Synlett 2024; 35(14): 1629-1647
DOI: 10.1055/a-2191-6011
account

Polyconjugated Materials for Printed (Opto)Electronics: Introducing Sustainability

Mauro Sassi
,
Sara Mattiello
,
Alice Fappani
,
Luca Beverina

This work was supported by the Italian Ministry of University and Research (MIUR; Grant Dipartimenti di Eccellenza-2017 ‘Materials for Energy’) through the grant PRIN2017 BOOSTER (project no. 2017YXX8AZ).
 


This work is dedicated to the memory of Prof. Giorgio A. Pagani.

Abstract

This account summarizes the impact that the quest for sustainability is having on the field of organic polyconjugate molecules and polymers for plastic (opto)electronics. While at the proof-of-concept level, the design criteria as well as the preferred synthetic strategies to access new and improved materials have been dominated by the need for performance. The ongoing transition from the lab environment to the industrial scale imposes strict limitations on the cost and overall environmental impact of new materials. We here summarize our efforts on the development of new design criteria and synthetic strategies aimed at improving sustainability – without compromising performance – in organic polyconjugated molecules. The article is composed of three sections: Introduction and motivation, sustainability through improved synthetic methods and through improved design.

Outline

1 Introduction and Motivation

2 Sustainable Reaction Methods

2.1 The New Tools: Reactions in Aqueous Solution of Surfactants

2.2 Intrinsically More Sustainable Reactions: Direct Arylation

2.3 Sustainable Multistep Protocols: Combining Micellar, Solventless, and Mechanochemical Methods

3 Sustainability as a Design Criterion: De Novo Design

4 Conclusion


#

Biographical Sketches

Zoom Image

Mauro Sassi was born in 1982 in Varese, Italy. In 2007, he graduated in Materials Science from the University of Milano-Bicocca under the supervision of Prof. Giorgio A. Pagani. In 2011, he received a PhD in Materials Science from the same University with a thesis on high-performance heteroaromatic electrochromic organic materials under the supervision of Prof. Luca Beverina. Since 2019, he has worked as an Assistant Professor in Organic Chemistry at the University of Milano-Bicocca. His research focuses on the synthesis of conjugated materials for organic electronics, OPV, and rechargeable organic batteries with an emphasis on improving scalability and sustainability. He has coauthored over 50 publications in peer-reviewed journals and seven patent applications. In his spare time, he enjoys playing the French horn and hiking.

Zoom Image

Sara Mattiello graduated in Materials Science at the University of Milano-Bicocca in 2014. She obtained her PhD from the same University in 2018 working under the supervision of Prof. Luca Beverina, and she is now an Assistant Professor. Her research interests are focused on photochemistry in microheterogeneous environments, formulation chemistry, and green chemistry applied to organic electronics. In her spare time, she loves painting, practicing yoga, and role-playing games.

Zoom Image

Alice Fappani was born in 1997 in Como (Italy). She graduated in Chemical Sciences and Technologies at the University of Milano Bicocca in 2021, after a thesis on micellar catalysis and an industrial internship working with lubricant formulations, with Prof. Luca Beverina as her academic tutor for both activities. In the following year, she was granted a fellowship in the same research group, further developing formulation chemistry topics within the collaboration between the University and a cosmetic company (Intercos SpA). She has been a Ph.D. student since 2022 at the Materials Science Department of the University of Milano Bicocca. Her doctoral project, carried out in the group of her longstanding mentor, is focused on colloidal synthesis of perovskites in the form of stable and printable inks for solar applications and is developed with the collaboration and sponsorship of a leading company in the Italian energetic field (ENI). Besides science, her interests are reading, cultural trips, and Shaolin practice.

Zoom Image

Luca Beverina was born in 1975 in Milano. He graduated and got a PhD in Materials Science from the University of Milano-Bicocca. After a postdoc period at the Georgia Institute of Technology of Atlanta (GA, USA), he got a permanent position at the University of Milano-Bicocca as an Assistant Professor in Organic Chemistry. Since 2019, he has held the position of Full Professor of Organic Chemistry at the same University, where he is also a Delegate of the Rector for the valorization of Intellectual Property. His research focuses on the design and synthesis of conjugated materials for (opto)electronics and biophotonics. In collaboration with leading international research groups, he develops molecules and polymers for OLEDs, OFETs, photovoltaic, electrochromic, and sensor applications, amongst others. In the last 10 years, he has dedicated increasing efforts to the application of green chemistry to the field of organic electronics, raising awareness of the need to improve sustainability alongside performance. He has coauthored over 150 publications in peer-reviewed journals and 20 patent applications. He is an avid Grand Sumo supporter and enjoys trekking and climbing.

1

Introduction and Motivation

Organic polyconjugated materials for plastic (opto)electronics have come a long way in the last 20 years.[1] [2] [3] [4] [5] [6] [7] [8] [9] Initially considered as a possible cheaper, but likely low-performance, alternative to existing technologies based on inorganic established semiconductors such as silicon and germanium, they progressively came into their own by showing a portfolio of properties making them much more than simple replacements of existing materials. First and foremost, the flexibility of organic chemistry allows the design of a vast array of materials sharing the same cheap and available components (C, H, N, S, O, P) and possessing optoelectronic properties tailored to suit a specific application. Organic field effect transistors (OFETs),[1] electro-optic modulators,[10] [11] [12] organic solar cells (OSCs),[3] [7] [13] [14] organic light-emitting devices (OLEDs),[15] [16] sensors of many different kinds,[17] [18] [19] [20] rechargeable batteries and supercapacitors,[21] organic light-emitting transistors (OLETs),[22] [23] light-emitting electrochemical cells (LEECs),[24] and thermoelectric generators (TEGs)[20] [25] [26] are just a few examples of the vast array of technologies benefiting from the unique properties of organic semiconductors. Aside from optoelectronic features, organic materials are also advantageous in terms of compatibility with low-cost solution-processing techniques (including printing) and mechanical flexibility.[27] [28] [29] [30] Several research groups are developing organic devices and whole integrated circuits that can be directly coupled with fabrics, foodstuffs,[31] [32] and even human skin.[9] , [33] [34] [35] [36] All such applications are unique aspects of organic devices and are under development exclusively because performing materials connecting all advantages are becoming increasingly available.

The whole field is still dominated by the quest for performance, even though, at least for the big three flagship organic devices (OLEDs, OFETs, and OSCs), the numbers are already within the range required to make a convincing case for their industrialization.

Possibly because of the relative maturity of the technology, the OSC community was the first to grow an increasing awareness of the hurdles the technology was about to face while transitioning from the proof-of-concept lab discovery stage to the industrial environment.[37] [38] [39] Some concerns regarding the engineering aspects of the technology, such as scaling up of the fabrication procedures (very likely involving a change in the deposition methods), the need for standardization and reproducibility, the specific hurdles presented by large-area devices, encapsulation to improve durability, integration and so on. But on a more fundamental level, researchers realized that the majority of performing materials under consideration for scaling up were too complex to make, syntheses were inefficient, and the amount of toxic waste produced was unacceptable.[39,40]

In the nearly 20 years of my independent career, I have been involved in some of the most relevant applications of polyconjugated molecules and polymers, starting from nonlinear optics[41] [42] [43] [44] and biophotonics,[45–48] later moving to printed electronic and optoelectronic devices.[49–53] Squaraine derivatives – themselves not the easiest compounds to make, nor the more sustainable from the standpoint of the chemistry involved – constituted the backbone of my research for almost as long.[54] [55] A progressive strengthening of collaborations with different companies working on printed (opto)electronics caused a progressive shift from the development of new materials to the improvement of the synthesis and processing of existing and performing materials.[56]

The purpose of this account is to show that much can already be done to dramatically improve the way well-known and performing materials can be made and processed. We will show that emerging cross-coupling methods such as direct arylation can significantly improve atom economy and reduce the number of steps required to construct polyconjugated materials.[57] [58] We will also show examples of the remarkable improvement of known reactions when performed under solventless or near-solventless conditions, both under standard heating and mechanochemical conditions. Mostly, we will focus on the new kid on the block for polyconjugated materials: micellar catalysis. Such a new paradigm, heralded by the group of Prof. Lipshutz, positively revolutionized the field of synthetic chemistry in the last 15 years and is progressively becoming a key asset for the synthesis of polyconjugated molecules.[59–64] The key features of the micellar approach are also showing the potential to revolutionize processing alongside synthesis.[65,66] Finally, we will comment on the role of design as a tool to select materials that are intrinsically sustainable and reasonably efficient over alternatives that might be very efficient but show no potential for sustainable and economically feasible industrialization.[67] [68]


# 2

Sustainable Reaction Methods

Organic semiconductors pertain to very diverse classes and thus there are a vast variety of reactions involved in their preparation. Focusing on the most characteristic feature of such compounds – an extended conjugated framework – it is easy to realize that olefination, cross-coupling reactions involving sp2 hybridized carbon and nitrogen atoms, and the Sonogashira coupling of acetylene derivatives and aryl halides get the lion’s share. Amongst cross-coupling reactions, the Stille coupling of aryl stannanes and aryl halides remains the preferred method, particularly when the introduction of thiophene rings is required. Toxicity concerns as well as purification issues, progressively boosted the importance of the Suzuki–Miyaura (SM) coupling; still requiring an aryl halide but enabling the substitution of highly toxic stannanes with the corresponding more benign boronic acids, esters, and trifluoroborates.

The common trait of both reactions is the need for activation of one of the two partners, frequently requiring the use of organolithium reagents or a further palladium-catalyzed reaction. Starting with the seminal works of Fagnou,[57] an increasingly large portion of performing polyconjugate materials became available via a variety of protocols for direct heteroarylation (DHA). DHA is the reaction of a (hetero)aryl halide with another (hetero)aryl derivative through direct regioselective activation of a C–H bond of the latter, thus avoiding the need for its preliminary functionalization. The appeal of the method on both sustainability and cost of the final materials is obvious and the literature on the subject is vast.[69] [70] [71] [72] [73] [74] [75] [76] [77]

We became involved in the development of green-chemistry-compliant methods for the synthesis of polyconjugate materials due to an increasing commitment to technological transfer. Companies becoming involved in printed (opto)electronics rapidly realized that popular materials are too complex, too expensive, and are prepared through routes that produce way too much waste.[39] [40] [78] [79] Successful scaling up of promising materials becomes much easier if sustainability is introduced as a criterion already at the design level and not as an afterthought once the molecular structure has already been defined.

Over the years, taking vast inspiration from other researchers working in the field as well as from leading figures developing sustainable synthetic methods to produce natural compounds and drugs, we developed tools to tackle all such issues. In the following paragraphs we will use our contributions as a guideline to highlight the pivotal roles played by micellar catalysis, solventless and mechanochemical reactions, direct arylation reactions, and rational design in improving the sustainability and efficiency of relevant multistep protocols, leading to both established and new materials.

2.1

Micellar Catalysis

2.1.1

General Features of the Approach

The key concept of micellar catalysis is that poor and even negligible solubility in water is not a limit to carrying out organic reactions in such solvent. Quite the contrary, under appropriate conditions, such reactions can outperform identical ones performed in standard organic solvents in terms of yield, purity, and reaction time, while also working at lower temperatures.[80] [81] [82] [83] [84] [85] [86] [87]

Such a remarkable result is made possible using suitable surfactants, along with reagents and catalysts. Most surfactants show a self-assembly behavior in water above a certain critical concentration (generally referred to as the critical micellar concentration) leading to nanostructured association colloids. The spherical micelle is the most common, but others such as wormlike micelles, vesicles, and lamellae can also be formed.[88] The common trait of all such nanostructures is the presence of a hydrophobic internal domain that can behave as a reservoir for lyophilic molecules (as is the case for most oil-in-water formulations) or as a nanoreactor when mutually reactive species are present and colocalized within such compartment. Over the last thirty years, such features enabled a plethora of organic transformations to be carried out in water, involving water-insoluble reagents and products. The literature on the subject is vast and can be roughly divided into two distinct periods. During the first one, researchers focused on mechanistic studies and performed reactions with the concentration of the reagent below the maximum additive concentration (MAC).[89] The latter is the upper limit concentration that still leads to a homogeneous colloidal solution; under such conditions, all chemical species are quantitatively embedded within the association colloids. Interest in micellar methods gained significant momentum in the last 15 years, mostly thanks to the intuition of Prof. Lipshutz’s group who demonstrated that micellar methods can be exceedingly efficient even well above MAC, to the point that reactions performed at 0.2–0.5 M formal concentration are now common.[59] Quite clearly, interest in micellar methods as a superior preparative tool exploded, eventually surpassing the boundaries of natural and pharmaceutical compounds, and impacting on organic semiconductors.[80]


# 2.1.2

Micelles and Conjugated Materials

We entered the field of micellar catalysis in an unusual way. Back in 2016, we were developing a formulation strategy enabling the harvesting of the triplet excited states of platinum octaethylporphyrin by 9,10-diphenylanthracene (9,10-DPA). Without going into too much detail about the peculiar photophysical process we were after, we were looking for methods to confine in as little as possible volume a large quantity of 9,10-DPA molecules, along with a few porphyrins, while working in a physiological environment.[90] Micelles turned out to be the ideal solution. Amongst all possible surfactants, we selected Kolliphor EL (K-EL) – formerly known as Cremophor EL and FDA-approved for use on human patients – due to its existing track record in injectable formulations.[91]

As expected, K-EL gave stable formulations in water at the required reciprocal concentration of porphyrin and DPA. What we did not expect was that the triplet harvesting process was efficient in air-equilibrated water. Oxygen is an exceedingly efficient triplet quencher and is obviously present in air-equilibrated water. We expected to need further encapsulation with suitable barrier layers to keep the oxygen away from the micellar core. It took us a while to realize that the ricinolein double bonds present in the lipophilic domains of K-EL react readily with oxygen and that the PEG portion present in the hydrophilic corona prevents further oxygen diffusion. We immediately realized the potential of this very peculiar feature in reactions requiring the use of oxygen-sensitive species. Being greatly involved in the synthesis of polyconjugated derivatives, we tested the use of micellar aqueous solutions of K-EL for the sustainable synthesis of a variety of conjugated biaryl building blocks of potential interest for the preparation of organic semiconductors.[92]

Zoom Image
Scheme 1 K-EL enabled Suzuki–Miyaura couplings of a series of biaryls. All reactions were performed under standard laboratory conditions.

We did not expect K-EL to perform similarly to the state-of-the-art designer surfactants introduced by Lipshutz[85] [86] , [93] [94] [95] [96] and Handa,[97] [98] amongst others. We focused instead on the role of such surfactant in keeping oxygen away from the reaction site and thus enabling efficient cross-coupling, even under standard laboratory conditions. As shown in Scheme [1], we managed to prepare a series of aromatic and heteroaromatic derivatives in good to excellent yield while working at room temperature, in water, and under air.[92] Even more importantly, we demonstrated that – in the case of solid products that could be isolated by suction filtration – we could recycle the reaction medium up to three times without the addition of a fresh aliquot of palladium catalyst, with comparable performance. Working under essentially analogous conditions, we prepared molecular materials of interest for the development of organic solar cells.[99]


# 2.1.2

Micellar Solutions and Emulsions

Micellar reactions proved to be as versatile for conjugated materials as already demonstrated for natural products and drugs. Yet, like in previous reports, the use of water alone did not universally provide suitable results. When dealing with materials with a particularly poor solubility or when extensive phase segregation during the reaction is prevented through mixing, the use of small amounts of organic solvents of relatively low toxicity proved to be a flexible fallback strategy.[100] [101] [102] [103] We were positively surprised to observe that we could keep running cross-coupling, palladium-mediated reactions under a standard laboratory atmosphere also in the presence of small amounts of toluene; that is, under emulsion conditions while still working with K-EL in this case as the emulsifier.[104]

We were able to carry out efficient SM coupling on a series of thermally sensitive latent pigment derivatives of interest for the fabrication of organic transistors, solar cells, and luminescent solar collectors.[105] [106] [107] As shown in Scheme [2], our procedure proved to be superior with respect to standard protocols performed in organic solvents and under an inert atmosphere.

Zoom Image
Scheme 2 SM reactions performed on latent pigment derivatives in water/toluene emulsions in the presence of K-EL

The use of 10 vol% of toluene in 2 wt% solutions of K-EL was the key to the next step in the drive to improve sustainability in organic materials: polymerizations. The most obvious difference between a small-molecule organic semiconductor and a conjugated polymer is the dimensions of the single macromolecule. Whereas a molecule is generally way smaller than a spherical micelle – the hydrodynamic diameter can vary, but generally fits within the 5–20 nm range – a polymer is of comparable if not larger size. This characteristic makes the use of the standard micellar model unrealistic as the growing macromolecule cannot be confined within a single micelle and thereby keep growing. Indeed, we observed that reactions performed in aqueous solutions of surfactants quickly evolved to massive phase segregation and ultimately gave low molecular weight materials, which was not the case for control experiments performed in organic solvents.[108] Figure [1b] shows one example of such extensive phase segregation phenomena. The addition of 10 vol% toluene to the reaction mixture improved dramatically the formulative state and the outcome of the polymerization. We were able to produce materials with molecular weight and dispersity comparable to or outperforming those obtained in the control experiments, leading to materials having suitable optical properties for optoelectronic applications and improved stability.[108] It should be noted that our strategy is significantly different from the well-established mini-emulsion polymerization, also based on the use of an oil-in-water emulsion. The amount of toluene we employed is very small: roughly 20 mg for every 100 mg of the monomer mixture. In contrast, in standard mini-emulsion approaches both the monomers and the growing polymer are always dissolved in the organic solvent, thus requiring at least two orders of magnitude higher amount of the former.[109] [110] In quantitative terms, we could produce two well-established and performing polymeric semiconductors – PF8BT and PF8T2 – with properties at least as good as controls made under homogeneous conditions but with an E-Factor[84] (a popular green chemistry metric corresponding to the weight ratio between waste and purified product) reduced by one order of magnitude (from few hundred to 40–50 depending on the specific polymer).

Zoom Image
Figure 1 Surfactant-enhanced polymerizations under micellar and emulsion conditions

We later realized that the use of polymeric branched emulsifiers such as K-EL in polymerization reactions has a negative impact on transport properties. During polymerization, some of the surfactant remains entrapped in a sort of interpenetrated polymeric network, leading to insulating residues within the semiconducting material that are very difficult to remove without resorting to organic-solvent-intensive purifications such as extractions and reprecipitations.

To avoid this rather serious issue for successful implementation in organic transistors – one of the main application areas of organic semiconductors – we had to move from both micellar and emulsion conditions and develop a different approach based on dispersion. However, before directly tackling polymerizations, we had to gear up the approach while still working with small molecules and explicitly address the main thorns in the side of micellar catalysis: reproducibility and scaling up.


# 2.1.3

What To Do When It Does Not Work

Micellar reactions are generally very efficient, sustainable, and applicable to a wide range of compounds. In some specific documented cases, either they do not work, or they behave unpredictably. In the particularly appropriate wording of Dr. Gallou, reactions can be capricious in terms of reproducibility and scalability.[111] One reaction in particular, the SM coupling of (4-bromophenyl)methyl acetate with 2-thienylboronic acid (Table [1]) gave us serious headaches. While performing the reaction in K-EL 2 wt% in water, we obtained dramatically different results depending on the shape of the reaction vessel (test tube or round-bottom flask) and the stirring speed. Just to give an example, running the reaction in a test tube at a stirring speed of the magnetic bar of 800 rpm gave 80% yield; the same reaction performed at 1000 rpm gave a 45% yield. Table [1] summarizes the details of the rather frustrating experimental campaign. It should be noted that the erratic behavior is not only typical of industrial surfactants like K-EL but is also observed with designer surfactants such as TPGS-750-M.[89]

Table 1 SM Coupling of (4-Bromophenyl)methyl Acetate with 2-Thienylboronic, Performed under Various Formulation Conditions

Entry

Setup

Stirring speed (rpm)

Reaction medium

Conv. (%)

1

A

800

2 wt% K-EL in H2O

80

2

A

1000

2 wt% K-EL in H2O

43

3

B

800

2 wt% K-EL in H2O

15

4

C

1000

2 wt% K-EL in H2O

65

5

C

1000

2 wt% K-EL in H2O

68

6

A

800

2 wt% TPGS-750M in H2O

50

7

A

1000

2 wt% TPGS-750M in H2O

67

8

D

1000

2 wt% K-EL in H2O

94

9

A

800

2 wt% TPGS-750M in H2O

25

10

A

1000

2 wt% TPGS-750M in H2O

72

11

A

800

(2 wt% K-EL in H2O)/acetone 9:1

70

12

A

1000

(2 wt% K-EL in H2O)/acetone 9:1

61

Our take on the issue is based on the realization that when working far above MAC, the reaction mixture cannot be described as a micellar solution because the amount of the hydrophobic phase far exceeds that of the surfactant. For liquid organics, the mixture is an emulsion; for solids, it is a dispersion. Both disperse systems are out of equilibrium microheterogeneous environments and require an energy input – in the form of heating and/or stirring – to maintain homogeneity. As even the most amateurish cook has experienced at least once in a lifetime, a good mayonnaise (a disperse system well above MAC) requires stirring but the way the mixture is stirred and for how long makes all the difference between a nice and creamy seasoning and a yellowish slime. In our case, the stirring speed and vessel geometries influenced the various transport phenomena active within the multiphase reaction system, leading to unpredictable results. When performed in plain water but under turbo-emulsification instead of magnetic bar stirring, the reaction gave a quantitative yield.

Performing reactions with a turbo-emulsifier might be efficient but it is far from sustainable due to the energy intake required. The closest disperse system to a thermodynamically stable micellar solution is the so-called spontaneous emulsion; i.e., an emulsion requiring minimal stirring to be efficiently formed. Such systems are used for example for the remediation of accidental oil spillage on open water, where efficient stirring is obviously not practical. The 4:1 wt/wt mixture of the industrial surfactant Tween 80 and lecithin (funnily enough an important ingredient of homemade mayonnaise) is capable of reducing the water/oil interface tension to almost zero, thus triggering the self-emulsification process.[112] The use of such a surfactant mixture (TL82) effectively tamed the erratic behavior of our reaction, leading to identical results irrespective of the setup and stirring speed of the reaction.[89]

The TL82 made even more for us. The main narrative of micellar chemistry is the use of the hydrophobic effect to perform efficient and sustainable chemistry in water using water-insoluble reagents. We decided to push the concept even further, challenging reagents that are insoluble in both water and organic solvents: halogenated pigments. The solubility in common organic solvents of the most problematic of such derivatives (the Class 3 derivatives of Figure [2]) is below 10–6 M. The only prior contribution describing a SM coupling on such halides required a high-temperature mechanochemical reaction, raising concern in terms of safety and scalability. We developed a safe, scalable, and simple procedure requiring the use of an aqueous solution of TL82 as the reaction medium in the absence of any organic solvent.[113] The process proved to be very efficient for bromides and demonstrated that, under the appropriate conditions, surfactant-enhanced reactions are as efficient with dispersions as they are with emulsion and micellar solutions.

Zoom Image
Figure 2 Substrates for SM couplings on organic pigments carried out under dispersion conditions

Our work with pigments represented the missing link we were looking for to carry out efficient polymerizations not only in the absence of organic solvents during synthesis but during processing as well.[66]

As we previously discussed, the main obstacle to obtaining polymers of high molecular weight was the lack of homogeneity and, hence, of efficient mass transport at increasing conversion.[108] The use of small amounts of organic solvents was instrumental in improving molecular weights but caused another problem we could not resolve. In standard microemulsion polymerizations, both the growing polymer and the reagents are dissolved in the organic solvent at least up to high molecular weights. The surfactant is only present at the interphase between the water and oil phase and does not contaminate the final material. Conversely, in our micellar polymerizations, we used very small amounts of cosolvent as a mixing aid for all components of the reaction mixture. While growing, the polymer embeds the surfactant in an interpenetrated polymer network. Complete removal of the surfactant is still possible, but this is very solvent-intensive as it requires multiple dissolution and precipitation steps.

The issue became evident while employing micellar polymerized materials in field-effect transistors. The performance of such devices is critically dependent on purity, and we observed a dramatic impact of trace amounts of surfactants reducing the characteristic mobility of over two orders of magnitude with respect to carefully purified control samples. Figure [3] shows a comparison between the NMR spectra of different samples of PF8T2 polymer prepared via standard and micellar polymerization. The presence of trace amounts of surfactant is evident. We speculated that running the polymerization in dispersion conditions (with TL82) instead of emulsion (K-EL + toluene) could solve the issue. Indeed, not only did we obtain a better overall polymerization in terms of yield and molecular weights, but we also realized that the homogeneity of the reaction mixture during polymerization could be further exploited to directly process the polymer from water instead of going through the usual isolation, purification, and molecular weight refinement in the Soxhlet extractor. On doing so, we adapted a procedure originally introduced by Turner, consisting of the purification of the reaction mixture by dialysis, followed by dilution and direct deposition in the film form via spin coating.[65] Figure [4] shows a schematic representation of the overall process.

Zoom Image
Figure 3 NMR characterization of samples of PF8T2, obtained with organic solvent (Ref, blue trace), micellar (Mic, orange trace) vs. dispersion (Disp, red trace) polymerization. The peak highlighted with the arrow at δ = 3.70 ppm corresponds to the PEG chain of K-EL. The cartoon represents the interpenetrated polymer network comprising PF8T2 and K-EL. Full details on the assignment are discussed in ref.[66] Figure reprinted with permission from Ceriani et al.[66]
Zoom Image
Figure 4 Direct processing of aqueous inks of PF8T2 polymer prepared by dispersion polymerization

The devices we obtained – never employing organic solvents – featured performances comparable to those of analogue polymers prepared and processed from organic solvents. In particular, the field effect mobility measured on aqueous vs. organic solvents devices were 1.27 × 10–3 and 0.5 × 10–3 cm2V–1s–1, comparable values within the experimental error.[66]


# 2.1.4

Introducing Selectivity

Aside from our product-oriented efforts, we studied on a more fundamental level the influence of the specific substrate/surfactant interactions to complement efficiency with selectivity. As frequently happens, an unexpected (and unwanted) result triggered our attention towards such effects. The SM coupling of 4,7-dibromo-5,6-difluoro-2,1,3-benzothiadiazole (DBBF) and thiophene-2-boronic acid is of particular interest to produce relevant donor–acceptor building blocks of polyconjugated materials.[114] [115] [116] [117] [118] The reaction is reasonably efficient in organic solvents and can also be performed satisfactorily via direct arylation.[118] Being interested in both the material itself and in the further validation of our methods, we tested the reaction under micellar conditions using K-EL as the surfactant. The coupling is surprisingly inefficient and essentially limited to the conversion of the bromide into the monosubstituted derivative TBF, irrespective of the excess of boronic acid employed.[119] Under the very same experimental conditions, different boronic acids and/or dibromide derivatives gave near quantitative conversion of the double arylation product. We observed that the reaction of TBF with excess thiophene-2-boronic acid gives the double arylation products in 2% yield only, confirming that the formation of the target DTBT is not favored while working with aqueous K-EL (Scheme [3]).

Zoom Image
Scheme 3 (a) SM reaction between DBF and thiophene-2-boronic acid. (b) SM reaction between TBF and thiophene-2-boronic acid

We speculated that the double arylation product had a particularly strong interaction with K-EL, leading to saturation of the reaction sites (the lipophilic pocket of the association colloid). To perturb such strong, and undesired, interaction, we blended K-EL with lipophilic (Span 80) and hydrophilic (Tween 80) additional surfactants, observing that a more polar environment gave preferential monoarylation over double arylation. This observation made it possible to prepare DBBF in 97% yield with an extraordinarily small E-factor of 4.8. We also devised a two-step protocol for the preparation of asymmetrically substituted double arylation products requiring a first monoarylation carried out in a polar surfactant mixture (K-EL/Tween 80, 7:3), followed by a second arylation performed in a more balanced mixture obtained by further addition of the nonpolar Span 80, along with a second aliquot of a different aryl-boronic acid.[119]

We further expanded the concept of selective reagent vs. product interactions with the surfactant in a study devoted to the preparation of a series of [1]benzothieno[3,2-b][1]benzothiophene (BTBT) arylation derivatives.[120] Such molecules are amongst the most efficient molecular semiconductors to date and – prior to our involvement – are generally accessible through inefficient and high-waste synthetic routes.[121] [122] [123] [124] [125] [126] Industrial and designer surfactants differ in the details of chemical composition but share the same set of noncovalent interactions while interacting with other components of the reaction mixture: dipole–dipole and van der Waals. The association behavior of organic semiconductors is dominated by π-stack interactions; as such, we deviated from standard design guidelines for the amphiphile, introducing, in the lipophilic region, groups featuring an extended conjugation that is thus capable of forming πp-stacking alongside van der Waals interactions. As molecules characterized by extended conjugation can possess both electron-deficient and electron-rich character, we compared the performance of two amphiphiles featuring complementary electronic character – PiNap-750M and BTBT-750M (Figure [5]) – in promoting the SM coupling of 2-bromo-[1]benzothieno[3,2-b][1]benzothiophene and phenylboronic acid using the industrial surfactant K-EL as the benchmark.

Zoom Image
Figure 5 Molecular structure of the aromatic designer surfactants BTBT-750M and PiNap-750M

K-EL does not efficiently promote the reaction, likely due to the extremely low solubility of the bromide. As expected, the addition of a little toluene improved both conversion and yield, but only upon raising the reaction temperature from 25 to 80 °C (Table [2]). Both ‘semiconducting’ surfactants gave good results without the need for a cosolvent and while working at room temperature, yet they strongly differ in the required reaction time: 48 h for BTBT-750M and just 1 h for PiNap-750M. The latter is expected to form stronger donor–acceptor ππ–ππ interactions with all BTBT electron-rich derivatives, thus promoting rapid and efficient incorporation into the lipophilic region of the association colloid. As the product is not fully planar due to the torsion angle between the BTBT core and the phenyl ring, the interaction will be stronger with the bromide, thus leading to swift and full conversion, as observed.

Table 2 SM Reactions on BTBT Derivatives

Entry

Medium

T (°C)

t (h)

Yield (%)

1

K-EL 2%

25

6

26

2

K-EL 2%/THF 9:1 v/v

25

6

50

3

K-EL 2%/toluene 9:1 v/v

25

6

45

4

K-EL 2%/toluene 9:1 v/v

80

1

90

5

BTBT-750M 2%

25

24

73

6

PiNAP-750M 2%

25

1

78

7

BTBT-750M 2%

25

48

97

8

PiNAP-750M 2%

25

1

97

The use of designer surfactants featuring selective interaction with reagents and/or products of different transformations is a general strategy that goes well beyond the manufacturing of organic semiconductors, as exemplified by the body of work of Prof. Handa and its signature surfactant FI-750-M.[98] , [127] [128] [129]


# 2.1.5

When the Micellar Route is the Only Route

In most cases, micellar protocols offer a more sustainable route for the preparation of known products. In other cases, they provide the only possible access to materials of a particularly challenging nature such as persistent, luminescent radicals. (3,5-Dichloro-4-pyridyl)bis(2,4,6-trichlorophenyl)methyl radical (PyBTM), is a luminescent persistent radical that exhibits a remarkable luminescence quantum yield of 26% at room temperature in a polymethylmethacrylate (PMMA) matrix.[130] [131] [132] Such relatively efficient emission is associated with an uncharacteristically large separation between the absorption and emission spectra (Stokes shift), ensuring minimal reabsorption of the emitted light even at very high concentrations. Being interested in applications in luminescent solar concentrators, we developed a series of PyBTM derivatives featuring extended conjugation and thus improved photon harvesting.[133] As PyBTM is an octachloro triarylmethane derivative, we devised a late-stage functionalization approach involving SM coupling of the former with several different arylboronic acids. In this case, the micellar approach (in this case carried out in K-EL solutions and the presence of 10 vol% toluene) was not simply better because of the homogeneous phase one performed in organic solvents: it was the only one that enabled isolation of the product in synthetically useful amounts (Scheme [4]).

Zoom Image
Scheme 4 Synthetic pathway leading to the luminescent radicals PyPBTM, Py-PNBTM, and PyPhBTM

We attribute the success of our approach to the particularly high effective concentration that can be reached while benefiting from the hydrophobic effect. Recently, we demonstrated that it is possible to prepare more heavily functionalized persistent radicals using the same chemistry, eventually reaching unprecedented stabilities and luminescence efficiencies.[134]

Similarly, when we targeted the highly conjugated, branched spirobifluorene derivative 4DPA – a molecule enabling for the first time the observation of intramolecular triplet–triplet annihilation – the micellar approach was the only one capable of providing the required 4 SM couplings leading to the target derivative in useful yields Scheme [5].[135]

Zoom Image
Scheme 5 Synthesis of the 4DPA derivative

Micellar methods allow the preparation of relatively complex molecules in high yield and on an industrially relevant scale, as we demonstrated with the synthesis of the fluorescent material MPDTBOP (5,6-diphenoxy-4,7-bis[5-(2,6-dimethylphenyl)-2-thienyl]benzo[c]1,2,5-thiadiazole), which we made in 78% overall yield in five steps, four of them using water as the solvent.[56]


#
# 2.2

Intrinsically More Sustainable Reactions: Direct Arylation

Micellar methods (and the correlated emulsion and dispersion routes) provide a very versatile tool to dramatically improve existing reaction protocols. Recent years have witnessed intense research efforts on the development of intrinsically more efficient and sustainable arylation methods not requiring the activation of one of the two reaction partners and collectively known as ‘direct arylation methods’. By far, the most popular is the palladium-mediated direct coupling of an aryl halide (iodides and bromides giving the best results) and a second arene, having obvious advantages in terms of atom economy and number of synthetic steps required.[57] Direct arylation methods are becoming increasingly competitive with Stille and SM, also for the manufacturing of polyconjugated molecules and polymers.[136] [137] [138] [139] [140] [141] [142] We exploited the direct (hetero)arylation polymerization (DHAP) approach for the synthesis of a polymeric latent pigment derivative, aimed at mimicking the transport properties of the very popular electron-transporting material poly{[N,N′-bis(2-octyldodecyl)-naphthalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,5′-(2,2′-bithiophene)} (P(NDI2OD-T2).[143] [144] [145] [146] [147]

As shown in Scheme [6], the PNTET2 polymer features four ester residues per repeating unit. In this form, both the starting monomers and the polymer are very soluble. By carefully optimizing reaction conditions (catalytic system, stoichiometry, reaction temperature, time, and base) we were able to obtain a number average molar mass (M n) in the order 100 Kg/mol with a dispersity of 2.8.

Zoom Image
Scheme 6 Synthesis via direct arylation of the latent pigment polymer PNTET2

Once the PNTET2 polymer is heated above 350 °C, the alcohol residues eliminate as the corresponding mixture of alkene and the naphthalene tetraester unit converts into the highly electron-withdrawing naphthalene dianhydride. Therefore, as shown in Scheme [7], the polymer converts from the flexible, soluble, and light-yellow PNTET2 form to the deep-blue, insoluble, and brittle PNDAT2 form. Such processes opens the way to the preparation of organic heterostructures through an orthogonal solvent approach.[105]

Zoom Image
Scheme 7 Thermally activated cleavage of the ester functionalities of polymer PNTET2 to give polymer PNDAT2

We also used DHAP for the preparation of a particularly challenging synthetic target: the glycol chain functionalized N-type semiconductor p(NDI-C4-TEGMe-T2).[148] Glycol chains are becoming increasingly popular as solubilizing chains for all semiconducting materials developed to work in contact with an aqueous environment, such as electrolyte-gated transistors. The said polymer is generally prepared via a Stille coupling, requiring the use of toxic aryl stannanes.[149] We developed a DHAP polymerization approach based on the conditions established by the group of Sommer (Scheme [8]).[76]

Zoom Image
Scheme 8 DHAP on a glycol chain functionalized n-type polymer

Although the M n obtained using DHAP vs. Stille polymerization was lower, we succeeded in developing a stable electrolyte-gated organic transistor using our polymer and printing as the processing method; this is also a sustainable protocol.[148] Due to the increasing success of the method in leading laboratories dedicated to the synthesis of polyconjugated materials, we attempted to merge the direct arylation and micellar approaches. Unfortunately, one of the very few weaknesses of DHA is the need for high temperatures; reactions are generally performed above 100 °C. This does not represent a limit for the use of water as the solvent per se. The use of pressurized reaction vessels enables reactions to be safely carried out in an aqueous environment well above 100 °C. Unfortunately, nonionic surfactants such as those more common in micellar chemistry possess cloud points below 100 °C. In short, at temperatures between 80 and 100 °C the surfactants become water-insoluble, preventing the formation of association colloids. Just like in the case of micellar solution reactions performed on challenging substrates, the solution to this problem is the use of a small amount of water-insoluble cosolvents such as toluene. Opposite to micellar solutions, the formation of emulsions is promoted at high temperatures, as all interfacial tensions are reduced. We thus further exploited the 10 vol% toluene in 2 wt% K-EL emulsion conditions to develop a DA method that connected the benefit of sustainability with the operational simplicity of working under air.[150] Our method benefited from the seminal work of the Leclerc group who demonstrated the compatibility of DHA with both water and oxygen, with the extra benefit of a sizeable reduction in the amount of organic solvents required.[151] The key evolutionary step of our method with respect to the previously reported approaches was the use of a cationic co-surfactant – the thermally stable Aliquat HTA-1 – and neodecanoic acid in the presence of the more widely employed pivalic acid as a ligand for palladium. The neodecanoic acid increases the lipophilicity of the catalytic system, a feature of great importance while working under micellar and emulsion conditions. Scheme [9] shows several representative examples of the conjugated building blocks we prepared, along with an estimate of the corresponding E-factors.

Zoom Image
Scheme 9 Examples of DHA reactions performed under emulsion conditions

Alongside palladium-mediated cross-coupling reactions, photoredox direct arylation protocols are gaining increasing attention, particularly when they are compatible with direct sunlight irradiation as the activation method.[152] [153] [154] [155] [156] [157] [158] [159] Photoredox arylation protocols require the same reagents as palladium mediated arylation: an aryl halide (mostly bromides) and an activated, generally electron-rich, arene. The role of the palladium catalyst is, in this case, played by a photoredox mediator (Cat), which is a molecule that – once excited in the corresponding first excited state (Cat*) – can convert the arylhalide (Ar-X) into the corresponding neutral radical (Ar) plus one equivalent of halogenide (X) (Scheme [10]). The excited photosensitizer is immediately reduced to the ground state by a sacrificial tertiary amine (NR3), consequently forming the corresponding radical cation (R3N+•). The Ar radical can either intercept the coupling partner (T), with subsequent formation of the product (Ar–T) or react with R3N+• to give the hydrodehalogenation side product (Ar–H). The reaction selectively gives Ar–T only if T is used in large excess (typically 20–50 equivalents), which clearly limits the synthetic usefulness. In the absence of NR3, only the relatively diluted radical Ar–T can reduce the Cat* intermediate, thus seriously slowing down the reaction. The use of poorly soluble tertiary amines ensures both the presence of the required reducing agent (necessary to regenerate the catalyst) and the swift removal (by precipitation) of the unwanted corresponding radical cation, thus establishing a trade-off between the selectivity and reaction speed.[160]

Zoom Image
Scheme 10 Catalytic cycle of a photoredox direct arylation

Nevertheless, the method remains attractive as the photosensitizer can be an organic molecule, thus enabling a metal-free direct arylation process. The heteroaromatic 10-(4-methoxy)phenyl-10H-phenothiazine (PTh-OMe) molecule is amongst the most efficient metal-free mediators.[152] Performing such photoredox reactions in water is not only advantageous from the point of view of sustainability but it is also more efficient overall as the protocol becomes selective for arylation over dehalogenation, even when the excess of coupling partner is kept at a minimum.[158] [161] The improved selectivity is made possible by the efficient removal of the R3N+• intermediate from the reaction site as soon as it is formed from the corresponding amine. In a microheterogeneous environment such as micellar solutions and emulsions, amines with different structures will be partitioned in the water vs. oil phase according to the corresponding polarity. In all cases the R3N+• species formed while regenerating the photocatalyst will localize in water because of its ionic nature, thus leaving the oil phase where both the target arene and bromide are localized. The selective partitioning of amines and corresponding radical cations leads to the observed arylation vs. dehalogenation selectivity.

In fact, provided that an amine with a suitable partition is selected, the sacrificial donor will be available within the micelles, but the corresponding R3N+• species will be immediately removed from the reaction site, thus preventing the hydrodehalogenation pathway. We further improved on the method by incorporating the photocatalyst within the structure of the surfactant (derivative S-PTh shown in Scheme [11]). According to this method, pyrrole and indole can be efficiently arylated with a variety of bromides in good to excellent yield, while completely avoiding the formation of the hydrodehalogenation byproduct and without resorting to either a large excess of pyrrole or the use of specialized amines.[161] We recently further investigated the role of partitioning in micellar reactions using a computational approach. The most relevant finding was that lipophilic species do not localize in the core of the micelle but rather at the boundary between the lipophilic and hydrophilic regions.[162] This is the compartment where the sudden removal of the R3N+• species from the reactive sites would be most efficient, thus further supporting our interpretation of the observed selectivity.[162]

Zoom Image
Scheme 11 Synthetic route and chemical structure of the photoredox active S-PTh surfactant and examples of S-PTh mediated photoredox arylations of pyrrole and indole

# 2.3

Sustainable Multistep Protocols

Organic semiconductors are often complex targets, requiring multistep synthetic approaches. The overall improvement of a multistep protocol requires the combination of different approaches, including, but not limited to, micellar catalysis and direct arylation. Over the last few years, we have developed several mechanochemical and solventless approaches that – along with our signature surfactant enhanced methods – enabled the efficient and green-chemistry-compliant preparation of complex derivatives, such as the well-known and commercially available hole conductor 2,2′,7,7′-tetrakis(N,N-di(4-methoxyphenyl)amino)-9,9′-spirobifluorene(Spiro-OMeTAD).[163]

The established synthesis of Spiro-OMeTAD is representative of the synthetic protocols frequently employed in the manufacturing of organic semiconductors. As such chemicals are mostly employed in proof-of-concept studies, very little attention is paid to the use of hazardous chemicals, to the amount of waste produced, or to the overall efficiency of the process. The target is the preparation of the product and not the optimization of the process, with consequences on the retail price and overall environmental impact. As shown in Scheme [12], the synthesis of the key intermediate 9,9′-spirobifluorene requires the use of hazardous Grignard or organolithium reagents on aryl halides that are troublesome to access.[164] [165] The subsequent bromination requires the use of a chlorinated solvent, and it is followed by a palladium-catalyzed Buchwald–Hartwig amination with excess 4,4′-dimethoxydiphenylamine in refluxing toluene.[166,167] The overall yield is 18%, after a laborious chromatographic purification, further impacting the overall consumption of organic solvent and leading to an E-Factor of 5300.[163] We completely redesigned the whole process, starting from an alternative preparation of the key 9,9′-spirobifluorene intermediate.

Zoom Image
Scheme 12 Example of established access to Spiro-OMeTAD (top) vs. the green-chemistry-compliant approach we developed (bottom).

The reaction of fluorene and dibenzothiophene-S,S-dioxide in the presence of KN-(SiMe3)2 as the base in anhydrous dioxane at 150 °C for 16 h was reported to give 9,9′-spirobifluorene in 45% yield after chromatographic purification.[168] As dibenzothiophene-S,S-dioxide can be prepared by the quantitative oxidation of dibenzothiophene by hydrogen peroxide in the presence of catalytic H2WO4, we deemed it promising and we performed a thorough optimization, eventually increasing the yield to 83% by using excess molten fluorene as the solvent and KOH as the base. No chromatographic purification was required. For the bromination, we opted for a micellar catalysis approach using an aqueous 4 wt% solution of the anionic surfactant Dowfax 3B2 as the reaction medium and Br2/H2O2 as the brominating agent. Extractive crystallization from a little chloroform gave the pure 2,2′-7,7′-isomer in 82% yield. For the final BH step we adopted a near-solventless approach, carrying out the reaction using a 25 wt% amount of toluene with respect to the whole reaction mixture, K3PO4 as the base, and Pd(AcO)2/X-Phos as the catalyst. We isolated the pure product in 75% yield after cold filtration through a plug of SiO2/carbon. The overall yield of the process was 50%, with an E-factor of 555, essentially one order of magnitude lower than the previously established method.[163] We later developed a whole series of new HTL materials that could be prepared by using the same BH conditions. One of them is particularly remarkable because it is a polymer with a molecular weight in line with the best-reported results in the literature for BH polymerized materials.[169]

Zoom Image
Scheme 13 Examples of tandem micellar B-H amination and acid-promoted cyclization syntheses of fluorinated acridines.

Another valuable approach to improve efficiency and sustainability involves carrying out multiple steps (each of them performed under sustainable conditions) without isolating the intermediates.

Building upon our previous experience, we developed a synthesis of fluorinated acridines – up to that point considered particularly challenging substrates – via sequential micellar BH amination/cyclization of aryl bromides.[170] As shown in Scheme [13], we challenged seven different fluorinated bromobenzaldehydes with 2-amino anisole using a 2 wt% K-EL aqueous solution as the reaction medium. The crude amination product was not isolated but directly cyclized with trifluoroacetic acid to give the final acridines in 70–90% yield. We also performed control experiments using toluene as the solvent for the BH amination, in all cases observing significantly lower yields. Some of the acridines thus prepared are potential antibiotics against resistant bacterial strains.[171]


#
# 3

Sustainability as a Design Criterion

As we have shown, sustainability can be pursued by improving the synthetic protocols according to all the available green-chemistry-compliant tools. A more drastic approach involves the design of new and different active compounds that are specifically devised to be both performing and sustainable. In other terms, the requirement for sustainability does not come as an extra feature once the target compound has been developed and successfully validated but becomes a design criterion as important as those dictated by the structure–property relationships developed in the specific area of interest. We applied this concept to a series of fluorophores developed for luminescent solar concentrators (LSCs).[172] [173] Such devices are slabs of high optical quality materials (PMMA is the most common choice) doped with luminescent compounds and acting as planar waveguides in building integrated photovoltaics.[172] To be both efficient and sustainable by design, the ideal luminophore must feature: (1) accessibility from abundant raw materials through efficient and sustainable synthetic processes; (2) low oxygen sensitivity; (3) thermal- and photo-chemical stability; (3) highly efficient fluorescence; (4) a wide Stokes shift, to minimize reabsorption losses; and, finally, (5) perfect compatibility with industrially processed polymeric matrices. Amongst the vast plethora of possible structural motifs offering high luminescence and high Stokes shift, we selected the [1]benzothieno[3,2-b][1]benzothiophene (BTBT) core due to its simple synthetic access (E-factor around 10), availability of efficient routes for late-stage functionalization and high thermal, chemical, and photochemical stability.[124] Pristine BTBT is essentially non-fluorescent, electron-rich, and, thus, potentially oxygen-sensitive and very poorly soluble. We thus developed a mechanochemical synthesis route giving access to a class of BTBT derivatives simultaneously featuring a very low E-factor, high thermal- and photochemical stability, near unity luminescence efficiency, and minimal reabsorption losses, owing to a large Stokes shift.[68] Crucially, such derivatives were also perfectly dispersible into PMMA produced via state-of-the-art industrial routes. Indeed, we fabricated LSCs with an optical power efficiency (OPE) as high as 3.0% (corresponding to an optical quantum efficiency [OQE] of 54%), which was the highest efficiency reported at the time for large-area transparent LSC devices Scheme [14].[68]

Zoom Image
Scheme 14 Mechanochemical synthesis of BTBT-Ox, BTBT-Ox2, and C8-BTBT-Ox2-C8 derivatives

In short, we treated BTBT with Oxone (potassium peroxymonosulfate) as the oxidizer under mechanochemical conditions in a ball miller. Remarkably, we could selectively isolate the S,S-dioxide (BTBT-Ox) versus the S,S,S′,S′-tetraoxide (BTBT-Ox2) simply by changing the amount of Oxone we employed (Scheme [15]B). The reaction produces no organic waste, yet, in this case, the excess of Oxone must be considered in the estimate of the still remarkably small E-factor of 22. A similar procedure leads to the oxidation products of BTBT derivatives bearing solubilizing chains.

We pushed such a ‘de novo’ design approach even further, while looking for similarly luminescent materials, operating at a lower emission wavelength.[67] A crucial characteristic of such molecules is the presence of strong charge-transfer interactions in a donor–acceptor–donor structure constituted by easily accessible building blocks that can be combined via sustainable chemistry approaches. We selected the 4,7-dithien-2-yl-2,1,3benzothiadiazole (DTB) structural motif as it features efficient emission, high Stokes shift, and high molar absorptivity.[174] DTB is a simplified analogue of the better performing but synthetically challenging diaryloxybenzoheterodiazole derivative, requiring the use of 4,7-dibromo-5,6-difluoro-2,1,3-benzothiadiazole (DBBF) as the key starting material.[175] Based on such a core, we designed five conjugation extended derivatives that we could prepare by exclusively employing green-chemistry-compliant approaches. Prior to the synthesis, we evaluated the potentialities of the different structures via a computational approach, confirming that all of them had potential. Scheme [15] shows the synthesis and corresponding yield and E-factors for all derivatives. The only step we did not perform under micellar catalysis was the bromination, which was carried out in the benign benzotrifluoride solvent. Derivatives TT1 and TT3 performed particularly well for the target application. We prepared thin-film LSCs and performed a comparative characterization with respect to devices containing the perylene dye LR305, the acknowledged standard in the field. Our intrinsically sustainable luminophores gave superior results, particularly for large-area devices. To fully appreciate the relevance of the result, it should be noted that the E-factor for the preparation of LR305 on a comparable scale is greater than 1000.[67]

Zoom Image
Scheme 15 Synthesis of low optical gap materials via green-chemistry-compliant approaches

# 4

Conclusion

For the best part of my independent career, I have designed, synthesized, and tested new conjugated materials with the aim of developing structure–property relationships. Along the way, with the invaluable help of talented students and collaborators, we made a few performing materials that brought us in contact with companies for scaling up and technical transfer. What at the time we believed to be the end of the journey, turned out to be a new starting point as we had to go back to the drawing board to design completely different synthetic approaches and new materials. Starting from the adaptation of existing micellar chemistry protocols to the synthesis of small, conjugated molecules in water, we developed our own surfactants and methods. We rapidly gained access to derivatives of increasing synthetic complexity, eventually extending the approach to the preparation of polymers. In doing so, we did not restrain ourselves to the use of aqueous surfactant solutions as reaction medium, but we also considered plain water and even solventless approaches. We are still barely scratching the surface of the vast reservoir of green chemistry methods developed in other fields, yet we already managed to reduce by an order of magnitude the amount of waste associated with the production of known and efficient molecular and polymeric organic semiconductors. The next challenge will be to extend the green chemistry approaches to processing, a topic where we recently demonstrated very promising proof-of-concept results.


#
#

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

Over the last eight years, the development of green methods for the synthesis and processing of conjugated materials has been one of the core topics of our group. None of the results that we have described would have been possible without the help of several very talented PhD students and postdoctoral researchers. All of them are gratefully acknowledged, with a special mention for Erika Ghiglietti, Myles Rooney, Chiara Ceriani, Adiel Calascibetta, Alessandro Sanzone, Francesca Pallini, Anita Zucchi, and Sara Mecca.


Corresponding Author

L. Beverina
University of Milano-Bicocca, Department of Materials Science and INSTM
Via R. Cozzi 55, 20125 Milano
Italy   

Publication History

Received: 21 September 2023

Accepted after revision: 13 October 2023

Accepted Manuscript online:
13 October 2023

Article published online:
20 November 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom Image
Zoom Image
Zoom Image
Zoom Image
Zoom Image
Scheme 1 K-EL enabled Suzuki–Miyaura couplings of a series of biaryls. All reactions were performed under standard laboratory conditions.
Zoom Image
Scheme 2 SM reactions performed on latent pigment derivatives in water/toluene emulsions in the presence of K-EL
Zoom Image
Figure 1 Surfactant-enhanced polymerizations under micellar and emulsion conditions
Zoom Image
Figure 2 Substrates for SM couplings on organic pigments carried out under dispersion conditions
Zoom Image
Figure 3 NMR characterization of samples of PF8T2, obtained with organic solvent (Ref, blue trace), micellar (Mic, orange trace) vs. dispersion (Disp, red trace) polymerization. The peak highlighted with the arrow at δ = 3.70 ppm corresponds to the PEG chain of K-EL. The cartoon represents the interpenetrated polymer network comprising PF8T2 and K-EL. Full details on the assignment are discussed in ref.[66] Figure reprinted with permission from Ceriani et al.[66]
Zoom Image
Figure 4 Direct processing of aqueous inks of PF8T2 polymer prepared by dispersion polymerization
Zoom Image
Scheme 3 (a) SM reaction between DBF and thiophene-2-boronic acid. (b) SM reaction between TBF and thiophene-2-boronic acid
Zoom Image
Figure 5 Molecular structure of the aromatic designer surfactants BTBT-750M and PiNap-750M
Zoom Image
Scheme 4 Synthetic pathway leading to the luminescent radicals PyPBTM, Py-PNBTM, and PyPhBTM
Zoom Image
Scheme 5 Synthesis of the 4DPA derivative
Zoom Image
Scheme 6 Synthesis via direct arylation of the latent pigment polymer PNTET2
Zoom Image
Scheme 7 Thermally activated cleavage of the ester functionalities of polymer PNTET2 to give polymer PNDAT2
Zoom Image
Scheme 8 DHAP on a glycol chain functionalized n-type polymer
Zoom Image
Scheme 9 Examples of DHA reactions performed under emulsion conditions
Zoom Image
Scheme 10 Catalytic cycle of a photoredox direct arylation
Zoom Image
Scheme 11 Synthetic route and chemical structure of the photoredox active S-PTh surfactant and examples of S-PTh mediated photoredox arylations of pyrrole and indole
Zoom Image
Scheme 12 Example of established access to Spiro-OMeTAD (top) vs. the green-chemistry-compliant approach we developed (bottom).
Zoom Image
Scheme 13 Examples of tandem micellar B-H amination and acid-promoted cyclization syntheses of fluorinated acridines.
Zoom Image
Scheme 14 Mechanochemical synthesis of BTBT-Ox, BTBT-Ox2, and C8-BTBT-Ox2-C8 derivatives
Zoom Image
Scheme 15 Synthesis of low optical gap materials via green-chemistry-compliant approaches