Efficient Photochemical Vapor Generation from Low Concentration Formic Acid Media

Herein, we report on surprisingly efficient photochemical vapor generation (PVG) of Ru, Re, and especially Ir, achieved from very dilute HCOOH media employing a thin-film flow-through photoreactor operated in flow injection mode. In the absence of added metal ion sensitizers, efficiencies near 20% for Ir and approximately 0.06% for Ru and Re occur in a narrow range of HCOOH concentrations (around 0.01 M), significantly higher than previously reported from conventionally optimized HCOOH concentrations (1–20 M). A substantial enhancement in efficiency, to around 9 and 1.5%, could be realized for Ru and Re, respectively, when 0.005 M HCOONa served as the PVG medium. The addition of metal ion sensitizers (particularly Cd2+ and Co2+) to 0.01 M HCOOH significantly enhanced PVG efficiencies to 17, 2.2, and 81% for Ru, Re, and Ir, respectively. Possible mechanistic aspects occurring in dilute HCOOH media are discussed, wherein this phenomenon is attributed to the action of 185 nm radiation available in the thin-film flow-through photoreactor. An extended study of PVG of Fe, Co, Ni, As, Se, Mo, Rh, Te, W, and Bi from both dilute HCOOH and CH3COOH was undertaken, and several elements for which a similar phenomenon appears were identified (i.e., Co, As, Se, Te, and Bi). Although use of dilute HCOOH media is attractive for practical analytical applications employing PVG, it is less tolerant toward dissolved gases and interferents in the liquid phase due to the likely lower concentrations of free radicals and aquated electrons required for analyte ion reduction and product synthesis.

Instrumentation.Standard solutions were introduced into a stream of the photochemical medium with the aid of an injection valve (0.5 mL sample volume).Delivery at an arbitrary flow rate to the photoreactor was undertaken using a peristaltic pump (Reglo Digital, Ismatec) which was also used to evacuate waste from the gas-liquid separator (GLS).All connecting tubing was made of PTFE (i.d. 1 mm) with the exception of the Tygon pump tubing.The thinfilm flow-through photoreactor was a 19W low-pressure mercury discharge lamp (Jitian Instruments Co., Beijing, China) internally fitted with three efficiently irradiated lengths of synthetic quartz tubing (total volume ≈0.72 mL) as well as two short quartz segments on either end of the photoreactor (≈0.25 mL) which serve as exterior inlet and outlet sample connection ports.These segments are not efficiently irradiated and are thus not taken into consideration when irradiation times are calculated.The effluent stream was mixed with a flow of Ar carrier and directed to a plastic GLS made from a polypropylene centrifuge vial of 15 mL internal volume.A 400 mL min -1 flow of Ar carrier was used to provide efficient release of generated S-4 volatile species and their transport from the GLS to the ICPMS.The outlet of the GLS was connected via PTFE tubing (2 mm i.d.× 40 cm long) to an Agilent 8900 ICPMS/MS via an ultra-high matrix introduction (UHMI) port located downstream of a Scott double-pass spray chamber.Carrier liquid (2% (m/v) HNO3) was mixed with an internal standard (IS) solution prepared in 2% (m/v) HNO3 and was concurrently introduced into the spray chamber via a MicroMist nebulizer (Burgener Research Inc., Mississauga, Canada).The liquid carrier channel was equipped with a manual injection valve (0.5 mL sample loop volume) and this arrangement was exclusively utilized for estimation of overall PVG efficiency (see Section "Procedure, Data Evaluation, and Conventions").

S-5
Detection was achieved using an Agilent 8900 triple quadrupole ICPMS operating in timeresolved analysis and single quadrupole modes.Optimal plasma settings for PVG measurements and selected analyte and IS isotopes monitored are summarized in Table S1.A maximum of three analytes was monitored along with at least one IS selected on the basis of similar m/z and 1 st ionization potential of the analytes during tests of PVG of other analytes from dilute media.Introduction of a He gas flow (4.1 mL min -1 ) into the reaction/collision cell served to exclude eventual contributions from any polyatomic interferences to measured baseline intensity, e.g., 40 Ar 16 O + on 56 Fe + , 40 Ar 18 O 1 H + on 59 Co + , 40 Ar 38 Ar + on 78 Se + , and particularly when high concentrations of metal ion sensitizers (in mg L -1 ) were utilized for PVG, i.e., 40 Ar 61 Ni + and 38 Ar 63 Cu + on 101 Ru + ).This approach was found essential when low PVG efficiencies of investigated analytes were encountered.Uncertainties are represented as combined SD.

RF
The reduction capabilities of the dilute HCOOH medium were also examined with different oxidation states of the elements.Ir 3+ and Ir 4+ were chosen for this purpose, employing standards prepared from solid hexachloroiridate(III) and hexachloroiridate(IV).No significant difference in peak area response was obtained with Ir 3+ and Ir 4+ standards when 0.01 M, 0.1 M, 1 M, and 10 M HCOOH were employed for PVG at a sample flow rate of 1.5 mL min -1 , hence, the reduction capability is sufficient to reduce both species to Ir 0 .
Other dilute media were also examined for PVG of Ru, Re, and Ir.Absolutely no response was obtained for Ru and Ir using 0.005-1.0M CH3COOH while an 83-fold lower response was observed for Re at 0.01 M CH3COOH in comparison to that measured using 0.01 M HCOOH.
Addition of 0.001 M CH3COOH to a 0.01 M HCOOH medium led to around 20% decrease in PVG efficiency for both Ru and Re but not for Ir.The decrease in PVG efficiency to around 50% of that achieved with neat 0.01 M HCOOH was identified at 0.005 M CH3COOH for Ru, 0.01 M CH3COOH for Re and 0.02 M CH3COOH for Ir.To some extent, this may also be attributed to a sequestering of CO needed for synthesis of the volatile metal carbonyl species due to its reaction with H3C • to yield acetyl radicals, 1 altering the chemical synthesis of the desired products.

S-8
PVG from Dilute HCOONa Media.Uncertainties are represented as combined SD.

S-9
Impact of Metal Ion Sensitizers.Using a dilute HCOOH medium, the effect of the individual metal ion sensitizers on overall PVG efficiency of Ru, Re, and Ir was examined (Figures S6A-C).The metal ions (Cd 2+ , Co 2+ , Cu 2+ , Fe 2+ , Mn 2+ , and Ni 2+ ) were chosen on the basis of their previously reported efficacy with PVG of various analytes 2,3 and were added only to the (mixed) analyte standards prepared in 0.01 M HCOOH and not to the carrier medium, which contained only 0.01 M HCOOH and into which flow the standard was injected.
PVG of Re was impacted the most by the addition of Cd 2+ ions, wherein a concentration of 5 mg L -1 Cd 2+ increased the response by approximately 32-fold.In addition to Cd 2+ , significant positive effects were identified for Fe 2+ and possibly also for Co 2+ ions, but these were much smaller than those arising with Cd 2+ .Both Mn 2+ and Ni 2+ ions exhibited negative effects in the tested range of concentrations (10-500 mg L -1 for Mn 2+ and 1-50 mg L -1 Ni 2+ ) as the response measured with the addition of sensitizers was always lower than that without added sensitizer (not evident in Figure S6B).

Figure S4 .
Figure S4.Influence of HCOONa fraction in the photochemical medium on PVG efficiency of Ru, Re, and Ir at a sample flow rate 2 mL min -1 (IT = 22 s).The sum of concentrations of HCOOH and HCOONa was always 0.01 M. 10 µg L -1 Ru and Re and 0.1 µg L -1 Ir were employed for PVG.Uncertainties are represented as combined SD.

Figure S5 .
Figure S5.Influence of sample flow rate using 0.005 M HCOONa as the photochemical medium on PVG efficiency of Ru, Re, and Ir determined with 5 µg L -1 Ru and Re and 0.1 µg L -1 Ir.

Figure S6 .
Figure S6.Effect of various metal ion concentrations on PVG efficiency of A) Ru, B) Re, and C) Ir using 0.01 M HCOOH and sample flow rate 2 mL min -1 (IT = 22 s). 10 µg L -1 Ru and Re and 0.1 µg L -1 Ir were employed for PVG.Uncertainties are represented as combined SD.The range 0.5 to 2000 mg L -1 is presented on a logarithmic scale.