Synthesis of 5-Fluoroalkylated Pyrimidine Nucleosides via Negishi Cross-Coupling

Abstract

5-fluoroalkylated pyrimidine nucleosides (1) have potential as therapeutic agents and molecular imaging agents targeting HSV1-tk suicide gene therapy. Thus, straightforward preparation of 5-fluoroalkylated nucleoside derivatives has been developed. Reported herein are the first examples of Pd-catalyzed Negishi cross-coupling of 3-N-benzoyl-3′,5′-di-O-benzoyl-5-iodo-2′-deoxyuridine (2a) and 3-N-benzoyl-3′,5′-di-O-benzoyl-5-iodo-2′-deoxy-2′-fluoroarabinouridine (2b) with unactivated Csp³ fluoroalkylzinc bromides. This paper demonstrates the first synthesis of six 5-fluoroalkyl-2′-deoxy pyrimidine nucleoside derivatives with three to five methylene-chain lengths (5). Furthermore, this methodology has been extended to create a series of thirteen 5-alkyl substituted nucleosides, including the target nucleosides 5 and 5-silyloxypropyl and 5-cyanobutyl derivatives.

Introduction

The convenient access to functionalized nucleosides has been driven by the demonstrated pharmacological activity of nucleoside analogues against a broad spectrum of biological targets.ⁱ In particular, several 5-substituted pyrimidine nucleosides show potent and selective anti-herpes activity.ⁱⁱ The selectivity of pyrimidine nucleosides for herpes-infected cells is due to the specific phosphorylation of nucleoside substrates by the virus-encoded thymidine kinase, HSV1-TK. Parallel to this paradigm, the HSV1-tk gene expressing the HSV1-TK proteinⁱⁱⁱ has been exploited an important strategy for the suicide gene therapy of cancer where transduced cancer cells expressing the viral protein are made more sensitive to nucleoside therapy.^iv Our research interest is in the development of ¹⁸F-radiolabeled^v 5-fluoroalkylated pyrimidine nucleosides ([¹⁸F]1a–f) (Figure 1) that can be used as probes for the non-invasive in vivo molecular imaging of HSV1-TK in genetically transduced cells to correlate with the effectiveness of gene therapy and to monitor chemotherapy.^vi These radioactive probes have been designed to allow for: (i) rapid and efficient preparation of [¹⁸F]1a–f by convenient radiochemistry, (ii) assessment of the stability effect of the 2′-hydrogen vs. 2′-fluorine on the glycosidic bond in vivo, and (iii) selective phosphorylation by HSV1-TK and thus accumulation in HSV1-TK expressing cells. However, in order to validate the preparation of the ¹⁸F-labeled nucleosides, a practical synthesis of the nonradioactive molecules 1a–f first needs to be developed for testing the binding affinity to HSV1-TK. This research effort is also motivated by the potential antiviral and antitumor activity of these novel nucleosides.

FIGURE 1.

Structure of 5-[¹⁸F]Fluoroalkyl Nucleosides as Molecular Imaging Agents for HSV1-tk Gene Therapy

Several synthetic approaches allow access to 5-fluoroalkyl pyrimidine nucleosides. The classical approach calls for the preparation of 5-substituted uracils from corresponding 5-alkylbarbituric acids^vii followed by glycosylation of the uracils with reactive pentose derivatives.^viii This methodology suffers from typically low yields due to multiple steps, and extensive purification due to insufficient anomeric selectivity during the coupling step. A straightforward method to C-5 modified pyrimidine nucleosides is the palladium-catalyzed cross-coupling reactions of 5-iodouridine derivatives with terminal alkynes,^ix or by means of an organometallic reagent of mercury,^x stannane,^xi boron,^xic or zinc,^xii for example. During our studies on the synthesis of HSV1-TK molecular imaging probes with a pyrimidine core, 3-N-benzoyl-3′,5′-di-O-benzoyl-5-iodo-2′-deoxyuridine (2a) and 3-N-benzoyl-3′,5′-di-O-benzoyl-5-iodo-2′-deoxy-2′-fluoroarabinouridine (2b), were found to be the key synthetic intermediates in metal-catalyzed cross-coupling reactions. These intermediates were readily accessible from 5-iodo-2′-deoxyuridine,^xiii,xiv which is commercially available or from 5-iodo-2′-deoxy-2′-fluoroarabinouridine^xv obtained following a modified literature procedure.^xvi

Our initial approach to 5-fluoroalkyl nucleosides was to follow the palladium-catalyzed Sonogashira coupling route that has been reported by several groups.^ix,xvii One example is shown in Scheme 1 for the preparation of the tribenzoyl-protected 5-fluoropropyl-2′-deoxyuridine 5a. The propargylated derivative 3 was obtained in very good yield, and the critical fluorination step, using DAST, afforded the fluoroalkyne 4 in moderate yields.^xviii However, this synthetic route was plagued by the significant defluorination observed during hydrogenation of 4 using Pd/C as the catalyst, resulting in inseparable mixture of products 5a and 5h (70:30 ratio, 89 %). Our observation mirrored that of a report that mono-fluoro-substituted olefins, such as fluoroethene and 3-fluoropropene, tend to eliminate fluorine during the course of coupling reactions with palladium.^xix First hydrogenating 3, then fluorinating the saturated alcohol 6 with DAST, avoided the fluorine-elimination route, although the limitation to this approach was the very low yield of 5a (< 14 %). TBAF was employed as an alternative fluorination method. However, reaction of mesylate 7a with TBAF did not afford the fluorinated nucleoside 5a. The disappointing results of this strategy, in particular, that of the hydrogenation step to 5a and the fluorination step to 5a, obviated the need to look at other strategies that allowed for direct coupling of the intact fluoroalkyl moiety.^xx

SCHEME 1.

Synthesis of 5-(3-Fluoropropyl) Pyrimidine Nucleoside 5a via Sonogashira Coupling Approach

The palladium-catalyzed Negishi cross-coupling reactions of organohalides with organozinc reagents are another well-established and powerful method to forming carbon-carbon bonds.^xxi The ease in preparation,^xxii the high functional group tolerance, and reactivity of organozinc reagents^xxid increases their synthetic utility. Contingent upon the preparation of the unactivated and saturated fluoroalkylzinc bromides from commercially available fluoroalkyl bromides, we chose the Negishi method to prepare our target nonradioactive nucleosides 1a–f.

The first examples of Negishi coupling reactions employing 5-iodo-2′-deoxyuridine nucleosides described the Pd- or Ni- catalyzed Csp²–Csp² coupling of alkynyl-, indolyl- and thienylzinc reagents to 5-iodo-3′,5′-di-O-bis-trimethylsily-protected 2′-deoxyuridine in 7–50 % yields.^xii A trifluoroisopropenylzincate with a Csp² center was also successfully coupled to 5-iodo-3′,5′-di-O-benzoyl-protected 2′-deoxyuridine using Pd(PPh₃)₄ in 70 % yield.^xxiii Most recent reports indicate zincated nucleosides as an entry into 5-aryl-substituted uridines (Csp²–Csp² coupling).^xxiv However, to our knowledge, there exist no examples of Csp²–Csp³ cross-coupling reactions between 5-iodo-2′-deoxyuridine derivatives and unactivated and fully saturated alkylzincates. Reported herein are our efforts towards a convenient and efficient procedure for the preparation of 5-fluoroalkylated pyrimidine nucleosides 1a–f by Negishi cross-coupling.

Results and Discussion

Our latest efforts to realize the 1a–f by the Negishi cross-coupling approach focused on a rapid one-step preparation of tribenzoyl-protected 5-fluoroalkyl nucleosides 5a′–f from the strategic intermediates 2a,b (Scheme 2) and commercially available saturated fluoroalkyl bromides. Again, although compounds 5a′–f themselves would be inactive against HSV1-TK, they represent the key precursors towards the final target nucleosides 1a–f.

SCHEME 2.

Synthesis of 5-Fluoroalkylated Pyrimidine Nucleosides 5a–f Using Saturated Fluoroalkylzincates

At the onset of our investigation, efforts were made to prepare the critical zinc reagent for subsequent coupling. Following an important report by Huo on the mild and facile preparation of alkylzinc bromides from alkyl bromides and common zinc powder with catalytic amounts of iodine,^xxiia,xxv we observed high conversion of 4-fluorobutyl bromide to (4-fluorobutyl)zinc bromide in our model reaction.^xxvi With the zinc reagent in hand, experimentation began by monitoring the effect of catalyst on the Negishi cross-coupling reaction of 2a with (4-fluorobutyl)zinc bromide, using the coupling conditions as described by Huo.^xxiia Nine catalysts were selected for this analysis and the results of these experiments are summarized in Table 1. The Pd(PPh₃)₄ catalyst, which successfully coupled the activated isopropenylzinc reagent with the dibenzoyl-protected 5-iodo-2′-deoxyuridine,^xxiii as described previously, afforded trace amounts of the coupling product 5c (entry 1, Table 1).^xxvii Employing a variety of catalytic systems shown to be effective in Negishi coupling reactions, including Cl₂Ni(PPh₃)₂,^xxiia and Pd(dppf)Cl₂,^xxviii afforded the product in negligible yields (entries 2 and 3, Table 1). In addition, the Pd-N-heterocyclic carbene (NHC) catalyst, PEPPSI-IPr,^xxix did not yield any appreciable product (entry 4, Table 1).

TABLE 1.

Effect of catalyst on the Negishi coupling of (4-fluorobutyl)zinc bromide with 2a

Entry	Catalytic System	Yielda(%)
		5c	8	2a
1	Pd(PPh3)4 (5 mol %)	2	3	95
2	Cl2Ni(PPh3)2 (5 mol %)	0	23	77
3	Pd(dppf)Cl2 (5 mol %); CuI (6 mol %)	1	30	69
4	PEPPSI-IPr (5 mol %); LiBr (2 equiv.)	1	9	90
5	Pd(P(t-Bu3)2 (5 mol %)	47b	31	12
6	Pd(dba)2 (0.1 mol %); S-Phos (0.2 mol %)	5	10	85
7	Pd2dba3 (0.1 mol %); JohnPhos (0.2 mol %)	4	10	86
8	Pd2dba3 (0.1 mol %); PCy2-JohnPhos (0.2 mol %)	0	2	98
9	Pd2dba3 (0.1 mol %); (R,S)-JosiPhos (0.2 mol %)	0	4	96

^aHPLC yield;

^bIsolated yield: 43%.

Unlike the previous catalysts, experiments with Pd(P(t-Bu)₃)₂^xxx gave very encouraging results (entry 5, Table 1). We observed the coupling product 5c in 47 % yield. The major by-product isolated is the de-iodinated nucleoside 8, in which the iodide is replaced by hydrogen. The reason for this exchange has not been investigated. However, the major side product 8 would most likely arise from competitive β-hydride elimination of the nucleoside-Pd(II)-alkyl intermediate as opposed to the desired reductive elimination of the same intermediate Pd(II) species to yield the coupled product. Subsequent reductive elimination of the nucleoside-Pd-H intermediate would form the de-iodinated product 8. It may be suggested that acceleration of the reductive elimination step of the second intermediate Pd(II) species is critical to maximizing the yield of these reactions.

Bulky biphenylphosphine ligands are reported as improving Pd-catalyzed coupling yields by suppressing the tendency for destructive β-hydride elimination thus favoring reductive elimination.^xxxi Contrary to expectations, bulky Buchwald ligands such as S-Phos,^xxivb,xxxii cyclohexylJohnPhos (PCy₂-JohnPhos),^xxxii JohnPhos,^xxxiii and the Solvias ligand (R,S)-JosiPhos^xxxiv did not improve the coupling yields (entries 6–9, Table 1). Apparently, for most of the catalytic systems employed, oxidative addition was the rate-limiting step in the Negishi cross-coupling reactions as evidenced by the substantial recovery of starting material 2a.

These initial experiments established that in our system, Pd(P(t-Bu)₃)₂ is the most effective catalyst in achieving cross-coupling between 2a and (4-fluorobutyl)zinc bromide. Having obtained the coupling product we focused on further optimizing the reaction by increasing the consumption of the starting material 2a and by increasing the ratio between product 5c and de-iodinated nucleoside 8 (i.e. reductive elimination vs. β-hydride elimination).

Upon a more detailed investigation into the optimization of the coupling conditions several interesting observations were made (Table 2). A 1.6:1 ratio of Zn reagent/nucleoside resulted in a reasonably good yield within 1 h (entries 1–4, Table 2). Even more impressively, an increase in the stoichiometry increased the rate of reaction to afford the same product ratios in half the time (entries 5–12, Table 2) and lead to maximal consumption of the starting material. The reaction time is a critical factor due to the HPLC detection of degradation products within 30 min of initiation of the reaction, which typically increased in relative intensity with longer reaction times. The data also suggested that the stability of the nucleoside in the reaction mixture was dependent on the Zn reagent/nucleoside ratio since degradation products were observed to a greater degree and isolated yields are lower with larger excesses of Zn reagent (entries 8, 12, and 13, Table 2). Regardless of the stoichiometry of reagents used, no more than 53 % of 5c could be obtained in any of the coupling conditions investigated. Furthermore, the ratio of 5c to 8 remained more or less constant, and there typically remained trace amounts of starting material 2a. As such, based on isolated yields, entry 6 of Table 2 was chosen as the optimized Negishi cross-coupling condition.

TABLE 2.

Optimization of Negishi cross-coupling reaction using Pd(P(t-Bu)₃)₂ under various conditions

Entry	Equiv Zn Reagent	Time	Yielda (%)
			5cb	8	2a
1	1.6	15 min	40	25	35
2	1.6	30 min	48	31	30
3	1.6	45 min	47	32	13
4	1.6	1 h	47 (43)	31	12
5	3	15 min	43	21	36
6	3	30 min	56 (53)	40	4
7	3	45 min	55	32	7
8	3	1 h	57 (43)	36	6
9	5	15 min	34	23	38
10	5	30 min	52 (48)	33	5
11	5	45 min	56	36	7
12	5	1 h	50 (44)	32	5
13	8	1 h	65 (26)	28	4

^aHPLC yield;

^bValues in parentheses represent isolated yield.

Preparation of a Series of 5-Alkylated Nucleoside Derivatives

To demonstrate the utility and versatility of the optimized Negishi cross-coupling methodology, a series of 5-alkylated nucleosides derivatives, including our desired 5-fluoroalkyl nucleosides 5a′–f, was prepared using a variety of alkylzinc reagents (Table 3). A variety of zinc reagents could be successfully coupled to the nucleosides 2a or 2b in appreciable yields. As expected, (4-fluorobutyl)zinc and (5-fluoropentyl)zinc bromides were prepared in high yields (> 0.74 M) following the Huo method,^xxiia and subsequent coupling to 2a or 2b afforded the coupling products 5c–5f in moderate yields (entries 4–7, Table 3). Interestingly, attempts at preparing the short-chained (3-fluoropropyl)zinc and (2-fluoroethyl)zinc bromides were unsuccessful (entries 1 and 8, Table 3). Using the more reactive 3-iodo-1-fluoropropane, we obtained the corresponding alkylzinc iodide reagent, albeit in lower yields as compared to the longer chain fluoroalkylzinc bromides.^xxxv Cross-coupling of the zinc reagent with 2a,b afforded the coupled products 5a′,b after 24 h in very low yield (< 8 %) (entry 2 and 3, Table 3). In an analogous fashion, short chain alkanes were coupled to 2a through an alkylzinc iodide intermediate (entries 9–11, Table 3). The nucleosides 5g–i were generated in low yields (15–23 %) but it is important to note that they were obtained in yields higher than those observed with the short chain fluoroalkyl zincates. This result suggests that the fluorine moiety confers instability to the desired zinc intermediate thus leading to low (or no) yields of the zinc reagent and very low isolated yields of the coupling product.

TABLE 3.

Preparation of Library of 5-Alkylated Nucleosides by Pd(P(t-Bu)₃)₂-Catalyzed Negishi Cross-Coupling of Nucleoside 2a or 2b with Zincated Alkyl Halides (R′Zn-Halide)^a

Entry	Nucleoside Substrate	Zincatea (R′ Zn-Halide)	Nucleoside Product	Product Yieldb (%)
1	2a		5a′ (X = H; R′ = (CH2)3F)	Not obtainedc
2	2a		5a′ (X = H; R′ = (CH2)3F)	8
3	2b		5b (X = F; R′ = (CH2)3F)	5
4	2a		5c (X = H; R′ = (CH2)4F)	53
5	2b		5d (X = F; R′ = (CH2)4F)	43
6	2a		5e (X = H; R′ = (CH2)5F)	50
7	2b		5f (X = F; R′ = (CH2)5F)	53
8	2a		(X = H; R′ = (CH2)2F)	Not obtainedc
9	2a		5g (X = H; R′ = Bu)	15
10	2a		5h′ (X = H; R′ = Pr)	17
11	2a		5i (X = H; R′ = Et)	23
12	2a		5g′ (X = H; R′ = Bu)	44
13	2a		5i′ (X = H; R′ = Et)	23
14	2a		5j (X = H; R′= (CH2)5OCOCH3)	33
15	2a		5k (X = H; R′ = (CH2)3COOEt)	39
16	2a		5l (X = H; R′ = (CH2)3CN)	29
17	2a		5m (X = H; R′ = (CH2)3OTBS)	30

^aZinc reagents were prepared following Huo method;^22a Iodoalkylzinc reagents employed for entries 2, 3, 9–11; Commercial dialkylzinc reagents (3 equiv) employed for entries 12 and 13.

^bIsolated yields based on nucleoside substrate 2a/2b.

^cBromoalkylzinc reagents could not be prepared following the Huo method,^22a thus coupling to 2a or 2b was not performed.

The efficiency of Negishi cross-coupling reactions employing dialkylzinc derivatives^xxxvi led us to explore the effect of using the dialkylzinc in place of the alkylzinc halides for cross-coupling reactions with 2a. When using the commercially available (n-Bu)₂Zn, the nucleoside derivative 5g′ was easily prepared, and in much greater yield (44 %) as compared to utilizing the monoalkylzinc reagent (15 %) (entry 12, Table 3). There was no difference in the isolated yield of 5i′ (23 %) when Et₂Zn vs. EtZnI was used (entry 13, Table 3). Of note, however, is that either zinc reagent successfully generated nucleoside 5i/5i′, which upon deprotection of the benzoyl groups would afford the potent antiherpes agent 5-ethyl-2′-deoxyuridine (EDU, Edoxuridine, Aedurid®).^ii,xxxvii Employment of the Negishi cross-coupling approach using dialkylzinc reagents^xxxviii thus represents an alternative for synthesizing 5-alkylated nucleosides.

We extended our methodology to include other readily accessible functionalized alkylzinc reagents (entries 14–17, Table 3). Of particular interest was the successful coupling reaction employing (3-tert-butyldimethylsiloxypropyl)zinc bromide to afford 5m (entry 15, Table 3). As elaborated in Scheme 3, silylated coupling products, via the OTBS moiety, offer an alternative route to rapidly obtaining mesylate precursors for subsequent ¹⁸F-radiolabeling to achieve our target HSV1-TK molecular imaging probes [¹⁸F]1a–f.^xxxix We have since prepared the series of six mesylate precursors 7a–f that upon labeling with [¹⁸F]KF and Kryptofix (K[2,2,2]), followed by deprotection, have successfully afforded no-carrier added [¹⁸F]1a–f in good yields and in short preparation times. Details of the radiolabeling studies and biological characterizations of [¹⁸F]1a–f will be published elsewhere.

SCHEME 3.

Synthesis of 5-[¹⁸F]Fluoroalkylated Pyrimidine Nucleosides [¹⁸F]1a–f

Deprotection

Access to the bioactive nucleosides targeting HSV1-TK was exemplified using the above-prepared nucleoside derivatives 5c and 5d. The 3-N-benzoyl and 3′,5′-di-O-benzoyl-protecting groups were hydrolyzed by treatment of 5c,d with a standard solution of sodium methoxide in anhydrous methanol for 15 min under reflux. The corresponding unprotected nucleosides 5-fluorobutyl-2′-deoxyuridine (1c) and 5-fluorobutyl-2′-deoxy-2′-fluoroarabinouridine (1d) were obtained in very high yields (> 95 %) demonstrating that benzoyl removal was compatible with the presence of the 5-fluoroalkyl moieties. This procedure allowed for the rapid and efficient two-step preparation of the bioactive compounds 1c from 2a and 1d from 2b in 52 % and 41 % overall yields.

Conclusions

Our studies demonstrate the first Pd(P(t-Bu)₃)₂-catalyzed cross-coupling reactions between fully protected 5-iodo-2′-deoxyuridine nucleosides and unactivated alkylzinc reagents. Furthermore, this methodology represents a significant improvement over conventional syntheses to 5-alkyl substituted pyrimidine nucleosides that typically involve multiple steps affording products in lower overall chemical yields.^vii,viii As such, our approach provides easier access to biologically important 5-alkyl substituted pyrimidine nucleosides, including our new series of 5-fluoroalkylated pyrimidine nucleosides. We are currently expanding the scope of this reaction to include more elaborate zinc reagents.

Experimental Section

Typical Procedure for Cross-Coupling of 2a or 2b with Organozinc Species

In a flame-dried 5-mL two-neck flask equipped with a stir bar was added 2a (0.08 mmol) or 2b (0.08 mmol), Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) and anhyd DMA (1 mL) under argon. To the reaction mixture, alkylzinc reagent^xxiia (0.6–0.9M solution in DMA, 0.23 mmol) or dialkylzinc reagent (230 μL of 1 M in heptane, 0.23 mmol) was added dropwise and the reaction stirred at room temperature for 30 min. The reaction was monitored by HPLC (Method A) and judged to be complete. The reaction mixture was quenched with saturated NH₄Cl (5 mL) and extracted with EtOAc (3 × 5 mL). The combined organic extracts were washed with brine, dried over MgSO₄, filtered and concentrated. The crude product mixture was purified by silica gel preparative thin layer chromatography (PTLC) using of 20 % to 30 % EtOAc in hexanes.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(3-fluoropropyl)-2′-deoxyuridine (5a′)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (3-fluoropropyl)zinc iodide (375 μL of 0.6 M solution in DMA, 0.23 mmol) afforded 4 mg of 5a′, 8 % yield. ¹H NMR (200 MHz, CDCl₃) δ 8.10–7.90 (m, 6H), 7.70–7.42 (m, 9H+1H), 6.45 (dd, 1H, J₁ = 5.5 Hz and J₂ = 8.4 Hz), 5.70–5.66 (m, 1H), 4.79 (dd, 1H, J₁ = 12.3 Hz and J₂ = 25.1 Hz), 4.77 (dd, 1H, J₁ = 12.2 and J₂ = 26.0 Hz), 4.58 (m, 1H), 4.33 (dt, 2H, J₁ = 5.7 Hz and J₂ = 47.3 Hz), 2.78 (dd, 1H, J₁ = 5.2 Hz and J₂ = 14.1 Hz), 2.47–2.32 (m, 1H,), 2.18–2.09 (m, 2H), 1.69–1.39 (m, 2H). ¹³C NMR (50 MHz, CDCl₃) δ 168.8, 166.2, 162.3, 149.4, 135.3, 135.0, 135.96, 131.7, 130.6, 130.0, 129.7, 129.5, 129.4, 129.2, 129.1, 128.8, 114.9, 85.6, 83.3 (d, J = 164.7 Hz), 83.2, 75.1, 64.5, 38.4, 29.2 (d, J = 19.9 Hz), 23.8 (d, J = 5.6 Hz). HRMS calcd for C₃₃H₂₉FN₂O₈Na ([M+Na]⁺) 623.1806, found 623.1830. HPLC A: t_R = 10.5 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(3-fluoropropyl)-2′-deoxy-2′-fluoroarabinouridine (5b)

2b(51 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (3-fluoropropyl)zinc iodide (375 μL of 0.6 M solution in DMA, 0.23 mmol) afforded 3 mg of 5b, 5 % yield. ¹H NMR (200 MHz, CD₃CN) δ 8.12–7.94 (m, 6H), 7.79–7.43 (m, 9H+1H), 6.31 (dd, 1H, J₁ = 3.2 Hz and J₂ = 20.7 Hz), 5.69 (ddd, 1H, J₁ = 1.0 Hz, J₂ = 3.6 Hz and J₃ = 19.0 Hz), 5.44 (ddd, 1H, J₁ = 1.0 Hz, J₂ = 3.3 Hz and J₃ = 50.2 Hz), 4.79 (dd, 1H, J₁ = 12.1 Hz and J₂ = 24.6 Hz), 4.77 (dd, 1H, J₁ = 12.1 Hz and J₂ = 25.8 Hz), 4.65–4.59 (m, 1H), 4.37 (dt, 2H, J₁ = 5.9 Hz and J₂ = 47.4 Hz), 2.34–2.27 (m, 2H), 1.76–1.70 (m, 2H). ¹³C NMR (50 MHz, CD₃CN) δ 170.4, 167.1, 166.4, 163.5, 150.1, 138.4, 138.3, 136.6, 135.0, 134.6, 132.6, 131.4, 130.9, 130.5, 130.1, 129.8, 114.2, 94.4 (d, J = 189.8 Hz), 85.7 (d, J = 16.5 Hz), 84.4 (d, J = 161.6 Hz), 81.5, 77.9 (d, J = 30.3 Hz), 64.3, 30.1 (d, J = 19.5 Hz), 23.8 (d, J = 5.7 Hz). HRMS calcd for C₃₃H₂₉F₂N₂O₈ ([M+H]⁺) 619.1892, found 619.1896. HRMS calcd for C₃₃H₂₈F₂N₂O₈ Na ([M+Na]⁺) 641.1711, found 641.1710. HPLC A: t_R = 12.1 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(4-fluorobutyl)-2′-deoxyuridine (5c)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (4-fluorobutyl)zinc bromide (250 μL of 0.9 M solution in DMA, 0.23 mmol) afforded cross-coupling product 5c as a white solid (24 mg, 53 % yield). ¹H NMR (200 MHz, CDCl₃) δ 8.10–7.89 (m, 6H), 7.70–7.39 (m, 9H+1H), 6.46 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.8 Hz), 5.68 (dt, 1H, J₁ = 1.5 Hz and J₂ = 6.5 Hz), 4.78 (dd, 1H, J₁ = 12.2 Hz and J₂ = 28.8 Hz), 4.77 (dd, 1H, J₁ = 12.2 Hz and J₂ = 29.5 Hz), 4.59–4.55 (m, 1H), 4.35 (dt, 2H, J₁ = 5.7 Hz and J₂ = 47.3 Hz), 2.78 (ddd, 1H, J₁ = 1.3 Hz and J₂ = 5.4 Hz and J₃ = 14.2 Hz), 2.47–2.32 (m, 1H), 2.22–2.12 (m, 2H), 1.69–1.39 (m, 4H). ¹³C NMR (50 MHz, CDCl₃) δ 168.9, 166.2, 162.4, 149.4, 135.3, 134.6, 134.0, 131.8, 130.6, 130.0, 129.7, 129.5, 129.4, 129.2, 129.1, 128.8, 128.3, 115.7, 85.6, 83.8 (d, J = 163.9 Hz), 83.1, 75.2, 64.5, 38.3, 30.1 (d, J = 19.7 Hz), 26.9, 24.6 (d, J = 5.1 Hz). HRMS calcd for C₃₄H₃₂FN₂O₈ ([M+H]⁺) 615.2143, found 615.2148. HPLC A: t_R = 11.6 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(4-fluorobutyl)-2′-deoxy-2′-fluoroarabinouridine (5d)

2b (51 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (4-fluorobutyl)zinc bromide (281 μL of 0.8 M solution in DMA, 0.23 mmol) afforded 20 mg of 5d, 43 % yield. ¹H NMR (200 MHz, CD₃CN) δ 8.13–7.94 (m, 6H), 7.79–7.49 (m, 9H+1H), 6.32 (dd, 1H, J₁ = 3.1 Hz and J₂ = 20.8 Hz), 5.70 (dd, 1H, J₁ = 3.2 Hz and J₂ = 18.8 Hz), 5.45 (dd, 1H, J₁ = 2.4 Hz and J₂ = 50.2 Hz), 4.80 (dd, 1H, J₁ = 12.1 Hz and J₂ = 29.2 Hz), 4.78 (dd, 1H, J₁ = 12.1 Hz and J₂ = 30.6 Hz), 4.65–4.62 (m, 1H), 4.36 (dt, 2H, J₁ = 6.0 Hz and J₂ = 47.4 Hz), 2.25–2.18 (m, 2H), 1.64–1.46 (m, 4H). ¹³C NMR (50 MHz, CD₃CN) δ 170.4, 166.4, 163.5, 150.1, 138.2, 138.1, 136.5, 134.9, 134.6, 134.2, 132.6, 131.4, 130.8, 130.5, 130.1, 129.8, 129.6, 114.7, 94.4 (d, J = 189.7 Hz), 85.6 (d, J = 16.5 Hz), 84.9 (d, J = 160.9 Hz), 81.4, 77.8 (d, J = 30.3 Hz), 64.2, 30.5 (d, J = 19.5 Hz), 27.2, 25.1 (d, J = 5.4 Hz). HRMS calcd for C₃₄H₃₁F₂N₂O₈ ([M+H]⁺) 633.2048, found 633.2062. HPLC A: t_R = 13.2 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(5-fluoropentyl)-2′-deoxyuridine (5e)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (5-fluoropentyl)zinc bromide (250 μL of 0.9 M solution in DMA, 0.23 mmol) afforded 24 mg of 5e, 50 % yield. ¹H NMR (200 MHz, CDCl₃) δ 8.10–7.90 (m, 6H), 7.66–7.43 (m, 9H), 7.36 (s, 1H), 6.47 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.8 Hz), 5.70–5.67 (m, 1H), 4.78 (dd, 1H, J₁ = 12.2 Hz and J₂ = 28.7 Hz), 4.76 (dd, 1H, J₁ = 12.2 Hz and J₂ = 29.5 Hz), 4.5 (m, 1H), 4.34 (dt, 2H, J₁ = 5.7 Hz and J₂ = 47.2 Hz), 2.78 (dd, 1H, J₁ = 5.3 Hz and J₂ = 14.2 Hz), 2.40 (ddd, 1H, J₁ = 6.3 Hz and J₂ = 8.5 Hz and J₃ = 14.5 Hz), 2.18–1.97 (m, 2H), 1.72–1.27 (m, 6H). ¹³C NMR (50 MHz, CDCl₃) δ 168.9, 166.2, 162.4, 149.4, 135.3, 134.4, 134.0, 131.8, 130.6, 130.0, 129.7, 129.5, 129.4, 129.2, 129.1, 128.8, 116.0, 85.5, 84.0 (d, J = 163.5 Hz), 83.1, 75.2, 64.6, 38.3, 30.2 (d, J = 19.5 Hz), 28.3, 27.2, 25.0 (d, J = 5.1 Hz). HRMS calcd for C₃₅H₃₄FN₂O₈ ([M+H]⁺) 629.2299, found 629.2298. HPLC A: t_R = 13.7 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(5-fluoropentyl)-2′-deoxy-2′-fluoroarabinouridine (5f)

2b (51 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (5-fluoropentyl)zinc bromide (268 μL of 0.84 M solution in DMA, 0.23 mmol) afforded 26 mg of 5f, 53 % yield. ¹H NMR (200 MHz, CD₃CN) δ 8.13–7.95 (m, 6H), 7.75–7.50 (m, 9H+1H,), 6.33 (dd, 1H, J₁ = 3.0 Hz and J₂ = 20.8 Hz), 5.70 (dd, 1H, J₁ = 3.0 Hz and J₂ = 18.9 Hz), 5.46 (dd, 1H, J₁ = 2.6 Hz and J₂ = 50.2 Hz), 4.80 (dd, 1H, J₁ = 12.1 Hz and J₂ = 31.0 Hz), 4.78 (dd, 1H, J₁ = 12.1 Hz and J₂ = 32.5 Hz), 4.65–4.61 (m, 1H), 4.38 (dt, 2H, J₁ = 6.0 Hz and J₂ = 47.5 Hz), 2.21–2.14 (m, 2H), 1.71–1.28 (m, 6H). ¹³C NMR (50 MHz, CD₃CN) δ 170.4, 166.4, 163.5, 150.1, 138.0, 137.9, 136.5, 134.9, 134.6, 134.2, 132.6, 131.3, 130.8, 130.6, 130.5, 130.1, 129.83, 129.81, 129.6, 115.0, 94.4 (d, J = 189.8 Hz), 85.6 (d, J = 16.3 Hz), 85.0 (d, J = 160.8 Hz), 81.5, 77.8 (d, J = 30.3 Hz), 64.2, 30.8 (d, J = 19.3 Hz), 28.9, 27.4, 25.4 (d, J = 5.6 Hz). HRMS calcd for C₃₅H₃₃F₂N₂O₈ ([M+H]⁺) 647.2205, found 647.2202. HPLC A: t_R = 15.2 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-butyl-2′-deoxyuridine (5g/5g′)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with butylzinc iodide (281 μL of 0.8 M solution in DMA, 0.23 mmol) afforded 7 mg of 5g, 15 % yield. Employing (n-Bu)₂Zn (230 μL of 1 M in heptane, 0.23 mmol) afforded 20 mg of 5i′, 44 % yield. ¹H NMR of 5g (200 MHz, CDCl₃) δ 8.10–7.90 (m, 6H), 7.69–7.43 (m, 9H) 7.34 (s, 1H), 6.47 (dd, 1H, J₁ = 5.5 Hz and J₂ = 8.7 Hz), 5.70–5.67 (m, 1H), 4.78 (dd, 1H, J₁ = 12.2 Hz and J₂ = 28.5 Hz), 4.76 (dd, 1H, J₁ = 12.2 Hz and J₂ = 29.3 Hz), 4.57–4.56 (m, 1H), 2.77 (dd, 1H, J₁ = 5.0 Hz and J₂ = 13.9 Hz), 2.47–2.33 (m, 1H), 2.10–2.05 (m, 2H), 1.29–1.12 (m, 4H), 0.81 (t, 3H, J = 7.0 Hz). ¹³C NMR of 5g (50 MHz, CDCl₃) δ169.0, 166.2, 162.4, 149.5, 135.2, 134.1, 134.0, 131.9, 130.6, 130.0, 129.8, 129.5, 129.4, 129.2, 129.0, 128.8, 116.5, 85.5, 83.1, 75.2, 64.6, 38.3, 30.8, 27.0, 22.5, 13.8. HRMS calcd for C₃₄H₃₂N₂O₈Na ([M+Na]⁺) 619.2056, found 619.2048. HPLC A: t_R = 16.1 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-propyl-2′-deoxyuridine (5h′)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with propylzinc iodide (281 μL of 0.8 M solution in DMA, 0.23 mmol) afforded 8 mg of 5h′, 17 % yield. ¹H NMR (200 MHz, CDCl₃) δ 8.10–7.90 (m, 6H), 7.69–7.43 (m, 9H) 7.32 (s, 1H), 6.47 (dd, 1H, J₁ = 5.5 Hz and J₂ = 8.7 Hz), 5.71–5.68 (m, 1H), 4.78 (dd, 1H, J₁ = 12.3 Hz and J₂ = 32.6 Hz), 4.76 (dd, 1H, J₁ = 12.3 Hz and J₂ = 33.4 Hz), 4.57 (m, 1H), 2.77 (dd, 1H, J₁ = 4.7 Hz and J₂ = 14.1 Hz), 2.47–2.32 (m, 1H), 2.11–2.01 (m, 2H), 1.42–1.31 (m, 2H), 0.77 (t, 3H, J = 7.2 Hz). ¹³C NMR (50 MHz, CDCl₃) δ 169.0, 166.2, 162.4, 149.5, 135.2, 134.2, 134.0, 131.9, 130.6, 130.0, 129.7, 129.5, 129.4, 129.2, 129.0, 128.8, 116.2, 85.4, 83.0, 75.2, 64.5, 38.3, 29.2, 21.8, 13.8. HRMS calcd for C₃₃H₃₀N₂O₈Na ([M+Na]⁺) 605.1900, found 605.1944. HPLC A: t_R = 13.5 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-ethyl-2′-deoxyuridine (5i/5i′)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with ethylzinc iodide (312 μL of 0.72 M solution in DMA, 0.23 mmol) afforded 10 mg of 5i, 23 % yield. Employing Et₂Zn (230 μL of 1 M in heptane, 0.23 mmol) afforded 10 mg of 5i′, 23 % yield. ¹H NMR of 5i (200 MHz, CDCl₃) δ 8.09–7.91 (m, 6H), 7.69–7.43 (m, 9H), 7.33 (s, 1H), 6.47 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.6 Hz), 5.70–5.67 (m, 1H), 4.79 (dd, 1H, J₁ = 12.2 Hz and J₂ = 29.2 Hz), 4.77 (dd, 1H, J₁ = 12.2 Hz and J₂ = 29.9 Hz), 4.57–4.56 (m, 1H), 2.77 (dd, 1H, J₁ = 5.1 Hz and J₂ = 13.9 Hz), 2.40 (ddd, 1H, J₁ = 6.7 Hz and J₂ = 8.3 Hz and J₃ = 14.5 Hz), 2.15 (q, 2H, J = 7.3 Hz), 0.93 (t, 3H, J = 7.4 Hz). ¹³C NMR of 5i (50 MHz, CDCl₃) δ 169.0, 166.21, 166.16, 162.4, 149.5, 135.2, 134.0, 133.7, 131.9, 130.7, 130.4, 130.0, 129.8, 129.6, 129.4, 129.2, 129.0, 128.8, 128.7, 117.7, 85.5, 83.1, 75.2, 64.5, 38.3, 20.6, 13.0. HRMS calcd for C₃₂H₂₈N₂O₈Na ([M+Na]⁺) 591.1743, found 591.1739. HPLC A: t_R = 10.7 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(5-acetoxypentyl)-2′-deoxyuridine (5j)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (5-acetoxypentyl)zinc bromide (281 μL of 0.8 M solution in DMA, 0.23 mmol) afforded 17 mg of 5j, 33 % yield. ¹H NMR (200 MHz, CDCl₃) δ8.41–7.90 (m, 6H), 7.69–7.43 (m, 9H), 7.36 (s, 1H), 6.46 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.6 Hz), 5.70–5.67 (m, 1H), 4.78 (dd, 1H, J₁ = 12.2 Hz and J₂ = 30.5 Hz), 4.76 (dd, 1H, J₁ = 12.2 Hz and J₂ = 31.4 Hz), 4.58–4.57 (m, 1H), 4.00 (t, 2H, J = 6.6 Hz), 2.78 (dd, 1H, J₁ = 5.2 Hz and J₂ = 14.1 Hz), 2.48–2.33 (m, 1H), 2.18–2.05 (m, 2H), 2.05 (s, 3H), 1.59–1.21 (m, 6H). ¹³C NMR (50 MHz, CDCl₃) δ 171.3, 169.0, 166.2, 162.4, 149.4, 135.3, 134.4, 134.0, 131.8, 130.6, 130.0, 129.8, 129.6, 129.4, 129.2, 129.1, 128.8, 116.1, 85.6, 83.1, 75.2, 64.5, 38.3, 28.5, 27.2, 25.8, 21.2. HRMS calcd for C₃₇H₃₆N₂O₁₀Na ([M+Na]⁺) 691.2248, found 691.2268. HPLC A: t_R = 11.8 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(4-ethoxy-4-oxobutyl)-2′-deoxyuridine (5k)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (4-ethoxy-4-oxobutyl)zinc bromide (300 μL of 0.75 M solution in DMA, 0.23 mmol) afforded 19 mg of 5k, 39 % yield. ¹H NMR (200 MHz, CDCl₃) δ 8.09–7.91 (m, 6H), 7.65–7.41 (m, 9H+1H), 6.44 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.6 Hz), 5.69–5.66 (m, 1H), 4.78 (dd, 1H, J₁ = 12.2 Hz and J₂ = 22.4 Hz), 4.76 (dd, 1H, J₁ = 12.2 Hz and J₂ = 23.2 Hz), 4.57–4.56 (m, 1H), 4.1 (q, 2H, J = 7.1 Hz), 2.77 (dd, 1H, J₁ = 5.1 Hz and J₂ = 13.8 Hz), 2.42 (ddd, 1H, J₁ = 6.5 Hz and J₂ = 8.3 Hz and J₃ = 14.5 Hz), 2.21–2.14 (m, 2H), 2.17 (t, 2H, J = 7.2 Hz), 1.77–1.67 (m, 2H), 1.25 (t, 3H, J = 7.1 Hz). ¹³C NMR (50 MHz, CDCl₃) δ173.2, 168.9, 166.2, 162.3, 149.4, 135.2, 135.0, 133.9, 131.8, 130.7, 130.0, 129.8, 129.6, 129.4, 129.2, 129.0, 128.8, 128.3, 115.2, 85.7, 83.1, 75.2, 64.5, 38.3, 33.9, 26.8, 23.9, 14.5. HRMS calcd for C₃₆H₃₄N₂O₁₀Na ([M+Na]⁺) 677.2111, found 677.2093. HRMS calcd for C₃₆H₃₅N₂O₁₀ ([M+H]⁺) 655.2292, found 655.2272. HPLC A: t_R = 11.4 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(3-cyanopropyl)-2′-deoxyuridine (5l)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (3-cyanopropyl)zinc bromide (256 μL of 0.88 M solution in DMA, 0.23 mmol) afforded 13 mg of 5l, 29 % yield. ¹H NMR (200 MHz, CDCl₃) δ8.11–7.91 (m, 6H), 7.67–7.43 (m, 9H +1H), 6.43 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.5 Hz), 5.69–5.66 (m, 1H), 4.80 (dd, 1H, J₁ = 12.2 Hz and J₂ = 27.3 Hz), 4.79 (dd, 1H, J₁ = 12.2 Hz and J₂ = 28.6 Hz), 4.59 (m, 1H), 2.81 (dd, 1H, J₁ = 5.3 Hz and J₂ = 14.2 Hz), 2.47–2.36 (m, 1H), 2.33–2.08 (m, 4H), 1.79–1.69 (m, 2H). ¹³C NMR (50 MHz, CDCl₃) δ 168.7, 166.2, 162.3, 149.3, 135.8, 135.4, 134.1, 134.0, 131.7, 130.7, 130.0, 129.8, 129.6, 129.5, 129.2, 128.9, 119.4, 113.6, 85.9, 83.4, 75.2, 53.6, 38.5, 26.9, 24.2, 16.9. HRMS calcd for C₃₄H₂₉N₃O₈Na ([M+Na]⁺) 630.1852, found 630.1874. HPLC A: t_R = 8.5 min.

3-N-benzoyl-3′,5′-di-O-benzoyl-5-(3-tert-butyldimethylsilyloxypropyl)-2′-deoxyuridine (5m)

2a (50 mg, 0.08 mmol) and Pd(P(t-Bu)₃)₂ (1.9 mg, 0.004 mmol) with (3-tert-butyldimethylsilyloxypropyl)zinc bromide (321 μL of 0.7 M solution in DMA, 0.23 mmol) afforded 25 mg of 5m, 47 % yield. ¹H NMR (200 MHz, CD₃CN) δ 8.09–7.90 (m, 6H), 7.69–7.42 (m, 9H), 7.40 (s, 1H), 6.43 (dd, 1H, J₁ = 5.4 Hz and J₂ = 8.7 Hz), 5.67–5.64 (m, 1H), 4.77 (dd, 1H, J₁ = 12.2 Hz and J₂ = 14.4 Hz), 4.75 (dd, 1H, J₁ = 12.2 Hz and J₂ = 14.6 Hz), 4.59–4.55 (m, 1H), 3.51 (t, 2H, J = 6.2 Hz), 2.77 (ddd, 1H, J₁ = 1.4 Hz and J₂ = 5.4 Hz and J₃ = 14.2 Hz), 2.40 (ddd, 1H, J₁ = 6.6 Hz and J₂ = 8.6 Hz and J₃ = 14.4 Hz), 2.30–2.19 (m, 2H), 1.66–1.52 (m, 2H), 0.88 (s, 9H), 0.03 (s, 6H). ¹³C NMR (50 MHz, CDCl₃) δ 169.0, 166.2, 162.4, 149.4, 135.2, 134.4, 134.0, 133.9, 131.9, 130.7, 130.0, 129.8, 129.6, 129.4, 129.2, 129.0, 128.8, 115.9, 85.8, 83.1, 75.2, 64.5, 62.6, 38.3, 31.5, 26.2, 24.1, 18.5, -5.1. HRMS calcd for C₃₉H₄₄N₂O₉SiNa ([M+Na]⁺) 735.2174, found 735.2688. HRMS calcd for C₃₉H₄₅N₂O₉Si ([M+H]⁺) 713.2894, found 713.2881. HPLC A: t_R = 15.1 min.

Typical Procedure for Deprotection of Tribenzoyl-Protected Nucleosides

In vial equipped with a stir bar was added tribenzoylprotected nucleosides. Dry 0.5 N NaOMe in MeOH solution (3 mL) was added dropwise. Reaction vial was sealed and then stirred at 80°C for 15 min. Reaction judged to be complete by HPLC (MeOH/H₂O (30/70)). After cooling to room temperature, reaction was quenched with 1N HCl (1.5 mL) and the solvent removed in vacuo. The crude product residue was purified by silica gel PTLC using 10 % MeOH in dichloromethane.

5-(4-fluorobutyl)-2′-deoxyuridine (1c)

The deprotection procedure described above, using 5c (23.4 mg, 0.04 mmol) afforded 11.5 mg of 1c, 99 % yield. ¹H NMR (200 MHz, CD₃OD) δ8.06 (s, 1H), 6.49 (t, 1H, J₁ = 6.7 Hz), 4.64 (dt, 2H, J₁ = 5.6 Hz and J₂ = 47.6 Hz), 4.64–4.57 (m, 1H), 4.12 (t, 1H, J₁ = 3.1 Hz), 3.98 (dd, 1H, J₁ = 12.1 Hz and J₂ = 16.8 Hz), 3.96 (dd, 1H, J₁ = 12.1 Hz and J₂ = 17.3 Hz), 2.59–2.37 (m, 2H+2H), 2.00–1.81 (m, 4H,). ¹³C NMR (50 MHz, CD₃OD) δ 166.1, 152.4, 138.6, 115.6, 89.1, 86.6, 84.8 (d, J = 162.8 Hz), 72.4, 62.9, 41.5, 31.2 (d, J = 19.6 Hz), 27.5, 25.6 (d, J = 5.1 Hz). HRMS calcd for C₁₃H₂₀FN₂O₅ ([M+H]⁺) 303.1356, found 303.1349. HPLC B: t_R = 7.6 min.

5-(4-fluorobutyl)-2′-deoxy-2′-fluoroarabinouridine (1d)

The deprotection procedure above, using 5d (12.7 mg, 0.021 mmol) afforded 6.1 mg of 1d, 95 % yield. ¹H NMR (200 MHz, CD₃CN with D₂O) δ 7.53 (s, 1H), 6.13 (dd, 1H, J₁ = 4.1 Hz and J₂ = 16.4 Hz), 5.02 (dt, 2H, J₁ = 3.9 Hz and J₂ = 52.2 Hz), 4.43 (dt, 2H, J₁ = 5.9 Hz and J₂ = 47.3 Hz), 4.24–4.20 (m, 1H), 3.90–3.70 (m, 1H+2H), 2.32–2.25 (m, 2H), 1.78–1.45 (m, 4H). ¹³C NMR (50 MHz, CD₃CN with D₂O) δ 164.2, 151.2, 138.0 (d, J = 3.0 Hz), 114.2, 96.8 (d, J = 190.4 Hz), 85.0 (d, J = 160.7 Hz), 84.5 (d, J = 4.6 Hz), 84.0 (d, J = 16.6 Hz), 74.4 (d, J = 24.5 Hz), 61.5, 30.6 (d, J = 19.4 Hz), 27.1, 25.0 (d, J = 5.4 Hz). HRMS calcd for C₁₃H₁₉F₂N₂O₅ [(M+H)⁺] 321.1262, found 321.1253. HPLC B: t_R = 10.9 min.