Environmental assessment of a bottom-up hydrolytic synthesis of TiO₂ nanoparticles

Abstract

A green metrics evaluation of the bottom-up hydrolytic sol–gel synthesis of titanium dioxide (TiO₂) nanoparticles has been performed by following two different approaches, namely, EATOS software and LCA methodology. Indeed, the importance of engineered nanomaterials is increasing worldwide in many high-technological applications. Due to the as yet completely un-established environment and human health impact of nano-sized materials, the possibility of at least choosing a greener synthetic strategy through an accurate comparison of detailed environmental assessments will soon be of absolute importance in both the small and large scale production of these advanced inorganic materials. The present LCA study has been carried out following an ecodesign approach, in order to limit the environmental impacts and protect human health. The results of LCA analysis suggest that the highest environmental impact is mainly due to energy and the titanium isopropoxide precursor used in the synthesis process. Concurrently, software EATOS has been employed to calculate the environmental parameters that account for the environmental and social costs related to all the chemicals involved in the analyzed synthesis. As the EATOS approach is based purely on synthetic chemical mechanism considerations, thus neglecting any energy contributions, and its results cannot be directly compared to those arising from LCA analysis. However, similar and comparable outcomes are obtained by simply neglecting the energy contributions, broadening the application fields of the combined EATOS-LCA approach to the inorganic synthesis of engineered nanomaterials, highlighting the great potential of their synergy.

---

1. Introduction

Several evaluating parameters related to the environmental and human health impacts of a particular chemical process are gaining increasing interest and consideration alongside the traditionally employed ones, such as the yield, time and cost, as demanded by Green Chemistry, Green Engineering and Process Intensification developing philosophies.^1–5

At these latter regarding different metrics have been proposed over the last few decades,^6,7 among which the E-factor⁸ and the mass index (MI),^9,10 which are expressed respectively as eqn (1) and (2), have been the most studied and applied, leading to the design and development of the software, EATOS (Environmental Assessment Tool for Organic Syntheses) by Eissen and Metzger.¹¹

Indeed, this free of charge software¹² allows the utilization of easily available data for calculations of the abovementioned mass metrics, and has been applied to several studies because it also allows a comparison of different synthetic strategies for the obtainment of a particular target compound.^11,13–17

The most significant limitation of every mass metric, including EATOS, is the lack of any energy analysis together with its intrinsic character of being usually devoted to the gate-to-gate boundaries of a particular research laboratory or manufacturing plant. This means that several fundamental steps into the life cycle of employed chemicals, such as the extraction of raw materials, production, transportation, sales, distribution, use, and their final fate, are not considered by such mass metrics.^7,18

All of these latter shortcomings can be overcome by applying the Life Cycle Assessment (LCA) methodology, which is based on a cradle-to-grave approach.¹⁹ Owing to the intrinsic highly comprehensive nature as well as to the difficulty in finding the necessary inventory data for the LCA, its use has been mainly limited to large scale production processes, instead of the early stages of research for innovative and greener synthetic routes, in which EATOS has already demonstrated its applicability.¹⁵ Nevertheless, numerous strategies have been proposed to simplify the LCA approach thus rendering it easily applicable to a laboratory scale. Among these artifices, for example, the possibility of substituting chemicals that are absent into the LCA software database with some analogues has been reported.^20,21 In particular, in those works, different synthetic strategies have been evaluated and compared using both EATOS and LCA approaches. Noteworthy is the fact that similar assessments resulted, even though they are based two significantly different approaches.

Although the diffusion of the green chemistry developing philosophy usually involves less environmentally-friendly organic reactions, the significant environmental impact that nanotechnology related research activities are increasingly engendering should be considered, particularly considering that the environmental and human health effects of nano-sized materials are not yet fully established. In this last perspective, among the key issues related to the minimisation of the impact of nanotechnology on the environmental and on human health, which have been recently identified by Albrecht et al.²² a life cycle assessment and green chemistry metrics have been reported to be essential.

Therefore, the possibility of a greener approach for the synthesis of target engineered nanomaterials (being understood the desired particles size and shape) should be strongly recommended in order to pursue increasingly sustainable development.

LCA studies related to different nanomaterials have recently been reviewed²³ including also a few studies necessarily related to titanium dioxide,^24–27 which is the most studied and applied semiconductor and photocatalyst, owing to its unique physicochemical properties.^28,29 In general, the preparation routes for this material produce anatase, rutile, or an amorphous solid, depending on the experimental synthetic conditions employed. Nanosized anatase is the most attractive crystalline form of titanium dioxide for advanced and high-technological applications mainly because of its higher stability and photocatalytic activity than rutile and brookite.

Consequently, the obtainment of anatase nanoparticles with high purity and a precisely controlled structure and particle size is the main purpose of optimizing synthetic methods, and many approaches have been proposed, including inert gas condensation,³⁰ sol–gel method^31,32 and hydrothermal synthesis.³³ Among these, the sol–gel technique is most frequently applied for the synthesis of anatase nanoparticles with sizes ranging from 5 nm to several micrometers and a large variety of crystal shapes.

Aim of the present work was to apply two different green metrics evaluation tools, namely EATOS software and a detailed LCA study, to the hydrolytic sol–gel synthesis of TiO₂ nanoparticles, produced according to the method patented by one of the most important Italian companies, a supplier of chemicals for building and further industrial sectors,^34,37 and following an ecodesign approach. The obtained results were then compared to evaluate the applicability of these combined approaches to the inorganic synthesis of engineered nanomaterials.

Indeed, similar to what has recently begun to be accepted by a part of the organic chemistry community, an environmental impact evaluation should always accompany any new synthetic strategy proposed for the synthesis of inorganic engineered nanomaterials, and the same should also be valid for already established and recognized preparation procedures with the present manuscript representing a pioneering work in that precise direction. Being the recently developed synthetic approach for the obtainment of engineered nanomaterials, based on a meticulous reaction mechanism (thus no longer based on prolonged solid state diffusion-controlled high temperature treatments), it should not be surprising that an environmental assessment of their chemical synthesis can be, as in the case of the present work, evaluated using EATOS software, i.e., a tool originally developed for organic syntheses.

2. Experimental

The environmental performance of the hydrolytic synthesis of anatase TiO₂ nanoparticles was evaluated using the LCA methodology (according to the ISO 14040/44)^35,36 and EATOS tool, according to the reaction mechanism considerations reported in the following paragraph.

2.1. Hydrolytic sol–gel synthesis of TiO₂ nanoparticles

The hydrolytic sol–gel synthesis of anatase TiO₂ nanoparticles was performed according to the procedure recently patented and actually employed by Colorobbia S.p.A., one of the most important Italian suppliers of chemicals for the building and industrial sectors for the preparation of aqueous TiO₂ suspensions.^34,37 It is well known^37,38 that in the hydrolytic sol–gel syntheses of oxide nanomaterials, water acts both as a solvent and as a true reactant, with the reaction mechanism involving subsequent hydrolysis and condensation reactions. All of the possible hydrolytic reactions to which the organic metal precursors can undergo are reported in eqn (3)–(6), in which titanium isopropoxide is considered a titanium precursor. Eqn (7) can alternatively be used to summarise and group all the previous reactions.

Ti(OiPr)₄ + H₂O → Ti(OiPr)₃OH + iPrOH (3)

Ti(OiPr)₃OH + H₂O → Ti(OiPr)₂(OH)₂ + iPrOH (4)

Ti(OiPr)₂(OH)₂ + H₂O → TiOiPr(OH)₃ + iPrOH (5)

TiOiPr(OH)₃ + H₂O → Ti(OH)₄ + iPrOH (6)

Ti(OiPr)₄ + nH₂O → Ti(OiPr)_4−n(OH)_n + niPrOH (7)

The next condensation reactions can obviously occur between both hydroxylated metal species, leading to a Ti–O–Ti bond with the release of a water molecule (eqn (8)), and between a hydroxide and an alkoxide leading to the formation of a Ti–O–Ti bond with the release of an isopropyl alcohol molecule (eqn (9)).

≡Ti–OH + HO–Ti≡ → ≡Ti–O–Ti≡ + H₂O (8)

≡Ti–OiPr + HO–Ti≡ → ≡Ti–O–Ti≡ + iPrOH (9)

The chemical reactivity of the metal alkoxide towards hydrolysis and condensation mainly depends on the electrophilic nature of the metal atom, its ability to increase the coordination number, the steric hindrance of the alkoxy group, as well as the polarity, the dipole moment and the acidity of the solvent.³⁹ Therefore, the major problem of aqueous sol–gel methods based on the hydrolysis and condensation of a molecular precursor is control over the reactions rates. For most transition metal oxide precursors, these reactions are too fast, resulting in the loss of morphological and structural control over the final oxide material. One of the possibilities of decreasing and adjusting the reactivity of the precursor is the use of organic additives, which can act as chelating ligands that would modify the reactivity of the precursors.

In this specific case, for each Ti–O–Ti bridge formed, a water or an alcohol molecule is generated, and considering the product TiO₂ as the repetitive unit of the inorganic network and taking into account the octahedral coordination of Ti atoms and the trigonal-planar one of each oxygen atom, the stoichiometric reaction describing the overall hydrolytic sol–gel synthesis of TiO₂ nanoparticles can be expressed as eqn (10).

Ti(OiPr)₄ + 2H₂O → TiO₂ + 4iPrOH (10)

The experimental synthetic procedure followed³⁷ exploited the use of concentrated nitric acid and Triton X-100 to control the alkoxide reactivity as well as the hydrolysis and condensation reactions rates; the latter also acting as dispersant agent for the as synthesised nanocrystals. The previous reaction (eqn 10), however, assumes that the complete hydrolysis of the metal precursor occurs first, leading to the formation of four molecules of isopropyl alcohol, with subsequent condensation reactions occurring only between the tetra hydroxylated species, leading to the liberation of two water molecules for each TiO₂ repetitive unit formed.

To be as accurate as possible when comparing the two green metrics evaluation procedures (i.e. EATOS and LCA), eqn (11) was considered. This is slightly different from eqn (10) because eqn (11) considers the water in the reaction products.

Ti(OiPr)₄ + 4H₂O → TiO₂ + 2H₂O + 4iPrOH (11)

2.2. Description of the life cycle of the bottom-up hydrolytic synthesis of TiO₂ nanoparticles

Fig. 1 shows the production process of the considered bottom-up hydrolytic synthesis of a TiO₂ nanoparticles suspension. In particular, the process consists of two main steps: the sol preparation and the addition of auxiliary materials to prevent the flocculation of titanium dioxide. Because titanium isopropoxide, which is required for the sol preparation, has been not present in LCA databases, the representation of its synthesis route from titanium tetrachloride has been included in the present LCA study.⁴³ The same also applies to the synthesis of the TiCl₄ precursor, which a further step backward to its preparation from ilmenite.⁴³

Fig. 1 Bottom-up hydrolytic synthesis of the TiO₂ nanoparticles life cycle.

2.2.1. Sol preparation. The chemicals used to produce anatase TiO₂ nanoparticles suspension are listed in Table 1. Titanium isopropoxide (TIP) and water were introduced in a closed 200 l volume glass reactor with nitric acid (HNO₃) as a catalyst. This mixture was vigorously stirred for 16 hours at 80 °C, until it gradually transformed into a translucent sol by peptization.

Table 1 Percent composition of chemicals used to produce the nano-TiO₂ suspension (main product) and co-products

Chemicals	wt%
Titanium isopropoxide	23.22
Water	73.40
Nitric acid 63%	2.38
Triton X-100	1
Total	100
Recycled isopropanol (co-product)	12
Remaining 88%	%
TiO2 nanoparticles suspension (main product)	85.71
H2O (co-product)	14.29
Total	100

---

2.2.2. Addition of auxiliary material. After sol production, TX-100 was added to the colloidal suspension to avoid flocculation of the obtained nanoparticles. After 8 hours stirring at 25 °C, anatase nanoparticles were generated, together with isopropyl alcohol and water as co-products (according also to eqn (11)).

2.2.3. Maintenance operations. This study evaluated the maintenance operations at the end of the process, particularly the cleaning the reactor with water. It has been assumed that 1% of the TiO₂ nanoparticles suspension remains on the reactor walls, so that the water used to clean the reactor is contaminated by nanoparticles. A reverse osmosis filtration process was used to purify the as-contaminated water.

2.2.4. Energy consumption. In the sol preparation step, the heat consumption to warm up the solution at 80 °C and keep the temperature constant for 16 hours was evaluated. Similarly, the electric energy consumption to mix the sol for 16 hours and mix the TX-100 and sol for 8 hours was assessed. Moreover, the electricity used for the vacuum system and the reverse osmosis filtration was considered. The electrical energy supply was assumed to be the Italian mix electrical energy generated by Ecoinvent database.⁴⁰

2.2.5. Emissions and vacuum systems: assumptions and general considerations. The studied production system was a closed reactor, which avoids the emissions of nanoparticles to the air; hence, the only emissions to be assessed are those occurring in the cleaning water (during the maintenance operations).

The following assumptions were considered:

• During the production and purification steps, the release of HNO₃ into the air occurs. In particular, 1% of the total amount of material has been assumed to be emitted.

• During the maintenance operations, particularly in the purification process, the % of the TiO₂ nanoparticles suspension produced and composed of TiO₂, HNO₃, TX-100 and H₂O was assumed to be released into the cleaning water.

Owing to the above-mentioned emissions, a reverse osmosis filter equipped with a scrubber system and a scrubber system solely were considered in the maintenance operations and production process, respectively. The scope was the reduction of the emissions to water and air. In particular, the scrubber system can capture the HNO₃ emissions, while the titanium dioxide nanoparticles present in the obtained aqueous suspension are entrapped by the reverse osmosis filter. The efficiency of the filter in the production process is 99.97%. The installation of a reverse osmosis filter with high efficiency reduces the environmental impact according to the ecodesign approach.

Regarding the end of life treatments, the part of filter that captured the emissions, together with the entrapped chemical substances, were disposed of in a residual material landfill.

2.2.6. Mass balance. Table 1 shows the mass balance of the chemical synthesis; in particular 12% of the total was recovered as isopropanol (i.e. a reaction co-product) after the synthetic procedure. The remaining 88% was composed of 85.71% TiO₂ nanoparticles suspension in water (main product) and 14.29% of recoverable water considered as a co-product.

The physical–chemical properties of the TiO₂ nanoparticles suspension are reported in Table 2. In particular, the amount of TiO₂ nanoparticles in the aqueous suspension results was 6 wt%, which was used to calculate the chemical reaction effective yield of 69.34%, which will be considered in the following EATOS calculations (see section 4).

Table 2 Physical and chemical properties of the nanoTiO₂ suspension

Physical and chemical properties	Amount
TiO2 concentration (%w/w)	6
Density (g ml^−1)	1.15
Viscosity 20 °C (mPa s^−1)	2
Nanoparticle size (nm)	30
Polydispersity index (pdI)	0.25
pH	5.5

---

3. Life cycle assessment (LCA)

3.1. Aim definition

The aim of this LCA study was to assess the environmental impacts of the hydrolytic sol–gel synthesis of a titanium dioxide nanoparticles suspension produced according to the procedure recently patented and actually employed by Colorobbia S.p.A. for the preparation of aqueous TiO₂ suspensions,^34,37 to identify its environmental loads.

3.2. System, functional unit and function of the system

The represented system is a multi-output process, which is based on the mass allocation criteria. The functional unit is represented by both the main product, i.e., 0.75425 kg of the TiO₂ nanoparticles suspension and two co-products, namely 0.12 kg of recovered isopropanol and 0.12575 kg of recovered water.

3.3. System boundaries

Fig. 1 highlights the system boundaries for the analysis, which includes the upstream phases, from raw material extraction and chemicals production, to the finished product packaging, thus obtaining “a cradle to the gate” overview. The production, maintenance and disposal of facilities, as well as the environmental burdens related to the production and disposal of packaging materials and other auxiliary materials have been also included in the present study. Emissions into the air and water, as well as disposal of the part of filter that captures the nanoparticles or the volatile chemical compounds, such as nitric acid, were taken into account. The transportation of chemical materials, facilities and packaging materials to the company that used them and the transportation of waste materials to a treatment facility have been also considered. In addition, the energy consumption required in all life cycle steps of the TiO₂ nanoparticles suspension production were also evaluated.

3.4. Data quality and impact assessment methodology

The data for the study was collected both directly from the authors of the patented procedure (primary data)³⁷ and from scientific literature (secondary data). In particular, the primary data, referring to the optimized method for the preparation of aqueous dispersion of TiO₂ nanoparticles,^34,37 consisted of the composition and physical/chemical properties of the TiO₂ nanoparticles suspension and energy consumption. Where the data was somehow missing, the study was completed based on the secondary data obtained from the Ecoinvent database that exploited them to model the background processes (land use, materials production, fuel and electricity production, and transport). The analysis is performed using the SimaPro 7.3.3 software and IMPACT 2002+ evaluation method to assess the environmental impacts.

3.5. Life cycle inventory

In the LCA study, the quality and credibility of the results largely depend on the quality of the data included in the Life Cycle Inventory (LCI) stage. In accordance with the ILCD Handbook,⁴¹ the inventory must state, in a specific and reliable way, all the inputs in the form of material and energy resources and all the outputs in the form of air emissions, emissions into water and soil, as well as the solid waste that is generated, for each of the stages of the life cycle of the system being studied (according to ISO/TS 14048).⁴² The inventory data was modelled in SimaPro 7.3.3, taking the Ecoinvent database v2.2 as a reference to configure the inventory of some chemicals (i.e. nitric acid and TX-100), natural gas, electricity, heat, transport, infrastructure, machinery, and waste treatments. In those cases, where the chemicals were missing, such as the case of titanium isopropoxide (reagent) and titanium tetrachloride (titanium isopropoxide precursor), they were created using literature data.⁴³ The same approach was followed for the vacuum system and the purification processes.

3.6. Impact assessment

The life cycle impact assessment (LCIA) results were modeled using the IMPACT 2002+ method with Simapro 7.3.3 to determine the environmental impacts related to the emissions released and the resources consumed in the system under study.⁴⁴ This impact assessment method covers more impact categories than other methods and includes more substances, but the following additions and modifications were implemented to describe the system considered in a more representative manner:

• Land use was estimated by considering the basic indicators of both land occupation and transformation. In the present study Transformation, to forest intensive, normal, Transformation, to forest intensive and Transformation, to arable were introduced.

• Mineral extraction was characterised in consideration of some additional resources, such as silver, gravel, sand, lithium, bromine and water in ground, derived from the category Minerals of Eco-indicator 99 with the same characterisation factors.⁴⁵

• A radioactive waste category was added. In particular, both this kind of waste and its occupied volume were evaluated considering the same characterization and normalization factors of EPID 2003 method.⁴⁶ This category allows the possible damage of the electric energy mix, which also includes the electricity generated by nuclear plants, to be taken into account. This latter kind of energy produces radioactive waste, which needs to be safely managed and disposed.

Concerning the toxicity of nanoparticles, it is known that there are currently several major uncertainties and knowledge gaps with regard to the behavior, chemical and biological interactions and toxicological properties of engineered nanomaterials.⁴⁷ Despite this, a preliminary attempt to assess the toxicity of TiO₂ nanoparticles released in water on freshwater ecosystem and human health has been done. In particular, the IMPACT 2002+ method has been further exploited for this purpose as follows:

- Particulates, <100 nm, in freshwater (a representative substance of the damage on freshwater ecosystem) has been introduced in a new impact category named nanoTiO₂ ecotoxicity in freshwater, which refers to a new damage category with the same name and with a characterization factor calculated by Salieri et al.⁴⁸

- NanoTiO₂Human toxicity, in freshwater (a representative substance of a local damage on human health, considering the area of Emilia Romagna region in the north-east of Italy) has been introduced in a new impact category called NanoTiO₂carcinogens in freshwater, which refers to a new damage category with the same name and with a new calculated damage assessment factor.⁴⁹ This factor has been determined considering the Anandanl and Kumar study, which evaluated the target organ toxicity dose (TTD) value for human liver cell toxicity in stream water of 8.33 μg L⁻¹ for the TiO₂ nanoparticles.⁵⁰ The new damage factor was determined by the Eco-indicator 99 calculation method for carcinogenic substances. In particular, it consists of three main steps: fate analysis, effect analysis and damage analysis. It considers an emission released into freshwater during the purification of water contaminated by nanoparticles of 1 kg per year, a water body volume of Reggio Emilia province of 9 × 10⁶ m³ and a calculated nanoparticles concentration of 1.111 × 10⁻⁷ kg m⁻³. The damage assessment factor obtained was 7.14 × 10⁻⁶ DALY kg⁻¹ of TiO₂ nanoparticles released into water.

4. EATOS calculations

The list of starting substances considered in the assessment made using EATOS and the categories to which they have been considered to belong are reported in Table 4. Table 4 summarizes the yield of the synthetic procedure as well as the useful amounts of obtained coupled products; the latter accounting for their possibility to be recovered and eventually recycled. All of the amounts reported in the two above-mentioned tables are in accordance with the functional unit considered in LCA analysis (i.e. the amount of TiO₂ obtained from 1 kg of starting materials).

The amount of solvent has been considered to be completely recyclable (see Table 3), while the useful quantities of isopropyl alcohol and the remaining water (i.e. the second coupled product) have been settled in accordance to what is indicated by the patented procedure³⁷ and off course to the amounts used in the LCA assessment, as reported in Table 1. In detail, 120 g constituted the effectively recovered amount of isopropyl alcohol, while the maximum useful amount of the coupled product water has been established according to the yield of the reaction, assuming thus a maximum value of 69.34% of that theoretically obtainable. This yield value corresponds to a water amount which is significantly lower with respect to what is indicated (Table 1) and considered in the LCA framework. The reason needs to be found in the fact that what is considered by LCA as the water co-product, indeed refers to the amount of water recovered after the synthetic procedure, the latter comprising both the effective water co-product and part of the water originally intended as the solvent.

Table 3 List of starting substances used for the EATOS environmental assessment of the hydrolytic sol–gel synthesis of TiO₂ nanoparticles, considering the stoichiometric reaction reported in eqn (11)

Substance	Category	Molecular weight (g mol^−1)	Quantity (g)
Ti(OiPr)4	Key substrate	284.2308	232.2
H2O	Substrate	18.0152	58.87
H2O	Solvent	18.0152	675.13 (recyclable quantity = 100%)
HNO3, 63%	Catalyst	63.0128	23.8
Triton X-100	Auxiliary material	646.8572	10

---

Table 4 List of the product and coupled products considered in the EATOS environmental assessment of the hydrolytic sol–gel synthesis of TiO₂ nanoparticles, considering the stoichiometric reaction reported in eqn (9)

Substance	Category	Molecular weight (g mol^−1)	Useful quantity (% or g)	Yield (%), referred to the key substrate
TiO2	Product	79.8788	—	69.34
H2O	Coupled product	18.0152	69.34%	—
C3H8O	Coupled product	60.0956	120 g	—

---

The software EATOS allows calculating four important environmental parameters: the mass index MI (eqn (2)), the environmental factor E (eqn (1)) together with EI_in and EI_out corresponding to the former ones in which each substance quantity is multiplied by its specific total weighting factor, Q_tot, which represents the mean value calculated among the different weighting factors Q_i values. In detail, the weighting factors allow evaluating and examining the specific chemical reaction with particular regard to the potential environmental and human health relevance of each substance employed.⁵¹ The index i in the Q_i notation, accounts for the i^th weighting category among the claiming of resources, risk, human toxicity, chronic toxicity, ecotoxicology, ozone creation, air pollution, accumulation, degradability, greenhouse effect, ozone depletion, nitrification, and acidification. Each Q_i can assume values ranging from 1 to 10 according to specific classifications internally made by the software algorithms. The EATOS software also allows selecting the opportune significance to associate with each weighting category.

The relevant information that determines the Q_i value for that particular substance can be easily found in the Material Safety Data Sheet (MSDS) from the supplier. In particular, in this work, the MSDS considered those available on the Sigma Aldrich website,⁵² and regarding the price, the largest amount available was selected.

5. Results and discussion

The Life Cycle Impact Assessment (LCIA) was performed by both a midpoint and endpoint assessment. The midpoint indicators are considered to be linked to the cause-effect chain (environmental mechanism) of an impact category. Common examples of midpoint characterization factors include the ozone depletion potentials, global warming potentials, and photochemical ozone (smog) creation potentials. The endpoint indicators are instead considered to be linked to the cause-effect chain for all categories of impacts (e.g., human health impacts, in terms of disability adjusted life years for carcinogenicity, climate change, ozone depletion, photochemical ozone creation, or impacts in terms of changes in biodiversity, etc.).⁵³

5.1. LCA results

The single score damage is 988.87 μPt for 1 kg of TiO₂ nanoparticles suspension produced. As Fig. 2 shows, the electric energy consumption to mix the colloidal solution (sol) for 16 hours (indicated as E1) and the one necessary to mix the sol and TX-100 for 8 hours (indicated as E2) are the two contributions that are mainly responsible for the total damage (38.76% and 19.38% respectively), followed by titanium isopropoxide (TIP) production (13.65%) and by the heat consumption (indicated as H₂) to maintain the solution at 80 °C (10.33%). A midpoint category interpretation was conducted because it is more appropriate to evaluate the environmental impacts of the various substances counted in the life cycle inventory. The impact categories considered are global warming, non-renewable energy, mineral extraction, carcinogens, non-carcinogens, respiratory inorganics, respiratory organics, aquatic ecotoxicity, terrestrial ecotoxicity, ionising radiation, ozone layer depletion, terrestrial acidification, aquatic acidification, aquatic eutrophication, radioactive waste, land occupation, nanoTiO₂ ecotoxicity in freshwater, and nanoTiO₂ human toxicity.

Fig. 2 Evaluation by a single score of 1 kg of a bottom-up hydrolytic synthesis of TiO₂ nanoparticles, where TIP = titanium isopropoxide; TX-100 = Triton X-100; PA = packaging of raw materials; T = transport of raw materials; H1 = heat to warm up the solution at 80 °C; H2 = heat to maintain the solution at 80 °C; E1 = electric energy to mix the sol for 16 hours; E2 = electric energy to mix the sol and TX-100 for 8 hours; GR = glass reactor; VS = vacuum system; WP = water purification.

Table 5 summarises the outcomes of the midpoint and endpoint analysis that are explained in detail hereafter. In particular, for midpoint analysis, the contributions of the main impact categories on the total damage are reported.

Table 5 Characterised LCIA results

Midpoint results			Endpoint results
Impact category	Unit	Total	Damage category	Unit	Total
Carcinogens	kg C2H3Cl eq.	4.43 × 10^−2	Human health	DALY	1.68 × 10^−6
Non-carcinogens	kg C2H3Cl eq.	1.63 × 10^−2
Respiratory inorganics	kg PM2.5 eq.	2.13 × 10^−3
Ionizing radiation	Bq C-14 eq.	48
Ozone layer depletion	kg CFC-11 eq.	3.95 × 10^−7
Respiratory organics	kg C2H4 eq.	1.06 × 10^−3
Aquatic ecotoxicity	kg TEG water	1.70 × 10^2	Ecosystem quality	PDF m^2 year	0.503
Terrestrial ecotoxicity	kg TEG soil	52.8
Terrestrial acid/nutri	kg SO2 eq.	4.39 × 10^−2
Land occupation	m^2 org.arable	2.91 × 10^−2
Aquatic acidification	kg SO2 eq.	1.27 × 10^−2
Aquatic eutrophication	kg PO4 P-lim	8.17 × 10^−4
Global warming	kg CO2 eq.	2.86	Climate change	kg CO2	2.863
Non-renewable energy	MJ primary	55	Resources	MJ primary	55.15
Mineral extraction	MJ surplus	1.53 × 10^−1
Radioactive waste	kg	6.37 × 10^−5	Radioactive waste	kg	6.37 × 10^−5
NanoTiO2 ecotoxicity in freshwater	kg	6.63 × 10^−9	NanoTiO2 ecotoxicity in freshwater	PAF m^3 day	1.86 × 10^−9
NanoTiO2 carcinogens in freshwater	kg	6.63 × 10^−9	NanoTiO2 carcinogens in freshwater	DALY	4.73 × 10^−14

---

Fig. 3 shows that the most significant contribution to the total damage is due to the non-renewable energy impact category, which is primarily affected by natural gas (52.31%), crude oil (25.58%) and hard coal (11.23%) emissions. For all types of emissions, the electric energy consumption to mix the sol for 16 hours is the process that produces the major environmental load (33.09%, 25.5%, and 49.18% respectively). In particular, in this process, the emissions are mainly caused by the natural gas and crude oil production and the hard coal mining respectively.

Fig. 3 Evaluation by the impact categories of 1 kg of a bottom-up hydrolytic synthesis of TiO₂ nanoparticles, where C = Carcinogens; NC = Non-Carcinogens; RI = Respiratory Inorganics; IR = Ionizing Radiation; OLD = Ozone Layer Depletion; RO = Respiratory Organics; AE = Aquatic Ecotoxicity; TE = Terrestrial Ecotoxicity; TA = Terrestrial acid/nutri; LO = Land Occupation; AA = Aquatic Acidification; AE = Aquatic Eutrophication; GW = Global Warming; NRE = Non-renewable Energy; ME = Mineral Extraction; RW = Radioactive Waste; NE = NanoTiO₂ ecotoxicity in freshwater; NC = NanoTiO₂ carcinogens in freshwater.

The second major contribute to the total damage is generated by the global warming impact category, mainly due to GHG (greenhouse gas) emissions (96.17%), which are for the 41.24% belonging to the electric energy consumption to mix the sol for 16 hours and in particular that generated by the natural gas used to produce the electricity.

In respiratory inorganics, the major contributions are due to the following emissions to the air: 37.3% of NO_x, 28.91% of SO₂ and 18.61% of particulates <2.5 μm. All of these are mainly due to the electric energy consumption to mix the sol for 16 hours (38.77%, 49.65% and 45.54% respectively), and in particular, that generated for the first two (NO_x and SO₂) during heavy fuel oil combustion while the latter is during hard coal combustion.

In the radioactive waste impact category, the volume occupied by low-active radioactive waste contributes 65.31% of the total damage due to the electric energy production by nuclear power plants.

Regarding the nanoTiO₂ ecotoxicity in freshwater and nanoTiO₂ carcinogens in freshwater impact categories, the damage is totally due to the release of 6.63 μg of particulates, <100 nm (anatase TiO₂ nanoparticles) and 6.63 μg nanoTiO₂human toxicity in freshwater during the purification of contaminated water, which are obtained by the maintenance operation (reactor cleaning, see paragraph 2.2.3).

The endpoint analysis highlights (Fig. 4) that the total damage is affected for 23.9% the human health (2.36 × 10⁻⁴ Pt), for 36.7% the resources (3.63 × 10⁻⁴ Pt), for 29.24% the climate change (2.89 × 10⁻⁴ Pt), for 3.71% the ecosystem quality (3.67 × 10⁻⁵ Pt), for 6.44% the radioactive waste (6.37 × 10⁻⁵ Pt), for 3.36 × 10⁻⁶% the nanoTiO₂ ecotoxicity in freshwater (3.32 × 10⁻¹¹ Pt) and for 6.75 × 10⁻⁷% the nanoTiO₂ carcinogens in freshwater (6.67 × 10⁻¹² Pt). In the latter two cases, the damage is restrained because, as mentioned before, the LCA study has been set in an ecodesign approach, thus installing a specific filter for nanoparticles with high efficiency (99.97%) to limit the nanoparticle releases.

Fig. 4 Evaluation by the damage categories of 1 kg of a bottom-up hydrolytic synthesis of TiO₂ nanoparticles, where TIP = titanium isopropoxide; TX-100 = Triton X-100; PA = packaging of raw materials; T = transport of raw materials; H1 = heat to warm up the solution at 80 °C; H2 = heat to maintain the solution at 80 °C; E1 = electric energy to mix the sol for 16 hours; E2 = electric energy to mix the sol and TX-100 for 8 hours; GR = glass reactor; VS = vacuum system; WP = water purification.

Fig. 5 shows that the variation of the sole filter efficiency from 99.97% to 95% leads to an increase in the damage in nanoTiO₂ ecotoxicity in freshwater and nanoTiO₂ carcinogens in freshwater impact categories equal to 165.67%.

Fig. 5 Comparison of the environmental impact values (Pt) for 99.97% and 95% filter efficiency.

Furthermore, life cycle cost analysis was carried out. The internal and external costs were assessed. The first takes into account the costs of all input materials and the latter considers the environmental costs caused by the hydrolytic synthesis of TiO₂ nanoparticles.

The results for 1 kg of the TiO₂ nanoparticles suspension show that internal and environmental costs values are 29.125 euro and 1.289 euro respectively, which are mainly caused by electric energy consumption.

5.2. EATOS results

The results from the EATOS software in terms of the evaluation of histograms for the metrics MI, E-factor, EI_in and EI_out are reported in Fig. 6. The green metric EI_out parameter, expressed in Potential Environmental Impact (PEI; the larger its value, the worse the process will impact on the environment) per kg of product, accounts for the environmental damage potentially induced by the reaction waste. In particular, as detailed in Table 6, its total value of 9.6195 is mainly attributed to coupled products, because only a limited amount of isopropyl alcohol is recovered. The category by-products accounts for the fact that as the yield of the synthetic process is equal to 69.34, some unspecified by-products need to be considered.

Fig. 6 EATOS results.

Table 6 Contributions of all the substances to the metrics, S-1, E, EI_in and EI_out

Category	Substance	S^−1 (kg of starting material/kg of product)	E (kg of waste/kg of product)	EI_in (PEI kg^−1)	EI_out (PEI kg^−1)
Substrates	Titanium isopropoxide and water	6.4728	0.0402	13.4997	—
Coupled products	Isopropanol and water	—	3.4604	—	7.5233
By-products	Unspecified	—	1.9723	—	1.18
Auxiliaries	Triton X-100	0.2278	0.2278	0.3418	0.2848
Catalysts	Nitric acid	0.3416	0.3416	1.3665	0.5124
Impurities	Unspecified	0.1587	0.1587	0.3968	0.119
Sewage/water	Water + unspecified	0.2006	0.2006	0.1003	—
Solvents	Water	0.7691	0.7691	0.3845	—
Total		8.1706	7.1707	16.0896	9.6195

---

More interestingly, in the particular framework of a direct comparison with the LCA results, the potential environmental and human health relevance of each substance employed in the total inward materials flux related to the hydrolytic synthesis of anatase TiO₂ nanoparticles suspension should be considered. In this regard, the green metrics parameter EI_in reaches a total value of 16.0896 PEI kg⁻¹, which is 83.9% due to the titanium isopropoxide precursor, 8.5% to the catalyst, 2.1% to auxiliary materials, and 2.4% to the solvent (i.e. water). Noteworthy is that this latter finding is in agreement with the previously discussed LCA results. Indeed, by simply neglecting the energy contributions, the LCA outcomes show that the total damage is mainly caused by titanium isopropoxide for 87.98% and by nitric acid for 7.22%.

According to this latter approximation, Fig. 7 reports the comparison between the LCA and EATOS outcomes in terms of the main factors affecting the single score damage and EI_in respectively, revealing a tremendously similar trend.

Fig. 7 EATOS and LCA (excluding energy consumptions) results comparison.

It needs to be specified that for the present EATOS analysis, the weighting categories considered have been those in which the corresponding values for at least one substance have been found and consequently inserted, i.e. human toxicity, chronic toxicity, ecotoxicology, and accumulation (to which an influence of 25% has been assigned). The possibility of Q = 0 assignment has also been considered.

A further possibility offered by the EATOS software, is to calculate and present the different entry items, such as substrates, solvents, etc., by the so-called economic index, i.e. COST INDEX (CI). The different contributions to the CI (expressed in eur kg⁻¹) deriving from the different input materials used for synthesis under consideration are detailed in Table 7.

Table 7 Detailed quantification of the different contributions to the CI, for the hydrolytic sol–gel synthesis of anatase TiO₂ nanoparticles investigated with the EATOS tool

Substance category	CI (EUR kg^−1)
Substrates	454.948
Auxiliaries	8.9115
Catalysts	17.6381
Impurities	13.6392
Solvent	7.9986
Sewage/water	2.0866
Total	505.222

---

6. Conclusions

The environmental assessment of the bottom-up hydrolytic sol–gel synthesis of anatase TiO₂ nanoparticles was concurrently performed by the software EATOS and by LCA methodology and, and despite the different philosophies on which they are based, similar conclusions can under some conditions be drawn.

EATOS software was originally developed for organic chemists and mainly applies to fine chemical processes, relying on the data that is easily available from the material safety data sheets. On the contrary, life cycle assessment is a much more detailed and versatile method, suitable to assess industrial scale procedures, but it is barely adaptable to small scale ones,⁵⁴ owing to its intrinsic time-consuming nature and to the difficulty to source necessary inventory data. In particular, for chemical processes, the databases to which the LCA methodology refers to, are still significantly inadequate. Therefore, when the required chemical is not included in a database, it is necessary either to consider chemicals with similar properties, among those comprised in the databases, or to recreate the synthesis of that particular substance ex novo.

Despite its immediacy, EATOS suffers from the major strong approximation of neglecting any energetic contribution. The typical limitations and drawbacks of both EATOS and LCA highlight their complementary character, which has been extensively exploited in environmental and human health assessments of several organic reactions.

The present work represents, to the best of authors’ knowledge, the first example, in which the synergy between LCA methodology and EATOS software has been applied to a green metrics evaluation of the inorganic synthesis of engineered nanomaterials.

In detail, the studied hydrolytic sol–gel synthesis route generates a suspension of TiO₂ nanoparticles with high purity and crystallinity according to the patented procedure.³⁷ The evaluated ecodesign approach (based on the installation of 99.97% efficiency filter and the use of closed reactor) together with the aqueous solvent employed and the low processing temperature, contribute to the good fit of this synthetic protocol in a green chemistry perspective.

The LCA results showed that most environmental loads are generated by the electric energy consumption (58.14%), followed by titanium isopropoxide (13.65%) and heat consumption to maintain the solution at 80 °C (10.33%). A better environmental performance can be achieved by the following:

• using renewable energy sources (e.g. solar power, geothermal, biomass, etc.) to reduce the non-renewable energy consumption.

• using microwave dielectric heating of reaction mixture, which has recently been significantly employed not only in organic chemistry but also in inorganic synthesis of a wide variety of engineered nanomaterials,⁵⁵ because of its intrinsic advantages over conventional heating techniques based on heat transfer mechanisms (rather than on energy transfer ones).

• substituting titanium tetraisopropoxide with a different metal oxide precursor that is more environmentally friendly and less costly.

In this regard, the use of EATOS software alone, can furnish the first reliable approximation of an environmental assessment of synthetic strategies employing different metal oxide precursors. Indeed, in the present study, neglecting any energy contributions, the comparison between the results obtained using LCA methodology and EATOS software resulted in similar conclusions. In particular, they had very similar environmental impacts with the more affecting factor being titanium isopropoxide, followed by nitric acid with the aqueous solvent and auxiliary triton X-100 showing similar close contributions.

In conclusion EATOS provides fast and functional results that are comparable to the LCA results when energy consumption is neglected. Furthermore, EATOS is a free of charge and user-friendly software, so it is particularly suitable when a choice among different synthetic strategies needs to be considered for the preparation of the desired engineered nanomaterial.

This study represents the first example of the application of EATOS software (originally developed for the synthesis of fine chemicals) to the inorganic synthesis of transition metal oxide nanostructures, and the natural progression of the research activities in this direction, actually in progress, will involve a comparison among the most widely employed synthetic strategies for obtaining the desired inorganic nanocrystals. Nevertheless, a deeper investigation by a complete LCA study should be carried out because the LCA methodology provides more detailed and accurate results. Indeed midpoint and endpoint analysis give complete knowledge on (i) the impact and damage categories that mainly affect the total environmental and human health impacts of the studied process, (ii) the substances that cause these impacts and (iii) the compartments (air, soil, water, raw) that result in more damage.

The main conclusion of the present work, which be a recommendation to inorganic chemists and materials scientists worldwide, is to always combine an environmental assessment with any new proposed strategy for the synthesis of engineered nanomaterials, so that the strict requirement of using the most environmentally friendly procedure could very soon accompany traditional requests of a desired size and shape.

The assessment by means of easily available and accessible EATOS tool, although able to furnish trustworthy indications, should be integrated by a complete life cycle assessment study, thereby allowing a consideration of the impacts of the energy consumption, which usually represents the most impacting parameter of the entire process.

Acknowledgements

Authors thank Colorobbia S.p.A. for their support in the Life Cycle Inventory Analysis. This study has been supported by “ARACNE” project, Bando Regione Emilia Romagna, Italy.

References

1. P. T. Anastas and J. C. Warner, in Green Chemistry: Theory and Practice, Oxford University Press, New York, 1998.
2. P. T. Anastas and J. B. Zimmerman, Design through the 12 principles of green engineering, Environ. Sci. Technol., 2003, 37, 94A–101A.
3. T. Van Gerven and A. Stankiewicz, Structure, energy, synergy, time-the fundamentals of process intensification, Ind. Eng. Chem. Res., 2009, 48, 2465–2474.
4. A. Stankiewicz and J. A. Moulijn, Process intensification: transforming chemical engineering, Chem. Eng. Prog., 2000, 96, 22–34.
5. J. F. Jenck, F. Angterberg and M. J. Droescher, Products and processes for a sustainable chemical industry: a review of achievements and prospects, Green Chem., 2004, 6, 544–556.
6. D. J. C. Constable, A. D. Curzons and V. L. Cunningham, Metrics to “green” chemistry-which are the best?, Green Chem, 2002, 4, 521–527.
7. C. Jiménez-Gonzalez, D. J. C. Constable and C. S. Ponder, Evaluating the “Greenness” of chemical processes and products in the pharmaceutical industry-a green metrics primer, Chem. Soc. Rev., 2012, 41, 1485–1498.
8. R. A. Sheldon, The E factor: fifteen years on, Green Chem., 2007, 9, 1273–1283.
9. T. Hudlicky, D. A. Frey, L. Koroniak, C. D. Claeboe and L. E. Brammer Jr., Toward a “reagent-free” synthesis-tandem enzymatic and electrochemical methods for increased effective mass yield (EMY), Green Chem, 1999, 1, 57–59.
10. C. Jimenez-Gonzalez, C. S. Ponder, Q. B. Broxterman and J. B. Manley, Using the right green yardstick: why process mass intensity is used in the pharmaceutical industry to drive more sustainable processes, Org. Process Res. Dev., 2011, 15, 912–917.
11. M. Eissen and J. O. Metzger, Environmental performance metrics for daily use in synthetic chemistry, Chem. – Eur. J., 2002, 8, 3581–3585.
12. M. Eissen and J. O. Metzger, EATOS, Environmental Assessment Tool for Organic Syntheses; the software can be obtained free of charge via http://www.chemie.uni-oldenburg.de/oc/metzger/eatos.
13. M. Eissen, K. Hungerbuehler, S. Dirks and J. Metzger, Green Chem., 2003, 5, G25–G27.
14. A. Corradi, C. Leonelli, A. Rizzuti, R. Rosa, P. Veronesi, R. Grandi, S. Baldassari and C. Villa, New “green” approaches to the synthesis of pyrazole derivatives, Molecules, 2007, 12, 1482–1495.
15. S. Protti, D. Dondi, M. Fagnoni and A. Albini, Assessing photochemistry as a green synthetic method. Carbon-carbon bond forming reactions, Green Chem., 2009, 11, 239–249.
16. C. Villa, R. Rosa, A. Corradi and C. Leonelli, Microwaves-mediated preparation of organoclays as organic-/bio-inorganic hybrid materials, Curr. Org. Chem., 2011, 15, 284–295.
17. W. Piang-Siong, P. de Caro, C. Lacaze-Dufaure, A. S. C. Sing and W. Hoareau, Effect of catalytic conditions on the synthesis of new aconitate esters, Ind. Crops Prod., 2012, 35, 203–210.
18. M. G. T. C. Ribeiro and A. A. S. C. Machado, Grenness of chemical reactions-limitations of mass metrics, Green Chem., Lett. Rev., 2013, 6, 1–18.
19. L. Tufvesson, P. Tufvesson, J. Woodley and P. Borjesson, Life cycle assessment in green chemistry: overview of key parameters and methodological concerns, Int. J. Life Cycle Assess., 2013, 18, 431–444.
20. D. Ravelli, S. Protti, P. Neri, M. Fagnoni and A. Albini, Photochemical technologies assessed: the case of rose oxide, Green Chem., 2011, 13, 1876–1884.
21. D. Ravelli, D. Dondi, M. Fagnoni and A. Albini, Titanium dioxide photocatalysis: an assessment of the environmental compatibility for the case of the functionalization of heterocyclics, Appl. Catal., B, 2010, 99, 442–447.
22. M. A. Albrecht, C. W. Evans and C. L. Raston, Green chemistry and the health implications of nanoparticles, Green Chem., 2006, 8, 417–432.
23. R. Hischier and T. Walser, Life cycle assessment of engineered nanomaterials: state of the art and strategies to overcome existing gaps, Sci. Total Environ., 2012, 425, 271–282.
24. N. Osterwalder, C. Capello, K. Hungerbuhler and W. J. Stark, Energy consumption during nanoparticle production : how economic is dry synthesis, J. Nanopart. Res., 2006, 8, 1–9.
25. G. F. Grubb and B. R. Bakshi, Energetic and environmental evaluation of titanium dioxide nanoparticles, IEEE International Symposium on Electronics and the Environment, IEEE, San Francisco, USA, 2008.
26. G. F. Grubb and B. R. Bakshi, Life cycle of titanium dioxide nanoparticle production, J. Ind. Ecol., 2010, 15, 81–95.
27. G. F. Grubb and B. R. Bakshi, Appreciating the role of thermodynamics in LCA improvement analysis via an application to titanium dioxide nanoparticles, Environ. Sci. Technol., 2011, 45, 3054–3061.
28. D. P. Macwan, P. N. Dave and S. Chaturvedi, A review on nano-TiO₂ sol-gel type syntheses and its applications, J. Mater. Sci., 2011, 46, 3669–3686.
29. Y. F. Chen, C. Y. Lee, M. Y. Yeng and H. T. Chiu, The effect of calcination temperature on the crystallinity of TiO₂ nanopowders, J. Cryst. Growth, 2003, 274, 363–370.
30. J. Rubio, J. L. Oteo, M. Villegas and P. Duran, Characterization and sintering behavior of submicrometre titanium dioxide spherical particles obtained by gas-phase hydrolysis of titanium tetrabutoxide, J. Mater. Sci., 1997, 32, 643–652.
31. S. M. Gupta and M. Tripathi, A review on the synthesis of TiO₂ nanoparticles by solution route, Cent. Eur. J. Chem., 2012, 10(2), 279–294.
32. A. B. Corradi, F. Bondioli, B. Focher, A. M. Ferrari, C. Grippo, E. Mariani and C. Villa, Conventional and microwave-hydrothermal synthesis of TiO₂ nanopowders, J. Am. Ceram. Soc., 2005, 88, 2639–2641.
33. P. Alphonse, A. Varghese and C. Tendero, Stable hydrosols for TiO₂ coatings, J. Sol-Gel Sci. Technol., 2011, 56(3), 250–263.
34. http://www.colorobbia.it/ .
35. ISO 14040, Environmental management – Life cycle assessment – Principles and framework, International Standards Organization, ISO, 2006.
36. ISO 14044, Environmental management – Life cycle assessment – Requirements and guidelines, International Standards Organization, ISO, 2006.
37. G. Baldi, M. Bitossi and A. Barzanti, Method for the preparation of aqueous dispersions of TiO₂ in the form of nanoparticles, and dispersions obtainable with this method, US Pat., 0317959 A1, 2008.
38. M. Niederberger and G. Garnweitner, Organic reaction pathways in the nonaqueous synthesis metal oxide nanoparticles, Chem. – Eur. J., 2006, 12, 7282–7302.
39. M. Niederberger and N. Pinna, Metal Oxide Nanoparticles in Organic Solvents: Synthesis, Formation, Assembly and Application, Springer-Verlag, London, 2009, ch. 2, pp. 7–18.
40. Life Cycle Inventories, Ecoinvent Database, Version 2.0, 2010. http://www.ecoinvent.ch/.
41. EUROPEAN COMMISSION JOINT RESEARCH CENTRE, Institute for Environment and Sustainability, Analysis of existing Environmental Impact Assessment Methodologies for use in Life Cycle Assessment, ILCD Handbook, European Union, 1st edn, 2010.
42. ISO/TS 14048, Environmental management – Life cycle assessment – Data documentation format, International Standards Organization, ISO, 2002.
43. L. Piccolo, B. Calcagno and M. Ghirga, Method for preparing titanium tetrachloride, US Pat., 3787556, 1974.
44. O. Jolliet, M. Margni, R. Charles, S. Humbert, J. Payet, G. Rebitzer and R. Rosenbaum, IMPACT 2002+: A new life cycle impact assessment methodology, Int. J. Life Cycle Assess., 2003, 8(6), 324–330.
45. M. Goedkoop and R. Spriensma, The Eco-indicator 99 – A damage oriented method for Life cycle Assessment, Methodology Report, Pré Consultants B.V., Amersfoort, 2001.
46. J. Potting and M. Hauschild, The EDIP 2003 methodology – Background for spatial differentiation in life cycle impact assessment, Methodology Report, Danish Environmental Protection Agency, Copenhagen, 2004.
47. C. Som, M. Bergesb, Q. Chaudhryc, M. Dusinskad, T. F. Fernandese, S. I. Olsenf and B. Nowack, The importance of life cycle concepts for the development of safe nanoproducts, Toxicology, 2010, 269, 160–169.
48. B. Salieri, Ph.D. Thesis, University of Bologna, 2013.
49. M. Pini, P. Neri and A. M. Ferrari, Determination of the potential damage and benefits of TiO₂ nanoparticles by Life Cycle Assessment methodology, Internal Technology Report RT-07, University of Modena and Reggio Emilia, 2013.
50. A. Anandanl and A. Kumar, Exposures to TiO₂ and Ag Nanoparticles: What are Human Health Risks?, Sci. Soc., 2011, 9(2), 155–162.
51. Environmental Assessment Tool for Organic Syntheses, EATOS, user manual version 1.1, by M. Eissen and J.O. Metzger, available online free of charge at http://www.metzger.chemie.uni-oldenburg.de/eatos/eatosmanual.pdf.
52. The Materials Safety Data Sheet (MSDS) of all the substances considered in this study can be found at the website: http://www.sigma-aldrich.com.
53. J. C. Bare, P. Hofstetter, D. W. Pennington and A. H. Udo de Haes, Life Cycle Impact Assessment Workshop Summary Midpoints versus Endpoints: The Sacrifices and Benefits, Int. J. Life Cycle Assess., 2000, 5(6), 319–326.
54. D. Ravelli, S. Protti, P. Neri, M. Fagnonia and A. Albini, Photochemical technologies assessed: the case of rose oxide, Green Chem., 2011, 13, 1876.
55. M. Baghbanzadeh, L. Carbone, P. D. Cozzoli and C. O. Kappe, Microwave-assisted synthesis of colloidal inorganic nanocrystals, Angew. Chem., Int. Ed., 2011, 50, 11312–11359.

---