral evalúa los posibles orígenes de los radicales hidrogenados NH, CH y OH en plasmas inducidos mediante excitación con pulsos láseres ultracortos. Con este fin, junto con la espectroscopia de emisión óptica, se emplea como herramienta para el trazado de las rutas de reacción el marcado isotópico selectivo de compuestos orgánicos con átomos de deuterio. Este estudio se beneficia de los mayores desplazamientos isotópicos en las longitudes de onda de emisión experimentados por las bandas moleculares en comparación con los desplazamientos de líneas de emisión de átomos o iones. Mientras que la diferencia de masa entre isótopos tiene un efecto bastante pequeño en las transiciones electrónicas de los átomos, los niveles de energía rotacionales y vibracionales en las moléculas se ven afectados en mayor medida. Esto significa que con el uso de espectrómetros con un poder de resolución medio pueden monitorizarse simultáneamente las emisiones de isotopólogos. Resumen 16 De igual modo, las emisiones relativas de CH también fueron más intensas en fs-LIBS. El análisis exhaustivo de las respuestas ópticas expuso que la intensidad de la señal de CH parece ser una combinación de la abundancia de fragmentos C2 y la presencia de grupos OH en los compuestos, más allá de relacionarse directamente con la presencia de enlaces nativos C-H. Por tanto, parece razonable pensar que la reacción química radicales C2 y OH pueda ser la principal ruta responsable de la producción de CH. Además, se evaluaron y propusieron las posibles rutas para la generación de radicales NH y OH. De nuevo se computaron las emisiones relativas respecto al total de las emisiones de radicales hidrogenados (CH, NH y OH). Atendiendo a las señales ópticas de plasmas inducidos con pulsos de fs, se observaron emisiones de NH más elevadas en los compuestos con N y, especialmente, en aquellos con enlaces N-H. Las emisiones de NH en compuestos sin N fueron vinculadas a la reacción entre el N atmosférico y el H liberado de los sólidos ablacionados. Estas emisiones resultaron bastante menos intensas, no sólo por la ausencia de N en las moléculas, sino principalmente por la mayor afinidad del N atmosférico por los átomos de C liberados para formar CN. De forma análoga, también se observó una relación directa entre la presencia de grupos hidroxilo en los sólidos moleculares y la intensidad de las emisiones de OH (por ejemplo, para el pentaeritritol y para el ácido L-ascórbico). Por el contrario, al igual que ocurre con los otros radicales diatómicos, el mayor grado de atomización en ns-LIBS hace que las posibles relaciones entre las emisiones ópticas y la estructura molecular de los compuestos resulte mucho menos evidente. Los resultados demuestran que las emisiones ópticas derivadas de la ablación láser con pulsos ultracortos reflejan mejor la estructura de los sólidos moleculares que las emanadas de la excitación con pulsos de ns. Esta mejor correspondencia se debe a un efecto combinado de dos fenómenos como son la menor atomización de los compuestos y la baja interferencia de la atmósfera circundante. Por lo tanto, fs-LIBS puede considerarse como una herramienta mucho más selectiva para el análisis de sólidos moleculares, en comparación con ns-LIBS. Resumen 17 2. Aplicaciones de LIBS al análisis y caracterización de sólidos moleculares  Discriminación de materiales nanométricos de carbono. Caracterización óptica de grafeno Desde el descubrimiento del grafeno en 2004, la comunidad científica ha realizado grandes esfuerzos en la mejora de los procesos para su fabricación y en la implantación de métodos para su caracterización físico-química. El grafeno, que es descrito como una capa grafítica monocristalina estable en condiciones ambiente, posee excepcionales propiedades: rigidez mecánica, dureza y elasticidad, una gran área superficial, alta conductividad térmica y eléctrica, una baja absorción óptica y una gran capacidad para su funcionalización química. Todas estas características lo convierten en un material muy atractivo para la fabricación de nuevos dispositivos así como para mejorar las cualidades de infinidad de productos fotónicos y electrónicos. Es por ello que el control de la calidad del grafeno, como componente esencial de dichos dispositivos, es un aspecto fundamental a controlar que precisa de técnicas analíticas capaces de diferenciar entre los productos de partida y los productos finales. El uso de LIBS para esta aplicación se sustenta en los recientes progresos experimentados por la técnica en la caracterización de compuestos de carbono con similar composición química. En este trabajo las emisiones ópticas de plasmas de 3 compuestos de carbono puro (grafito, grafeno multicapa y grafeno de pocas capas) y de un material de partida en la síntesis de grafeno como es el óxido de grafeno han sido estudiadas en profundidad. Las muestras empleadas, originariamente en forma de polvo o copos, fueron prensadas y analizadas mediante LIBS empelando una configuración experimental en modo local y una excitación con pulsos láser de ns. En cuanto a los resultados experimentales, las respuestas ópticas de todos los compuestos manifestaron líneas atómicas de emisión de C, H, O y N, así como señales asociadas a los sistemas moleculares CN y C2. Las emisiones de H, O, N y CN en los plasmas de materiales con una composición del 100% de átomos de carbono son consecuencia de la interacción entre los plasmas Resumen 18 y su atmósfera circundante. En el caso del óxido de grafeno, se observaron altas emisiones de H que fueron atribuidas a la presencia de algunos grupos funcionales, tales como –OH y –COOH, en dicho material. Además, dichos grupos funcionales también justifican las emisiones intermedias de O exhibidas por este material, que difieren de las bajas emisiones para el resto de compuestos debidas únicamente a la contribución del aire. Por otra parte, la presencia de los grupos funcionales también repercute en otras emisiones remitidas por los plasmas de óxido de grafeno. La inclusión de grupos funcionales genera una mezcla de hibridaciones sp2y sp3de los átomos de C, disminuyendo el número enlaces dobles de C y por tanto la generación de dímeros de carbono por fragmentación directa. Como consecuencia de este cambio de estructura del plano base del material se detectan intensidades despreciables para las emisiones de C2. Una de las principales diferencias observadas en el proceso de ablación láser de los materiales en cuestión fue los distintos valores para los umbrales de formación de sus plasmas. Los resultados revelaron un incremento de dichos umbrales en función del número de capas de grafeno que conforman el material. Este hecho puede justificarse de acuerdo con los valores de difusividad térmica y área específica de los materiales, que también aumentan con el número de capas grafíticas. Por tanto, es posible concluir que el número de capas de grafeno que conforman cada material, así como su distinta disposición, afecta directamente al proceso de ablación. También se llevó a cabo una evaluación del efecto de la fluencia láser (en el rango comprendido entre 10 J·cm-2 y 250 J·cm-2) en los patrones de emisión de estas muestras empleando parámetros temporales fijos para la adquisición de la señal LIBS (un tiempo de retraso de 1,5 μs y un tiempo de integración de 0,5 μs). Más allá de un incremento de la intensidad de la señal al aumentar la dosis de energía dispensada como consecuencia de su proporcionalidad con la tasa de ablación, las respuestas ópticas de los distintos compuestos no manifestaron diferencias significativas. Debe señalarse que las intensidades de algunas señales de emisión para el óxido de Resumen 19 grafeno se apartaban de las del resto de compuestos, por las diferencias estequiométricas anteriormente comentadas. Por el contrario, la monitorización del progreso temporal de las señales moleculares permitió identificar distintas pautas de emisión para los 4 materiales analizados. Entre ellas, la más destacable es la significativa diferencia en la velocidad de extinción de las emisiones moleculares, principalmente de C2, cuando los plasmas son inducidos con una alta fluencia láser (aprox. 250 J·cm-2). Mientras que en el óxido de grafeno estas emisiones experimentan una tendencia decreciente desde los primeros instantes posteriores a la formación del plasma, las intensidades de señal para el resto de materiales de carbono puro manifiestan un decaimiento mucho más sostenido. En estos materiales, el decaimiento de las señales moleculares parece ir acorde con el tipo de fragmentos de carbono (Cn) liberados durante su ablación. De este modo el grafito y el grafeno multicapa revelan un decaimiento más rápido de sus señales moleculares en comparación con el grafeno de pocas capas. Esta circunstancia se justifica por sus menores umbrales de ablación, que facilitan una mayor atomización y por tanto una liberación de fragmentos de carbono de menor tamaño que son disociados más rápidamente. De hecho, las diferencias entre estos materiales en la velocidad de extinción de la señal de C2 son más acentuadas en comparación con las reportadas a partir del comportamiento temporal del radical CN. Esta conducta temporal no sólo permite corroborar que las emisiones de C2 están directamente relacionadas con el grado de fragmentación de las estructuras grafíticas, sino que además puede actuar como un testigo marcador en la discriminación de materiales grafíticos con distinto número de capas. En resumen, los resultados han demostrado que los umbrales de formación de plasma son distintos como consecuencia de diferentes propiedades de las estructuras de carbono, tales como su número de capas y la presencia de grupos funcionales. A su vez, estos umbrales de ablación de los materiales de carbono son los que definen el patrón de ruptura de dichas estructuras que, Resumen 20 posteriormente, repercuten directamente en el comportamiento temporal de las emisiones moleculares, permitiendo así la diferenciación de estos materiales.  Nuevas estrategias quimiométricas aplicadas al reconocimiento de residuos explosivos El desarrollo de la tecnología LIBS para la detección de residuos explosivos en el escenario de una posible amenaza terrorista viene motivado por la necesidad de disponer de una herramienta analítica que proporcione sensibilidad y fiabilidad suficientes en la interrogación rápida de objetos sospechosos. Sin duda, la capacidad para detectar e identificar la presencia de sustancias explosivas es vital para poder prevenir cualquier acto terrorista. En este contexto, se incluyen en la presente Tesis Doctoral tres trabajos de investigación en los que las respuestas espectrales proporcionadas por la tecnología LIBS se han sometido a un tratamiento con potentes métodos quimiométricos para el reconocimiento de residuos explosivos. Aunque los trabajos han sido realizados a escala de laboratorio, pretenden abordar situaciones de riesgo en escenarios reales. Por ello, más allá del análisis de estos materiales a granel, se afronta la problemática de la detección de restos explosivos localizados a nivel de trazas sobre superficies de distinta naturaleza como consecuencia de su manipulación. Los tres estudios experimentales desarrollados comparten la metodología analítica y la herramienta quimiométrica, el aprendizaje de máquinas supervisado (supervised machine learning). Se pretende de esta forma obtener el máximo rendimiento de la limitada información proporcionada por la técnica LIBS, mejorando la sensibilidad y selectividad del método en el reconocimiento de residuos explosivos. La metodología consiste en el diseño y entrenamiento de un modelo de clasificación a partir de un conjunto de datos previamente conocidos (espectros LIBS de residuos de naturaleza conocida) para la posterior identificación y asignación de datos desconocidos (espectro LIBS de un residuo desconocido) a una determinada clase (residuo explosivo o residuo inocuo). Resumen 21 El equipo LIBS experimental empleado en estos tres trabajos de investigación fue el mismo, manteniéndose incluso las mismas condiciones operacionales (energía del pulso láser, condiciones focales, parámetros temporales de adquisición de la señal, etc.). Cabe destacar el uso de un láser pulsado de nanosegundos de Nd:YAG y un espectrómetro Czerny-Turner miniaturizado de cuatro canales. A) Detección de residuos explosivos de naturaleza orgánica sobre soportes de aluminio En este primer trabajo se abordó la situación más sencilla que puede encontrarse en el análisis LIBS de residuos en una superficie, esto es, cuando la composición elemental del residuo difiere de la composición de la superficie. Se realizó en primer lugar un estudio en profundidad de las emisiones ópticas de los plasmas de 9 compuestos orgánicos depositados como residuo sobre superficies de aluminio. Como residuos explosivos se consideró el conjunto formado por dinitrotolueno (DNT), trinitrotolueno (TNT), ciclotrimetilentrinitramina o hexógeno (RDX) y pentaeritritol tetranitrato (PETN). Como residuos inocuos también de naturaleza orgánica, cuya composición elemental y patrones de emisión son muy similares a los de las muestras explosivas, y que pueden actuar como potenciales confusantes se optó por materiales como la mantequilla, el aceite de oliva, el aceite de motor, la gasolina y la crema de manos. Una vez obtenido un volumen considerable de datos espectroscópicos de todos estos residuos localizados sobre superficies de aluminio, se procedió al tratamiento quimiométrico de la información LIBS. El método empleado se basó en la proyección de las intensidades de emisión de las señales espectrales más relevantes –C, CN, C2, H, N y O– en distintos subespacios bidimensionales con el ánimo de identificar las características espectrales que proporcionaban la mejor distinción entre los dos tipos de residuos. Para poder examinar de una forma rápida todos los gráficos de dispersión construidos, se calculó un parámetro numérico denominado Región Discriminante Residual –RDR– definido por la siguiente ecuación: Resumen 22     Este parámetro se fundamenta en el grado de solapamiento que existe entre el área ocupada por la nube de dispersión proyectada por los residuos explosivos –definida por la envolvente convexa de todos los datos que la constituyen– y el área ocupada por la nube de dispersión generada por los residuos inocuos. El valor que puede tomar dicho parámetro oscila entre 0 y 2. Así, un valor de 2 para el RDR indica que las dos nubes de datos se encuentran totalmente solapadas, de tal modo que las dos características espectrales involucradas en dicho subespacio bidimensional no permiten diferenciar entre los dos tipos de residuos. Por el contrario, un valor para RDR de 0, revela la mejor diferenciación entre residuos explosivos e inocuos. Una vez identificado el gráfico más eficaz para la distinción de residuos de acuerdo con su naturaleza, se procedió a diseñar, construir y entrenar distintos algoritmos –clasificadoresbasados en distintos principios estadísticos y matemáticos. Entre los modelos utilizados para hacer frente a la clasificación destacan los basados en densidades normales y en la regla del vecino más cercano, los basados en la estimación por núcleos o ventanas de Parzen, y los fundamentados en modelos de mezcla de funciones gaussianas y en las máquinas de vectores soporte. En el caso particular del reconocimiento de residuos depositados sobre aluminio se identificaron 3 diagramas que proporcionaron una adecuada separación entre clases de residuos. Se trataba de lo gráficos que involucraban la emisión atómica de C, N u O frente a la señal de H. Cabe señalar aquí que las emisiones de H, N y O pueden tener origen nativo o bien venir derivadas del proceso de ionización secundaria de aire. Esta circunstancia dificulta poder establecer si los residuos se diferencian directamente por su distinta estequiometria o bien por su estado físico y el distinto comportamiento de sus correspondientes plasmas con la atmósfera que los rodea. Resumen 23 Pese a que las señales moleculares CN y C2 han servido en estudios previos como activos útiles para la detección de explosivos, en este caso no mostraron tener utilidad en la clasificación de los residuos. En particular, las emisiones de C2 resultaban inservibles ya que en el rango espectral donde se manifiestan (desde 450 nm hasta 550 nm) también se exhibe el sistema de bandas de emisión del AlO, generado por la reacción del Al procedente de la muestra con el oxígeno ambiental, produciéndose un solapamiento entre ellas. Atendiendo a las estadísticas de clasificación reportadas por los diferentes clasificadores diseñados y entrenados en los 3 gráficos anteriormente mencionados, los resultados determinaron que el clasificador construido en base a la estimación por ventanas de Parzen sobre el gráfico que proyectaba las intensidades de emisión de las señales de O y H resultó ser el más adecuado, proporcionando las tasas de falsos positivos y de falsos negativos más bajas, inferiores al 5 %. Finalmente, se procedió a evaluar la sensibilidad de la metodología en el reconocimiento de estos residuos explosivos. En este caso particular, más allá de establecer la sensibilidad como el límite de detección (LOD) de acuerdo a la definición 3 de la IUPAC –concentración de un elemento capaz de proporcionar una señal de emisión neta de al menos 3 veces la desviación estándar del fondo espectral– se trató de identificar cual era la mínima intensidad de señal a partir de la cual se podía acometer la distinción entre los dos tipos de residuos. Para ello se propuso un parámetro alternativo basado en el establecimiento de unas coordenadas espaciales (x, y) de las variables involucradas en los gráficos de dispersión como límites de corte para la identificación correcta de los residuos. De este modo, las respuestas espectrales de aquellos residuos interrogados que proporcionen unas valores de intensidad para las señales O y H superiores a los valores (x, y) establecidos permiten su correcta identificación. En caso contrario, esto es, si los valores de intensidad se encuentran por debajo existen dudas acerca de la fiabilidad en el reconocimiento. Para una mejor interpretación de los resultados las coordenadas se normalizaron a la máxima respuesta del detector. En particular para el gráfico O vs H  , se obtuvieron unos valores para las Resumen 24 coordenadas muy próximos a (0, 0). Dichos valores mínimos para el límite de corte indicaban prácticamente una sensibilidad máxima en la distinción de ambos tipos de residuos. En definitiva, el clasificador diseñado puede asignar a la clase correspondiente con suficiente exactitud cualquier residuo cuya respuesta espectral ofrezca una mínima intensidad para las señales de oxígeno e hidrógeno. Los hallazgos de esta investigación ponen de relieve la capacidad de LIBS en conjunción con el uso de máquinas de aprendizaje supervisado para el reconocimiento de residuos explosivos bajo condiciones experimentales definidas, esto es, localizados sobre soportes de aluminio. Por ello, el siguiente paso relacionado con esta línea de investigación consistió en evaluar la validez de esta misma estrategia en distintas situaciones –idénticos residuos pero emplazados en otro tipo de superficies–, tal y como se describe a continuación. B) Detección de residuos explosivos de naturaleza orgánica depositados en superficies poliméricas La identificación de trazas de explosivos orgánicos depositados sobre superficies poliméricas mediante LIBS es un problema mucho más complejo que el ejemplo descrito con anterioridad. En esta situación, el sustrato tiene la misma composición química que el residuo y ambos contribuyen a los patrones finales de emisión con idénticas señales (C, H, O, N, CN y C2). Desafortunadamente, éste era un importante escollo a superar para el uso de esta tecnología en futuras aplicaciones reales. No obstante, el uso de métodos quimiométricos para el tratamiento de datos multivariables ha contribuido a superar este obstáculo. Mediante su uso es posible desvelar las mínimas diferencias existentes entre los espectros de emisión óptica de materiales explosivos e inocuos, pese a la dificultad añadida por la contribución de emisiones relativas al sustrato. En este trabajo se consideraron los mismos nueve residuos empleados en el estudio realizado con las láminas de aluminio –DNT, TNT, RDX y PETN, como explosivos, y mantequilla, Resumen 25 aceite de oliva, aceite de motor, gasolina y crema de manos, como confusantes. Estos residuos fueron depositados sobre tres soportes poliméricos –teflón, nylon y polietileno– por lo que resultaron un total de 27 muestras binarias. Para la selección de las variables espectrales que proporcionaban la mejor distinción de los residuos según su clase (es decir, el/los mejor/es diagrama/s de dispersión 2D) se computó el correspondiente valor del parámetro RDR descrito anteriormente. Los resultados revelaron que, independientemente del soporte sobre el que se localizara el residuo, los diagramas que involucraban CN y C2 como variables discriminatorias eran los más adecuados para la distinción entre tipos de residuos por proporcionar los valores más bajos de RDR. De acuerdo con la dispersión de las nubes de datos en estos gráficos, los residuos explosivos manifestaban mayores emisiones de CN y menores señales de C2 en comparación con los residuos inocuos. Para tratar de acrecentar la capacidad de discriminación, se evaluó el rendimiento de gráficos tridimensionales, añadiendo una tercera dimensión a los anteriores diagramas de dispersión. En estos diagramas 3D, el cálculo del parámetro RDR involucraba los valores del volumen ocupado por las nubes de dispersión en lugar de su área. Dichos volúmenes fueron también definidos a partir de las correspondientes envolventes convexas de las nubes de datos. El análisis de los resultados desveló menores valores de RDR para los diagramas de dispersión 3D en comparación con los reportados por los gráficos 2D, justificando así el uso de los primeros al proporcionar una mejor separación entre residuos explosivos e inocuos, en particular diagramas 3D formados por combinaciones de las intensidades de emisión de CN, C2, H y O. Cabe resaltar que las emisiones ópticas de los plasmas inducidos a partir de los diferentes residuos mostraron un comportamiento similar para los distintos soportes. Por ello se consideró la construcción de un clasificador global a partir del mismo diagrama de dispersión 3D para la detección de residuos con independencia del sustrato (es decir, válido para residuos depositados en teflón, nylon o polietileno). Si bien al agrupar las tres situaciones el gráfico construido reportaba un valor de RDR ligeramente Objectives 32 Therefore, the applied research that will be conducted during the development of the present Doctoral Thesis is aimed at shedding light on the chemistry of laser-induced plasmas of molecular solids as well as the design and implementation of new strategies to exploit molecular excited assets for sensing and characterization of organic materials in real-world applications. So, quite a few particular goals shall be pursued to cope with these broad objectives. Firstly, to complement the already existing knowledge on chemistry of organic plasmas, the origin and the potential routes leading to the generation of less investigated molecular species must be unveiled. To date, research has shown a particular concern on disclosing the origins of CN and C2 radicals, which are normally the most representative molecular emissions in plasmas of organics. Thus, for moving forward, the disclosure of origins and formation pathways of hydrogenated radicals (OH, NH and CH), which have been scarcely explored in laser-induced plasmas, will be pursued. Since the vast majority of LIBS applications are performed in air at atmospheric pressure, our interests will be focused in the applicability of LIBS to atmospheric conditions to meet general analytical challenges. Another important aspect to be considered with regard to selectivity and specificity of LIBS from organic plasmas are the likely relationships between the final optical emission patterns and the chemical structure of organics. The knowledge and identification of the spectral features which are truly representative of the structure of the compounds is of prime importance to exploit the use of LIBS as identification tool of organic samples. Although several studies have dealt with the analysis of organics by LIBS, there are not conclusive findings concerning the influence of the molecular structures of organics on their optical emission patterns. In this connection, it will be the subject of a systematic research within the development of the present Doctoral Thesis the diagnostics of laserinduced plasmas of a wide set of organic compounds with structural and compositional particularities, in order to reveal possible relations between the different molecular emissions –CN, C2, OH, NH and CH– and the chemical structure of organic compounds. Objectives 33 It is well known that operational parameters in LIBS, such as the excitation wavelength, the laser pulse length and the laser pulse energy, significantly affect the ablation process, the generation of the subsequent plasma and, hence, the optical emission signals displayed.4 Therefore, another goal that is raised for this Doctoral Thesis is the judgment of the effects of operational conditions on the emission spectroscopy of molecular solids. Since LIBS applications in many disciplines use nanosecond lasers for driving the ablation sampling process, our interests will be focused on scrutinizing the influence of different radiation wavelengths to the molecular emissions. At the same time, since the use of femtosecond lasers is now much more widespread, the potential effects on sensing and characterizing molecular solids by LIBS attributed to the laser pulse duration shall be evaluated. In addition to all these fundamental studies on plasmas of organic compounds, some investigations in the present Doctoral Thesis must be routed to the exploitation of this background knowledge towards different LIBS applications. Currently, LIBS technology is implemented in a wide range of real-world applications that benefit from its attractive features. Notwithstanding this, research in the present Doctoral Thesis will attempt to address the most cutting edge applications involving organic materials. On the one hand, the functionality of LIBS to the scrutiny of graphene –a material with exceptional properties and multiple applications that has been set to revolutionize the material world– and other nano-sized carbon materials will be evaluated. This investigation is motivated by the growing interest in the production of graphene and the need of analytical tools to control the quality of the materials involved in its manufacturing process.5 The optical characterization of graphene and its differentiation from other carbon-based materials with similar chemical composition is the first step for a future implementation of LIBS in this application field. Additionally, another field of study that has generated much interest in the use of LIBS is the detection of energetic materials. Certainly, investigations aimed at developing LIBS for security Objectives 34 applications have been encouraged by the increasing number of terrorist attacks during the last decades.6 The detection of explosives by LIBS provides the possibility of real-time and in-situ analysis of potential threats also minimizing the risk for the operator when using remote or standoff sensors. For this reason, the application of LIBS in the area of homeland security and in potential military field operations must be addressed. While some research has proved the potential use of LIBS for the detection of explosive materials, some pre-existing limitations of the technique have been also exposed. The main hindrance when applying LIBS for the recognition of organic explosives (typical chemical formula CCHHNNOO) lies in a biased loss of the molecular information of the target after laser ablation. Furthermore, emission signals that draw the optical fingerprint of explosive substances are not only scarce but also identical to those of innocuous organic materials. Hence, the efficiency of LIBS as analytical tool for the detection of explosive materials needs to be increased. To attain this, the design and implementation of advanced chemometric strategies to extract as much as possible information from this type of LIBS data shall be conducted. The main purpose of this research is the development of a fast, sensitive and reliable methodology providing the most accurate distinction between explosives and harmless compounds. Since the end-use of this methodology is intended to application in a real scenario of terrorist threat, the performance of the approach shall be tested to label minimal contaminations of explosive materials during recognition of traces deposited on solid surfaces. Objectives 35 REFERENCES 1 L. Radziemski, D. Cremers, A brief history of laser-induced breakdown spectroscopy: From the concept of atoms to LIBS 2012, Spectrochim. Acta Part B 87 (2013) 3–10. 2 F. J. Fortes, J. Moros, P. Lucena, L. M. Cabalín, J. J. Laserna, Laser-induced breakdown spectroscopy, Anal. Chem. 85 (2013) 640–669. 3 J. M. Anzano, I. B. Gornushkin, B. W. Smith, J. D. Winefordner, Laser-induced plasma spectroscopy for plastic identification, Polym. Eng. Sci. 40 (2000) 2423–2429. 4 A. Bogaerts, Z. Chen, Effect of laser parameters on laser ablation and laser-induced plasma formation: A numerical modeling investigation, Spectrochim. Acta Part B 60 (2005) 1280–1307. 5 W. Choi, I. Lahiri, R. Seelaboyina, Y. S. Kang, Synthesis of graphene and its applications: a review, Crit. Rev. Solid State 35 (2010) 52–71. 6 J. L. Gottfried, F. C. DeLucia Jr., C. A. Munson, A. W. Miziolek, Laser-induced breakdown spectroscopy for detection of explosives residues: a review of recent advances, challenges, and future prospects, Anal. Bioanal. Chem. 395 (2009) 283–300. 37 INTRODUCTION 1. An overview of laser-induced breakdown spectroscopy The laser-induced breakdown spectroscopy (LIBS) is a technique based on the detection and analysis of the spectrally resolved optical emissions of atoms, ions and small molecules present in plasmas generated by the pulsed laser ablation of samples.1 This technique exhibits several distinctive capabilities that have contributed to its fast expansion in numerous areas of applied chemistry and physics. Notable among them are, the potential to examine raw –without any previous preparation– samples, its flexibility to analyze any state of the matter –solids, liquids, and gases–, the potential to perform a simultaneous multielemental detection, its suitability for real-time monitoring of the elemental composition of samples, the potential for fast analysis, its adaptability upon field deployable instruments, and its handiness to operate at remote distances.2 The origin of LIBS dates from 1962, a couple of years after the pulsed ruby laser was invented by Maiman,3 when the first publication revealing laser-induced plasmas as a spectral source was presented by Brech and Cross.4 Shortly after, the technique was first used in the field of surface analysis5 and to examine metallic samples by pulsed Q-switched ruby laser as unique excitation source.6 Furthermore, during the sixties, optically induced breakdown when monopulse Introduction 38 ruby laser radiation was focused in air7 and in water8 was observed. Unfortunately, despite its promising start, the expansion of LIBS as analytical technique was slowed down over many years because of its limitations as quantitative method. It was not until the 1980s that, with the introduction of time-integrated and time-resolved spectroscopic measurements by Loore and Radziemski,9,10 the scientific community again focused attention and reactivated investigations in LIBS. Since then, LIBS has not stopped growing, not only in obtaining more knowledge about its performance but also addressing a lot of real analytical challenges. A substantial part of the research interests has been focused on improving the sensitivity and selectivity of the technique. To this end, the application of different experimental approaches for the excitation, as the use of a double-pulse (DP-LIBS),11 the handling of a scheme based on the tuning of the laser radiation of a pulse to the wavelength of a resonant atomic transition of the matrix atoms of the plasma (resonance-enhanced LIBS, RELIBS),12-14 the delivering of multi-pulses (MP-LIBS),15 and the operation with ultrashort laser pulses16 has been considered. On the other hand, improvements in LIBS instrumentation have allowed the development of field-deployable sensors which may be used in a wide variety of applications.17 In this connection, standoff sensors have been constructed with the goal of analyzing distant targets in environments where physical access is not possible or may present a risk to the operator.18 In the last years, the design of remote instruments based on a fiber optic cable to operate in aquatic media for analyzing submerged archaeological findings in sea is also a clear indicative fact of the progress and the potential of LIBS.19 Lastly, the effective deployment and implementation as crowning of the technology has been evidenced with its flexibility and adaptability to a new environment as the surface of Mars for geochemistry exploration in this planet.20 Nowadays, LIBS has become a very trendy analytical technique expanding to numerous application fields such as industry,21 homeland security,22,23 geology,24,25 biomedicine,26 cultural heritage,27 and planetary exploration.28,29 However, the current research in LIBS is not only focused Introduction 39 towards the starting-up and the development of laser-based analytical applications30 but also on the knowing, the understanding, and the unveiling of the fundamental aspects of the ablation physical process and the chemistry of the plasma.31 Particularly, a broad diversity of chemical reactions may occur in plasmas generated from the laser ablation of molecular solids. In this regard, the elucidation of the routes leading to the generation of diatomic radicals, and the emitting key species involved, are being, in recent years, subjects of interest for several research groups.32-39 This is because the knowledge on origins of molecular emissions and connection of these with the molecular structure of the ablated compounds is a key factor to implement LIBS as characterization tool of molecular solids. Introduction 40 2. Fundamentals of LIBS Laser-induced plasmas are non-stationary bodies whose stages of generation, evolution, and extinction are framed within a defined temporal scale. Figure 1 illustrates a simple sketch of the progress of plasmas generated with nanosecond (ns) laser pulses under air atmosphere at atmospheric pressure. It all begins with the delivering of a pulsed laser beam that is tightly focused onto the target surface. As seen, during the first instants of the laser-surface interaction (only a few nanoseconds after its shock), a fraction of the delivered radiative energy enters the target. This absorbed radiation produces a high energy density at the focal point that leads to a heating and a vaporization of a very thin and slightly deep layer from the surface, followed by a mass removal process called ablation.40 Several processes such as phase transitions and material ejection occur. If the energy provided by the incoming laser radiation is sufficiently high, heated material in the vicinity of the surface induces stress waves of sufficient magnitude to plastically deform it. Consequently, a thermal expansion at the surface occurs, and a weak plasma that propagates away from the target surface is generated; the plasma ignition has begun. Then, the arrival of the tailing part of the laser pulse contributes to various interaction-based phenomena such as electron capture, collisional excitation and ionization, recombination processes and fragmentation of higher clusters, occurring inside the plume.41 Consequently, due to the absorption of that remaining energy, a plasma plume with elevated both temperature and electron density, and sown with electrons, ions, atoms as well as small molecules and clusters is created. After the end of the laser pulse, the plasma plume continues its expansion into the surrounding atmosphere and interacts with it. A shock wave, which propagates along the direction of plasma expansion, is also created. As the plasma evolves its temperature decreases and the excited species relax producing spontaneous emission of radiation. The emitted light is spectrally and temporally resolved into a spectrograph providing analytical information of the constituent elements Introduction 41 of the ablated sample. Finally, the plasma condenses itself and disappears in the order of few milliseconds after the laser impact.17,42,43 Figure 1. Representative time-scale for the evolution sequence of a plasma induced by a ns laser pulse. A comprehensive description of all processes following the laser-matter interaction will be discussed in the next subsections. It must be noted that mechanisms will be described for the general case of ns-LIBS. The characteristics of ultrafast laser ablation will be discussed as a particular case since the time-scale for the ablation process and the plasma evolution differ from the ns regime. 2.1. Ablation process In the context of spectroscopy, ablation is the process of mass removal following the absorption of laser radiation by a solid material. 40 Several mechanisms such as thermal diffusion, melting, and intense evaporation of the sample are involved in laser ablation. The process of mass ejection produces a crater on the sample surface. This crater is usually circular; however, its shape depends on the laser beam profile (commonly Gaussian or flat-top) and the focusing optics. In addition to the imprint of the laser impact other features can be distinguished on the crater Introduction 48 2.4. Decay and emission of the plasma Shortly after the end of the laser pulse the plasma possess an extremely high temperature and an elevated electron density. These conditions give rise to intense continuum emission which dominates the spectral response in the first instance. Hence, conventional spectral measurements are registered between some nanoseconds and few microseconds elapsed from the incidence of the laser pulse onto the target. As the plasma evolves, it cools off, the temperature and electron density decrease sufficiently, and ionic and atomic emission lines begin to be resolved within the spectral signal. While ionic lines are normally observed along just some hundreds of nanoseconds, the atomic signals have a larger life time and their emissions can be extended during a few microseconds. The prevalence of these emissions over time is closely tied to ablation conditions, since these influence the plasma properties and the density of its population species. At the later stages of the life time of the plasma, some emissions associated to small molecules, clusters and particles can also emerge. Some of the possible mechanisms leading to the formation of these species are molecular fragmentation, recombination processes, vapor condensation, liquid sample ejection, phase explosion (occurring when a material is heated beyond its limit of thermodynamic stability) and spallation (photomechanical fracture due to thermo-elastic stress wave).64-67 Typically, the light emitted from the expanding plasma is dispersed using a spectrograph and subsequently gathered and analyzed. The spectroscopic analysis of all the optical emissions from the excited species enables identification of the elemental composition of the ablated material. Furthermore, the gathered emission spectrum provides information regarding the temperature and particle density of the plasma itself, as discussed below. The temporal behavior of laser-induced emissions is exemplified by Figure 3. The timeresolved spectral signals of graphite plasmas generated in air at atmospheric pressure are depicted. Immediately following plasma ignition, that is a for a delay time below 0.5 µs, all emissions are entirely masked by the high continuum –spectrum not shown. As the plasma evolves, the emissions Introduction 49 associated to native C atoms from the sample, new formed species –CN–, as well as elements coming from the secondary ionization of the surrounding air (H, N and O) stick out above a broad nonspecific background. As seen, a gate delay of about 1.5 μs is required for almost entirely suppress such continuum emission. Later, as the plasma cools down, emissions of C, H, N and O become extinct while CN and C2 molecular band systems dominates the spectrum. Figure 3. Temporal behavior of optical features in emission spectra gathered from graphite plasmas induced in ambient air at atmospheric pressure. This figure has been adapted from similar one featured in the reference 42. 2.5. Parameters for plasma characterization As commented above, the plasma generated contains atoms and ions in different excited states, free electrons and radiation. Considering that properties of the plasma define emitted light, there is no question that such light constitutes a faithful reflection of the several species populating the plasma68 and the different physical-chemical processes occurring.69 In addition, the analysis of this light allows diagnostics of plasma properties through the computation of a pair of important plasma parameters from the LIBS spectra: the electron density (ne) and the temperature (T). Brief comments on these parameters are provided below. Introduction 50 Electron density Electron density, in general, specifies the thermodynamic equilibrium states of the plasma.70 Optical emission spectroscopy methods may be employed for its determination. One of the methods employed to calculate the electron density is based on the broadening of emission lines due to the Stark effect (collisions of the emitting atoms with electrons and ions). For non-hydrogenic ions, the Stark broadening is dominated by electron impact. Then, the full width at half maximum (FWHM) of lines, Δλ1/2, is given by the following expression: / 2󰇡 10󰇢  󰇟Eq.2.6󰇠 where ne is the electron number density (cm-3) and the coefficient W is the electron impact parameter. Plasma temperature Plasma temperature determines the strength of the different distribution functions describing the plasma state.71 Several methods based on the line emission of neutral atoms and ions as well as on the line-to-continuum ratio are nowadays available to determine the temperature of plasmas (T). The requirements of local thermodynamic equilibrium (LTE) and optically thin plasmas have to be satisfied to apply the greater number of methods. However, alternative methods, which allow calculating the plasma temperature without making any LTE assumption, have been also proposed by several authors.31 A general approach for temperature calculation is based on the use of Boltzmann plots72. For plasma in LTE, the Boltzmann's law relates the total density N(T) of a neutral atom or ion to the population of an excited level73,74, as expressed in equation 2.7:    4󰇛󰇜 󰇛󰇜  󰇟Eq.2.7󰇠 where Imn is the relative line intensity, λmn is the wavelength, Amn is the transition probability, gm is the statistical weight for the upper level; Em is the energy of the excited level, T is the temperature, kB and h are Boltzmann and Planck constants, repectively, and U(T) is the partition function. Introduction 51 Taking natural logarithm in the Boltzmann equation, such method allows calculating the excitation temperature by plotting ln(Imn/gmAmn) vs Em. Once known the slope of this plot, which is equal to (1/kBT), the plasma temperature (T) can be estimated. Although the range of plasma temperatures may broadly change depending on both the LIBS operational conditions and the sample nature, values from 6000 K to 15000 K are characteristically reached.75,76 Commonly, plasma temperature is calculated within a specific integration window of the optical signal and an average value of such temperature is provided. However, the plasma temperature undergoes temporal variations. That is, it normally drops as the gate delay increases from the plasma formation until its extinction. These decay trends may also differ as a function of the operational conditions and gas environments. On the other hand, the vibrational (Tvib) and rotational (Trot) temperatures of excited radicals may be determined from their molecular emission bands. Several procedures such as the use of Boltzmann plots75 and a fitting of experimental spectra74 to synthetic ones may be employed. As an example, LIFBASE free software program may be used to simulate spectra of the OH, CH and CN radicals, among other species.77 Local thermodynamic equilibrium Meaningful application of theoretical Boltzmann–Maxwell and Saha–Eggert expressions that relate fundamental plasma parameters and concentration of species essentially requires the existence of local thermodynamic equilibrium (LTE) in the plasma.78 Laser-induced plasmas are characterized by strong radiative losses and by high expansion velocities. As a result, these plasmas can be in strong non-equilibrium. When plasma is in thermodynamic equilibrium (TE), each process is balanced by its inverse process, which is known as the principle of detailed balancing. The system reaches a uniform steady state of thermodynamic equilibrium when there is an overall balance in the collision processes suffered by the particles. However, when the energy lost by radiative processes is smaller than that involved in collisions between species that govern transitions and chemical Introduction 52 reactions, the LTE is set.76,79 For satisfying this condition a sufficiently high electron density has to be reached. The most popular criterion usually invoked as a proof of the existence of LTE in the plasma is the McWhirter criterion.80 Even though this criterion is known to be a necessary but not a sufficient condition to insure LTE. The McWhirter criterion allows calculating the critical electron density (ne) for which the plasma is within LTE using the following equation: 1.610/󰇛∆󰇜  󰇟Eq.2.8󰇠 where T is the plasma temperature (K) and ΔE is the higher energy difference (eV) of the levels whose populations are given by LTE conditions. This approach assumes that the collisional rates are at least ten times the radiative rates within the plasma. The critical electron density for LIBS plasmas is usually estimated in the range of 1015–1016 cm−3. 2.6. Influence of operational parameters in laser-induced plasmas Without forgetting that nature and specific characteristics of laser-produced plasmas are closely tied to the type of material, the whole chain of possible events and processes occurring during plasma formation intimately depend on the radiative attributes, that is, laser pulse operating parameters like its energy (E), wavelength (λ), duration (τ) and shape. At the same time, the pressure and composition of the background atmosphere play an important role in the subsequent plasma evolution.1 From a theoretical point of view, the ideal for attributing the changes observed on the ablation process and the optical spectra would be to have reliability that they are associated to the variation of one parameter. Unfortunately, the complex interplay between different processes generating the LIBS plume and the dominance of some over others as a function of the conditions makes difficult to ascertain which effects are most responsible for the observations. The following sections briefly encompass a discussion on the main effects when varying operational parameters in LIBS. Introduction 53 2.6.1. Laser pulse energy Far beyond the absolute energy per pulse, two magnitudes are interchangeably used to describe the energetic regime for laser ablation: fluence (energy per unit area, J cm-2) and irradiance (power per unit area, GW cm-2). Both are referred to the total laser pulse energy deposited on the target surface per unit of area, therefore experimental diagnosis of these parameters requires a judicious estimation of the spot size over which the laser beam is focused. The disparity between the two terms is that irradiance contemplates the laser pulse duration whilst the fluence is a timeintegrated measurement of the applied energy. All processes occurring as a consequence of the laser-sample interactions such as heating, melting and vaporization of the sample are naturally influenced by the energy dosage provided.81 As a general trend, increasing laser fluence/irradiance above the ablation threshold of the sample (minimal fluence/irradiance leading to a mass removal process) entails larger both crater depth and ablated mass. However, when laser energy input on the sample considerably exceeds such threshold, very high density plasmas are generated. Consequently, this density may partially or entirely shield the sample from the laser radiation; not the full energy is transferred from the laser pulse to the original material. Hence, a saturation of the ablation rate can be detected.82,83 The energetic regime in which ablation occurs also affects the geometrical aspect (size and shape), the dynamical behavior as well as the inherent properties (temperature and species number density) of expanding plasmas. Generally, the size and the expansion velocity of plasmas proportionally increase with the laser energy. With regard to the shape, at low irradiance, the expansion of plasma, containing little ablated matter and having low internal energy, prevails in a radial direction over that in a longitudinal one. Thus, the plume core remains "attached" to the sample surface, having a more disk-like shape. In contrast, at high irradiance, the plume is capable of pushing the surrounding air far enough in front of itself in order to expand into a hemispherical Introduction 54 shape because of the higher content of ablated matter and its larger internal energy.84 In turn, an increase is detected for T and Ne with the raise of the irradiance.85 All these effects of laser irradiance also became evident on the surface morphology of the interrogated target. Disparity between plasmas produced from the same sample but at distinct irradiation regimes are considered to originate in the different mass removal mechanisms occurring, as identified through various kinds of structures such as ripples, cones, cavities, and waves like ridges, at the center and peripheral regions of the generated crater print.86 2.6.2. Laser pulse wavelength Excitation laser wavelength (λ) also plays an important role on plasma generation, laserplasma coupling, plasma expansion dynamics and confinement, plasma properties and crater generation.87 The differential effects that laser pulse wavelengths cause on the generated plasmas have not been only evaluated experimentally88-92 but also modeled theoretically.81,93 Typically, harmonics of different orders (λ = 532, 355, 266 and 213 nm) from the fundamental wavelength of a Nd:YAG laser (λ = 1064 nm) have been used to investigate the influence of the laser wavelength in LIBS experiments. The energy of incident photons affects the laser-target coupling and the ensuing laser-plasma interactions that, in turn, modify the morphology of the plasma plumes. In line with this, Boueri et al.50 captured shadowgraph pictures of plasmas induced from Nylon. Spherical shape for the shock wave expansion of plasmas generated by UV radiation was observed. Conversely, a preferential propagation in the direction of the laser incidence resulted from IR excitation. In good agreement with this work, cylindrical plumes were observed following the laser ablation of aluminum with 1064 nm radiation. The cylindrical morphology of plumes denotes the dominance of IB absorption by plasmas at long laser wavelength.87 Similarly, in the laser ablation of polymers, lower plasma thresholds and more intense emissions for IR excitation Introduction 55 than those for UV were observed. These results were attributed to higher plasma temperatures for the former case since absorption by the plasma due to IB increases with wavelength.94 Particularly, in the field of laser ablation of organics there is still an ongoing discussion about the variety of mechanisms for material removal that are active depending on the particular excitation wavelength, for example, whether in addition to photothermal processes, photochemical reactions or even photophysical and mechanical processes are relevant.95 In a first step, it is generally recognized that for ns laser pulses, the energy of the laser photons is exploited for electronic excitation. The following steps are still under discussion. In photochemical processes, electronic excitation by the optical absorption of high energy photons results in direct bond breaking.96-98 The molecules come into a dissociative electronic excited state since the chemical bonds breakdown due to the expulsive force between atoms. Hence, when the energy of the incident photons is higher than the energy of a specific bond, this may be directly broken by just absorbing one photon. For example, a typical C-N bond, with a dissociation energy of 3.04 eV, may be broken by absorbing one UV photon (λ = 266 nm) with an energy of 4.66 eV. However, when using IR radiation at 1064 nm (1.17 eV), a multiphoton absorption process would be required for the photodissociaton of this bond. In this case, photothermal processes are more probably to occur. The electronic excitation is thermalized on a picosecond (ps) timescale, and the progressive heating of the target results in thermal bond breaking.99-101 Finally, if both thermal and non-thermal mechanisms significantly contribute to the overall processing rate, the process is denoted as photophysical. Two independent channels of bond breaking102,103 or different bond breaking energies for ground-state and electronically excited-state chromophores104,105 are supposed for this mechanism. Consequently, together with differences in the morphology of laser-produced plasmas, dissimilarities between the ablation craters printed as a function of the irradiation wavelength have been observed. Just for instance, the laser ablation of polyvinyl chloride with 266 nm excitation generated clear craters on the polymer surface, whereas with 355 nm wavelength larger thermal damage was produced on the Introduction 56 material; circumstances also denoting the dominance of photochemical and photothermal processes, respectively.90 2.6.3. Laser pulse duration The pulse duration (  ) is often defined as a full width at half maximum (FWHM), that is, the width of the time interval within which the power is at least half the peak power. Any focused laser pulse, whatever its duration, meets the required conditions in ablation since usually the rate of energy deposition greatly exceeds the rate of energy redistribution and dissipation. As a result, extremely high temperatures are attained in those regions where energy absorption occurs. Since the mechanisms of energy dissipation are variable, changes in the temporal width of laser pulses lead to fundamental differences of the ablation process.106 A differentiation between ultrashort (<1 ps) or short (>1 ps) ablation regimes is widely accepted. Similarly, the terms ns-LIBS and fs-LIBS are often found in literature.107 In this section, a concise description of the main effects of the laser pulse duration in laser-induced plasmas will be accomplished. Figure 4 shows a schematic illustration where the main features of fs-LIBS and ns-LIBS, at the different stages of the plasma evolution, are compared. Commercially available fs and ns lasers typically generate pulse durations of tens of fs and few ns, respectively. Hence, effects of laser pulse durations of 60 fs and 8 ns (18 µm and 2.4 m in length, respectively) have been discussed as hypothetical cases in this example. In a first stage, when incident light strikes the target surface, the photons of the laser pulse, regardless its length, are absorbed by the material. For fs-LIBS, the absorption of radiation and the material heating is extremely fast. Indeed, the thermal conduction from the laser-impact area to the lattice is minimized because the pulse duration is shorter than the timescale of photon-electron lattice interactions.108 However, with ns excitation, there is a longer interaction time between the laser beam and the sample. Therefore, thermal effects generally take Introduction 57 places during the ablation process. As observed, at early times of plasma formation, the closing tail of the ns pulse may also interact with the expanding plasma and reheat it, thereby increasing its temperature and electron density. Conversely, laser-plasma interactions do not exist during ultrashort excitation since the fs pulse delivery finishes before the plasma onset. In this context, phenomena such as the reheating of the plasma and its shielding effect are expected to have negligible contributions.50 In particular, reheating of the plasma is the responsible cause of the significantly higher intensity of continuum emission identified in ns-LIBS spectra as compared to that detected in fs-LIBS ones.109,110 Numerous studies have been conducted in the last years to assess the substantial differences between ns and fs laser-produced plasmas.80,111-113 For instance, concerning plasma geometry, ns laser-induced plasmas typically show spherical expansion whereas fs excitation produces narrower plasmas, which preferentially expand normal to the sample surface. Furthermore, decay rates of electron density and temperature have been found faster in plasmas generated with fs pulses as compared to those obtained for ns laser-generated plasmas. Finally, as indicated in Figure 4, substantial differences in shape as well as volume have been noticed in the morphology of craters formed by lasers with different pulse widths (t = ∞).114 Introduction 64 The main components are depicted: the pulsed laser that generates the powerful light packages used to produce the plasma plume; the optical system for delivering a focused laser pulse to the target sample, comprising a laser wavelength optical mirror and a focusing lens; the target holder; the optical system for collecting the light emitted by the created plasma and for guiding it towards the detection device, comprising a collecting lens and an optical fiber; the detection system consisting of a spectrograph to spectrally resolve the light and a detector –intensified charge-coupled device– to record the light; and the computer, with the specific control software, for acquisition and storage of the spectral information. The following sections encompass a brief description of these key pieces in a LIBS experiment. Introduction 65 3.1. Excitation source A laser system of high energy as excitation source is the key component within the LIBS framework. In this section, basic principles of laser operation and different laser sources used in LIBS experiments will be briefly described. The term "LASER" is an acronym for Light Amplification by Stimulated Emission of Radiation. The mechanism of stimulated emission produces photons with the same frequency, same phase, same sense of polarization and that propagates in the same direction. Figure 6 sketches the main elements of a laser system to describe the principle behind laser action. A laser basically consists of a pumping source that supplies energy to an active medium placed on an optical resonator. The pumping source perturbs the level population distribution in the active medium with transitions from the ground state level to the upper level that ultimately results in population inversion. Gas discharge laser lamps and direct-current arc lamps are routinely used as optical pump sources for solid state lasers, whereas gas lasing media is typically optically pumped using radiofrequency. Once the population inversion is reached, the spontaneous emission photons that are emitted along the optical axis of the resonator are responsible for initiating the formation of an amplified light wave. This wave is continuously reflected forward and backward between the two reflective mirror parallel surfaces that bound the optical resonator. For each round trip in the optical resonator, the light wave passes through the active medium twice and experiences amplification as long as the active medium exposes the population inversion. Furthermore, the shape and separation of these mirrors define the spatial distribution of the light wave inside the laser. However, while one of these mirrors reflects almost one hundred per cent of the light wave the other is partially transparent to allow its output. The light wave lost in the resonator that passes through the output coupler leads to produce the laser radiation.132 Introduction 66 The distinctive properties that characterize laser radiation are the brightness, the tunability and high chromaticity, high directionality, temporal and spatial coherence, and controlled polarization. Figure 6. Schematic diagram of a typical pulsed laser system. A multitude of laser types with different wavelengths, energies, pulse durations, and light beam profiles are applied for laser ablation studies and LIBS applications, depending on the requirements of each research. The following paragraphs shortly outline the main lasers used in LIBS. 3.1.1. Nanosecond lasers Nowadays, the most widespread excitation sources used for LIBS measurements are solidstate lasers, flash lamp-pumped, with Nd:YAG as laser medium and operated in the Q-switch mode to generate high-energy laser pulses with durations in the nanosecond range. These lasers are composed by an yttrium aluminum garnet crystal host (Y3Al5O12) were neodymium atoms (triply ionized –Nd3+–) are embedded to form the laser active medium. The Nd:YAG laser constitutes a four-level system. The pumping of the active medium, which is normally found rod-shaped, is accomplished by the action of Xe flash lamps, which are capable to produce microsecond to Introduction 67 millisecond duration pulses of broadband light of high radiant intensities at high repetition rates. The flash lamps may simply be arranged parallel to the rod. However, in the most common configuration the flash lamp is a tube located at one of the focus of a mirrored cavity, consisting of an elliptical cross-section perpendicular to the rod's axis, with the laser rod located at the other focus of the cavity. Under this configuration, the full emission from lamp is gathered by the rod. An alternative for pumping the Nd:YAG rod is the use of a semiconductor diode laser pumps (diode pumped solidstate –DPSS– laser). In these kind lasers, the pumping is done either by an array of diodes placed on the side the rod or by a fiber-coupled diode laser that illuminates the endface of the laser rod.133 At the same time, for obtaining short and highenergy pulses, the most common operational manner for pulsed lasers is Q-switched mode. While the pumping action is continuous, the light is allowed to reflect in the mirrors only for a short time to achieve the laser action. Such short periodic intervals are controlled by an acousto–optic coupler. An oscillating electric signal drives a transducer to vibrate, which creates sound waves in the crystal, thereby causing it to act as a grating for the incoming light. Thus, when there is no density change along the light path through the crystal, the optical amplification occurs. In contrast, laser action is prevented if light is scattered. The Q-switching allows generating laser pulses of few nanoseconds (5-20 ns) with peak energies from 10 mJ to 1 J. Although the output laser wavelength for the normal operational mode of Nd:YAG laser is 1064 nm, additional output shorter wavelengths may be easily generated with a variety of non-linear conversion techniques – harmonics generation. Non-linear crystals (e.g. potassium dihydrogen phosphate –KDP– or Barium borate –BBO–) are used to frequency doubling (532 nm), to triple the frequency (355 nm), and to frequency quadruplicating (266 nm) of the laser output. Although some larger wavelengths may be tuned, this conversion leads to a notable decrease of the laser energy.132 Introduction 68 3.1.2. Femtosecond lasers While lasers with nanosecond pulse duration are the standard “workhorse” for a LIBS setup, the discovery of high fluence solid state materials like Ti:Sapphire together with the invention in the late 1980s of the chirped pulsed amplification (CPA) technique, which is based on the stretching in time of the laser pulses prior to their amplification, led to remarkable progress in the development of ultrahigh peak power lasers.134,135 The basic operating scheme of CPA is illustrated in Figure 7. CPA-based fs laser systems consist of a seeding laser –also known as femtosecond oscillator–, a stretcher, an amplifier and a compressor, which act as detailed below: (1) As a first building block, the seeding laser is the responsible component to generate fs pulses. Although in the past few years new laser materials that can produce fs-class pulses have emerged, the shortest laser pulses (<10 fs) are typically emitted by Ti:sapphire lasers through the most effective modern short-pulse generation mechanism –Kerr-lens-mode-locking (KLM).136,137 These stable oscillators produce a train of fs pulses at a high repetition rate (ca. 10 MHz), necessary to support mode-locking, but with a limited energy per pulse of tens or hundreds of nanojoules. However, they are tunable over as much as 400 nm (680-1080 nm). (2) The output of the megahertz-repetition-rate oscillator is then chirped. The chirping process consists on the pass of the ultrafast pulses through a dispersive medium (for example a tilted grating) to be stretched up to a longer duration –nanosecond pulse width– to avoid damage to the optical components of the amplifier stage caused by the high peak power. In summary, chirping process spreads the wavelengths in the pulse over a much longer interval, reducing the peak power before amplification. (3) In the amplifier the laser pulse energy is enlarged to several milijoules. The bandwidth of the amplifier must be large enough to accommodate the full spectrum of the short pulse. The amplifier of course requires optical pumping to sustain gain and amplification. Introduction 69 (4) The pulses ejected from the amplification stage, usually via an electro-optical modulator, become recompressed close to the original pulse duration or less. The compressor unit reverses the dispersion process, squeezing the amplified light in time to recreate the ultrashort pulse, this time with much higher amplitude. Before fs laser pulses are used in any experiment, diagnostics in order to fully characterize their temporal profile and their spectral phase is routinely performed.138 In order to gain better insight into the operation principle of fs lasers, the reader is referred now to the references 134-138. Figure 7. Concept of chirped pulse amplification (CPA) based laser. The invention of the CPA technique in 1985 was a revolutionary breakthrough in femtosecond laser systems, resulting in the appearance of powerful femtosecond laser pulses and in the imminent availability of powerful commercial laser systems. Ultrashort pulse duration gave numerous benefits with respect to ns-LIBS: low ablation threshold, reduced sample damage, small ablated mass, improved spatial resolution, and absence of fractionation vaporization.107 As discussed above (Section 2.6.3), the mechanisms leading to energy absorption and target ablation are entirely different for nsand fs-LIBS. The fs laser pulses are so short that processes like ionization, sample heating, and vaporization all do not occur until the end or after the laser pulse. Hence, all laser Introduction 70 energy is deposited directly onto the sample before ablation occurs. The differences in laser-sample and laser-plasma interactions as well as in ablation mechanisms during interaction between ultrafast and nanosecond pulses with the sample lead to significant differences in crater shape and volume. fs lasers provide higher precision during ablation and minimized heat-affected zone on the material as compared to ns lasers. Disparities between expansion dynamics of generated plumes have been also detected. While fs-generated plasmas expand with a strong forward bias in direction normal to the sample surface, a spherical expansion is noticed for ns-produced plasmas. Such plasma expansion processes dictates distinct plasma lifetime, plasma emission and analytical figures of merit due to the distinctive duration of laser pulses. For all these reasons, fs laser pulses are predestined to providing a brighter future in numerous LIBS applications. 3.2. Optical system The optical system comprises all active optical components used in a LIBS setup. Thus, all these optic components may be subdivided by their function, subject to the action during emission spectroscopy assisted by laser ablation in which they are involved. We refer to focusing optics, which is intended to guide and focus the laser pulses onto the matter, to produce the plasma plume; and to the collecting optics, responsible for gathering and conducting the emitted light from the plasma up to the detection device. The following sections encompass a brief description of this optical system in a LIBS experiment. 3.2.1. Focusing optics Beyond some high-reflectivity mirrors to the laser-wavelength used to guide laser pulses along optical path from the laser to the sample, to attain and exceed the threshold for the plasma formation (typically greater than 108 W·cm-2) at the location of interaction with the sample, laser Introduction 71 pulses are usually focused down to a very small area by means of a single lens (in general a planeconvex lens and sometimes bi-convex or best form –spherical or cylindrical– lenses). Each laser ablation setup usually contains a customized focusing lens in relation to the specific experiment requirements. Some critical features of this lens are transmission efficiency at the irradiation wavelength, the type of coating, the damage threshold, and the focal length. This last is particularly significant since it dictates the distance between the lens and the sample for which the laser pulse is focused to a highly small area.139,140 Thus, when a rectangular beam with a plane wavefront of finite diameter  is focused by the lens onto a plane, the individual parts of the beam that pass through the lens converge and are imagined as point radiators of a new wave front. Since these radiators interfere with each other on the focal plane, constructive and destructive superpositions take place, and a distribution of the light energy with the central maximum containing about 86% of the total power in the beam is attained. The smallest focal diameter () limited by this diffraction phenomena of the wavefront, usually defined by the points where the intensity has fallen to (1/e2) of the central value, is given by the equation 3.1.   󰇛2 ·  · 󰇜/ 󰇟Eq. 3.1󰇠 where  is the focal length of the lens,  is the laser wavelength and  the diameter of the laser beam. For a plane front circular beam, a correction factor of 1.22 is introduced and equation 3.2 becomes:   󰇛2 · 1.22 ·  · 󰇜/ 󰇟Eq. 3.2󰇠 In the case of multimode laser beams the smallest focal diameter will be even larger since the beam comes from a cavity having several off-axis modes of vibration rather than from an apparent point source. This further correction for a multimode lasing regime considers transverse Introduction 72 electromagnetic modes of laser beam. However, since the number of radial zero fields, , are more critical to the focal spot size than the number of angular zero fields, , and that of longitudinal fields, , a new part is introduced into the equation 3.3. The new expression reads as follows:   2·1.22·· ·󰇛2  1󰇜 󰇟Eq. 3. 3󰇠 Although focal spot size is related to beam focus, wavelength, diameter and mode, there are other factors that affect focal spot diameter such as spherical aberration and thermal lensing effects. In general, any lab bench LIBS setup uses a single focusing lens. However, in the case of field-deployable LIBS sensors, operating at standoff and remote configurations, for which the lens-tosample distance may vary, optical arrangements composed by several lenses (e.g. beam expanders and telescopes) are used to adjust the laser beam focus.42,140 3.2.2. Collecting optics With regard to the gathering of the radiation emitted from the plasma up to the detection device, several configurations can be considered in a LIBS experiment. In an illustrative way, Figure 8 sketches three optical layouts used in plasma light collection. Figure 8A shows the simplest scheme to gather plasma emission. A converging lens is placed orthogonal –possible option is also an angular arrangement–to the direction of laser incidence. Alternatively, Figure 8B depicts the simplest configuration in which the use of a single lens is exploited. The collection of light from the plasma materializes with the lens used to focus the laser beam. For this configuration a dichroic mirror is needed. As seen, this mirror allows passing the incident radiation wavelength but reflects the plasma light gathered by the lens. This configuration allows observing emission of the plasma shock front during its propagation, but it is not an arrangement to observe the axial and radial directions of the plasma expansion. Finally, Figure 8C shows optical layout using a spherical mirror Introduction 73 instead of a lens to collect the plasma emission. Thus, all wavelengths emanating from the distant plasma that strike the mirror parallel to its principal axis are brought to a focus at the same convergence point. Accordingly, the use of this optic element suppresses any chromatic aberration. Figure 8. Frequent optical arrangements for plasma light collection in LIBS measurements. Plasma light collected by using any of these configurations may be focused either directly on the entrance slit of a spectrograph or on the tip of an optical fiber.42 Spectroscopic techniques extensively use optical fibers to pass radiation emitted to the detection device because of their excellent versatility. They are made of synthetic fused silica (amorphous silicon dioxide) and may be doped with trace elements to change their optical properties. A relatively large core and a high numerical aperture (NA) allow the fiber to gather the complete plasma volume. Core diameters of the optical fiber from 50 µm to 1000 µm are frequently used for LIBS measurements. Together with the size of the focused image of the plasma, the choice of the adequate optical fiber also involves to consider the wavelength range of interest. In this context, solarization resistant fibers are needed to observe deep-UV light (below 230 nm). Introduction 80 4.1. Optical emissions in plasmas of molecular solids The big challenge in the analysis of molecular solids by LIBS relies on the high similarity of the emission spectra gathered. Emission lines from atoms or ions of C, H, N and O as well as sequences of CN and C2 molecular bands are the most representative signals featured in LIBS spectra of organic compounds. To put the discussion into context, Figure 11 depicts LIBS spectra of polyethylene (PE) and polytetrafluoroethylene (PTFE) induced with visible laser radiation (532 nm) under air atmosphere. As seen, despite that the two polymer materials have different elemental composition, the same emission signals are featured within their spectra. The differences observed between both spectral responses are basically slight changes in the relative intensities of the optical emissions. Such a circumstance suggests that differentiation between organic compounds by LIBS needs to be tackled using a comprehensive framework of relative emissions from atoms and small radicals rather than addressed from the simple detection of all these specific spectral features. Faced with this challenge, knowledge of all possible species that populate the plasma as well as the plausible formation routes from which they may arise; subjects that can significantly help to substantially improve selectivity of LIBS for sensing molecular solids. The main features of optical emission spectra of organics as well as the models proposed for the formation of diatomic radicals in laser-induced plasmas will be discussed in the following sections. Introduction 81 Figure 11. LIBS spectra of polyethylene (top) and polytetrafluoroethylene (bottom) from plasmas induced with ns pulses (λ=532 nm) in air atmosphere. Emissions of atoms and ions from organic compounds As already stated, optical emissions revealed by LIBS spectra of organic compounds are mainly associated to atoms and ions of carbon, hydrogen, nitrogen and oxygen. 161 The electron transitions and the associated wavelengths for the most sensitive emission lines of such species are listed in Table 1. Introduction 82 Table 1. Main electron transitions and the associated wavelengths for emission lines of C, H, N and O. Note: optical emissions from H corresponding to the Balmer series are identified as H-delta, H-gamma, H-beta and H-alpha (n, radial quantum number). (I) and (II) denotes atoms and single-ionized atoms, respectively. Some molecular solids contain heteroatoms other than oxygen and nitrogen, like are phosphorus, sulfur, and the halogens –fluorine, chlorine and bromine. The ability of LIBS for the detection of such elements has been proved.162-165 Notwithstanding this, since these elements have not been subject of study in the performed investigations, they are not emphasized in the present Doctoral Thesis. Emissions of diatomic species from organic compounds The most representative molecular emissions in LIBS are associated to C2 and CN radicals. Concerning C2 radicals, the laser ablation of the molecular solids may produce carbon dimers, often in an emissive excited state, through a variety of processes, being remarkable among them the direct release of native C=C bonds from the original molecule and a recombination of C atoms. While the C2 molecule boasts seven triplet and six singlet electronic states giving rise to nine band Specie Transition Wavelength (nm) C (I) 2s22p3s → 2s22p2 193.09 C (I) 2s22p3s → 2s22p2 247.86 C (II) 2s24s → 2s23p 392.07 C (II) 2s24f → 2s23d 426.73 Hδ (I) 2 → 6 (n → n) 410.17 Hγ (I) 2 → 5 434.05 Hβ (I) 2 → 4 486.14 Hα(I) 2 → 3 656.28 742.36 N (I) 2s22p2(3P)3p → 2s22p2(3P)3s 744.23 746.83 777.19 O(I) 2s22p3(4S°)3p → 2s22p3(4S°)3s 777.42 777.54 Introduction 83 systems, which are spread over the entire range of the spectrum –from the vacuum ultraviolet, through the visible, up to the infrared spectral region–, the most easily excited of these band systems, and consequently with the most intense emissions, is the Swan band system, which arises from the d3πg–a3πu transition and contains six sequences lying between 420 and 720 nm.41,166-168 Within these band sequences (Δ, their specific vibrational transitions can be resolved depending on the ability of the detection system to distinguish between wavelength intervals. Even sometimes, the rotational transitions of each vibrational level are also exhibited. To graphically contextualize this, Figure 12 depicts the emission spectra, ranging from 430 nm to 570 nm, of four sequences (Δ= 2, 1, 0, -1) of C2 Swan system from graphite plasmas according to the resolution ability of the detection system. As seen, in the low-resolution spectrum the vibrational transitions of each sequence are hardly distinguished, whereas the overall structural appearance of well-developed vibrational peaks is clearly discernible in the high-resolution optical response. Those spectra that feature all these transitions allow determining the rotational and vibrational temperatures (Trot and Tvib) of the radical from the Boltzmann plot method169 or calculating them by matching the strength of lines in optical spectra synthesized with semi-automatic program packages with those in the corresponding experimental ones.170 Introduction 84 Figure 12. Optical responses of C2 Swan Band System reached when graphite plasma emissions illuminate diffraction gratings with different total number of grooves on their surface, A) 150 grooves mm-1 and B) 1800 grooves mm-1. As far as cyano species is concerned, CN emissions are generally featured in LIBS of carbon-containing organic samples when plasmas are induced under air atmosphere. The Violet system associated to the B 2 ∑ + -X 2 ∑ - transition, whose main sequences covers from ca. 350 to 430 nm, governs the spectrum. Figure 13 shows the spectral emission of the sequences Δ= +1, 0 and - 1 for the CN species from plasmas of polyethylene. As observed, in general, under typical operational conditions for LIBS analysis the Δ= 0 band sequence is the most sensitive. As in the case of C 2 , if all transitions are clearly resolved, the spectral analysis allows inferring of the rovibrational temperatures of CN radical that may be representative of the plume temperature if thermodynamic equilibrium exists. 171-173 Introduction 85 Figure 13. High resolution optical responses of CN Violet Band System from polyethylene plasma. Together with CN and C 2 species, other diatomic molecules are present in the excited vapor. Depending on the ablation conditions, temporal settings and the instrumentation used for light detection, less sensitive emission bands associated to OH, CH, NH, N 2 and NO excited molecules may also be featured in LIBS spectra of organics. 120,174 Beyond these observable species, the existence of others radicals, which either do not emit in the UV-Vis-IR region of the electronic spectrum or are intermediate reaction products with extremely short lifetime, cannot be ruled out. Indeed, the combination of LIBS with mass spectrometry based techniques has allowed detecting a higher number of molecular fragments (e.g. C 2 H + , C 2 H 2+ , C 3 , CO + , HCNH + ). Particularly, laser ionization mass spectrometry (LIMS) 175 and ion flow tube mass spectrometry (SIFT-MS) 33,176 have been successfully applied to study the fragmentation and recombination mechanisms in plasmas of organic explosives. Introduction 86 Table 2. Optical emissions for the band-heads of the sequences contained in the most intense electronic transitions of some diatomic molecules.177 Notwithstanding this, investigations described in the present Doctoral Thesis focus only on the most regular diatomic species that exhibit optical response. Table 2 reports on emission wavelengths for labeling the band-heads of the sequences contained in the most intense electronic transitions of the emitting species concerned. To show on paper the typical appearance of molecular emission fingerprint, Figure 14 depicts the fs-LIBS spectrum of terephthalic acid in which the spectral features of the most significant diatomic radicals that seed the laser-produced plasma are exhibited. As seen, alongside the most common CN and C2 optical features, emissions associated to the OH, CH, and NH radicals also emerge in the spectral response. While the most plausible routes to produce CN and C2 species in plasmas of molecular solids have been under active investigation and are practically elucidated,178,179 the origins of hydrogenated species and their role within the plasma chemistry have Radical System Transition / Sequence Wavelength (nm) OH 2 A∑+–X2π (0–0) Δ= 0 308.9 NH 335.8 A3π–X3∑- (0–0) Δ = 0 (1–1) Δ = 0 336.9 CH 2 CΣ+–X2П (0–0) Δ = 0 314.3 2 BΣ-–X2П (0–0) Δ = 0 387.1 2 AΔ–X2π (0–0) Δ= 0 431.4 CN Violet B2∑+–2X∑+ (1–0) Δ = 1 358.7 (0–0) Δ = 0 388.2 (0–1) Δ = -1 421.6 C2 Swan d3πg–a3πu (2–0) Δ = 2 438.3 (1–0) Δ = 1 473.7 (0–0) Δ = 0 516.5 (0–1) Δ = -1 563.5 (0–2) Δ = -2 619.1 (0–3) Δ = -3 667.7 Introduction 87 been not addressed yet in depth. This circumstance has motivated the investigations in progress described in the Chapter 2 of the present Doctoral Thesis. Figure 14. fs-LIBS spectrum from terephthalic acid plasma induced in air at atmospheric pressure. 4.2. Formation pathways of excited diatomic radicals Despite that the potential origins of excited diatomic species in LIBS organic plasmas have been studied in considerable depth, they continue to be debated still today. The direct release as fragments from the parent molecule and the occurrence of different chemical processes (recombination, substitution...) between species within the produced plasmas have been postulated as the feasible sources of such excited radicals. Furthermore, all these chemical routes tend to coexist rather to exclusively progress; while it is true that there is a whole series of factors influence that favor some mechanisms over others. Particularly, the ablation conditions, which define the primitive species that populate the plasma, cause chemical reactions to progress at a different extent. Some insights into this context are discussed below. Introduction 88 Regarding the C2 radicals, its formation process has been studied according to the laser fluence used to ablate the sample. At low fluence, the laser energy delivered produces the partial fragmentation of the parent molecule. Then, molecular fragments with two or more carbon atoms follow a progressive dissociation as indicated in reaction 1: Cn + M → Cn-2 +C2 [R. 1] In contrast, a larger atomization of the molecular solid occurs when increasing the laser energy. Then, recombination of carbon atoms to generate C2 dimers becomes an important route by means of a three-body reaction [R.2]:39,181 C + C + M → C2 [R. 2] Although itemizing all the reactions that may lead to the generation of carbon dimers is out of the scope of this section, some additional routes involving common species in organic plasmas for the generation of C2 are reported. In this vein, the reactions 3 and 4, with CN and CH radicals as chemical precursors, have been also postulated to produce dicarbon radicals:182,183 CN + C → C2 + N [R. 3] CH + CH → C2 + H + H In general, the high excitation temperatures reached in laser-induced plasmas suggest a large atomization of the molecular solids.184 However, the molecular structure of each compound certainly plays an important role in its mechanism of decomposition and therefore in the nature of the output species. Indeed, the nature of the C-to-C linkage (simple, double, triple, or aromatic) within the molecule probably affects the proliferation of C2 fragments. Within this framework, some [R. 4] Introduction 89 experimental results have revealed greater readings of C2 emissions for unsaturated and aromatic compounds; outcomes that may correlate with the strength of the chemical bonds.32,38 To contextualize the discussion, Table 3 displays the average bond dissociation energy –bond enthalpy– for the most usual carbon bonds in organic compounds.185 Table 3. Mean bond enthalpies (kJ mol-1) As seen, the dissociation energy of a double carbon bond is almost twice the strength of a simple bond. Hence, it seems rather likely that C2 emissions observed from plasmas of unsaturated and aromatic compounds are mainly linked with the direct release of C-to-C double bonds from the parent molecule. In contrast, the easiest breakage of C-to-C simple bonds in saturated molecules suggests a recombination mechanism to produce C2 species, if any. In summary, the dissociation energy of the chemical bond also influences on how and in what quantity the carbon dimers may be produced. However, the coexistence of the two processes to generate C2 radicals should not be discarded. In parallel, a great number of investigations have been focused on the possible sources of cyano radicals (CN). The CN radical is a very stable radical compared to other diatomic molecules, as can be deducted from information reported in Table 3, where bond dissociation energies of the most common diatomic species seeding organic plasmas are reported. As seen, CN molecule exhibits the highest dissociation energy. Such a circumstance underscores the large affinity between Chemical bond Bond enthalpy (kJ mol-1) Chemical bond Bond enthalpy (kJ mol-1) C–C 348 C–N 305 C=C 612 C=N 613 C≡C 837 C≡N 890 C–C (in benzene) 518 C–F 484 C–H 412 C–Cl 338 Introduction 96 by LIBS spectra of organic compounds are lines of C, H, O and N, and sequences of molecular bands, primarily of the CN and C2 molecules. Therefore, it is spectral information that requires powerful data treatment tools to address the enormous detection and identification challenges being posed. As a solution, in almost the totally of the cases at hand, LIBS is coupled to chemometric techniques. Basically, chemometrics performs calculations on the emission measurements of chemical excited species using statistic principles, mathematical models and computational methods, thereby allowing extraction a larger amount of information from a raw data set. The combination of chemometric methods with LIBS data has been mainly focused to the detection and identification of unknown targets. In this vein, pattern recognition methods that are based on the labeling of an input measurement (e.g. spectrum), are commonly used. When the models developed for classification are trained from previously labeled data, methods are referred as supervised learning techniques. Conversely, when classification models are trained from unlabeled data, techniques are termed as unsupervised learning methods.207 Regardless the statistical technique used, the same procedure is accomplished for predictive applications in the supervised pattern recognition technique. The framework of this approach is schematized in Fig. 16. First, LIBS information is acquired under fixed and well-defined operational conditions. This original data can be preprocessed, if applicable. For instance, spectral normalization, a background subtraction, or a signal smoothing are common processes to apply. In addition, in general, prior to the chemometric modeling, the data set acquired is reduced by removing non-informative variables. Thus, only those spectral features coherent with the specific chemical constituents of organic compounds are extracted to create a recognition classifier. To this end, supervised recognition method operates in two chronologically non-overlapping stages. First, a learning stage is applied to the information supplied by a labeled training data set. During this phase, the algorithm acquires the necessary knowledge to find a regularity in patterns of the training data set and to derive a rule that is able to differentiate among the regarded classes and, therefore, to Introduction 97 assign, on future, any unknown sample to one of these classes. Once completed this phase, a testing stage starts. Without any attempts to acquire new knowledge, the algorithm proceeds to classify some known samples –labeled test data–, which had not initially been included in the training data set, as members of the considered classes. This step allows evaluating the classificatory abilities of the designed model. Once the performance is assured, the model may be routinely applied to future predictions. Figure 16. Schematic sequence for class identification by supervised pattern recognition. A large variety of chemometric methods have been combined with LIBS to cope with classification of organic samples. Among them, the most outstanding are listed below: ‐ method of normalized coordinates (MNC)208 ‐ principal components analysis (PCA)209 ‐ linear discriminant analysis (LDA)210 ‐ partial least squares-discriminant analysis (PLS-DA)209,211 ‐ soft independent modeling of class analogy (SIMCA)209,212 and ‐ neural networks (NN)212,213 Introduction 98 Because it is not the subject of this Doctoral Thesis, the intricacies of all these chemometric techniques are not significantly detailed. For more detailed information, readers are referred to specific references.214,215 In contrast, in the focusing of the issue before us, the next sections, disaggregated by application fields, will provide a brief review to the state of the art in LIBS analysis of organic compounds. 5.1. Detection of explosive materials Currently, the design of sensors and the development of techniques to detect an explosive material in real-time regardless its location, is of crucial importance due to the latent threat of a terrorist attack.216 In this sense, LIBS disposes of a series of characteristics, which places it at the forefront of explosive sensing techniques, as compared to other analytical techniques. LIBS copes with difficult analytical challenges since hardly alters the target interrogated, has capability of operating in-situ (field-deployable sensors), in real time, and rapidly and with appropriate sensitivity. Furthermore, LIBS sensors may be miniaturized and designed to work in a contactless way at long distances, somewhat which minimizes the risk for the operator.217 Research in explosives detection by LIBS arose in 2006 when residues of organic explosives, located on aluminum supports, were detected at distances up to 45 m. In this work, the discrimination between organic explosives (TNT, RDX, PETN, Composition B and C4) and harmless compounds was accomplished by means of an algorithm based on intensity ratios of representative emission signals.218 Since then, the capabilities of LIBS for the detection of energetic materials have been extensively explored.23,219 In this field, LIBS has reached new heights with the detection of explosive residues from human fingerprints left on solid surfaces at distances up to 30 m.220 Indeed, Introduction 99 LIBS sensors designed to this matter have become more robust, and currently, some of them offer a great number of features like target pointing, scanning and wide-ranging capabilities.221 Beyond the conventional operating mode, LIBS has also coped with the recognition of explosives materials under different configurations. The uses of double pulse configurations (DPLIBS)222-224 and ultrashort excitation (fs-LIBS)225-227 have demonstrated to have several advantages, thereby reporting noteworthy results. DP-LIBS has proved an improvement on the selectivity to the explosives identification, since the contributions of atmospheric emissions, mainly O and N, on the final emission pattern are reduced. This is a crucial factor when O and N emission lines are used as discriminant variables in the chemometric models. Moreover, the sensitivity of the method gets better because emission intensities increase, probably, due to the thermal reheating of the plasma.223 Concerning the effect of ultrashort laser pulses in the recognition of explosives by LIBS, few studies have been done. Such a circumstance is justified on the complexity involved in working with fs-LIBS due to both the big size of femtosecond lasers as well as by their more restrictive operational conditions. All this hinders its assembling in portable sensors. In any case, several studies have evidenced possible advantages of fs-LIBS for the identification of organic explosives as compared to ns-LIBS. Some of them are a reduced interrogation of the substrate when explosive residues are investigated, the minimization of the background continuum within the optical responses, as well as, like in the case of DP-LIBS, a negligible atmospheric entrainment. However, these benefits tend to disappear for large laser fluences.226 A comparison between fsand ns-LIBS of trinitrotoluene (TNT) residues deposited on aluminum supports found larger molecular emissions (C2 and CN), which can be considered as tracers for this explosive, in the case of the ultrashort pulse regime.225 Moreover, the molecular formation dynamics of few explosives like NTO, HMX, and RDX in distinct atmospheres have been scrutinized228 and emission signatures resulting from femtosecond and nanosecond laser-induced plasmas have been compared. Results revealed that the CN-to-C emission intensity ratio in these nitrogen-containing compounds was higher for ultrashort irradiation Introduction 100 as compared to that attained for the ns regime; a distinctive character that may be beneficial for identification purposes.227 In a real-world scenario of a terrorist attack, finding clues lies on the sensing of traces deposited on solid surfaces rather than detecting a bulk material in itself. This operational framework has guided the strategic focus in many of the research works to investigate laser ablation of binary samples instead of the laser sensing of pure explosives. Such binary targets raise several possibilities to occur. The explosive residues may be localized in a large variety of surfaces with widely varying properties, being the chemical composition, due to the subject that we are concerned with, the most relevant feature. Always depending on the thickness, in general, a thin layer of an explosive residue is ablated together with the support behind it. Hence, detection of organic traces located on the surface of a metallic surface is relatively straightforward. The spectrally-resolved emissions of the organic compound are easily traceable among those of the metallic species.229 Indeed, the mere identification of carbon related emissions (e.g. C, CN and C2) indicates the presence of the organic substance. Conversely, if the support and the residue have the same nature (i.e. both are organics) the detection of explosive traces becomes a more difficult task. As both of them exhibit an analogous elemental composition, they contribute with identical signals to the final emission spectrum. Spectral information is expected to differ only in the relative intensity of the observed emissions, regardless of whether the residue exists on the support or not. Once residues have been detected, special case is the subsequent distinction between harmful and innocuous materials, whatever the binary target. The challenge lies on the differentiation between residues that share similar elemental composition; a problem that becomes worse when surface has a related chemical composition. Some solutions have been put forward in this regard. On the one hand, the choice of an appropriate set of experimental conditions, such as the irradiance level and a fine tuning of the focal conditions of the laser pulses, enlarges the sensitivity of a surface analysis, and hence allows a selective inspection of the residue by the absence of any contribution Introduction 101 from the support.230 An alternative option is to tackle the differentiation with spectral data treatments and the chemometric techniques. Since a significant number of investigations conducted during the development of the present Doctoral Thesis project are focused on the design and implementation of chemometric strategies to the LIBS information for recognizing explosive residues, the reader is urged to browse on the specific chapters 5, 6, and 7 for further detailed information. As a summary, it can be pointed out that supervised learning methods (SLMs) have been considered to address three distinct realworld situations: the recognition of organic explosives located on aluminum supports, the recognition of organic explosives located on polymeric substrates, and the recognition of explosive fingerprints on objects for courier services. It should be emphasized that performance of this chemometric framework has been very successfully tested, validated, and also transferred from a laboratory scale to a remote recognition of explosive fingerprints at the Laser Laboratory of University of Málaga.231 5.2. Sorting of polymers The use of LIBS as analytical tool for the classification of polymers lies in the need of alternative methods for recovering plastic wastes within the first stage of a recycling process. As mentioned, the great speed in analysis, the nondestructive nature, and the ability for real-time monitoring of LIBS are features that bring also many benefits for the sorting process of plastics.147 As in the case of explosives recognition, different strategies have been used to process the LIBS information in order to distinguish between polymers. Among them, the correlation with spectral libraries by the method of normalized coordinates (MNC),208,232 the analysis of spectral emission ratios,233. the application of multivariate methods234 and artificial neural networks (ANNs)213,235 have been considered for identifying polymer samples. Polymers such as polyethylene terephthalate Introduction 102 (PET), polyethylene (PE), polypropylene (PP), and polystyrene (PS), which compose the most common objects that we use in our everyday, have been discriminated by means LIBS. On the other hand, LIBS instruments implemented for in-line measurements have proven to effectively identify polymer wastes with high content in chlorine (e.g. polyvinyl chloride, PVC)236 In the same way, emission lines of F have been used as tracers on the identification of fluorine containing polymers (e.g. polytetrafluoroethylene, PTFE).237 Despite that the determination of the elemental composition of polymers allows an efficient sorting of plastics, the waste management process for recycling also seeks to remove potentially toxic plastics. This challenging issue can be addressed through the examination of the particular composition of plastic additives. In this context, LIBS has been applied to analyze the content of toxic heavy metals involved in pigments (Pb, Cd, Hg …) of technical polymers238 and of plastic parts of toys 239 as well as to identify flame retardants (Br, Cl …) in several plastics.240 In this line of enquiry, the most cutting-edge application of LIBS to plastic recycling processes lies on the classification of polymeric fragments from obsolete mobile phones. To this end, chemometric models for plastic samples classification have been designed on the basis of emission lines of elements representative of the coatings, like Si, Al, Ti, Na, K, and Ag, rather than on emission features related to organics.241 Undoubtedly, the feasibility of LIBS for the analysis of polymers seems more than evident in view of the numerous researches dealing with this issue. However, it is still far from being implemented as a reference method of analysis in an industrial environment. 5.3. Identification of microbiological samples The identification of microbiological samples like bacterial spores, molds, pollens, and nerve agent simulants is an extended application of LIBS in the fields of food security, environmental Introduction 103 inspection and citizen security. The similar elemental composition of the samples is again the main challenge in the use of LIBS for the classification of biological species. Hence, on this occasion, researchers have also had recourse to the use of chemometric tools for deciphering spectral information to afford discrimination among the different biomaterials. From the mere use of several cumulative intensity ratios 149 up until implementation of sophisticated multivariate methods 242-244, including the application of simple principal-components analysis245, have been considered to deal with the differentiation of several types of biomaterials simulants of a chemical warfare. In parallel, the analysis of bacterial samples is important to ensure food safety and streamline the food-monitoring process. In this context, multivariate regression analysis of LIBS spectra has been employed to distinguish the live bacterial pathogens Escherichia coli O157:H7 and Salmonella enterica, on various foods (eggshell, milk, bologna, ground beef, chicken, and lettuce) and food-processing surfaces (metal drain strainer and cutting board).246 Also regarding issues relating to human security, neural networks have been used for the Identification and discrimination of bacterial strains of same species causing hospital acquired infections, as they exhibit resistance to antibiotics, through detection of the small changes in their atomic composition, as a result of their mutations and genetic variations.152 Beyond the development of the predictive models, some efforts have been also focused in the use of alternative operative conditions to deal with these issues. In this regard, bacterial samples (Escherichia coli and Bacillus subtilis) have been analyzed by LIBS using femtosecond laser pulses. Some specific features, like the negligible displaying of emissions of N and O from the excited surrounding air, the favored emissions of intramolecular bonds as compared to the atomic ones, and the different kinetic behavior of molecular band head intensities of native CN bonds released by sample and of new formed CN from carbon recombination with atmospheric nitrogen, have been attained using this ablation regime. All this converts fs-LIBS in a very attractive technique for analyzing biological samples, since the outputs provide valuable data that allows an unambiguous Introduction 104 distinction for a specific bacterium either from a natural environment or from other biological species.112 5.4. Food industry LIBS has also demonstrated the potential to be a game changer in the food industry, which has also benefited from the product discrimination attainable with conventional LIBS sensors. In this context, LIBS combined with neural networks (NNs) has been applied to the identification, quality control, traceability, and adulteration detection of extra virgin olive oils with a certainty of more than 95%.247 Furthermore, the level of C2 emissions compared to atomic carbon emissions has been used to correlate the content of saturated fatty acids in oils in order to differentiate between palm oil and a group formed by corn, olive and sunflower oil.248 Another attractive application is the analysis of milk samples. LIBS measurements in low pressure He atmosphere allowed obtaining a good correlation between CN emission levels and the protein content in milk. Such emission feature was used to quantify the protein concentration in the samples tested.249 In parallel, a qualitative analysis of the composition, including not only proteins but also trace elements, of maternal milk and commercial infant formulas was performed by Z. Abdel-Salam et al. The relative concentration of proteins was also related to CN and C2 emissions.250 Finally, LIBS has been also employed to detect residues of pesticides (chlorpyrifos) located on fruit surfaces. In this case, P, S and Cl emission lines were used as main analytical assets to differentiate between clean and contaminated fruits.251 Introduction 105 5.5. Pharmaceutical industry While up until only a few years ago, LIBS was largely unfamiliar to the pharmaceutical industry, several research works have evidenced the possibility of using this technique for qualitative and quantitative analysis of pharmaceutical products that are mainly formed by an organic matrix. The capabilities of LIBS for a rapid quantitative analysis of multi-component pharmaceutical tablets was first illustrated using the atomic line emissions from some specific hetero-atoms such as phosphorus, chlorine, and fluorine contained in the pharmaceutical active ingredients of solid dosage form. Since none of these elements were present in the inactive matrix, they could conveniently be used to tag and quantify the drug content. It was also shown that using internal standardization of the corresponding atomic line emission with the C emission line at 247 nm the analytical performance of LIBS, namely in terms of sensitivity, linearity, precision and reduction of matrix effects was significantly improved. Furthermore, it was demonstrated that LIBS was able to measure the concentration of magnesium stearate (used as lubricant in some tablets); a task which had not been possible with High performance liquid chromatography technique –routinely used in the pharmaceutical industry.252 Concerning the organic analysis, several analgesic pills —paracetamol, ibuprofen, and acetylsalicylic acid— have been indentified based on the use of spectral libraries and intensity ratios involving C, C2, H, N and O emissions.253 In addition, conventional chemometric tools (SIMCA and PLS-DA) and a nonlinear classification method as support vector machines (SVM) have been applied to the discrimination of pharmaceutical samples.254 A comparative study evidenced that the utilization of SVM provides important improvements over classical chemometric approaches for the discrimination of such samples in terms of accuracy and sensitivity.255 Beyond chemical analysis, LIBS has shown to be a valuable tool in this application field to cope with other type of essays to assess the quality of pharmaceutical products. In this context, even though it proceeds through evaluation of optical emissions from representative elements, LIBS has Introduction 112 62 Y. B. Zel’dovich, Y. P. Reizer, Physics of shock waves and high temperature dynamics phenomena, Dover publications, Inc., Mineola, New York, Academic (2002). 63 J. J. Chang and B. E. Warner, Laser-plasma interaction during visible-laser ablation of metals, Appl. Phys. Lett. 69 (1996) 473–475. 64 C. Porneala, D. A. Willis, Observation of nanosecond laser-induced phase explosion in aluminum, Appl. Phys. Lett. 89 (2006) 211121. 65 E. Leveugle, D.S. Ivanov, L.V. Zhigilei, Photomechanical spallation of molecular and metal targets: Molecular dynamics study, Appl Phys A-Mater. 79 (2004) 1643–1655. 66 A. A. Ionin, S. I. Kudryashov, L. V. Seleznev, Near-critical phase explosion promoting breakdown plasma ignition during laser ablation of graphite, Phys. Rev. E 82 (2010) 016404. 67 S. S. Harilal, A. Hassanein, M. Polek, Late-time particle emission from laser-produced graphite plasma, J. Appl. Phys. 110 (2011) 053301. 68 F. Capitelli, F. Colao, M. R. Provenzano, R. Fantoni, G. Brunetti, N. Senesi, Determination of heavy metals in soils by laser induced breakdown spectroscopy, Geoderma 106 (2002) 45−62. 69 I. B. Gornushkin, A. Ya. Kazakov, N. Omenetto, B. W. Smith, J. D. Winefordner, Experimental verification of a radiative model of laser-induced plasma expanding into vacuum, Spectrochim. Acta Part B 60 (2005) 215−230. 70 G. Cristoforetti, E. Tognoni, L. A. Gizzi, Thermodynamic equilibrium states in laser-induced plasmas: From the general case to laser-induced breakdown spectroscopy plasmas, Spectrochim. Acta Part B 90 (2013) 1–22 71 H. R. Griem, Plasma spectroscopy, McGraw-Hill, Inc., New York (1964). 72 J. A. Aguilera, C. Aragón, Multi-element, Saha–Boltzmann and Boltzmann plots in laser-induced plasmas, Spectrochim. Acta Part B 62 (2007) 378–385. 73 S. Rai, A. K. Rai and S. N. Thakur, Identification of nitrocompounds with LIBS, Appl. Phys. B: Lasers Opt. 91 (2008) 645–650. 74 M. Dong, X. Mao, J. J. González, J. Lu, R. E. Russo, Time-resolved LIBS of atomic and molecular carbon from coal in air, argon and helium, J. Anal. At. Spectrom. 27 (2012) 2066–2075. 75 D. A. Rusak, B. C. Castle, B. W. Smith, J. D. Winefordner, Excitational, vibrational, and rotational temperatures in Nd:YAG and XeCl laser-induced plasmas, Spectrochim. Acta Part B 52 (1997) 1929– 1935. 76 C. Aragón, J. A. Aguilera, Characterization of laser induced plasmas by optical emission spectroscopy: A review of experiments and methods, Spectrochim. Acta Part B 63 (2008) 893–916. 77 J. Luque and D. R. Crosley, LIFBASE: database and spectral simulation (version 2.1.1), SRI International Report MP 99– 009, 1999. Introduction 113 78 G. Cristoforetti, A. De Giacomo, M. Dell'Aglio, S. Legnaioli, E. Tognoni, V. Palleschi, N. Omenetto, Local Thermodynamic Equilibrium in Laser-Induced Breakdown Spectroscopy: Beyond the McWhirter criterion, Spectrochim. Acta Part B 65 (2010) 86–95. 79 C. Tendero, C. Tixier, P. Tristant, J. Desmaison, P. Leprince, Atmospheric pressure plasmas: A review, Spectrochim. Acta Part B 61 (2006) 2–30. 80 R. W. P. McWhirter, in: Eds. R.H. Huddlestone, S.L. Leonard, Plasma diagnostic techniques, Academic Press, New York (1965) pp. 201–264. 81 A. Bogaerts, Z. Chen, Effect of laser parameters on laser ablation and laser-induced plasma formation: A numerical modeling investigation, Spectrochim. Acta Part B 60 (2005) 1280–1307. 82 J. M. Vadillo, J. M. Fernández Romero, C. Rodríguez, J. J. Laserna, Effect of plasma shielding on laser ablation rate of pure metals at reduced pressure, Surf. Interface Anal. 27 (1999) 1009–1015. 83 G. Cristoforetti, S. Legnaioli, V. Palleschi, E. Tognoni, P. A. Benedetti, Observation of different mass removal regimes during the laser ablation of an aluminum target in air, J. Anal. At. Spectrom. 23 (2008) 1518–1528. 84 X. Bai, Q. Ma, M. Perrier, V. Motto-Ros, D. Sabourdy, L. Nguyen, A. Jalocha, J. Yu Experimental study of laser-induced plasma: Influence of laser fluence and pulse duration, Spectrochim. Acta Part B 87 (2013) 27–35. 85 S. S. Harilal, C. V. Bindhu, Riju C. Issac, V. P. N. Nampoori, and C. P. G. Vallabhan, Electron density and temperature measurements in a laser produced carbon plasma, J. Appl. Phys. 82 (1997) 2140– 2146. 86 M. Akram, S. Bashir, A. Hayat, K. Mahmood, R. Ahmad, M. Khaleeq-U-Rahaman, Effect of laser irradiance on the surface morphology and laser induced plasma parameters of zinc, Laser Part. Beams 32 (2014) 119–128. 87 A. E. Hussein, P. K. Diwakar, S. S. Harilal, A. Hassanein, The role of laser wavelength on plasma generation and expansion of ablation plumes in air, J. Appl. Phys. 113 (2013) 143305. 88 M. P. Mateo, G. Nicolás, A. Yañez, Characterization of inorganic species in coal by laser-induced breakdown spectroscopy using UV and IR radiations, Appl. Surf. Sci. 254 (2007) 868–872. 89 C. Barnett, E. Cahoon, J. R. Almirall, Wavelength dependence on the elemental analysis of glass by laser induced breakdown spectroscopy, Spectrochim. Acta Part B 63 (2008) 1016–1023. 90 W. Q. Lei, Q. L. Ma, V. Motto-Ros, X. S. Bai, L. J. Zheng, H. P. Zeng, J. Yu, Effect of ablation photon energy on the distribution of molecular species in laser-induced plasma from polymer in air, Spectrochim. Acta Part B 73 (2012) 7–12. 91 N. M. Shaikh, M. S. Kalhoro, A. Hussain, M. A. Baig, Spectroscopic study of a lead plasma produced by the 1064 nm, 532 nm and 355 nm of a Nd:YAG laser, Spectrochim. Acta Part B 88 (2013) 198–202. 92 T. Moscicki, J. Hoffman, Z. Szymanski, The effect of laser wavelength on laser-induced carbon plasma, J. Appl. Phys. 114 (2013) 083306. Introduction 114 93 V. Morel, A. Bultel, Theoretical study of the formation mechanism of laser-induced aluminum plasmas using Nd:YAG fundamental, second or third harmonics, Spectrochim. Acta Part B 94–95 (2014) 63–70. 94 Q. Wang, P. Jander, C. Fricke-Begemann, R. Noll, Comparison of 1064 nm and 266 nm excitation of laser-induced plasmas for several types of plastics and one explosive, Spectrochim. Acta Part B 63 (2008) 1011–1015. 95 N. S. Allen, Photochemistry and photophysics of polymers materials, Ch. 14; Photoablation of polymers materials, John Wiley and Sons, Inc.; New Jersey (2010). 96 S. Lazare, V. Granier, Ultraviolet laser photoablation of polymers: a review and recent results, Laser Chem. 10 (1989) 25–40. 97 E. Sutcliffe, R. Srinivasan, Dynamics of UV laser ablation of organic polymer surfaces, J. Appl. Phys. 60 (1986) 3315–3322. 98 G. D. Mahan, H. S. Cole, Y. S. Liu, H. R. Philipp, Theory of polymer ablation, Appl. Phys. Lett. 53 (1988) 2377–2379. 99 S. R. Cain, F. C. Burns, C. E. Otis, On single-photon ultraviolet ablation of polymeric materials, J. Appl. Phys. 71 (1992) 4107–4117. 100 S. R. Cain, A photothermal model for polymer ablation: chemical modification, J. Phys. Chem. 97 (1993) 7572–7577. 101 B. Luk’yanchuk, N. Bityurin, M. Himmelbauer, N. Arnold, UV-laser ablation of polyimide: from long to ultra-short laser pulses, Nucl. Instrum. Methods Phys. Res. B 122 (1997) 347–355. 102 V. Srinivasan, M. A. Smrtic, S. V. Babu, Excimer laser etching of polymers, J. Appl. Phys. 59 (1986) 3861–3867. 103 H. Schmidt, J. Ihlemann, B. Wolff-Rottke, K. Luther, J. Troe, Ultraviolet laser ablation of polymers: Spot size, pulse duration, and plume attenuation effects explained, J. Appl. Phys. 83 (1998) 5458–5468. 104 B. Luk’yanchuk, N. Bityurin, S. Anisimov, D. Bauerle, The role of excited species in UV-laser materials ablation - Part I: Photophysical ablation of organic polymers, Appl. Phys. A: Mater. Sci. Process. 57 (1993) 367–374. 105 B. Luk’yanchuk, N. Bityurin, S. Anisimov, N. Arnold, D. Bauerle, The role of excited species in ultravioletlaser materials ablation III. Non-stationary ablation of organic polymers, Appl. Phys. A: Mater. Sci. Process. 62 (1996) 397–401. 106 R. E. Russo, X. Mao, J. J. González, J. Yoo, Femtosecond vs. Nanosecond laser pulse duration for laser ablation chemical analysis, Spectroscopy 28 (2013) 24–35. 107 T. A. Labutin, V. N. Lednev, A. A. Ilyin, A. M. Popov, Femtosecond laser-induced breakdown spectroscopy, J. Anal. At. Spectrom. 31 (2016) 90–118. 108 M. D. Shirk, P. A. Molian, A review of ultrashort pulsed laser ablation of materials, J. Laser Appl. 10 (1998) 18–28. Introduction 115 109 J. B. Sirven, B. Bousquet, L. Canioni, L. Sarger, Time-resolved and time-integrated single-shot laserinduced plasma experiments using nanosecond and femtosecond laser pulses, Spectrochim. Acta Part B 59 (2004) 1033–1039. 110 J. R. Freeman, S. S. Harilal, P. K. Diwakar, B. Verhoff, A. Hassanein, Comparison of optical emission from nanosecond and femtosecond laser produced plasma in atmosphere and vacuum conditions, Spectrochim. Acta Part B 87 (2013) 43-50. 111 B. N. Chichkov, C. Momma, S. Nolte, F. von Alvensleben, A. Tünnermann, Femtosecond, picosecond and nanosecond laser ablation of solids, Appl. Phys. A, 63 (1996) 109-115. 112 M. Baudelet, L. Guyon, J. Yu, J. P. Wolf, T. Amodeo, E. Frejafon, P. Laloi, Femtosecond time-resolved laser-induced breakdown spectroscopy for detection and identification of bacteria: A comparison to the nanosecond regime, J. Appl. Phys. 99 (2006) 084701. 113 B. Verhoff, S. S. Harilal, J. Freeman, P. K. Diwakar, A. Hassanein, Dynamics of femtoand nanosecond laser ablation plumes investigated using optical emission spectroscopy, J. Appl. Phys. 112 (2012) 093303. 114 K. H. Leitz, B. Redlingshofer, Y. Reg, A. Otto, M. Schmidt, Metal ablation with short and ultrashort laser pulses, Phys. Proc. 12 (2011) 230–238. 115 G. M. Fuge, M. N. R. Ashfold, S. J. Henley, Studies of the plume emission during the femtosecond and nanosecond ablation of graphite in nitrogen, J. Appl. Phys. 99 (2006) 014309. 116 J. R. Scott, A. J. Effenberger Jr., J. J. Hatch, Influence of Atmospheric Pressure and Composition on LIBS, in: S. Musazzi, U. Perini (Eds.), Laser-induced breakdown spectroscopy, Springer-Verlag, Berlin (2014) pp. 91–116. 117 Z. Chen, D. Bleiner, A. Bogaerts, Effect of ambient pressure on laser ablation and plume expansion dynamics: A numerical simulation, J. Appl. Phys. 99 (2006) 063304. 118 S. Mehrabian, M. Aghaei, S. H. Tavassoli, Effect of background gas pressure and laser pulse intensity on laser induced plasma radiation of copper samples, Phys. Plasmas 17 (2010) 043301. 119 K. R. Chen, J. N. Leboeuf, R. F. Wood, D. B. Geohegan, J. M. Donato, C. L. Liu, A. A. Puretzky, Modeling of dynamical processes in laser ablation, Appl. Surf. Sci. 96–98 (1996) 14–23. 120 T. Delgado, J. M. Vadillo, J. J. Laserna, Pressure effects in laser-induced plasmas of trinitrotoluene and pyrene by laser-induced breakdown spectroscopy (LIBS), Appl. Spectrosc. 68 (2014) 33–38. 121 S. Yalçin, Y. Y. Tsui, R. Fedosejevs, Pressure dependence of emission intensity in femtosecond laserinduced breakdown spectroscopy, J. Anal. At. Spectrom. 19 (2004) 1295–1301. 122 N. Farid, S. S. Harilal, H. Ding, A. Hassanein, Emission features and expansion dynamics of nanosecond laser ablation plumes at different ambient pressures, J. Appl. Phys. 115 (2014) 033107. 123 A. J. Effenberger Jr., J. R. Scott, Effect of atmospheric conditions on LIBS spectra, Sensors 10 (2010) 4907–4925. Introduction 116 124 Y. I. Lee, K. Song, H. K. Cha, J. M. Lee, M. C. Park, G. H. Lee, J. Sneddon, Influence of atmosphere and irradiation wavelength on copper plasma emission induced by excimer and Q-switched Nd:YAG laser ablation, Appl. Spectrosc. 51 (1997) 959–964. 125 C. Colón, A. Alonso-Medina, Application of a laser produced plasma: Experimental Stark widths of single ionized lead lines, Spectrochim. Acta Part B. 61 (2006) 856–863. 126 N. Konjević, M. Ivković, S. Jovićević, Spectroscopic diagnostics of laser-induced plasmas, Spectrochim. Acta Part B. 65 (2010) 593–602. 127 Y. Iida, Effects of atmosphere on laser vaporization an excitation processes of solid samples, Spectrochim. Acta Part B 45 (1990) 1353–1367. 128 J. A. Aguilera, C. A. Aragón, Comparison of the temperatures and electron densities of laser-produced plasmas obtained in air, argon, and helium at atmospheric pressure, Appl. Phys. A Mater. 69 (1999) S475–S478. 129 E. Vors, C. Gallou, L. Salmon, Laser-induced breakdown spectroscopy of carbon in helium and nitrogen at high pressure, Spectrochim. Acta Part B 63 (2008) 1198–1204. 130 W. Sdorra, K. Niemax, Basic investigations for laser microanalysis. III. Application of different buffer gases for laser-produced sample plumes, Mikrochim. Acta 107 (1992) 319–327. 131 M. Kuzuya, H. Matsumoto, H. Takechi, O. Mikami, Effect of laser energy and atmosphere on the emission characteristics of laser-induced plasmas, Appl. Spectrosc. 47 (1993) 1659–1664. 132 H. H. Telle, A. González Ureña, R. J. Donovan, Laser chemistry: spectroscopy, dynamics and applications, John Wiley & Sons; Chichester (2007). 133 W. Koechner, Solid-state laser engineering, Springer Series in Optical Sciences, New York (2006). 134 D. Strickland, G. Mourou, Compression of amplified chirped optical pulses, Opt. Commun. 56 (1985) 219–221. 135 A. Dubietis, G. Jonusauskas, A. Piskarskas, Powerful femtosecond pulse generation by chirped and stretched pulse parametric amplification in BBO crystal, Opt. Commun. 88 (1992) 437–440. 136 S. Yefet, A. Pe’er, A review of cavity design for Kerr lens mode-locked solid-state lasers, Appl. Sci. 3 (2013) 694–724. 137 D. E. Spence, P. N. Kean, W. Sibbett, 60-fs Pulse generation from a self-mode-locked Ti:sapphire Laser, Opt. Lett. 16 (1991) 42–44. 138 M. Wollenhaupt, A. Assion, T. Baumert, Short and ultrashort laser pulses. Chapter 12 in Springer Handbook of lasers and optics, F. Träger (Ed.), Springer, Berlin (2012) pp 1047–1094. 139 W. M. Steen, J. Mazumder, Laser material processing, 4th Edition, Springer, London (2010). 140 C. E. Webb, J. D. C. Jones, Handbook of laser technology and applications: Applications, Taylor & Francis; UK (2003). Introduction 117 141 H. Becker-Ross, S. V. Florek, Echelle spectrometers and charge-coupled devices, Spectrochim. Acta Part B 52 (1997) 1367–1375 142 C. Pasquini, J. Cortez, L. M. C. Silva, F. B. Gonzaga, Laser induced breakdown spectroscopy, J. Braz. Chem. Soc. 18 (2007) 463–512. 143 S. Musazzi, U. Perini, LIBS instrumental techniques, in: S. Musazzi, U. Perini (Eds.), Laser-Induced breakdown spectroscopy, Springer-Verlag, Berlin (2014) pp. 59-89. 144 S. Laville, M. Sabsabi, F. R. Doucet, Multi-elemental analysis of solidified mineral melt samples by laserinduced breakdown spectroscopy coupled with a linear multivariate calibration, Spectrochim. Acta Part B 62 (2007) 1557–1566. 145 D. L. Death, A. P. Cunningham, L. J. Pollard, Multi-element analysis of iron ore pellets by laser-induced breakdown spectroscopy and principal components regression, Spectrochim. Acta Part B 63 (2008) 763–769. 146 S. Yao, J. Lu, J. Li, K. Chen, J. Li, M. Dong, Multi-elemental analysis of fertilizer using laser-induced breakdown spectroscopy coupled with partial least squares regression, J. Anal. At. Spectrom. 25 (2010) 1733–1738. 147 J. M. Anzano, C. Bello-Gálvez, R. J. Lasheras, Identification of polymers by means of LIBS, in: S. Musazzi, U. Perini (Eds.), Laser-Induced breakdown spectroscopy, Springer-Verlag, Berlin (2014) pp. 421–438. 148 S. Morel, N. Leone, P. Adama, J. Amouroux, Detection of bacteria by time-resolved laser-induced breakdown spectroscopy, Appl. Optics 42 (2003) 6184–6191. 149 M. Baudelet, J. Yu, M. Bossu, J. Jovelet, J. P. Wolf, Discrimination of microbiological samples using femtosecond laser-induced breakdown spectroscopy, Appl. Phys. Letters 89 (2006) 163903. 150 S. J. Rehse, N. Jeyasingham, J. Diedrich, S. Palchaudhuri, A membrane basis for bacterial identification and discrimination using laser-induced breakdown spectroscopy, J. Appl. Phys. 105 (2009) 102034. 151 D. Marcos-Martínez, J. A. Ayala, R. C. Izquierdo-Hornillos, F. J. M De Villena, J. O. Cáceres, Identification and discrimination of bacterial strains by laser induced breakdown spectroscopy and neural networks, Talanta 84 (2011) 730–737. 152 S. Manzoor, S. Moncayo, F. Navarro-Villoslada, J. A. Ayala, R. Izquierdo-Hornillos, F. J. Manuel de Villena, J. O. Cáceres, Rapid identification and discrimination of bacterial strains by laser induced breakdown spectroscopy and neural networks, Talanta 121 (2014) 65–70. 153 R. Srinivasan, W. J. Leigh, Ablative photodecomposition: action of far-ultraviolet (193 nm) laser radiation on poly (ethylene terephthalate) films, J. Am. Chem. Soc. 104 (1982) 6784–6785. 154 Y. Iida, E. S. Yeung, Optical monitoring of laser-induced plasma derived from graphite and characterization of the deposited carbon film, Appl. Spectrosc. 48 (1994) 945–950. 155 S. Acquaviva, M. L. Giorgi, Temporal and spatial analysis of plasmas during graphite laser ablation in low-pressure N2, Appl. Surf. Sci. 197-198 (2002) 21–26. Introduction 118 156 S. Amoruso, G. Ausanio, M. Vitiello, X. Wang, Infrared femtosecond laser ablation of graphite in high vacuum probed by optical emission spectroscopy, Appl. Phys. A 81 (2005) 981–986. 157 K. F. Al-Shboul, S. S. Harilal, A. Hassanein, Spatio-temporal mapping of ablated species in ultrafast laser-produced graphite plasmas, Appl. Phys. Lett. 100 (2012) 221106. 158 A. Portnov, S. Rosenwaks, I. Bar, Emission following laser-induced breakdown spectroscopy of organic compounds in ambient air, Appl. Optics 42 (2003) 2835–2842. 159 A. Portnov, S. Rosenwaks, I. Bar, Identification of organic compounds in ambient air via characteristic emission following laser ablation, J. Lumin. 102 (2003) 408–413. 160 R. K. Thareja, R. K. Dwivedi, K. Ebihara, Interaction of ambient nitrogen gas and laser ablated carbon plume: Formation of CN, Nucl. Instrum. Meth. B 192 (2002) 301–310. 161 M. Tran, Q. Sun, B. W. Smith, J. D. Winefordner, Determination of C:H:O:N ratios in solid organic compounds by laser-induced plasma spectroscopy, J. Anal. At. Spectrom. 16 (2001) 628–632. 162 E. A. P. Cheng, R. D. Fraser, J. G. Eden, Detection of trace concentrations of column III and V hydrides by laser-induced breakdown spectroscopy, Appl. Spectrosc. 45 (1991) 949–952. 163 L. Dudragne, Ph. Adam, J. Amouroux, Time-resolved laser-induced breakdown spectroscopy: application for qualitative and quantitative detection of fluorine, chlorine, sulfur, and carbon in air, Appl. Spectrosc. 52 (1998) 1321–1327. 164 M. Tran, Q. Sun, B. W. Smith, J. D. Winefordner, Determination of F, Cl, and Br in solid organic compounds by laser-induced plasma spectroscopy, Appl. Spectrosc. 55 (2001) 739–744. 165 S. Kaski, H. Häkkänen, J. Korppi-Tommola, Determination of Cl/C and Br/C ratios in pure organic solids using laser-induced plasma spectroscopy in near vacuum ultraviolet, J. Anal. At. Spectrom. 19 (2004) 474–478. 166 R. B. King, Relative transition probabilities of the Swan bands of carbon, Astrophys. J. 108 (1948) 429– 433. 167 L. L. Danylewych, R. W. Nicholls, Intensity measurements on the C2 (d3Πg-a3Πu) Swan band system. I. Intercept and partial band methods, Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences 339 (1974) 197–212. 168 W. Weltner Jr., R. J. Van Zee, Carbon molecules, ions, and clusters, Chem. Rev. 89 (1989) 1713–1747. 169 S. Pellerin, K. Musiol, O. Motret, B. Pokrzywka, J. Chapelle, Application of the (0,0) Swan band spectrum for temperature measurements, J. Phys. D: Appl. Phys. 29 (1996) 2850–2865. 170 S. Acquaviva, Simulation of emission molecular spectra by a semi-automatic program package: the case of C2 and CN diatomic molecules emitting during laser ablation of a graphite target in nitrogen environment, Spectrochim. Acta Part A, 60 (2004) 2079–2086. 171 J. O. Hornkohl, C. Parigger, J. W. L. lewis, Temperature measurements from CN spectra in a laserinduced plasma, J. Quant. Spectrosc. Radiat. Transfer, 46 (1991) 405–411. Introduction 119 172 S. S. Harilal, R. C. Issac, C. V. Bindhu, P. Gopinath, V. P. N. Nampoori, C. P. G. Vallabhan, Time resolved study of CN band emission from plasma generated by laser irradiation of graphite, Spectrochim. Acta Part A 53 (1997) 1527–1536. 173 B. Német, K. Musiol, I. Sánta, J. Zachorowski, Time-resolved vibrational and rotational emission analysis of laser-produced plasma of carbon and polymers, J. Mol. Struct. 511–512 (1999) 259–270. 174 J. J. Camacho, L. Díaz, J. P. Cid, J. M. L. Poyato, Time-resolved spectroscopic diagnostic of the CO2 plasma induced by a high-power CO2 pulsed laser, Spectrochim. Acta Part B 66 (2011) 698–705. 175 T. Delgado, J. M. Vadillo, J. J. Laserna, Laser-induced plasma spectroscopy of organic compounds. Understanding fragmentation processes using ion-photon coincidence measurements, J. Anal. Atom. Spectrom. 28 (2013) 1377–1384. 176 M. Civiš, S. Civiš, K. Sovová, K. Dryahina, P. Španěl, M. Kyncl, Laser Ablation of FOX-7: Proposed mechanism of decomposition, Anal. Chem. 83 (2011) 1069–1077. 177 L. Wallace, Band-head wavelengths of C2, CH, CN, CO, NH, NO, O2, OH, and their ions, Astrophys. J. Suppl. S. 7 (1962) 165–290. 178 M. Dong, G. C.Y. Chan, X. Mao, J. J. Gonzalez, J. Lu, R. E. Russo, Elucidation of C2 and CN formation mechanisms in laser-induced plasmas through correlation analysis of carbon isotopic ratio, Spectrochim. Acta Part B 100 (2014) 62–69. 179 R. Glaus, J. Riedel, I. Gornushkin, Insight into the formation of molecular species in laser-induced plasma of isotopically labeled organic samples, Anal. Chem. 87 (2015) 10131–10137. 180 T. Delgado, J. M. Vadillo, J. J. Laserna, Primary and recombined emitting species in laser-induced plasmas of organic explosives in controlled atmospheres, J. Anal. At. Spectrom. 29 (2014) 1675–1685. 181 S. S. Harilal, R. C. Issac, C. V. Bindhu, V. P. N. Nampoori, C. P. G. Vallabhan, Spatial analysis of C2 band emission from laser produced plasma, Plasma Sources Sci. Technol. 6 (1997) 317–322. 182 M. W. Slack, Kinetics and thermodynamics of the CN molecule. III. Shocktube measurements of CN dissociation rates, J. Chem. Phys. 64 (1976) 228–236. 183 T. Kruse, P. Roth, Kinetics of C2 reactions during high-temperature pyrolysis of acetylene, J. Phys. Chem. A 101 (1997) 2138–2146. 184 S. Zhang, X. Wang, M. He, Y. Jiang, B. Zhang, W. Hang, B. Huang, Laser-induced plasma temperature, Spectrochim. Acta Part B 97 (2014) 13–33. 185 http://people.chem.umass.edu/botch/Chem122S08/Chapters/Ch6/BondEnthalpy/BondEnthalpy.html 186 B. de B. Darwent, Bond dissociation energies in simple molecules, http://www.nist.gov/data/nsrds/NSRDS-NBS31.pdf. 187 C. Vivien, J. Hermann, A. Perrone, C. Boulmer-Leborgne, A. Luches, A study of molecule formation during laser ablation of graphite in low-pressure nitrogen, J. Phys. D: Appl. Phys. 31 (1998) 1263–1272. 188 S. Abdelli-Messaci, T. Kerdja, A. Bendib, S. Malek, Emission study of C2 and CN in laser created carbon plasma under nitrogen environment, J. Phys. D: Appl. Phys. 35 (2002) 2772–2778. Introduction 120 189 M. Baudelet, L. Guyon, J. Yu, J. P. Wolf, T. Amodeo, E. Fréjafon, P. Laloi, Spectral signature of native CN bonds for bacterium detection and identification using femtosecond laser-induced breakdown spectroscopy, Appl. Phys. Lett. 88 (2006) 063901. 190 N. A. Zakorina, G. S. Lazeeva, A. A. Petrov, Spectroscopic determination of the isotopic composition of boron trifluoride, J. Nucl. Energy 21 (1967) 309–313. 191 G. S. Lazeeva, V. M. Nemets, A. A. Petrov, Spectral-isotopic method of determination of gas-forming elements in organic and inorganic substances, Spectrochim. Acta Part B 36 (1981) 1233–1242. 192 R. E. Russo, A. A. Bol'shakov, X. Mao, C. P. McKay, D. L. Perry, O. Sorkhabi, Laser ablation molecular isotopic spectrometry, Spectrochim. Acta Part B 66 (2011) 99–104. 193 A. A. Bol'shakov, X. Mao, J. J. Gonzalez, R. E. Russo, Laser ablation molecular isotopic spectrometry (LAMIS): current state of the art, J. Anal. At. Spectrom. 31 (2016) 119–134. 194 C. R. Burnet, Isotope shift in the atomic spectrum of carbon, Phys. Rev., 80 (1950) 494. 195 J. P. Nicklas, Isotope shift in neutral carbon, Phys. Rev., 95 (1954) 1469–1471. 196 K. H. Kurniawan, T. J. Lie, M. M. Suliyanti, R. Hedwig, M. Pardede, D. P. Kurniawan, Quantitative analysis of deuterium using laser-induced plasma at low pressure of helium, Anal. Chem. 78 (2006) 5768–5773. 197 J. R. Holmes, Isotope shift in some lines of nitrogen, Phys. Rev. 63 (1942) 41–46. 198 L. W. Parker, J. R. Holmes, Isotope Shift in Neutral Oxygen, J. Opt. Soc. Am., 43 (1953) 103–109. 199 M. Dong, X. Mao, J. J. Gonzalez, J. Lu, R. E. Russo, Carbon isotope separation and molecular formation in laser-induced plasmas by laser ablation molecular isotopic spectrometry, Anal. Chem. 85 (2013) 2899–2906. 200 A. Sarkar, X. Mao, G. C. Y. Chan, R. E. Russo, Laser ablation molecular isotopic spectrometry of water for 1D2/1H1 ratio analysis, Spectrochim. Acta Part B 88 (2013) 46–53. 201 P. Bollmark, I. Kopp, B. Rydh, Rotational analysis of the ND A3πi-X3Σ- (0,0) band, J. Mol. Spectrosc. 34 (1970) 487–499. 202 M. Zachwieja, W. Szajna, R. Hakalla, The A2Δ–X2π band system of the CD radical, J. Mol. Spectrosc. 275 (2012) 53–60. 203 A. A. Bol′shakov, X. Mao, J. Jain, D. L. McIntyre, R. E. Russo, Laser ablation molecular isotopic spectrometry of carbon isotopes, Spectrochim. Acta Part B 113 (2015) 106–112. 204 H. Hou, G. C.Y. Chan, X. Mao, V. Zorba, R. Zheng, R. E. Russo, Femtosecond laser ablation molecular isotopic spectrometry for zirconium isotope analysis, Anal. Chem. 87 (2015) 4788–4796. 205. X. Mao, A. A. Bol'shakov, D. L. Perry, O. Sorkhabi, R. E. Russo, Laser ablation molecular isotopic spectrometry: parameter influence on boron isotope measurements, Spectrochim. Acta Part B 66 (2011) 604–609. 206 X. Mao, A. A. Bol'shakov, I. Choi, C. P. McKay, D. L. Perry, O. Sorkhabi, R. E. Russo, Laser ablation molecular isotopic spectrometry: strontium and its isotopes, Spectrochim. Acta Part B 66 (2011) 767– 775. 207 A. K. Jain, R. P. W. Duin, J. Mao, Statistical pattern recognition: A review, IEEE Transactions on Pattern Analysis and Machine Intelligence 22 (2000) pp. 4–37. Introduction 121 208 A. Ferrero, P. Lucena, R. G. Herrera, A. Doña, R. Fernández-Reyes, J. J. Laserna, Libraries for spectrum identification: Method of normalized coordinates versus linear correlation, Spectrochim. Acta Part B 63 (2008) 383–388. 209 J. B. Sirven, B. Sallé, P. Mauchien, J. L. Lacour, S. Maurice, G. Manhès, Feasibility study of rock identification at the surface of Mars by remote laser-induced breakdown spectroscopy and three chemometric methods, J. Anal. At. Spectrom. 22 (2007) 1471–1480. 210 M. J. C. Pontes, J. Cortez, R. K. H. Galvão, C. Pasquini, M. C. U. Araújo, R. M. Coelho, M. K. Chiba, M. F de Abreu, B. E. Madari, Classification of Brazilian soils by using LIBS and variable selection in the wavelet domain, Anal. Chim. Acta 642 (2009) 12–18. 211 J. L. Gottfried, F. C. De Lucia Jr., A. W. Miziolek, Discrimination of explosive residues on organic and inorganic substrates using laser-induced breakdown spectroscopy, J. Anal. At. Spectrom. 24 (2009) 288–296. 212 F-Y. Yueh, H. Zheng, J. P. Singh, S. Burgess, Preliminary evaluation of laser-induced breakdown spectroscopy for tissue classification, Spectrochim. Acta Part B 64 (2009) 1059–1067. 213 M. Boueri, V. Motto-Ros, W. Q. Lei, Q. L. Ma, L. J. Zheng, H. P. Zeng, J. Yu, Identification of polymer materials using laser-induced breakdown spectroscopy combined with artificial neural networks, Appl. Spectrosc. 65 (2011) 307–314. 214 S. Moncayo, S. Manzoor, F. Navarro-Villoslada, J. O. Cáceres, Evaluation of supervised chemometric methods for sample classification by laser induced breakdown spectroscopy, Chemometr. Intell. Lab. 146 (2015) 354–364. 215 M. J. Adams, Chemometrics in Analytical Spectroscopy, 2nd Edition, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 OWF, UK (2004). 216 J. Yinon, Counterterrorist Detection Techniques of Explosives, Amsterdam, The Netherland (2007). 217 S. Singh, Sensors-An effective approach for the detection of explosives, J. Hazard Mater. 144 (2007) 15–28. 218 C. López-Moreno, S. Palanco, J. J. Laserna, F. De Lucia Jr., A. W. Miziolek, J. Rose, R. A. Walters, A. I. Whitehouse, Test of a stand-off laser-induced breakdown spectroscopy sensor for the detection of explosive residues on solid surfaces, J. Anal. At. Spectrom. 21 (2006) 55–60. 219 J. Moros, F. J. Fortes, J. M. Vadillo, J. J. Laserna, LIBS detection of explosives in traces, in: S. Musazzi, U. Perini (Eds.), Laser-induced breakdown spectroscopy, Springer-Verlag, Berlin (2014) pp. 349–375. 220 P. Lucena, I. Gaona, J. Moros, J. J. Laserna, Location and detection of explosive-contaminated human fingerprints on distant targets using standoff laser-induced breakdown spectroscopy, Spectrochim. Acta Part B 85 (2013) 71–77. 221 M. I. Gaona Fernández, (2014) Standoff sensing technology based on laser-induced breakdown spectroscopy: advanced targeting, surveillance and reconnaissance in security and architectural heritage applications. Unpublished PhD thesis. University of Málaga. Retrieved from the institutional repository of the University of Málaga at https://riuma.uma.es/xmlui/handle/10630/8213. 222 F. C. De Lucia Jr., J. L. Gottfried, C. A. Munson, A. W. Miziolek, Double pulse laser-induced breakdown spectroscopy of explosives: Initial study towards improved discrimination, Spectrochim. Acta Part B 62 (2007) 1399–1404. Chapter 1 129 Sensing Signatures Mediated by Chemical Structure of Molecular Solids in Laser-Induced Plasmas Jorge Serrano, Javier Moros, and J. Javier Laserna* Universidad de Ma laga, Departamento de Química Analítica, 29071 Ma laga, Espana ABSTRACT: Laser ablation of organic compounds has been investigated for almost 30 years now, either in the framework of pulse laser deposition for the assembling of new materials or in the context of chemical sensing. Various monitoring techniques such as atomic and molecular fluorescence, timeof-flight mass spectrometry, and optical emission spectroscopy have been used for plasma diagnostics in an attempt to understand the spectral signature and potential origin of gasphase ions and fragments from organic plasmas. Photochemical and photophysical processes occurring within these systems are generally much more complex than those suggested by observation of optical emission features. Together with laser ablation parameters, the structural and chemical−physical properties of molecules seem to be closely tied to the observed phenomena. The present manuscript, for the first time, discusses the role of molecular structure in the optical emission of organic plasmas. Factors altering the electronic distribution within the organic molecule have been found to have a direct impact on its ensuing optical emissions. The electron structure of an organic molecule, resulting from the presence, nature, and position of its atoms, governs the breakage of the molecule and, as a result, determines the extent of atomization and fragmentation that has proved to directly impact the emissions of CN radicals and C2dimers. Particular properties of the molecule respond more positively depending on the laser irradiation wavelength, thereby redirecting the ablation process through photochemical or photothermal decomposition pathways. It is of paramount significance for chemical identification purposes how, despite the large energy stored and dissipated by the plasma and the considerable number of transient species formed, the emissions observed never lose sight of the original molecule. Received: November 7, 2014 Accepted: February 10, 2015 Published: February 10, 2015 Article pubs.acs.org/ac © 2015 American Chemical Society DOI: 10.1021/acs.analchem.5b00212 Anal. Chem. 2015, 87, 2794−2801 137 Chapter 2 Exploring the formation routes of diatomic hydrogenated radicals using fs laser-induced breakdown spectroscopy of deuterated molecular solids Chapter 2 139 Exploring the formation routes of diatomic hydrogenated radicals using femtosecond laserinduced breakdown spectroscopy of deuterated molecular solids Jorge Serrano, Javier Moros and J. Javier Laserna* In recent years, laser-induced breakdown spectroscopy (LIBS) has expanded beyond multielemental analysis capability by exploring molecular solids and optical emissions from diatomic radicals. Despite many efforts to this end, the greatest emphasis has been placed on the study of the dynamics and the optical and temporal behavior of CN and C 2 emissions in order to elucidate the main chemical reactions occurring in the plasma. Other diatomic species, whose emission bands also appear in the visible and near-ultraviolet regions of the LIBS spectrum, remain virtually unexplored. This research focuses, for the first time, on elucidating the formation pathways of CH, NH, and OH radicals in femtosecond laserproduced plasmas of molecular solids. As the origin of diatomic species is not easily traceable to a single formation route due to the occurrence of primary and secondary processes within the plasma plume, isotopic labeling of some molecules has been used as a diagnostic tool. Deuterated isotopologues of urea, terephthalic acid and anthracene have been evaluated. The dominant routes have been identified through a comparison of the observed isotopic patterns. The findings reveal the direct release of native molecular bonds as a significant source that populates femtosecond laser-produced plasmas with NH radicals. Additional reactions of native hydrogen from molecules and atmospheric nitrogen also contribute to form such species. Oxygen from the atmosphere competes with nitrogen in these reactions to populate plasmas with OH radicals. Alternatively, the direct connection between C 2 and CH emissions verify that this radical derives from diatomic carbon fragments as precursors. These new findings entail a further step towards the use of molecular emissions as diagnostic tools in the analytical applications of laser-induced plasmas of organic compounds. Received 25th May 2015 Accepted 25th September 2015 DOI: 10.1039/c5ja00192g www.rsc.org/jaas This journal is © The Royal Society of Chemistry 2015 J. Anal. At. Spectrom.,2015,30, 2343–2352 JAAS PAPER View Article Online View Journal | View Issue 149 Chapter 3 Molecular signatures in femtosecond laser-induced organic plasmas: comparison with nanosecond laser ablation Chapter 3 151 Phys. Chem. Chem. Phys., 2016, 18, 2398-2408 This journal is ©the Owner Societies 2016 Cite this: Phys.Chem. Chem. Phys., 2016, 18,2398 Molecular signatures in femtosecond laser-induced organic plasmas: comparison with nanosecond laser ablation Jorge Serrano, Javier Moros and J. Javier Laserna* During the last few years, laser-induced breakdown spectroscopy (LIBS) has evolved significantly in the molecular sensing area through the optical monitoring of emissions from organic plasmas. Large efforts have been made to study the formation pathways of diatomic radicals as well as their connections with the bonding framework of molecular solids. Together with the structural and chemical–physical properties of molecules, laser ablation parameters seem to be closely tied to the observed spectral signatures. This research focuses on evaluating the impact of laser pulse duration on the production of diatomic species that populate plasmas of organic materials. Differences in relative intensities of spectral signatures from the plasmas of several organic molecules induced in femtosecond (fs) and nanosecond (ns) ablation regimes have been studied. Beyond the abundance and origin of diatomic radicals that seed the plasma, findings reveal the crucial role of the ablation regime in the breakage pattern of the molecule. The laser pulse duration dictates the fragments and atoms resulting from the vaporized molecules, promoting some formation routes at the expense of other paths. The larger amount of fragments formed by fs pulses advocates a direct release of native bonds and a subsequent seeding of the plasma with diatomic species. In contrast, in the ns ablation regime, the atomic recombinations and single displacement processes dominate the contribution to diatomic radicals, as long as atomization of molecules prevails over their progressive decomposition. Consequently, fs-LIBS better reflects correlations between strengths of emissions from diatomic species and molecular structure as compared to ns-LIBS. These new results entail a further step towards the specificity in the analysis of molecular solids by fs-LIBS. Received 23rd October 2015 Accepted 27th November 2015 DOI: 10.1039/c5cp06456b www.rsc.org/pccp PCCP PAPER 163 Chapter 4 Potential of laser-induced breakdown spectroscopy for discrimination of nano-sized carbon materials. Insights on the optical characterization of graphene