Abstract. CgHIsCIaN202P, monoclinic, P21/c, a = 13.722 (3), b = 13.929 (2), c = 8.0212 (13)./k, fl = 100.49 (1) °, Z = 4, D x = 1.426 Mg m -3, T = 268 K; R = 0.053 for 2413 observed reflections. The ring is in the chair conformation with the phosphoryI O axial and, the bis(chloroethylamine) group equatorial.Introduetlon. Cyclophosphamide (CP) and its congeners ifosfamide (IP) and trofosfamide (TP) are effective antitumor agents in clinical and experimental use. The first step in the metabolic activation of CP and its analogs is C(4)-hydroxylation by the mixed-function oxidases of liver microsomes (Arnold, Bourseaux & Brock, 1958;Brock & Hohorst, 1963;Cohen & Jao, 1970). The resulting 4-hydroxy derivative then decomposes to yield acrolein and phosphoramide mustard with the latter likely the ultimate alkylating metabolite (Colvin, Padgett & Fenselau, 1973, Connors, Cox, Farmer, Foster & Jarman, 1974Struck, Kirk, Witt & Laster, 1975). As part of our study of the molecular structures of CP analogs and derivatives we report here the structure of TP. A preliminary report of this work was presented at the Symposium on the Metabolism and Mechanism of Action of Cyclophosphamide (Camerman, Smith & Camerman, 1976).CP: R t = R 2 = CH2CH2CI, R 3 = H IP: R t = R 3 = CH2CH2CI, R 2 = H TP: R t = R 2 = R 3 = CH2CH2CITrofosfamide was recrystallized from an ether-water mixture. The clear crystals turned yellow and then reddish-brown on exposure to X-rays. In order to retard the apparent decomposition, the crystal was 0567-7408/81/040957-03501.00immersed in a 25 mm diameter stream of very dry, cool (268 K) air during data collection. Systematic absences hOl, l =/: 2n and 0/d), k =/= 2n confirm the space group P2~/c (C~k, No. 14). Unit-cell parameters were determined by least-squares refinement from the 20, X, q) angles of 18 reflections in the 20range from 31 to 48 °. Intensity data were collected on an automated four-circle diffractometer (Nb-filtered Mo radiation, 2=0.71069 A), and all independent hkl and hki reflections having 20 < 50 ° (corresponding to a minimum interplanar spacing of 0.84 A) were measured using a crystal of dimensions 0.87 x 0.73 x 0.75 mm. The 0-20 scan method was employed with stationary 10 s background measurements taken on both sides of each reflection scan. Reflection intensities, I, were computed as P -B(t./tb) and the standard deviation (I) = [P -B(tp/tb) 2 + (~.03I) 2] 1/2 where P and B are the total counts accumulated during the scan period, tp, and the total background counting time,_t0, respectively. Three standard reflections (1,11,0, 119, 1-0,1,[) measured every 200 reflections showed a steady decline in intensity to an average minimum intensity 87% of their initial value. A quadratic decomposition curve, fit by least squares to a plot of monitor reflection intensity versus reflection number, was used to calculate scale factors as a function of the serial order of data collection. 26 medium to strong intensities were recollected at a lower X-ray tube current to obtain a measure of r, the correct...