Electrochemically Mediated Reversible Addition–Fragmentation Chain-Transfer Polymerization

Abstract

An electrochemically mediated reversible addition-fragmentation chain-transfer polymerization (eRAFT) of (meth)acrylates was successfully carried out via electroreduction of either benzoyl peroxide (BPO) or 4-bromobenzenediazonium tetrafluoroborate (BrPhN₂⁺) which formed aryl radicals, acting as initiators for RAFT polymerization. Direct electroreduction of chain transfer agents was unsuccessful since it resulted in the formation of carbanions by a two-electron transfer process. Reduction of BrPhN₂⁺ under a fixed potential showed acceptable control, but limited conversion due to the generation of a passivating organic layer grafted on the working electrode surface. However, using fixed current conditions, easier to implement than fixed potential conditions, conversions > 80% were achieved. Well-defined homopolymers and block copolymers with a broad range of targeted degrees of polymerization were prepared.

Graphical abstract

1. Introduction

Reversible deactivation radical polymerization (RDRP) methods have been employed to prepare polymers with pre-determined molecular weight (MW), low dispersity (Ð), controlled architecture, and preserved chain-end functionality.¹ Reversible radical trapping or degenerative chain-transfer assure concurrent growth of all polymer chains, and are employed to extend the lifetime of propagating chains from seconds to hours or even days.^1,2 Two most often used RDRP methods include atom transfer radical polymerization (ATRP)^3–6 and reversible addition-fragmentation radical polymerization (RAFT) polymerization.^7,8

ATRP is based on converting propagating radicals (P_n^•) into dormant alkyl halide species (P_n-X) by transition metal complexes, usually X-Cu^II/L (L = ligand) (Figure 1A).^3,5 The resulting Cu^I/L complex reactivates the dormant species, P_n-X. The dynamic exchange between active and dormant states enables the simultaneous and uniform growth of all chains with pre-determined molecular weights and high chain-end fidelity.

Figure 1.

(A) Mechanism of eATRP. (B) CV of 2×10⁻³ M [Br-Cu^II(tris(2-methylpyridyl)amine]⁺ in DMSO + 0.1 Et₄NBF₄, at T = 25°C and scan rate 0.2 V s⁻¹.

In ATRP, the active Cu^I/L catalyst can be (re)generated by (photo)chemical or electrochemical methods.^9–13 Electrochemically mediated ATRP (eATRP) relies on the well-defined and reversible redox behavior of Cu/L complexes (Figure 1B), which allows for a precise control of the polymerization. For example, the [Cu^I]/[Cu^II] ratio can be set by adjusting the applied potential, which in turn determines the polymerization rate;^9,14 eATRP can be switched “on” and “off” by shifting the potential during polymerization.^9,15 Moreover, Cu can be easily recycled by electroplating onto the working electrode (WE), and then oxidized back into solution and re-used multiple times.^14,16 eATRP has been applied to various monomers,^9,15,17 preparing copolymers with different architectures^18,19 in a range of reaction media.^20,21 Computational simulations were also conducted for eATRP, elucidating the key parameters.²² The application of Fe-based catalyst broadened the application of eATRP.²³

Another important RDRP procedure is RAFT polymerization (Figure 2),^8,24–27 which is based on degenerative transfer, and is mediated by chain transfer agents (CTAs) such as dithioesters, dithiocarbamates, trithiocarbonates, or xanthates.²⁵ RAFT polymerization is compatible with many vinyl monomers and solvents, including both homogeneous and heterogeneous conditions.²⁸

Figure 2.

Mechanism of RAFT polymerization.

A continuous supply of radicals (i.e. “initiation”) is essential in RAFT due to inevitable termination process. Initiation can be accomplished by thermal decomposition of radical sources such as azobisisobutyronitrile (AIBN) between 50 and 70°C. However, such temperatures may be not suitable for some functional monomers²⁹ or for biological or thermo-responsive systems.³⁰ Thus, alternative radical generation methods were used for RAFT at ambient temperature, including zerovalent metals,^31–33 redox reactions,^34,35 γ-rays,³⁶ and light via the iniferter³⁷ or the photoinduced electron transfer processes.^38,39

In this work, we investigated the use of electrical current for RAFT polymerization, inspired by successful eATRP procedure and by modification of electrodes in the presence of RAFT agents.^40–42 However, an electrochemically mediated RAFT (eRAFT) is more challenging than eATRP. Cu/L complexes for ATRP have a well-defined and reversible redox behavior,⁴³ whereas the electrochemical reactivity of RAFT agents is mostly unexplored, and could result in irreversible redox processes that cannot be exploited to generate radicals. Thus, we first studied the redox properties of common CTAs (Figure 3B), which were found to be unsuitable for the direct electrogeneration of radicals by reduction on common electrodes. To circumvent this limitation, we employed electroreduction of common radical initiators, such as benzoyl peroxide (BPO), or a diazonium salt, 4-bromobenzenediazonium tetrafluoroborate (BrPhN₂⁺). As depicted in Figure 3A, reduction of both compounds generated radicals at ambient temperature, triggering controlled eRAFT polymerization of methyl methacrylate (MMA), n-butyl acrylate (BA), and tert-butyl acrylate (tBA).

Figure 3.

Structures of (A) radical sources and (B) CTAs used in this work. (C) Schematic representation of eRAFT polymerization, where Y is a chain-transfer moiety, M = monomer, k_p = propagation rate constant.

Reduction of diazonium salts occurred at more positive potentials and was further investigated under both fixed potential and fixed current conditions. The rate of radical generation was controlled by electrical current or potential, providing well-defined polymers with variable degrees of polymerizations (DPs) and good retention of chain-end functionality.

2. Results and Discussion

Electrochemical properties of RAFT chain transfer agents

Cyclic voltammetry (CV) of CPAD (4-cyano-4-(phenylcarbonothioylthio)pentanoic acid) and DDMAT (2-(dodecylthiocarbonothioylthio)-2-methylpropionic acid) gave irreversible reduction peaks at −1.00 V and −1.25 V vs. saturated calomel electrode (SCE), Figure 4. These values are comparable to the ones obtained in a previous report.⁴⁴ Quantitative reduction of CPAD by electrolysis at −1.4 V vs. SCE required two electrons per CPAD molecule (Figure S1). Both CV and electrolysis suggested that the weak C–S bond was irreversibly cleaved, with reduction of both ensuing fragments to anions, without generating radicals (Eq. 1). A similar behavior was observed for the reduction of alkyl halides used as ATRP initiators, which are also characterized by a weak bond, C–Br.⁴⁵ This mechanism is fundamentally different from the one-electron photoreduction in the photoinduced electron transfer (PET)-RAFT process.³⁹ Indeed, bulk electroreduction of the CTAs in the presence of monomer did not yield any polymer, confirming that the reduction process did not produce any long-lived free radicals. Therefore, electroreduction of radical initiators was tested in place of the electroreduction of CTAs.

Figure 4.

CV of 10⁻³ M CPAD and DDMAT in DMF + 0.1 M n-Bu₄NPF₆. T = 25°C, v = 0.1 V s−¹.

Eq. 1

Electrochemical Properties of Benzoyl Peroxide

Benzoyl peroxide (BPO) is a thermal initiator with a half-life time of ~1 h at 90°C.⁴⁶ However, BPO is essentially stable at ambient temperature and without application of any electrical current, giving < 1% of MMA conversion after 20 h (Table 1, entry 1). CV showed that BPO is irreversibly reduced at cathodic peak potential E_pc,_BPO = − 0.83 V vs. SCE (Figure 5). The peak significantly overlapped with the reduction of CPAD (E_pc,CPAD = − 1.00 V vs. SCE), which could be problematic due to concurrent electroreduction of CTA.

Table 1.

eRAFT by Electroreduction of BPO.^a

Entry	M	[M]/[CPAD]/[BPO]	Eapp	Time (h)	Conv (%)	kpapp b (h−1)	Mn,th c (×10−3)	Mn,app (×10−3)	Mw/Mn
1	MMA	500/0/1	-	20	1	-	-	-	-
2	MMA	4670/0/1	Epc,BPO − 0.08 V	20	38	0.023	-	71.1	1.78
3	MMA	500/1/1	Epc,BPO − 0.17 V	20	<5	-	-	-	-
4	MMA	500/1/1	Epc,BPO − 0.05 V	20	22	0.013	10.9	7.2	1.17
5	MMA	500/1/1	Epc,BPO + 0.34 V	20	0	-	-	-	-
6d	BA	500/1/1	Epc,BPO	20	25	0.015	10.0	7.6	1.16

^a[MMA] = 4.67 M (in DMF, 50% v/v), V_tot = 30 mL, [Et₄NPF₆] = 0.1 M, T = 25 °C.

^bThe slope of the ln([M]₀/[M]) vs. time plot.

^cM_n,th = [M]/[I] × M_M × conversion + M_CTA.

^dBA and DDMAT replaced MMA and CPAD.

Figure 5.

CV of (A) 7×10⁻³ M CPAD, (B) 10⁻² M BPO, and (C) 7×10⁻³ M CPAD + 10⁻² M BPO in DMF + 0.1 M Et₄NPF₆, T = 25°C, v = 0.1 V s⁻¹. The circles represent E_app during polymerization.

The overall electro-reductive cleavage of BPO to two anionic fragments requires two electrons (Eq. 2–3).⁴⁷ However, it involves the formation of more stable radical intermediates, which could initiate a polymerization in the presence of monomer (Eq. 4). Indeed, electroreduction of BPO at E_app = E_pc,BPO − 0.08 V in the presence of MMA gave polymers in good yield via free radical polymerization (FRP, Figure S2 and Table 1, entry 2).

$${(\text{RO})}_2+{\mathrm{e}}^{-}={\text{RO}}^{·}+{\text{RO}}^{-}$$ Eq. 2

$${\text{RO}}^{·}+{\mathrm{e}}^{-}={\text{RO}}^{-}$$ Eq. 3

$${\text{RO}}^{·}+\mathrm{M}={\mathrm{P}}_{\mathrm{n}}{ }^{·}$$ Eq. 4

eRAFT with BPO

In the presence of CPAD and BPO, MMA was polymerized at three different applied potentials shown by the circles in Figure 5 (entries 3–5 in Table 1). At E_app = E_pc,BPO − 0.17 V, < 5% conversion was detected. The potential was too negative, resulting in decomposition of the CTA. At E_app = E_pc,BPO − 0.05 V, controlled polymerization with a linear semi-logarithmic kinetics and low Ð was observed (Figure 6), but M_n was 30% lower than theoretical value due to chains generated by BPO “electro-decomposition”. At E_app = E_pc,BPO + 0.34 V, reduction of BPO was insufficient and no polymer was generated. The large overlap between the reduction waves of CPAD and BPO narrowed the useful potential window to conduct an eRAFT, thus only E_app ≈ E_pc,BPO could be successfully applied.

Figure 6.

eRAFT of MMA by electroreduction of BPO. (A) Semilogarithmic kinetic plot and (B) gel permeation chromatography (GPC) of the obtained polymer, at E_app = E_pc,BPO − 50 mV (Table 1, entry 4). Reaction conditions: [MMA] = 4.67 M (in DMF, 50% v/v), V_tot = 30 mL, [MMA]/[CPAD]/[BPO] = 500/1/1, [Et₄NPF₆] = 0.1 M, T = 25 °C.

BA was also polymerized in the presence of DDMAT and BPO, at E_app = E_pc,BPO (Table 1, entry 6). Despite the smaller overlap between the reduction waves of DDMAT and BPO, slow polymerization and M_n lower than theoretical values were observed. The limited potential window available with BPO prompted us to use a radical source with a more positive potential.

Electrochemical properties of BrPhN₂⁺

Reduction of BrPhN₂⁺ gave an irreversible reduction peak at about −0.1 V vs. SCE, which is ~ 1 V more positive than reduction of CPAD (Figure 7). Therefore, BrPhN₂⁺ could be reduced without affecting the CTAs. The reduction of BrPhN₂⁺ is a well-known process that generates very reactive bromophenyl radicals (BrPh^•). They are so reactive that they can quickly graft onto any electrode surface, forming a multi-layered coating of branched bromobenzenes.⁴⁸ However, some radicals can escape to solution to initiate and sustain the RAFT polymerization. Only a small amount of radicals is required to sustain the process because radical concentration is typically low (< 10⁻⁸ M). Therefore, we tested electrogeneration of radicals via reduction of BrPhN₂⁺.

Figure 7.

CV of 10⁻³ M BrPhN₂⁺ and 10⁻³ M CPAD. The circle represent E_app during polymerization.

Electrogeneration of Radicals by BrPhN₂⁺

FRP of BA was conducted by reducing BrPhN₂⁺ at $E_{\text{app}}=E_{{\text{pc,BrPHN}}_2{ }^{+}}$ on a Pt foil electrode (Table 2, entry 1). PBA was formed reaching 75% monomer conversion in 16 h and high molecular weight, M_n = 681,000 (Figure S5), typical for FRP. This confirmed that electroreduction of BrPhN₂⁺ generated aryl radicals that could initiate polymerization significantly faster than with BPO.

Table 2.

eRAFT of BA and MMA by reduction of BrPhN₂⁺ under potentiostatic conditions.^a

Entry	M	Time (h)	[M]/[DDMAT]/[BrPhN2+]	Qb (C)	Conv (%)	kpapp c (h−1)	Mn,thd (×10−3)	Mn,app (×10−3)	Mw/Mn
1	BA	16	500/0/1	−0.5	75	0.087	----	681.0	1.44
2	BA	4	500/1/10	−11.6	60	0.248	38.9	18.9	6.23
3	BA	20	500/1/1	−7.0	48	0.183	31.1	24.4	1.27
4	BA	20	500/1/0.5	−6.0	25	0.079	16.3	13.8	1.15
5e	MMA	48	500/1/10	−0.5	40	0.017	20.0	19.8	1.20

^a[BA] = 3.49 M (in DMF, 50% by v/v), V_tot = 8 mL, WE = Pt mesh (except for entry 1: Pt foil), E_app = −0.1 V vs. SCE. E_app was close to E_pc of the irreversible reduction peak of BrPhN₂⁺.

^bConsumed charge, calculated from cathodic current profile recorded during electrolysis.

^cThe slope of the ln([M]₀/[M]) vs. time plot. k_app for entries 2–4 was measured in the first 4 h.

^dM_n,th = [M]/[CTA] × M_M × conversion + M_CTA.

^eMMA = 4.67 M in DMF (Figure S7).

Potentiostatic eRAFT with BrPhN₂⁺

Potentiostatic eRAFT was conducted under a fixed $E_{\text{app}}=E_{{\text{pc,BrPHN}}_2{ }^{+}}\approx -0.1\mathrm{V}$vs. SCE, using DDMAT as chain transfer agent for BA polymerization. Figure 8 shows the influence of the [DDMAT]/[BrPhN₂⁺] ratio on eRAFT. The rate of polymerization increased with [BrPhN₂⁺] due to faster reduction, which led to a higher concentration of radicals. However, the addition of too much diazonium salt was detrimental to polymerization control. The final M_w/M_n was 6.2 for [DDMAT]/[BrPhN₂⁺] = 1/10 (Table 2, entry 2), whereas M_w/M_n was 1.3 for [DDMAT]/[BrPhN₂⁺] = 1/1.

Figure 8.

eRAFT of BA under potentiostatic conditions. (A) Polymerization kinetics; (B) MW and Ð evolution with conversion. Reaction conditions: [BA] = 3.49 M (in DMF, 50% by v), [BA]/[DDMAT] = 500/1, [Et₄NPF₆] = 0.1 M, T = 25 °C, E_app = − 0.1 V vs. SCE.

Despite the better control over Ð, the final MW was 15% lower than the theoretical value, suggesting that reduction of BrPhN₂⁺ generated more chains than defined by DDMAT. Continuous generation of new chains during polymerization caused some low MW tailing and higher dispersity. Decreasing the [DDMAT]/[BrPhN₂⁺] to 0.5 improved control but polymerization was slower (Figure 8).

For [BrPhN₂⁺]/[DDMAT] = 1/1 and 1/0.5, the reaction stopped after ca. 5 h at limited conversion (Figure 8A). This could be due to insufficient current flowing from the working electrode. Indeed, at fixed E_app the applied current quickly decayed (Figure 9A) due to electrografting of the insulating aryl compounds, as discussed in the next section.

Figure 9.

eRAFT by electroreduction of 7×10–3 M BrPhN₂⁺; (A) current profile during chronoamperometry at $E_{\text{app}}=E_{{\text{pc,BrPhN}}_2{ }^{+}}$ (Table 2, entry 3), and (B) potential between WE and CE during chronopotentiometry at I_app = −50 uA (Table 4, entry 3).

eRAFT of MMA was also examined at fixed E_app and in the presence of CPAD. Despite a slow polymerization rate, the MW of the resulting PMMA correlated well with theoretical values and Ð was 1.2 (Table 2, entry 5 and Figure S7). Interestingly, almost identical polymerization results were obtained for three different working electrodes, Pt, carbon felt, or graphite (Figure S7-S9). Polymers with higher Ð were obtained when using a Cu electrode (Figure S10).

Electrografting of aryl compounds

The low current and slow polymerization rate observed under potentiostatic conditions prompted us to analyze the electrografting of BrPhN₂⁺ on the surface of the WE. A variety of surfaces such as carbon, metals, and semiconductors can be grafted via a reduction of diazonium salts.^39,40 They are important agents for surface modification.^49–55 Even in the presence of monomer, a large fraction of aryl radicals was grafted onto the electrode, while some of them initiated polymerization. Although polymeric radicals cannot directly graft onto the electrode, they could covalently couple to the grafted aryl rings in the 3 or 5 position, thus increasing the thickness of organic layer.^50,56

The layer grafted during eFRP of BA (conditions as in Table 32, entry 1) was analyzed by X-ray photoelectron spectroscopy (XPS). XPS of the Pt WE showed an increase in the presence of C and Br, demonstrating grafting of an organic layer containing bromoaryl groups (Table 3 and Figure S11). The Pt content on the surface strongly decreased, indicating almost complete coverage; about 3% of Pt remained exposed, allowing for continuous electroreduction throughout the process. The macroscopic appearance of the electrode did not change after polymerization, indicating the formation of a thin layer.

Table 3.

XPS of Pt foil electrode before and after electrochemical reduction of BrPhN₂⁺.^a

Atom number %	O	C	Pt	Br
Before reaction	37.6	34.4	38.0	0
After 2 h reaction	20.7	74.9	2.8	1.6

^aThe large amount of adventitious O and C on the surface prior to deposition is due to the process employed to clean and activate the Pt electrode surface (see supporting information).

The Pt surface was also investigated by CV of a redox probe, ferrocene, which was analyzed before and after reduction of the diazonium salt (Figure 10). After deposition, ΔE_p, the difference between anodic and cathodic peak potentials increased. This indicated the presence of a layer that decreased the rate of electron transfer to ferrocene on the surface. The resistance caused by the organic layer dissipated some of the applied voltage, lowering the true applied potential in the polymerization, so that true potentiostatic conditions were not applied. This explains the decreasing polymerization rate in Figure 8A. In conclusion, the electrografting of a poorly-conductive layer on the electrode slowed BrPhN₂⁺ reduction, leading to current decay and slower radical generation.

Figure 10.

CV of 10⁻² M ferrocene (red) before and (blue) after grafting of aryl compounds. The grafting was performed by CV of 7×10⁻³ M BrPhN₂⁺ between +0.3 V and −0.3 V vs. SCE. WE = Pt disk in DMF + 0.1 M Et₄NPF₆ at 25°C.

Galvanostatic eRAFT with BrPhN₂⁺

Due to electrografting, potentiostatic condition for eRAFT resulted in limited conversion. Therefore, polymerizations under a fixed current, i.e., galvanostatic conditions, were tested to provide a continuous supply of radicals. The setup was simple, with application of a single current step during each polymerization (in comparison, a galvanostatic eATRP requires a few different current steps).^19,57 In eRAFT with BrPhN₂⁺, however, the formation of a resistive organic layer required the application of a negative E_app to supply constant current during polymerization (Figure 9B). Despite the negative E_app, reduction of monomer or solvent is not significant (see Figure S16–17).

Different applied currents (I_apps) were tested, which affected rate of radical generation (Table 4, entries 1–3 and Figure S12). Increasing I_app from −50, to −100, and to −200 μA gave more linear semilogarithmic kinetic plots, and conversion reached 80% in 10 h with the highest I_app (Figure 11). This is a noticeable improvement compared to eRAFT under potentiostatic conditions, indicating that a fixed current can provide a more efficient supply of radicals.

Table 4.

Galvanostatic eRAFT with BrPhN₂⁺.^a

Entry	Time (h)	Iapp (μA)	Qb (C)	Conv (%)	kpapp c (h−1)	Mn,thd (×10−3)	Mn,app (×10−3)	Mw/Mn
1	10	−200	−7.2	80	0.186	51.5	27.9	1.41
2	10	−100	−3.6	73	0.160	47.1	32.4	1.25
3	10	−50	−1.8	58	0.099	37.5	33.4	1.18

^aGeneral reaction conditions: [BA] = 3.49 M (in DMF, 50% by v/v), V_tot = 8 mL, [BA]/[DDMAT]/[BrPhN₂⁺] = 500/1/1, [Et₄NPF₆] = 0.1 M, T = 25°C

^bCalculated from cathodic current profile recorded during electrolysis.

^cThe slope of the ln([M]₀/[M]) vs. time plot.

^dM_n,th = [M]/[CTA] × M_M × conversion + M_CTA.

Figure 11.

Galvanostatic eRAFT of BA with different I_apps. Reaction conditions: [BA] = 3.49 M (in DMF, 50% by v), [BA]/[DDMAT]/[BrPhN₂⁺] = 500/1/1, [Et₄NPF₆] = 0.1 M, T = 25 °C. (A) Kinetic plot, (B) evolution of MW and Ð, (C) GPC traces for I_app = − 50 μA.

I_app influenced also MW and Ð. At each applied current M_n,app < M_n,th, indicating the formation of new chains during polymerization. Higher I_app generated a larger numbers of chains compared to the number defined by the chain transfer agent, and thus resulted in lower M_n,app.⁵⁸ Nevertheless, polymers with different degrees of polymerization were successfully prepared, in the range 100 – 400, with < 30% deviation from theoretical DP and M_w/M_n ≈ 1.2 (Table S1 and Figure S13-S14).

Chain extension

To evaluate the chain-end functionality of the obtained PBA, a PBA-macroCTA was chain-extended with tBA (Figure 12). The PBA-macroCTA was prepared via galvanostatic eRAFT with BrPhN₂⁺ (details in the supporting information) and had M_n,app = 12,900 (M_n,th = 13,200) and M_w/M_n = 1.19. The chain extension was carried out by both eRAFT and traditional RAFT polymerization. In the latter case, a thermal initiator, AIBN, was used at 65°C (Table 5).

Figure 12.

Chain extension of PBA-macroCTA with tBA by (A) eRAFT at room temperature and (B) RAFT polymerization initiated by AIBN at 65°C.

Table 5.

Chain extension of PBA-macroCTA by eRAFT and conventional RAFT polymerization.

Entry	Time (h)	[tBA]/[PBA-macroCTA]/[I]	Iapp (μA)	Qa (C)	Conv (%)	kpapp b (h−1)	Mn,thc (×10−3)	Mn,app (×10−3)	Mw/Mn
1d	5	500/1/1	−50	−0.9	22	0.040	26.8	20.2	1.27
2e	2	500/1/0.2	----	----	25	0.135	28.8	22.4	1.29

^aCalculated from cathodic current profile recorded during electrolysis.

^bThe slope of the ln([M]₀/[M]) vs. time plot.

^cM_n,th = [M]/[CTA] × M_M × conversion + M_CTA.

^dReaction conditions: [tBA] = 3.4 M (in DMF, 50% by v/v), V_tot = 8 mL, I = BrPhN₂⁺, [Et₄NPF₆] = 0.1 M, T = 25°C, WE = Pt mesh, I_app = −100 μA.

^eReaction conditions: [tBA] = 3.4 M (in DMF, 50% by v/v), V_tot = 8 mL, I = AIBN, T = 65°C.

Figure 12 shows the GPC traces of the macroCTA and the resulting block copolymers (PBA-b-PtBA). The block copolymer obtained from eRAFT had M_n,app = 20,200 (M_n,th = 26,800) and M_w/M_n = 1.27, while the product obtained by extension with conventional RAFT polymerization had M_n,app = 22,400 (M_n,th = 28,800) and M_w/M_n = 1.29. In both cases, a clean shift of MWD with conversion was observed. Chain extension by eRAFT showed a larger fraction of low-MW chains, further indicating that new chains were initiated during polymerization by the eRAFT process. Nevertheless, chain extension by traditional RAFT showed that the initiator prepared by eRAFT retained most of its chain-end functionality.

3. Conclusions

An eRAFT polymerization was mediated by reduction of BPO or BrPhN₂⁺ at a Pt electrode. The two initiators had very different reduction peak potentials, −0.83 V and −0.10 V vs. SCE, respectively. The CTAs were irreversible reduced at peak potential ~ −1 V vs. SCE via a two-electron reduction without formation of long-lived radical species; they could not be directly used to produce radicals by electroreduction.

Electroreduction of BPO initiated a conventional free radical polymerization, and was used for a successful but slow eRAFT at potentials E_app ≈ E_pc,BPO. Reduction of BPO partially overlapped with reduction of the CTAs, limiting the use of the peroxide for eRAFT polymerization.

Reduction of BrPhN₂⁺ occurred at much more positive potentials than reduction of the CTAs, resulting in a more effective generation of radicals to initiate a RAFT polymerization. An undesired electrografting of the aryl radicals on the electrode surface decreased its conductivity.

Well-defined polymers were prepared by applying either fixed potential or fixed current conditions. eRAFT with BrPhN₂⁺ can be carried out under galvanostatic conditions at a single current value, utilizing a simple current generator at ambient temperature. The amount of new chains generated by reduction of BrPhN₂⁺ was lowered by decreasing I_app or by decreasing the amount of diazonium salt relative to the CTA. Lower I_apps produced better controlled polymers but at slower rates. The chain-end functionality was well-maintained, as confirmed by synthesis of block copolymers with narrow molecular weight distribution.