An efficient synthesis of C-6 aminated 3-bromoimidazo[1,2-b]pyridazines

Abstract Treatment of 3-bromo-6-chloroimidazo[1,2-b]pyridazine with a variety of 1° or 2° alkylamines (2.0 equiv) with CsF (1.0 equiv), and BnNEt3Cl (10 mol %) in DMSO at 100 °C (24 hours) gave the corresponding C-6 aminated products in excellent isolated yields (79–98%; ave. yield = 92%). These conditions worked well for simple unfunctionalized 1° alkylamines, 1° amines with methylene- or ethylene-linked aromatic or heteroaromatic functionality (e.g. benzene, thiophene, furan, indole, and pyridine), and 2° alkylamines such as pyrrolidine, morpholine, and piperidine were also well tolerated. The method is cost-effect, providing C-6 aminated products in consistently high yields, while at the same time utilizing far less toxic fluoride (10 fold less), than previous fluoride-promoted aminations. Graphical Abstract


Results and discussion
We reasoned that a more efficient method could be developed which overcame the significant drawbacks and limitations associated with using expensive palladium catalyst/ ligand systems, and/or excess fluoride.Methods which decrease overall synthetic costs while at the same time reducing toxic wastestreams [22] (such as those associated with using 10 molar equivalents of toxic fluoride) [23] are in high demand.The known reactivity rates of arylhalides in SNAr (X ¼ F > Cl > Br > I), [24] suggested that more efficient in situ conversion of 9 to 12 using a more organic-soluble fluoride source (i.e., CsF vs KF) [25] might meet both critical needs, and provide significant advantages relative to a two-step synthesis involving isolation of 12 (Figure 3).Ideally, such a method would provide 12 in situ, since affordable methods for its preparation and isolation are low yielding (50%), [26] and one-pot conversion of 9 to C-6 amination products 11 would be significantly more cost effective (by a factor of approximately 100-500 based on commercial availability of 12 and/or synthetic precursors). [27]Here, we report a new method that avoids expensive transition metal catalysts, requires as little as 1.5-2.0equiv. of amine, utilizes only 1.0 equiv. of fluoride (CsF) in the presence of 10 mol% of BnEt 3 NCl-and does not require isolation of intermediate 12. Heating 9 with amines 10 at 100 � C in DMSO, 24 h, CsF (1.0 equiv.)and BnEt 3 NCl (10 mol%) gave products 11ap or SGI-1776 in consistently high yields (Table 1).Average yields for these reactions were 92%-over double the yield obtained for the optimization reaction which lacked  1. 24), and approximately double the yields obtained in the absence of phase transfer catalyst BnEt 3 NCl (46%-52%, Table 2, entries 4 and 5).In order to verify the logical assumption that the purpose of the 10-fold excess fluoride used in earlier methods, [21] was to force in situ formation of the more reactive C-6 fluorinated intermediate 12, we prepared compound 12 via our optimized reaction conditions vide infra (but in the absence of any nucleophilic amine); then monitored its conversion to 11a via 1 H NMR (Figure 3).Remarkably, the reaction with BuNH 2 reached 50% completion after only 12 min at 23 � C, and no starting material 12 remained after 14 min heating at 100 � C. (Spectra shown in Figure 3 were obtained at 23 � C (C) and 100 � C (D), respectively).Preliminary kinetic analyses were performed on both a control reaction (Run 1: compound 9, plus 2.0 equiv.BuNH 2 10a, no fluoride, no BnEt 3 NCl) and various reactions designed to test the effect of our optimized reaction components (Run 2: compound 9, 2.0 equiv.BuNH 2 10a, 1.0 equiv.CsF, no BnEt 3 NCl) (Run 3: compound 9, 2.0 equiv.BuNH 2 10a, 1.0 equiv.CsF, 10 mol% BnEt 3 NCl), and (Run 4: compound 9, 4.0 equiv.BuNH 2 10a, 1.0 equiv.CsF, 10 mol% BnEt 3 NCl) (Table 3). 2 nd order rates for these reactions reveal an accelerating effect with increasing fluoride concentration.Notably, the reaction involving only compound 9 and BuNH 2 (Run 1: no fluoride) was 3-fold slower than the reaction containing our optimized reaction conditions (Run 3).A modest acceleration due to added BuNH 2 was observed in Run 4 (25% relative to Run 3) which is consistent with accelerations attributed to the contribution of k 4 [amine] in previously reported SNAr reactions thought to proceed via a twostep addition-elimination mechanism with aryl or heteroaryl fluorides. [28]A plausible mechanism that is consistent with our observations is illustrated in Scheme 1.Recent theoretical work suggests that the involvement of discreet Meisenheimer-type complexes such as I, depends strongly on the structure of the reacting heteroaryl substrate, and azaheteroaryl-chlorides undergo SNAr via a one-step kinetic process without forming addition-intermediates, whereas reactions of the corresponding azaheteroaryl-fluorides proceed via mechanistic pathways that can best be described as a "continuum between stepwise [addition-elimination] and concerted". [29]Key results that support our proposed mechanism were obtained from optimization experiments performed with model Table 1.C-6 amino-substituted products.a-e a 100 mg scale.b 1.0 g scale.c 2.0 equiv.amine.d Isolated yields.e Full characterization data including 1 H, 13 C NMR, HRMS, IR, and HPLC traces are provided in the Supporting Information.
substrate butylamine (10a, Table 2), and include the following: (1) treatment of 9 with 10a (2.0 equiv.) in the absence of fluoride source gave product 11a in only 44% yield (Table 2, entry 24); (2) yields for 11a increased to approx.50% when fluoride (CsF or KF, 1.0 equiv.) was included (Table 2, entries 4 and 5); (3) the yield for 11a nearly doubled to 94% in the presence of CsF (1.0 equiv.)and 10 mol% of the phase transfer catalyst BnNEt 3 Cl (Table 2, entry 21); (4) yields for 11a were higher with 0.1-0.5 equiv.CsF than in the complete absence of fluoride source (Table 2, entries 22-24), and (5) satisfactory yield (i.e., approx.75%) from the reaction utilizing the less organic-soluble [25] KF required 10-fold excess KF in the absence of BnNEt 3 Cl (Table 2, entry 3).The fact that yields were only slightly increased when using sub-stoichiometric quantities of CsF (0.5 and 0.1 equiv., respectively, Table 2, entries 22 and 23) is consistent with our kinetic experiments that revealed a 3-fold slower rate for reaction of 9 in the  absence of fluoride.Reversible formation of intermediate 12 from 9, is also consistent with the fact that conversion of 9 to 12 in the absence of nucleophilic amine (Figure 3A, (A)), was incomplete, even after heating for 24h at 100 � C in the presence of 1.0 equiv CsF and 10 mol% BnNEt 3 Cl (see Supporting Information).Required use of 10fold excess of the less organic-soluble [25] KF (Table 2, entry 3) or 1.0 equiv. of CsF with 10 mol% BnNEt 3 Cl (Table 2, entry 21) supports our conclusion that increased concentrations of the poorly soluble fluoride anion, leads to enhanced in situ formation of intermediate 12 which will react more rapidly with nucleophilic amine in the SNAr reaction, consistent with the well known SNAr reactivity order for aryl and heteroaryl halides: (F > Cl > Br > I). [24]Solid-liquid phase transfer catalysis would be expected to enhance effective concentrations of organic-soluble F -in the reaction mixture (M þ F -, Scheme 1), and the lower yield observed when 100 mol% BnNEt 3 Cl was utilized (Table 2, entry 7), is consistent with earlier reports demonstrating less efficient solid-liquid phase transfer catalysis when stoichiometric quantities of a phase transfer catalyst are used as opposed to using catalytic quantities (i.e., 10 mol%). [30]Hence, efficient phase transfer catalyzed conversion of 9 to the more reactive 12 reasonably explains the substantially higher yields we obtained utilizing only 1.0 equiv.CsF/10 mol% BnNEt 3 Cl and our optimized conditions.Products were obtained in 79%-98% yield, and the reaction worked well for a variety of both 1 � and 2 � alkylamines (Table 1).Unfortunately, arylamine substrates did not perform well under these conditions, as manifested by the low conversion to C-6 aminated products when aniline or 2-aminothiazole were used as  Scheme 1. Plausible mechanism consistent with optimization and preliminary kinetic experiments. [33]odel arylamine nucleophiles (approx.25%-30% conversion, as determined by TLC and mass spectrometry). [31]timization Development of our optimized conditions was carried out utilizing butylamine (10a) and is summarized in Table 2.In our initial experiments (entries 1--4), we established that excess KF is required for yields >46%.The yield for 11a was slightly higher when using 1.0 equiv.CsF compared to 1.0 equiv.KF (52%, entry 5; compared to 46%, entry 4; respectively), but was still lower than desirable.Evaluation of the effect of solvents, bases, and number of equivalents of amine revealed that (1) in the presence of 100 mol% BnNEt 3 Cl, approximately equivalent yields could be obtained with only 1.0 equiv. of CsF, regardless of the presence or absence of added base (Table 2, entries 7-11); and (2) DMSO was the solvent of choice for this reaction (entries 11-19).The nature of the base did not seem to have a profound effect (Table 2, entries 7-11), and we ultimately found that optimal yields could be obtained utilizing only 10 mol% BnNEt 3 Cl, 1.0 equiv.CsF, and 2.0 equiv. 1 � or 2 � alkylamine (Table 2, entries 21-23).Since utilization of only 1.0 equiv. of fluoride greatly reduces the toxic waste stream required for high-yield preparation of C-6 aminated products, we applied these optimized conditions to a broad range of nucleophilic alkylamines.As illustrated in Table 1, the substrate scope is quite broad, thus establishing this method as a low cost, environmentally friendly, method that should find wide use within the synthetic organic community.The C-3 bromo functionality could potentially be converted to a wide variety of targets utilizing conventional cross-coupling methodologies, thus providing high yield access to a range of useful C-6/C-3 bis-substituted targets.

Summary
In summary, we have developed an efficient, low cost, metal-free, and environmentally friendly method for C-6 amination of the 3-bromoimidazo[1,2-b]pyridazine heterocycle.
In addition to avoiding expensive air and moisture sensitive transition metal catalysts, the method also avoids drawbacks associated with previous methods for C-6 amination of imidazo[1,2-b]pyridazines, which drawbacks include high reaction temperatures (120 � C-130 � C), large excess of amines (3.0-5.0 equiv.), and large excess of toxic fluoride (10.0 equiv.).The method gives exceptionally high yields for simple unfunctionalized 1 � alkylamines amines (e.g.butylamine, isopropylamine, and cyclohexylamine), and 1 � alkylamines with methylene-or ethylene-linked aromatic or heteroaromatic functionality (e.g.benzene, thiophene, furan, indole, and pyridine) were also well tolerated.2 � Alkylamines also coupled well under these conditions (e.g.pyrrolidine, morpholine, and piperidine), and gram-scale reactions gave products in comparable yields for representative reactions (79%-92%, pyrrolidine, butylamine, and piperidine, respectively).Unfortunately, arylamines were not good substrates under these conditions, with only approximately 30% conversion observed for model arylamines such as aniline and 2aminothiazole.Importantly, yields for alkylamines were from 30% to 40% higher than those obtained via significantly more expensive, air and moisture sensitive palladium catalyzed Buchwald-Hartwig coupling, as was impressively illustrated for the synthesis of (PIM-1)-selective inhibitor SGI-1776, which was obtained in 91% yield with our method (Table 1), compared to 50% yield in the original patent literature (Figure 1). [19,32]Our method avoids expensive transition metal catalysts and can be performed using DMSO without careful exclusion of air or rigorous drying/deoxygenation of solvent and starting materials (in contrast to the very unforgiving inert-atmosphere conditions typically required for transition metal catalyzed aminations).Thus, our method is synthetically more tractable, environmentally friendly, higher yielding, and more cost effective, than previously reported methods.

Experimental
Experimental details for preparing 11a are provided below, along with 1 H and 13 C NMR characterization data.Full experimental details and characterization data for all compounds in Table 1 (11a-p, SGI-1776), and compound 12 (Figure 3), are provided as Supporting Information (see below).

Figure 3 .
Figure 3. (A) Preparation of 12 using our optimized conditions; followed by (B) SNAr conversion of isolated 12 to 11a (monitored by 1 H NMR): (C) 1 HNMR after 12 min at 23 � C; (D) 1 H NMR after 14 additional min at 100 � C. (Note: The chemical shifts for both the NH and upfield doublet for 11a are shifted in spectrum D as would be expected at 100 � C. Product(s) 11a obtained in this and all other kinetic reactions (runs 1-4) were identical in every respect to authentic material (confirmed by HRMS and 1 H NMR at 23 � C). a Remaining mass balance was unreacted 9.
scale.b 0.21 M in DMSO-d 6 .c 2nd order rates determined from linear regression of 1/[9] vs. time (as determined by D integration of peaks from both 9 and 11a, normalized to [9] at t ¼ 0).

Table 2 .
Optimization experiments.a