关于「溶出模型族:Noyes–Whitney、Weibull 与九个时间点能分辨什么」,《草本多糖溶出动力学》给出的定量结论是什么?

结论:Noyes–Whitney 方程描述的是固体从表面溶掉,与多糖从颗粒内部扩散出来不是一回事;它适用于提取物粉末的复溶:表面积缩小拖慢尾巴,剂量超过饱和容量时停在 0.667,近饱和时延长搅拌无用。Weibull 形状指数与尺度无关,携带几何信息,但也随颗粒分布与取样表漂移:同一个球,换一张取样表就从 0.641 变成 0.577。九个时间点足以排除一级动力学,却在大约四分之一的模拟里分不清扩散解与 Weibull。相似因子宽松,半径变大两成仍得 57.3。模型选好之后,下一个问题是把三根杠杆同时画出来,看工艺参数的整张地形图,并把「道地」写成一个可以检验的统计问题。这是第五章的内容。

数据口径:本书非医疗建议,不宣称任何功效,只讨论得率、浓度、溶出分数与扩散系数这些物理化学量;书中参数为示意,全部计算由书中程序生成;自我试验前应先咨询执业医师。 第一章显示,示意参数下 30 分钟延长到 60 分钟得率乘以 1.411,60 分钟延长到 120 分钟只乘以 1.169;在 90 度、半径 1.5 毫米、30 分钟的基线上,时间加倍使溶出分数增加 0.154,升温十度只增加 0.056,半径减半增加 0.317。

English version

对照数据表

下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 0.577, 0.641

curvebetatau(min)worst |dF|points
sphere a = 1.0 mm0.56117.60.0396507
sphere a = 2.0 mm0.56170.20.0396507
sheet l = 1.0 mm0.618103.20.0584458
first-order k = 0.02961.00033.80.0000294
sieve mixture (p22)0.47220.90.0169624
beta is scale-free: sphere 1 mm and 2 mm share it; tau scales a^2
lab schedule 5 to 120 min, sphere a = 1.0 mm: beta = 0.641 (7 points)
lab schedule 5 to 120 min, sphere a = 2.0 mm: beta = 0.577 (9 points)
查看完整验算程序
# Chapter 4. p42: what Weibull's shape exponent says about mechanism.
# Weibull release: F = 1 - exp(-(t/tau)^beta). Linearised:
#   ln(-ln(1 - F)) = beta ln t - beta ln tau.
# We fit it by least squares to exact curves (points with 0.05 <= F <=
# 0.95, 50 points per doubling of t): sphere and plane sheet (Crank),
# first-order, and a sieve mixture (p22). D = 0.003 mm^2/min; sizes are
# illustrative (示意). Last lines: a typical laboratory schedule.
import math

D = 0.003
SIEVE = [(0.25, 0.10), (0.5, 0.20), (1.0, 0.35), (2.0, 0.25),
         (3.0, 0.10)]


def sphere(tau, nmax=4000):
    if tau <= 0.0:
        return 0.0
    terms = []
    for n in range(1, nmax + 1):
        e = math.exp(-n * n * math.pi ** 2 * tau) / (n * n)
        terms.append(e)
        if e < 1e-17:
            break
    return 1.0 - 6.0 / math.pi ** 2 * math.fsum(terms)


def sheet(tau, nmax=4000):
    if tau <= 0.0:
        return 0.0
    terms = []
    for n in range(nmax):
        m = 2 * n + 1
        e = 8.0 / (m * m * math.pi ** 2) * math.exp(
            -m * m * math.pi ** 2 * tau / 4.0)
        terms.append(e)
        if e < 1e-17:
            break
    return 1.0 - math.fsum(terms)


def fit(curve):
    xs, ys = [], []
    for k in range(-500, 701):
        t = 2.0 ** (k / 50.0)
        f = curve(t)
        if 0.05 <= f <= 0.95:
            xs.append(math.log(t))
            ys.append(math.log(-math.log(1.0 - f)))
    n = len(xs)
    mx, my = math.fsum(xs) / n, math.fsum(ys) / n
    b = math.fsum((x - mx) * (y - my) for x, y in zip(xs, ys)) / \
        math.fsum((x - mx) ** 2 for x in xs)
    a = my - b * mx
    tau = math.exp(-a / b)
    worst = max(abs(math.exp(-math.exp(a + b * x)) - math.exp(-math.exp(y)))
                for x, y in zip(xs, ys))
    return b, tau, worst, n


curves = [
    ("sphere a = 1.0 mm", lambda t: sphere(D * t)),
    ("sphere a = 2.0 mm", lambda t: sphere(D * t / 4.0)),
    ("sheet l = 1.0 mm", lambda t: sheet(D * t)),
    ("first-order k = 0.0296", lambda t: 1 - math.exp(-0.0296 * t)),
    ("sieve mixture (p22)", lambda t: math.fsum(
        w * sphere(D * t / a ** 2) for a, w in SIEVE))]
print("curve                     beta    tau(min)  worst |dF|  points")
for name, c in curves:
    b, tau, w, n = fit(c)
    print("%-24s  %.3f   %7.1f    %.4f     %4d" % (name, b, tau, w, n))
print("beta is scale-free: sphere 1 mm and 2 mm share it; tau scales a^2")
LAB = (5, 10, 15, 20, 30, 45, 60, 90, 120)
for name, scale in (("sphere a = 1.0 mm", 1.0), ("sphere a = 2.0 mm", 4.0)):
    xs = [math.log(t) for t in LAB if 0.05 <= sphere(D * t / scale) <= 0.95]
    ys = [math.log(-math.log(1.0 - sphere(D * t / scale))) for t in LAB
          if 0.05 <= sphere(D * t / scale) <= 0.95]
    mx, my = math.fsum(xs) / len(xs), math.fsum(ys) / len(ys)
    b = math.fsum((x - mx) * (y - my) for x, y in zip(xs, ys)) / \
        math.fsum((x - mx) ** 2 for x in xs)
    print("lab schedule 5 to 120 min, %s: beta = %.3f (%d points)"
          % (name, b, len(xs)))

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《草本活性多糖菲克扩散溶出动力学、道地石斛提取曲线与复方等效线图协同母典》第 4 章「溶出模型族:Noyes–Whitney、Weibull 与九个时间点能分辨什么」(版本 1.0.0). 5D Library. DOI 待发布

DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-289 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026289,
  author    = {Fan, Yafu},
  title     = {Fickian Release Kinetics of Herbal Polysaccharides: Extraction Curves for Dendrobium and Isobologram Analysis of Combinations},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论