当前位置:首页 >> 能源/化工 >>

膜分离方面的英文文献翻译,后面带英文


有机溶剂纳滤膜在迷迭香的抗氧化剂萃取物的浓缩中的应用
摘要 目前的研究揭示了纳滤膜应用的潜在效益,在植物精华提炼过程中。咖啡和迷迭香的酸在乙醇中和迷迭香精华在乙醇中的模型溶剂 的纳滤膜,能在实验室的错流系统中被实现。Duramem
TM

chemical engineering research and design 318–327

/>
89

(2011)

200 型的纳滤膜被挑选出来基于挑选试验。实验数据和模型预测论证了在

纯净迷迭香萃取物的浓缩中半透错流渗透过程的功效。观测到在滞留物中抗氧化剂的能力没有重大的损失在获得的提取物浓度的程 度和过程中,这能允许滞留物直接应用,作为防腐剂和功能性原料在食品,化妆品和药品中。挑选出来的膜的能力,这种能力用来 分离在提取物中的酚酸从更高的分子量氧化剂混合物,也被进行研究。在渗透中,干的固体含量被发现非常低,所以这允许它在提 取过程中直接重复利用,因此带来了新的经济效益。

1 介绍 在最近的 10 年中,人们对有生物活性和对人体健康有潜在好处的自然混合物的检测、萃取和分离的新技术的发 展越来越感兴趣。调味料和芳香性的植物已经成为在这种兴趣中特别的一种,因为它们富含维生素,黄酮类,萜 类,类胡萝卜素,植物雌激素,矿物质等,这导致了很强的抗氧化性,抗菌性和反炎症的活性。在这个方向中的 最大挑战是合适植物的选择和 自然抗氧化剂生产技术的发展, 这种氧化剂能作为防腐剂和功能性原料在食品, 化妆品和药品中应用。 这个挑战也可以满足日益增长的顾客对自然食品的需求和克服对合成产品的使用的严格的 限制。与这些植物隔离的物质的抗氧化剂和自由基清除的活性能和它们高在治疗中的潜力和对癌的,心血管的和 炎症疾病的抑制作用联系起来。 人们已经进行了相当多的研究,致力于发展可靠的分离,定量和定性分析方法对芳香性植物和调味剂的提取 物。 例如迷迭香, 鼠尾草, 百里香, 留兰香, 牛至等, 这些在传统中已经对不同疾病的治疗中被用做偏方来使用。 许多研究者已经指出, 主要的起作用的组分对这些植物的高的抗氧化剂活性是酚酸和双萜。 从上面提到的植物中, 迷迭香被选为代表性的例子在这个研究中。野生的迷迭香属地榆是典型的植物,在地中海地区的干燥温暖区域。 由于它在各种地基上对土地的低敏感性,它成为一种常见的家庭植物在世界的许多地方。几个世纪来,迷迭香已 经被用作偏方和调味品的添加剂。著名的生物活性迷迭香酸的第一次分离被 Scarpati and Oriente (1958)从这种植 物,引起了研究兴趣在它的合成和实际应用中。当今迷迭香已经表现出抗癌抗菌素,抗病毒,抗菌,反炎症和抗 氧化的性能。它很明显的对人体有益,已经吸引了许多公司来合成基于迷迭香提取物的产品,这些提取物作为功 能性原料在食品,化妆品和营养品领域中获得应用。 Cuvelier et al.研究了 24 迷迭香实验性装置和经济提取物的组成和抗氧化活性。27 种组成物被检测和分析出 来通过 HPLC 保留时间,UV 和大量的光谱。它们中的 22 中物质被定义了。作者研究了酚醛树脂合成物和提取物 的抗氧化活性的关系,而且它们合成物的分析表明起主要作用的是迷迭香酸,鼠尾草酚和鼠尾草酸,这些在所有 的测试组分有很高浓度的存在。一些有很高原子量的酚类化合物也被用来显示抗氧化的活性。最近,研究者们揭 示了鼠尾草酸的潜在能力对于抑制神经变性的疾病,例如中风老年痴呆和帕金森等,这些疾病通常是由于自由基 过量引起的。不像许多药品和自然的抗氧化剂,鼠尾草酸被发现能够穿过脑血管壁障并能被用于所谓的(PAT) 。 活性成分通常从植物材料中提取出来的,通过固液抽提法。在这个模型中,萃取的最佳溶剂的选择是要求复 原程度的最大化,这是对与得到的最重要的分类化合物来说的。由于溶剂的毒性和低活性,最初的提取物通常转 变成粉末或浓缩铀类,这些能被商品所利用。通常利用在大气压或真空中蒸发,这能导致抗氧化活性相当数量的 损失,由于鼠尾草酸,鼠尾草酚,迷迭香酸和他们的衍生物对热得敏感性。 在一些文献中有机溶剂纳滤技术被讨论,作为传统蒸发法的替代方法。尽管,纳滤应用到草本精华提取过程 中的潜在效益已经被很好的概述,但是很少的文献能够显示它们能够到这种程度,既他们能够在现实中被实现。 期望的草药溶剂系统的选择和合适膜的利用时克服这个问题的主要挑战。但是伴有协调分子量界限(MWCO)的 有机稳定纳滤膜的最近发展和商业利用已经扩展了(OSN)特点,对于在这一领域的分子分离。初步膜的测试系 统和模型混合的权威评价的使用表明,似乎是个合理的方法,但并没有在这一研究领域中被利用。已经发现存在 一个协同作用在天然提取物的活性成分之间,这导致了更高的活性与这些物质单独存在时相比。因此,当我们关 注这个工作中的迷迭香萃取物,纳滤膜的成功应用应该有能力使最初迷迭香萃取物的富集,这些萃取物包括迷迭 香酸,双萜和黄酮类。与此同时,必须能提供一些低分子量的芳香剂或可能对人身体有害的酸的彻底去除,这些 能被检测出来在一些低浓度迷迭香素产品中。这能被实现,通过选择合适的(MWCO)膜,使迷迭香酸和双萜完 全的抑制, 并且使酚酸尽可能低的被抑制。 这些膜能被广泛利用因为 CA 连同 RA 能被检测和量化在有机溶剂提取

chemical engineering research and design

89

(2011)

318–327 物,这些提取物是从一些其他的芳香性植物例如鼠尾草,百里香,留兰香,香草和薰衣草。对纳滤膜的附加条件

是对迷迭香试剂有好的兼容性,对检测系统高的流量和少的污染。所选择的膜能够被商业运用在合理的价格。这 个工作的首要目标是选择高效的膜和操作条件,对于基于用模型溶剂纳滤实验的迷迭香萃取物的浓缩来说。接着 的目的是估计扩大的必要因素,对于所选择的膜和操作条件,并且证明错流过程的实用性,在实验室条件下,用 迷迭香萃取物。最后的目的是评估纳滤产品的抗氧化活性,在实际应用中。 2 材料和方法 2.1 膜和化学品 购买到粗糙的干的迷迭香树叶。植物材料在萃取前放置在密封的口袋中。所有溶剂都是 HPLC 等级的并提供 下列药品:无水乙醇,甲醇,乙腈和冰醋酸,去离子水。迷迭香酸和咖啡酸和(DPPH)从**公司购买。DuramemTM 有机溶剂抗交联聚合物纳滤膜被**膜公司提供。生产商详细说明,膜的 MWCO 是 200,300,500Da。 2.2 迷迭香和咖啡酸在模型溶剂和迷迭香萃取物的分析 相当多数量的文献关于特别兴趣的自然萃取物包含化合物的加工和应用,已经用 HPLC 技术检测他们的量。 在这个工作中, 迷迭香和咖啡酸的含量被测试, 通过一个装备有二极管检测器的泵系统。 分析条件是 25 摄氏度, 0.7ml/min 的流速,200nm 波长和 5ul 注射量。流动相是溶剂 A 和溶剂 B 的混合物。为了样品的反复分析,一个 持续 95 分钟的方法被提出来, 这种方法伴有各样的溶剂梯度根据下面的条件: (1)0–90min, B: 0–100%; (2) 90–91min, B: 100–100%; (3) 91–94min,B: 100–0%; (4) 94–95min, B: 0–0%.当 RA 和 CA 的模型溶液被分析时, 一个短的方法在 35 分钟内被完成,这种方法应用用固定相组成分别为(1) 0–30min, B: 0–33.3%; (2) 30–31min, B: 33.3–33.3%; (3)31–34min, B: 33.3–0%; (4) 34–35min, B: 0–0%.迷迭香和咖啡酸的峰在样品的光谱中被标记出来,通过 HPLC 的保 留时间和标准进行对比。RE 产品的典型光谱图见图 2 。所有上面所提的 HPLC 方法被修正考虑到 RA 和 CA,这种 修正是通过分析一系列的标液, 这种标液用浓度的 7 个等级, 这些浓度包括了 0-1g/L 的浓度范围, 对所有酸来说。 2.3 抗氧化活性的检测 各种检测方法对自然产物的抗氧化效能的评估在文献中已经被提出。在这些中,DPPH 激发清除实验被认为 是最简单有效的, 尽管这些激发化合物是稳定的而且不能在瞬间产生。 它经常被用来对 RE 类抗氧化活性的检测。 氧化化合物能减少 DPPH*到 DHHP-H,导致它的吸收光谱下降。DPPH*显示了 2 个最大普值。从他们中 517nm 被 用来选择作为检测波长, 为了避免迷迭香模型的影响。 实验按照下面的方法进行, DPPH*的原始乙醇溶液被准备, 调节它的浓度使它在 517nm 达到大约 0.9AU 的吸收量。在 4 个玻璃瓶中,每一个含有 3ml 的 DPPH*溶液,不同 的被测抗氧化剂产品的体积被加入到玻璃瓶中,这样是为了得到在样品中 RE 或 BHT 的不同浓度。在其中一个玻 璃瓶中没有加入抗氧化剂,而是加入酒精到 3.1ml。混合后,所有的样品放置在黑暗中,而且它们的吸光度被检 测用 UNICAM HELIOS GammaUV/vis spectrophotometer.每种样品的自由基清除活性接着被估计用消失的百分比或 残留的 DPPH*的含量,定义为

这里 As 和 Ab 分别是分析和空白的吸收峰。 氧化剂的能力也可以被检测通过 EC50 值或功能指数。 被检测的抗 氧化剂产物样品被一式三份进行分析。变化系数,通过标准差和平均算法量的比值被收集,被发现是少于 4%。 2.4 迷迭香萃取物的制备 系统的研究关于迷迭香酸, 鼠尾草酸和鼠尾草酚从迷迭香树叶的萃取动力学, 并没有在文献中被发现。 最近, Baskan 等人研究了水,甲醇,乙腈和丙酮作为萃取溶剂的潜在作用,对于咖啡酸和迷迭香酸从鼠尾草样品中。 所有混合物的最高浓度被得到,在甲醇溶剂中。Hernandez-Hernandez 等人用氯仿和乙醇来萃取迷迭香酸,鼠尾 草酸和鼠尾草酚,从牛至和迷迭香中。乙醇萃取有了很高的迷迭香酸收率比氯仿溶剂,因为迷迭香酸在极性溶剂 中有很高的可溶性,然而鼠尾草酸和鼠尾草酚的收率并没有什么不同。Bonoli 等人得到了鼠尾草酚,鼠尾草酸和 迷迭香酸的最高回收量,从新鲜的迷迭香树叶中,当氯仿,异丙醇和甲醛用作萃取溶剂,同时,在氯仿中迷迭香 酸和鼠尾草酸并没有存在。很明显,醇类是合适的萃取溶剂对于从这些植物中提取的有抗氧化活性的天然的酚醛

chemical engineering research and design

89

(2011)

318–327 树脂化合物。考虑到食品和制药的应用,乙醇被选择作为萃取溶剂,因为它的低毒性和得到的萃取组分可以直接

使用。迷迭香萃取物通过把碎的迷迭香树叶用无水乙醇在一个密封的玻璃瓶中混合得到。固液比是 1:10 。萃取 系统在黑暗中 25 摄氏度下搅拌 9 小时。这个混合物接着成功透过滤纸和聚丙烯注射器过滤器,用来去除任何固 体颗粒,这些固体能导致膜污染在萃取过程中或破坏 HPLC 柱。滤液被冷藏储存直到被使用。分离的样品仍然被 放在室温下的黑暗中,用来做比较。 3.1 过滤装置 纳滤实验被进行在连续或半间歇错流模型下进行,使用 METcell 错流系统。实验的所有部分由水和有机溶剂 的耐腐蚀材料组成,而且设计工作压力为 69bar,温度为 50 摄氏度。有效地过滤区域能被增加或减少,通过改变 错流过滤单元的个数。最大和最小的反应体积是 800ml 和 100ml,滞留体积是 40ml 。这个系统可以提供更多可扩 展的数据,与死端过滤相比,尤其是对于高浓度和粘度的溶剂,这些溶剂可以形成浓差极化层,例如一些天然萃 取物。错流实验的结果能提供流量和抑制数据,这些数据与工业膜组件的模式一致。在进行每个实验前,原料溶 剂被通入原料管 1.系统被密封,套管 1 和气罐 16 通过关闭阀 15 和气体单元 14 来实现,这样能实现安全和方便 的压力控制。操作温度维持在 25±1 摄氏度,通过热电偶 7,温度控制器 6 和加热器 5 来实现。 3.2 膜条件 根据生产说明,在膜利用前要对膜进行预处理。在操作压力下用纯乙醇冲洗纳滤膜来去除膜上残留的试剂。 当收集到至少 50ml 透过液时,可视为冲洗完成。在操作压力下进行膜的预处理,由于原始膜结构的连续压实, 导致了随着时间通量下降,直到达到稳定的流量。因此,纯净溶剂的过滤在定容模型中持续循环进行,在管 10 中收集,通过泵返回到管 1,直到渗透液流量达到稳定。接着,使溶剂流出并且测试的溶剂流入 1 中。 3.3 过滤模型和实验测量值 膜的性能被评估通过 2 个主要因素:渗透流量和目标化合物的截留量。渗透量通过测量每次每单位面积的渗 透体积得到。瞬时平行样品从渗透和滞留物中会发生在过滤过程中,而且它们的浓度用来决定检测化合物的截留 性。 在膜的筛选试验中,一种模型溶剂,由咖啡酸和迷迭香酸的混合物溶解在无水乙醇中,浓度分别是 0.1g/L 和 0.5g/L , 被用于纳滤膜的恒定体积模型通过 10 和管 1 间的渗透循环。 真实萃取过程的膜的功效用这两种化合 物的截留量和被评估,因为他们是在 RE 化合物中发现的氧化化合物的 2 种主要典型组分。尽管 RA,鼠尾草酸和 鼠尾草酚以可观的浓度在酒精的迷迭香萃取液中存在, 在模型溶剂中被使用, RA 基于它的它的相对高的抗氧化性, 相近的原子量和相似的化学结构。0.5g/L 的浓度被选用来代表它在 REs 中大约的含量,在文献中。根据 Cuvelier 等人的研究, 与 RA 的浓度比在检测的 24 个萃取物中是从 0 到 0.7 。 CA 目前的实验, 这一比率是在 0.2 , 0.1g/L 与 CA在模型溶剂中相一致。对模型溶剂的研究能简化分析和集中筛选。所有的测试实验被用不同的基材和同样的 料量进行至少2次,目的是解释实验的技术错误和膜性能的可能的不一致。对于挑选的膜,四重的实验被进行在 20和40bar, 通过用同一个膜的四个不同的取样片在错流单元中。 每个单元的渗透的样品被独立的进行分析。 得到的各种系数是对 CA 的截流量是 1.0%,对 RA 的截留量是 0.29,对渗透流量是 2.2 在 20bar 下,这表明,测 量值的高的准确性。 在这个情况下,当间歇膜过滤过程被进行,目的是浓缩新生的迷迭香浓缩物,最初的 130mL 被通入到管 1 中。对系统进行加压后,渗透物被收集进入 10 中。1 中的体积继续保持恒定通过从 11 中加入新的萃取液,萃取 的速率等于流速。在过滤中,瞬间样品被从滞留物和渗透物中得到,并对其进行咖啡酸和迷迭香酸含量的分析。 这个过程结束当得到渗透的 2 倍的渗透体积。 实验技术的准确性和 RA 可能的退化性通过来基于 RA 或小部分的干 固体含量物料平衡的偏差用下面的公式来控制。 4. 用新生萃取物作为稀释剂的恒定体积间歇渗透过程模型 在这一部分中,我们提出了半透膜过滤过程的简单数学模型,在这个过滤过程中滞留物被循环回进料器,同 时透过液被分离收集。原料和透过液的流动速率可以相同也可以不同,这取决于固定体积或变体积模型。我们采 用固定体积模型,这种模型符合目前的实验工作。得到的等式是基于著名的系统中物料平衡的概念。 5 结果和讨论 表 1 的数据是得到的截留量和透过量的流量,在最初的模型溶剂筛选过程中,模型溶剂伴有所有可能的膜。 最合适的膜应该提供高的迷迭香酸的截留量和相对高的渗透流量。这会保证抗氧化化合物的很少的损失,并由高 透过流量的和高的产量。 从表中能发现 DM500 的截留量和 DM300 的基本相同, 但是在相同压力下有更低的流量。

chemical engineering research and design

89

(2011)

318–327 DM200 的 RA 截留量几乎是完全的,但是它也存在低得流量与 DM300 相比。对于 DM200 和 DM300 来说,流量

成比例的增加随着压力的增加从 20 到 40bar ,因此在 40bar 时,这两种膜的选择结束。基于模型的预测,一种 能推断在真实过程中 5 倍的浓度将会导致用 DM300 原始萃取组分大约 17%积累渗透物的浓度。 在第一种情况下, 加入的溶剂量会被作为萃取溶剂循环,这是非常不经济的。因此 DM200 被选择应用到实验室模型的 Res 的萃取 过程。它的从主要抗氧化混合物的组分中分离迷迭香酸能力,通过迷迭香酸和咖啡酸的截留量的不同而被检测。 根据恒定体积和模型方法的截留量,使用 DM200 和无水乙醇作为稀释剂在 20bar 下的过滤过程会导致 CA 和 RA 的原始量的大约 50%和 3%的减少, 在经过 11 次过滤后。 由于需要的稀释剂的体积来实现从产品中去除 50%的 CA, 这个过程似乎是经济和合理的,只有当相当多的 CA 存在在产品中。 表 2 概述了得到的新鲜萃取物的组成和它们的结构, 在它们经历过恒定体积用新鲜萃取剂作为稀释剂的过滤 过程。RE 固体残渣中迷迭香酸的浓度与文献中报道的迷迭香中迷迭香酸含量相近。根据等式 2,Ra 和干燥固体 的物料平衡的偏差被收集。 在干燥的萃取物中的相等 RA 的浓度, 透过量和很少的干燥固体含量在透过量中表明, RE 中的相当数量存在的所有组分被膜截留了。因此对于 RA 得到的结果能被用于所有的萃取物中的成分。 表 3 反映了主要过程特点对透滤体积数量的改变。 很明显, 截留系数保持稳定并与从筛选试验中得到的相等, 在实验误差之内。结果是对 DM200 的模型预测和实验结果有很好的一致性。在渗透过程中,在原始量的 15%的 负偏差之内渗透量流星缓慢的减少。尽管流量低大约 30%比 RA 的模型溶剂得到的量,它仍然保持很高与文献中 报道的量。比较低的量是允许的因为在真实萃取中干固体的量比模型溶剂中得到的量多,并且浓差极化和渗透压 的重要影响应该被考虑。 得到的 RE 部分的抗氧化活性的研究被用来估计它们的潜在抗氧化能力。图 5 中显示,RE 组分与 DPPH*反应 迅速,小于 20 分钟,而且接着反应达到了峰值。为了对比它们丁基衍生物 BHT,作为抗氧化组分的参照,反应 缓慢而且在 129 分钟后达到了稳定。为了消除时间的影响和对产品的抗氧化活性进行对比研究,样品被允许在 60 分钟后进行反应,进行在所有样品的进一步测量。 在图 6 中,新生 RE 的抗氧化活性与 BHT 相比较。实验数据表现的很合适,通过第二顺序等式有很高的回归 系数,分别是 0.995 和 0.999 。基于这些统计数据,EC50 的值是 0.019g/L 对于 BHT 和 0.025g/L 对于 RE。因此, 这两种抗氧化剂的活性指数分别是 52 和 40,这与先前报道的数据相近。这些值表明 Res 可以代替通常的合成的 抗氧化剂。表 7 表明,关于目前自由基湮灭与加在样品中的被测抗氧化活性物质的干燥固体的体积或含量相对的 实验数据很好的符合一个线性等式,有很高的回归系数(R)0.99) 。这就允许使用直线的斜率作为一个参数 来对抗氧化活性进行评估。表 8a 显示,表 7 中直线的斜率,各族组分的每单元的干重自由基湮灭的百分率。与 新鲜萃取物相比,来自这一浓度的干固体的活性几乎没有下降。 室温下黑暗中储存 10 天后,同样新鲜的 RE 样 品被发现有显著的下降,这证明萃取液对氧化物的敏感性。渗透物的干组分有几乎相近的抗氧化活性与浓缩液相 比。因此,我们可以认为通过膜的不同抗氧化组分的截留或透过没有选择性。透过液和浓缩液的比较低的抗氧化 活性可以被归咎于一些最敏感的组分的氧化性,由于在萃取过程中暴露在空气中的氧气中。 当考虑到得到的乙醇迷迭香精华的直接应用,来自固液湿法萃取的 OSN 截留量的潜在效益取决于最终产品的浓 度。表 8b 显示,通过纳滤膜的原始萃取体积的三倍减少量导致了抗氧化活性增强了大约 3 倍,每个体积单元。 后来渗透量的值几乎没有减少表明抗氧化剂化合物含量很少。这与 RA 的完全截留发现和渗透量中很少的干固体 组分相一致。这也能表明不进行进一步处理的渗透液能作为萃取试剂被重复利用。 6 结论 来自迷迭香萃取过程的 OSN 的应用的最近研究在下面显示 DM200 型阻有机溶剂的纳滤膜能被用来迷迭香萃取物的浓缩, 由于它的合理的渗透流量和对迷迭香酸和植物 中其他抗氧化组分的几乎全部截留的功效。 当DM200型膜被使用时, 半间歇的错流过滤过程被证明是有效的, 对于新鲜迷迭香精华的浓缩。 浓缩液, 透过液的自由基清除能力与新鲜萃取液的性能的对比,表明在过滤的过程中抗氧化活性的量没有大的损失。 模型溶剂的使用能合理化选择最合适的膜。 尽管在做这个工作中,只进行了乙醇迷迭香精华的过程,但是它提供了一个现存技术中 OSN 的实施的通用 框架,这个技术是通过固液萃取法使来自各种植物的迷迭香酸的分离。 得到的萃取液的浓度和滞留物的抗氧化活性能允许它直接应用,这些应用包括:在食品中作为防腐剂和在食 品和药物中作为生物活性添加剂,并能停止萃取蒸发的必要性。

chemical engineering research and design

89

(2011)

318–327 在萃取过程中,透过液能被直接冲进利用,因此这带来了额外的经济效益。

来自迷迭香酸的咖啡的部分分离能被实现,通过在 20bar 的压力下,使用伴有 200Da 的 MWCO 的 DM 膜, 进行乙醇溶剂的纳滤过程。以相同目的的膜的应用来萃取含有这两种酸的溶剂需要实验的验证。

chemical engineering research and design 318–327

89

(2011)

Application of organic solvent nano?ltration for concentration of antioxidant extracts of rosemary (Rosmarinus of?ciallis L.)
D. Peshev
a b c

a,?

, L.G. Peeva b , G. Peev a , I.I.R. Baptista b , A.T. Boam c
UK

Department of Chemical Engineering, University of Chemical Technology and Metallurgy, 1756 So?a, Bulgaria Department of Chemical Engineering and Chemical Technology, Imperial College London, London SW7 2AZ, Membrane Extraction Technology Ltd., London HA0 4PE, UK

a b s t r a c t
The present investigation revealed the potential bene?ts of nano?ltration application in herbal extracts processing. Nano?ltrations of modelling solution of caffeic and rosmarinic acid in ethanol and ethanolic rosemary extract were carried out in a laboratory cross-?ow system. DuramemTM 200 nano?ltration membrane was selected based on screening experiments. The experimental data and model predictions demonstrated the ef?cacy of a semi-batch cross-?ow dia?ltration process for concentration of fresh rosemary extracts. The observed absence of signi?cant loss of antioxidant capacity in the retentate during the process and the degree of extract concentration achieved may allow retentate direct application as preservative and functional ingredient in the foods, cosmetics, neutraceuticals and medicines. The capability of the selected membrane to separate monophenolic acids from higher molecular weight antioxidant compounds in the extracts was also discussed. The dry solids content in the permeate was found suf?ciently low as to permit its direct re-use in the extraction process thus bringing additional economical bene?ts. ? 2010 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved. Keywords: Organic solvent nano?ltration; Rosemary extract; Antioxidant activity

1. Introduction
In the last decade there has been an increased interest in the development of new techniques for detection, extrac- tion and separation of natural compounds with biological activity and potential bene?ts for human health (Naczak and Shahidi, 2006; Cuvelier et al., 1996). Spices and aromatic herbs have been a particular object of this interest due to their high content of vitamins, ?avonoids, terpenoids, carotenoids, phytoestrogens, minerals, etc., resulting in a strong antioxi- dant, antimicrobial and anti-in?ammatory activity. The main challenge in this direction is the selection of suitable herbs and development of technologies for production of natural antioxidants that can be used as preservatives and func-

tional ingredients in the foods, cosmetics, neutraceuticals and medicines. This challenge also answers increasing con- sumers’ demand for ―natural‖ foods (Hernandez-Hernandez et al., 2009; Riznar et al., 2006) and overcomes the tighter restrictions for use of synthetic products. The antioxidant and radical scavenging activity (Suhaj, 2006; Moreno et al., 2006) of the substances isolated from these plants can be linked to their high potential in treatment and prevention of cancerous, cardiovascular and in?ammatory diseases (Kingherbs, 2009; Morrison, 2009). A considerable amount of research work has been devoted to developing reliable methods for separation, identi?cation and quantitative analysis of extracts of aromatic herbs and spices such as rosemary (Rosmarinus of?ciallis L.) (Cuvelier et

Abbreviations: AU, absorption unit; BHT, 3,5-di-tert-4-butylhydroxytoluene; CA, caffeic acid; DM, DuramemTM ; DPPH?, 2,2-diphenyl1-pikryl-hidrazyl; DS, dry solids; MW, molecular weight; MWCO, molecular weight cut-off; OSN (SRNF), organic solvent nano?ltration (solvent resistant nano?ltration); RA, rosmarinic acid; RE, rosemary extract; SR, solid residue. ? Corresponding author. Tel.: +359 2 8163293; fax: +359 2 8685488. E-mail addresses: d.peshev@uctm.edu (D. Peshev), l.peeva@imperial.ac.uk (L.G. Peeva), georgiapeev@gmail.com (G. Peev), i.baptista@imperial.ac.uk (I.I.R. Baptista), atb@membrane-extraction-technology.com (A.T. Boam). Received 19 December 2009; Received in revised form 7 June 2010; Accepted 2 July 2010 0263-8762/$ – see front matter ? 2010 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.cherd.2010.07.002

chemical engineering research and design 318–327

89

(2011)

319

Nomenclature A C J M N R t V active membrane area (m2 ) concentration (kg m?3 ) permeate ?ux (m3 m?2 s?1 ) mass of dry solids (kg) number of permeated dia?ltration volumes; N = JAt/V rejection coef?cient time (s) dia?ltration volume (volume of feed solution) (m3 )

Subscripts b mixed mean in permeate E in the extract i for i-th component p in permeate r in retentate

al., 1996; Razborcek et al., 2007; Wang et al., 2004; Bonoli et al., 2003), sage (Salvia of?cialis L.) (Cuvelier et al., 1996; Baskan et al., 2007; Wang et al., 2004), thyme (Thymus vulgaris L.) (Kruma et al., 2008; Wang et al., 2004), spearmint (Mentha spicata L.) (Wang et al., 2004), oregano (Origanum vulgare L.) (Kruma et al., 2008), etc., which have been traditionally used as folk remedies for treatment of different diseases. Many researchers have pointed out that the main compounds responsible for the high antioxidant activity of these plants are phenolic acids (rosmarinic acid, carnosic acid, etc.) and diterpenes (carnosol, rosmanol, etc.) (Cuvelier et al., 1996; Hernandez-Hernandez et al., 2009; Moreno et al., 2006; Kingherbs, 2009; Morrison, 2009; Ly et al., 2006; Romano et al., 2009) (Fig. 1). From the above men- tioned plants, rosemary has been chosen as a representative example in this study. The wild form of Rosmarinus of?cinalis L. is typical for the dry warm regions of the Mediterranean area. Due to its low susceptibility to cultivation on every kind of substrate it has become a common household plant in many parts of the world. For many centuries rosemary has been used in folk medicine and for ?avoring food additive. The fact that the well known biologically active rosmarinic acid was iso- lated for the ?rst time by Scarpati and Oriente (1958) from this plant has given rise to the research interest in its composition and practical application. Nowadays rosemary is known to exhibit antitumor, antiviral, antibacterial, anti-in?ammatory and antioxidant activity. Its unambiguous health bene?ts has attracted many companies to commercialize products based on rosemary extracts ?nding application as functional ingredients in the food, cosmetics and nutraceuticals, e.g. AquaROXR ,

VivOXR , INOLENSR (Vitiva); EssenRoseR (Fenchem Biotek Ltd.); Rosemary-EcoR (Centerchem Inc.), etc. Cuvelier et al. (1996) investigated the composition and antioxidant activity of 24 rosemary pilot-plant and commer- cial extracts. Twenty-seven compounds were detected and characterized by HPLC retention time, UV and mass spectrum. Twenty-two of them were identi?ed. The authors investigated the relationship between phenolic composition and antiox- idant activity of the extracts and their component analysis indicated that the main contribution was from rosmarinic acid, carnosol and carnosic acids, which were present in high concentrations in all investigated commercial extracts. Some of the phenolic compounds with higher molecular weight (?avonoids) were also shown to exhibit antioxidant activ- ity. Recently researchers revealed the potential of carnosic acid for prevention of neurodegenerative diseases such as stroke, Alzheimer’s and Parkinson’s diseases, usually caused by an excess of free radicals. Unlike many drugs and nat- ural antioxidants, carnosic acid was found to be able to cross the blood–brain barrier and also to be used in the so called pathologically-activated-therapy (PAT) (Satoh et al., 2008; Lipton, 2007). The active constituents are commonly isolated from the plant material by solid–liquid extraction (Naczak and Shahidi, 2006; Suhaj, 2006). Selection of the most appropriate solvent for extraction is required to maximize the degree of recovery for the most important phenolic compounds contained in the matrix. Due to solvent toxicity and/or low activity, the primary extracts are typically converted into powder or concentrated oils to be available as commercial products (Cuvelier et al., 1996; Suhaj, 2006). Usually evaporation at normal pressure or in vacuum is applied, which can cause considerable loss of antioxidant activity due to thermal susceptibility of carnosic acid, carnosol, rosmarinic acid and theirs derivatives (Baskan et al., 2007; Bonoli et al., 2003). In several publications (Geens et al., 2007; Tzibranska et al., 2009; Vandezande et al., 2008) organic solvent nano?ltration (OSN or SRNF) technology has been discussed as an alternative of the traditional evaporation. Although the potential bene- ?ts of nano?ltration application to herbal extracts processing have been well outlined very few studies have shown the extent to which they can be practically achieved (Vincze and Vatai, 2004; Tylkowski et al., 2010). The selection of prospective herb/solvent systems and the availability of suitable mem- branes have been the main challenges at overcoming this lack. However, the recent development of organic stable nano?ltra- tion membranes with tuned molecular weight cut-off (MWCO) and commercial availability has broaden OSN opportunities for molecular separation in this ?eld (See-Toh et al., 2008). The use of expert card for evaluation of the system and mod- elling mixture for a preliminary membranes testing suggested

chemical engineering research and design 318–327

89

(2011)

319

Fig. 1 – Chemical structure of the main antioxidant compounds identi?ed in rosemary extracts.

320

chemical engineering research and design

89

(2011)

318–327

in (Tzibranska et al., 2009), seems a rational approach but has not yet been utilized in this research area. It has been sug- gested that there is a synergistic effect between the active constituents in natural extracts resulting in higher activity as compared to the activity of the pure compounds themselves (Cuvelier et al., 1996; Hernandez-Hernandez et al., 2009). Therefore, as we focus on rosemary extracts in this work, successful application of nano?ltration technology should be capable to enrich the initial rosemary extracts in rosmarinic acid, diterpenes (carnosol, carnosic acid, rosmanol, etc.) and ?avonoids, all with MW > 300 Da. At the same time it should provide suf?cient removal of some lower molecular weight (MW < 200 Da) aromatic (such as vanillic) or probably harmful to humans health (such as caffeic) monophenolic acids (Fig. 1), which have been detected at low concentrations in some commercial rosemary extracts (Cuvelier et al., 1996). This can be achieved by selecting membranes with suitable MWCO pro- viding complete rejection of rosmarinic acid and diterpenes and as low as possible rejection of monophenolic acids. These membranes could have wider application because CA has also been detected and quanti?ed together with RA in organic solvent extracts (including alcoholic) from some other aro- matic herbs such as sage, thyme, spermint, balm, self-heal and lavender (Wang et al., 2004; Razborcek et al., 2007; Kruma et al., 2008). Additional requirements for the nano?ltration membrane are good compatibility with the extracting agent, high ?ux and negligible fouling with the investigated systems. The selected membranes should be commercially available at reasonable price. The ?rst objective of this work was to select ef?cient membranes and operating conditions for concentration of ethanolic rosemary extracts based on nano?ltration experi- ments using modelling solution. The following objective was to estimate, for the selected membrane and operating condi- tions, the necessary parameters for scale-up and to prove the process performance in cross-?ow mode, at laboratory scale, using rosemary extracts. The ?nal aim was to evaluate the antioxidant activity of the nano?ltration products regarding their use in the practice.

obtained from a curve of rejection versus molecular weight of styrene oligomers dissolved in acetone) of the membranes was 200, 300 and 500 Da (MET, 2009).

2. methods
2.1. chemicals

Materials and

Materials and

Coarsely chopped dried rosemary leaves were obtained from the trade (Schwartz, Thame road, Haddenham, Bucks, HP17 8LB). The plant material was purchased in hermetically sealed packaging and opened just before extraction. All the sol- vents used were HPLC grade and supplied as follows: absolute ethanol (Fisher Scienti?c), methanol, acetonitrile and glacial acetic acid (Sigma–Aldrich), deionized water (Purite Ana- lyst water puri?cation system). Rosmarinic acid (>97%, HPLC grade), caffeic acid (>98%, HPLC grade) and 2,2-diphenyl-1pikryl-hidrazyl (DPPH? ) were purchased from Sigma–Aldrich UK, and BHT (>99%) from Fluka. DuramemTM organic solvent resistant crosslinked polyimide nano?ltration membranes (See-Toh et al., 2007) used during the runs were provided by Membrane Extraction Technology Ltd. (UK). As speci- ?ed by the manufacturer, the MWCO (de?ned as MW at which 90% rejection is

2.2. Analysis of rosmarinic and caffeic acid in the modelling solutions and rosemary extracts
Considerable number of works related to processing or application of natural extracts containing compounds of particular interest (rosmarinic acid, carnosic acid, carnosol, ?avones, etc.) have used HPLC technique for their quanti?cation (Cuvelier et al., 1996; Hernandez-Hernandez et al., 2009; Moreno et al., 2006; Kruma et al., 2008; Wang et al., 2004; Canelas and Costa, 2007; Ly et al., 2006). In this work the content of rosmarinic and caffeic acid was measured by a quaternary pump Agilent 1100/1200 HPLC system equipped with UV/Vis diode detector. The separation was carried out in a 15 mm × 4.6 mm Hypersil ODS column (SUPELCO), 5 m particle size. The analysis conditions were 25 ? C, 0.7 mL/min ?ow rate, 330 nm wavelength detection and 5 L injection volume. The mobile phase was a mixture of solvent A (75 mL acetonitrile + 420 mL deionised water + 4.25 mL acetic acid) and solvent B (methanol). For analysis of RE samples a method with 95 min duration was applied with variable solvents gradient according to the following conditions: (1) 0–90 min, B: 0–100%; (2) 90–91 min, B: 100–100%; (3) 91–94 min, B: 100–0%; (4) 94–95 min, B: 0–0%. When modelling solution of RA and CA was analyzed a shorter method completed in 35 min was applied with mobile phase composition as follows: (1) 0–30 min, B: 0–33.3%; (2) 30–31 min, B: 33.3–33.3%; (3) 31–34 min, B: 33.3–0%; (4) 34–35 min, B: 0–0%. The peaks of rosmarinic and caffeic acid in the spectrum of the RE samples were identi?ed by comparison of the HPLC retention times with those of their standards. Typical chromatograms for RE products are shown in Fig. 2 (extract, concentrate, perme- ate). Both above mentioned HPLC methods were calibrated with respect to RA and CA by analyzing series of standard solutions, using seven levels of concentration which covered the concentration range 0–1 g/L for both acids. The coef?- cients of linear correlation for all calibration graphs were over 0.999.

antioxidant was added and it was used as blank assay. All the vials were adjusted to 3.1 mL with ethanol. After a mixing, all the samples were kept in dark place and their absorbance

2.3. determination

Antioxidant

capacity

A variety of tests for assessment of the antioxidant potency of natural products has been suggested in the literature (Suhaj, 2006). Amongst them the DPPH ? (2,2-diphenyl-1- picrylhydrazyl) radical scavenging assay can be considered as simple and reliable, since the radical compound is stable and does not have to be generated instantaneously. It has been commonly used for determination of the antioxidant activ- ity of REs (Moreno et al., 2006; Romano et al., 2009; Almela et al., 2006). The antioxidant compounds reduce the DPPH ? radi- cal to DPPH-H decreasing its spectrophotometric absorbance. DPPH ? shows two spectral maxima (at 328 and 517 nm) (Almela et al., 2006). From them 517 nm was selected for determination in the present work as to avoid interference with the rosemary matrix. The assay was developed as follows. Initial ethanolic solution of DPPH? was prepared adjusting its concentration (approximately 0.1 mM) as to reach absorbance of approximately 0.9 AU at 517 nm. In 4 glass vials, each one containing 3 mL of the DPPH? solution, different volumes (but less than 0.1 mL) of the investigated antioxidant product were added in order to obtain different concentration of RE or commercial antioxidant (BHT) in the sample. In one of them no

chemical engineering research and design

89

(2011)

318–327

321

%quenched DPPH? , were found to be less than 4% in all cases.

Fig. 2 – Typical chromatograms for rosemary extract nano?ltration products and fresh rosemary extract. was measured periodically using UNICAM HELIOS Gamma UV/vis spectrophotometer. The free radical scavenging activ- ity of each product was then evaluated as percent quenched or remnant DPPH? de?ned as: %quenched DPPH ? As ) ?= 100(Ab Ab DPPH

%remnant DPPH? = 100 ? %quenched ?

where As and Ab are the absorbances of the analyzed and blank sample, respectively. The antioxidant power was also charac- terized by either the EC50 value (the concentration necessary to reduce 50% of the initial DPPH? present in the sample) or the ef?cacy index (1/EC50 ). The samples of the investigated antioxidant products were analyzed in triplicate. The varia- tion coef?cients, calculated as the percentage ratio between the standard deviations and the mean arithmetic values for

the results from cross-?ow ?ltration experiments provided ?ux and rejection data more consistent with industrial scale

2.4. preparation

Rosemary

extracts

A systematic study on the extraction kinetics of rosmarinic acid, carnosic acid and carnosol from rosemary leaves was not found in the literature. Recently Baskan et al. (2007) investigated the potential of water, methanol, acetonitrile and acetone as extraction solvents for carnosic acid and ros- marinic acid from sage samples. The highest concentration of both compounds was established in methanol extracts. Hernandez-Hernandez et al. (2009) used both chloroform and ethanol to extract rosmarinic acid, carnosic acid and carnosol from oregano and rosemary. Ethanol extraction provided sig- ni?cantly higher rosmarinic acid yield than chloroform one due to the higher solubility of rosmarinic acid in polar sol- vents, while carnosol and carnosic acid yields showed no signi?cant difference. Bonoli et al. (2003) obtained the highest recoveries of carnosol, carnosic and rosmarinic acid from fresh rosemary leaves when chloroform, isopropanol and methanol were used as extraction solvents, respectively, while no ros- marinic and carnosic acids were present in the chloroform extracts. Evidently alcohols are suitable solvents for extraction of natural phenolic compounds with antioxidant activity from these plants. As we consider food and pharmaceutical appli- cations, ethanol is selected as extraction solvent in this study due to its low toxicity and possibility for direct application of the obtained extract fractions. Rosemary extracts were obtained by mixing dry chopped rosemary leaves with absolute ethanol in a tightly closed glass vessel. The solid to liquid ratio was 1:10 (w/w). The extraction system was stirred for 9 h at 25 ? C in a dark place. The mixture was then successively ?ltered through Wathman 1 (Mediumfast) ?lter paper and 0.45 m pore size polypropylene syringe ?lter (WathmanTM ) as to remove any ?ne solid particles which could initiate membrane fouling during the ?ltration or damage the HPLC column. The ?ltrate was stored in a freezer until utilization. Separate samples were kept in a dark place at room temperature for comparative purposes.

3. Description of the nano?ltration experimental set-up and modes of operation
3.1. equipment Filtration

Nano?ltration experiments were carried out in continuous or semi-batch cross-?ow modes using METcell cross-?ow system (Membrane Extraction Technology Ltd., UK) (MET, 2009) (Fig. 3). All parts of the equipment were made of aqueous and organic solvent resistant materials and designed to oper- ate at up to 69 bar and 50 ? C. The effective ?ltration area can be easily increased or decreased by changing the number of cross-?ow ?ltration cells, /8/(14 cm2 /cell), connected in series up to six. The maximum and minimum working volumes were 800 mL and 100 mL, respectively and the hold-up volume (for assembly of three cells) was 40 mL. This system can provide more scalable data as compared to the dead-end ?ltration units, particularly with concentrated or viscous solutions that generate signi?cant concentration polarization layers, such as some natural extracts. Due to the intensive mixing ensured through the speci?c eddy pattern ?ow con?guration of the retentate stream in the ?ltration cells and the high retentate ?ow rate of 1.2 L/min generated by the gear pump/3/,

322

chemical engineering research and design

89

(2011)

318–327

Fig. 3 – Schematic diagram of the experimental set-up: 1—feed vessel; 2—retentate sampling port/draining port; 3—gear pump for liquid recirculation; 4—retentate line; 5—heater; 6—temperature controller; 7—thermocouple; 8—2.5 in. MET cross-?ow cell; 9—permeate sampling port; 10—permeate collection vessel; 11—feed tank; 12—triple valve; 13—HPLC pump; 14—MET Gas Unit; 15—reducing valve; 16—nitrogen cylinder. membrane elements. Before every experiment the feed solution was supplied into the feed vessel/1/. The system was pressurized by connecting/1/to gas cylinder/16/through reducing valve/15/and MET Gas Unit/14/, which allowed safe and easy pressure control. Operating temperature was maintained constant (25 ± 1 ? C) by the set of termocouple/7/, temperature controller/6/and heater/5/. get compounds. Permeate ?ux was obtained by measuring the volume permeating per unit ?ltration area per unit time. Instantaneous parallel samples from permeate and retentate were taken during the ?ltrations and their concentrations were used to determine the rejections of the investigated com- pounds: Cp,i Cr,i

3.2.

Membrane conditioning

Ri = 1 ?

(1)

According to the manufacturer instructions (MET, 2009) a preliminary treatment of the nano?ltration membrane is required before utilization. Removal of the preserving agent was conducted by performing nano?ltration with pure ethanol at the aimed operating pressure. The washing was considered suf?cient when at least 50 mL of permeate was collected through the permeate collection ports/9/for every cell. The preliminary conditioning of the membrane at the operating pressure is usually attributed to initial continuous compaction of the membrane structure, resulting in steep decrease of the ?ux with the time until a steady permeate ?ux is achieved (Gibbins et al., 2002). Therefore, the ?ltration of pure solvent was continued in constant volume mode through recirculation of the permeate, collected into the vessel/10/, by the HPLC pump/13/back to the feed vessel/1/until the permeate ?ux became stable. Then the system was drained and the test solution was loaded into/1/.

3.3. measurements

Filtration

modes and

experimental

Membrane performances were evaluated with respect to two main parameters: permeate ?ux and rejection of certain tar-

During the membrane screening tests a modelling solution, mixture of caffeic acid (180.16 Da) and rosmarinic acid (360.31 Da) dissolved in absolute ethanol with concentrations of 0.1 g/L and 0.5 g/L, respectively, was subjected to nano?ltra- tion at constant volume mode by recirculation of the permeate collected in/10/to the feed vessel/1/. The ef?cacy of the mem- branes for processing of real extracts was evaluated with respect to the rejection of these two compounds as far as they are representative for the two main groups of antioxi- dant compounds found in the REs (see Fig. 1). Although RA, carnosic acid and carnosol are usually present in comparable concentrations in alcoholic rosemary extracts (Bonoli et al., 2003; Almela et al., 2006; Moreno et al., 2006), RA was used in the modelling solution due to its relatively higher resistance to oxidation, nearly the same molecular weight and similar chemical structure. The concentration of 0.5 g/L was chosen as to represent approximately its content in REs reported in the literature. According to Cuvelier et al. (1996), the CA to RA concentration ratio in the investigated 24 extracts ranged from 0 to 0.7. In the present work this ratio was ?xed at 0.2, corresponding to 0.1 g/L CA in the modelling solution. Working with modelling solutions simpli?ed the analysis and fastened

chemical engineering research and design

89

(2011)

318–327

323

Table 1 – Performance of DuramemTM series with solution of caffeic and rosmarinic acid. Operating pressure, bar Membrane Rejection, % Caffeic acid
20 DuramemTM DuramemTM DuramemTM DuramemTM DuramemTM DuramemTM 200 300 500 200 300 500 93.9 87.8 82.0 96.1 90.7 91.5

Permeate ?ux, L m?2 h?1

Rosmarinic acid
99.7 94.0 94.5 99.5 94.7 97.4 15.4 24.4 21.4 25.7 41.1 31.3

40

the screening (Tzibranska et al., 2009). All screening tests were performed at least twice for every membrane using different sheets from one and the same batch aiming to account for the experimental technique error as well as for a possible nonuniformity of the membrane properties. Average values for the obtained parameters were considered. With the selected membrane (DuramemTM 200) a fourfold test was carried out at 20 and 40 bar by using four different coupons of the same membrane in the cells of the cross-?ow rig. The permeate samples taken from every cell were analyzed independently. The variation coef?cients obtained were 1.0 (0.32)% for the rejection of CA, 0.29 (0.35)% for the rejection of RA and 2.2 (2.9)% for the permeate ?ux at 20 (40) bar, respectively, demon- strating the high accuracy of the measurements. In the case when a semi-batch membrane ?ltration process was carried out aiming to concentrate fresh rosemary extract, initial volume of 130 mL was charged into the feed vessel/1/. After pressurizing the system permeate was collected separately into/10/. The volume in/1/was kept constant by feeding fresh rosemary extract from/11/with a ?ow rate equal to this of the permeate stream. During the ?ltration instantaneous samples were taken from the retentate and permeate and analyzed for caffeic and rosmarinic acid content. The process was completed when two dia?ltration volumes (2× 130 mL) of permeate were collected. The accuracy of the experimental techniques and possible degradation of RA were controlled by the deviation from the mass balance with respect to RA or dry solids content in the fractions using the following expressions:

According to the above mentioned assumption, the system of equations governing the process is:

V

dCr,i = JA(CE,i ? Cp,i ) dt

(4)

Cp,i = (1 ? Ri )Cr,i

(5)

By substituting Eq. (5) in (4), the latter was integrated analytically at initial condition t=0 Cr,i = CE,i the

and Eq. (6) was obtained for the evolution of retentate concentration of the component i: 1 ? Ri eN(Ri ?1) Cr,i = CE,i .

1 ? Ri

(6)

The so called cumulative permeate concentration is also of practical importance: Cb,i = (N + 1)CE,i ? Cr,i N (7)

5. discussion

Results and

(N + 1)CF,RA ? (NCb,RA + Cr,RA ) ErrRA% = 100 (2)

(N + 1)CF,RA

Table 1 contains data for rejection and permeate ?ux obtained during the preliminary screening ?ltrations of modelling solution with all available membranes from the DuramemTM series. The most suitable membrane should provide high rejection of rosmarinic acid and relatively high permeate ?ux. This will guarantee minor loss of antioxidant compounds (rosmarinic acid as reference) with the permeate stream and high

ErrDS% =

(N + 1)MF,DS ? (NMb,DS + Mr,DS ) 100. (N + 1)MF,DS

(3)

4. Modelling of constant volume semi-batch dia?ltration process with fresh extract as diluent
In this section we provide a simple mathematical descrip- tion of a semi-batch membrane ?ltration process in which the retentate is recirculated to the feed tank and the permeate is collected separately (Fig. 3). The ?ow rate of the feed and permeate could be either equal or not, which corresponds to constant volume or variable volume operation mode, respec- tively. We assume constant volume operational mode, which complies with the presented experimental work. The equa- tions obtained are based on the well known conception of the mass balance in the system.

Fig. 4 – Model predictions and experimental data for the retentate and cumulative permeate concentrations of rosmarinic acid during a single stage constant volume dia?ltration with fresh extract as diluent.

324

chemical engineering research and design

89

(2011)

318–327

Table 2 – Composition of rosemary extract fractions after three times concentration (N = 2).a . Fraction
Permeate Fresh extract Concentrate
a

Concentration of rosmarinic acid, kg/m3
0.00195 0.118 0.346

Concentration of dry solids, kg/m3
0.192 11.0 32.3

RA to dry solids ratio, kg/kg
0.0101 0.0107 0.0107

Caffeic acid was not detected in any one of the fractions.

throughput, respectively. As can be seen from the table DM 500 showed nearly the same rejection as DM 300 but lower ?ux at both pressures. The RA rejection of DM 200 is almost complete (>99%), but this is at the expense of lower ?ux (approximately 37%) as compared to DM 300. For DM 200 and DM 300 the ?ux increased almost proportionally with increas- ing the pressure from 20 to 40 bar, therefore the choice was done between these two membranes operating at 40 bar. Based on the model predictions (Eqs. (6) and (7) and Fig. 4) one can conclude that in a real process ?ve times concentration (N = 4, corresponding to 0.589 g/L RA in the retentate) would result in cumulative permeate concentration approximately 17% of the initial extract content when using DM 300. With DM 200 it would be only 1.5%. In the ?rst case an additional solvent recovery unit will be required prior to recycle it as extraction solvent, which is not economically attractive. Therefore DM 200 was selected to perform a laboratory scale dia?ltration process of REs. Its capability to separate the monophenolic acids from the group of the major antioxidant compounds can be assessed by the difference in the rejection of rosmarinic and caffeic acids. According to a constant volume and rejection modelling approach a dia?ltration process at 20 bar employing DM 200 and pure ethanol as diluent would result in approxi- mately 50% and 3% reduction of the initial amount of CA and RA, respectively after the permeation of eleven dia?ltration volumes (N = 11). Due to the great volume of diluent required to achieve 50% removal of CA from the product the process seems to be economically unattractive and reasonable only when considerable amounts (with respect to its/their toxicity; Hirose et al., 1997) of CA/monophenolic acids are present in the product. Table 2 summarizes the composition of the obtained fresh extract and its fractions after it was subjected to constant volume dia?ltration with fresh extract as diluent. The concentration of rosmarinic acid in the solid residue of the RE (approximately 1 g RA/100 g SR) is close to those reported in the literature for ethanolic/methanolic extracts of rosemary (Hernandez-Hernandez et al., 2009; Moreno et al., 2006; Almela et al., 2006). The deviation from the mass balance with respect to RA and dry solids content calculated according to Eqs. (2) and (3), respectively was 1.2% and 1.0%. The equal concen- trations of RA in the dried extract and permeate as well as the negligible dry solids content in the permeate indicated that all the compounds present in considerable amounts in

Fig. 5 – Kinetics of the characteristic radical scavenging reaction of rosemary extract, its fractions and BHT with DPPH? . the RE were rejected by the membrane. Therefore the results obtained for RA can be readily transferred to all such ingredients in the extracts. Table 3 represents the change in the main process characteristics versus the number of dia?ltration volumes. Evidently the rejection coef?cient remains constant and equal to that obtained from the screening tests within the experimental error. As a result there is a good agreement between the model prediction for DM 200 and the experimental data shown in Fig. 4. The permeate ?ux decreases slightly during the ?ltra- tion remaining within 15% negative deviation from the initial value. Although the ?ux is approximately 30% lower than the one obtained with the modelling solution of RA, it still remains suf?ciently high when compared to the values reported in

Table 3 – DuramemTM 200 performance during rosemary extract concentration at 40 bar. Number of dia?ltration volumes
0 1 1.5 2

Flux,

L m?2 h?1
17.9 16.5 NA 15.3

Rejection, %
99.3 99.2 99.2 99.2

Fig. 6 – Experimental data and curve ?tting for DPPH? quenching for 60 min by fresh rosemary extract and

BHT versus concentration of antioxidant in the samples.

chemical engineering research and design

89

(2011)

318–327

325

Fig. 7 – Experimental data and line ?tting for percent quenched free radical for the volume (b) of the extract and its fractions added in the sample. the literature for similar applications (Vincze and Vatai, 2004; Geens et al., 2007; Tylkowski et al., 2010). The lower value is reasonable as far as the dry solids content in the real extract is more than the one in the modelling solution (0.6 kg m ? 3 ) and more signi?cant effect of concentration polarization and osmotic pressure (Vandezande et al., 2008; Baker, 2004) should be expected. Investigation on the antioxidant capacity of the obtained RE fractions was carried out to assess their potential applications. As illustrated in Fig. 5 RE fractions react rapidly with DPPH ? in less than 20 min and then the reaction approaches a plateau. In contrast with them the butylated derivative BHT, used as a reference antioxidant compound, reacts slowly and reaches a steady state after 120 min (not shown). As to eliminate the impact of the time and carry out an accurate comparative study on the antioxidant capacity of the prod- ucts, the samples were allowed to react for a standard time of 60 min in all further measurements. In Fig. 6 the antioxidant activity of the fresh RE is compared to that of BHT. The experimental data can be well ?tted by second order equations with high regression coef?cients of 0.995 and 0.999, respectively. Based on these correlations the EC50 values were found to be 0.019 g/L for BHT and 0.025 g/L for RE. Hence the ef?cacy indexes (1/EC50 ) of these two antioxidants are 52 and 40—close to previously reported data on BHT and REs, obtained following DPPH ? free radical scavenging activity assay (Moreno et al., 2006; Almela et al., 2006). These values demonstrate that REs could replace conventional synthetic antioxidants. As illustrated in Fig. 7, the experimental data on percent quenched free radical plotted versus the dry solids amount

60 min versus the dry solids amount (a) or

(Fig. 7a) or volume (Fig. 7b) of the investigated antioxidant product added in the sample can be ?tted by a linear equa- tion with high coef?cient of correlation (R > 0.99). This allowed to use the slopes of the lines as a parameter for assessment of the antioxidant activity. Fig. 8a shows the slopes of the lines from Fig. 7a, i.e. the percent quenched free radical per unit dry weight from the respective fraction. There is a negligible decrease in the activity of the dry solids from the concen- trate in comparison with the fresh extract. A more signi?cant decrease (approximately 20%) was observed for a sample from the same fresh RE after 10 days storage in the dark at room temperature evidencing the extract susceptibility to oxidation. The dry substance of the permeate has approximately the same speci?c antioxidant power as the concentrate. Therefore no selective permeation or rejection of different antioxidant compounds through the membrane can be considered. The slightly lower speci?c antioxidant power of the concentrate and permeate can be attributed to oxidation of some of the most susceptible constituents due to exposure to oxygen from the ambient air during the extract processing. When consid- ering direct application of the obtained ethanolic rosemary extracts, potential bene?ts from integration of OSN with the solid–liquid extraction can be expected for concentration of the end product. Fig. 8b illustrates that three times reduction of the initial extract volume through nano?ltration results in approximately three times enhancement of the antioxidant activity per unit volume. The negligible value of the latter for the permeate indicates traces of antioxidant compounds. This result is in agreement with the ?ndings for the almost complete rejection of RA and negligible dry solids content in the permeate (Table 2). It also suggests that the perme-

Fig. 8 – Antioxidant activity per mg dry solids (a) or

L of extract fraction (b): 1—fresh RE; 2—concentrate; 3—RE kept 10 days

at room temperature; 4—permeate.

326

chemical engineering research and design

89

(2011)

318–327

ate could be re-used as extracting agent without any further treatment.

6. Conclusions
The present investigation on application of OSN in processing of extracts from rosemary has shown that: ? Organic solvent resistant nano?ltration membrane DuramemTM 200 can be used for concentration of etanolic rosemary extracts due to its reasonable permeate ?ux and almost complete rejection of rosmarinic acid and the others antioxidant constituents of the herb; ? A semi-batch cross-?ow dia?ltration process has been proved ef?cient for concentration of fresh rosemary extracts when DuramemTM 200 was employed. The comparison of the free radical scavenging activity of concentrate, permeate and fresh extract demonstrated no signi?cant loss of antioxidant capacity in the retentate during the ?ltration; ? The use of a modelling solution can streamline the selection of the most appropriate membrane; ? Although in this work ethanolic rosemary extracts were processed only, it provides a general framework for imple- mentation of OSN in existing technologies for isolation of rosmarinic acid from a variety of plants by solid–liquid extraction; ? The degree of extract concentration achieved and the antioxidant activity of the retentate found may allow its direct application as preservative in the food technology or biologically active ingredient in neutraceuticals and medicines, and ceases the necessity for extract evaporation; ? The permeate could be re-used directly in the extraction process thus bringing additional economical bene?ts; ? A partial separation of caffeic from rosmarinic acid can be achieved by nano?ltration of their alchoholic

Baskan, S., Oztekin, N., Erim, F.B., 2007. Determination of carnosic acid and rosmarinic acid in sage by capillary electrophoresis. Food Chem. 101, 1748–1752. Bonoli, M., Pelillo, M., Lercker, G., 2003. Fast separation and determination of carnosic acid and rosmarinic acid in different rosemary (Rosmarinus of?cinalis) extracts by capillary zone electrophoresis with ultra violet-diode array detection. Chromatographia 57 (7/8), 505–512.

solution utilizing DuramemTM membrane with MWCO of 200 Da at a pressure of 20 bar. The membrane application with the same purpose to plant extracts containing these two acids needs experimental con?rmation.

Acknowledgement
The research leading to these results has received fund- ing from the European Community’s Seventh Framework Programme (FP/2007-2013) under grant agreement No. PIAP- GA-2008-218068.

References
Almela, L., Sanchez-Munoz, B., Fernandez-Lopez, J.A., Roca, M.J., Rabe, V., 2006. Liquid chromatograpic–mass spectrometric analysis of phenolics and free radical scavenging activity of rosemary extract from different raw material. J. Chromatogr. A 1120, 221–229. Baker, R.W., 2004. Membrane technology and applications. In: Concentration Polarization, second edition. John Wiley & Sons, Chichester, England, pp. 161–178.

Canelas, V., Costa, C.T., 2007. Quantitative HPLC analysis of rosmarinic acid in extracts of Melissa of?cinalis and spectrophotometric measurement of their antioxidant activities. J. Chem. Educ. 84 (9), 1502–1504. Cuvelier, M.E., Richard, H., Berset, C., 1996. Antioxidative activity and phenolic composition of pilot-plant and commercial extracts of Sage and Rosemary. J. Am. Oil Chem. Soc. 73 (5), 645–652. Geens, J., De Witte, B., Van Der Bruggen, B., 2007. Removal of API’s (active pharmaceutical ingredients) from organic solvents by nano?ltration. Sep. Sci. Technol. 42, 2435–2449. Gibbins, E., D’Antonio, M., Nair, D., White, L.S., Santos, L.M.F., Vankelekom, I.F.G., Livingston, A.G., 2002. Observations on solvent ?ux and solute rejection across solvent resistant nano?ltration membranes. Desalination 147, 307–313. Hernandez-Hernandez, E., Ponce-Alquicira, E., Jramillo-Flores, M.E., Guerrero-Legarreta, I., 2009. Antioxidant effect rosemary (Rosmarinus of?cialis L.) and oregano (Origanum vulgare L.) extracts on TBARS and color of model raw pork batters. Meat Sci. 81, 410–417. Hirose, M., Takesada, Y., Tanaka, H., Tamano, S., Kato, T., Shirai, T., 1997. Carcinogenicity of antioxidants BHA, caffeic acid, sesamol, 4-methoxyphenol and catechol at low doses, either alone or in combination, and modulation of their effects in a rat medium-term multi-organ carcinogenesis model. Carcinogenesis 19 (1), 207–212. Kingherbs, 2009. http://www.made-in-china.com/showroom/ kingherbs/product-detailOoeJcKYHlgVw/China-RosmarinicAcid-Rosemary-Extract.html (accessed November 29, 2009). Kruma, Z., Andjelkovic, M., Verhe, R., Kreicbergs, V., 2008. Phenolic compounds in basil, oregano and thyme. In: Foodbalt Proceedings, http://llufb.llu.lv/conference/foodbalt/ 2008/Foodbalt-Proceedings-2008-99-103.pdf (accessed November 29, 2009). Lipton, S.A., 2007. Pathologically activated therapeutics for neuroprotection. Nat. Rev. Neurosci. 8, 803–808. Ly, T.N., Shimoyamada, M., Yamauchi, R., 2006. Isolation and characterization of rosmarinic acid oligomers in Celastrus hindsii benth leaves and their antioxidative activity. J. Agric. Food Chem. 54 (11), 3786–3793. MET, 2009. http://www.membrane-extraction-technology.com/ (accessed November 29, 2009). Moreno, S., Scheyer, T., Romano, C.S., Vojnov, A.A., 2006. Antioxidant and antimicrobial activities of rosemary extracts linked to their polyphenol composition. Free Radical Res. 40 (2), 223–231. Morrison, M., 2009. The Bene?ts of the Use of Lemon Balm in Herbal Preparations, Chemical Constituents Lemon Balm., http://www.herballegacy.com/Morrison Chemical.html (accessed November 29, 2009). Naczak, M., Shahidi, F., 2006. Phenolics in cereals, fruits and vegetables: occurrence, extraction and analysis (review). J. Pharm. Biomed. Anal. 41, 1523–1542. Razborcek, M.I., Voncina, D.B., Dolecek, V., Voncina, E., 2007. Determination of major phenolic acids, phenolic diterpenes and triterpenes in rosemary (Rosmarinus of?cinalis L.) by gas chromatography and mass spectrometry. Acta Chim. Slov. 54, 60–67. Riznar, K., Celan, S., Knez, Z., Skerget, M., Bauman, D., Glaser, R., 2006. Antioxidant and antimicrobial activity of rosemary extract in chicken frankfurters. J. Food Sci. 71 (7), C425–C429. Romano, C.S., Abadi, K., Repetto, V., Vojnov, A.A., Moreno, S., 2009. Synergistic antioxidant and antibacterial activity of rosemary plus butylated derivatives. Food Chem. 115, 456–461. Satoh, T., Kosaka, K., Itoh, K., Kobayashi, A., Yamamoto, M., Shimojo, Y., Kitajima, Ch., Cui, J., Kamins, J., Okamoto, S., Izumi, M., Shirasawa, T., Lipton, S.A., 2008. Carnosic acid, a catechol-type electrophilic compound, protects neurons both in vitro and in vivo through activation of the Keap1/Nrf2 pathway via S-alkylation of targeted cysteines on Keap 1. J. Neurochem. 104, 1116–1131.

chemical engineering research and design

89

(2011)

318–327

327

Scarpati, M., Oriente, G., 1958. Isolamente e constituzione dell’acido rosmarinico (dal Ric. Sci. 28, 2329–2333. See-Toh, Y.H., Lim, F.W., Livingston, A.G., 2007. solvents. J. Membr. Sci. 301, 3–10. See-Toh, Y.H., Silva, M., Livingston, A.G., 2008.

Rosmarinus off.).

Polymeric membranes for nano?ltration in polar aprotic

Controlling molecular weight cut-off curves for highly solvent

stable organic solvent nano?ltration (OSN) membranes. J. Membr. Sci. 324, 220–232. Suhaj, M., 2006. Spice antioxidants isolation and their antiradical activity: a review (critical review). J. Food

Compos. Anal. 19, 531–537. Tylkowski, B., Trusheva, B., Bankova, V., Giambertini, M., Peev, G., Nikolova, A., 2010. Extraction of biologically active +


相关文章:
膜分离方面的英文文献翻译,后面带英文
膜分离方面的英文文献翻译,后面带英文_能源/化工_工程科技_专业资料。有机溶剂纳滤膜在迷迭香的抗氧化剂萃取物的浓缩中的应用摘要 目前的研究揭示了纳滤膜应用的...
外文文献翻译
外文文献翻译_文学_高等教育_教育专区。从甜叶菊中得到甜味剂的膜分离方案 摘要: 在现有的方法中,从甜叶菊中提取和精制甜菊糖苷涉及到很多操作步骤,其中就包括有...
中英文参考文献格式1
中英文参考文献格式 中文参考文献格式参考文献(即引文出处)的类型以单字母方式...论文集 〔1〕郭宏,王熊,刘宗林.膜分离技术在大豆分离蛋白生产中综合利用的研究...
外文翻译
关键词:脱氮除磷去除;反硝化聚磷 1 引言膜生物反应器是一种膜分离技术与生物...参考文献 [1] BAE T H,HAN S S,TAK T M.Membrane sequencing batch ...
外文翻译
在这方面,氯消耗的减少和处理后的水中形成的三卤...(论文)设计外文翻译 理之后处理过的水被加到空洞的...膜分离单元可以操作持续 20-24 小时根据混凝剂剂量...
毕业外文翻译
(论文)外文翻译 污水经过二级处理后,仍含有极细微的...RO 和 ED 利用膜分离水中溶解的盐类及矿物质。 IX...英文个人简历模板 创意简历模板汇集 推理型题分析与总结...
文献翻译 - 副本
文献翻译 - 副本_工学_高等教育_教育专区。水处理...在被污染后,Na+和有机物的分离效率没有变化。 ...我们之前的研究显示了沸石过滤在除 盐方面是一种...
中英文文献检索报告《BMED在废水处理中的应用》
7. 娄玉峰; 张伟华 In 均相离子荷电膜分离技术在湿法冶金行业中的应用, 第...三、 文献分析(摘要翻译) 1. Eisaman, M. D.; Alvarado, L.; Larner, ...
膜分离技术翻译文档maxueyi
2页 免费 技术传播视野下的翻译研... 4页 免费 翻译文档 12页 免费喜欢此文档的还喜欢 膜分离技术研究进展+文献... 13页 免费 膜分离技术在给水处理中......
毕业论文外文翻译视频的前期拍摄与后期制作(适用于毕业论文外文翻译中英文对照)
(适用于毕业论文外文翻译中英文对照)_工学_高等教育...当主要的拍摄工作完成后, 就到了后期制作阶段。传统...个方面: 1、没有夹帧 就是完整镜头连接,中间没有...
更多相关标签:
英文文献翻译网站 | 参考文献翻译成英文 | 中文文献翻译成英文 | 英文文献翻译 | 英文文献翻译软件 | 翻译英文文献算抄袭吗 | 翻译英文文献的软件 | 参考文献英文翻译 |