A recent mechanistic study of transition-metal nanocluster formation and agglomeration (Besson, C.; Finney, E. E.; Finke, R. G.
J. Am. Chem. Soc. 2005,
127, 8179) identified two types of agglomeration for the first time, specifically bimolecular agglomeration of nanoclusters B (i.e., B + B → C; rate constant
k3) and a new step of autocatalytic agglomeration of nanoclusters B with larger already somewhat agglomerated nanoclusters and/or bulk metal, C (i.e., B + C → 1.5C; rate constant
k4). Herein, this two-step, parallel-path agglomeration mechanism is independently tested by the temperature-induced agglomeration of the prototype preformed, pre-isolated P
2W
15Nb
3O
629−-stabi
lized Ir(0)
~900 transition-metal nanoclusters undergoing cyclohexene hydrogenation and concomitant agglomeration. The resulting
k3 and
k4 rate constants are measured as a function of the temperature, yielding the previously unavailable Δ
H and Δ
S for each type of nanocluster agglomeration under the specific reaction conditions, which include the presence of cyclohexene and H
2, Δ
H3 = 6.2(3) kcal/mol, Δ
S3 = −46(2) eu and Δ
H4 = 18(1) kcal/mol and Δ
S4 = −2.5(2) eu (standard state = 1 M). The interesting activation parameters suggest that the
k3 agglomeration step may be associatively activated, while the
k4 step appears to be dissociatively activated, for reasons discussed in the main text. Also reported is the attempted agglomeration of preformed, isolated Ir(0)
~900 nanoclusters with added salt, [Bu
4N][BF
4]. Reported primarily in the Supporting Information are extensive efforts attempting to achieve the in-principle ideal goal of measuring agglomeration kinetics (
k3 and
k4) simultaneously with nucleation and growth kinetics (
k1 and
k2) in the presence of pyridine as a known agglomerating agent. Overall, the successful kinetic measurements of the
k3 and
k4 agglomeration steps for pre-isolated Ir(0)
~900 nanoclusters provided herein are significant in seven ways: (i) they independently verify the only recently discovered two types of agglomeration, bimolecular (k
3) and autocatalytic (k
4) agglomeration; (ii) the temperature dependence of the k
3 and k
4 processes provide the first activation parameters for these processes and yield the previously unavailable insights of their apparently associatively activated (
k3) and dissociatively activated (
k4) natures, at least under the reaction conditions with olefin and hydrogen present; and (iii) the two rate constants
k3 and
k4 define the term “nanocluster stabi
lity” in solution rigorously and in an experimentally testable way for the first time. In addition, (iv) the present studies serve as “proof of concept” that the measurement of agglomeration rate constants is a viable way to rank quantitatively and therefore distinguish transition-metal nanocluster stabi
lizers; (v) the results show that added salts, such as [Bu
4N][BF
4], are ineffective in agglomerating at least highly stabi
lized P
2W
15Nb
3O
629−-
ligated Ir(0)
~900 nanoclusters; and (vi) the results give some insight into the relative sizes of agglomerated C (≥Ir(0)
~2000) versus that of the starting Ir(0)
~900 nanoclusters, B. Finally, (vii) the results also make apparent that the study of nanocluster agglomeration, by additional physical methods and using other, to-be-developed agglomerants, as a preferred way to quantitate nanocluster stabi
lity, remains as an important research challenge. Some thoughts about what additional physical methods may provide the best avenues for future studies are briefly discussed in the Summary.